List of Contributors
Numbers in parentheses indicate the page(s) on which the authors’ contributions begin.
Steven A...
111 downloads
3765 Views
55MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
List of Contributors
Numbers in parentheses indicate the page(s) on which the authors’ contributions begin.
Steven A. Abrams (811)
USDA/ARS Children’s Nutrition Research Center, Houston, Texas 77030, USA John S. Adams (341, 1379)
Division of Endocrinology, Diabetes, and Metabolism, University of California, 8700 Beverly Blvd, Los Angeles, Cedars-Sinai Medical Center, Los Angeles, California 90048, USA Judith E. Adams (967)
Clinical Radiology, Imaging Science and Biomedical Engineering, Stopford Building, The University, Manchester M13 9PT, United Kingdom Luciano Adorini (631, 1511, 1833)
BioXell SpA, 20132 Milano, Italy Paul H. Anderson (711)
Institute of Medical and Veterinary Science, Adelaide, South Australia, Australia Gerald J. Atkins (711)
Hanson Institute, Adelaide, South Australia, Australia Jane E. Aubin (649)
Department of Molecular and Medical Genetics, Faculty of Medicine, University of Toronto, Toronto, Ontario M5S 1A8, Canada Isabelle Bailleul-Forestier (599)
INSERM E110, Université Paris VII, IFR58, Cordeliers Biomedical Institute, 75270 Paris Cedex 06, France Julia Barsony (363)
Laboratory of Cellular Biochemistry and Biology, NIDDK/NIH, Bethesda, Maryland 20892-0850, USA Thomas K. Barthel (219)
Department of Biochemistry and Molecular Biophysics, University of Arizona, Tucson, Arizona 85721, USA
Norman H. Bell (789)
Department of Medicine, Medical University of South Carolina, Charleston, South Carolina 29425, USA Ariane Berdal (599)
INSERM E110, Université Paris VII, IFR58, Cordeliers Biomedical Institute, 75270 Paris Cedex 06, France Joel J. Bergh (751)
Department of Biological Sciences, University of Delaware, Newark, Delaware 19716, USA Jacqueline L. Berry (1293)
University of Manchester, Vitamin D Research Group, Department of Medicine, Manchester Royal Infirmary, Manchester M13 9WL, UK Daniel D. Bikle (609)
Endocrine Research Unit, Veterans’ Affairs Medical Center, University of California-San Francisco, San Francisco, California 94121-1598, USA John P. Bilezikian (1355)
Department of Medicine, College of Physicians and Surgeons, Columbia University, New York, New York 10032, USA Ernst Binderup (1489)
Biological Research, Leo Pharma, DK-2750 Ballerup, Denmark Lise Binderup (1489)
Biological Research, Leo Pharma, DK-2750 Ballerup, Denmark Nicholas J. Bishop (803)
Academic Department of Child Health, University of Sheffield, Sheffield Children’s Hospital, Sheffield S10 2TH, United Kingdom Ilse Bogaerts (135)
Laboratory Analytische Chemie, Van Evenstraat 4; B-3000 Leuven, Belgium
xiv Ricardo L. Boland (883)
Department de Biología, Bioquímica & Farmacía, Universidad Nacional del Sur, San Juan 670, (8000) Bahía Blanca, Argentina Adele L. Boskey (477)
Mineralized Tissues Laboratory, Hospital for Special Surgery, Affiliated with Weil College of Cornell Medical School, New York, New York 10021, USA Roger Bouillon (135, 429, 1763)
Legendo, Onderwijs en Navorsing 902, U.Z. Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium Barbara D. Boyan (575)
Department of Biomedical Engineering at Georgia Tech and Emory University, Georgia Institute of Technology, Atlanta, Georgia 30332, USA Philippe Brachet (1779)
INSERM U 643, Centre Hospitalier Universitaire, 30 bd Jean Monnet, 44093 Nantes, France Alex J. Brown (1313, 1449)
Renal Division, Washington University School of Medicine, St. Louis, Missouri 63110, USA Edward M. Brown (551)
Division of Endocrinology, Diabetes and Hypertension, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts 02115, USA Carsten Carlberg (313)
Department of Biochemistry, University of Kuopio, PFIN-70211 Kuopio, Finland Geert Carmeliet (429)
Legendo, Onderwijs en Navorsing 902, U.Z. Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium Thomas O. Carpenter (1049)
Department of Pediatrics, Yale University, School of Medicine, New Haven, Connecticut 06520-8064, USA Marie-Claire Chapuy (1085)
INSERM Unit 403, Faculty Laennec and Department of Rheumatology and Bone Disease, Edouard Herriot Hosptial, Lyon, France Fredriech K. W. Chan (1355)
Department of Medicine, Queen Elizabeth Hospital, Hong Kong Tai C. Chen (1599)
Vitamin D, Skin, and Bone Research Laboratory, Boston University Medical Center, Boston, Massachusetts 02188, USA Sylvia Christakos (721)
Department of Biochemistry and Molecular Biology, University of Medicine and Dentistry of New Jersey, New Jersey Medical School, Newark, New Jersey 07103, USA
LIST
OF
CONTRIBUTORS
Margaret Clagett-Dame (1543)
Department of Biochemistry, University of Wisconsin–Madison, Madison, Wisconsin 53706, USA Thomas L. Clemens (899)
Department of Cell Biology and Physiology, Department of Internal Medicine, University of Cincinnati School of Medicine, Cincinnati, Ohio 45276, USA Je-Yong Choi (327)
Department of Biochemistry, Kyungpook National University, Daegu, Korea Fredric L. Coe (1339)
Nephrology Section, The University of Chicago Pritzker School of Medicine, Chicago, Illinois 60637, USA Kay Colston (1663)
Department OGEM, St. George’s Hospital Medical School, London, SW17 0RE, United Kingdom Juliet E. Compston (951)
Department of Medicine, Level 5, University of Cambridge, School of Clinical Medicine, Addenbrooke’s Hospital, Cambridge, England CB2 2QQ, United Kingdom Nancy E. Cooke (117)
Department of Genetics and Medicine, University of Pennsylvania, Philadelphia, Pennsylvania 19104-6149, USA Clara Crescioli (1833)
Endocrinology Unit, Department of Clinical Physiopathology, University of Florence, Florence 50139, Italy Heide S. Cross (1709)
Department of Pathophysiology, Medical, University of Vienna, A-1090 Vienna, Währingergürtel 18-20, Austria Michael Danilenko (1635)
Department of Clinical Biochemistry, Faculty of Health Science, Ben-Gurion University of the Negev, Beer-Sheva 84105, Israel Jean-Luc Davideau (599)
INSERM E110, Université Paris VII, IFR58, Cordeliers Biomedical Institute, 75270 Paris Cedex 06, France Michael Davies (1293)
Vitamin D Research Group, University Department of Medicine, Manchester Royal Infirmary, Manchester, M13 9WL, United Kingdom Hector F. DeLuca (3, 1543)
Department of Biochemistry, University of Wisconsin–Madison, Madison, Wisconsin 53706, USA Marie B. Demay (341)
Endocrine Unit, Department of Medicine, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts 02114, USA
LIST
OF
xv
CONTRIBUTORS
Puneet Dhawan (721)
Department of Biochemistry and Molecular Biology, University of Medicine and Dentistry of New Jersey, New Jersey Medical School, Newark, New Jersey 07103, USA Carlos Encinas Dominguez (219)
Department of Biochemistry and Molecular Biophysics, University of Arizona, Tucson, Arizona 85721, USA Diane R. Dowd (291)
Department of Pharmacology–W334, Case Western Reserve University, Cleveland, Ohio 44106, USA Marc K. Drezner (1159)
Department of Medicine, Endocrinology, Diabetes, and Metabolism Section, University of Wisconsin–Madison, Madison, Wisconsin 53792, USA Thomas W. Dunlop (313)
Department of Biochemistry, University of Kuopio, PFIN-70211 Kuopio, Finland Adriana S. Dusso (1313)
Renal Division, Washington University School of Medicine, St. Louis, Missouri 63110, USA Richard Eastell (1101)
University of Sheffield Clinical Sciences Centre, Northern General Hospital, Sheffield South Yorkshire S5 7AU, United Kingdom Michael J. Econs (1189)
Indiana University School of Medicine, Department of Medicine and Medical and Molecular Genetics, Indianapolis, Indiana 46202, USA John Eisman (193)
Bone and Mineral Program, Garvan Institute of Medical Research, St. Vincent’s Hospital, Sydney, New South Wales 2010, Australia Sol Epstein (1253)
Division of Endocrinology, Diabetes, and Bone Diseases, Department of Medicine, Mount Sinai School of Medicine, New York, New York 10029, USA Erik Fink Eriksen (1805)
Osteoporosis Team of Lilly Research Laboratories, Lilly Corp Center, Indianapolis, Indiana 46285, USA Luis M. Esteban (193)
Bone and Mineral Program, Garvan Institute of Medical Research, St. Vincent’s Hospital, Sydney, New South Wales 2010, Australia Dan Faibish (477)
Mineralized Tissues Laboratory, Hospital for Special Surgery, Affiliated with Weil College of Cornell Medical School, New York, New York 10021, USA
Yue Fang (1121)
Department of Internal Medicine, Genetic Laboratory, Erasmus Medical Centre, NL-3015 GE Rotterdam, The Netherlands Mary C. Farach-Carson (751)
Department of Biological Sciences, University of Delaware, Newark, Delaware 19716, USA Murray J. Favus (1339)
Section of Endocrinology, University of Chicago, Chicago, Illinois 60637, USA David Feldman (1207, 1679)
Stanford University School of Medicine, Division of Endocrinology, Stanford, California 94305-5103, USA David Findlay (711)
Department of Orthopedic Surgery and Trauma, University of Adelaide, Adelaide 5000, South Australia, Australia Christian Frank (313)
Department of Biochemistry, University of Kuopio, PFIN-70211 Kuopio, Finland Leonard P. Freedman (263)
Department of Molecular Endocrinology and Bone Biology, Merck Research Laboratories, West Point, Pennsylvania 19486-0004, USA Ryuji Fujiki (305)
University of Tokyo, Institute of Molecular and Cellular Biosciences, 1-1-1 Yayoi-cho, Bunkyo-ku, Tokyo, 113-0032, Japan Masafumi Fukagawa (1821)
Division of Nephrology and Dialysis Center, Kobe University School of Medicine, 7-5-2 Kusunoki-cho, Chuo-ku, Kobe 650-0017, Japan Robert F. Gagel (687)
Section of Endocrinology, University of Texas M.D. Anderson Cancer Center, Houston, Texas, USA Emmanuel Garcion (1779)
INSERM U 646, 10 rue André Boquel, 49100 Angers, France Edith M. Gardiner (193)
Bone and Mineral Program, Garvan Institute of Medical Research, St. Vincent’s Hospital, Sydney, New South Wales 2010, Australia Marielle Gascon-Barré (47)
Département de Pharmacologie, Faculté de Médecine, Université de Montréal, and Centre de recherche de l’Université de Montréal, Montréal, Québec H2X 1P1, Canada Edward Giovannucci (1617)
Harvard School of Public Health, Department of Nutrition, Boston, Massachusetts 02115, USA Henning Glerup (1805)
Aarhus Kommunehospital, Dept V, Noerrebrogade 44, DK-8000 Aarhus C, Denmark
xvi Francis H. Glorieux (1197)
Genetics Unit, Shriners Hospital for Children, Departments of Surgery, Pediatrics, and Human Genetics, McGill University, Montréal, Québec H3G 1A6, Canada Wagn O. Godtfredsen (1489)
Medicinal Chemistry, Leo Pharma, DK-2750 Ballerup, Denmark David Goltzman (737)
Department of Medicine, McGill University and McGill University Health Center, Montréal, Québec H3A 1A1, Canada Soraya Gutierrez (327)
Departamento de Biología Molecular, Universidad de Concepción, Concepción, Chile Conny Gysemans (1763)
Legendo, Onderwijs en Navorsing 902, U.Z. Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium Bernard P. Halloran (823)
Division of Endocrinology, Veterans Affairs Medical Center, San Francisco, California 94121, USA Carol A. Haussler (219)
Department of Biochemistry and Molecular Biophysics, University of Arizona, Tucson, Arizona 85721, USA Mark R. Haussler (219)
Department of Biochemistry and Molecular Biophysics, University of Arizona, Tucson, Arizona 85721, USA Robert P. Heaney (773)
Creighton University, Omaha, Nebraska 68131, USA Johan Heersche (649)
Faculty of Dentistry, University of Toronto, Toronto M5S 1A8, Ontario, Canada Helen L. Henry (69)
Department of Biochemistry, University of California–Riverside, Riverside, California 92521, USA Pamela A. Hershberger (1741)
Department of Pharmacology, University of Pittsburgh, Pittsburgh, Pennsylvania 15213, USA Martin Hewison (1379)
Division of Medical Sciences, The University of Birmingham, Queen Elizabeth Medical Centre, Birmingham, B15 2TH, United Kingdom Richard A. Heyman (1557)
X-Ceptor Therapeutics, San Diego, California 92121, USA Kanji Higashio (665)
Research Center for Genomic Medicine, Saitama Medical School, Saitama, 350-1241, Japan
LIST
OF
CONTRIBUTORS
Michael F. Holick (37, 1511, 1791)
Vitamin D, Skin, and Bone Research Laboratory; Department of Medicine; Endocrinology, Nutrition and Diabetes Section; Boston Medical Center and Boston University School of Medicine, Boston, Massachusetts 02118, USA Bruce W. Hollis (931)
Departments of Pediatrics, Biochemistry, and Molecular Biology, Medical University of South Carolina, Charleston, South Carolina 29425, USA Ronald L. Horst (15)
U.S. Department of Agriculture, Agricultural Research Service, National Animal Disease Center, Ames, Iowa 50010-0070, USA Jui-Cheng Hsieh (219)
Department of Biochemistry and Molecular Biophysics, University of Arizona, Tucson, Arizona 85721, USA Karl L. Insogna (1049)
Department of Medicine, Yale University School of Medicine, New Haven, CT 06520-8064, USA Elizabeth T. Jacobs (219)
College of Medicine, University of Arizona, Tucson, Arizona 85721, USA Amjad Javed (327)
Department of Cell Biology, University of Massachusetts Medical School, Worcester, Massachusetts 01655, USA Glenville Jones (1423)
Department of Biochemistry, Queen’s University, Kingston, Ontario K7L 3N6, Canada Candace S. Johnson (1741)
Department of Pharmacology & Therapeutics, Roswell Park Cancer Institute, Buffalo, New York 14263, USA Peter W. Jurutka (219)
Department of Biochemistry and Molecular Biophysics, University of Arizona, Tucson, Arizona 85721, USA Mehmet Kahraman (1405)
Department of Chemistry, Johns Hopkins University, Baltimore, Maryland 21218-2685, USA S. Kaleem Zaidi (327)
Department of Cell Biology, University of Massachusetts Medical School, Worcester, Massachusetts 01655, USA Heidi J. Kalkwarf (839)
Division of General and Community Pediatrics, Children’s Hospital Medical Center, Cincinnati, Ohio 45229, USA
LIST
OF
xvii
CONTRIBUTORS
Shigeaki Kato (305)
University of Tokyo, Institute of Molecular and Cellular Biosciences, 1-1-1 Yayoi-cho, Bunkyo-ku, Tokyo, 113-0032, Japan Anne-Marie Kissmeyer (1489)
Biological Research, Leo Pharma, DK-2750 Ballerup, Denmark Hirochika Kitagawa (305)
University of Tokyo, Institute of Molecular and Cellular Biosciences, 1-1-1 Yayoi-cho, Bunkyo-ku, Tokyo, 113-0032, Japan Lilia M. C. Koberle (1355)
Health Sciences Department, Federal University, Sao Carlos, Brazil H. Phillip Koeffler (1727)
Hematology/Oncology Division, Cedars-Sinai Medical Center, Los Angeles, California 90048, USA Ruth Koren (761)
Felsenstein Medical Research Center, Beilinson Campus, Rabin Medical Center, Petah Tikva 49100, Israel Barbara E. Kream (703)
Department of Medicine, University of Connecticut Health Center, Farmington, Connecticut 06030-1850, USA Richard Kremer (737)
Department of Medicine, McGill University and McGill University Health Center, Montréal, Québec H3A 1A1, Canada Aruna V. Krishnan (1679)
Stanford University School of Medicine, Division of Endocrinology, Stanford, California 94305-5103, USA Noboru Kubodera (1525)
Department of Product Planning, Chugai Pharmaceutical Co. Ltd., 2-1-9 Kyobashi, Chuo-ku, Tokyo, 104-8301, Japan Rajiv Kumar (515)
Departments of Medicine, Biochemistry and Molecular Biology and Mayo Proteomics Research Center, Divisions of Nephrology, and Endocrinology, Diabetes, Metabolism, and Nutrition, Mayo Clinic and Foundation, Rochester, Minnesota 55905-0002, USA Kiyoshi Kurokawa (1821)
Research Center for Advanced Science and Technology, The University of Tokyo, Tokyo, Japan Christopher J. Laing (117)
Department of Genetics and Medicine, University of Pennsylvania, Philadelphia, Pennsylvania 19104-6144, USA Jacques Lemire (1753)
Pediatric Nephrology, University of California-San Diego, La Jolla, California 92093-0831, USA
Frédéric Lézot (599)
INSERM E110, Université Paris VII, IFR58, Cordeliers Biomedical Institute, 75270 Paris Cedex 06, France Yan Chun Li (721, 871, 1511)
Department of Medicine/GI Section, University of Chicago, Chicago, Illinois 60637, USA Jane B. Lian (327)
Department of Cell Biology, University of Massachusetts Medical School, Worcester, Massachusetts 01655, USA Alexander C. Lichtler (703)
Department of Genetics and Developmental Biology, The University of Connecticut Health Center, Farmington, Connecticut 06030, USA Paul Lips (1019)
Department of Endocrinology, Vrijie University Medical Center, Amsterdam, 1007 MB, The Netherlands Yan Liu (721)
Department of Biochemistry and Molecular Biology, New Jersey Medical School, Newark, New Jersey 07103, USA Paul MacDonald (291)
Department of Pharmacology–W334, Case Western Reserve University, Cleveland, Ohio 44106, USA Hubert Maehr (1511)
BioXell, Inc., Hoffmann-La Roche, Inc., Nutley, New Jersey 07110-1199, USA Mario Maggi (1833)
Andrology Unit, Department Clinical Physiopathology, University of Florence, 50139 Florence, Italy Peter J. Malloy (1207)
Stanford University School of Medicine, Division of Endocrinology, Stanford, California 94305-5103, USA David J. Mangelsdorf (863)
Department of Pharmacology, Howard Hughes Medical Institute, University of Texas Southwestern Medical Center at Dallas, Dallas, Texas 75390-9050, USA Chantal Mathieu (1763)
Legendo, Onderwijs en Navorsing 902, U.Z. Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium Brian May (85)
School of Molecular and Biomedical Science, University of Adelaide, South Australia 5005, Australia Andrew P. Mee (1293)
Vitamin D Research Group, University Department of Medicine, Manchester Royal Infirmary, Manchester, M13 9WL, United Kingdom
xviii Pierre J. Meunier (1085)
INSERM Unit 403, Faculty Laennec and Department of Rheumatology and Bone Disease, Edouard Herriot Hospital, Lyon, France Toshimi Michigami (851)
Department of Environmental Medicine, Osaka Medical Center and Institute for Maternal and Child Health, Osaka, Japan Martin Montecino (327)
Departamento de Biología Molecular, Universidad de Concepción, Concepción, Chile Dino Moras (279)
Département de Biologie et de Génomique Structurales, CNRS/INSERM/Université Louis Pasteur 1, BP 10142, 67404 Illkirch Cedex, France Roberta Morosetti (1727)
Pediatric Oncology Division, Catholic University of Rome, Rome, Italy Howard A. Morris (711)
Hanson Institute, Adelaide, South Australia, Australia Daniel L. Motola (863)
Department of Pharmacology, Howard Hughes Medical Institute, University of Texas Southwestern Medical Center at Dallas, Dallas, Texas 75390-9050, USA Josephia Muindi (1741)
Department of Medicine, Roswell Park Cancer Institute, Buffalo, New York 14263, USA Shigeo Nakajima (851)
Department of Developmental Medicine (Pediatrics), D-5, Osaka University Graduate School of Medicine, Osaka 565-0871, Japan Tally Naveh-Many (537)
Minerva Center for Calcium and Bone Metabolism, Hebrew University Hadassah Medical Center, Ein Karem, Jerusalem, 91120, Israel Philippe Naveilhan (1779)
INSERM U 643, Centre Hospitalier Universitaire, 30 bd Jean Monnet, 44093 Nantes, France Isabelle Neveu (1779)
INSERM U 643, Centre Hospitalier Universitaire, 30 bd Jean Monnet, 44093 Nantes, France Anthony W. Norman (381)
Department of Biochemistry, University of California, Riverside, Riverside, California 92521-0129, USA Anders Nykjaer (153)
Institute of Medical Biochemistry, University of Aarhus, Ole Worms Allee, DK-8000 Aarhus C, Denmark James O’Kelly (1727)
Hematology/Oncology Division, Cedars-Sinai Medical Center, Los Angeles, California 90048, USA
LIST
OF
CONTRIBUTORS
John L. Omdahl (85)
Office of Research, University of New Mexico School of Medicine, Albuquerque, New Mexico 87131-5166, USA Peter Ordentlich (1557)
X-Ceptor Therapeutics, San Diego, California 92121, USA Keiichi Ozono (851)
Department of Developmental Medicine (Pediatrics), D-5, Osaka University Graduate School of Medicine, Osaka 565-0871, Japan A. Michael Parfitt (497, 1029)
Division of Endocrinology and Center for Osteoporosis and Metabolic Bone Disease, University of Arkansas for Medical Science, Little Rock, Arkansas 72205, USA Donna M. Peehl (1679)
Stanford University School of Medicine, Division of Endocrinology, Stanford, California 94305-5103, USA Sara Peleg (1471)
Department of Endocrine Neoplasia and Hormonal Diseases, Unit 435, University of Texas, M.D. Anderson Cancer Center, Houston, Texas 77030-4009, USA Xiaorong Peng (721)
Department of Biochemistry and Molecular Biology, University of Medicine and Dentistry of New Jersey, New Jersey Medical School, Newark, New Jersey 07103, USA John M. Pettifor (1065)
Department of Pediatrics, Chris Hani Baragwanath Hospital, Mineral Metabolism Research Unit, P.O. Bertsham, Johannesburg, Gauteng 2013, South Africa J. Wesley Pike (167, 1207, 1403, 1543)
Department of Biochemistry, University of Wisconsin–Madison, Madison, Wisconsin 53706, USA Elizabeth A. Platz (1617)
Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland 21205, USA Lori A. Plum (1543)
Department of Biochemistry, University of Wisconsin–Madison, Madison, Wisconsin 53706, USA Shirwin Pockwinse (327)
Department of Cell Biology, University of Massachusetts Medical School, Worcester, Massachusetts 01655, USA Huibert A.P. Pols (1121, 1571)
Department of Internal Medicine, Erasmus Medical Centre, NL-3015 GD Rotterdam, The Netherlands
LIST
OF
xix
CONTRIBUTORS
Anthony A. Portale (453, 823)
Department of Pediatrics, University of California – San Francisco, San Francisco, California 94121, USA Gary H. Posner (1405)
Department of Chemistry, Johns Hopkins University, Baltimore, Maryland 21218-2685, USA Mehrdad Rahmaniyan (789)
Department of Medicine, Medical University of South Carolina, Charleston, South Carolina 29425, USA Amiram Ravid (761)
Felsenstein Medical Research Center, Beilinson Campus, Rabin Medical Center, Petah Tikva 49100, Israel G. Satyanarayana Reddy (15, 1511)
Brown University, Department of Chemistry, Providence, Rhode Island, USA Jörg Reichrath (1791)
The Saarland University Hospital, Department of Dermatology, Kirrberger Str., 66421 Homburg/Saar, Germany Timothy A. Reinhardt (15)
U.S. Department of Agriculture, Agricultural Research Service, National Animal Disease Center, Ames, Iowa 50010-0070, USA Alfred A. Reszka (263)
Department of Molecular Endocrinology and Bone Biology, Merck Research Laboratories, West Point, Pennsylvania 19486-0004, USA B. Lawrence Riggs (1101)
Division of Endocrinology, Mayo Clinic and Mayo Foundation, Rochester, Minnesota 55905, USA Natacha Rochel (279)
Département de Biologie et de Génomique Structurales, CNRS/INSERM/Université Louis Pasteur 1, BP 10142, 67404 Illkirch Cedex, France Mishaela R. Rubin (1355)
Department of Medicine, College of Physicians and Surgeons, Columbia University, New York, New York 10027, USA Andrew F. Russo (687)
Department of Physiology and Biophysics, University of Iowa, Iowa City, Iowa 52242, USA Philip Sambrook (1239)
Institute of Bone & Joint Research, University of Sydney, Royal North Shore Hospital, Sydney 2065, Australia Adina E. Schneider (1253)
Division of Endocrinology, Diabetes, and Bone Diseases, Department of Medicine, Mount Sinai School of Medicine, New York, New York 10029, USA
Gary G. Schwartz (1599)
Comprehensive Cancer Center, Wake Forest University School of Medicine, Winston-Salem, North Carolina 27157, USA Zvi Schwartz (575)
Department of Periodontics, Hebrew University Hadassah Faculty of Dental Medicine, Jerusalem, Israel Jiali Shen (327)
Department of Cell Biology, University of Massachusetts Medical School, Worcester, Massachusetts 01655, USA Nirupama K. Shevde (167, 1543)
Department of Biochemistry, University of Wisconsin–Madison, Madison, Wisconsin 53706, USA Rafal R. Sicinski (1543)
Department of Chemistry, University of Warsaw, ul. L. Pasteura 2, Warsaw 02-093, Poland Justin Silver (537)
Minerva Center for Calcium and Bone Metabolism, Hebrew University Hadassah Medical Center, Ein Karem, Jerusalem, 91120, Israel Eduardo A. Slatopolsky (1313, 1449)
Renal Division, Washington University School of Medicine, St. Louis, Missouri 63110, USA Bonny L. Specker (839)
Martin Program in Human Nutrition, South Dakota State University, Brookings, South Dakota 57007, USA René St-Arnaud (105, 1197)
Genetics Unit, Shriners Hospital for Children, Montréal, Quebec H3G 1A6, Canada Gary S. Stein (327)
Department of Cell Biology, University of Massachusetts Medical School, Worcester, Massachusetts 01655, USA Janet L. Stein (327)
Department of Cell Biology, University of Massachusetts Medical School, Worcester, Massachusetts 01655, USA Paula H. Stern (565)
Department of Molecular Pharmacology, Northwestern University Feinberg School of Medicine, Chicago, Illinois 60611, USA George P. Studzinski (1635)
UMD-New Jersey Medical School, Newark, New Jersey 07103, USA Tatsuo Suda (665)
Research Center for Genomic Medicine, Saitama Medical School, Hidaka-shi, Saitama 350-1241, Japan
xx
LIST
Amelia L.M. Sutton (291)
Department of Pharmacology–W334, Case Western Reserve University, Cleveland, Ohio 44106, USA Peter Tebben (515)
Departments of Medicine, Biochemistry and Molecular Biology and Mayo Proteomics Research Center, Divisions of Nephrology, and Endocrinology, Diabetes, Metabolism, and Nutrition, Mayo Clinic and Foundation, Rochester, Minnesota 55905-0002, USA Harriet S. Tenenhouse (453)
Departments of Pediatrics and Human Genetics, McGill University and Montréal Children’s Hospital Research Institute, Montréal, Québec, H3Z 2Z3, Canada Michelle L. Thatcher (219)
Department of Biochemistry and Molecular Biophysics, University of Arizona, Tucson, Arizona 85721, USA Susan Thys-Jacobs (1355)
Department of Medicine, College of Physicians and Surgeons, Columbia University, New York, New York 10027, USA Dwight A. Towler (899)
Department of Medicine, Washington University School of Medicine, St. Louis, Missouri 63110, USA Donald L. Trump (1741)
Department of Medicine, Roswell Park Cancer Institute, Buffalo, New York 14263, USA André G. Uitterlinden (1121)
Genetic Laboratory, Department of Internal Medicine, Erasmus Medical Centre, NL-3015 GD Rotterdam, The Netherlands Milan R. Uskokovi´c (1511)
BioXell, Inc., Nutley, New Jersey 07110-1199, USA Sami Väisänen (313)
Department of Biochemistry, University of Kuopio, PFIN-70211 Kuopio, Finland Hugo Van Baelen (135)
Legendo, Onderwijs en Navorsing 902, U.Z. Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium Sophie Van Cromphaut (429)
Legendo, Onderwijs en Navorsing 902, U.Z. Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium Johannes P.T.M. van Leeuwen (1571)
Department of Internal Medicine, Erasmus MC, 3000 DR Rotterdam, The Netherlands
OF
CONTRIBUTORS
Joyce B.J. van Meurs (1121)
Genetic Laboratory, Department of Internal Medicine, Erasmus Medical Centre, NL-3015 GD Rotterdam, The Netherlands Andre J. van Wijnen (327)
Department of Cell Biology, University of Massachusetts Medical School, Worcester, Massachusetts 01655, USA Christel Verboven (135)
Laboratory Analytische Chemie, Van Evenstraat 4; B-3000 Leuven, Belgium Reinhold Vieth (995)
Pathology and Laboratory Medicine, Mount Sinai Hospital, Toronto, Ontario, M5G 1X5, Canada Robert H. Wasserman (411)
Department of Biomedical Sciences, VRT-08-20, Cornell University, Ithaca, New York 14853, USA JoEllen Welsh (1663)
Department Biological Science, University Notre Dame, Notre Dame, Indiana 46556, USA G. Kerr Whitfield (219)
Department of Biochemistry and Molecular Biophysics, University of Arizona, Tucson, Arizona 85721, USA Michael P. Whyte (913)
Center for Metabolic Bone Disease and Molecular Research, Shriners Hospital for Children, St. Louis, Missouri 63131, USA and Division of Bone and Mineral Diseases, Washington University School of Medicine at Barnes-Jewish Hospital, St. Louis, Missouri 63110, USA. Thomas Willnow (153)
Division of Molecular Cardiovascular Research, Max-Delbrueck-Center for Molecular Medicine, Robert-Roessle-Strasse 10, D-13125 Berlin, Germany Didier Wion (1779)
INSERM U318, Centre Hospitalier Michallon, 38043 Grenoble cedex 09, France Hisataka Yasuda (665)
Center for Experimental Medicine, Institute of Medical Science, University of Tokyo, Tokyo 108-8639, Japan Daniel Young (327)
Department of Cell Biology, University of Massachusetts Medical School, Worcester, Massachusetts 01655, USA Chi Zhang (291)
Department of Pharmacology–W334, Case Western Reserve University, Cleveland, Ohio 44106, USA
Preface to the 2nd Edition
Those interested in the vitamin D field will not be surprised that this second edition is considerably larger than the first edition. A great deal of progress has been made since the first edition was published in 1997. However, our goal in planning this updated version remains the same. We have endeavored to provide investigators, clinicians, and students with a comprehensive, definitive, and up-to-date compendium of the diverse scientific and clinical aspects of vitamin D, each area covered by experts in the field. Our hope for the second edition is that this book will continue to serve as both a resource for current researchers, as well as a guide to assist those in related disciplines to enter the realm of vitamin D research. We hope that this book will illuminate the vitamin D field and help investigators identify areas where new research is needed as well as educate them about what is currently known. We believe that the first edition succeeded in stimulating interactions between researchers and clinicians from different disciplines and that it facilitated collaborations. As we move from basic science and physiology to the use of vitamin D and its analogs as pharmacological agents to treat various diseases, the need for cross-collaborations between researchers and clinicians from different disciplines will increase. We hope that this new volume will continue to be a valuable resource that plays a role in this advancement and stimulates and facilitates these interactions. Enormous progress in the study of vitamin D has been made in the approximately eight years since the first edition was written and we hope that this book has contributed in some way to this progress. The first edition proved to be highly valuable to its readers and the chapters have been cited frequently as authoritative reviews of the field. However, it became clear to us that the time was ripe to organize a second edition. Building on the original, we hope this second edition
will incorporate all of the progress made in the field since the first edition was published so that our objective of an up-to-date compendium containing everything you wanted to know about vitamin D will continue. The second edition is essentially a new and reinvigorated book. We have changed the symbol on the cover to reflect its updated content and the field's continued evolution into the molecular world. This new edition now includes 104 chapters. In order to cover the growth of new information on vitamin D, we have had to publish this new edition in two volumes. In addition, the book has undergone some major remodeling. There are 33 completely new chapters and 18 other chapters have had major changes in authorship and are totally rewritten. While approximately half of the chapters have some of the same authors, all have had major updates and many have new co-authors with new perspectives. We have endeavored to attract the leading investigators in each field to author the chapter covering their area. We are especially pleased with the roster of authors who have written for the second edition. They really are the leaders in their respective fields. We wish to express our thanks to Tari Paschall, Judy Meyer, and Sarah Hajduk as well as the rest of the Elsevier/Academic Press staff for their expertise and indispensable contribution to bringing this revised edition to fruition. Most of all, we thank the authors for their contributions. We hope that our readers will find this updated volume useful and informative and that it will contribute to the burgeoning growth of the vitamin D field. DAVID FELDMAN J. WESLEY PIKE FRANCIS H. GLORIEUX
Preface to the 1st Edition
Our reasons for deciding to publish an entire book devoted to vitamin D can be found in the rapid and extensive advances currently being made in this important field of research. Enormous progress in investigating many aspects of vitamin D, from basic science to clinical medicine, has been made in recent years. The ever-widening scope of vitamin D research has created new areas of inquiry so that even workers immersed in the field are not fully aware of the entire spectrum of current investigation. Our goal in planning this book was to bring the diverse scientific and clinical fields together in one definitive and up-to-date volume. It is our hope that this compendium on vitamin D will serve as both a resource for current researchers and a guide to stimulate and assist those in related disciplines to enter this field of research. In addition, we hope to provide clinicians and students with a comprehensive source of information for the varied and extensive material related to vitamin D. The explosion of information in the vitamin D sphere has led to new insights into many different areas, and in our treatment of each subject in this book we have tried to emphasize the recent advances as well as the established concepts. The classic view of vitamin D action, as a hormone limited to calcium metabolism and bone homeostasis, has undergone extensive revision and amplification in the past few years. We now know that the vitamin D receptor (VDR) is present in most tissues of the body and that vitamin D actions, in addition to the classic ones, include important effects on an extensive array of other target organs. To cover this large number of subjects, we have organized the book in the following manner: Section I, the enzymes involved in vitamin D metabolism and the activities of the various metabolites; Section II, the mechanism of action of vitamin D, including rapid, nongenomic actions and the role of the VDR in health and disease; Section III, the
effects of vitamin D and its metabolites on the various elements that constitute bone and the expanded understanding of vitamin D actions in multiple target organs, both classic and nonclassic; Section IV, the role of vitamin D in the physiology and regulation of calcium and phosphate metabolism and the multiplicity of hormonal, environmental, and other factors influencing vitamin D metabolism and action; and Sections V and VI, the role of vitamin D in the etiology and treatment of rickets, osteomalacia, and osteoporosis and the pathophysiological basis, diagnosis, and management of numerous clinical disorders involving vitamin D. The recent recognition of an expanded scope of vitamin D action and the new investigational approaches it has generated were part of the impetus for developing this volume on vitamin D. It has become clear that in addition to the classic vitamin D actions, a new spectrum of vitamin D activities that include important effects on cellular proliferation, differentiation, and the immune system has been identified. This new information has greatly expanded our understanding of the breadth of vitamin D action and has opened for investigation a large number of new avenues of research that are covered in Sections VII and VIII of this volume. Furthermore, these recently recognized nonclassic actions have led to a consideration of the potential application of vitamin D therapy to a range of diseases not previously envisioned. This therapeutic potential has spawned the search for vitamin D analogs that might have a more favorable therapeutic profile, one that is less active in causing hypercalcemia and hypercalciuria while more active in a desired application such as inducing antiproliferation, prodifferentiation, or immunosuppresion. Since 1α,25-dihydroxyvitamin D [1,25(OH)2D] and its analogs are all presumed to act via a single VDR, a few years ago most of us probably would have thought that it was impossible to achieve a
xxiv separation of these activities. Yet today, many analogs that exhibit different profiles of activity relative to 1,25(OH)2D have already been produced and extensively studied. The development of analogs with an improved therapeutic index has opened another large and complex area of vitamin D research. This work currently encompasses three domains: (1) the design and synthesis of vitamin D analogs exhibiting a separation of actions with less hypercalcemic and more antiproliferative or immunosuppressive activity, (2) the interesting biological question of the mechanism(s) by which these analogs achieve their differential activity, and (3) the investigation of the potential therapeutic applications of these analogs to treat various disease states. These new therapeutic applications, from psoriasis to cancer, from immunosuppression to neurodegenerative diseases, have drawn into the field an expanded population of scientists and physicians interested in vitamin D. Our goal in editing this book was to create a comprehensive resource on vitamin D that would be of use to a mix of researchers in different disciplines. To achieve this goal, we sought authors who had contributed greatly to their respective fields of vitamin D research. The book has a large number of chapters to accommodate
PREFACE
TO THE
1ST EDITION
many contributors and provide expertise in multiple areas. Introductory chapters in each section of the book are designed to furnish an overview of that area of vitamin D research, with other chapters devoted to a narrowly focused subject. Adjacent and closely related subjects are often covered in separate chapters by other authors. While this intensive style may occasionally create some redundancy, it has the advantage of allowing the reader to view the diverse perspectives of the different authors working in overlapping fields. In this regard, we have endeavored to provide many crossreferences to guide the reader to related information in different chapters. We express our thanks to Jasna Markovac (Editor-inChief), for encouraging us to develop this book and guiding us through the process; to Tari Paschall (Acquisitions Editor) and the Academic Press staff for their diligence, expertise, and patience in helping us complete this work. Most of all, we thank the authors for their contributions that have made this book possible. DAVID FELDMAN J. WESLEY PIKE FRANCIS H. GLORIEUX
Abbreviations
AA AC ACE ACF ACTH ADH ADHR ADP AHO AI AIDS Aj.AR ALP ANG II ANP APC APD AR ARC ATP ATRA AUC Bmax BARE bFGF BFU BGP BLM BMC BMD BMI BMP BMU bp BPH
arachadonic acid adenyl cyclase angiotensin converting enzyme activation frequency adrenocorticotropin antidiuretic hormone (vasopressin) autosomal dominant hypophosphatemic rickets adenosine diphosphate Albright’s hereditary osteodystrophy adequate intake acquired immunodeficiency syndrome adjusted apposition rate alkaline phosphatase angiotensin II atrial natriuretic peptide antigen presenting cell aminohydroxypropylidene bisphosphonate androgen receptor activator recruited cofactor adenosine triphosphate all-trans-retinoic acid area under the curve maximum number of binding sites bile acid response element basic fibroblast growth factor burst-forming unit bone Gla protein (osteocalcin) basal lateral membrane bone mineral content bone mineral density body mass index bone morphogenetic protein basic multicellular unit base pairs benign prostatic hyperplasia
BSA BUA [Ca2+]i
bovine serum albumin bone ultrasound attentuation internal calcium ion molar concentration CaBP calcium binding protein CAD coronary artery disease CaM calmodulin cAMP cyclic AMP CaSR or CaR calcium sensing receptor CAT chloramphenicol acetyltransferase CBG corticosteroid-binding globulin CBP competitive protein binding assay CC chief complaint CDCA chenodeoxycholic acid CDK or Cdk cyclin-dependent kinase cDNA complementary DNA CDP collagenase-digestible protein Cdx-2 caudal-related homeodomain protein CFU colony-forming unit cGMP cyclic GMP CGRP calcitonin gene-related peptide CHF congestive heart failure CK-II casein kinase-II CLIA competitive chemiluminescence immunoassay cM centimorgans Cm. Ln. cement line CNS central nervous system CPBA competitive protein binding assays cpm counts per minute CRE cAMP response element CREB cAMP response element binding protein CRF chronic renal failure CsA cyclosporin A CSF colony-stimulating factor
xxvi CT CTR CTX CVC CYP CYP24 DAG DBD DBP DBP DC DCA DCT DEXA or DXA 7-DHC DHEA DHT DIC DMSO DR DRIP DSP E1 E2 EAE EBT EBV EC EC50 or ED50 ECaC ECF EDTA EGF ELISA EMSA EP1 ER ERE ERK Et FACS FAD FCS FDA FFA FIT
ABBREVIATIONS
calcitonin or computerized tomography calcitonin receptor cerebrotendinous xanthomatosis calcifying vascular cells cytochrome P450 cytochrome P450, 24-hydroxylase diacylglycerol DNA binding domain diastolic blood pressure vitamin D binding protein dendritic cell deoxycholic acid distal convoluted tubule dual energy X-ray absorptiometry 7-dehydrocholesterol dehydroepiandrosterone dihydrotachysterol or dihydrotestosterone disseminated intravascular coagulation dimethyl sulfoxide direct repeat vitamin D receptor interacting protein dental sialoprotein estrone estradiol experimental autoimmune encephalitis electron-beam computed technology Epstein-Barr virus endothelial cells effective concentration (dose) to cause 50% effect epithelium calcium channel extracellular fluid ethylenediaminetetraacetic acid epidermal growth factor enzyme-linked immunosorbent assay electrophoretic mobility shift assay PG receptor-1 estrogen receptor or endoplasmic reticulum estrogen response element extracellular signal-regulated kinase endothelin fluorescence-activated cell sorting or sorter flavin adenine dinucleotide fetal calf serum U.S. Food and Drug Administration free fatty acid Fracture Intervention Trial
FMTC FP FRAP FS FSK FXR g g G0, G1, G2 GAG GC-MS G-CSF GDNF GFP GFR GH GHRH GIO GM-CSF GnRH GR GRE GRTH HAT HDAC HEK HHRH HIV HNF HPI HPLC HPV hr HRE HSA Hsp HSV HVDRR HVO IBMX IC50 ICA ICMA IDBP IDDM IDM IFN
familial medullary thyroid carcinoma formation period fluorescence recovery after photobleaching Fanconi syndrome forskolin farnesol X receptor gram acceleration due to gravity gap phases of the cell cycle glycosaminoglycan gas chromatography-mass spectrometry granulocyte colony-stimulating factor glial cell-derived neurotrophic factor green fluorescent protein glomerular filtration rate growth hormone growth hormone-releasing hormone glucocorticoid induced osteoporosis granulocyte-macrophage colonystimulating factor gonadotropin-releasing hormone glucocorticoid receptor glucocorticoid response element generalized resistance to thyroid hormone histone acetyltransferase histone deacetylase human embryonic kidney hereditary hypophosphatemic rickets with hypercalciuria human immunodeficiency virus hepatocyte nuclear factor history of present illness high-performance liquid chromatography human papilloma virus hour hormone response element human serum albumin heat-shock protein herpes simplex virus hereditary vitamin D-resistant rickets hypovitaminosis D osteopathy isobutylmethylxanthine concentration to inhibit 50% effect intestinal calcium absorption immunochemiluminometric assay intracellular vitamin D–binding protein insulin-dependent diabetes mellitus infants of diabetic mothers interferon
xxvii
ABBREVIATIONS
Ig IGFBP IGF-I, -II IGF-IR IL i.m. IMCal i.p. IP3 IRMA IU IUPAC i.v. JG JNK Kd Km kb kbp kDa KO LBD LCA LDL Li. Ce. LIF LNH LOD LPS LT LXR M M MAPK Mab MAR MAR MARRS MCR M-CSF MEN2 MGP MHC min MLR Mlt MR MRI mRNA MTC NADH
immunoglobulin IGF binding protein insulin-like growth factor type I, II IGF-I receptor interleukin (e.g., IL-1, IL-1β, etc.) intramuscular intestinal membrane calcium binding complex intraperitoneal inositol trisphosphate immunoradiometric assay international units International Union of Pure and Applied Chemists intravenous juxtaglomerular c-Jun NH2-terminal kinase dissociation constant Michaelis constant kilobases kilobase pairs kilodaltons knock out ligand binding domain lithocholic acid low-density lipoprotein lining cell leukemia inhibitory factor late neonatal hypocalcemia logarithm of the odds lipopolysaccharide leukotriene liver X receptor mitosis phase of cell cycle molar mitogen-activated protein kinase monoclonal antibody matrix attachment region mineral apposition rate membrane-associated rapid response steroid metabolic clearance rate macrophage colony-stimulating factor multiple endocrine neoplasia type 2 matrix Gla protein major histocompatibility complex minute mixed lymphocyte reaction mineralization lag time mineralcorticoid receptor magnetic resonance imaging messenger ribonucleic acid medullary thyroid carcinoma nicotinamide adenine dinucleotide
NADPH
nicotinamide adenine dinucleotide phosphate NAF nuclear accessory factor NBT nitroblue tetrazolium NcAMP nephrogenous cAMP NCP noncollagen protein NGF nerve growth factor NHANES III National Health and Nutrition Examination Survey III NHL Non-Hodgkin’s lymphoma NIDDM non-insulin-dependent diabetes mellitus NIH National Institutes of Health NK cell natural killer cell NLS nuclear localization signal NMR nuclear magnetic resonance NPT sodium/phosphate cotransporter NR nuclear receptor Ob osteoblast Oc osteocalcin or osteoclast OCIF osteoclastogenesis inhibitory factor (same as OPG) OCT 22-oxacalcitriol ODF osteoclast differentiation factor (same as RANKL) 1α-OHD3 1α-hydroxyvitamin D3 25OHD3 25-hydroxyvitamin D3 1,25(OH)2D3 1α,25-dihydroxyvitamin D3 24,25(OH)2D3 24,25-dihydroxyvitamin D3 OHO oncogenic hypophosphatemic osteomalacia Omt osteoid maturation time OPG osteoprotegerin OPN osteopontin OSM oncostatin M OVX ovariectomy Pi inorganic phosphate PA2 phospholipase A2 PAD peripheral arterial vascular disease PAM pulmonary alveolar macrophage PBL peripheral blood lymphocyte PBS phosphate-buffered saline PC phophatidyl choline PCNA proliferating cell nuclear antigen PCR polymerase chain reaction PCT proximal convoluted tubule PDDR pseudovitamin D deficiency rickets PDGF platelet-derived growth factor PEIT percutaneous ethanol injection therapy PHEX phosphate regulating gene with homologies to endopeptidases on the X chromosome PG prostaglandin
xxviii PHA PHP PIC PKA PKC PKI PLA2 PLC PMA PMCA PMH p.o. poly(A) PPAR PR PRA PRL PSA PSI PT PTH PTHrP PTX PUVA QCT QSAR 9-cis-RA RA RA RANK RANKL RAP RAR RARE RAS RBP RCI RDA RFLP RIA RID RNase ROCs ROS RPA RRA RT-PCR RXR
ABBREVIATIONS
phytohemagglutinin pseudohypoparathyroidism preinitiation complex protein kinase A protein kinase C protein kinase inhibitor phospholipase A2 phospholipase C phorbol 12-myristate 13-acetate plasma membrane calcium pump past medical history oral polyadenosine peroxisome proliferator-activated receptor progesterone receptor plasma renin activity prolactin prostate-specific antigen psoriasis severity index parathyroid parathyroid hormone parathyroid hormone-related peptide parathyroidectomy psoralen-ultraviolet A quantitative computerized tomography quantitative structure-activity relationship 9-cis-retinoic acid retinoic acid rheumatoid arthritis receptor activator NF-κB receptor activator NF-κB ligand receptor-associated protein retinoic acid receptor retinoic acid response element renin–angiotensin system retinol-binding protein relative competitive index recommended dietary allowance restriction fragment length polymorphism radioimmunoassay receptor interacting domain ribonuclease receptor operated calcium channels reactive oxygen species ribonuclease protection assay radioreceptor assay reverse transcriptase-polymerase chain reaction retinoid X receptor
RXRE SBP SD SDS SE SEM SH SHBG SLE SOS Sp1 SPF SRC-1 SSCP SV40 SXA t1/2 T3 T4 TBG TBP TC TF TFIIB TG TGF TIO TK TLR TmP or TmPi TNF TPA TPN TPTX TR TRAP TRAP TRP TRE TRE TRH Trk TSH TSS UF US USDA UTR UV VDDR-I
retinoid X receptor response element systolic blood pressure standard deviation sodium dodecyl sulfate standard error standard error of the mean social history sex hormone–binding globulin systematic lupus erythematosus speed of sound selective promoter factor 1 sun protection factor steroid receptor coactivator-1 single strand conformational polymorphism simian virus 40 single energy X-ray absorptiometry half-time triiodothyronine thyroxine thyroid-binding globulin TATA binding protein tumoral calcinosis tubular fluid general transcription factor IIB transgenic transforming growth factor tumor-induced osteomalacia thymidine kinase toll-like receptor tubular absorptive maximum for phosphorus tumor necrosis factor 12-O-tetradecanoylphorbol-13-acetate total parenteral nutrition thyroparathyroidectomized thyroid hormone receptor tartrate-resistant acid phosphatase thyroid hormone receptor associated proteins transient receptor potential thyroid hormone response element TPA response element thyrotropin-releasing hormone tyrosine kinase thyrotropin transcription start site ultrafiltrable fluid ultrasonography U.S. Department of Agriculture untranslated region ultraviolet vitamin D-dependent rickets type I (see PDDR)
xxix
ABBREVIATIONS
VDDR-II VDR VDRE VDRL VEGF VERT VICCs VSMC
vitamin D-dependent rickets type II (see HVDRR) vitamin D receptor vitamin D response element vitamin D receptor ligand vascular endothelial growth factor Vertebral Efficacy with Risedronate Therapy studies voltage-insensitive calcium channels vascular smooth muscle cell
VSSCs WHI WRE WSTF XLH XRD ZEB
voltage-senstive calcium channels Women’s Health Initiative Wilms’ tumor gene, WT1, responsive element Williams syndrome transcription factor X-linked hypophosphatemic rickets X-ray diffraction zinc finger, E box-binding transcription factor
Approximate Normal Ranges for Serum Values in Adultsa Measure
SI Units
Ionized calcium Total calcium Phosphorous, inorganic 25(OH)D 1,25(OH)2D
1.12–1.32 mmol/L 2.17–2.52 mmol/L 0.77–1.49 mol/L 24.9–169.5 nmol/L 60–108 pmol/L
Conventional Units
Conversion Factorb
4.5–5.3 mg/dL 8.7–10.1 mg/dL 2.4–4.6 mg/dL 10–68 ng/mL 25–45 pg/mL
0.2495 0.2495 0.3229 2.496 2.40
Approximate Normal Ranges for Serum Values in Childrena Measure Ionized calcium Total calcium Phosphorous, inorganic 25(OH)D 1,25(OH)2D aNormal
SI Units
Conventional Units
Conversion Factorb
1.19–1.29 mmol/L 2.25–2.63 mmol/L 1.23–1.62 mol/L 34–91 nmol/L 65–134 pmol/L
4.8–5.2 mg/dL 9.0–10.5 mg/dL 3.8–5.0 mg/dL 14–37 ng/mL 27–56 pg/mL
0.2495 0.2495 0.3229 2.496 2.40
ranges differ in various laboratories and these values are provided only as a general guide. factor X conventional units = SI units
bConversion
Useful Equivalencies of Different Units Vitamin D Calcium Phosphorus
1 µg = 40 IU 1 mmol = 40 mg 1 mmol = 30 mg
CHAPTER 1
Historical Perspective HECTOR F. DELUCA Department of Biochemistry, University of Wisconsin-Madison, Madison, Wisconsin
I. Discovery of the Vitamins II. Discovery That Vitamin D Is Not a Vitamin III. Isolation and Identification of Nutritional Forms of Vitamin D
IV. Discovery of the Physiological Functions of Vitamin D V. Discovery of the Hormonal Form of Vitamin D References
I. DISCOVERY OF THE VITAMINS
animals whereas purified materials could not [3]. Funk had found similar results for the prevention of neuritis and reasoned that there were “vital amines” present in foods from natural sources and actually provided the basis for the term “vitamins” used later to describe essential micronutrients [5].
A. Early Nutritional Views The field of nutrition was largely dominated in the nineteenth century by German chemists, led by Justus von Liebig [1]. They taught that adequacy of the diet could be described by an analysis of protein, carbohydrate, fat, and mineral. Thus, a diet containing 12% protein, 5% mineral, 10–30% fat, and the remainder as carbohydrate would be expected to support normal growth and reproduction. This view remained largely unchallenged until the very end of the 19th century and the beginning of the 20th century [2–5]. However, evidence opposing this view began to appear. One of the first was the famous study of Eijkman who studied prisoners in the Dutch East Indies maintained on a diet of polished rice [6]. A high incidence of the neurological disorder beri-beri was recorded in these inmates. Eijkman found that either feeding whole rice or returning the hulls of the polished rice could eliminate beri-beri. Eijkman reasoned that polished rice contained a toxin that was somehow neutralized by the rice hulls. Later, a colleague, Grijns [7], revisited the question and correctly demonstrated that hulls contained an important and required nutrient that prevented beri-beri. Other reports revealed that microorganic nutrients might be present. The development of scurvy in mariners was a common problem. This disease was prevented by the consumption of limes on British ships (hence, the term “Limey” to describe British sailors) and sauerkraut and fruits on other ships. This led Hoist and Frohlich to conclude that scurvy could be prevented by a nutrient present in these foods [8]. Experiments by Lunin, Magendie, Hopkins, and Funk showed that a diet of purified carbohydrate, protein, fat, and salt is unable to support growth and life of experimental animals [2–5]. This suggested that some unknown or vital factor present in natural foods was missing from the purified diets. Hopkins developed a growth test in which natural foods were found to support rapid growth of experimental VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX,
B. McCollum and Osborne and Mendel’s Discovery of Vitamin A and B Complex A key experiment demonstrating essential micronutrients was one carried out at the Wisconsin Agricultural Experiment Station, engineered by Stephen Moulton Babcock and carried out by E. B. Hart supported by McCollum and Steenbock [9]. Herds of dairy cows were maintained on a diet composed individually only of corn, oats, or wheat or were fed a mixture of all of these grains, all receiving the same amount of carbohydrate, protein, fat, and salts and all providing equal analysis according to the German chemists [1]. The animals on the corn diet did very well, produced milk in large amounts, and reproduced normally. Those on the wheat diet failed to thrive and soon were unable to reproduce or lactate. The oat group was found to be intermediate between the corn and wheat groups, and the mixture approximated the growth and reproduction found with corn. Yet all these diets had the same proximate analysis. The conclusion of the Wisconsin Experiment Station study was that there are unknown nutrients present in corn and not found in wheat that are essential for life and reproduction. This led E. B. Hart, Chairman of Biochemistry at Wisconsin, to conceive that a search for these nutrients must begin. Professor McCollum was asked to search for these nutrients using small experimental animals. McCollum and Davis demonstrated there was present in butter fat a substance that prevented xerophthalmia and was also required for growth. They termed this “a lipin-soluble growth factor” [10]. McCollum later named this factor “vitamin A” [11]. Copyright © 2005, Elsevier, Inc. All rights reserved.
4 This substance was absent from lard and other fats but was found in large amounts in cod liver oil. In constructing the diets, McCollum obtained the carbohydrates and salts from milk whey, which, unknown to him, supplied the vitamin B complex group of micronutrients that permitted him to observe a vitamin A deficiency. McCollum at Wisconsin [11] and Osborne and Mendel [12] at the Connecticut Experiment Station carried out experiments in which cod liver oil was used as a source of fat in the diet, but the minerals were supplied from pure chemicals mixed to approximate the mineral composition of milk. Starch or sugar was used as the carbohydrate. These animals developed a different group of symptoms, namely, neuritis, which could be cured by the provision of the milk components. McCollum and Osborne and Mendel correctly concluded that this activity was due to a different micronutrient called “vitamin B.” This ushered in the concept of the organic micronutrients known as vitamins.
C. History of Rickets The disease rickets was very likely known in antiquity but was described in the 15th century as revealed by later writings. Whistler first provided a clear description of rickets in which the skeleton was poorly mineralized and deformed [13]. Rickets undoubtedly in ancient times appeared only on rare occasions and hence was not considered a problem. However, at the end of the 19th century, the industrial revolution had taken place: A highly agrarian population had become urbanized, and smoke from the industrial plants polluted the atmosphere. Thus, in low-sunlight countries such as England, rickets appeared in epidemic proportions. In fact, it was known as “the English disease” [14]. Some reports of the beneficial action of cod liver oil had appeared. However, they were not given scientific credence. With the discovery of the vitamins, Sir Edward Mellanby in Great Britain began to reason that rickets might also be a disease caused by a dietary deficiency [15]. Mellanby fed dogs a diet composed primarily of oatmeal, which was the diet consumed where the incidence of rickets was the highest (i.e., Scotland). McCollum inadvertently maintained the dogs on oatmeal indoors and away from ultraviolet light. The dogs developed severe rickets. Learning from the experiments of McCollum, Mellanby provided cod liver oil to cure or prevent the disease. Mellanby could not decide whether the healing of rickets was due to vitamin A known to be present in the cod liver oil or whether it was a new and unknown substance. Therefore, the activity of healing rickets was first attributed to vitamin A.
HECTOR F. DELUCA
D. Discovery of Vitamin D McCollum, who had moved to Johns Hopkins from Wisconsin, continued his experiments on the fat-soluble materials. McCollum used aeration and heating of cod liver oil to destroy the vitamin A activity or the ability to support growth and to prevent xerophthalmia [16]. However, cod liver oil treated in this manner still retained the ability to cure rickets. McCollum correctly reasoned that the activity in healing rickets was due to a new and heretofore unknown vitamin that he termed “vitamin D.” On the basis of the experiments of McCollum and of Mellanby, vitamin D became known as an essential nutrient.
II. DISCOVERY THAT VITAMIN D IS NOT A VITAMIN At the same time that Sir Edward Mellanby was carrying out the experiments in dogs, Huldshinsky [17] and Chick et al. [18] independently found that rickets in children could be prevented or cured by exposing them to sunlight or to artificially induced ultraviolet light. Thus, the curious findings were that sunlight and ultraviolet light somehow equaled cod liver oil. These strange and divergent results required resolution. Steenbock and Hart had noted the importance of sunlight in restoring positive calcium balance in goats [19]. At Wisconsin, with McCollum carrying out experiments in small experimental animals (i.e., rats), Steenbock was required to work with larger animals. Steenbock then began to study goats because they would consume less materials and could serve as better experimental animals than cows. Steenbock began to study the calcium balance of lactating goats and found that those goats maintained outdoors in the sunlight were found to be in positive calcium balance, whereas those maintained indoors lost a great deal of their skeletal calcium to lactation [19]. Steenbock and Hart, therefore, noted the importance of sunlight on calcium balance. This work then undoubtedly led Steenbock to realize that the ultraviolet healing properties described by Huldschinsky might be related to the calcium balance experiments in goats. By irradiating the animals and diets, Steenbock and Black found that vitamin D activity could be induced and rickets could be cured [20]. A similar finding was reported soon thereafter by Hess and Weinstock [21]. Steenbock then traced this to the nonsaponifiable fraction of the lipids in foods [22]. He found that ultraviolet light activated an inactive substance to become a vitamin D–active material. Thus, ultraviolet light could be used to irradiate foods, induce vitamin D activity, and fortify foods to eliminate rickets as a major
5
CHAPTER 1 Historical Perspective
medical problem. This discovery also made available a source of vitamin D for isolation and identification.
III. ISOLATION AND IDENTIFICATION OF NUTRITIONAL FORMS OF VITAMIN D From irradiation of mixtures of plant sterols, Windaus and colleagues isolated a material that was active in healing rickets [23]. This substance was called “vitamin D1,” but its structure was not determined. Vitamin D1 proved to be an adduct of tachysterol and vitamin D2, and thus vitamin D1 was actually an error in identification. The British group led by Askew was successful in isolating and determining the structure of the first vitamin D, vitamin D2 or ergocalciferol, from irradiation of plant sterols [24]. A similar identification by the Windaus group confirmed the structure of vitamin D2 [25]. Windaus and Bock also isolated the precursor of vitamin D3 from skin, namely, 7-dehydrocholesterol [26]. Furthermore, 7-dehydrocholesterol was synthesized [27] and converted to vitamin D3 (cholecalciferol) as identified by the Windaus group [28]. Thus, the structures of nutritional forms of vitamin D became known (Fig. 1). Windaus and Bock, having isolated 7-dehydrocholesterol from skin, provided the presumptive evidence that vitamin D3 is the form of vitamin D produced in skin, a discovery that was later confirmed by the chemical identification of vitamin D3 in skin by Esvelt et al. [29] and of a previtamin D3 in skin by Holick et al. [30]. Synthetic vitamin D as produced by the irradiation process replaced the irradiation of foods as a means of fortifying foods with vitamin D and was also rapidly applied to a variety of diseases including rickets and tetany and in the provision to domestic animals such as chickens, cows, and pigs. Windaus’ group provided chemical syntheses of the vitamin D compounds, confirming their structures and
thus ending the era of the isolation and identification of nutritional forms of vitamin D and making them available for the treatment of disease. For his contributions, Windaus received the 1938 Nobel Prize in chemistry.
IV. DISCOVERY OF THE PHYSIOLOGICAL FUNCTIONS OF VITAMIN D A. Intestinal Calcium and Phosphorus Absorption Besides bone mineralization, the earliest discovered function of vitamin D is its important role in the absorption and utilization of calcium. The first report of this finding was in the early 1920s by Orr and colleagues [31]. Kletzien et al. [32] demonstrated that vitamin D plays an important role in the utilization of calcium from the diet, and a number of experiments were carried out on the utilization of calcium and phosphorus from cereal diets. Nicolaysen was responsible, however, for demonstrating unequivocally the role of vitamin D in the absorption of calcium and independently of phosphorus from the diet [33]. Nicolaysen also followed the early work of Kletzien et al. [32] in which animals adapted to a low calcium diet were better able to utilize calcium than were animals on an adequate calcium diet. This work was confirmed by Nicolaysen, who postulated the existence of an “endogenous factor” that would inform the intestine of the skeletal needs for calcium [34]. This endogenous factor later proved to be largely the active form of vitamin D, 1,25-dihydroxyvitamin D3 [1,25(OH)2D3] [35]. Strong support for this concept was provided by the studies of Ribovich et al. [36] that showed animals maintained on a constant exogenous source of 1,25(OH)2D3 are unable to change intestinal calcium transport in response to changes in dietary calcium levels.
B. Mobilization of Calcium from Bone
FIGURE 1 Nutritional forms of vitamin D.
For many years, investigators have attempted to show that vitamin D plays a role directly on the mineralization process of the skeleton. However, early work by Rowland and Kramer [37], later work by Lamm and Neuman [38], and more recent work by Underwood and DeLuca [39] demonstrated very clearly that vitamin D does not play a significant role in the actual mineralization process of the skeleton but that the failure to mineralize the skeleton in vitamin D deficiency is due to inadequate levels of calcium and phosphorus in the plasma. Thus, the action of vitamin D in mineralizing
6 the skeleton and in preventing hypocalcemic tetany is the elevation of plasma calcium and phosphorus [40]. These discoveries laid to rest the concept of a role of vitamin D in mineralization. However, Carlsson [41] and Bauer et al. [42] were the first to realize that a major function of vitamin D is to induce the mobilization of calcium from bone when required. Thus, in animals on a low-calcium diet, the rise in serum calcium induced by vitamin D is the result of actual mobilization of calcium from bone [43]. This important function is known to be essential for the provision of calcium to meet soft-tissue needs, especially those of nerves and muscle, on a minute-to-minute basis when it is in insufficient supply from the diet. It is likely that the function of vitamin D in mobilizing calcium from bone is an osteoclastic-mediated process [44]. It is clear, however, that both vitamin D and parathyroid hormone are required for this function [45]. Furthermore, it is clear that vitamin D plays an important role in osteoclasticmediated bone resorption [46], which is certainly the first event in bone remodeling and an essential event in bone modeling [47].
C. Renal Reabsorption of Calcium and Phosphorus A final site of vitamin D action to elevate plasma calcium is in the distal renal tubule. Although experiments were suggestive of a role for vitamin D in increasing renal tubule reabsorption of calcium, a clear demonstration of this did not occur until the 1980s at the hands of Yamamoto et al. [48]. The renal tubule reabsorbs 99% of the filtered calcium even in the absence of vitamin D. However, reabsorption of the last 1% of the filtered load requires both vitamin D and parathyroid hormone. Thus, these agents work in concert in the renal reabsorption of calcium as well as in the mobilization of calcium from bone. Both agents are required to carry out this function.
D. Discovery of New Functions of Vitamin D With discovery of the receptor for the vitamin D hormone (described in Section V,G below) came the surprising result that this receptor could be found in a variety of tissues not previously appreciated as targets of vitamin D action. It localizes in the distal renal tubule cells, enterocytes of the small intestine, bone lining cells, and osteoblasts in keeping with its known role in calcium metabolism [49,50]. However, its appearance
HECTOR F. DELUCA
in tissues such as parathyroid gland, islet cells of the pancreas, cells in bone marrow (i.e., promyelocytes), lymphocytes, and certain neural cells raised the question of whether the functions of vitamin D might be broader than previously anticipated [49,50]. As a result of those findings, new functions of vitamin D have been found. For example, vitamin D plays a role in causing differentiation of promyelocytes to monocytes and the subsequent coalescing of the monocytes into multinuclear osteoclast precursors and ultimately into active osteoclasts [51,52]. Suppression of parathyroid cell growth and suppression of parathyroid hormone gene expression represent other new vitamin D actions [53,54]. In keratinocytes of skin, vitamin D appears to play a role in suppression of growth and in cellular differentiation [55]. Likely, discoveries of many new functions of 1,25(OH)2D3 will be made and are well on their way, as described in later chapters of this volume.
V. DISCOVERY OF THE HORMONAL FORM OF VITAMIN D A. Early Work of Kodicek The true pioneer of vitamin D metabolism was Egan MA Kodicek working at the Dunn Nutritional Laboratory in Cambridge U.K. Kodicek used a bioassay at first to study the fate of the vitamin D molecule and found that much vitamin D was converted to biologically inactive products [56]. Clearly, however, this approach of assaying vitamin D activity following administration of known doses of vitamin D was of limited value in determining metabolism.
B. Radiolabeled Vitamin D Experiments Professor Kodicek then began to synthesize radiolabeled vitamin D2. Unfortunately, the degree of labeling was not sufficient to permit the administration of truly physiological doses of vitamin D. Nevertheless, Professor Kodicek continued investigations into this important area. At the conclusion of 10 years of work, he concluded that vitamin D was active without metabolic modification and that the metabolites that were found were biologically inactive [57]. This conclusion was reached even as late as 1967, when it was concluded that vitamin D3 itself was the active form of vitamin D in the intestine [58]. However, chemical synthesis of vitamin D3 of high specific activity in the laboratory of the author proved to be of key importance in the demonstration of biologically active metabolites [59].
7
CHAPTER 1 Historical Perspective
By providing a truly physiological dose of vitamin D, it could be learned that the vitamin D itself disappeared and instead polar metabolites could be found in the target tissues before those tissues responded [60]. The polar metabolites proved to be more biologically active and acted more rapidly than vitamin D itself [61]. Thus, presumptive evidence of conversion of vitamin D to active forms had been obtained as early as 1967.
C. Isolation and Identification of the Active Form of Vitamin D By 1968, the first active metabolite of vitamin D was isolated in pure form and chemically identified as 25-hydroxyvitamin D3 (25OHD3) [62]. Its structure was confirmed by chemical synthesis [63] that provided it for study to the medical and scientific world. For a couple of years, 25OHD3, was visualized as the active form of vitamin D. However, when it was synthesized in radiolabeled form, it was found to be rapidly metabolized to yet more polar metabolites [64]. By this time, the Kodicek laboratory reawakened their interest in metabolism of vitamin D and began to study the metabolism of lα-tritium-labeled vitamin D [65]. Furthermore, polar metabolites of vitamin D were found by Haussler, Myrtle, and Norman [66]. The Wisconsin group labeled these metabolites as peak 5 [64], the Norman group called it peak 4B [66], and Lawson, Wilson, and Kodicek described it as peak P [65]. Kodicek et al. claimed that the metabolite of vitamin D found in intestine was deficient in tritium at the 1-position [65]. However, Myrtle et al. reported that peak 4B did not lose its tritium [67]. Thus, the suggestion of a modification at the 1-position could not be confirmed. The DeLuca group, however, isolated the active metabolite from intestines of 1600 chickens given radiolabeled vitamin D, and, by means of mass spectrometric techniques and specific chemical reactions, the structure of the active form of vitamin D in the intestine was unequivocally demonstrated as 1,25(OH)2D3 [68]. Of great importance was the finding of Fraser and Kodicek that the peak P metabolite could be produced by homogenates of chicken kidney and that anephric animals are unable to produce the peak P metabolite [69]. They correctly concluded that the site of synthesis of the active form of vitamin D is the kidney. The Wisconsin group then chemically synthesized both 1α25(OH)2D3 [70] and 1β,25(OH)2D3 [71] and provided unequivocal proof that the active form is 1α,25(OH)2D3. Furthermore, this group was able to synthesize lαOHD3, an important
analog that assumed great importance as a therapeutic agent throughout the world [72].
D. Proof That 1,25(OH)2D3 Is the Active Form of Vitamin D Proof that 1,25(OH)2D3 and not 25OHD3 is the active form was provided by experiments in which anephric animals respond to 1,25(OH)2D3 by increasing intestinal absorption of calcium and bone calcium mobilization, whereas animals receiving 25OHD3 at physiological doses did not [73–75]. Furthermore, the experiment of nature, namely, vitamin D–dependency rickets type I, an autosomal recessive disorder, provided final proof [76]. This disease could be corrected by physiological doses of synthetic 1,25(OH)2D3, whereas large amounts of vitamin D3 or 25OHD3 were needed to heal the rickets. The exact defect in this disease is now clearly known and is described elsewhere in this volume. 25OHD3 at pharmacological doses likely acts as an analog of the final vitamin D hormone, 1,25(OH)2D3 (Fig. 2).
E. Discovery of the Vitamin D Endocrine System Immediately after the identification of 1,25(OH)2D3 as the active form of vitamin D came studies in which it could be shown that animals on a low-calcium diet produce large quantities of 1,25(OH)2D3, whereas those on a high-calcium diet produce little or no 1,25(OH)2D3 [77]. A reciprocal arrangement was found for the metabolite 24R,25(OH)2D3. Thus, when calcium is needed, production of 1,25(OH)2D3 is markedly stimulated and the 24-hydroxylation degradation reaction is suppressed. When adequate calcium is present, production of 1,25(OH)2D3 is shut off and the 24-hydroxylation reaction is turned on. This discovery also satisfactorily provided evidence that 1,25(OH)2D3 is the likely endogenous factor originally described by Nicolaysen et al. [34]. The next important step was the demonstration that it is parathyroid hormone that activates 1α-hydroxylation of 25OHD3 in the kidney [78]. Thus, parathyroidectomy eliminates the hypocalcemic stimulation of 1α-hydroxylation and suppression of 24-hydroxylation, whereas administration of parathyroid hormone restores that capability. Fraser and Kodicek also provided evidence that, in intact chickens, injection of parathyroid hormone stimulated the 1α-hydroxylation reaction [79]. Thus, the basic vitamin D endocrine system was largely discovered and reported in the early 1970s, being completed by 1974.
8
HECTOR F. DELUCA
OH
Liver
Kidney
Microsomes (Mitochondria)
Mitochondria
HO
HO Vitamin D3
OH
HO 25-hydroxyvitamin D3
OH 1α,25-dihydroxyvitamin D3
FIGURE 2 Activation of the vitamin D molecule.
F. Other Metabolites of Vitamin D During the course of identification of 1,25(OH)2D3, 21,25(OH)2D3 was reported as a metabolite, as was 25,26(OH)2D3 [80,81]. However, the identification of 21,25(OH)2D3 was in error and was corrected to 24,25(OH)2D3, with the correct stereochemistry as 24R,25(OH)2D3 [82]. Over the late 1970s and early 1980s, as many as 30 metabolites of vitamin D were identified [83]. These are covered in other chapters in this volume. Of great importance was the use of the fluoro derivatives of vitamin D to illustrate that the only activation pathway of vitamin D is 25-hydroxylation followed by 1-hydroxylation [84]. Thus, 24-difluoro25OHD3 supported all known functions of vitamin D for at least two generations of animals [85]. 24-Difluoro25OHD3 cannot be 24-hydroxylated. Furthermore, other fluoro derivatives such as 26,27-hexafluoro-25OHD3 [86] and 23-difluoro-25OHD3 [87] are all fully biologically active, illustrating that 26-hydroxylation, 24-hydroxylation, and 23-hydroxylation are not essential to the function of vitamin D.
G. Discovery of the Vitamin D Receptor Zull and colleagues provided evidence that the function of vitamin D is blocked by transcription and protein inhibitors [88]. Thus, it became clear very early that a nuclear activity is required for vitamin D to carry out its functions. This work confirmed and extended the earlier work of Eisenstein and Passavoy [89]. With the discovery of the active forms of vitamin D came new attention to the idea that vitamin D may work through a nuclear mechanism. Thus, Haussler et al. reported vitamin D compounds to be associated with chromatin [66]. However, these experiments did not
exclude the possibility that the vitamin D compounds might be bound to the nuclear membrane. The first clear demonstration of the existence of a vitamin D receptor was at the hands of Brumbaugh and Haussler [90]. Furthermore, the experiments of Kream et al. [91] provided strong and unequivocal evidence of the existence of a nuclear receptor for 1,25(OH)2D3. Intense efforts toward purification of the receptor and its study appeared with the knowledge that it is a receptor protein with a molecular weight of approximately 55,000. In 1987, a partial cDNA sequence for the chicken vitamin D receptor was determined [92]. This was followed by isolation of the full coding sequence for the human [93] and rat [94,95] receptors. Cloning of the cDNAs encoding the vitamin D receptor in a human and mouse permitted the isolation of the gene encoding the vitamin D receptor [96–98]. The human gene was completely described and the mouse promoter was isolated and shown to be a TATA-less Sp1driven promoter [98]. The human gene appears to have alternate promoters [96,97]. Two groups have prepared receptor Null mutant mice, permitting extensive experiments with vitamin D receptorless animals [99,100]. From a historical point of view, one of the most important discoveries was vitamin D-dependency rickets type II [101], which is now known to be due to a defect in the receptor gene [102,103] (discussed in Chapter 11). This discovery essentially provided receptor knockout experiments in humans, allowing unequivocal proof of the essentiality of the vitamin D receptor for the functions of vitamin D. The nature of the receptor and how it functions are described in subsequent chapters along with current thinking on the molecular mechanism of action of 1,25(OH)2D3. However, the discovery of vitamin D responsive elements in the osteocalcin, osteopontin, preproparathyroid, and 24-OHase genes represent important
CHAPTER 1 Historical Perspective
historical developments [104–107]. This led to a consensus sequence and, most important, the development of the 3, 4, 5-rule of Umesono et al. [108]. It is now clear that vitamin D–responsive elements represent two imperfect repeat sequences separated by three nonspecified nucleotides. The vitamin D receptor will bind to these response elements, but it requires the presence of another nuclear factor, which proved to be the retinoid-X receptor (RXR) [109,110]. It is quite clear that the vitamin D receptor forms a heterodimer on the vitamin D responsive elements with the RXR protein on the 5′ arm of the responsive element and the vitamin D receptor on the 3′ segment [111]. The work of Rosenfeld, Glass, and colleagues [112] has demonstrated that the RXR protein when complexed with the vitamin D receptor on the responsive elements will not accept an RXR ligand, thus acting as a silent partner. Details of what is known concerning the role of the vitamin D receptor in transcription are described fully in subsequent chapters.
Acknowledgment This work was supported in part by a fund from the Wisconsin Alumni Research Foundation.
References 1. von Liebig J 1957 Animal Chemistry or Organic Chemistry in Its Application to Physiology and Pathology. In: Glass HB (ed) A History of Nutrition. The Riverside Press, Cambridge, MA. 2. Lunin N 1881 Über die Bedeutung der anorganischen Salze für die Ernährung des Tieres. Z Physiol Chem 5:31–39. 3. Hopkins G 1912 Feeding experiments illustrating the importance of accessory food factors in normal dietaries. J Physiol 44:425–460. 4. Magendie F 1816 Ann Chim Phys. 3:86–87. In: Glass HB (ed) 1957 A History of Nutrition by McCollum EV. The Riverside Press, Cambridge MA. 5. Funk C 1911 The chemical nature of the substance that cures polyneurities in birds produced by a diet of polished rice. J Physiol (London) 43:395–402. 6. Eijkman C 1897 Ein beri-beri Anliche der Hühner. Virchows Arch 148:523–527. 7. Grijns G 1957 Concerning polyneuritis geneeskundig gallinarum tijschrift voor ned indie 1901. In: McCollum EV, Glass HB (eds) A History of Nutrition. Houghton Mifflin, Boston, p. 216. 8. Hoist A. Frolich T 1907 Experimental studies relating ship-beriberi to scurvy. II. On the etiology of scurvy. J Hyg 7: 634–671. 9. Hart EB, McCollum EV, Steenbock H, Humphrey GC 1911 Physiological effect on growth and reproduction of rations balanced from restricted sources. Wis Agric Exp Sta Res Bull 17:131–205. 10. McCollum EV, Davis M 1913 The necessity of certain lipins in the diet during growth. J Biol Chem 25:167–175.
9 11. McCollum EV, Simmons N, Pitz W 1916 The relation of the unidentified dietary factors, the fat-soluble A and watersoluble B of the diet to the growth promoting properties of milk. J Biol Chem 27:33–38. 12. Osborne TB, Mendel LB 1917 The role of vitamins in the diet. J Biol Chem 31:149–163. 13. Whistler D 1645 De morbo puerli anglorum, quem patrio ideiomate indigenae vocant “the rickets” (Lugduni, Batavorum 1645). As cited by Smerdon GT,1950 Daniel Whistler and the English Disease. A translation and biographical note. J Hist Med 5:397–415. 14. Hess A 1929 The history of rickets. In: Rickets, Including Osteomalacia and Tetany. Lea & Febiger, Philadelphia, pp. 22–37. 15. Mellanby E 1919 An experimental investigation on rickets. Lancet 1:407–412. 16. McCollum EV, Simmonds N, Becker JE, Shipley PG 1922 An experimental demonstration of the existence of a vitamin which promotes calcium deposition. J Biol Chem 53:293–298. 17. Huldshinsky K 1919 Heilung von Rachitis durch Künstlich Hohen-sonne. Deut Med Wochenschr 45:712–713. 18. Chick H, Palzell EJ, Hume EM 1923 Studies of rickets in Vienna 1919–1922. Medical Research Council, Special Report No. 77. 19. Steenbock H, Hart EB 1913 The influence of function on the lime requirements of animals. J Biol Chem 14:59–73. 20. Steenbock H, Black A 1924 Fat-soluble vitamins. XVII. The induction of growth-promoting and calcifying properties in a ration by exposure to ultraviolet light. J Biol Chem 61:405–422. 21. Hess AF, Weinstock M 1924 Antirachitic properties imparted to lettuce and to growing wheat by ultraviolet irradiation. Proc Soc Exp Biol Med 22:5–6. 22. Steenbock H, Black A 1925 Fat-soluble vitamins. XXIII. The induction of growth-promoting and calcifying properties in fats and their unsaponifiable constituents by exposure to light. J Biol Chem 64:263–298. 23. Windaus A, Linsert O 1928 Vitamin D1 . Ann Chem 465:148. 24. Askew FA, Bourdillon RB, Bruce HM, Jenkins RGC, Webster TA 1931 The distillation of vitamin D. Proc R Soc 8107:76–90. 25. Windaus A, Linsert O, Luttringhaus A, Weidlich G 1932 Crystalline-vitamin D2. Ann Chem 492:226–241. 26. Windaus A, Bock F 1937 Über das Provitamin aus dem Sterin der Schweineschwarte. Z Physiol Chem 245:168–170. 27. Windaus A, Lettre H, Schenck F 1935 7-Dehydrocholesterol. Ann Chem 520:98–107. 28. Windaus A, Schenck F, von Werder F 1936 Über das antirachitisch wirksame Bestrahlungs-produkt aus 7-Dehydrocholesterin. Hoppe-Seylers Z Physiol Chem 241:100–103. 29. Esvelt RP, Schnoes HK, DeLuca HF 1978 Vitamin D3 from rat skins irradiated in vitro with ultraviolet light. Arch Biochem Biophys 188:282–286. 30. Holick MF, MacLaughlin JA, Clark MB, Holick SA, Potts JT Jr, Anderson RR, Blank IH, Parrish JA, Elias P 1980 Photosynthesis of previtamin D3 in human skin and the physiologic consequences. Science 210:203–205. 31. Orr WJ, Holt LE Jr, Wilkins L, Boone FH 1923 The calcium and phosphorus metabolism in rickets, with special reference to ultraviolet rat therapy. Am J Dis Children 26:362–372. 32. Kletzien SWF, Templin VM, Steenbock H, Thomas BH 1932 Vitamin D and the conservation of calcium in the adult. J Biol Chem 97:265–280. 33. Nicolaysen R 1937 Studies upon the mode of action of vitamin D. III. The influence of vitamin D on the absorption of calcium and phosphorus in the rat. Biochem J 31:122–129. 34. Nicolaysen R, Eeg-Larsen N, Malm OJ 1953 Physiology of calcium metabolism. Physiol Rev 33:424–444.
10 35. Boyle IT, Gray RW, Omdahl JL, DeLuca HF 1972 Calcium control of the in vivo biosynthesis of 1,25-dihydroxyvitamin D3: Nicolaysen’s endogenous factor. In: Taylor S (ed) Endocrinology 1971. Heinemann Medical Books, London, pp. 468–476. 36. Ribovich ML, DeLuca HF 1975 The influence of dietary calcium and phosphorus on intestinal calcium transport in rats given vitamin D metabolites. Arch Biochem Biophys 170:529–535. 37. Rowland J, Kramer B 1921 Calcium and phosphorus in the serum in relation to rickets. Am J Dis Children 22:105–119. 38. Lamm M, Neuman WF 1958 On the role of vitamin D in calcification. Arch Pathol 66:204–209. 39. Underwood JL, DeLuca HF 1984 Vitamin D is not directly necessary for bone growth and mineralization. Am J Physiol 246:E493–E498. 40. DeLuca HF 1967 Mechanism of action and metabolic fate of vitamin D. Vit Horm 25:315–367. 41. Carlsson A 1952 Tracer experiments on the effect of vitamin D on the skeletal metabolism of calcium and phosphorus. Acta Physiol Scand 26:212–220. 42. Bauer GCH, Carlsson A, Lindquist B 1955 Evaluation of accretion, resorption, and exchange reactions in the skeleton. Kungl Fysiograf Sallskapets I. Lund Forhandlingar 25:3–18. 43. Blunt JW, Tanaka Y, DeLuca HF 1968 The biological activity of 25-hydroxycholecalciferol, a metabolite of vitamin D3. Proc Natl Acad Sci USA 61:1503–1506. 44. Morony S, Capparelli C, Lee R, Shimamoto G, Boone T, Lacey DL, Dunstan CR 1999 A chimeric form of osteoprotegerin inhibits hypercalcemia and bone resorption induced by IL-1β, TNF-α, PTH, PTHrP, and 1,25(OH)2D3. J Bone Miner Res 14:1478–1485. 45. Garabedian M, Tanaka Y, Holick MF, DeLuca HF 1974 Response of intestinal calcium transport and bone calcium mobilization to 1,25-dihydroxyvitamin D3 in thyroparathyroidectomized rats. Endocrinology 94:1022–1027. 46. Raisz LG, Trummel CL, Holick MF, DeLuca HF 1972 1,25-Dihydroxycholecalciferol: A potent stimulator of bone resorption in tissue culture. Science 175:768–769. 47. Frost HM 1966 Bone Dynamics in Osteoporosis and Osteomalacia. Henry Ford Hospital Surgical Monograph Series. Thomas, Springfield, IL. 48. Yamamoto M, Kawanobe Y, Takahashi H, Shimazawa E, Kimura S, Ogata E 1984 Vitamin D deficiency and renal calcium transport in the rat. J Clin Invest 74:507–513. 49. Stumpf WE, Sar M, Reid FA, Tanaka Y, DeLuca HF 1979 Target cells for 1,25-dihydroxyvitamin D3 in intestinal tract, stomach, kidney, skin, pituitary and parathyroid. Science 206:1188–1190. 50. Stumpf WE, Sar M, DeLuca HF 1981 Sites of action of 1,25 (OH)2 vitamin D3 identified by thaw-mount autoradiography. In: Cohn DV, Talmage RV, Matthews JL (eds) Hormonal Control of Calcium Metabolism. Excerpta Medica, Amsterdam, pp. 222–229. 51. Suda T, Takahashi N, Martin TJ 1992 Modulation of osteoclast differentiation. Endocr Rev 13:66–80. 52. Suda T 1992 The role of 1α,25-dihydroxyvitamin D3 in the myeloid cell differentiation. Proc Soc Exp Biol Med 191:214–220. 53. Demay MB, Kiernan, MS, DeLuca HF, Kronenberg HM 1992 Sequences in the human parathyroid hormone gene that bind the 1,25-dihydroxyvitamin D3 receptor and mediate transcriptional repression in response to 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 89:8097–8101. 54. Silver J 1994 Regulation of parathyroid hormone production by 1α,25-(OH)2D3 and its analogues. Their therapeutic usefulness in secondary hyperparathyroidism. In: Vitamin D and
HECTOR F. DELUCA
55.
56.
57. 58. 59. 60. 61. 62. 63. 64. 65.
66. 67. 68.
69. 70. 71. 72. 73.
Its Analogues. The Second International Forum on Calcified Tissue and Bone Metabolism. Chugai Pharmaceutical, Tokyo, pp. 60–63. Smith EL, Walworth NC, Holick MF 1986 Effect of 1α,25-dihydroxyvitamin D3 on the morphologic and biochemical differentiation of cultured human epidermal keratinocytes grown in serum-free conditions. J Invest Dermatol 86:709–714. Kodicek E 1956 Metabolic studies on vitamin D. In: Wolstenholme GWE, O’Connor CM (eds) Ciba Foundation Symposium on Bone Structure and Metabolism. Little, Brown, and Co., Boston, pp. 161–174. Kodicek E 1960 The metabolism of vitamin D. In: Umbreit W, Molitor H (eds) Proceedings of the Fourth International Congress of Biochemistry. Pergamon, London, Vol. 11, pp. 198–208. Haussler MR, Norman AW 1967 The subcellular distribution of physiological doses of vitamin D3. Arch Biochem Biophys 118:145–153. Neville PF, DeLuca HF 1966 The synthesis of [1,2-3H]vitamin D3 and the tissue localization of a 0.25 µg (10 IU) dose per rat. Biochemistry 5:2201–2207. Lund J, DeLuca HF 1966 Biologically active metabolite of vitamin D3 from bone, liver, and blood serum. J Lipid Res 7:739–744. Morii H, Lund J, Neville PF, DeLuca HF 1967 Biological activity of a vitamin D metabolite. Arch Biochem Biophys 120:508–512. Blunt JW, DeLuca HF, Schnoes HK 1968 25-Hydroxycholecalciferol. A biologically active metabolite of vitamin D3. Biochemistry 7:3317–3322. Blunt JW, DeLuca HF 1969 The synthesis of 25-hydroxycholecalciferol. A biologically active metabolite of vitamin D3. Biochemistry 8:671–675. DeLuca HF 1970 Metabolism and function of vitamin D. In: DeLuca HF, Suttie JW (eds) The Fat-Soluble Vitamins. Univ. of Wisconsin Press, Madison, pp. 3–20. Lawson DEM, Wilson PW, Kodicek E 1969 Metabolism of vitamin D. A new cholecalciferol metabolite, involving loss of hydrogen at C-l, in chick intestinal nuclei. Biochem J 115:269–277. Haussler MR, Myrtle JF, Norman AW 1968 The association of a metabolite of vitamin D3 with intestinal mucosa chromatin in vivo. J Biol Chem 243:4055–4064. Myrtle JF, Haussler MR, Norman AW 1970 Evidence for the biologically active form of cholecalciferol in the intestine. J Biol Chem 245:1190–1196. Holick MF, Schnoes HK, DeLuca HF, Suda T, Cousins RJ 1971 Isolation and identification of 1,25–dihydroxycholecalciferol. A metabolite of vitamin D active in intestine. Biochemistry 10:2799–2804. Fraser DR, Kodicek E 1970 Unique biosynthesis by kidney of a biologically active vitamin D metabolite. Nature 228: 764–766. Semmler EJ, Holick MF, Schnoes HK, DeLuca HF 1972 The synthesis of 1α,25-dihydroxycholecalciferol—A metabolically active form of vitamin D3. Tetrahedron Lett 40:4147–4150. Paaren HE, Schnoes HK, DeLuca HF 1977 Synthesis of 1α,25-dihydroxyvitamin D3. J Chem Soc Chem Commun, 890–892. Holick MF, Semmler EJ, Schnoes HK, DeLuca HF 1973 1αHydroxy derivative of vitamin D3: A highly potent analog of 1α,25-dihydroxyvitamin D3. Science 180:190–191. Boyle IT, Miravet L, Gray RW, Holick MF, DeLuca HF 1972 The response of intestinal calcium transport to 25-hydroxy and 1,25-dihydroxyvitamin D in nephrectomized rats. Endocrinology 90:605–608.
11
CHAPTER 1 Historical Perspective
74. Holick MF, Garabedian M, DeLuca HF 1972 1,25Dihydroxycholecalciferol: Metabolite of vitamin D3 active on bone in anephric rats. Science 176:1146–1147. 75. Wong RG, Norman AW, Reddy CR, Coburn JW 1972 Biologic effects of 1,25-dihydroxycholecalciferol (a highly active vitamin D metabolite) in acutely uremic rats. J Clin Invest 51:1287–1291. 76. Fraser D, Kooh SW, Kind HP, Holick MF, Tanaka Y, DeLuca HF 1973 Pathogenesis of hereditary vitamin D dependent rickets: An inborn error of vitamin D metabolism involving defective conversion of 25-hydroxyvitamin D to 1α,25dihydroxyvitamin D. N Engl J Med 289:817–822. 77. Boyle IT, Gray RW, DeLuca HF 1971 Regulation by calcium of in vivo synthesis of 1,25-dihydroxycholecalciferol and 21,25-dihydroxycholecalciferol. Proc Natl Acad Sci USA 68:2131–2134. 78. Garabedian M, Holick MF, DeLuca HF, Boyle IT 1972 Control of 25-hydroxycholecalciferol metabolism by the parathyroid glands. Proc Natl Acad Sci USA 69:1673–1676. 79. Fraser DR, Kodicek E 1973 Regulation of 25-hydroxycholecalciferol-1-hydroxylase activity in kidney by parathyroid hormone. Nature (New Biol) 241:163–166. 80. Suda T, DeLuca HF, Schnoes HK, Ponchon G, Tanaka Y, Holick MF 1970 21,25-Dihydroxycholecalciferol. A metabolite of vitamin D3 preferentially active on bone. Biochemistry 9:2917–2922. 81. Suda T, DeLuca HF, Schnoes HK, Tanaka Y, Holick MF 1970 25,26-Dihydroxycholecalciferol, a metabolite of vitamin D3 with intestinal calcium transport activity. Biochemistry 9:4776–4780. 82. Tanaka Y, Frank H, DeLuca HF, Koizumi N, Ikekawa N 1975 Importance of the stereochemical position of the 24-hydroxyl to biological activity of 24-hydroxyvitamin D3. Biochemistry 14:3293–3296. 83. DeLuca HF, Schnoes HK 1983 Vitamin D: Recent advances. Annu Rev Biochem 52:411–439. 84. Brommage R, DeLuca HF 1985 Evidence that 1,25dihydroxyvitamin D3 is the physiologically active metabolite of vitamin D3. Endocr Rev 6:491–511. 85. Brommage R, Jarnagin K, DeLuca HF, Yamada S, Takayama H 1983 1- but not 24-hydroxylation of vitamin D is required for skeletal mineralization in rats. Am J Physiol 244: E298–E304. 86. Tanaka Y, Pahuja DN, Wichmann JK, DeLuca HF, Kobayashi Y, Taguchi T, Ikekawa N 1982 25-Hydroxy-26,26,26,27,27,27hexafluorovitamin D3: Biological activity in the rat. Arch Biochem Biophys 218:134–141. 87. Nakada M, Tanaka Y, DeLuca HF, Kobayashi Y, Ikekawa N 1985 Biological activities and binding properties of 23,23difluoro-25-hydroxyvitamin D3 and its lα-hydroxy derivative. Arch Biochem Biophys 241:173–178. 88. Zull JE, Czarnowska-Misztal E, DeLuca HF 1965 Actinomycin D inhibition of vitamin D action. Science 149:182–184. 89. Eisenstein R, Passavoy M 1964 Actinomycin D inhibits parathyroid hormone and vitamin D activity. Proc Soc Exp Biol Med 117:77–79. 90. Brumbaugh PF, Haussler MR 1973 lα,25-Dihydroxyvitamin D3 receptor: Competitive binding of vitamin D analogs. Life Sci 13:1737–1746. 91. Kream BE, Reynolds RD, Knutson JC, Eisman JA, DeLuca HF 1976 Intestinal cytosol binders of 1,25-dihydroxyvitamin D3 and 25-hydroxyvitamin D3. Arch Biochem Biophys 176: 779–787. 92. McDonnell DP, Mangelsdorf DJ, Pike JW, Haussler MR, O’Malley BW 1987 Molecular cloning of complementary
93.
94. 95. 96.
97. 98. 99.
100.
101.
102.
103.
104.
105.
106.
107.
DNA encoding the avian receptor for vitamin D. Science 235:1214–1217. Baker AR, McDonnell DP, Hughes M, Crisp TM, Mangelsdorf DJ, Haussler MR, Pike JW, Shine J, O’Malley BW 1988 Cloning and expression of full-length cDNA encoding human vitamin D receptor. Proc Natl Acad Sci USA 85: 3294–3298. Burmester JK, Maeda N, DeLuca HF 1988 Isolation and expression of rat 1,25-dihydroxyvitamin D3 receptor cDNA. Proc Natl Acad Sci USA 85:1005–1009. Burmester JK, Wiese RJ, Maeda N, DeLuca HF 1988 Structure and regulation of the rat 1,25-dihydroxyvitamin D3 receptor. Proc Natl Acad Sci USA 85:9499–9502. Miyamoto K, Kesterson RA, Yamamoto H, Nishiwaki E, Tatsumi S, Taketani Y, Morita K, Pike JW, Takeda E 1997 Structural organization of the human vitamin D receptor chromosomal gene and its promoter. Mol Endocrinol 11:1165–1179. Jehan F, DeLuca HF 1997 Cloning and characterization of the mouse vitamin D receptor promoter. Proc Natl Acad Sci USA 94:10138–10143. Jehan F, DeLuca HF 2000 The mouse vitamin D receptor is mainly expressed through an Sp1-driven promoter in vivo. Arch Biochem Biophys 377:273–283. Yoshizawa T, Handa Y, Uematsu Y, Takeda S, Sekine K, Yoshihara Y, Kawakami T, Arioka K, Sato H, Uchiyama Y, Masushige S, Fukamizu A, Matsumoto T, Kato S 1997 Mice lacking the vitamin D receptor exhibit impaired bone formation, uterine hypoplasia and growth retardation after weaning. Nat Genet 16:391–396. Li YC, Pirro AE, Amling M, Delling G, Baron R, Bronson R, Demay MB 1997 Targeted ablation of the vitamin D receptor: An animal model of vitamin D-dependent rickets type II with alopecia. Proc Natl Acad Sci USA 94:9831–9835. Brooks MH, Bell NH, Love L, Stern PH, Orfei E, Queener SF, Hamstra AJ, DeLuca HF 1978 Vitamin D-dependent rickets Type II. Resistance of target organs to 1,25-dihydroxyvitamin D. N Engl J Med 298:996–999. Hughes MR, Malloy PJ, Kieback DG, Kesterson RA, Pike JW, Feldman D, O’Malley BW 1988 Point mutations in the human vitamin D receptor gene associated with hypocalcemic rickets. Science 242:1702–1706. Wiese RJ, Goto H, Prahl JM, Marx SJ, Thomas M, Al-Aqeel A, DeLuca HF 1993 Vitamin D–dependency rickets Type II: Truncated vitamin D receptor in three kindreds. Mol Cell Endocrinol 90:197–201. Demay MB, Gerardi JM, DeLuca HF, Kronenberg HM 1990 DNA sequences in the rat osteocalcin gene that bind the 1,25-dihydroxyvitamin D3 receptor and confer responsiveness to 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 87:369–373. Noda M, Vogel RL, Craig AM, Prahl J, DeLuca HF, Denhardt DT 1990 Identification of a DNA sequence responsible for binding of the 1,25-dihydroxyvitamin D3 receptor and 1,25-dihydroxyvitamin D3 enhancement of mouse secreted phosphoprotein 1 (Spp1, osteopontin) gene expression. Proc Natl Acad Sci USA 87:9995–9999. Demay MB, Kiernan MS, DeLuca HF, Kronenberg HM 1992 Sequences in the human parathyroid hormone gene that bind the 1,25-dihydroxyvitamin D3 receptor and mediate transcriptional repression in response to 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 89:8097–8101. Zierold C, Darwish HM, DeLuca HF 1995 Two vitamin D response elements function in the rat 1,25-dihydroxyvitamin D 24-hydroxylase promoter. J Biol Chem 270:1675–1678.
12 108. Umesono K, Murakami KK, Thompson CC, Evans RM 1991 Direct repeats as selective response elements for the thyroid hormone, retinoic acid, and vitamin D3 receptors. Cell 65:1255–1266. 109. Yu VC, Delsert C, Andersen B, Holloway JM, Devary OV, Naar AM, Kim SY, Boutin J-M, Glass CK, Rosenfeld MG 1991 RXRβ: A coregulator that enhances binding of retinoic acid, thyroid hormone, and vitamin D receptors to their cognate response elements. Cell 67:1251–1266. 110. Munder M, Herzberg IM, Zierold C, Moss VE, Hanson K, Clagett-Dame M, DeLuca HF 1995 Identification of the porcine intestinal accessory factor that enables DNA
HECTOR F. DELUCA
sequence recognition by vitamin D receptor. Proc Natl Acad Sci USA 93:2796–2799. 111. Jin CH, Pike JW 1996 Human vitamin D receptor-dependent transactivation in Saccharomyces cerevisiae requires retinoid X receptor. Mol Endocrinol 10:196–205. 112. DiRenzo J, Soderstrom M, Kurokawa R, Ogliastro M-H, Ricote M, Ingrey S, Horlein A, Rosenfeld MG, Glass CK 1997 Peroxisome proliferator-activated receptors and retinoic acid receptors differentially control the interactions of retinoid X receptor heterodimers with ligands, coactivators, and corepressors. Mol Cell Biol 17: 2166–2176.
CHAPTER 2
Vitamin D Metabolism RONALD L. HORST AND TIMOTHY A. REINHARDT U.S. Department of Agriculture, Agricultural Research Service, National Animal Disease Center, Ames, Iowa
G. SATYANARAYANA REDDY Brown University, Department of Chemistry, Providence, Rhode Island
I. Introduction II. Vitamin D Metabolism III. Vitamin D Toxicity
IV. Species Variation in Vitamin D Metabolism and Action V. Conclusions References
I. INTRODUCTION
II. VITAMIN D METABOLISM
In 1919, when the field of experimental nutrition was still in its infancy, Sir Edward Mellanby conducted a classic experiment that for the first time associated the supplementation of various growth-promoting fats with the prevention of rickets [1]. He credited the cure to the presence of a fat-soluble substance called vitamin A. McCollum et al. [2], however, later discovered that the factor responsible for healing rickets was distinct from vitamin A. McCollum named this new substance vitamin D. It was also during this period when scientists realized that there were two antirachitic factors with distinct structures. As discussed by Norman [3], the first factor to be identified was designated vitamin D2 (also known as ergocalciferol), whereas the structure of vitamin D3 (cholecalciferol) became evident some 4 to 5 years later. Vitamins D3 and D2 are used for supplementation of animal and human diets in the United States. Vitamin D3 is the form of vitamin D that is synthesized by vertebrates, whereas vitamin D2 is the major naturally occurring form of the vitamin in plants. Animals that bask in the sun such as amphibia, reptiles, and birds therefore synthesize sufficient endogenous vitamin D3 to meet their daily needs. However, herbivores may have evolved utilizing vitamin D2 as their predominant source. This chapter focuses on the general control and function of key enzymes involved in the regulation of vitamin D2 and vitamin D3 metabolism. Species differences in vitamin D metabolism, as well as vitamin D toxicity, are also discussed. The reader is also directed toward a number of additional reviews regarding vitamin D metabolism and action [3–8]. This chapter gives an overview of vitamin D metabolism; critical steps are discussed in further detail in the subsequent chapters of this section. Metabolism of vitamin D analogs is covered in Chapter 81.
A. Overview
VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX,
Vitamin D refers to a group of compounds that possess antirachitic activity. Technically vitamin D is classified as a secosteroid. Secosteroids are those in which one of the rings has been broken; in vitamin D, the 9,10 carbon–carbon bond of ring B is broken, and it is indicated by the inclusion of “9,10-seco” in the official nomenclature. There are several known nutritional forms of vitamin D, however, the best known examples are cholecalciferol (vitamin D3), which is produced in the skin, and ergocalciferol (vitamin D2), which is derived from plant tissues (Fig. 1). Therefore, when reference is made to vitamin D, the lack of a subscript usually implies either vitamin D2 or vitamin D3. The vitamin Ds are named according to the rules of the International Union of Pure and Applied Chemists (IUPAC) for steroid nomenclature [9]. Because vitamin D is derived from a steroid, the structure retains its numbering from the parent compound cholesterol. Configurations at asymmetric centers are designated by using the R and S notation applying the sequence-rule procedure [10]. Configuration of the double bonds are notated E for “entgegen” or trans, and Z for “zuzammen” or cis [11]. Thus the official name of vitamin D3, by relation to cholesterol, is 9,10-seco(5Z,7E)5,7,10(19) cholestatriene-3β-ol, and the official name of vitamin D2 is 9,10-seco(5Z,7E)-5,7,10(19), 22-ergostatetraene-3β-ol. Contemporary views categorize vitamin D3 not as a vitamin but, rather, as a prosteroid hormone. This concept is supported by the fact that in mammals vitamin D3 is derived from 7-dehydrocholesterol (the precursor of cholesterol) present in the skin. The direct action of sunlight on 7-dehydrocholesterol results in cleavage Copyright © 2005, Elsevier, Inc. All rights reserved.
16
RONALD L. HORST, TIMOTHY A. REINHARDT, AND G. SATYANARAYANA REDDY
FIGURE 1
Important nutritional forms of vitamin D.
of the B ring of the steroid structure that on thermoisomerization yields vitamin D3 (see Chapter 3). The significance of vitamin D as a prosteroid hormone became clearer in 1967 when Morii et al. [11a] isolated a new metabolite of vitamin D3 from rats that was as effective as vitamin D3 in healing rickets, raising blood calcium, and increasing intestinal calcium transport. This compound acted more rapidly than vitamin D3, requiring only 8 to 10 hr after oral administration to initiate its response. This metabolite was identified as 25-hydroxyvitamin D3 (25OHD3) [11b]. The liver was demonstrated to be important in the production of this most abundant circulating form of vitamin D3 that, under normal conditions, is present at 20 to 50 ng/ml [4]. Shortly following the discovery of 25OHD3, a number of laboratories showed that this metabolite is specifically hydroxylated at the lα-position in the kidney to yield 1,25-dihydroxyvitamin D3 [1,25(OH)2D3] [11c–13]. The latter metabolite is now generally accepted as the hormonally active form of vitamin D3. Its importance is evidenced by the genetic disorder pseudo vitamin D–deficiency rickets (PDDR), which is caused by mutations in the 1α-hydroxylase gene. PDDR results in the inability to produce 1,25(OH)2D leading to severe rickets (see Chapters 71 and 72). In normal human plasma, 1,25(OH)2D3 circulates at approximately 1000-fold lower concentrations than 25OHD3 and is generally present at 20 to 65 pg/ml [14]. This simplistic picture outlined for vitamin D3 activation is complicated by the fact that vitamin D3 can be oxidatively metabolized to a variety of products. Most of these numerous metabolites have no identifiable
biological function, and indeed many have been isolated from animals fed abnormally high amounts of vitamin D3. Nevertheless, the evidence collected to date indicates that 25-hydroxylated vitamin D3 metabolites are preferentially metabolized at the side chain. In particular, carbon centers C-23, C-24, and C-26 are readily susceptible to further oxidation. Figure 2 illustrates products of these oxidative pathways. As indicated, these pathways are shared by both 25OHD3 and 1,25(OH)2D3, and their physiological importance is still a matter of controversy. For example, there is evidence that 24,25-dihydroxyvitamin D3 [24,25(OH)2D3] may function to stimulate bone mineralization [15,16], suppress parathyroid hormone (PTH) secretion [17], and maintain embryonic development [18]. For the most part, however, the C-24 hydroxylation and other side-chain modifications are generally considered to be catabolic in nature and play a key role in maintaining vitamin D homeostasis [19]. Although these side-chain oxidative pathways yield metabolites that are considered “nonfunctional,” the presence of these compounds in circulation could pose serious problems in the analysis for 25OHD3 and 1,25(OH)2D3 [20]. Further complicating the issue of understanding vitamin D activation, catabolism, and metabolite analysis is the presence of vitamin D2. Vitamin D2 has been shown to contribute significantly to the overall vitamin D status in humans and other mammals consuming supplemental vitamin D2 [21–23]. Vitamin D2 can also be metabolized in a similar fashion to produce several metabolites analogous to the vitamin D3 endocrine system, including the hormonally
17
CHAPTER 2 Vitamin D Metabolism
OH
−H2 O
OH (1α)
OH
(3β)HO
OH
OH(1α)
(3α)HO
1,25(OH)2D3
OH(1α)
1,25(OH)2-3-epi-D3 −H2O
OH
OH (1α)
FIGURE 2 Pathways of vitamin D3 metabolism.
active form of vitamin D2, 1,25-dihydroxyvitamin D2 [1,25(OH)2D2] [24]. Simple inspection of the side chain, however, would imply that differences between metabolism of vitamin D2 and vitamin D3 may exist. The presence of unsaturation at carbon centers C-22/C-23, along with the additional methyl group at C-24, would seem to preclude the existence of the same metabolic pathways for the two vitamins. Figure 3 outlines some of the known pathways of vitamin D2 metabolism that have been shown to date. Deviations in the vitamin D2 and vitamin D3 pathways are discussed in detail in the following sections.
B. 25-Hydroxylase The 25-hydroxylation of vitamin D is the initial step in vitamin D activation. The enzyme responsible for production of this metabolite is located in the liver (see Chapter 4). Extrahepatic sources of 25-hydroxylation have been described [25]; however, experiments with hepatectomized rats provided evidence that the liver is the major, if not the sole, physiologically relevant site of 25-hydroxylation of vitamin D [26]. Subsequent studies
also described the existence of the 25-hydroxylase in both liver mitochondria and microsomes [27–31]. In early work, the microsomal enzyme was described as an enzyme of low capacity and high affinity and, therefore, the enzyme of greatest physiological importance [31]. In contrast, the mitochondrial enzyme was described as a high-capacity, low-affinity enzyme thought to be relevant only under conditions of high vitamin D concentration such as vitamin D toxicity [32]. Early evidence that the microsomal enzyme was the physiologically relevant enzyme came from experiments that suggested this enzyme could be regulated by vitamin D status [31]. It is now clear that liver production of 25-hydroxyvitamin D (25OHD) is not significantly regulated. 25OHD production is primarily dependent on substrate concentration. An important consequence of this lack of physiological regulation of 25OHD is that measurement of blood 25OHD is an excellent measure of vitamin D nutritional status. The purification and cloning of putative liver 25-hydroxylases have been reviewed several times [33–37]. Examination of the literature shows that most of the focus is on the mitochondrial 25-hydroxylase
18
RONALD L. HORST, TIMOTHY A. REINHARDT, AND G. SATYANARAYANA REDDY
FIGURE 3 Pathways of vitamin D2 metabolism.
designated CYP27A1, which is a cytochrome P450 capable of C-26(27) hydroxylation of sterols involved in bile acid synthesis and the 25-hydroxylation of vitamin D3. The rat, rabbit, and human enzymes have been cloned [38–41]. The CYP27A1 clone has been expressed
in COS cells [39,42], and its activity has been isolated from the mitochondria of these cells. The expressed enzyme was found to 27-hydroxylate cholestanetriol and 25-hydroxylate vitamin D3. However, CYP27A1 does not 25-hydroxylate vitamin D2 [39]. Rather, CYP27A1
CHAPTER 2 Vitamin D Metabolism
was found to 24-hydroxylate and 26(27)-hydroxylate vitamin D2. These activities could explain the presence of 24-hydroxyvitamin D2 (24OHD2), 1,24-dihydroxyvitamin D2 [1,24(OH)2D2], and 24,26-dihydroxyvitamin D2 [24,26(OH)2D2] [43–45] in the plasma of rats and cows. Since rats fed vitamin D–deficient diets and supplemented with physiological amounts of vitamin D2 have 25OHD2 as their predominant monohydroxylated vitamin D2 metabolite in the plasma [44] and targeted disruption of CYP27A1 does not decrease serum 25OHD3 [46], CYP27A1 is likely not the physiologic enzyme responsible for the 25-hydroxylation of vitamin D. The rat liver microsomal 25-hydroxylase (CYP2C11) has also been studied, but it has been shown to be male-specific [47]. Data have also been presented indicating that microsomes do [48] or do not [49] possess 25-hydroxylase activity. Therefore, conclusions regarding the importance of the CYP27A1 and other ostensible microsomal 25-hydroxylases require additional research. Data obtained studying pig liver 25-hydroxylation of vitamin D3 [36,37,48,50,51] suggest that a third liver 25-hydroxylase exists that is microsomal in origin. In the pig, this enzyme, CYP2D25, is present equally in males and females and is markedly different from CYP27A1 and CYP2C11 based on a terminal amino acid sequence [51]. Most important is the finding that this pig microsomal enzyme 25-hydroxylates vitamin D2 and vitamin D3 equally. The 25-hydroxylation of vitamin D is not yet completely understood. Several enzymes may play a role in the 25-hydroxylation of vitamin D. Whether one enzyme is more physiologically relevant than others remains to be determined. Studies in primary cultures of pig hepatocytes suggest that both CYP2D25 and CYP27A1 can play a role in 25-hydroxylation of vitamin D3 [51]. Nevertheless, it is clear that mammals can use vitamin D2 as a sole source of vitamin D. Therefore, any 25-hydroxylase proclaimed as the key enzyme(s) in the 25-hydroxylation of vitamin D must be capable of 25-hydroxylating vitamin D2 as well as vitamin D3. If mammals, for example, possessed an enzyme with specificity for the vitamin D2 side chain, this enzyme may be missed due to the almost exclusive use of vitamin D3 or vitamin D3 analogs as substrates for the 25-hydroxylating reaction.
C. 1α-Hydroxylase (CYP27B1) In the late 1960s, 25OHD3 was believed to be the metabolically active form of vitamin D. However, the presence of a more polar metabolite, which accumulated
19 in the intestinal mucosa chromatin of chicks administered 3H-labeled vitamin D3, suggested a new candidate for the active form of vitamin D [52]. Subsequent work by Lawson et al. [53] showed that during the formation of this metabolite the 1α-3H was lost. This led them to suggest that the new metabolite had an oxygen function inserted at C-1 in addition to the hydroxyl group at C-25. The enhanced biological activity of this new metabolite was evident before its structure could be determined [54–56]. Fraser and Kodicek [56] demonstrated that nephrectomy abolished production of the new metabolite, and this active vitamin D compound was synthesized by kidney mitochondria. In 1971, three laboratories identified the active form of vitamin D as 1,25(OH)2D3 [11,12,57]. Subsequently, the vitamin D2 form was also isolated and identified [24]. The CYP27B1 is located in the inner mitochondrial membrane of the proximal convoluted tubule cells of the kidney [58] and is discussed in detail in Chapter 5. Extrarenal sites of 1α-hydroxylation have been reported in bone, liver, placenta, macrophages, and skin [59]. The physiological significance of these sites on systemic calcium metabolism is in doubt, as nephrectomy and/or severe renal failure results in very low to undetectable circulating 1,25(OH)2D3 levels [60]. The regulation of 1,25(OH)2D3 production is reciprocally regulated with respect to 24,25(OH)2D3 [61]. Hypocalcemia caused by calcium-deficient diets, vitamin D deficiency, or pathological factors results in increased production of 1,25(OH)2D3 [61–67]. This hypocalcemic-mediated induction of 1,25-dihydroxyvitamin D [1,25(OH)2D] production is secondary to increased PTH. Administration of PTH to animals results in increased 1,25(OH)2D3 production [64,67,68]. PTH treatment in vitro induces CYP27B1 in renal slices [63] and cultured kidney cells [67,69] and is cAMP dependent [63,67,69,70]. Thyroparathyroidectomy (TPTX) or parathyroidectomy (PTX) results in the loss of the ability to synthesize 1,25(OH)2D3. In humans, acute administration of PTH or primary hyperparathyroidism results in increased production of 1,25(OH)2D [14,71], which is evidenced by elevations in plasma 1,25(OH)2D. However, in animal studies where PTH was administered chronically to goats and calves, a transient rise in plasma 1,25(OH)2D was observed followed by a rapid decline to nearly undetectable levels [72,73]. These results could be attributed to hypercalcemic feedback on the renal CYP27B1. When plasma calcium in these animals reached 13 mg/dl, 1,25(OH)2D production appeared to cease. This same group conducted similar experiments in rats and showed that chronic PTH infusion did not result in a reduction of plasma 1,25(OH)2D3, but rather a
20
RONALD L. HORST, TIMOTHY A. REINHARDT, AND G. SATYANARAYANA REDDY
modest rise [74]. Clearly, there are varying degrees of direct calcium-mediated feedback on the renal CYP27B1. It is possible that species and age affect the set points at which plasma calcium becomes a direct negative regulator of 1,25(OH)2D3 production. In contrast to the indirect role of plasma calcium in inducing 1,25(OH)2D3 production, the role of plasma phosphate appears more direct. As plasma phosphate declines, animals shift from 24,25(OH)2D production to increased 1,25(OH)2D3 production [75,76]. Since phosphate-deficient animals are hypercalcemic, serum PTH is down and therefore cannot be providing the signal to increase 1,25(OH)2D3 production. Furthermore, TPTX phosphate-deficient animals produce 1,25(OH)2D3 similarly to intact phosphate-deficient animals [66]. Gray [77] demonstrated that hypophysectomy abolished the increase in plasma 1,25(OH)2D concentrations that normally accompanied dietary phosphate deprivation. Gray demonstrated that growth hormone or triiodothyronine replacement to hypophysectomized rats restored elevations in plasma 1,25(OH)2D associated with low dietary phosphorus, therefore suggesting a permissive role of these hormones in regulation of the renal CYP27B1 during phosphorus deficiency. A direct negative effect of 1,25(OH)2D on its own production has been reported. The inhibitory effect of 1,25(OH)2D3 on renal CYP27B1 activity occurs both in vivo and in vitro [64,78]. This repressive activity is probably indirect [79]. In vivo, this effect may be partially mediated through the ability of 1,25(OH)2D3 to inhibit PTH secretion [80]. Similarly, Beckman et al. [81] showed that vitamin D toxicity mildly inhibited renal CYP27B1 activity while low-calcium diets significantly induced the CYP27B1. They further demonstrated that administration of toxic doses of vitamin D to animals fed a calcium-deficient diet reduced CYP27B1 activity by 90%. This result occurred in spite of the fact that these animals were hypocalcemic and had serum PTH levels equal to those of control animals receiving calcium-deficient normal vitamin D diets. These data suggest that high plasma concentration of vitamin D metabolites may act directly to suppress CYP27B1 activity. There are many additional factors such as calcitonin (CT), acidosis, sex steroids, prolactin, growth hormone, glucocorticoids, thyroid hormone, and pregnancy that are potential regulators of 1,25(OH)2D production. One of the most recent and interesting is the requirement for the endocytic receptor megalin in the proximal tubular cells to allow uptake of 25OHD for 1α-hydroxylation [82]. Discussion of these is beyond the scope of this general review of vitamin D metabolites. The characteristics and regulation of the CYP27B1 are described further in several reports [37,83–85].
The cloning of the renal CYP27B1 has been achieved [86–89]. Studies with CYP27B1 knockout mice [90,91] suggest that these animals have abnormalities similar to those observed in PDDR. These knockout models will undoubtedly provide new insight into the functions of 1,25(OH)2D and are the subject of further review in Chapter 7.
D. 24-Hydroxylase (CYP24A1) The 24-hydroxylation of 25OHD3 and 1,25(OH)2D3 to form 24,25(OH)2D3 [92] and 1,24,25(OH)3D3 [93,94] is the primary mechanism and the first step in a metabolic pathway to inactivate and degrade these vitamin D metabolites. It now appears that CYP24A1 is ubiquitous and may be present in every cell and tissue that contains the vitamin D receptor (VDR). In the kidney, CYP24A1 is found on the inner mitochondrial membrane of the renal tubules [95]. The primary regulators of renal CYP24A1 activity are PTH and 1,25(OH)2D3. Normal and TPTX animals receiving injections or infusions of 1,25(OH)2D3 show marked increases in both renal CYP24A1 mRNA levels and activity [63,74,96,97]. Administration of PTH partially or completely blocks expression of CYP24A1 mRNA and activity in these animals [63,68,74,96,97]. PTH acts on the kidney via adenylate cyclase and cAMP, and it has been shown that infusions of cAMP in vivo block l,25(OH)2D3mediated inductions of the renal CYP24A1 [96,98]. Animals on calcium-deficient diets have elevated plasma 1,25(OH)2D concentrations, which are accompanied by suppressed or undetectable renal CYP24A1 activity [78,96], as well as reduced VDR concentrations [99]. The reasons for the inability of 1,25(OH)2D3 to up-regulate renal CYP24A1 during calcium deficiency are not clear. Iida et al. [100] have proposed that the down-regulation of renal VDR during calcium deficiency may be responsible for preventing the l,25(OH)2D3mediated induction of renal CYP24A1. In vivo studies by Reinhardt and Horst [74], however, suggest that under these conditions PTH is probably the more important mediator of renal CYP24A1 regulation rather than downregulation of VDR. In their experiments, Reinhardt and Horst [74] showed that 1,25(OH)2D3 treatment of animals on normal calcium diets resulted in significant up-regulation of renal CYP24A1 as well as VDR. However, when PTH was infused simultaneously with 1,25(OH)2D3, VDR up-regulation was still observed (albeit to a lesser degree), whereas CYP24A1 up-regulation was completely blocked. The importance of PTH in preventing the l,25(OH)2D3-mediated up-regulation of the renal CYP24A1 is also apparent
21
CHAPTER 2 Vitamin D Metabolism
by observation in aged rats. With advancing age, renal PTH receptors are down-regulated [101,102], while VDR remains unchanged [103]. The reduction in renal PTH receptors makes the kidney less responsive to PTH [64], which is associated with significant elevations in CYP24A1 mRNA [103,104]. These data suggest that renal responsiveness to PTH, not a decline in VDR, is the major physiological regulator of the renal CYP24A1. In the intestine, 1,25(OH)2D3 is the primary regulator of CYP24A1. In vivo administration of 1,25(OH)2D3 rapidly induces intestinal CYP24A1 activity [105]. This activity peaks by 6 hr postinjection, and rapidly declines thereafter to control levels 24 hr postinjection. Time-course experiments show that CYP24A1 mRNA peaks 4 to 6 hr post-injection and then rapidly disappears [104]. This is in contrast to the renal CYP24A1 mRNA, which peaks 12 to 24 hr post-l,25(OH)2D3 treatment and declines much more slowly. Shinki et al. [96] proposed that the intestinal CYP24A1 was 100 times more sensitive to 1,25(OH)2D3 stimuli than the renal CYP24A1. However, they examined CYP24A1 mRNA only 3 hr after a 1,25(OH)2D3 dose. Because renal CYP24A1 requires an additional 6 to 12 hr to reach peak expression, they likely underestimated the true sensitivity of the kidney to a 1,25(OH)2D3 dose. In contrast to their effect in the kidney, TPTX, PTH administration, or cAMP infusion does not affect intestinal expression of CYP24A1 induced by 1,25(OH)2D3 [96]. Animals fed low-calcium diets, with the associated secondary hyperparathyroidism and high plasma 1,25(OH)2D3 concentrations, have marked inductions of both intestinal CYP24A1 mRNA and activity [96,105]. Another contrast between intestinal and renal CYP24A1 expression is seen in the aging rat model. Intestinal CYP24A1 mRNA and activity decline or change very little in the aged animal. This contrasts to the largely increased expression of renal CYP24A1 observed in the aged animal [103]. Calcitonin has been shown to be a potent suppressor of intestinal CYP24A1 expression [106]. In these experiments, Beckman et al. [81,106] showed that vitamin D toxicity was a potent inducer of CYP24A1 mRNA and enzyme expression in both intestine and kidney. They also showed that if the hypervitaminosis D3-induced hypercalcemia was prevented by feeding low-calcium diets, the intestinal CYP24A1 expression was enhanced four-fold over hypercalcemic animals receiving the same toxic doses of vitamin D3 but consuming a normal calcium diet. These observations prompted the examination of the possibility that CT released in response to the hypercalcemia may have suppressed the induced expression of the intestinal CYP24A1. In their series of experiments, Beckman et al. [106] clearly demonstrated that CT was a potent suppressor of intestinal CYP24A1 activity. Conceivably, the CT-mediated suppression of
CYP24A1 activity could enhance l,25(OH)2D-mediated activities by prolonging its half-life. This could exacerbate conditions that manifest hypercalcemia, such as hypervitaminosis D, by preventing 24-hydroxylation and catabolism of active vitamin D metabolites. The CYP24A1 has been purified [107,108] and cloned [107,109] and the clone has been expressed [107,109]. Analysis of the amino acid sequence of the rat and human CYP24A1s showed that the sequences were 90% similar. The 21-amino acid heme binding region was found to be 100% identical [109]. Ohyama et al. [110] isolated the gene encoding the rat CYP24A1. This single-copy gene was approximately 15 kb and was composed of 12 exons. Several putative vitamin D response elements have been identified and are currently under study. Details of the purification and cloning of the CYP24A1 have been reviewed previously [35], and additional review of the molecular analysis and regulation of the CYP24A1 can be found in Chapter 6 in this book.
E. Physiological Role of CYP24A1 The major site for 24-hydroxylation appears to be the kidney. This is based on the observation that nephrectomy reduced or eliminated plasma 24,25(OH)2D3 [111]. However, nephrectomy also eliminates the production of 1,25(OH)2D3, the primary stimulator of CYP24A1. Therefore, the possibility remains that 24,25(OH)2D3 may reappear in plasma of nephrectomized subjects treated with therapeutic doses of 1,25(OH)2D3. Some studies suggest that 25OHD3 is the primary substrate for CYP24A1 [112]; however, it is now generally accepted that CYP24A1 is distributed throughout the body and that 1,25(OH)2D3, rather than 25OHD3, is the preferred substrate for CYP24A1 [96,113,114]. Napoli et al. [115] and Napoli and Horst [116] identified the formation of 24-oxo-1,25(OH)2D3 and 24-oxo-l,23,25-trihydroxyvitamin D3 [24-oxo-1,23, 25(OH)3D3] from intestinal homogenates incubated with physiological amounts of 1,25(OH)2D3. The formation of 24-oxo-1,23,25(OH)3D3 from 1,25-(OH)2D3 was enhanced by treatment of experimental animals with exogenous 1,25-(OH)2D3. They suggested that 24-hydroxylation, followed by C-23 oxidation, most likely represents a mechanism for terminating the cellular action of 1,25(OH)2D3. In a review, Haussler [117] proposed a model for the cellular action of 1,25(OH)2D3 in which he suggested that receptor-mediated, selfinduced catabolism of 1,25(OH)2D3 modulates the action of 1,25(OH)2D3. The work of Lohnes and Jones [118], Reddy and Tserng [119], and Makin et al. [120] provided further support for this proposal by showing
22
RONALD L. HORST, TIMOTHY A. REINHARDT, AND G. SATYANARAYANA REDDY
the ubiquitous presence of catabolic pathways initiated by CYP24A1 in 1,25(OH)2D3 target tissues and the complete destruction of 1,25(OH)2D3 by these pathways. In fact, CYP24A1 has been shown to do more than just initiate this catabolic cascade. Akiyoshi-Shibata et al. [121] expressed the rat CYP24A1 cDNA in Escherichia coli. They found that this enzyme not only 24-hydroxylates but catalyzes the dehydrogenation of the 24-OH group and performs 23-hydroxylation resulting in 24-oxo-l,23,25(OH)3D3 production. Only the cleavage at C-23/C-24 resulting in the 24,25,26,27tetranor-1OH,23COOHD3 was not demonstrated. However, it is now recognized from the work of several groups that CYP24A1 is capable of the complete catabolism of 25OHD3 and 1,25(OH)2D3 via a fivestep reaction process that includes 24-hydroxylation, 24-oxidation, 23-hydroxylation, side-chain cleavage, and subsequent production of the final degradative product, calcitroic acid [121–125]. Direct evidence that self-induced metabolism of 1,25(OH)2D3 suppresses the action of 1,25(OH)2D3 on target cells was reported by Pols et al. [126,127] and Reinhardt and Horst [128,129]. Both laboratories showed that ketoconazole inhibited l,25(OH)2D3induced metabolism. This inhibition resulted in increased specific accumulation of 1,25(OH)2D3 in target cells and a significant increase in the cellular half-life of l,25(OH)2D3-occupied VDR [129]. A result of blocking the self-induced metabolism of 1,25(OH)2D3 was up-regulation of the VDR. Reinhardt and Horst [128] extended these studies by demonstrating that selfinduced metabolism of 1,25(OH)2D3 in target cells limits the response of target cells to a primary 1,25(OH)2D3 stimulus by reducing occupancy of VDR by 1,25(OH)2D3 and by preventing VDR up-regulation. Additionally, their data showed that entry of 1,25(OH)2D3 into the cell is restricted due to extensive metabolism of the 1,25(OH)2D3. In whole-cell VDR assays, hormone was degraded so rapidly that VDR binding was prevented. Reinhardt et al. [130] confirmed the inhibitory effects of self-induced induction of CYP24A1 on the cellular action of 1,25(OH)2D3 in vivo by demonstrating that ketoconazole potentiates the 1,25(OH)2D3 up-regulation of VDR in rat intestine and bone. Clearly, one of the primary roles of CYP24A1 catabolic pathway is terminating the actions of 1,25(OH)2D3. Recent work using CYP24A1 knockout mice has provided additional evidence for the role of CYP24A1 in the regulation of 1,25(OH)2D activity via catabolism [19]. These mice were unable to clear 1,25(OH)2D, as was evidenced by abnormally high circulating concentrations of this metabolite. These data provided support to the contention that the primary role of CYP24A1 is
initiating the catabolic pathway and terminating the actions of 1,25(OH)2D. The role of CYP24A1 as an enzyme responsible for the production of a biologically active compound, i.e., 24,25(OH)2D3, has been controversial. The work of St-Arnaud et al. [19] has also addressed part of this issue. Deficient mineralization of intramembranous bone was found in CYP24A1-ablated mice. The genetic cross of CYP24A1-ablated mice with VDR-ablated mice rescued the defective bone phenotype, strongly suggesting that the high concentrations of 1,25(OH)2D acting through VDR and not 24,25(OH)2D was the cause of the bone defect. Constitutive expression of CYP24A1 in transgenic rats surprisingly resulted in low plasma 24,25(OH)2D and 25OHD with no effect on plasma 1,25(OH)2D [131,132]. These rats developed albumineria and hyperlipidemia and suffered from reduced bone mass. The authors demonstrated that excreted albumin appeared to compete for the binding and reabsorption of the DBP-25-OHD3 complex with megalin, resulting in a loss of 25OHD3 into the urine and subsequent reduction of plasma 24,25(OH)2D3. Supplementation of these rats with 25OHD3 prevented the bone loss without changing plasma 1,25(OH)2D.
F. Other Vitamin D3 Derivatives Functionalized at C-24 In a series of experiments conducted by Wichmann et al. [133,134], a number of 24-hydroxylated derivatives were isolated from plasma of chicks made toxic with vitamin D3. These metabolites included 24OHD3, 23,24,25-trihydroxyvitamin D3 [23,24,25(OH)3D3], and 24,25,26-trihydroxyvitamin D3 [24,25,26(OH)3D3]. These metabolites have not been described in animals receiving physiological amounts of vitamin D3, and their biological significance is unknown; however, it is likely that 23,24,25(OH)3D3 and 24,25,26(OH)3D3 are metabolites of 24,25(OH)2D3. 24,25(OH)2D3 is also the probable precursor to the formation of the side chain cleavage product, 25,26,27-tri-norvitamin D324-carboxylic [135]. This metabolite has been shown to be a product of in vitro kidney perfusion using 25OHD3 as substrate. The analogous pathway, however, could not be demonstrated using 1,25(OH)2D3 as substrate (S. Reddy, personal communication, 1996). Another metabolite isolated in the experiments of Wichmann et al. [133] was 23-dehydro-25OHD3. The immediate precursor, site(s), and biological activity of this compound are unknown. Plausible sources of the 23-dehydro compound are dehydration of 24,25(OH)2D3
CHAPTER 2 Vitamin D Metabolism
or 23,25-dihydroxyvitamin D3 [23,25(OH)2D3]. It is not certain if any of the metabolites are important under physiological conditions.
G. 23-Hydroxylation The discovery of a C-23 oxidative pathway emerged much later than the other pathways and was ushered in by the identification of 23(S),25(R)25OHD3-26,23lactone [136,137], 23(S),25(R)1,25(OH)2D3-26,23lactone [138], and their respective precursors 23(S),25(OH)2D3 and 1,23(S),25(OH)3D3 [139,140]. To date, there has been no specific 23-hydroxylase identified for the vitamin D system. Rather, like other side-chain modifications, 23-hydroxylation is likely carried out by CYP24A1 [122]. The compound, 25OHD3-26,23-lactone, can be detected in plasma from normal rats, pigs, and chicks [140,141]. However, in several species, this metabolite is not expressed unless animals are consuming excessive amounts of vitamin D3 [142]. This metabolite has unique activity in that it is three- to fivefold more competitive than 25OHD3 for binding to the plasma vitamin D-binding protein (DBP) [136]. It was, therefore, the first modification of 25OHD3 that led to enhanced binding to the plasma DBP. The metabolite, 1,25(OH)2D3-26,23-lactone, has also been demonstrated under normal conditions [143], with elevated plasma concentrations occurring during exogenous administration of pharmacological amounts of 1,25(OH)2D3 [144]. The major locus for formation of C-23 hydroxylated derivatives appears to be the kidney. Horst and Littledike [142] and Napoli et al. [140] demonstrated that nephrectomy eliminated or greatly impaired the biosynthesis of 25OHD326,23-lactone when animals were treated with excess vitamin D3 or 25OHD3. They showed that this response was due to the inability of the animals to synthesize 23(S),25(OH)2D3. However, when 23(S),25(OH)2D3 was given to nephrectomized animals, the synthesis of 25OHD3-26,23-lactone was restored. These data suggested that C-23-hydroxylation occurred predominantly, but not exclusively, in the kidney, whereas extrarenal tissues are quantitatively important in the pathway leading to 25OHD3-26,23-lactone synthesis, which includes formation of the lactone intermediates 23,25,26(OH)3D3 [145] and 25OHD3-26,23-lactol [146]. Although ambiguities remain regarding the biological effects of C-24 oxidation, 23-hydroxylation appears to clearly be a deactivation event. 23-Hydroxylation is the first side-chain modification of 25OHD3 noted to substantially reduce its affinity for the plasma DBP [147]. 23-Hydroxylation of 1,25(OH)2D3 also leads to its
23 increased plasma clearance and reduced VDR binding and biological activity [148]. The role of 23-hydroxylation as a primary oxidation event for the further metabolism of 25OHD3 and 1,25(OH)2D3 is relatively minor to its role in the further metabolism of vitamin D3 metabolites that have been previously oxidized at C-24. In other words, very little production of 23,25(OH)2D3 or 1,23,25(OH)3D3 would be expected under normal conditions. Rather, the convergences of the C-24 and C-23 oxidative pathways would lead predominantly to the formation of 24-keto-1,23,25(OH)2D3 and 24-keto-1,23,25(OH)3D3 [116], which subsequently cleave to form C-23 acids [119,149]. Therefore, as indicated in Fig. 2, the C-23 oxidative pathway can lead to two different patterns of side-chain modifications for both 25OHD3 and 1,25(OH)2D3. One pathway, which is relatively minor under physiological conditions and more predominant during hypervitaminosis D3, leads through 23-hydroxylation to formation of the lactones, whereas the other more physiologically significant pathway leads through 24-hydroxylation to 23-hydroxylation of 24-keto metabolites. Other oxidized C-23 metabolites that have been identified include 23-keto derivatives of 25OHD3 and 1,25(OH)2D3. 23-Keto-25-hydroxyvitamin D3 was synthesized in vitro from 23(S)25(OH)2D3 and 23(R),25(OH)2D3 and has unique properties in that it binds with twofold higher affinity than 25OHD3 for binding sites on the plasma DBP [150]. This affinity should be compared to that for 23(S),25(OH)2D3, which binds with 6- to 10-fold less affinity. 23-Keto25OHD3 is also about fourfold more competitive than 25OHD3 for binding to the VDR. 23-Ketonization is, therefore, the first example of a side-chain modification enhancing the affinity of 25OHD3 for the VDR. This high affinity of 23-keto-25OHD3 for VDR prompted biosynthesis of 23-keto-l,25(OH)2D3 to determine if this modification might enhance binding of 1,25(OH)2D3 to VDR. Horst et al. [151] prepared this metabolite by incubating 23-keto-25OHD3 in kidney homogenates prepared from vitamin D–deficient chicks. The major metabolite was isolated and identified as 23-keto-1,25(OH)2D3 and was shown to possess about 40% the activity of 1,25(OH)2D3 for VDR binding. 23-Ketonization of 1,25(OH)2D3, therefore, reduced the affinity of 1,25(OH)2D3 to VDR rather than increased binding. 23-Keto metabolites do not appear to be synthesized under physiologic conditions, as Napoli et al. [115] could not demonstrate the presence of these metabolites from rat intestinal homogenates incubated with 1,25(OH)2D3 or from intestinal extracts from rats dosed with 1,25(OH)2D3.
24
RONALD L. HORST, TIMOTHY A. REINHARDT, AND G. SATYANARAYANA REDDY
H. C-26 Hydroxylation 26-Hydroxylation of 25OHD3 and 1,25(OH)2D3 produces 25,26(OH)2D3 [152] and 1,25,26(OH)3D3 [153], respectively. The natural product was originally assigned the 25(R) configuration. However, Partridge et al. [154] gave the assignment as 25(S). Ikekawa et al. [155] later discovered that the naturally occurring 25,26(OH)2D3 actually existed as a mixture of 25(S) and 25(R) isomers. Although this assignment seems somewhat trivial, it was important in unraveling a controversy that existed regarding the physiological precursor to the in vivo synthesis of 25OHD3-26,23-lactone. Hollis et al. [156] demonstrated that 25,26(OH)2D3 isolated from in vivo sources could act as a precursor to the formation of the 25OHD3-26,23-lactone. Subsequent research, however, suggested that synthetic 25(S),26(OH)2D3 (which at the time was thought to be the natural configuration) did not act as precursor to the formation of 25OHD3-26,23-lactone [157], but synthetic 25(R),26(OH)2D3 could act as a precursor [140,158]. As naturally occurring 25,26(OH)2D3 is a mixture of the R and S isomers, this research validated the conclusion of Hollis et al. [156] suggesting that formation of 25OHD3-26,23-lactone could indeed proceed through 25,26(OH)2D3. This pathway has been shown to be relatively minor [159,160], with the major pathway to 25OHD3-26,23-lactone synthesis proceeding through 23(S),25(OH)2D3 [140,145,158]. The major locus for the 26-hydroxylase is unknown. Blood concentrations of 25,26(OH)2D3 are not depressed in nephrectomized humans or pigs [142,161,162]. Therefore, production of these metabolites must take place at extrarenal sources. 26-Hydroxylase activity has, however, been demonstrated in microsomes isolated from rat and pig kidneys [163]. The only extrarenal source was reported in liver mitochondria [164]. The physiological role of the C-26 oxidative pathway remains elusive. However, 25,26(OH)2D3 and 1,25,26(OH)3D3 have been shown to possess biological activity with regard to stimulating bone calcium resorption and intestinal calcium absorption, albeit to a lesser degree than either 25OHD3 or 1,25(OH)2D3 [152,165]. Therefore, it seems unlikely that 26-hydroxylation is essential for calcium uptake from the gut or release of calcium from bone.
I. C-3 Epimerization Reddy et al. [159,160] have reported the metabolism of 1,25(OH)2D3 in primary cultures of neonatal human keratinocytes and rat osteosarcoma cells into the novel A-ring modified metabolite, 1,25(OH)2-3-epi-D3.
This epimer is formed as a result of the change in the orientation of the C-3 hydroxy group from β to α. Other investigators also confirmed this finding [166,167]. Epimerization of hydroxy groups is a wellknown phenomenon in bile acid metabolism [168] and the reaction is conducted by bile acid hydroxysteroid dehydrogenase. Figure 4 outlines the proposed pathways of C-3 epimerization of 1,25(OH)2D3 as described by Reddy et al. [159]. Although there are two potential pathways for the production of 1,25(OH)2-3-epi-D3 from 1,25(OH)2D3, the most likely pathway is through keto intermediates. The C-3 epimerization has been shown to play a major role in hormone activation and inactivation in other steroid systems [169]. Indeed, 1,25(OH)2-3-epi-D3 binds to the cellular 1,25(OH)2D3 receptor (VDR) with less affinity [170,171] and has minimal activity at activating intestinal calcium transport and bone calcium resorption than 1,25(OH)2D3 [170]. On the other hand, 1,25(OH)2-epi-D3 is equipotent to 1,25(OH)2D3 at suppressing parathyroid hormone secretion in bovine parathyroid cells [171] and at inhibiting keratinocyte proliferation [170,172]. While 1,25(OH)2-3-epi-D3 does undergo side-chain metabolism [159,171], its conversion to C-23 and C-24 oxidized metabolites occurs at a slower rate than 1,25(OH)2D3 [171]; therefore, enhanced metabolic stability of the 1,25(OH)2-3-epi-D3 has been proposed as a possible explanation for the high in vitro activity in spite of its reduced binding affinity for VDR [171]. Thus, the enzyme(s) responsible for C-3 epimerization appears to play an important role not only in the regulation of intracellular concentration of 1,25(OH)2D3 but also in the formation of metabolites with a different biological activity profile in specific target tissues. These differences in intracellular concentration of 1,25(OH)2D3 and its metabolites from one tissue to another may be one possible explanation for the well-known tissuespecific actions of 1,25(OH)2D3.
J. Unique Aspects of Vitamin D2 Metabolism The 24 position of vitamin D2, in contrast to the similar position in vitamin D3, can be considered to be highly reactive. It is a tertiary carbon as well as an allylic position, and the formation of a reactive intermediate (radical, cation) at this position would be highly stabilized. The proximity of this reactive center to the 25 position would afford the possibility of C-24-hydroxylation of vitamin D2, but the presence of the C-24 methyl would preclude further oxidation to C-24-keto compounds as is known to occur in vitamin D3 metabolism. Jones et al. [43] were the first to demonstrate C-24 oxidation when they isolated 24OHD2
25
CHAPTER 2 Vitamin D Metabolism
FIGURE 4
Possible pathways of epimerization at C-3 of 1,25(OH)2D3 (adapted from [159]).
from the plasma of male rats treated with 100 IU of radiolabeled vitamin D2. Engstrom and Koszewski [173] have determined that production of 24OHD2 can occur in liver homogenates from a variety of species, and actually exceeds the formation of 25OHD2. Horst et al. [44] have shown that the concentration of 24OHD2 in plasma was about 20% that of 25OHD2 in rats receiving physiological doses of vitamin D2, and was equivalent to 25OHD2 in rats receiving pharmacological doses of vitamin D2. They also demonstrated that 1α-hydroxylation of 24OHD2 to form 1,24(OH)2D2 represented a minor but significant pathway for vitamin D2 activation. In their experiments, they determined that 1,24(OH)2D2 rivaled both 1,25(OH)2D2 and 1,25(OH)2D3 in biopotency. Both 24OHD2 and 25OHD2 and their 1α-hydroxylated metabolites can undergo subsequent hydroxylation to form 24,25-dihydroxyvitamin D2 [24,25(OH)2D2] and 1,24,25-trihydroxyvitamin D2 [1,24,25(OH)3D2]. The formation of 1,24,25(OH)3D2 represented an unequivocal deactivation of the vitamin D2 molecule [174]. Conversely, the comparable vitamin D3 metabolite, 1,24,25(OH)3D3, maintains significant biological activity and must undergo further side-chain oxidation to be rendered totally inactive [174]. Although vitamin D2 is known to undergo side-chain oxidation, only recently has evidence emerged suggesting that vitamin D2 (like vitamin D3) undergoes side-chain cleavage.
The paucity of information regarding vitamin D2 sidechain metabolism is primarily due to the lack of appropriate radiolabeled and cold substrates. Zimmerman et al. [175] have used [9,11-3H]-1,25-(OH)2D2 and cold 1,25(OH)2D2 as substrates to determine if 1,25-(OH)2D2 undergoes side-chain cleavage. Similar experiments were done using [3α-3H]-25OHD2 as substrate [176]. In these metabolism experiments it became clear that aqueous-soluble metabolites were being produced from 1,25(OH)2D2 [175] and 25OHD2 [176]. Furthermore, Zimmerman et al. [175] demonstrated that calcitroic acid was the major aqueous-soluble metabolite being produced from cell culture and kidney perfusion experiments. In a more recent study Horst et al. [177] studied the metabolism of 1,25(OH)2D3 and 1,25(OH)2D2 using a purified rat CYP24A1 system. As expected, 1,23(OH)2-24,25,26,27-tetranor D and calcitroic acid were the major lipid- and aqueous-soluble metabolites, respectively, when 1,25(OH)2D3 was used as substrate. However, when 1,25-(OH)2D2 was used as substrate, 1,24,25(OH)3D2 was the major lipid-soluble metabolite with no evidence for the production of either 1,23(OH)2-24,25,26,27-tetranor D or calcitroic acid. Apparently, the CYP24A1 was able to 24-hydroxylate 1,25(OH)2D2, but was unable to cause further changes that would result in side-chain cleavage. When the analog 1,25-(OH)2-22-ene-D3 was used as substrate, CYP24A1 was again able to effect 24-hydroxylation
26
RONALD L. HORST, TIMOTHY A. REINHARDT, AND G. SATYANARAYANA REDDY
but not side-chain cleavage (R. Horst, S. Reddy, and J. Omdahl, 2003, unpublished data). Sunita Rao et al. [178] demonstrated that 1,25(OH)2-22-ene-D3 could be metabolized to calcitroic acid by RWLue-4 cells and rat kidney. They suggested that the 1,25(OH)2-22-ene-D3 was first hydroxylated at C-24, followed by further oxidation to 1,25(OH)2-24-oxo-22-ene-D3 prior to side-chain, double-bond reduction to form 1,25(OH)224-oxo-D3. The 1,25(OH)2-24-oxo-D3 was then further metabolized to calcitroic acid, presumably by CYP24A1. The compound 1,25(OH)2D4 [a.k.a. 22,23 dihydro1,25-(OH)2D2] has also been shown to undergo sidechain oxidation similar to that of 1,25(OH)2D2 in vitro [179] and metabolized to calcitroic acid in vivo [180]. These results suggest that metabolism of 1,25(OH)2D2 to calcitroic acid clearly involves enzymes other than CYP24A1. The compounds 25,26-dihydroxyvitamin D2 [25,26(OH)2D2] and 1,25,26-trihydroxyvitamin D2 [1,25,26(OH)3D2] have been chemically synthesized [181,182]. The metabolite 25,26(OH)2D2 has also been tentatively, but not exhaustively, identified from in vivo sources [141,183], but it could not be demonstrated in kidneys perfused with 25OHD2 [184]. Similarly, 1,25,26(OH)3D2 could not be demonstrated [185]. However, it is clear that 26-hydroxylation does occur when vitamin D2 metabolites have been previously 24-hydroxylated. For example, when 24,25(OH)2D2 and 1,24,25(OH)3D2 were used as precursors, the metabolites 24,25,26(OH)3D2 and 1,24,25,26(OH)4D2, respectively, were produced in significant amounts [184,185]. Similarly, Koszewski et al. [45] and Jones et al. [186] demonstrated that C-26 hydroxylation was the major metabolic pathway for the further metabolism of 24OHD2 and 1,24(OH)2D2. Oxidation at C-24, therefore, appears to be a prerequisite for C-26 oxidation of vitamin D2 compounds. A similar situation also appears to exist for C-28 oxidation as demonstrated by Reddy and Tserng [184,185]. In their experiments, they isolated 24,25,28(OH)3D2 and 1,24,25,28(OH)4D2 from rat kidney perfusions and, through the use of various substrates, were able to show C-28 hydroxylation of vitamin D2 metabolites occurs only after C-24 hydroxylation [187].
III. VITAMIN D TOXICITY A. Overview The first documented reports of vitamin D intoxication were made in the late 1920s by Kreitmeir and Moll [188] and Putscher [189]. These cases resulted from the ingestion of large quantities of vitamin D in
the diet. Vitamin D intoxication, however, has never been reported following prolonged sunlight exposure. Holick et al. [190] suggested that nature has provided various control points that prevent the overproduction of vitamin D3 by the skin. The most important point of control is the diversion of vitamin D3 production from 7-dehydrocholesterol to non–biologically active overirradiation products such as lumisterol and tachysterol. In addition, these authors suggested that skin pigmentation and latitude were also significant determinants (albeit to a lesser degree) that limit the cutaneous production of vitamin D3. As discussed earlier, once vitamin D is in circulation, the conversion to 25OHD is relatively uncontrolled. Normally, 25OHD circulates at 30 to 50 ng/ml in most species [141]. However, when vitamin D is given in excess, plasma 25OHD can be elevated to concentrations of 1000 ng/ml or greater [191,192], while plasma 1,25(OH)2D remains at or below normal concentrations [193]. When circulating at very high concentrations, 25OHD can compete effectively with 1,25(OH)2D for binding to the VDR. Therefore, during vitamin D toxicosis, 25OHD can induce actions usually attributed to 1,25(OH)2D [193]. High circulating 25OHD can, therefore, explain how humans with low circulating concentrations of 1,25(OH)2D can show signs of vitamin D toxicity [193], and why anephric humans [who are incapable of producing 1,25(OH)2D] can become vitamin D toxic [194]. Clinical aspects of vitamin D toxicity are discussed in Chapter 78. Although it is generally accepted that 1,25(OH)2D is reduced during hypervitaminosis, a notable exception to this generalization is the ruminant [195,196]. Horst and co-workers [195,197] have shown that vitamin D3 intoxication initiated by giving 15 million IU of vitamin D3 intramuscularly (i.m.) results in significant elevations in plasma 1,25(OH)2D3. In contrast, pigs given the same i.m. dose showed a reduction in plasma 1,25(OH)2D3, as was observed in other species [197]. Therefore, elevations in plasma 1,25(OH)2D may play a significant role in the pathogenesis of vitamin D toxicity in ruminants.
B. Differences in Toxicity between Vitamins D2 and D3 Most research dealing with utilization of vitamins D2 and D3 assumes that these two forms are equally potent in most mammals. However, when large and potentially toxic doses were administered orally to rhesus monkeys [198] and horses [199], or were used in treating childhood osteodystrophy [200], vitamin D2
27
CHAPTER 2 Vitamin D Metabolism
presented fewer hypercalcemic side effects than vitamin D3. In addition, 1OHD2, which is as effective as the 1OHD3 in standard bioassays, was shown to be five- to 15-fold less toxic than 1OHD3 [201]. Studies by Horst et al. [44] provide some insight into the difference between vitamin D2 and vitamin D3 toxicity. They demonstrated that under physiological conditions the predominant monohydroxylated form of both vitamins D2 and D3 is 25OHD. In the vitamin D2–dosed rats, 24OHD2 accounted for approximately 20% of the monohydroxylated metabolites, whereas 24OHD3 could not be detected in the vitamin D3–dosed rats. When a modest superphysiological dose (800 IU/day) of vitamin D3 was given to rats, 25OHD3 remained the predominant metabolite in vitamin D3–dosed rats and was present at 26.3 ng/ml. Under these conditions, there was still no evidence for the presence of 24OHD3. However, when the same amount of vitamin D2 was given, the concentrations of 24OHD2 (14.1 ng/ml) nearly matched those of 25OHD2 (15.9 ng/ml). Interestingly, the combined concentrations of 24OHD2 and 25OHD2 in the vitamin D2–dosed animals (~30 ng/ml) was similar to the 25OHD3 concentration (26.3 ng/ml) in the vitamin D3–dosed rat. In standard assays, 25OHD2 and 25OHD3 are equipotent at displacing 3H-1,25(OH)2D3 from the calf thymus VDR. However, 24OHD2 has been shown to have at least a twofold lower affinity for binding to the calf thymus VDR (R. L. Horst, unpublished data, 1996). Therefore, the reduced toxicity of vitamin D2 is probably a result of diverting metabolism away from the production of 25OHD2 in favor of 24OHD2, which has a relatively limited affinity for VDR (a step necessary for the initiation of a biological response). Further differences between vitamins D2 and D3 were noted in their ability to up-regulate the VDR. Beckman et al. [202] found that VDR was significantly more enhanced in animals fed excess vitamin D3 relative to those animals receiving an equivalent amount of vitamin D2. Increased VDR would potentially accentuate toxic side effects by enhancing the responsiveness of intestinal tissue to the elevated 25OHD.
C. Factors Affecting Toxicity The severity of the effects and pathogenic lesions in vitamin D intoxication depend on such factors as the type of vitamin D (vitamin D2 versus vitamin D3), the dose, the functional state of the kidneys, and the composition of the diet. Vitamin D toxicity is enhanced by a rich dietary supply of calcium and phosphorus, and it is reduced when the diet is low in calcium and phosphorus [203,204]. Toxicity is also reduced when the
vitamin is accompanied by high intakes of vitamin A or by thyroxine injections [205]. The route of administration also influences toxicity. Parenteral administration of 15 million IU of vitamin D3 in a single dose caused toxicity and death in many pregnant dairy cows [195]. On the other hand, oral administration of 20 to 30 million IU of vitamin D2 daily for 7 days resulted in little or no toxicity in pregnant dairy cows [206]. Napoli et al. [207] have shown that rumen microbes are capable of metabolizing vitamin D to the inactive 10-keto-19-norvitamin D3. Parenteral administration would circumvent the deactivation of vitamin D by rumen microbes and may partially explain the difference in toxicity between oral and parenteral vitamin D. Various measures have been used in human medicine for treatment of vitamin D toxicity. These measures are mainly concerned with management of hypercalcemia. Vitamin D withdrawal is obviously indicated. It is usually not immediately successful, however, owing to the long plasma half-life of vitamin D (5 to 7 days) and 25OHD (20 to 30 days). This is in contrast to the short plasma half-life of 1OHD3 (1 to 2 days) and 1,25(OH)2D3 (4 to 8 hr). Because intestinal absorption of calcium contributes to hypercalcemia, a prompt reduction in dietary calcium is indicated. Sodium phytase, an agent that reduces intestinal calcium absorption, has also been used successfully in vitamin D toxicity management in monogastrics [208]. This treatment would be of little benefit to ruminants because of the presence of rumen microbial phytases. There have also been reports that CT [209], glucagon [210], etidronate [211], and glucocorticoid therapy [212] reduce serum calcium levels or prevent the calcinosis resulting from vitamin D intoxication (see Chapter 78).
IV. SPECIES VARIATION IN VITAMIN D METABOLISM AND ACTION Most concepts of vitamin D metabolism and function have been developed with the rat and/or chick as experimental models. Studying vitamin D metabolism is hampered by the paucity of data on the normal circulating levels of vitamin D metabolites in birds, mammals, and reptiles under normal conditions. Most recent research has focused on the analysis of 25OHD and 1,25(OH)2D as indicators of vitamin D status or aberrant physiological states. Table I summarizes the concentrations of the two metabolites that have been reported for several species by various laboratories. Close inspection of the information suggests that some mammals (mole rat, wild wood vole, horse, and wild wood mouse) and aquatic species (lamprey, carp,
28
RONALD L. HORST, TIMOTHY A. REINHARDT, AND G. SATYANARAYANA REDDY
TABLE I Plasma 25-Hydroxyvitamin D and 1,25-Dihydroxyvitamin D Concentrations in Several Species of Animals Concentration Species
25OHD (ng/ml) 1,25(OH)2D (pg/ml) Ref.
Human Rhesus monkey Rhesus monkey Marmoset Marmoset Wild woodmouse Wild bank vole Mole rat Lamprey Shark Leopard shark Horned shark Carp Bastard halibut Atlantic cod Bullfrog (mature) Soft-shelled turtle Turkey Chicken Cow Sheep Pig Horse aND,
32 188 50 90 64 <5 <5 <2 NDa ND 56 33 ND ND <2 2 16 26 27 43 27 76 6
31 207 95 400 640 <10 <5 17 274 87 3 6 174 192 59 21 12 52 21 38 36 60 19
[162] [231] [232] [232] [233] [234] [234] [213] [235] [235] [236] [236] [235] [235] [237] [235] [235] [141] [141] [141] [141] [141] [238]
Not done.
halibut, and bullfrog) appear to have very low or undetectable concentrations of 25OHD, and yet these animals appeared to be normal with no evidence of vitamin D deficiency. It is questionable whether some of these species have a requirement for vitamin D. The damara mole rat, for example, is a subterranean herbivore that in its natural habitat has no access to any obvious source of vitamin D and consumes a diet of roots and tubers [213]. These animals exhibit a high apparent calcium absorption efficiency (91%) and, like the horse and rabbit, actually use renal calcium excretion as the major regulator of calcium homeostasis [214,215]. In studies with rabbits consuming adequate amounts of calcium, it is very difficult to develop any overt or histological signs of vitamin D deficiency, and vitamin D may play a minor, if any, role in normal day-to-day functions in these animals [216].
In the wild, most animals do not have a dietary need for vitamin D, as sufficient vitamin D3 can be synthesized in the skin on irradiation by sunlight. However, indoor confinement of humans and other animals has resulted in the diet becoming the main source of vitamin D, leading to considerable research to determine the amount of dietary vitamin D required to substitute for lack of exposure to sunlight. Photochemically produced vitamin D3 enters the circulation and becomes immediately available, whereas dietary vitamin D3 may undergo modifications prior to becoming available for use by the body. One species where significant modification of vitamin D occurs before absorption is the ruminant. Within 24 hr, as much as 80% of vitamin D can undergo metabolism in vitro in rumen incubation media [217]. At least four metabolites are produced by the rumen microbes [217,218]. Two of these metabolites have been identified as the cis (5Z) and trans (5E) isomers of 10-keto-19-norvitamin D3 (Fig. 2) [207]. The trans isomer has also been identified in cow plasma (R. L. Horst, unpublished data, 1983). Neither compound has agonistic activity with regard to promoting bone calcium resorption [219] or intestinal calcium absorption [207]. Rather, this novel metabolism is likely a detoxification process, as evidenced by the ability of ruminants to tolerate large oral doses of vitamin D3 that would be toxic if given parenterally. The presence of the rumen, therefore, represents a major control point in vitamin D metabolism that may differ from monogastrics. Such a control point may have survival value, because the ruminant evolved as a grazing animal with the opportunity for long periods of sunlight exposure, as well as consumption of large quantities of irradiated plants. If left uncontrolled, such a combination might result in vitamin D toxicity. Shortly after the discovery of vitamin D, it seemed apparent that vitamins D2 and D3 had similar biological activities in most mammals and that birds and New World monkeys discriminated against vitamin D2 in favor of vitamin D3 [220,221]. More recent research, fostered by the discovery of sensitive analytical techniques and the availability of high specific activity 3H-labeled vitamin D species, indicated that differences in the metabolism of vitamins D2 and D3 in mammals are perhaps widespread. Most notable were the apparent discrimination against vitamin D2 by pigs [222], cows [218], and humans [223] and the apparent preference for vitamin D2 by rats [222,224]. Vitamin D and its metabolites are transported in the blood of vertebrates attached to a specific protein commonly known as the vitamin D binding protein or DBP [225]. Baird et al. [226] have shown that protein binding increases the solubility of steroids and that the
29
CHAPTER 2 Vitamin D Metabolism
metabolic clearance rate of steroids is in part dependent on their binding to specific plasma proteins. Affinity of metabolites to the plasma transport proteins may, therefore, provide a means for determining which species would utilize vitamin D2 poorly. For example, if the binding protein showed lower affinity toward 25OHD2 relative to 25OHD3, then one would predict that 25OHD2 would be removed from the circulation faster than 25OHD3. This is indeed the case for the chick. Hoy et al. [227] showed that chick discrimination against vitamin D2 was probably a result of enhanced clearance of the vitamin D2 metabolites 25OHD2 and 1,25(OH)2D2, and that the enhanced clearance was associated with weaker binding of these vitamin D2 metabolites (relative to the vitamin D3 forms) to DBP. In one of the most comprehensive studies reported to date, Hay and Watson [228] studied the affinities of DBP for 25OHD2 and 25OHD3 in 63 vertebrate species. They found that the DBP in fish, reptiles, and birds discriminated against 25OHD2 in favor of 25OHD3, which is consistent (at least in birds) with the discrimination against vitamin D2. One notable exception to this hypothesis, however, is the New World monkey. Hay and Watson [228] found that in New World monkeys, the plasma transport protein has equal affinity for 25OHD2 and 25OHD3, which is inconsistent with the well-documented discrimination against vitamin D2. Factors other than affinity of the binding protein for 25OHD are, therefore, important in determining how efficiently the different forms of vitamin D can be utilized by animals. Another example of species discrimination against the different vitamin D forms is in the rat. However, in this species, discrimination is against vitamin D3 in favor of vitamin D2 [222]. The rat DBP is known to have equal affinity for 25OHD2 and 25OHD3, but a lower affinity for vitamin D2 relative to vitamin D3 [229]. Reddy et al. [230] suggested that the lower affinity for vitamin D2 resulted in its enhanced availability for liver 25-hydroxylation. Hence, in the presence of DBP, more 25OHD2 was made relative to 25OHD3 when equal amounts of vitamin D2 or vitamin D3 substrate were perfused into rat livers. This observation is consistent with the higher circulating concentrations of 25OHD2 observed in acute experiments with vitamin D–deficient rats dosed with equal amounts of vitamins D2 and D3 [222]. In the experiments conducted by Reddy et al. [230], if binding protein was eliminated from the perfusion media, equal amounts of 25OHD2 and 25OHD3 were synthesized. Collectively, these data suggest that discrimination against the different forms of vitamin D could likely result from variations in the affinity of DBP for the parent compound and/or one or
more of their metabolites. Regardless of the mechanism for discrimination, it appears that these differences are present to afford animals the most efficient utilization of the most abundant antirachitic agents available in their environment.
V. CONCLUSION Vitamin D metabolism still remains an exciting area of research with much more to be learned. Critical questions remain unanswered regarding complete elucidation of the vitamin D2 metabolic pathway, and species differences between vitamins D2 and D3 metabolism are still virtually unexplored. The introduction of vitamin D analogs has also resulted in a totally different set of issues regarding their metabolism, tissue kinetics, mechanism of action, and potential therapeutic uses.
References 1. Mellanby E 1919 An experimental investigation on rickets. Lancet I 4985:407–412. 2. McCollum EV, Simmonds N, Becker JE, Shipley PG 1922 Studies on experimental rickets. XXI. An experimental demonstration of the existence of a vitamin which promotes calcium deposition. J Biol Chem 53:293–312. 3. Norman AW 1979 Vitamin D: The calcium homeostaic hormone. In: Darby WJ (ed) Nutrition: Basic and Applied Science. Academic Press, New York. 4. Napoli JL, Horst RL 1984 Vitamin D metabolism. In: Kumar R (ed) Vitamin D: Basic and Applied Aspects. Martinus Nijhoff, Boston, pp. 91–123. 5. Lawson DEM 1978 Vitamin D. Academic Press, London. 6. Kodichek E 1974 The story of vitamin D from vitamin to hormone. Lancet I 7853:325–329. 7. Haussler MR, McCain TA 1977 Basic and clinical concepts related to vitamin D metabolism and action. N Engl J Med 297:974–983, 1041–1050. 8. Deluca HF, Schnoes HK 1976 Metabolism and action of vitamin D. Annu Rev Biochem 45:631–666. 9. IUPAC 1969 IUPAC Commission on the Nomenclature of Organic Chemistry and IUPAC–IUB Commission on Biochemical Nomenclature. Revised tentative rules for nomenclature of steroids. Biochem J 113:5–28. 10. Cahn RS, Ingold C, Prelog V 1966 Specification of molecular chirality. Angew Chem, Int Ed Engl 5:385–415. 11. Blackwood JE, Gladys CL, Loening KL, Petrarca AE, Rush JE 1968 Unambiguous specification of stereoisomerism about a double bond. J Am Chem Soc 90:509–510. 11a. Morii H, Lund J, Neville PF, DeLuca HF 1967 Biological activity of a vitamin D metabolite. Arch Biochem Biophys 120:508–512. 11b. Blunt JW, DeLuca HF, Schnoes HK 1968 25-Hydroxycholecalciferol: A biologically active metabolite of vitamin D3. Biochemistry 7:3317–3322. 11c. Norman AW, Myrtle JF, Midgett RJ, Nowicki HG 1971 1,25-Dihydroxycholecalciferol: Identification of the
30
12.
13.
14.
15. 16. 17.
18. 19.
20. 21.
22. 23.
24. 25. 26. 27. 28. 29.
RONALD L. HORST, TIMOTHY A. REINHARDT, AND G. SATYANARAYANA REDDY
proposed active form of vitamin D3 in the intestine. Science 173:51–54. Lawson DEM, Fraser DR, Kodicek E, Morris HR, Williams DH 1971 Identification of 1,25-dihydroxycholecalciferol, a new kidney hormone controlling calcium metabolism. Nature 230:228–230. Holick MF, Schnoes HK, DeLuca HF, Suda T, Cousins RJ 1971 Isolation and identification of 1,25-dihydroxycholecalciferol. A metabolite of vitamin D active in intestine. Biochemistry 10:2799–2804. Broadus AE, Horst RL, Lang R, Littledike ET, Rasmussen H 1980 The importance of circulating 1,25-dihydroxyvitamin D in the pathogenesis of hypercalciuria and renal-stone formation in primary hyperparathyroidism. N Engl J Med 302:421–426. Bordier P, Rasmussen H, Marie P, Miravet L, Gueris J, Ryckwaert A 1978 Vitamin D metabolism and bone mineralization in man. J Clin Endocrinol Metab 46:284–294. Ornoy A, Goodwin D, Noff D, Edelstein S 1978 24,25Dihydroxyvitamin D is a metabolite of vitamin D essential for bone formation. Nature 276:517–519. Canterbury JM, Lerman S, Claflin AJ, Henry H, Norman A, Reiss E 1977 Inhibition of parathyroid hormone secretion by 25-hydroxycholecalciferol and 24,25-dihydroxycholecalciferol in the dog. J Clin Invest 78:1375–1383. Henry HL, Norman AW 1978 Vitamin D: Two dihydroxylated metabolites are required for normal chicken egg hatchability. Science 201:835–837. St-Arnaud R, Arabian A, Travers R, Barletta F, Raval-Pandya M, Chapin K, Depovere J, Mathieu C, Christakos S, Demay MB, Glorieux FH 2000 Deficient mineralization of intramembranous bone in vitamin D-24-hydroxylase-ablated mice is due to elevated 1,25-dihydroxyvitamin D and not to the absence of 24,25-dihydroxyvitamin D. Endocrinology 141:2658–2666. Horst RL, Littledike ET, Riley JL, Napoli JL 1981 Quantitation of vitamin D and its metabolites and their plasma concentrations in five species of animals. Anal Biochem 116:189–203. Hartwell D, Hassager C, Christiansen C 1987 Effect of vitamin D2 and vitamin D3 on the serum concentrations of 1,25(OH)2D2 and 1,25(OH)2D3 in normal subjects. Acta Endocrinol 115:378–384. Hollis BW, Pittard WB 1984 Relative concentrations of 25-hydroxyvitamin D2/D3 and 1,25-dihydroxyvitamin D2/D3 in maternal plasma at delivery. Nutr Res (N Y) 4:27–32. Reinhardt TA, Horst RL, Orf JW, Hollis BW 1984 A microassay for 1,25-dihydroxyvitamin D not requiring high performance liquid chromatography: Application to clinical studies. J Clin Endocrinol Metab 58:91–98. Jones G, Schnoes HK, DeLuca HF 1975 Isolation and identification of 1,25-dihydroxyvitamin D2. Biochemistry 14:1250–1256. Tucker G, Gaganon R, Haussler M 1973 Vitamin D3-25hydroxylase: Tissue occurrence and apparent lack of regulation. Arch Biochem Biophys 155:47–57. Ponchon G, DeLuca HF 1969 Metabolites of vitamin D3 and their biological activity. J Nutr 99:157–167. Bhattacharyya MH, DeLuca HF 1973 The regulation of rat liver calciferol-25-hydroxylase. J Biol Chem 248: 2969–2973. Bhattacharyya MH, DeLuca HF 1973 Comparative studies on the 25-hydroxylation of vitamin D3 and dihydrotachysterol. J Biol Chem 248:2974–2977. Bjorkhem I, Holmberg I 1979 On the 25-hydroxylation of vitamin D3 in vitro studied with a mass fragmentographic technique. J Biol Chem 254:9518–9524.
30. Bjorkhem I, Holmberg I 1978 Assay and properties of a mitochondrial 25-hydroxylase active on vitamin D3. J Biol Chem 253:842–849. 31. Madhok TC, DeLuca HF 1979 Characteristics of the rat liver microsomal enzyme system converting cholecalciferol into 25-hydroxycholecalciferol. Evidence for the participation of cytochrome P-450. Biochem J 184:491–499. 32. Bjorkhem I, Holmberg I, Oftebro H, Pedersen JI 1980 Properties of a reconstituted vitamin D3 25-hydroxylase from rat liver mitochondria. J Biol Chem 255:5244–5249. 33. Okuda K 1992 Vitamin D3 hydroxylase—its occurrence, biological importance and biochemistry. In: Ruckpaul K, Rein H (eds) Cytochrome P-450 Dependent Biotransformation of Endogenous Substrates: Frontiers in Biotransformation, Vol. 6. Akademic Verlag, Berlin, pp.114–147. 34. Okuda K 1994 Liver mitochodrial p450 involved in cholesterol catabolism and vitamin activation. J Lipid Res 35:361–372. 35. Okuda K-I, Usui E, Ohyama Y 1995 Recent progress in enzymology and molecular biology of enzymes involved in vitamin D metabolism. J Lipid Res 36:1641–1652. 36. Postlind H, Axen E, Bergman T, Wikvall K 1997 Cloning, structure, and expression of a cDNA encoding vitamin D3 25-hydroxylase. Biochem Biophys Res Commun 241: 491–497. 37. Wikvall K 2001 Cytochrome P450 enzymes in the bioactivation of vitamin D to its hormonal form (review). Int J Mol Med 7:201–209. 38. Andersson S, Davis DL, Dahlbach H, Jornvall H 1989 Cloning, structure, and expression of the mitochondrial cytochrome P-450 sterol 26-hydroxylase, a bile acid biosynthetic enzyme. J Biol Chem 264:8222–8229. 39. Guo Y-D, Strugnell S, Back DW, Jones G 1993 Transfected human liver cytochrome P-450 hydroxylates vitamin D analogs at different side-chain positions. Proc Natl Acad Sci USA 90:8668–8672. 40. Cali J, Russell D 1991 Characterization of human sterol 27-hydroxylase. J Biol Chem 266:7774–7778. 41. Usui E, Noshiro M, Okuda K 1990 Molecular cloning of cDNA for vitamin D3 25-hydroxylase from rat liver mitochondria. FEBS Lett 262:135–138. 42. Usui E, Noshiro M, Ohyama Y, Okuda K 1990 Unique property of liver mitochondrial P450 to catalyze the two physiologically important reactions involved in both cholesterol catabolism and vitamin D activation. FEBS Lett 274: 175–177. 43. Jones G, Schnoes HK, Levan L, DeLuca HF 1980 Isolation and identification of 24-hydroxyvitamin D2 and 24,25-dihydroxyvitamin D2. Arch Biochem Biophys 202:450–457. 44. Horst RL, Koszewski NJ, Reinhardt TA 1990 1α-Hydroxylation of 24-hydroxyvitamin D2 represents a minor physiological pathway for the activation of vitamin D2 in mammals. Biochemistry 29:578–582. 45. Koszewski NJ, Reinhardt TA, Napoli JL, Beitz DC, Horst RL 1988 24,26-Dihydroxyvitamin D2: A unique physiological metabolite of vitamin D2. Biochemistry 27:5785–5790. 46. Rosen H, Reshef A, Maeda N, Lippoldt A, Shpizen S, Triger L, Eggertsen G, Bjorkhem I, Leitersdorf E 1998 Markedly reduced bile acid synthesis but maintained levels of cholesterol and vitamin D metabolites in mice with disrupted sterol 27-hydroxylase gene. J Biol Chem 273: 14805–14812. 47. Hayashi S, Usui E, Okuda K 1988 Sex-related difference in vitamin D3 25-hydroxylase of rat liver mitochondria. J Biochem (Tokyo) 103:863–866.
CHAPTER 2 Vitamin D Metabolism
48. Hosseinpour F, Wikvall K 2000 Porcine microsomal vitamin D(3) 25-hydroxylase (CYP2D25). Catalytic properties, tissue distribution, and comparison with human CYP2D6. J Biol Chem 275:34650–34655. 49. Saarem K, Bergseth S, Oftebro H, Petersen JI 1984 Subcellular localization of vitamin D3 25-hydroxylation in human liver. J Biochem (Tokyo) 259:10936–10946. 50. Axen E, Bergman T, Wikvall K 1994 Microsomal 25-hydroxylation of vitamin D2 and vitamin D3 in pig liver. J Steroid Biochem Mol Biol 51:97–106. 51. Hosseinpour F, Ibranovic I, Tang W, Wikvall K 2003 25-Hydroxylation of vitamin D3 in primary cultures of pig hepatocytes: evidence for a role of both CYP2D25 and CYP27A1. Biochem Biophys Res Commun 303:877–883. 52. Haussler MR, Myrtle JF, Norman AW 1968 The association of a metabolite of vitamin D3 with intestinal mucosa chromatin in vivo. J Biol Chem 243:4055–4064. 53. Lawson DEM, Wilson PW, Kodicek E 1969 Metabolism of vitamin D. A new cholecalciferol metabolite, involving loss of hydrogen at C-1, in chick intestinal nuclei. Biochem J 115:269–277. 54. Kodicek E, Lawson DEM, Wilson PW 1970 Biological activity of a polar metabolite of vitamin D3. Nature 228: 763–764. 55. Haussler MR, Boyce DW, Littledike ET, Rasmussen H 1971 A rapidly acting metabolite of vitamin D3. Proc Natl Acad Sci USA 68:177–181. 56. Fraser DR, Kodicek E 1970 Unique biosynthesis by kidney of a biologically active vitamin D metabolite. Nature 228:764–766. 57. Holick MF, Schnoes HK, Deluca HF 1971 Identification of 1,25-dihydroxycholecalciferol, a form of vitamin D3 metabolically active in the intestine. Proc Natl Acad Sci USA 68:803–804. 58. Paulson SK, Deluca HF 1985 Subcellular location and properties of rat renal 25-hydroxyvitamin D3-1α-hydroxylase. J Biol Chem 260:11488–11492. 59. Armbrecht HJ, Okuda K, Wongsurawat N, Nemani RK, Chen ML, Boltz MA 1992 Characterization and regulation of the vitamin D hydroxylases. J Steroid Biochem 43: 1073–1081. 60. Gray RW, Weber HP, Dominguez JH, Lemann J, Jr. 1974 The metabolism of vitamin D3 and 25-hydroxyvitamin D3 in normal and anephric humans. J Clin Endocrinol Metab 39:1045–1056. 61. DeLuca HF 1974 Vitamin D: The vitamin and the hormone. Fed Proc 33:2211–2219. 62. Boyle I, Gray R, Deluca HF 1971 Regulation by calcium of in vivo synthesis of 1,25-dihydroxycholecalciferol and 21,25-dihydroxycholecalciferol. Proc Natl Acad Sci USA 68:2131–2137. 63. Armbrecht HJ, Wongsurawat N, Zenser TV, Davis BB 1984 Effect of PTH and 1,25-(OH)2D3 on renal 25(OH)D3 metabolism, adenylate cyclase, and protein kinase. Am J Physiol 246:E102–E107. 64. Armbrecht HJ, Wongsurawat N, Zenser TV, Davis BB 1982 Differential effects of parathyroid hormone on the renal 1,25-dihydroxyvitamin D3 and 24,25-dihydroxyvitamin D3 production of young and adult rats. Endocrinology 111:1339–1344. 65. Horst RL, Eisman JA, Jorgensen NA, DeLuca HF 1977 Adequate response of plasma 1,25-dihydroxyvitamin D to parturition in paretic (milk fever) dairy cows. Science 196:662–663.
31 66. Hughes MR, Brumbaugh PF, Haussler MR, Wergedal JE, Baylink DJ 1975 Regulation of serum 1α,25-dihydroxyvitamin D3 by calcium and phosphate in the rat. Science 190:578–580. 67. Henry HL 1979 Regulation of the hydroxylation of 25-hydroxyvitamin D3 in vivo and in primary cultures of chick kidney cells. J Biol Chem 254:2722–2729. 68. Omdahl JL 1978 Interaction of the parathyroid and 1,25dihydroxyvitamin D3 in the control of renal 25-hydroxyvitamin D3 metabolism. J Biol Chem 253:8474–8487. 69. Henry HL 1981 Insulin permits parathyroid hormone stimulation of 1,25-dihydroxyvitamin D3 production in cultured kidney cells. Endocrinology 108:733–735. 70. Armbrecht HJ, Forte LR, Wongsurawat N, Zenser TV, Davis BB 1984 Forskolin increases 1,25-dihydroxyvitamin D3 production by rat renal slices in vitro. Endocrinology 114:644–649. 71. Aarskog D, Aksnes L 1980 Acute response of plasma 1,25-dihydroxyvitamin D to parathyroid hormone. Lancet 1(8164):362–363. 72. Hove K, Horst RL, Littledike ET, Beitz DC 1984 Infusions of parathyroid hormone in ruminants: Hypercalcemia and reduced plasma 1,25-dihydroxyvitamin D concentrations. Endocrinology 114:897–903. 73. Hustmyer FG, Beitz DC, Goff JP, Nonnecke BJ, Horst RL, Reinhardt TA 1994 Effects of in vivo administration of 1,25-dihydroxyvitamin D3 on in vitro proliferation of bovine lymphocytes. J Dairy Sci 77:3324–3330. 74. Reinhardt TA, Horst RL 1990 Parathyroid hormone down-regulates 1,25-dihydroxyvitamin D receptors (VDR) and VDR messenger ribonucleic acid in vitro and blocks homologous up-regulation of VDR in vivo. Endocrinology 127:942–948. 75. Henry HL, Midgett RJ, Norman AW 1974 Regulation of 25-hydroxyvitamin D3-1-hydroxylase in vivo. J Biol Chem 249:7584–7592. 76. Tanaka Y, DeLuca HF 1973 The control of 25-hydroxyvitamin D metabolism by inorganic phosphorus. Arch Biochem Biophys 154:566–574. 77. Gray RW 1987 Evidence that somatomedins mediate the effect of hypophosphatemia to increase serum 1,25-dihydroxyvitamin D3 levels in rats. Endocrinology 121:504–512. 78. Armbrecht HJ, Zenser TV, Davis BB 1982 Modulation of renal production of 24,25- and 1,25-dihydroxyvitamin D3 in young and adult rats by dietary calcium, phosphorus, and 1,25-dihydroxyvitamin D3. Endocrinology 110: 1983–1988. 79. Brenza HL, DeLuca HF 2000 Regulation of 25-hydroxyvitamin D3 1α-hydroxylase gene expression by parathyroid hormone and 1,25-dihydroxyvitamin D3. Arch Biochem Biophys 381:143–152. 80. Slatopolsky E, Lopez-Hilker S, Delmez J, Dusso A, Brown A, Martin KJ 1990 The parathyroid-calcitriol axis in health and chronic renal failure. Kidney Int 38:S41–S47. 81. Beckman MJ, Johnson JA, Goff JP, Reinhardt TA, Beitz DC, Horst RL 1995 The role of dietary calcium in the physiology of vitamin D toxicity: Excess dietary vitamin D3 blunts parathyroid hormone induction of kidney 1-hydroxylase. Arch Biochem Biophys 319:535–539. 82. Leheste JR, Melsen F, Wellner M, Jansen P, Schlichting U, Renner-Muller I, Andreassen TT, Wolf E, Bachmann S, Nykjaer A, Willnow TE 2003 Hypocalcemia and osteopathy in mice with kidney-specific megalin gene defect. FASEB J 17:247–249.
32
RONALD L. HORST, TIMOTHY A. REINHARDT, AND G. SATYANARAYANA REDDY
83. Kumar R 1984 Metabolism of 1,25-dihydroxyvitamin D3. Physiol Rev 64:478–504. 84. Henry HL 1992 Vitamin D hydroxylases. J Cell Biochem 49:4–9. 85. Portale AA, Miller WL 2000 Human 25-hydroxyvitamin D-1α-hydroxylase: cloning, mutations, and gene expression. Pediatr Nephrol 14:620–625. 86. Fu GK, Portale AA, Miller WL 1997 Complete structure of the human gene for the vitamin D 1α-hydroxylase, P450c1α. DNA Cell Biol 16:1499–1507. 87. Glorieux FH, St-Arnaud R 1998 Molecular cloning of (25-OH D)-1α-hydroxylase: an approach to the understanding of vitamin D pseudo-deficiency. Recent Prog Horm Res 53:341–349; discussion 350. 88. Shinki T 1998 Cloning and expression of 25-hydroxyvitamin D3-1α-hydroxylase gene. Nippon Rinsho 56:1440–1446. 89. Yoshida T, Yoshida N, Nakamura A, Monkawa T, Hayashi M, Saruta T 1999 Cloning of porcine 25-hydroxyvitamin D3 1α-hydroxylase and its regulation by cAMP in LLC-PK1 cells. J Am Soc Nephrol 10:963–970. 90. Panda DK, Miao D, Tremblay ML, Sirois J, Farookhi R, Hendy GN, Goltzman D 2001 Targeted ablation of the 25-hydroxyvitamin D 1α-hydroxylase enzyme: evidence for skeletal, reproductive, and immune dysfunction. Proc Natl Acad Sci USA 98:7498–7503. 91. Dardenne O, Prud’homme J, Arabian A, Glorieux FH, St-Arnaud R 2001 Targeted inactivation of the 25-hydroxyvitamin D3-1α-hydroxylase gene (CYP27B1) creates an animal model of pseudovitamin D-deficiency rickets. Endocrinology 142:3135–3141. 92. Holick MF, Schnoes HK, DeLuca HF, Gray RW, Boyle IT, Suda T 1972 Isolation and identification of 24,25-dihydroxycholecalciferol, a metabolite of vitamin D3 made in the kidney. Biochemistry 11:4251–4255. 93. Kumar R, Schnoes HK, DeLuca HF 1978 Rat intestinal 25-hydroxyvitamin D3- and 1α,25-dihydroxyvitamin D324-hydroxylase. J Biol Chem 253:3804–3809. 94. Kumar R, Harnden D, DeLuca HF 1976 Metabolism of 1,25-dihydroxyvitamin D3: Evidence for side-chain oxidation. Biochemistry 15:2420–2423. 95. Iwata K, Yamamoto A, Satoh S, Ohyama Y, Tashiro Y, Setoguchi T 1995 Quantitative immunoelectron microscopic analysis of the localization and induction of 25-hydroxyvitamin D3 24-hydroxylase in rat kidney. J Histochem Cytochem 43:255–262. 96. Shinki T, Jin CH, Nishimura A, Nagai Y, Ohyama Y, Noshiro M, Okuda K, Suda T 1992 Parathyroid hormone inhibits 25-hydroxyvitamin D3-24-hydroxylase mRNA expression stimulated by 1α,25-dihydroxyvitamin D3 in rat kidney but not in intestine. J Biol Chem 267:13757–13762. 97. Nishimura A, Shinki T, Jin CH, Ohyama Y, Noshiro M, Okuda K, Suda T 1994 Regulation of messenger ribonucleic acid expression of 1α,25-dihydroxyvitamin D3-24-hydroxylase in rat osteoblasts. Endocrinology 134:1794–1799. 98. Shigematsu T, Horiuchi N, Ogura Y, Miyahara T, Suda T 1986 Human parathyroid hormone inhibits renal 24-hydroxylase activity of 25-hydroxyvitamin D3 by a mechanism involving adenosine 3′,5′-monophosphate in rats. Endocrinology 118: 1583–1589. 99. Goff JP, Reinhardt TA, Beckman MJ, Horst RL 1990 Contrasting effects of exogenous 1,25-dihydroxyvitamin D [1,25-(OH)2D] versus endogenous 1,25-(OH)2D, induced by dietary calcium restriction, on vitamin D receptors. Endocrinology 126:1031–1035.
100. Iida K, Shinki T, Yamaguchi A, DeLuca HF, Kurokawa K, Suda T 1995 A possible role of vitamin D receptors in regulating vitamin D activation in the kidney. Proc Natl Acad Sci USA 92:6112–6116. 101. Hanai H, Liang CT, Cheng L, Sacktor B 1989 Desensitization to parathyroid hormone in renal cells from aged rats is associated with alterations in G-protein activity. J Clin Invest 83:268–277. 102. Hanai H, Brennan DP, Cheng L, Goldman ME, Chorev M, Levine MA, Sacktor B, Liang CT 1990 Downregulation of parathyroid hormone receptors in renal membranes from aged rats. Am J Physiol 259:F444–F450. 103. Johnson JA, Beckman MJ, Pansini-Porta A, Christakos S, Bruns ME, Beitz DC, Horst RL, Reinhardt TA 1995 Age and gender effects on 1,25-dihydroxyvitamin D3-regulated gene expression. Exp Gerontol 30:631–643. 104. Matkovits T, Christakos S 1995 Variable in vivo regulation of rat vitamin D-dependent genes (osteopontin, Ca,Mg-adenosine triphosphatase, and 25-hydroxyvitamin D3 24-hydroxylase): Implications for differing mechanisms of regulation and involvement of multiple factors. Endocrinology 136: 3971–3982. 105. Goff JP, Reinhardt TA, Engstrom GW, Horst RL 1992 Effect of dietary calcium or phosphorus restriction and 1,25-dihydroxyvitamin D administration on rat intestinal 24-hydroxylase. Endocrinology 131:101–104. 106. Beckman MJ, Goff JP, Reinhardt TA, Beitz DC, Horst RL 1994 In vivo regulation of rat intestinal 24-hydroxylase: Potential new role of calcitonin. Endocrinology 135: 1951–1955. 107. Ohyama Y, Noshiro M, Okuda K 1991 Cloning and expression of cDNA encoding 25-hydroxyvitamin D3 24-hydroxylase. FEBS Lett 278:195–198. 108. Ohyama Y, Hayashi S, Okuda K 1989 Purification of 25-hydroxyvitamin D3 24-hydroxylase from rat kidney mitochondria. FEBS Lett 255:405–408. 109. Chen K-S, DeLuca HF 1995 Cloning of the human 1α,25dihydroxyvitamin D-3 24-hydroxylase gene promoter and identification of two vitamin D-responsive elements. Biochim Biophys Acta 1263:1–9. 110. Ohyama Y, Noshiro M, Eggertsen G, Gotoh O, Kato Y, Bjorkhem I, Okuda K 1993 Structural characterization of the gene encoding rat 25-hydroxyvitamin D3 24-hydroxylase. Biochemistry 32:76–82. 111. Horst RL, Littledike ET, Gray RW, Napoli JL 1981 Impaired 24,25-dihydroxyvitamin D production in anephric human and pig. J Clin Invest 67:274–280. 112. Taniguchi T, Eto TA, Shiotsuki H, Sueta H, Higashi S, Iwamura T, Okuda KI, Setoguchi T 2001 Newly established assay method for 25-hydroxyvitamin D3 24-hydroxylase revealed much lower Km for 25-hydroxyvitamin D3 than for 1α,25-dihydroxyvitamin D3. J Bone Miner Res 16:57–62. 113. Jones G, Tenenhouse HS 2002 1,25(OH)2D, the preferred substrate for CYP24. J Bone Miner Res 17:179–181. 114. Burgos-Trinidad M, DeLuca HF 1991 Kinetic properties of 25-hydroxyvitamin D- and 1,25-dihydroxyvitamin D-24hydroxylase from chick kidney. Biochim Biophys Acta 1078:226–230. 115. Napoli JL, Pramanik BC, Royal PM, Reinhardt TA, Horst RL 1983 Intestinal synthesis of 24-keto-1,25-dihydroxyvitamin D3. A metabolite formed in vivo with high affinity for the vitamin D cytosolic receptor. J Biol Chem 258:9100–9107. 116. Napoli JL, Horst RL 1983 C(24)- and C(23)-oxidation, converging pathways of intestinal 1,25-dihydroxyvitamin D3
33
CHAPTER 2 Vitamin D Metabolism
117. 118. 119. 120.
121.
122.
123.
124.
125. 126.
127.
128. 129.
130.
131.
132.
metabolism: Identification of 24-keto-1,23,25-trihydroxyvitamin D3. Biochemistry 22:5848–5853. Haussler MR 1986 Vitamin D receptors: Nature and function. Annu Rev Nutr 6:527–562. Lohnes D, Jones G 1987 Side chain metabolism of vitamin D3 in osteosarcoma cell line UMR-106. Characterization of products. J Biol Chem 262:14394–14401. Reddy GS, Tserng KY 1989 Calcitroic acid, end product of renal metabolism of 1,25-dihydroxyvitamin D3 through C-24 oxidation pathway. Biochemistry 28:1763–1769. Makin G, Lohnes D, Byford V, Ray R, Jones G 1989 Target cell metabolism of 1,25-dihydroxyvitamin D3 to calcitroic acid: Evidence for a pathway in kidney and bone involving 24-oxidation. Biochem J 262:173–180. Akiyoshi-Shibata M, Sakaki Y, Ohyama Y, Noshiro M, Okuda K, Yabusaki Y 1994 Further oxidation of hydroxycalcidiol by calcidiol 24-hydroxylase. Eur J Biochem 224:335–343. Sakaki T, Sawada N, Komai K, Shiozawa S, Yamada S, Yamamoto K, Ohyama Y, Inouye K 2000 Dual metabolic pathway of 25-hydroxyvitamin D3 catalyzed by human CYP24. Eur J Biochem 267:6158–6165. Sakaki T, Sawada N, Nonaka Y, Ohyama Y, Inouye K 1999 Metabolic studies using recombinant Escherichia coli cells producing rat mitochondrial CYP24. CYP24 can convert 1α,25-dihydroxyvitamin D3 to calcitroic acid. Eur J Biochem 262:43–48. Inouye K, Sakaki T 2001 Enzymatic studies on the key enzymes of vitamin D metabolism; 1α-hydroxylase (CYP27B1) and 24-hydroxylase (CYP24). Biotechnol Annu Rev 7:179–194. Beckman MJ, Tadikonda P, Werner E, Prahl J, Yamada S, DeLuca HF 1996 Human 25-hydroxyvitamin D3-24-hydroxylase, a multicatalytic enzyme. Biochemistry 35:8465–8472. Pols HAP, van Leeuwen JPTM, Schilte JP, Visser TJ, Birkenhager JC 1988 Heterologous up-regulation of the 1,25-dihydroxyvitamin D3 receptor by parathyroid hormone (PTH) and PTH-like peptide in osteoblast-like cells. Biochem Biophys Res Commun 156:588–594. Pols HAP, Birkenhager JC, Schilte JP, Visser TJ 1988 Evidence that the self-induced metabolism of 1,25-dihydroxyvitamin D3 limits the homologous up-regulation of its receptor in rat osteosarcoma cells. Biochim Biophys Acta 970:122–129. Reinhardt TA, Horst RL 1989 Self-induction of 1,25-dihydroxyvitamin D3 metabolism limits receptor occupancy and target tissue responsiveness. J Biol Chem 264:15917–15921. Reinhardt TA, Horst RL 1989 Ketoconazole inhibits selfinduced metabolism of 1,25-dihydroxyvitamin D3 and amplifies 1,25-dihydroxyvitamin D3 receptor up-regulation in rat osteosarcoma cells. Arch Biochem Biophys 272:459–465. Reinhardt TA, Horst RL, Engstrom GW, Atkins KB 1988 Ketoconazole potentiates 1,25-(OH)2D3-directed upregulation of 1,25-(OH)2D3-receptors in rat intestine and bone. In: Norman AW, Schaefer K, Grigoleit H-G, Herrath DV (eds) Vitamin D: Molecular, Cellular and Clinical Endocrinology. Walter de Gruyter, Berlin, pp. 233–234. Hosogane N, Shinki T, Kasuga H, Taketomi S, Toyama Y, Suda T 2003 Mechanisms for the reduction of 24,25-dihydroxyvitamin D3 levels and bone mass in 24-hydroxylase transgenic rats. FASEB J 17:737–739. Kasuga H, Hosogane N, Matsuoka K, Mori I, Sakura Y, Shimakawa K, Shinki T, Suda T, Taketomi S 2002 Characterization of transgenic rats constitutively expressing
133.
134.
135.
136.
137.
138.
139. 140.
141.
142. 143.
144.
145. 146.
147.
vitamin D-24-hydroxylase gene. Biochem Biophys Res Commun 297:1332–1338. Wichmann J, Schnoes HK, DeLuca HF 1981 Isolation and identification of 24(R)-hydroxyvitamin D3 from chicks given large doses of vitamin D3. Biochemistry 20: 2350–2353. Wichmann JK, Schnoes HK, DeLuca HF 1981 23,24, 25-Trihydroxyvitamin D3, 24,25,26-trihydroxyvitamin D3, 24-keto-25-hydroxyvitamin D3, and 23-dehydro-25-hydroxyvitamin D3: New in vivo metabolites of vitamin D3. Biochemistry 20:7385–7391. DeLuca HF, Schnoes HK 1979 Recent developments in the metabolism of vitamin D3. In: Norman AW (ed) Vitamin D: Basic Research and Its Clinical Application. Walter de Gruyter, Berlin, pp. 445–458. Horst RL 1979 25-OHD3-26,23-lactone: A metabolite of vitamin D3 that is 5 times more potent than 25-OHD3 in the rat plasma competitive protein binding radioassay. Biochem Biophys Res Commun 89:286–293. Wichmann JK, DeLuca HF, Schnoes HK, Horst RL, Shepard RM, Jorgensen NA 1979 25-Hydroxyvitamin D3 26,23lactone: A new in vivo metabolite of vitamin D. Biochemistry 18:4775–4780. Ohnuma N, Norman AW 1982 Production in vitro of 1α,25dihydroxyvitamin D3-26,23-lactone from 1α,25-dihydroxyvitamin D3 by rat small intestinal mucosa homogenates. Arch Biochem Biophys 213:139–147. Napoli JL, Horst RL 1983 (23S)-1,23,25-Trihydroxycholecalciferol, an intestinal metabolite of 1,25-dihydroxycholecalciferol. Biochem J 214:261–264. Napoli JL, Pramanik BC, Partridge JJ, Uskokovic MR, Horst RL 1982 23S,25-Dihydroxyvitamin D3 as a circulating metabolite of vitamin D3. Its role in 25-hydroxyvitamin D326,23-lactone biosynthesis. J Biol Chem 257:9634–9639. Horst RL, Littledike ET 1982 Comparison of plasma concentrations of vitamin D and its metabolites in young and aged domestic animals. Comp Biochem Physiol 73B: 485–489. Horst RL, Littledike ET 1980 25-OHD3-26,23 lactone: Demonstration of kidney-dependent synthesis in the pig and rat. Biochem Biophys Res Commun 93:149–154. Ishizuka S, Ohba T, Norman AW 1988 1,25-(OH)2D3-26,23lactone is a major metabolite of 1,25-(OH)2D3 under physiological conditions. In: Norman AW, Schafer K, Grigoleit H-G, Herrath DV (eds) Vitamin D: Molecular, Cellular and Clinical Endocrinology. Walter de Gruyter, Berlin, pp. 143–144. Ishizuka S, Ishimoto S, Norman AW 1984 Isolation and identification of 1α,25-dihydroxy-24-oxovitamin D3, 1α,25dihydroxyvitamin D3 26,23-lactone, and 1α,24(S),25-trihydroxyvitamin D3: In vivo metabolites of 1α,25-dihydroxyvitamin D3. Biochemistry 23:1473–1478. Napoli JL, Horst RL 1982 23,25,26-Trihydroxycholecalciferol. A 25-hydroxycholecalciferol-26,23-lactone precursor. Biochem J 206:173–176. Yamada S, Nakayama K, Takayama H, Shinki T, Takasaki Y, Suda T 1984 Isolation, identification, and metabolism of (23S,25R)-25-hydroxyvitamin D3 26,23-lactol. A biosynthetic precursor of (23S,25R)-25-hydroxyvitamin D3 26,23-lactone. J Biol Chem 259:884–889. Horst RL, Pramanik BC, Reinhardt TA, Shiuey SJ, Partridge JJ, Uskokovic MR, Napoli JL 1982 Binding properties of 23S,25dihydroxyvitamin D3: An in vivo metabolite of vitamin D3. Biochem Biophys Res Commun 106:1006–1011.
34
RONALD L. HORST, TIMOTHY A. REINHARDT, AND G. SATYANARAYANA REDDY
148. Horst RL, Wovkulich PM, Baggiolini EG, Uskokovic MR, Engstrom GW, Napoli JL 1984 (23S)-1,23,25-Trihydroxyvitamin D3: Its biologic activity and role in 1,25-dihydroxyvitamin D3 26,23-lactone biosynthesis. Biochemistry 23:3973–3979. 149. Esvelt RP, Schnoes HK, DeLuca HF 1979 Isolation and characterization of 1α-hydroxy-23-carboxytetranorvitamin D: A major metabolite of 1,25-dihydroxyvitamin D3. Biochemistry 18:3977–3983. 150. Horst RL, Reinhardt TA, Pramanik BC, Napoli JL 1983 23-Keto-25-hydroxyvitamin D3: A vitamin D3 metabolite with high affinity for the 1,25-dihydroxyvitamin D specific cytosol receptor. Biochemistry 22:245–250. 151. Horst RL, Reinhardt TA, Napoli JL 1982 23-Keto-25hydroxyvitamin D3 and 23-keto-1,25-dihydroxyvitamin D3: Two new vitamin D3 metabolites with high affinity for the 1,25-dihydroxyvitamin D3 receptor. Biochem Biophys Res Commun 107:1319–1325. 152. Suda T, DeLuca HF, Schnoes HK, Tanaka Y, Holick MF 1970 25,26-Dihydroxycholecalciferol, a metabolite of vitamin D3 with intestinal calcium transport activity. Biochemistry 9:4776–4780. 153. Reinhardt TA, Napoli JL, Pramanik B, Littledike ET, Beitz DC, Partridge JJ, Uskokovic MR, Horst RL 1981 1α,25,26-Trihydroxyvitamin D3: An in vivo and in vitro metabolite of vitamin D3. Biochemistry 20:6230–6235. 154. Partridge JJ, Shiuey SJ, Chadha HK, Baggiolini ET, Bolount JE, Uskokovic MR 1981 Synthesis and structure proof of a vitamin D3 metabolite, 25(S),26-dihydroxycholecalciferol. J Am Chem Soc 103:1253–1255. 155. Ikekawa N, Koizumi N, Ohshima E, Ishizuka S, Takeshita T, Tanaka Y, DeLuca HF 1983 Natural 25,26-dihydroxyvitamin D3 is an epimeric mixture. Proc Natl Acad Sci USA 80:5286–5288. 156. Hollis BW, Roos BA, Lambert PW 1980 25,26Dihydroxycholecalciferol: A precursor in the renal synthesis of 25-hydroxycholecalciferol-26,23-lactone. Biochem Biophys Res Commun 95:520–528. 157. Napoli JL, Horst RL 1981 25,26-Dihydroxyvitamin D3 is not a major intermediate in 25-hydroxyvitamin D3-26,23-lactone formation. Arch Biochem Biophys 212:754–758. 158. Tanaka Y, DeLuca HF, Schnoes HK, Ikekawa N, Eguchi T 1981 23,25-Dihydroxyvitamin D3: A natural precursor in the biosynthesis of 25-hydroxyvitamin D3-26,23-lactone. Proc Natl Acad Sci USA 78:4805–4808. 159. Reddy GS, Muralidharan KR, Okamura WH, Tserng KY, McLane JA 2001 Metabolism of 1α,25-dihydroxyvitamin D3 and its C-3 epimer 1α,25-dihydroxy-3-epi-vitamin D3 in neonatal human keratinocytes. Steroids 66:441–450. 160. Siu-Caldera ML, Sekimoto H, Weiskopf A, Vouros P, Muralidharan KR, Okamura WH, Bishop J, Norman AW, Uskokovic MR, Schuster I, Reddy GS 1999 Production of 1α,25-dihydroxy-3-epi-vitamin D3 in two rat osteosarcoma cell lines (UMR 106 and ROS 17/2.8): Existence of the C-3 epimerization pathway in ROS 17/2.8 cells in which the C-24 oxidation pathway is not expressed. Bone 24:457–463. 161. Horst RL, Shepard RM, Jorgensen NA, DeLuca HF 1979 The determination of 24,25-dihydroxyvitamin D and 25,26-dihydroxyvitamin D in plasma from normal and nephrectomized man. J Lab Clin Med 93:277–285. 162. Shepard RM, Horst RL, Hamstra AJ, DeLuca HF 1979 Determination of vitamin D and its metabolites in plasma from normal and anephric man. Biochem J 182: 55–69.
163. Bergman T, Postlind H 1991 Characterization of mitochondrial cytochromes P-450 from pig kidney and liver catalysing 26-hydroxylation of 25-hydroxyvitamin D3 C27 steroids. Biochem J 276:427–432. 164. Napoli JL, Okita RT, Masters BS, Horst RL 1981 Identification of 25,26-dihydroxyvitamin D3 as a rat renal 25-hydroxyvitamin D3 metabolite. Biochemistry 20:5865–5871. 165. Tanaka Y, Schnoes HK, Smith CM, DeLuca HF 1981 1,25,26-Trihydroxyvitamin D3 isolation, identification and biological activity. Arch Biochem Biophys 210: 104–109. 166. Astecker N, Reddy GS, Herzig G, Vorisek G, Schuster I 2000 1α,25-Dihydroxy-3-epi-vitamin D3 a physiological metabolite of 1α,25-dihydroxyvitamin D3: its production and metabolism in primary human keratinocytes. Mol Cell Endocrinol 170:91–101. 167. Masuda S, Kamao M, Schroeder NJ, Makin HL, Jones G, Kremer R, Rhim J, Okano T 2000 Characterization of 3-epi-1α,25-dihydroxyvitamin D3 involved in 1α,25-dihydroxyvitamin D3 metabolic pathway in cultured cell lines. Biol Pharm Bull 23:133–139. 168. Hylemon PB 1985 Metabolism of bile acids in intestinal microflora. In: Danielsson H, Sjoevall J (eds) Sterols and Bile Acids, Elsevier, pp. 331–344. 169. Penning TM, Bennett MJ, Smith-Hoog S, Schlegel BP, Jez JM, Lewis M 1997 Structure and function of 3α-hydroxysteroid dehydrogenase. Steroids 62:101–111. 170. Norman AW, Bouillon R, Farach-Carson MC, Bishop JE, Zhou LX, Nemere I, Zhao J, Muralidharan KR, Okamura WH 1993 Demonstration that 1β,25-dihydroxyvitamin D3 is an antagonist of the nongenomic but not genomic biological responses and biological profile of the three A-ring diastereomers of 1α,25-dihydroxyvitamin D3. J Biol Chem 268: 20022–20030. 171. Brown AJ, Ritter C, Slatopolsky E, Muralidharan KR, Okamura WH, Reddy GS 1999 1α,25-dihydroxy-3-epi-vitamin D3, a natural metabolite of 1α,25-dihydroxyvitamin D3, is a potent suppressor of parathyroid hormone secretion. J Cell Biochem 73:106–113. 172. Schuster I, Astecker N, HE, Hertig G, Reddy GS, Schmid J, Vorisek G 1997 Vitamin D metabolism in human keratinocytes and biological role of products. In: Norman AW, Bouillon R, Thomasset M (eds) Vitamin D, Chemistry, Biology and Clinical Applications of the Steroid Hormone. Walter de Gruyter, New York, pp. 551–558. 173. Engstrom GW, Koszewski NJ 1989 Metabolism of vitamin D2 in pig liver homogenates: Evidence for a free radical reaction. Arch Biochem Biophys 270:432–440. 174. Horst RL, Reinhardt TA, Ramberg CF, Koszewski NJ, Napoli JL 1986 24-Hydroxylation of 1,25-dihydroxyergocalciferol: An unambiguous deactivation process. J Biol Chem 261:9250–9256. 175. Zimmerman DR, Reinhardt TA, Kremer R, Beitz DC, Reddy GS, Horst RL 2001 Calcitroic acid is a major catabolic metabolite in the metabolism of 1α-dihydroxyvitamin D2. Arch Biochem Biophys 392:14–22. 176. Rao LS, Ray R, Holick MF, Horst RL, Uskokovic MR, Reddy GS 2000 Metabolism of [3α-3H] 25-hydroxyvitamin D2 in kidneys isolated from normal and vitamin D2-intoxicated rats. J Nutr Sci Vitaminol (Tokyo) 46:222–229. 177. Horst RL, Omdahl JA, Reddy S 2003 Rat cytochrome P450C24 (CYP24) does not metabolize 1,25-dihydroxyvitamin D2 to calcitroic acid. J Cell Biochem 88: 282–285.
CHAPTER 2 Vitamin D Metabolism
178. Sunita Rao D, Balkundi D, Uskokovic MR, Tserng K, Clark JW, Horst RL, Satyanarayana Reddy G 2001 Double bond in the side chain of 1α,25-dihydroxy-22-ene-vitamin D3 is reduced during its metabolism: studies in chronic myeloid leukemia (RWLeu-4) cells and rat kidney. J Steroid Biochem Mol Biol 78:167–176. 179. Byford V, Strugnell S, Coldwell R, Schroeder N, Makin HL, Knutson JC, Bishop CW, Jones G 2002 Use of vitamin D4 analogs to investigate differences in hepatic and target cell metabolism of vitamins D2 and D3. Biochim Biophys Acta 1583:151–166. 180. Tachibana Y, Tsuji M 2001 Study on the metabolites of 1α,25-dihydroxyvitamin D4. Steroids 66:93–97. 181. Williams DH, Morris DS, Gilhooly MA, Norris AF 1982 Synthesis of 25(R)-hydroxycholecalciferol-26,23(S)-lactone and of 25(S),26-dihydroxyergocalciferol. In: Norman AW, Schaefer K, Herrath DV, Grigoleit H-G (eds) Vitamin D: Chemical, Biochemical and Clinical Endocrinology of Calcium Metabolism. Walter de Gruyter, Berlin, pp. 1067–1072. 182. Mazur Y, Segev D, Jones G 1982 Synthesis and determination of absolute configuration of kidney metabolites of vitamin D2. In: Norman AW, Schaefer K, Herrath DV, Grigoleit H-G (eds) Vitamin D: Chemical, Biochemical and Clinical Endocrinology of Calcium Metabolism. Walter de Gruyter, Berlin, pp. 1101–1106. 183. Hay AWM, Jones G 1979 The elution profile of vitamin D2 metabolites from Sephadex LH-20 columns. Clin Chem 25:473–475. 184. Reddy GS, Tserng KY 1990 24,25,28-Trihydroxyvitamin D2 and 24,25,26-trihydroxyvitamin D2: Novel metabolites of vitamin D2. Biochemistry 29:943–949. 185. Reddy GS, Tserng KY 1986 Isolation and identification of 1,24,25-trihydroxyvitamin D2, 1,24,25,28-tetrahydroxyvitamin D2, and 1,24,25,26-tetrahydroxyvitamin D2: New metabolites of 1,25-dihydroxyvitamin D2 produced in rat kidney. Biochemistry 25:5328–5336. 186. Jones G, Byford V, Makin H, Kremer R, Rice R, deGraffenried LA, Knutson JC, Bishop CW 1996 Antiproliferative activity and target cell catabolism of the vitamin D analog 1α24(S)-(OH)2D2 in normal and immortalized human epidermal cells. Biochem Pharmacol 52:133–140. 187. Rao DS, Siu-Caldera ML, Uskokovic MR, Horst RL, Reddy GS 1999 Physiological significance of C-28 hydroxylation in the metabolism of 1α,25-dihydroxyvitamin D2. Arch Biochem Biophys 368:319–328. 188. Kreitmeir H, Moll T 1928. Hypervitaminose durch grosse Dosen Vitamin D Munch. Med. Wochenschr 75:637–639. 189. Putscher W 1929 Über Vigantolschädigung der Niere einum Kinde. Z Kinderheilkd 48:269. 190. Holick MF, MacLaughlin JA, Doppelt SH 1981 Regulation of cutaneous previtamin D3 photosynthesis in man: Skin pigment is not an essential regulator. Science 211:590–593. 191. Clark MB, Potts JT 1977 25-Hydroxyvitamin D3 regulation. Calcif Tissue Res 22(Supplement):29–34. 192. Shepard RM, DeLuca HF 1980 Plasma concentrations of vitamin D3 and its metabolites in the rat as influenced by vitamin D3 or 25-hydroxyvitamin D3 intakes. Arch Biochem Biophys 202:43–53. 193. Hughes MR, Baylink DJ, Jones PG, Haussler MR 1976 Radioligand receptor assay for 25-hydroxyvitamin D2/D3 and 1α,25-dihydroxyvitamin D2/D3: Application to hypervitaminosis D. J Clin Invest 58:61–70. 194. Counts SJ, Baylink DJ, Shen FH, Sherrard DJ, Hickman RO 1975 Vitamin D intoxication in an anephric child. Ann Intern Med 82:196–200.
35 195. Littledike ET, Horst RL 1982 Vitamin D3 toxicity in dairy cows. J Dairy Sci 65:749–759. 196. Reinhardt TA, Conrad HR 1980 Mode of action of pharmacological doses of cholecalciferol during parturient hypocalcemia in dairy cows. J Nutr 110:1589–1594. 197. Horst RL, Reinhardt TA 1983 Vitamin D metabolism in ruminants and its relevance to the periparturient cow. J Dairy Sci 66:661–678. 198. Hunt RD, Garcia FG, Walsh RJ 1972 A comparison of the toxicity of ergocalciferol and cholecalciferol in Rhesus monkeys (Masaca mulatta). J Nutr 102:975–986. 199. Harrington DD, Page EH 1983 Acute vitamin D3 toxicosis in horses: Case reports and experimental studies of the comparative toxicity of vitamins D2 and D3. J Am Vet Med Assoc 182:1358–1369. 200. Hodson EM, Evans RA, Dunstan CR, Hills E, Wong SYP, Rosenberg AR, Roy LP 1985 Treatment of childhood renal osteodystrophy with calcitriol or ergocalciferol. Clin Nephrol 24:192–200. 201. Sjoden G, Smith C, Lingren U, DeLuca HF 1985 1α-hydroxyvitamin D2 is less toxic than 1α-hydroxyvitamin D3 in the rat. Proc Soc Exp Biol Med 178:432–436. 202. Beckman MJ, Horst RL, Reinhardt TA, Beitz DC 1990 Up-regulation of the intestinal 1,25-dihydroxyvitamin D receptor during hypervitaminosis D: A comparison between vitamin D2 and vitamin D3. Biochem Biophys Res Commun 169:910–915. 203. Mortensen JT, Brinck P, Binderup L 1993 Toxicity of vitamin D analogues in rats fed diets with standard or low calcium contents. Pharmacol Toxicol 72:124–127. 204. Hines TG, Jacobson NL, Beitz DC, Littledike ET 1985 Dietary calcium and vitamin D: Risk factors in the development of atherosclerosis in young goats. J Nutr 115: 167–178. 205. Payne JM, Manston R 1967 The safety of massive doses of vitamin D3 in the prevention of milk fever. Vet Rec 81:214–216. 206. Hibbs JW, Pounden WD 1955 Studies on milk fever in dairy cows. IV. Prevention by short-time, prepartum feeding of massive doses of vitamin D. J Dairy Sci 38:65–72. 207. Napoli JL, Sommerfeldt JL, Pramanik BC, Gardner R, Sherry AD, Partridge JJ, Uskokovic MR, Horst RL 1983 19-Nor-10-ketovitamin D derivatives: Unique metabolites of vitamin D3, vitamin D2, and 25-hydroxyvitamin D3. Biochemistry 22:3636–3640. 208. Recker RR, Schoenfeld JL, Slatopolsky R, Goldsmith R, Brickman A 1979 25-Hydroxyvitamin D (calcidiol) on renal osteodystophy: Long term results of a six center trial. In: Norman AW, Schaefer K, Herrath DV, Grigoleit HG, Coburn JW, DeLuca HF, Mawer EB, Suda T (eds) Vitamin D: Basic Research and Its Clinical Application. Walter de Gruyter, Berlin, pp. 869–875. 209. West TET, Joffe M, Sinclair L, O’Riordan JLH 1971 Treatment of hypercalcemia with calcitonin. Lancet, 675–678. 210. Ulbrych-Jablonska A 1972 The hypocalcemic effect of glucagon in cases of hypercalcemia. Helv Paediatr Acta 27:613–615. 211. Kingma JG, Roy PE 1990 Effect of ethane-I-hydroxyl, I-diphosphonate on arterial calcinosis induced by hypervitaminosis D: a morphologic investigation. J Exp Path (Oxford) 71:145–153. 212. Streck WF, Waterhouse C, Haddad JG 1979 Glucocorticoid effects in vitamin D intoxication. Arch Intern Med 139: 974–977.
36
RONALD L. HORST, TIMOTHY A. REINHARDT, AND G. SATYANARAYANA REDDY
213. Buffenstein R, Skinner DC, Yahav S, Moodley GP, Cavaleros M, Zachen D, Ross FP, Pettifor JM 1991 Effect of oral cholecalciferol supplementation at physiological and supraphysiological doses in naturally vitamin D3-deficient subterranean damara mole rats (Cryptomys damarensis). J Endocrinol 131:197–202. 214. Cheeke PR, Amberg JW 1973 Comparative calcium excretion by rats and rabbits. J Anim Sci 37:450–454. 215. Schryver HF, Hintz HF, Lowe JE 1974 Calcium and phosphorus in the nutrition of the horse. Cornell Vet 64:493–515. 216. Bourdeau JE, Schwer-Dymerski DA, Stern PH, Langman CB 1986 Calcium and phosphorus metabolism in chronically vitamin D-deficient laboratory rabbits. Miner Electrolyte Metab 12:176–185. 217. Sommerfeldt JL, Horst RL, Littledike ET, Beitz DB 1979 In vitro degradation of cholecalciferol in rumen fluid. J Dairy Sci 62(Supplement 1):192. 218. Sommerfeldt JL, Napoli JL, Littledike ET, Beitz DC, Horst RL 1983 Metabolism of orally administered [3H]ergocalciferol and [3H]cholecalciferol by dairy calves. J Nutr 113:2595–2600. 219. Stern PH, Horst RL, Gardner R, Napoli JL 1985 10-Keto or 25-hydroxy substitution confer equivalent in vitro boneresorbing activity to vitamin D3. Arch Biochem Biophys 236:555–558. 220. Hunt RD, Garcia FG, Hegsted DM, Kaplinsky N 1967 Vitamin D2 and vitamin D3 in New World primates: Influence on calcium absorptions. Science 158:943–945. 221. Steenbock H, Kletzien SWF, Haplin JG 1932 Reaction of chicken to irradiated ergosterol and irradiated yeast as contrasted with natural vitamin D of fish liver oils. J Biol Chem 97:249–264. 222. Horst RL, Napoli JL, Littledike ET 1982 Discrimination in the metabolism of orally dosed ergocalciferol and cholecalciferol by the pig, rat and chick. Biochem J 204:185–189. 223. Sebert JL, Garabedian M, deChasteigner C, Defrance D 1991 Comparative effects of equal doses of vitamin D2 and vitamin D3 for the correction of vitamin D deficiency in the elderly. In: Norman AW, Bouillon R, Thomasset M (eds) Vitamin D: Gene Regulation, Structure–Function Analysis and Clinical Application. Walter de Gruyter, Berlin, pp. 765–766. 224. Reddy GS, Ray R, Holick MF 1990 Serum 1,25-dihydroxyvitamin D2 levels are two times higher than 1,25-dihydroxyvitamin D3 levels in vitamin D-deficient rats, dosed acutely with equal amounts of vitamin D2 and vitamin D3. J Bone Miner Res 5(Supplement 2):S265. 225. Haddad JG, Walgate J 1976 25-Hydroxyvitamin D transport in human plasma. J Biol Chem 251:4803–4809. 226. Baird DT, Horton R, Longcope C, Tait JF 1969 Steroid hormone dynamics under steady state conditions. Recent Prog Horm Res 25:611–664. 227. Hoy DA, Ramberg CF Jr., Horst RL 1988 Evidence that discrimination against ergocalciferol by the chick is the
228. 229. 230.
231.
232.
233.
234.
235.
236.
237.
238.
result of enhanced metabolic clearance rates for its monoand dihydroxylated metabolites. J Nutr 118:633–638. Hay AWM, Watson G 1977 Vitamin D2 in vertebrate evolution. Comp Biochem Physiol 56B:375–380. Horst RL, Reinhardt TA, Beitz DC, Littledike ET 1981 A sensitive competitive protein binding assay for vitamin D in plasma. Steroids 37:581–592. Reddy GS, Napoli JL, Hollis BW 1984 Preferential 25-hydroxylation of vitamin D2 vs vitamin D3 by rat liver is dependent upon vitamin D-binding protein. Calcif Tissue Int 36:524 (Abstract). Arnaud SB, Young DR, Cann C, Reinhardt TA, Henrickson R 1985 Is hypervitaminosis D normal in the rhesus monkey? In: Norman AW, Schaefer K, Grigoleit H-G, Herrath DV (eds) Vitamin D. Chemical, Biochemical and Clinical Update. Walter de Gruyter, Berlin, pp. 585–586. Shinki T, Shiina Y, Takahashi N, Tanioka Y, Koizumi H, Suda T 1983 Extremely high circulating levels of 1α,25dihydroxyvitamin D3 in the marmoset, a New World monkey. Biochem Biophys Res Commun 114:452–457. O’Loughlin PD, Morris HA 1991 Duodenal transport of calcium in the marmoset. In: Norman AW, Bouillon R, Thomasset M (eds) Vitamin D. Gene Regulation, Structure– Function Analysis and Clinical Application. Walter de Gruyter, Berlin, pp. 345–346. Shore RF, Hayes ME, Balment RJ, Mawer EB 1988 Serum 25(OH)D3 and 1,25(OH)2D3 levels in wild and laboratorybred wood mice and bank voles. In: Norman AW, Schaefer K, Grigoleit H-G, Herrath DV (eds) Vitamin D. Molecular, Cellular and Clinical Endocrinology. Walter de Gruyter, Berlin, pp. 633–634. Kobayashi T, Takeuchi A, Okano T 1991 An evolutionary aspect in vertebrates from the viewpoint of vitamin D3 metabolism. In: Norman AW, Bouillon R, Thomasset M (eds) Vitamin D. Gene Regulation, Structure–Function Analysis and Clinical Application. Walter de Gruyter, Berlin, pp. 679–680. Glowacki J, Deftos LJ, Mayer E, Norman AW, Henry H 1982 Chondrichthyes cannot resorb implanted bone and have calcium-regulating hormones. In: Norman AW, Schaefer K, Herrath DV, Grigoleit H-G (eds) Vitamin D. Chemical, Biochemical and Clinical Endocrinology of Calcium Metabolism. Walter de Gruyter, Berlin, pp. 613–615. Sundell K, Bishop JE, Björnsson BT, Horst RL, Hollis BW, Norman AW 1991 Organ distribution of a high affinity receptor for 1,25(OH)2 vitamin D3 in the Atlantic cod (Gadus morhua), a marine teleost. In: Norman AW, Bouillon R, Thomasset M (eds) Vitamin D. Gene Regulation, Structure– Function Analysis and Clinical Application. Walter de Gruyter, Berlin, pp. 114–115. Maenpaa PH, Koskinen T, Koskinen E 1988 Serum profiles of vitamins A, E and D in mares and foals during different seasons. J Anim Sci 66:1418–1423.
CHAPTER 3
Photobiology of Vitamin D MICHAEL F. HOLICK
I. II. III. IV.
Vitamin D, Skin, and Bone Research Laboratory; Department of Medicine; Endocrinology, Nutrition and Diabetes Section; Boston Medical Center and Boston University School of Medicine, Boston, Massachusetts
Introduction Historical Perspective Photobiology of Vitamin D Role of Sunlight and Dietary Vitamin D in Bone Health, Overall Health, and Well-Being
I. INTRODUCTION Vitamin D is neither a vitamin nor a hormone when adequate exposure to sunlight is available to promote the synthesis of vitamin D in the skin [1]. First produced in ocean-dwelling phytoplankton and zooplankton, vitamin D has been made by life forms on earth for almost 1 billion years [2]. Although the physiological function of vitamin D in these lower life forms is uncertain, it is well recognized that the photosynthesis of vitamin D became critically important for landdwelling vertebrates that required a mechanism to increase the efficiency of absorption of dietary calcium. The major physiological function of vitamin D in vertebrates is to maintain extracellular fluid concentrations of calcium and phosphorus within a normal range. Vitamin D accomplishes this by increasing the efficiency of the small intestine to absorb dietary calcium and phosphorus, and by stimulating the mobilization of calcium and phosphorus stores from the bone [3].
II. HISTORICAL PERSPECTIVE It was the lack of appreciation of the beneficial effect of sunlight in preventing rickets that caused the widespread endemic outbreak of this bone-deforming disease in the 17th through 19th centuries. As early as 1822, Sniadecki [4] had suggested that the high incidence of rickets that occurred in Warsaw, Poland, in the early 1800s was likely caused by the lack of adequate exposure to sunlight. This was based on his observations that whereas children who lived in Warsaw had a very high incidence of rickets, children who lived in the rural areas outside of the city did not suffer the same malady. Seventy years later Palm [5] also recognized that lack of sunlight was a common denominator associated with the high incidence of rickets in children VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
V. Sunlight, Vitamin D, and Skin Cancer VI. Conclusion Acknowledgment References
living in the inner cities when compared to children living in underdeveloped countries. Both Sniadecki [4] and Palm [5] encouraged systematic sunbathing as a means of preventing and curing rickets, but their recommendations went unheeded. Huldschinsky [6] was the first to clearly demonstrate that it was exposure of the skin to ultraviolet radiation that was responsible for the antirachitic activity of sunlight that had been postulated by Sniadecki [4] and Palm [5]. He exposed four rachitic children to radiation from a mercury arc lamp and reported dramatic reversal of rickets within 4 months. He further reasoned that the radiation responsible for producing the antirachitic factor was the same radiation that caused melanization of the skin. The intimate relationship between sunlight and its antirachitic properties in humans was finally demonstrated in 1921 when Hess and Unger [7] took rachitic children and exposed them to sunlight on the roof of a New York City hospital and demonstrated that within 4 months the rachitic lesions had healed. The concept that rickets was caused by a nutritional deficiency was supported by the observation in 1827 by Bretonneau, who treated a 15-month-old child with acute rickets with cod liver oil and noted the incredible speed at which the patient was cured [8]. His student Trousseau used liver oils from a variety of fish and aquatic mammals including herring, whales, and seals for the treatment of rickets and osteomalacia. In the early 20th century, Mellanby [9] conducted classic studies where he demonstrated that rachitic beagles that were fed cod liver oil were cured of their bone-deforming disease. Originally, it was thought that the active antirachitic factor in cod liver oil was vitamin A. However, McCollum et al. [10] heated cod liver oil to destroy the vitamin A activity and demonstrated that the preparation maintained its antirachitic activity. He, therefore, coined the term “vitamin D” for the antirachitic factor. Copyright © 2005, Elsevier, Inc. All rights reserved.
38 The appreciation that exposure to sunlight could cure rickets in children prompted Steenbock and Black [11] and Hess and Weinstock [12] in 1924 to expose a variety of foods and other substances including vegetable oils, animal feed, lettuce, wheat, and grasses to ultraviolet radiation. They found that this process imparted antirachitic activity to all of the substances. These observations led Steenbock [13] to conclude that there would be great utility in irradiation of food substances for the prevention and cure of rickets in children. This concept initially led to the addition of provitamin D (ergosterol or 7-dehydrocholesterol) to milk and its subsequent irradiation to impart antirachitic activity. Once vitamin D (D without a subscript refers to either D2 or D3) was inexpensive to produce, vitamin D was directly added to milk. This simple concept led to the eradication of rickets as a significant health problem in the United States and other countries that followed this practice.
MICHAEL F. HOLICK
SUN
SUN
SUN H HO 7-DEHYDROCHOLESTROL
SUN H
HO
LUMISTEROL3 SUN HO
PREVITAMIN3 CH3
H BLOOD VESSEL
OH TACHYSTEROL3
DBP HO SUN
VITAMIN3
SUN
SUN
III. PHOTOBIOLOGY OF VITAMIN D A. Photosynthesis of Vitamin D3 When human skin is exposed to sunlight, it is the solar ultraviolet B photons with energies between 290 and 315 nm that are responsible for causing the photolysis of 7-dehydrocholesterol (provitamin D3; the immediate precursor in the cholesterol biosynthetic pathway) to previtamin D3 [1,14] (Fig. 1). This photochemical process occurs in the plasma membrane of skin cells; as a result, the thermodynamically unstable cis,cis isomer of previtamin D3 is rapidly transformed by a rearrangement of double bonds to form vitamin D3 [15] (Fig. 2). Approximately 50% of previtamin D3 is converted to vitamin D3 within 2 hr. As vitamin D3 is formed in the membrane, its open flexible structure is likely jettisoned from the plasma membrane into the extracellular space. Once vitamin D3 enters the extracellular fluid space, it is attracted to the vitamin D binding protein (DBP) in the circulation, and thus enters the dermal capillary bed.
B. Regulation of Vitamin D Synthesis by Sunlight In the late 1960s, a theory was popularized that human skin pigmentation evolved to play a critically important role in regulating the cutaneous photosynthesis of vitamin D3. Loomis [16] speculated that melanin pigmentation in humans evolved for the purpose of preventing excessive production of vitamin D3 in the skin due to chronic excessive exposure to
OH SUPRASTEROL I
HO SUPRASTEROL II OH 5,6-TRANSVITAMIN D
FIGURE 1 Photochemical events that lead to the production and regulation of vitamin D3 in the skin. Reprinted with permission from Holick [1]. DBP is the plasma vitamin D binding protein.
sunlight especially in peoples who live near the equator. Although it is likely that increased pigmentation was important for the prevention of skin cancer, there is little evidence that skin pigmentation evolved to prevent vitamin D intoxication. The reason for this is that once previtamin D3 is formed in the skin, it either can isomerize to vitamin D3 or can absorb solar ultraviolet B radiation and isomerize into biologically inactive photoisomers lumisterol and tachysterol [17] (Fig. 1). Furthermore, when vitamin D3 is made, it can either enter into the circulation or absorb solar ultraviolet photons and isomerize to at least three photoproducts including suprasterol I, suprasterol II, and 5,6-transvitamin D3 [18] (Fig. 1). Thus, sunlight itself can regulate the total output of vitamin D3 in the skin by causing the photodegradation of previtamin D3 and vitamin D3. This is the likely explanation for why there are no reported cases of vitamin D intoxication–induced hypercalcemia from chronic excessive exposure to sunlight. Melanin, on the other hand, provides the body with an effective natural sunscreen. Thus, increased melanin pigmentation will reduce the efficiency of the sun-mediated photosynthesis of previtamin D3 [19]. This is particularly
39
CHAPTER 3 Photobiology of Vitamin D
OH
s-trans,s-cis-PRE-D3
SUN
C A
D
B
HO
HEXANE
PRO-D3
19 C D 10 9 3 5 6 7 HO
T1/2= 91 HRS
s-cis,s-cis-PRE-D3 25°C
CH3
SUN HO
VITAMIN D3 T1/2= 8 HRS
SKIN
HO PRO-D3
HO s-cis,s-cis-PRE-D3
FIGURE 2
Photolysis of provitamin D3 (pro-D3) into previtamin D3 (pre-D3) and its thermal isomerization to vitamin D3 in hexane and in lizard skin. In hexane, pro-D3 is photolyzed to s-cis,s-cis-pre-D3. Once formed, this energetically unstable conformation undergoes a conformational change to the s-trans,s-cis-pre-D3. Only the s-cis, s-cis-pre-D3 can undergo thermal isomerization to vitamin D3. The s-cis,s-cis conformer of pre-D3 is stabilized in the phospholipid bilayer by hydrophilic interactions between the 3/3-hydroxyl group and the polar head of the lipids, as well as by van der Waals interactions between the steroid ring and side-chain structure and the hydrophobic tail of the lipids. These interactions significantly decrease the conversion of the s-cis,s-cis conformer to the s-trans,s-cis conformer, thereby facilitating the thermal isomerization of s-cis,s-cis-pre-D3 to vitamin D3. Reprinted with permission from Holick et al. [15].
important in Blacks who live in northern latitudes and who ingest very little dietary vitamin D. It is the likely explanation for why Blacks have lower circulating concentrations of 25-hydroxyvitamin D3 (25OHD) and are more prone to developing vitamin D deficiency [20] (see Chapter 33).
C. Influence of Latitude, Season, and Time of Day on Vitamin D Synthesis It was recognized at the beginning of the 20th century that the incidence of rickets was much higher during the winter and early spring months and that rickets was less prevalent during the summer and autumn [21]. An evaluation of the effect of season on the cutaneous production of vitamin D3 in Boston revealed that during the summer months of June and July, 7-dehydrocholesterol was most efficiently converted to previtamin D3 [22] (Fig. 3). There was a gradual decline in the production of previtamin D3 after August, and there was essentially no previtamin D3 formed in human skin
after November (Fig. 3). Previtamin D3 photosynthesis commenced in the middle of March. To evaluate the influence of latitude on the cutaneous production of vitamin D3, similar studies were conducted in Edmonton, Canada (52°N), Los Angeles (34°N), and San Juan, Puerto Rico (18°N). In Edmonton, the photosynthesis of previtamin D3 essentially ceased by mid-October and did not resume until mid-April. However, in Los Angeles and San Juan, the production of previtamin D3 in the skin occurred throughout the year (Fig. 3). The cutaneous production of vitamin D3 was also evaluated in Boston every hour from the time the sun rose to the time the sun set in the middle of the month on cloudless days for a full year. It was found that during the summer, sunlight was capable of producing vitamin D3 in the skin from 0700 to as late as 1700 hr Eastern Standard Time (EST) (Fig. 4). However, as the zenith angle of the sun increased in the spring and autumn, previtamin D3 photosynthesis in the skin began at approximately 0900 and ceased at approximately 1600 EST [1].
40
MICHAEL F. HOLICK
D. Influence of Sunscreen Use, Melanin, Clothing, Glass, and Plastics on Vitamin D Synthesis
25
20
15
10
5
0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
FIGURE 3
Photosynthesis of previtamin D3 after exposure of 7-dehydrocholesterol (7-DHC) to sunlight. Measurements were as follows: in Boston (42°N) after 1 hr () and 3 hr () and total photo products (previtamin D3, lumisterol, and tachysterol) after 3 h in Boston (); in Edmonton, Canada (52°N), after 1 hr (); in Los Angeles (34°N) () and Puerto Rico (18°N) in January (). Reprinted with permission from Webb et al. [22].
Boston Edmonton Bergen
A
% Previtamin D3 formation
8
Any substance such as melanin, clothing, or a sunscreen that absorbs ultraviolet B radiation will reduce the cutaneous production of vitamin D3 [1,23–25]. Sunscreen use is now widely accepted as an effective method for diminishing the damaging effects due to chronic excessive exposure to sunlight. Sunscreens are of great benefit because they can prevent sunburn, skin cancer, and skin damage associated with exposure to sunlight. The solar radiation that is most responsible for causing damage to the skin is the ultraviolet B (UVB) radiation. Because the major function of sunscreens is to absorb solar UVB radiation on the surface of the skin before it penetrates into the deeper viable layers, sunscreen use diminishes the total number of UVB photons that can reach the 7-dehydrocholesterol stores in the viable epidermis to form previtamin D3. Topical application of a sunscreen will substantially diminish or completely prevent the cutaneous production of previtamin D3 [25]. The topical application of
8
C
6
6
4
4
2
2
0
0 Jan
Mar
May
Jul
Sep
8:00
Nov
Buenos Aires Johannesburg Cape Town Ushuaia
10
% Previtamin D3 formation
% Previtamin D3 formation
B
8 6 4 2
10:00 12:00 14:00 16:00 18:00
Hour of Day
Month 10
Boston Edmonton Bergen
% Previtamin D3 formation
% Photoproducts from 7-DHC
30
Buenos Aires Johannesburg Cape Town Ushuaia
D
8 6 4 2 0
0 Jan
Mar
May
Jul
Month
Sep
Nov
8:00
10:00 12:00 14:00 16:00 18:00
Hour of Day
FIGURE 4 Influence of season, time of day, and latitude on the synthesis of previtamin D3 in Northern (A and C) and southern hemispheres (B and D). The hour indicated in C and D is the end of the 1-hr exposure time. (Reproduced with permission from Chen TC 1998 The Photobiology of Vitamin D. In: Holick MF (ed) Vitamin D—Physiology, Molecular Biology and Clinical Applications. Humana Press, Totowa, NJ, pp. 17–37.)
41
CHAPTER 3 Photobiology of Vitamin D
Serum Vitamin D (nmol/L)
80
Without Sunscreen
60 40
With Sunscreen
20 0
−2
0
2
4
8 6 Days
10
12
14
16
FIGURE 5 Effect of sunscreen on vitamin D status. Circulating concentrations of vitamin D were measured in young adults after application of a cream either with a sun protection factor of 8 or without sunscreen (topical placebo cream) following a single exposure to one minimal erythemal dose of simulated sunlight. Reprinted with permission from Matsuoka et al. [23].
60
Serum 25-Hydroxyvitamin D (ng/ml)
Serum Vitamin D3 (ng/ml)
a sunscreen with a sun protection factor (SPF) of 8, followed by whole-body exposure to one minimal erythemal dose of simulated sunlight, prevented any significant increase in the blood levels of vitamin D3 in healthy young adult volunteers. On the other hand, without the sunscreen, there was a marked 10- to 20-fold increase in circulating concentrations of vitamin D3 [23] (Fig. 5). Similarly, since clothing absorbs ultraviolet B radiation, the wearing of clothing on body surfaces prevents the cutaneous production of vitamin D3 in those covered surfaces [24] (Fig. 6). Thus, in cultures that require the covering of almost the entire body with
clothing, such as Bedouins living in the Negev Desert, despite the intense sunlight environment, the Bedouin women have an increased risk of vitamin D deficiency and osteomalacia and their children are more prone to developing vitamin D–deficiency rickets [26]. Although sunscreens and clothing significantly diminish the cutaneous production of vitamin D, for children and young adults there is little concern about their developing vitamin D deficiency by practicing sun protection. The main reason is that the casual everyday limited exposure of sunlight to the face and hands is sufficient to provide the vitamin D requirement. However, elderly people, who are often very concerned about their health and appearance, will consistently apply a topical sunscreen on all sun-exposed areas and/or wear clothing over most sun-exposed areas before going outdoors. Since aging significantly decreases the capacity of the skin to produce vitamin D3 because of the marked decline in 7-dehydrocholesterol in the epidermis, chronic sunscreen use by elderly people can increase the risk of vitamin D deficiency [1,23] (Fig. 7). Because glass and Plexiglas efficiently absorb most if not all ultraviolet B photons, exposure of the skin to sunlight through glass or Plexiglas will not produce any vitamin D3 (Fig. 8).
80 60 40 20 0
None
Summer
Autumn
Summer and Sunscreen
PA IL 40
30
20
10
0 Without Sunscreen
FIGURE 6 Effect of clothing and sunscreen on vitamin D status. Circulating concentrations of vitamin D were measured in human subjects who wore either no clothing, summer-type clothing, autumn-type clothing, or summer-type clothing and sunscreen 24 hr after a whole-body exposure to one minimal erythemal dose of ultraviolet B radiation. Reprinted with permission from Matsuoka et al. [24].
50
FIGURE 7
With Sunscreen
Effect of chronic sunscreen use on vitamin D status. Circulating concentrations of 25-hydroxyvitamin D were measured in adults from Pennsylvania (PA) and Illinois (IL) who always wore a sunscreen or never wore a sunscreen. Reprinted with permission from Matsuoka et al. [25].
42
MICHAEL F. HOLICK
% Previtamin D3 Synthesis
7 6
5.98%
5 4 3 2 1 0
No Shield
ND Glass
ND Plexiglas
ND Plastic
FIGURE 8 Prevention of provitamin D3 formation by UV radiation by common light-shielding materials. The effects of glass, plastic, or Plexiglas (DuPont Chemical Company, Memphis, TN) placed between the simulated sunlight source and the provitamin D3 were measured. ND, Not detectable. Reprinted with permission from Holick [1].
IV. ROLE OF SUNLIGHT AND DIETARY VITAMIN D IN BONE HEALTH, OVERALL HEALTH, AND WELL-BEING Vitamin D deficiency is finally being recognized as an epidemic for all age groups in most industrialized countries [27–35]. Ninety to 95% of our vitamin D requirement comes from exposure to sunlight [36]. Most experts agree that in the absence of sun exposure, 1000 IU of vitamin D is required to satisfy the body’s vitamin D need [37]. If children and adults received sensible exposure to sunlight during the spring, summer and fall, they are able to store vitamin D in their body fat and call upon it during the winter when the sun is incapable of producing vitamin D in the skin. However, because of the increased concern about the risk of skin cancer from sun exposure and because both children and adults have fewer outdoor activities, there is a resurgence of vitamin D deficiency for the population at large. Nesby-O’Dell et al. [38] reported that 42% of African-American women aged 15–49 years were found to be vitamin D–deficient throughout the United States at the end of the winter. Tangpricha et al. reported that 32% of healthy young men and women aged 18–29 years were vitamin D deficient at the end of the winter in the Boston area [31]. Sullivan et al. [35] reported that 48% of young girls aged 9–11 years were vitamin D deficient at the end of the winter and 17% were vitamin D deficient at the end of the summer in Maine. Vitamin D deficiency has also become a major health concern for pregnant women and infants who receive their sole nutrition from breastfeeding [54].
Rickets, which was considered to be a disease of the 19th century, has become much more common, especially in children of color who are solely breastfed [52]. For preadolescent and adolescent children, vitamin D deficiency will prevent them from attaining their genetically programmed peak bone mass. For adults, it will precipitate and exacerbate osteoporosis, increase risk of fracture, and cause the painful bone disease osteomalacia [39–44]. It is now also recognized that insufficient exposure to sunlight and vitamin D deficiency increases risk of many common cancers, including colon, breast, prostate, ovarian, and esophageal cancers and non-Hodgkin’s and Hodgkin’s lymphoma [45–49]. In addition, it increases risk of children developing type 1 diabetes [50] and also puts adults at increased risk of developing hypertension and having heart disease [51–55].
V. SUNLIGHT, VITAMIN D, AND SKIN CANCER There is no question that chronic excessive exposure to sunlight and sun-burning experiences, especially during childhood and young adulthood, increases risk of squamous and basal cell carcinoma and wrinkling [56]. There is, however, little evidence that sensible limited exposure to sunlight to satisfy the body’s requirement for vitamin D will substantially increase risk of developing either basal or squamous cell carcinoma [36]. Furthermore, if basal or squamous cell carcinoma develops and is detected early, this type of cancer is easily curable. Melanoma, on the other hand, is often a deadly skin cancer. Although it has been suggested that melanoma is associated with sun exposure, it should be noted that most melanomas occur on non—sun-exposed areas, i.e., back of leg, abdomen, etc., and rarely occur on the face [57]. Furthermore, a recent study in Scandinavia suggested that melanoma is associated with increased number of nevi (moles), color of the hair, i.e., red and blonde, and number of sun-burning experiences [58]. They also reported that there may be an association with frequent exposure to a tanning bed. However, the data were not statistically significant and the authors concluded that there was only a suggestion of an association. To put this into perspective, it is estimated that approximately 1200 people per year will die in the United States of squamous and basal cell carcinomas that are clearly associated with excessive sun exposure. However, the lack of adequate sun exposure and/or adequate vitamin D nutrition puts at risk more than 150,000 lives for dying of many common cancers [59,60]. There has been no estimate of the toll that vitamin D
CHAPTER 3 Photobiology of Vitamin D
deficiency has in mortality rates due to heart disease and complications of type 1 diabetes, which has now been clearly linked to vitamin D deficiency.
VI. CONCLUSION The skin has a huge capacity to produce vitamin D [61]. Children and adults exposed to natural or artificial ultraviolet B radiation can satisfy their vitamin D requirement [62–66]. Exposure of a healthy adult in a bathing suit to one minimal erythemal dose (a light pinkness to the skin) of sunlight or tanning bed UV radiation is equivalent to taking between 10,000 IU and 20,000 IU of vitamin D [36,61]. Because melanin is such an effective sunscreen, African Americans require five to 50 times the exposure that a white person requires to satisfy their body’s vitamin D requirement [19]. Although aging markedly diminishes the capacity of the skin to produce vitamin D3, because its capacity is so high, elders exposed to sunlight can still raise their blood levels of 25(OH)D into a satisfactory range [67]. Thus, there needs to be a reexamination of the message that any exposure to sunlight requires some type of sunscreen or sun protection. This extreme position has pervaded the psyche of the population at large. It unfortunately has put many people at risk for vitamin D deficiency and many of the serious chronic diseases that have been associated with inadequate sun exposure and vitamin D deficiency. Sensible exposure to sunlight, typically no more than 5 to 10 minutes a day between 10 A.M. and 3 P.M. during seasons when vitamin D can be produced in the skin will satisfy most people’s vitamin D requirement [60,61]. Taking a multivitamin containing 400 IU of vitamin D will satisfy approximately 40% of the body’s requirement. Thus, additional supplementation and/or foods that contain vitamin D are required to satisfy the 1000 IU of vitamin D that most experts believe is necessary to raise the blood levels of 25(OH)D above 30 ng/ml, which is considered to be a healthy level.
Acknowledgment This work was supported in part by the following grants: General Clinical Research Center M01RR00533, R01 AR36963 and the ITA’s UV Foundation.
References 1. Holick MF 1994 McCollum Award Lecture, Vitamin D—New horizons for the 21st century. Am J Clin Nutr 60:619–630.
43 2. Holick MF 1989 Phylogenetic and evolutionary aspects of vitamin D from phytoplankton to humans. In: Pang PKT, Schreibman MP (eds) Vertebrate Endocrinology: Fundamentals and Biomedical Implications, Vol. 3. Academic Press, Orlando, FL, pp. 7–43. 3. Holick MF 1995 Vitamin D: Photobiology, metabolism, and clinical applications. In: DeGroot LJ, Besser M, Burger HG, Jameon JL, Loriaux DL, Marshall JC, Odell WD, Potts JT, Rubenstein AH (eds) Endocrinology, 3rd ed. Chapter 59. Saunders, Philadelphia, pp. 990–1013. 4. Sniadecki J 1939 cited by W Mozolowski, Jerdrzej Sniadecki (1768–1883) on the cure of rickets. Nature 143:121. 5. Palm TA 1980 The geographic distribution and etiology of rickets. Practitioner 45:270–279, 421–442. 6. Huldschinsky K 1919 Curing rickets by artificial UV-radiation (Heilung von Rachitis durch kunstliche Hohensonne). Deut Med Wochenschr 45:712–713 (in German). 7. Hess AF, Unger LF 1921 Cure of infantile rickets by sunlight. JAMA 77:39. 8. Mayer J 1957 Armand Trousseau and the arrow of time. Nutr Rev 15:321–323. 9. Mellanby T 1918 The part played by an accessory factor in the production of experimental rickets. J Physiol 52:11–14. 10. McCollum EF, Simmonds N, Becker JE, Shipley PG 1922 Studies on experimental rickets; and experimental demonstration of the existence of a vitamin which promotes calcium deposition. J Biol Chem 53:293–312. 11. Steenbock H, Black A 1924 The reduction of growth-promoting and calcifying properties in a ration by exposure to ultraviolet light. J Biol Chem 61:408–411. 12. Hess AF, Weinstock M 1924 Antirachitic properties imparted to inert fluids and green vegetables by ultraviolet irradiation. J Biol Chem 62:301–313. 13. Steenbock H 1924 The induction of growth-prompting and calcifying properties in a ration exposed to light. Science 60:224–225. 14. MacLaughlin JA, Anderson RR, Holick MF 1982 Spectral character of sunlight modulates photosynthesis of previtamin D3 and its photoisomers in human skin. Science 216:1001–1003. 15. Holick MF, Tian XQ, Allen M 1995 Evolutionary importance for the membrane enhancement of the production of vitamin D3 in the skin of poikilothermic animals. Proc Natl Acad Sci USA 92:3124–3126. 16. Loomis F 1967 Skin-pigment regulation of vitamin D biosynthesis in man. Science 157:501–506. 17. Holick MF, MacLaughlin JA, Doppelt SH 1981 Factors that influence the cutaneous photosynthesis of previtamin D3. Science 211:590–593. 18. Webb AR, deCosta BR, Holick MF 1989 Sunlight regulates the cutaneous production of vitamin D3 by causing its photodegradation. J Clin Endocrinol Metab 68:882–887. 19. Clemens TL, Adams JS, Henderson SL, Holick MF 1982 Increased skin pigment reduces the capacity of the skin to synthesize vitamin D. Lancet 1:74–76. 20. Bell NH, Greene A, Epstein S, Oexmann MJ, Shaw W, Shary J 1985 Evidence for alteration of the vitamin D endocrine system in Blacks. J Pediatr 76:470–473. 21. Kassowitz M 1987 Tetany and autointoxication in infants (Tetani and autointoxication in kindersalter). Wien Med Presse 97:139 (in Dutch). 22. Webb AR, Kline L, Holick MF 1988 Influence of season and latitude on the cutaneous synthesis of vitamin D3: Exposure to winter sunlight in Boston and Edmonton will not promote vitamin D3 synthesis in human skin. J Clin Endocrinol Metab 67:373–378.
44 23. Matsuoka LY, Ide L, Wortsman J, MacLaughlin JA, Holick MF 1987 Sunscreens suppress cutaneous vitamin D3 synthesis. J Clin Endocrinol Metab 64:1165–1168. 24. Matsuoka LY, Wortsman J, Dannenberg MJ, Hollis BW, Lu Z, Holick MF 1992 Clothing prevents ultraviolet-B radiationdependent photosynthesis of vitamin D3. J Clin Endocrinol Metab 75:1099–1103. 25. Matsuoka LY, Wortsman J, Hanifan N, Holick MF 1988 Chronic sunscreen use decreases circulating concentrations of 25-hydroxyvitamin D: A preliminary study. Arch Dermatol 124:1802–1804. 26. Sedrani SH, Al-Arabi KM, Abanny A, Elidrissy A 1990 Frequency of vitamin D deficiency rickets in Riyadh. In: Study of Vitamin D Status and Factors Leading to Its Deficiency in Saudi Arabia. King Saudi Univ. Press, Riyadh, pp. 281–285. 27. Gloth FM, Gundberg CM, Hollis BW, Haddad HG, Tobin JD 1995 Vitamin D deficiency in homebound elderly persons. JAMA 274:1683–1686. 28. Malabanan A, Veronikis IE, Holick MF 1998 Redefining vitamin D insufficiency. Lancet 351:805–806. 29. Lips P 2001 Vitamin D deficiency and secondary hyperparathyroidism in the elderly: consequences for bone loss and fractures and therapeutic implications. Endocr Rev 22:477–501. 30. Chapuy MC, Preziosi P, Maaner M, Arnaud S, Galan P, Hercberg S, et al. 1997 Prevalence of vitamin D insufficiency in an adult normal population. Osteopor Int 7:439–443. 31. Tangpricha V, Pearce EN, Chen TC, Holick MF 2002 Vitamin D insufficiency among free-living healthy young adults. Am J Med 112:659–662. 32. Kreiter SR, Schwartz RP, Kirkman HN, Charlton PA, Calikoglu AS, Davenport M 2000 Nutritional rickets in African American breast-fed infants. J Pediatr 137:2–6. 33. Welch TR 2000 Vitamin D deficient rickets: the reemergence of a once-conquered disease J Pediatr 137:143–145. 34. Thomas MK, Lloyd-Jones DM, Thadhani RI, Shaw AC, Deraska DJ, Kitch BT 1998 Hypovitaminosis D in medical inpatients. N Engl J Med 338:777–783. 35. Sullivan SS, Rosen CJ, Chen TC, Holick MF 2003 Seasonal changes in serum 25(OH)D in adolescent girls in Maine. J Bone Mine Res (Proceedings for Twenty-Fifth Annual Meeting of the American Society for Bone and Mineral Research), S407. 36. Holick MF 2003 Vitamin D: a millennium perspective. J Cell Biochem 88:296–307. 37. Heaney RP, Davies KM, Chen TC, Holick MF, Barger-Lux MJ 2003 Human serum 25-hydroxycholecalciferol response to extended oral dosing with cholecalciferol. Am J Clin Nutr 77:204–210. 38. Nesby-O’Dell S, Scanlon KS, Cogswell ME, Gillespie C, Hollis BW, Looker AC 2002 Hypovitaminosis D prevalence and determinants among African American and white women of reproductive age: third national health and nutrition examination survey, 1988–1994. Am J Clin Nutr 76:187–192. 39. Holick MF 2001 Sunlight “D”ilemma: risk of skin cancer or bone disease and muscle weakness. Lancet 357:4–6. 40. Glerup H, Mikkelsen K, Poulsen L, Hass E, Overbeck S, Andersen H et al. 2000 Hypovitaminosis D myopathy without osteomalacic bone involvement. Calcif Tissue Int 66:419–424. 41. Chapuy MC, Schott AM, Garnero P, Hans D, Delmas PD, Meunier J, et al. 1996 Healthy elderly French women living at home have secondary hyperparathyroidism and high bone turnover in winter. J Clin Endocrinol Metab 81:1129–1133. 42. Malabanan AO, Turner AK, Holick MF 1998 Severe generalized bone pain and osteoporosis in a premenopausal black female: Effect of vitamin D replacement. J Clin Densitometr 1:201–204.
MICHAEL F. HOLICK
43. Rimaniol J, Authier F, Chariot P 1994 Muscle weakness in intensive care patients: initial manifestation of vitamin D deficiency. Intensive Care Med 20:591–592. 44. Bischoff HA, Stähelin HB, Dick W, Akos R, Knecht M, Salis C, et al. 2003 Effect of vitamin D and calcium supplementation on falls: a randomized controlled study. J Bone Miner Res 18:343–351. 45. Garland CF, Garland FC, Shaw EK, Comstock GW, Helsing KJ, Gorham ED 1989 Serum 25-hydroxyvitamin D and colon cancer: Eight-year prospective study. Lancet 18:1176–1178. 46. Garland CF, Garland FC, Gorham ED, Raffa J 1992 Sunlight, Vitamin D, and Mortality from Breast and Colorectal Cancer in Italy. Biologic Effects of Light. Walter de Gruyter, New York, pp. 39–43. 47. Hanchette CL, Schwartz GG 1992 Geographic patterns of prostate cancer mortality. Cancer 70:2861–2869. 48. Ahonen MH, Tenkanen L, Teppo L, Hakama M, Tuohimaa P 2000 Prostate cancer risk and prediagnostic serum 25-hydroxyvitamin D levels (Finland). Cancer Causes Control 11:847–852. 49. Grant WB 2002 An ecologic study of dietary and solar ultraviolet-B links to breast carcinoma mortality rates. Am Cancer Soc 94:272–281. 50. Hypponen E, Laara E, Jarvelin M-R, Virtanen SM 2001 Intake of vitamin D and risk of type 1 diabetes: a birth-cohort study. Lancet 358:1500–1503. 51. Rostand SG 1979 Ultraviolet light may contribute to geographic and racial blood pressure differences. Hypertension 30:150–156. 52. Li YC, Kong J, Wei M, Chen ZF, Liu SQ, Cao LP 2002 1,25dihydroxyvitamin D3 is a negative endocrine regulator of the renin–angiotensin system. J Clin Invest 110: 229–238. 53. Scragg R, Jackson R, Holdaway IM, Lim T, Beaglehole R 1990 Myocardial infarction is inversely associated with plasma 25-hydroxyvitamin D3 levels: a community-based study. Int J Epidemiol 19:559–563. 54. Zitterman A, Schulze Schleithoff S, Tenderich C, Berthold H, Koefer R, Stehle P 2003 Low vitamin D status: a contributing factor in the pathogenesis of congestive heart failure? J Am Coll Cardiol 41:105–112. 55. Holick MF 2002 Sunlight and vitamin D, both good for cardiovascular health [editiorial] J Gen Intern Med 17: 733–735. 56. Housman TS, Feldman SR, Williford PM, Fleischer Jr AB, Goldman ND, Acostamadiedo JM 2003 Skin cancer is among the most costly of all cancers to treat for the Medicare population. J Am Acad Dermatol 48:425–429. 57. Garland CF, Garland FC, Gorham ED. Rising trends in melanoma. An hypothesis concerning sunscreen effectiveness. Ann Epidemiol 3:103–110. 58. Veierod MB, Weiderpass E, Thorn M, Hansson J, Lund E, Armstrong B, Adami H-O 2003 A prospective study of pigmentation, sun exposure, and risk of cutaneous malignant melanoma in women. J Nat Can Inst 95:1530–1538. 59. Grant WB 2002 An estimate of premature cancer mortality in the U.S. due to inadequate doses of solar ultraviolet-B radiation. Cancer 70:2861–2869. 60. Holick MF, Jenkins M 2003 The UV Advantage. iBooks, New York 61. Holick MF 2002 Vitamin D: the underappreciated D-lightful hormone that is important for skeletal and cellular health. Curr Opin Endocrinol Diabetes 9:87–98. 62. Chel VGM, Ooms ME, Popp-Snijders C, Pavel S, Schothorst AA, Meulemans CCE, et al. 1998 Ultraviolet irradiation corrects vitamin D deficiency and suppresses
CHAPTER 3 Photobiology of Vitamin D
secondary hyperparathyroidism in the elderly. J Bone Miner Res 13:1238–1242. 63. Chuck A, Todd J, Diffey B 2001 Subliminal ultraviolet-B irradiation for the prevention of vitamin D deficiency in the elderly: a feasibility study. Photochem Photoimmun Photomed 17:168–171. 64. Lund B, Sorensen OH 1979 Measurement of 25-hydroxyvitamin D in serum and its relation to sunshine, age, and vitamin D intake in the Danish population. Scand J Clin Lab Invest 39:23–30.
45 65. Adams JA, Clemens TL, Parrish JA, Holick MF 1982 Vitamin-D synthesis and metabolism after ultraviolet irradiation of normal and vitamin-D-deficient subjects. N Engl J Med 306:722–725. 66. Koutkia P, Lu Z, Chen TC, Holick MF 2001 Treatment of vitamin D deficiency due to Crohn’s disease with tanning bed ultraviolet B radiation. Gastroenterology 121:1485–1488. 67. Reid IR, Gallagher DJA, Bosworth J 1985 Prophylaxis against vitamin D deficiency in the elderly by regular sunlight exposure. Age Ageing 15:35–40.
CHAPTER 4
The Vitamin D 25-Hydroxylase MARIELLE GASCON-BARRÉ
Département de Pharmacologie, Faculté de Médecine, Université de Montréal, and Centre de recherche de l’Université de Montréal, Montréal, Québec Canada
I. Introduction II. Hepatic Uptake of Vitamin D III. The Monooxygenases Active in the Hydroxylation of Vitamin D at C-25 IV. The Microsomal Enzymes V. The Mitochondrial Enzyme
VI. Cerebrotendinous Xanthomathosis VII. Ontogeny of the Vitamin D 25-Hydroxylases VIII. Sex Differences in the Hydroxylation of Vitamin D at C-25 IX. Conclusions References
I. INTRODUCTION
mitochondrial enzymes, believed by many to be more physiologically relevant than the mitochondrial entity. In this chapter, we review the most relevant research area on the D 25-hydroxylases and address the specificity and regulation of each enzyme in the context of its dynamic functioning, including uptake, ontogeny, sex-related differences, and hepatic and intrahepatic regionalization.
25-Hydroxyvitamin D (25OHD)* is the first hydroxylated metabolite of vitamin D (D) and the immediate precursor of the fully active and hormonal form of the vitamin, 1α,25-dihydroxyvitamin D [1,25(OH)2D]. It was discovered by DeLuca and his group, who rapidly identified the liver as the first site of activation of D3 [1–3]. Over the past 35 years, the enzyme systems involved in the C-25 hydroxylation of D3, D2, and several of their analogs have been the object of intense studies by groups in North America, Europe, and Japan. The research has allowed the identification of two intrahepatic organelles, the smooth endoplasmic reticulum (microsomes) and the mitochondrium, as sites possessing fully active but distinct D 25-hydroxylases. The mitochondrial enzyme has been cloned [4–6] and its identity as a D3 25-hydroxylase established with certainty. Moreover, its presence and activity has been positively identified in all species studied including the human [7]. The microsomal enzyme received the attention of early workers in the field. It has been identified clearly in the pig, where the enzyme has been cloned and clearly shown to hydroxylate D3 and D2 at C-25 [8]. Lately, a new microsomal D 25-hydroxylase species has also been cloned and its gene transcript shown to be present in mice and humans [9]. The latter is also reported to be active in the C-25 hydroxylation of both D3 and D2. In most species, early work has shown that the microsomal D3 25-hydroxylase is an enzyme also active in the oxidation of several endogenous and exogenous substances and, based on the enzyme kinetics of the respective microsomal and *The
term vitamin D includes both vitamin D3 and vitamin D2.
VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
II. HEPATIC UPTAKE OF VITAMIN D A. The First Prerequisite Step toward Its Activation The hepatic sequestration of vitamin D is the first prerequisite step toward its activation and subsequent C-25 hydroxylation by the liver. Under normal conditions, the fractional hepatic uptake of D3 during a single pass across the rat and dog liver lies between 40 and 60% [10–12], an uptake higher than that observed for all hydroxylated metabolites of the vitamin [10,13,14]. Its hepatic clearance is estimated to be 357 ml min−1 in dogs [11]. Studies in which total uptake has been investigated for periods varying from 18 sec to 70 min have revealed that the liver does not accumulate significantly more D3 than that observed during the first pass across the organ [13,15–19] suggesting a steady state liver/ serum concentration ratio of 0.4 to 0.6 at physiological concentrations of the vitamin. In addition, data available indicate that there is no stringent regulation of uptake by the vitamin D status [13]. Both hepatocytes and nonparenchymal liver cells are able to sequester D3 and the mitochondrial D3 25-hydroxylase gene transcript has been found in sinusoidal endothelial, stellate, and Kupffer cells of the liver [20]. In a 1981 study, Dueland et al. [16], however, found that only hepatocytes were able to transform the vitamin into 25OHD3. Copyright © 2005, Elsevier, Inc. All rights reserved.
48 The uptake of D3 is not regioselective within the hepatic acinus and not perturbed by the destruction of either the periportal (proximal) or perivenous (distal or pericentral) region [21], indicating that its extraction takes place according to the concentration gradient along the acinus. This observation suggests that the intrahepatic concentration of the vitamin should be higher in periportal than in perivenous hepatocytes. This notion may be important in light of the reported intraacinar localization of the mitochondrial D3 25-hydroxylase, the kinetic parameters of the mitochondrial and microsomal enzymes, and the influence of regioselective liver diseases on the biosynthesis of 25OHD3 (see Chapter 75 for review). In addition, the hepatic extraction of D3 has been shown to be independent of its hepatic venous or arterial route of delivery [22,23], indicating that circulating vitamin D of endogenous or exogenous origin should have similar hepatic availability. In addition, manipulation of the diet in order to obtain different proportions of putative D3 carrier proteins or lipoproteins in vivo was found not to significantly affect the hepatic extraction of D3 when evaluated in isolated perfused liver preparations [22], an observation also made in vitro by Ziv et al. [19]. However, Haddad et al. [15] found that chylomicron remnants or LDL optimized the hepatic uptake of the vitamin in vitro.
III. THE MONOOXYGENASES ACTIVE IN THE HYDROXYLATION OF VITAMIN D AT C-25 Early studies rapidly established that the liver was the site of hydroxylation of D3 at C-25 [3,22,24]. The first experimental evidence obtained indicated that the D3 25-hydroxylase was localized in both the microsomal and mitochondrial fractions of liver homogenates [25]. However, it was soon reported that the enzyme was instead found almost entirely in liver microsomes and that mitochondria exhibited no or very low activity toward D3 [26]. It was also rapidly established that oxygen-18 from molecular oxygen was incorporated at C-25, strongly suggesting that the microsomal enzyme was a mixed-function oxidase [27]. Firm demonstration for the presence of a mitochondrial D3 25-hydroxylase only came several years later [7]. In 1978, studies by Suda’s group using isolated–perfused rat liver preparations demonstrated that the activation of 1αOHD3 was not regulated (Km 2.0 µM) compared to that of the natural substrate, and that two distinct Km values could be defined for the C-25 hydroxylation of D3 (5.6 nM and 1 µM), suggesting the participation of more than one D3 25-hydroxylase [28].
MARIELLE GASCON-BARRÉ
Evaluation of enzyme activity revealed that 25OHD3 production was also present in extrahepatic tissue [29,30]. At that time, support for the extrahepatic presence of a D3 25-hydroxylase came from the in vivo studies of Ponchon et al. [2] and Olson et al. [3], who reported that hepatectomy markedly reduced but did not eliminate the formation of 25OH[3H]D3 after [3H]D3 injection. They, indeed, observed that 10% of the radioactivity appearing in the plasma of intact rats was found as [3H]25OHD3 in hepatectomized animals 4 hr after injection of the parent compound. These observations have now been confirmed at the molecular level with the demonstration of the presence of the mitochondrial D3 25-hydroxylase in numerous tissues and organs (Table I). In the pig, the microsomal D3 25-hydroxylase gene transcript and its associated protein have also been shown to be present not only in the liver but also in several other organs including the kidney [31,32], whereas the gene expression of a mouse and human microsomal D 25-hydroxylase has been reported to be mainly present in the liver and testis (Table II) [9]. The contribution of the extrahepatic enzymes to the circulating concentrations of 25OHD under normal physiological conditions is not, however, presently known, but several laboratories have clearly shown that the mitochondrial and the porcine microsomal enzymes are active in extrahepatic tissues. These observations indicate that the enzymes participate in the local production of 25OHD and possibly, as well, in the production of 1α,25(OH)2D as both enzymes have been reported to also harbor 1α-hydroxylase activity as indicated in Tables III and IV.
IV. THE MICROSOMAL ENZYMES DeLuca’s group was the first to report on the subcellular localization of the D3 25-hydroxylase. Indeed, in the mid-1970s, Bhattacharyya and DeLuca [26,33,34] observed that the D3 25-hydroxylase was present in the microsomal fraction of rat liver and that it was influenced by the vitamin D status and required the presence of a cytosolic factor for optimum activity. They also noticed that the 25-hydroxylation of the synthetic compound dihydrotachysterol (DHT3) was not regulated compared to that of the natural substrate D3 [34]. In 1979, Madhok and DeLuca [35] confirmed the presence of the enzyme in rat liver microsomes and soon after, purification and reconstitution of the enzyme was achieved in Björkhem’s [36] and DeLuca’s [37] laboratories. It was shown to be a cytochrome P450 (P450, CYP) mixed-function oxidase requiring NADPH and NADPH cytochrome c/P450 reductase. Yoon and DeLuca [37] also reported that a soluble
49
CHAPTER 4 The Vitamin D 25-Hydroxylase
TABLE I Expression Sites of the Mitochondrial CYP27A1 Species
Location
Humana
Liver (adult and fetal)
Mouse
Rabbit
Rat
Reference
Cali and Russell, 1991 [6] Gascon-Barré et al. [141] Kidney (adult and fetal) Gascon-Barré et al., 2001 [141] HepG2 hepatoma cells Guo et al., 1993 [87] Fibroblastsc Cali and Russell 1991 [6] Macrophages Hansson et al., 2003 [165] Stroma cells of the ovary Jabara et al., 2003 [166] Arterial endothelial cells Javitt et al., 2002 [167] Keratinocytes Lehmann et al., [168] Parathyroid glands Schuessler et al. [169] Parathyroid tumors Correa et al., [170] Intestine (fetal jejunum Theodoropoulos et al. [120] and colon) Liver Ichikawa et al., 1995 [171] Duodenum Calvaria Lung Skin Long bone Spleen Osteoblasts (primary culture) Liver Andersson et al., 1989 [4] Duodenum Adrenal Lung Kidney Spleen Liverd Usui et al., 1990 [5] Twisk et al., 1995c [118] Liver sinusoidal Theodoropoulos et al. [20] endothelial, stellate and Kupffer cells Intestine Theodoropoulos et al. [119] Ovaries Su et al., 1990 [86] Kidney Mullick et al., 1995 [91] Axén et al., 1995 [112]
cytosolic factor was necessary for full reconstitution of enzyme activity (see Table III). In 1983, Andersson et al. [38] purified to homogeneity a rat microsomal cytochrome P450 active on the C-25 hydroxylation of the bile acid intermediates 5β-cholestane-3α,7α,12α-triol (cholestane-triol, C-triol), 5β-cholestane-3α,7α-diol (C-diol) as well as D3, 1αOHD3 but not D2. A year later, Hayashi et al. [39] purified and partially sequenced a P450 active on the 25-hydroxylation of D3 and suggested that an inhibitor was removed from microsomes during the purification steps as previously suggested by others [36,40]. Moreover, contrary to earlier observations, the reconstituted enzyme did not require the presence of a cytoplasmic factor. The partial amino acid sequence obtained indicated that the enzyme was different from P450 (PB-1) and P450 (MC-1) (CYP2B1 and 1A1, respectively, according to P450 nomenclature [41]). The same group later obtained an N-terminal amino acid sequence and reported that their purified P450 was also active on several endo- and xenobiotics as already reported by others [38,42] as well as D3, 1αOHD3 but
TABLE II Species CYP2D25 Pig
CYP2R1 Mouse
In addition to these sites, enzyme activity has been shown in a human macrophages [172] and b bovine endothelial cells [173]. c SV40 transformed; d in situ hybridization illustrates a preferential perivenous localization within the acinus [118].
Human
Expression Sites of the Microsomal CYP2D25 and CYP2R1 Location
Reference
Liver Kidney Adrenals Brain Heart Intestine Lung Muscle Spleen Thymus
Hosseinpour and Wikvall [52]
Liver Testis Kidney Brain Epididymis Lung Spleen Heart Muscle Skin Liver
Cheng et al. [9]
Cheng et al. [9]
50
MARIELLE GASCON-BARRÉ
TABLE III Species
Sex
Substrate
The Microsomal Vitamin D 25-Hydroxylases Hydroxylation Site
Characteristics
Reference
Km: 180 nMa (D-depleted) 444 nMa (D-repleted) Km: 44 nMa 360 nMa 108 to 205 pmol 30 min−1 0.75 nmol P450−1b
Delvin et al., 1978 [174]
Rat
D3
C-25
Rat
D3 DHT3 D3
C-25 C-25 C-25
D3 D3 1αOHD3 D2 Benzphetamine D3
C-25 C-25 C-25 C-25 Demethylation C-25
D3 1αOHD3 D2 C-triole C-diol f Ethylmorphine D3
C-25 C-25 C-25 C-25 C-25 Demethylation C-25
D3 1αOHD3 Testosterone
Rat
Dehydroepiandrosterone Benzphetamine D3
C-25 C-25 C-2α C-16α C-16α
7.68 pmol hr −1 mg pro−1b,c 4.3 pmol min −1 nmol P450−1b 3 pmol min −1 nmol P450−1b 2 pmol min −1 nmol P450−1b 921 pmol min −1 nmol P450−1b 3.5–4 pmol hr −1106 hepatocytesd Km: 4 µM (D-depleted) 6 µM (D-repleted) 335 pmol min−1 nmol P450−1b 1000 pmol min−1 nmol P450−1b <10 pmol min−1 nmol P450−1b 2600 pmol min−1 nmol P450−1b 360 pmol min−1 nmol P450−1b 21,000 pmol min−1 nmol P450−1b 2.3 nmol min−1 mg pro−1b (152 pmol min−1 nmol P450) 0.21 nmol min−1 nmol P450−1b 1.73 nmol min−1 nmol P450−1b 9.34 nmol min−1 nmol P450−1b 8.36 nmol min−1 nmol P450−1b 14.6 nmol min−1 nmol P450−1b
Demethylation C-25
15.6 nmol min−1 nmol P450−1b 195 pmol min−1 nmol P450−1b
Rat
C-triol Testosterone D3 C-triol* C-diol** D3
C-25 C-16α C-25 C-25 C-25 C-25
580 pmol min−1 nmol P450−1b 1283 pmol min−1 nmol P450−1b 9 pmol min−1 nmol P450−1b 90 pmol min−1 nmol P450−1b 188 pmol min−1 nmol P450−1b 335 pmol min−1 mg pro−1b (167 pmol min−1 nmol P450−1)
D3
C-25
1αOHD3 1αOHD3 D3
C-25 C-25 C-25
6 pmol min−1 mg pro−1b (4 pmol min−1 nmol P450−1) 0.152 nmol min mg pro−1a 0.053 nmol min mg pro−1a 37 pmol min 106 hepatocytesd 29 pmol min 106 hepatocytesd
D3
C-25 C-25 C-25
Rat (phenobarbital pretreated) Rat Beef
Rat
Rat
Rat Rat
Rat Rat (D depleted) (1,25(OH)2D3treated) Rat Human
,
1αOHD3
0.46 pmol hr1 mg pro−1a,c 0.40 pmol hr1 mg pro−1a 0.8–4.7 pmol min−1 mg proa
Madhok and DeLuca, 1979 [35] Björkhem et al., 1979 [36]
Yoon and DeLuca, 1979 [37] Hiwatashi et al., 1980 [42]
Dueland et al., 1981 [16]
Andersson et al., 1983 [38]
Hayashi et al., 1984 [39] Hayashi et al., 1986 [43]
Andersson and Jörnvall, 1986 [45]
Dählback and Wikvall, 1987 [46]
Saarem and Pedersen, 1987 [47] Benbrahim et al., 1988 [175]
Thierry-Palmer et al., 1995 [49]
Continued
51
CHAPTER 4 The Vitamin D 25-Hydroxylase
TABLE III Species
Sex
Pig
,
Piglet (normal) Piglet (PDDR) Piglet (normal) Piglet (PDDR) Piglet (PDDR)
,
Castrated
COS cells (pCYP2D25)
Substrate
Hydroxylation Site
D3 D2 1αOHD3 1αOHD2 25OHD3
C-25 C-25 C-25 C-25 C-1α
1αOHD3 D3
C-25
D2 1αOHD3 25OHD3 1αOHD3 25OHD3 25OHD3 D3 D2
pCYP2D25
m/hCYP2R1
The Microsomal Vitamin D 25-Hydroxylases—Cont’d
,
Characteristics 188 pmol min−1 nmol P450−1 e 98 pmol min−1 nmol P450−1e 756 pmol min−1 nmol P450−1 f 842 pmol min−1 nmol P450−1 f 6 pmol min−1 nmol P450−1e 6 pmol min−1 nmol P450−1e 3 pmol min−1 nmol P450−1 f 1 pmol min−1 nmol P450−1 f 760 pmol min−1 nmol P450−1a 200 pmol min−1 nmol P450−1a Km: 100 nM 110 pmol min−1 nmol P450−1a 475 pmol 48 h−1 mg cell pro−1g 45 pmol 48 h−1 mg cell pro−1g 57 pmol min−1 nmol CYP2D25−1 12 pmol min−1 nmol CYP2D25−1 45 pmol min−1 nmol CYP2D25−1 ND ND
C-25 C-1α C-25 C-1α C-26 C-25 C-25
Reference Axén et al. [51]
Axén et al. [53]
Hosseinpour and Wikvall [52] Hosseinpour et al. [32] Hosseinpour et al. [32] Araya et al. [57]
Cheng et al. [9]
a Liver f Partially
microsomes. b Reconstituted system. cIn the presence of a cytosolic factor. d Freshly isolated hepatocytes. e Partially purified liver microsomal P450. purified kidney microsomal P450. g Kidney pCYP2D25. *5β-cholestane-3α,7α,12α-triol; **5β-cholestane, 3α,7α,12α-diol. ND, Not determined.
TABLE IV Species
Sex
Rat (untreated)
Substrate D3 1αOHD3 DHT3 C-triol* D3 D3
The Mitochondrial Vitamin D 25-Hydroxylase Hydroxylation Site
Rat
D3 D2 D3
C-25 C-25 C-25 C-26(27) C-25 C-25 D3 C-triol* C-25 C-25 C-25 C-25 C-25
Rat
D3
C-25
(Phenobarbital treated) (Rachitagenic diet) Human Human Rat Human
and —
D3
C-25 Rabbit
D3 1αOHD3 25OHD3 C-triol*
C-25 C-25 C-1α C-27
Characteristics 0.02 nmol min−1 nmol P450−1b 0.116 nmol min−1 nmol P450−1b 0.01 nmol min−1 nmol P450−1b 8.7 nmol min−1 nmol P450−1b 0.03 nmol min−1 nmol P450-1b 0.064 nmol min−1 nmol P450−1b 0.16 nmol min−1 nmol P450−1b 40 nmol min−1 nmol P450−1b Km: 10−5 Ma Km: 2 × 10−5 Ma,c 10 pmol min−1 mg pro−1a 2 pmol min−1 mg pro−1a 0.093 nmol hr −1 mg pro−1a 0.435 nmol hr −1 mg pro−1a 550 pmol min −1 mg pro−1a (305 pmol min nmol P450−1a) 525 pmol min nmol P450−1a 350 pmol min nmol P450−1a 395 pmol min nmol P450−1b 1200 pmol min nmol P450−1b < 10 pmol min nmol P450−1b 2500 pmol min nmol P450−1b
Reference Björkhem et al., 1980 [110]
Oftebro et al., 1981 [83] Saarem et al., 1984 [84] Saarem and Pedersen, 1985 [85] Holmberg et al., 1986 [176] Saarem and Pedersen, 1987 [35] Dählback and Wikvall, 1987 [46]
Dählback and Wikvall, 1988 [111]
Continued
52
MARIELLE GASCON-BARRÉ
TABLE IV Species
Sex
Substrate
The Mitochondrial Vitamin D 25-Hydroxylase—Cont’d Hydroxylation Site
Rat
1αOHD3
C-25
Rat Rat
hCYP27A1
C-triol* D3 1αOHD3 C-triol* D2 D3 DHD3 1αOHD2 1αOHD3 1αOHDHT3 D3 1αOHD3 25OHD3 25OHD3 C-triol* 25OHD3
hCYP27A1
D3
C-27 C-25 C-25 C-27 C-25 C-25 C-25 C-25 C-25 C-25 C-25 C-25 C-1α C-27 C-27 C-1α C-27 C-25
25OHD3
C-1α
1αOHD3
C-25 (major) C-24 (minor) C-26 (minor) C-24 (major) C-26 (minor) C-24 (major) C-26 (major) C-25 (minor) C-26 (major) C-25 (minor)
Cos cellsc
E. colid
HepG2
1αOHD2 1αOHD4 24-epi1α OHD3 ∆22-1αOHD3
Characteristics
Reference
Km: 54 µMb 3.8 nmol min−1 nmol P450−1b Km: 6.3 µMb 0.36 nmol min−1 nmol P450−1b 1.4 nmol min−1 nmol P450−1b 36 nmol min−1 nmol P450−1b N/Db 342 pmol hr−1106 cellsb 344 pmol hr−1106 cellsb 16 pmol hr−1106 cellsb 1328 pmol hr−1106 cellsb 1383 pmol hr−1106 cellsb 72 pmol min−1 nmol P450−1b 189 pmol min−1 nmol P450−1b 24 pmol min−1 nmol P450−1b 19 pmol min−1 nmol P450−1b 4337 pmol min−1 nmol P450−1b 5 pmol hr nmol hr min−1 CYP27A1−1 4.5 pmol hr nmol hr min−1 CYP27A1−1 Km = 3.2 µM Vmax = 0.27 mol CYP27A1−1 min−1 Km = 3.5 µM Vmax = 0.021 mol CYP27A1−1 min−1
Masumoto et al., 1988 [80] Okuda et al., 1988 [76] Ohyama et al., 1991 [81]
Guo et al., 1993 [87]
Axén et al., 1994 [99]
Araya et al., [57] Sawada et al., [101]
Byford et al., [105]
C-24 (major) C-26 (major) C-25 (minor)
a Mitochondria; b Reconstituted system. c In the presence of a cytosolic factor. **5β-cholestane, α,7α,12α-diol. N/D, Not detectable.
not D2 [43], further demonstrating the broad spectrum of activity of the enzyme. Sex-related differences have also been observed in the microsomal C-25 hydroxylation of D3. Indeed, Hayashi et al. [43] isolated a D3 25-hydroxylase exhibiting the same N-terminal amino acid sequence as CYP2C11 [44], a P450 known to exist in male rat liver microsomes but not in those of females. This observation may explain why microsomes (or P450 purified from the microsomal fraction) obtained from female livers have consistently been shown, as indicated in Table III, to exhibit significantly less activity in the C-25 hydroxylation of D3 than preparations
d Transfected
with hCYP27. *5β-cholestane-3α,7α,12α-triol;
obtained from their male counterparts [43,45–48] except, possibly, at low substrate concentrations as reported by Thierry-Palmer et al. [49]. Whether Yoon and DeLuca [37] and Andersson et al. [38] reported on the same isoenzyme as Hayashi et al. [43] is not known, as no amino acid sequence was proposed in the former studies, but Andersson and Jörnvall [45] later proposed a partial sequence suggestive of the malespecific testosterone 16α-hydroxylase, which is now known to be a P450 compatible with the male-specific CYP2C11. Using male pig microsomes, Segura-Aguilar [50] reported that the C-25 hydroxylation of D3 is also
CHAPTER 4 The Vitamin D 25-Hydroxylase
catalyzed by a peroxidase as indicated by the use of t-butyl hydroperoxide as electron donor in the absence of NADPH and NADPH-P450 reductase. The Km of the enzyme was found to be 30 µM, a concentration much higher than the reported values in the presence of NADPH and NADPH-P450 (see Table III).
A. The Porcine CYP2D25 Purification of the porcine microsomal D3 25-hydroxylase was achieved between 1992 and 1994 by Axén et al. [31,51]. At the time, the protein was only found in liver and kidney. However, the monoclonal antibody raised against the porcine enzyme also recognized a protein of similar molecular weight in human liver but, to date, the pig homolog of the enzyme has not been cloned in any other species. The purified enzyme was shown to catalyze the C-25 hydroxylation of vitamin D3 and vitamin D2 and both compounds competitively inhibited the C-25 hydroxylation of one another. Moreover, ketoconazole, a nonselective P450 inhibitor, also inhibited the hydroxylation of D3 as well as of D2 [51]. The enzyme has been shown to have a high affinity for D compounds and converts D3 into 25OHD3 with a reported Km of 100 nM, a concentration considered to be within the physiological range. Its order of efficiency at saturating concentrations (25 µM) is reported to be 1αD3 > D3 > D2 with 750, 200 and 110 pmol/nmol P450/min respectively [52]. A year later, Axén [31] showed that the enzyme also catalyzed the hydroxylation of 25OHD3 at C-1α, indicating that the enzyme was not only a D3 25-hydroxylase but also a microsomal 1α-hydroxylase. Interestingly, the enzyme was also shown to exhibit similar 1α-hydroxylation activity in the liver of normal pigs and of pigs suffering from pseudovitamin D deficiency rickets (PDDR) (see Chapter 71), but not in the kidney where the 1α-hydroxylation was significantly decreased by the PDDR genotype (Table III) [53]. The molecular cloning of the porcine hepatic microsomal D3 25-hydroxylase was achieved by Wikvall and his group in 1997 [8], and the gene was subsequently shown to map to chromosome 5 [54]. The gene transcript codes for a protein of 500 amino acids and has a predicted molecular weight of 56,374. At that time, Northern blot analysis revealed that the gene was found in high abundance in the liver, but the analysis also showed a clearly detectable signal in the kidney. Subsequent studies using RT-PCR revealed that the transcript is expressed in several tissues including the intestine, lung, thymus, adrenals, and the brain (Table II) [52]. Transfection in simian COS cells results in the synthesis of an enzyme that is recognized by a monoclonal antibody raised against purified D 25-hydroxylase and
53 catalyzes the C-25 hydroxylation of D. The gene transcript exhibits a 70-80% identity with members of the CYP2D family of P450 and has been termed CYP2D25. CYP2D25 has unique properties among the CYP2D enzymes. Indeed, although it metabolizes the drug tolterodine, a muscarinic receptor antagonist and a substrate for the human CYP2D6, into its 5-hydroxymethylated metabolite, tolterodine also inhibits the 25-hydroxylation of D3 while quinidine, an inhibitor of CYP2D6, does not significantly inhibit the C-25 hydroxylation of D3 [52]. Moreover, it was found that the residues in substrate recognition site 3 (SRS-3) are important determinants for its function as a D 25-hydroxylase but not for its activity in the metabolism of tolterodine [55]. Surprisingly, cyclosporine A, a known inhibitor of the mitochondrial D3 25-hydroxylase, was also shown to significantly inhibit the C-25 hydroxylation activity of CYP2D25 [52]. Phorbol 12-myristate 13-acetate (PMA) down-regulates the expression of CYP2D25 and the subsequent C-25 hydroxylation of 1αOHD3, suggesting the involvement of protein kinase C in the regulation of CYP2D25 [56]. The molecular cloning of the microsomal D3 25-hydroxylase in pig kidney was achieved by Hosseinpour et al. in 2002 [32]. Genomic Southern analysis suggests the presence of a single gene for CYP2D25 in the pig and the DNA sequence and deduced peptide sequence of the renal enzyme are homologous to those of the hepatic CYP2D25. Immunocytochemistry indicates that the enzyme is expressed almost exclusively in cells of the cortical tubules as illustrated in Fig. 1. The kidney enzyme has been shown to catalyze not only the hydroxylation of D3 at C-25 but also that of 25OHD3 at C-1α [32,57]. Moreover, recombinant CYP2D25 not only hydroxylates D3 at C-25 but also converts 25OHD3 into both 1α,25(OH)2D3 and 25,26(OH)2D3 with rates of conversion of 57, 12, and 45 pmol/nmol CYP2D25/min, respectively [57]. Microsomes obtained from human livers have been shown to hydroxylate 1αOHD3 at C-25. They, however, exhibit a wide range of activities (0.7 to 4.7 pmol/ min/mg protein). Despite a high degree of sequence identity between CYP2D25 and the human CYP2D6 (77% in its primary structure), no correlation between the microsomal CYP2D6 content and the capacity to hydroxylate 1αOHD3 at C-25 could be demonstrated [52]. In addition, the same group showed that CYP2D6 overexpressed in insect microsomes did not reveal any C-25 hydroxylation activity. Collectively, these observations clearly indicate that CYP2D6 is not a D 25-hydroxylase in human and definitive identification of the microsomal D3 25-hydroxylase, in species other than the pig, will require further kinetic, immunological, and molecular characterization. However, as indicated later, the recent identification of
54
MARIELLE GASCON-BARRÉ
A
chromosome 11, and the mouse CYP2R1 is reported to be 89% identical in sequence to the human enzyme. The enzyme is sexually dimorphic, is present in highest abundance in the liver and testis (Table II), and is active on both D2 and D3 compounds including 1αOHD3. Its sequence identity with other microsomal D 25-hydroxylases (CYP2D25, CYP2C11) has been reported to vary between 33% and 39% [9]. Biochemical characterization of the enzyme has, however, not yet been published and hence its substrate specificity is not yet known.
B
V. THE MITOCHONDRIAL ENZYME A. The C27 Sterol 27-Hydroxylase
FIGURE 1 Localization of the microsomal D3 25-hydroxylase in pig kidney. Immunohistochemical localization of CYP2D25 in sections of pig kidney incubated with polyclonal antibodies raised against the hepatic CYP2D25. Immunostaining (in red) was seen in the cells of proximal tubules but not in distal tubules (A and B). No staining was detected in glomeruli (B). Control sections not treated with primary antibodies, but otherwise treated in the same way, did not show immunostaining (not shown in figure). Reproduced from Biochimica et Biophysica Acta, 2002, volume 1580, 140 [32], by copyright permission of Elsevier Science B.V.
an orphan receptor of the CYP2 family as a putative D 25-hydroxylase may soon elucidate the role of the endoplasmic reticulum in the hepatic activation of vitamin D in humans.
B. The Human and Mouse CYP2R1 Using a cDNA library from CYP27A1 null mouse liver, Cheng et al. [9] have identified an evolutionarily conserved orphan cytochrome P450 termed CYP2R1 and demonstrated that the biochemical properties of the enzyme were consistent with a D 25-hydroxylase. The CYP2R1 gene transcript has characteristic sequence features associated with cytochromes P450 of the endoplasmic reticulum. It is conserved through evolution from fish to human [58], where it is located on
Bile acid biosynthesis represents the most important pathway in the metabolism and excretion of cholesterol. Although several enzymes participate in the oxidation of the cholesterol molecule, the C27 sterol hydroxylase catalyzes the first and probably the ratelimiting step in the oxidation of the side chain in the “acidic” bile acid biosynthesis pathway [59]. In accordance with IUPA sterol nomenclature [60] and the stereochemistry of the initial reaction catalyzed by the enzyme [61–63], Cali and Russell [6] termed the enzyme sterol 27-hydroxylase (EC 1.14.13.15), although the enzyme exhibits low substrate specificity and has been referred to as the 26-hydroxylase in most early publications [64]. CYP27A1 has now been adopted as the gene symbol. After considerable debate on the subcellular localization of the 27-hydroxylase [44,65,66], the enzyme was established in 1973 to be located in the inner membrane matrix of liver mitochondria and shown to display properties similar to that of P450, catalyzing hydroxylation in the mitochondrial membrane of the adrenal glands [67,68]. These observations were later confirmed by others [69–71]. Thus, sterol 27-hydroxylase is a mixed-function oxidase requiring the two mitochondrial-specific protein cofactors, ferrodoxin (adrenodoxin), an iron–sulfur protein [72], and ferrodoxin (adrenodoxin) reductase, a FAD-containing enzyme [73]. The enzyme catalyzes the C-27 hydroxylation of 5β-cholestane-3α,7α,12β-triol [70], 5β-cholestane3α,7α-diol [65], cholesterol [74–76], and 7-ketocholesterol [77], as well as several other substrates involved in bile acid biosynthesis. Wikvall [78] purified the enzyme to apparent homogeneity from rabbit liver mitochondria. Okuda et al. [76] using female rat liver mitochondria achieved final purification of the 27-hydroxylase and noticed that the enzyme exhibited no activity toward xenobiotics using benzphetamine, 7-ethoxycoumarin, and benzo[a]pyrene as substrates.
55
CHAPTER 4 The Vitamin D 25-Hydroxylase
In 1975, Björkhem et al. [79] observed the presence of the 27-hydroxylase in human liver mitochondria, and a few years later, the same group was the first to report that rat and human mitochondria catalyzed the C-25 hydroxylation of D3 but, at that time, attributed the D3 25-hydroxylation activity to an enzyme different from the cholestane-triol 27-hydroxylase [7].
B. The Mitochondrial D3 25-Hydroxylase Is the C27 Sterol 27-Hydroxylase Masumoto et al. [80] purified to homogeneity the D3 25-hydroxylase from rat liver mitochondria and characterized it as a P450 catalyzing the C-25 hydroxylation of both D3 and 1αOHD3. These researchers proposed that a single enzyme was involved in the C-27 hydroxylation of 5β-cholestane-3α,7α,12α-triol and in the C-25 hydroxylation of D3, although the preparation exhibited much lower activity toward D3 compounds than toward the C-27 hydroxylation of bile acid intermediates as illustrated in Table IV [76]. Ohyama et al. [81] pursued the hypothesis they earlier put forward [80] and found that D3 competitively inhibited the 27-hydroxylation of C-triol while C-triol inhibited D3 25-hydroxylation. They concluded, based on the criteria proposed by Haldane and Dixon [82], that both substrates were catalyzed at a common active site of a single protein. Meanwhile Oftebro et al. [83] and Saarem et al. [84,85] confirmed the original observation of Björkhem et al. [7] that a D3 25-hydroxylase was present in human liver mitochondria and that C27 sterol 27-hydroxylation as well as D3 25-hydroxylation activities were located probably exclusively in the inner mitochondrial membrane matrix [84,85]. Cloning of the rabbit [4], rat [5,86], human [6,87], and pig [8] CYP27A1 (chromosomal location: 2q33qter [88]) was achieved between 1989 and 1997. Cloning of the rabbit 27-hydroxylase by Andersson and co-workers [4] illustrated the hydrophobic nature of the protein (with 36% of the amino acids being either aromatic or hydrophobic side chains), and the presence of a conserved cysteine residue at position 444 that is believed to be the ligand for the heme ion in P450 enzymes. Andersson’s work also confirmed earlier biochemical observations indicating extrahepatic enzyme activity with the demonstration that CYP27A1 mRNAs were present in multiple tissues as compiled in Table I. The predicted amino acid sequence of the rabbit [4], rat [5], and human [6] mitochondrial D3 25-hydroxylase contains a mitochondrial specific presequence of 32–36 amino acids on which the destination of the hemoprotein is dependent, as illustrated by Sakaki et al. [89] using a modified CYP27A1 whose
mitochondrial targeting signal had been replaced by a microsomal N-terminal targeting signal. The construct, indeed, resulted in the microsomal localization of a fully active enzyme, providing adequate electron transfer was made available to the enzyme. The mature protein contains 444 to 501 amino acids depending on the species involved and the rat enzyme molecular weight has been estimated to be 51,182 [5]. Northern analyses demonstrate two mRNA sizes of 1.9 and 2.3 kb for CYP27A1 in liver and fibroblast cells [6,86]. Characterization of the 2.3-kb mRNA indicated that the sequence was identical to the 1.9-kb mRNA in its protein-coding region, except for a 400-nucleotide extended sequence at its 5′ end [90]. The rat CYP27A1 gene contains 11 exons of 80-415 nucleotides that are separated by 10 introns of 83 bases to 10kb [91]. The gene has been shown to be regulated by several hormones as will be discussed later [90–94]. The protein sequence of the human enzyme has been reported to be 72% identical to the rat and 81% identical to the rabbit CYP27A1 [5,6]. Heterologous cell systems confirmed the requirement for adrenodoxin and adrenodoxin reductase and not NADPH P450 reductase [4] as well as the identity of sterol 27-hydroxylase and D3 25-hydroxylase. Indeed, Usui et al. [95] prepared an expression plasmid encoding the rat liver D3 25-hydroxylase and transfected it in simian COS cells. They observed both sterol C-27 and D3 C-25 hydroxylation activities in a solubilized mitochondrial fraction supplemented with adrenodoxin and NADPH-adrenodoxin reductase. Akiyoshi-Shibata et al. [96] and Sakaki et al. [89] expressed the cDNA encoding the precursor protein of the rat liver D3 25-hydroxylase in Saccharomyces cerevisiae and found that the mitochondrial fraction exhibited both D3 25-hydroxylase and C-triol 27-hydroxylase activities. These data clearly confirmed the identity of the two enzymes and ruled out the possibility that the simian cell line contained endogenous 27-hydroxylase.
C. Specificity of the CYP27A1 CYP27A1 is a high-capacity enzyme involved in the hydroxylation of bile acid intermediates, but its substrate specificity is broad and exceeds the field of bile acid biosynthesis. As discussed ealier, the purified enzyme has been shown to be active on both the C-25 and C-27 hydroxylation of cholesterol [44,64,68], and to hydroxylate 5β-cholestane-3α,7α-diol and 5β-cholestane- 3α,7α,12α-triol at C-27. The enzyme has also been shown to carry out the oxidation of the product of the reaction, 5β-cholestane-3α,7α,12α,27-tetrol, to
56 3α,7α,12α-trihydroxy-5β-cholestanoic acid [97], as well as to substitute for the classical pathway of bile acid biosynthesis involving an NAD-coupled alcohol dehydrogenase. In a manner similar to other mitochondrial P450, CYP27A1 contains a proline-rich sequence located in the amino-terminal region of the molecule that serves to reduce the tendency of the polypeptide to misfold prior to heme binding [98]. The rabbit, rat, human, pig, and bovine enzyme hydroxylates D3 compounds at C-25 as well as at many other positions on the side chain of the molecule [5,86,87,99–101]. Indeed, CYP27A1 has been shown to catalyze the C-1α, 24, 25, 26, and 27 hydroxylation of D3, D2 and related compounds [87,100–105]. In addition, the pig and human kidney CYP27A1 catalyzes the formation of a metabolite expected to be, according to its mass spectrometric properties, 4β,25(OH)2D3 [57], a metabolite already reported by others in the rat [106,107]. While D3 substrates were found to be preferably hydroxylated at C-25, several groups have observed that when the substrate exhibits the ergocalciferol side chain such as that found in D2 or 1αOHD2, a predominance of 24-hydroxy metabolites occurs with some production of 26- and 27-hydroxylated products. These metabolites have also been found in vivo in rat, cow, chicken, and human [87,102,103,108,109]. Surprisingly, D2 is not hydroxylated at position 25 while 1αOHD2 is only poorly hydroxylated as indicated in Table IV. In addition, several laboratories have now shown that the enzyme prefers 1α-hydroxylated analogs of D3 over their nonhydroxylated counterparts, and this includes the natural substrate D3, which is hydroxylated at C-25 several times less efficiently than 1αOHD3 (see Table IV) [81,87,110,111]. The observation that the 27-hydroxylase is far more efficient at hydroxylating compounds exhibiting a 1α-hydroxyl group than toward the natural substrate is in full accordance with Ohyama et al. [81], who observed that the activity of the enzyme toward bile acid intermediates was proportional to the number of hydroxyl groups on the nucleus. Björkhem [64] also raised the hypothesis that the binding of nonpolar substrates such as vitamin D3 or cholesterol to the active site of the enzyme may differ from the corresponding binding of more polar substrates such as 5β-cholestane-3α,7α,12α-triol. Jones and co-workers [104] have also shown that lengthening the side chain of the vitamin D molecule is quite well tolerated in that the enzyme is able to continue to efficiently hydroxylate D3 compounds including its analogs. These studies led to the proposal of a model of interaction between enzyme and substrate whereby the 25-hydroxylase appears to be directed to its terminal hydroxylation site by the distance from the end of the side chain. These data were also taken to
MARIELLE GASCON-BARRÉ
suggest that the substrate binding pocket of CYP27A1 can tolerate a variety of steroidal shapes either by accommodating only the terminal carbons of the side chain, or by having a specific cleft for the side chain within a broader pocket for the vitamin D/steroidal ring structure. Using several side-chain analogs of D3, the same group showed that in the CYP27A1-containing hepatoma cell HepG2, 1αOHD3 is mainly hydroxylated at C-25 with only minor C-24 and C-26 hydroxylations. 1αOHD2, however, is mainly hydroxylated at C-24 with minor hydroxylation at C-26 while both 1αOHD4 and ∆22-1αOHD3 are mainly hydroxylated at C-24 and C-26 with only minor production of their 25-hydroxylated products. These observations led the authors to conclude that the C-24 methyl group plays a crucial role in the determination of the CYP27A1directed C-24 hydroxylation. They also concluded, based on their studies with 1αOHD4 and ∆22-1αOHD3, that the C22–C23 double bond also contributes to the altered metabolism observed with D2 or D2-derived compounds [105]. Lately, Wikvall and co-workers [57,99,112] have reported that the 27-hydroxylase purified from pig and rabbit livers as well as recombinant human CYP27A1 expressed in Escherichia coli or monkey COS cells was also able to catalyze the 1α-hydroxylation of 25OHD3, albeit at a much lower rate than that observed for the conversion of D3 into 25OHD3. Interestingly, 25OHD3 1α-hydroxylase activity has been reported previously in pig [113] and fish [114] as well as in fetal and adult rat livers [115,116]. Whether or not the reported activity was due to the 27-hydroxylase is unknown as the adult rat liver 1α-hydroxylation activity was found not in the mitochondrium but in the microsomal fraction [116].
D. Regulation of CYP27A1 and Enzyme Activity CYP27A1 mRNA t1/2 has been reported to be between 13 and 24 hr in the rat liver as well as in rat hepatocytes [20,117]. In addition, many hormones, endogenous products, hepatocyte nuclear factors (HNFs), and xenobiotics have been shown to regulate the gene encoding CYP27A1 as well as its gene product. They include bile acids, growth hormone, IGF-1, glucocorticoids, insulin, 1,25(OH)2D3, thyroxine, cyclosporine A, protease inhibitors, fibrates, lipopolysaccharides (LPSs), TNFα, IL-1β, HNF1, HNF4α, and the physiological state of the animal [20,86,91,93,94,118–126]. In the rat [92,93] but not the rabbit [127], CYP27A1 is highly sensitive to the prevailing concentrations of bile acids with increases in enzyme activity, steady-state
CHAPTER 4 The Vitamin D 25-Hydroxylase
mRNA levels, and the rate of gene transcription following interruption of the enterohepatic circulation of bile acids [93,118] while increases in bile acid concentrations are known to down-regulate CYP27A1 [93,126,128,129]. 7α(OH)-4-Cholesten-3-one is an inhibitor of CYP27A1 in the pig and, in a manner similar to the microsomal CYP2D25, PMA down-regulates the expression of CYP27A1, an observation suggesting the participation of protein kinase C in the regulation of its transcription [56]. Bile acids have been reported to suppress CYP27A1 transcription by decreasing its binding to HNF-1, a transcription factor expressed in liver, intestine, pancreas, and kidney [129,130]. Similarly, HNF4α, which is a strong stimulator of CYP27A1 gene transcription, is also down-regulated by bile acids. It has, therefore, been postulated that the inhibitory effect of bile acids on CYP27A1 could be mediated through a repression of HNF4α gene transcription and protein expression rather than through a direct effect on the bile acid response element (BARE). Cholic acid has also been reported to affect CYP27A1 mRNA stability [122]. In accordance with the latter observations, Bolt et al. [131] have shown that cholestasis induced by bile-duct ligation in rats significantly inhibited the 25-hydroxylation of D3 and that addition of bile salts to liver homogenates also had an inhibitory effect on the enzyme. By contrast, biliary cirrhosis does not seem to directly affect the total hepatic output of 25OHD3 in vivo. Indeed, Plourde et al. [132] observed by directly sampling the hepatic effluent after intraportal D3 injection that the 25-hydroxylation of D3 was similar in dogs with interruption of the enterohepatic circulation by choledococystostomy anastomosis and biliary cirrhosis secondary to bile-duct ligation. These observations still leave open any definitive conclusion as to the effect of bile salts on the C-25 hydroxylation of D3 by the mitochondrial and/or microsomal enzyme(s) in vivo (see Chapter 75). The role of insulin in CYP27A1 was illustrated when bile acid production via the sterol 27-hydroxylase was shown to be inhibited up to 58% after 24 hr incubation of hepatocytes with 140 nmol/liter insulin [94]. The decrease in enzyme activity could be explained by a concomitant reduction in CYP27A1 (–62%) mRNA level as well as in transcriptional activity (–75%) [94]. Thus physiological concentrations of insulin seem to down-regulate CYP27A1 gene transcription through a direct effect of the hormone on the hepatocytes [94], although the exact locus of interaction of insulin within the gene has not been identified. Avadhani’s group first reported regulation of CYP27A1 by pituitary-regulated steroids, growth hormone, and the diurnal rhythm with a twofold increase in enzyme activity observed in the mid-dark compared
57 to the mid-light period [93]. They subsequently showed that CYP27A1 is regulated by glucocorticoids as illustrated by a sevenfold induction in mRNA levels in isolated hepatocytes exposed to dexamethasone [117]. The response to the drug is now known to be due to a dexamethasone response element located in the promoter region of the CYP27A1 gene, which has been mapped to positions between −792 and −1095 bp [122,126]. In addition to glucocorticoids up-regulating the enzyme, other drugs known to regulate CYP27A1 include cyclosporine A, protease inhibitors, and fibrates (see also Chapters 73 and 74). At the enzyme level, cyclosporine A has been shown to inhibit the C-25 hydroxylation of both D3 and cholesterol through a direct inhibition of the gene encoding the sterol 27-hydroxylase [133,134]. Indeed, a cyclosporine A response element has been identified on the CYP27A1 gene promoter at 1087–4000 bp upstream of the putative ATG start site [122]. Protease inhibitors have also been shown to decrease the C-25 hydroxylation of D3 in a human hepatocyte cell line, suggesting that patients on antiviral therapy may have an increased risk of 25OHD insufficiency [123]. The fibrate hypolipidemic drugs, on the other hand, have been reported to decrease bile acid synthesis in rat hepatocytes and to suppress CYP27A1 mRNA and enzyme activity by transcriptional and posttranscriptional mechanisms involving the peroxisome proliferator–activated receptor α (PPARα) [124]. The presence of regulatory mechanisms related to the vitamin D status as modulators of the 25-hydroxylation of the vitamin has long been a widely accepted concept [30,33]. In addition, in the early 1980s, Baran and Milne [135] and Bell et al. [136] put forward the hypothesis that 1,25(OH)2D3 might be an inhibitor of the D3 25-hydroxylase. Later studies indicated that 1,25(OH)2D3 administration influenced the in vivo handling of D3 by accelerating its metabolism as well as by increasing its metabolic and biliary clearances, but the molecular mechanisms involved in the regulation of CYP27A1 by the D status were investigated much later [12,137–139]. Axén et al. [112] first reported that CYP27A1 located in kidney and liver was affected by 1,25(OH)2D3 administration, but they showed that CYP27A1 mRNA was decreased to a greater extent in the kidney than in the liver. On the other hand, Theodoropoulos et al. [20,119] investigated the effect of D3 nutritional status as well as that of 25OHD3 and 1,25(OH)2D3 on the rat hepatic and intestinal CYP27A1 and its associated enzyme activity. The hepatic CYP27A1 gene transcript was not found to be sensitive to the D3 or 25OHD3 status but was highly sensitive to 1,25(OH)2D3 exposure with a 60% decrease in its steady-state mRNA levels
58
MARIELLE GASCON-BARRÉ
and a parallel decrease in its protein and enzyme activity (Figs. 2 and 3) [20]. The effect of 1,25(OH)2D3 was shown to be mediated by a significant decrease in the CYP27A1 transcription rate, whereas its mRNA t1/2 remained unchanged. By contrast, intestinal CYP27A1 mRNA levels were shown to be sensitive to D3 and 25OHD3 as well as 1,25(OH)2D3 exposure. 1,25(OH)2D3 administration resulted in a decrease in both intestinal CYP27A1 mRNA t1/2 and transcription rate, and CYP27A1 enzyme activity was shown to be significantly decreased after 1,25(OH)2D3 administration in vivo [119]. In addition, in the human fetal intestine kept in primary culture, 1,25(OH)2D3 was shown to decrease CYP27A1 mRNA levels by 29% after 48 hr of incubation in the presence of the hormone [120]. Collectively, these data indicate that 1,25(OH)2D3
significantly influences CYP27A1 mRNA levels and enzyme activity in the major D-metabolizing organs as well as in target organs. It also illustrates that organspecific differences are present in the regulation of the CYP27A1 gene transcript and response to D3 or its metabolites. Interestingly, Chen and Chiang [126] have suggested that the vitamin D receptor (VDR) (which has now been shown to be present in the liver [140]) could be a mediator in the inhibition of CYP27A1 by conjugated bile acids.
E. Intraacinar Localization of CYP27A1 Several studies have indicated that the CYP27A1 gene transcript is heterogeneously distributed within the
A CYP27A
18S
B CYP27A/18S (Arbitary Units)
1.0
0.8
0.6
0.4
0.2
0.0 0
1
3
5
7
Length of repletion with 1,25 (OH)2D3 (days)
FIGURE 2
Influence of 1,25(OH)2D3 administration to vitamin D–depleted rats on the hepatic D3 25-hydroxylase gene transcript rCYP27A1. Steady-state levels of hepatic rCYP27A1 mRNA levels in hypocalcemic D-depleted rats repleted with 1,25(OH)2D3 after 1 to 7 days of repletion (28 pmol/day). (A) Representative Northern blot analyses of the CYP27A1 gene transcript. (B) Densitometric analyses. Data are presented as means ± SEM. Statistically significant differences between group means were analyzed by ANOVA, with individual contrasts evaluated by the Bonferroni post hoc test. Main effect P <0.001. Statistically significant different from D-depleted rats: *P<0.0001, **P<0.0001; n = 3 animals/group. Reproduced from the American Journal of Physiology (Endocrinology and Metabolism), 2003, volume 284, p. E143 [20] by copyright permission of the American Physiological Society.
59
CHAPTER 4 The Vitamin D 25-Hydroxylase
A
D
G
B
E
H
C
F
I
FIGURE 3 Influence of 1,25(OH)2D3 administration to vitamin D depleted rats on the hepatic intra-acinar distribution of the D3 25-hydroxylase gene transcript CYP27A1. Representative photomicrograph of rat liver sections obtained after in situ hybridization with an rCYP27A antisense riboprobe. A 10-cycle RT-PCR amplification was used. Liver sections were obtained from D-depleted rats (A and B), normal rats (D and E), and 7-day 1,25(OH)2D3-repleted rats (28 pmol/day) (G and H). Negative in situ hybridization control sections with the rCYP27A sense riboprobe are presented in C, F, and I for livers obtained from D-depleted, normal, and 1,25(OH)2D3-repleted rats, respectively. Reproduced from the American Journal of Physiology (Endocrinology and Metabolism), 2003, volume 284, p. E144 [20] by copyright permission of the American Physiological Society.
hepatic acinus with higher perivenous than periportal mRNA levels in both rat and human livers [118,141]. In the human liver, in situ hybridization illustrates a clear CYP27A1 label in hepatocytes that increases in intensity in the perivenous region of the hepatic acinus [141]. In addition, regulation studies in the rat have shown increased CYP27A1 gene transcription, mRNA levels, and production of 27-hydroxycholesterol as well as a recruitment of hepatocytes expressing CYP27A1 toward the mid- to periportal region of the liver acinus when the enterohepatic circulation was interrupted by a bile salt sequestrant [93,118]. These data clearly indicate that CYP27A1 is subject to feedback regulation (in intensity and intrahepatic regionalization) at the transcription level by bile acids returning via the portal blood [93,118]. A study on the effect of the vitamin D status has, however, shown that in the vitamin D–depleted rat liver, 1,25(OH)2D3 decreased CYP27A1 in both the periportal and perivenous regions (Fig. 3) [20]. Thus the impact of 1,25(OH)2D3 on the CYP27A1
gene transcript is more widely distributed across the hepatic acinus than that of bile acids.
VI. CEREBROTENDINOUS XANTHOMATOSIS Mutations that affect either the level of expression of CYP27A1 or its primary sequence lead to cerebrotendinous xanthomatosis (CTX) [142,143]. CTX is an autosomal recessive inherited inborn error of bile acid biosynthesis leading to defects in sterol metabolism and storage characterized by xanthomatosis, atherosclerosis, progressive neurological dysfunction, and dementia. In these patients, however, both abnormal [143–146] and normal [147] vitamin D and/or calcium metabolism have been reported, which gives credence to the presence of more than one enzyme active in the 25-hydroxylation of D in humans, or to a contra-up-regulation of CYP27A1 in cases of mild mutations [148]. Indeed, some CTX
60 patients have been shown to retain residual 27-hydroxylase activity [44], while the human cell line Hep3B has been shown to efficiently activate 1αOHD3 at C-25 despite the absence of detectable CYP27A1 mRNA when evaluated by Northern analysis [149,150]. Rosen et al. [151] were the first to report on a mouse with a disrupted CYP27A1 (CYP27A1−/−). Surprisingly, despite several biochemical markers indicating disruption of bile acid and cholesterol metabolism, no CTXrelated pathological abnormalities were observed in the mice, an observation also reported by others [152]. In addition, CYP27A1−/− mice were shown to have supranormal 25OHD3 (106 compared to 43 ng/ml in CYP27A1+/+) and normal 1,25(OH)2D3 serum concentrations. The latter observations suggest that in the mouse, as in many other species, D3 25-hydroxylase(s) other than CYP27A1 are active in the endogenous synthesis of 25OHD3, an observation that has been confirmed by the identification of CYP2R1 as a D 25hydroxylase in the liver of CYP27A1−/− mice [9]. These observations also suggest that, in the CYP27A1−/− mice, an up-regulation of these enzyme(s) takes place. However, marked species differences in the mechanisms controlling the synthesis of bile acid and cholesterol metabolism have been documented [77,152–154]. Indeed, in contrast to patients suffering from CTX, CYP27A1−/− mice exhibit a dramatic increase in CYP3A activity. This leads to a marked up-regulation in the sidechain oxidation of bile acid biosynthesis, an increase not observed in CTX despite accumulation of cholestane 3α,7α,12α-triols [154]. The increase in CYP3A is secondary to the elevation of toxic bile acid intermediates in the CYP27A1−/− mouse which are potent activators of the xenobiotic/pregnane X receptor (PXR) in the mouse but only weak activators of PXR in humans. The increase in PXR acts as a feed-forward regulatory loop by which toxic bile acid intermediates activate PXR and induce their own metabolism through a PXRmediated induction of CYP3A. Because of the reported differences in bile acid and cholesterol metabolism between a mouse and human with disrupted CYP27A1, the CYP27A1−/− mouse may, therefore, not be an adequate model of the CTX-related defects in D3 anabolism observed in patients stricken with the disease. These observations also illustrate that D 25-hydroxylases other than CYP27A are most likely present in human as suggested by the putative role of the microsomal CYP2R1 in the C-25 hydroxylation of the vitamin and the capacity of CYP3A4 to act as a D 25-hydroxylase [9,58]. In additional support for this hypothesis, a study reported by Byford et al. [105] has indicated that the human CYP27A1-negative cell line Hep 3B is able to transform [3H]1αOHD4 into its 25-hydroxylated
MARIELLE GASCON-BARRÉ
metabolite at low nanomolar concentration, further suggesting the presence of a D 25-hydroxylase other than CYP27A1 in human.
VII. ONTOGENY OF THE VITAMIN D 25-HYDROXYLASES The fetal rat liver can adequately sequester D3 [155]. Moreover, when compared to values obtained in the late fetal period, D3 uptake decreased significantly during the first 2 weeks post partum and then dramatically increased (sixfold) between days 14 and 22 post partum when it reached an uptake capacity similar to that previously observed in the adult rat. The liver of 18- (3 days before birth), 19-, and 22-day-old rat fetuses hydroxylates D3 at C-25 and the value was shown to remain unchanged at day 1 after birth [156,157]. In one study [156], the microsomal 25-hydroxylase activity was shown to sharply increase from day 2 to 11 and to remain unchanged until day 60 of chronological age, whereas in another study [157], maximal enzyme activity was reached only near weaning. These authors also noticed that a cytosolic factor was necessary for optimum 25-hydroxylase activity in fetuses, neonates, and adult rats. In that context, it has been suggested that cytosol contains a cellular D binding protein [158] that could facilitate access of the substrate to the enzyme as proposed by Tsankova et al. [159]. In the pig, expression of the microsomal CYP2D25 has been shown to be similar in the newborn and adult liver, but its expression was found to be lower in the newborn than in the adult kidney [32]. The data suggest a tissue-specific developmental regulation of CYP2D25 in the pig. In contrast, the CYP27A1 gene transcript can easily be detected in the human fetal liver, kidney, jejunum, and colon. Moreover, the CYP27A1 gene transcript was shown to be regulated by 1,25(OH)2D3 in the fetal intestine [120,141]. However, in 16- to 20-week-old human fetuses, the level of the CYP27A1 gene transcript was found to be approximately 40% lower in the fetal liver than in the adult liver [141], but the level of expression of the fetal human kidney CYP27A1 gene transcript was similar to that of the adult kidney [141]. In human, 25OHD has been shown to be lower in cord than in maternal serum, although a significant relationship was observed between the two sites [160]. Premature infants have low circulating 25OHD concentrations compared to infants born at term [161,162], which rapidly increases, however, in the subsequent 4 days after birth [162], indicating fully operational D 25-hydroxylase activity in preterm infants.
61
CHAPTER 4 The Vitamin D 25-Hydroxylase
VIII. SEX DIFFERENCES IN THE HYDROXYLATION OF VITAMIN D AT C-25 Despite the observation that the circulating 25OHD concentrations are not significantly different in males and females, several studies indicate that both the microsomal and the mitochondrial enzyme activities may be gender specific and/or regulated by sex hormones. Indeed, CYP2C11, which exhibits D3 25-hydroxylase activity, is known to be male specific in the rat and rabbit but not in human [88]. Not surprisingly, the enzyme activity reported in the rat has been largely shown to be higher in males than females [45,46]. In addition, Hayashi et al. [48] concluded, based on immunoprecipitation techniques, that the D3 25-hydroxylase found in the male rat liver was a different entity from that found in the female, suggesting that a microsomal P450 distinct from CYP2C11 is also able to catalyze the 25-hydroxylation of vitamin D in rodents. By contrast, the mitochondrial enzyme CYP27A1 has been shown in the human and rat liver to be more abundant in females than in males with a 43% higher expression in female than in male rat livers, although the regulation mechanisms of the enzyme by sex hormones is not presently known [47,141,163]. In addition, Saarem and Pedersen [47] observed that both enzyme activities increased when estradiol was injected into male rats while testosterone decreased the activities of both enzymes in female rats. Addya et al. [163] also reported that the relative 25-hydroxylase levels were reduced in castrated females but increased in castrated males. Although regulatory studies have not yet been carried out on the microsomal CYP2R1, investigation of the levels of its gene transcript in male and female mouse livers indicates the absence of sexual dimorphism [9].
IX. CONCLUSIONS In liver, both microsomes and mitochondria harbor monooxygenase systems active on the 25-hydroxylation of the endogenous substrate D3, but also on several analogs of the vitamin. This allows the design of molecules with therapeutic potential that can be activated by a single hydroxylation step by the liver. The mitochondrial sterol 27-hydroxylase is a multifunctional enzyme exhibiting high capacity with a reported Km for D3 several times higher (micromolar range) than the circulating steady-state concentration of the substrate in vivo. The dynamic liver regulation of enzyme distribution associated with uptake of the substrate favoring selective acinar accumulation seems
to allow, however, the necessary conditions within the hepatocyte microenvironment for efficient transformation of D into 25OHD in vivo. In addition, the presence of CYP27A1 in multiple extrahepatic tissues and organs is suggestive of an enzyme system displaying pleiotrophic functions that are able to support the C-25 hydroxylation of the vitamin in target organs. Whether or not these sites participate in the toxicity of the vitamin, or act, in a tissue-specific manner, for the in situ delivery of 25OHD to the renal and/or the extra-renal 1α-hydroxylase as observed in vitro, remains to be investigated. The microsomal D 25-hydroxylase was originally thought to be a high-specificity low-capacity enzyme. It has now been shown that the enzymes described to date also have low substrate specificity and exhibit regulation by sex hormones and preferential intraacinar localization to the perivenous area (CYP2C11) [164]. The porcine hepatic and kidney CYP2D25 has been characterized and shown to be a high-affinity enzyme and to hydroxylate D3 and D2 compounds at both C-25 and C-1α. The human and mouse CYP2R1 has also been reported to activate D3 and D2 compounds. However, the relative importance of the microsomal and mitochondrial enzyme in the C-25 hydroxylation of vitamin D is not presently known. Indeed, Hosseinpour et al. [56] in a study in pig hepatocytes that possess both CYP2D25 (microsomes) and CYP27A1 (mitochondria) activities found that both enzymes equally participate in the 25-hydroxylation of D3, whereas Cheng et al. [9] reported that CYP27A1 was more abundantly expressed than CYP2R1 (microsomes) in the mouse liver, but the comparative contribution of each enzyme to the formation of 25OHD is not yet known. In conclusion, the liver is essential to the normal homeostasis of vitamin D. Indeed, it activates the vitamin of endogenous and exogenous origin, and efficiently exports 25OHD to the systemic circulation where its concentration represents the best marker of the vitamin D nutritional status. The presence of D microsomal and mitochondrial 25-hydroxylases (which also exhibit C-1α activity) in extrahepatic tissues also raises the hypothesis of an auto/para/intracrine system of D action in D target organs.
X. ADDENDUM While this book was in press, the biological relevance of CYP2R1 as a D 25-hydroxyalse was illustrated by the elucidation by Cheng et al. [177] of a molecular defect in the CYP2R1 gene which was found to be responsible
62 for low serum 25OHD and the classical symptoms of rickets. The affected individual had a normal CYP27A1 gene but was found to be homozygous for a transition mutation in exon 2 of the CYP2R1 gene which resulted in the elimination of D 25-hydroxylase enzyme activity. These observations establish CYP2R1 as a biologically important D 25-hydroxylase and illustrate, for the first time, that a mutation in the CYP2R1 gene which abolishes enzyme activity can lead to a rare molecular autosomal recessive disorder, a selective 25OHD deficiency.
References 1. Blunt JW, DeLuca HF, Schnoes HK 1968 25-hydroxycholecalciferol. A biologically active metabolite of vitamin D3. Biochemistry 7:3317–3322. 2. Ponchon G, Kennan AL, DeLuca HF 1969 “Activation” of vitamin D by the liver. J Clin Invest 48:2032–2037. 3. Olson EB Jr., Knutson JC, Bhattacharyya MH, DeLuca HF 1976 The effect of hepatectomy on the synthesis of 25-hydroxyvitamin D3. J Clin Invest 57:1213–1220. 4. Andersson S, Davis DL, Dahlback H, Jornvall H, Russell DW 1989 Cloning, structure, and expression of the mitochondrial cytochrome P-450 sterol 26-hydroxylase, a bile acid biosynthetic enzyme. J Biol Chem 264: 8222–8229. 5. Usui E, Noshiro M, Okuda K 1990 Molecular cloning of cDNA for vitamin D3 25-hydroxylase from rat liver mitochondria. FEBS Lett 262:135–138. 6. Cali JJ, Russell DW 1991 Characterization of human sterol 27-hydroxylase. A mitochondrial cytochrome P-450 that catalyzes multiple oxidation reaction in bile acid biosynthesis. J Biol Chem 266:7774–7778. 7. Bjorkhem I, Holmberg I 1978 Assay and properties of a mitochondrial 25-hydroxylase active on vitamin D3. J Biol Chem 253:842–849. 8. Postlind H, Axén E, Bergman T, Wikvall K 1997 Cloning, structure, and expression of a cDNA encoding vitamin D3 25-hydroxylase. Biochem Biophys Res Commun 241: 491–497. 9. Cheng JB, Motola DL, Mangelsdorf DJ, Russell DW 2003 De-orphanization of cytochrome P450 2R1. A microsomal vitamin D 25-hydroxylase. J Biol Chem 278: 38084–38093. 10. Gascon-Barré M, Huet PM 1982 Role of the liver in the homeostasis of calciferol metabolism in the dog. Endocrinology 110:563–570. 11. Gascon-Barré M, Vallières S, Huet PM 1986 Influence of phenobarbital on the hepatic handling of [3H]vitamin D3 in the dog. Am J Physiol (Gastrointest Liver Physiol) 251:G627–G635. 12. Gascon-Barré M, Gamache M 1991 Contribution of the biliary pathway to the homeostasis of vitamin D3 and of 1,25-dihydroxyvitamin D3. Endocrinology 129:2335–2344. 13. Rojanasathit S, Haddad JG 1976 Hepatic accumulation of vitamin D3 and 25-hydroxyvitamin D3. Biochim Biophys Acta 421:12–21. 14. Gascon-Barré M, Vallières S, Huet PM 1986 Uptake of the hormone 1,25-dihydroxyvitamin D3 by the dog liver. Can J Physiol Pharmacol 64:699–702.
MARIELLE GASCON-BARRÉ
15. Haddad JG, Jennings AS, Choon Aw T 1988 Vitamin D uptake and metabolism by perfused rat liver: Influences of carrier proteins. Endocrinology 123:498–504. 16. Dueland S, Holmberg I, Berg T, Pedersen JI 1981 Uptake and 25-hydroxylation of vitamin D3 by isolated rat liver cells. J Biol Chem 256:10430–10434. 17. Dueland S, Helgerud P, Pedersen JI, Berg T, Drevon CA 1983 Plasma clearance, transfer, and distribution of vitamin D3 from intestinal lymph. Am J Physiol 245:E326–E331. 18. Gascon-Barré M, Elbaz H, Therrien-Ferland D 1985 Sequestration and microsomal C-25 hydroxylation of [3H]vitamin D3. Metabolism 34:244–250. 19. Ziv E, Bar-On H, Silver J 1985 Vitamin D3 uptake by the isolated perfused rat liver from lipoprotein fractions is separate from cholesterol and triglyceride uptake. Eur J Clin Invest 15:95–99. 20. Theodoropoulos C, Demers C, Petit JL, Gascon-Barré M 2003 High sensitivity of the rat hepatic vitamin D3-25 hydroxylase CYP27A to 1,25-dihydroxyvitamin D3 administration. Am J Physiol 284:E138–E147. 21. Gascon-Barré M, Vallières S, Benbrahim N 1992 C-25 hydroxylation of vitamin D3 in the periportal and perivenous region of the hepatic acinus. Am J Physiol (Endocrinol Metab) 262:E810–E817. 22. Gascon-Barré M, Huet PM, St-Onge Brault G, Brault A, Kassissia I 1988 Liver extraction of vitamin D 3 is independent of its venous or arterial route of delivery. Studies in isolated-perfused rat liver preparations. J Pharmacol Exp Ther 245:975–981. 23. Horsting M, DeLuca HF 1969 In vitro production of 25-hydroxycholecalciferol. Biochem Biophys Res Commun 36:251–256. 24. Ponchon G, DeLuca HF 1969 The role of the liver in the metabolism of vitamin D. J Clin Invest 48:1273–1279. 25. DeLuca HF 1971 Metabolism and mechanism of action of 25-hydroxycholecalciferol. In: Nicholes G Jr., Wasserman RH (eds) Cellular Mechanisms for Calcium and Homeostasis. Academic Press, New York, pp. 426–430. 26. Bhattacharyya MH, DeLuca HF 1974 Subcellular location of rat liver calciferol-25-hydroxylase. Arch Biochem Biophys 160:58–62. 27. Madhok TC, Schnoes HK, DeLuca HF 1978 Incorporation of oxygen-18 into the 25-position of cholecalciferol by hepatic cholecalciferol 25-hydroxylase. Biochem J 175: 479–482. 28. Fukushima M, Nishii Y, Suzuki M, Suda T 1978 Comparative studies on the 25-hydroxylations of cholecalciferol and 1α-hydroxycholecalciferol in perfused rat liver. Biochem J 170:495–502. 29. Tucker GI, Gagnon RE, Haussler MR 1973 Vitamin D3-25hydroxylase: Tissue occurrence and apparent lack of regulation. Arch Biochem Biophys 155:47–57. 30. Bhattacharyya MH, DeLuca HF 1974 The regulation of calciferol-25-hydroxylase in the chick. Biochem Biophys Res Commun 59:734–741. 31. Axén E 1995 Purification from pig kidney of a microsomal cytochrome P450 catalyzing 1α-hydroxylation of 25-hydroxyvitamin D3. FEBS Lett 375:277–279. 32. Hosseinpour F, Norlin M, Wikvall K 2002 Kidney microsomal 25- and 1α-hydroxylase in vitamin D metabolism: catalytic properties, molecular cloning, cellular localization and expression during development. Biochim Biophys Acta 1580:133–144. 33. Bhattacharyya MH, DeLuca HF 1973 The regulation of rat liver calciferol-25-hydroxylase. J Biol Chem 248:2969–2973.
CHAPTER 4 The Vitamin D 25-Hydroxylase
34. Bhattacharyya MH, DeLuca HF 1973 Comparative studies on the 25-hydroxylation of vitamin D3 and dihydrotachysterol. J Biol Chem 248:2974–2977. 35. Madhok TC, DeLuca HF 1979 Characteristics of the rat liver microsomal enzyme system converting cholecalciferol into 25-hydroxycholecalciferol. Biochem J 184:491–499. 36. Björkhem I, Hansson R, Holmberg I, Wikvall K 1979 25-Hydroxylation of vitamin D3 by a reconstituted system from rat liver microsomes. Biochem Biophys Res Commun 90:615–622. 37. Yoon PS, DeLuca HF 1980 Resolution and reconstitution of soluble components of rat liver microsomal vitamin D3-25hydroxylase. Arch Biochem Biophys 203:529–541. 38. Andersson S, Holmberg I, Wikvall K 1983 25-hydroxylation of C27-steroids and vitamin D3 by a constitutive cytochrome P-450 from rat liver microsomes. J Biol Chem 258:6777–6781. 39. Hayashi S, Noshiro M, Okuda K 1984 Purification of cytochrome P-450 catalyzing 25-hydroxylation of vitamin D3 from rat liver microsomes. Biochem Biophys Res Commun 121:994–1000. 40. Holmberg I 1984 Inhibition of reconstituted vitamin D3 25hydroxylase by a protein fraction from rat liver microsomes. Biochem Biophys Res Commun 123:1209–1214. 41. Nelson DR, Kamataki T, Waxman DJ, Guengerich FP, Estabrook RW, Feyereisen R, Gonzalez FJ, Coon MJ, Gunsalus IC, Gotoh O, Okuda K, Nebert DW 1993 The P450 superfamily: Update on new sequences, gene mapping, accession numbers, early trivial names of enzymes, and nomenclature. DNA Cell Biol 12:1–51. 42. Hiwatashi A, Ichikawa Y 1980 Purification and organspecific properties of cholecalciferol 25-hydroxylase system: Cytochrome P-450D25-linked mixed function oxidase system. Biochem Biophys Commun 97:1443–1449. 43. Hayashi S, Noshiro M, Okuda K 1986 Isolation of a cytochrome P-450 that catalyzes the 25-hydroxylation of vitamin D3 from rat liver microsomes. J Biochem 99:1753–1763. 44. Okuda KI 1994 Liver mitochondrial P450 involved in cholesterol catabolism and vitamin D activation. J Lipid Res 35:361–372. 45. Andersson S, Jornvall H 1986 Sex differences in cytochrome P-450-dependent 25-hydroxylation of C27-steroids and vitamin D3 in rat liver microsomes. J Biol Chem 261: 16932–16936. 46. Dählback H, Wikvall K 1987 25-hydroxylation of vitamin D3 in rat liver: Roles of mitochondrial and microsomal cytochrome P-450. Biochem Biophys Res Commun 142: 999–1005. 47. Saarem K, Pedersen JI 1987 Sex differences in the hydroxylation of cholecalciferol and of 5β-cholestane-3α,7α,12αtriol in rat liver. Biochem J 247:73–78. 48. Hayashi SI, Usui E, Okuda K 1988 Sex-related difference in vitamin D3 25-hydroxylase of rat liver microsomes. J Biochem 103:863–866. 49. Thierry-Palmer M, Free AL, Nagappan PR, Butler AR 1995 Cholecalciferol 25-hydroxylation is similar in liver microsomes from male and female rats when cholecalciferol concentration is low. J Nutr 125:104–111. 50. Segura-Aguilar J 1996 Peroxidase activity of liver microsomal vitamin D 25-hydroxylase and cytochrome P450 1A2 catalyzes 25-hydroxylation of vitamin D3 and oxidation of dopamine to aminochrome. Biochem Mol Med 58:122–129. 51. Axén E, Bergman T, Wikvall K 1994 Microsomal 25-hydroxylation of vitamin D2 and vitamin D3 in pig liver. J Steroid Biochem Mol Biol 51:97–106.
63 52. Hosseinpour F, Wikvall K 2000 Porcine microsomal vitamin D3 25-hydroxylase (CYP2D25): Catalytic properties, tissue distribution and comparison with human CYP2D6. J Biol Chem 275:34650–34655. 53. Axén E, Harmeyer J, Wikvall K 1998 Renal and hepatic 1α-hydroxylation of 25-hydroxyvitamin D3 in piglets suffering from pseudo vitamin D-deficiency rickets, type I. Biochim Biophys Acta 1407:234–242. 54. Ciobanu DC, Gill RK, Rothschild MF, Bell NH 2000 Rapid communication: Porcine vitamin D-25-hydroxylase maps to chromosome 5′. J Anim Sci 78:3193–3194. 55. Hosseinpour F, Hidestrand M, Ingelman-Sundberg M, Wikvall K 2001 The importance of residues in substrate recognition site 3 for the catalytic function of CYP2D25 (vitamin D 25-hydroxylase). Biochem Biophys Res Commun 288:1059–1063. 56. Hosseinpour F, Ibranovic I, Tang W, Wikvall K 2003 25-Hydroxylation of vitamin D3 in primary cultures of pig hepatocytes: evidence for a role of both CYP2D25 and CYP27A1. Biochem Biophys Res Commun 303:877–883. 57. Araya Z, Hosseinpour F, Bodin K, Wikvall K 2003 Metabolism of 25-hydroxyvitamin D3 by microsomal and mitochondrial vitamin D3 25-hydroxylases (CYP2D25 and CYP27A1): a novel reaction by CYP27A1. Biochim Biophys Acta 1632:40–47. 58. Nelson DR 2003 Comparison of P450s from human and fugu: 420 million years of vertebrate P450 evolution. Arch Biochem Biophys 409:18–24. 59. Bjorkhem I 1985 Mechanism of bile acid biosynthesis in mammalian liver. In: Danielsson H, Sjovall J. (eds) Sterols and Bile Acids. Elsevier Science, New York, pp. 231–278. 60. Popjak G, Edmond J, Anet FAL, Easton NR, Jr. 1977 Carbon-13 NMR studies on cholesterol biosynthesized from [13C]mevalonates. J Am Chem Soc 99:931–935. 61. Danielsson H 1960 On the oxidation of 3α,7α,12α-trihydroxycoprostane by mouse and rat liver homogenates. Acta Chem Scand 14:348–352. 62. Shefer S, Cheng FW, Batta AK, Dayal B, Tint GS, Salen G, Mosbach EH 1978 Stereospecific side chain hydroxylations in the biosynthesis of chenodeoxycholic acid. J Biol Chem 253:6386–6392. 63. Atsuta Y, Okuda K 1981 On the stereospecificity of cholestanetriol 26-monooxygenase. J Biol Chem 256: 9144–9146. 64. Björkhem I 1992 Mechanism of degradation of the steroid side chain in the formation of bile acids. J Lipid Res 33:455–471. 65. Björkhem I, Gustafsson J 1973 Omega-hydroxylation of steroid side-chain in biosynthesis of bile acids. Eur J Biochem 36:201–212. 66. Cronholm T, Johansson G 1970 Oxidation of 5β-cholestane3α,7α,12α-triol by rat liver microsomes. Eur J Biochem 373–381. 67. Taniguchi S, Hoshita N, Okuda K 1973 Enzymatic characteristics of CO-sensitive 26-hydroxylase system for 5β-cholestane-3α,7α,12α-triol in rat-liver mitochondria and its intramitochondrial localization. Eur J Biochem 40:607–617. 68. Bjorkhem I, Gustafsson J 1974 Mitochondrial omegahydroxylation of the cholesterol side chain. J Biol Chem 249:2528–2535. 69. Okuda KI, Weber P, Ullrich V 1977 Photochemical action spectrum of the CO-inhibited 5β-cholestane-3α,7α,12α-triol 26-hydroxylase system. Biochem Biophys Res Commun 74:1071–1076.
64 70. Sato R, Atsuta Y, Imai Y, Taniguchi S, Okuda KI 1977 Hepatic mitochondrial cytochrome P-450: isolation and functional characterization. Proc Natl Acad Sci USA 74:5477–5481. 71. Pedersen JI, Oftebro H, Vanngard T 1977 Isolation from bovine liver mitochondria of a soluble ferredoxin active in a reconstituted steroid hydroxylation reaction. Biochem Biophys Res Commun 76:666–673. 72. Okamura T, John ME, Zuber MX, Simpson ER, Waterman MR 1985 Molecular cloning and amino acid sequence of the precursor form of bovine adrenodoxin: Evidence for a previously unidentified COOH-terminal peptide. Proc Natl Acad Sci USA 82:5705–5709. 73. Hanukoglu I, Gutfinger T 1989 cDNA sequence of adrenodoxin reductase. Identification of NADP-binding sites in oxidoreductases. Eur J Biochem 180:479–484. 74. Ugele B, Locher M, Burger HJ, Gebhardt R 1986 Is there a heterogeneity of liver parenchyma in taurocholate uptake? In: Paumgartner G, Stiehl A, Gerok W (eds) Bile Acids and the Liver. MTP Press Ltd., Boston, pp. 153–160. 75. Pedersen JI, Godager HK 1978 Purification of NADPHferredoxin reductase from rat liver mitochondria. Biochim Biophys Acta 525:28–36. 76. Okuda K, Masumoto O, Ohyama Y 1988 Purification and characterization of 5β-cholestane-3α,7α,12α-triol 27-hydroxylase from female rat liver mitochondria. J Biol Chem 263:18138–18142. 77. Lyons MA, Maeda N, Brown AJ 2002 Paradoxical enhancement of hepatic metabolism of 7-ketocholesterol in sterol 27-hydroxylase-deficient mice. Biochim Biophys Acta 1581:119–126. 78. Wikvall K 1984 Hydroxylations in biosynthesis of bile acids. Isolation of a cytochrome P-450 from rabbit liver mitochondria catalyzing 26-hydroxylation of C27-steroids. J Biol Chem 259:3800–3804. 79. Björkhem I, Gustafsson J, Johansson G, Persson B 1975 Biosynthesis of bile acids in man. Hydroxylation of the C27-steroid side chain. J Clin Invest 55:478–486. 80. Masumoto O, Ohyama Y, Okuda K 1988 Purification and characterization of vitamin D 25-hydroxylase from rat liver mitochondria. J Biol Chem 263:14256–14260. 81. Ohyama Y, Masumoto O, Usui E, Okuda K 1991 Multi-functional property of rat liver mitochondrial cytochrome P-450. J Biochem 109:389–393. 82. Dixon M, Webb EC 1979 The investigation of enzyme specificity. In: Dixon M, Webb EC (eds), Enzymes. Longman, London, pp. 232–234. 83. Oftebro H, Saarem K, Bjorkhem I, Pedersen JI 1981 Side chain hydroxylation of C27-steroids and vitamin D3 by a cytochrome P-450 enzyme system isolated from human liver mitochondria. J Lipid Res 22:1254–1264. 84. Saarem K, Bergseth S, Oftebro H, Pedersen JI 1984 Subcellular localization of vitamin D3 25-hydroxylase in human liver. J Biol Chem 259:10936–10960. 85. Saarem K, Pedersen JI 1985 25-hydroxylation of 1α-hydroxyvitamin D-3 in rat and human liver. Biochim Biophys Acta 840:117–126. 86. Su P, Rennert H, Shayiq RM, Yamamoto R, Zheng Y-M, Addya S, Strauss JFI, Avadhani NG 1990 A cDNA encoding a rat mitochondrial cytochrome P450 catalyzing both the 26-hydroxylation of vitamin D3: Gonadotropic regulation of the cognate mRNA in ovaries. DNA Cell Biol 9:657–665. 87. Guo YD, Strugnell S, Back DW, Jones G 1993 Transfected human liver cytochrome P-450 hydroxylates vitamin D analogs at different side-chain positions. Proc Natl Acad Sci USA 90:8668–8672.
MARIELLE GASCON-BARRÉ
88. Nebert DW, McKinnon RA 1994 Cytochrome P450: Evolution and functional diversity. In: Boyer JL, Ockner RK (eds) Progress in Liver Diseases. WB Saunders, Philadelphia, pp. 63–97. 89. Sakaki T, Akiyoshi-Shibata M, Yabusaki Y, Ohkawa H 1992 Organella-targeted expression of rat liver cytochrome P450c27 in yeast. Genetically engineered alteration of mitochondrial P450 into a microsomal form creates a novel functional electron transport chain. J Biol Chem 267: 16497–16502. 90. Shayiq RM, Avadhani NG 1992 Sequence complementarity between the 5′-terminal regions of mRNAs for rat mitochondrial cytochrome P-450c27/25 and a growth hormone– inducible serine protease inhibitor. A possible gene overlap. J Biol Chem 267:2421–2428. 91. Mullick J, Addya S, Sucharov C, Avadhani NG 1995 Localization of a transcription promoter within the second exon of the cytochrome P-450c27/25 gene for the expression of the major species of two-kilobase mRNA. Biochemistry 34:13729–13742. 92. Wiese RJ, Uhland-Smith A, Ross TK, Prahl JM, DeLuca HF 1992 Up-regulation of the vitamin D receptor in response to 1,25-dihydroxyvitamin D3 results from ligand-induced stabilization. J Biol Chem 267:20082–20086. 93. Vlahcevic ZR, Jairath SK, Heuman DM, Stravitz RT, Hylemon PB, Avadhani NG, Pandak WM 1996 Transcriptional regulation of hepatic sterol 27-hydroxylase by bile acids. Am J Physiol 270:G646–G652. 94. Twisk J, Hoekman FM, Lehmann EM, Meijer P, Mager WH, Princen HMG 1995 Insulin suppresses bile acid synthesis in cultured rat hepatocytes by down-regulation of cholesterol 7α-hydroxylase and sterol 27-hydroxylase gene transcription. Hepatology 21:501–510. 95. Usui E, Noshiro M, Ohyama Y, Okuda K 1990 Unique property of liver mitochondrial P450 to catalyze the two physiologically important reactions involved in both cholesterol catabolism and vitamin D activation. FEBS Lett 274:175–177. 96. Akiyoshi-Shibata M, Usui E, Sakaki T, Yabusaki Y, Noshiro M, Okuda K, Ohkawa H 1991 Expression of rat liver vitamin D3 25-hydroxylase cDNA in Saccharomyces cerevisiae. FEBS Lett 280:367–370. 97. Holmberg-Betsholtz I, Lund E, Bjorkhem I, Wikvall K 1993 Sterol 27-hydroxylase in bile acid biosynthesis. Mechanism of oxidation of 5β-cholestane-3α,7α,12α,27-tetrol into 3α,7α,12α-trihydroxy-5β-cholestanoic acid. J Biol Chem 268:11079–11085. 98. Kusano K, Kagawa N, Sakaguchi M, Omura T, Waterman MR 2001 Importance of a prolife-rich sequence in the amino-terminal region for correct folding of mitochondrial and soluble microbial p450s. J Biochem 129:271–277. 99. Axén E, Postlind H, Sjoberg H, Wikvall K 1994 Liver mitochondrial cytochrome P450 CYP27 and recombinant-expressed human CYP27 catalyze 1α-hydroxylation of 25-hydroxyvitamin D3. Proc Natl Acad Sci USA 91: 10014–10018. 100. Arlazoroff A, Roitberg B, Werber E, Shidlo R, Berginer VM 1991 Epileptic seizure as a presenting symptom of cerebrotendinous xanthomatosis. Epilepsia 32:657–661. 101. Sawada N, Sakaki T, Ohta M, Inouye K 2000 Metabolism of vitamin D3 by human CYP27A1. Biochem Biophys Res Commun 273:977–984. 102. Horst RL, Koszewski NJ, Reinhardt TA 1990 1α-Hydroxylation of 24-hydroxyvitamin D2 represents a minor physiological pathway for the activation of vitamin D2 in mammals. Biochemistry 29:578–582. 103. Mawer EB, Jones G, Davies M, Still PE, Byford V, Schroeder NJ, Makin HLJ, Bishop CW, Knutson JC 1998
65
CHAPTER 4 The Vitamin D 25-Hydroxylase
104.
105.
106. 107.
108. 109. 110. 111. 112.
113.
114. 115.
116. 117.
118.
119.
Unique 24-hydroxylated metabolites represent a significant pathway of metabolism of vitamin D2 in humans: 24-hydroxyvitamin D2 and 1,24-dihydroxyvitamin D2 detectable in human serum. J Clin Endocrinol Metab 83:2156–2166. Dilworth FJ, Scott I, Green A, Strugnell S, Guo YD, Roberts EA, Kremer R, Calverley MJ, Makin HLJ, Jones G 1995 Different mechanisms of hydroxylation site selection by liver and kidney cytochrome P450 species (CYP27 and CYP24 involved in vitamin D metabolism. J Biol Chem 270: 16766–16774. Byford V, Strugnell S, Coldwell R, Schroeder N, Makin HLJ, Knutson JC, Bishop CW, Jones G 2002 Use of vitamin D4 analogs to investigate differences in hepatic and target cell metabolism of vitamins D2 and D3. Biochim Biophys Acta 1583:151–166. Thierry-Palmer M, Gray TK, Napoli JL 1988 Ring hydroxylation of 25-hydroxycholecalciferol by rat renal microsomes. J Steroid Biochem 29:623–628. Rao DS, Dayal R, Siu-Caldera ML, Horst RL, Uskokovic MR, Tserng KY, Reddy GS 1999 Isolation and identification of 4,25-dihydroxyvitamin D2: A novel A-ring hydroxylated metabolite of vitamin D2. J Steroid Biochem Mol Biol 71:63–70. Jones G, Schnoes HK, Levan L, DeLuca HF 1980 Isolation and identification of 24-hydroxyvitamin D2 and 24,25-dihydroxyvitamin D2. Arch Biochem Biophys 202:450–457. Wichmann JK, Schnoes HK, DeLuca HF 1981 Isolation and identification of 24(R)-hydroxyvitamin D3 from chicks given large doses of vitamin D3. Biochemistry 20:2350–2353. Bjorkhem I, Holmberg I, Oftebro H, Pedersen JI 1980 Properties of a reconstituted vitamin D3 25-hydroxylase from rat liver mitochondria. J Biol Chem 255:5244–5249. Dählback H, Wikvall K 1988 25-Hydroxylation of vitamin D3 by a cytochrome P-450 from rabbit liver mitochondria. Biochem J 252:207–213. Axén E, Postlind H, Wikvall K 1995 Effects on CYP27 mRNA expression in rat kidney and liver by 1α,25dihydroxyvitamin D3, 1α-hydroxylase activity. Biochem Biophys Res Commun 215:136–141. Hollis BW 1990 25-hydroxyvitamin D3–1α-hydroxylase in porcine hepatic tissue: Subcellular localization to both mitochondria and microsomes. Proc Natl Acad Sci USA 87:6009–6013. Takeuchi A, Okano T, Kobayashi T 1991 The existence of 25-hydroxyvitamin D3-1α-hydroxylase in the liver of carp and bastard halibut. Life Sci 48:275–282. Takeuchi A, Okano T, Sekimoto H, Kobayashi T 1994 The enzymatic formation of 1α,25-dihydroxyvitamin D3 from 25-hydroxyvitamin D3 in the liver of fetal rats. Comp Biochem Physiol [C] 109C:1–7. Negrea LA, Slotapolsky E, Dusso AS 1995 1,25-dihydroxyvitamin D synthesis in rat liver microsomes. Horm Metabol Res 27:461–464. Stravitz RT, Vlahcevic ZR, Russell TL, Heizer ML, Avadhani NG, Hylemon PB 1996 Regulation of sterol 27-hydroxylase and an alternative pathway of bile acid biosynthesis in primary cultures of rat hepatocytes. J Steroid Biochem Mol Biol 57:337–347. Twisk J, Hoekman MFM, Mager WH, Moorman AFM, De Boer PAJ, Scheja L, Princen HMG, Gebhardt R 1995 Heterogenous expression of cholesterol 7α-hydroxylase and sterol 27-hydroxylase genes in the rat liver lobulus. J Clin Invest 95:1235–1243. Theodoropoulos C, Demers C, Mirshahi A, Gascon-Barré M 2001 1,25-dihydroxyvitamin D3 downregulates the rat
120.
121.
122.
123. 124.
125. 126. 127.
128.
129.
130. 131. 132. 133.
134.
135.
intestinal vitamin D3-25-hydroxylase CYP27A. Am J Physiol 281:E315–E325. Theodoropoulos C, Demers C, Delvin E, Ménard D, GasconBarré M 2003 Calcitriol regulates the expression of the gene encoding all three vitamin D3 hydroxylases and the drug metabolizing enzyme CYP3A4 in the human fetal intestine. Clin Endocrinol 58:489–499. Memon RA, Moser AH, Shigenaga JK, Grunfeld C, Feingold KR 2001 In vivo and in vitro regulation of sterol 27-hydroxylase in the liver during the acute phase response. Potential role of hepatocyte nuclear factor-1. J Biol Chem 276:30118–30126. Segev H, Honigman A, Rosen H, Leitersdorf E 2001 Transcriptional regulation of the human sterol 27-hydroxylase gene (CYP27) and promoter mapping. Atherosclerosis 156:339–347. Cozzolino M, Vidal M, Arcidiacono MV, Tebas P, Yarasheski KE, Dusso AS 2003 HIV-protease inhibitors impair vitamin D bioactivation to 1,25-dihydroxyvitamin D. AIDS 17:513–520. Post SM, Duez H, Gervois PP, Staels B, Kuipers F, Princen HMG 2001 Fibrates suppress bile acid synthesis via peroxisome proliferator—Activated receptor-α-mediated downregulation of cholesterol 7α-hydroxylase and sterol 27-hydroxylase expression. Arterioscler Thromb Vasc Biol 2001:1840–1845. Araya Z, Tang W, Wikvall K 2003 Hormonal regulation of the human sterol 27-hydroxylase gene CYP27A1. Biochem J 372:529–534. Chen W, Chiang JYL 2003 Regulation of human sterol 27-hydroxylase gene (CYP27A1) by bile acids and hepatocyte nuclear factor 4α (HNF4α). Gene 313:71–82. Araya Z, Sjoberg H, Wikvall K 1995 Different effects on the expression of CYP7 and CYP27 in rabbit liver by cholic acid and cholestyramine. Biochem Biophys Res Commun 216: 868–873. Nakajima S, Hsieh JC, MacDonald PN, Galligan MA, Haussler CA, Whitfield GK, Haussler MR 1994 The C-terminal region of the vitamin D receptor is essential to form a complex with a receptor auxiliary factor required for high affinity binding to the vitamin D-responsive element. Mol Endocrinol 8:159–172. Rao YP, Vlahcevic ZR, Stravitz RT, Mallonee DH, Mullick J, Avadhani NG, Hylemon PB 1999 Down-regulation of the rat hepatic sterol 27-hydroxylase gene by bile acids in transfected primary hepatocytes: possible role of hepatic nuclear factor 1α. J Steroid Biochem Mol Biol 70:1–14. Cereghini S 1996 Liver-enriched transcription factors and hepatocyte differentiation. FASEB J 10:267–282. Bolt MJG, Sitrin MD, Favus MJ, Rosenberg IH 1981 Hepatic vitamin D 25-hydroxylase: inhibition by bile duct ligation or bile salts. Hepatology 1:436–440. Plourde V, Gascon-Barré M, Willems B, Huet PM 1988 Severe cholestasis leads to vitamin D depletion without perturbing its C-25 hydroxylation in the dog. Hepatology 8:1577–1585. Princen HMG, Meijer P, Wolthers BG, Vonk RJ, Kuipers F 1991 Cyclosporine A blocks bile acid synthesis in cultured hepatocytes by specific inhibition of chenodeoxycholic acid synthesis. Biochem J 275:501–505. Dahlback-Sjoberg H, Bjorkhem I, Princen HMG 1993 Selective inhibition of mitochondrial 27-hydroxylation of bile acid intermediates and 25-hydroxylation of vitamin D3 by cyclosporin A. Biochem J 2933:203–206. Baran DT, Milne ML 1983 1,25-Dihydroxyvitamin D-induced inhibition of 3H-25 hydroxyvitamin D production by the rachitic rat liver in vitro. Calcif Tissue Int 35:461–464.
66 136. Bell NH, Shaw S, Turner RT 1984 Evidence that 1,25-dihydroxyvitamin D3 inhibits the hepatic production of 25-hydroxyvitamin D in man. J Clin Invest 74: 1540–1544. 137. Clements MR, Davies M, Hayes ME, Hickey CD, Lumb GA, Mawer EB, Adams PH 1992 The role of 1,25-dihydroxyvitamin D in the mechanism of acquired vitamin D deficiency. Clin Endocrinol 37:17–27. 138. Haddad P, Gascon-Barré M, Brault G, Plourde V 1986 Influence of calcium or 1,25-dihydroxyvitamin D3 supplementation on the hepatic microsomal and in vivo metabolism of vitamin D3 in vitamin D-depleted rats. J Clin Invest 78:1529–1537. 139. Halloran BP, Bikle DD, Levens MJ, Castro ME, Globus RK, Holton E 1986 Chronic 1,25-dihydroxyvitamin D3 administration in the rat reduces the serum concentration of 25-hydroxyvitamin D by increasing metabolic clearance rate. J Clin Invest 78:622–628. 140. Gascon-Barré M, Demers C, Mirshahi A, Néron S, Zalzal S, Nanci A 2003 The normal liver harbors the vitamin D nuclear receptor in nonparenchymal and biliary epithelial cells. Hepatology 37:1034–1042. 141. Gascon-Barré M, Demers C, Ghrab O, Theodoropoulos C, Lapointe R, Jones G, Valiquette L, Ménard D 2001 Expression of CYP27A, a gene encoding a vitamin D-25 hydroxylase in human liver and kidney. Clin Endocrinol 54:107–115. 142. Cali JJ, Hsieh CL, Francke U, Russell DW 1991 Mutations in the bile acid biosynthetic enzyme sterol 27-hydroxylase underlie cerebrotendinous xanthomatosis. J Biol Chem 266:7779–7783. 143. Leitersdorf E, Reshef A, Meiner V, Levitzki R, Schwartz SP, Dann EJ, Berkman N, Cali JJ, Klapholz L, Berginer VM 1993 Frameshift and splice-junction mutations in the sterol 27-hydroxylase gene cause cerebrotendinous xanthomatosis in Jews of Moroccan origin. J Clin Invest 91:2488–2496. 144. Berginer VM, Shany S, Alkalay D, Berginer J, Dekel S, Salen G, Tint GS, Gazit D 1993 Osteoporosis and increased bone fractures in cerebrotendinous xanthomatosis. Metab Clin Exp 42:69–74. 145. Leitersdorf E, Safadi R, Meiner V, Reshef A, Bjorkhem I, Friedlander Y, Morkos S, Berginer VM 1994 Cerebrotendinous xanthomatosis in the Israeli Druze: molecular genetics and phenotypic characteristics. Am J Hum Genet 55:907–915. 146. Berginer VM, Shany S, Alkalay D, Berginer J, Dekel S, Salen G, Tint GS, Gazit D 1993 Osteoporosis and increased bone fractures in cerebrotendinous xanthomatosis. Metabolism 42:69–74. 147. Kuriyama M, Fujiyama J, Kubota R, Nakagawa M, Osame M 1993 Osteoporosis and increased bone fractures in cerebrotendinous xanthomatosis [Letter to editor]. Metab Clin Exp 42:1497. 148. Sawada N, Sakaki T, Kitanaka S, Kato S, Inouye K 2001 Structure-function analysis of CYP27B1 and CYP27A1. Studies on mutants from patients with vitamin D–dependent rickets type I (VDDR-I) and cerebrotendinous xanthomatosis (CTX). Eur J Biochem 268:6607–6615. 149. Guo YD, Strugnell S, Jones G 1991 Identification of a human liver mitochondrial cytochrome P-450 cDNA corresponding to the vitamin D3-25-hydroxylase. J Bone Miner Res 6:S120. 150. Strugnell S, Byford V, Makin HLJ, Moriarty RM, Gilardi R, LeVan LW, Knutson JC, Bishop CW, Jones G 1995 1α,24(S)dihydroxyvitamin D2: A biologically active product of 1α-hydroxyvitamin D2 made in the human hepatoma Hep3B. Biochem J 310:233–241.
MARIELLE GASCON-BARRÉ
151. Rosen H, Reshef A, Maeda N, Lippoldt A, Shpizen S, Triger L, Eggertsen G, Björkhem I, Leitersdorf E 1998 Markedly reduced bile acid synthesis but maintained levels of cholesterol and vitamin D metabolites in mice with disrupted sterol 27-hydroxylase gene. J Biol Chem 273:14805–14812. 152. Honda A, Salen G, Matsuzaki Y, Batta AK, Xu G, Leitersdorf E, Tint GS, Erickson SK, Tanaka N, Shefer S 2001 Side chain hydroxylations in bile acid biosynthesis catalyzed by CYP3A are markedly up-regulated in CYP27−/− mice but not in cerebrotendinous xanthomatosis. J Biol Chem 276: 34579–34585. 153. Björkhem I, Araya Z, Rudling M, Angelin B, Einarsson C, Wikvall K 2002 Differences in the regulation of the classical and the alternative pathway for bile acid synthesis in human liver. No coordinate regulation of CYP7A1 and CYP27A1. J Biol Chem 277:26804–26807. 154. Goodwin B, Gauthier KC, Umetani M, Watson MA, Lochansky MI, Collins JL, Leitersdorf E, Mangelsdorf DJ, Kliewer SA, Repa JJ 2003 Identification of bile acid precursors as endogenous ligands for the nuclear xenobiotic pregnane X receptor. Proc Natl Acad Sci USA 100: 223–228. 155. Martial J, Plourde V, Gascon-Barré M 1985 Sequestration of vitamin D3 by the fetal and neonatal rat liver. Biol Neonate 48:21–28. 156. Plourde V, Haddad P, Gascon-Barré M 1985 Microsomal C-25 hydroxylation of [3H]-vitamin D3 by the fetal and neonatal rat liver. Pediatr Res 19:1206–1209. 157. Thierry-Palmer M, Cullins S, Rashada S, Gray TK, Free A 1986 Development of vitamin D3 25-hydroxylase activity in rat liver microsomes. Arch Biochem Biophys 250:120–127. 158. Miller ML, Ghazarian JG 1980 Studies on vitamin D3 metabolism. Discrete liver cytosolic binding proteins for vitamin D3 and 25-hydroxyvitamin D3. Biochem Biophys Res Commun 96:1619–1625. 159. Tsankova V, Visentin M, Cantoni L, Carelli M, Tacconi MT 1996 Peripheral benzodiazepine receptor ligands in rat liver mitochondria: Effect on 27-hydroxylation of cholesterol. Eur J Pharmacol 299:197–203. 160. Delvin EE, Glorieux FH, Salle BL, David L, Varenne JP 1982 Control of vitamin D metabolism in preterm infants: feto-maternal relationships. Arch Dis Child 57:754–757. 161. Hillman LS, Hoff N, Salmons S, Martin L, McAlister WH, Haddad J 1985 Mineral homeostasis in very premature infants: Serial evaluation of serum 25-hydroxyvitamin D, serum minerals, and bone mineralization. J Pediatr 106:970–980. 162. Salle BL, Glorieux FH, Delvin EE, David LS, Meunier G 1983 Vitamin D metabolism in preterm infants. Serial serum calcitriol values during the first four days of life. Acta Paediatr Scand 72:203–206. 163. Addya A, Zheng YM, Shayiq RM, Fan J, Avadhani NG 1991 Characterization of a female-specific hepatic mitochondrial cytochrome P-450 whose steady-state level is modulated by testosterone. Biochemistry 30:8323–8330. 164. Bühler R, Lindros KO, Nordling A, Johansson I, IngelmanSundberg M 1992 Zonation of cytochrome P450 isozyme expression and induction in rat liver. Eur J Biochem 204: 407–412. 165. Hansson M, Ellis E, Hunt MC, Schmitz G, Babiker A 2003 Marked induction of sterol 27-hydroxylase activity and mRNA levels during differentiation of human cultured monocytes into macrophages. Biochim Biophys Acta 1593:283–289. 166. Jabara S, Christenson LK, Wang CY, McAllister JM, Javitt NB, Dunaif A, Strauss JF 2003 Stromal cells of the
CHAPTER 4 The Vitamin D 25-Hydroxylase
167. 168.
169.
170.
171.
human postmenopausal ovary display a distinctive biochemical and molecular phenotype. J Clin Endocrinol Metab 88:484–492. Javitt NB 2002 25R,26-hydroxycholesterol revisited: Synthesis, metabolism, and biologic roles. J Lipid Res 43: 665–670. Lehmann B, Tiebel O, Meurer M 1999 Expression of vitamin D3 25-hydroxylase (CYP27) mRNA after induction by vitamin D3 or UVB radiation in keratinocytes of human skin equivalents. A preliminary study. Arch Dermatol Res 291:507–510. Schuessler M, Astecker N, Herzig G, Vorisek G, Schuster I 2001 Skin is an autonomous organ in synthesis, two-step activation and degradation of vitamin D3: CYP27 in epidermis completes the set of essential vitamin D3-hydroxylases. Steroids 66:399–408. Correa P, Segersten U, Hellman P, Åkerström WG 2002 Increased 25-hydroxyvitamin D3 1α-hydroxylase and reduced 25-hydroxyvitamin D3 24-hydroxylase expression in parathyroid tumors—New prospects for treatment of hyperparathyroidism with vitamin D. J Clin Endocrinol Metab 87:5826–5829. Ichikawa F, Sato K, Nanjo M, Nishii Y, Shinki T, Takahashi N, Suda T 1995 Mouse primary osteoblasts express vitamin D3 25-hydroxylase mRNA and convert 1α-hydroxyvitamin D3 into 1α,25-dihydroxyvitamin D3. Bone 16:129–135.
67 172. Bjorkhem I, Andersson O, Diczfalusy U, Sevastik B, Xiu RJ, Duan C, Lund E 1994 Atherosclerosis and sterol 27-hydroxylase: Evidence for a role of this enzyme in elimination of cholesterol from human macrophages. Proc Natl Acad Sci USA 91:8592–8596. 173. Reiss AB, Martin KO, Javitt NB, Martin DW, Grossi EA, Galloway AC 1994 Sterol 27-hydroxylase: high levels of activity in vascular endothelium. J Lipid Res 35:1026–1030. 174. Delvin EE, Arabian A, Glorieux FH 1978 Kinetics of liver microsomal cholecalciferol 25-hydroxylase in vitamin Ddepleted and -repleted rats. Biochem J 172:417–422. 175. Benbrahim N, Dubé C, Vallières S, Gascon-Barré M 1988 The calcium ionophore A23187 is a potent stimulator of the vitamin D3-25 hydroxylase in hepatocytes isolated from normocalcemic vitamin D-depleted rats. Biochem J 255: 91–97. 176. Holmberg I, Berlin T, Ewerth S, Bjorkhem I 1986 25 hydroxylase activity in subcellular fractions from human liver. Evidence for different rates of mitochondrial hydroxylation of vitamin D2 and D3. Scand J Clin Lab Invest 46: 785–790. 177. Cheng JB, Levine MA, Bell NH, Mangelsdorf DJ, Russel DW 2004 Genetic evidence that the human CYP2R1 enzyme is a key vitamin D 25-hydroxylase. Proc Natl Acad Sciences (USA) 101:771–775.
CHAPTER 5
The 25-Hydroxyvitamin D 1α-Hydroxylase HELEN L. HENRY
Department of Biochemistry, University of California at Riverside, Riverside, California
I. Occurrence and Characteristics of 25OHD3-1α-Hydroxylase II. Characteristics of the Proteins Involved in the 1α-Hydroxylation of 25OHD3 III. Cytochrome P4501α: Cloning and Gene Structure
IV. Regulation of kidney 1α-Hydroxylase Activity and Gene Expression V. Summary References
I. OCCURRENCE AND CHARACTERISTICS OF 25OHD3 1α-HYDROXYLASE
exclusively in the proximal tubules. In situ hybridization studies of cultures of embryonic mouse kidneys confirm the presence of 25OHD3 1α-hydroxylase (CYP1α) in tubular epithelium, but not collecting ducts or glomeruli [8].
A. The Kidney as the Source of Circulating 1α,25(OH)2D3 It is now well accepted that vitamin D is a precursor of the sterol hormone 1α,25-dihydroxyvitamin D3 [1α,25(OH)2D3]. The general pathway of production of 1α,25(OH)2D3 is shown in Fig. 1. It has been appreciated for some time [1,2] that the kidney is the major site of production of circulating 1α,25(OH)2D3, although as described later and discussed more thoroughly elsewhere in this volume, many other tissues and cell types have been shown to produce 1α,25(OH)2D3 from 25-hydroxyvitamin D3 (25OHD3). Within the kidney, it was established early on by microdissection studies that in the fetal rabbit [3] and in the vitamin D–deficient rat [4] and chick [5], the proximal tubules are the region of the most robust activity of the 1α-hydroxylase. With the cloning of the cDNA for the cytochrome P450 component of the 1α-hydroxylase (see Section III) has come the ability to measure its mRNA and protein levels along the nephron. Since these determinations are more sensitive than the measurement of enzyme activity, localization studies can now be carried out under conditions of vitamin D sufficiency and normal mineral status. Thus in vitamin D–sufficient mice and humans, mRNA and/or protein has been identified by in situ hybridization or immunohistochemical staining in the more distal portions of the nephron along with relatively low expression in the proximal tubules [6,7]. These observations suggest that, while the 1α-hydroxylase occurs throughout the nephron, its regulation varies such that the effects of vitamin D status and abnormal phosphorus metabolism (see Section IV) occur primarily if not VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
B. Extrarenal Production of 1α,25(OH)2D3 Along with the demonstration that the kidney contained the enzymatic capability to produce 1α,25(OH)2D3 other work was suggesting that this metabolic step was largely confined to this organ. Physiological observations such as studies in nephrectomized animals [9] and clinical experience with patients suffering from chronic renal failure [10,11] strongly supported this concept. Thus, the kidney was the focus of investigation of the 1α-hydroxylase for the decade following its identification as the site of 1,25(OH)2D3 synthesis. In terms of the regulation of serum levels of 1α,25(OH)2D3 , the kidney remains the focus of study and most of the remainder of this chapter is devoted to the renal production of 1α,25(OH)2D3. Some extrarenal sites of 1α,25(OH)2D3 will be mentioned briefly here and discussed more thoroughly in Chapter 79. It should be noted that in general the extrarenal 1α-hydroxylation of 25OHD3 does not respond to the regulatory influences involved in calcium homeostasis that control the renal enzyme activity. In the early 1980s the first reports appeared that suggested that cells outside the kidney could produce 1α,25(OH)2D3 from 25(OH)D3. These reports described studies in calvarial cells [12,13] and were followed by those of 1α-hydroxylase and 24R-hydroxylase activity in chondrocytes [14,15]. These descriptions of 1α-hydroxylase activity in bone were followed by the demonstration that keratinocytes could also produce 1α,25(OH)2D3 [16,17]. Copyright © 2005, Elsevier, Inc. All rights reserved.
70
HELEN L. HENRY
FIGURE 1 Basic metabolic pathway for the major circulating vitamin D metabolites. The two renal hydroxylases that metabolize 25OHD3 are, in general, regulated in a reciprocal fashion to one another.
During this same time period, the hypercalcemia of sarcoidosis was traced to the production of 1α,25(OH)2D3 by pulmonary macrophages [18], and there followed many reports of the production of 1α,25(OH)2D3 by cells derived from the hematopoietic system [19–24]. These sites of the autocrine and/or paracrine production and action of 1α,25(OH)2D3 are discussed in more detail in Chapter 79. Reports that placental decidual cells [25,26] also contain 1α-hydroxylase activity were complicated by the fact the production of 1,25(OH)2D3 by trophoblastic tissues can produce 1α,25(OH)2D3 by a nonenzymatic process [27]. With the availability of the nucleotide and amino acid sequences for CYP1α, the presence of both the message and the protein in placental tissue [28,29] has now been confirmed. Other tissues in which evidence for CYP1α transcription and/or expression exists include the adrenal medulla, brain, pancreas, colon [29], parathyroid glands [30], prostate cells [31], and non-small-cell lung carcinoma [32].
C. Species Distribution The 1α-hydroxylase, not surprisingly, occurs widely among vertebrate species. None of the species examined in the original species comparison [33] have since been shown to lack the renal 1α-hydroxylase.
Since there are no reports of 1α-hydroxylase activity in invertebrates it can be tentatively concluded that the ability to produce 1α,25(OH)2D3 arose early in vertebrate evolution. There has been a special interest in the occurrence of vitamin D and its metabolites in fish, due in part to the historical importance of cod liver oil as a source of vitamin D and in part to the question of how fish produce vitamin D under conditions of limited ultraviolet light. Plasma levels of vitamin D metabolites have been measured in several species of fish [34–36]. Fish liver and kidney homogenates have been reported to convert 25OHD3 to 1α,25(OH)2D3 [33,36,37]. Both phytoplankton and zooplankton were reported to have abundant quantities of provitamin D and vitamin D and were suggested to be the dietary source of vitamin D for fish [38], although these authors have concluded that vitamin D and its metabolites are unimportant to mineral metabolism in the freshwater fish tilapia [39]. Thus, although there is no question that fish contain vitamin D and its metabolites and 1α-hydroxylase activity, the physiological role of these in fish remains unclear. Another interesting comparative aspect of vitamin D metabolism is posed by the naked mole rat. This nocturnal, cave-dwelling animal with no access to ultraviolet light nevertheless has the ability to synthesize both 1α,25(OH)2D3 and receptors for the hormone [40,41]. Thus, although these animals are usually vitamin D
71
CHAPTER 5 The 25-Hydroxyvitamin D 1α-Hydroxylase
ultimately derived from NADPH, generated intramitochondrially from tricarboxylic acid cycle intermediates. In vitro assays of 1α-hydroxylase activity in isolated mitochondria have used either malate [42] or other citric acid cycle intermediates [43] as a source of reducing equivalents. The current model for mitochondrial steroid hydroxylases involves the sequential transfer of electrons from NADPH to NADPH-ferredoxin reductase, a 50-kDa protein that, in the case of adrenal and gonadal steroid hydroxylases, is loosely associated with the inner mitochondrial membrane [44]. The electrons are then passed to ferredoxin, an 11-kDa mitochondrial matrix protein that shuttles between the NADPH-ferredoxin reductase and the terminal component of the hydroxylase machinery, cytochrome P450 (so named because of its distinct spectral characteristics when carbon monoxide is bound to it). As discussed below, each of these three proteins has been the subject of study in the context of the 1α-hydroxylation of 25OHD3.
deficient, they maintain the enzymatic machinery necessary to produce the active hormonal form and the receptor necessary to respond to it.
II. CHARACTERISTICS OF THE PROTEINS INVOLVED IN THE 1α-HYDROXYLATION OF 25OHD3 A. Mixed-Function Oxidases Enzymes that hydroxylate endogenous steroids in a stereospecific manner are called mixed-function oxidases because they reduce one atom of molecular oxygen to water and one atom to the hydroxyl group to be stereospecifically incorporated into the steroid (Fig. 2). Such reactions are important in the pathways of the production of androgens, estrogens, progestins, mineralocorticoids, and glucocorticoids. Indeed, the first step in the production of these steroid hormones, the cleavage of six carbons from the side chain of cholesterol, results from two successive hydroxylations at C-20 and C-22. Some mixed-function oxidases are microsomal (e.g., the 17α-hydroxylase/C-17-C-20 lyase and the 21-hydroxylase), and some (e.g., cholesterol side-chain cleavage and the 11β-hydroxylase) are located in the inner mitochondrial membrane. The 25OHD3 1α-hydroxylase belongs to the latter group of steroid hydroxylases. For mitochondrial mixed-function oxidases, the electrons for the reduction of molecular oxygen are
1. CYTOCHROME P450
The evidence that the 1α-hydroxylase is a member of the class of cytochrome P450-dependent steroid hydroxylases was obtained in the mid-1970s and consists of the following: (1) Solubilized preparations of chick kidney mitochondria were shown to contain cytochrome P450 and to carry out 1α-hydroxylation of 25OHD3 [45]; (2) more directly, it was shown that inhibition of 1α-hydroxylase activity in isolated chick kidney mitochondria by carbon monoxide was reversed specifically by light of 450 nm [42].
R–OH + H2O NADPH
FPox
NHI–Fe
2+
P450–Fe
3+
R Reductase
NADP
FPred
Ferredoxin
NHI–Fe 3+
Cytochrome P450
P450–Fe 2+
P450–Fe 2+−O2
O2
CO P450–Fe 2+−CO
FIGURE 2 Electron transport chain for mitochondrial steroid hydroxylases. The general class of mitochondrial mixed-function oxidases, of which the 25OHD3 1α-hydroxylase is a member, consists of three components in or associated with the inner mitochondrial membrane. Cytochrome P450 reduces molecular oxygen to water and to the hydroxyl group to be incorporated into the steroid; it confers the stereospecificity of the hydroxylation reaction. In its reduced form, cytochrome P450 binds carbon monoxide and in this form absorbs light of 450 nm. The flavoprotein (FP) is NADPH-ferredoxin reductase, and the nonheme iron (NHI) protein is ferredoxin. Ox, oxidized; red, reduced.
72
HELEN L. HENRY
There have been sporadic reports of the partial purification of CYP1α. An apparently electrophoretically pure bovine kidney cytochrome P450, with a molecular weight of 57,000, retained putative 1α-hydroxylase activity [46], but there was no confirmation of the identity of the product of 25OHD3 hydroxylation. In the early 1990s, there were two reports purporting to have purified the 1α-hydroxylase cytochrome P450 from mitochondria from either vitamin D–replete or vitamin D–deficient chick kidneys [47,48]. The purified proteins had molecular weights of 59,000 and 57,000, but amino acid compositions and N-terminal sequences, when compared with the four mammalian CYP1α sequences that are now known, make it unlikely that either of these proteins was indeed the 1α-hydroxylase cytochrome P450. More recently, a 54-kDa protein with 1α-hydroxylase activity and appropriate spectral characteristics was isolated from chick kidney [49] but no sequence information about this protein was given. In another study with pig kidney mitochondria, 1α-hydroxylase activity was partially purified [50] and separated from steroid side chain–hydroxylating enzymes [51], but in neither report were any physical characteristics of the protein given. In summary, although the primary sequence of CYP1α has been deduced from cloned cDNA from several mammalian species, purified preparations of the protein itself are still not routinely obtained. The primary sequence of the cytochrome P450 that catalyzes the 1α-hydroxylation of 25OHD3, as deduced from its cDNA sequence, reveals that it is structurally related to the mitochondrial sterol side chain hydroxylases and therefore it has been given the systematic name of CYP27B1 (see [52] for a listing of human cytochrome P450s). In this chapter, the protein and its nucleic acid sequences will be referred to by its familiar and functionally more descriptive name of CYP1α. 2.
FERREDOXIN
a. Characteristics of Mitochondrial Ferredoxin Vertebrate mitochondrial ferredoxins are nonheme iron– sulfur proteins of 114–128 amino acids, which on electrophoresis generally migrate at 11–12 kDa. cDNA analysis indicates a mitochondrial leader sequence of 58–62 amino acids, depending on the species [53]. The ferredoxin involved in the mitochondrial hydroxylation of endogenous steroids is a matrix protein that shuttles between NADPH-ferredoxin reductase and cytochrome P450 to deliver, one at a time, the two electrons required for the reduction of molecular oxygen. Ferredoxins from mammalian adrenal glands (adrenodoxins) have been studied most extensively [54,55], although the protein has been isolated from several other steroidogenic tissues [56].
As is the case for ferredoxins in other steroidogenic tissues, chick kidney ferredoxin is a mitochondrial matrix protein [57]. This submitochondrial localization is consistent with its function as an electron-carrying shuttle between the NADPH-ferredoxin reductase and CYP1α. Reconstitution assays in a number of laboratories have demonstrated the absolute requirement of 1α-hydroxylase activity for ferredoxin [58–59]. Chick kidney ferredoxin has been overexpressed in Escherichia coli and shown to be approximately twice as active as bovine adrenal ferredoxin in supporting 1α-hydroxylation in a reconstituted assay system [60], further illustrating the similarities that the 1α-hydroxylase shares with other mitochondrial mixed-function oxidases that hydroxylate endogenous steroids. cDNA cloning has shown that the pig kidney and adrenal ferredoxin share identical amino acid sequences [61]; similarly, in the chick kidney and testis ferredoxins share the same primary structure [62,63]. These observations suggest that a single gene encodes ferredoxin for the various tissues in which it is expressed. Southern blot analysis of bovine [64] and chick genomic DNA [62] supports the existence of a single ferredoxin gene in these species. In the human, two ferredoxin genes have been identified, but their coding regions for the mature protein are identical [65]. Taken together the available evidence indicates that a single ferredoxin protein serves as the electron shuttle for all mitochondrial cytochrome P450s that hydroxylate endogenous steroids in the species that have been examined thus far. b. Regulation of Mitochondrial Ferredoxin Levels Since the rate of hydroxylation of a steroid by cytochrome P450 will be governed to some extent by the supply of electrons to reduce molecular oxygen, a regulatory component at this step is possible, although such a process would clearly lack some specificity in cells that can carry out more than one hydroxylation. Mammalian ferredoxins and their mRNAs have been shown to be developmentally and hormonally regulated [66–69]. For example, hypophysectomy decreases and replacement treatment with adrenocorticotropin (ACTH) increases adrenal ferredoxin mRNA levels in the rat [67,70]. In a study of several cell types, ferredoxin mRNA was stimulated by appropriate pituitary tropic hormones, possibly through a cAMP-dependent mechanism, and ferredoxin mRNA levels decreased in the testis and adrenal during fetal development [66]. Chick kidney mitochondrial ferredoxin mRNA is up-regulated approximately 40% in the vitamin D– deficient state [62]. Although kidney mitochondrial ferredoxin is also presumably required for the 24Rhydroxylation of vitamin D, which is, in general, regulated reciprocally to the 1α-hydroxylase, 1α-hydroxylase activity in vitamin D deficiency is much higher than is
73
CHAPTER 5 The 25-Hydroxyvitamin D 1α-Hydroxylase
24R-hydroxylase activity in the vitamin D–replete chick. Therefore, increased mRNA levels of ferredoxin mRNA in vitamin D deficiency are not likely to be involved in the reciprocal regulation of these two enzymes, but rather are related to the increased demand of maximal 1α-hydroxylase activity brought about by vitamin D deficiency. Ferredoxin phosphorylation has been suggested as a posttranslational mechanism for the regulation of its steroid hydroxylation activity in the adrenal gland and kidney. Although adrenal ferredoxin can be phosphorylated in vivo by cAMP-dependent protein kinase, no convincing evidence of phosphorylated adrenal ferredoxin in vivo has appeared, nor has a role been established for such a phosphoferredoxin in the physiological regulation of adrenal or gonadal steroidogenesis. On the other hand, ferredoxin from primary cultures of chick kidney cells has been shown to exist as a phosphoprotein and to be rapidly dephosphorylated by treatment of the cells with the phorbol ester 12-O-tetradecanoylphorbol-13-acetate (TPA) [57]. This dephosphorylation is accompanied by decreased 1α-hydroxylase activity, supporting for a role of ferredoxin phosphorylation and dephosphorylation in the regulation of 1-hydroxylation of 25OHD3. However, the molecular mechanisms involved in such a process remain to be determined. 3. NADPH-FERREDOXIN REDUCTASE
The flavoprotein NADPH-ferredoxin reductase, which accepts electrons from intramitochondrially generated NADPH, has been purified from pig kidney mitochondria [71] and does not appear to differ significantly in size (~52 kDa), enzymatic characteristics, or immunologically from the reductases from other mammalian tissues that carry out hydroxylations in the production of endogenous steroids such as the corpus luteum and the adrenal gland [72,73]. cDNAs for bovine and human adrenal NADPH-ferredoxin reductase have been cloned and shown to code for a mitochondrial targeting sequence in addition to the mature protein [74].
B. Accessory Proteins In addition to the three components of the enzymatic machinery that hydroxylates 25OHD3 at the 1α position, certain other cellular proteins have been implicated in the process of moving the substrate from the circulation to the site of metabolism in the inner mitochondrial membrane. Although the steroid hormones and their precursors are lipophilic and might be thought to simply diffuse through the cellular membranes
to the site of their metabolism, specific proteins to direct their intracellular movements are quite likely. One group of candidates for this type of protein, currently termed the intracellular vitamin D binding proteins, or IDBPs, was originally found in vitamin D(and other steroid hormone)-resistant New World primates [75]. The constitutive expression of binding proteins for 25OHD3 in these animals was initially thought to be the causative factor in their vitamin D resistance and/or the response to it, leading to high circulating levels of the hormone. More recently, these proteins have been cloned and sequenced and found to be homologous to human heat shock proteins of the hsp-70 family [76,77], and they are now thought to play an active role in the delivery of the substrate 25OHD3 to the 1α-hydroxylase, rather than acting to sequester the substrate as originally thought. In addition, these proteins have been shown to bind DNA and may play a function in the regulation of 24R-hydroxylase gene expression (see Chapter 21). A somewhat more definitive role in the uptake of 25OHD3 from the circulation and delivery to the 1α-hydroxylase has been proposed for the endocytotic receptor megalin, a member of the low-density lipoprotein receptor family [78]. These authors initially observed abnormal excretion of 25OHD3 in megalin knockout mice and showed that complexes of the vitamin D binding protein, DBP, and 25OHD3 are filtered by the glomerulus and reabsorbed by megalin in the proximal tubule of the kidney. In megalin knockout mice, the mRNA levels for both the 1α-hydroxylase and the 24R-hydroxylase were altered in a way consistent with severe vitamin D deficiency (that is, elevated 1α-hydroxylase and decreased 24R-hydroxylase activity; see Section IV), confirming the inability of the kidney to produce 1α,25(OH)2D3 to regulate these two enzymes [79,80]. See Chapter 10 for a more thorough discussion of the role of megalin in vitamin D metabolism.
III. CYTOCHROME P4501α: CLONING AND GENE STRUCTURE Reports of the cloning of the cDNA encoding the cytochrome P450 catalyzing the 1α-hydroxylation of 25(OH)D3 appeared from a number of laboratories in 1997. The mouse cDNA was cloned from kidneys of mice lacking the VDR, which resulted in very high expression levels of 1α-hydroxylase mRNA [81]. At approximately the same time, the rat kidney [82,83], human keratinocyte [84], and human kidney [85] sequences for the 1α-hydroxylase cytochrome P450 were published. These were followed by the porcine kidney
74 coding sequence [86]. For the mouse, rat, and human sequences, the identification of the cloned cDNA as that encoding the 1α-hydroxylase cytochrome P450 was confirmed by transfection into cells capable of supporting steroid hydroxylation (cell lines derived from kidney or testis having endogenously expressed ferredoxin and ferredoxin reductase) and demonstration that the product formed from added 25(OH)D3 was 1α,25(OH)2D3. The porcine sequence was identified on the basis of its 88% homology with the other three mammalian sequences. The sequences of cDNAs encoding the mammalian CYP1αs predict proteins with 501 to 508 amino acids. The structures of the human [85,87,88] and mouse [89] CYP27B1 genes have been determined. The mammalian CYP27B1 gene is divided into nine exons, seven
HELEN L. HENRY
of roughly equal size (170–200 bp encoding 57–67 amino acids each) and two smaller ones (26 and 37 amino acids). The putative sterol binding region (Fig. 3) is encoded by exon VI and the heme-binding region, centering around the cysteine residue that is invariant in all mitochondrial P450s that hydroxylate endogenous steroids, is encoded by exon VIII. The intron size ranges from approximately 80 to 660 bp, so altogether, the CYP27B1 gene is small in size relative to some other members of the mitochondrial cytochrome P450s, although the overall intron/exon organization is highly conserved among the genes coding for these proteins. The gene for CYP1α has been mapped to human chromosome 12 [83,87] and linkage analysis shows that it lies
FIGURE 3 Deduced amino acid sequences of CYP1α from several mammalian species. The human sequence is given in its entirety and in the other sequences, identical amino acids are indicated by dashes. Box A is the putative sterol binding site and Box B is the heme-binding region, containing the cysteine that is invariant in all cytochrome P450s.
75
CHAPTER 5 The 25-Hydroxyvitamin D 1α-Hydroxylase
very close to the defect for human vitamin D–dependent rickets (VDDR type I) (see Chapter 71). In the mouse, the CYP1α gene is on chromosome 10, the region equivalent to the localization of the human gene [89].
IV. REGULATION OF KIDNEY 1α-HYDROXYLASE ACTIVITY AND GENE EXPRESSION In numerous whole-animal and cell-culture model systems it is now well established that the two most important regulators of the activity of the 1α-hydroxylase are 1α,25(OH)2D3 itself and parathyroid hormone. In addition, there has been evidence for the involvement of calcitonin, dietary mineral levels, and hormones of other endocrine systems in the regulation of renal 25OHD3 metabolism.
A. 1α,25(OH)2D3 It was recognized at the time of the localization of the production of 1,25(OH)2D3 in the kidney that, in both avian and mammalian species, vitamin D–deficient animals have higher 1α-hydroxylase activity than do vitamin D–replete animals [1,90–92]. The ability of 1,25(OH)2D3 to replicate this effect on 1α-hydroxylase has been demonstrated in chick [93,94] and mammalian [95] cells in vitro. Both in vivo and in cell culture, the activity of the 24-hydroxylase is regulated by 1α,25(OH)2D3 in a reciprocal fashion to the 1α-hydroxylase. When 1,25(OH)2D3 is removed from the cell culture medium [93] or when its administration in vivo is terminated [90], 1α-hydroxylase activity returns to the high levels characteristic of the vitamin D–deficient state, indicating the ready reversibility of the inhibitory effect of 1,25(OH)2D3 on kidney 1α-hydroxylase activity. In a wide variety of
experimental conditions the effect of 1α,25(OH)2D3 on 1α-hydroxylase activity observed in intact animals in vivo can be duplicated by experiments in primary cultures of kidney cells. From these types of experiments, however, the mechanism of the down-regulation of 1α-hydroxylase activity by 1α,25(OH)2D3 has not been elucidated. The availability of the cDNA and genomic sequences for CYP1α has opened up new avenues of exploration of this question. There is no question that steadystate CYP1α mRNA levels are down-regulated by 1α,25(OH)2D3 in the kidneys of intact animals [96,97] and that this down-regulation requires the presence of the VDR [81,96]. When steady-state CYP1α mRNA levels are measured in cultured cells, however, the results are mixed and may depend on the cell type used, as shown in Table I. In fact, of the cell types tested, only in the MCT cells did 1α,25(OH)2D3 downregulate the endogenous expression of the mRNA for the 1α-hydroxylase cytochrome P450. Of the nonresponsive cells, both the LLC-PK1 and the HKC-8 cells have been reported to express 1α-hydroxylase activity [98,99]. However, in none of the studies in Table I were data regarding the effect of 1α,25(OH)2D3 on 1α-hydroxylase enzymatic activity presented under the same conditions in which the cells were apparently unresponsive. Thus, it is not known whether the failure to see down-regulation of CYP1α mRNA by 1α,25(OH)2D3 was due to dissociation of its effects on the transcription and activity of the cytochrome P450 or whether these cells, for unknown reasons, failed to respond to 1α,25(OH)2D3 at both the RNA and protein levels. In addition to studies of the effects of 1α,25-dihydroxyvitamin D3 on the steady-state levels of CYP1α mRNA, several laboratories have prepared and studied reporter constructs using the 5′-flanking region of the gene. In one report such a construct was active in cells derived from proximal renal tubules but not in other
TABLE I Summary of Effects of 1α,25(OH)2 D3 on Steady-State CYP1α mRNA levels Whole Animals Species Rat Mouse Mouse-VDR-KO
Effect of 1α,25(OH)2D3 ↓ ↓ →
Cell Culture Reference
Cell Line
Effect of 1α,25(OH)2D3 treatment
Reference
[96,97] [96] [81,96]
MCT cells LLC-PK1 AOK-B50 HKC-8
↓ → → →
[96,151] [86] [152] [152]
Cell types referred to are MCT:SV-40-transformed mouse proximal tubule cells [151]; LLC-PK1; AOK-B50, a derivative of LLC-PK1 stably transfected with opossum PTH/PTHrP receptor [153]; and HKC-8, SV-40-transformed human kidney cell line [154].
76
HELEN L. HENRY
TABLE II
Summary of Effects of 1α,25(OH)2D3 on CYP1α Promoter Activity
Promoter (species-kb)
Cell type
Effect
Reference
m-1.7 m-1.6 h-1.1 h-0.5
AOK-B50 OK AOK-B50 MCT
→ ↓ → ↓
[155] [156] [88] [151]
Cell types referred to are the same as in Table 1, with the addition of the opossum kidney cell line, OK.
cell types [100]. A summary of the results of this type of study is shown in Table II. No definitive vitamin D regulatory element has been identified from sequence analysis of the promoter region of mammalian CYP1α genes, although functional analysis suggests that such a regulatory site may lie within 0.5 kb of the transcriptional start site [96]. As is the case with studies of the regulation by 1α,25(OH)2D3 of endogenous CYP1α mRNA levels, it is clear from Table II that studies of the regulation of promoter activity in various cell types have yielded mixed results and have not as yet revealed whether 1α,25(OH)2D3 down-regulates 1α-hydroxylase activity through a decrease in transcription of its mRNA, through an indirect transcriptional mechanism, or through posttranscriptional and/or posttranslational processes. It is possible that more than one mechanism is involved and that the model system employed will play a role in which is revealed.
B. Parathyroid Hormone In whole animal studies, parathyroidectomy diminishes and administration of PTH increases 1α-hydroxylase activity in kidney tissue, in both avian and mammalian species [101–105]. PTH has also been shown to stimulate 1α-hydroxylase activity in cultured avian [106–108] and mammalian [109,110] kidney cells. The stimulatory effect of PTH on 1α-hydroxylase activity is mediated at least in part by the cAMP signaling pathway, as its effects can be mimicked by forskolin [109,111]. Although specific proteins with altered phosphorylation states in response to PTH have been identified [112], it has not yet been determined whether they play a physiological role in the regulation of the 1α-hydroxylation of 25OHD3. In addition to the cAMP signaling pathway, protein kinase C is also involved in the regulation of the renal
metabolism of 25OHD3 by both the 1α-hydroxylase and the 24-hydroxylase. In chick kidney cells, the phorbol ester protein kinase C (PKC) activator TPA decreases 1α-hydroxylase activity transiently [57] and over a 4-hr incubation period [113]. The latter decrease in 1α-hydroxylase activity is associated with increased 24-hydroxylase activity. The abilities of TPA and 1,25(OH)2D3 to decrease 1α-hydroxylase activity in cultured chick kidney cells are additive, indicating that they operate through distinct mechanisms. In perfused rat kidney proximal tubules, on the other hand, PKC activation by TPA resulted in a transient increase in 1α-hydroxylase activity and a sustained increase when the calcium ionophore A23187 was added with TPA [114]. PTH has been reported to stimulate the PKC pathway in rat renal cells [115]. The relative physiological importance of the cAMP and PKC signaling pathways in the control of 1α-hydroxylase activity has not yet been fully resolved. Furthermore, the fact that TPA activation of PKC decreases 1α-hydroxylase activity in chick kidney cells, both in the short term and over longer times, while increasing its activity in rat preparations, suggests that the involvement of PKC in the regulation of 1α-hydroxylase may be more complex than initially appreciated. The availability of reagents with which to study the regulation of steady-state CYP1α mRNA levels and promoter activity has led to the demonstration that PTH exerts at least part of its effects on 1α-hydroxylase activity at the transcriptional level. In addition, the role of cyclic AMP in this process has been confirmed in these studies, which are summarized in Tables III and IV. In every cell type studied, PTH stimulated the endogenous levels of CYP1α mRNA and, when tested, this effect was enhanced by cyclic AMP. Only in the MCT cells, however, did 1α,25(OH)2D3 attenuate the effect of PTH, similar to the results seen when the ability of 1α,25(OH)2D3 alone to down-regulate CYP1α mRNA was tested (Table I). PTH also stimulated the TABLE III Summary of the Regulation of Steady-State CYP1α mRNA Levels by PTH, cAMP, and Forskolin Effect of Cell line LLC-PK1 AOK-B50 HKC-8 MCT
PTH
cAMP/FSK
1α,25(OH)2D3a
Reference
↑ ↑ ↑ ↑
↑ ↑ nd nd
→ → → ↓
[86] [152] [152] [96]
a The effect of 1α,25(OH) D refers to whether the steroid attenuates 2 3 (↓) or has no effect on (→) stimulation of CYP1α mRNA levels by PTH or cAMP. nd, Not determined.
77
CHAPTER 5 The 25-Hydroxyvitamin D 1α-Hydroxylase
activity of a variety of CYP1α promoter constructs and this effect was attenuated, although only modestly in two cases, by 1α,25(OH)2D3. The observation that the effect of 1α,25(OH)2D3 is not particularly robust may account for the lack of effect on endogenous levels of 1α,25(OH)2D3 mRNA. Taken together, the results very strongly indicate that PTH stimulation of 1α-hydroxylase activity in renal cells is mediated, at least in part, by a transcriptional effect on CYP1α gene expression. Furthermore, they support the concept that a balance between PTH and 1α,25(OH)2D3 determines this level of expression. That this transcription effect may not tell the whole story, however, is suggested by a report that in adult rats, PTH administration increases CYP1α mRNA levels in the kidney severalfold, but has no effect on 1α-hydroxylase activity. Clearly the regulation of the renal production of 1α,25(OH)2D3 is complex, occurs at more than one step, and is an area that deserves more detailed investigation.
C. Other Regulators of Renal 1α,25(OH)2D3 Synthesis
levels of this message and directly influences the transcriptional activity of the CYP1α gene promoter. Shinki et al. [97] showed increased steady-state CYP1α mRNA levels when normocalcemic rats were injected with calcitonin. Interestingly, this increase in CYP1α mRNA was preceded by a marked decrease in VDR mRNA, leaving open the possibility that part of the effect of calcitonin is to render the kidney 1α-hydroxylase less sensitive to feedback inhibition by 1α,25(OH)2D3. In a second report of the stimulation of steady-state levels of CYP1α mRNA by calcitonin in rat kidney [96] it was also observed that simultaneous administration of 1α,25(OH)2D3 abrogated this effect in intact rats but not in the VDR-KO mouse model that was also used. A direct stimulatory effect of calcitonin on CYP1α mRNA levels in the MCT cells, a cell line derived from mouse proximal tubules [96], and in LLC-PK1 cells [122] has been reported. Finally, the human CYP1α promoter fused to CAT reporter is up-regulated by calcitonin in MCT cells, further underscoring a role for this calcium-regulating hormone in the renal metabolism of 25OHD3. 2. DIETARY PHOSPHORUS
1. CALCITONIN
The role of calcitonin in the regulation of renal vitamin D metabolism has a long and somewhat controversial history. Although there is a substantial number of reports suggesting that calcitonin plays a role in the regulation of vitamin D metabolism of the intact animal [97,116–118], evidence that these effects are exerted directly on the renal cell that carries out the 1α-hydroxylation of 25OHD3 has been not been consistently obtained. Thus, when added to isolated renal cell preparations in vitro calcitonin has been reported to stimulate [119], inhibit [120], or have no effect [121] on 1α-hydroxylase activity. More recent studies measuring CYP1α mRNA levels, however, strongly suggest that calcitonin does stimulate the steady-state
In mammalian animal models, low dietary phosphorus leads to increased serum concentrations of 1α,25(OH)2D3 and increased 1α-hydroxylase activity measured in renal preparations [123–126]. In humans as well, dietary phosphorus is inversely correlated with serum levels of 1α,25(OH)2D3 [127–130]. It has been demonstrated that dietary phosphorus restriction in mice leads to an increase of renal proximal tubule production of 1α,25(OH)2D3 as well as increases in both the mRNA and protein levels of CYP1α [7], providing strong evidence that the effect of phosphorus is exerted, at least in part, on the level of transcription of CYP1α. In addition, these authors also provided evidence that reduced catabolism (due to reduced 24Rhydroxylase activity) of 1α,25(OH)2D3 may also play a role in the increased serum levels of 1α,25(OH)2D3 seen when dietary phosphorus is low. 3. OTHER STEROID HORMONES
TABLE IV Promoter (species-kb) m-1.7 m-1.6 h-1.1 h-0.3 h-0.5
Summary of Effects of PTH on CYP1α Promoter Activity Cell type
AOK-B50 OK AOK-B50 AOK-B50 MCT
PTH/FSK/ cAMP 1α,25(OH)2D3 Reference ↑ ↑ ↑ ↑ ↑
↓ (marginal) ↓ (modest) ↓ nd nd
[155] [156] [88] [157] [151]
Steroid hormones of other endocrine systems have been tested for their effects, or lack thereof, on 1,25(OH)2D3 production in vivo or in cell culture. These include estrogen and the synthetic glucocorticoid dexamethasone. The stimulatory effects of estrogen on 1α-hydroxylase activity measured in extracts of kidney tissue from chicks and Japanese quail [131–133] could not be confirmed as direct ones in cell culture [134]. Similarly, although the negative influence of glucocorticoids on bone health and the vitamin D endocrine system is well documented in a number
78
HELEN L. HENRY
of animal models and clinical studies (see Chapter 72), it is apparent that these effects cannot be attributed in a substantial way to direct alteration of vitamin D metabolism. The effects of glucocorticoids on plasma levels of 1α,25(OH)2D3 are variable [135–137], and dexamethasone exhibited a slight inhibitory effect on 1α-hydroxylase activity in cultured chick kidney cells [138]. A more recent study in mice showed that under conditions of low dietary calcium only, treatment of animals slightly reduced 1α-hydroxylase mRNA and activity in the kidney [139]. Thus, although interactions between other endocrine systems and the regulation of mineral metabolism by 1α,25(OH)2D3 undoubtedly play a role in calcium homeostasis and bone mineral metabolism, they do not apparently take place in a significant way at the level of 1α,25(OH)2D3 production in the kidney. 4. ANTIMYCOTICS
The antimycotics ketoconazole and miconazole inhibit 1α-hydroxylase in cell culture [140] and were subsequently shown to substantially decrease serum levels of 1,25(OH)2D3 and active calcium transport in rats [141]. Clinical studies have suggested that there are lowered serum 1,25(OH)2D3 levels in individuals being treated with these antifungal agents [142]. In one report, ketoconazole was used to ameliorate hypercalcemia resulting from acute active tuberculosis [143]. These effects of the antimycotics are similar to those known for other steroid hormones since they occur at the level of steroid hydroxylation [144,145]. Currently the primary interest in these agents in the vitamin D field is focused on their inhibition of the 24R-hydroxylase and the catabolic pathway for 1α,25(OH)2D3 that it initiates [146–148]. 5. THE KLOTHO GENE PRODUCT
A mouse homozygous mutation in a single gene, named klotho, displays multiple disorders that are similar to those seen in human aging [149]. In addition, these mice and a similar strain that is null for the klotho gene [150] have greatly elevated plasma 1α,25(OH)2D3. The levels of mRNA encoding CYP1α are also elevated in the kidneys of these mice. The regulatory responses to 1α,25(OH)2D3, PTH, and calcitonin appear to be intact in these mice, suggesting that there may be an alternate regulatory circuit involved in governing 1α,25(OH)2D3 hydroxylase activity.
V. SUMMARY The 25(OH)D3-1α-hydroxylase is a member of the family of cytochrome P450–dependent enzymes,
specifically those mitochondrial enzymes that metabolize endogenous steroids such as cholesterol side-chain cleavage systems in the adrenal cortex and gonads and the adrenal 11β-hydroxylase. The CYP1α activity is widely distributed in vertebrates. Although the kidney is still considered to be the primary site of circulating levels of 1α,25(OH)2D3, it is clearly produced in many other cells as well, mostly as part of paracrine/ autocrine systems. The past several years has seen considerable advances in the field of vitamin D metabolism with the cloning of CYP1α so that the tools are now available to study its regulation at the molecular level. The elucidation of these mechanisms and how they differ in the kidney and extrarenal sites of 1α,25(OH)2D3 production is just one of the exciting areas in which we can expect new ideas and information in the future.
References 1. Fraser DR, Kodicek E 1970 Unique biosynthesis by kidney of a biologically active vitamin D metabolite. Nature 288:764–766. 2. Norman AW 1971 Evidence for a new kidney produced hormone, 1,25-dihydroxycholecalciferol, the proposed biologically active form of vitamin D. Am J Clin Nutr 24:1346–1351. 3. Akiba T, Endou H, Koseki C, Sakai F 1980 Localization of 25-hydroxyvitamin D3-1α-hydroxylase activity in the mammalian kidney. Biochem Biophys Res Commun 94:313–318. 4. Kawashima H, Torikai S, Kurokawa K 1981 Localization of 25-hydroxyvitamin D3 1α-hydroxylase and 24-hydroxylase along the rat nephron. Proc Natl Acad Sci USA 78:1199–1203. 5. Brunette MG, Chan M, Ferriere C, Roberts KD 1978 Site of 1,25(OH)2 vitamin D3 synthesis in the kidney. Nature 276:287–289. 6. Zehnder D, Bland R, Walker EA, Bradwell AR, Howie AJ, Hewison M, Stewart PM 1999 Expression of 25-hydroxyvitamin D3-1α-hydroxylase in the human kidney. J Am Soc Nephrol 10:2465–2473. 7. Zhang MY, Wang X, Wang JT, Compagnone NA, Mellon SH, Olson JL, Tenenhouse HS, Miller WL, Portale AA 2002 Dietary phosphorus transcriptionally regulates 25-hydroxyvitamin D-1alpha-hydroxylase gene expression in the proximal renal tubule. Endocrinology 143:587–595. 8. Yamagata M, Kimoto A, Michigami T, Nakayama M, Ozono K 2001 Hydroxylases involved in vitamin D metabolism are differentially expressed in murine embryonic kidney: application of whole mount in situ hybridization. Endocrinology 142:3223–3230. 9. Hartenbower DL, Stella FJ, Norman AW, Friedler RM, Coburn JW 1977 Impaired vitamin D metabolism in acute uremia. J Lab Clin Med 90:760–766. 10. Mawer EB, Taylor CM, Backhouse J, Lumb GA, Stanbury SW 1973 Failure of formation of 1,25-dihydroxycholecalciferol in chronic renal insufficiency. Lancet 1:626–628. 11. Eisman JA, Hamstra AJ, Kream BE, DeLuca HF 1976 1,25Dihydroxyvitamin D in biological fluids: A simplified and sensitive assay. Science 193:1021–1023. 12. Turner RT, Puzas JE, Forte MD, Lester GE, Gray TK, Howard GA, Baylink, DJ 1980 In vitro synthesis of 1α,25dihydroxycholecalciferol and 24,25-dihydroxycholecalciferol
CHAPTER 5 The 25-Hydroxyvitamin D 1α-Hydroxylase
13.
14.
15.
16. 17. 18. 19.
20.
21.
22. 23.
24.
25. 26.
27.
28.
by isolated calvarial cells. Proc Natl Acad Sci USA 77: 5720–5724. Puzas JE, Turner RT, Howard GA, Brand JS, Baylink DJ 1987 Synthesis of 1,25-dihydroxycholecalciferol and 24,25-dihydroxycholecalciferol by calvarial cells: Characterization of the enzyme systems. Biochem J 245:333–338. Pedrozo HA, Boyan BD, Mazock J, Dean DD, Gomez R, Schwartz Z 1999 TGFβ1 regulates 25-hydroxyvitamin D3, 1α- and 24-hydroxylase activity in cultured growth plate chondrocytes in a maturation-dependent manner. Calcif Tissue Int 64:50–56. Schwartz Z, Brooks B, Swain L, Del Toro F, Norman AW, Boyan BD 1992 Production of 1,25-dihydroxyvitamin D3 and 24,25-dihydroxyvitamin D3 by growth zone and resting zone chondrocytes is dependent on cell maturation and is regulated by hormones and growth factors. Endocrinology 130:2495–2504. Bikle DD, Nemanic MK, Gee E, Elias P 1986 1,25Dihydroxyvitamin D3 production by human keratinocytes. J Clin Invest 78:557–566. Bikle DD, Nemanic MK, Whitney JO, Elias PW 1986 Neonatal human foreskin keratinocytes produce 1,25dihydroxyvitamin D3. Biochem 25:1545–1548. Adams JS, Sharma OP, Gacad MA, Singer FR 1983 Metabolism of 25-hydroxyvitamin D3 by cultured pulmonary alveolar macrophages in sarcoidosis. J Clin Invest 72:1856–1860. Reichel H, Koeffler HP, Tobler A, Norman AW 1987 1α,25Dihydroxyvitamin D3 inhibits γ-interferon synthesis by normal human peripheral blood lymphocytes. Proc Natl Acad Sci USA 84:3385–3389. Reichel H, Bishop JE, Koeffler HP, Norman AW 1987 25-Dihydroxyvitamin D3 metabolism by lipopolysaccharidestimulated normal human macrophages. J Clin Endocr Metab 64:1–9. Adams JS, Singer FR, Gacad MA, Sharma OP, Hayes MJ, Vouros P, Holick MF 1985 Isolation and structural identification of 1,25-dihydroxyvitamin D3 produced by cultured alveolar macrophages in sarcoidosis. J Clin Endocr Metab 60:960–966. Mawer EB, Hayes ME 1992 Renal and extra-renal synthesis of 1,25-dihydroxy vitamin D. Prog Endocrinol 87:382–386. Cadranel J, Garabedian M, Milleron B, Guillozo H, Akoun G, Hance AJ 1990 1,25(OH)2D3 production by T lymphocytes and alveolar macrophages recovered by lavage from normocalcemic patients with tuberculosis. J Clin Invest 85: 1588–1593. Pryke AM, Duggan C, White CP, Posen S, Mason RS 1990 Tumor necrosis factor-α induces vitamin D-1-hydroxylase activity in normal human alveolar macrophages. J Cell Physiol 142:652–656. Delvin EE, Arabian A 1987 Kinetics and regulation of 25-hydroxycholecalciferol 1-α−hydroxylase from cells isolated from human term decidua. Eur J Biochem 164:659–662. Zerwekh JE, Breslau NA 1986 Human placental production of 1,25-dihydroxyvitamin D3: Biochemical characterization and production in normal subjects and patients with pseudohyperparathyroidism. J Clin Endocr Metab 62:192. Hollis BW, Iskersky VN, Chang MK 1989 In vitro metabolism of 25-hydroxyvitamin D3 by human trophoblastic homogenates, mitochondria, and microsomes: Lack of evidence for the presence of 25-hydroxyvitamin D3-1α-hydroxylase and 24R-hydroxylase. Endocrinology 125:1224–1230. Diaz L, Sanchez I, Avila E, Halhali A, Vilchis F, Larrea F 2000 Identification of a 25-hydroxyvitamin D3 1α-hydroxylase gene transcription product in cultures of human syncytiotrophoblast cells [In Process Citation]. J Clin Endocrinol Metab 85: 2543–2549.
79 29. Zehnder D, Bland R, Williams MC, McNinch RW, Howie AJ, Stewart PM, Hewison M 2001 Extrarenal expression of 25-hydroxyvitamin D3-1α−hydroxylase. J Clin End Metab 86: 888–894. 30. Segersten U, Correa P, Hewison M, Hellman P, Dralle H, Carling T, Akerstrom G, Westin G 2002 25-hydroxyvitamin D(3)-1alpha-hydroxylase expression in normal and pathological parathyroid glands. J Clin Endocrinol Metab 87:2967–2972. 31. Schwartz GG, Whitlatch LW, Chen TC, Lokeshwar BL, Holick MF 1998 Human prostate cells synthesize 1,25-dihydroxyvitamin D3 from 25-hydroxyvitamin D3. Cancer Epidemiol Biomarkers Prev 7:391–395. 32. Jones G, Ramshaw H, Zhang A, Cook R, White J, Petkovich M 1999 Expression and activity of vitamin D-metabolizing cytochrome P450s (CYP1α and CYP24) in human nonsmall cell lung carcinomas. Endocrinology 140:3303–3310. 33. Henry HL, Norman AW 1975 Presence of renal 25-hydroxyvitamin-D-1-hydroxylase in species of all vertebrate classes. Comp Biochem Physiol 50B:441–434. 34. Nahm TH, Lee SW, Fausto A, Sonn Y, Avioli LV 1979 25-(OH)D, a circulating vitamin D metabolite in fish. Biochem Biophys Res Commun 89:396–402. 35. Egaas E, Lambertsen G 1979 Naturally occurring vitamin D3 in fish products analyzed by high performance liquid chromatography using vitamin D2 as an internal standard. J Am Oil Chem Soc 56:188. 36. Takeuchi A, Okano T, Kobayashi T 1991 The existence of 25-hydroxyvitamin D3-1α-hydroxylase in the liver of carp and bastard halibut. Life Sci 48:275–282. 37. Brown PB, Robinson EH 1992 Vitamin D studies with channel catfish (Ictalurus punctatus) reared in calcium-free water. Comp Biochem Physiol 103A:213–219. 38. Rao DS, Raghuramulu N 1996 Food chain as origin of vitamin D in fish. Comp Biochem Physiol 114A:15–19. 39. Rao DS, Raghuramulu N 1999 Vitamin D3 and its metabolites have no role in calcium and phosphorus metabolism in Tilapia mossambica. J Nutr Sci Vitaminol (Tokyo) 45:9–19. 40. Buffenstein R, Sergeev IN, Pettifor JM 1993 Vitamin D hydroxylases and their regulation in a naturally vitamin D– deficient subterranean mammal, the naked mole rat (Heterocephalus glaber). J Endcrinol 138:59–64. 41. Sergeev IN, Buffenstein R, Pettifor JM 1993 Vitamin D receptors in a naturally vitamin D-deficient subterranean mammal, the naked mole rat (Heterocephalus glaber): Biochemical characterization. Gen Comp Endocrinol 90:338–345. 42. Henry HL, Norman AW 1974 Studies on calciferol metabolism. IX. Renal 25-hydroxyvitamin D3-1-hydroxylase. Involvement of cytochrome P-450 and other properties. J Biol Chem 249: 7529–7535. 43. Hagenfeldt Y, Pedersen JI, Bjorkhem I 1988 Properties of guinea-pig kidney 25-hydroxyvitamin D3-1α-hydroxylase assayed by isotope dilution-mass spectrometry. Biochem J 250:521–526. 44. Churchill PF, deAlvare LR, Kimura T 1978 Topological studies of the steroid hydroxylase complexes in bovine adrenocortical mitochondria. J Biol Chem 253:4924–4929. 45. Ghazarian JG, Jefcoate CR, Knutson JC, Orme-Johnson W, DeLuca HF 1974 Mitochondrial cytochrome P450. A component of chick kidney 25-hydroxycholecalciferol1α−hydroxylase. J Biol Chem 249:3026–3033. 46. Hiwatashi A, Nishii Y, Ichikawa Y 1982 Purification of cytochrome P-450-1α (25-hydroxyvitamin D3-1α hydroxylase) of bovine kidney mitochondria. Biochem Biophys Res Commun 105:320–327.
80 47. Burgos-Trinidad M, Ismail R, Ettinger RA, Prahl JM, DeLuca HF 1992 Immunopurified 25-hydroxyvitamin D 1α-hydroxylase and 1,25-dihydroxyvitamin D 24-hydroxylase are closely related but distinct enzymes. J Biol Chem 267:3498–3505. 48. Moorthy B, Mandel ML, Ghazarian JG 1991 Amino-terminal sequence homology of two chick kidney mitochondrial proteins immunoisolated with monoclonal antibodies to the cytochrome P450 of the 25-hydroxyvitamin D3-1α hydroxylase. J Bone Miner Res 6:199–204. 49. Wakino S, Meguro M, Suzuki H, Saruta T, Ogishima T, Ishimura Y, Shinki T, Suda T 1996 Evidence for 54-kD protein in chicken kidney as a cytochrome P450 with a high molecular activity of 25-hydroxyvitamin D3 1α-hydroxylase. Gerentology 42:67–77. 50. Gray RW, Ghazarian JG 1989 Solubilization and reconstitution of kidney 25-hydroxyvitamin D3 1α- and 24-hydroxylases from vitamin D-replete pigs. Biochem J 259:561–568. 51. Postlind H 1990 Separation of the cytochromes P-450 in pig kidney mitochondria catalyzing 1α-, 24- and 26-hydroxylations of 25-hydroxyvitamin D3. Biochem Biophys Res Commun 168:261–266. 52. Nelson DR 1999 Cytochrome P450 and the individuality of species. Arch Biochem Biophys 369:1–10. 53. Okamura T, John ME, Zuber MX, Simpson ER, Waterman MR 1985 Molecular cloning and amino acid sequence of the precursor form of bovine adrenodoxin: evidence for a previously unidentified COOH-terminal peptide. Proc Natl Acad Sci USA 82:5705–5709. 54. Sakihama N, Hiwatashi A, Miyatake A, Shin M, Ichikawa Y 1988 Isolation and purification of mature bovine adrenocortical ferredoxin with an elongated carboxyl end. Arch Biochem Biophys 264:23–29. 55. Cupp JR Vickery LE 1989 Adrenodoxin with a COOHterminal deletion (des 116-128) exhibits enhanced activity. J Biol Chem 264:1602–1607. 56. Coughlan VM, Cupp JR, Vickery LE 1988 Purification and characterization of human placental ferredoxin. Arch Biochem Biophys 264:376–382. 57. Tang C, Kain SR, Henry HL 1993 The phorbol ester 12-O-tetradecanoyl-phorbol-13-acetate stimulates the dephosphorylation of mitochondrial ferredoxin in cultured chick kidney cells. Endocrinology 133:1823–1829. 58. Yoon PS, DeLuca HF 1980 Purification and properties of chick renal mitochondrial ferredoxin. Biochemistry 19:2165–2171. 59. Driscoll WJ, Omdahl JL 1989. Characterization and N-terminal amino acid sequence of multiple ferredoxins in kidney and adrenal mitochondria. Eur J Biochem 185:181–187. 60. Tang C, Henry HL 1993 Overexpression in Escherichia coli and affinity purification of chick kidney ferredoxin [published erratum appears in J Biol Chem 1993 268(15):11474]. J Biol Chem 268:5069–5076. 61. Omdahl JL, Wilson K, Swerdlow H, Driscoll WJ 1992 Molecular cloning and immunological characterization of porcine kidney ferredoxin. Arch Biochem Biophys 293:213–218. 62. Blanchard RK, Henry HL 1996 Chick kidney ferredoxin: Complementary DNA cloning and vitamin D effects on mRNA levels. Comp Biochem Physiol [B] 114B:337–344. 63. Kagimoto K, McCarthy JL, Waterman MR, Kagimoto M 1988 Deduced amino acid sequence ot mature chick testis ferredoxin. Biochem Biophys Res Commun 155:379–383. 64. Kagimoto M, Kagimoto K, Simpson ER, Waterman MR 1988 Transcription of the bovine adrenodoxin gene produces two species of mRNA of which only one is translated into adrenodoxin. J Biol Chem 263:8925–8928.
HELEN L. HENRY
65. Chang C-Y, Wu D-A, Mohandas TK, Chung B-C 1990 Structure, sequence, chromosomal location and evolution of the human ferredoxin gene. DNA Cell Biol 9:205–212. 66. Voutilainen R, Picado-Leonard J, DiBlasio AM, Miller WL 1988 Hormonal and developmental regulation of adrenodoxin messenger ribonucleic acid in steroidogenic tissues. J Clin Endocrinol Metab 66:383–388. 67. John ME, John TS, Boggaram V, Simpson ER, Waterman MR 1986 Transcriptional regulation of steroid hydroxylase genes by corticotropin. Proc Natl Acad Sci USA 83:4715–4719. 68. Kramer RE, Anderson CM, Peterson JA, Simpson ER, Waterman MR 1982 Adrenodoxin biosynthesis by bovine adrenal cells in monolayer culture. J Biol Chem 257: 14921–14925. 69. Miller WL 1988 Molecular biology of steroid hormone synthesis. Endocr Rev 9:295–318. 70. Simpson ER, Waterman MR 1983 Regulation by ACTH of steroid hormone biosynthesis in the adrenal cortex. Can J Biochem Cell Biol 61:692–707. 71. Gnanaiah W, Omdahl JL 1986 Isolation and characterization of pig kidney mitochondrial ferredoxin: NADP+ oxidoreductase. J Biol Chem 261:12649–12654. 72. Chu JW, Kimura T 1973 Studies on the adrenal steroid hydroxylases. Molecular and catalytic properties of adrenodoxin reductase (a flavoprotein). J Biol Chem 248:2089–2094. 73. Brandt ME, Vickery LE 1992 Expression and characterization of human mitochondrial ferredoxin reductase in Escherichia coli. Arch Biochem Biophys 294:735–740. 74. Solish SB, Picado-Leonard J, Morel Y, Kuhn RW, Mohandas TK, Hanukoglu I, Miller WL 1988 Human adrenodoxin reductase: Two mRNAs encoded by a single gene on chromosome 17cen q25 are expressed in steroidogenic tissues. Proc Natl Acad Sci USA 85:7104–7108. 75. Gacad MA, Adams JS 1993 Identification of a competitive binding component in vitamin D-resistant New World primate cells with a low affinity but high capacity for 1,25-dihydroxyvitamin D3. J Bone Miner Res 8:27–35. 76. Wu S, Ren S., Chen H, Chun RF, Gacad MA, Adams JS 2000 Intracellular vitamin D binding proteins: Novel facilitators of vitamin D-directed transactivation. Mol Endocrinol 14: 1387–1397. 77. Wu S, Chun R, Gacad MA, Ren S, Chen H, Adams JS 2002 Regulation of 1,25-dihydroxyvitamin D synthesis by intracellular vitamin D binding protein-1. Endocrinology 143:4135–4138. 78. Nykjaer A, Dragun D, Walther D, Vorum H, Jacobsen C, Herz J, Melsen F, Christensen EI, Willnow, TE 1999 An endocytic pathway essential for renal uptake and activation of the steroid 25-(OH) vitamin D3. Cell 96:507–515. 79. Hilpert J, Wogensen L, Thykjaer T, Wellner M, Schlichting U, Orntoft TF, Bachmann S, Nykjaer A, Willnow TE 2002 Expression profiling confirms the role of endocytic receptor megalin in renal vitamin D3 metabolism. Kidney Int 62:1672–1681. 80. Leheste JR, Melsen F, Wellner M, Jansen P, Schlichting U, Renner-Muller I, Andreassen TT, Wolf E, Bachmann S, Nykjaer A, Willnow TE 2003 Hypocalcemia and osteopathy in mice with kidney-specific megalin gene defect. FASEB J 17:247–249. 81. Takeyama K, Kitanaka S, Sato T, Kobori M, Yanagisawa J, Kato S 1997 25-Hydroxyvitamin D3 1α-hydroxylase and vitamin D synthesis. Science 277:1827–1830. 82. Shinki T, Shimada H, Wakino S, Anazawa H, Hayashi M, Saruta T, DeLuca HF, Suda T 1997 Cloning and expression of rat 25-hydroxyvitamin D3-1α-hydroxylase cDNA. Proc Natl Acad Sci USA 94:12920–12925.
CHAPTER 5 The 25-Hydroxyvitamin D 1α-Hydroxylase
83. St-Arnaud R, Messerlian S, Moir JM, Omdahl JL, Glorieux FH 1997 The 25-hydroxyvitamin D 1α-hydroxylase gene maps to the pseudovitamin D deficiency rickets (PDDR) disease locus. J Bone Miner Res 12:1552–1559. 84. Fu GK, Lin D, Zhang MY, Bikle DD, Shackleton CH, Miller WL, Portale AA 1997 Cloning of human 25-hydroxyvitamin D1α-hydroxylase and mutations causing vitamin D-dependent rickets type 1. Mol Endocrinol 11:1961–1970. 85. Monkawa T, Yoshida T, Wakino S, Shinki T, Anazawa H, DeLuca HF, Suda T, Hayashi M, Saruta T 1997 Molecular cloning of cDNA and genomic DNA for human 25-hydroxyvitamin D3 1α-hydroxylase. Biochem Biophys Res Commun 239:527–533. 86. Yoshida T, Yoshida N, Nakamura A, Monkawa T, Hayashi M, Saruta T 1999 Cloning of porcine 25-hydroxyvitamin D3 1α-hydroxylase and its regulation by cAMP in LLC-PK1 cells. J Am Soc Nephrol 10:963–970. 87. Fu GK, Portale A, Miller WL 1997 Complete structure of the human gene for the vitamin D 1α−hydroxylase, P450c1α. DNA Cell Biol 16:1961–1970. 88. Kong XF, Zhu XH, Pei YL, Jackson DM, Holick MF 1999 Molecular cloning, characterization, and promoter analysis of the human 25-hydroxyvitamin D3-1α-hydroxylase gene. Proc Natl Acad Sci USA 96:6988–6993. 89. Panda DK, Kawas SA, Seldin MF, Hendy GN, Goltzman D 2001 25-Hydroxyvitamin D 1α-hydroxylase: Structure of the mouse gene, chromosomal assignment, and developmental expression. J Bone Miner Res 16:46–56. 90. Henry HL, Midgett RJ, Norman AW 1974 Studies on calciferol metabolism X: Regulation of 25-hydroxyvitamin D3-1-hydroxylase, in vivo. J Biol Chem 249:7584–7592. 91. Colston KW, Evans IM, Spelsberg TC, Macintyre I 1977 Feedback regulation of vitamin D metabolism by 1,25-dihydroxycholecalciferol. Biochem J 164:83–89. 92. Takahashi N, Miyaura C, Horiuchi N, Suda T 1981 Regulation and its refractoriness of 25-hydroxyvitamin D3 metabolism in vitamin D deficiency. J Nutr Sci Vitaminol 27:429–437. 93. Henry HL 1979 Regulation of the hydroxylation of 25hydroxyvitamin D3 in vivo and in primary cultures of chick kidney cells. J Biol Chem 254:2722–2729. 94. Spanos E 1983 Regulation of the renal 25OHD3 hydroxylases by vitamin-D3 metabolites and analogs in primary chick kidney cell cultures. Acta Endocrinol 104:25–26. 95. Turner RT, Bottemiller BL, Howard GA, Baylink DJ 1980 In vitro metabolism of 25-hydroxyvitamin D3 by isolated rat kidney cells. Proc Natl Acad Sci USA 77:1537–1540. 96. Murayama A, Takeyama K, Kitanaka S, Kodera Y, Kawaguchi Y, Hosoya T, Kato S 1999 Positive and negative regulations of the renal 25-hydroxyvitamin D3 1α-hydroxylase gene by parathyroid hormone, calcitonin, and 1α,25(OH)2D3 in intact animals. Endocrinology 140:2224–2231. 97. Shinki T, Ueno Y, DeLuca HF, Suda T 1999 Calcitonin is a major regulator for the expression of renal 25-hydroxyvitamin D3-1alpha-hydroxylase gene in normocalcemic rats. Proc Natl Acad Sci USA 96:8253–8258. 98. Condamine L, Vztovsnik F, Friedlander G, Menaa C, Garabédian M 1994 Local action of phosphate depletion and insulin-like growth factor 1 on in vitro production of 1,25dihydroxyvitamin D3 by cultured mammalian kidney cells. J Clin Invest 94:1673–1679. 99. Bland R, Walker EA, Hughes SV, Stewart PM, Hewison M 1999 Constitutive expression of 25-hydroxyvitamin D3-1αhydroxylase in a transformed human proximal tubule cell line: evidence for direct regulation of vitamin D metabolism by calcium. Endocrinology 140:2027–2034.
81 100. Yoshida T, Yoshino J, Hayashi M, Saruta T 2002 Identification of a renal proximal tubular cell-specific enhancer in the mouse 25-hydroxyvitamin D 1alpha-hydroxylase gene. J Am Soc Nephrol 13:1455–1463. 101. Fraser DR, Kodicek E 1973 Regulation of 25-hydroxycholecalciferol-1-hydroxylase activity in kidney by parathyroid hormone. Nature New Biol 241:163–166. 102. Baksi SN, Kenny AD 1979 Acute effects of parathyroid extract on renal vitamin D hydroxylases in Japanese quail. Pharmacology 18:169–174. 103. Booth BE, Tsai HC, Morris RC 1985 Vitamin-D status regulates 25-hydroxyvitamin-D3-1-α-hydroxylase and its responsiveness to parathyroid hormone in the chick. J Clin Invest 75:155–161. 104. Walker AT, Stewart AF, Korn EA, Shiratori T, Mitnick MA, Carpenter TO 1990 Effect of parathyroid hormone-like peptides on 25-hydroxyvitamin D-1α-hydroxylase activity in rodents. Am J Physiol 258:E297–E303. 105. Armbrecht HJ, Boltz MA, Hodam TL 2003 PTH increases renal 25(OH)D3-1alpha-hydroxylase (CYP1alpha) mRNA but not renal 1,25(OH)2D3 production in adult rats. Am J Physiol Renal Physiol 284:F1032–F1036. 106. Spanos E, Brown DJ, Macintyre I 1979 Regulation of 25(OH)-D3 metabolism by parathyroid hormone in primary chick kidney cell cultures. FEBS Lett 105:31–34. 107. Henry HL 1981 Insulin permits parathyroid hormone stimulation of 1,25-dihydroxyvitamin D3 production in cultured kidney cells. Endocrinology 108:733–735. 108. Rost CR, Bikle DD, Kaplan RA 1981 In vitro stimulation of 25-hydroxycholecalciferol 1α-hydroxylation by parathyroid hormone in chick kidney slices: Evidence for a role for adenosine 3′5′-monophosphate. Endocrinology 108:1002–1006. 109. Fukase M, Birge SJ, Rifas L, Avioli LV, Chase LR 1982 Regulation of 25 hydroxyvitamin D3-1-hydroxylase in serumfree monolayer culture of mouse kidney. Endocrinology 110: 1073–1075. 110. Kramer R, Goltzman D 1982 Parathyroid hormone stimulates mammalian 25-hydroxyvitamin D3-1α-hydroxylase in vitro. Endocrinology 110:294–296. 111. Henry HL 1985 Parathyroid hormone modulation of 25-hydroxyvitamin D3 metabolism by cultured chick kidney cells is mimicked and enhanced by forskolin. Endocrinology 116:503–507. 112. Noland TA, Henry HL 1983 Protein phosphorylation in chick kidney: Response to parathyroid hormone, cyclic AMP, calcium and phosphatidyl serine. J Biol Chem 258: 538–546. 113. Henry HL, Luntao EM 1989 Interactions between intracellular signals involved in the regulation of 25-hydroxyvitamin D3 metabolism. Endocrinology 124:2228–2234. 114. Ro HK, Tembe V, Favus MJ 1992 Evidence that activation of protein kinase C can stimulate 1,25-dihydroxyvitamin D3 secretion by rat proximal tubules. Endocrinology 131: 1424–1428. 115. Dunlay R, Hruska KA 1990 PTH receptor coupling to phospholipase C is an alternative pathway of signal transduction in bone and kidney. Am J Physiol 258:F223–F231. 116. Galante L, Colston KW, MacAuley SJ, Macintyre I 1972 Effect of calcitonin on vitamin D metabolism. Nature 238: 271–273. 117. Nesbitt T, Lobaugh B, Drezner MK 1987 Calcitonin stimulation of renal 25-hydroxyvitamin-D-1α-hydroxylase activity in hypophosphatemic mice—evidence that the regulation of calcitriol production is not universally abnormal in x-linked hypophosphatemia. J Clin Invest 79:15–19.
82 118. Horiuchi N, Takashita H, Matsumoto T, Takahashi N, Shimazawa E, Suda T, Ogata E 1979 Salmon calcitonininduced stimulation of 1α,25-dihydroxycholecalciferol synthesis in rats involving a mechanism independent of adenosine 3′:5′:-cyclic monophosphate. Biochem J 184:269–275. 119. Larkins RG, MacAuley SJ, Rapoport A, Martin TJ, Tullock BR, Macintyre I 1974 Effects of nucleotides,hormones, ions, and 1,25-dihydroxycholecalciferol on 1,25-dihydroxycholecalciferol production in isolated chick renal tubules. Clin Sci Mol Med 46:569–582. 120. Rasmussen H, Wong M, Bikle DD, Goodman DBP 1972 Hormonal control of the renal conversion of 25-hydroxycholecalciferol to 1,25-dihydroxycholecalciferol. J Clin Invest 51:2502–2504. 121. Henry HL, Noland TA, Al-Abdaly FA, Cunningham NS, Luntao EM, Amdahl LD 1982 Responses of cultured chick kidney cells to parathyroid hormone and calcitonin. In: Norman AW, Schaefer K, Herrath DV, Grigoleit H-G (eds.) Vitamin D: Chemical, Biochemical, and Clinical Endocrinology of Calcium Metabolism. Walter de Gruyter, Berlin. 122. Yoshida N, Yoshida T, Nakamura A, Monkawa T, Hayashi M, Saruta T 1999 Calcitonin induces 25-hydroxyvitamin D3 1αhydroxylase mRNA expression via protein kinase C pathway in LLC-PK1 cells. J Am Soc Nephrol 10:2474–2479. 123. Hughes MR, Brumbaugh PF, Hussler MR, Wergedal JE, Baylink DJ 1975 Regulation of serum 1alpha,25-dihydroxyvitamin D3 by calcium and phosphate in the rat. Science 190:578–580. 124. Baxter LA, DeLuca HF 1976 Stimulation of 25-hydroxyvitamin D3-1α-hydroxylase by phosphate depletion. J Biol Chem 251:3158–3161. 125. Gray RW, Napoli JL 1983 Dietary phosphate deprivation increase 1,25-dihydroxyvitamin D3 synthesis in rat kidney in vitro. J Biol Chem 258:1152–1155. 126. Rader JI, Baylink DJ, Hughes MR, Safilian EF, Haussler MR 1979 Calcium and phosphorus deficiency in rats: Effect on PTH and 1,25-dihydroxyvitamin D3. Am J Physiol 236: E118–E122. 127. Gray RW, Wilz DR, Caldas AE, Lemann J 1977 The importance of phosphate in regulating plasma 1,25-(OH)2-vitamin D levels in humans: Studies in healthy subjects, in calcium-stone formers and in patients with primary hyperparathyroidism. J Clin Endocrinol Metab 45:299–306. 128. Maierhofer WJ, Gray RW, Lemann J 1984 Phosphate deprivation increases serum 1,25-(OH)2-vitamin-D concentrations in healthy men. Kidney Int 25:571–575. 129. Portale AA, Halloran BP, Murphy MM, Morris FC 1986 Oral intake of phosphorus can determine the serum concentration of 1,25-dihydroxyvitamin D by determining its production rate in humans. J Clin Invest 77:7–12. 130. Portale AA, Halloran BP, Murphy MM, Morris RC, Jr. 1986 Oral intake of phosphorus can determine the serum concentration of 1,25-dihydroxyvitamin D by determining its production rate in humans. J Clin Invest 77:7–12. 131. Castillo L, Tanaka Y, DeLuca HF, Sunde ML 1977 The stimulation of 25-hydroxyvitamin D3-1α-hydroxylase by estrogen. Arch Biochem Biophys 179:211–217. 132. Baksi SN, Kenny AD 1980 Estradiol-induced stimulation of 25-hydroxyvitamin D3-1-hydroxylase in vitamin D-deficient Japanese quail. Pharmacology 20:298–303. 133. Pike JW, Spanos E, Coloston KW, Macintyre I, Haussler MR 1978 Influence of estrogen on renal vitamin D hydroxylases and serum 1α,25-(OH)2D3 in chicks. Am J Physiol 235: E338–E343.
HELEN L. HENRY
134. Henry HL 1981 25-OH-D3 metabolism in kidney cell cultures: Lack of a direct effect of estradiol. Am J Physiol 240:E119–E124. 135. Findling JW, Adams ND, Lemann J Jr., Gray RW, Thomas CJ, Tyrrell JB 1982 Vitamin D metabolites and parathyroid hormone in Cushing’s syndrome: Relationship to calcium and phosphorus homeostasis. J Clin Endocrinol Metab 54: 1039–1044. 136. Chesney RW, Hamstra AJ, Mazess RB, DeLuca HF, O’Reagan S 1978 Reduction of serum 1,25-dihydroxyvitamin D3 in children receiving glucocorticoids. Lancet 2:1123–1125. 137. Seeman E, Kumar R, Hunder GG, Scott M, Heath H, Riggs BL 1980 Production, degradation, and circulating levels of 1,25Dihydroxyvitamin D in health and in chronic glucocorticoid excess. J Clin Invest 66:664–669. 138. Henry HL 1986 Effect of dexamethasone on 25-hydroxyvitamin D3 metabolism in chick kidney cell cultures. Endocrinology 118:940–944. 139. Akeno N, Matsunuma A, Maeda T, Kawane T, Horiuchi N 2000 Regulation of vitamin D-1α-hydroxylase and 24-hydroxylase expression by dexamethasone in mouse kidney. J Endocrinol 164:339–348. 140. Henry HL 1985 Effect of ketoconazole and miconazole on 25-hydroxyvitamin D3 metabolism by cultured chick kidney cells. J Steroid Biochem Mol Biol 23:991–994. 141. Boass A, Toverud SU 1995 Suppression of circulating calcitriol and duodenal active Ca transport by ketoconazole in pregnant rats. Am J Physiol Endocrinol Metab 269: E934–E939. 142. Glass AR, Eil C 1986 Ketoconazole-induced reduction in serum 1,25-dihydroxyvitamin D. J Clin Endocr Metab 63: 766–769. 143. Saggese G, Bertelloni S, Baroncelli GI, Di Nero G 1993 Ketoconazole decreases the serum ionized calcium and 1,25-dihydroxyvitamin D levels in tuberculosis-associated hypercalcemia. Am J Dis Child 147:270–273. 144. Kowal J 1983 The effect of ketoconazole on steroidogenesis in cultured mouse adrenal cortex tumor cells. Endocrinology 112:1541–1543. 145. Loose DS, Kan PB, Hirst MA, Marcas RA, Feldman D 1983 Ketoconazole blocks adrenal steroidogenesis by inhibiting cytochrome P450-dependent enzymes. J Clin Invest 71: 1495–1499. 146. Zhao J, Tan BK, Marcelis S, Verstuyf A, Bouillon R 1996 Enhancement of antiproliferative activity of 1α,25-dihydroxyvitamin D3 (analogs) by cytochrome P450 enzyme inhibitors is compound- and cell-type specific. J Steroid Biochem Mol Biol 57:197–202. 147. Peehl DM, Skowronski RJ, Leung GK, Wong ST, Stamey TA, Feldman D 1994 Antiproliferative effects of 1,25-dihydroxyvitamin D3 on primary cultures of human prostatic cells. Cancer Res 54:805–810. 148. Peehl DM, Seto E, Feldman D 2001 Rationale for combination ketoconazole/vitamin D treatment of prostate cancer. Urology 58:123–126. 149. Yoshida T, Fujimori T, Nabeshima Y 2002 Mediation of unusually high concentrations of 1,25-dihydroxyvitamin D in homozygous klotho mutant mice by increased expression of renal 1α-hydroxylase gene. Endocrinology 143:683–689. 150. Tsujikawa H, Kurotaki Y, Fujimori T, Fukuda K, Nabeshima YI 2003 Klotho, a gene related to a syndrome resembling human premature aging, functions in a negative regulatory circuit of vitamin D endocrine system. Mol Endocrinol.
CHAPTER 5 The 25-Hydroxyvitamin D 1α-Hydroxylase
151. Murayama A, Takeyama K, Kitanaka S, Kodera Y, Hosoya T, Kato S 1998 The promoter of the human 25-hydroxyvitamin D3 1α−hydroxylase gene confers positive and negative responsiveness to PTH, calcitonin, and 1α,25(OH)2D3. Biochem Biophys Res Commun 249:11–16. 152. Brenza HL, DeLuca HF 2000 Regulation of 25-hydroxyvitamin D3 1α-hydroxylase gene expression by parathyroid hormone and 1,25-dihydroxyvitamin D3 [In Process Citation]. Arch Biochem Biophys 381:143–152. 153. Bringhurst FR, Juppner H, Guo J, Urena P, Potts JT Jr Kronenberg HM, Abou-Samra AB, Segre GV 1993 Cloned, stably expressed parathyroid hormone (PTH)/PTH-related peptide receptors activate multiple messenger signals and biological responses in LLC-PK1 kidney cells. Endocrinology 132:2090-2098. 154. Racusen LC, Monteil C, Sgrignoli A, Lucskay M, Marouillat S, Rhim JG, Morin JP 1997 Cell lines with extended in vitro
83 growth potential from human renal proximal tubule: characterization, response to inducers, and comparison with established cell lines. J Lab Clin Med 129:318–329. 155. Brenza HL, Kimmel-Jehan C, Jehan F, Shinki T, Wakino S, Anazawa H, Suda T, DeLuca HF 1998 Parathyroid hormone activation of the 25-hydroxyvitamin D3-1α-hydroxylase gene promoter. Proc Natl Acad Sci USA 95:1387–1391. 156. Armbrecht HJ, Hodam TL, Boltz MA 2003 Hormonal regulation of 25-hydroxyvitamin D3-1alpha-hydroxylase and 24-hydroxylase gene transcription in opossum kidney cells. Arch Biochem Biophys 409:298–304. 157. Gao XH, Dwivedi PP, Choe S, Alba F, Morris HA, Omdahl JL, May BK 2002 Basal and parathyroid hormone induced expression of the human 25-hydroxyvitamin D 1alpha-hydroxylase gene promoter in kidney AOK-B50 cells: role of Sp1, Ets and CCAAT box protein binding sites. Int. J Biochem Cell Biol 34:921–930.
CHAPTER 6
The 25-Hydroxyvitamin D 24-Hydroxylase JOHN OMDAHL BRIAN MAY
Office of Research, University of New Mexico School of Medicine, Albuquerque, NM 87131-5166 School of Molecular and Biomedical Science, University of Adelaide, South Australia 5005, Australia
I. Background II. Enzyme Structure and Function III. Cellular Expression and Regulation
I. BACKGROUND A. Enzyme Function and Regulated Expression Vitamin D is a secosteroid whose biological function is dependent upon its metabolic activation and turnover. These metabolite pathways contain specific hydroxylase enzymes that are members of the cytochrome P450 superfamily of mixed-function monooxygenases. Bioactivation of vitamin D involves the sequential actions of 25-hydroxylase and 1-hydroxylase enzymes leading to the synthesis of the hormonally active secosteroid 1,25-dihydroxyvitamin D [1,25(OH)2D]* (Fig. 1). These two enzymes are discussed in Chapter 4 (vitamin D 25-hydroxylase) and Chapter 5 (25-hydroxyvitamin D 1α-hydroxylase) and will be mentioned in this chapter only on a comparative basis to 25-hydroxyvitamin D24(R)-hydroxylase cytochrome P450c24 (CYP24), the enzyme that directs the side-chain metabolism of 25-hydroxylated vitamin D metabolites, which leads to their terminal physiological inactivation and turnover. Most cellular actions of vitamin D are mediated through the secosteroid hormone 1,25(OH)2D and involve the transcription of vitamin D–dependent genes. These regulatory processes involve the coordinated modulation and coupling of rapid signal-transduction pathways with slower acting ligand-dependent transcription factors [1]. In both cases, the secosteroid ligand binds to a ligand-specific receptor. The rapid response receptor is located in the cellular membrane of target tissues and initiates rapid signaling responses through a receptor that has been referred to as the membrane-associated receptor [e.g., mVDR or
*The term vitamin D will be used to denote vitamin D3 unless stated otherwise. VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
IV. Molecular Aspects References
membrane-associated rapid response steroid (MSSRS) receptor complex] [2,3]. The hormone receptor for the transcription process involves the nuclear vitamin D receptor (nVDR), which is a ligand-dependent transcription factor that functions as a VDR:1,25(OH)2D heterodimeric complex with the cis-retinoic acid: RXR complex (i.e., VDR-RXR) to regulate vitamin D–dependent genes associated with development and homeostasis (see Chapters 13–17 for details). Cellular and ambient levels 1,25(OH)2D are regulated through the hormone’s synthesis and degradation. The regulated 1α-hydroxylase (i.e., P4540c1 or CYP27B1) directs the hormone’s biosynthesis. In a complementary manner, the regulatory 24(R)-hydroxylase enzyme (i.e., P450c24 or CYP24A1) mediates degradation of 1,25(OH)2D through side-chain oxidation reactions. Expression of the two enzymes is most commonly regulated in a diametric manner that functions to maintain the hormone’s level at a proper set point. For example, low-calcium states are associated with an elevation in parathyroid hormone (PTH) that functions to up-regulate P450c1 (i.e., CYP27B1) while down-regulating P450c24 (i.e., CYP24A1) expression. The associated increase in 1,25(OH)2D functions to stimulate calcium absorption and correct the low-calcium state. Control of 1,25(OH)2D overproduction and the attendant development of a hypercalcemic state is autoregulated through a negative feedback system whereby 1,25(OH)2D down-regulates its oversynthesis (i.e., suppresses CYP27B1 expression) while up-regulating expression of CYP24A1. Of particular importance is the action of CYP24A1 to catabolically inactivate 1,25(OH)2D and thereby function to counterregulate the hormone’s cellular and ambient concentration. The regulatory and expression linkages between CYP27B1 and CYP24A1 are fundamental to understanding the vitamin D regulatory pathway and will serve as a point of reference in this chapter as well as other Copyright © 2005, Elsevier, Inc. All rights reserved.
86
JOHN OMDAHL AND BRIAN MAY
Liver Liver UV light 7-dehydro cholesterol
Skin Skin
P450c25
Pre-Vitamin D
Diet
25(OH)D3 Kidney Kidney
Target Target Tissues Tissues P450c24
P450c1 Calcitroic acid
1,25(OH)2D3 24,25(OH)22D33
1,25(OH)22D33 RXR VDR
( )
P450c24
P450c1
1,24,25(OH)33D33 VDRE
Genes
P450c24 Calcitroic acid
Excretion
FIGURE 1
Activation and molecular pathways for vitamin D. Integrative schematic of vitamin D synthesis, metabolism and molecular action.
sections in this book in which various aspects of vitamin D metabolism are discussed extensively.
B. Enzyme Cellular Expression The 24(R)-hydroxylase enzyme displays a broad tissue distribution wherein it is expressed in nearly all cells. A number of examples are discussed in the regulatory section in the second half of this chapter. The major route of enzyme induction involves the combined action of 1,25(OH)2D and VDR to increase transcription of the CYP24A1 gene as discussed in the latter two sections of this chapter. Recognizing the importance of 1,25(OH)2D in the induction of CYP24A1, the enzyme is not expressed in vitamin D– deficient animals. Consequently, the cellular expression of CYP24A1 is linked tightly to the vitamin D status and coexpression of VDR. Because of the ubiquitous distribution of VDR [4,5], most cells express a basal level of CYP24A1 or respond to increased 1,25(OH)2D levels by inducing biosynthesis of the enzyme, particularly in kidney and small intestine [6–8]. Consequently, the cellular level of CYP24A1 is particularly influenced by the concentration of VDR and the level of 1,25(OH)2D, the latter reflecting both locally produced hormone [9] and hormone acquired from the circulation. Hence the broad cellular induction of CYP24A1 by 1,25(OH)2D can be envisioned as an effector arm of the negative feedback system that functions to maintain
proper hormone level and avoids the buildup of a potentially toxic secosteroid. Kidney is the major endocrine organ for 1,25(OH)2D and therefore there has been considerable interest in the regulation of renal CYP24A1 expression. Using the rat developmental model, Matkovits and Christakos [6] reported low levels of renal mRNA at birth that remained unchanged for several weeks and then increased substantially over the next 20 months, thereby establishing kidney as the major site of basal enzyme expression. In recent studies, quantitation of rat kidney mRNA levels using a real time RT-PCR protocol [10] confirmed this developmental profile for CYP24A1 mRNA expression (PH Anderson, S Iida, BK May, HA Morris, submitted). A small fall in renal CYP24A1 mRNA level was seen between 3 and 9 weeks of age followed by a continuous increase until 24 months. In a comparative measurement, CYP27B1 (i.e., 1αhydroxylase) mRNA was highest at 3 weeks of age and then decreased particularly between 3 and 15 weeks, followed by a further slight decrease over 24 months. The high production of renal 1,25(OH)2D during earlier development would ensure adequate circulating 1,25(OH)2D and hence calcium for bone growth. Serum levels of 1,25(OH)2D are determined by the expression levels of both CYP27B1 and CYP24A1 in the kidney [10]. CYP24A1 is expressed in both proximal and distal tubules [11–13], although induced enzyme expression occurs to a higher level in the proximal tubule [13].
87
CHAPTER 6 The 25-Hydroxyvitamin D 24-Hydroxylase
Mechanistic and regulatory studies of 1,25(OH)2D3 action and CYP24A1 expression have relied on a host of cell-culture models. Investigations with cultured cells have revealed induction of the CYP24A1 gene in many different cell types. Enzyme activity and/or CYP24A1 mRNA levels have been detected in numerous primary and transformed cultures that include intestinal, osteoblast, monocyte/macrophage, keratinocyte, and cancer cells [14–25]. In most of these studies, cells incubated in a serum-free medium expressed little if any 24(R)-hydroxylase activity or CYP24A1 mRNA, but enzyme and message levels were highly induced by 1,25(OH)2D treatment. In addition to 1,25(OH)2D, other more cell-type-selective regulators of CYP24A1 gene expression have been studied, and include PTH, calcitonin, phorbol ester, dexamethasone, and gammainterferon (see Section III of this chapter). The induction of CYP24A1 mRNA in primary cultures of rat renal tubular cells was prevented by actinomycin D, verifying that the increase in CYP24A1 mRNA level was due to gene transcription [15]. Cycloheximide also inhibited the induction of CYP24A1 mRNA in these cells, demonstrating a continued need for protein synthesis [15,19], which could involve both signal transduction and transcription regulatory proteins.
C. Enzyme Properties P450c24 acquired its name 25-hydroxyvitamin D 24(R)-hydroxylase based upon its action to 24(R)hydroxylate, the prehormone 25-hydroxyvitamin D3 [25OHD] to 24(R),25-dihydroxyvitamin D [24,25 (OH)2D] [26]. The P450c24 enzyme was subsequently shown to 24(R)-hydroxylate 1,25(OH)2D3 to 1,24,25 (OH)3D3 [27]. See Figure 2. The CYP24A1 in some species expresses predominantly 23(S)-hydroxylase activity [11,28]; therefore, CYP24A1 isoenzymes direct both C23 and C24 oxidative pathways. In both instances, the dominant C23 or C24 hydroxylation activity directs the initial step associated with the enzyme’s side-chain oxidative reactions associated with the inactivation of 25-hydroxylated vitamin D metabolites as discussed in the next section on enzyme pathways. The CYP24A1 enzyme is a protoporphyrin IX hemoprotein (~52 kDa) that binds oxygen and directs the regiospecific and stereospecific hydroxylation and oxidation of secosteroid substrate molecules using electrons derived from NADPH [29] (Fig. 2). The vitamin D enzyme is distinct from the 24(S)-hydroxylase enzyme involved with cholesterol metabolism [30] (i.e., CYP2R1). CYP24A1 is an inner mitochondrial membrane enzyme that is part of a mini–electron
Cyp24A1red
Cyp24A1ox
25 (OH)D3 + O2 [1,25 (OH)2D3]
24,25 (OH)2D3 + H2O [1,24,25 (OH)3D3]
FIGURE 2
Terminal reaction of CYP24A1. The use of molecular oxygen by reduced CYP24A1 to 24(R)-hydroxylate 25-hydroxyvitamin D substrates. red, reduced; ox, oxidized.
transport chain consisting of NADPH-ferredoxin reductase, ferredoxin (an iron–sulfur protein), and the terminal cytochrome CYP24A1 [31] (Fig. 2). Electrons are delivered through a specific protein:protein interaction between the enzyme and the electron donor ferredoxin [32]. Since the vitamin D 24(R)-hydroxylase enzyme is less than 40% similar to other P450s, it was assigned to a new P450 family and given the cytochrome P450 abbreviation CYP24A1. Consistent with its unique protein sequence, CYP24A1 does not react immunologically with several liver P450 enzymes or mitochondrial P450c27 (CYP27A1) that expresses 25-hydroxylase activity and is a family member with the 25-hydroxyvitamin D 1α-hydroxylase enzyme (i.e., CYP27).
D. Enzyme Pathways CYP24A1 directs the biosynthesis of 24,25(OH)2D and 1,24,25(OH)3D. These two vitamin D metabolites express less (i.e., 24,25(OH)2D) or approximately equivalent (i.e., 1,24,25(OH)3D) biological activity compared to 1,25(OH)2D. A biological role for 24,25 (OH)2D has been studied most extensively [2,3] because of its higher cellular and ambient concentration. Both metabolites are initial reactants in the C24-oxidation pathway that leads to the side-chain oxidation of 25-hydroxy metabolites to terminal C23-carboxylic acid end products [33–35] (Fig. 3). An investigation with partially purified 24(R)-hydroxylase enzyme from pig kidney documented copurification of the 23- and 24-hydroxylase activities [31], which was consistent with one enzyme expressing both activities. All catalytic activities in the C24-oxidation pathway were subsequently shown to be expressed by recombinant rat CYP24A1 [28,36]. The multifunctional enzyme initially oxidizes the 24-hydroxyl group to a 24-keto (oxo) functionality. Subsequent steps involve the 23(S)-hydroxylation of the 24-oxo metabolites and
88
JOHN OMDAHL AND BRIAN MAY
to favor an initial C23(S)- or C24(R)-hydroxylation. It is evident, therefore, that CYP24A1 is a multicatalytic enzyme [28,36–38] that is capable of directing all reactions in the side-chain oxidation of 25-hydroxylated vitamin D metabolites to cyclic lactone (C23 pathway) [38] or C23 carboxylic-acid (C24 pathway) [37], or both [28,38] end products (Fig. 3). In this regard, CYP24A1 is the central enzyme involved with determining the biological half-life of 1,25(OH)2D and its therapeutic analogs.
II. ENZYME STRUCTURE AND FUNCTION A. Cytochrome P450 Motif and Function
FIGURE 3
Multicatalytic action of CYP24A1. Reactions of the C23 and C24 oxidative pathways catalyzed by CYP24A1 with lactol and carboxylic acid end products.
oxidative cleavage between C23 and C24 followed by oxidation of the C23 hydroxyl, via a C23 aldehyde intermediate, to 24,25,26,27-tetranor-23-COOH-vitamin D or calcitroic acid from 25(OH)2D and 1,25(OH)2D substrates, respectively (Fig. 3). However, the initial hydroxylation occurs at C23 in several CYP24A1 isoforms from different species, which result in an alternate pathway that leads to a side-chain cyclic lactone as final product (Fig. 3). For example, rat CYP24A1 expresses predominantly the C24-oxidation pathway, whereas human [28] and guinea pig [11] enzymes express significant levels of C23-hydroxylase activity that results in the synthesis of both lactone and carboxylicacid end products. Expression of predominantly 23(S)or 24(R)-hydroxylase activity in different isoforms would appear to be determined by point-residue differences in the substrate binding and catalytic center that direct substrate alignment in the heme active site so as
Cytochromes P450 constitute a superfamily of hemecontaining enzymes that direct the monooxygenation of hydrophobic substrates. More than 1000 genes have been identified and divided into 70 families [39] that metabolize preferred substrates ranging from polycyclic aromatic hydrocarbons to fatty acids and steroids. Although all P450s have a similar structural core, they have unique amino acid sequences that determine the enzyme’s substrate recognition and mediate regiospecific and stereospecific monooxygenation reactions within the active site. Molecular oxygen used in the monooxygenase reactions is activated through its interaction with the iron atom in the active site hemeprotoporphyrin IX molecule. In most instances, the oxygen molecule is activated to an oxyferryl functionality that stereospecifically directs the oxygenation/ hydroxylation of substrates that enter through a surface oriented substrate-access channel and subsequently align with the heme molecule in the enzyme’s central active site.
B. Molecular Modeling Only a limited number of P450 enzymes have been crystallized and characterized by X-ray diffraction analysis [40–44]. However, because of the molecules’ conserved structural attributes, it is possible to use molecular-modeling software (e.g., Insight-II and Sybyl) to develop enzyme similarity models that closely resemble the enzyme’s native conformation. Using the Insight-II Modeler program (Accelrys Inc.) and crystallographic coordinate values for P450-BM3 and P450-2C5, it has been possible to derive a similarity model for rat P450c24 (CYP24A1) (A Annalara, S Graham, JL Omdahl, unpublished data). Most of the substrate-interactive residues in CYP24A1 are located predominantly in the B′, F, G, and I helices, and the F-G loop.
89
CHAPTER 6 The 25-Hydroxyvitamin D 24-Hydroxylase
FIGURE 4
Model of rat CYP24A1 active site. A similarity model for CYP24A1 active site that shows alignment of the side-chain C24 with the heme (red) and alignment of the A-ring (adjacent to the 1,25(OH)2D notation) with F-249 in the F-helix.
Substrate docking of 1,25(OH)2D in the CYP24A1 model placed the C24 position in close proximity (2 Å) to the heme oxyferryl group with the A-ring of the molecule oriented toward the F-helix (Fig. 4). The A-ring displayed close interactions (3 Å) with met-246 and phe-249 of the F-helix. A prominent role for phe-249 in binding 1,25(OH)2D was determined though site-directed mutagenesis as discussed in the following section. The CYP24A1 similarity model is being used in conjunction with site-directed mutagenesis to map the full spectrum of substrate-binding sites within the substrate-access channel and active site.
C. Metabolic, Kinetic, and Spectroscopic Analysis 1. METABOLIC ANALYSIS
Consistent with the P450 nature of CYP24A1, the enzyme’s activity is impeded by P450 azole inhibitors [45–47]. The CYP24A1 purified from rat kidney [29] was shown to metabolize 25(OH)D to 24,25(OH)2D and 1,25(OH)2D to 1,24,25(OH)3D. Subsequent studies using bacterial-expressed recombinant enzyme demonstrated the action of CYP24A1 to side-chain metabolize 1,25(OH)2D [37]. The role of CYP24A1 to
direct side-chain oxidation reactions beyond the initial 24(R)-hydroxylation was established using recombinant enzyme preparations from bacteria and baculovirus [28,36–38]. Most interestingly, the rat enzyme was demonstrated to metabolize all steps in the C24-oxidation pathway leading to C23-carboxylic acid end products [36,37]. In contrast to the rat enzyme, human CYP24A1 [28,38] was found to catalyze both the C24 and C23 oxidation pathways that resulted in the tetranor C23-carboxylic acid and the 26,23-lactone end products, respectively (Fig. 3). Evident from these studies was the multicatalytic activity of CYP24A1 isozymes to direct all reactions in the C24 and C23 oxidative pathways. In contrast to previous recombinant CYP24A1 studies that used enzyme membrane preparations or detergent extracts (i.e., impure enzyme), a more recent investigation succeeded in purifying bacterial recombinant rat CYP24A1 (A Annalora, JL Omdahl, unpublished data). The purified enzyme expressed all steps in the C24 oxidation pathways, which demonstrated clearly the multicatalytic activity of CYP24A1. 2. KINETIC ANALYSIS
The rate of substrate metabolism by CYP24A1 has been analyzed using cellular, detergent extracted, and
90
JOHN OMDAHL AND BRIAN MAY
purified enzyme preparations. Km values from cellular and detergent extract preparations [19,48–51] span a large range (0.01–1.5 µM) in which Km values for 1,25(OH)2D are 1/4 to 1/30 that of 25(OH)D, which supported the concept that CYP24A1 has a higher preference for 1,25(OH)2D hormone. However, a more recent study [52] with solubilized enzyme determined the Km for 25(OH)D to be 0.6 µM and that of 1,25(OH)2D was 35-fold higher (20.9 µM), which indicated an enzyme preference for 25(OH)D instead of 1,25(OH)2D. The diversity of Km values is due in part to the sequential–multicatalytic properties of CYP24A1 and the combined effect of electron transfer from the co-redox ferredoxin partner and dissociation of intermediate reactants from the enzyme. Using purified rat CYP24A1, it has been possible to focus on the first step of the multicatalytic reaction path and to regulate the rate of electron transfer to the enzyme (A Pastuezyn, JL Omdahl, unpublished data). Using these conditions and those that favor the 24(R)-hydroxylation step, we have determined Km values for 25(OH)D and 1,25(OH)2D that are less than 1 µM in which 25(OH)D had a slightly lower Km and greater turnover number (A Pastuezyn, JL Omdahl, unpublished data). Also evident from the kinetic analysis with pure enzyme was the slow reaction kinetics for CYP24A1 [i.e., turnover number (TN) of 2–20 min−1] [29], compared to the higher rate observed for another mitochondrial P450 from the bile-acid pathway (i.e., CYP27A1, TN = 40–50 min−1) [53,54]. 3. SPECTRAL ANALYSIS
CYP24A1 displays a typical type-I mitochondrial P450 CO-reduced difference spectrum with a 450-nm Soret band that shifts to 420 nm upon enzyme inactivation. Because of the presence of a heme molecule in the enzyme’s active site, it is possible to follow the absolute spectral shift from 420 to 390 nm that occurs upon addition of substrate (i.e., substrate-induced spectral shift) and determine the dissociation constant (Kd) for substrate with enzyme [55]. Such binding techniques revealed a high specificity of the enzyme for vitamin D secosteroids containing a 1α- and/or 25-hydroxyl group (A Annalora, JL Omdahl et al., submitted). Both 1α-hydroxyvitamin D and 1,25(OH)2D displayed the lowest Kd (highest binding affinity) for CYP24A1. However, only 1,25(OH)2D was capable of being 24(R)-hydroxylated [29], which demonstrated the necessity of a C25-hydroxyl group for the physiological side-chain 24(R)-hydroxylation and subsequent oxidation reactions catalyzed by CYP24A1. In a related manner, the side-chain metabolism of 1,25(OH)2D2 to calcitroic acid is impeded by the vitamin D2 side-chain
structure modifications (i.e., double bond at C22 and methyl group at C24) [35,56]. The Kd for 25(OH)D is sixfold higher than observed for 1,25(OH)2D. But because of its lower Km, higher turnover number, and 1000-fold higher ambient level, the conversion of 25(OH)D to 24,25(OH)2D is a favored reaction. In addition, the 24(R)-hydroxylated metabolite has a decreased affinity for CYP24A1 (A Annalora, JL Omdahl, submitted), which favors its dissociation from enzyme, especially in cells where electron transfer is limiting. The combined effects result in a significant serum level of 24,25(OH)2D (2–5 ng/ml) that is 1/5 to 1/10 the concentration of the parent prehormone 25(OH)D. In contrast, the ambient level of 1,24,25(OH)3D is insignificant because of the metabolite’s low cellular concentration and greater affinity to remain bound to CYP24A1 (i.e., lower Kd) for further oxidative metabolism. Introduction of a single mutation in the F-helix of CYP24A1 has a profound impact on substrate binding and perturbation of the heme spin-state in the active site [57] (A Annalora, JL Omdahl, unpublished data). Mutation of phenylalanine 249 to threonine (F249T) caused a three- to sixfold increase in the Kd for 25(OH)D and 1,25(OH)2D, respectively. Based upon the CYP24A1 similarity model, phenylalanine 249 is projected to interact with the A-ring of 1,25(OH)2D when bound in the active site (Fig. 4). Loss of this hydrophobic interaction through the F249T mutation prevents the hormone from binding properly in the substrate pocket. As a consequence, the last oxidative sequence in the C24 pathway (i.e., 24,25,26,27-tetranor-1,23(OH)2D to calcitroic acid) is suppressed because of the misalignment of substrate side-chain in the active site. Mutagenesis of phenylalanine 249 to alanine (F249A) or tyrosine (F249Y) also confounded the substrate binding. However, in this instance the mutation affected an early step in the C24 pathway involving conversion of 1,24,25(OH)3D to 24-oxo-1,25(OH)2D. It is evident, therefore, that the combination of molecular modeling and site-directed mutagenesis can be used to identify amino acid residues that are involved with the multiple catalytic activities associated with CYP24A1.
III. CELLULAR EXPRESSION AND REGULATION A. CYP24A1 Knockout and Transgenic Animals CYP24A1 cDNA clones have been isolated for rat [58], human [59], chicken [60], mouse [61], and pig [62].
91
CHAPTER 6 The 25-Hydroxyvitamin D 24-Hydroxylase
The human CYP24A1 gene is located on chromosome 20 [63], but there are no known diseases that can be attributed to this gene, although aberrant overexpression is potentially oncogenic as discussed in a later section. The exon–intron structures have been determined for the rat [64] and mouse [61] CYP24A1 genes, which are “typical genes” of length about 15 kb. The biological functions of CYP24A1 gene expression in vivo have been pursued in conventional CYP24A1 knockout mice and also in transgenic rats constitutively expressing CYP24A1. Knockout mice were generated using embryonic stem cell technology so that the heme-binding domain of the mouse CYP24A1 gene was deleted ensuring that no functional CYP24A1 protein was produced [65,66]. Homozygous null mice exhibited a high circulating level of 1,25(OH)2D in keeping with the lack of catabolism through the C23/24 oxidation pathway. There was abnormal bone histology characterized by excessive undermineralized bone matrix and this raised the possibility that a deficiency of the CYP24A1 oxidation product 24,25(OH)2D may be responsible. While 24,25(OH)2D has been considered an inactive metabolite, there is evidence for a biological function of 24,25(OH)2D in bone physiology, particularly in the development of growth plate chondrocytes [67,68]. However, the mouse phenotype was rescued when the CYP24A1 null mice were mated with vitamin D receptor (VDR) knockout mice [69], demonstrating that the bone abnormality was not due to a deficiency of 24,25(OH)2D but most likely to elevated 1,25(OH)2D, which requires VDR for its biological actions [70,71]. In a different approach, constitutive expression of CYP24A1 has been achieved in the tissues of transgenic rats with the introduction of a rat CYP24A1 cDNA sequence under the control of the viral MLV-LTR promoter [72,73]. It was anticipated that characteristics of this transgenic rat would resemble those observed in CYP27B1 knockout mice where the lack of production of 1,25(OH)2D results in hypocalcemia, hyperparathyroidism, growth retardation, and rickets [74,75]. Unexpectedly, after 8 weeks the transgenic rats developed albuminuria and hyperlipidemia with a significant reduction in circulating 25(OH)D and 24,25(OH)2D levels accompanied by a lowering of bone mineral density [72,73]. On the other hand, the plasma concentration of 1,25(OH)2D in the transgenic rats was the same as in wild type, most likely because of up-regulation of CYP27B1. Excess renal excretion of albumin is probably responsible for the reduction in plasma 25(OH)D due to its prevention of 25(OH)D reuptake by renal tubular cells [73]. Infusion of transgenic rats with 25(OH)D prevented the mineralization defect by making the circulating 1,25(OH)2D
more available [73]. It is not clear how constitutive expression of CYP24A1 alters renal function, resulting in albuminuria and marked reduction of plasma 25(OH)D. Overall, these studies are significant and suggest that 1,25(OH)2D is involved in functions other than regulation of vitamin D metabolism. Thus far, there have been no reports on the ablation or overexpression of CYP24A1 in selected tissues, which could shed light on tissue specific roles for CYP24A1.
B. Regulation of Enzyme Expression 1. KIDNEY
The most important physiological regulators of renal CYP24A1 expression in the adult are 1,25(OH)2D and PTH [76–78]. These hormones will therefore influence the production of renal 1,25(OH)2D and hence the level of circulating hormone. Renal CYP24A1 gene transcription is induced by 1,25(OH)2D in a VDR-dependent fashion. This action of 1,25(OH)2D occurs chiefly in the proximal convoluted tubules [79] and will assist in the maintenance of 1,25(OH)2D levels, in particular in the prevention of hypercalcemia. In marked contrast, the peptide hormone PTH functions through its cellular receptor to suppress renal CYP24A1 expression [50] when circulating levels of calcium are low [76–78]. This inhibition occurs through alteration of CYP24A1 mRNA stability and alterations in CYP24A1 gene transcription [62,80,81] (R Serda and JL Omdahl, unpublished). However, the PTH-mediated inhibition of CYP24A1 expression is not observed in the intestine that lacks PTH receptor [50]. PTH-differentially regulates CYP24A1 in proximal and distal convoluted tubular cells with PTH inhibiting 1,25(OH)2D-stimulated expression of CYP24A1 in proximal convoluted tubules, but potentiating the initial induction of CYP24A1 by 1,25(OH)2D in distal convoluted tubular cells [79]. It has been proposed that in distal tubules a coordinated action of PTH and 1,25(OH)2D stimulates calcium reabsorption, but when levels of 1,25(OH)2D are high, PTH stimulates the induction of CYP24A1 by 1,25(OH)2D, thereby lowering cellular 1,25(OH)2D and avoiding excessive calcium uptake by these tubules [79]. In addition to 1,25(OH)2D and PTH, the peptide hormone calcitonin [82–84] is observed to also contribute to regulation of renal CYP24A1 expression. Calcitonin is secreted by the thyroid parafollicular C cells in response to elevated serum calcium levels and is well known to have hypocalcemic actions that result from an inhibition of osteoclastic bone resorption and stimulation of renal calcium excretion [83,85]. Whether these actions contribute significantly to the
92
JOHN OMDAHL AND BRIAN MAY
maintenance of ambient calcium levels in human adults is uncertain [83]. Alternatively, calcitonin could contribute to calcium homeostasis through regulating 1,25(OH)2D levels. Renal CYP27B1 expression in response to calcitonin has been studied. For example, calcitonin induces CYP27B1 expression in kidney LLC-PK1 and MCT proximal tubule cells [86,87], but there is no effect on CYP27B1 levels in the proximal convoluted HKC-8 cells [88]. In distal parts of the nephron, calcitonin stimulates CYP27B1 under normocalcemic conditions [89]. However, the physiological rationale for any induction of CYP27B1 by calcitonin is unclear, since calcitonin is produced in response to high serum calcium. Recent studies have established for the first time that expression of the rat CYP24A1 promoter can be substantially induced by calcitonin in both transiently and stably transfected kidney HEK293 cells (X-P Gao, P Dwivedi, JL Omdahl, H Morris, and B May, submitted). These findings raise the possibility that such a stimulatory action by calcitonin could be important in vivo for modulating renal 1,25(OH)2D production. 2. INTESTINE
Intestinal cells do not express PTH or calcitonin receptors. Circulating 1,25(OH)2D stimulates intestinal calcium import, and the induction of intestinal CYP24A1 by the hormone is likely to be an important factor in determining the level of circulating secosteroid [76]. In this regard, induction of CYP24A1 in the intestine is more responsive to acute changes in 1,25(OH)2D compared with the kidney [90]. 3. BONE
An important role for CYP24A1 in bone development was suggested from the phenotype of CYP24A1 null mice that were found to exhibit abnormal intramembranous bone formation as discussed earlier. This phenotype may result from a local toxic level of 1,25(OH)2D due to the loss of CYP24A1. In earlier studies, Nishimura et al. [91] showed that 1,25(OH)2D induces expression of CYP24A1 in osteoblast cell lines and observed that there is a greater induction by 1,25(OH)2D in the less mature osteoblastic cell lines. This data has been confirmed by the quantification of CYP24A1 mRNA [10] in both rat and human osteoblast-like cell lines (P Anderson, H Morris, and B May, unpublished data). Nonstimulated levels of CYP24A1 mRNA are very low but can be increased dramatically up to 400,000-fold by 1,25(OH)2D. This 1,25(OH2)D–dependent stimulation of CYP24A1 expression appears blunted in fully differentiated, mature cells. Locally produced 1,25(OH)2D, acting in an autocrine/paracrine fashion, is now considered to
have an antiproliferative and prodifferentiation effect on many cells types, both normal and cancerous [92]. It is therefore an attractive hypothesis that 1,25(OH)2D may modulate CYP24A1 expression during osteoblast development so that low levels of 1,25(OH)2D are maintained during the proliferative phase but are increased during the mineralization phase. In support of this concept, the addition of 1,25(OH)2D to proliferative osteoblasts in culture inhibits proliferation but addition to more mature osteoblasts promotes expression of genes involved in differentiation and mineralization [93]. Osteoblasts also express PTH receptor, but there is no evidence that PTH lowers expression of 1,25(OH)2D-induced CYP24A1 as seen in the kidney [94]. In fact, in rat osteoblastic UMR-106 cells, PTH synergises with 1,25(OH)2D to initially enhance CYP24A1 mRNA and protein levels by a mechanism that involves PKA activity [94,95], although mRNA expression is lowered at later time intervals (R Serda and JL Omdahl, unpublished data). It is possible that the early synergistic increase in CYP24A1 in response to PTH and 1,25(OH)2D serves to ensure lowered cellular levels of 1,25(OH)2D commensurate with early osteoblast development. In a manner similar to PTH, the potent glucocorticoid dexamethasone markedly stimulates CYP24A1 mRNA expression and CYP24A1 activity in 1,25(OH)2Dtreated UMR-106 cells [96]. In vivo, glucorticoids markedly affect bone metabolism, albeit in a complex fashion, and an enhanced catabolism of 1,25(OH)2D in the presence of glucorticoid in osteoblasts could influence net bone resorption by counteracting the anabolic effect of 1,25(OH)2D in stimulating mineralization in these cells [96]. 4. MONOCYTES/MACROPHAGES
Immune cell types such as monocytes and macrophages express both CYP27B1 and CYP24A1 with the locally synthesized 1,25(OH)2D exerting its effects on the immune system through both autocrine and paracrine mechanisms [97]. The production of 1,25(OH)2D expression by monocytes or macrophages is markedly enhanced by exposure of cells to the cytokine INF-gamma [98]. This response mimics the autoimmune disease sarcoidosis where there is an abnormally high production of 1,25(OH)2D by diseaseactivated macrophages with associated hypercalcemia. INF-gamma-stimulated production of 1,25(OH)2D results both from up-regulation of CYP27B1 expression [99,100] and from a novel repression of CYP24A1 induction by 1,25(OH)2D that suppresses the ability of 1,25(OH)2D to control its own synthesis. The molecular aspects of this repression have been defined [101] and are discussed later.
93
CHAPTER 6 The 25-Hydroxyvitamin D 24-Hydroxylase
5. KERATINOCYTE
As mentioned, 1,25(OH)2D can influence the differentiation of many cells in culture [92] and keratinocytes are a well studied system in which locally synthesized 1,25(OH)2D acts in an autocrine/paracrine fashion to inhibit cell proliferation and to stimulate terminal keratinocyte differentiation [102,103]. As well as expressing CYP27B1 for local 1,25(OH)2D production, keratinocytes have the capacity to metabolise 1,25(OH)2D through induction of CYP24A1 [45]. At pharmacological levels, 1,25(OH)2D and its analogs can be used in the treatment of hyperproliferative diseases such as psoriasis, but their effectiveness is restricted by expression of CYP24A1, and this has led to a search for selective inhibitors of CYP24A1 [45]. It appears that a constitutive expression of CYP27B1 enzyme in keratinocytes synthesizes local 1,25(OH)2D in which this level of 1,25(OH)2D functions to finely tune CYP24A1. Interestingly, the level of CYP27B1 enzyme activity in keratinocytes markedly declines when CYP24A1 activity is induced by 1,25(OH)2D [45,104]. This inhibition of CYP27B1 enzyme activity is not due to a decrease in CYP27B1 mRNA but rather reflects an increased CYP24A1 directed catabolism of both substrate 25(OH)D and the product 1,25(OH)2D of CYP27B1 [45,104]. This finding that CYP27B1 activity is not directly regulated by 1,25(OH)2D differs from the situation in kidney, where excess 1,25(OH)2D appears to inhibit transcription of the CYP27B1 gene [105,106]. 6. CANCER CELLS
CYP24A1 may be an important player in contributing to the dysregulation of cell growth through lowering the local cellular 1,25(OH)2D concentration. In this regard, CYP24A1 has been suggested as a potential oncogene, since the gene was found coamplified with a known oncogene in human breast cancer tissue [107]. Elevated expression of CYP24A1 mRNA has been noted in highly malignant human colon carcinomas and may be due to amplification of the CYP24A1 gene [108]. A study of different prostate cancer cell lines showed that the growth inhibitory properties of 1,25(OH)2D correlated inversely with the level of CYP24A1 [109]. For example, the human prostate cancer DU-145 cell line expresses high levels of CYP24A1 in response to 1,25(OH)2D, and there is only minimal inhibition of cell proliferation. However, the antiproliferative activity of 1,25(OH)2D on these cells can be significantly restored through inhibition of CYP24A1 activity [110]. The high levels of induction of CYP24A1 observed in the prostate cancer cell lines [109] do not correlate with VDR amounts and the mechanism is not known. A significantly lower prevalence of prostate, mammary, and
colon cancers in Asia compared with Western industrialized countries has been linked to typical Asian diets with high amounts of soy products rich in the phytoestrogens, particularly isoflavonoids, suggesting that these compounds may prevent cancer [111]. Genestein, a major isoflavonoid, has been investigated recently. Administration of genestein or soy protein decreased expression of CYP24A1 mRNA in mouse colon [111] while genestein was also found to be a potent inhibitor of CYP24A1 mRNA expression in DU-145 cells [112].
IV. MOLECULAR ASPECTS In keeping with the widespread tissue distribution of CYP24A1, expression of the gene can be influenced by a broad range of regulatory agents. Molecular studies have focused on the control of CYP24A1 expression by 1,25(OH)2D, PTH, calcitonin, phorbol ester, dexamethasone, and gamma interferon, and their actions are summarized in Fig. 5. Control of CYP24A1 expression by these regulators appears to occur predominantly through modulation of gene transcription, although PTH is a notable exception in which it can alter CYP24A1 mRNA stability.
A. 1,25(OH)2D and the Rat CYP24A1 Promoter 1,25(OH)2D is the most potent inducer of CYP24A1 expression in kidney, osteoblast and other cell types. Investigations into the architecture of the rat CYP24A1 proximal promoter region have permitted a detailed understanding of the 1,25(OH)2D inductive process. Functional transcription factor binding sites have been identified that underlie basal and induced expression (see Figs. 6 and 7). Of central importance are two vitamin D response elements (VDREs) present on the non-coding strand and about 100 bp apart [113–118]. The proximal VDRE-1 at −136/−150 (5′-AGGTGAgtgAGGGCG-3′) and the distal VDRE-2 at −244/−258 (5′-GGTTCAgcgGGTGCG-3′) are DR3-type elements comprising two hexameric half-site direct repeats with a spacing of three nucleotides. Such a tandem arrangement of two VDREs remains unique to the CYP24A1 promoter and underlies transcriptional synergy between the VDREs and the high level of 1,25(OH)2D induction [117]. Electrophoretic mobility shift assays employing specific antibodies have shown that each VDRE binds VDR together with its partner RXR. Earlier experiments revealed than an oligonucleotide encompassing VDRE-1 bound the VDR/RXR heterodimer
94
JOHN OMDAHL AND BRIAN MAY
TPA (Dex) 1,25D
mVDR
calcitonin
CTR
KIDNEY
PKC
mVDR ERKs
PKA/PKC
P P AP1-like, Ets-1
P
RXR, Ets-1 other factors
PKA
Sp1, NF-Y
endonuclease
CYP24 mRNA RXR VDR
RXR VDR
VDRE-2
VDRE-1
PTHR
PTH
Induced CYP24 Gene
CYP24 protein
P mitochondria
Stat-1 VDR
NUCLEUS P Stat-1
IFN-gamma receptor
MONOCYTES/ MACROPHAGES
INF-gamma
FIGURE 5 Transcription regulation model in kidney and monocyte/macrophage cells. The transcription factors on the CYP24A1 promoter are shown for the regulatory agents phorbol ester (TPA), dexamethasone (dex), 1,25(OH)2D acting on either the nuclear VDR (nVDR) at the membrane or a novel membrane-inserted VDR (mVDRm), calcitonin acting on the calcitonin receptor (CTR), PTH acting on the PTH receptor (PTHR), and interferon gamma acting on the interferon gamma receptor. It is proposed that PTH action results in degradation of the CYP24A1 mRNA through the action of an endonuclease.
to a much lesser extent than an oligonucleotide encompassing VDRE-2 [117]. However, mutagenesis of each site within a –298 bp CYP24A1 promoter construct demonstrated that VDRE-1 contributes far more strongly to the 1,25(OH)2D inductive response in all cell types examined including COS-1, kidney HEK-293, UMR106, and ROS17/2.8 cell lines. This suggested that transcription factors bound nearby to VDRE-1 could enhance its activity.
Subsequently, a functional binding site for the ubiquitous transcription factor Ets-1 (Ets binding site, EBS; see Fig. 6) was identified downstream of VDRE-1 at −119/−128 on the noncoding strand and was required for maximal 1,25(OH)2D induction in COS-1 cells [1,119]. The EBS cooperates with VDRE-1 but not VDRE-2, presumably reflecting the close proximity of the VDRE-1 [1]. Direct protein–protein interaction between Ets-1 and VDR has been demonstrated by the
RNAP II machinery −298
RXR VDR
VDRE-2
RXR VDR
AP-1 like VDRE-1 EBS
GC
CCAAT
FIGURE 6 Cis-binding elements in CYP24A1 promoter. Functional binding sites for transcription factors within the first –298 bp of the rat CYP24A1 promoter. The arrows indicate that the sequences for VDRE-1, VDRE-2, and EBS are on the noncoding strand.
95
CHAPTER 6 The 25-Hydroxyvitamin D 24-Hydroxylase
two-hybrid assay [1]. As shown in Fig. 7B, Ets-1 bound at the EBS probably interacts with liganded VDR/RXR on VDRE-1 [1] and also with coactivator complexes [78] bound to liganded VDR/RXR. The EBS does not contribute to the low basal CYP24A1 promoter expression observed in transfected cells in the absence of added 1,25(OH)2D [119]. A repressor complex bound to unliganded VDR/RXR on the CYP24A1 promoter maintains this low expression in the absence of secosteroid [120]. The presence of this repressor complex may sterically prevent the binding of Ets-1 (Fig. 7A). When 1,25(OH)2D binds to VDR, it is proposed that the repressor complex dissociates and is replaced first by a coactivator complex with histone acetyltransferase activity that remodels
A
Basal
AP-1
Ets-1
X
X
chromatin [78,121] and then by a mediator complex, as described later. Since the EBS does not contribute to CYP24A1 promoter-directed basal expression detected in cells, it was clear that other transcription factor binding sites must be present. A GC box at −101/−114 (5′-CCCGCCCC-3′) immediately downstream of EBS and an inverted CCAAT box further downstream at –51/–62 (5′CATTGGC-3′) are functionally active (see Fig. 5; X-H Gao, P Dwivedi, and B May, unpublished data). Supershift analysis employing isolated nuclear extracts suggest that the ubiquitous factors Sp1 and NF-Y [122] bind to the GC and CCAAT box, respectively, with no binding detected of C/EBP alpha or beta to the CCAAT motif. Mutagenesis experiments established
HDAC? Repressor complex RXR VDR
VDRE-2
B
RNAP II machinery
Repressor complex RXR VDR
AP-1 like VDRE-1 EBS
weak GC
CCAAT
1,25D-induction
HAT
HAT CBP/p300
CBP/p300
p160
p160
RXR VDR
RXR VDR
VDRE-2
AP-1 like VDRE-1 EBS
RNAP II machinery
GC
CCAAT
C
RNAP II machinery
mediator complex RXR VDR
VDRE-2
RXR VDR
AP-1 like VDRE-1 EBS
Strong GC
CCAAT
FIGURE 7 Proposed binding of regulatory trans-proteins. (A) In the absence of 1,25(OH)2D, it is proposed that a repressor complex binds a VDR/RXR heterodimer on each VDRE. This complex may repress through modifying chromatin via histone deacetylase activity (HDAC) and/or preventing access to the AP-1 like binding protein and Ets-1. Low basal expression is driven by Sp1 and NF-Y bound at the GC and CCAAT box sites, respectively, through interaction with the RNA polymerase II machinery. (B) Following ligand binding to VDR, it is proposed that coactivators p160 and CBP/p300 are recruited and through histone acetyl transferase activity (HAT) modify chromatin and enhance binding of AP-1 like protein and Ets-1. Synergy may result from an interaction of all transcription factors with CBP except NF-Y and from interaction of the two coactivator complexes. (C) The two coactivator complexes are proposed to be replaced with a coactivator mediator complex that interacts with and activates RNA polymerase II and hence stimulates CYP24A1 gene transcription.
96
JOHN OMDAHL AND BRIAN MAY
(−) 1,25 (OH)2D + 1,25 (OH)2D UMR-106 cells 60x 10.00
36x
5.00
33x
38x 24x
6x
1.00 0.50
Mutated AP-1 like
Mutated EBS
Mutated GC + CCAAT
Mutated CCAAT
Mutated GC
WTpCYP24-298-luc
Mutated AP-1 like
Mutated EBS
Mutated GC + CCAAT
Mutated CCAAT
0
Mutated GC
0.25
WT pCYP24-298-luc
Relative Luciferase Activity
15.00
FIGURE 8
Mutational analysis of CYP24A1 promoter during induction of 1,25(OH)2D. Expression of transiently transfected wild type and mutant –298 bp CYP24A1 promoter constructs in UMR-106 cells. Cells were treated with 1,25(OH)2D at 10−7M. The relative luciferase activity represents the mean ±SD firefly luciferase/Renilla luciferase ratios of triplicate samples.
that the GC and CCAAT boxes predominantly are responsible for basal expression of the CYP24A1 promoter in the HEK-293 and UMR-106 cell lines examined. In UMR-106 cells, basal expression is substantially lowered when the CCAAT and GC boxes are mutated either individually or together (Fig. 8). A similar contribution from each site to basal expression is observed in stably transfected HEK-293T cells (Fig. 9). Sp1 and NF-Y could drive basal expression through interactions with the general transcription machinery (see Fig. 7A), notably with the general transcription factor TFIID bound at the TATA box of the CYP24A1 promoter. The GC and CCAAT box elements are important for induction of the CYP24A1 promoter by 1,25(OH)2D. In UMR-106 cells, when the GC and CCAAT box sites are both inactivated, 1,25(OH)2D-induced CYP24A1 promoter activity is reduced from about 60-fold to 24-fold (Fig. 9). The GC and CCAAT box sites are not responsive per se to 1,25(OH)2D, since mutagenesis of both VDREs in the CYP24A1 promoter abolishes induction by 1,25(OH)2D. In UMR-106 cells, mutagenesis of EBS has no effect on basal expression but lowers 1,25(OH)2D induction from 60-fold to 36-fold (Fig. 8). Another sequence further upstream of VDRE-1 at –163/−171, 5′-TGTCGGTCA-3′, has recently been characterized through site-directed mutagenesis of the
CYP24A1 promoter and shown to markedly stimulate 1,25(OH)2D induction, particularly in osteoblastic cell lines (P Dwivedi, J Kaplan, JL Omdahl, H Morris, and B May, in preparation). This site resembles an AP-1 site (consensus 5′-TGAG/CTCA-3′) and is referred to here as AP-1 like (as shown in Figs. 6 and 7). In UMR106 cells, inactivation of this sequence markedly lowers 1,25(OH)2D induction of the CYP24A1 promoter from 60-fold to six-fold (Fig. 8), and further analysis has established that the site cooperates with VDRE-1 and not VDRE-2. The AP-1 like sequence does not contribute to basal CYP24A1 promoter expression, perhaps because of the presence of the repressor complex bound on VDRE-1 (Figs. 8 and 7A). Within the AP-1 like motif, the sequence 5′-CGGTCA-3′ resembles a VDRE halfsite and has attracted attention previously when it was shown that mutagenesis of this sequence lowered 1,25(OH)2D induction [114,118]. Overall, the rat CYP24A1 proximal promoter (see Fig. 6) comprises a weak VDRE-1 encircled by compensatory transcription factors and a strong upstream VDRE-2 apparently devoid of nearby transcription factors. A model for the nuclear genomic action of 1,25(OH)2D that leads to stimulation of transcription of the CYP24A1 promoter can be proposed (Fig. 7). When 1,25(OH)2D binds to VDR located at each VDRE, the
97
CHAPTER 6 The 25-Hydroxyvitamin D 24-Hydroxylase
−calcitonin
50.00
+calcitonin
Luciferase Activity / ug Protein
14x
45.00
2.8x
5.00
1.2x
Mutated GC
Mutated CCAAT
WTpCYP24-298-luc
Mutated GC
Mutated CCAAT
WT pCYP24-298-luc
0
FIGURE 9 Mutational analysis of CYP24A1 promoter during induction by calcitonin. Expression of wild-type and mutant –298 bp CYP24A1 promoter constructs in stably transfected HEK-293T cells expressing the calcitonin receptor. Cells were treated overnight with calcitonin (10 nM). The fire fly luciferase activity was normalised to the amount of cellular protein and the values represent the mean ± SD of triplicate samples.
repressor complex is dissociated and a p160 coactivator is recruited on each VDRE, together with other coactivators such as CBP/p300 and SKIP/NCoA-62 [78,121], so that cooperative interactions between the coactivator complexes result in transcriptional synergy. At VDRE-1, it is envisaged that recruitment of a coactivator complex by liganded VDR/RXR is enhanced by the AP-1 like protein, by Ets-1, and also by Sp1. The coactivator complex through its intrinsic histone acetyl transferase activity could modify chromatin permitting access to factors and facilitating assembly of the transcription machinery. Finally, in a proposed two-step mechanism, the coactivator complex would be replaced by the multiprotein mediator complex DRIP (VDR interacting proteins) that interacts with and activates RNAPII [78]. There is evidence that the induction of CYP24A1 mRNA by 1,25(OH)2D can be blocked by the protein synthesis inhibitor cycloheximide, implicating the obligatory production of a regulatory protein [81,123]. This protein is apparently not VDR or RXR, suggesting that for induction to proceed optimally, synthesis of a transcription factor or a coactivator component or a protein that stabilizes CYP24A1 mRNA is required [81].
B. 1,25(OH)2D and Mitogen-Activated Protein Kinase Activities In addition to its classical genomic action in the nucleus, it is now clear that 1,25(OH)2D also functions at the plasma membrane to rapidly activate downstream signaling pathways critical for transcription of the CYP24A1 promoter [1,119]. Such a nongenomic action is also a feature of steroid hormones, which is currently receiving considerable attention [124]. The importance of phosphorylation events in the induction of the CYP24A1 promoter by 1,25(OH)2D was first suggested by studies in COS-1 cells cotransfected with VDR, where overexpression of H-Ras increased the inductive response and also potentiated the effect of exogenously expressed wild-type Ets-1 acting through the EBS [119]. This response agreed with the proposal that the functionality of Ets-1 was dependent on phosphorylation of threonine residue 38. A mutant form of Ets-1 where threonine residue 38 was mutated to alanine lowered the level of induction of the CYP24A1 promoter to that observed with mutated EBS [1]. Also, a dominant negative Ras mutant inhibited 1,25(OH)2D induction to a level greater than that observed following
98 EBS inactivation. This demonstrated that the action of Ras on the CYP24A1 promoter was likely to be channeled not only through Ets-1 but also through other components on the promoter. Since MAP kinase activities are downstream effectors of Ras, a detailed investigation was undertaken into the possible role of these kinases in the activation of the rat CYP24A1 promoter by 1,25(OH)2D in VDRtransfected COS-1 cells [1]. Rapid stimulation of ERK1/2 and ERK5 activities by 1,25(OH)2D occurred in these cells with maximal activities peaking at 5–10 min and inhibition of activity directed by a Ras dominant negative clone. These data raised the possibility that ERK1/2 and ERK5 could be involved in the regulation of the CYP24A1 promoter. Studies with specific pharmacological inhibitors and dominant negative mutants established that in 1,25(OH)2D treated COS-1 cells, inhibition of the ERK1/2 module markedly lowered induction of the CYP24A1 promoter while the p38 and JNK modules were not required [1]. Stimulation of ERK5 activity by 1,25(OH)2D was required for maximal induction in which the secosteroid’s action was effected through phosphorylation of Ets-1. Activated ERK1 phosphorylated RXRalpha at serine 260, and when COS-1 cells were transfected with the mutant RXRalphaS260A clone (in which serine was replaced by alanine) induction of the CYP24A1 promoter was inhibited. VDR was not phosphorylated by either kinase activity. These studies established a critical role in these cells for the Ras-ERK1/2-RXRalpha signaling pathway for induction of the CYP24A1 promoter with an enhancing contribution from the Ras-ERK5 pathway through phosphorylation of Ets-1. An analysis of the signaling pathways that contribute to 1,25(OH)2D induction of the CYP24A1 promoter using specific pharmacological inhibitors and dominant negative clones has revealed that the contribution of ERK modules and other signaling molecules is dependent on the cell type. For example, in HEK-293 cells 1,25(OH)2D induction is dependent on JNK activity and there is little activation of ERK1/2. The contribution to promoter induction of Src and PI3Kinase activities also is cell type dependent. PKC (see Section III,B,4) is important for 1,25(OH)2D induction in most cell types examined. Overall these studies show that the nongenomic signaling pathways activated by 1,25(OH)2D are vital for CYP24A1 promoter activity with PKC/Ras/ ERK modules of particular importance. The precise role of the signaling pathways and the downstream targets on the CYP24A1 promoter (i.e. AP-1-like sequence, VDR/RXR, Ets-1, Sp1 and CCAAT-box binding protein) in the different cell types will require an in-depth analysis. As discussed previously [1], it is interesting that whereas overexpressed Ras facilitates
JOHN OMDAHL AND BRIAN MAY
1,25(OH)2D induction of the CYP24A1 promoter activity in COS-1 cells through phosphorylation of serine 160 in RXR, in Ras-transformed keratinocytes 1,25(OH)2D-dependent phosphorylation of serine 160 in RXR attenuates 1,25(OH)2D transactivation [125]. The molecular basis for this difference in the action of phosphorylated RXR is not known. As discussed earlier, the inhibitory action of genestein on CYP24A1 mRNA expression supports the proposal that soy consumption may protect against cancer through elevating local 1,25(OH)2D production. Genestein is a well-known tyrosine kinase inhibitor and the observed inhibitory effect on CYP24A1 mRNA expression is likely to be at the nongenomic level.
C. 1,25(OH)2D and the Nongenomic Response The rapid activations of ERK1/2 and ERK5 and other signaling molecules by 1,25(OH)2D, as a prelude to CYP24A1 promoter stimulation, are consistent with the nongenomic actions reported for steroid hormones [124]. Studies of the estrogen receptor (ER) provide a model in which ER is recruited to the plasma membrane by interaction with caveolin proteins that serve as scaffolds in caveolae, discrete domains of the plasma membrane [126,127]. Signaling molecules including various G proteins, Src, PI3Kinase, and Raf are also recruited to these scaffolds so that the ER in response to ligand can rapidly modulate a variety of signaling pathways within a confined space. There is evidence for the ER interacting directly with Src, and PI3K and stimulating PKC and PKA activities through various G proteins. Studies on the mechanism underlying the nongenomic action of 1,25(OH)2D are less extensive than those of estrogen/ER pathway. There is evidence that nongenomic actions of 1,25(OH)2D include G-protein stimulation of adenylate cyclase, phospholipase C beta [77] and phospholipase C gamma, the last through the tyrosine kinases Src and PI3K [128]. By analogy with ER, it can be envisaged that the classical VDR is recruited to the plasma membrane where it activates signaling cascades in response to 1,25(OH)2D. Indeed, nongenomic activities are lost from osteoblast cultures derived from homozygous mice expressing a mutant VDR lacking the first zinc finger necessary for DNA binding [129]. Presumably conformational changes in the mutant VDR interfere with membrane localization or binding to the downstream signaling molecules (i.e., function as dominant-negative VDR). On the other hand, it is possible that a unique VDR may participate in nongenomic signaling. A novel progesterone receptor has been identified with seven putative transmembrane domains, thus resembling the classic G protein coupled
99
CHAPTER 6 The 25-Hydroxyvitamin D 24-Hydroxylase
receptors [130]. This raises the possibility that a novel membrane form of VDR, perhaps together with the classical VDR acting at the plasma membrane, contribute to nongenomic biological responses. In that regard, it is interesting that PKC activation by 1,25(OH)2D is still observed in osteoblast cultures isolated from VDR null mice where the second zinc finger of the DNA binding domain has been deleted [131,132]. Although these authors suggest that the nongenomic response is independent of the nuclear VDR activity, it is still possible that the absence of the second zinc finger alters downstream interactions necessary for PKC activation. Other evidence has shown that the ligand specificity for nongenomic activities in osteoblasts is distinct from that for nuclear gene activation [133].
lowered synergy from 7.1-fold to 3.3-fold and 1.1-fold, respectively. Hence both sites play a role, but the AP-1 like sequence plays a predominant role in TPA sensitivity. It can be proposed (see Fig. 5) that activated PKC in response to TPA further stimulates the activity of MAP kinase signaling pathways above that achieved by 1,25(OH)2D alone, leading to the phosphorylation and activation of AP-1 like protein and Ets-1 and overall enhancement of 1,25(OH)2D induction. Whether the nonresponsiveness of the CYP24A1 promoter to phorbol ester action in osteoblasts mirrors the lack of activation of MAP kinase pathways by TPA remains to be determined.
E. Glucocorticoid and Protein Kinase C D. Phorbol Ester and Protein Kinase C The stimulation of CYP24A1 expression by phorbol ester treatment and the involvement of PKC activity has been of interest for many years [76]. The synergistic action of phorbol ester and 1,25(OH)2D appears to be cell type specific. In mouse renal tubules, the phorbol ester 12-O-tetradecanoylphorbol-13-acetate (TPA), also referred to as PMA, increases CYP24A1 activity in a PKC-dependent fashion [134]. TPA also markedly increased the level of CYP24A1 mRNA in 1,25(OH)2Dtreated rat renal cells [15,135] and in rat intestinal epithelial cells [136,137]. However, synergy is not observed in osteoblast cells. TPA does not affect either CYP24A1 mRNA levels or CYP24A1 promoter activity in the presence or absence of 1,25(OH)2D [94,137,138] in rat osteoblastic UMR-106 cells, although some TPA stimulation of CYP24A1 mRNA in these cells in the presence of 1,25(OH)2D has been reported at earlier times after treatment [96]. Synergy also is not observed in rat osteoblastic ROS17/2.8 cells transfected with rat –298 bp CYP24A1 promoter constructs (P Dwivedi and B May, unpublished). The molecular mechanism by which TPA stimulates 1,25(OH)2D-induced CYP24A1 expression has attracted interest but has remained elusive [76]. Since it is well known that TPA can stimulate the activity of the transcription factor AP-1 and more recently Ets proteins [139], studies have been undertaken to determine the possible roles of the EBS and AP-1 like sequences in the CYP24A1 promoter in kidney 293T cells (P Dwivedi, B Nutschey, and B May, submitted). TPA alone gives minimal induction of the CYP24A1 promoter, but in the presence of 1,25(OH)2D there is marked synergy (about 7.1-fold) and this was localized to the first –186 bp of promoter sequence. Within this region, mutagenesis individually of the AP-1 like sequence and EBS
Dexamethasone markedly enhances 1,25(OH)2D induction of CYP24A1 mRNA in UMR-106 cells and to a lesser extent in renal LLC-PK1 cells with the effect of dexamethasone being blocked by a PKC inhibitor [96]. In this study, dexamethasone markedly increased c-fos mRNA and protein but only slightly induced the abundance of c-jun mRNA in UMR-106 cells. It was proposed that dexamethasone enhances CYP24A1 mRNA expression in osteoblast cells via activation of AP-1 in a process that requires PKC activity. It would be relevant to determine if dexamethasone acts through the AP-1 like sequence identified in the CYP24A1 promoter.
F. Parathyroid Hormone In recent studies employing opossum kidney (OK) cells that express both endogenous VDR and PTH receptors, Armbrecht et al. [140] showed that PTH (acting through cAMP) moderately increases expression of a reporter construct containing −580 bp of rat CYP24A1 promoter. The mechanism by which this cAMP-mediated increase in transcription occurs is not known but may involve increased VDR [138]. As expected, 1,25(OH)2D also stimulated promoter expression. The apparent paradox that PTH stimulates renal CYP24A1 promoter expression, yet substantially lowers CYP24A1 activity in vivo, has been resolved [62,80,81]. It has been demonstrated that PTH markedly destabilizes the CYP24A1 mRNA, reducing the half-life by fourfold as determined in AOK-B50 cells [81]. The precise mechanism underlying this cAMP-dependent degradation is not known but could involve PKA-directed phosphorylation of a CYP24A1 mRNA binding protein whose loss permits degradation by a specific endonuclease (Fig. 5). Interestingly, while the 3′ untranslated region is a common target for regulating mRNA stability, the studies
100 of Zierold et al. [80] showed that neither the 3′ nor the 5′ untranslated regions of the CYP24A1 mRNA contain the instability element, suggesting that such a site is located in the coding region. Alteration of mRNA stability is considered a rapid response mechanism for modulating gene expression. Hence in the hypocalcemic state, PTH production by the parathyroid gland will result in destabilization of renal CYP24A1 mRNA with a rapid increase in the production of 1,25(OH)2D by the kidney through decreased removal of 25(OH)D and 1,25(OH)2D. In contrast to this, it is accepted that the major effect of PTH on renal CYP27B1 expression is through increased promoter activity [140,141].
G. Calcitonin As discussed earlier, recent data have shown that calcitonin substantially increases expression of a rat CYP24A1 –298 bp promoter construct in both transiently and stably transfected kidney HEK-293 cells expressing the calcitonin receptor. This induction is strongly dependent on the GC and CCAAT box sites as demonstrated in stably transfected HEK-293 cells (Fig. 9) in which calcitonin induces the transfected promoter 14-fold. However, this induction is reduced to 2.8-fold and 1.2-fold when the CCAAT box and GC boxes are mutated, respectively. Pharmacological inhibitor studies indicated the involvement of both PKA and PKC activities in calcitonin stimulation but not basal expression. Although calcitonin is known to activate ERK1 and ERK2 in these cells [142], these ERK activities do not participate in calcitonin induction, since the ERK inhibitor PD98059 and the dominant negative clone ERK1K71R did not affect calcitonin induction. The precise roles of PKA and PKC remain to be determined, but it seems likely that they act through Sp1 and NF-Y (Fig. 5). On the basis of this new data, it is tempting to suggest that calcitonin in vivo contributes to the control of circulating calcium levels through regulating 1,25(OH)2D. Such an action of calcitonin could be important in the hypercalcemic state when a transient increase in serum calcium triggers the production of calcitonin. In this regard, rats consuming a high-calcium diet markedly increase renal CYP24A1 and VDR mRNA levels, although the hormone regulation is unclear [10].
H. Interferon (INF)-Gamma Increased production of 1,25(OH)2D by INF-gamma activated monocytes/macrophages is associated with a marked inhibition of CYP24A1 up-regulation
JOHN OMDAHL AND BRIAN MAY
by 1,25(OH)2D. A novel mechanism for this antagonistic action by INF-gamma has been elucidated [101]. Basal expression of a human 24-hydroxylase promoter construct encompassing the two VDREs was not affected by INF-gamma, but 1,25(OH)2D-stimulated expression was inhibited. These workers established a sequence of events where, following the binding of INF-gamma to its receptor on the cell surface, the transcription factor Stat1 is activated by phosphorylation and in its homodimeric phosphorylated form enters the nucleus and binds liganded VDR (Fig. 5). This interaction abrogates the binding of liganded VDR/RXR to the VDREs in the CYP24A1 promoter, thereby suppressing 1,25(OH)2D activation. It is interesting that INFgamma stimulated CYP27B1 gene expression is not inhibited by high levels of 1,25(OH)2D in these cells [99], although studies suggesting a negative regulation of CYP27B1 gene expression by 1,25(OH)2D have been reported in the kidney [143]. Perhaps a genomic-acting inhibitory mechanism exists only in kidney cells.
Acknowledgments The authors recognize the expert assistance of Stephanie Finlayson, Andrezej Pastuszyn, Rita Serda, and Andrew Annolara, who greatly facilitated the preparation and composition of this chapter. Also available were discussions with Howard Morris, Prem Dwivedi, and Satya Reddy on subject matter contained in this chapter.
References 1. Dwivedi PP, Hii CS, Ferrante A, et al. 2002 Role of MAP kinases in the 1,25-dihydroxyvitamin D3-induced transactivation of the rat cytochrome P450C24 (CYP24) promoter. Specific function for Erk1/Erk2 and Erk5. J Biol Chem 277:29643–29653. 2. Nemere I, Pietras RJ, Blackmore PF 2003 Membrane receptors for steroid hormones: Signal transduction and physiological significance. J Cell Biochem 88:438–445. 3. Farach-Carson MC, Nemere I 2003 Membrane receptors for vitamin D steroid hormones: potential new drug targets. Curr Drug Targets 4:67–76. 4. Pike JW 1991 Vitamin D3 receptors: structure and function in transcription. Annu Rev Nutr 11:189–216. 5. Hannah SS, Norman AW 1994 1 Alpha,25(OH)2 vitamin D3regulated expression of the eukaryotic genome. Nutr. Rev 52:376–382. 6. Matkovits T, Christakos S 1995 Variable in vivo regulation of rat vitamin D-dependent genes (osteopontin, Ca,Mg-adenosine triphosphatase, and 25-hydroxyvitamin D3 24-hydroxylase): implications for differing mechanisms of regulation and involvement of multiple factors. Endocrinology 136: 3971–3982.
101
CHAPTER 6 The 25-Hydroxyvitamin D 24-Hydroxylase
7. Armbrecht HJ, Boltz MA 1991 Expression of 25-hydroxyvitamin D 24-hydroxylase cytochrome P450 in kidney and intestine. Effect of 1,25-dihydroxyvitamin D and age. FEBS Lett 292:17–20. 8. Shinki T, Jin CH, Nishimura A, et al. 1992 Parathyroid hormone inhibits 25-hydroxyvitamin D3-24-hydroxylase mRNA expression stimulated by 1 alpha,25-dihydroxyvitamin D3 in rat kidney but not in intestine. J Biol Chem 267:13757–13762. 9. Zehnder D, Bland R, Williams MC, et al. 2001 Extrarenal expression of 25-hydroxyvitamin D(3)-1 alpha-hydroxylase. J Clin Endocrinol Metab 86:888–894. 10. Anderson PH, O’Loughlin PD, May BK, Morris HA 2003 Quantification of mRNA for the vitamin D metabolising enzymes CYP27B1 and CYP24 and VDR in kidney using real-time reverse transcriptase-polymerase chain reaction. J Mol Endocrinol (in press). 11. Pedersen JI, Hagenfeldt Y, Bjorkhem I 1998 Assay and properties of 25-hydroxyvitamin D3 23-hydroxylase. Evidence that 23,25-dihydroxyvitamin D3 is a major metabolite in 1,25-dihydroxyvitamin D3-treated or fasted guinea pigs. Biochem J 250:527–532. 12. Kumar R, Schaefer J, Grande JP, Roche PC 1994 Immunolocalization of calcitriol receptor, 24-hydroxylase cytochrome P-450, and calbindin D28k in human kidney. Am J Physiol 266:F477–F485. 13. Iwata K, Yamamoto A, Satoh S, et al. 1995 Quantitative immunoelectron microscopic analysis of the localization and induction of 25-hydroxyvitamin D3 24-hydroxylase in rat kidney. J Histochem Cytochem 43:255–262. 14. Henry HL, Dutta C, Cunningham N, et al. 1992 The cellular and molecular regulation of 1,25(OH)2D3 production. J Steroid Biochem Mol Biol 41:401–407. 15. Chen ML, Boltz MA, Armbrecht HJ 1993 Effects of 1,25dihydroxyvitamin D3 and phorbol ester on 25-hydroxyvitamin D3 24-hydroxylase cytochrome P450 messenger ribonucleic acid levels in primary cultures of rat renal cells. Endocrinology 132:1782–1788. 16. Matsumoto T, Kawanobe Y, Ogata E 1985 Regulation of 24,25-dihydroxyvitamin D-3 production by 1,25-dihydroxyvitamin D-3 and synthetic human parathyroid hormone fragment 1-34 in a cloned monkey kidney cell line (JTC-12). Biochim Biophys Acta 845:358–365. 17. Tomon M, Tenenhouse HS, Jones G 1990 Expression of 25-hydroxyvitamin D3-24-hydroxylase activity in Caco-2 cells. An in vitro model of intestinal vitamin D catabolism. Endocrinology 126:2868–2875. 18. Kane KF, Langman MJ, Williams GR 1996 Antiproliferative responses to two human colon cancer cell lines to vitamin D3 are differently modified by 9-cis-retinoic acid. Cancer Res 56:623–632. 19. Inaba M, Burgos-Trinidad M, DeLuca HF 1991 Characteristics of the 25-hydroxyvitamin D3- and 1,25-dihydroxyvitamin D3-24-hydroxylase(s) from HL-60 cells. Arch Biochem Biophys 284:257–263. 20. Skowronski RJ, Peehl DM, Feldman D 1995 Actions of vitamin D3, analogs on human prostate cancer cell lines: comparison with 1,25-dihydroxyvitamin D3. Endocrinology 136:20–26. 21. Makin G, Lohnes D, Byford V, Ray R, Jones G 1989 Target cell metabolism of 1,25-dihydroxyvitamin D3 to calcitroic acid. Evidence for a pathway in kidney and bone involving 24-oxidation. Biochem J 262:173–180. 22. Eil C, Liberman UA, Marx SJ 1986 The molecular basis for resistance to 1,25-dihydroxyvitamin D: studies in cells
23.
24. 25. 26.
27. 28. 29.
30.
31.
32.
33. 34.
35.
36.
37.
38. 39.
cultured from patients with hereditary hypocalcemic 1,25(OH)2D3-resistant rickets. Adv Exp Med Biol 196: 407–422. Morgan JW, Reddy GS, Uskokovic MR, et al. 1994 Functional block for 1 alpha,25-dihydroxyvitamin D3-mediated gene regulation in human B lymphocytes. J Biol Chem 269:13437–13443. Naveilhan P, Neveu I, Baudet C, et al. 1993 Expression of 25(OH) vitamin D3 24-hydroxylase gene in glial cells. NeuroReport 5:255–257. Chen ML, Heinrich G, Ohyama YI, et al. 1994 Expression of 25-hydroxyvitamin D3-24-hydroxylase mRNA in cultured human keratinocytes. Proc Soc Exp Biol Med 207:57–61. Holick MF, Schnoes HK, DeLuca HF, et al. 1972 Isolation and identification of 24,25-dihydroxycholecalciferol, a metabolite of vitamin D made in the kidney. Biochemistry 11:4251–4255. Holick MF, Kleiner-Bossaller A, Schnoes HK, et al. 1973 1,24,25-Trihydroxyvitamin D3. A metabolite of vitamin D3 effective on intestine. J Biol Chem 248:6691–6696. Beckman MJ, Tadikonda P, Werner E, et al. 1996 Human 25-hydroxyvitamin D3-24-hydroxylase, a multicatalytic enzyme. Biochemistry 35:8465–8472. Ohyama Y, Okuda K 1991 Isolation and characterization of a cytochrome P-450 from rat kidney mitochondria that catalyzes the 24-hydroxylation of 25-hydroxyvitamin D3. J Biol Chem 266:8690–8695. Lund EG, Guileyardo JM, Russell DW 1999 cDNA cloning of cholesterol 24-hydroxylase, a mediator of cholesterol homeostasis in the brain. Proc Natl Acad Sci USA 96: 7238–7243. Gray RW, Omdahl JL, Ghazarian JG, Horst RL 1990 Induction of 25-OH-vitamin D3 24- and 23-hydroxylase activities in partially purified renal extracts from pigs given exogenous 1,25-(OH)2D3. Steroids 55:395–398. Pikuleva IA, Cao C, Waterman MR 1999 An additional electrostatic interaction between adrenodoxin and P450c27 (CYP27A1) results in tighter binding than between adrenodoxin and p450scc (CYP11A1). J Biol Chem 274:2045–2052. Reddy GS, Tserng KY 1989 Calcitroic acid, end product of renal metabolism of 1,25-dihydroxyvitamin D3 through C-24 oxidation pathway. Biochemistry 28:1763–1769. Esvelt RP, Schnoes HK, DeLuca HF 1979 Isolation and characterization of 1 alpha-hydroxy-23-carboxytetranorvitamin D: a major metabolite of 1,25-dihydroxyvitamin D3. Biochemistry 18:3977–3983. Horst RL, Reinhardt TA, Ramberg CF, Koszewski NJ, Napoli JL 1986 24-Hydroxylation of 1,25-dihydroxyergocalciferol. An unambiguous deactivation process. J Biol Chem 261:9250–9256. Akiyoshi-Shibata M, Sakaki T, Ohyama Y, et al. 1994 Further oxidation of hydroxycalcidiol by calcidiol 24-hydroxylase. A study with the mature enzyme expressed in Escherichia coli. Eur J Biochem 224:335–343. Sakaki T, Sawada N, Nonaka Y, Ohyama Y, Inouye K 1999 Metabolic studies using recombinant Escherichia coli cells producing rat mitochondrial CYP24. CYP24 can convert 1α,25-dihydroxyvitamin D3 to calcitroic acid. Eur J Biochem 262:43–48. Sakaki T, Sawada N, Komai K, et al. 2000 Dual metabolic pathway of 25-hydroxyvitamin D3 catalyzed by human CYP24. Eur J Biochem 267:6158–6165. Peterson JA, Graham SE 1998 A close family resemblance: The importance of structure in understanding cytochromes P450. Structure 6:1079–1085.
102 40. Poulos TL, Finzel BC, Howard AJ 1987 High-resolution crystal structure of cytochrome P450cam. J Mol Biol 195:687–700. 41. Ravichandran KG, Boddupalli SS, Hasermann CA, Peterson JA, Deisenhofer J 1993 Crystal structure of hemoprotein domain of P450BM-3, a prototype for microsomal P450’s. Science 261:731–736. 42. Hasemann CA, Ravichandran KG, Peterson JA, Deisenhofer J 1994 Crystal structure and refinement of cytochrome P450terp at 2.3 A resolution. J Mol Biol 236:1169–1185. 43. Brown AJ, Finch JL, Lopez-Hilker S, et al. 1990 New active analogues of vitamin D with low calcemic activity. Kidney Int Suppl 29:S22–S27. 44. Wester MR, Johnson EF, Marques-Soares C, et al. 2003 Structure of a substrate complex of mammalian cytochrome P450 2C5 at 2.3 Å resolution: Evidence for multiple substrate binding modes. Biochemistry 42:6370–6379. 45. Schuster I, Egger H, Astecker N, et al. 2001 Selective inhibitors of CYP24: Mechanistic tools to explore vitamin D metabolism in human keratinocytes. Steroids 66:451–462. 46. Schuster I, Egger H, Bikle D, et al. 2001 Selective inhibition of vitamin D hydroxylases in human keratinocytes. Steroids 66:409–422. 47. Peehl DM, Seto E, Feldman D 2001 Rationale for combination ketoconazole/vitamin D treatment of prostate cancer. Urology 58:123–126. 48. Bayanov AA, Brunt AR 1999 Role of hypoxia and constitutionally different resistance to hypoxia/stress as the determiners of individual profile of cytochrome P450 isozyme activity. Gen Pharmacol 33:355–361. 49. Tenenhouse HS, Yip A, Jones G 1988 Increased renal catabolism of 1,25-dihydroxyvitamin D3 in murine X-linked hypophosphatemic rickets. J Clin Invest 81:461–465. 50. Shinki T, Jin CH, Nishimura A, et al. 1992 Parathyroid hormone inhibits 25-hydroxyvitamin D3-24-hydroxylase mRNA expression stimulated by 1 alpha,25-dihydroxyvitamin D3 in rat kidney but not in intestine. J Biol Chem 267:13757–13762. 51. Burgos-Trinidad M, DeLuca HF 1991 Kinetic properties of 25-hydroxyvitamin D- and 1,25-dihydroxyvitamin D24-hydroxylase from chick kidney. Biochim Biophys Acta 1078:226–230. 52. Taniguchi T, Eto TA, Shiotsuki H, et al. 2001 Newly established assay method for 25-hydroxyvitamin D3 24hydroxylase revealed much lower Km for 25-hydroxyvitamin D3 than for 1α,25-dihydroxyvitamin D3. J Bone Miner Res 16:57–62. 53. Pikuleva IA, Bjorkhem I, Waterman MR 1997 Expression, purification, and enzymatic properties of recombinant human cytochrome P450c27 (CYP27). Arch Biochem Biophys 343:123–130. 54. Pikuleva IA, Puchkaev A, Bjorkhem I 2001 Putative helix F contributes to regioselectivity of hydroxylation in mitochondrial cytochrome P450 27A1. Biochemistry 40: 7621–7629. 55. Schenkman JB, Remmer H, Estabrook RW 1967 Spectral studies of drug interaction with hepatic microsomal cytochrome. Mol Pharmacol 3:113–123. 56. Horst RL, Omdahl JA, Reddy S 2003 Rat cytochrome P450C24 (CYP24) does not metabolize 1,25-dihydroxyvitamin D2 to calcitroic acid. J Cell Biochem 88:282–285. 57. Omdahl JL, Bobrovnikova EV, Annalora A, Chen P, Serda R 2003 Expression, structure–function, and molecular modeling of vitamin D P450s. J Cell Biochem 88:356–362. 58. Ohyama Y, Noshiro M, Okuda K 1991 Cloning and expression of cDNA encoding 25-hydroxyvitamin D3 24-hydroxylase. FEBS Lett 278:195–198.
JOHN OMDAHL AND BRIAN MAY
59. Chen KS, Prahl JM, DeLuca HF 1993 Isolation and expression of human 1,25-dihydroxyvitamin D3 24-hydroxylase cDNA. Proc Natl Acad Sci USA 90:4543–4547. 60. Jehan F, Ismail R, Hanson K, DeLuca HF 1998 Cloning and expression of the chicken 25-hydroxyvitamin D3 24-hydroxylase cDNA. Biochim Biophys Acta 1395:259–265. 61. Akeno N, Saikatsu S, Kawane T, Horiuchi N 1997 Mouse vitamin D-24-hydroxylase: molecular cloning, tissue distribution, and transcriptional regulation by 1α,25-dihydroxyvitamin D3. Endocrinology 138:2233–2240. 62. Zierold C, Reinholz GG, Mings JA, Prahl JM, DeLuca HF 2001 Regulation of the porcine 1,25-dihydroxyvitamin D3-24-hydroxylase (CYP24) by 1,25-dihydroxyvitamin D3 and parathyroid hormone in AOK-B50 cells. Arch Biochem Biophys 381:323–327. 63. Hahn CN, Baker E, Laslo P, et al. 1993 Localization of the human vitamin D 24-hydroxylase gene (CYP24) to chromosome 20q13.2—>q13.3. Cytogenet Cell Genet 62: 192–193. 64. Ohyama Y, Noshiro M, Eggertsen G, et al. 1993 Structural characterization of the gene encoding rat 25-hydroxyvitamin D3 24-hydroxylase. Biochemistry 32:76–82. 65. St Arnaud R, Arabian A, Glorieux FH 1996 Abnormal bone development in mice deficient for the vitamin D 24-hydroxylase gene. J Bone Miner Res 11:S126. 66. St Arnaud R 1999 Targeted inactivation of vitamin D hydroxylases in mice. Bone 25:127–129. 67. Boyan BD, Sylvia VL, Dean DD, Schwartz Z 2001 24, 25-(OH)2D3 regulates cartilage and bone via autocrine and endocrine mechanisms. Steroids 66:363–374. 68. Schwartz Z, Ehland H, Sylvia VL, et al. 2002 1α,25-dihydroxyvitamin D3 and 24R,25-dihydroxyvitamin D3 modulate growth plate chondrocyte physiology via protein kinase Cdependent phosphorylation of extracellular signal-regulated kinase 1/2 mitogen-activated protein kinase. Endocrinology 143:2775–2786. 69. St Arnaud R, Arabian A, Travers R, et al. 2000 Deficient mineralization of intramembranous bone in vitamin D-24hydroxylase-ablated mice is due to elevated 1,25-dihydroxyvitamin D and not to the absence of 24,25-dihydroxyvitamin D. Endocrinology 141:2658–2666. 70. Haussler MR, Whitfield GK, Haussler CA, et al. 1998 The nuclear vitamin D receptor: biological and molecular regulatory properties revealed. J Bone Miner Res 13:325–349. 71. Issa LL, Leong GM, Eisman JA 1998 Molecular mechanism of vitamin D receptor action. Inflamm Res 47:451–475. 72. Kasuga H, Hosogane N, Matsuoka K, et al. 2002 Characterization of transgenic rats constitutively expressing vitamin D-24-hydroxylase gene. Biochem Biophys Res Commun 297:1332–1338. 73. Hosogane N, Shinki T, Kasuga H, et al. 2003 Mechanisms for the reduction of 24,25-dihydroxyvitamin D3 levels and bone mass in 24-hydroxylase transgenic rats. FASEB J 17:737–739. 74. Panda DK, Miao D, Tremblay ML, et al. 2001 Targeted ablation of the 25-hydroxyvitamin D 1α-hydroxylase enzyme: Evidence for skeletal, reproductive, and immune dysfunction. Proc Natl Acad Sci USA 98:7498–7503. 75. Dardenne O, Prud’homme J, Arabian A, Glorieux FH, St Arnaud R 2001 Targeted inactivation of the 25-hydroxyvitamin D(3)-1(α)-hydroxylase gene (CYP27B1) creates an animal model of pseudovitamin D-deficiency rickets. Endocrinology 142:3135–3141. 76. Omdahl JL, May BK 1997 The 25-hydroxyvitamin D 24-hydroxylase. In: Feldman D, Glorieux F, Pike J (eds) Vitamin D. Academic Press, San Diego pp. 69–85.
103
CHAPTER 6 The 25-Hydroxyvitamin D 24-Hydroxylase
77. Omdahl JL, Morris HA, May BK 2002 Hydroxylase enzymes of the vitamin D pathway: Expression, function, and regulation. Annu Rev Nutr 22:139–166. 78. Sutton AL, MacDonald PN 2003 Vitamin D: More than a “bone-a-fide” hormone. Mol Endocrinol 17:777–791. 79. Yang W, Friedman PA, Kumar R, et al. 1999 Expression of 25(OH)D3 24-hydroxylase in distal nephron: coordinate regulation by 1,25(OH)2D3 and cAMP or PTH. Am J Physiol 276:E793–E805. 80. Zierold C, Mings JA, DeLuca HF 2001 Parathyroid hormone regulates 25-hydroxyvitamin D3-24-hydroxylase mRNA by altering its stability. Proc Natl Acad Sci USA 98: 13572–13576. 81. Zierold C, Mings JA, DeLuca HF 2003 Regulation of 25hydroxyvitamin D3-24-hydroxylase mRNA by 1,25-dihydroxyvitamin D3 and parathyroid hormone. J Cell Biochem 88:234–237. 82. Pondel M 2000 Calcitonin and calcitonin receptors: bone and beyond. Int J Exp Pathol 81:405–422. 83. Brown EM, MacLeod RJ 2001 Extracellular calcium sensing and extracellular calcium signaling. Physiol Rev 81:239–297. 84. Inzerillo AM, Zaidi M, Huang CL 2002 Calcitonin: The other thyroid hormone. Thyroid 12:791–798. 85. Jones G, Strugnell SA, DeLuca HF 1998 Current understanding of the molecular actions of vitamin D. Physiol Rev 78:1193–1231. 86. Yoshida N, Yoshida T, Nakamura A, et al. 1999 Calcitonin induces 25-hydroxyvitamin D3 1α-hydroxylase mRNA expression via protein kinase C pathway in LLC-PK1 cells. J Am Soc Nephrol 10:2474–2479. 87. Murayama A, Takeyama K, Kitanaka S, et al. 1998 The promoter of the human 25-hydroxyvitamin D3 1α-hydroxylase gene confers positive and negative responsiveness to PTH, calcitonin, and 1α,25(OH)2D3. Biochem Biophys Res Commun 249:11–16. 88. Bland R, Zehnder D, Hughes SV, et al. 2001 Regulation of vitamin D-1α-hydroxylase in a human cortical collecting duct cell line. Kidney Int 60:1277–1286. 89. Shinki T, Ueno Y, DeLuca HF, Suda T 1999 Calcitonin is a major regulator for the expression of renal 25-hydroxyvitamin D3-1α-hydroxylase gene in normocalcemic rats. Proc Natl Acad Sci USA 96:8253–8258. 90. Henry HL 2001 The 25(OH)D3/1α,25(OH)2D3-24(R)-hydroxylase: A catabolic or biosynthetic enzyme? Steroids 66:391–398. 91. Nishimura A, Shinki T, Jin CH, et al. 1994 Regulation of messenger ribonucleic acid expression of 1α,25-dihydroxyvitamin D3-24-hydroxylase in rat osteoblasts. Endocrinology 134:1794–1799. 92. Ylikomi T, Laaksi I, Lou YR, et al. 2002 Antiproliferative action of vitamin D. Vitam Horm 64:357–406. 93. Gurlek A, Kumar R 2001 Regulation of osteoblast growth by interactions between transforming growth factor-beta and 1α,25-dihydroxyvitamin D3. Crit Rev Eukaryot Gene Expr 11:299–317. 94. Armbrecht HJ, Hodam TL 1994 Parathyroid hormone and 1,25-dihydroxyvitamin D synergistically induce the 1,25dihydroxyvitamin D-24-hydroxylase in rat UMR-106 osteoblast-like cells. Biochem Biophys Res Commun 205:674–679. 95. Armbrecht HJ, Hodam TL, Boltz MA, et al. 1998 Induction of the vitamin D 24-hydroxylase (CYP24) by 1,25-dihydroxyvitamin D3 is regulated by parathyroid hormone in UMR106 osteoblastic cells [see comments]. Endocrinology 139: 3375–3381. 96. Kurahashi I, Matsunuma A, Kawane T, Abe M, Horiuchi N 2002 Dexamethasone enhances vitamin D-24-hydroxylase
97. 98.
99. 100.
101.
102. 103. 104. 105.
106.
107. 108.
109.
110.
111.
112.
expression in osteoblastic (UMR-106) and renal (LLC-PK1) cells treated with 1α,25-dihydroxyvitamin D3. Endocrine 17: 109–118. Griffin MD, Xing N, Kumar R 2003 Vitamin D and its analogs as regulators of immune activation and antigen presentation. Annu Rev Nutr 23:117–145. Dusso AS, Kamimura S, Gallieni M, et al. 1997 γ-Interferoninduced resistance to 1,25-(OH)2 D3 in human monocytes and macrophages: A mechanism for the hypercalcemia of various granulomatoses. J Clin Endocrinol Metab 82:2222–2232. Monkawa T, Yoshida T, Hayashi M, Saruta T 2000 Identification of 25-hydroxyvitamin D3 1α-hydroxylase gene expression in macrophages. Kidney Int 58:559–568. Overbergh L, Decallonne B, Valckx D, et al. 2000 Identification and immune regulation of 25-hydroxyvitamin D-1-α-hydroxylase in murine macrophages. Clin Exp Immunol 120:139–146. Vidal M, Ramana CV, Dusso AS 2002 Stat1-vitamin D receptor interactions antagonize 1,25-dihydroxyvitamin D transcriptional activity and enhance stat1-mediated transcription. Mol Cell Biol 22:2777–2787. Bikle DD, Ng D, Tu C, Oda Y, Xie Z 2001 Calcium- and vitamin D-regulated keratinocyte differentiation. Mol Cell Endocrinol 177:161–171. Bikle DD, Tu CL, Xie Z, Oda Y 2003 Vitamin D regulated keratinocyte differentiation: Role of coactivators. J Cell Biochem 88:290–295. Xie Z, Munson SJ, Huang N, et al. 2002 The mechanism of 1,25-dihydroxyvitamin D3 autoregulation in keratinocytes. J Biol Chem 277:36987–36990. Murayama A, Takeyama K, Kitanaka S, et al. 1999 Positive and negative regulations of the renal 25-hydroxyvitamin D3 1α-hydroxylase gene by parathyroid hormone, calcitonin, and 1α,25(OH)2D3 in intact animals. Endocrinology 140:2224–2231. Murayama A, Kim M, Unno K, Takeyama K, Kato S 2001 A novel mechanism of the vitamin D dependent transcriptional repression through the human 25-hydroxyvitamin D3 1α-hydroxylase in nVDRE. J Bone Miner Res 16:S229. Albertson DG, Ylstra B, Segraves R, et al. 2000 Quantitative mapping of amplicon structure by array CGH identifies CYP24 as a candidate oncogene. Nat Genet 25:144–146. Bareis P, Bises G, Bischof MG, Cross HS, Peterlik M 2001 25-hydroxy-vitamin D metabolism in human colon cancer cells during tumor progression. Biochem Biophys Res Commun 285:1012–1017. Miller GJ, Stapleton GE, Hedlund TE, Moffat KA 1995 Vitamin D receptor expression, 24-hydroxylase activity, and inhibition of growth by 1α,25-dihydroxyvitamin D3 in seven human prostatic carcinoma cell lines. Clin Cancer Res 1:997–1003. Ly LH, Zhao XY, Holloway L, Feldman D 1999 Liarozole acts synergistically with 1α,25-dihydroxyvitamin D3 to inhibit growth of DU 145 human prostate cancer cells by blocking 24-hydroxylase activity. Endocrinology 140: 2071–2076. Kallay E, Adlercreutz H, Farhan H, et al. 2002 Phytoestrogens regulate vitamin D metabolism in the mouse colon: Relevance for colon tumor prevention and therapy. J Nutr 132: 3490S–3493S. Farhan H, Wahala K, Adlercreutz H, Cross HS 2002 Isoflavonoids inhibit catabolism of vitamin D in prostate cancer cells. J Chromatogr B Anal Technol Biomed Life Sci 777:261–268.
104 113. Ohyama Y, Ozono K, Uchida M, et al. 1994 Identification of a vitamin D-responsive element in the 5′-flanking region of the rat 25-hydroxyvitamin D3 24-hydroxylase gene. J Biol Chem 269:10545–10550. 114. Hahn CN, Kerry DM, Omdahl JL, May BK 1994 Identification of a vitamin D responsive element in the promoter of the rat cytochrome P450(24) gene. Nucleic Acids Res 22:2410–2416. 115. Zierold C, Darwish HM, DeLuca HF 1994 Identification of a vitamin D-response element in the rat calcidiol (25-hydroxyvitamin D3) 24-hydroxylase gene. Proc Natl Acad Sci USA 91:900–902. 116. Zierold C, Darwish HM, DeLuca HF 1995 Two vitamin D response elements function in the rat 1,25-dihydroxyvitamin D 24-hydroxylase promoter. J Biol Chem 270:1675–1678. 117. Kerry DM, Dwivedi PP, Hahn CN, et al. 1996 Transcriptional synergism between vitamin D-responsive elements in the rat 25-hydroxyvitamin D3 24-hydroxylase (CYP24) promoter. J Biol Chem 271:29715–29721. 118. Ohyama Y, Ozono K, Uchida M, et al. 1996 Functional assessment of two vitamin D-responsive elements in the rat 25-hydroxyvitamin D3 24-hydroxylase gene. J Biol Chem 271:30381–30385. 119. Dwivedi PP, Omdahl JL, Kola I, Hume DA, May BK 2000 Regulation of rat cytochrome P450C24 (CYP24) gene expression. Evidence for functional cooperation of rasactivated Ets transcription factors with the vitamin D receptor in 1,25-dihydroxyvitamin D3-mediated induction. J Biol Chem 275:47–55. 120. Dwivedi PP, Muscat GE, Bailey PJ, Omdahl JL, May BK 1998 Repression of basal transcription by vitamin D receptor: evidence for interaction of unliganded vitamin D receptor with two receptor interaction domains in RIP13∆1. J Mol Endocrinol. 20:327–335. 121. Barry JB, Leong GM, Church WB, et al. 2003 Interactions of SKIP/NCoA-62, TFIIB, and retinoid X receptor with vitamin D receptor helix H10 residues. J Biol Chem 278:8224–8228. 122. Mantovani R 1999 The molecular biology of the CCAATbinding factor NF-Y. Gene 239:15–27. 123. Zierold C, Mings JA, Prahl JM, Reinholz GG, DeLuca HF 2002 Protein synthesis is required for optimal induction of 25-hydroxyvitamin D3-24-hydroxylase, osteocalcin, and osteopontin mRNA by 1,25-dihydroxyvitamin D3. Arch Biochem Biophys 404:18–24. 124. Falkenstein E, Tillmann HC, Christ M, Feuring M, Wehling M 2000 Multiple actions of steroid hormones—a focus on rapid, nongenomic effects. Pharmacol Rev 52:513–556. 125. Solomon C, White JH, Kremer R 1999 Mitogen-activated protein kinase inhibits 1,25-dihydroxyvitamin D3-dependent signal transduction by phosphorylating human retinoid X receptor α. J Clin Invest 103:1729–1735. 126. Razandi M, Alton G, Pedram A, et al. 2003 Identification of a structural determinant necessary for the localization and function of estrogen receptor α at the plasma membrane. Mol Cell Biol 23:1633–1646. 127. Razandi M, Oh P, Pedram A, Schnitzer J, Levin ER 2002 ERs associate with and regulate the production of caveolin: implications for signaling and cellular actions. Mol Endocrinol 16:100–115. 128. Buitrago C, Gonzalez PV, De Boland AR 2002 Nongenomic action of 1α,25(OH)2-vitamin D3. Activation of muscle cell PLC gamma through the tyrosine kinase c-Src and PtdIns 3-kinase. Eur J BioChem 269:2506–2515.
JOHN OMDAHL AND BRIAN MAY
129. Erben RG, Soegiarto DW, Weber K, et al. 2002 Deletion of deoxyribonucleic acid binding domain of the vitamin D receptor abrogates genomic and nongenomic functions of vitamin D. Mol Endocrinol. 16:1524–1537. 130. Hammes SR 2003 The further redefining of steroid-mediated signaling. Proc Natl Acad Sci USA 100:2168–2170. 131. Wali RK, Kong J, Demay MB, et al. 2001 The Vitamin D receptor is not required for the rapid activation of PKC and rise in intracellular calcium induced by 1,25-dihydroxyvitamin D3 in mouse osteoblasts. J Bone Miner Res 16:S229. 132. Wali RK, Kong J, Sitrin MD, Bissonnette M, Li YC 2003 Vitamin D receptor is not required for the rapid actions of 1,25-dihydroxyvitamin D3 to increase intracellular calcium and activate protein kinase C in mouse osteoblasts. J Cell BioChem 88:794–801. 133. Norman AW, Henry HL, Bishop JE, et al. 2001 Different shapes of the steroid hormone 1α,25(OH)2-vitamin D3 act as agonists for two different receptors in the vitamin D endocrine system to mediate genomic and rapid responses. Steroids 66:147–158. 134. Mandla S, Boneh A, Tenenhouse HS 1990 Evidence for protein kinase C involvement in the regulation of renal 25-hydroxyvitamin D3–24-hydroxylase. Endocrinology 127: 2639–2647. 135. Armbrecht HJ, Chen ML, Hodam TL, Boltz MA 1997 Induction of 24-hydroxylase cytochrome P450 mRNA by 1,25-dihydroxyvitamin D and phorbol esters in normal rat kidney (NRK-52E) cells. J Endocrinol 153:199–205. 136. Koyama H, Inaba M, Nishizawa Y, Ohno S, Morii H 1994 Protein kinase C is involved in 24-hydroxylase gene expression induced by 1,25(OH)2D3 in rat intestinal epithelial cells. J Cell BioChem 55:230–240. 137. Armbrecht HJ, Boltz MA, Hodam TL, Kumar VB 2001 Differential responsiveness of intestinal epithelial cells to 1,25-dihydroxyvitamin D3—role of protein kinase C. J Endocrinol 169:145–151. 138. Yang W, Hyllner SJ, Christakos S 2001 Interrelationship between signal transduction pathways and 1,25(OH)2D3 in UMR106 osteoblastic cells. Am J Physiol Endocrinol Metab 281:E162-E170. 139. Reddy SP, Vuong H, Adiseshaiah P 2003 Interplay between proximal and distal promoter elements is required for squamous differentiation marker induction in the bronchial epithelium: role for ESE-1, Sp1, and AP-1 proteins. J Biol Chem 278:21378–21387. 140. Armbrecht HJ, Hodam TL, Boltz MA 2003 Hormonal regulation of 25-hydroxyvitamin D3-1α-hydroxylase and 24-hydroxylase gene transcription in opossum kidney cells. Arch Biochem Biophys 409:298–304. 141. Gao XH, Dwivedi PP, Choe S, et al. 2002 Basal and parathyroid hormone induced expression of the human 25-hydroxyvitamin D 1α-hydroxylase gene promoter in kidney AOK-B50 cells: role of Sp1, Ets and CCAAT box protein binding sites. Int J Biochem Cell Biol 34: 921–930. 142. Raggatt LJ, Evdokiou A, Findlay DM 2000 Sustained activation of Erk1/2 MAPK and cell growth suppression by the insert-negative, but not the insert-positive isoform of the human calcitonin receptor. J Endocrinol 167:93–105. 143. Takeyama K, Kitanaka S, Sato T, et al. 1997 25-Hydroxyvitamin D3 1α-hydroxylase and vitamin D synthesis. Science 277:1827–1830.
CHAPTER 7
Mutant Mouse Models of Vitamin D Metabolic Enzymes RENÉ ST-ARNAUD
Genetics Unit, Shriners Hospital for Children, and McGill University, Montreal (Quebec), Canada
I. Introduction II. Hepatic 25-Hydroxylation and cyp27A1 Knockout III. 24-Hydroxylation and cyp24A1 Knockout
I. INTRODUCTION A. Note on Cytochrome P450 Nomenclature Since 1987, efforts have been made to encourage scientists worldwide to avoid “home-made” designations for naming cytochrome P450s in order to avoid confusing the nomenclature system and the scientific literature. This chapter will follow the guidelines provided in the relevant publications [1,2] and the recently updated names listed on the Cytochrome P450 Homepage Web site (http://drnelson.utmem.edu/mouse.master.table.html). Human genes will be italicized in capital letters, while mouse genes will be in lower case italics. Proteins are referred to in capital, nonitalicized characters (Table I).
B. Vitamin D Metabolism and Mutant Mouse Models of Vitamin D Metabolic Enzymes Vitamin D, produced endogenously in the skin upon exposure to ultraviolet light (sunlight) or ingested in the diet must be metabolized twice to be activated and function as a key regulator of mineral ion homeostasis. Vitamin D, bound to the vitamin D binding protein, DBP, is transported to the liver where vitamin D 25-hydroxylases (CYP27A1 and/or CYP2R1) add a hydroxyl group on carbon 25 to produce 25-hydroxyvitamin D [25(OH)D] (see Chapter 4). The 25(OH)D metabolite also circulates in the bloodstream bound to DBP (see Chapter 8). It must be further hydroxylated in the kidney to gain hormonal bioactivity. Hydroxylation at position 1α by the enzyme 25-hydroxyvitamin D1α-hydroxylase (CYP27B1) converts 25(OH)D to 1α,25-dihydroxyvitamin D [1,25(OH)2D], the active, hormonal form of vitamin D that regulates mineral homeostasis, skeletal homeostasis, and cellular differentiation (see Chapter 5). Among several target genes, the 1,25(OH)2D hormone induces in target cells the VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
IV. 1α-Hydroxylation and cyp27B1 Knockout V. Summary and Perspectives References
expression of the gene encoding the key effector of its catabolic breakdown: 25-hydroxyvitamin D-24-hydroxylase (CYP24A1). This ensures attenuation of the 1,25(OH)2D biological signal inside target cells and helps regulate vitamin D homeostasis. All vitamin D hydroxylases characterized to date belong to the superfamily of cytochrome P450 enzymes. They are heme-containing, mixed-function oxidases that use molecular oxygen as a terminal electron acceptor. They require the accessory electron transfer proteins, ferredoxin and ferredoxin reductase, to accept reducing equivalents from nicotinamide adenine dinucleotide phosphate (NADPH) and stereospecifically hydroxylate vitamin D metabolites. Molecular genetics–based research has allowed dramatic progress in our understanding of vitamin D metabolism and function. Several mice strains, deficient for the vitamin D metabolic enzymes CYP27A1, CYP27B1, and CYP24A1, have been engineered. This chapter will review the phenotype of the mutant strains, highlight the key findings derived from the analysis of the strains, and discuss the future avenues for research.
II. HEPATIC 25-HYDROXYLATION AND cyp27A1 KNOCKOUT Until recently, only one cytochrome P450 molecule had been cloned and shown to hydroxylate vitamin D at position 25. This was the bifunctional CYP27A1 sterol 27-hydroxylase that derives its name for its ability to both 27-hydroxylate the side chains of cholesterolderived intermediates involved in bile acid biosynthesis and 25-hydroxylate vitamin D [3]. However, several pieces of experimental evidence argued for the existence of a second, physiologically relevant vitamin D 25-hydroxylase enzyme (reviewed in [4]). First, mice with a disrupted cyp27A1 gene have reduced bile acid Copyright © 2005, Elsevier, Inc. All rights reserved.
106
RENÉ ST-ARNAUD
TABLE I Cytochrome P450 Nomenclature Used Common name
Human gene symbol
Mouse gene symbol
Protein
CYP27A1 CYP2R1 CYP27B1 CYP24A1
cyp27A1 cyp2R1 cyp27B1 cyp24A1
CYP27A1 CYP2R1 CYP27B1 CYP24A1
Vitamin D-25-hydroxylase (mitochondrial) Vitamin D-25-hydroxylase (microsomal) 25-Hydroxyvitamin D-1α-hydroxylase 25-Hydroxyvitamin D-24-hydroxylase
synthesis, but normal vitamin D metabolite levels [5]. Second, patients with the inherited disease cerebrotendinous xanthomatosis, caused by mutations in CYP27A1 [6,7], exhibit normal vitamin D metabolism and no obvious 25(OH)D or 1,25(OH)2D deficiency [8]. Third, early studies using perfused rat liver revealed kinetics of vitamin D metabolism that supported two 25-hydroxylase activities: a high-affinity, low-capacity enzyme and a low-affinity, high-capacity form (now presumed to be CYP27A1) [9]. The subcellular localizations of these two enzymatic activities are thought to differ, CYP27A1 being a mitochondrial enzyme [10,11], while the high-affinity, low-capacity enzyme is associated with the endoplasmic reticulum (microsomes) [12]. Next, the promoter of CYP27A1 is regulated by bile acids, but not by vitamin D metabolites [13], a finding that is not consistent with the weak but demonstrated regulation of the hepatic 25-hydroxylase observed following vitamin D intake in previously vitamin D–depleted animals [14]. Last, cells transfected with an expression vector for CYP27A1 are not able to 25hydroxylate vitamin D2 [15]. The 25-hydroxy-D2 metabolite is synthesized in vivo, however, suggesting that an enzyme distinct from CYP27A1 is involved in its synthesis.
A. Phenotype of cyp27A1-Deficient Mice The CYP27A1 enzyme is important for bile acid biosynthesis and regulation of cholesterol homeostasis [10,11]. Mutations in the human gene cause cerebrotendinous xanthomatosis (CTX), a lipid storage disorder leading to premature atherosclerosis and progressive neurological dysfunction [6,7]. Mice deficient for cyp27A1 have been engineered [5]. Homozygous cyp27A1−/− animals have larger livers and larger adrenals [16], decreased synthesis and excretion of bile acids [5], and hypertriglyceridemia [16]. The increased formation of 25-hydroxylated bile alcohols and cholestanol observed in CTX patients was not observed in cyp27A1−/− mice, and no CTX-related pathological abnormalities were evident [5]. At present, it remains
unclear whether the cyp27A1-deficient mouse is a valid animal model for CTX. Interestingly, cyp27A1-null mice had normal serum concentrations of 1,25(OH)2D and slightly elevated levels of 25(OH)D in their blood [5]. These results strengthen the hypothesis that a distinct vitamin D 25-hydroxylase enzyme plays a physiological role in vitamin D metabolism.
B. Identification of Novel Vitamin D-25-Hydroxylases An expression-based screening strategy has been used to demonstrate that CYP2R1 has vitamin D 25-hydroxylase activity. The experimental strategy used a cDNA library from cyp27A1-null mice to avoid isolating clones encoding the sterol 27-hydroxylase (CYP27A1). Expression of CYP2R1 in transfected cells led to transcriptional activation through the vitamin D receptor when vitamin D2, vitamin D3, or 1α(OH)D3 was added to cells [17]. The results obtained strongly suggest that CYP2R1 is the microsomal vitamin D 25-hydroxylase. Mice with mutations in the cyp2R1 gene are being engineered. It will prove interesting to analyze vitamin D metabolism and circulating levels of 25(OH)D and 1,25(OH)2D in cyp2R1-null mice. Based on specific activity determination, it was estimated that the microsomal enzyme is responsible for about 30% of the total 25-hydroxylation activity, while the mitochondrial enzyme is responsible for the remaining 70% [18]. This suggests a redundancy of function that could lead to no detectable vitamin D–related phenotype in cyp2R1-deficient mice. Altering the vitamin D 25-hydroxylase activity in mice in vivo would then require crossing the cyp27A1-null and cyp2R1-deficient strains and measuring circulating vitamin D metabolites levels as a function of cyp27A1 and cyp2R1 gene dosage. Purified recombinant CYP3A4 was also recently shown to 25-hydroxylate the 1α(OH)D2 substrate in vitro, with lower enzymatic activity toward 1α(OH)D3 or vitamin D2 [19]. Interestingly, CYP3A4 expression
107
CHAPTER 7 Mutant Mouse Models of Vitamin D Metabolic Enzymes
vectors were devoid of 25-hydroxylase activity in the transfection-based strategy that led to the identification of CYP2R1 as a putative microsomal vitamin D 25-hydroxylase [17]. CYP3A4 regulates the metabolism of more than half of all commonly used drugs. It remains to be determined whether CYP3A4 plays any physiological role in vitamin D metabolism or whether its activity is limited to the conversion of the prodrug 1α(OH)D2.
III. 24-HYDROXYLATION AND cyp24A1 KNOCKOUT The CYP24A1 enzyme catalyzes the addition of an hydroxyl group on carbon 24 of the vitamin D secosteroid backbone. It can utilize several substrates: 25(OH)D, to produce 24,25-dihydroxyvitamin D [24,25(OH)2D]; 1,25(OH)2D, to produce 1,24,25trihydroxyvitamin D [1,24,25(OH)3D]; and even 1α-hydroxyvitamin D [1α(OH)D], to generate 1α,24dihydroxyvitamin D [1,24(OH)2D]. The 24,25(OH)2D metabolite is the most abundant dihydroxylated metabolite in the circulation. The C-24 oxidation pathway catalyzed by the CYP24A1 molecule comprises five enzymatic steps that involve successive hydroxylation/oxidation reactions at carbons 24 and 23, followed by cleavage of the secosteroid at the C-23/C-24 bond and subsequent oxidation of the cleaved product to calcitroic acid [20,21]. The hypothesis that the main role of the C-24 oxidation pathway is attenuation of the 1,25(OH)2D biological signal inside target cells was tested in vitro using cytochrome P450 inhibitors. Blocking P450 activity by treatment of cells with ketoconazole inhibits catabolism and results in 1,25(OH)2D accumulation and extended hormone action [22]. This hypothesis was also tested and confirmed in vivo by engineering cyp24A1-deficient mice.
A. Phenotype of cyp24A1-Deficient Mice Exons 9 and 10 of the cyp24A1 gene, encoding the heme-binding domain [23], were deleted using homologous recombination in embryonic stem (ES) cells. This strategy ensured that a true null allele was created. Heterozygous animals were phenotypically normal and fertile, and the engineered mutation was transmitted to the progeny with the expected Mendelian ratio [24]. Fifty percent of cyp24A1−/− mice die before 3 weeks of age (Table II) [24,25]. Analysis of macrophage function ruled out impaired responses to infection as the cause for post-natal death. The perinatal lethality is
TABLE II
Incomplete Penetrance of the cyp24A1−/− Perinatal Lethality Phenotype
Number of cyp24A1−/− embryos 578
Alive at weaning
Dead at weaning
274 (47%)
304 (53%)
most likely a consequence of hypercalcemia secondary to hypervitaminosis D, since the inactivation of the cyp24A1 gene in mice impaired the ability of the animals to clear 1,25(OH)2D. Bolus and chronic 1,25 (OH)2D administration resulted in a marked elevation in serum 1,25(OH)2D levels in the mutant animals [24]. Chronic 1,25(OH)2D administration in cyp24A1−/− mutants resulted in histological changes consistent with hypervitaminosis D in the kidney: cortical tubular dilation, necrotic debris, and calcification (nephrocalcinosis). The inability to regulate 1,25(OH)2D and calcium homeostasis presumably leads to fatal hypercalcemia. Indeed, extremely high levels of circulating 1,25(OH)2D and calcium were measured in runted animals that died before weaning [24]. Another condition where vitamin D homeostasis is disregulated in cyp24A1-null mice is pregnancy. Circulating 1,25(OH)2D levels rise in the late stages of gestation in rodents [26]. Since gestating cyp24A1−/− females cannot regulate this increase through CYP24A1-mediated catabolism, serum 1,25(OH)2D concentrations rise to abnormally high levels in pregnant cyp24A1−/− dams [24]. This is accompanied by concomitant elevation of the circulating calcium concentration [24] that sometimes reaches fatal levels. Since half of the mutant progeny appear unaffected by the cyp24A1 deficiency, these animals most likely use alternate means of regulating vitamin D homeostasis. We have measured clearance and metabolism of labeled 1,25(OH)2D in cyp24A1−/− survivors and heterozygote controls. These experiments have confirmed that cyp24A1-null mice have impaired clearance of 1,25(OH)2D. Surprisingly, cyp24A1−/− mice appear to lack not only 24-hydroxylated metabolites but also 1,25(OH)2D-26,23-lactone [27], supporting the view that the CYP24A1 enzyme is also responsible for hydroxylating 1,25(OH)2D at C-23 and C-26 [28,29]. These surprising findings suggest that the cyp24A1null survivors adapt to the impaired vitamin D catabolism not by using an alternative catabolic route, but by limiting the synthesis of the active compound. The survival of some cyp24A1−/− mutant animals to adulthood has also allowed experiments designed to address the effect of perturbing vitamin D metabolism during development. Bone development is
108 abnormal in homozygous mutants born of homozygous females [24,30]. Bone forms via two mechanisms during embryogenesis: intramembranous bone formation and endochondral bone formation [31,32]. The intramembranous formation allows the growth of the surface of flat bones as well as the thickening of cortical bone, while endochondral formation provides for the lengthening of long bones. The major difference between the two is the presence or absence of an intermediate cartilaginous phase. Intramembranous bone formation occurs when mesenchymal precursor cells proliferate and differentiate directly into osteoblasts that produce a mucoprotein matrix, called osteoid, in which collagen fibrils are enmeshed. The osteoblasts then begin to mineralize the osteoid by depositing inorganic crystals of calcium phosphate on, between, and within the collagen fibers, forming a primary immature bone tissue called woven bone. Woven bone is progressively replaced by mature, lamellar bone. In lamellar bone, the matrix is compact and the collagen fibers are long and highly ordered. Endochondral bone formation entails the conversion of a cartilaginous template, the “anlagen,” into bone. Mesenchymal cells condense and differentiate into chondroblasts that secrete the cartilaginous matrix. The chondroblasts embed themselves within the matrix they produce and become chondrocytes. Because of the malleable nature of the cartilaginous matrix, the chondrocytes can continue to proliferate in their lacunae. The embryonic anlagen, a template of the future bone piece, is avascular. During its early development, a ring of woven bone (the bony collar) is formed by intramembranous ossification in the future midshaft area. This calcified woven bone is then invaded by vascular tissue, and this process of angiogenesis brings to the bone rudiment the precursors of the osteoclastic lineage. The osteoclasts excavate the hematopoietic bone marrow cavity while new osteoblasts are recruited to replace the cartilage scaffold with bone matrix. At the extremities of long bones (epiphyses), longitudinal growth occurs by a similar process of endochondral bone formation at the growth plates [32]. Growth-plate chondrocytes undergo several steps of maturation from proliferating chondrocytes, forming characteristic orderly columns, to nonproliferating, hypertrophic cells. The cartilaginous matrix becomes mineralized just below the hypertrophic zone of the growth plate and the chondrocytes then undergo apoptosis. Once calcified, the cartilage matrix is partially resorbed by osteoclasts. Following resorption, osteoblasts differentiate to form a layer of woven bone on top of the mineralized cartilage remnants of the longitudinal septa. This is the first activation–resorption– formation sequence of remodeling the cartilage into
RENÉ ST-ARNAUD
woven bone. Further down in the growth plate, this immature bone will undergo subsequent remodeling to replace the woven bone and calcified cartilage remnants with lamellar bone. An extensive literature demonstrates that cells from the growth plate respond to 24,25(OH)2D in vitro in a cell maturation–dependent manner (reviewed in [33]). The less differentiated cells of the resting zone respond to 24,25(OH)2D. The more mature cells of the prehypertrophic and hypertrophic compartments respond primarily to 1,25(OH)2D. Interestingly, treatment of resting zone chondrocytes with 24,25(OH)2D induces a change in maturation state [34], supporting the hypothesis that 24,25(OH)2D plays a role in cartilage development. It is interesting to note that no significant disruption of growth plate organization was noted in cyp24A1-null mice and that all cell types could readily be identified (Fig. 1), with the expression of differentiation markers such as type II collagen and type X collagen confirmed (St-Arnaud and Beaulieu, unpublished results). Our data suggest that the absence of CYP24A1 activity does not affect growth plate development and
FIGURE 1 Normal growth plate architecture in cyp24A1-null mice. Sagittal sections through the vertebrae of cyp24A1+/− (upper panel) and cyp24A1−/− (lower panel) littermates. Sections were stained by the Goldner method. See color plate.
CHAPTER 7 Mutant Mouse Models of Vitamin D Metabolic Enzymes
that 24,25(OH)2D is not required for chondrocyte maturation in vivo. It remains formally possible, however, that a redundant endocrine system may be able to compensate for the function of the 24,25(OH)2D metabolite in living animals. Histological examination of the bones from homozygous cyp24A1−/− pups born from homozygous cyp24A1-null dams revealed an accumulation of osteoid at sites of intramembranous ossification, however [24,30]. Control heterozygote littermates showed normal bone structure. Two major hypotheses can be formulated to account for the phenotype of the cyp24A1-deficient embryos from cyp24A1 mutant mice: (1) perturbation of 1,25(OH)2D catabolism through the inactivation of the C-24 oxidation pathway affects bone development; (2) metabolites of vitamin D hydroxylated at position 24 are essential for bone development. To differentiate between these possibilities, the cyp24A1-deficient animals were bred to mice carrying an inactivating mutation of the vitamin D receptor (VDR) gene [35]. If elevated 1,25(OH)2D levels, acting through the vitamin D receptor, were responsible for the observed phenotype, then mice lacking the receptor and the cyp24A1 gene should not show the aberrant intramembranous bone development. Using this elegant genetic strategy, incontrovertible evidence was obtained that expression of the VDR is required for the impaired mineralization phenotype of the cyp24A1-deficient animals: double mutant homozygotes (cyp24A1−/− and VDR−/−) showed normal intramembranous bone formation at all sites examined [24]. This demonstrates that elevated 1,25(OH)2D levels during gestation affect mineralization and suggest that impaired vitamin D metabolism during development perturbs bone formation.
B. Future Studies The cyp24A1-deficient strain represents a useful tool to address the still controversial question of the putative biological activity of vitamin D metabolites hydroxylated at the 24 position. Does it remain possible that 24,25(OH)2D exerts biological activity in specific tissues and that genetic redundancy prevented the identification of this activity in cyp24A1-null mice? Using the cyp24A1-ablated mice as a source of costochondral chondrocytes would confirm or refute the importance of local production of vitamin D metabolites in the maturation of chondrocytic cells in this culture system [36]. Published results also suggest that 24,25(OH)2D may be important during fracture repair. The circulating levels of 24,25(OH)2D increase during fracture repair
109
in chicks because of an increase in renal CYP24A1 activity [37]. When the effect of various vitamin D metabolites on the mechanical properties of healed bones was tested, the strength of healed bones was maximized in animals fed 24,25(OH)2D3 [38]. These results support a role for 24,25(OH)2D as an essential vitamin D metabolite for fracture repair. We have started to use the murine model of distraction osteogenesis [39] to compare fracture repair between wild-type mice and cyp24A1−/− mutant mice. If a defect in fracture repair is observed in cyp24A1-null mice, we will attempt rescue using 24,25(OH)2D, 1,24,25(OH)3D, 1,24(OH)2D, or other vitamin D analogs.
IV. 1α-HYDROXYLATION AND cyp27B1 KNOCKOUT The 25-hydroxyvitamin D-1α-hydroxylase enzyme (CYP27B1) converts 25(OH)D to 1,25(OH)2D, the biologically active metabolite of vitamin D. Mutations in the CYP27B1 gene cause pseudo–vitamin D deficiency Rickets (PDDR), a rare autosomal disease characterized by growth retardation, failure to thrive, rickets, and osteomalacia [40,41] (see Chapter 71). Biochemical analysis of serum from affected patients reveals hypocalcemia, secondary hyperparathyroidism, and undetectable levels of 1,25(OH)2D3.
A. Phenotype of cyp27B1-Deficient Mice An animal model of PDDR was engineered independently by two laboratories using targeted inactivation of the cyp27B1 gene in mice [42,43]. The engineered mutation was transmitted to the progeny with the expected Mendelian ratio. Heterozygous animals had no discernable phenotype and were fertile. Homozygous mutant animals were phenotypically normal at birth but exhibited retarded growth as measured by weight gain [43] and femur length at 8 weeks [44]. Serum analysis of homozygous mutant animals confirmed that they were hypocalcemic, hypophosphatemic, and had undetectable circulating levels of 1,25(OH)2D3 [42,43]. Serum parathyroid hormone levels were highly increased as early as weaning [42] and remained elevated thereafter [42,43]. The parathyroid glands also increased in size in young adult cyp27B1 homozygous mutant mice [42,43]. From a biochemical standpoint, the only difference between the cyp27B1−/− mice and patients with PDDR is that patients with PDDR have normal serum levels of 25(OH)D [45–47] and 24,25(OH)2D [48], whereas
110 elevated levels of 25(OH)D [42,43] and very low levels of 24,25(OH)2D [42] are seen in cyp27B1 mutant mice. The 1,25(OH)2D hormone is the main in vivo regulator of the expression of the CYP24A1 enzyme that catalyzes the synthesis of 24,25(OH)2D [49]. Therefore, low circulating 24,25(OH)2D levels and undetectable cyp24A1 expression in cyp27B1−/− mice would be anticipated. The discrepancy with the human disease remains to be explained but could result from species differences. This observed inhibition of cyp24A1 expression in mice, combined with the targeted ablation of the CYP27B1 enzyme, leads to a metabolic block in mutant animals and an accumulation of the unprocessed 25(OH)D substrate [42]. The expression of vitamin D–dependent genes was also analyzed in cyp27B1-null mice. Expression of the calbindin-D9k gene was completely inhibited in the intestine of mutant mice [42,43] and severely downregulated in the kidney [43]. A reduction in the levels of mRNA for calbindin-D28k and the vitamin D receptor in kidney tissue was observed, whereas renal expression of cyp24A1 was almost completely abolished [42,43]. Interestingly, expression levels of the osteocalcin and osteopontin genes in tibia were unaffected in mutant animals [42]. The osteocalcin gene promoter contains several positive and negative regulatory elements [50]. Similarly, osteopontin gene expression is regulated by many agents acting via diverse signaling pathways in specific cell types [51]. Although 1,25(OH)2D3 has been shown to regulate the expression of these genes both in vitro and in vivo [50,51], the unaltered mRNA levels for both genes measured in cyp27B1-null animals suggest that the vitamin D hormone is not the main modulator of the steady-state expression of these genes in murine bone tissue. Histological analysis of the bones from 3-week-old mutant animals confirmed the evidence of rickets [42]. At the age of 7 to 8 weeks, femurs from cyp27B1-ablated mice present a severe disorganization in the architecture of the growth plate and marked osteomalacia [42,43]. The X-ray features of the bones from cyp27B1−/− animals also matched the clinical manifestations of PDDR. Contact radiography of femurs from mutant animals revealed diffuse osteopenia (hypomineralization) and rachitic metaphyseal changes [52]. Inactivation of cyp27B1 therefore produced a bone phenotype that is identical to the human genetic disease PDDR, confirming that mutations in CYP27B1 underlie this disorder. The animal model was further used to compared efficacy of clinical rescue regimens (see later discussion). The vitamin D hormone has been shown to have immunomodulatory effects, and local synthesis of 1,25(OH)2D could play an important autocrine or paracrine role in the differentiation or function of immune cells [53–55]. Interestingly, cyp27B1-deficient
RENÉ ST-ARNAUD
mice exhibited a reduction in CD4- and CD8-positive peripheral T lymphocyte populations [43]. Large CD4positive ectopic lymph nodes were detected in the neck of the cyp27B1-null and cyp27B1+/− heterozygous animals [43]. These findings support an important role for CYP27B1 in the regulation of immune function. This phenotype has never been reported in vitamin D receptor (VDR) knockout strains [35,56,57], which raises the hypothesis that 1,25(OH)2D may not exert its immunomodulatory effects via the VDR or that CYP27B1 influences immune function via metabolites other than 1,25(OH)2D [43]. The fertility and reproductive capacity of vitamin D– deficient rats is impaired, and normal reproductive effectiveness can be rescued by normalizing calcemia in males but not females, where 1,25(OH)2D is required for reproductive function [58,59]. cyp27B1-deficient females showed uterine hypoplasia, decreased ovarian size, compromised folliculogenesis, and infertility [43]. A similar phenotype was observed in one strain of VDR-deficient mice [57]. These results establish a critical role for cyp27B1 function in female reproduction.
B. Rescue of the Phenotype 1. 1,25(OH)2D REPLACEMENT THERAPY
The treatment of choice for PDDR patients is replacement therapy with 1,25(OH)2D3 [45]. It results in rapid and complete correction of the abnormal phenotype, restoring normocalcemia, eliminating secondary hyperparathyroidism and features of rickets. The restoration of bone mineral content is equally rapid and histological evidence of healing has been documented [45]. This therapeutic regimen was used to rescue the PDDR phenotype of cyp27B1−/− mice [60]. Animals received daily subcutaneous injections of 1,25(OH)2D3. A “rescue” concentration of 500 pg per g body weight (BW) was administered from 21 days of age until 35 days of age, followed by a “maintenance” dose of 100 pg per g BW (s.c.) from 35 days of age until they reached the age of 60 days. This treatment normalized biochemical parameters, restored the biomechanical properties of bones, and cured rickets and osteomalacia, as assessed by histomorphometric analyses [60]. PDDR patients receiving 1,25(OH)2D3-replacement therapy have normal serum concentrations of the metabolite [45], whereas chronic treatment with 1,25(OH)2D3 did not normalize circulating levels of the metabolite in cyp27B1−/− mice [60]. This difference is most likely due to the induction of the CYP24A1 enzyme observed in 1,25(OH)2D3-treated cyp27B1null mutants [60], and further highlights the species differences in vitamin D metabolic pathways that was noted in the first description of the animal model for
111
CHAPTER 7 Mutant Mouse Models of Vitamin D Metabolic Enzymes
PDDR [42,43], with regard to the regulation of cyp24A1 gene expression and circulating 25(OH)D3 or 24,25(OH)2D3 metabolite levels. 2. 1α(OH)D REPLACEMENT THERAPY
Before 1,25(OH)2D3 became available as a drug, clinicians used the monohydroxylated analog 1α(OH)D3 for treating PDDR patients [61]. This compound, which requires only liver hydroxylation at the 25 position to fully mimic 1,25(OH)2D3, is still used in Japan and some European countries. We have used a rescue regimen similar to the treatment described earlier to treat cyp27B1-deficient mice with 1α(OH)D3. The treatment was as effective as the injection protocol with 1,25(OH)2D3 to normalize calcemia, phosphatemia, and reverse the secondary hyperparathyroidism in cyp27B1−/− mice, without deleterious effects in heterozygous controls. The replacement therapy with 1α(OH)D3 also normalized growth and corrected the bone phenotype of the mutant mice, effectively curing rickets and osteomalacia and normalizing histomorphometric indices. Experiments are in progress to compare the efficacy of other vitamin D analogs, such as 1α(OH)D2 or 1,24(S)(OH)2D2 [62], in rescuing the PDDR phenotype of cyp27B1 knockout animals. 3. HIGH-CALCIUM, HIGH-PHOSPHORUS, HIGH-LACTOSE DIET
Mutation in the vitamin D receptor (VDR) gene results in a second form of vitamin D–related rickets called hereditary vitamin D–resistant rickets (HVDRR),
characterized by hypocalcemia, secondary hyperparathyroidism, rickets, osteomalacia, and alopecia [63] (see Chapter 72). Several laboratories have developed a valid mouse model for this type of hereditary rickets [35,56,57]. Rescue of rickets has been successfully performed using a 2% calcium diet, 1.25% phosphorus and 20% lactose, previously reported to prevent hypocalcemia in vitamin D–deficient rats [64]. This diet completely normalized all biochemical and histomorphometric parameters in VDR-ablated animals [65,66]. We rescued the PDDR phenotype of mice deficient for the cyp27B1 gene by feeding them with the highcalcium diet [67]. Homozygous mutant and heterozygous control animals were fed the rescue diet (2% calcium, 1.25% phosphorus, and 20% lactose) from 3 weeks of age until sacrifice at 8.5 weeks of age. Blood biochemistry analysis revealed that the rescue diet corrected the hypocalcemia and secondary hyperparathyroidism. Despite the restoration of normocalcemia, cyp27B1−/− (and heterozygous control) animals fed the rescue diet initially gained weight less rapidly than control mice fed normal mouse chow. Although cyp27B1−/− mice fed the rescue diet eventually reached the same weight as control animals, the treatment did not entirely correct bone growth as femur size remained significantly smaller than control (Fig. 2). Bone histology and histomorphometry confirmed that the rickets and osteomalacia were cured. The rescue diet also restored the biomechanical properties of the bone tissue within normal parameters. These results demonstrate that
Growth rescue in cyp27B1-deficient mice *
20
**
control
* Femur length (mm)
1,25(OH)2D3 rescue High Ca2+ diet rescue 10
0 +/−
−/−
The high-calcium rescue diet only partially restores bone growth of cyp27B1−/− mice. Femurs were measured at sacrifice (60 days old). cyp27B1−/− animals had significantly reduced femur lengths, which was only partially corrected by the high-calcium diet (High Ca2+ diet rescue). Replacement therapy with 1,25(OH)2D3 (1,25(OH)2D3 rescue) normalized femur growth. *, p < 0.05; **, p < 0.01.
FIGURE 2
112 correction of the abnormal mineral ion homeostasis by feeding with a high-calcium rescue diet is effective to rescue the PDDR phenotype of cyp27B1 mutant mice. This treatment, however, does not appear as effective as 1,25(OH)2D3 replacement therapy since bone growth remained impaired. In addition to the partial rescue of femur growth, the high-calcium rescue diet increased growth plate thickness in cyp27B1−/− and cyp27B1+/− animals [67]. This phenomenon was not observed in VDR-ablated mice receiving similar treatment [65,66]. A major difference between the two strains of animals is the ability to synthesize 1,25(OH)2D3, both systemically and locally. It has been hypothesized that local production of 1,25(OH)2D3 is important for the proliferation and maturation of chondrocytes in the growth plate, and that some of these effects involve nongenomic mechanisms of action for 1,25(OH)2D3 [33]. Autocrine synthesis and nongenomic actions of 1,25(OH)2D3 would remain active in VDR-ablated animals, whereas they are affected in cyp27B1-deficient mice. This could explain the differences in the response of the two strains to the dietary manipulation. The targeted cyp27B1 mouse model that we have engineered [42] represents an invaluable tool to characterize the putative autocrine/ paracrine roles of 1,25(OH)2D3 in the growth plate under conditions of normal mineral ion homeostasis (see later discussion).
C. Future Studies Local production of 1,25(OH)2D could play an important autocrine or paracrine role in the differentiation or function of osteoblasts, chondrocytes, keratinocytes, and macrophages [33,68,69]. The targeting strategy used to engineer the cyp27B1-deficient strain, based on the Cre/lox methodology, allows tissuespecific inactivation of the targeted gene. The technology utilizes the Cre recombinase from the P1 bacteriophage. This enzyme catalyzes recombination between two 34-base-pair recognition elements, called loxP sites, causing excision of the intervening sequences. Transgenic mice lines expressing Cre under the control of tissue-specific promoters are crossed with mice harboring loxP sites to achieve excision at will (Fig. 3) (reviewed in [70]). The cyp27B1 gene was cloned from a 129 SV mouse genomic library to construct a targeting vector in which exon 8, encoding the heme binding domain [71], was flanked by a 5′-loxP recognition site and by a 3′-loxP-neo-loxP selection cassette. Following homologous recombination at the cyp27B1 locus in ES cells and injection of targeted ES cells in host blastocysts, a mouse strain in which the cyp27B1
RENÉ ST-ARNAUD
locus is “floxed” (flanked by loxP) was established. The conditional cyp27B1 allele provides an invaluable genetic tool to analyze the putative autocrine/paracrine roles that have been hypothesized for 1,25(OH)2D in various cell types. It may be argued that a significant portion of the active 1,25(OH)2D metabolite found in target tissues is derived systemically. Against the background of circulating levels of the hormone, any local drop may be too subtle an effect to be easily detectable. However, a recent finding in the field of sex steroids is that a large proportion of androgens and estrogens are synthesized locally in peripheral target tissues such as skin or adipose tissue [72]. Despite “central” steroidogenesis in the gonads, peripheral target tissues utilize “on the spot” intracellular hormone synthesis/action to regulate their function [73]. Moreover, chondrocytes are compartmentalized away from the circulation, and local synthesis of vitamin D metabolites may contribute significantly to the regulation of the proliferation, differentiation, and function of growth plate cells. The 1,25(OH)2D hormone has been shown to stimulate macrophage function. Treatment of macrophages with 1,25(OH)2D enhances macrophage chemotaxis [74] and increases phagocytic killing of bacteria [75]. Crossing the floxed cyp27B1 mice with mice expressing Cre in macrophages under the control of the macrophage scavenger receptor A gene promoter [76] would inhibit autocrine synthesis of 1,25(OH)2D by macrophages. One possible phenotype that could be predicted following this genetic manipulation would be increased susceptibility to microbial infection. The 1,25(OH)2D hormone is likely to be an autocrine or paracrine factor for epidermal differentiation [77] since keratinocytes make 1,25(OH)2D [78], express the VDR [79], and respond to 1,25(OH)2D with changes in proliferation and differentiation [79,80]. Indeed, our results show that the expression of keratinocyte differentiation markers is reduced in the skin of cyp27B1 knockout animals and that recovery of the epidermal barrier (assessed by the transepidermal water loss assay) [81] is significantly delayed in CYP27B1-deficient skin [82]. Tissue-specific inactivation of cyp27B1 in floxed mice crossed to keratin 5-Cre [83] or keratin 14-Cre transgenicmice [84] would allow examination of the contribution of systemic versus local production of 1,25(OH)2D on skin differentiation. Chondrocytes have been shown to express cyp27B1 both in vivo [71] and in vitro [36,85]. This has led investigators to hypothesize that local production of 1,25(OH)2D could play an important autocrine or paracrine role in the differentiation or function of these cells [33]. To test this putative role of the vitamin D
CHAPTER 7 Mutant Mouse Models of Vitamin D Metabolic Enzymes
113
FIGURE 3 Tissue-specific gene inactivation using the Cre/lox strategy. Floxed (flanked by loxP) mice are crossed with transgenic mice in which the Cre recombinase is under the control of a tissue-specific promoter (TSP). In this example, loxP sites (red triangles) flank exon 8 of the target gene and the tissue-specific promoter is expressed in bone tissue. Expression of the Cre transgene in bone cells will cause excision of exon 8 between the two loxP sites (bottom left), while the target gene will remain intact in other tissues due to lack of Cre expression (bottom right panel).
hormone in growth plate cells, we have initiated cyp27B1 gene ablation in chondrocytes [86] by crossing the floxed cyp27B1 strain with transgenic mice expressing the Cre recombinase under the control of the type II collagen promoter/enhancer (Col2-Cre) [87]. Because 1,25(OH)2D regulates the growth and terminal differentiation of hypertrophic chondrocytes in vitro [33,88,89], we hypothesize that these parameters will be perturbed in vivo following the genetic manipulation that we will engineer. This should be manifested as increased chondrocyte proliferation and an abnormal architecture of the growth plate, which could even affect the primary spongiosa. Neovascularization and VEGF expression could be reduced [90]. The expression of differentiation markers could be perturbed in Col2-Cre; 1α−/fl mice.
V. SUMMARY AND PERSPECTIVES Mutant mouse models have allowed significant progress in our understanding of vitamin D metabolism and function. The observation that cyp27A1−/− (mitochondrial 25-OHase) mice have normal vitamin D homeostasis suggested that a distinct cytochrome P450 enzyme could be involved in the 25-hydroxylation of vitamin D under physiological conditions. An expressionbased cloning strategy using a cDNA library prepared from the liver of cyp27A1−/− animals was then used to identify the microsomal 25-hydroxylating enzyme, CYP2R1. It will prove interesting to analyze vitamin D metabolism and circulating levels of 25(OH)D and 1,25(OH)2D in cyp2R1-null mice and cyp27A1-cyp2R1 compound mutants to determine the relative contributions
114 of the two enzymes in vivo. The cyp24A1-deficient strain will be used to test the putative role of vitamin D metabolites hydroxylated at position 24 in mammalian fracture repair. The confirmation of such a role for 24,25(OH)2D would open up interesting avenues for clinical research using vitamin D analogs. Finally, the cyp27B1-deficient mice represent a valid animal model for pseudo–vitamin D deficiency rickets. The conditional ablation strategy used to engineer these animals will help elucidate the hypothesized autocrine/ paracrine roles of 1,25(OH)2D in target tissues such as cartilage, skin, and macrophages.
References 1. McAlpine PJ, Van Cong N, Boucheix C, Pakstis AJ, Doute RC, Shows TB 1987 The 1987 Catalog of mapped genes and report of the nomenclature committee. Cytogenet Cell Genet 46:29–101. 2. Shows TB, et al., McAlpine PJ, Boucheix C, Collins FS, Conneally PM, Frezal J, Gershowitz H, Goodfellow PN, Hall JG 1987 Guidelines for human gene nomenclature. An international system for human gene nomenclature (ISGN, 1987). Cytogenet Cell Genet 46:11–28. 3. Okuda KI, Usui E, Ohyama Y 1995 Recent progress in enzymology and molecular biology of enzymes involved in vitamin D metabolism. J Lipid Res 36:1641–1652. 4. Jones G, Strugnell SA, DeLuca HF 1998 Current understanding of the molecular actions of vitamin D. Physiol Rev 78:1193–1231. 5. Rosen H, Reshef A, Maeda N, Lippoldt A, Shpizen S, Triger L, Eggertsen G, Bjorkhem I, Leitersdorf E 1998 Markedly reduced bile acid synthesis but maintained levels of cholesterol and vitamin D metabolites in mice with disrupted sterol 27-hydroxylase gene. J Biol Chem 273:14805–14812. 6. Salen G, Shefer S, Cheng FW, Dayal B, Batta AK, Tint GS 1979 Cholic acid biosynthesis: The enzymatic defect in cerebrotendinous xanthomatosis. J Clin Invest 63:38–44. 7. Cali JJ, Hsieh CL, Francke U, Russell DW 1991 Mutations in the bile acid biosynthetic enzyme sterol 27-hydroxylase underlie cerebrotendinous xanthomatosis. J Biol Chem 266: 7779–7783. 8. Kuriyama M, Fujiyama J, Kubota R, Nakagawa M, Osame M 1993 Osteoporosis and increased bone fractures in cerebrotendinous xanthomatosis. Metabolism 42:1497–1498. 9. Fukushima M, Suzuki Y, Tohira Y, Nishii Y, Suzuki M 1976 25-Hydroxylation of 1α-hydroxyvitamin D3 in vivo and in perfused rat liver. FEBS Lett 65:211–214. 10. Cali JJ, Russell DW 1991 Characterization of human sterol 27-hydroxylase. A mitochondrial cytochrome P-450 that catalyzes multiple oxidation reactions in bile acid biosynthesis. J Biol Chem 266:7774–7778. 11. Andersson D, Davis DL, Dahlback H, Jornvall H 1989 Cloning, structure and expression of the mitochondrial cytochrome P-450 sterol 26-hydroxylase, a bile acid biosynthetic enzyme. J Biol Chem 264:8222–8229. 12. Madhok TC, DeLuca HF 1979 Characteristics of the rat liver microsomal enzyme system converting cholecalciferol into 25-hydroxycholecalciferol. Evidence for the participation of cytochrome p-450. Biochem J 184:491–499.
RENÉ ST-ARNAUD
13. Vlahcevic ZR, Jairath SK, Heuman DM, Stravitz RT, Hylemon PB, Avadhani NG, Pandak WM 1996 Transcriptional regulation of hepatic sterol 27-hydroxylase by bile acids. Am J Physiol 270:G646–652. 14. Bhattacharyya MH, DeLuca HF 1973 The regulation of rat liver calciferol-25-hydroxylase. J Biol Chem 248:2969–2973. 15. Guo YD, Strugnell S, Back DW, Jones G 1993 Transfected human liver cytochrome P-450 hydroxylates vitamin D analogs at different side-chain positions. Proc Natl Acad Sci USA 90:8668–8672. 16. Repa JJ, Lund EG, Horton JD, Leitersdorf E, Russell DW, Dietschy JM, Turley SD 2000 Disruption of the sterol 27-hydroxylase gene in mice results in hepatomegaly and hypertriglyceridemia. Reversal by cholic acid feeding. J Biol Chem 275:39685–39692. 17. Cheng JB, Motola DL, Mangelsdorf DJ, Russell DW 2003 De-orphanization of cytochrome P450 2R1: A microsomal vitamin d 25-hydroxylase. J Biol Chem 278:38084–38093. 18. Bjorkhem I, Holmberg I 1978 Assay and properties of a mitochondrial 25-hydroxylase active on vitamin D3. J Biol Chem 253:842–849. 19. Gupta RP, Hollis BW, Patel SB, Bell NH 2003 CYP3A4 is a human microsomal vitamin D-25-hydroxylase. J Bone Miner Res (in press). 20. Makin G, Lohnes D, Byford V, Ray R, Jones G 1989 Target cell metabolism of 1,25-dihydroxyvitamin D3 to calcitroic acid. Evidence for a pathway in kidney and bone involving 24-oxidation. Biochem J 262:173–180. 21. Reddy GS, Tserng KY 1989 Calcitroic acid, end product of renal metabolism of 1,25-dihydroxyvitamin D3 through C-24 oxidation pathway. Biochemistry 28:1763–1769. 22. Reinhardt TA, Horst RL 1989 Ketoconazole inhibits selfinduced metabolism of 1,25-dihydroxyvitamin D3 and amplifies 1,25-dihydroxyvitamin D3 receptor up-regulation in rat osteosarcoma cells. Arch Biochem Biophys 272:459–465. 23. Ohyama Y, Noshiro M, Eggertsen G, Gotoh O, Kato Y, Bjorkhem I, Okuda K 1993 Structural characterization of the gene encoding rat 25-hydroxyvitamin D3 24-hydroxylase. Biochemistry 32:76–82. 24. St-Arnaud R, Arabian A, Travers R, Barletta F, Raval-Pandya M, Chapin K, Depovere J, Mathieu C, Christakos S, Demay MB, Glorieux FH 2000 Deficient mineralization of intramembranous bone in vitamin D-24-hydroxylase-ablated mice is due to elevated 1,25-dihydroxyvitamin D and not to the absence of 24,25-dihydroxyvitamin D. Endocrinology 141:2658–2666. 25. St-Arnaud R 1999 Targeted inactivation of vitamin D hydroxylases in mice. Bone 25:127–129. 26. Kovacs CS, Kronenberg HM 1997 Maternal-fetal calcium and bone metabolism during pregnancy, puerperium, and lactation. Endocr Rev 18:832–872. 27. Masuda S, Byford V, Arabian A, Sakai Y, Demay MB, St-Arnaud R, Jones G 2004 Altered pharmacokinetics of 1α,25-dihydroxyvitamin D3 and 25-hydroxyvitamin D3 in blood and tissues of the vitamin D-24-hydroxylase null mouse. Submitted. 28. Miyamoto Y, Shinki T, Yamamoto K, Ohyama Y, Iwasaki H, Hosotani R, Kasama T, Takayama H, Yamada S, Suda T 1997 1α,25-dihydroxyvitamin D3-24-hydroxylase (CYP24) hydroxylates the carbon at the end of the side chain (C-26) of the C-24-fluorinated analog of 1α,25-dihydroxyvitamin D3. J Biol Chem 272:14115–14119. 29. Beckman MJ, Tadikonda P, Werner E, Prahl J, Yamada S, DeLuca HF 1996 Human 25-hydroxyvitamin D3-24-hydroxylase, a multicatalytic enzyme. Biochemistry 35:8465–8472.
CHAPTER 7 Mutant Mouse Models of Vitamin D Metabolic Enzymes
30. St-Arnaud R, Glorieux FH 1997 Vitamin D and bone development. In: Feldman D, Glorieux FH, Pike JW (eds) Vitamin D. Academic Press, San Diego, pp. 293–303. 31. Zelzer E, Olsen BR 2003 The genetic basis for skeletal diseases. Nature 423:343–348. 32. Kronenberg HM 2003 Developmental regulation of the growth plate. Nature 423:332–336. 33. Boyan BD, Dean DD, Sylvia VL, Schwartz Z 1997 Cartilage and vitamin D: genomic and nongenomic regulation by 1,25(OH)2D3 and 24,25(OH)2D3. In: Feldman D, Glorieux FH, Pike JW (eds) Vitamin D. Academic Press, San Diego, pp. 395–422. 34. Schwartz Z, Dean D, Walton J, Brooks B, Boyan B 1995 Treatment of resting zone chondrocytes with 24,25-dihydroxyvitamin D3 [24,25-(OH)2D3] induces differentiation into a 1,25-(OH)2D3-responsive phenotype characteristic of growth zone chondrocytes. Endocrinology 136:402–411. 35. Li YC, Pirro AE, Amling M, Delling G, Baron R, Bronson R, Demay MB 1997 Targeted ablation of the vitamin D receptor: an animal model of vitamin D-dependent rickets type II with alopecia. Proc Natl Acad Sci USA 94:9831–9835. 36. Schwartz Z, Brooks B, Swain L, Del Toro F, Norman A, Boyan B 1992 Production of 1,25-dihydroxyvitamin D3 and 24, 25-dihydroxyvitamin D3 by growth zone and resting zone chondrocytes is dependent on cell maturation and is regulated by hormones and growth factors. Endocrinology 130:2495–2504. 37. Seo EG, Norman AW 1997 Three-fold induction of renal 25-hydroxyvitamin D3-24-hydroxylase activity and increased serum 24,25-dihydroxyvitamin D3 levels are correlated with the healing process after chick tibial fracture. J Bone Miner Res 12:598–606. 38. Seo EG, Einhorn TA, Norman AW 1997 24R,25-dihydroxyvitamin D3: An essential vitamin D3 metabolite for both normal bone integrity and healing of tibial fracture in chicks. Endocrinology 138:3864–3872. 39. Tay BK, Le AX, Gould SE, Helms JA 1998 Histochemical and molecular analyses of distraction osteogenesis in a mouse model. J Orthop Res 16:636–642. 40. Miller WL, Portale AA 1999 Genetic causes of rickets. Curr Opin Pediatr 11:333–339. 41. St-Arnaud R, Glorieux FH 2000 Hereditary defects in vitamin D metabolism and action. In: DeGroot LJ, Jameson JL (eds) Endocrinology, 4th ed., Vol. 2. W.B. Saunders, Philadelphia; pp. 1154–1168. 42. Dardenne O, Prud’homme J, Arabian A, Glorieux FH, St-Arnaud R 2001 Targeted inactivation of the 25-hydroxyvitamin D3-1α-hydroxylase gene (CYP27B1) creates an animal model of pseudovitamin D-deficiency rickets. Endocrinology 142:3135–3141. 43. Panda DK, Miao D, Tremblay ML, Sirois J, Farookhi R, Hendy GN, Goltzman D 2001 Targeted ablation of the 25-hydroxyvitamin D 1α-hydroxylase enzyme: Evidence for skeletal, reproductive, and immune dysfunction. Proc Natl Acad Sci USA 98:7498–7503. 44. St-Arnaud R, Demay MB 2003 Vitamin D biology. In: Glorieux FH, Juppner H, Pettifor JM (eds) Pediatric Bone: Biology & Diseases. Elsevier/Academic Press, San Diego, pp. 193–216. 45. Delvin EE, Glorieux FH, Marie PJ, Pettifor JM 1981 Vitamin D dependency: replacement therapy with calcitriol. J Pediatr 99: 26–34. 46. Rosen JF, Finberg L 1972 Vitamin D-dependent rickets: actions of parathyroid hormone and 25-hydroxycholecalciferol. Pediatr Res 6:552–562.
115
47. Scriver CR, Reade TM, DeLuca HF, Hamstra AJ 1978 Serum 1,25-dihydroxyvitamin D levels in normal subjects and in patients with hereditary rickets or bone disease. N Engl J Med 299:976–979. 48. Mandla S, Jones G, Tenenhouse HS 1992 Normal 24-hydroxylation of vitamin D metabolites in patients with vitamin Ddependency rickets type I. Structural implications for the vitamin D hydroxylases. J Clin Endocrinol Metab 74:814–820. 49. Shinki T, Jin CH, Nishimura A, Nagai Y, Ohyama Y, Noshiro M, Okuda K, Suda T 1992 Parathyroid hormone inhibits 25-hydroxyvitamin D3-24-hydroxylase mRNA expression stimulated by 1α,25-dihydroxyvitamin D3 in rat kidney but not in intestine. J Biol Chem 267:13757–13762. 50. Lian JB, Stein GS, Stein JL, van Wijnen AJ 1998 Osteocalcin gene promoter: Unlocking the secrets for regulation of osteoblast growth and differentiation. J Cell Biochem Suppl 31:62–72. 51. Denhardt DT, Noda M 1998 Osteopontin expression and function: role in bone remodeling. J Cell Biochem Suppl 31:92–102. 52. St-Arnaud R, Dardenne O, Glorieux FH 2001 Etiologie moléculaire des rachitismes vitamino-dépendents héréditaires. Médecine/Sciences 17:1289–1296. 53. Abe E, Miyaura C, Sakagami H, Takeda M, Konno K, Yamazaki T, Yoshiki S, Suda T 1981 Differentiation of mouse myeloid leukemia cells induced by 1α,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 78:4990–4994. 54. Nunn JD, Katz DR, Barker S, Fraher LJ, Hewison M, Hendy GN, O’Riordan JL 1986 Regulation of human tonsillar T-cell proliferation by the active metabolite of vitamin D3. Immunology 59:479–484. 55. Provvedini DM, Rulot CM, Sobol RE, Tsoukas CD, Manolagas SC 1987 1α,25-Dihydroxyvitamin D3 receptors in human thymic and tonsillar lymphocytes. J Bone Miner Res 2:239–247. 56. Van Cromphaut SJ, Dewerchin M, Hoenderop JG, Stockmans I, Van Herck E, Kato S, Bindels RJ, Collen D, Carmeliet P, Bouillon R, Carmeliet G 2001 Duodenal calcium absorption in vitamin D receptor-knockout mice: functional and molecular aspects. Proc Natl Acad Sci USA 98:13324–13329. 57. Yoshizawa T, Handa Y, Uematsu Y, Takeda S, Sekine K, Yoshihara Y, Kawakami T, Arioka K, Sato H, Uchiyama Y, Masushige S, Fukamizu A, Matsumoto T, Kato S 1997 Mice lacking the vitamin D receptor exhibit impaired bone formation, uterine hypoplasia and growth retardation after weaning. Nat Genet 16:391–396. 58. Halloran BP, DeLuca HF 1980 Effect of vitamin D deficiency on fertility and reproductive capacity in the female rat. J Nutr 110:1573–1580. 59. Kwiecinski GG, Petrie GI, DeLuca HF 1989 Vitamin D is necessary for reproductive functions of the male rat. J Nutr 119:741–744. 60. Dardenne O, Prudhomme J, Hacking SA, Glorieux FH, St-Arnaud R 2003 Rescue of the pseudo-vitamin D deficiency rickets phenotype of CYP27B1-deficient mice by treatment with 1,25-dihydroxyvitamin D3: biochemical, histomorphometric, and biomechanical analyses. J Bone Miner Res 18: 637–643. 61. Reade TM, Scriver CR, Glorieux FH, Nogrady B, Delvin E, Poirier R, Holick F, DeLuca HF 1975 Response to crystalline 1α-hydroxyvitamin D3 in vitamin D dependency. Pediatr Res 9:593–599. 62. Horst R, Prapong S, Reinhardt T, Koszewski N, Knutson J, Bishop C 2000 Comparison of the relative effects of 1,24dihydroxyvitamin D(2) [1,24-(OH)(2)D(2)], 1,24-dihydroxyvitamin D(3) [1,24-(OH)(2)D(3)], and 1,25-dihydroxyvitamin D(3)
116
63. 64. 65.
66.
67.
68.
69. 70. 71.
72. 73.
74.
75.
76.
77.
[1,25-(OH)(2)D(3)] on selected vitamin D-regulated events in the rat. Biochem Pharmacol 60:701–708. Malloy PJ, Pike JW, Feldman D 1999 The vitamin D receptor and the syndrome of hereditary 1,25-dihydroxyvitamin Dresistant rickets. Endocr Rev 20:156–188. Kollenkirchen U, Fox J, Walters MR 1991 Normocalcemia without hyperparathyroidism in vitamin D-deficient rats. J Bone Miner Res 6:273–278. Li YC, Amling M, Pirro AE, Priemel M, Meuse J, Baron R, Delling G, Demay MB 1998 Normalization of mineral ion homeostasis by dietary means prevents hyperparathyroidism, rickets, and osteomalacia, but not alopecia in vitamin D receptor-ablated mice. Endocrinology 139:4391–4396. Amling M, Priemel M, Holzmann T, Chapin K, Rueger JM, Baron R, Demay MB 1999 Rescue of the skeletal phenotype of vitamin D receptor-ablated mice in the setting of normal mineral ion homeostasis: formal histomorphometric and biomechanical analyses. Endocrinology 140:4982–4987. Dardenne O, Prud’homme J, Hacking SA, Glorieux FH, St-Arnaud R 2003 Correction of the abnormal mineral ion homeostasis with a high-calcium, high-phosphorus, highlactose diet rescues the PDDR phenotype of mice deficient for the 25-hydroxyvitamin D-1α-hydroxylase (CYP27B1). Bone 32:332–340. Hewison M, O’Riordan JLH 1997 Immunomodulatory and cell differentiation effects of vitamin D. In: Feldman L, Glorieux FH, Pike JW (eds) Vitamin D. Academic Press, San Diego, pp. 447–462. Bikle D 1997 Vitamin D and skin. In: Feldman D, Glorieux FH, Pike JW (eds) Vitamin D. Academic Press, San Diego, pp. 379–394. Jiang R, Gridley T 1997 Gene targeting: Things go better with Cre. Curr Biol 7:R321–323. Panda DK, Al Kawas S, Seldin MF, Hendy GN, Goltzman D 2001 25-hydroxyvitamin D 1α-hydroxylase: structure of the mouse gene, chromosomal assignment, and developmental expression. J Bone Miner Res 16:46–56. Chen W, Thiboutot D, Zouboulis CC 2002 Cutaneous androgen metabolism: basic research and clinical perspective. J Invest Dermatol 119:992–1007. Labrie F, Luu-The V, Lin SX, Simard J, Labrie C, El-Alfy M, Pelletier G, Bélanger A 2000 Intracrinology: role of the family of 17β−hydroxysteroid dehydrogenases in human physiology and disease. J Mol Endocrinol 25:1–16. Girasole G, Wang JM, Pedrazzoni M, Pioli G, Balotta C, Passeri M, Lazzarin A, Ridolfo A, Mantovani A 1990 Augmentation of monocyte chemotaxis by 1α,25-dihydroxyvitamin D3. Stimulation of defective migration of AIDS patients. J Immunol 145:2459–2464. Rook GA, Taverne J, Leveton C, Steele J 1987 The role of gamma-interferon, vitamin D3 metabolites and tumour necrosis factor in the pathogenesis of tuberculosis. Immunology 62:229–234. Horvai A, Palinski W, Wu H, Moulton KS, Kalla K, Glass CK 1995 Scavenger receptor A gene regulatory elements target gene expression to macrophages and to foam cells of atherosclerotic lesions. Proc Natl Acad Sci USA 92:5391–5395. Pillai S, Bikle DD, Elias PM 1988 Vitamin D and epidermal differentiation: Evidence for a role of endogenously produced
RENÉ ST-ARNAUD
78. 79. 80.
81. 82.
83.
84.
85.
86.
87.
88.
89.
90.
vitamin D metabolites in keratinocyte differentiation. Skin Pharmacol 1:149–160. Bikle DD, Nemanic MK, Whitney JO, Elias PW 1986 Neonatal human foreskin keratinocytes produce 1,25-dihydroxyvitamin D3. Biochemistry 25:1545–1548. Pillai S, Bikle DD, Elias PM 1988 1,25-Dihydroxyvitamin D production and receptor binding in human keratinocytes varies with differentiation. J Biol Chem 263:5390–5395. Smith EL, Walworth NC, Holick MF 1986 Effect of 1α,25dihydroxyvitamin D3 on the morphologic and biochemical differentiation of cultured human epidermal keratinocytes grown in serum-free conditions. J Invest Dermatol 86: 709–714. Vinson LJ, Singer EJ, Kowbler WR, Lehman MD, Masurat T 1965 The nature of the epidermal barrier and some factors influencing skin permeability. Tox Appl Pharm 7:7–19. Bikle DD, Chang S, Crumrine D, Elalieh H, Man M-Q, Choi EH, Dardenne O, Xie Z, St-Arnaud R, Feingold K, Elias PM 2004 25-hydroxyvitamin D 1 alpha hydroxylase is required for optimal epidermal differentiation and permeability barrier homeostasis. J Invest Dermatol 122:984–992. Tarutani M, Itami S, Okabe M, Ikawa M, Tezuka T, Yoshikawa K, Kinoshita T, Takeda J 1997 Tissue-specific knockout of the mouse Pig-a gene reveals important roles for GPI-anchored proteins in skin development. Proc Natl Acad Sci USA 94:7400–7405. Vasioukhin V, Degenstein L, Wise B, Fuchs E 1999 The magical touch: genome targeting in epidermal stem cells induced by tamoxifen application to mouse skin. Proc Natl Acad Sci USA 96:8551–8556. Pedrozo HA, Boyan BD, Mazock J, Dean DD, Gomez R, Schwartz Z 1999 TGFβ1 regulates 25-hydroxyvitamin D3 1αand 24-hydroxylase activity in cultured growth plate chondrocytes in a maturation-dependent manner. Calcif Tissue Int 64:50–56. St-Arnaud R, Dardenne O, Prud’homme J, Hacking SA, Glorieux FH 2003 Conventional and tissue-specific inactivation of the 25-hydroxyvitamin D-1α-hydroxylase (CYP27B1). J Cell Biochem 88:245–251. Terpstra L, Prud’homme J, Arabian A, Takeda S, Karsenty G, Dedhar S, St-Arnaud R 2003 Reduced chondrocyte proliferation and chondrodysplasia in mice lacking the integrin-linked kinase in chondrocytes. J Cell Biol 162:139–148. Gerstenfeld LC, Kelly CM, Von Deck M, Lian JB 1990 Comparative morphological and biochemical analysis of hypertrophic, non-hypertrophic and 1,25(OH)2D3 treated non-hypertrophic chondrocytes. Connect Tissue Res 24: 29–39. Kato Y, Shimazu A, Iwamoto M, Nakashima K, Koike T, Suzuki F, Nishii Y, Sato K 1990 Role of 1,25-dihydroxycholecalciferol in growth-plate cartilage: inhibition of terminal differentiation of chondrocytes in vitro and in vivo. Proc Natl Acad Sci USA 87:6522–6526. Lin R, Amizuka N, Sasaki T, Aarts MM, Ozawa H, Goltzman D, Henderson JE, White JH 2002 1α,25-dihydroxyvitamin D3 promotes vascularization of the chondro-osseous junction by stimulating expression of vascular endothelial growth factor and matrix metalloproteinase 9. J Bone Miner Res 17:1604–1612.
CHAPTER 8
Vitamin D–Binding Protein CHRISTOPHER J. LAING AND NANCY E. COOKE Departments of Genetics and Medicine, University of Pennsylvania, Philadelphia, Pennsylvania
I. Introduction II. Vitamin D Binding Protein and Its Gene III. Functional Features of Vitamin D Binding Protein
IV. Concluding Remarks References
I. INTRODUCTION
been cloned and sequenced. The amino acid structure of DBP was first predicted from these cDNA sequences. Human DBP mRNA contains 1690 nucleotides and encodes a 458-amino-acid secreted protein. There is a 16-amino-acid signal peptide, based upon alignment of the sequenced amino terminus of the mature protein [15,16] with the cDNA-predicted sequence of the primary translation product. The predicted mature DBP proteins in rat, mouse, and rabbit are each 460 amino acids long, two amino-acids longer than the human sequence, whereas that of T. scripta is 466 amino acids in length (see Section II,B,3). There is an N-linked glycosylation consensus sequence in human, rat, and mouse but not rabbit or T. scripta DBP. Periodic positioning of 28 cysteine residues predicts a characteristic secondary structure demarcated by internal disulfide bonds and defines the presence of three internally homologous domains [15] (Fig. 1). The crystal structure of DBP has now been described by several groups [17–21]. The three domains of DBP consist entirely of α-helices, 10 in domain 1, nine in domain 2, and four in domain 3. For a more detailed description of the tertiary structure of the DBP molecule, refer to Chapter 9.
Originally identified in 1959 by serum electrophoresis as a polymorphic protein, vitamin D binding protein (DBP) was referred to as the group-specific component of serum (Gc globulin) [1]. At that time its function was not known, although it became useful in population genetics [2] and forensic medicine [3]. DBP is structurally related to albumin and α-fetoprotein, and the DBP gene is a member of the linked albumin and α-fetoprotein gene family. Binding, solubilization, and serum transport of the vitamin D sterols have been considered the predominant functions of DBP, although there is now evidence that DBP may have a spectrum of biological activities, including roles in inflammation and the immune system. The following sections review the general structural and functional features of DBP. A more detailed review of how some of these structural features may explain DBP’s functional activities can be found in Chapter 9. Chapter 10 reviews the relationships among DBP, megalin, and cubilin in the regulation of a key step in vitamin D metabolism.
II. VITAMIN D BINDING PROTEIN AND ITS GENE A. Structural Features Serum DBP is a polymorphic, monomeric, serum α-globulin of approximately 58 kDa, its size being dependent upon glycosylation state [4]. The DNA sequences have been determined for the human [5,6] and rat [7] DBP structural genes. Rat and human DBP genes span 35 and 42 kb, respectively, and both contain 13 exons. The signal sequence of the protein is encoded on the first exon, whereas the last exon contains the entire noncoding 3′-untranslated region. Since the initial reports of human DBP cDNA structure [8,9], DBP cDNAs from the rat [10], mouse [11], rabbit [12], chicken [13], and a turtle, Trachemys scripta [14], have VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND, GLORIEUX
B. Evolutionary Relationships 1. CHROMOSOMAL LOCALIZATION AND LINKAGES
The human DBP gene was assigned to chromosome 4 by analysis of somatic cell hybrids and sublocalized to 4q11-q13 by in situ hybridization to metaphase chromosome spreads [22]. This sublocalization overlaps with the known positions of albumin and α-fetoprotein [23]. These three genes have been assigned to chromosome 13 in the rat [24] and to chromosome 5 in the mouse [11]. The rodent chromosomes encoding DBP are syntenic with human chromosome 4, thus demonstrating conservation of the DBP–α-fetoprotein–albumin linkage in three species. A linkage between DBP and albumin Copyright © 2005, Elsevier, Inc. All rights reserved.
118
CHRISTOPHER J. LAING AND NANCY E. COOKE
L
NH2
T
S
L
V
L
Q S
A
K
S C E
E
K
T
K N K
E
Y D R
G R E
C K
E
F
S H
L
G K
E D
F
A C C A
E
T
L
S
V
V
E
K
D P D C Y D
T
R
T
S
A
L
S
V S
L
V
L
Y
S R
K
V Q E
F
T G S
P
F G A
E
S N S
P
F
P V
G
DOMAIN I
L
E I
A S
V
Y
E R K
C C E L
C M
D N
T
P
E
T
P
F
E Q P Q H K
L
C E
A
F
R K D P
K
E
Y
W
S
T
C C S G V M S
L
Y
S
K
P
T
V C
F
A N Q F M
T G P H
A A E
Y
S
T
N
Y G
Y
S
V
L
L
S
L
P
A Q
S R
L
S N
L
I
K
L
A Q K
A
L
P
L
V D E
L
D A
T
V
K
L
C D N
L
S
T
K
T
K
E Q C C D E
F
K
S
L
K
E R
L
Q
L
K
V R N S
L
T
T
L
L
S
L
C S Q Y
A
A
Y G E
K
K
D E
T
A
H
V S
E C C K
S
L
I
N
T
I
S
E D C M A
K
E
L
P
E H
T
P
A N
DOMAIN II
A M D V K N
T
P
L
F
V C
T
Y
F
P
Q A
A
E
V D P
L
E
L
T
K
V M D K
Y
T
F
E G
L
S
K
L
T
P
E
L V
K
S
N A
K G P
L
L
K
E Q G K D
I
F
S
S
L
E
C A D Y
S
E N
T
F
T
E
Y
K
K
K
I
S C C N S
A
F
D S R K N V
L
K
A
P
L
E
I
D A
I
L
COOH
M P
D V C D P G N E D
V D C C S
T
T
C
F
E
L
S R R
L
F
V
E
P
L
L
A
E R
L
K
A
L
E
K
T
A D P
T H
K L K L
DOMAIN III N S
P
Y C D S
E
L
K N
P
FIGURE 1 Human DBP, primary structure. The amino acid sequence is indicated using single-letter amino acid abbreviations. Fewer than 25% of the amino acids are different in human, mouse, and rat DBP (shaded) [8–11]. The smaller circles mark the positions of amino acid residues present in albumin and α-fetoprotein, but absent in DBP [8]. The three internal domains defined by amino acid sequence similarities are indicated at left. Disulfide bond positions are marked, and the position of a disulfide bond present in domain 1 of DBP, but absent in both albumin and α-fetoprotein, is indicated by a distinctive line between residues 13 and 59. The positions of interruptions to the coding sequence by introns are indicated by arrowheads.
119
CHAPTER 8 Vitamin D–Binding Protein
20 kb upstream from the 5′ end of the α-fetoprotein gene [31,32]. The rat α-albumin gene is located 10 kb downstream from the α-fetoprotein locus, so that the order of the four genes from 5′ to 3′ with respect to the direction of albumin transcription is the same in the rat as in humans. Similar spatial relationships between the genes have also been reported in the mouse [33]. Thus, DBP is a member of a linked multigene family encoding at least four serum proteins, each of which is predominantly expressed in the fetal and/or adult liver (see Section III,A).
has also been noted in horses and in chickens [25,26]. A fourth gene in the DBP/albumin/α-fetoprotein family has been described by two groups and named afamin by one group [27] and α-albumin by the other [28]. α-Albumin is selectively expressed in the liver during the late stages of development, suggesting that it may be a phylogenetic intermediate between α-fetoprotein and albumin and may share some of the functions of these proteins [28,29]. Physical and meiotic mapping has determined that the organization of the gene family in humans is centromere–DBP–albumin–α-fetoprotein–α-albumin– telomere, the transcriptional orientation of DBP being opposite to the other members of the gene family (Fig. 2). The separation between DBP and the remaining members of the gene family is at least 1.5 Mb [30]. In contrast to the substantial distance between DBP and albumin, the albumin and α-fetoprotein genes are tightly linked in the mammalian species studied so far. In the rat and human, albumin is located only 14 to
2. EVOLUTION WITHIN THE GENE FAMILY
There is a striking pattern of structural similarity among the proteins and genes of the DBP, albumin, α-fetoprotein, and α-albumin family [6–12,15,16,20, 21,28]. The four mRNAs and encoded proteins have sequence similarities of approximately 40% at the nucleotide level and 25% at the amino acid level. The positioning of disulfide bond–forming cysteine residues
IN SITU LOCALIZATION DBP ALB/AFP/αALB
GENETIC LINKAGE (cM) IL8
1.9
ALB
1.0
DBP
>1.5 Mb
630 kb αALB AFP ALB
PHYSICAL MAP
(100 kb)
280 MY 300 MY HUMAN CHROMOSOME 4 560-600 MY
TRIPLE DOMAIN
SINGLE DOMAIN
FIGURE 2 The linked DBP multigene family. The chromosomal locations of DBP, albumin (ALB), α-fetoprotein (AFP), and α-albumin (αALB) as determined by in situ localization and recombination analyses are indicated [22,157]. Physical mapping has indicated the arrangement and transcriptional orientations of the genes to be as shown [30]. The evolutionary relationships among the members of the DBP/ALB/AFP-αALB multigene family, as predicted from amino acid and nucleic acid similarities, are indicated in the cladistic diagram [34–37]. Times of divergence are indicated in millions of years (MY). The physical maps and cladogram are not drawn to scale.
120 is nearly identical in all four proteins, predicting a shared three-domain secondary structure. Despite the amino acid sequence similarities between DBP and albumin, and the similarity in individual domain topologies, the orientation of the three domains is very different in the two proteins. This possibly contributes to the unique binding properties of DBP [20,21]. The carboxy-terminal domain of DBP is 124 amino acids shorter that those of albumin and α-fetoprotein. This is due to the loss of the exons homologous to 12 and 13 in the 15-exon albumin and α-fetoprotein genes, resulting in the 13-exon DBP gene [6,7]. With the exception of these two missing exons, the positions at which the introns interrupt the coding regions of the three genes are highly similar (Fig. 1). These striking structural parallels firmly establish the shared ancestry of the DBP, albumin, and α-fetoprotein genes. A number of models have been proposed to account for the evolution of the DBP multigene family and for the shared, triplicated internal domain structure. In one model, the ancestral internal domain is encoded by four exons that subsequently triplicated, creating the present gene encoding a three-domain protein. This domain triplication likely predated vertebrate evolution [34,35]. The separation of a unique DBP gene from this precursor is estimated to have occurred 560–600 million years ago [36,37] (Fig. 2). It has been postulated that albumin duplicated from the ancestral gene and began to diverge about 280 million years ago, just after the time of the amphibian/reptile divergence about 350 million years ago. The absence of a larval α-fetoprotein gene in amphibians and the existence of α-fetoprotein in chickens is consistent with this conclusion [36]. Duplication and subsequent evolution of the α-albumin and α-fetoprotein genes followed this [37]. Evidence of this divergence may be observed in a duplicated albumin gene in the adult amphibian and some reptiles [36,38], and the presence of α-fetoprotein in reptiles. It has been estimated that DBP has evolved at the slowest rate, and that α-fetoprotein and α-albumin have evolved at the fastest rates of the four proteins, although all have evolved at a more rapid rate than proteins of the globin family [36,37]. There has been significant invasion of noncoding sequences by repetitive DNA elements. Intron expansion in the DBP gene has been significantly greater compared with the other family members, providing further evidence of its longer period of separation from the ancestral gene [39]. Regardless of the precise details of this series of evolutionary events, DBP appears to be the oldest member of this ancient and well-conserved gene family. 3. DBP IN VERTEBRATE EVOLUTION
A specific protein carrier for vitamin D compounds has been identified in examples of all major vertebrate
CHRISTOPHER J. LAING AND NANCY E. COOKE
phyla [40–44]. The presence of a serum DBP in some species of cartilaginous fish (Cyprinus carpio), but not in others (Tilapia nilotica), suggests that the emergence of a discrete DBP may have occurred during the evolution of this phylum [41]. Overall the structures of the DBP gene, mRNA, and protein have been well conserved. There is over 70% amino acid homology among human, mouse, rat, rabbit, and T. scripta DBP. Furthermore, the affinity of reptilian, avian, and mammalian DBPs for the major vitamin D metabolite, 25-hydroxyvitamin D3, are all in the nanomolar range. However, despite this high degree of both structural and functional conservation, significant interspecies variations in DBP concentration and in DBP–ligand affinity exist. These variations appear to be more dependent on ecological variables such as geographic distribution and diet, than on phylogenetic distance [43,44]. Only mammalian DBP demonstrates equal affinity for metabolites of vitamin D2 and vitamin D3, suggesting that the ability to efficiently handle vitamin D2 metabolites may have conferred some advantage on emerging mammalian vertebrates [44]. An unusual circumstance has been observed in the emydid turtle T. scripta, where DBP is also able to transport thyroxine on a unique binding site. The DBP of this species constitutes the major carrier of vitamin D sterols, but its serum concentration is regulated by thyroxine. Comparative studies suggest that this dual binding is unique to members of Emydidae [45,46]. The case of emydid turtles notwithstanding, DBP is structurally and functionally highly conserved among vertebrates, and this, along with the apparent evolutionary responsiveness of its functional features to ecological factors, suggests an important physiological role for this protein.
C. Polymorphisms of DBP/Gc-Globulin Protein and Gene DBP (or group-specific component, Gc) was originally characterized in humans by serum electrophoresis as the product of two autosomal, codominant alleles (Gc1 and Gc2). Isoelectric focusing allowed the further characterization of slow (Gc1S) and fast (Gc1F) subtypes of Gc1 [47,48], resulting in six common phenotypes. The protein isoforms produced by the Gc1F and Gc1S alleles represent a mixed population containing or lacking a single N-acetylneuraminic acid on a threonine residue at position 420 [49]. In contrast, the Gc2 allele produces a single protein that contains a nonglycosylated lysine at position 420. The DBP gene contains an Alu middle repetitive element located at the end of DBP intron 8 resulting in four gene polymorphisms created by differences in the size of the Alu poly(A) tract. Each of these polymorphisms is equally distributed among
121
CHAPTER 8 Vitamin D–Binding Protein
the three common alleles [50]. Although the major alleles can be distinguished at the DNA level by restriction digestions or PCR-single-strand conformation polymorphism analysis [51], protein electrophoretic techniques remain the most usual initial approach to assigning protein isoform phenotypes. Although commonly used for forensic purposes in the past [3,52], DBP phenotypying is now more frequently used to map population dynamics and to assess phenotypic susceptibility to disease (see later discussion). In addition to the three common alleles, there are also more than 124 rare variant alleles described worldwide. The geographic occurrences of these variants often correspond to patterns of human population migrations and thus are of anthropological interest [2,53–56]. The molecular bases for some of these rare variants have been determined by sequencing of exons amplified by the polymerase chain reaction [57–59]. A number of genetic studies have suggested the presence of a DBP-null allele [60,61], but these have not been clearly distinguished from low-expressing “pseudo-silent” alleles [62]. The lack of a naturally occurring DBP-null homozygote in the human population suggested that such a genotype would be embryonically lethal. A homozygous, DBP-deficient mouse line has, however, been generated that is viable and fertile [63]. DBP polymorphisms have also been identified in nonhuman primates [64,65], rodents [66], a variety of ungulates [25,66–70], domestic cats [71], marsupials [72], and birds [26,73,74]. The high degree of polymorphic variation in the human population and its widespread occurrence among vertebrates make the DBP locus among the most polymorphic known. There is some suggestion that differences in serum concentration of DBP may be attributable to phenotypic isoform [75], and that different isoforms of DBP may exhibit affinity differences for vitamin D metabolites [76,77]. However, the biological significance, if any, of these differences remains to be established. DBP isoforms have also been reported to be associated with differences in susceptibility to a wide variety of diseases and syndromes, including multiple sclerosis [78,79], autoimmune thyroid disease [80], chronic obstructive pulmonary disease [81,82], low bone mineral density in women [83], and male osteoporosis [84]. A protein that is structurally related to DBP, and thus a possible variant of DBP, has been identified as a potential serum marker for prostate cancer [85]. The nature of this protein and its relationship to DBP await its isolation and further characterization. The associations between DBP isoform and disease are, for the most part, limited to specific geographic areas or ethnic groups. It is not clear whether they apply to the human population in general, or what, if any, the biological bases and implications of these associations may be.
III. FUNCTIONAL FEATURES OF VITAMIN D BINDING PROTEIN A. Synthesis and Turnover of DBP The circulating DBP concentration in adult mammals and birds is in the micromolar range [43,44]. DBP concentration increases in the postnatal period and can be influenced by a number of factors and processes. Most studies have failed to link vitamin D status or disturbances of mineral homeostasis with alterations to DBP concentration [86–89], although there is some evidence that there may be a correlation between ligand concentration and DBP concentration in the elderly [90]. DBP concentration increases during pregnancy [91,92], and under the influence of female sex hormones in humans and birds [93–95]. In rodents, DBP concentration increases in response to testosterone and is reported to be higher in males than in females [96]. Malnutrition results in a decrease in circulating DBP concentration [97], as does liver failure [91,98], pronounced proteinuria [99,100], and increased actin–DBP complexing resulting from tissue necrosis or damage (see Section III,C,2). The specific mechanisms by which these factors and processes regulate DBP concentration are not well understood. Like albumin and other serum proteins, DBP is produced primarily in the liver, and the increase in postnatal DBP concentration to adult levels is thought to reflect maturation of the liver protein synthesis processes [86,101]. It has been estimated that 10 mg/kg/day of DBP is produced in humans [102]. The relatively low concentration of hepatic DBP mRNA suggests that the DBP transcript, like that of albumin, may be very stable and exhibit a long cytosolic half-life [10]. Spliced DBP transcripts have been identified in a range of extrahepatic sites, although at concentrations approximately 100- to 1000-fold lower than those found in the liver [101]. The role of these extrahepatic transcripts is not known, as they are not thought to be in enough abundance to significantly influence the circulating concentration of DBP, and indeed it has not yet been determined whether they are translated. DBP production is up-regulated following its increased clearance in response to trauma or hepatic necrosis [103–105] (see Section III,C,2), raising the possibility that its synthesis is regulated by a positive feedback mechanism. A range of growth factors and steroids have been demonstrated to influence DBP production by cells in culture [106,107], although the mechanisms by which these factors regulate DBP synthesis are largely unknown. It is thought that the postnatal regulation of DBP production is largely at the level of gene transcription, as is the case for α-fetoprotein and albumin [108]. Recent work has identified a locus-control
122 region for the human DBP gene, located within its large first intron. The expression of a full-length human DBP transgene in mice is abolished by deletion of this region [109, and Hiroki et al., submitted, 2004]. Expression of the DBP gene is partially under the control of HNF-1α and HNF-1β, which bind to three cis elements near its proximal promoter. HNF-1α has a major stimulatory effect whereas HNF-1β is a transdominant inhibitor of HNF-1α stimulation [110]. Both are members of the POU-homeodomain family of transcription regulators [111]. Protein kinase C has been implicated in the regulation of α-fetoprotein and albumin gene expression in the liver; however, it has not been demonstrated to regulate DBP expression [112]. The regulatory mechanisms for DBP clearance from the circulation are not known. Clearance of DBP from the plasma in the rabbit, like that of albumin, is described by a multiexponential curve with an estimated t1/2 of 1.7 days. This is less than half that of albumin with a t1/2 of 5 days, but over five times faster than that of its major ligand, 25-hydroxyvitamin D3 (25(OH)D3), suggesting that ligand recycling is likely [113]. DBP is primarily cleared by the kidneys as a result of its uptake by megalin in the proximal tubule epithelium [102,114–116] (see Section III,B,3 of this chapter and Chapter 10 of this volume). DBP–actin complexes are cleared up to three times more rapidly than unliganded DBP, primarily by hepatic filtration [117,118] (see Section III,C,2). Its association with 25(OH)D3, on the other hand, has not been reported to affect the clearance rate of DBP [113]. It has been suggested that glycosylation state may be responsible for altering clearance mechanisms of other plasma proteins, including sex hormone–binding globulin (SHBG) [119]; however, differences in glycosylation of DBP (see Section II,C) have not been reported to influence its clearance from the circulation [113]. The structural similarities between, and shared ancestry of albumin, α-fetoprotein, and DBP originally suggested overlapping regulatory mechanisms of production and turnover. It is now clear, however, that regulation of DBP synthesis and catabolism may be quite independent, and these processes await further characterization.
B. Molecular Interactions DBP associates with a number of biological molecules, including vitamin D sterols, actin, fatty acids, and various cell membrane components. These interactions will be discussed in this section, and their proposed physiological significance considered in Section C. Some of these interactions are summarized in Table I.
CHRISTOPHER J. LAING AND NANCY E. COOKE
TABLE I Features and Binding Characteristics of DBP Features Isoelectric point Electrophoretic migration Size Plasma concentration Plasma half-life Daily production rate Altered plasma levels Increased Decreased Vitamin D sterol binding Plasma capacity Normal sterol occupancy Affinity (KD) ~nanomolar ~micromolar Clearance by Actin binding Plasma capacity Normal actin occupancy Affinity (KD) DBP-actin complexes Clearance by Cell associations Renal epithelium Cell “receptor” Internalization Physiology Lymphocytes Cell “receptor” Internalization Physiology Monocytes/macrophages Cell “receptor” Internalization Physiology Neutrophils Cell “receptor” Internalization Physiology
4.5–4.8 Post-albumin, inter-α-globulin 58 kDa, single-chain glycoprotein 4–8 µM (232–464 mg/liter) 2.5–3.0 days ~10 mg/kg Estrogen, pregnancy Nephrotic syndrome, liver disease, malnutrition mol/mol (2.4 mg D sterol/liter) <5% 25OHD, 24,25(OH)2D 1,25(OH)2D, vitamin D Renal proximal tubule epithelial cells mol/mol (270 mg/liter) Trace—? mg/liter Nanomolar Seen with tissue injuries, inflammation Hepatic phagocytic cells, sinusoidal endothelium
Megalin, cubilin Yes Protein clearance, delivery of ligand ?, IgG–Fc receptor complex ? ?, conversion to DBP-maf ? ? ?, activation of phagocytic function Chondroitin sulfate proteoglycan ?, no ?, enhancement of C5a chemotaxis
1. VITAMIN D LIGANDS
As well as having weak, nonspecific associations with lipoproteins and albumin, vitamin D sterols in the extracellular fluid form a specific, high-capacity, high-affinity association with DBP [120–122]. All naturally occurring and synthetic vitamin D sterols are bound by a single binding site, but with variable affinity [reviewed in 123–126]. Among the naturally occurring vitamin D metabolites, DBP shows the greatest affinity for 25(OH)D3 and 24,25-dihydroxyvitamin D3 (24,25(OH)2D3), followed by 1,25dihydroxyvitamin D3 (1,25(OH)2D3) and parent
123
CHAPTER 8 Vitamin D–Binding Protein
vitamin D3 [4,113,42,127,128] (Table I). The vitamin D–binding site of DBP was first localized to the amino terminus of the protein using biochemical techniques such as proteolysis [129,130] (Fig. 1). Molecular modeling and labeling studies have identified important contact points and molecular orientations of a variety of vitamin D sterols within the vitamin D–binding pocket [124]. The vitamin D–binding site of DBP imposes rigid steric restrictions on its ligands. The binding pocket of the nuclear vitamin D receptor shares this feature, which contrasts with many other ligand–protein interactions such as the estrogen receptor-binding site, within which ligands are relatively free to rotate [125]. However, the actual orientations of vitamin D ligands within the DBP and vitamin D receptor-binding pockets are very different. In the latter case, the sterol is completely surrounded by the receptor protein, whereas there is indication that the DBP binding site remains “open” and that part of the sterol molecule protrudes outside of the pocket [124–126,131] (see Chapter 15). Elucidation of the crystal structure of 25(OH)D3liganded DBP has provided a more detailed understanding of the molecular interaction between these two molecules [19,20]. These structural features will pave the way to understanding how different ligands are oriented within the vitamin D–binding space and are discussed in further detail in Chapter 9 of this volume. SHBG is known to bind estradiol and C19-androgen metabolites in different orientations resulting in distinct conformational changes in the binding protein, and possibly providing a structural explanation for differences in the biological activities of these ligands [132,133]. Whether similar binding differences are observed among the various vitamin D ligands of DBP remains to be fully determined. Vitamin D binding increases the chemical [134] and thermal [122] stability of DBP, diminishes its isoelectric point [134,135], and is now known to result in rotation of DBP domain I structures [20]. The association of vitamin D ligands with DBP is thought to facilitate their dispersal throughout the body, and their delivery to sites of metabolism and action (see Section III,C,1). 2. ACTIN
The ability of DBP to form a complex with actin was first observed during cytosolic contamination of preparations of DBP-containing extracellular fluid [136,137]. DBP binds globular, but not filamentous actin in a specific, saturable, high-affinity manner (Table I). It possesses a single binding site for actin that was originally localized to the C-terminal end of the molecule [130,138] (Fig.1), but which more recently, through elucidation of the crystal structure of DBP, has been identified as involving components
of all three DBP domains [17–19,21] (for more detail see Chapter 9). Binding of actin involves rotations of the domains and flexion of some of the intradomain extended structures within the DBP molecule [17]. The extent, if any, to which the binding of actin influences the affinity of DBP for vitamin D, or vice versa, remains unclear, although evidence thus far suggests that the binding sites are structurally and functionally independent [17,19,21,139]. DBP–actin complexes have been identified in normal human plasma and have been found to increase in concentration during pregnancy and following incidents of tissue damage and necrosis. This has prompted the proposal that DBP acts in concert with other actin-binding proteins, particularly gelsolin, as a component of an extracellular actin scavenging system [140–143] (see Section III,C,2). 3. CELL ASSOCIATIONS
DBP can associate with the plasmalemma of a wide variety of cells, including renal proximal tubule epithelium [114], lymphocytes [144,145], monocytes [144–147], neutrophils [148], yolk-sac endodermal cells [149], trophoblastic cells [150,151], and spermatocytes [152]. Despite the demonstration of DBP transcripts in a number of tissues other than liver (see Section III,A), it is generally thought that this cellsurface DBP is more likely to be acquired, at least partially, from the extracellular fluid, rather than completely produced de novo [147]. A search for a specific cellsurface DBP receptor has been thus far unsuccessful, although various nonspecific cell-surface structures have been found to interact with DBP (Table I). In the renal proximal tubule epithelium, megalin, a member of the low-density lipoprotein receptor gene family, and its co-receptor cubilin are responsible for binding a wide variety of molecules, including DBP [114,115,153]. DBP is thought to bind to the IgG-Fc receptor complex of lymphocytes [154], and an elastase-sensitive chondroitin sulfate proteoglycan has been implicated as a nonsaturable DBP-binding structure on neutrophil surfaces. This chondroitin sulfate, although not specifically identified, is thought to bind multiple DBP molecules [148,155]. Although a surface receptor on monocytes has yet to be identified, the existence of a monoclonal antibody for DBP that does not recognize monocyte-associated protein [146,156], suggests a discrete structural interaction of DBP with the cell surface. The fate and function of cell-associated DBP are not fully understood in most instances, although megalin-complexed DBP is thought to deliver 25-hydroxyvitamin D3 to intracellular metabolic enzymes (see Section III,C,1,b), and neutrophilassociated DBP is thought to be involved in regulation of chemotaxis (see Section III,C,2).
124 C. Physiological Roles of DBP DBP was originally thought to be an acute phase reactant until it was found to bind vitamin D sterols and the identity of Gc-globulin and DBP were confirmed [122,157,158]. However, the subsequent discovery of its range of seemingly unrelated ligand and cell associations has prompted the suggestion that DBP is a multifunctional protein [159]. Although actions of DBP in addition to the physiological economy of vitamin D have been well documented (see Section III,B), many of the biological roles suggested for DBP are speculative, or supported by circumstantial evidence. The reported functions of DBP fall into two main categories: those involving the physiological economy and function of vitamin D and calcium homeostasis, and those involved in inflammatory processes and the immune response. It is still not clear whether these functions of DBP are independent or are physiologically linked. 1. DBP AND THE PHYSIOLOGICAL ECONOMY FUNCTION OF VITAMIN D
AND
a. The Role of Plasma Proteins in Sterol Transport and Function The sterols, and their biologically functional subset, the steroid hormones, are a group of biological molecules that are structurally related to cholesterol. They include the sex hormones, corticosteroids, thyroid hormones, retinoids, and vitamin D molecules. These molecules function in a diverse range of tissues, which necessitates distribution throughout the body. Sterols are highly lipophilic, and as such disperse poorly in the aqueous environment of the extracellular space. They do, however, readily and rapidly diffuse across cell membranes. These shared limitations necessitate transport mechanisms that allow the dispersal of sterols through the circulation, and their controlled access to cells. These requirements are met partially by nonspecific carriers such as albumin, but more importantly by a heterogeneous group of ligandspecific protein carriers. In addition to DBP, this group of proteins includes ex-hormone binding globulin (SHBG), corticosteroid-binding globulin (CBG), thyroidbinding globulin (TBG), and retinol-binding protein (RBP). These carrier proteins, while structurally unrelated, share many common functional features, reflective of the shared biochemical and physiological features of their sterol ligands. This section summarizes these shared features, along with the changing views of the functional roles of carrier proteins in steroid physiology. The specific case of DBP and vitamin D transport is considered in Section III,C,1,b. The vast majority of circulating steroids (over 90%, and in many cases up to 99%) are found in association
CHRISTOPHER J. LAING AND NANCY E. COOKE
with their specific carrier proteins. The free hormone theory was developed to describe the manner in which these proteins disperse their ligands throughout the extracellular space and make them available to target cells [160,161]. According to this theory, the ease with which the steroids can diffuse across cell membranes means that it is the small, free (or unbound) fraction of steroid in the extracellular fluid that is biologically available to cells. As such, sterol carrier proteins are thought to function by sequestering their ligands and making them unavailable to cells (and hence ensuring their continued transit in the circulation). The association between a particular protein carrier and its ligand obeys the law of mass action, where bound and “free” species exist in the following equilibrium: Ligand + BP ↔ Ligand−BP The factors that determine the proportion of free (or bioavailable) sterol are the concentration of each of the equilibrium species and the affinity of the ligand for its carrier (as well as those of other ligands competing for the same site). The metabolic clearance rates of the sex steroids and the corticosteroids are inversely proportional to their affinities for their respective specific plasma binding proteins [162,163], providing experimental support for the free hormone theory. However, the free hormone hypothesis has been challenged by a number of observations that suggest it may not fully explain the relationship between sterols, their binding proteins, and target cells. One of the most important is the direct, tissue-specific association of binding proteins with cell membranes. A specific cell membrane receptor for SHBG has been well described in a number of cell types. The binding of specific ligands to the SHBG–receptor complex initiates second-messenger cascades resulting in alterations to cell growth and differentiation, and secretion of specific proteins [reviewed in 164–166]. Thus SHBG not only acts to deliver its ligands to target cells, but directly participates in a biologically functional process that is either independent of or that augments the classic genomic actions of steroid hormones [reviewed in 167]. It has been recognized that SHBG expression occurs in a range of extrahepatic sites, many of which are sex-steroid target cells and coexpress the SHBG receptor. This suggests that SHBG may act as an autocrine or paracrine factor [168,169]. Although less well described, and more hotly contested, evidence has been presented which suggests that CBG and RBP may also bind to cell membrane receptors in a tissue-specific fashion. The physiological consequences of such associations are still largely speculative [170–174].
CHAPTER 8 Vitamin D–Binding Protein
Our original views of sterol binding proteins ascribed to them a primarily passive role in the retention of their ligands in the circulation, thus creating a pool from which controlled cellular uptake could occur. They were thought to maintain adequate stores of their ligands in the face of intermittent supply, and to buffer against potentially toxic overload. This role for the binding proteins is certainly physiologically relevant. However, more recent descriptions of receptors and cell interactions for these proteins suggest a more diverse and active function in the physiological trafficking of sterols. The functional similarities within this structurally diverse group of binding proteins suggest that some of these themes may also be shared by DBP. b. The Case of Vitamin D–Binding Protein The free hormone theory has been suggested to describe the DBP–vitamin D–cell interaction [91,160,175], and a number of lines of experimental evidence support this idea. The metabolic clearance rates of vitamin D metabolites and analogs are inversely proportional to their affinity for DBP [123,176], and DBP has been shown to limit the uptake and action of 1,25(OH)2D3 in cells grown in vitro [177,178]. Analogs of 1,25(OH)2D3 with a low affinity for DBP stimulate vitamin D– responsive gene expression more rapidly than those with higher affinity [179], and mice not expressing DBP demonstrate a more rapid genomic response to 1,25(OH)2D3 than their wild-type littermates [63]. It has been proposed that the pathophysiology of vitamin D toxicity may involve the dislodging from DBP and resultant increased bioavailability of 1,25(OH)2D3 by the more abundant and avidly bound 25(OH)D3 [63,180]. Several experimental observations, however, suggest that the free hormone hypothesis may not fully explain the role of DBP in vitamin D transport and function. Its high plasma concentration and its high ligand affinity (Table I) ensure that DBP binds between 85% and 98% of each circulating vitamin D sterol, depending upon the metabolite, while the remainder is considered “free” or in association with low-affinity carriers [91,127,181]. On the other hand, only 5% of DBP molecules are occupied by a vitamin D sterol ligand [130]. Using a range of estimates of “free” 1,25(OH)2D3 concentration [91,127,176,181,182], the concentration of bioavailable 1,25(OH)2D3 in the extracellular fluid is still approximately an order of magnitude less than the dissociation constant of the intracellular vitamin D receptor. Thus if only the free fraction of hormone were available for cellular uptake, there would be very little likelihood of high receptor occupancy. This suggests that proteinmediated ligand uptake by cells may play an important role in the physiology of vitamin D metabolism and function.
125 Megalin–DBP complexes are internalized by the renal proximal tubular cells, and DBP is thought to undergo proteolysis after receptor and ligand scavenging. This results in the delivery of vitamin D sterols to intracellular binding sites such as metabolic enzymes and the vitamin D receptor, via specific cytosolic chaperones, members of the heat-shock protein-70 family [183]. This process constitutes an important step in the conversion of 25(OH)D3 to the biologically active 1,25(OH)2D3 [114,116] (see Chapter 10). The association of DBP with other cell types may involve further specific or nonspecific receptors (see Section III,C,2). These receptors may also be involved in ligand internalization, or may regulate cell signaling as has been described for cell surface receptors for SHBG (see Section III,C,1,a). 1,25(OH)2D3 and 24,25(OH)2D3 can exert effects in a variety of cells via rapid, nongenomic pathways thought to involve novel cell membrane receptors [184,185] (see Chapter 23). This is a theme that has also been described for sex hormones, which can stimulate cell-signaling cascades directly via sex-steroid membrane receptors, or via SHBG receptors [167]. It is tantalizing to speculate that some of the non-genomic actions of vitamin D metabolites may be mediated via interactions of DBP with cell surface receptors. The demonstration of DBP transcripts in a range of extrahepatic tissues (see Section III,A) suggests that DBP produced at these sites may be involved in functions not directly related to the dispersal of vitamin D ligands in the circulation and may act as an autocrine or paracrine factor as has been demonstrated for SHBG (see Section III,C,1,a). Further work is required to elucidate the mechanisms by which DBP regulates the cellular uptake, metabolism, and biological activities of its ligands. However, it is clear that whether the bioavailability of vitamin D is dependent upon the “free hormone” model, or protein-mediated pathways, or a combination of the two, DBP plays an important role in regulating the physiological economy of its sterol ligands. When fed a normal diet, mice in which the DBP locus has been inactivated do not demonstrate any specific phenotype that distinguishes them from wild-type littermates [63]. This “mild” phenotype is in contrast to mouse models in which the genes for the vitamin D receptor [186] or enzymes of metabolism of vitamin D [187,188] have been inactivated. Rather mild phenotypic differences have also been observed in mouse models and natural human mutations resulting in no or low expression of CBG [189,190], TBG [191,192], and RBP [193], suggesting that the mechanisms by which sterols arrive at and enter cells are likely to be tissue specific and to possess a certain element of redundancy. It has been suggested that the nonspecific carrier
126 proteins, such as albumin, may play a more than trivial role in the transport and processing of sterols [194]. However, in the case of both DBP–null [63] and RBPnull [193] mice, the absence of specific binding protein renders the animal more susceptible to derangements associated with abnormal intake or production of the respective ligand. The intermittent nature of vitamin D production or intake suggests that an important evolutionary advantage may be conferred upon organisms possessing an efficient plasma binding protein for the vitamin D sterols. That interspecies variations in DBP concentration and 25OHD3 affinity correlate more closely with ecological factors than with phylogenetic distance (see Section II,B,3) suggests that the properties of this protein have been important determinants of the conservation or activity of its vitamin D ligands during vertebrate evolution. Furthermore, there is some evidence to suggest that the concentration of DBP may be altered during times of increased calcium demand, such as at the initiation of lactation [44]. Possible links among malnutrition, DBP concentration, and the development of rachitic syndromes [97,195,196] also support the important role played by DBP in the regulation of the physiological economy of its vitamin D ligands. 2. THE ROLE OF DBP IN INFLAMMATORY PROCESSES SYSTEM
AND IN THE IMMUNE
Inflammation is a complex tissue reaction to physical injury, to infection, or to a local immune response and is characterized by changes to the local vasculature and the accumulation of fluid, biologically active proteins, and white blood cells. The pathogenesis and development of both local and systemic inflammation is intricately linked to these vascular and cellular responses. DBP has been associated with a number of steps within the inflammatory process, including the prevention of thromboembolic events in the microvasculature by actin, the stimulation of chemotaxis by phagocytic neutrophils, and the activation and stimulation of phagocytic function by macrophages. In this section, the evidence for a physiological role for DBP in the inflammatory process and in immune function is discussed. Tissue damage or necrosis, either due to or as a sequel to a primary insult, is an early component of the general inflammatory process and results in the release of intracellular proteins such as filamentous actin. Free actin in the circulation results in the development of thromboemboli in the microvasculature and the disruption of normal coagulation events [197]. This leads to tissue ischemia and multiple organ dysfunction syndrome. Actin may also be directly toxic to pulmonary
CHRISTOPHER J. LAING AND NANCY E. COOKE
endothelial cells [198]. DBP and another plasma protein, gelsolin, have been hypothesized to act in a coordinated fashion to dismantle actin filaments, sequester monomeric actin, and facilitate its clearance from the circulation [18,21,117,118,197]. In support of this hypothesis, circulating DBP concentrations are decreased and DBP–actin complexes increased in a variety of circumstances associated with tissue necrosis or cell disruption, such as fulminant hepatic necrosis [104,105,199], septic shock [200], and multiple trauma [103,201,202]. This reduced concentration of DBP is thought to be the result of the rapid removal of DBP–actin complexes in the liver [104,203] (see Section III,A). During recovery from liver disease or tissue trauma, plasma DBP concentration rises above the normal range, perhaps indicating a compensatory stimulation of DBP synthesis, which has been suggested to result from cytokine activity [102–105]. The priority for maintenance of plasma DBP concentration by the liver is similar to that of prothrombin index, and higher than that for plasma albumin concentration [98]. Nevertheless, the actin scavenging system appears to be saturable [141] and an association between low plasma DBP concentration and the development of multiple organ dysfunction has been documented in patients with fulminant hepatic necrosis and following severe injury [199,202]. An early response to tissue damage is the recruitment and activation of phagocytic cells. In the immune system, chemotaxis involves the directional migration of cells in response to an extracellular cue, usually in the form of a chemical gradient. A wide variety of chemoattractants have been identified, with varying degrees of specificity. A serum enhancer of C5a-stimulated chemotaxis of leukocytes was subsequently identified to be DBP. The mechanism(s) by which DBP enhances neutrophil response to C5a are unknown, but several studies document that such an effect exists [155,204–207]. Preincubation of DBP with neutrophils is sufficient for enhanced C5a-stimulated chemotaxis, which is apparently mediated via DBP–cell surface proteoglycan interaction [148,205,207] (see Section III,B,3). Although cell membrane–associated proteoglycans have been suggested to facilitate the uptake of their ligands [208], DBP bound to the surface of neutrophils does not appear to be internalized [155]. Some cell-surface proteoglycans are known to possess cytoplasmic domains that interact with components of the cytoskeleton [208], raising the possibility that the DBP–chondroitin sulfate complex is involved in cell signaling. It is now known that a neutrophil-associated serine protease, elastase, establishes the steady-state binding of DBP to neutrophil surfaces by cleaving and thus regulating the number of available chondroitin-sulfate binding sites. The establishment of steady-state binding is essential
127
CHAPTER 8 Vitamin D–Binding Protein
for cochemotactic activity, suggesting that the mechanism by which DBP acts as a cochemotaxin may involve dynamic proteoglycan signaling in the cytosol, or that localization of the DBP–receptor complex to active areas of the cell membrane is required [155]. Antigenic stimulants cause the activation of phagocytic function in leukocytes and other cells via several pathways. In one such pathway, cell death and inflammation increase the activity of membrane phospholipases, resulting in the accumulation of lysophospholipids and other phospholipid degradation products, which are potent stimulators of macrophage activation [209,210]. It has been demonstrated ex vivo and in vitro that this activation may occur through an intermediate macrophage-activating factor, the partially deglycosylated DBP (DBP-maf). DBP is converted to DBP-maf via the inducible, stepwise actions of B-lymphocytic β-galactosidase and T-lymphocytic sialidase [211–213]. DBP-maf activity is specific to Fcγ-mediated phagocytosis and does not appear to be involved in complementprimed phagocytic processes that can be stimulated by such factors as lipopolysaccharide [210]. DBP-maf activated macrophages demonstrate significant tumoricidal activity [214–216]. The specific stimulants for DBP-maf generation and the particular phagocytic pathways that it activates suggest that the principal role of DBP-maf may be in the immune response to neoplasia. The exact mechanism by which DBP-maf acts is unknown; however, the complete deglycosylation of DBP-maf by the removal of the terminal alphaN-acetylgalactosamine (GalNAc) moiety results in diminished macrophage-activating activity. Furthermore, the accumulation of tumor-derived alpha-N-acetylgalactomanidase in advanced cancer patients has been linked to immunosuppression. This suggests that DBPmaf may stimulate Fcγ translocation to the cell surface via GalNAc-specific lectin on the macrophage surface [217–219]. DBP maf has also been reported to stimulate bone-resorptive activities of osteoclasts in vitro [220]. It has been proposed that a defect in lymphocytic β-galactosidase and/or some other deficiency in DBPmaf generation may be involved in the pathogenesis of some osteopetrosis syndromes [221–223]. Most recently, it has been suggested that DBP–maf may have direct antiangiogenic effects on endothelial cells and be able to inhibit the angiogenic actions of a number of tumor-derived growth factors. This activity is proposed to operate via the CD36 receptor, expressed on both macrophages and endothelial cells [224,225], although the nature of this proposed activity awaits more thorough characterization. Given the seemingly diverse and integral roles played by DBP in inflammatory and immune processes, it was surprising that mice in which DBP expression had
been disrupted were viable [63]. Indeed, DBP-null mice show no obvious phenotype that can be related to deficient immune function. No differences in phagocytic cell recruitment in response to intraperitoneal treatment with thioglycollate medium [226] or to infection with Listeria monocytogenes or Toxoplasma gondii [White et al., submitted, 2004.] could be attributed to the inactivated DBP allele. DBP-null mice do not demonstrate osteopetrotic lesions [63], causing some uncertainty about the obligatory nature of DBP-maf in osteoclastic function. The reasons for these apparent discrepancies between in vitro or ex vivo actions of DBP and observations in the DBP-null mouse are as yet unclear. The processes of immunity and inflammation reflect complex genetic determinants. Strain differences in peritoneal recruitment of leukocytes have been identified in mice [226], and species differences are likely to be more profound still. Furthermore, phagocyte site specificity, as well as the nature of the target particles, may differentially determine the nature and regulation of phagocytic processes [227]. Therefore, the lack of observed phenotypic difference between DBP-null and wild-type mice in relation to inflammation and immunity may reflect species-, strain-, tissue-, and/or pathologyspecific factors. Immune defects related to the inactivated DBP allele, if any, may not become apparent without producing the appropriate specific challenges to the immune system. Alternatively, it is possible that the “non–vitamin D binding” actions of DBP may not represent an obligate role for DBP in inflammation and immunity, or may be biologically redundant. A third possibility is that they may be involved in an as-yetunidentified integrated physiological role for DBP. Further experimentation is required to resolve these points.
IV. CONCLUDING REMARKS Several important technical steps have advanced our understanding of the structural and functional features of DBP. The elucidation of the sequences of both the full-length DBP gene and its messenger RNA in a variety of vertebrates has revealed the relationship between this protein and the albumin/α-fetoprotein multigene family, and their place in vertebrate evolution. Comparisons of the primary structures of these proteins have identified differences in DBP that pointed toward a structural explanation for its unique functional activities. Descriptions of the crystal structure of DBP in complex with two of its ligands, 25(OH)D3 and actin, have extended this explanation to the tertiary structural level. The generation of a mouse line with an inactivated DBP locus has provided an important tool for the study
128 of the integrated physiology of DBP. It has enabled the demonstration of the important role played by DBP in the turnover and actions of its vitamin D ligands. Further investigation of non–vitamin D functions of DBP using this mouse model are underway. While our understanding of its actions and their molecular bases is increasing, further work is required to clarify the physiological role(s) of DBP in the whole organism. Functional similarities reported among the other sterol binding proteins may hold important clues for future investigations of the physiology of DBP.
References 1. Hirschfeld J, Jonsson B, Rasmusson M 1959 Immune electrophoretic demonstration of qualitative differences in human sera and their relation to the haptoglobins. Acta Pathol Microbiol 47:160–168. 2. Cleve H, Constans J 1988 The mutants of the vitamin D–binding protein: more than 120 variants of the Gc/DBP system. Vox Sang 54:215–225. 3. Westwood WA, Werret DJ 1986 Group specific component: a review of the isoelectric focusing methods and auxiliary methods available for the separation of its phenotypes. Forensic Sci Int 32:135–150. 4. Haddad JG, Walgate J 1976 25-Hydroxyvitamin D transport in human plasma. Isolation and partial characterisation of calcifidiol-binding protein. J Biol Chem 251:4803–4809. 5. Braun A, Kofler A, Morawietz S, Cleve H 1993 Sequence and organization of the human vitamin D–binding protein gene. Biochim Biophys Acta 1216:385–394. 6. Witke FW, Gibbs PEM, Zielinski R, Yang F, Bowman BH, Dugaiczyk A 1993 Complete structure of the human Gc gene: differences and similarities between members of the albumin gene family. Genomics 16:751–754. 7. Ray K, Wang XK, Zhao M, Cooke NE 1991 The rat vitamin D binding protein (Gc globulin) gene: structural analysis, functional and evolutionary correlations. J Biol Chem 266: 6221–6229. 8. Cooke NE, David EV 1985 Serum vitamin D binding protein is a third member of the albumin and α-fetoprotein gene family. J Clin Invest 76:2420–2424. 9. Yang F, Brune JL, Naylor SL, Cupples RL, Naberhaus KH, Bowman BH 1985 Human group-specific component (Gc) is a member of the albumin family. Proc Natl Acad Sci USA 82:7994–7998. 10. Cooke NE 1986 Rat vitamin D binding protein: determination of the full length primary structure from cloned cDNA. J Biol Chem 261:3441–3450. 11. Yang F, Bergeron JM, Linehan LA, Lalley PA, Sakaguchi AY, Bowman BH 1990 Mapping and conservation of the group-specific component gene in mouse. Genomics 7: 509–516. 12. Osawa M, Tsuji T, Yukawa N, Saito T, Takeichi S 1994 Cloning and sequence analysis of cDNA encoding rabbit vitamin D binding protein (Gc globulin). Biochem Mol Biol Int 34:1003–1009. 13. Burnside J, Sofer L 1999 Gallus gallus vitamin D binding protein cDNA. GenBank direct submission; accession number AF121350.
CHRISTOPHER J. LAING AND NANCY E. COOKE
14. Hunt JL, Licht P 1998 Identification and structural characterization of a novel member of the vitamin D binding protein family. Comp Biochem Physiol 121B:397–406. 15. Schoentgen F, Metz-Boutigue MH, Jolles J, Constans J, Jolles P 1986 Complete amino acid sequence of human vitamin D binding protein (group-specific component): evidence of a three fold internal homology as in serum albumin and alphafetoprotein. Biochim Biophys Acta 871:189–198. 16. Litwiller R, Fass R, Kumar R 1987 The amino acid sequence of the NH2-terminal portion of the rat and human vitamin D binding protein: evidence for a high degree of homology between rat and human vitamin D binding protein. Life Sci 38:2179–2184. 17. Head JF, Swamy N, Ray R 2002 Crystal structure of the complex between actin and human vitamin D binding protein at 2.5Å resolution. Biochemistry 41:9015–9020. 18. Otterbein LR, Cosio C, Graceffa P, Dominguez R 2002 Crystal structures of the vitamin D–binding protein and its complex with actin: structural basis of the actin-scavenger system. Proc Natl Acad Sci USA 99:8003–8008. 19. Swamy N, Head JF, Weitz D, Ray R 2002 Biochemical and preliminary crystallographic characterization of the vitamin D sterol- and actin-binding by human vitamin D–binding protein. Arch Biochem Biophys 402:14–23. 20. Verboven C, Rabijns A, De Maeyer M, Van Baelen H, Bouillon R, De Ranter C 2002 A structural basis for the unique binding features of the human vitamin D–binding protein. Nature Struct Biol 9:131–136. 21. Verboven C, Bogaerts I, Waelkens E, Rabijns A, Van Baelen H, Bouillon R, De Ranter C 2003 Actin–DBP: the perfect structural fit? Acta Crystallogr D Biol Crystallogr 59:263–273. 22. Cooke NE, Williard HF, David EV, George DL 1986 Direct regional assignment of the gene for vitamin D binding protein (Gc-globulin) to human chromosome 4q11–q13 and identification of an associated DNA polymorphism. Hum Genet 73:225–229. 23. Harper ME, Dugaiczyk A 1983 Linkage of the evolutionarilyrelated serum albumin and α-fetoprotein genes with q11–q22 of human chromosome 4. Am J Hum Genet 35:565–572. 24. Cooke NE, Levan G, Szpirer J 1987 The rat vitamin D binding protein (Gc-globulin) gene is syntenic with the rat albumin and α-fetoprotein genes on chromosome 14. Cytogenet Cell Genet 44:98–100. 25. Weitkamp LR, Allen PZ 1979 Evolutionary conservation of equine Gc alleles and of mammalian Gc/albumin linkage. Genetics 92:1347–2272. 26. Juneja RK, Sandberg K, Andersson L, Gahne B 1982 Close physical linkage between albumin and vitamin D binding protein (Gc) loci in chicken: a 300 million year old linkage group. Genet Res Cambridge 40:95–98. 27. Lichenstein HS, Lyons DE, Wurfel MM, Johnson DA, McGinley MD, Leidli JC, Trollinger DB, Mayer JP, Wright SD, Zukowski MM 1994 Afamin is a new member of the albumin, α-fetoprotein, and vitamin D–binding protein gene family. J Biol Chem 269:18149–18154. 28. Belanger L, Roy S, Allard D 1994 New albumin gene 3′ adjacent to the α1-fetoprotein locus. J Biol Chem 269:5481–5484. 29. Gabant P, Forrester L, Nichols J, Van Reeth T, De Mees C, Pajack B, Watt A, Smitz J, Alexandre H, Szpirer C, Szpirer J 2002 Alpha-fetoprotein, the major fetal serum protein, is not essential for embryonic development but is required for female fertility. Proc Natl Acad Sci USA 99:12865–12870. 30. Song Y-H, Naumova AK, Liebhaber SA, Cooke NE 1999 Physical and meiotic mapping of the region of human chromosome 4q11–q13 encompassing the vitamin D binding
CHAPTER 8 Vitamin D–Binding Protein
31. 32.
33. 34. 35. 36.
37. 38. 39. 40. 41. 42. 43.
44. 45.
46.
47. 48.
49.
protein DBP/Gc-globulin and albumin multigene cluster. Genome Res 9:581–587. Urano Y, Sakai M, Watanabe K, Tamaoki T 1984 Tandem arrangement of the albumin and alpha-fetoprotein genes in the human genome. Gene 32:255–261. Chevrette M, Guertin M, Turcotte B, Belanger L 1987 The rat α1-fetoprotein gene: characterization of the 5′-flanking region and tandem organization with the albumin gene. Nucleic Acids Res 15:1338–1339. The Jackson Laboratory, Bar Harbor, Maine 2003 Mouse Genome Informatics Web Site. URL: http://www.informatics.jax. org. August. Eiferman FA, Young PR, Scott RW, Tilghman SM 1981 Intragenic amplification and divergence in the mouse α-fetoprotein gene. Nature 294:713–718. Gibbs EM, Dugaiczyk A 1987 Origin of structural domains of the serum albumin gene family and a predicted structure of the gene for vitamin D–binding protein. Mol Biol Evol 4:364–379. Haefliger DN, Moskaitis JE, Schoenberg DR, Wahli W 1989 Amphibian albumins as members of the albumin, alpha-fetoprotein, vitamin D–binding protein multigene family. J Mol Evol 29:344–354. Gibbs PEM, Witke WF, Dugaiczyk A 1998 The molecular clock runs at different rates among closely related members of a gene family. J Mol Evol 46:552–561. Brown MA, Chambers GK, Licht P 1997 Purification and partial amino acid sequences of two distinct albumins from turtle plasma. Comp Biochem Physiol 118B:367–374. Nishio H, Hamdi HK, Dugaiczyk A 1999 Genomic expansion across the albumin gene family on human chromosome 4q is directional. Biol Chem 380:1431–1434. Hay AWM, Watson G 1976 The plasma transport of 25-hydroxycholecalciferol in fish, amphibians, reptiles and birds. Comp Biochem Physiol 53:167–172. Van Baelen H, Allewaert K, Bouillon R 1988 New aspects of the plasma carrier for 25-hydroxycholecalciferol in vertebrates. Ann NY Acad Sci 538:60–68. Laing CJ, Fraser DR 1999 The vitamin D system in iguanian lizards. Comp Biochem Physiol 123B:373–379. Laing CJ, Shea GM, Fraser DR 1999 Variation in the concentration and binding affinities of the plasma vitamin D–binding protein. In: Danks J, Dacke C, Flik G, Gay C (eds) Calcium Metabolism: Comparative Endocrinology. BioScientifica, Bristol, UK, pp. 93–98. Laing CJ 2000 Comparative Biology of Plasma Vitamin D Binding Protein. Ph.D. thesis. Department of Animal Science, University of Sydney, Sydney, Australia (http://adt.caul.edu.au). Licht P 1994 The relation of the dual thyroxine/vitamin D– binding protein (TBP/DBP) of emydid turtles to the vitamin D– binding proteins of other vertebrates. Gen Comp Endocrinol 94:215–224. Whitworth DJ, Hunt L, Licht P 2000 Widespread expression of the mRNA encoding a novel vitamin D/thyroxine dual binding protein in the turtle Trachemys scripta. Gen Comp Endocrinol 118:354–358. Constans J, Viau M 1977 Group-specific component: evidence for two subtypes of the Gc1 gene. Science 198:1070–1071. Hoste B 1979 Group specific component (Gc) and transferrin (Tf) subtypes ascertained by isoelectric focusing. A simple non-immunological staining procedure for Gc. Hum Genet 50:75–79. Cleve H, Patutschnick W 1979 Neuraminidase treatment reveals sialic acid differences in certain genetic variants of the Gc system (vitamin D binding protein). Hum Genet 47:193–198.
129 50. Braun A, Bichlmaier R, Muller B, Cleve H 1993 Molecular evaluation of an Alu repeat including a polymorphic variable poly(dA) (AluVpA) in the vitamin D binding protein (DBP) gene. Hum Genet 90:526–532. 51. Dissing J, Thymann M, Hopkinson D 2003 Simultaneous detection of ACP1 and GC genotypes using PCR/SSCP. Ann Hum Genet 67:81–85. 52. Thymann M 1981 Gc subtypes determined by agarose isoelectric focusing. Distribution in Denmark, and application to paternity cases. Hum Hered 31:214–221. 53. Lai LYC, Constans J, Archer GT 1990 A new cathodal Gc variant in Australia. Ann Hum Biol 17:526–532. 54. Chen LZ, Easteal S, Board PG, Summers KM, Bhatia KK, Kirk RL 1990 Albumin–vitamin D–binding protein haplotypes in Asian-Pacific populations. Hum Genet 85:89–97. 55. Kravchuk OI, Balanovskii OP, Nurbaev SD, Makarov SV, Spitsyn VA, Ginter EK 1998 Genogenography of the aboriginal population of Marii El (from data on immunobiochemical polymorphism). Genetika 34:1542–1554. 56. Yin J, Zhang G, Liu J 2000 Genetic polymorphism of groupspecific component/vitamin-D–binding protein subtypes in six populations from north-eastern China. Anthropol Anzeiger 58:193–198. 57. Kofler A, Braun A, Jenkins T, Serjeanston SW, Cleve H 1995 Characterization of mutants of the vitamin D–binding protein/ group specific component: GC aborigine (1A1) from Australian aborigines and South African blacks, and 2A9 from south Germany. Vox Sang 68:50–54. 58. Yuasa I, Kofler A, Braun A, Bichlmaier R, Kammerer S, Cleve H 1995 A novel sequence polymorphism in exon 8 of the human vitamin D–binding protein (GC) gene in an African population. Jap J Hum Genet 40:149–151. 59. Yuasa I, Kofler A, Braun A, Umetsu K, Bichlmaier R, Kammerer S, Cleve H 1995 Characterization of mutants of the vitamin D–binding protein/group-specific component: molecular evolution of GC*1A2 and GC*1A3, common in some Asian populations. Hum Genet 95:507–512. 60. Dykes D, Polesky H, Cox E 1981 Isoelectric focusing of Gc (vitamin D binding globulin) in parentage testing. Hum Genet 58:174–175. 61. Yoshifumi Y, Nishimoto H, and Ikemoto S 1989 Interstitial deletion of the proximal long arm of chromosome 4 associated with father-child incompatibility within the Gc-system: probable reduced gene dosage effect and partial piebald trait. Am J Med Genet 32:520–523. 62. Vavrusa B, Cleve H, Constans J 1983 A deficiency mutant of the Gc system. Hum Genet 65:102–107. 63. Safadi FF, Thornton P, Magiera H, Hollis BW, Gentile M, Haddad JG, Liebhaber SA, Cooke NE 1999 Osteopathy and resistance to vitamin D toxicity in mice null for vitamin D binding protein. J Clin Invest 103:239–251. 64. Cleve H, Patutschnick W 1979 Different phenotypes of the group-specific component (Gc) in chimpanzees. Hum Genet 50:217–220. 65. Tsuchida S, Nagai A, Ikemoto S 1990 Polymorphisms of the group-specific component (Gc) in three species of Tamarins (Sanguinus oedipus, S. labiatus, S. mystax). Jpn J Vet Sci 52:649–651. 66. Imawari M, Akanuma Y, Muto Y, Itakura H, Kosaka K 1980 Isolation and partial characterization of two immunologically similar vitamin D–binding proteins in rat serum. J Biochem 88:349–360. 67. Gahne B, Juneja RK 1978 Polymorphic post-albumin of cattle and horse plasma identified as vitamin D binding
130
68. 69. 70. 71. 72. 73. 74. 75. 76.
77. 78.
79. 80.
81.
82.
83.
84.
protein (Gc protein). Anim Blood Groups Biochem Genet 9:37–40. Tate ML, Dodds KG, Thomas KJ, McEwan KM 1992 Genetic polymorphism of plasminogen and vitamin D binding protein in red deer (Cervus elephus L). Anim Genet 23:303–313. Vankan DM, Bell K 1992 Genetic polymorphism of plasma vitamin D–binding protein (Gc) in Australian goats. Anim Genet 23:457–462. Shin JA, Yang YH, Kim HS, Yun YM, Lee KK 2002 Genetic polymorphism of the serum proteins of horses in Jeju. J Vet Sci 3:255–263. Juneja RK, Niini T, Larsson HEB, Sandberg K 1991 Genetic polymorphism of six plasma proteins in the domestic cat. J Hered 82:178–181. Arthur H, Bell K, Cooper DW 1998 Plasma protein polymorphisms in the tammar wallaby (Macropus eugenii). Aust J Zool 46:193–199. Juneja RK, Gahne B, Kuryl J, Gasparka J 1982 Genetic polymorphism of the vitamin D–binding protein and a pre-transferrin in chicken plasma. Hereditas 98:89–96. Shibata T, Abe T 1996 Linkage between the loci for serum albumin and vitamin D binding protein (Gc) in the Japanese quail. Anim Genet 27:195–197. Lauridsen AL, Vestergaard P, Nexo E 2001 Mean serum concentration of vitamin D-binding protein (Gc globulin) is related to the Gc phenotype in women. Clin Chem 47:753–756. Braun A, Brandhofer A, Cleve H 1990 Interaction of the vitamin D–binding protein (group-specific component) and its ligand 25-hydroxy-vitamin D3: binding differences of the various genetic types disclosed by isoelectric focusing. Electrophoresis 11:478–483. Arnaud J, Constans J 1993 Affinity differences for vitamin D metabolites associated with the genetic isoforms of the human serum carrier protein (DBP). Hum Genet 92:183–188. Arnason A, Jensson O, Skaftadottir I, Birgisdottir B, Gudmundsson G, Johannesson G 1980 HLA types, Gc protein and other genetic markers in multiple sclerosis and two other neurological diseases in Iceland. Acta Neurol Scand 62(Suppl 78):39–40. Niino M, Kikuchi S, Fukazawa T, Yabe I, Tashiro K 2002 No association of vitamin D–binding protein gene polymorphisms in Japanese patients with MS. J Neuroimmunol 127:177–179. Pani MA, Regulla K, Segni M, Hoffmann S, Hufner M, Pasquino AM, Usadel KH, Badenhoop K 2002 A polymorphism within the vitamin D–binding protein gene is associated with Graves’ disease but not with Hashimoto’s thyroiditis. J Clin Endocrinol Metab 87:2564–2567. Ishii T, Keicho N, Teramoto S, Azuma A, Kudoh S, Fukuchi Y, Ouchi Y, Matsuse T 2001 Association of Gc-globulin variation with susceptibility to COPD and diffuse panbronchitis. Eur Respir 18:753–757. Sandford AJ, Chagani T, Weir TD, Connett JE, Anthonisen NR, Pare PD 2001 Susceptibility genes for rapid decline of lung function in the lung health study. Am J Resp Crit Care Med 163:469–473. Ezura Y, Nakajima T, Kajita M, Ishida R, Inoue S, Yoshida H, Suzuki T, Shiraki M, Hosoi T, Orimo H, Emi M 2003 Association of molecular variants, haplotypes, and linkage disequilibrium within the human vitamin D–binding protein (DBP) gene with postmenopausal bone mineral density. J Bone Miner Res 18:1642–1649. Papiha SS, Allcroft LC, Kanan RM, Francis RM, Datta HK 1999 Vitamin D binding protein gene in male osteoporosis: association of plasma DBP and bone mineral density with
CHRISTOPHER J. LAING AND NANCY E. COOKE
85.
86.
87. 88. 89. 90.
91.
92.
93. 94.
95. 96. 97. 98.
99.
100. 101.
(TAAA)(n)-Alu polymorphism in DBP. Calcif Tissue Int 65:262–266. Hlavaty JJ, Partin AW, Shue MJ, Mangold LA, Derby J, Javier T, Kelley S, Stieg A, Briggman JV, Hass GM, Wu Y-J 2003 Identification and preliminary clinical evaluation of a 50.8-kDa serum marker for prostate cancer. Urology 61:1261–1265. Haddad JG, Walgate J 1976 Radioimmunoassay of the binding protein for vitamin D and its metabolites in human serum. Concentrations in normal subjects and patients with disorders of mineral homeostasis. J Clin Invest 67:1550–1560. Bouillon R, Van Baelen H, DeMoor P 1977 The measurement of vitamin D–binding protein in human serum. J Clin Endocrinol Metab 45:225–231. Imawari M, Goodman DS 1977 Immunological and immunoassay studies of the binding protein for vitamin D and its metabolites in human serum. J Clin Invest 59:432–442. Parviainen M, Harmoinen A, Visakorpi J 1982 Immunonephelometric quantification of serum vitamin D–binding globulin. Scand J Clin Lab Invest 42:571–575. Rejnmark L, Lauridsen AL, Vestergaard P, Heickendorff L, Andreasen F, Mosekilde L 2002 Diurnal rhythm of plasma 1,25-dihydroxyvitamin D and vitamin D–binding protein in postmenopausal women: relationship to plasma parathyroid hormone and calcium and phosphate metabolism. Eur J Endocrinol 146:635–642. Bikle DD, Gee E, Halloran B, Haddad JG 1984 Free 1,25dihydroxyvitamin D levels in serum from normal subjects, pregnant subjects, and subjects with liver disease. J Clin Invest 74:1966–1971. Abbas SK, Care AD, Van Baelen H, Bouillon R 1987 Plasma vitamin D–binding protein and free 1,25-dihydroxyvitamin D3 index in pregnant ewes and their fetuses in the last month of gestation. J Endocrinol 115:7–12. Fraser DR, Emtage JS 1976 Vitamin D in the avian egg. Its molecular identity and mechanism of incorporation into yolk. Biochem J 160:671–682. Cheema C, Grant BF, Marcus R 1989 Effects of estrogen on circulating “free” and total 1,25-dihydroxyvitamin D and on parathyroid–vitamin D axis in postmenopausal women. J Clin Invest 83:537–542. Harris SS, Dawson-Hughes B 1998 The association of oral contraceptive use with plasma 25-hydroxyvitamin D levels. J Am Coll Nutr 17:282–284. Bouillon R, Vandoren G, Van Baelen H, DeMoor P 1978 Immunochemical measurement of the vitamin D–binding protein in rat serum. Endocrinology 102:1710–1715. Laing CJ, Fraser DR 2002 Changes with malnutrition in the concentration of plasma vitamin D–binding protein (DBP) in growing rats. Br J Nutr 88:133–139. Schiodt FV, Bondesen S, Muller K, Rasmussen A, Hjortrup A, Kirkegaard P, Hansen BA, Tygstrup N, Ott P 1999 Reconstitution of the actin-scavenger system after orthotopic liver transplantation for end-stage liver disease: a prospective and longitudinal study. Liver Transplant Surg 5:310–317. Nyomba BL, Bouillon R, Lissens W, Van Baelen H, De Moor P 1985 1,25-Dihydroxyvitamin D and vitamin D–binding protein are both decreased in streptozotocin-diabetic rats. Endocrinology. 116:2483–2488. Auwerx J, De Keyser L, Bouillon R, De Moor P 1986 Decreased free 1,25-dihydroxycholecalciferol index in patients with the nephrotic syndrome. Nephron 42:231–235. McLeod JF, Cooke NE 1989 The vitamin D–binding protein, α-fetoprotein, albumin multigene family: detection of transcripts in multiple tissues. J Biol Chem 264:21760–21769.
CHAPTER 8 Vitamin D–Binding Protein
102. Kawakami M, Blum CB, Ramakrishan R, Dell RB, Goodman DS 1981 Turnover of the plasma binding protein for vitamin D and its metabolites in normal human subjects. J Clin Endocrinol Metab 53:1110–1116. 103. Dahl B, Schiodt FV, Rudolph S, Ott P, Kiaer T, Heslet L 2001 Trauma stimulates the synthesis of Gc-globulin. Inten Care Med 27:394–399. 104. Schiodt FV, Ott P, Tygstrup N, Dahl B, Bondesen S 2001 Temporal profile of total, bound, and free Gc-globulin after acetaminophen overdose. Liver Transplant 7:732–738. 105. Schiodt FV, Clemmesen JO, Bondesen S, Dahl B, Ott P 2001 Increased turnover of Gc-globulin in patients with hepatic encephalopathy. Scand J Gastroenterol 36:998–1003. 106. Egawa T, Ito H, Nakamura H, Yamamoto H, Kishimoto S 1992 Hormonal regulation of vitamin D–binding protein production by a human hepatoma cell line. Biochem Int 28:551–557. 107. Guha C, Osawa M, Werner PA, Galbraith RM, Paddock GV 1995 Regulation of human Gc (vitamin D–binding) protein levels: hormonal and cytokine control of gene expression in vitro. Hepatology 21:1675–1681. 108. Danan JL, Ribiero V, Lechner MC, Pompom D 1991 Expression of the rat vitamin D–binding protein gene. In: Norman AW, Bouillon R, Thomasset M (eds) Vitamin D. Gene Regulation, Structure-Function Analysis and Clinical Application. Proceedings of the Eighth Workshop on Vitamin D. Walter de Gruyter, New York, pp. 697–698. 109. Hiroki T, Song Y-H, Naumova AK, Liebhaber SA, Cooke NE 2000 A human vitamin D binding protein gene with limited flanking sequences is sufficient for liver-specific and copy number-dependent expression in transgenic mice. The Endocrine Society, 82nd Annual Meeting, Toronto, June 21–25, (poster). 110. Song Y-H, Ray K, Liebhaber SA, Cooke NE 1998 Vitamin D– binding protein gene transcription is regulated by the relative abundance of hepatocyte nuclear factors 1α and 1β. J Biol Chem 273:28408–28418. 111. Chi YI, Frantz JD, Oh BC, Hansen L, Dhe-Paganon S, Shoelson SE 2002 Diabetes mutations delineate an atypical POU domain in HNF-1alpha. Mol Cell 10: 1129–1135. 112. Roncero C, Ventura J-J, Sanchez A, Bois-Joyeux B, Mesa ML, Thomassin H, Danan J-L, Benito M, Fabregat I 1998 Phorbol esters down-regulate α-fetoprotein gene expression without affecting growth in fetal hepatocytes in primary culture. Biochim Biophys Acta 1402:151–164. 113. Haddad JG, Fraser DR, Lawson DEM 1981 Vitamin D binding protein. Turnover and fate in the rabbit. J Clin Invest 67:1550–1560. 114. Nykjaer A, Dragun D, Walther D, Vorum H, Jacobsen C, Herz J, Melsen F, Christensen EI 1999 An endocytic pathway essential for renal uptake and activation of the steroid 25-(OH) vitamin D3. Cell 96:507–515. 115. Nykjaer A, Fyfe JC, Kozyraki R, Leheste JR, Jacobsen C, Nielsen MS, Verroust PJ, Aminoff M, De la Chapelle A, Moestrup SK, Ray R, Gliemann J, Willnow TE, Christensen EI 2001 Cubilin dysfunction causes abnormal metabolism of the steroid hormone 25(OH) vitamin D3. Proc Natl Acad Sci USA 98:13895–13900. 116. Leheste JR, Melsen F, Wellner M, Jansen P, Schlichting U, Renner-Muller I, Andreassen TT, Wolf E, Bachmann S, Nykjaer A, Willnow T 2003 Hypocalcemia and osteopathy in mice with kidney-specific megalin gene defect. FASEB J 17:247–249.
131 117. Goldschmidt-Clermont P, Van Baelen H, Bouillon R, Shook TE, Williams MH, Nel AE, Galbraith RM 1988 Role of groupspecific component (vitamin D binding protein) in clearance of actin from the circulation in the rabbit. J Clin Invest 81:1519–1527. 118. Dueland S, Nenseter MS, Drevon CA 1991 Uptake and degradation of filamentous actin and vitamin D–binding protein in the rat. Biochem J 274:237–241. 119. Cousin P, Dechaud H, Grenot C, Lejeune H, Hammond GL, Pugeat M 1999 Influence of glycosylation on the clearance of recombinant human sex hormone-binding globulin from rabbit blood. J Steroid Biochem Mol Biol 70:115–121. 120. Innes Chalk KJ, Kodicek E 1961 The association of 14Clabelled vitamin D2 with rat serum proteins. Biochem J 79:1–7. 121. Rikkers H, DeLuca HF 1967 An in vivo study of the carrier proteins of 3H-vitamins D3 and D4 in rat serum. Am J Physiol 213:380–386. 122. Bouillon R, Van Baelen H, Rombauts W, DeMoor P 1976 The purification and characterization of the human serum binding protein for the 25-hydroxycholecalciferol (transcalciferin). Identity with group specific component. Eur J Biochem 66:285–291. 123. Bouillon R, Verstuyf A, Zhao J, Tan B, Van Baelen H 1996 Nonhypercalcemic vitamin D analogs: interactions with the vitamin D–binding protein. Horm Res 45:117–121. 124. Swamy N, Dutta A, Ray R 1997 Roles of the structure and orientation of ligands and ligand mimics inside the ligandbinding pocket of the vitamin D–binding protein. Biochemistry 36:7432–7436. 125. Norman AW, Ishizuka S, Okamura WH 2001 Ligands for the vitamin D endocrine system: different shapes function as agonists and antagonists for genomic and rapid response receptors or as a ligand for the plasma vitamin D binding protein. J Steroid Biochem Mol Biol 76:49–59. 126. Carlberg C 2003 Molecular basis of the selective activity of vitamin D analogues. J Cell Biochem 88:274–281. 127. Vieth R 1994 Simple method for determining specific binding capacity of vitamin D–binding protein and its use to calculate the concentration of “free” 1,25-dihydroxyvitamin D. Clin Chem 40:435–441. 128. Teegarden D, Meredith SC, Sitrin MD 1991 Determination of the affinity of vitamin D metabolites to serum vitamin D binding protein using an assay employing lipid-coated polystyrene beads. Anal Biochem 199:293–299. 129. Ray R, Bouillon R, Van Baelen H, Holick MF 1991 Photoaffinity labelling of human serum vitamin D binding protein and chemical cleavages of the labelled protein: identification of an 11.5 kDa peptide containing the putative 25-hydroxyvitamin D3 binding site. Biochemistry 30:6638–7642. 130. Haddad JG, Hu YZ, Kowalski MA, Laramore C, Ray K, Robzyk P, Cooke NE 1992 Identification of the sterol- and actin binding domains of plasma vitamin D binding protein (Gc globulin). Biochemistry 31:7174–7181. 131. Addo JK, Swamy N, Ray R 2002 The C19 position of 25-hydroxyvitamin D3 faces outward in the vitamin D sterolbinding pocket of vitamin D–binding protein. Bioorg Med Chem Lett 12:279–281. 132. Avvakumov GV, Grishkovskaya I, Muller YA, Hammond GL 2002 Crystal structure of human sex hormone-binding globulin in complex with 2-methoxyestradiol reveals the molecular basis for high affinity interactions with C-2 derivatives of estradiol. J Biol Chem 277:45219–45225.
132 133. Grishkovskaya I, Avvakumov GV, Hammond GL, Catalano MG, Muller YA 2002 Steroid ligands bind human sex hormone-binding globulin in specific orientations and produce distinct changes in protein conformation. J Biol Chem 277:32086–32093. 134. Surarit R, Svasti J 1980 Effect of ligand binding on the conformation of human plasma vitamin D–binding protein (group specific component). Biochem J 191:401–410. 135. Koszelak S, McPherson A, Bouillon R, Van Baelen H 1985 Crystallisation and preliminary X-ray analysis of the vitamin Dbinding protein from human serum. J Steroid Biochem 23:1077–1078. 136. Cooke N, Walgate J, Haddad J 1979 Human serum binding protein for vitamin D and its metabolites. II. Specific, high affinity association with a protein in nucleated tissue. J Biol Chem 254:5965–5971. 137. Van Baelen H, Bouillon R, De Moor P 1980 Vitamin D–binding protein (Gc-globulin) binds actin. J Biol Chem 255:2270–2272. 138. Buch S, Gremm D, Wegner A, Mannherz HG 2002 Binding of a C-terminal fragment (residues 369 to 435) of vitamin D– binding protein to actin. Biol Chem 383:1621–1631. 139. McLeod JF, Kowalski MA, Haddad JG 1989 Interactions among serum vitamin D binding protein, monomeric actin, profilin and profilactin. J Biol Chem 264:1260–1267. 140. Emerson DL, Arnaud P, Galbraith RM 1983 Evidence of increased Gc:actin complexes in pregnant serum: a possible result of trophoblast embolism. Am J Reprod Immunol 4:185–189. 141. Goldschmidt-Clermont PJ, Galbraith RM, Emerson DL, Werner PAM, Nel AE, Lee WM 1985 Accurate quantitation of native Gc in serum and estimation of endogenous GcG-actin complexes by rocket immunoelectrophoresis. Clin Chim Acta 148:173–183. 142. Haddad JG, Harper KD, Guoth M, Pietra GG, Sanger JW 1990 Angiopathic consequences of saturating the plasma scavenger system for actin. Proc Natl Acad Sci USA 87:1381–1385. 143. Lee WM, Galbraith RM 1992 The extracellular actinscavenger system and actin toxicity. New Engl J Med 326: 1335–1341. 144. Petrini M, Galbraith RM, Werner, PAM, Emerson DL, Arnaud P 1984 Gc (vitamin D binding protein) binds to cytoplasm of all human lymphocytes and is expressed on B-cell membranes. Clin Immunol Immunopathol 31:282–295. 145. Machii T, Kimura H, Ueda E, Chujo T, Morikta T, Katagiri S, Tagawa S, Kitani T 1986 Distribution of Gc protein (vitamin D binding protein) on the surfaces of normal human lymphocytes and leukemic lymphocytes. Acta Haematol 75:26–29. 146. McLeod JF, Kowalski MA, Haddad JG 1986 Characterization of a monoclonal antibody to human serum vitamin D binding protein (Gc globulin): recognition of an epitope hidden in membranes of circulating monocytes. Endocrinology 119:77–83. 147. Guoth M, Murgia A, Smith RM, Prystowsky MB, Cooke NE, Haddad JG 1990 Cell surface vitamin D binding protein (Gc globulin) is acquired from plasma. Endocrinology 127:2313–2321. 148. DiMartino SJ, Kew RR 1999 Initial characterisation of the vitamin D binding protein (Gc-globulin) binding site on the neutrophil plasma membrane: evidence for a chondroitin sulfate proteoglycan. J Immunol 163:2135–2142. 149. Danan JL, Delorme AC, Ripoche MA, Bouillon R, Van Baelen H, Mathieu H 1985 Presence of immunoreactive vitamin D–binding protein in rat yolk sac endodermal cells. Endocrinology 117:243–247.
CHRISTOPHER J. LAING AND NANCY E. COOKE
150. Emerson DL, Werner PA, Cheng MH, Galbraith RM 1985 Presence of Gc (vitamin D–binding protein) and interactions with actin in human placental tissue. Am J Reprod Immunol Microbiol 7:15–21. 151. Nestler JE, McLeod JF, Kowalski MA, Strauss JF, Haddad JG 1987 Detection of vitamin D binding protein on the surface of cytotrophoblasts isolated from human placentae. Endocrinology 120:1996–2002. 152. Yu HM, Li XJ, Kadam AL, Cheng CY, Koide SS 1994 Human testis vitamin D binding protein involved in infertility. Arch Androl 33:119–128. 153. Moestrup SK, Verroust PJ (2001) Megalin- and cubilinmediated endocytosis of protein-bound vitamins, lipids, and hormones in polarized epithelia. Ann Rev Nutr 21:407–428. 154. Petrini M, Galbraith RM, Emerson DL, Nel AE, Arnaud P 1985 Structural studies of T lymphocyte Fc receptors. Association of Gc protein with IgG binding to Fc gamma. J Biol Chem 260:1804–1810. 155. DiMartino SJ, Shah AB, Trujillo G, Kew RR 2001 Elastase controls the binding of the vitamin D–binding protein (Gc-globulin) to neutrophils: a potential role in the regulation of C5a co-chemotactic activity. J Immunol 166:2688–2694. 156. Osawa M, Kimura A, Yukawa N, Seto Y, Saito T, Tsuji T, Takeichi S 1998 Epitope analysis of monoclonal antibodies to human Gc globulin (vitamin D–binding protein). Biochem Mol Biol Int 44:293–303. 157. Weitkamp LR, Rucknagel DL, Gershowitz H 1966 Genetic linkage between structural loci for albumin and group specific component (Gc). Am J Hum Genet 18:559–571. 158. Daiger SP, Schanfield MS, Cavalli-Sforza LL 1975 Human group-specific component (Gc) proteins bind vitamin D and 25-hydroxyvitamin D. Proc Natl Acad Sci USA 72:2076–2080. 159. White P, Cooke N 2000 The multifunctional properties and characteristics of vitamin D binding protein. Trends Endocrinol Metab 11:320–321. 160. Mendel CM 1989 The free hormone hypothesis: a physiologically based mathematical model. Endocr Rev 10:232–274. 161. Hammond GL 2002 Access of reproductive steroids to target tissues. Obstet Gynecol Clin N Am 29:411–423. 162. Siiteri PK, Murai JT, Hammond GL, Nisker JA, Raymoure WJ, Kuhn RW 1982 The serum transport of steroid hormones. Rec Prog Horm Res 38:457–503. 163. Vermeullen A 1988 Physiology of the testosterone binding globulin in man. Ann NY Acad Sci 538:103–111. 164. Nakhla AM, Leonard J, Hryb DJ, Rosner W 1999 Sex hormone–binding globulin receptor signal transduction proceeds via a G protein. Steroids 64:213–216. 165. Rosner W, Hryb DJ, Khan MS, Nakhla AM, Romas NA 1999 Androgen and estrogen signaling at the cell membrane via G-proteins and cyclic adenosine monophosphate. Steroids 64:100–106. 166. Rosner W, Hryb DJ, Khan MS, Nakhla AM, Romas NA 1999 Sex hormone–binding globulin mediates steroid hormone signal transduction at the plasma membrane. J Steroid Biochem Mol Biol 69:481–485. 167. Heinlein CA, Chang C 2002 The roles of androgen receptors and androgen-binding proteins in nongenomic androgen actions. Mol Endocrinol 16:2181–2187. 168. Hryb DJ, Nakhla AM, Kahn SM, St George J, Levy NC, Romas NA, Rosner W 2002 Sex hormone-binding globulin in the human prostate is locally synthesized and may act as an autocrine/paracrine effector. J Biol Chem 277: 26618–26622.
CHAPTER 8 Vitamin D–Binding Protein
169. Kahn SM, Hryb DJ, Nakhla AM, Romas NA, Rosner W 2002 Sex hormone-binding globulin is synthesized in target cells. J Endocrinol 175:113–120. 170. Orchinik M, Hastings N, Witt D, McEwen BS 1997 Highaffinity binding of corticosterone to mammalian neuronal membranes: possible role of corticosteroid binding globulin. J Steroid Biochem Mol Biol 60:229–236. 171. Hodam JR, Creek KE 1998 Comparison of the metabolism of retinol delivered to human keratinocytes either bound to serum retinol-binding protein or added directly to the culture medium. Exp Cell Res 238:257–264. 172. Sundaram M, Sivaprasadarao A, DeSousa MM, Findlay JBC 1998 The transfer of retinol from serum retinol-binding protein to cellular retinol-binding protein is mediated by a membrane receptor. J Biol Chem 273:336–3342. 173. Noy N Retinoid-binding proteins: mediators of retinoid action. Biochem J 348:481–495. 174. Breuner CW, Orchinik M 2002 Plasma binding proteins as mediators of corticosteroid action in vertebrates. J Endocrinol 175:99–112. 175. Bouillon R, Van Assche F, Van Baelen H, Heyns W, DeMoor P 1981 Influence of the vitamin D–binding protein on the serum concentration of 1,25-dihydroxyvitamin D—significance of the free 1,25-dihydroxyvitamin D concentration. J Clin Invest 67:589–596. 176. Dilworth FJ, Calverly MJ, Makin HLJ, Jones G 1994 Increased biological activity of 20-epi-1,25-dihydroxyvitamin D3 is due to reduced catabolism and altered protein binding. Biochem Pharmacol 47:987–993. 177. Vanham G, Van Baelen H, Tan BK, Bouillon R 1988 The effect of vitamin D analogs and of vitamin D-binding protein on lymphocyte proliferation. J Steroid Biochem 29:381–386. 178. Vargas S, Bouillon R, Van Baelen H, Raisz LG 1990 Effects of vitamin D-binding protein on bone resorption stimulated by 1,25-dihydroxyvitamin D3. Calcif Tissue Int 47:164–168. 179. Furuichi T, Kawata S, Asoh Y, Kumaki K, Ohyama Y 1998 Differential time course of induction of 1α,25-dihydroxyvitamin D3-24-hydroxylase mRNA expression in rats by 1α, 25dihydroxyvitamin D3 and its analogs. Life Sci 62:453–459. 180. Pettifor JM, Bikle DD, Cavaleros M, Zachen D, Kamdar MC, Ross FP 1995 Serum levels of free 1,25-dihydroxyvitamin D in vitamin D toxicity. Ann Intern Med 122:511–513. 181. Van Hoof HJ, Swinkels LM, Ross HA, Sweep CG, Benraad TJ 1998 Determination of non-protein-bound plasma 1,25dihydroxyvitamin D by symmetric (rate) dialysis. Anal Biochem 258:176–183. 182. Silver J, Fainaru M 1979 Transport of vitamin D sterols in human plasma: effect of excess vitamin D, 25-hydroxyvitamin D and 1,25-dihydroxyvitamin D. Eur J Clin Invest 9:433–438. 183. Adams JS, Chen H, Chun RF, Nguyen L, Wu S, Ren S-Y, Barsony J, Gacad MA 2003 Novel regulators of vitamin D action and metabolism: lessons learned at the Los Angeles Zoo. J Cell Biochem 88:308–314. 184. Larsson D, Nemere I, Sundell K 2001 Putative basal lateral membrane receptors for 24,25-dihydroxyvitamin D3 in carp and Atlantic cod enterocytes: characterization of binding and effects on intracellular calcium regulation. J Cell Biochem 83:171–186. 185. Norman AW, Olivera CJ, Silva FRMB, Bishop JE 2002 A specific binding protein/receptor for 1α,25-dihydroxyvitamin D3 is present in an intestinal caveolae membrane fraction. Biochem Biophys Res Commun 298:414–419.
133 186. Donohue MM, Demay MB 2002 Rickets in VDR null mice is secondary to decreased apoptosis of hypertrophic chondrocytes. Endocrinology 143:3691–3694. 187. St-Arnaud R, Arabian A, Glorieux FH 1996 Abnormal bone development in mice deficient for the vitamin D 24-hydroxylase gene. J Bone Miner Res 11:S126. 188. St-Arnaud R, Dardenne O, Prud’homme J, Hacking SA, Glorieux FH 2003 Conventional and tissue-specific inactivation of the 25-hydroxyvitamin D-1-hydroxylase (CYP27B1). J Cell Biochem 88:245–251. 189. Torpy DJ, Bachmann AW, Grice JE, Fitzgerald SP, Phillips PJ, Whitworth JA, Jackson RV 2001 Familial corticosteroidbinding globulin deficiency due to a novel null mutation: association with fatigue and relative hypotension. J Clin Endocrinol Metab 86:3692–3700. 190. Joyner JM, Hutley LJ, Bachmann AW, Torpy DJ, Prins JB 2003 Greater replication and differentiation of preadipocytes in inherited corticosteroid-binding globulin deficiency. Am J Physiol Endocrinol Metab 284:E1049–E1054. 191. Domingues R, Bugalho MJ, Garrao A, Boavida JM, Sobrinho L 2002 Two novel variants in the thyroxine-binding globulin (TBG) gene behind the diagnosis of TBG deficiency. Eur J Endocrinol 146:485–490. 192. Su C-C, Wu Y-C, Chiu C-Y, Won JG-S, Jap T-S 2003 Two novel mutations in the gene encoding thyroxine-binding globulin (TBG) as a cause of complete TBG deficiency in Taiwan. Clin Endocrinol 58:409–414. 193. Vogel S, Piantedosi R, O’Byrne SM, Kako Y, Quadro L, Gottesman ME, Goldberg IJ, Blaner WS 2002 Retinol-binding protein-deficient mice: biochemical basis for impaired vision. Biochemistry 41:15360–15368. 194. Baker ME 2002 Albumin, steroid hormones and the origin of vertebrates. J Endocrinol 175:121–127. 195. Soliman AT, Madina EH, Morsi MR 1996 Radiological, biochemical, and hormonal changes in malnourished children with rachitic manifestations. J Trop Pediatr 42:34–37. 196. Lulseged S, Fitwi G 1999 Vitamin D deficiency rickets: sociodemographic and clinical risk factors in children seen at a referral hospital in Addis Ababa. East Afr Med J 76: 457–461. 197. Vasconcellos CA, Lind SE 1993 Coordinated inhibition of actin-induced platelet aggregation by plasma gelsolin and vitamin D–binding protein. Blood 82:3648–3657. 198. Erukhimov JA, Tang Z-L, Johnson BA, Donahoe MP, Razzack JA, Gibson KF, Lee WM, Wasserloos KJ, Watkins SA, Pitt BR 2000 Actin-containing sera from patients with adult respiratory distress syndrome are toxic to sheep pulmonary epithelial cells. Am J Respir Crit Care Med 162: 288–294. 199. Schiodt FV, Ott P, Bondesen S, Tygstrup N 1997 Reduced serum Gc-globulin concentrations in patients with fulminant hepatic failure: association with multiple organ failure. Crit Care Med 25:1366–1370. 200. Lee WM, Reines D, Watt GH, Cook JA, Wise WC, Halushka PV, Galbraith RM 1989 Alterations in Gc levels and complexing in septic shock. Circ Shock 28:249–255. 201. Osawa M, Yukawa N, Saito T, Huang XL, Kusakabe T, Takeichi S 1998 Increased complex formation of Gc globulin with actin in plasma from human cadavers. Forensic Sci Int 96:39–45. 202. Dahl B, Schiodt FV, Ott P, Wians F, Lee WM, Balko J, O’Keefe GE 2003 Plasma concentration of Gc-globulin is associated with organ dysfunction and sepsis after injury. Crit Care Med 31:152–156.
134 203. Dahl B, Schiodt FV, Kiaer T, Ott P, Bondesen S, Tygstrup N 1998 Serum Gc-globulin in the early course of multiple trauma. Crit Care Med 26:285–289. 204. Metcalf JP, Thompson AB, Gossman GL, Nelson KJ, Koyama S, Rennard SI, Robbins RA 1991 Gc globulin functions as a co-chemotaxin in the lower respiratory tract. Am Rev Respir Dis 143:844–849. 205. Kew RR, Mollison KW, Webster RO 1992 Binding of Gc globulin (DBP) to C5a or C5a des Arg is not necessary for co-chemotactic activity. J Leukocyte Biol 58:55–58. 206. Kew RR, Fisher JA, Webster RO 1995 Co-chemotactic effect of Gc-globulin (DBP) for C5a: transient conversion into an active co-chemotaxin by neutrophils. J Immunol 155: 5369–5374. 207. Binder R, Kress A, Kan G, Herrmann K, Kirschfink M 1999 Neutrophil priming by cytokines and vitamin D binding protein (Gc-globulin): impact on C5a-mediated chemotaxis, degranulation and respiratory burst. Mol Immunol 36:885–892. 208. Belting M 2003 Heparan sulfate proteoglycan as a plasma membrane carrier. Trends Biochem Sci 28:145–151. 209. Yamamoto N, Ngwenya BZ 1987 Activation of mouse peritoneal macrophages by lysophospholipids and ether derivatives of neutral lipids and phospholipids. Cancer Res 47:2008–2013. 210. Yamamoto N, St Claire DA, Homma S, Ngwenya BZ 1988 Activation of mouse macrophages by alkylglycerols, inflammation products of cancerous tissues. Cancer Res 48:6044–6049. 211. Naraparaju VR, Yamamoto N 1994 Roles of β-galactosidase of B lymphocytes and sialidase of T lymphocytes in inflammation-primed activation of macrophages. Immunol Lett 43:143–148. 212. Yamamoto N 1996 Structural definition of a potent macrophage activating factor derived from vitamin D3-binding protein with adjuvant activity for antibody production. Mol Immunol 33:1157–1164. 213. Mohamad SB, Nagasawa H, Uto Y, Hori H 2002 Preparation of Gc protein-derived macrophage activating factor (GcMAF) and its structural characterization and biological activities. Anticancer Res 22:4297–4300. 214. Yamamoto N, Naraparaju VR, Neil RN, Suyama H, Nakazato H 1997 Therapeutic efficiency of vitamin D3-binding protein-derived macrophage activating factor for prostate, breast, and colon cancers. Proc Am Assoc Cancer Res 38:31. 215. Koga Y, Naraparaju VR, Yamamoto N 1999 Antitumor effect of vitamin D–binding protein-derived macrophage activating factor on Ehrlich ascites tumor-bearing mice. Proc Soc Exp Biol Med 220:20–26.
CHRISTOPHER J. LAING AND NANCY E. COOKE
216. Yamamoto N, Urade M 2002 Therapeutic efficacy of Gc protein derived macrophage activating factor for adenocarcinoma. Proc Am Assoc Cancer Res 43:914. 217. Ono T, Ichikawa H, Asami R, Yamamoto N 1995 Enzymatically modified Gc globulin induces the translocation of FcγRI and FcγRII from intracellular storage compartments to the cell surface in murine peritoneal macrophage. Jpn J Inflamm 15:293–299. 218. Yamamoto N, Naraparaju VR, Asbell SO 1996 Deglycosylation of serum vitamin D3–binding protein leads to immunosuppression in cancer patients. Cancer Res 56:2827–2831. 219. Mohamad SB, Nagasawa H, Uto Y, Hori H 2002 Tumor cell alpha-N-acetylgalactosaminidase activity and its involvement in GcMAF-related macrophage activation. Comp Biochem Biophys 132A:1–8. 220. Swamy N, Ghosh S, Schneider GB, Ray R 2001 Baculovirusexpressed vitamin D–binding protein-macrophage activating factor (DBP-maf) activates osteoclasts and binding of 25-hydroxyvitamin D3 does not influence this activity. J Cell Biochem 81:535–546. 221. Schneider GB, Benis KA, Flay NW, Ireland RA, Popoff SN 1995 Effects of vitamin D–binding protein-macrophage activating factor (DBP-MAF) infusion on bone resorption in two osteopetrotic mutations. Bone 16:657–662. 222. Yamamoto N, Naraparaju VR 1996 A defect in β-galactosidase of B-lymphocytes in the osteopetrotic (op/op) mouse. Immunol Lett 50:35–40. 223. Datta HK, Cook DB, Kanan RM 1999 Evidence for continuous basal generation of Gc-MAF: absence in infantile osteopetrosis and restoration after bone marrow transplant. Blood 93:4026–4027. 224. Kanda S, Mochizuki Y, Miyata Y, Kanetake H, Yamamoto N 2002 Effects of vitamin D3-binding protein–derived macrophage activating factor (GcMAF) on angiogenesis. J Nat Cancer Inst 94:1311–1319. 225. Kisker O, Onizuka S, Becker CM, Fannon M, Flynn E, D’Amato R, Zetter B, Folkman J, Ray R, Swamy N, PineShepherd S 2003 Vitamin D binding protein-macrophage activating factor (DBP-maf) inhibits angiogenesis and tumor growth in mice. Neoplasia 5:32–40. 226. White P, Liebhaber SA, Cooke NE 2002 129X1/SvJ mouse strain has a novel defect in inflammatory cell recruitment. J Immunol 168:869–874. 227. Reichner JS, Fitzpatrick PA, Wakshull E, Albina JE 2001 Receptor-mediated phagocytosis of rat macrophages is regulated differentially for opsonized particles and nonopsonized particles containing β-glucan. Immunology 104:198–206.
CHAPTER 9
New Aspects of DBP ILSE BOGAERTS AND CHRISTEL VERBOVEN Lab. Analytische Chemie, Van Evenstraat 4; B-3000 Leuven, Belgium
HUGO VAN BAELEN AND ROGER BOUILLON Legendo, Onderwijs en Navorsing 902, Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium.
I. Introduction II. Three-Dimensional Structure of Vitamin D–Binding Protein III. Three-Dimensional Structures of Vitamin D–Binding Protein in Complex with Ligands
IV. A Structural Explanation for the Unique Functions of Vitamin D–Binding Protein within Its Gene Family References
SUMMARY
N-terminal part of domain I. The crystal structure of the DBP–actin complex was determined at 2.4 Å resolution. This structure reveals that the fast-growing side of actin monomers is blocked completely through a perfect structural fit with DBP, demonstrating how DBP effectively interferes with actin-filament formation. Comparisons of the DBP structures with the structure of serum albumin, another family member, reveal a similar topology but also significant differences in overall as well as local folding. These structural differences explain the unique vitamin D3–binding and actin-binding properties of DBP.
The crystal structure of human serum vitamin D– binding protein (DBP) and the complex of hDBP and actin were solved. Despite the common genetic origin and similarities in overall gene/protein structure, the 3D structure of DBP differs substantially from that of human serum albumin. The crystal structure of DBP allowed the positioning of the surface grove in domain I responsible for binding of all vitamin D metabolites. This allows us to explain or predict the relative affinity of vitamin D metabolites and analogs for DBP. This is of particular importance since DBP affinity explains, at least in part, differences in pharmacokinetics and selective profiles of vitamin D analogs. The resolution of the DBP surface interacting with G actin explains the tight binding between these two proteins and also allows us to explain the molecular basis of the DBP actin scavenging system, which is an important protection against tissue damage–induced multiple organ failure. Finally, the binding site of DBP has been identified for FFA, but other putative interactions of DBP with macrophages (DBP-MAF) and complement need further work.
I. INTRODUCTION The human serum vitamin D–binding protein (DBP) has many physiologically important functions, ranging from transporting vitamin D3 metabolites, binding and sequestering globular actin and binding fatty acids to functioning in the immune system. The crystal structure of DBP in complex with the vitamin D3 metabolite 25-hydroxyvitamin D3 was solved, as well as the structure of DBP in complex with a vitamin D3 analog. Both structures reveal the vitamin D–binding site in the VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
II. THREE-DIMENSIONAL STRUCTURE OF VITAMIN D–BINDING PROTEIN A. Overall Features of the Vitamin D–Binding Protein Structure The crystal structure of human DBP was solved to 2.3-Å resolution [1]. The two DBP molecules present in the asymmetric unit (further referred to as DBP-A and DBP-B) are not identical. One of the major differences between them is that the DBP-B molecules are in complex with 25OHD3,whereas the DBP-A molecules are not. As in human serum albumin (HSA) [2,3], DBP has an all α-helical structure and contains three structurally similar domains (Fig. 1). DBP and the other members of its protein family (human serum albumin, afamin, and α-fetoprotein) are postulated to have evolved from a progenitor that arose from the triple repeat of a 192-amino acid sequence [4]. This threedomain structure has been preserved in DBP; however, the third repeat is largely truncated at the C terminus. Copyright © 2005, Elsevier, Inc. All rights reserved.
136
ILSE BOGAERTS, CHRISTEL VERBOVEN, HUGO VAN BAELEN, AND ROGER BOUILLON
DOMAIN I DOMAIN I
DOMAIN II
DOMAIN II
FIGURE 1
DOMAIN III
DOMAIN II
Stereo view of the overall fold of the DBP structure.
The HSA structure [2,3] indicates that each domain consists of 10 α-helices. In the DBP structure, the first domain (residues 1–191) has this α-helical arrangement. However, the second domain (residues 192–378) has a similar topology but helix 7 is replaced by a coil folding (Figs. 2, 3), and the third domain (residues 379–458) contains only helices 1–4. The three domains of DBP do not pack in a spherical manner but adopt a rather peculiar shape with two large grooves A more detailed comparison of both DBP-A and DBP-B structures reveals differences in local and overall folding. All pairwise Cα atoms in DBP-A and DBP-B superimpose with a root mean square (rms) deviation of 1.63 Å. Superimposing all common Cα atoms of both domains I, II, and III gives, however, mean rms deviations of 0.646 Å, 0.981 Å, and 0.416 Å, respectively. The higher value for the superposition of the overall DBP-A and DBP-B structures is caused by slightly different orientations of the respective domains in both structures. This difference of DBP-B may be
caused by the presence of 25OHD3, because a similar phenomenon is observed with the HSA structure. All three HSA domains assume a different orientation upon binding of free fatty acids [5]. Differences between DBP-A and DBP-B are also observed at the temperature factor level. The main difference is a much higher B-factor of the N terminus of domain I in the DBP-A molecule (91.2 compared to 76.8 for DBP-B). 25OHD3 is bound to helices 1–6 of domain I (see later discussion); its presence explains the higher stability in this region of DBP-B compared to the same region in DBP-A. In addition to the electron density for the 25OHD3 molecule, elongated residual density is observed that is able to accommodate oleic acid, a free fatty acid (FFA). DBP has been reported to bind unsaturated FFAs such as oleic acid [6–8]. Because the DBP25OHD3 crystals were obtained from a pooled human serum sample, the FFAs were expected to cocrystallize with DBP. Although DBP can be postulated to have several different FFA binding sites analogous with HSA [5,9], only one site was detected in DBP-A, and two fatty acids were seen in DBP-B. Because the FFAs present in this crystal are acquired from plasma rather than deliberately added to the protein sample, all possible FFA binding pockets may not be occupied.
B. Comparison with the Structure of Serum Albumin, a Member of Its Gene Family
FIGURE 2
The structure of domain II illustrating its typical topology. The numbering of the helices is kept the same as the similar HSA structure, although the small helix 7 is not present in this second domain of DBP.
Based on the presence of sequence homology and nearly identical disulfide bridge pattern in DBP, human serum albumin (HSA), α-fetoprotein, and afamin, the overall folds of these proteins are believed to be homologous [11]. Despite the sequence similarity, the only function DBP shares with the other family members is its fatty acid binding ability [6,7].
137
CHAPTER 9 New Aspects of DBP
FIGURE 3 Sequence alignment of DBP sequences of various species (DBP-H: human; DBP-RB: rabbit; DBP-R: rat; DBP-M: mouse) and the human sequences belonging to the same protein family (albu-H: human serum albumin; AFAM-H: human afamin, FETA-H: human α-fetoprotein). DBP has 23% sequence identity with serum albumin, 20% with α-fetoprotein, and 21% with afamin. Conserved residues are marked in blue; identical residues, green; and homologous residues, purple. Triangles indicate the Cys residues responsible for the typical disulfide bridge pattern observed in this protein family. For ALBU-H, AFAM-H, and FETO-H, only the sequence that aligns with the DBP sequence is shown. The secondary structure elements of DBP are also shown. The α-helices, shown at the top, have the same numbering as the similar HSA structure, although α-helix 7 of domain II has a coil folding in DBP. This figure was made with ALSCRIPT [10]. See color plate.
Superposition of the respective domains of the DBP and HSA structures shows similar topologies (Fig. 4a). Although the folding within each corresponding domain shows some parallels, the global orientation of the three domains in both molecules is strikingly different (Fig. 4b), resulting in two totally different structures. Superposition of all the aligned Cα atoms of DBP and HSA [3] gives an rms deviation as high as 10.1 Å (438 atoms superimposed). A comparison between the separate domains also indicates remarkable differences. The rms deviation between the aligned Cα atoms of domain I is 4.0 Å (175 atoms superimposed), 4.2 Å for domain II (185 atoms superimposed), and 3.9 Å for domain III
(78 atoms superimposed). These rms deviations are even higher when HSA is bound to fatty acids.
III. THREE-DIMENSIONAL STRUCTURES OF VITAMIN D–BINDING PROTEIN IN COMPLEX WITH LIGANDS A. The DBP–Vitamin D Complex 1. THE VITAMIN D BINDING SITE
Residual electron density that can accommodate 25OHD3 is observed in DBP-B close to the biochemically
138
ILSE BOGAERTS, CHRISTEL VERBOVEN, HUGO VAN BAELEN, AND ROGER BOUILLON
A
B
DOM III HSA
DOM III HSA DOM I HSA
DOM I HSA
DOM I DBP
DOM I DBP
NN
N N
DOM III DBP DOM III DBP
DOM II DBP DOM II DBP
DOM II HSA
DOM II HSA
FIGURE 4 Structural differences between DBP and serum albumin. (A) Stereo representation of the superimposed domains II of DBP (light gray) and HSA (dark grey). The helices of HSA are labeled. The small helix 7 is not present in this particular DBP domain. (B) Stereo view of the overall folds of the DBP and HSA (medium gray) molecules based on the superposition of the domains II. DBP domain I is shown in dark grey, DBP domains II and III in light grey.
identified vitamin D–binding residues (35–49) [12] (Fig. 5b). Residues belonging to helices 1–6 of domain I form the complete vitamin D–binding site (Fig. 5a,b). This binding site designation was further confirmed by the elucidation of the structure of DBP in complex with the JY analog, 22-(m-hydroxyphenyl)-23,24,25,26,27pentanor vitamin D3 (Table I). The electron density map of the DBP-JY structure clearly illustrates the presence of residual density in the N-terminal part of domain I for both the DBP-A and DBP-B molecules (Fig. 5c). This density can accommodate the JY analog. In contrast with the DBP-25(OH)D3 structure (Fig. 5b), the shape of the electron density in the sidechain region of the vitamin D3 molecule illustrates the presence of an aromatic feature (Fig. 5c). In the DBP-JY structure, the vitamin D–binding regions of both DBP-A
and DBP-B have similar average B-values, which is in contrast to the DBP-25(OH)D3 structure. The vitamin D–binding site is lined predominantly by hydrophobic residues (Fig. 5b), allowing favorable interactions with the hydrophobic vitamin D3 ligand. Binding of 25(OH)D3 is further improved by hydrogen bond formation of the 25-hydroxyl with Tyr 32 (O–H…O distance 2.85 Å), and the 3-hydroxyl with Ser 76 (O…H–O distance 2.81 Å) and Met 107 (O–H…S distance 3.01 Å). For DBP-JY, the same hydrogen bonds can be formed for 3-hydroxyl, but the hydroxyl of the phenolic substituent is hydrogen bonded to Glu 15. This hydroxyl of JY is at the same position as the oxygen of a water molecule in the DBP-25(OH)D3 structure (Fig. 5b,c). However, the detailed structural features of 25(OH)D3 or JY cannot be seen in the
139
CHAPTER 9 New Aspects of DBP
density of the structures, probably because not all of the sites present in the DBP-25(OH)D3 and DBP-JY crystals are fully occupied, but it may also be ascribed to mobility of the bound ligand. The vitamin D–binding site of DBP-B in the DBP-25(OH)D3 structure is
estimated to be 55% occupied by 25(OH)D3. For the DBP-JY structure, the estimated occupancy of JY in DBP-A is 0.66, whereas that for DBP-B is 0.72. Despite the low detail in the observed electron density, the proposed orientation of the molecule in the
A
B
F24
V12 WAT WAT
V51 E8
L31
Y32
T72
L47 435
576
F36 579
M107
FIGURE 5 The conformation of vitamin D metabolites and analogs in the vitamin D–binding site of DBP. (A) Stereo view of the location of the vitamin D–binding site in the DBP molecule. Helices 1–6 of domain I of DBP-A (light blue) and DBP-B (pink) from the DBP-25OHD3 structure are superimposed based on the common Cα atoms in helix 2. The DBP-B molecule is the one in which the loop at the top and helix 1(H1) are complete. The 25OHD3 is shown in a ball-and-stick representation. This superposition illustrates that the disordered residues of the DBP-A molecule belong to the region that forms the vitamin D–binding site. Two water molecules are present in the vitamin D–binding site (red balls). (B) View of the simulated annealing Fo−Fc omit map of DBP in the DBP-25OHD3 structure showing the electron density for the omitted 25OHD3 molecule. The map is contoured at 3.0 σ. Residues closer than 4.2 Å to the ligand are shown. Two water molecules (WAT) are present in the vitamin D–binding site (red balls). (Continued)
140
ILSE BOGAERTS, CHRISTEL VERBOVEN, HUGO VAN BAELEN, AND ROGER BOUILLON
C
E15
F24 L27
V12 E8
JY T72
WAT
V51 L47
S28 L31 Y32 K35 F36
L75 S76 S79 M107 D
FIGURE 5
Cont’d (C) Same as (B), but showing the electron density for the omitted JY molecule in the DBP-JY structure. The one remaining water molecule in the vitamin D–binding site is a red ball. (D) Stereo view of the ball-and-stick representation of the modeled vitamin D3 analogs in the vitamin D–binding site. 25OHD3 is shown in gray; 1,25(OH)2D3, dark blue; JX, pink; JY, purple; ZAF, yellow; ZAG, light blue; ZAQ, red; and ZAU, green. For ZAF, ZAG, and ZAQ, the presence of the substituent allows for extra hydrophobic interactions with DBP residues, probably resulting in the somewhat higher affinity compared to 1,25(OH)2D3. See color plate.
binding site is consistent with extensive biochemical and modeling data and, therefore, is assumed to be correct. The increased binding of the JY analog to DBP compared to the binding of 25(OH)D3 could result from favorable stacking of the JY aromatic side chain with Phe 24 (Fig. 5c). The aromatic residues Tyr 32, Phe 36, and Tyr 68 further stabilize the JY aromatic side chain. The JX analog, 22-(P-hydroxyphenyl)23,24,25,26,27-pentanor vitamin D3 (Table I), has an approximately threefold higher affinity for DBP than 25(OH)D3. Modeling of JX shows that the p-hydroxyl
allows for extra stacking with Phe 24 (Fig. 5d). This hydroxyl also makes a tight hydrogen bond with Ser 28. Although the modeling of 1,25(OH)2D3 illustrates that this molecule can make the same hydrogen bonds with DBP as 25(OH)D3 (Fig. 5d), it also shows that the axial 1-hydroxyl causes steric hindrance with Met 107. This may explain the lower DBP binding affinity observed for 1,25(OH)2D3. The vitamin D–binding site is a cleft; consequently, certain parts of the vitamin D3 molecule have no interaction with DBP—for example, the β-side of the C- and D-rings of
141
CHAPTER 9 New Aspects of DBP
TABLE I Structural Formulas of Vitamin D3 Metabolites and Analogs and Their Affinity for DBP 18
R1
12
R3
17 13
11
16
D
C 14
9
15
8
7
H
6 19
5 10
4 A
1
3 HO Vitamin D analog
R2
2
R2
R1 21
25OHD
24
22 20
23
25 27
R3
H
Affinity for DBP
H
667
H
1
H
H
2111
H
H
1314
26 OH 21 1,25(OH)2D3
22 20
24 23
25 27
OH
26 OH
JX OH OH JY
21 ZAF
ZAG
21
24
22 20
23
22 20
25 27
OH
2
OH
1.5
26 OH 24
23
25 27 26 OH
21 ZAQ 21 ZAU
22 20
24 23
22 20
OH
CH2=CH
OH
H
26 OH 24
23
25 27
25 27
3.8 1
To illustrate the relative affinity of the ligands for the vitamin D–binding protein, the reported Relative Competitive Index values [13] are shown. All reported values are relative to the result for 1,25(OH)D3, which is normalized to 1. The ZAU analog has the same structural formula as 1,25(OH)D3 except for the methyl group on C-13, which is replaced by a vinyl group in the β-position.
142
ILSE BOGAERTS, CHRISTEL VERBOVEN, HUGO VAN BAELEN, AND ROGER BOUILLON
the vitamin D3 molecule (Table I, Fig. 5a–c). Analogs substituted in these parts of the molecule display affinities similar to the affinity between 1,25(OH)2D3 and DBP (Table I). Furthermore, modeling of the analogs ZAG (11β-phenyl-1α,25-dihydroxyvitamin D3), ZAF (11α-phenyl-1α,25-dihydroxyvitamin D3), ZAQ (11αvinyl-1α,25-dihydroxyvitamin D3), and ZAU (13β-vinyl18-nor-1α,25-dihydroxyvitamin D3) illustrates that large substituents, such as a vinyl group or a phenyl ring, can easily fit in the binding site (Fig. 5d). These observations confirm the proposed orientation of 25OHD3 in the vitamin D–binding site. 2. DIFFERENCES BETWEEN THE VITAMIN D–BINDING SITE AND THE VITAMIN D POCKET OF THE VITAMIN D RECEPTOR
The effects of the vitamin D3 hormone and its analogs are mediated through binding to an intracellular vitamin D receptor (VDR) [13]. The comparison of the DBP vitamin D–binding site with that of the VDR structure [14] clearly illustrates that they are dissimilar. In VDR, the vitamin D–binding site is a closed pocket formed in the inner structure of the receptor, whereas in DBP it is a cleft located at the surface of the molecule and partly in contact with the surrounding solvent. Moreover, no parallel for both of these physiologically important binding sites is found in ligand conformation, hydrogen bonding, or hydrophobic interactions. In the VDR pocket, the A-ring of the vitamin D3 ligand has a B-chair conformation (3-hydroxyl axial), whereas in the DBP cleft the A-ring has an A-chair conformation (3-hydroxyl equatorial). In both binding sites, the C5–C6–C7–C8 torsion angle of the ligand is nonplanar. The angle is 149° in DBP and −149° in VDR. The torsion angles responsible for the side-chain orientation of the vitamin D3 molecule are different as well (C13–C17–C20–C22 = −77° in DBP and 89° in VDR; C17–C20–C22–C23 = −70° in DBP and −156° in VDR). 3. BIOLOGICAL IMPLICATIONS
The active vitamin D3 hormone, 1,25(OH)2D3, is used in the treatment of renal osteodystrophy, hypoparathyroidism, and osteoporosis. Its analogs are used successfully in the treatment of psoriasis [15] and are currently being explored for the treatment of hyperproliferative disorders [16] and the prevention of autoimmune disorders [17,18]. However, treatment with 1,25(OH)2D3 is limited because of severe side effects (such as hypercalciurea, hypercalcemia, and increased bone resorption) caused by the required pharmacological doses [13]. In order to separate the antiproliferative and prodifferentiating effects from the calcemic and skeletal effects,
a whole generation of chemically modified 1,25(OH)2D3 molecules (analogs) has been synthesized, but with poor success rates. Many of the analogs, having a more selective activity profile, display a decreased binding affinity for DBP [19–21]. Recent progress in certain aspects of the functional mechanism of 1,25(OH)2D3 provides new perspectives for the rational design of vitamin D3 analogs with a more selective profile. For example, structural analysis of the ligand binding domain of VDR [14,22] revealed the three-dimensional arrangement of its vitamin D–binding pocket. The elucidation of the three dimensional architecture of DBP’s vitamin D–binding site provides a solid basis for detailed structural studies on DBP–vitamin D analog complexes that will lead to the better understanding of the selective action mechanisms of 1,25(OH)2D3 and its analogs. The information contained in the three-dimensional architectures of both the DBP and VDR vitamin D–binding sites may lead to the development of compounds with a more original chemical structure and a more tissue- or gene-specific biological action.
B. The DBP–Actin Complex 1. ROLE OF DBP IN THE ACTIN-SCAVENGER SYSTEM
Actin is a highly abundant intracellular protein present in all eukaryotic cells and has a pivotal role in muscle contraction as well as in cell movements. Actin also has an essential function in maintaining and controlling cell shape and architecture: It is the essential building block of the microfilament system, a cytoskeletal structure that complements two other cytoskeletal structures (the microtubules and the intermediate filaments). In low-salt buffers actin exists as a monomeric protein (globular actin, G-actin), but it polymerizes under physiological salt conditions into a double helical 10-nm-thick filament structure (filamentous actin, F-actin). In vivo, the equilibrium between G-actin and F-actin is controlled mainly by the action of several actin-binding proteins with distinct activities (e.g., gelsolin or profilin). In conditions involving severe cell injury, such as trauma, shock, sepsis, and fulminant hepatic necrosis, large quantities of actin are released in the systemic circulation. The presence of actin filaments in blood, leading to an increase in blood viscosity, is dangerous and may be fatal [23]. In addition, actin can promote clot formation by its ability to aggregate platelets [24]. Therefore, the presence of a protective system, preventing actin polymerization and ensuring actin’s fast elimination, is essential.
143
CHAPTER 9 New Aspects of DBP
The vitamin D–binding protein is an actin-binding protein [11,25] and acts as an actin-sequestering agent in extracellular space. DBP and gelsolin—the only other plasma protein that binds actin avidly—play a crucial role in the clearance of actin filaments from the circulation, a process known as the “actin-scavenger system” [23]. Gelsolin has three actin-binding properties: (i) it severs F-actin; (ii) it caps the end of F-actin; and (iii) it nucleates actin filament assemblies. Gelsolin will shift the equilibrium between actin assembly–disassembly toward actin depolymerization, but will basically not prevent actin reassembly. DBP, by forming a complex with G-actin, prevents the (re-)formation of actin filaments. Therefore DBP and gelsolin have complementary functions: Upon severing the actin filaments by gelsolin, the generated globular actin is locked in its monomeric state by DBP. The clearance of this DBP–actin complex is even substantially faster compared to the clearance of free DBP [26–29]. The complementary action of both proteins is further illustrated by the previous observation that both DBP and gelsolin are required to inhibit actinstimulated platelet aggregation [24]. 2. GENERAL FEATURES OF THE DBP–ACTIN STRUCTURE
The DBP–actin structure described here was solved by X-ray diffraction by Verboven et al. [30]. The final model of the DBP–actin structure converged to an R factor of 19.90% and an Rfree value of 25.35%. It contains an actin, DBP, and ATP molecule, one calcium ion, and 290 water molecules. The structure of the DBP–actin complex reveals that G-actin binds in one of the DBP grooves, mainly formed by helix 10 of domain I, helix 6 of domain II, and helix 3 of domain III, allowing DBP and actin to fit as two pieces of a jigsaw puzzle (Fig. 6). The surface area buried at the interface of this complex is as large as 3600 Å2. Numerous intermolecular hydrogen bonds, hydrophobic contacts, and electrostatic interactions (Table II) stabilize the complex. The actin monomer consists of two domains with each domain further subdivided in two subdomains (Fig. 6). Subdomain 1 comprises residues 1–32, 70–137, and 338–375; subdomain 2, residues 33–69; subdomain 3, residues 138–180 and 270–337; and subdomain 4, residues 181–269. Biochemical experiments delimited the DBP interface by the residues 360–372 of the actin subdomain 1 [31]. The DBP–actin structure demonstrates that actin residues of subdomains 1 and 3 constitute the DBP binding interface (Fig. 6, Table II). As in actin–gelsolin segment 1 (actin-GS1) [32] and in actin–gelsolin segment 4 to 6 (actin-GS4-6) [33], the binding of DBP occurs at the actin cleft formed at the
SUBDOMAIN2
SUBDOMAIN4
SUBDOMAIN1 SUBDOMAIN3
FIGURE 6 Overall architecture of the DBP–actin structure. Actin binds in one of the large grooves present in the DBP structure. The three domains of DBP and the four subdomains in actin are shown in different colors. See color plate.
interface of subdomains 1 and 3. The exposed hydrophobic residues on helix 341–349 of actin subdomain 1 are thereby occluded from solvent. Even more striking is that in all these complexes the apolar patch (actin residues 341–349) is masked by a solventexposed hydrophobic face of a helix present in gelsolin segment 1, in gelsolin segment 4, and in DBP. Upon complexation with DBP, the changes in the actin structure are restricted to some small regions rather than affecting the general fold. This is reflected in the small rms differences obtained from the superposition of the DBP–actin structure with other actin complexes: actin-GS1 [32], actin-DNase I [34], and uncomplexed actin [35]. Similar to other actin complex structures [32,33], the structural differences are mainly caused by folding changes in subdomains 2 and 4, whereas, despite their interaction with DBP, the backbone fold of subdomains 1 and 3 is hardly altered. Only the side chains of the residues at the interface adopt different orientations to ensure the favorable contacts with the neighboring DBP residues. The ATP molecule, bound in the cleft between the subdomains 2 and 4, and nearly all of the ATP surrounding residues in DBP–actin have the same orientation and
144
ILSE BOGAERTS, CHRISTEL VERBOVEN, HUGO VAN BAELEN, AND ROGER BOUILLON
TABLE II
Intermolecular Contacts between Actin and DBP Residues
α-Actin
DBP
Residue
Atom
Residue
Atom
Distance (Å)
LYS 113 LYS 113 TYR 143 TYR 143 ALA 144 SER 145 SER 145 GLY 146 GLY 146 GLY 146 GLY 146 ARG 147 ARG 147 ARG 147 ARG 147 ARG 147 ARG 147 THR 148 THR 148 TYR 166 TYR 166 TYR 166 TYR 166 GLU 167 GLU 167 GLU 167 TYR 169 TYR 169 LEU 171 HIS 173 HIS 173 MET 176 GLU 276 TYR 279 TYR 279 ASN 280 ASN 280 ASN 280 MET 283 MET 283 MET 283 LYS 284
NZ NZ CG OH O O C O CA O O NH2 NH1 NH1 NH1 NH1 NH1 CG2 CG2 CE2 OH CD2 CD2 O OE1 CG CE1 OH CD1 NE2 CD2 SD CG OH CD2 OD1 ND2 OD1 O CB CE NZ
SER 278 ASP 282 LEU 184 LEU 188 THR 180 ASN 125 THR 180 ASN 125 THR 180 PHE 183 ARG 187 GLU 122 PRO 123 THR 124 ASN 125 ILE 128 ARG 187 ARG 187 LEU 188 LEU 195 THR 198 VAL 294 PHE 298 LYS 191 SER 194 LEU 195 GLU 286 CYS 295 PHE 298 PHE 298 PRO 300 THR 398 THR 400 PHE 399 THR 400 THR 398 PHE 399 THR 400 TYR 394 PHE 399 SER 434 SER 395
OG OD2 CD1 CD2 CG2 OD1 CG2 OD1 O CB NE O O C ND2 CD1 NH2 CB CD2 CD1 CG2 CG2 CE2 NZ OG CD1 OE1 SG CD2 O CG CG2 CG2 CE2 CG2 CB N OG1 OH CB CB O
3.51 3.49 3.64 3.87 3.21 3.45 3.87 3.14 3.65 3.57 3.25 3.59 2.92 3.64 2.84 3.80 3.23 3.93 3.85 4.07 3.25 4.07 3.72 2.45 2.63 3.34 3.85 3.28 3.81 4.02 3.32 4.05a 4.20 3.67b 4.01a 3.19 3.51 3.00 3.61 3.89 3.45 3.76 (Continued)
145
CHAPTER 9 New Aspects of DBP
TABLE II Intermolecular Contacts between Actin and DBP Residues—Cont’d α-Actin
DBP
Residue
Atom
Residue
LYS 284 ASP 286 ASP 286 ASP 286 ILE 287 ILE 287 ILE 287 ILE 287 ASP 288 ILE 289 ARG 290 ARG 290 LYS 291 LYS 291 LYS 291 ALA 295 LYS 328 LYS 328 ILE 345 LEU 346 LEU 349 LEU 349 LEU 349 THR 351 THR 351 MET 355 HIS 371
NZ OD1 OD1 OD1 CD1 CG2 CG2 CD1 OD2 CD1 NH2 NH2 CD NZ NZ CB NZ NZ CG2 CD1 CD1 CD1 CD1 OG1 CG2 CE CB
ASN 397 TYR 297 ASN 440 ASN 440 PHE 117 SER 434 ASN 435 PRO 442 ARG 202 PHE 298 TYR 394 SER 434 PRO 118 TYR 120 TYR 151 VAL 121 PRO 123 GLU 127 LEU 184 LEU 188 LEU 184 LYS 185 LEU 188 LYS 185 GLN 189 LYS 287 GLN 285
β- or γ-Actin
Atom O CE2 CG ND2 CE1 O OD1 CD NE CZ OH O O O OH CG1 CA OE1 CD2 CD1 CG CA CD1 CG CD CE OE1
Distance (Å) 3.30 3.74 3.60 2.97 3.19a 3.48b 3.79b 3.26a 3.26 4.03 4.07 2.87 3.36 2.79 2.83 3.58 3.94 3.19 3.70 4.06 4.06 4.13 3.99 4.19 3.74 4.20 2.89
DBP
Residue
Atom
Residue
Atom
Distance (Å)
CYS 272 PHE 279 VAL 287 VAL 287
SG CE2 CG1 CG1
THR 400 PHE 399 SER 434 ASN 435
OG1 CD2 O OD1
3.36c 3.55 3.40 3.66
a
These interactions are absent in DBP–β-actin or DBP–γ-actin complexes. residues are mutated in β- or γ-actin and have different interactions with neighboring DBP residues. Their interactions are shown in the second part of the table. cThis interaction is not present in DBP–α-actin. bThese
146
ILSE BOGAERTS, CHRISTEL VERBOVEN, HUGO VAN BAELEN, AND ROGER BOUILLON
conformation as found for all other known ATP–actin structures. Although DBP interacts with the C-terminal part of actin, located in subdomain 1, the C-terminal actin residues 373 to 375 are disordered in DBP–actin and show no electron density. Similar to actin-GS1 [32] and actin-GS4–6 [33], residues 41 to 50 (i.e., the loop that forms a β-sheet with DNase I in the actinDNase I structure [34] are also disordered in our structure. According to limited proteolysis experiments performed with DBP [12] the actin binding site is located in its C-terminal part, and more specifically residues 350 to 403 (of domains II and III) are involved in actin binding. However, our DBP–actin structure shows that all three DBP domains interact with actin (Fig. 6, Table II). Besides residues 394–395 and 397–400, both part of the biochemically identified actin-binding region (residues 350–403), an additional 35 residues from other DBP regions also ascertain the interaction. Only minor folding differences are observed between the DBP conformations in the presence or absence of actin. The rms difference of 1.86 Å, obtained from superimposing all common α-carbons of the DBP–actin and of the DBP (PDB code 1J78) [1] structures, can mainly be ascribed to a different orientation of domain I and to a different folding in some loop regions. The domain by domain superposition gives lower rms difference values: 1.40 Å, 1.27 Å and 0.46 Å for domains I, II, and III, respectively. The larger rms difference value for domain I is caused by the interaction with actin and by crystal packing differences (Fig. 7). In order to allow the interaction between solvent-exposed hydrophobic DBP residues on the C-terminal helix of domain I (amino acid region 180–188) and the hydrophobic cleft formed by actin subdomains 1 and 3, the C-terminal part of DBP domain I has to readjust its orientation slightly (Fig. 7). In agreement with previous biochemical observations [25,36], the DBP–actin structure proves that the bound actin and the observed folding differences in domain I do not hinder the binding of vitamin D3 compounds to the vitamin D–binding site of DBP (Fig. 7). All presently missing residues in the DBP structure (98–104; 318–322; 361–364), belonging to regions that are poorly defined in the electron density map, are located opposite of the actin binding site. The DBP residues involved in actin binding are relatively well conserved between different species in all presently known DBP sequences. Together with the highly conserved amino acid sequence of actin throughout evolution (e.g., human and rabbit skeletal muscle actin are 100% identical) it explains why the
Neighboring DBP
Actin
25OHD3
Domain I DBP-25OHD3 Domain I DBP-actin
FIGURE 7
The superposition of the domains I of DBP–actin and of the DBP–25-hydroxyvitamin D3 (DBP–25OHD3) complex [1]. Domain I of DBP–actin is shown in dark grey, actin in light grey, domain I of DBP-25OHD3 in medium grey, and a neighboring DBP (at the top) present in the DBP–25OHD3 crystal is also in dark gray. Some of the interacting residues are shown in ball-and-stick representation as well as the 25-hydroxyvitamin D3. The C-terminal part of domain I (helices 7–10) of DBP–actin is shifted toward the actin to allow favorable contacts between DBP helix 10 and the cleft between actin subdomains 1 and 3. Owing to the presence of a neighboring molecule in the DBP-25OHD3 crystal, the N-terminal part of domain I (helices 1–6; i.e., the vitamin D binding site) is shifted toward this neighbor to make intermolecular hydrogen bonds and hydrophobic contacts. This superposition also illustrates that the binding of a vitamin D3 compound to DBP in the presence of actin is possible.
actin-binding property of DBP is universally observed in vertebrates. In comparison with α-actin, a twofold decrease of DBP binding affinity for the nonmuscle actin isoforms, β-actin and γ-actin, was observed [36]. The actin in this DBP–actin structure originates from rabbit muscle, i.e., α-actin. Among the α/β or α/γ sequence differences, the residues Met/Leu 176, Ala/Cys 272, Tyr/Phe 279, and Ile/Val 287 are located in the DBP binding site and the residues Asn/Thr 297, Met/Leu 299, and Thr/Ser 358 are situated in its surroundings. However, in silico replacement of these residues in the DBP–actin structure does not result in any unfavorable contacts (Table II), and there is no indication that the binding interface should fold differently to accommodate the replaced amino acids. Even, the interaction surface between β- or γ-actin and DBP, using the modeled DBP–actin β- or γ-complex for the calculations, remains similar (3609 Å2 compared to 3607 Å2 for DBP–α-actin). The lower affinity of DBP for both β- and γ-actin can only be ascribed to the loss of hydrophobic interactions between DBP residues 117 and 442 and the β or γ-actin residue 287, between DBP residue 398
147
CHAPTER 9 New Aspects of DBP
and the β- or γ-actin residue 176, and between DBP residue 400 and the β- or γ-actin residue 279 (Table II). 3. THE DBP–ACTIN BINDING INTERFACE COMPARED OTHER ACTIN COMPLEXES
TO
Several structures of G-actin in complex with an actin-binding protein have already been elucidated: actin–DNase I [34], actin–profilin [37], actin–GS1 [32], and actin–GS4-6 [33]. The actin surface buried by the complex formation with DBP is 3600 Å2, whereas the buried surfaces for the actin–DNase I, actin–GS1, actin–GS4-6, and actin–profilin complexes are 1800 Å2, 2100 Å2, 2100 Å2, and 2000 Å2, respectively. From all these complexes, DBP has the largest binding interface, indicating that the DBP–actin interaction is not merely fortuitous. The DNase I binding site of actin, consisting of residues from subdomains 2 and 4, is located opposite to the DBP binding site (subdomains 1 and 3). This corroborates the observation that the ternary complex of DBP, DNase I, and actin can be formed in vitro [25,38]. The interaction surfaces of profilin, gelsolin segment 1, and gelsolin segments 4–6 are all located at the same side of actin as the DBP binding site and are all partially overlapping. This structural comparison also demonstrates how DBP bound to actin obstructs its further interaction with profilin, as previously observed [39]. Gelsolin, consisting of six segments of 120 to 130 amino acids each (S1 to S6), has been reported to nucleate growth through the binding of two actin monomers in an appropriate orientation. DBP can prevent re-formation of actin filaments by reducing the effective G-actin concentration. The observation that DBP is capable of displacing one actin from the ternary gelsolin:actin (1:2) complex may also play a role in inhibiting the nucleation [40]. However, the current understanding of the structural organization of the gelsolin:actin (1:2) complex does not permit us to explain how DBP can remove one actin from it. For, in the proposed model of this ternary complex [33], the DBP binding interface is blocked in both actins either by segment 1 or by segment 4 of gelsolin. However, another study, involving cross-linking experiments with different gelsolin constructs, indicated that a ternary complex with the two actins in appropriate orientation for filament growth [41] is formed only with gelsolin segments 2–6, whereas with the whole gelsolin the actin monomers are in an antiparallel orientation as in the lower actin dimer. It is noteworthy that DBP, gelsolin segment 1, and gelsolin segment 4–6 have a hydrophobic helix that interacts with the hydrophobic patch in the cleft between subdomains 1 and 3 of actin. It can be postulated that other proteins binding actin at
the barbed end side and having an appropriate helix with solvent accessible hydrophobic residues, will bind to actin in such an orientation that this helix is also located in the cleft between subdomains 1 and 3. This may be the case for the N-terminal helix of the actin sequestering β-thymosins. Such an orientation of this helix would explain the cross-links between thymosin β 4 and actin: Lys 3–Glu 167 and Lys 18–Asp 1 [42]. Moreover, it was previously suggested that thymosin β 4 interacts with actin through a patch of hydrophobic residues located on the N-terminal helix [43]. This hypothesis also supports the previously suggested idea that the α3-helix of cofilin [44], an actin-severing protein, could interact with actin in a similar manner to its corresponding α-helix in gelsolin. 4. THE EXCELLENT ACTIN-SEQUESTERING PROPERTIES OF DBP
Actin monomers polymerize into double helical filaments (Fig. 8a) twisting around each other [45]. In contrast to other types of polymerization, the actin polymerization is not linear but involves a nucleation step, with a nucleus consisting of three subunits. The resulting microfilaments are polarized: the affinity for addition of new actin monomers is different (approximately 10-fold) at both ends. This can result, at intermediate actin monomer concentrations, in a polymerization of the filament at one end (the “fast growing” or “barbed” end) and a simultaneous depolymerization at the other end (the “slow growing” or “pointed” end) [46]. DBP binds G-actin at the side corresponding to the barbed or fast-growing end of actin in the F-actin filament. It thereby occludes the actin residues 166 to 169 and 286 to 289, all responsible for the longitudinal interactions between actin subunits within one F-actin strand [45]. Moreover, both regions are part of a hydrophobic pocket, which is postulated to be essential for the insertion of a hydrophobic plug from an actin subunit of the opposite actin strand. Although DBP does not make an immediate contact with actin residues 110 to 112, which are involved in interactions with an actin subunit of the opposite F-actin strand, its large volume sterically hinders the approach of the “opposite strand” subunit. The presence of DBP on G-actin therefore prevents the binding of two actin subunits: one that belongs in F-actin to the same strand, the other to the opposite strand (Fig. 8a). Hence, DBP not only blocks elongation of actin at this side, but also prohibits actin nucleation. The presence of DBP on actin seems not to hinder the interaction with other actin molecules at the pointed end side (Fig. 8a). Filament growth is, however, impossible. In conditions where the molar G-actin concentration is less than or equal to that of DBP, all G-actin subunits will be
148
ILSE BOGAERTS, CHRISTEL VERBOVEN, HUGO VAN BAELEN, AND ROGER BOUILLON
A
(-)
B
(+) FIGURE 8 The binding of DBP to actin prevents actin-filament formation. (A) Superposition of the actin from the DBP–actin complex with the central actin subunit of the F-actin model [45] (present in PDB file 1alm) [58]. The five actin subunits have different grey scales. The presence of DBP (dark grey) in the DBP–actin complex prevents the binding of two subunits, which would belong in an actin filament to opposite strands: the two bottom actin subunits shown in light grey. For clarity reasons, the actin from the DBP–actin is not shown. The + and − signs indicate the fast-growing (barbed) and slow-growing (pointed) ends of F-actin, respectively. (B) Different DBP–actin complexes with the DBP always shown in dark grey and the actin in the same grey scales as the central actin subunit and the pointed-end actin subunits of (A). The presence of the bulky DBP on all actin subunits prohibits the approach of these actin subunits to another actin subunit.
captured and blocked by DBP (Fig. 8b). In addition, the presence of DBP prevents the formation of the so-called lower dimer. In this actin dimer model, the two actin molecules are in antiparallel orientation, a prerequisite for enabling actin polymerization [47]. 5. IMPLICATIONS FOR THE ACTIN-SCAVENGER SYSTEM
The observed perfect fit between actin and DBP, as illustrated here, provides the structural basis for the important role of DBP in the extracellular actin-scavenger system. Furthermore, although DBP
and gelsolin are both present in large concentrations (µM) in blood, their capacity to scavenge actin may still be overwhelmed during massive cell injury. Indeed, the saturation of the actin-scavenger system leads to thrombi formation and microangiopathy [48], and excessive amounts of actin in the circulation may lead to a condition resembling “multiple organ dysfunction syndrome.” This syndrome was found to be associated with reduced serum DBP levels [49,50]. Moreover, DBP serum concentrations are reduced in some patients with hepatic failure [49,51]. The DBP concentrations not only have some value in predicting
149
CHAPTER 9 New Aspects of DBP
survival from hepatic failure [52,53] but also may identify high-risk patients after multiple trauma [54]. Consequently, the DBP–actin structure, revealing the DBP residues essential for the interaction with actin and for the actin sequestering, is of major importance for the development of a drug that could be applied in the pathological conditions just mentioned. Despite intensive analysis no DBP-deficient individual has been identified, leading to the suggestion that certain functions of DBP might be essential for survival. However, viable DBP knockout mice have been generated [55]. These mice were only used for analysis of DBP’s role in vitamin D metabolism and action. Although DBP knockout mice are viable in normal laboratory conditions, it might well be that the role of DBP in the actin-scavenger system is confined to cases of severe tissue damage in which DBP is essential for survival. This would be in line with our striking structural observations and the perfect conservation of DBP–actin interaction during the evolution of vertebrates. In the actin–GS1 [32], actin–GS4-6 [33], actin– profilin [37], and DBP–actin complexes and in actin covalently linked to tetramethylrhodamine-5-maleimide [56] (actin-TMR), actin filament formation is blocked at the barbed-end side. The structures of all these complexes illustrate that all these actin ligands block the cleft between subdomains 1 and 3. Even a small molecule such as TMR is able to prevent filament formation. Intracellular sequestering proteins such as thymosin β4 (5 kDa) and profilin (15 kDa) are rather small compared to their extracellular counterpart, DBP (51 kDa). These remarkable differences in size of the actin interaction surfaces between the extra- and intracellular sequestering proteins might be related to the fact that in the intracellular compartment an equilibrium between G-actin and F-actin has to be maintained. In extracellular space F-actin formation has to be prevented at any cost. Therefore, DBP, a better actin-sequestering protein, is present in high concentrations in the extracellular compartment. 6. COMPARISON OF THE THREE REPORTED DBP–ACTIN STRUCTURES
Although our DBP–actin structure was obtained from a crystal grown in a different space group (P21) to that of two other very recently reported DBP–actin structures (space group P212121; PDB codes 1kxp [35] and 1lot [57]), the structures are very similar. The Cα atoms of our DBP–actin structure and the 1lot structure superimpose with an rms difference of 0.791 Å (781 atoms superimposed), while the Cα atoms of our DBP–actin structure and the 1kxp structure superimpose
with an rms difference of 0.709 Å (777 atoms superimposed). The differences between these three structures mainly reside in the loop regions and the N and C termini and are mainly the result of different lattice contacts. In the paper of Otterbein and co-workers, the possibility of a conformational change within the C-terminal actin residues 365–375 is proposed, since in their structure (1kxp) this region is poorly defined in the electron-density map. However, the seven extra residues 365–371 in our structure assume a similar conformation as in the other reported actin structures. The 1kxp structure and our structure have Mg2+-ATP at their nucleotide-binding site, while the 1lot structure has Ca2+-ATP. The presence of a different ion does not alter the conformation of ATP or the conformations of the surrounding amino acids. Only the water coordination of the ions is different: Ca2+ has a pentagonal bipyramidal coordination with five water molecules and two O atoms from the ATP β- and γ-phosphates, compared with the nearly octahedral coordination of Mg2+, which has one less water molecule. In our DBP–actin structure and the 1lot structure, His 73 of the actin molecule is methylated, while in the 1kxp structure this is not the case. Although the peptide bond between Glu 72 and His 73 of actin has the same orientation in 1lot as in our structure, there is no hydrogen bond in 1lot between the main-chain carbonyl group of Glu 72 and the side chain of Arg 183. In 1kxp, the orientation of this peptide bond is different, making such a hydrogen bond impossible.
IV. A STRUCTURAL EXPLANATION FOR THE UNIQUE FUNCTIONS OF VITAMIN D–BINDING PROTEIN WITHIN ITS GENE FAMILY The unique functions of DBP within this protein family must arise from its distinctive fold. The different helical arrangement observed in domain I of DBP allows for the binding of vitamin D3 ligands. Superposition of helix 2 of domain I of DBP with that of HSA illustrates that helix 4 of HSA makes a closer contact with helices 2 and 3, leaving no room for a vitamin D3 molecule. Compared to HSA, α-fetoprotein, and afamin, the third domain of DBP is largely truncated at the C terminus. Moreover, the comparison of the DBP and HSA structures revealed remarkably differing orientations of the three domains in both structures [1]. The shorter domain III of DBP together with its distinctive fold allows DBP to bind actin in contrast to the other family members (Fig. 9).
150
ILSE BOGAERTS, CHRISTEL VERBOVEN, HUGO VAN BAELEN, AND ROGER BOUILLON
A
DBP domain I Actin HSA domain III
HSA domain I
DBP domain III DBP domain II HSA domain II
B
FIGURE 9
Superposition of the DBP from the DBP-actin structure with the HSA structure [2], illustrating that HSA cannot bind actin. (A) Stereoview of the overall structure of HSA (in gray) positioned on the DBP–actin structure based on the superposition of domains II of DBP and HSA. Only the actin subdomains 1 and 3 (light grey) are shown. Domain I of DBP is shown in dark grey, domain II, in grey and domain III in medium grey. The orientation of domains I–III in DBP and HSA is completely different. Consequently, the residues forming the actin-binding interface in DBP have a totally different arrangement in HSA. (B) Close-up view of (A) illustrating that owing to their different orientation, HSA helices 9 and 10 of domain I and helices 1–3 of domain III collide with parts of the actin structure.
References 1. Verboven C, Ragijns A, De Maeyer M, Van Baelen H, Bouillon R, De Ranter C 2002 A structural basis for the unique binding features of the human vitamin D-binding protein. Nat Struct Biol 9:131–136. 2. He XM, Carter DC 1992 Atomic-structure and chemistry of human serum-albumin. Nature 358:209–215. 3. Sugio S, Kashima A, Mochizuki S, Noda M, Koboyashi K 1999 Crystal structure of human serum albumin at 2.5 angstrom resolution. Protein Eng 12:439–446.
4. Cooke NE, Haddad JG 2003 In: “Vitamin D” Feldman D, Glorieux FH, Pike JW (eds) Vitamin D; 1st ed. Academic Press, Orlando, FL, pp. 87–101. 5. Curry S, Mandelkow H, Brick P, Franks N 1998 Crystal structure of human serum albumin complexed with fatty acid reveals an asymmetric distribution of binding sites. Nat Struct Biol 5:827–835. 6. Williams MH, Vanalstyne EL, Galbraith RM 1988 Evidence of a novel association of unsaturated fatty-acids with Gc (vitamin-D-binding protein). Biochem Biophys Res Commun 153:1019–1024.
CHAPTER 9 New Aspects of DBP
7. Ena JM, Esteban C, Perez MD, Uriel J, Calvo M 1989 Fattyacids bound to vitamin-D-binding protein (Dbp) from human and bovine sera. Biochem Int 19:1–7. 8. Bouillon R, Xiang DZ, Convents R, Vanbaelen H 1992 Polyunsaturated fatty-acids decrease the apparent affinity of vitamin-D metabolites for human vitamin D–binding protein. J Steroid Biochem Mol Biol 42:855–861. 9. Bhattacharya AA, Grune T, Curry S 2000 Crystallographic analysis reveals common modes of binding of medium and long-chain fatty acids to human serum albumin. J Mol Biol 303:721–732. 10. Barton GJ 1993 Alscript—a tool to format multiple sequence alignments. Protein Eng 6:37–40. 11. White P, Cooke N 2000 The multifunctional properties and characteristics of vitamin D-binding protein. Trends Endocrinol Metab 11:320–327. 12. Haddad JG, Hu YZ, Kowalski MA, Laramore C, Ray K, Robzyk P, Cooke NE 1992 Identification of the sterol-binding and actin-binding domains of plasma vitamin-D bindingprotein (Gc-globulin). Biochemistry 31:7174–7181. 13. Bouillon R, Okamura WH, Norman AW 1995 Structure– function relationships in the vitamin-D endocrine system. Endocr Rev 16:200–257. 14. Rochel N, Wurtz JM, Mitschler A, Klaholz B, Moras D 2000 The crystal structure of the nuclear receptor for vitamin D bound to its natural ligand. Mol Cell 5:173–179. 15. Ashcroft DM, Po ALW, Williams HC, Griffiths CEM 2000 Systematic review of comparative efficacy and tolerability of calcipotriol in treating chronic plaque psoriasis. Br Med J 320:963–967. 16. van den Bemd GJCM, Pols HAP, van Leeuwen JPTM 2000 Anti-tumor effects of 1,25-dihydroxyvitamin D-3 and vitamin D analogs. Curr Pharm Design 6:717–732. 17. Casteels K, Waer M, Laureys J, Valckx D, Depovere J, Bouillon R, Mathieu C 1998 Prevention of autoimmune destruction of syngeneic islet grafts in spontaneously diabetic nonobese diabetic mice by a combination of a vitamin D-3 analog and cyclosporine. Transplantation 65:1225–1232. 18. Cantorna MT, Hayes CE, DeLuca HF 1996 1,25-Dihydroxyvitamin D-3 reversibly blocks the progression of relapsing encephalomyelitis, a model of multiple sclerosis. Proc Natl Acad Sci USA 93:7861–7864. 19. Dusso AS, Negrea L, Gunawardhana S, Lopezhilker S, Finch J, Mori T, Nishii Y, Slatopolsky E, Brown AJ 1991 On the mechanisms for the selective action of vitamin-D analogs. Endocrinology 128:1687–1692. 20. Bouillon R, Allewaert K, Xiang DZ, Tan BK, Vanbaelen H 1991 Vitamin-D analogs with low affinity for the vitamin-D binding-protein—Enhanced in vitro and decreased in vivo activity. J Bone Miner Res 6:1051–1057. 21. Brown AJ 2000 Mechanisms for the selective actions of vitamin D analogues. Curr Pharm Design 6:701–716. 22. Tocchini-Valentini G, Rochel N, Wurtz JM, Mitschler A, Moras D 2001 Crystal structures of the vitamin D receptor complexed to superagonist 20-epi ligands. Proc Natl Acad Sci USA 98:5491–5496. 23. Lee WM, Galbraith RM 1992 Mechanisms of disease—the extracellular actin-scavenger system and actin toxicity. N Engl J Med 326:1335–1341. 24. Vasconcellos CA, Lind SE 1993 Coordinated inhibition of actin-induced platelet-aggregation by plasma gelsolin and vitamin-D-binding protein. Blood 82:3648–3657. 25. Van Baelen H, Bouillon R, De Moor P 1980 Vitamin-Dbinding protein (Gc-globulin) binds actin. J Biol Chem 255: 2270–2272.
151 26. Lind SE, Smith DB, Janmey PA, Stossel TP 1986 Role of plasma gelsolin and the vitamin-D-binding protein in clearing actin from the circulation. J Clin Invest 78:736–742. 27. Goldschmidt-Clermont PJ, Van Baelen H, Bouillon R, Shook TE, Williams MH, Nel AE, Galbraith RM 1988 Role of group-specific component (vitamin-D binding-protein) in clearance of actin from the circulation in the rabbit. J Clin Invest 81:1519–1527. 28. Dueland S, Nenseter MS, Drevon CA 1991 Uptake and degradation of filamentous actin and vitamin-D-binding protein in the rat. Biochem J 274:237–241. 29. Herrmanssdoerfer AJ, Heeb GT, Feustel PJ, Estes JE, Keenan CJ, Minnear FL, Selden L, Giunta C, Flor JR, Blumenstock FA 1993 Vascular clearance and organ uptake of G- and F-actin in the rat. Am J Physiol 265:G1071–G1081. 30. Verboven C, Bogaerts I, Waelkens E, Rabijns A, Van Baelen H, Bouillon R, De Ranter C 2003 Actin-DBP: the perfect structural fit? Acta Crystallogr D Biol Crystallogr 59: 263–273. 31. Houmeida A, Hanin V, Constans J, Benyamin Y, Roustan C 1992 Localization of a vitamin-D-binding protein-interaction site in the COOH-terminal sequence of actin. Eur J Biochem 203:499-503. 32. McLaughlin PJ, Gooch JT, Mannherz HG, Weeds AG 1993 Structure of gelsolin segment-1-actin complex and the mechanism of filament severing. Nature 364:685–692. 33. Robinson RC, Mejillano M, Le VP, Burtnick LD, Yin HL, Choe S 1999 Domain movement in gelsolin: a calciumactivated switch. Science 286:1939–1942. 34. Kabsch W, Mannherz HG, Suck D, Pai EF, Holmes KC 1990 Atomic structure of the actin–DNase-I complex. Nature 347:37–44. 35. Otterbein LR, Cosio C, Graceffa P, Dominguez R 2002 Crystal structures of the vitamin D-binding protein and its complex with actin: structural basis of the actin-scavenger system. Proc Nat Acad Sci USA 99:8003–8008 36. Mcleod JF, Kowalski MA, Haddad JG 1989 Interactions among serum vitamin-D binding-protein, monomeric actin, profilin, and profilactin. J Biol Chem 264:1260–1267. 37. Schutt CE, Myslik JC, Rozycki MD, Goonesekere NCW, Lindberg U 1993 The structure of crystalline profilin betaactin. Nature 365:810–816. 38. Goldschmidt-Clermont PJ, Galbraith RM, Emerson DL, Marsot F, Nel AE, Arnaud P 1985 Distinct sites on the the G-actin molecule bind group-specific component and deoxy-ribonuclease-I. Biochem J 228:471–477. 39. Goldschmidt-Clermont PJ, Vanalstyne EL, Day JR, Emerson DL, Nel AE, Lazarchick J, Galbraith RM 1986 Groupspecific component (vitamin-D binding-protein) prevents the interaction between G-actin and profilin. Biochemistry 25: 6467–6472. 40. Janmey PA, Stossel TP, Lind SE 1986 Sequential binding of actin monomers to plasma gelsolin and its inhibition by vitamin-D-binding protein. Biochem Biophy Res Commun 136:72–79. 41. Hesterkamp T, Weeds AG, Mannherz HG 1993 The actin monomers in the ternary gelsolin–2-actin complex are in an antiparallel orientation. Eur J Biochem 218:507–513. 42. Safer D, Sosnick TR, Elzinga M 1997 Thymosin β(4) binds actin in an extended conformation and contacts both the barbed and pointed ends. Biochemistry 36:5806–5816. 43. Van Troys M, Dewitte D, Goethals M, Carlier MF, Vandekerckhove J, Ampe C 1996 The actin binding site of thymosin β 4 mapped by mutational analysis. EMBO J 15:201–210.
152
ILSE BOGAERTS, CHRISTEL VERBOVEN, HUGO VAN BAELEN, AND ROGER BOUILLON
44. Fedorov AA, Lappalainen P, Fedorov EV, Drubin DG, Almo SC 1997 Structure determination of yeast cofilin. Nat Struct Biol 4:366–369. 45. Holmes KC, Popp D, Gebhard W, Kabsch W 1990 Atomic model of the actin filament. Nature 347:44–49. 46. Wegner A 1976 Head to tail polymerization of actin. J Mol Biol 108:139–150. 47. Steinmetz MO, Stoffler D, Hoenger A, Bremer A, Aebi U 1997 Actin: from cell biology to atomic detail. J Struct Biol 119:295–320. 48. Vercaeren I, Winderickx J, Devos A, Peeters B, Heyns W 1993 An effect of androgens on the length of the poly(A)-tail and alternative splicing cause size heterogeneity of the messenger ribonucleic acids encoding cystatin-related protein [corrected and republished with original paging, article originally printed in 1992 Endocrinology 131(6):2496–2502]. Endocrinology 132:2496–2502. 49. Schiodt FV, Ott P, Bondesen S, Tygstrup N 1997 Reduced serum Gc-globulin concentrations in patients with fulminant hepatic failure: association with multiple organ failure. Crit Care Med 25:1366–1370. 50. Dahl P, Schiodt FV, Klaer T, Ott P, Bondesen S, Tygstrup N 1998 Serum Gc-globulin in the early course of multiple trauma. Crit Care Med 26:285–289. 51. Schiodt FV, Bondesen S, Tygstrup N 1995 Serial measurements of serum Gc-globulin in acetaminophen intoxication. Eur J Gastroenterol Hepatol 7:635–640.
52. Lee WM, Galbraith RM, Watt GH, Hughes RD, McIntire DD, Hoffman BJ, Williams R 1995 Predicting survival in fulminant hepatic-failure using serum Gc protein concentrations. Hepatology 21:101–105. 53. Schiodt FV, Bondesen S, Petersen I, Dalhoff K, Ott P, Tygstrup N 1996 Admission levels of serum Gc-globulin: predictive value in fulminant hepatic failure. Hepatology 23:713–718. 54. Dahl B, Schiodt FV, Nielsen M, Kiaer T, Williams JG, Ott P 1999 Admission level of Gc-globulin predicts outcome after multiple trauma. Injury—Int J Care Injured 30: 275–281. 55. Safadi FF, Thornton P, Magiera H, Hollis BW, Gentile M, Haddad JG, Liebhaber SA, Cooke NE 1999 Osteopathy and resistance to vitamin D toxicity in mice null for vitamin D binding protein. J Clin Invest 103:239–251. 56. Otterbein LR, Graceffa P, Dominguez R 2001 The crystal structure of uncomplexed actin in the ADP state. Science 293:708–711. 57. Head JF, Swamy N, Ray R 2002 Crystal structure of the complex between actin and human vitamin D–binding protein at 2.5 angstrom resolution. Biochemistry 41: 9015–9020. 58. Mendelson R, Morris EP 1997 The structure of the actomyosin subfragment 1 complex: results of searches using data from electron microscopy and x-ray crystallography. Proc Natl Acad Sci USA 94:8533–8538.
CHAPTER 10
Endocytic Pathways for 25-(OH) Vitamin D3 THOMAS E. WILLNOW ANDERS NYKJAER
Division of Molecular Cardiovascular Research Max-Delbrueck-Center for Molecular Medicine D-13125 Berlin, Germany Institute of Medical Biochemistry, University of Aarhus, DK-8000 Aarhus C, Denmark
I. Introduction II. Renal Endocytosis of 25-(OH) Vitamin D3
III. Conclusions References
I. INTRODUCTION
II. RENAL ENDOCYTOSIS OF 25-(OH) VITAMIN D3
In the 1970s, Fraser and Kodicek [1] and DeLuca and co-workers [2] were the first to demonstrate that the kidney is responsible for the conversion of 25-(OH) vitamin D3 to a more polar metabolite that acts as a potent regulator of calcium transport in the intestinal mucosa. Shortly thereafter, this metabolite was identified as 1,25-(OH)2 vitamin D3 [3,4]. Conversion was shown to take place in the epithelial cells of the proximal convoluted tubules (PCT) [5,6] by the action of 25-hydroxyvitamin D3-1α-hydroxylase, a cytochrome P450 enzyme located in the mitochondria [7]. A large body of work has firmly established the pivotal role of the PCT in the conversion of 25-(OH) vitamin D3 to 1,25-(OH)2 vitamin D3, and the regulatory steps involved in this process [reviewed in 8,9]. The route, however, that delivers the precursor to PCT cells for metabolism has been less well characterized. In the circulation, 25-(OH) vitamin D3 is transported bound to the carrier vitamin D–binding protein (DBP), also known as group-specific component of serum (Gc-globulin) [10]. Similar to other steroids, 25-(OH) vitamin D3 can dissociate from its carrier and passively diffuse through the plasma membrane of target cells. In PCT, the uptake of the free vitamin is believed to proceed from the circulation via the basolateral site of the epithelium [11,12]. Recent studies have identified an alternative route for delivery of 25-(OH) vitamin D3 to the PCT. It involves endocytosis of 25-(OH) vitamin D3/DBP complexes via endocytic receptors expressed on the luminal surface of the epithelium. This chapter describes the role of endocytic pathways for 25-(OH) vitamin D3/DBP complexes in vitamin D metabolism, and the molecular components involved in this process. VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
A. Physiology 1. 25-(OH) VITAMIN D3/DBP COMPLEXES ARE SUBJECT TO GLOMERULAR FILTRATION
Patients with nephrotic syndrome suffer from hypocalcemia due to low serum levels of 25-(OH) vitamin D3 and 1,25-(OH)2 vitamin D3. Hypovitaminosis D is caused by an increased urinary loss of 25-(OH) vitamin D3 bound to DBP, depleting the plasma vitamin D pool [13–15]. The same phenotype can be observed in nephrotic rats [16]. Because no 25-(OH) vitamin D3 or DBP is present in the urine of control subjects, increased glomerular permeability in nephrotic syndrome was held responsible for an abnormal filtration of 25-(OH) vitamin D3/DBP complexes. In the healthy kidney, 25-(OH) vitamin D3/DBP complexes were believed to be excluded from glomerular filtration. This assumption has to be revised because a number of studies now demonstrate that circulating 25-(OH) vitamin D3/DBP complexes are normally filtered through the glomerulus and retrieved by PCT cells from the primary urine. Foremost, this fact has been demonstrated in patients suffering from proximal tubular resorption deficiencies (i.e, Fanconi syndrome) [17] such as in Itai–Itai disease (cadmium-induced tubulopathy) [18], Imerslund–Grasbeck syndrome (IGS) (see Section II,C,1,b) [19] or multiple myeloma [20]. In these individuals, the glomerular filtration barrier is intact and only low-molecular-weight metabolites are lost into the urine because of the inability of the PCT to retrieve physiological ligands from the tubule lumen. Patients with tubular resorption deficiencies of various etiologies excrete DBP (Fig. 1A) and 25-(OH) Copyright © 2005, Elsevier, Inc. All rights reserved.
154
THOMAS E. WILLNOW AND ANDERS NYKJAER
A
1
B
2
3
4
5
6
7
8
9
FIGURE 1 Excretion of DBP in patients and in mice with tubular resorption deficiencies. (A) Western blot analysis demonstrates the presence of DBP in urine samples from patients with multiple myeloma (light chain disease) (lane 1), cystinosis (lane 3), Fanconi–Bickel syndrome (lane 4), and Imerslund–Grasbeck syndrome (lane 5), but not from a healthy control subject (lane 2). (Samples provided by D. Müller, Humboldt University Berlin.) (B) DBP excretion is also seen in urine samples from megalindeficient (lanes 6 and 8) but not from wild-type mice (lanes 7 and 9).
vitamin D3 [20], indicating that the vitamin–carrier complex is subject to normal filtration. Thus, the urinary loss of 25-(OH) vitamin D3/DBP in nephrotic syndrome may be explained in part by the excessive amounts of filtered high-molecular-weight proteins that saturate tubular retrieval pathways. Further evidence for a physiological role of glomerular filtration in renal vitamin D metabolism stems from the fact that glomerular filtration rates directly correlate with plasma 1,25(OH)2 vitamin D3 levels [21]. 2. MEGALIN, AN ENDOCYTIC RECEPTOR FOR 25-(OH) VITAMIN D3/DBP
The identity of the tubular retrieval pathway for 25-(OH) vitamin D3/DBP and its significance for vitamin D metabolism was uncovered through studies on megalin, an endocytic receptor abundantly expressed on the brush border (luminal) surface of the PCT cells (Fig. 2) [22,23]. Megalin is a scavenging receptor with broad ligand specificity that constitutes the main pathway for tubular reabsorption of filtered plasma proteins [24]. Mice genetically deficient for this receptor suffer from low-molecular-weight proteinuria and excrete the receptor ligands [20,25–27]. Among other proteins, megalin knockout mice lose DBP into the urine, which is normally retrieved by this receptor (Fig. 1B). No uptake of DBP is seen in megalin-deficient kidneys whereas endocytosis of DBP can easily be detected in renal tissue from control animals (Fig. 2). Both mice with an obligate or with a conditional kidney-specific megalin gene defect were used to evaluate the pathophysiological consequences of urinary
loss of DBP. These studies have highlighted the importance of tubular retrieval pathways for renal and systemic vitamin D metabolism [20,25,28,29]. Megalin-deficient mice excrete approximately 0.3 mg of DBP per day [25]. Because the same amount of protein is retrieved from the glomerular filtrate in a normal kidney in 24 hr, glomerular filtration likely accounts for most of the plasma turnover of the carrier protein [30]. Massive urinary loss of DBP in megalin knockout mice results in a concomitant loss of 25-(OH) vitamin D3 bound to the carrier protein (Table I). The total 25-(OH) vitamin D3 loss amounts to 0.3 to 0.5 ng in 24 hr [25]. On a vitamin D3–enriched diet, the plasma levels of 25-(OH) vitamin D3 and 1,25-(OH)2 vitamin D3 are reduced by more than 70% in the obligate and by 50% in the conditional knockout line (Table I). The less severe phenotype in the latter mouse model is due to the residual megalin activity that is retained in animals with kidney-specific megalin gene deletion (10% of normal levels) [29]. Owing to the dietary supplementation with vitamin D3, megalin-deficient mice can balance the constant urinary loss of 25-(OH) vitamin D3 and exhibit normal plasma calcium, phosphorus, and PTH levels. However, when placed on a vitamin D–depleted chow, the animals suffer from hypocalcemia and hyperparathyroidism (Table I) [29]. Their bones are characterized by a massive increase in osteoid surfaces and by a dramatic reduction in inactive and mineralizing surfaces and in total bone mineral content (Table II). Such features of osteomalacia as a consequence of hypovitaminosis D are also observed in mice with inactivation of the DBP gene [31], highlighting
155
Anti-DBP
Anti-megalin
CHAPTER 10 Endocytic Pathways for 25-(OH) Vitamin D3
FIGURE 2 Immunodetection of megalin and DBP in murine PCT. In wild-type (+/+) PCT cells, megalin (arrowheads) is present on the luminal cell surface where it internalizes filtered DBP, detectable in intracellular vesicles (arrows). No uptake of the carrier is seen in PCT of megalin knockout mice (−/−).
TABLE I Endocrine Parameters in Wild-Type Mice (Control) and in Animals with Obligate (Megalin−/−) or Kidney-Specific Megalin Gene Defect (Conditional) Genotypes Parameters Vitamin D–enriched diet Urine 25-OH D3 (nM/mM creatinine) Plasma 25-OH D3 (nM) 1,25-(OH)2 D3 (pM) PTH (pM) Calcium (mM) Phosphorous (mM) Vitamin D–depleted diet Plasma PTH (pM) Calcium (mM) Phosphorous (mM)
Control
n
Conditional
n
Megalin−/−
n
n.d.
8
10.9 ± 1.9
13
10.4 ± 4.1
6
—
92.8 ± 8.3 120.1 ± 7.0 52.1 ± 4.0 2.3 ± 0.04 3.2 ± 0.21
9 5 5 5 5
48.1 ± 6.4 66.4 ± 14.6 67.0 ± 6.3 2.2 ± 0.04 3.0 ± 0.08
8 4 5 5 5
27.2 ± 8.3 43.2 ± 17.5 68.5 ± 1.8 — —
4 4 3
0.0006 0.009 0.08 0.6 0.56
72.6 ± 4.0 2.2 ± 0.03 3.0 ± 0.08
10 10 10
88.3 ± 1.8 1.8 ± 0.1 3.1 ± 0.1
5 9 9
— — —
P
0.04 0.03 0.3
Data are the mean ± SEM. Statistical significance of differences between values in controls and conditional knockout mice are indicated (P). n, number of animals; n.d., not detectable.
156
THOMAS E. WILLNOW AND ANDERS NYKJAER
TABLE II Bone Morphometric Analysis of Wild Type Mice (Control) or Animals with Kidney-Specific Megalin Gene Defect (Conditional) on a Vitamin D–Depleted Diet (Mean ± SEM; n, number of animals) Parameters Bone mineral content in femur and tibia (mg) Osteoid surface (% of total surface) Inactive surfaces (% of total surface) Mineralizing surface (% of total surface)
Control (n = 5)
Conditional (n = 4)
P
50.4 ± 3.3 3.0 ± 1.7 92.4 ± 1.7 31.8 ± 6.0
29.5 ± 5.0 67.2 ± 17.9 18.7 ± 16.1 2.2 ± 0.9
< 0.01 0.02 < 0.01 < 0.01
the close functional relationship of the endocytic receptor megalin and the carrier DBP in vitamin D metabolism. Further support for a crucial role of the megalin retrieval pathway in uptake of 25-(OH) vitamin D3 comes from studies in rats treated with megalin antagonists. When receptor antagonists are infused into PCT of the rat, no conversion of 25-(OH) vitamin D3 to 1,25-(OH)2 vitamin D3 is seen, indicating that the megalin receptor pathway provides PCT cells with the precursor to produce the active hormone 1,25-(OH)2 vitamin D3 [25].
B. Cell Biology According to current hypotheses, free steroid hormones enter cells by passive diffusion through the
BN16 cells
35 Celluar uptake (% of added tracer)
plasma membrane. In contrast, steroids bound to transport proteins are considered biologically inactive and blocked from cell entry by their carriers [32]. This concept does not apply to the secosteroid 25-(OH) vitamin D3 complexed with DBP as shown in rat and mouse models [25, 29] and in cultured cells [19]. When Brown Norway rat choriocarcinoma (BN16) cells are incubated with 25-(OH) vitamin D3/DBP complexes more than 30% of the complexes are taken up within 4 hr (Fig. 3). The cellular uptake of the steroid is dependent on megalin, which is abundantly expressed in these cells, and can be blocked by a megalin antagonist, the receptor-associated protein (RAP) [33], or by anti-megalin antibodies. Very little uptake of complexed 3H-25-(OH) vitamin D3 is seen in keratinocytes that lack megalin expression (Fig. 3). Similar results have been obtained in other cell lines not expressing the megalin, indicating
30 25-OH-D3 /125I-DBP 25
3H-25-OH -D/DBP 3
20 15 10
Keratinocytes
5 0
FIGURE 3
buffer
α-megalin lgG
sheep lgG
RAP
buffer
Cellular uptake of 25-(OH) vitamin D3/DBP complexes. Brown Norway rat choriocarcinoma (BN16) cells and keratinocytes were incubated for 4 hr with 25-(OH) vitamin D3/DBP complexes labeled on either the steroid (3H-25-(OH) vitamin D3/DBP, open bars) or the protein moiety (25-(OH) vitamin D3/125I-DBP, closed bars). Cellular uptake of the carrier and the vitamin can be detected in megalin-expressing BN16 cells but not in keratinocytes that lack the receptor. Megalin-mediated uptake of complexed 25-(OH) vitamin D3 is blocked by the megalin antagonist receptor-associated protein (RAP) and by sheep anti-megalin antibodies, but not by control sheep IgG.
157
CHAPTER 10 Endocytic Pathways for 25-(OH) Vitamin D3
FIGURE 4
This immunoelectron micrograph depicts internalized DBP (arrows) in a lysosome of a murine PCT cell. (Provided by E. I. Christensen, University of Aarhus.)
that the efficient uptake of metabolites complexed with DBP requires the presence of megalin [34–37]. Like most internalized proteins, DBP is transported to lysosomes for degradation (Fig. 4) [25,38]. The intracellular route taken by 25-(OH) vitamin D3 through the endocytic compartments to the mitochondria is unknown and probably involves intracellular 25-(OH) vitamin D binding proteins (IDBP) (see Chapter 21) [39].
autoantigen in an induced glomerular nephritis model in rats, the passive Heymann nephritis [40]. In this animal model, autoantibodies directed against megalin form immunodeposits in the glomerular wall, causing severe tissue damage and renal failure [40, 41]. Megalin is a giant 600-kDa cell surface protein. Its gene covers 235,000 base pairs on the human chromosome 2q24-q31 [42] and consists of 79 exons [43]. The protein is highly conserved among species and can even be found in primitive organisms such as nematodes [44]. Sequencing of the rat and human megalin cDNA revealed a close structural similarity of this receptor with members of the LDL receptor gene family, a group of multifunctional endocytic receptors [43,45,46]. Similar to other family members, megalin is characterized by a number of structural domains required to perform endocytosis (Fig. 5). Its extracellular domain consists of 36 cysteine-rich complement-type repeats,
Complement-type repeat EGF YWTD
NH2
EGF precursor homology domain
CUB domain EGF-type repeat Amphipathic hellx
Membrane anchor NPXY motif
COOH
C. Molecular Biology Investigations in rats treated with megalin antagonists [25] or in mice with induced megalin gene defect [25,29] have established a central role of this receptor in renal uptake of 25-(OH) vitamin D3 metabolites in vivo. Recent studies have uncovered the molecular details of this pathway. Besides the receptor megalin itself, it involves co-receptors (cubilin), cellular adaptors (disabled-2), and components of the endocytic machinery (voltage-gated chloride channel-5), as well as extracellular (DBP) and intracellular 25-(OH) vitamin D3-binding proteins (IDBP). DBP and IDBP will only be mentioned briefly in this chapter because they are covered elsewhere in this textbook (Chapters 8 and 21, respectively). Other components of the pathway will be described in the following sections. 1. RECEPTORS AND CO-RECEPTORS
a. Megalin Megalin, also known as glycoprotein (gp) 330 or low-density lipoprotein (LDL) receptorrelated protein-2, was initially identified as the
NH2
Membrane
Megalin
COOH
Membrane
Cubilin
FIGURE 5 (Left panel) Structure of megalin. The extracellular domain of the receptor consists of clusters of complement-type repeats and epidermal growth factor (EGF) presursor homology domains. The clusters of complement-type repeats are the site of ligand binding. YWTD repeats in the EGF precursor homology domains are responsible for pH-dependent release of ligands. A single transmembrane domain mediates membrane attachment. NPXY motifs in the cytoplasmic tail are involved in internalization via coated pits and in cytosolic adaptor binding. (Right panel) Structure of cubilin. The cubilin polypeptide consists of 27 CUB domains, a cluster of 8 EGF repeats, and an amphipathic helix. CUB domains are involved in ligand binding. The amino-terminal amphipathic helix mediates peripheral membrane attachment.
158
bil in
ali n
Cu
α-
αM eg
organized in four clusters of seven to 11 repeats. The second cluster represents the main site of ligand binding [47]. All clusters of complement-type repeats are separated by epidermal growth factor (EGF) precursor homology domains composed of EGF-type repeats and tyrosine-tryptophan-threonine-aspartate (YWTD)containing spacer regions. The later sites are involved in pH-dependent release of ligands in endosomes [48]. A 22-amino-acid transmembrane domain mediates membrane attachment of the receptor. The cytoplasmic tail contains three asparagine-proline-X-tyrosine (NPXY) elements that represent signals for coated-pit internalization [49] and binding sites for cytosolic adapter proteins such as Disabled (Dab)-2 [50]. Megalin is expressed in many absorptive epithelia of embryonic and adult tissues, the predominant sites of expression being the yolk sac and the neuroepithelium of early embryos and the PCT and the small intestine in adults [24]. In these tissues, megalin expression is restricted to the apical cell surface and to endosomal compartments in line with a role of this receptor in apical clearance pathways [23]. Many ligands have been identified that bind to megalin in vitro and that are taken up by the receptor into cultured cells or in vivo. By and large, these ligands represent three major classes of molecules: (1) proteases and protease/inhibitor complexes, (2) lipoproteins, and (3) vitamins or hormones bound to carrier proteins [reviewed in 24,51]. Besides 25-(OH) vitamin D3/DBP, the receptor reabsorbs complexes of retinol (vitamin A) with the retinol-binding protein [26] and vitamin B12 with the carrier transcobalamin [27] from the primary urine into PCT, indicating a role as a general retrieval receptor for filtered metabolites. The function of the receptor in other epithelia is less well characterized. Forebrain defects in mice with obligate megalin gene deletion argue for an important role of the protein in brain development [52]. Megalin binds most ligands with moderate affinities (Kd 100–500 nM). Ligand binding is dependent on the presence of calcium that is required to stabilize the three-dimensional structure of the complement-type repeats [53]. DBP binds to megalin with a Kd of 108 nM regardless of whether 25-(OH) vitamin D3 is bound to the carrier or not [25]. b. Cubilin When Nykjaer et al. [19] used DBPaffinity chromatography to purify membrane proteins from renal tissues that bind this carrier, two DBPinteracting proteins were identified (Fig. 6). One was megalin, the other cubilin, a peripheral membrane protein expressed in the PCT and other epithelia [54]. Cubilin was initially identified as the receptor for uptake of intrinsic factor (IF)/vitamin B12 complexes in the
THOMAS E. WILLNOW AND ANDERS NYKJAER
DBP-Column
250
96 68 45 30 1
2
3
4
5
6
7
FIGURE 6 The kidney expresses two DBP receptors. Two proteins can be recovered by DBP-affinity chromatography from renal membrane extracts. Lanes 1–5 depict elution fractions from a DBPcolumn separated by SDS–polyacrylamide gel electrophoresis and stained with silver nitrate. Two high-molecular-weight DBP-binding proteins elute from the column, megalin and cubilin. Lanes 6 and 7 are immunoblots of the fraction in lane 4 incubated with antimegalin (lane 6) or anti-cubilin antisera (lane 7).
intestine and is therefore also known as the intrinsic factor/cobalamin receptor [55]. The human cubilin gene is located on chromosome 10p12.1 [56]. The sequence gives rise to a 460-kDa protein. So far, cubilin has been cloned from humans, rats, and dogs [56–58]. It received its name from a cluster of so called CUB domains that make up most of the receptor polypeptide (Fig. 5). CUB domains are 120 amino acids modules found in components of the complement system, such as C1r/C1s, in the EGF related sea urchin protein UEGF and in the bone morphogenic protein, BMP-1 (hence the name CUB). In addition to 27 CUB domains, cubilin possesses a short amino terminal amphipathic helix and 8 EGF-like repeats [57]. In contrast to prototype endocytic receptors such as megalin, cubilin lacks a membrane anchor and a cytosolic tail and is believed to associate with the plasma membrane via its hydrophobic amino terminal helix [59]. The protein directly binds to megalin and recycles through the endocytic compartments as a co-receptor complex (see Section II,C,1,c). In mice genetically deficient for megalin, cubilin biosynthesis and intracellular transport are significantly impaired [60]. Therefore, PCTs lacking megalin are also functionally deficient for cubilin [60, 61]. Besides in the PCT, cubilin is expressed in several absorptive epithelia including the yolk sac and the ileum.
159
CHAPTER 10 Endocytic Pathways for 25-(OH) Vitamin D3
At the ultrastructural level, cubilin with megalin colocalizes in the apical part of epithelial cells, in particular in clathrin-coated pits, in dense apical tubules, and in endosomes, in agreement with the requirement of the receptor to associate with megalin [62]. Investigations into the physiological role of cubilin have been facilitated by the identification of natural mutations in the corresponding gene in patients [63]. Defects in the cubilin gene were found in individuals suffering from megaloblastic anemia 1 or ImerslundGrasbeck syndrome (IGS), a rare recessive disorder characterized by a reduced number of red blood cells displaying an increased cell volume [63,64]. Consistent with a crucial role of endocytic receptors in tubular retrieval of filtered metabolites in the kidney, several plasma proteins have been identified that are cleared from the primary urine via cubilin. These ligands include DBP [19], transferrin [65], and albumin [61], as well as Clara cell secretory protein (CCSP), a transporter for steroid hormones [60]. As a consequence of the cubilin gene defect, patients with IGS exhibit lowmolecular-weight proteinuria and excrete ligand/carrier complexes such as 25-(OH) vitamin D3/DBP [19]. c. The Two-Receptor Model Based on studies in a number of experimental systems, Nykjaer and Willnow proposed a two-receptor model for the binding of 25-(OH) vitamin D3/DBP to the PCT cell surface and for the subsequent cellular uptake of the vitamin/carrier complexes (Fig. 7) [19,66]. According to this model, complexes of DBP and 25-(OH) vitamin D3 are reabsorbed from the glomerular filtrate by association of DBP with megalin, followed by endocytic uptake. Alternatively, the complexes are bound by cubilin first and internalized via interaction of this receptor
FIGURE 7 Two-receptor model for renal uptake of 25-(OH) vitamin D3/DBP. Details of this model are given in the text.
with megalin. Because cubilin acts as a surrogate binding site for DBP on the PCT surface, cubilin deficiency partially affects renal reabsorption of 25-(OH) vitamin D3, whereas DBP or megalin gene defects completely eliminate the uptake pathway. This two-receptor model is confirmed by the moderate vitamin D deficiency observed in cubilin-deficient patients and dogs as compared to DBP or megalin knockout mice [19,25,29,31]. 2. DISABLED-2
To recycle through the endocytic compartments, receptors require the assistance of scaffold or adaptor proteins, cytoplasmic factors that bind to the tail of the receptors and regulate their trafficking. Some adaptor proteins are common to most endocytic receptors such as AP-2 that localize receptors to the clathrin coat of endocytic vesicles [67]. Other adaptors are specific for certain receptors or even cell types. Disabled (Dab)-2 is an adaptor protein required for proper function of megalin in PCT. It harbors a phosphotyrosine-binding (PTB) domain with which it attaches to an NPXY motif in the megalin tail sequences [50]. Lack of Dab-2 expression in knockout mice results in impaired tubular endocytosis and in excretion of DBP [68]. The exact mode of action of Dab-2 still remains to be elucidated. However, findings obtained in Dab-2-deficient mice suggest a direct role of the adaptor in megalin trafficking and in endocytosis of 25-(OH) vitamin D3/DBP complexes. 3. VOLTAGE-GATED CHLORIDE CHANNEL-5
Additional support for a role of endocytic pathways in the renal metabolism of 25-(OH) vitamin D3 stems from studies in patients with Dent’s disease, a rare X-linked hereditary disorder characterized by disturbances in calcium homeostasis and bone metabolism [69]. This disease is caused by mutations in the gene encoding voltage-gated chloride channel-5 (ClC-5), a Cl− transporter expressed in endosomes of PCTs [70–72]. ClC-5 is believed to be responsible for sustaining chloride conductance required for import of H+ and acidification of the endosomes. Thus, ClC-5 defects in patients result in tubular endocytic dysfunction and in the inability to reabsorb filtered plasma proteins from the lumen of the PCT [73]. One explanation for defects in calcium homeostasis in ClC-5-deficient individuals comes from observations in mice lacking this ion channel. In ClC-5 knockout animals, endocytic malfunction results in disruption of megalin-mediated uptake of 25-(OH) vitamin D3/DBP complexes, in urinary loss of these metabolites, and in a three-fold decrease in plasma levels of 25-(OH) vitamin D3 and 1,25-(OH)2 vitamin D3 [74].
160
THOMAS E. WILLNOW AND ANDERS NYKJAER
III. CONCLUSION In recent years, a number of genetic models affecting renal vitamin D metabolism in patients and laboratory animals have been elucidated. Taken together, studies in these models have identified the existence of an endocytic pathway for 25-(OH) vitamin D3 metabolites in the PCT (Fig. 8). In this pathway, circulating 25-(OH) vitamin D3 metabolites are directed to the kidney through association with DBP, internalized from the glomerular filtrate by megalin and cubilin, two DBPbinding proteins that form a two-receptor complex, and delivered to lysosomal compartments. Recycling of the receptors megalin and cubilin requires Dab-2, a cytoplasmic adaptor for megalin. Inside the cells, DBP is degraded while 25-(OH) vitamin D3 is transferred into the cytoplasm for further activation and resecretion. The intracellular mechanisms that control the release of the vitamin D metabolite from the endocytic compartments and its transport to mitochondria, the site of hydroxylation, are unknown. However, the low solubility of 25-(OH) vitamin D3 strongly argues for the presence of intracellular transport proteins that facilitate the targeting of the steroid to the correct organelle. Such intracellular transport proteins have been identified for lipophilic compounds including vitamin A (cellular retinol–binding proteins) [75] and vitamin E (α-tocopherol transfer protein) [76]. Studies by Adams and co-workers indicate that two members of the constitutively expressed heat shock protein-70 family, designated intracellular vitamin D-binding protein-1 and -2 (IDBP-1, -2) serve similar functions for vitamin D [39,77].
FIGURE 8
The existence of an endocytic pathway for 25-(OH) vitamin D3 is intriguing as it is clearly distinct from the free diffusion of the metabolite into PCT. Uptake proceeds from the luminal rather than the basolateral side of the epithelium. Cell entry occurs in the DBP-bound, but not in the free form of the hormone. There is no doubt that this endocytic pathway plays a crucial role in the retrieval of filtered vitamin D molecules from the glomerular filtrate in order to prevent uncontrolled urinary loss of essential metabolites. The contribution of this pathway to production of 1,25-(OH)2 vitamin D3 in the kidney still remains to be clarified. However, the small fraction of free 25-(OH) vitamin D3 available for passive diffusion (0.003% of total) [78] and the large amount of precursor converted by the kidney every day (0.3 to 0.5 ng) [79] strongly argue for the existence of uptake pathways whereby the kidney gains access to the pool of bound 25-(OH) vitamin D3. The expression of megalin in other epithelia such as the small intestine, breast, and prostate that also exhibit 25-hydroxyvitamin D3-1α-hydroxylase activities may even indicate a role of this endocytic pathway in extrarenal conversion of 25-(OH) vitamin D3. In conclusion, the findings discussed in this chapter do not contradict other routes (such as free diffusion) for uptake of 25-(OH) vitamin D3 into the kidney (or other tissues). In fact, residual levels of 1,25-(OH)2 vitamin D3 are found in megalin- and DBP-deficient mice [29,31]. Whether they result from renal or extrarenal conversion of the precursor is unclear at present. Whatever the source of 1,25-(OH)2 vitamin D3 may be, this pool of the hormone is not able to sustain normal calcium and bone metabolism under dietary vitamin D restriction.
Endocytic pathway for renal uptake and conversion of 25-(OH) vitamin D3/DBP complexes. Details of this pathway are discussed in the text.
161
CHAPTER 10 Endocytic Pathways for 25-(OH) Vitamin D3
References 1. Fraser DR, Kodicek E 1970 Unique biosynthesis by kidney of a biological active vitamin D metabolite. Nature 228:764–766. 2. Gray R, Boyle I, DeLuca HF 1971 Vitamin D metabolism: the role of kidney tissue. Science 172:1232–1234. 3. Lawson DE, Fraser DR, Kodicek E, Morris HR, Williams DH 1971 Identification of 1,25-dihydroxycholecalciferol, a new kidney hormone controlling calcium metabolism. Nature 230:228–230. 4. Holick MF, Schnoes HK, DeLuca HF, Suda T, Cousins RJ 1971 Isolation and identification of 1,25-dihydroxycholecalciferol. A metabolite of vitamin D active in intestine. Biochemistry 10:2799–2804. 5. Brunette MG, Chan M, Ferriere C, Roberts KD 1978 Site of 1,25(OH)2 vitamin D3 synthesis in the kidney. Nature 276:287–289. 6. Kawashima H, Torikai S, Kurokawa K 1981 Localization of 25-hydroxyvitamin D3 1α-hydroxylase and 24-hydroxylase along the rat nephron. Proc Natl Acad Sci USA 78:1199–1203. 7. Paulson SK, DeLuca HF 1985 Subcellular location and properties of rat renal 25-hydroxyvitamin D3-1α-hydroxylase. J Biol Chem 260:11488–11492. 8. Bland R, Zehnder D, Hewison M 2000 Expression of 25-hydroxyvitamin D3-1α-hydroxylase along the nephron: new insights into renal vitamin D metabolism. Curr Opin Nephrol Hypertens 9:17–22. 9. Brown AJ, Dusso A, Slatopolsky E 1999 Vitamin D. Am J Physiol 277:F157–175. 10. Daiger SP, Schanfield MS, Cavalli-Sforza LL 1975 Groupspecific component (Gc) proteins bind vitamin D and 25-hydroxyvitamin D. Proc Natl Acad Sci USA 72:2076–2080. 11. Haddad JG 1995 Plasma vitamin D–binding protein (Gc-globulin): multiple tasks. J Steroid Biochem Mol Biol 53:579–582. 12. White P, Cooke N 2000 The multifunctional properties and characteristics of vitamin D–binding protein. Trends Endocrinol Metab 11:320–327. 13. Goldstein DA, Oda Y, Kurokawa K, Massry SG 1977 Blood levels of 25-hydroxyvitamin D in nephrotic syndrome. Studies in 26 patients. Ann Intern Med 87:664–667. 14. Barragry JM, France MW, Carter ND, Auton JA, Beer M Boucher BJ, Cohen RD 1977 Vitamin-D metabolism in nephrotic syndrome. Lancet 24:629–632. 15. Colston K, Williams NJ, Cleeve HJ 1985 Studies on vitamin D binding protein in the nephrotic syndrome. Clin Chem 31:718–721. 16. Mizokuchi M, Kubota M, Tomino Y, Koide H 1992 Possible mechanism of impaired calcium and vitamin D metabolism in nephrotic rats. Kidney Int 42:335–340. 17. Harrison HE 1953 The Fanconi syndrome. J Chron Dis 7: 346–355. 18. Teranishi H, Kasuya M, Aoshima K, Kato T, Migita S 1983 Demonstration of vitamin D–binding protein (Gc-globulin) in the urine of Itai-itai disease patients. Toxicol Lett 15:7-12. 19. Nykjaer A, Fyfe JC, Kozyraki R, Leheste JR, Jacobsen C, Nielsen MS, Verroust PJ, Aminoff M, de la Chapelle A, Moestrup SK, Ray R, Gliemann J, Willnow TE, Christensen EI 2001 Cubilin dysfunction causes abnormal metabolism of the steroid hormone 25(OH) vitamin D3. Proc Natl Acad Sci USA 98:13895–13900. 20. Leheste JR, Rolinski B, Vorum H, Hilpert J, Nykjaer A, Jacobsen C, Aucouturier P, Moskaug JO, Otto A, Christensen EI,
21.
22.
23.
24. 25.
26.
27.
28.
29.
30. 31.
32. 33. 34. 35. 36.
Willnow TE 1999 Megalin knockout mice as an animal model of low molecular weight proteinuria. Am J Pathol 155: 1361–1370. St John A, Thomas MB, Davies CP, Mullan B, Dick I, Hutchison B, van der Schaff A, Prince RL 1992 Determinants of intact parathyroid hormone and free 1,25-dihydroxyvitamin D levels in mild and moderate renal failure. Nephron 61: 422–427. Orlando RA, Farquhar MG 1993 Identification of a cell line that expresses a cell surface and a soluble form of the gp330/receptor-associated protein (RAP) Heymann nephritis antigenic complex. Proc Natl Acad Sci USA 90:4082–4086. Christensen EI, Nielsen S, Moestrup SK, Borre C, Maunsbach AB, de Heer E, Ronco P, Hammond TG, Verroust P 1995 Segmental distribution of the endocytosis receptor gp330 in renal proximal tubules. EJCB 66:349–364. Christensen EI, Willnow TE 1999 Essential role of megalin in renal proximal tubule for vitamin homeostasis. J Am Soc Nephrol 10:2224–2236. Nykjaer A, Dragun D, Walther D, Vorum H, Jacobsen C, Herz J, Melsen F, Christensen EI, Willnow TE 1999 An endocytic pathway essential for renal uptake and activation of the steroid 25-(OH) vitamin D3. Cell 96:507–515. Christensen EI, Moskaug JO, Vorum H, Jacobsen C, Gundersen TE, Nykjaer A, Blomhoff R, Willnow TE, Moestrup SK 1999 Evidence for an essential role of megalin in transepithelial transport of retinol. J Am Soc Nephrol 10:685–695. Birn H, Willnow TE, Nielsen R, Norden AG, Bonsch C, Moestrup SK, Nexo E, Christensen EI 2002 Megalin is essential for renal proximal tubule reabsorption and accumulation of transcobalamin-B12. Am J Physiol Renal Physiol 282: F408–F416. Hilpert J, Wogensen L, Thykjaer T, Wellner M, Schlichting U, Orntoft TF, Bachmann S, Nykjaer A, Willnow TE 2002 Expression profiling confirms the role of endocytic receptor megalin in renal vitamin D3 metabolism. Kidney Int 62: 1672–1681. Leheste JR, Melsen F, Wellner M, Jansen P, Schlichting U, Renner-Muller I, Andreassen TT, Wolf E, Bachmann S Nykjaer A, Willnow TE 2003 Hypocalcemia and osteopathy in mice with kidney-specific megalin gene defect. FASEB J 17: 247–249. Haddad JG, Fraser DR, Lawson DE 1981 Vitamin D plasma binding protein. Turnover and fate in the rabbit. J Clin Invest 67:1550–1560. Safadi FF, Thornton P, Magiera H, Hollis BW, Gentile M, Haddad JG, Liebhaber SA, Cooke NE 1999 Osteopathy and resistance to vitamin D toxicity in mice null for vitamin D binding protein. J Clin Invest 103:239–251. Mendel CM 1989 The free hormone hypothesis: a physiologically based mathematical model. Endocr Rev 10:232–274. Christensen EI, Gliemann J, Moestrup SK 1992 Renal tubule gp330 is a calcium binding receptor for endocytic uptake of protein. J Histochem Cytochem 40:1481–1490. Bikle DD, Gee E 1989 Free, and not total, 1,25-dihydroxyvitamin D regulates 25-hydroxyvitamin D metabolism by keratinocytes. Endocrinology 124:649–654. Vanham G, Van Baelen H, Tan BK, Bouillon R 1988 The effect of vitamin D analogs and of vitamin D–binding protein on lymphocyte proliferation. J Steroid Biochem 29:381–386. Adams JS 1984 Specific internalization of 1,25-dihydroxyvitamin D3 by cultured intestinal epithelial cells. J Steroid Biochem 20:857–862.
162 37. Manolagas SC, Deftos LJ 1980 Studies of the internalization of vitamin D3 metabolites by cultured osteogenic sarcoma cells and their application to a non-chromatographic cytoreceptor assay for 1,25-dihydroxyvitamin D3. Biochem Biophys Res Commun 95:596–602. 38. Keenan MJ, Holmes RP 1991 The uptake and metabolism of 25-hydroxyvitamin D3 and vitamin D binding protein by cultured porcine kidney cells (LLC-PK1). Int J Biochem 23:1225–1230. 39. Wu S, Ren S, Chen H, Chun RF, Gacad MA, Adams JS 2000 Intracellular vitamin D binding proteins: novel facilitators of vitamin D–directed transactivation. Mol Endocrinol 14: 1387–1397. 40. Kerjaschki D, Miettinen A, Farquhar MG 1987 Initial events in the formation of immune deposits in passive Heymann nephritis. gp330-anti-gp330 immune complexes form in epithelial coated pits and rapidly become attached to the glomerular basement membrane. J Exp Med 166:109–128. 41. Yamazaki H, Ullrich R, Exner M, Saito A, Orlando RA, Kerjaschki D, Farquhar MG 1998 All four putative ligandbinding domains in megalin contain pathogenic epitopes capable of inducing passive Heymann nephritis. J Am Soc Nephrol 9:1638–1644. 42. www.ncbi.nlm.nih.gov/cgi-bin/Entrez 43. Saito A, Pietromonaco S, Loo AK, Farquhar MG 1994 Complete cloning and sequencing of rat gp330/“megalin,” a distinctive member of the low density lipoprotein receptor gene family. Proc Natl Acad Sci USA 11 9725–9729. 44. Yochem J, Greenwald I 1993 A gene for a low density lipoprotein receptor-related protein in the nematode Caenorhabditis elegans. Proc Natl Acad Sci USA 90: 4572–4576. 45. Hjalm G, Murray E, Crumley G, Harazim W, Lundgren S, Onyango I, Ek B, Larsson M, Juhlin C, Hellman P, Davis H, Akerstrom G, Rask L, Morse B 1996 Cloning and sequencing of human gp330, a Ca2+-binding receptor with potential intracellular signaling properties. Eur J Biochem 239:132–137. 46. Willnow TE, Nykjaer A, Herz J 1999 Lipoprotein receptors: new roles for ancient proteins. Nat Cell Biol 1:E157–E162. 47. Orlando RA, Exner M, Czekay R-P, Yamazaki H, Saito A, Ullrich R, Kerjaschki D, Farquhar MG 1997 Identification of the second cluster of ligand-binding repeats in megalin as a site for receptor–ligand interactions. Proc Natl Acad Sci USA 94:2368–2373. 48. Rudenko G, Henry L, Henderson K, Ichtchenko K, Brown MS, Goldstein JL, Deisenhofer J 2002 Structure of the LDL receptor extracellular domain at endosomal pH. Science 298: 2353–2358. 49. Kibbey RG, Rizo J, Gierasch LM, Anderson RG 1998 The LDL receptor clustering motif interacts with the clathrin terminal domain in a reverse turn conformation. J Cell Biol 142:59–67. 50. Oleinikov AV, Zhao J, Makker SP 2001 Cytosolic adaptor protein Dab2 is an intracellular ligand of endocytic receptor gp600/ megalin. Biochem J 347:613–621. 51. Nykjaer A, Willnow TE 2002 The low-density lipoprotein receptor gene family: a cellular Swiss army knife? Trends Cell Biol 12:273–280. 52. Willnow TE, Hilpert J, Armstrong SA, Rohlmann A, Hammer RE, Burns DK, Herz J 1996 Defective forebrain development in mice lacking gp330/megalin. Proc Natl Acad Sci USA 93:8460–8464. 53. Fass D, Blacklow S, Kim PS, Berger JM 1997 Molecular basis of familial hypercholesterolaemia from structure of LDL receptor module. Nature 388:691–693.
THOMAS E. WILLNOW AND ANDERS NYKJAER
54. Christensen EI, Birn H, Verroust P, Moestrup SK 1998 Membrane receptors for endocytosis in the renal proximal tubules. Int Rev Cytol 180:237–284. 55. Seetharam B, Levine JS, Ramasamy M, Alpers DH 1988 Purification, properties, and immunochemical localisation of a receptor for intrinsic factor–cobalamin receptor in the rat kidney. J Biol Chem 263:4443–4449. 56. Kozyraki R, Kristiansen M, Silahtaroglu A, Hansen C, Jacobsen C, Tommerup N, Verroust PJ, Moestrup SK 1998 The human intrinsic factor vitamin B12 receptor, cubilin: molecular characterization and chromosomal mapping of the gene to 10p within the autosomal recessive megaloblastic anemia (MGA1) region. Blood 91:3593–3600. 57. Moestrup SK, Kozyraki R, Kristiansen M, Kaysen JH, Rasmussen HH, Brault D, Pontillon F, Goda FO, Christensen EI, Hammond TG, Verroust PJ 1998 The intrinsic factor– vitamin B12 receptor and target of teratogenic antibodies is a megalin-binding peripheral membrane protein with homology to developmental proteins. J Biol Chem 273:5235–5242. 58. Xu D, Kozyraki R, Newman TC, Fyfe JC 1999 Genetic evidence of an accessory activity required specifically for cubilin brush-border expression and intrinsic factor–cobalamin absorption. Blood 94:3604–3606. 59. Kristiansen M, Kozyraki R, Jacobsen C, Nexo E, Verroust PJ, Moestrup SK 1999 Molecular dissection of the intrinsic factorvitamin B12 receptor, cubilin, discloses regions important for membrane association and ligand binding. J Biol Chem 274:20540–20544. 60. Burmeister R, Boe IM, Nykjaer A, Jacobsen C, Moestrup SK, Verroust P, Christensen EI, Lund J, Willnow TE 2001 A tworeceptor pathway for catabolism of Clara cell secretory protein in the kidney. J Biol Chem 276:13295–13301. 61. Birn H, Fyfe JC, Jacobsen C, Mounier F, Verroust PJ, Orskov H, Willnow TE, Moestrup SK, Christensen EI 2000 Cubilin is an albumin binding protein important for renal tubular albumin reabsorption. J Clin Invest 105:1353–1361. 62. Sahali D, Mulliez N, Chatelet F, Dupuis R, Ronco P, Verroust P 1988 Characterization of a 280-kD protein restricted to the coated pits of the renal brush border and the epithelial cells of the yolk sac. Teratogenic effect of specific monoclonal antibodies. J Exp Med 167:213–218. 63. Aminoff M, Carter JE, Chadwick RB, Johnson C, Grasbeck R, Abdelaal MA, Broch H, Jenner LB, Verroust PJ, Moestrup SK, de la Chapelle A, Krahe R 1999 Mutations in CUBN, encoding the intrinsic factor–vitamin B12 receptor, cubilin, cause hereditary megaloblastic anaemia 1. Nat Genet 21: 309–313. 64. Imerslund O 1960 Idiopathic chronic megaloblastemic anemia in children. Acta Paediatr 49:208–209. 65. Kozyraki R, Fyfe J, Verroust PJ, Jacobsen C, Dautry-Varsat A, Gburek J, Willnow TE, Christensen EI, Moestrup SK 2001 Megalin-dependent cubilin-mediated endocytosis is a major pathway for the apical uptake of transferrin in polarized epithelia. Proc Natl Acad Sci USA 98:12491–12496. 66. Willnow TE, Nykjaer A 2002 Pathways for kidney-specific uptake of the steroid hormone 25-hydroxyvitamin D3. Curr Opin Lipidol 13:255–260. 67. Bansal A, Gierasch LM 1991 The NPXY internalization signal of the LDL receptor adopts a reverse-turn conformation. Cell 67:1195–1201. 68. Morris SM, Tallquist MD, Rock CO, Cooper JA 2002 Dual roles for the Dab2 adaptor protein in embryonic development and kidney transport. EMBO J 21:1555–1564. 69. Pook MA, Wrong O, Wooding C, Norden AG, Feest TG, Thakker RV 1993 Dent’s disease, a renal Fanconi syndrome
CHAPTER 10 Endocytic Pathways for 25-(OH) Vitamin D3
70.
71.
72.
73.
with nephrocalcinosis and kidney stones, is associated with a microdeletion involving DXS255 and maps to Xp11.22. Hum Mol Genet 2:2129–2134. Lloyd SE, Pearce SH, Fisher SE, Steinmeyer K, Schwappach B, Scheinman SJ, Harding B, Bolino A, Devoto M, Goodyer P, Rigden SP, Wrong O, Jentsch TJ, Craig IW, Thakker RV 1996 A common molecular basis for three inherited kidney stone diseases. Nature 379:445–449. Igarashi T, Gunther W, Sekine T, Inatomi J, Shiraga H, Takahashi S, Suzuki J, Tsuru N, Yanagihara T, Shimazu M, Jentsch TJ, Thakker RV 1998 Functional characterization of renal chloride channel, CLCN5, mutations associated with Dent’s Japan disease. Kidney Int 54:1850–1860. Yamamoto K, Cox JP, Friedrich T, Christie PT, Bald M, Houtman PN, Lapsley MJ, Patzer L, Tsimaratos M, van’t Hoff WG, Yamaoka K, Jentsch TJ, Thakker RV 2000 Characterization of renal chloride channel (CLCN5) mutations in Dent’s disease. J Am Soc Nephrol 11:1460–1468. Gunther W, Luchow A, Cluzeaud F, Vandewalle A, Jentsch TJ 1998 ClC-5, the chloride channel mutated in Dent’s disease, colocalizes with the proton pump in endocytotically active kidney cells. Proc Natl Acad Sci USA 95:8075–8080.
163 74. Piwon N, Gunther W, Schwake M, Bosl MR, Jentsch TJ 2000 ClC-5 Cl−-channel disruption impairs endocytosis in a mouse model for Dent’s disease. Nature 408:369–373. 75. Napoli JL 1996 Biochemical pathways of retinoid transport, metabolism, signal transduction. Clin Immunol Immunopathol 80:S52–62. 76. Ouahchi K, Arita M, Kayden H, Hentati F, Ben Hamida M, Sokol R, Arai H, Inoue K, Mandel JL, Koenig M 1995 Ataxia with isolated vitamin E deficiency is caused by mutations in the alpha-tocopherol transfer protein. Nat Genet 9:141–145. 77. Adams JS, Chen H, Chun RF, Nguyen L, Wu S, Ren SY, Barsony J, Gacad MA 2003 Novel regulators of vitamin D action and metabolism: lessons learned at the Los Angeles zoo. J Cell Biochem 88:308–314. 78. Bikle DD, Gee E, Halloran B, Kowalski MA, Ryzen E, Haddad JG 1986 Assessment of the free fraction of 25-hydroxyvitamin D in serum and its regulation by albumin and the vitamin D–binding protein. J Clin Endocrinol Metab 63:954–959. 79. Bikle DD, Siiteri PK, Ryzen E, Haddad JG 1985 Serum protein binding of 1,25-dihydroxyvitamin D: a reevaluation by direct measurement of free metabolite levels. J Clin Endocrinol Metab 61:969–975.
CHAPTER 11
The Vitamin D Receptor J. WESLEY PIKE AND NIRUPAMA K. SHEVDE Department of Biochemistry, University of Wisconsin–Madison, Madison, Wisconsin
I. II. III. IV.
Introduction Discovery of the Vitamin D Receptor Characterization of the Vitamin D Receptor Structural Gene for the Vitamin D Receptor
I. INTRODUCTION Many significant advances have been made during the past several decades in our understanding of the mechanisms through which steroid hormones function within target cells [1–6]. It is now well known that both the sex and adrenal steroids, thyroid hormone, retinoic acid (RA), and 1,25-dihydroxyvitamin D3 (1,25(OH)2D3) all bind to specific intracellular receptors. These proteins in turn transduce the environmental information carried by the small hormonal signals to the cell nucleus resulting in modifications in gene expression. Virtually all of the genes that encode nuclear receptors have now been cloned, revealing them to be part of a large gene family of proteins with highly related structural similarity. Interestingly, members of this family of nuclear receptor genes also transduce vertebrate signals generated by other forms of retinoic acid such as 9-cis-RA, fatty acids, nutritional metabolites, and invertebrate signaling molecules such as the insect molting hormone ecdysone [6] as well as the classical hormones. Perhaps most importantly, the cloning of the nuclear receptor family of genes has made available critical reagents that have led to important new insights into the structures of the proteins, an understanding of how hormonal signals activate these factors, delineation of the mechanisms by which they interact with the regulatory regions of target genes, and the discovery that numerous additional transcription factor complexes are essential for downstream gene modulation. The idea that vitamin D might function as a steroidlike hormone emerged in 1968 [7] and actually predated the documented discovery of 1,25(OH)2D3, the active hormonal form of vitamin D3 [8–10]. The lipophilic nature of the small vitamin D molecule and its capacity to localize in target tissues, as well as the fact that vitamin D responses were sensitive to transcriptional inhibitors, provided an initial basis for this hypothesis. Additional studies also defined the exquisite and highly complex nature of the regulation VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
V. Functional Analysis of the Vitamin D Receptor VI. The Human Vitamin D–Receptor Chromosomal Gene VII. Concluding Comments References
of 1,25(OH)2D3 synthesis and production by the renal enzyme 25-hydroxyvitamin D3 1α-hydroxylase (see Chapter 5) [11]. Nevertheless, the discovery of a “binding protein” in specific target tissues that appeared to mediate nuclear localization of what eventually was determined to be the active vitamin D hormone, 1,25(OH)2D3 was perhaps the most significant [12]. Indeed, further characterization of this macromolecule over the years, and its eventual cloning in 1987 [13], confirmed the original hypothesis that 1,25(OH)2D3 was a steroid-like hormone. Its proposed mechanism in general terms is simply depicted in Fig. 1. These studies also showed that the actions of 1,25(OH)2D3 were directly linked to the gene regulatory properties of this “binding protein” or vitamin D receptor (VDR) [13–15], as alluded to in the figure. The structure and function of the VDR confirmed 1,25(OH)2D3 as an authentic steroid hormone and ushered in a new era of research on the molecular mechanisms underlying its regulation of gene expression. In this chapter, we provide an historical overview as well as contemporary summary of the VDR’s central role in mediating the actions of the vitamin D hormone. We describe the discovery of the receptor, a summary of its tissue and cellular distribution, and its biochemical characteristics. We also provide an overview of the VDR’s structural organization as well as how it functions to regulate gene expression. Finally, we describe the structural organization for the human VDR chromosomal gene. The reader is referred to the many additional chapters in the Mechanism of Action section of this book for more depth into each of the many aspects related to VDR structure, biochemistry, and biological function.
II. DISCOVERY OF THE VITAMIN D RECEPTOR Evidence for the existence of a binding protein or receptor (i.e., the VDR) was first provided by Haussler and Norman in 1969 [12]. These studies demonstrated Copyright © 2005, Elsevier, Inc. All rights reserved.
168
J. WESLEY PIKE AND NIRUPAMA K. SHEVDE
DB P-1 ,25 (O H)
2D 3
1,
25
(O H D3 )2
Target cell Nucleus R
R
mRNA
Tissue Specific Cellular Response
Mineral Homeostasis Bone Metabolism Vitamin D3 Catabolism Immunomodulation Differentiation Receptor Autoregulation
FIGURE 1
Model for the molecular mechanism of action of the vitamin D hormone. 1,25(OH)2D3 dissociates from serum vitamin D binding protein (DBP) or other carrier proteins, enters the cell, and interacts with the vitamin D receptor (VDR). Activation by the ligand leads to binding of the VDR to promoters within responsive genes and subsequent modulation of gene expression. Biological responses are indicated.
a preferential uptake of active vitamin into the intestinal chromatin fraction of chickens. They also provided further evidence that the process was saturable at low concentrations of administered active vitamin and was likely mediated by a specific protein component. Following these pioneering studies, more definitive evidence for the VDR began to emerge from the studies of Brumbaugh and Haussler [16] and Lawson and Wilson [17], who established the protein nature of the receptor via proteolytic digestion studies and equilibrium sedimentation analysis. Studies by Brumbaugh and Haussler [18] also established that the protein bound 1,25(OH)2D3 with high nanomolar affinity, an important prerequisite for mediating vitamin D action. Nevertheless, it was experiments carried out in vitro in 1975 that provided the most compelling evidence that
the protein might function as a nuclear receptor by demonstrating that cytosol-derived VDR could bind to chromatin fractions in a hormone-sensitive fashion [18]. Collectively, these findings provided the impetus for further exploration into the properties, structure, and function of the VDR.
III. CHARACTERIZATION OF THE VITAMIN D RECEPTOR A. Tissue Distribution of the VDR The VDR was first discovered in extracts prepared from chicken intestine, perhaps as a result of the considerable role the chicken as a model system had played in delineating the function of vitamin D in
169
CHAPTER 11 The Vitamin D Receptor
calcium homeostasis. Despite this, the known actions of vitamin D extended well beyond those in the intestine and were certainly not limited to the avian species. Target tissues of vitamin D included the kidney, bone, and the parathyroid glands and accordingly, the VDR was eventually identified in those tissues in avian as well as mammalian species [19–23]. The discovery of the VDR in mammalian intestine and in most of the listed tissues, nevertheless, required efforts over several years. Kream et al. [24] provided evidence for the existence of a VDR in rat intestinal tissue. Its properties were similar to those of the chicken receptor, although equilibrium sedimentation experiments suggested that rat receptor might exhibit a smaller molecular weight. Importantly, the discovery of receptors in these as well as other tissues was facilitated by tritiumlabeled 1,25(OH)2D3, whose availability with increasing radioisotopic specific activity provided the enhanced sensitivity necessary to detect VDR levels in tissues other than the intestine where it was most abundant. Thus, high specific radioactivity 1,25(OH)2D3 was a major determinant in the discovery and subsequent characterization of the VDR. 1,25(OH)2D3 with very high specific radioactivity also enabled detection of VDR in many unexpected target tissues as well as in cultured cell lines. The search for receptors in these tissue and cell sources was prompted by emerging evidence that the biological effects of 1,25(OH)2D3 extended beyond that of calcium and phosphorus homeostasis to include a role in cellular proliferation and differentiation [25,26]. Accordingly, VDRs were discovered in tissues such as pancreas [20,27], placenta [20], pituitary [28], ovary [29], testis [30], mammary gland [31], and heart [32]. Cultured cell lines that expressed the VDR were similarly extensive and now include cells of fibroblastic, chondrocytic, osteoblastic, myoblastic, hematopoietic, and lymphopoietic origin as well as cells derived from normal kidney, intestine, and skin. Of more fundamental nature, they also include cells of both tumorigenic and nontumorigenic origin [33–36]. A partial listing of vitamin D target tissues and cells is documented in Table I. The biological role of 1,25(OH)2D3 in many of these tissues and cells is considered more extensively in many additional chapters in this book. Perhaps the most interesting biological effect of vitamin D metabolites, however, is their ability to regulate proliferation and differentiation [37]. These effects of 1,25(OH)2D3 are not unlike those of most steroid hormones. Most importantly, however, the actions highlight a potential therapeutic role for 1,25(OH)2D3 and its analogs in cancer treatment, an area considered in more depth in Chapters 84–88 and 93–97, and in the regulation of the immune system [38,39], considered in Chapters 36, 98, and 99.
TABLE I Cellular and Tissue Distribution of the VDR System Gastrointestinal Hepatic Renal Cardiovascular Endocrine Exocrine Reproductive
Immune Respiratory Musculoskeletal
Epidermis/appendage Central nervous system Connective tissue
Tissue Esophagus, stomach, small intestine, large intestine, colon Liver parenchyma cells Kidney, urethra Cardiac muscle Parathyroid gland, thyroid, adrenal, pituitary Parotid gland, sebaceous gland Testis, ovary, placenta, uterus, endometrium, yolk sac, avian chorioallantoic membrane, avian shell gland Thymus, bone marrow, B cells, T cells Lung alveolar cells Bone osteoblasts and osteocytes, cartilage chondrocytes, striated muscle Skin, breast, hair follicles Brain neurons Fibroblasts, stroma
Antiproliferative effects of vitamin D are currently being exploited in the treatment of human psoriasis, a hyperproliferative disorder of the skin [40] that is another important target of vitamin D action [37,41] outlined in more detail in Chapter 101. Indeed, the clear presence of the VDR in most tissues in avian and mammalian species indicates that the biological effects of vitamin D are likely to be very widespread. While the analysis of tissue extracts led to the discovery of many VDR-containing sites, it was the use of scintillant-enhanced autoradiography employed early on by Zile et al. [42] and Jones and Haussler [43] and later by Stumpf et al. [44,45] that led to a thorough appreciation of the wide cellular distribution of vitamin D target tissues. These studies revealed such targets as the duodenum, ileum, and jejunum of the rat intestine, stomach, kidney, skin, and pituitary as well as several neurons in the brain. While the presence of 1,25(OH)2D3 (or a metabolite of the administered compound) in these tissues through autoradiographic means does not provide direct evidence of the presence of VDR, the bulk of these targets were confirmed when an antibody to the VDR was developed some years later enabling the use of immunocytochemistry [46–49]. Most recently, studies using Northern blot [13], in situ hybridization [50], or RT-PCR [51] analyses have been employed to detect the VDR mRNA in cells and tissues.
170 The availability of probes that enabled these analyses closely followed the cloning of the VDR in 1987 and 1988 (to be discussed in Section IV). Although this RNA detection approach has been useful, it often ignores the important underlying assumption that VDR mRNA transcripts correlate directly with VDR protein. The use of these nucleic techniques has been helpful, however, in the study of the transcriptional regulation of VDR production.
B. Subcellular Distribution of the VDR Although the site of action of ligand-bound VDR in the nucleus is clear, the subcellular location of unliganded VDR has remained controversial over several decades. Early studies suggested that the unliganded VDR was cytosolic (see Fig. l and Brumbaugh and Haussler [18]). Studies in the early 1980s, however, suggested that the VDR might be loosely associated with the nuclear fraction [52,53]. Indeed, the application of immunocytochemistry supports this largely nuclear localization of the apo form of the VDR, an intracellular residence site consistent with that of most other nuclear receptors. It is important to note, however, that the presence of the unliganded VDR within the nucleus does not imply that the VDR must be bound to regulatory sites of action on 1,25(OH)2D3 responsive genes. To the contrary, recent chromatin immunoprecipitation assays to be discussed later suggest that the VDR is not bound to these regulatory sites until activated by 1,25(OH)2D3 [54,55], supporting the need to identify these alternative localization sites for VDR in the nucleus in the absence of hormone. Perhaps not surprisingly, the controversy regarding the intracellular site of residency of unliganded VDR has reemerged recently. In studies using fluorescent probes and discussed in greater detail in Chapter 22, Barsony and colleagues [57,58] provide evidence that recombinantly derived VDR distributes into both the cytoplasmic and nuclear compartments in the absence of ligand and is imported into the nucleus via several essential components, including specific nuclear import signals within the VDR molecule, interaction with RXR, and interaction with nuclear import proteins lining the nuclear pores. Moreover, the VDR actively cycles between the two compartments. With the qualification that the VDR is expressed in relatively high concentrations in the cell as a chimeric protein fused to green fluorescent protein (GFP), the data seem to support strongly active receptor movement between both cellular compartments. Given these caveats, however, it is likely that the controversy will continue.
J. WESLEY PIKE AND NIRUPAMA K. SHEVDE
C. Biochemical Properties and Organization of the VDR The VDR is expressed in low concentrations in target tissues and cultured cells, a property consistent with its role as a potent transcription regulatory molecule [58,59]. Estimates of receptor abundance range from under 200 to over 25,000 copies of VDR/cell (10 to 100 fmol/mg protein) depending on the cell type or cell line examined, and up to 1 pmol/mg protein in certain tissue extracts. Differences in VDR abundance suggest that cells with higher VDR content may be more highly responsive to 1,25(OH)2D3 than those with lower levels. Little evidence supports this hypothesis, however, and in fact recent data suggest that the expression levels of many of the comodulatory complexes discussed in Chapters 14, 16, and 17 may be more significant determinants of cellular response than the receptor concentration itself [60]. Other factors may also contribute to the level of response to 1,25(OH)2D3, including cellular capacity to internalize and metabolize 1,25(OH)2D3, differential activation events that involve the VDR directly (phosphorylation or other posttranslational modifications), and finally the accessibility and individual activation properties of specific gene targets. These, as well as additional events, are likely to contribute significantly to the biological responsiveness and cellular sensitivity to 1,25(OH)2D3. Numerous physical and functional properties of the VDR were defined following the protein’s discovery in 1973, several of which are documented in Table II. Sedimentation analysis revealed a protein of 3 to 3.7 S that exhibited an elongated shape. Gel filtration estimates of protein size ranged from 50,000 to 70,000 Da
TABLE II
Biochemical Properties of the VDR and Its RNA Transcript
VDR protein Molecular weight: 50 kDa (human) to 60 kDa (chicken) Sedimentation coefficient (S): 3.2 (human, rat) to 3.7 (chicken) Equilibrium dissociation constant: 1–2 × 10–10 M Isoelectric point: 6.2 Phosphoprotein: Serine phosphorylated at residue 208 (human) in response to 1,25(OH)2D3 Lability: Proteolytically sensitive Tissue abundance: 10 to 1000 fmol of VDR/mg protein hVDR mRNA 4.8 kb (human), 3.0 (chicken)
171
CHAPTER 11 The Vitamin D Receptor
depending on species. The most important characteristic of the VDR, however, was its capacity to bind 1,25(OH)2D3 with both high affinity and selectivity [61–65]. Accordingly, the VDR displayed an equilibrium dissociation constant (KD) of approximately 10−10 M for the natural ligand 1,25(OH)2D3 and bound this ligand’s precursors as well as less active metabolites with substantially lower affinities [63,64]. The contribution of both the 25-hydroxyl and the 1α-hydroxyl groups on the 1,25(OH)2D3 molecule to specific, high-affinity binding to VDR has been studied extensively [64] and will be discussed later. An important discovery in the late 1970s was the observation that the VDR could bind to DNA, an anticipated property of a protein that activates transcription [58]. While a more thorough understanding of VDR DNA binding emerged following the identification of specific DNA-binding sites (vitamin D response elements or VDREs) within gene promoters, as discussed later in this chapter and other chapters, the finding that the VDR bound to nonspecific DNA set the stage for the ensuing studies aimed at understanding the structural organization and function of the VDR. Perhaps the most important observation made of VDR DNA binding properties was the finding that while VDR exhibited an obvious “affinity” for DNA in the absence of 1,25(OH)2D3, a feature unlike that of most steroid receptors, its “affinity” for DNA quantitatively increased following complex formation with 1,25(OH)2D3 [65,66]. The latter property implied that the structure of the VDR or perhaps the composition of the active receptor was transformed in the presence of the hormonal ligand. Recent X-ray diffraction studies to be described later in this chapter as well as in Chapter 15 define the nature of the effects of ligand on the VDR and its DNA binding capabilities at the atomic level [67–69]. Purification of the VDR initially from chicken intestine [14,58] and later from porcine intestine [15] led to the development of monoclonal antibodies to the VDR. These reagents proved essential in further immunological characterization of the VDR, provided alternative receptor isolation methods, and ultimately offered the means whereby the VDR gene was cloned. Western blotting analysis revealed for the first time that the precise molecular mass of VDR varied from species to species, ranging from 60,000 Da in the chicken [70] to approximately 50,000 Da in humans [71]. These relative sizes were confirmed through molecular cloning. Immunological analyses of this type also revealed that the VDR comprised A and B isoforms in the chicken but a single form in mammalian-derived tissues [70]. The two forms of the chicken VDR are
now known to arise from the existence of alternative start sites in the mRNA, although the functional consequence of two such proteins with differing aminoterminal extensions remains unknown [72,73]. Finally, immunoprecipitation of the VDR from in vitro translated tissue mRNA revealed that whereas the chicken receptor protein was larger than that from mammalian cells, the mRNA transcripts were considerably smaller [71]. The differential size of these transcripts has been confirmed through the use of hybridization techniques incorporating the cDNA for the VDR as a probe. Initial insights into the structural organization of the VDR emerged using techniques such as limited proteolytic digestion together with radiolabeled 1,25(OH)2D3 and immunologic probes. Initial studies by Pike [74] suggested that the VDR epitope for a specific monoclonal antibody to the VDR lay immediately adjacent to the domain necessary for DNA binding. Additional studies by Allegretto and co-workers [75,76] employing limited digestion with trypsin led to the realization that to the l,25(OH)2D3-bound VDR could be cleaved into two domains, one a large fragment of 30 to 40 kDa that retained prebound 1,25(OH)2D3 and a second of 16 to 20 kDa that could bind DNA independently. Most importantly, the findings suggested that the two functional domains (ligand-binding and DNA-binding domains) were separated by a proteolytically sensitive “hinge.” The inability following trypsin digestion to identify the domain responsible for ligand binding of hormone-free receptor argued strongly for a ligandinduced change in receptor conformation. This finding substantiated earlier observations of increased VDR stability in tissue extracts following complex formation with ligand. Additional preliminary evidence indicated that the DNA binding domain was located amino terminal to the hormone binding domain [77]. As described in the next section, these rudimentary insights into the overall structural organization of the VDR were proven correct via the subsequent cloning of the receptor’s structural gene.
IV. STRUCTURAL GENE FOR THE VITAMIN D RECEPTOR A. Cloning of the VDR The biochemical purification of the steroid receptors in the early 1980s led to the development of immunological probes that facilitated not only further characterization of the proteins, but the cloning of their structural genes as well. As a result, the receptors for virtually all the known steroid, thyroid, and
172 vitamin hormones were cloned during the latter half of the 1980s [2,3]. Interestingly, sequence analysis of the first cloned receptor cDNAs revealed regions that exhibited significant sequence similarity with the DNA binding domain of an unrelated transcription factor TFIIIA. In addition, sequence similarities within this specific region among the several nuclear receptors suggested the possibility that related transcripts might be recovered through low-stringency hybridization screening [3]. Accordingly, this led to the recovery of numerous additional cDNAs for unknown nuclear receptors or orphan receptors. For many of these, cognate ligands, if they exist, remain currently unidentified. However, investigation of several of these unknown receptors led to the eventual discovery of novel ligands and the identification of new hormonal systems. Examples include the discovery of retinoid X receptor (RXR) and its ligand 9-cis RA [78], peroxisome proliferator-activated receptor gamma (PPARγ) and one of its ligands prostaglandin J2 [79], farnesol–X receptor (FXR) and its ligand farnesol [80], the liver X receptor (LXR) and its oxysterol ligands [81,82], and several others. Interestingly, the sequence of the DNA binding domain of the glucocorticoid receptor (GR) was first noted to exhibit a high degree of similarity with the viral oncogene v-erbA [83]. The cloning of c-erbA led to the discovery that this gene encoded the first of two thyroid hormone receptor (TR) genes [84]. The development of the anti-VDR monoclonal antibody 9A7 enabled McDonnell et al. [13] to screen randomly primed chicken intestinal cDNA expression libraries and eventually to recover a single cDNA clone that encoded a zinc finger containing DNA binding domain. This domain exhibited high homology to the DNA binding domain of GR. The presence of this domain and its reactivity to the 9A7 antibody (known to interact in or near the DNA binding domain of the VDR [65]) led to initial confidence that this and several other isolated cDNAs represented a portion of the transcript encoding the chicken VDR. Subsequent hybridization-selected in vitro translation techniques using these clones substantiated the authenticity of the cDNA clones [13]. The recovery of this cDNA allowed the subsequent rapid recovery of full-length VDR cDNA transcripts from human [85] tissue sources. The rat intestinal VDR was cloned shortly thereafter by Burmester et al. [86]. Recovery of a cDNA transcript from the human leukemia HL-60 cell line [87] revealed virtual identity to that of the original human VDR cDNA cloned by Baker et al. [85]. Sequences of the VDR have been reported from mouse [88], Japanese quail [89], and Xenopus [90] and have appeared in both the human and mouse genomic data bases [91,92]. The cloning of the VDRs from various
J. WESLEY PIKE AND NIRUPAMA K. SHEVDE
sources and species was a milestone in the vitamin D field. Perhaps as important, it provided the first direct evidence of a structural relationship between the VDR and other bona fide members of the steroid receptor family of genes [93]. The VDR is highly conserved across tissue sources and species. The overall sequence similarity of rat, mouse, and avian receptors to that of human VDR is 79, 86, and 66%, respectively. This homology rises to above 95% within specific domains such as the DNA binding domain (see Chapter 13). There is, nevertheless, considerable variability in the number of residues amino terminal to the DNA binding domain. This region varies from 21 amino acids in the human VDR (the smallest of the VDRs) to approximately 57 amino acids in the chicken protein [85,89]. In addition, the chicken gene exhibits one alternative start site (the first at nucleotide 47 and the second at nucleotide 67, which corresponds to the most 5′ ATG in the human VDR transcript identified by Baker et al. [85]) that accounts for the two forms of the VDR found in the avian species [72,73]. The human VDR cDNA sequence also presents as a transcript with two potential start sites, one lying only three codons downstream of the first [85]. Whether both initiation codons are used in this transcript is unclear. However, an important polymorphism in the VDR gene has been demonstrated in the human population [94,95]. This polymorphism is a single-base-pair transition that occurs in the first initiation site reported by Baker et al. [85] (ATG to ACG). The result of this polymorphism in humans is the production of a smaller VDR of 424 amino acids that occurs in homozygous form in approximately 37% of the human population and in heterozygous form (one allele containing both initiation sites and the second containing only the downstream site) in approximately 48% of the population [96] (Chapter 72). Whether the two proteins exhibit different biochemical properties and/or different functions remains an open question, although it does appear that the smaller form may be more active [96–98]. This has been suggested to occur via a tighter contact with the transcription factor TFIIB [96]. Association studies indicate that this polymorphism may be linked to bone mineral density [97], although these data are not particularly robust [99–101] (Chapter 68).
B. VDR Is a Member of the Intracellular Receptor Gene Family The cloning of glucocorticoid [83] and estrogen [102] receptors in 1985 and 1986 were the first of a series of successful efforts to clone the intracellular receptor genes. More than 150 cloned members of this
173
CHAPTER 11 The Vitamin D Receptor
intracellular receptor gene family [5] have been identified, making it the largest transcription factor family currently known. It is in fact noteworthy that the cloning of this family of genes was so successful that publication of the sequence of both the mouse and human genomes in 2002 did not reveal a single uncloned and unidentified functional nuclear receptor gene [91,92]. The evolutionary success of this family of transcription factors is highlighted both by its abundance and by its involvement in an incredibly wide range of biological processes that include growth, development, differentiation, and adult homeostasis. The reader is referred to other reviews on the steroid receptor family for details [1–6] as well as summaries in Chapters 13–16.
A
C. Structural Domains of the VDR 1. OVERVIEW
The cloning of the estrogen receptor (ER) in 1986 and the GR shortly thereafter revealed that these proteins were made up of distinct regions or domains, leading to their designation as domains A, B, C, D, E, and F [102]. As illustrated in Fig. 2A, the highly conserved DNA binding domain is designated the C domain and provides a central focus for the remainder of the molecules. Accordingly, the nonconserved A/B domain comprises all residues amino terminal to the DNA binding domain and the D domain or hinge region comprises residues lying downstream of C. The carboxyterminal region that contains the ligand binding
NR A/B
B
C 1
24
D 89
E/F 232
427
41 6 42 3
1
9
34
31
27
23
2
C
2
hVDR
427
44 8 45 5
7
5
37
35
0 31
27
0
E/F Domain VDR
456
40 8 41 5
8
6
33
31
1 27
23
4
TRβ
462
90 7 91 4
0
8
83
80
1 76
72
1
RARα
934
ER
53 8 54 5
45 7
43 5
38 9
34 9
PR
595
FIGURE 2 Functional domain structure of the nuclear receptor superfamily. (A) The nuclear receptors (NR) are separated into four regions designated A/B, C, D, and E/F. (B) Residue boundaries of corresponding regions within the human VDR. (C) Three regions of sequence similarity and residue boundaries within the E/F domain of several nuclear receptors. Shown in addition to VDR are thyroid receptor β, (TRβ), retinoid receptor α (RARα), progesterone receptor (PR), and estrogen receptor (ER). Functions associated with the domains include transactivation (A/B), DNA binding (C), flexible hinge (D), and dimerization, ligand binding, transactivation, and repression (E/F) (see text for details).
174 domain is termed the E or E/F domain. Three stretches of sequence similarity exist in the E domain for virtually all of the nuclear receptor gene family. The F domain exhibits extensive variability and is not conserved, much like that of the A/B domain. The domain structure of mammalian VDRs is depicted in Fig. 2B. As can be seen, the A/B domain is very short relative to other members of the nuclear receptor family, representing one of the most abbreviated of the entire gene family. The C region comprises the DNA binding domain of the VDR [103]. The D domain is a highly flexible region within the VDR and links both its DNA-binding and its hormone-binding components [103]. It is not conserved among VDRs from other species or among other members of the steroid receptor family [85,86]. This region is, however, essential for intact hormone binding, a feature unlike that of most other steroid receptors. The hinge region may be more complex than that found in other receptors in that it contains an additional segment that corresponds to a specific exon in the chromosomal gene [104]. Interestingly, however, this region can be removed without compromising 1,25(OH)2D3 binding or transcriptional function [67,105]. Finally, the E/F domain contains the 1,25(OH)2D3 binding region of the VDR [103] and, equally important, serves as a highly complex protein–protein interface for a series of additional proteins that are integral to receptor activity [106–108]. This region has been termed activation function 2 or AF-2. The ability of the VDR to bind both DNA and additional transcription factors highlights its central role in nucleating the formation of transcriptionally active complexes at specific sites of gene regulation. Since the general role of all members of the receptor gene family is similar, it is not surprising that the E region is moderately conserved across the entire gene family and contains several subdomains that are even more highly conserved, as depicted in Fig. 2C. The F domain is absent in the VDR. 2. DNA BINDING DOMAIN
The DNA binding domain of the VDR is encoded in domain C. This domain consists of two similar modules, each consisting of a zinc-coordinated finger structure. The zinc atoms are individually coordinated in a tetrahedral fashion through four highly conserved cysteine residues that serve to stabilize the finger structure itself. Importantly, the finger modules are structurally unrelated to the zinc fingers found in TFIIIA wherein the zinc atom is coordinated through two cysteines and two histidines [109,110]. Surprisingly, although the two zinc modules of the VDR appear highly related to each other structurally, they are not equivalent topologically because the chirality of the residues that
J. WESLEY PIKE AND NIRUPAMA K. SHEVDE
coordinate the zinc atoms in each module is different [110]. The function of each of these modules in DNA binding is also different. Whereas the amino-terminal module directs specific DNA binding in the major grove of the DNA binding site, the carboxy-terminal module serves as a dimerization interface for interaction with partner proteins [111,112]. It seems likely that the two exons that encode these modules (see Section VI) evolved separately, although it is possible that each derived from a common ancestral gene through duplication and then diverged later. The threedimensional structure of the DNA-binding domain of several of the receptors has been determined through both nuclear magnetic resonance (NMR) spectroscopy and X-ray crystallography. The salient features of the finger structures are the presence of carboxy-terminal α-helices essential to interaction with DNA. Thus, our understanding of the structural organization of these modules as well as the mechanisms by which they function to interact with DNA and other proteins is well advanced [113–118]. 3. LIGAND BINDING DOMAIN
The E region of the VDR represents a multifunctional domain that exerts absolute regulatory control over the DNA-binding as well as transcription-modifying properties of the VDR. Historically, this region and comparable regions found in other receptors were explored for common structural features as well as for individual functional properties. As indicated in the overview in this section, several regions of homology were identified. Since cognate ligands differ substantially, it appeared likely that these regions of similarity represented motifs integral to the receptors’ role in transactivation. Closer scrutiny revealed the presence of multiple heptad repeats that were located in the ligand binding domain (LBD) of most members of the gene family including that of the VDR [119]. Despite this and many other important biochemical and molecular biologic discoveries, it is clear that the greatest insight into LBD structure emerged as a result of X-ray crystallographic studies initiated by Moras and colleagues with the retinoid receptors RARγ and RXRα [120,121]. As of this date, the three-dimensional structures of the LBDs of most of the known ligand-binding members of the nuclear receptor family including VDR have been determined [67–69,122–126]. With the exception of RXRα, the structures of the LBDs from most of the receptors have been solved in the presence of their cognate ligands. Thus, the bulk of our information of LBD structure in the absence of hormone derives from the apo structure of RXRα. RXRs play a unique role as unliganded heterodimer partners for many members of the nuclear receptor family [5,6].
175
CHAPTER 11 The Vitamin D Receptor
three-dimensional structures of numerous members of the steroid receptor superfamily [67–69,121,122–126,128]. As with many of the steroid receptors, the threedimensional structure of the LBD of the VDR has also been determined [67–69]. As this is the subject of Chapter 15, only a brief summary of the structure of the VDR will be described here. Rochel et al. [67] first crystallized the human VDR LBD in complex with 1,25(OH)2D3 as well as other ligands [68]. A recent report by Vanhooke and colleagues has confirmed and extended this important work using the rat VDR [69]. Essential to the successful crystallography effort of both groups was the removal of a portion of the D region that prevented crystal formation. The three-dimensional structure of the LBD of the rat VDR is depicted in Fig. 3. The human and rat VDR structures revealed the presence of thirteen α-helices and three β-sheets that together form at its core a highly hydrophobic ligand binding pocket. Surprisingly, the overall volume of this pocket is much greater than that seen for many of the other steroid receptors. 1,25(OH)2D3 binds in this pocket in an extended configuration with the A ring in the β chair conformation and the 1α-OH group equatorial. The ligand is anchored by six hydrogen-bonding interactions, the 1α-OH group with Ser233, Arg270, the 3β-OH group with Tyr143 and Ser274, and the 25-OH group with His301 and His393 (using the rat VDR
Accordingly, it seems unlikely that the LBD of this receptor will faithfully mimic the other apo-receptor forms. Nevertheless, the structure has revealed the presence of 12 α-helices (H1–H12) arranged in an anti-parallel α-helical sandwich. Perhaps most interesting is the location of H12, which projects away from the body of the RXRα LBD [120]. As will be discussed hereafter and in other chapters in more detail, this projection is in contrast to the location of H12 in virtually all the liganded or holo-receptor forms wherein H12 is packed tightly against the body of the LBD [67–69,121,122–126]. This presumptive change in nuclear receptor structure in the AF-2 region in response to hormone, together with many other changes evident in the overall conformation of the receptor, appear integral to its ability to attract DNA binding partners and to recruit protein complexes essential for transcriptional activation. It is important to note here, however, that the idea that ligand induced significant changes in receptor structure emerged many years earlier from both receptor stability and proteolytic digestion studies [75,76, and references therein]. The latter analyses have been refined through the use of metabolically labeled nuclear receptors derived from in vitro translated receptor cDNAs [127] and are described in detail in Chapters 18 and 83. The reader is referred for details to the many primary publications that report the
A
DNA Binding
B
AF2
N
C 1
21
88
240
Hinge Region
423 Ligand Binding AF2
Deletion
N
C 116
165 211 240
Hinge Region
DRIP205
423 Ligand Binding
6H
N-KNHPMLMNLL KDH-C 625
637
NR Box II
FIGURE 3 Structural organization of the rat VDR in complex with l,25(OH)2D3. (A) Schematic organization of the specific functional domains of the wild-type rat VDR (upper panel) and a ligand binding domain/6x-histidine fusion protein (middle panel) used to prepare rat VDR LBD crystals as described by Vanhooke et al. [69]. Shown in the lower panel is the DRIP205 NR box II peptide that was cocrystallized with the rat VDR LBD. (B) Ribbon representation of the rat VDR LBD in complex with l,25(OH)2D3 and the DRIP205 coactivator peptide.
176 sequence numbering). Several of the LBD helices, including H9 and H10, provide a binding surface for the VDR’s DNA binding partner RXR (see later discussion). Importantly, H12 is packed against the body of the LBD such that it seals the ligand binding site and along with H3 and H4 creates the AF-2 hydrophobic cleft that serves as a docking site for transcriptional comodulators of VDR action (to be discussed later as well as in detail in Chapters 13–18). Several of the residues that form H12 make direct contact with the ligand. Importantly, the recent elucidation of the wildtype VDR from the zebrafish has confirmed that the removal of the “insertion region” located in the hinge region is without consequence to the overall structure of the receptor [129]. 4. ACTIVATION FUNCTIONS
As indicated earlier, hormone binding leads to the creation of a functional AF-2. The core of this functional domain appears to be residues associated with H12 such as human residues 416 to 423 [130,131], although early mutagenesis studies indicated more extensive involvement of components located upstream in the E/F region [7]. Indeed, H3 and H4 are also involved. The hydrophobic cleft formed by these three helices provides an essential binding site for single proteins that mediate linkage to more complex sets of factors that include the p160 histone acetyltransferase (HAT) coactivator complexes [132] and the D-receptor interacting protein (DRIP) complex (Mediator-D complex) [133], both of which are discussed hereafter and in detail in Chapters 14 and 16. Proteins such as the p160 coactivators SRC-1, SRC-2, and SRC-3, and DRIP205 interact directly within the ligand-induced H12 cleft via a short helical LxxLL motif that is common to these comodulators [134]. Indeed, studies have confirmed that LxxLL-containing peptides that bind to either the VDR or RXR function as dominant negative regulators of VDR action [135]. Vanhooke et al. [69] has provided three-dimensional structural insight into the interaction between the rat VDR and an LxxLL motif through cocrystallization of the rat VDR LBD with an LxxLL peptide derived from the second coactivator motif of DRIP205. This group observed that the LxxLL peptide folds into a short α-helix that binds to the surface groove formed by H3, H4, H5, and H12. The LxxLL interaction buries some 500 square angstroms of VDR surface area and involves direct contact with residues of VDR H3, H4, and H12 including Pro412, Leu413, and Glu416 of H12. These interactions form a “charge clamp” that has been described previously for both PPARγ and its coactivator SRC-1 [125]. The side chains of many of the LxxLL motif are buried within the pocket and surrounded by hydrophobic residues.
J. WESLEY PIKE AND NIRUPAMA K. SHEVDE
Nonspecific amino acids project directly into the solvent and away from the binding cavity. These findings support a strong structural basis for the interaction of the VDR with at least one coactivator and may lead to new insights with regard to VDR–coactivator interaction and stability. 5. LIGAND BINDING AND DIFFERENTIAL ACTIVATION
Since 1,25(OH)2D3 acts as a small molecular switch to activate the VDR, it should not be surprising that the binding of ligand would induce sweeping conformational changes in the VDR LBD. Changes in conformation are also likely to be sensitive to the overall structure of the ligand, supporting the hypothesis that synthetic analogs of natural hormonal ligands may be capable of inducing unique changes in receptor, thereby altering their biological function. Indeed, synthetic ligands of the ER as well as other nuclear receptors have been identified that produce both unique receptor conformations as identified by their crystal structures [123,136,137] and that manifest surprising tissue selectivity [138,139]. Numerous vitamin D analogs have also been created. Interestingly, although these compounds manifest unusual actions both in vivo and in vitro (see Chapters 82–88), none have demonstrated their ability to induce a conformation within the VDR that is different from that induced by 1,25(OH)2D3 via crystallography [68,69]. Thus, the molecular basis for their potentially useful therapeutic actions remains unclear.
V. FUNCTIONAL ANALYSIS OF THE VITAMIN D RECEPTOR A. Osteocalcin Regulation as a Model for VDR-Regulated Transcription Osteocalcin (OC) is a relatively abundant noncollagenous protein found in bone. Genetic ablation of the OC gene in mice suggests that this osteoblast-specific protein may contribute to the density and structural integrity of bone [140]. Despite the uncertainty surrounding the function of OC, its expression at the transcriptional level is regulated by a number of cytokines, growth factors, and systemic hormones, one of which is 1,25(OH)2D3 [141,142]. This fact together with the cloning of the human OC gene and its promoter in 1986 [143] provided the raw materials for investigation of 1,25(OH)2D3 action on gene expression at the molecular level. Initial efforts focused on determining if 1,25(OH)2D3 could stimulate the activity of the human OC gene promoter linked to a reporter gene when introduced
177
CHAPTER 11 The Vitamin D Receptor
into bone cells via transfection [103,144,145]. The results suggested that 1,25(OH)2D3 strongly induced this gene promoter, indicating that the effects were direct [146]. Kerner et al. [144] localized the cis-acting element within the promoter subsequently to a region approximately 500 bp upstream of the transcriptional start site. This study and an additional one by Ozono et al. [147] precisely mapped the first vitamin D response element (VDRE) to a directly repeated hexanucleotide sequence separated by three base pairs. Subsequent studies by others using the rat OC gene promoter resulted in a similar identification [148–150]. These studies provided the first insights into DNA sequences within genes that mediated vitamin D-inducible activity. Many genes have been explored since that time for their sensitivity to 1,25(OH)2D3, including genes for mouse osteopontin (OPN) [151], mouse calbindinD28K [152], rat calbindin D9K [153], rat [154,155] and human [156] 25-hydroxyvitamin D3 24-hydroxylase (two VDREs), human p21 [157], and a host of others. A partial list of the sequences that have been shown to mediate vitamin D action together with their locations within the promoters is documented in Table III and elsewhere in this book. These sequences reveal that a “typical” VDRE comprises two hexanucleotide repeats separated by a 3-bp spacer. Although the sequence of the “spacer” is not generally conserved, recent studies have suggested that these base pairs may influence VDR binding [158]. Nevertheless, the general half-site consensus sequence is AGGTCA or GGTTCA, although considerable variability is apparent. Chapter 18 describes many of these VDREs in more detail. It also identifies additional putative VDREs in gene promoters that are an “atypical” configuration; however, the validity of these elements in the context of their natural promoters will require additional investigation.
TABLE III Gene Rat osteocalcin Human osteocalcin Mouse osteopontin Rat calbindin-D9K Mouse calbindin-D28K Rat 24 hydroxylase Human 24 hydroxylase Rat pit 1 Human p21
The efforts to define VDREs were part of a larger effort by numerous investigators to learn about steroid receptor response elements in general, with regard to specificity as well as to receptor selectivity [159]. These overall efforts led to our current understanding that nuclear receptor binding sites can be categorized into three groups: palindromic half-sites that interact with receptors for the sex steroids; directly repeated half-sites with variable spacing that interact with receptors for 1,25(OH)2D3, retinoic acid, thyroid hormone, and a number of other ligands; and single half-sites that mediate the actions of receptors such as NGF-IB that act as monomers [6]. The reader is referred elsewhere [1–6] for a complete review of the nature of these DNA binding sites. The binding of the VDR to endogenous gene promoters in living cells has been examined by several investigators including Sierra et al. [160], Zhang et al. [161]; and Pike and colleagues [54,55] using chromatin immunoprecipitation techniques. In these assays, cultured cells are treated with 1,25(OH)2D3 for specific periods of time, and the cells are then harvested and treated briefly with formaldehyde to fix the proteins that are bound to DNA. Cellular chromatin is then isolated, sonicated to produce small random fragments of 400–500 bp, and precipitated with antibodies to the protein of interest. The coprecipitated DNA is then isolated and its abundance evaluated by PCR using primers designed to amplify specific regions of gene promoters that are of interest. The concentration of promoter DNA correlates directly with immunoprecipitated protein. Importantly, 1,25(OH)2D3 stimulated the appearance of the VDR on several genes promoters including those for OC [160] as well as Cyp24 and OPN [55]. The results of a similar experiment for Cyp24 and OPN promoters are seen in Fig. 4. Although this
Location and Sequence of Positive Natural Vitamin D Response Elements Location −460/−446 −499/−485 −757/−743 −489/−475 −198/−183 −150/−136 (Proximal) −258/−244 (Distal) −169−/155 (Proximal) −291/−277 (Distal) −67/−52 −779/−765
Nucleotide sequence GGGTGA atg AGGACA GGGTGA acg GGGGCA GGTTCA cga GGTTCA GGGTGT cgg AAGCCC GGGGGA tgtg AGGAGA AGGTGA gtg AGGGCG GGTTCA gcg GGTGCG AGGTGA gcg AGGGCG AGTTCA ccg GGTGTG AGTTCA gcga AGTTCA AGGGAG att GGTTCA
178
J. WESLEY PIKE AND NIRUPAMA K. SHEVDE
A
Mouse CYP24 promoter
VDRE2
VDRE1
−236
ORF control
−51
+1
+11055
+11252
Mouse osteopontin promoter
VDRE
−854
ORF control +1
−658
B
+5181
CYP24
1,25 (OH)2D3
−
+
+5380
Osteopontin −
+
−
+
−
+
αVDR αRXR -Ab INPUT promoter
ORF
promoter
ORF
FIGURE 4 1,25(OH)2D3-induced association of the VDR and RXR to the promoters for Cyp24 and OPN. (A) Schematic diagram of the murine Cyp24 and OPN promoters. The locations of the VDREs within the two promoters are indicated with +1 representing the start site of transcription. Primer sites utilized to amplify either the VDRE in the promoter region or the open reading frame (ORF) are indicated. (B) VDR and RXR binding to the OPN or Cyp24 promoters as assessed by chromatin immunoprecipitation. For details, see [54].
approach does not permit the exact identification of the binding site for the VDR on a target gene, the results support localization of the VDR to a region within the promoter that contains a previously identified VDRE. This type of study is consistent with virtually all previous in vitro work that has led to the identification of VDREs in the OC, Cyp24, and OPN gene promoters.
B. VDR Binding to Specific DNA Involves RXR High-affinity binding of the VDR to DNA in vitro requires two regions, the DNA binding domain and the LBD. This conclusion is drawn from numerous mutagenesis experiments that revealed that amino acid alterations either in the zinc finger modules [104,162–168] or in the VDR LBD [103,107,130,169] were capable of blocking DNA binding. Interestingly, the molecular basis for altered DNA binding in the two regions is different. In the first case, changes in the VDR DNA binding domain prevent interaction directly with DNA
(see Chapter 13 for details of specific DNA binding). In the second, however, changes in LBD residues block the ability of the receptor to form the protein dimers that are essential for the high-affinity VDRE binding [107,169]. The discovery that the VDR must form a dimer to bind DNA was suggested by the structural nature of VDREs, which were composed of two repeated halfsites. Surprisingly, however, it was found that the VDR bound to DNA not as a homodimer but rather as a heterodimer as illustrated in the cartoon in Fig. 5. Liao et al. [170] and Sone et al. [162,171,172] both observed that while VDR derived from mammalian cell extracts was fully capable of binding to DNA in vitro, receptor produced either by in vitro transcription/translation or through recombinant means in nonmammalian cells failed to produce DNA binding-competent VDR. The inactivity of the VDR could be rescued, however, through the addition of general mammalian cell extract, suggesting the presence of a VDR DNA-binding facilitator. This factor(s) was discovered in a variety of tissue
179
CHAPTER 11 The Vitamin D Receptor
CBP/p300 PCAF
DRIP
SRC1
HAT 1,25(OH)2D3
RXR
Chromatin
Basal machinery
VDR
VDRE DR3
FIGURE 5
Contemporary model for VDR- and RXR-mediated gene regulation. The VDR and RXR bind to the promoter of a target gene as a heterodimer in response to 1,25(OH)2D3. Transcriptionally active complexes that include the SRC/CBP/p300 HAT coactivator complex or the DRIP205/Mediator complex or perhaps others are subsequently recruited to facilitate changes in chromatin architecture and to enhance RNA polymerase II entry. Both may be needed to modulate gene expression.
and cell sources [172] and termed nuclear accessory factor (NAF); its presence was subsequently confirmed by others [173,174]. NAFs were also discovered for other nuclear receptors including those for thyroid hormone and retinoic acid. In 1991 and 1992, Yu et al. [175], Leid et al. [176], Zhang et al. [177], and Kliewer et al. [178] all discovered that RXR, a previously cloned member of the nuclear receptor family, was the likely NAF protein partner. Three isotypes of RXR (α, β, and γ) are known and may be present in a given cell type. Considerable effort has gone into defining the nature and function of RXR-mediated DNA binding, and the reader is referred to the many reviews that deal with this complex issue [2,5]. Nevertheless, it is now clear that both VDR and RXR subunits are necessary not only to enable high-affinity DNA binding, but also to contribute to the activation process (see Chapter 13). Jin et al. [107,169] utilized an extensive series of internal deletions of the VDR to define two regions essential for interaction with RXR. These regions coincide with the two moderately conserved subregions of homology located in the E domain of the VDR and are associated directly with H9 and H10 of the crystal structure of the RXR LBD [120]. Although there has been some disagreement with regard to domain requirements for dimerization of other receptors [179,180], few have been cocrystallized together with
RXR, which would allow a more precise understanding of dimerization at the atomic level. The interaction surfaces appear to be large, however, and are widely believed to involve H9 and H10 of each protein. Unfortunately, the deletion analysis described by Jin et al. [107] was unable to distinguish between structural abnormalities induced by the deletion and the actual dimerization domain itself. Thus, elucidation of the three-dimensional structure of VDR-RXR LBD heterodimer will be required to define precisely the interaction between VDR and RXR. Despite these in vitro studies, simple experiments designed to unequivocally demonstrate an absolute requirement for RXR in VDR function in intact cells have not been carried out, primarily because of the ubiquitous expression of endogenous RXR in mammalian cell lines. This remains important since it has been proposed that the activation of the osteopontin gene by 1,25(OH)2D3 involves only VDR homodimers [181]. It is also important because the ability of 1,25(OH)2D3 to suppress several growth factor/cytokines genes such as those for IL-2 [182], GM-CSF [183], and interferon γ [184] as well as human PTH [185] may not require RXR. This suggests that the need for RXR in vitamin D gene regulation may be gene promoter–specific. Jin and Pike [106] utilized yeast, which do not express VDR and RXR′, to recreate the VDR transcriptional
180 response unit and to test for the requirement of RXRs on a typical VDRE. While VDR exhibited little capacity to activate a chimeric gene promoter containing a VDRE derived from the OC genes in the absence of RXR, the addition of RXR dramatically increased VDR-mediated activation. Despite this study, perhaps the most definitive assessment of RXR participation has come from studies by Barsony and colleagues [186] using immunofluorescence and Pike and colleagues [54,55] using the chromatin immunoprecipitation techniques. Prufer et al. [186] discovered that both VDR and RXR colocalize in response to 1,25(OH)2D3 within the nucleus in a VDR DNA-binding domain-dependent manner as discrete complexes that are believed to represent sites of transcription. As depicted in Fig. 4, the studies by Kim et al. [55] showed unequivocally that treatment of cells with 1,25(OH)2D3 leads to the rapid colocalization of both VDR and RXR on a similar fragment of the Cyp24 gene promoter, a gene that contains typical VDREs and is believed to require a VDR/RXR heterodimer for activation. Although the resolution of binding in these intact cells does not permit direct determination of heterodimer formation, this conclusion is a reasonable one given the earlier in vitro biochemical interaction studies. Interestingly, 1,25(OH)2D3 also promotes the colocalization of both VDR and RXR to the osteopontin promoter as seen in Fig. 4 [55]. These data indicate that VDR homodimer-mediated activation of the osteopontin gene is unlikely.
C. Polarity of DNA Binding The asymmetric nature of natural VDREs (direct repeats) coupled with the heteromeric nature of the receptor activation unit (VDR-RXR) indicates that the two receptor subunits must bind to the VDRE with a defined polarity. Studies by Jin and Pike [106] and Freedman and co-workers [164] addressed this question in detail. Through the use of chimeric receptors and chimeric response elements, it is now clear that RXR binds to the upstream 5′ half-element and the VDR binds to the downstream 3′ half-element of VDREs oriented on the DNA sense strand as illustrated in Fig. 5. This organization is consistent with the relative polarity noted for both RXR-TR [180] and RXR-RAR heterodimers [187,188] bound to their respective response elements.
D. Transactivation by the VDR and RXR Involves Coregulators Studies by McDonnell et al. [103] first demonstrated that the responsiveness of a transfected human
J. WESLEY PIKE AND NIRUPAMA K. SHEVDE
OC promoter to 1,25(OH)2D3 in VDR-negative cell lines such as CV-1 or COS7 required co-introduction of VDR. This biological assay enabled an extensive examination of the domains of the VDR that are required for transactivation and revealed the not surprising crucial nature of both the DNA binding and heterodimerization functions in transactivation. They also revealed the critical role of an intact AF-2 function in the context of normal DNA binding and heterodimer function for subsequent transactivation. Interestingly, this assay also established the inactive nature of many mutant VDRs that were identified in the human syndrome of hereditary l,25(OH)2D3-resistant rickets (HVDRR) considered in more detail in Chapter 72 [104,167,168]. Activation of the VDR involves a l,25(OH)2D3dependent conformational change within the LBD, which as indicated earlier creates a functional AF-2 domain. The protein components that interact within the AF-2 region of the nuclear receptors in general include the SWI/SNF complex that functions to remodel chromatin in an ATP-dependent manner [189,190], the p300/CBP coactivator complex [191,192] that contains nuclear receptor interacting proteins SRC1, SRC2, and/or SRC3 [132] and function to acetylate histones [193, 194], mediator complex that contains the nuclear receptor interacting protein DRIP205 and functions to facilitate entry of RNA polymerase II [133,195,196], and likely other [197] complexes. The potential interactions of the SRC-1/CBP and the DRIP complexes with the VDR/RXR heterodimer are illustrated in Fig. 5. The exact role of each of these enzyme-containing complexes during transcription is unclear as is the temporal order in which they are recruited to receptor-bound promoters during activation. The VDR has been shown to interact directly with specific components of each of these complexes in vitro [197–200], and the addition of many of these factors to cells via transfection assays strongly enhances l,25(OH)2D3-induced transcription. Direct evidence that the VDR recruits many of these factors to target promoters in living cells has only recently emerged, however, through the use of chromatin immunoprecipitation assays. Zhang et al. [161] and Kim et al. [55], both have demonstrated that l,25(OH)2D3 induces the recruitment of SRC-1 to the Cyp24 promoter. Kim et al. [55] also showed that SRC-2 and SRC-3 were similarly recruited, although the temporal pattern of recruitment differed among these coactivators. This group also observed that the HAT containing integrators p300 and CBP as well as mediator component DRIP205 were also recruited to the Cyp24 promoter in response to l,25(OH)2D3 [55]. Importantly, the consequence of this recruitment in the case of DRIP205 appears to be association of RNA polymerase II [55], whereas in the case of the p160 complex they result in the acetylation of histone 4 (Kim and Pike, unpublished).
181
CHAPTER 11 The Vitamin D Receptor
of the RXR H12 region by Thompson et al. [209]. The concept that RXR is not a “silent partner” but rather a transcriptionally active partner of the VDR has been reinforced also through studies by Pathrose et al. [135], who used RXR-selective LxxLL peptides to demonstrate that blockade of RXR recruitment of coactivators also leads to inhibition of transcription induced by l,25(OH)2D3. Finally, recent studies by Bettoun et al. [210] also support the contribution of RXR to VDR-induced transcription. Thus, it seems likely that both subunits of the VDR heterodimer complex are essential to the regulation of gene expression by l,25(OH)2D3.
These studies provide clear evidence not only that the VDR interacts with these coregulators on specific genes in living cells but that the recruitment of these complexes leads to functional consequences with regard to gene promoter status and transcriptional activity. A number of proteins also interact with the VDR in a ligand-independent manner in regions outside that of AF-2. These include the protein NCoA62, a coactivator that may play a role in mRNA transcript splicing [201]; WINAC [202]; Alien [203]; hairless [204]; SUG1, a 19S proteasomal component that may play a role in VDR degradation [205]; E6-AP, an E3 ligase that may play a role in VDR processing [Pike et al., unpublished]; and corepressors such as NCoR [206] and SMRT [207]. Little is known of how these proteins interact with the VDR or of the domain(s) within the VDR that mediates the interactions. The VDR also interacts with the core promoter transcription factor TFIIB [108,208], suggesting that the biochemical mechanisms by which the VDR contributes regulatory inputs into the basal transcriptional apparatus will be multifaceted and complex. The reader is referred to Chapters 13–17 for additional details regarding the interaction of the VDR with potential comodulators and the core transcription factor TFIIB.
F. VDR Dynamics Apart from the involvement of VDR and RXR as well as numerous coregulators in transcription, recent studies have suggested that the process is highly dynamic. Indeed, general studies in cultured cells using chromatin immunoprecipitation and other assays have indicated that the binding of nuclear receptors to target gene promoters is cyclic in nature [211–213]. Accordingly, hormone-induced binding of the receptor to gene promoters exhibited cycles with a periodicity ranging from 30 to 40 min. The recruitment of coregulator complexes is also cyclic, suggesting the possibility of a spatial and temporal order to the transcriptional process that is necessary for gene regulation [214]. Whether this cycling process is fundamental to the mechanism whereby genes are activated and transcription is maintained by transcription factors is currently unclear. The binding of the VDR and RXR to the Cyp24 and OPN promoters in response to l,25(OH)2D3 is also cyclic, as seen in Fig. 6 [55]. This binding is accompanied in parallel fashion by coactivators such as the p160 family of genes, CBP, and p300 as well as
E. Direct Involvement of RXR in Transactivation One question that has emerged from many of the functional studies of VDR is whether RXR contributes directly to the recruitment of the coactivators described earlier. This is particularly relevant since RXR also contains a functional AF-2 which interacts in vitro with most of the coactivator complexes described earlier. That RXR contributes directly to VDR-mediated transactivation has been demonstrated through mutagenesis
B
CYP24
Relative intensity, ×10−3
VDR
120
RXR pol II
80 40 0
OPN 120
VDR RXR pol II
80
40
Time (min)
165
150
135
120
105
90
75
60
45
30
15
0
165
150
135
120
105
90
75
60
45
30
15
0 0
Relative intensity, ×10−3
A
Time (min)
FIGURE 6 Localization of VDR, RXR, and RNA pol II to the (A) Cyp24 and (B) OPN promoters in response to 1,25(OH)2D3 as a function of time. The results are a densitometric depiction of the PCR products detected following amplification of either Cyp24 or OPN gene promoter DNA. For details, see Kim et al. [55].
182
J. WESLEY PIKE AND NIRUPAMA K. SHEVDE
Coactivator complexes
VDRE
TATA
Ligand-driven complex assembly
1,25(OH)2D3
RXR
VDR PCAF
HAT complex
RXR VDR
CBP/p300 p160s RXR VDR
Proteasome-driven disassembly
HISTONE
VDRE
HISTONE
Coactivator complexes
DRIP complex
DRIP 205 RXR VDR VDRE
TATA
FIGURE 7
Model for the 1,25(OH)2D3-driven assembly of transcriptionally active regulatory complexes on target genes and proteasome-directed disassembly of these complexes during signal processing. The order of recruitment and the influence of the gene itself on this process are currently unknown.
by the receptor interacting component of the DRIP complex, DRIP205 [55]. While current efforts are in place to determine the fundamental role of cycling in nuclear receptor activation, it is clear that this process may well enable precise real-time cellular monitoring of ambient hormone concentrations. A model for l,25(OH)2D3-driven assembly of VDR/RXR cofactor complexes on DNA and their proteasome mediated disassembly drawn from Kim et al. [55] is seen in Fig. 7.
G. General Conclusions An understanding of the functional activity of the VDR has emerged as a direct result of biochemical, molecular biological, and structural investigation. While insights gained from the three-dimensional structure of the VDR and other nuclear receptors have been exceptionally revealing, they have benefited enormously from numerous earlier biochemical and molecular biologic studies that have provided a wealth of information regarding the general domain structure and function of the nuclear receptor family. It is, in fact, a testimony to the effectiveness of these latter approaches that such great insights have emerged and that the crystallography has in many cases been largely confirmatory. The three approaches have, however, been highly complementary and have greatly extended our understanding of nuclear receptor structure and function.
VI. THE HUMAN VITAMIN D–RECEPTOR CHROMOSOMAL GENE A. Organization of the Gene The cloning of the VDR structural gene in 1987 initiated over a decade of highly productive research on the mechanism of action of vitamin D. Interestingly, unlike many of the other members of the nuclear receptor family, which exist as products of multiple genes, the VDR itself remains the apparent product of a single gene. Indeed, the recently published sequence analysis of both the mouse and human genomes appears to confirm the existence of only one VDR [91,92]. The human VDR gene is reported to lie on chromosome 12 [215]. The initial organization of the intron–exon structure of human VDR chromosomal gene corresponding to the sequence of the VDR reported by Baker et al. [85] was determined in 1988 [104]. Additional efforts have defined the complete structure of the gene [216]. Since the VDR gene is considered in some depth in Chapter 12, it will only be described briefly in this section. Restriction mapping of several lambda clones and a series of four recovered human cosmid clones coupled to nucleotide sequence analysis of relevant portions of these clones revealed a gene spanning over 75 kb of DNA. Eight exons comprise the coding sequence of the VDR protein. The first of these is exon 2, which contains the most proximal 3 bp of the 5′ noncoding
183
CHAPTER 11 The Vitamin D Receptor
sequence, the translation start site, and nucleotide sequence that encodes the first zinc finger module. Exon 3, which lies approximately 15 kb downstream, encodes the second zinc finger module. Exons 4, 5, and 6 encode the D region or hinge. Exons 6, 7, 8, and 9 encode a portion of the hinge and the carboxy-terminal E/F region as well as approximately 3200 nucleotides of 3′ noncoding sequence. The relationship between these exons, the full-length mRNA, and the VDR protein itself is illustrated in Fig. 8. It is clear that the human chromosomal gene for the VDR is not unlike other steroid receptor genes in size and exon organization. Insight into the 5′ end of this large gene and its promoter has also emerged over the past decade, revealing the promoter and regulatory regions of the human VDR to be highly complex (Fig. 8). This seems to be untrue of the rodent VDR genes where a single promoter is likely responsible for VDR expression [217]. Two short exons lie upstream of exon 2 and account for the known 5′ noncoding sequence reported by Baker et al. [85]. These exons, termed 1a and 1c, are 77 and 81 bp in length. A promoter (P1) lies immediately upstream of exon 1a and is clearly characterized by its GC-rich nature and the absence of a TATA box. Interestingly, an exon of 121 bp not found in the originally reported sequence of Baker et al. [85] and termed exon 1b is located 4.5 kb downstream of exon 1a. Variable use of exons 1b and 1c leads to the production
−281 −204
of alternatively spliced mRNAs whose nature and function remain unknown. More recently, additional upstream exons and their associated promoters have been ascribed to the human VDR gene [218]. These additions, as well as the transcripts and protein products that they produce [219], are described more completely in Chapter 12. Since the relative abundance of the bulk of these transcripts is exceedingly low, their contribution to overall VDR expression and to the various potential forms of the VDR proteins that are produced remains controversial. Clearly, additional work will be required to establish the bona fides of these novel forms of the VDR. Nevertheless, the presence of these transcripts suggests unique and possibly tissue-selective regulation of VDR expression. The identity of regulatory control elements in the human gene as well as those in the mouse gene is only now beginning to emerge.
B. Polymorphisms within the VDR Gene Genetic polymorphisms have been identified within the human VDR gene. The first represents a C to T transition in the translation initiation site located in exon 2 [94,95]. The presence of a C in this position results in the initiation of translation at a site three codons downstream and in the production of a gene
−83 −2
FIGURE 8 Structural organization of the human chromosomal VDR gene. The human VDR gene locus (DNA) is composed of as many as 14 exons (1a–1f, 2 through 9) spanning over 75 kb of DNA. A 10-kb scale is indicated at the right. The location of exons relative to the mRNA transcript of ~4800 nucleotides (mRNA) and the encoded VDR protein of 427 amino acids (hVDR) is illustrated. With regard to the hVDR mRNA, negative numbers indicate 5′ noncoding nucleotides and positive numbers indicate protein encoding nucleotides beginning with +1 indicated by Baker et al. [85]. Numbers below the hVDR protein indicate the amino acid residue boundaries of shaded homology domains. Regions of functionality are designated A/B, C, D, and E/F as in Figs. 2 or 3.
184 product three amino acids shorter. The distribution of this polymorphism in the human population and a potential relationship between the frequency of this polymorphism and bone mineral density were discussed in an earlier section of this chapter, and additional studies have been carried out which explore this issue in greater detail [97–101]. One finding has been the observation that the smaller form of the VDR appears to be more active transcriptionally [96–98]. Additional polymorphisms, although they are not located in regions of the gene that might directly affect the structure of the protein, have also been defined within the introns located between exons 7 and 8 as well as within the 3′ noncoding region of exon 9 [220, see Chapter 68 and references therein]. These polymorphisms appear also to correlate with bone mineral density in several human populations and are hypothesized to be predictive for osteoporosis. At presence, these conclusions are highly controversial, however, and are not widely supported by numerous recent studies that have been carried out over the past several years. This topic and associated references are considered in depth in Chapter 68.
VII. CONCLUDING COMMENTS The basic elements of the mechanism of action of vitamin D have been defined. The pace of exploration into the actions of vitamin D has accelerated enormously since the mid-1980s, largely as a result of the molecular cloning of the VDR in 1987 but also as a result of the availability of cloned vitamin D target genes. As described in this chapter, we have gained considerable insight into the structure of the VDR and its compartmentalization into definable functional domains. The now-available crystal structure of the VDR LBD has placed much of this biochemical and molecular biologic work into a clearer perspective. The availability of recombinant clones has allowed investigation of the interaction of the VDR with vitamin D–inducible gene promoters and definition of VDREs. Further studies have revealed that the VDR requires a protein partner for DNA binding in the form of RXR, a central regulator of several nuclear receptors. Although additional research will be necessary, a considerable understanding of the role of receptor and its ligand in the regulation of transcription has begun to emerge. These insights are currently being utilized to gain an understanding of the tissue-selective mechanisms of action of a new generation of vitamin D analogs under consideration as therapeutic agents for a broad range of indications that include skin diseases, immunologic disorders, and cancer. The cloning of the
J. WESLEY PIKE AND NIRUPAMA K. SHEVDE
VDR also enabled the recovery of its chromosomal gene. Subsequent investigation into the nature of the human syndrome of hereditary l,25(OH)2D3-resistant rickets (HVDRR) was made possible through characterization of the VDR gene itself [104,216] and has revealed the underlying cause to be mutations in the gene that produce dysfunctional receptor protein (described in Chapter 72). This discovery, together with the genetic ablation of the VDR gene in mice (see Chapter 20 for details) that clearly mimics physiologically the syndrome of HVDRR, confirm the central role of the VDR in the regulation of mineral metabolism. In view of our current understanding of VDR action on target gene promoters, future studies are likely to focus on the downstream roles of co-regulator complexes in mediating VDR action at the individual gene promoter level. Questions that need to be answered include those that address the individual roles of coregulator complexes in transcription, the order in which they are recruited to individual promoters, and whether features of the promoter itself play a deterministic role in the recruitment process and cellular response. Elucidation of these processes will be essential to the eventual understanding of the unique actions of vitamin D analogs that are potentially useful therapeutically. Also likely to emerge from these studies will be a better understanding of the regulation of the VDR at both the transcriptional and posttranslational levels and the contribution of this regulation to the mechanism of vitamin D action. Perhaps more important than these molecular details is the likelihood that we will achieve a better understanding of how vitamin D controls directly as well as indirectly the expression of broad networks of genes that are responsible in turn for cellular and tissue activities such as those involving proliferation and differentiation. Thus, although significant progress has been made in the past few years there is still much work that needs to be done.
References 1. Beato M 1989 Gene regulation by steroid hormones. Cell 56:335–344. 2. Evans RM 1988 The steroid and thyroid hormone receptor super-family. Science 240:889–895. 3. O’Malley BW 1990 The steroid receptor superfamily: More excitement predicted for the future. Mol Endocrinol 4:363–369. 4. Beato M, Herrliche P, Schutz G 1995 Steroid hormone receptors: Many actors in search of a plot. Cell 83:851–857. 5. Mangelsdorf DJ, Evans RM 1995 The RXR heterodimer and orphan receptors. Cell 83:841–850. 6. Mangelsdorf DJ, Thummel C, Beato M, Herrliche P, Schultz G, Umesono K, Blumberg B, Kastner P, Mark M, Chambon P,
185
CHAPTER 11 The Vitamin D Receptor
7. 8.
9.
10.
11. 12. 13.
14.
15.
16.
17. 18. 19. 20.
21. 22. 23. 24. 25.
Evans RM 1995 The nuclear receptor superfamily: The second decade. Cell 83:835–839. Norman AW 1968 The mode of action of vitamin D. Biol Rev 243:4055–4064. Lawson DEM, Wilson PW, Kodicek E, Morrison HR, Williams DH 1971 Identification of 1,25-dihydroxycholecalciferol, a new kidney hormone controlling calcium metabolism. Nature 230:228–230. Norman AW, Myrtle JF, Midgett RJ, Nowicki HG, Williams V, Popjak G 1971 1,25-Dihydroxycholecalciferol: Identification of the proposed active form of vitamin D in the intestine. Science 173:51–54. Holick MF, Schnoes HK, DeLuca HF, Suda T, Cousins RJ 1971 Isolation and identification of 1,25-dihydroxycholecalciferol. A metabolite of vitamin D active in intestine. Biochemistry 10:2799–2804. Fraser DR, Kodicek E 1970 Unique biosynthesis by kidney of a biologically active vitamin D metabolite. Nature 228:764–766. Haussler MR, Norman AW 1969 Chromosomal receptor for a vitamin D metabolite. Proc Natl Acad Sci USA 62:155–162. McDonnell DP, Mangelsdorf DJ, Pike JW, Haussler MR, O’Malley BW 1987 Molecular cloning of complementary DNA encoding the avian receptor for vitamin D. Science 235:1214–1217. Pike JW, Marion SL, Donaldson CA, Haussler MR 1983 Serum and monoclonal antibodies against the chick intestinal receptor for 1,25-dihydroxyvitamin D3: Generation by a preparation enriched in a 64,000 dalton protein. J Biol Chem 258:1289–1296. Dame MC, Pierce EA, Prahl JM, Hayes CE, DeLuca HF 1986 Monoclonal antibodies to the porcine intestinal receptor for 1,25-dihydroxyvitamin D3: Interaction with distinct receptor domains. Biochemistry 25:4523–4534. Brumbaugh PF, Haussler MR 1974 1,25-Dihydroxycholecalciferol receptors in intestine. Temperature-dependent transfer of the hormone to chromatin via a specific receptor. J Biol Chem 249:1258–1262. Lawson DEM, Wilson PW 1974 Intranuclear localization and receptor proteins for 1,25-dihydroxycholecalciferol in chick intestine. Biochem J 144:573–583. Brumbaugh PF, Haussler MR 1975 Specific binding of 1,25dihydroxycholecalciferol to nuclear components of chick intestine. J Biol Chem 250:1588–1594. Brumbaugh PF, Hughes MR, Haussler MR 1975 Cytoplasmic and nuclear components for 1,25-dihydroxyvitamin D3 in chick parathyroid glands. Proc Natl Acad Sci USA 72:4871–4875. Pike JW, Gooze LL, Haussler MR 1980 Biochemical evidence for 1,25-dihydroxyvitamin D receptor macromolecules in parathyroid, pancreas, pituitary, and placental tissues. Life Sci 26:407–414. Kream BE, Jose M, Yamada S, DeLuca HF 1977 A specific high affinity binding macromolecule for 1,25-dihydroxyvitamin D3 in fetal bone. Science 197:1086–1088. Chandler JS, Pike JW, Haussler MR 1979 1,25-Dihydroxyvitamin D3 receptors in rat kidney cytosol. Biochem Biophys Res Commun 90:1057–1063. Colston K, Feldman D 1980 Nuclear translocation of the 1,25-dihydroxycholecalciferol receptor in mouse kidney. J Biol Chem 255:7510–7513. Kream BE, Yamada Y, Schnoes HK, DeLuca HF 1977 Specific cytosol-binding protein for 1,25-dihydroxyvitamin D3 in rat intestine. J Biol Chem 254:9488–9491. Abe E, Miyaura C, Sakagami H, Takeda M, Konno K, Yamazaki T, Yashiki S, Suda T 1981 Differentiation of mouse
26.
27. 28. 29. 30. 31. 32. 33. 34.
35.
36. 37.
38. 39.
40. 41. 42. 43. 44.
myeloid leukemia cells induced by 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 78:4990–4994. Shavit ZB, Teitlebaum SL, Reitsma P, Hall A, Pegg LE, Trial J, Kahn AJ 1983 Induction of monocytic differentiation and bone resorption by 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 80:5907–5911. Christakos S, Norman AW 1981 Biochemical characterization of 1,25-dihydroxyvitamin D3 receptors in chick pancreas and kidney cytosols. Endocrinology 108:140–149. Haussler MR, Manolagas SC, Deftos L 1980 Evidence for a 1,25-dihydroxyvitamin D3 receptor-like macromolecule in rat pituitary. J Biol Chem 255:5007–5010. Dokoh S, Donaldson CA, Marion SL, Pike JW, Haussler MR 1983 The ovary: A target organ for 1,25-dihydroxyvitamin D3. Endocrinology 112:200–206. Merke J, Kreusser W, Bier B, Ritz E 1983 Demonstration and characterization of a testicular receptor for 1,25-dihydroxyvitamin D3 in the rat. Eur J Biochem 130:303–308. Colston K, Hirst M, Feldman D 1980 Organ distribution of the cytoplasmic 1,25-dihydroxycholecalciferol receptor in various mouse tissues. Endocrinology 107:1916–1922. Walters MR, Wicker DC, Riggle PC 1986 1,25-Dihydroxyvitamin D3 receptor identified in the rat heart. J Mol Cell Cardiol 18:67–72. Colston K, Colston MJ, Fieldsteel AH, Feldman D 1982 1,25-Dihydroxyvitamin D receptors in human epithelial cancer cell lines. Cancer Res 42:856–859. Colston K, Colston MJ, Feldman D 1981 1,25-Dihydroxyvitamin D and malignant melanoma: The presence of receptors and inhibition of cell growth in culture. Endocrinology 108:1083–1086. Eisman JA, Martin TJ, MacIntyre I, Frampton RJ, Moseley JM, Whitehead R 1980 1,25-Dihydroxyvitamin D3 receptors in a cultured human breast cancer cell line (MCF-7). Biochem Biophys Res Commun 93:9–15. Haussler MR 1986 Vitamin D receptor: Nature and function. Annu Rev Nutr 6:527–562. Feldman D, Chen T, Hirst M, Colston K, Karasek M, Cone C 1980 Demonstration of 1,25-dihydroxyvitamin D receptors in human skin biopsies. J Clin Endocrinol Metab 51: 1463–1465. Prowedini DM, Tsoukas CD, Deftos LJ, Manolagas SC 1983 1,25-Dihydroxyvitamin D3 receptors in human leucocytes. Science 221:1181–1183. Bhalla AK, Amento EP, Clemens TL, Holick MF, Krane SM 1983 Specific high affinity receptors for 1,25-dihydroxyvitamin D3 in human peripheral blood mononuclear cells: Presence in monocytes and induction in T lymphocytes following activation. J Clin Endocrinol Metab 57:1308–1310. Krageballe K 1992 Treatment of psoriasis with calcipotriol and other vitamin D analogues. J Am Acad Dermatol 27: 1001–1008. Simpson RU, DeLuca HF 1980 Characterization of a receptorlike protein for 1,25-dihydroxyvitamin D3 in rat skin. Proc Natl Acad Sci USA 77:5822–5827. Zile M, Barsness EC, Yamada S, Schnoes HK, DeLuca HF 1976 Localization of 1,25-dihydroxyvitamin D3 in intestinal nuclei in vivo. Arch Biochem Biophys 186:15–24. Jones PG, Haussler MR 1979 Scintillation autoradiography localization of 1,25-dihydroxyvitamin D3 in chick intestine. Endocrinology 104:313–321. Stumpf WE, Sar M, Reid FA, Tanaka Y, DeLuca HF 1979 Target cells for 1,25-dihydroxyvitamin D3 in intestinal tract, stomach, kidney, skin, pituitary, and parathyroids. Science 206:1188–1190.
186 45. Stumpf WE, Sar M, Clark SA, DeLuca HF 1982 Brain target sites for 1,25-dihydroxyvitamin D3. Science 215:1403–1405. 46. Clemens TL, Garrett KP, Zhou XY, Pike JW, Haussler MR, Dempster DW 1988 Immunocytochemical localization of the 1,25-dihydroxyvitamin D3 receptor in target cells. Endocrinology 122:1224–1230. 47. Berger U, Wilson P, McClelland RA, Colston K, Haussler MR, Pike JW, Coombes RC 1988 Immunocytochemical detection of 1,25-dihydroxyvitamin D3 receptors in normal human tissues. J Clin Endocrinol Metab 67:607–613. 48. Milde P, Merke J, Ritz E, Haussler MR, Rauterberg EW 1989 Immunocytochemical detection of 1,25-dihydroxyvitamin D3 receptors by monoclonal antibodies: A comparison of four immunoperoxidase methods. J Histochem Cytochem 37:1609–1617. 49. Berger U, Wilson P, McClelland RA, Colston K, Haussler MR, Pike JW, Coombes RC 1987 Immunocytochemical detection of 1,25-dihydroxyvitamin D3 receptors in breast cancer. Cancer Res 47:6793–6799. 50. Nevah-Many T, Marx R, Keshet E, Pike JW, Silver J 1990 Regulation of 1,25-dihydroxyvitamin D3 receptor gene expression by 1,25-dihydroxyvitamin D3 in the parathyroid in vivo. J Clin Invest 86:1968–1975. 51. Mocharla H, DeTogni P, Yu XP, White P, Jilka RJ, Manolagas SC 1994 In search for an association between vitamin D receptor polymorphism and VDR mRNA expression in peripheral blood mononuclear cells of normal volunteers. J Bone Miner Res 9(Suppl. 1):S416. 52. Walters MR, Hunziker W, Norman AW 1980 Unoccupied 1,25-dihydroxyvitamin D3 receptors. Nuclear/cytosol ratios depend on ionic strength. J Biol Chem 255:6799–6805. 53. Walters MR, Hunziker W, Norman AW 1981 1,25-Dihydroxyvitamin D3 receptors: Intermediates between triiodothyronine and steroid hormone receptors. Trends Biochem Sci 6:268–271. 54. Yamamoto H, Shevde NK, Warrier A, Plum LA, DeLuca HF 2003 2-Methylene-19-nor-(20S)-1,25-dihydroxyvitamin D3 potently stimulates gene-specific DNA binding of the vitamin D receptor in osteoblasts. J Biol Chem 278: 31756–31765. 55. Kim S, Yamamoto H, Shevde NK, Pike JW 2004 1,25Dihydroxyvitamin D3 stimulates dynamic vitamin D receptor/ retinoid x receptor DNA-binding and coactivator recruitment in intact osteoblasts. J Bone Min Res 18(Suppl 2): 530 (Abst). 56. Racz A, Barsony J 1999 Hormone dependent translocation of vitamin D receptors is linked to transactivation. J Biol Chem 274:19352–19360. 57. Prufer K, Barsony J 2002 Retinoid X receptor dominates the nuclear import and export of the unliganded vitamin D receptor. Mol Endocrinol 16:1738–1751. 58. Pike JW, Haussler MR 1979 Purification of chicken intestinal receptor for 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 76:5488–5494. 59. Haussler MR, Pike JW, Chandler JS, Manolagas SC, Deftos LJ 1981 Molecular actions of 1,25-dihydroxyvitamin D3: New cultured cell models. Ann NY Acad Sci 372:502–517. 60. Katzenellenbogen JA, O’Malley BW, Katzenellenbogen BS 1996 Tripartite steroid hormone receptor pharmacology: interaction with multiple effector sites as a basis for the celland promoter-specific action of these hormones. Mol Endocrinol 6:119–131. 61. Mellon WS, DeLuca HF 1979 An equilibrium and kinetic study of 1,25-dihydroxyvitamin D3 binding to chicken intestinal cytosol employing high specific activity l,25-dihydroxy[3H-26,27] vitamin D3. Arch Biochem Biophys 197:90–95.
J. WESLEY PIKE AND NIRUPAMA K. SHEVDE
62. Wecksler WR, Norman AW 1980 A kinetic and equilibrium binding study of 1,25-dihydroxyvitamin D3 with its cytosol receptor from chick intestinal mucosa. J Biol Chem 255:3571–3574. 63. Kream BE, Jose JL, DeLuca HF 1977 The chick intestinal cytosol binding protein for 1,25-dihydroxyvitamin D3: A study of analog binding. Arch Biochem Biophys 179:462–468. 64. Wecksler WR, Okamura WH, Norman AW 1978 Quantitative assessment of the structural requirements for the interaction of 1,25-dihydroxyvitamin D3 with its chick intestinal mucosa receptor system. J Steroid Biochem 9:929–937. 65. Pike JW, Haussler MR 1983 Association of 1,25-dihydroxyvitamin D3 with cultured 3T6 mouse fibroblast. J Biol Chem 258:8554–8560. 66. Hunziker W, Walters MR, Bishop JE, Norman AW 1983 Unoccupied and in vitro and in vivo occupied 1,25-dihydroxyvitamin D3 intestinal receptors. Multiple biochemical forms and evidence for transformations. J Biol Chem 258:8642–8648. 67. Rochel N, Wurtz JM, Mitschler A, Klaholz B, Moras D 2000 The crystal structure of the nuclear receptor for vitamin D bound to its natural ligand. Mol Cell 5:173–179. 68. Tocchini-Valentini G, Roche N, Wurtz JM, Mitschler A, Moras D 2001 Crystal structures of the vitamin D receptor complexed to superagonist 20-epi ligands. Proc Natl Acad Sci USA 98:5491–5496. 69. Vanhooke JL, Benning MM, Bauer CB, Pike JW, DeLuca HF 2004 Molecular structure of the rat vitamin D receptor ligand binding domain complexed with 2-carbon-substituted vitamin D3 hormone analogs and an LxxLL-containing coactivator peptide. Biochemistry, 43:4101–4110. 70. Pike JW, Sleator N, Haussler MR 1987 Chicken intestinal receptor for 1,25-dihydroxyvitamin D3. Immunologic characterization and homogeneous isolation of a 60,000-dalton protein. J Biol Chem 262:1305–1311. 71. Mangelsdorf DJ, Pike JW, Haussler MR 1987 Avian and mammalian receptors for 1,25-dihydroxyvitamin D3: In vitro translation to characterize size and hormone-dependent regulation. Proc Natl Acad Sci USA 84:354–358. 72. Pike JW, Kesterson RA, Scott RA, Kerner SA, McDonnell DP, O’Malley BW 1988 Vitamin D receptors: Molecular structure of the protein and its chromosomal gene. In: Normal AW, Schaefer K, Grigoleit H-G, Herrath DV (eds) Vitamin D: Molecular, Cellular and Clinical Endocrinology, de Gruyter, Berlin, pp. 215–224. 73. Elaroussi MA, Prahl JM, DeLuca HF 1992 The avian vitamin D receptors: Primary structures and their origins. Proc Natl Acad Sci USA 91:11596–11600. 74. Pike JW, 1984 Monoclonal antibodies to chick intestinal receptors for 1,25-dihydroxyvitamin D3: Interaction and effects of binding on receptor function. J Biol Chem 259:1167–1173. 75. Allegretto EA, Pike JW 1985 Trypsin cleavage of chick 1,25dihydroxyvitamin D3 receptors. Generation of discrete polypeptides which retain hormone but are unreactive to DNA and monoclonal antibody. J Biol Chem 260: 10139–10145. 76. Allegretto EA, Pike JW, Haussler MR 1987 Immunological detection of unique proteolytic fragments of the chick 1,25-dihydroxyvitamin D3 receptor. Distinct 20-kDa DNA binding and 45-kDa hormone-binding species. J Biol Chem 262:1312–1319. 77. Allegretto EA, Pike JW, Haussler MR 1987 C-Terminal proteo-lysis of the avian 1,25-dihydroxyvitamin D3 receptor. Biochem Biophys Res Commun 147:479–485.
187
CHAPTER 11 The Vitamin D Receptor
78. Heyman RA, Mangelsdorf DJ, Dyck JA, Stein RB, Eichele G, Evans RM, Thaller C 1992 9-cis-Retinoic acid is a high affinity ligand for the retinoid X receptor. Cell 68:397–406. 79. Forman BM, Tontonoz P, Chen C, Brun RP, Speigelman BM, Evans RM 1995 15-Deoxy-∆12,14-prostaglandin J2 is a ligand for the adipocyte determination factor PPARy. Cell 83:803–812. 80. Forman BM, Goode E, Chen J, Oro AE, Bradley DJ, Perlman T, Noonan DJ, Burka LT, McMorris T, Lamph WW, Evans RM, Weinberger C 1995 Identification of a nuclear receptor that is activated by farnesol metabolites. Cell 81:687–693. 81. Janowski BA, Willy PJ, Devi TR, Falck JR, Mangelsdorf DJ 1996 An oxysterol signalling pathway mediated by the nuclear receptor LXR alpha. Nature 383:728–731. 82. Chawla A, Repa JJ, Evans RM, Mangelsdorf DJ 2001 Nuclear receptors and lipid physiology: opening the X-files. Science 294:1866–1870. 83. Hollenberg SM, Weinberger C, Ong ES, Cerelli G, Ono AE, Lebo R, Thompson EB, Rosenfeld MG, Evans RM 1985 Primary structure and expression of a functional human glucocorticoid receptor cDNA. Nature 318:635–641. 84. Weinberger C, Thompson CC, Ong ES, Lebo R, Gruol DJ, Evans RM 1986 The c-erbA gene encodes a thyroid hormone receptor. Nature 324:641–646. 85. Baker AR, McDonnell DP, Hughes MR, Crisp TM, Mangelsdorf DJ, Haussler MR, Shine J, Pike JW, O’Malley BW 1988 Molecular cloning and expression of human vitamin D receptor complementary DNA: Structural homology with thyroid hormone receptor. Proc Natl Acad Sci USA 85:3294–3298. 86. Burmester JK, Maeda N, DeLuca HF 1988 Isolation and expression of rat 1,25-dihydroxyvitamin D receptor cDNA. Proc Natl Acad Sci USA 85:1005–1009. 87. Goto H, Chen KS, Prahl JM, DeLuca HF 1992 A single receptor identical to that for intestinal T47D cells mediates the action of 1,25-dihydroxyvitamin D3 in HL-60 cells. Biochim Biophys Acta 1132:103–108. 88. Kamei Y, Kawada T, Fukuwatari T, Ono T, Kato S, Sugimoto E 1995 Cloning and sequence of the gene encoding the mouse vitamin D receptor. Gene 152:281–282. 89. Elaroussi MA, Prahl JM, DeLuca HF 1992 The avian vitamin D receptors: Primary structures and their origins. Proc Natl Acad Sci USA 91:11596–11600. 90. Li YC, Bergwitz C, Juppner H, Demay MB 1996 Cloning and characterization of the vitamin D receptor from Xenopus laevis. J Bone Miner Res 11(Suppl. 1):S161. 91. Ciana P, Vegeto E, Beato M, Chambon P, Gustafsson JA, Parker M, Wahli W, Maggi A 2002 Looking at nuclear receptors from the heights of Erice. Workshop on nuclear receptor structure and function. EMBO Rep 3:125–129. 92. Enmark E, Gustafsson JA 2001 Comparing nuclear receptors in worms, flies and humans. Trends Pharmacol Sci 22: 611–615. 93. McDonnell DP, Pike JW, O’Malley BW 1988 Vitamin D receptor: A primitive steroid receptor related to the thyroid hormone receptor. J Steroid Biochem 30:41–47. 94. Kesterson RA, Pike JW 1991 Unpublished. 95. Saijo T, Ito M, Takeda E, Huq AHM, Naito E, Yokoto I, Sone T, Pike JW, Kuroda Y 1991 A unique mutation in vitamin D receptor gene in three Japanese patients with vitamin D– dependent rickets type II: Utility of single-strand conformation polymorphism analysis for heterozygous carrier detection. Am J Hum Genet 49:668–673. 96. Jurutka PW, Remus LS, Whitfield GK, Thompson PD, Hsieh JC, Zitzer H, Tavakkoli P, Galligan MA, Dang HT, Haussler CA,
97.
98.
99.
100.
101.
102. 103.
104.
105. 106. 107. 108. 109. 110. 111.
112. 113.
Haussler MR 2000 The polymorphic N terminus in human vitamin D receptor isoforms influences transcriptional activity by modulating interaction with transcription factor IIB. Mol Endocrinol 14:401–420. Arai H, Miyamoto K, Taketani Y, Yamamoto H, Iemori Y, Morita K, Tonai T, Nishisho T, Mori S, Takeda E 1997 A vitamin D receptor gene polymorphism in the translation initiation codon: effect on protein activity and relation to bone mineral density in Japanese women. J Bone Miner Res 12:915–921. Gross C, Krishnan AV, Malloy PJ, Eccleshall TR, Zhao XY, Feldman D 1998 The vitamin D receptor gene start codon polymorphism: a functional analysis of FokI variants. J Bone Miner Res 13:1691–1699. Chiu KC, Chuang LM, Yoon C 2001 The vitamin D receptor polymorphism in the translation initiation codon is a risk factor for insulin resistance in glucose tolerant Caucasians. BMC Med Genet 2:2. Kubota M, Yoshida S, Ikeda M, Okada Y, Arai H, Miyamoto K, Takeda E 2001 Association between two types of vitamin D receptor gene polymorphism and bone status in premenopausal Japanese women. Calcif Tissue Int 68:16–22. Whitfield GK, Remus LS, Jurutka PW, Zitzer H, Oza AK, Dang HT, Haussler CA, Galligan MA, Thatcher ML, Encinas Dominguez C, Haussler MR 2001 Functionally relevant polymorphisms in the human nuclear vitamin D receptor gene. Mol Cell Endocrinol 177:145–159. Green S, Walter P, Kumar V, Krust A, Bornert JM, Argos P, Chambon P 1986 Human oestrogen receptor cDNA: Sequence expression and homology to v-erbA. Nature 320:134–139. McDonnell DP, Scott RA, Kerner RA, O’Malley BW, Pike JW 1989 Functional domains of the human vitamin D receptor regulate osteocalcin gene expression. Mol Endocrinol 3:635–644. Hughes MR, Malloy PJ, Kieback DG, Kesterson RA, Pike JW, Feldman D, O’Malley BW 1988 Point mutations in the human vitamin D receptor gene associated with hypocalcemic rickets. Science 242:1702–1705. Rochel N, Tocchini-Valentini G, Egea PF, Juntunen K, Garnier JM, Vihko P, Moras D 2001 Eur J Biochem 268:971–979. Jin CH, Pike JW 1996 Human vitamin D receptor dependent transactivation in Saccharomyces cerevisiae requires retinoid X receptor. Mol Endocrinol 10:196–205. Jin CH, Kerner SA, Hong MH, Pike JW 1996 Transcriptional activation and dimerization functions of the human vitamin D receptor. Mol Endocrinol 10:945–957. MacDonald PN, Sherman DR, Dowd DR, Jefcoat SC Jr, DeLisle RK 1995 The vitamin D receptor interacts with general transcription factor IIB. J Biol Chem 270:4748–4752. Berg JM 1989 DNA binding specificity and steroid receptors. Cell 57:1065–1068. Berg JM 1988 Proposed structure for the zinc-binding domains from transcription factor IIIA and related proteins. Proc Natl Acad Sci USA 85:99–102. Mader S, Kumar V, deVereneuil H, Chambon P 1989 Three amino acids of the oestrogen receptor are essential to its ability to distinguish an oestrogen from a glucocorticoid responsive receptor. Nature 338:271–274. Umesono K, Evans RM 1989 Determinants of target gene specificity for steroid/thyroid hormone receptors. Cell 57:1139–1146. Schwabe JWR, Neuhaus D, Rhodes D 1990 Solution structure of the DNA binding domain of the oestrogen receptor. Nature 348:458–461.
188 114. Hard T, Kellenbach E, Boelens R, Maler BA, Dahlman K, Freedman LP, Carlstedt-Duke J, Yamamoto KR, Gustafsson JA, Kaptein R 1990 Solution structure of the glucocorticoid receptor DNA binding domain. Science 249:157–160. 115. Luisi BF, Xu W, Otwinowski Z, Freedman LP, Yamamoto KR, Sigler PB 1991 Crystallographic analysis of the interaction of the glucocorticoid receptor with DNA. Nature 352: 497–505. 116. Schwabe JWR, Chapman L, Finch JT, Rhodes D 1993 The crystal structure of the oestrogen receptor DNA binding domain bound to DNA: How receptors discriminate between their response elements. Cell 75:567–578. 117. Lee MS, Kliewer SA, Provencal J, Wright PE, Evans RM 1993 Structure of the retinoid X receptor α DNA binding domain: A helix required for homodimeric DNA binding. Science 260:1117–1121. 118. Rastinejad F, Perlmann T, Evans RM, Sigler PB 1995 Structural determinants of nuclear receptor assembly on DNA direct repeats. Nature 375:203–211. 119. Forman BM, Yang CR, Au M, Casanova J, Ghysdael J, Samuels HH 1989 A domain containing leucine-zipper-like motifs mediate novel in vivo interactions between the thyroid hormone and retinoic acid receptors. Mol Endocrinol 3:1610–1626. 120. Bourguet W, Ruff D, Chambon P, Gronemeyer H, Moras D 1995 Crystal structure of the ligand binding domain of the human nuclear receptor RXRa. Nature 375:377–382. 121. Renaud JP, Natacha R, Ruff M, Vivat V, Chambon P, Gronemyer H, Moras D 1995 Crystal structure of the RARy ligand binding domain bound to all-trans-retinoic acid. Nature 378:681–689. 122. Wagner RL, Apriletti JW, McGrath ME, West BL, Baxter JD, Fletterick RJ 1995 A structural role for hormone in the thyroid hormone receptor. Nature 378:690–697. 123. Shiau AK, Barstad D, Loria PM, Cheng L, Kushner PJ, Agard DA, Greene GL 1998 The structural basis of estrogen receptor/coactivator recognition and the antagonism of this interaction by tamoxifen. Cell 95:927–937. 124. Williams S, Sigler P 1998 Atomic structure of progesterone complexed with its receptor. Nature 393:392–396. 125. Nolte RT, Wisely GB, Westin S, Cobb JE, Lambert MH, Kurokawa R, Rosenfeld MG, Willson TM, Glass CK Milburn MV 1998 Ligand binding and co-activator assembly of the peroxisome proliferator-activated receptor-gamma. Nature 395:137–143. 126. Downes M, Verdecia MA, Roecker AJ, Hughes R, Hogenesch JB, Kast-Woelbern HR, Bowman ME, Ferrer JL, Anisfeld AM, Edwards PA, Rosenfeld JM, Alvarez JG, Noel JP, Nicolaou KC, Evans RM 2003 A chemical, genetic, and structural analysis of the nuclear bile acid receptor FXR. Mol Cell 11:1079–1092. 127. Peleg S, Sastry M, Collins ED, Bishop JE, Norman AW 1995 Distinct conformational changes induced by 20-epi analogues of 1,25-dihydroxyvitamin D3 are associated with enhanced activation of the vitamin D receptor. J Biol Chem 270: 10551–10558. 128. Quack M, Carlberg C 1999 Selective recognition of vitamin D receptor conformations mediates promoter selectivity of vitamin D analogs. Mol Pharmacol 55:1077–1087. 129. Moras D 2003 Twelfth Workshop on Vitamin D Maastricht, The Netherlands, June 2003, p. 3 (Abstract). 130. Whitfield GK, Hsieh JC, Nakajima S, MacDonald PN, Thompson PD, Jurutka PW, Haussler CA, Haussler MR 1995 A highly conserved region in the hormone-binding domain
J. WESLEY PIKE AND NIRUPAMA K. SHEVDE
131.
132.
133. 134.
135.
136.
137.
138.
139.
140.
141. 142. 143. 144.
145.
of the human vitamin D receptor contains residues vital for heterodimerization with retinoid X receptor and for transcriptional activation. Mol Endocrinol 9:1166–1179. Danielian PS, White R, Lees JA, Parker MG 1992 Identification of a conserved region required for hormone dependent transcriptional activation by steroid hormone receptors. EMBO J 11:1025–1033. Xu J, O’Malley BW 2002 Molecular mechanisms and cellular biology of the steroid receptor coactivator (SRC) family in steroid receptor function. Rev Endocr Metab Disord 3:185–192. Rachez C, Freedman LP 2001 Mediator complexes and transcription. Curr Opin Cell Biol 13:274–280. McInerney EM, Rose DW, Flynn SE, Westin S, Mullen TM, Krones A, Inostroza J, Torchia J, Nolte RT, Assa-Munt N, Milburn MV, Glass CK, Rosenfeld MG 1998 Determinants of coactivator LXXLL motif specificity in nuclear receptor transcriptional activation. Genes Dev 12:3357–3368. Pathrose P, Barmina O, Chang CY, McDonnell DP, Shevde NK, Pike JW 2002 Inhibition of 1,25-dihydroxyvitamin D3dependent transcription by synthetic LXXLL peptide antagonists that target the activation domains of the vitamin D and retinoid X receptors. J Bone Miner Res 17:2196–2205. [Erratum in 2003 J Bone Miner Res 18:390.] Brzozowski AM, Pike AC, Dauter Z, Hubbard RE, Bonn T, Engstrom O, Ohman L, Greene GL, Gustafsson JA, Carlquist M 1997 Molecular basis of agonism and antagonism in the oestrogen receptor. Nature 389:753–758. Shiau AK, Barstad D, Radek JT, Meyers MJ, Nettles KW, Katzenellenbogen BS, Katzenellenbogen JA, Agard DA, Greene G 2002 Structural characterization of a subtypeselective ligand reveals a novel mode of estrogen receptor antagonism. Nature Struc Biol 9:359–364. Tzukerman MT, Esty A, Santiso-Mere D, Danielian P, Parker MG, Stein RB, Pike JW, McDonnell DP 1994 Human estrogen receptor transcriptional capacity is determined by both cellular and promoter context and mediated by two functionally distinct intramolecular regions. Mol Endocrinol 8:21–30. McDonnell DP, Clemm DL, Hermann T, Goldman ME, Pike JW 1995 Analysis of estrogen receptor function in vitro reveals three distinct classes of anti-estrogens. Mol Endocrinol 9:659–669. Ducy P, Desbois C, Boyce B, Pinero G, Story B, Dunstan C, Smith E, Bonadio J, Goldstein S, Gundberg C, Bradley A, Karsenty G 1996 Increased bone formation in osteocalcindeficient mice. Nature 382:448–452. Price PA, Baukol SA 1980 1,25-Dihydroxyvitamin D3 increases synthesis of the vitamin D-dependent bone proteins by osteosarcoma cells. J Biol Chem 225:11660–11663. Pan LC, Price PA 1986 1,25-Dihydroxyvitamin D3 stimulates transcription of bone gla protein. J Bone Miner Res 1(Suppl. 1):A20. Celeste AJ, Rosen V, Buecker JL, Kriz R, Wang EA, Wozney JM 1986 Isolation of the human gene for bone gla protein utilizing mouse and rat cDNA clones. EMBO J 5:1885–1890. Kerner SA, Scott RA, Pike JW 1989 Sequence elements in the human osteocalcin gene confer basal activation and inducible response to hormonal vitamin D3. Proc Natl Acad Sci USA 86:4455–4459. Morrison NA, Shine J, Fragonas J-C, Verkest V, McMenemy ML, Eisman JA 1989 1,25-Dihydroxyvitamin D3 responsive element and glucocorticoid repression in the osteocalcin gene. Science 246:1158–1161.
CHAPTER 11 The Vitamin D Receptor
146. Yoon K, Rutledge SJC, Buenaga RF, Rodan GA 1988 Characterization of the rat osteocalcin gene: Stimulation of promoter activity by 1,25-dihydroxyvitamin D3. Biochemistry 27:8521–8526. 147. Ozono K, Liao J, Scott RA, Kerner SA, Pike JW 1990 The vitamin D responsive element in the human osteocalcin gene: Association with a nuclear proto-oncogene enhancer. J Biol Chem 265:21881–21888. 148. Demay MB, Gerardi JM, DeLuca HF, Kronenberg HM 1990 DNA sequences in the rat osteocalcin gene that bind the 1,25-dihydroxyvitamin D receptor and confer responsiveness to 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 87:369–373. 149. Lian JB, Stewart C, Puchacz E, Mackowiak S, Shalhoub V, Colart D, Sambetti G, Stein G 1989 Structure of the rat osteocalcin gene and regulation of vitamin D–dependent expression. Proc Natl Acad Sci USA 86:1143–1147. 150. Terpening CM, Haussler MR, Jurutka PW, Galligan MA, Komm BS, Haussler MR 1991 The vitamin D responsive element in the rat bone gla protein gene is an imperfect direct repeat that cooperates with other cis elements in 1,25dihydroxyvitamin D3 mediated transcriptional activation. Mol Endocrinol 5:373–385. 151. Noda M, Vogel RL, Craig AM, Prahl J, DeLuca HF, Denhardt D 1990 Identification of a DNA sequence responsible for binding of the 1,25-dihydroxyvitamin D3 receptor and 1,25-dihydroxyvitamin D3 enhancement of mouse secreted phosphoprotein 1 (Supp-1 or osteopontin) gene expression. Proc Natl Acad Sci USA 87:9995–9999. 152. Gill RK, Christakos S 1993 Identification of sequence elements in the mouse calbindin-D28K gene that confer 1,25-dihydroxyvitamin D- and butyrate inducible responses. Proc Natl Acad Sci USA 90:2984–2988. 153. Darwish HM, DeLuca HF 1992 Identification of a 1,25-dihydroxyvitamin D3 response element in the 5’ flanking region of the rat calbindin D-9K gene. Proc Natl Acad Sci USA 89:603–607. 154. Ohyama Y, Ozono K, Uchida M, Shinki T, Kato S, Suda T, Yamamoto O, Noshiro M, Kato Y 1994 Identification of a vitamin D responsive element in the 5′-flanking region of the rat 25-hydroxyvitamin D3 24-hydroxylase gene. J Biol Chem 269:10545–10550. 155. Zierold C, Darwish HM, DeLuca HF 1995 Two vitamin D response elements function in the rat 1,25-dihydroxyvitamin D3–24-hydroxylase promoter. J Biol Chem 269: 1675–1678. 156. Chen KS, DeLuca HF 1995 Cloning of the human 1,25dihydroxyvitamin D3–24-hydroxylase gene promote and identification of two vitamin D responsive elements. Biochim Biophys Acta 1263:1–9. 157. Liu M, Lee M-H, Cohen M, Freedman LP 1996 Transcriptional activation of the p21 gene by vitamin D3 leads to the differentiation of the myelomonocytic cell line U937. Genes Dev 10:142–153. 158. van den Bemd GJ, Jhamai M, Staal A, van Wijnen AJ, Lian JB, Stein GS, Pols HA, van Leeuwen JP 2002 A central dinucleotide within vitamin D response elements modulates DNA binding and transactivation by the vitamin D receptor in cellular response to natural and synthetic ligands. J Biol Chem 277:14539–14546. 159. Umesono K, Murikami KK, Thompson CC, Evans RM 1991 Direct repeats as selective response elements for the thyroid hormone, retinoic acid, and vitamin D3 receptors. Cell 65:1225–1266.
189 160. Sierra J, Villagra A, Paredes R, Cruzat F, Gutierrez S, Javed A, Arriagada G, Olate J, Imschenetzky M, Van Wijnen AJ, Lian JB, Stein GS, Stein JL, Montecino M 2003 Regulation of the bone-specific osteocalcin gene by p300 requires Runx2/Cbfa1 and the vitamin D3 receptor but not p300 intrinsic histone acetyltransferase activity. Mol Cell Biol 23:3339–3351. 161. Zhang C, Dowd DR, Staal A, Gu C, Lian JB, van Wijnen AJ, Stein GS, MacDonald PN 2003 Nuclear coactivator-62 kDa/ Ski-interacting protein is a nuclear matrix-associated coactivator that may couple vitamin D receptor-mediated transcription and RNA splicing. J Biol Chem 278:35325–35336. 162. Sone T, Kerner SA, Pike JW 1991 Vitamin D receptor interaction with specific DNA. Association as a 1,25-dihydroxyvitamin D3-modulated heterodimer. J Biol Chem 266: 23296–23305. 163. Freedman LP, Towers T 1991 DNA binding properties of the vitamin D receptor zinc fingers region. Mol Endocrinol 5:1815–1826. 164. Towers T, Luisi BL, Asianov A, Freedman LP 1993 DNA target selectivity by the vitamin D receptor: Mechanism for dimer binding to an asymmetric repeat element. Proc Natl Acad Sci USA 90:6310–6314. 165. Lemon B, Freedman LP 1996 Selective effects of ligands on vitamin D3 receptor and retinoid X receptor mediated gene activation in vivo. Mol Cell Biol 16:1006–1016. 166. Nishikawa J, Kitaura M, Matsumoto M, Imagawa M, Nishihara T 1994 Difference and similarity of DNA sequence recognized by VDR homodimer and VDR/RXR heterodimer. Nucleic Acids Res 22:2902–2907. 167. Sone T, Scott R, Hughes M, Malloy P, Feldman D, O’Malley BW, Pike JW 1989 Mutant vitamin D receptors which confer hereditary resistance to 1,25-dihydroxyvitamin D3 in humans are transcriptionally inactive in vitro. J Biol Chem 264: 20230–20234. 168. Hughes MR, Malloy PJ, O’Malley BW, Pike JW, Feldman D 1991 Genetic defects of the 1,25-dihydroxyvitamin D receptor. J Receptor Res 11:699–716. 169. Nakajima S, Hsieh JC, MacDonald PN, Galligan MA, Haussler CA, Whitfield GK, Haussler MR 1994 The C-terminal region of the vitamin D receptor is essential to form a complex with a receptor auxiliary factor required for high affinity binding to the vitamin D responsive element. Mol Endocrinol 8:159–172. 170. Liao J, Ozono K, Sone T, McDonnell DP, Pike JW 1990 Vitamin D receptor interaction with specific DNA requires a nuclear protein and 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 87:9751–9755. 171. Sone T, McDonnell DP, O’Malley BW, Pike JW 1990 Expression of the human vitamin D receptor in Saccharomyces cerevisiae: Purification properties and generation of polyclonal antibodies. J Biol Chem 265:21997–22003. 172. Sone T, Ozono K, Pike JW 1991 A 55-kilodalton accessory factor facilitates vitamin D receptor DNA binding. Mol Endocrinol 5:1578–1586. 173. MacDonald PN, Haussler CA, Terpening CM, Galligan MA, Reeder MC, Whitfield GK, Haussler MR 1991 Baculovirusmediated expression of the human vitamin D receptor: Functional characterization, vitamin D response element interactions and evidence for a receptor auxiliary factor. J Biol Chem 266:18808–18813. 174. Ross TK, Moss VE, Prahl JM, DeLuca HF 1992 A nuclear protein essential for binding of rat 1,25-dihydroxyvitamin D3 receptor to its response elements. Proc Natl Acad Sci USA 89:256–260.
190 175. Yu V, Delsert C, Andersen B, Holloway JM, Devary OV, Naar AM, Kirn SY, Boutin JM, Glass CK, Rosenfeld MG 1991 RXR/3: A coregulator that enhances binding of retinoic acid, thyroid hormone, and vitamin D receptors to their cognate response elements. Cell 67:1251–1266. 176. Leid M, Kastner P, Lyons R, Nakshatro H, Saunders M, Zacharewski T, Chen J-Y, Staub A, Gamier J-M, Mader S, Chambon P 1992 Purification, cloning and RXR identity of the HeLa cell factor with which RAR or TR heterodimerizes to bind target sequences efficiently. Cell 68:377–395. 177. Zhang XK, Hoffman B, Tran PB, Graupner G, Pfahl M 1992 Retinoid X receptor is an auxiliary protein for thyroid hormone and retinoic acid receptors. Nature 355:441–446. 178. Kliewer SA, Umesono K, Mangelsdorf DJ, Evans RM 1992 Retinoid X receptor interacts with nuclear receptors in retinoic acid, thyroid, and vitamin D signaling. Nature 355:446–449. 179. Perlmann T, Umesono K, Rangarajan PN, Forman BM, Evans RM 1996 Two distinct dimerization interfaces differentially modulate target gene specificity of nuclear hormone receptors. Mol Endocrinol 10:958–966. 180. Perlmann T, Rangarajan PN, Umesono K, Evans RM 1993 Determinants for selective RAR and TR recognition of direct repeat HREs. Genes Dev 7:1411–1422. 181. Freedman LP, Arce V, Perez Fernandez R 1994 DNA sequences that act as high affinity targets for the vitamin D3 receptor in the absence of the retinoid X receptor. Mol Endocrinol 8:265–273. 182. Alroy I, Towers TL, Freedman LP 1995 Transcriptional repression of the interleukin-2 gene by vitamin D3: direct inhibition of NFATp/AP-1 complex formation by a nuclear hormone receptor. Mol Cell Biol 15:5789–5799. 183. Towers TL, Staeva TP, Freedman LP 1999 A two-hit mechanism for vitamin D3-mediated transcriptional repression of the granulocyte-macrophage colony-stimulating factor gene: vitamin D receptor competes for DNA binding with NFAT1 and stabilizes c-Jun. Mol Cell Biol 19:4191–4109. 184. Staeva-Vieira TP, Freedman LP 2002 1,25-Dihydroxyvitamin D3 inhibits IFN-gamma and IL-4 levels during in vitro polarization of primary murine CD4+ T cells. J Immunol 168: 1181–1189. 185. Mackey SL, Heymont JL, Kronenberg HM, Demay MB 1996 Vitamin D receptor binding to the negative human parathyroid hormone vitamin D response element does not require the retinoid x receptor. Mol Endocrinol 10:298–305. 186. Prufer K, Racz A, Lin GC, Barsony J 2000 Dimerization with retinoid X receptors promotes nuclear localization and subnuclear targeting of vitamin D receptors. J Biol Chem 275:41114–41123. 187. Kurokawa B, Yu VC, Naar A, Kyakumoto S, Han Z, Silverman S, Rosenfeld MG, Glass CK 1993 Differential orientation of the DNA-binding domain and carboxy terminal dimerization interface regulate binding site selection by nuclear receptor heterodimers. Genes Dev 7:1423–1435. 188. Sucov HM, Murakami KK, Evans RM 1990 Characterization of an autoregulated response element in the mouse retinoic acid receptor type beta gene. Proc Natl Acad Sci USA 87: 5392–5396. 189. Fyodorov DV, Kadonaga JT 2001 The many faces of chromatin remodeling: SWItching beyond transcription. Cell 106:523–525. 190. Narlikar GJ, Fan HY, Kingston RE 2002 Cooperation between complexes that regulate chromatin structure and transcription. Cell 108:475–487. 191. Chakravarti D, LaMorte VJ, Nelson MC, Nakajima T, Schulman IG, Juguilon H, Montminy M, Evans RM 1996
J. WESLEY PIKE AND NIRUPAMA K. SHEVDE
192. 193. 194.
195.
196.
197.
198.
199. 200.
201.
202.
203.
204.
205.
206.
Role of CBP/P300 in nuclear receptor signalling. Nature 383:99–103. Ogryzko VV, Schiltz RL, Russanova V, Howard BH, Nakatani Y 1996 The transcriptional coactivators p300 and CBP are histone acetyltransferases. Cell 87:953–959. Imhof A, Yang XJ, Ogryzko VV, Nakatani Y, Wolffe AP, Ge H 1997 Acetylation of general transcription factors by histone acetyltransferases. Curr Biol 7:689–692. Spencer TE, Jenster G, Burcin MM, Allis CD, Zhou J, Mizzen CA, McKenna NJ, Onate SA, Tsai SY, Tsai MJ, O’Malley BW 1997 Steroid receptor coactivator-1 is a histone acetyltransferase. Nature 389:194–198. Jiang YW, Veschambre P, Erdjument-Bromage H, Tempst P, Conaway JW, Conaway RC, Kornberg RD 1998 Mammalian mediator of transcriptional regulation and its possible role as an end-point of signal transduction pathways. Proc Natl Acad Sci USA 95:8538–8543. Rachez C, Lemon BD, Suldan Z, Bromleigh V, Gamble M, Naar AM, Erdjument-Bromage H, Tempst P, Freedman LP 1999 Ligand-dependent transcription activation by nuclear receptors requires the DRIP complex. Nature 398:824–828. Masuyama H, Brownfield CM, St-Arnaud R, MacDonald PN 1997 Evidence for ligand-dependent intramolecular folding of the AF-2 domain in vitamin D receptor-activated transcription and coactivator interaction. Mol Endocrinol 11:1507–1517. Kraichely DM, Collins JJ 3rd, DeLisle RK, MacDonald PN 1999 The autonomous transactivation domain in helix H3 of the vitamin D receptor is required for transactivation and coactivator interaction. J Biol Chem 274:14352–14358. Leo C, Li H, Chen JD 2000 Differential mechanisms of nuclear receptor regulation by receptor-associated coactivator 3. J Biol Chem 275:5976–5982. Issa LL, Leong GM, Barry JB, Sutherland RL, Eisman JA 2001 Glucocorticoid receptor-interacting protein-1 and receptor-associated coactivator-3 differentially interact with the vitamin D receptor (VDR) and regulate VDR-retinoid X receptor transcriptional cross-talk. Endocrinology 142:1606–1615. Baudino TA, Kraichely DM, Jefcoat SC Jr, Winchester SK, Partridge NC, MacDonald PN 1998 Isolation and characterization of a novel coactivator protein, NCoA-62, involved in vitamin D–mediated transcription. J Biol Chem 273: 16434–16441. Kitagawa H, Fujiki R, Yoshimura K, Mezaki Y, Uematsu Y, Matsui D, Ogawa S, Unno K, Okubo M, Tokita A, Nakagawa T, Ito T, Ishimi Y, Nagasawa H, Matsumoto T, Yanagisawa J, Kato S, 2003 The chromatin-remodeling complex WINAC targets a nuclear receptor to promoters and is impaired in Williams syndrome. Cell 113:905–917. Polly P, Herdick M, Moehren U, Baniahmad A, Heinzel T, Carlberg C 2000 VDR-Alien: a novel, DNA-selective vitamin D3 receptor-corepressor partnership. FASEB J 14:1455–1463. Hsieh JC, Sisk JM, Jurutka PW, Haussler CA, Slater SA, Haussler MR, Thompson CC 2003 Physical and functional interaction between the vitamin D receptor and hairless corepressor, two proteins required for hair cycling. J Biol Chem 278:38665–38674. Masuyama H, MacDonald PN 1998 Proteasome-mediated degradation of the vitamin D receptor (VDR) and a putative role for SUG1 interaction with the AF-2 domain of VDR. J Cell Biochem 71:429–440. Perissi V, Staszewski LM, McInerney EM, Kurokawa R, Krones A, Rose DW, Lambert MH, Milburn MV, Glass CK,
CHAPTER 11 The Vitamin D Receptor
207.
208.
209.
210.
211. 212.
Rosenfeld MG 1999 Molecular determinants of nuclear receptor-corepressor interaction. Genes Dev 13:3198–3208. Tagami T, Lutz WH, Kumar R, Jameson JL 1998 The interaction of the vitamin D receptor with nuclear receptor corepressors and coactivators. Biochem Biophys Res Commun 253:358–363. Blanco JCG, Wang I-M, Tsai SY, Tsai M-J, O’Malley BW, Jurutka PW, Haussler MR, Ozato K 1995 Transcription factor TFIIB and the vitamin D receptor cooperatively active ligand-dependent transcription. Proc Natl Acad Sci USA 92:1535–1539. Thompson PD, Remus LS, Hsieh JC, Jurutka PW, Whitfield GK, Galligan MA, Encinas Dominguez C, Haussler CA, Haussler MR 2001 Distinct retinoid X receptor activation function-2 residues mediate transactivation in homodimeric and vitamin D receptor heterodimeric contexts. J Mol Endocrinol 27:211–227. Bettoun DJ, Burris TP, Houck KA, Buck DW 2nd, Stayrook KR, Khalifa B, Lu J, Chin WW, Nagpal S 2003 Retinoid X receptor is a nonsilent major contributor to vitamin D receptor-mediated transcriptional activation. Mol Endocrinol 17:2320–2328. Shang Y, Hu X, DiRenzo J, Lazar MA, Brown M 2000 Cofactor dynamics and sufficiency in estrogen receptorregulated transcription. Cell 103:843–852. Reid G, Hubner MR, Metivier R, Brand H, Denger S, Manu D, Beaudouin J, Ellenberg J, Gannon F 2003 Cyclic, proteasomemediated turnover of unliganded and liganded ERalpha on responsive promoters is an integral feature of estrogen signaling. Mol Cell 11:695–707.
191 213. Shang Y, Myers M, Brown M 2003 Formation of the androgen receptor transcription complex. Mol Cell 9:601–610. 214. Burakov D, Crofts LA, Chang CP, Freedman LP 2002 Reciprocal recruitment of DRIP/mediator and p160 coactivator complexes in vivo by estrogen receptor. J Biol Chem 277:14359–14362. 215. Szpirer S, Szpirer C, Riviere M, Levan G, Maryen P, Cassiman JJ, Wiese R, DeLuca HF 1991 The SP1 transcription factor gene (SP1) and the 1,25-dihydroxyvitamin D receptor gene (VDR) are colocalized on human chromosomal arm 12q and rat chromosome 7. Genomics 11: 168–173. 216. Miyamoto K, Kesterson RA, Yamamoto H, Taketani Y, Nishiwaki E, Tatsumi S, Inoue Y, Morita K, Takeda E, Pike JW 1997 Structural organization of the human vitamin D receptor chromosomal gene and its promoter. Mol Endocrinol 11:1165–1179. 217. Jehan F, DeLuca HF 2000 The mouse vitamin D receptor is mainly expressed through an Sp1-driven promoter in vivo. Arch Biochem Biophys 377:273–283. 218. Crofts LA, Hancock MS, Morrison NA, Eisman JA 1998 Multiple promoters direct the tissue-specific expression of novel N-terminal variant human vitamin D receptor gene transcripts. Proc Natl Acad Sci USA 95:10529–10534. 219. Sunn KL, Cock TA, Crofts LA, Eisman JA, Gardiner EM 2001 Novel N-terminal variant of human VDR. Mol Endocrinol 15:1599–1609. 220. Morrison NA, Qi JC, Tokita A, Kelly PJ, Crofts L, Nguyen TV, Sambrook PN, Eisman JA 1994 Prediction of bone density from vitamin D receptor alleles. Nature 367:284–287.
CHAPTER 12
Vitamin D Receptor Promoter and Regulation of Receptor Expression LUIS M. ESTEBAN, JOHN A. EISMAN, AND EDITH M. GARDINER Bone and Mineral Research Program, Garvan Institute of Medical Research, Sydney, Australia
I. Introduction II. Tissue Distribution III. The Vitamin D Receptor Gene Locus
IV. Regulation of Vitamin D Receptor Expression and Abundance V. Concluding Remarks References
I. INTRODUCTION
Current understanding of this regulation is often confounded by inconsistencies between findings in different experimental systems, which may relate to the particular species, tissue, or cell type represented by the model under study. This chapter therefore describes current understanding of the complexities of VDR gene regulation, with the goal of providing a foundation for interpretation of recent findings and a basis for further exploration.
The cellular and physiological actions of 1,25dihydroxyvitamin D [1,25(OH)2D3] are mediated primarily through the nuclear vitamin D receptor (VDR) as described in Chapter 11, although there is in vitro evidence for rapid actions of the hormone that appear to be mediated by a distinct nonnuclear receptor, as described in Chapter 23. Alteration in circulating levels of 1,25(OH)2D3 is a primary mechanism by which receptorexpressing cells are subjected to physiological hormone responses, as described in Chapters 46, 49, 65, and 66. Another important regulatory mechanism, however, is variation in cellular content of functional receptor, by alteration in VDR abundance or by post-translational modification. Such regulation allows multiple target cells to have graded responses to 1,25(OH)2D3 in a given hormonal environment, depending on the VDR levels. This chapter addresses the transcriptional mechanisms by which VDR abundance is regulated. Regulation of receptor function by post-translational mechanisms is discussed in Chapter 13. The nuclear VDR gene structure is complex, with several functioning promoters and alternatively spliced 5′ coding and noncoding exons in some species. There are consensus transcription factor binding elements and response motifs in the promoter regions. In some cases, there is experimental evidence that these motifs are functional and respond to known regulators of VDR transcription. However, there are also consensus element matches that suggest previously unsuspected regulatory inputs and await experimental investigation. Moreover, in some cases the promoter regions lack consensus response elements relevant to signal transduction pathways that are known to regulate VDR expression, suggesting indirect mechanisms of regulation. Study of the transcriptional control of the VDR gene therefore provides both expected and unexpected insights into the physiological regulation and function of this receptor. VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
II. TISSUE DISTRIBUTION Specific receptors for 1,25(OH)2D3 were initially identified in the intestine of the chick by sucrose density gradient separation and assays of binding to [3H]-1,25(OH)2D3 [1,2]. Such receptors were subsequently identified in rat and human intestine, rat and chick fetal calvaria, chick kidney and pancreas, and rat skin [3–7], with sedimentation coefficients of 3.2–3.7 S. This experimental approach failed to detect receptors in brain, lung, thymus, adipose tissue, skeletal muscle, uterus, and myocardium, leading to the initial impression that there was a selective organ distribution. In addition to these studies, which detected VDR primarily in calciotropic tissues, early reports also indicated the presence of a receptor with physicochemical properties similar to those of chicken intestine receptor in the breast cancer cell line MCF-7 [8] and in a number (23 of 33) of cultured human cancer cell lines [9], including seven of 10 malignant melanoma lines and eight of nine colonic carcinoma lines. In the 1980s monoclonal antibodies generated against VDR protein isolated from chicken and pig intestine, or against synthetic peptides from the human VDR [10–12], were used to detect VDR in protein extracts from a variety of tissues and cell lines, and to clone the chicken VDR cDNA [13,14]. Copyright © 2005, Elsevier, Inc. All rights reserved.
194 A monoclonal antibody to the chick intestinal receptor detected nuclear VDR in histological sections of normal human epithelial tissues, including liver, kidney, thyroid, adrenal, gastrointestinal tract, breast, and skin [15]. A quantitative immunoradiometric assay (IRMA) indicated the presence of VDR protein in rat tissues [16], with the highest level in proximal small intestine and colon, moderate levels in ileum and kidney, and lower amounts in tissues of the vitamin D endocrine system, e.g., bone, thyroid/parathyroid, and skin. The IRMA proved more sensitive than the earlier sedimentation and radioligand approach mentioned above, revealing even lower amounts of VDR protein in the nonclassical vitamin D target tissues lung, heart, stomach, spleen, and liver, but not in cerebrum, cerebellum, or skeletal muscle [16]. More recent studies have detected VDR in pituitary, activated B and T lymphocytes, ovary, and testes, as well as in brain [17]. The availability of even more sensitive techniques such as reverse-transcription polymerase chain reaction (RT-PCR) has made it clear that VDR is expressed even in liver, which previously was considered to have very low or even no VDR [18]. VDR mRNA and protein were very low in hepatocytes, but Kupffer, stellate, and endothelial cells expressed functional VDR that mediated a positive CYP24 promoter response to 1,25(OH)2D3, suggesting that selective hepatic cell populations are also targets for the vitamin D endocrine system [18]. Based on these findings, it appears that there are more than 50 vitamin D target tissues, with the actions of 1,25(OH)2D3 extending far beyond its role in calcium homeostasis. The levels of VDR vary greatly between tissues and even between different cell types within a given tissue. Regulation of VDR expression therefore plays a fundamental role in determining physiological vitamin D responsiveness.
III. THE VITAMIN D RECEPTOR GENE LOCUS The many biological actions of vitamin D and its derivatives and the varied distribution of its receptor indicate complex regulation, which includes precise control of VDR and its abundance in a tissue- and species-specific manner. Part one of this section is focussed on the structure of the 5′ end of the VDR gene in chicken and mammals, including human, mouse, and rat, with detailed descriptions of VDR N-terminal variants that may provide functional diversity to the vitamin D endocrine system. The second part of this section presents evidence for the complex regulation of the VDR through its promoter structure and the presence of numerous transcription factor response elements.
LUIS M. ESTEBAN, JOHN A. EISMAN, AND EDITH M. GARDINER
The third part briefly reviews evidence related to the effects of common DNA sequence polymorphisms on VDR expression.
A. VDR Gene Structure 1. STRUCTURE OF THE VDR LOCUS
VDR is widely but not ubiquitously distributed. Its transcription has been analyzed primarily in mammalian and avian species, although it is present in all vertebrates and to date cDNAs have been isolated from more than 18 species including fish, amphibians, and reptiles. The human and rat mRNAs are 4.4 kb [19,20], in contrast with the approximately 2.7 kb avian [21,22] and the two Xenopus laevis (2.2 and 1.8 kb) transcripts [23]. However the sizes of the encoded receptor proteins are somewhat comparable across species, at 422 amino acid residues in Xenopus, 424 residues in rat, 427 and 477 residues in human, and 437 and 451 residues in chicken. There are eight coding exons (exons 2 to 9 according to human VDR notation [24] and a variable number of untranslated 5′ exons. The functional start codon of the mouse locus is in exon 3 [25], which is equivalent to human and rat exon 2 and chicken exon 4 (Fig. 1). Notably, in the chicken locus, translation is initiated not in exon 4, but instead in exon 3 at one of two initiation codons, giving rise to two isoforms that differ by 14 amino acids [22]. There are two known untranslated exons upstream of the first coding exons in chicken, mouse, and rat, which are highly conserved at the nucleotide level among mammalian but not chicken loci. The human VDR locus is more complex, with six transcribed exons upstream of the initiation codon in exon 2. Alternative promoter usage and alternative splicing of exons 1A, 1B, and 1C [24] and 1D, 1E, and 1F [27] give rise to 14 different human VDR transcripts [27]. Exons 1A and 1C are equivalent to mouse exons 1 and 2 [25]. Exon 1D is of particular interest, as it contains a translation initiation codon that is in frame with the conventional human VDR open reading frame [27]. An alternatively spliced transcript has been detected that includes exons 1D and 1C and encodes an N-terminally extended receptor isoform, VDRB1, which is present in human tissues and cell lines [28] (discussed further hereafter). A second exon 1Dcontaining transcript lacking exon 1C and encoding a shorter isoform VDRB2 has been detected, but to date the existence of this putative receptor has not been confirmed. Exon 1D–containing transcripts have not been reported in other species, although the DNA sequence is evolutionarily conserved in mouse and rat [29], with
195
CHAPTER 12 VDR Promoter and Abundance
ATG
ATG hVDRB1
1A 1F
1E
1D
1B
1C
2
Human hVDR
ATG 1A
1D-like
1C
2
Rat
ATG 1
1D-like
2
3
Mouse
ATG 1
2
3
4
Chicken
FIGURE 1
Structure of the 5′ flanking region of mammalian and chicken VDR gene loci. Arrows indicate exons with the initiation codon; two start codons are present in chicken exon 3. Alternative splicing produces a number of transcripts in the human VDR; two of them, which give rise to the original VDR and the VDRB1 isoform, are shown. Dotted line on the chicken locus indicates sequence not yet determined.
an overall homology of 75% with the human sequence. Chicken exon 2 has low homology with human exon 1D (50%) and its position relative to exon 1 is not conserved compared to the relative positions of human exons 1A and 1D (Fig. 1). Primer extension analysis of mouse VDR transcripts did give rise to products with a different sequence 5′ to exon 2, perhaps originating from an upstream promoter or a differentially spliced transcript [30]. To date, there is no evidence from homology searches of the mouse and rat genome databases that the remaining human exons (1B, 1E, and 1F) are present in rodents. However, they do appear to be present in other primates, as genomic sequence from chimpanzee (Pan troglodytes) includes all the exons with a 98–100% identity to human sequences [29]. The evolutionary conservation of these exons and surrounding sequences suggests an important biological role for the N-terminal extension (whether for modulation of VDR expression or function). Future studies on other mammals may reveal whether these isoforms are a general phenomenon or limited to species evolutionarily closer to humans. 2. N-TERMINAL NUCLEAR RECEPTOR VARIANTS
The VDR is a member of the nuclear receptor (NR) superfamily of ligand-inducible transcription factors. It has a modular structure typical of the NRs, including a DNA binding domain (DBD) and a C-terminal ligand binding domain (LBD), linked by a flexible hinge
(for details see Chapter 11). There is a less wellconserved A/B domain at the N terminus, varying in length from 23 amino acids in the VDR to more than 600 amino acids in the mineralocorticoid receptor [31]. There is little or no A/B region homology among the NRs, in part due to the existence of N-terminal variants or some receptors. These isoforms can arise from separate genes or from a single gene by alternative splicing, usage of different promoters, and/or distinct translational start sites [32–38]. Such variants contribute to functional diversity. As an example, the two human progesterone receptor isoforms differ in promoter specificity [39] and ligand responses [40]. There are also functional differences, as the progesterone A receptors can dominantly inhibit the B receptors [41,42] and other steroid receptor superfamily members [43]. Such distinct functional characteristics of N-terminal variant nuclear receptors may contribute to modulation of different physiological responses. 3. VITAMIN D RECEPTOR VARIANTS
The broad range of physiological vitamin D responses indicates functional diversity, and although there is only one VDR locus in each species [on human chromosome (Chr) 12, mouse Chr 15, and rat Chr 7], there is some evidence for VDR isoforms. In flounder (Paralichthys olivaceus), two VDR subtypes originate from different transcripts, with nucleotide and predicted amino acid sequence variations distributed
196
LUIS M. ESTEBAN, JOHN A. EISMAN, AND EDITH M. GARDINER
sometimes termed VDRA [28,29]. These two human VDR isoforms have been detected together at similar levels in human kidney tissue and in kidney, intestine, and bone cell lines [28], with the identity of the larger protein confirmed using a VDRB1-specific antiserum. In addition, the VDRB1-encoding transcript was also detected in parathyroid adenoma tissue and in skin and mammary gland cell lines [27]. VDRB1 exhibits differential transactivation activity [29] that appears to be due, at least in part, to a strong ligand-independent transcriptional activation domain in the A/B region [50]. The differential accumulation of VDRB1 into nuclear speckles may also contribute to transactivation differences [28] although a mechanism for such an effect has not been established. The distinct transactivation properties of VDRB1 would be consistent with a role in diversification of the vitamin D physiological response. The relative abundance of VDR and VDRB1 in each tissue or cell type may therefore be regulated. The complexity of the 5′ flanking region of the human VDR locus suggests mechanisms that may provide such regulation (discussed in Section III,A,1 and later).
throughout the receptor [44]. Flounder VDRa shows a more limited tissue distribution than VDRb and, to date, no functional differences between the two have been reported. VDRs from chicken and quail exhibit two isoforms (A and B) that differ by 14 amino acids and are translated from a single mRNA species. The chicken VDR gene contains 3 in-frame ATG sites for translation initiation, with the third in a position in exon 4 equivalent to the mammalian start codon site (Fig. 1). The first ATG is in a suboptimal context (TCCATGT), whereas the second is a better fit for the optimal consensus sequence (AGCATGG) [22,45]. Mutational analysis indicates that the two translated avian VDR isoforms arise from the first and second ATG sites by a “leaky scanning” mechanism [22], but no differential functions have been identified for these two chicken receptors. Similarly, the physiological significance of a dominant-negative C-terminally truncated rat VDR resulting from aberrant splicing leading to retention of intron 8 is not known [46]. Variants of human VDR have been identified. A common start codon polymorphism that produces a VDR with a three-amino-acid N-terminal truncation [47] has been associated with elevated transactivation activity [48,49]. A human VDR isoform with a longer N-terminal extension is generated from an exon 1D-containing transcript, as described earlier (See Section III,A,1) [27,28]. Other than its 50-amino-acid N-terminal extension, the VDRB1 protein is identical to the originally described VDR [19] (Fig. 2), which is
A /B
DBD
B. VDR Promoter To analyze the regulation of the VDR gene, its 5′ flanking region has been cloned and its promoter activity characterized in mouse [25,30], human [24,27,51], and chicken [52]. In these species the flanking sequences
Hinge
LBD
AF-2
427 aa
50
hVDR
427 aa
ATG
hVDR B1
ATG
TGA
Fok I 1A 1D
1B
1C
2
3
4 5 6
7 8
9
FIGURE 2 Structure of the human VDR isoforms. Receptor protein domains are shown on top, exon structure and alternative 5′ splicing patterns below. Arrows indicate the start codons for VDR (exon 2) and VDRB1 isoform (exon 1D). VDRB1 differs from VDR only in its N-terminal extension of 50 amino acids. *FokI polymorphism gives rise to a protein of 424 amino acids. White boxes indicate the exons that encode the N-terminal extension of VDRB1 isoform. Striped boxes represent untranslated exons.
197
CHAPTER 12 VDR Promoter and Abundance
5′ to exon 1 (mouse and chicken) or exon 1A (humans) (see Fig. 1) are highly conserved [52]. Their properties are the focus of the following sections. 1. NONHUMAN VDR PROMOTERS
The sequence of the mouse VDR 5′ flanking DNA contains characteristics of a GC-rich promoter including four tandem Sp1 sites and two upstream CCAAT boxes in opposite orientation (Fig. 3). Such TATA-less promoters are not uncommon among the steroid/ thyroid hormone receptor superfamily [53]. A 1.5-kb segment of the mouse 5′-flanking region activated transcription of a reporter gene in rat osteosarcoma ROS 17/2.8 and mouse fibroblastic NIH3T3 cell lines [25]. Deletion analysis indicated that the sequence from 54 to 177 bp upstream of the transcription start site provides most of the promoter activity and a region 177 to 481 bp upstream appears to have a repression function [30]. The first, second, and fourth Sp1 sites between 56 and 112 bp upstream of the transcription start site are necessary for mVDR promoter activity, whereas the Sp1–3 site appears to be inhibitory [30]. The Sp1 site proximal to exon 1 (Sp1-1) (position 1435, Fig. 3) overlaps a perfect consensus site for two closely related zinc finger transcription factors, Krox-20 and Krox-24 [54]. Both factors were originally noted for their importance in the nervous system [55,56]; however, Krox20−/− mice exhibit abnormal bone formation with a defect in endochondral ossification [57]. This phenotype is consistent with Krox20 expression in a subpopulation of growth plate hypertrophic chondrocytes and in differentiating osteoblasts, two cell types that are known to be responsive to vitamin D and to express the VDR (see Chapters 32 and 33). Thus, the effect of the Krox20 deletion on bone formation may relate to a defect in VDR expression in cells normally responsive to 1,25(OH)2D3. To date, however, VDR levels in these mice and the function of the Krox binding site in the mouse VDR promoter have not been tested. The mouse VDR promoter contains an AP-1 site (position 1123, Fig. 3) that differs by one nucleotide from one that overlaps the VDRE in the human osteocalcin promoter [58]. AP-1 is a target for the stressactivated protein kinases p38 and JNK, two major MAPK (mitogen-activated protein kinase) pathways. The p38 and JNK pathways and the c-Jun component of AP-1 were required for VDR stimulation by stress activation, which further sensitized human breast cancer MCF-7 cells to vitamin D3-induced growth inhibition [59]. There is also an imperfect AP-1 site in the mouse VDR promoter (position 1281, Fig. 3), similar to a binding site for the osteoblast-specific transcription factor NMP-2 (nuclear matrix protein 2) in the rat osteocalcin promoter [60]. Although these similarities
are suggestive, its role in regulating mouse VDR transcription is not known. No consensus vitamin D-, cAMP-, glucocorticoid-, or estrogen-responsive element was identified in the mouse VDR promoter 1.5-kb region by computer analysis, although there are a putative cyclic AMP-response element binding protein/Ets transcription factor (CREB/ETF) binding site at position 944 and an AP-2 site at position 1043 (Fig. 3) [25]. Treatment with forskolin, a cyclic AMP agonist known to increase endogenous VDR protein and mRNA content (discussed further in Section IV,B,4,1) [61], resulted in a three to fivefold increase in promoter-reporter activity, although the CREB/ETF site itself was not tested for function [25]. Further analysis of the mouse promoter demonstrated the presence of two E-boxes (at positions 1034 and 1129, Fig. 3), both with the consensus sequence ‘CACCTG’ target for the ZEB transcription factor, and both mediating up-regulated VDR expression in vitro [62]. In this region there are also five potential binding sites for the transcription factor c-MYB that appear to function additively with ZEB [62]. In this context it is notable that both ZEB and VDR knockout mice exhibit growth retardation and skeletal abnormalities [63,64]. A WT1, Wilms’ tumor gene, responsive element (WRE) overlapping the Sp1–4 site (position 1387, Fig. 3) is functional and able to mediate WT1-enhanced VDR promoter activity in transient transactivation assays, consistent with elevation of endogenous VDR upon overexpression of WT1 in colon carcinoma and kidney cell lines [65,66]. The rat VDR promoter has not yet been experimentally examined, but high DNA sequence homology (over 85%) with the mouse promoter suggests it is likely to function similarly (Fig. 3). Three of the functional Sp1 sites of the mouse promoter are perfectly conserved in the rat 5′ region whereas the fourth, Sp1–2, is divergent at two nucleotides. The rat Krox20/24 site differs from the mouse sequence by one nucleotide. The chicken VDR promoter is also structurally similar to mouse, lacking a TATA box and possessing several GC-rich Sp1 sites. Deletion studies indicated that the proximal Sp1 sites are strong elements to drive gene expression, with upstream promoter sequences having a negative effect on gene expression [52]. This promoter also has an AP2 site, and a Krox20/24 site outside the proximal promoter region [52]. Thus far, the regulatory contributions of these sites to chicken VDR expression have not been elucidated. 2. HUMAN VDR PROMOTERS
Regulation of the expression of human VDR appears to be more complex than the rodent and avian VDR with six untranslated exons 5′ to the exon that
198
LUIS M. ESTEBAN, JOHN A. EISMAN, AND EDITH M. GARDINER
FIGURE 3 Alignment of the promoter regions of mouse, rat, and human VDR. The first exon is underlined (exon 1 of mouse and exon 1A of human and rat). Bent arrows indicate major transcriptional start sites. Superscript stars on shaded nucleotides indicate the beginning of the published cDNAs [19,20,26]. Additional sequence in the first exon was determined: for mVDR (mouse) by primer extension [30] and 5′ RACE (Esteban et al., unpublished); for rVDR (rat) by 5′ RACE (Esteban et al., unpublished); for hVDR (human) by 5′ RACE [24]. In the 600 bp upstream of the first exon there is 88% identity between mouse and rat and about 63% between rodents and human sequence. Alignment with chicken VDR promoter is not included because its identity with mammals is only about 40%, although the GC-rich region containing the Sp1 sites is highly conserved [30]. Specific transcription factor responsive elements are indicated only for mouse VDR which has been more extensively studied. Other sites that are also present in the human and chicken sequence [24,52] and in the promoter of human exon 1C [51] are discussed in the text. Alignment is based on mouse VDR promoter coordinates (gb#AF017779, 1574 bp) and the homologous region of rVDR promoter from gb#AC119476 and hVDR promoter from gb#AY342401 (nt 235-2019). Asterisks indicate sequence identity in all three promoters.
199
CHAPTER 12 VDR Promoter and Abundance
E-Box c-MYB
Sp1 1A
WRE
Forskolin, ATRA, 17β-E2, Phytoestrogens, Dex ATG
ATG hVDR
1F
1E 1A1D
2
1B 1C
Cdx-2
ATRA
FIGURE 4 Regulatory regions of human VDR locus. White circles indicate regions with functional promoter activity, bent arrows on exons 1A, 1C, 1D, and 1F are transcriptional start sites. Two major initiation codons on exons 1D (for VDRB1) and exon 2 (for VDR) are indicated by ATG above downward arrows. Experimentally confirmed transcription factor response elements are: WRE (WT1 responsive element), Sp1 transcription factor, Cdx-2 (caudal-related homeodomain). Also two E-boxes (ZEB responsive element) and several c-MYB responsive elements present in the promoter region 5′ to exon 1A are functional in mouse [62], but have not been tested in the human sequence. Other untested consensus response elements for other transcription factors [24] are not shown. Forskolin, ATRA (all-trans-retinoic acid), 17β-E2 (17β-estradiol), phytoestrogens, and Dex (dexamethasone) induce the promoter 5′ to exon 1C, but response elements have not been defined.
encodes the originally described initiation codon, exon 2 (Fig. 4) [24,27]. Up to 14 variant transcripts assembled from multiple combinations of the six upstream exons were initially identified in human kidney [27]. Their expression is regulated by up to three distinct promoters: a promoter upstream of exon 1A and equivalent to those described for rodents and chicken [24,27], a second promoter upstream of exon 1F [27], and a third promoter upstream of exon 1C [51]. The proximal 5′-flanking region upstream of exon 1A of the human VDR gene is highly conserved with chicken and mouse loci, both in its lack of a TATA box initiator and the presence of a GC-rich region with putative Sp1 and transcription factor binding sites (Fig. 4). As in mouse, proximal elements from 103 to 464 bp upstream of the transcription start site displayed high promoter activity, whereas elements further upstream repressed transcription [24]. Also comparable to the mouse, deletion of the Sp1 sites proximal to the start of the mRNA resulted in a 10-fold drop in promoter activity, indicating an important role in promoter function. This same promoter may also drive the expression of exon 1D transcripts, as available evidence indicates that upstream sequences within 300 bp of exon 1D lack promoter activity and may, in fact, contain a suppressor element [27]. Further studies are required to clarify whether the exon 1D promoter does overlap the 1A promoter and/or transcribed exon sequence. A few transcription factor binding sites have been identified in the human 1A promoter region. At position –3731 to –3720 bp relative to the exon 1A transcription start site, there is a polymorphic binding site for
the caudal-related homeodomain protein Cdx-2 (see Section III,C), which activates VDR gene transcription by binding to this element [67,68]. As it is an intestinespecific transcription factor, Cdx-2 may contribute to tissue-specific regulation of the VDR. As in the mouse VDR promoter, there are two E-box targets for the ZEB transcription factor and potential binding sites for the transcription factor c-MYB [62]. There are also several WREs: one overlapping the most distal Sp1 site [66] and three more putative distal sites that are able to bind WT1 protein [69]. The only functional WRE of the latter three, 308 bp upstream of the exon 1A transcription start site (Figs. 3 and 4), appears to be distinct from the WT1-response element defined for the mouse promoter. A distal promoter of the human VDR gene, more than 9 kb upstream of exons 1A and 1D, drives the expression of exon 1F–containing transcripts. Whereas exon 1A and 1D transcripts appear to be ubiquitous, the distribution of exon 1F–containing transcripts is more restricted, detected to date only in human kidney, parathyroid tissue, and the intestinal carcinoma cell line LIM 1863 [27]. As these sources represent major calciotropic target tissues, the distal promoter may require tissue- or cell-specific regulatory factors to mediate the restricted gene expression pattern. Interestingly, although not translated, exon 1F–containing transcripts are underexpressed in parathyroid adenomas and in hyperplastic glands of secondary hyperparathyroidism in comparison with normal parathyroid glands [70]. In addition to those initiating at exons 1F, 1A, and 1D, human VDR transcripts that initiate at exon 1C
200
LUIS M. ESTEBAN, JOHN A. EISMAN, AND EDITH M. GARDINER
have been detected in MCF-7 breast cancer cells [51]. These transcripts appear to arise from a third human VDR promoter upstream of exon 1C, as evidenced by the promoter activity of a 1.3 kb 5′ flanking segment in reporter gene assays [51]. Characteristic promoter features in this segment include GC-rich regions and a TATA box at –29 bp, as well as putative AP-2 and Sp1 transcription factor binding sites and several consensus glucocorticoid response elements (GREs). The GREs appear to be regulated according to cell type, as dexamethasone induced promoter activity in the SUM159PT but not in the MCF-7 breast cancer cell line [51]; however, specific functional GRE(s) in this region have not been defined. There are no consensus vitamin D (VDRE), retinoic acid (RARE), or estrogen (ERE) responsive elements in the functional 1.3-kb exon 1C promoter fragment, nor is there a clear forskolin responsive element [51]. This is somewhat surprising, as transcriptional regulation by this region is enhanced in the estrogen receptor (ER)-positive MCF-7 cell line by treatment with forskolin, retinoic acid, and 17β-estradiol. The latter response appears to be mediated by the estrogen receptor, as the response to estradiol is absent in the ER-negative SUM159PT cell line [51]. This promoter is also responsive to up-regulation by phytoestrogens in the ER positive human breast cancer cell lines T47D and MCF-7 [71]. A retinoic acid responsive region downstream of exon 1C may be relevant for regulation of endogenous VDR expression [24], and the effect of forskolin may be mediated through the AP-2 sites upstream of the human exon 1C [51], as some studies have implicated AP-2 elements in mediating cAMP responsiveness of other gene promoters [72,73]. Further studies are required to clarify the mechanisms of these 1C promoter hormone responses, which are of particular interest as there is currently no evidence to suggest that either the exon 1F or the exon 1A promoter is hormonally regulated [24,25,27,30].
1F
ATG 1A 1E 1D
Cdx-2
1B
ATG 2 1C
Fok I
C. VDR Polymorphisms Polymorphisms in the VDR locus have been studied in relation to skeletal biology, renal metabolic and immune functions, and cancer risk and responsiveness in a variety of populations worldwide. The most commonly studied polymorphisms and a more recently described polymorphism in a transcription factor binding site are briefly discussed here. A polymorphism has been described in the Cdx-2 binding site of the human promoter (see Section III,B,2), which resulted in lower promoter activity (70% of control transcription level) and impaired binding to Cdx-2 protein (Fig. 5). This single nucleotide polymorphism may have significant impact on target-cell responsiveness to 1,25(OH)2D3 by affecting VDR abundance. It has been associated with lower bone mineral density (BMD) at the lumbar spine in postmenopausal Japanese women [68]. Presence of a polymorphic FokI restriction site in exon 2 (the ‘f’ allele) actually affects receptor structure, allowing production of a normal-length human VDR (427 aa), whereas absence of the site (the ‘F’ allele) produces an N-terminally shortened 424 aa protein [47]. The ‘F’ allele has been associated with elevated transactivation activity rather than differences in VDR expression per se [48,49,74]. Specific FokI alleles have been correlated with differences in bone density, fracture incidence, nutritional rickets, risk of osteosarcoma, auto-immune hepatitis, urolithiasis, insulin sensitivity and drug clearance rates [75–81]. Restriction site polymorphisms for the enzymes BsmI and ApaI have been described in the last intron of the human VDR gene and a silent TaqI site in exon 9 [82,83] as well as a polyadenosine (A) microsatellite within the 3′ untranslated region (3′ UTR) [84]. These polymorphisms have been associated with differences in BMD in a number of genetic and population based studies [85–90], but not all have observed such associations [91,92]. Although unlikely to lead to changes in
TGA 3
45
6
78
Bsm I Apa I
9
Poly(A) Taq I microsatellite
FIGURE 5 Polymorphisms in human VDR gene. Single-nucleotide polymorphism has been identified in the Cdx-2 site, restriction site polymorphisms in exon 2, intron 8, and exon 9, and poly(A) microsatellite in 3′ UTR. Only the FokI polymorphism affects protein structure. The BsmI, ApaI, TaqI, and poly(A) polymorphisms are in linkage disequilibrium in some populations [74,84].
201
CHAPTER 12 VDR Promoter and Abundance
the receptor structure, these polymorphisms could affect mRNA accumulation and hence VDR protein level and function. A minigene representing a haplotype lacking the BsmI and ApaI sites but including a functional TaqI site (‘BBAAtt’) produced higher reporter gene activity than the ‘bbaaTT’ haplotype in transient transfection assays, suggesting that the ‘BBAAtt’ haplotype accumulated more VDR mRNA and protein [85]. Consistent with this finding, alleles ‘b’, ‘a’, and ‘T’ were associated with significantly lower VDR and higher PTH mRNA levels in patients with primary hyperparathyroidism [93]. Similarly, the poly(A) microsatellite ‘L’ (long) allele in the 3′ UTR was associated with higher VDR activity than the ‘S’ (short) allele in human fibroblast cell lines [74]. Allelespecific variation in mRNA stability or transcriptional regulation may be tissue or cell type specific, however, as the opposite pattern was observed in myeloid leukemia and prostate cancer cell lines, with a 30% reduction in RT-PCR product derived from the small ‘t’ allele [94]. Tissue specific allele effects may also explain the lack of relationship between VDR alleles and receptor density in duodenal mucosa of healthy premenopausal women, despite an association with bone mineral content (BMC) [95], and also the lack of association between VDR genotype and protein concentration in duodenum [96], peripheral blood mononuclear cells [97], and cultured skin fibroblasts [98]. Thus, although there appears to be no association between the BsmI, ApaI, and TaqI polymorphisms and VDR levels in many tissues and cell types studied, genotype may affect receptor levels in some tissues or under certain pathological conditions. Recent studies have indicated genetic interactions between various VDR alleles, with combined haplotypes sometimes regulating polymorphism effects that were not detected in single polymorphism analysis [74].
IV. REGULATION OF VITAMIN D RECEPTOR EXPRESSION AND ABUNDANCE An important mechanism for the modulation of cellular responsiveness to 1,25(OH)2D3 is mediated by the regulation of receptor abundance. Developmental regulation occurs, with receptors for 1,25(OH)2D3 appearing in the neonatal rat intestine shortly before weaning, coincident with a switch to vitamin D–dependent calcium absorption [99]. Target cell responsiveness is also regulated, with the ability of 1,25(OH)2D3 to stimulate 24-hydroxylase and inhibit collagen synthesis in mouse osteoblastic cells proportional to its receptor
content [100,101]. Receptor abundance also varies with cellular differentiation, exemplified in a series of clonal rat osteosarcoma cell lines of increasing osteoblasticlike character (ROS 24/1, ROS 2/3, and ROS 17/2.8), in which receptor levels correlated with the magnitude of 1,25(OH)2D3-associated inhibition of growth, transition from cuboid to spindle-like morphology, and suppression of colony formation in soft agar [102]. The following section focuses on a variety of studies which have shown that complex physiological signals involved in the regulation of VDR abundance vary among species and between the various 1,25(OH)2D3 target organs. To facilitate comprehension of this complex story, main findings are summarized in Table I.
A. Homologous Regulation VDR expression is affected by the presence of its own ligand, 1,25(OH)2D3, with ligand-mediated autoregulation of VDR documented in several experimental systems. Using a chicken cDNA probe, a receptor mRNA species in mouse 3T6 fibroblasts was substantially increased by prior exposure to 1,25(OH)2D3 [14]. In the pig kidney cell line LLC-PK1 there was homologous up-regulation by 1,25(OH)2D3 and other vitamin D analogs with a substantial increase in the number of 1,25(OH)2D3 receptors without altering the affinity of receptor for the hormone. Treatment of these cells with the RNA synthesis inhibitor actinomycin D showed that the increase in receptors was in part dependent on RNA synthesis, and in part on receptor occupancy [103]. Similar results were observed in rat osteosarcoma ROS 17/2 cells, in which there was a receptor increase upon treatment with 1,25(OH)2D3, with roughly 50–60% of the increase prevented by inhibition of RNA polymerase II activity with α-amanitin [104]. Similarly, human breast cancer cells increased specific binding of [3H]1,25(OH)2D3 upon hormone treatment, which was prevented by actinomycin D treatment [105]. Thus, although pharmacological inhibitors of transcription partially inhibit VDR induction, other 1,25(OH)2D3 responsive mechanisms also appeared to be involved in homologous regulation of receptor abundance. The typical response to 1,25(OH)2D3 treatment is characterized by an ascending phase, which represents hormone association with receptor and reaches a maximum at 90–120 min, followed by a rapid descending phase that is closely associated with a decrease of medium 1,25(OH)2D3 due to the metabolism of the hormone [106]. Treatment of the osteosarcoma cell line UMR-106 with ketoconazole (an inhibitor of 1,25(OH)2D3 metabolism) resulted, after a lag-time of about
202
LUIS M. ESTEBAN, JOHN A. EISMAN, AND EDITH M. GARDINER
TABLE I Brief Summary of Regulators of VDR Abundance
1,25(OH)2D3
Calcium Phosphorus
Fatty acids Glucocorticoids
Estrogens
Retinoic acid
Growth factors Cell cycle/ differentiation Growth hormone (GH) PTH
PKA activators
PKC activators
aD 3
Effect on VDR levels
Comment
Up-regulation in calcemic tissues (kidney, intestine, parathyroid gland), and no change in other tissues Up-regulation in cell lines
In normocalcemic animals
[113,114,116–120]
Mainly due to stabilization of the VDR by D3a Hypocalcemia prevents upregulation by D3 In low-phosphate diet
[124,125]
In hypocalcemia, down-regulation in kidney and parathyroid gland, and no effect in intestine In rat and chick increase of intestinal VDR. In kidney no change Parathyroid gland had no change Modulation of differentiation and increase of VDR by butyrate Species-dependent. Increased VDR in rat and decreased in mouse Intestine up-regulation Kidney down-regulation Testis, heart, lung, no change Increase of VDR in intestine and enhancement of Ca2+ absorption No change in kidney VDR increase in osteoblastic cells Up-regulation of hVDR exon 1C prom Up-regulation of VDR is tissue- and species-dependent Up-regulation of hVDR exon 1C prom Up-regulation of VDR and stimulation of proliferation. In vivo, tissue-specific Proliferating cells, VDR high Quiescent cells, VDR low Increased Ca2+ absorption and intestinal CaBP-9K, but no change in VDR Reduction in kidney VDR increase in osteoblastic cells VDR down-regulated in ROS17/2.8 and decreased up-regulatory effect of D3 VDR up-regulated by forskolin and other PKA activators Blocking Ca2+ channels reduces cAMP induction Phorbol esters down-regulate VDR in osteoblastic cells In rat kidney and epithelial cells potentiation of D3 action VDR phosphorylation
is 1,25(OH)2D3.
References
[119,127,129,130] [112,132]
With high-phosphate diet Synergistic effect of D3 and butyrate Effect probably not VDR specific, but on general development Effects are tissue-dependent and cell type-dependent
[133] [135,136,139]
Some effects of 17β-estradiol on Ca2+ are independent of D3
[155–158,162]
ER-dependent Synergistic effect of ATRA and D3
[51] [126,152,171–174]
[141,142,144] [109,143,150–152]
By EGF, insulin, IGF-I
[24,51] [179–182]
No up-regulation by D3 in differentiated cells May be IGF-I mediated effect
[100,101,185, 187] [190–192,194]
cAMP mediated By PTH
[196,198,199] [200,201]
Via cAMP elevation
[61,182,198,203]
When activated by PTH Via PKC activation
[204] [183,205]
Via PKC activation VDR-PKC cross-regulation
[207–211,214]
CHAPTER 12 VDR Promoter and Abundance
90 min, in a sharp increase of receptor accumulation, which was blocked by coincubation with cycloheximide and actinomycin D [107]. Most of these findings were based on receptor quantification by high specific activity tritiated ligand and, therefore, non-hormone-binding or unoccupied forms of the receptor may not have been characterized. Western blots using the monoclonal antibody IVG8C11, which has high cross-reactivity with porcine and other mammalian VDR molecules [11,108], avoided this potential problem and detected up-regulation of 1,25(OH)2D3 receptors in human HL-60 promyelocytic leukemic cells line at 12 hr and down-regulation at 48 and 72 hr in the continuous presence of hormone [13]. Furthermore, 1,25(OH)2D3 stabilized VDR mRNA levels in human MG-63 osteosarcoma cells, increasing its apparent half-life by 11 hr and increasing its VDR protein content [109]. However, the effect of 1,25(OH)2D3 on VDR levels appears to be cell specific, as treatment with the hormone down-regulated VDR mRNA level in the human megakaryoblastic leukemia cell line HIMeg, but caused no change in the human osteosarcoma cell line HOS-8603 [110]. Similarly, treatment with 1,25(OH)2D3 did not alter VDR protein expression in three phenotypically distinct osteosarcoma cell lines [111]. In general, these in vitro studies indicate that in many but not all cell line models, receptor abundance is increased by transcriptional and translational mechanisms. Specificity of VDR up-regulation by 1,25(OH)2D3 has also been observed in numerous in vivo studies, with outcomes highly dependent upon the tissue examined, the method by which 1,25(OH)2D3 levels were manipulated, and perhaps also upon species. Basal expression of chicken intestinal VDR appears not to be dependent upon 1,25(OH)2D3, as VDR mRNA concentrations were similar in vitamin D–deficient and normal chick duodena [112]. In vitamin D–deficient rats injected with 1,25(OH)2D3, there was significant up-regulation of renal VDR after 5 days of treatment, again with minimal effect on duodenal VDR [113]. Duration of treatment influenced results, with shorter treatment period more likely to yield an increase in VDR mRNA levels. In rat parathyroid gland there was a maximum increase of VDR mRNA at 6 hr and only a marginal increase at 24 hr [114], whereas daily dosing for 5 days did not alter the level of VDR mRNA in either the intestine or kidney of vitamin D–deficient rats [115]. Furthermore, mRNA levels of the receptor in the intestine of vitamin D-deficient rats increased 10-fold above deficiency levels at 6 and 12 hr after an intravenous dose of 1,25(OH)2D3, returning to predosing levels at 24 hr [116]. The protein response in this study was less marked, as there was only a twofold increase of the VDR at 12 hr, as detected by IRMA, with
203 no change detected by ligand-binding assay [116]. In normocalcemic rats, the administration of vitamin D resulted in a 5-fold increase in kidney receptor concentration, whereas in hypocalcemic animals, vitamin D treatment did not change receptor levels [117,118], perhaps reflecting a homeostatic increase in plasma 1,25(OH)2D3 levels in the latter group. A consistent set of observations was reported with exogenous administration of 1,25(OH)2D3 in rats fed a normal diet, which resulted in a 1.5-fold increase in duodenal and a 3-fold increase in renal VDR content [119]. By contrast, rats fed a low-calcium diet in the same study exhibited plasma 1,25(OH)2D3 concentrations similar to those treated with exogenous 1,25(OH)2D3, but duodenal VDR content was not up-regulated and renal VDR was 20–38% lower than in rats with normal calcium intake. Thus, VDR levels in kidney and intestine may be homeostatically regulated by dietary calcium, but not through the elevation of plasma 1,25(OH)2D3 levels [119] (see Section IV,B,1,a). Yet homologous VDR up-regulation was not a universal property of 1,25(OH)2D3 target tissues, as homologous up-regulation of VDR occurred under a wide variety of experimental conditions in rat kidney but not in non-calcium homeostatic VDRcontaining tissues such as lung, heart, or testis [120]. Clearly, the mechanisms of homologous up-regulation of VDR must be complex. Many studies addressing mechanism have focused on VDR transcript levels in cell line models after 1,25(OH)2D3 treatment. In addition to those studies mentioned above, increased VDR mRNA was also observed after treatment in mouse fibroblast 3T6 cells [121], the human osteosarcoma cell line MG-63 [109], and HL-60 cells [122]. Accumulation of VDR transcripts appears to occur in part through increased transcription, but also to a substantial extent through mRNA stabilization, as evidenced by studies using RNA polymerase II inhibitors [103–105]. Studies suggestive of stabilization of the VDR protein by its ligand include radio-ligand binding studies in human breast cancer and pig LLC-PK1 cell lines [105,123], but not in the rat osteosarcoma cell line ROS 17/2 [104]. More recent studies using IRMA indicated that VDR protein levels were elevated 3-fold after 8 hr of treatment with no change in mRNA levels in 3T6 mouse fibroblasts and IEC-6 rat intestinal epithelial cells, as well as in rat intestine and kidney [124]. Notably, with inhibition of protein synthesis using cycloheximide, 1,25(OH)2D3 markedly reduced degradation of previously formed receptor. Consistent findings were reported in the rat osteosarcoma ROS 17/2.8, and in the human breast cancer T47-D cell lines, in which administration of 1,25(OH)2D3 did not change receptor mRNA level, but protein did accumulate as a result of increased receptor half-life [125,126]. Thus, in addition
204 to increasing transcription and mRNA stabilization, it appears that 1,25(OH)2D3 can increase VDR content in vivo or in cell culture by stabilizing the receptor. This effect on protein stability may be due to hormoneinduced conformational changes that directly affect receptor degradation or that enhance its functional interactions with RXR, DNA, or transcriptional co-regulators and thereby prolong receptor half-life. As ligand effects on nuclear receptor complex formation, stabilization and turnover are a very active area of investigation (described in more detail in Chapters 13 to 16), it is likely that this aspect of VDR regulation could be significantly clarified in the near future.
B. Heterologous Regulation VDR expression can be modulated by numerous physiological stimuli such as dietary composition, steroid hormones and retinoids, growth factors, peptide hormones, and second messenger activators. This section will consider the effects of these agents in regulating VDR levels to clarify the interrelationships between their activities in calcium homeostasis and cell proliferation and differentiation. 1. DIETARY COMPOSITION
A primary function of the vitamin D hormone system is control of calcium and phosphorus homeostasis, in part through 1,25(OH)2D3 regulation of its own receptor, as described in Section IV,A. As the levels of 1,25(OH)2D3 in vivo are tightly linked with the levels of dietary calcium and phosphorus (discussed in detail in Chapters 24 to 26), this section will briefly discuss mechanistic aspects of the effects of dietary calcium and phosphorus on VDR expression. Fatty acids also play a role of in the regulation of VDR levels, in part through interactions between the VDR and other nuclear receptors and transcriptional cofactors, as described in Chapter 16. Recent studies that reveal a role for dietary lipids in the regulation of VDR abundance are also briefly discussed here. a. Calcium A number of studies have shown that calcium levels have no impact on the VDR levels in the intestine, but do regulate its abundance in kidney and parathyroid gland, as under conditions of severe hypocalcemia and resulting hyperparathyroidism, little or no VDR is present in these tissues. For example, in rat kidney, in the absence of vitamin D, total VDR protein was increased 2-fold by an increase in serum calcium [117], indicative of a vitamin D–independent regulatory pathway. Also, although vitamin D increased renal VDR levels 5-fold at normal serum calcium levels, this was not the case in hypocalcemic animals [117].
LUIS M. ESTEBAN, JOHN A. EISMAN, AND EDITH M. GARDINER
Complementing these VDR protein studies, dietary calcium restriction decreased renal VDR mRNA levels in mice, even after treatment with 1,25(OH)2D3, whereas with dietary calcium supplementation, treatment with 1,25(OH)2D3 caused VDR mRNA to increase 8- to 10-fold in the vitamin D-deficient group [127]. These findings are also consistent with reduced renal VDR protein content after 1,25(OH)2D3 treatment of rats with low dietary calcium intake compared to normocalcemic animals [118,119,128,129]. Thus, hypocalcemia markedly impairs homologous up-regulation of renal VDR by 1,25(OH)2D3, suggesting interaction between distinct 1,25(OH)2D3- and calcium-responsive mechanisms in this tissue. By contrast to these renal changes, duodenal VDR was not up-regulated after 1,25(OH)2D3 treatment in rats fed a low-calcium diet [118,119,129]. These findings suggest that in rodent models, down-regulation of VDR under hypocalcemic conditions is tissue-specific: intestinal VDR is present at both low and high serum calcium levels, whereas kidney VDR is absent or markedly reduced under hypocalcemic conditions, resulting in abolition of the 1,25(OH)2D3 response. A high-calcium diet prevented a fall in VDR mRNA levels in the parathyroid gland of rats with hyperparathyroidism, as did 1,25(OH)2D3 treatment [130]; this coordinate regulation of VDR expression by calcium and 1,25(OH)2D3 in kidney and parathyroid gland was not associated with changes in intestinal VDR mRNA. There are species differences, however, as there was a significant decrease in VDR mRNA in the duodenum of chicks fed a low-calcium diet in the presence of increased serum 1,25(OH)2D3 [112]. Thus, complex regulation of VDR levels under dietary deficiencies is likely to involve other factors or hormones such as parathyroid hormone for modulation of the 1,25(OH)2D3 response. Other physiological changes affecting calcium homeostasis are pregnancy and lactation, which both up-regulate intestinal calcium absorption. However, rat intestinal VDR mRNA did not change during gestation and lactation, although intestinal VDR content increased 2-fold in late pregnancy and lactation, suggesting a posttranscriptional mechanism of VDR up-regulation [131]. In general, these studies indicate that both 1,25(OH)2D3 and calcium are required for VDR production in the kidney above the basal level, whereas calcium has only a minor effect on intestinal VDR level. Downregulation of renal VDR in response to hypocalcemia may be a protective measure to block 1,25(OH)2D3mediated suppression of the 1α-hydroxylase and induction of the 24-hydroxylase, which would result in a net increase in serum 1,25(OH)2D3 levels [129].
205
CHAPTER 12 VDR Promoter and Abundance
The mechanisms by which 1,25(OH)2D3 and calcium regulate renal VDR expression levels may be indirect or posttranscriptional, as no vitamin D response elements have been detected in either mouse or human VDR promoter, and no response to 1,25(OH)2D3 has been observed in reporter-gene analyses, suggesting that 1,25(OH)2D3 does not directly activate expression of its own receptor [24,25,27,51]. b. Phosphorus Dietary phosphorus restriction upregulates VDR in a tissue-specific and time-dependent manner [132]. A low-phosphorus diet resulted in rapid increase of serum 1,25(OH)2D3 and a 3.5-fold increase in rat intestinal VDR by day 3, as quantitated by ligand binding assay, followed by a decrease to a plateau of 2.5-fold of control levels by day 5 to 10. This finding is consistent with an increase in VDR mRNA in duodenum of chicks fed a low-phosphorus diet [112]. By contrast, a low-phosphorus diet did not up-regulate kidney or splenic monocyte/macrophage VDR in rat [132]. In another study, rats on a high-phosphorus diet did not exhibit altered levels of VDR in the parathyroid gland, although PTH mRNA levels were elevated and there was prevention of both hypocalcemia and low calcitriol levels [133]. The mechanisms by which phosphorus may regulate VDR mRNA levels and functional receptor abundance have not yet been elucidated. c. Fatty Acids Understanding of the roles of dietary fatty acids in regulation of gene expression is increasing [134] and evidence that lipids regulate VDR levels is accumulating. In rats, an ovariectomy-associated increase in intestinal VDR protein declined after longterm dietary supplementation with essential fatty acids [135]. As lipids activate transcription by the nuclear receptor peroxisome proliferator-activated receptor (PPAR), their effects on VDR levels may be indirect, via increased competition by PPAR for the VDR heterodimerization partner RXR [134]. Fatty acid regulation of protein kinases, particularly PKC, may also modulate VDR action and expression levels (see Section IV,B,4). Butyrate, a short-chain fatty acid and normal constituent of colon content, enhanced nuclear VDR protein content in Caco-2 cells in a time- and dosedependent manner; its prodrug, tributyrin, significantly increased VDR-mRNA level (2.5-fold) and VDR binding activity [136]. This response may mediate the synergistic effect of butyrate compounds with 1,25(OH)2D3 in induction of differentiation in the Caco-2 cell line [137]. Similar butyrate-induced increases have been observed in some [138,139], but not all [139] cell culture models, perhaps reflecting tissue and species specificity. To date, however, there is limited evidence for butyrate regulation of VDR abundance or 1,25(OH)2D3 function in vivo.
2. STEROIDS AND RETINOIDS
In addition to 1,25(OH)2D3, other hormones that regulate VDR abundance are steroids and retinoids. Like calcitriol, these classes of hormones also act through nuclear receptors to regulate transcription of target genes that include the VDR gene. This section will focus on glucocorticoid, estrogen, and retinoid effects on VDR abundance, as each is clinically important, particularly in skeletal health. a. Glucocorticoids In prolonged exposure to elevated glucocorticoid levels such as in Cushing’s syndrome or in glucocorticoid therapy, glucocorticoid actions appear to oppose vitamin D effects (see Chapter 73). The result is often osteopenia, with major calcium homeostatic changes in intestine and bone contributing to accelerated bone loss [140]. The present section will focus on the role of glucocorticoids in the regulation of VDR abundance as a possible basis for this apparent antagonism between glucocorticoid and vitamin D actions. In vivo treatment with the glucocorticoid hydrocortisone has been reported to increase intestinal VDR mRNA in rat if administered between 15 and 17 days postpartum, although administration before day 14 or on days 19–21 postpartum did not affect VDR expression [141,142]. The sensitive period between days 15 and 17 postpartum coincides with a glucocorticoid-sensitive period of rat intestinal development, however, and may reflect a general glucocorticoid effect on intestinal maturation rather than a specific effect on VDR mRNA, as it is accompanied by an increase in actin mRNA [142]. The effect of glucocorticoids does appear to be tissue specific, as dexamethasone injection caused VDR up-regulation in rat intestine and down-regulation in kidney, but had no effect on receptor level in testis, heart, and lung [143]. Species differences may also exist, as intestinal VDR was decreased by hydrocortisone administration in mouse [144], whereas changes in VDR were not observed in chicken intestine after treatment with the glucocorticoid compound triamcinolone [145]. There is also species variation in bone cell response. In short-term rat calvarial organ cultures glucocorticoids prevented a fall in VDR levels [146] and in rat osteosarcoma cell cultures glucocorticoids stimulated 1,25(OH)2D3 receptor levels [147,148]. By contrast, in mouse osteoblastic cells glucocorticoids caused a decrease in VDR abundance [149]. Given these interspecies differences, extrapolation from animal models to humans about glucocorticoid effects on VDR is problematic. For mechanistic studies, human cells and tissue models are particularly important, even though outcomes can vary in a single cell line. In one study using MG-63 osteosarcoma cells, dexamethasone produced a time- and dose-dependent
206 decrease in VDR mRNA [150], whereas in another study dexamethasone increased the amount of VDR mRNA, although VDR protein levels were not affected [109]. Similarly, VDR protein levels in cultured human keratinocytes and fibroblasts were not altered after treatment with corticosteroids [151]. Evidence for functional interaction comes from MCF-7 breast cancer and Caov-4 ovarian cancer cell lines, in which dexamethasone augmented calcitriol effects, although VDR levels were not assessed in this study [152]. Thus, it seems that the effects of glucocorticoids on the 1,25(OH)2D3 response in human cells are tissue-specific and may not relate to VDR levels per se in all cases. Although the effects of glucocorticoids on VDR might be mediated by direct action of the glucocorticoid receptor on the promoter of the VDR gene, no glucocorticoid responsive elements (GREs) have been found in the exon 1 promoter region (1.5 kb) of the mouse or exon 1A upstream region of the human VDR locus (discussed in Section III,B) [24,25,27]. However, in the 5′ flanking region of human exon 1C there are putative GREs [51]. Dexamethasone did not induce activity of this promoter in the estrogen-dependent MCF-7 breast cancer cell line, although there was upregulation in the estrogen-independent SUM159PT breast cancer cell line [51], consistent with cell type specificity of glucocorticoid effects on VDR abundance. The presence of multiple promoters, as in the human VDR locus, or indirect regulation via other mechanisms (e.g., as in estrogen-dependent versus estrogenindependent breast cancer cell lines, or perhaps by competition for transcription cofactors; see Chapter 16) may account for the apparent differences in the effects of glucocorticoids on VDR regulation. Furthermore, glucocorticoids may interact with other hormones that regulate VDR abundance, such as PTH (see Section IV,B,3,c). b. Estrogens Actions of the steroid hormone estrogen are mediated by the estrogen receptor (ER), another member of the nuclear receptor superfamily. Reductions in estrogen levels at the menopause are associated with decreased bone mineral density, which is thought to occur as a result of calcium homeostatic changes. One such change is a reduction in intestinal calcium absorption [153] which is improved by estrogen replacement [154]. Although 17β-estradiol (E2) may positively affect intestinal calcium absorption independent of vitamin D [155], the actions attributed to estrogens might also be mediated at least in part by regulation of 1,25(OH)2D3 or VDR levels, as 1,25(OH)2D3 is the principal hormonal regulator of intestinal calcium absorption. Animal models support this concept, as estrogenstimulated increases in VDR expression and response were observed in rat duodenum [156,157]. Other studies
LUIS M. ESTEBAN, JOHN A. EISMAN, AND EDITH M. GARDINER
have found that estrogen and phytoestrogens such as genistein increase VDR transcript level, protein expression, and endogenous 1,25(OH)2D3 bioactivity in rat colonic mucosa and colon carcinoma cell lines [158,159]. Thus, some estrogen activities in the colonic mucosa such as growth suppression could be mediated, at least in part, by increased responsiveness to endogenous 1,25(OH)2D3. However, in another study intestinal VDR levels in immature rat were unaffected by estrogen, although there were increases in VDR abundance in uterus and liver but a decrease in kidney [160,161]. Thus, conclusions about the regulation of the VDR in one tissue cannot be extrapolated to other tissues. There have been a number of studies of estrogen effects on VDR abundance in bone-derived cells because of the relevance to postmenopausal osteoporosis. VDR expression in mouse bone decreased with aging and estrogen deprivation but recovered after E2 supplementation [162]. In vitro findings are consistent, as E2 treatment doubled total cellular VDR-binding capacity in the ROS 17/2.8 rat osteosarcoma cell line [163], increased VDR protein in human osteosarcoma cell lines [164], and increased VDR mRNA levels in the human MG-63 and N-976 osteoblastic cell lines [109,162]. Correlation between VDR mRNA and protein changes was not entirely consistent [109], although E2 treatment did potentiate the antiapoptotic effect of 1,25(OH)2D3 in the latter study [162]. Interaction between vitamin D and estrogen regulation has been observed in VDR knockout mice, in which VDR appears essential for the regulation of CYP19 (aromatase cytochrome P450, a key enzyme in estrogen biosynthesis) of both female and male gonads [165]. To date, estrogen responsive elements (EREs) have not been detected in the promoter region of human, mouse, or chicken VDR genes. However, the human exon 1C promoter region (see Section III,B,2) was up-regulated by 17β-estradiol and the phytoestrogens resveratrol and genistein, despite the apparent lack of an ERE in the promoter region [51,71]. As for glucocorticoids, estradiol effects on VDR abundance vary according to species and tissue. Evidence indicates that there are significant interactions between estradiol and vitamin D responses in bone and intestine, with some renal involvement as well. It is not clear whether these findings reflect a particular experimental focus on calcium homeostatic tissues, or a special role for an estrogen–vitamin D axis in calcium regulation. Studies of this interaction in other non-calcium-homeostatic tissues should resolve this question. c. Retinoic Acid Vitamin A and its active metabolites retinol and retinoic acid, known for their involvement in vision, reproduction, embryonic development, and epithelial growth and differentiation, are also
207
CHAPTER 12 VDR Promoter and Abundance
necessary for normal bone metabolism. The actions of the retinoic acids are mediated by nuclear receptors closely related to the VDR. There are two families of nuclear retinoid receptors, the retinoic acid receptor RAR, activated by all-trans-retinoic acid (ATRA) and/or 9-cis-retinoic acid (9-cis RA), and the retinoid X receptor RXR, activated specifically by 9-cis RA [166]. In addition, RXR can act as DNA binding partner for a number of other nuclear receptors, including those for the thyroid hormone (TR), as well as the VDR and the RAR [167,168]. Vitamin A and 1,25(OH)2D3 share a number of physiological effects, including stimulation of bone resorption and effects on bone cell differentiation [169] and on keratinocyte proliferation and differentiation [170]; thus, vitamin A effects on bone and other tissues may be mediated through VDR regulation. All-trans-retinoic acid produced a dose-dependent increase in 1,25(OH)2D3 binding in rat osteosarcoma cells after 24 hr that was inhibited by actinomycin D [171]. This increase in binding was due to increased VDR protein, rather than modification of receptor affinity for the hormone. Induction of VDR by retinoic acid was greater in proliferating than nonproliferating tumorigenic cells (ROS 17/2A and UMR-106M), but not in the nontumorigenic cell line RCJ 1.20 [172]. ATRA caused a reduction in VDR abundance in rat osteoblastic cells, but an increase in receptor level in mouse osteoblasts that required protein and RNA synthesis. These opposite effects of ATRA on mouse and rat cells suggest that species-dependent factors modulate the action of retinoids in mammalian cells [173]. Findings in human osteosarcoma (MG-63) and breast cancer (T47-D) cell lines were similar to those in mouse osteoblasts with an ATRA-induced increase in VDR mRNA and its half-life although, at least in the MG-63 cells, there was no change in VDR protein levels [109,126]. Other studies have focused on the synergistic effects of retinoic acid and 1,25(OH)2D3 as an indication of their interactions and cross-regulation of their activities. In breast and ovarian cancer cells (MCF-7 and NIH:OVCAR3), and in human promyeloid leukemia cells (HL-60), ATRA and vitamin D3 analogs have additive or synergistic effects with respect to inhibition of cell growth, DNA synthesis, and the induction of differentiation [152,174]. In addition, ATRA cooperates with 1,25(OH)2D3 to promote differentiation of HL60 cells to monocytes [175]. By contrast, although ATRA and 1,25(OH)2D3 each inhibited proliferation of normal human keratinocytes, in combination they had antagonistic rather than synergistic effect [176]. The action of retinoic acid may be through direct action of RAR on the VDR gene promoter, as the ability
of retinoic acid to up-regulate the VDR is mediated, at least in part, at the transcriptional level. There are sites that may confer retinoic acid responsiveness upstream and downstream of the human VDR exon 1C promoter [24,51] (see Section III,B,2), but retinoic acid response regions have not been identified in the rodent VDR loci. It appears, therefore, that the retinoic acid regulation of VDR may be tissue and species specific. It is possible that, like glucocorticoid and estrogen effects, retinoids may affect VDR abundance by indirect mechanisms such as changing levels or availability of RXR or of shared transcriptional co-regulators. If so, there may be specificity in the mechanism, as another hormone with a receptor that heterodimerizes with RXR, triiodothyronine, increased VDR content in MG63 osteosarcoma cells [177], and the progesterone analog R5020 increased VDR in the T47-D breast cancer cell line [126]. 3. PEPTIDE HORMONES
In addition to the steroid and retinoid hormones, peptide hormones also play important roles in determining VDR level. They include the polypeptide growth factors, growth hormone, and the calciotropic hormones PTH and PTHrP. This section focuses on the effects of these compounds, which act via membrane receptors. a. Growth Factors In general, 1,25(OH)2D3 inhibits cell proliferation and induces differentiation (Chapter 92), yet VDR levels are often correlated with cellular proliferation in vivo and in vitro. The level of VDR in primary mouse osteoblastic cell cultures appeared to be regulated in accordance with the rate of cell division [178], suggesting that vitamin D regulation may be modulated by cell cycle–associated changes in VDR content. Mitogenic growth factors such as epidermal growth factor (EGF), insulin, and insulin-like growth factor I (IGF-I), which stimulated the proliferation of NIH-3T3 mouse fibroblastic and MCF-7 human breast cancer cell lines and rat growth plate chondrocytes, also caused significant increases in VDR protein and mRNA levels [179,180]. Furthermore, subcutaneous administration of EGF increased intestinal VDR in the proximal segments of the small intestine but not in kidney of suckling rats [181]. By contrast, in UMR-106 osteoblastic cells, EGF reduced the number of 1,25(OH)2D3 binding sites without altering binding affinity, and attenuated PTH-induced VDR up-regulation [182]. Similar down-regulation of VDR expression by basic fibroblast growth factor (bFGF), EGF, and Ha-ras transformation was observed in mouse NIH3T3 fibroblasts and HC-11 mammary epithelial cells [183,184]. Changes in signaling by mitogenic growth factors may also be the cause of decreases in VDR in confluent cultures of the mouse MMB-1 osteoblastic cell line and
208 in mouse primary osteoblastic cell culture [100,101], although this is not a universal finding [101]. Involvement of growth factors in decreased VDR abundance during cellular differentiation is also likely [185,186], as growth factor deprivation can force cellular quiescence, with restoration of VDR expression upon cell cycle reentry [187]. In this context, although the mechanism is not known, it is interesting to note that apigenin, a flavonoid that induces cellular growth arrest with concomitant inhibition of intracellular signalling and decreased protooncogene expression, potently inhibited VDR mRNA and protein expression in human keratinocytes [188]. Effects of growth factors on VDR abundance therefore appeared to be determined by cell type or tissue origin, by the state of confluence, and by the intracellular signaling response. The differences in the effects of growth factors on VDR levels in proliferation and postconfluent cultures is consistent with a role for 1,25(OH)2D3 in control of cell cycle and progression of cells down a differentiation pathway. However, the mechanisms by which VDR abundance is controlled during cell cycle and the function of 1,25(OH)2D3 in the regulation of cellular differentiation remain as yet unanswered questions. b. Growth Hormone Postmenopausal osteoporosis is an age-related disease that is often associated with intestinal calcium malabsorption [189], and 1,25(OH)2D3 is its main hormonal regulator. Calcium malabsorption can be corrected by estrogen therapy [153] and intestinal absorption of calcium is also increased by growth hormone (GH). GH increased intestinal absorption of calcium in pigs and similarly increased intestinal calcium absorption and expression of the calcium binding protein CaBP 9K, but not plasma concentration of 1,25(OH)2D3, in aged female rats [190,191]. The stimulation of calcium absorption by GH may result, in part, from up-regulation of intestinal VDR [192] and possibly ER [193], which up-regulates intestinal VDR, as noted in Section IV,B,2,b. Although GH also increased intestinal calcium absorption and serum 1,25(OH)2D3 in pig, GH treatment did not change intestinal VDR levels though it did reduce renal VDR in this species [194]. It is still unclear whether the GH effect on VDR occurs at the level of transcription, translation, or message stability, and whether GH effects on VDR are mediated indirectly through IGF-I [191], ER [193], or changes in serum calcium [190]. c. Parathyroid Hormone Parathyroid hormone (PTH) plays a central role in the regulation of calcium homeostasis, in part by regulating the synthesis of the other major calciotropic hormone, 1,25(OH)2D3. PTH is a primary inducer of the 1α-hydroxylation of 25(OH)D3 in kidney to produce 1,25(OH)2D3 [195],
LUIS M. ESTEBAN, JOHN A. EISMAN, AND EDITH M. GARDINER
which, in turn, feeds back to control PTH production and regulate VDR in the parathyroid gland (see Chapters 5 and 30). The PTH/PTHrP receptor and VDR are both present in kidney and osteoblasts. The mechanism of PTH action is mediated through activation of the PKA and PKC signal transduction systems [196–198], both of which regulate VDR levels in vitamin D target cells (discussed in Section IV,B,4). Furthermore, PTH-induced increases in intracellular calcium may also regulate VDR levels. This diversity of PTH second messenger responses leads to differential effects on VDR levels and interactions with the vitamin D response pathway that appear to vary according to the test system. VDR content in the preosteoblastic UMR-106 rat osteogenic sarcoma cell line and the MC3T3-E1 mouse preosteoblastic line was increased by PTH in a time- and dose-dependent manner via cAMP activation [196,198,199]. By contrast, PTH or PTHrP treatment of the more mature rat osteoblastic osteosarcoma cell line ROS17/2.8 led to decreased VDR mRNA levels and 1,25(OH)2D3 binding [200,201]. The differences between these osteoblastic models may relate to the different stages of differentiation, perhaps reflecting cell type–specific variation in factors that regulate VDR levels. In concept, this would be consistent with the observations that a PTH-dependent decrease in VDR in ROS 17/2.8 was mediated by the PKA response [201], whereas a PTH-dependent increase in VDR in rat chondrocytic cultures was mediated by the PKC response pathway [202] (discussed further hereafter). Interactions in vivo are also complex, as PTH inhibited an increase in intestinal VDR levels after 5-day coinfusion of rats with 1,25(OH)2D3 [200]. There was also a partial inhibition of renal VDR levels in this study, accompanied by abolition of CYP24 response to the 1,25(OH)2D3 treatment. Tissue-specific variation in PTH effects on VDR levels presumably relate to the activation of various signaling pathways, depending on the specific target cell type and its state of differentiation (discussed further in the next section). 4. ACTIVATORS OF SECOND MESSENGERS
The various physiological inputs that affect VDR levels involve complex interactions among the intracellular signal transduction pathways. This section briefly addresses the regulation of VDR by the second messengers protein kinase A (PKA) and protein kinase C (PKC). PKA activators include forskolin or cAMP analogs, whereas PKC activators include the phorbol esters, such as phorbol myristate acetate (PMA) and phorbol-12,13-dibutyrate. Each can affect VDR levels in isolation, and the evidence suggests that the VDRmodulating activities of the PKA and PKC pathways
CHAPTER 12 VDR Promoter and Abundance
can be mutually antagonistic. Given their central role in the integration of cellular responses to extracellular molecules, these pathways are key in the mechanisms that regulate VDR levels in target cells. a. Protein Kinase A Early studies showed that cellular response to PTH was via activation of the cAMP-dependent protein kinase A (PKA) pathway, which preceded PTH-induced increases in VDR levels in UMR-106 cells [196]. Subsequent mechanistic studies in NIH3T3 cells confirmed that the adenylate cyclase activator forskolin led to a substantial time-dependent increase in VDR abundance [61]. Similar increases by forskolin have been observed in the preosteoblastic UMR-106 and MC3T3-E1 cell lines [198,199], but not in more mature osteoblastic ROS17/2.8 cells, in which forskolin caused a decrease in VDR level and suppressed homologous up-regulation of VDR [201]. PKA activation by forskolin also increased VDR mRNA in the human HOS-8603 osteosarcoma cell line and a megakaryoblastic leukemia line [110], and a prostaglandin E2-induced increase in VDR in HL-60 human promyelocytic cell line also appeared to be via cAMP [203]. The cAMP activation effect on 1,25(OH)2D3 binding in UMR-106 cells was not affected by the increase of intracellular calcium using a calcium agonist or a calcium ionophore; however, cAMP-mediated activation was reduced by use of calcium channel blockers to lower intracellular calcium [204]. The increase in VDR levels by PKA activation thus appears to involve elevation of intracellular calcium. This is particularly interesting, as intracellular calcium is increased in some cell lines by 1,25(OH)2D3, suggesting a mechanism for synergy between the 1,25(OH)2D3 and PKA responses. b. Protein Kinase C Inhibition of VDR levels in mouse NIH3T3 fibroblastic cells by basic FGF (see Section IV,B,3,a) activated protein kinase C (PKC) [183], consistent with the finding that activation of PKC by phorbol esters down-regulated VDR mRNA and protein levels in a time- and dose-dependent manner. Staurosporine, an inhibitor of PKC, blocked the effects of phorbol esters on VDR levels. Activation of PKC also overcame the stimulation of VDR expression by mitogenic serum factors [183]. Kinetics of the VDR response to PKC activation by phorbol esters was time dependent and complex, however, with a decrease and subsequent increase in VDR mRNA levels in the preosteoblastic UMR-106 rat osteosarcoma cell line [205], but a consistent time- and dose-dependent increase in the ROS17/2.8 line, representing a more mature osteoblastic stage [206]. Further, pretreatment with phorbol esters prior to 1,25(OH)2D3 treatment resulted in a synergistic 10-fold induction of VDR mRNA in the latter cell line [206]. Such synergy in regulation of VDR may explain the functional potentiation
209 of 1,25(OH)2D3 responses by PKC activation in primary rat kidney cell [207] and intestinal epithelial cell cultures [208]. In the latter, induction of the vitamin D target gene CYP24 by pharmacological doses of 1,25(OH)2D3 was potentiated by phorbol ester treatment. Similarly, 1,25(OH)2D3 induced phosphorylation of osteopontin and induced tumorigenic transformation of JB6C141.5a epidermal cells only in the presence of phorbol esters [209]. There appears to be a complex relationship between PKC and VDR actions involving cross-regulation of both molecules [210,211]. Direct modulation of the phosphorylation state of the VDR by PKC has been suggested [212,213], with phosphorylation at serine-51 by PKC altering VDR regulation of transcriptional transactivation [214]. Taken together, these observations suggest that VDR function is regulated by functional interactions between PKC and 1,25(OH)2D3. Intracellular calcium levels may play a mechanistic role in this interaction, as both PKC activation and 1,25(OH)2D3 can increase intracellular calcium. There is evidence to support this concept, as elevation in intracellular calcium by a calcium ionophore decreased VDR levels in NIH3T3 cells in a time-dependent manner, which was further decreased after further PKC activation [183]. c. PKA and PKC Interactions Simultaneous activation of PKC and PKA responses in NIH3T3 cells abolished the PKA-mediated increase in VDR protein and mRNA [61], indicating mutual antagonism between these intracellular signal transduction pathways in the regulation of VDR levels. This concept is supported by subsequent observations that PKC activation by phorbol esters in UMR-106 and ROS 17/2.8 cells blocked a PTH- and forskolin-induced up-regulation of VDR [205] and similar findings in MC3T3-E1 and UMR-106–01 osteosarcoma cells [198]. Thus, the evidence suggests cell type-specific and timedependent effects of PKC and PKA pathway activation on VDR levels, as well as interaction between these pathways. The relative strengths of the PKA and PKC signal transduction responses will, of course, vary as relevant extracellular stimuli change. The net effects of the multiple signaling inputs on VDR abundance will depend on downstream changes in VDR transcription and on VDR posttranslational modification. Changes in VDR transcription may result from activation of transcription cofactors that act on the VDR promoter, such as CREB activation by PKA after PTH stimulation. Posttranslational effects that might affect VDR function could include PKC phosphorylation of VDR, which would alter VDR transactivation activity and could have secondary effects on VDR production by homologous or heterologous pathways. The availability
210 of VDR promoter sequences and cloned promoter regions from several species will facilitate dissection of the mechanisms responsible for second-messenger effects on VDR abundance.
V. CONCLUDING REMARKS A wide range of tissues respond to 1,25(OH)2D3 with a variety of actions including calcium homeostatic changes, regulation of cellular proliferation and differentiation, and regulation of vitamin D metabolism. Abundance of the nuclear VDR is a major determinant of the strength of these 1,25(OH)2D3 responses. Factors that influence VDR content include the hormone 1,25(OH)2D3 itself, dietary calcium, phosphorus and fatty acid content, steroid and retinoid hormones, peptide hormones including growth factors and the major calciotropic hormone PTH, and other activators of secondmessenger systems. The existence of multiple promoters in the human locus and of variant VDR isoforms in human and nonhuman species suggests other mechanisms by which nuclear VDR responses to these regulators may vary. The responses to the range of regulators vary in a species-specific manner and exhibit tissue and cell-type specificity. The mechanisms that provide specificity to VDR regulation are not fully understood but may relate to the state of target cell proliferation or differentiation, cell type–specific differences in intracellular second-messenger responses to a variety of activators, and variation in the complement of downstream nuclear proteins available for the regulation of gene transcription in various target cells. There is still much work needed to elucidate the details of these mechanisms and their importance in the regulation of cell proliferation and differentiation and in the physiological vitamin D response. The complex regulation of VDR expression by this variety of effectors in fulfilling the multiple biological roles of 1,25(OH)2D3, coupled with the increasing understanding of the complex promoter structure and regulation of the VDR locus, make this gene an ideal model for further mechanistic investigation. Importantly, understanding of the regulation of VDR abundance may also have clinical utility, as target cell responses to 1,25(OH)2D3 and vitamin D analog treatment may be limited by levels of functional VDR and thus subject to enhancement by therapeutic increases in VDR availability.
References 1. Brumbaugh PF, Haussler MR 1975 Specific binding of 1α,25-dihydroxycholecalciferol to nuclear components of chick intestine. J Biol Chem 250:1588–1594.
LUIS M. ESTEBAN, JOHN A. EISMAN, AND EDITH M. GARDINER
2. Kream BE, Reynolds RD, Knutson JC, Eisman JA, DeLuca HF 1976 Intestinal cytosol binders of 1,25-dihydroxyvitamin D3 and 25-hydroxyvitamin D3. Arch Biochem Biophys 176:779–787. 3. Kream BE, Yamada S, Schnoes HK, DeLuca HF 1977 Specific cytosol-binding protein for 1,25-dihydroxyvitamin D3 in rat intestine. J Biol Chem 252:4501–4505. 4. Kream BE, Jose M, Yamada S, DeLuca HF 1977 A specific high-affinity binding macromolecule for 1,25-dihydroxyvitamin D3 in fetal bone. Science 197:1086–1088. 5. Christakos S, Norman AW 1979 Studies on the mode of action of calciferol. XVIII. Evidence for a specific high affinity binding protein for 1,25 dihydroxyvitamin D3 in chick kidney and pancreas. Biochem Biophys Res Commun 89:56–63. 6. Wecksler WR, Ross FP, Mason RS, Norman AW 1980 Biochemical properties of the 1α, 25-dihydroxyvitamin D3 cytosol receptors from human and chicken intestinal mucosa. J Clin Endocrinol Metab 50:152–157. 7. Simpson RU, DeLuca HF 1980 Characterization of a receptorlike protein for 1,25-dihydroxyvitamin D3 in rat skin. Proc Natl Acad Sci USA 77:5822–5826. 8. Eisman JA, Martin TJ, MacIntyre I, Moseley JM 1979 1,25-dihydroxyvitamin-D-receptor in breast cancer cells. Lancet 2:1335–1336. 9. Frampton RJ, Suva LJ, Eisman JA, Findlay DM, Moore GE, Moseley JM, Martin TJ 1982 Presence of 1,25-dihydroxyvitamin D3 receptors in established human cancer cell lines in culture. Cancer Res 42:1116–1119. 10. Pike JW, Marion SL, Donaldson CA, Haussler MR 1983 Serum and monoclonal antibodies against the chick intestinal receptor for 1,25-dihydroxyvitamin D3. Generation by a preparation enriched in a 64,000-dalton protein. J Biol Chem 258:1289–1296. 11. Dame MC, Pierce EA, Prahl JM, Hayes CE, DeLuca HF 1986 Monoclonal antibodies to the porcine intestinal receptor for 1,25-dihydroxyvitamin D3: interaction with distinct receptor domains. Biochemistry 25:4523–4534. 12. Tuohimaa P, Blauer M, Jaaskelainen T, Itkonen A, Lindfors M, Mahonen A, Palvimo J, Vilja P, Maenpaa PH 1992 Characterization of human 1,25-dihydroxyvitamin D3 receptor anti-peptide antibodies. J Steroid Biochem Mol Biol 43: 649–657. 13. Lee Y, Inaba M, DeLuca HF, Mellon WS 1989 Immunological identification of 1,25 dihydroxyvitamin D3 receptors in human promyelocytic leukemic cells (HL-60) during homologous regulation. J Biol Chem 264: 13701–13705. 14. McDonnell DP, Mangelsdorf DJ, Pike JW, Haussler MR, O’Malley BW 1987 Molecular cloning of complemetary DNA encoding the avian receptor for vitamin D. Science 235:1214–1217. 15. Berger U, Wilson P, McClelland RA, Colston K, Haussler MR, Pike JW, Coombes RC 1988 Immunocytochemical detection of 1,25-dihydroxyvitamin D receptors in normal human tissues. J Clin Endocrinol Metab 67:607–613. 16. Sandgren ME, Bronnegard M, DeLuca HF 1991 Tissue distribution of the 1,25-dihydroxyvitamin D3 receptor in the male rat. Biochem Biophys Res Commun 181:611–616. 17. Walters MR 1992 Newly identified actions of the vitamin D endocrine system. Endocr Rev 13:719–764. 18. Gascon-Barre M, Demers C, Mirshahi A, Neron S, Zalzal S, Nanci A 2003 The normal liver harbors the vitamin D nuclear receptor in nonparenchymal and biliary epithelial cells. Hepatology 37:1034–1042.
211
CHAPTER 12 VDR Promoter and Abundance
19. Baker AR, McDonnell DP, Hughes M, Crisp TM, Mangelsdorf DJ, Haussler MR, Pike JW, Shine J, O’Malley BW 1988 Cloning and expression of full-length cDNA encoding human vitamin D receptor. Proc Natl Acad Sci USA 85:3294–3298. 20. Burmester JK, Wiese RJ, Maeda N, DeLuca HF 1988 Structure and regulation of the rat 1,25-dihydroxyvitamin D3 receptor. Proc Natl Acad Sci USA 85:9499–9502. 21. Elaroussi MA, Prahl JM, DeLuca HF 1994 The avian vitamin D receptors: primary structures and their origins. Proc Natl Acad Sci USA 91:11596–11600. 22. Lu Z, Hanson K, DeLuca HF 1997 Cloning and origin of the two forms of chicken vitamin D receptor. Arch Biochem Biophys 339:99–106. 23. Li YC, Bergwitz C, Juppner H, Demay MB 1997 Cloning and characterization of the vitamin D receptor from Xenopus laevis. Endocrinology 138:2347–2353. 24. Miyamoto K, Kesterson RA, Yamamoto H, Taketani Y, Nishiwaki E, Tatsumi S, Inoue Y, Morita K, Takeda E, Pike JW 1997 Structural organization of the human vitamin D receptor chromosonal gene and its promoter. Mol Endocrinol 11:1165–1179. 25. Jehan F, DeLuca HF 1997 Cloning and characterization of the mouse vitamin D receptor promoter. Proc Natl Acad Sci USA 94:10138–10143. 26. Kamei Y, Kawada T, Fukuwatari T, Ono T, Kato S, Sugimoto E 1995 Cloning and sequencing of the gene encoding the mouse vitamin D receptor. Gene 152:281–282. 27. Crofts LA, Hancock MS, Morrison NA, Eisman JA 1998 Multiple promoters direct the tissue-specific expression of novel N-terminal variant human vitamin D receptor gene transcripts. Proc Natl Acad Sci USA 95:10529–10534. 28. Sunn KL, Cock TA, Crofts LA, Eisman JA, Gardiner EM 2001 Novel N-terminal variant of human VDR. Mol Endocrinol 15:1599–1609. 29. Gardiner EM, Esteban LM, Fong C, Allison SJ, Flanagan JL, Kouzmenko AP, Eisman JA 2004 Vitamin D receptor B1 and exon 1D: functional and evolutionary analysis. J Steroid Biochem Mol Biol 89–90:233–238. 30. Jehan F, DeLuca HF 2000 The mouse vitamin D receptor is mainly expressed through an Sp1-driven promoter in vivo. Arch Biochem Biophys 377:273–283. 31. Arriza JL, Weinberger C, Cerelli G, Glaser TM, Handelin BL, Housman DE, Evans RM 1987 Cloning of human mineralocorticoid receptor complementary DNA: structural and functional kinship with the glucocorticoid receptor. Science 237:268–275. 32. Kuiper GG, Gustafsson JA 1997 The novel estrogen receptor-β subtype: potential role in the cell- and promoter-specific actions of estrogens and anti-estrogens. FEBS Lett 410:87–90. 33. Tremblay GB, Tremblay A, Copeland NG, Gilbert DJ, Jenkins NA, Labrie F, Giguere V 1997 Cloning, chromosomal localization, and functional analysis of the murine estrogen receptor β. Mol Endocrinol 11:353–365. 34. Lazar MA 1993 Thyroid hormone receptors: multiple forms, multiple possibilities. Endocr Rev 14:184–193. 35. Hoopes CW, Taketo M, Ozato K, Liu Q, Howard TA, Linney E, Seldin MF 1992 Mapping of the mouse Rxr loci encoding nuclear retinoid X receptors RXRα, RXRβ, and RXRγ. Genomics 14:611–617. 36. Giguere V, Lyn S, Yip P, Siu CH, Amin S 1990 Molecular cloning of cDNA encoding a second cellular retinoic acidbinding protein. Proc Natl Acad Sci USA 87:6233–6237. 37. Dreyer C, Krey G, Keller H, Gival F, Helftenbein G, Wahli W 1992 Control of the peroxisomal β-oxidation pathway by
38.
39.
40.
41.
42.
43.
44.
45. 46.
47.
48.
49.
50.
a novel family of nuclear hormone receptors. Cell 68: 879–887. Conneely OM, Kettelberger DM, Tsai MJ, Schrader WT, O’Malley BW 1989 The chicken progesterone receptor A and B isoforms are products of an alternate translation initiation event. J Biol Chem 264:14062–14064. Kastner P, Krust A, Turcotte B, Stropp U, Tora L, Gronemeyer H, Chambon P 1990 Two distinct estrogenregulated promoters generate transcripts encoding the two functionally different human progesterone receptor forms A and B. EMBO J 9:1603–1614. Giangrande PH, McDonnell DP 1999 The A and B isoforms of the human progesterone receptor: two functionally different transcription factors encoded by a single gene. Recent Prog Horm Res 54:291–313. Vegeto E, Shahbaz MM, Wen DX, Goldman ME, O’Malley BW, McDonnell DP 1993 Human progesterone receptor A form is a cell- and promoter-specific repressor of human progesterone receptor B function. Mol Endocrinol 7:1244–1255. Tung L, Mohamed MK, Hoeffler JP, Takimoto GS, Horwitz KB 1993 Antagonist-occupied human progesterone B-receptors activate transcription without binding to progesterone response elements and are dominantly inhibited by A-receptors. Mol Endocrinol 7:1256–1265. McDonnell DP, Shahbaz MM, Vegeto E, Goldman ME 1994 The human progesterone receptor A-form functions as a transcriptional modulator of mineralocorticoid receptor transcriptional activity. J Steroid Biochem Mol Biol 48: 425–432. Suzuki T, Suzuki N, Srivastava AS, Kurokawa T 2000 Identification of cDNAs encoding two subtypes of vitamin D receptor in flounder, Paralichthys olivaceus. Biochem Biophys Res Commun 270:40–45. Kozak M 1996 Interpreting cDNA sequences: some insights from studies on translation. Mamm Genome 7:563–574. Ebihara K, Masuhiro Y, Kitamoto T, Suzawa M, Uematsu Y, Yoshizawa T, Ono T, Harada H, Matsuda K, Hasegawa T, Masushige S, Kato S 1996 Intron retention generates a novel isoform of the murine vitamin D receptor that acts in a dominant negative way on the vitamin D signaling pathway. Mol Cell Biol 16:3393–3400. Saijo T, Ito M, Takeda E, Huq AH, Naito E, Yokota I, Sone T, Pike JW, Kuroda Y 1991 A unique mutation in the vitamin D receptor gene in three Japanese patients with vitamin Ddependent rickets type II: utility of single-strand conformation polymorphism analysis for heterozygous carrier detection. Am J Hum Genet 49:668–673. Arai H, Miyamoto K, Taketani Y, Yamamoto H, Iemori Y, Morita K, Tonai T, Nishisho T, Mori S, Takeda E 1997 A vitamin D receptor gene polymorphism in the translation initiation codon: effect on protein activity and relation to bone mineral density in Japanese women. J Bone Miner Res 12:915–921. Jurutka PW, Remus LS, Whitfield GK, Thompson PD, Hsieh JC, Zitzer H, Tavakkoli P, Galligan MA, Dang HT, Haussler CA, Haussler MR 2000 The polymorphic N terminus in human vitamin D receptor isoforms influences transcriptional activity by modulating interaction with transcription factor IIB. Mol Endocrinol 14:401–420. Flanagan JL, Sunn KL, Leong GM, Fong C, Kouzmenko AP, Eisman JA, Gardiner EM 2002 Enhanced transactivation activity of vitamin D receptor B1 associated with focal nuclear accumulation and cofactor binding. J Bone Miner Res suppl 1 17:S216.
212 51. Byrne IM, Flanagan L, Tenniswood MP, Welsh J 2000 Identification of hormone-responsive promoter immediately upstream of exon 1c in the human vitamin D receptor gene. Endocrinology 141:2829–2836. 52. Lu Z, Jehan F, Zierold C, DeLuca HF 2000 Isolation and characterization of the chicken vitamin D receptor gene and its promoter. J Cell Biochem 77:92–102. 53. Gronemeyer H, Laudet V 1995 Trancription factors 3: nuclear receptors. Protein Profile 2:1173–1308. 54. Lemaire P, Vesque C, Schmitt J, Stunnenberg H, Frank R, Charnay P 1990 The serum-inducible mouse gene Krox-24 encodes a sequence-specific transcriptional activator. Mol Cell Biol 10:3456–3467. 55. Schneider-Maurouny S, Topilko P, Seitandou T, Levi G, Cohen-Tannoudji M, Pournin S, Babinet C, Charnay P 1993 Disruption of Krox-20 results in alteration of rhombomeres 3 and 5 in the developing hindbrain. Cell 75:1199–1214. 56. Topilko P, Schneider-Maunoury S, Levi G, Baron-Van Evercooren A, Chennoufi AB, Seitanidou T, Babinet C, Charnay P 1994 Knox-20 controls myelination in the peripheral nervous system. Nature 371:796–799. 57. Levi G, Topilko P, Schneider-Maunoury S, Lasagna M, Mantero S, Cancedda R, Charnay P 1996 Defective bone formationin Krox-20 mutant mice. Development 122:113–120. 58. Schule R, Umesono K, Mangelsdorf DJ, Bolado J, Pike JW, Evans RM 1990 Jun-Fos and receptors for vitamins A and D recognize a common response element in the human osteocalcin gene. Cell 61:497–504. 59. Qi X, Pramanik R, Wang J, Schultz RM, Maitra RK, Han J, DeLuca HF, Chen G 2002 The p38 and JNK pathways cooperate to trans-activate vitamin D receptor via c-Jun/AP-1 and sensitize human breast cancer cells to vitamin D3-induced growth inhibition. J Biol Chem 277:25884–25892. 60. Merriman HL, van Wijnen AJ, Hiebert S, Bidwell JP, Fey E, Lian J, Stein J, Stein GS 1995 The tissue-specific nuclear matrix protein, NMP-2, is a member of the AML/CBF/ PEBP2/runt domain transcription factor family: interactions with the osteocalcin gene promoter. Biochemistry 34: 13125–13132. 61. Krishnan AV, Feldman D 1992 Cyclic adenosine 3′,5′monophosphate up-regulates 1,25-dihydroxyvitamin D3 receptor gene expression and enhances hormone action. Mol Endocrinol 6:198–206. 62. Lazarova DL, Bordonaro M, Sartorelli AC 2001 Transcriptional regulation of the vitamin D3 receptor gene by ZEB. Cell Growth Differ 12:319–326. 63. Takagi TMH, Kondoh H, Higashi Y 1998 δEF1, a zinc finger and homeodomain transcription factor, is required for skeleton patterning in multiple lineages. Development 125:21–31. 64. Yoshizawa T, Handa Y, Uematsu Y, Takeda S, Sekine K, Yoshihara Y, Kawakami T, Arioka K, Sato H, Uchiyama Y, Masushige S, Fukamizu A, Matsumoto T, Kato S 1997 Mice lacking the vitamin D receptor exhibit impaired bone formation, uterine hypoplasia and growth retardation after weaning. Nat Gene 16:391–396. 65. Wagner KD, Wagner N, Sukhatme VP, Scholz H 2001 Activation of vitamin D receptor by the Wilms’ tumor gene product mediates apoptosis of renal cells. J Am Soc Nephrol 12:1188–1196. 66. Maurer U, Jehan F, Englert C, Hubinger G, Weidmann E, DeLuca HF, Bergmann L 2001 The Wilms’ tumor gene product (WT1) modulates the response to 1,25-dihydroxyvitamin D3 by induction of the vitamin D receptor. J Biol Chem 276: 3727–3732.
LUIS M. ESTEBAN, JOHN A. EISMAN, AND EDITH M. GARDINER
67. Yamamoto H, Miyamoto K, Li B, Taketani Y, Kitano M, Inoue Y, Morita K, Pike JW, Takeda E 1999 The caudalrelated homeodomain protein Cdx-2 regulates vitamin D receptor gene expression in the small intestine. J Bone Miner Res 14:240–247. 68. Arai H, Miyamoto KI, Yoshida M, Yamamoto H, Taketani Y, Morita K, Kubota M, Yoshida S, Ikeda M, Watabe F, Kanemasa Y, Takeda E 2001 The polymorphism in the caudalrelated homeodomain protein Cdx-2 binding element in the human vitamin D receptor gene. J Bone Miner Res 16: 1256–1264. 69. Lee TH, Pelletier J 2001 Functional characterization of WT1 binding sites within the human vitamin D receptor gene promoter. Physiol Genomics 7:187–200. 70. Correa P, Akerstrom G, Westin G 2002 Exclusive underexpression of vitamin D receptor exon 1f transcripts in tumors of primary hyperparathyroidism. Eur J Endocrinol 147:671–675. 71. Wietzke JA, Welsh J 2003 Phytoestrogen regulation of a Vitamin D3 receptor promoter and 1,25-dihydroxyvitamin D3 actions in human breast cancer cells. J Steroid Biochem Mol Biol 84:149–157. 72. Duan C, Clemmons DR 1995 Transcription factor AP-2 regulates human insulin-like growth factor binding protein-5 gene expression. J Biol Chem 270:24844–24851. 73. Johnson W, Jameson JL 1999 AP-2 (activating protein 2) and Sp1 (selective promoter factor 1) regulatory elements play distinct roles in the control of basal activity and cyclic adenosine 3′,5′-monophosphate responsiveness of the human chorionic gonadotropin-β promoter. Mol Endocrinol 13:1963–1975. 74. Whitfield GK, Remus LS, Jurutka PW, Zitzer H, Oza AK, Dang HT, Haussler CA, Galligan MA, Thatcher ML, Encinas-Dominguez C, Haussler MR 2001 Functionally relevant polymorphisms in the human nuclear vitamin D receptor gene. Mol Cell Endocrinol 177:145–159. 75. Masi L, Cimaz R, Simonini G, Bindi G, Stagi S, Gozzini A, Malentacchi C, Brandi ML, Falcini F 2002 Association of low bone mass with vitamin d receptor gene and calcitonin receptor gene polymorphisms in juvenile idiopathic arthritis. J Rheumatol 29:2225–2231. 76. Fischer PR, Thacher TD, Pettifor JM, Jorde LB, Eccleshall TR, Feldman D 2000 Vitamin D receptor polymorphisms and nutritional rickets in Nigerian children. J Bone Miner Res 15:2206–2210. 77. Ruza E, Sotillo E, Sierrasesumaga L, Azcona C, PatinoGarcia A 2003 Analysis of polymorphisms of the vitamin D receptor, estrogen receptor, and collagen Iα1 genes and their relationship with height in children with bone cancer. J Pediatr Hematol Oncol 25:780–786. 78. Vogel A, Strassburg CP, Manns MP 2002 Genetic association of vitamin D receptor polymorphisms with primary biliary cirrhosis and autoimmune hepatitis. Hepatology 35:126–131. 79. Chen WC, Chen HY, Lu HF, Hsu CD, Tsai FJ 2001 Association of the vitamin D receptor gene start codon Fok I polymorphism with calcium oxalate stone disease. BJU Int 87:168–171. 80. Chiu KC, Chuang LM, Yoon C 2001 The vitamin D receptor polymorphism in the translation initiation codon is a risk factor for insulin resistance in glucose tolerant Caucasians. BMC Med Genet 2:2. 81. Kishi S, Yang W, Boureau B, Morand S, Das S, Chen P, Cook EH, Rosner GL, Schuetz EG, Pui CH, Relling MV 2004 Effects of prednisone and genetic polymorphisms on etoposide disposition in children with acute lymphoblastic leukemia. Blood 103:67–72.
CHAPTER 12 VDR Promoter and Abundance
82. Morrison NA, Yeoman R, Kelly PJ, Eisman JA 1992 Contribution of trans-acting factor alleles to normal physiological variability: Vitamin D receptor gene polymorphisms and circulating osteocalcin. Proc Natl Acad Sci USA 89:6665–6669. 83. Faraco JH, Morrison NA, Baker A, Shine J, Frossard PM 1989 Apa I dimorphism at the human vitamin D receptor gene locus. Nucleic Acids Res 17:2150. 84. Ingles SA, Haile RW, Henderson BE, Kolonel LN, Nakaichi G, Shi CY, Yu MC, Ross RK, Coetzee GA 1997 Strength of linkage disequilibrium between two vitamin D receptor markers in five ethnic groups: implications for association studies. Cancer Epidemiol Biomarkers Prev 6:93–98. 85. Morrison NA, Qi JC, Tokita A, Kelly PJ, Crofts L, Nguyen TV, Sambrook PN, Eisman JA 1994 Prediction of bone density from vitamin D receptor alleles. Nature 367: 284–287. 86. Yamagata Z, Miyamura T, Iijima S, Asaka A, Sasaki M, Kato J, Koizumi K 1994 Vitamin D receptor gene polymorphism and bone mineral density in healthy Japanese women. Lancet 344:1027. 87. Ferrari S, Rizzoli R, Chevalley T, Slosman D, Eisman JA, Bonjour JP 1995 Vitamin-D-receptor-gene polymorphisms and change in lumbar-spine bone mineral density. Lancet 345:423–424. 88. Spector TD, Keen RW, Arden NK, Morrison NA, Major PJ, Nguyen TV, Kelly PJ, Baker JR, Sambrook PN, Lanchbury JS, Eisman JA 1995 Influence of vitamin D receptor genotype on bone mineral density in postmenopausal women: a twin study in Britain. Bri Med J 310:1357–1360. 89. Krall EA, Parry P, Lichter JB, Dawson-Hughes B 1995 Vitamin D receptor alleles and rates of bone loss: influences of years since menopause and calcium intake. J Bone Miner Res 10:978–984. 90. Grundberg E, Brandstrom H, Ribom EL, Ljunggren O, Kindmark A, Mallmin H 2003 A poly adenosine repeat in the human vitamin D receptor gene is associated with bone mineral density in young Swedish women. Calcif Tissue Int (in press). 91. Hustmyer FG, Peacock M, Hui S, Johnston CC, Christian J 1994 Bone mineral density in relation to polymorphism at the vitamin D receptor gene locus. J Clin Invest 94: 2130–2134. 92. Melhus H, Kindmark A, Amer S, Wilen B, Lindh E, Ljunghall S 1994 Vitamin D receptor genotypes in osteoporosis. Lancet 344:949. 93. Carling T, Rastad J, Akerstrom G, Westin G 1998 Vitamin D receptor (VDR) and parathyroid hormone messenger ribonucleic acid levels correspond to polymorphic VDR alleles in human parathyroid tumors. J Clin Endocrinol Metab 83:2255–2259. 94. Verbeek W, Gombart AF, Shiohara M, Campbell M, Koeffler HP 1997 Vitamin D receptor: no evidence for allelespecific mRNA stability in cells which are heterozygous for the Taq I restriction enzyme polymorphism. Biochem Biophys Res Commun 238:77–80. 95. Barger-Lux MJ, Heaney RP, Hayes J, DeLuca HF, Johnson ML, Gong G 1995 Vitamin D receptor gene polymorphism, bone mass, body size, and vitamin D receptor density. Calcif Tissue Int 57:161–162. 96. Kinyamu HK, Gallagher JC, Knezetic JA, DeLuca HF, Prahl JM, Lanspa SJ 1997 Effect of vitamin D receptor genotypes on calcium absorption, duodenal vitamin D receptor concentration, and serum 1,25(OH)2D3 dihydroxyvitamin D levels in normal women. Calcif Tissue Int 60:491–495.
213 97. Mocharla H, Butch AW, Pappas AA, Flick JT, Weinstein RS, De Togni P, Jilka RL, Roberson PK, Parfitt AM, SC M 1997 Quantification of vitamin D receptor mRNA by competitive polymerase chain reaction in PBMC: lack of correspondence with common allelic variants. J Bone Miner Res 12:726–733. 98. Gross C, Musiol IM, Eccleshall TR, Malloy PJ, Feldman D 1998 Vitamin D receptor gene polymorphisms: analysis of ligand binding and hormone responsiveness in cultured skin fibroblasts. Biochem Biophys Res Commun 242:467–473. 99. Halloran BP, DeLuca HF 1981 Appearance of the intestinal cytosolic receptor for 1,25-dihydroxyvitamin D3 during neonatal development in the rat. J Biol Chem 256:7338–7342. 100. Walters MR, Rosen DM, Norman AW, Luben RA 1982 1,25-Dihydroxyvitamin D receptors in an established bone cell line. Correlation with biochemical responses. J Biol Chem 257:7481–7484. 101. Chen TL, Li JM, Ye TV, Cone CM, Feldman D 1986 Hormonal responses to 1,25-dihydroxyvitamin D3 in cultured mouse osteoblast-like cells: modulation by changes in receptor level. J Cell Physiol 126:21–28. 102. Dokoh S, Donaldson CA, Haussler MR 1984 Influence of 1,25-dihydroxyvitamin D3 on cultured osteogenic sarcoma cells: correlation with the 1,25-dihydroxyvitamin D3 receptor. Cancer Res 44:2103–2109. 103. Costa EM, Hirst MA, Feldman D 1985 Regulation of 1,25-dihydroxyvitamin D3 receptors by vitamin D analogs in cultured mammalian cells. Endocrinology 117:2203–2210. 104. Pan LC, Price PA 1987 Ligand-dependent regulation of the 1,25-dihydroxyvitamin D3 receptor in rat osteosarcoma cells. J Biol Chem 262:4670–4675. 105. Sher E, Frampton RJ, Eisman JA 1985 Regulation of the 1,25-dihydroxyvitamin D3 receptor by 1,25-dihydroxyvitamin D3 in intact human cancer cells. Endocrinology 116:971–977. 106. Pols HAP, Schilte HP, Visser TJ, Birkenhager JC 1987 Effect of ketoconazole on metabolism and binding of 1,25dihydroxyvitamin D3 by intact rat osteogenic sarcoma cells. Biochim Biophys Acta 931:115–119. 107. Pols HAP, Birkenhager JC, Schilte JP, Visser TJ 1988 Evidence that the self-induced metabolism of 1,25-dihydroxyvitamin D3 limits the homologous up-regulation of its receptor in rat osteosarcoma cells. Biochim Biophys Acta 970:122–129. 108. Dame MC, Pierce EA, DeLuca HF 1985 Identification of the porcine intestinal 1,25-dihydroxyvitamin D3 receptor on sodium dodecyl sulfate/polyacrylamide gels by renaturation and immunoblotting. Proc Natl Acad Sci USA 82: 7825–7829. 109. Mahonen A, Maenpaa PH 1994 Steroid hormone modulation of vitamin D receptor levels in human MG-63 osteosarcoma cells. Biochem Biophys Res Commun 205:1179–1186. 110. Song LN 1996 Demonstration of vitamin D receptor expression in a human megakaryoblastic leukemia cell line: regulation of vitamin D receptor mRNA expression and responsiveness by forskolin. J Steroid Biochem Mol Biol 57:265–274. 111. Bland R, Sammons RL, Sheppard MC, Williams GR 1997 Thyroid hormone, vitamin D and retinoid receptor expression and signalling in primary cultures of rat osteoblastic and immortalised osteosarcoma cells. J Endocrinol 154:63–74. 112. Meyer J, Fullmer CS, Wasserman RH, Komm BS, Haussler MR 1992 Dietary restriction of calcium, phosphorus, and vitamin D elicits differential regulation of the mRNAs for avian intestinal calbindin-D28k and the 1,25-dihydroxyvitamin D3 receptor. J Bone Miner Res 7:441–448.
214 113. Costa EM, Feldman D 1986 Homologous up-regulation of the 1,25(OH)2 vitamin D3 receptor in rats. Biochem Biophys Res Commun 137:742–747. 114. Naveh-Many T, Marx R, Keshet E, Pike JW, Silver J 1990 Regulation of 1,25-dihydroxyvitamin D3 receptor gene expression by 1,25-dihydroxyvitamin D3 in the parathyroid in vivo. J Clin Invest 86:1968–1975. 115. Huang YC, Lee S, Stolz R, Gabrielides C, Pansini-Porta A, Bruns ME, Bruns DE, Miffin TE, Pike JW, Christakos S 1989 Effect of hormones and development on the expression of the rat 1,25-dihydroxyvitamin D3 receptor gene. Comparison with calbindin gene expression. J Biol Chem 264:17454–17461. 116. Strom M, Sandgren ME, Brown TA, DeLuca HF 1989 1,25-dihydroxyvitamin D3 up-regulates the 1,25-dihydroxyvitamin D3 receptor in vivo. Proc Natl Acad Sci USA 86:9770–9773. 117. Sandgren ME, DeLuca HF 1990 Serum calcium and vitamin D regulate 1,25-dihydroxyvitamin D receptor concentration in rat kidney in vivo. Proc Natl Acad Sci USA 87:4312–4314. 118. Zineb R, Zhor B, Odile W, Marthe RR 1998 Distinct, tissuespecific regulation of vitamin D receptor in the intestine, kidney, and skin by dietary calcium and vitamin D. Endocrinology 139:1844–1852. 119. Goff JP, Reinhardt TA, Beckman MJ, Horst RL 1990 Contrasting effects of exogenous 1,25-dihydroxyvitamin D [1,25-(OH)2D] versus endogenous 1,25-(OH)2D, induced by dietary calcium restriction, on vitamin D receptors. Endocrinology 126:1031–1035. 120. Gensure RC, Antrobus SD, Fox J, Okwueze M, Talton SY, Walters MR 1998 Homologous up-regulation of vitamin D receptors is tissue specific in the rat. J Bone Miner Res 13:454–463. 121. Mangelsdorf DJ, Pike JW, Haussler MR 1987 Avian and mammalian receptors for 1,25-dihydroxyvitamin D3: in vitro translation to characterize size and hormone-dependent regulation. Proc Natl Acad Sci USA 84:354–358. 122. Pan P, Reddy K, Lee S, Studzinski GP 1991 Differentiationrelated regulation of 1,25-dihydroxyvitamin D3 receptor mRNA in human leukaemia cells HL-60. Cell Prolif 24:159–170. 123. Costa EM, Feldman D 1987 Measurement of 1,25-dihydroxyvitamin D3 receptor turnover by dense amino acid labeling: changes during receptor up-regulation by vitamin D metabolites. Endocrinology 120:1173–1178. 124. Wiese RJ, Uhland-Smith A, Ross TK, Prahl JM, DeLuca HF 1992 Up-regulation of the vitamin D receptor in response to 1,25-dihydroxyvitamin D3 results from ligand-induced stabilization. J Biol Chem 267:20082–20086. 125. Arbour NC, Prahl JM, DeLuca HF 1993 Stabilization of the vitamin D receptor in rat osteosarcoma cells through the action of 1,25-dihydroxyvitamin D3. Mol Endocrinol 7:1307–1312. 126. Davoodi F, Brenner RV, Evans SR, Schumaker LM, Shabahang M, Nauta RJ, Buras RR 1995 Modulation of vitamin D receptor and estrogen receptor by 1,25(OH)2vitamin D3 in T-47D human breast cancer cells. J Steroid Biochem Mol Biol 54:147–153. 127. Healy KD, Zella JB, Prahl JM, DeLuca HF 2003 Regulation of the murine renal vitamin D receptor by 1,25-dihydroxyvitamin D3 and calcium. Proc Natl Acad Sci USA 100: 9733–9737. 128. Uhland-Smith A, DeLuca HF 1993 The necessity for calcium for increased renal vitamin D receptor in response to 1,25dihydroxyvitamin D. Biochim Biophys Acta 1176:321–326.
LUIS M. ESTEBAN, JOHN A. EISMAN, AND EDITH M. GARDINER
129. Beckman MJ, DeLuca HF 2002 Regulation of renal vitamin D receptor is an important determinant of 1α,25-dihydroxyvitamin D3 levels in vivo. Arch Biochem Biophys 401: 44–52. 130. Brown AJ, Zhong M, Finch J, Ritter C, Slatopolsky E 1995 The roles of calcium and 1,25-dihydroxyvitamin D3 in the regulation of vitamin D receptor expression by rat parathyroid glands. Endocrinology 136:1419–1424. 131. Zhu Y, Goff JP, Reinhardt TA, Horst RL 1997 Pregnancy and lactation increase vitamin D–dependent intestinal membrane calcium adenosine triphosphatase and calcium binding protein messenger ribonucleic acid expression. Endocrinology 139:3520–3524. 132. Sriussadaporn S, Wong MS, Pike JW, Favus MJ 1995 Tissue specificity and mechanism of vitamin D receptor up-regulation during dietary phosphorus restriction in the rat. J Bone Miner Res 10:271–280. 133. Hernandez A, Concepcion MT, Rodriguez M, Salido E, Torres A 1996 High phosphorus diet increases preproPTH mRNA independent of calcium and calcitriol in normal rats. Kidney Int 50:1872–1878. 134. Jump DB, Clarke SD, Thelen A, Liimatta M, Ren B, Badin MV 1997 Dietary fat, genes, and human health. Adv Exp Med Biol 422:167–176. 135. Leonard F, Haag M, Kruger MC 2001 Modulation of intestinal vitamin D receptor availability and calcium ATPase activity by essential fatty acids. Prostaglandins Leukot Essent Fatty Acids 64:147–150. 136. Gaschott T, Stein J 2003 Short-chain fatty acids and colon cancer cells: the vitamin D receptor—butyrate connection. Recent Results Cancer Res 164:247–257. 137. Gaschott T, Wachtershauser A, Steinhilber D, Stein J 2001 1,25-Dihydroxycholecalciferol enhances butyrate-induced p21(Waf1/Cip1) expression. Biochem Biophys Res Commun 283:80–85. 138. Costa EM, Feldman D 1987 Modulation of 1,25-dihydroxyvitamin D3 receptor binding and action by sodium butyrate in cultured pig kidney cells (LLC-PK1). J Bone Miner Res 2:151–159. 139. Maiyar AC, Norman AW 1992 Effects of sodium butyrate on 1,25-dihydroxyvitamin D3 receptor activity in primary chick kidney cells. Mol Cell Endocrinol 84:99–107. 140. Kim CH, Cheng SL, Kim GS 1999 Effects of dexamethasone on proliferation, activity, and cytokine secretion of normal human bone marrow stromal cells: possible mechanisms of glucocorticoid-induced bone loss. J Endocrinol 162: 371–379. 141. Massaro ER, Simpson RU, DeLuca HF 1983 Glucocorticoids and appearance of 1,25-dihydroxyvitamin D3 receptor in rat intestine. Am J Physiol 244:E230–E235. 142. Lee S, Szlachetka M, Christakos S 1991 Effect of glucocorticoids and 1,25-dihydroxyvitamin D3 on the developmental expression of the rat intestinal vitamin D receptor gene. Endocrinology 129:396–401. 143. Gensure RC, Fish JM, Walters MR 1993 Dexamethasone downregulates vitamin D receptors in rat kidney, unmasking a high affinity binding site. Biochem Biophys Res Commun 195:1139–1144. 144. Hirst M, Feldman D 1982 Glucocorticoids down-regulate the number of 1,25-dihydroxyvitamin D3 receptors in mouse intestine. Biochem Biophys Res Commun 105:1590–1596. 145. Wilhelm F, Norman AW 1985 Influence of triamcinolone, estradiol-17β and testosterone on 1,25-dihydroxyvitamin D3 binding performances to its chick intestinal receptor. J Steroid Biochem 23:913–918.
CHAPTER 12 VDR Promoter and Abundance
146. Manolagas SC, Anderson DC, Lumb GA 1979 Glucocorticoids regulate the concentration of 1,25-dihydroxycholecalciferol receptors in bone. Nature 277:314–315. 147. Manolagas SC, Abare J, Deftos LJ 1984 Glucocorticoids increase the 1,25(OH)2D3 receptor concentration in rat osteogenic sarcoma cells. Calcif Tissue Int 36:153–157. 148. Chen TL, Cone CM, Morey-Hilton E, Feldman D 1983 1α,25-dihydroxyvitamin D3 receptors in cultured rat osteoblast-like cells. Glucocorticoid treatment increases receptor content. J Biol Chem 258:4350–4355. 149. Chen TL, Cone CM, Morey-Holton E, Feldman D 1982 Glucocorticoid regulation of 1,25(OH)2-vitamin D3 receptors in cultured mouse bone cells. J Biol Chem 257:13564–13569. 150. Godschalk M, Levy JR, Downs RW 1992 Glucocorticoids decrease vitamin D receptor number and gene expression in human osteosarcoma cells. J Bone Miner Res 7:21–27. 151. Midorikawa K, Sayama K, Shirakata Y, Hanakawa Y, Sun L, Hashimoto K 1999 Expression of vitamin D receptor in cultured human keratinocytes and fibroblasts is not altered by corticosteroids. J Dermatol Sci 21:8–12. 152. Saunders DE, Christensen C, Williams JR, Wappler NL, Lawrence WD, Malone JM, Malviya VK, Deppe G 1995 Inhibition of breast and ovarian carcinoma cell growth by 1,25-dihydroxyvitamin D3 combined with retinoic acid or dexamethasone. Anticancer Drugs 6:562–569. 153. Gallagher JC, Riggs BL, Eisman JA, Hamstra A, Arnaud SB, DeLuca HF 1979 Intestinal calcium absorption and serum vitamin D metabolites in normal subjects and osteoporotic patients. J Clin Invest 64:729–736. 154. Gallagher JC, Riggs BL, DeLuca HF 1980 Effect of estrogen on calcium absorption and serum vitamin D metabolites in postmenopausal osteoporosis. J Clin Endocrinol Metab 51:1359–1364. 155. Ten Bolscher M, Netelenbos JC, Barto R, Van Buuren LM, Van der vijgh WJ 1999 Estrogen regulation of intestinal calcium absorption in the intact and ovariectomized adult rat. J Bone Miner Res 14:1197–1202. 156. Liel Y, Shany S, Smirnoff P, Schwartz B 1999 Estrogen increases 1,25-dihydroxyvitamin D receptors expression and bioresponse in the rat duodenal mucosa. Endocrinology 140:280–285. 157. Colin EM, Van Den Bemd GJ, Van Aken M, Christakos S, De Jonge HR, DeLuca HF, Prahl JM, Birkenhager JC, Buurman CJ, Pols HA, Van Leeuwen JP 1999 Evidence for involvement of 17β-estradiol in intestinal calcium absorption independent of 1,25-dihydroxyvitamin D3 level in the Rat. J Bone Miner Res 14:57–64. 158. Schwartz B, Smirnoff P, Shany S, Y L 2000 Estrogen controls expression and bioresponse of 1,25-dihydroxyvitamin D receptors in the rat colon. Mol Cell Biochem 203:87–93. 159. Lechner D, Cross HS 2003 Phytoestrogens and 17β-estradiol influence vitamin D metabolism and receptor expression— relevance for colon cancer prevention. Recent Results Cancer Res 164:379–391. 160. Levy J, Zuili I, Yankowitz N, Shany S 1984 Induction of cytosolic receptors for 1α,25-dihydroxyvitamin D3 in the immature rat uterus by oestradiol. J Endocrinol 100:265–269. 161. Duncan WE, Glass AR, Wray HL 1991 Estrogen regulation of the nuclear 1,25-dihydroxyvitamin D3 receptor in rat liver and kidney. Endocrinology 129:2318–2324. 162. Duque G, Abdaimi KE, Macoritto M, Miller MM, Kremer R 2002 Estrogens (E2) regulate expression and response of 1,25-dihydroxyvitamin D3 receptors in bone cells: changes with aging and hormone deprivation. Biochem Biophys Res Commun 299:446–454.
215 163. Liel Y, Kraus S, Levy J, Shany S 1992 Evidence that estrogens modulate activity and increase the number of 1,25-dihydroxyvitamin D receptors in osteoblast-like cells (ROS 17/2.8). Endocrinology 130:2597–2601. 164. Ishibe M, Nojima T, Ishibashi T, Koda T, Kaneda K, Rosier RN, Puzas JE 1995 17β-estradiol increases the receptor number and modulates the action of 1,25-dihydroxyvitamin D3 in human osteosarcoma-derived osteoblast-like cells. Calcif Tissue Int 57:430–435. 165. Kinuta K, Tanaka H, Moriwake T, Aya K, Kato S, Seino Y 2000 Vitamin D is an important factor in estrogen biosynthesis of both female and male gonads. Endocrinology 141:1317–1324. 166. Allenby G, Bocquel MT, Saunders M, Kazmer S, Speck J, Rosenberger M, Lovey A, Kastner P, Grippo JF, Chambon P, Levin A 1993 Retinoic acid receptors and retinoid X receptors: interactions with endogenous retinoic acids. Proc Natl Acad Sci USA 90:30–34. 167. Yu VC, Delsert C, Andersen B, Holloway JM, Devary OV, Naar AM, Kim SY, Boutin JM, Glass CK, Rosenfeld MG 1991 RXRβ: a coregulator that enhances binding of retinoic acid, thyroid hormone, and vitamin D receptors to their cognate response elements. Cell 67:1251–1266. 168. Kliewer SA, Umesono K, Manglesdorf DJ, Evans RM 1992 Retinoid X receptor interacts with nuclear receptors in retinoic acid, thyroid hormone and vitamin D3 signalling. Nature 355:446–449. 169. Navia JM, Harris SS 1980 Vitamin A influence on calcium metabolism and calcification. Ann N Y Acad Sci 355:45–57. 170. Bikle DD, Pillai S 1993 Vitamin D, calcium, and epidermal differentiation. Endocr Rev 14:3–19. 171. Petkovich PM, Heersche JNM, Tinker DO, Jones G 1984 Retinoic acid stimulates 1,25-dihydroxyvitamin D3 binding in rat osteosarcoma cells. J Biol Chem 259:8274–8280. 172. Lee KL, Petkovich PM, Heersche JN 1988 The effects of sodium butyrate on the retinoic acid-induced changes in 1,25-dihydroxyvitamin D3 receptors in tumorigenic and nontumorigenic bone derived cell lines. Endocrinology 122:2399–2406. 173. Chen TL, Feldman D 1985 Retinoic acid modullation of 1,25(OH)2 vitamin D3 receptors and bioresponse in bone cells: species differences between rat and mouse. Biochem Biophys Res Commun 132:74–80. 174. Dore BT, Uskokovic MR, Momparler RL 1993 Interaction of retinoic acid and vitamin D3 analogs on HL-60 myeloid leukemic cells. Leuk Res 17:749–757. 175. Brown G, Bunce CM, Rowlands DC, Williams GR 1994 Alltrans retinoic acid and 1α,25-dihydroxyvitamin D3 co-operate to promote differentiation of the human promyeloid leukemia cell line HL60 to monocytes. Leukemia 8:806–815. 176. Gibson DF, Bikle DD, Harris J 1998 All-trans retinoic acid blocks the antiproliferative prodifferentiating actions of 1,25dihydroxyvitamin D3 in normal human keratinocytes. J Cell Physiol 174:1–8. 177. Mahonen A, Pirskanen A, Maenpaa PH 1991 Homologous and heterologous regulation of 1,25-dihydroxyvitamin D-3 receptor mRNA levels in human osteosarcoma cells. Biochim Biophys Acta 1088:111–118. 178. Chen TL, Feldman D 1981 Regulation of 1,25-dihydroxyvitamin D3 receptors in cultured mouse bone cells. Correlation of receptor concentration with the rate of cell division. J Biol Chem 256:5561–5566. 179. Krishnan AV, Feldman D 1991 Stimulation of 1,25-dihydroxyvitamin D3 receptor gene expression in cultured cells by serum and growth factors. J Bone Miner Res 6:1099–1107.
216 180. Klaus G, Weber L, Rodriguez J, Fernandez P, Klein T, Grulich-Henn J, Hugel U, Ritz E, Mehls P 1998 Interaction of IGF-I and 1α,25(OH)2D3 on receptor expression and growth stimulation in rat growth plate chondrocytes. Kidney Int 53:1152–1161. 181. Bruns DE, Krishnan AV, Feldman D, Gray RW, Christakos S, Hirsch GN, Bruns ME 1989 Epidermal growth factor increases intestinal calbindin-D9k and 1,25-dihydroxyvitamin D receptors in neonatal rats. Endocrinology 125:478–485. 182. van Leeuwen JPTM, Pols HAP, Schilte JP, Visser TJ, Birkenhager JC 1991 Modulation by epidermal growth factor of the basal 1,25(OH)2D3 receptor level and the heterologous up-regulation of the 1,25(OH)2D3 receptor in clonal osteoblast-like cells. Calcif Tissue Int 49:35–42. 183. Krishnan AV, Feldman D 1991 Activation of protein kinase-C inhibits vitamin D receptor gene expression. Mol Endocrinol 5:605–612. 184. Escaleira MT, Brentani MM 1999 Vitamin D3 receptor (VDR) expression in HC-11 mammary cells: regulation by growthmodulatory agents, differentiation, and Ha-ras transformation. Breast Cancer Res Treat 54:123–133. 185. Zhao X, Feldman D 1993 Regulation of vitamin D receptor abundance and responsiveness during differentiation of HT-29 human colon cancer cells. Endocrinology 132:1808–1814. 186. Hewison M, Freeman L, Hughes SV, Evans KN, Bland R, Eliopoulos AG, Kilby MD, Moss PAH, Chakraverty R 2003 Differential regulation of vitamin D receptor and its ligand in human monocyte-derived dendritic cells. J Immunol 170: 5382–5390. 187. Segaert S, Degreef H, Bouillon R 2000 Vitamin D receptor expression is linked to cell cycle control in normal human keratinocytes. Biochem Biophys Res Commun 279:89–94. 188. Segaert S, Courtois S, Garmyn M, Degreef H, Bouillon R 2000 The flavonoid apigenin suppresses vitamin D receptor expression and vitamin D responsiveness in normal human keratinocytes. Biochem Biophys Res Commun 268:237–241. 189. Gallagher JC 1990 The pathogenesis of osteoporosis. Bone Miner 9:215–227. 190. Denis I, Thomasset M, Pointillart A 1994 Influence of exogenous porcine growth hormone on vitamin D metabolism and calcium and phosphorus absorption in intact pigs. Calcif Tissue Int 54:489–492. 191. Fleet JC, Bruns ME, Hock JM, Wood RJ 1994 Growth hormone and parathyroid hormone stimulate intestinal calcium absorption in aged female rats. Endocrinology 134: 1755–1760. 192. Chen C, Noland KA, Kalu DN 1997 Modulation of intestinal vitamin D receptor by ovariectomy, estrogen and growth hormone. Mech Ageing Dev 99:109–122. 193. Chen C, Kalu DN 1998 Modulation of intestinal estrogen receptor by ovariectomy, estrogen and growth hormone. J Pharmacol Exp Ther 286:328–333. 194. Goff JP, Caperna TJ, Steele NC 1990 Effects of growth hormone administration on vitamin D metabolism and vitamin D receptors in the pig. Domest Anim Endocrinol 7:425–433. 195. Garabedian M, Holick MF, DeLuca HF, Boyle IT 1972 Control of 25-hydroxycholecalciferol metabolism by parathyroid glands. Proc Natl Acad Sci USA 69:1673–1676. 196. Pols HAP, van Leeuwen JP, Schilte JP, Visser TJ, Birkenhager JC 1988 Heterologous up-regulation of the 1,25-dihydroxyvitamin D3 receptor by parathyroid hormone (PTH) and PTH-like peptide in osteoblast-like cells. Biochem Biophys Res Commun 156:588–594. 197. Chakravarthy BR, Durkin JP, Rixon RH, Whitfield JF 1990 Parathyroid hormone fragment [3-34] stimulates protein kinase C (PKC) activity in rat osteosarcoma and murine
LUIS M. ESTEBAN, JOHN A. EISMAN, AND EDITH M. GARDINER
198.
199.
200.
201.
202.
203.
204.
205.
206.
207.
208.
209.
210. 211.
T-lymphoma cells. Biochem Biophys Res Commun 171: 1105–1110. Krishnan AV, Cramer SD, Bringhurst FR, Feldman D 1995 Regulation of 1,25-dihydroxyvitamin D3 receptors by parathyroid hormone in osteoblastic cells: role of second messenger pathways. Endocrinology 136:705–712. van Leeuwen JPTM, Birkenhager JC, Vink-van Wijngaarden T, van den Bemd GJCM, Pols HAP 1992 Regulation of 1,25dihydroxyvitamin D3 receptor gene expression by parathyroid hormone and cAMP-agonists. Biochem Biophys Res Commun 185:881–886. Reinhardt TA, Horst RL 1990 Parathyroid hormone downregulates 1,25-dihydroxyvitamin D receptors (VDR) and VDR messenger ribonucleic acid in vitro and blocks homologous up-regulation of VDR in vivo. Endocrinology 127:942–948. Sriussadaporn S, Wong MS, Whitfield JF, Tembe V, Favus MJ 1995 Structure–function relationship of human parathyroid hormone in the regulation of vitamin D receptor expression in osteoblast-like cells (ROS 17/2.8). Endocrinology 136: 3735–3742. Klaus G, von Eichel B, May T, Hugel U, Mayer H, Ritz E, Mehls O 1994 Synergistic effects of parathyroid hormone and 1,25-dihydroxyvitamin D3 on proliferation and vitamin D receptor expression of rat growth cartilage cells. Endocrinology 135:1307–1315. Smith SJ, Green LM, Hayes ME, Mawer EB 1999 Prostaglandin E2 regulates vitamin D receptor expression, vitamin D-24-hydroxylase activity and cell proliferation in an adherent human myeloid leukemia cell line (Ad-HL60). Prostaglandins Other Lipid Mediat 57:73–85. van Leeuwen JPTM, Birkenhager JC, Schilte JP, Buurman CJ, Pols HAP 1990 Role of calcium and cAMP in heterologous up-regulation of the 1,25-dihydroxyvitamin D3 receptor in an osteoblast cell line. Cell Calcium 11:281–289. van Leeuwen JPTM, Birkenhager JC, Buurman CJ, van den Bemd GJCM, Bos MP, Pols HAP 1992 Bidirectional regulation of the 1,25-dihydroxyvitamin D3 receptor by phorbol esteractivated protein kinase-C in osteoblast-like cells: interaction with adenosine 3′,5′-monophosphate-induced up-regulation of the 1,25-dihydroxyvitamin D3 receptor. Endocrinology 130:2259–2266. Reinhardt TA, Horst RL 1994 Phorbol 12-myristate 13-acetate and 1,25-dihydroxyvitamin D3 regulate 1,25-dihydroxyvitamin D3 receptors synergistically in rat osteosarcoma cells. Mol Cell Endocrinol 101:159–165. Chen ML, Boltz MA, Armbrecht HJ 1993 Effects of 1,25dihydroxyvitamin D3 and phorbol ester on 25-hydroxyvitamin D3 24-hydroxylase cytochrome P450 messenger ribonucleic acid levels in primary cultures of rat renal cells. Endocrinology 132:1782–1788. Armbrecht HJ, Hodam TL, Boltz MA, Chen ML 1993 Phorbol ester markedly increases the sensitivity of intestinal epithelial cells to 1,25-dihydroxyvitamin D3. FEBS Lett 327:13–16. Chang PL, Prince CW 1993 1α,25-Dihydroxyvitamin D3 enhances 12-O-tetradecanoylphorbol-13-acetate-induced tumorigenic transformation and osteopontin expression in mouse JB6 epidermal cells. Cancer Res 53:2217–2220. Martell RE, Simpson RU, Taylor JM 1987 1,25-Dihydroxyvitamin D3 regulation of phorbol ester receptors in HL-60 leukemia cells. J Biol Chem 262:5570–5575. van Leeuwen JPTM, Birkenhager JC, M vdBGJC, Buurman CJ, Staal A, Bos M, Pols HAP 1992 Evidence for the functional involvement of protein kinase C in the action of 1,25-dihydroxyvitamin D3 in bone. J Biol Chem 267:12562–12569.
CHAPTER 12 VDR Promoter and Abundance
212. Desai RK, van Wijnen AJ, Stein JL, Stein GS, Lian JB 1995 Control of 1,25-dihydroxyvitamin D3 receptor–mediated enhancement of osteocalcin gene transcription: effects of perturbing phosphorylation pathways by okadaic acid and staurosporine. Endocrinology 136:5685–5693. 213. Hara H, Yasunami Y, Adachi T 2002 Alteration of cellular phosphorylation state affects vitamin D receptor–mediated
217 CYP3A4 mRNA induction in Caco-2 cells. Biochem Biophys Res Commun 296:182–188. 214. Hsieh JC, Jurutka PW, Nakajima S, Galligan MA, Haussler CA, Shimizu Y, Shimizu N, Whitfield GK, Haussler MR 1993 Phosphorylation of the human vitamin D receptor by protein kinase C. Biochemical and functional evaluation of the serine 51 recognition site. J Biol Chem 268:15118–15126.
CHAPTER 13
Nuclear Vitamin D Receptor: Structure-Function, Molecular Control of Gene Transcription, and Novel Bioactions G. KERR WHITFIELD, PETER W. JURUTKA, CAROL A. HAUSSLER, JUI-CHENG HSIEH, THOMAS K. BARTHEL, ELIZABETH T. JACOBS,* CARLOS ENCINAS DOMINGUEZ, MICHELLE L. THATCHER, AND MARK R. HAUSSLER Department of Biochemistry and Molecular Biophysics, and *Arizona Cancer Center; College of Medicine, University of Arizona, Tucson, Arizona I. Introduction II. Gene Targets and Biological Actions of the Vitamin D Receptor III. The Vitamin D Receptor as a Member of the Nuclear Receptor Superfamily IV. Structure-Function of the Vitamin D Receptor
V. Mechanisms of Vitamin D Receptor-Mediated Control of Gene Expression VI. Implications of Vitamin D Receptor-Mediated Signaling for Human Health and Disease VII. Summary and Perspectives References
I. INTRODUCTION
found to be a 3.3–3.7S protein (depending on species) that bound 1,25(OH)2D3 with high affinity (Kd = 0.1–1 nM) [8,9] and displayed a pharmacologic profile for binding various D metabolites and analogs that is consistent with their relative biological activities [10,11]. Avian intestinal VDR was shown to be a DNA-binding protein, and this property was exploited in the first purification of VDR [12]. An enriched avian VDR preparation, containing one major and several minor proteins of molecular mass 50–70 kDa [12,13], was then utilized to generate monoclonal antibodies to the VDR that cross-reacted with all species of vitamin D receptor tested [14]. Finally, via monoclonal antibody-based screening of a λGT-11 expression vector library, a partial cDNA clone of avian VDR was obtained [15], revealing the presence of the classical (Cys2–Cys2)2 zinc finger motif of the chromatinbound steroid hormone receptors [16]. Independent biochemical evidence demonstrated the importance of reduced cysteine thiol groups for DNA association [17] and implicated zinc in maintaining the DNA binding integrity of VDR [18]. Subsequently, fulllength human [19], rat [20], and avian [21] VDRs were cloned and their sequences deduced, allowing for overexpression and functional studies of the type discussed in this chapter and elsewhere in this volume. It is notable that the cloning and classification of VDR
A. Historical Aspects and Overview of the Vitamin D Receptor The existence of the vitamin D receptor (VDR) was originally appreciated when it was observed that administration of radioactive vitamin D3 at physiologic doses to rachitic chickens elicited the selective nuclear localization of a bioactive metabolite of the parent vitamin in the target intestine [1]. The binding of this D-metabolite in the nucleus was specific, saturable, and confined to the deoxynucleoprotein chromatin subfraction [1]. Subsequent salt extraction experiments resulted in the discovery of the chromosomal protein that comprises the vitamin D receptor [2]. The vitamin D metabolite associated with VDR was shown to be more polar than 25-hydroxyvitamin D3 (25(OH)D3) [3,4] and at least five times as potent and more rapidly acting than either vitamin D3 or 25(OH)D3 [4]. These results set the stage for the chemical identification [5,6] of the VDR ligand as 1α,25-dihydroxyvitamin D3 (1,25(OH)2D3) (Fig. 1), a metabolite produced mainly in the kidney [7]. When isotopically labeled 1,25(OH)2D3 could be produced in quantity, biochemical and pharmacological characterization of VDR was accomplished. VDR was VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
Copyright © 2005, Elsevier, Inc. All rights reserved.
220
G. KERR WHITFIELD
VDR
ET AL .
RXR OH
HO
OH
1α,25-Dihydroxyvitamin D3
COOH
9-cis-Retinoic acid
FIGURE 1 Chemical structures of the classic natural ligands for the protein subunits of the VDR–RXR heterodimer. 1α,25(OH)2D3 is a high-affinity ligand for VDR (Kd < 1 nM) [8,9]. The vitamin A metabolite 9-cis-retinoic acid is a natural ligand for RXR with a Kd of approximately 10 nM [26].
as a DNA-binding transcription factor activated by ligand not only validated the discovery of VDR as a chromosomal protein [2], but also verified the initial report of its uncharacterized hormonal ligand localized in nuclear chromatin [1].
a heterodimer of the 1,25(OH)2D3-liganded VDR and unoccupied RXR.
B. The Vitamin D Receptor Signals as a Binary Complex with Retinoid X Receptor
A. Vitamin D Metabolism and Actions in Target Tissues
More recent research on VDR included a key demonstration [22–24] that high-affinity and specific DNA binding by VDR requires the presence of a co-receptor, now known to be any of the three retinoid X receptor (RXR) isoforms found in human tissues [25]. Furthermore, it was demonstrated that VDR-RXR heterodimer activation of transcription occurs on a DNA responsive element in the natural context of the rat osteocalcin promoter in intact bone cells, and that 9-cis-retinoic acid (9-cis RA) (Fig. 1), one of the natural ligands of the RXR partner [26], operates, at least in the context of this promoter, to suppress VDR-mediated stimulation of transcription by 1,25(OH)2D3 [27]. Therefore, RXR is the functional partner of VDR in a binary heterocomplex that is switched positively by the vitamin D hormone and negatively by 9-cis RA or a synthetic RXR-specific ligand [27–32]. Apparently, 9-cis RA attenuates 1,25(OH)2D3 responsiveness by diverting RXRs away from VDR-mediated transcription toward RXR homodimers and other RXR-dependent transcriptional pathways [33,34] (see Section V,A). From these and other studies with different vitamin D responsive genes (reviewed by Haussler et al. [35]), it appears that the active signal transduction complex that mediates the biological effects of vitamin D is
II. GENE TARGETS AND BIOLOGICAL ACTIONS OF THE VITAMIN D RECEPTOR
An overview of the formation of 1,25(OH)2D3 and its actions in vertebrates with a mineralized skeleton is presented in Fig. 2. The hormonal precursor and parent compound, vitamin D3, can either be obtained in the diet or formed from 7-dehydrocholesterol in skin via a nonenzymatic, UV light-dependent reaction. Vitamin D3 is then transported to the liver, where it is hydroxylated at the 25 position of the side chain to produce 25(OH)D3, the major circulating form of vitamin D3. The final step in the production of the hormonal form occurs mainly in the kidney, via a tightly regulated 1α-hydroxylation reaction (Fig. 2). The cytochrome P450-containing (CYP) enzymes that catalyze 25- and 1α-hydroxylations are mitochondrial CYP27A1 and CYP27B1, respectively, as well as a newly characterized microsomal 25-hydroxylase, CYP2R1 [36]. These reactions are discussed in more detail in Chapters 4–5 of this volume. As depicted in Fig. 2, 1,25(OH)2D3 can circulate to various target tissues to exert its actions which, in the conception of the authors of this chapter, are largely mediated by the nuclear VDR and its RXR heterodimeric partner. Many of the long-recognized functions of 1,25(OH)2D3 involve the regulation of calcium and phosphate metabolism, raising the blood levels of these ions to facilitate bone mineralization. Thus, in the small
221
CHAPTER 13 Nuclear Vitamin D Receptor
1,25 VDR Skin
Macrophage
Involucrin
Liver CYP27A1/2R1
Immune system
24-hydroxylated D Metabolites
Circulating 25(OH)D3
VDR Circulating 1α,25(OH)2D3
25(OH)D3 1α,25(OH)2D3 CYP27B1 −
PTH
Diet
1,25 Carbonic VDR anhydrase
1,25
CYP24
Differentiation
Vitamin D3
Kidney
Vitamin D catabolism
Keratinocyte
Ca2+
1,25
To nuclear vitamin D receptor
PO3− 4
VDR
1,25 +
VDR
Blood Ca•PO4
Parathyroid
Absorption Other 1,25(OH)2D3VDR target systems: • Central nervous • Endocrine (thyroid, pancreas, etc.) • Reproductive tract • Muscle • Some tumors
Osteoclast precursor
Reabsorption
Ca2+
1,25 VDR
Mineralization of collagen
PO3− 4 Small intestine
Bone Osteoblast
1,25 VDR
Resorption OPG decoy − 1,25 receptor RANKL VDR
+ IL-6
Osteoblast
+
RANK RANKL
Low Ca2+
PTH Synthesis − 1,25 VDR
c-fms
M-C
SF
Osteocalcin Osteopontin + Cell differentiation and fusion
PO3− 4 Ca2+ Osteoclast
FIGURE 2 Vitamin D metabolism and its biologic actions via the nuclear VDR. Depicted at the upper left are the sources of vitamin D3, which are either dietary or via a UV light-mediated photolysis of 7-dehydrocholesterol in the epidermis (see Chapter 3). The central bar (shaded) depicts the enzymatic conversion of vitamin D3 to the 25-hydroxy and then to the 1α,25-dihydroxy derivative (1α,25(OH)2D3), which is the active, hormonal form. These and subsequent hydroxylation reactions are catalyzed by CYPs (see text). The 1,25(OH)2D3 hormonal form (shown as 1,25 inside a white semicircle) then circulates to the various target tissues (selected target tissues identified in italics), where it binds to the nuclear VDR (depicted in white letters inside a black semicircle). Examples of target genes that are either transcriptionally activated (no symbol shown) or repressed (– symbol in a circle) are listed inside boxes.
intestine, 1,25(OH)2D3 binds to VDR, eliciting regulation of genes involved in calcium and phosphate absorption. In the kidney, 1,25(OH)2D3 conserves blood calcium and phosphate by promoting their reabsorption from the glomerular filtrate. The parathyroid gland also expresses VDR, and when the receptor is liganded with 1,25(OH)2D3, parathyroid hormone (PTH) synthesis is suppressed by a direct action on gene transcription [37]. This negative feedback loop, which limits the stimulation of CYP27B1 by PTH under low calcium conditions (Fig. 2), serves to curb the bone-resorbing effects of PTH in anticipation of 1,25(OH)2D3-mediated increases in both intestinal calcium absorption and bone resorption, thus preventing hypercalcemia. The direct boneresorbing effects of 1,25(OH)2D3 are mediated first in osteoblasts, where the liganded VDR promotes transcription of the gene encoding receptor activator of NF-κB ligand (RANKL) [38]. RANKL, displayed on the surface of osteoblasts, is detected by a cognate receptor
(RANK) on osteoclast precursor cells to promote their differentiation into bone-resorbing osteoclasts. This process also requires the liganding of c-fms, an osteoclast precursor surface receptor for macrophage colony stimulating factor (M-CSF). As illustrated in Fig. 2, 1,25(OH)2D3-triggered osteoclastogenesis can be inhibited by osteoprotegerin (OPG) in its capacity as a decoy receptor for RANKL [39]. Other aspects of bone biology regulated by 1,25(OH)2D3 shown in Fig. 2 include the transcriptional activation in osteoblasts of the genes encoding osteocalcin [40] and osteopontin [41], two proteins involved in bone remodeling, as well as the paracrine activation of existing osteoclasts via induction of osteoblast-derived cytokines such as IL-6 [42]. In addition to bone mineral homeostasis, 1,25(OH)2D3 also has effects on differentiation of certain cell types in skin [43] and in the immune system (Fig. 2) [44]. Examples of genes regulated in these tissues are involucrin in keratinocytes [45] and carbonic anhydrase
222 in macrophages [46]. Interestingly, the skin and the immune system are now recognized as extrarenal sites of CYP27B1 action to produce 1,25(OH)2D3 locally for autocrine and paracrine effects [47,48] (not shown in Fig. 2; see also Chapter 5). As depicted in Fig. 2 using the kidney as an example, an important mechanism by which the 1,25(OH)2D3VDR-mediated signal is terminated in all target tissues is the catalytic action of CYP24, an enzyme that initiates the process of 1,25(OH)2D3 catabolism (see Chapters 6 and 7). The CYP24 gene is transcriptionally activated by 1,25(OH)2D3 [49,50]. In addition, the 1α-hydroxylase (1α-OHase) CYP27B1 gene appears to be repressed by 1,25(OH)2D3 in a short negative feedback loop to limit the production of 1,25(OH)2D3 [51]. Sections III and IV of this volume contain a more extensive discussion of 1,25(OH)2D3 target tissues, which, as indicated in the box at the lower left of Fig. 2, include numerous cell types in addition to small intestine, bone, kidney, parathyroid, skin, and the immune system. Finally, insights into novel roles for the VDRRXR heterodimer in the colon and the hair follicle are discussed in Section VI of this chapter.
B. VDR-Regulated Genes and Vitamin D-Responsive Elements The vitamin D target cells/tissues discussed above all express significant levels of VDR and one or more of its RXR heteropartners. After liganding with 1,25(OH)2D3, the next step in the bioactivity of VDRRXR is the recognition of DNA sequences in the genes that are transcriptionally regulated. Table I displays the sequences of vitamin D-responsive elements (VDREs), which are binding sites for the VDR-RXR heterodimer in the 5′ upstream regions of a variety of vitamin D target genes. Many of the VDREs listed in Table I consist of a direct hexanucleotide repeat with a spacer of three nucleotides (DR3); such VDREs can mediate either positive (upper panel) or negative (lower panel) transcriptional regulation of target genes. In positive DR3 VDREs, VDR has been shown to occupy the 3′ halfelement, with RXR residing on the 5′ half-site [28]. Below the positive DR3 VDRE sequences in Table I is shown an “optimal” VDRE that was experimentally determined via binding of randomized oligonucleotides to a VDR-RXR heterodimer [52,53]. The conclusion from random selection is in general agreement with the repertoire of natural VDREs and defines the optimal VDRE as a direct repeat of two six-base half-elements that resemble estrogen responsive element (ERE) halfsites, i.e., AGGTCA, separated by a spacer of three nucleotides, denoted as a DR3. The highest affinity
G. KERR WHITFIELD
ET AL .
3′ (VDR) half-site is PGTTCA, where P is a purine base, and the highest affinity 5′ (RXR) half-site is PGGTCA. In addition, the random selection results suggest that the guanine at position 3 of the spacer (shown in lowercase in Table I) is important for VDR binding, an observation that is consistent with the finding [23,54] that this base is partially protected by RXR-VDR in methylation interference assays. In addition, a positive VDRE from the proximal region of the human CYP3A4 gene (central panel, Table I) represents an unusual configuration in that the two hexanucleotide half-element repeats are everted relative to one another, with a spacer of six base pairs (ER6) [55,56]. However, an examination of the 5′ halfsite complement, namely gAGTTCA, reveals that it is a perfect match to the 3′ VDR binding half-site in the optimal DR3 VDRE, including the first base pair of the spacer. Likewise, the 3′ half-site of the ER6 VDRE represents a perfect match to the 5′ RXR binding site in the optimal DR3 VDRE. Whether the ER6-bound VDR and RXR can interact with each other in the same fashion as on a DR3 element, and, indeed, whether VDR and RXR occupy the expected positions in the ER6, are questions for future research. Nevertheless, it should be emphasized that the use of an ER6 element by VDR has precedent with other receptors. In fact, this same element in the CYP3A4 gene has been shown to function as a responsive element for the pregnane X receptor (PXR), which also binds DR3 elements [57–59]. ER6-type elements also bind thyroid hormone receptor (TR)-RXR heterodimers [60,61], for which the normal responsive elements in other genes are direct repeats with a spacer of four nucleotides (DR4s) [60]. In contrast to the unique ER6 element, in which both half-elements coincide with the randomly selected sequences, the half-sites that exist in natural DR3 elements usually contain one to three bases that do not match the optimal VDRE. These variant bases occur with approximately equal distribution between the two half-sites. The variant bases, shown with single underline in Table I and tabulated below the random selection DR3, are usually purines. One notable exception is a thymidine in the third position of the 5′ (RXR) half-site, actually causing this site to match the optimal 3′ (VDR) half-site. The multiple sequence variations in natural VDREs may provide a spectrum of affinities for the VDR-RXR heterodimer, thus enabling these elements to respond to differing concentrations of the receptors (or their ligands) [62]. Another possibility, for which increasing evidence is accumulating, is that variant VDRE sequences induce unique conformations in the VDR-RXR complex, thereby promoting association of the heterodimer with distinct subsets of coactivators [63],
223
CHAPTER 13 Nuclear Vitamin D Receptor
TABLE I DNA Sequences of Selected Natural VDREs from Vitamin D-Regulated Genes Tissue/cell
Type of VDRE
5′-Half
Spacer
3′-Half
Reference
Intestine Intestine Keratinocyte Kidney Many targets Many targets Myelomonocytic line U937 Osteoblast Osteoblast Osteoblast Osteoclast precursor, macrophage Osteoclast precursor, macrophage
Positive DR3 (gene) human CYP3A4 distal rat CYP3A23 human involucrin human NPT2 rat CYP24 proximal rat CYP24 distal human p21 human osteocalcin rat osteocalcin mouse osteopontin chicken carbonic anhydrase chicken integrin β3
GGGTCA AGTTCA GGCAGA GGGGCA AGGTGA GGTTCA AGGGAG GGGTGA GGGTGA GGTTCA AGGGCA GAGGCA
gca tga tct gca gtg gcg att acg atg cga tgg gaa
AGTTCA AGTTCA GGCAGA AGGGCA AGGGCG GGTGCG GGTTCA GGGGCA AGGACA GGTTCA AGTTCG GGGAGA
[56,98,108] [56,316] [45] [317] [49] [50] [88] [318] [40,319] [41] [46] [320]
Random selection Common deviations Uncommon deviations
Optimal DR3 in positive DR3s in positive DR3s
PGGTCA –– TGG– –ACAAG
nng nna nnt
PGTTCA ––GG–G ––CAG–
[52,53]
Intestine
Positive ER6 (gene) human CYP3A4 proximal
TGAACT caaagg AGGTCA
[55,56]
Parathyroid Parathyroid Osteoblast Osteoblast Kidney Tumors, other targets Tumors, other targets (Pre-)osteoblast
Negative DR3 (gene) human PTH chicken PTH mouse osteocalcin rat bone sialoprotein PKA inhibitor rat PTHrP proximal rat PTHrP distal rat Cbfa1
GGTTCA GGGTCA GGGCA A AGGGTT ATGTTG AGGTTA GGTGGA AGTACT
[37] [321] [322] [323] [324] [301] [302] [325]
aag gga atg tat ctg ctc gag gtg
CAGACA GGGTGT AGGACA AGGTCA AGGTCA AGTGAA GGGTGA AGGTCA
Each VDRE consists of 5′ and 3′ half-elements (column 3) that bind RXR-VDR and mediate transcriptional regulation of genes (column 2) in target cells (column 1). Bases in positive VDREs that differ from the optimal sequence obtained by random selection experiments (center row) are single underlined, unless they are unique to negative VDREs, in which case they are double underlined. P denotes a purine nucleotide, i.e., A or G; n is any base.
or permitting differential actions in the context of diverse tissues [62]. When the DR3 sequences that mediate negative transcriptional regulation are examined, it is apparent that these, too, display a considerable degree of deviation from the optimal VDRE sequence. Not surprisingly, the base differences occurring in negative VDREs do not correspond to those found in positive VDREs. Those variant bases that exist only in negative VDREs are double underlined in Table I and often include pyrimidines, in addition to the predominant purine substitutions that occur in both positive and negative
VDREs (single underlined). Exactly how the uniquely negative acting bases influence the VDR-RXR heterodimer to mediate transcriptional repression is not known. However, Koszewski and colleagues [64] were able to demonstrate that the negative VDRE in the chicken PTH gene can be converted into a positive element simply by changing the most 3′ two bases from GT to CA. The success of this experiment implies that the VDRE sequence alone can determine whether the bound VDR-RXR heterodimer will activate or repress transcription. These results can also be interpreted as further evidence that the VDRE sequence can induce
224 significant conformational changes in the VDR-RXR heterodimer that are sufficient to promote binding to corepressors, rather than coactivators. Another novel aspect of negative VDREs is that the variant bases are not equally distributed between halfsites (Table I). Thus, PTH genes have unique negative variations exclusively in the 3′ element, while other negatively regulated genes have variations primarily in the 5′ half-element. The significance of this asymmetry for negative regulation, or even possibly the polarity of VDR and RXR binding to the two half-sites, is not currently known. Clues may be found by examining the all-trans-retinoic acid receptor (RAR)-RXR heterocomplex, which acquires ligand-dependent repressor activity when associated in reverse polarity (RXR on the 3′ half-site) with a direct repeat with a spacer of one nucleotide (DR1) retinoic acid responsive element (RARE) [65]. It should be noted that one negative VDRE in Table I, namely the distal VDRE in the rat parathyroid hormonerelated peptide (PTHrP) gene, has no base difference that would identify it as a negative element, i.e., all variations observed in this VDRE are also evident in positive VDREs. However, this same gene possesses a proximal VDRE that contains three base variants occurring only in negative VDREs (Table I). This observation highlights another area that is not well understood with regard to VDREs, namely the means by which multiple VDREs in the same gene function together to effect transcriptional regulation. The fact that three 1,25(OH)2D3-regulated genes (CYP24, CYP3A4, and PTHrP) with multiple VDREs have been characterized to date suggests that probing the functional interaction of two or more VDREs in the same gene is an important area for further study. The conclusion, from analyzing a number of VDREs, is that the docking sequence for VDR-RXR on DNA consists of clearly defined DR3 and ER6 motifs in the promoter region of vitamin D controlled genes. Regulation of the expression of many of these genes by 1,25(OH)2D3 in traditional vitamin D target tissues such as kidney, bone, parathyroid, skin and the hematopoietic system has been well established [66], and the combination of Fig. 2 and Table I paints a picture of the biological breadth of VDR-mediated gene control. Obviously, the 1,25(OH)2D3-VDR/RXR complex modulates the expression of a multitude of genes, and recent microarray studies [67–71] suggest that the number of vitamin D-regulated genes is much greater than the well-characterized bioresponses discussed herein. As described next, the use of VDR null animals has both validated the obligatory role of VDR in 1,25(OH)2D3 signaling, and revealed which VDR responses are the most pathophysiologically relevant.
G. KERR WHITFIELD
ET AL .
C. Consequences of VDR Ablation in Rodents and Humans Important insight into the biological impact of VDR to mediate the myriad of gene regulatory effects of 1,25(OH)2D3 discussed earlier can be gained by examining the phenotype of mice in which the VDR has been ablated (VDR knockouts), as well as the phenotype of human patients with hereditary hypocalcemic vitamin D-resistant rickets (HVDRR). HVDRR is a consequence of defects in the VDR gene on human chromosome 12 [72–75]. Clinically significant HVDRR is an autosomal recessive disorder resulting in severe bowing of the lower extremities, short stature, and often alopecia [74]. The serum chemistry in HVDRR reveals hypocalcemia, secondary hyperparathyroidism, elevated alkaline phosphatase, variable hypophosphatemia, and markedly increased 1,25(OH)2D3. Many of these sequelae are the expected manifestations of the loss of 1,25(OH)2D3 regulation in the target genes outlined above, such as PTH, RANKL, and CYP24. Most aspects of HVDRR, with the exception of alopecia, mimic those of classic vitamin D-deficiency rickets, suggesting that VDR not only mediates all of the bone mineral homeostatic actions of vitamin D, but may also participate in hair follicle function in a manner independent of vitamin D status. As detailed later in this chapter, individual natural VDR mutations observed in HVDRR have been very instructive in the elucidation of structure/ function relationships within the receptor, especially with regard to DNA, ligand, and coactivator binding. Several strains of homozygous VDR knockout mice have been created [76–78], and their phenotype has been characterized under various conditions. Somewhat surprisingly, VDR−/− mice, except for hair loss, appeared phenotypically normal until weaning, after which they began to display the suite of symptoms characteristic of HVDRR, namely hypocalcemia, hyperparathyroidism, and elevated 1,25(OH)2D3 levels [76,77]. One early observation with VDR knockout animals was that most aspects of the VDR−/− phenotype, except for alopecia, could be reversed if blood calcium levels were normalized by dietary manipulations [79]. The so-called “rescue diet” contained calcium, phosphorus, and lactose supplements to restore crucial bone mineral ions not properly assimilated after weaning [80]. The successful rescue, in relation to bone mineral homeostasis, of VDR-null mice suggests that the dominant biological action of 1,25(OH)2D3-VDR is to promote intestinal calcium and phosphate absorption, especially when there is physiologic stress from the requirement for bone mineralization under conditions of low calcium availability. With respect to alopecia, it was observed that lack of VDR prevented hair follicles from entering the anagen phase of the hair cycle [81], but transgenic
CHAPTER 13 Nuclear Vitamin D Receptor
VDR expression targeted to keratinocytes in the background of a VDR-null animal was successful in restoring hair cycling [82]. These results not only confirmed the essential role of VDR in maintaining hair (and skin) health, but also pinpointed keratinocytes as an important cellular target for VDR action in skin. More recent studies using VDR−/− animals fed the rescue diet to normalize calcium have revealed additional roles of VDR that are not simply secondary effects of maintaining serum calcium and phosphate levels. Separate studies have found that VDR−/− mice, when compared to normal mice: (a) exhibit uterine hypoplasia [77], probably linked to reduced expression of aromatase (CYP19) [83]; (b) display hyperplasia of the mammary glands [84]; (c) have blunted insulin synthesis in pancreatic beta cells [85]; (d) have reduced synthesis of macrophage IL-18 and STAT4 in T helper subset type 1 (Th1) lymphocytes, leading to an impaired Th1 response [86]; and (e) are markedly more sensitive to the skin carcinogen 7,12-dimethylbenzanthracene [87]. These exciting recent results indicate that beyond bone mineral integrity and hair cycling, VDR signaling significantly affects development in certain tissues as well as the function of the endocrine pancreas. VDR action is clearly immunomodulatory, and VDR likely mediates the reported anticancer effects of vitamin D. In the context of chemoprevention, 1,25(OH)2D3 is known to possess potent antiproliferative, prodifferentiation, proapoptotic, and/or cell cycle arrest activities in epithelial cells [56]. The molecular mechanisms underlying these effects may be related to the ability of liganded VDR to arrest cells at the G1 stage via induction of cell cycle regulatory proteins such as p21 [88] (Table I) and p27 [89], to control cell growth transcription factors such as c-myc [90] and c-fos [91], or to elicit apoptosis by repression of the Bcl-2 antiapoptotic factor [92]. However, an additional intriguing possibility is that VDR is chemopreventative by inducing detoxification enzymes, e.g., CYP3A, since several of the genes encoding these CYPs in mammals possess VDREs (Table I). Thus, as discussed in depth later in this chapter, a newly recognized action of liganded VDR is that of inducing CYPs for xenobiotic detoxification.
III. THE VITAMIN D RECEPTOR AS A MEMBER OF THE NUCLEAR RECEPTOR SUPERFAMILY A. VDR in the Context of the 48 Human Nuclear Receptors A proper understanding of the molecular actions of VDR requires an appreciation of its genetic placement within the nuclear receptor superfamily. The human
225 genome sequence [93] encodes 48 nuclear receptors [94], and interrelationships between these receptors are illustrated in Fig. 3. All human nuclear receptors, along with those from other species, have been placed into six groups according to a unified system of nomenclature [95]. As with previous analyses using a subset of human receptors [95], VDR groups with the TRs, RARs, the peroxisome proliferator-activated receptors (PPARs), the reverb orphan receptors, and the retinoic acid receptor-related orphan receptors (RORs) in a single clade that has been named group 1 [95]. As illustrated on the right side of Fig. 3, a significant feature shared by group 1 receptors is that many of them form active heterodimeric complexes with the RXRs from group 2 [96]. The closest relatives of VDR within group 1 are PXR and the constitutive androstane receptor (CAR), which, with VDR, form subgroup 1I (Fig. 3) (full designations are NR1I1 for VDR, NR1I2 for PXR, and NR1I3 for CAR) [95]. The next closest group of receptors to VDR is subgroup 1H, including the liver X receptors (LXRs) and the farnesoid X receptor (FXR). One notable feature of the 1H and 1I receptors is the ability of these receptors to recognize multiple ligands [97]. The diversity of ligands for PXR is especially broad and includes not only endogenous steroids, but also an array of other lipophilic compounds such as the secondary bile acid lithocholic acid (LCA) and the antibiotic rifampicin, as well as xenobiotics such as hyperforin, the active ingredient of St. John’s wort [97]. Until recently, VDR was thought to possess a single ligand, the 1,25(OH)2D3 hormone. However, it is now known that mammalian VDR has at least two additional ligands (Fig. 4), namely LCA and its derivative, 3-ketolithocholic acid (3-ketoLCA) [98]. This finding is of considerable interest to human medicine, since it is known that LCA is a secondary bile acid with significant carcinogenic potential [98–100], and its binding to VDR could trigger the detoxification of this ligand via the induction of CYP3A4 mediated by the VDREs in this gene (Table I). The structures of both LCA and 3-ketoLCA are shown in Fig. 4, along with the structures for prototypical PXR, FXR, and LXR ligands. It is evident when comparing these compounds that the ligand binding profile of VDR, expanded to include LCA and its 3-keto derivative, now resembles the ligand profile of the closest VDR relatives in the nuclear receptor superfamily, especially when it is noted that both PXR [101,102] and FXR [98,103] are also activated by LCA to some extent. It is now appreciated that most group 1I and 1H members, with the exception of CAR and the two LXRs, can recognize bile acid ligands. Therefore, based upon an analysis of its evolutionary position among the nuclear receptors and recent data
226
G. KERR WHITFIELD
TRα 1A TRβ RARα RARβ 1B RARγ PPARα PPARβ 1C PPARγ reverbα 1D reverbβ RORα RORβ 1F RORγ VDR PXR 1I CAR LXRβ LXRα 1H FXR HNF4 2A HNF4γ RXRα 2B RXRβ RXRγ TR2 2C TR4 TLX 2E PNR COUP COUP2 2F EAR2 DAX 0B SHP GCNF1 6A ERα 3A ERβ ERRα ERRβ 3B ERRγ GR MR 3C PR AR FTZ1 5A LRH1 NURR1 4A NOR1 NGFIB Outgroup (Cnidarian)
RXR
RXR
RXR
Group 1
RXR
RXR
RXR RXR
Group 2
nb
Group 0
or
Group 6
Group 3
Group 5 Group 4
AmNR2
FIGURE 3 Human nuclear receptors categorized into subgroups (0B-6A) according to the unified nomenclature scheme [95] with subgroup classifications shown to the right of the receptor abbreviations. This cladogram was created in Clustal W [326] using protein sequences obtained from GenBank. Abbreviations for Group 1 receptors and for RXR are defined in the text, and accession numbers are taken from Whitfield et al. [104]. Other abbreviations are HNF4, hepatocyte nuclear factor-4; TR2/4, testis-specific receptors 2/4; TLX, human homolog of Drosophila tailless; PNR, photoreceptor-specific nuclear receptor; COUP, chicken ovalbumin upstream promoter transcription factor; EAR2, erbA-related receptor-2; DAX, DSS-AHC region on human X chromosome; SHP, small heterodimer partner; GCNF1, germ cell nuclear factor-1; ERα/β, estrogen receptors α/β; ERRα/β/γ, estrogen-receptor related receptors α/β/γ; GR, glucocorticoid receptor; MR, mineralocorticoid receptor; PR, progesterone receptor; AR, androgen receptor; (Continued)
ET AL .
227
CHAPTER 13 Nuclear Vitamin D Receptor
VDR COOH
COOH
O
HO Lithocholic acid
PXR
3-Ketolithocholic acid
CH3
FXR
C=O
COOH
OH
HO O 5β-Pregnane-3,20-dione
LXR
Chenodeoxycholic acid
OH O
HO HO 22(R)-Hydroxycholesterol
24,25-Epoxycholesterol
FIGURE 4 Novel VDR ligands compared with ligands for the related receptors PXR, FXR, and LXR. The top panel shows the structures of LCA and its derivative, 3-ketoLCA, which are newly recognized ligands for human VDR [98]. The central two panels display the structures of 5β-pregnane-3,20-dione and chenodeoxycholic acid (CDCA), which have been reported as natural ligands for PXR [109] and human FXR [290,291], respectively. The nuclear receptors LXRα and LXRβ bind to oxysterols such as the two shown in the lower panel [327].
with bile acids, it is apparent that VDR binds ligands beyond the vitamin D congeners that regulate bone mineral homeostasis, and it is conceivable that VDR associates with novel lipophilic ligands to effect other extraosseous bioresponses (see Section VI,B).
FIGURE 3 Continued
B. New Insights from the Evolution of VDR Recently we cloned a high-affinity receptor for 1,25(OH)2D3 from the sea lamprey [104], which, along with other lampreys, hagfishes, and jawless fishes,
FTZ1, fushi tarazu 1 factor homolog; LRH1, liver receptor homolog-1; NURR1, nur related protein-1; NOR1, neural orphan receptor-1; NGF1B, nerve growth factor-induced protein-1B. AmNR2 is Acropora millepora nuclear receptor-2, a protein from coral used as an outgroup for constructing the cladogram. To the right of the subgroup classification is a schematic depiction of the typical configuration of the corresponding receptor (monomer or dimer) that binds to its cognate DNA responsive element. Many group 1 receptors, including VDR, bind as heterodimers with RXR (primary receptor shown as solid oval, with heterodimeric partner indicated as box containing the letters RXR). In addition, RXR forms homodimers on its responsive element. Many other receptors form homodimers on their respective responsive elements, illustrated as paired solid ovals. Receptors that bind DNA as monomers are depicted with single solid ovals. The group 0B receptors do not have a DBD and, hence, do not bind directly to DNA (indicated by the letters “nb”).
228 constitutes the most ancient lineage among extant vertebrates. Because lampreys lack both a calcified skeleton/teeth and hair, this was an unexpected finding given the dominant role of VDR in bone mineralization and hair cycling in mammals. Lamprey VDR, which displays approximately 60% amino acid sequence identity with other known vertebrate VDRs, is expressed in juvenile skin, mouth, and larval protospleen, but is not well expressed in juvenile intestine [104]. The low expression in intestine, combined with the fact that lampreys lack a calcified skeleton and teeth, strongly suggests that VDR is not a major factor in intestinal calcium absorption in this species. However, the presence of lamprey VDR in skin implies that VDR has some role in this tissue, analogous to the action of mammalian VDR in keratinocytes as discussed earlier. Another clue to the function of VDR in lampreys might be the observation that, among several VDREs tested, only the VDRE from the human CYP3A4 gene was able to confer significant 1,25(OH)2D3-dependent activation of a reporter gene [104]. These findings may offer insight into the function of VDR in the lamprey, indicating this evolutionarily ancient receptor could perform the role of inducing CYPs for xenobiotic detoxification, and likely provide a link to the present and a novel action of VDR akin to that of its close evolutionary cousins. Indeed, the common thread connecting the 1I receptors, VDR, PXR and CAR, is the induction of CYP enzymes that participate in xenobiotic detoxification, as summarized in Fig. 5A. A major target for VDR and PXR in humans is CYP3A4 [55,56,98], for which the detoxification substrates include LCA [105]. Initial studies focused on VDR liganded to 1,25(OH)2D3 as a regulator of CYP3A4 [55,56], but more recent experiments have revealed that LCA is also capable of binding VDR to up-regulate expression of human CYP3A4, or its equivalent in rats (CYP3A23) or mice (CYP3A11) [98]. There is also evidence that other CYP enzymes may be VDR targets [106]. In fact, many of the traditional targets of VDR-mediated regulation of vitamin D metabolism, namely 1α- and 24-hydroxylases, are CYPs (CYP27B1 and CYP24, respectively, according to standard CYP nomenclature) [107]. These studies therefore demonstrate that, in addition to its well-known role in calcium homeostasis, VDR also participates in the regulation of intestinal CYP genes in a manner similar to its closest relative, PXR [55,56]. As summarized in Fig. 5A, VDR and PXR may regulate some of the same genes, since PXR also utilizes DR3 and ER6 responsive elements in DNA for transcriptional activation [108]. The differential transcriptional effects of VDR and PXR may reside not in the genes that are regulated, but rather in the overlapping,
G. KERR WHITFIELD
ET AL .
yet distinct, ligand profile (Fig. 5A, second column); for example, PXR does not respond to 1,25(OH)2D3 [109]. It should be noted that CAR, the third human receptor in subgroup 1I, is also implicated in CYP regulation [110,111]. Whereas CAR normally binds DR4 elements (Fig. 5A), this relatively uncharacterized receptor may exhibit some cross over binding to DR3 elements; conversely, PXR may cross over to DR4 elements [112]. There is some preliminary evidence that VDR may also cross over to DR4 elements [106], as it was reported that VDR can activate transcription from at least two elements that strongly resemble degenerate DR4s [38,113]. Another aspect to consider in the differential activities of VDR, PXR, and CAR is the tissue distributions of these receptors. In humans, VDR is expressed in a large number of tissues (reviewed in Haussler et al. [35]; see also Fig. 2), including small intestine, colon, kidney, skin, certain immune cells (e.g., macrophages and Th1 lymphocytes), prostate, pancreas, ovary, and heart, but is present only in very low levels in mammalian liver [114]. The expression patterns of PXR and CAR are more limited; PXR is found predominantly in liver, small intestine, and colon [59], and CAR is restricted mainly to liver [115]. Thus, it would appear that VDR likely plays a significant role in CYP induction in intestine, colon, kidney, skin, and possibly also prostate and ovary, but defers to PXR and CAR in liver, a major site of xenobiotic detoxification and excretion. VDR shares fewer common characteristics with the more distantly related subgroup 1H receptors. As with PXR and CAR, liver is a predominant site of action for FXR and LXRα, along with other tissues such as adipose [116,117]. However, in contrast to the VDR/PXR/CAR grouping, xenobiotic detoxification does not appear to be a primary function of these receptors. Rather, as shown in Fig. 5A, FXR and the LXRs appear to be important regulators of cholesterol metabolism to bile acids and transport of cholesterol metabolites [118,119]. The general conclusion based upon examination of the evolutionary position of VDR among the 48 nuclear receptors is that it is extremely closely related to PXR, both structurally and functionally. In one respect, VDR appears to have evolved as a “specialty” regulator of intestinal calcium absorption and hair growth in terrestrial animals, providing both a mineralized skeleton for locomotion in a calciumscarce environment, and physical protection against the harmful UV radiation of the sun. Yet VDR also has retained its PXR-like ability to effect xenobiotic detoxification via CYP induction. VDR could complement PXR by serving as a guardian of epithelial cell integrity, especially at environmentally or xenobiotically exposed sites such as skin, intestine, and kidney.
229
CHAPTER 13 Nuclear Vitamin D Receptor
IV. STRUCTURE-FUNCTION OF THE VITAMIN D RECEPTOR VDR possesses the classical (Cys2–Cys2)2 zinc finger motif for DNA binding, and a multifunctional C-terminal domain for ligand binding, RXR heterodimerization, and transcriptional coactivator binding, but lacks the long N-terminal extension typical of the traditional steroid hormone receptors [120]. Figure 5B displays in schematic form the structural similarities between the group 1I and 1H receptors, highlighting the major functional domains for DNA-, ligand-, and heterodimer-binding. The percent identities within these functional domains mirror the evolutionary classification of these receptors, with PXR showing the highest identities to VDR in all three functional regions. The group 1H receptors (LXR and FXR) exhibit identities that are somewhat lower in the DNA binding and heterodimerization domains when compared to VDR, but display relatively strong identities in the ligand binding domain (LBD) (at 27–34%, approximating that of PXR and CAR), perhaps reflecting the fact that all 1I and 1H receptors recognize cholesterol derivatives with intact (or nearly intact) side chains (see Fig. 4). The human thyroid hormone receptor α (TRα) and human RXRα are included in Fig. 5B for comparison. Note that amino acid identities remain relatively high in the DNA binding and heterodimerization domains, even when VDR is compared to RXR, a group 2 receptor. However, identities in the LBD decline precipitously when TRα and particularly RXRα, which displays only 5% identity with VDR in the LBD, are compared to VDR. These low identities are likely a consequence of the fact that TR and RXR both bind nonsterol ligands. A unique feature of VDR and PXR that is probably related to their ligand binding profile is the fact that both of these receptors have especially lengthy amino acid segments between the DNA binding domain (DBD) and the highly conserved regions of the LBD (residues “missing” in the other receptors in this area relative to VDR are indicated with numbers preceded by a minus sign in Fig. 5B). The presence of additional residues following the DBD in VDR and PXR has been proposed to confer upon these two receptors more spacious binding pockets for accommodating a range of ligands, such as xenobiotics [121,122].
A. The Zinc Finger DNA Binding Domain of hVDR Throughout the nuclear receptor superfamily, the DBD is the region of highest sequence similarity, with
Ligands
A VDR 1I
PXR CAR
oxysterols
DR4
oxysterols
DR4
1° bile acids 1 (e.g., CDCA)
IR1,ER8
LXRα FXR
Genes
xenobiotic xenobiotics DR3,ER6 LCA, 3-ketoLCA (DR4, ER8) transport, CYPs xenobiotic DR4 xenobiotics transport, CYPs (DR3)
LXRβ 1H
Elements
1,25(OH)2D3 Ca++ homeostasis, LCA, 3-ketoLCA DR3,ER6 CYPs, hair cycle
cholesterol/lipid homeostasis cholesterol/lipid homeostasis bile acid homeostasis
Ligand binding Heterodimerization
B
VDR
DBD −5
+17
PXR
1I
−13
2B
RXRα
FIGURE 5
=34%
=41%
=30%
=41%
−46
=27%
=37%
−43
=16%
=35%
=5%
=31%
−47
−24
+86 51% +126
TRα
=49% −4
53%
FXR
1A
=27%
−56
+97
LXRα LXRβ
=50%
64%
CAR
1H
=35%
67%
56% +15 49% −73
+134 47%
Comparative features of VDR and closely related nuclear receptors from subgroups 1I and 1H. Abbreviations for each receptor are defined in the text. (A) Comparison of cognate ligands, DNA-responsive elements, and target gene function. Ligand abbreviations are defined in the text and the legend to Fig. 4. DNA responsive elements are classified as direct repeats with spacers of three or four nucleotides (DR3, DR4), elements in which the two half-site repeats are everted with a spacers of 6 or 8 nucleotides (ER6, ER8), or inverted repeats with a spacer of one nucleotide (IR1). Under target genes, CYP refers to cytochrome P450containing enzymes that are involved in detoxification of xenobiotics, or in the case of the CYP24 gene, degradation of 1,25(OH)2D3 and 25(OH)D3. (B) Schematic representations of each receptor from panel A along with the human TRα and RXRα. The subgroup designation of each receptor is listed at the left (see Fig. 3). The zinc finger DBD is indicated by a solid black box with white lettering that includes the percent amino acid identity in the DBD between the indicated receptor and human VDR. Other domains for which percent identities were computed are those that mediate ligand binding (three subdomains, lightly shaded) and heterodimerization (two subdomains indicated by hatched boxes). The remaining regions of the receptors are depicted as white boxes with numbers indicating a greater (+) or lesser (−) quantity of amino acids in those segments compared with human VDR.
230 identities ranging from 67% between VDR and PXR, to 47% between VDR and RXRα (Fig. 5Β). This high degree of similarity is a consequence of the fact that all nuclear receptors, with the exception of those that do not directly bind DNA, possess a DBD composed of two zinc fingers of the Cys2–Cys2 type. As illustrated schematically in Fig. 6A, the two zinc finger modules in hVDR are followed in the primary sequence by a C-terminal extension (CTE), which contains a long α-helix. These subregions are color coded in Fig. 6A for recognition in Fig. 6B–D (see color plate insert). The following subdomains are highlighted: a portion of the N-terminal domain (only two residues, 22 and 23, shown in dark blue), two zinc fingers (colored green and light blue, respectively) separated by an intervening region (brown), and a CTE shown in yellow. Nuclear localization signals are depicted in purple and naturally occurring inactivating mutations that alter amino acids in this region are denoted in red. Coordinated zinc atoms are shown as black circles with white lettering. The entire DBD, extending from positions 22 to 110 in the human VDR amino acid sequence, has been crystallized [123], and representative views of this region, constructed in Protein Explorer [124] using the published coordinates, are shown in Fig. 6B–D. It is important to consider that the hVDR DBD fragment was crystallized by Shaffer and Gewirth as a homodimer on three different DR3 VDREs [123]; the views shown here (Figs. 6B–D) are from VDR DBD fragment homodimers bound to the mouse osteopontin VDRE (Table I) [41]. While the native VDR functions not as a homodimer, but as an RXR heterodimer as described below and detailed elsewhere in this chapter, it is reasonable to assume that the DNA contacts established by homodimeric VDR DBD fragments approximate those of at least the core zinc finger region of the VDR partner when it binds to a VDRE as part of the full-length VDR-RXR heterodimer. Accordingly, as depicted in Fig. 6B, which illustrates only the 3′ half-element of the VDRE with the docked hVDR DBD fragment, the first zinc finger (green in color plate insert) of hVDR possesses an α-helix on the C-terminal side containing amino acid residues (specifically E42, K45, R49, and R50) that contact VDRE 3′ half-element bases in the major groove of DNA. Recent data [123,125] indicate that hVDR residues E42 and R49 make contact with the central two base pairs of the canonical AGTTCA VDRE 3′ half-element, whereas K45 and R50 contact the second and fifth G-C base pairs, respectively, in this half-element. Accordingly, the α-helix that contains these four hVDR base-pair contact residues (E42, K45, R49 and R50) is referred to as the specific DNA
G. KERR WHITFIELD
ET AL .
recognition α-helix and extends from E42 to K53. As depicted schematically in Fig. 6A, E42 forms part of a previously reported proximal box (P-box), which consists of three amino acids that distinguish the nuclear receptors which bind to AGGTCA ERE-type half-elements (EGG P-box) from those that bind to AGAACA GREtype half-elements (GSV P-box) [125]. Thus, exchange of only these three amino acids can switch ER homodimers to recognize GRE inverted repeats and, conversely, GR homodimers to recognize ERE inverted repeats [126]. However, for RXR heterodimerizing receptors such as VDR, an additional level of DNAbinding specificity is manifest, with an arginine at position 49 (Fig. 6A) dictating heterodimeric binding to direct repeat hormone responsive elements (homodimerizing nuclear receptors possess a lysine in the position corresponding to R49 in hVDR) [125]. Finally, in hVDR, the terminal residue of the DNA recognition α-helix, K53 (Fig. 6A), is also required for VDRE binding, and along with the CTE, confers VDR, as well as its closely related receptor PXR, with selective DR3 docking capacity [125]. Because heterodimerizing nuclear receptors such as VDR employ residues beyond the P-box to generate selective responsive element motif binding, we have grouped the highly conserved R49 along with the K53 residue in hVDR (or their positional equivalents in other receptors) into a new “specificity” box (S-box) [125]. The S-box therefore complements and extends the traditional P-box, and together they endow RXR heterodimerizing nuclear receptors with the capacity to specifically recognize their cognate direct repeat DNA responsive elements. An additional α-helix situated on the C-terminal side of the second zinc finger (blue in color plate insert) is positioned to interact with the DNA phosphate backbone (Fig. 6B), and there is extensive contact between residues at one end of this α-helix (specifically R73, R74, and R80) and DNA phosphates. The combined energy provided by the DNA recognition and phosphate backbone binding α-helices is apparently what is required for specific VDRE binding by VDR, and this conclusion is supported by the location of VDR point mutations found in patients with HVDRR [74] as illustrated in red in Fig. 6B (see color plate insert). All of these VDR-inactivating point mutations occur in the two zinc fingers, and many are localized to DNA base- or phosphate backbone-contact residues (e.g., K45, R50, R73, and R80). The consequences of the other HVDRR mutations in the DBD are less obvious, but still instructive. The planar side chain of phenylalanine 47 interacts extensively with the aliphatic portion of arginine 80 on one side and serine 51 on the other. Replacement of this residue with a branched chain amino acid (F47I) may
231
CHAPTER 13 Nuclear Vitamin D Receptor
A
DBD
CTE
D
Ligand binding/Heterodimerization
46 47
35
33
50
73
Ser51
Second Zn Finger
First Zn Finger
P-box E42 G43 G46 Zn 22
51 49-50 P 53-55
24 45 46 47
DNA Binding Mutations
Zn
82-83 102-104
60 80
50 R49 K53 S-box
89
C
Positively charged residues that mediate nuclear localization of hVDR
Natural mutations in hVDR that inhibit DNA binding and cause vitamin D resistance
B
110
CTE
Nuclear Localization
22
22 73
Zn
Zn Zn
82
33 45 35
46
Zn
83
N37
80 47
RXR
DNA 50
54
110 Heterodimeric orientation of CTE
110
DNA
K91 50 49 E92
Homodimeric orientation of CTE
110
103 102
55
53 104 103 102 104
Homodimeric orientation of CTE
110
Heterodimeric orientation of CTE
FIGURE 6 Structure–function relationships in the human VDR DBD (see color plate insert and text for details).
produce steric hindrances in this critical region. Glycines 33 and 46 are oriented toward the phosphate backbone; their mutation to negatively charged (and bulkier) aspartates (G33D and G46D) would likely lead to repulsion and/or steric hindrance, preventing proper DNA contact. Finally, histidine 35 forms a watermediated hydrogen bond complex with both a DNA phosphate and with lysine 45; its mutation to glutamine (H35Q) presumably disrupts these interactions. Therefore, with respect to the core zinc finger DBD of hVDR and its contact with the VDRE, the interpretation of the X-ray structure of the VDR-DNA cocrystal is consistent with the observed vitamin D-resistant rachitic and alopecic phenotype of the patients with DBD-inactivating VDR mutations. This is perhaps the most compelling evidence not only for the obligatory
role for VDR in vitamin D action, but also for the absolute requirement for DNA binding as a mechanism for VDR signaling. The depiction of the hVDR DBD fragment co-crystal with a VDRE in Fig. 6B includes two alternative representations for the positioning of the CTE. The CTE location in the VDR homodimer-VDRE structure is shown at the lower left [123]. An alternative positioning of the CTE, which is taken from a model of a VDRDBD/RXR-DBD heterodimer on a DR3 VDRE that was constructed by Rastinejad and colleagues [127] based on X-ray crystallographic coordinates of a TRDBD/RXR-DBD bound to a DR4 TRE, is illustrated at the lower right. The heterodimeric positioning of the CTE suggested by Rastinejad et al. [127] would be more consistent with a VDR-RXR heterodimer rather
232 than a VDR-VDR homodimer as the physiologic mediator of 1,25(OH)2D3 action [27,32,54,128–131]. The heterodimeric orientation of the VDR CTE is also more reconcilable with mutagenesis results [132] for the hVDR CTE, which indicate a crucial role for CTE residues in VDRE binding, including clusters of basic residues at amino acids 102–104 and 109–111. Although the X-ray crystallographic coordinates of the VDR/RXR heterodimer docked to a DR3 VDRE are not available, the model as presented by Rastinejad et al. [127] places the VDR CTE in a position nearly parallel to the DNA phosphate backbone, extending to the right in Fig. 6B in a 3′ direction relative to the VDRE sense strand. This heterodimeric positioning of the CTE would allow VDR basic residues in the 109–111 cluster to contact the DNA phosphate groups, unlike in the VDR homodimer cocrystal of Shaffer and Gewirth [123], in which the VDR CTE crosses the DNA phosphate backbone at nearly a right angle, and VDR residues 109–111 are placed too far from the phosphate backbone for any contact to occur. Therefore, pending the cocrystallization of a VDR-RXR heterodimer on a VDRE, we conclude [125] that, whereas the hVDR DBD fragment homodimer crystal [123] provides valid results for the DNA base and phosphate backbone contacts of the core zinc finger residues, it distorts the positioning of the CTE, which likely conforms instead to the heterodimeric model of RXR-VDR [127]. In order to bind to DNA and regulate gene expression, transcription factors require nuclear translocation after biosynthesis on the polyribosomes, and nuclear localization is dependent upon various arrangements of basic amino acids in short stretches within the primary amino acid sequence [133]. Four such segments of positively charged amino acids that mediate hVDR nuclear translocation [134,135] exist in the DBD as designated in purple in Fig. 6A (see color plate insert). Figure 6C contains the same view of the VDR DBD as in Figure 6B, but again highlights (in purple) the hVDR basic residues that mediate nuclear localization [134,135]. These residues align vertically on one surface of hVDR when the CTE is in the heterodimeric orientation, perhaps creating an interaction surface for nuclear import factors. Some of the same hVDR amino acids that contact DNA also participate in nuclear localization, namely the basic residue clusters at positions 49–50 [134] and 102–104 [135]. The coidentity of DNA binding and nuclear localization residues is not a general feature of transcriptionally active proteins, but has been reported previously for a variety of DNA- and RNA-binding proteins [136]. Two additional basic clusters are also involved in hVDR nuclear translocation, located at residues 53–55 [134] and 82–83 [135]. Exactly how these positively charged residues mediate
G. KERR WHITFIELD
ET AL .
nuclear localization is not known, although the fundamental nuclear membrane machinery, including importin, has been characterized for this general process [137]. Interestingly, Barsony and colleagues [138] have confirmed the essential role of hVDR residues 49–50 and 53–55 [134], and demonstrated that positionally equivalent basic residues in RXR are also important in nuclear localization, especially of the unliganded VDR-RXR heterodimer [138] (see also Chapter 22). Thus, it appears that a quasi-stable, hormone-unoccupied VDR-RXR heterodimer is the species of receptor that localizes in the nucleus of vitamin D target cells, probably bound nonspecifically to DNA with relatively low affinity. The weak association between unliganded VDR and its RXR partner may be facilitated by hydrogen bonding of N37 in hVDR to residues in the second zinc finger in RXR, as well as by salt bridges between K91 and E92 in hVDR to oppositely charged amino acids in the second zinc finger of RXR (N37, K91, and E92 are shown in orange in Fig. 6C (see color plate insert)). These three residues located on the 5′ side of the tertiary structure of the VDR DBD on the VDRE constitute a proposed [127] RXR interface (as suggested by the orange arrow, Fig. 6C) that is likely buttressed when the VDR-RXR heterodimer is bound to DNA. However, as discussed in the next section, a much stronger dimerization interface between VDR and RXR, that is 1,25(OH)2D3dependent, exists in the LBD of VDR. A final structure-function aspect of the VDR DBD relates to regulation of nuclear localization and DNA binding by phosphorylation. Serine 51, located near the C terminus of the first zinc finger and within the base recognition α-helix, is a site for phosphorylation by protein kinase C (PKC) [139]. Whereas some investigators have observed a potentiation of VDR activity after PKC activation in intact cells [140], probably as a consequence of phosphorylation of a VDR coactivator, results from our laboratory indicate that direct phosphorylation of human VDR by PKC at serine 51 has negative effects on both nuclear localization and DNA binding [139]. Figure 6D (inset at upper right) presents a crystallographic view of a portion of the first zinc finger of hVDR bound to a DR3 VDRE [123]. It is evident that the serine 51 side chain faces toward a DNA phosphate group (in color plate, phosphorus atom is shaded orange; oxygens are magenta). We hypothesize that phosphorylation of the serine 51 side chain which is located only 4.7 nm from one of the phosphate oxygens, not only introduces a steric hindrance to this positioning of the VDR DNA binding helix, but also creates a charge repulsion between the two negative phosphate groups that would prevent the DNA-binding α-helix of VDR from contacting the VDRE. Presumably, introduction
CHAPTER 13 Nuclear Vitamin D Receptor
of a negative charge at position 51 would also neutralize the nearby basic nuclear translocation residues. Supporting this conclusion is our observation [139] that replacing serine 51 with a constitutive negatively charged amino acid, namely aspartate, precludes both nuclear localization and VDRE binding by hVDR. Therefore, although the biological significance of PKCcatalyzed phosphorylation of hVDR is unknown, this modification clearly modulates the activity of the receptor, perhaps attenuating it under certain conditions of cell growth and differentiation.
B. Ligand Binding/Heterodimerization/ Transactivation Domains of VDR and Comparison with PXR To date, an X-ray crystallographic structure of the DBD and the LBD together in the same receptor protein has not been reported, and therefore it is not possible to visualize how the DBD and the ligand binding/ heterodimerization domains are arranged relative to one another. However, an X-ray crystallographic structure of the VDR ligand binding/heterodimerization domain bound to 1,25(OH)2D3 has been obtained by Rochel et al. [121]. As shown schematically in Fig. 7A, and in a view created in Protein Explorer [124] in Fig. 7B, the crystallized VDR structure contains two discontinuous segments, from residues 118 to 164 and from 216 to 425, spliced together to form a peptide that could be purified and crystallized [121] (the position of the deletion is indicated by a white box with crossed lines in Fig. 7A and by red highlighting of residues 164 and 216 on either side of the deletion in the Fig. 7B color plate insert). Despite the discontinuity, the structure of the VDR ligand binding/heterodimerization domain displayed in Fig. 7B bears a strong resemblance to the X-ray crystal structures of the corresponding regions of other nuclear receptors [141], consisting of what has been described as a “sandwich” of 12–13 α-helices with a centrally located hydrophobic ligand binding pocket (see also Chapter 15). Within the general structure of the α-helical sandwich, the ninth and 10th helices (depicted in green in Fig. 7, see color plate insert) have been shown in several nuclear receptors to contain heterodimer contacts [142]. The surface formed by these helices in VDR appears similar to that observed in crystals of the LBD of RARα, RXRα, PPARγ, and ER [143–146]. Moreover, there is a human HVDRR patient harboring a VDR in which arginine 391 (helix 10) is mutated to cysteine (colored blue in Fig. 7B, see color plate insert), and in vitro experiments have confirmed that
233 RXR heterodimerization by the R391C mutant hVDR is severely impaired [147]. The R391C HVDRR patient, like all those with DNA-binding mutations in hVDR (see Fig. 6AB), displayed an alopecic phenotype, indicating that RXR heterodimerization is functionally linked to VDRE binding, and that both of these actions of VDR are required for hair cycling. Independently supporting an obligatory role for the RXRα heteropartner in hair cycling is the observation that mice with an RXRα conditional knockout in skin exhibit alopecia indistinguishable from that in VDRnull animals [128]. Crystallographic studies with a number of nuclear receptors have revealed that helices 3, 5, and 12 combine to form a platform for transcriptional coactivator binding [148–152], illustrated for hVDR in yellow in Fig. 7B (see color plate insert). The involvement of helix 12 in VDR action is also supported by the identification of an HVDRR patient, in whom glutamate 420 is altered to lysine [153]. Further testing has demonstrated that natural mutant E420K hVDR [153], as well as its synthetic counterpart, E420A [154], have abrogated transactivation capacity caused by their inability to interact with the comodulators steroid receptor coactivator-1 (SRC-1) and vitamin D receptor interacting protein 205 (DRIP205). The third natural hVDR mutant highlighted in Fig. 7B is an arginine 274 to leucine alteration that leads to a loss in binding of the 1,25(OH)2D3 ligand [147]. Arginine 274 is located in helix 5, and directly contacts the 1,25(OH)2D3 ligand via hydrogen bonding of the 1α-hydroxyl moiety [121]. Thus, the three natural point mutations in the hVDR LBD highlighted in Fig. 7B represent loss-of-function alterations for each of the three molecular actions of this domain, namely ligand binding, heterodimerization, and transactivation, with the respective disruption of contact by hVDR with 1,25(OH)2D3, RXR, and coactivators. In summary, just as clinically significant inactivating point mutations in the hVDR DBD prove the obligatory role of DNA binding in VDR action (Fig. 6B), the three HVDRR associated point mutations in the C-terminal domain of hVDR highlighted in Fig. 7B illuminate the necessity for ligand binding, RXR heterodimerization, and coactivator contact in order for VDR to transduce the signal for the prevention of rickets. Although the crystal structure of the unoccupied LBD of hVDR has not been solved, it is generally accepted [141] that unliganded nuclear receptors are configured such that the helix 12/activation function 2 (AF2) is in the “open” orientation, directed away from the α-helical sandwich and allowing for entry of the ligand into a hydrophobic pocket (Fig. 8). Upon ligand binding with 1,25(OH)2D3, helix 12 of VDR, which contains two ligand
234
G. KERR WHITFIELD
A
DBD
ET AL .
Ligand binding/Heterodimerization
CTE
Transactivation 165
215
118
425 H1
deleted
H3 H5 βstrands
H9
H10 H12
hVDR
hPXR/SXR
B
C Helix 1 Helix 9
Helix 5
C 274 420 Hel
Helix 3
391
Helix 10
RXR
r
ato tiv
c oa
ix 1
2
SR12813
1,25(OH)2D3
∆165-215
Proposed strand region deleted from hVDR
177-198 (unstructured)
FIGURE 7 Structure-function of the human VDR ligand binding/heterodimerization/transactivation domain compared to that of hPXR/SXR (see color plate insert and text for details).
contact residues (valine 418 and phenylalanine 422), is repositioned to the agonist bound “closed” orientation and interacts with helices 3, 5, and 11 [121] to create a coactivator platform (Fig. 7B). 1,25(OH)2D3 ligand binding also alters the conformation of the LBD such that the surface presented by helices 9 and 10 is more favorable for association with RXR (Fig. 7B). Facilitation by 1,25(OH)2D3 of strong heterodimerization of VDR with RXR has been demonstrated [27,32,147], and the resulting VDR-RXR heterocomplex is capable of high-affinity interaction with VDREs. Therefore VDR liganding with the 1,25(OH)2D3 hormone generates both RXR heterodimerization for specific DNA binding and coactivator docking for transcriptional activation. One fascinating aspect of the action of the VDR LBD is the following observation. Point mutations in hVDR that abrogate DNA binding and RXR heterodimerization lead to both rickets and alopecia; however,
neither 1,25(OH)2D3 ligand binding (e.g., R274L) nor coactivator interaction (e.g., E420K) mutations elicit alopecia in conjunction with the vitamin D-resistant rachitic phenotype. Since vitamin D deficiency or ablation of the 1α-hydroxylase enzyme that generates the 1,25(OH)2D3 hormone [155,156] also causes rickets without alopecia, it is clear that the 1,25(OH)2D3 ligand binding function of VDR can be dissociated from the action of the receptor to regulate the hair cycle. Similarly, transcriptional activation by VDR is not required for hair growth. It therefore appears that either the unliganded VDR-RXR heterodimer bound to DNA is sufficient for control of the hair cycle, or a novel (non-vitamin D) lipophilic ligand operates in skin to activate VDR. As discussed later in this chapter, because VDR mediation of the hair cycle does not require coactivators, it is possible that this VDR function is executed by gene repression, and a nuclear receptor corepressor such as the hairless protein (Hr) [157]
235
CHAPTER 13 Nuclear Vitamin D Receptor
may participate in VDR signal transduction in the hair cycle. The availability of a PXR structure [122] permits interesting comparisons between the ligand binding/ heterodimerization domain of VDR (complexed with 1,25(OH)2D3) and that of PXR (bound to synthetic ligand SR12813), its closest relative in the nuclear receptor superfamily. The overall structures of the domains of these two receptors are nearly superimposable [121,122] (compare Figs. 7B and 7C with matching views created in Protein Explorer [124]). The PXR LBD has a particularly large ligand binding pocket (1150 Å3) [158]; the ligand binding pocket of VDR, even with the deletion of residues 165–215, is also large (approx. 700 Å3) compared with other nuclear receptors that have been crystallized (approx. 400 Å3 [121] but ranging as high as 800 Å3 for LXRβ) [159]. This suggests that like PXR and LXR, VDR may be able to accommodate a variety of lipophilic ligands beyond 1,25(OH)2D3, including LCA/3-ketoLCA. Of additional interest is the fact that the PXR crystal, which was created from a continuous stretch of sequence from positions 142–431 of overexpressed human PXR [122], contains residues that correspond to most of the 165–215 region that is absent in the VDR LBD crystal. Thus the PXR X-ray crystallographic solution can serve as a model for the VDR structure in this region. Of particular note are the additional two β-strands in PXR (illustrated in red in color plate insert), for which the corresponding hypothetical positions in VDR are suggested by the dotted outline in Fig. 7B. Indeed, inclusion of these two β-strands, plus associated turns, in this speculative hVDR LBD structure could allow for the further expansion of the ligand binding pocket. The assertion has been made that this region (amino acids 165–215 in hVDR) may not be necessary for the transactivation function of VDR [160]. However, a mutation was reported in an HVDRR patient in which alteration of cysteine 190 to tryptophan results in loss of hormone binding by hVDR [161]. Although it is possible that the substitution of cysteine 190 by the more nonpolar and much bulkier tryptophan side chain somehow disrupts the general tertiary structure of the VDR LBD, this finding could be interpreted as an indication that the deleted region (165–215) does in fact serve some important function that is not detected by limited in vitro testing of the mutated receptor performed with only the 1,25(OH)2D3 ligand at high (10− 7 M) concentrations [160]. Therefore, as suggested by the close evolutionary relationship between VDR and PXR (Figs. 3 and 5B), as well as the ability of both receptors to bind bile acids (Fig. 5A) and functionally overlap to induce CYP
detoxification enzymes (Fig. 5A), X-ray crystallographic structures of the VDR and PXR ligand binding domains (Fig. 7B,C) strongly support the concept that VDR potentially possesses an as yet uncharacterized array of novel ligands. The repertoire of new VDR ligands could approach that of PXR if one considers the relatively large size of the hydrophobic pocket in the hypothetical expanded VDR LBD architecture pictured in Fig. 7B.
V. MECHANISMS OF VITAMIN D RECEPTOR-MEDIATED CONTROL OF GENE EXPRESSION A. Initial Biochemical Steps in the Activation of VDR Signaling As described earlier in Figs. 1–7 and Table I, several lipophilic ligands and target gene DNA sequences have been defined for VDR, and the structures of the 1,25(OH)2D3 hormone-occupied LBD as well as the VDRE-docked DBD have been elucidated. However, the precise molecular mechanism whereby VDR ligands signal the control of gene transcription has not been fully characterized. This process is best understood for VDR mediation of 1,25(OH)2D3-stimulated transcription, where RXR heterodimerization constitutes an obligatory initial step in the VDR activation pathway. At physiologic receptor concentrations and ionic strength, VDR will bind to VDREs only in the presence of both 1,25(OH)2D3 and RXR [32], and in an in vitro transcription system containing native chromatin, 1,25(OH)2D3 will transactivate via VDR only when RXR is included [129]. Figure 8 illustrates in hypothetical schematic fashion how the hormonal ligand could be influencing VDR to interact more efficaciously with its heterodimeric partner, with a VDRE, and with coactivators. Based upon experiments with VDR [66,130,162], as well as insight from the mode of action of other nuclear receptors [163–165], the key event in the allosteric model presented in Fig. 8 is the binding of a ligand, either the 1,25(OH)2D3 ligand for VDR or 9-cis RA for RXR. According to this model, the ligand that arrives initially sets the tone for the dimerization and transcriptional events that follow. If 1,25(OH)2D3 binds first, the liganded VDR is proposed to follow the pathway to VDR-RXR heterodimerization, indicated by bold arrows in Fig. 8. There are several steps that apparently are set in motion by the ligand binding event. These include high-affinity heterodimeric interaction of VDR with RXR and the recognition by the VDR-RXR heterodimer of specific
236
G. KERR WHITFIELD
ET AL .
Basal state: 9-cis RA
RXR
Platform for coactivators
1,25(OH)2D3 ligand
VDR
1,25(OH)2D3
Protein kinase
P
VDR
RXR
Non-specific DNA ‘‘sliding’’ 5'
AGGTCA
9-cis Retinoic acid ligand
Homodimeric RXR diversion
3nt AGTTCA
DR3
RXR
3'
VDRE
Coactivator recruitment
VDR Coactivator(s)
Platform for coactivators 9-cis RA
RXR
LXXLL
9-cis RA
LXXLL
RXR
1,25(OH)2D3
RXR
5'
AGGTCA 1nt AGGTCA DR1 RXRE
P
VDR
3' 5'
Coactivator recruitment
9-cis RA-stimulated transactivation via RXR-RXR
AGGTCA
3nt AGTTCA
DR3
VDRE
3'
1,25(OH)2D3-stimulated transactivation via RXR-VDR
FIGURE 8
Allosteric model of VDR-RXR activation after binding 1,25(OH)2D3, or diversion of the obligatory RXR partner by its 9-cis-retinoic acid ligand (see text for details).
VDREs upstream of 1,25(OH)2D3 target genes [27,32]. The presence of the 1,25(OH)2D3 ligand in the binding pocket also results in a dramatic conformational change in the position of helix 12 at the C terminus of VDR, bringing it to the “closed” position to serve in its AF2 role as part of a platform for coactivator binding [154,166,167]. Further, as depicted in Fig. 8 and supported experimentally [130,131,162], the liganding of VDR conformationally influences its RXR heteropartner to cause the AF2 region of RXR to pivot into the “closed” or active position. The VDR-RXR heterodimer is now poised to bind coactivators, and the allosteric repositioning of the RXR AF2 also greatly reduces the affinity of RXR for its 9-cis RA ligand. With RXR serving as a subordinate partner, VDR is referred to as a nonpermissive primary receptor within the heterodimer,
because RXR cannot bind 9-cis RA when heterodimerized to liganded VDR [130,131,162]. The conformational changes elicited in VDR by the above described interactions with ligand, RXR and/or DNA have the added effect of converting VDR into a more efficient substrate for one or more serine protein kinases [54]. The most relevant phosphorylation appears to be catalyzed by casein kinase II (CK2) on hVDR serine 208 [168,169], an event that has been shown to potentiate the transcriptional activity of the VDR-RXR heterodimer [170], likely by enhancing interactions with coactivators [171]. As discussed earlier (Fig. 7), the coactivator platform in other nuclear receptors has been shown in crystallographic studies to consist of helix 12 (the AF2) in concert with helices 3 and 5 [148–152], and it has
CHAPTER 13 Nuclear Vitamin D Receptor
been suggested that most nuclear receptors have similar interaction domains [141,142], with the possible exception of the progesterone receptor [146,172]. These crystallographic studies indicate that coactivators of the p160 class, such as SRC-1, utilize one or more LXXLL motifs for binding to nuclear receptors (where L = leucine and X = any residue). There is evidence from site-directed mutagenesis experiments that helix 12 and helix 3 of VDR are indeed involved in coactivator contact [154,173]; also, the above-mentioned natural mutation in VDR in which glutamate 420 is altered to lysine has been found not only to confer the HVDRR phenotype in a human patient, but also to abolish binding of SRC-1 to VDR in experiments with transfected cells [153]. As the “subordinate” partner receptor bound to the 5′ half-site of the VDRE, RXR nonetheless has been proven to be essential for transactivation by the 1,25(OH)2D3-liganded VDR-RXR heterodimer [130,131,162]. Accordingly, the depiction in Fig. 8 illustrates the AF2s of VDR and RXR simultaneously contacting the same coactivator. This is a plausible scenario, since SRC-1 and other p160 coactivators possess three distinct LXXLL domains that are known to interact with nuclear receptors [174,175]. However, also pictured in Fig. 8 (by the dashed vertical line dividing the coactivator) is the possibility that VDR and RXR contact separate coactivators at this stage [130,131,176], the combination of which is required for the full transcriptional response. In summary, the initial events in positive signaling by VDR are triggered by 1,25(OH)2D3 liganding, which generates the VDR-RXR heterodimer that both binds to the VDRE and recruits bridging coactivators that attract and stimulate the transcription machinery. The events described in the preceding paragraphs are all predicated on the binding of 1,25(OH)2D3, and not 9-cis RA, to the unliganded VDR-RXR heterodimer. Should 9-cis RA bind first, the scenario likely becomes quite different. In this case, as suggested in Fig. 8 (lower left), the VDR-RXR heterodimer would dissociate, leaving RXR to form homodimers that could then seek out retinoid X receptor responsive elements (RXREs) in DNA. The most common reported motif for an RXRE is a DR1 [25]. In this pathway (narrow arrows on the left side of Fig. 8), the binding of the RXR ligand would preempt action by 1,25(OH)2D3 in favor of retinoid signaling, and this model is consistent with observations in many systems using 1,25(OH)2D3- and retinoid-dependent reporter genes [27,30,32]. However, there are contrasting reports that, in certain regulatory or cellular contexts [177,178], or with different VDREs [179], 1,25(OH)2D3 and retinoid ligands can act synergistically to effect transcriptional activation. As will be
237 discussed later in relation to Fig. 10, synergistic activation of gene expression by 1,25(OH)2D3 and retinoids could be the result of distinct VDRE sequences allosterically imparting conformational variations on the VDRRXR heterodimer and/or the presence of cell-specific transcriptional comodulators.
B. A Sequential, Cyclic Working Model for Transactivation by 1,25(OH)2D3-Liganded VDR-RXR In the process of dissecting the complete mechanism whereby 1,25(OH)2D3-liganded VDR effects transcriptional activation, numerous VDR-interacting proteins (VIPs) have been identified [35,66,180]. VDR possesses interaction surfaces in the DBD (Fig. 6C) and LBD (Fig. 7B) for RXR, the most fundamental VIP. The 1,25(OH)2D3-liganding of VDR conformationally influences both subunits in the VDR-RXR heterodimer to recognize VDREs and become a binary nucleus for the attraction of coactivators to vitamin D regulated promoters (Fig. 8). The balance of the transactivation process is incompletely understood, but appears to involve numerous other VIPs that remodel chromatin, participate in transcriptional initiation, and modify VDR. Some VIPs directly contact VDR, and others act as part of supercomplexes tethered to VDR. By analogy to the molecular actions of the TR [181] and ER [182], and the observations that the regions of VDR required for interactions with p160 coactivators like SRC-1 [166], vitamin D receptor interacting proteins such as DRIP205 [167], and proteasomelinked TRIP1/SUG1 [183] physically overlap and include the AF2/helix 12 domain, we propose that these VIPs function via a sequential or combinatorial [184,185] rather than simultaneous mechanism. Figure 9 represents a current conception of the 1,25(OH)2D3-liganded VDR-RXR transcriptional activation cycle, focusing on sequential complexes formed on the VDR-RXR heterodimer [186]. A starting point for the cyclic model illustrated in Fig. 9 is that of the unliganded and loosely associated VDR-RXR heterodimer nonspecifically bound to DNA, with transcription idling in the basal state (Fig. 9, top). This depiction is similar to the basal state of the VDR-RXR heterodimer in Fig. 8, but also includes a corepressor bound to the unliganded VDR-RXR. Corepressors have a higher affinity for unliganded VDR [187], in which the AF2 domains of VDR and RXR are presumably in the open, inactive position [166]. Some corepressors, such as silencing mediator for retinoic acid and thyroid hormone receptors (SMRT)
238 [188], and the hairless gene product, Hr [157,189], associate with proteins possessing histone deacetylase (HDAC) activity that represses chromatin [188]. A final member of the apo-VDR complex is basal transcription factor IIB (TFIIB), which interacts with unliganded VDR at two sites [190–193] (indicated as I and II in Fig. 9). Thus, analogous to the findings of Li et al. [194], who showed that TR-RXR exists in a complex with both a corepressor (NCoR) and coactivator (SRC3/ACTR), we postulate that inactive VDR-RXR is associated with both a corepressor and an activator (TFIIB) of transcription. As detailed in Fig. 8, a number of transformations occur upon switching VDR to the active state by binding of the 1,25(OH)2D3 ligand, including higher affinity heterodimeric association between VDR and RXR, specific recognition of VDREs by the liganded heterodimer, repositioning of the AF2 domains in both VDR and RXR, and serine phosphorylation of VDR. These changes convert VDR and RXR into platforms for coactivator binding and, concomitant with 1,25(OH)2D3 liganding, the corepressor/HDAC complex dissociates in order to permit the docking of a coactivator(s). The coactivator shown in Fig. 9 (right) is SRC-1, which belongs to the p160 class of coregulators that interact via LXXLL domains with a hydrophobic crevice created by contact between the AF2/helix 12 and helices 3 and 5 in both VDR and RXR (as discussed above and pictured for VDR in Fig. 7B). The coactivator complex, illustrated on the right in Fig. 9, apparently contains at least three additional proteins. NCoA-62 [193,195], a nuclear receptor coactivator also known as Ski-interacting protein, or SKIP, is believed to interact directly with VDR, but at a site distinct from the AF2 domain, and forms a ternary complex with VDR and SRC-1 [193,196]. CREB binding protein (CBP) or its relative, p300, is a cointegrator that is next incorporated into the activation complex by interaction with SRC-1 [197]. CBP/p300 does not bind directly to VDR and only associates with the receptor when the SRC-1 bridge is present [186]. CBP/p300, in turn, attracts the multisubunit Brahma/ SWI2-related gene 1 associated factor (BAF), or the closely related polybromo- and BAF-containing complex (PBAF) [198], both of which can serve as the mammalian counterpart [199] to the yeast SWI/SNF complex [185]. The net result of ligand-driven assembly of the VDR-RXR/SRC-1/NCoA-62/CBP-p300/PBAF supercomplex (depicted on the right in Fig. 9) is that chromatin remodeling factors have been recruited to the promoter of a 1,25(OH)2D3-regulated gene. There is an alternative view of the interaction of PBAF with nuclear receptors such as RAR [200] and
G. KERR WHITFIELD
ET AL .
VDR [201] (Fig. 9, top), in which the ATP-driven chromatin remodeling activity of PBAF is required prior to RXR-heterodimeric binding to the RARE and VDRE, respectively, and that the remodeling SWI/SNF-like complex actually attracts the unliganded nuclear receptor heterodimer to the promoter site instead of vice versa (Fig. 9, right). Indeed, the chromatin remodeling complex WINAC that is defective in Williams syndrome [201], an inherited hypercalcemic disorder caused by a deletion in chromosome 7, has been reported [201] to target unliganded VDR to its promoter sites. WINAC [201], PBAF [198], and possibly other complexes that utilize ATP hydrolysis to rearrange nucleosomal arrays to render DNA accessible to VDR-RXR are clearly required for 1,25(OH)2D3 stimulation of transcription in the context of native chromatin [198]. However, in vivo footprinting of the rat osteocalcin gene promoter reveals that only in the presence of the 1,25(OH)2D3 ligand is the VDRE occupied by VDR [202]. Therefore, whether PBAF-like complexes facilitate VDRE recognition by VDR prior to liganding will require further proof, although roles for SWI/SNF at multiple steps in transactivation have been reported [203], including recruiting histone acetyl transferase (HAT) complexes. SWI/SNF has even been proposed to execute disassembly of nuclear receptor-coregulator complexes, a role also postulated for chaperone proteins such as p23 (not illustrated in Fig. 9) [204]. Disassembly of transcription factor complexes could facilitate “turning of the cycle” in Fig. 9 and generate rapid, dynamic interactions between VDR and its array of co-regulators. In fact, recent studies of the RNA polymerases suggest that rather than consisting of stable holocomplexes, transcription (or transcriptional control) units are assembled in a stochastic fashion from freely diffusible subunits that only transiently associate [205]. Employing fluorescence recovery after photobleaching, it was observed that the liganded glucocorticoid receptor [206] and its interacting protein (GRIP1) coactivator [207] quickly exchange between a DNA promoter and the nucleoplasm. Thus, the transcriptional activation complexes pictured in Fig. 9 may be dynamically assembled and reassembled by a “hit and run” mechanism instead of remaining as holo-complexes throughout the turning of the cycle. The task of PBAF, along with the SRC-1/CBP or SRC-1/p300 complex, is to derepress the chromatin in the vicinity of the VDRE (Fig. 9). This is accomplished by the combined action of BAF/PBAF to remodel the chromatin [198] by sliding nucleosomes along DNA, and of the SRC-1 and CBP/p300 complex to reacetylate nucleosomal histones via HAT activities [208] intrinsic to the coactivators. Histone acetylation and
239
CHAPTER 13 Nuclear Vitamin D Receptor
Basal state
Corepressor AF2
AF2
I
VDR
RXR
Hormone catabolism
HDAC
TFIIB Corepressor/HDAC
PBAF SWI/SNF
TFIIB
II Nonspecific DNA
Repressed chromatin
Corepressor/HDAC
Protein phosphatase
CYP24
SRC-1,CBP NCoA-62
1,25(OH)2D3
Chromatin remodeling
TRIP1/SUG1 SRC-1,CBP NCoA-62, TRIP1/SUG1 TFIIB,PBAF
VDR degradation in 26S proteasome
1,25(OH)2D3Ligand Protein kinase
CBP/P300
RXR
ubiquitination
AF2
N SRC-1 Co LXXLL LXXLL A62 AF2 AF2
P
1,25
VDR
U
Transcriptional recycling
VDR
3nt
TFIIB
3' 5'
Repressed chromatin
VDR target gene
monoubiquitination
SRC-1,CBP PBAF(SWI/SNF)
Mediator complex D
NCoA-62 DRIP205/Mediator RNA pol II
R
I P
PIC s
DRIP205/Mediator RNA pol II
DRIP205 NR1
NR2
AF2
AF2
P
RNA Pol II TFIIB
1,25
RXR VDR 5'
3'
Autoacetylation
Histone deacetylation
N Co A62
3nt
VDRE Specific DNA sequence
TRIP1/SUG1
Spliceosomemediated hnRNA processing
P
1,25
RXR VDR 5'
Chromatin remodeling
Histone acetylation
TRIP1/SUG 1 AF2
VDR poly-
VDRE Specific DNA sequence
FIGURE 9
PBAF SWI/SNF
3nt
3'
VDRE Specific DNA sequence
De-repressed chromatin
Transcriptional initiation
VDR target gene
The transcriptional life cycle of VDR, depicting a model for the stages in transactivation of an idealized 1,25(OH)2D3regulated gene (clockwise from top of figure). At twelve o’clock is shown the unliganded VDR-RXR heterodimer including its proposed interactions with TFIIB and with a corepressor. The corepressor, which binds to VDR when its AF2 is in the “inactive” configuration, acts by attracting HDACs that repress chromatin. Also shown is PBAF, which according to one view [201] may remodel chromatin to attract the VDR-RXR heterodimer to the promoter. Illustrated at three o’clock are the postulated initial alterations that transpire after binding of 1,25(OH)2D3 to VDR. These changes include one or more hormone-dependent phosphorylations of VDR (indicated by a circled “P” against a dark background), along with reconfiguration of the AF2 domains in VDR and RXR. The VDR-RXR heterodimer is now bound to a VDRE adjacent to a target gene. Changes in the conformation and phosphorylation status of the VDR-RXR heterodimer serve to dissociate the corepressor/HDAC complex and promote the binding of a coactivator complex that includes not only SRC-1, but also CBP/p300, nuclear receptor coactivator-62 (NCoA-62), and a SWI/SNF chromatin remodeling complex anchored by PBAF. This coactivator complex derepresses chromatin in the vicinity of the VDR target gene by catalyzing histone acetylation and chromatin remodeling. At six o’clock is illustrated the transcriptional initiation complex. Steps that are postulated to occur between three and six o’clock are (i) autoacetylation of SRC-1 and CBP/p300, leading to their dissociation from the VDR-RXR heterodimer, taking PBAF with them; (ii) binding of DRIP205 to the AF2 of VDR (and also RXR), an event that attracts other DRIPs in the Mediator complex; and (iii) delivery of TFIIB to the RNA Pol II transcription machine (indicated by a rightward arrow). The complex pictured at nine o’clock is postulated to form after the transcription machine has moved away from the promoter, with DRIP205/Mediator dissociating from the VDR-RXR heterodimer to be replaced by TRIP1, a mammalian ortholog of the yeast SUG1 protein. NCoA-62 also dissociates, and may be involved in processing of the primary heterogeneous nuclear RNA (hnRNA) transcript by the spliceosome [210]. VDR is then subject to ubiquitination by TRIP1 and its associated ubiquitin ligation complex (an attached ubiquitin is indicated by a circled “U”). The structure illustrated at nine o’clock now has three alternative fates. The “transcriptional recycling” fate encompasses dissociation of TRIP1, either with or without monoubiquitination of VDR, and reassociation of SRC-1, CBP/p300, NCoA-62, TFIIB and PBAF to regenerate the complex at three o’clock for a rapid reinitiation of transcription. A second fate would be dissociation and metabolic elimination of the 1,25(OH)2D3 ligand via catabolism by CYP24, leading to recovery of the unliganded complex at twelve o’clock after dissociation of TRIP1 and reassociation of TFIIB and corepressor. A third fate (receptor degradation), which may occur after several rounds of VDR ubiquitination and transcriptional initiation, would be transfer of the poly-ubiquitinated VDR to the 26S proteasome for degradation, effectively terminating the VDR transcriptional life cycle.
240 methylation [184] appear to reduce internucleosomal interaction in chromatin, converting it to transcriptionally active euchromatin. Acetylation by HATs is proposed to continue until coactivators undergo selfacetylation [182] and dissociate from VDR along with BAF/PBAF to free sites on the receptor for recruiting DRIPs and the RNA polymerase II preinitiation complex. The next complex to form with the VDR-RXR heterodimer (shown at the bottom of Fig. 9) is initiated by the attraction of DRIP205 to the AF2s of both VDR and RXR [167]. This occurs on a similar or overlapping docking site to that vacated by the p160-class coactivator (SRC-1) in the previous complex. The physical impossibility of both SRC-1 and DRIP205 interacting simultaneously with the AF2 platforms of VDR and RXR is, in fact, one of the chief arguments for the sequential model presented herein (see also Rachez et al. [167]). DRIP205 then attracts a Mediator complex that is still incompletely characterized, but which contains numerous other DRIPs that bridge the VDR/RXR/NCoA-62/DRIP205 complex to the RNA polymerase II (RNA Pol II) transcription machine [209]. When DRIP205 associates with VDR, we hypothesize that it displaces TFIIB and delivers it to RNA Pol II [191,192], facilitating the formation of the transcriptional preinitiation complex (PIC). After transcription begins and the initiation complex dissociates, DRIP205/Mediator is likely released from VDR, as is NCoA-62, which remains in the vicinity of the induced gene and appears to participate in spliceosomemediated heterogeneous nuclear RNA processing [210]. Once transcriptional initiation has occurred, and DRIP205/Mediator has departed the VDR-RXR platform, the absence of coactivators with either HAT or chromatin remodeling activity would presumably expose the promoter DNA and its nucleosomes in the region of the VDRE to the repressive action of histone deacetylases. Several possibilities now exist for the fate of the liganded VDR-RXR AF2 platform. Evidence from chromatin immunoprecipitation assays [211] indicates rapid oscillatory binding of p160 and DRIP coactivators to VDR, suggesting that the next event in the VDR transcriptional cycle is the rebinding of the SRC-1, NCoA-62, CBP/p300/PBAF supercomplex, basically “short-cutting,” or recycling, back to the chromatin remodeling stage shown at the right of Fig. 9. Because 1,25(OH)2D3-liganding of VDR has been reported to protect the receptor from proteasome-mediated degradation [212], sustained transcriptional activation by VDR-RXR via alternating recruitment of p160 and DRIP205 coactivators is a conceivable scenario. On the other hand, in other eukaryotic transcription systems, for example the srb10 kinase Mediator subunit
G. KERR WHITFIELD
ET AL .
which signals ubiquitination of the transcriptional activator Gcn4, and the MED8 subunit that assembles into Elongin BC-based E3 ubiquitin ligase [213], Mediator attraction to the PIC is coupled to the subsequent recruitment of ubiquitination machinery. Thus, as hypothesized on the left side of Fig. 9, in the case of VDR action, the vacant VDR-RXR AF2 platform may become occupied by TRIP1 (the mammalian ortholog of the yeast SUG1 protein [183]) as a consequence of DRIP205/Mediatorcoupled attraction of the ubiquitination machinery. TRIP1/SUG1 is a documented VDR-interacting protein [183,186] that stimulates VDR proteolysis in a process blocked by the 26S proteasome inhibitor, MG-132. If VDR follows the GR pattern of successive mono-, di-, tri-, and tetra-ubiquitination, in which the linkage of fewer than four ubiquitin moieties precludes association of the proteasome degradation machinery [214], then like GR, VDR will be capable of reengaging HAT/DRIP/PIC complexes and completing additional rounds of transcription [215]. In this model, activated VDR-RXR would traverse the transcription cycle pictured in Fig. 9 clockwise until the first ubiquitination event, and then “short circuit” the cycle by making four rounds of the lower half of the cycle prior to tetra-ubiquitination and subsequent degradation in the 26S proteasome. A final alternative fate of the TRIP1-VDR-RXR complex would be triggered by the dissociation of both TRIP1 and the 1,25(OH)2D3 ligand. The loss of ligand leads to a conformational change in VDR and RXR, in which the AF2 domains pivot to the open inactive position, leading to reassociation of corepressor with the unliganded VDR-RXR heterodimer. Unliganded VDR would now be available as a target for a protein phosphatase to remove the phosphate(s) on VDR that was covalently attached by the hormone-dependent protein kinase(s). All of these changes would return the VDR-RXR heterodimer to the state appearing at the top of Fig. 9, completing a full turn of the proposed transcription activation cycle. Termination of the VDR transcriptional cycle is postulated to be accompanied by catabolism of the 1,25(OH)2D3 ligand via CYP24 (24-hydroxylase) initiation of its conversion to calcitroic acid [47]. This enzyme is, in fact, encoded by one of the genes that is transcriptionally activated by the liganded VDR-RXR heterodimer to promote destruction of the ligand as a mechanism for self-limiting the transcriptional activation signal [216] (see also Chapter 6). The transcriptional activation cycle model for VDR illustrated in Fig. 9 is presented as a working hypothesis because the order in which each co-regulator participates has not been determined experimentally, although firm evidence exists that each VDR-interacting player binds to the receptor, at least under in vitro conditions.
CHAPTER 13 Nuclear Vitamin D Receptor
Thus, the order of events depicted in Fig. 9 will not be absolutely defined until real-time chromatin immunoprecipitation assays of the type carried out by Pike and colleagues [211] are completed for VDR on several of its endogenous target gene promoters. The cycle shown in Fig. 9 is a basic pattern subject to many variations depending upon the specific target cell and gene regulated by VDR. A schematic conception of tissue- and promoter-specific influences on the transcriptional activation cycle is shown in Fig. 10. As discussed earlier (Table I), there are many variations in the sequences of natural VDREs, and there are indications that these sequence differences may allosterically confer the VDRE-bound heterodimer with distinct conformations that attract different sets of coactivators [62,63]. The VDRE may also be located adjacent to binding sites for other transcription factors (termed cis-modulators in Fig. 10), and there are reported cases of interactions between the bound VDR-RXR heterodimer and nearby DNA-bound transmodulator proteins. Prominent examples are the human osteocalcin and involucrin promoters, in which the VDRE is located adjacent to an AP-1 binding site for the jun/fos heterodimer [45,217]. Given the obvious complexity of many other natural VDRE-containing promoters, there are likely additional examples that remain to be characterized, for instance in the promoters of the rat osteocalcin [218] and mammalian CYP24 genes [219]. As outlined earlier, phosphorylation of VDR by various serine kinases can be either stimulatory when catalyzed by CK2 at serine 208 [170], or inhibitory when catalyzed by PKC at serine 51 [139]. Moreover, we have recently provided evidence [54] that phosphorylation of hVDR by an unidentified kinase may be required for high-affinity heterodimeric binding of the receptor to a VDRE. There are other kinases that phosphorylate VDR, such as PKA [220], and because protein kinase expression is likely cell specific, the resulting phosphorylation pattern of VDR may vary from one target cell to another. Phosphorylation of RXR has been studied less extensively, but there are indications that RXR is also phosphorylated [221–223]. This posttranslational modification of RXR presumably could influence its transcriptional activity [219,221–223], but whether phosphorylation of RXR plays a role in signaling by VDR-RXR, as suggested by Fig. 10, remains to be clarified. Specific studies of RXR as a heterodimeric partner are complicated by the fact that there are three distinct RXR genes in mammals [224], with each gene producing multiple isoforms [225]. On a superficial level, all of these RXR isoforms seem interchangeable as heterodimeric partners for VDR. However, results from several laboratories [28,177,226] strongly suggest that RXR isoforms may exhibit differential abilities to
241 participate in transactivation in heterodimeric context with VDR. The VDR-mediated transactivation models presented in this chapter (Figs. 8 and 9) consider the 1,25(OH)2D3occupied VDR-RXR heterodimer to be refractory to stimulation by 9-cis RA, a natural RXR ligand. However, there have been reports of synergism between 1,25(OH)2D3 and 9-cis RA in the activation of certain genes [227]. The above-mentioned possibility of alternative RXR conformations when the VDR-RXR heterodimer is bound to different VDREs raises the possibility that some of these conformations might allow for 9-cis RA to bind to RXR and to participate in reconfiguring the RXR AF2 for coactivator binding. For many years 1,25(OH)2D3 was considered to be the sole physiologic ligand for VDR. However, with the recognition of at least two new natural bile acid ligands for VDR (Fig. 4), plus the availability of numerous chemical analogs of 1,25(OH)2D3, it is possible to screen ligands that selectively activate the VDR-RXR heterodimer [228–230]. The ability of some analogs to mimic certain transcriptional actions of 1,25(OH)2D3 (e.g., in skin differentiation) while avoiding the hypercalcemic effects has unveiled exciting possibilities for the use of these compounds in therapeutic interventions for hyperproliferative skin disorders [231], for autoimmune disorders [232], or as potential antitumor agents [233]. The basis for selective actions of VDR ligands is presumably that they, like the VDREs (which can also be considered as ligands for VDR-RXR), stabilize distinct conformations of the heterodimer that can undergo some, but not all, of the full spectrum of coactivator/corepressor interactions. This strategy would be analogous to the use of selective estrogen receptor modulators (SERMs) in targeting estrogen receptor antagonist and agonist actions to specific estrogen target tissues [234]. As has been emphasized throughout this chapter (especially Figs. 8 and 9), in order for VDR to control the expression of genes, it must interact with coregulator proteins that possess a variety of activities that transduce the signal to RNA polymerase II. Many of these proteins associate with the AF2/helix 12 of VDR, or at least influence the helix 3/5/12 platform. Clearly, the allosteric regulators of VDR just outlined impinge on the ability of the receptor to attract the sequence of comodulators illustrated from left to right across the top of Fig. 10. The several categories of comodulators that associate with VDR, classified according to their actions, are corepressors (Hr/HDACs), SWI/SNF chromatin remodelers (PBAF/WINAC), HAT coactivators/ cointegrators (SRC-1/CBP/p300), DRIP Mediators and basal transcription factors that assist in the assembly of the PIC containing RNA polymerase II
242
G. KERR WHITFIELD
ET AL .
Cell specific comodulator expression Chromatin repression: Hr/HDACs
CoChromatin remodeling: integrators: CBP/ PBAF/WINAC p300 (SWI/SNF)
Coactivators: SRC-1 GRIP1/TIF2 ACTR
Mediator/Basal factors: DRIP 205 TAFs TFIIB DRIPs
HATs
Receptor degradation: TRIP-ubiquitination 26S proteasome
NCoA-62 RXR ligands Rexinoids
Cell specific P phosphorylation
AF-2
AF-2
RXR isoforms
VDR polymorphisms
Transmodulator
VDR ligands Drug analogs
P
Cell specific phosphorylation
Transmodulator DNA
Cismodulator
VDRE sequences
Cismodulator
Promoter specific
FIGURE 10 Model showing potential allosteric regulatory inputs (large single- and double-headed, open arrows) to VDR-RXR signaling as mediated by a liganded heterodimeric receptor–DNA complex. The VDR-RXR heterodimer bound to a VDRE is considered to be the central entity into which various allosteric inputs converge. The basic program of transcriptional activation by liganded VDR presented in Figs. 8 and 9 can be modulated by variations in VDRE sequences in different target promoters (see Table I), or by adjoining DNA sequences to which may be bound other (possibly cellspecific) proteins that serve as “transmodulators” of VDR action (see text for examples). VDR activity may be intentionally modified by drug analogs, for which selective actions are believed to be the result of their ability to induce conformations of VDR that differ from the 1,25(OH)2D3-dependent conformation, and may promote VDR interactions with distinct sets of coactivators. The same principle may apply to alternative natural VDR ligands, such as LCA. Another possible molecular scenario for activation of the VDR-RXR heterodimer may be the ability of certain VDREs or ligands to permit the binding of rexinoid ligands to RXR. This possibility, plus the existence of three distinct RXR genes, each producing multiple isoforms, some cell-specific, further amplifies the variations in signaling by the VDR-RXR heterodimer. All of these allosteric inputs to VDR and its RXR coreceptor may influence subsequent steps in transcriptional regulation, including the binding of various coactivators or corepressor complexes (shown at the top), some of which may be cell-specific.
(DRIP205/TAFs/TFIIB), coactivators that participate in downstream hnRNA splicing (NCoA-62), and receptor proteolysis cofactors (TRIP1-SUG1/26S proteasome). Perusing the list of comodulators at the top of Fig. 10, one is impressed with the sheer number of protein–protein interactions that occur in target cells in order for the liganded VDR-RXR-VDRE supercomplex to repress (upper left) or stimulate (top center) DNA transcription. This complex VDR signaling network would be analogous to the estrogen receptor signaling network [235] that involves a multitude of estrogen receptor interacting proteins.
The exact set of comodulators recruited by the activated VDR-RXR heterodimer may constitute another tissue-specific permutation influencing the nature of transcriptional activation. Many coactivators, including SRC-1, GRIP1 (SRC-2), and ACTR (SRC-3), have distinct tissue distributions. Studies with knockout mice suggest that there are coactivator redundancies in many target tissues [236]. Nevertheless, the presence of varying relative and/or limiting concentrations of certain coactivators may have significant, yet incompletely characterized, effects in a given tissue [237]. The final variable illustrated in Fig. 10 that influences
CHAPTER 13 Nuclear Vitamin D Receptor
VDR-RXR signaling is common polymorphisms in the hVDR gene [238]. As discussed in Section VI,A, both the quality and quantity of VDR-mediated transactivation can be affected by polymorphic variants in the hVDR gene. Therefore, as pictured in integrated fashion in Fig. 10, the VDR-RXR “nucleus” that recruits comodulators for signal transduction is a multifaceted entity that is differentially influenced in an allosteric manner (open arrows in Fig. 10) by hVDR gene polymorphisms, variable VDR lipophilic ligands, cell-specific VDR and RXR phosphorylations perhaps resulting from cross-talk actions of cell surface-acting hormones, unique transmodulators associated with cis elements adjacent to VDREs, and gene promoter-selective VDRE DNA sequence platforms.
VI. IMPLICATIONS OF VITAMIN D RECEPTOR-MEDIATED SIGNALING FOR HUMAN HEALTH AND DISEASE A. VDR Polymorphisms: Association with Disease Risk and Functional Consequences A potentially significant factor in signaling by the VDR-RXR heterodimer is the existence of common polymorphic variants in the hVDR gene, a summary of which is presented in Fig. 11. There are essentially three major locations of polymorphic variation in hVDR: (i) a dimorphic site toward the 5′ end of the gene between exons 1D and 1G within a docking sequence for Cdx-2, a caudal-related homeobox protein; (ii) a dimorphic sequence at the translation start site in exon II; and (iii) a cluster of linked sites in or near exon IX. The characteristics, frequencies, and functional consequences of these three variable loci are discussed next. Cdx-2 plays an important role in intestinal development and differentiation [239], and has been reported to participate in intestine-specific expression of VDR [240] as well as the sucrase-isomaltase complex (SI) [241]. A binding site for this factor, namely the sequence TTTAY (where Y is a pyrimidine [242]) has been identified near the predominant hVDR gene promoter located downstream of exon IA (the Cdx-2 binding site is designated by a black dot in the top register of Fig. 11 between exons ID and IG). Within this binding site is a dimorphic (A/G) variation that results in a change in the Cdx-2 binding site on the complementary strand from TTTAT to TTTAC (see inset, lower left of Fig. 11). The binding site with the G variant was shown by gel mobility shift experiments to have a slightly lower affinity for the Cdx-2 factor [243], implying that
243 intestinal hVDR expression could be suboptimal with the G allele of the receptor gene. Consistent with this in vitro result and conclusion, bone mineral density (BMD) in the lumbar spine was observed to be 12% lower in Japanese women with the GG genotype compared with AA homozygotes [243]. The distribution of the A versus G version of the VDR gene was found to be approximately 45% versus 55% in a Japanese population [243], indicating that the apparently less active G allele predominates in this ethnic group. The translational start site polymorphism in exon II of the hVDR gene produces two alternative forms of the hVDR protein, with one isoform three amino acids longer at the N terminus than the other. The three additional residues in the longer f form are Met-Glu-Ala (Fig. 11, center inset). The approximate worldwide gene frequencies of the f and F alleles are 35% and 65%, respectively [244–247]. Interestingly, the F hVDR allele is associated with enhanced BMD in three different populations of women [244,245,248]. In concert with these clinical observations, the shorter F species of hVDR is also more transcriptionally active [192,244]. Studies in our laboratory revealed that the shorter F hVDR interacts more avidly with TFIIB, and that mutation of the glutamate at position 2 of the longer f isoform to an uncharged residue (alanine) restores the transactivation ability of the longer species to that of the shorter one [192]. Since the TFIIB interaction at this site (site II in the depiction of TFIIB at the top of Fig. 9) appears to be dependent on the presence of two basic residues (Arg-18 and Arg-22) in the hVDR N-terminal domain, we have speculated that the molecular mechanism whereby the negatively charged Glu-2 residue attenuates TFIIB binding could involve either an intermolecular repulsion between f hVDR and presumed negative residues in TFIIB that bridge to hVDR Arg-18/22, or a nonproductive intramolecular interaction of Glu-2 in f hVDR with the Arg-18/22 TFIIB site II, thus precluding TFIIB contact [192]. The common 3′ polymorphisms in the hVDR gene are more complex and difficult to interpret with respect to their functional impact. There are almost two dozen reported hVDR gene polymorphisms in or near exon IX [249] that are in linkage disequilibrium, none of which result in a change in the encoded protein (Fig. 11). Also, unlike the Cdx-2 variations, none of these common 3′ polymorphisms in the hVDR gene are known to occur in the binding site for a factor that could affect receptor transcription. However, one of the areas of variation, a singlet(A) repeat microsatellite approximately 1 kb upstream of the poly(A) tail in the 3′ UTR [250] (Fig. 11, right inset), could conceivably affect the stability or translatability of the VDR mRNA. We have postulated [238] that the human
244
G. KERR WHITFIELD
ET AL .
Alternative Splicing/promoter usage (Tissue specificity?)
IF
IE IAID IG IB
IC II
DNA-bonding zinc fingers
Untranslated regions
Ligand-binding domain
III
IV V VI
VII VIII
IX
q13-14 BsmI (intron)
5′ UTR Chr 12
Alternative translation start
Linkage: ♦ BAtS or ♦ baTL
Cdx-2 binding site ♦ Cdx-2 tissue-specific enhancer for intestine ♦ binding site is polymorphic:
Clinical association: ♦ GG homozygote lower BMD
ApaI TaqI (A)S or (A)L (intron) (silent)
Clinical association: ♦ ff homozygote lower BMD Functional observation: ♦ f isoform less active in vitro
Functional observation: ♦ G allele less active in vitro
3′ UTR
Length difference in a singlet(A) repeat: S: Short (A) repeat = (A)10-15 L: Long (A) repeat = (A)17-24 Clinical associations: ♦ lower Ca absorption, BMD (BAt) ♦ higher rate of bone loss (B) ♦ risk factor for prostate (TL), breast cancers (b) ♦ risk factor for psoriasis (A)
FIGURE 11
Summary of common polymorphic variations in the human VDR gene that may influence receptor activity. At the top is the arrangement of VDR exons [277]. Exons shown as black rectangles contain coding sequences; hatched rectangles represent the 5′ and 3′ untranslated regions (UTRs). Exon IX contains the terminal portion of the coding region along with the entire 3′ UTR [277]. A black dot indicates the position of a binding site for the intestinal Cdx-2 transcription factor [241] that has been reported to affect VDR expression in the intestine [240]. The composition of the 5′ UTR is complex [328], but a predominant form of VDR mRNA appears to be produced from the promoter just upstream of exon IA and contains exons IA and IC spliced to exons II–IX (this promoter and two alternative promoters are indicated by arrows above exons IF, IA, and ID) [277]. There are two variants at the Cdx-2 binding site, two variants in exon II, and numerous polymorphic forms in or near exon IX. The two Cdx-2 binding site variants are proposed to have different affinities for the Cdx-2 intestine-specific transcription factor [243]. SI is the abbreviation for sucraseisomaltase, another gene with a binding site for Cdx-2. The two exon II sequence variants dictate alternative translational start sites in the VDR mRNA, beginning at methionine-1 (M1) or methionine-4 (M4); these alternative forms are detected by the presence ( f) or absence (F) of a FokI restriction site in the genomic DNA (or in cDNA made from the hVDR mRNA). The multiple polymorphic forms at the 3′ end of the gene form a linkage group, with the two most common haplotypes (BAtS and baTL) shown in the black inset at right. None of these 3′ gene variations, including the two TaqI forms, result in any change in the VDR protein sequence. However, we have postulated that the L/S polymorphic forms, which affect the length of a singlet(A) repeat in the 3′ UTR, may influence translatability of the hVDR mRNA (see text). Text boxes at the bottom indicate clinical and functional correlations that have been observed for the indicated hVDR polymorphic variants.
poly(A) binding protein (PABP) binds to this singlet(A) repeat, especially in its longer versions. It has been reported that mammalian and yeast PABPs can enhance translatability of mRNAs via an interaction with other proteins that bind at the 5′ end of the message [251,252]. Usually, PABPs bind to the poly(A) tail of mRNAs, with each monomer occupying approximately 27 adenylate residues [253]. Intriguingly, further studies with human PABP suggest that as few as 11 consecutive As can bind to PABP, with 25 As giving maximum
affinity [254]. Thus, we have speculated that binding of PABP may be enhanced in long (L) alleles (17–24 As) versus short (S) alleles (10–15 As) of the hVDR gene. Greater binding of PABP to L alleles would then lead to more efficacious hVDR mRNA translation via a superior interaction with factors such as EF-4B [251]. This proposal is consistent with our preliminary observation that hVDR L alleles generate a higher concentration of transcriptionally active receptor than do S alleles [238]. The gene frequencies of hVDR L alleles
CHAPTER 13 Nuclear Vitamin D Receptor
(with the remainder being S) vary in human populations from a high of 91% in Asians to a low of 59% in Caucasians [255], indicating that the apparently more active L allele and its linked 3′ polymorphisms predominate worldwide. Given our determination that the L alleles, like the F alleles, produce VDR with a higher activity than the respective alternative alleles S and f, we have devised a system for classifying VDR genotypes at both loci using a single variable [238]. Accordingly, the genotype FFLL is assigned an allele score of 4, with the opposite ffSS genotype receiving a score of 0. Intermediate genotypes receive a score that equals the sum of the more active alleles, L and F; for example, a genotype of FfLS would receive a score of 2. Using allele scores, the average VDR activity in a group of patients can be estimated, with clinical implications as discussed next. The ability of allele score to predict the VDR activity in an individual patient is limited by the fact that there appears to be a more active and a less active subgroup within each allele score grouping [238]. This observation strongly suggests that at least one more functionally significant VDR gene polymorphism remains to be discovered. There is, in fact, a third functionally significant polymorphism within the Cdx-2 binding site between exons 1D and 1G; however, this polymorphism would be predicted to affect VDR activity only in intestine, not in skin fibroblasts that were the subject of our investigation [238]. A significant amount of literature is devoted to identifying correlations between hVDR gene polymorphisms and various clinical parameters or metabolic disorders, for instance BMD or prostate cancer risk [249]. Without reviewing this entire body of work (the reader is referred to Chapter 66 and to other recent reviews [249,256,257]), we offer several generalizations as follows. A broad summary of the accumulating data implicates the less transcriptionally active ff and BB/tt (BAtS haplotype) homozygous alleles as risk factors for low BMD [248], although these are not universal findings and the field of hVDR gene polymorphisms is somewhat controversial [258,259]. Nevertheless, an intuitive interpretation that less active hVDR alleles constitute a risk factor for skeletal hypomineralization is consistent with the physiologic role of VDR in calcium metabolism and bone biology (Fig. 2). The studies probing cancer prevalence versus hVDR gene polymorphisms are more difficult to reconcile and do not offer the physiological correlations supplied by bone mineral homeostasis. In the case of cancer incidence, it appears that the more active hVDR alleles are actually risk factors, especially for prostate and breast cancer [250,260–263]. This is counterintuitive considering the reported antiproliferative and
245 prodifferentiation roles of VDR in prostate [264,265], skin [43], immune cells [88,266,267], and other tissues [268–270]. There are exceptions to the more transcriptionally active baTL hVDR haplotype (see Fig. 11, inset at right center for linkage of alleles in this 3′ region of the gene) having an association with higher cancer incidence; for example the baTL genotype correlates with a lower susceptibility to the initiation and progression of malignant melanoma [271], as well as with a lower breast cancer risk of Latinas living in the United States [272]. Finally, the baTL haplotype associates with a lower incidence of the hyperproliferative skin disease psoriasis [273]. Nevertheless, the majority of data in the case of both prostate [250,260,262] and breast [263] cancer indicate that the more active F/baTL genotypes are procarcinogenic. An explanation of these findings is still elusive, but one mechanism to account for this paradox could be the known immunomodulatory influence of 1,25(OH)2D3-liganded VDR, which consists (in part) of a modification of T-cell function through suppression of the Th1 response [274]. This action of VDR is beneficial in combating graft rejection and autoimmune diseases, but it could conceivably have a negative impact on immune surveillance against malignant cells and tumors [275,276]. Another speculative explanation for the association of more active hVDR polymorphisms with prostate and breast cancer is that VDR, by its action to induce CYP enzymes which can catalyze the formation as well as degradation of carcinogenic metabolites, may be facilitating the synthesis of androgenic and estrogenic compounds that promote cell proliferation in the corresponding endocrine cancer, but as yet there is no evidence of this occurrence. An omission in the literature on hVDR polymorphisms and their pathophysiologic impact is that many investigators limit their analyses to only the F/f or the 3′ UTR polymorphisms in isolation, instead of determining the complete genotype of the patient cohort. Examining the genotype at only one site cannot accurately predict the bioactivity of the hVDR produced from that allele, and will thus give an incomplete picture of the true impact of these polymorphisms for human disease. Published results in which only one site is analyzed (or multiple 3′ UTR sites without either the FokI or the Cdx-2 sites) tend to be inconsistent and even contradictory. This, it could be argued, may be a contributing factor to the current state of controversy in the field. Recently, monitoring the transcriptional potency of endogenous hVDR in cultured human skin fibroblasts, we determined that a combination of FFLL homozygous alleles produced hVDR with the highest transcriptional activity (up to 30-fold more potent than ffSS), and genetic mixtures containing
246 the less active hVDR alleles ( f and S) correspondingly compromised the transcriptional potency of hVDR in heterozygotes such as FfLS [238]. Thus, because 5′ and 3′ hVDR polymorphisms can counteract one another in terms of functional activity and clinical impact, the influence of combined polymorphisms should be examined in the future, perhaps utilizing the allele score concept [238] to account for genetic variation at all relevant sites. The ultimate genotype analysis would be to determine complete hVDR gene haplotypes [249]. This analysis presents considerable technical difficulties; indeed, it may be quite impractical to include the FokI restriction endonuclease site determining f versus F and the 3′ cluster of variants, given the large distance between them in the genome (the FokI site in exon II is >30 kb from the BsmI and ApaI sites between exons VIII and IX and nearly 40 kb from the singlet(A) repeat in exon IX) [277]. Nevertheless, results from such studies, even if limited in scope, could provide valuable insights. Several groups have sought to identify polymorphic variants in VDR target genes that are likely associated with human disease, including osteocalcin [278], CYP19 [279], the cyclin-dependent kinase inhibitors p21 and p27 [280], and CYP3A4 [281]. This approach is promising, especially when the target gene is actually involved in a physiologic process related to the disease in question, for example CYP19 (aromatase) in the etiology of osteoporosis in postmenopausal women [279]. In fact, a complete assessment of the genetic component for disease risk would ideally account for variations not only in VDR, but also in RXR, VDR coactivators, and relevant target genes. Finally, disorders with newly reported relationships to VDR polymorphisms include: type I diabetes mellitus (higher risk with genotypes At and Bt) [282], susceptibility to the hemorrhagic form of dengue fever (with the t genotype displaying greater resistance) [283], anemia in hemodialysis patients (the BB genotype has lower hemoglobin levels) [284], and various dental indices (AA patients had the highest rates of alveolar bone loss) [285]. Many of these disorders have an immune component, and renewed interest in the immune effects of the VDR-RXR signaling system has received impetus from the recent observation that VDR knockout mice, contrary to early reports of a normal immune system, display an abnormal T-cell response [86].
B. Emerging Clinically Relevant Actions of VDR in the Colon and Hair Follicle In addition to the noncalcemic functions of VDR that have been discussed earlier in this chapter, such as
G. KERR WHITFIELD
ET AL .
modulation of the immune system and prevention of epithelial cell cancers, two striking new extraosseous actions of VDR have come into focus recently, as illustrated in Fig. 12. One of these is to drive the hair cycle in skin, which is unique in that, as explained later, it requires neither a vitamin D-derived ligand [155] nor the transactivation function of VDR [153] (Fig. 12B). Therefore, VDR impinges on hair cycle signal transduction in keratinocytes either in an unliganded form or perhaps occupied by an undiscovered lipophilic ligand present in skin, and does so apparently by repressing a VDR target gene(s) [157]. This gene repression in the hair follicle may be mediated by the recently demonstrated physical and functional interaction between VDR-RXR and the Hr nuclear receptor corepressor product of the hairless gene [157]. The second novel noncalcemic action of VDR, the detoxification of carcinogenic LCA in the colon (Fig. 12A), entails the binding of the newly discovered LCA ligand to VDR to induce CYPs [98]. Thus VDR has emerged as a secondary bile acid sensor as well as calcemic endocrine nuclear receptor [98]. VDR and vitamin D have been implicated in the prevention of colon cancer through epidemiologic observations that regions of the world with the lowest sunlight exposure coincide with high incidences of both rickets and colon cancer [286]. Furthermore, vitamin D supplementation reduces dietary fat-promoted colon carcinogenesis in rats [287], and VDR-null mice exhibit precocious hyperproliferation in the colon descendens where VDR is normally expressed in levels that rival those in skin and small intestine [288]. The paradigm whereby VDR/vitamin D suppresses colon cancer has not been fully characterized, and could involve the ability of VDR to arrest cells at the G1 stage of the cell cycle via induction of p21 [88] and p27 [289], to repress cell growth transcription factors such as c-myc [90] and c-fos [91], or to elicit apoptosis by diminishing the Bcl-2 anti-apoptotic factor [92]. A newly discovered mechanism through which VDR could prevent colon cancer, particularly that induced by high-fat Western diets, is that of chemoprevention by inducing CYP3A4 to detoxify LCA in the colon cells [56,98]. Figure 12A illustrates the role for VDR-RXR in the regulation of intestinal CYP enzymes, in particular CYP3A4. This concept is based upon the recent appreciation that the secondary bile acid, LCA, and its 3-keto metabolite, are novel VDR ligands [98]. Figure 12A also depicts the hepatic synthesis of primary bile acids, focusing on chenodeoxycholic acid, the precursor of LCA. The rate limiting enzyme in this biosynthetic process, CYP7A1, is under positive as well as negative control by VDR-related receptors LXR and FXR, respectively [290]. High liver cholesterol leads
247
CHAPTER 13 Nuclear Vitamin D Receptor
A
Oxysterols
+
LIVER
LXR CYP 7A1 Cholesterol FXR − Chenodeoxycholic acid
Gut bacteria 7-dehydroxylation COO−
LCA HO
H VDR
RXR
LCA
+
Colonocyte
CYP3A4 LCA 6αOH LCA
COLON Colon xenobiotic detoxification
B
ABC transporter 6αOH LCA in lumen
Noggin
Wnt ligand Fz Dsh
BMP4
PTHrP
BMPR
PTHR
Keratinocyte ? β-catenin
SMADs
?
1 BULB
HAIR FOLLICLE
Dermal papilla
2
RXRα
Shh, Hoxc13, msx-1, msx-2
Telogen
FIGURE 12
Hr
Lef1/TCF
novel ligand VDR ?
Anagen
Catagen
Two novel extraosseous actions of the VDR-RXR heterodimer: xenobiotic detoxification in the colon and control of hair cycling. (A) Role of VDR-RXR in detoxification of the secondary bile acid, LCA. The rate-limiting step in the formation of bile acids from cholesterol in liver is catalyzed by CYP7A1 [116]. One of the major primary bile acids, chenodeoxycholic acid, can be converted by gut bacteria to LCA, which may accumulate in the colon since it is not effectively reabsorbed. In the colonocytes, LCA-bound VDR (or 1,25(OH)2D3liganded VDR, not shown), heterodimerized with RXR, activates transcription of the gene encoding CYP3A4. The CYP3A4 enzyme then converts LCA to its 6α-hydroxy form, which is a substrate for the ABC efflux transporter. (B) Proposed action of the VDR-RXR heterodimer in the hair follicle. Keratinocytes or their precursors in the bulge of the follicle are thought to receive signals that stimulate the follicle to exit a resting phase (telogen) and enter a phase of active growth (anagen). Recent research [295,299,300] has identified a number of factors that influence this transition, and a simplified diagram is presented here. Abbreviations not defined in the text are Wnt, ortholog of Drosophila wingless and mouse int-1; Lef1, lymphoid enhancer factor-1; TCF, T cell-specific factor; and msx-1 and msx-2, orthologs of Drosophila muscle-specific homeobox protein. Factors that are membrane receptors are boxed. Solid arrows indicate activation and dotted lines ending in a solid perpendicular line denote inhibition. The black circled 1 and 2 in the BMP/SMAD pathway represent sequential negative and positive signaling as described in the text. The precise point(s) at which VDR-RXR or the RXR-VDR-Hr complex acts has not been characterized, and it is not known whether VDR functions unliganded or occupied by a novel, non-vitamin D ligand.
248 to the production of oxysterols (Fig. 4), which are known to bind to LXR and induce the CYP7A1 gene that catalyzes the first step in bile acid synthesis via 7α-hydroxylation [116]. An opposing regulation of CYP7A1 occurs as a result of high levels of primary bile acids in liver (e.g., chenodeoxycholic acid), which bind to FXR (Fig. 4) and, via an indirect mechanism, inhibit the transcription of CYP7A1 [291]. Formation of the secondary bile acid, LCA, by intestinal bacteria is effected by 7-dehydroxylation of chenodeoxycholic acid [292,293]. LCA then travels to the colon, where its potentially toxic or carcinogenic effects [99,100] are proposed to be ameliorated by the catabolic action of CYP3A4 in the colonocyte (Fig. 12A) [105]. The CYP3A4 gene is transactivated by LCA-bound [98] (or 1,25(OH)2D3bound [55,56]) VDR-RXR, leading to 6α-hydroxylation of LCA [105] and its secretion via an ATP binding cassette (ABC) transporter into the colonic lumen for excretion [290]. Regulation of CYP3A4 was previously known to occur via PXR [59], but experiments with knockout mice have clearly demonstrated that another receptor, presumably VDR [98], is capable of mediating this regulation in PXR−/− mice [55,98]. Therefore, we propose that, in addition to its antiproliferation/ prodifferentiation and proapoptotic actions, VDR specifically operates to prevent high dietary fat-induced colon cancer by detoxifying LCA. This chemoprotective action of VDR via CYP3A4 induction occurs when the receptor is occupied by its alternative ligands, LCA or 1,25(OH)2D3, explaining both how LCA is detoxified, and why this process is amplified in the presence of supplemental vitamin D. A second noncalcemic site of VDR action, in this case clinically relevant to the prevention of hair loss, is the hair follicle (Fig. 12B). As described earlier, alopecia in VDR knockout mice cannot be corrected by excess dietary calcium [79]. Also, vitamin D- or 1,25(OH)2D3deficiency, the latter caused by ablation of the CYP27B1 1α-OHase enyzme [155,156], does not confer alopecia. Thus, the action of VDR to trigger normal hair cycling is independent of vitamin D status, indicating that VDR functions in this capacity either in the unliganded conformation, or occupied by a nonvitamin D-derived lipid constitutively present in skin. Interestingly, conditional knockout of RXRα in skin [128] elicits alopecia indistinguishable from that occurring in VDR-null mice, implying that the functional VDR unit, as with 1,25(OH)2D3-dependent actions, is the VDR-RXR binary complex. Further experiments utilizing VDR knockout animals [294], and with VDR−/− mice in which VDR expression is transgenically targeted to keratinocytes [82], have demonstrated that the absence of VDR function renders hair follicles unable to
G. KERR WHITFIELD
ET AL .
enter the growth (anagen) phase. Therefore, as depicted in Fig. 12B, VDR appears to function in keratinocytes to trigger the telogen (resting phase) to anagen (growth phase) transition in the hair cycle. How does VDR impinge upon hair cycle signaling? According to the hypothetical model presented in Fig. 12B, VDR-RXR could intervene at any of a number of steps in the complex process of anagen initiation. Findings mainly from the laboratory of Fuchs et al. [295] have revealed that the Lef/TCF family of transcription factors, complexed with β-catenin, play a crucial role in the transmodulation of several genes which encode proteins that lead to the initiation of anagen. The list of genes directly or indirectly activated by the Lef/TCF/βcatenin complex includes sonic hedgehog (Shh), Hoxc13, msx-1, and msx-2 [295]. This complex also activates hair specific genes such as keratin 14, and represses E-cadherin, the latter effect diminishing adherens junctions to facilitate hair follicle formation [296]. As illustrated in Fig. 12B, control of Lef1/TCF transcription factor includes at least two upstream pathways, one originating with the binding of an extracellular Wnt ligand to a cell surface receptor of the Frizzled (Fz) family. This signal, acting through the Disheveled factor (Dsh), leads to a stabilization of β-catenin and formation of an active complex with Lef1/TCF. The second pathway involves an inhibitory factor, bone morphogenetic protein-4 (BMP4), that functions in a negative manner to block synthesis of Lef1/TCF, and this inhibition can be relieved by association with noggin, an extracellular ligand secreted by the dermal papilla that acts as an antagonist at the BMP receptor (BMPR) [297,298]. Based upon recent experiments [298] in which the BMPR was knocked out, a second function of BMP/SMAD signaling has been revealed subsequent to its antagonism by noggin to permit Lef1 accumulation in hair follicle matrix cells. Once Lef1 has accumulated in response to the noggin antagonist, this is followed by a positive effect of BMP/SMAD to somehow facilitate the cooperative action of β-catenin and Lef1 to modulate transcription in the cortex cells that generate the hair shaft. Therefore, these pathways all converge in the genesis of the transcriptionally active Lef1/TCF/ β-catenin complex (Fig. 12B). To date, VDR has been studied only in relation to β-catenin in colon cancer cells, and paradoxically, the receptor attenuates β-catenin signaling by competing with Lef1/TCF for β-catenin binding [270]. However, the potential actions of VDRRXR to influence either the Wnt or BMP/noggin pathways in keratinocytes have not as yet been probed, and the receptor heterodimer could affect any number of steps in hair-cycle signal transduction as postulated in Fig. 12B.
249
CHAPTER 13 Nuclear Vitamin D Receptor
A third pathway initiated by yet another extracellular ligand, PTHrP, may be involved in triggering anagen, although recent reports indicate that PTHrP accelerates entry into catagen (apoptosis of the lower follicle) rather that preventing entry into anagen [299,300]. Regardless of which hypothesis for PTHrP action is correct, effective hair growth during anagen would require either inhibition of PTHrP synthesis and secretion, or repression of its receptor (PTHR), in target cells, and as postulated in Fig. 12B, VDR could act as a repressor at either of these steps. Interestingly, transcription of PTHrP is directly inhibited by the 1,25(OH)2D3-occupied VDR-RXR heterodimer through negative VDREs [301,302] (Table I), rendering PTHrP as one of the most attractive candidate targets for repression by VDR in hair follicle signaling. The mechanism whereby VDR-RXR triggers the hair cycle remains unknown. However, naturally occurring point mutations in hVDR that confer alopecia as well as vitamin D-resistant rickets exist in the hVDR DNA binding [74,303,304] and RXR heterodimerization [147] domains, yet mutations that specifically compromise either 1,25(OH)2D3 ligand [147,305] or coactivator [153] contacts yield rickets without alopecia. Thus, neither 1,25(OH)2D3 ligand contacts nor receptor-mediated transactivation is required for VDR to drive the hair cycle; only the RXR heterodimerization and DNA binding functions of VDR appear to be necessary for this action. A tonic inhibitor of hair growth has been hypothesized to exist in telogen skin [306], and a reasonable molecular scenario is that VDR-RXR functions to repress this inhibitor, which may or may not be BMP4 or PTHrP (Fig. 12B). Mechanisms of gene repression by VDR are not well understood (see Section II), and VDR does not associate strongly with traditional nuclear receptor corepressors such as N-CoR and SMRT [307–309]. Recently, VDR has been observed to interact physically and functionally with a novel corepressor, the Hr product of the hairless (hr) gene [157]. Hr recruits HDACs to repress the actions of both TR [310] and ROR [311], and it dramatically inhibits both basal and ligand-stimulated VDRRXR-mediated transcription [157]. Intriguingly, similar to knockout of the genes encoding either VDR or RXRα, inactivating mutations in the mammalian hr gene result in congenital hair loss in both mice [312] and humans [313,314]. Remarkably, the hair loss phenotype elicited by specific mutations in hVDR resembles the atrichia resulting from mutations in the hr gene [315], suggesting that the VDR, RXRα, and Hr proteins all impact a common signaling pathway, perhaps as a trimeric complex (Fig. 12B). Coexpression of hVDR and Hr mRNAs in the matrix and outer root sheath cells of the
mouse skin hair follicle [157] supports this hypothesis. What remains to be determined is where the putative RXRα-VDR-Hr complex impinges on hair cycle signal transduction to trigger the telogen-to-anagen transition, and several possibilities are illustrated in Fig. 12B. Elucidating the molecular role of VDR in controlling the hair cycle clearly represents an important challenge, one that may lead us to fundamental new insight into novel gene repressive mechanisms of VDR action. Finally, understanding this mechanism may facilitate the development of VDR mimetics to prevent or treat hair loss.
VII. SUMMARY AND PERSPECTIVES The vitamin D receptor is a unique member of the nuclear receptor superfamily in the realm of ligand binding, mediating both the endocrine control of bone mineral metabolism and the detoxification of endobiotics such as LCA. VDR functions as a heterodimer with unliganded RXR, with the binary protein complex required for recognition and high-affinity association with VDREs in the promoters of regulated genes, as well as for the recruitment of transcriptional comodulator proteins that govern the activity of RNA polymerase II. As a nuclear receptor that heterodimerizes with RXR, VDR therefore possesses the molecular character both of its evolutionarily closest endocrine relative, the thyroid hormone receptor, and of its closest xenobiotic detoxifying receptor, PXR. The biology of VDR signaling primarily involves (i) stimulation of intestinal calcium and phosphate absorption to prevent rickets/ osteomalacia, (ii) enhancement of bone remodeling via osteoblast-induced osteoclast maturation, (iii) differentiation of skin cells and maintenance of the hair cycle, (iv) repression/induction of CYP enzymes for 1,25(OH)2D3 hormone synthesis/degradation as well as for the promotion of secondary bile acid detoxification, (v) modulation of the immune system, and (vi) potential anticancer actions via the control of epithelial cell growth, differentiation, and apoptosis. Based upon biochemical mutagenesis experiments, the X-ray crystallographic characterization of the DNAand ligand-binding domains, and naturally occurring mutations that confer on patients the vitamin D-resistant phenotype of rickets and/or alopecia, the structurefunction of VDR is reasonably well understood. Amino acid residues in an α-helical region of the zinc finger domain of VDR that contact specific DNA base pairs in VDRE docking sites have been identified, as have nuclear localization signals in this domain. The more structurally complex LBD of VDR consists of a sandwich
250 of α-helices surrounding a hydrophobic pocket that accommodates lipophilic ligands, as well as surface facets for the attraction of numerous interacting proteins including the RXR dimerization partner. The VDR-RXR heterodimer, which is allosterically influenced by 1,25(OH)2D3 and other VDR ligands, associates with (i) ATP-dependent SWI/SNF chromatin remodeling complexes (e.g., PBAF and WINAC) that apparently slide nucleosomes along DNA and are required for both induction and repression; (ii) p160 platform coactivators (e.g., SRC-1) that recruit CBP/p300 to effect HAT modification of chromatin that disrupts internucleosomal interactions to convert heterochromatin to transcriptionally active euchromatin; and (iii) DRIP205/Mediator complexes that bridge to the C-terminal domain of RNA polymerase II plus TFIIB, which links VDR-RXR to transcriptional initiation. Alternatively, VDR-RXR may associate with corepressors such as Hr that attract HDACs to return chromatin structure to the transcriptionally inactive heterochromatin state in the vicinity of gene promoters negatively regulated by VDR. The functional activity of hVDR appears to be influenced by common gene polymorphisms, with certain 5′ promoter (Cdx-2), translation start site (F/f ), and 3′-UTR (L/S) alleles yielding higher expression of (Cdx-2 A-allele and 3′-UTR L-allele), or intrinsically more transcriptionally active (translation start site F-allele), hVDR. Such polymorphic hVDR variants affect bone mineral density and may relate to the epidemiology of osteoporosis, as well as significantly alter the risk of certain epithelial cancers. In conclusion, in the time elapsed since the last edition of this volume, we have witnessed an explosion of new information enhancing our understanding of VDR, including crystallographic views of both the DBD and ligand binding/heterodimerization domains, availability of the complete human genome sequence, and thorough characterizations of VDR and RXR knockout mice. However, many questions still remain unanswered. A major structure-function breakthrough would be the X-ray crystallographic solution of fulllength VDR, or at least of a closely related nuclear receptor, so that we can begin to understand the interactions between the DBD and the ligand binding/ heterodimerization domains. A second major realm of future investigation is the identification of additional VDR-regulated genes, particularly those supporting intestinal calcium and phosphate transport, as the molecular mechanism whereby VDR exerts this primary action is currently uncharacterized. An equally exciting endeavor will be the discovery of additional natural VDR or RXR ligands, especially in the more recently identified and studied VDR target tissues such as the hair follicle/skin and immune systems. The availability
G. KERR WHITFIELD
ET AL .
of this information would allow for molecular investigations of the VDR transcriptional control cycle (Fig. 10) in the presence of novel ligands and/or in the context of other promoters in addition to those of the rat osteocalcin or mouse osteopontin genes that have heretofore been the focus of VDR transactivation studies. Such experiments will likely confirm and extend our understanding of the variety of conformations and coregulator associations that the VDR-RXR heterodimer is capable of achieving. Moreover, insights from the approaches just outlined may then enable us to comprehend the interactions between two or more VDREs in the same vitamin D regulated promoter, or between the VDR-RXR heterodimer and other transactivators that are bound to the same promoter. Finally, all of these concepts need to be applied to the poorly understood topic of negative regulation by the VDRRXR heterodimer. Transrepression by VDR is likely complex, and no doubt involves a variety of mechanisms, but its characterization will increase our knowledge of such fundamental VDR tenets as suppression of PTH and the 1α-OHase enzyme in relation to calcium/ phosphate homeostasis, control of PTHrP and other hair cycle-related genes with respect to preventing alopecia, and regulation of Bcl-2 and other genes controlling cell growth and apoptosis that may play a role in the action of VDR as a nuclear receptor sentinel in preventing epithelial cell cancers. Clearly, there are many challenges ahead in our molecular probing of VDR and, undoubtedly, a few more surprises as well.
Acknowledgments Supported by NIH grants to M.R.H. The authors thank other members of our laboratory, Hope T. L. Dang, Jamie Dawson, Neal Hall, Magdalena J. Kaczmarska, and Stephanie A. Slater, for their contributions.
References 1. Haussler MR, Myrtle JF, Norman AW 1968 The association of a metabolite of vitamin D3 with intestinal mucosa chromatin in vivo. J Biol Chem 243:4055–4064. 2. Haussler MR, Norman AW 1969 Chromosomal receptor for a vitamin D metabolite. Proc Natl Acad Sci USA 62:155–162. 3. Myrtle JF, Haussler MR, Norman AW 1970 Evidence for the biologically active form of cholecalciferol in the intestine. J Biol Chem 245:1190–1196. 4. Haussler MR, Boyce DW, Littledike ET, Rasmussen H 1971 A rapid acting metabolite of vitamin D3. Proc Natl Acad Sci USA 68:177–181. 5. Holick MF, Schnoes HK, DeLuca HF, Suda T, Cousins RJ 1971 Isolation and identification of 1,25-dihydroxycholecalciferol, a metabolite of vitamin D active in the intestine. Biochemistry 10:2799–2804.
251
CHAPTER 13 Nuclear Vitamin D Receptor
6. Lawson DEM, Fraser DR, Kodicek E, Morris HR, Williams DH 1971 Identification of 1,25-dihydroxycholecalciferol, a new kidney hormone controlling calcium metabolism. Nature 230:228–230. 7. Fraser DR, Kodicek E 1970 Unique biosynthesis by kidney of a biologically active vitamin D metabolite. Nature 228: 764–766. 8. Brumbaugh PF, Haussler MR 1974 1α,25-Dihydroxycholecalciferol receptors in intestine. I. Association of 1α,25dihydroxycholecalciferol with intestinal mucosa chromatin. J Biol Chem 249:1251–1257. 9. Brumbaugh PF, Haussler MR 1974 1α,25-Dihydroxycholecalciferol receptors in intestine. II. Temperature-dependent transfer of the hormone to chromatin via a specific cytosol receptor. J Biol Chem 249:1258–1262. 10. Brumbaugh PF, Haussler MR 1973 1α,25-Dihydroxyvitamin D3 receptor: Competitive binding of vitamin D analogs. Life Sci 13:1737–1746. 11. Norman AW, Procsal DA, Okamura WH, Wing RM 1975 Structure-function studies of the interaction of the hormonally active form of vitamin D3, 1α,25-dihydroxy-vitamin D3, with the intestine. J Steroid Biochem 6:461–467. 12. Pike JW, Haussler MR 1979 Purification of chicken intestinal receptor for 1,25-dihydroxyvitamin D. Proc Natl Acad Sci USA 76:5485–5489. 13. Pike JW, Donaldson CA, Marion SL, Haussler MR 1982 Development of hybridomas secreting monoclonal antibodies to the chicken intestinal 1α,25-dihydroxyvitamin D3 receptor. Proc Natl Acad Sci USA 79:7719–7723. 14. Pike JW, Marion SL, Donaldson CA, Haussler MR 1983 Serum and monoclonal antibodies against the chick intestinal receptor for 1,25-dihydroxyvitamin D3. J Biol Chem 258:1289–1296. 15. McDonnell DP, Mangelsdorf DJ, Pike JW, Haussler MR, O’Malley BW 1987 Molecular cloning of complementary DNA encoding the avian receptor for vitamin D. Science 235:1214–1217. 16. Evans RM 1988 The steroid and thyroid hormone receptor superfamily. Science 240:889–895. 17. Pike JW 1981 Evidence for a reactive sulfhydryl in the DNA binding domain of the 1,25-dihydroxyvitamin D3 receptor. Biochem Biophys Res Commun 100:1713–1719. 18. Haussler MR, Mangelsdorf DJ, Yamaoka K, Allegretto EA, Komm BS, Terpening CM, McDonnell DP, Pike JW, O’Malley BW 1988 Molecular characterization and actions of the vitamin D hormone receptor. In: Ringold G, ed. Steroid Hormone Action, Vol. 75. Alan R. Liss, New York, pp. 247–262. 19. Baker AR, McDonnell DP, Hughes MR, Crisp TM, Mangelsdorf DJ, Haussler MR, Pike JW, Shine J, O’Malley BW 1988 Cloning and expression of full-length cDNA encoding human vitamin D receptor. Proc Natl Acad Sci USA 85:3294–3298. 20. Burmester JK, Wiese RJ, Maeda N, DeLuca HF 1988 Structure and regulation of the rat 1,25-dihydroxyvitamin D3 receptor. Proc Natl Acad Sci USA 85:9499–9502. 21. Lu Z, Hanson K, DeLuca HF 1997 Cloning and origin of the two forms of chicken vitamin D receptor. Arch Biochem Biophys 339:99–106. 22. Sone T, Ozono K, Pike JW 1991 A 55-kilodalton accessory factor facilitates vitamin D receptor DNA binding. Mol Endocrinol 5:1578–1586. 23. MacDonald PN, Haussler CA, Terpening CM, Galligan MA, Reeder MC, Whitfield GK, Haussler MR 1991 Baculovirusmediated expression of the human vitamin D receptor: functional characterization, vitamin D response element
24.
25. 26. 27.
28. 29.
30. 31. 32.
33. 34.
35.
36. 37.
38.
39.
interactions, and evidence for a receptor auxiliary factor. J Biol Chem 266:18808–18813. Ross TK, Moss VE, Prahl JM, DeLuca HF 1992 A nuclear protein essential for binding of rat 1,25-dihydroxyvitamin D3 receptor to its response elements. Proc Natl Acad Sci USA 89:256–260. Mangelsdorf DJ, Evans RM 1995 The RXR heterodimers and orphan receptors. Cell 83:841–850. Heyman RA, Mangelsdorf DJ, Dyck JA, Stein RB, Eichele G, Evans RM, Thaller C 1992 9-Cis retinoic acid is a high affinity ligand for the retinoid X receptor. Cell 68:397–406. MacDonald PN, Dowd DR, Nakajima S, Galligan MA, Reeder MC, Haussler CA, Ozato K, Haussler MR 1993 Retinoid X receptors stimulate and 9-cis retinoic acid inhibits 1,25-dihydroxyvitamin D3-activated expression of the rat osteocalcin gene. Mol Cell Biol 13:5907–5917. Jin CH, Kerner SA, Hong MH, Pike JW 1996 Transcriptional activation and dimerization functions in the human vitamin D receptor. Mol Endocrinol 10:945–957. Cheskis B, Freedman LP 1994 Ligand modulates the conversion of DNA-bound vitamin D3 receptor (VDR) homodimers into VDR-retinoid X receptor heterodimers. Mol Cell Biol 14:3329–3338. Lemon BD, Freedman LP 1996 Selective effects of ligands on vitamin D3 receptor- and retinoid X receptor-mediated gene activation in vivo. Mol Cell Biol 16:1006–1016. Cheskis B, Freedman LP 1996 Modulation of nuclear receptor interactions by ligands: Kinetic analysis using surface plasmon resonance. Biochemistry 35:3309–3318. Thompson PD, Jurutka PW, Haussler CA, Whitfield GK, Haussler MR 1998 Heterodimeric DNA binding by the vitamin D receptor and retinoid X receptors is enhanced by 1,25-dihydroxyvitamin D3 and inhibited by 9-cis retinoic acid: evidence for allosteric receptor interactions. J Biol Chem 273:8483–8491. MacDonald PN, Dowd DR, Haussler MR 1994 New insight into the structure and functions of the vitamin D receptor. Sem Nephrol 14:101–118. Haussler MR, Jurutka PW, Hsieh J-C, Thompson PD, Selznick SH, Haussler CA, Whitfield GK 1995 New understanding of the molecular mechanism of receptor-mediated genomic actions of the vitamin D hormone. Bone 17(Suppl 2): 33S–38S. Haussler MR, Whitfield GK, Haussler CA, Hsieh J-C, Thompson PD, Selznick SH, Encinas Dominguez C, Jurutka PW 1998 The nuclear vitamin D receptor: biological and molecular regulatory properties revealed. J Bone Miner Res 13:325–349. Cheng JB, Motola DL, Mangelsdorf DJ, Russell DW 2003 De-orphanization of cytochrome P450 2R1: A microsomal vitamin D 25-hydroxylase. J Biol Chem 278:38084–38093. DeMay MB, Kiernan MS, DeLuca HF, Kronenberg HM 1992 Sequences in the human parathyroid hormone gene that bind the 1,25-dihydroxyvitamin D3 receptor and mediate transcriptional repression in response to 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 89:8097–8101. Kitazawa R, Kitazawa S 2002 Vitamin D3 augments osteoclastogenesis via vitamin D-responsive element of mouse RANKL gene promoter. Biochem Biophys Res Commun 290:650–655. Yasuda H, Shima N, Nakagawa N, Yamaguchi K, Kinosaki M, Mochizuki S, Tomoyasu A, Yano K, Goto M, Murakami A, Tsuda E, Morinaga T, Higashio K, Udagawa N, Takahashi N, Suda T 1998 Osteoclast differentiation factor is a ligand for
252
40.
41.
42. 43. 44. 45.
46. 47. 48.
49.
50.
51.
52.
53.
54.
G. KERR WHITFIELD
osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc Natl Acad Sci USA 95:3597–3602. Terpening CM, Haussler CA, Jurutka PW, Galligan MA, Komm BS, Haussler MR 1991 The vitamin D-responsive element in the rat bone gla protein is an imperfect direct repeat that cooperates with other cis-elements in 1,25dihydroxyvitamin D3-mediated transcriptional activation. Mol Endocrinol 5:373–385. Noda M, Vogel RL, Craig AM, Prahl J, DeLuca HF, Denhardt DT 1990 Identification of a DNA sequence responsible for binding of the 1,25-dihydroxyvitamin D3 receptor and 1,25-dihydroxyvitamin D3 enhancement of mouse secreted phosphoprotein 1 (Spp-1 or osteopontin) gene expression. Proc Natl Acad Sci USA 87:9995–9999. Horowitz MC 1993 Cytokines and estrogen in bone: anti-osteoporotic effects. Science 260:626–627. Bikle DD, Pillai S 1993 Vitamin D, calcium and epidermal differentiation. Endocr Rev 14:3–19. Manolagas SC, Yu X-P, Girasole G, Bellido T 1994 Vitamin D and the hematolymphopoietic tissue: A 1994 update. Sem Nephrol 14:129–143. Bikle DD, Ng D, Oda Y, Hanley K, Feingold K, Xie Z 2002 The vitamin D response element of the involucrin gene mediates its regulation by 1,25-dihydroxyvitamin D3. J Invest Dermatol 119:1109–1113. Quelo I, Machuca I, Jurdic P 1998 Identification of a vitamin D response element in the proximal promoter of the chicken carbonic anhydrase II gene. J Biol Chem 273:10638–10646. Omdahl JL, Morris HA, May BK 2002 Hydroxylase enzymes of the vitamin D pathway: expression, function, and regulation. Ann Rev Nutr 22:139–166. Adams JS, Singer FR, Gacad MA, Sharma OP, Hayes MJ, Vouros P, Holick MF 1985 Isolation and structural identification of 1,25-dihydroxyvitamin D3 produced by cultured alveolar macrophages in sarcoidosis. J Clin Endocrinol Metab 60:960–966. Ohyama Y, Ozono K, Uchida M, Shinki T, Kato S, Suda T, Yamamoto O, Noshiro M, Kato Y 1994 Identification of a vitamin D-responsive element in the 5′ flanking region of the rat 25-hydroxyvitamin D3 24-hydroxylase gene. J Biol Chem 269:10545–10550. Zierold C, Darwish HM, DeLuca HF 1994 Identification of a vitamin D-responsive element in the rat calcidiol (25-hydroxyvitamin D3) 24-hydroxylase gene. Proc Natl Acad Sci USA 91:900–902. Murayama A, Takeyama K, Kitanaka S, Kodera Y, Kawaguchi Y, Hosoya T, Kato S 1999 Positive and negative regulations of the renal 25-hydroxyvitamin D3 1α-hydroxylase gene by parathyroid hormone, calcitonin, and 1α,25(OH)2D3 in intact animals. Endocrinology 140:2224–2231. Colnot S, Lambert M, Blin C, Thomasset M, Perret C 1995 Identification of DNA sequences that bind retinoid X receptor1,25(OH)2D3-receptor heterodimers with high affinity. Mol Cell Endocrinol 113:89–98. Nishikawa J, Kitaura M, Matsumoto M, Imagawa M, Nishihara T 1994 Difference and similarity of DNA sequence recognized by VDR homodimer and VDR/RXR heterodimer. Nucleic Acids Res 22:2902–2907. Jurutka PW, MacDonald PN, Nakajima S, Hsieh J-C, Thompson PD, Whitfield GK, Galligan MA, Haussler CA, Haussler MR 2002 Isolation of baculovirus-expressed human vitamin D receptor: DNA responsive element interactions and phosphorylation of the purified receptor. J Cell Biochem 85:435–457.
ET AL .
55. Thummel KE, Brimer C, Yasuda K, Thottassery J, Senn T, Lin Y, Ishizuka H, Kharasch E, Schuetz J, Schuetz E 2001 Transcriptional control of intestinal cytochrome P-4503A by 1alpha,25-dihydroxy vitamin D3. Mol Pharmacol 60: 1399–1406. 56. Thompson PD, Jurutka PW, Whitfield GK, Myskowski SM, Eichhorst KR, Encinas Dominguez C, Haussler CA, Haussler MR 2002 Liganded VDR induces CYP3A4 in small intestinal and colon cancer cells via DR3 and ER6 vitamin D responsive elements. Biochem Biophys Res Commun 299:730–738. 57. Bertilsson G, Heidrich J, Svensson K, Asman M, Jendeberg L, Sydow-Backman M, Ohlsson R, Postlind H, Blomquist P, Berkenstam A 1998 Identification of a human nuclear receptor defines a new signaling pathway for CYP3A induction. Proc Natl Acad Sci USA 95:12208–12213. 58. Blumberg B, Sabbagh W, Jr., Juguilon H, Bolado J, Jr., van Meter CM, Ong ES, Evans RM 1998 SXR, a novel steroid and xenobiotic-sensing nuclear receptor. Genes Dev 12:3195–3205. 59. Lehmann JM, McKee DD, Watson MA, Willson TM, Moore JT, Kliewer SA 1998 The human orphan nuclear receptor PXR is activated by compounds that regulate CYP3A4 gene expression and cause drug interactions. J Clin Invest 102:1016–1023. 60. Hartong R, Wang N, Kurokawa R, Lazar MA, Glass CK, Apriletti JW, Dillmann WH 1994 Delineation of three different thyroid hormone-response elements in promoter of rat sarcoplasmic reticulum Ca2+ ATPase gene. J Biol Chem 269: 13021–13029. 61. Baniahmad A, Steiner C, Kohne AC, Renkawitz R 1990 Modular structure of a chicken lysozyme silencer: involvement of an unusual thyroid hormone receptor binding site. Cell 61:505–514. 62. van Den Bemd GJ, Jhamai M, Staal A, van Wijnen AJ, Lian JB, Stein GS, Pols HA, van Leeuwen JP 2002 A central dinucleotide within vitamin D response elements modulates DNA binding and transactivation by the vitamin D receptor in cellular response to natural and synthetic ligands. J Biol Chem 277:14539–14546. 63. Staal A, van Wijnen AJ, Birkenhäger JC, Pols HAP, Prahl J, DeLuca H, Gaub M-P, Lian JB, Stein GS, van Leeuwen JPTM, Stein JL 1996 Distinct conformations of vitamin D receptor/ retinoid X receptor-α heterodimers are specified by dinucleotide differences in the vitamin D-responsive elements for the osteocalcin and osteopontin genes. Mol Endocrinol 10:1444–1456. 64. Koszewski NJ, Ashok S, Russell J 1999 Turning a negative into a positive: vitamin D receptor interactions with the avian parathyroid hormone response element. Mol Endocrinol 13:455–465. 65. Kurokawa R, Söderström M, Hörlein A, Halachmi S, Brown M, Rosenfeld MG, Glass CK 1995 Polarity-specific activities of retinoic acid receptors determined by a co-repressor. Nature 377:451–454. 66. Jurutka PW, Whitfield GK, Hsieh J-C, Thompson PD, Haussler CA, Haussler MR 2001 Molecular nature of the vitamin D receptor and its role in regulation of gene expression. Rev Endocr Metab Disord 2:203–216. 67. Lyakhovich A, Aksenov N, Pennanen P, Miettinen S, Ahonen MH, Syvala H, Ylikomi T, Tuohimaa P 2000 Vitamin D induced up-regulation of keratinocyte growth factor (FGF7/KGF) in MCF-7 human breast cancer cells. Biochem Biophys Res Commun 273:675–680. 68. Farach-Carson MC, Xu Y 2002 Microarray detection of gene expression changes induced by 1,25(OH)2D3 and a Ca2+
CHAPTER 13 Nuclear Vitamin D Receptor
69. 70. 71.
72.
73. 74. 75.
76.
77.
78.
79.
80.
81. 82.
83.
influx-activating analog in osteoblastic ROS 17/2.8 cells. Steroids 67:467–470. Krishnan AV, Peehl DM, Feldman D 2003 Inhibition of prostate cancer growth by vitamin D: Regulation of target gene expression. J Cell Biochem 88:363–371. Li X, Zheng W, Li YC 2003 Altered gene expression profile in the kidney of vitamin D receptor knockout mice. J Cell Biochem 89:709–719. Swami S, Raghavachari N, Muller UR, Bao YP, Feldman D 2003 Vitamin D growth inhibition of breast cancer cells: gene expression patterns assessed by cDNA microarray. Breast Cancer Res Treat 80:49–62. Feldman D, Malloy PJ 1990 Hereditary 1,25-dihydroxyvitamin D resistant rickets: molecular basis and implications for the role of 1,25(OH)2D3 in normal physiology. Mol Cell Endocrinol 72:C57–62. Faraco JH, Morrison NA, Baker A, Shine J, Frossard PM 1989 ApaI dimorphism at the human vitamin D receptor gene locus. Nucleic Acids Res 17:2150. Malloy PJ, Pike JW, Feldman D 1999 The vitamin D receptor and the syndrome of hereditary 1,25-dihydroxyvitamin Dresistant rickets. Endocr Rev 20:156–188. Rut AR, Hewison M, Kristjansson K, Luisi B, Hughes MR, O’Riordan JLH 1994 Two mutations causing vitamin D resistant rickets: modelling on the basis of steroid hormone receptor DNA-binding domain crystal structures. Clin Endocrinol (Oxf) 41:581–590. Li YC, Pirro AE, Amling M, Delling G, Baron R, Bronson R, Demay MB 1997 Targeted ablation of the vitamin D receptor: an animal model of vitamin D-dependent rickets type II with alopecia. Proc Natl Acad Sci USA 94: 9831–9835. Yoshizawa T, Handa Y, Uematsu Y, Takeda S, Sekine K, Yoshihara Y, Kawakami T, Arioka K, Sato H, Uchiyama Y, Masushige S, Fukamizu A, Matsumoto T, Kato S 1997 Mice lacking the vitamin D receptor exhibit impaired bone formation, uterine hypoplasia and growth retardation after weaning. Nat Genet 16:391–396. Van Cromphaut SJ, Dewerchin M, Hoenderop JG, Stockmans I, Van Herck E, Kato S, Bindels RJ, Collen D, Carmeliet P, Bouillon R, Carmeliet G 2001 Duodenal calcium absorption in vitamin D receptor-knockout mice: functional and molecular aspects. Proc Natl Acad Sci USA 98:13324–13329. Li YC, Amling M, Pirro AE, Priemel M, Meuse J, Baron R, Delling G, Demay MB 1998 Normalization of mineral ion homeostasis by dietary means prevents hyperparathyroidism, rickets, and osteomalacia, but not alopecia in vitamin D receptor-ablated mice. Endocrinology 139:4391–4396. Amling M, Priemel M, Holzmann T, Chapin K, Rueger JM, Baron R, Demay MB 1999 Rescue of the skeletal phenotype of vitamin D receptor-ablated mice in the setting of normal mineral ion homeostasis: formal histomorphometric and biomechanical analyses. Endocrinology 140:4982–4987. Sakai Y, Demay MB 2000 Evaluation of keratinocyte proliferation and differentiation in vitamin D receptor knockout mice. Endocrinology 141:2043–2049. Chen CH, Sakai Y, Demay MB 2001 Targeting expression of the human vitamin D receptor to the keratinocytes of vitamin D receptor null mice prevents alopecia. Endocrinology 142: 5386–5389. Kinuta K, Tanaka H, Moriwake T, Aya K, Kato S, Seino Y 2000 Vitamin D is an important factor in estrogen biosynthesis of both female and male gonads. Endocrinology 141:1317–1324.
253 84. Zinser G, Packman K, Welsh J 2002 Vitamin D3 receptor ablation alters mammary gland morphogenesis. Development 129:3067–3076. 85. Zeitz U, Weber K, Soegiarto DW, Wolf E, Balling R, Erben RG 2003 Impaired insulin secretory capacity in mice lacking a functional vitamin D receptor. FASEB J 17:509–511. 86. O’Kelly J, Hisatake J, Hisatake Y, Bishop J, Norman A, Koeffler HP 2002 Normal myelopoiesis but abnormal T lymphocyte responses in vitamin D receptor knockout mice. J Clin Invest 109:1091–1099. 87. Zinser GM, Sundberg JP, Welsh J 2002 Vitamin D3 receptor ablation sensitizes skin to chemically induced tumorigenesis. Carcinogenesis 23:2103–2109. 88. Liu M, Lee MH, Cohen M, Bommakanti M, Freedman LP 1996 Transcriptional activation of the Cdk inhibitor p21 by vitamin D3 leads to the induced differentiation of the myelomonocytic cell line U937. Genes Dev 10:142–153. 89. Yang ES, Burnstein KL 2003 Vitamin D inhibits G1 to S progression in LNCaP prostate cancer cells through p27Kip1 stabilization and Cdk2 mislocalization to the cytoplasm. J Biol Chem 278:46862–46868. 90. Saunders DE, Christensen C, Wappler NL, Schultz JF, Lawrence WD, Malviya VK, Malone JM, Deppe G 1993 Inhibition of c-myc in breast and ovarian carcinoma cells by 1,25-dihydroxyvitamin D3, retinoic acid and dexamethasone. Anticancer Drugs 4:201–208. 91. Candeliere GA, Jurutka PW, Haussler MR, St-Arnaud R 1996 A composite element binding the vitamin D receptor, retinoid X receptor α, and a member of the CTF/NF-1 family of transcription factors mediates the vitamin D responsiveness of the c-fos promoter. Mol Cell Biol 16:584–592. 92. Elstner E, Linker-Israeli M, Umiel T, Le J, Grillier I, Said J, Shintaku IP, Krajewski S, Reed JC, Binderup L, Koeffler HP 1996 Combination of a potent 20-epi-vitamin D3 analogue (KH 1060) with 9-cis-retinoic acid irreversibly inhibits clonal growth, decreases bcl-2 expression, and induces apoptosis in HL-60 leukemic cells. Cancer Res 56:3570–3576. 93. Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, Baldwin J, Devon K, Dewar K, Doyle M, FitzHugh W, Funke R, Gage D, Harris K, Heaford A, Howland J, Kann L, Lehoczky J, LeVine R, McEwan P, McKernan K, Meldrim J, Mesirov JP, Miranda C, Morris W, Naylor J, Raymond C, Rosetti M, Santos R, Sheridan A, Sougnez C, Stange-Thomann N, Stojanovic N, Subramanian A, Wyman D, Rogers J, Sulston J, Ainscough R, Beck S, Bentley D, Burton J, Clee C, Carter N, Coulson A, Deadman R, Deloukas P, Dunham A, Dunham I, Durbin R, French L, Grafham D, Gregory S, Hubbard T, Humphray S, Hunt A, Jones M, Lloyd C, McMurray A, Matthews L, Mercer S, Milne S, Mullikin JC, Mungall A, Plumb R, Ross M, Shownkeen R, Sims S, Waterston RH, Wilson RK, Hillier LW, McPherson JD, Marra MA, Mardis ER, Fulton LA, Chinwalla AT, Pepin KH, Gish WR, Chissoe SL, Wendl MC, Delehaunty KD, Miner TL, Delehaunty A, Kramer JB, Cook LL, Fulton RS, Johnson DL, Minx PJ, Clifton SW, Hawkins T, Branscomb E, Predki P, Richardson P, Wenning S, Slezak T, Doggett N, Cheng JF, Olsen A, Lucas S, Elkin C, Uberbacher E, Frazier M, Gibbs RA, Muzny DM, Scherer SE, Bouck JB, Sodergren EJ, Worley KC, Rives CM, Gorrell JH, Metzker ML, Naylor SL, Kucherlapati RS, Nelson DL, Weinstock GM, Sakaki Y, Fujiyama A, Hattori M, Yada T, Toyoda A, Itoh T, Kawagoe C, Watanabe H, Totoki Y, Taylor T, Weissenbach J, Heilig R, Saurin W, Artiguenave F, Brottier P, Bruls T, Pelletier E, Robert C, Wincker P, Smith DR, Doucette-Stamm L, Rubenfield M, Weinstock K, Lee HM, Dubois J, Rosenthal A,
254
94. 95. 96. 97.
98.
99. 100. 101.
102.
103.
104.
G. KERR WHITFIELD
Platzer M, Nyakatura G, Taudien S, Rump A, Yang H, Yu J, Wang J, Huang G, Gu J, Hood L, Rowen L, Madan A, Qin S, Davis RW, Federspiel NA, Abola AP, Proctor MJ, Myers RM, Schmutz J, Dickson M, Grimwood J, Cox DR, Olson MV, Kaul R, Shimizu N, Kawasaki K, Minoshima S, Evans GA, Athanasiou M, Schultz R, Roe BA, Chen F, Pan H, Ramser J, Lehrach H, Reinhardt R, McCombie WR, de la Bastide M, Dedhia N, Blocker H, Hornischer K, Nordsiek G, Agarwala R, Aravind L, Bailey JA, Bateman A, Batzoglou S, Birney E, Bork P, Brown DG, Burge CB, Cerutti L, Chen HC, Church D, Clamp M, Copley RR, Doerks T, Eddy SR, Eichler EE, Furey TS, Galagan J, Gilbert JG, Harmon C, Hayashizaki Y, Haussler D, Hermjakob H, Hokamp K, Jang W, Johnson LS, Jones TA, Kasif S, Kaspryzk A, Kennedy S, Kent WJ, Kitts P, Koonin EV, Korf I, Kulp D, Lancet D, Lowe TM, McLysaght A, Mikkelsen T, Moran JV, Mulder N, Pollara VJ, Ponting CP, Schuler G, Schultz J, Slater G, Smit AF, Stupka E, Szustakowski J, Thierry-Mieg D, Thierry-Mieg J, Wagner L, Wallis J, Wheeler R, Williams A, Wolf YI, Wolfe KH, Yang SP, Yeh RF, Collins F, Guyer MS, Peterson J, Felsenfeld A, Wetterstrand KA, Patrinos A, Morgan MJ, Szustakowki J, de Jong P, Catanese JJ, Osoegawa K, Shizuya H, Choi S, Chen YJ 2001 Initial sequencing and analysis of the human genome. Nature 409:860–921. Robinson-Rechavi M, Carpentier AS, Duffraisse M, Laudet V 2001 How many nuclear hormone receptors are there in the human genome? Trends Genet 17:554–556. Laudet V, Auwerx J, Gustafsson J-A, Wahli W 1999 A unified nomenclature system for the nuclear receptor superfamily. Cell 97:161–163. Escriva H, Delaunay F, Laudet V 2000 Ligand binding and nuclear receptor evolution. Bioessays 22:717–727. Moore LB, Maglich JM, McKee DD, Wisely B, Willson TM, Kliewer SA, Lambert MH, Moore JT 2002 Pregnane X receptor (PXR), constitutive androstane receptor (CAR), and benzoate X receptor (BXR) define three pharmacologically distinct classes of nuclear receptors. Mol Endocrinol 16:977–986. Makishima M, Lu TT, Xie W, Whitfield GK, Domoto H, Evans RM, Haussler MR, Mangelsdorf DJ 2002 Vitamin D receptor as an intestinal bile acid sensor. Science 296: 1313–1316. Nagengast FM, Grubben MJ, van Munster IP 1995 Role of bile acids in colorectal carcinogenesis. Eur J Cancer 31A:1067–1070. Hamada K, Umemoto A, Kajikawa A, Seraj MJ, Monden Y 1994 In vitro formation of DNA adducts with bile acids. Carcinogenesis 15:1911–1915. Staudinger JL, Goodwin B, Jones SA, Hawkins-Brown D, MacKenzie KI, LaTour A, Liu Y, Klaassen CD, Brown KK, Reinhard J, Willson TM, Koller BH, Kliewer SA 2001 The nuclear receptor PXR is a lithocholic acid sensor that protects against liver toxicity. Proc Natl Acad Sci USA 98:3369–3374. Xie W, Radominska-Pandya A, Shi Y, Simon CM, Nelson MC, Ong ES, Waxman DJ, Evans RM 2001 An essential role for nuclear receptors SXR/PXR in detoxification of cholestatic bile acids. Proc Natl Acad Sci USA 98:3375–3380. Parks DJ, Blanchard SG, Bledsoe RK, Chandra G, Consler TG, Kliewer SA, Stimmel JB, Willson TM, Zavacki AM, Moore DD, Lehmann JM 1999 Bile acids: natural ligands for an orphan nuclear receptor. Science 284:1365–1368. Whitfield GK, Dang HTL, Schluter SF, Bernstein RM, Bunag T, Manzon LA, Hsieh G, Dominguez CE, Youson JH, Haussler MR, Marchalonis JJ 2003 Cloning of a functional
105. 106.
107.
108. 109.
110. 111.
112.
113.
114.
115.
116. 117.
118. 119. 120.
ET AL .
vitamin D receptor from the lamprey (Petromyzon marinus), an ancient vertebrate lacking a calcified skeleton and teeth. Endocrinology 144:2704–2716. Araya Z, Wikvall K 1999 6α-Hydroxylation of taurochenodeoxycholic acid and lithocholic acid by CYP3A4 in human liver microsomes. Biochim Biophys Acta 1438:47–54. Drocourt L, Ourlin JC, Pascussi JM, Maurel P, Vilarem MJ 2002 Expression of CYP3A4, CYP2B6, and CYP2C9 is regulated by the vitamin D receptor pathway in primary human hepatocytes. J Biol Chem 277:25125–25132. Nelson DR, Koymans L, Kamataki T, Stegeman JJ, Feyereisen R, Waxman DJ, Waterman MR, Gotoh O, Coon MJ, Estabrook RW, Gunsalus IC, Nebert DW 1996 P450 superfamily: update on new sequences, gene mapping, accession numbers and nomenclature. Pharmacogenetics 6:1–42. Goodwin B, Redinbo MR, Kliewer SA 2002 Regulation of CYP3A gene transcription by the pregnane X receptor. Annu Rev Pharmacol Toxicol 42:1–23. Kliewer SA, Moore JT, Wade L, Staudinger JL, Watson MA, Jones SA, McKee DD, Oliver BB, Willson TM, Zetterstrom RH, Perlmann T, Lehmann JM 1998 An orphan nuclear receptor activated by pregnanes defines a novel steroid signaling pathway. Cell 92:73–82. Honkakoski P, Sueyoshi T, Negishi M 2003 Drug-activated nuclear receptors CAR and PXR. Ann Med 35:172–182. Guo GL, Lambert G, Negishi M, Ward JM, Brewer HB, Jr., Kliewer SA, Gonzalez FJ, Sinal CJ 2003 Complementary roles of farnesoid X receptor, pregnane X receptor, and constitutive androstane receptor in protection against bile acid toxicity. J Biol Chem 278:45062–45071. Maglich JM, Stoltz CM, Goodwin B, Hawkins-Brown D, Moore JT, Kliewer SA 2002 Nuclear pregnane X receptor and constitutive androstane receptor regulate overlapping but distinct sets of genes involved in xenobiotic detoxification. Mol Pharmacol 62:638–646. Gill RK, Christakos S 1993 Identification of sequence elements in mouse calbindin-D28K gene that confer 1,25dihydroxyvitamin D3- and butyrate-inducible responses. Proc Natl Acad Sci USA 90:2984–2988. Clemens TL, Garrett KP, Zhou X-Y, Pike JW, Haussler MR, Dempster DW 1988 Immunocytochemical localization of the 1,25-dihydroxyvitamin D3 receptor in target cells. Endocrinology 122:1224–1230. Baes M, Gulick T, Choi HS, Martinoli MG, Simha D, Moore DD 1994 A new orphan member of the nuclear hormone receptor superfamily that interacts with a subset of retinoic acid response elements. Mol Cell Biol 14:1544–1551. Repa JJ, Mangelsdorf DJ 1999 Nuclear receptor regulation of cholesterol and bile acid metabolism. Curr Opin Biotechnol 10:557–563. Makishima M, Okamoto AY, Repa JJ, Tu H, Learned RM, Luk A, Hull MV, Lustig KD, Mangelsdorf DJ, Shan B 1999 Identification of a nuclear receptor for bile acids. Science 284:1362–1365. Edwards PA, Kast HR, Anisfeld AM 2002 BAREing it all: the adoption of LXR and FXR and their roles in lipid homeostasis. J Lipid Res 43:2–12. Lu TT, Repa JJ, Mangelsdorf DJ 2001 Orphan nuclear receptors as eLiXiRs and FiXeRs of sterol metabolism. J Biol Chem 276:37735–37738. Whitfield GK, Jurutka PW, Haussler CA, Haussler MR 1999 Steroid hormone receptors: evolution, ligands and molecular basis of biologic function. J Cell Biochem Suppls 32/33: 110–122.
CHAPTER 13 Nuclear Vitamin D Receptor
121. Rochel N, Wurtz JM, Mitschler A, Klaholz B, Moras D 2000 The crystal structure of the nuclear receptor for vitamin D bound to its natural ligand. Mol Cell 5:173–179. 122. Watkins RE, Wisely GB, Moore LB, Collins JL, Lambert MH, Williams SP, Willson TM, Kliewer SA, Redinbo MR 2001 The human nuclear xenobiotic receptor PXR: structural determinants of directed promiscuity. Science 292:2329–2333. 123. Shaffer PL, Gewirth DT 2002 Structural basis of VDR-DNA interactions on direct repeat response elements. EMBO J 21:2242–2252. 124. Martz E 2002 Protein explorer: easy yet powerful macromolecular visualization. Trends Biochem Sci 27:107–109. 125. Hsieh J-C, Whitfield GK, Jurutka PW, Haussler CA, Thatcher ML, Thompson PD, Dang HTL, Galligan MA, Oza AK, Haussler MR 2003 Two basic amino acids C-terminal of the P-box specify functional binding of the vitamin D receptor to its rat osteocalcin DNA responsive element. Endocrinology 144:5065–5080. 126. Mader S, Kumar V, de Verneuil H, Chambon P 1989 Three amino acids of the oestrogen receptor are essential to its ability to distinguish an oestrogen from a glucocorticoidresponsive element. Nature 338:271–274. 127. Rastinejad F, Perlmann T, Evans RM, Sigler PB 1995 Structural determinants of nuclear receptor assembly on DNA direct repeats. Nature 375:203–211. 128. Li M, Indra AK, Warot X, Brocard J, Messaddeq N, Kato S, Metzger D, Chambon P 2000 Skin abnormalities generated by temporally controlled RXRalpha mutations in mouse epidermis. Nature 407:633–636. 129. Lemon BD, Fondell JD, Freedman LP 1997 Retinoid X receptor:vitamin D3 receptor heterodimers promote stable preinitiation complex formation and direct 1,25-dihydroxyvitamin D3-dependent cell-free transcription. Mol Cell Biol 17:1923–1937. 130. Thompson PD, Remus LS, Hsieh J-C, Jurutka PW, Whitfield GK, Galligan MA, Encinas Dominguez C, Haussler CA, Haussler MR 2001 Distinct retinoid X receptor activation function-2 residues mediate transactivation in homodimeric and vitamin D receptor heterodimeric contexts. J Mol Endocrinol 27:211–227. 131. Pathrose P, Barmina O, Chang CY, McDonnell DP, Shevde NK, Pike JW 2002 Inhibition of 1,25-dihydroxyvitamin D3-dependent transcription by synthetic LXXLL peptide antagonists that target the activation domains of the vitamin D and retinoid X receptors. J Bone Miner Res 17:2196–2205. 132. Hsieh J-C, Whitfield GK, Oza AK, Dang HTL, Price JN, Galligan MA, Jurutka PW, Thompson PD, Haussler CA, Haussler MR 1999 Characterization of unique DNA binding and transcriptional activation functions in the carboxylterminal extension of the zinc finger region in the human vitamin D receptor. Biochemistry 38:16347–16358. 133. Cyert MS 2001 Regulation of nuclear localization during signaling. J Biol Chem 276:20805–20808. 134. Hsieh J-C, Shimizu Y, Minoshima S, Shimizu N, Haussler CA, Jurutka PW, Haussler MR 1998 Novel nuclear localization signal between the two DNA-binding zinc fingers in the human vitamin D receptor. J Cell Biochem 70:94–109. 135. Luo Z, Rouvinen J, Mäenpää PH 1994 A peptide C-terminal to the second Zn finger of human vitamin D receptor is able to specify nuclear localization. Eur J Biochem 223: 381–387. 136. LaCasse EC, Lefebvre YA 1995 Nuclear localization signals overlap DNA- or RNA-binding domains in nucleic acidbinding proteins. Nucleic Acids Res 23:1647–1656.
255 137. Stewart M, Baker RP, Bayliss R, Clayton L, Grant RP, Littlewood T, Matsuura Y 2001 Molecular mechanism of translocation through nuclear pore complexes during nuclear protein import. FEBS Lett 498:145–149. 138. Prufer K, Racz A, Lin GC, Barsony J 2000 Dimerization with retinoid X receptors promotes nuclear localization and subnuclear targeting of vitamin D receptors. J Biol Chem 275:41114–41123. 139. Hsieh J-C, Jurutka PW, Nakajima S, Galligan MA, Haussler CA, Shimizu Y, Shimizu N, Whitfield GK, Haussler MR 1993 Phosphorylation of the human vitamin D receptor by protein kinase C: Biochemical and functional evaluation of the serine 51 recognition site. J Biol Chem 268:15118–15126. 140. Yang W, Hyllner SJ, Christakos S 2001 Interrelationship between signal transduction pathways and 1,25(OH)2D3 in UMR106 osteoblastic cells. Am J Physiol Endocrinol Metab 281:E162–170. 141. Moras D, Gronemeyer H 1998 The nuclear receptor ligandbinding domain: structure and function. Curr Opin Cell Biol 10:384–391. 142. Weatherman RV, Fletterick RJ, Scanlan TS 1999 Nuclearreceptor ligands and ligand-binding domains. Ann Rev Biochem 68:559–581. 143. Bourguet W, Vivat V, Wurtz JM, Chambon P, Gronemeyer H, Moras D 2000 Crystal structure of a heterodimeric complex of RAR and RXR ligand-binding domains. Mol Cell 5:289–298. 144. Bourguet W, Ruff M, Chambon P, Gronemeyer H, Moras D 1995 Crystal structure of the ligand-binding domain of the human nuclear receptor RXR-α. Nature 375:377–382. 145. Gampe RT, Montana VG, Lambert MH, Miller AB, Bledsoe RK, Milburn MV, Kliewer SA, Willson TM, Xu HE 2000 Asymmetry in the PPARgamma/RXRalpha crystal structure reveals the molecular basis of heterodimerization among nuclear receptors. Mol Cell 5:545–555. 146. Tanenbaum DM, Wang Y, Williams SP, Sigler PB 1998 Crystallographic comparison of the estrogen and progesterone receptor’s ligand binding domains. Proc Natl Acad Sci USA 95:5998–6003. 147. Whitfield GK, Selznick SH, Haussler CA, Hsieh J-C, Galligan MA, Jurutka PW, Thompson PD, Lee SM, Zerwekh JE, Haussler MR 1996 Vitamin D receptors from patients with resistance to 1,25-dihydroxyvitamin D3: point mutations confer reduced transactivation in response to ligand and impaired interaction with the retinoid X receptor heterodimeric partner. Mol Endocrinol 10:1617–1631. 148. Feng W, Ribeiro RC, Wagner RL, Nguyen H, Apriletti JW, Fletterick RJ, Baxter JD, Kushner PJ, West BL 1998 Hormonedependent coactivator binding to a hydrophobic cleft on nuclear receptors. Science 280:1747–1749. 149. Darimont BD, Wagner RL, Apriletti JW, Stallcup MR, Kushner PJ, Baxter JD, Fletterick RJ, Yamamoto KR 1998 Structure and specificity of nuclear receptor-coactivator interactions. Genes Dev 12:3343–3356. 150. Nolte RT, Wisely GB, Westin S, Cobb JE, Lambert MH, Kurokawa R, Rosenfeld MG, Willson TM, Glass CK, Milburn MV 1998 Ligand binding and co-activator assembly of the peroxisome proliferator-activated receptor-γ. Nature 395:137–143. 151. Stehlin C, Wurtz JM, Steinmetz A, Greiner E, Schule R, Moras D, Renaud JP 2001 X-ray structure of the orphan nuclear receptor RORbeta ligand-binding domain in the active conformation. EMBO J 20:5822–5831. 152. Mi LZ, Devarakonda S, Harp JM, Han Q, Pellicciari R, Willson TM, Khorasanizadeh S, Rastinejad F 2003 Structural
256
153.
154.
155.
156.
157.
158.
159.
160.
161.
162.
163.
164. 165.
G. KERR WHITFIELD
basis for bile acid binding and activation of the nuclear receptor FXR. Mol Cell 11:1093–1100. Malloy PJ, Xu R, Peng L, Clark PA, Feldman D 2002 A novel mutation in helix 12 of the vitamin D receptor impairs coactivator interaction and causes hereditary 1,25-dihydroxyvitamin D-resistant rickets without alopecia. Mol Endocrinol 16:2538–2546. Jurutka PW, Hsieh J-C, Remus LS, Whitfield GK, Thompson PD, Haussler CA, Blanco JCG, Ozato K, Haussler MR 1997 Mutations in the 1,25-dihydroxyvitamin D3 receptor identifying C-terminal amino acids required for transcriptional activation that are functionally dissociated from hormone binding, heterodimeric DNA binding and interaction with basal transcription factor IIB, in vitro. J Biol Chem 272: 14592–14599. Panda DK, Miao D, Tremblay ML, Sirois J, Farookhi R, Hendy GN, Goltzman D 2001 Targeted ablation of the 25-hydroxyvitamin D 1alpha-hydroxylase enzyme: evidence for skeletal, reproductive, and immune dysfunction. Proc Natl Acad Sci USA 98:7498–7503. Dardenne O, Prud’homme J, Arabian A, Glorieux FH, St-Arnaud R 2001 Targeted inactivation of the 25-hydroxyvitamin D3-1α-hydroxylase gene (CYP27B1) creates an animal model of pseudovitamin D-deficiency rickets. Endocrinology 142:3135–3141. Hsieh JC, Sisk JM, Jurutka PW, Haussler CA, Slater SA, Haussler MR, Thompson CC 2003 Physical and functional interaction between the vitamin D receptor and Hairless corepressor, two proteins required for hair cycling. J Biol Chem 278:38665–38674. Watkins RE, Maglich JM, Moore LB, Wisely GB, Noble SM, Davis-Searles PR, Lambert MH, Kliewer SA, Redinbo MR 2003 2.1 A crystal structure of human PXR in complex with the St. John’s wort compound hyperforin. Biochemistry 42:1430–1438. Hoerer S, Schmid A, Heckel A, Budzinski RM, Nar H 2003 Crystal structure of the human liver X receptor beta ligandbinding domain in complex with a synthetic agonist. J Mol Biol 334:853–861. Rochel N, Tocchini-Valentini G, Egea PF, Juntunen K, Garnier JM, Vihko P, Moras D 2001 Functional and structural characterization of the insertion region in the ligand binding domain of the vitamin D nuclear receptor. Eur J Biochem 268:971–979. Malloy PJ, Hughes MR, Pike JW, Feldman D 1991 Vitamin D receptor mutants and hereditary 1,25-dihydroxyvitamin D resistant rickets. In: Norman AW, Bouillon R, Thomasset M (eds) Vitamin D: Gene Regulation, Structure-Function Analysis and Clinical Application. Walter de Gruyter, New York, pp. 116–124. Bettoun DJ, Burris TP, Houck KA, Buck DW, 2nd, Stayrook KR, Khalifa B, Lu J, Chin WW, Nagpal S 2003 Retinoid X receptor is a non-silent major contributor to vitamin D receptor-mediated transcriptional activation. Mol Endocrinol 17:2320–2328. Warnmark A, Treuter E, Wright A, Gustafsson JA 2003 Activation functions 1 and 2 of nuclear receptors: molecular strategies for transcriptional activation. Mol Endocrinol 17:1901–1909. Thompson EB, Kumar R 2003 DNA binding of nuclear hormone receptors influences their structure and function. Biochem Biophys Res Commun 306:1–4. McKenna NJ, O’Malley BW 2002 Combinatorial control of gene expression by nuclear receptors and coregulators. Cell 108:465–474.
ET AL .
166. Masuyama H, Brownfield CM, St-Arnaud R, MacDonald PN 1997 Evidence for ligand-dependent intramolecular folding of the AF-2 domain in vitamin D receptor-activated transcription and coactivator interaction. Mol Endocrinol 11: 1507–1517. 167. Rachez C, Gamble M, Chang CP, Atkins GB, Lazar MA, Freedman LP 2000 The DRIP complex and SRC-1/p160 coactivators share similar nuclear receptor binding determinants but constitute functionally distinct complexes. Mol Cell Biol 20:2718–2726. 168. Jurutka PW, Hsieh J-C, MacDonald PN, Terpening CM, Haussler CA, Haussler MR, Whitfield GK 1993 Phosphorylation of serine 208 in the human vitamin D receptor: the predominant amino acid phosphorylated by casein kinase II, in vitro, and identification as a significant phosphorylation site in intact cells. J Biol Chem 268:6791–6799. 169. Hilliard GM, Cook RG, Weigel NL, Pike JW 1994 1,25Dihydroxyvitamin D3 modulates phosphorylation of serine 205 in the human vitamin D receptor: Site-directed mutagenesis of this residue promotes alternative phosphorylation. Biochemistry 33:4300–4311. 170. Jurutka PW, Hsieh J-C, Nakajima S, Haussler CA, Whitfield GK, Haussler MR 1996 Human vitamin D receptor phosphorylation by casein kinase II at ser-208 potentiates transcriptional activation. Proc Natl Acad Sci USA 93: 3519–3524. 171. Barletta F, Freedman LP, Christakos S 2002 Enhancement of VDR-mediated transcription by phosphorylation: correlation with increased interaction between the VDR and DRIP205, a subunit of the VDR-interacting protein coactivator complex. Mol Endocrinol 16:301–314. 172. Williams SP, Sigler PB 1998 Atomic structure of progesterone complexed with its receptor. Nature 393:392–396. 173. Kraichely DM, Collins JJ, 3rd, DeLisle RK, MacDonald PN 1999 The autonomous transactivation domain in helix H3 of the vitamin D receptor is required for transactivation and coactivator interaction. J Biol Chem 274:14352–14358. 174. Heery DM, Kalkhoven E, Hoare S, Parker MG 1997 A signature motif in transcriptional co-activators mediates binding to nuclear receptors. Nature 387:733–736. 175. Ding XF, Anderson CM, Ma H, Hong H, Uht RM, Kushner PJ, Stallcup MR 1998 Nuclear receptor-binding sites of coactivators glucocorticoid receptor interacting protein 1 (GRIP1) and steroid receptor coactivator 1 (SRC-1): multiple motifs with different binding specificities. Mol Endocrinol 12:302–313. 176. Jurutka PW, Remus LS, Whitfield GK, Galligan MA, Haussler CA, Haussler MR 2000 Biochemical evidence for a 170-kilodalton, AF-2-dependent vitamin D receptor/retinoid X receptor coactivator that is highly expressed in osteoblasts. Biochem Biophys Res Commun 267:813–819. 177. Kephart DD, Walfish PG, DeLuca H, Butt TR 1996 Retinoid X receptor isotype identity directs human vitamin D receptor heterodimer transactivation from the 24-hydroxylase vitamin D response elements in yeast. Mol Endocrinol 10:408–419. 178. Li XY, Xiao JH, Feng X, Qin L, Voorhees JJ 1997 Retinoid X receptor-specific ligands synergistically upregulate 1,25dihydroxyvitamin D3-dependent transcription in epidermal keratinocytes in vitro and in vivo. J Invest Dermatol 108: 506–512. 179. Kato S, Sasaki H, Suzawa M, Masushige S, Tora L, Chambon P, Gronemeyer H 1995 Widely spaced, directly repeated PuGGTCA elements act as promiscuous enhancers for different classes of nuclear receptors. Mol Cell Biol 15:5858–5867.
257
CHAPTER 13 Nuclear Vitamin D Receptor
180. Sutton AL, MacDonald PN 2003 Vitamin D: More than a “bone-a-fide” hormone. Mol Endocrinol 17:777–791. 181. Sharma D, Fondell JD 2000 Temporal formation of distinct thyroid hormone receptor coactivator complexes in HeLa cells. Mol Endocrinol 14:2001–2009. 182. Shang Y, Hu X, DiRenzo J, Lazar MA, Brown M 2000 Cofactor dynamics and sufficiency in estrogen receptorregulated transcription. Cell 103:843–852. 183. Masuyama H, MacDonald PN 1998 Proteasome-mediated degradation of the vitamin D receptor (VDR) and a putative role for SUG1 interaction with the AF-2 domain of VDR. J Cell Biochem 71:429–440. 184. Rosenfeld MG, Glass CK 2001 Coregulator codes of transcriptional regulation by nuclear receptors. J Biol Chem 276:36865–36868. 185. Huang ZQ, Li J, Sachs LM, Cole PA, Wong J 2003 A role for cofactor-cofactor and cofactor-histone interactions in targeting p300, SWI/SNF and mediator for transcription. EMBO J 22:2146–2155. 186. Encinas Dominguez C 2002 Insights into the biochemical life cycle of the vitamin D receptor: protein and DNA interactions that transduce the signal for gene expression. Biochemistry and Molecular Biophysics. University of Arizona, Tucson, pp. 175–176 [Ph.D. thesis]. 187. Dwivedi PP, Muscat GE, Bailey PJ, Omdahl JL, May BK 1998 Repression of basal transcription by vitamin D receptor: evidence for interaction of unliganded vitamin D receptor with two receptor interaction domains in RIP13delta1. J Mol Endocrinol 20:327–335. 188. Lusser A, Kadonaga JT 2003 Chromatin remodeling by ATPdependent molecular machines. Bioessays 25:1192–1200. 189. Potter GB, Zarach JM, Sisk JM, Thompson CC 2002 The thyroid hormone-regulated corepressor hairless associates with histone deacetylases in neonatal rat brain. Mol Endocrinol 16:2547–2560. 190. Blanco JCG, Wang I-M, Tsai SY, Tsai M-J, O’Malley BW, Jurutka PW, Haussler MR, Ozato K 1995 Transcription factor TFIIB and the vitamin D receptor cooperatively activate ligand-dependent transcription. Proc Natl Acad Sci USA 92:1535–1539. 191. Masuyama H, Jefcoat SC, Jr., MacDonald PN 1997 The N-terminal domain of transcription factor IIB is required for direct interaction with the vitamin D receptor and participates in vitamin D-mediated transcription. Mol Endocrinol 11:218–228. 192. Jurutka PW, Remus LS, Whitfield GK, Thompson PD, Hsieh J-C, Zitzer H, Tavakkoli P, Galligan MA, Dang HT, Haussler CA, Haussler MR 2000 The polymorphic N terminus in human vitamin D receptor isoforms influences transcriptional activity by modulating interaction with transcription factor IIB. Mol Endocrinol 14:401–420. 193. Barry JB, Leong GM, Church WB, Issa LL, Eisman JA, Gardiner EM 2003 Interactions of SKIP/NCoA-62, TFIIB, and retinoid X receptor with vitamin D receptor helix H10 residues. J Biol Chem 278:8224–8228. 194. Li X, Kimbrel EA, Kenan DJ, McDonnell DP 2002 Direct interactions between corepressors and coactivators permit the integration of nuclear receptor-mediated repression and activation. Mol Endocrinol 16:1482–1491. 195. Baudino TA, Kraichely DM, Jefcoat SC, Jr., Winchester SK, Partridge NC, MacDonald PN 1998 Isolation and characterization of a novel coactivator protein, NCoA-62, involved in vitamin D-mediated transcription. J Biol Chem 273:16434–16441. 196. Zhang C, Baudino TA, Dowd DR, Tokumaru H, Wang W, MacDonald PN 2001 Ternary complexes and cooperative
197.
198. 199. 200.
201.
202.
203. 204. 205. 206. 207.
208.
209. 210.
211.
212.
213.
interplay between NCoA-62/Ski-interacting protein and steroid receptor coactivators in vitamin D receptor-mediated transcription. J Biol Chem 276:40614–40620. Lee YH, Koh SS, Zhang X, Cheng X, Stallcup MR 2002 Synergy among nuclear receptor coactivators: selective requirement for protein methyltransferase and acetyltransferase activities. Mol Cell Biol 22:3621–3632. Lemon B, Inouye C, King DS, Tjian R 2001 Selectivity of chromatin-remodelling cofactors for ligand-activated transcription. Nature 414:924–928. Becker PB, Horz W 2002 ATP-dependent nucleosome remodeling. Ann Rev Biochem 71:247–273. Dilworth FJ, Fromental-Ramain C, Yamamoto K, Chambon P 2000 ATP-driven chromatin remodeling activity and histone acetyltransferases act sequentially during transactivation by RAR/RXR in vitro. Mol Cell 6:1049–1058. Kitagawa H, Fujiki R, Yoshimura K, Mezaki Y, Uematsu Y, Matsui D, Ogawa S, Unno K, Okubo M, Tokita A, Nakagawa T, Ito T, Ishimi Y, Nagasawa H, Matsumoto T, Yanagisawa J, Kato S 2003 The chromatin-remodeling complex WINAC targets a nuclear receptor to promoters and is impaired in Williams syndrome. Cell 113:905–917. Breen EC, van Wijnen AJ, Lian JB, Stein GS, Stein JL 1994 In vivo occupancy of the vitamin D responsive element in the osteocalcin gene supports vitamin D-dependent transcriptional upregulation in intact cells. Proc Natl Acad Sci USA 91:12902–12906. Fry CJ, Peterson CL 2002 Transcription. Unlocking the gates to gene expression. Science 295:1847–1848. Freeman BC, Yamamoto KR 2002 Disassembly of transcriptional regulatory complexes by molecular chaperones. Science 296:2232–2235. Vermeulen W, Houtsmuller AB 2002 The transcription cycle in vivo. A blind watchmaker at work. Mol Cell 10:1264–1266. McNally JG, Muller WG, Walker D, Wolford R, Hager GL 2000 The glucocorticoid receptor: rapid exchange with regulatory sites in living cells. Science 287:1262–1265. Becker M, Baumann C, John S, Walker DA, Vigneron M, McNally JG, Hager GL 2002 Dynamic behavior of transcription factors on a natural promoter in living cells. EMBO Rep 3:1188–1194. Demarest SJ, Martinez-Yamout M, Chung J, Chen H, Xu W, Dyson HJ, Evans RM, Wright PE 2002 Mutual synergistic folding in recruitment of CBP/p300 by p160 nuclear receptor coactivators. Nature 415:549–553. Rachez C, Freedman LP 2001 Mediator complexes and transcription. Curr Opin Cell Biol 13:274–280. Zhang C, Dowd DR, Staal A, Gu C, Lian JB, van Wijnen AJ, Stein GS, MacDonald PN 2003 Nuclear coactivator-62 kDa/Ski-interacting protein is a nuclear matrix-associated coactivator that may couple vitamin D receptor-mediated transcription and RNA splicing. J Biol Chem 278:35325–35336. Kim S, Yamamoto H, Shevde NK, Pike JW 2003 1,25Dihydroxyvitamin D3 stimulates dynamic vitamin D3 receptor/ retinoid receptor DNA-binding and coactivator recruitment in intact osteoblasts. J Bone Miner Res 18(Suppl 2):S30 (Abstract 1110). Li XY, Boudjelal M, Xiao JH, Peng ZH, Asuru A, Kang S, Fisher GJ, Voorhees JJ 1999 1,25-Dihydroxyvitamin D3 increases nuclear vitamin D3 receptors by blocking ubiquitin/ proteasome-mediated degradation in human skin. Mol Endocrinol 13:1686–1694. Conaway RC, Brower CS, Conaway JW 2002 Emerging roles of ubiquitin in transcription regulation. Science 296: 1254–1258.
258 214. Wang X, Pongrac JL, DeFranco DB 2002 Glucocorticoid receptors in hippocampal neurons that do not engage proteasomes escape from hormone-dependent down-regulation but maintain transactivation activity. Mol Endocrinol 16: 1987–1998. 215. Gonzalez F, Delahodde A, Kodadek T, Johnston SA 2002 Recruitment of a 19S proteasome subcomplex to an activated promoter. Science 296:548–550. 216. Ohyama Y, Ozono K, Uchida M, Yoshimura M, Shinki T, Suda T, Yamamoto O 1996 Functional assessment of two vitamin D-responsive elements in the rat 25-hydroxyvitamin D3 24-hydroxylase gene. J Biol Chem 271:30381–30385. 217. Aslam F, McCabe L, Frenkel B, van Wijnen AJ, Stein GS, Lian JB, Stein JL 1999 AP-1 and vitamin D receptor (VDR) signaling pathways converge at the rat osteocalcin VDR element: requirement for the internal activating protein-1 site for vitamin D-mediated trans-activation. Endocrinology 140:63–70. 218. Sierra J, Villagra A, Paredes R, Cruzat F, Gutierrez S, Javed A, Arriagada G, Olate J, Imschenetzky M, Van Wijnen AJ, Lian JB, Stein GS, Stein JL, Montecino M 2003 Regulation of the bone-specific osteocalcin gene by p300 requires Runx2/ Cbfa1 and the vitamin D3 receptor but not p300 intrinsic histone acetyltransferase activity. Mol Cell Biol 23:3339–3351. 219. Dwivedi PP, Hii CS, Ferrante A, Tan J, Der CJ, Omdahl JL, Morris HA, May BK 2002 Role of MAP kinases in the 1,25dihydroxyvitamin D3-induced transactivation of the rat cytochrome P450C24 (CYP24) promoter. Specific functions for ERK1/ERK2 and ERK5. J Biol Chem 277:29643–29653. 220. Hsieh J-C, Dang HTL, Galligan MA, Whitfield GK, Jurutka PW, Thompson PD, Haussler CA, Haussler MR 2001 Phosphorylation of the human vitamin D receptor by protein kinase A downregulates 1,25(OH)2D3-dependent transactivation by reducing retinoid X receptor β heterodimerization. J Bone Miner Res 16(Suppl 1):S231 (Abstract F517). 221. Bastien J, Adam-Stitah S, Plassat JL, Chambon P, RochetteEgly C 2002 The phosphorylation site located in the A region of retinoic X receptor alpha is required for the antiproliferative effect of retinoic acid (RA) and the activation of RA target genes in F9 cells. J Biol Chem 277:28683–28689. 222. Adachi S, Okuno M, Matsushima-Nishiwaki R, Takano Y, Kojima S, Friedman SL, Moriwaki H, Okano Y 2002 Phosphorylation of retinoid X receptor suppresses its ubiquitination in human hepatocellular carcinoma. Hepatology 35:332–340. 223. Lee HY, Suh YA, Robinson MJ, Clifford JL, Hong WK, Woodgett JR, Cobb MH, Mangelsdorf DJ, Kurie JM 2000 Stress pathway activation induces phosphorylation of retinoid X receptor. J Biol Chem 275:32193–32199. 224. Mangelsdorf DJ, Borgmeyer U, Heyman RA, Zhou JY, Ong ES, Oro AE, Kakizuka A, Evans RM 1992 Characterization of three RXR genes that mediate the action of 9-cis retinoic acid. Genes Dev 6:329–344. 225. Ruau D, Duarte J, Ourjdal T, Perriere G, Laudet V, RobinsonRechavi M. 2004 Update of NUREBASE: nuclear hormone receptor functional genomics. Nucleic Acids Res. 32 Database issue:D165–167. 226. Thompson PD, Hsieh J-C, Whitfield GK, Haussler CA, Jurutka PW, Galligan MA, Tillman JB, Spindler SR, Haussler MR 1999 The vitamin D receptor displays DNA binding and transactivation as a heterodimer with the retinoid X receptor, but not with the thyroid hormone receptor. J Cell Biochem 75:462–480. 227. Reinhardt TA, Koszewski NJ, Omdahl J, Horst RL 1999 1,25-Dihydroxyvitamin D3 and 9-cis-retinoids are synergistic
G. KERR WHITFIELD
228. 229. 230.
231. 232. 233.
234. 235. 236.
237. 238.
239. 240.
241. 242.
243.
244.
ET AL .
regulators of 24-hydroxylase activity in the rat and 1, 25dihydroxyvitamin D3 alters retinoic acid metabolism in vivo. Arch Biochem Biophys 368:244–248. Masuda S, Jones G 2003 Vitamin D analogs-drug design based on proteins involved in vitamin D signal transduction. Curr Drug Targets Immune Endocr Metabol Disord 3:43–66. Issa LL, Leong GM, Sutherland RL, Eisman JA 2002 Vitamin D analogue-specific recruitment of vitamin D receptor coactivators. J Bone Miner Res 17:879–890. Liu YY, Nguyen C, Peleg S 2000 Regulation of ligandinduced heterodimerization and coactivator interaction by the activation function-2 domain of the vitamin D receptor. Mol Endocrinol 14:1776–1787. Scott LJ, Dunn CJ, Goa KL 2001 Calcipotriol ointment. A review of its use in the management of psoriasis. Am J Clin Dermatol 2:95–120. Adorini L 2002 Immunomodulatory effects of vitamin D receptor ligands in autoimmune diseases. Int Immunopharmacol 2:1017–1028. Vegesna V, O’Kelly J, Said J, Uskokovic M, Binderup L, Koeffler HP 2003 Ability of potent vitamin D3 analogs to inhibit growth of prostate cancer cells in vivo. Anticancer Res 23:283–289. Riggs BL, Hartmann LC 2003 Selective estrogen-receptor modulators—mechanisms of action and application to clinical practice. N Engl J Med 348:618–629. McDonnell DP, Norris JD 2002 Connections and regulation of the human estrogen receptor. Science 296:1642–1644. Gehin M, Mark M, Dennefeld C, Dierich A, Gronemeyer H, Chambon P 2002 The function of TIF2/GRIP1 in mouse reproduction is distinct from those of SRC-1 and p/CIP. Mol Cell Biol 22:5923–5937. Xu J, Li Q 2003 Review of the in vivo functions of the p160 steroid receptor coactivator family. Mol Endocrinol 17:1681–1692. Whitfield GK, Remus LS, Jurutka PW, Zitzer H, Oza AK, Dang HTL, Haussler CA, Galligan MA, Thatcher ML, Encinas Dominguez C, Haussler MR 2001 Functionally relevant polymorphisms in the human nuclear vitamin D receptor gene. Mol Cell Endocrinol 177:145–159. Freund JN, Domon-Dell C, Kedinger M, Duluc I 1998 The Cdx-1 and Cdx-2 homeobox genes in the intestine. Biochem Cell Biol 76:957–969. Yamamoto H, Miyamoto K, Li B, Taketani Y, Kitano M, Inoue Y, Morita K, Pike JW, Takeda E 1999 The caudal-related homeodomain protein Cdx-2 regulates vitamin D receptor gene expression in the small intestine. J Bone Miner Res 14:240–247. Suh E, Chen L, Taylor J, Traber PG 1994 A homeodomain protein related to caudal regulates intestine-specific gene transcription. Mol Cell Biol 14:7340–7351. van Wering HM, Moyer L, Grand RJ, Krasinski SD 2002 Novel interaction at the Cdx-2 binding sites of the lactasephlorizin hydrolase promoter. Biochem Biophys Res Commun 299:587–593. Arai H, Miyamoto KI, Yoshida M, Yamamoto H, Taketani Y, Morita K, Kubota M, Yoshida S, Ikeda M, Watabe F, Kanemasa Y, Takeda E 2001 The polymorphism in the caudalrelated homeodomain protein Cdx-2 binding element in the human vitamin D receptor gene. J Bone Miner Res 16: 1256–1264. Arai H, Miyamoto K-I, Taketani Y, Yamamoto H, Iemori Y, Morita K, Tonai T, Nishisho T, Mori S, Takeda E 1997 A vitamin D receptor gene polymorphism in the translation initiation codon: effect on protein activity and relation to
CHAPTER 13 Nuclear Vitamin D Receptor
245.
246.
247.
248.
249.
250.
251. 252. 253. 254. 255.
256. 257.
258.
259.
260.
bone mineral density in Japanese women. J Bone Miner Res 12:915–921. Ferrari S, Rizzoli R, Manen D, Slosman D, Bonjour JP 1998 Vitamin D receptor gene start codon polymorphisms (FokI) and bone mineral density: interaction with age, dietary calcium, and 3′-end region polymorphisms. J Bone Miner Res 13:925–930. Ames SK, Ellis KJ, Gunn SK, Copeland KC, Abrams SA 1999 Vitamin D receptor gene Fok1 polymorphism predicts calcium absorption and bone mineral density in children. J Bone Miner Res 14:740–746. Gennari L, Becherini L, Mansani R, Masi L, Falchetti A, Morelli A, Colli E, Gonnelli S, Cepollaro C, Brandi ML 1999 FokI polymorphism at translation initiation site of the vitamin D receptor gene predicts bone mineral density and vertebral fractures in postmenopausal Italian women. J Bone Miner Res 14:1379–1386. Gross C, Eccleshall TR, Malloy PJ, Villa ML, Marcus R, Feldman D 1996 The presence of a polymorphism at the translation initiation site of the vitamin D receptor gene is associated with low bone mineral density in postmenopausal Mexican-American women. J Bone Miner Res 11: 1850–1855. Uitterlinden AG, Fang Y, Bergink AP, van Meurs JB, van Leeuwen HP, Pols HA 2002 The role of vitamin D receptor gene polymorphisms in bone biology. Mol Cell Endocrinol 197:15–21. Ingles SA, Ross RK, Yu MC, Irvine RA, La Pera G, Haile RW, Coetzee GA 1997 Association of prostate cancer risk with genetic polymorphisms in vitamin D receptor and androgen receptor. J Natl Cancer Inst 89:166–170. Gallie DR 1998 A tale of two termini: a functional interaction between the termini of an mRNA is a prerequisite for efficient translation initiation. Gene 216:1–11. Mangus DA, Evans MC, Jacobson A 2003 Poly(A)-binding proteins: multifunctional scaffolds for the post-transcriptional control of gene expression. Genome Biol 4:223. Baer BW, Kornberg RD 1980 Repeating structure of cytoplasmic poly(A)-ribonucleoprotein. Proc Natl Acad Sci USA 77:1890–1892. Deo RC, Bonanno JB, Sonenberg N, Burley SK 1999 Recognition of polyadenylate RNA by the poly(A)-binding protein. Cell 98:835–845. Ingles SA, Haile RW, Henderson BE, Kolonel LN, Nakaichi G, Shi CY, Yu MC, Ross RK, Coetzee GA 1997 Strength of linkage disequilibrium between two vitamin D receptor markers in five ethnic groups: implications for association studies. Cancer Epidemiol Biomarkers Prev 6:93–98. Ralston SH 2002 Genetic control of susceptibility to osteoporosis. J Clin Endocrinol Metab 87:2460–2466. Gennari L, Becherini L, Falchetti A, Masi L, Massart F, Brandi ML 2002 Genetics of osteoporosis: role of steroid hormone receptor gene polymorphisms. J Steroid Biochem Mol Biol 81:1–24. Hustmyer FG, Peacock M, Hui S, Johnston CC, Christian J 1994 Bone mineral density in relation to polymorphism at the vitamin D receptor gene locus. J Clin Invest 94: 2130–2134. Houston LA, Grant SFA, Reid DM, Ralston SH 1996 Vitamin D receptor polymorphism, bone mineral density, and osteoporotic vertebral fracture: Studies in a UK population. Bone 18:249–252. Taylor JA, Hirvonen A, Watson M, Pittman G, Mohler JL, Bell DA 1996 Association of prostate cancer with vitamin D receptor gene polymorphism. Cancer Res 56:4108–4110.
259 261. Feldman D 1997 Androgen and vitamin D receptor gene polymorphisms: the long and short of prostate cancer risk. J Natl Cancer Inst 89:109–111. 262. Xu Y, Shibata A, McNeal JE, Stamey TA, Feldman D, Peehl DM 2003 Vitamin D receptor start codon polymorphism (FokI) and prostate cancer progression. Cancer Epidemiol Biomarkers Prev 12:23–27. 263. Ruggiero M, Pacini S, Aterini S, Fallai C, Ruggiero C, Pacini P 1998 Vitamin D receptor gene polymorphism is associated with metastatic breast cancer. Oncol Res 10:43–46. 264. Peehl DM, Skowronski RJ, Leung GK, Wong ST, Stamey TA, Feldman D 1994 Antiproliferative effects of 1,25-dihydroxyvitamin D3 on primary cultures of human prostatic cells. Cancer Res 54:805–810. 265. Blutt SE, Allegretto EA, Pike JW, Weigel NL 1997 1,25dihydroxyvitamin D3 and 9-cis-retinoic acid act synergistically to inhibit the growth of LNCaP prostate cells and cause accumulation of cells in G1. Endocrinology 138: 1491–1497. 266. Mangelsdorf DJ, Koeffler HP, Donaldson CA, Pike JW, Haussler MR 1984 1,25-dihydroxyvitamin D3-induced differentiation in a human promyelocytic leukemia cell line (HL-60): Receptor-mediated maturation to macrophage-like cells. J Cell Biol 98:391–398. 267. Freedman LP 1999 Transcriptional targets of the vitamin D3 receptor-mediating cell cycle arrest and differentiation. J Nutr 129(2S Suppl):581S-586S. 268. Yabushita H, Hirata M, Noguchi M, Nakanishi M 1996 Vitamin D receptor in endometrial carcinoma and the differentiation-inducing effect of 1,25-dihydroxyvitamin D3 on endometrial carcinoma cell lines. J Obstet Gynaecol Res 22:529–539. 269. Giuliano AR, Franceschi RT, Wood RJ 1991 Characterization of the vitamin D receptor from the Caco-2 human colon carcinoma cell line: Effect of cellular differentiation. Arch Biochem Biophys 285:261–269. 270. Palmer HG, Gonzalez-Sancho JM, Espada J, Berciano MT, Puig I, Baulida J, Quintanilla M, Cano A, de Herreros AG, Lafarga M, Munoz A 2001 Vitamin D3 promotes the differentiation of colon carcinoma cells by the induction of E-cadherin and the inhibition of beta-catenin signaling. J Cell Biol 154:369–387. 271. Hutchinson PE, Osborne JE, Lear JT, Smith AG, Bowers PW, Morris PN, Jones PW, York C, Strange RC, Fryer AA 2000 Vitamin D receptor polymorphisms are associated with altered prognosis in patients with malignant melanoma. Clin Cancer Res 6:498–504. 272. Ingles SA, Garcia DG, Wang W, Nieters A, Henderson BE, Kolonel LN, Haile RW, Coetzee GA 2000 Vitamin D receptor genotype and breast cancer in Latinas (United States). Cancer Causes Control 11:25–30. 273. Park BS, Park JS, Lee DY, Youn JI, Kim IG 1999 Vitamin D receptor polymorphism is associated with psoriasis. J Invest Dermatol 112:113–116. 274. Lemire JM, Archer DC, Beck L, Spiegelberg HL 1995 Immunosuppressive actions of 1,25-dihydroxyvitamin D3: preferential inhibition of Th1 functions. J Nutr 125(6 Suppl): 1704S-1708S. 275. Vogel G 2002 Asthma research. Missing gene takes mice’s breath away. Science 295:253. 276. Pages F, Berger A, Henglein B, Piqueras B, Danel C, Zinzindohoue F, Thiounn N, Cugnenc PH, Fridman WH 1999 Modulation of interleukin-18 expression in human colon carcinoma: consequences for tumor immune surveillance. Int J Cancer 84:326–330.
260 277. Miyamoto K-i, Kesterson RA, Yamamoto H, Taketani Y, Nishiwaki E, Tatsumi S, Inoue Y, Morita K, Takeda E, Pike JW 1997 Structural organization of the human vitamin D receptor chromosomal gene and its promoter. Mol Endocrinol 11:1165–1179. 278. Yamada Y, Ando F, Niino N, Shimokata H 2003 Association of polymorphisms of interleukin-6, osteocalcin, and vitamin D receptor genes, alone or in combination, with bone mineral density in community-dwelling Japanese women and men. J Clin Endocrinol Metab 88:3372–3378. 279. Zmuda JM, Cauley JA, Danielson ME, Ferrell RE 2003 Vitamin D receptor and aromatase gene interaction and bone mass in older African-American women. Metabolism 52:521–523. 280. Kibel AS, Suarez BK, Belani J, Oh J, Webster R, BrophyEbbers M, Guo C, Catalona WJ, Picus J, Goodfellow PJ 2003 CDKN1A and CDKN1B polymorphisms and risk of advanced prostate carcinoma. Cancer Res 63:2033–2036. 281. Tayeb MT, Clark C, Haites NE, Sharp L, Murray GI, McLeod HL 2003 CYP3A4 and VDR gene polymorphisms and the risk of prostate cancer in men with benign prostate hyperplasia. Br J Cancer 88:928–932. 282. Pani MA, Knapp M, Donner H, Braun J, Baur MP, Usadel KH, Badenhoop K 2000 Vitamin D receptor allele combinations influence genetic susceptibility to type 1 diabetes in Germans. Diabetes 49:504–507. 283. Loke H, Bethell D, Phuong CX, Day N, White N, Farrar J, Hill A 2002 Susceptibility to dengue hemorrhagic fever in Vietnam: evidence of an association with variation in the vitamin D receptor and Fc gamma receptor IIa genes. Am J Trop Med Hyg 67:102–106. 284. Erturk S, Kutlay S, Karabulut HG, Keven K, Nergizoglu G, Ates K, Bokesoy I, Duman N 2002 The impact of vitamin D receptor genotype on the management of anemia in hemodialysis patients. Am J Kidney Dis 40:816–823. 285. Inagaki K, Krall EA, Fleet JC, Garcia RI 2003 Vitamin D receptor alleles, periodontal disease progression, and tooth loss in the VA dental longitudinal study. J Periodontol 74:161–167. 286. Garland CF, Garland FC, Gorham ED 1999 Calcium and vitamin D. Their potential roles in colon and breast cancer prevention. Ann N Y Acad Sci 889:107–119. 287. Pence BC, Buddingh F 1988 Inhibition of dietary fat-promoted colon carcinogenesis in rats by supplemental calcium or vitamin D3. Carcinogenesis 9:187–190. 288. Kallay E, Pietschmann P, Toyokuni S, Bajna E, Hahn P, Mazzucco K, Bieglmayer C, Kato S, Cross HS 2001 Characterization of a vitamin D receptor knockout mouse as a model of colorectal hyperproliferation and DNA damage. Carcinogenesis 22:1429–1435. 289. Kawa S, Nikaido T, Aoki Y, Zhai Y, Kumagai T, Furihata K, Fujii S, Kiyosawa K 1997 Vitamin D analogues up-regulate p21 and p27 during growth inhibition of pancreatic cancer cell lines. Br J Cancer 76:884–889. 290. Chawla A, Repa JJ, Evans RM, Mangelsdorf DJ 2001 Nuclear receptors and lipid physiology: opening the X-files. Science 294:1866–1870. 291. Lu TT, Makishima M, Repa JJ, Schoonjans K, Kerr TA, Auwerx J, Mangelsdorf DJ 2000 Molecular basis for feedback regulation of bile acid synthesis by nuclear receptors. Mol Cell 6:507–515. 292. Hirano S, Nakama R, Tamaki M, Masuda N, Oda H 1981 Isolation and characterization of thirteen intestinal microorganisms capable of 7 alpha-dehydroxylating bile acids. Appl Environ Microbiol 41:737–745.
G. KERR WHITFIELD
ET AL .
293. Takamine F, Imamura T 1995 Isolation and characterization of bile acid 7-dehydroxylating bacteria from human feces. Microbiol Immunol 39:11–18. 294. Sakai Y, Kishimoto J, Demay MB 2001 Metabolic and cellular analysis of alopecia in vitamin D receptor knockout mice. J Clin Invest 107:961–966. 295. Fuchs E, Merrill BJ, Jamora C, DasGupta R 2001 At the roots of a never-ending cycle. Dev Cell 1:13–25. 296. Barrandon Y 2003 Developmental biology: A hairy situation. Nature 422:272–273. 297. Jamora C, DasGupta R, Kocieniewski P, Fuchs E 2003 Links between signal transduction, transcription and adhesion in epithelial bud development. Nature 422:317–322. 298. Kobielak K, Pasolli HA, Alonso L, Polak L, Fuchs E 2003 Defining BMP functions in the hair follicle by conditional ablation of BMP receptor IA. J Cell Biol 163:609–623. 299. Peters EM, Foitzik K, Paus R, Ray S, Holick MF 2001 A new strategy for modulating chemotherapy-induced alopecia, using PTH/PTHrP receptor agonist and antagonist. J Invest Dermatol 117:173–178. 300. Cho YM, Woodard GL, Dunbar M, Gocken T, Jimenez JA, Foley J 2003 Hair-cycle-dependent expression of parathyroid hormone-related protein and its type I receptor: evidence for regulation at the anagen to catagen transition. J Invest Dermatol 120:715–727. 301. Falzon M 1996 DNA sequences in the rat parathyroid hormone-related peptide gene responsible for 1,25-dihydroxyvitamin D3-mediated transcriptional repression. Mol Endocrinol 10:672–681. 302. Kremer R, Sebag M, Champigny C, Meerovitch K, Hendy GN, White J, Goltzman D 1996 Identification and characterization of 1,25-dihydroxyvitamin D3-responsive repressor sequences in the rat parathyroid hormone-related peptide gene. J Biol Chem 271:16310–16316. 303. Hirst MA, Hochman HI, Feldman D 1985 Vitamin D resistance and alopecia: a kindred with normal 1,25-dihydroxyvitamin D binding, but decreased receptor affinity for deoxyribonucleic acid. J Clin Endocrinol Metab 60:490–495. 304. Hughes MR, Malloy PJ, Kieback DG, Kesterson RA, Pike JW, Feldman D, O’Malley BW 1988 Point mutations in the human vitamin D receptor gene associated with hypocalcemic rickets. Science 242:1702–1705. 305. Kristjansson K, Rut AR, Hewison M, O’Riordan JLH, Hughes MR 1993 Two mutations in the hormone binding domain of the vitamin D receptor cause tissue resistance to 1,25-dihydroxyvitamin D3. J Clin Invest 92:12–16. 306. Botchkarev VA, Botchkareva NV, Nakamura M, Huber O, Funa K, Lauster R, Paus R, Gilchrest BA 2001 Noggin is required for induction of the hair follicle growth phase in postnatal skin. FASEB J 15:2205–2214. 307. Hörlein AJ, Näär AM, Heinzel T, Torchia J, Gloss B, Kurokawa R, Ryan A, Kamei Y, Soderstrom M, Glass CK, Rosenfeld MG 1995 Ligand-independent repression by the thyroid hormone receptor mediated by a nuclear receptor co-repressor. Nature 377:397–404. 308. Tagami T, Lutz WH, Kumar R, Jameson JL 1998 The interaction of the vitamin D receptor with nuclear receptor corepressors and coactivators. Biochem Biophys Res Commun 253:358–363. 309. Wong CW, Privalsky ML 1998 Transcriptional silencing is defined by isoform- and heterodimer-specific interactions between nuclear hormone receptors and corepressors. Mol Cell Biol 18:5724–5733. 310. Potter GB, Beaudoin GM, 3rd, DeRenzo CL, Zarach JM, Chen SH, Thompson CC 2001 The hairless gene mutated in
261
CHAPTER 13 Nuclear Vitamin D Receptor
311.
312. 313.
314.
315.
316.
317.
318.
319.
congenital hair loss disorders encodes a novel nuclear receptor corepressor. Genes Dev 15:2687–2701. Moraitis AN, Giguere V, Thompson CC 2002 Novel mechanism of nuclear receptor corepressor interaction dictated by activation function 2 helix determinants. Mol Cell Biol 22:6831–6841. Stoye JP, Fenner S, Greenoak GE, Moran C, Coffin JM 1988 Role of endogenous retroviruses as mutagens: the hairless mutation of mice. Cell 54:383–391. Ahmad W, Faiyaz ul Haque M, Brancolini V, Tsou HC, ul Haque S, Lam H, Aita VM, Owen J, deBlaquiere M, Frank J, Cserhalmi-Friedman PB, Leask A, McGrath JA, Peacocke M, Ahmad M, Ott J, Christiano AM 1998 Alopecia universalis associated with a mutation in the human hairless gene. Science 279:720–724. Cichon S, Anker M, Vogt IR, Rohleder H, Putzstuck M, Hillmer A, Farooq SA, Al-Dhafri KS, Ahmad M, Haque S, Rietschel M, Propping P, Kruse R, Nothen MM 1998 Cloning, genomic organization, alternative transcripts and mutational analysis of the gene responsible for autosomal recessive universal congenital alopecia. Hum Mol Genet 7:1671–1679. Miller J, Djabali K, Chen T, Liu Y, Ioffreda M, Lyle S, Christiano AM, Holick M, Cotsarelis G 2001 Atrichia caused by mutations in the vitamin D receptor gene is a phenocopy of generalized atrichia caused by mutations in the hairless gene. J Invest Dermatol 117:612–617. Barwick JL, Quattrochi LC, Mills AS, Potenza C, Tukey RH, Guzelian PS 1996 Trans-species gene transfer for analysis of glucocorticoid-inducible transcriptional activation of transiently expressed human CYP3A4 and rabbit CYP3A6 in primary cultures of adult rat and rabbit hepatocytes. Mol Pharmacol 50:10–16. Taketani Y, Segawa H, Chikamori M, Morita K, Tanaka K, Kido S, Yamamoto H, Iemori Y, Tatsumi S, Tsugawa N, Okano T, Kobayashi T, Miyamoto K, Takeda E 1998 Regulation of type II renal Na+-dependent inorganic phosphate transporters by 1,25-dihydroxyvitamin D3. Identification of a vitamin D-responsive element in the human NAPi-3 gene. J Biol Chem 273:14575–14581. Kerner SA, Scott RA, Pike JW 1989 Sequence elements in the human osteocalcin gene confer basal activation and inducible response to hormonal vitamin D3. Proc Natl Acad Sci USA 86:4455–4459. DeMay MB, Gerardi JM, DeLuca HF, Kronenberg HM 1990 DNA sequences in the rat osteocalcin gene that bind the 1,25dihydroxyvitamin D3 receptor and confer responsiveness to
320.
321.
322.
323.
324.
325.
326.
327.
328.
1,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 87: 369–373. Cao X, Ross FP, Zhang L, MacDonald PN, Chappel J, Teitelbaum SL 1993 Cloning of the promoter for the avian integrin β3 subunit gene and its regulation by 1,25-dihydroxyvitamin D3. J Biol Chem 268:27371–27380. Liu SM, Koszewski N, Lupez M, Malluche HH, Olivera A, Russell J 1996 Characterization of a response element in the 5′-flanking region of the avian (chicken) PTH gene that mediates negative regulation of gene transcription by 1,25dihydroxyvitamin D3 and binds the vitamin D3 receptor. Mol Endocrinol 10:206–215. Lian JB, Shalhoub V, Aslam F, Frenkel B, Green J, Hamrah M, Stein GS, Stein JL 1997 Species-specific glucocorticoid and 1,25-dihydroxyvitamin D responsiveness in mouse MC3T3-E1 osteoblasts: Dexamethasone inhibits osteoblast differentiation and vitamin D down-regulates osteocalcin gene expression. Endocrinology 138:2117–2127. Kim RH, Li JJ, Ogata Y, Yamauchi M, Freedman LP, Sodek J 1996 Identification of a vitamin D3-response element that overlaps a unique inverted TATA box in the rat bone sialoprotein gene. Biochem J 318:219–226. Rowland-Goldsmith MA, Henry HL 1997 Structure and regulatory elements of the vitamin D down-regulated gene for cAMP dependent protein kinase inhibitor (PKI). In: Norman AW, Bouillon R, Thomasset M (eds) Vitamin D: Chemistry, Biology and Clinical Applications of the Steroid Hormone. Vol. 1. Riverside: University of California, Printing and Reprographics, Riverside, CA, pp. 300–301. Drissi H, Pouliot A, Koolloos C, Stein JL, Lian JB, Stein GS, van Wijnen AJ 2002 1,25-(OH)2-Vitamin D3 suppresses the bone-related Runx2/Cbfa1 gene promoter. Exp Cell Res 274:323–333. Thompson JD, Higgins DG, Gibson TJ 1994 CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22:4673–4680. Janowski BA, Grogan MJ, Jones SA, Wisely GB, Kliewer SA, Corey EJ, Mangelsdorf DJ 1999 Structural requirements of ligands for the oxysterol liver X receptors LXRalpha and LXRbeta. Proc Natl Acad Sci USA 96:266–271. Sunn KL, Cock TA, Crofts LA, Eisman JA, Gardiner EM 2001 Novel N-terminal variant of human VDR. Mol Endocrinol 15:1599–1609.
CHAPTER 14
Vitamin D Receptor Cofactors: Function, Regulation, and Selectivity LEONARD P. FREEDMAN AND ALFRED A. RESZKA Department of Molecular Endocrinology and Bone Biology, Merck Research Laboratories, West Point, Pennsylvania
I. Introduction II. The Vitamin D Receptor and the Basal Machinery of Transcription III. Coactivators IV. Integration of Signaling Pathways
V. Molecular Basis for Tissue-Selective Vitamin D Receptor Ligands VI. Conclusions References
I. INTRODUCTION
multistep process initiated at the promoter region of expressed genes. It is catalyzed by RNA polymerase II (RNA Pol II) and requires the assembly of general transcription factors (GTFs) including TFIIA, -B, -D, -E, -F, -H, at the promoter (illustrated in Fig. 1; for review see [1–4]). This process is regulated by a combination of transcription factors recruited to a given promoter by direct DNA binding to specific response elements. DNA-bound factors mediate protein–protein interactions with components of the transcription machinery, ultimately targeting the recruitment and/or control of RNA Pol II. Several direct contacts have been identified between VDR or other nuclear receptors and the basal transcription apparatus. TFIIB was reported to interact directly with VDR-LBD. Interestingly, this interaction does not include the transactivation motif (AF-2) of VDR [5,6]. Moreover, the ligand may have a distinct effect on this interaction, depending on the cell type [5,7,8]. Taken together these data may suggest the requirement of additional targets for VDR transcription activity. For example, another basal factor, TFIIA, can also directly bind to VDR. This effect is strongly stimulated by ligand and occurs in the context of VDR bound to a promoter DNA template [9]. VDR also binds several TBP-associated factors TAFs (TATA box-binding protein [TBP] associated factors) that comprise the basal factor TFIID. TAFII135 and TAFII55 bind to VDR, RAR, and TR and enhance their activity [10,11]. TAFII28 has a potentiating or repressive effect on VDR and ER activity, depending on the cell type (COS versus HeLa cells, respectively) [12]. Finally, VDR binds to a number of newly discovered factors that appear to bridge or recruit other activities important for the activation process, as described at length later.
Vitamin D3 receptor (VDR) regulates transcription in direct response to its hormonal ligand, 1α,25(OH)2D3. Ligand binding leads to the recruitment of coactivators, defined as proteins that potentiate the activity of specific transcription factors. Many of these cofactors, as part of large complexes, act to remodel chromatin through ATP-dependent subunits or intrinsic histone-modifying activities. In addition, other ligand-recruited complexes appear to act more directly on the transcriptional apparatus. This suggests that transcriptional regulation by VDR and other nuclear receptors may involve a process of both chromatin alteration and direct recruitment of key initiation components at regulated promoters. This chapter will review the major cofactor complexes found to have significant effects on VDR function. The vast number of these coactivators and coactivator complexes provide potential targets for the relatively specific effects that have been achieved with some clinically important ligand analogs of nuclear receptors (for example, raloxifene for the estrogen receptor). This has spurred a tremendous amount of interest and effort in the quest to generate highly selective nuclear receptor modulators that would confer as drugs desirable effects on target tissues in the absence of potentially deleterious side effects. We will review efforts to achieve this kind of profile with vitamin D analogs, with an emphasis on how such selective ligands might function at the level of cofactor regulation.
II. THE VITAMIN D RECEPTOR AND THE BASAL MACHINERY OF TRANSCRIPTION Transcriptional activation of genes regulated by nuclear receptors and other transcription factors is a VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
Copyright © 2005, Elsevier, Inc. All rights reserved.
264 III. COACTIVATORS A. The SRC/p160 Family of Coactivators Over the past several years, a growing number of proteins have been identified as coactivators for various nuclear receptors (for review see [3,13,14]). Many of these putative coactivators have been cloned by yeast two-hybrid or GST pull-down assays by virtue of their interaction with members of the nuclear receptor family, and their ability to potentiate transcriptional activity. Among the many nuclear receptor coactivators characterized thus far, a homologous family of proteins has emerged. It has been alternatively named SRC, NCoA, or more generically, p160, based on one of its first identified members, the 160-kDa protein steroid receptor coactivator-1 (SRC-1) [15]. The SRC/p160 family comprises three types of factors, on the basis of their homologies, including SRC-1/NCoA-1, GRIP1/ TIF2/NCoA-2, and pCIP/RAC3/ACTR/AIB1/TRAM-1 (for review, see [16,17]). The original enzymatic activity found to be common to these factors is histone acetyltransferase (HAT) activity, which catalyzes the acetylation of lysine residues at the N-terminal tails of histones [18,19]. Acetylation is thought to destabilize the interactions between DNA and the histone cores that form its nucleosomal structure in the nucleus. The SRC/p160 family members would thereby act as coactivators by loosening the repressive effect of chromatin on gene expression. In the original publications describing these proteins, ligand-dependent binding to VDR was only documented with the coactivator ACTR, but subsequent analyses have shown that VDR is also the target of GRIP1/TIF-2 and SRC-1 [20,21] (also our observations). Specific inactivation of SRC-1 by gene targeting reveals that there may be at least partial functional redundancy between the different SRC/p160 family members, since only a partial resistance to hormonal response is observed in SRC-1(−/−) mice. This phenomenon is concomitant with increased mRNA levels of other coactivators such as TIF2, perhaps compensating for the loss of SRC-1 [22]. The coactivator effects of SRC-1 on nuclear receptors have also been demonstrated in vitro in the presence of chromatin assembled templates [23]. Under these conditions, SRC-1 strongly potentiates the ligand-dependent activity of the progesterone receptor (PR-B). Interestingly, this potentiation also occurs to a certain extent on naked DNA, in the absence of chromatin. These results may suggest a dual effect of SRC-1, both on chromatin remodeling and on other activities or interactions yet to be identified. Beyond this homologous family, several proteins that have been found to act as co-regulators of VDR
LEONARD P. FREEDMAN AND ALFRED A. RESZKA
cannot be classified in any of the previous categories. NCoA-62 /SKIP, identified by yeast two-hybrid, coactivates VDR in a ligand-dependent fashion [24]. ChIP analyses reveal that NCoA-62 associates with liganded VDR after the binding of SRC-1 [25]. Interestingly, NCoA-62 was also shown to interact with components of the spliceosome, and a dominant negative NCoA-62 inhibits splicing of 1α,25(OH)2VD3 -induced transcripts, suggesting that it somehow couples VDR-mediated transcription to the process of RNA splicing. TIF1 was found to interact with VDR and other nuclear receptors [26]. TIF1 also interacts with heterochromatin-associated proteins and has a kinase activity that can phosphorylate several general transcription factors (TFIIEa, TAFII28, and TAFII55), suggesting a novel mechanism of transcription regulation for nuclear receptors [27].
B. Large Coactivator Assemblies with Multifunctional HAT Activity The SRC/p160 family of coactivators not only binds to nuclear receptors, such as VDR, but also recruits and forms complexes with CBP/p300 [19]. The cointegrators CBP and p300 also bind to a large panel of transcription factors [28], and although they do bind to nuclear receptors, this interaction is much weaker than is seen with SRC/p160. Indeed, they appear to interact with nuclear receptors cooperatively along with SRC/ p160 and other components such as p/CIP and PCAF, together forming a larger coactivator complex [13,29]. Alternatively, CBP and SRC-1 may be stabilized by a specific RNA coactivator, SRA (steroid receptor RNA activator). Interestingly, SRA was found to be part of a 600–700 kDa ribonucleoprotein structure that includes SRC-1 [30]. CBP/p300 display HAT activity [31,32], which plays an integral role in chromatin remodeling. The fact that different components of the complex possess HAT activity suggests some sort of a cooperative effect or/and an increased array of specificities. Recent reports suggest additional functions for CBP’s HAT activity besides chromatin remodeling. CBP/p300 can acetylate nonhistone proteins, such as the transcription factor p53, thus enhancing its DNA binding activity [33]. Components of the basal machinery (TFIIEa, TFIIF) are also acetylated by p300, PCAF, and TAFII250, but the effects of this modification have not yet been elucidated [34]. Intriguingly, CBP/p300 can regulate the association between ACTR and the estrogen receptor by directly acetylating ACTR at two lysines near to one of its NR boxes, thereby disrupting its association with ER [35]. Currently, our view of how SRC/p160 functions is primarily to act as means of recruiting
265
CHAPTER 14 Vitamin D Receptor Cofactors: Function, Regulation, and Selectivity
CBP/p300 to a nuclear receptor. It is CBP/p300 rather than SRC/p160 that appears to be the primary source of HAT activity.
C. The Human Mediator Complex: DRIP 1. IDENTIFICATION OF THE COMPLEX
Besides the SRC/p160 coactivators, several laboratories, including our own, identified a novel type of nuclear receptor coactivator complex, alternatively called DRIP, TRAP, ARC, NAT, or mammalian Mediator (reviewed
in [36]) (Table I), with each name depending on the purification process and the activators tested as targets. These complexes are required for activation of transcription in vitro, in different transcription assays involving purified components of the transcription machinery. At the time of their respective discoveries, each of these complexes was thought to be specific for a distinct transcription factor. DRIP (vitamin D receptor interacting proteins) [37,38] and ARC (activator recruited cofactor) [39,40] complexes were originally purified from nuclear extracts by in vitro pull-down assays using GST fusions with VDR-LBD or using the activation motifs of several
TABLE I Subunit Composition of General Coactivator Complexes. DRIP
DRIP250 DRIP240 DRIP205 DRIP150 DRIP130 DRIP100 DRIP97 DRIP92 DRIP77 DRIP70–2
ARC CBP/p300 ARC250 ARC240 ARC205 ARC150 ARC130 ARC105/TIG-1 ARC100 ARC92 ARC77 ARC70 ARC42
SMCC/TRAP
TRAP240 TRAP230 TRAP220 TRAP170 TRAP150 TRAP100 TRAP97 TRAP95 TRAP93 TRAP80
CRSP
p230 CRSP200 CRSP150 CRSP130
ARC36 ARC34
DRIP33
ARC33 ARC32 p28
p150 p140/hSur2
p95 p90 CRSP77 CRSP70
Mediator (mouse)
Mediator (Yeast S.c.)
p160a p160b Rgr1/p110
Nut1 Gal11 Rgr1
Ring3/p96a p96b
Sin4 Srb4 Med1
p78 p70 p55
hSrb10 DRIP36 DRIP34
NAT
hMed7 hMed6 hTRF hSrb11 hSoh1
hSrb7 hNut2
CRSP34 CRSP33
p56/Cdk8 p45 p37 p36 p33 p31/Cycl.C p30 p23 p22 p21 p17 p14
p34 Med7/p36 Med6/p32 TRF/p28a p28b
Srb7/p21
Subunits with highlight are equivalent in the different complexes, or homologs between mammalian and yeast complexes. Subunits in the same lanes have similar molecular weights regardless of any homology.
Med2 Pgd1/Hrs1 Srb10 Med4 Med7 Srb5 Med6 Med8 Srb11 Rox3/Ssn7 Srb2 Med9/Cse2 Srb7 Med10/Nut2 Med11 Srb6
266 transcription factors (SREBP-1a, NF-kB, and VP16), respectively. The TRAP complex (TR associated proteins [41,42]) was isolated by coimmunoprecipitation of epitope-tagged TR stably expressed in HeLa cells. TRAP was subsequently found to be identical to SMCC (Srb/Med-containing cofactor complex [43], a complex purified by coimmunoprecipitation with antibodies directed against epitope-tagged Srb10. The cloning of these component subunits by independent groups revealed the near-identity of their sequences, suggesting that these different complexes might actually constitute a single, universal one. Importantly, some subunits do differ from complex to complex. For example, TRAP150 has no homology with any of its candidate counterparts. In addition, DRIP/ARC/CRSP130, also identified as hSur2 [44], has not been identified in the TRAP complex. The various subunit compositions are summarized in Table I. Other complexes exist that have close identities with the DRIP, ARC, and TRAP complexes. CRSP (cofactor required for Sp1 activation) [45], purified by multiple chromatographic steps, appears to be a subset of nine subunits of the DRIP/ARC complex and might represent a stable core of subunits or a conserved subcomplex among various functionally related complexes. The CRSP complex differs, however, by its two unrelated 34- and 70-kDa subunits (Table I). The homology of CRSP70’s N terminus with the elongation factor TFIIS is a unique feature among all the complexes described so far. Despite its apparently limited number of subunits, CRSP potentiates the activity of the transcription factor Sp1 in vitro. The NAT complex (negative regulator of activated transcription) was identified by coimmunoprecipitation out of HeLa nuclear extracts with an antibody against hSrb10/CDK8 [46]. NAT shares many common subunits with DRIP, ARC, and TRAP. However, when the NAT complex was tested in vitro in a purified transcription assay in the presence of RNA Pol II, general transcription factors (TFIIA, to -H), and a cofactor activity PC4 (see later discussion), it exhibited a repressing effect on transcription driven by various activators, without any influence on basal transcription. This unexpected result suggests that these complexes may not only have an activation potential, but also repressive activities on transcription in vitro, as discussed later. The mouse Mediator complex was identified [47] through biochemical purification out of nuclear extracts of murine cells. Its name reflects its homologies with a yeast Mediator counterpart (Table I; see later discussion). 2. FUNCTIONALITY OF THE DRIP COMPLEX
The functional role of DRIP and related complexes in human cells can be postulated on the basis of a series
LEONARD P. FREEDMAN AND ALFRED A. RESZKA
of classic studies in yeast (for review, see [48,49]). Genetic and biochemical analyses in yeast revealed the importance of several types of factors for transcription of target genes by RNA Pol II in response to activators [49]. They include the products of SRB genes, identified in a genetic screen as suppressors of the effect of truncations within the CTD of RNA Pol II [50]. Srb proteins point to the importance of CTD phosphorylation in the transition between transcription initiation at the promoter and elongation of the RNA [51] (transcript, and references therein). Biochemical fractionation resulted in the isolation of Mediator as a multisubunit complex [52] that contained Srb proteins, and another set of components named Med proteins [53,54]. Detailed functional studies in yeast revealed the importance of both Srb and Med components of the Mediator complex in either general transcription (Srb4 [55]), or activation of transcription by specific factors both in vivo and in vitro, such as Gcn4, Gal4, or VP16 (observed for Med2, Med6, Pgd1/Hrs1, and Sin4 [53,54]). This raised the definition of the yeast Mediator as a “global transcription coactivator” [56]. The existence of human homologs of a number of yeast Mediator proteins suggested that a corresponding complex would exist in higher organisms. Mammalian Mediator was first identified in mice by Kornberg’s group [47]. Its homology to the yeast Mediator is also suggested by some of its functional characteristics: the complex can bind the CTD of RNA Pol II, and it stimulates in vitro CTD phosphorylation, catalyzed by TFIIH. Some of its subunits have yeast counterparts (Rgr1, Med6, Med7, Srb7), but several other subunits lack any yeast homologs. Interestingly some, but not all, of the mouse Mediator subunits are also present within DRIP, ARC, TRAP/SMCC, and NAT (see Table I). That DRIP, ARC, and TRAP/SMCC complexes all contain Med components [40,43,57] may suggest that these complexes are likely to be human homologs of the Mammalian Mediator. However, a side-by-side comparison of the respective subunits within human and mouse complexes favors a scenario where “mammalian Mediator” encompasses subcomplexes purified as DRIP, ARC, and TRAP/SMCC (Table I). A larger diversity of complexes in higher eukaryotes relative to yeast was previously suggested for the RNA Pol II holoenzyme on the basis of its different subunit compositions in mammalian preparations (i.e., variations in a subset of components) [48]. The potential differences in composition of the DRIP, ARC, TRAP/SMCC, NAT, and mammalian Mediator complexes may reflect this diversity. Whatever the case, the presence of Srb/Med subunits in DRIP, ARC, and TRAP/SMCC strongly suggests that at least part of how this complex functions is through recruitment of RNA Pol II. Indeed, experiments demonstrate
CHAPTER 14 Vitamin D Receptor Cofactors: Function, Regulation, and Selectivity
that DRIP bound to liganded VDR indeed can recruit RNA Pol II and associated subunits [58]. As mentioned previously, activities of all these complexes were tested in highly purified transcription assays in vitro, in response to specific activators. The DRIP complex strongly potentiated ligand-dependent VDR-RXR transcription on DNA templates assembled into chromatin, but interestingly it had little or no effect on the same transcription in the absence of chromatin (naked templates). None of the previously identified SRC/p160 coactivators have been found to be part of the purified DRIP complex [38], and the absence of any HAT activity in the DRIP complex suggests that it may contain distinct chromatin remodeling activities, or may recruit them. The ARC complex tested on chromatin-assembled templates exhibits cooperativity between different activators, such as Sp1 and SREBP-1a, on the same template promoter, in conditions where ARC has no effect on the same activators tested individually [40]. This is an interesting demonstration of the ability of these complexes to integrate multiple transcription pathways into a synergistic effect on gene activation. TRAP/SMCC also enhances activatordependent transcription in a purified in vitro system with RNA Pol II, GTFs, and PC4, but only in the absence of or with limiting concentrations of TFIIH [43]. In a transcription system including both PC4 and TFIIH, the SMCC complex can repress transcription. This effect was specific for the presence of activators such as Gal4-AH, since no strong effect was observed on basal transcription. By analogy with the effect of Mediator in yeast on phosphorylation of RNA Pol II CTD, the kinase activity of SMCC was tested in vitro. Although the CTD was phosphorylated, the major substrate was the cofactor PC4, a single-stranded DNA binding protein that is required for activated transcription [59,60]. PC4 can interact in vitro with TFIIA and VP16 [60]. More importantly, PC4 binding and transcription activities are lost upon phosphorylation [59,60]. The conservation of PC4 in yeast (as Tsp1) highlights its general requirement among eukaryotic systems [61]. Interestingly, the repression effect by SMCC was also observed when a CTD-less form of RNA Pol II was used. Based on this, the authors suggested that regulation of transcription by SMCC could be mediated in concert with PC4, but independently of modifications of the RNA Pol II CTD [43].
D. ATP-Dependent Remodeling Cofactors In addition to HAT and HDAC-containing activities, complexes found biochemically to contain ATPdependent chromatin remodeling activities, which
267
could be expected to produce more significant chromatin rearrangements, have also been linked to steroid and nuclear receptors. Yeast genetics originally identified a group of gene products, SNF/Swi, which reduced expression of the SUC2 and HO genes involved in sucrose fermentation and mating-type switching, while among suppressors of these genes, SIN or switchindependent mutations, were identified to be histones and other chromatin-associated proteins. SNF2/Swi2 homologs were subsequently found in Drosophila (Brahma) and mammalian cells (hbrm and BRG-1) [62]. Purified multisubunit SNF/Swi complexes were later found to contain DNA-dependent ATPase activity intrinsic to SNF2/Swi2, which simultaneously disrupted nucleosomes and enhanced transcription factor binding [63,64]. Purification of hbrm and BRG-1 revealed they are present in distinct mammalian complexes associated with selective BRG-1 associated factors (BAFs) [65]. ATP-dependent remodeling by these complexes has been shown to be independent of histone tails and is both persistent upon removal of ATP and reversible upon readdition of SNF/Swi [66,67]. An early connection of SNF/Swi to nuclear receptormediated transcription was revealed by the loss of steroid-induced reporter gene activation by GR and ER in Swi-1, -2, and -3-mutant yeast strains [68]. Additionally, when overexpressed in cells, hbrm was found to cooperate with transfected GR in mammalian cell lines devoid of endogenous hbrm [69] and ER was observed to interact with both hbrm and BRG-1 in yeast in twohybrid experiments [70]. Furthermore, conditional BRG-1 knockout F9 cell lines were found to be inviable, while heterozygote lines could not proliferate yet remained sensitive to RA-induced differentiation [71]. One extensively examined model for ligand-dependent perturbations of chromatin structure in preinitiation complex (PIC) formation, as well as antagonism between the progesterone and glucocorticoid receptors (PR and GR), has been the integrated composite MMTV promoter/enhancer (reviewed in [72]). Transiently transfected MMTV DNA is activated by PR ligands; however, random integration of this same DNA into the mammalian genome with concomitant packaging into a phased array of chromatin eliminated PR-dependent gene activation by preclusion of binding site accessibility, but also induced progestin-specific repression of glucocorticoid-mediated transcription from MMTV, suggesting competition for common molecular targets. One report has described a model for this competition mediated through BRG-1 [73]. Interestingly, both progestins and anti-progestins inhibited GR stimulation of integrated MMTV DNA and decreased its association with limiting amounts of BRG-1, as assessed by coimmunoprecipitation. These same compounds were unable
268 to reduce transactivation of transiently transfected promoters, nor were they able to destabilize GR bound to the p160 coactivators NCoA-1/SRC-1 and NCoA2/GRIP-1, as well as with p300. Thus the observed squelching was a function of recruitment of remodeling complex(es) and not through association of other coactivator complexes, and may reflect an initial binding threshold for target site accessibility. In addition to SNF/Swi, other ATP-dependent remodeling complexes have been isolated, including the distinct yeast RSC complex, and a related family of complexes which share homologs of the SNF2/Swi2related ISWI ATPase, namely the Drosophila complexes NURF, ACF, and CHRAC and the human RSF complex (recently reviewed in [74]). The ISWI-containing complexes were initially purified biochemically based on their abilities to catalyze the assembly of nucleosomal DNA (ACF [75]), physiologically space salt or polyglutamate-deposited nucleosomes (ACF [75], CHRAC [76], RSF [77]), enhance transcription factor binding (NURF [78]), and increase transcription from nucleosomal templates in vitro (NURF [79], RSF [77]). Recently, another distinct ATP-dependent remodeling activity was found associated with an HDAC-containing complex, NURD, which also possesses a subunit with limited homology to N-CoR, MTA-1 [80,81]. Intriguingly, antibodies directed against the CHD4/Mi-2 ATPase subunit of NURD partially relieve ligand-independent TR/RXR repression of the TRβA promoter in Xenopus oocytes by microinjection experiments [80]. Recently, two new chromatin remodeling complexes have been characterized through profound functional effects on VDR. VDR has been found to associate with a novel complex called WINAC, which shares subunits of the SWI/SNF and ISWI complexes [82]. WINAC interacts directly with VDR, albeit in a ligandindependent manner, through one constituent subunit, the Williams syndrome transcription factor (WSTF). Interestingly, the WSTF gene is deleted in patients with Williams syndrome, a neurodevelopmental disorder that, in addition to causing cognitive disorders, also results in congenital heart disease. Kitagawa and coworkers found, using chromatin immunoprecipitation (ChIP) assays, that WINAC is recruited to both negative and positive VDREs by VDR, and that WSTF overexpression potentiated VDR-mediated transactivation or transrepression. Presumably, this occurs through the ability of the WINAC complex to enhance both the assembly and disassembly of nucleosomal arrays. As it contains BRG1, WINAC functions in an ATP-dependent fashion. A second chromatin-remodeling complex recently described also appears to be important for VDR transcriptional function, at least in a purified, reconstituted
LEONARD P. FREEDMAN AND ALFRED A. RESZKA
transcription system. Using conventional biochemical purification, Lemon and co-workers [83] purified an activity that is required for ligand-dependent transactivation by VDR using a chromatin-dependent reconstituted in vitro transcription assay. This activity was identified as a multisubunit complex called PBAF, which shares many, but not all, of the same subunits as SWI/SNF-A. Using antibody depletion, it was found that only PBAF, but not SWI/SNF, supported VDRdependent transactivation, suggesting that at least in this experimental system, which is dependent on the presence of the DRIP complex, these two chromatin remodeling complexes are not functionally equivalent, nor does one complement the activity of the other. The ever-expanding number of chromatin-remodeling, histone-modifying, and Pol II–recruiting complexes that interact with and are utilized by nuclear receptors in general and VDR in particular makes this a particularly challenging scientific question to study. It also suggests that this wide repertoire could be the basis for tissue selectivity by VDR analogs and other nuclear receptor modulators (see next section). However, it also suggests that the process of transcriptional regulation involves multiple steps mediated by multiple complexes conferring distinct functions, whereby VDR and other nuclear receptors serve as ligand-regulated platforms for these complexes to the DNA in the proximity of regulated gene promoters. Some chromatin immunoprecipitation experiments are consistent with this so-called ordered-recruitment model, but it may also turn out that specific promoters (or more accurately, specific VDREs) impose the recruitment of a selective repertoire of cofactor complexes through allosteric effects perpetuated through DNA-bound VDR. The delineation of tissue selectivity by VDR ligands, then, will be an exceedingly complicated paradigm to define, but it may be that this complexity is actually what nature had in mind to achieve maximal regulatory flexibility.
IV. INTEGRATION OF SIGNALING PATHWAYS The WINAC, SRC-1/p160, and DRIP complexes represent three unrelated protein complexes, together carrying several distinct activities that have been shown to potentiate transcription activity of VDR (and other nuclear receptors). These activities may function together to provide a synergistic effect of 1,25(OH)2D3mediated activation, or to provide specificity of targeting to its regulation. VDR transcription regulation may require a combination of chromatin remodeling activities, as well as efficient recruitment of the RNA Pol II
269
CHAPTER 14 Vitamin D Receptor Cofactors: Function, Regulation, and Selectivity
motif in the DRIP complex as a combination of two subunits, and might explain the presence of two unrelated activation functions in the same receptor. In the TRAP/SMCC complex, both p53 and VP16 interact via TRAP80, but binding of VP16 and TR (which binds TRAP220) to the TRAP complex is not mutually exclusive, suggesting that this complex may support activation by a combination of distinct transcription factors. The transactivation domain (conserved region 3, CR3) of E1A interacts directly with hSur-2, which corresponds to DRIP130 [44]. This defines yet another novel and distinct target motif for an activator within the complexes. Several other classes of transcription factors have been found to functionally interact with DRIP subunits, and presumably the entire complex. One noteworthy example is interferon-activated ISGF3 transcription factor. ISGF3 is a heterotrimeric complex comprising STAT1, STAT2, and a DNA-binding subunit, IRF9. Upon induction by interferons, resulting in tyrosine phosphorylation of STATs, the complex enters the nucleus, binds to interferon–stimulated response elements, and activates transcription of regulated promoters. Lau and co-workers [88] have reported that ISGF3stimulated transcription is dependent on direct STAT2 interactions with DRIP150. This interaction leads to the recruitment of other DRIP subunits as well as RNA Pol II to ISGF3 target promoters, as demonstrated by chromatin immunoprecipitation. These results
machinery, via several of its basal factors. We have suggested a stepwise model that combines these distinct activities (Fig. 1), where WINAC and/or CBP/p160 coactivator complexes might be required for chromatin remodeling, followed by the direct recruitment of the transcription machinery by the DRIP complex. Definitive experimental analysis has yet to confirm this view [14,84]. In vitro, however, the SRC/p160 and DRIP coactivators appear to have no real intrinsic difference in their ability to interact with nuclear receptors. They both utilize the AF-2 of nuclear receptors and bind with similar affinities (TRAP220 versus TIF2 for TR [85]; DRIP205 and GRIP1 for VDR [86]. Thus we could envision a cooperative model where both CBP/p160 and DRIP complexes simultaneously occupy a promoter, and through their combined actions facilitate activation of transcription (Fig. 1). The observation that CBP can acetylate ACTR, leading to the latter’s dissociation from liganded ER [35], suggests a mechanism for the sequential model, whereby the first complex functions to acetylate histones and disrupt chromatin structure, whereupon it itself dissociates from the receptor, allowing the DRIP complex to bind and act at the level of direct recruitment. Recent studies have revealed multiple binding motifs for coactivators within the DRIP and TRAP complexes. For example, GR interacts with the DRIP complex via both AF-1 and AF-2 motifs with DRIP150 and DRIP205, respectively [87]. These results define the GR binding
DRIP/mediator (direct recruitment) p53
HAT complexes (nuclesome modification) SWI/SNF complexes (chromatin remodeling) CBP P300
BAF60 BAF155
STAT GR AR
80 150
SRC-1 p160
E1A 130
230
Med6 205
100
NFκB SREBP Sp1
Med7
PCAF
BAF53 BAF180
VP16
BRG/hBRM
HAT
ATP
RXR-VDR CTD TFIIH TFIIE TFIIA
Nuclesomes
FIGURE 1
DR3
RNA pol II
TFIID TFIIB
VDR associations with coactivator complexes
TFIIF
270 implicate an entirely distinct signaling system as functioning through the DRIP complex, and suggest that diverse pathways, such as the vitamin D and interferon signaling systems, may intersect or compete for limiting cofactors (such as the DRIP complex) by sharing subunits required for their optimal transactivation function.
V. MOLECULAR BASIS FOR TISSUE-SELECTIVE VITAMIN D RECEPTOR LIGANDS A. Established Paradigms The molecular basis for tissue selectivity of various VDR ligands is an area of considerable clinical interest. Most intriguing is our expanding understanding that mechanisms conferring tissue selectivity for one type of nuclear receptor may not fully apply to another. The concept of a tissue-selective ligand for a nuclear receptor was first used to describe ligands for the estrogen receptor (ER) that maintained certain estrogenic effects in the absence of others. To a great extent, such ligands, SERMs such as tamoxifen and raloxifene, act as antagonists in certain tissues and cell types (e.g., breast) while conferring some degree of agonism in others (e.g., bone). For the most part, receptor agonism by these ligands is less than that seen with the natural ligand, estradiol. Ligands for the VDR differ in the respect that superagonists have been described in addition to antagonists. For the superagonists, no equivalent ER or PR paradigm exists to help explain the genesis of such activity. For the antagonists, we find that, as with SERMs, some residual agonism is left intact, suggesting VDR modulator-like activity. Interestingly, the molecular basis for partial VDR agonism seems to differ from that described for ERα. Earlier work with ERα sought to explain tissue selectivity through X-ray crystallography analysis. Crystallography comparing the structure of ERα when liganded with estradiol or raloxifene revealed that the position of helix 12 differed, depending on which ligand (agonist or antagonist) was bound into the LBD [89]. Similar differences are predicted to exist for the progesterone receptor when bound to progesterone and the selective progesterone receptor modulator (SPRM), RU486 [90]. The shifting of H12 after modulator (SERM, SPRM, etc.) binding changes the topology of the AF-2 domain and alters the ability of the receptor to recruit coactivators. It also enhances the ability of the receptor to recruit co-repressors, and thus suppresses gene transcription. Although crystallographic analyses of the VDR have been reported, none have yet examined the
LEONARD P. FREEDMAN AND ALFRED A. RESZKA
conformation of the receptor in the presence of an antagonist. The X-ray structure in the presence of the natural ligand, 1α,25(OH)2D3, shows that VDR has a somewhat larger ligand-binding pocket (697 Å3) than that for ERα (369 Å3) or PR (422 Å3) [91]. This means that 1α,25(OH)2D3 occupies a smaller proportion of the pocket (56%) than with either ERα (63%) or PR (67%). This may suggest that a number of different ligands could bind to VDR and allow the induction of several conformations upon ligand binding. Although we do not have information regarding the structure of VDR in the presence of an antagonist, data do exist for the binding of superagonists [92]. Interestingly, regardless of whether it is bound by 1α,25(OH)2D3 or by either of two superagonists (MC1288 or KH1060), VDR seems to respond in a similar fashion. The overall protein structure is identical and the A- to D-ring moieties of each ligand form identical contacts, bound in identical orientations, with key residues within the VDR LBD. Conclusions based on this work include that superagonist-bound VDR might be more stable and exhibit a longer half-life than seen with the natural ligand. Other work suggests that superagonists can selectively enhance coactivator recruitment. This does resemble the selective effects of ERα and PR ligands on coactivator and corepressor recruitment, although several fundamental differences separate VDR from these nuclear receptors, as discussed below. Selective coactivator recruitment in and of itself is only part of how SERMs and SPRMs elicit tissuespecific responses. To a certain extent, a ligand can elicit only one conformation of the receptor, and this confers a specific affinity of the receptor for each individual coactivator and co-repressor. After this, the altered responses of nuclear receptors to these partial agonists vary cell-type to cell-type, based on the relative levels of expression of each of the coregulators. Thus in a cell with high levels of the right coactivator, e.g.,SRC-1, an agonistic response might be seen, whereas in other cell types, reduced expression of the required coactivator can result in better co-repressor recruitment and resulting transcriptional repression. This can also be achieved through higher levels of co-repressor expression within a given cell type. As noted above for ERα, the altered positioning of H12 upon binding to raloxifene changes the relative affinity of the receptor for the various co-regulators. The clearest evidence for ligand-induced and tissue-selective recruitment of coactivators was generated using chromatin immunoprecipitation (ChIP) analyses [93]. Using MCF-7 and Ishikawa cells as models for breast and uterine stimulation, estradiol was compared to both tamoxifen and raloxifene. Whereas both SERMs display similar antagonism of breast cancer growth, tamoxifen is
CHAPTER 14 Vitamin D Receptor Cofactors: Function, Regulation, and Selectivity
considered to be quite uterotropic. Not surprisingly, both breast cancer antagonists displayed a similar lack of ability to recruit coactivators, such as SRC-1, GRIP1, and AIB1, to estrogen-responsive promoters in MCF-7 breast cancer cells. In addition, both effectively recruited the SMRT and NCoR corepressors, along with two histone deacetylases, all of which act to suppress gene transcription. Interestingly, and consistent with their divergent effects on the uterus, raloxifene maintained this general profile in Ishikawa cells, whereas the response to tamoxifen in these cells resembled that of the natural agonist, estradiol. However, of the two promoters tested (c-Myc and cathepsin D) only the proliferative c-Myc promoter showed the agonist recruitment profile. How then is selectivity generated? In the case of MCF-7 versus Ishikawa cells, the selectivity can be traced back to the relative expression of the coactivator, SRC-1, whose expression is diminished in MCF-7 cells (versus that seen in Ishikawa) [93]. Increased expression (via transfection) of SRC-1, but not GRIP1 or AIB1, enhanced the ability of tamoxifen, but not raloxifene, to induce gene expression off of the c-Myc (and IGF-I) promoter. This type of selective transcriptional response, as influenced by altered ratios of coactivators and co-repressors, was also reported for PR in its response to RU486 [94]. Following the observation that RU486 acts as an agonist in T47D (breast cancer) cells and an antagonist in HeLa (cervical carcinoma) cells, it was observed that the T47D cell line has diminished expression of both NCoR and SMRT corepressors, although expression levels of SRC-1 appeared to be normal. Consistent with its ability to convert tamoxifen into an agonist in MCF-7 cells, overexpression of SRC-1 in HeLa cells can convert RU486 into a partial agonist. RU486 can also be converted into an antagonist in T47D cells, where it otherwise displays partial agonism, by the overexpression of SMRT. This results in the suppression of RU486, but not progesterone, recruitment of SRC-1, to PR and, consequently, suppressed ligand-induced transcription. Distinguished from the SPRM-like activities of RU486, the PR pure antagonist, ZK98, is refractive to SRC-1 overexpression and maintains its pure antagonist profile. Antagonism for this ligand does not require overexpression of any co-repressor in these cells.
B. The VDR Paradigm(s) 1. ANTAGONISTS/PARTIAL AGONISTS
With a paradigm in place for two well-characterized nuclear receptors, it seems logical that VDR should act in an identical fashion in response to partial agonists.
271
Agonists and superagonists should favor binding of coactivators over co-repressors, while partial antagonists should favor recruitment of co-repressors. To some extent this paradigm holds true, except where co-repressors are concerned. The binding of VDR to co-repressors in the unliganded state is relatively weak, and the physiological role for this interaction, if any, remains unknown. Indeed, mammalian two-hybrid analyses directly comparing VDR to thyroid hormone receptor (TR) showed that TR recruitment of SMRT is up to 26-fold stronger than that for VDR, while the recruitment of NCoR was an incredible 335-fold stronger [95]. One might expect that the binding of an antagonist ligand to VDR could enhance recruitment of either of these co-repressors, but this has not been seen. Indeed, the antagonists, ZK159222 and TEI-9647, actually displace a weakly bound NCoR in GST pulldown analyses. A similar response has also been reported for the Gemini ligand, which displaces NCoR binding in both GST pull-down and mammalian two-hybrid analyses (Gonzales et al., in preparation, 2003). Interestingly, however, overexpression of NCoR along with VDR and RXR has been found to partially suppress Gemini and EB1436-mediated transcription to a greater extent than that seen with the natural ligand, 1,25(OH)2VD3 [96]. This provides some evidence for ligand-specific effects of a co-repressor on VDRmediated gene transcription. Unresolved is the means by which this transcriptional repression is mediated, given the displacement of the co-repressor in the presence of the Gemini ligand. Meanwhile other recent studies have shown that the hairless co-repressor, which suppresses VDR-mediated transcription in the presence of 1,25(OH)2VD3, also binds to the VDR [97]. Consistent with its ability to suppress agonist-induced transcription, the hairless co-repressor binds VDR with equal affinity in the presence or absence of the natural ligand. Whether or not the hairless co-repressor could play a role in partial antagonism by vitamin D analogs remains to be established. 2. SUPERAGONISTS AND SELECTIVE COACTIVATOR RECRUITMENT
In considering the actions of VDR superagonists, several mechanisms seem to play a role, each changing, depending on the ligand tested (Fig. 2). For the superagonist ligand, numerous effects have been described relating to enhanced VDR-RXR heterodimer formation and enhanced recruitment of several of the coactivators. The formation of a VDR-RXR heterodimer in and of itself is an interesting aspect of VDR biology [98]. For many years, RXR was considered a silent partner for VDR, as transcription off of the heterodimer was dependent on VDR ligands alone. Recently, however,
272
LEONARD P. FREEDMAN AND ALFRED A. RESZKA
CBP P300 CBP P300
GRIP-1
80
PCAF 150
PCAF CBP P300
AIB-1
230
130 Med6
SRC-1 205
100
PCAF
Med7
CTD
DR3
Selective and partial agonists Natural ligand and selective/super agonists Super agonists RXR
FIGURE 2
VDR
O Ligand
Ligand-dependent coactivator complex recruitment to VDR
several lines of evidence have suggested that RXR plays a prominent role in VDR biology in both determining subcellular localization and, indeed, participating in the control of gene transcription. For instance, the dimerization domain of RXR encodes a nuclear localization sequence that is required for VDR translocation to the nucleus [99]. Mutation of this sequence traps VDR in the cytoplasm when in the unliganded state. Interestingly, however, 1,25(OH)2VD3 triggers the localization of NLS-deficient RXR-VDR into the nucleus. Paradoxically, it also enhances the export rate of the normal heterodimer from the nucleus. Within the nucleus, RXR also contributes to 1,25(OH)2VD3-induced transcription. This has been demonstrated by the ability of RXR-specific LXXLL motif peptides to suppress 1,25(OH)2VD3-induced transcription [100]. Very recently, VDR ligands have been shown to elicit a “phantom ligand effect” on RXR and thus to enable the recruitment of p160 family coactivators to RXR itself [101]. This could play a role in superagonist activity by certain VDR ligands, as suggested by the ability of certain of these ligands to induce heterodimer formation at concentrations well below that required by the natural ligand. This has
been demonstrated for the 20-epi ligands, MC1627, MC1288, and 2-methylene-19-nor-(20S)-1,25-dihydroxyvitamin D3 (2-MD), which bind VDR with equal affinity to 1,25(OH)2VD3, yet induce heterodimer formation at concentrations 10 to 100-fold lower [102–105]. It remains unknown whether or not the RXR moiety of the VDR heterodimer recruits coactivators under these circumstances. Enhanced heterodimer formation may contribute to the enhanced transcriptional activity of MC1288, but this is likely only part of the story (Fig. 2). Perhaps more important is the ability of this ligand to enhance recruitment of the DRIP complex to VDR [102]. Interestingly, while MC1288 induces enhanced DRIP205 recruitment, recruitment of SRC-1 and GRIP-1 coactivators is no greater than that seen with the natural ligand [102,103]. Meanwhile, 2-MD shows enhanced recruitment of DRIP205, SRC-1, and GRIP-1 [105], whereas MC1627 shows enhanced DRIP205 and GRIP-1 recruitment. In the context of myeloid cell differentiation, which is potently induced by both MC1288 and MC1627, it is the recruitment of DRIP205 that best correlates with the phenomenon. With regard to 2-MD, it remains unknown whether or not any specific
CHAPTER 14 Vitamin D Receptor Cofactors: Function, Regulation, and Selectivity
coactivator recruitment is responsible for enhanced bone anabolic activity. Selective recruitment of DRIP205 and p160 family members can also be regulated as part of the normal differentiation process, as demonstrated in the keratinocyte [106]. In this study, Vitamin D3 analogs were not examined, but by inference, analogs displaying selective coactivator recruitment are likely to show enhanced tissue selectivity. Using GST-VDR as bait, coactivator complexes were purified from both proliferating and differentiated keratinocytes in a 1α,25(OH)2VD3 dependent fashion. Interestingly, whereas DRIP205 and associated proteins predominated in the proliferating cells, AIB-1 (SRC-3) and GRIP-1 predominated after differentiation. This was associated with reciprocal changes in coactivator expression, DRIP205 declining with differentiation and AIB-1 increasing. Although overexpression of either cofactor could enhance transcription in the proliferation phase, only AIB-1 was effective postdifferentiation. We see this as intriguing, since ligands such as MC1288 can enhance recruitment of DRIP-205 during their induction of differentiation, as described earlier. Putatively, such selective coactivator recruitment, coupled with selective expression in the keratinocyte, could enhance the differentiation response in this system. Indeed, MC1288 suppresses proliferation and stimulates keratinocyte differentiation at concentrations 600- and 700-fold lower than seen with the natural ligand, respectively [107]. The enhanced potency of MC1288 in the keratinocyte versus myeloid cells, whereby MC1288 is approximately 100-fold more potent than 1α,25(OH)2VD3, may suggest a somewhat synergistic effect of the ligand specificity with the heightened and selective expression of DRIP-205 in the proliferating keratinocyte. That VDR should show different capacities to recruit one coactivator versus another in a ligand-dependent state is consistent with somewhat different means of recruitment for each coactivator. Indeed, a direct comparison of GRIP-1 and AIB-1 showed that different sequences within VDR are required to recruit each coactivator [108]. Furthermore, RXR seems to play a more active role in recruitment of AIB-1 than GRIP-1 in VDR-RXR-mediated transcription. Nonetheless, there is greater evidence for enhanced GRIP-1 recruitment by VDR ligands than for any other coactivator. Indeed, whereas several ligands can induce an interaction between VDR and SRC-1, GRIP-1, and AIB-1, OCT (22-oxa 1α,25(OH)2VD3) only poorly recruits either SRC-1 or AIB-1, while its recruitment of GRIP-1 is comparable to that of the natural ligand [109]. Consistent with this, GRIP-1, but not SRC-1, overexpression
273
substantially enhances OCT-induced transcription in a fashion reminiscent of SERM effects on ERα, as described earlier. Analyses of a broad panel of vitamin D analogs showed that GRIP-1 was more consistently recruited at levels closer to that of the natural ligand than was AIB-1 [104]. For these ligands, transcriptional activity was enhanced by overexpression of GRIP-1, whereas, AIB-1 overexpression actually suppressed transcription. To date, we know of only one example of cell type–dependent coactivator recruitment by VDR modulators. This type of activity was demonstrated for the noncalcemic analog Ro-26–9228, which recruits GRIP-1 in human osteoblasts, but not in CaCo-2 cells [110]. In this study, the coactivator was tested in vitro as a GST-fusion protein; therefore the differential recruitment of GRIP-1 to VDR in these two different cell types could not be attributed to altered expression of the coactivators. Since similar selective recruitment of DRIP-205 and SRC-1 was also reported, it seems that VDR is somehow modified by its expression in these different cell types. This also distinguishes this analog from numerous others, which show more or less selective recruitment of GRIP-1 over other tested coactivators. In summary, the VDR-RXR heterodimer seems to have its own mechanisms for the genesis of VDRMlike activity. As with ERα and PR, VDR can display somewhat selective coactivator recruitment in the presence of a broad range of tissue-selective ligands. Although the data are limited, the preponderance of evidence suggests that co-repressor recruitment is not at all involved in the genesis of VDRM activity. Binding of NCoR and SMRT is weak in the unliganded state, and this is further reduced by the binding of agonist and antagonist ligands. Coactivator-dependent effects on superagonism are also seen with certain ligands. These ligands display both enhanced VDRRXR heterodimer formation and enhanced recruitment of DRIP205 and, perhaps, GRIP-1 in a ligand-specific manner (Fig. 2). Overall, this suggests that the modulation of VDR transcriptional activity is coactivatorspecific. This sets a new paradigm for VDR that is distinct from that of ERα and PR, which respond to coactivator and co-repressors, depending on relative expression levels.
VI. CONCLUSIONS Transcriptional regulation by 1,25(OH)2D3 can be dissected into several functional activities that are mediated by VDR. Generally, transactivation requires that a repressed state of chromatin has to be disrupted
274 in a given target gene’s regulatory regions, together with the ability to promote productive elongation of RNA products by the RNA Pol II machinery at the site of transcription initiation. Several candidates for such activities that are recruited by nuclear receptors in direct response to ligand binding have been identified, as has been described in this review. The HAT activitycontaining coactivators (SRC/p160 family, CBP/p300, PCAF, etc.) appear at face value to act primarily in the disruption of chromatin through histone modifications, although a series of additional, provocative targets corresponding to ATP-dependent chromatin remodelers have more recently been identified. The regulation of RNA Pol II and its ability to respond to activators appears to be mediated by a distinct type of cofactor complex (DRIP/ARC/TRAP/SMCC/Mediator) whose functions are not yet completely elucidated. Given the fact that many classes of activators interact with DRIP beyond VDR and nuclear receptors, this complex must be considered as a regulatory panel for RNA Pol II rather than an exclusive target for nuclear receptors. DRIP may be therefore viewed as a downstream target of multiple transcription activators, perhaps conferring on RNA Pol II the ability to simultaneously integrate multiple signaling pathways onto a single promoter in vivo. This model may provide a key for an interpretation of some unresolved mechanisms involving the simultaneous cross-talk of several signaling systems.
References 1. Zawel L, Reinberg D 1995 Common themes in assembly and function of eukaryotic transcription complexes. Annu Rev Biochem 64:533–561. 2. Roeder RG 1996 The role of general initiation factors in transcription by RNA polymerase II. Trends Biochem Sci 21:327–335. 3. Lemon BD, Freedman LP 1999 Nuclear receptor cofactors as chromatin remodelers. Curr Opin Genet Dev 9:499–504. 4. Lee TI, Young RA 2000 Transcription of eukaryotic proteincoding genes. Ann Rev Genet 34:77–137. 5. Blanco JC, Wang IM, Tsai SY, Tsai MJ, O’Malley BW, Jurutka PW, Haussler MR, Ozato K 1995 Transcription factor TFIIB and the vitamin D receptor cooperatively activate ligand-dependent transcription. Proc Natl Acad Sci USA 92:1535–1539. 6. Jurutka P, Hsieh J, Remus L, Whitfield G, Thompson P, Haussler C, Blanco J, Ozato K, Haussler M 1997 Mutations in the 1,25-dihydroxyvitamin D3 receptor identifying C-terminal amino acids required for transcriptional activation that are functionally dissociated from hormone binding, heterodimeric DNA binding, and interaction with basal transcription factor IIB, in vitro. J Biol Chem 272:14592–14599. 7. Masuyama H, Jefcoat SC, Jr., MacDonald PN 1997 The N-terminal domain of transcription factor IIB is required for direct interaction with the vitamin D receptor and participates in vitamin D-mediated transcription. Mol Endocrinol 11:218–228.
LEONARD P. FREEDMAN AND ALFRED A. RESZKA
8. MacDonald P, Sherman D, Dowd D, Jefcoat SJ, DeLisle R 1995 The vitamin D receptor interacts with general transcription factor IIB. J Biol Chem 270:4748–4752. 9. Lemon BD, Fondell JD, Freedman LP 1997 Retinoid X receptor:vitamin D3 receptor heterodimers promote stable preinitiation complex formation and direct 1,25-dihydroxyvitamin D3-dependent cell-free transcription. Mol Cell Biol 17:1923–1937. 10. Mengus G, May M, Carre L, Chambon P, Davidson I 1997 Human TAF(II)135 potentiates transcriptional activation by the AF-2s of the retinoic acid, vitamin D3, and thyroid hormone receptors in mammalian cells. Genes Dev 11:1381–1395. 11. Lavigne AC, Mengus G, Gangloff YG, Wurtz JM, Davidson I 1999 Human TAF(II)55 interacts with the vitamin D3 and thyroid hormone receptors and with derivatives of the retinoid X receptor that have altered transactivation properties. Mol Cell Biol 19:5486–5494. 12. May M, Mengus G, Lavigne AC, Chambon P, Davidson I 1996 Human TAF(II28) promotes transcriptional stimulation by activation function 2 of the retinoid X receptors. EMBO J 15:3093–3104. 13. Torchia J, Glass C, Rosenfeld MG 1998 Co-activators and co-repressors in the integration of transcriptional responses. Curr Opin Cell Biol 10:373–383. 14. Freedman LP 1999 Increasing the complexity of coactivation in nuclear receptor signaling. Cell 97:5–8. 15. Onate SA, Tsai SY, Tsai MJ, O’Malley BW 1995 Sequence and characterization of a co-activator for the steroid hormone receptor superfamily. Science 270:1354–1357. 16. McKenna NJ, Xu J, Nawaz Z, Tsai SY, Tsai MJ, O’Malley BW 1999 Nuclear receptor coactivators: multiple enzymes, multiple complexes, multiple functions. J Steroid Biochem Mol Biol 69:3–12. 17. Xu L, Glass CK, Rosenfeld MG 1999 Coactivator and corepressor complexes in nuclear receptor function. Curr Opin Genet Dev 9:140–147. 18. Spencer TE, Jenster G, Burcin MM, Allis CD, Zhou JX, Mizzen CA, McKenna NJ, Onate SA, Tsai SY, Tsai MJ, Omalley BW 1997 Steroid receptor coactivator-1 is a histone acetyltransferase. Nature 389:194–198. 19. Chen HW, Lin RJ, Schiltz RL, Chakravarti D, Nash A, Nagy L, Privalsky ML, Nakatani Y, Evans RM 1997 Nuclear receptor coactivator ACTR is a novel histone acetyltransferase and forms a multimeric activation complex with P/CAF and CBP/p300. Cell 90:569–580. 20. Hong H, Kohli K, Garabedian MJ, Stallcup MR 1997 GRIP1, a transcriptional coactivator for the AF-2 transactivation domain of steroid, thyroid, retinoid, and vitamin D receptors. Mol Cell Biol 17:2735–2744. 21. Masuyama H, Brownfield CM, StArnaud R, MacDonald PN 1997 Evidence for ligand-dependent intramolecular folding of the AF-2 domain in vitamin D receptor-activated transcription and coactivator interaction. Mol Endocrinol 11:1507–1517. 22. Xu J, Qiu Y, DeMayo FJ, Tsai SY, Tsai MJ, O’Malley BW 1998 Partial hormone resistance in mice with disruption of the steroid receptor coactivator-1 (SRC-1) gene. Science 279:1922–1925. 23. Liu Z, Wong J, Tsai SY, Tsai MJ, O’Malley BW 1999 Steroid receptor coactivator-1 (SRC-1) enhances ligand-dependent and receptor-dependent cell-free transcription of chromatin. Proc Natl Acad Sci USA 96:9485–9490. 24. Baudino TA, Kraichely DM, Jefcoat SC, Jr., Winchester SK, Partridge NC, MacDonald PN 1998 Isolation and characterization of a novel coactivator protein, NCoA-62, involved
CHAPTER 14 Vitamin D Receptor Cofactors: Function, Regulation, and Selectivity
25.
26.
27.
28. 29. 30.
31. 32. 33. 34. 35. 36. 37.
38.
39.
40. 41. 42.
in vitamin D–mediated transcription. J Biol Chem 273: 16434–16441. Zhang C, Dowd DR, Staal A, Gu C, Lian JB, van Wijnen AJ, Stein GS, MacDonald PN 2003 Nuclear coactivator-62 kDa/Ski-interacting protein is a nuclear matrix-associated coactivator that may couple vitamin D receptor-mediated transcription and RNA splicing. J Biol Chem 278: 35325–35336. Le Douarin B, vom Baur E, Zechel C, Heery D, Heine M, Vivat V, Gronemeyer H, Losson R, Chambon P 1996 Liganddependent interaction of nuclear receptors with potential transcriptional intermediary factors (mediators). Phil Trans R Soc Lond B Biol Sci 351:569–578. Fraser RA, Heard DJ, Adam S, Lavigne AC, Le Douarin B, Tora L, Losson R, Rochette-Egly C, Chambon P 1998 The putative cofactor TIF1alpha is a protein kinase that is hyperphosphorylated upon interaction with liganded nuclear receptors. J Biol Chem 273:16199–16204. Goldman PS, Tran VK, Goodman RH 1997 The multifunctional role of the co-activator CBP in transcriptional regulation. Recent Prog Horm Res 52:103–119. McKenna NJ, Nawaz Z, Tsai SY, Tsai M-J, O’Malley BW 1998 Distinct steady-state nuclear receptor coregulator complexes exist in vivo. Proc Natl Acad Sci USA 95:11697–11702. Lanz RB, McKenna NJ, Onate SA, Albrecht U, Wong J, Tsai SY, Tsai MJ, O’Malley BW 1999 A steroid receptor coactivator, SRA, functions as an RNA and is present in an SRC-1 complex. Cell 97:17–27. Bannister AJ, Kouzarides T 1996 The CBP co-activator is a histone acetyltransferase. Nature 384:641–643. Ogryzko VV, Schiltz RL, Russanova V, Howard BH, Nakatani Y 1996 The transcriptional coactivators p300 and CBP are histone acetyltransferases. Cell 87:953–959. Gu W, Roeder RG 1997 Activation of p53 sequence-specific DNA binding by acetylation of the p53 C-terminal domain. Cell 90:595–606. Imhof A, Yang XJ, Ogryzko VV, Nakatani Y, Wolffe AP, Ge H 1997 Acetylation of general transcription factors by histone acetyltransferases. Curr Biol 7:689–692. Chen H, Lin RJ, Xie W, Wilpitz D, Evans RM 1999 Regulation of hormone-induced histone hyperacetylation and gene activation via acetylation of an acetylase. Cell 98:675–686. Rachez C, Freedman LP 2000 Mediator complexes and transcription. Curr Opin Cell Biol 13:274–280. Rachez C, Suldan Z, Ward J, Chang CP, Burakov D, Erdjument-Bromage H, Tempst P, Freedman LP 1998 A novel protein complex that interacts with the vitamin D3 receptor in a ligand-dependent manner and enhances VDR transactivation in a cell-free system. Genes Dev 12:1787–1800. Rachez C, Lemon BD, Suldan Z, Bromleigh V, Gamble M, Naar AM, Erdjument-Bromage H, Tempst P, Freedman LP 1999 Ligand-dependent transcription activation by nuclear receptors requires the DRIP complex. Nature 398:824–828. Naar AM, Beaurang PA, Robinson KM, Oliner JD, Avizonis D, Scheek S, Zwicker J, Kadonaga JT, Tjian R 1998 Chromatin, TAFs, and a novel multiprotein coactivator are required for synergistic activation by Sp1 and SREBP-1a in vitro. Genes Dev 12:3020–3031. Naar AM, Beaurang PA, Zhou S, Abraham S, Solomon W, Tjian R 1999 Composite co-activator ARC mediates chromatindirected transcriptional activation. Nature 398:828–832. Fondell JD, Ge H, Roeder RG 1996 Ligand induction of a transcriptionally active thyroid hormone receptor coactivator complex. Proc Natl Acad Sci USA 93:8329–8333. Ito M, Yuan CX, Malik S, Gu W, Fondell JD, Yamamura S, Fu ZY, Zhang X, Qin J, Roeder RG 1999 Identity between
43.
44. 45. 46.
47.
48. 49. 50. 51.
52.
53.
54.
55. 56.
57.
58.
59.
275
TRAP and SMCC complexes indicates novel pathways for the function of nuclear receptors and diverse mammalian activators. Mol Cell 3:361–370. Gu W, Malik S, Ito M, Yuan CX, Fondell JD, Zhang X, Martinez E, Qin J, Roeder RG 1999 A novel human SRB/MEDcontaining cofactor complex, SMCC, involved in transcription regulation [published erratum appears in Mol Cell 1999 Apr;3(4):following 541]. Mol Cell 3:97–108. Boyer TG, Martin ME, Lees E, Ricciardi RP, Berk AJ 1999 Mammalian Srb/Mediator complex is targeted by adenovirus E1A protein [see comments]. Nature 399:276–279. Ryu S, Zhou S, Ladurner AG, Tjian R 1999 The transcriptional cofactor complex CRSP is required for activity of the enhancer-binding protein Sp1. Nature 397:446–450. Sun X, Zhang Y, Cho H, Rickert P, Lees E, Lane W, Reinberg D 1998 NAT, a human complex containing Srb polypeptides that functions as a negative regulator of activated transcription. Mol Cell 2:213–222. Jiang YW, Veschambre P, Erdjument-Bromage H, Tempst P, Conaway JW, Conaway RC, Kornberg RD 1998 Mammalian mediator of transcriptional regulation and its possible role as an end-point of signal transduction pathways. Proc Natl Acad Sci USA 95:8538–8543. Parvin JD, Young RA 1998 Regulatory targets in the RNA polymerase II holoenzyme. Curr Opin Genet Dev 8:565–570. Myer VE, Young RA 1998 RNA polymerase II holoenzymes and subcomplexes. J Biol Chem 273:27757–27760. Koleske AJ, Young RA 1994 An RNA polymerase II holoenzyme responsive to activators [see comments]. Nature 368:466–469. Hengartner CJ, Myer VE, Liao SM, Wilson CJ, Koh SS, Young RA 1998 Temporal regulation of RNA polymerase II by Srb10 and Kin28 cyclin-dependent kinases. Mol Cell 2:43–53. Kim YJ, Bjorklund S, Li Y, Sayre MH, Kornberg RD 1994 A multiprotein mediator of transcriptional activation and its interaction with the C-terminal repeat domain of RNA polymerase II. Cell 77:599–608. Lee YC, Min S, Gim BS, Kim YJ 1997 A transcriptional mediator protein that is required for activation of many RNA polymerase II promoters and is conserved from yeast to humans. Mol Cell Biol 17:4622–4632. Myers LC, Gustafsson CM, Bushnell DA, Lui M, ErdjumentBromage H, Tempst P, Kornberg RD 1998 The Med proteins of yeast and their function through the RNA polymerase II carboxy-terminal domain. Genes Dev 12:45–54. Thompson CM, Young RA 1995 General requirement for RNA polymerase II holoenzymes in vivo. Proc Natl Acad Sci USA 92:4587–4590. Myers LC, Gustafsson CM, Hayashibara KC, Brown PO, Kornberg RD 1999 Mediator protein mutations that selectively abolish activated transcription [see comments]. Proc Natl Acad Sci USA 96:67–72. Rachez C, Lemon BD, Suldan Z, Bromleigh V, Gamble M, Naar AM, Erdjument Bromage H, Tempst P, Freedman LP 1999 Ligand-dependent transcription activation by nuclear receptors requires the DRIP complex. Nature 398:824–828. Chiba N, Suldan Z, Freedman LP, Parvin JD 2000 Binding of liganded vitamin D receptor to the vitamin D receptor interacting protein coactivator complex induces interaction with RNA polymerase II holoenzyme. J Biol Chem 275:10719–10722. Kretzschmar M, Kaiser K, Lottspeich F, Meisterernst M 1994 A novel mediator of class II gene transcription with homology to viral immediate-early transcriptional regulators. Cell 78:525–534.
276 60. Ge H, Zhao Y, Chait BT, Roeder RG 1994 Phosphorylation negatively regulates the function of coactivator PC4. Proc Natl Acad Sci USA 91:12691–12695. 61. Henry NL, Bushnell DA, Kornberg RD 1996 A yeast transcriptional stimulatory protein similar to human PC4. J Biol Chem 271:21842–21847. 62. Carlson M, Laurent BC 1994 The SNF/SWI family of global transcriptional activators. Curr Opin Cell Biol 6:396–402. 63. Cote J, Quinn J, Workman JL, Peterson CL 1994 Stimulation of GAL4 derivative binding to nucleosomal DNA by the yeast SWI/SNF complex. Science 265:53–60. 64. Imbalzano AN, Kwon H, Green MR, Kingston RE 1994 Facilitated binding of TATA-binding protein to nucleosomal DNA. Nature 370:481–485. 65. Wang W, Xue Y, Zhou S, Kuo A, Cairns BR, Crabtree GR 1996 Diversity and specialization of mammalian SWI/SNF complexes. Genes Dev 10:2117–2130. 66. Schnitzler G, Sif S, Kingston RE 1998 Human SWI/SNF interconverts a nucleosome between its base state and a stable remodeled state. Cell 94:17–27. 67. Guyon JR, Narlikar GJ, Sif S, Kingston RE 1999 Stable remodeling of tailless nucleosomes by the human SWI-SNF complex. Mol Cell Biol 19:2088–2097. 68. Yoshinaga SK, Peterson CL, Herskowitz I, Yamamoto KR 1992 Roles of SWI1, SWI2, and SWI3 proteins for transcriptional enhancement by steroid receptors. Science 258: 1598–1604. 69. Muchardt C, Yaniv M 1993 A human homologue of Saccharomyces cerevisiae SNF2/SWI2 and Drosophila brm genes potentiates transcriptional activation by the glucocorticoid receptor. EMBO J 12:4279–4290. 70. Ichinose H, Garnier JM, Chambon P, Losson R 1997 Liganddependent interaction between the estrogen receptor and the human homologues of SWI2/SNF2. Gene 188:95–100. 71. Sumi-Ichinose C, Ichinose H, Metzger D, Chambon P 1997 SNF2beta-BRG1 is essential for the viability of F9 murine embryonal carcinoma cells. Mol Cell Biol 17:5976–5986. 72. Beato M, Sanchez Pacheco A 1996 Interaction of steroid hormone receptors with the transcription initiation complex. Endocr Rev 17:587–609. 73. Fryer CJ, Archer TK 1998 Chromatin remodelling by the glucocorticoid receptor requires the BRG1 complex. Nature 393:88–91. 74. Kornberg RD, Lorch Y 1999 Chromatin-modifying and -remodeling complexes. Curr Opin Genet Dev 9:148–151. 75. Ito T, Bulger M, Pazin MJ, Kobayashi R, Kadonaga JT 1997 ACF, an ISWI-containing and ATP-utilizing chromatin assembly and remodeling factor. Cell 90:145–155. 76. Varga-Weisz PD, Wilm M, Bonte E, Dumas K, Mann M, Becker PB 1997 Chromatin-remodelling factor CHRAC contains the ATPases ISWI and topoisomerase II. Nature 388: 598–602. 77. LeRoy G, Orphanides G, Lane WS, Reinberg D 1998 Requirement of RSF and FACT for transcription of chromatin templates in vitro. Science 282:1900–1904. 78. Tsukiyama T, Wu C 1995 Purification and properties of an ATPdependent nucleosome remodeling factor. Cell 83:1011–1020. 79. Mizuguchi G, Tsukiyama T, Wisniewski J, Wu C 1997 Role of nucleosome remodeling factor NURF in transcriptional activation of chromatin. Mol Cell 1:141–150. 80. Xue Y, Wong J, Moreno GT, Young MK, Cote J, Wang W 1998 NURD, a novel complex with both ATP-dependent chromatin-remodeling and histone deacetylase activities. Mol Cell 2:851–861.
LEONARD P. FREEDMAN AND ALFRED A. RESZKA
81. Zhang Y, LeRoy G, Seelig HP, Lane WS, Reinberg D 1998 The dermatomyositis-specific autoantigen Mi2 is a component of a complex containing histone deacetylase and nucleosome remodeling activities. Cell 95:279–289. 82. Kitagawa H, Fujiki R, Yoshimura K, Mezaki Y, Uematsu Y, Matsui D, Ogawa S, Unno K, Okubo M, Tokita A, Nakagawa T, Ito T, Ishimi Y, Nagasawa H, Matsumoto T, Yanagisawa J, Kato S 2003 The chromatin-remodeling complex WINAC targets a nuclear receptor to promoters and is impaired in Williams syndrome. Cell 113:905–917. 83. Lemon B, Inouye C, King DS, Tjian R 2001 Selectivity of chromatin-remodelling cofactors for ligand-activated transcription. Nature 414:924–928. 84. Fondell JD, Guermah M, Malik S, Roeder RG 1999 Thyroid hormone receptor-associated proteins and general positive cofactors mediate thyroid hormone receptor function in the absence of the TATA box-binding protein-associated factors of TFIID. Proc Natl Acad Sci USA 96:1959–1964. 85. Treuter E, Johansson L, Thomsen JS, A Wr, Leers J, Pelto-Huikko M, Sjberg M, Wright AP, Spyrou G, Gustafsson J 1999 Competition between thyroid hormone receptor-associated protein (TRAP) 220 and transcriptional intermediary factor (TIF) 2 for binding to nuclear receptors. Implications for the recruitment of trap and p160 coactivator complexes. J Biol Chem 274:6667–6677. 86. Rachez C, Gamble M, Chang C-PB, Atkins GB, Lazar MA, Freedman LP 2000 The DRIP complex and SRC-1/p160 coactivators share similar nuclear receptor binding determinants but constitute functionally distinct complexes. Mol Cell Biol 20:2718–2726. 87. Hittelman AB, Burakov D, JA Ii-L, Freedman LP, Garabedian MJ 1999 Differential regulation of glucocorticoid receptor transcriptional activation via AF-1-associated proteins. EMBO J 18:5380–5388. 88. Lau JF, Nusinzon I, Burakov D, Freedman LP, Horvath CM 2003 Role of metazoan mediator proteins in interferonresponsive transcription. Mol Cell Biol 23:620–628. 89. Brzozowski AM, Pike AC, Dauter Z, Hubbard RE, Bonn T, Engstrom O, Ohman L, Greene GL, Gustafsson JA, Carlquist M 1997 Molecular basis of agonism and antagonism in the oestrogen receptor. Nature 389:753–758. 90. Williams SP, Sigler PB 1998 Atomic structure of progesterone complexed with its receptor. Nature 393:392–396. 91. Rochel N, Wurtz JM, Mitschler A, Klaholz B, Moras D 2000 The crystal structure of the nuclear receptor for vitamin D bound to its natural ligand. Mol Cell 5:173–179. 92. Tocchini Valentini G, Rochel N, Wurtz JM, Mitschler A, Moras D 2001 Crystal structures of the vitamin D receptor complexed to superagonist 20-epi ligands. Proc Natl Acad Sci USA 98:5491–5496. 93. Shang Y, Brown M 2002 Molecular determinants for the tissue specificity of SERMs. Science 295:2465–2468. 94. Liu Z, Auboeuf D, Wong J, Chen JD, Tsai SY, Tsai MJ, O’Malley BW 2002 Coactivator/corepressor ratios modulate PR-mediated transcription by the selective receptor modulator RU486.Proc Natl Acad Sci USA 99:7940–7944. 95. Tagami T, Lutz WH, Kumar R, Jameson JL 1998 The interaction of the vitamin D receptor with nuclear receptor corepressors and coactivators. Biochem Biophys Res Commun 253:358–363. 96. Herdick M, Carlberg C 2000 Agonist-triggered modulation of the activated and silent state of the vitamin D(3) receptor by interaction with co-repressors and co-activators. J Mol Biol 304:793–801.
CHAPTER 14 Vitamin D Receptor Cofactors: Function, Regulation, and Selectivity
97. Hsieh J-C, Sisk JM, Jurutka PW, Haussler CA, Slater SA, Haussler MR, Thompson CC 2003 Physical and functional interaction between the vitamin D receptor and hairless corepressor, two proteins required for hair cycling. J Biol Chem M304886200. 98. Cheskis B, Freedman LP 1994 Ligand modulates the conversion of DNA-bound vitamin D3 receptor (VDR) homodimers into VDR-retinoid X receptor heterodimers. Mol Cell Biol 14:3329–3338. 99. Prufer K, Barsony J 2002 Retinoid X receptor dominates the nuclear import and export of the unliganded vitamin D receptor. Mol Endocrinol 16:1738–1751. 100. Pike JW, Pathrose P, Barmina O, Chang CY, McDonnell DP, Yamamoto H, Shevde NK 2003 Synthetic LXXLL peptide antagonize 1,25-dihydroxyvitamin D3-dependent transcription. J Cell Biochem 88:252–258. 101. Bettoun DJ, Burris TP, Houck KA, Buck DW, II, Stayrook KR, Khalifa B, Lu J, Chin WW, Nagpal S 2003 Retinoid X receptor is a non-silent major contributor to vitamin D receptor-mediated transcriptional activation. Mol Endocrinol me.2003-0148. 102. Yang W, Freedman LP 1999 20-Epi analogues of 1,25-dihydroxyvitamin D3 are highly potent inducers of DRIP coactivator complex binding to the vitamin D3 receptor. J Biol Chem 274:16838–16845. 103. Liu YY, Nguyen C, Peleg S 2000 Regulation of ligandinduced heterodimerization and coactivator interaction by the activation function-2 domain of the vitamin D receptor. Mol Endocrinol 14:1776–1787. 104. Issa LL, Leong GM, Sutherland RL, Eisman JA 2002 Vitamin D analogue-specific recruitment of vitamin D receptor coactivators. J Bone Miner Res 17:879–890.
277
105. Yamamoto H, Shevde NK, Warrier A, Plum LA, DeLuca HF, Pike JW 2003 2-Methylene-19-nor-(20S)-1,25-dihydroxyvitamin D3 potently stimulates gene-specific DNA binding of the vitamin D receptor in osteoblasts. J Biol Chem 278:31756–31765. 106. Oda Y, Sihlbom C, Chalkley RJ, Huang L, Rachez C, Chang C-PB, Burlingame AL, Freedman LP, Bikle DD 2003 Two distinct coactivators, DRIP/Mediator and SRC/p160, are differentially involved in VDR transactivation during keratinocyte differentiation. Mol Endocrinol me.2003-0063. 107. Gniadecki R 1997 Effects of 1,25-dihydroxyvitamin D3 and its 20-epi analogues (MC 1288, MC 1301, KH 1060), on clonal keratinocyte growth: evidence for differentiation of keratinocyte stem cells and analysis of the modulatory effects of cytokines. Br J Pharmacol 120:1119–1127. 108. Issa LL, Leong GM, Barry JB, Sutherland RL, Eisman JA 2001 Glucocorticoid receptor-interacting protein-1 and receptor-associated coactivator-3 differentially interact with the vitamin D receptor (VDR) and regulate VDR-retinoid X receptor transcriptional cross-talk. Endocrinology 142: 1606–1615. 109. Takeyama K, Masuhiro Y, Fuse H, Endoh H, Murayama A, Kitanaka S, Suzawa M, Yanagisawa J, Kato S 1999 Selective interaction of vitamin D receptor with transcriptional coactivators by a vitamin D analog. Mol Cell Biol 19: 1049–1055. 110. Peleg S, Ismail A, Uskokovic MR, Avnur Z 2003 Evidence for tissue- and cell-type selective activation of the vitamin D receptor by Ro-26–9228, a noncalcemic analog of vitamin D3. J Cell Biochem 88:267–273.
CHAPTER 15
Vitamin D Nuclear Receptor Ligand-Binding Domain Crystal Structures NATACHA ROCHEL AND DINO MORAS Département de Biologie et de Génomique Structurales, IGBMC, CNRS/INSERM/Université Louis Pasteur, Illkirch, France
I. II. III. IV. V.
Introduction Biological Properties of hVDR∆ Solution Studies Crystal Structure of hVDR∆ Bound to 1α,25(OH)2D3 Mutant Analysis
I. INTRODUCTION All nuclear receptors (NRs) are modular proteins that harbor one DNA-binding domain and one ligandbinding domain (LBD) (Fig. 1) [1,2]. NRs act as agonistinduced factors that enhance or suppress transcription of their target genes. Certain receptors can act as silencers of transcription in the absence of ligands or the presence of certain antagonists. Agonists induce a change in the structure of the NR that allows interaction with coactivators that can acetylate histones, which prepare target gene promoters through decondensation of the corresponding chromatin [3]. Following decondensation, a second complex, called TRAP (thyroid receptor–associated proteins) or DRIP (Vitamin D receptor–interacting proteins) appears to establish linkage to the basal transcriptional machinery [4,5]. The ligand-binding domain of nuclear receptors harbors ligand-dependent activation function or AF-2, a major interface for dimerization with RXR and interface for coactivators as well as co-repressors. This domain is highly structured and encodes several functions in a ligand-dependent manner. Detailed molecular insights into the structure–function of nuclear receptors have been gained by the elucidation of the crystal structures of the LBD alone or in complexes with agonists, antagonists, and co-regulator peptides (Fig. 2). The first 3D structures reported for NR LBDs were those of the unliganded RXRα [6], the all-trans-retinoic acid–bound RARγ [7], and the agonist-bound thyroid receptor TRβ [8]. Unliganded receptors are referred to as apo receptor forms while liganded referred to as holo forms. To date, the crystal structures of 23 distinct NR LBDs VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
VI. VII. VIII. IX.
Crystal Structure of zVDR Bound to 1α,25(OH)2D3 Structure of hVDR Complexed to Superagonist Ligands Structure of zVDR in Complex with Gemini Conclusion References
and four LBD heterodimers have been reported. The later concern the EcR LBD bound to an antagonist and RXRα LBD bound to phospholipid ligand [9] and LXβ and RXRβ bound to agonists ligands [10]. The general fold of nuclear receptors consists of a threelayered α-helical sandwich. The helices have been designated H1 to H12, according to the first crystal structures of RXR and RAR [11]. The different crystal structures of apo and holo forms of RXR LBDs as well as extensive NR mutagenesis has demonstrated that the LBD undergoes a major conformational change upon ligand binding (Fig. 2). Ligand-induced conformational changes for NRs have been likened to the “mousetrap mechanism” described for RAR [7]. Upon ligand binding, helix H11 is repositioned in the continuity of helix H10, and helix H12 swings to seal the binding cavity while the ω-loop flips over underneath helix H6 carrying along the N-terminal part of helix H3. Helix 12, also referred as the activating domain AD of the AF-2 function, stabilizes ligand binding by contributing to the hydrophobic environment, in some cases making additional contacts with the ligand. The structural data reveal that H12, when folded back onto the core of the LBD, forms a hydrophobic cleft together with other surface-exposed residues that accommodates the “NR box” of coactivators such as members of the SRC-1/TIF2 family [12–14]. VDRs have been characterized from mammals [15–17], birds [18], Xenopus laevis [19], Paralichtus olivaceus [20], zebrafish (GenBank accession number AAF21427), and recently from lampreys [21]. Sequence analysis of the VDR subfamily members reveals that VDR presents a large insertion domain at the N-terminal Copyright © 2005, Elsevier, Inc. All rights reserved.
280
NATACHA ROCHEL AND DINO MORAS
A/B
C
D
E
N
F AD
C
DNA binding dimerization
DIMERIZATION NLS LIGAND BINDING
AF-1 Activation function ligand independent
AF-2 Activation function ligand-dependent interaction with cofactor
FIGURE 1 Structural and functional organization of nuclear receptors. NRs consist of six domains (A–F) based on regions of conserved sequence and function.
N
H9
H8
H1
H4 H10
H5
H3 H3 H7
H2
H12 9-cisRA
H11 H11
H6
part of the LBD in the peptide connecting helices H1 to H3. The length of this connecting region varies between 72 and 81 residues compared to 15–25 residues for the other nuclear receptors. This insertion region, for which no biological function has been found, is poorly conserved between the VDR species with only 9% identity. One mutation associated with the genetic disease vitamin D–resistant rickets type II has been found in this region (Cys190Trp of hVDR) with no effect on ligand binding [22]. A phosphorylation site also has been identified at Ser208 [23]. Secondary structure prediction reveals that this region is not structured. Among the NR LBD crystal structures solved, it has been reported frequently that the region connecting helices H1 to H3 is poorly ordered and shows little if any secondary structure. In hVDR, this region is sensitive to proteolysis with a trypsin cleavage site at the C terminus of Arg174 [24]. In order to stabilize the overall structure of the hVDR LBD, a VDR LBD mutant has been engineered by removing 50 residues in the region connecting helices H1 to H3 [25,26]. This VDR mutant, hVDR∆ (118-427∆166-216), stabilized the protein by lowering the number of conformations adopted by the insertion region. This hVDR∆ has been fully characterized in solution and compared to the wild-type hVDR.
H12
II. BIOLOGICAL PROPERTIES OF hVDR∆
FIGURE 2 Superimposition of unliganded (green and yellow) and liganded (blue and red) hRXRα LBD monomers. The main conformational differences affect helices H3, H11 and H12. The arrows show the main structural changes upon ligand binding. Adapted from Fig. 4A of Egea PF, Mitschler A, Rochel N, Ruff M, Chambon P, Moras D 2000 Crystal structure of the human RXRalpha ligand-binding domain bound to its natural ligand: 9-cis retinoic acid. EMBO J 19:2592–2601 [48].
The ability of hVDR∆ to bind ligands was determined by Scatchard analysis. This VDR mutant exhibits the same binding affinity for 1α,25(OH)2D3 and 1α,25(OH)2D3 analogs as the hVDR wild type [25,26]. Functional studies of the hVDR∆ and hVDR wild type were performed in order to compare their transactivation properties. In a first system, the hVDR LBD mutant and wild type were fused to the DNA binding domain of the yeast activator Gal4. The chimeric proteins were
CHAPTER 15 Vitamin D Nuclear Receptor Ligand-Binding Domain Crystal Structures
expressed by transient transfection in COS cells and transactivation was measured with a Gal4-responsive reporter. Both proteins exhibit comparable transactivation [25]. In a second system, the full-length hVDR mutant or wild type was cotransfected with a luciferase reporter plasmid containing the osteopontin gene VDRE under the control of the thymidine kinase promoter [26]. Similar results with regard to activation were obtained for the two VDR constructs, confirming the functional integrity of the hVDR∆ protein.
III. SOLUTION STUDIES Like the wild-type receptor, the hVDR∆ behaves as a monomeric species and is able to heterodimerize with RXR. The hVDR∆/RXRs LBDs complexed to 1α,25(OH)2D3 and 9-cis-retinoic acid form a stable species and the two partners are able to bind their ligands as seen by native electrospray ionization mass spectrometry and analytical ultracentrifugation [26]. Low-resolution structural information on the hVDR structure has been obtained by a small-angle X-ray scattering study [26,27]. The scattering function is time and space averaged. The radii of gyration (Fig. 3) of hVDR and hVDR∆ complexed to 1α,25(OH)2D3 were measured and are 26.2 Å and 23.4 Å, respectively. The difference of 2.8 Å, which would correspond to an increase in volume of 35% for a spherical object, is large and would suggest an appended insertion domain weakly connected to the main core LBD. The longest distance in the protein was also estimated in this study. A difference of 15 Å between the Dmax of hVDR and hVDR∆ suggests again that this insertion region may be positioned along the axis collinear to a N- to C-terminal axis including helix H12. This solution study confirms that the insertion domain is mobile and weakly connected to the rest of the LBD, which is itself structured similarly to that of the hVDR wild type.
IV. CRYSTAL STRUCTURE OF hVDR∆ BOUND TO 1α,25(OH)2D3 Deletion of the hinge region insertion in hVDR∆ stabilizes the protein and allowed the crystallization of the hVDR–1α,25(OH)2D3 complex [25]. The crystal structure was solved at 1.8 Å by a combination of molecular replacement using a homology model based on the retinoic acid receptor RARγ [7] and isomorphous replacement with a mercurial derivative. A higher resolution (1.5 Å) data set was collected later [28]. The overall topology of the hVDR∆ LBD (Fig. 4) is that of the canonical LBD with 13 α helices sandwiched in
281
three layers and a three-stranded β sheet. Helices H1 and H3 are connected by two small helices H2 and H3n, H3n replacing the ω loop of the RARγ structure [7]. The truncation of the hVDR∆ construct is positioned just before H3n as shown in Fig. 4. The VDR LBD structure is closely related to that of the holo hRARγ LBD structure [7] with a root mean square deviation (rmsd) of the superimposed structure of 1.2 Å over 179 residues. The connection between helices H1 and H3 follows a path between H3 and the tip of the β-sheet similar to that of the ERα structure [29]. The tip of the β sheet is consequently shifted outward and enlarges the ligand-binding cavity. The β sheet tip is stabilized by hydrogen bonds with residues of the H2-H3n loop. Helix 12 is in the agonist position and stabilized by two interactions with the ligand. Helix 12 is also stabilized by several hydrophobic contacts with residues of H3, H5, and H11, and two polar interactions with residues of H3 and H4 (Fig. 5). Some of these residues contact the ligand, thus indicating an additional indirect ligand control of the position of helix H12. In the hVDR∆ structure, a strong crystalline contact is observed between helix H3n and helices H3, H4, and H12 of a symmetrically related molecule, with H3n mimicking the coactivator SRC-1 peptide contacts [12–14]. Active vitamin D resides in a chair B conformation with the 19-methylene “up” and the 1α-OH and 3β-OH groups in an equatorial and axial orientation, respectively. A chair A conformation of the A-ring would disrupt the hydrogen bonds formed by the hydroxyl and the protein. The conjugated triene system connecting A-ring to C- and D-rings accommodates an almost trans conformation with the C6–C7 bond exhibiting a torsion angle of –149° that deviates significantly from the planar geometry, which results in the curved shape of the ligand bound to the receptor. The deviation explains the lack of biological activity of the analogs with a trans or a cis conformation of the C6–C7 bond [30]. The α face of the C-ring is lined by Trp286, whereas the methyl C18 on the β face points toward Val234 (H3). The ligand-binding pocket is lined by hydrophobic residues (Fig. 6). The elongated ligand embraces helix H3 with its A-ring oriented toward the C termini of helix H5 and the 25-OH close to helices H7 and H11. The 1-OH group forms two hydrogen bonds with Ser237 (H3) and Arg274 (H5), while the 3-OH group is hydrogen-bonded to Ser278 (H5) and Tyr143 (loop H1-H2). The 25-OH moiety forms two hydrogen bonds with His305 (loop H6-H7) and His397 (H11). The triene is tightly fitted in a hydrophobic channel sandwiched between Ser275 (loop H5-β) and Trp286 (β1) on one side and Leu233 (H3) on the other side. The aliphatic chain at position 17 of the D-ring adopts
282
NATACHA ROCHEL AND DINO MORAS
8.5
In I(Q)
A
7.5 6.5 5.5 0
0.002
0.004
Q2 (Å 8
B
9
C
7
8
In I(Q)
In I(Q)
6 5 4 3 2
0.006
−2)
7 6 5 4 3
1 0
0.05
0.1
Q (Å
D
0.15
0.2
0.05
−1)
0.1
Q (Å
E
1 0.9
0.9
0.8
0.8
0.7
0.7
0.6
0.6
0.5
0.2
0.25
0.5
0.4
0.4
0.3
0.3
0.2
0.2
0.1
0.1
0
0.15
−1)
1
P(r)
P(r)
0
0.25
0 0
50
r (Å)
0
50
r (Å)
FIGURE 3 Solution scattering analysis of hVDR∆ (▲) and hVDR wild type (◆) bound to 1α,25(OH)2D3. (A) Guinier analysis. (B) and (C) X-ray scattering curves of hVDR∆ and wild type, respectively. (D) and (E) Distance distribution function P(r). Adapted from Fig. 6 of Rochel N, Tocchini-Valentini G, Egea PF, Juntunen K, Garnier JM, Vihko P, Moras D 2001 Functional and structural characterization of the insertion region in the ligand-binding domain of the vitamin D nuclear receptor. Eur J Biochem 268:971–979 [26].
an extended conformation parallel to the C13–C18 bond with the C13–C17–C20–C22 torsion angle close to 90°; it is surrounded by hydrophobic residues. The ligand binding cavity of hVDR∆ is large (697 Å3) with the ligand occupying only 56% of this volume. A channel of water molecules near position 2 of the A-ring makes an additional space that can accommodate ligands with a methyl group at position 2. The fourfold increase in binding affinity of the 2α−methyl analog is in agreement with this observation [31]. Additional space around the aliphatic chain is also observed.
V. MUTANT ANALYSIS In order to investigate structurally and functionally important amino acid interactions within the ligandbinding pocket of the wild-type hVDR in the presence of several synthetic vitamin D analogs, Vaisanen et al. [32] have combined side-directed alanine mutagenesis with limited proteolytic digestion, electrophoretic mobility shift assay, and reporter gene assay. They have shown that structurally different agonists have distinct ligand–receptor interactions and that residues
283
CHAPTER 15 Vitamin D Nuclear Receptor Ligand-Binding Domain Crystal Structures
H4 H1 H9
H8
K264
H4 H10 H3
H5 E420 H3
H12
H11
S235
H12
H2 H7
D232 H6
T415
L414
Loop6–7
H2
H3n M412
R154
Insertion Domain 220ELTVS216................164GGGND160
FIGURE 4 Overall fold of hVDR∆ ligand-binding domain. The helices are represented as cylinders and β sheets as arrows. The whole structure is colored in gray except the helix H12 in purple. The ligand is depicted in yellow. The insertion region location is shown in green. The reconnected residues are indicated, together with their sequence numbering. Adapted from Fig. 1 of Rochel N, Wurtz JM, Mitschler A, Klaholz B, Moras D 2000 The crystal structure of the nuclear receptor for vitamin D bound to its natural ligand. Mol Cell 5:173–179 [25]. (See color plate).
FIGURE 5 Intramolecular interactions of helix H12 in VDR. The backbone of the protein is colored in gray except for helix H12 in purple. The side chains residues involved in polar stabilization of H12 are shown together with the hydrogen bonds (green dots). The ligand is depicted in yellow and red for the oxygen atoms. Adapted from Fig. 3 of Rochel N, Wurtz JM, Mitschler A, Klaholz B, Moras D 2000 The crystal structure of the nuclear receptor for vitamin D bound to its natural ligand. Mol Cell 5:173–179 [25]. (See color plate).
F150
F150
S278
O3
R274
Y147
Y143
Y147
Y143
S278
O3
R274
Y293
Y293 S237
S237 O1 L233
S275
I271
W286
I271 M272
V234
V234
V418
S275
O1 L233
W286
I268
V418
L313
M272 L230
I268
V300
L227
L313 V300
L227 F422
F422 H397 O25
H397 O25 L309
L309 Y401
H305
Y401
H305
FIGURE 6 Stereo view of 1α,25(OH)2D3 in the hVDR binding pocket. The ligand molecule is shown in the experimental electron density omit map contoured at 1.0 standard deviation. The hydroxyl groups are depicted as red spheres. Water molecules are shown as purple spheres. Hydrogen bonds are shown as green dotted lines. Carbon atoms are colored in gray, and oxygen and nitrogen atoms are colored in red and blue, respectively. Adapted from Fig. 3 of Rochel N, Wurtz JM, Mitschler A, Klaholz B, Moras D 2000 The crystal structure of the nuclear receptor for vitamin D bound to its natural ligand. Mol Cell 5:173–179 [25]. (See color plate).
284 H229, D232, E269, F279, and Y295 are critical for the agonist conformation [32]. Using a two-dimensional alanine scanning mutational analysis, Choi et al. [33] have studied the interactions between VDR and various vitamin D ligands. Eighteen alanine mutants of residues forming the ligand-binding pocket were studied. These residues form either hydrogen bonds with the ligands (Y143, S237, R274, S278, H305, H397) or hydrophobic interactions (D144, L233, V234, I238, I268, I271, S275, C288, W286, V300, Q400, Y401). The transactivation potencies of these mutants with the natural hormone and 11 other ligands were evaluated through functional studies with luciferase reporter assay and a mouse osteopontin response element. The comparison of six mutants of hVDR∆ and wild-type hVDR has shown that both LBPs are similar, thus providing additional evidence to support identical structures of VDR mutant and wild type. The importance of these residues in the interactions with specific ligands was evaluated in this study. Eight residues (Y143, D144, L233, I271, R274, W286, H397, Y401) have been shown to be essential for transactivation by vitamin D ligands, and mutations to less bulky hydrophobic residues increase the potency of 20-epi ligands [33].
VI. CRYSTAL STRUCTURE OF zVDR BOUND TO 1α,25(OH)2D3 In order to validate the conclusions made on the hVDR∆ structure and to find another crystal packing, the VDR from zebrafish (Danio rerio) was also studied. The zVDR exhibits 100% identity in the LBP lining residues and its transactivation potency is 50% of that of hVDR [34]. This activity is consistent with that of Xenopus laevis VDR or lamprey VDR, which show 50% [19] and 25% [21] of the activity of the hVDR, respectively. The zVDR LBD exhibits 69% identity and 79% similarity in its sequence with that of the hVDR LBD, while the insertion region (191–252 of zVDR) exhibits only 34% identity and 47% similarity. The zVDR LBD in complex with 1α,25(OH)2D3 was crystallized in presence of a coactivator peptide that contains the second NR box of SRC-1 [34]. The zVDR LBD (Fig. 7) adopts the canonical active conformation observed for agonist-bound receptors. The SRC-1 peptide (687-696 HKILHRLLQE) forms an amphipatic α-helix interacting with a hydrophobic cleft on the LBD surface. These interactions are similar to those described for other NRs [12–14]. In particular, Glu446 from H12 [Glu420] forms hydrogen bonds with the backbone amide nitrogen of Leu690 and Leu691.
NATACHA ROCHEL AND DINO MORAS
H1
H12
FIGURE 7 Overall fold of zVDR ligand-binding domain. The helices are represented as cylinders and β sheets as arrows. The whole structure is colored in gray and the SRC-1 peptide in dark gray. The ligand 1α,25(OH)2D3 is depicted in yellow.
At the other end, Lys274 from H3 [Lys246] forms a hydrogen bond to the main chain oxygen of Leu694. The structures of zVDR LBD and hVDR∆ LBD complexes with 1α,25(OH)2D3 are similar with a root mean square deviation of 0.72 Å over 236 main-chain atoms. The binding pocket is identical and the natural ligand adopts the same conformation and forms the same interactions with the protein in the two structures (Fig. 8).
H11
FIGURE 8 Superimposition of the LBP of hVDR∆-1α,25(OH)2D3
in green and zVDR-1α,25(OH)2D3 in yellow. The side-chain residues forming the binding pocket are shown together with the hydrogen bonds (dots).
CHAPTER 15 Vitamin D Nuclear Receptor Ligand-Binding Domain Crystal Structures
Notably the channel of water molecules at position 2 of the ligand, which was observed in the hVDR structure, is also observed in the zVDR. The differences observed when superimposing the two structures are small and primarily involve the loops. The ligandbinding pocket is conserved in both sequence and structure. The VDR-specific insertion region between helices H2 and H3 present in the zVDR construct and deleted in the hVDR mutant construct is not visible in the electron density map. This probably reflects the mobility of this region resulting in a static disorder. In the crystal structure of a VDR subfamily member, PXR [35,36], which also contains an insertion region between helices H1 and H3, the loop is also partially unresolved. The structural similarities between zVDR wild-type and hVDR∆ LBDs bound to 1α,25(OH)2D3 validate the overall biological relevance of the mutant hVDR structure.
VII. STRUCTURE OF hVDR COMPLEXED TO SUPERAGONIST LIGANDS Among the several synthetic analogs of vitamin D, the 20-epi compounds [37], which exhibit an inverted stereochemistry at position 20 in the flexible aliphatic chain, have attracted much attention. They are potent growth inhibitors and inducers of cell differentiation, while showing an affinity similar to that of 1α,25(OH)2D3 for VDR [37,38]. KH1060 (1α,25-dihydroxy20epi-22oxa-24,26,27-trihomovitamin D3 [37]), a member of this 20-epi family, exhibits similar properties with decreased calcemic side effects. The 20-epi compounds induce VDR-dependent transcription at concentrations at least 100-fold lower than the natural ligand and provoke antiproliferative activity several orders of magnitude higher than the natural ligand [37–40]. The differences in biological activity of 1α,25(OH)2D3 and the 20-epi molecules in general, and KH1060 in particular, are known to be VDR-LBD dependent. However, they are not yet understood. The ability of 20-epi analogs complexed to VDR to induce transcription appears to correlate with the ability of these compounds to promote coactivator interaction [40]. Differing proteolytic digestion patterns of VDR in the presence of the natural ligand and the 20-epi compounds have been interpreted to reflect large conformational changes in the receptor upon binding of the later class of molecules. In order to investigate the binding mode of the 20-epi analogs to the VDR-LBD, we have determined the high-resolution crystal structures of the hVDR∆ in complex with MC1288 (20-epi-1α,25(OH)2D3 [37]) and KH1060 and compared these structures to that obtained with the natural ligand [28] (Fig. 9).
285
When compared to hVDR∆-1α,25(OH)2D3 complex, the atomic models show a rms deviation on Cα atoms of 0.08 Å and 0.14 Å for hVDR∆-MC1288 complex and hVDR∆−KH1060 complex, respectively. Variations involve only some side chains located at the surface of the protein. In opposition to the view that the 20-epi analogs induce a different agonist conformation, the overall conformation and especially the position of helix H12 are strictly maintained in all three complexes. Furthermore, the ligand-binding cavity is unique and conserved for all three complexes. The rms deviations of all atoms forming the ligand pocket are 0.09 Å for hVDR∆-MC1288 complex and 0.12 Å for the hVDR∆-KH1060 complex. The sizes of the three ligands are 381 Å3, 375 Å3, and 392 Å3 for 1α,25(OH)2D3, MC1288, and KH1060, whereas the volume of the ligand pocket remains unaltered in the three complexes (660 Å3). The ligands occupy only 57% of the volume of the pocket for the VDR1α,25(OH)2D3 and VDR-MC1288 complexes and 59% for the VDR-KH1060 complex. The A, seco-B, and C/D rings form identical contacts as previously described for the 1α,25(OH)2D3-VDR complex. The hydroxyl groups make the same hydrogen bonds: 1-OH with Ser237 and Arg274, 3-OH with Tyr143 and Ser278, and 25-OH with His305 and His397. The deletion of 1-OH or 25-OH leads to the largest changes with a significant decrease in binding (1/1000), while that of the 3-OH has a smaller effect (1/20) [41]. The specific interactions observed in the three ligand–protein complexes involve the hydrophobic contacts of the 17β-aliphatic chains (Fig. 9B). When comparing the natural ligand and MC1288, the main difference is the positioning of the methyl group C21 which results in different contacts with Val300, Leu309, and His397. In MC1288, the C21 moiety is closer to His397. Other protein–ligand contacts differ, but to a lesser extent they involve the methyl groups at positions 23 (His305), 24 (His397), and 27 (His397 and Val418). In these two complexes, the carbon C22 makes no contact at a distance closer than 4.2 Å. In the case of KH1060, the methyl group C21 is quite close to that of the corresponding atom in 1α,25(OH)2D3. Note that the oxygen atom at position 22 of KH1060 forms a van der Waals contact with Val300. The major differences observed between KH1060 and the two other ligands are the tighter and more numerous ligand–protein contacts. The methyl groups C26α and C27α, specific to KH1060, form additional contacts with H3, loop 6–7, H11, and H12. A weak density is observed for the C26α methyl group, suggesting a structural disorder, whereas the C27α methyl group is clearly defined. In the three complexes, the ligands adopt an elongated conformation (Fig. 9C) similar to that described
286
NATACHA ROCHEL AND DINO MORAS
A
A
3-OH 25-OH
B
1-OH
FIGURE 9 Crystal structures of the VDR LBD complexed to 1α,25(OH)2D3, MC1288, and KH1060. (A) Experimental electron density omit map contoured at 2.0 standard deviation of (A) 1α,25(OH)2D3, (B) MC1288, and (C) KH1060. (B) Closeup view of KH1060 in the ligand-binding pocket. Secondary structure features are represented in blue (α-helices) and green (β-strands). The ligand is colored in yellow with the oxygen atoms in red. The volume of the cavity is represented in gray. (C) Superposition of 1α,25(OH)2D3 (yellow), MC1288 (green), and KH1060 (blue) ligands after superimposed VDR complexes. Oxygen atoms are colored in red. Adapted from Fig. 2 and 3 of Tocchini-Valentini G, Rochel N, Wurtz JM, Mitschler A, Moras D 2001 Crystal structures of the vitamin D receptor complexed to superagonist 20-epi ligands. Proc Natl Acad Sci USA 98:5491–5496 [28]. (See color plate).
C
B
C
1-OH
3-OH A D 25-OH C
CHAPTER 15 Vitamin D Nuclear Receptor Ligand-Binding Domain Crystal Structures
in the hVDR∆-1α,25(OH)2D3 structure. In the 20-epi complexes, the aliphatic chain is less constrained, thus allowing alternative conformations of the 17β-chain. The distances between the 25-OH group and the 1-OH and 3-OH are similar for the three complexes. These represent the anchoring points that must be maintained in order to obtain an active conformation. The aliphatic side chain of 1α,25(OH)2D3 adopts an extended conformation parallel to the C13–C18 bond with the C13–C17–C20–C22 torsion angle close to 90°. The two 20-epi analogs use different strategies to fit into the cavity. KH1060 accommodates its longer chain by adopting an eclipsed conformation around the O22–C23 bond, whereas MC1288 adopts a gauche conformation and has to compensate for a chain that is too short to maintain the 25-OH interactions because of the C20 inverted stereochemistry. Our preliminary docking experiments of the 20-epi analogs correctly positioned the methyl C21 in the same cavity as 1α,25(OH)2D3, but failed to give the correct geometry of the aliphatic chains. The crystallographic structures provide the exact information regarding these changes and the details of the interactions with the protein. Compared to the natural ligand, the aliphatic chains of the 20-epi analogs are lining the opposite side of the pocket. The different orientation of the chains is reflected by the different values of the torsion angles C16–C17–C20–C22 and C17–C20–C22(O22 for KH1060)–C23. Farther down the chain, while 1α,25(OH)2D3 adopts an extended conformation, the MC1288 and KH1060 exhibit gauche and eclipsed conformations, respectively, to properly orient the 25 hydroxyl group. The energetically unfavorable eclipsed conformation observed in KH1060 (C20–O22–C23–C24 of KH1060 equal to 16°) is made possible by the strong interactions formed by O22 with C12 (2.9 Å) and C18 (3.24 Å). A methylene group at this position as in MC1301 would be shifted away from the D-ring. A crude modeling analysis suggests that the methylene group would move away from C12, adopting a gauche (+) conformation instead of the eclipsed one observed for KH1060. As a result, additional contacts within the protein pockets are formed (Val300) that may contribute to the higher affinity of MC1301; other contributions such as solvation cannot be excluded. Similarly, the naturally occurring hydroxylation of C24α in KH1060 and C23 in 1α,25(OH)2D3 would form steric clashes with His305, affecting the hydrogen bond network with 25-OH. Based on the crystal structures this activity would depend on the capacity of the pocket to accommodate these additional groups. Both conformations and additional interactions of the 20-epi analogs are predicted to afford higher stability and longer half-life of the active complexes. Indeed,
287
within 3 hr, 60% of 1α,25(OH)2D3 dissociates from the VDR complex, whereas only 5–20% of MC1288 is dissociated [38]. Furthermore, receptors lacking the C-terminal helix 12 exhibit different dissociation behavior for the 20-epi analogs that are still capable of interacting with a mutated receptor [38,39]. Limited proteolytic digestion studies show that the 20-epi analog–receptor complexes are more resistant to digestion, suggesting that they are more stable. Additionally, an important role might be played by the rate of assimilation, i.e., how synthetic agonists are metabolized as compared to the natural ligand in different cells types [42,43].
VIII. STRUCTURE OF zVDR IN COMPLEX WITH GEMINI Among more than 3000 synthetic analogs of 1α,25(OH)2D3 synthesized to increase the potency and specificity of the physiological effects of vitamin D, Gemini (1α,25-dihydroxy-21-(3-hydroxy-3-methylbutyl)vitamin D3) is an interesting molecule with two identical side chains branching at carbon 20. Although Gemini binds less efficiently to VDR (38% compared to 1α,25(OH)2D3 [44]), its transactivation potency is similar to that of 1α,25(OH)2D3 in ROS cells [44] and 10-fold higher in HeLa and COS-7 cells [45]. It has been shown that in presence of an excess of co-repressor, the VDR–Gemini complex shifts from an agonist to an inverse agonist conformation through the recruitment of N-Cor and mediates repression [45,46]. Its stereochemistry and its 25% increase in volume when compared to the natural ligand create a major docking problem using the original LBP as a template. The packing constraints of the crystal form obtained for the hVDR∆ discriminate complexes with even small conformational changes near the ligand-binding pocket. In order to overcome this problem and obtain crystals of complexes that could not be crystallized with hVDR∆, we used the zVDR LBD construct. The zVDR in complex with Gemini and a coactivator peptide was crystallized and the structure solved at resolution 2.6 Å [34]. At the backbone level, the overall structure of zVDR-Gemini is almost identical to that of zVDR1α,25(OH)2D3 with a rms deviation of 0.37 Å over 249 main-chain atoms. However, the binding of the ligand’s second side chain results in a significant adaptation of the binding pocket that would have been difficult to predict [47]. A new pocket is created by the combined effect of a backbone shift and a side-chain conformational reorientation. The positions of helices H11 and H12 are unchanged and the peptide coactivator makes the same interactions with the protein, underlying the similar agonist character of both structures.
288 The most striking effect that emerges from this study is the “formation” of a new channel that extends the original pocket. This ligand-dependent pocket emphasizes the adaptability of the LBD and confirms the induced-fit mechanism of ligand binding. Despite large discrepancies in the shape and size of the ligand pocket, the overall structure of VDR and more importantly the position of H12 and of the coactivator peptide remain unaltered, consistent with the agonist character of Gemini. Thus the molecular mechanism of the agonism of Gemini is probably identical to that of 1α,25(OH)2D3.
IX. CONCLUSION The crystal structures of VDR LBD explain most features of ligand binding. These structures show the adaptability of the ligands and of the pocket in some specific cases. In all experimentally determined structures, the ligands were agonists and the protein conformation, notably H12, is conserved. The available structural information does not permit an understanding of the functional and structural role of the insertion domain. The loop 2-3, truncated in the hVDR and only partially visible in the zVDR structures, apparently does not influence the function of this nuclear receptor. Since this unique feature of VDR has not been lost during evolution, the question of its function remains an interesting one. However, the next step of structural studies should address the more general problem of the full-length receptor structure in different functional states.
Acknowledgment This work was supported by grants from CNRS, INSERM, Ministére de la Recherche et de la Technologie and by the Genopole and SPINE programs.
References 1. Moras D, Gronemeyer H 1998 The nuclear receptor ligandbinding domain: structure and function. Curr Opin Cell Biol 10:384–391. 2. Chambon P 1996 A decade of molecular biology of retinoic acid receptors. FASEB J 10:940–954. 3. McKenna NJ, Lanz RB, O’Malley BW 1999 Nuclear receptor coregulators: cellular and molecular biology. Endocr Rev 20:321–344. 4. Glass CK, Rosenfeld MG 2000 The coregulator exchange in transcriptional functions of nuclear receptors. Genes Dev 14:121–141.
NATACHA ROCHEL AND DINO MORAS
5. Freedman LP 1999 Increasing the complexity of coactivation in nuclear receptor signaling. Cell 97:5–8. 6. Bourguet W, Ruff M, Chambon P, Gronemeyer H, Moras D 1995 Crystal structure of the ligand-binding domain of the human nuclear receptor RXR-alpha. Nature 375:377–382. 7. Renaud JP, Rochel N, Ruff M, Vivat V, Chambon P, Gronemeyer H, Moras D 1995 Crystal structure of the RARgamma ligand-binding domain bound to all-trans retinoic acid. Nature 378:681–689. 8. Wagner RL, Apriletti JW, McGrath ME, West BL, Baxter JD, Fletterick RJ 1995 A structural role for hormone in the thyroid hormone receptor. Nature 378:690–697. 9. Billas IM, Iwema T, Garnier JM, Mitschler A, Rochel N, Moras D 2003 Nature 426:91–96. 10. Svensson S, Ostberg T, Jacobsson M, Norstrom C, Stefansson K, Hallen D, Johansson IC, Zachrisson K, Ogg D, Jendeberg L 2003 EMBO J. 22:4625–4633. 11. Wurtz JM, Bourguet W, Renaud JP, Vivat V, Chambon P, Moras D, Gronemeyer H 1996 A canonical structure for the ligand-binding domain of nuclear receptors. Nat Struc Biol 3:87–94. 12. Darimont BD, Wagner RL, Apriletti JW, Stallcup MR, Kushner PJ, Baxter JD, Fletterick RJ, Yamamoto KR 1998 Structure and specificity of nuclear receptor-coactivator interactions. Genes Dev 12:3343–3356. 13. Nolte RT, Wisely GB, Westin S, Cobb JE, Lambert MH, Kurokawa R, Rosenfeld MG, Willson TM, Glass CK, Milburn MV 1998 Ligand binding and co-activator assembly of the peroxisome proliferator-activated receptor-gamma. Nature 395:137–143. 14. Shiau AK, Barstad D, Loria PM, Cheng L, Kushner PJ, Agard DA, Greene GL 1998 The structural basis of estrogen receptor/coactivator recognition and the antagonism of this interaction by tamoxifen. Cell 95:927–937. 15. Baker AR, McDonnell DP, Hughes M, Crisp TM, Mangelsdorf DJ, Haussler MR, Pike JW, Shine J, O’Malley BW 1988 Cloning and expression of full-length cDNA encoding human vitamin D receptor. Proc Natl Acad Sci USA 85: 3294–3298. 16. Burmester JK, Maeda N, DeLuca HF 1988 Isolation and expression of rat 1,25-dihydroxyvitamin D3 receptor cDNA. Proc Natl Acad Sci USA 85:9499–9502. 17. Kamei Y, Kawada T, Fukuwatari T, Ono T, Kato S, Sugimoto E 1995 Cloning and sequencing of the gene encoding the mouse vitamin D receptor. Gene 152:281–282. 18. Lu Z, Hanson K, DeLuca HF 1997 Cloning and origin of the two forms of chicken vitamin D receptor. Arch Biochem Biophys 339:99–106. 19. Li YC, Bergwitz C, Juppner H, Demay MB 1997. Cloning and characterization of the vitamin D receptor from Xenopus laevis. Endocrinology 138:2347–2353. 20. Suzuki T, Suzuki N, Srivastava AS, Kurokawa T 2000 Identification of cDNAs encoding two subtypes of vitamin D receptor in flounder, Paralichthys olivaceus. Biochem Biophys Res Commun 270:40–45. 21. Whitfield GK, Dang HT, Schluter SF, Bernstein RM, Bunag T, Manzon LA, Hsieh G, Dominguez CE, Youson JH, Haussler MR, Marchalonis JJ 2003 Cloning of a functional vitamin D receptor from the lamprey (Petromyzon marinus), an ancient vertebrate lacking a calcified skeleton and teeth. Endocrinology 144:2704–2716. 22. Malloy PJ, Pike JW, Feldman D 1999 The vitamin D receptor and the syndrome of hereditary 1,25-dihydroxyvitamin D-resistant rickets. Endocrinol Rev 20:156–188.
CHAPTER 15 Vitamin D Nuclear Receptor Ligand-Binding Domain Crystal Structures
23. Jurutka PW, Hsieh JC, Nakajima S, Haussler CA, Whitfield GK, Haussler MR 1996 Human vitamin D receptor phosphorylation by casein kinase II at Ser-208 potentiates transcriptional activation. Proc Natl Acad Sci USA 93:3519–24. 24. Vaisanen S, Juntunen K, Itkonen A, Vihko P, Maenpaa PH 1997 Conformational studies of human vitamin-D receptor by antipeptide antibodies, partial proteolytic digestion and ligand binding. Eur J Biochem 248:156–162. 25. Rochel N, Wurtz JM, Mitschler A, Klaholz B, Moras D 2000 The crystal structure of the nuclear receptor for vitamin D bound to its natural ligand. Mol Cell 5:173–179. 26. Rochel N, Tocchini-Valentini G, Egea PF, Juntunen K, Garnier JM, Vihko P, Moras D 2001 Functional and structural characterization of the insertion region in the ligand-binding domain of the vitamin D nuclear receptor. Eur J Biochem 268:971–979. 27. Juntunen K, Rochel N, Moras D, Vihko P 1999 Large-scale expression and purification of the human vitamin D receptor and its ligand-binding domain for structural studies. Biochem J 344:297–303. 28. Tocchini-Valentini G, Rochel N, Wurtz JM, Mitschler A, Moras D 2001 Crystal structures of the vitamin D receptor complexed to superagonist 20-epi ligands. Proc Natl Acad Sci USA 98:5491–5496. 29. Brzozowski AM, Pike AC, Dauter Z, Hubbard RE, Bonn T, Engstrom O, Ohman L, Greene GL, Gustafsson JA, Carlquist M 1997 Molecular basis of agonism and antagonism in the oestrogen receptor. Nature 389:753–758. 30. Norman AW, Okamura WH, Hammond MW, Bishop JE, Dormanen MC, Bouillon R, van Baelen H, Ridall AL, Daane E, Khoury R, Farach-Carson MC 1997 Comparison of 6-s-cisand 6-s-trans-locked analogs of 1α,25-dihydroxyvitamin D3 indicates that the 6-s-cis conformation is preferred for rapid nongenomic biological responses and that neither 6-s-cis- nor 6-s-trans-locked analogs are preferred for genomic biological responses. Mol Endocrinol 11:1518–1531. 31. Fujishima T, Liu Z, Miura D, Chokki M, Ishizuka S, Konno K, Takayama H 1998 Synthesis and biological activity of 2-methyl-20-epi analogues of 1α,25-dihydroxyvitamin D3. Bioorg Med Chem Lett 8:2145–2148. 32. Vaisanen S, Ryhanen S, Saarela JT, Perakyla M, Andersin T, Maenpaa PH 2002 Structurally and functionally important amino acids of the agonistic conformation of the human vitamin D receptor. Mol Pharmacol 62:788–794. 33. Choi M, Yamamoto K, Itoh T, Makishima M, Mangelsdorf DJ, Moras D, DeLuca HF, Yamada S 2003 Interaction between vitamin D receptor and vitamin D ligands. Two-dimensional alanine scanning mutational analysis. Chem Biol 10:261–270. 34. Ciesielski F, Rochel N, Mitschler A, Kouzmenko A, Moras D 2004 J Steroid Biochem Mol Biol. 89–90:55–59. 35. Watkins RE, Wisely GB, Moore LB, Collins JL, Lambert MH, Williams SP, Willson TM, Kliewer SA, Redinbo MR 2001 The human nuclear xenobiotic receptor PXR: structural determinants of directed promiscuity. Science 292:2329–2333.
289
36. Watkins RE, Maglich JM, Moore LB, Wisely GB, Noble SM, Davis-Searles PR, Lambert MH, Kliewer SA, Redinbo MR 2003 2.1 A crystal structure of human PXR in complex with the St. John’s wort compound hyperforin. Biochemistry 42: 1430–1438. 37. Binderup L, Latini S, Binderup E, Bretting C, Calverley M, Hansen K 1991 20-Epi-vitamin D3 analogues: A novel class of potent regulators of cell growth and immune responses. Biochem Pharmacol 42:1569–1575. 38. Peleg S, Sastry M, Collins ED, Bishop JE, Norman AW 1995 Distinct conformational changes induced by 20-epi analogues of 1α,25-dihydroxyvitamin D3 are associated with enhanced activation of the vitamin D receptor. J Biol Chem 270: 10551–10558. 39. Liu YY, Collins ED, Norman AW, Peleg S 1997 Differential interaction of 1α,25-dihydroxyvitamin D3 analogues and their 20-epi homologues with the vitamin D receptor. J Biol Chem 272:3336–3345. 40. Yang W, Freedman LP 1999 20-Epi analogues of 1,25dihydroxyvitamin D3 are highly potent inducers of DRIP coactivator complex binding to the vitamin D3 receptor. J Biol Chem 274:16838–16845. 41. Bouillon R, Okamura WH, Norman AW 1995 Structurefunction relationships in the vitamin D endocrine system. Endocr Rev 16:200–257. 42. Dilworth FJ, Williams GR, Kissmeyer AM, Nielsen JL, Binderup E, Calverley MJ, Makin HL, Jones G 1997 The vitamin D analog, KH1060, is rapidly degraded both in vivo and in vitro via several pathways: principal metabolites generated retain significant biological activity. Endocrinology 138: 5485–5496. 43. Shankar VN, Dilworth FJ, Makin HL, Schroeder NJ, Trafford DJ, Kissmeyer AM, Calverley MJ, Binderup E, Jones G 1997 Metabolism of the vitamin D analog EB1089 by cultured human cells: redirection of hydroxylation site to distal carbons of the side chain. Biochem Pharmacol 53:783–793. 44. Norman AW, Manchand PS, Uskokovic MR, Okamura WH, Takeuchi JA, Bishop JE, Hisatake JI, Koeffler HP, Peleg S 2000 Characterization of a novel analogue of 1α,25(OH)2vitamin D3 with two side chains: interaction with its nuclear receptor and cellular actions. J Med Chem 43:2719–2730. 45. Bury Y, Herdick M, Uskokovic MR, Carlberg C 2001 Gene regulatory potential of 1α,25-dihydroxyvitamin D3 analogues with two side chains. J Cell Biochem 81:179–190. 46. Carlberg C 2003 Molecular basis of the selective activity of vitamin D analogues. J Cell Biochem 88:274–281. 47. Väisänen S, Peräkylä M, Kärkkäinen JI, Uskokovic R, Carlberg C 2003 Structural evaluation of the agonistic action of a vitamin D analog with two side chains binding to the nuclear vitamin D receptor. Mol Pharmacol 63:1230–1237. 48. Egea PF, Mitschler A, Rochel N, Ruff M, Chambon P, Moras D 2000 Crystal structure of the human RXRα ligand-binding domain bound to its natural ligand: 9-cis retinoic acid. EMBO J 19:2592–2601.
CHAPTER 16
Comodulators of Vitamin D Receptor–Mediated Gene Expression DIANE R. DOWD, AMELIA L. M. SUTTON, CHI ZHANG, AND PAUL N. MACDONALD Department of Pharmacology, Case Western Reserve University, Cleveland, Ohio
I. Introduction II. Coactivators III. Co-repressors
IV. Conclusion—Integrated Model of Comodulator Activity References
I. INTRODUCTION
with vitamin D response elements (VDREs) in the promoters of target genes. VDREs are characterized by two direct hexameric repeats with a spacer of three nucleotides (DR-3 elements) (Chapter 14). Although purified recombinant VDR interacts weakly with DR-3 elements, nuclear extracts containing one or more accessory factors were shown to dramatically enhance the binding affinity of the VDR to these VDREs [8,9]. Subsequent studies revealed that the nuclear accessory factor was the retinoid X receptor, or RXR (Fig. 1) [10,11]. RXR is currently well accepted as the biologically relevant heterodimeric partner for the VDR and other class II nuclear receptors (NRs) (reviewed in [12]). The association of 1,25(OH)2D3 with VDR promotes both the heterodimerization with RXR and high-affinity binding to DR-3 VDREs, with the VDR occupying the 3′ half-site and the RXR occupying the 5′ halfsite [13,14]. In this manner, the interaction of the liganded VDR/RXR with a VDRE confers target gene selectivity and ultimately influences the rate of RNA polymerase II (RNA Pol II)–directed transcription (Chapter 14). About 15 years ago, a phenomenon termed “squelching” was observed in which the ligand binding domain (LBD) of one NR interfered with the transcriptional activation mediated by a second NR [15,16]. The theory was that there are limiting quantities of accessory factors or adapter proteins that interact with the receptor’s LBD and are necessary for NR-mediated transcription. Thus, the LBD of one NR competes with the other intact liganded receptor for binding to these proteins. Indeed, it is becoming increasingly clear that additional protein–protein interactions are required to regulate 1,25(OH)2D3-dependent, VDR/RXR-mediated transcription. The regulation of VDR-mediated transcription likely involves a complex series of macromolecular interactions occurring in a temporally coordinated fashion. Association between the liganded
Nearly a century has passed since the discovery of fat-soluble vitamin D as a micronutrient that is essential to maintain appropriate bone mineralization. In 1970, the bioactive form of vitamin D was isolated and identified as 1,25-dihydroxyvitamin D3 (1,25(OH)2D3) [1–3]. Its main physiological role is to promote the intestinal absorption of dietary calcium and phosphate. Thus, 1,25(OH)2D3 ensures that the serum concentration of these ions is optimal for normal mineralization of the collagen matrix. Moreover, 1,25(OH)2D3 may directly affect bone remodeling by causing osteoblasts to terminally differentiate into osteocytes and deposit calcified matrix [4]. 1,25(OH)2D3 also promotes the differentiation of precursor cells into mature osteoclasts, which function to resorb bone and maintain appropriate bone remodeling [5]. In addition to its traditional role in maintaining calcium and phosphate homeostasis, the vitamin D endocrine system is also involved in a number of other physiological processes including blood pressure regulation, immune function, mammary gland development, and hair follicle cycling (reviewed in [6]). 1,25(OH)2D3 is generated by two sequential hydroxylations of vitamin D3 (cholecalciferol), a secosteroid precursor that is obtained in the diet or produced in the skin upon exposure to UV light (Chapter 2). 1,25(OH)2D3 is transported in the serum bound to the serum vitamin D binding protein (Chapters 8 and 9). 1,25(OH)2D3 dissociates from this transport protein and, because of its lipophilic nature, is thought to enter the cell by passive diffusion. Once inside the cell, 1,25(OH)2D3 is bound selectively by the vitamin D receptor or VDR (Chapter 11). VDR was first cloned in 1987 and was found to be a member of the superfamily of nuclear receptors (NRs) that regulates gene expression in a 1,25(OH)2D3-dependent manner [7]. The binding of 1,25(OH)2D3 to VDR enhances the association of VDR VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
Copyright © 2005, Elsevier, Inc. All rights reserved.
292
DIANE R. DOWD, AMELIA L. M. SUTTON, CHI ZHANG, AND PAUL N. MACDONALD
RXR
VDR
D
D RXR VDR
VDRE
D RXR VDR
VDRE
FIGURE 1 1,25(OH)2D3-induced heterodimerization between VDR and RXR. VDR binds to its ligand, 1,25(OH)2D3, heterodimerizes with RXR, and binds to a DR-3 type VDRE in the promoter of 1,25(OH)2D3-responsive genes. VDRs occupy the proximal or 3′ half-site while RXR occupies the distal or 5′ half-site.
VDR/RXR heterodimer and other transcriptional components may be classified into two general categories: general transcription factors and the comodulatory proteins. The interaction of VDR with the first group results in direct contacts with the preinitiation complex (PIC), which may facilitate assembly or recruitment of the PIC and thereby stimulate transcription by RNA Pol II. TFIIB and TATA binding-protein-associated factors (TAFs) are examples of this class of transcription factor, which modulates the activity of liganded VDR. In addition to direct contacts with the general transcription machinery, the liganded VDR is also linked to the transcriptional PIC by the NR comodulatory factors. NR comodulators are proteins that interact directly with NRs and modulate, either positively or negatively, their abilities to regulate their transcriptional activity. NR comodulators are classified either as coactivators or co-repressors, and they aid in the induction or repression, respectively, of ligand/receptormediated transcription. Coactivator proteins function to augment transcription by one of three proposed mechanisms. First, coactivators may function as macromolecular bridges between the liganded receptor and the general transcriptional machinery. In this capacity, they may recruit components of the PIC, aid in the
assembly of the PIC, or promote the stability of the complex. Second, some coactivators possess histone acetyltransferase (HAT) activity or recruit HAT activity to the promoter, which may loosen chromatin packaging, thereby making it more accessible to the basal transcriptional machinery. A third possible mechanism of coactivator function is to increase the rate of coupling between RNA Pol II–directed transcription and more distal events such as transcription elongation and RNA processing. For example, several coactivator proteins express domains that are consistent with RNA processing proteins, thus pointing to a role in both activated transcription and RNA splicing mechanisms (Section II, E). Thus, these multifunctional proteins may be able to interact with the nuclear receptor, general transcription factors, chromatin remodeling activities, and the splicing machinery to couple transcription to RNA processing and dramatically influence the rate at which a gene product is expressed. NR co-repressors are generally defined as proteins that interact with unliganded receptors to repress basal expression of hormone-responsive genes. They are distinguished from other general transcriptional repressors that interfere with NRs through distinct mechanisms such as binding to the receptor or to DNA to disrupt NR response-element binding. In this chapter, NR co-repressors are discussed as nuclear factors that interact with the unliganded NRs and directly modify histones or recruit modifying enzymes, such as histone deacetylases (HDACs), to maintain chromatin in a tightly packaged state and silence transcription from the promoter in an NR-dependent manner. Research spanning the past two decades reveals that 1,25(OH)2D3-mediated transcription is more complex than the simple binding of the receptor to DNA and the recruitment of RNA Pol II to initiate transcription. VDR/RXR-activated transcription involves complex interactions that may occur in a spatially distinct and temporally coordinated fashion to increase the rate at which 1,25(OH)2D3-responsive genes are transcribed and at which the resulting RNA transcript is processed. An integrated model of coactivator/co-repressor function (Fig. 2) proposes that type II NRs maintain a transcriptionally inactive state at a promoter by recruiting co-repressors and their associated HDAC activity. Upon binding ligand, the co-repressors either dissociate from the receptor or are displaced from the complex by the entrance of coactivator molecules. These coactivators then acetylate core histones while recruiting other coactivators and transcription factors, thereby creating a transcriptionally permissive environment at the promoter. At later stages, they may also promote the recruitment or stability of the RNA splicing machinery to enhance the rate at which mature RNAs are made and subsequently translated. The focus of this chapter
CHAPTER 16 Comodulators of Vitamin D Receptor–Mediated Gene Expression
cor
epr
ess
RXR VDR
ors
Ac
VDRE TATA
D
D RXR
coa
ctiv ato
rs
VDR RNA Pol II
VDRE
TBP TATA Ac Ac Ac Ac
FIGURE 2
Model of comodulator activity on VDR-mediated gene expression. The VDR/RXR heterodimer is loosely bound to a VDRE in the absence of ligand. In this state, it may interact with co-repressors, which function to keep gene transcription repressed, in part, by keeping histone proteins deacetylated. Upon binding ligand, the co-repressor is released and replaced by coactivators. The coactivators then function to remodel the chromatin and aid in the recruitment of RNA Pol II and other key components of the preinitiation complex.
is on the roles of the comodulator proteins of vitamin Ddependent transcription in the induction and repression of 1,25(OH)2D3-regulated gene expression.
II. COACTIVATORS The binding of 1,25(OH)2D3 to the VDR initiates an orchestrated cascade of protein assembly ultimately leading to transcriptional activation of select target genes. Ligand-induced coactivator recruitment to the VDR/RXR heterodimer is one of the early events in this transactivation process. These initial interactions between VDR and coactivators are the seed for the assembly of intricate multiprotein complexes that remodel the chromatin structure, recruit the core transcriptional machinery, and induce expression of 1,25(OH)2D3-regulated genes. Thus, understanding the molecular details of 1,25(OH)2D3-induced assembly of VDR and various coactivator complexes is central to the process of 1,25(OH)2D3/VDR-activated transcription.
A. Mechanism of Interaction VDR and other NRs exhibit a modular structure with three principle domains [17]. The amino terminus
293
of most NRs contains a transactivation function, although for VDR, this region is small and its function is poorly understood. Adjacent to this domain is the DNA-binding domain, which contains nine strictly conserved cysteine residues forming two zinc-binding modules that mediate high-affinity and highly selective binding to DNA. The large carboxyl-terminal ligandbinding domain (LBD) of VDR is organized into 12 alpha helices [18]. The LBD mediates association of VDR not only with 1,25(OH)2D3 [17], but also with its heterodimeric partner RXR [10,11,19] and comodulatory proteins [20]. Helix 12 of the VDR LBD contains the ligand-dependent activation function-2 (AF-2), which is essential for transactivation mediated by the VDR [19–21]. The molecular mechanisms involved in AF-2-dependent transactivation mediated by VDR are becoming increasingly clear based on structural analysis of VDR and other NRs. As 1,25(OH)2D3 nestles into the ligand-binding cavity of VDR, the AF-2 helix undergoes a subtle yet important conformational change. Helix 12 folds over the LBD [22] and, together with helices 3, 4, and 5, creates a hydrophobic crevice that selectively interacts with a complementary leucine-rich hydrophobic domain on many nuclear receptor coactivators (Fig. 3; [23]). This domain in the coactivators is termed the NR box and is composed of the consensus core sequence LXXLL [24]. The NR box forms an amphipathic α-helix whose orientation in the hydrophobic crevice of VDR is probably stabilized by a “charge clamp” composed of two conserved charged residues, one in the AF-2 helix 12 (E420) and the other in helix 3 of the VDR (K246) [25]. Besides the three highly conserved leucines, other residues in the NR box are also important for high-affinity binding and selectivity of coactivators for individual nuclear receptors [26–28]. While many coactivators interact with NRs via LXXLL motifs, others do not. Those coactivators that lack LXXLL motifs appear to interact with VDR and other nuclear receptors in an AF-2-independent manner (discussed in Section II,D). Thus, it is apparent that NRs can bind multiple coactivators using distinct motifs.
B. SRC Family of Coactivators The first nuclear receptor coactivator identified was steroid receptor coactivator-1 (SRC-1) [29]. SRC-1 was initially identified as a progesterone receptor–interacting protein that selectively enhanced hormone-dependent transcription. Subsequent studies have shown that SRC-1 is important for the transactivation of nearly all nuclear receptors including VDR [20,30,31]. SRC-1 constitutes the founding member of the p160 or SRC family, which also includes GRIP-1/TIF2/SRC-2 [32,33] and p/CIP/RAC3/ACTR/AIB-1/TRAM-1/SRC-3 [34–38].
294
DIANE R. DOWD, AMELIA L. M. SUTTON, CHI ZHANG, AND PAUL N. MACDONALD
FIGURE 3 The vitamin D receptor undergoes a conformational change upon binding hormone. The VDR binds ligand, and this induces a conformational change in helix 12 of the AF-2 domain. Helix 12 folds to trap the ligand in the binding pocket. This change also creates a hydrophobic cleft or surface on VDR composed of helices H3, H4, H5, and H12 that LXXLL motifs in coactivator proteins use for docking to the VDR.
SRCs are composed of several domains that are highly conserved among the coactivator family (Fig. 4). A basic helix–loop–helix domain and a PAS domain exists in the amino terminus of SRCs. The PAS domain is characteristic of the Per/Arnt/Sim family of transcription
factors. This amino-terminal region exhibits intrinsic transcriptional activity when tethered to a heterologous DNA-binding domain and is thought to mediate protein–protein interactions with other transcription factors, perhaps including other PAS proteins [39]. Three LXXLL-containing NR boxes are located in the mid-region of the SRCs and these mediate the liganddependent interaction with NRs [24]. Although these three NR boxes were first presumed to be functionally redundant, subsequent mutational studies showed that the individual LXXLL motifs and surrounding sequences confer some degree of receptor selectivity [27,40,41]. In the case of SRC-3, NR box III appears to be most important for interaction with VDR [42]. The carboxyl termini of SRCs contain a second autonomous transactivation domain [39]. This domain is characterized by a glutamine-rich region common to many transcriptional activators. The large carboxylterminal region of SRC-1 also mediates interactions with the transcription factors CREB-binding protein (CBP)/p300 [27,43], p300/CBP-associated factor (p/CAF) [44], and coactivator-associated arginine methyltransferase 1 (CARM1) [45]. These interactions involve additional LXXLL motifs distinct from the NR boxes described above. The precise mechanisms of transcriptional activation by SRCs are not entirely clear, but most studies point to the central role of SRCs in modifying the histone acetylation state and chromatin structure of hormoneresponsive promoters [46–48]. Once targeted to the appropriate promoter through interactions with NRs, SRCs are thought to remodel the chromatin, creating a template that is more accessible to the transcriptional machinery. This remodeling occurs through the covalent addition of acetyl groups onto the carboxyl-terminal lysine residues of histones. This acetyl modification weakens the electrostatic interaction between the positively charged histone tails and the negatively charged phosphate backbone of the DNA, thus inducing decondensation of the chromatin (reviewed in [49]). Such HAT activity is contained both in the SRCs themselves [36,47] and in several of the other transcriptional proteins that SRCs recruit, such as p300/CBP [50,51] and p/CAF [52]. In addition, p300/CBP binds directly to nuclear receptors and, together with SRC coactivators, they synergistically stimulate transcription [53,54]. Furthermore, SRCs also promote histone methylation, and presumably loosening of the chromatin structure, through their interactions with CARM1 [45]. Murine knockout models of each of the SRCs have been generated [55–57]. These models show a degree of functional redundancy in the SRC family, especially in the development and maintenance of the female reproductive system. However, the individual SRCs
CHAPTER 16 Comodulators of Vitamin D Receptor–Mediated Gene Expression
295
FIGURE 4
Schematic of the conserved SRC coactivator domains. bHLH, basic helix–loop–helix domain; PAS, region characteristic of the Per/Arnt/Sim family of transcription factors; NR boxes, nuclear receptor interaction boxes containing the LXXLL motifs; CBP interaction, region that interacts with CREB-binding protein; Q-rich, region rich in glutamine residues.
may also have distinct physiological functions perhaps related to nuclear hormone receptor selectivity. For example, deletion of SRC-2 results in aberrant spermatogenesis and testicular defects, whereas deletion of either SRC-1 or SRC-3 has no effect on the male reproductive tract [56]. Mutation of SRC-3 causes decreased estrogen-mediated protection against vascular damage [57]. Since the impact of these genetic deletions on VDR/ 1,25(OH)2D3-targeted systems has not been investigated, it is unknown which SRCs are important in vivo in VDR-mediated transcription. However, in vitro evidence suggests that SRC-2 may be a favored coactivator for VDR since SRC-2 is expressed in 1,25(OH)2D3 target cells such as osteoblasts and preferentially augments 1,25(OH)2D3-induced expression of the osteocalcin gene [58].
C. Mediator-D Complex In addition to SRCs, a large multiprotein complex called DRIP (vitamin D receptor interacting proteins) is a coactivator for VDR and other nuclear receptors [59,60]. DRIP shares many components with the transcriptional coactivator complexes thyroid hormone receptor (TR) activating proteins, activator recruited cofactor, and the mammalian Mediator complex [61–63]. In fact, these four complexes are most probably one in the same. These complexes interact with a variety of mammalian transcriptional activators, most notably NRs, SREBP-1a, and E1A. Thus, they likely have fundamental roles in activator-induced transcriptional processes well beyond the VDR and other NRs. Because of the considerable similarity between proteins comprising these various complexes, a unified nomenclature was proposed that
uses the Mediator complex as the basis for naming the multiple complexes and subunits [64]. Using this nomenclature, DRIP is referred to as Mediator-D. Mediator-D is proposed to interact with the liganded VDR bound to its enhancer element and then to recruit RNA Pol II, thereby acting as a bridge between VDR and the PIC [64,65]. Although mediator complexes may recruit RNA Pol II to the promoter, the polymerase is not tightly bound, thereby allowing for its release and the efficient initiation of transcription. Moreover, yeast mediator does not interact with the hyperphosphorylated form of RNA Pol II involved in transcriptional elongation [66]. These data suggest that mediator complexes are associated with RNA Pol II only during preinitiation and the transition to elongation. Moreover, in vitro transcription studies indicate that following the release of RNA Pol II, mediator remains bound to the promoter to function in the reinitiation of a second round of transcription [67]. Chromatin immunoprecipitation studies on hormone-responsive promoters allude to an even more dynamic cycling process of unknown function [68]. While the Mediator-D complex efficiently stimulates VDR-mediated transcription in vitro on chromatinized templates, Mediator-D and other mammalian mediator complexes do not contain detectable HAT activity [60]. However, recent evidence from yeast mediator complexes suggests they may assist in maintaining chromatin in a hyperacetylated, open conformation [69]. This raises the question of when and how the recruitment of chromatin remodeling complexes and mediator complexes takes place. A model that addresses this question proposes the sequential recruitment of coactivators and is described in more detail later (Section IV).
296
DIANE R. DOWD, AMELIA L. M. SUTTON, CHI ZHANG, AND PAUL N. MACDONALD
10 36 240 17 cdk8 95 130 230 33 97 150 220 78 34 70 100 D VDR
FIGURE 5 Model of the Mediator-D complex interacting with the liganded VDR. Med-220 uses an LXXLL NR box motif for docking to the VDR.
The Mediator-D complex is composed of at least 10 different proteins anchored by Med220, which interacts directly with ligand-activated VDR/RXR heterodimers through the second of two LXXLL motifs (Fig. 5, [70]). While Med220 promotes interaction with the VDR, another component of Mediator-D, Med130, may stimulate assembly of the PIC on the promoter [71]. The interaction between VDR and Med220 is enhanced by phosphorylation of VDR, and this correlates with an increase in 1,25(OH)2D3-mediated transcription [72]. Further support for the importance of this interaction comes from studies from Yang and Freedman [73] using 20-epi analogs of 1,25(OH)2D3. They showed a link between the increased transcriptional activity of these higher potency 20-epi compounds and increased binding of VDR to Med220 when compared to the effects of the 1,25(OH)2D3 ligand. This enhanced binding and transcriptional activity was also reflected in an increase in cellular differentiation and a decrease in proliferation. No differences were observed between 1,25(OH)2D3 and the 20-epi analogs when comparing the ability of these ligands to promote binding of VDR to SRC-2, thus highlighting the importance of mediator complexes in agonist-induced VDR activity. Genetic ablation of Med220 in mice causes embryonic lethality [74]. Moreover, fibroblasts derived from these animals exhibit attenuated thyroid hormone– stimulated transcription, but these cells retain normal retinoic acid responses, suggesting both receptor selectivity and a functional redundancy of NR coactivators in vivo [74]. Heterozygous mutants display growth retardation, impaired transcription and hypothyroidism, underscoring the primary importance of this mediator complex in TR-regulated processes and possibly other NR-mediated transcriptional events. Since VDRmediated transcriptional processes have not been examined in this model, it is unclear whether Med220 is required for 1,25(OH)2D3 action in vivo.
D. NCoA62/SKIP NCoA62 is a VDR and NR coactivator that was isolated as a VDR-interacting protein by a yeast two-hybrid screen [75]. Expression of NCoA62 in mammalian cells enhances vitamin D-, retinoic acid-, estrogen-, and glucocorticoid-activated transcription. NCoA62 is structurally and mechanistically distinct from SRCs and Mediator-D. However, it is highly related to Bx42, a Drosophila melanogaster nuclear protein putatively involved in ecdysone-stimulated transcription [76]. NCoA62 was independently identified as a factor that interacts with the Ski oncoprotein and was termed Ski-interacting protein, or SKIP [77]. Thus, it is referred to as NCoA62/SKIP throughout this chapter. NCoA62/SKIP has subsequently been shown to interact with a diverse array of transcription regulatory factors including CBF1 and silencing mediator for retinoic acid and thyroid hormone receptors (SMRT) [78], Smad2 and Smad3 [79], poly(A)-binding protein 2 [80], human papillomavirus (HPV-16) E7 oncoprotein [81], MIBP1, a member of the major histocompatibility complex binding protein family [82], and the retinoblastoma tumor suppressor protein [83]. These studies highlight a potential common role for NCoA62/SKIP in a more general aspect of target gene regulation by different classes of transcription factors. Importantly, the interaction between NCoA62/SKIP and VDR is independent of the AF-2 domain [75]. Moreover, ligand is not required for the interaction in vitro, although ligand does enhance the interaction. Unlike the SRC family of coactivators, NCoA62/SKIP does not contain obvious LXXLL motifs. NCoA62 also exhibits a marked preference for binding to the VDR-RXR heterodimer relative to VDR alone [84]. The region of NCoA62/SKIP between residues 274 and 342 mediates the interaction with the VDR-RXR heterodimer, and this region is referred to as the receptor interacting domain or RID (Fig. 6). The highly charged C-terminal domain is responsible for NCoA62/SKIP coactivator function in NR-mediated transcription. This region expresses an autonomous transactivation domain that is referred to as TAD-1 (reviewed in [85]). The nuclear localization sequence (NLS) of NCoA62/SKIP is contained in the C-terminal residues 531–536 and is necessary and sufficient for nuclear targeting of NCoA62/SKIP [86]. A strictly conserved LPXP motif is present in all reported NCoA62/SKIP orthologs from a variety of organisms [87], but the function of this domain or region is unknown at present. The mechanisms through which NCoA62/SKIP may function to augment VDR-mediated transcription are becoming more apparent. For example, selective,
CHAPTER 16 Comodulators of Vitamin D Receptor–Mediated Gene Expression
297
FIGURE 6 Schematic of NCoA62/SKIP. LPXP, a strictly conserved motif of unknown function; SNW, region containing a central conserved serine, asparagine, and tryptophan, as well as the receptor interaction domain (RID); TAD-1, transcription activation domain; NLS, nuclear localization sequence.
multiprotein complexes form between VDR, RXR, NCoA62/SKIP, and SRC coactivators [84]. NCoA62 and SRCs contact different domains within the VDRRXR heterodimer. Indeed, both NCoA62/SKIP, and SRC simultaneously interact with liganded VDR to form a ternary complex in vitro, which suggests that liganded VDR can act as a bridge to recruit both SRCs and NCoA62/SKIP to the promoter complex (Fig. 7) [84]. Coexpression of NCoA62/SKIP and SRCs leads to a synergistic induction of VDR-mediated transcription, and protein interference studies indicate that both coactivators are required for optimal VDR-mediated transcription [84]. Perhaps the most promising clues to the significance and mechanism of NCoA62/SKIP in VDR-activated transcription are found in recent chromatin immunoprecipitation (ChIP) approaches [86]. These studies show that NCoA62/SKIP is physically recruited in a 1,25(OH)2D3-dependent manner to the 1,25(OH)2D3responsive regions of native VDR target genes in osteoblasts including the osteocalcin and 24-hydroxylase promoters. Interestingly, NCoA62/SKIP is associated with the promoter region after the entry of both VDR and SRC, suggesting it may function at more distal steps of the transactivation process compared to the SRCs. As mentioned previously, the SRCs exhibit histone acetyltransferase (HAT) activity and recruit other proteins such as CBP/P300 that possess HAT
NCoA62/ D SRCs RXR VDR
VDRE
FIGURE 7
CBP/ p300 PIC
Ac
Ac Ac Ac Ac
Hypothetical model for VDR-SRC-NCoA62/SKIP complexes in VDR-activated transcription.
activity [47,88]. No significant HAT activity has been detected for NCoA62/SKIP (Zhang and MacDonald, unpublished observations). Thus, one possible interpretation is that NCoA62/SKIP enters the promoter region after the chromatin remodeling step. Additionally, NCoA62/SKIP is also known to interact with basal transcription factors such as TFIIB [89], suggesting that NCoA62/SKIP may also be involved in connecting VDR to the general transcription machinery. Regardless, NCoA62/SKIP and SRCs apparently function through different mechanisms to enhance VDR-mediated transcription, and the function of both distinct classes of coactivators is needed for appropriate vitamin D– activated target gene expression.
E. Nuclear Receptor Coactivators: Potential Links to RNA Processing In addition to chromatin remodeling and recruiting the transcription machinery, NR coactivators have been implicated in more distal steps of gene expression including transcription elongation and RNA processing. In this regard, NR coactivators have been proposed to be important coupling factors linking transcription and RNA processing [90]. The coupling concept is that RNA processing events (e.g., RNA capping, polyadenylation, and splicing) are physically connected to RNA Pol II and that RNA processing begins on the nascent RNA emerging from the polymerase complex (reviewed in [91–94]). Putative candidates that may aid in this coupling process are predicted to associate with and alter the functional properties of proteins involved in both transcription and RNA processing. In fact, many NR coactivator proteins contain domains or regions that are associated with RNA processing proteins, making some NR coactivators strong candidates as transcription-splicing couplers [90]. Indeed, Auboeuf et al. have demonstrated that progesterone receptor (PR)-activated transcription influences splicing decisions of alternatively spliced transcripts in a PR- and
298
DIANE R. DOWD, AMELIA L. M. SUTTON, CHI ZHANG, AND PAUL N. MACDONALD
progesterone response element–dependent manner [90]. NR coactivators such as CoAA and PGC-1 have been shown to alter RNA processing events in NRmediated transcription [90,95]. Thus, the idea put forth in these papers is that activated steroid receptors bind to target gene promoters and recruit factors that are involved in both transcription regulation and splicing regulation. This is thought to ensure efficient coordination between RNA synthesis and the final mature mRNA. NCoA62/SKIP may also play a central role in coupling transcription to mRNA slicing. Several observations support this hypothesis. First, endogenous NCoA62/SKIP is recruited to native VDR-responsive promoters in a 1,25-(OH)2D3-dependent manner in osteoblast-like target cells proving its physical association with the transcriptional regulatory machinery at the responsive promoter regions [86]. Second, studies characterizing protein components of the spliceosome and interchromatin granules, where many splicing factors are present, identified NCoA62/SKIP as a component that is associated with the splicing machinery in general and at discrete steps in the splicing mechanism [86,96–98]. Finally, a dominant-negative NCoA62/SKIP mutant effectively interferes with appropriate splicing of transcripts derived from a VDR-activated mini-gene cassette [86]. These studies indicate that disrupting the function of endogenous NCoA62/SKIP inhibits 1,25(OH)2D3-activated transcription, in part, by interfering with appropriate splicing of the mRNA transcripts. This supports the hypothesis that NCoA62/SKIP may couple VDR-mediated transcription to RNA splicing and ensure appropriate, efficient processing of vitamin D– regulated mRNA transcripts.
F. Other Coactivators In addition to these three classes of coactivators, several other proteins interact with VDR in a liganddependent manner and stimulate VDR-mediated transactivation. As mentioned earlier, VDR directly associates with several components of the transcriptional machinery. For example, VDR selectively binds to TFIIB [99,100], a component of the basal transcriptional complex whose entry is rate-limiting in the preinitiation complex formation. The LBD and amino-terminal regions of VDR are involved in contacting an aminoterminal region of TFIIB [99–101], and these two proteins function cooperatively to enhance ligand-dependent transcription [99]. VDR also interacts with TFIID and TFIIA, two essential components of the basal transcriptional machinery [102,103]. TFIID consists of the TATA-box-binding protein and associated factors
or TAFs. VDR associates with the TFIID subunit hTAF(II)135, and TAF(II)135 selectively potentiates VDR-mediated transcription. Direct interaction of VDR with these general factors further strengthens the link between VDR and the core transcriptional machinery and thus represents a putative mechanism in VDR-activated transcription. VDR also contacts Smad3, a transcription factor activated by TGF-β signaling [104]. Smad3 interacts with VDR in a ligand-dependent manner and stimulates VDR-mediated transcription. Additionally, SRC-1 enhances the binding of Smad3 to VDR, suggesting cooperation between these two coactivators. Given its function in two diverse signaling networks, Smad3 may serve as a bridge between 1,25(OH)2D3 and TGF-β signal transduction pathways. Interestingly, NCo62/SKIP serves as a coactivator for Smad-dependent transcription, lending additional support for the concept of crosstalk between these two pathways [79]. An additional binding partner for VDR is the helix–loop–helix transcription factor Ets-1 [105]. This interaction occurs between the DNA-binding domains of both proteins. Unlike other VDR coactivators, Ets-1 can stimulate VDR-mediated transcription in the absence of ligand on select promoters, such as the prolactin promoter.
III. CO-REPRESSORS The role of the co-repressor in regulating gene expression is generally the converse of that of the coactivator (reviewed in [106]). Co-repressors interact with DNA-bound transcription factors and play essential roles in silencing or repressing transcription. In the case of NR-regulated gene expression, co-repressors generally function to lower basal promoter activity in the presence of unliganded receptor; however, in some cases, repression can occur in the presence of NRs bound to antagonists.
A. SMRT and NCoR Perhaps the best characterized NR co-repressors are the ubiquitously expressed proteins SMRT [107] and nuclear receptor co-repressor (NCoR) [108]. SMRT and NCoR co-repressors strongly repress basal promoters in systems regulated by TR and the retinoic acid receptor (RAR). While the SRC family of coactivators interact with NRs via NR boxes (described earlier), SMRT and NCoR co-repressor interaction is mediated through co-repressor NR (CoRNR) boxes [109]. CoRNR boxes are generally composed of the
CHAPTER 16 Comodulators of Vitamin D Receptor–Mediated Gene Expression
sequence Φ-X-X-Φ-Φ, where Φ is a hydrophobic residue and X is any amino acid. Mutation of these sequences abolishes interaction of the co-repressors with unliganded NR [109]. In contrast to the SRC family of coactivators, SMRT and NCoR interaction with VDR is localized to the ligand binding domain but is independent of the AF-2 domain [110,111]. In fact, deletion of the AF-2 helix enhances co-repressor binding [107]. Although VDR is in the same NR family as TR and RAR, the silencing effects of SMRT and NCoR on VDR-regulated templates are weaker than in the other systems [112]. This is most probably due to a reduced binding of SMRT and NCoR to VDR relative to that of TR and RAR. It is possible that there may be antagonists for the VDR that could enhance the binding of the receptor to the co-repressors and thus increase transcriptional silencing. This means of interaction has been documented for co-repressor recruitment to the ER [113,114] and the PR [115]. The mechanism of repression of the SMRT/NCoR class involves interaction with Sin3, a protein shown to be part of a complex that includes HDAC1 and 2 [116–118]. Thus, the Sin3 proteins Sin3A and Sin3B recruit HDAC activity to target gene promoters and their subsequent activities result in compact or tightly wound nucleosomal packaging that ultimately represses basal expression of the promoter.
B. Hairless As mentioned earlier, the VDR plays a role in mammalian hair follicle cycling and disruption of the VDR gene, in some cases, leads to total alopecia (for review see [119]). Inactivation of the hairless gene (hr) also results in a similar phenotype [120]. Hr can repress transcription mediated by VDR, TR and RARrelated orphan receptor, and this discovery provides a molecular basis for its role in the maintenance of hair growth [121–123]. Although Hr lacks homology with SMRT and NCoR, it functions in a similar manner. Like these other co-repressors, Hr binds to VDR and the other NRs in a region localized to the central portion of the ligand binding domain and independent of AF-2 [123], the docking site for some coactivators. This interaction requires two Φ-X-X-Φ-Φ hydrophobic motifs in Hr [123]. Although Hr does not exhibit histone deacetylase activity, it is present in complexes with multiple HDACs. Hr also is found in matrixassociated deacetylase bodies together with SMRT and multiple HDACs [121]. Whereas SMRT and NCoR are ubiquitously expressed, the expression of Hr is restricted primarily to brain and skin, and thus may serve a more specialized role (reviewed in [124]).
299
C. Alien In 1999, Dressel et al. described a novel co-repressor, Alien, which is unrelated to the SMRT/NCoR family of co-repressors [125]. In these early studies, Alien interacted with TR in the absence of ligand and repressed TR-mediated basal transcription. Alien does not interact with the glucocorticoid receptor, RXR, or RAR. In this system, the mechanism of Alien-mediated silencing appears to be through the recruitment of Sin3A and its associated histone deacetylase activity. Alien did not interact with SMRT or NCoR and thus represents an independent pathway to Sin3 recruitment. Polly et al. [126] extended these findings to demonstrate that Alien also interacts with VDR and represses basal transcription from a DR-3 VDRE, but not an atypical IP9-type VDRE, suggesting a level of specificity for repression. The interaction of Alien with unliganded VDR is independent of the AF-2 domain of VDR and appears to use a different interaction surface than does NCoR. In contrast to the studies by Dressel, a deacetylase inhibitor had little effect on the ability of Alien to repress transcription. Thus, in the VDR system, Alien appears to be using alternate molecular pathways to exert its corepressor effect.
IV. CONCLUSION—INTEGRATED MODEL OF COMODULATOR ACTIVITY Gene expression is a complex process involving the coordinated repression and activation of transcription. According to the present model of nuclear hormone signaling, unliganded VDR, like other NRs, are DNAbound and complexed with co-repressor molecules that keep chromatin condensed and the promoter inaccessible to the transcription machinery. Upon binding 1,25(OH)2D3, the co-repressors are displaced by coactivators to begin the transcriptional process. To date, more than 30 nuclear receptor coactivators have been identified and recent effort focuses on distinguishing the redundant and unique functions of these proteins. Immunodepletion and dominant negative approaches indicate that SRCs, Mediator-D, and NCoA62/SKIP have different functions and all are required for robust VDR-mediated transcription. However, the temporal and spatial details of how these diverse coactivators and other accessory transcription factors are assembled onto a 1,25(OH)2D3-responsive promoter remain to be determined. Chromatin immunoprecipitation (ChIP) assays have proven instrumental in providing a limited picture of how endogenous VDR and coactivators convene on a native promoter template by capturing transcriptional complexes assembled on the promoter
300
DIANE R. DOWD, AMELIA L. M. SUTTON, CHI ZHANG, AND PAUL N. MACDONALD
at various times. ChIP assays demonstrate that both nuclear receptors and their coactivators cycle on and off the promoter [68,86,127]. Upon ligand addition, VDR, or other nuclear receptors, enters the transcriptional complex first, followed by SRCs (Fig. 8; [68,86]). SRCs likely loosen the chromatin structure by both bringing intrinsic HAT activity and recruiting extrinsic HAT activity to the complex, through their interaction with p300/CBP. SRCs then dissociate, allowing for binding
LX
SRCs
XL
D RXR VDR
L
p300 VDRE
Ac
p300 SRCs Med 220
L
XL
LX
D RXR VDR
RNA Pol II TBP
VDRE NCoA62/ SKIP
Ac Ac Ac Ac
spliceosome NCoA62/ SKIP
References
Med 220
L
VDRE
XL LX
D RXR VDR
Ac Ac Ac Ac
TBP
RNA Pol II Ac Ac Ac Ac
of the Mediator-D multimeric complex [68,128]. Mediator-D is thought to recruit the PIC and RNA Pol II holoenzyme to initiate transcription of target genes [65]. NCoA62/SKIP also enters the VDR complex following SRC-1 [86], but the relative entrance of NCoA62/SKIP and Mediator-D to the transcriptional complex is currently unknown. It is possible that SRCs help to recruit NCoA62/SKIP to the complex as suggested by the data that these two proteins can form a stable ternary complex with VDR [84]. Once NCoA62/SKIP is bound, it may target the spliceosome complex to the actively transcribed gene [86], allowing for efficient splicing of the nascent transcript. Taken together, these ChIP studies suggest a temporal model of coactivator action in which all three major classes of coactivators enter the complex at distinct times and provide three different functions: chromatin remodeling, recruitment of the core transcriptional machinery, and coupled transcriptional activation and RNA splicing. Importantly, one must be cautious not to oversimplify what is obviously a hugely complex process that requires a variety of macromolecular machines and coordination of many complex pathways. The model presented in Fig. 8 is meant to represent one global means to integrate the known actions of comodulator proteins, but the reality will likely turn out to be a much more complicated scenario.
Ac Ac Ac Ac
FIGURE 8 Model of the sequential occupation of the promoter site with coactivators, the PIC and the spliceosome. First, ligandactivated VDR/RXR binds to VDREs in target genes and recruits coactivators such as SRCs and p300/CBP. The HAT activity of these coactivators loosens the chromatin structure, allowing for a more transcriptionally permissive environment. Second, SRCs and p300/CBP dissociate, allowing for binding of the Mediator-D multimeric complex. This complex is thought to recruit RNA Pol II and the core transcriptional machinery to initiate active transcription of the target gene. Finally, NCoA62/SKIP binds to the VDR/RXR complex and tethers the splicing machinery to the activated promoter, allowing for immediate splicing of the nascent transcript.
1. Fraser DR, Kodicek E 1970 Unique biosynthesis by kidney of a biologically active vitamin D metabolite. Nature 228: 764–766. 2. Kodicek E, Lawson DEM, Wilson PW 1970 Biological activity of a polar metabolite of vitamin D3. Nature 228: 763–765. 3. Holick MF, Schnoes HK, DeLuca HF, Suda T, Cousins RJ 1971 Isolation and identification of 1,25-dihydroxycholecalciferol: a metabolite of vitamin D active in the intestine. Biochemistry 10:2799–2804. 4. Owen TA, Aronow MS, Barone LM, Bettencourt B, Stein GS, Lian JB 1991 Pleiotropic effects of vitamin D on osteoblast gene expression are related to the proliferative and differentiated state of the bone cell phenotype: dependency upon basal levels of gene expression, duration of exposure, and bone matrix competency in normal rat osteoblast cultures. Endocrinology 128:1496–504. 5. Bar-Shavit Z, Teitelbaum SL, Reitsma P, Hall A, Pegg LE, Trial S, Kahn AJ 1983 Induction of moncytic differentiation and bone resorption by 1,25(OH)2D3. Proc Natl Acad Sci USA 80:5908–5911. 6. Sutton AL, MacDonald PN 2003 Vitamin D: more than a “bone-a-fide” hormone. Mol Endocrinol 17:777–791. 7. McDonnell DP, Mangelsdorf DJ, Pike JW, Haussler MR, O’Malley BW 1987 Molecular cloning of complementary DNA encoding the avian receptor for vitamin D. Science 235:1214–1217. 8. Liao J, Ozono K, Sone T, McDonnell DP, Pike JW 1990 Vitamin D receptor interaction with specific DNA requires
CHAPTER 16 Comodulators of Vitamin D Receptor–Mediated Gene Expression
9.
10.
11.
12. 13.
14. 15. 16.
17. 18. 19.
20.
21.
22.
23.
24.
a nuclear protein and 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 87:9751–9755. MacDonald PN, Haussler CA, Terpening CM, Galligan MA, Reeder MC, Whitfield GK, Haussler MR 1991 Baculovirusmediated expression of the human vitamin D receptor. Functional characterization, vitamin D response element interactions and evidence for a receptor auxiliary factor. J Biol Chem 266:18808–18813. MacDonald PN, Dowd DR, Nakajima S, Galligan MA, Reeder MC, Haussler CA, Ozato K, Haussler MR 1993 Retinoid X receptors stimulate and 9-cis retinoic acid inhibits 1,25-dihydroxyvitamin D3-activated expression of the rat osteocalcin gene. Mol Cell Biol 13:5907–5917. Kliewer SA, Umesono K, Mangelsdorf DJ, Evans RM 1992 Retinoid X receptor interacts with nuclear receptors in retinoic acid, thyroid hormone, and vitamin D3 signalling. Nature 355:446–449. Mangelsdorf DJ, Evans RM 1995 The RXR heterodimers and orphan receptors. Cell 83:841–850. Kurokawa R, Yu VC, Naar A, Kyakumoto S, Han Z, Silverman S, Rosenfeld MG, Glass CK 1993 Differential orientations of the DNA-binding domain and carboxy-terminal dimerization interface regulate binding site selection by nuclear receptor heterodimers. Genes Dev 7:1423–1435. Lemon BD, Freedman LP 1996 Selective effects of ligands on vitamin D3 receptor- and retinoid X receptor-mediated gene activation in vivo. Mol Cell Biol 16:1006–1016. Tasset D, Tora L, Fromental C, Scheer E, Chambon P 1990 Distinct classes of transcriptional activating domains function by different mechanisms. Cell 62:1177–1187. Meyer M-E, Gronemeyer H, Turcotte B, Bocquel M-T, Tasset D, Chambon P 1989 Steroid hormone receptors compete for factors that mediate their enhancer function. Cell 57:433–442. McDonnell DP, Scott RA, Kerner SA, O’Malley BW, Pike JW 1989 Functional domains of the human vitamin D3 receptor regulate osteocalcin gene expression. Mol Endocrinol 3:635–644. Rochel N, Wurtz JM, Mitschler A, Klaholz B, Moras D 2000 The crystal structure of the nuclear receptor for vitamin D bound to its natural ligand. Mol Cell 5:173–179. Nakajima S, Hsieh J-C, MacDonald PN, Galligan MA, Haussler CA, Whitfield GK, Haussler MR 1994 The C-terminal region of the vitamin D receptor is essential to form a complex with a receptor auxiliary factor required for high affinity binding to the vitamin D–responsive element. Mol Endocrinol 8:159–172. Masuyama H, Brownfield CM, St.-Arnaud R, MacDonald PN 1997 Evidence for ligand-dependent intramolecular folding of the AF-2 domain in vitamin D receptor-activated transcription and coactivator interaction. Mol Endocrinol 11:1507–1517. Danielian PS, White R, Lees JA, Parker MG 1992 Identification of a conserved region required for hormone dependent transcriptional activation by steroid hormone receptors. EMBO J 11:1025–1033. Renaud J-P, Rochel N, Ruff M, Vivat V, Chambon P, Gronemeyer H, Moras D 1995 Crystal structure of the RAR-γ ligand-binding domain bound to all-trans retinoic acid. Nature 378:681–689. Feng W, Ribeiro RC, Wagner RL, Nguyen H, Apriletti JW, Fletterick RJ, Baxter JD, Kushner PJ, West BL 1998 Hormone-dependent coactivator binding to a hydrophobic cleft on nuclear receptors. Science 280:1747–1749. Heery DM, Kalkhoven E, Hoare S, Parker MG 1997 A signature motif in transcriptional co-activators mediates binding to nuclear receptors. Nature 387:733–736.
301
25. Nolte, RT, Wisely GB, Westin S, Cobb JE, Lambert MH, Kurokawa R, Rosenfeld MG, Willson TM, Glass CK, Milburn MV 1998 Ligand binding and co-activator assembly of the peroxisome proliferator–activated receptor-gamma. Nature 395:137–143. 26. Darimont BD, Wagner RL, Apriletti JW, Stallcup MR, Kushner PJ, Baxter JD, Fletterick RJ, and Yamamoto KR 1998 Structure and specificity of nuclear receptor–coactivator interactions. Genes Dev 12:3343–3356. 27. McInerney EM, Rose DW, Flynn SE, Westin S, Mullen TM, Krones A, Inostroza J, Torchia J, Nolte RT, Assa-Munt N, Milburn MV, Glass CK, Rosenfeld MG 1998 Determinants of coactivator LXXLL motif specificity in nuclear receptor transcriptional activation. Genes Dev 12:3357–3368. 28. Pathrose P, Barmina O, Chang CY, McDonnell DP, Shevde NK, Pike JW 2002 Inhibition of 1,25-dihydroxyvitamin D3dependent transcription by synthetic LXXLL peptide antagonists that target the activation domains of the vitamin D and retinoid X receptors. J Bone Miner Res 17:2196–2205. 29. Onate SA, Tsai SY, Tsai M-J, O’Malley BW 1995 Sequence and characterization of a coactivator for the steroid hormone receptor superfamily. Science 270:1354–1357. 30. Kraichely DM, Collins JJ III, DeLisle RK, MacDonald PN 1999 The autonomous transactivation domain in helix H3 of the vitamin D receptor is required for transactivation and coactivator interaction. J Biol Chem 274:14352–14358. 31. Gill RK, Bell NH 2000 Steroid receptor co-activator-1 mediates 1,25-dihydroxyvitamin D3–stimulated alkaline phosphatase in human osteosarcoma cells. Calcif Tissue Int 66:370–374. 32. Hong H, Kohli K, Trivedi A, Johnson DL, Stallcup MR 1996 GRIP1, a novel mouse protein that serves as a transcriptional coactivator in yeast for the hormone binding domains of steroid receptors. Proc Natl Acad Sci USA 93:4948–4952. 33. Voegel JJ, Heine MJ, Zechel C, Chambon P, Gronemeyer H 1996 TIF2, a 160 kDa transcriptional mediator for the liganddependent activation function AF-2 of nuclear receptors. EMBO J 15:3667–3675. 34. Li H, Gomes PJ, Chen JD 1997 RAC3, a steroid/nuclear receptor–associated coactivator that is related to SRC-1 and TIF2. Proc Natl Acad Sci USA 94:8479–8484. 35. Anzick SL, Kononen J, Walker RL, Azorsa DO, Tanner MM, Guan XYGS, Kallioniemi OP, Trent JM, Meltzer PS 1997 AIB1, a steroid receptor coactivator amplified in breast and ovarian cancer. Science 277:965–968. 36. Chen H, Lin RJ, Schiltz RL, Chakravarti D, Nash A, Nagy L, Privalsky ML, Nakatani Y, Evans RM 1997 Nuclear receptor coactivator ACTR is a novel histone acetyltransferase and forms a multimeric activation complex with P/CAF and CBP/p300. Cell 90:569–580. 37. Takeshita A, Cardona GR, Koibuchi N, Suen CS, Chin WW 1997 TRAM-1, a novel 160-kDa thyroid hormone receptor activator molecule, exhibits distinct properties from steroid receptor coactivator-1. J Biol Chem 272:27629–27634. 38. Torchia J, Rose DW, Inostroza J, Kamei Y, Westin S, Glass CK, Rosenfeld MG 1997 The transcriptional co-activator p/CIP binds CBP and mediates nuclear-receptor function. Nature 387:677–684. 39. Onate SA, Boonyaratanakornkit V, Spencer TE, Tsai SY, Tsai MJ, Edwards DP, O’Malley BW 1998 The steroid receptor coactivator-1 contains multiple receptor interacting and activation domains that cooperatively enhance the activation function 1 (AF1) and AF2 domains of steroid receptors. J Biol Chem 273:12101–12108. 40. Ding XF, Anderson CM, Ma H, Hong H, Uht RM, Kushner PJ, Stallcup MR 1998 Nuclear receptor-binding sites of coactivators
302
41.
42. 43.
44.
45. 46.
47.
48. 49. 50. 51. 52. 53.
54.
55.
56.
57.
DIANE R. DOWD, AMELIA L. M. SUTTON, CHI ZHANG, AND PAUL N. MACDONALD
glucocorticoid receptor interacting protein 1 (GRIP1) and steroid receptor coactivator 1 (SRC-1): multiple motifs with different binding specificities. Mol Endocrinol 12:302–313. Leers J, Treuter E, Gustafsson JA 1998 Mechanistic principles in NR box–dependent interaction between nuclear hormone receptors and the coactivator TIF2. Mol Cell Biol 18: 6001–6013. Leo C, Li H, Chen JD 2000 Differential mechanisms of nuclear receptor regulation by receptor-associated coactivator 3. J Biol Chem 275:5976–5982. Voegel JJ, Heine MJ, Tini M, Vivat V, Chambon P, Gronemeyer H 1998 The coactivator TIF2 contains three nuclear receptor-binding motifs and mediates transactivation through CBP binding-dependent and -independent pathways. EMBO J 17:507–519. Korzus E, Torchia J, Rose DW, Xu L, Kurokawa R, McInerney EM, Mullen TM, Glass CK, Rosenfeld MG 1998 Transcription factor–specific requirements for coactivators and their acetyltransferase functions. Science 279:703–707. Chen D, Ma H, Hong H, Koh SS, Huang SM, Schurter BT, Aswad DW, Stallcup MR 1999 Regulation of transcription by a protein methyltransferase. Science 284:2174–2177. Utley RT, Ikeda K, Grant PA, Cote J, Steger DJ, Eberharter A, John S, Workman JL 1998 Transcriptional activators direct histone acetyltransferase complexes to nucleosomes. Nature 394:498–502. Spencer TE, Jenster G, Burcin MM, Allis CD, Zhou J, Mizzen CA, McKenna NJ, Onate SA, Tsai SY, Tsai MJ, O’Malley BW 1997 Steroid receptor coactivator-1 is a histone acetyltransferase. Nature 389:194–198. Chen H, Lin RJ, Xie W, Wilpitz D, Evans RM 1999 Regulation of hormone-induced histone hyperacetylation and gene activation via acetylation of an acetylase. Cell 98:675–686. Workman JL, Kingston RE 1998 Alteration of nucleosome structure as a mechanism of transcriptional regulation. Annu Rev Biochem 67:545–579. Ogryzko VV, Schiltz RL, Russanova V, Howard BH, Nakatani Y 1996 The transcriptional coactivators p300 and CBP are histone acetyltransferases. Cell 87:953–959. Bannister AJ, Kouzarides T 1996 The CBP co-activator is a histone acetyltransferase. Nature 384:641–643. Yang XJ, Ogryzko VV, Nishikawa J, Howard BH, Nakatani Y 1996 A p300/CBP-associated factor that competes with the adenoviral oncoprotein E1A. Nature 382:319–324. Kamei Y, Xu L, Heinzel T, Torchia J, Kurokawa R, Gloss B, Lin S-C, Heyman RA, Rose DW, Glass CK, Rosenfeld MG 1996 A CBP integrator complex mediates transcriptional activation and AP-1 inhibition by nuclear receptors. Cell 85: 403–414. Chakravarti D, LaMorte VJ, Nelson MC, Nakajima T, Schulman IG, Juguilon H, Montminy M, Evans RM 1996 Role of CBP/P300 in nuclear receptor signalling. Nature 383:99–103. Xu J, Qiu Y, DeMayo FJ, Tsai SY, Tsai MJ, O’Malley BW 1998 Partial hormone resistance in mice with disruption of the steroid receptor coactivator-1 (SRC-1) gene. Science 279:1922–1925. Gehin M, Mark M, Dennefeld C, Dierich A, Gronemeyer H, Chambon P 2002 The function of TIF2/GRIP1 in mouse reproduction is distinct from those of SRC-1 and p/CIP. Mol Cell Biol 22:5923–5937. Yuan Y, Liao L, Tulis DA, Xu J 2002 Steroid receptor coactivator-3 is required for inhibition of neointima formation by estrogen. Circulation 105:2653–2659.
58. Issa LL, Leong GM, Barry JB, Sutherland RL, Eisman JA 2001 Glucocorticoid receptor-interacting protein-1 and receptorassociated coactivator-3 differentially interact with the vitamin D receptor (VDR) and regulate VDR-retinoid X receptor transcriptional cross-talk. Endocrinology 142:1606–1615. 59. Rachez C, Suldan Z, Ward J, Chang CP, Burakov D, Erdjument-Bromage H, Tempst P, Freedman LP 1998 A novel protein complex that interacts with the vitamin D3 receptor in a ligand-dependent manner and enhances VDR transactivation in a cell-free system. Genes Dev 12: 1787–1800. 60. Rachez C, Lemon BD, Suldan Z, Bromleigh V, Gamble M, Naar AM, Erdjument-Bromage H, Tempst P, Freedman LP 1999 Ligand-dependent transcription activation by nuclear receptors requires the DRIP complex. Nature 398:824–828. 61. Fondell JD, Ge H, Roeder RG 1996 Ligand induction of a transcriptionally active thyroid hormone receptor coactivator complex. Proc Natl Acad Sci USA 93:8329–8333. 62. Jiang YW, Veschambre P, Erdjument-Bromage H, Tempst P, Conaway JW, Conaway RC, Kornberg RD 1998 Mammalian mediator of transcriptional regulation and its possible role as an end-point of signal transduction pathways. Proc Natl Acad Sci USA 95:8538–8543. 63. Naar AM, Beaurang PA, Zhou S, Abraham S, Solomon W, Tjian R 1999 Composite co-activator ARC mediates chromatindirected transcriptional activation. Nature 398:828–832. 64. Rachez C, Freedman LP 2001 Mediator complexes and transcription. Curr Opin Cell Biol 13:274–280. 65. Chiba N, Suldan Z, Freedman LP, Parvin JD 2000 Binding of liganded vitamin D receptor to the vitamin D receptor interacting protein coactivator complex induces interaction with RNA polymerase II holoenzyme. J Biol Chem 275: 10719–10722. 66. Sun X, Zhang Y, Cho H, Rickert P, Lees E, Lane W, Reinberg D 1998 NAT, a human complex containing Srb polypeptides that functions as a negative regulator of activated transcription. Mol Cell 2:213–222. 67. Yudkovsky N, Ranish JA, Hahn S 2000 A transcription reinitiation intermediate that is stabilized by activator. Nature 408:225–229. 68. Burakov D, Crofts LA, Chang CP, Freedman LP 2002 Reciprocal recruitment of DRIP/mediator and p160 coactivator complexes in vivo by estrogen receptor. J Biol Chem 277: 14359–14362. 69. Lorch Y, Beve J, Gustafsson CM, Myers LC, Kornberg RD 2000 Mediator-nucleosome interaction. Mol Cell 6:197–201. 70. Rachez C, Gamble M, Chang CP, Atkins GB, Lazar MA, Freedman LP 2000 The DRIP complex and SRC-1/p160 coactivators share similar nuclear receptor binding determinants but constitute functionally distinct complexes. Mol Cell Biol 20:2718–2726. 71. Cantin GT, Stevens JL, Berk AJ 2003 Activation domainmediator interactions promote transcription preinitiation complex assembly on promoter DNA. Proc Natl Acad Sci USA 100:12003–12008. 72. Barletta F, Freedman LP, Christakos S 2002 Enhancement of VDR-mediated transcription by phosphorylation: correlation with increased interaction between the VDR and DRIP205, a subunit of the VDR-interacting protein coactivator complex. Mol Endocrinol 16:301–314. 73. Yang W, Freedman LP 1999 20-Epi analogues of 1,25-dihydroxyvitamin D3 are highly potent inducers of DRIP coactivator complex binding to the vitamin D3 receptor. J Biol Chem 274:16838–16845.
CHAPTER 16 Comodulators of Vitamin D Receptor–Mediated Gene Expression
74. Ito M, Yuan CX, Okano HJ, Darnell RB, Roeder RG 2000 Involvement of the TRAP220 component of the TRAP/SMCC coactivator complex in embryonic development and thyroid hormone action. Mol Cell 5:683–693. 75. Baudino TA, Kraichely DM, Jefcoat SC Jr, Winchester SK, Partridge NC, MacDonald PN 1998 Isolation and characterization of a novel coactivator protein, NCoA-62, involved in vitamin D-mediated transcription. J Biol Chem 273: 16434–16441. 76. Wieland C, Mann S, Besser H, Saumweber H 1992 The Drosophila nuclear protein Bx42, which is found in many puffs on polytene chromosomes, is highly charged. Chromosoma 101:517–525. 77. Dahl R, Wani B, Hayman MJ 1998 The Ski oncoprotein interacts with Skip, the human homolog of Drosophila Bx42. Oncogene 16:1579–1586. 78. Zhou S, Fujimuro M, Hsieh JJ, Chen L, Miyamoto A, Weinmaster G, Hayward SD 2000 SKIP, a CBF1-associated protein, interacts with the ankyrin repeat domain of NotchIC to facilitate NotchIC function. Mol Cell Biol 20:2400–2410. 79. Leong GM, Subramaniam N, Figueroa J, Flanagan JL, Hayman MJ, Eisman JA, Kouzmenko AP 2001 Ski-interacting protein interacts with Smad proteins to augment transforming growth factor-beta-dependent transcription. J Biol Chem 276:18243–18248. 80. Kim YJ, Noguchi S, Hayashi YK, Tsukahara T, Shimizu T, Arahata K 2001 The product of an oculopharyngeal muscular dystrophy gene, poly(A)-binding protein 2, interacts with SKIP and stimulates muscle-specific gene expression. Hum Mol Genet 10:1129–1139. 81. Prathapam T, Kuhne C, Banks L 2001 The HPV-16 E7 oncoprotein binds Skip and suppresses its transcriptional activity. Oncogene 20:7677–7685. 82. Fukuda S, Yamasaki Y, Iwaki T, Kawasaki H, Akieda S, Fukuchi N, Tahira T, Hayashi K 2002 Characterization of the biological functions of a transcription factor, c-myc intron binding protein 1 (MIBP1). J Biochem (Tokyo) 131:349–357. 83. Prathapam T, Kuhne C, Banks L 2002 Skip interacts with the retinoblastoma tumor suppressor and inhibits its transcriptional repression activity. Nucleic Acids Res 30:5261–5268. 84. Zhang C, Baudino TA, Dowd DR, Tokumaru H, Wang W, MacDonald PN 2001 Ternary complexes and cooperative interplay between NCoA-62/Ski-interacting protein and steroid receptor coactivators in vitamin D receptor-mediated transcription. J Biol Chem 276:40614–40620. 85. MacDonald PN, Baudino TA, Tokumaru H, Dowd DR, Zhang C 2001 Vitamin D receptor and nuclear receptor coactivators: crucial interactions in vitamin D–mediated transcription. Steroids 66:171–176. 86. Zhang C, Dowd DR, Staal A, Gu C, Lian JB, Van Wijnen AJ, Stein GS, MacDonald PN 2003 NCoA62/SKIP is a nuclear matrix-associated coactivator that may couple vitamin D receptor-mediated transcription and RNA splicing. J Biol Chem 278:35325–35336. 87. Kostrouchova M, Housa D, Kostrouch Z, Saudek V, Rall JE 2002 SKIP is an indispensable factor for Caenorhabditis elegans development. Proc Natl Acad Sci USA 99:9254–9259. 88. Orgryzko VV, Schiltz RL, Russanova V, Howard BH, Nakatani Y 1996 The transcriptional coactivators p300 and CBP are histone acetyltransferases. Cell 87:953–959. 89. Barry JB, Leong GM, Church WB, Issa LL, Eisman JA, Gardiner EM 2003 Interactions of SKIP/NCoA-62, TFIIB, and retinoid X receptor with vitamin D receptor helix H10 residues. J Biol Chem 278:8224–8228.
303
90. Auboeuf D, Honig A, Berget SM, O’Malley BW 2002 Coordinate regulation of transcription and splicing by steroid receptor coregulators. Science 298:416–419. 91. Manley JL 2002 Nuclear coupling: RNA processing reaches back to transcription. Nat Struct Biol 9:790–791. 92. Orphanides G, Reinberg D 2002 A unified theory of gene expression. Cell 108:439–451. 93. Proudfoot NJ, Furger A, Dye MJ 2002 Integrating mRNA processing with transcription. Cell 108:501–512. 94. Maniatis T, Reed R 2002 An extensive network of coupling among gene expression machines. Nature 416:499–506. 95. Monsalve M, Wu Z, Adelmant G, Puigserver P, Fan M, Spiegelman BM 2000 Direct coupling of transcription and mRNA processing through the thermogenic coactivator PGC-1. Mol Cell 6:307–316. 96. Makarov EM, Makarova OV, Urlaub H, Gentzel M, Will CL, Wilm M, Luhrmann R 2002 Small nuclear ribonucleoprotein remodeling during catalytic activation of the spliceosome. Science 298:2205–2208. 97. Mintz PJ, Patterson SD, Neuwald AF, Spahr CS, Spector DL 1999 Purification and biochemical characterization of interchromatin granule clusters. EMBO J 18:4308–4320. 98. Neubauer G, King A, Rappsilber J, Calvio C, Watson M, Ajuh P, Sleeman J, Lamond A, Mann M 1998 Mass spectrometry and EST-database searching allows characterization of the multi-protein spliceosome complex. Nat Genet 20:46–50. 99. Blanco JCG, Wang I-M, Tsai SY, Tsai M-J, O’Malley BW, Jurutka PW, Haussler MR, Ozato K 1995 Transcription factor TFIIB and the vitamin D receptor cooperatively activate ligand-dependent transcription. Proc Natl Acad Sci USA 92:1535–1539. 100. MacDonald PN, Sherman DR, Dowd DR, Jefcoat SC, DeLisle RK 1995 The vitamin D receptor interacts with general transcription factor IIB. J Biol Chem 270:4748–4752. 101. Jurutka PW, Remus LS, Whitfield GK, Thompson PD, Hsieh JC, Zitzer H, Tavakkoli P, Galligan MA, Dang HT, Haussler CA, Haussler MR 2000 The polymorphic N terminus in human vitamin D receptor isoforms influences transcriptional activity by modulating interaction with transcription factor IIB. Mol Endocrinol 14:401–420. 102. Lemon BD, Fondell JD, Freedman LP 1997 Retinoid X receptor:vitamin D3 receptor heterodimers promote stable preinitiation complex formation and direct 1,25-dihydroxyvitamin D3-dependent cell-free transcription. Mol Cell Biol 17:1923–1937. 103. Mengus G, May M, Carre L, Chambon P, Davidson I 1997 Human TAF(II)135 potentiates transcriptional activation by the AF-2s of the retinoic acid, vitamin D3, and thyroid hormone receptors in mammalian cells. Genes Dev 11: 1381–1395. 104. Yanagisawa J, Yanagi Y, Masuhiro Y, Suzawa M, Watanabe M, Kashiwagi K, Toriyabe T, Kawabata M, Miyazono K, Kato S 1999 Convergence of transforming growth factor-beta and vitamin D signaling pathways on SMAD transcriptional coactivators. Science 283:1317–1321. 105. Tolon RM, Castillo AI, Jimenez-Lara AM, Aranda A 2000 Association with Ets-1 causes ligand- and AF2-independent activation of nuclear receptors. Mol Cell Biol 20:8793–8802. 106. Rosenfeld MG, Glass CK 2001 Coregulator codes of transcriptional regulation by nuclear receptors. J Biol Chem 276:36865–36868. 107. Chen JD, Evans RM 1995 A transcriptional co-repressor that interacts with nuclear hormone receptors. Nature 377: 454–457.
304
DIANE R. DOWD, AMELIA L. M. SUTTON, CHI ZHANG, AND PAUL N. MACDONALD
108. Horlein AJ, Naar AM, Heinzel T, Torchia J, Gloss B, Kurokawa R, Ryan A, Kamei Y, Soderstrom M, Glass CK, Rosenfeld MG 1995 Ligand-independent repression by the thyroid hormone receptor mediated by a nuclear receptor co-repressor. Nature 377:397–404. 109. Hu X, Lazar MA 1999 The CoRNR motif controls the recruitment of corepressors by nuclear hormone receptors. Nature 402:93–96. 110. Nagy L, Kao HY, Love JD, Li C, Banayo E, Gooch JT, Krishna V, Chatterjee K, Evans RM, Schwabe JW 1999 Mechanism of corepressor binding and release from nuclear hormone receptors. Genes Dev 13:3209–3216. 111. Perissi V, Staszewski LM, McInerney EM, Kurokawa R, Krones A, Rose DW, Lambert MH, Milburn MV, Glass CK, Rosenfeld MG 1999 Molecular determinants of nuclear receptor-corepressor interaction. Genes Dev 13:3198–3208. 112. Tagami T, Lutz WH, Kumar R, Jameson JL 1998 The interaction of the vitamin D receptor with nuclear receptor corepressors and coactivators. Biochem Biophys Res Commun 253:358–363. 113. Smith CL, Nawaz Z, O’Malley BW 1997 Coactivator and corepressor regulation of the agonist/antagonist activity of the mixed antiestrogen, 4-hydroxytamoxifen. Mol Endocrinol 11:657–666. 114. Jackson TA, Richer JK, Bain DL, Takimoto GS, Tung L, Horwitz KB 1997 The partial agonist activity of antagonistoccupied steroid receptors is controlled by a novel hinge domain-binding coactivator L7/SPA and the corepressors N-CoR or SMRT. Mol Endocrinol 11:693–705. 115. Zhang X, Jeyakumar M, Petukhov S, Bagchi MK 1998 A nuclear receptor corepressor modulates transcriptional activity of antagonist-occupied steroid hormone receptor. Mol Endocrinol 12:513–524. 116. Nagy L, Kao HY, Chakravarti D, Lin RJ, Hassig CA, Ayer DE, Schreiber SL, Evans RM 1997 Nuclear receptor repression mediated by a complex containing SMRT, mSin3A, and histone deacetylase. Cell 89:373–380. 117. Heinzel T, Lavinsky RM, Mullen T-M, Soderstrom M, Laherty CD, Torchia J, Yang WM, Brard G, Ngo SD, Davie JR, Seto E, Eisenman RN, Rose DW, Glass CK, Rosenfeld MG 1997 A complex containing N-CoR, mSin3 and histone deacetylase mediates transcriptional repression. Nature 387:43–48.
118. Alland L, Muhle R, Hou H, Potes J, Chin L, Schreiber-Agus N, Depinho RA 1997 Role for N-CoR and histone deacetylase in Sin3-mediated transcriptional repression. Nature 387: 49–55. 119. Malloy PJ, Pike JW, Feldman D 1999 The vitamin D receptor and the syndrome of hereditary 1,25-dihydroxyvitamin D– resistant rickets. Endocr Rev 20:156–188. 120. Miller J, Djabali K, Chen T, Liu Y, Ioffreda M, Lyle S, Christiano AM, Holick M, Cotsarelis G 2001 Atrichia caused by mutations in the vitamin D receptor gene is a phenocopy of generalized atrichia caused by mutations in the hairless gene. J Invest Dermatol 117:612–617. 121. Potter GB, Beaudoin GM 3rd, DeRenzo CL, Zarach JM, Chen SH, Thompson CC 2001 The hairless gene mutated in congenital hair loss disorders encodes a novel nuclear receptor corepressor. Genes Dev 15:2687–2701. 122. Moraitis AN, Giguere V, Thompson CC 2002 Novel mechanism of nuclear receptor corepressor interaction dictated by activation function 2 helix determinants. Mol Cell Biol 22:6831–6841. 123. Hsieh JC, Sisk JM, Jurutka PW, Haussler CA, Slater SA, Haussler MR, Thompson CC 2003 Physical and functional interaction between the vitamin D receptor and hairless corepressor, two proteins required for hair cycling. J Biol Chem 278:38665–38674. 124. Panteleyev AA, Paus R, Ahmad W, Sundberg JP, Christiano AM 1998 Molecular and functional aspects of the hairless (hr) gene in laboratory rodents and humans. Exp Dermatol 7:249–267. 125. Dressel U, Thormeyer D, Altincicek B, Paululat A, Eggert M, Schneider S, Tenbaum SP, Renkawitz R, Baniahmad A 1999 Alien, a highly conserved protein with characteristics of a corepressor for members of the nuclear hormone receptor superfamily. Mol Cell Biol 19:3383–3394. 126. Polly P, Herdick M, Moehren U, Baniahmad A, Heinzel T, Carlberg C 2000 VDR-Alien: a novel, DNA-selective vitamin D3 receptor-corepressor partnership. FASEB J 14:1455–1463. 127. Shang Y, Hu X, DiRenzo J, Lazar MA, Brown M 2000 Cofactor dynamics and sufficiency in estrogen receptorregulated transcription. Cell 103:843–852. 128. Sharma D, Fondell JD 2002 Ordered recruitment of histone acetyltransferases and the TRAP/Mediator complex to thyroid hormone-responsive promoters in vivo. Proc Natl Acad Sci USA 99:7934–7939.
CHAPTER 17
Promoter Targeting of Vitamin D Receptor through a Chromatin Remodeling Complex SHIGEAKI KATO,* RYOJI FUJIKI, AND HIROCHIKA KITAGAWA* Institute of Molecular and Cellular Biosciences, University of Tokyo, Japan, and *SOREST, Japan Science and Technology, Saitama, Japan
I. Introduction II. Chromatin Remodeling Is a Prerequisite for Transcriptional Controls by the Vitamin D Receptor III. Purification and Identification of Williams Syndrome Transcription Factor as a Vitamin D Receptor Interactant IV. Purification and Identification of a Novel Williams Syndrome Transcription Factor Complex Associating with Vitamin D Receptor V. WINAC Is a Novel Multifunctional ATP-Dependent Chromatin Remodeling Complex That Rearranges
a Nucleosome Array around a Vitamin D Responsive Element in Vitro VI. Williams Syndrome Transcription Factor Coactivated the Ligand-Induced Transactivation Function of Vitamin D Receptor VII. Molecular Mechanism of VDR Promoter Targeting of Vitamin D Receptor by WINAC and Cooperative WINAC Function with the Co-regulator Complexes References
I. INTRODUCTION
These co-regulatory complexes contain enzymes that modify histones through acetylation, which results in rearrangement of nucleosome arrays [9]. However, prior to such necleosome array reorganization by co-regulator complexes, chromatin remodeling is believed to be coupled with the promoter targeting of VDR, like all of the DNA binding transcription factors (Fig. 1).
The calciotropic hormone 1,25(OH)2D3, the active form of vitamin D3, regulates calcium homeostasis as well as cellular proliferation and differentiation [1]. Most biological actions of 1,25(OH)2D3 are believed to be mediated through transcriptional controls of a particular set of target genes by the vitamin D receptor (VDR) [2,3]. VDR is a member of the nuclear receptor (NR) gene superfamily, which acts as a ligand-inducible transcriptional factor [4]. Like other members of the nuclear receptor superfamily, VDR structure is divided into several functional domains, with the most highly conserved DNA-binding domain (C) located centrally and the less highly conserved ligand-binding domain (E) located at the C-terminal end [5]. Most nuclear receptors harbor both an N-terminal activation function 1 (AF-1) and a C-terminal AF-2 domain [6]. The VDR, however, appears to lack the significant N-terminal AF-1 function because of its relatively short A/B domain. In the promoters of target genes that are controlled by liganded VDR, VDR/RXR heterodimer recognizes and directly binds to cognate vitamin D responsive elements (VDREs) [6]. Liganded VDR bound upon the VDREs induces the recruitment of a number of histone acetyltransferase (HAT) and non-HAT coactivators and coactivator complexes to activate transcription [7,8]. VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
II. CHROMATIN REMODELING IS A PREREQUISITE FOR TRANSCRIPTIONAL CONTROLS BY THE VITAMIN D RECEPTOR At transcriptional initiation sites in promoters, rearrangement of the nucleosome array is thought to be indispensable for controlling transcription of sequencespecific regulators such as VDR through chromatin remodeling and histone modifications [9,10]. These complexes (Fig. 2) modify the chromatin configuration in a highly regulated manner. One class contains several discrete subfamilies of transcription coregulatory complexes that harbor either HAT or histone deacetylase (HDAC) activities to covalently modify histones through acetylation [11]. In NR ligand-induced transactivation processes, the complexes containing HDAC first act to corepress transactivation of unliganded NRs. Upon ligand binding, two HAT complexes, p160/CBP and TRRAP/GCN5, are Copyright © 2005, Elsevier, Inc. All rights reserved.
306
SHIGEAKI KATO, RYOJI FUJIKI, AND HIROCHIKA KITAGAWA
Co-activator complexes
Co-repressor complexes
HAT complexes
HDAC complexes HDAC1/2
PCAF complex
mSin3
p300
NCoR/ SMRT
non-HAT co-activators
p160 family protein
DRIP/TRAP complexes Chromatin remodeling
Chromatin remodeling
VDR RXR
VDRE
FIGURE 1 VDRs coordinate the activities of chromatin remodeling complexes to facilitate. Because the promoter is assembled into a nucleosomal structure that is transcriptonally inactive the chromatin environment limits the accessibility of DNA-binding transcription factor and their co-regulators to promoter. To conquer this difficulty, an interaction of VDR with unknown chromatin remodeling complexes reorganizes nucleosomal arrays at VDR targeting promoters, which results in stabilized binding of transcription co-regulator.
A SWI/SNF family
ISWI family
ATPase
Bromo
hSWI/SNF complex Brg1/ BAF250 hBrm BAF60a/b/c BAF155 BAF170 BAF47 Ini1 BAF50
B
ATPase
SANT
hSpt3 hPCAF TAF30 hAda3 PAF65β
chromo
ATPase
hNuRD complex
hAcf1
MTA2 HDAC1 MBD3
hISWI
RbA MTA1 p46 Mi-2 RbA HDAC2 p48
HDAC complexes CBP/p160 complex
TAF15 PAF65α hAda2 TRRAP
PHD
hACF complex
HAT complexes hPCAF complex
Mi-2 family
CBP/ p300 p160 family protein
mSin3 complex HDAC1SAP18 RbA p46 mSin3 SAP30 HDAC2 RbA p48
TAF31
FIGURE 2 Chromatin remodeling complexes. (A) ATP-dependent chromatin remodeling complexes. These types of complexes can be divided into three major classes based on catalytic ATPase subunit: Brg1/hBrm, hISWI, and Mi-2. Shown here are some representative members of the SWI/SNF, ISWI, and Mi-2 families of human ATP-dependent chromatin remodeling complexes. (B) Some representative members of the histone modifying complexes. Purple, ATPase; red, histone acetyltransferase; green, histone deacetylase. (See color plate).
CHAPTER 17 Promoter Targeting of Vitamin D Receptor through a Chromatin Remodeling Complex
recruited with dissociation of the HDAC complexes, and coactivate the NR function [12–14]. Another class of complexes contains chromatin remodeling complexes that use ATP hydrolysis to rearrange nucleosomal arrays in a noncovalent manner, thereby making chromosomal DNA accessible for sequence-specific regulators such as the VDR [11]. Besides of transcriptional controls, these ATP-dependent chromatin remodeling complexes are supposed to act on DNA repair and DNA replication. These complexes are further classified into subfamilies based on the major catalytic components, the ATPases (Brg1/hBrm, hISWI, and Mi-2) [15,16]. Recently, the ligand-induced transactivation function of VDR in vitro has been shown to require a SWI2/SNF2-type chromatin remodeling complex containing pBAF180 [17].
III. PURIFICATION AND IDENTIFICATION OF WILLIAMS SYNDROME TRANSCRIPTION FACTOR AS A VITAMIN D RECEPTOR INTERACTANT To search for novel co-regulatory complexes that interact with VDR, HeLa cell nuclear extracts were incubated with a chimeric VDR-hinge-ligand binding domain region (VDR-DEF) fused to glutathione-Stransferase (GST) in the presence or absence of 1,25(OH)2D3 (Fig. 3A). Proteins that interacted with VDR-DEF were separated by SDS-PAGE and silver stained (Fig. 3B). Mass spectrometry coupled with the apparent molecular weights of the different proteins
GST
P11
kDa
HNE
HeLa nuclear extract (HNE)
VDR(D3+)
B
VDR(D3−)
A
TRRAP/PAF400 P11 Phosphocellulose column DRIP240/TRAP230 200 GST
DRIP205/TRAP220 WSTF TRAP130
116 GST–VDR(D3−)
GST–VDR(D3+)
WINAC
WINAC DRIP/TRAPs
307
DRIP100/TRAP100 97 DRIP70 66
WSTF TRAP220 TIF2 Brg1 hSNF2h
FIGURE 3 Purification and identification of human proteins interacting with 1,25(OH)2D3–unbound and bound VDR. (A) Purification scheme for VDR interacting proteins. The eluted fraction from a P11 phosphocellulose column was incubated with immobilized GST-VDR(DEF) in the absence or presence of 1,25(OH)2D3 (D3, 10−6 M). (B) Identification of ligand-independent and -dependent VDR interacting proteins. In the upper panel, fractions were subjected to SDS-PAGE, followed by silver staining. Total HeLa S3 nuclear extract [HNE] (lane 1), a fraction eluted from the P11 column [P11] (lane 2), fractions from GST [GST] (lane 3), and unliganded and liganded GST-VDR(DEF) columns [VDR(−);VDR(+)] (lanes 4 and 5) were examined by mass spectrometry. Identified proteins are indicated at the right side of the panel. The lower panel shows Western blot analysis using specific antibodies shown in the panel.
308
SHIGEAKI KATO, RYOJI FUJIKI, AND HIROCHIKA KITAGAWA
associated with the ligand activated VDR-DEF led to the identification of several known components of the DRIP/TRAP/SMCC complex (Fig. 3B) that were in agreement with previous observations [12,14]. One ligand-independent VDR-specific interactant turned out as the Williams syndrome transcription factor (WSTF)/ WBSCR9/BAZ1B [18,19]. WSTF is a candidate gene whose product is potentially responsible for diverse WS disorders [18,20]. WSTF is highly homologous to hACF1 as one of the Bromodomain-adjacent-to-Zn finger (BAZ) family proteins [19], and hACF1 is a partner of hSNF2h (a Drosophila ISWI homolog), which forms well-characterized ISWI-based chromatin remodeling complexes (see Fig. 2) [21]. These facts raise the possibility that a WSTF-containing complex is an ATPdependent chromatin complex.
than 670 kDa contained WSTF, indicating that WSTF forms a stable nuclear complex. With mass fingerprinting, we identified all the components of the purified complex containing WSTF (Fig. 4C) and designated this complex as WINAC (WSTF including nucleosome assembly complex) [22]. WINAC consists of at least 13 components, but unexpectedly, does not contain hSNF2h as an ATPase subunit (Fig. 4C). Rather, the SWI/SNF type ATPases (Brg1 and hBrm) and several BAF components are shared with the SWI2/SNF2-based complexes [9]. Interestingly, WINAC appears to harbor three additional components (TopoIIb, FACTp140, and CAF-1p150) [23–25], which have not been observed in any of the known ATP-dependent chromatin remodeling complexes, but are supposed to be involved in other nuclear processes such as DNA repair and replication.
IV. PURIFICATION AND IDENTIFICATION OF A NOVEL WILLIAMS SYNDROME TRANSCRIPTION FACTOR COMPLEX ASSOCIATING WITH VITAMIN D RECEPTOR
V. WINAC IS A NOVEL MULTIFUNCTIONAL ATP-DEPENDENT CHROMATIN REMODELING COMPLEX THAT REARRANGES A NUCLEOSOME ARRAY AROUND A VITAMIN D RESPONSIVE ELEMENT IN VITRO
WSTF-containing complexes were purified by sequential columns (Fig. 4A) and fractionated on glycerol density gradients (Fig. 4B, upper panel). The fractionated complex with a molecular weight of greater
A
B Nuclear extract from f–WSTF cells GST
An ATP-dependent chromatin assembly reaction is clearly induced by WINAC as assessed by a standard micrococcal nuclease assay (Fig. 5A), indicating that
C kDa 200 116
BAF250 Brg1/hBrm TopoIIb WSTF
97 GST–VDR(D3−)
Glycerol density gradient
66
BAF170 BAF155 FACT p140 CAF–1 p150
WSTF Brg1
aFLAG M2 resin
WINAC
VDR GST–VDR BAF60a BAF57 BAF53 Ini1 WINAC
FIGURE 4 Purification and identification of a human WSTF-containing multiprotein complex “WINAC”. (A) Purification scheme of WINAC from MCF7 stable transformants. (B) Fractionation of purified complexes on glycerol density gradient. In the lower panel, Western blot analysis of each fraction using specific antibodies is shown. (C) The purified complex was subjected to SDS-PAGE, followed by silver staining and identified by mass spectrometry (indicated in the left of the panel).
309
CHAPTER 17 Promoter Targeting of Vitamin D Receptor through a Chromatin Remodeling Complex
A
ATP WINAC
B
+ − + + − +
Proximal probe
Distal probe
C
Proximal Probe
GAL–VDR −
+
−
+
−
+
−
+
GAL–ERα
+
+
−
−
−
−
+
+
−
−
+
+
GAL–PPAR −
−
+
+
−
+
−
+
WINAC
WINAC
FIGURE 5 WINAC as an ATP-dependent chromatin remodeling complex. (A) Chromatin reconstitution activity of WINAC. The reacted samples were subjected to partial micrococcal nuclease digestion. The molecular mass marker is the 200-bp ladder. (B) Chromatin disruption by WINAC is specifically VDR dependent. Oligonucleotide probe corresponds to either a sequence between the GAL4 sites and the RNA start site (Proximal Probe) or 900 bp upstream of the start site (Distal Probe). (C) WINAC is unable to affect the chromatin architecture through other nuclear receptors (ER and PPAR).
Brg1/hBrm in WINAC serves as an ATPase for this ATP-dependent chromatin remodeling process. WINAC appeared to have a chromatin assembly activity (data not shown) similar to RSF [22,26]. We then examined the ability of WINAC to disrupt nucleosome arrays through VDR bound VDRE since the known ATP-dependent chromatin remodeling complexes are able to recognize a nucleosomal array around binding sites of a sequence specific regulator [27]. Disruption of the nucleosome arrays in the vicinity of the GAL4 binding site was induced only when both VDR and WINAC were present (Fig. 5B), while other regions appeared unaffected in the nucleosome array. Reflecting VDR-specific nucleosome disruption by WINAC (among several tested receptors) (Fig. 5C), WINAC potentiated ligand-induced transactivation in vitro only for VDR, but not for either ERα or PPARγ (data not shown)[22].
VI. WILLIAMS SYNDROME TRANSCRIPTION FACTOR COACTIVATED THE LIGAND-INDUCED TRANSACTIVATION FUNCTION OF VITAMIN D RECEPTOR In a reporter assay, 1,25(OH)2D3 (10−9 M) was effective in inducing VDR AF-2 transactivation function, and WSTF coactivated this ligand-induced AF-2
function of VDR, but not ERα (Fig. 6A). Coactivation by either Brg1 or hBrm was expectedly detected in VDR and ERα (Fig. 6A) [28,29]; however, such coactivator-like activity of WSTF was found only for VDR, but not for ERα, even in the presence of Brg1/hBrm (Fig. 6A). ChIP analysis of the positive VDRE in the human 25-hydroxyvitamin D3 24-hydroxylase gene promoter revealed that associations of VDR and the WINAC components with the promoter did not require ligand binding. In contrast, ligand binding to VDR appeared to be a prerequiste for occupancy of known coactivators TRAP220 and TIF2 on the promoter (Fig. 6B), together with ligand-induced histone H4 acetylation (data not shown) [22] though ligand-induced TRAP220 and TIF2 occupancy was cyclic (data not shown), as already described [28]. Such ligand-dependent and -independent recruitment of factors to the promoter and histone modification were robustly attenuated by WSTF-RNAi expression (Fig. 6B). As the VDR/RXR heterodimer also represses transcription in a ligand-dependent manner through a negative VDRE (nVDRE), the action of WSTF in ligandinduced transrepression was examined in a naturally occurring nVDRE in the human 25-hydroxyvitamin D3 1α-hydroxylase [1 (OH)ase gene promoter [30]. ChIP analysis revealed that VDR and WINAC appear to bind to the nVDRE in a ligand-independent manner, while cyclic ligand-induced recruitments of N-CoR and HDAC2 were
310
SHIGEAKI KATO, RYOJI FUJIKI, AND HIROCHIKA KITAGAWA
B Transcription activity (fold)
A GAL4-VDR-DEF
WSTF
− − − − − +
− − − − +
− − − − +
− − − − − − + − − − − − − + + +
+ − + − − +
− + + − − +
− − − + − +
− + − + − +
− − − − + +
− + − − + +
− − + − − + − − − − −
WSTF-RNAi − − − + − − + − + − + Brg1
− − − − + + + − − − −
TRAP220
− − − − − − − + + − −
TIF2 E2
− − − − − − − − − + + − + + + + + + + + + +
Input
Input
VDR
VDR
fWSTF
fWSTF
Brg1
Brg1
TRAP220
NCoR
TIF2
HDAC2
hSNF2H
hSNF2H
D3
−
+
−
+
+
D3
−
+
−
+
+
WSTF-RNAi
−
−
−
−
+
WSTF-RNAi
−
−
−
−
+
D GAL4-ERα-DEF
1α(OH)ase gene promoter negative distal VDRE
Transcription activity (fold)
− − − − − −
Transcription activity (fold)
WSTF WSTF-RNAi Brg1 TRAP220 TIF2 D3
C 24(OH)ase gene promoter positive distal VDRE
WSTF WSTF-RNAi Brg1 NCoR HDAC2 D3
− − − − − −
− − − − − +
− − − − +
− − − − +
− − − − − − + − − − − − − + + +
+ − + − − +
− + + − − +
− − − + − +
− + − + − +
− − − − + +
− + − − + +
FIGURE 6
Ligand-dependent promoter targeting of co-regulators through WINAC-VDR association. (A) VDR-specific facilitation of coactivator accessibility by WINAC. MCF7 cells were transfected with the expression vectors of a luciferase reporter plasmid containing the GAL4 upstream activation sequence (UAS) [17mer(×2)] driven by the β-globin promoter (0.5 µg). PML-CMV (2 ng), GAL4-DBD-VDR-DEF (0.2 µg), GAL4-DBD-ERa-DEF (0.2 µg), pDNA3-FLAG-WSTF (+; 0.1 µg: ++; 0.3 µg), pSV-Brg1 (0.2 µg), pSV-hBrm (0.2 µg), pcDNA3-TRAP220 (0.3 µg), pcDNA3-TIF2 (0.3 µg), siRNA (+; 0.1 µg: ++; 0.2 µg) of WSTF-RNAi, or control RNAi or their combinations were transfected as indicated in the panels in the absence or presence of ligand (10−9 M). Bars in each graph show the fold change in luciferase activity relative to the activity of the receptor transactivation in the presence of ligand. (B, C) ChIP analysis on the 24(OH)ase promoter and 1α(OH)ase promoter of WSTF stable transformants. Soluble chromatin was prepared from WSTF stable transformants treated with 1,25(OH)2D3 (10−9 M) for 45 min and immunoprecipitated with indicated antibodies. (D) The coregulator-like actions of WSTF on the naturally occurring negative vitamin D response elements. MCF7 cells were transfected with the expression vectors of either the luciferase reporter plasmid containing a human 1α(OH)ase promoter containing a negative VDRE and the factors shown in (A) or together with pcDNA3-N-CoR (0.3 µg), pcDNA3-HDAC2 (0.3 µg).
observed (Fig. 6C). Ligand-dependent repression was dependent on the expression levels of WSTF (Fig. 6D). Thus, it is likely that WINAC association with VDR facilitates targeting of a putative co-repressor complex to the nVDRE. Thus, these findings indicate that WINAC rearranges the nucleosome array around the positive and negative VDREs, thereby facilitating accessibility of the coregulatory complexes accessible to VDR for further transcription controls (Fig. 7).
VII. MOLECULAR MECHANISM OF VDR PROMOTER TARGETING OF VITAMIN D RECEPTOR BY WINAC, AND COOPERATIVE WINAC FUNCTION WITH THE CO-REGULATOR COMPLEXES Both WSTF and the ATPase subunits coactivated the ligand-induced VDR transactivation, a finding similar
311
CHAPTER 17 Promoter Targeting of Vitamin D Receptor through a Chromatin Remodeling Complex
WINAC WSTF
VDR RXR
VDR RXR
HDAC complexes HAT complexes
HDAC1/2
WINAC
mSin3
PCAF complex
NCoR/ WSTF SMRT VDR RXR
Gene activation
WSTF
VDR RXR
Target gene
gene represseion
WINAC
p300
p160 family protein
1α,25(OH)2D3
Target gene
FIGURE 7
Model of WINAC action in chromatin reorganization. WINAC mediates the recruitment of unliganded VDR to VDR target sites in promoters, whereas subsequent binding of co-regulators requires ligand binding. This recruitment order exemplifies that an interaction of a sequence-specific regulator with a chromatin remodeling complex can organize nucleosomal arrays at specific local sites in order to make promoters accessible for co-regulators.
to the reported coactivator-like actions of the SWI2/SNF2-type complex components, ATPases and BAF57, for the ligand-induced ERα transactivation [28,29]. Notably, VDR coactivation by the liganddependent NR coactivators (TIF2 and TRAP220) was abrogated by WSTF-RNAi expression. However, such WSTF coactivator-like actions were not observed for ERα and the other receptors tested (data not shown), nor were they reduced by WSTF-RNAi expression, supporting the observed direct and selective interaction of WSTF with VDR, though it remains unclear whether promoter targeting of VDR requires only WINAC or the other chromatin remodeling complexes. More remarkably, we found that WSTF could potentiate the ligand-induced transrepression of VDR on the 1α(OH)ase negative VDRE (Fig. 6D), where ablation of endogenous WSTF by RNAi expression led to a significant reduction in ligand-induced co-repressor recruitment (Fig. 6C). Thus, ligand-independent association of WINAC and VDR in VDR target promoters appears to facilitate local nucleosomal array accessibility for ligand-dependent co-regulators, following histone tail modifications by the recruited co-regulator complexes [11] (Fig. 7). From our ChIP analysis, VDR appears to be selectively targeted through WINAC to the promoters in a ligand-independent fashion or following recruitment of co-regulator complexes. This ligand-independent association of VDR with the target promoter is surprisingly distinct from the ligand-induced promoter targeting of many other steroid receptors [28,29]. It is possible that other nonsteroidal receptors such as
RAR, RXR, and TR may associate with their target promoters in the absence of hormone, and as-yetunidentified chromatin remodeling complexes may assist the promoter targeting. As HAT and HDAC complexes appear not to associate with unliganded VDR bound upon the tested promoters, WINAC targeting to the VDR target promoters does not appear to require specific histone tail modifications by other co-regulators. Thus, it is possible that WINAC binding to promoters facilitates VDR recognition and specific binding to VDREs, through nucleosomal mobilization by WINAC, presumably cooperating with the other chromatin complexes [17]. Alternatively, once VDR binds to VDREs during nonspecific chromatin remodeling, WINAC might be recruited to VDR on specific promoters to engage in local nucleosome reorganization. The latter possibility coincides well with a recent report regarding a sequence-specific regulator, SATB1 [16]. As a result of WINAC recruitment, the local chromatin structure near VDREs may transit into an active chromosomal state that appears competent for receipt of both coactivator and co-repressor complexes dependent on the VDRE sequences and the tertiary positions of DNA-bound VDR. This scheme may not be considered, however, with VDREs composed of unrelated DNA sequences with the canonical VDREs. Regardless, our hypothesis is not consistent with current understanding that the chromatin remodeling complexes are recruited only after acetylation/ deacetylation of histone tails by the coregulatory complexes [11]. However, the orders of the complex targetings are supposed to be dependent on regulator type
312
SHIGEAKI KATO, RYOJI FUJIKI, AND HIROCHIKA KITAGAWA
and promoter context [31,32]. This point should be addressed with VDR together with WINAC.
activator interactions with the ATM-related Tra1 subunit. Science 292:2333–2337. Fyodorov DV, Kadonaga JT 2001 The many faces of chromatin remodeling: SWItching beyond transcription. Cell 106:523–525. Yasui D, Miyano M, Cai S, Varga-Weisz P, Kohwi-Shigematsu T 2002 SATB1 targets chromatin remodeling to regulate genes over long distances. Nature 419:641–645. Lemon B, Inouye C, King DS, Tjian R 2001 Selectivity of chromatin-remodeling cofactors for ligand-activated transcription. Nature 414:924–928. Lu X, Meng X, Morris CA, Keating MT 1998 A novel human gene, WSTF, is deleted in Williams syndrome. Genomics 54:241–249. Jones MH, Hamana N, Nezu J, Shimane M 2000 A novel family of bromodomain genes. Genomics 63:40–45. Peoples RJ, Cisco MJ, Kaplan P, Francke U 1998 Identification of the WBSCR9 gene, encoding a novel transcriptional regulator, in the Williams–Beuren syndrome deletion at 7q11.23. Cytogenet Cell Genet 82:238–246. Poot RA, Dellaire G, Hulsmann BB, Grimaldi MA, Corona DF, Becker PB, Bickmore WA, Varga-Weisz PD 2000 HuCHRAC, a human ISWI chromatin remodelling complex contains hACF1 and two novel histone-fold proteins. EMBO J 19: 3377–3387. Kitagawa H, Fujiki R, Yoshimura K, Mezaki Y, Uematus Y, Matsui D, Ogawa S, Unno K, Okubo M, Tokita A, Nakagawa T, Ito T, Ishimi Y, Nagasawa H, Matsumoto T, Yanagisawa J, Kato S 2003 The chromatin remodeling complex WINAC targets a nuclear receptor to promoters and is impaired in Williams Syndrome. Cell 113:1–13. Smith S, Stillman B 1989 Purification and characterization of CAF-I, a human cell factor required for chromatin assembly during DNA replication in vitro. Cell 58:15–25. Varga-Weisz PD, Wilm M, Bonte E, Dumas K, Mann M, Becker PB 1997 Chromatin-remodelling factor CHRAC contains the ATPases ISWI and topoisomerase II. Nature 388:598–602. LeRoy G, Orphanides G, Lane WS, Reinberg D 1998 Requirement of RSF and FACT for transcription of chromatin templates in vitro. Science 282:1900–1904. Loyola A, LeRoy G, Wang YH, Reinberg D 2001 Reconstitution of recombinant chromatin establishes a requirement for histonetail modifications during chromatin assembly and transcription. Genes Dev 15:2837–2851. Ito T, Bulger M, Pazin MJ, Kobayashi R, Kadonaga JT 1997 ACF, an ISWI-containing and ATP-utilizing chromatin assembly and remodeling factor. Cell 90:145–155. Shang Y, Hu X, DiRenzo J, Lazar MA, Brown M 2000 Cofactor dynamics and sufficiency in estrogen receptor-regulated transcription. Cell 103:843–852. Belandia B, Orford RL, Hurst HC, Parker MG 2002 Targeting of SWI/SNF chromatin remodelling complexes to estrogenresponsive genes. EMBO J 21:4094–4103. Murayama A, Takeyama K, Kitanaka S, Kodera Y, Hosoya T, Kato S 1998 The promoter of the human 25-hydroxyvitamin D3 1α-hydroxylase gene confers positive and negative responsiveness to PTH, calcitonin, and 1α,25(OH)2D3. Biochem Biophys Res Commun 249:11–16. Lomvardas S, Thanos D 2002 Modifying gene expression programs by altering core promoter chromatin architecture. Cell 110:261–271. Soutoglou E, Talianidis I 2002 Coordination of PIC assembly and chromatin remodeling during differentiation-induced gene activation. Science 295:1901–1904.
Acknowledgment We sincerely thank all collaborators for the WINAC projects, and the laboratory members for helpful discussions and technical support. We are grateful to Miss Y. Nagasawa for preparation of the manuscript.
References
15. 16. 17. 18. 19. 20.
1. Deluca HF 1986 The metabolism and functions of vitamin D. Adv Exp Med Biol 196:361–375. 2. Bouillon R, Okamura WH, Norman AW 1995 Structurefunction relationships in the vitamin D endocrine system. Endocr Rev 16:200–257. 3. Walters MR 1992 Newly identified actions of the vitamin D endocrine system. Endocr Rev 13:719–764. 4. Mangelsdorf DJ, Thummel C, Beato M, Herrlich P, Schutz G, Umesono K, Blumberg B, Kastner P, Mark M, Chambon P, et al. 1995 The nuclear receptor superfamily: the second decade. Cell 83:835–839. 5. Huges MR, Malloy PJ, Kieback DG, Kesterson RA, Pike JW, Feldman D, O’Malley BW 1988 Point mutations in the human vitamin D receptor associated with hypocalcemic rickets. Science 242:1702–1705. 6. Haussler MR, Whitfield GK, Haussler CA, Hsieh JC, Thompson PD, Selznick SH, Dominguez CE, Jurutka PW 1988 The nuclear vitamin D receptor: biological and molecular regulatory properties revealed. J Bone Miner Res 13: 325–349. 7. Rachez C, Lemon D, Suldan Z, Bromleigh V, Gamble M, Naar M, Erdjument H, Tempst P, Freedman P 1999 Ligand-dependent transcription activation by nuclear receptors requires the DRIP complex. Nature 398:824–828. 8. Glass CK, Rosenfeld MG 2000 The co-regulator exchange in transcriptional functions of nuclear receptors. Genes Dev 14:121–141. 9. Narlikar GJ, Fan HY, Kingston RE 2002 Cooperation between complexes that regulate chromatin structure and transcription. Cell 108:475–487. 10. Emerson BM 2002 Specificity of gene regulation. Cell 109:267–270. 11. Hassan AH, Prochasson P, Neely KE, Galasinski SC, Chandy M, Carrozza MJ, Workman, JL 2002 Function and selectivity of bromodomains in anchoring chromatinmodifying complexes to promoter nucleosomes. Cell 111: 369–379. 12. Yanagisawa J, Kitagawa H, Yanagida M, Wada O, Ogawa S, Nakagomi M, Oishi H, Yamamoto Y, Nagasawa H, McMahon SB, Cole MD, Tola L, Takahashi N, Kato S 2002 Nuclear receptor function requires a TFTC-type histone acetyltransferase complex. Mol Cell 9:553–562. 13. Lanz RB, McKenna NJ, Onate SA, Albrecht U, Wong J, Tsai SY, Tsai MJ, O’Malley BW 1999 A steroid receptor coactivator, SRA, functions as an RNA and is present in an SRC-1 complex. Cell 97:17–27. 14. Brown CE, Howe L, Sousa K, Alley SC, Carrozza MJ, Tan S, Workman JL 2001 Recruitment of HAT complexes by direct
21.
22.
23. 24.
25. 26.
27. 28. 29. 30.
31. 32.
CHAPTER 18
Molecular Basis of the Diversity of Vitamin D Target Genes CARSTEN CARLBERG, THOMAS W. DUNLOP, CHRISTIAN FRANK, AND SAMI VÄISÄNEN Department of Biochemistry, University of Kuopio, Kuopio, Finland
I. Molecular Basis of the Genomic Actions of 1,25(OH)2D3 II. Classical Vitamin D–Receptor Binding Sites
III. Complex Vitamin D–Receptor Binding Sites IV. Conclusion References
I. MOLECULAR BASIS OF THE GENOMIC ACTIONS OF 1,25(OH)2D3
the p160-family, such as SRC-1, TIF2, and RAC3 [8]. These CoAs link the ligand-activated VDR to enzymes displaying histone acetyltransferase (HAT) activity that cause chromatin opening. Subsequently, ligand-activated VDR changes rapidly from the CoAs of the p160-family and those of the DRIP/TRAP family. The latter are part of a mediator complex of approximately 15 proteins that build a bridge to the basal transcription machinery [9]. In this way ligand-activated VDR fulfills two tasks, opening chromatin and activating transcription. The LBD of the VDR can be stabilized by 1,25(OH)2D3 or its analogs in an agonistic, antagonistic, or inverse agonistic conformation [10]. The position of helix 12 is the most critically important feature of these conformations, because it determines the distance that separates the charge clamp amino acids K246 in helix 3 and E420 in helix 12 that are both essential for VDR– CoA interaction (see Chapters 13–15). Most VDR ligands have been identified as agonists and only a few as pure or partial antagonists [11]. Two side-chain analogs, such as Gemini and its derivatives, have conditional inverse agonistic properties [12]. Antagonists induce CoR dissociation from the VDR but completely or partially prevent CoA interaction and thus transactivation [13]. At supramolar CoR concentrations inverse agonists actively recruit CoRs to the VDR and thus mediate repression of 1,25(OH)2D3 target genes. These ligand-triggered protein–protein interactions are the central molecular events of nuclear 1,25(OH)2D3 signaling.
A. Ligand-Triggered Protein–Protein Interactions of the VDR The vitamin D3 receptor (VDR) is the only nuclear protein that binds with high affinity to the biologically most active vitamin D metabolite, 1α,25-dihydroxyvitamin D3 (1,25(OH)2D3) [1]. The VDR is one of 11 members of the nuclear receptor (NR) superfamily that function as classical endocrine receptors. These include the receptors for the nuclear hormones retinoic acid, thyroid hormone, estradiol, progesterone, testosterone, cortisol, and aldosterol, which bind their specific ligand with a Kd of 1 nM or less [2]. Like most members of the NR superfamily, the VDR contains two zinc finger structures forming a characteristic DNA-binding domain (DBD) of 66 amino acids [3] and a carboxyterminal ligand-binding domain (LBD) of approximately 300 amino acids formed by 12 α-helices [4] (see Chapters 11–15). Ligand binding causes a conformational change within the LBD, whereby helix 12, the most carboxy-terminal α-helix, closes the ligand-binding pocket via a “mousetrap-like” intramolecular folding [5]. The LBD is also involved in a variety of interactions with nuclear proteins, such as other members of the NR superfamily, coactivator (CoA) and co-repressor (CoR) proteins [6]. CoR proteins, such as NCoR, SMRT, and Alien, link nonliganded, DNA-bound VDR to enzymes with histone deacetylase activity that cause chromatin condensation [7]. This provides VDR with intrinsic repressive properties comparable to those of retinoic acid and thyroid hormone receptors (RARs and T3Rs). The conformational change within VDR’s LBD after binding of 1,25(OH)2D3 or one of its agonistic analogs results in replacement of CoR by a CoA protein of VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
B. DNA Binding of the VDR An essential prerequisite for a direct modulation of transcription via 1,25(OH)2D3-triggered protein– protein interactions is the location of activated VDR close to the basal transcriptional machinery. This is achieved Copyright © 2005, Elsevier, Inc. All rights reserved.
314
CARSTEN CARLBERG, THOMAS W. DUNLOP, CHRISTIAN FRANK, AND SAMI VÄISÄNEN
through the specific binding of the VDR to a 1,25(OH)2D3 response element (VDRE) in the regulatory region of a primary 1,25(OH)2D3 responding gene [14]. The DBD of the VDR contacts the major grove of a hexameric sequence, referred to as core binding motif, with the consensus sequence RGKTSA (R = A or G, K = G or T, S = C or G). The affinity of monomeric VDR to a single core binding motif is not sufficient for the formation of a stable protein–DNA complex and thus VDR requires formation of homo- and/or heterodimeric complexes with a partner NR in order to allow efficient DNA binding [15]. In most cases the heterodimeric partner of VDR is the retinoid X receptor (RXR), another NR family member. The protein–DNA complex of a VDR-RXR heterodimer sitting on a VDRE therefore can be considered as a molecular switch for primary 1,25(OH)2D3 responding genes. Most NRs are able to dimerize in solution via their LBDs, but the DBDs dimerize only in the presence of DNA [16]. The DBD and the LBD of all NRs are linked by a hinge region of 35 to 50 amino acid residues that form a long α-helical structure according to the crystal structure of the DNA-bound T3R-RXR heterodimer [17]. The loop between this α-helix and the second zinc finger of the DBD contains a short six-amino-acid residue region, referred to as a T-box, which has been suggested to provide a dimerization interface for the interaction with the DBD of RXR [18]. Steric constrains allow dimerization of DBDs only on response elements (REs) with properly spaced core binding motifs. Modeling of the DBDs of VDR and RXR on DNA [17] suggested that an asymmetric arrangement, i.e., head-to-tail, as a direct repeat with three intervening nucleotides (DR3) provides the most efficient interface of the core DBDs. This fits with the 3-4-5 rule of Umesono et al. [19], in which VDR-RXR heterodimers should show optimal binding to DR3-type REs, whereas other NRs prefer DR4-type REs [e.g., T3R, constitutive androstane receptor (CAR) and pregnane X receptor (PXR)] and DR5-type REs (e.g., RARs). On DR3-, DR4-, and DR5-type REs, the different heterodimers bind with the same polarity, in which RXR always binds to the upstream hexamer and the partner receptor, e.g., VDR, to the downstream hexamer [20,21]. This specific and directed dimerization of the DBDs appears to be the major discriminative parameter for selective RE recognition.
II. CLASSICAL VITAMIN D–RECEPTOR BINDING SITES A. Classification of DR3-type REs Numerous studies (e.g., [15,22]) have confirmed Umesono’s suggestion [19] that VDR binds well to
DR3-type REs. Most of these studies also demonstrated that VDR preferentially or even exclusively forms heterodimers with RXR on these REs. DR3-type REs are therefore widely accepted as the classical structure of a VDRE. Every transcriptionally responsive primary 1,25(OH)2D3 target gene has to contain at least one VDRE in its promoter region and these VDREs are generally located relatively close to the transcription start site (TSS) of these genes. It is assumed that matrix attachment regions (MARs) subdivide genomic DNA into units of an average length of 100 kB containing the coding region of at least one gene [23]. DNA looping should be able to bring any DNA site within the same chromatin unit close to the basal transcriptional machinery that is assembled on the TSS (Fig. 1). This model suggests that also very distant sequences can serve as VDREs and that even sequences downstream of the TSS could serve as functional VDR binding sites. The list of the presently known natural VDREs (Table I) indicates that most of them have a DR3-type structure and are located within the first 1000 bp of promoter sequence upstream of the TSS. In addition, this VDRE list suggests that the consensus VDR core binding motif is RGKTSA, although some natural hexameric sequences show a significant deviation from this consensus sequence (Table I). However, one has to take into account that all these VDREs had been identified before the genomes of Homo sapiens and other species had been sequenced, and only limited promoter sequences were available. Moreover, only a very few of these VDREs, such as that of the rat osteocalcin gene, are understood in their promoter context, i.e., in the context of chromatin organization and flanking binding sites for other transcription factors (TFs). Therefore, the present VDRE list (Table I) has to be taken with some precaution. Results from in silico promoter screening and a more detailed understanding of chromatin organization of primary 1,25(OH)2D3 target genes will revise this list soon. On the basis of presently published data, the strongest VDR-RXR heterodimer binding DR3-type VDREs has been identified within the rat atrial natriuretic factor (ANF) promoter [24], the mouse and pig osteopontin promoter [25,26], and the chicken carbonic anhydrase II promoter [27]. These four elements were categorized into class I (see Table I). The DR3-type VDREs of the human and rat 24-hydroxylase (CYP24) promoter [28,29], the human Na+-dependent inorganic phosphate transporter type II promoter [30], the rat osteocalcin promoter [31], the human parathyroid hormone (PTH) promoter [32], and the first VDRE of the rat PTH related peptide (PTHrP) promoter [33] demonstrated only 6–30% of the binding strength of the rat ANF VDRE and were grouped into class II. The binding of VDR-RXR heterodimers to the ten
315
CHAPTER 18 Diversity of Vitamin D Target Genes
VDRE R X R
MAR
V D R
TSS Basal transcriptional machinery
Coding region
? V D R MAR
R X R
VDRE
FIGURE 1
Schematic structure of a chromatin unit. A chromatin unit is a region between two MARs and often contains only one gene. DNA looping should permit that every DNA sequence within the same chromatin unit to be located near the basal transcriptional machinery. The occurrence of VDREs downstream of a coding region has not yet been proven experimentally.
DR3-type VDREs in classes I and II was found to be enhanced by 1,25(OH)2D3 by a factor of 2 to 5 [34]. In contrast, the second VDRE of the rat PTHrP promoter [35], the rat calbindin D9k promoter [36] and the quail slow myosin heavy chain promoter [37], the human growth hormone promoter [38], the chicken integrin β3 promoter [39], the chicken PTH promoter [40], and the human p21WAF1/CIP1 promoter [41] were grouped together into class III, as they showed less than 5% of the affinity for VDR-RXR heterodimers compared to the rat ANF VDRE and no significant ligand inducibility. According to these stringent in vitro criteria, the core sequences of the class III members cannot be considered as functional VDREs. The in vitro binding affinity of VDR-RXR heterodimers to VDREs was shown to be proportional to their in vivo functionality in transiently transfected cells, i.e., to mediate in a heterologous promoter context induction of reporter gene activity [21,42]. In this assay system class III VDREs did not show any in vivo functionality. The hexameric sequences of the class III VDREs show a significant deviation from the RGKTSA consensus and explains their relatively low ability of complex formation with VDR-RXR heterodimers. However, it is possible that class III VDREs may gain responsiveness to 1,25(OH)2D3 in their natural promoter context through the help of flanking partner proteins. Moreover, the functionality of a 1,25(OH)2D3 responding gene will also
depend on a potential cooperative action of two or more VDREs, such as in the case of the CYP24 gene [43].
B. Do DR3-Type REs Differ in Their Function? Since the pleiotropic physiological effects of 1,25(OH)2D3 are based in the final analysis on transcriptional regulation of primary 1,25(OH)2D3 responding genes, the genes should be explained through activation of VDR-RXR heterodimers bound to DR3-type VDREs. Therefore, there have been attempts to explain the various effects of 1,25(OH)2D3 by a multiplicity of 1,25(OH)2D3 signaling pathways that are based on DR3-type VDREs [44]. This leads to the question of whether each DR3-type VDRE has an individual functionality that may explain the specific effects of 1,25(OH)2D3 or whether all DR3-type VDREs function in the same way. The affinity of the known natural DR3-type VDREs for VDR-RXR heterodimers in vitro (at a fixed protein– DNA ratio) seems to be their major discriminating parameter, which allowed the grouping of the VDREs into classes as discussed earlier (Table I). VDR-RXR heterodimers appear to form identical complexes on the 10 VDREs of classes I and II, since indistinguishable VDR-RXR heterodimer conformations were observed on these VDREs [34]. This is in contrast to a report that VDR-RXR heterodimers take different conformations
316
CARSTEN CARLBERG, THOMAS W. DUNLOP, CHRISTIAN FRANK, AND SAMI VÄISÄNEN
TABLE I Natural VDRCS Gene
Species
Class I DR3-type VDREs ANF Osteopontin Osteopontin Carbonic anhydrase II
Rat Mouse Pig Chicken
AGAGGTCATGAAGGACA AAGGTTCACGAGGTTCA ATGGGTCATATGGTTCA GAAGGGCATGGAGTTCG
−907 −759 −2261 −62
[24] [26] [25] [27]
Human Human Human
CGAGGTCAGCGAGGGCG GGAGTTCACCGGGTGTG CAGGGGCAGCAAGGGCA
−171 −291 −1977
[28] [28] [30]
Human Human Rat Rat Rat Rat
ATGGTTCAAAGCAGACA CCGGGTGAACGGGGGCA CTGGGTGAATGAGGACA CGAGGTGAGTGAGGGCG AGGGTTCAGCGGGTGCG TAAGGTTACTCAGTGAA
−122 −500 −457 −152 −259 −805
[32] [52] [31] [29] [29] [35]
Class III DR3-type VDREs p21 Growth hormone Bone sialo protein PTH related peptide Calbindin D9k Integrin β3 PTH Slow myosin heavy chain
Human Human Rat Rat Rat Chicken Chicken Quail
GTAGGGAGATTGGTTCA TGGGGTCAACAGTGGGA GAAGGGTTTATAGGTCA AGGGTGGAGAGGGGTGA GAGGGTGTCGGAAGCCC GCGAGGCAGAAGGGAGA GAGGGTCAGGAGGGTGT GAAGGACAAAGAGGGGA
−779 −59 −30 −1107 −490 −772 −76 −801
[41] [38] [78] [33] [36] [39] [40] [37]
DR4-type VDREs Pit-1 Calbindin D28k
Rat Mouse
GAAGTTCAGCGAAGTTCA CTGGGGGATGTGAGGAGA
−683 −200
[48] [95]
DR6-type VDREs Osteocalcin Phospholipase C-γ1 CYP24 Fibronectin
Human Human Rat Mouse
TTTGGTGACTCACCGGGTGA GCAGGTCAGACCACTGGACA CGGGTCGAGCCCAGGGTTCA CCGGGTGACGTCACGGGGTA
−514 −805 −231 −152
[51] [56] [55] [54]
ER9-type VDREs Calbindin D9k Osteocalcin c-fos p21
Human Rat Mouse Mouse
TGCCCTTCCTTATGGGGTTCA TGCACTGGGTGAATGAGGACA TGACCCTGGGAACCGGGTCCA TGACCTGAAAGTGGAAGGTGA
−147 −461 −482 −3811
[21] [21] [57] [96]
Complex VDREs Osteocalcin Osteocalcin c-fos
Human Rat Mouse
TTTGGTGACTCACCGGGTGAACGGGGGCA TGCACTGGGTGAATGAGGACA AGGTGAAAGATGTATGCCA AGACGGGGGTTGAAAG
−514 −461 −178
[42] [21] [77]
Class II DR3-type VDREs CYP241 CYP242 Na+-dependent inorganic Phosphate transporter type II PTH Osteocalcin Osteocalcin CYP241 CYP242 PTH related peptide
Sequence
Position
Reference
The core sequence of the different types of VDREs and their position in relation to the TSS are indicated. Hexameric core binding motifs are in bold and deviations from the consensus sequence RGKTSA are underlined.
317
CHAPTER 18 Diversity of Vitamin D Target Genes
on the mouse osteopontin and the rat osteocalcin DR3type VDRE [44]. In that study, nuclear extracts from bone cells provided protein complexes on the two VDREs that were differentially recognized by antibodies. One could assume that the VDR-RXR heterodimers in these complexes interact with different types of cofactors as has been described for estrogen REs [45]. However, VDRRXR heterodimers formed on the different DR3-type VDREs display no significant differences in their interaction with a given CoA or CoR protein. The ligand sensitivity of VDR-RXR heterodimers bound to the different DR3-type VDREs showed no significant deviation from the average value of 0.1 nM. This suggests that the ligand-triggered VDR-RXRVDRE complex formation does not depend on ligand concentration or VDR conformation [34]. In addition, VDR-RXR heterodimers show no significant differences in their 1,25(OH)2D3-triggered interaction with CoA and CoR proteins. These findings suggest that heterodimers are able to differentiate between individual DR3-type VDREs on the basis of the different protein-DNA affinities. However, when the complexes are formed, they appear to function in exactly the same fashion. This means that at least comparative in vitro data cannot support any indications of multiple DR3type VDRE-mediated 1,25(OH)2D3 signaling pathways. Investigations of VDR-RXR heterodimers and their conformations have provided multiple new insights on the functionality of these molecular switches that may well explain their function in living cells [7,46]. This makes it likely that the lack of variation of the in vitro functionality of DR3-type VDREs (from positively as well as negatively regulated genes) can be transferred to the in vivo situation. This facilitates further investigation of the principles of DR3-type VDRE-mediated gene regulation, since an observation that has been made with one specific DR3-type VDRE can be generalized for the whole family of DR3-type VDREs. However, an explanation for the multiplicity of 1,25(OH)2D3 signaling has to be found at a different level.
C. Other DR-Type VDREs Since a variety of DR3-type VDREs seems unable to explain the pleiotropic physiological action of 1,25(OH)2D3 and the dissociated profile that some synthetic 1,25(OH)2D3 analogs display, other VDRE structures such as direct repeats with four and six intervening nucleotides (DR4 and DR6) or everted repeats with nine spacing nucleotides (ER9) may offer an alternative explanation. A comparison of the individual VDRE core sequences (Table I) with their classification according to the affinity for VDR-RXR heterodimers
suggests that the degree of deviation from the core binding motif consensus sequence RGKTSA [47] is proportional to the loss of in vitro functionality [34]. Interestingly, the DR4-type RE of the rat pit-1 gene [48], which contains two perfect core binding motifs, was found to be stronger than any known natural DR3-type VDRE [34]. Does this suggest that DR4-type REs are even better VDREs than DR3-type structures? One has to consider that a DR4-type REs is also recognized by the heterodimeric complexes of T3R, CAR, PXR, and other orphan NRs with RXR [49,50], whereas the same complexes bind to DR3-type REs less tightly than VDR-RXR heterodimers. The competitive situation on DR4-type REs may therefore be the reason why in vivo VDR-RXR heterodimers still prefer DR3-type REs. Moreover, VDR-RXR heterodimers bind to DR4-type REs in the same conformation as to DR3-type REs [50], i.e., there seems to be no differential action of VDR on these elements due to a differential complex formation with RXR. The VDRE of the human osteocalcin promoter was the first identified natural binding site for the VDR [51,52]. It was initially described as a DR6-type structure, but later a third cryptic hexamer was identified at a distance of three nucleotides, so that the whole VDRE is more likely a complex DR6/DR3-type RE (Table I). The DR6 part of the VDRE has been shown to bind VDR homodimers [15] and VDR-RAR heterodimers [53], whereas the DR3 part weakly binds VDR-RXR heterodimers. Other examples of a DR6-type VDREs that bind VDR homodimers and VDR-RAR heterodimers have been identified in the promoters of mouse fibronectin [54], rat CYP24 [55], and human phospholipase C [56]. Their functionality remains to be determined.
D. Everted Repeats Dimerization facilitates cooperative, high-affinity interaction of NRs, such as VDR and RXR, with specific hexameric core binding motifs. As schematically depicted in Fig. 2, VDR-RXR heterodimers bind to DR-type VDREs in a nonsymmetrical head-to-tail tandem arrangement. In contrast, like all palindromic sequences, ER-type VDREs are per se symmetric, since the heterodimeric partner receptors bind in a tail-to-tail arrangement. However, natural ER9-type REs were found to be sufficiently asymmetric in their core binding sequences, in order to allow a polarity-determined binding of heterodimers [21]. So far, ER9-type VDREs have been identified within the promoter regions of the human calbindin D9k gene [21], the mouse c-fos gene [57], and the rat osteocalcin gene [21]. Interestingly, the last VDRE shares one core binding motif with the
318
CARSTEN CARLBERG, THOMAS W. DUNLOP, CHRISTIAN FRANK, AND SAMI VÄISÄNEN
Direct repeats
Everted repeat 9
3 or 4
DBD
Hinge
LBD Helix 12
Helix 12
RXR
VDR
RXR
VDR
FIGURE 2 DNA complex formation of VDR-RXR heterodimers. The DNA binding of VDR-
RXR heterodimers to DR- or ER-type VDREs is schematically indicated. Some critical α-helices within the LBD of VDR and the DBDs of VDR and RXR are indicated by cylinders.
DR3-type VDRE identified by Demay et al. [31], and there appears to be a complex ER9/DR3-type RE. On DR3-type VDREs, both receptor DBDs are located at roughly the same side of the DNA (tilted by 51.4°), whereas on ER9-type VDREs the DBDs are at nearly opposite sites of the DNA (tilted by 154.3°) [18]. Moreover, the distance between the DBDs along the axis of the DNA is threefold greater on ER9-type VDREs than on DR3-type VDREs (Fig. 2). Computer modeling and experimental studies have both shown that the T-box of VDR contributes to the dimerization interface of the extended VDR DBD with the RXR DBD, i.e., on a DR3-type VDRE the VDR DBD directly contacts the RXR DBD, whereas a direct contact of both DBDs is not possible on an ER9-type VDRE. The lack of DBD dimerization on an ER9-type RE suggests that DNA-driven cooperativity between the partner DBDs is unlikely. In fact, the spacing of the two core binding motifs is less stringent than on DRtype REs, as ER7-, ER8-, and ER10-type structures are also able to bind dimeric VDR complexes [58,59]. However, the Kd value for the binding of VDR-RXR heterodimers to DR3- and ER9-type REs has been determined to be in a similar range of 0.5 to 1 nM [21]. According to our model of multiple 1,25(OH)2D3 signaling pathways [60], the pleiotropic function of 1,25(OH)2D3 is based on a variety of dimeric VDR complexes bound to different types of VDREs. The model assumes that each of these VDR-VDRE
complexes may be representative for a group of primary 1,25(OH)2D3 target genes that are involved in the regulation of a distinct proportion of the pleiotropic actions of this nuclear hormone. In support of the model, some 1,25(OH)2D3 analogs have shown the tendency to preferentially activate VDR-RXR heterodimers that are bound to ER9-type VDREs [61], whereas other analogs prefer DR3-type VDRE-bound VDR complexes [62]. The selective biological profile of the analog EB1089, i.e., having both potent antiproliferative potential and a reduced calcemic effect [63], was associated with a higher selectivity (approximately 15 times) to activate ER9-type VDREs compared to DR3-type VDREs [61]. Because of the relatively low number of known primary 1,25(OH)2D3 target genes with characterized VDREs, this idea has to be proven. However, the genes of mouse c-fos [57] and human and mouse p21WAF1/CIP1 [14] each contain an ER9-type VDRE in their regulatory regions. A VDRE-selective in vitro stabilization of VDR-RXR heterodimers was demonstrated by ligand-dependent gel shift assays and showed that EB1089 mediated the stabilization of VDR-RXR heterodimers on an ER9-type VDREs at approximately eightfold lower concentrations than on a DR3-type VDRE [64]. In contrast, the natural hormone 1,25(OH)2D3 showed no significant selectivity. Taken together, the selective activation of ER9-type VDREs by EB1089, i.e., the observation of promoter selectivity of a 1,25(OH)2D3 analog, seems to be based on the enhanced DNA binding affinity of
319
CHAPTER 18 Diversity of Vitamin D Target Genes
a subset of all VDR-RXR heterodimers to this type of VDRE.
E. Do Hexameric Core Binding Motifs Provide Sufficient Specificity for DNA Binding of VDR-RXR Heterodimers? It has been known for some time that some NRs, such as T3R, which can also bind as monomers to DNA, have a clear preference for certain 5′-flanking sequences to the hexameric core binding motif [65,66]. In addition, it had been demonstrated that in particular the dinucleotide sequence 5′-flanking of the downstream hexamer of a RE have a significant effect on the complex formation of NR heterodimers, such as RARRXR, T3R-RXR, and VDR-RXR [66–68]. This provides the sequence of the spacer between the hexameric motifs of natural REs with a critical role for determining the specific recognition of the RE and the regulation of the respective gene. The dinucleotide AG, as found, e.g., in front of the downstream motif of the DR4-type RE of the rat pit-1 gene, seems to be optimal for VDRRXR heterodimers [68]. This preference should be independent of the type of RE and apply to all VDRE types. The crystal structure of DNA-bound T3R-RXR heterodimers [17] demonstrated a contact of amino acids of the carboxy-terminal extension of the T3R-DBD with the two 5′-flanking nucleotides of the downstream core binding motif. Similar assumptions could be made for the VDR. Interestingly, variations of the dinucleotide sequence 5′-flanking to the upstream core binding motif also have a lower but still significant influence on complex formation and functional activity of VDR-RXR heterodimers [68]. This suggests that RXR also specifically contacts 5′-flanking nucleotides to its binding motif. In summary, 5′-flanking sequences should be considered as an integral part of a RE, so that more likely octameric motifs instead of hexameric sequences are specific VDR binding sites. A reanalysis of previously characterized VDREs (Table I) for their 5′-flanking sequences will provide a better understanding of their relative strength. Even more important is the possibility of a more accurate prediction of putative VDREs from large amounts of sequence data now available from the human genome.
F. VDR Homodimers and RXR-Independent 1,25(OH)2D3 Signaling VDR homodimers were found to bind and initiate ligand-driven transactivation from DR3- and DR6-type VDREs [15,69]. Moreover, homodimers of VDR DBDs
have even been crystallized on DR3-type REs [3] confirming the view that they may have some physiological relevance. The in vitro binding of VDR homodimers on idealized DR3- and DR6-type REs is not as high as that of VDR-RXR heterodimers on DR3-type REs [42,70]. However, in vitro binding affinity of VDR-RXR heterodimers to natural DR3-type VDRE of class III is also weak [34] and VDR homodimers may be specifically stabilized by CoA proteins [71]. In vitro complex formation experiments as well as yeast or mammalian two-hybrid experiments leave no doubts that within the NR superfamily, RXR is the most efficient protein– protein interaction partner of the VDR. This interaction preference is evolutionarily conserved and also applies to the 15 closest members of VDR within its superfamily [72]. Moreover, at least one of the three RXR subtypes seems to be expressed in every mammalian cell, which makes investigations of putative RXRindependent 1,25(OH)2D3 signaling pathways difficult. In vitro as well as in a Drosophila cell system the formation of VDR homodimers has been shown [15]. Evidence for VDR-RAR and VDR-T3R heterodimers has also been reported [42,53,59]. The concentration of protein was very high, however. VDR-RAR and VDR-T3R heterodimers showed preference for DR5and DR6-type REs [42] and have also been observed on ER7- to ER10-type REs [58,59]. Taken together, on most VDREs and in the vast majority of physiological situations, VDR-RXR heterodimers seem to mediate 1,25(OH)2D3 signaling. There may be exceptions, however. In these cases, their in vitro relevance needs to be studied in the context of the natural promoter in a chromatin environment.
III. COMPLEX VITAMIN D–RECEPTOR BINDING SITES A. Simple versus Complex VDREs With the exception of the VDREs in the human and rat osteocalcin gene, all natural VDREs are formed by only two hexameric core binding motifs in a DR3, DR4, DR6, and ER9 arrangement (Table I). These latter targets may be therefore considered as “simple” VDREs. The question is, whether the occurrence of one simple VDRE within a promoter is sufficient for gene responsiveness to 1,25(OH)2D3. Because of its optimized 5′-flanking dinucleotide and core binding motif sequences the DR4-type RE of the rat pit-1 gene is the most efficient known VDRE in vitro [34,50]. However, the chromatin in the region of the pit-1 gene promoter containing this RE seems to be closed in the adult rat, so that the responsiveness of the gene to 1,25(OH)2D3
320
CARSTEN CARLBERG, THOMAS W. DUNLOP, CHRISTIAN FRANK, AND SAMI VÄISÄNEN
is lower than expected [73]. This indicates that a high in vitro binding affinity of VDR-RXR heterodimers for a VDRE is not sufficient for responsiveness to 1,25(OH)2D3. When the promoter region that contains the VDRE is covered by condensed chromatin, VDRRXR heterodimers are unable to bind there. This makes sufficiently decondensed chromatin an essential prerequisite for a functional VDRE. Chromatin decondensation is achieved by the activity of HATs, which are recruited to their local chromatin target by CoA proteins. In turn, these CoAs are transiently attracted to a promoter region by ligand-activated NRs and other active TFs. Therefore, the more TF binding sites a given promoter region contains and the more of these TFs are expressed in the respective cell, the higher is the chance that this area of the promoter gets locally decondensed. One example is the VDRE of the rat osteocalcin gene, which is flanked on both sides with a binding site for the TF Runx2/Cbfa1 [74]. By contacting CoA proteins and HATs Runx2/Cbfa1 seems to mediate the opening of chromatin locally, which allows efficient binding of VDR-RXR heterodimers to this decondensed region to occur. This mechanism suggests that VDREs are better targets for VDR-RXR heterodimers, if other TFs are bound to the same chromatin region. In this respect, promoter context and cell-specific expression of other TFs may be of greater importance to VDRE functionality and specificity than its in vitro binding profile. If this idea is true, it will apply to other members of the NR superfamily and will question the validity of isolated simple VDREs. Therefore, in the future VDREs will have to be understood as complex structures with multiple TF binding sites. Some of these TF binding sites will be other NR core binding motifs. The DR6/DR3 and ER9/DR3 structures of the VDREs of the human and rat osteocalcin genes, respectively, are the first examples of complex VDREs. These two complex VDREs show only limited homology to each other, although they are derived from orthologous genes. This suggests that for an important primary 1,25(OH)2D3 responding gene such as osteocalcin, there may be limited evolutionary pressure for a specific VDRE structure. It seems to be more important to guarantee an efficient binding of the VDR to a promoter in competition with the tight packaging of nucleosomes. It is interesting to note that the complex VDRE of the human osteocalcin gene is overlaid by a binding site of the TF AP-1 [75], which provides the RE with an increased activity. These types of REs are also observed for other NRs and often referred to as composite REs [76]. Another interesting example of a complex/composite VDRE has been reported in the mouse c-fos promoter [77]. Within this VDRE three hexameric core binding motifs are forming a DR7/DR7 structure, which contains an internal
binding site for the TF NF-1. Additional examples are the VDRE of the human and mouse fibronectin gene, which contains an internal binding site for the TF CREB [54], or the VDRE of the rat bone sialo protein, which also seems to bind the general TF TBP [78].
B. RE Clusters In mammals, the most responsive known primary 1,25(OH)2D3 target gene is CYP24, which gene product is the key enzyme involved in the catabolism of 1,25(OH)2D3. Activation of this gene provides a negative feedback loop mechanism controlling the level of the hormonal ligand. The reason for the strong responsiveness of the CYP24 gene appears to be that it contains two DR3-type VDREs separated by a distance of less than 100 bp in close proximity to the TSS. These DR3-type VDRE clusters are evolutionarily conserved between humans and rodents [43,79]. The sequences of both VDREs are not optimal and are categorized into class II (Table I). However, these less optimal structures are more than compensated by the fact that they are located in close vicinity to each other and to the TSS of the gene. In addition, binding sites for the TF Ets-1 have been characterized within the CYP24 promoter and seem to interfere synergistically with the two VDREs, i.e., the cluster of VDR and Ets binding sites form another type of complex VDRE. Such types of RE clusters are also known for other genes that are primary targets of additional members of the NR superfamily, such as VDR’s close relatives CAR and PXR. Investigation of natural CAR and PXR target gene, such as CYP2B6 or CYP3A4, respectively [80,81], indicate that a single RE is insufficient for mediating the regulatory role of the receptors and that more likely at least two CAR or PXR REs in close proximity to each other are required. In the case of CAR, these multiple RE clusters are called phenobarbital response enhancer modules (PBREMs), because they mediate the responsiveness to the CAR activator phenobarbital. The CYP2B6 gene contains two DR4-type REs with an additional binding site for the TF NF-1 [80], whereas the PBREM of the UDPglucuronosyltransferase 1A1 gene is formed by three CAR REs [82]. In contrast, the CYP3A4 gene contains an ER6-type RE proximal to the TSS and a more distal DR3/ER6 cluster [81]. Taken together, these assemblies represent two or more simple NR REs together with binding sites for other types of TFs. Their overall function seems to follow the same rules, i.e., the greater number of NRs and other TFs bind to such promoter regions, the greater is the chance that they induce histone acetylation and chromatin decondensation.
CHAPTER 18 Diversity of Vitamin D Target Genes
Another interesting observation in relation to the RE clusters is that they are rather promiscuously bound by related members of the NR superfamily, such as VDR, CAR, and PXR. VDR was shown to activate the CYP2B6, CYP3A4, and CYP2C9 genes by replacing CAR and PXR within the respective RE clusters [83]. In particular, the CYP3A4 gene, which contains a DR3-type RE within its RE cluster, was shown to be stimulated effectively by 1,25(OH)2D3 and can be considered as a primary 1,25(OH)2D3 responding gene. This suggests that 1,25(OH)2D3 and the VDR may have an impact on the metabolism of prescription drugs, of which 60% are metabolized by the CYP3A4 enzyme [84]. In this context, it is interesting to note that the VDR seems to be able to act as an intestinal bile acid sensor, because certain bile acids have been identified as low-affinity VDR ligands [85]. These actions are also mediated by CYP3A4. It is very likely that future investigations will reveal more of these 1,25(OH)2D3 target genes, thus enlarging the physiological importance of this hormone to an even greater extent.
C. Negative VDREs A relatively undercharacterized aspect of 1,25(OH)2D3 signaling is the mechanism of down-regulation of 1,25(OH)2D3 responsive genes. Microarray experiments have indicated that nearly half of all primary 1,25(OH)2D3 responding genes are down-regulated by the hormone, but few of them have been studied in more detail. It is obvious that only genes which show basal activity can be down-regulated, i.e., these genes exhibit basal activity due to other TFs binding to their promoter. There are several different models that attempt to explain how 1,25(OH)2D3 and the VDR can mediate down-regulation of genes, but the common theme is that VDR counteracts the activity of specific TFs. In the situation where these activating TFs are other NRs or TFs that bind to composite NR REs, VDR could simply compete for DNA binding sites [86,87]. In a similar way, VDR could also compete for binding to partner proteins, such as RXR, or for common CoAs, such as SRC-1 or p300 [88]. In all these situations the down-regulating effects of the VDR should be of general impact, i.e., the mechanism could apply to other genes in the same way. So far, however, no general down-regulating effects of 1,25(OH)2D3 have been reported. The binding of 1,25(OH)2D3 to the VDR results in a ligand-dependent conformational change, which leads to an exchange of proteins that bind to the LBD of the VDR, i.e., altering protein-protein interaction profiles but no changes in VDR-DNA interaction properties.
321 In most cases, however, 1,25(OH)2D3-dependent down-regulation of a gene involves specific DNA binding sites on the specific promoter, which are referred to as negative VDREs. Experiments suggest that DNA binding of the VDR to down-regulated genes such as the PTH gene, does not involve RXR [89,90], whereas other studies, such as on the ANF gene [91], come to opposite conclusions. As discussed above, there are also alternative partner proteins for the VDR, such as RAR or VDR itself, and one might speculate that cell- and gene-specific TFs could help the VDR in binding efficiently to DNA. Investigation of the 1,25(OH)2D3-mediated repression of the PTHrP gene indicates that protein kinases may be involved in the down-regulation of this gene [92]. Some of the natural VDREs in Table I were identified within genes that are down-regulated by 1,25(OH)2D3 and are therefore more likely to be negative VDREs. These sequences are not revealing, however, because in the absence of a natural promoter context and normal chromatin organization, these VDREs behave like positive VDREs. This again highlights the fact that promoter context, the chromatin status, and the cluster of neighboring TFs are important for the function of a simple VDRE. Thus simple VDREs cannot act as negative VDREs. In the case of the calcitonin gene, two separate promoter regions are suggested to be responsible for the down-regulation of gene transcription by 1,25(OH)2D3 [93] (see Chapter 39). Together, this information indicates that mechanisms governing 1,25(OH)2D3-mediated down-regulation is complex. The concept that VDRE clusters together with and other TF binding sites regulate primary 1,25(OH)2D3 responding genes suggests that a promoter may contain both negative and positive VDREs. The activities of the different VDREs are determined by the promoter context and may not be simultaneously active. One might imagine that prior to stimulation with 1,25(OH)2D3 only the negative VDREs bind the VDR and recruit CoRs. This would actively condense the chromatin on a particular promoter region. The addition of ligand induces the release of CoR proteins and reduces chromatin density. The VDR may then be transiently released from the negative VDRE and bind to a positive VDRE, which may be uncovered through 1,25(OH)2D3-dependent local nucleosome acetylation. The VDR then interacts with CoAs of the DRIP/TRAP mediator protein complex on this positive VDRE leading to transient transcriptional activation. After a certain period of time, newly synthesized, unliganded VDR again binds to the negative VDRE, which initiates chromatin closing and inactivation of the positive VDRE. In this or even more complex scenarios, the balance between negative and positive
322
CARSTEN CARLBERG, THOMAS W. DUNLOP, CHRISTIAN FRANK, AND SAMI VÄISÄNEN
VDREs could explain the time course of the activation of primary 1,25(OH)2D3 responding genes.
have to be included in more effective versions of in silico screening software.
D. Whole-Genome Screening for Putative VDREs
IV. CONCLUSION
Considering both orientations of DNA, the NR core binding motif RGKTSA should be found every 256 bp of genomic DNA. Dimeric assemblies of such hexamers should show up as direct repeats every 65,536 bp of promoter sequence and as everted repeats every 32,768 bp. Since VDR-RXR heterodimers bind comparably well to DR3-, DR4-, and ER9-type REs, an in silico screening is expected to identify a putative VDR binding site every 16,384 bp. This would predict nearly 200,000 putative VDREs within the human genome, and as a consequence on average every gene would contain several VDREs in its promoter and should be responsive to 1,25(OH)2D3. Similar calculations apply to other members of the NR superfamily, and for TFs with a shorter specific binding site even higher numbers would be predicted. A realistic number of 1,25(OH)2D3 responding genes is far less than this, perhaps a few hundred. The number of VDR molecules varies from hundreds to several thousand molecules. These calculations make it obvious that not every putative VDR binding site is used in nature in any cell at any given time. The most obvious reason is that most of these sequences are effectively covered by nucleosomes, so that they are not accessible to the VDR. This applies in particular to those sequences that are isolated from other NR or TF binding sites or lie distant from the promoter. This perspective strongly discourages the idea that isolated, simple VDREs may be functional in vivo. Therefore, the presently identified simple VDREs (Table I) may be parts of more complex VDREs as already demonstrated for the CYP24 and osteocalcin gene. An effective in silico prediction of novel VDREs has to focus on the identification of complex VDREs. Unfortunately, presently available in silico screening software, such as NUBISCAN [94], is unable to identify complex VDREs. Other available programs such as TRANSFAC have the capability to identify larger numbers of REs for different TFs, making the identification of potential complex REs possible. However, even these programs are unable to predict chromatin condensation states and nucleosome positioning, which is essential information in determining the likelihood that a given RE lies in an accessible region of a promoter. General parameters, such as nucleosome positioning and interspecies homology screening of regulatory sequences as well as the binding sites of all other TFs and their cell-specific expression patterns,
The sequencing of the complete human genome and the genomes of other species, i.e., the availability of all regulatory sequences, enables a more mature understanding of the diversity of 1,25(OH)2D3 target genes. Perhaps the idea of simple isolated VDREs, such as those listed in Table I, should shift to the concept of complex VDREs, of which the simple VDRE represents the core. Depending on the temporal presence of cellspecific TFs, these complex REs may act positively or negatively in respect to 1,25(OH)2D3. The coordinated action of these different types of VDREs could explain the individual response of target genes to 1,25(OH)2D3.
Acknowledgment The Academy of Finland (grant 50319) supported the work.
References 1. Sutton AL, MacDonald PN 2003 Vitamin D: more than a “bone-a-fide” hormone. Mol Endocrinol 17:777–791. 2. Chawla A, Repa JJ, Evans RM, Mangelsdorf DJ 2001 Nuclear receptors and lipid physiology: opening the X-files. Science 294:1866–1870. 3. Shaffer PL, Gewirth DT 2002 Structural basis of VDR–DNA interactions on direct repeat response elements. EMBO J 21:2242–2252. 4. Rochel N, Wurtz JM, Mitschler A, Klaholz B, Moras D 2000 Crystal structure of the nuclear receptor for vitamin D bound to its natural ligand. Mol Cell 5:173–179. 5. Moras D, Gronemeyer H 1998 The nuclear receptor ligandbinding domain: structure and function. Curr Opin Cell Biol 10:384–391. 6. Glass CK, Rosenfeld MG 2000 The coregulator exchange in transcriptional functions of nuclear receptors. Genes Dev 14:121–141. 7. Polly P, Herdick M, Moehren U, Baniahmad A, Heinzel T, Carlberg C 2000 VDR-Alien: a novel, DNA-selective vitamin D3 receptor-corepressor partnership. FASEB J 14:1455–1463. 8. Leo C, Chen JD 2000 The SRC family of nuclear receptor coactivators. Gene 245:1–11. 9. Rachez C, Lemon BD, Suldan Z, Bromleigh V, Gamble M, Näär AM, Erdjument-Bromage H, Tempst P, Freedman LP 1999 Ligand-dependent transcription activation by nuclear receptors requires the DRIP complex. Nature 398:824–828. 10. Carlberg C 2003 Molecular basis of the selective activity of vitamin D analogues. J Cell Biochem 88:274–281. 11. Toell A, Gonzalez MM, Ruf D, Steinmeyer A, Ishizuka S, Carlberg C 2001 Different molecular mechanisms of vitamin D3 receptor antagonists. Mol Pharmacol 59:1478–1485.
CHAPTER 18 Diversity of Vitamin D Target Genes
12. Macias Gonzalez M, Samenfeld P, Peräkylä M, Carlberg C 2003 Corepressor excess shifts the two side chain vitamin D analogue Gemini from an agonist to an inverse agonist of the vitamin D receptor. Mol Endocrinol 17:2028–2038. 13. Väisänen S, Ryhänen S, Saarela JT, Peräkylä M, Andersin T, Maenpää PH 2002 Structurally and functionally important amino acids of the agonistic conformation of the human vitamin D receptor. Mol Pharmacol 62:788–794. 14. Carlberg C, Polly P 1998 Gene regulation by vitamin D3. Crit Rev Eukaryot Gene Expr 8:19–42. 15. Carlberg C, Bendik I, Wyss A, Meier E, Sturzenbecker LJ, Grippo JF, Hunziker W 1993 Two nuclear signalling pathways for vitamin D. Nature 361:657–660. 16. Glass CK 1994 Differential recognition of target genes by nuclear receptor monomers, dimers, and heterodimers. Endocr Rev 15:391–407. 17. Rastinejad F, Perlmann T, Evans RM, Sigler PB 1995 Structural determinants of nuclear receptor assembly on DNA direct repeats. Nature 375:203–211. 18. Quack M, Szafranski K, Rouvinen J, Carlberg C 1998 The role of the T-box for the function of the vitamin D receptor on different types of response elements. Nucleic Acids Res 26: 5372–5378. 19. Umesono K, Murakami KK, Thompson CC, Evans RM 1991 Direct repeats as selective response elements for the thyroid hormone, retinoic acid, and vitamin D3 receptors. Cell 65: 1255–1266. 20. Quelo I, Kahlen JP, Rascle A, Jurdic P, Carlberg C 1994 Identification and characterization of a vitamin D3 response element of chicken carbonic anhydrase-II. DNA Cell Biol 13:1181–1187. 21. Schräder M, Nayeri S, Kahlen JP, Müller KM, Carlberg C 1995 Natural vitamin D3 response elements formed by inverted palindromes: polarity-directed ligand sensitivity of vitamin D3 receptor–retinoid X receptor heterodimer-mediated transactivation. Mol Cell Biol 15:1154–1161. 22. MacDonald PN, Dowd DR, Nakajima S, Galligan MA, Reeder MC, Haussler CA, Ozato K, Haussler MR 1993 Retinoid X receptors stimulate and 9-cis retinoic acid inhibits 1,25-dihydroxyvitamin D3-activated expression of the rat osteocalcin gene. Mol Cell Biol 13:5907–5917. 23. Hancock R 2000 A new look at the nuclear matrix. Chromosoma 109:219–225. 24. Kahlen JP, Carlberg C 1996 Functional characterization of a 1,25-dihydroxyvitamin D3 receptor binding site found in the rat atrial natriuretic factor promoter. Biochem Biophys Res Commun 218:882–886. 25. Zhang Q, Wrana JL, Sodek J 1992 Characterization of the promoter region of the porcine opn (osteopontin, secreted phosphoprotein 1) gene. Eur J Biochem. 207:649–659. 26. Noda M, Vogel RL, Craig AM, Prahl J, DeLuca HF, Denhardt DT 1990 Identification of a DNA sequence responsible for binding of the 1,25-dihydroxyvitamin D3 receptor and 1,25-dihydroxyvitamin D3 enhancement of mouse secreted phosphoprotein 1 (Spp-1 or osteopontin) gene expression. Proc Natl Acad Sci USA 87:9995–9999. 27. Quelo I, Machuca I, Jurdic P 1998 Identification of a vitamin D response element in the proximal promoter of the chicken carbonic anhydrase II gene. J Biol Chem 273: 10638–10646. 28. Chen KS, DeLuca HF 1995 Cloning of the human 1α,25dihydroxyvitamin D3 24-hydroxylase gene promoter and identification of two vitamin D-responsive elements. Biochem Biophys Acta 1263:1–9.
323 29. Zierold C, Darwish HM, DeLuca HF 1994 Identification of a vitamin D-response element in the rat calcidiol (25-hydroxyvitamin D3) 24-hydroxylase gene. Proc Natl Acad Sci USA 91:900–902. 30. Taketani Y, Segawa H, Chikamori M, Morita K, Tanaka K, Kido S, Yamamoto H, Iemori Y, Tatsumi S, Tsugawa N, Okano T, Kobayashi T, Miyamoto K, Takeda E 1998 Regulation of type II renal Na+-dependent inorganic phosphate transporters by 1,25-dihydroxyvitamin D3: identification of a vitamin Dresponsive element in the human NaPi-3 gene. J Biol Chem 273:14575–14581. 31. Demay MB, Gerardi JM, DeLuca HF, Kronenberg, HM 1990 DNA sequences in the rat osteocalcin gene that bind the 1,25-dihydroxyvitamin D3 receptor and confer responsiveness to 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 87:369–373. 32. Demay MB, Kieran MS, DeLuca HF, Kronenberg HM 1992 Sequences in the human parathyroid hormone gene that bind the 1,25-dihydroxyvitamin D3 receptor and mediate transcriptional repression in response to 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 89:8097–8101. 33. Kremer R, Sebag M, Champigny C, Meerovitch K, Hendy GN, White J, Goltzman D 1996 Identification and characterization of 1,25-dihydroxyvitamin D3-responsive repressor sequences in the rat parathyroid hormone-related peptide gene. J Biol Chem 271:16310–16316. 34. Toell A, Polly P, Carlberg C 2000 All natural DR3-type vitamin D response elements show a similar functionality in vitro. Biochem J 352:301–309. 35. Falzon M 1996 DNA sequences in the rat parathyroid hormone-related peptide gene responsible for 1,25-dihydroxyvitamin D3-mediated transcriptional repression. Mol Endocrinol 10:672–681. 36. Darwish HM, DeLuca HF 1992 Identification of a 1,25dihydroxyvitamin D3-response element in the 5′-flanking region of the rat calbindin D9k gene. Proc Natl Acad Sci USA 89:603–607. 37. Wang GF, Nikovits W, Schleinitz M, Stockdale FE 1998 A positive GATA element and a negative vitamin D receptorlike element control atrial chamber-specific expression of a slow myosin heavy-chain gene during cardiac morphogenesis. Mol Cell Biol 18:6023–6034. 38. Alonso M, Segura C, Dieguez C, Perez-Fernandez R 1998 High-affinity binding site to the vitamin D receptor DNA binding domain in the human growth hormone promoter. Biochem Biophys Res Commun 247:882–887. 39. Cao X, Ross FP, Zhang L, MacDonald PN, Chappel J, Teitelbaum SL 1993 Cloning of the promoter for the avian integrin β3 subunit gene and its regulation by 1,25-dihydroxyvitamin D3. J Biol Chem 268:27371–27380. 40. Liu SM, Koszewski N, Lupez M, Malluche HH, Olivera A, Russell J 1996 Characterization of a response element in the 5′-flanking region of the avian (chicken) PTH gene that mediates negative regulation of gene transcription by 1,25dihydroxyvitamin D3 and binds the vitamin D3 receptor. Mol Endocrinol 10:206–215. 41. Liu M, Lee MH, Cohen M, Bommakanti M, Freedman LP 1996 Transcriptional activation of the Cdk inhibitor p21 by vitamin D3 leads to the induced differentiation of the myelomonocytic cell line U937. Genes Dev 10:142–153. 42. Schräder M, Müller KM, Carlberg C 1994 Specificity and flexibility of vitamin D signaling. Modulation of the activation of natural vitamin D response elements by thyroid hormone. J Biol Chem 269:5501–5504.
324
CARSTEN CARLBERG, THOMAS W. DUNLOP, CHRISTIAN FRANK, AND SAMI VÄISÄNEN
43. Kerry DM, Dwivedi PP, Hahn CN, Morris HA, Omdahl JL, May BK 1996 Transcriptional synergism between vitamin Dresponsive elements in the rat 25-hydroxyvitamin D3 24-hydroxylase (CYP) promoter. J Biol Chem 271:29715–29721. 44. Staal A, van Wijnen AJ, Birkenhäger JC, Pols HAP, Prahl J, DeLuca HF, Gaub MP, Lian JB, Stein GS, van Leeuwen JPTM, Stein JL 1996 Distinct conformations of vitamin D receptor/ retinoid X receptor-a heterodimers are specified by dinucleotide differences in the vitamin D-responsive elements of the osteocalcin and osteopontin genes. Mol Endocrinol 10:1444–1456. 45. Wood JR, Greene GL, Nardulli AM 1998 Estrogen response elements function as allosteric modulators of estrogen receptor conformations. Mol Cell Biol 18:1927–1934. 46. Herdick M, Bury Y, Quack M, Uskokovic M, Polly P, Carlberg C 2000 Response element- and coactivator-mediated conformational change of the vitamin D3 receptor permits sensitive interaction with agonists. Mol Pharmacol 57:1206–1217. 47. Carlberg C 1995 Mechanisms of nuclear signalling by vitamin D3. Interplay with retinoid and thyroid hormone signalling. Eur J Biochem 231:517–527. 48. Rhodes SJ, Chen R, DiMattia GE, Scully KM, Kalla KA, Lin SC, Yu VC, Rosenfeld MG 1993 A tissue-specific enhancer confers Pit-1-dependent morphogen inducibility and autoregulation on the pit-1 gene. Genes Dev 7:913–932. 49. Mangelsdorf DJ, Evans RM 1995 The RXR heterodimers and orphan receptors. Cell 83:841–850. 50. Quack M, Carlberg C 2000 Ligand-triggered stabilization of vitamin D receptor/retinoid X receptor heterodimer conformations on DR4-type response elements. J Mol Biol 296: 743–756. 51. Morrison NA, Shine J, Fragonas JC, Verkest V, McMenemey ML, Eisman JA 1989 1,25-Dihydroxyvitamin D-responsive element and glucocorticoid repression in the osteocalcin gene. Science 246:1158–1161. 52. Ozono K, Liao J, Kerner SA, Scott RA, Pike JW 1990 The vitamin D–responsive element in the human osteocalcin gene. Association with a nuclear proto-oncogene enhancer. J Biol Chem 265:21881–21888. 53. Schräder M, Bendik I, Becker-Andre M, Carlberg C 1993 Interaction between retinoic acid and vitamin D signaling pathways. J Biol Chem 268:17830–17836. 54. Polly P, Carlberg C, Eisman JA, Morrison NA 1996 Identification of a vitamin D3 response element in the fibronectin gene that is bound by a vitamin D3 receptor homodimer. J Cell Biochem 60:322–333. 55. Kahlen JP, Carlberg C 1994 Identification of a vitamin D receptor homodimer-type response element in the rat calcitriol 24-hydroxylase gene promoter. Biochem Biophys Res Commun 202:1366–1372. 56. Xie Z, Bikle DD 1997 Cloning of the human phospholipase C-γ1 promoter and identification of a DR6-type vitamin Dresponsive element. J Biol Chem 272:6573–6577. 57. Schräder M, Kahlen JP, Carlberg C 1997 Functional characterization of a novel type of 1α,25-dihydroxyvitamin D3 response element identified in the mouse c-fos promoter. Biochem Biophys Res Commun 230:646–651. 58. Schräder M, Müller KM, Becker-André M, Carlberg C 1994 Response element selectivity for heterodimerization of vitamin D receptors with retinoic acid and retinoid X receptors. J Mol Endocrinol 12:327–339. 59. Schräder M, Müller KM, Nayeri S, Kahlen JP, Carlberg C 1994 VDR-T3R receptor heterodimer polarity directs ligand sensitivity of transactivation. Nature 370:382–386. 60. Carlberg C 1996 The concept of multiple vitamin D signaling pathways. J Investig Dermatol Symp Proc 1:10–14.
61. Nayeri S, Danielsson C, Kahlen JP, Schräder M, Mathiasen IS, Binderup L, Carlberg C 1995 The anti-proliferative effect of vitamin D3 analogues is not mediated by inhibition of the AP-1 pathway, but may be related to promoter selectivity. Oncogene 11:1853–1858. 62. Danielsson C, Mathiasen IS, James SY, Nayeri S, Bretting C, Hansen CM, Colston KW, Carlberg C 1997 Sensitive induction of apoptosis in breast cancer cells by a novel 1,25-dihydroxyvitamin D3 analogue shows relation to promoter selectivity. J Cell Biochem 66:552–562. 63. Mørk Hansen C, Mäenpää PH 1997 EB1089—a novel vitamin D analog with strong antiproliferative and differentiation inducing effects on target cells. Biochem Pharmacol. 54:1173–1179. 64. Quack M, Carlberg C 1999 Selective recognition of vitamin D receptor conformations mediates promoter selectivity of vitamin D analogs. Mol Pharmacol 55:1077–1087. 65. Carlberg C, Hooft van Huijsduijnen R, Staple JK, DeLamarter JF, Becker-Andre M 1994 RZRs, a new family of retinoid-related orphan receptors that function as both monomers and homodimers. Mol Endocrinol 8:757–770. 66. Schräder M, Becker-Andre M, Carlberg C 1994 Thyroid hormone receptor functions as monomeric ligand-induced transcription factor on octameric half-sites. Consequences also for dimerization. J Biol Chem 269:6444–6449. 67. Mader S, Leroy P, Chen JY, Chambon P 1993 Multiple parameters control the selectivity of nuclear receptors for their response elements. J Biol Chem 268:591–600. 68. Quack M, Frank C, Carlberg C 2002 Differential nuclear receptor signalling from DR4-type response elements. J Cell Biochem 86:601–612. 69. Cheskis B, Freedman LP 1994 Ligand modulates the conversion of DNA-bound vitamin D3 receptor (VDR) homodimers into VDR-retinoid X receptor heterodimers. Mol Cell Biol 14:3329–3338. 70. Schräder M, Wyss A, Sturzenbecker LJ, Grippo JF, LeMotte P, Carlberg C 1993 RXR-dependent and RXR-independent transactivation by retinoic acid receptors. Nucleic Acids Res 21:1231–1237. 71. Takeshita A, Ozawa Y, Chin WW 2000 Nuclear receptor coactivators facilitate vitamin D receptor homodimer action on direct repeat hormone response elements. Endocrinology 141:1281–1284. 72. Robinson-Rechavi M, Garcia HE, Laudet V 2003 The nuclear receptor superfamily. J Cell Sci 116:585–586. 73. Castillo AI, Jimenez-Lara AM, Tolon RM, Aranda A 1999 Synergistic activation of the prolactin promoter by vitamin D receptor and GHF-1: role of coactivators, CREB-binding protein and steroid hormone receptor coactivator-1 (SRC-1). Mol Endocrinol 13:1141–1154. 74. Sierra J, Villagra A, Paredes R, Cruzat F, Gutierrez S, Javed A, Arriagada G, Olate J, Imschenetzky M, Van Wijnen AJ, Lian JB, Stein GS, Stein JL, Montecino M 2003 Regulation of the bonespecific osteocalcin gene by p300 requires Runx2/Cbfa1 and the vitamin D3 receptor but not p300 intrinsic histone acetyltransferase activity. Mol Cell Biol 23:339–3351. 75. Schüle R, Umesono K, Mangelsdorf DJ, Bolado J, Pike JW, Evans RM 1990 Jun-Fos and receptors for vitamins A and D recognize a common response element in the human osteocalcin gene. Cell 61:497–504. 76. Miner JN, Yamamoto KR 1992 The basic region of AP-1 specifies glucocorticoid receptor activity at a composite response element. Genes Dev 6:491–2501. 77. Candeliere GA, Jurutka PW, Haussler MR, St-Arnaud R 1996 A composite element binding the vitamin D receptor, retinoid X receptor α, and a member of the CTF/NF-1 family of
325
CHAPTER 18 Diversity of Vitamin D Target Genes
78.
79. 80.
81.
82.
83.
84.
85. 86.
87.
transcription factors mediates the vitamin responsiveness of the c-fos promoter. Mol Cell Biol 16:584–592. Kim RH, Li JJ, Ogata Y, Yamauchi M, Freedman LP, Sodek J 1996 Identification of a vitamin D3-response element that overlaps a unique inverted TATA box in the rat bone sialoprotein gene. Biochem J 318:219–226. Zierold C, Darwish HM, DeLuca HF 1995 Two vitamin D response elements function in the rat 1,25-dihydroxyvitamin D 24-hydroxylase promoter. J Biol Chem 270:1675–1678. Honkakoski P, Zelko I, Sueyoshi T, Negishi M 1998 The nuclear orphan receptor CAR-retinoid X receptor heterodimer activates the phenobarbital-responsive enhancer module of the CYP2B gene. Mol Cell Biol 18:5652–5658. Lehmann JM, McKee DD, Watson MA, Willson TM, Moore JT, Kliewer SA 1998 The human orphan nuclear receptor PXR is activated by compounds that regulate CYP3A4 gene expression and cause drug interactions. J Clin Invest 102:1016–1023. Sugatani J, Kojima H, Ueda A, Kakizaki S, Yoshinari K, Gong QH, Owens IS, Negishi M, Sueyoshi T 2001 The phenobarbital response enhancer module in the human bilirubin UDP-glucuronosyltransferase UGT1A1 gene and regulation by the nuclear receptor CAR. Hepatology 33:1232–1238. Drocourt L, Ourlin JC, Pascussi JM, Maurel P, Vilarem MJ 2002 Expression of CYP3A4, CYP2B6, and CYP2C9 is regulated by the vitamin D receptor pathway in primary human hepatocytes. J Biol Chem 277:25125–25132. Thummel KE, Brimer C, Yasuda K, Thottassery J, Senn T, Lin Y, Ishizuka H, Kharasch E, Schuetz J, Schuetz E 2001 Transcriptional control of intestinal cytochrome P-4503A by 1α,25-dihydroxy vitamin D3. Mol Pharmacol 60:1399–1406. Makishima M, Lu TT, Xie W, Whitfield GK, Domoto H, Evans RM, Haussler MR, Mangelsdorf DJ 2002 Vitamin D receptor as an intestinal bile acid sensor. Science 296:1313–1316. Towers TL, Freedman LP 1998 Granulocyte-macrophage colonystimulating factor gene transcription is directly repressed by the vitamin D3 receptor: implications for allosteric influences on nuclear receptor structure and function by a DNA element. J Biol Chem 273:10338–10348. Towers TL, Staeva TP, Freedman LP 1999 A two-hit mechanism for vitamin D3-dediates transcriptional repression of the
88. 89.
90.
91.
92.
93.
94.
95.
96.
granulocyte-macrophage colony-stimulating factor gene: vitamin D receptor completes for DNA binding with NFAT1 and stabilizes c-Jun. Mol Cell Biol 19:4191–4199. Polly P, Carlberg C, Eisman JA, Morrison NA 1997 1α,25Dihydroxyvitamin D3 receptor as a mediator of transrepression of retinoid signaling. J Cell Biochem 67:287–296. Mackey SL, Heymont JL, Kronenberg HM, Demay MB 1996 Vitamin D receptor binding to the negative human parathyroid hormone response element does not require the retinoid X receptor. Mol Endocrinol 10:298–305. Nishishita T, Okazaki T, Ishikawa T, Igarashi T, Hata K, Ogata E, Fujita T 1998 A negative vitamin D response DNA element in the human parathyroid hormone-related peptide gene binds to vitamin D receptor along with Ku antigen to mediate negative gene regulation by vitamin D. J Biol Chem 273: 10901–10907. Chen S, Costa CHRM, Nakamura K, Ribero RCJ, Gardner DG 1999 Vitamin D–dependent suppression of human atrial natriuretic peptic gene promoter activity requires heterodimer assembly. J Biol Chem 274:11260–11266 Okazaki T, Nishimori S, Ogata E, Fujita T 2003 Vitamin D– dependent recruitment of DNA-PK to the chromatinized negative vitamin D response element in the PTHrP gene is required for gene repression by vitamin D. Biochem Biophys Res Commun 304:632–637. Peleg S, Abruzzese RV, Cooper CW, Gagel RF 1993 Downregulation of calcitonin gene transcription by vitamin D requires two widely separated enhancer sequences. Mol Endocrinol 7:1999–1008. Podvinec M, Kaufmann MR, Handschin C, Meyer UA 2002 NUBIScan, an in silico approach for prediction of nuclear receptor response elements. Mol Endocrinol 16: 1269–1279. Gill RK, Christakos S 1993 Identification of sequence elements in mouse calbindin-D28k gene that confer 1,25-dihydroxyvitamin D3- and butyrate-inducible responses. Proc Natl Acad Sci USA 90:2984–2988. Carlberg C 1997 Critical analysis of 1α,25-dihydroxyvitamin D3 response elements. In: Norman AW, Bouillon B, Thomasset M (eds) Proc 10th Workshop on Vitamin D, pp. 268–275.
CHAPTER 19
Intranuclear Organization of the Regulatory Machinery for Vitamin D–Mediated Control of Skeletal Gene Expression GARY S. STEIN, JANE B. LIAN, MARTIN MONTECINO,* JANET L. STEIN, ANDRE J. VAN WIJNEN, AMJAD JAVED, JE-YONG CHOI,** S. KALEEM ZAIDI, SORAYA GUTIERREZ,* JIALI SHEN, SHIRWIN POCKWINSE, AND DANIEL YOUNG Department of Cell Biology and Cancer Center, University of Massachusetts Medical School, Worcester, Massachusetts; ∗Departamento de Biologia Molecular, Universidad de Concepcion, Concepcion, Chile; ∗∗Department of Biochemistry, Kyungpook National University, Daegu, Korea I. Introduction II. Requirements for Physiologically Responsive Control of Skeletal Gene Expression In Vivo III. Gene Expression within the Three-Dimensional Context of Nuclear Architecture IV. Chromatin Remodeling Facilitates Vitamin D–Mediated Promoter Accessibility and Integration of Regulatory Activities V. Nuclear Microenvironments: Accommodating The Rules That Govern In Vivo Transcriptional Control
VI. Scaffolding of Regulatory Elements for Combinatorial Control of Gene Expression VII. Intranuclear Trafficking of Skeletal Regulatory Factors to Subnuclear Sites That Support Transcription: “To Be in the Right Place at the Right Time” VIII. The Regulated and Regulatory Parameters of Subnuclear Organization References
I. INTRODUCTION
factors abrogate competency for vitamin D control of skeletal gene expression during development and fidelity of gene expression in tumor cells.
The architecturally associated subnuclear organization of nucleic acids and cognate regulatory factors illustrates functional interrelationships between nuclear structure and gene expression. Mechanisms that contribute to the spatial distribution of transcription factors within the dynamic three-dimensional context of nuclear architecture control the sorting of regulatory information and the transcriptionally competent or repressed chromatin configuration of gene promoters as well as the assembly and activities of sites within the nucleus that support gene expression. Vitamin D control of gene expression serves as a paradigm for experimentally addressing mechanisms that govern the intranuclear targeting of regulatory factors to nuclear domains where chromatin remodeling and transcription of developmental as well as tissue-specific genes occur. We will provide an overview of molecular, cellular, biochemical, and in vivo genetic approaches that provide insight into the trafficking of regulatory factors that mediate vitamin D–controlled gene expression to transcriptionally active subnuclear sites. Examples will be presented that suggest modifications in the intranuclear targeting of transcription VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
II. REQUIREMENTS FOR PHYSIOLOGICALLY RESPONSIVE CONTROL OF SKELETAL GENE EXPRESSION IN VIVO Bone formation during development and skeletal remodeling throughout life requires the complex and interdependent expression of cell growth and phenotypic genes (reviewed in [1,2]). There is a requirement for responsiveness to a broad spectrum of regulatory cues that transduce physiological signals from the extracellular matrix to sites within the nucleus where genes that mediate skeletogenesis reside [3–6]. As our understanding of gene regulatory mechanisms expands, it becomes increasingly evident that there are unique parameters of transcriptional control that support the transient activation and suppression of genes for skeletal development and bone homeostasis. Other mechanisms Copyright © 2005, Elsevier, Inc. All rights reserved.
328 are invoked for long-term obligations to gene expression that sustain the specialized properties of the bone cells. Vitamin D serves as a principal modulator of skeletal gene transcription necessitating an understanding of interfaces between activity of this steroid hormone with regulatory cascades that are functionally linked to the regulation of skeletal genes [7]. There is growing appreciation for the repertoire of factors that influence gene expression for commitment to the osteoblast lineage. It is well documented that sequentially expressed genes support progression of osteoblast differentiation through developmental transition points where responsiveness to phosphorylationmediated regulatory cascades determine competency for establishing and maintaining the structural and functional properties of bone cells [2,4,8,9]. The catalog of promoter elements and cognate regulatory proteins that govern skeletal gene expression offers essential but insufficient insight into mechanisms that are operative in intact cells. Gene promoters serve as regulatory infrastructure by functioning as blueprints for responsiveness to the flow of cellular regulatory signals. But to access the specific genetic information necessitates understanding transcriptional control of skeletal genes within the context of the subnuclear organization of nucleic acids and regulatory proteins. Explanations are required for (1) convergence of multiple regulatory signals at promoter sequences; (2) the integration of regulatory information at independent promoter domains; (3) selective utilization of redundant regulatory pathways; (4) thresholds for initiation or down-regulation of transcription with limited intranuclear representation of promoter elements and regulatory factors; (5) mechanisms that render the promoters of cell growth and phenotypic genes competent for protein–DNA and protein–protein interactions in a physiologically responsive manner; (6) the composition, organization, and assembly of sites within the nucleus that support transcription; and (7) the intranuclear trafficking of regulatory proteins to transcriptionally active foci.
III. GENE EXPRESSION WITHIN THE THREE-DIMENSIONAL CONTEXT OF NUCLEAR ARCHITECTURE Evidence is accumulating that the architectural organization of nucleic acids and regulatory proteins within the nucleus support functional interrelationships between nuclear structure and gene expression (Fig. 1 and 2). There is increasing acceptance that components of nuclear architecture are functionally linked to the organization and sorting of regulatory information in a manner that permits selective utilization [10–20]. The primary
GARY S. STEIN,
ET AL .
level of nuclear organization, the representation and ordering of genes and promoter elements, provides alternatives for physiological control. The molecular organization of regulatory elements, the overlap of regulatory sequences within promoter domains, and the multipartite composition of regulatory complexes increase options for responsiveness. Chromatin structure and nucleosome organization reduce distances between regulatory sequences, facilitate crosstalk between promoter elements, and render elements competent for interactions with positive and negative regulatory factors [21]. The components of higher order nuclear architecture that includes nuclear pores [22,23], the nuclear matrix, and subnuclear domains contribute to the subnuclear distribution and activities of genes and regulatory factors (reviewed in [24–27]). Compartmentalization of regulatory complexes is illustrated by focal organization of PML bodies [28–30], Runx bodies [11,12,16,31,32], and the nucleolus, chromosomes [33], as well as by the punctate intranuclear distribution of sites for replication [34–36], DNA repair, transcription [37–42], and the processing of gene transcripts [26,43,44]. There is emerging recognition that nuclear structure and function are causally interrelated. With mounting evidence for organization of nucleic acids and regulatory proteins into subnuclear domains that are associated with components of nuclear architecture, the perception of a dichotomy between nuclear architecture and control of gene expression is difficult to justify. Rather, it is necessary to design experiments to define mechanisms that direct genes and regulatory factors to sites within the nucleus where localization integrates regulatory parameters of gene expression and establishes microenvironments with boundaries between regulatory complexes that are required for fidelity of activity. The bone-specific osteocalcin gene and skeletalrestricted Runx2 (AML3/Cbfa1/PEBP24) transcription factor serve as paradigms for obligatory relationships between nuclear structure with physiological control of skeletal gene expression [32,45–48]. The modularly organized promoter of the bone-specific osteocalcin gene contains proximal and distal regulatory elements that support basal, and tissue-specific as well as growth factor, homeodomain, signaling protein, and steroid hormone responsive transcriptional control (reviewed in [45,46,49–55]) (see Fig. 3). Modulation of osteocalcin gene expression during bone formation and remodeling requires physiologically responsive accessibility of these proximal and upstream promoter sequences to regulatory and coregulatory proteins as well as protein–protein interactions that integrate independent promoter domains. The nuclear matrixassociated Runx transcription factors contribute to the control of skeletal gene expression by sequence-specific
CHAPTER 19 Intranuclear Organization of the Regulatory Machinery for Vitamin D
329
FIGURE 1 Multiple levels of chromatin architecture within the nucleus. Higher order chromatin organization in the nucleus results from sequential packaging of DNA from a linear double helix (upper panel). The sequence of the Runx binding element is shown. A loop domain emanating from the highly compact chromatin is schematically illustrated (lower panel). Elements at the base of the loop structure designated MARs [matrix attachment regions, or alternatively locus control regions (LCR) or scaffold attachment regions (SCA)] mediate association of these genomic domains with the nuclear scaffold. Genes within the loop domain undergo local chromatin remodeling to support transcriptional activation or suppression. The nuclear matrix provides anchorage for both nucleic acids and regulatory as well as co-regulatory factors that control transcription.
binding to promoter elements of target genes and serving as scaffolds for the assembly and organization of co-regulatory proteins that mediate biochemical and architectural control of promoter activity.
IV. CHROMATIN REMODELING FACILITATES VITAMIN D–MEDIATED PROMOTER ACCESSIBILITY AND INTEGRATION OF REGULATORY ACTIVITIES It is well recognized that genomic DNA is packaged as chromatin. These “bead on a string” structures designated nucleosomes are structurally remodeled to accommodate requirements for transcription, emphasizing the extent to which architectural organization of genes is causally related to functional activity. The
identification and characterization of proteins that catalyze histone acetylation, deacetylation, methylation, and phosphorylation [56–63], as well as the SWI/SNFrelated proteins [56,64–68] that facilitate chromatin remodeling and potentially the accessibility of promoter sequences to regulatory and coregulatory factors, represent an important dimension in control of the structural and functional activities of genes and promoter regulatory elements [64,69–75]. Relationships of regulatory signaling pathways to enhance activities that modulate gene, chromatin, and chromosome organization can now be directly investigated. Additional levels of specificity are provided by structural modifications of gene promoters that influence competency for factor interactions. Simply stated, changes in the architectural properties of promoter elements determine effectiveness of gene regulatory sequences as substrates for interactions with regulatory factors. The regulatory and regulated parameters of chromatin
330
GARY S. STEIN,
ET AL .
Structural components
Apoptosis
Survivin
Replication sites
RPA
Chromosomal territories
Chromosomes
Nuclear envelope
PML bodies
Nucleoli
Transcription SWI/SNF complex sites
BRCA1
CAF-1
Replication and repair
RUNX domains
VDR SC 35 Domains
Coiled bodies
TLE
Transcription
Splicing FIGURE 2 Components of nuclear architecture are functionally linked to the organization and sorting of regulatory information. Nuclear functions are organized into distinct, nonoverlapping subnuclear domains. Nuclear matrix (shown in the center as revealed by electron microscopy), the underlying network of anastomizing network of filaments and fibers provides structural basis for the functional compartmentalization of nuclear functions. Immunofluorescence microscopy of the nucleus has revealed the distinct subnuclear distribution of vital nuclear processes, including (but not limited to) DNA replication sites and proteins involved in replication such as CAF-1 and RPA; DNA damage/repair as shown by BRCA1; chromatin remodeling, e.g., mediated by the SWI/SNF complex; structural parameters of the nucleus, such as the nuclear envelope, chromosomes, and chromosomal territories; chromatin organization and tissue specific transcriptional control, for example Runx, TLE, and VDR domains; and RNA synthesis and processing involving, for example, transcription sites; SC35 domains, coiled bodies, and nucleoli as well as proteins involved in cell survival e.g., survivin. Subnuclear PML bodies of unknown function have been examined in numerous cell types. All these domains are associated with the nuclear matrix. (See color plate).
remodeling and the rate limiting steps in the relevant signaling cascades are being actively pursued and will unquestionably provide insight into skeletal gene regulatory mechanisms from structural and functional perspectives. The chromatin organization of the osteocalcin gene illustrates dynamic remodeling of a promoter to accommodate requirements for phenotype-related developmental and steroid hormone responsive activity.
Nuclease digestion and ligation-mediated PCR analysis as well as in vitro nucleosome reconstitution studies establish the placement of nucleosomes in the proximal basal/tissue specific domain and at the upstream vitamin D–responsive element, blocking accessibility of these promoter sequences to regulatory proteins in immature bone cells when this skeletal-restricted gene is suppressed [60,76–80]. In response to developmental and skeletal regulatory signals the striking
CHAPTER 19 Intranuclear Organization of the Regulatory Machinery for Vitamin D
A
Inactive Osteocalcin Gene
B
Basal Transcription OC Gene Distal DHS
C
331
Proximal DHS
Vitamin D Induced Transcription of OC Gene Distal DHS
D
Proximal DHS
Proximal DNasel HS
Nuclear Matrix Distal DNasel HS
FIGURE 3 Remodeling of the Osteocalcin gene promoter during developmental progression of the osteoblast phenotype. The transcriptionally silent rat Osteocalcin gene is schematically illustrated with nucleosomes placed in the proximal tissue-specific and distal enhancer region of the promoter (A). Factors that support basal tissue-specific transcription are recruited to the OC gene promoter and are organized in proximal and distal promoter domains. Modifications in chromatin structure that mediate assembly of the regulatory machinery for the nuclease hypersensitive sites reflect OC gene transcription. A positioned nucleosome resides between the proximal basal and distal enhancer regions of the promoter (B). In response to vitamin D, chromatin remodeling renders the upstream VDRE competent for binding the VDR/RXR heterodimer with its cognate element (C). Higher order chromatin organization permits cross talk between basal transcription machinery and the vitamin D receptor complex that involves direct interactions of the vitamin D receptor, Runx2, and TFIIB (D).
removal of a nucleosome and modifications in chromatin structure render the proximal promoter of the OC gene accessible to regulatory and co-regulatory proteins that support basal level activity [48,48,60,78]. Vitamin D enhancement of osteocalcin gene transcription is associated with removal of the nucleosome at the upstream vitamin D–responsive element that permits binding of the vitamin D receptor–RXR heterodimer [48,60,78,81]. The retention of a nucleosome
between the proximal and upstream enhancer domain reduces distance between the basal and vitamin D–responsive element and supports a promoter configuration that is conducive to protein–protein interactions between the vitamin D receptor and the basal TFIIB transcription factor [82–84]. Interaction of the vitamin D receptor at the distal promoter region of the bone-specific osteocalcin gene requires nucleosomal remodeling [85].
332 Thus, insight into control of skeletal gene expression can be obtained from an understanding of mechanisms that alter osteocalcin gene chromatin organization under biological conditions. Site-directed mutagenesis of osteocalcin genes that are genetically integrated in stable cell lines has established that RUNX elements flanking the proximal and upstream promoter sequences are responsible for developmental and vitamin D– induced chromatin remodeling [48]. The recent demonstration of functional interactions between the VDR and flanking Runx proteins is consistent with linkage of VDRE organization and Runx regulatory elements [86]. Reduced CpG methylation is associated with transcriptional activation of the bone-specific osteocalcin gene in osteoblasts [87]. In vitro and in vivo genetic approaches have demonstrated that RUNX2 controls developmental and steroid hormone–responsive chromatin reconfiguration of the osteocalcin gene promoter [48,80]. Chromatin immunoprecipitation analyses have shown that developmental and vitamin D–linked remodeling of osteocalcin gene promoter organization is accompanied by acetylation of histones in the proximal basal and upstream vitamin D responsive element domains [88,89]. This posttranslational modification of histone proteins reduces the tenacity of histone DNA interactions in a manner that is conducive to an open chromatin organization with increased access to regulatory factors. The most compelling evidence for a functional involvement of chromatin organization in skeletal gene expression is the obligatory relationship of dynamic changes in the biochemical and structural properties of osteocalcin gene promoter organization with competency for bone tissue-restricted and enhanced transcription in response to vitamin D [48]. Yet, despite the cogent support for a central role of chromatin remodeling in transcriptional control of the osteocalcin gene, there are open-ended questions. It is not justifiable to extrapolate from these findings to conclude that all genes that are activated and suppressed during skeletogenesis employ identical mechanisms. From a broader biological perspective there are multiple levels of control that must be mechanistically characterized to explain physiologically responsive regulation of chromatin structure within restricted and global genomic contexts.
V. NUCLEAR MICROENVIRONMENTS: ACCOMMODATING THE RULES THAT GOVERN IN VIVO TRANSCRIPTIONAL CONTROL Key components of the basal transcription machinery and several tissue-specific transcription factor complexes
GARY S. STEIN,
ET AL .
are functionally compartmentalized as specialized subnuclear domains [11–13,16,27,39,40,90–102]. Such compartmentalization may, at least in part, accommodate biological constraints on the control of transcription in nuclei of intact bone cells. The low representation of promoter regulatory elements and cognate transcription factors necessitates a subnuclear organization of nucleic acids and regulatory proteins that supports threshold concentrations for the activation and repression of gene expression. From an historical perspective, compartmentalization of the regulatory machinery for ribosomal genes in nucleoli and the organization of chromosomes during mitosis provide paradigms for intranuclear localization of genes and regulatory complexes. During the past several years there has been growing recognition that the organization of nucleic acids and regulatory proteins is functionally linked to the assembly, organization, and activity of gene regulatory machinery. Cellular, molecular, biochemical, and genetic evidence indicates an obligatory relationship between sites within the nucleus where regulatory complexes reside and fidelity of transcriptional control. The biological relevance for the intranuclear distribution of regulatory complexes is directly reflected by aberrant nuclear structure–gene expression interrelationships that are associated with perturbations in skeletal development [18] and leukemia [16].
VI. SCAFFOLDING OF REGULATORY COMPONENTS FOR COMBINATORIAL CONTROL OF GENE EXPRESSION Functional interrelationships between nuclear structure and gene expression are strikingly reflected by dual recognition of regulatory proteins, such as RUNX transcription factors, for interactions with both promoter elements and co-regulatory proteins; such interactions modulate the structural and functional properties of targeted genes at microenvironments within the nucleus. Sequence-specific interactions with promoter elements result in placement of RUNX proteins at strategic sites where they provide scaffolds for protein–protein interactions that mediate the organization of machinery for a broad spectrum of regulatory requirements. These include histone modifications and chromatin remodeling that establish competency for transcription factor binding and genomic conformations that interface activities at proximal and upstream promoter domains, as well as the integration of regulatory cues from signaling pathways that activate or suppress gene expression in a physiologically responsive manner. As a consequence, the RUNX proteins are posttranslationally modified
CHAPTER 19 Intranuclear Organization of the Regulatory Machinery for Vitamin D
(e.g., phosphorylated) to further influence the extent to which they engage in regulatory activity. The complexity of RUNX regulatory proteins that assemble as supercomplexes of transcriptional regulatory factors illustrates the potential impact on skeletalrelated gene expression. Recent documentation that RUNX proteins are components of a stable complex that includes basal transcription factors, chromatin remodeling factors, and histone-modifying factors indicates the scope of RUNX-mediated combinatorial control. A key component of the RUNX complex is the p300/CBP coactivator that functions as a transcriptional adaptor. Interactions with several transcription factors results in the formation of multimolecular complexes that regulate expression of a broad spectrum of genes [103]. p300 contains a domain with intrinsic histone acetyltransferase (HAT) activity [104,105], which has been implicated in chromatin structure alterations associated with modulation of gene expression [106]. p300 interacts with additional proteins containing HAT activity that include P/CAF, SRC-1, and ACTR. A basis is thereby provided for formation of large multiprotein complexes that contribute multiple HAT activities with options for specificity [107–111]. It has been established that RUNX2 and p300 are components of the same nuclear complexes in osteoblastic cells [112]. Furthermore, when recruited to the osteocalcin gene promoter by RUNX2, p300 stimulates both basal and vitamin D-enhanced osteocalcin promoter activity. Thus interactions of RUNX2 with p300 supports assembly of multisubunit complexes with several HAT-containing proteins at a series of regulatory regions of the bone-specific osteocalcin gene promoter. In a parallel manner, Kitabayashi et al. [113] have shown that in myeloid cells RUNX1, a homolog of the bonespecific RUNX2, interacts with p300 and together they up-regulate myeloid-specific genes. It was also determined that a C-terminal region of the Runt domain in both RUNX1 and RUNX2, is critical for their interactions with p300 [112,113]. Considering the high degree of homology between these two members of the RUNX transcription factor family it is likely that the structural determinants for RUNX interactions with p300 are conserved. In addition to functioning as transcriptional activators, RUNX proteins suppress gene expression (transcription). Repression requires the recruitment of transcriptional repressors and co-repressors with histone deacetylase activity (HDACs) to promoter regulatory elements of genes that are down-regulated. Combinatorial control that dampens transcription is illustrated by interaction of RUNX2 with the transcriptional co-repressors TLE/ Groucho through a conserved VWRPY domain located at the C terminus of the protein, which represses the
333
expression of the bone-sialo protein (BSP) gene in osteoblastic cells [32]. Another example of combinatorial control that results in transcriptional suppression by RUNX2 is down-regulation of the p21CIP/WAF promoter in fibroblastic and osteoblastic cells. Here HDAC6 interacts with a second repression domain that also resides in the C-terminal region of RUNX2 and is recruited to chromatin by RUNX2 [114]. Taken together, these results are consistent with combinatorial control that is mediated by RUNX-dependent recruitment of coactivator and co-repressors proteins that are associated with and organized as multiprotein complexes to activate or repress target genes in a physiologically responsive manner. p300 can also be recruited to gene promoters by the transcription factor C/EBP [115,116]. Interestingly, a C/EBP-responsive regulatory element has been identified in the proximal promoter region of the rat OC gene adjacent to the RUNX2 site C [117]. C/EBPβ physically interacts with RUNX2 and synergistically activates the osteocalcin promoter [117], suggesting that both proteins form a complex with p300 and together up-regulate basal tissue-specific transcription. C/EBPβ has additionally been shown to interact with ATPdependent chromatin remodeling complexes of the SWI/SNF family [118], recruiting these complexes to promoter sequences and activating cell-specific expression. Taken together, these findings indicate that RUNX factors engage in protein–DNA and protein–protein interactions that collectively determine the composition and organization of promoter regulatory complexes. The inclusion of chromatin remodeling activity in these multisubunit complexes provides a biochemical basis for conformational modifications of promoter elements as well as combinatorial specificity for transcription. Transcription factors that function as scaffolds for interaction with co-regulatory proteins provide an architectural basis for accommodating the combinatorial requirements of biological control. Combinatorial control supports replication, transcription, and repair by two mechanisms. Context-dependent combinations and permutations of regulatory proteins are assembled into multipartite complexes that increase specificity. Scaffold-associated protein–DNA and protein–protein interactions permit integration of regulatory activities. Nuclear microenvironments are thereby organized, with gene promoters as focal points, where threshold concentrations of regulatory macromolecules are attained. The complexity that is achieved by these architecturally organized oligomeric factors can maximize options for responsiveness to diverse regulatory requirements for transient and long term biological control.
334 VII. INTRANUCLEAR TRAFFICKING OF SKELETAL REGULATORY FACTORS TO SUBNUCLEAR SITES THAT SUPPORT TRANSCRIPTION: ‘‘TO BE IN THE RIGHT PLACE AT THE RIGHT TIME’’ There is a need to gain insight into mechanisms that direct skeletal factors to subnuclear sites where regulatory events occur. Association of osteoblast, myeloid, and lymphoid RUNX transcription factors that mediate tissue-specific transcription with the nuclear matrix has permitted direct examination of mechanisms for targeting regulatory proteins to transcriptionally active subnuclear domains [12,45–47,95,119–126]. Both biochemical and immunofluorescence analyses have shown that RUNX transcription factors exhibit a punctate nuclear distribution that is associated with the nuclear matrix in situ [11,12,127,128]. Taken together, these observations are consistent with the concept that the nuclear matrix is functionally involved in gene localization and in the concentration and subnuclear localization of regulatory factors [12,93,94,129–132]. The initial indication that nuclear matrix association of RUNX factors is required for maximal activity was provided by the observation that transcriptionally active RUNX proteins associate with the nuclear matrix but inactive C-terminally truncated RUNX proteins do not [6,12,32,128,133] (Fig. 3). This localization of RUNX was established by biochemical fractionation and in situ immunofluorescence as well as by green fluorescent protein tagged RUNX proteins [31] in living cells. Colocalization of RUNX1, 2, and 3 at nuclear matrix–associated sites indicates a common intranuclear targeting mechanism may be operative for the family of RUNX transcription factors [31,32,127]. Variations in the partitioning of transcriptionally active and inactive RUNX between subnuclear fractions permitted development of a strategy to identify a region of the RUNX transcription factors that directs the regulatory proteins to nuclear matrix–associated foci. A series of deletions and internal mutations was constructed and assayed for competency to associate with the nuclear matrix by Western blot analysis of biochemically prepared nuclear fractions and by in situ immunostaining following transfection into intact cells. Association of osteogenic and hematopoietic RUNX proteins with the nuclear matrix is independent of DNA binding and requires a nuclear matrix targeting signal, a 31-amino-acid segment near the C terminus that is distinct from nuclear localization signals [12]. The nuclear matrix targeting signal functions autonomously and is necessary as well as sufficient to direct the transcriptionally
GARY S. STEIN,
ET AL .
active RUNX transcription factors to nuclear matrix– associated sites where gene expression occurs [12]. These findings indicate mechanisms involved in the selective trafficking of proteins to specialized domains within the nucleus where they become components of functional regulatory complexes. At least two trafficking signals appear to be required for subnuclear targeting of RUNX transcription factors; the first supports nuclear import (nuclear localization signal) and a second mediates association with the nuclear matrix (nuclear matrix targeting signal). The multiplicity of determinants for nuclear localization and alternative splicing of RUNX messenger RNA may provide the requisite complexity to support targeting to specific sites within the nucleus in response to diverse biological conditions. Furthermore, because gene expression by RUNX involves contributions by factors and co-regulatory proteins that include CBFβ [46,134–139] and C/EBP [117,140], Groucho/TLE [32,141,142], HES, and SMAD [6,143], RUNX may facilitate recruitment of these factors to the nuclear matrix.
A. Properties of Transcriptionally Active Subnuclear Compartments Association of genes and cognate factors with the nuclear matrix may support the formation and/or activities of nuclear domains that facilitate transcriptional control [10,45,55,132,144–151]. Results from our laboratory indicate that the association of RUNX transcription factors with the nuclear matrix is obligatory for activity [11,18]. The promoter recognition function of the runt homology domain of RUNX, and thus the consequential interactions with RUNX-responsive genes, is essential for formation of transcriptionally active foci containing RUNX and RNA polymerase II that are nuclear matrix associated [11]. Additionally, the nuclear matrix targeting signal supports transactivation when associated with an appropriate promoter, and transcriptional activity of the nuclear matrix targeting signal depends on association with the nuclear matrix [11]. Taken together, targeting of RUNX transcription factors to the nuclear matrix is important for their function and transcription. However, components of the nuclear matrix that function as acceptor sites remain to be established. Characterization of such nuclear matrix components will provide an additional dimension to characterizing molecular mechanisms associated with gene expression—the targeting of regulatory proteins to specific spatial domains within the nucleus. An initial indication of transcription factor interactions with the nuclear matrix is provided by crystal structure of the
CHAPTER 19 Intranuclear Organization of the Regulatory Machinery for Vitamin D
RUNX nuclear matrix targeting signal that was determined by X-ray diffraction analysis at 2.7 Å [127,152].
B. Subnuclear Targeting and Integration of Signaling Pathways Gene expression during skeletal development and bone remodeling is controlled by a broad spectrum of regulatory signals that converge at promoter elements to activate or repress transcription in a physiologically responsive manner. The subnuclear compartmentalization of transcription machinery necessitates a mechanistic explanation for directing signaling factor to sites within the nucleus where gene expression occurs under conditions that support integration of regulatory cues. The interactions of YAP and SMAD co-regulatory proteins with C-terminal segments of the RUNX2 transcription factor permit assessment of requirements for recruitment of cSRC and BMP/TGFb-mediated signals to skeletal target genes. Our findings indicate that nuclear import of YAP and SMAD co-regulatory factors is agonist dependent. However, there is a stringent requirement for fidelity of RUNX subnuclear targeting for recruitment of these signaling proteins to transcriptionally active subnuclear foci. Our results demonstrate that the interactions and spatial–temporal organization of RUNX and SMAD as well as YAP co-regulatory proteins are essential for assembly of transcription machinery that supports expression of skeletal genes [6,128]. Competency for intranuclear trafficking of RUNX proteins has similarly been functionally linked with the subnuclear localization and activity of TLE/Groucho co-regulatory proteins [32]. These findings are consistent with proteins serving as a scaffold for interactions with co-regulatory proteins that contribute to biological control.
C. In Vivo Consequences of Aberrant Intranuclear Trafficking of RUNX Transcription Factors Using RUNX2 and its essential role in osteogenesis as a model, we investigated the fundamental importance of fidelity of subnuclear localization for tissue differentiating activity by deleting the intranuclear targeting signal via homologous recombination. Mice homozygous for the deletion (RUNX2∆C) do not form bone because of perturbed maturation or arrest of osteoblasts. Heterozygotes do not develop clavicles, but are otherwise normal. These phenotypes are
335
indistinguishable from those of the RUNX2 homozygous and heterozygous null mutants, indicating that the intranuclear targeting signal is a critical determinant for function. The expressed truncated RUNX2∆C protein enters the nucleus and retains normal DNA-binding activity, but shows complete loss of intranuclear targeting. These results establish that the multifunctional N-terminal region of the RUNX2 protein is not sufficient for biological activity. Our results demonstrate that subnuclear localization of RUNX factors in specific foci together with associated regulatory functions is essential for control of RUNXdependent genes involved in tissue differentiation during embryonic development [18]. The importance of subnuclear localization of RUNX transcription factors for biological control is further indicated by compromised subnuclear organization and activity of RUNX1 hematopoietic regulatory proteins in acute myelogenous leukemia [16].
VIII. THE REGULATED AND REGULATORY PARAMETERS OF SUBNUCLEAR ORGANIZATION Multiple lines of evidence suggest that components of nuclear architecture contribute both structurally and enzymatically to control gene expression during osteoblast differentiation. Sequences have been identified that direct RUNX transcription factors to nuclear matrix–associated sites that support transcription in a cell cycle–dependent manner [153]. Insight is thereby provided into mechanisms linked to the assembly and activities of subnuclear domains where transcription occurs. In a restricted sense, the foundation has been provided for experimentally addressing intranuclear trafficking of gene regulatory factors and control of association with the nuclear matrix to establish and sustain domains that are competent for transcription. The unique sequences [11,12] and crystal structure for the 31-amino-acid nuclear matrix targeting signal of RUNX transcription factors [127,152] support specificity for localization at intranuclear sites where the regulatory machinery for gene expression is assembled, rendered operative, and/or suppressed. In a broader context, there is a growing appreciation for involvement of nuclear architecture in a dynamic and bidirectional exchange of gene transcripts and regulatory factors between the nucleus and cytoplasm, as well as between regions and structures within the nucleus [13,27,35,154–156]. It would be presumptuous to propose a single model to account for the specific pathways that direct transcription
336 factors to sites within the nucleus that support transcription. However, findings suggest that parameters of nuclear architecture functionally interface with components of transcriptional control. The involvement of nuclear matrix–associated transcription factors with recruitment of regulatory components to modulate transcription remains to be defined. Working models that serve as frameworks for experimentally addressing components of transcriptional control within the context of nuclear architecture can be compatible with mechanisms that involve architecturally or activitydriven assembly of transcriptionally active intranuclear foci. The diversity of targeting signals must be established to evaluate the extent to which regulatory discrimination is mediated by encoded intranuclear trafficking signals. It will additionally be important to biochemically and mechanistically define the checkpoints, which are operative during subnuclear distribution of regulatory factors, and the editing steps, which are invoked to ensure that structural and functional fidelity of nuclear domains, where replication and expression of genes occur. There is emerging recognition that placement of regulatory components of gene expression must be temporally and spatially coordinated to optimally mediate biological control. It is realistic to anticipate that further understanding of mechanisms that position genes and regulatory factors for establishment and maintenance of the bone cell phenotype will clarify nuclear structure–function interrelationships that are operative during osteoblast differentiation and vitamin D modulation of regulatory activity.
GARY S. STEIN,
4. 5.
6. 7.
8. 9.
10. 11.
12.
13. 14.
Acknowledgments 15.
Results presented in this chapter were in part supported by grants for the National Institutes of Health (AR45688, PO1CA82834, DE12528, AR39588, AR45689, PO1AR48818, FIRCA R03 TW00990, FONDECYT 1030479/DK32520). The authors appreciate the editorial assistance of Elizabeth Bronstein and Karen Concaugh with the preparation of this manuscript.
16.
17.
18.
References 1. Bilezikian JP, Raisz LG, Rodan GA 2002 Principles of Bone Biology. Academic Press, San Diego. 2. Stein GS, Lian JB, Montecino M et al. 2002 Involvement of nuclear architecture in regulating gene expression in bone cells. In: Bilezikian JP, Raisz LG, Rodan GA (eds) Principles of Bone Biology. 1. Academic Press, San Diego, pp. 169–188. 3. Stein GS, van Wijnen AJ, Stein JL, Lian JB, Montecino M 1997 Nuclear architecture in developmental transcriptional
19. 20. 21.
ET AL .
control of cell growth and tissue-specific genes. In: Bird RC, Stein GS, Lian JB, Stein JL (eds) Nuclear Structure and Gene Expression. Academic Press, New York, pp. 178–214. Lian JB, Stein GS 1999 The cells of bone. In: Seibel MJ, Robins SP, Bilezikian JP (eds) Dynamics of Bone and Cartilage Metabolism. Academic Press, San Diego, pp. 165–185. Xiao G, Gopalkrishnan RV, Jiang D et al. 2002 Bone morphogenetic proteins, extracellular matrix, and mitogenactivated protein kinase signaling pathways are required for osteoblast-specific gene expression and differentiation in MC3T3-E1 cells. J Bone Miner Res 17:101–110. Zaidi SK, Sullivan AJ, van Wijnen AJ et al. 2002 Integration of Runx and Smad regulatory signals at transcriptionally active subnuclear sites. Proc Natl Acad Sci USA 99:8048–8053. Lian JB, Staal A, van Wijnen AJ, Stein JL, Stein GS 1999 Biologic and molecular effects of vitamin D on bone. In: Holick MF (ed) Vitamin D: Molecular Biology, Physiology, and Clinical Applications. Humana Press, Totowa, NJ, pp. 175–193. Aubin JE, Liu F 1996 The osteoblast lineage. In: Bilezikian JP, Raisz LG, Rodan GA (eds) Principles of Bone Biology. Academic Press, San Diego, pp. 51–68. Schinke T, Karsenty G 2002 Transcriptional control of osteoblast differentiation and function. In: Bilezikian JP, Raisz LG, Rodan GA (eds) Principles of Bone Biology. 1. Academic Press, San Diego, pp. 83–91. Berezney R, Mortillaro M, Ma H et al. 1996 Connecting nuclear architecture and genomic function. J Cell Biochem 62:223–226. Zeng C, McNeil S, Pockwinse S et al. 1998 Intranuclear targeting of AML/CBFα regulatory factors to nuclear matrixassociated transcriptional domains. Proc Natl Acad Sci USA 95:1585–1589. Zeng C, van Wijnen AJ, Stein JL et al. 1997 Identification of a nuclear matrix targeting signal in the leukemia and bonerelated AML/CBFα transcription factors. Proc Natl Acad Sci USA 94:6746–6751. Lamond AI, Earnshaw WC 1998 Structure and function in the nucleus. Science 280:547–553. Ma H, Siegel AJ, Berezney R 1999 Association of chromosome territories with the nuclear matrix. Disruption of human chromosome territories correlates with the release of a subset of nuclear matrix proteins. J Cell Biol 146:531–542. Ma H, Samarabandu J, Devdhar RS et al. 1998 Spatial and temporal dynamics of DNA replication sites in mammalian cells. J Cell Biol 143:1415–1425. McNeil S, Zeng C, Harrington KS et al. 1999 The t(8;21) chromosomal translocation in acute myelogenous leukemia modifies intranuclear targeting of the AML1/CBFalpha2 transcription factor. Proc Natl Acad Sci USA 96:14882–14887. McNeil S, Guo B, Stein JL et al. 1998 Targeting of the YY1 transcription factor to the nucleolus and the nuclear matrix in situ: the C-terminus is a principal determinant for nuclear trafficking. J Cell Biochem 68:500–510. Choi J-Y, Pratap J, Javed A et al. 2001 Subnuclear targeting of Runx/Cbfa/AML factors is essential for tissue-specific differentiation during embryonic development. Proc Natl Acad Sci USA 98:8650–8655. DeFranco DB 2002 Navigating steroid hormone receptors through the nuclear compartment. Mol Endocrinol 16: 1449–1455. Gasser SM 2002 Visualizing chromatin dynamics in interphase nuclei. Science 296:1412–1416. Strahl BD, Grant PA, Briggs SD et al. 2002 Set2 is a nucleosomal histone H3-selective methyltransferase that mediates transcriptional repression. Mol Cell Biol 22:1298–1306.
CHAPTER 19 Intranuclear Organization of the Regulatory Machinery for Vitamin D
22. Iborra FJ, Jackson DA, Cook PR 2000 The path of RNA through nuclear pores: apparent entry from the sides into specialized pores [In Process Citation]. J Cell Sci 113 (Pt 2): 291–302. 23. Mattaj IW, Englmeier L 1998 Nucleocytoplasmic transport: the soluble phase. Annu Rev Biochem 67:265–306. 24. Berezney R, Jeon KW 1995 Structural and Functional Organization of the Nuclear Matrix. Academic Press, New York. 25. Penman S 1995 Rethinking cell structure. Proc Natl Acad Sci USA 92:5251–5257. 26. Misteli T 2000 Cell biology of transcription and pre-mRNA splicing: nuclear architecture meets nuclear function. J Cell Sci 113:1841–1849. 27. Stein GS, van Wijnen AJ, Stein JL et al. 2000 Intranuclear trafficking of transcription factors: implications for biological control. J Cell Sci 113:2527–2533. 28. Dyck JA, Maul GG, Miller WH et al. 1994 A novel macromolecular structure is a target of the promyelocyte-retinoic acid receptor oncoprotein. Cell 76:333–343. 29. Grande MA, van der Kraan I, van Steensel B et al. 1996 PML-containing nuclear bodies: their spatial distribution in relation to other nuclear components. J Cell Biochem 63:280–291. 30. Melnick A, Licht JD 1999 Deconstructing a disease: RARalpha, its fusion partners, and their roles in the pathogenesis of acute promyelocytic leukemia. Blood 93:3167–3215. 31. Harrington KS, Javed A, Drissi H et al. 2002 Transcription factors RUNX1/AML1 and RUNX2/Cbfa1 dynamically associate with stationary subnuclear domains. J Cell Sci 115:4167–4176. 32. Javed A, Guo B, Hiebert S et al. 2000 Groucho/TLE/R-Esp proteins associate with the nuclear matrix and repress RUNX (CBFα/AML/PEBP2α) dependent activation of tissuespecific gene transcription. J Cell Sci 113:2221–2231. 33. Ma H, Siegel AJ, Berezney R 1999 Association of chromosome territories with the nuclear matrix. Disruption of human chromosome territories correlates with the release of a subset of nuclear matrix proteins. J Cell Biol 146:531–542. 34. Leonhardt H, Rahn HP, Cardoso MC 1998 Intranuclear targeting of DNA replication factors. J Cell Biochem Suppl 30–31:243–249. 35. Wei X, Samarabandu J, Devdhar RS et al. 1998 Segregation of transcription and replication sites into higher order domains. Science 281:1502–1505. 36. Cook PR 1999 The organization of replication and transcription. Science 284:1790–1795. 37. Ciejek EM, Tsai MJ, O’Malley BW 1983 Actively transcribed genes are associated with the nuclear matrix. Nature 306: 607–609. 38. van Steensel B, Jenster G, Damm K, Brinkmann AO, van Driel R 1995 Domains of the human androgen receptor and glucocorticoid receptor involved in binding to the nuclear matrix. J Cell Biochem 57:465–478. 39. Htun H, Barsony J, Renyi I, Gould DL, Hager GL 1996 Visualization of glucocorticoid receptor translocation and intranuclear organization in living cells with a green fluorescent protein chimera. Proc Natl Acad Sci USA 93:4845–4850. 40. Stenoien D, Sharp ZD, Smith CL, Mancini MA 1998 Functional subnuclear partitioning of transcription factors. J Cell Biochem 70:213–221. 41. Kimura H, Tao Y, Roeder RG, Cook PR 1999 Quantitation of RNA polymerase II and its transcription factors in an HeLa cell: little soluble holoenzyme but significant amounts of polymerases attached to the nuclear substructure. Mol Cell Biol 19:5383–5392.
337
42. Verschure PJ, van Der Kraan I, Manders EM, van Driel R 1999 Spatial relationship between transcription sites and chromosome territories. J Cell Biol 147:13–24. 43. Misteli T, Spector DL 1999 RNA polymerase II targets pre-mRNA splicing factors to transcription sites in vivo. Mol Cell 3:697–705. 44. Smith KP, Moen PT, Wydner KL, Coleman JR, Lawrence JB 1999 Processing of endogenous pre-mRNAs in association with SC-35 domains is gene specific. J Cell Biol 144: 617–629. 45. Merriman HL, van Wijnen AJ, Hiebert S et al. 1995 The tissue-specific nuclear matrix protein, NMP-2, is a member of the AML/CBF/PEBP2/runt domain transcription factor family: interactions with the osteocalcin gene promoter. Biochemistry 34:13125–13132. 46. Banerjee C, Hiebert SW, Stein JL, Lian JB, Stein GS 1996 An AML-1 consensus sequence binds an osteoblast-specific complex and transcriptionally activates the osteocalcin gene. Proc Natl Acad Sci USA 93:4968–4973. 47. Ducy P, Zhang R, Geoffroy V, Ridall AL, Karsenty G 1997 Osf2/Cbfa1: a transcriptional activator of osteoblast differentiation. Cell 89:747–754. 48. Javed A, Gutierrez S, Montecino M et al. 1999 Multiple Cbfa/AML sites in the rat osteocalcin promoter are required for basal and vitamin D responsive transcription and contribute to chromatin organization. Mol Cell Biol 19:7491–7500. 49. Demay MB, Gerardi JM, DeLuca HF, Kronenberg HM 1990 DNA sequences in the rat osteocalcin gene that bind the 1,25-dihydroxyvitamin D3 receptor and confer responsive to 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 87:369–373. 50. Markose ER, Stein JL, Stein GS, Lian JB 1990 Vitamin Dmediated modifications in protein-DNA interactions at two promoter elements of the osteocalcin gene. Proc Natl Acad Sci USA 87:1701–1705. 51. Bortell R, Owen TA, Bidwell JP et al. 1992 Vitamin Dresponsive protein-DNA interactions at multiple promoter regulatory elements that contribute to the level of rat osteocalcin gene expression. Proc Natl Acad Sci USA 89:6119–6123. 52. Hoffmann HM, Catron KM, van Wijnen AJ et al. 1994 Transcriptional control of the tissue-specific, developmentally regulated osteocalcin gene requires a binding motif for the Msx family of homeodomain proteins. Proc Natl Acad Sci USA 91:12887–12891. 53. Tamura M, Noda M 1994 Identification of a DNA sequence involved in osteoblast-specific gene expression via interaction with helix–loop–helix (HLH)-type transcription factors. J Cell Biol 126:773–782. 54. Ducy P, Karsenty G 1995 Two distinct osteoblast-specific cis-acting elements control expression of a mouse osteocalcin gene. Mol Cell Biol 15:1858–1869. 55. Guo B, Odgren PR, van Wijnen AJ et al. 1995 The nuclear matrix protein NMP-1 is the transcription factor YY1. Proc Natl Acad Sci USA 92:10526–10530. 56. Workman JL, Kingston RE 1998 Alteration of nucleosome structure as a mechanism of transcriptional regulation. Annu Rev Biochem 67:545–579. 57. Zhang Y, Sun ZW, Iratni R et al. 1998 SAP30, a novel protein conserved between human and yeast, is a component of a histone deacetylase complex. Mol Cell 1:1021–1031. 58. Ayer DE 1999 Histone deacetylases: transcriptional repression with SINers and NuRDs. Trends Cell Biol 9:193–198. 59. Kornberg RD, Lorch Y 1999 Chromatin-modifying and -remodeling complexes. Curr Opin Genet Dev 9:148–151.
338 60. Montecino M, Frenkel B, van Wijnen AJ et al. 1999 Chromatin hyperacetylation abrogates vitamin D-mediated transcriptional upregulation of the tissue-specific osteocalcin gene in vivo. Biochemistry 38:1338–1345. 61. Safadi FF, Rane SG, Gentile MA et al. 1999 Cancellous osteopenia in mice lacking cyclin dependent kinase-4 (CDK-4). J Bone Miner Res 14:S164. 62. Agalioti T, Lomvardas S, Parekh B et al. 2000 Ordered recruitment of chromatin modifying and general transcription factors to the IFN-beta promoter. Cell 103:667–678. 63. Hassan AH, Neely KE, Workman JL 2001 Histone acetyltransferase complexes stabilize swi/snf binding to promoter nucleosomes. Cell 104:817–827. 64. Imbalzano AN 1998 Energy-dependent chromatin remodelers: complex complexes and their components. Crit Rev Eukaryot Gene Expr 8:225–255. 65. Lorch Y, Cairns BR, Zhang M, Kornberg RD 1998 Activated RSC-nucleosome complex and persistently altered form of the nucleosome. Cell 94:29–34. 66. Peterson CL, Zhao Y, Chait BT 1998 Subunits of the yeast SWI/SNF complex are members of the actin-related protein (ARP) family. J Biol Chem 273:23641–23644. 67. Schnitzler G, Sif S, Kingston RE 1998 Human SWI/SNF interconverts a nucleosome between its base state and a stable remodeled state. Cell 94:17–27. 68. de la Serna I, Carlson KA, Imbalzano AN 2001 Mammalian SWI/SNF complexes promote MyoD-mediated muscle differentiation. Nat Genet 27:187–190. 69. de la Serna I, Imbalzano AN 2002 Unfolding heterochromatin for replication. Nat Genet 32:560–562. 70. Carrozza MJ, Utley RT, Workman JL, Cote J 2003 The diverse functions of histone acetyltransferase complexes. Trends Genet 19:321–329. 71. Neely KE, Workman JL 2002 Histone acetylation and chromatin remodeling: which comes first? Mol Genet Metab 76:1–5. 72. Fan HY, He X, Kingston RE, Narlikar GJ 2003 Distinct strategies to make nucleosomal DNA accessible. Mol Cell 11:1311–1322. 73. Narlikar GJ, Fan HY, Kingston RE 2002 Cooperation between complexes that regulate chromatin structure and transcription. Cell 108:475–487. 74. Jaskelioff M, Peterson CL 2003 Chromatin and transcription: histones continue to make their marks. Nat Cell Biol 5:395–399. 75. Peterson CL 2003 Transcriptional activation: getting a grip on condensed chromatin. Curr Biol 13:R195–R197. 76. Breen EC, van Wijnen AJ, Lian JB, Stein GS, Stein JL 1994 In vivo occupancy of the vitamin D responsive element in the osteocalcin gene supports vitamin D–dependent transcriptional upregulation in intact cells. Proc Natl Acad Sci USA 91:12902–12906. 77. Montecino M, Lian J, Stein G, Stein J 1994 Specific nucleosomal organization supports developmentally regulated expression of the osteocalcin gene [abstract]. J Bone Miner Res 9:S352. 78. Montecino M, Lian J, Stein G, Stein J 1996 Changes in chromatin structure support constitutive and developmentally regulated transcription of the bone-specific osteocalcin gene in osteoblastic cells. Biochemistry 35:5093–5102. 79. Paredes R, Gutierrez J, Gutierrez S et al. 2002 Interaction of the 1alpha,25-dihydroxyvitamin D3 receptor at the distal promoter region of the bone-specific osteocalcin gene requires nucleosomal remodelling. Biochem J 363:667–676. 80. Gutierrez J, Sierra J, Medina R et al. 2000 Interaction of CBFα/AML/PEBP2α transcription factors with nucleosomal
GARY S. STEIN,
81.
82.
83. 84.
85.
86.
87.
88.
89.
90. 91. 92.
93.
94.
95.
96.
ET AL .
sequences requires flexibility in the translational positioning of the histone octamer and exposure of the Cbfα site. Biochemistry 39:13565–13574. Paredes R, Gutierrez J, Gutierrez S et al. 2002 Interaction of the 1alpha,25-dihydroxyvitamin D3 receptor at the distal promoter region of the bone-specific osteocalcin gene requires nucleosomal remodelling. Biochem J 363:667–676. Blanco JCG, Wang I-M, Tsai SY et al. 1995 Transcription factor TFIIB and the vitamin D receptor cooperatively activate ligand-dependent transcription. Proc Natl Acad Sci USA 92:1535–1539. MacDonald PN, Sherman DR, Dowd DR, Jefcoat SC, Jr., DeLisle RK 1995 The vitamin D receptor interacts with general transcription factor IIB. J Biol Chem 270:4748–4752. Guo B, Aslam F, van Wijnen AJ et al. 1997 YY1 regulates VDR/RXR mediated transactivation of the vitamin D responsive osteocalcin gene. Proc Natl Acad Sci USA 94: 121–126. Paredes R, Gutierrez J, Gutierrez S et al. 2002 Interaction of the 1alpha,25-dihydroxyvitamin D3 receptor at the distal promoter region of the bone-specific osteocalcin gene requires nucleosomal remodelling. Biochem J 363:667–676. Paredes R, Arriagada G, Cruzat F, Villagra A, Olate J, Zaidi SK et al. 2003 The Runx2 transcription factor interacts with the 1α,25-dihydroxy vitamin D3 receptor and facilitates 1α,25-dihydroxy vitamin D3–dependent upregulation of the rat OC gene expression in osteoblastic cells. Unpublished. Villagra A, Gutierrez J, Paredes R et al. 2002 Reduced CpG methylation is associated with transcriptional activation of the bone-specific rat osteocalcin gene in osteoblasts. J Cell Biochem 85:112–122. Shen J, Montecino MA, Lian JB et al. 2002 Histone acetylation in vivo at the osteocalcin locus is functionally linked to vitamin D dependent, bone tissue-specific transcription. J Biol Chem 277:20284–20292. Shen J, Hovhannisyan H, Lian JB et al. 2003 Transcriptional induction of the osteocalcin gene during osteoblast differentiation involves acetylation of histones h3 and h4. Mol Endocrinol 17:743–756. Robinson SI, Nelkin BD, Vogelstein B 1982 The ovalbumin gene is associated with the nuclear matrix of chicken oviduct cells. Cell 28:99–106. Stief A, Winter DM, Stratling WH, Sippel AE 1989 A nuclear DNA attachment element mediates elevated and positionindependent gene activity. Nature 341:343–345. Schaack J, Ho WY, Freimuth P, Shenk T 1990 Adenovirus terminal protein mediates both nuclear matrix association and efficient transcription of adenovirus DNA. Genes Dev 4:1197–1208. Dworetzky SI, Wright KL, Fey EG et al. 1992 Sequencespecific DNA-binding proteins are components of a nuclear matrix-attachment site. Proc Natl Acad Sci USA 89: 4178–4182. van Wijnen AJ, Bidwell JP, Fey EG et al. 1993 Nuclear matrix association of multiple sequence-specific DNA binding activities related to SP-1, ATF, CCAAT, C/EBP, OCT-1, and AP-1. Biochemistry 32:8397–8402. Banerjee C, McCabe LR, Choi J-Y et al. 1997 Runt homology domain proteins in osteoblast differentiation: AML-3/CBFA1 is a major component of a bone specific complex. J Cell Biochem 66:1–8. Bangs P, Burke B, Powers C et al. 1998 Functional analysis of Tpr: identification of nuclear pore complex association and nuclear localization domains and a role in mRNA export. J Cell Biol 143:1801–1812.
CHAPTER 19 Intranuclear Organization of the Regulatory Machinery for Vitamin D
97. Ma H, Samarabandu J, Devdhar RS et al. 1998 Spatial and temporal dynamics of DNA replication sites in mammalian cells. J Cell Biol 143:1415–1425. 98. Tang Y, Getzenberg RH, Vietmeier BN et al. 1998 The DNAbinding and τ2 transactivation domains of the rat glucocorticoid receptor constitute a nuclear matrix targeting signal. Mol Endocrinol 12:1420–1431. 99. Misteli T, Spector DL 1999 RNA polymerase II targets pre-mRNA splicing factors to transcription sites in vivo. Mol Cell 3:697–705. 100. Wei X, Somanathan S, Samarabandu J, Berezney R 1999 Three-dimensional visualization of transcription sites and their association with splicing factor-rich nuclear speckles. J Cell Biol 146:543–558. 101. Kimura H, Tao Y, Roeder RG, Cook PR 1999 Quantitation of RNA polymerase II and its transcription factors in an HeLa cell: little soluble holoenzyme but significant amounts of polymerases attached to the nuclear substructure. Mol Cell Biol 19:5383–5392. 102. DeFranco DB 2002 Navigating steroid hormone receptors through the nuclear compartment. Mol Endocrinol 16: 1449–1455. 103. Goodman RH, Smolik S 2000 CBP/p300 in cell growth, transformation, and development. Genes Dev 14:1553–1577. 104. Bannister AJ, Oehler T, Wilhelm D, Angel P, Kouzarides T 1995 Stimulation of c-Jun activity by CBP: c-Jun residues Ser63/73 are required for CBP induced stimulation in vivo and CBP binding in vitro. Oncogene 11:2509–2514. 105. Ogryzko VV, Schiltz RL, Russanova V, Howard BH, Nakatani Y 1996 The transcriptional coactivators p300 and CBP are histone acetyltransferases. Cell 87:953–959. 106. Spencer VA, Davie JR 1999 Role of covalent modifications of histones in regulating gene expression. Gene 240:1–12. 107. Chakravarti D, LaMorte VJ, Nelson MC et al. 1996 Role of CBP/P300 in nuclear receptor signalling. Nature 383:99–103. 108. Yang XJ, Ogryzko VV, Nishikawa J, Howard BH, Nakatani Y 1996 A p300/CBP-associated factor that competes with the adenoviral oncoprotein E1A. Nature 382:319–324. 109. Chen H, Lin RJ, Schiltz RL et al. 1997 Nuclear receptor coactivator ACTR is a novel histone acetyltransferase and forms a multimeric activation complex with P/CAF and CBP/p300. Cell 90:569–580. 110. Spencer TE, Jenster G, Burcin MM et al. 1997 Steroid receptor coactivator-1 is a histone acetyltransferase. Nature 389: 194–198. 111. Torchia J, Rose DW, Inostroza J et al. 1997 The transcriptional co-activator p/CIP binds CBP and mediates nuclearreceptor function. Nature 387:677–684. 112. Sierra J, Villagra A, Paredes R et al. 2003 Regulation of the bone-specific osteocalcin gene by p300 requires Runx2/Cbfa1 and the vitamin D3 receptor but not p300 intrinsic histone acetyl transferase activity. Mol Cell Biol 23:3339–3351. 113. Kitabayashi I, Yokoyama A, Shimizu K, Ohki M 1998 Interaction and functional cooperation of the leukemiaassociated factors AML1 and p300 in myeloid cell differentiation. EMBO J 17:2994–3004. 114. Westendorf JJ, Zaidi SK, Cascino JE et al. 2002 Runx2 (Cbfa1, AML-3) interacts with histone deacetylase 6 and represses the p21(CIP1/WAF1) promoter. Mol Cell Biol 22:7982–7992. 115. Oelgeschlager M, Janknecht R, Krieg J, Schreek S, Luscher B 1996 Interaction of the co-activator CBP with Myb proteins: effects on Myb-specific transactivation and on the cooperativity with NF-. EMBO J 15:2771–2780.
339
116. Mink S, Haenig B, Klempnauer KH 1997 Interaction and functional collaboration of p300 and C/EBPbeta. Mol Cell Biol 17:6609–6617. 117. Gutierrez S, Javed A, Tennant D et al. 2002 CCAAT/ enhancer-binding proteins (C/EBP) β and δ Activate osteocalcin gene transcription and synergize with Runx2 at the C/EBP element to regulate bone-specific expression. J Biol Chem 277:1316–1323. 118. Kowenz-Leutz E, Leutz A 1999 A C/EBP beta isoform recruits the SWI/SNF complex to activate myeloid genes. Mol Cell 4:735–743. 119. Bae SC, Yamaguchi-Iwai Y, Ogawa E et al. 1993 Isolation of PEBP2αB cDNA representing the mouse homolog of human acute myeloid leukemia gene, AML1. Oncogene 8:809–814. 120. Meyers S, Downing JR, Hiebert SW 1993 Identification of AML-1 and the t(8;21) translocation protein (AML-1/ETO) as sequence-specific DNA-binding proteins; the runt homology domain is required for DNA binding and protein–protein interactions. Mol Cell Biol 13:6336–6345. 121. Meyers S, Lenny N, Sun W-H, Hiebert SW 1996 AML-2 is a potential target for transcriptional regulation by the t(8;21) and t(12;21) fusion proteins in acute leukemia. Oncogene 13:303–312. 122. Meyers S, Lenny N, Hiebert SW 1995 The t(8;21) fusion protein interferes with AML-1B-dependent transcriptional activation. Mol Cell Biol 15:1974–1982. 123. Wang S, Wang Q, Crute BE et al. 1993 Cloning and characterization of subunits of the T-cell receptor and murine leukemia virus enhancer core-binding factor. Mol Cell Biol 13:3324–3339. 124. Nuchprayoon I, Meyers S, Scott LM et al. 1994 PEBP2/CBF, the murine homolog of the human myeloid AML1 and PEBP2β/CBFβ proto-oncoproteins, regulates the murine myeloperoxidase and neutrophil elastase genes in immature myeloid cells. Mol Cell Biol 14:5558-5568. 125. Frank R, Zhang J, Uchida H et al. 1995 The AML1/ETO fusion protein blocks transactivation of the GM-CSF promoter by AML1B. Oncogene 11:2667–2674. 126. Satake M, Nomura S, Yamaguchi-Iwai Y et al. 1995 Expression of the Runt domain-encoding PEBP2 alpha genes in T cells during thymic development. Mol Cell Biol 15:1662–1670. 127. Tang L, Guo B, van Wijnen AJ et al. 1998 Preliminary crystallographic study of the glutathione S-transferase fused with the nuclear matrix targeting signal of the transcription factor AML-1/CBFα2. J Struct Biol 123:83–85. 128. Zaidi SK, Sullivan AJ, van Wijnen AJ et al. 2002 Functional integration of RUNX2 and BMP2 responsive regulatory signals at transcriptionally active subnuclear sites in osseous and non-osseous cells [abstract]. J Bone Miner Res 17:5439. 129. Bidwell JP, van Wijnen AJ, Fey EG et al. 1993 Osteocalcin gene promoter-binding factors are tissue-specific nuclear matrix components. Proc Natl Acad Sci USA 90:3162–3166. 130. Blencowe BJ, Nickerson JA, Issner R, Penman S, Sharp PA 1994 Association of nuclear matrix antigens with exoncontaining splicing complexes. J Cell Biol 127:593–607. 131. Mancini MA, Shan B, Nickerson JA, Penman S, Lee WH 1994 The retinoblastoma gene product is a cell cycle-dependent, nuclear matrix-associated protein. Proc Natl Acad Sci USA 91:418–422. 132. Nickerson JA, Blencowe BJ, Penman S 1995 The architectural organization of nuclear metabolism. In: Berezney R, Jeon KW (eds) Structural and Functional Organization of the Nuclear Matrix.162A. Academic Press, New York, pp. 67–123.
340 133. Choi J-Y, Javed A, Zaidi SK et al. 1999 Structure-function determinants for subnuclear targeting and transactivation in the Cbfa1/AML3 transcription factor [abstract]. J Bone Miner Res 14:S469. 134. Ogawa E, Inuzuka M, Maruyama M et al. 1993 Molecular cloning and characterization of PEBP2β, the heterodimeric partner of a novel Drosophila runt-related DNA binding protein PEBP2α. Virology 194:314–331. 135. Giese K, Kingsley C, Kirshner JR, Grosschedl R 1995 Assembly and function of a TCR alpha enhancer complex is dependent on LEF-1-induced DNA bending and multiple protein-protein interactions. Genes Dev 9:995–1008. 136. Mao S, Frank RC, Zhang J, Miyazaki Y, Nimer SD 1999 Functional and physical interactions between AML1 proteins and an ETS protein, MEF: implications for the pathogenesis of t(8;21)-positive leukemias. Mol Cell Biol 19:3635–3644. 137. Xie XQ, Pardali E, Holm M, Sideras P, Grundstrom T 1999 AML and Ets proteins regulate the I alpha1 germ-line promoter. Eur J Immunol 29:488–498. 138. Miller J, Horner A, Stacy T et al. 2002 The core-binding factor beta subunit is required for bone formation and hematopoietic maturation. Nat Genet 32:645–649. 139. Kundu M, Javed A, Jeon JP et al. 2002 Cbfbeta interacts with Runx2 and has a critical role in bone development. Nat Genet 32:639–644. 140. Zhang DE, Hetherington CJ, Meyers S et al. 1996 CCAAT enhancer-binding protein (C/EBP) and AML1 (CBFα2) synergistically activate the macrophage colony-stimulating factor receptor promoter. Mol Cell Biol 16:1231–1240. 141. Levanon D, Goldstein RE, Bernstein Y et al. 1998 Transcriptional repression by AML1 and LEF-1 is mediated by the TLE/Groucho corepressors. Proc Natl Acad Sci USA 95:11590–11595. 142. Javed A, Barnes GL, Jassanya BO et al. 2001 runt homology domain transcription factors (Runx, Cbfa, and AML) mediate repression of the bone sialoprotein promoter: evidence for promoter context-dependent activity of Cbfa proteins. Mol Cell Biol 21:2891–2905. 143. Zhang YW, Yasui N, Ito K et al. 2000 A RUNX2/PEBP2αA/ CBFA1 mutation displaying impaired transactivation and Smad interaction in cleidocranial dysplasia. Proc Natl Acad Sci USA 97:10549–10554. 144. Chen HY, Sun JM, Hendzel MJ, Rattner JB, Davie JR 1996 Changes in the nuclear matrix of chicken erythrocytes that accompany maturation. Biochem J 320:257–265.
GARY S. STEIN,
ET AL .
145. Nardozza TA, Quigley MM, Getzenberg RH 1996 Association of transcription factors with the nuclear matrix. J Cell Biochem 61:467–477. 146. Stein GS, Stein JL, Lian JB, van Wijnen AJ, Montecino M 1996 Functional interrelationships between nuclear structure and transcriptional control: contributions to regulation of cell cycle- and tissue-specific gene expression. J Cell Biochem 62:198–209. 147. Alvarez M, Long H, Onyia J et al. 1997 Rat osteoblast and osteosarcoma nuclear matrix proteins bind with sequence specificity to the rat type I collagen promoter. Endocrinology 138:482–489. 148. Davie JR 1997 Nuclear matrix, dynamic histone acetylation and transcriptionally active chromatin. Mol Biol Rep 24:197–207. 149. Grande MA, van der Kraan I, de Jong L, van Driel R 1997 Nuclear distribution of transcription factors in relation to sites of transcription and RNA polymerase II. J Cell Sci 110:1781–1791. 150. Jackson DA 1997 Chromatin domains and nuclear compartments: establishing sites of gene expression in eukaryotic nuclei. Mol Biol Rep. 24:209–220. 151. Lindenmuth DM, van Wijnen AJ, Hiebert S et al. 1997 Subcellular partitioning of transcription factors during osteoblast differentiation: developmental association of the AML/CBFα/PEBP2α-related transcription factorNMP-2 with the nuclear matrix. J Cell Biochem 66: 123–132. 152. Tang L, Guo B, Javed A et al. 1999 Crystal structure of the nuclear matrix targeting signal of the transcription factor AML-1/PEBP2αB/CBFα2. J Biol Chem 274: 33580–33586. 153. Young DW, Zaidi SK, Stein JL, Lian JB, van Wijnen AJ, Stein GS 2002 The dynamic spatial distribution of a tissuespecific transcription factor during mitosis. Unpublished. 154. Wei X, Somanathan S, Samarabandu J, Berezney R 1999 Three-dimensional visualization of transcription sites and their association with splicing factor-rich nuclear speckles. J Cell Biol 146:543–558. 155. Stein GS, Montecino M, van Wijnen AJ, Stein JL, Lian JB 2000 Nuclear structure–gene expression interrelationships: implications for aberrant gene expression in cancer. Cancer Res 60:2067–2076. 156. Gasser SM 2002 Visualizing chromatin dynamics in interphase nuclei. Science 296:1412–1416.
CHAPTER 20
Mouse Models of Vitamin D Receptor Ablation MARIE B. DEMAY
I. II. III. IV.
Endocrine Unit, Department of Medicine, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts
Introduction Effect on Growth and Mineral Ion Homeostasis Effect on Vascular System Effect on Reproduction
I. INTRODUCTION The absence of functional vitamin D receptors (VDRs) has been shown to be the molecular basis for the human disease HVDRR (see Chapter 72). Data from affected patients and their families have provided significant insight into the role of the VDR in vivo. However, the limitations of human studies have precluded detailed investigations of the abnormalities in mineral ion homeostasis and determinants of alopecia in these kindreds. Four laboratories have independently generated mice with targeted ablation of the VDR [1–4]. The resultant phenotype mirrors that seen in humans affected with HVDRR. The generation of these mice lacking functional VDRs has permitted detailed investigations into abnormalities in calcium absorption, bone formation, and parathyroid function that cannot be performed in humans. Furthermore, they have permitted novel investigations into the role of the VDR in nontraditional target organs such as the skin and vascular system.
II. EFFECT ON GROWTH AND MINERAL ION HOMEOSTASIS Mice homozygous for targeted ablation of the VDR are born with the expected Mendelian frequency, demonstrating that the VDR is not essential for embryonic development. The pups are phenotypically normal and indistinguishable from their wild-type and heterozygous littermates until the third week of life [1–4]. From 21 days of age, however, the knockout mice maintain a body weight approximately 10–15% lower than that of their sex-matched control littermates until approximately 6 to 8 months of age. From that point on, they VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
V. Effect on the Immune System VI. Integument VII. Conclusions References
gradually begin to lose weight, and by 1 year of age are approximately half the size of their sex-matched littermates (Fig. 1). Correlating with normal immunoreactive PTH (iPTH) levels, the knockout mice are normocalcemic the first 2 weeks of life, although, by 21 days, secondary hyperparathyroidism develops, presumably as a consequence of impaired intestinal calcium absorption [1]. This time of onset of impaired intestinal calcium absorption correlates well with data obtained from studies in vitamin D–deficient rats [5]. These investigations demonstrated that intestinal calcium absorption was vitamin D–independent the first 2 weeks of life, then was gradually replaced by a vitamin D–dependent transport system, which predominated by 35 days. Studies in 10-week-old VDR-null mice confirmed a marked impairment in duodenal calcium absorption [3]. This was accompanied by a significant reduction in mRNAs encoding the calcium channels that are thought to regulate calcium entry into the enterocyte, CaT1 and ECaC (see Chapters 24 and 25). There was also a dramatic decrease in the expression of calbindin 9K, a 1,25-dihydroxyvitamin D– inducible gene thought to play a role in intracellular calcium transfer [2–4, 6]. This suggests that the 1,25-dihydroxyvitamin D–induced expression of these calcium channels and of calbindin 9K is a key determinant of intestinal calcium absorption and that this hormonal regulation requires the presence of the nuclear VDR. Interestingly, although ECaC and CaT1 mRNA levels in the kidney are unaffected by VDR status [3], renal calbindin 9K levels are dramatically decreased in the VDR-null mice [2–4, 6], correlating with an increase in renal calcium clearance. These data suggest that calbindin 9K deficiency likely plays an important role in the pathophysiology of renal calcium loss in the VDR-null mice. Copyright © 2005, Elsevier, Inc. All rights reserved.
342
MARIE B. DEMAY
FIGURE 1 Phenotype of VDR-null mice. At 1 year of age, the VDR-null mice (foreground) are significantly smaller than their wild-type littermates and demonstrate severe alopecia.
As a result of the PTH–mediated increase in urinary phosphorus clearance, the VDR knockout mice develop hypophosphatemia and by 21 days of age have serum phosphorus levels in the range of 6.5 mg/dl versus 10.5 mg/dl in control littermates. Initially, the secondary hyperparathyroidism is able to compensate for the decrease in intestinal calcium absorption and the knockout mice maintain normal serum ionized calcium levels until day 28. However, from that point onward, ionized calcium levels are decreased by 30%: those of the VDR knockout mice stabilizing at approximately 1.0 mmol/liter versus 1.3 mmol/liter in control littermates [1–4]. At 15 days of age, there is no detectable abnormality in parathyroid function in the VDR knockout mice. They have normal serum iPTH levels and the size of their parathyroid glands is not increased [1], suggesting that the transcription-repressing effects of 1,25-dihydroxyvitamin D on the PTH gene and its antiproliferative effects on parathyroid cells observed in vitro (see Chapter 30) are not required for normal parathyroid development or do not require the actions of the nuclear VDR. However, in association with the development of impaired intestinal calcium absorption, secondary hyperparathyroidism is observed in the VDR-null mice. There is a gradual onset of parathyroid hyperplasia due to increased parathyroid cellular proliferation, evidenced by a twofold increase in the
number of parathyroid cells expressing proliferating cell nuclear antigen at 35 days of age [7]. By 70 days of age, when the iPTH levels are approximately 16-fold elevated, the VDR knockout mice have a 10-fold increase in parathyroid glandular volume, accompanied by an increase in PTH mRNA levels, as assessed by in situ hybridization analyses [1]. When the VDR-null mice are treated with a special diet that enables maintenance of normal mineral ion homeostasis, hyperparathyroidism is not observed, nor is parathyroid glandular hyperplasia [7] (Fig. 2). This suggests that the nuclear VDR is not required for normal parathyroid function when serum calcium and phosphorus levels are normal. Perhaps the role of the VDR in the parathyroid cell is important as a response to hypercalcemic or hypocalcemic stress, or alternatively its role may be redundant. The lack of parathyroid abnormalities in normocalcemic VDR-null mice does, however, suggest that hypocalcemia is the major stimulus to PTH gene transcription and parathyroid glandular proliferation when the actions of 1,25-dihydroxyvitamin D are impaired. Analogous to findings in humans with VDR mutations and in states of vitamin D deficiency during growth (see Chapter 63), VDR-null mice develop rickets and osteomalacia [1–4]. The skeletons of 2-week-old VDR knockout mice are indistinguishable from those of their control littermates, both radiologically and histologically, demonstrating that the VDR is not required
343
CHAPTER 20 Mouse Models of Vitamin D Receptor Ablation
A B C FIGURE 2 Parathyroid hyperplasia in the VDR-null mice is prevented by normalizing mineral ions. (A) Parathyroid gland from a 70-day-old wild-type mouse; (B) from a 70-day-old VDR-null littermate fed regular chow; (C) from a 70-day-old VDR-null mouse with normal mineral ion homeostasis. The asterisk marks the parathyroid gland. (See color plate).
for skeletal development. However, expansion of the hypertrophic chondrocyte layer of the growth plate is evident histologically at 21 days [8], and by 35 days of age, there is a dramatic expansion in this layer with disruption of the columnar alignment of the hypertrophic chondrocytes (Fig. 3). Histological analyses at this time also reveal osteomalacia, characterized by a decrease in bone mineralization and an increase in osteoid volume. Histomorphometric analyses of the skeletons of 70-day-old mice reveal a 30-fold increase in osteoid volume in the VDR knockout mice, compared to that observed in their wild-type and heterozygous littermates. This is associated with an increase
A
B
in bone volume/tissue volume, a consequence of an increase in both trabecular thickness and trabecular number (Table I). As a result of the dramatic impairment in bone mineralization, the bones of the VDR knockout mice demonstrate reduced stiffness and decreased strength during biomechanical testing [9]. Analyses of cellular content reveal an increase in osteoblast number, presumably due to increased bone turnover associated with secondary hyperparathyroidism. Despite the marked secondary hyperparathyroidism, osteoclast number is not significantly increased in the VDR-null mice [9]. Since osteoclast differentiation in response to PTH has been shown to be normal in vitro
C
FIGURE 3 Prevention of rickets by normalizing mineral ion homeostasis. (A) Tibial growth plate from a 35-day-old wild-type mouse; (B) from a 35-day-old VDR-null littermate fed regular chow, demonstrating marked expansion of the hypertrophic chondrocyte layer with distortion of the normal columnar structure; (C) from a 35-day-old VDR-null mouse with normal mineral ion homeostasis. (See color plate).
344
MARIE B. DEMAY
TABLE I Histomorphometric Parameters in VDR Knockout Mice On regular chow Wild-type OS/BS OV/BV BV/TV
20.9 + 2.9% 1.7 + 0.16% 10.3 + 2.4%
Knockout 86.4 + 2.8% 51.3 + 3.4% 54.5 + 2.9%
On diet that normalizes mineral ions Wild-type
Knockout
22.2 + 3.6% 1.4 + 0.3% 13.1 + 1.4%
23.1 + 1.7% 1.2 + 0.2% 14.3 + 1.2%
OS, osteoid surface; BS, bone surface; OV, osteoid volume; BV, bone volume; TV, tissue volume. Adapted from [9].
in the VDR-null mice [10], an increase in osteoclast number would have been anticipated. The inappropriately normal osteoclast number may be due to an impairment in osteoclast attachment to bone in the VDR-null mice, since the osteoclast αvβ3 integrin is highly induced by 1,25-dihydroxyvitamin D [11,12]. Alternatively, since osteoclasts are unable to resorb osteoid, their attachment to this unmineralized matrix may be impaired. As a consequence of the dramatic decrease in bone mineralization in the VDR knockout mice, their bones demonstrate reduced stiffness and strength on biomechanical testing compared to those of control littermates [9]. When normal mineral ion homeostasis is maintained in the VDR-null mice, all of these skeletal abnormalities are prevented [9]. Notably, there is normalization of the mineral apposition rate and osteomalacia is not observed. There is normalization of osteoid volume and of bone volume/tissue volume (Table I), suggesting that this abnormality in the hypocalcemic VDR-null mice is a consequence of an increase in matrix synthesis due to the anabolic effects of PTH, coupled with impaired bone resorption, since osteoclasts cannot resorb unmineralized matrix. It is notable that in the setting of normal serum calcium and phosphorus levels, histomorphometric analyses and biomechanical testing demonstrate that there is no detectable consequence of VDR ablation. These studies suggest that the VDR is not required for skeletal homeostasis. However, they do not exclude the possibility that the role of the VDR in the skeleton is redundant and that, in its absence, a second receptor or alternative homeostatic pathway performs its normal functions. Notable in this respect is the effect of 1,25-dihydroxyvitamin D on the regulation of RANK ligand production by the osteoblast. Although 1,25-dihydroxyvitamin D has been shown to play an important role in inducing the synthesis of this key regulator of osteoclast differentiation [13], histomorphometric analyses in the normocalcemic VDR-null mice reveal normal osteoclast numbers and
resorption surfaces. However, in vitro studies demonstrate that osteoblasts from VDR-null mice cannot support osteoclastogenesis when cocultured with normal spleen cells and 1,25-dihydroxyvitamin D, although when these cocultures are performed with PTH and interleukin 1α, osteoclasts with resorbing activity were formed [10]. These data clearly demonstrate that the VDR plays a key role in the skeleton; however, in its absence, other regulatory molecules, including cytokines, hormones, or alternative receptors, may be called upon to maintain skeletal homeostasis. Notable in this respect is the notion that the rapid actions of 1,25-dihydroxyvitamin D, such as the rapid increase in intracellular calcium and activation of second messengers, are mediated by an alternative receptor (see Chapters 23 and 33). However, studies in osteoblasts isolated from mice lacking the DNA binding domain of the VDR demonstrate that the rapid increase in intracellular calcium in response to hormone is abolished [4]. These data suggest that at least some of the rapid actions of this steroid hormone are dependent upon the presence of a functional nuclear receptor and challenge the hypothesis that the actions of a membrane receptor are responsible for the maintenance of skeletal homeostasis in the normocalcemic mice lacking functional nuclear receptors. The presence of a normal growth plate prior to the development of impaired mineral ion homeostasis suggests that the expansion of the late hypertrophic chondrocyte layer, characteristic of rickets, is a consequence of impaired intestinal calcium absorption rather than the absence of a functional receptor in the chondrocytes of the VDR-null mice. Furthermore, maintenance of normal mineral ion homeostasis prevents the development of rachitic changes [7] (Fig. 3), demonstrating that, in an ideal metabolic environment, the VDR is not essential for the development or maintenance of a normal growth plate. Since extracellular calcium promotes the expression of markers of terminal chondrocyte differentiation and increases the production of mineralized matrix in chondrocytic cell lines [14], studies
345
CHAPTER 20 Mouse Models of Vitamin D Receptor Ablation
were performed to address whether impaired mineral ion homeostasis led to abnormalities in chondrocyte differentiation. These studies demonstrated that the growth plate changes were not associated with a disruption in the acquisition of markers of chondrocyte differentiation, as assessed by in situ hybridization of long bones [8]. Interestingly, expansion of the late hypertrophic chondrocyte layer was observed at a time when the VDR null mice were normocalcemic, but had hypophosphatemia due to secondary hyperparathyroidism. Although no detectable change in chondrocyte proliferation was observed, there was marked impairment in apoptosis of the late hypertrophic chondrocytes in the hypophosphatemic VDR null mice [8]. In vitro studies have demonstrated that inorganic phosphate induces chondrocyte apoptosis in a time and dosedependent manner [15–17], suggesting that the development of hypophosphatemia is a critical event in the etiology of rickets in the VDR-null mice.
III. EFFECT ON VASCULAR SYSTEM Based on epidemiological studies demonstrating an inverse relationship between plasma 25-hydroxyvitamin D levels and blood pressure, investigations were performed in the VDR knockout mice to determine whether, in fact, these mice were predisposed to developing hypertension. The mean blood pressure in the VDR knockout mice was found to be 20 mmHg higher than that of wild-type mice [18]. This was associated with an increase in renin mRNA levels and in plasma angiotensin II levels. Despite an increase in these parameters under basal conditions, salt loading, a suppressor of renin expression, was capable of decreasing renin mRNA levels in the VDR knockout mice. However, the VDR knockout mice still maintained higher renin mRNA levels than controls treated in the same fashion. These data suggest that the regulatory mechanisms required for normal salt and water balance are operative in the VDR knockout mice, but that VDR ablation leads to an increase in basal renin activity. To determine whether this increase in renin was due to hormone deficiency, wild-type mice were rendered vitamin D deficient. Interestingly, a 50% elevation in renin mRNA levels was observed in the vitamin D–deficient mice with functional VDRs, demonstrating that this was a ligand-dependent effect [18]. Studies in cell models demonstrated that overexpression of the VDR in renal cells resulted in a decrease in renin mRNA levels and that regulatory regions in the renin gene could mediate transcriptional repression in response to 1,25-dihydroxyvitamin D (see Chapter 54).
IV. EFFECT ON REPRODUCTION Initial evaluation of the reproductive potential of VDR-null mice revealed variability among the knockout lines generated. Female mice with deletion of the second zinc finger were reported to be fertile [1], whereas those with deletion of the first zinc finger had infertility associated with uterine hypoplasia and impaired folliculogenesis [2]. Subsequent studies revealed that male VDR-null mice from this same line also had reproductive dysfunction, characterized by a decreased sperm count and a reduction in sperm motility. Aromatase activity in the ovary and testis were found to be reduced and evaluation of gonadotropins revealed hypergonadotropic hypogonadism, pointing to a defect at the level of the gonad, rather than reproductive dysfunction secondary to decreased body mass or intercurrent illness [19]. However, studies in mice lacking the DBD of the VDR showed normal uterine, testicular, and seminal vesicle weight as well as normal circulating levels of testosterone or estradiol [4]. Folliculogenesis and spermatogenesis were not impaired in this line of mice, nor was fertility. Interestingly, subsequent analyses revealed that calcium supplementation partially reversed the decrease in aromatase activity and normalized estrogen levels in the mice with ablation of the first zinc finger [19]. When VDR-null females from this same line of mice were fed a high-calcium diet, they had normal-sized litters and normal pup survival [20]; their fertility was shown to be directly related to the calcium content of the diet [21]. These studies once again point to the important contribution of abnormal mineral ion homeostasis to the phenotype of the VDR-null mice and raise the question as to what specific abnormality is responsible for the development of hypergonadotropic hypogonadism in the hypocalcemic mice with targeted ablation of the first zinc finger. Studies of mammary gland development have demonstrated an enhancement of growth and budding in the absence of a functional VDR. The glands of the VDR-null females were shown to have an increase in the number of secondary branch points and as well as in the number of terminal end buds [22]. The lack of VDR also enhanced the proliferative response to estrogen and progesterone. The effect of 1,25-dihydroxyvitamin D on growth of mammary glands in vitro demonstrated a receptor-dependent suppression of the sex steroid– induced branching enhancement in the wild-type mice, but not in those lacking a functional VDR [22]. These observations suggest that hormone-dependent effects of the VDR serve to attenuate mammary development and raise the question of the long-term consequences
346
MARIE B. DEMAY
of VDR ablation or vitamin D deficiency on mammary gland development and involution (see Chapter 93).
V. EFFECT ON THE IMMUNE SYSTEM A number of investigations have demonstrated that 1,25-dihydroxyvitamin D induces the expression of several genes with immunomodulatory properties in vitro as well as having immunomodulatory effects in vivo (see Chapters 98 and 99). These studies have demonstrated that 1,25-dihydroxyvitamin D and its analogs are able to prevent the development of autoimmune disorders and synergize with agents that prevent graft rejection in experimental models. Vitamin D–deficient animals have impaired chemotaxis and cell-mediated immunity, but the postulated immunomodulatory properties of the ligand-bound serum vitamin D binding protein raise the question as to what role the nuclear VDR plays in the immune dysfunction observed in vitamin D deficiency (see Chapters 8 and 9). Studies in the VDR-null mice have shown modest consequences of VDR ablation, most of which seem to be reversed by normalization of mineral ion homeostasis [23]. Splenocyte proliferation in response to antiCD3 stimulation is mildly impaired, as is macrophage chemotaxis [23]; however, macrophage phagocytosis and killing are not affected. Leucocyte and lymphocyte subset composition is also normal. In vivo rejection of transplanted islet cells was comparable in the knockout and wild-type mice, demonstrating that absence of the VDR did not enhance graft rejection [23]. Since vitamin D analogs have been shown to delay the onset of diabetes in NOD mice, the effect of VDR status on susceptibility to diabetes was examined. Interestingly, the
A
B
VDR-null mice were protected from low-dose streptozotocin-induced diabetes. All these immune defects, as well as protection from streoptozotocininduced diabetes, were reversed in the setting of normal mineral ion homeostasis [23]. VDR-null mice were shown to have an increase in mature splenic dendritic cells, consistent with the known effects of 1,25-dihydroxyvitamin D on inhibiting dendritic cell maturation [24]. Although studies in cells from the VDR-null mice have demonstrated that the VDR is required for the effects of 1,25-dihydroxyvitamin D on the differentiation of bone marrow progenitors into macrophages and monocytes [25], the lack of significant in vivo consequences of VDR ablation suggest that the role of the nuclear VDR in the immune system is redundant.
VI. INTEGUMENT Analogous to what is seen in numerous HVDRR kindreds, the VDR knockout mice develop progressive alopecia. Clinical evidence of alopecia is seen as early as the 4th week of life, beginning in the periorbital region and progressing dorsally, eventually resulting in total loss of fur by 3 to 4 months of age [1–4]. Histological examination of the skin reveals dilatation of hair follicles with formation of dermal cysts (Fig. 4). This histological phenotype is analogous to that seen in mice with mutations in the hairless gene, in which progressive loss of fur is seen from 2 weeks of age [26]. However, expression of hairless mRNA in the skin and keratinocytes of neonatal VDR null mice is not decreased [27], suggesting that, if these two genes are in a common pathway, the VDR is likely to be downstream of hairless.
C
FIGURE 4 Normalization of mineral ion homeostasis does not prevent skin changes in the VDR-null mice. (A) Skin section from a 70-day-old wild-type mouse; (B) from a 70-day-old VDR-null littermate fed regular chow, demonstrating dermal cysts and dilatation of hair follicles; (C) from a 70-day-old VDR-null mouse with normal mineral ion homeostasis. (See color plate).
CHAPTER 20 Mouse Models of Vitamin D Receptor Ablation
Because normalization of mineral ion homeostasis failed to prevent the development of the skin phenotype in the VDR-null mice [7] (Fig. 4), studies were undertaken to address whether the loss of fur was a consequence of an abnormality in keratinocytes, which give rise to the epidermal component of the hair follicle. Numerous studies have demonstrated that 1,25-dihydroxyvitamin D plays a significant role in decreasing keratinocyte proliferation and promoting keratinocyte differentiation (see Chapter 35). Studies in primary keratinocytes isolated from neonatal VDR knockout mice failed to demonstrate abnormal keratinocyte proliferation or impaired differentiation [27]. However, investigations in growing mice demonstrated a decrease in expression levels of epidermal differentiation markers, including involucrin, profilaggrin, and loricrin, from birth to 3 weeks of age [28]. Taken together, these data suggest that the effects of the VDR on keratinocytes are not essential during development, but that they are required for keratinocyte differentiation and skin homeostasis postnatally. Hair follicle development begins at embryonic day 14.5 and is dependent on reciprocal interactions between the epidermal and mesodermal components of the hair follicle. This morphogenic period lasts until the 3rd week of life, which marks the end of the first hair cycle. After this period, the hair follicle continues to cycle, characterized by a phase of rapid growth (anagen), followed by a regression phase (catagen), a quiescent phase (telogen), and reentry into anagen to generate a new hair shaft. Regulation of this postmorphogenic hair cycling is also thought to require continued epidermal–mesodermal communication. Abnormalities in the dermal papilla, the mesodermal component of the hair follicle, or in epidermal–mesodermal communication, required for the maintenance of the normal hair cycle, could also lead to the development of alopecia. Depilation of fur in control mice results in initiation of a new hair cycle with regrowth of fur. Such studies, performed at 18 days of age, a time when there is no detectable histological abnormality in the skin of the VDR-null mice, demonstrated that the VDR-null mice had a defect in anagen initiation [27]. The question remained as to whether this was due to a cellular or a metabolic defect. Because 1,25-dihydroxyvitamin D down-regulates its own biosynthesis by repressing the 25-hydroxyvitamin D-1α-hydroxylase gene [29] and increases its metabolism by up-regulating the 24-hydroxylase gene through VDR-dependent actions [30, 31], VDR-ablated mice have very high levels of 1,25-dihydroxyvitamin D even when mineral ion levels are normalized. Since 1,25-dihydroxyvitamin D has been shown to inhibit keratinocyte proliferation and to promote keratinocyte
347 differentiation [32], very high levels of this hormone or its metabolites may lead to alopecia by toxic interactions with an alternative receptor. However, VDR null mice with undetectable circulating levels of both 25-hydroxyvitamin D and 1,25-dihydroxyvitamin D still developed alopecia, demonstrating that this phenotype was not a consequence of toxic levels of circulating hormone [33]. Interestingly, wild-type littermates with undetectable circulating levels of 25-hydroxyvitamin D and 1,25-dihydroxyvitamin D did not develop alopecia [33]. This observation suggested that absence of receptor per se rather than absence of ligand-dependent receptor effects was responsible for the alopecia in the VDR-null mice and that different skin phenotypes are seen in absence of ligand versus absence of receptor. Since the VDR is widely expressed under normal circumstances and is missing from all cells of the VDR-null mice, it was not evident whether lack of VDR expression in keratinocytes, dermal papilla cells, or other organs of the VDR null-mice was responsible for the abnormality in the hair cycle. Hair reconstitution assays, in which implantation of a mixture of keratinocytes and activated dermal papilla cells into a nude mouse host recapitulates the process of hair follicle morphogenesis [34] leading to a functional hair follicle, provided a method to address whether the VDR was required in either the epidermal or mesodermal component of the hair follicle, or both. By implanting wild-type or VDR-null dermal papilla cells with wildtype or VDR-null keratinocytes, both isolated from neonatal mice in which the hair cycle is still in the morphogenic period, hair follicles were generated, with epidermal and mesodermal components differing in VDR status (wild-type versus null) [33]. These studies confirmed that the VDR was not required for hair follicle morphogenesis. However, hair follicles reconstituted with VDR-null keratinocytes were unable to sustain postmorphogenic hair cycles. In contrast, when dermal papilla cells of either genotype were mixed with wildtype keratinocytes, a normal response to anagen induction, including generation of new hair shafts, was observed. These studies demonstrated that expression of the VDR in the keratinocytes was essential for normal postmorphogenic hair cycling. To address whether keratinocyte-specific expression of the VDR was sufficient to maintain hair follicle homeostasis, transgenic mice expressing the VDR under the regulation of a keratinocyte-specific promoter were generated [35, 36]. These mice were crossbred to the VDR-null mice to generate mice that expressed the VDR only in keratinocytes. These mice did not develop alopecia and had a normal response to anagen-initiating stimuli, providing definitive proof that expression of
348 the VDR in keratinocytes was both necessary and sufficient for the maintenance of postmorphogenic hair cycling. Future investigations will be required to determine which regions of the VDR are required for skin homeostasis, whether the actions are dependent on an unique subset of nuclear receptor coactivators or an intact ligand binding domain, and, ultimately, what molecular pathway is disrupted by VDR ablation.
VII. CONCLUSIONS Studies in mice lacking functional VDRs have clearly demonstrated that the major roles of this nuclear receptor in vivo are the promotion of intestinal calcium absorption and the maintenance of skin homeostasis. Interestingly, effects of the VDR on two traditional target organs, the parathyroid and the skeleton, seem to be redundant, in that no abnormalities are observed in these organs in VDR-null mice with normal mineral ions. The effects of the VDR on intestinal calcium absorption seem to be classical ligand-dependent effects, and significant progress has been made in the identification of genes involved in VDR-mediated intestinal calcium absorption. In contrast, data point to potentially hormone-independent effects of the VDR on the hair cycle, and the molecular pathways that mediate the actions of this nuclear receptor in the skin are yet to be identified.
References 1. Li YC, Pirro AE, Amling M, Delling G, Baron R, Bronson R, Demay MB 1997 Targeted ablation of the vitamin D receptor: an animal model of vitamin D-dependent rickets type II with alopecia. Proc Natl Acad Sci USA 94:9831–9835. 2. Yoshizawa T, Handa Y, Uematsu Y, Takeda S, Sekine K, Yoshihara Y, Kawakami T, Alioka K, Sato H, Uchiyama Y, Masushige S, Fukamizu A, Matsumoto T, Kato S 1997 Mice lacking the vitamin D receptor exhibit impaired bone formation, uterine hypoplasia and growth retardation after weaning. Nat Genet 16:391–396. 3. VanCromphaut S, Dewerchin M, Hoenderop J, Stockmans I, VanHerck E, Kato S, Bindels R, Collen D, Carmeliet P, Bouillon R, Carmeliet G 2001 Duodenal calcium absorption in vitamin D receptor-knockout mice: Functional and molecular aspects. Proc Natl Acad Sci USA 98:13324–13329. 4. Erben R, Soegiarto D, Weber K, Zeitz U, Lieberherr M, Gniadecki R, Moller G, Adamski J, Balling R 2002 Deletion of deoxyribonucleic acid binding domain of the vitamin D receptor abrogates genomic and nongenomic functions of vitamin D. Mol Endocrinol 16:1524–1537. 5. Dostal LA, Toverud SU 1984 Effect of vitamin D3 on duodenal calcium absorption in vivo during early development. Am J Physiol 246:G528–G534.
MARIE B. DEMAY
6. Li YC, Pirro AE, Demay M 1998 Analysis of vitamin Ddependent calcium-binding protein messenger ribonucleic acid expression in mice lacking the vitamin D receptor. Endocrinology 139:847–851. 7. Li YC, Amling M, Pirro AE, Priemel M, Meuse J, Baron R, Delling G, Demay MB 1998 Normalization of mineral ion homeostasis by dietary means prevents hyperparathyroidism, rickets, and osteomalacia, but not alopecia in vitamin D receptor–ablated mice. Endocrinology 139:4391–4396. 8. Donohue MM, Demay MB 2002 Rickets in VDR null mice is secondary to decreased apoptosis of hypertrophic chondrocytes. Endocrinology 143:3691–3694. 9. Amling M, Priemel M HT, Chapin K, Rueger JM, Baron R, Demay MB 1999 Rescue of the skeletal phenotype of vitamin D receptor ablated mice in the setting of normal mineral ion homeostasis: Formal histomorphometric and biomechanical analyses. Endocrinology 140:4982–4987. 10. Takeda S, Yoshizawa T, Nagai Y, Yamato H, Fukumoto S, Sekine K, Kato S, Matsumoto T, Fujita T 1999 Stimulation of osteoclast formation by 1,25-dihydroxyvitamin D requires its binding to vitamin D receptor (VDR) in osteoblastic cells: Studies using VDR knockout mice. Endocrinology 140:1005–1008. 11. Cao X, Ross FP, Zhang L, MacDonald PN, Chappel J, Teitelbaum SL 1993 Cloning of the promoter for the avian integrin β3 subunit gene and its regulation by 1,25-dihydroxyvitamin D3. J Biol Chem 268:27371–27380. 12. Medhora MM, Teitelbaum S, Chappel J, Alvarez J, Mimura H, Ross FP, Hruska K 1993 1α,25-Dihydroxyvitamin D3 up-regulates expression of the osteoclast integrin αvβ3. J Biol Chem 268:1456–1461. 13. Yasuda H, Shima N, Nakagawa N, Yamaguchi K, Kinosaki M, Mochizuki S, Tomoyasu A, Yano K, Goto M, Murakami A, Tsuda E, Morinaga T, Higashio K, Udagawa N, Takahashi N, Suda T 1998 Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc Natl Acad Sci USA 95:3597–3602. 14. Chang W, Tu C, Pratt S, Chen T, Shoback D 2002 Extracellular Ca2+-Sensing receptors modulate matrix production and mineralization in chondrogenic RCJ3.15.18 cells. Endocrinology 143:1467–1474. 15. Adams C, Mansfield K, Perlot R, Shapiro I 2001 Matrix regulation of skeletal cell apoptosis: Role of calcium and phosphate ions. J Biol Chem 276:20316–20322. 16. Mansfield K, Rajpurohit R, Shapiro I 1999 Extracellular phosphate ions cause apoptosis of terminally differentiated epiphyseal chondrocytes. J Cell Physiol 179:276–286. 17. Mansfield K, Teixeira C, Adams C, Shapiro I 2001 Phosphate ions mediate chondrocyte apoptosis through a plasma membrane transporter mechanism. Bone 28:1–8. 18. Li YC, Kong J, Wei M, Chen ZF, Liu SQ, Cao LP 2002 1,25Dihydroxyvitamin D3 is a negative endocrine regulator of the renin-angiotensin system. J Clin Invest 110:229–238. 19. Kinuta K, Tanaka H, Moriwake T, Aya K, Kato S, Seino Y 2000 Vitamin D is an important factor in estrogen biosynthesis of both female and male gonads. Endocrinology 141:1317–1324. 20. Johnson LE, DeLuca HF 2001 Vitamin D receptor null mutant mice fed high levels of calcium are fertile. J Nutr 131: 1787–1791. 21. Johnson LE, DeLuca HF 2002 Reproductive defects are corrected in vitamin D-deficient female rats fed a high calcium, phosphorus and lactose diet. J Nutr 132:2270–2273.
CHAPTER 20 Mouse Models of Vitamin D Receptor Ablation
22. Zinser G, Packman K, Welsh J 2002 Vitamin D3 receptor ablation alters mammary gland morphogenesis. Development 129:3067–3076. 23. Mathieu C, VanEtten E, Gysemans C, Decallonne G, Kato S, Laureys J, Depovere J, Valckx D, Verstuyf A, Bouillon R 2001 In vitro and in vivo analysis of the immune system of vitamin D receptor knockout mice. J Bone Miner Res 16:2057–2065. 24. Xing N, L-Maldonado ML, Bachman LA, McKean DJ, Kumar R, Griffin MD 2002 Distinctive dendritic cell modulation by vitamin D3 and glucocorticoid pathways. Biochem Biophys Res Commun 297:645–652. 25. O’Kelly J, Hisatake J, Hisatake Y, Bishop J, Norman A, Koeffler HP 2002 Normal myelopoiesis but abnormal T lymphocyte responses in vitamin D receptor knockout mice. J Clin Invest 109:1091–1099. 26. Mann SJ 1971 Hair loss and cyst formation in hairless and rhino mutant mice. Anat Rec 170:485–499. 27. Sakai Y, Demay MB 2000 Evaluation of keratinocyte proliferation and differentiation in vitamin D receptor knockout mice. Endocrinology 141:2043–2049. 28. Xie Z, Komuves L, Yu QC, Elalieh H, Ng DC, Leary C, Chang S, Crumrine D, Yoshizawa T, Kato S, Bikle DD 2002 Lack of the vitamin D receptor is associated with reduced epidermal differentiation and hair follicle growth. J Invest Dermatol 118:11–16. 29. Takeyama K, Kitanaka S, Sato T, Kobori M, Yanagisawa J, Kato S 1997 25-Hydroxyvitamin D3 1α-hydroxylase and vitamin D synthesis. Science 277:1827–1830.
349 30. Kerry DM DP, Hahn CN, Morris HA, Omdahl JL, May BK 1996 Transcriptional synergism between vitamin D-responsive elements in the rat 25-hydroxyvitamin D 24-hydroxylase (CYP24) promoter. J Biol Chem 22:29715–29721. 31. Ohyama Y, Ozono K, Uchida M, Shinki T, Kato S, Suda T, Yamamoto O, Noshiro M, Kato Y 1994 Identification of a vitamin D–responsive element in the 5′-flanking region of the rat 25-hydroxyvitamin D3 24-hydroxylase gene. J Biol Chem 269:10545–10550. 32. Bikle DD, Pillai S 1993 Vitamin D, calcium, and epidermal differentiation. Endocr Rev 14:3–19. 33. Sakai Y, Kishimoto J, Demay MB 2001 Metabolic and cellular analysis of alopecia in vitamin D receptor knockout mice. J Clin Invest 107:961–966. 34. Kishimoto J, Ehama R, Wu L, Jiang S, Jiang N, Burgeson RE 1999 Selective activation of the versican promoter by epithelial-mesenchymal interactions during hair follicle development. Proc Natl Acad Sci USA 96:7336–7341. 35. Chen C, Sakai Y, Demay MB 2001 Targeting expression of the human vitamin D receptor to the keratinocytes of vitamin D receptor null mice prevents alopecia. Endocrinology 142: 5386–5389. 36. Kong J, Li XJ, Gavin D, Jiang Y, Li YC 2002 Targeted expression of human vitamin D receptor in the skin promotes the initiation of the postnatal hair follicle cycle and rescues the alopecia in vitamin D receptor null mice. J Invest Dermatol 118:631–638.
CHAPTER 21
Intracellular Vitamin D Response Element Binding Proteins JOHN S. ADAMS
Burns and Allen Research Institute and the Division of Endocrinology, Diabetes and Metabolism, Cedars-Sinai Medical Center, The David Geffen School of Medicine at UCLA, Los Angeles, California
I. Introduction II. New World Primates III. The Biochemical Nature of Vitamin D Resistance in New World Primates IV. New World Primate–like Vitamin D Resistance in Man V. Characterization of the Human Response Element Binding Protein (REBiP)
I. INTRODUCTION Since the publication of the first edition of Vitamin D, much has been learned regarding the cellular machinery for intracellular vitamin D trafficking (see Chapters 10 and 22), genomic action (see Chapters 11–20), and metabolism (see Chapters 4–7). This chapter will chronicle the discovery of a family of heterogeneous nuclear ribonucleoproteins (hnRNPs) that constitute one such previously unrecognized mode of control over the transactivation of vitamin D–regulated genes. As so often happens in science and medicine, this discovery evolved from the molecular analysis of a successful “experiment of nature.” The story begins with this author’s investigation of a man-made disturbance in vitamin D homeostasis in several species of New World primates resident at the Los Angeles Zoo in the mid1980s and concludes with the discovery of a previously unrecognized form of human vitamin D–resistant rickets.
II. NEW WORLD PRIMATES A. Early Primate Evolution In the Eocene period the great southern hemispheric landmass, Pangea, ruptured. This tectonic event resulted in the American landmass and Madagascar moving away from Africa. This continental separation occurred early in the process of primate evolution, trapping primordial primates in South America, Africa, and Madagascar, respectively. As a consequence, the three major primate infraorders, platyrrhines or New World primates, catarrhines or Old World primates, and lemurs, evolved independently of one another [1] (Fig. 1). VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
VI. Heterogeneous Nuclear Ribonucleoproteins (hnRNPs) VII. Compensation for the Dominant-Negative Acting, Response Element Binding Proteins VIII. Conclusion Reference
Unlike Old World primates, including our own species, which have populated virtually every landmass on our planet, New World primates have remained confined to Central and South America for the past 50 million years. Compared to Old World primates, especially some of the terrestrial species such as gorilla, New World primates are smaller in stature, a trait well-suited to their lifestyle as plant-eating arboreal sunbathers residing in the canopy of the periequatorial rain forests of the Americas.
B. Simian Bone Disease The appearance of generalized metabolic bone disease in captive primates has been recognized for the past 150 years [2]. The disease, which has not been well studied from a histopathological standpoint, carries the clinical and radiological stigmata of rickets and osteomalacia [3]. Compared to Old World primates reared in captivity, New World primates or platyrrhines are particularly susceptible to the disease. The disorder affects primarily young, growing animals and results in muscle weakness, skeletal fragility, and in many instances death of the affected individual. Rachitic bone disease of this sort has long presented a problem to veterinarians caring for captive platyrrhines, particularly in North American and European zoos [4], because death of preadolescent and adolescent primates prior to sexual maturity severely limits on-site breeding programs. Because the disease was reported to be ameliorated either by the oral administration of vitamin D3 in large doses or by ultraviolet B irradiation of affected Copyright © 2005, Elsevier, Inc. All rights reserved.
352
JOHN S. ADAMS
Old World
New World
Madagascar
Hominoidea Platyrrhini
Superorder
Catarrhini
ANTHRIPOIDEA PRIMATES
Lemuridae
Infraorder
PROSIMII
Suborder Order
FIGURE 1 New World primate evolution. Shown in geographic terms is the independent evolution of the three primate infraorders, Platyrrhini, Catarrhini, and Lemuridae, in South America (the New World), Africa (the Old World), and Madagascar, respectively. The superorder Hominoidea represents the human precursors.
primates, it was presumed to be caused by vitamin D deficiency [4]. The frequent occurrence of rickets and osteomalacia in New World primates was also ascribed to the relative inability of platyrrhines to effectively employ vitamin D2 in their diet [5]; a similar observation had been made for chickens [6]. Using assay technology that does not discriminate between 25-hydroxylated vitamin D2 and vitamin D3 metabolites, investigators [7] determined that 25-hydroxyvitamin D (25-OHD) levels were two- to threefold higher when platyrrhines were dosed with supplemental vitamin D3 than with vitamin D2. These data suggested that 25-hydroxylation of vitamin D substrate in New World primates was much more effective when vitamin D3 was employed as substrate. However, in the same study two species of Old World primates demonstrated similar discrimination against vitamin D2, in favor of vitamin D3. In summary, these results seemed to indicate that all subhuman primates, whether Old or New World, were relatively resistant to vitamin D2 in terms of its ability to engender an increase in serum levels of 25-OHD. Finally, Hay [8] suggested that New World primates may transport 25-OHD in the serum by means and via proteins that are dissimilar from those encountered in Old World primate species. This hypothesis was disproven by Bouillon et al. [9], who showed that the
vitamin D binding protein was the major carrier of 25-OHD in their serum of both New and Old World primates. The question of why platyrrhines were more susceptible to vitamin D deficiency than were catarrhines began to be answered with the detection of extraordinarily high circulating levels of the active vitamin D metabolite, 1,25-dihydroxyvitamin D, in New World primates [10–12]. These data confirmed that New World primates were resistant to the vitamin D hormone.
C. Outbreak of Rickets in the New World Primate Colonies of the Los Angeles Zoo The index case in the original studies was a preadolescent New World primate of the Emperor tamarin species (Fig. 2A). When investigated radiographically (Fig. 2B), this tamarin and those like him displayed classical rickets complete with growth retardation and metaphyseal cupping and fraying characteristic of rickets. In order to investigate this rachitic syndrome, blood and urine was collected from involved monkeys as well as from control, nonrachitic New and Old World primates. As shown in Fig. 3, that comparison yielded a biochemical phenotype that was most remarkable for
353
CHAPTER 21 Response Element Binding Proteins
A
B
FIGURE 2 A rachitic New World primate resident of the Los Angeles Zoo. (A) A preadolescent emperor tamarin with rickets. (B) The characteristic rachitic “cupping” and “fraying” of the tibial metaphysis (arrows) of this monkey.
an elevated serum 1,25-(OH)2D in rachitic New World primates [11]. In fact, with the exception of nocturnal primates in the genus Aotus, New World primates in all genera had vitamin D hormone levels ranging to 100fold higher than that observed in Old World primates including man [13–16]. In the initial analysis New World primates affected with rickets were those with the lowest 1,25-(OH)2D levels, while their healthy counterparts were those with the highest serum 1,25-(OH)2D. These data were interpreted to mean that most New World primate genera were naturally resistant to the vitamin D hormone, and that the resistant state could be compensated by maintenance of high 1,25-(OH)2D levels. If this was true, then an increase in the serum 1,25-(OH)2D concentration in affected primates should result in biochemical
compensation for the resistant state and resolution of their rachitic bone disease. When rachitic New World primates were exposed to 6 months of artificial sunlight, both serum 25-hydroxyvitamin D (25-OHD) and product 1,25-(OH)2D levels rose significantly (P ≤ 0.02), resulting in cure of rickets [15]. In summary, New World primates are periequitorial sunbathers for a reason. As depicted by the oversized arrows in a simplified scheme of vitamin D synthesis and metabolism in Fig. 4, New World primates require much cutaneous vitamin D synthesis in order to push their 25-OHD and 1,25-(OH)2D levels high enough to effectively interact with the VDR. The question remained: why were these primates resistant to all but the highest levels of the vitamin D hormone?
7-DHC
Rachitic Phenotype
Cholesterol
UVB Blood calcium.............
Slightly decreased
Urine calcium..............
Slightly decreased
Blood phosphate................
Normal
Urine phosphate.................
Normal
Serum creatinine................
Normal
Liver function......................
Normal
25-OHD..............................
Normal
1,25-(OH)2D.......................
Very high
FIGURE 3 Biochemical phenotype of rachitic New World primates. Demonstrated are the biochemical indices of bone health in New World primates suffering from rickets compared to developmental age- and sex-matched, nonrachitic Old World primates. The outstanding characteristic is a 1,25-dihydroxyvitamin D (1,25(OH)2D) level two to three orders of magnitude greater than that observed in Old World primates, including man.
Vitamin D3
25-OHase 25-OHD3
24,25-(OH)2D3
1-OHase 1,25-(OH)2D3
1,24,25-(OH)3D3
Calcitroic acid
VDR
FIGURE 4 Simplified scheme of vitamin D synthesis and metabolism in New World primates. The bold arrows describe the means by which high 1,25-dihydroxyvitamin D3 (1,25-(OH)2D3) levels are achieved and maintained in New World primates compared to their Old World and human primate counterparts.
354
JOHN S. ADAMS
III. THE BIOCHEMICAL NATURE OF VITAMIN D RESISTANCE IN NEW WORLD PRIMATES In order to answer the question of resistance in New World primates, cultured fibroblasts and immortalized cell lines from both resistant and hormone-responsive New and Old World primates were used to track, step by step, the path taken by the vitamin D hormone from the serum vitamin D binding protein (DBP) in the blood in route to the nucleus and transactivation of hormone-responsive genes [11,13–22]. It was shown that the movement of hormone from DBP, across the cell membrane and through the cytoplasm and nuclear membrane was indistinguishable from that observed in Old World primate cells. It was also determined that the ability of the New World primate VDR to bind 1,25-(OH)2D3 or 1,25-(OH)2D2 and induce receptor dimerization with the retinoid X receptor (RXR) was normal. In fact, when removed from the intranuclear environment and in distinction to previous reports [12], the VDR in New World primates was similar to the Old World primate VDR in all biochemical and functional respects [22]. That which was not the same was the reduced ability of VDR-RXR complex to bind to its cognate cis element and transactivate the expression of genes. In order to elucidate nuclear receptor events in New World primate cells, the nuclei of New World primate cells were isolated and extracted. In addition to the VDR-RXR, it was determined that these extracts contained a second protein bound to VDREs. This protein
A
was coined the vitamin D response element binding protein or VDRE-BP [21]. In electrophoretic mobility shift assay (EMSA) using the VDRE as probe, Old World primate cell extract contained only the VDR-RXR bound to the VDRE probe, while the New World primate extract contained two probe-reactive bands, one compatible with the VDR-RXR and a second, more pronounced VDRE-BP-VDRE band. This VDRE-BP-VDRE binding reaction was specific, as the VDRE-BP was competed from VDRE probe by the addition of excess unlabeled VDRE. These data suggested that VDRE-BP might function as a dominant-negative inhibitor of receptorresponse element binding by competing in trans with receptor, “knocking it off” the VDRE (Fig. 5), thus preventing VDR-RXR binding to the VDREs. When recombinant human VDR and RXR were permitted to interact in EMSA with increasing amounts of nuclear extract from vitamin D-resistant cells containing a VDRE-BP or from normal vitamin D–responsive cells, the addition of more control extract only amplified the VDR-RXR-retarded probe on the gel. By contrast, increasing amounts of the hormone-resistant extract competed away receptor–probe binding in favor VDRE-BP–probe binding. To date, two distinct VDRE-BPs have been identified, purified, cloned, and characterized in New World primates [21]. Both are members of the heterogeneous nuclear ribonucleoprotein A (hnRNPA) family of single strand mRNA binding proteins [23]. However, as just pointed out, VDRE-BPs can also bind specifically to double-strand DNA. In fact, it is by virtue of their ability to bind DNA that they can be distinguished from
B 1,25-D
RXR
VDR
VDRE
Wildtype
VDRE-BP
VDRE
Dominant-negative
FIGURE 5 The dominant-negative action of the New World primate vitamin D response element binding protein (VDRE-BP). (A) The “wild-type” events by which the retinoid X receptor (RXR) and hormone-liganded vitamin D receptor (VDR) dimer pair interact in trans with the vitamin D response element (VDRE) to regulate transcription of 1,25-dihydroxyvitamin D (1,25-D)-responsive genes. (B) The proposed “dominant-negative” event leading to competition for binding to the VDRE between the VDRE-BP and RXR-VDR; the net result is a blockade of transcriptional regulation.
355
CHAPTER 21 Response Element Binding Proteins
A. Index Case
Luciferase activity (arbitrary units)
500
*p < 0.001 250
0 OWP
+VDRE-BP
NWP
FIGURE 6 Squelching of RXR-VDR-directed, VDRE-reporterdriven transactivation. Shown is the significant decrease in VDREdirected luciferase reporter activity, assayed in relpicate (n = 4), in a subclone of Old World primate (OWP) cells after stable transfection with the New World primate VDRE-BP-2 (solid bar) compared to reporter activity in vector-alone transfected OWP cells (open bar) and New World primate (NWP) cells (shaded bar). Data are shown as luciferase units relative to an internal β-galactosidase standard.
traditional co-repressor proteins [24]. When overexpressed, they can effectively squelch VDR-directed transactivation. Depicted in Fig. 6 is VDRE-directed reporter activity in a subclone of wild-type Old World primate cells stably overexpressing the New World primate VDRE-BP-2 as well as in naturally hormone-resistant New World primate cells. Stable overexpression of VDRE-BP-2 squelched luciferase activity substantially compared to untransfected, wild-type cells to levels observed in hormone-resistant New World primate cells that naturally overexpress the protein; VDRE-BP-1 overexpression reduced, but not significantly, transactivation. This is strong confirmatory evidence that when overexpressed in vivo, VDRE-BP-2 is the cause of vitamin D resistance in these monkeys.
IV. NEW WORLD PRIMATE–LIKE VITAMIN D RESISTANCE IN MAN The Adams laboratory has been studying a vitamin D– resistant state in New World primates for more than 15 years with the expectation that it would provide valuable insight into how steroid hormones regulate gene expression in human primates. Prior to this time it was unknown whether there existed a human homolog to the vitamin D–resistant state in New World primates.
In 1993 a patient with the classical signs of type II hereditary vitamin D–resistant rickets (HVDRRII), including alopecia, was reported [25]. The biochemical phenotype of this patient included hypocalcemia (2.03 mmol/L corrected for albumin [normal range, 2.25–2.55 mmol/L]); raised serum alkaline phosphatase (1101 U/L [normal range, <300 U/L]); and raised circulating levels of 1,25-(OH)2D [466–650 pmol/L (normal range, 48–156 pmol/L]). Despite this, the patient had normal VDR expression; sequence analyses indicated that the coding regions, as well as the 5′ and 3′ untranslated regions of the VDR gene, were normal. Furthermore, when extracted from cells, VDR from the patient displayed normal binding capacity and affinity for the active form of vitamin D, 1,25-(OH)2D3. Transfection of VDR cDNA from the patient into receptor-deficient (human) CV-1 cells resulted in normal transactivation in response to 1,25-(OH)2D3 [25]. However, in the subject’s own cells the VDR was incapable of stable nuclear localization and transactivation after exposure to hormone 1,25-(OH)2D3. These data suggested that an extra-VDR factor was interfering with the nuclear translocation and transactivation of the VDR in cells from this patient. The patient was without evidence of any other form of steroid–thyroid–retinoid hormone resistance, and correction of rachitic bone disease was accomplished with high-dose 1,25-(OH)2D3 treatment (12 µg/day) and calcium supplements (1–3 g/day), although the alopecia persisted. It was recently discovered [26] that the underlying cause of insensitivity to 1,25-(OH)2D in this patient resulted from over-expression of a vitamin D response element binding protein (REBiP) with similarity to hnRNPA proteins that cause receptornormal hormone resistance in New World primates.
B. Suppression of RXR-VDR-Mediated Transactivation and VDRE Binding in HVDRR Cell Lines To investigate the presence of such an extra-VDR factor in cells from this HVDRR patient, promoterreporter assays were carried out using normal and HVDRR fibroblasts as transfection recipients [26]. Data confirmed the suppression of VDRE-mediated transactivation in HVDRR cells in the presence or absence of wild-type VDR. Furthermore, in contrast to control cells, HVDRR fibroblasts showed no induction of luciferase activity following treatment with 10 nM 1,25-(OH)2D3. Based on these data shown, as was the case in hormoneresistant New World primate cells, we hypothesized that cells from the HVDRR patient constitutively overexpress
356 a nuclear protein that competes with the VDR-RXR for binding to heterodimer cis recognition sequences (Fig. 5). To test this hypothesis subsequent assays were carried out using EBV-transformed HVDRR and control B-lymphocytes, which could be cultured in large volumes to facilitate maximal protein recovery from nuclear extracts. In contrast to control lymphocytes, cells from the HVDRR patient did not demonstrate any significant antiproliferative response to increasing doses of 1,25-(OH)2D3 following mitogen (PMA) activation. To investigate this further, EMSAs were carried out with recombinant human VDR and RXR and the 3-nucleotide-spaced direct repeat VDRE (VDRE-DR3; cis sequence AGGTCAcagAGGTCA) probe in the presence or absence of increasing amounts of nuclear extract from control or HVDRR cells. Data showed that addition of the HVDRR nuclear extract competitively displaced RXR-VDR binding to the VDRE in a dosedependent fashion. In contrast, nuclear extracts from control cells were without effect. These results were confirmed by densitometric analysis of EMSA band density performed in triplicate; significant (p < 0.05), stepwise diminishment of probe-VDR-RXR complex formation with increasing doses of HVDRR NE (4, 8, and 12 µg) was achieved compared to equivalent control extracts.
JOHN S. ADAMS
Nuclear extracts from HVDRR and control cells were subjected to SDS-PAGE and the separated proteins analyzed for their ability to bind radiolabeled VDRE probe and RXRE probe in Southwestern blots. A VDRE-DR3 and RXRE-reactive protein was observed only in the nuclear extracts from the HVDRR cells. In view of the RNA binding capacity of hnRNPs, further studies were carried out to assess REBiP binding to single-stranded nucleotide sequences [26]. When HVDRR nuclear extracts were incubated with a single-strand radiolabeled DNA probe consisting of the upper strand of the RXRE, the REBiP complex could only be competed away using an excess of unlabeled single- or doublestranded RXRE. Results presented above suggested that resistance to 1,25-(OH)2D3 in the HVDRR patient correlated with overexpression of an REBiP(s). To confirm a functional link between the cis binding of REBiP and VDR-mediated transactivation, further reporter studies were carried out using the VDR-positive HKC-8 human renal cell line cotransfected with expression constructs for human hnRNPA1 or hnRNPA2 and with the NWP dominant-negative-acting hnRNP-related VDRE-BP2, which show strong homology to one another [21]. Both of these REBiPs suppressed VDREdirected transcription to a similar degree (Fig. 7).
V. CHARACTERIZATION OF THE HUMAN RESPONSE ELEMENT BINDING PROTEIN (REBiP) A. REBiP Binding in Cis The cis-element specificity of the HVDRR cell extract-VDRE EMSA complex was assessed by competition analyses. Data confirmed that RXRE and VDRE competed away the hormone response element–binding complex; however, consensus sequences for other types of cis elements such as the ERE showed no displacement of the response element binding complex [26]. The peptide nature of the complex component(s) was confirmed using HVDRR nuclear extracts that had been digested by trypsin; unlike untreated HVDRR extracts, these preparations were unable to compete away the EMSA complex formed by recombinant VDR-RXR [26]. Furthermore, VDRE probe binding by HVDRR nuclear extracts was competed away by the addition of antihnRNPA1 and -A2 antibody, suggesting that, in a fashion similar to that in New World primate cells (Fig. 5), HVDRR cells expressed an hnRNP-related REBiP capable of interacting with the VDRE. Characterization of the REBiP by Western blot analysis showed increased expression of hnRNPA1 and -A2 immunoreactive proteins in HVDRR cells.
** p < 0.001 Relative VDRE-luciferase activity
**
Vector
Human hnRNPA1
**
Primate VDRE-BP2
FIGURE 7 Equivalent squelching of VDRE-directed luciferase activity by human and New World primate hnRNP-related response element binding proteins. Shown are vitamin D receptor (VDR)-positive human kidney cells (HKC-8) transfected with a VDRE-driven reporter plasmid alone (solid bar) or cotransfected with expression constructs for human hnRNPA1 (dark-shaded bar) or New World primate VDRE-BP2 (light-shaded bar). Cells were precultured in the presence of 10 nM 1,25-(OH)2D3 for 24 hr. Data are shown as luciferase units relative to an internal β-galactosidase standard. Values are the mean ± standard deviation of three separate assays.
357
CHAPTER 21 Response Element Binding Proteins
B. Lessons Learned from Overexpression of REBiP In summary, we have described a new form of human sterol/steroid hormone insensitivity that results from overexpression of a dominant-negative-acting response element binding protein or REBiP. In this case the phenotype of vitamin D resistance could not be attributed to mutant forms or abnormal amounts of either the VDR or its principal binding partners, RXRα and RXRβ, and there was no clinical evidence for malfunction of other steroid hormone receptors. The fact that REBiP is both a single- and double-strand nucleic acid binding protein and not a receptor binding protein indicates that vitamin D resistance in this patient was not due to abnormal expression of a VDR accessory protein. As with New World primates, the underlying cause of insensitivity to 1,25-(OH)2D3 appears to be overexpression of an hnRNPA, with both the human hnRNPA1 and the New World primate VDRE-BP2 equivalent able to squelch VDRE-mediated transactivation. The fact that the hypocalcemia and rachitic bone disease in this HVDRR patient were responsive to high-dose 1,25-(OH)2D3 treatment [25] indicates that vitamin D resistance is not absolute, but instead is likely to be determined by the relative abundance of the hnRNPA-related REBiP and competent (i.e., 1,25(OH)2D3-liganded) RXR-VDR heterodimer present in the target cell. For example, if the balance in cis element binding favored REBiP, because of either a relative abundance of this protein and/or a relative lack of the competitive RXR-VDR, then one would predict a 1,25-(OH)2D3-reversible vitamin D-resistant phenotype in vivo. In fact, this prediction was compatible with the patient’s response to treatment with high-dose 1,25-(OH)2D3; hypocalcemia and rachitic bone disease were corrected as long as the serum 1,25-(OH)2D3 level remained high [25]. These data also suggest that it is the VDRE-DR3 cis element that legislates the antirachitic action of the hormone. In contrast, failure of the patient’s alopecia to be improved with an increase in serum 1,25-(OH)2D3 levels suggests that (1) the VDRE-DR3 does not participate in control of genes rendering the hairless phenotype; (2) maternal levels of 1,25-(OH)2D3 are not high enough for fetal hair development; and/or (3) as suggested by previous data from HVDRR patients and VDR-ablated animals, the role of 1,25-(OH)2D3-VDR in hair follicle development involves only a prenatal mechanism. That the REBiP complex could not be competed out with an ERE or YY1-RE suggests that the hnRNPA family members have greater specificity for direct repeat half-sites. This is further supported by the elucidation of a double-strand DNA target sequence for hnRNPA1 that includes two cryptic direct repeat half-sites
(ggctggtcttgaactcctgA/GCTCAAA/GGTGAtcctcc; [27]) and no overlap with established RNA target sequences for hnRNPA1 [23]. Furthermore, studies from Chen et al. have shown that the estrogen resistance-causing estrogen response element binding protein in New World primates belongs to the distinct hnRNPC family (see Section VII) and possesses an entirely different target sequence on double-strand DNA. This provides an explanation for the highly specific phenotype observed with the HVDRR patient.
VI. HETEROGENEOUS NUCLEAR RIBONUCLEOPROTEINS (hnRNPs) A. hnRNPs: The Classical View The hnRNPs were initially recognized as proteins that interacted in trans with specific cis elements in the 3′ end of pre-mRNAs [28]. This interaction of the hnRNPs with the pre-mRNA begins cotranscriptionally and may persist as an mRNA–protein complex (mRNP) through the process of splicing, posttranscriptional modification, and nuclear export to the ribosome. The classic view [23] held that these proteins were critical in the “stabilization” of those mRNAs that were destined for translation; on the other hand, the absence of these stabilizing factors routed mRNAs to destruction in advance of translation. However, the data presented here for both the New World primate VDRE-BPs and the human REBiPs indicate that some members of the hnRNP family of proteins, at least when overexpressed, can also alter the process of transcription through a direct protein–DNA interaction. This kind of information and that from a variety of other sources [28] now clearly demonstrates that the hnRNPs are more than just single strand RNA binding proteins that bind to 3′ end of pre-mRNAs, “stabilizing” those transcripts for eventual translation.
B. hnRNPs as Multifunctional Proteins Whereas work presented so far points to a role of hnRNPs in transcription, the work of others is beginning to show that these proteins subserve a number of functions both inside and outside of the eukaryotic cell nucleus. There are at least four protein machines in the nucleus and another in the cytoplasm of the human cell dedicated to the process of making transcripts of genes and converting those transcripts to proteins: the nucleosome, transcription apparatus, spliceosome, and exosome in the nucleus and the ribosome outside of the nuclear compartment. There is now growing evidence that the hnRNPs are involved in the operation of each of these [29].
358
JOHN S. ADAMS
The nucleosome is the basic structural element in chromatin [30]. It consists of 146 bp of DNA wound tightly around a core of histones. The nucleosomes and their chromatin are differentially compacted into highly condensed heterochromatin and euchromatin, the latter of which is less densely packed and representative of the sections of the genome that are accessible to transcription activators. It is the insertion of single-strand and double-strand DNA binding proteins, including members of the hnRNP family [31,32], that initiates decompaction and provides accessibility to the regulatory elements of genes destined for transcriptional modulation. The transcription apparatus is classically viewed as the collection of proteins associated with RNA polymerase II (RNAPII). Included among these proteins are a number of transcription factors that interact with specific cis elements in the promoter regions of genes targeted for transcription. These transcription factors include the sterol/sterol receptors as well as members of the hnRNP family of single/double-strand DNA proteins [20,21,33] which can compete with the receptor proteins for the same cis element. DNA strand separation is a crucial element of the transcription process. Strand separation is initiated by RNAPII as it is tethered to the promoter by a bridge of co-regulator proteins [34]. Once initiation is fixed, the process of “elongation” ensues and the co-regulator tether is broken or dislodged from its DNA anchor site in the promoter. The RNAPII is then free to make its way down the coding strand of the DNA template (Fig. 8). This chain of events makes it possible for the human hnRNP-related REBiP which can occupy the VDRE in the promoter (at left in Fig. 8) and squelch hormonedirected transactivation [26], to also interact with the C terminal domain (CTD) (middle of Fig. 8) of RNAPII,
1,25-D
RNAP
Pre-mRNA
CTD
Co-activator RXR
TATA
attracting the RNAPII to the transcription start site. By successfully competing in cis for the VDRE, the liganded VDR-RXR can dislodge the REBiP, permitting the coactivator tether to assemble and interact with RNAPII. Once initiation of transcription is fixed, it is possible for the REBiP to reoccupy the VDRE releasing the VDR-RXR-anchored tether and permitting elongation (of right in Fig. 8). On the other hand, if the REBiP remains bound to the CTD of RNAPII, then by virtue of its ability to bind single-strand RNA it has the potential to influence the processing, export, and translation of the transcript by binding to its 3′ UTR (see later discussion). The spliceosome is one of the largest of the nuclear machines. It functions to recognize the intron–exon junctions in eukaryotic genes, remove the introns and then variably re-“splice” the transcribed exons in a cellspecific manner. The spliceosome comprises 145 distinct proteins. At least 30 of these proteins, including 12 in the hnRNP family, have recognized functions outside of the realm of splicing [35]. The function of these many spliceosome-associated hnRNPs is unknown. It is almost certain that their ability to recognize both RNA and other proteins is key in demarcating the scission and resplice junctions in pre-mRNA as part of the so-called exon–exon junction complex or EJC. There is now convincing evidence that the processes of transcription, splicing, 3′ polyadenylation and 5′ capping are functionally coupled and temporally coincident [36] with the CTD of RNAPII proposed to serve as a platform for the ordered assembly of the different families of pre-mRNA processing machines ([29]; see Fig. 8). An especially relevant example is the recent study of Aboeuf and co-workers [37]. They showed that the liganded glucocorticoid receptor (GR) exerted coincident control over RNAPII and alternative splicing of
REBiP
VDR
RNAP CTD
REBiP 5′
REBiP RNAP 3′
CTD REBiP
FIGURE 8 Simplified schematic of the potential for serial interaction of the vitamin D-resistance-causing, human vitamin D–resistant rickets (HVDRR) response element binding protein (REBiP) with the VDRE in the promoter of a 1,25(OH)2D-VDR-RXR-regulated gene. Shown is the transition of REBiP interaction with the VDRE in the basal state (at left), the C-terminal domain (CTD) of the RNA polymerase II (RNAP; middle), and 3′ untranslated end of the pre-mRNA transcript of that gene (at right). Arrows indicate the direction of transcription.
CHAPTER 21 Response Element Binding Proteins
the transcript initiated under the influence of the ligand. It is possible that VDRE-reactive REBiP and VDR-RXR coordinately alter transcription and splicing in a similar fashion. The exosome constitutes the protein machinery responsible for the export of mRNA from the nucleus [38]. These events do not occur in isolation. Rather they are functionally linked to upstream events in RNA splicing and processing and downstream processes such as nonsense-mediated messenger decay [39]. The hnRNPs are proposed to be principal players in the exosome, acting as a connector between the 3′ end of the mRNA and constituent proteins of the nuclear pore [40]. Hence, it is possible, as described in Fig. 8, that the hnRNPA-related VDRE-BPs and REBiP, by virtue of their protein- and nucleic acid-binding capacity, could be passed from promoter to RNAPII- to 3′ end of mRNA to exosome and contribute to the delivery of that same gene’s mRNA to the cytoplasm. These hypotheses are yet to be confirmed but are under active investigation. The ribosome and the process of translation is the final stop for an mRNA. hnRNPs serve as ribosome recognition proteins [41]. They participate in parking transcripts with tRNAs anchored in the 30s subunit of the ribosome. Therefore, it is possible that the influence of the human REBiPs and the nonhuman primate homolog VDRE-BPs can extend all the way from the promoter of a gene to the translation of that gene’s transcript into a functional protein; as previously noted it is now clear that the process of making and translating RNA is not the result of a series of independent processes, but the outcome of functionally-linked simultaneous events. It is also clear, as demonstrated by the hnRNPs, that proteins thought once to be involved with specific single steps in the making, processing and translating RNA are not limited to single steps in that chain of events. Finally, as exemplified by the 154 proteins making up the spliceosome machinery of which only half have a clearly defined splicing function [35], it is also clear that there remains much to be learned about how these various complex machines described above function in concert with one another.
VII. COMPENSATION FOR THE DOMINANT-NEGATIVE ACTING, RESPONSE ELEMENT BINDING PROTEINS A. Intracellular Vitamin D Binding Proteins During the process of discovery of the VDRE-BPs in New World primate cells, it was also observed that
359 these cells were extraordinarily efficient at accumulating 25-hydroxylated vitamin D metabolites in the cytoplasmic space. Accumulation here was the result of expression of a second set of resistance-associated proteins. These intracellular vitamin D bindings proteins [42,43], or IDBPs as they have come to be called, exhibit both high capacity and high affinity for 25-hydroxylated vitamin D metabolites. In fact, among all of the vitamin D metabolites that have been tested, the IDBP(s) purified from vitamin D–resistant New World primate cells bind 25-OHD3 and 25-OHD2 best [14,43]; in a competitive displacement assay using radioinert 25-OHD3 as competitor and [3H]25-OHD3 as labeled ligand, the concentration of metabolite required to achieve half-maximal displacement of labeled hormone (EC50) was <1 nM. Although normally present in Old World primate (including human) cells, these proteins can be overexpressed some 50-fold in New World primate cells. These IDBPs are highly homologous to proteins in the heat shock protein-70 family [44]. The first four members of this family, IDBP-1, -2, -3 and -4, have been cloned and characterized [21,45]. They bear a high degree of sequence identity with constitutively expressed human heat shock protein-70, endoplasmic reticulum–targeted glucose regulated protein-78 (grp78), mitochondrial-targeted grp 75, and heat-shock inducible heat shock protein-70, respectively. The general domain structure of the IDBPs [44] is shown in Fig. 9A. They all contain an ATP-binding-ATPase domain ahead of a protein-protein interaction domain. Some like IDBP-2 and IDBP-3 also harbor an N-terminal organelle-targeting domain. Preliminary studies indicate that the vitamin D ligand binding domain resides in the middle of the molecule near the C-terminal aspect of the ATPase domain [46]. What are these IDBPs doing inside the hormoneresistant New World primate cell? Two countervailing hypotheses were considered to explain the function of these proteins (Fig. 9B). One hypothesis held that these IDBPs were “sink” molecules that worked in cooperation with the VDRE-BP in the nucleus to exert vitamin D resistance by disallowing access of the hormone to the VDR and the nucleus of the cell. The opposing hypothesis held that these were “swim” molecules that actually promoted the delivery of ligand to the vitamin D receptor, improving the ability of the VDR to dimerize and bind to DNA, antagonizing the actions of the VDRE-BP that was overexpressed in New World primate cells. In order to determine which of these hypotheses was correct, the most abundant of these IDBPs, IDBP-1, was stably overexpressed in vitamin Dresponsive (i.e., wild-type) Old World primate cells and shown to be pro-transactivating [46]; the endogenous
360
JOHN S. ADAMS
A
B
ATPase
Hormone
‘‘swim’’
Vitamin D binding Organelle targeting
Protein:protein interaction
“sink’’ RXR Target cell
VDR
VDRE
FIGURE 9 General domain structure of the hsp70-related intracellular vitamin D binding proteins (IDBPs) and their proposed function in vitamin D-resistant New World primate cells. (A) Domain structure of hsp70-like proteins. All contain an ATP-binding-ATPase domain ahead of two protein–protein interaction domains. Some also harbor an N-terminal organelle-targeting domain. Molecular mapping studies place the vitamin D sterol binding domain in the ATPase domain of the protein. (B) The countervailing “sink” and “swim” functional hypotheses for these IDBPs.
transcriptional activity of three different 1,25-(OH)2Dresponsive genes, the vitamin D-24-hydroxylase, osteopontin, and osteocalcin genes, in Old World primate wild-type cells was significantly enhanced. It was concluded from these studies, at least for the function of transactivation, that IDBP-1 was a “swim” molecule for the vitamin D hormone, promoting delivery of ligand to the VDR; a clear role for IDBP-2 and IDBP-3 in this regard has not yet been determined. Considering the facts that New World primates are required to maintain very high serum levels of 1,25(OH)2D in order to avert rickets (Fig. 2), it was hypothesized that the IDBPs, which are known to bind 25-OHD even better than 1,25-(OH)2D, will also promote the synthesis of the active vitamin D metabolite via promotion of the 25-OHD-1-hydroxylase. When human kidney cells expressing the 25-OHD-1-hydroxylase gene were stably transfected with IDBP-1 and incubated with substrate 25-OHD3, 1,25-(OH)2D3 production went up four- to eightfold compared to untransfected, wildtype cells [47]. This increase in specific 25-OHD1-hydroxylase activity occurred independent of a change in expression of the 25-OHD-1-hydroxylase gene [46]. In fact, current data (see next section) now strongly indicate that this increase in hormone production is the result of the ability of IDBPs to promote the delivery of substrate 25-OHD to the inner mitochondrial membrane and the 25-OHD-1-hydroxylase stabled there.
B. A New Model for Intracellular Vitamin D Trafficking Dogma has held that sterol/steroid hormones such as vitamin D, by nature of their lipid solubility, diffuse through the plasma membrane of the target cell and
“Ping-Pong” around the cell interior until they encounter another specific binding protein such as 25-OHD-1-hydroxylase or the VDR with which to bind. Recent results, developed from a compendium of confocal imaging studies with fluorescently labeled IDBPs and vitamin D metabolites as well as with protein–protein interaction experiments [45], indicate that the hormone does not haphazardly “Ping-Pong” around the cell interior. Rather, the hormone enters the cell and is distributed to specific intracellular destinations by a series of protein–protein interactions which involve the hsp70 family of IDBPs. For example, we now know from the work of Willnow et al. (see Chapter 10; [48,49]) that vitamin Ds can enter target cells via internalized vesicles. The vitamin D stays bound to DBP, which is in turn bound by megalin and cubulin, members of the LDL superfamily of proteins. Once inside the cell there is interaction between the C-terminal domain of megalin, which protrudes into the cytoplasm, and the N-terminal domain of IDBP-1 [45]. In fact, if one then incubates IDBP-1 overexpressing cells with a fluorescently labeled 25-hydroxylated vitamin D metabolite, one will observe a significant increase in the uptake of the labeled hormone [45].
VIII. CONCLUSION As our knowledge base increases, the mechanisms underlying the control of vitamin D action and metabolism within target cells continue to expand, sometimes in unexpected directions. The discovery of the hnRNP-related vitamin D response element binding proteins in subhuman (VDRE-BPs) in human (REBiPs) primate is one such unexpected direction. The potential that these multifunctional nucleic acid binding proteins can alter hormone-directed transcription of genes in
361
CHAPTER 21 Response Element Binding Proteins
addition to controlling the posttranscriptional fate of mRNAs transcribed from those genes indicates that the traditional views of the various nuclear machines (i.e., transcriptosome, sliceosome, etc.) working independently of one another is no longer tenable. Adding to the untenable nature of these traditional views is the ability of these multifunctional hnRNPs to recruit the expression of other classes of proteins, such as the hsp70 family of intracellular vitamin D binding proteins, not previously known to be important in the chaperoning of small molecules, to further modify to the cellular responsiveness to vitamin Ds. Although originally discovered as elements in the cellular vitamin D regulatory system, the hnRNP and hsp family of proteins are also active in the control of estrogen action. Chen and colleagues [20,33] have discovered proteins in the hnRNP C family that act in a dominant-negative mode to squelch estrogen–estrogen response element (ERE)-directed transactivation. In a manner similar to that observed for the VDRE-BPs and REBiPs with the VDRE, these estrogen response element binding proteins (ERE-BPs) act in trans to prevent estrogen receptor (ER) dimer binding to the ERE. Most recently, Chen et al. [26] have also discovered an intracellular estrogen binding protein (IEBP). This protein is in the hsp27 family. Although the co-overexpression of IEBP is observed in the presence of ERE-BP overexpression, the role the IEBP plays, if any, in ERERE-directed transactivation is still under analysis.
11.
12. 13.
14. 15. 16. 17.
18.
19.
20.
References 1. Pilbeam D 1984 The descent of hominoids and hominids. Sci Am 250:84–96. 2. Bland Sutton JB 1884 Observation on rickets etc. in wild animals. J Anat 18:363–397. 3. Krook L, Barrett RB 1962 Simian bone disease—A secondary hyperparathyroidism. Cornell Vet 52:459–492. 4. Hershkovitz P 1977 Living New World Monkeys (Platyrrhini): With an Introduction to Primates. University of Chicago Press, Chicago. 5. Hunt RD, Garcia FG, Hegsted DM 1967 A comparison of vitamin D2 and D3 in New World primates. I. Production and regression of osteodystrophia fibrosa. Lab Anim Care 17:222–234. 6. Steenbock H, Kletzein SWF, Halpin JG 1932 The reaction of the chicken to irradiated ergosterol and irradiated yeast as contrasted with the natural vitamin D of fish liver oils. J Biol Chem 97:249–266. 7. Marx SJ, Jones G, Weinstein RS, Chrousos GP, Renquist DM 1989 Differences in mineral metabolism among nonhuman primates receiving diets with only vitamin D3 or only vitamin D2. J Clin Endocrinol Metab 69:1282–1290. 8. Hay AW 1975 The transport of 25-hydroxycholecalciferol in a New World monkey. Biochem J 151:193–196. 9. Bouillon R, van Baelen H, de Moor P 1976. The transport of vitamin D in the serum of primates. Biochem J 159:463–466. 10. Shinki T, Shiina Y, Takahashi N, Tanioka Y, Koizumi H, Suda T 1983 Extremely high circulating levels of 1 alpha,
21.
22.
23. 24. 25.
26.
27.
25-dihydroxyvitamin D3 in the marmoset, a New World monkey. Biochem Biophys Res Commun 114:452–457. Adams JS, Gacad MA, Baker AJ, Gonzalez B, Rude RK 1985 Serum concentrations of 1,25-dihydroxyvitamin D3 in platyrrhini and catarrhini: A phylogenetic appraisal. Am J Primatol 19: 219–224. Liberman UA, de Grange D, Marx SJ 1985 Low affinity of the receptor for 1-alpha,25-dihydroxyvitamin D3 in the marmoset, a New World monkey. FEBS Lett 182:385–388. Adams JS, Gacad MA, Baker AJ, Kheun G, Rude RK 1985 Diminished internalization and action of 1,25-dihydroxyvitamin D3 in dermal fibroblasts cultured from New World primates. Endocrinology 116:2523–2527. Gacad MA, Adams JS 1991 Endogenous blockade of 1,25dihydroxyvitamin D–receptor binding in New World primate cells. J Clin Invest 87:996–1001. Gacad MA, Deseran MW, Adams JS 1992 Influence of ultraviolet B radiation on vitamin D3 metabolism in New World primates. Am J Primatol 28:263–270. Gacad MA, Adams JS 1992 Specificity of steroid binding in New World primate B95–8 cells with a vitamin D-resistant phenotype. Endocrinology 131:2581–2587. Adams JS, Gacad MA 1988 Phenotypic diversity of the cellular 1,25-dihydroxyvitamin D3–receptor interaction among different genera of New World primates. J Clin Endocrinol Metab 66:224–229. Gacad MA, Adams JS 1993 Identification of a competitive binding component in vitamin D-resistant New World primate cells with a low affinity but high capacity for 1,25-dihydroxyvitamin D3. J Bone Miner Res 8:27–35. Arbelle JE, Chen H, Gacad MA, Allegretto EA, Pike JW, Adams JS 1996 Inhibition of vitamin D receptor–retinoid X receptor-vitamin D response element complex formation by nuclear extracts of vitamin D–resistant New World primate cells. Endocrinology 137:786–789. Chen H, Arbelle JE, Gacad MA, Allegretto EA, Adams JS 1997 Vitamin D and gonadal steroid-resistant New World primate cells express an intracellular protein which competes with the estrogen receptor for binding to the estrogen response element. J Clin Invest 99:669–675. Chen H, Hu B, Allegretto EA, Adams JS 2000 The vitamin D response element–binding protein. A novel dominant-negative regulator of vitamin D–directed transactivation. J Biol Chem 275:35557–35564. Chun RF, Chen H, Boldrick L, Sweet C, Adams JS 2001 Cloning, sequencing, and functional characterization of the vitamin D receptor in vitamin D–resistant New World primates. Am J Primatol 54:107–118. Dreyfuss G, Matunis MJ, Pinol-Roma S, Burd CG 1993 hnRNP proteins and the biogenesis of mRNA. Annu Rev Biochem 62:289–321. Horwitz KB, Jackson TA, Bain DL, Richer JK, Takimoto GS, Tung L 1996 Nuclear receptor coactivators and corepressors. Mol Endocrinol 10:1167–1177. Hewison M, Rut AR, Kristjansson K, Walker RE, Dillon MJ, Hughes MR, O’Riordan JL 1993 Tissue resistance to 1,25-dihydroxyvitamin D without a mutation of the vitamin D receptor gene. Clin Endocrinol (Oxf) 39: 663–670. Chen H, Hewison M, Hu B, Adams JS 2003 Heterogeneous nuclear ribonucleoprotein (hnRNP)-binding to hormone response elements: a novel cause of vitamin D resistance. Proc Natl Acad Sci USA 100:6109–6114. Donev RM, Doneva TA, Bowen WR, Sheer D 2002 HnRNPA1 binds directly to double-stranded DNA in vitro within a 36 bp sequence. Mol Cell Biochem 233:181–185.
362 28. Dreyfuss G, Kim VN, Kataoka N 2002 Messenger-RNAbinding proteins and the messages they carry. Nat Rev Mol Cell Biol 3:195–205. 29. Orphanides G, Reinberg D 2002 A unified theory of gene expression. Cell 108:439–451. 30. Horn PJ, Peterson CL 2002 Molecular biology. Chromatin higher order folding—wrapping up transcription. Science 297: 1824–1827. 31. Michelotti GA, Michelotti EF, Pullner A, Duncan RC, Eick D, Levens D 1996 Multiple single-stranded cis elements are associated with activated chromatin of the human c-myc gene in vivo. Mol Cell Biol 16:2656–2669. 32. Martens JH, Verlaan M, Kalkhoven E, Dorsman JC, Zantema A 2002 Scaffold/matrix attachment region elements interact with a p300-scaffold attachment factor A complex and are bound by acetylated nucleosomes. Mol Cell Biol 22: 2598–2606. 33. Chen H, Hu B, Gacad MA, Adams JS 1998 Cloning and expression of a novel dominant-negative-acting estrogen response element-binding protein in the heterogeneous nuclear ribonucleoprotein family. J Biol Chem 273:31352–31357. 34. Yin YW, Steitz TA 2002 Structural basis for the transition from initiation to elongation transcription in T7 RNA polymerase. Science 298:1387–1395. 35. Zhou Z, Licklider LJ, Gygi SP, Reed R 2002 Comprehensive proteomic analysis of the human spliceosome. Nature 419:182–185. 36. Proudfoot NJ, Furger A, Dye MJ 2002 Integrating mRNA processing with transcription. Cell 108:501–512. 37. Auboeuf D, Honig A, Berget SM, O’Malley BW 2002 Coordinate regulation of transcription and splicing by steroid receptor coregulators. Science 298:416–419. 38. Hilleren P, McCarthy T, Rosbash M, Parker R, Jensen TH 2001 Quality control of mRNA 3′-end processing is linked to the nuclear exosome. Nature 413:538–542.
JOHN S. ADAMS
39. Reed R, Hurt E 2002 A conserved mRNA export machinery coupled to pre-mRNA splicing. Cell 108:523–531. 40. Nakielny S, Dreyfuss G 1999 Transport of proteins and RNAs in and out of the nucleus. Cell 99:677–690. 41. Zhu J, Mayeda A, Krainer AR 2001 Exon identity established through differential antagonism between exonic splicing silencer-bound hnRNP A1 and enhancer-bound SR proteins. Mol Cell 8:1351–1361. 42. Gacad MA, Chen H, Arbelle JE, LeBon T, Adams JS 1997 Functional characterization and purification of an intracellular vitamin D–binding protein in vitamin D–resistant new world primate cells. Amino acid sequence homology with proteins in the hsp-70 family. J Biol Chem 272:8433–8440. 43. Gacad MA, Adams JS 1998 Proteins in the heat shock-70 family specifically bind 25-hydroxyvitamin D3 and 17betaestradiol. J Clin Endocrinol Metab 83:1264–1267. 44. Hartl FU 1996 Molecular chaperones in cellular protein folding. Nature 381:571–579. 45. Adams JS, Chen H, Chun RF, Nguyen L, Wu S, Ren SY, Barsony J, Gacad MA 2003 Novel regulators of vitamin D action and metabolism: Lessons learned at the Los Angeles Zoo. J Cell Biochem 88:308–314. 46. Wu S, Ren S, Chen H, Chun RF, Gacad MA, Adams JS 2000 Intracellular vitamin D binding proteins: novel facilitators of vitamin D-directed transactivation. Mol Endocrinol 14:1387–1397. 47. Wu S, Chun R, Gacad MA, Ren S, Chen H, Adams JS 2002 Regulation of 1,25-dihydroxyvitamin D synthesis by intracellular vitamin D binding protein-1. Endocrinology 143:4135. 48. Nykjaer A, Dragun D, Walther D, Vorum H, Jacobsen C, Herz J, Melsen F, Christensen EI, Willnow TE 1999 An endocytic pathway essential for renal uptake and activation of the steroid 25-(OH) vitamin D3. Cell 96:507–515. 49. Christensen EI, Willnow TE 1999 Essential role of megalin in renal proximal tubule for vitamin homeostasis. J Am Soc Nephrol 10:2224–2236.
CHAPTER 22
Vitamin D Receptor and Retinoid X Receptor Subcellular Trafficking JULIA BARSONY
Laboratory of Cell Biochemistry and Biology, National Institutes of Diabetes, Digestive and Kidney Diseases, NIH, Bethesda, Maryland
I. Evolving Concepts of Receptor Localization II. Nucleocytoplasmic Trafficking of Vitamin D Receptor and Retinoid X Receptor
III. Intranuclear Trafficking of Vitmain D Receptor and Retinoid X Receptor IV. Summary and Future Directions References
I. EVOLVING CONCEPTS OF RECEPTOR LOCALIZATION
proposed by E. V. Jensen (Fig. 1A) for the mechanism of steroid receptor actions in the 1960s [11]. Later, antibodies were raised against steroid receptors and were used for immunocytochemical detection. Results of these studies indicated that even the unliganded nuclear receptors reside exclusively in the nucleus. This initiated a dramatic shift in the paradigm of nuclear receptor localization. The evidence for receptor translocation was widely criticized as artifactual, and the preferred model advocated that both unliganded and liganded nuclear receptors are anchored to target gene sequences [12] (Fig. 1B). Opponents of this static model demonstrated that detergent treatment of cell membranes, which was necessary for getting the antibodies inside the cell, washed out many of the cytoplasmic receptors [13], and they considered the exclusively nuclear receptor localization an artifact. Soon after the cloning of VDR, antibodies were generated and used for the detection of VDR immunoreactivity in fixed cells [14]. Not too surprisingly, VDR immunoreactivity was initially detected in the nucleus both before and after calcitriol treatment [14], which was consistent with the model depicted in Fig. 1B. Subsequently, we developed a more sensitive immunocytology technique by replacing chemical fixation with microwave fixation. This fixation allowed antibody access to receptors without detergent use. Our studies with this method showed unliganded VDR in the cytoplasm and an apparent translocation of cytoplasmic VDR into the nucleus after calcitriol treatment [15], consistent with the model depicted in Fig. 1A. This concept was further supported by our studies with a biologically active fluorescent labeled calcitriol in living cells. Results of these experiments also indicated that high affinity binding sites reside in the cytoplasm and accumulate in the nucleus over time [16], which is consistent with the model depicted in Fig. 1A. Debate about VDR localization
It is well established that vitamin D receptor (VDR) mediates most, if not all, actions of calcitriol in target cells. Vitamin D receptors (VDRs) belong to the type II nuclear receptor family of proteins, which all act as heterodimers with retinoid X receptors (RXRs) to regulate transcription of their target genes. VDR undergoes a rapid activation process upon binding of calcitriol. This activation process involves multiple protein–protein interactions, conformational change, phosphorylation, and interaction with specific target gene sequences. An important aspect of VDR functioning is the temporal and spatial aspect of activation events. These events include protein–protein interactions: dimerization with RXR, interactions with coactivators [1], co-repressors [2,3], chaperones [4], kinases [5,6], calreticulin [7], and with nuclear import and export receptors [8]. Several of these protein interactions influence the binding of VDR and RXR to target genes and the interaction with the transcription machinery. Other protein interactions influence VDR localization and stability. Our understanding of the mechanisms of VDR functions has changed considerably over the past few years, in part due to new insights into VDR and RXR subcellular localization and trafficking. Even before the identification of VDR in the early 1980s, there was great interest in locating the receptor for calcitriol as a way to gain insight into the mechanism of calcitriol actions. Early studies with biochemical methods, such as radioligand and cell fractionation, detected the unliganded receptor for calcitriol mostly in the cytosolic fractions of target cells, whereas the liganded VDR was recovered mostly in nuclear fractions [9,10]. These findings supported the model of VDR translocation, which was similar to the model
364
JULIA BARSONY
RXR/VDR
VDR
A
Translocation
RXR/VDR
B
Anchored
RXR/VDR
RXR/VDR
C
Shuttling
FIGURE 1 Models of VDR/RXR Localization. (A) Translocation model. Model depicts the RXR in the nucleus and the unliganded VDR in both the cytoplasm and the nucleus. It stipulates that binding of calcitriol (red diamond) initiates VDR translocation from the cytoplasm into the nucleus, where dimerization of VDR with RXR would take place. (B) Anchored model. Model depicts both the unliganded and the liganded VDR and the RXR in the nucleus, attached to specific DNA recognition sequences. It assumes that ligand binding does not change receptor localization but initiates VDR recruitment of RXR and coactivators and an interaction with the transcription machinery. (C) Shuttling model. Model depicts both unliganded and liganded VDR forming dimers with RXR in the cytoplasm and in the nucleus. It assumes that change in the shuttling speed of VDR after hormone binding reflects the formation of a distinct macromolecular complex. See CD-ROM for color.
persisted until recently, and was only resolved when the use of green fluorescent protein (GFP) chimeras and modern imaging methods revealed the highly dynamic nature of VDR, RXR, and most other nuclear receptors. Transcriptionally active fluorescent protein chimeras of VDR were first generated in our laboratory [17] and later in other laboratories [18,19]. We detected the unliganded GFP-VDR both in the cytoplasm and the nucleus, and showed that after calcitriol addition GFPVDR accumulates in the nucleus [17]. A subsequent study found more unliganded GFP-VDR in the nucleus after fixation than we did in living cells [18]. Nonetheless, this study [18] and another study by Sunn et al. [19] both detected hormone-induced GFP-VDR translocation into the nucleus. These initial experiments with GFP-VDR, like previous studies using immunocytology, observed the steady-state distribution of receptors and appeared to favor the translocation model (Fig. 1A). More recently, we conducted kinetic analysis of VDR trafficking using photobleaching and in vitro transport assays. Results of these experiments revealed that unliganded VDR and liganded VDR both shuttle constantly between the cytoplasm and the nucleus [20] and suggested a novel dynamic model for VDR and RXR localization (Fig. 1C). This kinetic analysis revealed that the nucleocytoplasmic exchange rate is faster for the liganded GFP-VDR than for the unliganded GFPVDR [21], indicating that a faster nuclear import of the liganded VDR could give the appearance of hormoneinduced VDR translocation. Almost certainly, different protein–protein interactions explain the difference between the transport characteristics of the unliganded and the liganded VDR [22]. The most likely interacting partner of VDR that could influence hormonedependent change in VDR distribution is RXR. Dimerization with RXR is essential for both constitutive and calcitriol-dependent functions of VDR. Recently this has been elegantly demonstrated by knockout mouse studies [23–26]. Therefore, the subcellular localization of RXR and the location of dimerization are important aspects of calcitriol actions. Historically, immunocytology experiments led to the general consensus favoring exclusively nuclear localization of RXR. Also in living cells, microscopy showed fluorescent protein chimeras of RXR predominantly in the nucleus [27,28]. The translocation model for VDR actions was consistent with these findings (Fig. 1A), and assumed that VDR dimerizes with RXR in the nucleus after hormone binding. However, three lines of evidence contradicted this model. First, our fluorescence resonance energy transfer (FRET) experiments demonstrated the presence of VDR/RXR heterodimer not only in the nucleus, but also in the cytoplasm [27] (Fig. 2). Second, coexpression experiments showed that the wild-type RXR could bring
365
CHAPTER 22 VDR and RXR Subcellular Trafficking
A
B
FIGURE 2
Fluorescence Energy Transfer (FRET) Visualizes Interactions of Unliganded GFP-VDR and RXR-BFP. FRET signal is generated when the green VDR and blue RXR are closer than 50 Å. High FRET intensities are labeled with green pseudocolor and low background intensities with blue pseudocolor. Cells expressed fluorescent protein chimeras of wild-type VDR and either the wild-type RXR-BFP (A) or the heterodimerization incompetent mutant of RXR-BFP (B). Cells were cultured without serum or calcitriol before the experiment. FRET is most abundant in the nuclei, but also visible in the cytoplasm. Images were taken from living cells by confocal microscope (Zeiss LSM410). Bars = 10 µm. See CD-ROM for color.
into the nucleus a nuclear import mutant of VDR [20] (see detail in Section II). Third, photobleaching and in vitro transport assays demonstrated that unliganded and liganded RXR rapidly shuttle between the cytoplasm and the nucleus [21] (Fig. 3). Taken together these findings supported the novel, dynamic model for VDR/RXR localization (Fig. 1C). According to this model, both VDR and RXR constantly interact with large protein complexes to facilitate ongoing nuclear import and export and intranuclear trafficking.
II. NUCLEOCYTOPLASMIC TRAFFICKING OF VITAMIN D RECEPTOR AND RETINOID X RECEPTOR A. Nuclear Import and Export Mechanisms Nucleocytoplasmic transport of macromolecules through the nuclear pore complex (NPC) is a fundamental cellular process requiring energy. Whereas the NPC permits unrestrained passage of small molecules (up to 9 nm in diameter) through a central aqueous channel, transport of most proteins (over 20–40 kDa) and RNA can only pass through the NPC in both directions by receptor-mediated and Ran-GTP gradient
dependent active transport [29–32]. Rapid progress in cell biology has revealed the major components of this transport machinery (reviewed in [33,34]). The soluble transport receptors specifically recognize their cargoes. In addition, they often act as chaperone proteins to prevent aggregation of basic proteins [35]. Multiple classes of nuclear transport receptors exist, including the family of importin-β-like proteins (importins/exportins, also known as karyopherins) [36]. There are more than 20 members of this family and they mediate both nuclear import and export of proteins and RNA. The small nuclear transport factor 2 (NTF2/p10) imports the small GTPase Ran into the nucleus [37]. Proteins of the third transport receptor class, a heterodimer of TAP/NXF and p15/NXT, mediate mRNA export from the nucleus. All transport receptors interact with components of the NPC and contain a phenylalanine/glycine-rich motif. How do these transport receptors work? Transport receptors specifically bind their cargo in the originating compartment and then release them in the destination compartment. The regulation of this process by RanGTPase is best understood for transport mediated by importins and exportins (Fig. 4). Both importins and exportins bind to Ran-GTP through their conserved amino-terminal domains. The concentration of RanGTP is high in the nucleus, whereas the concentration of Ran-GDP is high in the cytoplasm. This gradient is maintained by Ran-GTPase activating protein and
366
JULIA BARSONY
Pre-bleach
Post-bleach
Recovery
Time (min)
GFP-VDR Vehicle
15–30
GFP-VDR Calcitriol
5–15
YFP-RXR Vehicle
20–30
FIGURE 3 Fluorescence recovery after photobleaching demonstrates that VDR and RXR shuttle between the cytoplasm and nucleus. COS-7 cells expressing GFP-VDR (upper and middle panels) were treated with vehicle (0.1% ethanol) or calcitriol (5 nM) for 1 hr before microscopy. COS7 cells expressing YFP-RXR (lower panels) were treated with vehicle. Cells with two nuclei were selected (left) and a 2 µm2 spot (4×4 pixel) within one of the nuclei (arrow) was exposed to a focused laser beam at full power for 1 sec/pixel. This intense illumination caused irreversible photobleaching (darkening) of all the receptors within the exposed nucleus, whereas the receptors in the other nucleus remained bright. After photobleaching, serial images were taken for measurements of fluorescence recovery. The images taken immediately after photobleaching are shown in the center panels and images taken at complete recovery are shown in the right panels. Range of recovery time (right) was calculated from 10 independent experiments. This is the time for reaching equilibrium of brightness between the two nuclei of the cell, which occurs due to the passage of bright and dark receptors through the cytoplasm. Bars, 10 µm. (See color plate).
Ran guanine nucleotide exchange factor. The gradient of Ran-GTP and its role in mediating cargo release from import/export receptors have been confirmed in living cells using FRET [38]. Nuclear import is mediated by import receptors, such as importin β (Fig. 4). Importins bind either directly or indirectly through adapter proteins to their cargo proteins in the cytoplasm without binding to Ran. Importin α is the most important adapter protein, which binds to basic amino acid–rich sequences on their cargo proteins, so-called nuclear localization sequences (NLSs). Importin α then binds to proteins of the importin β family, and the import receptor complex interacts with proteins of the nuclear pore complex in a Ran-dependent manner. It was reported that large transport cargoes might need more than one receptor
molecule for efficient translocation, because otherwise partitioning of the transport complex into the hydrophobic phase of the NPC may be hindered [39]. After crossing the nuclear pore complex, cargoes are released in the nucleus upon Ran-GTP binding. The importin/Ran-GTP complex is then recycled back to the cytoplasm. Export from the nucleus is mediated by export receptors (Fig. 4). Ran-GTP induces cargo binding to export receptors exclusively in the nucleus. After export to the cytoplasm, GTP hydrolysis by Ran-GAP and RanBP1/2 allows separation of the exportin/cargo/RanGTP complex. Ran is then recycled back to the nucleus by the importer NTF2. Nuclear export of proteins often involves multiple mechanisms; many of these mechanisms are poorly understood. The best-characterized
367
CHAPTER 22 VDR and RXR Subcellular Trafficking
GTP hydrolysis GDP
Cytoplasm
Importin
Exportin
NPC Nucleus Exportin
Importin GTP
GTP
Ran-GTP
GDP
Ran-GDP
GTP
Ran-GTP binding
Cargo
FIGURE 4 Schematic representation of nuclear import and export mechanisms. Ran-GTP gradient confers directionality of nuclear import and export receptor movement across the nuclear pore complex (NPC). Adapter proteins, such as importin α, recognize nuclear localization signals (NLS) on their cargo in the cytoplasm. This complex is stabilized by the binding of the import receptors importin β, which interacts with the importin β-binding (IBB) domain of the cargoloaded importin α. This triple import complex then docks to the cytoplasmic side of the NPC via importin β and translocates to the nuclear side of the NPC. There binding of Ran-GTP terminates translocation, leading to the displacement of importin α–cargo protein from the import complex and the formation of an importin β–Ran-GTP complex. The C-terminal-IBB-domain of the cargofree importin α binds to the export receptor CAS cooperatively with Ran-GTP, and is recycled back to the cytoplasm. The Ran-GTP bound importin β translocates back to the cytoplasm either alone or with cargoes that can directly bind to importin β. In the cytoplasm, the activities of Ran-binding protein 1 (RanBP1) and the Ran-specific GTPase-activating protein 1 (Ran-GAP) result in the hydrolysis of GTP on Ran, which leads to the disassembly of export complexes. CAS recycles back to the nucleus by passing through the NPC. Another import receptor, NTF2, together with its dimerization partner takes Ran into the nucleus. Export receptors bind with their cargo and with Ran-GTP in the nucleus. After this complex crosses the NPC, hydrolysis of GTP leads to the disassembly of the export complex, and exportin recycles back to the nucleus. See CD-ROM for color.
export pathway uses the export receptor called chromosomal region maintenance 1 protein (Crm-1) or exportin 1. Cargoes for Crm-1 are proteins that carry a leucinerich nuclear export signal (NES) [40] and RNA [41]. Leptomycin B (LMB) is a specific inhibitor of Crm-1 mediated export; it prevents binding of cargo proteins to Crm-1 [42]. Even without an NES, proteins can be exported from the nucleus by Crm-1 using adaptor proteins [43]. In addition, other Ran-independent export mechanisms carry mRNA and ribosomal subunits
out of the nucleus [44]. Newly identified exportins are CAS, an importin family member that binds to the large NES in the importin family of proteins, and exportin-t, which is responsible for tRNA export [45]. Calreticulin could also be an export receptor. It binds to specific sequences in the DNA-binding domains of nuclear receptors, and this binding is essential for the efficient export of these receptors [8,46]. However, the direct interaction of calreticulin with the nuclear pore complex has not been demonstrated, leaving open the possibility
368
JULIA BARSONY
that it is not an export receptor but an adaptor protein. Adding to the complexity of nuclear import and export mechanisms is indirect transport, utilizing a “piggyback” mechanism through cargo association with other exported proteins or RNA [47]. The recent rapid progress in the identification of import and export mechanisms for nuclear receptors is fueled by the understanding that receptor interactions with the import and export machinery represent a critical mechanism that regulates their transcriptional activities. During the past 2 years, several kinetic studies with fluorescent chimeras of nuclear receptors demonstrated that most of them shuttle constantly between cytoplasm and nucleus [48]. Their import is mediated through binding to importins. One or two bipartite NLSs have been found in the DNA-binding domain (DBD) of most nuclear receptors. In addition to the NLSs in the DBD, constitutive NLS was identified in the hinge region of the androgen receptor (AR) [49]. Ligandregulated NLS was found in the ligand-binding domain (LBD) of the glucocorticoid receptor (GR) [50] and the AR [51]. Several nuclear proteins utilize other import pathways; their arginine-rich NLS binds directly to importin β [52]. Finally, indirect nuclear import through a piggyback mechanism is also possible. Export of nuclear receptors is still not well defined. Early studies indicated that the export of progesterone receptors and estrogen receptors are mediated by passive diffusion [53]. More recent studies in living cells with fluorescent protein chimeras revealed that the export of nuclear receptors is also mediated by active transport. First, we demonstrated that Crm-1 mediates export of VDR. This prompted studies that revealed the Crm-1-mediated export of estrogen receptor (ER), TR, and retinoic acid receptor (RAR) [54]. GR, AR, and progesterone receptor (PR) did not appear to be DBD 24
exported by Crm-1 [46,55,56]. however, sequences important for the seemingly NLS-dependent export of GR, AR, and PR have been identified in the LBD [56]. Other reports indicated that calreticulin regulate export by interacting with conserved amino acids in the DBD of nuclear receptors. In addition to direct export by export receptors, nuclear receptors may be exported by a piggyback mechanism through association with other shuttling proteins or RNA [47]. Such indirect export has been reported for the RXR, via the phosphorylationdependent export of the orphan nuclear receptor NGFI-B that contains leucine-rich NESs [57]. Another example is the 14-3-3 protein that interacts with RIP140 co-repressor and exports it from the nucleus, thus enhancing transcriptional activity of GR [58]. Because RIP140 binds to both VDR [59] and RXR [60], this mechanism may be important for the regulation of calcitriol action. Current research is likely to elucidate novel export pathways of nuclear receptors and to resolve existing controversies. The complexities of nuclear import and export mechanisms for nuclear receptors provide a versatile regulatory pathway of hormone actions. Ongoing research will elucidate the spatial–temporal relationship between receptors and co-regulators [61,62] and the regulation and the functional significance of receptor export [63,64].
B. Interactions of VDR and RXR with Nuclear Import and Export Receptors Using wheat germ agglutinin in permeabilized cells, we were the first to demonstrate that VDR interacts with the NPC [15]. Subsequently, putative nuclear localization sequences were identified within the VDR (Fig. 5).
Hinge region 89
LBD 190
N
AF-2 427 C
NLS1
NLS2
NLS3
NLS4 154RPPVRVNDGGGSHPSRPNSR
69TKDNRRH 102RKREMILKRK
45KGFFRRSMKRKA
FIGURE 5 Schematic representation of hVDR functional domains and the putative nuclear localization sequences (NLSs). Amino acids representing the NLSs are shown in bold letters. Mutations of NLS2, NLS3, and NLS4 did not cause cytoplasmic retention of the receptor in living cells; their location in the VDR is marked with pink. See CD-ROM for color.
CHAPTER 22 VDR and RXR Subcellular Trafficking
The bipartite NLS between the two zinc-fingers of the DBD is homologous to the NLS found in steroid receptors (NLS1). Mutations of this region resulted in cytoplasmic retention of both unliganded and liganded VDR [65]. We later confirmed that this is a functional NLS in living cells by generating mutations in NLS1 of the GFP-VDR (K53Q, R54G, K55E) [66]. Another cluster of basic amino acids (NLS2) in the second zinc finger is homologous to the NLS found in the orphan nuclear receptor TR2 [67]. Mutations of these basic amino acids to lysine did not cause cytoplasmic retention of GFP-VDR (Barsony J. and Prufer K., unpublished); thus this segment is unlikely to function as NLS. Two more clusters of basic amino acids within the VDR appeared initially to function as NLSs. Michigami and colleagues showed that a short segment of 20 amino acids in the hinge-region of VDR (NLS4; 154–173) enabled cytoplasmic GFP-tagged alkaline phosphatase to translocate to nuclei [18]. Surprisingly, in the context of the intact receptor we were unable to confirm that this segment functions as NLS using mutational analysis. Mutations at amino acids 154–158 (R154G, P155A, P156A, R158G) and at amino acids 169–173 (R169G, P170A, R173G) did not prevent nuclear import of GFPVDR (Barsony J. and Prufer K., unpublished). Another peptide representing amino acids 102–111 of VDR has also been shown to enable nuclear accumulation of the fluorescein-labeled IgG [68]. We deleted these amino acids 102–110 (NLS3) of the GFP-VDR without significant impairment of the nuclear import of the unliganded VDR (Fig. 6). In addition to constitutive NLSs, VDR segments that are important for the calcitriol-induced redistribution have been defined (Fig. 6). We have shown that liganddependent acceleration of GFP-VDR import is disturbed by deletion or mutations in the activation-function 2 (AF2) domain [17,21]. This finding indicated that coactivators could play a role in the regulation of VDR import. Deletion of NLS3 also abolished hormonedependent nuclear accumulation of GFP-VDR (Fig. 6). Coincidentally, the residues 105–109 within this putative NLS of VDR binds to the hsp70 [69]. The loss of hsp70 binding to VDR has no effect on hormone binding and coactivator interactions [69]. Therefore, this finding could indicate a role for hsp70 in the regulation of nuclear import of the liganded VDR. RXR also interacts with import receptors. Figure 7 shows a putative NLS in the first zinc finger of RXR, which is a homolog of the NLS1 of VDR. To test the functional significance of this NLS, we introduced point mutations at amino acids 160, 161, 164, and 165 (K160Q, R161G, R164G, and K165Q) [20]. Microscopy showed that the mutant nls1YFP-RXR is retained in the cytoplasm (Fig. 7). An additional NLS is located between amino acids 181 and 186 within the second
369 zinc finger of RXR (NLS2). Mutations in this segment (R182E, R184E) caused cytoplasmic retention of the nls2YFP-RXR (Fig. 7). Taken together, these findings show that both VDR and RXR interact with nuclear import receptors via multiple binding sites. We used coexpression experiments with wild-type and NLS mutants of VDR and RXR to study the impact of dimerization on receptor nuclear import (Fig. 8). Coexpression of wild-type RXR-BFP restored nuclear localization of nls1GFP-VDR, whereas coexpression of nls1RXR-BFP retained the wild-type GFP-VDR in the cytoplasm. Moreover, calcitriol treatment induced nuclear accumulation nls1RXR-BFP when it was coexpressed with GFP-VDR. These data indicate that RXR dominates the nuclear import of unliganded VDR, whereas liganded VDR dominate the nuclear import of RXR. In addition, these data strongly support the notion that VDR and RXR dimerize in the cytoplasm. Similar to other nuclear receptors, VDR and RXR also exit the nucleus through the NPC [21]. We studied this export by conventional permeabilization experiments and by monitoring fluorescence recovery after photobleaching (FRAP) (Fig. 3). This later technique allowed us to characterize the nucleocytoplasmic shuttling speed; the recovery half-times were 15–30 min for the unliganded VDR and 5–15 min for liganded VDR, and 20–30 min for unliganded RXR [20]. Results indicated that both VDR and RXR have to interact frequently with nuclear import and export receptors. Nuclear export of unliganded GFP-VDR was inhibited by LMB treatment, resulting in an exclusively nuclear receptor accumulation (Fig. 9). This finding indicates that the unliganded VDR is exported by Crm-1. Calcitriol treatment prevented LMB sensitivity of VDR, suggesting that this faster VDR export is mediated by a Crm-1 independent mechanism. We searched for the NES of VDR by mutating the leucine-rich sequences. Mutations of the leucines 320-325 to alanines in the LBD of VDR abrogated LMB sensitivity [21]. In addition, nuclear export of GFP-VDR was inhibited by mutations at the conserved calreticulin-binding site in the DBD (F47A, F48A). Nuclear export of RXR was also prevented by mutations in the calreticulin-binding site (F158A, F159A), but the RXR export was not inhibited by treatment with LMB [21]. Interestingly, RXR can also be exported by an indirect mechanism. Katagiri and colleagues showed that treatment with NGF induced phosphorylation of orphan nuclear receptor NGFI-B (also called nur77) at amino acid 105, and this phosphorylation induced NGFI-B export together with the export of the dimerization partner, RXR [57]. It is possible that RXR export by other dimerization partners could influence VDR nucleocytoplasmic trafficking. The research of nuclear export pathways is rapidly progressing and is likely to yield a better insight into
370
JULIA BARSONY
Vehicle
Calcitriol
GFP-VDR
FIGURE 6
Intact binding sites for coactivators and heat shock protein 70 are required for calcitriol-dependent nuclear accumulation of VDR. Leucines 417 and 420 were changed to alanines to create cofGFP-VDR, and amino acids 102-110 were deleted to create ∆hsp70GFP-VDR (also referred to as NLS3). COS-7 cells were transfected with expression vectors 24 hr before microscopy and were treated with 10 nM calcitriol for the last 1 hr in serum- and phenol red-free medium. Calcitriol induced nuclear accumulation of GFP-VDR, but failed to change the nucleocytoplasmic distribution of cofGFPVDR and ∆hsp70GFP-VDR. Images were taken from living cells by confocal microscope (Zeiss LSM410). Bars, 10 µm. See CD-ROM for color.
cof GFP-VDR
∆hsp 70GFP-VDR
DBD 140
Hinge 206
LBD
AF-2
265
467
N
C
FIGURE 7 Schematic represen-
NLS1 NLS2
tation of RXR functional domains and the putative nuclear localization sequences (NLSs). Images are shown from HEC293 cells stably expressing wild-type YFP-RXR or its mutants as indicated. Mutation in NLS1 and NLS2 both cause cytoplasmic localization of YFPRXR. Images were taken from living cells by confocal microscope (Zeiss LSM410). Bars, 10 µm. (See color plate).
YFP-RXR
NLS1YFP-RXR
NLS2YFP-RXR
371
CHAPTER 22 VDR and RXR Subcellular Trafficking
GFP-VDR
GFP-VDR+ nlsRXR-BFP
the mechanisms of VDR and RXR export in the near future.
C. Importance of VDR/RXR Shuttling for Transcriptional Activities Vehicle
Calcitriol
nlsGFP-VDR
nlsGFP-VDR +RXR-BFP
Vehicle
Calcitriol
In vivo, the importance of nuclear import and export for calcitriol actions has been demonstrated by experiments on fibroblasts from patients with hereditary resistance to calcitriol [15]. As expected, mutations in the NLSs of the VDR and RXR inhibited calcitriol-induced transcriptional activities. Luciferase reporter assays (Fig. 10) showed that coexpression of nls1GFP-VDR with nls1RXR-BFP completely blocked transcriptional activity, and even the nls1GFP-VDR expressed alone had minimal transcriptional activity. The activity of nls1GFP-VDR was substantially increased by coexpression of wild-type RXR. Coexpression of nls1RXR-BFP with wild-type GFP-VDR reduced calcitriol-induced transcription of 24Ohase/Luc. These results showed a close correlation between transcriptional activities and nuclear localization, indicating the physiological relevance of receptor compartmentalization [20]. Unexpectedly, inhibition of nuclear export also blocked transcriptional activity (Fig. 9C). Treatment of cells expressing GFP-VDR with LMB caused nuclear retention of receptors and prevented the effect of calcitriol on reporter gene expression. LMB was ineffective to change receptor export and transcriptional activity when it was added after calcitriol [20]. Based on these results we suggested for the first time that receptor export contributes to transcriptional activity of VDR. Subsequently, the importance of nuclear export for signaling was demonstrated for other proteins, such as Smad 1[70].
D. Regulation of VDR and RXR Nucleocytoplasmic Trafficking FIGURE 8
VDR and RXR nuclear import by “piggyback” mechanism. HEK293 cells expressed wild-type GFP-VDR alone or coexpressed GFP-VDR and mutant nlsRXR-BFP (upper panels). HEK293 cells expressed mutant nlsGFP-VDR alone or coexpressed nlsGFP-VDR and wild-type RXR-BFP (lower panels). Images show the green signal (GFP), whereas the blue (BFP) signal is not shown. Alone, both nlsGFP-VDR and nlsRXR-BFP sequestered in the cytoplasm. Coexpression of nlsRXR-BFP caused cytoplasmic retention of GFP-VDR. Calcitriol addition induced nuclear accumulation of GFP-VDR despite the presence of nlsRXR-BFP. Calcitriol failed to induce nuclear accumulation of nlsGFP-VDR. Coexpression of RXR-BFP allowed nuclear accumulation of nlsGFP-VDR. These findings suggest that RXR and VDR form dimers in the cytoplasm. Furthermore, these data indicate that RXR dominates the import of the unliganded VDR and liganded VDR dominates the import of RXR. Bars, 10 µm. See CD-ROM for color.
Nucleocytoplasmic partitioning is a regulatory tool for signaling utilized in a wide variety of species, including yeast, plants, and vertebrates. Several regulatory mechanisms have been described: oligomerization; multiple NLSs and NESs acting in concert; importin or exportin overexpression; increase or decrease in the cargo affinity to importins or exportins; cargo sequestration by binding to cytoplasmic or nuclear docking proteins; and cargo NLS or NES masking by conformational change, phosphorylation, acetylation, or protein binding. Various signaling pathways make use of this regulatory potential. Well-characterized examples
372
JULIA BARSONY
A
DBD
Hinge region
24
89
LBD 190
AF-2 427
N
C 47-48 Calreticulin
320-325 Crm-1 Vehicle
Leptomycin B
B
C
16
24-Ohase/Lucactivities Fold Induction
14 12 10 8 6 4 2 0 Vehicle
LeptomycinB
FIGURE 9 Nuclear export of VDR. (A) Schematic representation of VDR functional domains indicating the sequences important for export by calreticulin and by Crm-1 receptors. (B) Treatment with leptomycin B caused nuclear accumulation of GFP-VDR by preventing its Crm-1 mediated export. GL48 cells (a HEK293 derived line stably expressing GFP-VDR) were treated with vehicle (DMSO) or leptomycin B (2 nM for 3 hr) before microscopy. Images of living cells were taken by confocal microscope (Zeiss LSM410). Bar, 10 µm. (C) Inhibition of VDR export decreased calcitriol-induced transcriptional activity. ROSA I cells, which stably express a 24-hydroxylase-luciferase reporter, were first treated with either vehicle (DMSO) or with leptomycin B (2 nM for 3 hr and then treated with 10 nM calcitriol for 24 hr. Luminescence values are expressed as mean ± 1 S.D. See CD-ROM for color.
are NF-κB [71], MAP kinases [72], STAT [73], Smad1 [70], and p53 [74,75]. To date, very little is known about the regulation of VDR and RXR nucleocytoplasmic shuttling, and data regarding the regulation of other nuclear receptors are just starting to emerge. The first insight in the regulation of VDR nucleocytoplasmic trafficking came from coexpression experiments. We found that dimerization with RXR promotes nuclear import and nuclear retention of
VDR [76]. Similarly, dimerization with RXR has also been shown to increase nuclear retention of TR [28]. Calcium/calmodulin is one of the important regulators of nuclear import. This mechanism has been shown to play a critical role for NF-κB/Rel signaling. In most cells, binding of inhibitory I-kB protein (IκB) to NF-κB/Rel masks their NLS and thus NF-κB/Rel are sequestered in the cytoplasm. About 150 different signals can stimulate NF-κB/Rel signaling, including UV radiation,
373
CHAPTER 22 VDR and RXR Subcellular Trafficking
450000
400000
Normalized luciferase activity
350000
300000 250000
200000 150000
100000 50000
0 GFP-VDR
nls GFP-VDR
nls GFP-VDR + RXR-BFP
GFP-VDR + nls RXR-BFP
FIGURE 10
Nuclear import of VDR and RXR correlates with calcitriol-induced transcriptional activities. The mutant nlsGFP-VDR lost almost all transcriptional activities. Coexpression of RXR-BFP partially restored transcriptional activities of nlsGFP-VDR. Coexpression of nlsRXR-BFP with GFP-VDR decreased transcriptional activity of GFPVDR. COS-7 cells were transiently transfected with the receptor expression vectors as indicated, together with 24-hydroxylase-luciferase reporter and ß-galactosidase expression vectors, using Lipofectamine 2000 reagents as described [21]. Normalized luciferase activities from vehicle-treated cells are shown as open bars, and from calcitriol-treated (10 nM for 24 hr) cells are shown as shaded bars. Luminescence data were normalized with β-galactosidase activities, and data are expressed as mean ±1 S.D.
oxidative stress, inflammation, viruses, and bacterial lipopolysaccharides. Cellular stimulation results in phosphorylation, ubiquitination, and degradation of IkB, allowing the translocation of NF-κΒ/Rel to the nucleus and the action on target genes. Calmodulin (CaM) is a highly conserved, ubiquitously expressed Ca2+ binding protein that serves as a key mediator of intracellular Ca2+ signals. After signal-dependent release of Rel-c from the docking, CaM binds Rel-c adjacent to its NLS in a calcium-dependent manner and blocks its nuclear translocation [77]. This mechanism plays an important role in calcium-mediated inhibition of inflammatory response. Another example is the regulation of nuclear factor of activated T cell (NFAT) activity by calcium [78]. NFAT proteins are expressed in most immune system cells and regulate the transcription of cytokine genes critical for the immune response. The activity of NFAT proteins is tightly regulated by the
calcium/calmodulin-dependent protein phosphatase 2B and calcineurin (CaN). Dephosphorylation of NFAT by CaN is required for NFAT nuclear localization. Current immunosuppressive drugs such as cyclosporine A and FK506 block CaN activity, thus inhibiting nuclear translocation of NFAT and consequent cytokine gene transcription. We found that the ligand-dependent nuclear import of GR was not affected by depletion of intracellular calcium stores [79]. Nuclear import of VDR was also unperturbed by calcium depletion (Barsony J., unpublished). In contrast, the export of nuclear receptors is regulated by calcium-dependent binding of calreticulin to their DBD. Depletion of calcium caused nuclear retention of GR, whereas overexpression of calreticulin or cell fusion–induced release of calreticulin from the endoplasmic reticulum enhanced GR export [46]. Similar calcium-dependent regulation of VDR export is expected, since VDR
374 export is also mediated by a calreticulin binding sensitive pathway [20]. Docking in the cytoplasm can retard import and docking in the nucleus can retard export of proteins. Docking of unliganded steroid receptors in the cytoplasm is caused by their binding to heat shock protein 90 (hsp90), which masks their NLS [80]. Whereas hsp90 binding does not retain unliganded VDR and RXR in the cytoplasm, binding to microtubules could serve as potential docking mechanism for VDR in the cytoplasm [81]. There are indications for nuclear docking of steroid receptors. Tethering to the nuclear matrix by the small RING finger protein SNURF has been shown to retard the export of AR after hormone withdrawal [51]. A similar mechanism could be involved in the regulation of VDR export, as an association of VDR with the nuclear matrix has been described [82,83]. Our Triton X-100 permeabilization experiments indicated that about 2% of VDR is docked in the nucleus [21]. Protein interactions that facilitate export of nuclear receptors are emerging. Binding to 14-3-3 protein has been shown to facilitate nuclear export of GR, AR, ER, and TR, but not PPAR and RXR [58,84]. Increased expression of export receptors (calreticulin, Crm-1) has been shown to facilitate export of nuclear receptors [85]. Interaction with cell-cycle regulators could also be important. We observed cell-cycle regulated changes in VDR compartmentalization in HEK293 cells stably expressing GFP-VDR [86]. Nuclear import and export mechanisms that transport coactivators and corepressors in and out of the nucleus adds to the complexity of the regulation of VDR trafficking. For example, SRC-1 has been shown to shuttle between the cytoplasm and the nucleus, exported from the nucleus by Crm-1 [87]. SMRT phosphorylation and nuclear export is associated with inhibition of corepressor function [88]. These and other protein interactions are likely to be important for the control of VDR and RXR trafficking. Receptor modifications, such as acetylation and phosphorylation, represent additional potential regulatory mechanisms of VDR and RXR trafficking. Receptor acetylation at lysine is one of the mechanisms that control transcriptional activities. Important studies have demonstrated that a conserved lysine-rich motif within the hinge region/LBD of the ER and AR is directly acetylated [89,90]. Ligand-induced acetylation promoted coactivator binding and reduced co-repressor binding. Coincidently, the same sequence functions as hormone-dependent NLS of the AR. A potential acetylation signal sequence is also present in the VDR [89]. Therefore, this region could be a regulatory site for nuclear import as well. Phosphorylation is another potential regulatory mechanism of subcellular trafficking. For example, the nuclear retention of thyroid hormone
JULIA BARSONY
receptor alpha1 (TRα1) is mediated by receptor phosphorylation [91]. Exploration of these and other regulatory mechanisms for VDR and RXR trafficking is expected to reveal important and often cell type–specific pathways that modulate hormone actions.
III. INTRANUCLEAR TRAFFICKING OF VITAMIN D RECEPTOR AND RETINOID X RECEPTOR In our original studies using photobleaching in the nuclei of GFP-GR expressing cells, we found that the receptor is in rapid circulation within the nucleus (Fig. 3) [79,92]. Since then, several studies described intranuclear motion of various steroid receptors based on photobleaching experiments [61,62,93–95]. These studies characterized the speed of receptor motion by measuring the half-maximal time for fluorescence recovery, after photobleaching an area within the nucleus. Estrogen binding decreased the mobility of GFP-ER; the t1/2 was 1.6 sec for unliganded, 5.8 sec for estrogen bound [96]. In another study by Stenoien et al., the t1/2 for unliganded ER was less than 1 sec, for estrogen-bound ER 5–6 sec. A similar decrease in the intranuclear mobility of GFP-RAR has been observed after retinoic acid treatment, but not in the mobility of GFP-TR [96]. We found that calcitriol binding also decreased the intranuclear mobility of GFP-VDR; the t1/2 for the unliganded receptor was about 2 sec and for the liganded receptor was 5–7 sec (Barsony et al, unpublished). Hormone treatment induces focal accumulation of most nuclear receptors, including the VDR (Fig. 11) [27]. Mutational analysis indicated that the focal accumulation of GFP-VDR signifies binding to regulatory DNA sequences [76]. A mutation that prevented DNA binding of VDR prevented the formation of discrete bright foci in the nuclei of GFP-VDR–expressing cells (Fig. 11). In addition, microscopy showed that GFPVDR and RXR-BFP bind to the same DNA-binding sites (foci). In cells that coexpressed GFP-VDR and RXRBFP, calcitriol induced accumulation of both receptors in the same foci [97]. Moreover, FRET experiments confirmed that GFP-VDR and RXR-BFP bind to foci as heterodimers [97] (Fig. 11). We used two types of photobleaching techniques to study the dynamics of GFPVDR exchange between foci and areas of diffuse signal. For FRAP, a 0.2 µm2 area (less than 5% of total nuclear area) corresponding to a discrete bright spot within the nucleus was subjected to intense laser illumination for 0.5 sec, which caused the receptors in the illuminated area to completely lose fluorescence. Following this
375
CHAPTER 22 VDR and RXR Subcellular Trafficking
GFP-VDR
dna GFP-VDR
A
GFP-VDR + RXR-BFP
GFP-VDR + hd RXR-BFP
B
FIGURE 11 Calcitriol induces focal accumulation of GFP-VDR and RXR-BFP in the nucleus. (A) GFP-VDR and dnaGFP-VDR were stably expressed in COS-7 cells. Without calcitriol, GFP-VDR is diffusely distributed in the nucleus (not shown). Treatment with calcitriol (10 nM calcitriol for 1 hr) caused formation of bright foci (left). A mutation (R77Q) was introduced in the DNA-binding region of GFP-VDR to create dnaGFPVDR [76]. Calcitriol did not cause focal accumulation of dnaGFP-VDR (right), indicating that intranuclear foci signify DNA binding. (B) FRET signals from the nuclei of COS7 cells stably expressing either GFP-VDR and RXR-BFP (left) or GFP-VDR and a heterodimerization incompetent mutant hdRXR-BFP (right). Cells were treated with calcitriol (10 nM for 1 hr) before microscopy. The high-intensity FRET signal shows that GFP-VDR and RXR-BFP accumulate in the same foci (left). Low, background intensity signal was generated in nuclei of cells expressing GFP-VDR and hdRXR-BFP, indicating lack of dimerization. Images were taken from living cells by confocal microscope (Zeiss LSM410). Bars, 10 µm. (See color plate).
photobleaching, serial images were taken and fluorescence recovery was calculated in the area that was bleached. Data were obtained from at least 20 nuclei of COS-7 cells. For fluorescence loss in photobleaching (FLIP) measurements, a 0.4-µm2 area corresponding to diffuse nuclear signal distant from the selected bright spot was subjected to intense laser illumination for increasing times (0.5 sec to 1 min) in 1-sec increments. Loss of fluorescence was measured over the selected bright spot to detect the time-dependent decrease in
fluorescence intensities, as more and more receptors turned black in the nucleus because of photobleaching. As expected, shorter photoexposure caused loss of fluorescence outside the foci (6.5 sec), and longer photoexposure caused complete loss of fluorescence even in the foci (42 sec). From this difference, the exchange rate between receptor and DNA-binding sites can be calculated. Similar FLIP experiments were performed in COS-7 cells that express YFP-RXR (Barsony et al., unpublished).
376 Results of our photobleaching experiments are in good agreement with results of FRAP and FLIP experiments using GFP-GR. Two groups used fluorescent chimeras of steroid receptors in cells that contain enlarged well-defined DNA-binding sites for these receptors. Interaction of GFP-GR with a large tandem array of MMTV-LTR-ras-BPV have been studied by Hager and colleagues [62,95]. Measuring fluorescence loss in photobleaching (FLIP) and fluorescence recovery after photobleaching (FRAP) they found that the exchange rate between the array and receptors GFP-GR outside the array was about 5 sec [62]. In contrast, measuring the interaction of CFP-LacER with an array of lac operator elements Stenoien and colleagues [93] found that even 20 min of bleaching failed to result in loss of fluorescence over the array after estradiol treatment, suggesting that this receptor was immobilized on the array. Nonetheless, the same group found that coactivator interaction with this array was short; SRC recovery was observed within 20–30 sec, and CBP recovery within 20 sec. These results argued against the belief that steroid receptor complexes are statically bound to regulatory sites in hormone-responsive promoters, and suggested that receptors and their cofactors must interact with the response element repeatedly during transcriptional activation. The short duration of receptor–DNA interaction has been further supported by chromatin immunoprecipitation studies [98].
IV. SUMMARY AND FUTURE DIRECTIONS A significant advance during the past 3 years is the understanding that VDR, RXR, and most other nuclear receptors and co-regulators constantly shuttle in and out of the nucleus and remain in rapid motion within the nucleus. During this movement, VDR forms complexes with multiple protein machines in a sequential manner. Earlier, we learned which proteins can bind VDR, and now is the time to learn where and when these protein interactions take place. The ability to monitor receptor movement in living cells by fluorescent techniques in real time gives a new insight into the “rules of the game,” the hormonal effects. With multispectral imaging, several “players” can be monitored simultaneously, and their interaction can be detected using FRET. Expression of fluorescent protein chimeras of VDR and RXR is already permitting mechanistic understanding of nuclear import and export. Multiple nuclear localization signals have been identified in both receptors and in their coactivators and co-repressors. A better understanding of the export pathways is expected shortly.
JULIA BARSONY
Available data strongly suggest that transcriptional regulation is a highly dynamic, sequential process. The technologies that we developed have already been applied to studies on the trafficking of steroid receptors and other RXR partners, such as TR. Ongoing research is likely to elucidate the regulatory mechanisms that influence VDR and RXR trafficking in the near future. These mechanisms will generate a broad interaction “field” with multiple signaling processes converging to modulate calcitriol actions. In addition, our ability to monitor receptor–DNA interactions at a time resolution of seconds provides us with an opportunity to explore the dynamic assembly of transcription complexes at the regulatory sites.
References 1. MacDonald PN, Baudino TA, Tokumaru H, Dowd DR, Zhang C 2001 Vitamin D receptor and nuclear receptor coactivators: crucial interactions in vitamin D–mediated transcription. Steroids 66:171–176. 2. Polly P, Herdick M, Moehren U et al. 2000 VDR-Alien: A novel, DNA-selective vitamin D3 receptor–corepressor partnership. FASEB J 14:1455–1463. 3. Zhang C, Baudino TA, Dowd DR et al. 2001 Ternary complexes and cooperative interplay between NCoA-62/Ski-interacting protein and steroid receptor coactivators in vitamin D receptormediated transcription. J Biol Chem 276:40614–40620. 4. Lutz W, Kohno K, Kumar R 2001 The role of heat shock protein 70 in vitamin D receptor function. Biochem Biophys Res Commun 282:1211–1219. 5. Barletta F, Freedman LP, Christakos S 2002 Enhancement of VDR-mediated transcription by phosphorylation: Correlation with increased interaction between the VDR and DRIP205, a subunit of the VDR-interacting protein coactivator complex. Mol Endocrinol 16:301–314. 6. Jurutka PW, MacDonald PN, Nakajima S et al. 2002 Isolation of baculovirus-expressed human vitamin D receptor: DNA responsive element interactions and phosphorylation of the purified receptor. J Cell Biochem 85:435–457. 7. Wheeler DG, Horsford J, Michalak M, White JH, Hendy GN 1995 Calreticulin inhibits vitamin D3 signal transduction. Nucleic Acids Res 23:3268–3274. 8. Holaska JM, Black BE, Love DC et al. 2001 Calreticulin is a receptor for nuclear export. J Cell Biol 152:127–140. 9. Putkey JA, Wecksler WR, Norman AW 1978 The interaction of 1,25-dihydroxyvitamin D3 with its intestinal mucosa receptor: kinetic parameters and structural requirements. Lipids 13: 723–729. 10. Bloor JH, Dasmahapatra A, Weiser MM, Klohs WD 1982 Kinetics of subcellular distribution in rat intestine of 1,25dihydroxycholecalciferol administered in vivo. Evidence for concentration within 5 min into purified nuclei. Biochem J 208:567–575. 11. Jensen EV, Suzuki T, Kawashima T et al. 1968 A two-step mechanism for the interaction of estradiol with rat uterus. Proc Natl Acad Sci USA 59:632–638. 12. Gorski J, Welshons WV, Sakai D et al. 1986 Evolution of a model of estrogen action. Recent Prog Horm Res 42:297–329.
CHAPTER 22 VDR and RXR Subcellular Trafficking
13. Wikstrom AC, Bakke O, Okret S, Bronnegard M, Gustafsson JA 1987 Intracellular localization of the glucocorticoid receptor: Evidence for cytoplasmic and nuclear localization. Endocrinology 120:1232–1242. 14. Clemens TL, Garrett KP, Zhou XY et al. 1988 Immunocytochemical localization of the 1,25-dihydroxy vitamin D3 receptor in target cells. Endocrinology 122: 1224–1230. 15. Barsony J, Pike JW, DeLuca HF, Marx SJ 1990 Immunocytology with microwave-fixed fibroblasts shows 1α,25-dihydroxyvitamin D3–dependent rapid and estrogen-dependent slow reorganization of vitamin D receptors. J Cell Biol 111:2385–2395. 16. Barsony J, Renyi I, McKoy W 1997 Subcellular distribution of normal and mutant vitamin D receptors in living cells. Studies with a novel fluorescent ligand. J Biol Chem 272:5774–5782. 17. Racz A, Barsony J 1999 Hormone-dependent translocation of vitamin D receptors is linked to transactivation. J Biol Chem 274:19352–19360. 18. Michigami T, Suga A, Yamazaki M et al. 1999 Identification of amino acid sequence in the hinge region of human vitamin D receptor that transfers a cytosolic protein to the nucleus. J Biol Chem 274:33531–33538. 19. Sunn KL, Cock TA, Crofts LA, Eisman JA, Gardiner EM 2001 Novel N-terminal variant of human VDR. Mol Endocrinol 15:1599–1609. 20. Prufer K, Barsony J 2002 Retinoid X receptor dominates the nuclear import and export of the unliganded vitamin D receptor. Mol Endocrinol 16:1738–1751. 21. Prufer K, Barsony J 2002 Retinoid x receptor dominates the nuclear import and export of the unliganded vitamin d receptor. Mol Endocrinol 16:1738–1751. 22. Barsony J, Prufer K 2002 Vitamin D receptor and retinoid X receptor interactions in motion. Vitam Horm 65:345–376. 23. Kato S, Takeyama K, Kitanaka S et al. 1999 In vivo function of VDR in gene expression-VDR knock-out mice. J Steroid Biochem Mol Biol 69:247–251. 24. Mathieu C, Van Etten E, Gysemans C et al. 2001 In vitro and in vivo analysis of the immune system of vitamin D receptor knockout mice. J Bone Miner Res 16:2057–2065. 25. Chen CH, Sakai Y, Demay MB 2001 Targeting expression of the human vitamin D receptor to the keratinocytes of vitamin D receptor null mice prevents alopecia. Endocrinology 142: 5386–5389. 26. Li M, Indra AK, Warot X et al. 2000 Skin abnormalities generated by temporally controlled RXRalpha mutations in mouse epidermis. Nature 407:633–636. 27. Prufer K, Racz A, Lin GC, Barsony J 2000 Dimerization with retinoid X receptors promotes nuclear localization and subnuclear targeting of vitamin D receptors. J Biol Chem 275: 41114–41123. 28. Baumann CT, Maruvada P, Hager GL, Yen PM 2001 Nuclear cytoplasmic shuttling by thyroid hormone receptors. Multiple protein interactions are required for nuclear retention. J Biol Chem 276:11237–11245. 29. Mattaj IW, Englmeier L 1998 Nucleocytoplasmic transport: the soluble phase. Annu Rev Biochem 67:265–306. 30. Gorlich D, Kutay U 1999 Transport between the cell nucleus and the cytoplasm. Annu Rev Cell Dev Biol 15:607–660. 31. Weis K 2003 Regulating access to the genome: Nucleocytoplasmic transport throughout the cell cycle. Cell 112:441–451. 32. Gorlich D, Seewald MJ, Ribbeck K 2003 Characterization of Ran-driven cargo transport and the RanGTPase system by kinetic measurements and computer simulation. EMBO J 22:1088–1100.
377 33. Chook YM, Blobel G 2001 Karyopherins and nuclear import. Curr Opin Struct Biol 11:703–715. 34. Kuersten S, Ohno M, Mattaj IW 2001 Nucleocytoplasmic transport: Ran, beta and beyond. Trends Cell Biol 11:497–503. 35. Jakel S, Mingot JM, Schwarzmaier P, Hartmann E, Gorlich D 2002 Importins fulfil a dual function as nuclear import receptors and cytoplasmic chaperones for exposed basic domains. EMBO J 21:377–386. 36. Macara IG 2001 Transport into and out of the nucleus. Microbiol Mol Biol Rev 65:570–94, table. 37. Smith A, Brownawell A, Macara IG 1998 Nuclear import of Ran is mediated by the transport factor NTF2. Curr Biol 8:1403–1406. 38. Kalab P, Weis K, Heald R 2002 Visualization of a Ran-GTP gradient in interphase and mitotic Xenopus egg extracts. Science 295:2452–2456. 39. Ribbeck K, Gorlich D 2002 The permeability barrier of nuclear pore complexes appears to operate via hydrophobic exclusion. EMBO J 21:2664–2671. 40. Stade K, Ford CS, Guthrie C, Weis K 1997 Exportin 1 (Crm1p) is an essential nuclear export factor. Cell 90:1041–1050. 41. Fornerod M, Ohno M. 2002 Exportin-mediated nuclear export of proteins and ribonucleoproteins. Results Probl Cell Differ 35:67–91. 42. Ossareh-Nazari B, Bachelerie F, Dargemont C 1997 Evidence for a role of CRM1 in signal-mediated nuclear protein export. Science 278:141–144. 43. Trotta CR, Lund E, Kahan L, Johnson AW, Dahlberg JE 2003 Coordinated nuclear export of 60S ribosomal subunits and NMD3 in vertebrates. EMBO J 22:2841–2851. 44. Clouse KN, Luo MJ, Zhou Z, Reed R 2001 A Ran-independent pathway for export of spliced mRNA. Nat Cell Biol 3:97–99. 45. Yi R, Bogerd HP, Wiegand HL, Cullen BR 2002 Both ran and importins have the ability to function as nuclear mRNA export factors. RNA 8:180–187. 46. Holaska JM, Black BE, Rastinejad F, Paschal BM 2002 Ca2+-dependent nuclear export mediated by calreticulin. Mol Cell Biol 22:6286–6297. 47. Segal SP, Graves LE, Verheyden J, Goodwin EB 2001 RNARegulated TRA-1 nuclear export controls sexual fate. Dev Cell 1:539–551. 48. Bunn CF, Neidig JA, Freidinger KE et al. 2001 Nucleocytoplasmic Shuttling of the Thyroid Hormone Receptor alpha. Mol Endocrinol 15:512–533. 49. Zhou ZX, Sar M, Simental JA, Lane MV, Wilson EM 1994 A ligand-dependent bipartite nuclear targeting signal in the human androgen receptor. Requirement for the DNA-binding domain and modulation by NH2-terminal and carboxyl-terminal sequences. J Biol Chem 269:13115–13123. 50. Picard D, Yamamoto KR 1987 Two signals mediate hormonedependent nuclear localization of the glucocorticoid receptor. EMBO J 6:3333–3340. 51. Poukka H, Karvonen U, Yoshikawa N et al. 2003 The RING finger protein SNURF modulates nuclear trafficking of the androgen receptor. J Cell Sci 113(Pt 17):2991–3001. 52. Palmeri D, Malim MH 1999 Importin beta can mediate the nuclear import of an arginine-rich nuclear localization signal in the absence of importin alpha. Mol Cell Biol 19:1218–1225. 53. Guiochon-Mantel A, Lescop P, Christin-Maitre S et al. 1991 Nucleocytoplasmic shuttling of the progesterone receptor. EMBO J 10:3851–3859. 54. Maruvada P, Baumann CT, Hager GL, Yen PM 2003 Dynamic shuttling and intranuclear mobility of nuclear hormone receptors. J Biol Chem 278:12425–12432.
378 55. Liu J, DeFranco DB 2000 Protracted nuclear export of glucocorticoid receptor limits its turnover and does not require the exportin 1/CRM1-directed nuclear export pathway. Mol Endocrinol 14:40–51. 56. Saporita AJ, Zhang Q, Navai N et al. 2003 Identification and characterization of a ligand-regulated nuclear export signal in androgen receptor. J Biol Chem. 57. Katagiri Y, Takeda K, Yu ZX et al. 2001 Modulation of retinoid signalling through NGF-induced nuclear export of NGFI-B. Nat Cell Biol 2:435–440. 58. Zilliacus J, Holter E, Wakui H et al. 2001 Regulation of glucocorticoid receptor activity by 14-3-3-dependent intracellular relocalization of the corepressor rip140. Mol Endocrinol 15:501–511. 59. Masuyama H, Jefcoat SCJ, MacDonald PN 1997 The N-terminal domain of transcription factor IIB is required for direct interaction with the vitamin D receptor and participates in vitamin D–mediated transcription. Mol Endocrinol 11:218–228. 60. Wiebel FF, Steffensen KR, Treuter E, Feltkamp D, Gustafsson JA 1999 Ligand-independent coregulator recruitment by the triply activatable OR1/retinoid X receptor–alpha nuclear receptor heterodimer [In Process Citation]. Mol Endocrinol 13:1105–1118. 61. Stenoien DL, Mancini MG, Patel K et al. 2000 Subnuclear trafficking of estrogen receptor-alpha and steroid receptor coactivator-1. Mol Endocrinol 14:518–534. 62. Becker M, Baumann C, John S et al. 2002 Dynamic behavior of transcription factors on a natural promoter in living cells. EMBO Rep 3:1188–1194. 63. Xiao Z, Brownawell AM, Macara IG, Lodish HF 2003 A novel nuclear export signal in Smad1 is essential for its signaling activity. J Biol Chem 278:34245–34252. 64. Black BE, Holaska JM, Rastinejad F, Paschal BM 2001 DNA binding domains in diverse nuclear receptors function as nuclear export signals. Curr Biol 11:1749–1758. 65. Hsieh JC, Shimizu Y, Minoshima S et al. 1998 Novel nuclear localization signal between the two DNA-binding zinc fingers in the human vitamin D receptor. J Cell Biochem 70: 94–109. 66. Prufer K, Racz A, Lin GC, Barsony J 2000 Dimerization with retinoid X receptors promotes nuclear localization and subnuclear targeting of vitamin D receptors. J Biol Chem 275:41114–41123. 67. Yu Z, Lee CH, Chinpaisal C, Wei LN 1998 A constitutive nuclear localization signal from the second zinc-finger of orphan nuclear receptor TR2. J Endocrinol 159:53–60. 68. Luo Z, Rouvinen J, Maenpaa PH 1994 A peptide C-terminal to the second Zn finger of human vitamin D receptor is able to specify nuclear localization. Eur J Biochem 223:381–387. 69. Swamy N, Mohr SC, Xu W, Ray R 1999 Vitamin D receptor interacts with DnaK/heat shock protein 70: Identification of DnaK interaction site on vitamin D receptor. Arch Biochem Biophys 363:219–226. 70. Xiao Z, Brownawell AM, Macara IG, Lodish HF 2003 A novel nuclear export signal in Smad1 is essential for its signaling activity. J Biol Chem 278:34245–34252. 71. Birbach A, Gold P, Binder BR et al. 2002 Signaling molecules of the NF-kappa B pathway shuttle constitutively between cytoplasm and nucleus. J Biol Chem 277:10842–10851. 72. Aplin AE, Hogan BP, Tomeu J, Juliano RL 2002 Cell adhesion differentially regulates the nucleocytoplasmic distribution of active MAP kinases. J Cell Sci 115:2781–2790.
JULIA BARSONY
73. Meyer T, Marg A, Lemke P, Wiesner B, Vinkemeier U 2003 DNA binding controls inactivation and nuclear accumulation of the transcription factor Stat1. Genes Dev 17:1992–2005. 74 O’Keefe K, Li H, Zhang Y 2003 Nucleocytoplasmic shuttling of p53 is essential for MDM2-mediated cytoplasmic degradation but not ubiquitination. Mol Cell Biol 23:6396–6405. 75. Wadhwa R, Yaguchi T, Hasan MK et al. 2002 Hsp70 family member, mot-2/mthsp70/GRP75, binds to the cytoplasmic sequestration domain of the p53 protein. Exp Cell Res 274:246–253. 76. Prufer K, Racz A, Lin GC, Barsony J 2000 Dimerization with retinoid X receptors promotes nuclear localization and subnuclear targeting of vitamin D receptors. J Biol Chem 275:41114–41123. 77. Antonsson A, Hughes K, Edin S, Grundstrom T 2003 Regulation of c-Rel nuclear localization by binding of Ca2+/calmodulin. Mol Cell Biol 23:1418–1427. 78. Trevillyan JM, Chiou XG, Chen YW et al. 2001 Potent inhibition of NFAT activation and T cell cytokine production by novel low molecular weight pyrazole compounds. J Biol Chem 276:48118–48126. 79. Barsony J, Carroll J, McKoy W, Renyi I, Gould DL, Htun H et al. 1997 Intracellular Traffic of Glucocorticoid Receptors: Studies with Green Fluorescent Protein Chimeras in Living Cells. Springer-Verlag, New York. 80. Pratt WB, Dalman FC, Meshinchi S, Scherrer LC 1990 The relationship between glucocorticoid receptor binding to Hsp90 and receptor function. Nippon Naibunpi Gakkai Zasshi 66: 1185–1197. 81. Barsony J, McKoy W 1992 Molybdate increases intracellular 3′,5′-guanosine cyclic monophosphate and stabilizes vitamin D receptor association with tubulin-containing filaments. J Biol Chem 267:24457–24465. 82. Lian JB, Stein JL, Stein GS et al. 2001 Contributions of nuclear architecture and chromatin to vitamin D–dependent transcriptional control of the rat osteocalcin gene. Steroids 66:159–170. 83. Zhang C, Dowd DR, Staal A et al. 2003 Nuclear coactivator-62 kDa/Ski-interacting protein is a nuclear matrix– associated coactivator that may couple vitamin D receptor–mediated transcription and RNA splicing. J Biol Chem 278:35325–35336. 84. Kino T, Souvatzoglou E, De Martino MU et al. 2003 Protein 14-3-3sigma interacts with and favors cytoplasmic subcellular localization of the glucocorticoid receptor, acting as a negative regulator of the glucocorticoid signaling pathway. J Biol Chem 278:25651–25656. 85. Walther RF, Lamprecht C, Ridsdale A et al. 2003 Nuclear export of the glucocorticoid receptor is accelerated by cell fusion-dependent release of calreticulin. J Biol Chem 278: 37858–37864. 86. Prufer K, Schroder C, Racz A, Barsony J 2000 Cell Cycle Dependence of Vitamin D Receptor Expression. Vitamin D Workshop, Riverside, CA. 87. Amazit L, Alj Y, Tyagi RK et al. 2003 Subcellular localization and mechanisms of nucleocytoplasmic trafficking of steroid receptor coactivator-1. J Biol Chem 278: 32195–32203. 88. Hong SH, Privalsky ML 2000 The SMRT corepressor is regulated by a MEK-1 kinase pathway: inhibition of corepressor function is associated with SMRT phosphorylation and nuclear export. Mol Cell Biol 20:6612–6625.
CHAPTER 22 VDR and RXR Subcellular Trafficking
89. Wang C, Fu M, Angeletti RH et al. 2001 Direct acetylation of the estrogen receptor alpha hinge region by p300 regulates transactivation and hormone sensitivity. J Biol Chem 276:18375–18383. 90. Fu M, Wang C, Zhang X, Pestell R 2003 Nuclear receptor modifications and endocrine cell proliferation. J Steroid Biochem Mol Biol 85:133–138. 91. Nicoll JB, Gwinn BL, Iwig JS et al. 2003 Compartmentspecific phosphorylation of rat thyroid hormone receptor alpha1 regulates nuclear localization and retention. Mol Cell Endocrinol 205:65–77. 92. Hager GL, Smith CL, Fragoso G et al. 1998 Intranuclear trafficking and gene targeting by members of the steroid/nuclear receptor superfamily. J Steroid Biochem Mol Biol 65:125–132. 93. Stenoien DL, Patel K, Mancini MG et al. 2001 FRAP reveals that mobility of oestrogen receptor-alpha is ligand- and proteasome-dependent. Nat Cell Biol 3:15–23.
379 94. Akiyama TE, Baumann CT, Sakai S, Hager GL, Gonzalez FJ 2002 Selective intranuclear redistribution of PPAR isoforms by RXR alpha. Mol Endocrinol 16:707–721. 95. McNally JG, Muller WG, Walker D, Wolford R, Hager GL 2000 The glucocorticoid receptor: rapid exchange with regulatory sites in living cells. Science 287:1262–1265. 96. Maruvada P, Baumann CT, Hager GL, Yen PM 2003 Dynamic shuttling and intranuclear mobility of nuclear hormone receptors. J Biol Chem 278:12425–12432. 97. Barsony J, Prufer K 2002 Vitamin D receptor and retinoid X receptor interactions in motion. Vitam Horm Adv Res Appl 65:345–376. 98. Yamamoto H, Shevde NK, Warrier A et al. 2003 2-Methylene19-nor-(20S)-1,25-dihydroxyvitamin D3 potently stimulates gene-specific DNA binding of the vitamin D receptor in osteoblasts. J Biol Chem 278:31756–31765.
CHAPTER 23
1α,25(OH)2-Vitamin D3–Mediated Rapid and Genomic Responses Are Dependent upon Critical Structure–Function Relationships for Both the Ligand and Receptor(s) ANTHONY W. NORMAN
Department of Biochemistry and Division of Biomedical Sciences, University of California, Riverside, California
I. Introduction to Vitamin D3 and 1α,25(OH)2D3 II. Vitamin D Endocrine System III. Proteins with Ligand Binding Domains for 1α,25(OH)2D3
IV. Structure–Function Evaluation of Selected Rapid Responses Mediated by 1α,25(OH)2D3 V. Summary VI. Addendum References
I. INTRODUCTION TO VITAMIN D3 AND 1α,25(OH)2D3
certainty whether any or all of them occur via signal transduction pathways that do not involve immediate interaction of 1α,25(OH)2D3 with its nuclear receptor.
The secosteroid 1α,25(OH)2-vitamin D3 [1α,25(OH)2D3] initiates biological responses via both regulation of gene transcription as well as via rapid response pathways some of which include opening of voltage-gated Ca2+ and Cl− channels and involvement with protein kinase C (PKC) [1]. The general objective of this chapter is to provide an introduction to rapid or nongenomic responses mediated by 1α,25(OH)2D3. The principal focus will be to compare structure– function relationships for ligands for rapid responses, genomic responses mediated by the nuclear vitamin D receptor (VDR) with the plasma vitamin D binding protein (DBP). There is a problem of definition: what is implied by the term “rapid responses”? For the purposes of this chapter, “rapid responses” is used to describe biological responses to 1α,25(OH)2D3 that appear to occur too rapidly to be simply explained via interaction of the ligand with the vitamin D receptor (VDR) present in the cell nucleus. Frequently these responses are generated within 1–2 min to 15–45 min to, in some cases, 1–4 hr. However, since there is not yet available a detailed molecular mechanism(s) to describe all the observed rapid responses to 1α,25(OH)2D3, it is not known with VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
A. Structure of 1α,25(OH)2D3 1α,25(OH)2D3 is one of 37 known metabolites of vitamin D3 [2,3]. The structure of 1α,25(OH)2D3 is presented in Fig. 1A. The structure and chemistry of vitamin D compounds are related to those of the classical steroid hormones, i.e., the glucocorticoids, estrogens, androgens, progestins, and mineralocorticoids. However, there are three structural aspects of the vitamin D compounds that make them distinct from classic steroids: (a) The vitamin D3 compounds have a side chain comprising eight carbons, whereas the classical steroids have no side chain or only a two-carbon side chain; (b) the 9,10 carbon–carbon bond of the β-ring is broken in this family of related compounds and, accordingly, they are designated as being secosteroids; the term “seco” designates that one of the rings of the cyclopentanoperhydrophenanthrene ring structure of classical steroids has been broken; and (c) the vitamin D3 family of molecules is unusually conformationally flexible in contradistinction to the classic steroids (see later discussion). Copyright © 2005, Elsevier, Inc. All rights reserved.
382
ANTHONY W. NORMAN
A
22
21
24
18 17 12 11 13 14 9
5
4
16 15
8H 7
6
Side chain
25 OH
Rotation, around B-ring 6,7 bond
19 10
3 HO
B
A-ring chair–chair interconversion
OH
2
TOP VIEW (as shown in structure)
C
IN-PLANE VIEW
Side-chain Side-chain H H
HO
HO
H
H H
HO α-chair
D
21 18 12 11 13 14 9
6 4 3
510
8H 7
H
A-ring conformations (chair–chair inversion)
OH β-chair
22
24 26 20 23 25 OH 17 27 16
OH HO
15
19
H
A
19 HO
D
C 9
Fast
7 6
HO
OH 2 6-s-trans confirmation
6-s-cis confirmation
FIGURE 1 Structural aspects of 1α,25(OH)2D3 which contribute to its conformational flexibility. (A) Structure of 1α,25(OH)2D3 and an indication of the location of the three structural aspects of vitamin D secosteroids that contribute to the conformational flexibility of these molecules. (B) The dynamic single bond rotation of the cholesterol-like side chain of the hormone shown on the center panel (i.e., rotations about the six single bonds are indicated by the curved arrows). Energy minimization calculations to define the most stable orientation of the side chain indicate that there are 394 conformers within 4 kcal/mol of the global minimum (lowest energy state). The dots indicate the position in threedimensional space of the 25-hydroxyl group for the 394 minima; collectively these dots define the volume in space that the side chain occupies. The left panel illustrates a “top view” looking down on the C/D rings while the right panel presents an edge-on view of the C/D rings. A discussion of the consequences of the side-chain conformational mobility has been previously presented [4,5,116]. (C) The cyclohexane-like A-ring is free to interchange rapidly between a pair of chair-chair conformers effectively equilibrating the key 1α- and 3β-hydroxyls between equatorial and axial orientations [6,7]. Thus when the 1α-hydroxyl is axial, the 3β-hydroxyl is equatorial, and vice versa. (D) Rotational freedom about the 6–7 carbon–carbon single bond of the seco B-ring allows conformations ranging from the more steroid-like 6-s-cis conformation to the open and extended 6-s-trans form of the hormone [8].
CHAPTER 23 Function Relationships for Both the Ligand and Receptor(s)
B. Conformational Flexibility
II. VITAMIN D ENDOCRINE SYSTEM
The intact eight-carbon side-chain of vitamin D3 and related secosteroids can easily assume numerous shapes and positions by virtue of rotation about its many carbon–carbon single bonds; see Fig. 1B. The dots indicate the position in three-dimensional space of the 25-hydroxyl group for 394 readily identifiable side-chain conformations. A discussion of the consequences of the side-chain conformational mobility has been previously presented [4,5]. Breakage of the 9,10 carbon–carbon bond releases the A-ring from being held in place to the B-ring and, accordingly, the cyclohexane-like A-ring is free to rapidly interchange (many thousands of times per second) between a pair of chair–chair conformers (Fig. 1C). This has the consequence of changing the orientation of the key 1α- and 3β-hydroxyls between either the equatorial or axial orientation [6,7]. Rotational freedom about the 6–7 carbon–carbon bond of the seco B-ring allows a continuum of conformations ranging from the more steroid-like 6-s-cis conformation to the open and extended 6-s-trans form of the hormone [8]; these conformations are illustrated in Fig. 1D and also Fig. 11. It has been generally assumed for receptor–ligand interactions that the ligand is frozen in a single conformation dictated by both the structural constraints of the ligand and the three-dimensional architecture of the peptide chains that create the ligand binding domain (LBD) of the receptor(s). The ligands for the RAR and TR, as for the VDR, are all conformationally flexible, and the X-ray crystallographic structure for each receptor indicated that only one definitive conformer was present in their ligand binding domain [9,10]. This clearly demonstrates that the capability of steroid receptors to capture one ligand conformation from a large population of available flexible conformers is a general feature of several proteins. Certainly the most important polar functional group of 1α,25(OH)2D3 is the 1α-hydroxyl; it is this functional group that makes the molecule distinct from 25(OH)D3, which is the major circulating form of vitamin D3 present in the blood. 1α,25(OH)2D3 binds some 666-fold more tightly to the nuclear receptor than 25(OH)D3 [1]. Also, the orientation of the hydroxyl on carbon-1 is important to the nuclear receptor; thus the nuclear receptor binds 1α,25(OH)2D3 3000-fold better than 1β,25(OH)2D3 [1]. Thus there is no question that receptors for 1α,25(OH)2D3 must be able to distinguish both the presence or absence of this hydroxyl and its orientation (1α versus 1β) as well as to capture it to form a stable receptor–ligand complex.
A. General Description
383
The concept of the existence of the vitamin D endocrine system is now firmly established (see Fig. 2) [1,12,13]. This endocrine system includes four key components. First, the kidney functions as an endocrine gland to produce in a physiologically regulated manner the two secosteroid hormones 1α,25(OH)2D3 and 24R,25-dihydroxyvitamin D3 [24R,25(OH)2D3]. Second, a feedback loop regulates the secretion of parathyroid hormone (PTH) in response to the serum ionized Ca2+ level. A prime function of PTH is to interact in the proximal tubule of the kidney to increase or (in the face of falling PTH levels) decrease the activity of the 25(OH)D3-1α-hydroxylase, which, in turn, determines the output of 1α,25(OH)2D3 available to generate biological responses. A third key component in the operation of the vitamin D endocrine system is the plasma vitamin D binding protein (DBP), which carries vitamin D3 and all its metabolites to their various target organs. The DBP is known to have a specific ligand binding domain for vitamin D-related ligands that is different in specificity from the ligand binding domain of the nuclear vitamin D receptor [11]. Fourth and most importantly, target organs for 1α,25(OH)2D3 possess receptors for the hormones that, when occupied with ligand, initiate the signal transduction pathways leading to the appearance of biological responses. Many of these topics are reviewed in detail in other chapters of this book. Understanding how 1α,25(OH)2D3 can produce biological responses via regulation of gene transcription [14–16] as well as via rapid membrane-initiated pathways [17–19], some of which involve opening of voltage-gated Ca2+ channels [20] or protein kinase C (PKC) [21] (see Fig. 3), is one of the areas of current intensive research investigation. The regulation of gene transcription by 1α,25(OH)2D3 is known to be mediated by interaction of this ligand with a nuclear receptor protein, termed the VDRnuc, while the membrane-initiated responses are postulated to be mediated through interaction of the 1α,25(OH)2D3 with a protein receptor located on the external surface of the cell; this receptor is referred to as the VDRmem. There is also emerging evidence that the plasma membrane binding protein [VDRmem] may be associated with caveolae. Caveolae are flask-shaped invaginations on the plasma membrane that serve as platforms for accumulation of signal transduction proteins; the following citations will provide an entrée into this large literature [22–24].
384
ANTHONY W. NORMAN
VITAMIN D3
LIVER
Kidney as an endocrine gland 25(OH)2D3
25(OH) D3
24-HYDROXYLASE (+) 24R,25(OH)2D3 PLACENTAL PRODUCTION 1α,25(OH)2D3
(+)
(−) 1-HYDROXYLASE (+) 1α,25(OH)2D3 H
PARACRINE PRODUCTION OF 1α,25(OH)2D3
H
HO
OH H
OH
OH
H
OH
H
24R,25(OH)2D3
PARATHYROID HORMONE
(−) (−) SHORT LONG FEEDBACK FEEDBACK LOOP LOOP
HO
1α,25(OH)2D3
24R,25(OH)2D3
BLOOD BLOOD 1α,25(OH)2D3
FETUS Development
CELLULAR GROWTH & DEFFERENTION
RAPID ACTIONS
1α,25(OH)2D3 NUCLEAR RECEPTORS
BLOOD 24R,25(OH)2D3
CLASSIC TARGET ORGANS 1α,25(OH)2D3 or ANALOGS
Pi Ca2+
24R,25(OH)2D3 RECEPTORS
BONE INTESTINE KIDNEY
REABSORPTION OF Ca2+ & Pi ABSORPTION OF Ca2+ SELECTED BIOLOGICAL RESPONSES
MOBILIZATION / ACCRETION OF Ca2+ & Pi
FIGURE 2 Summary of the vitamin D endocrine system. The parent vitamin D3 is first converted in the liver to 25(OH)D3 and then in the kidney to one of two hormones, 1α,25(OH)2D3 or 24R,25(OH)2D3. 1α,25(OH)2D3 produces biological responses in its target organs via either interacting with a nuclear receptor (VDRnuc) or with a system which generates rapid responses and is believed to involve a membrane receptor (VDRmem) [1]. The precise biological role(s) of 24R,25(OH)2D3 are not yet defined; however, several studies suggest the presence of receptors in chondrocytes [59,117,118] and bone [61,119].
B. Tissues with the Nuclear VDR Table I summarizes the target organs which are known to possess the VDRnuc. The nuclear responses to 1α,25(OH)2D3 are generated in a manner homologous to that of classical steroid hormones, i.e., cortisol, progesterone, estradiol, testosterone, and aldosterone. In the general model, the hormone is produced in an endocrine gland in response to a physiological stimulus, then circulates in the blood bound to a protein carrier (the vitamin D binding protein or DBP), which delivers it to target tissues where the hormone enters the cell and interacts with a specific, high-affinity intracellular receptor(s). The receptor–hormone complex then localizes in the nucleus, undergoes some type of activation perhaps involving phosphorylation [25–29], and binds to a hormone response element (HRE) on the DNA to modulate the expression of hormone-sensitive genes. The modulation of gene transcription results in either the induction or the repression of specific mRNAs, ultimately resulting in changes in protein
expression needed to produce the required biological response [15,30,31]. High-affinity receptors for 1α,25(OH)2D3 have been identified in at least 35 target tissues [14,32,33], and more than 100 genes are known to be regulated by 1α,25(OH)2D3 [15]. The mechanism of the regulation of these genes may be direct, as described earlier, through HREs, or indirect. See the following reviews for further information [16,34,35]. Since the discovery of 1α,25(OH)2D3 in 1968 [36] and its VDRnuc in 1969 [37], research has focused on defining how these agents collaborate to generate biological responses; of course the principal focus was on understanding regulation of gene transcription. Table II summarizes the new biological actions of 1α,25(OH)2D3 that are known to be mediated by the VDRnuc. In many instances, the first hint of the biological process involved was derived from discovery and then biochemical description and characterization of the VDR in a new location. For example (see Table II), the observation that the VDR was present in the pancreas [38,39] and more particularly the β cells [40]
385
CHAPTER 23 Function Relationships for Both the Ligand and Receptor(s)
Natural steroid hormone
Conformationally flexible analogs
Conformationally restricted analogs
TABLE I Tissue Distribution of the Nuclear Receptor (VDR) for 1α,25(OH)2D3 Target cell
Nucleus
Receptors
Plasma membrane
Ligand-receptor complex
MAP kinase activation
DNA
cAMP PKC Ca2+ channel CI−channel (open)
“Cross talk”
Gene expression Up regulation
mRNA
Down regulation
mRNA
(minutes-hours-days)
(seconds-minutes)
BIOLOGICAL RESPONSES
FIGURE 3 Pathways for generation of biological responses by 1α,25(OH)2D3. In the genomic pathway, occupancy of the nuclear receptor for 1α,25(OH)2D3 (VDRnuc) by a ligand leads to an up- or down-regulation of genes subject to hormone regulation. In the membrane-initiated pathway, occupancy of a membrane receptor for 1α,25(OH)2D3 by a ligand can lead to activation of protein kinase C (PKC), adenylate cyclase, phospholipase C, and/or opening of voltage-gated Ca2+ or chloride (Cl−) channel openings, which are coupled to generation of the end biological response(s). Also, this figure emphasizes that rapid responses can engage in cross-talk to modulate genomic responses.
suggested that it was responsible for the production of calbindin. Another discovery was by Tsoukas et al. [41], who reported that B and T lymphocytes, when activated, both expressed a VDRnuc. This observation is the foundation of a major new field of research concerning the role of vitamin D on the immune system [42,43].
C. Rapid Responses Starting in approximately 1983, a series of observations were reported which suggested that some of the biological responses generated by 1α,25(OH)2D3 occurred too rapidly to be simply explained via genomic events.
Adipose Bone Brain Cancer cells (many) Colon Ganglion Hair follicle Kidney Lung Muscle (smooth) Osteoclast Pancreas β cell Parotid Placenta Skin Testis Thyroid Yolk sac
Reference a
Target cell
Reference a
[123] [125,126] [128–130] [131,133–138]
Adrenal Bone marrow Breast Cartilage
[124] [127] [131,132] [139–141]
[133,142] [145] [147] [38] [153,154] [108,157,158]
Eggshell gland Epididymis Intestine Liver Muscle (cardiac) Osteoblast
[143,144] [146] [148,149] [150–152] [155,156] [159,160]
[161,162] [38,39]
Ovary Parathyroid
[163,164] [165–167]
[168] [169] [175–179] [180,181] [184] [185]
Pituitary Prostate Stomach Thymus Uterus
[169–171] [172–174] [170] [182,183] [134,146]
aIn the interest of economy, only selected reference citations are included. bAn alternative approach to the biochemical characterization of receptors in a proposed target organ for 1α,25(OH)2D3 is the technique of autoradiography using high specific activity [3H]1,25(OH)2D3; W. D. Stumpf has published many papers proposing the presence of specific receptors for 1α,25(OH)2D3 based on his autoradiographic studies [170,186].
Accordingly, there is now active research on how 1α,25(OH)2D3 generates biological responses via both genomic and rapid pathways; a model is presented in Fig. 3. Table III presents a tabulation of the rapid responses mediated by 1α,25(OH)2D3. In comparison to our understanding of the interaction of 1α,25(OH)2D3 with its nuclear VDR and the plethora of details concerning regulation of gene transcription, it is clear at the time of preparation of this chapter that the field of nongenomic responses is still in its descriptive phase. As a consequence of many of these studies, it was originally proposed, based on this laboratory’s studies on transcaltachia, that some of the actions of 1α,25(OH)2D3 may be mediated at the cell membrane or by extranuclear subcellular components
386 TABLE II
ANTHONY W. NORMAN
New Biological Actions of 1α,25(OH)2D3
Target cells or organ actions related to VDRnuc Adipose Bone marrow Cancer cells
Cells with a VDRmem Immune system Intestine Leukemia NB4 cells
Osteoblasts Pancreas Pancreas B cells Vascular smooth muscle
Biological process Adipocyte metabolism Monocyte differentiation to macrophages Antiproliferation, apoptosis, cell differentiation Rapid responses Immunosuppression Transcaltachia Activation of MAP-kinase; cell differentiation Opening of Ca2+ and Cl− channels Insulin secretion Insulin secretion Cell migration; activation of phosphatidyl inositol-3′-kinase
TABLE III Organ/cell/system Cartilage cells
Colon Cross-talk between VDR and phosphatase Endothelial cells: skeletal muscle cells, leukemia cells Fibroblasts Intestine
Keratinocytes Kidney Lipid bilayer Liver Muscle Osteoblast
Parathyroid cells Promyelocytic leukemic cells
References [187] [188,189] [190–193]
See Table III [41,194] [44,88] [94,195]
[20,96] [40,196] [120,197] [121]
near the plasma membrane [44–46]. A candidate membrane receptor (VDRmem) for 1α,25(OH)2D3 has been identified [47]; see later discussion. An alternative possibility is that the nuclear vitamin D receptor (VDRnuc) may, under some circumstances, be associated with plasma membrane acceptor sites [48] or with caveolae in the plasma membrane [49]. As a point of comparison between 1α,25(OH)2D3 and the other steroid hormones, it is well known that they, too (estradiol, progesterone, testosterone, aldosterone, and cortisol), can all generate a wide array of rapid responses; the following references provide an entrée into this large literature [50–52]. As with 1α,25(OH)2D3, it is not clear for the other steroid hormones whether they all utilize novel membrane receptors or nuclear receptors associated with plasma membrane caveolae that are linked to the generation of rapid responses (or both). There is strong experimental evidence for both a progesterone G–protein coupled receptor [53,54] and a nuclear estrogen receptor associated with caveolae in endothelial cells that is linked to rapid responses mediated by nitric oxide (NO) [55–58]. Both of these physiological settings could be a paradigm for 1α,25(OH)2D3-mediated rapid responses.
Distribution of Rapid Responses to 1α,25(OH)2D3 Response studied
Reference
Activation of PLCβ Regulation of matrix vesicle PKC Arachidonic acid is an autocoid mediator of growth plate cells PKC effects Activation of Ser/Thr phosphatase P79 S6 kinase via interaction with catalytic subunit Activation of PI3-kinase
[198] [199–201] [202] [203] [204]
Accumulation of cGMP near VDRnuc Rapid transport of intestinal Ca2+ (transcaltachia); CaCo-2 cells, PKC, G proteins Activation of PKC Alter PKC subcellular distribution PKC effects Activation of highly purified PKC Increased intracellular Ca2+ and activation of PKC and MAP kinase PKC and Ca2+ effects ROS 17/2.8 cells, Ca2+ channel opening Cl− channel opening, UMR-106 cells Ca2+ channel opening by 24R,25(OH)2D3 Phospholipid metabolism Cytosolic Ca2+ Aspects of cell differentiation PKC effects
[207] [44,208,209] [210] [211] [212] [213] [214] [215,216] [217] [20] [218] [203] [219] [220] [221,222] [21,223]
[121,205,206]
387
CHAPTER 23 Function Relationships for Both the Ligand and Receptor(s)
Thus, each possesses a unique ligand binding domain which specifically recognizes secosteroids. More specifically, the ligand with highest affinity to DBP is 25(OH)D3. The physiological responsibility of DBP is to transport vitamin D and all its daughter metabolites through the blood compartment delivering appropriate metabolites to appropriate target organs. In contrast, the VDR has a more specific ligand binding domain in that it exclusively focuses on binding 1α,25(OH)2D3 (666-fold more tightly than 25(OH)D3).
III. PROTEINS WITH LIGAND BINDING DOMAINS FOR 1α,25(OH)2D3 A. Introduction There are only a limited number of proteins that have ligand binding domains for vitamin D3 secosteroids; these include the plasma transport protein or vitamin D– binding protein (DBP), the various cytochrome P450 hydroxylases that mediate the myriad of metabolic transformations of vitamin D3 and its metabolites, and receptors for 1α,25(OH)2D3, including the classic nuclear receptor (VDRnuc) and membrane receptors (VDRmem). There is also a report from Seo et al. [59] describing a putative membrane receptor for 24R,25(OH)2D3 [60,61]. An ongoing project in the laboratory of the author is to determine the shape of the vitamin D3 metabolite ligand(s) for all these classes of proteins. Since these proteins are not believed to be structurally related, it is not anticipated that there necessarily will be structural homology of the various ligand binding domains. In this chapter we will first detail structural comparison of the VDRnuc with the DBP. Figure 4 presents a comparison of the domains of the VDRnuc with the domains of DBP. The details are provided in the figure legend. It is to be emphasized that there is not any amino acid sequence homology between these two proteins.
B. Vitamin D Receptor (VDRnuc) Based on evaluation of the human genome database, it is known that there are a total of 48 members of the nuclear receptor superfamily [62]. In this superfamily, 28 receptors have an LBD which binds a small ligand molecule; of these 28 receptors, 8 are classical hormone receptors and 20 are orphan receptors [62]. The VDR belongs to the subfamily of nuclear receptors with an LBD which binds a classical hormone; this includes as ligands the glucocorticoids (GR), progestins (PR), estrogens (ER), aldosterone (MR for mineral corticoid), androgens (AR), thyroid hormone (TR), the hormonal forms of vitamin A (RAR), and the VDR [63,64].
A VDR
HINGE
DNA BINDING
H2N
1
24
A/B
89 118 C
“LOOP”
165
1α,25(OH)2D3 BINDING
215
COOH
404 427
D
E
F
B DBP
25(OH)D3 BINDING
H2N 1
ACTIN BINDING DOMAIN
110 I
192
COOH
368 II
458 III
FIGURE 4 Schematic models of the VDR and DBP. (A) The VDR comprises 427 amino acid residues that are divided into six domains (A–F). The numbers below the VDR indicate the amino acid residue boundaries for the various domains. The VDR belongs to a superfamily of nuclear receptors that have the same general A–F domain organization. The C domain, the most highly conserved, which contains the DNA-binding domain defines the superfamily; it contains two zinc finger motifs. The E domain or ligand binding domain (LBD) is less conserved and is responsible for binding 1α,25(OH)2D3 or its analogs and transcriptional activation. The A/B domain of the VDR is much smaller than other members of the superfamily. The portion of the intact VDR that was crystallized and subjected to X-ray crystallographic analysis included residues 118–427, but with deletion of the loop region of the hinge domain D, specifically residues 165–215 (see text). (B) DBP consists of 458 amino acid residues and is divided into three domains (I, II, and III). The numbers below the DBP indicate the amino acid residue boundaries for the various domains. Domains I, II, and III have been postulated to have evolved from a progenitor that arose from the triple repeat of a 192 amino acid sequence [78]. However, domain III is significantly truncated at the C terminus. The 25(OH)D3 binding cleft is associated with the first six α-helices or residues 1–110 of domain I. The actin binding property of DBP is associated with a portion of domains I and III, which clamp the actin while it rests on domain II.
388 A dramatic advance in understanding of the threedimensional structure of the LBD of steroid receptors has occurred over the past 7 years with the X-ray crystallographic structure determination of the LBD of five steroid hormone–related receptors. These include the LBDs of the TR, RAR, ER, PR, GR, MR, PPARγ (peroxisome proliferator activated receptor), and the closely related TR and RAR; this work has been reviewed by Weatherman et al. [65]. They all have a common structural organization consisting of six domains [65,66] with significant amino acid sequence homologies across all the domains (see Fig. 4A). The different domains act as distinct functional modules that can function independently of each other [67]; see Fig. 3A. Also, an X-ray structure is available for the LBD of the unoccupied 9-cis-retinoic acid receptor RXR [68]. Further, the ER LBD X-ray structures are known for a ligand (raloxifene) that can act as an antagonist of the transcriptional activation function of the nuclear receptor. In 2000, the crystal structure of a modified version of the ligand binding domain of the nuclear receptor for vitamin D, bound to 1α,25(OH)2D3, was determined at a 1.8-Å resolution [69]. A follow-up X-ray crystallographic report compared the VDR LBD and bound ligand for 1α,25(OH)2D3 with that of four superagonist analogs of 1α,25(OH)2D3 [70]. The structure of the LBD of the human VDRnuc spans amino acid residues 143–427, [COOH terminus] but without residues 165–215, which were in an undefined loop in the hinge region of domain D (see Fig. 4A). In comparison to other nuclear receptors, the length of the VDR loop region is significantly longer. The removal of the flexible insertion domain in the VDR LBD produced a more soluble protein that was more amenable to crystallization. The VDR LBD protein structure is very similar to the LBD of the five other X-ray crystallographic determined nuclear hormone receptor structures [65]. The VDR LBD structure consists of a three-stranded β-sheet and 12 α-helices that are arranged to create a three-layer sandwich that completely encompasses the ligand 1α,25(OH)2D3 in a hydrophobic core (see Fig. 6A,C). Comparison of panels A, B, and C of Fig. 5 strikingly illustrates the organizational structure of the VDR ligand binding domain. Panel A shows the conventional ribbon structure; VDR comprises 12 helical ribbons that collectively create a three-layered sandwich structure. However, when one converts the ribbon structure to a Corey–Pauling representation of all of the atoms associated with the VDR (Fig. 5B), it can be seen that no portion of the ligand [hormone 1α,25(OH)2D3] is exposed to the surface environment. Said differently, the 1α,25(OH)2D3 is buried in the interior of the ligand binding domain, as illustrated
ANTHONY W. NORMAN
in Fig.5C. The dominant structural features of the ligand molecule, the 1α-, 3β-, and 25-hydroxyl groups are all interacting in a steroid-specific manner with various hydrogen bonding partners provided by the R-groups of the appropriate amino acids of the primary amino acid sequence of the VDR. The functional groups of the amino acids comprising the interior surface of the VDR ligand cavity interact very selectively with the ligand 1α,25(OH)2D3 to form a stable ligand–LBD complex (Fig. 5C). 1α,25(OH)2D3 is positioned in the LBD so that the A-ring enters the binding cavity first and then moves to the deep interior of the receptor so that the trailing side chain and the 25-hydroxyl group are just inside the LBD, which is closed as a consequence of helix 12 moving from its open position (characteristic of the unoccupied VDR LBD). The 1α-hydroxyl participates in two hydrogen bonds with Ser237 (helix-3) and Arg274 (helix-5). The 3β-hydroxyl group participates in two hydrogen bonds with Ser278 (helix-5) and Tyr143 (helix-2). The 25-hydroxyl group is hydrogen bonded to both His305 (loop helices 6/7) and His397 (helix-11). The C6-C7 single bond exhibits a trans conformation that deviates by 30° from the planar extended form of the ligand (see Fig. 12). With respect to conformational flexibility of the vitamin D secosteroids, it is generally accepted that the molecular mobility is displayed in organic solvents as well as in an aqueous environment similar to that encountered in biological systems. It is generally assumed for receptor–ligand interactions that the ligand is frozen out in a single shape or conformation that is dictated both by the structural constraints of the ligand and by the threedimensional architecture of the peptide chains that create the ligand binding domain of the receptor.
C. Vitamin D Binding Protein (DBP) The vitamin D–binding protein (DBP), also known as group-specific component (GC-globulin) [71,72], is the serum protein that serves as the transporter and reservoir for the principal vitamin D metabolites throughout the vitamin D endocrine system [73,74]. These include 25(OH)D3, the major circulating metabolite [KD ~ 6 × 10−9 M] [75], and the steroid hormone 1α,25(OH)2D3 [KD ~ 6 × 10−8 M]. DBP can be up to 5% glycosylated and is known to be one of the most polymorphic proteins, with three common allelic variants and more than 124 rare variants known [76]. DBP’s plasma concentration (4–8 µM) is approximately 20-fold higher than that of the total circulating vitamin D metabolites (~10−7 M). DBP binds 88% of the total serum 25(OH)D3 and 85% of serum
389
CHAPTER 23 Function Relationships for Both the Ligand and Receptor(s)
A
B
C
H1
H8 H10/11 H5
H3
1,25D H7 1,25D H12 H12
FIGURE 5 Three-dimensional structure of the VDR ligand binding domain for 118–425 (∆165–215 is the loop deletion) with its bound ligand 1α,25(OH)2D3 as determined by X-ray crystallography. (A) Ribbon structure of the VDR LBD bound to 1α,25(OH)2D3 as its ligand. Each of the twelve helices [e.g., helix 1 (H1)] can be identified by individual colors (as well as the labels on panel 3A). The colors of the individual helices are as follows: H1, green; H2, gray; H3, light blue; H4, purple; H5, yellow; H6/α-sheet, pink; H7, red; H8, aqua; H9, bronze; H10/H11, gold; H12, black; ∆165–215 (loop deletion), rust. The white regions represent loops and other flexible regions of the receptor molecule. The ligand, 1α,25(OH)2D3 has its atoms colored so that oxygen is red and carbon blue. (B) The VDR LBD is shown as a Corey–Pauling–Koltun (CPK) space-filling model. The position of helix-12 (black) in the closed position [69] effectively sequesters the ligand from the external environment of the protein, indicated by the absence of visible carbon and oxygen atoms from 1α,25(OH)2D3 in this view. (C) A view of the interior of the ligand binding domain of the VDR. This panel represents a “slice” across the middle plain of the view of panel (B). Clearly visible in the interior of the VDR LBD is the agonist ligand 1α,25(OH)2D3. (See color plate.)
1,25(OH)2D3, yet only 5% of the total circulating DBP actually carries vitamin D metabolites [77]. The three-domain structure of DBP is shown in Fig. 6A. Domains I, II, and III have been postulated to have evolved from a progenitor that arose from the triple repeat of a 192 amino acid sequence [78]; however, domain III is significantly truncated at the C-terminus. The position of the vitamin D metabolite and actin binding domains are specified in domains I and portions of domains I, II, and III, respectively. The X-ray crystallographic structures of the human DBP with a bound ligand of either 25(OH)D3 or 22-(m-hydroxyphenyl)-23,24,25,26,27-pentanor vitamin D3 has been determined [79] (see Fig. 6). Both crystals contained two nonidentical DBP molecules present in the asymmetric unit [79]. 25(OH)D3 was absent in the two recent X-ray crystal structures of DBP complexed with an actin monomer [80,81]. A second biological function of DBP is to serve as a scavenger in the circulating system for actin [80–82]. In the DBP: actin complex, the actin is bound in a groove formed
by the three domains, where domain I and III flank the actin monomer as it sits on top of domain II. DBP differs structurally from other serum transport proteins such as SHBG [83], retinol binding protein (RBP) [84], and thyroid hormone binding globules (TBG) [85], which all are β-barrel proteins. Uteroglobulin (UG), a progesterone binding protein, is small (17-kDa, α-helical) and lacks any homology to DBP. The most striking difference between DBP and the proteins in its family as well as other serum binding proteins is the topology of the ligand binding domain. The N-terminal region of DBP, helix 1–helix 6 of domain I, forms the ligand binding domain (LBD) where 25(OH)D3 and other vitamin D metabolites bind. When bound to DBP, vitamin D sterols including 1α,25(OH)2D3 remain highly exposed to the external environment, which is not the case for internally sequestered ligands bound to SHBG, RBP, progesterone binding globins UG, and TBG. Although virtually one whole surface of the 25(OH)D3 molecule (Fig. 6B) is exposed to the
390
ANTHONY W. NORMAN
I
A H2N–
Vitamin D Binding Domain
II
III
–COOH
FIGURE 6 Three-dimensional structure of the vitamin D binding protein (DBP). (A) Amino acid sequence of the 476 amino acids that form the primary amino acid sequence of DBP. This protein consists of three domains that have high sequence homology and are believed to have occurred evolutionarily as a consequence of gene duplication. The bold black bars indicate the presence of 14 disulfide linkages. See CD-ROM for 6A in color. (Continued)
CHAPTER 23 Function Relationships for Both the Ligand and Receptor(s)
B
391
C
FIGURE 6 Cont’d (B) The X-ray structure of DBP. This figure illustrates the three domains (I, yellow; II, light blue; III, green) of the DBP in a ribbon structure representation. The atoms of the ligand, 25(OH)D3, are colored so that oxygen is red and carbon blue. The ligand binding domain of the DBP is a crevice located on the surface of domain I. (C) This illustrates the CPK space-filling structure of DBP, with white regions indicating flexible regions of the molecule. Virtually the entire top surface (β-face) of the 25(OH)D3 (blue and red) is exposed to the external environment. (See color plate for parts B and C).
external environment when bound to the DBP, the protein’s affinity for 25(OH)D3 still remains high (~6 × 10−9 M), presumably because of the relative strength of the protein–ligand interactions on the α-face of 25(OH)D3. When 25(OH)D3 is bound to DBP, the region below the plane of the A, C, and D rings (Fig. 1), with its LBD, 25(OH)D3, in the α-chair (3β-OH equatorial) conformation, lies in the cleft with the α-face and the entire side chain making favorable hydrophobic contacts with the LBD. The DBP/25(OH)D3 complex is further stabilized by hydrogen bonds made by the two hydroxyls of the ligand. The conformation of 25(OH)D3 when bound to the DBP cleft has an overall L-type shape (hook-like) with the exocyclic alkene (C19) 31° below the CD-ring plane (see Fig. 11) and the side chain oriented almost perpendicular to and below the trans-hydrindane (CD-ring) (see Fig. 12) [79]. The distance between C17-O25 in the DBP/25-OH complex is 3.8 Å; thus the 25-hydroxyl group of the ligand is not extended away from the side chain, but rather curled back toward Tyr32 (Fig. 12). Table IV presents the important similarities and differences in the protein structures of the VDR and DBP as well as the shapes of their ligands. It is clear that each binding protein (DBP and VDR) has its own individual specification for generating a stable protein– ligand complex that is appropriate for its role in the biological system.
D. Membrane Receptor for 1α,25(OH)2D3 [VDRmem] The concept of the existence of a membrane receptor for 1α,25(OH)2D3 has its origins in the study of the
process of 1α,25(OH)2D3-mediated response of transcaltachia, or the rapid hormonal stimulation of intestinal Ca2+ absorption in the perfused chick intestine [86]. The most potent agonist is the natural hormone 1α,25(OH)2D3, which can stimulate the transport of calcium across the intestine within 1–2 min [44,87]. The current view of this laboratory and others is that transcaltachia is initiated by 1α,25(OH)2D3 via a membrane-localized receptor, the VDRmem. One hallmark of rapid responses that will be described in detail in a later portion of this chapter is that a different shape of 1α,25(OH)2D3 is optimal for genomic responses (bowl shaped 6-s-trans 1α,25(OH)2D3) as contrasted with the ligand shape for rapid responses with a 6-s-cis-shaped 1α,25(OH)2D3 (see Fig. 12). Thus, 1α,25(OH)2D3, which is conformationally flexible, is the known optimal agonist for both genomic responses and rapid responses. However, a major breakthrough occurred with the chemical synthesis of 1α,25(OH)2D3-lumisterol D3 (JN). This compound is permanently locked in the 6-s-cis position and while it is a full agonist for rapid responses binds only weakly to the VDRnuc [88]. An important verification of the concept of the existence of the proposed VDRmem is the isolation and purification of such a protein. Our laboratory described in 1994 a protein present in the basal lateral membrane of chick intestinal mucosa cells, which displays a 4400-fold enrichment in binding activity for 1α,25(OH)2D3 [47]. We believed at that time that the purified protein represented the VDRmem. Since that time, there have been two additional efforts to further purify and characterize this protein. One involves the work of Farach-Carson and Nemere, who have reported the existence of a membrane associated response system (MARRS) that is present in chick intestinal
392
ANTHONY W. NORMAN
TABLE IV
Comparison of VDR and DBP Crystal Structures
Receptor parameter Mol. wt. (kDa) Number of amino acid residues of intact protein Number of residues in X-ray structure Location of LBD on the protein Ligand–protein contacts Ligand A ring conformation 3β-OH C5–C6–C7–C8 torsion angle Ligand general shape A ring position C17–C20–C22–C23 torsion angle Side-chain orientation Distance from C-17 to oxygen on C-25 Overall ligand shape
VDRnuc 51 427 158 Interior pocket Uniquea β-Chair Axial +211° Bowl-shaped 6-s-trans 30° above C/D ring −156° Extended 6.9 Å Bowl
DBP 58 458 458 Surface cleft Uniquea α-Chair Equatorial +149° Twisted 6-s-trans 31° below C/D ring −70° Curled down 3.8 Å Hook
a The descriptor unique is used to indicate that the amino acid residues of the protein involved with the stabilizing hydrogen bond contact points with the respective ligands, 1α,25(OH)2D3 for VDR and 25(OH)D3 for DBP, are totally different.
cells to be a candidate as a ligand binding protein capable of initiating rapid responses to 1α,25(OH)2D3 and phosphate absorption [19]. Recent work in our laboratory has focused on isolation of the intestinal, kidney, or lung membrane fractions that contain caveolae. Caveolae are flask-shaped membrane invaginations that are enriched in sphingolipids and cholesterol which are commonly found in both caveolae and/or lipid rafts [22–24]. The caveolae enriched membrane fraction (CMF) is isolated from chick or rat intestine, kidney, or lung tissue. We find that under in vitro conditions there is a saturable binding of high-specific-activity tritiated 1α,25(OH)2D3 in the CMF. This binding activity is steroid specific for 1α-hydroxylated analogs and for 1α,25(OH)2D3 has a KD equal to 1.4 nM [49]. Current studies are focusing on the physiological relationship of the CMF 1α,25(OH)2D3 and the generation of physiological responses.
IV. STRUCTURE–FUNCTION EVALUATION OF SELECTED RAPID RESPONSES MEDIATED BY 1α,25(OH)2D3 A. Molecular Tools for Study of 1α,25(OH)2D3 Mediated Rapid Responses Figure 7 presents structures of two families of 1α,25(OH)2D3 analogs that have been found to be useful to study both rapid responses and genomic responses
mediated by both 1α,25(OH)2D3. The natural hormone, 1α,25(OH)2D3, is an agonist of both rapid responses (VDRmem) as well as genomic responses (VDRnuc). In the upper row, the 6-s-cis lumisterol analogs (JN and MZ) are full agonists for rapid responses mediated by 1α,25(OH)2D3. In contrast, the 1β,25(OH)2D3 (HL) is an antagonist of rapid responses mediated by 1α,25(OH)2D3 [89]. In the lower row, analogs MK and ML have been found to be antagonists of the nuclear receptor for 1α,25(OH)2D3 (VDRnuc) blocking gene activation [90,91]. These analogs are structurally related to the vitamin D metabolite 1α,25(OH)2D3-2623-lactone D3 [92,93]. The analogs MK and ML have no agonist actions on rapid responses [94] . Thus, this family of molecular tools has facilitated study of both rapid responses and genomic response systems to identify crucial structure–function relationships. In the following paragraphs, three case studies will be presented that rely heavily on the use of these molecular tools.
B. Rapid Response Case Studies 1. CASE STUDY 1: PATCH CLAMP ANALYSIS OF ROS 17/2.8 CELLS
It has been demonstrated in osteoblasts that the steroid hormone 1α,25(OH)2D3 facilitates the opening of L-type Ca2+ channels as well as prolongation of the open state [20,95]; these responses occur in seconds after the addition of 1α,25(OH)2D3 and represent
393
CHAPTER 23 Function Relationships for Both the Ligand and Receptor(s)
RAPID RESPONSES VDRmem ANTAGONIST HO H
OH
H
OH
OH
H
H
1α,25(OH)2D3
H
OH
OH
H
H
HO
HO
H
OH OH
H
H
HO
1α,25(OH)2-Lumisterol3 JN
HO 1α,24(OH)2-Lumisterol3 MZ
1β,25(OH)2D3 HL
GENOMIC RESPONSES VDRNuc
H
H OH H
H
HO 1α,25(OH)2D3
OH
HO
H O
O O
ANTAGONISTS
O
OH
HO
OH
TEI-9647 MK
H
TEI-9648 ML
FIGURE 7 Molecular tools for study of rapid responses and genomic responses mediated by 1α,25(OH)2D3 and related analogs. The structures shown for 1α,25(OH)2D3 indicate (top) the 6-cis conformational shape and (bottom) the 6-trans conformations of 1α,25(OH)2D3. The natural hormone is a full agonist for all known rapid responses and genomic responses. In the remainder of the figure, the upper panel presents three structures of analogs in the 6-s-cis shape of 1α,25(OH)2D3 that have proven very useful for study of rapid responses. The analogs JN and MZ are agonists for mediation of rapid responses (see text in later figures). Analog HL or 1β,25(OH)2D3 is an antagonist only of rapid responses [88,89]. The structures shown on the bottom row (MK and ML) are analogs of a naturally occurring metabolite of 1α,25(OH)2D3, namely the 1α,25(OH)2-26,23-lactone-D3. In analogs MK the 25-hydroxyl group has been dehydrated to give a 26–27 double bond. These analogs have been shown to be antagonists of the nuclear VDR and are without effect on rapid responses [90,94].
breakthrough studies. In follow-up studies, Zanello and Norman [96] carried out a structure–function analysis in ROS 17/2.8 cells to identify the structural components of 1α,25(OH)2D3 that were crucial for opening the L-gated Ca2+ channels. In these studies we also learned that 1α,25(OH)2D3 was also able to open chloride channels ROS 17/2.8 cells. Figure 8A summarizes a structure–function study evaluating the ability of 1α,25(OH)2D3 and related analogs to open chloride channels. The results emphasize the structural specificity inherent in the receptor that binds these analogs to initiate rapid responses. Thus, the 6-s-cis-locked 1α,25(OH)2D3-lumisterol is a full agonist in comparison to 1α,25(OH)2D3 for opening of chloride channels. However, the companion 6-s-trans–locked analog 1α,25(OH)2-tachysterol3 is without agonist activity. These results emphasize that rapid responses are initiated only by vitamin D secosteroids that have a shape
similar to the 6-s-cis. Also, it is clear that other nonsecosteroids such as cholesterol or 17β estradiol are impotent in this system. Use of the analog HL [1β,25(OH)2D3] indicates that the membrane receptor has no agonistic response to this ligand. However, as shown by the combined use of 1β,25(OH)2D3 with 1α,25(OH)2D3, it was found that the beta isomer functions as an antagonist. This observation is studied in further detail and the results shown in Fig. 8B. Here a dose-response experiment employing 1α,25(OH)2D3 as an agonist was performed; as is typical of most rapid responses to this hormone, the dose-response curve is biphasic with maximal activity at about 5–20 nM with a fall-off in response at concentrations > 50 nM. However, in the combined presence of 1β,25(OH)2D3 and 1α,25(OH)2D3, it is clear that 1β,25(OH)2D3 is able to block all dose levels of the natural hormone. These results were in agreement with earlier studies
394 A
ANTHONY W. NORMAN
5
OH
B
H OH
4 3
Fold increase of the 2 currents
H
5
HO 1α,25(OH)2-Lumisterol3 JN
H OH H
1α,25(OH)2-Tachysterol3 JB
HO
OH
1 0 1α,25(OH)2D3 HL 1α,25D+HL JN 1β,25D
1α,25-D3 alone 1α,25-D3 +1β,25-D3
6
H
ANTAGONIST
Fold 4 increase 3 of chloride currents 2
HO
0 Cholesterol 17βEstradiol
OH H 1α,25(OH)2D3 HL
1
JB
H
OH
0.05
0.5 5 1α,25(OH)2D3 (nM )
50
FIGURE 8 Structure–function evaluation of ligand opening of chloride currents in ROS 17/2.8 cells. (A) Comparative affects of analogs of 1α,25(OH)2D3 and other steroids on chloride currents and ROS 17/2.8 cells. For further details see (AWN 660). (B) Fold increase of chloride currents mediated by 1α,25(OH)2D3 in the absence and presence of the antagonist 1 nM 1β,25(OH)2D3. This data resulted from patch clamping of single isolated cells. 1α,25(OH)2D3 or analogs HL [1β,25(OH)2D3] were added at least 3 min prior to the electrical measurements. For further details see [96].
where analog HL was also found to be capable of inhibiting 1α,25(OH)2D3 mediated transcaltachia [89]. In data not presented, the ability of 1β,25(OH)2D3 to bind to the VDRnuc was evaluated in comparison to 1α,25(OH)2D3; it can only bind approximately 0.5% as well as the natural hormone. From these studies, we conclude that the agonist effect of 1α,25(OH)2D3 in osteoblasts at the site of the membrane level seems to be determined by some structural features of the hormone molecule that may be crucial for its interaction with the putative membrane receptor in the cell surface. 2. CASE STUDY 2: 1α,25(OH)2D3 STIMULATION OF INSULIN SECRETION
Pancreatic β-cells have been found to have a receptor for vitamin D and to be a target of 1α,25(OH)2D3 [39,97]. Although there are many studies on the effect of 1α,25(OH)2D3 on insulin secretion [98–102], the precise role of vitamin D in β-cell function is not yet clear. Most investigation has been restricted to the chronic effects of 1α,25(OH)2D3 and its analogs on insulin secretion. There have been only a small number of investigations concerning the rapid nongenomic effect of 1α,25(OH)2D3 on pancreatic β-cell function. Among these studies, Sergeev and Rhoten [103] have reported that the administration of 1α,25(OH)2D3 evoked oscillations of intracellular Ca2+ in a pancreatic β-cell line within a few minutes. However, the detailed mechanism of the nongenomic effect of 1α,25(OH)2D3 through signal transduction via VDRmem in pancreatic β-cells remains unclear. In this study, we have compared the effect of two 1α,25(OH)2D3 analogs, 1α,25(OH)2lumisterol and 1β,25(OH)2D3, in the rapid action of 1α,25(OH)2D3 on insulin secretory mechanisms, and we find evidence of nongenomic stimulation
by vitamin D on insulin secretion in normal rat pancreatic β-cells. As shown in Fig. 9, 1α,25(OH)2lumisterol3 (JN) augmented the 16.7 nM glucose–induced insulin release from rat islets in a dose-dependent manner. This fact was primarily apparent in the second phase of insulin release induced by 16.7 nM glucose, which occurs after 10 min. Also in data not shown, the stimulation was reversible within 15 min after withdrawal of analog JN from the medium. Strikingly, the stimulatory effect of insulin release was completely abolished by addition of 1β,25(OH)2D3 (HL), whereas analog HL alone did not affect glucose-induced insulin release. This result strongly suggests that 1α,25(OH)2D3 and its analog JN function in a nongenomic fashion in the pancreas β-cell. In pancreatic β-cells, signals generated through intracellular glucose metabolism are believed to close the KATP channel on the plasma membrane. The resulting decrease in K+ conductance leads to depolarization of the membrane, with subsequent opening of a voltagedependent Ca2+ channel. Ca2+ influx through these channels then increases, leading to a rise in [Ca2+]i, which eventually triggers exocytosis of insulin granules [103,104]. The secretory enhancement by 1α,25(OH)2− lumisterol3 was observed also in the presence of diazoxide (an opener of KATP channels), a depolarizing concentration of K+, and 16.7 mM glucose (which is known to produce so-called KATP channel-independent insulin release through metabolic signaling derived from glucose) [105,106]. On the other hand, both insulin release and [Ca2+]i dynamics were not affected by 1α,25(OH)2lumisterol3 in the presence of a basal concentration (3.3 mM) of glucose, indicating that 1α,25(OH)2lumisterol3 itself is not a depolarizing agent,
CHAPTER 23 Function Relationships for Both the Ligand and Receptor(s)
Insulin release (ng/islet/ 30 min)
6
4
2
0 JN (M) 0 HL (M)
10−10
10−9
10−8
10−7
10−8 10−8
10−8
FIGURE 9 Glucose-induced insulin secretion from rat pancreatic islets stimulated by 6-s-cis 1α,25(OH)2D3 2-lumisterol (analog JN). In this study [120] islets of Langerhans were isolated from rats by collagenase digestion. Insulin release from intact islets was monitored using either static incubation or diffusion system. The results shown in this panel are obtained from a study of the concentration-dependent effect of 1α,25(OH)2-lumisterol (JN) on 16.7 nM glucose-induced insulin secretion from rat pancreatic islets. In some incubations, the analog of 1β,25(OH)2D3 (HL) was added simultaneously with the analog JN. Values are expressed at the means ± SE of five determinations in the same experiment. *P < 0.05 vs control (16.7 nM glucose alone).
like sulfonylurea, which induces KATP channel closure to produce membrane depolarization. 1α,25(OH)2lumisterol3, therefore, acts distal to membrane depolarization in glucose-induced insulin secretion, and its enhancement of insulin secretion could result from rising [Ca2+]i or augmentation of Ca2+ efficacy in the exocytotic system. However, in the presence of depolarizing concentration of K+ with diazoxide, 1α,25(OH)2lumisterol3 did not enhance insulin release without a high concentration of glucose, and even with a high concentration of glucose, 1α,25(OH)2lumisterol3 failed to enhance insulin release from diazoxidehyperpolarized islets. 1α,25(OH)2lumisterol3-enhanced insulin release, therefore, requires both activated glucose metabolism and plasma membrane depolarization, as suggested also by the observation that mannoheptulose prevents 1α,25(OH)2lumisterol3 from enhancing insulin secretion, even in the depolarizing condition. 3. CASE STUDY 3: ACTIVATION OF PHOSPHATIDYL INOSITOL 3-KINASE BY 1α,25(OH)2D3
Recent studies suggest that 1α,25(OH)2D3 plays an important role in the cardiovascular system through its receptors in the heart and in vascular smooth muscle cells (VSMC) [107,108]. In particular, 1α,25(OH)2D3
395
has been shown to regulate calcium homeostasis, modulate growth, and increase calcification in smooth muscle cells [109–111]. Indeed, the mitogenic role of 1α,25(OH)2D3 in VSMC has been previously reported [110–112], although its effect on the migration has not been investigated. Therefore, the purpose of this study was to determine whether 1α,25(OH)2D3 can promote VSMC migration, and if so, to determine whether the mechanism is mediated by the genomic or nongenomic effects of VDR. As shown in Fig. 10A, the effect of 1α,25(OH)2D3 on vascular smooth muscle cell (VSMC) migration was assessed using a modified Boyden transwell cell culture chamber. The addition of 1α,25(OH)2D3 to the lower compartment of the Boyden chamber induced a concentration-dependent increase of VSMC migration in the upper chamber over the time frame of 4 hr; in this assay there is actual physical movement of the VSMC cell, i.e., a chemotaxis effect happens quite rapidly. Indeed, the rapid response occurred over physiological concentration range of 0.1–100 nM 1α,25(OH)2D3. Overall, the migration rate increased as a consequence of the dose response studies from 19% to 42% over control levels. A role for PI3-kinase in VSMC migration has been previously reported [113,114]. As shown in Figure 10B, treatment of VSMC with 1α,25(OH)2D3 activated PI3kinase in a time course–dependent and dose-dependent fashion (upper two rows of Fig. 10). The PI3-kinase was dramatically activated over a concentration range of 0.12–1 nM 1α,25(OH)2D3. Shown in the bottom row of Fig. 10B, the two inhibitors of PI3-kinase, namely wortmannin and LY294002, both blocked activation by 1α,25(OH)2D3. In contrast, the RNA polymerase inhibitor, 5,6-dichloro-1-β-D-ribofuranosylben3imidazole (DRB), was without effect on 1α,25(OH)2D3 activation of PI3-kinase. In data not shown, DRB, which is an inhibitor of genomic responses, was shown to block gene transcription induced by TMF-α. In additional data not presented, using 1α,25(OH)2lumisterol (JN) duplicated all of the results shown in Fig. 10B. In the results shown in Fig.10C, a comparison of the effect of the two specific molecular tool antagonists, namely analog HL, rapid responses, and analog MK, genomic responses, on 1α,25(OH)2D3 activation of PI3-kinase. The rapid response measured as a consequence of 1α,25(OH)2D3 induced cell migration as well as the activation of PI3-kinase are shown; both rapid responses were blocked by 1β,25(OH)2D3 (HL), but were not affected by the genomic VDR receptor antagonist, analog MK. Collectively, these findings suggest there exists a rapid response membrane receptor that mediates the nongenomic actions of 1α,25(OH)2D3 VSMC and
396
1,25D3-induced cell migration (increased in %)
A
ANTHONY W. NORMAN
B
50
*
40
*
Time course PIP3
* C
30
10′
30′
60′
1
10
1,25D3
*
20
5′
Dose response PIP3
10
0
C
0
0.01
0.1 1 1,25D3 (nmol/L)
10
0.01
100
0.1
1,25D3 (nmol/L) Inhibitors PIP3
C
−
Wort
LY
−
DRB
1,25D3
C 1,25D3-induced cell migration (increase in %)
40
30
* *
20
10
0
−
Control
HL
MK
1,25D3 PIP3 C
−
HL
MK
1,25D3
FIGURE 10 Evaluation of 1α,25(OH)2D3 mediated rapid responses in vascular endothelial cells with concordant activation of phosphatidyl-inositol-3-kinase activation. (A) 1α,25(OH)2D3 induces vascular smooth muscle cell (VSMC) migration. The effect of increasing concentrations of 1α,25(OH)2D3 on VSMC migration is shown. The experiments were performed three times in triplicates. *P < 0.05 compared with unstimulated cells. (B) 1α,25(OH)2D3 induces PI3-kinase activity in VSMCs. Time course: Time-dependent 5–60 min increase in PI3-kinase activity induced by 1α,25(OH)2D3 (1 nmol/L) compared with unstimulated or control cells. Dose–response: Effect of increasing concentrations of 1α,25(OH)2D3 of PI3-kinase activity. Cells were stimulated for 10 min with 1α,25(OH)2D3 (0.01 to 10 nmol/liter) were left untreated. Inhibitors: Effect of wortmanin (WORT, 30 nmol/liter), LY294002 (L, 10 µmol/liter), or DRB (50 µmol/liter) on PI3-kinase activity induced by 1α,25(OH)2D3 (10 nmol/liter). All experiments in this panel were performed three times. Results taken from [120]. (C) Effect of 1β,25(OH)2D3 (HL) and (23S)-25-dehydro-1α-OH-D3-26,23-lactone (MK on 1α,25(OH)2D3induced VSMC migration and PI3-kinase activity. (Top) VSMC migration was induced by 1 nmol/liter of 1α,25(OH)2D3 in the presence or in the absence of 50 nmol/liter of HL or MK. Experiments were performed three times in triplicates. *P < 0.05 compared with unstimulated cells (control). (Bottom) Effect of HL or MK (50 nmol/liter) on PI3-kinase activity induced by 1α,25(OH)2D3 (1 nmol). Experiments were performed three times with comparable results (from the following study [121]).
397
CHAPTER 23 Function Relationships for Both the Ligand and Receptor(s)
PI3-kinase. The mechanism underlying this nongenomic response is not completely understood. The PI3kinase assays suggest that a signaling complex containing both the VDR and PI3-kinase is formed after the addition of 1α,25(OH)2D3. It is possible that this activation process could occur in the plasma membrane or caveolaeenriched membrane fraction; alternatively, it is possible that the activation process of PI3-kinase occurred in the cell cytosol. A similar observation to that discussed earlier for 1α,25(OH)2D3 was reported concerning the nongenomic/rapid response effects of estradiol and endothelial cells [115]. The estrogen receptor (ER) was also shown to activate PI3-kinase. However, additional evidence suggested that the ER formed a heterodimer with the p85α regulating subunit of PI3-kinase.
C. Summary of Structure–Function Studies Table V represents a structure–function summary analysis of the process of transcaltachia (the rapid hormonal stimulation of calcium absorption) as well as the three case studies (insulin secretion, smooth muscle cell migration, and opening of chloride channels). In each of these systems, the natural hormone 1α,25(OH)2D3 was a full agonist. Further, the 6-s-cis locked 1α,25(OH)2lumisterol3 was also a full agonist. In the two systems that were evaluated, transcaltachia [88] and chloride channel opening [96], the 6-s-trans restricted analog 1α,25(OH)2tachysterol3 was without any agonist effect for these rapid responses. Finally, 1β,25(OH)2D3 (analog HL) was a potent antagonist of rapid responses mediated by the natural hormone. Collectively, the structure–function studies carried out in these four systems with respect to rapid responses
clearly indicate that there is a membrane-associated receptor that can distinguish structural differences in potential ligand agonists (e.g., 6-s-cis vs 6-s-trans).
V. SUMMARY This chapter has focused on key structure–function relationships that bear on differences between the classical genomic responses mediated by the VDR in contrast to the putative VDR membrane receptor which is responsible for rapid responses. A third important player in the biological systems where both genomic and rapid responses occur is the plasma transport protein, the vitamin D binding protein (DBP). All of the experimental results summarized here emphasize that genomic responses and rapid responses are mediated by different shapes of the agonist hormone 1α,25(OH)2D3. Although the agonist 1α,25(OH)2D3 is conformationally flexible because of three facets of its structure (see Fig. 1), the dominant structural aspect relevant to the present discussion is the fact that it is a secosteroid. More specifically, the 9–10 carbon bond is broken in vitamin D–related compounds, thus imparting complete 360° capability to rotate around the 6–7 single carbon bond (see Fig. 11). This figure emphasizes by 90° increments the change in shapes moving from the planar-6-s-cis shape at 0° through 90° (α triene) to 180° (planar 6-s-trans shape), and to 270° (β-triene) and then finally 90° back to the starting point of 0°. In solutions (including the cell cytosol) the free 1α,25(OH)2D3 ligand would present literally millions of shapes to potential receptors. However the nature of receptor– ligand interaction is such that only one preferred shape will be selected to form a receptor–ligand complex.
TABLE V Summary of Case Studies: 1α,25(OH)2D3 Structure–Function Analysis of Rapid Responses Case studies Response → Agonist ↓ 1α,25(OH)2D3 conformationally flexible (JN) 6-s-cis locked (JB) 6-s-trans locked (HL) 1β,25(OH)2D3 Reference
Chick intestine: Transcaltachia (rapid Ca2+ absorption)
Osteoblasts: Opening Cl− channels & secretion
Pancreatic β-cells: Insulin secretion
Endothelial cells: Smooth muscle cell migration
Yes
Yes
Yes
Yes
Yes No Antagonism Norman, Bouillon, Farach-Carson [88]
Yes No Antagonism Zanello, Norman [96]
Yes — Antagonism Ishida, Norman [120]
Yes — Antagonism Liao, Norman [121]
398
ANTHONY W. NORMAN
360° Rotation around 6,7-single bond of 1α,25(OH)2D3 Planar 6-s-cis SC D H C H
VDRmem ligand agonist
A OH OH 0° 315°
HO β-Triene
SC D C
A H
OH 270°
45°
Rotation above CDplane
SC 90°
Free 1,25-D3
211°
180°
H H
α-Triene
A
Rotation below CDplane 135°
225°
D C
OH
HO
149°
SC D C
Bound 1,25-D3 VDR-nuc
OH
H H
Bound 25-D3 DBP-LBD
A
Planar 6-s-trans
OH
FIGURE 11 Biologically relevant ligand shapes of 1α,25(OH)2D3 for rapid responses, genomic responses, and binding to vitamin D binding protein (DBP). This illustrates the consequences of 360° rotation about the 6,7 carbon bond of 1α,25(OH)2D3. Four steps of successive 90° rotations are illustrated. Each intermediate ligand structure is a different shape, particularly with respect to the position of the critical 1α-hydroxyl and the orientation of the A-ring in relation to the plane of the C/D rings. The three arrows indicate the degrees of rotation about the 6,7 carbon bond of the ligand for the VDRmem (0°), the ligand for DBP (+149°), and the ligand for VDRnuc (+211°).
A
B
1α,25(OH)2D3 VDRnuc
C
1α,25(OH)2-lumisterol VDRmem
25(OH)D3 DBP
FIGURE 12 Space-filling and stick representations illustrating the clear differences in shapes for the optimal VDRmem ligand, 1α,25(OH)2-lumisterol (planar molecule; analog JN in Fig. 11), the optimal DBP 25(OH)D3 or 1α,25(OH)2D3 ligand (twisted 6-s-trans with the A-ring C10-C19 exocyclic alkene 30° below the plane of the C/D ring), and the optimal VDRnuc ligand 1α,25(OH)2D3 (bowl shaped twisted 6-s-trans with the A-ring 30° above the plane of the C/D ring). Each of the illustrated molecules has its 25-OH pointed to the right and the A-ring to the left. Also, there are significant differences in the orientation of the side chain for the VDRnuc (north) and DBP (south); the optimal side-chain orientation of the VDRmem ligand has not yet been precisely defined.
399
CHAPTER 23 Function Relationships for Both the Ligand and Receptor(s)
Careful inspection of Fig. 11 and 12 will identify the three optimal shapes of the conformationally flexible natural hormone 1α,25(OH)2D3. The optimal ligand shape for the VDRnuc is the bowl shaped 6-s-trans where the A ring is rotated 211° from that of the planar 6-s-trans 1α,25(OH)2D3. In contrast, the optimal ligand for the DBP is a structure which has been rotated 149° from that of the planar 6-s-cis ligand at 0°. Finally, the optimal ligand for the VDRmem is the planar 6-s-cis shape of 1α,25(OH)2D3. The contrast in shapes between the optimal ligand for VDRnuc, VDRmem, and DBP is summarized in Fig. 12. The difference in shapes is dramatic and emphasizes an opportunity as well as a challenge for shape-selective drug development. It is possible to carry out chemical syntheses of conformationally restricted analogs such that a product putative drug could be developed that only interacts with VDRnuc without significant actions
on VDRmem and vice versa. An overriding drug development challenge relates to whether any drug that might be optimal for VDRmem would also bind significantly to DBP, which might be a necessity for transport of the drug to its cellular sites of action. Figure 13 is a summary figure that attempts to integrate the signal transduction pathways which are used by the nuclear receptor (VDRnuc) and the membrane receptor (VDRmem) of 1α,25(OH)2D3. Figure 13 emphasizes the complexity of overlapping and interconnecting signal transduction pathways. Whereas there is a solid and secure foundation describing 1α,25(OH)2D3mediated genomic actions, 1α,25(OH)2D3-mediated rapid responses are still in their developmental phase. One extraordinary challenge will be to identify and fully biochemically characterize the VDRmem. An equally daunting task will be to define which signal transduction pathways mediate rapid responses in the
1α,25(OH)2D3 = RAPID RESPONSES Ligand shape matters
Affects target cell biology
1α,25(OH)2D3
PLASMA MEMBRANE Ca 2+
Cell nucleus
Ca 2+
Caveolae Signal transduction PI3K
Tissues with rapid responses Pancreas B cell Adipocytes Vascular smooth muscle Intestine Monocytes Osteoblasts
G PROTEIN
GENE EXPRESSION
Altered genomic responses Osteocalcin promoter 24-OHase promoter Alkaline phosphatase NB4 cell differentiation Micro array analysis
PKC
Phospholipase C
Systems PKC RAS/MAP Kinase
Second messengers Phosphoproteins “Cross-talk” RAF/MAP Kinase
PI3 ′Kinase
PtdIns-3,4,5-P3
FIGURE 13 Schematic model describing how the conformationally flexible 1α,25(OH)2D3 can interact with a nuclear receptor to generate genomic responses or a plasma membrane receptor to generate rapid responses. Binding of 1α,25(OH)2D3 to the membrane surface receptor may result in the activation of one or more second messenger systems, including phospholipase C, protein kinase C, G protein coupled receptors, or phosphatidyl-inositol-3-kinase (PI3). There are a number of possible outcomes including opening of the voltage-gated calcium channel or generation of the indicated second messengers. Some of these second messengers, particularly RAF/MAP kinase, may engage in cross-talk with a nucleus to modulate gene expression. Evidence has been presented that rapid responses can modulate the list of specific genes tabulated in the figure [122].
400
ANTHONY W. NORMAN
various cell systems. A further important question is to answer whether the VDRmem can communicate with the nucleus of the cell to modulate gene transcription. In data not shown but summarized in Fig. 13, there are at least five different systems where molecular biological evidence has been obtained for the process of crosstalk. In these studies, 6-s-cis locked analogs have been shown to modulate genomic responses at relatively early time intervals (2–4 hr). Without a doubt, it will be intriguing to monitor developments in this field in the future.
to the flickering gate properties of certain membrane channels that open and close rapidly. In contrast, over a long time-frame (steady state equilibrium), 1α,25(OH)2D3 occupancy of the genomic pocket is significantly favored over the alternative pocket because the increase in hydrophobic stability prolongs the half-life of the ligand in the genomic pocket. A related model to describe ligand/receptor induced dissociation of rapid from genomic responses was included in the comprehensive analysis of nongenotropic, sex-nonspecific signaling by the ER and AR [228].
VI. ADDENDUM
References
Recent evidence suggests that both the VDR and the estrogen receptor (ER) are localized to plasma membrane caveolae and they are also required for initiation of non-genomic rapid responses [224,225]. Using the atomic coordinates of the X-ray structure of the VDR and computer modeling, we have been able to identify the presence of a putative alternative LBD in the VDR that can accommodate via computer docking either the conformationally flexible 1,25(OH)2D3 or the 6-s-cis locked 1α,25(OH)2lumisterol; both are known to be full agonists for rapid responses [226]. As a consequence we have proposed that a conformational ensemble receptor model can give insight into the complex ligand-receptor relationships for the VDR that determine whether the genomic (G) or alternative (A) pocket becomes occupied by a ligand leading to either a genomic or rapid response [227]. In this model the unbound receptor macromolecules exist in the cytoplasm as multiple, equilibrating VDR conformations that follow the laws associated with standard statistical distributions. Thus, 1,25(OH)2D3 or a drug ligand would sample the existing population of VDR conformations available thus generating a new statistical distribution of the ligand-receptor conformers. The population of each VDR-ligand conformer is dependent on the ligand’s structure and chemistry and the subsequent receptor A- versus G-pocket stabilities. The genomic pocket portal is believed to be controlled by the position of helix-12 and a finite time interval will be required for the helix to move to an open position; thus, entrance to the genomic pocket is gated. While details of access to the alternative pocket are not yet known, it is proposed that given the increased loop character of the amino acid residues forming the lid of the alternative pocket, the energy barrier between the opened and closed states is far less than that required for helix-12 repositioning. This effectively makes the alternative pocket more accessible to the ligand than the genomic pocket. This mechanism may be analogous
1. Bouillon R, Okamura WH, Norman AW 1995 Structure– function relationships in the vitamin D endocrine system. Endocr Rev 16:200–257. 2. Henry HL, Norman AW 1991 Metabolism of vitamin D. In: Coe FL, Favus MJ (eds) Disorders of Bone and Mineral Metabolism. Raven Press, New York. 3. Henry HL, Norman AW 1984 Vitamin D: Metabolism and biological action. Ann Rev Nutr 4:493–520. 4. Midland MM, Plumet J, Okamura WH 1993 Effect of C20 stereochemistry on the conformational profile of the side chains of vitamin D analogs. Bioorg Med Chem Lett 3: 1799–1804. 5. Okamura WH, Palenzuela JA, Plumet J, Midland MM 1992 Vitamin D: Structure–function analyses and the design of analogs. J Cell Biochem 49:10–18. 6. Wing RM, Okamura WH, Pirio MR, Sine SM, Norman AW 1974 Vitamin D3: Conformations of vitamin D3, 1α,25dihydroxyvitamin D3, and dihydrotachysterol3. Science 186: 939–941. 7. Okamura WH, Norman AW, Wing RM 1974 Vitamin D: Concerning the relationship between molecular topology and biological function. Proc Natl Acad Sci USA 71:4194–4197. 8. Norman AW, Okamura WH, Farach-Carson MC, Allewaert K, Branisteanu D, Nemere I, Muralidharan KR, Bouillon R 1993 Structure–function studies of 1,25-dihydroxyvitamin D3 and the vitamin D endocrine system. 1,25-Dihydroxypentadeuterio-previtamin D3 (as a 6-s-cis analog) stimulates nongenomic but not genomic biological responses. J Biol Chem 268:13811–13819. 9. Renaud JP, Rochel N, Ruff M, Vivat V, Chambon P, Gronemeyer H, Moras D 1995 Crystal structure of the RARgamma ligand-binding domain bound to all-trans retinoic acid. Nature 378:681–689. 10. Wagner RL, Apriletti JW, McGrath ME, West BL, Baxter JD, Fletterick RJ 1995 A structural role for hormone in the thyroid hormone receptor. Nature 378:690–697. 11. Bishop JE, Collins ED, Okamura WH, Norman AW 1994 Profile of ligand specificity of the vitamin D binding protein for 1α,25(OH)2-vitamin D3 and its analogs. J Bone Miner Res 9:1277–1288. 12. Norman AW, Hurwitz S 1993 The role of the vitamin D endocrine system in avian bone biology. J Nutr 123:310–316. 13. Norman AW 1998 Receptors for 1α,25(OH)2D3: Past, present, and future. J Bone Miner Res 13:1360–1369. 14. Lowe KE, Maiyar AC, Norman AW 1992 Vitamin D-mediated gene expression. Crit Rev Eukar Gene Exp 2:65–109.
CHAPTER 23 Function Relationships for Both the Ligand and Receptor(s)
15. Hannah SS, Norman AW 1994 1α,25(OH)2-vitamin D3regulated expression of the eukaryotic genome. Nutr Rev 52:376–382. 16. Norman AW, Collins ED 2001 Vitamin D and gene expression. In: Moussa NN, Berdanier C (eds) Nutrient–Gene Interactions in Health and Disease. CRC Press, Orlando, FL. 17. Norman AW, Bishop JE, Bula CM, Olivera CJ, Mizwicki MT, Zanello LP, Ishida H, Okamura WH 2002 Molecular tools for study of genomic and rapid signal transduction responses initiated by 1α,25(OH)2-vitamin D3. Steroids 67:457–466. 18. Norman AW 2002 Nuclear receptor structure (VDR) and ligand specificities for genomic and rapid biological responses. In: Bilezikian JP, Raisz LG, Rodan GA (eds) Principles of Bone Biology. Academic Press, San Diego. 19. Nemere I, Farach-Carson MC 1999 Membrane receptors for steroid hormones: A case for specific cell surface binding sites for vitamin D metabolites and estrogens. Biochem Biophys Res Commun 248:443–449. 20. Caffrey JM, Farach-Carson MC 1989 Vitamin D3 metabolites modulate dihydropyridine-sensitive calcium currents in clonal rat osteosarcoma cells. J Biol Chem 264:20265–20274. 21. Berry DM, Antochi R, Bhatia M, Meckling-Gill KA 1996 1,25-Dihydroxyvitamin D3 stimulates expression and translocation of protein kinase Cα and Cδ via a nongenomic mechanism and rapidly induces phosphorylation of a 33-kDa protein in acute promyelocytic NB4 cells. J Biol Chem 271:16090–16096. 22. Anderson RG, Jacobson K 2002 A role for lipid shells in targeting proteins to caveolae, rafts, and other lipid domains. Science 296:1821–1825. 23. Isshiki M, Ying YS, Fujita T, Anderson RG 2002 A molecular sensor detects signal transduction from caveolae in living cells. J Biol Chem 277:43389–43398. 24. Razani B, Woodman SE, Lisanti MP 2002 Caveolae: from cell biology to animal physiology. Pharmacol Rev 54:431–467. 25. Ortí E, Bodwell JE, Munck A 1992 Phosphorylation of steroid hormone receptors. Endocr Rev 13:105–128. 26. Darwish HM, Burmester JK, Moss VE, DeLuca HF 1993 Phosphorylation is involved in transcriptional activation by the 1,25-dihydroxyvitamin D3 receptor. Biochim Biophys Acta Lipids Lipid Metab 1167:29–36. 27. Jurutka PW, Hsieh J-C, Haussler MR 1993 Phosphorylation of the human 1,25-dihydroxyvitamin D3 receptor by cAMP-dependent protein kinase, in vitro, and in transfected COS-7 cells. Biochem Biophys Res Commun 191: 1089–1096. 28. Hsieh JC, Jurutka PW, Nakajima S, Galligan MA, Haussler CA, Shimizu Y, Shimizu N, Whitfield GK, Haussler MR 1993 Phosphorylation of the human vitamin D receptor by protein kinase C. Biochemical and functional evaluation of the serine 51 recognition site. J Biol Chem 268:15118–15126. 29. Jurutka PW, Hsieh J-C, MacDonald PN, Terpening CM, Haussler CA, Haussler MR, Whitfield GK 1993 Phosphorylation of serine 208 in the human vitamin D receptor. J Biol Chem 268:6791–6799. 30. Freedman LP, Arce V, Perez Fernandez R 1994 DNA sequences that act as high affinity targets for the vitamin D3 receptor in the absence of the retinoid X receptor. Mol Endocrinol 8:265–273. 31. Liu M, Freedman LP 1994 Transcriptional synergism between the vitamin D3 receptor and other nonreceptor transcription factors. Mol Endocrinol 8:1593–1604. 32. Walters MR 1992 Newly identified actions of the vitamin D endocrine system. Endocr Rev 13:719–764.
401
33. Norman AW, Roth J, Orci L 1982 The vitamin D endocrine system: Steroid metabolism, hormone receptors and biological response (calcium binding proteins). Endocr Rev 3:331–366. 34. Carlberg C, Quack M, Herdick M, Bury Y, Polly P, Toell A 2001 Central role of VDR conformations for understanding selective actions of vitamin D3 analogues. Steroids 66:213–221. 35. Haussler MR, Whitfield GK, Haussler CA, Hsieh JC, Thompson PD, Selznick SH, Dominguez CE, Jurutka PW 1998 The nuclear vitamin D receptor: Biological and molecular regulatory properties revealed. J Bone Miner Res 13:325–349. 36. Haussler MR, Myrtle JF, Norman AW 1968 The association of a metabolite of vitamin D3 with intestinal mucosa chromatin, in vivo. J Biol Chem 243:4055–4064. 37. Haussler MR Norman AW 1969 Chromosomal receptor for a vitamin D metabolite. Proc Natl Acad Sci USA 62:155–162. 38. Christakos S, Norman AW 1981 Studies on the mode of action of calciferol XXIX—Biochemical characterization of 1,25-dihydroxyvitamin D3 receptors in chick pancreas and kidney cytosol. Endocrinology 108:140–149. 39. Ishida H, Norman AW 1988 Demonstration of a high affinity receptor for 1,25-dihydroxyvitamin D3 in rat pancreas. Mol Cell Endocrinol 60:109–117. 40. Roth J, Bonner-Weir S, Norman AW, Orci L 1982 Immunocytochemistry of vitamin D-dependent calcium binding protein in chick pancreas: Exclusive localization in B-cells. Endocrinology 111:2216–2218. 41. Tsoukas CD, Provvedini DM, Manolagas SC 1984 1,25Dihydroxyvitamin D3: A novel immunoregulatory hormone. Science 224:1438–1440. 42. Adorini L 2002 Immunomodulatory effects of vitamin D receptor ligands in autoimmune diseases. Int Immunopharmacol 2:1017–1028. 43. Mathieu C, Adorini L 2002 The coming of age of 1,25-dihydroxyvitamin D3 analogs as immunodulatory agents. Trends Mol Med 8:174–179. 44. Nemere I, Yoshimoto Y, Norman AW 1984 Studies on the mode of action of calciferol. LIV. Calcium transport in perfused duodena from normal chicks: enhancement with 14 minutes of exposure to 1α,25-dihydroxyvitamin D3. Endocrinology 115: 1476–1483. 45. Norman AW, Nemere I, Williams G, King M 1984 1α,25Dihydroxyvitamin D3 mediates biological responses both as a steroid hormone and as a membrane-active agent. In: Cohn DV, Potts JT Jr, Fujita T (eds) Endocrine Control of Bone and Calcium Metabolism. Elsevier Science, Amsterdam. 46. Nemere I, Zhou L-X, Norman AW 1993 Nontranscriptional effects of steroid hormones. Receptor 3:277–291. 47. Nemere I, Dormanen MC, Hammond MW, Okamura WH, Norman AW 1994 Identification of a specific binding protein for 1α,25-dihydroxyvitamin D3 in basal-lateral membranes of chick intestinal epithelium and relationship to transcaltachia. J Biol Chem 269:23750–23756. 48. Kim YS, MacDonald PN, Dedhar S, Hruska KA 1996 Association of 1α,25-dihydroxyvitamin D3-occupied vitamin D receptors with cellular membrane acceptance sites. Endocrinology 137:3649–3658. 49. Norman AW, Olivera CJ, Barreto Silva FR, Bishop JE 2002 A specific binding protein/receptor for 1α,25-dihydroxy D3 is present in an intestinal caveolae membrane fraction. Biochem Biophys Res Commun 298:414–419. 50. Losel RM, Falkenstein E, Feuring M, Schultz A, Tillmann HC, Rossol-Haseroth K, Wehling M 2003 Nongenomic steroid
402
51. 52.
53.
54.
55.
56. 57. 58.
59.
60.
61.
62. 63. 64.
65. 66.
67. 68. 69.
ANTHONY W. NORMAN
action: controversies, questions, and answers. Physiol Rev 83:965–1016. 2002 Proceedings of the Second International Meeting on Rapid Responses to Steroid Hormones. Steroid 67:413–548. Cato AC, Nestl A, Mink S 2002 Rapid actions of steroid receptors in cellular signaling pathways. Science STKE 2002 http://stke.sciencemag.org/cgi/content/full/sigtrans; 2002/138/re9:RE9. Zhu Y, Rice CD, Pang Y, Pace M, Thomas P 2003 Cloning, expression, and characterization of a membrane progestin receptor and evidence it is an intermediary in meiotic maturation of fish oocytes. Proc Natl Acad Sci USA 100:2231–2236. Zhu Y, Bond J, Thomas P 2003 Identification, classification, and partical characterization of genes in humans and other vertebrates homologous to a fish membrane progestin receptor. Proc Natl Acad Sci USA 100:2237–2242. Chambliss KL, Yuhanna IS, Mineo C, Liu P, German Z, Sherman TS, Mendelsohn ME, Anderson RG, Shaul PW 2000 Estrogen receptor alpha and endothelial nitric oxide synthase are organized into a functional signaling module in caveolae. Circ Res 87:E44–E52. Mendelsohn ME 2000 Nongenomic, ER-mediated activation of endothelial nitric oxide synthase: How does it work? What does it mean? Circ Res 87:956–960. Shaul PW 2000 Novel role of estrogen receptors in vascular endothelium. Semin Perinatol 24:70–74. Wyckoff MH, Chambliss KL, Mineo C, Yuhanna IS, Mendelsohn ME, Mumby SM, Shaul PW 2001 Plasma membrane estrogen receptors are coupled to endothelial nitricoxide synthase through Gai. J Biol Chem 276:27071–27076. Seo E-G, Kato A, Norman AW 1996 Evidence for a 24R,25(OH)2-vitamin D3 receptor/binding protein in a membrane fraction isolated from a chick tibial fracture-healing callus. Biochem Biophys Res Commun 225:203–208. Nemere I, Yazzie-Atkinson D, Johns DO, Larsson D 2002 Biochemical characterization and purification of a binding protein for 24,25-dihydroxyvitamin D3 from chick intestine. J Endocrinol 172:211–219. Boyan BD, Bonewald LF, Sylvia VL, Nemere I, Larsson D, Norman AW, Rosser J, Dean DD, Schwartz Z 2002 Evidence for distinct membrane receptors for 1α,25-(OH)2D3 and 24R,25-(OH)2D3 in osteoblasts. Steroids 67:235–246. Willson TM, Moore JT 2002 Genomics verus orphan nuclear receptors: A half-time report. Mol Endocrinol 16:1135–1144. Evans RM 1988 The steroid and thyroid hormone receptor superfamily. Science 240:889–895. Mangelsdorf DJ, Thummel C, Beato M, Herrlich P, Schütz G, Umesono K, Blumberg B, Kastner P, Mark M, Chambon P, Evans RM 1995 The nuclear receptor superfamily: The second decade. Cell 83:835–839. Weatherman RV, Fletterick RJ, Scanlon TS 1999 Nuclear receptor ligands and ligand-binding domains. Annu Rev Biochem 68:559–582. Krust A, Green S, Argos P, Kumar V, Walter P, Bornert JM, Chambon P 1986 The chicken oestrogen receptor sequence: Homology with v-erbA and the human oestrogen and glucocorticoid receptors. EMBO J 5:891–897. Beato M 1989 Gene regulation by steroid hormones. Cell 56:335–344. Bourguet W, Ruff M, Chambon P, Gronemeyer H, Moras D 1995 Crystal structure of the ligand-binding domain of the human nuclear receptor RXRα. Nature 375:377–382. Rochel N, Wurtz JM, Mitschler A, Klaholz B, Moras D 2000 The crystal structure of the nuclear receptor for vitamin D bound to its natural ligand. Mol Cell 5:173–179.
70. Tocchini-Valentini G, Rochel N, Wurtz JM, Mitschler A, Moras D 2001 Crystal structures of the vitamin D receptor complexed to superagonist 20-epi ligands. Proc Natl Acad Sci USA 98:5491–5496. 71. Daiger SP, Schanfield MS, Cavalli-Sforza LL 1975 Group-specific component (Gc) proteins bind vitamin D and 25-hydroxyvitamin D. Proc Natl Acad Sci USA 72: 2076–2080. 72. Delmez JA, Tindira C, Grooms P, Dusso A, Windus DW, Slatopolsky E 1989 Parathyroid hormone suppression by intravenous 1,25-dihydroxyvitamin D. A role for increased sensitivity to calcium. J Clin Invest 83:1349–1355. 73. Bouillon R, Van Baelen H 1981 Transport of vitamin D: Significance of free and total concentrations of the vitamin D metabolites. Calcif Tissue Int 33:451–453. 74. Haddad JG 1995 Plasma vitamin D–binding protein (Gc-globulin): Multiple tasks. J Steroid Biochem Mol Biol 53:579–582. 75. Haddad JG Jr, Walgate J 1976 25-Hydroxyvitamin D transport in human plasma. Isolation and partial characterization of calcifidiol-binding protein. J Biol Chem 251:4803–4809. 76. Song YH, Naumova AK, Liebhaber SA, and Cooke NE 1999 Physical and meiotic mapping of the region of human chromosome 4q11-q13 encompassing the vitamin D binding protein DBP/Gc-globulin and albumin multigene cluster. Genome Res 9:581–587. 77. Bouillon R, Van Baelen H, De Moor P 1977 The measurement of the vitamin D-binding protein in human serum. J Clin Endocr Metab 45:225–231. 78. Gibbs PEM, Dugaiczyk A 1987 Origin of structural domains of the serum albumin gene family and a predicted structure of the gene for vitamin D binding protein. Mol Biol Evolu 4:364–379. 79. Verboven C, Rabijns A, De Maeyer M, Van Baelen H, Bouillon R, De Ranter C 2002 A structural basis for the unique binding features of the human vitamin D-binding protein. Nat Struct Biol 9:131–136. 80. Head JF, Swamy N, Ray R 2002 Crystal structure of the complex between actin and human vitamin D–binding protein at 2.5 Å resolution. Biochemistry 41:9015–9020. 81. Otterbein LR, Cosio C, Graceffa P, Dominguez R 2002 Crystal structures of the vitamin D-binding protein and its complex with actin: Structural basis of the actin-scavenger system. Proc Natl Acad Sci USA 99:8003–8008. 82. Haddad JG, Hu YZ, Kowalski MA, Laramore C, Ray K, Robzyk P, Cooke NE 1992 Identification of the sterol- and actin-binding domains of plasma vitamin D binding protein (Gc-globulin). Biochem 31:7174–7181. 83. Grishkovskaya I, Avvakumov GV, Sklenar G, Dales D, Hammond GL, Muller YA 2000 Crystal structure of human sex hormone-binding globulin: steroid transport by a laminin G-like domain. EMBO J 19:504–512. 84. Zanotti G, Calderone V, Beda M, Malpeli G, Folli C, Berni R 2001 Structure of chicken plasma retinol-binding protein. Biochim Biophys Acta 1550:64–69. 85. Janssen OE, Lahner H, Grasberger H, Spring SA, Saller B, Mann K, Refetoff S, Einspanier R 2002 Characterization and primary structures of bovine and porcine thyroxine-binding globulin. Mol Cell Endocrinol 186:27–35. 86. Norman AW 1995 Transcaltachia (the rapid hormonal stimulation of intestinal calcium transport): A component of adaptation to calcium needs and calcium availability. Am Zool 35:483–489. 87. De Boland AR, Nemere I, Norman AW 1990 Ca2+-channel agonist Bay K8644 mimics 1,25(OH)2-vitamin D3 rapid
CHAPTER 23 Function Relationships for Both the Ligand and Receptor(s)
88.
89.
90.
91.
92.
93.
94.
95. 96.
97. 98.
99.
100.
enhancement of Ca2+ transport in chick perfused duodenum. Biochem Biophys Res Commun 166:217–222. Norman AW, Okamura WH, Hammond MW, Bishop JE, Dormanen MC, Bouillon R, Van Baelen H, Ridall AL Daane E, Khoury R, Farach-Carson MC 1997 Comparison of 6-s-cis and 6-s-trans locked analogs of 1α,25(OH)2-vitamin D3 indicates that the 6-s-cis conformation is preferred for rapid nongenomic biological responses and that neither 6-s-cis nor 6-s-trans locked analogs are preferred for genomic biological responses. Mol Endocrinol 11: 1518–1531. Norman AW, Bouillon R, Farach-Carson MC, Bishop JE, Zhou L-X, Nemere I, Zhao J, Muralidharan KR, Okamura WH 1993 Demonstration that 1β,25-dihydroxyvitamin D3 is an antagonist of the nongenomic but not genomic biological responses and biological profile of the three A-ring diastereomers of 1α,25-dihydroxyvitamin D3. J Biol Chem 268: 20022–20030. Bula CM, Bishop JE, Ishizuka S, Norman AW 2000 25-Dehydro-1α-hydroxyvitamin D3–26,23S-lactone antagonizes the nuclear vitamin D receptor by mediating a unique noncovalent conformational change. Mol Endocrinol 14:1788–1796. Ishizuka S, Miura D, Eguchi H, Ozono K, Chokki M, Kamimura T, Norman AW 2000 Antagonistic action of novel 1α,25-dihydroxyvitamin D3-26,23-lactone analogs on 25-hydroxyvitamin-D3-24-hydroxylase gene expression induced by 1α,25-dihydroxy-vitamin D3 in human promyelocytic leukemia (HL-60) cells. Arch Biochem Biophys 380:92–102. Ishizuka S, Ishimoto S, Norman AW 1984 Isolation and identification of 1α,25-dihydroxy-24-oxo-vitamin D3, 1α,25-dihydroxyvitamin D3-26,23-lactone and 1α,24(S),25trihydroxyvitamin D3: in vivo metabolites of 1α,25-dihydroxyvitamin D3. Biochemistry 23:1473–1478. Ishizuka S, Norman AW 1987 Metabolic pathways of 1α,25dihydroxyvitamin D3 to 1α,25-dihydroxyvitamin D3-26,23lactone: Stereo-retained and stereo-selective lactonization. J Biol Chem 262:7165–7170. Miura D, Manabe K, Gao Q, Norman AW, Ishizuka S 1999 1α,25-Dihydroxyvitamin D3–26,23-lactone analogs antagonize differentiation of human leukemia cells (HL-60 cells) but not of human acute promyelocytic leukemia cells (NB4 cells). FEBS Lett 460:297–302. Yukihiro S, Posner GH, Guggino SE 1994 Vitamin D3 analogs stimulate calcium currents in rat osteosarcoma cells. J Biol Chem 269:23889–23893. Zanello LP, Norman AW 1997 Stimulation by 1α,25(OH)2vitamin D3 of whole cell chloride currents in osteoblastic ROS 17/2.8 cells: A structure–function study. J Biol Chem 272:22617–22622. Clark SA, Stumpf WE, Sar M, DeLuca HF, Tanaka Y 1980 Target cells for 1,25-dihydroxyvitamin D3 in the pancreas. Cell Tissue Res 209:515–520. Kadowaki S, Norman AW 1984 Studies on the mode of action of calciferol (XLIX). Dietary vitamin D is essential for normal insulin secretion from the perfused rat pancreas. J Clin Invest 73:759–766. Labriji-Mestaghanmi H, Billaudel B, Garnier PE, Malaisse WJ, Sutter BC 1988 Vitamin D and pancreatic islet function. I. Time course for changes in insulin secretion and content during vitamin D deprivation and repletion. J Endocrinol Invest 11:577–584. Tanaka Y, Seino Y, Ishida M, Yamaoka K, Satomura K, Yabuuchi H, Imura H 1986 Effect of 1,25-dihydroxyvitamin D3
101. 102.
103. 104. 105.
106.
107. 108. 109.
110. 111.
112.
113.
114. 115.
116.
117.
403
on insulin secretion: Direct or mediated. Endocrinology 118(5): 1971–1976. Beaulieu C, Kestekian R, Havrankova J, Gascon-Barré M 1993 Calcium is essential in normalizing intolerance to glucose that accompanies vitamin D depletion in vivo. Diabetes 42:35–43. Lee S, Clark SA, Gill RK, Christakos S 1994 1,25Dihydroxyvitamin D3 and pancreatic beta-cell function: Vitamin D receptors, gene expression, and insulin secretion. Endocrinology 134:1602–1610. Sergeev IN, Rhoten WB 1995 1,25-Dihydroxyvitamin D3 evokes oscillations of intracellular calcium in a pancreatic β-cell line. Endocrinology 136:2852–2861. Henquin JC 1994 Cell biology of insulin secretion. In: Kahn CR, Weir GC (eds) Joslin’s Diabetes Mellitus. Lea & Febiger, Philadelphia. Gembal M, Gilon P, Henquin JC 1992 Evidence that glucose can control insulin release independently from its action on ATP-sensitive K+ channels in mouse B cells. J Clin Invest 89:1288–1295. Gembal M, Detimary P, Gilon P, Gao ZY, Henquin JC 1993 Mechanisms by which glucose can control insulin release independently from its action on adenosine triphosphate-sensitive K+ channels in mouse B cells. J Clin Invest 91:871–880. Walters MR, Wicker DC, Riggle PC 1986 1,25Dihydroxyvitamin D3 receptors identified in the rat heart. J Mol Cell Cardiol 18:67–72. Merke J, Hofmann W, Goldschm D, Ritz E 1987 Demonstration of 1,25(OH)2 vitamin-D3 receptors and actions in vascular smooth-muscle cells in vitro. Calcif Tissue Int 41:112–114. Inoue T, Kawashima H 1988 1,25-Dihydroxyvitamin D3 stimulates 45Ca2+-uptake by cultured vascular smooth muscle cells derived from rat aorta. Biochem Biophys Res Commun 152:1388–1394. Mitsuhashi T, Morris RC Jr, Ives HE 1991 1,25Dihydroxyvitamin D3 modulates growth of vascular smooth muscle cells. J Clin Invest 87:1889–1895. Jono S, Nishizawa Y, Shioi A, Morii H 1998 1,25Dihydroxyvitamin D3 increase in vitro vascular calcification by modulating secretion of endogenous parathyroid hormone-related peptide. Circulation 98:1302–1306. Koh E, Morimoto S, Nabata T, Takamoto S, Kitano S, Ogihara T 1990 Effects of 1,25-dihydroxyvitamin D3 on the syntheses of DNA and glycosaminoglycans by rat aortic smooth muscle cells in vitro. Life Sci 46:1545–1551. Duan C, Bauchat JR, Hsieh T 2000 Phosphatidylinositol 3-kinase is required for insulin-like growth factor-I-induced vascular smooth muscle cell proliferation and migration. Circ Res 86:15–23. Yamboliev IA, Chen J, Gerthoffer WT 2001 PI 3-kinases and Src kinases regulate spreading and migration of cultured VSMCs. Am J Physiol Cell Physiol 281:C709–C718. Simoncini T, Hafezi-Moghadam A, Brazil DP, Ley K, Chin WW, Liao JK 2000 Interaction of oestrogen receptor with the regulatory subunit of phosphatidylinositol-3-OH kinase. Nature 407:538–541. Okamura WH, Midland MM, Hammond MW, Rahman NA, Dormanen MC, Nemere I, Norman AW 1994 Conformation and related topological features of vitamin D: Structure– function relationships. In: Vitamin D, A pluripotent steroid hormone: structural studies, molecular endocrinology and clinical applications. Norman AW, Bouillon R, Thomasset M (eds) Walter de Gruyter, Berlin. Corvol M, Ulmann A, and Garabedian M 1980 Specific nuclear uptake of 24,25-dihydroxycholecalciferol, a vitamin D3 metabolite biologically active in cartilage. FEBS Lett 116:273–276.
404 118. Pedrozo HA, Schwartz Z, Rimes S, Sylvia VL, Nemere I, Posner GH, Dean DD, Boyan BD 1999 Physiological importance of the 1,25(OH)2D3 membrane receptor and evidence for a membrane receptor specific for 24,25(OH)2D3. J Bone Miner Res 14:856–867. 119. Kato A, Seo EG, Einhorn TA, Bishop JE, Norman AW 1998 Studies on 24R,25-dihydroxyvitamin D3: Evidence for a non-nuclear membrane receptor in the chick tibial fracturehealing callus. Bone 23:141–146. 120. Kajikawa M, Ishida H, Fujimoto S, Mukai E, Nishimura M, Fujita J, Tsuura Y, Okamoto Y, Norman AW, Seino Y 1999 An insulinotropic effect of vitamin D analog with increasing intracellular Ca2+ concentration in pancreatic β-cells through nongenomic signal transduction. Endocrinology 140: 4706–4712. 121. Rebsamen MC, Sun J, Norman AW, Liao JK 2002 1α,25dihydroxyvitamin D3 induces vascular smooth muscle cell migration via activation of phosphatidylinositol 3-kinase. Circ Res 91:17–24. 122. Norman AW, Song XD, Bishop JE, Okamura WH, Ishizuka S 2000 1α,25(OH)2-vitamin D3 mediated rapid and genomic responses in NB4 cells: Evidence for cross-talk from rapid responses to genomic effects. In: Vitamin D Endocrine System. Structural, Biological, Genetic and Clinical Aspects. University of California, Riverside, Riverside, CA. 123. Shi H, Norman AW, Okamura WH, Sen A, Zemel MB 2002 1α,25-Dihydroxyvitamin D3 inhibits uncoupling protein 2 expression in human adipocytes. FASEB J 16:1808–1810. 124. Clark SA, Stumpf WE, Bishop CW, DeLuca HF, Park DH, Joh TH 1986 The adrenal—a new target organ of the calciotropic hormone 1,25-dihydroxyvitamin-D3. Cell Tissue Res 243:299–302. 125. Kream BE, Jose M, Yamada S, DeLuca HF 1977 A specific high-affinity binding macromolecule for 1,25-dihydroxycholecalciferol in fetal bone. Science 197:1086–1088. 126. Chen TL, Hirst MA, Feldman D 1979 A receptorlike binding macromolecule for 1α,25-dihydroxycholecalciferol in cultured mouse bone cells. J Biol Chem 254: 7491–7494. 127. Bellido T, Girasole G, Passeri G, Yu X-P, Mocharla H, Jilka RL, Notides A, Manolagas SC 1993 Demonstration of estrogen and vitamin D receptors in bone marrow-derived stromal cells: Up-regulation of the estrogen receptor by 1,25-dihydroxyvitamin-D3. Endocrinology 133:553–562. 128. Walters MR, Fischette CT, Fetzer C, May B, Riggle PC, Tibaldo-Bongiorno M, Christakos S 1992 Specific 1,25dihydroxyvitamin D3 binding sites in choroid plexus. Eur J Pharmacol 13:309–311. 129. Stumpf WE, O’Brien LP 1987 1,25(OH)2 vitamin D3 sites of action in the brain. Histochemistry 87:393–406. 130. Bidmon H-J, Stumpf WE 1994 Distribution of target cells for 1,25-dihydroxyvitamin D3 in the brain of the yellow bellied turtle Trachemys scripta. Brain Res 640:277–285. 131. Colston KW, Berger U, Wilson P, Hadcocks L, Naeem I, Earl HM, Coombes RC 1988 Mammary gland 1,25-dihydroxyvitamin D3 receptor content during pregnancy and lactation. Mol Cell Endocrinol 60:15–22. 132. Berger U, McClelland RA, Wilson P, Greene GL, Haussler MR, Pike JW, Colston K, Easton D, Coombes RC 1991 Immunocytochemical determination of estrogen receptor, progesterone receptor, and 1,25-dihydroxyvitamin D3 receptor in breast cancer and relationship to prognosis. Cancer Res 51:239–244. 133. Harper KD, Iozzo RV, Haddad JG 1989 Receptors for and bioresponses to 1,25-dihydroxyvitamin D in a human
ANTHONY W. NORMAN
134.
135. 136.
137.
138. 139.
140.
141.
142. 143. 144.
145. 146. 147.
148. 149.
150. 151.
colon carcinoma cell line (HT-29). Metabolism 38(11): 1062–1069. Reichrath J, Rafi L, Müller SM, Mink D, Reitnauer K, Tilgen W, Schmidt W, Friedrich M 1998 Immunohistochemical analysis of 1,25-dihydroxyvitamin D3 receptor in cervical carcinoma. Histochem J 30:561–567. Miyaguchi S, Watanabe T 2001 The role of vitamin D3 receptor mRNA in the proliferation of hepatocellular carcinoma. Hepato-Gastroenterology 47:468–472. Kawa S, Yoshizawa K, Tokoo M, Imai H, Oguchi H, Kiyosawa K, Homma T, Nikaido T, Furihata K 1996 Inhibitory effect of 22-oxa-1,25-dihydroxyvitamin D3 on the proliferation of pancreatic cancer cell lines. Gastroenterology 110: 1605–1613. Folgueira MA, Federico MH, Roela RA, Maistro S, Katayama ML, Brentani MM 2000 Differential regulation of vitamin D receptor expression in distinct leukemic cell lines upon phorbol ester-induced growth arrest. Braz J Med Biol Res 33:559–568. Masood R, Nagpal S, Zheng T, Cai J, Tulpule A, Smith DL, Gill PS 2001 Kaposi sarcoma is a therapeutic target for vitamin D3 receptor agonist. Blood 96:3188–3194. Klaus G, Merke J, Eing H, Hügel U, Milde P, Reichel H, Ritz E, Mehls O 1991 1,25(OH)2D3 receptor regulation and 1,25(OH)2D3 effects in primary cultures of growth cartilage cells of the rat. Calcif Tissue Int 49:340–348. Balmain N, Hauchecorne M, Pike JW, Cuisinier-Gleizes P, Mathieu H 1993 Distribution and subcellular immunolocalization of 1,25-dihydroxyvitamin D3 receptors in rat epiphyseal cartilage. Cell Mol B 39:339–350. Berry JL, Farquharson C, Whitehead CC, Mawer EB 1996 Growth plate chondrocyte vitamin D receptor number and affinity are reduced in avian tibial dyschondroplastic lesions. Bone 19:197–203. Vandewalle B, Adenis A, Hornez L, Revillion F, Lefebvre J 1994 1,25-Dihydroxyvitamin D3 receptors in normal and malignant human colorectal tissues. Cancer Lett 86:67–73. Biroc SL 1969 A possible receptor site for a metabolite of vitamin D3 in the supernatant fraction of hen egg shell gland. [Journal unknown.] Striem S, Bar A 1991 Modulation of quail intestinal and egg shell gland calbindin (Mr 28,000) gene expression by vitamin D3, 1,25-dihydroxyvitamin D3 and egg laying. Mol Cell Endocrinol 75:169–177. Stumpf WE, Clark SA, O’Brien LP, Reid FA 1988 1,25(OH)2 vitamin D3 sites of action in spinal cord and sensory ganglion. Anat Embryo 177:307–310. Walters MR 1982 1,25-Dihydroxyvitamin D receptors in rat testes, epididymis, and uterus. Fed Proc 41(4):1165. Reichrath J, Schilli M, Kerber A, Bahmer FA, Czarnetzki BM, Paus R 1994 Hair follicle expression of 1,25-dihydroxyvitamin D3 receptors during the murine hair cycle. Brit J Dermatol 131:477–482. Haussler MR, Norman AW 1969 Chromosomal receptor for a vitamin D metabolite: A nutrition classic. Nutr Rev 43:181–183. Tsai HC, Norman AW 1972 Studies on calciferol metabolism. VIII. Evidence for a cytoplasmic receptor for 1,25-dihydroxyvitamin D3 in the intestinal mucosa. J Biol Chem 248:5967–5975. Duncan WE, Glass AR, Wray HL 1991 Estrogen regulation of the nuclear 1,25-dihydroxyvitamin D3 receptor in rat liver and kidney. Endocrinology 129:2318–2324. Segura C, Alonso M, Fraga C, Caballero T, guez C, Andrez F 1999 Vitamin D receptor ontogenesis in rat liver. Histochem Cell Biol 112:163–167.
CHAPTER 23 Function Relationships for Both the Ligand and Receptor(s)
152. Drocourt L, Ourlin JC, Pascussi JM, Maurel P, Vilarem MJ 2002 Expression of CYP3A4, CYP2B6, and CYP2C9 is regulated by the vitamin D receptor pathway in primary human hepatocytes. J Biol Chem. 277:25125–25132. 153. Nguyen TM, Guillozo H, Marin L, Dufour ME, Tordet C, Pike JW, Garabedian M 1990 1,25-Dihydroxyvitamin D3 receptors in rat lung during the perinatal period: Regulation and immunohistochemical localization. Endocrinology 127:1755–1762. 154. Miller CW, Morosetti R, Campbell MJ, Mendoza S, Koeffler HP 1997 Integrity of the 1,25-dihydroxyvitamin D3 receptor in bone, lung, and other cancers. Mol Carcinog 19: 254–257. 155. O’Connell TD, Simpson RU 1996 Immunochemical identification of the 1,25-dihydroxyvitamin D3 receptor protein in human heart. Cell Biol Int 20:621–624. 156. Weishaar RE, Simpson RU 1987 Vitamin D3 and cardiovascular function in rats. J Clin Invest 79:1–7. 157. Simpson RU, Thomas GA, Arnold AJ 1985 Identification of 1,25 dihydroxyvitamin D3 receptors and activities in muscle. J Biol Chem 260:8882–8891. 158. Boland R, Norman AW, Ritz E, Hasselbach W 1992 Presence of a 1,25-dihydroxyvitamin D3 receptor in chick skeletal muscle myoblasts. Biochem Biophys Res Commun 128: 305–311. 159. Chen TL, Feldman D 1981 Regulation of 1,25-dihydroxyvitamin D3 receptors in cultured mouse bone cells. J Biol Chem 256:5561–5566. 160. Walters MR, Rosen DM, Norman AW, Luben RA 1982 1,25Dihydroxyvitamin D receptors in an established bone cell line: Correlation with biochemical responses. J Biol Chem 257:7481–7484. 161. Reddy SV, Kurihara N, Menaa C, Landucci G, Forthal D, Koop BA, Windle JJ, Roodman GD 2001 Osteoclasts formed by measles virus-infected osteoclast precursors from hCD46 transgenic mice express characteristics of pagetic osteoclasts. Endocrinology 142:2898–2905. 162. Menaa C, Barsony J, Reddy SV, Cornish J, Cundy T, Roodman GD 2002 1,25-Dihydroxyvitamin D3 hypersensitivity of osteoclast precursors from patients with Paget’s disease. J Bone Miner Res 15:228–235. 163. Dokoh S, Donaldson CA, Marion SL, Pike JW, and Haussler MR 1983 The ovary: A target organ for 1α,25dihydroxyvitamin D3. Endocrinology 112(1):200–206. 164. Ahonen MH, Zhuang YH, Aine R, Ylikomi T, Tuohimaa P 2000 Androgen receptor and vitamin D receptor in human ovarian cancer: Growth stimulation and inhibition by ligands. Int J Cancer 86:40–46. 165. Henry HL, Norman AW 1975 Localization of 1,25dihydroxy-vitamin D3 in chick parathyroid glands. Biochem Biophys Res Commun 62:781–788. 166. Wecksler WR, Henry HL, Norman AW 1977 Studies on the mode of action of calciferol XIII: Subcellular localization of 1,25-dihydroxy vitamin D3 in chicken parathyroid glands. Arch Biochem Biophys 183:168–175. 167. Wecksler WR, Ross FP, Mason RS, Posen S, Norman AW 1980 Studies on the mode of action of calciferol. XXIV. Biochemical properties of the 1α,25-dihydroxyvitamin D3 cytoplasmic receptors from human and chick parathyroid glands. Arch Biochem Biophys 201:95–103. 168. Peterfy C, Tenenhouse A 1982 Vitamin D receptors in isolated rat parotid and acinar cells. Biochim Biophys Acta 721:158–163. 169. Pike JW, Gooze LL, Haussler MR 1980 Biochemical evidence for 1,25-dihydroxyvitamin D3 receptor macromolecules
170.
171. 172.
173.
174.
175. 176.
177. 178. 179.
180. 181.
182.
183. 184. 185.
186.
405
in parathyroid, pancreatic, pituitary, and placental tissues. Life Sci 26:407–414. Stumpf WE, Sar M, Reid FA, Tanaka Y, DeLuca HF 1979 Target cells for 1,25-dihydroxyvitamin D3 in intestinal tract, stomach, kidney, skin, pituitary, and parathyroid. Science 206:1188–1190. Haussler MR, Manolagas SC, Deftos LJ 1982 Receptor for 1,25-dihydroxyvitamin D3 in GH3 pituitary cells. J Steroid Biochem Mol Biol 16:15–19. Skowronski RJ, Peehl DM, Feldman D 1993 Vitamin D and prostate cancer: 1,25 Dihydroxyvitamin D3 receptors and actions in human prostate cancer cell lines. Endocrinology 132:1952–1960. Kivineva M, Bläuer M, Syvälä H, Tammela T, Tuohimaa P 1998 Localization of 1,25-dihydroxyvitamin D3 receptor (VDR) expression in human prostate. J Steroid Biochem Mol Biol 66:121–127. Miller GJ, Stapleton GE, Ferrara JA, Lucia MS, Pfister S, Hedlund TE, Upadhya P 1992 The human prostatic carcinoma cell line LNCaP expresses biologically active, specific receptors for 1α,25-dihydroxyvitamin D3. Cancer Res 52:515–520. Simpson RU, DeLuca HF 1980 Chacterization of a receptorlike protein for 1,25-dihydroxyvitamin D3 in rat skin. Proc Natl Acad Sci USA 77(10):5822–5826. Feldman D, Chen T, Cone C, Hirst M, Shani S, Benderli A, Hochberg Z 1982 Vitamin D resistant rickets with alopecia: Cultured skin fibroblasts exhibit defective cytoplasmic receptors and unresponsiveness to 1,25(OH)2D3. J Clin Endo 55(5):1020–1022. Pillai S, Bikle DD, Elias PM 1988 1,25-Dihydroxyvitamin D production and receptor binding in human keratinocytes varies with differentiation. J Biol Chem 263:5390–5395. Segaert S, Degreef H, Bouillon R 2000 Vitamin D receptor expression is linked to cell cycle control in normal human keratinocytes. Biochem Biophys Res Commun 279:89–94. Chen CH, Sakai Y, Demay MB 2001 Targeting expression of the human vitamin D receptor to the keratinocytes of vitamin D receptor null mice prevents alopecia. Endocrinology 142: 5386–5389. Habib FK, Maddy SQ, Gelly KJ 1990 Characterisation of receptors for 1,25-dihydroxyvitamin D3 in the human testis. J Steroid Biochem Mol Biol 35(2):195–199. Osmundsen BC, Huang HFS, Anderson MB, Christakos S, Walters MR 1989 Multiple sites of action of the vitamin D endocrine system: FSH stimulation of testis 1,25-dihydroxyvitamin D3 receptors. J Steroid Biochem Mol Biol 34(1–6): 339–343. Provvedini DM, Sakagami Y, Manolagas SC 1989 Distinct target cells and effects of 1α,25-dihydroxyvitamin D3 and glucocorticoids in the rat thymus gland. Endocrinology 124:1532–1538. Stumpf WE, Downs TW 1987 Nuclear receptors for 1,25(OH)2 vitamin D3 in thymus reticular cells studied by autoradiography. Histochemistry 87:367–369. Berg JP, Torjesen PA, Haug E 1991 Calcitriol receptors in rat thyroid follicular cells (FRTL-5). Acta Endocr (Copenh) 125:574–580. Danan J, Delorme A, Cuisinier-Gleizes P 1981 Biochemical evidence for a cytoplasmic 1α,25-dihydroxyvitamin D3 receptor-like protein in rat yolk sac. J Biol Chem 256: 4847–4850. Stumpf WE, Sar M 1975 Autoradiograph techniques for localizing steroid hormones. Methods Enzymol 36: 135–140.
406 187. Shi H, Norman AW, Okamura WH, Sen A, Zemel MB 2001 1α,25-Dihydroxyvitamin D3 modulates human adipocyte metabolism via nongenomic action. FASEB J 15: 2751–2753. 188. Abe E, Miyaura C, Sakagami H, Takeda M, Konno K, Yamazaki T, Yoshiki S, Suda T 1981 Differentiation of mouse myeloid leukemia cells induced by 1α,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 78:4990–4994. 189. Mangelsdorf DJ, Koeffler HP, Donaldson CA, Pike JW, Haussler MR 1984 1,25-Dihydroxyvitamin D3-induced differentiation in a human promyelocytic leukemia cell line (HL-60): Receptor-mediated maturation to macrophage-like cells. J Cell Biol 98:391–398. 190. Park WH, Seol JG, Kim ES, Hyun JM, Jung CW, Lee CC, Binderup L, Koeffler HP, Kim BK, Lee YY 2000 Induction of apoptosis by vitamin D3 analogue EB1089 in NCI-H929 myeloma cells via activation of caspase 3 and p38 MAP kinase. Br J Haematol 109:576–583. 191. Manolagas SC, Provvedi DM, Murray EJ, Tsoukas CD, Deftos LJ 1986 The antiproliferative effect of calcitriol on human peripheral-blood mononuclear-cells. J Clin Endocr Metab 63:394–400. 192. Elstner E, Lee YY, Hashiya M, Pakkala S, Binderup L, Norman AW, Okamura WH, Koeffler HP 1994 1α,25Dihydroxy-20-epi-vitamin D3: An extraordinarily potent inhibitor of leukemic cell growth in vitro. Blood 84:1960–1967. 193. Zhou JY, Norman AW, Akashi M, Chen D-L, Uskokovic MR, Aurrecoechea JM, Dauben WG, Okamura WH, and Koeffler HP 1991 Development of a novel 1,25(OH)2vitamin D3 analog with potent ability to induce HL-60 cell differentiation without modulating calcium metabolism. Blood 78:75–82. 194. van Etten E, Branisteanu DD, Verstuyf A, Waer M, Bouillon R, Mathieu C 2000 Analogs of 1,25-dihydroxyvitamin D3 as dose-reducing agents for classical immunosuppressants. Transplantation 69:1932–1942. 195. Song X, Bishop JE, Okamura WH, Norman AW 1998 Stimulation of phosphorylation of mitogen-activated protein kinase by 1α,25-dihydroxyvitamin D3 in promyelocytic NB4 leukemia cells: A structure–function study. Endocrinology 139:457–465. 196. Norman AW, Frankel BJ, Heldt AM, Grodsky GM 1980 Vitamin D deficiency inhibits pancreatic secretion of insulin. Science 209:823–825. 197. Cade C, Norman AW 1987 Rapid normalization/stimulation by 1,25(OH)2-vitamin D3 of insulin secretion and glucose tolerance in the vitamin D–deficient rat. Endocrinology 120:1490–1497. 198. Schwartz Z, Shaked D, Hardin RR, Gruwell S, Dean DD, Sylvia VL, Boyan BD 2003 1α,25(OH)2D3 causes a rapid increase in phosphatidylinositol-specific PLC-beta activity via phospholipase A2 dependent production of lysophospholipid. Unpublished. 199. Schwartz Z, Sylvia VL, Larsson D, Nemere I, Casasola D, Dean DD, Boyan BD 2002 1α,25(OH)2D3 regulates chondrocyte matrix vesicle protein kinase C (PKC) directly via G-protein-dependent mechanisms and indirectly via incorporation of PKC during matrix vesicle biogenesis. J Biol Chem 277:11828–11837. 200. Schwartz Z, Ehland H, Sylvia VL, Larsson D, Hardin RR, Bingham V, Lopez D, Dean DD, Boyan BD 2002 1α,25Dihydroxyvitamin D3 and 24R,25-dihydroxyvitamin D3 modulate growth plate chondrocyte physiology via protein
ANTHONY W. NORMAN
201.
202.
203.
204.
205. 206.
207. 208.
209.
210.
211.
212.
213. 214.
kinase C–dependent phosphorylation of extracellular signal– regulated kinase 1/2 mitogen-activated protein kinase. Endocrinology 143:2775–2786. Schwartz Z, Sylvia VL, Del Toro F, Hardin RR, Dean DD, Boyan BD 2000 24R,25-(OH)2D3 mediates its membrane receptor-dependent effects on protein kinase C and alkaline phosphatase via phospholipase A(2) and cyclooxygenase-1 but not cyclooxygenase-2 in growth plate chondrocytes. J Cell Physiol 182:390–401. Boyan BD, Sylvia VL, Curry D, Chang Z, Dean DD, Schwartz Z 1998 Arachidonic acid is an autocoid mediator of the differential action of 1,25-(OH)2D3 and 24,25-(OH)2D3 on growth plate chondrocytes. J Cell Physiol 176:516–524. Bissonnette M, Wali RK, Hartmann SC, Niedziela SM, Roy HK, Tien X-Y, Sitrin MD, Brasitus TA 1995 1,25Dihydroxyvitamin D3 and 12-O-tetradecanoyl phorbol 13-acetate cause differential activation of Ca2+-dependent and Ca2+-independent isoforms of protein kinase C in rat colonocytes. J Clin Invest 95:2215–2221. Bettoun DJ, Buck DW, Lu J, Khalifa B, Chin WW, Nagpal S 2002 A vitamin D receptor-Ser/Thr phosphatase-p70 S6 kinase complex and modulation of its enzymatic activities by the ligand. J Biol Chem 277:24847–24850. Buitrago C, Gonzalez PV, De Boland AR 2002 Nongenomic action of 1α,25(OH)2-vitamin D3. Eur J Biochem 269: 2506–2515. Hmama Z, Nandan D, Sly L, Knutson KL, Herrera-Velit P, Reiner NE 1999 1α,25-Dihydroxyvitamin D3–induced myeloid cell differentiation is regulated by a vitamin D receptor– phosphatidylinositol 3-kinase signaling complex. J Exp Med 190:1583–1594. Barsony J, Marx SJ 1991 Rapid accumulation of cyclic GMP near activated vitamin D receptors. Proc Natl Acad Sci USA 88:1436–1440. De Boland AR, Norman AW 1990 Evidence for involvement of protein kinase C and cyclic adenosine 3′,5′-monophosphatedependent protein kinase in the 1,25-dihydroxyvitamin D3– mediated rapid stimulation of intestinal calcium transport (transcaltachia). Endocrinology 127:39–45. De Boland AR, Norman AW 1990 Influx of extracellular calcium mediates 1,25-dihydroxyvitamin D3–dependent transcaltachia (the rapid stimulation of duodenal Ca2+ transport). Endocrinology 127:2475–2480. Khare S, Tien X-Y, Wilson D, Wali RK, Bissonnette BM, Scaglione-Sewell B, Sitrin MD, Brasitus TA 1994 The role of protein kinase-Cα in the activation of particulate guanylate cyclase by 1α,25-dihydroxyvitamin D3 in CaCo-2 cells. Endocrinology 135:277–283. Bissonnette M, Tien X-Y, Niedziela SM, Hartmann SC, Frawley BP Jr, Roy HK, Sitrin MD, Perlman RL, Brasitus TA 1994 1,25(OH)2 vitamin D3 activates PKC-α in Caco-2 cells: A mechanism to limit secosteroid-induced rise in [Ca2+]i. Am J Physiol Gastrointest Liver Physiol 267: G465-G475. Yada Y, Ozeki T, Meguro S, Mori S, Nozawa Y 1989 Signal transduction in the onset of terminal keratinocyte differentiation induced by 1α,25-dihydroxyvitamin D3: Role of protein kinase C translocation. Biochem Biophys Res Commun 163:1517–1522. Simboli-Campbell M, Franks DJ, Welsh JE 1992 1,25(OH)2D3 increases membrane associated protein kinase C in MDBK cells. Cell Signal 4:99–109. Slater SJ, Kelly MB, Taddeo FJ, Larkin JD, Yeager MD, McLane JA, Ho C, Stubbs CD 1995 Direct activation of
CHAPTER 23 Function Relationships for Both the Ligand and Receptor(s)
215.
216.
217.
218.
219.
220. 221.
protein kinase C by 1α,25-dihydroxyvitamin D3. J Biol Chem 270:6639–6643. Baran DT, Sorensen AM, Honeyman RW, Ray R, Holick MF 1990 1α,25-Dihydroxyvitamin D3–induced increments in hepatocyte cytosolic calcium and lysophosphatidylinositol: Inhibition by pertussis toxin and 1β,25-dihydroxyvitamin D3. J Bone Miner Res 5:517–524. Beno DWA, Brady LM, Bissonnette M, Davis BH 1995 Protein kinase C and mitogen-activated protein kinase are required for 1,25-dihydroxyvitamin D3–stimulated Egr induction. J Biol Chem 270:3642–3647. Sellés J, Boland R 1991 Evidence on the participation of the 3′,5′-cyclic AMP pathway in the non-genomic action of 1,25-dihydroxy-vitamin D3 in cardiac muscle. Mol Cell Endocrinol 82:229–235. Zanello LP, Norman AW 1996 1α,25(OH)2 vitamin D3– mediated stimulation of outward anionic currents in osteoblast-like ROS 17/2.8 cells. Biochem Biophys Res Commun 225:551–556. Bourdeau A, Atmani F, Grosse B, Lieberherr M 1990 Rapid effects of 1,25-dihydroxyvitamin D3 and extracellular Ca2+ on phospholipid metabolism in dispersed porcine parathyroid cells. Endocrinology 127:2738–2743. Sugimoto T, Ritter C, Ried I, Morrissey J, Slatopolsky E 1988 Effect of 1,25-dihydroxyvitamin D3 on cytosolic calcium in dispersed parathyroid cells. Kidney Int 33:850–854. Bhatia M, Kirkland JB, Meckling-Gill KA 1995 Monocytic differentiation of acute promyelocytic leukemia cells in response to 1,25-dihydroxyvitamin D3 is independent of nuclear receptor binding. J Biol Chem 270:15962–15965.
407
222. Bhatia M, Kirkland JB, Meckling-Gill KA 1996 1,25Dihydroxyvitamin D3 primes acute promyelocytic cells for TPA-induced monocytic differentiation through both PKC and tyrosine phosphorylation cascades. Exp Cell Res 222:61–69. 223. Biskobing DM, Rubin J 1993 1,25-Dihydroxyvitamin D3 and phorbol myristate acetate produce divergent phenotypes in a monomyelocytic cell line. Endocrinology 132:862–866. 224. Chambliss KL, Yuhanna IS, Anderson RG, Mendelsohn ME, Shaul PW 2002 ERβ has nongenomic action in caveolae. Mol Endocrinol 16:938–946. 225. Huhtakangas JA, Olivera CJ, Bishop JE, Zanello LP, Norman AW. The vitamin D receptor is present in caveolae-enriched plasma membranes and binds 1,25(OH)2-vitamin D3 in vivo and in vitro. Mol Endocrinol. In Press. 2004. 226. Mizwicki MT, Bula CM, Keidel D, Bishop JE, Zanello LP, Wurtz JM, Moras D, Norman AW 2004 Identification of a novel alternative ligand binding pocket in the nuclear vitamin D receptor and its functional importance in 1,25(OH)2vitamin D3 cellular signaling. Proc Nat Acad Sci USA. In Press. 227. Norman AW, Mizwicki MT, Norman DPG 2004 Steroid hormone rapid actions, membrane receptors and a conformational ensemble model. Nature Reviews Drug Discovery 3:27–41. 228. Kousteni S, Bellido T, Plotkin LI, O'Brien CA, Bodenner DL, Han L, Han K, DiGregorio GB, Katzenellenbogen JA, Katzenellenbogen BS, Roberson PK, Weinstein RS, Jilka RL, Manolagas SC 2001 Nongenotropic, sex-nonspecific signaling through the estrogen or androgen receptors: dissociation from transcriptional activity. Cell 104:719–730.
CHAPTER 24
Vitamin D and the Intestinal Absorption of Calcium: A View and Overview ROBERT H. WASSERMAN I. II. III. IV. V.
Department of Biomedical Sciences, Cornell University, Ithaca, New York
Introduction The Vitamin D Hormone Overview Transcellular Calcium Absorption Vitamin D and the Paracellular Path
I. INTRODUCTION Calcium is the fifth most prominent element in the human body, constituting 0.3% of the total number of atoms therein and about 2% by weight, with 99% of this calcium present in the skeleton in the form of highly insoluble calcium phosphate crystals [1]. The calcium mineral deposited into the organic matrix of bone contributes to the rigidity and strength of bone and is a storehouse of calcium that can be removed to maintain the circulating levels of calcium within the normal range of about 2.5 mM (10 mg per dl). Besides bone formation, calcium is essential for the functioning of a multitude of physiological and biochemical processes. As shown by Sydney Ringer in the latter part of the 19th century, the calcium ion is essential for heart contractility, for the development of fertilized eggs, and for cell adhesion. In addition, other functions requiring calcium include blood coagulation, neuromuscular excitability, nerve impulse transmission, the maintenance of the permeability of cell and cellular membranes, enzyme activation, and the secretion and functioning of hormones [2]. The only way that calcium ions gain access to the body to serve these functions is by the ingestion and absorption of dietary calcium. The secosteroid vitamin D is required in most vertebrate species for the optimization and regulation of the calcium absorption process. The importance of vitamin D in human and animal biology is emphasized by the consequences of its lack. A deficiency of vitamin D predisposes the young growing individual to the disease of rickets and, in the adult, to the disease of osteomalacia. Rickets is characterized VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
VI. Vesicular Transport of Calcium VII. Commentary on Segment-Specific Intestinal Absorption of Calcium VIII. Summary References
by hypocalcemia and/or hypophosphatemia, bone deformities of the long bones, scoliosis of the spine, and enlargement of the costrochondral junctions (“rachitic rosary”) [3]. In the adult, vitamin D deficiency leads to the disease of osteomalacia, which is a failure of mineralization of osteoid produced during bone remodeling. The earlier studies by investigators such as E. Mellanby, R. Nicolaysen, E. Kodicek, A. Carlsson, and B. Lindquist established that an important site of vitamin D action was the intestinal tract and the proper absorption of calcium (see Chapter 1 for historical notes). A number of uncertainties, as one might expect, surrounded these pioneer studies. Hypotheses were offered that vitamin D directly affected calcium absorption and secondarily phosphate absorption, or that the direct effect was on phosphate absorption, with calcium secondarily affected. No special physiological and biochemical mechanisms to explain how vitamin D effects the absorption of calcium were evident at those times. In recent years, a considerable amount of research has been accomplished that has advanced our understanding of these transport processes, but a full description of the absorption processes is still lacking. This chapter describes some of the earlier as well as more current information on calcium absorption, emphasizing the role of vitamin D. The amount of material to be included in this chapter was necessarily limited, and important contributions unfortunately had to be omitted. However, the reader is referred to various reviews [4–14] for more detailed discussions and contrasting views on the role of vitamin D in specific epithelial transport processes. Copyright © 2005, Elsevier, Inc. All rights reserved.
412
ROBERT H. WASSERMAN
II. THE VITAMIN D HORMONE
16
As discussed in detail in this volume, vitamin D, either from dietary sources or by endogenous synthesis following ultraviolet irradiation of the skin, is transformed into its most active hormonal form, 1,25-dihydroxyvitamin D [1,25(OH)2D], by two sequential hydroxylation reactions [14]. The first, 25-hydroxylation, takes place in the liver, and the second, lα-hydroxylation, occurs in the kidney. The rate of renal synthesis of 1,25(OH)2D3 is feedback regulated. The metabolite, 1,25(OH)2D3, often referred to as the vitamin D hormone, with PTH and calcitonin constitute the traditional calciotropic hormones that maintain serum calcium within narrow limits through effects, direct or indirect, on the intestine, kidney, and skeleton. In addition to its homeostatic function, the vitamin D hormone also influences a number of other biological systems, as discussed elsewhere in this volume. In intestine, kidney, and other calcium-transporting organs, 1,25(OH)2D3 elicits genomic effects, inducing the synthesis of specific proteins following its interaction with the nuclear vitamin D receptor (VDRnuclear) (refer to Chapters 11–14 for details). Nongenomic effects of 1,25(OH)2D3 have also been described which are proposed to operate through activation of a membraneassociated vitamin D receptor (VDRmembrane) (refer to Chapter 23 by A.W. Norman for details).
14
III. OVERVIEW A. Relation of Calcium Intake to Calcium Absorption The duality of the mechanism of calcium absorption was clearly demonstrated in earlier studies on the relationship between dietary calcium intake (or intraluminal calcium concentrations) and the amount or rate of calcium absorption. The form of this relationship is shown in Fig. 1 [15], and a theoretical description of this relationship is depicted in Fig. 2 [16]. The curve can be resolved into two components. One component of calcium absorption is saturable and follows Michaelis– Menten-type kinetics. The other mode, diffusional absorption, is nonsaturable, with the amount of absorption a direct function of the luminal concentration of calcium. The equation describing this phenomenon, in Michaelis–Menten terms, is as follows: Jms = {(Jmax)[Ca]}/{[Ca] + Kt} + D[Ca]
EFFLUX
12
Vitamin D3 10
Calcium flux (µ moles / cm. hour)
8
6 4
Rachitic
2
0
Rachitic 2
4 INFLUX
Vitamin D3
6 0
20
40 60 80 100 Intraluminal Ca conc. (mM)
120
140
FIGURE 1
Effect of vitamin D3 and intraluminal calcium concentration on the unidirectional fluxes of calcium across chick duodenum in vivo. Efflux represents transfer of calcium from lumen to blood and influx from blood to lumen. Note that the vitamin D increased both calcium outflux and influx and increased both the saturable and nonsaturable mode of absorption (see Fig. 2). Each point represents the mean of experiments in five to six chicks. (From Wasserman and Kallfelz [15]).
calcium absorbed by a saturable process; [Ca] = calcium concentration in lumen (or the dietary calcium intake); Kt = binding constant of calcium to the ratelimiting component of the saturable process; and D = proportionality factor that relates calcium absorption to calcium intake or luminal calcium concentrations. As examples, reference is made to two reports relating calcium absorption to calcium intake in the human. Wilkinson [17] collected data from the literature on calcium balance studies of healthy individuals. Their ages varied from 19 to 83 years. The equation relating calcium intake to net calcium absorbed in the Wilkinson study, as given by Nordin and Marshall [18], is:
(1)
y = [(491x)/(287 + x)] + 0.06x − 206 (mg/day) (2)
where Jms = total calcium flux across the intestinal membranes (mucosal to serosal); Jmax = maximum flux of
where x is dietary calcium and y is net calcium absorbed. The units are mg/day. The value of 206 mg/day
CHAPTER 24 Vitamin D and the Intestinal Absorption of Calcium: A View and Overview
5 I+II
Jms =
Ca Jms (Arbitrary units)
4
(Jmax )[Ca] + D[Ca] [Ca] + Kt I+II
3
II
Jms = D[Ca] II
2
I
Jms =
1
0
1
(Jmax )[Ca] [Ca] + Kt
2 3 4 Intraluminal Ca concentration (Arbitrary units)
5
I
6
FIGURE 2 Theoretical curves for calcium flux across intestinal membranes (mucosal to serosal). Curve I (JI) represents the active component of the process and shows typical saturation kinetics. The second component, curve II (JII), represents simple diffusion and is a function of the diffusion coefficient of the system (D) and the calcium concentration in the precursor compartment (Ca). The dashed line (I + II) represents the sum of the two separate processes and is not unlike experimentally observed results (see Fig. 1). (From Wasserman and Taylor [16].)
represents the fecal excretion of calcium when calcium intake is zero. Heaney et al. [19] examined the relationship between calcium intake and calcium absorption in 180 healthy individuals. Included in the study were nuns of average age of about 41 years and a group of men and women of about 30–52 years of age. The calcium absorption (Ca Abs) equation that gave the best fit included an exponential term for the saturable component, as shown below: Ca Abs = 0.3127[exp(−1.0539CaD)]CaD + 0.1541CaD (g/day)
(3)
where CaD = grams of calcium ingested per day (g/day). The diffusion, nonsaturable factor in the Heaney study differs from that of the Wilkinson study, which might be due to differences in experimental technique and the different makeup of the study groups.
B. Relative Absorption of Calcium by Segments of the Small Intestine The degree of calcium absorption in any particular segment of the gastrointestinal tract is determined by the rate of absorption in that segment and the transit
413
time spent in that segment. The latter is related to segment length. Using the rat as experimental animal, Marcus and Lengemann [20,20a] approached this question by first determining transit rate along the small intestine with the radioactive nonabsorbable indicator yttrium-91. The transit half-time in the duodenum was 6 min for either a liquid or a solid meal. The transit half-time increased with progression aborally, maximizing in the distal small intestine at 102 min for the solid meal and 120 min for the liquid dose. With this information and data obtained after dosing rats with 45Ca-labeled liquid or solid meals, the degree of absorption in the various segments of the small intestine was estimated. With the labeled liquid dose, the duodenum, jejunum, and ileum absorbed 15%, 23%, and 62% of the ingested calcium, respectively. With labeled solid dose, the duodenum, jejunum, and ileum absorbed 8%, 4%, and 88%, respectively. In both cases, the stomach absorbed 0%. The relative sites of absorption of radiostrontium in the rat, as determined by Cramer and Copp [21], yielded similar results, with 7% of the effective absorption occurring in the duodenum, 17% in the jejunum, 65% in the ileum, and 2% in the stomach. In the dog, the relative effect sites of calcium absorption in the small intestine were 4% in duodenum, 16% in the jejunum, and 80% in the ileum [22]. More recently, Duflos et al. [23] essentially confirmed the earlier work, finding the sojourn time in the ileum was considerably longer than in the duodenum by a factor of about 55, and that most of the calcium that was absorbed was absorbed in the ileum. Thus, in these mammals, the sites where most of the ingested calcium was absorbed were the lower segments of the small intestine, and only about 10% of calcium absorption takes place in the duodenum despite its capacity to absorb calcium at a more rapid rate than any other segment. Undoubtedly a similar situation pertains to the human. This emphasizes the need to understand more thoroughly the mechanisms of calcium absorption that occurs in the mammalian ileum where most of the dietary calcium is absorbed. This is discussed more fully in Section VII, A. In avian species, the intestinal regions where most of the calcium is absorbed are the jejunum and duodenum with the lowest relative absorption taking place in the lower regions of the small intestine [24,25].
C. Relevant Structural Features of the Enterocyte The absorption of nutrients, including calcium, occurs by their movement across the epithelium cell
414 layer of the intestinal tract, which is composed of absorbing enterocytes, goblet cells, and other cell types. The surface of the absorbing enterocyte that faces the intestinal lumen is the brush border membrane, the microvillus. Attached to it exteriorly is the glycocalyx, comprised of anionic glycoproteins and mucopolysaccharides. Within the core of the microvillus are microtubules that contribute to the maintenance of its structure and integrity. These microtubules extend into the dense terminal web region located subjacent to the microvillus. Beneath the terminal web and within the cytosolic compartment are the usual intracellular organelles: mitochondria, endoplasmic reticulum, and nucleus. The basolateral membrane defines the exterior basal and lateral cell boundaries. The lamina propria is the structure beneath the enterocyte which contains blood capillaries and the lymphatic vessels that transfer absorbed nutrients parenterally. With this brief description of the intestinal epithelium, it is apparent that the transcellular absorption of calcium is a complex process, requiring the movement of calcium from the intestinal lumen through the glycocalyx region, across the brush border membrane, into and through the cell interior (with the possibility of transient sequestration by intracellular organelles), followed by the mediated extrusion from the cell across the basolateral membrane into the extracellular fluid of the lamina propria. The capillaries resident in the lamina propria transfer the newly absorbed ion into the main circulatory system for distribution of calcium throughout the body proper. The nonsaturable, diffusional absorption of calcium moves through the paracellular path that exists between adjacent cells of the epithelial cellular membrane.
IV. TRANSCELLULAR CALCIUM ABSORPTION A. Thermodynamic Parameters The physicochemical parameters that pertain to transcellular calcium transport are the intracellular ion calcium concentration of the order of 10−7 M, and an intracellular electropotential of about −50 mV with respect to the intestinal lumen. Thus, the calcium ion, usually in the millimolar range within the intestinal lumen, can enter the enterocyte by diffusion down a steep electrochemical gradient. The blood side of the epithelial cellular membrane has a potential of about +6 mV with respect to the intestinal lumen, and therefore the electropotential gradient from cell interior to lamina propria is about +56 mV. There is also a considerable calcium concentration
ROBERT H. WASSERMAN
gradient with the intracellular Ca2+ at 10−7 M and extracellular unbound Ca2+ at 1.25 × 10−3 M Ca2+. Thus, an input of energy is required to transport intracellular Ca2+ uphill against a considerable electrochemical gradient into the extracellular fluid within the lamina propria.
B. Evidence for the Active Intestinal Transport of Calcium The active intestinal calcium transport across the intestine was first shown by Schachter [26], using the in vitro everted gut sac technique. Calcium in the mucosal (lumen) side was transported into the serosal compartment uphill against a concentration gradient. The active transport process was highly dependent on the vitamin D status of the donor animal. The presence of an active Ca2+ transport system was also demonstrated in the intact animal [16]. The unidirectional Ca2+ fluxes across the intestine of the rat were measured, and knowing the electrochemical potential difference across the duodenal segment, fluxes could be subjected to analysis by the method of Ussing (see Wasserman and Taylor [16] for a more complete discussion and references). The analyses showed that calcium was absorbed by an active, energy-requiring mechanism in both the duodenum and the lower segments of the rat small intestine. This was also demonstrated by others [27,28]. A similar approach was employed with chick intestine in situ and the presence of active calcium transport by chick duodenum was demonstrated [15]. In addition, it was shown that the uphill transport of Ca2+ was enhanced by vitamin D repletion of vitamin D–deficient chicks, verifying the earlier findings of Schachter [26].
C. Vitamin D and Calcium Entry, the First Step in Transcellular Calcium Absorption 1. VITAMIN D AND BRUSH BORDER PERMEABILITY
An increase in the permeability of the microvillar (brush border) membrane of the intestine by vitamin D was demonstrated more than 40 years ago with the in vitro everted gut sac procedure [26] and with the ligated loop procedure in the intact animal [15]. An enhanced calcium uptake by isolated brush border membrane vesicles from vitamin D–adequate animals as compared to vesicles from vitamin D–deficient animals was shown by several groups; among those were Bikle et al. [29] and Fontaine et al. [30].
CHAPTER 24 Vitamin D and the Intestinal Absorption of Calcium: A View and Overview
2. MOLECULAR BASES OF CALCIUM ENTRY
a. Calcium Entry Channels1 Hoenderop and colleagues [31], using expression cloning technology, identified a protein from rabbit kidney associated with maximal Ca2+ uptake by a Xenopus oocyte test system. The protein was appropriately named epithelial calcium channel (ECaC). Sequencing of the ECaC cDNA revealed a protein comprising 730 amino acids with a molecular weight of about 83 kDa. Topolographic analysis predicted the presence of three structural domains: a hydrophobic domain composed of six transmembrane segments, and two hydrophilic domains. The C-terminal hydrophilic domain contained two potential phosphorylation sites. The N-terminal domain had ankyrin binding repeats and several phosphorylation sites. Besides the kidney, ECaC mRNA was highly expressed in rabbit duodenum, less in jejunum, but not in the ileum. The apparent affinity constant for Ca2+ uptake by ECaC is about 0.2 mM. Peng et al. [32], using a similar Xenopus oocyte expression system, screened a rat duodenal library and identified a cDNA encoding another epithelial calcium transport protein (CaT1) that was closely related to ECaC. The CaT1 cDNA encodes a protein of 727 amino acids with a predicted relative mass of 83.2 kDa, similar to the Hoenderop ECaC protein. The topology of CaT1 was also similar. CaT1 mRNA, like ECaC mRNA, was expressed in duodenum, proximal jejunum but not in ileum or distal jejunum. CaT1 mRNA, unlike ECaC mRNA, was not expressed in kidney. The apparent affinity constant for Ca2+ uptake by CaT1 is about 0.4 mM. Screening a rat kidney cortex library with a CaT1 probe, Peng et al. (33) isolated a cDNA encoding a protein with 84.2% homology with rabbit kidney ECaC. This protein, CaT2, localized specifically in the kidney and was not expressed in intestine or any other tissue examined. The Michaelis constant (Km) was 0.66 mM when expressed in Xenopus oocytes. By in situ hybridization, CaT2 (also known as ECaC) was reported to colocalize with calbindin-D28K in distal renal tubules. Similar epithelial calcium channels were identified in human intestine, with CaT1 transcripts highly expressed in duodenum and less so in stomach and jejunum [34]. No CaT1 mRNA was expressed in the 1The epithelial calcium channel identified by Hoenderop et al. [31]
is known as ECaC (original designation), CaT2, and TRPV5. The channel identified by Peng et al. [32] is known as CaT1 (original designation), ECaC2, and TRPV6. The TRP designations indicate a close relationship to the transient receptor potential (TRP) channel proteins.
415
ileum, but very low levels were seen in ascending but not descending colon. Calcium transport studies done on oocytes injected with human CaT1 cDNA showed Ca2+ uptake that was concentration dependent, saturable, and with an apparent Michaelis constant (Km) of 0.25 mM. The apparent Michaelis constant of the rat CaT1 was higher, at 0.44 mM [32]. Muller et al. [35] also obtained the full-length cDNA of human ECaC (hECaC) with properties very similar to those of rabbit ECaC. hECaC was expressed in duodenum, jejunum, colon, and rectum, but again not in ileum. Barley et al. [36] identified a cDNA from human duodenum giving a predicted amino acid sequence with a 97% homology with rat CaT1. This protein, named ECaC2 (CaT1), was present in duodenum but not ileum, colon, or kidney. ECaC2 was not correlated with 1,25(OH)2D3 levels, suggesting that the formation of ECaC2 (CaT1) was not vitamin D–dependent in the human. This appears to conflict with the observations of others who have demonstrated a vitamin D dependency of channel synthesis. Caco-2 cells are derived from a colon adenocarcinoma cell line, which differentiates in culture and at confluence assumes the characteristics of a welldifferentiated small intestine–like absorbing cellular membrane [37]. These cells have a vitamin D receptor (VDR) and, in response to 1,25(OH)2D3, there is an increase in transepithelial calcium transport and an increase in calbindin-D9K expression [38,39]. Of particular interest here is the recent study on the expression of the epithelial calcium channel, CaT1, in response to 1,25(OH)2D3 in this cell line [38]. CaT1 mRNA was present at low levels in the absence of 1,25(OH)2D3. After the addition of the hormone to the culture system, an increase in CaT1 mRNA occurred at 2 hr and further elevated with time. The rapid increase in the 1,25(OH)2D3-mediated expression of CaT1 mRNA preceded that of calbindin-D9K mRNA. b. Other Factors Possibly Involved in Calcium Entry i. Brush border membrane lipids. Rasmussen and colleagues [40] observed that 1,25(OH)2D3 altered the composition of brush border phospholipids. Using labeled precursors, the vitamin D hormone stimulated the synthesis of phosphatidylcholine (PC) with an accompanying decrease in the synthesis of phosphatidylethanolamine (PE), a response to 1,25(OH)2D3 that was independently verified [41]. The 1,25(OH)2D3 mediated lipid modification closely corresponded with an increase of Ca2+ uptake by isolated brush border membrane vesicles. There was also an increase in the number of polyunsaturated fatty acids groups in the PC fractions, which could result in an increase membrane fluidity, but neither Putkey et al. [42] using electron
416 spin resonance nor Bikle et al. [43] using fluorescence polarization were unable to observe an effect of vitamin D on the fluidity state of the brush border membranes. On the other hand, Brasitus et al. [44], using specific probes, did find an early effect of 1,25(OH)2D3 on brush border properties, but an increase in overall Ca2+ absorption was not discernible in the Brasitus study until much later, at 5 hr after 1,25(OH)2D3. These findings are consistent with other observations that an early modification of brush border membrane properties might increase Ca2+ entry, but a less rapid vitamin D–dependent reaction is required before an increase in overall calcium absorption can take place (cf. IV,C,3). ii. Alkaline phosphatase. This microvillar enzyme hydrolyzes a number of organic ortho- and pyrophosphates, and vitamin D treatment of deficient animals increases the activity of this enzyme [45–49]. Relating vitamin D–enhanced calcium transport with increases in alkaline phosphatase has not yielded a consistent and meaningful correspondence. For these reasons, the role of this brush border enzyme in vitamin D– mediated calcium absorption is yet to be clearly defined. iii. Intestinal membrane calcium-binding protein (IMCal). A particulate calcium-binding complex was identified in brush border membranes isolated from rat intestinal mucosa by Kowarski and Schachter [50,51]. The calcium-binding component of the complex, termed IMCal, was resolvable from an enzymatically active component and shown to have a molecular weight of about 200,000 by gel filtration. It was composed of subunits with a molecular weight of about 20,500, similar in size to a vitamin D–dependent protein described by Miller et al. [52]. The concentration of IMCal, as measured by its calcium binding activity, varied with factors that influence the degree of calcium absorption, including dietary calcium levels, age, intestinal distribution, and vitamin D status [53]. The proposal was put forth that IMCal functioned by mediating the transfer of calcium from intestinal lumen across the brush border membrane into the cytosolic compartment. The molecular weight data clearly differentiates IMCal from the epithelial calcium channels, CaT1 and ECaC. Unfortunately, little information on IMCal has been forthcoming since the appearance of the early reports. iv. Calmodulin. This calcium-binding protein, identified by W. Y. Cheung in 1970, activates many enzyme systems. This protein is present in relatively high concentration in the microvillar region of the small intestine [54–56]. Although there is no evidence that 1,25(OH)2D3 affects the synthesis of calmodulin, there are reports that the vitamin D hormone does
ROBERT H. WASSERMAN
cause a redistribution of calmodulin in the intestinal cell, increasing its association with the brush border complex [57,58] and specifically with a 105-kDa protein [58]. The 105-kDa protein of Bikle is most likely the 110-kDa protein-calmodulin complex studied by Mooseker and Coleman [59]. The 110-kDa protein is myosin I, an ATPase mechanoenzyme that tethers the F-actin filaments to the microvillar membrane, and it was speculated that the actin–myosin I–calmodulin complex via a Ca2+-dependent reaction might influence microvillar membrane permeability [60]. v. Protein sulfhydryl (-SH) groups. Selenite binds tightly to sulfhydryl groups of proteins, and it was shown that selenite uptake by brush border vesicles was increased by 1,25(OH)2D3 [61]. This was a rapid effect; a significant increase in selenite uptake was seen at 10 min after intravenous administration of 1,25(OH)2D3 to rachitic chicks [62]. Using a fluorescent probe, it was shown that feeding diets deficient in calcium or phosphorus to vitamin D–adequate chicks resulted in about a 2.5-fold increase in the mucosal concentration of calbindin-D28K and about a 3.1-fold increase in reactive sulfhydryl groups of isolated brush border membranes [63]. In the presence of sodium dodecyl sulfate (SDS), the number of reactive groups was increased by a factor of about 4.7, indicating that most of the available sulfhydryl groups were inaccessible to reagent and most likely buried within the lipid phase of the brush border membrane. Proteins prominently labeled by the fluorescent sulfhydryl probe had molecule masses of about 43,000 and 110,000. The sulfhydryl content of the 110,000-Da protein was positively affected by 1,25(OH)2D3. 1,25(OH)2D3-mediated increases in sulfhydryl groups occurred in intestinal and kidney brush border membranes, but not mitochondria, intestinal basolateral membranes, erythrocyte membranes, or cardiac sarcolemma, indicating some degree of specificity [64]. Other hormones also increase the number of available number of sulfhydryl groups [64], and these effects are yet to be explained. 3. CALCIUM ENTRY IS SOMETIMES LIMITING, SOMETIMES NOT
The calcium entry step can, in some situations, be the limiting step in overall calcium absorption and not in others. An example of the latter, shown in Fig. 3, comes from the study of the time course of events that take place when 1,25(OH)2D3 is administered to a vitamin D–deficient animal as reported by Fullmer [65]. Using the in situ ligated duodenal loop procedure, 47Ca placed in the gut lumen was shown to accumulate in the mucosal tissue rather rapidly, within 30 min, after the 1,25(OH)2D3 dose (Fig. 3). Only later, at 3–4 hr after hormone, was
417
CHAPTER 24 Vitamin D and the Intestinal Absorption of Calcium: A View and Overview
80
a
40 b 30 20
b
10 0
0 1
2 4 Time after 1,25(OH)2D3 (hr)
2.4
A
70
8
FIGURE 3
Intestinal absorption and tissue retention of luminally administered 47Ca2+ in vitamin D–deficient chicks at varying times following a single dose of 1,25(OH)2D3. Data are means ± SEM for six chicks. Lowercase letters represent statistically significant results: a denotes significant difference from the zero time control (p < 0.05) in 47Ca2+ tissue retention, and b denotes significant difference ( p < 0.05) in 47Ca2+ absorption above the control value. (From Fullmer et al. [82]; reprinted with permission.)
Duodenal 47Ca Absorption (% Dose)
47Calcium
(% dose)
50
Absorbed Retained
2.0
Absorption
60
1.6
50
1.2
40
0.8
CaBP
30
0.4 Trace
20
B
50
0
47Ca Absorption
2.0
40
1.6
30
1.2
20
0.8 CaBP
10
0.4
Trace
0
0 0
there an increase in overall 47Ca absorption. The 3 to 4-hr lag period coincided with the period required for the appearance of calbindin-D28K, the vitamin D–induced high-affinity calcium-binding protein, as shown in Fig. 4. (The identification, properties, and possible functions of the calbindins are more fully described in Section IV,D,1 and by Christakos, et al. in Chapter 42.) To illustrate a circumstance in which calcium entry was the limiting step, vitamin D–deficient chicks were given per os a dose of vitamin D one-third of their requirement [41]. A small amount of calbindin was synthesized, considerably less than is seen under more normal conditions and less than needed for the maximal absorption of calcium. 1,25(OH)2D3 given to these animals resulted in an increase in calcium absorption at 2 hrs whereas 3 to 4 hr is required for the synthesis of calbindin-D28K to become evident (Fig. 4), indicating in this case that the limiting step is calcium entry. The studies by the Belgian group [66] on the behavior of VDR knockout (VDR-KO) mice bears on this question, as summarized by Bouillon et al. [11]. The VDR knockout mice have characteristics expected of vitamin D–resistant individuals. It was reported that the calbindin-D9K in the duodenum of the Leuven strain of VDR-KO mice on a normal calcium diet was about the same as in the wild-type mice, but the expression of duodenal ECaC2 (CaT1) was considerably downregulated, as was the absorption of calcium. In this
47Ca
CaBP (µg/mg protein)
a
1
2 4 Time (Hrs.) after 1,25(OH)2D3
8
FIGURE 4 Effect of 1,25(OH)2D3 on duodenal absorption of 47Ca and on CaBP synthesis in rachitic chicks. (A) Absorption period, 30 min; 1,25(OH)2D3, 1 µg per chick. (B) Absorption period, 15 min; 1,25(OH)2D3, 0.3 µg per chick. The data represent the mean ± SEM for 47Ca absorption for five or six chicks. 1,25(OH)2D3 significantly increased 47Ca absorption at 4 and 8 hr (above the time 0 control; P < 0.01). (From Wasserman et al. [41].)
situation, it was proposed that calcium entry via the calcium channel(s) was the rate-limiting step in transcellular calcium transport [11].
D. Cytosolic Transfer of Calcium, the Second Step in Transcellular Calcium Absorption Two current models have been proposed to explain the vitamin D–enhanced transcytosolic movement of calcium. One is based on the proposal that the vitamin D– induced, high-affinity, calcium-binding proteins, the calbindins, serve as intracellular facilitators of calcium diffusion. The other proposes that intracellular calcium is transported in membrane-bound vesicles, the vesicular transport model. Each model could contribute to the protection of the cell against Ca2+-mediated cytotoxicity by sequestration of in-transit cytosolic Ca2+.
418 1. THE CALBINDINS AND THE CALBINDIN-BASED MODEL OF INTESTINAL CALCIUM TRANSPORT
a. The Calbindins The calbindins were first identified in chick intestine in 1966 [67] and then in rat intestine in 1967 [68]. The calbindins have been shown to be widely distributed in many species and present in many different tissues, including kidney, pancreas, and brain [8,69]. Many properties and characteristics of the calbindins have been defined, and the induction of their synthesis by the vitamin D hormone was established [10,69,70,71]. The avian form, calbindin-D28K, has a molecular weight of about 30,000 and binds four Ca2+ ions per molecule with high affinity. The mammalian form, calbindin-D9K, has a molecular weight of about 10,000 and binds two Ca2+ ions with high affinity. The high-affinity binding sites of both proteins have an average association constant (Ka) of 2–4 × 106 M−1 and therefore a dissociation constant (Kd) of 2.5–5 × 10−7 M. The binding affinities of the proteins are in a range that are appropriate for these proteins to serve as effective intracellular Ca2+ buffers. The high affinity calcium binding sites of the calbindins are in the helix–loop–helix EF-hand configuration of Kretsinger and Nockolds [72]. At this site, each calcium atom is bound to five hydrophobic amino acid residues and one water molecule. Although the larger calbindin, calbindin D28K, has six EF hand motifs, only four participate in high-affinity calcium binding. b. Calbindin Synthesis The primary stimulus for the syntheses of calbindin-D9K and calbindin-D28K is interactions of 1,25(OH)2D3 with the nuclear vitamin D receptor (VDR). However, other factors appear to be able to participate in the synthetic process. For example, in vitamin D receptor (VDR)-ablated mice, intestinal calbindin-D28K mRNA was substantially reduced but was normalized by a high local concentration of calcium, phosphorus, and/or lactose (the so-called “rescue diet”) in the intestinal lumen [73]. As another example, in the shell gland of the laying hen, there is a high Ca2+ flux across the shell gland membrane of the laying hen in order to provide the large amount of calcium required for the formation of the egg shell. Bar et al. [74] obtained results consistent with a dependency of calbindin mRNA synthesis on the rate at which calcium is transported across the egg shell gland. 1,25(OH)2D3 in the egg shell gland system is apparently required for the translation of calbindin-D28K mRNA into the protein [75]. c. The Relation of Calbindin to Calcium Absorption Direct correlations between the mucosal concentration of calbindin and the efficiency of calcium absorption under a wide variety of experimental conditions have been demonstrated, giving support to an
ROBERT H. WASSERMAN
important role of the calbindins in vitamin D–mediated calcium absorption. In the avian species, early work on the calcium-binding protein in the chick intestine revealed its vitamin D dependency and the presence of the protein in all parts of the small intestine [76]. The concentration of calbindin-D28K followed the following sequence: duodenum > jejunum > ileum. The vitamin D– dependent increase in calcium absorption followed the same sequence, i.e., duodenum > jejunum > ileum [77]. This provided in this species a reasonable site-to-site correlation between the calcium-binding protein concentrations and the magnitude of the effect of vitamin D on the calcium absorption process. Another correlation is the temporal correspondence between a 1,25(OH)2D3-mediated increase in calcium absorption and appearance of calbindin-D28K in chick duodenum of vitamin-D-deficient chicks (Fig. 4). As reviewed elsewhere [69,78], there are corresponding increases in intestinal calbindin and calcium absorption during adaptation to low-calcium or low-phosphorus diets, during growth, and in the transition from nonlaying to egg-laying in hens. Both of these entities decrease during aging [79,80]. Cortisol treatment, which inhibits calcium absorption, appears to do so by depressing calbindin synthesis [49]. There are situations, however, where noncorrelations exist. One example is the persistence of calbindin in the cytosol after vitamin D–enhanced calcium absorption has descended toward the pre–vitamin D baseline [69]. Noncorrelations of this sort can be explained in terms of the multifunctional role of 1,25(OH)2D3 on calcium absorption, indicating that under certain circumstances, vitamin D–dependent reactions or factors other than the calbindins become limiting, such as calcium entry as discussed in Section IV,C,3. 2. FUNCTIONS OF THE CALBINDINS
There are at least three, perhaps four, functions proposed to be served by the calbindins. One is participation in the control of the opened–closed status of the microvillar calcium channel. Another is the acceleration of the diffusion of calcium through the cytosol, and the third, an increase in the availability of calcium for extrusion by calcium pumps at the basolateral membrane. The fourth is a direct stimulatory effect on calcium pump activity. a. Calbindin and Activation of the Calcium Entry Process A clue that the transfer of calcium across the microvillus membrane is controlled by calcium itself comes from the report of Sampson et al. [81]. Using electron microscopic and Ca autoradiographic techniques, Sampson and colleagues showed that in vitamin D– deficient rat, luminal Ca2+ can cross the microvillar membrane, binding predominately to the inner aspect
CHAPTER 24 Vitamin D and the Intestinal Absorption of Calcium: A View and Overview
of the microvillus. After vitamin D repletion, Ca2+ moves away from this region, some sequestered by mitochondria. Confirmation and additional detail came 20 years later when the movement of calcium during absorption was visualized by the use of a stable calcium isotope (44Ca2+) and ion microscopy [82]. Like the Sampson observations, it was shown in the vitamin D– deficient chick intestine that luminal Ca can enter the intestinal cell but remains bound to elements within the microvillar region. After vitamin D–repletion, absorbed Ca2+ also appeared to interact transiently with structures in this region before moving throughout the cell as a calbindin complex. The calcium ions initially associated with the intracellular elements of the microvillar membrane might have a role in the control of calcium entry. This view is consistent with the observations of Yue et al. [83], who showed that the transport function of the epithelial calcium channel, CaT1, was inactivated by high Ca2+ concentrations and activated by a decrease in intracellular Ca2+ concentrations. Feedback control of calcium channel activity by intracellular calcium was also put forth by Hoenderop et al. [12] and Peng et al. [13]. Kinase activation by cyclic nucleotides could also control channel activity by phosphorylation of sites present in the epithelial calcium channels. Possibly related to this scheme is the control of a CaT-like (CaT-L) calcium channel of selected exocrine tissues closely related to CaT1 and ECaC as reported by Niemeyer et al. [84]. This channel has a calmodulin binding site that is also a site for protein kinase C phosphorylation. It was shown that the calcium-dependent calmodulin binding to CaT-L facilitates channel inactivation, and protein kinase C–mediated phosphorylation at the calmodulin site restores channel activity. b. Calbindins and Cytosolic Diffusion of Calcium Ninety percent of the calbindin within the enterocyte is readily released from intestinal tissue upon homogenization and therefore assumed to be primarily in a soluble state. Because of its soluble nature and its Ca2+-binding characteristics, the idea that the calbindins could serve as facilitators of Ca2+ flux through the cytosol was put forth [85]. Levine and Williams [86] and Levine et al [87] suggested that calbindin acts as an “aqueous phase ionophore,” i.e., a diffusional facilitator, increasing the flux of Ca2+ from apical to the basal region of the cell by a factor of 102. Kretsinger et al. [88] provided a theoretical basis for calbindin to act as a cytosolic diffusional facilitator. With this mechanism of calbindin action, the total amount of Ca2+ moving from the apical to this basal region would be the sum of the concentrations of free Ca2+ and calbindinbound Ca2+.
419
Feher [89] and Feher et al. [90] tested the diffusional facilitator hypothesis by examining the rate of diffusion of Ca2+ in a three-compartment system, from a precursor (lumen) compartment through the center “cytosol” compartment and then into the product (“extracellular”) compartment. The compartments were separated by semipermeable membranes. By placing calbindin in the “cytosol” middle compartment, there was a significant increase in the rate of diffusion of Ca2+ as compared to albumin, giving experimental support to the diffusion facilitator hypothesis. Other observations provide additional support for this concept of calbindin function. Pansu et al. [91] had reported that theophylline inhibits Ca2+ binding to calbindin and depresses Ca2+ absorption with no observable effect on Ca2+ uptake by isolated brush border membranes or basolateral membrane vesicles. More recently, Koster et al. [92] incorporated a soluble Ca2+ ligand, BAPTA, into isolated rabbit cortical collecting tubules and showed that the intracellular BAPTA “fully mimicked” the stimulatory effect of CaBP on transcellular Ca2+ transport. Detailed analysis of the intestinal absorption system by Bronner et al. [7] and computer modelling by Feher et al. [93] provide additional support for the functioning of calbindins as diffusional facilitators.
E. Calcium Extrusion Across the Basolateral Membrane, the Third Step in Transcellular Calcium Absorption The transfer of Ca2+ from the cell interior to the extracellular space requires energy input in order to overcome the considerable electrochemical potential difference. As depicted in Fig. 5, two systems associated with the basolateral membrane are available to extrude Ca2+ against this considerable gradient, the ATP-dependent Ca2+ pump (Ca2+-ATPase) and a Na+/Ca2+ exchanger. Also depicted is the process of exocytosis, the terminal event of the vesicular transport model. 1. THE PLASMA MEMBRANE CALCIUM PUMP
An ATP-dependent plasma membrane Ca2+ pump (PMCA) was first identified in erythrocyte membranes, and the properties of PMCA have been summarized by Carafoli [94] and Penniston and Enyedi [95]. PMCA has a molecular weight of 130–140 kDa, transports one Ca2+ per ATP, and has a Km of approximately 0.2 µM in the presence of calmodulin. Isolated basolateral vesicle membranes were initially used to demonstrate the presence of an ATP-dependent uphill transporter in intestinal membranes [96–98].
420
ROBERT H. WASSERMAN
Ca-related events the basolateral membrane Ca
Ca Channel Ca Pump
Ca
Ca Ca-Calbindin
Ca
Ca
Diffusion Ca
Ca
Free Ca
Na/Ca Exchange
Na
Ca
Diffusion
Na
Na/K Pump Vesicular
K
Ca
Transport Ca
Exocytosis
FIGURE 5 Calcium-related events at the basolateral membrane of the enterocyte. These include the uphill transport by the plasma membrane calcium pump and the Na+/K+ exchange pump. The Na+/K+ pump maintains a transmembrane gradient of Na+ for the operation of the Na+/Ca2+ exchanger. Also shown is the vesicular mode of transport. From [10].
The transport capacity of the ATP-dependent Ca2+ uptake by rat basolateral vesicles varies with maturation and aging [80,99], intestinal segment [100], and crypt-to-villus axis [101]. The stimulatory effect of vitamin D on the ATP-dependent uptake of Ca2+ by isolated basolateral membranes was shown [96,98], and kinetic analysis indicated that 1,25(OH)2D increases the Vmax of the transport process by a factor of about 3, whereas the Km is unaffected by treatment [10]. Zelinski et al. [102] were the first to show that 1,25(OH)2D3 stimulated the synthesis of Ca2+ pump mRNA. This was followed by our studies done with an antibody produced against the erythrocyte plasma membrane Ca2+ pump that cross-reacts with a Ca2+ epitope in rat and avian intestine [103,104]. Immunohistochemical analysis demonstrated clearly that the Ca2+ pump was associated with the basolateral membrane [105]. Vitamin D stimulated the synthesis of the avian Ca2+ pump [105] and, as shown by Armbrecht et al. [106], also of the calcium pump in rat intestine. Adaptation of
vitamin D–replete chicks to either a low-calcium diet or a low-phosphorus diet also increased the amount of the Ca2+ pump units associated with the basolateral membranes [107]. The increase in PMCA gene expression by 1,25(OH)2D3 is due, at least in part, to a transcriptional event [108]. 2. MODULATORS OF CALCIUM PUMP ACTIVITY
Calmodulin, the ubiquitous high-affinity calcium binding protein, is known to stimulate the activity of the PMCA [94]. Several studies have shown that calbindin-D9K and calbindin-D28K, under certain conditions, can also stimulate the activity of the basolateral Ca2+ pump (refer to Wasserman et al. [104] for references and discussion). Timmermans et al. [109] verified that calbindin-D28K can stimulate PMCA activity of rat duodenal basolateral membrane vesicles, an effect shared by another high-affinity calcium-binding protein, parvalbumin.
CHAPTER 24 Vitamin D and the Intestinal Absorption of Calcium: A View and Overview
The exact mechanism by which these high-affinity calcium binding proteins other than calmodulin affect PMCA activity is not known. One possibility is a direct interaction with the PMCA molecule, in a calmodulinlike response. Another is a means of efficiently transferring Ca2+ to the calcium binding site of PMCA, similar to the stimulatory effect of the calcium chelator, EGTA, on PMCA activity (refer to Carafoli [94] for a discussion of the EGTA effect). Like the microvillar membrane, the lipid composition of the basolateral membranes is affected by vitamin D status [110]. There is a relative increase in cholesterol content of the basolateral membranes and an increase in the arachidonic acid content of membrane phospholipids. Calcium and phosphorus deficiencies also altered lipid composition and fluidity [111]. Such lipid changes could affect the operation of enzymes, including and the plasma membrane calcium pump in particular. 3. PLASMA MEMBRANE SODIUM–CALCIUM EXCHANGER
The extrusion of Ca2+ by the Na+/Ca2+ exchanger is linked to the downhill movement of extracellular Na+ into the cell. The exchanger is rheogenic, transferring three Na+ per one Ca2+, and therefore is responsive to the cytosolic negative electropotential with respect to the external fluid phase. The Ca2+ binding affinity of the Na+/Ca2+ exchanger was reported to vary from 0.1 to 8 µM in different systems [10], in a range appropriate for the extrusion of intracellular Ca2+ across the basolateral membrane. Ghijsen et al. [112] estimated that the Na+/Ca2+ exchanger in rat intestine might account for 20% of the Ca2+ extrusion capacity of the basolateral membrane. They also reported that vitamin D did not affect exchanger activity.
V. VITAMIN D AND THE PARACELLULAR PATH Calcium is also absorbed by the nonsaturable diffusional mode by way of the paracellular path that exists between adjacent cells of the intestinal cellular membrane. The permeability of the pathway is regulated by the proteins that make up the tight junctional complex at the apical region of the adjacent cells, and also by structures present within the intercellular space [113,114]. The proteins that form these structures are synthesized by the adjacent intestinal cells and cojoin within the intercellular space to serve as part of the “glue” holding the cells together. Proteins involved in the function of the tight junction complex have been well described, and these include primarily occludin and a protein that is a member of the claudin family [115].
421
Claudin has been specifically implicated in creating charge-selective channels across the tight junctional complex [116]. Another protein, cadherin, is associated with the adherens junctions located in the intercellular space below the apical tight junctional complex. These and other paracellular-associated proteins are discussed in detail in reviews by Anderson and Van Itallie [117] and Balda and Matter [118].
A. Thermodynamic Parameters The thermodynamic parameters that pertain to the calcium flow through the paracellular path include the electropotential difference across the intestinal epithelium of about +6 mV, and the calcium concentration in the product compartment, the lamina propria, which is assumed to be about 1.25 mM, is the same as in blood plasma. Using these values, the concentration of calcium in the intestinal lumen required for there to be net diffusion of calcium via the paracellular path has been estimated to be in the range of 2–6 mM Ca2+ [5,16].
B. Vitamin D Dependency of the Diffusional Process The diffusional, nonsaturated mode of calcium absorption is considered by some to be independent of vitamin D action [119–121]. There are, however, observations that are consistent with an enhancement of paracellular calcium absorption by vitamin D. The earlier studies in which unidirectional fluxes of calcium across the avian duodenum were measured showed that vitamin D, in addition to enhancing active Ca2+ transport, also increases the nonsaturable phase of calcium absorption (Fig. 1) [15]. Further, an increase in the diffusional flow of Ca2+ from plasma to intestinal lumen by vitamin D was also observed in situ in chick ileum and rat duodenum [122], supporting the view that vitamin D has a stimulatory effect on the diffusional process. Subsequently, a number of reports appeared that provided further support of the ability of vitamin D to enhance the paracellular absorption of calcium. For instance, Hurwitz and Bar [25] reported that vitamin D increased both outflux (lumen to plasma) and influx (plasma to lumen) in the intact chick “which provides some evidence that vitamin D acts to increase the permeability of the intestine for calcium.” Karbach [123] measured the fluxes of calcium across different segments of rat intestine, using the Ussing chamber technique. Both fluxes (mucosa to
422 serosa and serosa to mucosa) were increased by 1,25(OH)2D3 and, in Karbach’s words, there was a “stimulation of paracellular calcium flux in both directions in all three intestinal segments.” Studies on rat intestine by Jungbluth and Binswanger [124] yielded a similar view that, in the duodenum and jejunum, the “overall effect of 1,25(OH)2D3 on intestinal Ca and Pi transport is to increase both cell-mediated active mucosal-to-serosal transport and paracellular diffusional serosal-to-mucosa ion movement.” The work of Chirayath et al. [125], with the Caco-2 cell line, demonstrated a stimulation of diffusional transport of Ca2+ through the paracellular pathway by 1,25(OH)2D3. The transfer of Ca2+ across the confluent cell layer was inversely related to transepithelial resistance, i.e., directly related to conductance. The bidirectional fluxes of calcium were also increased. These observations and others reported by Chirayath et al. [125] appear to establish a vitamin D–dependent paracellular route in the confluent Caco-2 cells. Consistent with this are Caco-2 cell line experiments of Fleet et al. [126,127] in which 1,25(OH)2D3 increased the bidirectional calcium fluxes. Li et al. [128], in examining events occurring in vitamin D receptor–null mice, noted a 50% reduction in intestinal calbindin-D9K without a significant change in calcium absorption. These authors surmised that the passive paracellular pathway compensated for the “reduction in the calbindin-mediated transport” and put forth the view that the paracellular calcium transport is increased by 1,25(OH)2D3 in these mice. The mechanism by which 1,25(OH)2D3 alters the paracellular permeability of intestinal epithelium is not precisely known. Stenson et al. [129] have reported that activation of protein kinase C (PKC) increases paracellular permeability, and 1,25(OH)2D3 activation of PKC has been shown in various systems [130–132]. Perez et al. [133] suggest that epithelial permeability is controllable by intracellular mediators, such as Ca2+, cAMP, and protein kinase C, probably by affecting cytoskeleton activity.
VI. VESICULAR TRANSPORT OF CALCIUM The movement of luminally derived Ca2+ through the cytosolic compartment within membrane-bound structures was proposed by Jande and Brewer in 1974 [134] and by Warner and Coleman in 1975 [135] on the basis of histological and electron probe analysis, respectively. More recently, Nemere and Norman [136] also implicated vesicular transport as a means by which Ca2+ moves through the cell, to be released from the enterocyte by exocytosis.
ROBERT H. WASSERMAN
VII. COMMENTARY ON SEGMENT-SPECIFIC INTESTINAL ABSORPTION OF CALCIUM A. Duodenum and Ileum In attempts to understand the mechanism by which vitamin D increases calcium absorption, the duodenum of various species has attracted most research attention because of its highly active transport system. From this effort, the major vitamin D–dependent components of the active process have been identified. These include the epithelial calcium channels, the calbindins, and the plasma membrane calcium pump. But it is the ileum rather than the duodenum where perhaps 70–80% of the ingested calcium is absorbed (cf. III,B). Because of this, it is important to attempt to understand the processes by which this distal segment transports calcium, and different views have been expressed as to the mechanisms of ileal calcium absorption. On one hand, Pansu et al. [120] could not detect a saturable, active component of calcium absorption in the rat ileum, or a response of the segment to 1,25(OH)2D3, or the presence in this segment of CaBP, i.e., calbindinD9K. In a recent report, Bronner et al. [137] stated that “in the ileum … all calcium is absorbed by the passive route,” a route of absorption which these investigators consider to be “vitamin D-independent.” On the other hand, in situ intestinal perfusion studies in the rat disclosed evidence of the ability of all segments of the small intestine, the duodenum, jejunum, and ileum, to actively transport calcium, using unidirectional flux data and Ussing criteria [16]. Krawitt and Schedl [27], also using an in situ perfusion system, observed net movement of calcium from lumen to plasma in the proximal, middle, and distal segments of the small intestine, indicative of active calcium transport in each of these segments. In a study of patients with chronic renal disease, Vergne-Marini et al. [138] showed that 1,25(OH)2D3 treatment increased the unidirectional calcium flux ratio in the ileum from 1.3 to 2.3, the latter exceeding the theoretical Ussing flux ratio of 1.2, thereby providing evidence of an active transport system in the human ileum. Armbrecht et al. [80] verified the presence of calbindin-D9K in rat ileum and the dependency of its synthesis on 1,25(OH)2D3. Armbrecht et al. [106] further demonstrated that the concentration of the plasma membrane calcium pump in rat ileum is elevated by 1,25(OH)2D3, the response varying with age. Thus, although these are present in lesser quantities than in the duodenum, the ileum does contain two essential components of the vitamin D–dependent transcellular active transport system, calbindin-D9K and
Transcellular
Paracellular Vitamin D deficient
A
D
Ca
Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca Ca
Ca Ca Ca Ca
−D
Ca
Ca
Ca Ca Ca Ca Ca
Ca pump
Vitamin D adequate—Duodenum
B
E Ca pump
Ca Ca
CaT1 Ca
Ca
Ca
+D
Ca
Ca Ca
Ca
CaBP Ca
Ca
CaBP Ca
Ca
Ca
Vitamin D adequate—Ileum
C
F Ca pump
Ca Ca Ca
Ca
CaBP +D
Ca
Ca
Ca
Ca Ca
CaBP
Ca
FIGURE 6 Models of vitamin D–mediated processes of intestinal absorption of calcium. Transcellular: (A) In the absence of vitamin D, calcium entry is by simple diffusion across the microvillar membrane (in duodenum, a basal level of an epithelial calcium channel might be present). Entered calcium might transiently associate with the inner aspect of the microvillar complex before diffusing to the basolateral membrane, followed by extrusion into the lamina propria by the plasma membrane calcium pump, the latter also at basal levels. (B) In the duodenum after vitamin D repletion, the rate of calcium entry is accelerated because of the synthesis of epithelial calcium channels, mediated by 1,25(OH)2D3 action. Calcium entry is controlled by the calcium concentration in the microvillar region with high [Ca] inhibiting calcium entry and low [Ca] increasing channel permeability. A means of lowering [Ca] is by complexation to 1,25(OH)2D3induced calbindins, the high-affinity calcium-binding proteins. Calbindin (mammal, calbindin-D9K; avian, calbindin-D28K) also functions as a transcytosolic diffusion facilitators, increasing the rate of transfer of bound calcium to the basolateral membrane. Calcium, dissociated from calbindin in the microvicinity of the calcium pump, is transported uphill into the extracellular fluid and at an accelerated rate due to increased synthesis of pump units mediated by 1,25(OH)2D3. (C) An epithelial calcium channel has not been identified in the ileum, and therefore calcium entry into the enterocyte is by simple diffusion. After vitamin D, the ileal enterocyte contains the 1,25(OH)2D3-induced calbindin and PMCA, already present, is augmented by 1,25(OH)2D3. As designated by the lighter shading, calbindin and PMCA are at lower concentrations compared to the duodenum, accounting for the slower rate of active transport in the ileum. The relatively long ileal transit time largely compensates for the relatively slow rate of ileal calcium absorption. Paracellular: (D) In vitamin D deficiency, calcium is absorbable by the nonsaturable diffusional process via the paracellular path if the concentration of calcium in the intestinal lumen is sufficiently high to overcome the thermodynamic barrier. (E,F) Though controversial, the evidence summarized in the text supports a 1,25(OH)2D3-mediated increase in the calcium permeability of the paracellular pathway. Modified from [10].
424 the plasma membrane calcium pump. Not identified in the ileum, however, are the recently discovered epithelial calcium channels (Section IV,C,2,a). Perhaps there is a calcium channel in the rat ileum different from CaT1 or ECaC but yet to be identified. In any event, luminal calcium can diffuse into the enterocyte by simple diffusion across the microvillar membrane as shown in the ion microscopy studies of Fullmer et al. [82]. After entry, calcium can now be transferred through and out of the cell in the same manner as in the duodenum but at a slower rate. Despite this slower rate, ileal calcium absorption by the active process could make a significant impact on overall calcium absorption because of the relatively long transit time of calcium in the ileal segment.
B. Colon The absorption of dietary calcium in the human colon is a small percentage of totally absorbed calcium, perhaps 3–6%, depending on calcium intake [139]. In the rat, the colonic contribution to total absorption was about 8% [21]. The colon has the capability of actively transporting calcium by a vitamin D–mediated mechanism [140–143].
VIII. SUMMARY Vitamin D is required for the optimal absorption of calcium in order. Physiologically, the understanding of vitamin D action began with the demonstration circa 1960 that vitamin D stimulated the saturable, energydependent absorption of calcium. Biochemically, among the first components of the calcium transport system to be identified were vitamin D–induced high-affinity Ca binding proteins, the calbindins: the avian type, calbindinD28K, and the mammalian type, calbindin-D9K. One of the proposed function of the calbindins is to facilitate the transcytosolic movement of calcium. Other components of trans-cellular calcium absorption are the basolateral plasma membrane calcium pump, the number of pump units enhanced by vitamin D, and the recently identified epithelial calcium channels that facilitate calcium entry across the apical microvillar membrane. These channels also have a vitamin D-dependency and their permeability and their activity is feedback regulated, directly or indirectly, by intracellular calcium concentrations. With these three components, a model of the operation of active transcellular calcium absorption system is presented in Fig. 6. In brief, calcium within the intestinal lumen enters the enterocyte by simple diffusion across the lipid layer of the microvillar membrane or through the epithelial calcium channels. The calcium, perhaps transiently interacting with components of the
ROBERT H. WASSERMAN
microvillus region, binds to calbindin and moves through the cytosol in this bound form, or as the free ion. At the basolateral membrane, the calcium released from calbindin is transported actively into the extracellular space by the plasma membrane calcium pump. Calcium absorbed by a nonsaturable diffusion process, moves paracellularly across the intestinal cellular layer. Although this is not universally accepted, there is considerable evidence that vitamin D does increase this diffusional mode of calcium absorption. Thus, it appears that vitamin D, through its hormonal form, affects calcium absorption both by the active transport system and by the diffusional system. In addition, calcium is absorbed in the absence of vitamin D, although at a much slower rate and to a lesser extent than in the presence of vitamin D, and most often in insufficient amounts to prevent bone disease. Which mode of absorption dominates depends on a number of different variables. These variables include vitamin D status, dietary and luminal calcium concentrations, the vigor of the active transport system in any given segment, and the transit time in that segment. By whichever mechanism, there is experimental support for the view that, in normal, vitamin D–adequate individuals, most (or all) of the ingested calcium is absorbed by vitamin D– mediated processes which include both transcellular active transport and transparacellular diffusion.
Acknowledgments The author gives special thanks to Norma Jayne for her capable secretarial assistance and to the National Institutes of Health (DK-04652) for many years of research support.
References 1. Robertson WG 1988 Chemistry and biochemistry of calcium. In: Nordin BEC (ed) Calcium in Human Biology. SpringerVerlag, London. 2. Avioli LV, Birge SJ 1978 Mechanisms of calcium absorption: A reappraisal. In: Andreoli TE, Hoffman JF, Fanestil DD (eds) Physiology of Membrane Disorders. Plenum, New York, pp. 919–940. 3. Harrison HE 1975 Vitamin D malnutrition. Prog Food Nutr Sci 1:349–357. 4. Bronner F, Pansu D 1999 Nutritional aspects of calcium absorption. J Nutr 129(1):9–12. 5. Walling MW 1982 Regulation of intestinal calcium and inorganic phosphate absorption. In: Parsons JA (ed) Endocrinology of Calcium Metabolism. Raven Press, New York, pp. 87–101. 6. Favus MJ 1985 Factors that influence absorption and secretion of calcium in the small intestine and colon. Am J Physiol 248(1):G147–G157.
CHAPTER 24 Vitamin D and the Intestinal Absorption of Calcium: A View and Overview
7. Bronner F, Pansu D, Stein WD1986 An analysis of intestinal calcium transport across the rat intestine. Am J Physiol 250(1):G561–G569. 8. Christakos S, Gabrielides C, Rhoten WB 1989 Vitamin D– dependent calcium binding proteins: chemistry, distribution, functional considerations, and molecular biology. Endocr Rev 10(1):3–26. 9. Gross M, Kumar R 1990 Physiology and biochemistry of vitamin D-dependent calcium binding proteins. Am J Physiol 259:F195–F209. 10. Wasserman RH, Fullmer CS 1995 Vitamin D and intestinal calcium transport: facts, speculations and hypotheses. J Nutr 125(7 Suppl):1971S–1979S. 11. Bouillon R, Van Cromphaut S, Carmeliet G 2003 Intestinal calcium absorption: Molecular vitamin D mediated mechanisms. J Cell Biochem 88(2):332–339. 12. Hoenderop JG, Nilius B, Bindels RJ 2002 ECaC: The gatekeeper of transepithelial Ca2+ transport. Biochim Biophys Acta 1600(1–2):6–11. 13. Peng JB, Brown EM, Hediger MA 2003 Apical entry channels in calcium-transporting epithelia. News Physiol Sci 18(4):158–163. 14. Holick MF 2003 Vitamin D: A millenium perspective. J Cell Biochem 88:296–307. 15. Wasserman RH, Kallfelz FA 1962 Vitamin D3 and the unidirectional calcium fluxes across the rachitic chick duodenum. Am J Physiol 203:221–224. 16. Wasserman RH, Taylor AN 1969 Some aspects of the intestinal absorption of calcium, with special reference to vitamin D. In: Comar CL, Bronner F (eds) Mineral Metabolism, An Advanced Treatise, Vol. III. Academic Press, New York, pp. 321–403. 17. Wilkinson R 1976 Absorption of calcium, phosphorus, and magnesium. In: Nordin BEC (ed) Calcium Phosphate and Magnesium Metabolism. Churchill Livingstone, Edinburgh. 18. Nordin BEC, Marshall DH 1988 Dietary requirements for calcium. In: Nodin BEC (ed) Calcium in Human Biology Springer-Verlag, London. 19. Heaney RP, Saville PD, Recker RR 1975 Calcium absorption as a function of calcium intake. J Lab Clin Med 85(6):881–890. 19a. Marcus, CS, & Lengemann, FW 1962 Use of radioyttrium to study food movement in the small intestine of the rat. J Nutr 76:179–182. 20. Marcus CS, Lengemann FW 1962 Absorption of Ca45 and Sr85 from solid and liquid food at various levels of the alimentary tract of the rat. J Nutr 77:155–160. 21. Cramer CF, Copp DH 1959 Progress and rate of absorption of radiostrontium through intestinal tracts of rats. Proc Soc Exp Biol Med 102:514–517. 22. Cramer CF 1965 Sites of calcium absorption and the calcium concentration of gut contents in the dog. Can J Physiol and Pharmacol 43:75–78. 23. Duflos C, Bellaton C, Pansu D, Bronner F 1995 Calcium solubility, intestinal sojourn time and paracellular permeability codetermine passive calcium absorption in rats. J Nutr 125(9):2348–2355. 24. Hurwitz S, Bar A 1966 Rate of passage of calcium-45 and yttrium-91 along the intestine, and calcium absorption in the laying fowl. J Nutr 89:311–316. 25. Hurwitz S, Bar A 1972 Site of vitamin D action in chick intestine. Am J Physiol 222(3):761–767. 26. Schachter D 1962 Vitamin D and the active transport of calcium by the small intestine. In: Wasserman RH (ed) The Transfer of Calcium and Strontium across Biological Membranes. Academic Press, New York, pp. 197–210.
425
27. Krawitt EL, Schedl HP 1968 In vivo calcium transport by rat small intestine. Am J Physiol 214(2):232–236. 28. Younoszai MK, Urban E, Schedl HP 1973 Vitamin D and intestinal calcium fluxes in vivo in the rat. Am J Physiol 225(2):287–292. 29. Bikle DD, Morrissey RL, Zolock DT 1979 The mechanism of action of vitamin D in the intestine. Am J Clin Nutr 32(11):2322–2328. 30. Fontaine O, Matsumoto T, Goodman DB, Rasmussen H 1981 Liponomic control of Ca2+ transport: relationship to mechanism of action of 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 78(3):1751–1754. 31. Hoenderop JG, van der Kemp AW, Hartog A, van de Graaf SF, van Os CH, Willems PT, Bindels RJ 1999 Molecular identification of the apical Ca2+ channel in 1,25-dihydroxyvitamin D3-responsive epithelia. J Biol Chem 274(13): 8375–8378. 32. Peng JB, Chen XZ, Berger UV, Vassilev PM, Tsukaguchi H, Brown EM, Hediger MA 1999 Molecular cloning and characterization of a channel-like transporter mediating intestinal calcium absorption. J Biol Chem 274(32): 22739–22746. 33. Peng JB, Chen XZ, Berger UV, Vassilev PM, Brown EM Hediger MA 2000 A rat kidney-specific calcium transporter in the distal nephron. J Biol Chem 275(36):28186–28194. 34. Peng JB, Chen XZ, Berger UV, Weremowicz S, Morton CC, Vassilev PM, Brown EM, Hediger MA 2000 Human calcium transport protein CaT1. Biochem Biophys Res Commun 278(2):326–332. 35. Muller D, Hoenderop JG, Meij IC, van den Heuvel LP, Knoers NV, den Hollander AI, Eggert P, Garcia-Nieto V, Claverie-Martin F, Bindels RJ 2000 Molecular cloning, tissue distribution, and chromosomal mapping of the human epithelial Ca2+ channel (ECaC1). Genomics 67(1):48–53. 36. Barley NF, Howard A, O’Callaghan D, Legon S, Walters JR 2001 Epithelial calcium transporter expression in human duodenum. Am J Physiol 280(2):G285-G290. 37. Wood RJ, Tchack L, Taparia S 2001 1,25-Dihydroxyvitamin D3 increases the expression of the CaT1 epithelial calcium channel in the Caco-2 human intestinal cell line. BMC Physiol. 1(1):11. 38. Giuliano AR, Wood RJ 1991 Vitamin D–regulated calcium transport in Caco-2 cells: unique in vitro model. Am J Physiol 260(2 Pt 1):G207–G212. 39. Fleet JC, Wood RJ 1994 Identification of calbindin D-9k mRNA and its regulation by 1,25-dihydroxyvitamin D3 in Caco-2 cells. Arch Biochem Biophys 308(1):171–174. 40. Rasmussen H, Matsumoto T, Fontaine O, Goodman DB 1982 Role of changes in membrane lipid structure in the action of 1,25-dihydroxyvitamin D3. Fed Proc 41(1):72–77. 41. Wasserman RH, Brindak ME, Meyer SA, Fullmer CS 1982 Evidence for multiple effects of vitamin D3 on calcium absorption: Response of rachitic chicks, with or without partial vitamin D3 repletion, to 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 79(24):7939–7943. 42. Putkey JA, Spielvogel AM, Sauerheber RD, Dunlap CS, Norman AW 1982 Vitamin D–mediated intestinal calcium transport. Effects of essential fatty acid deficiency and spin label studies of enterocyte membrane lipid fluidity. Biochim Biophys Acta 688(1):177–190. 43. Bikle DD, Whitney J, Munson S 1984 The relationship of membrane fluidity to calcium flux in chick intestinal brush border membranes. Endocrinology 114(1):260–267. 44. Brasitus TA, Dudeja PK, Eby B, Lau K 1986 Correction by 1,25-dihydroxycholecalciferol of the abnormal fluidity
426
45. 46. 47.
48. 49.
50. 51.
52. 53. 54. 55.
56.
57. 58.
59. 60.
61. 62.
ROBERT H. WASSERMAN
and lipid composition of enterocyte brush border membranes in vitamin D-deprived rats. J Biol Chem 261(35): 16404–16409. Pileggi VJ, DeLuca HF, Steenbock H 1955 The role of vitamin D and intestinal phytase in the prevention of rickets in rats on cereal diets. Arch Biochem Biophys 58:194–204. Holdsworth ES 1970 The effects of vitamin D on the enzyme activities in the mucosal cells of the chick small intestine. J Membrane Biol 3:43–53. Norman AW, Mircheff AK, Adams TH, Spielvogel A 1970 Studies on the mechanism of action of calciferol. 3. Vitamin D– mediated increase of intestinal brush border alkaline phosphatase activity. Biochim Biophys Acta 215(2):348–359. Haussler M, Nagode LA, Rasmussen H 1970 Induction of intestinal brush border alkaline phosphatase by vitamin D: Identity with Ca-ATPase. Nature 228(277),1199–1201. Feher JJ, Wasserman RH 1979 Intestinal calcium-binding protein and calcium absorption in cortisol-treated chicks: effects of vitamin D3 and 1,25-dihydroxyvitamin D3. Endocrinology 104(2):547–551. Kowarski S, Schachter D 1975 Vitamin D–dependent, particulate calcium-binding activity and intestinal calcium transport. Am J Physiol 229:1198–1204. Kowarski S, Schachter D 1980 Intestinal membrane calciumbinding protein. Vitamin D–dependent membrane component of the intestinal calcium transport mechanism. J Biol Chem 255(22):10834–10840. Miller A, Ueng TH, Bronner F 1979 Isolation of a vitamin D– dependent, calcium-binding protein from brush border of rat duodenal mucosa. FEBS Lett 103:319–322. Schachter D, Kowarski S 1982 Isolation of the protein IMCal, a vitamin D–dependent membrane component of the intestinal transport mechanism for calcium. Fed Proc 41:84–87. Glenney JR Jr, Glenney P 1985 Comparison of Ca++regulated events in the intestinal brush border. J Cell Biol 100(3):754–763. Howe CL, Keller TCS, III, Mooseker MS, Wasserman RH 1982 Analysis of cytoskeletal proteins and Ca2+-dependent regulation of structure in intestinal brush borders from rachitic chicks. Proc Natl Acad Sci USA 79:1134–1138. Kaune R, Munson S, Bikle DD 1994 Regulation of calmodulin binding to the ATP extractable 110 kDa protein (myosin I) from chicken duodenal brush border by 1,25-(OH)2D3. Biochim Biophys Acta 1190(2):329–336. Bikle DD, Munson S, Chafouleas J 1984 Calmodulin may mediate 1,25-dihydroxyvitamin D–stimulated intestinal calcium transport. FEBS Lett 174(1):30–33. Bikle D, Munson S, Christakos S, Kumar R, Buckendahl P 1989 Calmodulin binding to the intestinal brush-border membrane: comparison with other calcium-binding proteins. Biochim Biophys Acta 1010(1):122–127. Mooseker MS, Coleman TR 1989 The 110-kDa proteincalmodulin complex of the intestinal microvillus (brush border myosin I) is a mechanoenzyme. J Cell Biol 108:2395–2400. Mooseker MS, Wolenski JS, Coleman TR, Hayden SM, Cheney RD, Espreafico E, Heintzelman MB, Peterson MD 1991 Structural and functional dissection of a membranebound mechanoenzyme brush border myosin I. Current Top Membr 33:31–55. Mykkanen HM, Wasserman RH 1989 Uptake of 75Se-selenite by brush border membrane vesicles from chick duodenum stimulated by vitamin D. J Nutr 119(2):242–247. Mykkanen HM, Wasserman RH 1990 Relationship of membrane-bound sulfhydryl groups to vitamin D–stimulated
63.
64.
65. 66.
67. 68. 69.
70. 71. 72. 73.
74.
75.
76.
77. 78. 79.
80.
uptake of [75Se] selenite by the brush border membrane vesicles from chick duodenum. J Nutr 120(8):882–888. de Talamoni NT, Mykkanen H, Wasserman RH 1990 Enhancement of sulfhydryl group availability in the intestinal brush border membrane by deficiencies of dietary calcium and phosphorus in chicks. J Nutr 120(10):1198–1204. Tolosa de Talamoni N, Mykkanen H, Cai Q, Wasserman RH 1991 Hormonal effects on the sulfhydryl groups associated with intestinal brush border membrane proteins. Biochim Biophys Acta 1094(2):224–230. Fullmer CS 1992 Intestinal calcium absorption: calcium entry. J Nutr 122(3 Suppl):644–650. Van Cromphaut SJ, Dewerchin M, Hoenderop JG, Stockmans I, Van Herck E, Kato S, Bindels RJ, Collen D, Carmeliet P, Bouillon R, Carmeliet G 2001 Duodenal calcium absorption in vitamin D receptor-knockout mice: Functional and molecular aspects. Proc Natl Acad Sci USA 98(23):13324–13329. Wasserman RH, Taylor AN 1966 Vitamin D3–induced calcium-binding protein in chick intestinal mucosa. Science 152:791–793. Kallfelz FA, Taylor AN, Wasserman RH 1967 Vitamin D– induced calcium-binding factor in rat intestinal mucosa. Proc Soc Exptl Biol Med 125:54–58. Wasserman RH, Fullmer CS 1982 Vitamin D–induced calcium-binding protein (CaBP). In: Cheung WY (ed) Calcium and Cell Function, Vol. 2. Academic Press, New York, pp. 175–216, Wasserman RH, Fullmer CS 1983 Calcium transport proteins, calcium absorption, and vitamin D. Annu Rev Physiol 45:375–390. Feher JJ, Wasserman RH 1979 Calcium absorption and calcium-binding protein: Quantitative relationship. Am J Physiol 236(5):E556-E561. Kretsinger RH, Nockolds CE 1973 Carp muscle calciumbinding protein. II. Structure determination and general description. J Biol Chem 248(9):3313–3326. Li YC, Pirro AE, Demay MB 1998 Analysis of vitamin D– dependent calcium-binding protein messenger ribonucleic acid expression in mice lacking the vitamin D receptor. Endocrinology 139(3):847–851. Bar A, Striem S, Vax E, Talpaz H, Hurwitz S 1992 Regulation of calbindin mRNA and calbindin turnover in intestine and shell gland of the chicken. Am J Physiol 262 (5 Pt 2):R800–R805. Striem S, Bar A 1991 Modulation of quail intestinal and egg shell gland calbindin (Mr 28,000) expression by vitamin D3, 1,25-dihydroxyvitamin D3 and egg laying. Mol Cell Endocrinol 75(2):169–177. Taylor AN, Wasserman RH 1967 Vitamin D3–induced calcium-binding protein: Partial purification, electrophoretic visualization, and tissue distribution. Arch Biochem Biophys 119(1):536–540. Wasserman RH 1962 Studies on vitamin D3 and the intestinal absorption of calcium and other ions in the rachitic chick. J Nutr 77:69–80. Wasserman RH, Fullmer CS 1989 On the molecular mechanism of intestinal calcium transport. Adv Exp Med Biol 249:45–65. Armbrecht HJ, Boltz MA, Christakos S, Bruns ME 1998 Capacity of 1,25-dihydroxyvitamin D to stimulate expression of calbindin changes with age in the rat. Arch Biochem Biophys 352(2):159–164. Armbrecht HJ, Boltz M, Strong R, Richardson A, Bruns ME, Christakos S 1989 Expression of calbindin-D decreases with
CHAPTER 24 Vitamin D and the Intestinal Absorption of Calcium: A View and Overview
81.
82.
83. 84.
85. 86.
87.
88.
89. 90.
91. 92.
93. 94. 95. 96.
97.
age in intestine and kidney. Endocrinology 125(6): 2950–2956. Sampson HW, Matthews JL, Martin JH, Kunin AS 1970 An electron microscopic localization of calcium in the small intestine of normal, rachitic, and vitamin-D-treated rats. Calcif Tissue Res 5(4):305–316. Fullmer CS, Chandra S, Smith CA, Morrison GH, Wasserman RH 1996 Ion microscopic imaging of calcium during 1,25-dihydroxyvitamin D–mediated intestinal absorption. Histochem Cell Biol 106(2):215–222. Yue L, Peng JB, Hediger MA, Clapham DE 2001 CaT1 manifests the pore properties of the calcium-release-activated calcium channel. Nature 410(6829):705–709. Niemeyer BA, Bergs C, Wissenbach U, Flockerzi V, Trost C 2001 Competitive regulation of CaT-like-mediated Ca2+ entry by protein kinase C and calmodulin. Proc Natl Acad Sci USA 98(6):3600–3605. Schachter D 1969 Toward a molecular description of active transport. In: Dowben RM (ed) Biological Membranes. Little, Brown, Boston, pp. 157–176. Levine BA, Williams RJP 1982 Calcium binding to proteins and other large biological anion centers. In: Cheung WY (ed) Calcium and Cell Function, Vol. II. Academic Press, New York, pp. 2–38. Levine BA, Williams RJP, Fullmer CS, Wasserman RH 1977 NMR studies of various calcium-binding proteins. In: Wasserman RH, Corradino RA, Carafoli E, Kretsinger RH, MacLennan DH, Siegel FL (eds) Calcium Binding Proteins and Calcium Function Elsevier/North-Holland, Amsterdam, pp. 29–37. Kretsinger RH, Mann JE, Simmonds JG 1982 Model of facilitated diffusion of calcium by the intestinal calcium binding protein. In: Norman AW, Schaefer K, von Herrath D, Grigoleit H-G (eds) Vitamin D: Chemical, Biochemical and Clinical Endocrinology of Calcium Metabolism. de Gruyter, Berlin, pp. 233–248. Feher JJ 1983 Facilitated calcium diffusion by intestinal calcium-binding protein. Am J Physiol 244:303–307. Feher JJ, Fullmer CS, Fritzsch GK 1989 Comparison of the enhanced steady-state diffusion of calcium by calbindin-D9K and calmodulin: Possible importance in intestinal calcium absorption. Cell Calcium 10:189–203. Pansu D, Bellaton C, Roche C, Bronner F 1989 Theophylline inhibits transcellular Ca transport in intestine and Ca binding by CaBP. Am J Physiol 257(6 Pt 1):G935-G943. Koster HPG, Hartog A, Van Os CH, Bindels RJM 1995 Calbindin-D28K facilitates cytosolic calcium diffusion without interfering with calcium signaling. Cell Calcium 18:187–196. Feher JJ, Fullmer CS, Wasserman RH 1992 Role of facilitated diffusion of calcium by calbindin in intestinal calcium absorption. Am J Physiol 262(2 Pt.1):C517–C526. Carafoli E 1991 Calcium pump of the plasma membrane. Physiol Rev 71(1):129–153. Penniston JT, Enyedi A 1998 Modulation of the plasma membrane Ca2+ pump. J Membrane Biol 165(2):101–109. Freedman RA, Weiser MM, Isselbacher KJ 1977 Calcium translocation by Golgi and lateral-basal membrane vesicles from rat intestine: Decrease in vitamin D–deficient rats. Proc Natl Acad Sci USA 74:3612–3616. Ghijsen WEJM, De Jong MD, Van Os CH 1982 ATP-dependent calcium transport and its correlation with Ca2+-ATPase activity in basolateral plasma membranes of rat duodenum. Biochim Biophys Acta 689:327–336.
427
98. Hildmann B, Schmidt A, Murer H 1982 Ca++-transport across basal-lateral plasma membranes from rat small intestinal epithelial cells. J Membrane Biol 65:55–62. 99. Ghisan FK, Leonard D, Pietsch J 1988 Calcium transport by plasma membranes of enterocytes during development: Role of 1,25(OH)2 vitamin D3. Pediatr Res 24:338–341. 100. Van Corven EJJM, de Jong MD, van Os CH 1986 Enterocyte isolation procedure specificially effects ATP-dependent Ca2+transport in small intestinal plasma membranes. Cell Calcium 7:89–99. 101. Walters JR, Weiser MM 1987 Calcium transport by rat duodenal villus and crypt basolateral membranes. Am J Physiol 252:G170-G177. 102. Zelinski JM, Sykes DE, Weiser MM 1991 The effect of vitamin D on rat intestinal plasma membrane Ca-pump mRNA. Biochem Biophys Res Commun 179:749–755. 103. Borke JL, Caride A, Verma AK, Penniston JT, Kumar R 1990 Cellular and segmental distribution of Ca2+-pump epitopes in rat intestine. Pflügers Arch 417(1):120–122. 104. Wasserman RH, Chandler JS, Meyer SA, Smith CA, Brindak ME, Fullmer CS, Penniston JT, Kumar R 1992 Intestinal calcium transport and calcium extrusion processes at the basolateral membrane. J Nutr 122:662–671. 105. Wasserman RH, Smith CA, Brindak ME, de Talamoni N, Fullmer CS, Penniston JT, Kumar R 1991 Vitamin D and mineral deficiencies increase the plasma membrane calcium pump of chicken intestine. Gastroenterology 102:886–894. 106. Armbrecht HJ, Boltz MA, Kumar VB 1999 Intestinal plasma membrane calcium pump protein and its induction by 1,25(OH)2D3 decrease with age. Am J Physiol 277: G41–G47. 107. Cai Q, Chandler JS, Wasserman RH, Kumar R, Penniston JT 1993 Vitamin D and adaptation to dietary calcium and phosphate deficiencies increase intestinal plasma membrane calcium pump gene expression. Proc Natl Acad Sci USA 90:1345–1349. 108. Pannabecker TL, Chandler JS, Wasserman RH 1995 Vitamin D– dependent transcriptional regulation of the intestinal plasma membrane calcium pump. Biochem Biophys Res Commun 213:499–505. 109. Timmermans JA, Bindels RJ, Van Os CM 1995 Stimulation of plasma membrane Ca2+ pump by calbindin-D28k and calmodulin is additive in EGTA-free solutions. J Nutr 125 (7 Suppl):1981S–1986S. 110. Alisio A, Canas F, de Bronia DH, Pereira R, Tolosa de Talamoni N 1997 Effect of vitamin D deficiency on lipid composition and calcium transport by basolateral membrane vesicles from chick intestine. Biochem Mol Biol Int 42(2):339–347. 111. Tolosa de Talamoni NG 1996 Calcium and phosphorous deficiencies alter the lipid composition and fluidity of intestinal basolateral membranes. Comp Biochem Physiol A Physiol 115(4):309–315. 112. Ghijsen WEJM, De Jong MD, Van Os CH 1983 Kinetic properties of Na+/Ca2+ exchange in basolateral plasma membranes of rat small intestine. Biochem Biophys Acta 730:85–94. 113. Wong V, Goodenough DA 1999 Paracellular channels! Science 285:62. 114. Tang VW, Goodenough DA 2003 Paracellular ion channel at the tight junction. Biophys J 84(3):1660–1673. 115. Itoh M, Furuse M, Morita K, Kubota K, Saitou M, Tsukita S 1999 Direct binding of three tight junction-associated MAGUKs, ZO-1, ZO-2, and ZO-3, with the COOH termini of claudins. J Cell Biol 147(6):1351–1363.
428 116. Colegio OR, Van Itallie CM, McCrea HJ, Rahner C, Anderson JM 2002 Claudins create charge-selective channels in the paracellular pathway between epithelial cells. Am J Physiol 283(1):C142–C147. 117. Anderson JM, Van Itallie CM 1995 Tight junctions and the molecular basis for regulation of paracellular permeability. Am J Physiol 269(4 Pg 1):G467–G475. 118. Balda MS, Matter K 2000 Transmembrane proteins of tight junctions. Semin Cell Dev Biol 11(4):281–289. 119. Pansu D, Bellaton C, Bronner F 1981 Effect of Ca intake on saturable and nonsaturable components of duodenal Ca transport. Am J Physiol 240(1):G32–G37. 120. Pansu D, Bellaton C, Roche C, Bronner F 1983 Duodenal and ileal calcium absorption in the rat and effects of vitamin D. Am J Physiol 244(6):G695–G700. 121. Nellans HN 1990 Intestinal calcium absorption. Interplay of paracellular and cellular pathways. Miner Electrolyte Metab 16(2–3):101–108. 122. Wasserman RH, Taylor AN, Kallfelz FA 1966 Vitamin D and transfer of plasma calcium to intestinal lumen in chicks and rats. Am J Physiol 211:419–423. 123. Karbach U 1992 Paracellular calcium transport across the small intestine. J Nutr 122:672–677. 124. Jungbluth H, Binswanger U 1989 Unidirectional duodenal and jejunal calcium and phosphorus transport in the rat: effects of dietary phosphorus depletion, ethane-1-hydroxy1,1-diphosphonate and 1,25-dihydroxycholecalciferol. Res Exp Med (Berl) 189(6):439–449. 125. Chirayath MV, Gajdzik L, Hulla W, Graf J, Cross HS, Peterlik M 1998 Vitamin D increases tight-junction conductance and paracellular Ca2+ transport in Caco-2 cell cultures. Am J Physiol 274(2 Pt 1):G389–G396. 126. Fleet JC, Wood RJ 1999 Specific 1,25(OH)2D3-mediated regulation of transcellular calcium transport in Caco-2 cells. Am J Physiol 276(4 Pt 1):G958–G964. 127. Fleet JC, Eksir F, Hance KW, Wood RJ 2002 Vitamin D– inducible calcium transport and gene expression in three Caco-2 cell lines. Am J Physiol 283(3):G618–G625. 128. Li YC, Bolt MJ, Cao LP, Sitrin MD 2001 Effects of vitamin D receptor inactivation on the expression of calbindins and calcium metabolism. Am J Physiol 281(3):E558–E564. 129. Stenson WF, Easom RA, Riehl TE, Turk J 1993 Regulation of paracellular permeability in Caco-2 cell monolayers by protein kinase C. Am J Physiol 265:G955–G962. 130. Corradino RA 1974 Embryonic chick intestine in organ culture: interaction of adenylate cyclase system and
ROBERT H. WASSERMAN
131.
132.
133. 134. 135. 136.
137. 138.
139. 140. 141. 142. 143.
vitamin D3–mediated calcium absorptive mechanism. Endocrinology 94(6):1607–1614. Wali RK, Baum CL, Sitrin MD, Brasitus TA 1990 1,25(OH)2 vitamin D3 stimulates membrane phosphoinositide turnover, activates protein kinase C, and increases cytosolic calcium in rat colonic epithelium. J Clin Invest 85:1296–1303. Obeid LM, Okazaki T, Karolak LA, Hannun YA 1990 Transcriptional regulation of protein kinase C by 1,25dihydroxyvitamin D3 in HL-60 cells. J Biol Chem 265, 2370–2374. Perez M, Barber A, Ponz F 1997 Modulation of intestinal paracellular permeability by intracellular mediators and cytoskeleton. Can J Physiol Pharmacol 75(4):287–292. Jande SS, Brewer LM 1974 Effects of vitamin D3 on duodenal absorptive cells of chicks. An electron microscopic study. Z Anat Entwickl-Gesch 144, 249–265. Warner RR, Coleman JR 1975 Electron probe analysis of calcium transport by small intestine. J Cell Biol 64: 54–74. Nemere I, Norman AW 1990 Transcaltachia, vesicular calcium transport, and microtubule-associated calbindin-D28K: Emerging views of 1,25-dihiydroxyvitamin D3–mediated intestinal calcium absorption. Miner Electrol Metab 16:109–114. Bronner F, Slepchenko B, Wood RJ, Pansu D 2003 Role of passive transport in calcium absorption. J Nutr 133:1426. Vergne-Marini P, Parker TF, Pak CY, Hull AR, DeLuca HF, Fordtran JS 1976 Jejunal and ileal absorption in patients with chronic renal disease. Effect of 1-alpha-hydroxycholecalciferol. J Clin Invest 57(4):861–866. Barger-Lux MJ, Heaney RP, Recker RR 1989 Time course of calcium absorption in humans: evidence for a colonic component. Calcif Tissue Int 44(5):308–311. Favus MJ, Kathpalia SC, Coe FL, Mond AE 1980 Effects of diet calcium and 1,25-dihydroxyvitamin D3 on colon calcium active transport. Am J Physiol 238(2):G75–G78. Favus MJ, Langman CB 1984 Effects of 1,25-dihydroxyvitamin D3 on colonic calcium transport in vitamin D-deficient and normal rats. Am J Physiol 246(3 Pt 1):G268–G273. Grimstead WC, Pak CY, Krejs GJ 1984 Effect of 1,25-dihydroxyvitamin D3 on calcium absorption in the colon of healthy humans. Am J Physiol 247(2 Pt 1):G189–G192. Petith MM, Wilson HD, Schedl HP 1979 Vitamin D dependence of in vivo calcium transport and mucosal calcium binding protein in rat large intestine. Gastroenterology 76(1):99–104.
CHAPTER 25
Intestinal Calcium Absorption: Lessons from Knockout Mice and Men ROGER BOUILLON, GEERT CARMELIET, AND SOPHIE VAN CROMPHAUT Laboratory for Experimental Medicine and Endocrinology, Katholieke Universiteit Leuven, Leuven, Belgium
I. Introduction II. The Epithelial Calcium Channels: Gatekeepers for Calcium Influx III. Active Intestinal Calcium Absorption: Vitamin D–Dependent Mechanisms
IV. Active Calcium Absorption During Reproduction V. Active Calcium Absorption and Corticosteroids VI. Summary References
I. INTRODUCTION
(KO) [12,13] and 1α(OH)ase KO mice [14,15] and normalized bone structure and strength. Based on these data, it is obvious that vitamin D action on bone is redundant, whereas the small intestine plays an indispensable role in 1,25(OH)2D action on calcium homeostasis and bone mineralization. Calcium is absorbed via two different pathways in the small intestine (Fig. 1). The passive, paracellular pathway through tight junctions is more important in distal parts of the small intestine [16]. The active form of intestinal calcium absorption is energy dependent and is most developed in duodenum. 1,25(OH)2D, through its genomic actions, is regarded as one of the major stimulators of this process, which requires the combined action of at least three groups of calcium transport proteins: (1) calcium influx, which was elusive until the epithelial calcium channels ECaC1 [17–22] and ECaC2 [23–25] were described as possible gatekeepers for calcium entering the cell; (2) intracellular calcium transfer by calbindin-D9K [26–28]; and (3) calcium extrusion by the plasma membrane calcium ATPase (PMCA1b) [29–31]. Furthermore, female mice of one VDR KO strain, namely the Leuven VDR KO, proved to have an estrus cycle and to be fertile. This offered the opportunity to investigate whether estrogens or hormonal changes during the female reproductive cycle influence gene expression regarding duodenal calcium absorption, and if so, whether these effects are vitamin D–VDR dependent or independent and have functional implications. This chapter describes new insights on calcium absorption, which were obtained from the evaluation of VDR [7,9,32–36], 1α(OH)ase [15], estrogen receptor alpha (ERKOα) and beta (ERKOβ) [32], and Na/Pi cotransporter (Npt2) [37] knockout mice.
Human diseases can often be mimicked by mouse genetic studies, which have become powerful tools in unraveling physiological processes at the molecular level. Inevitably, this approach has also been applied in mineral homeostasis research. 1,25(OH)2 vitamin D [1,25(OH)2D] is considered as one of the main hormones for calcium homeostasis and bone mineralization [1]. The genomic effects underlying part of these processes are mediated by the interaction of 1,25(OH)2D with the nuclear vitamin D receptor (VDR) in a ligand-dependent manner [2]. A specific hereditary defect in the gene coding for the 25-hydroxyvitamin D-1α hydroxylase [1α(OH)ase] or CYP27B1 impairs the final, critical step in the biosynthesis of 1,25(OH)2D resulting in vitamin D–dependent rickets type I (VDDR-I), also called pseudo–vitamin D deficiency (PDDR) [3] (see Chaper 71). Mutations in the VDR are responsible for vitamin D–dependent rickets type II (VDDR-II), also called hereditary hypocalcemic vitamin D–resistant rickets (HVDRR) [4] (see Chapter 72). Both of these diseases are transmitted as autosomal recessive traits. Clinical features and response to administrated 1,25(OH)2D are distinct. The phenotypes of mice with targeted ablation of 1α(OH)ase [5,6] or VDR [7–9] exhibit the clinical abnormalities observed in the VDDR-I and -II patients, respectively, and offer the opportunity to investigate the in vivo functions of 1,25(OH)2D and the molecular basis of its actions. The bone pathology in HVDRR can be cured by frequent intravenous calcium infusions [10] and prevented by high oral doses of calcium [11]. Consistent with this standard therapy, a high-calcium diet prevented rickets and hyperparathyroidism in VDR knockout VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
Copyright © 2005, Elsevier, Inc. All rights reserved.
430
ROGER BOUILLON,
3Na1+
ET AL .
NCX1
Ca2+ 1,25(OH)2D
VDR
ECaC2 ATP CaBP9K ADP ECaC1
+
−
PMCA1b
−
+
FIGURE 1 Active intestinal calcium absorption. Mechanisms and 1,25(OH)2D-VDR dependent effects.
II. THE EPITHELIAL CALCIUM CHANNELS: GATEKEEPERS FOR CALCIUM INFLUX A. Introduction The first step in transcellular calcium (re)absorption was not clearly defined until 1999. Moreover, this step was not regarded primordial because ion microscopy images revealed that luminal calcium entered the enterocyte of vitamin D–deficient chickens [38,39]. It became generally accepted that repletion with 1,25(OH)2D induces the synthesis of calbindin D, which subsequently facilitates the diffusion of calcium in a bound form from the brush border region through the cytosol to the vicinity of the PMCA1b [39]. However, in 1999, the epithelial calcium channels ECaC1 and ECaC2 (or calcium transport protein type 1, CaT1)—in the unified nomenclature renamed transient receptor potential vanilloid 5 (TRPV5) and TRPV6 [40]—were cloned from kidney [17] and small intestine [23], respectively. They are two new members of the Transient Receptor Potential Vanilloid (TRPV) family, and they are the most calcium selective of the TRP superfamily.
B. The Transient Receptor Potential (TRP) Channel Superfamily The TRP superfamily consists of a variety of calcium permeable cation channels, which may be involved in
calcium influx. Calcium influx across the plasma membrane is important for several physiologic processes, such as apoptosis, cell proliferation, and modulation of cell cycle, but also for ion homeostasis. These non-voltage-gated cation channels vary significantly in their selectivity of activation, but share structural features (amino-terminal ankyrin repeats, six transmembrane domains, a short hydrophobic stretch between S5 and S6 representing the pore region; Fig. 2) and the permeability for cations and calcium. The TRP superfamily can be divided into six subfamilies [24, 41, 42]. The first is composed of the classical TRPs: TRPC (C for canonical). The archetypical TRP, Drosophila TRP, belongs to the TRPC family and is required for phototransduction. Some TRPCs may be store-operated channels (SOCs), which are activated by release of calcium from internal stores. The mammalian TRPC proteins also are expressed in the central nervous system or are implicated in the pheromone response. Most members of the second subfamily TRPV (V for vanilloid) function in sensory physiology. These include vanilloid receptor VR1, human vanilloid receptor-like protein VRL-1, and OSM-9, which respond to heat, osmolarity, odorants, and mechanical stimuli. The third subfamily TRPN includes proteins with many ankyrin repeats, one of which, NOMPC, participates in mechanotransduction. The fourth subfamily, TRPM, is named after a putative tumor suppressor: melastatin. PKD2 and mucolipidin are the founding members of the TRPP and TRPML. Mutations in PKD2 and mucolipidin result in polycystic kidney disease and a severe neurodegenerative disorder, respectively.
431
CHAPTER 25 Calcium Absorption and KO Models
A Outside
Inside
A A A
COOH
NH2
B
Topside
FIGURE 2 ECaC architecture. The predicted topology of the ECaC1/2 monomer (A) and architecture of the ECaC1/2 (hetero)tetrameric channel complex (B). (A) Transmembrane segments are numbered and ankyrin repeats in the N-terminal tail are indicated with A. (B) Four monomeric ECaC1/2 subunits form a tetrameric complex; as a consequence, four aspartate residues (D542) form the calcium-binding site within the pore of the channel. (With permission, Cell Calcium, 2003, 33:504 [52].)
account for the basal calcium uptake by these cells. Noteworthy is the fact that ECaC2 mRNA is much more abundant in rodent placenta than ECaC1 mRNA, as assessed by quantitative RT-PCR (unpublished personal data) and its expression is not influenced by the genotype of the mother. Data on ECaC2 expression in placenta of VDR KO pups of VDR KO mothers are presently lacking. In mice, ECaC2 is expressed more ubiquitously than in humans, whereas ECaC1 expression is restricted to the kidney. The relative mRNA levels of ECaC1 and ECaC2 in tissues coexpressing these channels are different. In duodenum and other parts of the small intestine, ECaC2 mRNA level is much higher than that of ECaC1 mRNA [9] and ECaC1 seems to be the major isoform in kidney, although both isoforms are coexpressed in these tissues. Detailed investigations of protein levels and activity measurements will provide more clarity on this issue. In the mouse and human gastrointestinal tract, ECaC2 expression is restricted essentially to the epithelial cells [45]. In the duodenum, ECaC2 and ECaC1 proteins are present in a thin layer along the apical membrane of the villus tips [18,45]. As this is the major site for calcium absorption, ECaC1/2 expression patterns are in line with the role of these proteins in dietary calcium uptake. In the kidney, ECaC1 is localized in the apical domain of the epithelial cells lining the late part of the distal convoluted tubule and the connecting tubule [18,43]. In these nephron segments ECaC1 colocalizes with other proteins involved in calcium transport, such as calbindin-D28K, sodium-calcium exchanger (NCX), and PMCA1b [18].
C. Tissue Distribution of ECaC1 and ECaC2
D. Functional Properties of ECaC1 and ECaC2
The tissue distribution of ECaC1 and ECaC2 has been studied extensively by Northern blot, RT-PCR analysis, and immunohistochemistry [19–21,23,36,43,44]. In humans, both channels are coexpressed in the organs that mediate transcellular calcium transport, such as duodenum, jejunum, colon, and kidney, but also in exocrine tissues such as pancreas, prostate, mammary gland, sweat gland, and salivary gland. Epithelial calcium channels present in the placenta might be involved in placental transfer of maternal calcium. The placental syncytiotrophoblast layer carries out this crucial process in fetal development. The fact that mRNA of ECaC1 and ECaC2 is expressed in this layer may indicate that these calcium channels
When expressed in Xenopus laevis oocytes or human embryonic kidney (HEK) 293 cells, ECaC1 and ECaC2 mediate a saturable calcium uptake and manifest distinct electrophysiological features, including hyperpolarization-dependent calcium entry and calcium-dependent inactivation [17,46–49]. ECaC1 and ECaC2-expressing HEK293 cells display large inward currents that are strongly dependent on extracellular calcium and reverse at high positive membrane potentials [49]. The current/voltage relationship shows prominent inward rectification. Thus, under unstimulated physiological conditions, when the membrane potential is typically around –70 mV in the distal part of the nephron, ECaC1 constitutes a substantial
432 calcium conductance permitting basal calcium influx. Studies have shown that ECaC1 and ECaC2 exhibit a similar ion permeation sequence for divalent cations (Ca2+ > Sr2+ ~ Ba2+ > Mn2+) [49]. The characteristic pore region of ECaC1 and ECaC2 is unique for its high calcium selectivity. A single aspartatic residue in the pore region at position 542 (D542) is crucial for calcium permeation. Mutation of D542 to alanine abolishes calcium permeation, but does not affect the permeation of monovalent cations [48]. This amino acid is completely conserved during evolution in ECaC2 [21]. From an electrophysiological point of view ECaC1 and ECaC2 are highly similar, in line with the 75% homology at the amino acid level between these channels [20]. However, three distinctive differences have been found. Calcium-dependent inactivation of ECaC2 consists of a fast initial phase followed by a slower phase of inactivation, whereas ECaC1 displays only this slow inactivation behavior [50]. A second difference is the barium selectivity that results in a significantly higher Ba2+/Ca2+ current ratio for ECaC1 than for ECaC2. Finally, a pharmacological distinction can be made between ECaC1 and ECaC2 with ruthenium red, for which ECaC2 has a 100-fold lower affinity than ECaC1.
E. Molecular Architecture of the Epithelial Calcium Channels The oligomerization of ECaC1 and ECaC2 has only recently been unraveled [51]. Cross-linking studies, co-immunoprecipitation, and molecular mass determination of ECaC1/2 using sucrose gradient sedimentation have shown that ECaC1 and ECaC2 form homo- and heterotetrameric channel complexes (Fig. 2). This tetrameric architecture of ECaC1/2 implies that four of the aspartatic residues (D452) form a negatively charged ring that functions as a selectivity filter for calcium in analogy with voltage-gated calcium channels [52]. ECaC1 and ECaC2 are co-expressed in several tissues, which allows oligomerization of these channels in vivo, which subsequently may influence the channels properties. ECaC1 and ECaC2 exhibit different channel kinetics with respect to calcium-dependent inactivation, barium selectivity, and sensitivity for inhibition by ruthenium red. Concatemers were constructed consisting of four ECaC1 and ECaC2 subunits in a head-to-tail fashion. An increased number of ECaC1 subunits in such a concatemer displayed more ECaC1-like properties, indicating that the stoichiometry of ECaC1/2 heterotetramers influences the channel
ROGER BOUILLON,
ET AL .
properties [51]. Consequently, regulation of the relative expression levels of ECaC1 and ECaC2 may be a mechanism for fine-tuning calcium transport kinetics in ECaC1/2 expressing tissues.
F. Association of the S100A10–Annexin 2 Complex Is Required for Functional Expression Using yeast two-hybrid and GST pull-down assays, the S100A10 protein was identified as the first auxiliary protein of the epithelial calcium channels. S100A10 and annexin 2 were consistently detected in ECaC1 expressing distal convoluted and connecting tubules in the kidney [53]. Furthermore, the S100A10– annexin 2 complex was present along the brush border membrane of duodenum, which is in agreement with the ECaC2 localization [53]. Actually, S100A10 forms a heterotetrameric complex with annexin 2 and associates with the conserved sequence VATTV located in the C-terminal tail of ECaC1 and ECaC2. Of these five amino acids, the first threonine plays a crucial role. The activity of ECaC1 and ECaC2 was abolished when this particular threonine was mutated, demonstrating that this motif is essential for channel function. Malfunctioning of these mutant channels was accompanied by redistribution to a sub–plasma membrane area [53]. These findings suggest a role for the S100A10-annexin 2 heterotetramer in the trafficking process of ECaC1 and ECaC2 to the plasma membrane. Whether this involves facilitated translocation of ECaC1 and ECaC2 channels to the plasma membrane or enhancement of channel retention remains to be elucidated. Moreover, down-regulating annexin 2 expression using specific small interfering RNA inhibited ECaC1 and ECaC2 mediated currents in transfected HEK293 cells. Finally, quantitative PCR analysis of kidney samples from 1α(OH)ase KO mice showed that treatment with 1,25(OH)2D significantly up-regulated the mRNA abundance of S100A10, analogous to ECaC1 expression [53].
G. Summary In conclusion, the epithelial calcium channels are primarily expressed in calcium-transporting epithelia. Structural analysis and electrophysiological studies suggest that they play an essential role in calcium entry as a part of the calcium (re)absorption, thus contributing to calcium homeostasis.
433
CHAPTER 25 Calcium Absorption and KO Models
III. ACTIVE INTESTINAL CALCIUM ABSORPTION: VITAMIN D–DEPENDENT MECHANISMS A. Vitamin D Receptor Knockout Mice 1. GENERATION OF VDR-NULL MICE
Four different groups (Tokyo, Boston, Leuven, and Munchen, respectively [7–9,36]) created VDR null mice by ablation of either exon II/first zinc finger [7,9,36] or exon III/second zinc finger [8] in the DNA binding site of this nuclear receptor. The Leuven VDR-null mice were generated using the Cre-lox system. VDR null mice of the Boston, Leuven, and Munchen strain exhibit nearly normal survival rates for up to 6 months, whereas affected homozygotes of the Tokyo strain die within 15 weeks [7]. The differing survival curves are not completely related to diet or environmental variations, but may partly be related to strain differences: the Tokyo VDR-null mice housed in the facilities of the Leuven VDR-null mice still suffer from a more important failure to thrive compared with the Leuven strain [9]. However, all VDR-null mouse strains show growth retardation starting after weaning. 2. MINERAL HOMEOSTASIS AND RICKETS OF VDR-NULL MICE
VDR-null mice are born phenotypically normal, despite expression of VDR in fetal life from embryonic day 10.5 onward in several tissues of WT mice. After weaning, VDR-null mice develop hypocalcemia and hyperparathyroidism, as evidenced by elevated parathyroid hormone (PTH) levels in serum, an increase in the size of the parathyroid glands as well as an increase in PTH mRNA levels [7–9]. Concomitantly, hypophosphatemia develops and 1,25(OH)2D serum levels become 10- to 100-fold elevated, coincident with extremely low 24,25(OH)2D serum levels. These latter data are consistent with earlier findings that the activities of 1α(OH)ase and 24-hydroxylase are regulated negatively and positively, respectively, by 1,25(OH)2D [54,55]. Long bones showed a reduced length and the typical features of rickets, including widening of epiphyseal growth plates, thinning of the cortex, and cupping and widening of the metaphysis. Amling and co-workers [13] studied histomorphometric and biomechanical parameters in Boston VDR-null mice fed regular chow. The following observations were made in 70-day-old animals: (a) 85% of bone surface covered with osteoid; (b) a marked increase in bone volume, due to increased osteoid; (c) an increase in osteoblast number and increased surface of bone covered by osteoblasts; (d) normal osteoclast number and surface
of bone covered by osteoclasts; (e) impaired mineral apposition rate; (f) reduced stiffness and increased ultimate deformation; (g) expansion of length of growth plate (+48%) with disorganization of the chondrocyte columns and increased matrix volume. The time of onset of symptoms at 3 weeks of age fits with the observation that in neonatal rats intestinal calcium absorption occurs by a 1,25(OH)2D-independent mechanism and is gradually replaced by a 1,25(OH)2D-dependent mechanism [56]. Normalization of mineral ion homeostasis (calcemia and phosphatemia) by feeding these mice a diet enriched with 2% calcium, 20% lactose, and 1.25% phosphorus (the so-called “rescue diet”) prevented hyperparathyroidism and rickets [9,12,13,33]. Normal morphology in the growth cartilage and adjacent metaphysis and normal biomechanical competence of cortical bone were demonstrated. This is consistent with the prevention and treatment of human VDRR type II with high doses of calcium [10,11] (see Chapter 72). These data suggest that, although 1,25(OH)2D has been shown to have significant effects on the expression of genes by osteoblasts and osteoclasts, the principal action of VDR in skeletal growth, maturation, and remodeling is its role in intestinal calcium absorption (Fig. 1) and/or renal calcium reabsorption. 3. CALCIUM ABSORPTION IN VDR-NULL MICE: FUNCTIONAL ASPECTS
The beneficial effect of the rescue diet aroused the question whether intestinal calcium absorption was indeed impaired in VDR-null mice. Boston VDR-null exhibit a similar calcium disappearance from the whole intestinal tract, measured 1.5, 3, and 6 hr after administration, compared with WT littermates [34]. The amount of 45Ca in the sera and femurs at these time points is three to four times lower in these VDR-null mice, consistent with the hypocalcemia. Boston VDR-null mice display a normal urinary calcium excretion, which is to be considered inappropriately high taking the hypocalcemia into account. Urinary calcium excretion of the VDR-null mice becomes even twice as high as WT levels after 1 week of a rescue diet that normalizes serum calcium. This suggests that VDR inactivation has little effect on the total intestinal calcium absorption under these experimental conditions and that the VDR-null phenotype is due to a defect in renal calcium conservation. A similar or even higher urinary calcium/creatinine ratio is also noted in Leuven [32] and Tokyo VDR-null [33] mice on a normal diet, compared to the respective WT littermates. By contrast, this ratio is reduced to one fifth in Tokyo VDR-null mice on a rescue diet since weaning,
434
ROGER BOUILLON,
compared to Tokyo WT mice on the same high-calcium diet [33]. Moreover, the setup of the study in the Boston VDR-null mice allows only a crude estimation of calcium absorption, and cannot exclude a reduction in active duodenal calcium absorption. Marcus and co-workers determined that following administration of a single oral dose of tracer calcium, 62% was absorbed in the ileum, 23% in jejunum, and only 15% in duodenum. The amount absorbed depended on the residence time in the particular segment and the rate of absorption. The latter is highest in duodenum, followed by jejunum, and is lowest in ileum [57]. Furthermore, it is generally accepted that active duodenal calcium absorption rather intervenes in maintaining mineral homeostasis when the organism is deprived of a consistent and plentiful supply of calcium. This occurs in mammals mainly after weaning. As already previously mentioned, the VDR KO phenotype only becomes apparent after weaning [7–9] on a normal diet and is rescued by administration of the high-calcium diet. Hence, defects in active calcium
A
ECaC2
absorption could definitely have important implications for mineral homeostasis. Active calcium absorption was assessed in the Leuven and Tokyo VDR-null strain by measuring appearance of 45Ca in serum at several time points between 2 and 10 min after oral gavage [9] (Fig. 3B, right panel). The obtained area under the curve (AUCpo0–10 min) is used to quantify the absorption. This AUCpo0–10 min is, compared to the respective WT mice, reduced to one-third of normal, whether measured in the Leuven VDR-null mice (on a low-calcium diet for 1 week (AUCpo0–10 min WT: 13.3 ± 0.9 ∆µmol min vs. KO: 4.4 ± 0.4 ∆µmol min)) or in Tokyo VDR-null mice [kept on a normal calcium diet (AUCpo0–10 min WT: 7.0 ± 0.4 ∆µmol min vs. KO: 2.3 ± 0.3 ∆µmol min)]. On the other hand, calcium kinetics during the first 10 min after intravenous bolus in the Leuven VDR-null animals is indistinguishable from kinetics in WT mice (AUCiv1–10 min WT: 10.4 ± 0.6 ∆µmol min vs. KO: 9.2 ± 0.4 ∆µmol min) (Fig. 3B). Hence, low serum 45Ca concentrations in the VDR KO mice are not due to an enhanced clearance from the blood compartment.
Calbindin-D9K
*
ET AL .
PMCA1b
50
*
*§
* *§
0
2
100 * 50
1
*§
*§ * * *§
0
0
WT KO
WT KO
WT KO
WT KO
WT KO
WT KO
Low Ca
Normal
Rescue
Low Ca
Normal
Rescue
∆µMol serum calcium
B
WT Low Ca diet WT Normal diet WT Rescue diet
2.5
*
2.0
*
*
* *
1.5 1.0 0.5 0.0 0
* 2
*
*
*
*
4
6
8
10
Time (minutes)
∆µMol serum calcium
Diet:
3
150
mRNA (% WT Nl diet)
100
4
*
200
Calb/Tot prot (%)
650
mRNA (% WT Nl diet)
mRNA (% WT Nl diet)
700
100 *
50
*
*§
0 WT KO
WT KO
WT KO
Low Ca
Normal
Rescue
VDR WT per os VDR KO per os VDR WT i.v. VDR KO i.v.
2.5 2.0 1.5 1.0 0.5
x
x
x x
x
8
10
0.0 0
2
4
6
Time (minutes)
FIGURE 3 Effect of VDR inactivation and dietary intervention on active intestinal calcium absorption. (A) Gene expression patterns. (B) Functional aspects: in vivo calcium absorption assay. Calbindin-D9K expression: mRNA levels: rising right, solid bars, and rising left pattern; protein levels: diagonal cross-hatch, horizontal lines, and horizontal cross-hatch pattern. *P < 0.05 vs. WT normal diet, §P < 0.05 vs. KO normal diet, xP < 0.05 vs. WT low-calcium diet.
435
CHAPTER 25 Calcium Absorption and KO Models
Interestingly, the recorded absolute values in calcium appearance and area under the curve in the Tokyo strain are 50% lower than in the Leuven strain. The older age of these animals, the normal calcium diet, and strain differences are possible explanations for these findings [9]. Song et al. assessed active duodenal calcium absorption in Tokyo VDR-null mice by means of the in situ ligated loop technique [33]. In this study, active duodenal calcium absorption is strongly age dependent and falls to a third when comparing 2- and 3-month-old WT mice. Compared to WT mice, calcium absorption in VDR KO mice fed a normal calcium diet is reduced by 71% in 2-month-old mice (WT: 55.4 ± 3.8% vs KO: 16.3 ± 1.6%) and by 87% in 3-month-old mice (WT: 18.0 ± 0.9% vs KO: 2.3 ± 1.1%) [33]. Moreover, in the 3-month-old VDR KO mice calcium absorption is equal to the calculated rate of paracellular diffusion (13.8%/hr [58]). Feeding the high-calcium diet to the WT group caused a 65% reduction in the efficiency of intestinal calcium absorption in 2- and 3-month-old WT mice (WT 2 month: 20 ± 1.7%, WT 3 month: 6.3 ± 0.5%). This rescue diet induced relatively higher levels of calcium absorption in VDR KO mice, compared to KO mice on normal diet (KO 2 month: 18.4 ± 1.5%, KO 3 month: 10.0 ± 1.4%), which is more or less comparable to levels in VDR WT animals on the high-calcium diet [33]. The authors suggest that a constraint imposed upon transcellular intestinal calcium absorption in VDR KO mice fed the normal calcium diet is removed by the rescue diet. However, they also mention that between 2 and 3 months of age VDR KO mice do not gain any body weight when they are raised on a normal calcium diet; hence it is not excluded that the extremely low rate of calcium absorption seen at 3 months is caused by a general failure to thrive. These two different studies thus indicate that active duodenal calcium absorption is clearly reduced in Leuven [9] and Tokyo VDR KO [9,33] mice to about one-third of WT levels. Administration of a rescue diet also leads to a 60% decrease in active duodenal calcium absorption in WT mice [9,33]. Tokyo VDR WT and KO mice on this diet absorb comparable amounts of calcium. One may postulate that the values recorded on the high calcium diet are indicative of a basal level of calcium absorption, such as occurs when the calcium uptake and conservation mechanisms are shut down (in the case of vitamin D endocrine system) because the organism is supplied with plenty of calcium. Song et al. [33] calculated that this level of calcium absorption is too important to be exclusively passive. Further investigation will be necessary to determine whether, even when the drive for active calcium absorption is low, part of the calcium is taken up via a transcellular
pathway, which would be constitutively present and independent from vitamin D–VDR action. 4. CALCIUM ABSORPTION IN VDR-NULL MICE ON NORMAL CALCIUM DIET: MOLECULAR ASPECTS
Subsequently, attempts were made to unravel the molecular basis of this impaired active duodenal calcium absorption. Obviously, the three first targets to investigate were (a) ECaC2 and ECaC1; (b) calbindin-D9K; and (c) PMCA1b (Fig. 1). The expression of the epithelial calcium channels [9,33,36] and PMCA1b [9] was only quantified at RNA level. Calbindin-D9K expression was assessed at the RNA and protein levels (by Western blotting and RIA) [7,9,33–35]. Previous studies showed that 1,25(OH)2D repletion of vitamin–deficient chickens or rats led to an increase in calbindin-D9K [26–28] and PMCA1b expression [16,29–31]. Intriguingly, PMCA1b expression in Leuven and Tokyo VDR-null mice is comparable to WT littermates (Fig. 3A, Table I) [9]. The defect in calbindin-D9K expression is not consistent. Different groups established decreased duodenal calbindin-D9K RNA and protein levels in Tokyo VDR-null [7,9,33] (between 50 and 30% of WT levels; Table I). Initially, the same observation was made in Boston VDR-null mice [35] with a decrease of calbindin-D9K RNA expression from 50 to 15% between 2 and 11 weeks of age (Table I). A later study did not confirm these data: expression is reduced only in young animals, but by the age of 6 or 9 weeks the expression levels are comparable to those of WT littermates (Table I) [34]. Previously, it was already shown in Leuven VDR-null mice that the normal calbindin-D9K protein [with a slight decrease in calbindin-D9K RNA level (62%)] content could not account for the HVDRR phenotype of these mice (Fig. 3A, Table I) [9]. In contrast, duodenal ECaC2, and to a lesser degree ECaC1, mRNA levels are markedly reduced in the Leuven and the Tokyo VDR-null mice to less than 10% [9] and 5% [9,33] of their WT littermates, respectively (Fig. 3A, Table I). In duodenum of Munchen VDR-ablated mice, ECaC2 is also significantly decreased, but only by about threefold compared with WT mice (Table I) [36]. Nevertheless, the latter investigators used a different method for quantification of duodenal RNA samples. Data on ECaC2 expression in the Boston VDR-null mice are lacking at present. Hence, unexpectedly, among the calcium absorption mechanisms, only the expression of the epithelial calcium channels is dramatically and consistently downregulated in three VDR KO strains. At first glance, this seems contradictory to former functional and molecular data suggesting that regulation of calbindin-D9K
TABLE I Differences in Gene Expression Pattern of the Three Calcium Transporter Genes in the 4 Different VDR KO Strains on Varying Calcium Diets: An Overview of the Literature Mouse
Leuven VDR KO [9] (10 w)
Tokyo VDR KO [7,9,33]
Boston VDR KO [34,35]
Diet
Low Ca
Normal
Rescue
Normal
Rescue
ECaC2 mRNA
↓1/500
↓1/10
↓1/3
10 w: ↓1/30 [9] 2 m: ↓1/20 [33]
2 m: ↓1/26 [33]
Calb-D9K mRNA
↓1/20
↓62%
=
3 and 7w: ↓ 1/3 [7] 10 w: ↓1/3 [9] 2 m: ↓1/2 [33]
Calb-D9K protein
PMCA1b mRNA
↓1/5
↓1/2
=
=
=
↓1/2
Normal
10 w: ↓1/3 [9] 2 m: ↓1/2 [33]
2 m: =
[33]
2 m: ↓69% [33]
2→11w: [35] ↓1/2→1/6 3 w: 1/5 [34] 6–18 w: = [34] 3–9 w: 1/2 [34] 13–18 w: = [34]
10 w: = [9]
Expression levels are compared to levels of WT littermates under the same dietary conditions. Age: weeks (w) and months (m).
Rescue
5→11w: = [35]
Munchen VDR KO [36] (Age?) Normal
Rescue
↓1/3
=
437
CHAPTER 25 Calcium Absorption and KO Models
expression is rate-limiting for vitamin D–dependent active calcium absorption. Nonetheless, revision of the literature reveals some flaws in this theory, which could be completed with the following new notions. For instance, after 1,25(OH)2D injection, a lack of correlation was found between the early rise in calcium transport and the later increase in calbindin-D9K content [26,59]. Likewise, the first event uncovered on ion microscopic images following 1,25(OH)2D-repletion of vitamin D–deficient animals is an intensification in calcium accumulation at the brush border region [38]. There is a significant lag between the time of hormone administration and redistribution of calcium away from the brush border, which coincides with the induction of calbindin-D9K synthesis [38]. In normal, vitamin D replete mice, a single pharmacological injection of 1,25(OH)2D (2 µg/kg body weight) induced the expression of all three calcium transporters in duodenum (Table II), which correlated with a doubling of the functional calcium absorption [9]. Interestingly, when evaluating the duodenal gene expression pattern over time, ECaC2 expression was already maximal at 6 hr (sixfold increase), whereas calbindin-D9K protein levels only peaked at 24 hr (threefold induction) PMCA1b expression was raised, as well, but to a lesser extent ([9], Table II, and unpublished personal data). In agreement with this data, Wood and co-workers showed that treatment of the human colon adenocarcinoma cell line Caco-2 with 1,25(OH)2D rapidly up-regulated ECaC2 expression (4 hr: fourfold; 34 hr: 10-fold, without any detection of ECaC1), which preceded by several hours the doubling of calbindinD9K [60]. Finally, calcium absorption in vivo and calbindin-D9K protein levels correlated on the one hand, fairly well in several experimental conditions (Fig. 3) [9,33], on the other hand, this correlation could not be established in Tokyo VDR KO mice on a 0.5% calcium diet [33].
TABLE II 1,25(OH)2D Injection and Duodenal Gene Expression Patterns in WT Mice
ECaC2 mRNA Calbindin-D9K mRNA Calbindin-D9K RIA PMCA1b mRNA
Vehicle
1,25(OH)2D3 6 hr
1,25(OH)2D3 24 hr
1 ± 0.2 1 ± 0.2 0.9 ± 0.2 1 ± 0.1
6.5 ± 0.8* 2.2 ± 0.3* 0.8 ± 0.18 1.8 ± 0.2*
6.2 ± 2.2* 42.5 ± 6.7* 3.1 ± 0.1* 1.4 ± 0.2
Time-response differences measured in WT mice (n = 6/group). mRNA levels are expressed relative to levels in vehicle-treated mice. CalbindinD9K protein levels are expressed as % of total duodenal protein levels. *P < 0.05 vs. WT vehicle.
In summary, the study of gene expression patterns in duodena of VDR-null mice implies a crucial role for the epithelial calcium channels in active calcium absorption, probably combined with a more complex regulation driven by the interaction between calcium influx and intracellular calcium transfer (Fig. 1). In this way the calcium buffering capacity of the cell keeps step with the increase in apical calcium entry and serves to maintain a low calcium concentration in close proximity of the influx pathway. Indeed, increased intracellular free calcium concentration would lead to negative feedback regulation of ECaC2and ECaC1-mediated channel activity [49,61,62]. Hence, in the absence of sufficient calcium buffering capacity, ECaC2 and ECaC1 functional activity would decrease and nullify the increased expression of ECaC channels at the apical surface. This is a prerequisite for all cells involved in transcellular calcium transport since the increase in intracellular calcium concentration associated with enhanced apical calcium uptake would otherwise be cytotoxic. Further research, including the characterization of the phenotype of ECaC1, ECaC2, or calbindin-D9K null mice or calcium channel rescue of VDR KO mice, will be necessary to support this concept. 5. DIETARY INTERVENTIONS AND ADAPTATION CALCIUM ABSORPTION IN VDR WT AND KO MICE
OF INTESTINAL
It is generally accepted that the primary homeostatic mechanism for stimulating active calcium absorption in response to low dietary calcium intake is an increase in 1,25(OH)2D. Conversely, when calcium is administered abundantly, no drive for active calcium transport would be present. This pattern is indeed observed when in vivo calcium absorption is measured in WT Swiss mice. One week of 0.02% calcium diet doubles, min whereas 1 week of rescue diet halves, the AUC0–10 po when compared with values of Swiss mice on a normal calcium diet (Fig. 3B, left panel) [9]. The in situ ligated loop procedure in 3-month-old Tokyo WT mice demonstrates analogous effects: feeding a 2% calcium diet for 1 week suppresses calcium absorption by 75%, while feeding a 0.02% calcium diet for 1 week increases calcium absorption 2.3-fold (which was associated with concordant changes in plasma 1,25(OH)2D [33]. The influence of dietary calcium on duodenal expression levels of the calcium transporter genes was extensively studied in the Leuven strain (Fig. 3A, Table I) [9]. First, WT animals present with a significant 2.5-fold increase of calbindin-D9K protein content on the low-calcium diet and a more than 70% decrease in calbindin-D9K RNA and protein level on the rescue diet. In the KO animals consuming either high- or
438 low-calcium diet, calbindin-D9K RNA level and protein levels decrease to less than 40% of KO levels on a normal diet. Second, PMCA1b gene expression is significantly decreased (by two to three times) in WT mice on the low-calcium diet and in KO mice on both low- and high-calcium diet. Third, ECaC2 expression varies considerably in WT mice: a sixfold increase with calcium restriction and a 90% reduction with calcium abundance, compared to levels of WT mice on normal diet. Although ECaC2 expression is already severely impaired in VDR KO animals, it has still a tendency to further decline on both low- and highcalcium diet (Fig. 3A, Table I). Despite less pronounced differences and lower detection levels, a similar pattern is found for ECaC1 RNA levels in the different conditions [9]. Administration of the rescue diet results in normalization of ECaC2 RNA expression in the Munchen VDR KO [36], and of calbindin-D9K RNA levels in the Boston KO (Table I) [35]. Though calbindin-D9K levels are almost indistinguishable from WT levels, ECaC2 expression remains markedly lower in rescued Tokyo VDR KO animals, compared to WT littermates on the same diet [33] (Table I). On the whole, duodenal expression patterns in the different VDR KO models tend to resemble the WT profile on the rescue diet (except for ECaC2 expression in the Tokyo VDR KO). By contrast, lack of adaptation to the low-calcium diet accentuates the difference in duodenal gene expression pattern between Leuven VDR WT and KO animals, and especially, the discrepancy in ECaC2 expression (Table I). Hence, ECaC2 (and ECaC1) expression is not only strongly VDR-dependent and consistently downregulated in VDR-KO strains, but it fluctuates also markedly in WT animals according to dietary calcium-driven variations in 1,25(OH)2D [9,33]. This provides additional evidence that calcium influx, interacting with intracellular calcium transfer mechanisms, should be considered as the rate-limiting step in the process of vitamin D–dependent duodenal calcium absorption.
B. Deficiency in 1,25(OH)2D Production (vitamin D–Dependent Rickets Type I) 1. GENERATION OF 1α(OH)ASE-NULL MICE
The engineering of 1α(OH)ase-null mice was reported by two different groups. The 1α(OH)ase gene was ablated by targeting part of exon 7 and 8 [5], or exons 6 to 9 [6]. In both cases the heme-binding domain and part of or the complete hormone-binding
ROGER BOUILLON,
ET AL .
domain were knocked out. No exact data are available on life expectancy in these mice, but both strains demonstrate failure to thrive (see Chapter 7). 2. MINERAL HOMEOSTASIS IN 1α(OH)ASE-NULL MICE
Regarding calcium homeostasis both 1α(OH)ase null models are identical. They demonstrate hypocalcemia, severe secondary hyperparathyroidism, hypophosphatemia, elevated serum alkaline phosphatase and increased phosphaturia, analogous to the phenotype of the VDR-null mice [5,6]. However, 1α(OH)ase-null mice genuinely lack the capacity to synthesize 1,25(OH)2D: circulating levels of 1,25(OH)2D were undetectable, whereas circulating levels of 25(OH)D doubled [5,6]. Typical features of advanced rickets are observed histologically in bone of 1α(OH)ase-null mice: disorganization and widening of the columnar alignment of hypertrophic chondrocytes, resulting in increased width of the growth plate, impaired calcification of the hypertrophic cartilage, and accumulation of osteoid in trabecular and cortical bone [5,6]. This emphasizes the nonredundant role played by the renal 1α(OH)ase in producing the hormonally active metabolite of vitamin and shows that the pathology of the 1α(OH)ase-null mice is the equivalent of the human VDDR-I disease. The treatment of choice for VDRR-I is long-term replacement therapy with 1,25(OH)2D. This strategy was also applied on the 1α(OH)ase-null mice [15,63]; the effect of the rescue diet was also evaluated [14,15,64]. Either treatment with subcutaneous 1,25(OH)2D, or feeding the high-calcium and lactose diet effectively normalizes biochemical parameters (calcemia, hyperparathyroidism), restores the biomechanical properties of bones, corrects growth plate architecture, and cures rickets and osteomalacia in the 1α(OH)ase-null mice generated by the group of R. St-Arnaud [14,63]. Intriguingly, under the rescue diet, the overall size of the femur is the only parameter that is not normalized in this 1α(OH)ase KO strain [14]. Presumably, 1,25(OH)2D treatment stimulates active calcium absorption in these mice, whereas on the rescue diet calcium is delivered via the passive pathway. Passive calcium uptake may not be able to adequately meet the demands for calcium during rapid growth spurts in this animal model. However, the latter phenomenon is only observed in the 1α(OH)ase-null mice and not in the VDR KO mice. The team of Goltzman report also that despite normalization of calcemia and bone mineral content with the rescue diet, PTH levels remain elevated and the growth plate remains larger and more distorted in their specific 1α(OH)ase-null strain [64].
439
CHAPTER 25 Calcium Absorption and KO Models
It is not excluded that strain-specific effects intervene or that the background of the KO mice is different. As already mentioned, a basal level of active calcium transport may be present (as the three steps of transcellular calcium transport are still expressed in the respective KO mice, though at lower levels), which may be strain-related. When comparing VDR KO, 1α(OH)ase KO and compound VDR-1α(OH)ase KO mice, Goltzman and colleagues found that hypocalcemia correlates with the absence of VDR and bone mineralization correlates with ambient calcium levels, but that the increase in trabecular volume and osteoblast number preferentially correlates with the absence of 1α(OH)ase activity [64]. Moreover, the compound VDR–1α(OH)ase KO is definitely more affected; suffering from more extreme hypocalcemia and failure to thrive, compared with the single KO models [64]. Unfortunately, data regarding the duodenal phenotype of these compound VDR-1α(OH)ase KO mice are lacking at present. Subsequently, St-Arnaud et al. engineered a mouse with selective inactivation of the 1α(OH)ase enzyme in growth plate chondrocytes in order to investigate the putative autocrine/paracrine roles of 1,25(OH)2D in the growth plate under conditions of normal mineral ion homeostasis. First examination of the bones from these mutant animals does not reveal any gross abnormalities [65]. 3. CALCIUM ABSORPTION IN 1α(OH)ASE-NULL MICE
Precise histomorphometric analysis is required to assess the impact of chondrocyte-specific inactivation on 1α(OH)ase on the maturation and function of growth plate cells. However, the preliminary results, which could not distinguish a phenotype in these mice, and the therapeutic effect of the rescue diet in 1α(OH)ase KO mice point once again toward the intestine as major site of action of 1,25(OH)2D. Moreover, one can postulate that the defects in the duodenal expression pattern of 1α(OH)ase KO mice should be analogous to the findings in VDR KO mice. Unfortunately, comparative data concerning mRNA expression levels of the calcium transporter genes in 1α(OH)ase WT and KO mice are currently limited to assessment of calbindin-D9K: similar to the Northern analysis of duodenum extracts from the Tokyo and Boston VDR-null mice, RNA levels are clearly downregulated in both models of 1α(OH)ase-null mice [5,6]. Strangely enough, high dietary calcium intake significantly increases the expression of ECaC2 (10-fold) and calbindin-D9K (25-fold) [15]. The mechanism that underlies this vitamin D–independent calcium-regulated pathway is still unknown. Calcium response elements
and/or transcription factors could be involved in the calcium-mediated regulation of gene expression found in that study [15]. As already mentioned, in Leuven and Tokyo VDR WT and KO mice, the high-calcium diet reduces the expression levels of the calcium transporter genes [9,33], which is in line with the hypothesis that active calcium absorption is shut down under these conditions. The data regarding the Boston [35] and Munchen [36] VDR KO are equivocal: the respective authors simply stipulate that calbindin-D9K and ECaC2 levels, respectively, “normalize” compared to WT levels, but it is not clear if WT levels per sé are altered on the rescue diet. Supplementation of 1α-OHase KO mice with 1,25(OH)2D results in a tremendous increase (more pronounced than on the rescue diet) in duodenal mRNA levels of ECaC2 (50-fold), calbindin-D9K (165-fold), and PMCA1b (3.4-fold). ECaC1 mRNA only reaches detectable levels after 1,25(OH)2D administration [15]. These findings support the idea that 1,25(OH)2D is a significant regulator of both epithelial calcium channels in duodenum and that the 1,25(OH)2D–VDR axis stimulates active intestinal calcium absorption by increasing the rate of calcium influx across the intestinal brush border membrane (Fig. 1). At present, comparative data on duodenal expression patterns between WT, 1α(OH)ase KO, VDR KO, and compound VDR1α(OH)ase KO mice, as well as functional analysis of calcium absorption in these mice, are eagerly anticipated.
C. Disturbed Phosphate Homeostasis 1. GENERATION AND CHARACTERISTICS OF NPT2-NULL MICE AND HYP MICE
(See Chapter 26.) Disruption of the major renal sodium-phosphate (Na/Pi) cotransporter (Npt2) gene by targeted ablation of exons 6 to 10 results in increased urinary phosphate excretion, hypophosphatemia, and a unique age-related skeletal phenotype [66]. In addition, Npt2-null mice exhibit increased renal 1α(OH)ase activity, a threefold increase in serum concentrations of 1,25(OH)2D (consistent with the hypothesis that hypophosphatemia is an important signal for renal production of 1,25(OH)2D), hypercalcemia (WT mice 2.27 ± 0.07 mM, vs. null mice 2.99 ± 0.11 mM), and calciuria. Fasting normalizes serum and urinary calcium [37]. Mice with X-linked hypophosphatemia (Hyp) harbor a large deletion in the PHEX “Phosphate regulating gene with homology to Endopeptidases on the X chromosome” gene. They suffer from renal phosphate wasting (due to a 50% reduction in renal Npt2 expression)
440 and hypophosphatemia. However, intestinal calcium absorption in Hyp mice is impaired secondary to inappropriate renal production of 1,25(OH)2D [67], resulting in mild hypocalcemia (normal mice 2.35 ± 0.1 mM, vs Hyp mice 2.03 ± 0.01 mM) [37]. Whereas primary defects of the 1,25(OH)2D–VDR axis intervene in VDR and 1α(OH)ase KO mice, a perturbed phosphate homeostasis triggers the modifications in calcium homeostasis in Npt2-null and Hyp mice. Nonetheless, functional and molecular analysis of active calcium absorption in the latter mouse models are interesting to corroborate the hypothesis based on the VDR and 1α(OH)ase KO mice. 2. CALCIUM ABSORPTION IN NPT2-NULL AND HYP MICE
The absorption of 45Ca by isolated duodenal loops and the appearance of 45Ca in plasma 2 and 4 min after its administration are significantly (about 1.5-fold of WT levels) higher in Npt2-null mice and significantly (less than 35% of levels in normal mice) lower in Hyp mice. This is accompanied by a doubling of ECaC2, ECaC1, and calbindin-D9K mRNA abundance in Npt2-null mice, as assessed by ribonuclease protection assay [37]. On the other hand, the relative abundance of both epithelial calcium channel transcripts is significantly lower in Hyp mice (less than 55%) than in normal littermates [37]. Another study described already a reduction in immunoreactive calbindin-D9K abundance in these mice [68]. These findings clearly indicate that altered intestinal calcium absorption is responsible, at least in part, for the variations in calcium homeostasis in the Npt2-null and Hyp mice. In addition, they provide further support for the hypothesis that ECaC2, in combination with calbindin-D9K, mediates alterations in active intestinal calcium absorption (Fig. 1).
IV. ACTIVE CALCIUM ABSORPTION DURING REPRODUCTION A. Introduction Though 1,25(OH)2D is regarded as the major regulator of active duodenal calcium absorption, other steroid hormones, such as estrogens, or other reproductionrelated hormones may influence the process also. The mineralization of the fetal skeleton and rapid skeletal growth of the newborn require hormone-mediated adjustments in maternal calcium metabolism. This maternal adaptation, during pregnancy [69–72] as well as during lactation [73,74], presumably includes a variable combination of renal calcium conservation, increased calcium resorption from the skeleton, and enhanced
ROGER BOUILLON,
ET AL .
intestinal calcium absorption. Moreover, reduced intestinal calcium absorption is considered to contribute to the pathogenesis of postmenopausal osteoporosis [75–79]. Calcium absorption may be restored or stimulated during estrogen therapy. Most animal and human studies have been hampered by the fact that physiological or therapeutic changes in estrogen status may be accompanied by changes in total and/or free 1,25(OH)2D levels [69,79–82]. Hypothetically, alterations in intestinal calcium absorption could be related to variations in 1,25(OH)2D [70,76,83], to changes in intestinal responsiveness to 1,25(OH)2D [77,84], or to direct action of other hormones on the intestine [73,78,79]. Both estrogen receptor (ER) alpha and beta are present in the intestine [85], as well as the prolactin [86] and the PTH/PTH-related protein (PTH/PTHrP) receptor [87]. The Leuven VDR KO mouse [9] is a unique animal model because it combines a relatively normal fertility with the absence of 1,25(OH)2D genomic actions. Therefore, this animal model can contribute important new data to our understanding of the physiological role of estrogens in the regulation of intestinal calcium transport and by extension the molecular mechanisms underlying regulation of calcium transport in reproduction [88].
B. Gestation and Intestinal Calcium Absorption in VDR WT and KO Mice By the end of gestation (day 18.5), VDR WT and KO mice receiving a normal calcium diet display a slight but significant increase in serum calcium concentration, compared to the appropriate nonpregnant controls (Table III) [32,89]. The concept that cancellous bone is accumulated during gestation as a store for calcium release during subsequent lactation has been recognized for many years [69]. Acquisition of trabecular bone mass was substantiated not only in VDR WT mice, but also in the KO animals (Table III) [32,89]. In duodenum, a significant two-fold stimulation of PMCA1b, calbindin-D9K mRNA, and calbindin-D9K protein expression is seen only in pregnant WT mice (Fig. 4A). This can be assigned to a 2.9-fold increase in circulating 1,25(OH)2D levels [32], an effect that has also been described in rats [90]. In contrast, an impressive (12-fold) and consistent up-regulation of duodenal ECaC2 expression is noted in both pregnant VDR WT and KO animals [32] (Fig. 4A). Consequently ECaC2 expression in pregnant VDR KO mice reaches the level of WT control mice. Additionally, in WT mice, no significant influence of pregnancy on VDR mRNA levels is uncovered (Fig. 4A) [32]. Pregnancy in rats is associated with a doubling of intestinal absorption
441
CHAPTER 25 Calcium Absorption and KO Models
TABLE III
Reproduction and Calcium Homeostasis in VDR WT and KO Mice VDR WT mice
Control Pregnant Control Lactating Sham Ovx Estrogen
VDR KO mice
Serum Ca2+ (mg/dl)
Femur Ca2+ (mg/g dw)
Trabecular density (mg/cm3)
8.9 ± 0.1 10.0 ± 0.3a 9.9 ± 0.3 10.5 ± 0.2 8.7 ± 0.1 8.6 ± 0.1 8.9 ± 0.1
nd nd 20.7 ± 0.5 19.6 ± 0.2 20.3 ± 0.3 20.7 ± 0.3 22.3 ± 0.7c
133 ± 19 204 ± 15a 122 ± 18 86 ± 10 92 ± 11 52 ± 7c 792 ± 29c
Serum Ca2+ (mg/dl)
Femur Ca2+ (mg/g dw)
Trabecular density (mg/cm3)
8.1 ± 0.3a 9.1 ± 0.2b 9.3 ± 0.2 5.3 ± 0.4a,b 7.3 ± 0.4c 7.2 ± 0.3c 7.3 ± 0.4c
nd nd 19.4 ± 0.6 16.7 ± 0.7a,b 17.7 ± 0.6c 18.1 ± 0.6c 21.3 ± 0.6d
138 ± 16 188 ± 25a 130 ± 19 63 ± 13a,b 91 ± 18 79 ± 15 579 ± 61c,d
Dry weight (dw). aP < 0.05 vs WT control. bP < 0.05 vs. KO control. cP < 0.05 vs. WT sham. dP < 0.05 vs. KO sham.
500
600
P
C
P
*
5
*
3
300
§
150 0
C
L
300 200 100
* *
*§
C
L
0
100
* 50
1
*
*
0
0
S O E S O E ECaC2
ERKOα
C P * 0
50 0
0.4 0.2
0
0.0 WT KO
WT KO
50
100
C
P
150
300
100
200
50
100
0
C L
C L
*
*
0
0
C
L
100 75 50 25 0
S O E
50
* 100
25 50 0
0
S O E
Vitamin D status
* 40
200
0
C P
150 100
100
mRNA (% WT control)
0
S O E
0.6
80
1,25(OH)2D3(pg/ml)
mRNA (% WT control)
50
Calb-D9K PMCA1b
120
WT KO
100
300
150
2
150
S O E
mRNA (% WT)
2 1
*
200
400
0
4
450
C L
**
C
150
*
mRNA (% WT control)
*
0
0
*
mRNA (% WT sham)
§
150
C L
1
VDR WT 150
1,25(OH)2D3(pg/ml)
300
2
VDR KO
1,25(OH)2D3(pg/ml)
*
3
100
C P
1350
0
200
Vitamin D status
200
50 1,25(OH)2D3(pg/ml)
1500
C P
4
mRNA (% WT control)
0
*
5
300
mRNA (% WT sham)
§
150
* *
VDR WT Calb/Tot prot (%)
300
VDR KO
Calb/Tot prot (%)
mRNA (% WT control)
1200
mRNA (% WT control)
mRNA (% WT control) mRNA (% WT control)
1350
S O E
B
VDR WT
mRNA (% WT sham)
Ovx ± E2
mRNA (% WT sham)
Lactation
VDR KO
*
Calb/Tot prot (%)
VDR WT Gestation
PMCA1b
Calbindin-D9K
120 80
25 40
mRNA (% WT)
ECaC2
Calb/Tot prot (%)
A
0
0 WT
KO
FIGURE 4 Effect of reproduction on active intestinal calcium absorption: gene expression patterns. (A) Gestation, lactation, ovariectomy, and estrogen therapy in VDR WT and KO mice: duodenal gene expression pattern and 1,25(OH)2D-VDR status. (B) ERα WT and ERKOα mice: duodenal gene expression pattern and 1,25(OH)2D-VDR status. Calbindin-D9K expression: mRNA levels: solid bars; protein levels: horizontal lines. 1,25(OH)2D levels: rising right pattern. VDR mRNA levels: diagonal cross-hatch pattern. Control (C); Pregnant (P); Lactating (L); Sham (S); Ovariectomy (Ovx / O); Estrogen therapy (E). *P<0.05 vs. WT control or vs WT sham, § P < 0.05 vs. KO control or vs. KO sham.
442
ROGER BOUILLON,
+
−
−
ET AL .
+ 1,25(OH)2D
Ca2+
VDR ?
E2 Prolactin PTH-rP/PTH ATP
Placental lactogen
?
CaBP9K ECaC2
ADP
PMCA1b
FIGURE 5 Active intestinal calcium absorption and gestation. Gene expression pattern.
of calcium, even in the absence of vitamin D [91,92]. Accordingly, an increased ECaC2 expression in duodenum of VDR KO mothers may be considered part of the mechanisms to accommodate the calcium demands of late gestation, in view of the fact that calcemia and trabecular bone density are increased to the same extent as in pregnant WT littermates (Fig. 5). This model implies that hormonal adjustments of pregnancy may have a physiological role in promoting calcium absorption by direct action on the intestine, besides an indirect 1,25(OH)2D-dependent effect (Fig. 5) [69,70]. Possible candidate hormones involved in pregnancy-related adaptation of ECaC2 mRNA expression are, for instance, high circulating estrogens, PTHrP [69], and placental lactogen [93,94], as well as prolactin [69,95,96] (Fig. 5). On the other hand, prolactin injections do not increase calcium absorption in late pregnancy in rats, nor does suppression of prolactin secretion by bromocriptine reduce the high calcium transport ratios immediately before parturition [94]. Further studies are needed to evaluate the relative contribution of these pregnancy hormones and the functional repercussions of this changed gene expression on intestinal calcium absorption [88].
C. Lactation and Intestinal Calcium Absorption in VDR WT and KO Mice In order to minimize the stress of gestation-related calcium demand for VDR KO mothers, all mice were studied consuming a high-calcium diet during gestation.
Once the pups were delivered, the diet was switched to a normal calcium diet till analysis at day 19 of lactation. Pups of VDR KO mothers exhibit normal survival and growth during 19 days of lactation. During lactation maternal calcium requirements are high in the setting of a fall in estrogen levels. A PTHrP mediated temporary demineralization of the skeleton is postulated as the main compensatory mechanism to fulfill these needs [69]. In this mouse model bone loss is undoubtedly observed in lactating VDR WT mice, but especially in VDR KO mice (Table III). Moreover, calcemia and calcium content in femur are not affected in lactating WT mice, whereas lactating VDR KO mice suffer from severe hypocalcemia and low calcium content in femur, compared to their control counterparts (Table III). Nevertheless, molecular and functional analysis indicates that active intestinal calcium absorption is additionally enhanced in a VDR-independent manner. Duodenal ECaC2 expression is manifestly raised (13-fold) in lactating WT and KO mice compared to nonlactating littermates (Fig. 4A). Hence, as in pregnant VDR KO animals, ECaC2 expression levels of lactating VDR KO mice become similar to those of control WT mice. Calbindin-D9K RNA and protein levels are increased five and two times, respectively, in lactating WT animals (Fig. 4A). In lactating VDR KO mice, calbindin-D9K protein levels double (Fig. 4A). PMCA1b mRNA levels increase about 1.6-fold in lactating WT and KO mice [32]. This VDR-independent effect was supported by unaltered 1,25(OH)2D-levels in VDR WT animals (Fig. 4.A) [32], in contrast to
443
CHAPTER 25 Calcium Absorption and KO Models
+
−
−
+ 1,25(OH)2D
Ca2+
VDR
?
Prolactin PTH-rP/PTH Placental lactogen
? ATP CaBP9K ADP
ECaC2
PMCA1b
FIGURE 6 Active intestinal calcium absorption and lactation. Gene expression pattern.
∆µMol serum calcium
previous reports of elevated 1,25(OH)2D-levels in lactating rodents [73,97]. Likewise, a study in lactating rats reported that vitamin D deprivation does not appreciably affect enhanced calcium transport in duodenal sacs, but that vitamin D was required to prevent detrimental hypocalcemia [73]. See Fig. 6. Accordingly, the in vivo calcium absorption technique [9] validated the assumption that these changes in gene expression patterns reflect enhanced calcium absorption in this stage of lactation (unpublished personal data). Figure 7 illustrates the increase of calcium appearance in serum of lactating mothers, resulting in a 3.7-fold increase of area under the curve in lactating WT mice (AUCpo0–10 min lactating WT: 18.4 ± 1.7 ∆µmol min vs control WT: 4.9 ± 0.9 ∆µmol min) and a 2.2-fold increase in lactating KO mice (AUCpo0–10min lactating KO: 7.7 ± 1.4 ∆µmol min vs. control KO: 3.6 ± 0.3 ∆µmol min). 2.5
*
*
*
*
*
2.0 1.5
§
§
4 6 8 Time (minutes)
10
§
1.0
§
§
0.5
WT lactating KO lactating WT control KO control
0.0 0
2
FIGURE 7 Active intestinal calcium absorption and lactation. Functional aspects: in vivo calcium absorption assay. *P < 0.05 vs. WT control, §P < 0.05 vs. KO control.
Since estrogens are low during lactation [70], further studies are needed to unravel the observed vitamin D– independent stimulatory mechanism on calcium absorption (Fig. 6). Parathyroidectomy does not influence active calcium absorption during lactation [73], and in spite of elevated PTH levels, active duodenal calcium absorption is impaired in VDR KO mice [9]. Apart from prolactin [69,96], PTHrP levels remain high during lactation, and both are therefore candidate hormones [69,98]. When PTHrP is selectively knocked out from the lactating mammary gland, circulating PTHrP levels are decreased, but not absent, and bone mass is preserved [99]. Remarkably, plasma calcium levels are stable in these dams despite normal milk calcium concentrations, normal PTH levels, reduced 1,25(OH)2D and PTHrP levels. Apparently, dietary calcium serves as a more important source of the total calcium used for milk production than does skeletal calcium in mice. However, the duodenal phenotype of these lactating mice has not been described yet, and the exact stimulatory mechanism on calcium absorption during lactation still remains elusive [99]. Finally, ileal active calcium absorption in late lactation has previously been described in rats [94]; however, the presence of such an additional calcium absorption mechanism has not yet been verified in lactating VDR KO mice. In contrast to gestation, lactation is a setting in which not only the mRNA abundance of ECaC2 but also the expression of the two other calcium transporter proteins is positively influenced in VDR KO mice. In view of the prolonged high calcium needs during lactation, it is not totally unexpected that evolutionary
444
ROGER BOUILLON,
mechanisms have created more than one mechanism to stimulate calcium absorption (Fig. 5).
D. Direct Genomic Effects of Estrogens on Intestinal Calcium Absorption 1. ESTROGEN TREATMENT AND INTESTINAL CALCIUM ABSORPTION
In order to expand the notion of VDR-independent, reproduction-related stimulation of calcium absorption it is important to elaborate on estrogen effects on the intestine. Implantation of 17β-estradiol pellets during four weeks (following ovariectomy) results in a doubling of the uterine wet weight in VDR WT and KO animals, indicating that the administered estrogen dose was supraphysiological and that both genotypes have the same responsiveness to estrogens [32]. In duodenum a significant increase in ECaC2 mRNA expression is found, not only in VDR WT females (fourfold), but also in VDR KO females (eightfold). In contrast, neither expression of calbindin-D9K mRNA and protein, nor of PMCA1b mRNA varies significantly in estrogen-treated animals (Figs. 4A and 8). Last, duodenal VDR mRNA expression is not modified by estrogen status in VDR WT animals (Fig. 4A). Although the suggested increased dietary calcium absorption has not yet been confirmed by functional data, the manifest increase in bone mass (six- to eightfold increase in trabecular bone density), which is paralleled by markedly enhanced calcium content in femur
+
(10% more Ca in WT mice, and up to 20% in KO mice) provides indirect evidence for it (Table III). Although estrogen administration is known to affect the skeleton directly [100], the source for this increase in bone calcium is most likely dietary. Reduced renal calcium loss is less likely as the urinary calcium/creatinine ratio is not changed in VDR WT mice [32] or is even increased in VDR KO mice [32,34]. Furthermore, this VDR-independent induction of ECaC2 mRNA expression agrees with the observation that a pharmacological dose of estradiol-benzoate significantly increased duodenal calcium absorption during in vivo duodenal single pass perfusion in rats, an effect that was blocked by the ER antagonist ICI 182780 [101]. In addition, rat duodenal enterocytes in vitro respond directly to 17β-estradiol with enhanced calcium uptake that is suppressed by transcription and protein synthesis inhibitors [102]. Moreover, Weber et al. described an estrogen response element in the promoter sequence of the mouse ECaC2 gene [36]. Additionally, estradiolmediated induction of intestinal ECaC2 mRNA expression has also been reported in ovariectomized rats (fourfold) and in 1α(OH)ase KO mice (ninefold) [15]. Consequently, the physiologically high concentrations of estrogens during pregnancy may definitely play a role in VDR independent induction of ECaC2 expression. Finally, estrogen replacement therapy has a long history of use in the therapy of postmenopausal bone loss. Thus revisiting the physiology of estrogen action on calcium transport may trigger the development of new therapies targeted not only at the bone, but also at the intestine (Fig. 8) [88].
−
−
+ 1,25(OH)2D
Ca2+ ERβ
VDR
ERα
E2 ATP
CaBP9K ECaC2
ET AL .
ADP
PMCA1b
FIGURE 8 Active intestinal calcium absorption and estrogen treatment. Gene expression pattern.
445
CHAPTER 25 Calcium Absorption and KO Models
2. ESTROGEN DEFICIENCY IN VDR WT AND KO MICE
As pharmacological doses of estrogen interfere with intestinal calcium absorption, the effect of estrogen deficiency could not be overlooked either. Several human studies report a decreased efficiency of global calcium absorption in normal postmenopausal women [75,103], which is more pronounced in postmenopausal osteoporosis [76,104]. Data on active calcium absorption in humans are lacking so far. In VDR WT mice, ovariectomy causes bone loss as evidenced by the 43% reduction in trabecular bone density, compared to sham-operated littermates (Table III), while loss of ovarian function does not significantly alter this parameter in VDR KO mice. Total and free 1,25(OH)2D levels double in ovariectomized VDR WT mice, which may represent a compensatory mechanism for the lack of endogenous estrogens (Table III). Duodenal expression pattern of the three known calcium transporter genes is not altered by ovariectomy in VDR WT females, or in VDR KO females, though (Fig. 4A). These findings agree with a normal active calcium absorption in ovariectomized rats, as revealed previously by in vivo single-pass perfusion [101]. However, the observations in these rodent studies should be interpreted with caution in light of the limitations inherent to the design of the study: the validity of extrapolation toward older animals and human postmenopausal osteoporosis has yet to be proven. 3. ESTROGEN RECEPTOR ALPHA (ERKOα) AND BETA (ERKOβ) HULL MICE
Following ovariectomy, endogenous estrogen production, although significantly reduced, is not absent, as adrenal androgens can be aromatized to estrogens. By contrast, in ERKOα and ERKOβ mice, genomic actions of estrogens mediated by their respective ERα and ERβ, are completely abolished. ERKOα mice were obtained by targeted ablation of exon 2 [105] and ERKOβ mice via homologous recombination replacing exon 2 and part of exon 3 (encoding the first Zn finger) by the neo selection gene [32]. ERα mRNA expression is much more abundant in murine duodenum than ERβ (unpublished personal data), a pattern opposite to human and rat intestine [106]. This finding probably explains the unchanged duodenal gene expression pattern in ERKOβ mice [32]. In contrast, ERKOα mice displayed a significant reduction of duodenal ECaC2 mRNA expression to 45% of their WT littermates, while calbindin-D9K and PMCA1b levels were unaffected (Fig. 4B). Furthermore, serum calcium and 1,25(OH)2D3 levels, as well as duodenal VDR expression, were not different from those in WT littermates (Fig. 4B), pleading for direct genomic actions of estrogens on mouse duodenal ECaC2 mRNA expression, mediated by the ERα (Fig. 8).
V. ACTIVE CALCIUM ABSORPTION AND CORTICOSTEROIDS A. Corticosteroids and Intestinal Calcium Absorption Since besides 1,25(OH)2D, estrogens also have distinct VDR-independent effects on intestinal calcium absorption, it is appealing to explore the effects of a third group of steroid hormones on intestinal calcium absorption: corticosteroids. “Malabsorption of calcium” is considered to be a fairly consistent finding in steroid-treated patients [107–109] and animals [109–113]. It is demonstrable within the first 2 weeks of steroid treatment, when concentrations of vitamin D metabolites are either normal or increased. It is generally believed that this calcium malabsorption is due to reduced active calcium absorption. However, a molecular substrate for this malabsorption has never been irrefutably identified. Moreover, it is still unclear whether corticosteroids act through the 1,25(OH)2DVDR system, or whether they exert direct effects on the intestine via their own nuclear receptor.
B. Dexamethasone and Intestinal Calcium Absorption in WT Mice One previous study in vitamin D–deficient Swiss mice showed an important reduction in calbindin-D9K mRNA expression (−86%, compared with vehicletreated mice) after treatment with 2 mg/kg dexamethasone for 5 days [114]. This test was repeated with vitamin D–replete WT mice: no difference could be detected between dexamethasone or vehicle-treated mice at the molecular or at the functional level, not even when the mice were fed a 0.02% calcium diet for 1 week prior to sacrifice (Fig. 9). Because of this lack of an evident effect in WT animals using available techniques, evaluation in VDR KO mice was not further explored (unpublished personal data).
VI. SUMMARY Duodenal gene expression of the epithelial calcium channels clearly react more dynamically to changes in calcium intake [9], vitamin D status [9,33,37], estrogen administration, gestation, or lactation [32] than any of the other calcium transporter genes (Table IV). This is paralleled by similar functional changes in active intestinal calcium absorption, when assessed in VDR WT and KO animals, under varying dietary calcium intake [9,33] and at the end of lactation. Hence it seems that these intestinal calcium influx mechanisms
Lessons from Intervention Studies in VDR WT and KO Mice
VDR KO
6 hr
24 hr
Low calcium diet
↑↑ECaC2 ↑Calbindin-D9K ↑PMCA1b
↑↑ ECaC2 ↑↑ Calbindin-D9K ↑ PMCA1b
↑↑ ECaC2 ↑ Calbindin-D9K
no effect
no effect
↑ ↑ ↑ ↑
VDR WT
Dietary intervention Rescue diet ECaC2
Calbindin-D9K
(PMCA1b ECaC2 Calbindin-D9K PMCA1b
↑ ↑ ↑↑
1,25(OH)2D3
↑ ↑↑
TABLE IV
ECaC2
Reproduction Gestation
Lactation
E2 treatment
↑↑↑ ECaC2 ↑ Calbindin-D9K ↑ PMCA1b ↑↑↑ ECaC2
↑↑↑ ECaC2 ↑ Calbindin-D9K ↑ PMCA1b ↑↑↑ ECaC2 ↑ Calbindin-D9K ↑ PMCA1b
↑ ECaC2
Calbindin-D9K
PMCA1b
↑ ECaC2
VDR dependent
VDR independent and dependent
VDR in- and dependent
VDR independent
VDR independent
1,25(OH)2D3
1,25(OH)2D3 + ?
Hormone(s) ?
Hormone(s) ?
E2-ERα dependent
Importance of regulation is related to font size. VDR-dependent effects are underlined.
Factor(s) ?
447
CHAPTER 25 Calcium Absorption and KO Models
Calb-D9K PMCA1b
80
1.0
40
0.5
0
0.0 V D
V
D
B
1.5
∆µMol serum calcium
mRNA (% Vehicle)
ECaC2 120
Calb/Tot prot (%)
A
2.5 2.0 1.5 1.0
WT Vehicle WT Dexa
0.5 0.0 0
V D
2
4
6
8
10
Time (minutes)
FIGURE 9 Active intestinal calcium absorption and dexamethasone treatment. (A) Molecular aspects. (B) Functional aspects: in vivo calcium absorption assay. Vehicle (V); Dexamethasone (D).
are the feedback-controlled part of the actually known steps in the transcellular calcium transport system. At first sight, calcium entry was unlikely to be the critical or rate-limiting factor of active intestinal calcium absorption, because of the favorable electrochemical and concentration gradient between gastrointestinal lumen and intracellular milieu. Since the epithelial calcium channels are so tightly regulated by vitamin D, reproduction and estrogens, calcium entry appears to be the rate-limiting step. Such a natural selection mechanism is in fact obvious as it avoids the massive entry of calcium when subsequent intracellular calcium buffering and shuttling mechanisms would fail. In such circumstances high intracellular calcium concentrations would soon become lethal for the enterocyte. On the other hand, actual uptake of calcium from the gut into the body definitely requires intracellular calcium buffering and shuttling toward the basolateral membrane with subsequent extrusion from the enterocyte (Fig. 1, Table IV). Thus, it is not surprising and even crucial that calbindin-D9K and PMCA1b expression are also induced when active calcium absorption is activated. This, however, was observed to be at a lesser extent than ECaC2 expression. In addition, analysis of female VDR KO mice has provided irrefutable evidence for VDR-independent stimulation of the expression of calcium transporter genes by estrogens and lactation [32], whereas during gestation a combination of VDR-dependent and -independent mechanisms intervene [32] (Table IV). Presumably, active intestinal calcium absorption is a more complex process than the actually known three major steps. Apart from the S100A10-annexin 2 complex, regulatory elements/genes are presently enigmatic. Further unraveling of the calcium absorption process promises to be a fascinating challenge for further research. Moreover, future studies using ECaC1, ECaC2, and calbindin-D9K KO mice to examine intestinal calcium absorption will be indispensable to
provide in vivo evidence of the contribution of each of these proteins to the process of intestinal calcium transport [88].
References 1. Reichel H, Koeffler HP, Norman AW 1989 The role of the vitamin D endocrine system in health and disease. N Engl J Med 320:980–991. 2. Haussler MR, Whitfield GK, Haussler CA, Hsieh JC, Thompson PD, Selznick SH, Dominguez CE, Jurutka PW 1998 The nuclear vitamin D receptor: biological and molecular regulatory properties revealed. J Bone Miner Res 13:325–349. 3. Kato S, Yanagisawa J, Murayama A, Kitanaka S, Takeyama K 1998 The importance of 25-hydroxyvitamin D3 1α-hydroxylase gene in vitamin D–dependent rickets. Curr Opin Nephrol Hypertens 7:377–383. 4. Malloy PJ, Pike JW, Feldman D 1999 The vitamin D receptor and the syndrome of hereditary 1,25-dihydroxyvitamin D–resistant rickets. Endocr Rev 20:156–188. 5. Dardenne O, Prud’homme J, Arabian A, Glorieux FH, St Arnaud R 2001 Targeted inactivation of the 25-hydroxyvitamin D3-1α-hydroxylase gene (CYP27B1) creates an animal model of pseudovitamin D–deficiency rickets. Endocrinology 142:3135–3141. 6. Panda DK, Miao D, Tremblay ML, Sirois J, Farookhi R, Hendy GN, Goltzman D 2001 Targeted ablation of the 25-hydroxyvitamin D 1α-hydroxylase enzyme: evidence for skeletal, reproductive, and immune dysfunction. Proc Natl Acad Sci USA 98:7498–7503. 7. Yoshizawa T, Handa Y, Uematsu Y, Takeda S, Sekine K, Yoshihara Y, Kawakami T, Arioka K, Sato H, Uchiyama Y, Masushige S, Fukamizu A, Matsumoto T, Kato S 1997 Mice lacking the vitamin D receptor exhibit impaired bone formation, uterine hypoplasia and growth retardation after weaning. Nat Genet 16:391–396. 8. Li YC, Pirro AE, Amling M, Delling G, Baron R, Bronson R, Demay, MB 1997 Targeted ablation of the vitamin D receptor: an animal model of vitamin D–dependent rickets type II with alopecia. Proc Natl Acad Sci USA 94:9831–9835. 9. Van Cromphaut SJ, Dewerchin M, Hoenderop JG, Stockmans I, Van Herck E, Kato S, Bindels RJ, Collen D, Carmeliet P, Bouillon R, Carmeliet G 2001 Duodenal
448
10.
11. 12.
13.
14.
15.
16.
17.
18. 19.
20.
21. 22.
23.
24.
ROGER BOUILLON,
calcium absorption in vitamin D receptor-knockout mice: functional and molecular aspects. Proc Natl Acad Sci USA 98:13324–13329. Balsan S, Garabedian M, Larchet M, Gorski AM, Cournot G, Tau C, Bourdeau A, Silve C, Ricour C 1986 Long-term nocturnal calcium infusions can cure rickets and promote normal mineralization in hereditary resistance to 1,25-dihydroxyvitamin D. J Clin Invest 77:1661–1667. Hochberg Z, Tiosano D, Even L 1992 Calcium therapy for calcitriol-resistant rickets. J Pediatr 121:803–808. Li YC, Amling M, Pirro AE, Priemel M, Meuse J, Baron R, Delling G, Demay MB 1998 Normalization of mineral ion homeostasis by dietary means prevents hyperparathyroidism, rickets, and osteomalacia, but not alopecia in vitamin D receptor-ablated mice. Endocrinology 139:4391–4396. Amling M, Priemel M, Holzmann T, Chapin K, Rueger JM, Baron R, Demay MB 1999. Rescue of the skeletal phenotype of vitamin D receptor-ablated mice in the setting of normal mineral ion homeostasis: formal histomorphometric and biomechanical analyses. Endocrinology 140:4982–4987. Dardenne O, Prud’homme J, Hacking SA, Glorieux FH, St Arnaud R 2003 Correction of the abnormal mineral ion homeostasis with a high-calcium, high-phosphorus, highlactose diet rescues the PDDR phenotype of mice deficient for the 25-hydroxyvitamin D-1α-hydroxylase (CYP27B1). Bone 32:332–340. Van Abel M, Hoenderop JG, van der Kemp AW, van Leeuwen JP, Bindels RJ 2003 Regulation of the epithelial Ca2+ channels in small intestine as studied by quantitative mRNA detection. Am J Physiol Gastrointest Liver Physiol 285:G78–G85. Wasserman RH 1997 Vitamin D and the intestinal absorption of calcium and phosphorus. In: Feldman D, Glorieux FH, and Pike JW (eds) Vitamin D. Academic Press, San Diego, pp. 259–273. Hoenderop JG, van der Kemp AW, Hartog A, van de Graaf SF, van Os CH, Willems PH, Bindels RJ 1999 Molecular identification of the apical Ca2+ channel in 1:25-dihydroxyvitamin D3responsive epithelia. J Biol Chem 274:8375–8378. Hoenderop JG, Hartog A, Stuiver M, Doucet A, Willems PH, Bindels RJ 2000 Localization of the epithelial Ca2+ channel in rabbit kidney and intestine. J Am Soc Nephrol 11:1171–1178. Muller D, Hoenderop JG, Meij IC, van den Heuvel LP, Knoers NV, den Hollander AI, Eggert P, Garcia-Nieto V, Claverie-Martin F, Bindels RJ 2000 Molecular cloning, tissue distribution, and chromosomal mapping of the human epithelial Ca2+ channel (ECAC1). Genomics 67:48–53. Hoenderop JG, Muller D, Kemp AW, Hartog A, Suzuki M, Ishibashi K, Imai M, Sweep F, Willems PH, Os CH, Bindels RJ 2001 Calcitriol controls the epithelial calcium channel in kidney. J Am Soc Nephrol 12:1342–1349. Hoenderop JG, Nilius B, Bindels RJ 2003 Epithelial calcium channels: from identification to function and regulation. Pflügers Arch 446:304–308. Muller D, Hoenderop JG, Merkx GF, van Os CH, Bindels RJ 2000 Gene structure and chromosomal mapping of human epithelial calcium channel. Biochem Biophys Res Commun 275:47–52. Peng JB, Chen XZ, Berger UV, Vassilev PM, Tsukaguchi H, Brown EM, Hediger MA 1999 Molecular cloning and characterization of a channel-like transporter mediating intestinal calcium absorption. J Biol Chem 274:22739–22746. Peng JB, Brown EM, Hediger MA 2001 Structural conservation of the genes encoding CaT1:CaT2:and related cation channels. Genomics 76:99–109.
ET AL .
25. Peng JB, Hediger MA 2002 A family of calcium-permeable channels in the kidney: distinct roles in renal calcium handling. Curr Opin Nephrol Hypertens 11:555–561. 26. Thomasset M 1997 Calbindin-D9K. In: Feldman D, Glorieux FH, and Pike JW (ed) Vitamin D. Academic Press, San Diego, pp. 223–232. 27. Christakos S, Gabrielides C, Rhoten WB 1989 Vitamin D– dependent calcium binding proteins: chemistry, distribution, functional considerations, and molecular biology. Endocr Rev 10:3–26. 28. Dupret JM, Brun P, Perret C, Lomri N, Thomasset M, Cuisinier-Gleizes P 1987 Transcriptional and posttranscriptional regulation of vitamin D–dependent calciumbinding protein gene expression in the rat duodenum by 1,25-dihydroxycholecalciferol. J Biol Chem 262:16553–16557. 29. Cai Q, Chandler JS, Wasserman RH, Kumar R, Penniston JT 1993 Vitamin D and adaptation to dietary calcium and phosphate deficiencies increase intestinal plasma membrane calcium pump gene expression. Proc Natl Acad Sci USA 90:1345–1349. 30. Pannabecker TL, Chandler JS, and Wasserman RH 1995 Vitamin-D-dependent transcriptional regulation of the intestinal plasma membrane calcium pump. Biochem Biophys Res Commun 213:499–505. 31. Wasserman RH, Smith CA, Brindak ME, De Talamoni N, Fullmer CS, Penniston JT, Kumar R 1992 Vitamin D and mineral deficiencies increase the plasma membrane calcium pump of chicken intestine. Gastroenterology 102:886–894. 32. Van Cromphaut SJ, Rummens K, Stockmans I, Van Herck E, Dijcks FA, Ederveen AG, Carmeliet P, Verhaeghe J, Bouillon R, Carmeliet G 2003 Intestinal calcium transporter genes are upregulated by estrogens and the reproductive cycle through vitamin D receptor-independent mechanisms. J Bone Miner Res 18:1725–1736. 33. Song Y, Kato S, Fleet JC 2003 Vitamin D receptor (VDR) knockout mice reveal VDR-independent regulation of intestinal calcium absorption and ECaC2 and calbindin D9k mRNA. J Nutr 133:374–380. 34. Li YC, Bolt MJ, Cao LP, Sitrin MD 2001 Effects of vitamin D receptor inactivation on the expression of calbindins and calcium metabolism. Am J Physiol Endocrinol Metab 281:E558–E564. 35. Li YC, Pirro AE, Demay MB 1998 Analysis of vitamin D–dependent calcium-binding protein messenger ribonucleic acid expression in mice lacking the vitamin D receptor. Endocrinology 139:847–851. 36. Weber K, Erben RG, Rump A, Adamski J 2001 Gene structure and regulation of the murine epithelial calcium channels ECaC1 and 2. Biochem Biophys Res Commun 289:1287–1294. 37. Tenenhouse HS, Gauthier C, Martel J, Hoenderop JG, Hartog A, Meyer MH, Meyer RA Jr, Bindels RJ 2002 Na/Pi cotransporter (Npt2) gene disruption increases duodenal calcium absorption and expression of epithelial calcium channels 1 and 2. Pflügers Arch 444:670–676. 38. Fullmer CS, Chandra S, Smith CA, Morrison GH, Wasserman RH 1996 Ion microscopic imaging of calcium during 1,25-dihydroxyvitamin D–mediated intestinal absorption. Histochem Cell Biol 106:215–222. 39. Chandra S, Fullmer CS, Smith CA, Wasserman RH, Morrison GH 1990 Ion microscopic imaging of calcium transport in the intestinal tissue of vitamin D-deficient and vitamin D-replete chickens: a 44Ca stable isotope study. Proc Natl Acad Sci USA 87:5715–5719. 40. Montell C, Birnbaumer L, Flockerzi V, Bindels RJ, Bruford EA, Caterina MJ, Clapham DE, Harteneck C, Heller S, Julius D,
CHAPTER 25 Calcium Absorption and KO Models
41. 42. 43.
44.
45. 46. 47.
48.
49.
50.
51.
52. 53.
54.
55. 56. 57.
Kojima I, Mori Y, Penner R, Prawitt D, Scharenberg AM, Schultz G, Shimizu N, Zhu MX 2002 A unified nomenclature for the superfamily of TRP cation channels. Mol Cell 9:229–231. Montell C 2001 Physiology, phylogeny, and functions of the TRP superfamily of cation channels. Sci STKE 90:RE1. Hoenderop JG, Nilius B, Bindels RJ 2002 ECaC: the gatekeeper of transepithelial Ca2+ transport. Biochim Biophys Acta 1600:6–11. Loffing J, Loffing-Cueni D, Valderrabano V, Klausli L, Hebert SC, Rossier BC, Hoenderop JG, Bindels RJ, Kaissling B 2001 Distribution of transcellular calcium and sodium transport pathways along mouse distal nephron. Am J Physiol Renal Physiol 281:F1021–F1027. Peng JB, Chen XZ, Berger UV, Weremowicz S, Morton CC, Vassilev PM, Brown EM, Hediger MA 2000 Human calcium transport protein CaT1. Biochem Biophys Res Commun 278:326–332. Voets T, Nilius B 2003 The pore of TRP channels: trivial or neglected? Cell Calcium 33:299–302. Voets T, Janssens A, Prenen J, Droogmans G, Nilius B 2003 Mg2+-dependent gating and strong inward rectification of the cation channel TRPV6. J Gen Physiol 121:245–260. Watanabe H, Vriens J, Janssens A, Wondergem R, Droogmans G, Nilius B 2003 Modulation of TRPV4 gating by intra- and extracellular Ca2+. Cell Calcium 33: 489–495. Nilius B, Vennekens R, Prenen J, Hoenderop JG, Droogmans G, Bindels RJ 2001 The single pore residue Asp542 determines Ca2+ permeation and Mg2+ block of the epithelial Ca2+ channel. J Biol Chem 276:1020–1025. Vennekens R, Hoenderop JG, Prenen J, Stuiver M, Willems PH, Droogmans G, Nilius B, Bindels RJ 2000 Permeation and gating properties of the novel epithelial Ca2+ channel. J Biol Chem 275:3963–3969. Hoenderop JG, Vennekens R, Muller D, Prenen J, Droogmans G, Bindels RJ, Nilius B 2001 Function and expression of the epithelial Ca2+ channel family: comparison of mammalian ECaC1 and 2. J Physiol 537:747–761. Hoenderop JG, Voets T, Hoefs S, Weidema F, Prenen J, Nilius B, Bindels RJ 2003 Homo- and heterotetrameric architecture of the epithelial Ca2+ channels TRPV5 and TRPV6. EMBO J 22:776–785. den Dekker E, Hoenderop JG, Nilius B, Bindels RJ 2003 The epithelial calcium channels, TRPV5 & TRPV6: from identification towards regulation. Cell Calcium 33:497–507. van de Graaf SF, Hoenderop JG, Gkika D, Lamers D, Prenen J, Rescher U, Gerke V, Staub O, Nilius B, Bindels RJ 2003 Functional expression of the epithelial Ca2+ channels (TRPV5 and TRPV6) requires association of the S100A10annexin 2 complex. EMBO J 22:1478–1487. Murayama A, Takeyama K, Kitanaka S, Kodera Y, Hosoya T, Kato S 1998 The promoter of the human 25-hydroxyvitamin D3 1α-hydroxylase gene confers positive and negative responsiveness to PTH, calcitonin, and 1 alpha,25(OH)2D3. Biochem Biophys Res Commun 249:11–16. Omdahl J, May B 1997 The 25-hydroxyvitamin D 24-hydroxylase. In: Vitamin D, 1st Edition Feldman D, Glorieux FH, and Pike JW (eds) Academic Press, San Diego, pp. 69–85. Dostal LA, Toverud SU 1984 Effect of vitamin D3 on duodenal calcium absorption in vivo during early development. Am J Physiol 246:G528–G534. Marcus CS, Lengemann FW 1962 Absorption of Ca45 and Sr85 from solid and liquid food at various levels of the alimentary tract of the rat. J Nutr 77:155–160.
449 58. Bronner F, Yost JH 1985 Saturable and nonsaturable copper and calcium transport in mouse duodenum. Am J Physiol 249:G108–G112. 59. Bishop CW, Kendrick NC, Deluca HF 1983. Induction of calcium-binding protein before 1,25-dihydroxyvitamin D3 stimulation of duodenal calcium uptake. J Biol Chem 258:1305–1310. 60. Wood RJ, Tchack L, Taparia S 2001 1,25-Dihydroxyvitamin D3 increases the expression of the CaT1 epithelial calcium channel in the Caco-2 human intestinal cell line. BMC Physiol 1:11–18. 61. Yue L, Peng JB, Hediger MA, Clapham DE 2001 CaT1 manifests the pore properties of the calcium-release-activated calcium channel. Nature 410:705–709. 62. Nilius B, Prenen J, Vennekens R, Hoenderop JG, Bindels RJ, Droogmans G 2001 Modulation of the epithelial calcium channel, ECaC, by intracellular Ca2+. Cell Calcium 29:417–428. 63. Dardenne O, Prudhomme J, Hacking SA, Glorieux FH, St Arnaud R 2003 Rescue of the pseudo-vitamin D deficiency rickets phenotype of CYP27B1-deficient mice by treatment with 1,25-dihydroxyvitamin D3: biochemical, histomorphometric, and biomechanical analyses. J Bone Miner Res 18:637–643. 64. Panda DK, Miao D, Huo R, Hendy GN, Goltzman D 2003 Comparison of mice with deletion of the 25OH Vitamin D3 1αHydroxylase, of the vitamin D receptor and of both, suggest an in vivo role for VDR-dependent and independent regulation of calcium and skeletal homeostasis. J Bone Miner Res 17 (Suppl 1):1029. 65. St Arnaud R, Dardenne O, Prud’homme J, Hacking SA, Glorieux FH 2003 Conventional and tissue-specific inactivation of the 25-hydroxyvitamin D-1α-hydroxylase (CYP27B1). J Cell Biochem 88:245–251. 66. Beck L, Karaplis AC, Amizuka N, Hewson AS, Ozawa H, Tenenhouse HS 1998. Targeted inactivation of Npt2 in mice leads to severe renal phosphate wasting, hypercalciuria, and skeletal abnormalities. Proc Natl Acad Sci USA 95: 5372–5377. 67. Meyer RA Jr, Meyer MH, Gray RW, Bruns ME 1987 Evidence that low plasma 1,25-dihydroxyvitamin D causes intestinal malabsorption of calcium and phosphate in juvenile X-linked hypophosphatemic mice. J Bone Miner Res 2:67–82. 68. Bruns ME, Meyer RA Jr, Meyer MH 1984 Low levels of intestinal vitamin D-dependent calcium-binding protein in juvenile X-linked hypophosphatemic mice. Endocrinology 115:1459–1463. 69. Kovacs CS, Kronenberg HM 1997 Maternal–fetal calcium and bone metabolism during pregnancy, puerperium, and lactation. Endocr Rev 18:832–872. 70. Ritchie LD, Fung EB, Halloran BP, Turnlund JR, Van Loan MD, Cann CE, King JC 1998 A longitudinal study of calcium homeostasis during human pregnancy and lactation and after resumption of menses. Am J Clin Nutr 67:693–701. 71. Kent GN, Price RI, Gutteridge DH, Rosman KJ, Smith M, Allen JR, Hickling CJ, Blakeman SL 1991 The efficiency of intestinal calcium absorption is increased in late pregnancy but not in established lactation. Calcif Tissue Int 48:293–295. 72. Kent GN, Price RI, Gutteridge DH, Allen JR, Blakeman SL, Bhagat CI, St John A, Barnes MP, Smith M, Evans DV 1991 Acute effects of an oral calcium load in pregnancy and lactation: findings on renal calcium conservation and biochemical indices of bone turnover. Miner Electrolyte Metab 17:1–7.
450 73. Boass A, Toverud SU, Pike JW, Haussler MR 1981 Calcium metabolism during lactation: enhanced intestinal calcium absorption in vitamin D–deprived, hypocalcemic rats. Endocrinology 109:900–907. 74. Boass A, Toverud SU 1997 Enhanced nonsaturable calcium transport in the jejunum of rats during lactation, but not during pregnancy. J Bone Miner Res 12:1577–1583. 75. Heaney RP, Recker RR, Stegman MR, Moy AJ 1989 Calcium absorption in women: relationships to calcium intake, estrogen status, and age. J Bone Miner Res 4:469–475. 76. Gallagher JC, Riggs BL, Eisman J, Hamstra A, Arnaud SB, Deluca HF 1979 Intestinal calcium absorption and serum vitamin D metabolites in normal subjects and osteoporotic patients: effect of age and dietary calcium. J Clin Invest 64:729–736. 77. Gennari C, Agnusdei D, Nardi P, Civitelli R 1990 Estrogen preserves a normal intestinal responsiveness to 1,25-dihydroxyvitamin D3 in oophorectomized women. J Clin Endocrinol Metab 71:1288–1293. 78. O’Loughlin PD, Morris HA 1998 Oestrogen deficiency impairs intestinal calcium absorption in the rat. J Physiol 511:313–322. 79. Colin EM, Van Den Bemd GJ, Van Aken M, Christakos S, De Jonge HR, Deluca HF, Prahl JM, Birkenhager JC, Buurman CJ, Pols HA, van Leeuwen JP 1999 Evidence for involvement of 17β-estradiol in intestinal calcium absorption independent of 1,25-dihydroxyvitamin D3 level in the rat. J Bone Miner Res 14:57–64. 80. Kalu DN, Chen C 1999 Ovariectomized murine model of postmenopausal calcium malabsorption. J Bone Miner Res 14:593–601. 81. Chen C, Noland KA, Kalu DN 1997 Modulation of intestinal vitamin D receptor by ovariectomy, estrogen and growth hormone. Mech Ageing Dev 99:109–122. 82. Bouillon R, Van Assche FA, Van Baelen H, Heyns W, De Moor P 1981 Influence of the vitamin D-binding protein on the serum concentration of 1,25-dihydroxyvitamin D3. Significance of the free 1,25-dihydroxyvitamin D3 concentration. J Clin Invest 67:589–596. 83. Prince RL, Smith M, Dick IM, Price RI, Webb PG, Henderson NK, Harris MM 1991 Prevention of postmenopausal osteoporosis. A comparative study of exercise, calcium supplementation, and hormone-replacement therapy. N Engl J Med 325:1189–1195. 84. Liel Y, Shany S, Smirnoff P, Schwartz B 1999 Estrogen increases 1,25-dihydroxyvitamin D receptors expression and bioresponse in the rat duodenal mucosa. Endocrinology 140:280–285. 85. Campbell-Thompson M 1997 Estrogen Receptor alpha and beta expression in upper gastrointestinal tract with regulation of Trefoil Factor Family 2 mRNA levels in ovariectomized rats. Biochem Biophys Res Commun 240:478–483. 86. Nagano M, Chastre E, Choquet A, Bara J, Gespach C, Kelly PA 1995 Expression of prolactin and growth hormone receptor genes and their isoforms in the gastrointestinal tract. Am J Physiol 268:G431–G442. 87. Li H, Seitz PK, Thomas ML, Selvanayagam P, Rajaraman S, Cooper CW 1995 Widespread expression of the parathyroid hormone-related peptide and PTH/PTHrP receptor genes in intestinal epithelial cells. Lab Invest 73:864–870. 88. Christakos S, Prince R 2003 Estrogen, vitamin D, and calcium transport. J Bone Miner Res 18:1737–1739. 89. Rummens K, Van Cromphaut SJ, Carmeliet G, Van Herck E, van Bree R, Stockmans I, Bouillon R, Verhaeghe J 2003
ROGER BOUILLON,
90.
91. 92. 93. 94. 95.
96. 97. 98.
99.
100.
101.
102. 103. 104. 105.
106.
ET AL .
Pregnancy in mice lacking the vitamin D receptor: normal maternal skeletal response, but fetal hypomineralization rescued by maternal calcium supplementation. Pediatr Res 54:466–473. Zhu Y, Goff JP, Reinhardt TA, Horst RL 1998 Pregnancy and lactation increase vitamin D–dependent intestinal membrane calcium adenosine triphosphatase and calcium binding protein messenger ribonucleic acid expression. Endocrinology 139:3520–3524. Halloran BP, Deluca HF 1980 Calcium transport in small intestine during pregnancy and lactation. Am. J Physiol 239:E64–E68. Brommage R, Baxter DC, Gierke LW 1990 Vitamin D– independent intestinal calcium and phosphorus absorption during reproduction. Am J Physiol 259:G631–G638. Mainoya JR 1975 Effects of bovine growth hormone, human placental lactogen and ovine prolactin on intestinal fluid and ion transport in the rat. Endocrinology 96:1165–1170. Boass A, Lovdal JA, Toverud SU 1992 Pregnancy- and lactation-induced changes in active intestinal calcium transport in rats. Am J Physiol 263:G127–G134. Lotinun S, Limlomwongse L, Krishnamra N 1998 The study of a physiological significance of prolactin in the regulation of calcium metabolism during pregnancy and lactation in rats. Can J Physiol Pharmacol 76:218–228. Pahuja DN, Deluca HF 1981 Stimulation of intestinal calcium transport and bone calcium mobilization by prolactin in vitamin D-deficient rats. Science 214:1038–1039. Pike JW, Parker JB, Haussler MR, Boass A, Toverud SV 1979 Dynamic changes in circulating 1,25-dihydroxyvitamin D during reproduction in rats. Science 204:1427–1429. Lippuner K, Zehnder HJ, Casez JP, Takkinen R, Jaeger P 1996 PTH-related protein is released into the mother’s bloodstream during location: evidence for beneficial effects on maternal calcium-phosphate metabolism. J Bone Miner Res 11:1394–1399. VanHouten JN, Dann P, Stewart AF, Watson CJ, Pollak M, Karaplis AC, Wysolmerski JJ 2003 Mammary-specific deletion of parathyroid hormone-related protein preserves bone mass during lactation. J Clin Invest 112:1429–1436. Edwards MW, Bain SD, Bailey MC, Lantry MM, Howard GA 1992 17β Estradiol stimulation of endosteal bone formation in the ovariectomized mouse: an animal model for the evaluation of bone-targeted estrogens. Bone 13:29–34. Ten Bolscher M, Netelenbos JC, Barto R, Van Buuren LM, Van der vijgh WJ 1999 Estrogen regulation of intestinal calcium absorption in the intact and ovariectomized adult rat. J Bone Miner Res 14:1197–1202. Arjmandi BH, Salih MA, Herbert DC, Sims SH, Kalu DN 1993 Evidence for estrogen receptor–linked calcium transport in the intestine. Bone Miner 21:63–74. Heaney RP, Recker RR, Saville PD 1978 Menopausal changes in calcium balance performance. J Lab Clin Med 92:953–963. Gallagher JC, Aaron J, Horsman A, Marshall DH, Wilkinson R, Nordin BE 1973 The crush fracture syndrome in postmenopausal women. Clin Endocrinol Metab 2:293–315. Korach KS, Couse JF, Curtis SW, Washburn TF, Lindzey J, Kimbro KS, Eddy EM, Migliaccio S, Snedeker SM, Lubahn DB, Schomberg DW, Smith EP 1996 Estrogen receptor gene disruption: Molecular characterization and experimental and clinical phenotypes. Recent Prog Horm Res 51:159–186. Campbell-Thompson M, Lynch IJ, Bhardwaj B 2001. Expression of estrogen receptor (ER) subtypes and ERβ isoforms in colon cancer. Cancer Res 61:632–640.
CHAPTER 25 Calcium Absorption and KO Models
107. Reid IR 1997 Glucocorticoid osteoporosis—mechanisms and management. Eur J Endocrinol 137:209–217. 108. Shrivastava A, Lyon A, Mcintosh N 2000 The effect of dexamethasone on growth, mineral balance and bone mineralisation in preterm infants with chronic lung disease. Eur J Pediatr 159:380–384. 109. Smith BJ, Phillips PJ, Pannall PR, Cain HJ, Leckie WJ 1993 Effect of orally administered beclomethasone dipropionate on calcium absorption from the gut in normal subjects. Thorax 48:890–893. 110. Aloia JF, Semla HM, Yeh JK 1984 Discordant effects of glucocorticoids on active and passive transport of calcium in the rat duodenum. Calcif Tissue Int 36:327–331.
451 111. Glade MJ, Krook L, Schryver HF, Hintz HF 1982 Calcium metabolism in glucocorticoid-treated pony foals. J Nutr 112:77–86. 112. Weiler HA, Wang Z, Atkinson SA 1995 Dexamethasone treatment impairs calcium regulation and reduces bone mineralization in infant pigs. Am J Clin Nutr 61:805–811. 113. Yeh JK, Aloia JF 1986 Influence of glucocorticoids on calcium absorption in different segments of the rat intestine. Calcif Tissue Int 38:282–288. 114. Li H, Christakos S 1991 Differential regulation by 1,25-dihydroxyvitamin D3 of calbindin-D9k and calbindin-D28k gene expression in mouse kidney. Endocrinology 128: 2844–2852.
CHAPTER 26
Phosphate Homeostasis HARRIET S. TENENHOUSE ANTHONY A. PORTALE
I. II. III. IV. V.
Departments of Pediatrics and Human Genetics, McGill University and The Montreal Children’s Hospital Research Institute, Montreal, Quebec Department of Pediatrics, University of California San Francisco, San Francisco, California
Introduction Phosphate Homeostasis Intestinal Phosphate Absorption Renal Phosphate Transport Role of Phosphate in Regulation of Renal Vitamin D Metabolism
I. INTRODUCTION Inorganic phosphate (Pi) is fundamental to cellular metabolism and, in vertebrates, to skeletal mineralization. To accomplish these functions, transport systems have evolved to permit the efficient transfer of negatively charged Pi ions across hydrophobic membrane barriers. Ingested Pi is absorbed by the small intestine, deposited in bone, and filtered by the kidney where it is reabsorbed and excreted in amounts that are determined by the specific requirements of the organism. The kidney is a major determinant of Pi homeostasis by virtue of its ability to increase or decrease its Pi reabsorptive capacity to accommodate Pi need. Accordingly, significant advances have been made in our understanding of the molecular mechanisms involved in renal tubular Pi reabsorption and its hormonal regulation and modulation by dietary Pi intake. This chapter focuses on Pi chemistry and homeostasis, the cellular and molecular aspects of intestinal and renal Pi transport and their regulation, and Pi transport in bone. In addition, the role of novel Pi regulating genes, PHEX and FGF23, in renal Pi handling and vitamin D metabolism will be described. For a more detailed discussion of renal Pi wasting disorders in humans, the reader is referred to Chapters 69 and 70 in this volume.
II. PHOSPHATE HOMEOSTASIS
VI. Phosphate Transport in Bone VII. Disorders Causing Phosphate Deficiency VIII. Summary and Conclusions References
Pi is an important constituent of bone mineral, and thus in growing individuals, the balance of Pi must be positive to meet the needs of skeletal growth and consolidation; in the adult, Pi balance is zero. Pi deficiency results in osteomalacia in both children and adults.
B. Phosphate Chemistry Phosphate exists in the plasma in two forms, an organic form, consisting principally of phospholipids and phosphate esters, and an inorganic form [2]. Of the total plasma Pi concentration of approximately 14 mg/dl (4.52 mM), about 4 mg/dl (1.29 mM) is in the inorganic form. Of this, about 10% to 15% is protein bound and the remainder, which is freely filtered by the renal glomerulus, exists principally either as “free” Pi ions or as Pi complexed with sodium, calcium, or magnesium. In principle, four forms of inorganic Pi are present in biological solutions: H3PO4, H2PO4−, HPO42−, and PO43−, which are in a pH-dependent equilibrium. However, at physiological pH, only HPO42− and H2PO4− are present at significant concentrations. The ratio of the divalent to monovalent forms can be determined by the HendersonHasselbalch equation, pH = pKa + log (HPO42−/H2PO4−). The dissociation constant, pKa, for Pi is 6.8. Thus, at a pH of 7.4, the ratio of divalent to monovalent Pi anions is essentially 4:1. In clinical settings, only the inorganic orthophosphate form of Pi is routinely measured.
A. Phosphate Distribution in the Body Pi accounts for approximately 0.6% of body weight at birth and about 1% of body weight, or 600 to 700 g, in the adult [1]. Approximately 85% of body Pi is in the skeleton and teeth, approximately 15% is in soft tissue, and the remainder (∼0.3%) is in extracellular fluid. VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
C. Extracellular Phosphate Homeostasis In the adult in zero Pi balance, net intestinal absorption of Pi (dietary Pi minus fecal Pi) is approximately 60% to 65% of dietary intake. To satisfy the demands Copyright © 2005, Elsevier, Inc. All rights reserved.
454
HARRIET S. TENENHOUSE AND ANTHONY A. PORTALE
of rapid growth of bone and soft tissue, net intestinal absorption of Pi in infants is higher than in the adult and can exceed 90% of dietary intake [3,4]. Metabolic balance studies in normal adult humans reveal that over the customary range of dietary Pi, net absorption is a linear function of intake [5], with no indication of saturation. Thus, inadequate net Pi absorption results primarily from decreased Pi availability rather than from changes in the intrinsic capacity of intestinal Pi transport. A small amount of Pi is secreted into the intestinal lumen in digestive fluids. Absorbed Pi enters the extracellular Pi pool, which is in equilibrium with the bone and soft-tissue Pi pools. Pi is filtered at the glomerulus and is reabsorbed to a large extent by the renal tubule. In subjects in zero Pi balance, the amount of Pi excreted by the kidney is equal to the net amount absorbed by the intestine and in growing children is less than the net amount absorbed due to deposition of Pi in bone. An overall schema of Pi homeostasis is depicted in Fig. 1. Renal tubular reabsorption of Pi plays a central role in the regulation of plasma Pi concentration and Pi homeostasis. In response to a decrease in the extracellular
Pi concentration, urine excretion of Pi decreases promptly. This acute response reflects both a decrease in the filtered Pi load and an intrinsic adaptive increase in proximal tubule Pi reabsorption induced by hypophosphatemia or decreased dietary Pi intake (see Section IV,E,3). Hypophosphatemia also is a potent stimulus for the renal synthesis of 1,25-dihydroxyvitamin D3 (1,25(OH)2D) [6–8] (see Section V), and the resulting increase in serum 1,25(OH)2D concentration acts to stimulate intestinal absorption of Pi and calcium absorption and their mobilization from bone. Hypophosphatemia also can directly promote mobilization of Pi and calcium from bone. With an increase in plasma calcium concentration, PTH release is suppressed, which leads to a further decrease in renal Pi excretion but an increase in calcium excretion. These homeostatic adjustments result in an increase in extracellular Pi concentration toward normal values, with little change in serum calcium concentration. Conversely, in response to an increase in plasma Pi concentration, production of 1,25(OH)2D is decreased and release of PTH is increased. The effects of hyperphosphatemia on bone, kidney, and intestine are opposite to those
DIET 1400 mg
ECF Pi Absorption
Formation
1100 mg GUT
350 mg 550 mg
BONE
Secretion
Resorption
200 mg
350 mg
7000 mg
6100 mg
FECES 500 mg
KIDNEY
URINE 900 mg
FIGURE 1 Phosphate fluxes between body pools in the normal human adult in zero phosphate balance. Reprinted with permission from [189].
455
CHAPTER 26 Phosphate Homeostasis
occurring with hypophosphatemia, the net result being a decrease in plasma Pi concentration to normal values. In healthy subjects ingesting typical diets, the serum Pi concentration exhibits a circadian rhythm, characterized by a rapid decrease in early morning to a nadir shortly before noon, a subsequent increase to a plateau in late afternoon, and a small further increase to a peak shortly after midnight [9,10]. The amplitude of the rhythm (nadir to peak) is approximately 1.2 mg/dl (0.39 mM), or 30% of the 24-hr mean level. Restriction or supplementation of dietary Pi induces a substantial decrease or increase, respectively, in serum Pi concentrations during the late morning, afternoon, and evening, but induces less or no change in the morning fasting Pi concentration [10]. To minimize the impact of changes in dietary Pi on the serum Pi concentration, specimens for analysis should be obtained in the morning fasting state. Specimens obtained in the afternoon are more likely to be affected by diet and thus may be more useful to monitor the effect of manipulation of dietary Pi on serum Pi concentrations, as in patients with renal insufficiency receiving phosphorus-binding agents to treat hyperphosphatemia. Other factors can also affect the serum Pi concentration. Presumably because of movement of Pi into cells, the serum Pi concentration can be decreased acutely by intravenous infusion of glucose or insulin, ingestion of carbohydrate rich meals, acute respiratory alkalosis, or infusion or endogenous release of epinephrine. Serum Pi concentration can be increased acutely by metabolic acidosis and by intravenous infusion of calcium [11]. There are substantial effects of age on the fasting serum Pi concentration. In infants in the first 3 months of life, Pi levels are highest (4.8 to 7.4 mg/dl, mean 6.2 mg/dl [2 mM]) and decrease at age 1–2 years to 4.5 to 5.8 mg/dl (mean 5.0 mg/dl [1.6 mM]) [12]. In midchildhood, values range from 3.5 to 5.5 mg/dl (mean 4.4 mg/dl [1.42 mM]) and decrease to adult values by late adolescence [13,14]. In adult males, serum Pi is ∼3.5 mg/dl (1.13 mM) at age 20 years and decreases to ∼3.0 mg/dl (0.97 mM) at age 70 [14,15]. In women, the values are similar to those of men until after the menopause, when they increase slightly from ∼3.4 mg/dl (1.1 mM) at age 50 years to 3.7 mg/dl (1.2 mM) at age 70.
III. INTESTINAL PHOSPHATE ABSORPTION A. Cellular Aspects Dietary Pi is absorbed in the small intestine, with most of the absorption occurring in the duodenum and jejunum and minimal absorption in the ileum.
Pi absorption occurs by a nonsaturable, paracellular pathway as well as by a saturable, energy-requiring Na+-dependent process that has been localized to the mucosal surface. The movement of Pi from the small intestinal lumen into the blood involves transport across the mucosal brush border membrane against an electrochemical gradient, transport through the cytoplasm, and efflux across the serosal membrane [16]. A variety of approaches have been used to investigate Pi absorption by the small intestine. These include in vivo methods where the appearance of labeled Pi in the blood is monitored as a function of time after its oral administration, and in vitro studies where the Pi fluxes across purified mucosal or serosal membrane vesicle preparations are determined. Mucosal brush border membrane Pi transport is saturable, Na+-dependent, and driven by a Na+-gradient (outside > inside) that is maintained by the basolateral membrane–associated Na+,K+-ATPase. The exit of Pi at the serosal (basolateral) surface occurs down an electrochemical gradient and has not been well characterized. However, evidence suggests that the process is carrier-mediated, Na+-independent, and electrogenic. Mucosal Na/Pi cotransport is pH-dependent, with transport activity higher at pH 6 than at pH 7.4.
B. Regulation Pi transport across the mucosal membrane is regulated by 1,25(OH)2D and dietary intake of Pi [16]. Administration of the vitamin D hormone to either vitamin D–deficient or -replete animals elicits a significant increase in net Pi absorption that is associated with a corresponding increase in Na/Pi cotransport Vmax across the mucosal brush border membrane. These findings, combined with the demonstration that Na+-dependent Pi transport across the mucosal brush border membrane correlates with intestinal transepithelial Pi transport, uphold the notion that Pi transport across the mucosal membrane is the rate-limiting step in intestinal Pi absorption. The response of intestinal Pi transport to 1,25(OH)2D is dependent on protein synthesis, occurs several hours after its administration, and apparently occurs after weaning. In a pig model in which mutant animals exhibit defective renal synthesis of 1,25(OH)2D, intestinal brush border membrane Na/Pi cotransport Vmax is similar in newborn mutant and wild-type animals but is significantly reduced in weanling mutants when compared to age-matched wild-type animals [17]. Moreover, while 1,25(OH)2D had no effect on mucosal Pi transport in newborn mutants, administration of 1,25(OH)2D corrected the Pi transport defect in weanling mutants [17].
456 Intestinal Na/Pi cotransport is highly age-dependent and decreases significantly in the first 7 days post partum in both wild-type and mutant piglets [17]. In addition to the more frequently documented genomic mechanisms for 1,25(OH)2D-stimulated intestinal Pi absorption, there is increasing evidence for nonnuclear actions of the vitamin D hormone on duodenal Pi transport. Nemere demonstrated that low levels of 1,25(OH)2D elicit a rapid increase in Pi transport in the perfused duodenal loop of normal chicks [18]. The acute nature of the stimulation, which is apparent in 4–8 min after 1,25(OH)2D administration, suggests a signal transduction pathway independent of genomic regulation. Low-Pi diet also elicits an increase in mucosal Na/Pi cotransport. However, the response to Pi restriction is likely mediated by 1,25(OH)2D. It is well known that the renal synthesis of the vitamin D hormone is stimulated by hypophosphatemic states [8] and that the adaptive intestinal response to Pi restriction is blunted in vitamin D–deficient animals [19].
C. Molecular Mechanisms In an effort to identify intestinal Na/Pi cotransporters by expression cloning, mRNA was isolated from intestine of 1,25(OH)2D-treated rabbits and shown to stimulate Na/Pi cotransport when expressed in Xenopus oocytes [20]. In subsequent studies, a Pi uptake stimulator (PiUS) that increases Na/Pi cotransport in Xenopus oocytes was cloned from rabbit intestine [21]. However, since hydropathy analysis of the 425-amino-acid protein suggested that PiUS lacks transmembrane segments, it was not considered to be a Pi transporter per se. Recently, a full-length cDNA was generated from an EST clone that exhibited sequence homology with the most abundant, renal type II Na/Pi cotransporter [22] (see Section IV,C). When expressed in Xenopus oocytes, the cDNA induced Na/Pi cotransport that was electrogenic, with a pH dependence that resembled that of intestinal brush border membrane Na/Pi cotransport, i.e., higher at pH 6 that at pH 7.4 [22]. Based on its high homology to the renal type II transporter, it was designated type IIb (solute carrier series SLC34A2) and the renal isoform was renamed type IIa. These transporters will be referred to as NPT2a and NPT2b throughout the present chapter. The NPT2b/Npt2b genes map to human and mouse chromosome regions 4p15.2 and 5C1, respectively (Table I). NPT2b mRNA is expressed in a variety of tissues including small intestine, but not in kidney [22]. The demonstration that the NPT2b protein is localized to the apical
HARRIET S. TENENHOUSE AND ANTHONY A. PORTALE
TABLE I Nomenclature and Chromosomal Localization of Human and Mouse Na/Phosphate Cotransporter Genes Common name
Gene namea
NPT1d Npt1d NPT2a Npt2a NPT2b Npt2b NPT2c Npt2c GLVR-1, PIT-1 Glvr-1, Pit-1 RAM-1, PIT-2 Ram-1, Pit-2
SLC17A1 Slc17a1 SLC34A1 Slc34a1 SLC34A2 Slc34a2 NA NA SLC20A1 Slc20a1 SLC20A2 Slc20a2
Human locusb
Mouse locusc
6p22 13A3.2 5q35.2 13B2 4p15.2 5C1 NA 2A3 2q14.1 2F3 8p11.21 8A3
aGene
nomenclature is available at www.gene.ucl.ac.uk/nomenclature. gene loci are available at www.ensembl.org/homo. cMouse gene loci are available at www.ensembl.org/mus. dUpper and lower case refer to human and mouse genes, respectively. NA indicates that neither the gene name nor the gene locus is assigned. bHuman
membrane of enterocytes suggested that it plays a role in the absorption of Pi by the small intestine [22].
D. Molecular Mechanisms Governing NPT2b Regulation Western blotting and immunohistochemical analysis revealed that NPT2b protein expression in the small intestine is regulated by dietary Pi intake [23] (Table II). Chronic Pi deprivation elicits an increase in Na/Pi cotransport in small intestinal brush border membrane vesicles and a corresponding increase in NPT2b protein abundance in the intestinal apical membrane. In contrast, both parameters were decreased by chronically feeding a high-Pi diet and, under these conditions, NPT2b protein was no longer detectable in the intestinal apical membrane [23]. The correlation between apical Na/Pi cotransport and NPT2b protein expression in these studies suggested that NPT2b plays an important role in intestinal Pi absorption and its regulation by dietary Pi intake. Injection of cholecalciferol elicits comparable increases in small intestinal brush border membrane Na/Pi cotransport and apical abundance of NPT2b protein [23] (Table II). However, corresponding increases in intestinal NPT2b mRNA were not evident [23], suggesting that the effect of vitamin D cannot be
457
CHAPTER 26 Phosphate Homeostasis
TABLE II Factor Pi restriction High Pi intake 1,25(OH)2D EGF Glucocorticoids Aging
Factors Involved in the Regulation of Intestinal Pi Absorption and NPT2b Gene Expression Mucosal Na/Pi cotransport
NPT2b protein abundance
NPT2b mRNA abundance
Ref.
ND ND
NC NC NC,
[23] [23] [23,24] [25] [26] [24,26]
NC refers to no change in values. ND indicates that the measurement was not done.
explained by increased NPT2b gene transcription. In another study, however, the stimulation in Na/Pi cotransport elicited by 1,25(OH)2D was associated with comparable increases in both NPT2b mRNA and protein in 14-day-old rats [24]. In addition, in cultured intestinal epithelial cells, 1,25(OH)2D elicited an increase in NPT2b mRNA abundance (Table II) as well as NPT2b promoter activity [24]. Intestinal Na/Pi cotransport and apical NPT2b expression decreases with age in rats, either in the absence or presence of vitamin D [24] (Table II). Small intestinal NPT2b expression is also subject to regulation by epidermal growth factor (EGF) and glucocorticoids (Table II), hormones that are known to influence Pi homeostasis. EGF decreases NPT2b mRNA abundance in rat intestine and human intestinal cell (Caco-2) cultures [25]. EGF also inhibits NPT2b promoter activity in Caco-2 cells, and the latter response is abolished by actinomycin D, confirming transcriptional regulation [25]. The EGF response element in the NPT2b promoter was identified and evidence obtained for its interaction with the c-myb transcription factor [25]. Methylprednisolone, a glucocorticoid analog, also elicits a significant decrease in small intestinal brush border membrane Na/Pi cotransport and a concomitant decrease in NPT2b mRNA and protein expression [26] (Table II). The inhibition by methylprednisolone was evident in mice ranging from 14 days to 9 months of age, although apical transport activity and intestinal NPT2b expression were remarkably higher in the suckling mice than in adult animals [26] (Table II).
IV. RENAL PHOSPHATE TRANSPORT A. Physiology and Tubular Localization Considerable effort has been devoted to the study of Pi transport in the kidney, and much of the material
discussed here is covered in greater detail in review articles (see [27–34]). The proximal tubule is the major site of Pi reabsorption, with approximately 60% of the filtered load reclaimed in the proximal convoluted tubule and approximately 15% in the proximal straight tubule. In addition, a small but variable portion (<10%) of filtered Pi is reabsorbed in more distal segments of the nephron. Clearance studies in humans and experimental animals show that when the filtered load of Pi is progressively increased, Pi reabsorption increases until a maximum tubular reabsorptive rate for Pi, or TmP, is reached, after which Pi excretion increases in proportion to its filtered load. The measurement of TmP varies among individuals and within the same individual, due in part to variation in GFR. Thus the ratio, TmP/GFR, or the maximum tubular reabsorption of Pi per unit volume of GFR, is the most reliable quantitative estimate of the overall tubular Pi reabsorptive capacity [35] and can be considered to reflect the quantity of Na/Pi cotransporters available per unit of kidney mass. The serum Pi concentration at which Pi reabsorption is maximal is called the “theoretical renal Pi threshold”; this value is equal to the ratio, TmP/GFR, and closely approximates the normal fasting serum Pi concentration. Thus, the renal reabsorptive capacity for Pi is the principal determinant of the serum Pi concentration.
B. Cellular Aspects Transepithelial Pi transport is essentially unidirectional and involves uptake across the brush border membrane, translocation across the cell, and efflux at the basolateral membrane (Fig. 2). Pi uptake at the apical cell surface is the rate-limiting step in the overall Pi reabsorptive process and the major site of its regulation. As in the intestine, it is mediated by Na+-dependent
458
HARRIET S. TENENHOUSE AND ANTHONY A. PORTALE
Luminal
3Na+
Basolateral
type IIa
H2PO42−
Na+ type III Pi
xNa+ H2PO42−
type I
Pi A
Cl− / org. anions
−
Pi ?
FIGURE 2
Location of identified and postulated Na+-dependent and Na+-independent Pi transporters in the proximal tubule cell. Available data indicate that most of proximal Pi reabsorption occurs via the type IIa cotransporter (NPT2a), which is localized in the brush border membrane and is the major target for physiologic regulation of renal Pi reabsorption. Org. anions, organic anions; ?, as yet unidentified luminal Pi transporters; A− to the anion that is exchanged with Pi at the basolateral membrane. Reprinted with permission from [190].
Pi transporters that depend on the basolateral membrane-associated Na+,K+-ATPase. Na/Pi cotransport is electrogenic and sensitive to changes in pH, with 10- to 20-fold increases documented when the pH is raised from 6 to 8.5. Little is known about the translocation of Pi across the cell except that Pi anions rapidly equilibrate with intracellular inorganic and organic Pi pools. There are few data regarding the mechanisms involved in the efflux of Pi at the basolateral cell surface. The latter appears to be a passive process that is driven by the electrical gradient existing across the membrane and occurs via an anion exchange mechanism (Fig. 2).
C. Molecular Aspects cDNAs encoding Na/Pi cotransporters, designated type I (NPT1,SLC17A1) and type IIa (NPT2a, SLC34A1; formerly type II, NPT2), have been identified in mammalian kidney and share only 20% identity [36,37]. The NPT1 transporters are approximately 465 amino acids in length with seven to nine membranespanning segments, whereas the NPT2a transporters
comprise approximately 635 amino acids, with eight predicted membrane-spanning segments. The NPT1 and NPT2a transporters account for ∼85 and ∼13%, respectively, of renal Na/Pi cotransporter mRNAs [38]. The chromosomal localization of the NPT1 and NPT2a genes in human and mouse has been determined (Table I) and the human [39] and murine [40] genes cloned and characterized. NPT1 and NPT2a transcripts have been localized to the proximal tubule and NPT1 and NPT2a immunoreactive proteins to the brush border membrane of proximal tubular cells (Fig. 2). Whereas NPT1 is uniformly expressed in all segments of the proximal nephron, NPT2a expression is highest in the S1 segments of the proximal tubule (see [30]). Functional studies of NPT1 and NPT2a have been conducted in cRNA-injected oocytes and in cDNAtransfected renal cell lines (see [30]). Both NPT1 and NPT2a mediate high-affinity Na/Pi cotransport. However, their pH profiles differ significantly, that of NPT2a, and not NPT1, bearing close resemblance to the pH dependence of Na/Pi cotransport in isolated renal brush border membrane vesicles. Moreover, NPT1 exhibits broader substrate specificity than NPT2a and can mediate the transport of Cl− and organic anions. On the basis of these findings it was suggested that NPT1 not only mediates brush border membrane Na+-dependent Pi transport but also serves as an apical channel for Cl− transport and the excretion of anionic xenobiotics. The precise physiological role of NPT1 will thus require further study. Most recently, a Na/Pi cotransporter with homology to NPT2a and NPT2b was identified in rat, human [41], and mouse [42] kidney and designated NPT2c. Although the human and mouse NPT2c/Npt2c genes have been mapped, gene names have not yet been assigned (Table I). NPT2c is approximately one order of magnitude less abundant than NPT2a at the mRNA level, is expressed exclusively in the brush border membrane of proximal tubular cells, and is regulated by dietary Pi [41,42]. The relative abundance of Npt2c protein is significantly higher in kidneys of 22-day-old rats than in those of 60-day-old rats, suggesting that Npt2c is a growth-related renal Na/Pi cotransporter [41]. In addition, hybrid depletion studies suggested that Npt2c accounts for approximately 30% of Na/Pi cotransport in kidneys of Pi-deprived adult mice [42]. Further studies are necessary to determine the precise contribution of renal NPT2c to overall renal Pi reabsorption. The restricted tissue-specific expression of NPT1, NPT2a, and NPT2c is of interest. NPT1 is expressed primarily in kidney but has also been detected in liver and brain. NPT2a mRNA and protein have only been
459
CHAPTER 26 Phosphate Homeostasis
found in kidney and the multinucleated bone-resorbing cell, the osteoclast [43] (see Section VI). To date NPT2c has only been detected in kidney. Two additional Na/Pi cotransporters, designated type III, have been detected in mammalian kidney and account for <1% of renal Na/Pi cotransporter mRNAs. Both are cell surface viral receptors [gibbon ape leukemia virus (Glvr-1, Pit-1, SLC20A1) and murine amphotropic virus (Ram-1, Pit-2, SLC20A2)] that mediate high-affinity, electrogenic Na+-dependent Pi transport when expressed in oocytes and in mammalian cells [33,44]. Although Glvr-1 and Ram-1 show no sequence similarity to NPT1 or NPT2a, they share 60% sequence identity with a putative Pi permease of Neurospora crassa. Chromosomal localization of the two transporters in human and mouse is shown in Table I. Glvr-1 and Ram-1 are widely expressed in mammalian tissues (see Section VI) and may serve as “housekeeping” Na/Pi cotransporters. Both Glvr-1 and Ram-1 proteins are likely expressed in the basolateral membrane of polarized epithelial cells (Fig. 2) where they function to maintain cellular Pi homeostasis.
D. Structure–Function Studies of NPT2a NPT2a-mediated Na/Pi cotransport is electrogenic and involves the inward flux of three Na+ ions and one Pi anion (preferentially divalent) [45]. Binding of the first Na+ ion to the negatively charged carrier is followed by the interaction of two additional Na+ ions and Pi with the carrier [45,46]. The transfer of the fully loaded carrier is electroneutral. Electrogenicity is achieved by the reorientation of the empty carrier after the discharge of the Na+ and Pi ions to the cell interior [46]. Similar to other Na/solute cotransporters, there is a significant Na+ leak after interaction with the first Na+ ion [45]. However, the latter is of minimal physiological significance since the transporter is preferentially in its fully loaded transport cycle in the presence of Pi [45]. There is considerable information on the structures of the NPT2a and NPT2b transporters, and a schematic representation of NPT2a is shown in Fig. 3. Several different analytical approaches were used to arrive at this structure, and these are summarized as follows: (i) hydrophobicity predictions [37]; (ii) antibody accessibility combined with epitope insertion [47]; (iii) cysteine insertion and accessibility of permeant and impermeant sulfhydryl reagents [48,49]; and (iv) glycosylation studies [50]. These studies demonstrated several important features of NPT2a (Fig. 3): The transporter has a large extracellular loop that separates it into two domains. There is intramolecular homology
within the ICL1 and ECL3 domains. Both the NH2 and COOH termini are oriented intracellularly. The studies just described also addressed the functional significance of specific domains of NPT2a (Fig. 3). Cysteine-insertion studies suggested that ICL1 and ECL3 form an important part of a “permeation pore” that participates in both “cotransport” and “Na+ leak” function [48,49]. Chimera construction—based on different transport properties of NPT2a and NPT2b [22,37]— suggested the involvement of three amino acid residues in determining the pH dependence of NPT2a, namely increased transport at higher pH [51,52]. The chimera approach also suggested the importance of two basic amino acid residues in ICL3 for PTH-dependent internalization [53] (see Section IV,E,1). Deletion studies documented that the COOH terminus contains information for brush border membrane expression, i.e., a terminal PDZ-binding motif and a membrane internalization signal [54]. Both the NH2 and COOH termini portions of NPT2a are required for transport activity. However, functional studies revealed that cleavage of the NPT2a protein backbone, between the two glycosylation sites in the large extracellular loop, does not interfere with transport function [55,56]. It is assumed that under this condition a disulfide bridge within this large extracellular loop stabilizes the transporter (Fig. 3). Although NPT2a might be part of a multimeric complex (see later discussion), one transporter unit is sufficient to mediate Na/Pi cotransport [57].
E. Regulation of NPT2a Regulation of renal Pi reabsorption has been the subject of intense investigation (see Table III and [58–78]). The regulation is achieved primarily by alterations in the abundance of NPT2a protein in the brush border membrane of proximal tubular cells and occurs largely in the absence of changes in NPT2a mRNA (for review, see [30]). Changes in the amount of apical NPT2a protein are accomplished by either membrane insertion or retrieval/lysosomal degradation of the transporter. Although membrane trafficking of NPT2a is a key element in the regulation of renal Pi handling, little is known about the molecular signals that confer specificity to this process, i.e., why other transporters are not inserted into or retrieved from the membrane along with NPT2a. Several proteins, such as PDZK1 (NaPi-Cap1, diphor-1), NaPi-Cap2, and NHERF-1 [79], which interact with NPT2a, have been identified and likely play a key role in NPT2a trafficking in response to factors that stimulate or inhibit renal Pi transport [80].
460
HARRIET S. TENENHOUSE AND ANTHONY A. PORTALE
s s
ECL-3
ICL-1
ICL-3
NH2
TRL-COOH Glycosylation sites Na/Pi cotransport/Na+ slippage
PDZ interactions
α-helical pH-dependence PTH regulation/internalization Scaffolding/positioning
FIGURE 3 Structure–function relationship of the renal NPT2a Na/Pi cotransporter. See text for details. Reprinted with permission from [32].
1. PARATHYROID HORMONE
PTH is the major hormonal regulator of renal Pi reabsorption. PTH acts directly on proximal tubular cells and inhibits both apical and basolateral Na/Pi cotransport (Table III) by mechanisms that involve the internalization of cell surface NPT2a protein [60] and its subsequent lysosomal degradation [81]. Considerable progress has been made in the identification of the signaling pathways that mediate the PTH response and the proteins involved in the retrieval of NPT2a from the plasma membrane.
PTH binding to receptors on the basolateral membrane activates protein kinase A (PKA) and/or protein kinase C (PKC) signaling pathways, whereas its binding to apical receptors activates PKC [82]. The extracellular signal-regulated kinase/mitogen-activated protein kinase (ERK/MAPK) pathway also participates in PTH-induced signaling [83], and recent studies show that the PKA and PKC signaling pathways converge on the ERK/ MAPK pathway to internalize NPT2a protein [84]. Although the downstream targets for ERK/MAPKmediated phosphorylation remain unknown, changes
461
CHAPTER 26 Phosphate Homeostasis
in the phosphorylation state of NPT2a are not associated with its PTH-induced internalization [85]. Rather, it has been postulated that the phosphorylation of proteins that associate with NPT2a may determine its regulation. AKAP79, an A kinase anchoring protein [86], and RAP, a receptor-associated protein [87], have been shown to participate in the PTH-mediated retrieval of NPT2a from the plasma membrane of proximal tubular cells. In opossum kidney (OK) cells, AKAP79 associates with NPT2a and the regulatory and catalytic subunits of PKA, and this process is necessary for PKA-dependent inhibition of Na/Pi cotransport [86]. In RAP-deficient mice, PTH-induced internalization of NPT2a is significantly delayed, whereas regulation by dietary Pi (see Section IV,E,3) is not affected [87]. In contrast, NHERF-1, which associates with NPT2a and MAP-17 in mouse kidney [88], does not appear to play a role in PTH-induced internalization of NPT2a [89]. 2. OTHER HORMONAL REGULATORS
Other hormones also contribute to the regulation of proximal tubular Pi transport (Table III). Growth hormone, insulin-like growth factor-I, insulin, 1,25(OH)2D, and thyroid hormone all stimulate Pi reabsorption,
TABLE III Factor Pi restriction High Pi intake PTH IGF-1 Insulin 1,25(OH)2D T3 PTHrP ANF EGF TGF-α and -β Glucocorticoids Stanniocalcin FGF-23 sFRP-4 Aging
whereas PTH-related peptide, calcitonin, atrial natriuretic factor, epidermal growth factor, transforming growth factor-α, and glucocorticoids inhibit Pi reclamation (for review, see [30]). The increase in brush border membrane Na/Pi cotransport by thyroid hormone is associated with an increase in NPT2a mRNA [66], whereas both dexamethasone [73] and epidermal growth factor [69] decrease NPT2a mRNA abundance. Because thyroid hormone and glucocorticoids exert their effects on gene transcription via ligandactivated receptors that bind to response elements in the promoter region of target genes, it was suggested that these hormones exert their effects by modulating NPT2a gene transcription. However, neither thyroid hormone nor dexamethasone has an effect on NPT2a promoter–reporter gene expression (see [90]), suggesting that transcriptional mechanisms are not involved. Moreover, recent studies have been unable to detect a thyroid hormone–mediated increase in NPT2a mRNA (see [90]). Administration of 1,25(OH)2D to vitamin D–deficient rats elicits an increase in brush border membrane Na/Pi cotransport that is accompanied by an increase in renal NPT2a mRNA and protein abundance (Table III) [64]. Although these results are consistent with direct
Factors Involved in the Regulation of Renal Na/Pi Cotransport and NPT2a Gene Expression Na/Pi cotransporta
NPT2a gene expressionb
Ref.
ND ND ND ND ND ND ND
[58,59] [58,59] [60,61] [62] [63] [64] [65,66] [67] [68] [69,70] [71,72] [73] [74] [75,77] [76] [66,78]
aNa/P cotransport studies were performed in cultured proximal tubular cells or in renal brush border membrane vesicles i prepared from appropriate animal models. bRefers to changes in renal Npt2a mRNA or protein abundance. ND indicates that the measurement was not done. For other abbreviations, see text.
462 effects of 1,25(OH)2D on NPT2a-mediated renal Na/Pi cotransport, the effects may result from a 1,25(OH)2Ddependent decrease in PTH levels. However, the finding that 1,25(OH)2D increased the activity of a NPT2a promoter–luciferase reporter gene construct suggests a direct effect of the vitamin D hormone on NPT2a gene transcription [64]. Other hormones that regulate renal Pi handling include stanniocalcin, 5-hydroxytryptamine (5-HT), fibroblast growth factor (FGF)-23, and secreted frizzledrelated protein (sFRP)-4 (Table III). Stanniocalcin is a peptide hormone that counteracts hypercalcemia and stimulates Pi reabsorption in bony fish, and is also produced by humans. Infusion of stanniocalcin in rats stimulates renal Pi reabsorption and brush border membrane Na/Pi cotransport [74], suggesting a role for stanniocalcin in the maintenance of Pi homeostasis in mammals as well as fish. 5-HT is synthesized in the kidney and locally generated 5-HT was shown to interfere with PTH-mediated inhibition of renal Na/Pi cotransport [91]. These findings suggested that 5-HT is a paracrine modulator of renal Pi transport. FGF-23 is a secreted peptide contributing to renal Pi wasting in autosomal dominant hypophosphatemic rickets and oncogenic hypophosphatemic osteomalacia (see Sections VII,B and C). Administration of FGF-23 elicits a decrease in serum Pi in mice that is attributed to renal Pi wasting [92], a reduction in brush border membrane Na/Pi cotransport [75], and decreased renal Npt2a expression (Table III) [77]. In addition, rats receiving intrahepatic injection of FGF-23 cDNA develop hypophosphatemia and a significant decrease in both brush border membrane Na/Pi cotransport and NPT2c protein abundance [93]. FGF-23-dependent inhibition of NPT2a-mediated Na/Pi cotransport in OK cells occurs by activation of MAP kinase [94]. sFRP-4, which, like FGF-23, is highly expressed in tumors from patients with oncogenic hypophosphatemic osteomalacia [76] (see Section VII,C), has also been tested for its phosphaturic action. sFRP-4 induced a specific increase in the renal fractional excretion of Pi and hypophosphatemia when infused in rats and inhibited Na/Pi cotransport in vitro when added to OK cells (Table III) [76]. 3. REGULATION BY DIETARY PHOSPHATE
Dietary Pi intake is a key determinant of renal Pi handling. Pi deprivation elicits an increase in Pi reabsorption and brush border membrane Na/Pi cotransport. Both acute and chronic exposure to low dietary Pi induce an increase in transport Vmax as well as in NPT2a protein but not NPT2a mRNA [59] (Table III). These findings are consistent with recent data demonstrating
HARRIET S. TENENHOUSE AND ANTHONY A. PORTALE
that low dietary Pi has no effect on NPT2a promoter activity [90]. The acute response to Pi deprivation is mediated by microtubule-dependent recruitment of existing NPT2a protein to the apical membrane [95]. In contrast, exposure to high dietary Pi leads to the internalization of cell surface NPT2a protein into the endosomal compartment by a microtubule-independent mechanism [95]. Internalized NPT2a protein is then delivered to the lysosome, by a microtubuledependent process, for degradation [59]. A Pi response element (PRE) was identified in the mouse NPT2a promoter by DNA footprint analysis [96]. The PRE was shown to bind a mouse transcription factor, TFE3, and the expression of TFE3 is increased in the kidney in response to Pi deprivation [96]. On the basis of these results, it was suggested that TFE3 participates in the transcriptional regulation of the NPT2a gene by dietary Pi. Further work is necessary to assess the physiological significance of these findings.
F. Lessons from Mouse Models with Renal Phosphate Transport Defects 1. THE NPT2a KNOCKOUT MOUSE
The critical role of NPT2a in the maintenance of Pi homeostasis was clearly demonstrated in mice in which the Npt2a gene (lowercase refers to the mouse gene) was knocked out by targeted mutagenesis [97]. Mice homozygous for the disrupted gene (Npt2−/−) exhibit decreased renal Pi reabsorption, an ∼80% loss of brush border membrane Na/Pi cotransport, hypophosphatemia, an appropriate adaptive increase in the renal synthesis [98] and serum concentration of 1,25(OH)2D [97,98] and associated hypercalcemia, hypercalciuria, and hypoparathyroidism, and an agedependent skeletal phenotype [97,99]. Of interest was the demonstration that renal brush border membrane Na/Pi cotransport is not increased in Npt2−/− mice maintained on a low-Pi diet [100] and is not decreased in Npt2−/− mice following PTH administration [101]. These findings demonstrate unequivocally that NPT2a is the target for regulation of renal Pi handling by PTH and dietary Pi. Recent studies show that Npt2c, but not Npt1, is upregulated in Npt2−/− mice and that the 2.8-fold increase in renal brush border membrane Npt2c protein in the mutants is not accompanied by an increase in Npt2c mRNA [102]. The failure of Npt2−/− mice to respond to Pi restriction with a further increase in Npt2c protein suggests that Npt2c protein is maximally up-regulated and likely accounts for residual brush border membrane Na/Pi cotransport in Npt2−/− mice [102].
463
CHAPTER 26 Phosphate Homeostasis
2. THE X-LINKED HYP AND GY MOUSE MODELS
Hyp and Gy are spontaneous and radiation-induced mutations, respectively, on the mouse X chromosome, and mice harboring these mutations have served as valuable models to elucidate the pathophysiology of the homologous human disorder, X-linked hypophosphatemia (XLH) (Table IV) [103]. Mutations in Phex/ PHEX, a Pi regulating gene with homologies to endopeptidases on the X chromosome, have been identified in Hyp [104] and Gy [105] mice and XLH patients [106]. Like patients with XLH, Hyp and Gy mice are characterized by growth retardation, rickets and osteomalacia, hypophosphatemia, and renal defects in Pi transport and the regulation of vitamin D metabolism [103] (see Section V,C). In Hyp and Gy mice, the increased urinary Pi excretion can be attributed to a 50% decrease in renal brush border membrane Na/Pi cotransport Vmax [107,108] and a comparable reduction in the renal expression of Npt2a mRNA and immunoreactive protein [109–111]. In addition, renal expression of both Npt1 and Npt2c mRNA and protein is significantly decreased in Hyp mice when compared to that in normal littermates [102]. These data suggest that loss of Phex function is responsible for the down-regulation of all three renal Na/Pi cotransporters in the mutant strain. The Pi transport defect in Hyp mice is not intrinsic to the kidney but rather is dependent on a circulating factor, and this conclusion is based on several findings. Parabiosis of normal mice to Hyp mice leads to the development of hypophosphatemia, reduced renal Pi reabsorption, and a PTH-independent decrease in renal brush-border membrane Na/Pi cotransport in the normal mouse of the normal–Hyp pair [112,113]. Furthermore, renal cross-transplantation revealed that hypophosphatemia and renal Pi wasting persist in Hyp mice that
TABLE IV Disorder XLH Hyp Gy Ska1 ADHR OHO HHRH Npt2−/−
Serum Pi
have received a normal kidney, whereas normal serum Pi and normal renal Pi handling persist in normal mice that have received a Hyp kidney [114]. In further support for this hypothesis, serum from Hyp mice but not normal mice inhibited Na/Pi cotransport in primary mouse renal cell cultures [115]. Inhibition of Na/Pi cotransport in target renal cells was also elicited with conditioned medium derived from cultured Hyp osteoblasts, whereas no inhibition was apparent with conditioned medium from cultured normal osteoblasts [115]. The latter data suggest that the Hyp osteoblast is responsible for the release and/or modification of a humoral factor that accumulates in the absence of Phex function and inhibits Na/Pi cotransport in renal epithelial cells. A likely candidate for this factor is FGF-23 (see Section VII). Of interest is the recent report of a point mutation in the mouse Phex gene that was introduced by random mutagenesis with ethyl-nitrosourea [116]. The mutation, designated Ska1, causes skipping of exon 8 and results in clinical and biochemical phenotypes that are similar to those in Hyp mice and XLH patients, including hypophosphatemia, rickets, and osteomalacia (Table IV) [116].
V. ROLE OF PHOSPHATE IN THE REGULATION OF RENAL VITAMIN D METABOLISM A. Lessons from Mouse Models Dietary Pi intake and serum Pi concentration are critically important determinants of the renal metabolism of 1,25(OH)2D. Tanaka and DeLuca first demonstrated that rats maintained on a low-Pi diet
Features of Renal Phosphate Wasting Disorders in Humans and Mice Serum 1,25(OH)2D NC NC NC ND NC NC
Urine Ca/creatinine
Primary defect
Ref.
NC NC NC ND NC NC
PHEX mutation 3′Phex deletion 5′Phex deletion Phex mutation FGF23 mutation Tumor-mediated Unknown Npt2 deletion
[106] [104] [105] [116] [168] [188] [160] [97]
, an increase in values; , a decrease in values; NC, no change in values; ND, measurement not done.
464
HARRIET S. TENENHOUSE AND ANTHONY A. PORTALE
Studies in intact mice reveal that restriction of dietary Pi induces a rapid, sustained, approximately six- to eightfold increase in renal mitochondrial 1α-hydroxylase activity (Fig. 4A, C), which can be attributed to an approximately sevenfold increase in Vmax of the enzyme, with no measurable change in its Km [130]. The increase in 1α-hydroxylase activity can be attributed to an increase in the renal abundance of 1α-hydroxylase (P450c1α, CYP27B1) mRNA (Fig. 4B, D) and protein [130,131]. Immunohistochemical analysis demonstrates that the increase in P450c1α gene expression occurs exclusively in the proximal renal tubule and is due, at least in part, to increased transcription of the P450c1α gene [130]. Degradation of 1,25(OH)2D to calcitroic acid and cholacalcioic acid via the 24-oxidation pathway is initiated by the enzyme 25-hydroxyvitamin D-24-hydroxylase (24-hydroxylase, P450c24, CYP24). Pi restriction induces a significant decrease in renal 24-hydroxylase activity [132] and mRNA abundance [130,133].
synthesized predominately 1,25(OH)2D, and that this effect was not diminished by parathyroidectomy [6]. They further observed an inverse relationship between the synthesis of 1,25(OH)2D and the serum Pi concentration. Numerous studies in animals demonstrate that hypophosphatemia induced by dietary Pi restriction induces an increase in the production rate of 1,25(OH)2D measured in vivo [117,118] and in 1α-hydroxylase activity measured in vitro [7,8,119]. Studies in healthy human subjects are consistent with those in experimental animals: Restriction of dietary Pi induces an increase [10,120–123] and supplementation of dietary Pi [10,123,124] a decrease in the serum concentration of 1,25(OH)2D and in its in vivo production rate [123]. It is of interest that the stimulation of 1,25(OH)2D production by Pi depletion is abolished by hypophysectomy and is restored by administration of growth hormone or insulin-like growth factor I (IGF-1) [125–129], indicating that an intact growth hormone/IGF-1 axis is required for the effect; however, the molecular mechanisms are unknown.
40 # 20
0.02%
0.6%
1α-Hydroxylase activity (pg/mg/15 min)
125 100 75 50
#
25 # 0
FIGURE 4
WT Npt2−/−
800 600
#
400 200 0
1.6%
#
WT Hyp
1000 P450c1α mRNA abundance (% control)
1α-Hydroxylase activity (pg/mg/15 min)
WT Npt2−/−
0
C
B
60
0.02%
0.6%
1.6%
D P450c1α mRNA abundance (% control)
A
#
WT Hyp
600
400 #
200 # 0
0.02%
1% Dietary phosphorus
1.6%
0.02%
1% Dietary phosphorus
1.6%
Effects of dietary Pi on renal mitochondrial 1α-hydroxylase activity (A, C) and renal P450c1α mRNA abundance (B, D) in wildtype mice (white bars) and either Npt2−/− (A, B) or Hyp (C, D) mice (hatched bars). Mice were fed the low (0.02%), control (0.6–1%), or high (1.6%) Pi diet for 4-5 days. Renal mitochondria were prepared and incubated with 25-hydroxyvitamin D3 to determine 1α-hydroxylase activity, and renal total RNA was prepared to estimate the abundance of P450c1α mRNA, relative to β-actin mRNA, by ribonuclease protection assay. Bars depict mean ± SEM. *, Compared with the 0.6% or 1% Pi diet, within each species, P < 0.05. #, Compared with wild-type animals, within each diet group, P < 0.05.
CHAPTER 26 Phosphate Homeostasis
B. The Npt2a Knockout Mouse Mice homozygous for the disrupted Na/Pi cotransporter gene (Npt2−/−) have been used to address the question of the relationship between renal reabsorption of Pi, per se, and regulation of P450c1α gene expression. It has been postulated that normal transepithelial transport of Pi in the proximal renal tubule is essential for the regulation of renal 1,25(OH)2D synthesis by Pi. Beck et al. reported that renal Pi wasting and hypophosphatemia in Npt2−/− mice is associated with an increase in serum 1,25(OH)2D concentration [97]. The increase in serum 1,25(OH)2D in Npt2−/− mice can be attributed to an increase in renal 1α-hydroxylase activity and P450c1α mRNA abundance (Fig. 4A, B) [98]. Thus, renal 1α-hydroxylase activity and mRNA expression in Npt2−/− mice increase appropriately in response to the hypophosphatemia that results from renal Pi wasting. Moreover, restriction or supplementation of dietary Pi in Npt2−/− mice elicits a further increase or a decrease, respectively, in serum 1,25(OH)2D concentrations, in renal mitochondrial 1α-hydroxylase activity and in P450c1α mRNA abundance (Fig. 4A, B) [98]. In both wild-type and Npt2−/− mice, values of P450c1α mRNA abundance vary inversely and significantly with serum Pi concentrations. These studies provide unequivocal evidence that normal renal Na/Pi cotransport is not necessary for the regulation of renal 1,25(OH)2D production by Pi, and that renal Pi wasting in itself does not contribute to dysregulated 1,25(OH)2D production, thereby dispelling a previously held notion. These findings suggest that changes in serum Pi concentration per se are sufficient to initiate the signaling pathways involved in the up-regulation and down-regulation of the P4501α gene by manipulation of dietary Pi.
C. The X-Linked Hyp Mouse In contrast to the response to Pi depletion in wildtype and Npt2−/− mice, the hypophosphatemia that results from renal Pi wasting in XLH is attended by abnormal regulation of vitamin D metabolism. In patients with XLH, serum concentrations of 1,25(OH)2D are nominally within the normal range [134,135]; however, such values are inappropriately low given the attendant hypophosphatemia, and serum 1,25(OH)2D concentrations fail to increase with dietary Pi restriction [121]. Regulation of vitamin D metabolism is similarly disordered in the Hyp mouse model of XLH. In Hyp mice, the serum concentration of 1,25(OH)2D is not different from that in normal littermates and, therefore, is inappropriately low for the degree of hypophosphatemia [136,137].
465 Renal 1α-hydroxylase activity also is inappropriately low for the degree of hypophosphatemia [119], and Pi restriction induces a paradoxical decrease in both serum concentration of 1,25(OH)2D [136,137] and renal 1α-hydroxylase activity [138]. The molecular basis for the disordered regulation of vitamin D metabolism in Hyp mice was explored in wild-type and Hyp mice subjected to manipulation of dietary Pi. In wild-type mice a low-Pi diet elicits sixfold and threefold increases in renal 1α-hydroxylase activity and P450c1α mRNA expression, respectively, whereas in the Hyp strain the diet-induced changes were of similar magnitude but opposite in direction (Fig. 4C, D) [139]. A high-Pi diet was without effect in wild-type mice, whereas in Hyp mice the same diet induced threefold and twofold increases, respectively, in enzyme activity and P450c1α mRNA abundance (Fig. 4C, D) [139]. In wild-type mice, both renal 1α-hydroxylase activity and P450c1α mRNA abundance varied inversely and significantly with serum Pi concentrations, whereas in Hyp mice the relationship between both renal parameters and serum Pi concentration was direct [139]. The regulation of 24-hydroxylase also is abnormal in Hyp mice, with renal enzyme activity being twofold higher in Hyp than in wild-type littermates and paradoxically increasing further with Pi restriction [137]. In Hyp mice, Pi restriction induced a significant increase in renal P450c24 mRNA abundance [139–141], in contrast to the decrease observed in wild-type mice [139]. These findings suggest that the decrease in the serum concentration of 1,25(OH)2D in phosphorus-restricted Hyp mice can be attributed to an increase in its renal catabolism as well as to a reduction in its renal synthesis [139]. Nevertheless, current data demonstrate that regulation of both the P450c1α and P45024 genes by Pi is disordered in Hyp mice at the level of renal 1α-hydroxylase activity and renal P450c1α and P450c24 mRNA expression and suggest that loss of Phex function in Hyp mice gives rise to disordered transcriptional regulation of both the P450c1α and P450c24 genes. The extent to which posttranscriptional or posttranslational mechanisms contribute to regulation of expression of these genes remains to be determined.
VI. PHOSPHATE TRANSPORT IN BONE Pi is essential for the mineralization of extracellular matrix by osteogenic cells and is a vital element for osteogenic cell function. Thus, a considerable effort has been devoted to the characterization of Pi transport in osteoblasts and chondrocytes, the cells that produce
466 and secrete the extracellular matrix. Pi is also crucial for energy metabolism in osteoclasts, multinucleated cells involved in bone resorption. Recent progress on Pi transport in osteoblasts and osteoclasts is briefly described below (for review, see [142]).
A. Osteoblasts and Matrix Vesicles Studies of Pi transport in bone have primarily employed osteoblast cell lines, such as human osteosarcoma–derived SaOS-2 cells, rat osteosarcoma– derived UMR-106 and ROS 17/2.8 cells, and murine and rat calvarial–derived MC3T3-E1 and PyMS cells. Since these cell lines may differ, for example in their state of differentiation, the extent to which data generated from such in vitro studies reflects the in vivo situation remains to be determined. Na/Pi cotransport in cultured osteoblasts differs from that in renal proximal tubular cells in several respects [143]. Km values for Pi are higher in osteoblasts (300–500 µM vs 50–100 µM) and transport in osteoblasts is greater at acidic pH [143]. Moreover, several factors that inhibit Na/Pi cotransport in renal proximal tubular cells (Table III) stimulate transport in osteoblasts (Table V). In both cell types, Na/Pi cotransport is regulated by the extracellular Pi concentration, with transport increasing significantly as the Pi concentration in the culture medium is reduced [144] (Tables III and V). Na/Pi cotransport in osteoblasts is also stimulated by a variety of hormones and agents [145–154] (see Table V), which are of interest because of their potential therapeutic applications. The mechanisms involved in the stimulation of Na/Pi cotransport by many of these agents have been investigated. The actions of PTH and PTHrP are dependent on the cAMP-PKA pathway, whereas the stimulation by IGF-1, PDGF, bFGF, and fluoride is dependent on tyrosine phosphorylation, and the action of epinephrine, endothelin-1, and prostaglandin E2 is associated with the activation of protein kinase C– mediated phosphorylation. GLVR-1(PIT-1), the ubiquitous high-affinity type III Na/Pi cotransporter (see Section IV,C), is expressed in SaOS-2 cells [147]. The increase in Na/Pi cotransport elicited by IGF-1 in SaOS-2 cells is associated with a comparable increase in the relative abundance of GLVR-1 mRNA [147]. Furthermore, in cultured MC3T3-E1 cells, Glvr-1/Pit-1 mRNA abundance increased during cell differentiation and correlated with increases in osteocalcin mRNA and mineralization [155]. Although these findings suggest that Glvr1/Pit-1 mRNA may be used as a marker for osteoblast differentiation, the relative contribution of Glvr-1/Pit-1 to Pi transport in osteoblasts and the role of as yet
HARRIET S. TENENHOUSE AND ANTHONY A. PORTALE
TABLE V Factors Stimulating Na/Pi Cotransport in Cultured Osteoblasts Factor Pi restriction PTH IGF-1 1,25 (OH)2D PTHrPa TGF-αa PDGFa bFGFa Epinephrine Endothelin-1 Prostaglandin E2 Fluoride
Cell line
Ref.
MC3T3-E1 UMR-106 SaOS-2 PyMS UMR-106 UMR-106 MC3T3-E1 MC3T3-E1 MC3T3-E1 MC3T3-E1 PyMS UMR-106
[142,144] [61,145] [147] [148] [146] [146] [149] [150] [151] [152] [153] [154]
aPTHrP, parathyroid hormone related peptide; TGF-α, transforming growth factor-α; PDGF, platelet-derived growth factor; bFGF, basic fibroblast growth factor.
unidentified Na/Pi cotransporters in the osteoblast remain to be determined. Pi transport has also been examined in matrix vesicles, which are structures that derive from osteoblasts by budding from elongated tubular extensions that project from the plasma membrane of these cells [142]. Pi transport in matrix vesicles is Na+ dependent and reflects the transport activity of the cells from which they originated. It has been proposed that Pi accumulation inside the matrix vesicles is involved in the formation of nucleation sites that are necessary for initiation of the calcification process [142].
B. Osteoclasts Osteoclasts exhibit Na/Pi cotransport activity that is stimulated by the addition of bone particles to the culture medium and is inhibited by peptides with the RGDS (arg-gly-asp-ser) motif, suggesting that integrins and cell–matrix interactions are involved in the regulation of the transport process [156]. There is evidence that an NPT2a-like Na/Pi cotransporter that is present in osteoclasts is localized to discrete vesicles in unpolarized osteoclasts and to the basolateral membrane in actively resorbing cells [157]. Furthermore, the addition of phosphonoformic acid to osteoclast cultures led to inhibition of Na/Pi cotransport, diminished ATP production, and decreased bone resorption [157]. Taken together, these findings suggest that Na/Pi cotransport in osteoclasts is essential for energy metabolism and osteoclast-mediated bone resorption.
CHAPTER 26 Phosphate Homeostasis
Moreover, the data suggest that some of the Pi released from bone during the resorptive process might be utilized by the osteoclast to maintain cellular ATP content during the cylical processes of migration, attachment, and resorption. In light of the finding that NPT2a is expressed in the osteoclast, it was of interest to characterize the skeletal phenotype in mice homozygous for the disrupted Npt2a gene (see Section IV,F,1). Indeed, histomorphometric analysis revealed that Npt2−/− mice exhibit an age-dependent skeletal phenotype that is attributable not only to hypophosphatemia, secondary to the renal defect in brush border membrane Na/Pi cotransport, but also to an intrinsic osteoclast defect [99]. In addition to Npt2a, Pit-1 is expressed in the mouse osteoclast [43]. Furthermore, Npt2a and Pit-1 appear to associate with NHERF-1, ezrin, and actin, suggesting that such interactions play an important role in membrane sorting and regulation of Npt2a- and Pit-1mediated Na/Pi cotransport in the osteoclast [43].
VII. DISORDERS CAUSING PHOSPHATE DEFICIENCY Since Pi is sufficiently abundant in the diet, Pi deficiency is unlikely to develop except under unusual circumstances (for reviews, see [158,159]). The following review will be limited to well-characterized hypophosphatemic disorders associated with renal Pi wasting in humans. Three of the disorders, X-linked hypophosphatemia (XLH), autosomal dominant hypophosphatemic rickets (ADHR), and hereditary hypophosphatemic rickets with hypercalciuria (HHRH), are inherited, and the fourth, oncogenic hypophosphatemic osteomalacia (OHO), is acquired. Each is characterized by rickets and osteomalacia and, with the exception of HHRH, by abnormal regulation of renal vitamin D metabolism (Table IV). The discussion will focus primarily on the renal Pi transport phenotype and the underlying pathophysiology. For a more detailed review of these disorders, the reader is referred to Chapters 69 and 70 in this volume.
A. X-Linked Hypophosphatemia (XLH) The features that distinguish XLH from the other renal Pi wasting disorders are its X-linked dominant mode of inheritance, its higher prevalence (1 in 20,000), and the availability of two murine homologs (see Section IV,F,2), which serve as valuable models to study the pathophysiology of XLH [103,160]. The gene responsible for XLH [106] was identified and designated PHEX (formerly PEX) to depict a phosphate-regulating gene
467 with homology to endopeptidases on the X chromosome (Table IV). However, the mechanism whereby loss of PHEX function leads to renal Pi wasting and hypophosphatemia is not understood. The PHEX gene encodes a 749-amino-acid protein [106] that exhibits significant homology to the M13 family of zinc metallopeptidases, which are type II membrane glycoproteins, with a large extracellular domain containing 10 highly conserved cysteine residues and a zinc-binding motif, essential for conformational integrity and catalytic activity, respectively [161]. The 171 mutations in the PHEX gene identified in XLH patients, which are cataloged in a locus-specific database [162] available online (www.phexdb.mcgill.ca), are predicted to result in loss of function of the PHEX protein. Some disease-causing missense mutations in the PHEX gene give rise to proteins that remain trapped in the endoplasmic reticulum, where they are degraded [163,164]. In some cases, however, the trapped PHEX protein can be rescued from the endoplasmic reticulum [163] and exhibit full catalytic activity [164]. These findings provide a mechanism for loss of PHEX function and a basis for the development of novel therapeutic approaches for patients carrying such trafficking mutations. PHEX is expressed predominantly in osteoblasts, osteocytes, and odontoblasts, but not in kidney (see [103]). Although it is not immediately apparent how loss of PHEX function results in a decrease in renal Pi reabsorption, it was suggested that PHEX is involved in the inactivation of phosphaturic factors or the activation of a Pi-conserving hormone [104,106,160,165]. In either case, renal Na/Pi cotransporters would be downregulated and Pi wasting would ensue. Although recombinant PHEX can cleave PTHrP107-139 [166] and peptides derived from MEPE (matrix extracellular phosphoglycoprotein) and FGF-23 [167], the precise function of these PHEX substrates in vivo and their role in the regulation of brush border membrane Na/Pi cotransporters and renal Pi handling remain to be established. Patients with XLH are treated with oral Pi supplements and 1,25(OH)2D [160]. Since this form of therapy is often associated with undesirable side effects, the aim of future studies is to develop novel therapeutic approaches based on knowledge of the mechanism of action of the PHEX gene product.
B. Autosomal Dominant Hypophosphatemic Rickets (ADHR) ADHR is far less common than XLH, it exhibits male-to-male transmission consistent with autosomal dominant inheritance, and it is characterized by incomplete penetrance and variable age of onset [160].
468 The gene responsible for ADHR encodes a new member of the fibroblast growth factor (FGF) family, FGF23, a secreted peptide that is processed to amino- and carboxy-terminal peptides at a consensus proprotein convertase (furin) site, RHTR (ArgHisThrArg) [168] (Table IV). Missense mutations involving the two R residues in this proteolytic cleavage site have been identified in patients with ADHR [168] and were shown to abrogate FGF-23 processing [169–171]. FGF-23 induces hypophosphatemia and renal Pi wasting in vivo [77,92]. Moreover, the increase in renal Pi excretion elicited by the hydrolysis-resistant mutant FGF-23 is associated with decreased brush border membrane Na/Pi cotransport and decreased brush border membrane abundance of Npt2a [77] and Npt2c [93] proteins (see Section IV,E,2). Furthermore, the phosphaturic action of hydrolysis-resistant mutant FGF-23 is significantly greater than that of wild-type FGF-23 [77], and the amino- and carboxy-terminal processed FGF-23 fragments do not induce hypophosphatemia in vivo [171]. Although the latter studies attribute the phosphaturic action of FGF-23 only to the intact protein, differences in the pharmacokinetics of the intact and processed FGF-23 peptides may have contributed to these findings [171]. Regardless, the findings to date provide a mechanism for hypophosphatemia and renal Pi wasting in ADHR. The mechanisms through which FGF-23 affects Pi homeostasis remain unclear, since conflicting data are reported regarding its action on Na/Pi cotransport in OK cells [92,160,169]. In addition, the site of FGF-23 production is unknown, although it is abundantly expressed in tumors removed from patients with OHO, an acquired renal Pi wasting disorder with features of XLH and ADHR [92,169,172] (see Section VII,C). Therapy for ADHR is similar to that of XLH (see Section VII,A). However, once the mechanisms of FGF-23 processing, action, and degradation are well understood, novel therapeutic approaches will likely be developed.
C. Oncogenic Hypophosphatemic Osteomalacia (OHO) Also known as tumor-induced osteomalacia, OHO is an acquired and rare form of renal Pi wasting, with clinical and biochemical features of XLH and ADHR (Table IV). Patients with OHO also exhibit muscle weakness, fatigue, and fractures. Although several types of tumors are associated with this syndrome, the majority appear to arise from mesenchymal elements [173]. Evidence suggests that the phenotypic features of OHO result from a humoral phosphaturic factor, designated
HARRIET S. TENENHOUSE AND ANTHONY A. PORTALE
phosphatonin [174], that is secreted by the tumor, since removal of the tumor leads to correction of the clinical and biochemical phenotype. Consistent with this hypothesis is the demonstration that extracts from OHO tumors inhibit Na/phosphate cotransport in OK cells [175]. The analysis of abundantly and differentially expressed genes in OHO tumors led to the identification of phosphatonin candidates, including FGF-23 [92,169,172,176–178], sFRP-4 [76,178] and MEPE [176,179]. As discussed earlier (Section VII,B), patients with ADHR harbor mutations in FGF-23 that prevent its processing and degradation [169–171], and both wildtype and mutant forms of FGF-23 induce hypophosphatemia and renal Pi wasting in vivo [77,92,171]. Moreover, sFRP-4 decreases serum Pi levels and inhibits renal Pi reabsorption in rats and mice [76]. Thus, FGF-23 and sFRP-4 likely contribute to the pathogenesis of OHO. However, the role of MEPE in this disorder remains to be determined. Further work is necessary to establish whether the OHO tumor factor is related to the circulating phosphaturic factor responsible for the renal Pi wasting in the Hyp mouse model of XLH (see Section IV,F,2). Although the obvious therapy for patients with OHO is surgical removal of the tumor, the tumors are small, present in obscure areas, difficult to locate without sophisticated imaging procedures, and often inaccessible. Treatment with Pi, in combination with 1,25(OH)2D, can be initiated and continued until the tumor is removed [160].
D. Does Renal Phosphate Wasting in XLH, ADHR and OHO Involve a Common Metabolic Pathway? Several findings have led to the intriguing hypothesis that defects in a common metabolic pathway can account for impaired renal Pi reabsorption and hypophosphatemia in XLH, ADHR, and OHO. First, FGF-23 mutations that prevent its processing to amino- and carboxy-terminal fragments, and presumably result in the accumulation of FGF-23, cause ADHR [168]. Second, wild-type and mutant FGF-23 induce renal Pi wasting and hypophosphatemia in vivo [77,92,171]. Third, serum FGF-23 levels are greatly increased in OHO patients and are normalized after tumor excision, a procedure that corrects the clinical and biochemical abnormalities of the disorder [180,181]. Fourth, serum FGF-23 levels are increased in some XLH patients [180,181] and are increased >10-fold in Hyp mice when compared to levels in wild-type littermates [182]. Fifth, a peptide derived from FGF-23 is a PHEX
469
CHAPTER 26 Phosphate Homeostasis
substrate [167]. Although the evidence that FGF-23 is a contributing factor for impaired renal Pi reabsorption and hypophosphatemia in ADHR and OHO is strong, it is more difficult to explain the mechanism for increased serum FGF-23 levels in XLH. It is possible that loss of PHEX function leads to the accumulation of an FGF-23 peptide, confirmed to be a PHEX substrate [167], which in turn acts as a feedback inhibitor of FGF-23 processing by proprotein convertase. Further work is necessary to establish the validity of these hypotheses, as well as to determine the contribution of other phosphaturic factors such as sFRP-4 to the underlying pathophysiology and the mechanisms involved in the down-regulation of renal brush border membrane Na/Pi cotransporters.
E. Hereditary Hypophosphatemic Rickets with Hypercalciuria (HHRH) HHRH can be distinguished from XLH, ADHR, and OHO by the appropriately increased serum concentration of 1,25(OH)2D, in response to hypophosphatemia, and the associated hypercalciuria [160,183,184] (Table IV). Since Pi supplementation alone corrects all the clinical and biochemical abnormalities in HHRH, with the exception of the renal Pi leak, it was suggested that HHRH is a primary renal Pi wasting disorder [184]. Given that the biochemical features of HHRH closely resemble those of mice homozygous for the disrupted Npt2a gene (see Section IV,F,1) and that Npt2a is an important determinant of Pi homeostasis and a target for its regulation [30], we hypothesized that mutations in the human ortholog, NPT2a, may be responsible for HHRH. To test this hypothesis, the NPT2a coding region and a fragment of the NPT2a promoter were screened for mutations [185]. No putative diseasecausing mutations in the NPT2a gene were found [185]. Two single nucleotide polymorphisms (SNPs), a silent substitution in exon 7 and a nucleotide substitution in intron 4, were identified and neither segregated with HHRH in a large Bedouin kindred [185]. Furthermore, linkage analysis demonstrated that the two SNPs, as well as five microsatellite markers flanking NPT2a in the chromosome 5q35 region, were not linked to HHRH in the Bedouin kindred [185]. These data exclude NPT2a as a candidate gene for HHRH and suggest that mutations in another renal Na/Pi cotransporter, or a regulator thereof, is responsible for the disorder. It is of interest that heterozygous mutations in the NPT2a gene have been identified in two individuals with renal Pi wasting and hypophosphatemia associated with either urolithiasis or bone demineralization [186]. Although the authors suggest that one copy of
the mutant NPT2a gene is sufficient for the expression of the respective phenotypic abnormalities, more recent functional studies of the mutant NPT2a cDNAs in Xenopus oocytes failed to provide evidence for altered kinetics or dominant negative effects [187]. Thus, the heterozygous mutations in NPT2a likely cannot explain the renal Pi wasting and hypophosphatemia in the patients studied [186]. In support of this hypothesis is the demonstration that mice heterozygous for the disrupted Npt2a gene fail to exhibit a defect in renal brush border membrane Na/Pi cotransport [97].
VIII. SUMMARY AND CONCLUSIONS Several Na/Pi cotransporters have been identified in intestinal, renal, and bone cells. The type IIa Na/Pi cotransporter (Npt2a) is expressed in the brush border membrane of proximal tubular cells, where the bulk of filtered Pi is reabsorbed, and plays a crucial role in the maintenance of Pi homeostasis. Npt2a is a target for regulation by PTH and dietary Pi, both of which induce changes in Npt2a protein abundance in the brush border membrane. Several proteins essential to membrane insertion, retrieval, and degradation of Npt2a have recently been described. In addition, two novel Pi-regulating genes, PHEX and FGF23, which are mutated in patients with Mendelian Pi wasting disorders, have been identified. Future studies are necessary to uncover additional Na/Pi cotransporters in the intestine, kidney, and bone that play an important role in determining Pi economy and to define the precise mechanism whereby PHEX and FGF-23 regulate Pi homeostasis.
References 1. Nordin BEC 1976 Nutritional considerations. In: Nordin BEC (ed) Calcium, Phosphate and Magnesium Metabolism. Churchill Livingstone, New York, pp. 1–35. 2. Marshall RW 1976 Plasma fractions. In: Nordin BEC (ed) Calcium, Phosphate and Magnesium Metabolism. Churchill Livingstone, New York, pp. 162–185. 3. Rowe J, Rowe D, Horak E, Spackman T, Saltzman R, Robinson S, Philipps A, Raye J 1984 Hypophosphatemia and hypercalciuria in small premature infants fed human milk: evidence for inadequate dietary phosphorus. J Pediatr 104:112–117. 4. Giles MM, Fenton MH, Shaw B, Elton RA, Clarke M, Lang M, Hume R 1987 Sequential calcium and phosphorus balance studies in preterm infants. J Pediatr 110:591–598. 5. Wilkinson R 1976 Absorption of calcium, phosphorus and magnesium. In: Nordin BEC (ed) Calcium, Phosphate and Magnesium Metabolism. Churchill Livingstone, New York, pp. 36–112.
470 6. Tanaka Y, DeLuca HF 1973 The control of 25-hydroxyvitamin D metabolism by inorganic phosphorus. Arch Biochem Biophys 154:566–574. 7. Baxter LA, DeLuca HF 1976 Stimulation of 25-hydroxyvitamin D3-1α-hydroxylase by phosphate depletion. J Biol Chem 251:3158–3161. 8. Gray RW, Napoli JL 1983 Dietary phosphate deprivation increases 1,25-dihydroxyvitamin D3 synthesis in rat kidney in vitro. J Biol Chem 258:1152–1155. 9. Markowitz M, Rotkin L, Rosen JF 1981 Circadian rhythms of blood minerals in humans. Science 213:672–674. 10. Portale AA, Halloran BP, Morris RC Jr 1987 Dietary intake of phosphorus modulates the circadian rhythm in serum concentration of phosphorus. Implications for the renal production of 1,25-dihydroxyvitamin D. J Clin Invest 80:1147–1154. 11. Suki WN, Rouse D 1996 Renal transport of calcium, magnesium, and phosphate. In: Suki WN, Rouse D (eds) The Kidney. W.B. Saunders, Philadelphia, pp. 472–509. 12. Brodehl J, Gellissen K, Weber HP 1982 Postnatal development of tubular phosphate reabsorption. Clin Nephrol 17:163–171. 13. Arnaud SB, Goldsmith RS, Stickler GB, McCall JT, Arnaud CD 1973 Serum parathyroid hormone and blood minerals: interrelationships in normal children. Pediatr Res 7:485–493. 14. Greenberg BR, Winters RW, Graham JB 1960 The normal ranges of serum inorganic phosphorus and its utility as a discriminant in the diagnosis of congenital hypophosphatemia. J Clin Endocrinol Metab 20:364. 15. Keating FR Jr, Jones JD, Elveback LR, Randall RV 1969 The relation of age and sex to distribution of values in healthy adults of serum calcium, inorganic phosphorus, magnesium, alkaline phosphatase, total proteins, albumin, and blood urea. J Lab Clin Med 73:825–834. 16. Cross HS, Debiec H, Peterlik M 1990 Mechanism and regulation of intestinal phosphate absorption. Miner Electrolyte Metab 16:115–124. 17. Schroder B, Hattenhauer O, Breves G 1998 Phosphate transport in pig proximal small intestines during postnatal development: Lack of modulation by calcitriol. Endocrinology 139:1500–1507. 18. Nemere I 1996 Apparent nonnuclear regulation of intestinal phosphate transport: Effects of 1,25-dihydroxyvitamin D3, 24,25-dihydroxyvitamin D3, and 25-hydroxyvitamin D3. Endocrinology 137:2254–2261. 19. Danisi G, Murer H 1991 Inorganic phosphate absorption in small intestine. In: Field M, Frizzel RA (eds) Handbook of Physiology—The Gastrointestinal System IV. Oxford University Press, New York, pp. 323–336. 20. Yagci A, Werner A, Murer H, Biber J 1992 Effect of rabbit duodenal mRNA on phosphate transport in Xenopus laevis oocytes: dependence on 1,25-dihydroxy-vitamin-D3. Pflügers Arch 422:211–216. 21. Norbis F, Boll M, Stange G, Markovich D, Verrey F, Biber J, Murer H 1997 Identification of a cDNA/protein leading to an increased Pi-uptake in Xenopus laevis oocytes. J Membrane Biol 156:19–24. 22. Hilfiker H, Hattenhauer O, Traebert M, Forster I, Murer H, Biber J 1998 Characterization of a murine type II sodiumphosphate cotransporter expressed in mammalian small intestine. Proc Natl Acad Sci USA 95:14564–14569. 23. Hattenhauer O, Traebert M, Murer H, Biber J 1999 Regulation of small intestinal Na–Pi type IIb cotransporter by dietary phosphate intake. Am J Physiol 277:G756–G762. 24. Xu H, Bai L, Collins JF, Ghishan FK 2002 Age-dependent regulation of rat intestinal type IIb sodium-phosphate
HARRIET S. TENENHOUSE AND ANTHONY A. PORTALE
25.
26.
27.
28. 29. 30. 31. 32. 33. 34. 35.
36.
37.
38.
39.
40.
41. 42.
cotransporter by 1,25-(OH)2 vitamin D3. Am J Physiol Cell Physiol 282:C487–C493. Xu H, Inouye M, Hines ER, Collins JF, Ghishan FK 2003 Transcriptional regulation of the human NaPi-IIb cotransporter by EGF in Caco-2 cells involves c-myb. Am J Physiol Cell Physiol 284:C1262–C1271. Arima K, Hines ER, Kiela PR, Drees JB, Collins JF, Ghishan FK 2002 Glucocorticoid regulation and glycosylation of mouse intestinal type IIb Na-Pi cotransporter during ontogeny. Am J Physiol Gastrointest. Liver Physiol 283: G426–G434. Berndt TJ, Knox FG 1992 Renal regulation of phosphate excretion. In: Seldin DW, Giebisch G (eds) The Kidney, Physiology and Pathophysiology. Raven Press, New York, pp. 2511–2532. Tenenhouse HS 1997 Cellular and molecular mechanisms of renal phosphate transport. J Bone Mineral Res 12: 159–164. Tenenhouse HS 1999 Recent advances in epithelial sodiumcoupled phosphate transport. Curr Opin Nephrol Hypertens 8:407–414. Murer H, Hernando N, Forster I, Biber J 2000 Proximal tubular phosphate reabsorption: molecular mechanisms. Physiol Rev 80:1373–1409. Murer H, Hernando N, Forster I, Biber J 2003 Regulation of Na/Pi transporter in the proximal tubule. Annu Rev Physiol 65:531–542. Tenenhouse HS, Murer H 2003 Disorders of renal tubular phosphate transport. J Am Soc Nephrol 14:240–247. Collins JF, Bai L, Ghishan FK 2004 The SLC20 family of proteins: dual functions as sodium-phosphate cotransporters and viral receptors. Pflügers Arch, 447:647–652. Murer H, Forster I, Biber J 2004 The sodium phosphate cotransporter family SLC34. Pflügers Arch, 763–767. Bijvoet O 1976 The importance of the kidney in phosphate homeostasis. In: Avioli L, Bordier P, Fleisch H, Mary S, Slatopolsky E (eds) Phosphate Metabolism, Kidney and Bone. Armour-Montagu, Paris, p. 421. Werner A, Moore ML, Mantei N, Biber J, Semenza G, Murer H 1991 Cloning and expression of cDNA for a Na/Pi cotransport system of kidney cortex. Proc Natl Acad Sci USA 88: 9608–9612. Magagnin S, Werner A, Markovich D, Sorribas V, Stange G, Biber J, Murer H 1993 Expression cloning of human and rat renal cortex Na/Pi cotransport. Proc Natl Acad Sci USA 90: 5979–5983. Tenenhouse HS, Roy S, Martel J, Gauthier C 1998 Differential expression, abundance and regulation of Na+phosphate cotransporter genes in murine kidney. Am J Physiol 44:F527–F534. Hartmann CM, Hewson AS, Kos CH, Hilfiker H, Soumoumou Y, Murer H, Tenenhouse HS 1996 Structure of murine and human renal type II Na+-phosphate cotransporter genes (Npt2 and NPT2). Proc Natl Acad Sci USA 93: 7409–7414. Soumounou Y, Gauthier C, Tenenhouse HS 2001 Murine and human type I Na-phosphate cotransporter genes: structure and promoter activity. Am J Physiol Renal Physiol 281: F1082–F1091. Segawa H, Kaneko I, Takahashi A, Kuwahata M, Ito M, Ohkido I, Tatsumi S, Miyamoto K 2002 Growth-related renal type II Na/Pi cotransporter. J Biol Chem 277:19665–19672. Ohkido I, Segawa H, Yanagida R, Nakamura M, Miyamoto K 2003 Cloning, gene structure and dietary regulation of the
471
CHAPTER 26 Phosphate Homeostasis
43.
44.
45. 46. 47. 48.
49.
50. 51.
52.
53.
54. 55. 56.
57. 58.
59.
type-IIc Na/Pi cotransporter in the mouse kidney. Pflügers Arch 446:106–115. Khadeer MA, Tang Z, Tenenhouse HS, Eiden MV, Murer H, Hernando N, Weinman EJ, Chellaiah MA, Gupta A 2003 Na+-dependent phosphate transporters in the murine osteoclast: Cellular distribution and protein interactions. Am J Physiol Cell Physiol 284:1633–1644. Kavanaugh MP, Miller DG, Zhang W, Law W, Kozak SL, Kabat D, Miller AD 1994 Cell-surface receptors for gibbon ape leukemia virus and amphotropic murine retrovirus are inducible sodium-phosphate symporters. Proc Natl Acad Sci USA 91:7071–7075. Forster IC, Loo DD, Eskandari S 1999 Stoichiometry and Na+ binding cooperativity of rat and flounder renal type II Na+–Pi cotransporters. Am J Physiol 276:F644–F649. Forster I, Hernando N, Biber J, Murer H 1998 The voltage dependence of a cloned mammalian renal type II Na+/Pi cotransporter (NaPi-2). J Gen Physiol 112:1–18. Lambert G, Traebert M, Hernando N, Biber J, Murer H 1999 Studies on the topology of the renal type II NaPi-cotransporter. Pflügers Arch Eur J Physiol 437:972–978. Lambert G, Forster IC, Stange G, Kohler K, Biber J, Murer H 2001 Cysteine mutagenesis reveals novel structure–function features within the predicted third extracellular loop of the type IIa Na+/Pi cotransporter. J Gen Physiol 117:533–546. Kohler K, Forster IC, Stange G, Biber J, Murer H 2002 Identification of functionally important sites in the first intracellular loop of the NaPi-IIa cotransporter. Am J Physiol 282:F687–F696. Hayes G, Busch A, Lotscher M, Waldegger S, Lang F, Verrey F, Biber J, Murer H 1994 Role of N-linked glycosylation in rat renal Na/Pi-cotransport. J Biol Chem 269:24143–24149. de la Horra C, Hernando N, Lambert G, Forster I, Biber J, Murer H 2000 Molecular determinants of pH sensitivity of the type IIa Na/Pi cotransporter. J Biol Chem 275: 6284–6287. de la Horra C, Hernando N, Forster I, Biber J, Murer H 2001 Amino acids involved in sodium interaction of murine type II Na+–Pi cotransporters expressed in Xenopus oocytes. J Physiol 531:383–391. Karim-Jimenez Z, Hernando N, Biber J, Murer H 2000 A dibasic motif involved in parathyroid hormone-induced down-regulation of the type IIa NaPi cotransporter. Proc Natl Acad Sci USA 97:12896–12901. Karim-Jimenez Z, Hernando N, Biber J, Murer H 2001 Molecular determinants for apical expression of the renal type IIa Na+/Pi-cotransporter. Pflügers Arch 442:782–790. Ehnes C, Forster IC, Kohler K, Biber J, Murer H 2002 Functional studies on a split type II Na/Pi-cotransporter. J Membrane Biol 188:227–236. Kohl B, Wagner CA, Huelseweh B, Busch AE, Werner A 1998 The Na+-phosphate cotransport system (NaPi-II) with a cleaved protein backbone: implications on function and membrane insertion. J Physiol 508:341–350. Kohler K, Forster IC, Lambert G, Biber J, Murer H 2000 The functional unit of the renal type IIa Na+/Pi cotransporter is a monomer. J Biol Chem 275:26113–26120. Levi M, Lotscher M, Sorribas V, Custer M, Arar M, Kaissling B, Murer H, Biber J 1994 Cellular mechanisms of acute and chronic adaptation of rat renal phosphate transporter to alterations in dietary phosphate. Am J Physiol 267: F900–F908. Pfister MF, Hilfiker H, Forgo J, Lederer E, Biber J, Murer H 1998 Cellular mechanisms involved in the acute adaptation
60.
61.
62. 63. 64.
65. 66.
67.
68.
69. 70. 71.
72. 73.
74.
75.
of OK cell Na/Pi-cotransport to high- or low-Pi medium. Pflügers Arch 435:713–719. Kempson SA, Lotscher M, Kaissling B, Biber J, Murer H, Levi M 1995 Parathyroid hormone action on phosphate transporter mRNA and protein in rat renal proximal tubules. Am J Physiol 268:F784–F791. Lotscher M, Scarpetta Y, Levi M, Halaihel N, Wang H, Zajicek HK, Biber J, Murer H, Kaissling B 1999 Rapid downregulation of rat renal Na/Pi cotransporter in response to parathyroid hormone involves microtubule rearrangement. J Clin Invest 104:483–494. Caverzasio J, Bonjour JP 1993 Growth factors and renal regulation of phosphate transport. Pediatr Nephrol 7: 802–806. Abraham MI, McAteer JA, Kempson SA 1990 Insulin stimulates phosphate transport in opossum kidney epithelial cells. Am J Physiol 258:F1592–F1598. Taketani Y, Segawa H, Chikamori M, Morita K, Tanaka K, Kido S, Yamamoto H, Iemori Y, Tatsumi S, Tsugawa N, Okano T, Kobayashi T, Miyamoto K, Takeda E 1998 Regulation of type II renal Na+-dependent inorganic phosphate transporters by 1,25-dihydroxyvitamin D3. Identification of a vitamin D–responsive element in the human NaPi-3 gene. J Biol Chem 273:14575–14581. Sorribas V, Markovich D, Verri T, Biber J, Murer H 1995 Thyroid hormone stimulation of Na/Pi cotransport in opossum kidney cells. Pflügers Arch 431:266–271. Alcalde AI, Sarasa M, Raldua D, Aramayona J, Morales R, Biber J, Murer H, Levi M, Sorribas V 1999 Role of thyroid hormone in regulation of renal phosphate transport in young and aged rats. Endocrinology 140:1544–1551. Pizurki L, Rizzoli R, Moseley J, Martin TJ, Caverzasio J, Bonjour JP 1988 Effect of synthetic tumoral PTH-related peptide on cAMP production and Na-dependent Pi transport. Am J Physiol 255:F957–F961. Nakai M, Fukase M, Kinoshita Y, Fujita T 1988 Atrial natriuretic factor inhibits phosphate uptake in opossum kidney cells: as a model of renal proximal tubules. Biochem Biophys Res Commun 152:1416–1420. Arar M, Baum M, Biber J, Murer H, Levi M 1995 Epidermal growth factor inhibits Na-Pi cotransport and mRNA in OK cells. Am J Physiol 268:F309–F314. Arar M, Zajicek HK, Elshihabi I, Levi M 1999 Epidermal growth factor inhibits Na–Pi cotransport in weaned and suckling rats. Am J Physiol 276:F72–F78. Pizurki L, Rizzoli R, Caverzasio J, Bonjour JP 1990 Effect of transforming growth factor-α and parathyroid hormonerelated protein on phosphate transport in renal cells. Am J Physiol 259:F929–F935. Law F, Rizzoli R, Bonjour JP 1993 Transforming growth factor-β inhibits phosphate transport in renal epithelial cells. Am J Physiol 264:F623–F628. Levi M, Shayman JA, Abe A, Gross SK, McCluer RH, Biber J, Murer H, Lotscher M, Cronin RE 1995 Dexamethasone modulates rat renal brush border membrane phosphate transporter mRNA and protein abundance and glycosphingolipid composition. J Clin Invest 96:207–216. Wagner GF, Vozzolo BL, Jaworski E, Haddad M, Kline RL, Olsen HS, Rosen CA, Davison MB, Renfro JL 1997 Human stanniocalcin inhibits renal phosphate excretion in the rat. J Bone Miner Res 12:165–171. Saito H, Kusano K, Kinosaki M, Ito H, Hirata M, Segawa H, Miyamoto K, Fukushima N 2003 Human fibroblast growth factor-23 mutants suppress Na+-dependent phosphate
472
76.
77.
78.
79. 80. 81.
82.
83.
84.
85. 86.
87.
88.
89.
90. 91.
HARRIET S. TENENHOUSE AND ANTHONY A. PORTALE
co-transport activity and 1α,25-dihydroxyvitamin D3 production. J Biol Chem 278:2206–2211. Berndt T, Craig TA, Bowe AE, Vassiliadis J, Reczek D, Finnegan R, Jan De Beur S, Schiavi SC, Kumar R 2003 Secreted frizzled-related protein 4 is a potent tumor-derived phosphaturic agent. J Clin Invest, in press. Bai XY, Miao D, Goltzman D, Karaplis AC 2003 The autosomal dominant hypophosphatemic rickets R176Q mutation in FGF23 resists proteolytic cleavage and enhances in vivo biological potency. J Biol Chem 278:9843–9849. Traebert M, Lotscher M, Aschwanden R, Ritthaler T, Biber J, Murer H, Kaissling B 1999 Distribution of the sodium/phosphate transporter during postnatal ontogeny of the rat kidney. J Am Soc Nephrol 10:1407–1415. Gisler SM, Stagljar I, Traebert M, Bacic D, Biber J, Murer H 2001 Interaction of the type IIa Na/Pi cotransporter with PDZ proteins. J Biol Chem 276:9206–9213. Biber J 2001 Emerging roles of transporter–PDZ complexes in renal proximal tubular reabsorption. Pflügers Arch 443:3–5. Pfister MK, Ruf I, Stange G, Ziegler U, Lederer E, Biber J 1998 Parathyroid hormone leads to the lysosomal degradation of the renal type II Na/Pi cotransporter. Proc Natl Acad Sci USA 95:1909–1914. Traebert M, Volkl H, Biber J, Murer H, Kaissling B 2000 Luminal and contraluminal action of 1-34 and 3-34 PTH peptides on renal type IIa Na–Pi cotransporter. Am J Physiol 278:F792–F798. Lederer ED, Sohi SS, McLeish KR 2000 Parathyroid hormone stimulates extracellular signal-regulated kinase (ERK) activity through two independent signal transduction pathways: Role of ERK in sodium-phosphate cotransport. J Am Soc Nephrol 11:222–231. Bacic D, Schulz N, Biber J, Kaissling B, Murer H, Wagner CA 2003 Involvement of the MAPK-kinase pathway in the PTH-mediated regulation of the proximal tubule type IIa Na+/Pi cotransporter in mouse kidney. Pflügers Arch 446:52–60. Jankowski M, Hilfiker H, Biber J, Murer H 2001 The opossum kidney cell type IIa Na/Pi cotransporter is a phosphoprotein. Kidney Blood Press Res 24:1–4. Khundmiri SJ, Rane MJ, Lederer ED 2003 Parathyroid hormone regulation of type II sodium-phosphate cotransporters is dependent on an A kinase anchoring protein. J Biol Chem 278:10134–10141. Bacic D, Capuano P, Gisler SM, Pribanic S, Christensen EI, Biber J, Loffing J, Kaissling B, Wagner CA, Murer H 2003 Impaired PTH-induced endocytotic down-regulation of the renal type IIa Na+/Pi-cotransporter in RAP-deficient mice with reduced megalin expression. Pflügers Arch 446: 475–484. Pribanic S, Gisler SM, Bacic D, Madjdpour C, Hernando N, Sorribas V, Gantenbein A, Biber J, Murer H 2003 Interactions of MAP17 with the NaPi-IIa/PDZK1 protein complex in renal proximal tubular cells. Am J Physiol Renal Physiol. Lederer ED, Khundmiri SJ, Weinman EJ 2003 Role of NHERF-1 in regulation of the activity of Na-K ATPase and sodium-phosphate co-transport in epithelial cells. J Am Soc Nephrol 14:1711–1719. Hilfiker H, Hartmann CM, Stange G, Murer H 1998 Characterization of the 5′-flanking region of OK cell type II Na–Pi cotransporter gene. Am J Physiol 274:F197–F204. Hafdi Z, Couette S, Comoy E, Prie D, Amiel C, Friedlander G 1996 Locally formed 5-hydroxytryptamine stimulates
92.
93.
94.
95.
96.
97.
98.
99.
100.
101. 102.
103. 104.
105.
106.
phosphate transport in cultured opossum kidney cells and in rat kidney. Biochem J 320:615–621. Shimada T, Mizutani S, Muto T, Yoneya T, Hino R, Takeda S, Takeuchi Y, Fujita T, Fukumoto S, Yamashita T 2001 Cloning and characterization of FGF23 as a causative factor of tumor-induced osteomalacia. Proc Natl Acad Sci USA 98: 6500–6505. Segawa H, Kawakami E, Kaneko I, Kuwahata M, Ito M, Kusano K, Saito H, Fukushima N, Miyamoto KI 2003 Effect of hydrolysis-resistant FGF23-R179Q on dietary phosphate regulation of the renal type-II Na/Pi transporter. Pflügers Arch 446:585–592. Yamashita T, Konishi M, Miyake A, Inui K, Itoh N 2002 Fibroblast growth factor (FGF)-23 Inhibits renal phosphate reabsorption by activation of the mitogen-activated protein kinase pathway. J Biol Chem 277:28265–28270. Lotscher M, Kaissling B, Biber J, Murer H, Levi M 1997 Role of microtubules in the rapid regulation of renal phosphate transport in response to acute alterations in dietary phosphate content. J Clin Invest 99:1302–1312. Kido S, Miyamoto K, Mizobuchi H, Taketani Y, Ohkido I, Ogawa N, Kaneko, Harashima S, Takeda E 1999 Identification of regulatory sequences and binding proteins in the type II sodium/phosphate cotransporter NPT2 gene responsive to dietary phosphate. J Biol Chem 274:28256–28263. Beck L, Karaplis AC, Amizuka N, Hewson AS, Ozawa H, Tenenhouse HS 1998 Targeted inactivation of Npt2 in mice leads to severe renal phosphate wasting, hypercalciuria and skeletal abnormalities. Proc Natl Acad Sci USA 95: 5372–5377. Tenenhouse HS, Martel J, Gauthier C, Zhang MYH, Portale AA 2001 Renal expression of the sodium/phosphate cotransporter gene, Npt2; is not required for regulation of renal 1α-hydroxylase by phosphate. Endocrinology 142: 1124–1129. Gupta A, Tenenhouse HS, Hoag HM, Wang D, Khadeer MA, Namba N, Feng X, Hruska KA 2001 Identification of the type II Na+–Pi cotransporter (Npt2) in the osteoclast and the skeletal phenotype of Npt2−/− mice. Bone 29:467–476. Hoag HM, Martel J, Gauthier C, Tenenhouse HS 1999 Effects of Npt2 gene ablation and low-phosphate diet on renal Na/ phosphate cotransport and cotransporter gene expression. J Clin Invest 104:679–686. Zhao N, Tenenhouse HS 2000 Npt2 gene disruption confers resistance to the inhibitory action of PTH on renal Na-phosphate cotransport. Endocrinology 141:2159–2165. Tenenhouse HS, Martel J, Gauthier C, Segawa H, Miyamoto K 2003 Differential effects of Npt2a gene ablation and the X-linked Hyp mutation on renal expression of type IIc Na/Pi cotransporter. Am J Physiol, 285:F1277–F1278. Tenenhouse HS 1999 X-linked hypophosphatemia: A homologous disorder in humans and mice. Nephrol Dial Transplant 14:333–341. Beck L, Soumounou Y, Martel J, Krishnamurthy G, Gauthier C, Goodyer C, Tenenhouse HS 1997 Pex/PEX tissue distribution and evidence for a deletion in the 3′ region of the Pex gene in X-linked hypophosphatemic mice. J Clin Invest 99: 1200–1209. Strom TM, Francis F, Lorenz B, Boeddrich A, Econs M, Lehrach H, Meitinger T 1997 Pex gene deletions in Gy and Hyp mice provide mouse models for X-linked hypophosphatemia. Hum Mol Genet 6:165–171. HYP Consortium 1995 A gene (PEX) with homologies to endopeptidases is mutated in patients with X-linked hypophosphatemic rickets. Nat Genet 11:130–136.
CHAPTER 26 Phosphate Homeostasis
107. Tenenhouse HS, Klugerman AH, Neal JL 1989 Effect of phosphonoformic acid, dietary phosphate and the Hyp mutation on kinetically distinct phosphate transport processes in mouse kidney. Biochim Biophys Acta 984:207–213. 108. Tenenhouse HS, Meyer RA Jr, Mandla S, Meyer MH, Gray RW 1992 Renal phosphate transport and vitamin D metabolism in X-linked hypophosphatemic Gy mice: Responses to phosphate deprivation. Endocrinology 131:51–56. 109. Tenenhouse HS, Werner A, Biber J, Ma S, Martel J, Roy S, Murer H 1994 Renal Na+-phosphate cotransport in murine X-linked hypophosphatemic rickets: Molecular characterization. J Clin Invest 93:671–676. 110. Beck L, Meyer RA Jr, Meyer MH, Biber J, Murer H, Tenenhouse HS 1996 Renal expression of Na+-phosphate cotransporter mRNA and protein: Effect of the Gy mutation and low phosphate diet. Pflügers Arch 431:936–941. 111. Tenenhouse HS, Beck L 1996 Renal Na+-phosphate cotransporter gene expression in X-linked Hyp and Gy mice. Kidney Int 49:1027–1032. 112. Meyer RA Jr, Meyer MH, Gray RW 1989 Parabiosis suggests a humoral factor is involved in X-linked hypophosphatemia in mice. J Bone Miner Res 4:493–500. 113. Meyer RA Jr, Tenenhouse HS, Meyer MH, Klugerman AH 1989 The renal phosphate transport defect in normal mice parabiosed to X-linked hypophosphatemic mice persists after parathyroidectomy. J Bone Miner Res 4:523–532. 114. Nesbitt T, Coffman TM, Griffiths R, Drezner MK 1992 Cross-transplantation of kidneys in normal and Hyp mice: Evidence that the Hyp phenotype is unrelated to an intrinsic renal defect. J Clin Invest 89:1453–1459. 115. Lajeunesse D, Meyer RA Jr, Hamal L 1996 Direct evidence of a humorally-mediated inhibition of renal phosphate transport in the Hyp mouse: involvement of an osteoblast-derived factor. Kidney Int 50:1531–1538. 116. Carpinelli MR, Wicks IP, Sims NA, O’Donnell K, Hanzinikolas K, Burt R, Foote SJ, Bahlo M, Alexander WS, Hilton DJ 2002 An ethyl-nitrosourea-induced point mutation in phex causes exon skipping, x-linked hypophosphatemia, and rickets. Am J Pathol 161:1925–1933. 117. Hughes MR, Brumbaugh PF, Hussler MR, Wergedal JE, Baylink DJ 1975 Regulation of serum 1α,25-dihydroxyvitamin D3 by calcium and phosphate in the rat. Science 190:578–580. 118. Gray RW 1981 Control of plasma 1,25-(OH)2-vitamin D concentrations by calcium and phosphorus in the rat: effects of hypophysectomy. Calcif Tissue Int 33:485–488. 119. Lobaugh B, Drezner MK 1983 Abnormal regulation of renal 25-hydroxyvitamin D-1α-hydroxylase activity in the X-linked hypophosphatemic mouse. J Clin Invest 71:400–403. 120. Gray RW, Wilz DR, Caldas AE, Lemann J Jr 1977 The importance of phosphate in regulating plasma 1,25-(OH)2vitamin D levels in humans: studies in healthy subjects in calcium-stone formers and in patients with primary hyperparathyroidism. J Clin Endocrinol Metab 45:299–306. 121. Insogna KL, Broadus AL, Gertner JM 1983 Impaired phosphorus conservation and 1,25-dihydroxyvitamin D generation during phosphorus deprivation in familial hypophosphatemic rickets. J Clin Invest 71:1562–1569. 122. Maierhofer WJ, Gray RW, Lemann J Jr 1984 Phosphate deprivation increases serum 1,25-(OH)2-vitamin D concentrations in healthy men. Kidney Int. 25:571–575. 123. Portale AA, Halloran BP, Murphy MM, Morris RC Jr 1986 Oral intake of phosphorus can determine the serum concentration of 1,25-dihydroxyvitamin D by determining its production rate in humans. J Clin Invest 77:7–12.
473 124. Portale AA, Halloran BP, Morris RC Jr 1989 Physiologic regulation of the serum concentration of 1,25-dihydroxyvitamin D by phosphorus in normal men. J Clin Invest 83:1494–1499. 125. Gray RW 1981 Control of plasma 1,25-(OH)2-vitamin D concentrations by calcium and phosphorus in the rat: Effects of hypophysectomy. Calcif Tissue Int 33:485–488. 126. Gray RW, Garthwaite TL, Phillips LS 1983 Growth hormone and triiodothyronine permit an increase in plasma 1,25(OH)2D concentrations in response to dietary phosphate deprivation in hypophysectomized rats. Calcif Tissue Int 35: 100–106. 127. Gray RW, Garthwaite TL 1985 Activation of renal 1,25-dihydroxyvitamin D3 synthesis by phosphate deprivation: Evidence for a role for growth hormone. Endocrinology 116:189–193. 128. Halloran BP, Spencer EM 1988 Dietary phosphorus and 1,25-dihydroxyvitamin D metabolism: Influence of insulinlike growth factor I. Endocrinology 123:1225–1229. 129. Grieff M, Zhong M, Finch J, Ritter CS, Slatopolsky E, Brown AJ 1996 Renal calcitriol synthesis and serum phosphorus in response to dietary phosphorus restriction and anabolic agents. Am J Kidney Dis 28:589–595. 130. Zhang MY, Wang X, Wang JT, Compagnone NA, Mellon SH, Olson JL, Tenenhouse HS, Miller WL, Portale AA 2002 Dietary phosphorus transcriptionally regulates 25-hydroxyvitamin D-1α-hydroxylase gene expression in the proximal renal tubule. Endocrinology 143:587–595. 131. Yoshida T, Yoshida N, Monkawa T, Hayashi M, Saruta T 2001 Dietary phosphorus deprivation induces 25-hydroxyvitamin D3 1α-hydroxylase gene expression. Endocrinology 142:1720–1726. 132. Wu S, Grieff M, Brown AJ 1997 Regulation of renal vitamin D24-hydroxylase by phosphate: effects of hypophysectomy, growth hormone and insulin-like growth factor I. Biochem Biophys Res Commun 233:813–817. 133. Wu S, Finch J, Zhong M, Slatopolsky E, Grieff M, Brown AJ 1996 Expression of the renal 25-hydroxyvitamin D-24hydroxylase gene: regulation by dietary phosphate. Am J Physiol 271:F203–F208. 134. Scriver CR, Reade TM, DeLuca HF, Hamstra AJ 1978 Serum 1,25-dihydroxyvitamin D levels in normal subjects and in patients with hereditary rickets or bone disease. New Engl J Med 299:976–979. 135. Delvin EE, Glorieux FH 1981 Serum 1,25-dihydroxyvitamin D concentration in hypophosphatemic vitamin D–resistant rickets. Calcif Tissue Int 33:173–175. 136. Meyer RA Jr, Gray RW, Meyer MH 1980 Abnormal vitamin D metabolism in the X-linked hypophosphatemic mouse. Endocrinology 107:1577–1581. 137. Tenenhouse HS, Jones G 1990 Abnormal regulation of renal vitamin D catabolism by dietary phosphate in murine X-linked hypophosphatemic rickets. J Clin Invest 85:1450–1455. 138. Yamaoka K, Seino Y, Satomura K, Tanaka Y, Yabuuchi H, Haussler MR 1986 Abnormal relationship between serum phosphate concentration and renal 25-hydroxycholecalciferol-1α-hydroxylase activity in X-linked hypophosphatemic mice. Miner Electrol Metab 12:194–198. 139. Azam N, Zhang MY, Wang X, Tenenhouse HS, Portale AA 2003 Disordered regulation of renal 25-hydroxyvitamin D-1α-hydroxylase gene expression by phosphorus in X-linked hypophosphatemic (hyp) mice. Endocrinology 144: 3463–3468. 140. Roy S, Martel J, Ma S, Tenenhouse HS 1994 Increased renal 25-hydroxyvitamin D3-24-hydroxylase messenger ribonucleic
474
141.
142. 143. 144.
145. 146.
147.
148. 149.
150.
151.
152.
153.
154.
155.
156.
HARRIET S. TENENHOUSE AND ANTHONY A. PORTALE
acid and immunoreactive protein in phosphate-deprived Hyp mice: A mechanism for accelerated 1,25-dihydroxyvitamin D3 catabolism in X-linked hypophosphatemic rickets. Endocrinology 134:1761–1767. Roy S, Tenenhouse HS 1996 Transcriptional regulation and renal localization of 1,25-dihydroxyvitamin D3-24-hydroxylase gene expression: Effects of the Hyp mutation and 1,25dihydroxyvitamin D3. Endocrinology 137:2938–2946. Caverzasio J, Bonjour JP 1996 Characteristics and regulation of Pi transport in osteogenic cells for bone metabolism. Kidney Int 49:975–980. Caverzasio J, Selz T, Bonjour JP 1988 Characteristics of phosphate transport in osteoblastlike cells. Calcif Tissue Int 43:83–87. Montessuit C, Bonjour JP, Caverzasio J 1995 Expression and regulation of Na-dependent Pi transport in matrix vesicles produced by osteoblast-like cells. J Bone Miner Res 10: 625–631. Selz T, Caverzasio J, Bonjour JP 1989 Regulation of Nadependent Pi transport by parathyroid hormone in osteoblastlike cells. Am J Physiol 256:E93–100. Pizurki L, Rizzoli R, Caverzasio J, Bonjour JP 1991 Stimulation by parathyroid hormone-related protein and transforming growth factor-alpha of phosphate transport in osteoblast-like cells. J Bone Miner Res 6:1235–1241. Palmer G, Bonjour JP, Caverzasio J 1997 Expression of a newly identified phosphate transporter/retrovirus receptor in human SaOS-2 osteoblast-like cells and its regulation by insulin-like growth factor I. Endocrinology 138: 5202–5209. Veldman CM, Schlapfer I, Schmid C 1997 1α,25-dihydroxyvitamin D3 stimulates sodium-dependent phosphate transport in osteoblast-like cells. Bone 21:41–47. Zhen X, Bonjour JP, Caverzasio J 1997 Platelet-derived growth factor stimulates sodium-dependent Pi transport in osteoblastic cells via phospholipase Cgamma and phosphatidylinositol 3′ kinase. J Bone Miner Res 12:36–44. Suzuki A, Palmer G, Bonjour JP, Caverzasio J 2000 Stimulation of sodium-dependent phosphate transport and signaling mechanisms induced by basic fibroblast growth factor in MC3T3-E1 osteoblast-like cells. J Bone Miner Res 15:95–102. Suzuki A, Palmer G, Bonjour JP, Caverzasio J 2001 Stimulation of sodium-dependent inorganic phosphate transport by activation of Gi/o-protein-coupled receptors by epinephrine in MC3T3-E1 osteoblast-like cells. Bone 28:589–594. Masukawa H, Miura Y, Sato I, Oiso Y, Suzuki A 2001 Stimulatory effect of endothelin-1 on Na-dependent phosphate transport and its signaling mechanism in osteoblast-like cells. J Cell Biochem 83:47–55. Veldman CM, Schlapfer I, Schmid C 1998 Prostaglandin E2 stimulates sodium-dependent phosphate transport in osteoblastic cells via a protein kinase C–mediated pathway. Endocrinology 139:89–94. Burgener D, Bonjour JP, Caverzasio J 1995 Fluoride increases tyrosine kinase activity in osteoblast-like cells: regulatory role for the stimulation of cell proliferation and Pi transport across the plasma membrane. J Bone Miner Res 10:164–171. Nielsen LB, Pedersen FS, Pedersen L 2001 Expression of type III sodium-dependent phosphate transporters/retroviral receptors mRNAs during osteoblast differentiation. Bone 28:160–166. Gupta A, Miyauchi A, Fujimori A, Hruska KA 1996 Phosphate transport in osteoclasts: A functional and immunochemical characterization. Kidney Int 49:968–974.
157. Gupta A, Guo X, Alvarez UM, Hruska KA 1997 Regulation of sodium-dependent phosphate transport in osteoclasts. J Clin Invest 100:538–549. 158. Hodgson SF, Hurley DL 1993 Acquired hypophosphatemia. Endocrinol Metab Clin North Am 22:397–409. 159. Hruska KA, Lederer ED 1999 Hyperphosphatemia and hypophosphatemia. In: Favus MJ (ed) Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism. Lippincott Williams & Wilkins, Philadelphia, pp. 245–253. 160. Tenenhouse HS, Econs MJ 2001 Mendelian hypophosphatemias. In: Scriver CR, Beaudet AL, Sly WS, Valle D (eds) The Metabolic and Molecular Bases of Inherited Disease. McGraw-Hill, New York, pp. 5039–5067. 161. Turner AJ, Tanzawa K 1997 Mammalian membrane metallopeptidases: NEP, ECE, KELL and PEX. FASEB J 11:355–364. 162. Sabbagh Y, Jones AO, Tenenhouse HS 2000 PHEXdb, a locus-specific database for mutations causing X-linked hypophosphatemia. Hum Mut 16:1–6. 163. Sabbagh Y, Boileau G, DesGroseillers L, Tenenhouse HS 2001 Disease-causing missense mutation in the PHEX gene interfere with membrane targeting of the recombinant protein. Hum Mol Genet 10:1539–1546. 164. Sabbagh Y, Boileau G, Campos M, Carmona AK, Tenenhouse HS 2003 Structure and function of diseasecausing missense mutations in the PHEX gene. J Clin Endocrinol Metab 88:2213–2222. 165. Econs MJ, Francis F 1997 Positional cloning of the PEX gene: new insights into the pathophysiology of X-linked hypophosphatemic rickets. Am J Physiol 273:F489–F498. 166. Boileau G, Tenenhouse HS, DesGroseillers L, Crine P 2001 Characterization of PHEX endopeptidase catalytic activity: identification of parathyroid-hormone-related peptide 107-139 as a substrate and osteocalcin, PPi and phosphate as inhibitors. Biochem J 355:707–713. 167. Campos M, Couture C, Hirata IY, Juliano MA, Loisel TP, Crine P, Juliano L, Boileau G, Carmona AK 2003 Human recombinant PHEX has a strict S1′ specificity for acidic residues and cleaves peptides derived from FGF-23 and MEPE. Biochem J 373:271–279. 168. ADHR Consortium 2000 Autosomal dominant hypophosphataemic rickets is associated with mutations in FGF23. Nat Genet 26:345–348. 169. Bowe AE, Finnegan R, Jan de Beur SM, Cho J, Levine MA, Kumar R, Schiavi SC 2001 FGF-23 inhibits renal tubular phosphate transport and is a phex substrate. Biochem Biophys Res Commun 284:977–981. 170. White KE, Carn G, Lorenz-Depiereux B, Benet-Pages A, Strom TM, Econs MJ 2001 Autosomal-dominant hypophosphatemic rickets (ADHR) mutations stabilize FGF-23. Kidney Int. 60:2079–2086. 171. Shimada T, Muto T, Urakawa I, Yoneya T, Yamazaki Y, Okawa K, Takeuchi Y, Fujita T, Fukumoto S, Yamashita T 2002 Mutant FGF-23 responsible for autosomal dominant hypophosphatemic rickets is resistant to proteolytic cleavage and causes hypophosphatemia in vivo. Endocrinology 143:3179–3182. 172. White KE, Jonsson KB, Carn G, Hampson G, Spector TD, Mannstadt M, Lorenz-Depiereux B, Miyauchi A, Yang IM, Ljunggren O, Meitinger T, Strom TM, Juppner H, Econs MJ 2001 The autosomal dominant hypophosphatemic rickets (ADHR) gene is a secreted polypeptide overexpressed by tumors that cause phosphate wasting. J Clin Endocrinol Metab 86:497–500. 173. Kumar R 2000 Tumor-induced osteomalacia and the regulation of phosphate homeostasis. Bone 27:333–338.
475
CHAPTER 26 Phosphate Homeostasis
174. Econs MJ, Drezner MK 1994 Tumor-induced osteomalacia— unveiling a new hormone. N Engl J Med 330:1679–1681. 175. Jonsson KB, Mannstadt M, Miyauchi A, Yang IM, Stein G, Ljunggren O, Juppner H 2001 Extracts from tumors causing oncogenic osteomalacia inhibit phosphate uptake in opossum kidney cells. J Endocrinol 169:613–620. 176. Jan de Beur SM, Finnegan RB, Vassiliadis J, Cook B, Barberio D, Estes S, Manavalan P, Petroziello J, Madden SL, Cho JY, Kumar R, Levine MA, Schiavi SC 2002 Tumors associated with oncogenic osteomalacia express genes important in bone and mineral metabolism. J Bone Miner Res 17:1102–1110. 177. Fukumoto S, Yamashita T 2002 Fibroblast growth factor-23 is the phosphaturic factor intumor-induced osteomalacia and may be phosphatonin. Curr Opin Nephrol Hypertens 11:385–389. 178. Schiavi SC, Moe OW 2002 Phosphatnins: a new class of phosphate-regulating proteins. Curr Opin Nephrol Hypertens 11:423–430. 179. Rowe PS, de Zoysa PA, Dong R, Wang HR, White KE, Econs MJ, Oudet CL 2000 MEPE, a new gene expressed in bone marrow and tumors causing osteomalacia. Genomics 67:54–68. 180. Yamazaki Y, Okazaki R, Shibata M, Hasegawa Y, Satoh K, Tajima T, Takeuchi Y, Fujita T, Nakahara K, Yamashita T, Fukumoto S 2002 Increased circulatory level of biologically active full-length FGF-23 in patients with hypophosphatemic rickets/osteomalacia. J Clin Endocrinol Metab 87:4957–4960. 181. Jonsson KB, Zahradnik R, Larsson T, White KE, Sugimoto T, Imanishi Y, Yamamoto T, Hampson G, Koshiyama H, Ljunggren O, Oba K, Yang M, Miyauchi A, Econs M, Lavigne J, Jueppner H 2002 Fibroblast growth factor 23 in oncogenic osteomalacia and X-linked hypophosphatemia. N Engl J Med 348:1656–1663. 182. Yamazaki Y, Shimada T, Imai R, Hino R, Aono Y, Murakami J, Fukumoto S, Yamashita T 2003 Elevated circulatory and
183. 184.
185.
186.
187.
188.
189. 190.
expression level of fibroblast growth factor (FGF)-23 in hypophosphatemic mice. Bone 32, S88. Tieder M, Modai D, Samuel R, Arie R, Halabe A, Bab I, Gabizon D, Lieberman UA 1985 Hereditary hypophosphatemic rickets with hypercalciuria. N Engl J Med 312:611–617. Tieder M, Modai D, Shaked U, Samuel R, Arie R, Halabe A, Maor J, Weissgarten J, Averbukh Z, Cohen N, Edelstein S, Lieberman UA 1987 “Idiopathic” hypercalciuria and hereditary hypophosphatemic rickets. N Engl J Med 316:125–129. Jones AO, Tzenova J, Frappier D, Crumley M, Roslin NM, Kos CH, Tieder M, Langman CB, Proesmans W, Carpenter TO, Rice A, Anderson D, Morgan K, Fujiwara TM, Tenenhouse HS 2001 Hereditary hypophosphatemic rickets with hypercalciuria is not caused by mutations in the Na/Pi cotransporter NPT2 gene. J Am Soc Nephrol 12:507–514. Prie D, Huart V, Bakouh N, Planelles G, Dellis O, Gerard B, Hulin P, Benque-Blanchet F, Silve C, Grandchamp B, Friedlander G 2002 Nephrolithiasis and osteoporosis associated with hypophosphatemia caused by mutations in the type IIa sodium-phosphate cotransporter. N Engl J Med 347:983–991. Virkki LV, Forster IC, Hernando N, Biber J, Murer H 2003 Functional characterization of two naturally occurring mutations in the human sodium-phosphate cotransporter type IIa. J Bone Miner Res, 18:2135–2141. Kumar R 2002 New insights into phosphate homeostasis: fibroblast growth factor 23 and frizzled-related protein-4 are phosphaturic factors derived from tumors associated with osteomalacia. Curr Opin Nephrol Hypertens 11:547–553. Portale AA 1999 Calcium and phosphorus. In: Barrat TM, Avner ED, Harmon WE (eds) Pediatric Nephrology. Lippincott Williams and Wilkins, Baltimore, pp. 191–213. Murer H, Forster I, Hernando N, Lambert G, Traebert M, Biber J 1999 Posttranscriptional regulation of the proximal tubule NaPi-II transporter in response to PTH and dietary [Pi]. Am J Physiol 277:F676–F684.
CHAPTER 27
Mineralization DAN FAIBISH AND ADELE L. BOSKEY Hospital for Special Surgery, New York, New York, 10021; Affiliated with Weil College of Cornell Medical School, New York, New York
I. Introduction II. Mineralization and Mineral Properties in Model Systems with Vitamin D Alterations
III. Conclusions References
I. INTRODUCTION
Most biomineralization is epitaxial or heterogeneous— occurring on the surfaces of preexisting crystals or on protein and lipid templates that resemble the surface of the crystal [6]. These processes use much less energy than de novo mineralization, and require lower supersaturation. Crystal growth occurs as ions and ion clusters add on to the surface of the initial nuclei or other pre-existing crystals. Crystal growth requires less energy than nucleation and is limited in the case of biologic mineralization by the template upon which the crystals are deposited. Crystals can grow in all dimensions by the addition of ions [6]; by agglomeration [7], in which crystals accumulate, not always in an oriented fashion; or by secondary nucleation. Secondary nucleation, as is seen with hydroxyapatite crystals maturing in solutions in the presence of the dentin protein phosphophoryn [8], is a branching process in which new nuclei form on the surfaces of existing crystals, thus resulting in a new population of immature crystals. Among the vertebrate tissues, enamel contains the largest apatite crystals, but these crystals do contain carbonate [9] and other environmental contaminants (e.g., strontium, fluoride, lead [10]). Bone and tendon– bone insertions contain the smallest apatite mineral crystals [11], 6–10% carbonate [12], as well as adsorbed and incorporated citrate, fluoride, and other trace impurities. Bone apatite crystals are also hydroxide deficient [13], but are not totally devoid of hydroxyl groups [14]. Cementum and dentin mineral crystals have intermediate sizes and intermediate accumulation of foreign ions. In each of these tissues the crystals that are formed initially are smaller than those in the mature tissues because of both the growth of existing crystals and remodeling by osteoclasts [5]. In humans, mineral also deposits at abnormal (unexpected/dystrophic) locations [15]. This dystrophic/ pathologic mineral is frequently apatitic, but calcite (calcium carbonate) is found in pancreatic stones [16]; monosodium urate and sodium pyrophosphate, as well as apatite [15,17], can be deposited in cartilage; and
A. Definitions “Biologic mineralization” is the physicochemical process leading to deposition of inorganic crystals (minerals) on organic matrix within the cell or outside it. This term is more specific than “mineralization,” as biologic mineralization implies a relation between the organic matrix and the mineral. The cell-mediated biomineralization process describes events through which minerals are deposited within cells in an oriented fashion (for example, iron oxides and sulfides in magnetotactic bacteria [1]; silicates in diatoms [2]) or upon an extracellular matrix (for example, calcium carbonates in shells [3] and exoskeletons [4]) or calcium phosphates in bones and teeth [5]. The mineral in physiologically calcified vertebrate tissues is an analog of the geologic mineral, hydroxyapatite (Fig. 1). The physiologic hydroxyapatite crystals are much smaller than those found in geologic deposits and have stoichiometries different from the predicted 10Ca:6PO4 :2OH of the geologic mineral. For that reason biologic vertebrate mineral is often referred to as “apatite” or “apatitic,” meaning “like hydroxyapatite.” This chapter will focus on physiologic and dystrophic apatite formation in situ and in culture and how vitamin D affects the formation of this mineral. Crystalline deposits may form by several different mechanisms. De novo crystal deposition occurs when the solution supersaturation exceeds the solubility of the precipitating phase. Supersaturation refers to the ratio of the solution ion product to the solubility product of the phase in question. The process starts with “nucleation,” in which several ions or ion clusters come together in solution with the orientation they will have in the final crystal. This first step requires a great deal of energy and is facilitated by increasing the ion product, lowering the diffusion of ions in solution, etc. [5]. VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
Copyright © 2005, Elsevier, Inc. All rights reserved.
478
DAN FAIBISH AND ADELE L. BOSKEY
Ca10(PO4)6(OH)2
Na+ Mg++
CO3− −
F−
Sr++ K++ vacancies
FIGURE 1 The composition of the hydroxyapatite unit cell showing the major substituents occurring in biologic apatites and the ions for which they substitute.
brushite, oxalates, and uric acid occur in kidney and salivary stones (Table I). The dystrophic calcifications seen in arteries induced by vitamin D toxicity [18,19], those seen in the presence of hyperphosphatemia [20,21], and those seen in muscle calcifications [22], among others, are all apatitic. In some cases the tissue becomes bone-like and genes associated with osteogenesis are activated [23], as in vascular calcification.
B. Direct and Indirect Effects of Vitamin D and Vitamin D Metabolites on Mineralization 1. PHYSICAL CHEMISTRY OF MINERALIZATION
De novo apatite formation requires Ca2+, PO43−, and OH− ions to come together in the correct orientation with sufficient energy and in sufficient numbers to
TABLE I Dystrophic Mineral Deposits Mineral phase Apatite
Calcium carbonate (aragonite) Oxalates Pyrophosphates Urates
Found in
Affected by vitamin D
Blood vessels Kidney and bladder stones Salivary stones Muscle Skin Hyaline and articular cartilage Pancreatic stones
+ + + + + + 0
Kidney and salivary stones Soft tissues Hyaline and articular cartilage Hyaline and articular cartilage Kidney stones
0 0 0 0
+, Deposition in these tissues is accelerated in hypervitaminosis-D, 0, no effect.
form the first stable apatite crystal (nucleus). After this nucleus is formed, additional ions can add on to these small crystals (nuclei) causing crystal growth. As crystals become larger, new nuclei can branch off the surface (secondary nucleation), in a fashion analogous to glycogen formation. These new nuclei grow and form additional secondary nuclei in an exponential fashion. During biologic apatite formation, matrix molecules provide sites for accumulation of ions and templates to orient mineral growth. Some molecules may act as heterogeneous nucleators, facilitating the deposition of mineral. Others may bind to the crystals and regulate their shape and size [25]. Nucleation occurs at multiple sites along these templates, and crystal habit (shape) and size is regulated by the template and other matrix proteins that bind to the surface of the apatite crystals. In cartilage, bone, dentin, tendon, and ligaments, mineral crystals form on a collagenous matrix. This collagen is the “template” for initial mineral formation and crystal growth [24–26] and mineralization of the collagen matrix gives it increased flexibility and strength [27,28]. Differences in the distribution and properties of other extracellular matrix molecules associated with the fibrillar collagen influence the crystal size and the site of initial crystal deposition. In enamel, which does not have a collagen component, the enamelins are believed to act as nucleators [29], while amelogenin is thought to regulate crystal size and shape [30]. Tables II and III list matrix components from solution, cell culture, or animal models reviewed elsewhere [5] that have been shown to act as apatite nucleators and regulators of apatite crystal growth. Those proteins whose expression is known to be regulated by 1,25(OH)2 vitamin D3 are also indicated. It should be noted that the effects of vitamin D in regulating the expression of matrix molecules important for calcification are indirect effects. Cells control the mineralization process, both by regulation of calcium and phosphate efflux [31,32] and by production of collagen and noncollagenous proteins which guide and direct mineral deposition. There is also a physicochemical process by which supersaturated solutions lead sometimes to mineral deposition in unwanted sites. One example of this physicochemical effect is the hypervitaminosis D syndrome [18,19], where injections of vitamin D into animals cause elevations of circulating calcium and result in arterial and kidney calcification. 2. THE NATURE OF VERTEBRATE MINERAL
The mineral that forms in physiologically mineralized tissues—calcified cartilage, bone, dentin, and enamel, is an analogue of the geologic mineral, hydroxyapatite. The chemical formula for this mineral is Ca10(PO4)6(OH)2,
479
CHAPTER 27 Mineralization
TABLE II
Extracellular Matrix Proteins Associated with Mineralization and Regulated by Vitamin D: Solution Data Effect on mineralization
Ref.
Regulated by vitamin Da
Ref.
Template Inhibitor Inhibitor Inhibitor/nucleator Nucleator Weak nucleator/inhibitor Nucleator Nucleator Regulator of shape Inhibitor Inhibitor Nucleator Nucleator Inhibitor Inhibitor Hydrolyzes phosphate esters/phosphotransferase Hydrolyzes ATP, facilitates cellular Ca transport Degrade matrix molecules that inhibit mineralization
[24] [104] [91,110] [24] [97] [24] [8] [29] [30] [24] [142] [143] [144] [144] [24] [24] [145] [24]
+ + + ? + + + + + + + + ? ? + + + +
[92–95] [102,103] [110]
Protein Type I collagen Osteocalcin Osteopontin Osteonectin Bone sialoprotein (BSP) Dentin sialoprotein Dentin phosphophoryn Amelogenin Enamelin Matrix gla protein (MGP) Fetuin Proteolipid Biglycan Decorin Aggrecan Alkaline phosphatase ATPase Matrix metalloproteinases
[98–101] [150] [150] [150] [150] [151] [152] [117–119]
[153] [154] [145] [146–148]
a Regulation by vitamin D depends on concentration, cell type, and cell maturity; hence + indicates that there is an effect. Readers are referred to other chapters to see precise effects. ?, Unknown.
as there are two structural units in the smallest repeating component of the crystal structure (the unit cell). With the exception of enamel, the other physiologic mineral is deposited in an oriented fashion on a collagen template. Unlike geologic hydroxyapatite crystals
TABLE III Protein Type I collagen Osteocalcin Osteopontin Osteonectin Amelogenin Biglycan Alkaline phosphatase (bone specific) MMPs
which are quite large and visible to the naked eye, the physiologic mineral crystals are microscopic in size being <200Å in their longest dimension in bone, dentin, and related tissues, and ~400Å in the longest dimension in enamel. Because of the small crystal size (some
Extracellular Matrix Proteins Associated with Mineralization and Regulated by Vitamin D: Mineral Properties in Knockout, Transgenic, and Mutant Animals Model(s)
Phenotype
Osteogenesis imperfecta (mutant, knockout, knockin) Knockout Knockout Knockout Transgenic knockout Knockout Knockout
Brittle bones; smaller crystals Increased bone mineral content; larger crystals Increased bone mineral content, larger crystals Increased collagen maturity, larger crystals Defective enamel mineralization Decreased bone mineral content, larger crystals Poorly mineralized bone
Knockouts
Varies depending on enzyme and site
Ref. [154] [107] [111] [155] [156,157] [158] [159] [160]
480 crystals contain only a few unit cells), the physiologic mineral crystals contain a large number of surface impurities and vacancies. Bone apatite is rich in carbonate, which tends to substitute for phosphate in the lattice [33], and is deficient in hydroxide [13,14]; hence it is often referred to as “apatite.” Depending on diet and environment the bone apatite may contain magnesium, sodium, fluoride, strontium, potassium, acid phosphate, and citrate substituents (Fig. 1). The presence of such imperfections in bone, cementum, calcified cartilage, and similarly in the slightly larger dentin crystals in general reduces the crystallinity (crystal size and perfection) of the physiologic mineral deposits and tends to make the crystals more soluble than geologic hydroxyapatite [34,35]. This solubility is essential both to enable remodeling and to facilitate mineral homeostasis. The mechanisms of this homeostasis are highly dependent on vitamin D metabolites as reviewed in Section III of this book. Enamel apatite crystals are larger than bone and dentin mineral and tend to include different impurities. Enamel carbonate content is lower and carbonate is substituted mainly for hydroxide ions [33]. Fluoride is a frequent substituent, as it is included in dentifrices as a caries prophylactic, both to inhibit bacterial carbohydrate metabolism and to decrease enamel mineral solubility. This decrease in solubility is related to the hydrogen bond–induced stabilization of the apatite crystal structure [36]. 3. MINERALIZATION MECHANISMS IN BONE CARTILAGE
AND
In general, body fluids are supersaturated with respect to hydroxyapatite. In other words, the ion product [Ca]10 × [PO4]6 × [OH]2 exceeds the hydroxyapatite solubility product of 10−58 [37]. Yet biologic apatite only deposits physiologically in the so-called mineralized tissues. This is because the body fluids and tissues contain numerous mineralization inhibitors. These inhibitors prevent de novo precipitation, protect the cells from becoming engulfed in mineral, and also regulate the size and shape of the crystals that do form. Mineralization inhibitors include anionic molecules that can chelate calcium, e.g., citrate, ATP, pyrophosphate, glycosaminoglycans, and proteins that bind specifically to apatite crystals, e.g., osteonectin, fetuin, and albumin [24]. The deposition of mineral in bones and teeth occurs at specific sites where the barriers to crystal deposition are diminished either by elevating Ca × PO4 concentrations or by exposing matrix molecules or structures that facilitate mineral deposition. Mineral deposition can be facilitated when mineral crystals are already
DAN FAIBISH AND ADELE L. BOSKEY
present, in which case there is growth and proliferation of the initial crystals, when the surface structure of the matrix matches that of the precipitating phase, or when inhibitors of mineralization are removed [24]. Extracellular matrix vesicles [38], the site of initial mineral deposition in calcifying cartilage and mantle dentin (the first site of dentin mineral deposition), provide protected sites for the accumulation of calcium and phosphate ions [24,39]. Their membranes are also rich in enzymes such as ATPase, pyrophosphatase, alkaline phosphatase, and matrix metalloproteinases [38], enzymes that are required to disrupt mineralization inhibitors in the extracellular matrix [39]. However, the majority of physiologic mineral is believed not to be formed by a vesicle-mediated process. Most mineral forms at multiple discrete locations on type I collagen fibrils [40]. Collagen itself does not support de novo apatite formation; instead, specific anionic noncollagenous proteins can bind to and stabilize the initially formed crystals and regulate the mineralization process [24]. These noncollagenous proteins associate with the collagen fibrils, facilitating initiation of mineralization on the collagen template. In solutions and in cell culture, several of these proteins have been proven to be nucleators of apatite (Table II). These in vitro observations have been validated in animals in which these proteins are ablated (knocked out) or overexpressed (Table III). One of the challenges associated with deciphering the mechanism of biologic calcification is determining which proteins are absolutely essential for initiation of mineralization, and then finding the order in which these proteins are expressed.
C. Methods for Quantifying Tissue Mineralization Several important questions must be addressed when examining a mineralized/mineralizing tissue: Is the tissue mineralized? How much mineral is there? Is the mineral characteristic of physiologic mineral? (Is it a poorly crystalline apatite? Is it properly aligned with the collagen matrix? Is the composition of the mineral different from that in an age- and background-matched control tissue? What size are the crystals?) There are multiple techniques that can be used to address these questions. These techniques are illustrated by examples from vitamin D–related studies. 1. IS THE TISSUE MINERALIZED?
Radiographic methods [plain films, fine-focus radiographs, peripheral quantitative computerized tomography (pQCT), or even magnetic resonance
481
CHAPTER 27 Mineralization
imaging (MRI)] all show changes in opacity when mineral is present. For example, plain films of rachitic (vitamin D–deficient) growing animals show enlarged epiphyses and absence of mineral. Plain X-ray films are routinely used to reveal the presence of increased osteoid and bowing, characteristic of vitamin D–deficient osteomalacia. Vitamin D toxicity is similarly recognized radiographically as increased density in animals and man. More quantitative methods such as fine-focus radiographs, microcomputerized tomography (µCT), pQCT, and dual photon absorptiometry (DEXA) can provide more detail on bone mineral content and bone density. Analysis by µCT provides bone structural data including relative bone density, morphometric parameters such as trabecular number, trabecular thickness, bone volume fraction, and connectivity in three dimensions, in good but not perfect agreement with parameters measured by histomorphometric means [41]. Although there are not yet reports of µCT used to study vitamin D effects, increased availability of µCT should increase its application to metabolic bone diseases in animal models. Peripheral quantitative computerized tomography (pQCT) is related to µCT, enabling measurement of both cortical and trabecular density in peripheral sites. However, where µCT is limited to relatively small samples, whole bodies can be imaged by pQCT, although morphometric parameters cannot be determined. This technique has been applied to demonstrate bone loss in models of vitamin D–stimulated remodeling [42]. Vitamin D and calcium–deficient ovariectomized sheep characterized by quantitative computerized tomography (QCT) and mechanical testing showed decreased density and reduced mechanical strength when contrasted with ovariectomized sheep given a diet rich in calcium and vitamin D [43]. It is important to point out that in a large clinical study of elderly women, pQCT failed to find 1,25(OH)2 vitamin D3 or 25-OH vitamin D3 serum levels as predictors of bone mineral density [44]. 2. HOW MUCH MINERAL IS THERE?
The standard method for quantifying the amount of mineral in a tissue is gravimetric measurement of the weight of the residue left after the tissue is dried (110°C) and the organic components removed by ashing at 600°C. The ash weight, or ash density, if expressed per volume, reveals the total amount of mineral in the tissue. Chemical analysis of Ca and PO4 content has also been used to calculate mineral content, but because of the high organic phosphate content of these tissues these numbers may be inaccurate. Recently, infrared and Raman spectroscopic analysis of mineral to matrix
ratio has been shown to correlate linearly with ash weight of synthetic mixtures [45,46]. There are differences in mineral content of a given bone based on quantitative back scattered election imaging [47] and quantitative density fractionation of ground bone [48]. Nevertheless, quantitative and spatial information was first provided by the spectroscopic imaging techniques described hereafter. 3. IS THE MINERAL CHARACTERISTIC PHYSIOLOGIC MINERAL (IS IT A POORLY CRYSTALLINE APATITE)?
OF
The gold standard for determining the presence of bone mineral (or any mineral phase) is X-ray diffraction (XRD). The X-rays impingent on a powdered sample are reflected at specific angles characteristic of the spacing between lattice planes, and hence characteristic of the mineral phase in question (Fig. 2A). While the small crystal size and numerous imperfections in the physiologic apatite crystals make their diffraction patterns quite broad, they are still easily recognizable as apatite. Additionally, the broadening of individual peaks due to the small size and imperfections in a particular dimension can be used to calculate an average crystallite size and perfection. Electron microscopy (EM) shows the size and shape of the mineral crystals and selected area electron diffraction, analogous to X-ray diffraction, giving a characteristic pattern. The advantage of EM is that physiologic mineral crystals can be observed oriented parallel to the collagen fibrils, whereas dystrophic mineral may not always have that appearance [49]. Both XRD and electron diffraction have proven useful for characterizing mineral associated with matrices and matrix vesicles developed in culture [50]. The amount of mineral present can be quantified by neither X-ray diffraction nor electron diffraction. This can be achieved by quantitative analysis [chemical determination by microprobe or other chemical analysis] or by ashing the specimen (burning off the organic phase) to measure the gravimetric yield of mineral. Energy dispersive X-ray analysis (EDAX) can provide information on the chemical composition of the surface. Coupled to scanning electron microscopes, EDAX enables observation of the morphology of the mineral deposit and simultaneous determination of Ca:PO4 ratios of the mineral [51]. However, the precise determination of this ratio requires comparison with standards, a procedure not always followed. Information on mineral particle thickness and alignment can be obtained by scanning small-angle X-ray scattering (scanning-SAXS). The data obtained agree with the size predictions from infrared techniques [52]. As will be detailed later, scanning SAXS has been
482
DAN FAIBISH AND ADELE L. BOSKEY
A
211
300 112
002
202 210
102
0 20
22
24
26
28
30
32
34
36
Degrees 2θ
B
Phosphate
Amide I Amide II Carbonate
1800
1600
1400 Wave numbers
1200
1000
800
(cm−1)
FIGURE 2 (A) X-ray diffraction pattern of synthetic hydroxyapatite showing the major peaks used for mineral analysis. The broadening of the 002 (c-axis) peak (arrow) is routinely used to estimate crystallite size and perfection. (B) Infrared spectrum of bone mineral showing the bands characteristic of apatite and protein.
used to characterize the bone mineral in the transgenic mouse that overexpresses the vitamin D receptor [53]. Atomic force microscopy (AFM) [54,55] enables examination of the surface structure of a variety of materials and can be used to image individual bone crystals. This high-resolution technique depends on the use of a probe to visualize individual crystals. However, the organic matrix must be removed from the tissue for these measurements, and this is a difficult challenge that might cause breakdown of some crystals. To date there have been no AFM analyses of any of the bones or soft tissues of animals with vitamin D abnormalities.
The vibrations of atoms in molecules are determined both by the bonds in which they are located and by the environment surrounding these moieties. Vibrations that affect the dipole moment and those that are symmetrical, detected by infrared and Raman techniques, respectively, can be used both to identify molecular species and for quantitative analysis of mineral composition and content. As in X-ray and electron diffraction where the poorly mineralized biologic apatite has broad diffraction peaks, the IR and Raman spectra characteristic of physiologic apatite are also broadened (Fig. 2B). Analysis of the relative phosphate to protein peak area ratios (v1/amide I in
483
CHAPTER 27 Mineralization
Raman; v3/amide I in IR [46]) is used to calculate mineral content. Curve fitting or chemometric analysis of specific subbands of the compound spectral features can be used to reveal mineral crystal size and perfection (crystallinity [52,56–58]), acid phosphate content [56,57], and carbonate substitution [59,12]. Magnetic resonance imaging (MRI) and nuclear magnetic resonance (NMR) methods can also reveal the distribution of acid phosphate and features of the mineral crystal structure. Solid-state NMR spectroscopy has long been used for 1H, 13C, and 31P studies of bone mineral and calcium phosphates [60]. Magic angle spinning (MAS) and MAS coupled with cross polarization (CP-MAS) have been used to characterize initial mineral and matrix composition in developing bones [61–63]. Although there are no published studies of mineral characteristics observed by these techniques in the presence and absence of vitamin D, the high correlation between these techniques, X-ray diffraction, and spectroscopic and chemical analyses [64] indicates that NMR data would support the findings from XRD and IR described later. Quantitative backscattered electron imaging (qBEI) can be used to measure the amount of mineral in a localized region of biopsied bone. Gray-scale images are produced, and each pixel in the image is compared to osteoid and hydroxyapatite standards, from which calcium weight percent values are calculated. Data is compared based on pixel distributions of calcium peak weight percent (CaPeak). This technique has been used to describe the mineral distribution in the vitamin D receptor (VDR) knockout mice discussed later [65]. 4. METHODS FOR STUDYING IN VITRO MINERALIZATION
Studies in culture (in the presence of tissues or cells) have been extensively used to assess the function of vitamin D metabolites and the vitamin D receptor in the mineralization process. Although it is popular to use 10 mM β-glycerophosphate as a substrate [e.g., 66–68], as will be discussed later, 5–10 mM organic phosphate can cause ectopic mineral deposition, even in the absence of cells or matrix, as long as alkaline phosphatase activity is present [49,69]. Mineral can be characterized in the culture system by measuring changes in solution calcium or phosphate concentrations (often by the use of radiolabeled ions) and the amount of mineral in the matrix (chemical analysis, diffraction, or spectroscopic methods). More recently it has become very popular to quantify calcium and phosphate accumulation in culture by the use of histochemical stains. Sometimes these values are quantified by extracting the stained culture. One of the problems with this histochemical approach is that these stains
are often not specific for mineral. For example, von Kossa staining for phosphate [51,66] can equally label phosphate-containing matrix molecules, especially cell membranes [70]. Alizarin red, used to stain for calcium [51,66], can show the presence of this ion in calcium phosphates, but also calcium bound to proteoglycans. Thus conclusive proof of the presence of mineral in a culture experiment comes from diffraction methods, electron microscopy, or spectroscopic techniques as discussed earlier.
D. Direct and Indirect Effects of Vitamin D on Mineralization 1. PHYSICOCHEMICAL EFFECTS
Vitamin D’s “direct” effect is to increase serum Ca2+ levels by stimulating osteoclastic resorption, kidney Ca retention, and intestinal Ca absorption (see Section III of this book). This in turn leads to an elevation in local calcium concentration, and often an elevation in phosphate. The increased circulating levels of calcium and phosphate, though not causing marked hypercalcemia or hyperphosphatemia (because of the counter effects of parathyroid hormone), are thought to be sufficient to raise the serum levels enough to promote dystrophic calcification, especially in blood vessels [19,71] and kidneys [72]. Injection of vitamin D3 or 1,25-(OH)2 vitamin D3 into animals can cause formation of apatite deposits in articular and growth cartilage as well as bone [73]. There is also a case report showing that milk accidentally fortified with excessive vitamin D caused dental pulp calcification in a child [74], showing that mineralizing tissues other than bone or cartilage can be affected. The vitamin D–receptor knockout mouse has an impaired bone formation phenotype (see later discussion) that can be rescued by calcium treatment [75]. Similarly, vitamin D–deficient rickets in animals can be cured by increasing serum calcium (by lactose and calcium infusion) without addition of vitamin D [76,77]. This suggests that the increased mineralization associated with vitamin D may simply reflect increased calcium or calcium and phosphate concentrations. However, the mineral quality has not been examined in any of these calcium-treated models. From the perspective of physical chemistry, to precipitate a basic calcium phosphate such as bone apatite, the solution needs to have sufficient pH, and sufficient calcium and phosphate concentrations. In solution, at pH7.4, when the Ca × PO4 mM2 product exceeds 5.5 mM2, precipitation will occur even in the absence of a nucleator or a template for mineral deposition [78]. Thus, raising the local calcium or local phosphate concentration, even slightly, can cause
484 precipitation, all other factors being equivalent. This is extremely important when considering cell culture studies where ion concentrations are elevated beyond physiologic levels. Physiologic phosphate concentrations are 0.9–1.5 mM in most species [79]. β-Glycerophosphate, often used in mineralizing cultures and said to enhance osteoblastic differentiation, is a substrate for alkaline phosphatase. When 10 mM β-glycerophosphate is used in culture, the phosphate concentrations of the medium can be as high as 10 mM. With physiological calcium concentrations of 1.5-2.5mM, the Ca × PO4 concentration of the solution exceeds 5.5 mM2 and hence mineral will deposit in the absence of a matrix. As shown by Khouja et al. [69], the occurrence of this dystrophic mineralization simply demonstrates the presence of alkaline phosphatase activity, not of physiologic mineralization. Physiologic calcium concentrations are ~2.1–2.5 mM. Raising the concentration to 2.9 or 4.2 mM [67] with phosphate at even 2 mM will give a Ca × PO4 product of 5.8 mM2 or 8.4 mM2, greater than the value needed for de novo precipitation. Although there is altered gene expression caused by the presence of these elevated calcium concentrations [67], it has not yet been demonstrated that this altered gene expression is due to the presence of the calcium, rather than to the presence of the dystrophic mineral that surrounds the cells. We know that gene expression is altered when mineral surrounds the osteocyte [80], and in this case the mineral surrounding the osteocyte is physiologic mineral. Thus caution must be exercised when looking at gene expression in such cases. In situ there are more controls on calcium concentrations; thus the elevations in calcium in tissues of animals treated with vitamin D may not directly cause precipitation, but may increase the ion product sufficiently that initial mineralization is facilitated, as is crystal growth. In vitamin D–deficient animals (with rickets and/or osteomalacia) the mineral content of the bulk tissue, reflected by histochemical stains or ash weights, is decreased relative to controls (Fig. 3). The crystals present in the epiphysis and metaphysis, i.e., newly formed mineral crystals, tend to be larger than those in age- and gender-matched controls. It is not certain whether this is because osteoclast activity is increased or because in the absence of an appropriate matrix, existing crystals grow at the expense of new crystals being formed. Increased osteoclast activity has been noted in some cases of human hypophosphatemic oncogenic osteomalacia [81,82] and in vitamin D–deficient rats [83], hens [84], and pigs [85]. However, the effects of vitamin D metabolites on extracellular matrix formation and composition suggest that impaired formation may be a contributing factor.
DAN FAIBISH AND ADELE L. BOSKEY
2. EFFECTS OF VITAMIN D ON CELLS AND MATRIX MOLECULES
a. Introduction As reviewed elsewhere in this volume, vitamin D is a differentiation factor, facilitating the expression of mature chondrocyte and mature osteoblast activity, as well as osteoclastogenesis. Vitamin D alters the gene expression and protein synthesis of numerous matrix proteins which are regulators of biomineralization (Table II). This regulation occurs through both genomic and nongenomic pathways and is modulated by calcium influx. The vitamin D–regulated proteins include, but are not limited to, calcium-binding proteins and several enamel (e.g., enamelins, amelogenins), bone, and dentin proteins [e.g., dentin sialoglycoprotein (DSP) and dentin phosphoprotein (DPP)]. Vitamin D also increases expression of membrane-bound enzymes such as alkaline phosphatase and the kinases and phosphatases [86] that regulate the phosphorylation and dephosphorylation of intracellular and extracellular matrix proteins. The effects of vitamin D metabolites on cell differentiation are reviewed elsewhere in Section I of this book; however, it is important to note that in vitro 24,25-(OH)2 vitamin D stimulates chick chondrocyte [87] and rabbit osteoblast maturation [88], while 1,25(OH)2 vitamin D stimulates rat osteoblast differentiation in bone marrow stromal cells [89]. Since it is the mature chondrocyte and osteoblast that synthesize the mineralizable matrix and/or provide the enzymes necessary to modulate that matrix to facilitate mineralization, one indirect effect of vitamin D might be on the formation of the proper cells to allow mineralization. In cultured chick hypertrophic chondrocytes, for example, 1,25-(OH)2 vitamin D3 causes a dose-dependent decrease in activity of alkaline phosphatase, while 24,25(OH)2 vitamin D3 stimulates this activity [90]. The effect of 1,25(OH)2 vitamin D3 in stimulating kinase and phosphatase activity, and thereby regulating the extent of phosphorylation of osteopontin, is extremely interesting in light of data showing that the moderately phosphorylated bone osteopontin is an effective mineralization inhibitor [91]. In contrast, the highly phosphorylated milk osteopontin can induce hydroxyapatite formation in solution, and stabilize precursors of hydroxyapatite [Boskey, unpublished]. The activities of several of the other matrix proteins depend on their state of phosphorylation, stressing the importance of these enzymes. b. The Influence of Vitamin D on Bone Matrix Proteins Vitamin D modulates the production of several molecules (Table II) that are crucial for the synthesis and function of bone tissue, although there is no clear evidence that direct effects of vitamin D are required
485
CHAPTER 27 Mineralization
A
D-deficient
D-deficient
Control
Control
100
1.00
12
80
0.80
60
0.60
6 3 Ca
B
20
0.20
0
PO4
Control
0.40
Hyp
0.80 β002
80 Ash Weight
1.00
0.40
20
0.20 epiphysis metaphysis diaphysis
Control
D-deficient
80
0.80
60
0.60
40
0.40
20
0.20 metaphysis diaphysis
calvaria
Hyp
1.60 1.20 0.80 0.40 epiphysis metaphysis diaphysis
D-deficient
Control
D-deficient
2.00 1.60 1.20 0.80 0.40 0.00
0.00
0
diaphysis
0.00
Control 1.00
β002
Ash Weight
Control
epiphysis metaphysis diaphysis
100
epiphysis
D-deficient
2.00
0.00
0
C
Hyp
0.60
40
0.00
diaphysis
Control
100
60
epiphysis
Ca/PO4 (molar ratio)
0
40
Ca/PO4 (molar ratio)
9
β002
15 Ash Weight
Serum Chemistry
Control
metaphysis
diaphysis
metaphysis diaphysis calvaria
FIGURE 3 Mineral analyses in models of vitamin D deficiency. (A) Vitamin D–deficient rats. Male Sprague-Dawley rats (21 days old) were maintained in the dark on a vitamin-D and phosphate-deficient diet for 3 weeks. The first panel shows their serum Ca and PO4 levels. 1,25-(OH)2vitamin D levels were not detectable in the D-deficient animals. Tibias removed at sacrifice were used for analysis of ash weight and β002 (1/crystallinity). Reproduced with permission from Springer-Verlag [119]. (B) Female mice with hypo-phosphatemic rickets (Hyp) and their age- and sex-matched controls were 35 days old at sacrifice. Femora were used for analysis of ash weight, β002, and Ca/PO4 ratio of the ash. Reprinted with permission from Elsevier [117]. (C) Second-generation vitamin D–deficient rats and their age-matched controls were 7 weeks old at sacrifice. Femora were used for analysis of ash weight and β002 and Ca/PO4 ratio of the ash. Reprinted with permission from Elsevier [116].
for normal bone mineralization. The actions of these proteins have been described in detail elsewhere [24]; hence only a few representative proteins are described in this section. Readers are referred to references in Tables II and III for more detail. i. Collagen The scaffold of bone matrix is made predominantly of type I collagen molecules. These molecules are produced early in bone tissue formation, and their special morphology supports the mineral deposition in later stages. In addition, the collagen
molecules provide the elasticity characteristic of bone tissue, which is essential for its normal function in response to mechanical forces. The synthesis of collagen is a complex process, and it is regulated by several factors, one of them being vitamin D. The active metabolite, 1,25(OH)2 vitamin D3 downregulates the transcription of α1(I) collagen by osteoblasts in a variety of cells [92–95]. This decrease in type I collagen expression appears to be cell stage dependent, occurring in early bone nodules formed in
486
DAN FAIBISH AND ADELE L. BOSKEY
culture, but not in intermediate and mature ones [94]. Adding 1,25(OH)2 vitamin D3 after the completion of extracellular matrix formation has been reported to inhibit further mineralization in chick osteoblast cultures [95], suggesting that 1,25(OH)2 vitamin D3 affects the metabolic processes by which osteoblasts control mineralization, independent of its effects on the formation of the collagenous matrix. However, Matsumoto showed that 1,25(OH)2 vitamin D3 stimulated alkaline phosphatase activity in both early and late osteoblast cultures and reported enhanced mineral deposition based on EM and 45Ca uptake when 10−10 M 1,25(OH)2D3 was added after the start of mineralization [96]. Similarly (Fig. 4), in differentiating chick limb-bud mesenchymal cell micro-mass cultures, 1,25(OH)2D3 added continuously to cultures after the matrix had been formed and mineralization was started (day 9) increased the rate of mineral accumulation compared to mineralizing cultures that did not receive exogenous vitamin D. This increase was in contrast to cultures to which vitamin D was added continuously while cells were differentiating (day 5) or when the cells were first beginning to deposit a cartilage matrix (day 7).
Vitamin D also regulates the expression of the noncollagenous proteins in mineralized tissues (Table II). Many of these proteins are believed to play direct roles in the regulation of biomineralization [5,24], and thus they account in part for the indirect effect of vitamin D on mineralization. ii. Noncollagenous proteins The expression of bone sialoprotein (BSP), an in vitro apatite nucleator [97] and mineralization regulator [24], is suppressed by addition of 1,25(OH)2D3 to osteoblast cultures [68,95,98–101]. In osteoblast cultures without exogenous vitamin D supplementation, BSP is expressed at its highest levels immediately prior to onset of extracellular matrix mineralization [101]. Thus one wonders whether its suppression in the presence of 10−8 M 1,25(OH)2D3 may reflect a compensatory mechanism, preventing excessive initial calcification. Osteocalcin (OC) was one of the first matrix proteins whose expression was shown to be up-regulated by vitamin D [102]. OC is a member of a large family of hepatic and skeletal vitamin K–dependent proteins that undergo posttranslational modification and γ-carboxylation at key glutamic acid residues (Gla) and have mineral binding capacities. The mature 5.7-kDa
0.50 C 0.40
Ca : DNA
4P 0.30
4P +D day5
0.20
4P +D day7 4P +D day9
0.10
0.00 12
14
16 Days in Culture
19
21
FIGURE 4 Effect of 10−8 M 1,25(OH)2D3 on mineralization in differentiating chick limb bud mesenchymal cell micromass cultures. Data is expressed as µg Ca/µg DNA. Continuous vitamin D treatment was started at day 5 (when cartilage nodules formed), at day 7 (prior to the start of visible chondrocyte hypertrophy), or at day 9 (2 days before start of mineralization). Control cultures received a comparable volume (20 µl) of ethanol, the carrier for 1,25(OH)2D3. On days 12–21, Ca and DNA contents were determined in control and mineralizing cultures. The DNA contents of mineralizing and nonmineralizing cultures were not significantly different (not shown), and 1,25(OH)2 vitamin D3 treatment did not significantly accelerate proliferation. There was a slight increase in DNA content of cultures treated on day 5, but neither this increase nor the gradual increase in DNA content with time was significant, indicating that the cultures had already achieved their plateau phase of proliferation. In contrast, cultures treated on day 9 showed an increase in Ca content relative to the other cultures.
487
CHAPTER 27 Mineralization
protein contains three Gla residues and accumulates in bone as a result of its high affinity for hydroxyapatite [24]. Initial osteoblastic expression of OC occurs after the onset of extracellular matrix mineralization and increases with progressive mineralization and maturation of the osteoblast to a terminally differentiated state [68]. 1,25(OH)2D3 is a major stimulus for the transcription of human and rat OC, but it inhibits the transcription of the mouse OC gene [103], implying either a species difference or, more likely, differences in conditions under which expression was studied. OC can inhibit the formation of hydroxyapatite in vitro [104], a function that requires the presence of the Gla residues [105]. OC also is important for osteoclast recruitment [24]. OC-deficient mice have increased rates of bone formation despite a lack of detectable abnormalities in osteoclast numbers [106]. These mice also have increased mineral content and increased mineral crystallinity [107], supporting the role of OC in regulating mineral turnover. Thus OC has a role in the recruitment of osteoclasts to the surface of mineralized bone, contributing in this way to the regulation of both bone formation and resorption [101]. OC biosynthesis is suppressed throughout the life span of the culture, as a result of an arrested stage of cell differentiation when 1,25(OH)2 vitamin D3 is added to rat osteoblast cultures before the onset of mineralization. Once the cultures begin to mineralize and the cells are terminally differentiated, media OC concentrations are increased by continuous or acute application of the vitamin [108]. Osteopontin (OPN), a glycoprotein produced in a number of tissues [109], is consistently up-regulated by 1,25(OH)2 vitamin D3 in proliferating and differentiated mouse and rat osteoblasts [110]. In bone, OPN mediates autocrine–paracrine functions in the regulation of tissue formation. OPN is important for recruiting osteoclasts for bone remodeling [110] and it acts as a signaling protein in many tissues. OPN is an in vitro inhibitor of mineralization in muscle, cartilage, and bone [24,110]. In that light it is of interest to note that increasing levels of inorganic phosphate activate OPN expression [109].
II. MINERALIZATION AND MINERAL PROPERTIES IN MODEL SYSTEMS WITH VITAMIN D ALTERATIONS The effects on mineralization of vitamin D, 1,25(OH)2 vitamin D3 and other vitamin D metabolites varies with dose and type, as well as with the system being studied. Even in similar species effects depend on animal age, gender, route of delivery, and method of analysis.
In cell culture, results are also dependent on the maturity of cells or tissues, concentrations of metabolite, and whether the metabolite is given acutely or continuously. It is well known that vitamin D deficiency or even vitamin D insufficiency causes rickets (failure of growth plate cartilage to become mineralized) and osteomalacia (failure of osteoid to become mineralized) [112–114], and that vitamin D toxicity (hypervitaminosis D) causes dystrophic calcification [19,115]. But even in rodents the effects of vitamin D deficiency (or D insufficiency) and hypervitaminosis D can be variable. In cell and organ culture the same vitamin D metabolites have been reported both to prevent mineralization and to enhance it. Understanding the reasons for this diversity can provide a good deal of insight into mineralization mechanisms.
A. Animal Models and Human Diseases with Mineral Alterations 1. VITAMIN D DEFICIENCY
The mineral content (ash weight), crystallinity parameter (β002, which is inversely related to crystallite size and perfection), and Ca/P ratio of the mineral in the epiphyseal growth plates, metaphysis, diaphysis, and calvaria of several models of vitamin-D deficient rickets are compared in Fig. 3. In each model, the mineral content is decreased relative to age-matched controls, and the crystals tend to be larger, but Ca/P ratios are quite variable. The vitamin D–deficient rat (Fig. 3A [116]), the hypophosphatemic mouse (Fig. 3B [117]), and the second-generation vitamin D-deficient rat (Fig. 3C [118]) show distinct differences, although the D-deficient animals have negligible 1,25(OH)2vitamin D3 and reduced 25 (OH) vitamin D levels. In the secondgeneration vitamin D-deficient rats, born to vitamin D– deficient mothers [116], there are only small differences in bone ash weight, crystallinity, and Ca/P ratios when compared to age-matched controls. These differences are greater in comparably aged, comparable sex and background, first-generation vitamin D–deficient rats and chickens [118,119]. The initial electron microscopic studies of vitamin D– deficient bones [120] pointed out that the total number of extracellular matrix vesicles was not different from that in vitamin D–sufficient rats, but the vesicles contained less mineral and an extracelllular matrix devoid of mineral. Healing of the rickets was then associated with both matrix vesicle and extracellular matrix mineralization. In another classic study, EDAX was used to monitor changes in cellular and extracellular matrix ion concentrations in all zones of normal and rachitic growth plates [121]. Potassium levels were invariant,
488 and calcium was found throughout the extracellular matrix in both normal and rachitic growth plates. The phosphate levels in the extracellular matrix did not increase until mineralization commenced, and hence were lower in the rachitic matrix, leading the authors to point out the importance of phosphate sufficiency for mineralization. Since vitamin D is now known to regulate both calcium and phosphate levels ([122] and Section III of this book), these observations are no longer surprising, but are important in terms of the mineralization mechanisms discussed previously. In the hypophosphatemic mouse [117], in which failure to mineralize the growth plate and osteoid is due to a genetic abnormality in the endopeptidase known as PHEX (see Chapter 70), the Ca/P ratios are elevated, probably because of abnormalities in renal phosphate transport. In addition, there is less bone mineral and the mineral crystals that are present tend to be larger/more perfect. A mechanism can be postulated to explain these mineralization abnormalities in the bones of hypophosphatemic animals. It is suggested that vitamin D regulates mineralization by controlling degradation of matrix proteins based on the knowledge that a potential substrate for PHEX is matrix extracellular phosphoglycoprotein (MEPE) [123,124] and, by analogy, the structurally related phosphorylated matrix glycoproteins, OPN, BSP, and DMP-1, and that vitamin D causes down-regulation of PHEX expression [125]. In other words, vitamin D suppresses the breakdown of these PHEX substrates, enabling them to act as mineralization inhibitors. PHEX may also be important in producing nucleators by the degradation of these phosphoproteins. Although we know that many of these phosphoproteins can act both as nucleators and inhibitors [24], this mechanism is hypothetical and remains to be validated. The mineral properties of vitamin D insufficiency in human children have not been characterized. However, mineral in adults with osteomalacia has been subject to more detailed analysis (see later discussion). 2. VDR ALTERATIONS
The nuclear vitamin D receptor (VDR) binds 1,25(OH)2 vitamin D with high affinity and selectivity, as reviewed in Section II. The histologic appearance of the bone tissue of vitamin D receptor knockout (VDRKO) animals resembles that of an animal with vitamin D deficiency [75,126,127]. The mice have a normal skeleton at birth but develop hypocalcemia and hyperparathyroidism shortly after weaning, comparable to human vitamin D–resistant rickets type II [128,129] and Chapters 70 and 72). While bone volume was not decreased in these mice, the amount of unmineralized
DAN FAIBISH AND ADELE L. BOSKEY
bone (osteoid) was increased 15-fold compared to the controls [128]. However, the rachitic bone malformation and growth retardation in the VDR-KO can be rescued by dietary calcium supplementation or phosphorus restriction [129]. These animals have normal circulating levels of 25-OHvitamin D and elevated levels of 1,25(OH)2vitamin D. They have slight hypocalcemia and elevated PTH levels due to the inability of the vitamin D target organs to regulate calcium influx. There have not yet been characterizations of the mineral in these animals; however, it is anticipated that they will resemble the parameters in vitamin D–deficient bones. The VDR knockout animals have extended growth plates and an increased number of hypertrophic chondrocytes [128]. The hypertrophic chondrocytes express the correct relative amounts of phenotypic markers of chondrocyte maturity, type X collagen, vascular endothelial growth factor (VEGF), and osteopontin, but show decreased staining for annexin V–phosphatidylserine, an early marker of apoptosis [130]. This suggests that impaired mineralization might be due to the failure of chondrocytes to undergo apoptosis, but another possibility is that the calcium and phosphate levels are not sufficient to mineralize the growth plates, with apoptosis following rather than preceding mineralization. Additionally, both annexin and phosphatidylserine are membrane components of extracellular matrix vesicles; thus decreased staining for these components might suggest that chondrocytes are not producing vesicles, or that vesicles lacking these components are not functional. Osteoclast numbers are normal in the VDR knockout [75], arguing against excessive remodeling as the origin of the mineralization defect. More is known about the mineral characteristics in the vitamin D receptor transgenics. Gardiner et al. [53] overexpressed the vitamin D receptor under control of the osteocalcin promoter. The long bones in these VDR transgenic mice had increased cross-sectional area and increased strength with two- and threefold elevations of osteoblast VDR levels [53]. Based on qBEI analysis [65], cortical bone peak mineral content (CaPeak) was enhanced in a dose dependent manner with the mice with 3 × overexpression having values higher than controls (p < 0.02), and the transgenic mice with 2 × overexpression having intermediate values. In trabecular bone, the increase in CaPeak was significant for both twofold and threefold overexpression. Calcified cartilage and primary spongiosa showed parallel trends. There was also increased homogeneity of mineralization in the transgenic mice. However, FTIR microspectroscopy, XRD, and scanning-SAXS failed to show any significant alterations in the mineral crystal properties of the transgenic bones, although there was a trend for the
489
CHAPTER 27 Mineralization
Control Transgenic
Mineral : Matrix Ratio
12 9 6 3 0
CP
CC Cortical bone
CE
T Trabecular bone
FIGURE 5 Infrared microspectroscopy of the effect of three-fold VDR overexpression on the tibia of 4-month-old mice shows slight but not significant increase in mineral content (mineral:matrix ratio). Data were collected from cortical bone on the periosteal side (CP), in the center (CC), and on the endosteal side (CE), as well as within individual trabeculae (T).
4. THE 1-HYDROXYLASE KNOCKOUT
Mice that lack the 1-α-hydroxylase also present with hypocalcemia, rickets, and osteomalacia [134] and mimic human pseudovitamin D–deficiency rickets (PDDR) (see Chapter 71). Patients that lack this hydroxylase have secondary hyperparathyroidism, growth retardation, rickets, and osteomalacia [135]. The mineral in the bones of the PDDR mice has not been characterized, but their mechanical properties are typical of rachitic animals with significantly reduced stiffness and reduced ultimate load [136] (see Chapter 7). Treatment of these animals with high calcium, phosphate, and lactose diets rescues the phenotype, but the bones do not grow as well as the VDR knockout mice given the same diet [136], implying that the 1,25(OH)2 vitamin D may have properties beyond regulating serum calcium and phosphate levels. It is important to note that there were no differences in osteoclast number in the control and knockout mice given the normal or the enriched diet. 5. OSTEOPOROSIS AND OSTEOMALACIA IN HUMANS
mineral content (mineral:matrix ratio) to be increased in the transgenic mice (Fig. 5) while crystallinity was identical in both groups [65]. Thus VDR overexpression did not markedly change the mineral properties and did not mimic effects of vitamin D toxicity. In humans, hereditary 1,25(OH)2 vitamin D–resistant rickets (HVDRR), also known as vitamin D–dependent rickets type II, is a rare autosomal recessive disease that arises as a result of mutations in the gene encoding the VDR [131,132] (see Chapter 72). There have been no detailed mineral analyses in these patients. 3. VITAMIN D CARRIER PROTEIN KNOCKOUT
Vitamin D is carried to its receptor (VDR) in the target tissues by the vitamin D binding protein (DBP). The mineral properties in DBP knockout mice have not been reported, but they have an interesting phenotype [133]. DBP knockout mice have low levels of serum vitamin D metabolites but otherwise appear normal and show none of the bone abnormalities seen in vitamin D–deficiency rickets or osteomalacia. When stressed with excess exogenous vitamin D (1000 U/g body weight) to induce vitamin D toxicity, they did have elevated serum calcium levels, but did not show the toxic effects found in wild-type animals. Kidney mineral deposits were found in the wild-type but not in the knockout mice. Additionally, when stressed with a diet low in vitamin D, histomorphometry showed increased thickening of the osteoid seams in the DBP knockout mice, consistent with impaired mineralization and “hypovitaminosis D osteopathy.”
Many individuals with osteoporosis are Ca deficient, demonstrate vitamin D insufficiency, and show histochemical evidence of osteomalacia [137,138]. Osteoporotic bones without evidence of osteomalacia have decreased mineral content and increased mineral crystallinity based on FTIR microspectroscopic analysis [139,140]. Since it is likely that some of these osteoporotic patients could have osteomalacia, we tested the hypothesis that the mineral in the osteomalacic bone was different from normal. Two groups of methacrylateembedded iliac crest biopsies were examined by FTIRI. In this preliminary study, controls were seven female subjects, aged 36–57, without apparent bone disease and four female subjects with morphologically defined osteomalacia. In tissue sections from patients with osteomalacia, mineralization rates determined by histomorphometry were decreased, while osteoid parameters were increased, revealing defective primary mineralization. In those sections, the degree of mineralization measured by qBEI in old bone (calcified tissue) was slightly higher than in four of the controls, reflecting the normal evolution of the secondary mineralization of bone tissue. The FTIRI for the biopsies showed that mineral content is decreased, not significantly, in both cortical and trabecular bone of the osteomalacic patients, with slight but not significant increases in crystallinity (Fig. 6). No differences in mineral content, crystallinity, or collagen maturity parameters were found when 400 µm × 400 µm areas of mineralized regions in osteomalacic and control specimens were compared. These findings support the hypothesis that the
490
DAN FAIBISH AND ADELE L. BOSKEY
Control
Osteomalacia
5
4
3
2
1
0 Cortical Bone
Trabecular Bone
Cortical Bone
Mineral : Matrix
Trabecular Bone
Crystallinity
FIGURE 6 Infrared imaging comparing the mean mineral content (mineral: matrix ratio) and crystallinity in normal female patients to those in female patients with osteomalacia. Values are mean ± SD (n = 7 and 4, respectively) as averaged from three to five fields from each biopsy.
inferior mechanical properties of osteomalacic bone originate in the delayed pattern of primary mineralization and the resulting smaller amount of mineralized tissue.
III. CONCLUSIONS Vitamin D affects mineralization predominantly by regulating local calcium and phosphate levels. Additionally, because vitamin D has genomic and nongenomic effects on the mineralizing tissues’ cells, vitamin D’s indirect actions on membrane properties, enzyme activity, and matrix protein expression and phosphorylation can affect the mineral that is formed in its presence or absence. A recent report [141] indicates that an analog of 1,25-(OH)2 vitamin D3 stimulates osteoblast activity in vitro and in ovariectomized rats, arguing for a direct effect of vitamin D on bone cells beyond the increase in calcium and phosphate levels discussed in this chapter. The properties of the matrix formed with this new analog appear to go beyond the stimulation of matrix protein synthesis, but the properties of the mineral in animals treated with this drug are not known. Unfortunately there is a dearth of information on the mineral properties in most of the new animal models in which mineralization is altered. It is hoped that this chapter will encourage analyses of those tissues.
References 1. Glasauer S, Langley S, Beveridge TJ 2002 Intracellular iron minerals in a dissimilatory iron-reducing bacterium. Science 295:117–119. 2. Wetherbee R 2002 Biomineralization. The diatom glasshouse. Science 298:547. 3. Wilt FH 2002 Biomineralization of the spicules of sea urchin embryos. Zoolog Sci 19:253-61. 4. Pierce DC, Butler KD, Roer RD 2001 Effects of exogenous N-acetylhexosaminidase on the structure and mineralization of the post-ecdysial exoskeleton of the blue crab, Callinectes sapidus. Comp Biochem Physiol B Biochem Mol Biol 128:691–700. 5. Boskey AL 2001 Bone mineralization. In: Cowin SC (ed) Bone Biomechanics, 3rd ed. CRC Press, Boca Raton, FL, pp. 5.1–5.34. 6. Plate U, Arnold S, Stratmann U, Wiesmann HP, Hohling HJ. 1998 General principle of ordered apatitic crystal formation in enamel and collagen rich hard tissues. Connect Tissue Res 38:149–157. 7. Beshensky AM, Wesson JA, Worcester EM, Sorokina EJ, Snyder CJ, Kleinman JG 2001 Effects of urinary macromolecules on hydroxyapatite crystal formation. J Am Soc Nephrol 12:2108–16. 8. Boskey AL, Maresca M, Doty S, Sabsay B, Veis A 1990 Concentration-dependent effects of dentin phosphophoryn in the regulation of in vitro hydroxyapatite formation and growth Bone Miner 11:55–65. 9. Smith CE 1998 Cellular and chemical events during enamel maturation. Crit Rev Oral Biol Med 9:128–61. 10. Vernois V, Deschamps N, Revel G 1989 Characterization of human dental enamel by epithermal neutron activation analysis. J Trace Elem Electrolytes Health Dis 3:67–70.
CHAPTER 27 Mineralization
11. Handschin RG, Stern WB 1994 Crystallographic and chemical analysis of human bone apatite (Crista Iliaca) Clin Rheumatol 13(Suppl 1):75–90. 12. Ou-Yang H, Paschalis EP, Mayo WE, Boskey AL, Mendelsohn R 2001 Infrared microscopic imaging of bone: spatial distribution of CO32−. J Bone Miner Res 16: 893–900. 13. Loong CK, Rey C, Kuhn LT, Combes C, Wu Y, Chen S, Glimcher MJ 2000 Evidence of hydroxyl-ion deficiency in bone apatites: an inelastic neutron-scattering study. Bone 26:599–602. 14. Cho G, Wu Y, Ackerman JL 2003 Detection of hydroxyl ions in bone mineral by solid-state NMR spectroscopy. Science 300:1123–1127. 15. Boskey AL 2002 Pathogenesis of cartilage calcification: mechanisms of crystal deposition in cartilage. Curr Rheumatol Rep 4:245–251. 16. Gerbaud V, Pignol D, Loret E, Bertrand JA, Berland Y, Fontecilla-Camps JC, Canselier JP, Gabas N, Verdier JM 2000 Mechanism of calcite crystal growth inhibition by the N-terminal undecapeptide of lithostathine. J Biol Chem 275:1057–1064. 17. Fam AG 2000 What is new about crystals other than monosodium urate? Curr Opin Rheumatol 12:228–234. 18. Rajasree S, Umashankar PR, Lal AV, Sarma PS, Kartha CC 2002 1,25-Dihydroxyvitamin D3 receptor is upregulated in aortic smooth muscle cells during hypervitaminosis D. Life Sci 70:1777–1788. 19. Price PA, Buckley JR, Williamson MK 2001 The amino bisphosphonate ibandronate prevents vitamin D toxicity and inhibits vitamin D–induced calcification of arteries, cartilage, lungs and kidneys in rats. J Nutr 131:2910–2915. 20. Rodriguez-Cano L, Garcia-Patos V, Creus M, Bastida P, Ortega JJ, Castells A 1996 Childhood calcinosis cutis. Pediatr Dermatol 13:114–117. 21. Blay P, Fernandez-Martinez JM, Diaz-Lopez B 2001 Vertebral involvement in hyperphosphatemic tumoral calcinosis. Bone 28:316–318. 22. Claudepierre P, Rahmouni A, Bergamasco P, Montazel JL, Larget-Piet B, Chevalier X 1997 Misleading clinical aspects of hydroxyapatite deposits: a series of 15 cases. J Rheumatol 24:531–535. 23. Engelse MA, Neele JM, Bronckers AL, Pannekoek H, de Vries CJ 2001 Vascular calcification: expression patterns of the osteoblast-specific gene core binding factor alpha-1 and the protective factor matrix gla protein in human atherogenesis. Cardiovasc Res 52:281–289. 24. Gokhale J, Robey PG, Boskey AL 2001 The biochemistry of bone. In: Marcus R, Feldman D, Kelsey JC (eds) Osteoporosis, 2nd ed. Vol. 1. Academic Press, San Diego, pp. 107–189. 25. Boskey AL 1992 Mineral–matrix interactions in bone and cartilage. Clin Orthop 281:244–274. 26. Potter K, Leapman RD, Basser PJ, Landis WJ 2002 Cartilage calcification studied by proton nuclear magnetic resonance microscopy. J Bone Miner Res 17:652–660. 27. Burr DB 2002 The contribution of the organic matrix to bone’s material properties. Bone 31:8–11. 28. Landis WJ 1999 An overview of vertebrate mineralization with emphasis on collagen-mineral interaction. Gravit Space Biol Bull 12:15–26. 29. Hu CC, Hart TC, Dupont BR, Chen JJ, Sun X, Qian Q, Zhang CH, Jiang H, Mattern VL, Wright JT, Simmer JP 2000. Cloning human enamelin cDNA, chromosomal localization, and analysis of expression during tooth development. J Dent Res 79:912–919.
491 30. Moradian-Oldak J 2001 Amelogenins: assembly, processing and control of crystal morphology. Matrix Biol 20:293–305. 31. Stains JP, Weber JA, Gay CV 2002 Expression of Na+/Ca2+ exchanger isoforms (NCX1 and NCX3) and plasma membrane Ca2+ ATPase during osteoblast differentiation. J Cell Biochem 84:625–635. 32. Lundquist P, Ritchie HH, Moore K, Lundgren T, Linde A Phosphate and calcium uptake by rat odontoblast-like MRPC-1 cells concomitant with mineralization. J Bone Miner Res 17:1801–1813. 33. Penel G, Leroy G, Rey C, Bres E 1998 MicroRaman spectral study of the PO4 and CO3 vibrational modes in synthetic and biological apatites. Calcif Tissue Int 63:475–481. 34. Driessens FC 1988 Physiology of hard tissues in comparison with the solubility of synthetic calcium phosphates. Ann NY Acad Sci 523:131–136. 35. Barry AB, Baig AA, Miller SC, Higuchi WI 2002 Effect of age on rat bone solubility and crystallinity. Calcif Tissue Int 71:167–671. 36. Robinson C, Shore RC, Brookes SJ, Strafford S, Wood SR, Kirkham J 2000 The chemistry of enamel caries. Crit Rev Oral Biol Med 11:481–495. 37. Patel PR, and Brown WE 1975 Thermodynamic solubility product of human tooth enamel: powdered sample. J Dent Res 54:728–736. 38. Anderson HC 1995 Molecular biology of matrix vesicles. Clin Orthop 314:266–280. 39. Boskey AL, Boyan BD, Doty SB, Feliciano A, Greer K, Weiland D, Swain LD, Schwartz Z 1992 Studies of matrix vesicle-induced mineralization in a gelatin gel. Bone Miner 17:257–262. 40. Glimcher MJ 1987 The nature of the mineral component of bone and the mechanism of calcification. Instr Course Lect 36:49–69. 41. Uchiyama T, Tanizawa T, Muramatsu H, Endo N, Takahashi HE, Hara T 1997 A morphometric comparison of trabecular structure of human ilium between microcomputed tomography and conventional histomorphometry. Calcif Tissue Int 61:493–498. 42. Daci E, Verstuyf A, Moermans K, Bouillon R, Carmeliet G 2000 Bone resorption induced by 1α,25 dihydroxyvitamin D3 in vivo is not altered by inactivation of the plasminogen activator inhibitor 1. Bone 27:97–102. 43. Lill CA, Fluegel AK, Schneider E 2000 Sheep model for fracture treatment in osteoporotic bone: a pilot study about different induction regimens. J Orthop Trauma 14:559–565. 44. Boonen S, Cheng XG, Nijs J, Nicholson PH, Verbeke G, Lesaffre E, Aerssens J, Dequeker J 1997 Factors associated with cortical and trabecular bone loss as quantified by peripheral computed tomography (pQCT) at the ultradistal radius in aging women. Calcif Tissue Int 60:164–170. 45. Pienkowski D, Doers TM, Monier-Faugere MC, Geng Z, Camacho NP, Boskey AL, Malluche HH 1997 Calcitonin alters bone quality in beagle dogs. J Bone Miner Res 12: 1936–1943. 46. Carden A, Morris MD 2000 Application of vibrational spectroscopy to the study of mineralized tissues. J Biomed Opt 5:259–268. 47. Baker D, Kupke KG, Ingram P, Roggli VL, Shelburne JD 1985 Microprobe analysis in human pathology. Scan Electron Microsc 659–680. 48. Lundon K, Grynpas M 1993 The long-term effect of ovariectomy on the quality and quantity of cortical bone in the young cynomolgus monkey: a comparison of density fractionation and histomorphometric techniques. Bone 14:389–395.
492 49. Gronowicz G, Woodiel FN, McCarthy MB, Raisz LG 1989 In vitro mineralization of fetal rat parietal bones in defined serum-free medium: effect of beta-glycerol phosphate. J Bone Miner Res 4:313–324. 50. Tsukamoto Y, Fukutani S, Shin-Ike T, Kubota T, Sato S, Suzuki Y, Mori M 1992 Mineralized nodule formation by cultures of human dental pulp–derived fibroblasts. Arch Oral Biol 37:1045–1055. 51. Costa AM, Fernandes HM 2000 Long-term effects of parathyroid hormone, 1,25-dihydroxyvitamin D3, and dexamethasone on the cell growth and functional activity of human osteogenic alveolar bone cell cultures. Pharmacol Res 42:345–353. 52. Camacho NP, Rinnerthaler S, Paschalis EP, Mendelsohn R, Boskey AL, Fratzl 1999 Complementary information on bone ultrastructure from scanning small angle X-ray scattering and Fourier-transform infrared microspectroscopy. Bone 25:287–293. 53. Gardiner EM, Baldock PA, Thomas GP, Sims NA, Henderson KN, Hollis B, White CP, Sunn KL, Morrison NA, Walsh WR, Eisman JA 2000 Increased formation and decreased resorption of bone in mice with elevated vitamin D receptor in mature cells of the osteoblastic lineage. FASEB J 14:1908–1916. 54. Eppell SJ, Tong W, Katz JL, Kuhn L, Glimcher MJ 2001 Shape and size of isolated bone mineralites measured using atomic force microscopy. J Orthop Res 19:1027–1034. 55. Tong W, Glimcher MJ, Katz JL, Kuhn L, Eppell SJ 2003 Size and shape of mineralites in young bovine bone measured by atomic force microscopy. Calcif Tissue Int 72:592–598. 56. Miller LM, Vairavamurthy V, Chance MR, Mendelsohn R, Paschalis EP, Betts F, Boskey AL 2001 In situ analysis of mineral content and crystallinity in bone using infrared micro-spectroscopy of the nu(4) PO43− vibration. Biochim Biophys Acta 1527:11–19. 57. Gadaleta SJ, Paschalis EP, Betts F, Mendelsohn R, Boskey AL 1996 Fourier transform infrared spectroscopy of the solution-mediated conversion of amorphous calcium phosphate to hydroxyapatite: new correlations between X-ray diffraction and infrared data. Calcif Tissue Int 58:9–16. 58. Grynpas MD, Rey C 1992 The effect of fluoride treatment on bone mineral crystals in the rat. Bone 13:423–429. 59. Rey C, Renugopalakrishnan V, Collins B, Glimcher MJ 1991 Fourier transform infrared spectroscopic study of the carbonate ions in bone mineral during aging. Calcif Tissue Int 49:251–258. 60. Rothwell WP, Waugh JS, Yesinowski JP 1980 High-resolution variable-temperature P-31 NMR of solid calcium phosphates. J Am Chem Soc 102:2637–2643. 61. Wu Y, Ackerman JL, Kim HM, Rey C, Barroug A, Glimcher MJ 2002 Nuclear magnetic resonance spin-spin relaxation of the crystals of bone, dental enamel, and synthetic hydroxyapatites. J Bone Miner Res 17:472–480. 62. Wu Y, Ackerman JL, Strawich ES, Rey C, Kim HM, Glimcher MJ 2003 Phosphate ions in bone: identification of a calcium-organic phosphate complex by 31P solid-state NMR spectroscopy at early stages of mineralization. Calcif Tissue Int 72:610–626. 63. Wu Y, Glimcher MJ, Rey C, Ackerman JL 1994 A unique protonated phosphate group in bone mineral not present in synthetic calcium phosphates. Identification by phosphorus31 solid state NMR spectroscopy. J Mol Biol 244:423–435. 64. Bohic S, Rey C, Legrand A, Sfihi H, Rohanizadeh R, Martel C, Barbier A, Daculsi G. 2000 Characterization of the trabecular
DAN FAIBISH AND ADELE L. BOSKEY
65.
66.
67.
68.
69.
70.
71. 72. 73.
74. 75.
76.
77.
78. 79. 80.
rat bone mineral: Effect of ovariectomy and bisphosphonate treatment. Bone 26:341–348. Misof BM, Roschger P, Tesch W, Baldock PA, Valenta A, Meßmer P, Eisman JA, Boskey AL, Gardiner EM, Fratzl P, Klaushofer K 2003 Targeted overexpression of vitamin D receptor in osteoblasts increases calcium concentration without affecting structural properties of bone mineral crystals. Calcif Tissue Int June 6, e-publication. Balmain N, von Eichel B, Toury R, Belquasmi F, Hauchecorne M, Klaus G, Mehls O, Ritz E 1995 CalbindinD28K and -D9K and 1,25(OH)2 vitamin D3 receptor immunolocalization and mineralization induction in long-term primary cultures of rat epiphysial chondrocytes. Bone 17:37–45. Chang W, Tu C, Pratt S, Chen TH, Shoback D 2002 Extracellular Ca2+-sensing receptors modulate matrix production and mineralization in chondrogenic RCJ3.1C5.18 cells. Endocrinology 143:1467–1474. Lian JB, McKee MD, Todd AM, Gerstenfeld LC 1993 Induction of bone-related proteins, osteocalcin and osteopontin, and their matrix ultrastructural localization with development of chondrocyte hypertrophy in vitro. J Cell Biochem 52:206–219. Khouja HI, Bevington A, Kemp GJ, Russell RG 1990 Calcium and orthophosphate deposits in vitro do not imply osteoblastmediated mineralization: mineralization by betaglycerophosphate in the absence of osteoblasts. Bone 11:385–391. Bonewald LF, Harris SE, Rosser J, Dallas MR, Dallas SL, Camacho NP, Boyan B, Boskey A 2003 Von Kossa staining alone is not sufficient to confirm that mineralization in vitro represents bone formation. Calcif Tissue Int 72:537–547. Kingma JG Jr, Roy PE 1988 Ultrastructural study of hypervitaminosis D induced arterial calcification in Wistar rats. Artery 16:51–61. Rambeck WA, Strohle HG, Wetzel A, Zucker H 1981 Calcinogenic activity of vitamin D3 and vitamin D3 palmitate in rat and rabbit. Int J Vitam Nutr Res 51:359–364. Jiang YB, Wang YZ, Zhao J, Marchal G, Wang YX, Shen YL, Xing SZ, Zhang XY, Baert AL 1990 Metastatic calcification within bone. The main cause of osteosclerosis in hypervitaminosis D3. Radiologic–pathologic correlation. Invest Radiol 25:1188–1196. Giunta JL 1998 Dental changes in hypervitaminosis D. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 85:410–413. Amling M, Priemel M, Holzmann T, Chapin K. Rueger JM, Baron R, Demay MB 1999 Rescue of the skeletal phenotype of vitamin D receptor-ablated mice in the setting of normal mineral ion homeostasis: formal histomorphometric and biomechanical analyses. Endocrinology 140:4982–4987. Weinstein RS, Underwood JL, Hutson MS, DeLuca HF 1984 Bone histomorphometry in vitamin D–deficient rats infused with calcium and phosphorus. Am J Physiol 246: E499–E505. Holtrop ME, Cox KA, Carnes DL, Holick MF 1986 Effects of serum calcium and phosphorus on skeletal mineralization in vitamin D–deficient rats. Am J Physiol 251: E234–E240. Boskey AL 1989 Hydroxyapatite formation in a dynamic gel system: Effects of type I collagen, lipids, and proteoglycans. J Phys Chem 93:1628–1633. Giachelli CM, Jono S, Shioi A, Nishizawa Y, Mori K, Morii H 2001 Vascular calcification and inorganic phosphate. Am J Kidney Dis 38:S34–S377. Kato Y, Boskey A, Spevak L, Dallas M, Hori M, Bonewald LF 2001 Establishment of an osteoid preosteocyte-like cell
CHAPTER 27 Mineralization
81.
82.
83.
84. 85.
86.
87. 88.
89.
90.
91.
92.
93.
94.
95.
MLO-A5 that spontaneously mineralizes in culture. J Bone Miner Res 16:1622–1633. Shane E, Parisien M, Henderson JE, Dempster DW, Feldman F, Hardy MA, Tohme JF, Karaplis AC, Clemens TL 1997 Tumor-induced osteomalacia: clinical and basic studies. J Bone Miner Res 12:1502–1511. Siris ES, Clemens TL, Dempster DW, Shane E, Segre GV, Lindsay R, Bilezikian JP 1987 Tumor-induced osteomalacia. Kinetics of calcium, phosphorus, and vitamin D metabolism and characteristics of bone histomorphometry. Am J Med 82: 307–312. Holtrop ME, Cox KA, Clark MB, Holick MF, Anast CS 1981 1,25-Dihydroxycholecalciferol stimulates osteoclasts in rat bones in the absence of parathyroid hormone. Endocrinology 108:2293–2301. Newbrey JW, Truitt ST, Roland DA, Frost TJ, Untawale GG 1992 Bone histomorphometry in 1,25(OH)2D3- and vitamin D3– treated aged laying hens. Avian Dis 36:700–706. Denis I, Cournot G, Lacroix H, Colin C, Zerath E, Pointillart A 2000 In vivo bone metabolism and ex vivo bone marrow osteoprogenitors in vitamin D–deprived pigs. Bone 26: 491–498. Farach-Carson MC, Xu Y 2002 Microarray detection of gene expression changes induced by 1,25(OH)2D3 and a Ca2+ influxactivating analog in osteoblastic ROS 17/2.8 cells. Steroids 67:467–470. Binderman I, Somjen D 1984 24,25-Dihydroxycholecalciferol induces the growth of chick cartilage in vitro. Endocrinology 115:430–432. Nakamura T, Suzuki K, Hirai T, Kurokawa T, Orimo H 1992 Increased bone volume and reduced bone turnover in vitamin D–replete rabbits by the administration of 24R,25-dihydroxyvitamin D3. Bone 13:229–236. Erben RG, Scutt AM, Miao D, Kollenkirchen U, Haberey M 1997 Short-term treatment of rats with high dose 1,25-dihydroxyvitamin D3 stimulates bone formation and increases the number of osteoblast precursor cells in bone marrow. Endocrinology 138:4629–4635. Hale LV, Kemick ML, Wuthier RE 1986 Effect of vitamin D metabolites on the expression of alkaline phosphatase activity by epiphyseal hypertrophic chondrocytes in primary cell culture. J Bone Miner Res 1:489–495. Boskey AL, Maresca M, Ullrich W, Doty SB, Butler WT, Prince CW 1993 Osteopontin-hydroxyapatite interactions in vitro: inhibition of hydroxyapatite formation and growth in a gelatin-gel. Bone Miner 22:147–159. Mahonen A, Jukkola A, Risteli L, Risteli J, Maenpaa PH 1998 Type I procollagen synthesis is regulated by steroids and related hormones in human osteosarcoma cells. J Cell Biochem 68:151–163. Pavlin D, Bedalov A, Kronenberg MS, Kream BE, Rowe DW, Smith CL, Pike JW, Lichtler AC 1994 Analysis of regulatory regions in the COL1A1 gene responsible for 1,25-dihydroxyvitamin D3–mediated transcriptional repression in osteoblastic cells. J Cell Biochem 56:490–501. Bellows CG, Reimers SM, Heersche JN 1999 Expression of mRNAs for type-I collagen, bone sialoprotein, osteocalcin, and osteopontin at different stages of osteoblastic differentiation and their regulation by 1,25 dihydroxyvitamin D3. Cell Tissue Res 297:249–259. Broess M, Riva A, Gerstenfeld LC 1995 Inhibitory effects of 1,25(OH)2 vitamin D3 on collagen type I, osteopontin, and osteocalcin gene expression in chicken osteoblasts. J Cell Biochem 57:440–451.
493 96. Matsumoto T, Igarashi C, Takeuchi Y, Harada S, Kikuchi T, Yamato H, Ogata E 1991 Stimulation by 1,25-dihydroxyvitamin D3 of in vitro mineralization induced by osteoblast-like MC3T3-E1 cells. Bone 12:27–32. 97. Tye CE, Rattray KR, Warner KJ, Gordon JA, Sodek J, Hunter GK, Goldberg HA 2003 Delineation of the hydroxyapatitenucleating domains of bone sialoprotein. J Biol Chem 278:7949–7955. 98. Chen JJ, Jin H, Ranly DM, Sodek J, Boyan BD 1999 Altered expression of bone sialoproteins in vitamin D–deficient rBSP2.7Luc transgenic mice. J Bone Miner Res 14: 22122–22129. 99. Yang R, Gerstenfeld LC 1997 Structural analysis and characterization of tissue and hormonal responsive expression of the avian bone sialoprotein (BSP) gene. J Cell Biochem 64:77–93. 100. Sodek J, Kim RH, Ogata Y, Li J, Yamauchi M, Zhang Q, Freedman LP 1996 Regulation of bone sialoprotein gene transcription by steroid hormones. Connect Tissue Res 32:209–217. 101. White C, Gardiner E, Eisman J 1999 Tissue specific and vitamin D responsive gene expression in bone. Mol Biol Rep 25:45–61. 102. Price PA, Baukol SA 1981 1,25-Dihydroxyvitamin D3 increases serum levels of the vitamin K–dependent bone protein. Biochem Biophys Res Commun 99:928–935. 103. Chen TL, Fry D 1999 Hormonal regulation of the osteoblastic phenotype expression in neonatal murine calvarial cells. Calcif Tissue Int 64:304–309. 104. Boskey AL, Wians FH Jr, Hauschka PV 1985 The effect of osteocalcin on in vitro lipid-induced hydroxyapatite formation and seeded hydroxyapatite growth. Calcif Tissue Int 37:57–62. 105. Poser JW, Price PA 1979 A method for decarboxylation of gamma-carboxyglutamic acid in proteins. Properties of the decarboxylated gamma-carboxyglutamic acid protein from calf bone. J Biol Chem 254:431–436. 106. Ducy P, Desbois C, Boyce B, Pinero G, Story B, Dunstan C, Smith E, Bonadio J, Goldstein S, Gundberg C, Bradley A, Karsenty G 1996 Increased bone formation in osteocalcindeficient mice. Nature 382:448–452. 107. Boskey AL, Gadaleta S, Gundberg C, Doty SB, Ducy P, Karsenty G 1998 Fourier transform infrared microspectroscopic analysis of bones of osteocalcin-deficient mice provides insight into the function of osteocalcin. Bone 23:187–196. 108. Carpenter TO, Moltz KC, Ellis B, Andreoli M, McCarthy TL, Centrella M, Bryan D, Gundberg CM 1998 Osteocalcin production in primary osteoblast cultures derived from normal and Hyp mice. Endocrinology 139:35–43. 109. Giachelli CM and Steitz S 2000 Osteopontin: a versatile regulator of inflammation and biomineralization. Matrix Biol 19:615–622. 110. Lian JB, Shalhoub V, Aslam F, Frenkel B, Green J, Hamrah M, Stein GS, Stein JL 1997 Species-specific glucocorticoid and 1,25-dihydroxyvitamin D responsiveness in mouse MC3T3-E1 osteoblasts: Dexamethasone inhibits osteoblast differentiation and vitamin D down-regulates osteocalcin gene expression. Endocrinology 138:2117–2127. 111. Boskey AL, Spevak L, Paschalis E, Doty SB, McKee MD 2002 Osteopontin deficiency increases mineral content and mineral crystallinity in mouse bone. Calcif Tissue Int 71:145–154. 112. Berry JL, Davies M, Mee AP 2002 Vitamin D metabolism, rickets, and osteomalacia. Semin Musculoskelet Radiol 6:173–182.
494 113. Mawer EB, Davies M 2001 Vitamin D nutrition and bone disease in adults. Rev Endocr Metab Disord 2:153–164. 114. Pettifor JM 2002 Rickets. Calcif Tissue Int 70:398–399. 115. Evliyaoglu O, Berberoglu M, Ocal G, Adiyaman P, Aycan Z 2001 Severe hypercalcemia of an infant due to vitamin D toxicity associated with hypercholesterolemia. J Pediatr Endocrinol Metab 14:915–919. 116. Boskey AL, Di Carlo EF, Gilder H, Donnelly R, Weintroub S 1988 The effect of short-term treatment with vitamin D metabolites on bone lipid and mineral composition in healing vitamin D-deficient rats. Bone 9:309–318. 117. Boskey AL, Gilder H, Neufeld E, Ecarot B, Glorieux FH 1991 Phospholipid changes in the bones of the hypophosphatemic mouse. Bone 12:345–351. 118. Boskey AL, Dickson IR 1988 Influence of vitamin D status on the content of complexed acidic phospholipids in chick diaphyseal bone. Bone Miner 4:365–371. 119. Donnelly R, Bockman R, DiCarlo E, Betts F, Boskey A 1993 The effect of gallium nitrate on healing of vitamin D- and phosphate-deficient rickets in the immature rat. Calcif Tissue Int 53:400–410. 120. Anderson H.C, Cecil R, Sajdera SW 1975 Calcification of rachitic rat cartilage in vitro by extracellular matrix vesicles. Am J Pathol 79:237–254. 121. Shapiro IM, Boyde A 1984 Microdissection—elemental analysis of the mineralizing growth cartilage of the normal and rachitic chick. Metab Bone Dis Relat Res 5:317–326. 122. Xu H, Bai L, Collins JF, Ghishan FK 2002 Age-dependent regulation of rat intestinal type IIb sodium-phosphate cotransporter by 1,25-(OH)2 vitamin D3. Am J Physiol Cell Physiol 282:C487–C493. 123. Argiro L, Desbarats M, Glorieux FH, Ecarot B 2001 Mepe, the gene encoding a tumor-secreted protein in oncogenic hypophosphatemic osteomalacia, is expressed in bone. Genomics 74:342–351. 124. Campos M, Couture C, Hirata IY, Juliano MA, Loisel TP, Crine P, Juliano L, Boileau G, Carmona AK 2003 Human recombinant PHEX has a strict S1′ specificity for acidic residues and cleaves peptides derived from FGF-23 and MEPE. Biochem J [epub ahead of print] 125. Ecarot B, Desbarats M 1999 1,25-(OH)2D3 down-regulates expression of Phex, a marker of the mature osteoblast. Endocrinology 140:1192–1199. 126. Kato S, Takeyama K, Kitanaka S, Murayama A, Sekine K, Yoshizawa T 1999 In vivo function of VDR in gene expressionVDR knock-out mice. J Steroid Biochem Mol Biol 69: 247–251. 127. Masuyama R, Nakaya Y, Tanaka S, Tsurukami H, Nakamura T, Watanabe S, Yoshizawa T, Kato S, Suzuki K 2001 Dietary phosphorus restriction reverses the impaired bone mineralization in vitamin D receptor knockout mice. Endocrinology 142:494–497. 128. Li YC, Pirro AE, Amling M, Delling G, Baron R, Bronson R, Demay MB 1997 Targeted ablation of the vitamin D receptor: An animal model of vitamin D–dependent rickets type II with alopecia. Proc Natl Acad Sci USA 94:9831–9835. 129. Yoshizawa T, Handa Y, Uematsu Y, Takeda S, Sekine K, Yoshihara Y, Kawakami T, Arioka K, Sato H, Uchiyama Y, Masushige S, Fukamizu A, Matsumoto T, Kato S 1997 Mice lacking the vitamin D receptor exhibit impaired bone formation, uterine hypoplasia and growth retardation after weaning. Nat Gene 16:391–396. 130. Donohue MM, Demay MB 2002 Rickets in VDR null mice is secondary to decreased apoptosis of hypertrophic chondrocytes. Endocrinology 143:3691–3694.
DAN FAIBISH AND ADELE L. BOSKEY
131. Mechica JB, Leite MO, Mendonca BB, Frazzatto ES, Borelli A, Latronico AC 1997 A novel nonsense mutation in the first zinc finger of the vitamin D receptor causing hereditary 1,25dihydroxyvitamin D3–resistant rickets. J Clin Endocrinol Metab 82:3892–3894. 132. Cockerill FJ, Hawa NS, Yousaf N, Hewison M, O’Riordan JL, Farrow SM 1997 Mutations in the vitamin D receptor gene in three kindreds associated with hereditary vitamin D resistant rickets. J Clin Endocrinol Metab 82:3156–3160. 133. Safadi FF, Thornton P, Magiera H, Hollis BW, Gentile M, Haddad JG, Liebhaber SA, Cooke NE 1999 Osteopathy and resistance to vitamin D toxicity in mice null for vitamin D binding protein. J Clin Invest 103:239–251. 134. Panda D, Miao ML, Tremblay J, Sirois R, Farookhi GN, Hendy D, Goltzman D 2001 Targeted ablation of the 25-hydroxyvitamin D 1α-hydroxylase enzyme: Evidence for skeletal, reproductive, and immune dysfunction. Proc Natl Acad Sci USA 98:7498–7503. 135. Fraser D, Kooh SW, Kind HP, Holick MF, Tanaka Y, DeLuca HF 1973 Pathogenesis of hereditary vitamin-D-dependent rickets. An inborn error of vitamin D metabolism involving defective conversion of 25-hydroxyvitamin D to 1α,25-dihydroxyvitamin D. N Engl J Med 289:817–822. 136. Dardenne O, Prud’homme J, Hacking SA, Glorieux FH, St-Arnaud R 2003 Correction of the abnormal mineral ion homeostasis with a high-calcium, high-phosphorus, highlactose diet rescues the PDDR phenotype of mice deficient for the 25-hydroxyvitamin D-1α-hydroxylase (CYP27B1). Bone 32:332–340. 137. Nordin BE, Morris HA 1989 The calcium deficiency model for osteoporosis. Nutr Rev 47:65–72. 138. Gennari C 2001 Calcium and vitamin D nutrition and bone disease of the elderly. Public Health Nutr 4:547–559. 139. Paschalis EP, Betts F, DiCarlo E, Mendelsohn R, Boskey AL 1997 FTIR microspectroscopic analysis of human iliac crest biopsies from untreated osteoporotic bone. Calcif Tissue Int 61:487–492. 140. Gadeleta SJ, Boskey AL, Paschalis E, Carlson C, Menschik F, Baldini T, Peterson M, Rimnac CM 2000 A physical, chemical, and mechanical study of lumbar vertebrae from normal, ovariectomized, and nandrolone decanoate-treated cynomolgus monkeys (Macaca fascicularis). Bone 27:541–550. 141. Shevde NK, Plum LA, Clagett-Dame M, Yamamoto H, Pike JW, DeLuca HF 2002 A potent analog of 1α,25-dihydroxyvitamin D3 selectively induces bone formation. Proc Natl Acad Sci USA 99:13487–13491. 142. Schinke T, Amendt C, Trindl A, Poschke O, Muller-Esterl W, Jahnen-Dechent W 1996 The serum protein α2-HS glycoprotein/fetuin inhibits apatite formation in vitro and in mineralizing calvaria cells. A possible role in mineralization and calcium homeostasis. J Biol Chem 271:20789–20796. 143. Wu LN, Genge BR, Dunkelberger DG, LeGeros RZ, Concannon B, Wuthier RE 1997 Physicochemical characterization of the nucleational core of matrix vesicles. J Biol Chem 272:4404–4411. 144. Boskey AL, Spevak L, Doty SB, Rosenberg L 1997 Effects of bone CS-proteoglycans DS-decorin, and DS-biglycan on hydroxyapatite formation in a gelatin gel. Calcif Tissue Int 61:298–305. 145. Lidor C, Edelstein S 1987 Calcitriol increases Ca2+-ATPase activity. Biochem Biophys Res Commun 144:713–717. 146. Dean DD, Schwartz Z, Schmitz J, Muniz OE, Lu Y, Calderon F, Howell DS, Boyan BD 1996 Vitamin D regulation of metalloproteinase activity in matrix vesicles. Connect Tissue Res 35:331–336.
CHAPTER 27 Mineralization
147. Uchida M, Shima M, Chikazu D, Fujieda A, Obara K, Suzuki H, Nagai Y, Yamato H, Kawaguchi H 2001 Transcriptional induction of matrix metalloproteinase-13 (collagenase-3) by 1α,25-dihydroxyvitamin D3 in mouse osteoblastic MC3T3-E1 cells. J Bone Miner Res 16:221–230. 148. Lin R, Amizuka N, Sasaki T, Aarts MM, Ozawa H, Goltzman D, Henderson JE, White JH 2002 1α,25-Dihydroxyvitamin D3 promotes vascularization of the chondro-osseous junction by stimulating expression of vascular endothelial growth factor and matrix metalloproteinase 9. J Bone Miner Res 17:1604–1612. 149. Papagerakis P, MacDougall M, Berdal A 2002 Differential epithelial and mesenchymal regulation of tooth-specific matrix proteins expression by 1,25-dihydroxyvitamin D3 in vivo. Connect Tissue Res 43:372–375. 150. Farzaneh-Far A, Weissberg PL, Proudfoot D, Shanahan CM 2001 Transcriptional regulation of matrix gla protein. Z Kardiol 90:s38–s42. 151. Elfahime E, Felix JM, Koch B 1996 Regulation of corticosteroid-binding globulin synthesis by 1α,25-dihyroxyvitamin D3 (calcitriol), 9-cis-retinoic acid and triiodothyronine in cultured rat fetal hepatocytes. J Steroid Biochem Mol Biol 57:109–115. 152. Horton WE Jr, Balakir R, Precht P, Liang CT 1991 1,25Dihydroxyvitamin D3 down-regulates aggrecan proteoglycan expression in immortalized rat chondrocytes through a posttranscriptional mechanism. J Biol Chem 266:24804–24808. 153. Rickard DJ, Kazhdan I, Leboy PS 1995 Importance of 1,25dihydroxyvitamin D3 and the nonadherent cells of marrow for osteoblast differentiation from rat marrow stromal cells. Bone 16:671–678.
495 154. Camacho NP, Carroll P, Raggio CL 2003 Fourier transform infrared imaging spectroscopy (FT-IRIS) of mineralization in bisphosphonate-treated oim/oim mice. Calcif Tissue Int 72:604–609. 155. Boskey AL, Moore DJ, Amling M, Canalis E, Delany A 2003 Infrared analysis of the mineral and matrix in bones of osteonectin-null mice and their wildtype controls. J Bone Miner Res 18:1005–1011. 156. Paine ML, Zhu DH, Luo W, Bringas P Jr, Goldberg M, White SN, Lei YP, Sarikaya M, Fong HK, Snead ML 2000 Enamel biomineralization defects result from alterations to amelogenin self-assembly. J Struct Biol 132:191–200. 157. Gibson CW, Yuan ZA, Hall B, Longenecker G, Chen E, Thyagarajan T, Sreenath T, Wright JT, Decker S, Piddington R, Harrison G, Kulkarni AB 2001 Amelogenin-deficient mice display an amelogenesis imperfecta phenotype. J Biol Chem 276:31871–31875. 158. Xu T, Bianco P, Fisher LW, Longenecker G, Smith E, Goldstein S, Bonadio J, Boskey A, Heegaard AM, Sommer B, Satomura K, Dominguez P, Zhao C, Kulkarni AB, Robey PG, Young MF 1998 Targeted disruption of the biglycan gene leads to an osteoporosis-like phenotype in mice. Nat Genet 20:78–82. 159. Narisawa S, Frohlander N, Millan JL 1997 Inactivation of two mouse alkaline phosphatase genes and establishment of a model of infantile hypophosphatasia. Dev Dyn 208: 432–446. 160. Somerville RPT, Oblander SA, Apte SS. 2003 Matrix metalloproteinases: old dogs with new tricks. Genome Biology 4:216–238.
CHAPTER 28
Modeling and Remodeling: How Bone Cells Work Together A. M. PARFITT
I. II. III. IV.
Division of Endocrinology and Center for Osteoporosis and Metabolic Bone Disease, University of Arkansas for Medical Sciences
Introduction The Structural and Cellular Basis of Bone Growth The Purposes of Bone Remodeling The Basic Multicellular Unit as the Instrument of Bone Remodeling
I. INTRODUCTION For many years bone physiology has been thought of in terms of two opposing but otherwise unrelated kinds of cell, often depicted as sitting on either side of a seesaw. The assumption that osteoclasts and osteoblasts are independent still dominates the field, although it was demolished more than 30 years ago by Harold Frost [1], who recognized that the production and activity of these cells were coordinated in time and space in different modes for different biological purposes, the major modes being redistribution, repair, and replacement (Table I). During growth, bone is formed in one location and after a few weeks or months resorbed in a location that is different relative to the increased size and altered shape of the bone, not because there is anything wrong with it, but because it is no longer needed at that location. As bones grow in length, bone formed at the junction between the growth plate and the metaphysis is resorbed at the junction between the metaphysis and the diaphysis. As bones grow in width, bone formed beneath the periosteum is resorbed at the endosteum. At these various relative locations, resorption and formation continue with only brief interruptions for extended periods; the cells are operating in the modeling mode [2]. In fracture
Parts of this chapter have been previously published in Parfitt AM 1997 Genetic effects on bone mass and turnover-relevance to black/white differences. J Am Coll Nutr 16:325–333, and Parfitt AM 2004 New concepts of bone remodeling: A unified spatial and temporal model with physiological and pathophysiologic implications. In: Agarwal S, Stout S (eds) Kluwer Academic/Plenum New York (in press). Reproduced with permission of the publishers. VITAMIN D, 2ND EDITION FELDMAN, PIKE, AND GLORIEUX
V. Disordered Remodeling and Age-Related Bone Loss VI. Possible Targets for Vitamin D Action VII. Conclusions References
healing, damaged bone is removed and callus in the form of woven bone is laid down; the prime consideration is speed of production and the quality of the bone is of lesser importance. Dead bone can survive for thousands of years, but living bone gradually loses mechanical competence with increasing age and must be periodically replaced by new bone. The replacement mechanism in the adult skeleton, in which osteoclasts and osteoblasts are cooperating for a common purpose, is referred to as remodeling. The salient features of modeling and remodeling are compared in Table II. In modeling, bone is first formed and then resorbed soon after, but in remodeling, bone is first resorbed and then replaced at the same location. Growth is uninterrupted but remodeling occurs infrequently; episodes lasting a few months are usually separated by several years of quiescence. In both modeling and remodeling the rates of resorption and formation are both much higher than the differences between them, but in modeling the processes occur on different surfaces and any matching is genetic or systemic, whereas in remodeling the processes occur on the same surface, and there is local coupling. Modeling leads to net gain in bone mass, but remodeling usually is associated with net loss. Modeling and remodeling are different cellular mechanisms of bone turnover, which is the volume replacement of bone, usually expressed as a fractional rate constant (e.g., %/y), regardless of cellular mechanism. Modeling is the principal mode of bone cell coordination in the growing skeleton and remodeling is the principal mode in the mature skeleton, but this distinction is not absolute. The formation of secondary osteons in the long bone cortices by remodeling [1] and remodeling of cancellous bone in the central skeleton [3] both Copyright © 2005, Elsevier, Inc. All rights reserved.
498
A. M. PARFITT
TABLE I Supracellular Organization of Bone Cells Mode
Need
Redistribution (modeling) Repair (fracture callus) Replacement (remodeling)
Equal quality bone In a different place Any bone In the right place Better bone In the same place
begin during childhood and continue throughout life. After completion of longitudinal growth many bones continue to enlarge very slowly, either by maintenance at a slow rate of periosteal apposition in the modeling mode, or by very slow remodeling with prolonged interruptions [4]. However, remodeling is primarily a phenomenon of the endosteal envelope of the mature skeleton. In both modeling and remodeling osteoclastic bone resorption involves dissolution of the mineral followed soon after by depolymerization of glycosaminoglycans and digestion of collagen and other matrix proteins; at the light microscopic level, mineral and matrix appear to be removed simultaneously [1]. By contrast, osteoblastic bone formation occurs in two distinct stages separated in time and space. Bone matrix is first synthesized and organized, and exists as unmineralized osteoid for a variable time before the onset of mineralization, as explained in more detail in Chapter 63. Bone formed at a location where there is no existing bone is woven, with irregular distribution of collagen fibers, but bone laid down in apposition to an existing bone surface is lamellar and with regular orientation of collagen fibers [1].
TABLE II Comparison of Two Major Modes of Supracellular Organization of Bone Cells Feature Mode Purpose Context Sequence Timing Location Coupling Balance
Modeling
Remodeling
Redistribution Growth Youth F→R Continuous Different Systemic Net gain
Replacement Maintenance Maturity A → R →F Cyclical Same Local Net loss
F, Formation; R, resorption; A, activation. (For explanation of different definitions of this term, see text.)
II. THE STRUCTURAL AND CELLULAR BASIS OF BONE GROWTH During embryonic development the bones of the postcranial skeleton begin as cartilage models of particular size, shape, and location, based on genetic instructions acting via diffusible morphogens [5]. These are gradually replaced by bone as they enlarge, in accordance with the changing mechanical requirements that result from growth in length and increasing muscle strength [5]. The manner of replacement differs in detail between bones, but in a typical future long bone, islands of spiculated bone appear within the ends to form the future epiphyses and an elongated collar of more closely packed bone appears beneath the perichondrium to form the future diaphysis [6]. Cortical bone is mainly formed beneath the periosteum and is mainly resorbed at what has become the endosteum, so that diaphysial cortical bone represents a by-product of growth in width [7,8]. However, to maintain the flared waist-like shape of the metaphysis, there must also be localized endosteal apposition and periosteal resorption [8]. When growth stops, the most recently formed cortical bone becomes the permanent diaphysis. Endochondral ossification is described in Chapter 33, but some aspects especially pertinent to the understanding of Vitamin D effects are summarized here. Between the epiphysis and metaphysis of a growing long bone lies the cartilaginous growth plate (Fig. 1), which is organized vertically as transverse zones and horizontally as vertical columns of chondrocytes, embedded in a honeycomb of matrix [9–11]; in the proximal tibia of a young rat the plate is about 600 µm thick. The sequence of layers known as resting, proliferative, and hypertrophic zones, and the zones of provisional calcification and vascular invasion, represent a sequence of events that occur in the same absolute location, while the epiphysis retreats and the metaphysis approaches. At the upper (or distal) end of the growth plate new cartilage is produced and at the lower (or proximal) end old cartilage is replaced by new bone. For most of the growth period these processes occur at the same rate so that the width of the plate does not change. A typical chondrocyte arises from division of a stem cell in the resting zone, divides once or twice more [12], and increases in volume about 10-fold, all the while synthesizing new matrix [10]. After about 40 hr, corresponding to a growth rate of about 350 µm/day, the longitudinal cartilaginous septa become mineralized. Contrary to previous belief, this requires that the chondrocytes remain viable until mineralization in their vicinity is complete [10,13], when they undergo apoptosis. Ingrowth of capillaries brings macrophages that resorb the transverse unmineralized septa, and chondroclasts
499
CHAPTER 28 Modeling and Remodeling: How Bone Cells Work Together
Secondary ossification centre Primary spongiosum and resorption front
Articular cartilage
Epiphyseal growth plate
Metaphysis
Reserve zone (Germinal layer)
Proliferative zone
Prehypertrophic zone
Hypertrophic zone a) Maturation Diaphysis b) Degeneration (Apotosis)
c) Provisional calcification
FIGURE 1 The epiphyseal growth plate. The right-hand panel shows a detailed representation of chondrocytes in the individual zones of the growth plate. From Stevens and Williams [11], reprinted with permission.
that resorb about two-thirds of the mineralized septa. Osteoblasts migrate from the adjacent bone marrow and deposit loosely textured woven bone on the surface of the remaining one-third of the mineralized cartilage septa, to form the primary spongiosa [9–11]. Continued minimodeling [14] replaces the cartilage cores with bone to form the secondary spongiosa, and the trabeculae are progressively resorbed so that after a few weeks the metaphysis has been replaced by diaphyseal marrow [15]. As growth slows, the life span of the secondary spongiosa continues to lengthen, and when longitudinal growth ceases, the most recently formed spongiosa becomes the permanent metaphysis. Metaphyseal cancellous bone thus represents a by-product of growth in length. Bones of the central skeleton grow in a somewhat different manner. In the vertebral bodies cancellous bone formed beneath the two growth plates remains in continuity at the centrum, the adult bone resembling the fusion of two epiphyses and two metaphyses with no intervening diaphysis [16]. Unlike the long-bone metaphyses, newly made cancellous bone is not resorbed but persists into adult life and cancellous tissue fills the whole interior of the body. As in the long bones, the cortices grow outward by circumferential periosteal apposition, but unlike the long bones the inner region of the cortex is not completely resorbed, but is converted into cancellous bone that remains in
continuity with the cancellous bone produced at the growth plate [3]. The process of cancellization of the inner cortex, which is an important component of agerelated bone loss, begins during growth in the central skeleton. In the ilium (Fig. 2) the outer cortex behaves as in a vertebral body, but at the inner cortex the periosteal surface is resorbed and the endocortical surface undergoes formation, incorporating some existing trabeculae by compaction [3]. Growth of the pelvis entails movement of the entire ilium away from the midline. The final length of a bone will be influenced by the rate of growth and by the time during which growth continues prior to epiphyseal fusion. The rate of individual bone elongation is governed by the rate of cartilage cell proliferation in the resting zones, which is the primary target of growth hormone and its local mediator insulinlike growth factor-I (IGFI) [17], and growth continues until cartilage cell proliferation ceases, following which the epiphyses fuse [18]. To maintain the ratio of width to length within its narrow range, the rate and duration of periosteal apposition must be linked in some way to the rate and duration of cartilage cell proliferation. This is the main determinant of the final amount of bone in the bone, but several other mechanisms contribute to individual variation. The number of trabeculae formed will depend on the number of capillaries that invade
500
A. M. PARFITT
OUTER
INNER Ps
2y
Ec
20y Ps
8
6
Ec
Ec
4
Ps
Ec
Ps
2
0
mm
FIGURE 2 Model of iliac growth. Diagrammatic representation of iliac growth showing inferred changes in location of periosteal (Ps) and endocortical (Ec) surfaces of outer and inner cortices from age 2 to age 20. Left vertical dashed line indicates initial location of outer Ec surface and extent of expansion in cancellous width. All cancellous bone to the left of this line represents unresorbed cortical bone. Right vertical dashed line indicates final location of inner Ec surface. All initial cancellous bone to the right of this line was incorporated into the cortex by compaction. Reprinted from Parfitt et al. [3] with permission.
the zone of provisional calcification, and the thickness of trabeculae will reflect the balance between apoptotic death and survival of cartilage cells in the hypertrophic zone [19]. During transformation of the primary to the secondary spongiosa, many trabeculae are removed. The extent and timing of this process are genetically determined [20] and influenced by mechanical loading [21]. At the completion of growth, the distance from the articular cartilage to the metaphyseal–diaphyseal junction, which influences the total amount of metaphyseal cancellous bone, is broadly proportional to the length of the bone, but is also subject to individual variation. An important but inadequately studied question is the extent to which individual trabeculae become thicker during skeletal growth. In the peripheral skeleton, the life span of trabeculae is too short for their thickening to contribute to peak bone mass until a few years before growth ceases, but the situation is different in the central skeleton. In iliac bone biopsies in healthy subjects aged 2 to 17 years, there was a progressive increase in trabecular thickness of about 60 µm, the result of rapid remodeling with positive balance [3]. Interpretation of these data is complicated by the changing location of the biopsy site in relation to the acetabulum, but there is no substantial gradient in trabecular thickness in the adult ilium. Consequently, formation of initially thicker trabeculae with increasing age would not account for the data. Although this has not been demonstrated prospectively, it is probable that trabeculae do increase in thickness during growth. This inference is supported
by the modest prepubertal increase in 3D vertebral cancellous bone density determined by QCT, which accelerates during the pubertal growth spurt [22]. Since the region of interest is far from the growth plates, new trabeculae could not be formed and the increase in apparent density could only result from increased thickness of existing trabeculae. The final thickness of diaphyseal cortical bone is the difference between the thickness of bone formed at the periosteum and the thickness of bone resorbed at the endocortical surface. Periosteal apposition depends on osteoblast precursor proliferation, and as previously mentioned is tied to growth plate cartilage cell proliferation, most likely because the same genes control proliferation in response to hormonal stimulation. During the pubertal growth spurt, as well as peak height velocity, there is peak bone width velocity [23], driven by the surges in gonadal hormonal secretion [7] and potentiated by physical activity [5,7]. Bone is also added to the endocortical surface with narrowing of the marrow cavity, the only time in life when this occurs. But gain in bone width and cortical thickness can be partly disassociated from growth in length by other factors. For example, if net absorption of calcium is insufficient to meet the demands of bone growth, net endocortical resorption will increase and cortical width fail to increase or even decrease. But above a certain threshold, further increases in net absorption are not accompanied by increased calcium retention [24], indicating that endocortical resorption cannot be reduced below the level dictated by the growth timetable. No such limitation applies to physical activity, which amplifies all the effects of pubertal growth with increased endocortical as well as periosteal gain in the extremities [7], and increased trabecular thickness in the spine [25].
III. THE PURPOSES OF BONE REMODELING The reasons for bone replacement depend on the function of the bone that is replaced [26]. The primary function of bone is to resist mechanical loads; such structural bone includes all cortical bone, and cancellous bone in the long-bone metaphyses which transmits the loads on the synovial joints to the thick cortical bone above or below. All such bone is adjacent to yellow fatty marrow, in contrast to the red hematopoietic marrow adjacent to central cancellous bone [26]. The two kinds of marrow correspond to two kinds of cancellous bone, referred to as structural and metabolic to indicate their principal but not their only function (Table III); vertebral cancellous bone is load bearing, but contributes less to compressive strength than the cortical bone
CHAPTER 28 Modeling and Remodeling: How Bone Cells Work Together
TABLE III
Significance of Marrow Composition for Cancellous Bone
Feature
Yellow Marrow
Red Marrow
Bone type Location Principal Functions Cellularity Blood flow Turnover
Structural Peripheral Transmit loads Absorb energy Low Low Low
Metabolic Central Calcium homeostasis Support hematopoiesis High High High
surrounding it [27]. Note especially the difference in turnover—cortical bone in the ilium has higher turnover than cancellous bone in the extremities, so that biology trumps geometry [28]! Because structural and metabolic bones differ in their function, they differ in their reasons for replacement (Table IV). Structural bone, both cortical and cancellous, like all load-bearing materials, is subject to fatigue damage after a certain number of load-bearing cycles [29], but unlike manmade structures, it has its own mechanism of self-repair. There may also be nontargeted remodeling to maintain bone age below some upper limit, referred to here as spare, in the sense that its abrogation will not have immediate harmful effects, whereas the abrogation of targeted remodeling may allow fatigue damage to spread and accumulate into an overt fracture [30]. The reasons for the much higher turnover of metabolic cancellous bone are less well understood. The exchange of calcium ions at the bone surface, which is an essential element of plasma calcium homeostasis [31], is impaired by the gradual enlargement of the crystals at the expense of water, and it is likely that highly mineralized bone would need to be replaced. In the ilium, osteocyte viability declines TABLE IV
Reasons for Bone Replacement by Remodeling
Structural bone (cortical, fatty cancellous) Replace fatigue damaged bone — Targeted Prevent bone senescence — “Spare”a Metabolic bone (hematopoietic cancellous) Replace over-mineralized bone — Targeted Preserve osteocyte viability — ? Targeted ? Release of growth factors — “Spare” aSpare remodeling has also been referred to as redundant, surplus, stochastic, or nontargeted [30].
501
with age in interstitial bone but not in surface bone, most likely because its aging is prevented by remodeling [32]. Finally, the controlled release of growth factors stored in the bone may be necessary for optimum hematopoiesis. There are several other functions of remodeling. Bone resorption increases temporarily at night to compensate for the lack of intestinal calcium absorption [33], most likely because existing osteoclasts work a bit harder [31], and there must be some minimum number of osteoclasts for this mechanism to work. At times of temporary increase in calcium demand during growth, pregnancy, and lactation, there is a temporary increase in cortical porosity due to increased remodeling [34]. As mentioned earlier, during the adolescent growth spurt, remodeling accomplishes an increase in the thickness of trabeculae [3], an exception to the general rule that remodeling leads only to conservation or loss of bone. Remodeling discards bone made redundant by disuse [35], and malfunction of this mechanism may be one way in which estrogen deficiency has harmful effects [36]. Finally, remodeling participates in primary fracture healing in cortical bone, creating dowel pins between the two fragments [37].
IV. THE BASIC MULTICELLULAR UNIT AS THE INSTRUMENT OF BONE REMODELING All remodeling is carried out by temporary anatomic structures, first identified by Frost [38] and named by him basic multicellular units (or BMUs), to which, in the adult noninjured skeleton, all osteoclasts and osteoblasts belong. The BMU excavates and refills a tunnel through the bone [39], leaving a new osteon in its wake (Fig. 3). At the front is a team of multinucleated osteoclasts forming the cutting cone; behind them is a capillary that brings circulating monocytes to become preosteoclasts [40], surrounded by loose connective tissue containing nerve fibers and behind that successive teams of osteoblasts forming the closing cone. During its lifespan of many months, the BMU while traveling through the bone maintains the same spatial and temporal relationships between its constituent cells—osteoclasts, vascular endothelial cells, stromal connective tissue cells, and osteoblasts. Trying to understand bone remodeling while ignoring the structure of the BMU is like trying to understand endochondral ossification while ignoring the spatial and temporal relationships between the cells in the growth plate. Histologic sections most often intercept the BMU not longitudinally, as in Fig. 3, but transversely (Fig. 4). Each osteoblast team is assembled on the cement surface
502
A. M. PARFITT
FIGURE 3 Cortical BMU in longitudinal section. A typical BMU in human iliac cortical bone is shown (left panel) with a team of osteoclasts at the bottom, osteoid seams (stained purple) lined with osteoblasts above, and a capillary in between. The entire structure, which is approximately 1 to 2 mm long and 0.2 to 0.4 mm wide, is traveling downward at a rate that can be measured by the longitudinal distance between the tetracycline labels (middle panel) and which in humans is about 25 µm per day [39]. Schematic of coordinated movement in right panel. Middle and right panels reproduced by courtesy of Dr. Robert Schenk. (See color plate).
within a narrow window of time and in a narrow zone between the cutting and closing cones. Each team is like a necklace one cell deep, and about two to three new teams are formed each day. Each team refills the tunnel at a single cross-sectional location as completely as possible. New osteoblasts can join the team only while it is being assembled, not after it has begun to deposit new bone matrix [41]. As one goes further back the teams were born earlier and so are closer to completing their task (Fig. 3, right panel). As it moves through tissue space the BMU creates and leaves behind successive cross-sectional cycles of remodeling, each one slightly out of step with the one before, and at each cross-sectional location resorption moves centrifugally until the cavity is the right size, and formation moves centripetally until refilling of the cavity has been completed at that location. Until recently it was widely believed that the threedimensional organization of the BMU just described applied only to cortical bone, but it is now clear that a
similar organization applies to cancellous bone [39]. A cancellous BMU shows the same relationship between the osteoclastic cutting hemicone in front and the osteoblastic closing hemicone following behind (Fig. 5). Such structures are rarely observed since this requires precise orientation of the randomly generated section in three orthogonal planes, but considerable indirect evidence, reviewed in detail elsewhere [39], indicates that this is the normal, characteristic, manner of cancellous bone remodeling. The process is referred to as hemiosteonal because the BMU excavates a trench across the surface of the bone, rather than a tunnel through the bone, leaving in its wake what may be termed, with some geometrical latitude, a hemiosteon rather than an osteon. Osteonal and hemiosteonal remodeling both display the same relationships between cells that are maintained during BMU progression, the same localization of osteoblast assembly to the narrow zone between the cutting and closing cones or hemicones, and the same dependence on the circulation [42,43],
CHAPTER 28 Modeling and Remodeling: How Bone Cells Work Together
503
FIGURE 4
Cortical BMU in cross section. The bright field on the left shows a large cavity lined by purple staining osteoid within largely unstained mineralized bone. The fluorescence on the right shows that bone formation began eccentrically, as there are two tetracycline labels in the top left corner and only one around the rest of the circumference. The osteoblasts are plump and the osteoid seam (stained purple) is wide. In the lower left is another BMU in which bone formation at this location is almost completed. The cavity is much smaller, the osteoid seam much thinner, the cells much flatter, and the labels much closer together. The cutting cone of the BMU passed through the plane of section about 10 days ago in the right upper location and about 2–3 months ago in the left lower location. From the distribution of partly stained bone of low density that has been recently formed, the newer BMU has arisen adjacent to a not much older BMU in which secondary mineralization has not yet been completed, indicating a state of high bone turnover. Copyright A.M. Parfitt 2001: used with permission. (See color plate).
providing a source for osteoclast precursors and a conduit for ion exchange and intercellular signaling. As the BMU exists and moves in three-dimensional space, it has a beginning (called origination), a middle (called progression), and an end (called termination). Each BMU, representing a new remodeling project, begins on a quiescent surface in response to a target, which is a region of bone in need of replacement [26,30]. The site of origination must be reasonably close to a blood vessel, and in most cases it will be some distance from the target, to which the BMU must progress by excavating a trench in cancellous or a tunnel in cortical bone. Once the target has been reached, it will take some time for the machinery that has been cranked up to be cranked down again, so that the BMU will often progress for some distance beyond the target (Fig. 6). Such posttarget progression provides a convenient basis for spare remodeling, and the distance
traveled will depend on the availability of circulating osteoclast precursors, which are produced in the bone marrow under hormonal control, forming a pool whose size depends on the concentration in the blood [40]. During origination only osteoclasts are needed and during termination only osteoblasts are needed, but during progression, which is much the longest phase, both osteoclasts and osteoblasts are needed at the same time, but at different locations. How a new BMU gets started is still only dimly understood. The first step must be to recognize that some bone needs replacement. In cortical bone, most, perhaps all, remodeling is initiated by fatigue microdamage [44]. The necessary targeting is probably mediated by local death of osteocytes by apoptosis [45,46], although the signals involved are unknown. How, or even whether, targeting is achieved in metabolic bone is unknown [30]. Frost originally
504
A. M. PARFITT
FIGURE 5
Osteonal and hemiosteonal remodeling. (Upper panel) A representative cortical BMU in the dog, with the upper half shaded. (Lower panel) A cancellous BMU in a patient with secondary hyperparathyroidism, in which the spatial organization resembles the bottom half of the upper panel. Although it is possible that the appearance is specific to the particular disease state in which it was observed, reasons are given in [39] and in the text for believing that the structure is characteristic of cancellous bone remodeling in general, but is more likely to be observed when bone turnover is very high because the necessary circumstances will otherwise rarely occur. Called to author’s attention by Dr. David Baylink. Original provided by Dr. Robert Schenk and reproduced with permission. (See color plate).
defined activation as a stimulus to precursor cells to proliferate [1]. It is now often stated that the first step in remodeling is the activation of osteoclasts, but in the normal adult skeleton there are no osteoclasts waiting to be activated. Bone lining cells must be involved in site selection and it is these cells that are activated, to remove the thin endosteal membrane that normally separates them from the mineralized bone, and to retract, providing a foothold for the future BMU [39]. A new vessel then grows toward the site to allow access to circulating mononuclear precursors [40], which are presumably attracted by chemotaxis and fuse to form the first new osteoclasts [47], which begin resorbing in the right direction toward the target. In order for a BMU to progress through the bone or across the surface of the bone several conditions must be satisfied. First, access for circulating preosteoclasts must be maintained. In cortical bone this is achieved by neoangiogenesis, the central capillary keeping up with the cutting cone [42], but in cancellous bone it is
more likely achieved by vasculogenic mimicry, by lining cells that persist as a canopy over the remodeling site, forming a compartment that is in communication with a marrow sinusoid [43]. Second, signals for the arrival of new preosteoclasts must persist, analogous to the continued recruitment of leukocytes to sites of inflammation [46]. Third, the precursor cells that leave the circulation to form new osteoclasts must be replaced from the bone marrow. If all of these conditions are met, new cycles of remodeling, oriented transversely to the direction of BMU advance, will continue to be generated. Failure of one or more of these mechanisms presumably leads to BMU termination. The number of such cycles determines the total distance traveled by the BMU both toward and beyond its target [26,47], and the product of origination frequency and the average number of cycles per BMU corresponds to the histomorphometric index called activation frequency. This is the reciprocal of the average time between the onset of successive cycles of remodeling at the same location,
505
CHAPTER 28 Modeling and Remodeling: How Bone Cells Work Together
WEEKS QUIESCENCE
0
PRE-Oc 1
ORIGINATION
Ob
Oc 16
PROGRESSION
Ob
Oc
TERMINATION
32
QUIESCENCE
48
very old bone
old bone
new bone
FIGURE 6
Cancellous bone remodeling. Evolution of a single BMU through successive stages. BMU is shown as traveling from left to right, excavating a trench across the surface. It originates where a blood vessel arborizes over the bone surface (location denoted by vertical red line) and is directed to remove a region of bone that is too old to carry out its function, whether this function is primarily mechanical, as in the metaphyses of the long bones, or primarily metabolic, as in the axial skeleton [26]. The BMU progresses beyond this target as part of the stochastic or spare component of remodeling. Progression is sustained by the continued recruitment of new preosteoclasts (PRE-Oc), which need to be precisely targeted to the apex of the cutting hemicone, adjacent to the lining cells that cover the bone that is about to be resorbed. Termination occurs when the supply of PRE-Oc is turned off, but osteoblast recruitment continues until the trench is refilled. For further details see text. Copyright A. M. Parfitt (1995); used with permission. See color plate.
which is about 3 years in central cancellous bone and about 10 to 20 years at sites where bone turnover is lower. Activation frequency, which is a two-dimensional concept, is the best histologic index of the overall intensity of bone remodeling, but it is not a measure of the birth rate of new BMU, which is a three-dimensional concept. For example, one BMU that progresses for nine units of distance will have exactly the same effect on all histologic, radiokinetic, and biochemical indices of bone turnover as three BMUs that each progress for only three units of distance [26,47], but the biological significance will be quite different. The former represents a low demand for remodeling with a large surplus component, the latter a high demand for remodeling that is efficiently satisfied. Endocrine abnormalities such as estrogen deficiency, or parathyroid or thyroid hormone
excess, that increase the availability of osteoclasts can increase the extent of post target progression without increasing BMU origination, if the demand for remodeling has not changed [48]. Hormones are indifferent to bone structure and have no means of specifying where a new BMU will be formed. Bone remodeling resembles hematopoiesis in several respects. Both are examples of bone marrow cell renewal, in which changes in the number of functioning cells are more important than changes in individual cell activity. In both, the executive cells have a short life span and need to be continually replaced. In both, a basal rate of cell production can be increased on demand; in both, cell number depends not only on cell production but on the timing of cell death, which determines cell life span; and in both, dysregulation of cell number will lead to disease. The life span of the BMU
506 comprising separate stages of origination, progression, and termination, is measured in months, but the life span of osteoblasts while they are making bone is measured in weeks, and the life span of osteoclast nuclei is measured in days. All cells that originate in the marrow die after a rather short time, and all osteoclasts die by apoptosis [47], but osteoblasts, derived from local stromal cells, have a more complex fate [41]. About two-thirds die by apoptosis, so that the density of cells on the surface progressively declines, a process augmented by the incorporation of about 30% of the initial number of osteoblasts into bone as osteocytes. Because each osteoblast has to cover a progressively larger area, they become flatter and thinner and at the termination of bone formation at a particular location about 5% of the original number persist as the lining cells that cover all quiescent bone surfaces. Osteocytes may survive for decades, but eventually will either be removed as the bone is remodeled or themselves die by apoptosis [32].
V. DISORDERED REMODELING AND AGE-RELATED BONE LOSS Age-related bone loss is a near-universal characteristic of the human species, from which no group is known to be exempt. But whatever the proximate causes such as physical inactivity, altered nutrition, or sex hormone deficiency, and whatever the genetic susceptibilities or molecular mechanisms, all bone loss must be accountable for in terms of disordered remodeling. But first the distinction between its reversible and irreversible components must be clarified. Reversible bone loss is an inevitable consequence of an increase in bone remodeling, most clearly exemplified by the changes in rib cortical porosity during the antler growth cycle in deer, referred to as cyclic physiologic osteoporosis [49,50]. Some of the calcium needed for the antlers is borrowed from the skeleton by starting up a large number of new BMUs, and when they have run their course the debt is fully repaid. The aggregate volume of all the new holes is the remodeling space, but the reversible deficit also includes unmineralized bone matrix or osteoid, which has a short life span, and incompletely mineralized matrix in young recently formed bone, which has a long life span [34]. Because completion of mineralization takes a long time, mineral density increases steadily with the age of the bone. Increased remodeling reduces mean bone age and mean mineral density (Fig. 4) with converse changes when normal remodeling is restored. All three components of the reversible deficit contribute to temporary calcium needs during growth, pregnancy, and lactation [50].
A. M. PARFITT
It is useful to consider the mechanisms of irreversible bone loss in terms of the sequential changes at a single cross-sectional location that are initiated by a BMU as it travels perpendicular to the plane of section, even though this is an inaccurate description of the bone remodeling process in three dimensions (Fig. 7). The cycle begins with a quiescent surface covered by flat lining cells, continues with activation of the lining cells and recruitment of preosteoclasts, resorption to a certain depth, the reversal phase after the termination of resorption, osteoblast recruitment and bone formation, and in the ideal situation in a young healthy adult, complete refilling of the cavity and restoration of the bone surface to its previous location [51]. Because of the perpendicular movement of the bone surface, I call this the down and up model. In terms of this model, all irreversible bone loss is the result of focal imbalance between the depth of a resorption cavity and the depth of the new bone deposited within the cavity, which corresponds to the histologic measurement of wall thickness (Fig. 8). This focal imbalance can result either because the cavity is too deep or because the new bone is too shallow. The bone loss is irreversible because each cycle of remodeling, representing events within a two-dimensional slice of bone generated by the movement of the BMU in three dimensions, constitutes a transaction that once completed cannot be revoked, even though future transactions may be positive in response to treatment. Because of the three-dimensional organization of the BMU, bone resorption can be partitioned into horizontal and vertical components (Fig. 9). The horizontal component is the distance traveled by the BMU across the surface, which depends on the supply of osteoclast precursor cells, both production in the marrow and delivery by the circulation. The vertical component is the depth of the cavity, of which the main determinant is the timing of osteoclast apoptosis. Earlier apoptosis means a shallower cavity, later apoptosis a deeper cavity, with consequent perforation of trabecular plates and cancellization of the inner third of the cortex [52]. The timing of apoptosis is independent of birth rate, since delayed apoptosis can occur in the early stages of estrogen deficiency, in which osteoclast production is increased [53], or in glucocorticoid excess, in which it is decreased [54]. Unlike bone resorption, bone formation has only a vertical component, but like bone resorption, it is influenced by cell death as well as by cell birth. At each cross-sectional location, new osteoblasts assemble on the cement surface at the floor of the resorption cavity. They are columnar in shape and closely packed, and become progressively fewer, flatter, and less active until those remaining become lining cells (Fig. 4). In the elderly,
507
CHAPTER 28 Modeling and Remodeling: How Bone Cells Work Together
2. Activation Pre Oc.
3. Resorption 1. Quiescence
Oc.
Li.Ce. E.Lc. Old bone New bone 4. Reversal
Li.Ce. B.St.U.
Osteoid
?
Cm.Ln. Cm.Ln. 6. Quiescence
5. Formation Ob.
Cm.Ln.
FIGURE 7 Diagrammatic representation of the remodeling cycle in cancellous bone. Successive stages of quiescence, activation, resorption, reversal, formation, and back to quiescence at a single cross-sectional location generated by the movement of the BMU perpendicular to the plane of section are depicted, the so called down and up model. Li.Ce., Lining cell; Pre.Oc., preosteoclast; Oc., osteoclast; E.Lc., eroded lacuna; Cm.Ln., cement line; Ob., osteoblast; B.St.U., bone structural unit (hemiosteon). Refilling is assumed to be complete, and bone marrow lying above the lining cells is omitted for clarity. Reproduced from Parfitt [51] with permission of the publisher.
each team of osteoblasts on the cancellous surface makes less bone, partly because there are fewer initial members of the team and partly because some die sooner, which leads to trabecular and cortical thinning [52]. This can be the result of physical inactivity, estrogen deficiency, glucocorticoid excess, or aging per se. All bone loss occurs from an internal surface, one of the three subdivisions of the endosteal envelope, in contrast to the periosteal envelope, which may slowly gain bone, so that the decline in total bone mass is always smaller than the amount lost [55]. The absolute rate of bone loss from a surface, expressed as volume of bone per unit surface per unit time, depends on the average amount lost in each cycle of remodeling and how often the cycles occur (Fig. 10), which is why there is a general relationship between the rate of bone loss and the rate of bone turnover [56]. The rate of loss expressed as a percentage of the initial amount also depends on the surface-to-volume ratio, which is why the relative rate is slower in cortical than in cancellous bone, even though the absolute amount lost is
greater [57]. The absolute amount lost from a bone depends also on the absolute extent of surface, which declines with age in cancellous bone because whole structural elements are removed, and increases with age in cortical bone because subendocortical cavities become larger [52,58], which is why the fractional rate of loss tends to fall with age in cancellous bone but may increase with age in cortical bone [59]. A remarkable feature of age-related bone loss is its universality, affecting not only almost every person but almost every bone. Although different skeletal sites may lose bone at different rates in the short term, the wider the age range over which the data are collected, the more similar the rates become. For both central and peripheral sites, comprising various proportions of cortical and cancellous bone, the long term rates of bone loss measured cross sectionally are in the range of 1–1.5%/y [26]. Most bone lost with age is cortical and cortical, thinning is mainly the result of increased resorption depth [60,61], which is the two-dimensional reflection of deeper penetration by endocortical BMUs.
508
A. M. PARFITT
Normal – Focal Bone balance
Osteoclast mediated Bone loss
Osteoblast mediated Bone loss
Old bone
New bone
FIGURE 8 Possible mechanisms of focal remodeling imbalance. The upper panel shows a normal-depth resorption cavity on the left, completely refilled with new bone on the right. The middle panel shows a resorption cavity of excessive depth that is incompletely refilled by a normal amount of new bone. The bottom panel shows a resorption cavity of normal depth that is incompletely refilled by a subnormal amount of new bone. Note that the amount of bone lost as a result of the cycle, indicated by the clear area between the original surface location denoted by the interrupted line, and the new surface location, can be the same even though the cellular mechanism is quite different. Modified from Parfitt [52] with permission of the publisher.
Because the rates of fractional loss are so similar, the increase in resorption depth at different sites must be inversely related to the customary rate of turnover and positively related to the usual thickness of cortical bone at each site [26]. When bone loss is both generalized and sustained, as in normal aging, it appears that resorption depth at different sites increases to the extent necessary to bring about much the same rates of fractional bone loss, and to adjust for differences in bone turnover contingent on differences in marrow composition and for differences in local bone structure and geometry [26]. The only conceivable kind of explanation for such a phenomenon is biomechanical. All mechanical influences on bone remodeling are mediated by strain, the technical term for relative deformation of a structural material as the result of load bearing. Similar fractional rates of bone loss throughout the skeleton will produce similar proportional changes in the strains that occur in different bones as the result of the same pattern and intensity of
physical activity. The recruitment and activity of osteoclasts and osteoblasts are orchestrated by the mechanostat in such a way that strain is maintained within an acceptable range [36]. As a result of the sedentary lifestyle made possible by economic development, aging is in most persons accompanied by a progressive reduction in physical activity and muscle strength, of earlier onset and greater severity than is biologically mandated [62]. The risk of fracture should not be increased, as the reduced bone mass would remain appropriate to the reduced level of activity, but this does not take into account the age-related increase in the liability to falls, to which the mechanostat is blind. If this were the sole explanation for age-related bone loss, its magnitude should have been less in more physically active populations, but perhaps the mechanostat is reset, so that the bones respond not to actual but to erroneously perceived disuse [36], either because of estrogen deficiency, or as a consequence of the aging process itself [26].
509
CHAPTER 28 Modeling and Remodeling: How Bone Cells Work Together
(Direction of BMU progression)
Precursor cell supply (production and delivery)
Extent
Oc
Ob
Timing of apoptosis
Depth
FIGURE 9
Two aspects of bone resorption. A. BMU is traveling across the cancellous surface from left to right. At the apex the youngest osteoclast abuts the lining cell covering the surface that is about to be resorbed. The extent of resorption in the direction of progression depends on maintaining the supply of mononuclear osteoclast precursors that are produced in the bone marrow and delivered to the resorption site by the circulation. The depth of resorption, which determines the magnitude of the task set for the osteoblasts, depends on the timing of apoptosis. Copyright A. M. Parfitt, 2001, used with permission. See CD-ROM for color.
VI. POSSIBLE TARGETS FOR VITAMIN D ACTION
Acf.N
Acf X2
Acf X0.5
0
3 Time (years)
6
FIGURE 10 Remodeling determinants of bone loss. In normal adult human cancellous bone one cycle occurs about once every 3 years (upper panel). If the activation frequency is doubled, there will be four cycles in 6 years instead of two (middle panel), and if it is halved, there will be only one cycle (lower panel), with corresponding differences in the cumulative bone loss, indicated by the distances between the interrupted and solid lines after 6 years, even if the amount lost in each cycle remains constant. Differences in the magnitude of focal imbalance will lead to proportional differences in the rate of bone loss if there is no change in activation frequency. Copyright A. M. Parfitt, 2001, used with permission.
Vitamin D is essential for the mineralization of growth plate cartilage and osteoid as described in detail in Chapter 63. Apart from defective mineralization, the most consistent effect of vitamin D deficiency on bone remodeling is increased turnover due to secondary hyperparathyroidism, of which the pathogenesis is discussed in Chapter 30. Parathyroid hormone (PTH) increases osteoclast formation and activation frequency by mechanisms that are beyond the scope of this chapter [53], and this occurs despite the important role of calcitriol in osteoclastogenesis (Chapter 38). Serum calcitriol levels are normal or even increased in most patients with vitamin D deficiency [63] because the stimulatory effects of PTH, hypocalcemia, and hypophosphatemia on the renal 1α-hydroxylase more than offset the effect of substrate deficiency. Although bone formation rate is increased on all surfaces, the most clinically important effect is accelerated loss of cortical bone in both the peripheral and central skeleton, as a result of increased depth of resorption cavities on the
510 endocortical surface [64]. Further loss can be prevented by returning PTH secretion to normal by an appropriate combination of vitamin D and calcium. This contributes to the reduction in hip fracture risk by vitamin D [65], although improved muscle strength may be even more important (Chapter 102) [66]. Agents that increase osteoclast production in vitro, such as PTH, generally increase activation frequency and bone turnover in vivo, but such an effect has not been demonstrated for calcitriol. This could be because PTH secretion is inhibited (both directly and indirectly by an increase in plasma calcium), but calcitriol did not increase biochemical indices of bone turnover when given to patients with hypoparathyroidism [67]. PTH had its expected effects on bone turnover, and both treatments were effective in restoring normocalcemia, in part by increasing the level at which blood and bone equilibrate at quiescent bone surfaces [31]. A similar dissociation between bone turnover and calcemia was found with vitamin D treatment of hypoparathyroidism [68]. Why calcitriol and PTH, which have similar effects on osteoclast production in vitro, should have such dissimilar effects on bone turnover in vivo is an important topic for future research. Its importance resides not so much in the specific issue, but in the more general question of the relevance of in vitro studies to whole organism pathophysiology [69]. When vitamin D deficiency and secondary hyperparathyroidism occur in the context of intestinal malabsorption there is cancellous as well as cortical osteopenia [70]. This could be due to the added effect of calcium deficiency, although there may also be deficiency of some other nutrient not yet recognized as essential for bone health. A third possibility is that vitamin D deficiency might contribute more directly to the pathogenesis of osteoporosis. Since circulating calcidiol is a substrate for the extra renal synthesis of calcitriol in bone [71] and calcitriol has effects on osteoblasts such as stimulating the release of osteocalcin [72,73], it is conceivable that vitamin D deficiency could compromise osteoblast function even in the absence of secondary hyperparathyroidism, leading to incomplete refilling of resorption cavities and thinning of trabeculae to a greater extent than expected from aging alone. Although such a mechanism is speculative, it is in agreement with evidence that the level of vitamin D nutrition that is optimal for bone health is higher than is currently believed (Chapters 61,62) [74]. The hypercalcemia of vitamin D intoxication is often ascribed to increased intestinal calcium absorption together with reduced glomerular filtration rate, but effects of vitamin D on bone are also implicated. As mentioned earlier, calcitriol potentiates the effects
A. M. PARFITT
of PTH on setting the level of blood–bone equilibration [31], but whether this contributes to hypercalcemia is unclear. Increased bone resorption has been inferred in patients with vitamin D intoxication from several lines of evidence, including increased urinary hydroxyproline excretion [75], increased fasting urinary calcium excretion [76,77], negative calcium balance [78,79], and the response to bisphosphonate administration [76,77,80], but with minimal histologic verification. The single bone biopsy showed a sevenfold increase in osteoclast number but no increase in turnover; rather, there was defective mineralization and osteoid accumulation [75]. Serum levels of 25OHD are as high as 20 times normal, and levels of 1,25(OH)2D are about two to three times the normal mean [75,77]. In one case there was a close relationship between changes in plasma calcium and 1,25(OH)2D levels in response to glucocorticoid administration [75], but increased bone resorption has also occurred in an anephric patient [81]. 25-Hydroxy D levels may be high enough to activate calcitriol receptors in bone or to serve as substrate for the synthesis of calcitriol in bone; the metabolite responsible may differ between patients. Whether the increase in osteoclast number in a single case was due to increased recruitment or increased life span is unclear. The absence of increased turnover in this and other cases is more in keeping with increased activity of existing osteoclasts, which is a plausible in vivo expression of the in vitro effects of calcitriol on bone resorption (Chapter 32).
VII. CONCLUSIONS The remodeling system is highly complex [1,53] but its essential elements are few in number. First, for the primary purpose of damage repair, new BMUs must be initiated promptly and must travel rapidly in the right direction. Second, remodeling influences bone mass by the interplay of activation frequency, resorption depth, and wall thickness in different regions of the skeleton. Each of these elements depends on the recruitment, life span, and activity of teams of osteoclasts or osteoblasts, coordinated in part by signals from lining cells and osteocytes. It is regrettable that 80 years after the discovery of vitamin D and 30 years after its metabolism was elucidated, nothing definite is known about the direct effects of vitamin D or of any of its metabolites on any of the elements just enumerated—the necessary experiments have just not been conducted. A similar dearth of knowledge applies to most pharmaceutical agents [82]. This defect will not be remedied until there is more general recognition
CHAPTER 28 Modeling and Remodeling: How Bone Cells Work Together
that the role of disordered remodeling in the pathophysiology of metabolic bone disease depends on how the system actually works, and cannot be captured by the mantra-like repetition of “bone resorption” and “bone formation.”
20. 21.
References 1. Frost HM 1973 Bone remodeling and its relationship to metabolic bone diseases. Charles C. Thomas, Springfield, IL. 2. Parfitt AM 1997 Genetic effects on bone mass and turnoverrelevance to black/white differences. J Am Coll Nutr 16:325–333. 3. Parfitt AM, Travers R, Rauch F, Glorieux FH 1992 Structural and cellular changes during bone growth in healthy children. Bone 27:487–494. 4. Balena R, Shih M-S, Parfitt AM 1992 Bone resorption and formation on the periosteal envelope of the ilium: A histomorphometric study in healthy women. J Bone Miner Res 7:1475–1482. 5. Rauch F, Schoenau E. The developing bone: Slave or master of its cells and molecules? Pediatr Res 50:309–314. 6. Gardner E 1971 Osteogenesis in the human embryo and fetus. In: Bourne H (ed) The Biochemistry and Physiology of Bone, Vol. III, Development and Growth. Academic Press, New York, pp. 77–118. 7. Parfitt AM 1994 The two faces of growth: benefits and risks to bone integrity. Osteopor Int 4:382–398. 8. Sissons HA 1971 The growth of bone. In: Bourne GH (ed) The Biochemistry and Physiology of Bone, Vol. III. Development and Growth. Academic Press, New York, pp. 145–180. 9. Brighton CT 1978 Structure and function of the growth plate. Clin Orthop Rel Res 136:22–32. 10. Schenk RK, Hunziker EB 1991 Growth plate: Histophysiology, cell and matrix turnover. In: Glorieux FH (ed) Rickets, Nestle Nutrition Workshop Series Vol. 21. Raven Press, New York, pp. 63–78. 11. Stevens DA, Williams GR 1999 Hormone regulation of chondrocyte differentiation and endochondral bone formation. Mol Cell Endocrinol 151:195–204. 12. Kember NF, Kirkwood JR 1991 Cell kinetics and the study of longitudinal bone growth: a perspective. In: Dixon AD, Sarnat BG, Hoyte DA (eds) Fundamentals of Bone Growth: Methodology and Applications. CRC Press, Boca Raton, FL, pp. 153–162. 13. Boskey AL, Doty SB, Stiner D, Binderman I 1996 Viable cells are a requirement for in vitro cartilage calcification. Calcif Tissue Int 58:177–185. 14. Frost HM 1990 Structural adaptations to mechanical usage. 1. Redefining Wolff’s law: The bone remodeling problem. Anat Rec 226: 403–413. 15. Turner RT 1994 Cancellous bone turnover in growing rats: Time-dependent changes in association between calcein label and osteoblasts. J Bone Miner Res 9:1419–1424. 16. Doskocil EN, Valouch P, Pazderka V 1993 On vertebral body growth. Funct Dev. Morphol 3:149–155. 17. Daughaday WH, Rotwein P 1989 Insulin-like growth factors I and II: peptide, messenger ribonucleic acid and gene structures, serum and tissue concentrations. Endo Rev 10:68–91. 18. Parfitt AM 2002 Misconceptions (1): Epiphyseal fusion causes cessation of growth. Bone 30:337–339. 19. Amling M, Neff L, Tanaka S, Inoue D, Kuida K, Weir E, Philbrick WM, Broadus AE, Baron R 1997 Bcl-2 lies downstream
22. 23. 24. 25.
26.
27. 28. 29.
30. 31.
32. 33.
34. 35.
36. 37. 38. 39. 40.
511
to parathyroid hormone–related peptide in a signaling pathway that regulates chrondrocyte maturation during skeletal development. J Cell Biol 136:205–213. Schaffler MB, Wenzel TE, Fazzalari NL, Fyhrie DP 1993 Quantitative characteristics of cancellous bone morphogenesis. Trans Orthop Res Soc 18:150. Frost HM, Jee WSS 1994 Perspectives: A vital biomechanical model of the endochondral ossification mechanism. Anat Rec 240:435–446. Gilsanz V, Gibbens DT, Roe TF, et al. Vertebral bone density in children: Effect of puberty. Radiology 166:847–850. Garn SM 1970 The earlier gain and the later loss of cortical bone. Charles C. Thomas, Springfield, IL. Matkovic V, Heaney RP 1991 Calcium balance during human growth: Evidence for threshold behavior. Am J Clin Nutr 55:992–996. Block JE, Friedlander AL, Brooks GA, Steiger P, Stubbs HA, Genant HK 1989 Determinants of bone density among athletes engaged in weight-bearing and non-weight-bearing activity. J Appl Physiol 67:1100–1105. Parfitt AM. Skeletal heterogeneity and the purposes of bone remodeling: implications for the understanding of osteoporosis. In: Marcus R, Feldman D, Kelsey J (eds) Osteoporosis, 2nd ed. Academic Press, San Diego, pp. 433–447. Mazess RB 1990 Fracture risk: a role for compact bone. Calc Tiss Int 47:191–193. Parfitt AM 2002 Misconceptions (2): Turnover is always higher in cancellous than in cortical bone. 30:807–809. Burr DB, Forwood MR, Fyhrie DP, Martin RB, Schaffler MB, Turner CH 1997 Bone microdamage and skeletal fragility in osteoporotic and stress fractures. J Bone Miner Res 12: 6–15. Parfitt AM 2002 Targeted and non-targeted bone remodeling: Relationship to BMU origination and progression. Bone 30:5–7. Parfitt AM 1993 Calcium homeostasis. In: Mundy GR, Martin TJ (eds) Handbook of Experimental Pharmacology, Vol. 107. Physiology and Pharmacology of Bone. Springer-Verlag, Heidelberg, pp. 1–65. Qiu S, Palnitkar S, Rao DS, Parfitt AM 2002 Age and distance from the surface but not menopause reduce osteocyte variability in human cancellous bone. Bone 31:313–318. Lakatos P, Blumsohn A, Eastell R, Tarjan G, Shinoda H, Stern PH 1995 Circadian rhythm of in vitro bone-resorbing activity in human serum. J Clin Endocrinol Metab 80: 3185–3190. Parfitt AM 1980 Morphologic basis of bone mineral measurements. Transient and steady state effects of treatment in osteoporosis [Editorial]. Miner Electrolyte Metab 4:273–287. Parfitt AM 1990 Pharmacologic manipulation of bone remodelling and calcium homeostasis. In: Kanis JA (ed) Progress in Basic and Clinical Pharmacology, Vol. 4, Calcium Metabolism. S. Karger, Basel, pp. 1–27. Frost HM 1996 Perspectives: A proposed general model of the “mechanostat.” Anat Rec 244:139–147. Schenk R 1987 Cytodynamics and histodynamics of primary bone repair. In: Lane JM (ed) Fracture Healing. Churchill Livingstone, Edinburgh, pp. 23–32. Frost HM 1969 Tetracycline-based histological analysis of bone remodeling. Calcif Tissue Res 3:211–237. Parfitt AM 1994 Osteonal and hemi-osteonal remodeling: The spatial and temporal framework for signal traffic in adult human bone. J Cell Biochem 55:273–286. Parfitt AM 1998 Mini-Review. Osteoclast precursors as leucocytes: Importance of the area code. Bone 23:491–494.
512 41. Parfitt AM 1990 Bone-forming cells in clinical conditions. In: Hall BK (ed) Bone: A Treatise, Vol. 1, The Osteoblast and Osteocyte. Telford Press, Caldwell, NJ, pp. 351–429. 42. Parfitt AM 2000 Mini-Review. The mechanism of coupling— A role for the vasculature. Bone 26:319–323. 43. Parfitt AM 2001 The bone remodeling compartment: A circulatory function for bone lining cells. J Bone Miner Res 16:1583–1585. 44. Martin RB 2002 Is all cortical bone remodeling initiated by microdamage? Bone 30:8–13. 45. Verborgt O, Gibson GJ, Schaffler MB 2000 Loss of osteocyte integrity in association with microdamage and bone remodeling after fatigue in vivo. J Bone Miner Res 15:60–67. 46. Noble BS, Peet N, Stevens HY, Brabbs A, Mosley JR, Reilly GC, Reeve J, Skerry TM, Lanyon LE 2003 Mechanical loading: Biphasic osteocyte survival and targeting of osteoclasts for bone destruction in rat cortical bone. Am J Physiol Cell Physiol 284:C934–C943. 47. Parfitt AM, Mundy GR, Roodman GD, Hughes DE, Boyce B 1996 A new model for the regulation of bone resorption, with particular reference to the effects of bisphosphonates. J Bone Miner Res 11:150–159. 48. Parfitt AM 1996 Hormonal influences on bone remodeling and bone loss—application to the management of primary hyperparathyroidism. Ann Intern Med 125:413–415. 49. Banks WJ, Epling GP, Kainer RA, Davis RW 1968 Antler growth and osteoporosis. I. Morphological and morphometric changes in the costal compacta during the antler growth cycle. Anat Rec 162:387–398. 50. Parfitt AM 1981 Integration of skeletal and mineral homeostasis. In: DeLuca HF, Frost H, Jee W, Johnston C, Parfitt AM (eds) Osteoporosis: Recent Advances in Pathogenesis and Treatment. University Park Press, Baltimore, pp. 115–126. 51. Parfitt AM 1993 Morphometry of bone resorption: Introduction and overview. Bone 14:435–441. 52. Parfitt AM 1988 Bone remodeling: Relationship to the amount and structure of bone and the pathogenesis and prevention of fractures. In: Riggs BL, Melton LJ (eds) Osteoporosis— Etiology, Diagnosis and Management. Raven Press, New York, pp. 45–94. 53. Manolagas SC 2000 Birth and death of bone cells: basic regulatory mechanisms and implications for the pathogenesis and treatment of osteoporosis. Endocr Rev 21:115–137. 54. Weinstein RS, Jilka RJ, Parfitt AM, Manolagas SC 1998 Inhibition of osteoblastogenesis and promotion of apoptosis of osteoblasts and osteocytes by glucocorticoids. Potential mechanisms of their deleterious effects on bone. J Clin Invest 102:274–282. 55. Duan Y, Turner CH, Kim BT, Seeman E 2001 Sexual dimorphism in vertebral fragility is more the result of gender differences in age-related bone gain than bone loss. J Bone Miner Res 16:2267–2275. 56. Garnero P, Delmas PD 2001 Biochemical markers of bone turnover in osteoporosis. In: Marcus R, Feldman D, Kelsey J (eds) Osteoporosis, 2nd ed. Academic Press, New York, pp. 459–477. 57. Han Z-H, Palnitkar S, Rao DS, Nelson D, Parfitt AM 1996 Effect of ethnicity and age or menopause on the structure and geometry of iliac bone. J Bone Miner Res 11:1967–1975. 58. Martin RB 1972 The effects of geometric feedback in the development of osteoporosis. J Biomechanics 5:47–455. 59. Ensrud KE, Palermo L, Black DM, Cauley J, Jergas M, Orwoll ES, Nevin MC, Fox KM, Cummings SR 1995 Hip and calcaneal bone loss increase with advancing age: longitudinal
A. M. PARFITT
60.
61.
62. 63.
64. 65.
66. 67.
68. 69.
70.
71.
72.
73.
74.
results from the study of osteoporotic fractures. J Bone Miner Res 10:1778–1787. Han Z-H, Palnitkar S, Rao D Sudhaker, Nelson D, Parfitt AM 1997 Effects of ethnicity and age or menopause on the remodeling and turnover of iliac bone: Implications for mechanisms of bone loss. J Bone Miner Res 12:498–508. Parfitt AM 1990 The three organizational levels of bone remodeling: implications for the interpretation of biochemical markers and the mechanisms of bone loss. In: Christiansen C, Overgaard K (eds) Osteoporosis. Proceedings of Third International Symposium. Osteopress ApS, Denmark, pp. 429–434. Frost HM 1997 On our age-related bone loss: Insights from a new paradigm. J Bone Miner Res 12:1539–1546. Parfitt AM 1998 Osteomalacia and related disorders. In: Avioli LV, Krane SM (eds) Metabolic Bone Disease and Clinically Related Disorders, 3rd ed. Academic Press, San Diego, pp. 327–386. Parfitt AM 2003 Renal bone disease—A new conceptual framework for the interpretation of bone histomorphometry. Curr Opin Nephrol Hyperten, 12:387–408. Chapuy MC, Pamphile R, Paris E, Kempf C, Schlichting M, Arnaud S, Garnero P, Meunier PJ 2002 Combined calcium and vitamin D3 supplementation in elderly women: confirmation of reversal of secondary hyperparathyroidism and hip fracture risk: the Decalyos II study. Osteoporosis Int 13:257–264. Janssen HC, Samson MM, Verhaar HJ 2002 Vitamin D deficiency, muscle function and falls in elderly people. Am J Clin Nutr 75:611–615. Winer KK, Y
