VITAMINS AND HORMONES VOLUME 51
Editorial Board
FRANK CHYTIL
MARYF. DALLMAN JENNYP. GLUSKER
ANTHONYR. MEANS BERTW...
6 downloads
449 Views
27MB 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
VITAMINS AND HORMONES VOLUME 51
Editorial Board
FRANK CHYTIL
MARYF. DALLMAN JENNYP. GLUSKER
ANTHONYR. MEANS BERTW. O’MALLEY VERNL. SCHRAMM MICHAEL SPORN ARMENH. TASHJIAN,JR.
VITAMINS AND HORMONES ADVANCES IN RESEARCH AND APPLICATIONS
Editor-in-Chief
GERALDLITWACK Department of Pharmacology Jefferson Cancer Institute Thomas Jefferson University Medical College Philadelphia, Pennsylvania
Volume 51
ACADEMIC PRESS San Diego New York Boston London Sydney Tokyo Toronto
This book is printed on acid-free paper. @ Copyright 0 1995 by ACADEMIC PRESS, INC. A11 Rights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher.
Academic Press, Inc. A Division of Harcourt Brace & Company 525 B Street, Suite 1900, San Diego, California 92101-4495
United Kingdom Edition published by Academic Press Limited 24-28 Oval Road, London NWI 7DX
International Standard Serial Number: 0083-6729 International Standard Book Number: 0- 12-709851-8 PRINTED IN THE UNITED STATES OF AMERICA 95 96 9 7 9 8 99 0 0 Q W 9 8 7 6
5
4
3 2
I
Former Editors KENNETH V. THIMANN ROBERTS. HARRIS Newton, Massachusetts
JOHNA. LORRAINE University of Edinburgh Edinburgh, Scotland
PAULL. MUNSON University of North Carolina Chapel Hill, North Carolina
JOHNGLOVER University of Liverpool Liverpool, England
GERALD D. AURBACH Metabolic Diseases Branch National Institute of Diabetes and Digestive and Kidney Diseases National Institutes of Health Bethesda, Maryland
University of California Santa Cruz, California
IRA G. WOOL University of Chicago Chicago, Illinois
EGONDICZFALUSY Karolinska Sjukhuset Stockholm, Sweden
ROBERTOLSON School of Medicine State University of New York at Stony Brook Stony Brook, New York
DONALDB. MCCORMICK Department of Biochemistry Emory University School of Medicine Atlanta, Georgia
This Page Intentionally Left Blank
Contents PREFACE .................................................................
xi
CAMP-Dependent Regulation of Gene Transcription by cAMP Response Element-Binding Protein and cAMP Response Element Modulator
JOEL F. HABENER, CHRISTOPHER P. MILLER,AND MARIOVALLEJO Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CAMP-Dependent Signal Transduction Pathway CAMP-ResponsiveTranscription Factors CREB, CAMP Response Elements ..................... . . . . . . . . . . . . . . . . . . . . . Mechanisms of Transciptional Transactivation . . . . . . . . . . . . . . . . . . . . . . . The CREB and CREM Genes Are Multiexonic in Structure: Alternative Exon Splicing Generates a Complex Array of Isoproteins That Are Either Transactivators or Transrepressors . . . . . . . . . . . . . . . . . . VII. CAMP-Dependent Autoregulation of the Expression of the CREB and CREM Genes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII. Roles of CREB and CREM in the Physiological Regulation of Gene I. 11. 111. IV. V. VI.
.......................................... on Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IX. X. Oncogenic Forms of CREB, CREM, and ATF-1 . . . . . . . . . . . . . . . . . . . . . . . XI. Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References ..................................................
2 2 6 8 12
21 31 32 40 42 43 46
Multiple Facets of the Modulation of Growth by cAMP
PIERRE P. ROGER, SYLVIA REUSE,CARINEMAENHAUT, AND JACQUES E. DUMONT I. 11. 111. IV.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Negative Control of Cell Cycle Progression by cAMP . . . . . . . . . . . . . . . . . Positive Control of Cell Cycle Progression by cAMP . . . . . . . . . . . . . . . . . . Relationship between Growth and Differentiation Controls by cAMP . .
vii
59 73 83 118
...
Vlll
CONTENTS
V. A Role for Cytoskeleton Changes in Control of Growth by CAMP? . . . . . VI . CAMP and the Growth of Cancer Cells .............................. VII . Conclusions and Perspectives ....................................... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
122 127 140 144
Regulation of G-Protein-Coupled Receptors by Receptor Kinases and Arrestins RACHELSTERNE-MARR AND JEFFREY L . BENOVIC I . Introduction .... ................... I1. GRK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........... I11. Arrestins
........... .............
193 196 213 226 227
Vasopressin and Oxytocin: Molecular Biology and Evolution of the Peptide Hormones and Their Receptors EVITAMOHR.WOLFGANG MEYERHOF. AND DIETMAR RICHTER Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VP and OT Gene Expression and Regulation . . . . . . ............... VP and OT mRNA in Dendrites and Axons . . . . . . . ............... Somatic Recombination between the VP and OT Genes in Hypothalamic Neurons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Evolution of the Vertebrate VP/OT Gene Family ..................... VI . Nonapeptide Receptors ............................... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I. I1. I11. IV.
235 237 242 247 249 251 258
Structure and Functions of Steroid Receptors M. G. PARKER I. I1. I11. IV. V. VI .
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Physiological Responses to Estrogens ................................ Intracellular Localization of Estrogen Receptors ..................... Hormone Binding and Receptor Dimerization ........................ Target Gene Recognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transcriptional Activation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
267 268 272 272 274 275
CONTENTS
VII. Specific Gene Transcription by the Estrogen Receptor . . . . . . . . . . . . . . . . VIII. Mechanism of Antiestrogen Action .................................. IX. Cross-coupling with Other Signaling Pathways ...................... References ........................................................
ix 276 277 280 282
Phosphorylation and Steroid Hormone Action WENLONG BAIAND NANCY L. WEIGEL Introduction ....................................................... Phosphorylation of Steroid Hormone Receptors ...................... Regulation of Steroid Hormone Receptors by Phosphorylation . . . . . . . . . Interaction between Steroid Hormone Action and Signal Transduction Pathways ......................................................... V. Summary: A Model of Steroid Hormone Action ...................... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
I. 11. 111. IV.
289 290 296 301 305 307
Nucleocytoplasmic Shuttling of Steroid Receptors DONALD B. DEFRANCO, ANURADHA P. MADAN, YUTING TANG, UMAR. CHANDRAN, NIANXING XIAO,AND JUNYANC I. Introduction ....................................................... 11. Subcellular Localization of Steroid Receptors ........................ 111. Nucleocytoplasmic Shuttling of Steroid Receptors .................... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
315 316 323 333
'Transcriptional Regulation of the Genes Encoding the Cytochrome P-450 Steroid Hydroxylases
KEITH L. PARKER AND BERNARD P. SCHIMMER I. 11. 111. IV. V. VI.
Introduction . . . . . . ..................................... Overview of Steroid Cell-Selective Expression . . . . . . . . . . . . . . . . . . . . . . . . . . Hormone-Regulated Expression . . . . ..................... Perspectives and Future Directio Summary ............................ .......... References .......................................
339 355 362 363
X
CONTENTS
Stress and the Brain: A Paradoxical Role for Adrenal Steroids CAMERON, BRUCES. MCEWEN,DAVIDALBECK,HEATHER HELENM. CHAO,ELIZABETH GOULD,NICOLASHASTINGS, YASUKAZU KURODA, VICTORIA LUINE,ANAMARIAMAGARINOS, CHRISTINA R. MCKITTRICK, MILESORCHINIK, CONSTATINE PAVLIDES, PAUL VAHER,YOSHIFUMI WATANABE, AND NANCY WEILAND I. Introduction ....................................................... 11. Adrenal Steroids and Hippocampal Neuronal Atrophy 111. IV. V. VI. VII. Stress Effects on Cognitive Perfo VIII. Deregulation of the HPA Axis in IX. Effects of Exogenous Glucocorticoid Treatment on Cognitive Performance in Humans ........................................... ..................................... X. References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
371 374 377 380 387 388 391 393 394 395 396
Retinoids and Mouse Embryonic Development
T. MICHAELUNDERHILL, LORIE. KOTCH,AND ELWOOD LINNEY
INDEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
459
Preface The contents of Volume 51 are divided into two sections. The first covers cyclic AMP, kinases, and polypeptide hormones and the second part covers steroid hormone receptors, related genes, and members of the gene family. The volume starts with a n in-depth summary of cyclic AMP regulation of gene transcription by CREB and CREM from members of the laboratory of J. F. Habener. This is followed by a lengthy discussion of the effects of cyclic AMP on growth by members of the J. E. Dumont laboratory. Next, a discussion of the G-protein-coupled receptors and their regulation by kinases and arrestins by R. Sterne-Marr and J. L. Benovic. The first section is completed by a contribution on the structural biology and evolution of vasopressin and oxytocin from D. Richter’s laboratory. The s e c d section begins with a summary of the structure and function of steroid receptors by M. G. Parker and is followed by a discussion on phosphorylation and steroid hormone action from W. Bai and N. L. Weigel. Next is a work on the nucleocytoplasmic shuttling of steroid receptors from the D. B. DeFranco laboratory. There is a report on the transcriptional regulation of genes encoding cytochrome P450 steroid hydroxylases by K. L. Parker and B. P. Schimmer and this is followed by coverage of adrenal steroid action on stress and the brain from the B. S. McEwen laboratory. This section ends with a summary of retinoids and their role in mouse development by the E. Linney laboratory. This is a rather large volume and should provide a great deal of upto-date information for the researcher and student on several topics of current interest. I thank members of the Editorial Board for suggesting some of these topics and authors. Academic Press continues to be supportive and prompt in the publication of assembled volumes. I trust that this fourth volume completed under my guidance will set the tone for future numbers in this serial. GERALD LITWACK
xi
This Page Intentionally Left Blank
VITAMINS AND HORMONES, VOL. 51
CAMP-Dependent Regulation of Gene Transcription by CAMPResponse Element-Binding Protein and CAMP Response Element Modulator JOEL F. HABENER," CHRISTOPHER P. MILLER,*?' AND MARIO VALLEJOt "Laboratory of Molecular Endocrinology Massachusetts General Hospital Howard Hughes Medical Institute Harvard Medical School Boston, Massachusetts 02114 'Reproductive Endocrinology Unit Massachusetts General Hospital Boston. Massachusetts 02114
I. 11. 111. IV. V.
Introduction CAMP-Dependent Signal Transduction Pathway CAMP-Responsive Transcription Factors CREB, CREM, and ATF-1 CAMPResponse Elements Mechanisms of Transcriptional Transactivation A. Kinases B. Phosphatases C. Other Transactivational Domains of CREB D. Adapter Proteins That Couple CREB Transactivation to the Basal Polymerase I1 Transcriptional Complex VI. The CREB and CREM Genes Are Multiexonic in Structure: Alternative Exon Splicing Generates a Complex Array of Isoproteins That Are Either Transactivators or Transrepressors A. Exons Encode Functionally Distinct Domains B. Repressor Isoforms of CREM Are Generated by Several Different Mechanisms C. Alternati.ve Exon Splicing Appears to Provide a Mechanism by Which to Modulate the Transactivational Activities of CREB and CREM D. Unphosphorylated CREB Can Repress Gene Expression Mediated by Phosphorylated CREB E. Exon-Deleted Repressor Isoform of CREM Down-regulates Expression of the c-ros and c-iun Genes VII. CAMP-Dependent Autoregulation of the Expression of the CREB and CREM Genes VIII. Roles of CREB and CREM in the Physiological Regulation of Gene Transcription A. Testes B. Anterior Pituitary Gland C. Brain: Hypothalamus and Pineal Gland D. Possible Role of CREB in Memory 'Present address: Genetics Institute, Inc., Cambridge, Massachusetts 02140 1
Copyright 1 ) 1995 by Academic Press. Inc All rights of' reproduction in any form reserved
2
JOEL F. HABENER et al.
IX. CREB/CREM Autoregulation Network X. Oncogenic Forms of CREB, CREM, and ATF-1 XI. Future Directions References
I. INTRODUCTION Forty years have elapsed since the initial discovery of CAMP, the important second messenger and mediator of cellular signal transduction (reviewed by Sutherland, 1972). Subsequently followed the discoveries of the cellular receptors that serve as sensors of hormones and other extracellular signaling molecules, the transducers consisting of stimulatory GTP-binding proteins (Gilman, 19891, and the effector protein kinases activated by CAMP(Krebs, 1989) involved in the signal transduction cascade ultimately leading to the regulation of gene expression. It has been only during the past 7 years that the final mediators in the CAMP-dependent signaling cascade have been identified. These are the CAMP-responsive transcription factors, DNA-binding proteins whose functions are to stimulate or repress the transcription of target genes. At least three CAMP-response DNA-binding proteins have been clearly identified so far: CREB (CAMPresponse elementbinding protein), CREM (CAMP-response element modulator), and ATF-1 (activating transcription factor 1). The intent of this chapter is to describe selected aspects of our current understanding about the workings of these CAMP-responsive transcription factors, with the main emphasis on CREB and CREM. Additional perspectives on CREB and CREM are given in other publications (Roesler et al., 1988; Habener, 1990; Montminy etal., 1990; Meyer and Habener, 1992,1993; Hoeffler, 1992; Foulkes and Sassone-Corsi, 1992; DeGroot and Sassone-Corsi, 1993; Vallejo and Habener, 1994; Sassone-Corsi, 1994; Lalli and Sassone-Corsi, 1994; Vallejo et al., 1995). 11. CAMP-DEPENDENT SIGNAL TRANSDUCTION PATHWAY In almost all living organisms cells communicate by sending and receiving chemical signals in the form of neurotransmitters and hormones. These signals induce specific cellular responses, for example, changes in plasma membrane properties (ion channels or receptors), cellular growth and metabolism, or gene expression, depending on the nature of the signal and the specific cell type involved (Herschman,
CREB AND CREM REGULATION
3
1989; Karin, 1992).To elicit their actions, the signal neurotransmitter and hormone molecules must first bind to specific high-affinity cellular receptors that reside in the cytoplasm (e.g., steroid hormone receptors), in the nucleus (e.g., thyroid hormone and retinoid receptors), or on the cell surface (e.g., plasma hormone receptors) of target cells. Small lipophilic molecules such as steroid hormones, thyroid hormones, and retinoids can readily diffuse across plasma membranes. After gaining access to the interior of the cells, these hormones bind to and activate nuclear or cytoplasmic receptors (Fuller, 1991; Gronemeyer, 1992; Simons et al., 1992; O’Malley, 1990; Beato, 1989). Often, the ligand-bound activated receptors are sequence-specific DNAbinding proteins that regulate the transcription of specific sets of target genes. Larger and more complex ligands such as peptide hormones cannot diffuse across plasma membranes. These molecules bind to and activate receptors located on the plasma membranes of target cells (Dohlman et al., 1991).The activated cell surface receptors then transduce signals to the interior of the cell by mechanisms that involve coupling to guanine nucleotide-binding proteins (G proteins) and/or phosphorylation cascades (Gilman, 1987; Dohlman et al., 1991; Simon et al., 1991; Yarden and Ullrich, 1988; Ullrich and Schlessinger, 1990). In many cases these signal transduction pathways lead to the nucleus in order to regulate gene transcription. There are several wellcharacterized pathways by which signals can be transmitted from the cell surface to the nucleus (Fig. 1). These pathways can be conceptualized as having four components, or “messenger systems.” The first messenger is the hormone ligand, a macromolecule that binds to the cell surface receptor coupled to the G proteins. This complex is responsible for sensing the signal (receptor) and transmitting the signal (G protein) across the plasma membrane. The first messenger systems regulate intracellular levels of second messengers such as CAMP,diacylglycerol, and calcium (Ca2+).These regulatory substances are responsible for the transfer of information from the interior of the plasma membrane throughout the cell. The second messengers regulate the activities of the third component, effector molecules such as CAMP-dependentprotein kinase A (PKA),protein kinase C (PKC), and calmodulin-dependent protein kinase (CaMK).These kinases regulate by phosphorylations the activities of the final component in the signal transduction pathways, DNA-binding proteins such as CREB. CREB is one of several closely related transcription factors capable of mediating transcriptional regulation by cAMP (see below). The cAMP signaling pathway is one of the most important intracellular signal transduction pathways in eukaryotic cells (Habener,
4
JOEL F. HABENER et al
CAMP-
PKA
RIPTION
FIG.1. The cellular signal transduction pathways that modulate gene transcription. The steps in the signaling pathways are designated as first through fourth component messengers (see text). The first components are hormones and ligands (H,, H,, and H,) that bind to and activate cell surface receptors (Rl, R,, and R3), leading to elevation of intracellular levels of second messenger mediator molecules and subsequent activation of effector proteins (third messengers). The effector proteins are typically kinases that phosphorylate the fourth messengers, DNA-binding proteins. The DNA-binding phosphoprotein CREB is activated by the cAMP signaling pathway. DAG, diacylglycerol; PKC, protein kinase C; PKA, protein kinase A; CaMK, calmodulin-dependent protein kinase; P, phosphate.
1993) (Fig. 2). Elevation of intracellular cAMP occurs when plasma membrane receptors coupled to stimulatory G proteins (G,) are bound and activated by their specific ligands, leading to the activation of adenylyl cyclase (AC). AC catalyzes the conversion of ATP to CAMP, which then binds and activates PKA. In the absence of CAMP, the inactive PKA holoenzyme is a heterotetrameric protein complex consisting of two regulatory subunits (R) and two catalytic subunits (C). Thus far, two classes of R subunits (RI and RII), each with two isoforms (RIa, RIP, RIIa, and RIIP), and three isoforms of the C subunit (Ca,CP, and Cr) have been described. cAMP binds cooperatively to two sites on the R subunits within the holoenzyme, thereby liberating free active C subunits. In addition to modulating the activity of the C subunits, the RII subunits bind proteins known as A kinase-anchoring proteins (AKAPs). AKAPs are a family of proteins responsible for tethering type I1 PKA holoenzyme to specific cellular structures, including the nuclear matrix (Scott and McCartney, 1994).Recent experimental evidence indicating the presence of PKA holoenzyme within the nucleus is described later (Section V1,D). Nuclear compartmentalization of type I1 PKA may play a n important role in regulating the phosphorylation of nuclear proteins. It appears that the transactivational ac-
CREB AND CREM REGULATION
5
mRNA
t
Protein
FIG.2. CREB-mediated CAMP-dependent gene transcription. Hormones and ligands (H) that bind plasma membrane receptors coupled to stimulatory GTP-binding proteins (G,) induce an increase in intracellular cAMP levels by activation of adenylate cyclase (Ac), which then catalyze the conversion of ATP to CAMP. Increases in cAMP levels cause dissociation of protein kinase A (PKA) holoenzyme to liberate inactive regulatory subunits (R)and active catalytic subunits (C); the latter are translocated to the nucleus to phosphorylate (P+)transcription factors such as CREB. Nuclear translocation of PKA C subunits is regulated, whereas translocation of CREB is constitutive. CREB binds to cAMP response elements (CREs; e.g., 5'-TGACGTCA-3'), leading to increased gene transcription and subsequent protein synthesis. R, receptor and regulatory subunit of PKA.
tivity of CREB may be regulated, in part, at the level of the reversible translocation of the free active catalytic subunit of PKA from the cytoplasm to the nucleus (Hagiwara et al., 1993). CREB itself is constitutively transported into the nucleus, where it awaits phosphorylation. This situation is different from that of several other transcription factors, such as the thyroid hormone T, and glucocorticoid receptors, Nuclear factor-&, and CAAT box/enhancer-binding protein p (C/EBP-p). The transport of these proteins from the cytoplasm to the nucleus is activated by phosphorylation, although the exact mechanisms for this regulation are not fully understood. CREB contains a strong consensus site (RRPSY) for phosphorylation by PKA. This site lies within a portion of CREB known as the phosphorylation region (P box), or kinase-inducible domain (KID). Phosphorylation of the serine residue within this sequence potently activates the transactivation functions of CREB (Gonzalez and Montminy, 1989). Within the nucleus CREB recognizes and binds to DNA sequences typified by the consensus palindromic cAMP response element (CRE), 5'-TGACGTCA-3' (Section IV).
6
JOEL F. HABENER et al.
111. CAMP-RESPONSIVE TRANSCRIPTIONFACTORS CREB, CREM, AND ATF-1 Regulation of gene transcription is accomplished by the interaction of DNA-binding proteins with DNA regulatory sequences and with proteins of the general transcription machinery (Maniatis et al., 1987; Mitchell and Tjian, 1989; Ptashne, 1988). Several DNA-binding proteins are recognized to mediate CAMP-regulated gene transcription. These include CREB (Hoeffler et al., 1988; Gonzalez et al., 19891,CREM (Foulkes et al., 1991b), and ATF-1 (Rehfuss et al., 1991; Liu et al., 1993) (Fig. 3).As discussed later (Section VI), CREB, CREM, and ATF-1 exist as multiple isoforms due to the alternative splicing of exons during processing of primary RNA transcripts to the mature mRNAs. For example, two isoforms of CREB are ubiquitously expressed in most tissues, CREB327 and CREB341, but differ by an alternatively spliced exon (exon D) encoding 14 amino acids (HoeMer et al., 1990) (Fig. 3). CREB and CREM are members of a larger class of transcription factors, known as bZIP proteins because they possess a basic region responsible for DNA binding and a n adjacent leucine zipper required for dimerization. Dimerization is an absolute requirement for the binding
7 rDNA binding 1
1-transactivation cAMP+PKA CREE
ATF-1
Q1
domain lbZlPl
+p P-BOX
76%
Q2
BR
91%
ZIP
86%
FIG. 3. CAMP-responsive bZIP proteins CREB, CREM, and ATF-1. These three transcription factors constitute a distinct subfamily of bZIP proteins characterized by highly conserved basic regions (BR), leucine zippers (ZIP),and phosphorylation regions (P Box, also known as the kinase-inducible domain). The percentages of amino acid (aa) similarities to those of corresponding regions of CREB are indicated. CREM I and CREM I1 are generated by alternative exon splicing, resulting in proteins with different DNA binding domains. Locations of the glutamine-rich regions (Q1 and Q 2 ) and the protein kinase A (PKA) phosphorylation site, Serl*g, are also shown. P, phosphate.
CREB AND CREM REGULATION
7
of bZIP proteins to DNA. The basic region is approximately 30 amino acids in length, is well conserved, and contains a relatively high proportion of the positively charged amino acids lysine and arginine. The leucine zipper lies immediately C-terminal to the basic region and consists of a region of amino acids with leucines occurring at every seventh position. The leucine zipper forms an amphipathic a-helix, with hydrophobic residues (including leucines) along one face of the helix and hydrophilic residues along the opposite face. Dimerization of bZIP proteins occurs by formation of a parallel coiled-coil structure with the hydrophobic surfaces of two leucine zipper a-helices facing each other (Landschultz et al., 1988; Vinson et al., 1989). Dimerization brings the positively charged basic regions into a configuration that facilitates recognition of target DNA sequences through contacts of the basic regions with nucleotides in the major groove of the DNA helix. Dimers of bZIP proteins are thought to assume a Y-shaped structure in which the stem of the Y is formed by the juxtaposed leucine zippers and the arms by the basic regions. This structure, first proposed as the “scissors grip” model by Vinson et al. (1989),appears to be basically correct according to results from X-ray crystallographic analysis of the yeast bZIP protein GCN4 (O’Shea et al., 1991). In addition to CREB-related DNA-binding proteins, the bZIP transcription factor family also includes C/EBP-related proteins, Fos/Junrelated proteins, and several more distantly related factors (Johnson and McKnight, 1989; Meyer and Habener, 1993). The C/EBP-related factors are expressed during terminal cell differentiation, whereas the Fos/Jun-related proteins mediate early transcriptional responses to the activation of PKC and growth factor-activated Ras-dependent signaling pathways. The C/EBP- and Fos/Jun-related bZIP proteins can also bind CREs and closely related motifs, but they do so generally with lower affinities than CREB, CREM, or ATF-1. The Fos/Jun-related proteins bind preferentially to the closely related tumor promoter agent response elements (typified by the sequence 5‘-TGACTCA-3’), whereas C/EBP and related proteins prefer sequence elements more closely related t o the CCAAT motif. The functional significance of Fos/Jun or C/EBP interactions with CREs is not understood. In some circumstances, however, C/EBP activates CRE-mediated transcription independent of cAMP signaling (Park et al., 1993; Vallejo et al., 1995). Among all proteins within the bZIP family, CREB, CREM, and ATF-1 are unique in that they mediate transcriptional responses via the modulation of cAMP signaling pathways. CREB, CREM, and ATF-1 are structurally similar (Fig. 3). The amino acid sequences are highly conserved in their basic regions and leucine
8
JOEL F. HABENER et al
zippers, and all three have P boxes and glutamine-rich regions (Q1 and Q2) that function in transcriptional activation. The translational products of the CREM gene are unusual in that they consist of activators or inhibitors of CRE-mediated transcription, depending on whether exons C or G (encoding the glutamine-rich Q1 and Q2 regions; see Section VI,A) are spliced into the mature CREM mRNA (Foulkes et al., 1992). The absence of exons C and G results in CREM isoforms that bind CRE sequences but, since they lack the Q1 and Q2 regions, are unable to activate transcription, and therefore function as transcriptional repressors. The CREM gene products can also undergo alternative splicing of exons encoding two different bZIP domains (I and 11, Fig. 3) (Foulkes et al., 1992).The use of an alternative intronic promoter results in CAMP-inducible transcription of short mRNAs encoding potent repressors of CRE-mediated transcription known as inducible cAMP early repressors (ICERs) (Stehle et al., 1993; Molina et al., 1993). ICER proteins contain DNA binding and dimerization domains but lack glutamine-rich regions and P boxes. Consequently, they can bind CRE sequences and form homo- and heterodimers but cannot activate CRE-mediated transcription. To repress CRE-mediated transcription, ICER protein-s may bind CREs as nonfunctional homodimers or may dimerize with activators to form nonfunctional heterodimers. The CREM gene thus encodes most of the known repressors of CAMPinduced transcription. The role of alternative exon splicing in the interconversion of transactivator to transrepressor isoforms of CREB and CREM is discussed in more detail in Section V1,B. IV. cAMP RESPONSEELEMENTS Many genes whose transcription is regulated by cAMP contain the palindromic sequence 5’-TGACGTCA-3’, or close variations of it, in their promoter regions (Deutsch et al., 1988a,b; Roesler et al., 1988). This sequence is known as the CRE, and its integrity is required for transcriptional responses to CAMP. The perfect palindromic CRE sequence serves as a high-affinity binding site for CREB, CREM, and ATF-1. CREs also serve as binding sites for other bZIP proteins, including those related to C/EBP (Bakker and Parker, 1991; Park et al., 1990,1993; Vallejo et al., 19951, and Fos or J u n (Hai and Curran, 1991; Ryseck and Bravo, 1991). In these instances CREs may function as CAMP-independent transcriptional enhancers. Variant CREs include asymmetrical or atypical sequences that differ from the consensus motif by single or multiple nucleotide deletions
CREB AND CREM REGULATION
9
or substitutions. Examples of these include CRE-like sequences in the rat insulin (Philippe and Missoten, 19901,human enkephalin, and vasoactive intestinal polypeptide gene promoters (Deutsch et al., 1988a,b). Some of these sequences contain intact CRE half-sites (B’-CGTCA-3’) and are still capable of binding CREB and related proteins and mediating transcriptional regulation by CAMP, albeit to a lesser extent than for perfect palindromic CREs. Even among perfect palindromic CRE sequences there are variations in basal activities and relative responsiveness to CAMP-induced transcriptional activation. One reason for this variation is that sequences adjacent to the core palindromic octamer can influence the binding stability and/or transactivation functions of proteins bound to the core CRE (Deutsch et al., 1988a,b).The flanking sequences can serve as binding sites for additional DNA-binding proteins that interact with proteins bound to the core CRE octamer (Muro et al., 1992; Ikuyama et al., 1992; Miller et al., 1993). Additionally, flanking sequences can influence the stability of protein binding to the core octamer (Ryseck and Bravo, 1991). Results from DNase I footprinting and methylation interference experiments show that proteins in nuclear extracts and purified bacterially expressed CREB make base contacts that are well outside the core CRE octamer (Andrisani et al., 1988; Powers et al., 1989; Knepel et al., 1990; Vallejo et al., 1992). Consequently, it is not surprising that sequences flanking the core CRE can influence the binding or function of transcription factors bound to the CRE. All known bZIP transcription factors bind DNA as homodimers or heterodimers. Dimerization occurs in the absence of DNA, and results from experiments with synthetic peptides indicate that isolated leucine zippers of CREB form stable complexes similar to those of intact proteins (Yun et al., 1990). Additionally, from studies with Jun, Fos, GCN4, and C/EBP, it is known that the specificity and stability of the dimerization of bZIP proteins are determined by amino acids at particular positions within the leucine zipper a-helices (O’Shea et al., 1992; Scheurmann et al., 1991; Neuberg et al., 1991). Most bZIP proteins form homodimers capable of binding DNA. An important exception to this generalization is Fos, which does not form stable homodimers (Halazonetis et al., 1988; Smeal et al., 1989). In general, bZIP protein homodimers bind symmetrical CRE-like sequences, whereas heterodimers bind asymmetrical elements (Fig. 4). One criterion for establishing bZIP protein families (CREB, C/EBP, or Fos/Jun related) is the degree of amino acid similarity in the bZIP domains. To some extent, heterodimerization is permitted within these
10
qqp ($-()(-& JOEL F. HABENER et al.
TF-2
----c
--EX--
-TGACGNNN Asymmetric
Symmetric
CREE-P
CIEB
-
FIG. 4. Differential DNA binding of homo- and heterodimeric bZIP proteins to symmetrical and asymmetrical CAMP response elements (CREs). Homodimers of CREB, Jun, and ATF-2 bind to symmetrical CREs, whereas heterodimers of Jun:Fos, Jun:ATF-2, and CIATF:CIEBP preferentially bind asymmetrical or partial CREs. CREB:CREB homodimers also bind CRE half-sites. To bind asymmetric CREs CREB may have to be phosphorylated (P).
families because of the high similarity in the bZIP domains. Certain cross-family heterodimers can also form (Table I). For example, there is selective heterodimerization between members of the ATF and Fos/Jun families. Different heterodimers display DNA binding specificities distinguishable from each other and from those of their parental homodimers (Hai and Curran, 1991). CREB, ATF-2, and J u n bind symmetrical CREs as homodimers, and ATF-2 and ATF-3 form heterodimers with J u n . Ju n , but not Fos, heterodimerizes with ATF-2 (Macgregor et al., 1990). The binding specificity of ATF-2/Jun heterodimers is different from that of Jun/Fos heterodimers, indicating that the DNA binding specificity of J u n is modified by association with Fos or ATF-2. ATF-2iJun heterodimers have a preference for symmetrical CRE sites, but bind asymmetrical CREs and activator protein 1(AP-1) sites, whereas ATF-3/Jun heterodimers bind symmetrical and asymmetrical CREs and AP-1 sites with similar affinities (Hai and Curran, TABLE I TRANSCRIPTION FACTOR CROSS-TALK AMONG DIFFERENT bZIP FAMILIES OF PROTEINS Transcription factor Family 1
Family 2
Heterodimers
Jun CIEBP Jun Jun
Fos ATF C/EBP ATF
c-Jun:c-Fos CIEBP-PCIATF JunD:C/EBP-p c-Jun:ATF-2
11
CREB AND CREM REGULATION
1991). Recently, it has been shown that ATF-3 and JunD form heterodimers and act synergistically to stimulate CAMP-dependent transcription of the proenkephalin gene (Kobierski et al., 1991; Chu et al., 1994). In a similar manner different heterodimer combinations of Fos and J u n proteins bind with differing stabilities to various symmetrical and asymmetrical AP-1 elements and CREs (Ryseck and Bravo, 1991). Fos and J u n can heterodimerize with C/EBP-P to repress transcriptional activation by C/EBP-P (Hsu et al., 1994). In this circumstance, however, heterodimerization with Fos or J u n decreases the affinity of CIEBP-P for binding sites in the interleukin-6 promoter, leading to repression of transcription. One particularly interesting example of cross-family heterodimerization is the interaction between C/EBP-P and C/EBP-related ATF (C/ATF) (Vallejo et al., 1993) (Fig. 5). On the basis of the amino acid sequence of its bZIP domain, C/ATF is a member of the ATF/CREB family, yet it forms stable heterodimers with C/EBP-(3. Indeed, C/ATF was originally isolated in a search for novel CiEBP family members that could dimerize with C/EBP-P (Vallejo et al., 1993). C/ATF homodimers bind to symmetrical CREs and weakly activate transcription, but do not bind or transactivate asymmetrical CREs such as that of the gene encoding phosphoenolpyruvate carboxykinase, or CAAT-related DNA elements (Vallejo et al., 1993). C/ATF:C/EBP-P heterodimers bind to both symmetrical and asymmetrical CREs and activate transcription from both types of elements, but not from C/EBP sites. Therefore, C/ATF may function to redirect the binding of C/EBP-P from CAAT-related sequences to asymmetrical
Concentrations: Predominant: DNA-BPs: Elements
C/EBP>>C/ATF
CIATF=CIEBP
CIATFwwCIEBP
homodimers
heterodimers
homodimers
-1 CBS t
1
CRE-2r
CRE-1
f
TGCGCAAT TGACGCAG TGACGTCA Target genes containing different control elements
FIG. 5. Illustration of how DNA binding site preferences are determined by relative nuclear concentrations of CIEBP and C/ATF. When CiEBP levels exceed those of CIATF, formation of C/EBP hornodimers is favored, which preferentially bind CCAAT box-like DNA sequences (CBS). When CIATF levels are greater than those of CIEBP, C/ATF homodimers are favored, which bind symmetrical CRE sequences (e.g., CRE-1 of the rat proenkephalin gene promoter). At intermediate concentrations CiATF and CiEBP heterodirners are formed that preferentially bind asymmetrical CRE-like sequences (e.g., CRE-2 of the rat proenkephalin gene). BP, binding protein.
12
JOEL F. HABENER et al
CREs. A hypothetical model has been proposed in which the relative levels of C/ATF and C/EBP proteins determine the relative amounts of homo- and heterodimers present, and consequently dictate the promoter binding sites that are occupied and the genes that are transcriptionally activated (Fig. 5). When C/EBP levels exceed those of C/ATF, the formation of C/EBP homodimers is favored, leading to the activation of genes containing canonical C/EBP recognition sites (CCAAT box sites). Conversely, when C/ATF levels are greater than those of C/EBP, the formation of C/ATF homodimers will be favored, resulting in the activation of genes containing symmetrical CREs. Finally, when levels of C/ATF and C/EBP are similar, the formation of heterodimers will be favored, resulting in the activation of genes containing asymmetrical CREs.
V. MECHANISMS OF TRANSCRIPTIONAL TRANSACTIVATION The transactivational activities of CREB, CREM, and ATF-1 are directly dependent on their state and extent of phosphorylation. Thus, the interplay between the phosphorylating protein kinases and the dephosphorylating protein phosphatases is critical in determining the relative transactivation potential at any given moment. A. KINASES The PKA phosphorylation motif shared by CREB, CREMT, and ATF-1, RRPSY, is located within a region of about 60 amino acids known as the KID, or P box (Fig. 6). It has been demonstrated that Ser133 in CREB341, and the corresponding Serllg in the CREB327 isoform, present in the RRPSY motif are phosphoacceptor sites for phosphorylation by PKA (Gonzalez and Montminy, 1989). Ser133 in CREB corresponds to Serll7 in CREMT and Ser68 in ATF-1, respectively. Phosphorylation of Ser133 (Serllg) by PKA is an absolute requirement for the generation of CREB transactivational activity in response to CAMP signaling, as assessed by mutational analyses of CREB expressed in transfected cells as well as by direct microinjection of recombinant CREB into the nuclei of fibroblasts (Gonzalez and Montminy, 1989; Alberts et al., 1994a). Although phosphorylation converts Ser133 from a n uncharged into a negatively charged amino acid, it appears that the gain of a negative charge per se is not sufficient for CREB activation, because Ser133 cannot be substituted for by other
13
CREB AND CREM REGULATION
I
I
/
I
\
I
I
1
\ \
I I I I
1 1 1 1 '
'"GTDGVQGLQTLTMTNAA~ 1-CBP
binding domln-1
kTARl,,pblndlnp domsln+
FIG.6. The structure of the CREB protein. Depicted are the glutamine (&)-rich regions located at either side of the kinase-inducible domain (KID),or phosphorylation box (P BOX); the basic region that contains the DNA binding domain; and the leucine zipper (L) domain (ZIP) that mediates protein dimerization. The KID (amino acids 108-132) contains the CAMP-dependent protein kinase A (PKA) phosphorylation site (A-kinase box), and is also the region that binds to CREB-binding protein (CBP). The amino acid sequence of the hydrophobic cluster (HC) that interacts with the TATA box-associated factor TAFII,,o is also shown.
negatively charged residues (Gonzalez and Montminy, 1989). Therefore, the mechanism by which phosphorylation by PKA activates CREB remains unclear. It has been proposed that phosphorylation induces a n allosteric transition that results in a conformational change of CREB from an inactive into a n active configuration (Brindle et al., 1993; Gonzalez et al., 19911, perhaps facilitating direct interactions with other transcription factors or coactivator proteins, such as the CREB-binding protein (CBP; see Section V,C). Increasing evidence indicates that Serl33 provides a common phosphorylation site at which different signal transduction pathways converge. Thus, CREB mediates some of the transcriptional changes observed after stimulation by membrane depolarization and calcium influx into cells as a result of phosphorylation of Serl33 by Ca2+/ calmodulin-dependent kinases (e.g., CaMK) (Enslen et al., 1994; Sheng et al., 1990, 1991; Dash et al., 1991; Matthews, et al., 1994; Schwaninger et al., 1993; Bading et al., 1993). ATF-1 has also been shown to mediate Caz+-induced transcriptional responses (Liu, 1993), and CREM can be phosphorylated by CaMK (DeGroot et al., 1993b). Therefore, CREB, CREM, and ATF-1 are common targets that integrate signals conveyed by the activation of two different transduction pathways, one CAMPdependent and the other Caz+/calmodulin dependent. In addition, CaZ+-induced gene transcription in neuronal cells is enhanced by nitric oxide through mechanisms that may involve CREB
14
JOEL F. HABENER et al
phosphorylation by PKA (Peunova and Enikolopov, 19931, although the exact details of that interaction are unknown. Serl33 is phosphorylated in uitro by both CaMK-I1 and CaMK-IV (Matthews et al., 1994; Enslen et al., 1994). The cellular distribution of CaMK-IV is both cytoplasmic and nuclear, whereas CaMK-I1 is predominantly cytoplasmic. In addition, CaMK-I1 phosphorylates CREB on both Ser133 and Ser142, and phosphorylation at Ser142 inhibits the transcriptional capacity of CREB (Sun et al., 1994). These circumstances provide a n explanation for the observations that CREB-mediated transcriptional responses appear to be associated with CaMK-IV-dependent kinase activity (Matthews et al., 1994; Enslen et al., 1994). It is important to emphasize that both Ca2+- and CAMP-dependent pathways interact at different levels of the signal transduction cascade. For example, CAMP-induced phosphorylation regulates the activity of specific voltage-sensitive Ca2+ channels (Sculptoreanu et al., 19931, and in turn, Ca2+ affects the activity of certain phosphatases (Cohen, 1989) and AC isoforms (Katsushika et al., 1992; Yoshimura and Cooper, 1992). CREB is also phosphorylated by the activation of signal transduction pathways by growth factors such as transforming growth factor p l (Kramer et al., 1991) and nerve growth factor (NGF) (Ginty et al., 1994). Ginty et al. (1994) have recently reported that in PC12 cells, CREB mediates NGF-induced c-fos transcription via phosphorylation by a previously unidentified kinase. This CREB kinase (CREBK) is a single polypeptide with an apparent molecular mass of 105 kDa. The stimulation of CREBK by NGF (and related growth factors) is dependent on the stimulation of a Ras signaling pathway and does not involve the activation of Ca2+- or CAMP-dependent signaling pathways. The best-characterized Ras signaling pathway to date involves the activation of Ras after stimulation of cell surface receptor tyrosine kinases, which in turn results in the stepwise activation of the protein kinases Raf-1, mitogen-activated protein kinase (MAPK), and MAPK kinase (Blenis, 1993). NGF is only one of several growth factors that activate the Ras-MAPK pathway to phosphorylate several target proteins, including the transcription factors, c-Fos, c-Jun, and c-Myc (Blenis, 1993). However, NGF-induced Ras-dependent phosphorylation of CREB does not appear to involve activation of the MAPK pathway (Ginty et al., 1994). Therefore, the exact mechanisms of CREBK activation following NGF-induced stimulation of Ras remain to be determined. Interestingly, the CAMPand Ras-MAPK pathways interact at more proximal levels of the signal transduction cascade. Thus, it has been
15
CREB AND CREM REGULATION
observed in a number of cell types that activation of PKA by cAMP results in the inhibition of growth factor-induced MAPK activity (Burgering et al., 1993; Cook and McCormick, 1993; Graves et al., 1993; Sevetson et al., 1993; Wu et al., 1993). Studies carried out in uitro suggest that PKA does not directly inhibit MAPK and MAPK kinase activities (Graves et al., 1993). Rather, it appears that phosphorylation of Raf-1 by PKA is responsible for the observed blockade in the RasMAPK pathway (Cook and McCormick, 1993). Based on the observations described above, the following model is proposed (Fig. 7). Stimulation of cell surface receptor tyrosine kinases by NGF and other growth factors would result in the activation of Ras,
growth factor
depolarlzation
I cell membrane
I
I
I
I
SIGNALING
"
ZYS
FIG.7. Interactions among the Ras, Ca2+icalmodulin,and adenylyl cyclase-protein kinase A (PKA) signaling pathways. Signal transduction pathway cross-talk results in transcriptional responses from different sets of genes. Activation of growth factor receptors stimulates the Ras-mitogen-activated protein kinase (MAPK) pathway, resulting in the phosphorylation of transcription factors such as c-Jun, c-Fos, and c-Myc. In addition, a Ras-dependent CREB protein kinase (CREBK) that phosphorylates CREB is activated, although this pathway does not seem to involve MAP kinases. The Ras-MAPK pathway is also activated in response to Ca2+ entry triggered by membrane depolarization. Calcium entry activates Caz+icalmodulin-dependent protein kinases (CaMK), which phosphorylate CREB. In addition, Ca2+ alters the activity of some adenylyl cyclase (AC) isoforms, varying the amounts of cAMP that activate PKA to phosphorylate CREB. Finally, cAMP stimulation results in transcriptional responses mediated by transcription factors such as AP-2 and JunD, although direct phosphorylation of these transcription factors by PKA has not been demonstrated. MAPKK, MAPK kinase.
16
JOEL F. HABENER et al.
which would act as a branching point to distribute the signal through at least two different pathways: one leading to the phosphorylation of transcription factors such as c-Fos, c-Jun, and c-Myc via stimulation of MAPK activity, and the other leading to the phosphorylation of CREB (and perhaps its close homologs CREM and ATF-1) via stimulation of CREBK. In this manner growth factor activities would influence the relative transcription rates of a wide repertoire of genes important for cellular proliferation and/or differentiation. Subsequent activation of the CAMP-dependent pathway by different extracellular signals would attenuate the transcriptional activities of only a subset of growth factor-activated genes, without affecting (or perhaps potentiating) the transcriptional activities of genes regulated by CREB, which provides a point of convergence for both Ras-CREBK and CAMP-PKA signaling pathways. In addition, new genes would be recruited by the CAMPdependent activation of other transcription factors, such as AP-2, ATF-3, and JunD (Chu et al., 1994; Kobierski et al., 1991; Williams et al., 19881, although no direct evidence that PKA phosphorylates these transcription factors exists to date. Interestingly, the Ras-MAPK pathway is also activated by Ca2+ influx through voltage-sensitive channels (Rosen et al., 1994). All of these interactions at different levels among three different signaling pathways indicate the existence of a complex intracellular cross-talk that probably results in a fine regulation of gene expression. These precise mechanisms of response may allow cells to cope with the large constellation of environmental signals to which they are continuously exposed. Thus, by providing a common target for different intracellular signaling pathways, CREB may be directly involved in the regulation of a wide variety of cellular processes, including metabolic adaptations to the extracellular environment, proliferative responses, and differentiation events. In addition to phosphorylation of Ser133 by PKA, CaMK-IV, or CREBK, the KID also contains putative sites for phosphorylation by the processive protein kinases casein kinase I1 (CK-11) and glycogen synthase kinase-3 (GSK-3). Phosphorylation by these kinases is hierarchical, that is, it is facilitated by the prior phosphorylation of adjacent sites (Roach, 1991; Fiol et al., 1994). In the case of CREB, phosphorylation of Ser133 is the required first event. It has been suggested that CK-I1 and GSK-3 may play a functional role in regulating the transcriptional transactivation activity of CREB (Fiol et al., 1994). However, there is evidence suggesting that phosphorylation of CREM by CK-I and CK-I1 alters the binding capacity of CREMT (DeGroot et al., 199313).
CREB AND CREM REGULATION
17
B. PHOSPHATASES Under physiological conditions the degree of phosphorylation of a given protein in cells is the result of the regulated balance between the opposite effects of protein kinases and phosphatases. Accordingly, the transcriptional transactivation activity of CREB is down-regulated by phosphatase-mediated dephosphorylation. Protein phosphatase types 1 (PP-1)and 2A (PP-2A), two specific nuclear phosphatases, have been found to attenuate the CAMP-induced transcriptional activity of CREB (Wadzinski et al., 1993; Hagiwara et al., 1992). PP-1 dephosphorylates Ser133, and the attenuation by this phosphatase of the transcriptional responses induced by CREB is inhibited by a specific PP-1 inhibitor, 1-1(Hagiwara et al., 1992; Alberts et al., 1994b). However, different studies (Wheat et al., 1994; Wadzinski et al., 1993) have provided evidence that PP-2A is the primary enzyme involved in the inactivation of CREB by dephosphorylation. This notion is also supported by studies investigating the effects of phosphorylation on the binding of CREB to different high- and low-affinity CRE sites (Nichols et al., 1992). The notion that PP-2A is the enzyme that dephosphorylates CREB under physiological conditions derives from studies carried out by inhibiting PP-2A with simian virus 40 (SV40) small-tumor antigen, which inhibits the dephosphorylation of CREB and enhances CREB transactivation functions (Wheat et al., 1994). In contrast, 1-1,but not SV40 small-tumor antigen, was found to inhibit the dephosphorylation of and enhance transactivation by CREB (Alberts et al., 1994b). Thus, it remains unclear whether both PP-1 and PP-2A, or only one of them, is functionally important in the regulation of CREB activity induced by CAMP in cells. There is also evidence for the involvement of phosphatases in addition to PP-1 and PP-2A in regulating the phosphorylation of CREB. In uitro experiments show that Ca2+/calmodulin-dependentPP-2B (calcineurin) dephosphorylates CREB (Enslen et al., 1994). Calcineurin phosphatase activity may be required for CREB-mediated glucagon gene transcription, although this activity may occur via a n indirect mechanism (Schwaninger et al., 1993a,b). The overall functional significance of these apparently conflicting actions of protein phosphatases awaits further clarifications.
C. OTHERTRANSACTIVATIONAL DOMAINS OF CREB The transactivational domain of CREB contains two other regions with a relatively high content of glutamine residues. These regions,
18
JOEL F. HABENER et al.
termed Q1 and Q2, are located at either side of the KID (Fig. 6). Initial studies revealed that deletions of either of these two regions result in a marked reduction of CREB transcriptional activity (Brindle et al., 1993; Gonzalez et al., 1991; Quinn, 1993). In the absence of phosphorylation by PKA, the Q1 and Q2 domains are thought to be important for maintaining the basal activity of CREB. This activity may be due to interactions of the Q1 and/or Q2 domains with other transcription factors bound to neighboring sites located in proximity to a CRE. Consistent with this notion are the findings in the pancreatic islet cell lines Tu6 (Leonard et al., 1992) and RIN1027-B2 (Vallejo et al., 1995). CREB activity on the somatostatin gene promoter is dependent not on CAMP stimulation, but rather on interactions with another transcription factor(s) bound to the promoter located in the proximity of the CRE. The existence of glutamine-rich transactivation domains (Mitchell et al., 1989) has been documented in several transcription factors, such as S p l (Courey and Qian, 1988). These domains may provide interaction surfaces for the coupling of transcription factors with specific coactivator proteins associated with the RNA polymerase I1 complex and may result in the activation of transcription of target genes (Dynlacht et al., 1991; Hoey et al., 1993). Therefore, it is likely that the Q1 and Q2 regions in CREB are similarly involved. An interaction has been found between the Q2 domain of CREB and Drosophila TAF,,110 (Ferreri et al., 19941, one of the proteins associated with the basic transcriptional machinery (Hoey et al., 1993) (Fig. 8). It has been proposed that the Q1 and Q2 regions correspond to constitutive activator domains that become exposed for interactions with target coactivator proteins upon phosphorylation of the adjacent KID by PKA (Brindle et al., 1993; Ferreri et al., 1994; Quinn, 1993; Krajewski and Lee, 1994). Experiments carried out by Quinn (19931, in which the transcriptional transactivation activities of a number of deletion CREB mutants were tested, indicate that the transactivation domain of CREB is modular in structure, inasmuch as each of the activaticn domains (basal or inducible) can function independently of each other. Hybrid proteins consisting of the N-terminal transactivation domain of CREB (devoid of the DNA binding domain, amino acids 1248) fused to the DNA binding domain of B-cell activator protein (BSAP-1) that binds its regulatory element as a monomer, activate transcription constitutively independent of phosphorylation by PKA (Krajewski and Lee, 1994).By mutational deletion analyses, the transactivationai activity of the CREBIBSAP-1hybrid protein appears to be mediated by the glutamine-rich regions, not by the P boxlKID domain. Although the CREB/BSAP-1 fusion protein represents an artificial
CREB AND CREM REGULATION
inactive
19
e HC
P-BOX
Active
FIG. 8. The formation of protein complexes that mediate CREB basal transcriptional activity. CREB interacts via the hydrophobic cluster (HC) in the Q2 domain with the TATA box-associated factor TAF110. Note that this interaction can occur in the absence of CREB phosphorylation, hence generating basal transcriptional activity without CAMPstimulation. Contrast this situation (low-level transcriptional activity) with that illustrated in Fig. 9, in which CREB has been phosphorylated and interacts with the basal transcriptional machinery via a different set of basal proteins. TRX, transcription; TBP, TATA-box binding protein; TF, other transcription factors.
unphysiological model, the observations suggest that CREB monomers can couple to the basal transcriptional machinery, presumably by interactions of the glutamine-rich regions containing hydrophobic clusters with TAFlll10 (Fig. 9). Further, these findings suggest that CREB dimers may be required for productive transcriptional interactions with CBP (CREB-binding protein). D. ADAPTER PROTEINS THATCOUPLE CREB TRANSACTIVATION TO THE BASALPOLYMERASE I1 TRANSCRIPTIONAL COMPLEX
As discussed above, CREB may activate gene transcription not directly, but rather through interaction with another effector or coactivator proteins. Recently, a large (265-kDa) protein known as CBP that does not bind DNA has been identified (Chrivia et al., 1993) (Fig. 9A). Analysis of the amino acid sequence of CBP deduced from its cloned cDNA reveals the presence of at least three consensus phosphorylation sites for CaMK-I1 and one for PKA, as well as two putative zinc finger domains. In addition, the C-terminal region contains a glutamine-rich domain. The CREB binding domain, determined by deletional studies (Chrivia et al., 1993),is located within the N-terminal region (Fig. 9A).
20
JOEL F. HABENER et a1
xPKA
A ZF
ZF
y4
H2N-
COOH
1
2441
P-BOX
6 y2 ,~
TRX
Q1
P-BOX
Active (phosphorylate
RX
6 FIG. 9. Structure of the CREB-binding protein (CBP) and how CBP facilitates transcriptional activation by phosphorylated CREB. (A) The region of the molecule that contains the CREB binding domain (CREB BD) and the glutamine (Q)-rich region are shadowed. The location of the two putative zinc finger domains (ZF) and the protein kinase A (PKA) phosphorylation sites are also indicated. (B)The potential transcriptional adapter function of CBP. In the inactive state (top) unphosphorylated CREB is bound to the CRE in the promoter of the target gene, but cannot interact with the proteins that form the basic transcription (TRX) machinery assembled on the TATA box. Phosphorylation of CREB (and possibly also of CBP) by PKA triggers an interaction between CREB and CBP, which, in turn, interacts with TFIIB, forming a higher-hierarchy transcriptionally active complex (bottom). TBP, TATA-box binding protein; TF, other transcription factors; P, phosphate.
CBP interacts with CREB only when Ser133is phosphorylated. According to the proposed model that considers CBP as a coactivator, CBP is recruited to the CREB-DNA complex in the promoter of target genes upon phosphorylation of CREB. In this manner CBP mediates
CREB AND CREM REGULATION
21
CAMP-independent CREB-induced transcriptional responses. Microinjection of anti-CBP antibodies into the nuclei of cells inhibits transcriptional responses elicited by cAMP stimulation, suggesting that CBP is essential for the activation of transcription of CAMP-responsive genes. Interestingly, these studies also indicated that CBP interacts with c-Jun phosphorylated by Jun-kinase (Arias et al., 1994). Coactivator proteins are non-DNA-binding proteins that bridge DNA-bound transcription factors with protein components of the basal transcriptional machinery associated with RNA polymerase I1 (Dynlacht et al., 1991) As such a coactivator, CBP bound to phosphorylated CREB appears to interact directly with TFIIB (Kwok et al., 19941, one of the protein components of the basal RNA polymerase I1 transcription machinery (Buratowski, 1994) (Fig. 9B). The discovery of CBP has revealed interesting implications for the alterations in gene transcription elicited by certain viral proteins mediated via CRE-like elements. CBP contains regions with amino acid sequences homologous to those of equivalent regions in the protein p300, another recently identified nuclear protein that interacts with the adenovirus E1A oncoprotein and may mediate E l A-induced proliferation of cells (Arany et al., 1994). Both proteins share 85-95% similarity over several segments, one of which is also homologous to ADA2 (Arany et al., 1994; Chrivia et al., 19931, a coactivator protein found in yeast (Berger et al., 1992). Based on sequence similarities shared by p300 and CBP, it has been proposed (Arany et al., 1994) that interactions between CREB and p300, as well as between viral E1A (or its putative cellular counterpart) and CBP may occur in cells. In addition, CBP interacts with phosphorylated c-Jun (Arias et al., 19941, which activates gene transcription in response to mitogenic stimuli. Accordingly, these interactions among CREB, CBP, and p300 may provide a molecular substrate for the observed effects elicited by cAMP on cell proliferation and differentiation.
IN STRUCTURE: VI. THE CREB AND CREM GENESARE MULTIEXONIC ALTERNATIVE EXONSPLICING GENERATES A COMPLEX ARRAY OF ISOPROTEINS THATAREEITHER
TRANSACTIVATORS OR TRANSREPRESSORS The genes for the human and mouse CREBs have been isolated and partially sequenced. The genes are located on the long arm of human chromosome 2 mapped to 2q32.3-q34 (Hoeffler et al., 1990; Taylor et al., 1990) and to the proximal region of mouse chromosome 1 (Cole et
22
JOEL F. HABENER et al.
t
A exon: 5'-Flank A
B '? C
Alternatively Spllced Exons
D
Y
E
F
G
1 W H
I
GENE
PRO1rElN CREE341 4
ACTIVATION DOMAIN
I
1P b X , ] I '? I
CREE327 CREE-'?
,
*,STOP
CREB-Y CREE-W CREE4
S
FIG.10. The exonic organizaton of the CREB and CREM genes, mRNAs, and encoded proteins. (A) The CREB gene contains a t least 12 exons spanning more than 80 kb of DNA. The introns between exons A and B and H and I are very large (over 25 kb). The exons comprise functional modules in CREB. Exons E and F encode the phosphorylated domain and exons B, C, and G encode the glutamine-rich domains, important for tran-
23
CREB AND CREM REGULATION
al., 19921, which shares a large region of synteny with human chromosome 2q. The chromosomal location of the CREM gene has not yet been reported. The genes for ATF-1 and 2 ATF-2, related in structure to the CREB and CREM genes, are located on chromosomes 12q13 and 2q24.1-q32, respectively (Zucman et al., 1993; Diep et al., 1991).
FUNCTIONALLY DISTINCT DOMAINS A. EXONSENCODE The CREB genes consist of multiple exons, at least 12, spanning a n estimated 80-100 kb (Hoeffler et al., 1990; Ruppert et al., 1992). The structures of the CREM and ATF-1 genes have not yet been reported. However, several different cDNAs representing mRNAs encoded by these genes provide strong evidence that they are similar to CREB in their composite multiexonic structures (Foulkes et al., 1991b, 1992; Fkhfuss et al., 1991). The exons that make up the CREB and CREM genes are functionally modular in nature (Fig. 10). Exons E and F encode the P box, or KID, while exons C and G encode glutamine-rich regions important in mediating the transactivational activity imparted by phosphorylations in the P-box domain. Exons H and I encode the DNA binding domain consisting of the basic region and the leucine zipper dimerization sequences. The multiexonic structures of the CREB, CREM, and ATF-1 genes stand in sharp contrast to the intron-
~~
~~
~
~
~
~
scriptional transactivation. Transcription initiates from a non TATA box sequence (Inr). Exons H and I encode the bZIP domain involved in dimerization and DNA binding. LLLL, leucine repeat dimerization domain. Notably, exons 9,Y, D, and W are alternatively spliced in tissue- and development-specific patterns. The II sequence is not an alternatively spliced exon, but rather is an intron slippage alternative acceptor splice. The alternative splicing of exon D retains the open translational reading frame in CREB mRNAs, resulting in the formation of the two CREB isoforms, CREB327 and CREB341, differing by 14 amino acids. However, exons 9,Y,W, and the alternative II splice result in the premature termination of translation of the CREB mRNA. These isoforms of CREB are truncated a t the C-terminal end, and consist of CREBs devoid of the DNAbinding bZIP domain and the encoded nuclear translocation signal. It is believed that the formation of these isoforms serves to interrupt a positive autofeedback regulation of CREB on its own promoter. (B) The exons of CREM. Although the structure of the CREM gene is unknown, the multiplicity of alternatively spliced mRNAs (cDNAs) identified defines a multiexonic structure homologous to that of CREB. The CREM gene has two promoters: A 5' constitutively acting promoter, P1 (left), and a CAMP-regulated internal promoter, P2, located in the intron between exons G and X (y). The P2 promoter encodes a transrepressor isoform of CREM, ICER (inducible CAMP early repressor), consisting of the DNA-binding bZIP domain. Two bZIP domains are encoded by alternative splicing of exon I (I, and Ib). The nomenclature for the CREM isoforms is described in Greek letters on the left and according to their exonic composition on the right. S-CREM is formed by a n alternative translational initiation (ATG).
~
24
JOEL F. HABENER et a1
less genes encoding the related Jun and C/EBP subfamilies of bZIP proteins. The nomenclature describing the numerous CREB and CREM isoforms expressed as a consequence of alternative splicing of exons is confusing and requires clarification. Historically, the exons in the human CREB gene originally were reported and described by the letters A through I (Hoeffler et al., 1990). Subsequently, the peptide sequence corresponding to exon D of the CREB gene was designated a,because it was believed to be an a-helix (Yamamoto et al., 1990). In a later publication describing the structure of the mouse CREB gene, the authors designated the exons homologous to those of the human gene by the numbers 1 to 11 and Greek letters (Ruppert et al., 1992). The homology of the structure of the CREM gene to that of the CREB gene was apparently not recognized, so various forms of the cloned cDNAs representing alternatively spliced transcripts of the CREM gene were arbitrarily designated by Greek letters that bear no relationship t o the Greek letter designations given to the mouse CREB gene or to the corresponding letters A to I initially appended to the exons of the human CREB gene (see Fig. 10 for a partial explanation of the exon designations). Several of the exons of the CREB and CREM genes are alternatively spliced in the formation of mRNAs during the development and differentiation of specific tissue phenotypes. At least four alternatively spliced exons of the CREB genes have been identified so far (exons q , Y, D, and W) (Hoeffler et al., 1990; Waeber et al., 1991; Waeber and Habener, 1992; Ruppert et al., 1992). In addition, the mouse CREB gene undergoes an alternative splice site selection, known as intron slippage, between exons H and I, resulting in the translation of an R or H' sequence of amino acids (Ruppert et al., 1992). A remarkable property of the alternatively spliced exons in the generation of CREB mRNAs is that, with the exception of alternatively spliced exon D, which maintains the open translational reading frame of CREB, all of the other alternatively spliced exons (q,Y, and W) and the intron slippage isoform R introduce stop codons that result in the termination of translation. These alternative exon-splicing events cause termination of the translation of CREB proteins such that they do not have DNA-binding bZIP domains. Thus, these alternative splicing choices, resulting in the premature termination of translation, appear to be designed to inactivate the synthesis of transactivator forms of CREB. As discussed further below (Section VIII,A), it is believed that the premature termination of translation by the splicing in of exon W in Sertoli cells during spermatogenesis, and perhaps that of exons v' and
CREB AND CREM REGULATION
25
Y as well, may serve to interrupt a positive autofeedback loop of CREB acting on its own promoter. Transcripts of the CREM gene consist of combinations of alternatively spliced exons C, E, F, G, X, and I. Exon I consists of two alternatively spliced exons, I, and Ib, encoding alternative DNA binding and dimerization domains of CREM. The reasons for the two alternatively spliced bZIP domains are unknown. It is tempting to speculate, however, that they may provide diversity at the level of specificity of dimerization to other bZIP proteins and/or specificities in the recognition and binding to CREs with variations in their nucleotide sequences. Deletions of exons internal in the coding sequence markedly attenuate transactivational activities (Section V1.B). B. REPRESSOR ISOFORMS OF CREM ARE GENERATED BY SEVERAL DIFFERENT MECHANISMS An intriguing property ofthe expression ofthe CREB and CREM genes is the generation of multiple isoforms of the proteins due to mechanisms of alternative exon splicing and alternative utilization of translational start sites within the spliced mRNAs (Fig. 11).These mechanisms result in the interconversion of transcriptional activator to transcriptional repressor forms of the proteins and appear to be developmentally regulated in specific cell lineages, as discussed later (Section VIII). The internal exons in CREM, C through G, are all involved in encoding sequences important for the mediation of the transcriptional transactivation functions of CREM. Exons E and F encode P box, or KID, whose phosphorylation by CAMP-dependent kinases, Ca2f / calmodulin-regulated kinases, and Ras-activated kinases are critical for generating the transactivation functions. Exons C and G encode glutamine-rich regions with hydrophobic clusters that are also important in transactivation. Deletion of one or more of these exons results in a marked attenuation of the transactivational activities without alterations of the dimerization or DNA binding functions. Thus, these internally exon-deleted forms of CREM act as transrepressors by way of either competing for CRE sites as homodimers or forming less active heterodimers with transactivator isoforms of CREM and CREB (Foulkes and Sassone-Corsi, 1992; DeGroot and Sassone-Corsi, 1993). Alternative splicing of CREM takes place during the chronological maturation of spermatogenesis in the mouse (Foulkes et al., 1992). During the first 2 weeks of postnatal development, the repressor isoforms of CREM (CREMAC,G) are expressed in the maturing germ cells. Between 3 and 4 weeks of development, the C and G exons are
26
JOEL F. HABENER et al.
CREB
Alternative exon spllcing terminates translation and prevents translocation to nucleus.
SL- - - - - - - - - - - - ,-
P-BOX
CREM
.I-----?
:-Bl7-;-zlP--: NTS
Alternative exon splicing deletes Q-rich transactivation domains.
Q
Q (CREMa) BR I ZIP
Internal Translation (S-CREM) TX START
- - - - - - p.Box: ----- - - - r----I.
. I
1 BR I
ZIP
I
I
Internal Promoter (ICER)
FIG.11. Mechanisms for the generation of inactive and transrepressor isoforms of CREB and CREM. Splicing in of exons that terminate translation (exon W) inactivates CREB by preventing the synthesis of the DNA binding domain. In CREM, splicing out of exons important for transactivator functions (CREMa and CREMAC-G) creates repressor isoforms, as does alternative internal transcription (S-CREM) and/or translation (ICER). TRX, transcription; TX, translation; Q, glutamine; BR, basic region; ZIP, leucine zipper; NTS, nuclear translocation signal.
spliced into the CREM transcripts, resulting in the expression of transactivator isoforms of CREM. These important observations suggest that the target genes whose expression is regulated by CAMP signaling are switched from a repressed to an activated state during spermatogenesis by a mechanism of genetically and temporally programmed alternative splicing of CREM transcripts so as to switch the synthesis of CREM transrepressors to CREM transactivators of gene transcription. In addition to alternative splicing of exons in CREM, two other mechanisms are utilized to generate transrepressor isoforms of CREM: internal transcription/translationand internal translation. An alternatively used internal promoter exon that resides in the intron located between exons G and X is utilized to transcribe a 5’ truncated mRNA that encodes a small form of CREM called ICER (Foulkes et al., 1991b). ICERs consist of exons H and I, or 1, (with or without exon X) encoding the DNA binding domain devoid of any known transactivation do-
CREB AND CREM REGULATION
27
mains, and thereby are potent transrepressors of activator forms of CREM and CREB. The internal promoter of ICER, called P2, is upregulated by four closely located CREs activated by activator isoforms of CREB, CREM, and potentially other CRE-binding proteins, and is repressed by itself. As discussed in more detail in Section VIII,C, the temporally programmed interplay of the positive and negative regulation of CAMPresponsive target genes and the ICER promoter itself is important in controlling the 24-h circadian rhythm of melatonin synthesis in the pineal gland. The other mechanism, alternative internal translation, does not involve the alternative splicing of exons, but rather involves the alternative utilization of a n internal AUG codon to start translation, resulting in the synthesis of a n N-terminally truncated repressor similar to the exon-deleted and ICER CREM isoforms. Such alternative translation of CREM mRNA has been shown to take place during postnatal development of the rat brain (Delmas et al., 1992). It is curious that in the CREB gene alternative splicing in of exons q ,Y, and W, and the intron slippage splice, a,all result in the premature termination of the translation of the CREB mRNAs and the formation of C-terminal truncated forms of CREB lacking a DNA binding and dimerization domain with its encoded nuclear translocation signal. In contrast, all of the multiply alternatively spliced exons of CREM maintain the open translational reading frame. C. ALTERNATIVE EXONSPLICING APPEARS TO PROVIDE A MECHANISM BY WHICHTO MODULATE THE TRANSACTIVATIONAL ACTIVITIES OF CREB AND CREM As discussed earlier, most, if not all, isoforms of CREB and CREM, and also perhaps ATF-1, can freely dimerize with one another to form a large number of different combinations of CAMP-responsive DNAbinding bZIP proteins. Inasmuch as phosphorylation by CAMP-dependent PKA, and/or other protein kinases on the critical serine in exon E, and the integrity of the glutamine-rich exons C and G are required for the transactivation functions of CREB and CREM, impairment of phosphorylation or deletions of such exons that leave the DNA binding domain intact create repressor isoforms. These transactivationdeficient isoforms can dimerize with themselves and thereby serve as competitive inhibitors of the binding of transactivator forms of CREB and CREM to CREs. Alternatively, they can form heterodimers with transactivator isoforms, therefore reducing their relative transactivational activities when bound to DNA. Experimentally mutated forms
28
JOEL F. HABENER et al.
of CREB have been constructed in such a way as to impair transactivation (e.g., mutations in the PKA phosphorylation site) and to restrict heterodimerization combination pairs (e.g., mutations in the leucine zipper) (Loriaux et al., 1993). These studies show that a CREB dimer, in which one of the two monomeric members of the dimer is mutated in the phosphorylation site and is unable to be phosphorylated by PKA and the other monomer is not mutated and is phosphorylated, displays 50% of the activity of the wild-type fully phosphorylated homodimer. Similar experiments analyzing the relative transactivational activity of mutated dimers among CREB and CREM isoforms reveal similar additivity of activities: a dimer consisting of an exon-deleted CREM monomer and an intact CREB monomer gives 30-40% of the activity of the CREB homodimer (Loriaux et al., 1994). These observations underscore the complex nature of the combinatorial code at play in the regulation of gene transcription. Multiple different CRE-binding dimers consisting of various combinations of CREB and CREM isoforms can occur. The composition of the specific heterodimers, whether activator or repressor isoforms, determines their relative effectiveness to stimulate gene transcription. The prevalence of the specific heterodimers would be dependent on their relative concentrations in the nucleus, determined by the relative rates of their synthesis and degradation, which are likely to be influenced by both positive and negative autofeedback control on the promoters of the CREB and CREM genes.
D. UNPHOSPHORYLATED CREB CAN REPRESSGENEEXPRESSION MEDIATED BY PHOSPHORYLATED CREB Based on the considerations of the formation of heterodimers among different isoforms of CREB and CREM discussed above, at least three examples have been described in which unphosphorylated CREB (dephosphoCREB) antagonizes CREB and serves as a negative regulator of transcription (Fig. 12). Expression of dephosphoCREB in the pituitary glands of transgenic mice impairs their development (Struthers et al., 1991). A transgene consisting of a CREB isoform in which the Ser133 phosphorylated by PKA is mutated to an alanine targeted to the growth hormone-producing somatotrophs results in dwarfism due t o a marked developmental deficiency of somatotrophs, but not of other cell types. The mutated dephosphoCREB expressed by the transgene apparently antagonized the actions of wild-type CREB, resulting in a failure of the somatotrophs to develop. DephosphoCREB also represses transcriptional expression of the fos
CREB AND CREM REGULATION
P
29
P CREB I
CRE
TTRX I' CRE 1 '
PhosphoCREB strong activator
P
DephosphoCREB weak activator
P
Competition phosphoCREB vs. dephosphoCREB
Heterodimer of phosphoCREB I dephosphoCREB weak activator
FIG.12. Unphosphorylated CREB (dephosphoCREB) serves as a negative regulator of phosphoCREB. Because dephosphoCREB and phosphoCREB dimerize and bind CAMPresponsive element equivalently, and dephosphoCREB has markedly reduced transactivational activity compared to phosphoCREB, dephosphoCREB competitively inhibits transactivation of transcription. Inasmuch as protein kinase A activity in nuclei is believed to be rate limiting for the phospohrylation of CREB, the ratio of dephosphoCREB to phosphoCREB is an important determinant of the transactivational activity.
gene. An expression vector encoding CREB transfected into NIH 3T3 cells repressed the serum-stimulated expression of the fos gene (Ofir et al., 1991). This negative regulation by CREB was alleviated when CREB was phospohrylated on Ser133by cotransfection and expression of the catalytic subunit of PKA. Further, expression of the dephosphoCREB in which Serl33 was mutated to alanine also repressed serum stimulation of fos, but repression was not alleviated by PKA. These findings indicate that dephosphoCREB functions as a transrepressor of transcription of a gene (fos) encoding a transcription factor. Another example of negative regulation by dephosphoCREB was shown in its inhibition of the activation of transcription by C/EBP (Vallejo et al., 1995). The somatostatin gene is strongly up-regulated by cAMP via the interactions of CREB with a CRE located in the proximal region of its promoter. A somatostatin-producing cell line was discovered in which neither cAMP nor PKA could activate the somatostatin promoter. Rather, the unusually high constitutive basal transcription was driven by interactions of C/EBP with the CRE of the promoter. Further, it was shown that the PKA activity in these cells was severely inhibited by the overproduction of a n endogenous heat-stable inhibitor of PKA, resulting in a complete failure of the phosphorylation of CREB in response to the CAMP-PKA signaling pathway. Overexpression of
30
JOEL F. HABENER et al.
CREB transfected into the cells markedly attenuated transcription mediated by the CRE in the somatostatin promoter by competitively inhibiting the binding of C/EBP t o the CRE. These findings bring to light several potentially important insights. Namely, that transcription factors other than CREB, such as the C/EBP family of bZIP proteins, can bind to and activate CREs in the promoters of genes, that specific protein kinase inhibitors can markedly influence the activity of a signal transduction pathway, and that cells (the islet cell line) can survive and propagate in culture under conditions in which the nuclear CAMP-PKA pathway is essentially completely inactivated. Inasmuch as dephosphoCREB can serve as a negative regulator of phosphoCREB and other CRE-binding transactivators, the relative ratios of phosphoCREB and dephosphoCREB in the nucleus at any given moment will determine the relative magnitude of the CAMP-mediated transcriptional response. As discussed earlier, CREB is translocated constitutively t o the nucleus soon after completion of its synthesis ( t l i P of nuclear import, -15 min). In addition, nuclear CREB appears to be associated with nuclear chromatin at all times, presumably bound to CRE control elements of the DNA (Hagiwara et al., 1993). The regulated step in the phosphorylation of CREB is the availability of active catalytic subunit of PKA. The PKA holoenzyme type I1 is located in the perinuclear cytoplasm bound by specific AKAPs (Scott and McCartney, 1994). Upon activation by binding of cAMP to the regulator subunit, the catalytic subunit is released and translocates to the nucleus within 5-15 min (Hagiwara et al., 1993; Meinkoth et al., 1990; Nigg et al., 1985a).Holoenzyme probably also exists within the nucleus as well as in the perinuclear cytoplasm (Squint0 and Jungmann, 1989; Joachim and Schwoch, 1988; Schwoch and Freimann, 1986). AKAP-95 has been identified that is associated with the nuclear matrix and contains a zinc finger DNA binding domain (Coghlan et al., 1994). It seems likely that AKAP-95 tethers PKA holoenzyme to the nuclear matrix through interactions with the type I1 regulatory subunits (RIIa and RIIP). It has been observed that the addition of 8Br-CAMP to primary hepatocytes in uzuo results in the expression of nuclear PKA activity and the phospohrylation of CREB within 30-60 s (Gosse et al., 1995).These findings suggest that the stimulation of gene transcription in response to cAMP can occur very rapidly via activation of nuclear PKA, whereas longer, more sustained, responses occur by the recruitment to the nucleus of PKA catalytic subunits released from holoenzyme in the cytoplasm. The nuclear levels of PKA appear to be the rate-limiting step in
CREB AND CREM REGULATION
31
determining the ratio of phosphoCREB to dephosphoCREB. In studies of PC12 cells, it was estimated that a maximum of 40% of nuclear CREB is phosphorylated 30 min after the addition of forskolin (Hagiwara et al., 1993). Although the concentration of PKA in the nucleus (1.2 F M ) exceeds that of CREB (400 nM), the K , of CREB for PKA is relatively high (10 p M ) (Calbron et al., 1992). A tentative extrapolation of these findings suggests that in the absence or low levels of cAMP signaling, the vast majority of CREB is unphosphorylated and is serving a role as a repressor rather than as an activator of gene transcription. E. EXON-DELETED REPRESSOR ISOFORM OF CREM DOWN-REGULATES EXPRESSION OF THE c-fos AND c-jun GENES The alternatively spliced isoform of CREM that has the two glutamine-rich exons, C and G, deleted (CREMAC,G or CREMd is a repressor of cAMP stimluation of the c-fos gene in NIH 3T3 cells (Foulkes et al., 1991a). CREM does not antagonize serum-induced transcription of the c-fos gene, suggesting that dephosphoCREB and CREMAC,G act on the promoter of the c-fos gene by different mechanisms. Both CREB and CREMAG,C also down-regulate the expression of the c-jun gene in JEG-3 cells (Masquilier and Sassone-Corsi, 1992). The mechanism of the inhibition of the activation of c-jun is a direct competition by CREB and CREMAC,G of the binding of c-jun (and c-fos) to the TPA-response elements in the promoter of the human metallothionein IIA gene, not by dimerization of CREB and CREM with c-jun or c-fos. VII. CAMP-DEPENDENT AUTOREGULATION OF THE EXPRESSION OF THE CREB AND CREM GENES The transcription of both the CREB and CREM genes is autoregulated by their own encoded products, serving as important CAMPresponsive transcription factors. The CREB gene is positively autoregulated via three CREs that reside within the promoter consisting of 800 bp of sequence flanking the 5‘ end of the gene, that is, 5’ to exon A (Meyer et al., 1993; Walker et al., 1995) (Fig. 12). CREB produced by recombinant DNA techniques as well as CREB in extracts of cell nuclei bind to the CREs in the promoter of the CREB gene (Meyer et al., 1993; Walker et al., 1995). The CREM gene has an unusual dual-promoter structure. The 5’
32
JOEL F. HABENER et a1
A ATG
-1200
-600
+1
InR
B TRX
--I CRE HCRE
TRX
TX rMet
Ala Val
FIG.13. The human CREB and internal CREM (P2) gene promoters. (A) The 5’ flanking 1200 bp of the CREB promoter are enriched in G and C (70%). The transcription (TRX) is initiated from a n InR sequence and not a TATA box. Important control elements are three cAMP response elements (CREs) and three sites that bind the transcription factor Spl. The CREs lend autopositive regulation to the CREB gene. (B) The internal promoter (P2)of the CREM gene that encodes the transrepressor isoform (ICER) is upregulated by four closely spaced CREs. This mechanism provides an autoregulation as CREB and CREM transactivators activate the ICER promoter, resulting in the production of a dominant negative repressor (see also Fig. 10B). TX, translation.
flanking promoter (Pl) is constitutive and is not regulated by cAMP (Molina et al., 1993).Rather, the CREM gene contains a second promoter (P2) located within the intron between exons G and X that contains four closely spaced CREs [redesignated as CARES(Molina et al., 199311 (Fig. 13). The activation of the P2 promoter by transactivator forms of CREM and CREB results in the expression of a novel mRNA that encodes a transrepressor isoform of CREM called ICER, as mentioned earlier. Thus, in contrast to that of the CREB gene, cAMP signaling of the CREM gene results in the expression of a transrepressor of gene transcription. This fascinating interplay of positive and negative autoregulation of CREB and CREM transcription factors is discussed further in Section VIII in the context of their importance in the physiological regulation of gene expression. VIII. RQLESOF CREB AND CREM IN THE PHYSIOLOGICAL REGULATION OF GENETRANSCRIPTION Increasing evidence indicates that both CREB and CREM regulate certain cellular processes in several tissues of the intact organism, including the testes, the pituitary gland, and the brain. However, their exact physiological roles have not been completely established.
CREB AND CREM REGULATION
A.
33
TESTES
The CREB and CREM genes are highly expressed in the testes. The primary transcripts of these two genes undergo complex patterns of alternative splicing of exons during both the postnatal developmental maturation of spermatogenesis and the endogenous cycling of the seminiferous tubules in the adult animal. The alternative splicing results in the interconversions of transactivator, transrepressor, and inactive isoforms of CREB and CREM. The seminiferous tubules consist of the germ cells at various stages of development, the somatic Sertoli cells, and the intestinal Leydig cells. The pituitary gonadotropic hormones follicle-stimulating hormone (FSH) and luteinizing hormone (LH) stimulate CAMP-coupled receptors on the Sertoli and Leydig cells, respectively. The actions of FSH and LH, expressed from the pituitary gland during pubescence, activate the Sertoli cells to produce estrogenic hormones and other essential factors and activate the Leydig cells t o produce androgenic hormones, all of which are required to allow for the orderly maturation of the germ cells. Further, the dual actions of the FSH and LH gonadotropins are essential for driving the waves of germ cell differentiation in the testes of the adult animal. The program of germ cell differentiation takes 45-48 days to complete (Fig. 14A). Given the appropriate hormonal stimuli, the stem cell spermatogonia undergo three or four mitotic divisions to mature into spermatocytes, which then differerntiate through several additional stages (i.e., leptotene to zyotene to pachytene). The pachytene spermatocytes undergo two meiotic “divisions”: first segregation of diploid into haploid chromosomes, and then cell division into round spermatids, each of which contains a haploid complement of chromosomes. The round spermatids then undergo a process of morphogenesis to become elongated spermatids and finally mature spermatozoa that exit from the seminiferous tubules to the epididymis, where they are avaialable for fertilization. Spermatogenesis in the testes of rodents occurs in waves of approximately 12 days’ duration involving 14 cell association stages (stages I-XIV). Four such cycles (48 days) are required for the developmental maturation of stem cell spermatogonia to become mature spermatozoa (Fig. 14A). 1. Expression of the CREB Gene The CREB gene is expressed at high levels in both the Sertoli cells and the germ cells. Studies have been focused on the examination of the cyclical expression of CREB in Sertoli cells of the adult rat semi-
34
JOEL F. HABENER et a1
A I
i
'!
,
STAGES 12.9 days
I
+
B
4
z K
E 50 m
w K
0
I
11-111
+-12.9
VI VllC-d IX-XI XIII-: IV-V Vlla-b Vlll XI1 STAGE days-4
FIG.14. Cyclical expression of the CREB gene in Sertoli cells of rat seminiferous tubules during the 12-day spermatogenic cycle. (A) Depiction of the 14 germ cell association stages (I-XIV) in the seminiferous tubule of the adult rat. Shown is a diagrammatic cross-section of a tubule with the luminal and basal borders a t the top and bottom, respectively. The cells a t the bottom are the progenitor spermatogonia that give rise to spermatocytes above as they develop from left to right and progress up the staircase (bottom layer to top layer). The mature pachytene spermatocytes undergo two rapid meiotic divisions a t stage XIV (right, third cell layer from the bottom), resulting in the formation of the haploid round spermatids that then develop into mature spermatozoa. Approximately 12 days are required for each of the four layers of cells to traverse development from left to right (stages I-XIV). The entire time for development from
35
CREB AND CREM REGULATION
niferous tubule during the 12-day cell association cycles of maturation of the germ cells (Waeber et al., 1991) (Fig. 14). It has been proposed that cAMP generated by the action of FSH on Sertoli cells stimulates the expression of the CREB gene via CREs located in the promoter of the CREB gene. The positive autoregulation of the expression of CREB is interrupted by a switch in the processing of the CREB RNA transcripts, so as to splice an exon into the mRNA that terminates translation 5’ to the mRNA sequence encoding the bZIP DNA binding domain of CREB with its encoded nuclear translocation signal (Waeber and Habener, 1991).The C-terminal truncated isoforms of CREB, devoid of their nuclear translocation signal, become sequestered in the cytoplasm and are degraded. This mechanism of alternative splicing of exons serves as a means to inactivate the positive autoregulation of CREB. Such alternative splicing of at least two separate exons that result in the premature termination of translation of CREB mRNAs takes place in cell association stages V-VI of the seminiferous tubules of the rat (Waeber et al., 1991; Waeber and Habener, 1992). These exons appear to be spliced out during stages
“ f Bpp*r receptor
nucleus
y
\ -’u cytoplasm
A
STOP
---->
spermatogonia to fully mature spermatocyte takes approximately 48 days (four cycles of 12 days each). (B) CREB mRNA levels in Sertoli cells rise and fall during the 12-day cycle of spermatogenesis. The fluctuations in cAMP levels and follicle-stimulating hormone (FSH)binding sites precedes that of CREB mRNA, consistent with the autoregulation of CREB gene expression by FSH-mediated increases in CAMP.It has been proposed that the alternative splicing in of exon W, which stops translation before the DNA binding domain, interrupts the positive auto feedback loop of CREB gene transcription (see Fig. 10A). (C) Model depicting the FSH-regulated cyclical expression of CREB in Sertoli cells. PKA, protein kinase A; FSH-R, FSH receptor; NLS, nuclear localization signal; P, phosphoserine.
36
JOEL F. HABENER et al.
XIV-I, thereby allowing the synthesis of complete full-length CREB that can enter the nucleus and activate transcription (Fig. 14C).Notably, the mRNA encoding the FSH receptor in the testes appears also to be alternatively spliced by a similar mechanism, that is, the splicing in and out of exons that results in premature termination of translation N-terminal to the transmembrane-spanning domain that anchors the receptor in the plasma membrane. It has been proposed that the alternative splicing of the FSH receptor during the 12-day temporal cycles of the seminiferous tubule provides cyclical regulation of cAMP signaling in Sertoli cells (Walker et al., 1995). It is also possible that alternative splicing of the FSH receptor mRNAs may occur during the postnatal maturation of the testes and serve to activate FSH-mediated spermatogenesis at puberty. 2, Expression of the CREM Gene
As discussed above, the CREM gene encodes multiple regulators of the cAMP transcriptional response by alternative splicing (Foulkes et al., 1991b).In mice a developmental switch in CREM expression occurs during spermatogenesis, whereby alternative exon splicing converts CREM from a repressor to an activator (Foulkes et al., 1992).In premeiotic germ cells the CREMa (CREMAC,G)repressor isoform is expressed in low amounts. During the developmental transition of the spermatocytes through the pachytene stages VI-VIII, the glutamine-rich C and G exons are spliced into the CREM mRNAs, resulting in the translation of the CREMT(CREM I,) activator isoform (Fig. 10). The CREMTtransactivator accumulates to high amounts. During the early stages of spermiogenesis (stages X-XII), when the round spermatids elongate to spermatozoa, a new repressor isoform, CREMAC-G, is expressed (Walker et al., 1994). This repressor isoform of CREM (CREMAC-G) consists only of the N-terminal 38 amino acids of exon B and the 12 amino acids of exon X in frame with the exons encoding the DNAbinding bZIP domain (exons H and IJ. These circumstances suggest that during spermatogenesis, CREM gene expression undergoes a biphasic transition of isoforms: from repressor (prepachytene spermatocytes), to activator (postpachytene spermatocytes), to repressor (transition from round to elongated spermatids). These splicing-dependent changes in the transcriptional functions of CREM, along with the transitions in CREB splicing, are proposed to have an important role in programming the expression of CAMP-regulated target genes during spermatogenesis, although the detailed mechanisms involved are not yet understood (Walker et al., 1995). In the postnatal prepubertal rat FSH appears to be an important
CREB AND CREM REGULATION
37
functional switch necessary for the expression of CREM transactivator forms in the testes. Hypophysectomy of newly born rat pups, which eliminates production of FSH, results in the extinction of CREBT expression. Expression can be restored by the administration of exogenous FSH. The induction of CREMT by FSH appears to occur, to a large extent, by alternative utilization of a poly(A) addition site in the CREM mRNA, resulting in a marked stabilization of the mRNA (Foulkes et al., 1993). Analyses of CREM expression in the season-dependent modulation of spermatogenesis in hamsters have shown that FSH controls the switching on of CREM gene expression in the testes during the beginning of the experimental change from short to long photoperiods (summer), when FSH and LH levels are induced (Foulkes et al., 1993). B. ANTERIORPITUITARY GLAND In the anterior pituitary CREB may be involved in the regulation of the CAMP-dependentproliferation of somatotrophs. This notion is supported by loss-of-function experiments carried out by Struthers et al. (1991) in transgenic mice (described above) and by gain-of-function experiments carried out by Burton et al. (19911. These authors developed transgenic mice carrying a chimeric gene encoding an intracellular form of cholera toxin under the control of the growth hormone gene promoter. In these animals cholera toxin specifically expressed in somatotrophs irreversibly stimulates G, protein-mediated activation of AC, resulting in permanently elevated concentrations of CAMP.The phenotype of mice carrying this transgene is characterized by gigantism, hyperproliferation of somatotrophs, and pituitary hyperplasia (Burton et al., 1991).A clinical correlate of these observations is found in patients with pituitary adenomas due to constitutively active mutant forms of G, proteins (Landis et al., 1989).The possible existence of mutated CREB in other somatotroph adenomas has not been examined.
C. BRAIN:HYPOTHALAMUS AND PINEAL GLAND In the brain the distribution of transcripts encoding CREB and CREMT is diffuse, whereas the distribution of transcripts encoding repressor isoforms of CREM is restricted to specific areas (Mellstrom et al., 1993). This pattern of expression suggests that in the central nervous system the presence of CREM antagonists may determine region-specific differences in CREB- (or CREMT-)mediated responses
38
JOEL F. HABENER et a1
to cAMP stimulation. Peripheral hyperosmotic stimulation results in CREB phosphorylation in the hypothalamic supraoptic and paraventricular nuclei (Borsook et al., 1994). In addition, osmotic stimulation results in the induction of CREMa and CREMp in neurons of the supraoptic nucleus (Mellstrom et al., 1993). These findings suggest that these transcription factors are involved in the control of hypothalamic homeostatic mechanisms that maintain plasma osmolality. CREB and CREM have been implicated in the physiological mechanisms of the regulation of circadian rhythms (Ginty et al., 1993; Stehle et al., 1993). A pacemaker that controls circadian rhythms resides in the suprachiasmatic nucleus (SCN) of the hypothalamus. This nucleus receives inputs regarding light-dark cycles via direct retinohypothalamic projections that relay information received by the eye. In turn, the SCN controls the functional activity of noradrenergic inputs reaching the pineal gland, which produces the hormone melatonin. In this manner melatonin production and secretion follow the SCN-driven circadian rhythm. In the course of studies to elucidate the molecular mechanisms that synchronize the circadian pacemaker, it was found that in the SCN, CREB becomes rapidly phosphorylated in response to light (Ginty et al., 1993). This observation supports earlier studies indicating a n important role for cAMP in the regulation of circadian rhythms in the SCN (Murakami and Takahashi, 1983; Prosser and Gillette, 1989). At the level of the pineal gland, noradrenergic inputs controlled by the SCN activate AC-coupled p-adrenergic receptors, resulting in increased levels of cAMP in pinealocytes (Fig. 15). This elevation in cAMP levels induces transcription of the enzyme arylalkylamine N-acetyltransferase (NAT), which controls the synthesis of melatonin. Since CAMP-regulated NAT production is dependent on sympathetic stimulation controlled by the SCN, the concentration of NAT in pinealocytes exhibits a circadian rhymicity. What are the molecular bases of this circadian variation? One possible mechanism for this circadian variation is the CAMP-induced expression of ICER, a transcriptional repressor of CREB and CREM activator isoforms, under circadian control in the pineal gland. The expression of ICER in the pineal gland occurs in response to p-adrenergic stimulation, and consequently increases during the night and is lower during the day. Therefore, it is possible that elevations in cAMP levels in response to P-adrenergic stimulation at night results in the PKA-dependent phosphorylation of CREB (or CREMT),with subsequent transcriptional activation of the NAT gene. At the same time, CAMP-dependent activation of the CREM gene through the P2 promoter results in the production of ICER. After
CREB AND CREM REGULATION
39
FIG.15. Generation of a circadian rhythm in the pineal gland. pl-Adrenergic receptor stimulation driven by the suprachiasmatic nucleus (through an indirect pathway involving the superior cervical ganglion) results in the activation of adenylyl cyclase (AC) and production of CAMP. Activation of protein kinase A (PKA)follows due to the dissociation of the catalytic subunits from the CAMP-bound regulatory subunits. The catalytic subunits enter the nucleus of the pinealocyte, where it phosphorylates CREB. As a result, transcription of the gene encoding the melatonin-synthesizing enzyme NAT is activated. At the same time, transcription of the P2 promoter of the CREM gene is also activated, resulting in the production of the transcriptional repressor ICER. The levels of ICER gradually increase until they are high enough to compete for CRE binding with CREB, resulting in transcriptional inhibition of the NAT and CREM genes. In this manner, the activation triggered by stimulation of pl-adrenergic receptor returns to basal levels before the following cycle. G, G-protein; CREBP, phosphorylated CREB.
a lag time the intracellular levels of ICER become sufficiently high to compete with CREB for binding to the CRE, resulting in transcriptional repression of the NAT and CREM P2 promoters. Although the target genes regulated by CREB and CREM in the brain are unknown, these observations indicate that both of these transcription factors participate in neuronal signaling mechanisms in the hypothalamus, and may play an important role in the circadian mechanisms that control neuronal and hormonal responses to lightdark cycles.
D. POSSIBLE ROLEOF CREB IN MEMORY Emerging evidence indicates that CREB is also required for the normal process of higher brain functions such as memory consolidation (reviewed by Frank and Greenberg, 1994; Stevens, 1994). In an
40
JOEL F. HABENER et a1
effort to elucidate the physiological functions of CREB in uiuo, gene targeting via homologous recombination was used to inactivate the CREB gene (Hummler et al., 1994). Initially, the phenotype of homozygous mutant mice generated by this knockout of the CREB gene was found to be apparently normal, although these animals exhibit increased levels of CREM, indicating the existence of a considerable degree of functional redundancy to provide a back-up mechanism that compensates for the absence of CREB. A more detailed analysis of mice carrying the targeted mutation in the CREB gene, however, revealed a deficiency in long-term memory in these mice, although short-term memory is normal (Bourtchuladze et aZ., 1994). Interestingly, the fruitfly Drosophila homolog of CREB also appears to be involved in memory consolidation. This notion is indicated by experiments showing that the induced expression of a transgene that acts as a dominant negative regulator of Drosophila CREB results in a profound deficit in long-term memory associated with olfactory stimuli (Yin et al., 1994). This role of CREB in memory consolidation is presumably related to the CAMP-dependent transcriptional regulation of genes whose expression increases synaptic strengths in specific neuronal populations. It has been proposed that memories are encoded as patterns of synaptic strengths, and long-term potentiation (LTP) is a mechanism for the longlasting modification of synaptic strengths (Stevens, 1994). Consistent with the notion that CREB plays a role in memory consolidation by mediating LTP in the hippocampus, it has been found that LTP occurs via a CAMP-dependent mechanism (Huang et al., 19941, and that in mice carrying a targeted mutation of the CREB gene the stability of LTP is significantly reduced (Bourtchuladze et al., 1994). The hypothesis that CREB may be critically important for memory was initially derived from studies carried out in the sea slug Aplysia (Dash et al., 1990). Prolonged application or intracellular injection of cAMP into sensory neurons of Aplysia produced long-term increases in synaptic strength, suggesting that some of the gene products important for long-term facilitation are cAMP inducible. These findings in molusks, insects, and mammals identify CREB as a key component of an evolutionary conserved mechanism for memory consolidation.
IX. CREBXREM AUTOREGULATION NETWORK The promoters of the CREB and CREM genes contain several CRElike sequences. The binding of CREM and CREB proteins to these
CREB AND CREM REGULATION
41
sequences supports the concept that these genes are part of an autoregulatory network that mediates both positive and negative feedback regulation (Fig. 16). Transcriptional transactivation of the CREB and CREM genes in response to cAMP stimulation (Meyer et al., 1993; Molina et al., 1993) results in the synthesis of mRNAs encoding proteins (CREB and CREM isoforms) that, in turn, may alter their own CAMP-dependent transcriptional responses by determining the relative levels of transactivator or transrepressor isoforms. This CAMPdependent autoregulatory network also provides a mechanism for the regulated control of transcription rates of different target genes after cAMP stimulation. The idea of a positive feedback mechanism that stimulates the CAMP-dependent transcription of the CREB gene is supported by the existence of CRE sequences in its promoter and by transient transfection studies using reporter plasmids bearing segments of the CREB gene promoter (Meyer et al., 1993; Walker et al., 1995). According to this model, activation of the cAMP signaling pathway results in the
FIG.16. CREB-CREM autoregulatory network. Membrane receptor stimulation results in the activation of adenylyl cyclase (AC) and production of CAMP,which, in turn, activates protein kinase A (PKA). After translocating to the nucleus, PKA phosphorylates CREB, triggering the transcriptional activation of target genes (TG), including the gene encoding CREB itself (positive autoregulatory loop) and the CREM gene through the P2 promoter. Subsequent production of ICER results in the transcriptional repression of CREB-activated genes by competition with CREB for binding to the CRE. G, G-protein; P, phosphorylation.
42
JOEL F. HABENER et al.
stimulation of transcription of CREB and other target genes via phosphorylation of preexisting nuclear CREB or CREMT proteins. At the same time, transcription rates from the CREM gene P2 promoter can be also stimulated (Molina et al., 1993; Stehle et al., 19931, resulting in the production of the CREM repressor isoform, ICER. As the levels of ICER proteins increase, they occupy the CRE sites, repressing their own transcription, via an autoregulatory mechanism, as well as that of other CAMP-inducible genes, perhaps including the CREB gene (although this possibility has not been studied). Therefore, ICER acts as a delayed switch that resets the transcription rates of CAMP-inducible genes to basal levels. This hypothetical autoregulatory network may not be present in all cell types. Although the CREB gene seems t o be ubiquitously expressed, different CREM isoforms appear to be preferentially expressed in some tissues, including brain, pineal gland, and testes. Therefore, the relative concentrations of activator and repressor isoforms of these CREBs will determine the degree of expression of target genes in a spatial and temporal manner. In addition, these different isoforms interact to form homo- and hetoerdimers on the target CREs, adding another level of complexity to the regulatory pathway that controls CAMP-dependent gene expression.
X. ONCOGENEFORMS OF CREB, CREM, AND ATF-1 To date, the single example of an oncoprotein involving the CAMPresponsive factors is a fusion protein between ATF-1 and the EWS protein, an RNA-binding protein identified as a hybrid transcript in Ewing’s sarcoma that links its N-terminal domain to the ETs region of the DNA binding domain of the FLI-1 gene (Zucman et al., 1993) (Fig. 17).The expression of the EWS/ATF-1fusion gene results in a rare form of cancer called melanoma of soft parts. The expressed fusion gene consists of a balanced reciprocal translocation between chromosomes 12 and 22, which encode ATF-1 and EWS, respectively [t(12;22)(q13;q12)1. The N-terminal region of the EWS gene is fused to the C-proximal region of ATF-1 consisting of the DNA-binding bZIP domain. The fusion occurs in the intron between exons E and F that encode the P box of ATF-1. The N-terminal segment of the EWS RNA-binding protein is a potent transactivator of gene transcription, presumably targeted by the binding of the ATF-1 bZIP domain to CREs in the promoters of genes resulting in transformation. Several such fusion oncoproteins between
43
CREB AND CREM REGULATION NTPEWS
656
EWS NHz
COOH
I bZlP I
EWS/ATF-1 NHz
COOH
t
65
ATF-1
ATF-llEWS
FIG.17. Reciprocal translocation between chromosomes 12 and 22 [t(12;22) 1(13;12)1 causes melanoma of soft parts. The translocation results in the expression of a fusion protein consisting of the transactivating region (NTD-EWS) of the EWS RNA-binding protein with the C-proximaliDNA-binding (bZIP) domain of ATF1-1. BD, Binding domain; PKA, protein kinase A.
RNA-binding proteins and DNA binding domains of transcription factors have been described (reviewed in Rabbits, 1994). It is notable that, as yet, no subjects with oncogenic isoforms of CREB or CREM have been uncovered, particularly since the multiply spliced exons and large introns in these genes would theoretically predispose the subject to miss-splice mutations or recombinatorial translocations. Interestingly, several other bZIP proteins related to CREB and CREM, such as c-Jun, c-Fos, and c-Myc,were originally discovered as their oncogenic variants, in the guise of “tumor viruses” in chickens and rabbits. XI. FUTURE DIRECTIONS Tremendous strides have been made during the past several years in the enhancement of our understanding of the mechanisms by which cAMP signaling regulates gene transcription. The CAMP-responsive DNA-binding proteins CREB, CREM, and ATF-1 appear to comprise the fourth and final components of the cAMP signaling cascade in the control of gene expression. However, many questions remain unanswered and much more investigative research remains to be done to resolve these issues. Certain of the important questions that still require further experimentation are considered in this section. One important area of investigation is to decipher the protein-DNA recognition code by which homodimer and heterodimer combinations
44
JOEL F. HABENER et al.
of the CAMP-responsive DNA-binding proteins recognize and bind to composite CREs in the promoters of genes. Assuming that these proteins can freely dimerize with one another, the four known CAMPresponsive transcription factors, CREB, ATF-1, and the two alternatively spliced bZIP domains of CREM, can form 10 unique homoand heterodimer combinations (unique combinations = [ n x ( n + 1)]/2, where n represents the number of unique proteins). It is possible that there exist additional, as yet undiscovered, CAMP-responsive transcription families that can dimerize with CREB, CREM, or ATF-1. For example, the hypothetical discovery of two additional such proteins would increase the possible dimer combinations t o 21, thereby expanding even further the theoretical repertoire of different CRE motifs with which the dimers could interact. The number of dimer combinations is already theoretically even larger, given the evidence for the multiple alternatively spliced CREM isoforms (ICER, CREMa, CREMAC-G, and others; see Fig. 10B). The numbers could be even larger if circumstances are found in which CREB, CREM, and ATF-1 isoproteins are found to dimerize with other transcription factors, such as J u n and/or C/EBP members of the bZIP family of proteins. The formation of specific dimers of transcription factors within the nucleus predictably would depend on the concentrations of the proteins relative to one another, which, in turn, depend on their relative rates of formation and degradation. In many instances it has been established that the expression of genes encoding DNA-binding transcription factors is regulated by either autopositive or autonegative feedback control mechanisms, as has been found for CREB and for the ICER isoform of CREM. Important directions of investigation will be to decipher the mechanisms that control the complex autoregulatory networks of gene transcription, for example, elucidation of the feedback control mechanism by which targeted disruption of the CREB gene in mice leads to a 10-fold compensatory increase in the expression of the CREM gene. Current evidence indicates that, upon synthesis, CREB, and perhaps also CREM and ATF-1, is constitutively translocated to the nucleus. The nuclear translocation and activation of protein kinases and phosphatases appear to be the regulated steps in modulating the transactivation functions of CREB. An intriguing question regards the topographical compartmentalization of CREB and the kinases and phosphatases within the nucleus. Do all of these proteins comprise a ternary transcriptional complex? Is CREB bound more or less irreversibly to CREs of gene promoters and does it await interaction with protein kinases and phosphatases to regulate transcription? High-
CREB AND CREM REGULATION
45
resolution confocal immunofluorescence microscopy may help to resolve these questions. Another question relates to the mechanisms of the interactions of CREB with CBP and with TAFII,,,, and how they couple to the basal RNA polymerase I1 factors, TFIIB and TFIID, respectively. How do phosphorylations of CREB and CBP by PKA lead to the formation of transcriptionally productive interactions? What are the changes in the structures of the proteins in response to phosphorylations? Eventually, the structures may be provided by solving X-ray diffraction patterns derived from cocrystals of CREB and CBP binding domains. The rationale for the existence of two apparently separate mechanisms for the coupling of CREB to the basal polymerase I1 complex requires further investigation. The coupling of the CREB P box to CBP is entirely cAMP dependent, whereas the coupling of the glutaminerich domains of CREB to TAFII,,, appears t o occur independently of cAMP signaling. Do both mechanisms occur simultaneously? Are they utilized independently during development and/or in cells of different phenotypes? The highly complex multiexonic structures of CREB and CREM are unusual among transcription factors and require further elucidation. Are there yet additional undiscovered alternatively spliced exons in the CREB and CREM genes? How extensive is the alternative exon splicing among different tissues? Is alternative exon splicing developmentally or metabolically regulated to interconvert transactivator to transrepressor isoforms? Further investigations are needed of the biological consequences of alternative exon splicing of the CREB and CREM transcripts. Transcription of the CREB gene is up-regulated by cAMP to provide potent transactivator isoforms. Up-regulation of CREB appears to be interrupted at the cellular level by the splicing in of one or more exons that terminate translation. No such exons that terminate translation appear in CREM. Rather, in CREM the expression of the transactivator isoforms appears to be constitutively regulated, and cAMP upregulates a second promoter located in an internal alternatively utilized and spliced exon to produce a potent transrepressor isoform (ICER). It will be important to determine how ubiquitous the CAMPregulated expression of ICER is. Also, the functional importance of the two alternatively spliced exons encoding two different DNA binding domains of CREM remains to be elucidated. Further, a major future direction of research will be to determine how the alternative splicing of CREB and CREM is regulated. What are the factors and determinants involved in the regulation of the
46
JOEL F. HABENER et al.
specific RNA-splicing factors (splicosomes) that govern specific patterns of the alternative splicing of exons? Finally, it will be most exciting to determine the roles of CREB, CREM, and ATF-1 in development. Already, CREB functions have been implicated in spermatogenesis, pituitary development, and longterm memory consolidation. CREM is implicated as essential in spermatogenesis, the circadian regulation of melatonin synthesis in the pineal gland, and the cell division cycle. It is tempting to speculate that the ICER isoform of CREM may be a master down-regulator of the CAMP-dependent transcription of many essential genes in many different tissues in which gene expression in response to signaling is cyclical. ACKNOWLEDGMENTS We thank Doris A. Stoffers and Edward V. Maytin for critical reading of the manuscript and for helpful suggestions, and Townley G. Budde for preparation of the manuscript and the figures. J.F.H. is a n established investigator of the Howard Hughes Medical Institute. REFERENCES Alberts, A. S., Arias, J., Hagiwara, M., Montminy, M. R., and Feramisco, J . R. (1994a). Recombinant cyclic AMP response element binding protein (CREB) phosphorylated on Ser-133 is transcriptionally active upon its introduction into fibroblast nuclei. J . Biol. Chem. 269, 7623-7630. Alberts, A. S., Montminy, M. R., Shenolikar, S., and Feramisco, J . R. (199413).Expression of a peptide inhibitor of protein phosphatase 1 increases phosphorylation and activity of CREB in NIH 3T3 fibroblasts. Mol. Cell. Biol. 14, 4398-4407. Andrisani, 0. M., Pot, D. A,, Zhu, Z., and Dixon, J. E. (1988). Three sequence-specific DNA-protein complexes are formed with the same promoter element essential for expression of the rat somatostatin gene. Mol. Cell. Biol. 8, 1947-1956. Arany, Z., Sellers, W. R., Livingston, D. M., and Eckner, R. (1994). E1A-associated p300 and CREB-associated CBP belong to a conserved family of coactivators. Cell 77, 799-800. Arias, J., Alberts, A. S.,Brindle, P., Claret, F. X., Smeal, T., Karin, M., Feramisco, J., and Montminy, M. R. (1994).Activation of CAMP and mitogen responsive genes relies on a common nuclear factor. Nature 370, 226-229. Bading, H., Ginty, D. D., and Greenberg, M. E. (1993). Regulation of gene expression in hippocampal neurons by distinct calcium signaling pathways. Science 260,181-186. Bakker, O., and Parker, M. G. (1991). CAAT/enhancer binding protein is able to bind to ATFiCRE elements. Nucleic Acids Res. 19, 1213-1217. Beato, M. (1989). Gene regulation by steroid hormones. Cell 56, 535-544. Berger, S. L., Piba, B., Silverman, N., Marcus, G. A,, Agapite, J., Regier, J. L., Triezenberg, S. J., and Guarente, L. (1992). Genetic isolation of ADA2: A potential transcriptional adapter required for function of certain acidic activation domains. Cell 70,251-265. Berkowitz, L. A,, and Gilman, M. Z. (1990). ‘ h o distinct forms of active transcription factor CREB (CAMPresponse element binding protein). Proc. Nutl. Acad. Sci U.S.A. 87, 5258-5262.
CREB AND CREM REGULATION
47
Blendy, J., Hummler, E., Cole, T. J., Aguzzi, A,, Schmid, W., and Schutz, G. (1993). Targeted mutation of the cAMP response element binding protein CREB. 75th Annual Meeting of The Endocrine Society, Las Vegas, Nevada. Blenis, J. (1993). Signal transduction via the MAP kinases: Proceed a t your own RSK. Proc. Natl. Acad. Sci. U.S.A.90, 5889-5892. Borsook, D., Konradi, C., Falkowski, O., Comb, M., and Hyman, S. E. (1994). Molecular mechanisms of stress-induced proenkephalin gene regulation: CREB interacts with the proenkephalin gene in the mouse hypothalamus and is phosphorylated in response to hyperosmolar stress. Mol. Endocrinol. 8, 240-248. Bourtchuladze, R., Frenguelli, B., Blendy, J., Cioffi, D., Schutz, G., and Silva, A. J. (1994). Deficient long-term memory in mice with a targeted mutation of the CAMPresponsive element-binding protein. Cell 79, 59-68. Brindle, P., Linke, S., and Montminy, M. R. (1993).Protein-kinase-A-dependent activator in transcription factor CREB reveals new role for CREM repressors. Nature 364, 821-824. Buratowski, S. (1994). The basics of basal transcription by RNA polymerase 11. Cell 77, 1-3. Burgering, B. M. T., Pronk, G. J., van Weeren, P. C., Chardin, P., and Bos, J. L. (1993). cAMP antagonizes p2lras-directed activation of extracellular signal-regulated kinase 2 and phosphorylation of mSos nucleotide exchange factor. EMBO J . 12,42114220. Burton, F. H., Hasel, K. H., Bloom F. E., and Sutcliffe, J. G . (1991). Pituitary hyperplasia and gigantism in mice caused by a cholera toxin transgene. Nature 350, 74-77. Cao, Z., Umek, R. M., and McKnight, S. L. (1991). Regulated expression of three CIEBP isoforms during adipose conversion of 3T3-Ll cells. Genes Deu. 5, 1538-1552. Chen, B. P. C., Liang, G., Whelan, J., and Hai, T. (1994). ATF3 and ATF3DZip. Transcriptional repression versus activation by alternatively spliced isoforms. J. Biol. Chem. 269, 15819-15826. Chrivia, J. C., Kwok, R. P. S., Lamb, N., Hagiwara, M., Montminy, M., and Goodman, R. H. (1993). Phosphorylated CREB binds specifically to the nuclear protein CBP. Nature 365, 855-859. Chu, H. M., Tan, Y., Kiobierski, L. A., Balsam, L. B., and Comb, M. J. (1994).Activating transcription factor-3 stimulates 3’,5’-cyclic adenosine monophosphate-dependent gene expression. Mol. Endocrinol. 8, 59-68. Coghlan, V. M., Langeberg, L. K., Fernandez, A., Lamb, N. J . C., and Scott, J. D. (1994). Cloning and characterization of ADAP 95, a nuclear protein that associates with the regulatory subunit of type I1 CAMP-dependent protein kinase. J . Biol. Chem. 269, 7658-7665. Cohen, P. (1989). The structure and regulation of protein phosphatases. Aanu. Reu. Biochem. 58,453-508. Colbran, J. L., Roach, P. J., Fiol, C. J., Dixon, J. E., Andrisani, 0. M., and Corbin, J. D. (1992). CAMP-dependent protein kinase, but not the cGMP dependent enzyme, rapidly phosphorylates delta-CREB and a synthetic delta-CREB peptide. Biochem. Cell Biol. 70, 1277-1282. Cole, T. J., Copeland, N. G., Gilbert, D. J., Jenkins, N. A., Schutz, G., and Ruppert, S. (1992). The mouse CREB (CAMPresponsive element binding protein) gene: Structure, promoter analysis and chromosomal localization. Genomics 13, 974-982. Cook, S. J., and McCormick, F. (1993). Inhibition by cAMP of Ras-dependent activation of Raf. Science 262, 1069-1072. Courey, A. J., and Tjian, R. (1988).Analysis of S p l in vivo reveals multiple transcriptional domains, including a novel glutamine-rich activation motif. Cell 55, 887-898.
48
JOEL F. HABENER et al.
Dash, P. K., Hochner, B., and Kandel, E. R. (1990). Injection of the CAMP-responsive element into the nucleus of Aplysia sensory neurons blocks long-term facilitation. Nature 345, 718-721. Dash, P. K., Karl, K. A., Coplicos, M. A., Prywes, R., and Kandel, E. R. (1991). cAMP response element-binding protein is activated by Ca2+/calmodulin- a s well as CAMP-dependent protein kinase. Proc. Natl. Acad. Sci. U.S.A. 88, 5061-5065. DeGroot, R. P., and Sassone-Corsi, P. (1993). Hormonal control of gene expression: Multiplicity and versatility of cyclic adenosine 3',5'-monophosphate-responsivenuclear regulators. Mol. Endocrinol. 7, 145-153. DeGroot, R. P., Derua, R., Goris, J., and Sassone-Corsi, P. (1993a). Phosphorylation and negative regulation of the transcriptional activator CREM by p34cdc2. Mol. Endocrinol. 7, 1495-1501. DeGroot, R. P., den Hertog, J., Vandenheele, J. R., Goris, J., and Sassone-Corsi, P. (199313).Multiple and cooperative phosphorylation events regulate the CREM activator function. EMBO J . 12, 3903-3911. DeGroot, R. P., Ballou, L. M., and Sassone-Corsi, P. (1994). Positive regulation of the CAMP-responsive activator CREM by the p70 S6 kinase: An alternative route to mitogen-induced gene expression. Cell 79, 81-91. Delmas, V., Laoide, B. M., Masquilier, D., DeGroot, R. P., Foulkes, N. S., and SassoneCorsi, P. (1992). Alternative usage of initiation codons in mRNA encoding the CAMP-responsiveelement modulator generates regulators with opposite functions. Proc. Natl. Acad. Sci. U.S.A. 89, 4226-4230. Delmas, V., van der Hoorn, F., Mellstrom, B., Jegou, B., and Sassone-Corsi, P. (1993). Induction of CREM activator proteins in spermatids: Down-stream targets and implications for haploid germ cell differentiation. Mol. Endocrinol. 7, 1502-1514. Deutsch, P. J., Hoemer, J. P., Jameson, J. L., Lin, J. C., and Habener, J . F. (1988a). Structural determinants for transcriptional activation by CAMP-responsive DNA elements. J. Biol. Chem. 263, 18466-18472. Deutsch, P. J., HoeMer, J. P., Jameson, J . L., and Habener, J. F. (1988b).Cyclic AMP and phorbol ester-stimulated transcription mediated by similar DNA elements that bind distinct proteins. Proc. Natl. Acad. Sci. U.S.A. 85, 7922-7926. Diep, A,, Li, C., Klisak, I., Mohandas, T., Sparkes, R. S., Gaynor, R., and Lusis, A. J. (1991). Assignment of the gene for cyclic AMP-response element binding protein 2 (CREB2) to human chromosome 2q24.1-q32. Genomics 11,1161-1163. Dohlman,H. G.,Thorner, J.,Caron,M.G.,andLefkowitz,R. J. (1991).Modelsystemsforthe study ofseven-transmembrane-segment receptors.Annu.Rev. Biochem. 60,653-688. Dynlacht, B. D., Hoey, T., and Tjian, R. (1991). Isolation of coactivators associated with the TATA-bindingprotein that mediate transcriptional activation. Cell 66,563-576. Ellenberger, T. E., Brandl, C. J., Struhl, K., and Harrison, S. C. (1992).The GCN4 basic region leucine zipper binds DNA as a dimer of uninterrupted a-helices: Crystal structure of the protein-DNA complex. Cell 71, 1223-1237. Enslen, H., Sun, P., Brickey, D., Soderling, S. H., Klamo E., and Soderling, T. R. (1994). Characterization of Ca2+/calmodulin-dependentprotein kinase IV. Role in transcriptional regulation. J . Biol. Chem. 269, 15520-15527. Ferreri, K., Gill, G., and Montminy, M. R. (1994). The cAMP regulated transcription factor CREB interacts with a component of the TFIID complex. Proc. Natl. Acad. Sci. U.S.A. 91, 1210-1213. Fiol, C. J., Williams, J. S., Chou, C.-H., Wang, Q . M., Roach, P. J., and Andrisani, 0. M. (1994). Secondary phosphorylation of CREB341 a t Ser129 is required for CAMPmediated control of gene expression. J . Biol. Chem. 269, 32187-32193.
CREB AND CREM REGULATION
49
Foulkes, N. S., and Sassone-Corsi, P. (1992). More is better: Activators and repressors from the same gene. Cell 68,411-414. Foulkes, N. S., Laoide, B. M., Schlotter, F., and Sassone-Corsi, P. (1991a).Transcriptional antagonist CAMP-responsive element modulator (CREM) down-regulates c-fos CAMP-induces expression. Proc. Nutl. Acad. Sci. U.S.A. 88, 5448-5452. Foulkes, N. S., Borrelli, N. S., and Sassone-Corsi, P. (1991b).CREM gene: Use of alternative DNA-binding domains generates multiple antagonists of CAMP-induced transcription. Cell 64, 739-749. Foulkes, N. S., Mellstrom, B., Benusiglio, E., and Sassone-Corsi, P. (1992). Developmental switch of CREM function during spermatogenesis: From antagonist to activator. Nature 355, 80-84. Foulkes, N. S., Schlotter, F., Pevet, P., and Sassone-Corsi, P. (1993). Pituitary hormone FSH directs the CREM functional switch during spermatogenesis. Nature 362,264267. Frank, D. A,, and Greenberg, M. E. (1994).CREB: A mediator of long-term memory from mollusks to mammals. Cell 79, 5-8. Frankel, A. D., and Kim, P. S. (1991). Modular structure of transcription factors: Implications for gene regulation. Cell 65, 717-719. Fuller, P. J. (1991).The steroid receptor superfamily: Mechanisms of diversity. FASEB J . 5, 3092-3099. Gilman, A. G. (1987).G proteins: Transducers of receptor generated signals. Annu. Reu. Biochem. 56, 615-649. Ginty, D. D., Kornhauser, J. M., Thompson, M. A., Bading, H., Mayo, K. E., Takahashi, J. S., and Greenberg, M. E. (1993). Regulation of CREB phosphorylation in the suprachiasmatic nucleus by light and a circadian clock. Science 260, 238-241. Ginty, D. D., Bonni, A,, and Greenberg, M. E. (1994). Nerve growth factor activates a Ras-dependent protein kinase that stimulates c-fos transcription via phosphorylation of CREB. Cell 77, 713-725. Gonzalez, G. A,, and Montminy, M. R. (1989).Cyclic AMP stimulates somatostatin gene transcription by phosphorylation of CREB a t serine 133. Cell 59, 675-680. Gonzalez, G. A,, Yamamoto, K. K., Fischer, W. H., Karr, D., Menzel, P., Biggs, W., Vale, W. W., and Montminy, M. R. (1989).A cluster of phosphorylation sites on the cyclic AMP-regulated nuclear factor CREB predicted by its sequence. Nature 337, 747752. Gonzalez, G. A., Menzel, P., Leonard, J., Fischer, W. H., and Montminy, M. R. (1991). Characterization of motifs which are critical for activity of the cyclic AMPresponsive transcription factor CREB. Mol. Cell. Biol. 11, 1306-1312. Gosse, M. D., Walker, W. H., and Habener, J. F. (1995). Cyclic AMP activation of protein kinase-A and phosphorylation of nuclear transcription factor CREB in uiuo within one minute: Evidence for nuclear-located holoenzyme. Submitted. Graves, L. M., Bornfeldt, K. E., Raines, E. W., Potts, B. C., Macdonald, S. G., Ross, R., and Krebs, E. G. (1993). Protein kinase A antagonizes platelet-derived growth factorinduced signaling by mitogen-activated protein kinase in human arterial smooth muscle cells. Proc. Nutl. Acad. Sci. U.S.A. 90,10300-10304. Gronemeyer, H. (1992).Control of transcription activation by steroid hormone receptors. FASEB J . 6,2524-2529. Habener, J. F. (1990). Cyclic AMP response element binding proteins: A cornucopia of transcription factors. Mol. Endocrinol. 4, 1087-1094. Habener, J. F. (1993). Cyclic AMP second messenger signaling pathway. In “Endocrinology” (DeGroot RP, ed.), Ch. 5B, p. 77. Saunders, Philadelphia.
50
JOEL F. HABENER et a1
Hagiwara, M., Alberts, A., Brindle, P., Meinkoth, J., Feramisco, J., Deng, T., Karin, M., Shenolikar, S., and Montminy, M. (1992). Transcriptional attenuation following cAMP induction requires PP-1-mediated dephosphorylation of CREB. Cell 70, 105113. Hagiwara, M., Brindle, P., Harootunian, A,, Armstrong, R., Rivier, J.,Vale, W., Tsien, R., and Montminy, M. (1993). Coupling of hormonal stimulation and transcription via the cyclic AMP-responsive factor CREB is rate limited by nuclear entry of protein kinase A. Mol. Cell. Biol. 13, 4852-4859. Hai, T., and Curran, T. (1991). Cross-family dimerization of transcription factors FosiJun and ATFiCREB alters DNA binding specificity. Proc. Natl. Acad. Sci. U.S.A.88, 3720-3724. Hai, T., Liu, F., Coukos, W. J., and Green, M. R. (1989). Transcriptional factor ATF cDNA clones: An extensive family of leucine zipper proteins able to selectively form DNAbinding heterodimers. Genes Deu. 3, 2083-2090. Halazonetis, T. D., Georgopoulos, K., Greenberg, M. E., and Leder, P. (1988). c-Jun dimerizes with itself and with c-Fos, forming complexes of different DNA-binding affinities. Cell 55, 917-924. Herschman, H. R. (1989). Extracellular signals, transcriptional responses and cellular specificity. Trends Biochem. Sci. 14, 455-458. Hoeffler, J . P. (1992). Structureifunction relationships of CREBiATF proteins. J . Invest. Dermatol. 98, 21s-28s. Hoeffler, J. P., Meyer, T. E., Yung, Y., Jameson, J . L., and Habener, J. F. (1988). Cyclic AMP responsive DNA-binding protein: Structure based on a cloned placental cDNA. Science 242, 1430-1433. Hoeffler, J . P., Meyer, T. E., Waeber, G., and Habener, J . F. (1990). Multiple adenosine 3’,5’-monophosphate response element DNA-binding proteins generated by gene diversification and alternative exon splicing. Mol. Endocrinol. 4, 920-930. Hoey, T. R., Weinzierl, R. 0. J., Gill, G., Chen, J. L., Dynlacht, B. D., and Tjian, R. (1993). Molecular cloning and functional analysis of Drosophila TAFllO reveals properties expected of coactivators. Cell 72, 247-260. Hsu, W., Kerppola, T. K., Chen, P.-L., Curran, T., and Chen-Kiang, S. (1994). Fos and J u n repress transcription activation by NF-IL6 through association at the basic zipper region. Mol. Cell. Biol. 14, 268-276. Huang, Y.-Y., Li, X.-C., and Kandel, E. R. (1994). cAMP contributes to mossy fiber LTP by initiating both a covalently mediated early phase and macromolecular synthesisdependent late phase. Cell 79,69-79. Hummler, E., Cole, T. J., Blendy, J . A,, Ganss, R., Aguzzi, A., Schmid, W., Beermann, F., and Schutz, G. (1994). Targeted mutation of the CREB gene: Compensation within the CREB/ATF family of transcription factors. Proc. Natl. Acad. Sci. U.S.A. 91, 5647-5651. Hunter, T., and Karin, M. (1992). The regulation of transcription by phosphorylation. Cell 70, 375-387. Ikuyama, S.,Shimura, H., Hoeffler, J . P., and Kohn, L. D. (1992). Role of cyclic adenosine 3‘,5’-monophosphateresponse element in efficient expression of the rat thyrotropin receptor promoter. Mol. Endocrinol. 6, 1701-1715. Ingraham, H. A,, Chen, R., Mangalam, H. J., Elsholtz, H. P., Flynn, S. E., Lin, C. R., Simmonds, D. M., Swanson, L., and Rosenfeld, M. G. (1988). A tissue-specific transcription factor containing a homeodomain specifies a pituitary phenotype. Cell 55, 519-529.
CREB AND CREM REGULATION
51
Joachim, S., and Schwoch, G. (1988). Immunoelectron microscopic localization of catalytic and regulatory subunits of CAMP-dependent protein kinases in the parotid gland. Eur. J. Cell Biol. 46, 491-498. Johnson, P. F., and McKnight, S. L. (1989). Eukaryotic transcriptional regulatory proteins. Annu. Rev. Biochem. 58, 799-839. Karin, M. (1992). Signal transduction from cell surface to nucleus in development and disease. FASEB J . 6 , 2581-2590. Karpinski, B. A., Morle, G. D., Huggenvik, J., Uhler, M. D., and Leiden, J. M. (1992). Molecular cloning of human CREB-2: An ATF/CREB transcription factor that can negatively regulate transcription from the cAMP response element. Proc. Natl. Acad. Sci. U.S.A. 89, 4820-4824. Katsushika, S., Chen, L., Kawabe, J., Nilakantan, R., Halnon, N. F., Homcy, C. J., and Ishikawa, Y. (1992). Cloning and characterization of a sixth adenylyl cyclase isoform: Types V and VI constitute a subgroup within the mammalian adenylyl cyclase family. Proc. Natl. Acad. Sci.U.S.A. 89, 8774-8778. Kim, J., Tzamarias, D., Ellenberger, T., Harrison, S. C., and Struhl, K. (1993). Adaptability at the protein-DNA interface is a n important aspect of sequence recognition by bZIP proteins. Proc. Natl. Acad. Sci. U.S.A. 90, 4513-4517. Knepel, W., Chafitz, J. A., and Habener, J. F. (1990). Transcriptional activation of the rat glucagon gene by the cyclic AMP-responsive element in pancreatic islet cells. Mol. Cell. Biol. 10, 6799-6804. Knighton, D. R., Zheng, J., Eyck, L. F. T., Ashford, V. A., Xuong, N. H., Taylor, S., and Sowadski, J. M. (1991). Crystal structure of the catalytic subunit of cyclic adenosine monophosphate-dependent protein kinase. Science 253,407-414. Kobierski, L. A., Chu, H. M., Tan, Y., and Comb, M. J. (1991). CAMP-dependent regulation of proenkephalin by JunD and JunB: Positive and negative effects of AP-1 proteins. Proc. Natl. Acad. Sci. U.S.A. 88, 10222-10226. Konradi, C., Cole, R. L., Heckers, S., and Hyman, S. E. (1994). Amphetamine regulates gene expression in rat striatum via transcription factor CREB. J . Neurosci. 14, 5623-5634. Krajewski, W., and Lee, K. A. W. (1994). A monomeric derivative of the cellular transcription factor CREB functions as a constitutive activator. Mol. Cell. Biol. 14, 7204-7210. Kramer, I. M., Koornneef, I., Cole, R. L., Heckers, S., and Hyman, S. E. (1991). TGF-p1 induces phosphorylation of the cyclic AMP responsive element binding protein in ML-CC164 cells. EMBO J. 10, 1083-1089. Krebs, E. G. (1989). Role of the cyclic AMP-dependent protein kinase in signal transduction. JAMA, J. Am. Med. Assoc. 262, 1815-1818. Krupinski, J., Lehman, R., Frankenfield, C. D., Zwaagstra, J. C., and Watson, P. A. (1992). Molecular diversity in the adenylylcyclase family. Evidence for eight forms of the enzyme and cloning of type VI. J . Biol. Chem. 267, 24858-24862. Kwok, R. P. S., Lundblad, J . P., Chrivia, J . C., Richards, J. P., Bachinger, P., Brennan, R. G., Roberts, S. G. E., Green, M. R., and Goodman, R. H. (1994). Nuclear protein CBP is a coactivator for the transcription factor CREB. Nature 370, 223-226. Lalli, E., and Sassone-Corsi, P. (1994). Signal transduction and gene regulation: The nuclear response to CAMP. J. Biol. Chem. 269, 17359-17362. Lamph, W. W., Dwarki, V. J., Ofir, R., Montminy, M., and Verma, I. M. (1990). Negative and positive regulation by transcription factor cAMP response element-binding protein is modulated by phosphorylation. Proc. Natl. Acad. Sci. U.S.A. 87,4320-4324.
52
JOEL F. HABENER et al.
Landis, C. A., Masters, S. B., Spada, A., Pace, A. M., Bourne, H. R., and Vallar, L. (1989). GTPase inhibiting mutations activate the a chain of Gs and stimulate adenylyl cyclase in human pituitary tumors. Nature 340, 692-696. Johnson, P. F., and McKnight, S. L. (1988). The leucine zipper: A Landschulz, W. H., hypothetical structure common to a new class of DNA binding proteins. Science 240, 1759-1764. Laoide, B. M., Foulkes, N. S., Schlotter, F., and Sassone-Corsi, P. (1993). The functional versatility of CREM is determined by its modular structure. EMBO J . 12, 11791191. Lee, K. A. W., and Masson, N. (1993). Transcriptional regulation by CREB and its relatives. Biochim. Biophys. Acta 1174, 221-233. Leonard, J., Serup, P., Gonzalez, G., Edlund, T., and Montminy, M. (1992). The LIM family transcription factor Isl-1 requires cAMP response element binding protein to promote somatostatin expression in pancreatic islet cells. Proc. Natl. Acad. Sci. U.S.A. 89, 6247-6251. Liu, F., Thompson, M. A., Wagner, S., Greenberg, M. E., and Green, M. R. (1993). Activating transcription factor-1 can mediate Caz+= and CAMP-inducible transcriptional activation. J . Biol. Chern. 268, 6714-6720. Loriaux, M. M., Rehfuss, R. P., Brennan, R. G., and Goodman, R. H. (1993). Engineered leucine zippers show that hemiphosphorylated CREB complexes are transcriptionally active. Proc. Natl. Acad. Sci. U.S.A. 90, 9046-9050. Loriaux, M. M., Brennan, R. G., and Goodman, R. H. (1994). Modulatory function of CREB-CREMa heterodimers depends upon CREMu phosphorylation. J . Biol. Chem. 269,28839-28843. Macgregor, P. F., Abate, C., and Curran, T. (1990). Direct cloning of leucine zipper proteins: J u n binds cooperatively to the CRE and CRE-BP1. Oncogene 5,451-458. Maniatis, T., Goodbourn, S., and Fischer, J . A. (1987). Regulation of inducible and tissuespecific gene expression. Science 236, 1237-1244. Masquilier, D., and Sassone-Corsi, P. (1992). Transcriptional cross-talk: Nuclear factors CREM and CREB bind to AP-1 sites and inhibit activation by jun. J. Biol. Chern. 267,22460-22466. Matthews, R. P., Guthrie, C. R., Wailes, L. M., Zhao, X., Means, A. R., and McKnight, G. S. (1994). Calciumicalmodulin-dependent protein kinase types I1 and IV differentially regulate CREB-dependent gene expression. Mol. Cell. Biol. 14, 6107-6116. McCormick, A,, Brady, H., Theill, L. E., and Karin, M. (1990). Regulation of pituitaryspecific homeobox gene GHFl by cell-autonomous and environmental cues. Nature 345,829-832. Meinkoth, J . L., Ji, Y., Taylor, S. S., and Feramisco, J. R. (1990). Dynamics of the distribution of cyclic AMP-dependent protein kinase in living cells. Proc. Natl. Acad. Sci. U.S.A. 87, 9595-9599. Mellstrom, B., Naranjo, J. R., Foulkes, N. S., Lafarga, M., and Sassone-Corsi, P. (1993). Transcriptional response to cAMP in brain: Specific distribution and induction of CREM antagonists. Neuron 10, 655-665. Meyer, T. E., and Habener, J. F. (1992). Cyclic AMP response element binding protein CREB and modulator protein CREM are products of distinct genes. Nucleic Acids Res. 20, 6106. Meyer, T.E., and Habener, J. F. (1993). Cyclic adenosine 3’,5’-monophosphate response element binding protein (CREB) and related transcription-activating deoxyribonucleic acid binding proteins. Endocr. Rev. 14, 269-290. Meyer, T. E., Waeber, G., Lin, J., Beckmann, W., and Habener, J . F. (1993).The promoter
CREB AND CREM REGULATION
53
of the gene encoding 3’,5’-cyclic adenosine monophosphate (CAMP)response element binding protein contains cAMP respnose elements: Evidence for positive autoregulation of gene transcription. Endocrinology 132, 770-780. Miller, C. P., Lin, J. C., and Habener, J. F. (1993). Transcription of the rat glucagon gene by cyclic AMP response element binding protein CREB is modulated by adjacent CREB-associated proteins. Mol. Cell. Biol. 13, 7080-7090. Mitchell, P. J., and Tjian, R. (1989). Transcriptinal regulation in mammalian cells by sequence-specific DNA binding proteins. Science 245, 371-378. Molina, C. A,, Foulkes, N. S., Lalli, E., and Sassone-Corsi, P. (1993). Inducibility and negative regulation of CREM: An alternative promoter directs the expression of ICER, a n early response repressor. Cell 75, 875-886. Montminy, M. R., and Bilezikjian, L. M. (1987). Binding of a nuclear protein to the cyclicAMP response element of the somatostatin gene. Nature 328, 175-178. Murakami, N., and Takahashi, K. (1983). Circadian rhythm of adenosine 3‘,5‘monophosphate content in suprachiasmatic nucleus (SCN) and ventromedial hypothalamus (VMH) in the rat. Bruin Res. 276,297-304. Muro, A. F., Bernath, V. A., and Kornblihtt, A. R. (1992). Interaction of the -170 cyclic AMP response element with the adjacent CCAAT box in the human fibronectin gene promoter. J . Biol. Chem. 267, 12767-12774. Neuberg, M., Schuermann, M., and Muller, R. (1991). Mutagenesis of the DNA contact site in Fos protein: Compatibility with the scissors grip model and requirement for transformation. Oncogene 6, 1325-1333. Nichols, M., Weih, F., Schmid, W., DeVack, C., Kowenz-Leutz, E., Luckow, B., Boshart, M., and Schutz, G. (1992). Phosphorylation of CREB affects its binding to high and low affinity sites: Implciations for cAMP induced gene transcription. EMBO J . 11, 3337-3346. Nigg, E. A,, Hilz, H., Eppenberger, H. M., and Dutly, F. (1985a). Rapid and reversible translocation of the catalytic subunit of CAMP-dependent protein kinase type I1 from the Golgi complex to the nucleus. EMBO J . 4, 2801-2806. Nigg, E. A., Schafer, G., Hilz, H., and Eppenberger, H. M. (1985b). Cyclic-AMPdependent protein kinase type I1 is associated with the Golgi complex and with centrosomes. Cell 41, 1039-1051. Ofir, R., Dwarki, V. J.,Rashid, D., and Verma, I. M. (1991). CREB represses transcription of fos promoter: Role of phosphorylation. Gene Expression 1, 55-60. O’Malley, B. (1990). The steroid receptor superfamily: More excitement predicted for the future. Mol. Endocrinol. 4, 363-369. O’Neil, K. T., Hoess, R. H., and DeGrado, W. F. (1990). Design of DNA-binding peptides based on the leucine zipper motif. Science 249, 774-778. O’Shea, E. K., Rutkowski, R., and Kim, P. S. (1989). Evidence that the leucine zipper is a coiled coil. Science 243, 538-542. O’Shea, E. K., Klemm, J. D., Kim, P. S., and Alber, T. (1991). X-ray structure of the GCN4 leucine zipper, a two-stranded, parallel coiled coil. Science 254, 539-544. Park, E. A., Roesler, W. J., Liu, J., Klemm, D. J., Gurney, A. L., Thatcher, J. D., Shuman, J., Friedman, A., and Hanson, R. W. (1990). The role of the CCAATienhancerbinding protein in the transcriptional regulation of the gene for phosphoenolpyruvate carboxykinase (GTP). Mol. Cell. Biol. 10, 6264-6272. Park, E. A., Gurney, A. L., Nizielski, S. E., Hakimi, P., Cao, Z., Moorman, A,, and Hanson, R. W. (1993). Relative roles of CCAAT/enhancer-binding protein 0 and cAMP regulatory element-binding protein in controlling transcription of the gene for phosphoenolpyruvate carboxykinase (GTP). J. Biol. Chem. 268, 613-619.
54
JOEL F. HABENER et a1
Patel, L., Abate, C., and Curran, T. (1990). Altered protein conformation on DNA binding by Fos and Jun. Nature 347, 572-575. Peunova, N., and Enikolopov, G. (1993). Amplification of calcium-induced gene transcription by nitric oxide in neuronal cells. Nature 364, 450-453. Philippe, J., and Missotten, M. (1990).Functional characterization of a CAMP-responsive element of the rat insulin I gene. J . Biol. Chern. 265, 1465-1469. Philippe, J., Chick, W. L., and Habener, J. F. (1987). Multipotential phenotypic expression of genes encoding peptide hormones in rat insulinoma cell lines. J. Clin. Inuest. 79,351-358. Powers, A. C . , Tedeschi, F., Wright, K. E., Chan, J. S., and Habener, J . F. (1989). Somatostatin gene expression in pancreatic islet cells is directed by cell-specific DNA control elements and DNA-binding proteins. J . Biol. Chern. 264, 10048-10056. Prosser, R. A., and Gillette, M. U. (1989). The mammalian circadian clock in the suprachiasmatic nuclei is reset in vitro by CAMP.J . Neurosci. 9, 1073-1081. Ptashne, M. (1988). How eukaryotic transcriptional activators work. Nature 335, 683689. Quinn, P. G. (1993). Distinct activation domains within CAMPrespnose element-binding protein (CREB) mediate basal and CAMP-stimulated transcription. J . Biol. Chem. 268, 16999-17009. Quinn, W. G., and Tully, T. (1994). Induction of a dominant negative CREB transgene specifically blocks long-term memory in Drosophila. Cell 79, 49-58. Rabbits, T. H. (1994). Chromosomal translocations in human cancer. Nature 372, 143149. Fkhfuss, R. P., Walton, K. M., Loriaux, M. M., and Goodman, R. H. (1991). The CAMPregulated enhancer binding protein ATF-1 activates transcription in response to CAMP-dependent protein kinase A. J . Biol. Chem. 266, 18431-18434. Roach, P. J. (1991).Multiple and hierarchal protein phosphorylation. J . Biol. Chern. 266, 14139-14142. Roesler, W. J., Vandebark, G. R., and Hanson, R. W. (1988). Cyclic AMP and the induction of eukaryotic gene expression. J. Biol. Chem. 263,9063-9066. Ron, D., and Habener, J. F. (1992). CHOP, a novel developmentally regulated nuclear protein that dimerizes with transcription factors C/EBP and LAP and functions a s a dominant-negative inhibitor of gene transcription. Genes Dev. 6, 439-453. Rosen, L. B., Ginty, D. D., Weber, M. J., and Greenberg, M. E. (1994).Membrane depolarization and calcium influx stimulate MEK and MAP kinase via activation of Ras. Neuron 12,1207-1221. Ruppert, S., Cole, T. J., Boshart, M., Schmid, E., and Schutz, G. (1992). Multiple mRNA isoforms of the transcription activator protein CREB: Generation by alternative splicing and specific expression in primary spermatocytes. EMBO J. 11, 15031512. Ryseck, R. P., and Bravo, R. (1991). c-JUN, J U N B, and JUN D differ in their binding affinities to AP-1 and CRE consensus sequences: Effect of FOS proteins. Oncogene 6,533-542. Sassone-Corsi, P. (1994). Goals for signal transduction pathways: Linking up with transcriptional regulation. EMBO J . 13, 4717-4728. Schuermann, M., Hunter, J. B., Hennig, G., and Muller, R. (1991). Non-leucine residues in the leucine repeats of Fos and Jun contribute to the stability and determine the specificity of dimerization. Nucleic Acids Res. 19, 739-746. Schwaninger, M., Blume, R., Oetjen, E., and Knepel, W. (1993a). The immunosuppressive drugs cyclosporin A and FK506 inhibit calcineurin phosphatase activity and
CREB AND CREM REGULATION
55
gene transcription mediated through the CAMP-responsive element in a nonimmune cell line. Naunyn-Schmiedebergs Arch. Pharmacol. 348, 541-545. Schwaninger, M., Lux, G., Blume, R., Oetjen, E., Hidaka, H., and Knepel, W. (199313). Membrane depolarization and calcium influx induce glucagon gene transcription in pancreatic islet cells through the cyclic AMP-responsive element. J. Biol. Chem. 268,5168-5177. Schwoch, G., and Freimann, A. (1986). Quantitative changes in the catalytic and regulatory subunits of nuclear CAMP-dependent protein kinases type I and type I1 during isoproterenol-induced growth of the rat parotid gland. FEBS Lett. 197, 143. Scott, J. D. (1991). Cyclic nucleotide-dependent protein kinases. Pharmacacol. Ther. 50, 123-145. Scott, J. D., and McCartney, S. (1994). Localization of A-kinase through anchoring proteins. Mol. Endocrinol. 8, 5-11. Sculptoreanu, A., Scheuer, T., and Catterall, W. A. (1993). Voltage-dependent potentiation of L-type Ca2+ channels due to phosphorylation by CAMP-dependent protein kinase. Nature 364, 240-243. Sevetson, B. R., Kong, X., and Lawrence, J. C. (1993). Increasing CAMP attenuates activation of mitogen-activated protein kinase. Proc. Natl. Acad. Sci. U.S.A. 90, 10305-10309. Sheng, M., McFadden, G., and Greenberg, E. (1990). Membrane depolarization and calcium induce c-fos transcription via phosphorylation of transcription factor CREB. Neuron 4,571-572. Sheng, M., Thompson, M. A., and Greenberg, M. E. (1991). CREB: A Ca2+-regulated transcription factor phosphorylated by calmodulin-dependent kinases. Science 252, 1427-1430. Simon, M. I., Strathmann, M. P., and Gautam, N. (1991). Diversity of G proteins in signal transduction. Science 252, 802-808. Simons, S. S., Oshima, H., and Szapary, D. (1992). Higher levels of control: Modulation of steroid hormone-regulated gene transcription. Mol. Endocrinol. 6, 995-1002. Smeal, T., Angel, P., Meek, J., and Karin, M. (1989). Different requirements for formation of Jun:Jun and Jun:Fos complexes. Genes Deu. 3, 2091-2100. Squinto, S. P., and Jungmann, R. A. (1989). Modulation of nuclear cyclic AMPdependent protein kinase in dibutyryl cyclic AMP-treated rat H41IE hepatoma cells. Biochem. J. 260,673-682. Stehle, J. H., Foulkes, N. S., Molina, C. A., Simonneaux, V., Pevet, P., and Sassone-Corsi, P. (1993). Adrenergic signals direct rhythmic expression of transcriptional repressor CREM in the pineal gland. Nature 365, 314-320. Stevens, C. F. (1994). CREB and memory consolidation. Neuron 13, 769-770. Struthers, R. S., Vale, W. W., Arias, C., Sawchenko, P. E., and Montminy, M. R. (1991). Somatotroph hypoplasia and dwarfism in transgenic mice expressing a nonphosphorylatable CREB mutant. Nature 350, 622-624. Sum, P., Enslen, H., Myung, P. S., and Maurer, R. A. (1994). Differential activation of protein kinases type I1 and type IV involves CREB by CaZ+/calmodulin-dependent phosphorylation of a site that negatively regulates activity. Genes Deu. 8, 25272539. Sutherland, E. W. (1972). Studies on the mechanisms of hormone action. Science 177, 401-408. Taylor, A. K., Klisak, I., Mohondas, T., Sparks, R. S., Li, C., Gaymor, R., and Suis, A. J. (1990). Assignment of the human gene for CREB 1 to chromosome 2q32.3-q34. Genomics 7,416-421.
56
JOEL F. HABENER et al.
Ullrich, A., and Schlessinger, J. (1990). Signal transduction by receptors with tyrosine kinase activity. Cell 61, 203-212. Vallejo, M. (1994). Transcriptional control of gene expression by cAMP response element binding proteins. J . Neuroendocrinol. 6, 587-596. Vallejo, M., and Habener, J. F. (1994). Mechansims of transcriptional regulation by CAMP.In “Transcription: Mechansism and Regulation” (R. Conaway and J . Conaway, eds.), pp. 353-368. Raven, New York. Vallejo, M., Penchuk, L., and Habener, J . F. (1992). Somatostatin gene upstream enhancer element activated by a protein complex consisting of CREB, Isl-1-like, and a-CBF-like transcription factors. J . Biol. Chem. 267, 12876-12884. Vallejo, M., Ron, D., Miller, C. P., and Habener, J. F. (1993). CIATF, a member of the activating transcription factor family of DNA-binding proteins, dimerizes with CAAT/enhancer-binding proteins and directs their binding to cAMP response elements. Proc. Natl. Acad. Sci. U.S.A. 90, 4679-4683. Vallejo, M., Gosse, M. E., Beckman, W., and Habener, J. F. (1995). Impaired cyclic AMPdependent phsophorylation renders CREB a repressor of C/EBP-induced transcription of the somatostatin gene in a n insulinoma cell line. Mol. Cell. Biol. 15, 415424. Vinson, C. R., Sigler, P. B., and McKnight, S. L. (1989). Scissors grip model for DNA recognition by a family of leucine zipper proteins. Science 246, 911-922. Wadzinski, B. E., Wheat, W. H., Jaspers, S., Peruski, L. F., Lickteig, R. L., Johnson, G. L., and Klemm, D. J . (1993). Nuclear protein phosphatase 2A dephosphorylates protein kinase-A phosphorylated CREB and regulates CREB transcriptional stimulation. Mol. Cell. Biol. 13, 2822-2834. Waeber, G., and Habener, J . F. (1991). Nuclear translocation and DNA recognition signals colocalized within the bZIP domain of cyclic adenosine 3’,5’-monophosphate response element-binding protein CREB. Mol. Endocrinol. 5, 1431-1438. Waeber, G., and Habener, J. F. (1992). Novel testis germ cell-specific transcript of the CREB gene contains an alternatively spliced exon with multiple in-frame stop codons. Endocrinology 131,2010-2015. Waeber, G., Meyer, T. E., LeSieur, M., Hermann, H. L., Gerard, N., and Habener, J. F. (1991). Developmental stage-specific expression of the cyclic AMP response element binding protein CREB during spermatogenesis involves alternative exon splicing. Mol. Endocrinol. 5, 1418-1430. Walker, W. H., Sanborn, B. M., and Habener, J. F. (1994). An isoform of transcription factor CREM expressed during spermatogenesis lacks the phosphorylation domain and represses CAMP-induced transcription. Proc. Natl. Acad. Sci. U.S.A.91,1242312427. Walker, W. H.,Fucci, L., and Habener, J. F. (1995). Expression of the gene encoding transcritpion factor CREB: Regulation by FSH-induced cAMP signaling in primary rat Sertoli cells. Endocrinology. Weiss, M. A., Ellenberger, T., Wobbe, C. R., Lee, J. P., Harrison, S. C., and Struhl, K. (1990). Folding transition in the DNA-binding domain of GCN4 on specific binding to DNA. Nature 347, 575-578. Wheat, W. H., Fbesler, W. J., and Klemm, D. J . (1994). Simian virus 40 small tumor antigen inhibits dephosphorylation of protein kinase A-phosphorylated CREB and regulates transcriptional stimulation. Mol. Cell. Biol. 14, 5881-5890. Williams, S.C., Cantwell, C. A., and Johnson, P. F. (1991). A family of CiEBP-related proteins capable of forming covalently linked leucine zipper dimers in vitro. Genes Deu. 5, 1553-1567.
CREB AND CREM REGULATION
57
Williams, T., Admon, A., Luscher, B., and Tjian, R. (1988). Cloning and expression of AP-2, a cell-type-specific transcription factor that activates inducible enhancer elements. Genes Deu. 2, 1557-1569. Wu, J., Dent, P., Jelinek, T., Wolfman, A,, Weber, M. J., and Sturgill, T. W. (1993). Inhibition of the EGF-activated MAP kinase signaling pathway by adenosine 3’,5’monophosphate. Science 262, 1065-1068. Xing, L., and Quinn, P. G. (1994). Three distinct regions within the constitutive activation domain of CAMP regulatory element-binding protein (CREB) are required for transcription activation. J. Biol. Chem. 269, 28732-28736. Yamamoto, K. K., Gonzalez, G. A., Briggs, W. H., 111, and Montminy, M. R. (1988). Phosphorylation-induced binding and transcriptional efficacy of transcription factor CREB. Nature 334,494-498. Yamamoto, K. K., Gonzalez, G. A,, Menzel, P., Rivier, J., and Montminy, M. R. (1990). Characterization of a bipartite activator domain in transcription factor CREB. Cell 60,611-617. Yarden, Y., and Ullrich, A. (1988). Growth factor receptor tyrosine kinases. Annu. Reu. Biochem. 57,443-478. Yin, J. C. P., Wallach, J. S., Del Vecchio, M., Wilder, E. L., and Zhou, H. (1994). Induction of a dominant negative CREB transgene specifically blocks long-term memory in Drosophila. Cell 79, 49-58. Yoshimura, M., and Cooper, D. M. (1992). Cloning and expression of a calciuminhibitable adenylyl cyclase from NCB-20 cells. Proc. Nutl. Acad. Sci. U.S.A. 89, 6716-6720. Yun, Y., Dumoulin, M., and Habener, J. F. (1990). DNA-binding and dimerization domains of cyclic AMP-responsive protein CREB reside in the carboxyl-terminal 66 amino acids. Mol. Endocrinol. 4, 931-939. Zucman, J., Delattre, O., Desmaze, C., Epstein, A. L., Stenman, G., Speleman, F., Fletchers, C. D. M., Aurias, A., and Thomas, G. (1993). EWS and ATF-1 gene fusion induced by t( 12;22) translocation in malignant melanoma of soft parts. Nature 4, 341-345.
VITAMINS AND HORMONES, VOL. 51
Multiple Facets of the Modulation of Growth by cAMP PIERRE P. ROGER, SYLVIA REUSE, CARINE MAENHAUT, AND JACQUES E. DUMONT Instituie of Interdisciplinary Research Campus Erasme Free University of Brussels B-1070 Brussels, Belgium
I. Introduction A. General Considerations on Cell Cycle Controls B. Probes of the cAMP System: Pharmacological and Genetic Tools 11. Negative Control of Cell Cycle Progression by cAMP A. Early Work on Continuous Cell Lines B. Normal Cells C. Conclusions 111. Positive Control of Cell Cycle Progression by cAMP A. Recent Examples of CAMP-Mediated Positive Growth Control B. Synergism between cAMP and Other Mitogenic Factors C. Positive Regulation of Cell Cycle Progression by cAMP D. Biochemistry of Positive Control of Cell Cycle Progression by cAMP E. CAMP-Dependent and -Independent Mitogenic Pathways IV. Relationship between Growth and Differentiation Controls by cAMP V. A Role for Cytoskeleton Changes in Control of Growth by CAMP? VI. cAMP and the Growth of Cancer Cells A. Negative Modulation B. Escape from Negative Modulation C. cAMP as a Tumor Promoter D. Oncogenes Related to the cAMP Signaling Cascade VII. Conclusions and Perspectives References
I. INTRODUCTION Recent successes in elucidating the functions of cell-transforming proteins encoded by various oncogenes and the demonstration of antioncogenes have shed new light on the assumption that cancers mostly develop from alterations of growth control mechanisms normally involved in homeostasis, tissue repair, and development (Hunter, 1991). Delineating processes that control proliferation and differentiation of normal cells and elucidating how these mechanisms are modified or subverted in neoplasia are critical to our understanding of carcinogenesis. 59
Copyright 0 1995 by Academic Press, Inc. All rights of reproduction in any form reserved.
60
PIERRE P. ROGER et al.
A great part of our knowledge of growth control mechanisms is derived from the study of in vitro conventional models of cultured fibroblast-like cells, such as 3T3 cell lines (Baserga, 1985; Pardee, 1989; Rozengurt, 1986). These cells provide a well-defined and reproducible experimental material, although their immortality indicates that their regulatory circuits have been genetically altered. The proliferation of such cells is dependent on various external factors, including stimulatory or inhibitory growth factors, hormones, and neurotransmitters, and spatial restriction by cell-to-cellcontacts. Two groups of intracellular pathways have been shown t o relay the mitogenic signal brought by some of the primary stimuli. Membrane receptors for some growth factors (e.g., EGF,l PDGF, FGF, IGF-I, insulin and CSF-1) possess an intrinsic protein tyrosine kinase activity, as is the case for one class of transforming proteins encoded by oncogenes (Hunter, 1989; Cantley et al., 1991). Other membrane receptors are coupled via a GTP-binding protein to a phospholipase Cp that cleaves PIP, into diacylglycerol and IP, (Rozengurt, 1986; Berridge, 1987; Whitman and Cantley, 1988). Binding of some growth factors to their tyrosine kinase receptors also activates a phospholipase Cy by phosphorylation on a tyrosine residue. Diacylglycerol activates serinehhreonine PKCs and IP, mobilizes calcium from intracellular stores. The analogs of diacylglycerol, the tumor promoter phorbol esters, also activate C kinases and, in some cell types, proliferation. Intracellular Ca2+ has a well-accepted but poorly defined role in mitogenesis (Berridge, 1987; 'Abbreviations: ADF, actin depolymerizing factor; ACTH, adrenocorticotropic hormone; AKAP, A kinase-anchoring protein; AP, activator protein; ATF, activating transcription factors; bHLH-zip, basic region with a helix-loop-helix leucine zipper motif; C, catalytic subunit of PKA; CBP, CREB binding protein; CDK, cyclin-dependent kinase; CHO, Chinese hamster ovary; W P T , [8-(4-chlorophenyl)thio;CRE, CAMP-responsive element; CREB, CRE-binding protein; CREM, CRE modulator; CSF, colony-stimulating factor; EGF, epidermal growth factor; FGF, fibroblast growth factor; FSH, folliclestimulating hormone; GH, growth hormone; GHF-1, growth hormone transcription factor 1; GIP, gastric inhibitory protein; GRF, GH-releasing factor; hCG, human chorionic gonadotropin; HIV, human immunodeficiency virus; IGF-I, insulin-like growth factor I; IL, interleukin; IP,, inositol 1,4,5-triphosphate; LH, luteinizing hormone; LRF-7, liver regeneration factor-7; MAP, mitogen-activated protein; MDCK, Madin-Darby canine kidney; MPF, maturation-promoting factor; MSH, melanocyte-stimulating hormone; NGF, nerve growth factor; ODC, ornithine decarboxylase; PCNA, proliferating-cell nuclear antigen; PDGF, platelet-derived growth factor; PGE, prostaglandin E; PI, phosphatidylinositol; PIP,, PI 4,5-biphosphate; PKA, CAMP-dependentprotein kinase, or protein kinase A; PKC, protein kinase C; PTH, parathyroid hormone; R, regulatory subunit of PKA; RB, retinoblastoma protein; RSK, ribosomal S6 kinase; SRE, serum-responsive element; TGF-P, transforming growth factor P; TPA, 12-0-tetradecanoylphorbol-13acetate; TRE, TPA-responsive element; TSH, thyroid-stimulating hormone; VIP, vasointestinal peptide.
GROWTH MODULATION BY cAMP
61
Whitfield et al., 19871, but in keratinocytes Ca2+ instead induces differentiation (Pillai et al., 1990).Activation of the PIP, cascade is by no means a universal way to induce proliferation (Choi et al., 1990; Dean and Boynton, 1990; Margolis et al., 1990; Rasp6 et al., 1992). Even though they are clearly distinct in their initial part (receptor, transducer, and first intracellular signal) (Berridge, 1987; Chambard et al., 19871, the protein tyrosine kinase pathways and the PIP, cascade rapidly converge on several events, such as activation of Na+/H+exchange and several transporters, and on a signaling cascade that involves in sequence, from cell membrane to nucleus, the activation of c-Ha-Ras, c-raf, MAP kinase kinase, and MAP kinase, then p62TCF and the transcription of protooncogenes c-jun, c-fos, and c-myc, the products of which are also modulated by phosphorylation (for reviews see McCormick, 1993; Crews and Erikson, 1993; Davis, 1993; Muller et al., 1993) (Fig. 1). These early events are assumed t o be necessary for the entry of a quiescent cell into the cell cycle, and are therefore often called “early mitogenic events.” However, causative relationships with late commitment to DNA replication and cell division remain unclear. After years of studying “early mitogenic events,” we discover that these events are not followed by DNA synthesis and trigger another program in many cells (Raspe et al., 1992; Seuwen et al., 1990a). cAMP is the first identified intracellular second messenger of hormone action. In the 1970s and early 1980s the hypothesis that it may govern (mostly negatively) cell proliferation was considered fascinating and prompted intense scientific activity. As generalized by Pastan et al. (19751, “It seems reasonably well established that treating cells with cyclic AMP analogs inhibits growth. . . . The observation that cyclic AMP inhibits the growth of many types of cells in uitro has obvious implications for the chemotherapy of cancer.” Parallel research led t o the opposite conclusion, which was also generalized; “After having examined most of the large amount of evidence that has accumulated during the past 20 years, we conclude that cAMP is programmed to stimulate an unknown event or events leading to DNA synthesis in a wide variety of cells and possibly a later event(s) leading to mitosis and division” (Boynton and Whitfield, 1983). These and other attempts to generalize the role of intracellular signals (e.g., Berridge, 1975) led to fruitless arguments and confusion. They were soon confronted by the failure to demonstrate such a universal role, raising disinterest in the whole field: “Because they constitute an excellent illustration of the problem of premature generalizations in assigning growth control functions to intracellular components. . . , we can now say that while cyclic nucleotides play a very important role in a cell’s life processes, they have little t o do with cell
62
PIERRE P. ROGER et al.
I+ Ca+ t
I+
MAPKK
:+
FIG. 1. Model of the transfer of CAMP-independent mitogenic signals from cell membrane to nucleus (McCormick, 1993; Crews and Erickson, 1993; Davis, 1993; Muller et al., 1993). Growth factors such as EGF and PDGF bind to their tyrosine kinase (TK) membrane receptors, which results in receptor dimerization and trans autophosphorylation on tyrosine residues. Receptor tyrosine autophosphorylation allows binding to other proteins through SH (src homology) 2 domains, including phospholipase Cy (PLCy) and the GRP-2 adapter, in turn bound through its SH, domains to the SOS GDP/GTP exchanging factor that activates RAS. RAS activation stimulates RAF kinase which initiates a cascade of protein kinase activations by phosphorylation, which involves MAP kinase kinase (MAPKK), MAP kinase (MAPK), and S6 kinase (RSK). Nuclear translocation of MAP kinase allows phosphorylation and activation of several transcription factors, including c-Jun, c-Myc, and p62TCF,and transcriptional control of other genes, such as c-fos. Other mitogenic hormones bind to receptors with seven membrane domains and activate a phospholipase Cp (PLCP) and then protein kinase C (PKC) which also activates the same protein kinase cascade initiated by RAF. This model is complicated by the existence of several isoforms at almost each level. The model does not provide explanations for recent observations that tyrosine kinase-negative EGF receptor can stimulate MAP kinases (Campos-Gonzalez and Glenny, 1992; Selva et al., 1993;Eldredge et al., 1994) and that EGF receptors lacking all of the known tyrosine autophosphorylation sites can signal a limited MAP kinase activation and mitogenesis (Decker, 1993). On the other hand, clues to other growth factor-elicited signaling pathways are now appearing almost every month. DAG, diacylglycerol.
proliferation” (Baserga, 1985). The sufficient role of any intracellular signal was even questioned: “Before oncogenes came on the scene there appeared to be a chance that activating a few second messengers like CAMPor inositol phosphates and diacylglycerol might be enough to get cells cycling. This was pure wishful thinking. . .” (Hall, 1991). Such
GROWTH MODULATION BY cAMP
63
negative conclusions explain why the role of cAMP in growth control was often ignored in recent general reviews on this subject [e.g., that by Muller et al. (1993) or published in a recent Cell volume on cancer and proliferation controls (64, 249-350, 1991)l. Obviously, the significance of an intracellular signal, like that of an electronic switch, varies depending on the network in which it is involved, that is, on the cell program (Dumont et al., 1981). The confusion in the field results not from demonstrating a role for intracellular signals in the proliferation of specific cells, but from the generalization of such conclusions. The stimulation of proliferation through mechanisms dependent on cAMP is now well demonstrated (Dumont et al., 1989). Our aim in this chapter is to critically review the last 12 years’ advances as to the positive and negative roles of cAMP on the multiplication of normal and cancer cells. The few available data on the possible mechanisms underlying these controls are examined in depth, showing that (and hopefully how) cAMP may have positive or negative effects, depending on the program, that is, the differentiation of the cell type involved. The relationships of these effects to differentiation and tumor development are considered. Previous data, mostly on negative cAMP control of growth, have been extensively reviewed (Pastan and Johnson, 1974; Pastan et al., 1975; Friedman, 1976; Rebhun, 1977). The reader is especially referred to the reviews by Boynton and Whitfield (1983) and Christoffersen and Bronstad (1980) for a critical assessment of the evidence of negative and positive roles of cAMP in various stages of the cell cycle progression. The definitive evaluation of the overall roles played by cAMP in the proliferation of various cell types, as considered here, depends on the precise knowledge of their cycle (Section 1,A) as well as on carefully controlled manipulations of the cAMP signaling cascade (Section 1,B). A. GENERAL CONSIDERATIONS ON CELLCYCLE CONTROLS The basic mechanisms of cell division are probably well conserved in eukaryotic cells, as shown by the finding of homologous and similarly acting genes in yeasts and humans [Nurse, 1990; Nasmyth, 1990; Pelech et al., 1990; Koff et al., 1991; Muller et al., 1993; and several reviews on cell cycle regulation in Cell (79,547-582,1994)l. However, the controls exerted on these basic mechanisms may vary greatly among species, cell types, etc. (Baserga, 1990). The cell cycle of most eukaryotic cells has been operationally divided into four phases, generally described as successive, but which, in fact,
64
PIERRE P. ROGER et a1
are overlapping to an extent depending on the system and the experimental conditions. In the classical description a cell goes through a G, (gap-1)phase, during which it grows and prepares to replicate its chromosomes, then through an S (DNA-synthetic) phase, during which it replicates its chromosomes;then it pauses in a second gap, or G, phase, during which it prepares for the next phase: mitosis, cytokinesis, and division into two new cells. Some cells lack a detectable G, or even G, phase, indicating that the events normally performed during these phases could be achieved during the previous ones. The V79-8 tumor cell lacks a G, phase (Prescott, 19871, implying that the preparation for DNA replication is executed during the previous cycle (Cooper, 1979). In general, cells stop growing with a G, content of DNA, although, more rarely, G, arrest could also occur [it is the rule in some lower metazoans (David and Campbell, 197211. The duration of the G, phase is the most variable, and cell cycle progression is mostly dependent on external controls during this phase. Normal cells can rest in a quiescent, or resting, state (often called Go) outside the cell cycle, with a G, DNA content, from which they can be stimulated to reenter the cell cycle. Such quiescent cells undergo DNA replication after a long “prereplicative phase,” including the time necessary for entry into the cell cycle and the G, phase of the cell cycle (Baserga, 1985). G1 phases and prereplicative phases can be classified into at least four main categories (Boynton and Whitfield, 1983).Type 1is the basic minimum prereplicative phase of continuously cycling cells of established lines. Some of these cell lines, mostly of tumoral origin (HeLa, etc.), are unable to enter into the quiescent Go-like state. Like some prokaryotes, they cycle or they die. Type 2 is the prereplicative phase that follows stimulation of quiescent (Go) cells in starved or confluent culture of “normal” cell lines, which continuously cycle with a type l-like G, phase under unrestricted growth conditions (with serum or growth factors) (3T3 cell lines). The duration of the prereplicative phase of Go cells exceeds that of the type 1 G, phase of the same cells when continuously cycling. Clearly, the longer the cells stay in Go, the more labile and other components needed for the initiation of DNA synthesis are degraded and the longer it takes to initiate DNA synthesis upon stimulation (Baserga, 1985).In the model BALB/c-3T3 mouse embryo cell cultures the progression into the prereplicative phase is dependent on the synergistic cooperation of at least three hormones or growth factors present in serum: a short exposure to PDGF makes the Go cells “competent” to progress through the prereplicative phase in response t o the sequential addition of EGF and IGF-I, the latter controlling commit-
GROWTH MODULATION BY CAMP
65
ment for DNA replication (O’Keefe and Pledger, 1983). Only IGF-I is necessary for supporting a further type 1G, phase in cycling BALB/c3T3 cells, but the presence of PDGF and EGF during the previous cycle is required in order to prevent cells from entering Go after mitosis (Campisi and Pardee, 1984). Although the assignment of different roles for different growth factors might be of restricted application [in human fibroblasts PDGF and EGF are interchangeable, each being able to support all of the prereplicative development (Westermark and Heldin, 198511, the model supports a general concept of the prereplicative phase as a sequence of major regulatory events and points out that Go-G, transition, that is, entry into the cell cycle, requires specific events not involved in type 1 G, phase progression. A good example of events induced during GoG, transition is the rapid stimulation of the expression of protooncogenes such as c-fos, c-jun, and c-myc (Kaczmarek and Kaminska, 1989). Although c-myc expression is sharply and transiently induced by growth factors in quiescent cells, c-myc mRNA levels do not vary during cell cycle progression in continuously cycling cells (Thompson et al., 1985). Moreover, it is clear that continuously cycling cells just after mitosis still contain various relatively stable proteins required for cell cycle progression, DNA replication, and mitosis. Such proteins need resynthesizing during the type 2 prereplicative phase but are subject to only modest fluctuations during type 1 cycling [e.g., DNA polymerase a-associated primase (Tseng et al., 1989); PCNA/cyclin, the auxiliary protein required for DNA polymerase 6 activity (Morris and Mathews, 1989); and p34cdc2 kinase (Draetta et al., 1988)l. Whether type 2 prereplicative development is a good model for the stimulation of quiescent highly differentiated cells in uiuo, such as hepatocytes and kidney and thyroid cells, is still not clear. It is quite possible that the long prereplicative phase of such cells includes specific controls that reflect their differentiation. These cells, when stimulated, are already transcribing differentiation-related genes. Thus, their prereplicative phase does not require a triggering of protein, RNA, and lipid synthesis but includes a reorientation of the genetic expression program from differentiation toward more general activities required for cell cycle progression. This is simply a reformulation of the old concept of growth and differentiation as exclusive activities of the cell. On the other hand, in such differentiated cells repetitious stimulation by functional activators also leads to growth. These adaptive processes enhancing the functional capacity of the tissue might well follow distinct pathways in order to obey different constraints; that is, they may imply a third class (type 3).
66
PIERRE P. ROGER et al.
A fourth type is obviously the very long prereplicative phase following the activation of the small lymphocyte. This cell consists mainly of a small nucleus with highly condensed chromatin and a greatly reduced cytoplasm with few mitochondria, a poorly developed Golgi apparatus, and few if any, polyribosomes. Unlike the other types of prereplicative phase, the first stage of type 4 prereplicative phase includes decondensation of the chromatin, activation of the genome, and de nouo build-up of the mitochondria and the protein-synthetic apparatus. It is obvious that such different models of cell cycle progression and activation of quiescent cells might be regulated differently and at different levels.
B. PROBESOF THE cAMP SYSTEM: PHARMACOLOGICAL AND GENETIC TOOLS So far there is no evidence of cAMP growth regulation in mammalian cells by other means than its intracellular receptors, the PKAs. The demonstration that a given effect of an extracellular signal is mediated by cAMP and PKAs uses pharmacological arguments and genetic manipulation of the CAMP-PKA cascade. As reviewed earlier (Doskeland et al., 19911, if an effect B of agent A is mediated by cAMP and PKAs, 1. A should increase the activity of PKAs in a relevant compartment before or concomitantly with the appearance of B. A simply measurable index is the increase of cAMP concentration at the whole-cell level. It should be noted that while a lack of measurable increase of cAMP excludes a massive general PKA activation, it does not exclude a slight or local activation. Direct measurement of PKA activation within single cells is still at the pioneering study level (Adams et al., 1991, 1993). 2. If A acts by stimulating adenylate cyclase, B should be enhanced by inhibitors of cyclic nucleotide phosphodiesterases, and the action of A should be mimicked by agents that activate endogenous PKAs and by overexpression of the C subunit of PKA. 3. The action of A should be blocked by agents interfering with its stimulation of cAMP production or agents that lower cellular CAMP,by inhibitors of PKAs, and by agents acting downstream of the PKA. However, it should be considered that A may signal via another pathway than the cAMP system, but B may be observed only if the PKA retains its basal activity (i.e., cAMP has a
GROWTH MODULATION BY cAMP
67
permissive effect). In this case the effect of A will be counteracted by agents that inhibit the PKA below basal levels. Molecular mechanisms may be substrate-directed control (the residue phosphorylated by PKA becomes more or less accessible due either to phosphorylation of another residue or to the binding of a ligand) o r the control of phosphatases. Pharmacological and genetic tools used to manipulate the CAMPPKA cascade are summarized in Table I. Forskolin, a plant diterpene, activates vertebrate adenylate cyclase directly and enhances its response to activated G, (the GTP-binding protein transducer that stimulates adenylate cyclase), thus to the receptors that regulate G, (Seamon and Daly, 1986). Extracellular forskolin acts rapidly and is easy to wash off. It is thus particularly useful in inducing pulses of high-level cAMP of defined duration (Seamon and Daly, 1986; Roger et al., 1987a). However, this compound has other effects (Laurenza et al., 1989). For instance, independently of adenylate cyclase, it inhibits insulin-stimulated glucose transport (Joost and Steinfelder, 1987) and CAMP-stimulated L-type Ca2+ current (Boutjdir et a1., 1991);it influences Golgi apparatus function (Lippincott-Schwartz et al., 1991) and the gating of voltage-dependent K+ channels (Hoshi et al., 1988);and it desensitizes the acetylcholine nicotinic receptor (Wagoner and Pallotta, 1988). Analogs of forskolin may help to distinguish the effects on cyclase: 1,9-dideoxy-forskolin does not activate cyclase but reproduces other effects (Laurenza et al., 1989). Inhibition of glucose transport may be counteracted by raising glucose concentrations. Cholera toxin binds to membrane gangliosides through its B subunits. Its A unit is injected into the cytosol, from which it “ADPribosylates” the a, subunit of G,. ADP ribosylation inhibits the GTPase activity of a,,thus leading to permanent activation of G, and of adenylate cyclase (Moss and Vaughan, 1988). Cholera toxin is a useful agent to induce, after a delay, long-term steady elevations of CAMP.However, the B subunit of the toxin, binding to GM, ganglioside, may induce gene expression (Qureshi et al., 1991) or enhance or decrease proliferation per se, independently of cAMP (Spiegel et al., 1985; Spiegel and Fishman, 1987; Tetsumoto et al., 1988). Even the toxinmediated ADP ribosylation may not be entirely specific for G, (McCloskey, 1988). Pertussis toxin, after its internalization, ADP-ribosylates the subunit ai of Gi (the GTP-binding transducing protein that negatively controls adenylate cyclase).ADP-ribosylated Gi becomes inactive, thus relieving cyclase of any negative control (Foster and Kinney, 1985). In
68
PIERRE P. ROGER et al
TABLE I TOOLS USEDTO ASSESSTHE RQLEOF THE cAMP CASCADE Pharmacological Modulation of adenylate cyclase Activation Hormones, neurotransmitters, etc. Cholera toxin Forskolin Pertussis toxin Inhibition Hormones, neurotransmitters
Genetic
Activation Expression of constitutive activating receptor (A,) Expression of constitutive a , Expression of A subunit of cholera toxin
Inhibition Expression of mutated inactive a, (dominant negative phenotype) Antisense oligonucleotides of a, Modulation of cAMP phosphodiesterases Inhibition Activation Expression of yeast cAMP phosphodiesterase RO-201724, methylxanthines, etc. Modulation of PKAs Activation Activation Antisense oligonucleotides of R subunit of 8-chlorophenylthio-CAMP Sp CAMPS PKA Overexpression of C subunit of PKA Pairs of site-specific kinasespecific analogs C subunit of PKA Inhibition Inhibition Antisense oligonucleotides of C subunit Rp CAMPS Expression of mutated cAMP binding domain Inhibitors of protein kinase (inactivating C subunit) R subunit Overexpression of R subunit of PKA Overexpression of Walsh PKA inhibitor Activation Expression of constitutively activated transcription factor cDNA (CREB) Antisense oligonucleotides of CREM Inhibition Antisense oligonucleotides of CREB Expression of dominant inactivating mutations of CREB Expression of CREM
cells with such a tonic control, pertussis toxin therefore induces a prolonged elevation of cAMP levels. However, pertussis toxin also inhibits other Gi effects, for example, the activation in the PIP, phospholipase C cascade, and may inactivate other proteins (Sontag et al., 1991).
GROWTH MODULATION BY cAMP
69
The invasive adenylate cyclase toxins of Bordetella pertussis and Bacillus anthracis are powerful but rarely used tools to specifically elevate cellular cAMP levels in eukaryotic cells. The B . pertussis toxin seems to require activation by calmodulin and is rapidly proteolyzed by a process dependent on ATP binding. Removal of the toxin from the target cell medium results in a rapid loss of its intracellular activity, leading to a decay in the level of cellular cAMP (Hanski, 1989). Inhibitors of cyclic nucleotide phosphodiesterases by inhibiting the catabolism of cAMP also raise cAMP levels. However, due to the negative cooperativity of the phosphodiesterase system (Erneux et al., 19851, these inhibitors are much more efficient in increasing the cAMP response to activators of adenylate cyclase than in increasing basal levels of CAMP. Moreover, they also have other important CAMPindependent pharmacological effects: methylxanthines inhibit adenosine receptors and glucose transport (Steinfelder and Petho-Schramm, 1990), enhance Ca2+ release from sequestration sites in some cells (Huddart and Syson, 19751, and inhibit the induction by TSH and cAMP of ODC in thyroid cells (Mockel et al., 19801, a crucial step in proliferation induction. Preferably, more than one phosphodiesterase inhibitor should be tested, and it should be shown that the potencies of the drugs correlate with their known potencies against the major phosphodiesterases rather than, for example, their activity as adenosine antagonists. Analogs of CAMP are currently used mostly t o mimic-in some cases to inhibit-the effects of CAMP.Ideally, they should penetrate cells, be potent activators of PKA, and be resistant to hydrolysis without blocking the degradation of endogenous CAMP.For initial studies 8-CPTcAMP is preferred by many investigators because it has a generally high potency and is easily available commercially. However, it is also a potent inhibitor of cGMP-specific phosphodiesterase and therefore cGMP may increase in cells treated with this analog (Connolly et al., 1992). The effects of similar concentrations of the noncyclic nucleotides should be checked. Analogs with an S in the P ring come closest to meeting the above criteria. They activate (Sp CAMPS)or inhibit (Rp CAMPS)the action of cAMP (Rothermel et al., 1984; Erneux et al., 1986). However, sometimes, due to low penetration, high concentrations must be used. Commercial preparations of Rp CAMPSshould be used cautiously, since they have been reported to contain biologically active amounts of adenosine (Musgrave et al., 1993). A new Sp cAMP analog has been designed (Sp-5,6-dichloro-l-~-~-ribofuranosylbenzimidazole-3',5 '-monophosphorothioate), which associates the interesting properties of being a very potent and specific activator of PKAs
70
PIERRE P. ROGER et al
with a high lipophilicity and a strong resistance to phosphodiesterases (Sandberg et al., 1991). Butyrylated derivatives (mono- and dibutyryl-CAMP) have often been used in cell cycle control studies. The monobutyryl-CAMP derivative activates PKAs and inhibits some phosphodiesterases (Hsie et al., 1975). However, preparations are often contaminated by butyrate and anyway these analogs are hydrolyzed in the cells releasing butyrate (O’Neill et al., 1975; Coulson and Harrington, 19791, which has effects per se on cell proliferation and differentiation (Herlyn et al., 1988). Experiments with such analogs should therefore test butyrate and the nucleotide itself as controls. Martin and Kowalchyk (1981) have shown that in several cell lines the growth inhibition by cAMP derivatives does not require the 3’3’phosphodiester linkage. Phosphodiesterase and 5’-nucleotidase, which are present in serum-supplemented culture media and at the external surfaces of cells, hydrolyze cAMP and cAMP analogs generating the respective AMP and then adenosine derivatives, which have their own multiple receptors (Niles et al., 1979; Hargrove and Granner, 1982; Weisman et al., 1988; Van Lookeren Campagne et al., 1991). These derivatives are frequently toxic or cytostatic, even at low concentrations. When this occurs, the effects of cAMP and cAMP analogs are prevented by the addition of adenosine deaminase (Hargrove and Granner, 1982; Van Lookeren Campagne et al., 1991), inhibitors of adenosine transport such as dipyridamole (Weisman et al., 19881, or uridine (Niles et al., 1979; Hargrove and Granner, 1982; Weisman et al., 1988) and deoxycytidine (Albert et al., 1991). Adenosine toxicity has been ascribed to an inhibition of pyrimidine synthesis, probably by inhibiting ribonucleotide reductase activity and synthesis (Albert et al., 19911, which is bypassed by uridine addition (Hargrove and Granner, 1982). N6-benzyl-CAMP (Zorn et al., 19931, 8-amino-cAMP, and 7-deaza-cAMP, like 8-Cl-cAMP, are readily broken down into extremely toxic derivatives. The very strong growth-inhibiting properties of 8-C1-CAMP on several cancer cell lines, which is claimed to act through binding to type I1 PKA (Cho-Chung, 1989), were, in fact, ascribed to the toxicity of 8-C1-adenosine (Van Lookeren Campagne et al., 1991). 8-Br-cAMP, often cited as being hydrolysis resistant, is much less resistant than other analogs, such as N6-benzoyl-CAMP and N6-monobutyryl-c AMP. Recently, analogs of cAMP specific for the two sites of the two types of PKAs have been synthesized. If general activation of PKAs is desired, pairs of such analogs with preferential affinity for each of the
GROWTH MODULATION BY cAMP
71
cAMP sites (A and B) on the R subunit of PKA should be used for synergy, one member of the pair binding preferentially to site A and the other binding to site B. Among commercially available analogs, N6-benzoyl-CAMP and N6-monobutyryl-CAMP are site A selective and 8-methylamino-CAMP and 8-aminohexylamino-CAMP are site B selective. Demonstration of synergism suggests that the analogs really act via PKA activation. Other analogs allow specific stimulation of either PKA I or 11(Cho-Chung, 1989; Beebe et al., 1988; Van Sande et al., 1989). Although interpretation of the results in intact cells is complicated by the unknown penetration and degradation rates of these analogs, the great specificity of these compounds makes it very likely that results obtained with a specific pair do involve the corresponding PKA. No synergism would be expected if the analogs acted by phosphodiesterase inhibition, cGMP-dependentkinase activation, or adenosine action on its receptors. Inhibitors of specific protein kinases are commercially available (Hidaka et al., 1991). In general, the specificity of these compounds is only weak, a moderate increase in concentration (factor 5 ) being sufficient t o inhibit other kinases. An effect of such inhibitors might at least suggest that protein kinases, not other proteins such as channels, are involved (Kaupp, 1991). Investigation of the role of the cAMP cascade can also be carried out by the direct microinjection of the purified subunits of the PKAs themselves. Injection of the C subunit should reproduce the activation of the cascade (McClung and Kletzien, 1984; Roger et al., 1988a; Lamb et al., 1988), while injection of the R subunit should inhibit it. Such effects are indeed observed, but their duration is reduced by the fact that uncoupled subunits are degraded very quickly in the cell, which tends to reestablish the R and C subunit concentrations at their normal levels (Richardson et al., 1990). Exogenous free R and C subunits are thus short-lived, and for this reason will not induce the effects of longterm stimulation or inhibition of the cascade. The heat-stable inhibitor of PKA (Walsh and Glass, 1991; Fernandez et al., 1991; Kupperman et al., 1993) and antibodies against the C subunit of PKAs can also be used (Browne et al., 1987). When available, genetic tools might be the best to modulate, in the long term, the activity of the cAMP cascade. Mutations conferring constitutive activation of the cascade, whether at the level of receptors coupled to adenylate cyclase, such as the adenosine A2 receptor (Maenhaut et al., 19901, or at the level of G, (Landis et al., 1989; Zachary et al., 1990), have been shown to confer the proliferation phenotypes to
72
PIERRE P. ROGER et al.
transfected cells in which cAMP is a positive regulator of growth. To block CAMP-dependent responses, antisense oligonucleotides of the subunits of PKAs or CAMP-regulated transcription factors (CREB and CREM) and dominant negative mutated forms of these proteins have been used. For instance, antisense oligodeoxynucleotides targeted against the RII subunit of PKA I1 decrease the level of the protein and result in a great decrease in the differentiating antiproliferative action of cAMP analogs on HL60 leukemia cells (Tortora et al., 1990). Specific expression of mutated “killer” CREB in the somatotrophs of transgenic mice does not allow these cells to proliferate and induces dwarfism (Struthers et al., 1991). Expression of active recombinant fragments of protein kinase inhibitor or dominant mutations of RI has also been used to block “CAMP-dependent”gene transcription (Groveet al., 1987) and mitogenesis (N. Huang et al., 1994). Expression of mutated inactive a, also exerts a dominant negative phenotype (Osawa and Johnson, 1991). cAMP responses, including CAMP-dependent cell proliferation, are suppressed in mammalian cells expressing high levels of the yeast low-K, CAMP-phosphodiesterase gene (Kessin et al., 1992; N. Huang et al., 1994). Finally, cells with a mutation in one of the proteins of the cAMP cascade demonstrate, on the contrary, the role of the cascade in their proliferation. Tumor cells with mutations in the PKAs have been shown to be resistant to the inhibitory effects of the cAMP cascade (Bourne et al., 1975). Mutations that confer constitutive activity to the a, and ai2 subunits of GTP-binding proteins, which activates or inhibits adenylate cyclase, have been found. The fact that such mutations have been demonstrated in tumors suggests that in different cells the relief of either activation or inhibition of the cAMP cascade has an oncogenic potential (Lyons et al., 1990). However, constitutive ai2 and a, have other effects than the modulation of adenylate cyclase. Moreover, the effects of genetic tools might also be indirect. A mutation in a protein kinase might influence many features in the cell that are not directly involved in growth control but could indirectly influence it. In conclusion, there are many pharmacological and genetic tools to investigate the role of the cAMP cascade in the control of cell proliferation. None of these allows unambiguous conclusions. To be reached, such conclusions require the convergence of as many different experimental results as possible. Such results should preferably be obtained in one species, at one phase of development. Indeed, control pathways may vary from one species and from one stage of development to another (Dumont et al., 1981).
GROWTH MODULATION BY cAMP
73
11. NEGATIVE CONTROL OF CELLCYCLE PROGRESSION BY cAMP
A. EARLY WORKON CONTINUOUS CELLLINES The larger body of evidence in the 1970s suggesting a major negative regulatory role for cAMP on cell growth was derived from the study of neoplastic or nonneoplastic fibroblastic cell lines, mostly of rodent origin (types 1 and 2 prereplicative phases). The evidence has been extensively and critically reviewed (Pastan et al., 1975; Friedman, 1976; Rebhun, 1977; Boynton and Whitfield, 1983; Christoffersen and Bronstad, 1980). Thus, we restrict ourselves here to a brief outline of some major features of these studies. cAMP had been proposed as the intracellular mediator signaling the cell to stop growing under conditions that indeed limit its proliferation (e.g., culture confluence, cell starvation, or growth-inhibitory hormones) (Otten et al., 1972). The following criteria were generally accepted in order to establish cAMP as a specific negative regulator of cell cycle progression (Friedman, 1976). 1. The intracellular cAMP level should rise coincidentally with the entrance of cells into the quiescent state and, conversely, should fall as cells exit quiescence and enter the growing state. 2. Experimental elevations of cellular CAMP, at the appropriate time of the cell cycle, should induce cells to enter the quiescent state and prevent quiescent cells from entering the growing state in response to a growth stimulus. The first criterion was based on the now untenable assumption that cAMP could be the sole mediator of quiescence/cell cycle transitions. It depends on the accurate measurement of low basal cell cAMP concentration, which still remains a difficulty: cAMP basal contents vary little under conditions in which stimulated cAMP levels are shown to be inhibited; moreover, to evaluate concentrations, cell mass should be precisely estimated. In agreement with this criterion, the addition of serum, growth factors, or mitogenic proteases to quiescent cells with type 2 prereplicative phase often transiently decreases the levels of cAMP (Pastan et al., 1975). However, as thoroughly examined by Friedman (1976), Rebhun (19771, and Boynton and Whitfield (19831, these studies were challenged by other reports in which such changes were not observed and even opposite changes were demonstrated. Until recently, satisfying the second criterion also presented major difficulties. As discussed in Section I,B, the problem was to obtain
74
PIERRE P. ROGER et al.
elevations in intracellular cAMP in a specific manner and for the desired duration. Prolonged or too high an increase in cAMP concentration could inhibit cell cycle progression, even in systems in which cAMP is unequivocally mitogenic (Boynton and Whitfield, 1983). It should be emphasized that while most effects of cAMP are obtained by an increase in concentration by a factor of 2-4, cAMP enhancements by a factor of 20 are easily reached in the presence of the commonly used phosphodiesterase inhibitors (Miot et al., 1984). Elevating cellular cAMP levels does slow down or block cell cycle progression in a variety of cell lines. Moreover, many studies used brominated or butyrylated cAMP analogs with too often inadequate controls. Some results might be explained by non-CAMP-specific effects. For example, 8-Br-CAMP-resistant variants of GH pituitary cells are deficient in adenosine kinase (Martin and Fbnning, 19811,suggesting that these normal cells are, in fact, inhibited by the nonspecific generation of adenosine derivatives. Nevertheless, the fact that in some tumor cell lines mutants with altered PKAs are resistant to cAMP growth inhibition suggests that in these cells cAMP indeed acted through its normal effector IS49 lymphoma cells (Bourne et al., 1975), CHO cells (Singh et al., 19811, and Y1 adrenocortical tumor cells (Schimmer et al., 198611. Analyses of the kinetics of cell cycle progression in the presence of CAMP-elevating treatments showed that, depending on the experimental system, cAMP inhibits cell cycle progression at different phases. 1. Inhibition during S phase [as found in Reuber H35 hepatoma cells (Van Meeteren et al., 1982)l or at the initiation of S phase has little physiological meaning and might instead reflect a nonspecific perturbation of the balance between nucleotide precursors affecting DNA synthesis. It is reproduced by noncyclic analogs and reversed by uridine (Hargrove and Granner, 1982; Van Lookeren Campagne et al., 1991). 2. Inhibition of G, progression by high cellular cAMP seems to be widespread (Friedman, 1976). In some cell lines with a type 1 cell cycle, such as HeLa cells, it is the only effect of CAMP,whereas in other cell types (with a type 1 or 2 prereplicative phase) it coexists with inhibition (Weidman and Gill, 1977) or even stimulation of GI progression (Baptist et al., 1993). It gives sense to the decrease in endogenous cAMP often observed in late G,/mitosis (Friedman, 1976). These observations were extended recently, and some clues to the mechanism were provided (see Section II,B,6). The physiological meaning of a G, control by cAMP in response to external factors
GROWTH MODULATION BY cAMP
75
should be questioned in somatic cells, as, in general, these cells possess a G, DNA content. 3. In various cell lines cAMP has been reported to inhibit G, progression (Friedman, 1976). When analyzed, this inhibition was found at the different stages of the prereplicative phase that are the targets of positive regulation by serum or growth factors: at the GoG, transition (Heldin et al., 1989), in mid-GI (Leof et al., 1982)’or at the restriction or commitment point for DNA replication in late GI (Gill et al., 1980). In BALB/c-3T3 mouse embryo fibroblast-like cells, cAMP blocks the progression into the prereplicative phase at point V, 6 h before the onset of DNA synthesis. This negative control coexists with a positive effect of cAMP on the earlier acquisition of competence (O’Keefe and Pledger, 1983; Leof et al., 1982; Smets and Van b o y , 1987). Again, this illustrates very well that cAMP may exert both positive and negative influences at different stages of cell cycle progression. The established cell lines used for such analysis are often neoplastic and therefore defective in their proliferation controls. Therefore, in the next section we concentrate our attention on some well-described models of normal cells in which cAMP inhibition of growth might have a physiological meaning (Table 11). B. NORMAL CELLS
1. Fibroblasts PGE, inhibits the restimulation of Go-G, transition in quiescent lung fibroblasts by serum, PDGF, or macrophage-derived growth factor. This effect is reproduced by dibutyryl-CAMP (Fine and Goldstein, 1987). In agreement with such in uitro findings, a suppressive factor released by macrophages, which stimulates endogenous fibroblast PGE, production and cAMP formation, has been suggested to limit bleomycin-induced pulmonary fibrosis in hamsters (Clark et al., 1982, 1983). The inhibition of proliferation by cAMP has also been observed in CCL39 Chinese hamster fibroblasts (Magnaldo et al., 1989b), 10 T 1/2 cells (Matsukawa and Bertram, 1988)’ Syrian hamster embryo cells (Cowlen and Eling, 19921, and normal diploid human fibroblasts (Espinoza and Wharton, 1986). In these systems the inhibition is less potent in exponentially growing cells and mainly affects the restimulation of quiescent cells or the growth of the cells at high density (Magnaldo et al., 1989b; Matsukawa
76
PIERRE P. ROGER et al.
TABLE I1 "NORMAL" VERTEBRATE CELLSIN WHICHcAMP HASBEENFOUNDTO MEDIATE GROWTH INHIBITION Growth inhibitors PGE
PGE, adenosine, VIP, P-adrenergic PGE PGE, p-adrenergic, adenosine PGE PGE, interferon a1p ? ACTH Vasopressin, p-adrenergic ? ? ?
Glucagon
Cell systems Human fibroblasts,".* 10 T 112 cells,b rat-1 cells,b CCL39 Chinese hamster fibroblasts,b and BALBc-3T3 mouse embryo cells* Bovine aortic and rat cerebral microvascular endothelial cellsb Rat arterial smooth muscle cells* Human and rabbit aortic smooth muscle cellsh Murine and human T 1ymphocytesa.b Murine and human B lymphocytesh Mouse macrophage& BC3Hl muscle cell line and L6 myoblasts" Bovine and rat adrenocortical cells in culture" Rabbit kidney cellsb Human and rat astrocytes" Rat glial cells* Rat placental cells* Rat hepatocytes*
"Primary references can be found in the review by Boynton and Whitfield (1983). bSee text for details and references. c ? , Physiological growth inhibitor, not yet known.
and Bertram, 1988; Espinoza and Wharton, 1986). cAMP could affect mainly the Go-G, transition, but not the progression through G I . Different mechanisms have been proposed. In NIH 3T3 and rat-1 fibroblasts, cAMP inhibits the mitogenic signaling cascades converging on MAP kinase activation (Wu et al., 1993; Cook and McCormick, 1993; Hordijk et al., 1994). Analysis of various signaling intermediates indicates that cAMP interferes at a site downstream of p2lra5, but upstream of raf-1 kinase (Cook and McCormick, 19931,possibly involving a direct phosphorylation by PKA of Ser43 in the regulatory domain of raf-1 (Wu et al., 1993; Hafner et al., 1994). The inhibition by cAMP of the Go-G, transition also involves an inhibition of protooncogene expression: c-myc in human fibroblasts stimulated by PDGF (Heldin et al., 1989), c-myc and c-jun induced by TPA in 3T3 cells (Mechta et al., 1989), and c-jun in Syrian hamster embryo cells (Cowlen and Eling, 1992). However, in the latter two systems cAMP also inducesjunB.
GROWTH MODULATION BY cAMP
77
Induction of interferon pz (IL-6) gene expression by cAMP in human fibroblasts (Heldin et al., 1989; Zhang et al., 1988) and in other cells (Ray et al., 1989; Spangelo et al., 1990) could also contribute to explaining an inhibition of growth, sometimes by c-myc down-modulation (Zullo et al., 1985; Forsberg et al., 1988). However, antibodies neutralizing interferon p, fail to prevent the growth-inhibitory effects of cAMP agonists in human fibroblasts (Heldin et al., 1989). In AKR2B fibroblast cells cholera toxin inhibits TGF-P,-stimulated c-sis (PDGF) expression, and consequently its induction of c-myc expression (Howe et al., 1989).Another likely mechanism is the suppression of the expression of cyclin D1 by cAMP in human fibroblasts (Sewing et al., 1993). Whether this is a consequence of the inhibition by cAMP of raf-1 activity or c-myc expression remains unknown. On the other hand, the role of the several CAMP-dependent phosphorylations of cyclin D1 in human fibroblasts (Sewing and Muller, 1994) has not been defined. Pouyssegur and collaborators have even suggested that an inhibition of adenylate cyclase could contribute to the mitogenic activity of serotonin on CCL39 cells (Seuwen et al., 1988) or of a,-adrenergic agonists on CCL39 cells transfected with an a,-adrenergic receptor gene (Seuwen et al., 1990b). Similarly, lysophosphatidate induces proliferation of human fibroblasts and rat-1 cells seemingly by activating Gi2 (Van Corven et al., 1989). Oncogenic mutations activating G,2 inhibit cAMP accumulation (Wong et al., 1991) and induce high-density proliferation and neoplastic transformation of rat-1 fibroblasts (Pace et al., 1991) and NIH 3T3 cells (Hermouet et al., 1991). However, the activation of Gi2 may have other effects than inhibiting adenylate cyclase, such as the activation of phospholipase C, phospholipase A,, channels, or p2lras (Gupta et al., 1990; Van Corven et al., 1993). It is now clear that adenylate cyclase inhibition is not sufficient to explain the effect of ai2 on fibroblast growth (Hermouet et al., 1993; Stephens et al., 1993; Pouyssegur and Seuwen, 1992). 2. Vascular Endothelial and Smooth Muscle Cells Forskolin, cAMP analogs, and various phosphodiesterase inhibitors, including dipyridamole, at their therapeutic plasma concentrations, markedly inhibit the proliferation of bovine aortic (Leitman et al., 1986) and rat cerebral microvascular endothelial cells in culture (Kempski et al., 1987). However, in newborn human dermal microvascular endothelial cells (Davison and Karasek, 1981) and in fetal bovine aortic endothelial cells (Presta et al., 1989) cAMP is reported to be stimulatory.
78
PIERRE P. ROGER et a1
The serum- and PDGF-induced multiplication of vascular smooth muscle cells is inhibited by cAMP (Kempski et al., 1987; Nilsson and Olsson, 1984; Jonzon et al., 1985; Orekhov et al., 1986; HultgardhNilsson et al., 1988; Fukumoto et al., 1988) during the early (Nilsson and Olsson, 1984) and/or late prereplicative phase (Fukumoto et al., 1988). In rat arterial smooth muscle cells cAMP may mediate the growth-inhibitory effects of PGE (Nilsson and Olsson, 19841, adenosine (Jonzon et al., 19851, VIP (Hultgardh-Nilsson et al., 1988), and norepinephrine through P-receptors (Nakaki et al., 19861, as well as the growth-inhibitory effects of PGE in human aortic intima cells in primary culture (Orekhov et al., 1986). In these cells, as in fibroblasts (Cook and McCormick, 19931, cAMP inhibits the PDGF-induced MAP kinase and MAP kinase kinase cascade (Graves et al., 1993). However, inhibition of DNA synthesis cannot be attributed solely to this mechanism, since DNA synthesis is also blocked when forskolin is added several hours after PDGF (Graves et al., 1993). Moreover, unlike the situation in human fibroblasts, the inhibition of DNA synthesis by cAMP in adult rat smooth muscle cells is hardly explained by a n inhibition of early-response nuclear protooncogenes, including c-fos, c j u n , and c-myc (Hultgardh-Nilsson et al., 1994). 3. Immune System As reviewed by Kammer (1988), cellular cAMP in cells of the immune system has long been studied and considered as a potent and possibly physiologically activated immunosuppressive factor, a t least in part because of its negative influence on lymphocyte proliferation. Impairment of T-lymphocyte proliferation by HIV proteins was recently proposed to be mediated by the activation of the CAMP-PKA pathway in these cells (Hofmann et al., 1993). The view that CAMP-increasing factors or cAMP analogs inhibit the mitogenic activation of T lymphocytes is now well accepted. Prostaglandins of the E type and p-adrenergic factors, through CAMP,potently suppress the mitogenic activation by lectins or phorbol esters in murine thymic lymphocytes (Novogrodsky et al., 19831,various murine T-cell clones (Kim et al., 1988), and enriched preparations of human T lymphocytes (Chouaib et al., 1985;Iwaz et al., 1986;Maca, 1984; Ravid et al., 1990; Bartik et al., 1993). Adenosine, via CAMP, also inhibits mitogenesis in mice splenic T lymphocytes (Dos Reis et al., 1986).DNA synthesis stimulation by IL-2 is inhibited by cAMP in many systems (Maca, 1984; Dos Reis et al., 1986; Beckner and Farrar, 1986; Farrar et al., 19871, but it is reported to be relatively insensitive to cAMP in comparison to the activation by TPA in other T-cell systems (No-
GROWTH MODULATION BY cAMP
79
vogrodsky et al., 1983; Kim et al., 1988; Friedrich et al., 1989). PKA I, but not PKA 11,was reported to mediate the inhibitory effects of CAMP on human T-cell proliferation induced through the antigen-specific T-cell receptor-CD3 complex (Skalhegg et al., 1992). cAMP might inhibit T-lymphocyte activation and cell cycle progression by acting at different levels: on the mobilization of intracellular calcium elicited by lectins in human T cells (Chouaib et al., 1987; but see Lingk et al., 19901, on the CAMP-dependent phosphorylation of phospholipase Cy 1, which leads to inhibition of the activating tyrosine phosphorylation of this enzyme induced by the T-cell antigen-receptor complex (Park et al., 1992),on the activation of the Na+/H+antiport in murine thymocytes (Grinstein et al., 19871, etc. Interestingly, PKA I, but not PKA 11, interacts with the CD3 antigen-T-cell receptor complex (Skalhegg et al., 1994).These authors thus proposed a mechanism whereby CAMP, through PKA I-dependent phosphorylation of the T-cell receptor-CD3 complex or associated proteins (e.g., phospholipase Cyl), could inhibit antigen-activated T-cell proliferation. In all T-cell systems IL-2 production, which amplifies the activation process, is potently inhibited by cAMP (Jonzon et al., 1985; Chouaib et al., 1985; Iwaz et al., 1986; Averill et al., 1988; Mary et al., 1987;Novak and Rottenberg, 1990). This is not sufficient to explain the mitogenic inhibition by CAMP,since it is not completely overcome by the addition of IL-2 (Jonzon et al., 1985; Chouaib et al., 1985; Iwaz et al., 1986). Moreover cAMP suppresses the induction by IL-2 and colony-stimulating factors of c-myc protoncogene (Farrar et al., 1987) and ODC (Farrar et al., 1988; Farrar and Harel-Bellan, 1989) in the murine CT6 T-cell clone, and the late expression of the required transferrin receptor in human T cells (Chouaib et al., 1985). However, CAMP,like IL-2, induces the accumulation of IL-2 receptor mRNA (Farrar et al., 1987; Narumiya et al., 1987; Shirakawa et al., 1988) and the expression of fos and myb nuclear protooncogenes (Fsrrrar et al., 1987). In marked contrast with these studies, Takeshita et al., (1990) established a CAMP-dependent growing human T-cell line from an IL-2dependent cell line, and Shirakawa et al. (1988) found that IL-1 induces an increase in cAMP levels in murine thymocytes and in several cell lines and that forskolin and cAMP analogs mimic the comitogenic effect of IL-1 in the presence of lectin in thymocytes. As an apparent paradox, the mitogenic activation of human peripheral T lymphocytes involves the formation of new nuclear CREB complexes. This could imply a positive role for cAMP in the normal physiological regulation of T-lymphocyte activation (Wollberg et al., 1994). Similarly, the transcription of PCNA (the DNA polymerase S auxiliary protein) is in-
80
PIERRE P. ROGER et a1
duced by IL-2 in murine T lymphocytes, at least in part, through CREBs (D. Huang et al., 1994). Mitogenic and antimitogenic influences of CAMP or CAMP-controlled intermediates might well coexist in T lymphocytes and be diversely expressed, depending on the lymphocyte clone and the regulatory context. cAMP also suppresses the mitogenic activation of B lymphocytes (Kammer, 1988).Forskolin, like PGE and dibutyryl-CAMP,inhibits the stimulation by B-cell growth factor of DNA synthesis in human B lymphocytes (Muraguchi et al., 1984).This inhibition is a general finding, although some controversy remains as to the period of the cycle and the metabolic step involved (Hoffman, 1988; Simkin et al., 1987; Blomhoff et al., 1987; Holte et al., 1988).Inhibition of c-myc expression by a CAMP-induced block in transcription initiation was involved in several B-cell systems (Holte et al., 1988; Blomhoff et al., 1988; Slungaard et al., 1987; Andersson et al., 1994). However, some mitogenic pathways (e.g., IL-4) may escape the CAMP-induced inhibition (Vasquez et al., 1991; Kolb et al., 1993). Interferon a/pincreases cAMP levels and inhibits proliferation in a mouse macrophage-like cell line. Many interferon-resistant variants are also resistant to cholera toxin and have a defect in adenylate cyclase [Nagata et al., 1984); however, cAMP is not the mediator of the growth-inhibitory effects of interferon in mouse fibroblasts (Ebsworth et al., 1984) or smooth muscle cells (Fukumoto et al., 1988)I. The growth of murine macrophages is also reversibly inhibited by PGEz and cAMP (Vairo et al., 1990; Jackowski et al., 1990; Rock et al., 1992). The inhibition occurs in mid- or late G, and involves the inhibition of CDK4 activation and RB protein phosphorylation (Kato et al., 19941, which was caused either by a suppression of CYLUcyclin D gene expression (Cocks et al., 1992) or by an increased expression of the CDK inhibitor p27kW (Kato et al., 1994). In contrast with observations in B and T lymphocytes, early responses to mitogens, including Na+/H+ exchange and c-myc induction, are not affected by cAMP (Vairo et al., 1990; Rock et al., 1992). Thus, analysis of the recent literature confirms that cAMP inhibits lymphocyte and macrophage proliferation, but reveals that this activity is quite complex, may bear at different levels, affects differentially some activation pathways, and presents considerable variability in different systems. Positive effects of cAMP are even observed under certain conditions. 4. Other Systems Cholera toxin and forskolin inhibit the serum-stimulated proliferation of a nonfusing muscle cell line (BC3H1) (Kelvin et al., 1989).
GROWTH MODULATION BY cAMP
81
Forskolin inhibits the DNA synthesis in rat glial cells (Kempski et al., 1987) and rat placental cells (Soares et al., 1989) and mitosis of pig epidermal cells in a n in vitro outgrowth system (Takeda et al., 1983). In rabbit kidney vasopressin, isoproterenol, and dibutyryl-CAMP inhibit the synthesis of DNA in both cortical collecting tubules and cortical thick ascending limbs (Wilson and Horster, 1983). In rat hepatocytes in primary culture, although cAMP positively influences early GI progression, glucagon, forskolin, and cAMP analogs, in synergy with dexamethasone, transiently block DNA synthesis at a late GI stage (Vintermyr et al., 1989, 1993a; Thoresen et al., 1990). In contrast to what was observed in T lymphocytes (Skalhegg et al., 19921, this inhibition is mediated by both PKA isozymes (Vintermyr et al., 1993b). The late G, block by cAMP in hepatocytes is associated with a n inhibition of RB protein phosphorylation (Okamoto et al., 1993). 5. CAMP-Dependent Prophase Block of Meiosis in Oocytes There is at least one example of a well-demonstrated and welldocumented physiological G, inhibition by CAMP:the block of meiosis in amphibian oocytes (reviewed by Maller, 1987). The administration of progesterone, which induces meiotic maturation of Xenopus oocytes, induces a rapid decline in cAMP levels. Cholera toxin and other cAMP enhancers completely block the action of progesterone. The microinjection of the active C subunit of PKA inhibits the progression of meiosis, while a protein inhibitor of PKA induces meiosis per se (Darr et al., 1993), which demonstrates the role of the activated PKA. In fact, CAMP,through PKA activation, prevents the activation of p34cdc2 kinase by cyclin B (Rime et al., 1992) and the depolymerization of nuclear lamins (Molloy and Little, 1992). Thus, the decline of cAMP appears both necessary and sufficient t o overcome the prophase block. cAMP also maintains meiotic arrest before nuclear membrane breakdown in mammalian oocytes (Racowsky, 1984; Freter and Schultz, 1984; Bornslaeger et al., 1986; Homa et al., 1991; reviewed by Sato and Koide, 1987; Smith, 1989). 6. CAMP-Dependent Inhibition of G,-Mitosis Transition in Other Systems
The negative influence of cAMP on meiosis events in oocytes might well reflect a more widespread CAMP-dependent G2-mitosis block that also concerns somatic cells (see also Section 11,A).The microinjection of inhibitors of PKA in rat embryo fibroblasts induces early mitosis events such as chromatin condensation and, in cooperation with p34cdc2, nuclear envelope disassembly (Lamb et al., 1991). PKA inhibits the mitotic p34cdc2 kinase activity in fibroblast cell-free extracts
82
PIERRE P. ROGER et ~ l .
(Hohmann et al., 1993). In activated Xenopus egg extracts oscillations of PKA activity control the oscillations of p34cdc2 activity and successive mitoses (Grieco et al., 1994). The inhibition of p34cdc2 activity by PKA is due not to lack of cyclin expression but to a stabilization of Qr15 inhibitory phosphorylation of p34cdc2, which depends on an inhibition of cdc25 phosphatase activity through its dephosphorylation by a serine phosphatase (Grieco et al., 1994). In canine thyroid epithelial cells the potent effect of TSH and cAMP on cell cycle entry and S-phase initiation coexists with a relative inhibition of the cell cycle in G2 phase (Baptist et al., 1993). This block is at least partly dependent on continuous elevation of cAMP levels, since cessation of adenylate cyclase activation, once cells have reached S phase, hastens their entry into mitosis (Baptist et al., 1993). It is associated not only with a stabilization of the Tyr15 phosphorylated form of p34cdc2, but also with an especially high nuclear accumulation of both cyclin A and CDK2 (Baptist et al., 1995). Similar observations have been made in the case of G, delay induced by DNA-damaging treatments (O’Connor et al., 1993) and might explain inhibition of mitosis entry, since cyclin A (likely as a complex with CDK2) is required not only for S-phase progression, but also to ensure the dependence of mitosis on completion of DNA replication (Walker and Maller, 1991). Inhibition of PKA in response to unknown intracellular mechanisms could therefore be a necessary step at late stages of the cell cycle. The physiological meaning of a G, control by cAMP and PKA is speculative. G2 delays, by allowing DNA repair, enhance cell resistance to mutagenic treatments and ionizing radiations (Jung et al., 1994). Whether cAMP could contribute to signal G2 delays induced by DNA damage, or unreplicated DNA (as recently speculated by Grieco et al., 1994) is not known. C. CONCLUSIONS The existence of a negative control of the cell cycle by cAMP is well supported in a limited number of cell systems (Table 11). Both the Go/G, inhibition in some fibroblasts, vascular endothelial and smooth muscle cells, some lymphocytes, and macrophages and the prophase block prior to meiosis in the very special case of oocytes might have a physiological meaning. However, even in normal somatic cell systems, no generalization can be drawn about the mechanisms of this inhibition. In different systems cAMP can inhibit the cell cycle at each of its main control points: not only at the early Go-GI transition (i.e., on the initiation of the cell division program) involving the inhibition of early
GROWTH MODULATION BY cAMP
83
events such as the raf-1-MAP kinase signaling cascade and c-myc expression, but also in mid-G, or at the restriction point in late G, involving inhibition of cyclin D expression, CDK4 activity, and RB phosphorylation. The overall physiological significance of such latter effects, which do not involve inhibition of early mitogenic events, is not well understood. They may represent “safety exits” from the program of cell proliferation for cells exposed to unfavorable conditions. In this case the cells would revert to a Go stage. The late decision point in GI might also be a stage at which cAMP can change the definitive interpretation by the cell of early “mitogenic” events as truly mitogenic rather than functional or even apoptotic signals. Finally, in other cases the effects of cAMP bear on only some mitogenic cascades and should be regarded as belonging to the multiple cross-signaling between cascades existing in the cell.
111. POSITIVE CONTROL OF CELLCYCLE PROGRESSION BY cAMP Aside from the models discussed in Section II,B, most of the examples of negative growth control by cAMP are derived from established cell lines, mostly of tumoral origin. A strikingly converse situation is found for positive control of growth. Table I11 lists many examples of systems in which hormones, neurotransmitters, or growth factors produce an elevation of cAMP before enhancing proliferation and in which pharmacological activators of the cAMP production and cAMP analogs reproduce the mitogenic stimulation. Many of these cells are differentiated epithelial cells in primary culture or normal epithelial cell lines maintaining differentiation characteristics. A CAMP-positive control is also demonstrated for some differentiated nonepithelial systems. It should be pointed out that CAMP-mediated stimulations of proliferation are generally obtained with concentrations of cAMP that are within the physiological range (see, e.g., Ethier et al., 19891, while inhibitions are often found with high “pharmacological” doses of cAMP analogs. Although, in general, evidence is derived from in uitro culture models, in some cases evidence is also available that a mitogenic effect of cAMP could be effective in uiuo. Recently, a strong overall proliferative effect of VIP has been reported in whole incubated mouse embryos (Gressens et al., 1993).It is still not known whether this is mediated by CAMP,like most VIP actions, nor whether it involves the direct activation of regional VIP receptors or a growth factor secretion by the central nervous system. Previously, a stimulation of DNA synthesis by
TABLE I11 VERTEBRATE CELLSIN WHICHCAMPMEDIATES THE ACTION OF GROWTH-PROMOTING FACTORS Stimulants
p-Adrenergic Glucagon, P-adrenergic
Glucagon, PGE, P-adrenergic
Glucose
?
FSH TSH, hCG, thyroid-stimulating immunoglobulin P-Adrenergic, PGE
FSH, LH GRF ? ?
VIP MSH, ? P-Adrenergic, VIP P-Adrenergic ?
ACTH PTH, PGE ? 7
P-Adrenergic, PGE, IL-I ?
PGE PGE, adenosine, VIP PGE
Cell systems Rat parotid acinar cells in primary culture and in vivoa Rat hepatocytes in primary culture" (Miyazaki et al., 1992), T51B rat liver cells," and chick embryo liver cells. MDCK canine kidney cells,a mouse epithelial kidney cells in primary culture," and chick metanephric kidney cells in primary culture" Rat pancreatic islet B cells in primary culture (Rabinovitch et al., 1980) Rat embryo pancreatic epithelial cells (Filosa et al., 1975) Mouse primordial germ cells (De Felici et al., 1993) Immature rat Sertoli cells" Canine, rat, and human thyroid epithelial cells in primary culturec and FRTL5 rat thyroid cell line.' Murine, bovine, and human mammary epithelial c e l l s ~ ~ . ~ and rat prostate epithelial cells" (McKeehan et al., 1984; Nishi et al., 1988) Human, bovine, porcine, and rat ovarian granulosa cells.Murine and human somatotrophsc Rat and guinea pig Schwann Guinea pig enteric glial cellsc Rat embryo sympathetic neuroblasts' Mouse and human melanocytes in primary culturee and S91 mouse melanoma mutant cell line0 Mouse and human keratinocytes in primary culturea (Haegerstrand et al., 1989) Human epithelial cells from foreskin, cornea nasopharynx," and trachea (Willey et al., 1985) Mouse palatal epithelial cells in primary culture (Grove and Pratt, 1984) Rabbit adrenocortical cells in primary culture" Chick osteoblasts in culturec Rat chondrocytes in culture" Human dermal microvessel" and fetal bovine aortic (Presta et al., 1989) endothelial cells Rat thymic lymphoblastsa,c Canine marrow erythroid cells (Brown and Adamson, 1977) BALBc-3T3 mouse embryo cellsc Swiss 3T3 mouse embryo cells",c A10 rat embryo smooth muscle cell line.
"Primary references can be found in the review by Boynton and Whitfield (1983). b?, Physiological growth factor, not yet known. CSee text for details and references.
GROWTH MODULATION BY cAMP
85
cholera toxin or cAMP analogs was obtained in vivo for rodent epidermis (Kuroki, 19811, mammary cells (Silberstein et al., 1984), rat thyroid gland (Pisarev et al., 1970; Lewinski, 1980), adrenal cortex (Hornsby, 1985), rat parotid gland (Tsang et al., 1980), and small oviduct endometrium (Laugier et al., 1988). Such experiments do not prove that cAMP directly stimulates proliferation. cAMP may merely induce the secretion by one type of cell of a mitogenic factor for another type of cell. Indeed, the proliferative effect of cAMP on the adrenal cortex as a second messenger for ACTH seems to be indirect (reviewed by Hornsby, 1985; see also Mesiano et al., 1991). Nevertheless, constitutive activation of the cAMP cascade by genetic means in transgenic mice suggests that cAMP may be a more direct mitogenic stimulus in somatotrophs (Burton et ul., 1991) and thyroid cells (Ledent et ul., 1992). The consequences of an inappropriate cAMP increase in several human tissues can be deduced from the clinical pattern of McCuneAlbright syndrome (Levine, 19911, which is caused by a somatic mutation of the G,a gene in the early embryo and the expression of an activated G p protein that constitutively stimulates adenylate cyclase in multiple tissues (Weinstein et al., 1991; Schwindinger et al., 1992; Shenker et al., 1994). Abnormalities in McCune-Albright syndrome include GH-producing pituitary adenoma, hyperthyroidism and thyroid nodules, autonomous maturation associated with hyperplasia of ovarian follicles resulting in sexual precocity, some hyperplasia of testis Leydig cells, adrenocortical adenoma, thymic hyperplasia, gastrointestinal polyps, and polyostotic fibrous dysplasia. A. RECENTEXAMPLES OF CAMP-MEDIATED POSITIVE GROWTH CONTROL Much of the evidence of CAMP-mediated growth stimulation in the systems listed in Table I11 was already acquired before 1983 and was extensively reviewed and discussed by Boynton and Whitfield (1983). Therefore, we restrict our task here to updating their effort in a discussion of recent significant advances using present methodologies reported for only some systems. Some of the best-confirmed and most popular examples of CAMP-mediated stimulation of growth are no longer recalled in this section (including keratinocytes, hepatocytes, and parotid cells), in the absence of such new data. 1. Thyrocytes TSH, mainly through CAMP,stimulates the synthesis and secretion of thyroid hormones by thyroid follicular epithelial cells (thyrocytes),
86
PIERRE P. ROGER et al.
but also controls tissue growth. By relieving the feedback exerted by thyroid hormones on the pituitary, any treatment that decreases thyroid secretion induces a secretion of TSH. TSH, in turn, causes the hyperfunction, hypertrophy, and hyperplasia of the thyroid gland (Dumont, 1971). In hypophysectomized rats the injection of TSH increases the mitotic activity of the thyroid gland. This effect is reproduced by the injection of dibutyryl-CAMP (Pisarev et al., 1970; Lewinski, 1980). TSH also stimulates DNA replication and the proliferation of canine thyrocytes in primary culture as cell monolayers with serum and triggers DNA synthesis in the absence of serum (Roger et al., 1982, 1983; Roger and Dumont, 1984). Binding of TSH to receptors coupled to adenylate cyclase results in a rapid and sustained elevation of cellular cAMP levels. In this system the adenylate cyclase activators-cholera toxin and forskolin-and various cAMP analogs perfectly mimic the mitogenic effects of TSH (Roger et al., 1983; Roger and Dumont, 1984; Van Sande et al., 1989) as well as its functional and differentiation effects (Dumont, 1971; Passareiro et al., 1985; Roger et al., 1985; Gerard et al., 1989). This has been confirmed in a model of rat thyroid follicles in suspension culture (Wynford-Thomas et al., 1987) and, despite earlier negative reports (Valente et al., 1983a,b), in the FRTL5 rat thyroid cell line (Dere and Rapoport, 1986; J i n et al., 1986; Yun et al., 1986; Ealey et al., 1987; Tramontano et al., 1988). In the WRT rat thyroid cell line the TSH-induced DNA synthesis is inhibited by microinjection of a n antibody to G,a (Meinkoth et al., 1992). The direct mitogenic effect of TSH via cAMP was confirmed in normal human thyrocytes in primary culture (Roger et al., 1988b). In the absence of serum, in these different systems the mitogenic effects of TSH and cAMP require the comitogenic influence of insulin or IGF-I (Roger et al., 1983, 198713, 198813; Tramontano et al., 1988; Smith et al., 1986). The role of CAMP in thyroid growth in uitro and in uiuo has now been confirmed by manipulating the expression of the adenosine A2 receptor positively coupled to adenylate cyclase. This receptor, recently cloned in our laboratory (Libert et al., 1989), appears to be physiologically constitutive (Maenhaut et al., 19901, the concentration of adenosine normally present in many tissues being sufficient to activate it. Microinjection of the mRNA of this receptor in dog thyrocytes is sufficient to elicit DNA synthesis as well as differentiation expression (Maenhaut et al., 1990). Expression of this receptor in transgenic mice under the control of the thyroid-specific thyroglobulin promoter elicits goiter and hyperthyroidism and greatly enhances cell proliferation, that is, hyperfunction, differentiation expression, and growth (Ledent et al., 1992). In some human hyperfunctioning thyroid ade-
GROWTH MODULATION BY cAMP
87
nomas a mutational constitutive activation of G,a has been found (Lyons et al., 1990; Suarez et al., 1991; O’Sullivan et al., 1991). In patients affected by McCune-Albright syndrome, this mutation is associated with a nodular adenomatous goiter and hyperthyroidism (Weinstein et al., 1991). Moreover, in seven of 12 hyperfunctioning thyroid adenomas, new mutations of the TSH receptor, conferring constitutive activation of adenylate cyclase but not of phospholipase C, have been found (Parma et al., 1993). Thus, in thyroid cells the elevation of cAMP provides a sufficient relay for the induction of cell proliferation by TSH. 2. Somatotrophs cAMP has been proposed as a mediator for the stimulation by the hypothalamic GRF of GH secretion and gene transcription in pituitary somatotrophs (Barinaga et al., 1985). Forskolin mimics the mitogenic effect of GRF in rat somatotrophs in primary culture, while somatostatin, which reduces the GRF-stimulated rise in cAMP (Bilezikjian and Vale, 19831, attenuates this growth effect (Billestrup et al., 1986). In 40% of hyperfunctioning adenomas of the pituitary gland, an activating somatic mutation of the G,a has been found (Vallar et al., 1987; Landis et al., 1989, 1990; Lyons et al., 1990). This mutation also produces GH-secreting pituitary adenoma in McCune-Albright syndrome (Weinstein et al., 1991). Moreover, expression of the activating subunit of cholera toxin in somatotrophs of transgenic mice causes hyperfunctioning adenomas (i.e., hyperfunction, differentiation expression, and mitogenesis) (Burton et al., 1991). As in the thyroid gland, chronic constitutive activation of the cAMP cascade leads to stimulation of function, differentiation, and growth. Moreover, transgenic mice overexpressing in the somatotrophs a transcriptionally inactive mutant of CREB, which cannot be phosphorylated, exhibit atrophy of the pituitary gland, depletion of somatotrophs, and dwarfism (Strutters et al., 1991). 3. Mammary Epithelial Cells In mouse mammary epithelium ovariectomy reduces cAMP levels and DNA synthesis (Silberstein et al., 1984). Implants releasing cholera toxin, cAMP analogs, or forskolin or systemic injections of cholera toxin stimulate growth and morphogenesis of mouse mammary ducts (Silberstein et al., 1984; Sheffield et al., 1985). Cholera toxin also reinitiates growth in senescent mammary epithelium (Daniel et al., 1984). The earlier findings by Yang et al. (1980) and Taylor-Papadimitriou et al. (1980) involving cAMP in synergy with other factors in the stimula-
88
PIERRE P. ROGER et a1
tion of proliferation of cultivated mammary cells have been largely confirmed. PGE,, mimicked by cholera toxin, cAMP analogs, or phosphodiesterase inhibitors, potently synergizes with EGF, insulin, or IGF-I and induces in serum-free cultures the rapid proliferation of mammary epithelial cells from mice (Imagawa et al., 1988), rats (Ethier et al., 1987,1989),humans (Stampfer, 1982; Hammond et al., 1984), and calves (Shamay et al., 1990).
4. Bone Cells PTH stimulates DNA synthesis and proliferation in various osteoblastlike cells and chondrocytes in in vitro models (van der Plas et al., 1985; McDonald et al., 1986; Koike et al., 1990; Somjen et al., 1990).Whether the PTH mitogenic effect is direct and mediated by cAMP or Ca2+ remains controversial. cAMP analogs stimulate DNA synthesis in rat cartilage segments in uitro (Bomboy and Salmon, 1980). cAMP analogs and forskolin reproduce the PTH effect on the DNA synthesis of chick osteoblast-like cells in culture (van der Plas et al., 1985). These effects could be indirect, since in fetal rat osteoblasts in culture (McCarthy et al., 1990), PTH and cAMP stimulate the synthesis of IGF-I, a known mitogen for these cells. DNA synthesis and cAMP responses to PTH have recently been dissociated. In embryonic rabbit and chick chondrocytes PTH increases cAMP at 100-fold higher concentrations than those required for stimulation of cell division (Koike et al., 1990). Moreover, different fragments of the PTH molecule elicit the growth of cAMP responses in various models of chondroblasts and osteoblasts (Somjen et al., 1990). This evidence seems to exclude a direct mitogenic effect of cAMP in these cells. 5. Ovarian Follicular Granulosa Cells There is no doubt that gonadotropins acting through CAMP, such as FSH, enhance granulosa cell proliferation and differentiation in uiuo. In McCune-Albright syndrome the activating mutation of the G,a is found in hyperplastic ovarian follicles that undergo maturation independently of gonadotropins, resulting in the development of cystic ovaries and sexual precocity (Weinstein et al., 1991). However, in vitro studies are contradictory. No mitogenic effect of FSH or LH was demonstrated on dissociated human, bovine, porcine, or rat granulosa cells in culture (Savion et al., 1981; Hammond and English, 1987; Dorrington et al., 1988). Nevertheless, FSH very strongly potentiated the small stimulation of DNA synthesis by TGF-f3 in rat granulosa cell
GROWTH MODULATION BY cAMP
89
culture (Dorrington et al., 1988).By contrast, in intact hamster ovarian follicles incubated or cultivated in uztro, FSH or LH induced within 2 h an increase in DNA synthesis (Roy and Greenwald, 1986,1988, 1989). This peculiarly rapid effect on DNA synthesis, also observed by others (Pedersen, 1972; Monniaux, 19871, indicates that some granulosa cells are naturally arrested in late GI. It is reproduced by 8-Br-CAMP (Roy and Greenwald, 1988). Dibutyryl-CAMP also increases the granulosa cell number in isolated mouse follicles cultured in collagen gels (Carroll et al., 1991) and promotes the proliferation of chicken granulosa cells in culture (Yoshimura and Tamura, 1991). Whether the proliferation effects of FSH in uiuo or on intact follicles in uitro represent direct mitogenic effects remains unclear. It has been suggested that the stimulation of IGF-I production by FSH and cAMP in porcine granulosa cells could mediate, in part, the growth effects of these factors (Hsu and Hammond, 1987). However, FSH and cAMP synergize with IGF-I in the induction of DNA synthesis in serum-free cultures of rat granulosa cells (Bleg et al., 1992). On the other hand, a neutralizing antibody against EGF inhibits the stimulation of DNA synthesis by FSH and cAMP in hamster ovarian follicles, which indicates that it is mediated by an enhanced synthesis of EGF (Roy and Greenwald, 1991). 6. Adrenocortical Cells ACTH stimulates cAMP synthesis, hypertrophy, and hyperplasia of adrenocortical cells in animals (Hornsby, 1985). Macronodular hyperplasia of adrenal cells and adrenocortical adenoma that contain the activating mutation of G,a have been described in McCune- Albright syndrome (Weinstein et al., 1991). However, most authors have reported that ACTH and cAMP inhibit the proliferation of adrenocortical cells both in short-term cultures and in established lines from adrenocortical tumors (Hornsby, 1985). In contrast, Menapace et al., (1987) have reported that ACTH can directly stimulate DNA replication and multiplication of the parenchymal cells in primary cultures of rabbit adrenal cortex in a serum-free medium. The mitogenic effect of ACTH does not seem to be completely mimicked by CAMP,since cholera toxin and dibutyryl-CAMP stimulate DNA replication but not entry into mitosis in these cells (Menapace et al., 1987). Here, as in bone cells and ovarian granulosa cells, whether the mitogenic effect of cAMP is direct, or indirect through the induction of a growth factor such as FGF, remains an open question. FGF is induced by ACTH and cAMP analogs in the human fetal adrenal gland (Mesiano et al., 1991).Species differences in the mitogenic pathway of a hormone remain a possibility.
90
PIERRE P. ROGER et a1
7. Melanocytes Cultures of melanoma cells have long been established, providing interesting information in some lines on how MSH and its second messenger cAMP can stimulate cell multiplication (Pawelek et al., 1975). Since 1982 pure cultures of normal melanocytes have been established and used for analysis of growth and differentiation controls. Eisinger and Marco (1982) found that TPA and cholera toxin synergize to support the proliferation of normal human melanocytes. This was confirmed using a variety of substances increasing cAMP (Halaban, 1988; Abdel-Malek et al., 1992). cAMP is also permissive for the mitogenic activity of FGF (Halaban et al., 1988; Abdel-Malek et al., 1992). In fact, optimal growth of normal human melanocytes in serum-free conditions requires PKC activation , FGF, insulin, and MSH or other cAMP increasing factors (Herlyn et al., 1988; De Luca et al., 1993). Dibutyryl-CAMP also induces DNA synthesis and proliferation of isolated mouse embryo melanocytes (Mayer, 1982). The sustained proliferation of mouse melanoblasts in a serum-free medium depends on the presence of keratinocytes and is stimulated synergistically by dibutyryl-CAMP and FGF (Hirobe, 1992). 8 . Peripheral Nervous System Like melanocytes, Schwann cells and glial cells are derivatives of neural crest cells that appear in the early stages of the embryogenesis of vertebrates. In rodent embryos the proliferation of Schwann cells that provide support for axons is a n important part of peripheral nerve development. In adult rodents Schwann cells are stimulated to divide when a nerve is injured. The first observations by Raff et al. (1978), as expanded by Sobue et al. (1986), demonstrated that cAMP analogs, cholera toxin, and forskolin are potent inducers of DNA replication and proliferation of neonatal rat Schwann cells of the sciatic nerve cultured in the presence of serum. The increase in cellular cAMP is necessary to observe the mitogenic effects of PDGF and FGF and potentiates the growth effects of a glial growth factor and TGF-p (Ridley et al., 1989; Davis and Stroobant, 1990). In the neonatal guinea pig cholera toxin and dibutyryl-CAMP are mitogenic for Schwann cells, but also for enteric glial cells cultured with or without serum (Eccleston et al., 1987). The question of what controls Schwann cell cAMP levels in vivo is still unanswered, but a n axon-bound molecule could be involved, thus providing a mechanism by which axons can stimulate the proliferation of their support cells (Eccleston, 1992). VIP, through CAMP, stimulates the proliferation and survival of
GROWTH MODULATION BY cAMP
91
embryonic rat sympathetic neuroblasts cultured in a serum-free medium (Pincus et al., 1990a,b). 9. Thymocytes In sharp contrast with the well-demonstrated inhibitory effect of cAMP on lymphocyte activation and IL-2-induced mitogenesis, Shirakawa et al. (1988) reported that the comitogenic effect of IL-1 on phytohemagglutinin-induced murine thymocytes can be mediated by CAMP.IL-1 rapidly induces a sharp cAMP rise and a 4-h treatment with forskolin completely mimics the mitogenic effect of IL-1. 10. Nonepithelial Continuous Cell Lines Contrasting with data on normal diploid human fibroblasts and CCL39 Chinese hamster fibroblasts, cAMP positively influences the proliferation of fibroblast-like cell lines from mouse embryo. In different BALBc-3T3 cell clones the amplitude of the mitogenic response to PDGF or other growth factors is positively correlated with the basal levels of cAMP (Olashaw et al., 1984). In addition, although cAMP inhibits the progression of BALBc-3T3 cells in mid-G, (O’Keefe and Pledger, 19831, it potentiates the induction of competence by PDGF and TPA (Wharton et al., 1982; Smyth et al., 1992). As demonstrated by Fbzengurt (1986) and collaborators, the mitogenic effect of cAMP is still more conspicuous in some Swiss 3T3 mouse embryo cell lines. A variety of agents that promote cAMP accumulation in these cells, including PGE, (Rozengurt et al., 1983a), an adenosine agonist (Rozengurt, 1982a),ATP (N. Huang et al., 19941, and VIP (Zurier et al., 19881, induce DNA synthesis, acting synergistically with insulin, phorbol esters, and other growth factors. Interestingly, VIP also functions as a growth factor on whole mouse embryos in uitro (Gressens et al., 1993). Their mitogenic effects in Swiss 3T3 cells are mimicked by cholera toxin, forskolin, and cAMP analogs and potentiated by phosphodiesterase inhibitors (Fbzengurt et al., 1983a; Rozengurt, 1982a; Zurier et al., 1988). Moreover, these effects are inhibited in cells transfected with expression vectors encoding a mutated R subunit of PKA and a yeast low-K, phosphodiesterase (N. Huang et al., 1994). CAMP-elevating agents, among other factors, are required to support the sustained proliferation of Swiss 3T3 cells in a serum-free medium (Brooks et al., 1990). In these cells PGE, synthesis and secretion may mediate, in part, the growth effects of PDGF (Rozengurt et al., 1983b1, bombesin (Mehmet et al., 1990a), mastoparan (a peptide from wasp venom) (Gil et al., 19911, and IL-1 (Burch et al., 1989).These factors increase the synthesis and secretion of E-type prostaglandins,
92
PIERRE P. ROGER et al.
which, in an autocrine fashion, would stimulate cAMP synthesis (RQzengurt et al., 1983b; Mehmet et al., 1990a; Burch et al., 1989; Gil et al., 1991); however, the CAMP-mediated mitogenic activity of IL-1 in thymocytes appears to be direct (Shirakawa et al., 1988), and in FRTL5 thyroid cells IL-1 stimulates proliferation independently of cAMP (Mine et al., 1987). In contrast with observations of the primary culture of vascular smooth muscle cells, PGE, via CAMP,induces DNA synthesis in the quiescent A-10 rat embryo vascular smooth muscle cell line (Owen, 1986). However, it is inhibitory when these cells are asynchronously cycling in the presence of serum (Owen, 1986).
B. SYNERGISM BETWEEN cAMP AND OTHERMITOGENIC FACTORS Some form of cooperation is generally needed between different growth factors in order to achieve the stimulation of division of normal cells (de Asua et al., 1977; Pledger et al., 1978).This is not surprising, as the decision to divide should be submitted to several restrictions, that is, to several independent controls. A marked synergism among different growth factors often indicates that they control different mitogenic events, which must be executed in the right sequence to allow progression into the prereplicative phases and commitment to DNA replication and cell division. By analogy, the stepwise progression of tumors could reflect successive bypasses at different levels of growth regulation, that is, escape from different restrictions and controls. A different classification of multiple signals required for mitogenesis was presented by Rozengurt (1986). In the Swiss 3T3 cell line two or three synergistic signals may also be required for entry into S phase, but no temporal distinction is observed between them, and in some cases a single factor, such as PDGF or bombesin, is sufficient to elicit growth as it activates different signaling cascades (Rozengurt, 1986). The positive role of cAMP in cell cycle progression provides interesting examples of such synergistic cooperation. The full mitogenic effect of glucagon through cAMP requires both insulin and EGF in rat hepatocytes (McGowan et al., 1981; Friedman et al., 1981; Miyazaki et al., 1992). In melanocytes the growth effect of MSH and cAMP requires insulin and phorbol esters or FGF (Halaban, 1988). In human bronchial epithelial cells isoproterenol, through CAMP,potentiates the mitogenic stimulation by EGF and a pituitary extract (Willey et al., 1985). In human mammary epithelial cells cAMP stimulates growth only in the presence of EGF or serum (Taylor-Papadimitriou et al.,
GROWTH MODULATION BY cAMP
93
1980). The CAMP-dependent mitogenic activity of glucagon or PGE, in MDCK cells in a serum-free medium is observed in the presence of insulin or IGF-I (Taub et al., 19791, as are the effects of TSH in canine, rat, and human thyroid epithelial cells (Tramontano et al., 1988; Roger et al., 1987b, 1988b; Smith et al., 1986). In canine thyrocytes (Roger et al., 1987b) and mouse mammary cells (Imagawa et al., 1988; Ethier et al., 1987) a supplementary synergism is provided by EGF. The combination of EGF, insulin, and cAMP (cholera toxin) is also active in rat prostate epithelial cells (McKeehan et aZ.,1984; Nishi et al., 1988). Thus, the optimal proliferation of many different epithelial cells is obtained with similar combinations of a CAMP-stimulating hormone with general growth factors such as EGF or insulin (IGF-I).Such similitude could denote similar growth control mechanisms in these different epithelial cell types, as in the nonepithelial Swiss 3T3 cell line, in which the synergistic interactions between cAMP agonists and either EGF or insulin have been most thoroughly examined (Rozengurt, 1986). In these cells (Rozengurt, 198213) and in most experiments with canine thyrocytes, both insulin and cAMP are required t o initiate the prereplicative phase. However, in Swiss 3T3 cells, unlike canine thyrocytes, other growth factors, such as EGF, vasopressin, or phorbol esters, can also synergize with cAMP in the absence of insulin (Rozengurt, 1986). The increase in cellular cAMP also potentiates the mitogenic activity of several growth factors in rat Schwann cells (reviewed by Eccleston, 1992). In the analysis of such synergisms, we must distinguish the real signals triggering proliferation in uiuo from ubiquitous necessary comitogenic factors, that is, prove that a factor is involved in physiological control. For example, in the thyroid gland TSH is regulating; insulin, although necessary, is not. Glucose may be necessary for cell division in most systems, but it is only a signal for islet cells. Conversely, mice in which either the IGF-I or IGF-I1 receptor gene has been “knocked off” develop harmoniously, but at 60% of the normal rate.
C. POSITIVE REGULATION OF CELLCYCLE PROGRESSION BY cAMP As briefly discussed earlier, progression throughout the prereplicative phase and triggering of the deterministic part of the cell cycle (late GI, S, G,, and mitosis) depend on the sequential execution of several potentially rate-limiting events. Only some of these events may be regulated by CAMP, and their apparent importance for cell cycle progression may depend on both the cell type and its environment (experimental conditions imposed on the cell, that is, the condi-
94
PIERRE P. ROGER et a1
tions used for maintaining the cell in a quiescent state). In general, however, we are interested in the final influence on the commitment of cell cycle progression resulting from the integration of the controls a t the various restriction points, that is, on the overall proliferation effect. In density-inhibited BALB/c-3T3 cells cAMP potentiates the acquisition of competence induced by PDGF or FGF (Wharton et al., 1982), but sustained cAMP elevation impedes progression into the rest of the prereplicative phase [the cells are arrested at the V point in mid-G, (Leof et al., 1982)l. A 4-h elevation of CAMP also potentiates the induction of DNA synthesis by phytohemagglutinin observed after 48 h in rat thymocytes (Shirakawa et al., 1988). In cultured rat hepatocytes glucagon and CAMP-elevating agents enhance early in G,, but delay late in G,, the progression of the cycle (Vintermyr et al., 1989, 1993a; Thoresen et al., 1990; Miyazaki et al., 1992; Friedman et al., 1981). The latter effect requires higher cellular cAMP concentrations (Thoresen et al., 1990; Vintermyr et al., 1993a). By contrast, Boynton and Whitfield (1983) have reviewed the evidence for a late surge in cAMP preceding the onset of DNA replication. In T51B liver cells and in thymic lymphoblasts arrested near the end of G, by calcium deprivation, cAMP enhancers and cAMP itself can overcome this block and trigger DNA synthesis (Boynton and Whitfield, 1983). In some systems such as Swiss 3T3 cells (Rozengurt, 1982b1,Schwann cells (Sobue et al., 1986), and canine thyrocytes (Roger et al., 1987a), not only does cAMP initiate the prereplicative development, but a sustained elevation of cAMP is required before inducing DNA replication. In the latter system a late commitment point to DNA synthesis is still dependent on cAMP elevation by forskolin. Before this commitment point interruptions in forskolin presence, as short as 2h, delay the onset of DNA synthesis. Thus, in the presence of forskolin, canine thyrocytes progress toward S phase, but if this factor is withdrawn before the cells reach the commitment point, they rapidly regress to an earlier part of G,, from which they can be rescued by forskolin readdition (Roger et al., 1987a). Thus, in these cells the CAMP-dependent events, which are crucial for the progression through G, phase and for the commitment to DNA synthesis, are peculiarly labile. The fact that pulses of cycloheximide also bring canine thyrocytes back to the resting state (Roger et al., 1987a) shows that the proteins associated with passage through the restriction point in late GI are also labile, as was suggested by Pardee (1989). Canine thyrocytes should be considered as a system in which all of the major control points of prereplicative phase can be regulated by CAMP.
GROWTH MODULATION BY cAMP
95
Thus, the major control points in the prereplicative phase (initiation, progression, and S-phase commitment in late GI) can be dependent on CAMP augmentation, but many cells might be defective at some of these control points. It is interesting that in some non-CAMP-responsive cells clones of cells responding by proliferation to CAMPelevation may arise. If the positive control by CAMPat some steps may be lost in some cases, it can be induced de nouo in others (Pawelek et al., 1975; Takeshita et al., 1990). In contrast with the large body of evidence of the positive effects of CAMPon GI-phase progression and with the frequent inhibition of G, transit in established cell lines, there are sparse reports of a CAMPdependent G2 control of mitosis in uiuo. The p-adrenergic blocker propranolol prevents regenerating rat liver cells from entering mitosis without affecting their ability to initiate or complete DNA replication. It also inhibits the early prereplicative surge of cAMP that occurs shortly after partial hepatectomy, and CAMPinjection 2 h after operation reverses the mitosis-inhibiting action of propranolol. Rixon and Whitfield (1985) therefore proposed that an early CAMP-dependent prereplicative event determines mitosis rather than DNA replication. By contrast, Bybee and Tuffery observed a 3-fold increase in the appearance of metaphases in follicular cells of the thyroid gland and in acinar cells of the parotid and submaxillary glands, as early as 5 min after injection of TSH or isoproterenol in the rat (bdmond and Tuffery, 1981; Bybee and Tuffery, 1988, 1989). These results implicate the immediate recruitment by hormones acting through CAMPof a cell population resting in late G, or early mitosis in these tissues. However, we were unable to observe such an effect in primary cultures of thyrocytes stimulated by TSH (P. P. Roger, unpublished observations). On the other hand, Browne et al. reported that inhibition of PKAs by microinjection of a protein inhibitor delays the formation of the mitotic spindle in PHKl cells (Browne et al., 1987) and in fertilized sea urchin eggs (Browne et al., 1990). An increase in CAMPalso initiates meiosis in brittle star oocytes (Yamashita, 1988). Such effects would only have physiological relevance as controls in cells blocked in G, in their quiescent state. OF POSITIVE CONTROL OF CELLCYCLE D. BIOCHEMISTRY
PROGRESSION BY CAMP 1. Protein Kinase Activation PKAs are the major cellular receptors for CAMPin eukaryotes, and most of the effects of CAMPare assumed to be mediated through PKA
96
PIERRE P. ROGER et al.
activation and phosphorylation of specific protein substrates (Lohmann and Walter, 1984; Taylor et al., 1988). In the inactive form isozymes of PKA are tetramers composed of two C and R subunits. Two major types of isozymes, differing from each other only with respect to their R subunits (RI and RII) are found in most mammalian cells, but their relative amounts vary considerably from tissue to tissue or in different species for the same tissue and are very sensitive to physiological or pathological processes. Regulation is different in the two kinases, R being phosphorylated when PKA I1 is activated, while ATP mostly stabilizes PKA I and thus antagonizes activation. Isoforms of the RI and RII subunits, and of the C subunit, have been documented. The RIa, RIIa, and Ca subunits are expressed in most tissues, whereas RIP and Cp are found mainly in the brain and the testes. RIIP is most abundant in the brain, ovaries, and testes, whereas Cy has so far been found only in the testes. The biological significance of the coexistence of these distinct isozyme forms with a common C subunit is still largely unknown (Doskeland et al., 1993). However, differential anchoring of the R subunits leads to nonhomogeneous partitioning of C subunits and could cause a different protein substrate targeting of the two kinases. RI isoforms are primarily cytoplasmic, while RII isoforms are mainly localized on membranes, subcellular organelles, or the cytoskeleton through binding t o different AKAPs (for recent references see Scott et al., 1990; Ndubuka et al., 1993; Keryer et al., 1993). Distinct compartmentation of PKA isoenzymes should also be considered in view of recent evidence of compartmentalized accumulation of CAMP, which could explain the differential activation of PKA subtypes in response to different hormones (Scott and McCartney, 1994). Upon dissociation some C subunits migrate to the nucleus (Meinkoth et al., 1990). Heat-stable protein kinase inhibitors (aand P), which bind to a dissociated active C subunit and inhibit it, can suppress the effects of basal levels of cAMP on the activity of the protein kinases (Walsh et al., 1990). They also inhibit the nuclear translocation of the C subunits (Fantozzi et al., 1992). cGMP-dependent protein kinase can also be activated by cAMP in some special situations, explaining the cAMP effect on smooth muscle relaxation (Jiang et al., 1992). Although in the yeast Saccharomyces cereuisiae, the dependence of cell cycle progression upon activation of PKA is well demonstrated (Toda et al., 1987), there is little direct evidence that the activation of PKA mediates the CAMP-positive control of growth in mammalian cells. In several established cell lines marked variations of PKA isozyme relative amounts and activations were observed during cell cycle progression (reviewed by Boynton and Whitfield, 1983; Lohmann and
GROWTH MODULATION BY cAMP
97
Walter, 1984). This led some authors to postulate a specific role for either PKA I or I1 in the control of the cell cycle (Byus et al., 1977). However, these purely associative relationships can by no means indicate a causal link between the activation of PKA and replication controls. Furthermore, in these systems there is often no demonstration that cAMP plays a positive role in growth regulation. In some tumor cell lines PKAs were firmly implicated in the inhibition of proliferation by cAMP by the biochemical analysis of CAMPresistant mutants (see Section 11,A). Such evidence is lacking in the case of the positive control. Nevertheless, the elevated requirement for cAMP of a mutant melanoma cell line, as compared to its parental counterpart, is related to an elevated activation constant of PKA I for cAMP (Pawelek, 1979). Antisense oligonucleotides to the C subunit of PKA inhibit DNA synthesis of mouse mammary epithelial cells (Sheffield, 1991).The microinjection of the heat-stable PKA inhibitor inhibits, in part, the induction of DNA synthesis by TSH and 8-Br-CAMP in WRT rat thyroid cells (Kupperman et al., 1993). Martin and collaborators, by means of direct measurements of the activation of each PKA isozyme, showed that some hormones may selectively activate type I or I1 kinase. They correlated the stimulation of DNA synthesis by PGE, in the UMR 106 osteosarcoma cell line to the preferential activation of PKA I and the inhibition of proliferation in normal calvarial cells and in the human breast cancer cell line T47D to the specific activation of type I1 kinase by PGE, (Livesey et al., 1985; Livesey and Martin, 1988; Ng et al., 1983). In canine thyrocytes in primary culture, we more directly addressed the question of the involvement of PKA isozymes in the induction of DNA replication, using pairs of specific cAMP analogs that differentially modulate the activities of PKA I and 11. Before stimulation these cells contain comparable amounts of both kinase isozymes (Breton et al., 1989). DNA synthesis, as well as function and differentiation expression, present the same synergistic dependence on cAMP analogs as does the activation of PKA; this suggests that PKAs indeed mediate the cAMP action on these processes (Van Sande et al., 1989). Furthermore, in contrast to stimulation of function and differentiation expression, DNA replication seems more sensitive to PKA I activation (Van Sande et al., 1989). Consistently, the specific desensitization of this growth response (but not of the CAMP-dependent differentiation expression) is accompanied by the disappearance of PKA I, but not PKA 11, after several days of culture with TSH or forskolin (Breton et al., 1989; Roger et al., 1991). In the FRTL5 rat thyroid cell line it was also suggested that the TSH-regulated cell cycle progression involves mostly PKA I (Tortora et al., 1993).
98
PIERRE P. ROGER et al.
Thus, a predominant role of PKA I activation in the CAMPdependent stimulation of growth in several cell types is suggested. Recently, it was even reported that retroviral vector-mediated overexpression of RIa, but not RIIP, or C subunits of PKA enables a mammar y epithelial cell line to grow in a serum-free medium (Tortora et al., 1994). However, in some cells activation of PKA I instead inhibits proliferation [T lymphocytes (Skalhegg et al., 199211 or even leads to programmed cell death [myeloid leukemia cell line (Lanotte et al., 1991; Vintermyr et al., 1993b)l. Suggestions for specific roles of the subunits RI and RII of PKAs included different compartmentations, translocation or turnover of kinase isozymes (Nigg et al., 1985; Meinkoth et al., 1990; Weber and Hilz, 1986; Doskeland et al., 19931,or other functions of R subunits, unrelated to PKA activation, such as inhibition of protein phosphatases for RII (Khatra et al., 1985; Jurgensen et al., 1985; Vereb et al., 1986) or indirect activation of cGMP-dependent protein kinase for RI (Geahlen and Krebs, 1980). The work of Whitfield and collaborators on the peculiar model of T51B rat liver cells blocked in late G, by Ca2+ deprivation has been eventually explained by a clearly distinct feature of the involvement of PKA in cell cycle control. The early intriguing observation was the rapid initiation of DNA synthesis in such arrested cells by the addition of low concentrations of CAMP,which is membrane impermeant, as well as the finding of CAMPbinding sites at the outer face of the cell membrane (Boynton et al., 1985).These CAMPbinding sites are probably the R subunits of PKAs, since the ability of external CAMP to initiate DNA synthesis is mimicked by the addition of the C subunit of PKA and inhibited by the protein inhibitor of PKA. Kleine and Whitfield (1987) further showed that endogenous PKAs are accumulated at the outer surfaces of these cells during their progression into prereplicative phase initiated by serum factors. They proposed that Ca2+deprived cells fail to accumulate CAMP, thus preventing membrane PKA activation. CAMPwas presented as an autocrine G, progression factor with the external PKAs as its receptors (Kleine and Whitfield, 1987). This model clearly differs from the generally observed requirement for an elevation of intracellular CAMPin the CAMP-dependent cell proliferation. 2. Protein Phosphorylation Little is known about the nature and function of the proteins phosphorylated in response to the CAMPmitogenic stimulus. The fact that any interruption in CAMP signal introduces great delays in the onset of DNA synthesis in canine thyrocytes (Roger et al., 1987a) suggests
GROWTH MODULATION BY cAMP
99
that the phosphorylation of a specific protein(s) is necessary at the different stages of the prereplication phase. An 85-kDa CAMPdependent phosphoprotein was associated with late G, phase in B lymphocytes (Feuerstein et al., 1991). Various cytoskeleton proteins are phosphorylated in response to cAMP enhancers (see Section V), and the CAMP-dependent phosphorylation of vimentin was suggested to have a role in the stimulation of proliferation by cAMP in Swiss 3T3 cells (Escribano and Rozengurt, 1988). Upon dissociation of the cytoplasmic PKA I holoenzyme with CAMP, free C subunit progressively appears in the nucleus (Meinkoth et al., 1990) as a rate-limiting step for CAMP-dependent transcriptional induction (Hagiwara et al., 1993). Some transcription factors, such as the c-erbA-encoded thyroid hormone receptor (Goldberg et al., 19881, c-Fos (Abate et al., 1991; Tratner et al., 19921, and CREB (Yamamoto et al., 1988;Gonzalez and Montminy, 19891, are phosphorylated by PKA. The phosphorylation of c-Fos in its C terminus (either by PKC or by PKA) is required for transrepression activity of Fos on its own promoter and a mutation affecting this phosphorylation site enhances the c-fos transforming potential to a level comparable to that of v-fos (Tratner et al., 1992; Ofir et al., 1990). This phosphorylation could therefore be involved in the negative control of proliferation. By contrast, the CAMP-dependent phosphorylation of CREB and ATF-1 (Rehfuss et al., 1991) induces the transcription of genes containing CRE elements in their promoters. The induction of cell cycle-related proteins may thus take place a t the level of transcription and involve phosphorylation of CREB. Expression in the somatotroph of mutated nonphosphorylatable CREB indeed leads to pituitary atrophy and somatotroph depletion (Struthers et al., 1991). The same dominant negative mutant of CREB reduces, to some extent, PHIthymidine incorporation and proliferation in FRTL5 cells (Woloshin et al., 1992). PKA also phosphorylates the nuclear cyclin D1 at several sites, but the role of these phosphorylations remains to be determined (Sewing and Muller, 1994). 3. Gene Expression Balanced cell division requires duplication of all of the elements composing the cell. Therefore, in the preparation of cell division, all of the proteins should accumulate and be synthesized up to twice their initial level. It is not this aspect of protein metabolism that interests us in the mechanisms of cell proliferation, but rather the synthesized proteins that could represent signals or limiting factors in the whole process. a. Expression of Nuclear Protooncogenes. The protooncogenes of the
100
PIERRE P. ROGER et al.
myc, fos, and j u n families that encode nuclear phosphoproteins functioning as transcription factors play important roles in the regulation of cell growth, differentiation, and development (reviewed by Spencer and Groudine, 1991; De Pinho et al., 1991; Angel and Karin, 1991; Studzinski et al., 1991). Some of them, including c-myc, c-jun, and c-fos, pertain to the class of so-called immediate-early mitogenic genes, which are rapidly induced by growth factors and tumor promoters independently of protein synthesis (Herschman, 1991) during the mitogenic stimulation of a variety of cell types. They are also induced in nonmitogenic cascades (e.g., in the nervous system) and are therefore better called early-response genes. In fibroblast-like or leukemic cell lines antisense strategies or microinjections of neutralizing antibodies have suggested that c-myc (Spencer and Groudine, 1991; De Pinho et al., 1991),c-fos (Herschman, 19911, and thejun gene family (Kovary and Bravo, 1991) are necessary for the Go-S transition. While c-Fos and FosB activities are required mostly during the Go-G, transitions of Swiss 3T3 cells, Fra-1 and Fra-2 (Fos-related antigens) seem to be involved both in the Go-G, transition and in asynchronous growth (Kovary and Bravo, 1992). Fos and J u n proteins contain a leucine repeat that can mediate protein dimerization by the formation of the leucine zipper structure necessary for binding to DNA (Landschulz et al., 1988). The AP-1 transcription factor (Angel and Karin, 1991) is formed by a dimer of FosJ u n or Jun-Jun proteins. However, both Fos and Jun proteins may dimerize with other leucine zipper proteins. The c-Myc protein, whose C terminus contains a bHLH-zip, is also a DNA-binding transcription factor. The c-Myc protein dimerizes with another bHLH-zip protein, termed MAX, always present at significant levels in the cells, and the Myc-MAX complex, which is the active species of Myc, binds to DNA in a sequence-specific manner (Blackwood and Eisenman, 1991). Besides the function of these protooncogene proteins as transcription factors, it was suggested that c-Myc (Studzinski et al., 19911, and more recently c-Fos and c-Jun (AP-1 factor) (Murakami et al., 1991; Carter et al., 1991), may have more direct roles in DNA replication. CAMP-elevating factors induce c-fos mRNA transcription and accumulation in a wide variety of cell systems. In the FRTL5 thyroid cell lines (Colletta et al., 1986; Tramontano et al., 1986; Isozaki and Kohn, 1987) and in canine and human thyrocytes (Reuse et al., 1990; S. Reuse, unpublished observations) TSH, via CAMP,rapidly induces a transient elevation of c-fos mRNA. The effect of TSH in FRTL5 cells is transcriptional (Damante and Rapoport, 1988). Antisense c-Fos inhibits the TSH-dependent FRTL5 cell proliferation, suggesting that ex-
GROWTH MODULATION BY cAMP
101
pression of c-Fos is indeed necessary (Foti et al., 1990). Isoproterenol induces accumulation of c-fos mRNA and immunoreactive protein and acinar cell division in mouse submaxillary glands in uiuo (Barka et al., 1986). GRF induces c-fos expression and proliferation in rat somatotrophs (Billestrup et ul., 1987). cAMP also triggers c-fos expression and proliferation of H4IIE hepatoma cells (Squint0 et al., 1989). The expression of c-fos is rapidly induced in response to gonadotropins in rat ovarian granulosa cells (Delidow et al., 1990; Ness and Kasson, 1992) and porcine Leydig cells (Hall et al., 1991). 8-Br-CAMP is an inducer of c-fos in Swiss 3T3-Ll preadipocytes (Cornelius et al., 1991). In BALBc and Swiss 3T3 cell lines cAMP is a relatively poor inducer of c-fos mRNA and/or protein, but it potentiates the action of other CAMP-independent growth factors (Ran et al., 1986; Tsuda et al., 1986; Mehmet et al., 1988, 1990b; Mechta et al., 1989). In these different systems the CAMP-dependent induction of c-fos can therefore be part of a CAMP-dependent process of mitogenic activation. However, c-fos is similarly induced by cAMP in systems in which cAMP does not affect or even inhibit cell proliferation, including macrophages (Bravo et al., 1987b; Vairo et al., 19901,T lymphocytes (Farrar et al., 1987),PC12 pheochromocytoma cells (Greenberg et al., 19851,the R-5HT A5 epithelial cell line (Yeh et al., 19881, astrocytes (Gabellini et al., 19911, and NIH 3T3 cells (Fisch et al., 1989).The expression of c-fos has also been associated with CAMP-dependent induction of differentiation processes [e.g., in HL60 promyelocytic leukemia cells (Tsuda et al., 1987; Nakamura et al., 19901, a mastocytoma cell line (Goulding and Ralph, 1989), and the BC3H1 muscle cell line (Hu and Olson, 198811. Therefore, c-fos is more a gene of general early response to any stimulus than a specific mitogenic signal. The mechanism of the control of c-fos transcription in response to cAMP and other factors has been especially well studied (for reviews see Herschman, 1991; Angel and Karin, 1991). The major element responsive t o cAMP in the c-fos promoter is centered at position -60 and contains the sequence TGACGTTT, which, although different from the canonical CRE consensus (TGACGTCA), binds CREB (Foulkes et al., 1991b). CREB binds as a dimer to a CRE(s) and exerts its transcriptional regulatory function when phosphorylated by PKA. However, in corticotropic cells PKA regulates c-fos transcription by CREB-dependent and -independent mechanisms (Boutillier et al., 1992). Other sequences more upstream in the fos promoter (Fisch et al., 1989; Berkowitz et al., 1989) or in the transcribed region of the fos gene (Hartig et al., 1991) could also confer cAMP inducibility. A mechanism of down-regulation of the CAMP-dependent transcription of c-fos was
102
PIERRE P. ROGER et a1
recently proposed: the transcriptional antagonist CREM can bind to c-fos CRE and heterodimerize with activator CREB, thereby blocking cAMP induction (Foulkes et al., 1991b). The structure of the fosB gene is very similar to that of c-fos, including the CRE, SRE, and AP-1 elements in the 5’ upstream region, suggesting that c-fos and fosB are subjected to quite similar controls (Lazo et al., 1992). Data on the cAMP control of c-jun transcription are more disparate. In fact, in THP-1 monocytic leukemia cells, cAMP uncouples the transcription of c-fos and c-jun components of the AP-1 factor, c-jun being unresponsive to cAMP (Auwerx et al., 1990). It was proposed that the inhibition by cAMP of the induction by growth factors and TPA of c-jun transcription in HeLa cells, 3T3 cell lines, and Syrian hamster embryo cells could be related to growth-inhibitory effects of cAMP (Mechta et al., 1989; Angel et al., 1988; Janet et al., 1992; Cowlen and Eling, 1992). However, a similar inhibition of c-jun by cAMP was also observed during the mitogenic activation of canine thyrocytes (Reuse et al., 1991) and WRT rat thyroid cells (Tominaga et al., 1994) by TSH or forskolin. A similar inhibition of c-jun transcription was shown in rat Sertoli cells stimulated by FSH (Hamil et al., 1994). In porcine Leydig cells c-jun mRNA accumulation, unlike that of other protooncogenes, is not responsive to gonadotropins and cAMP (Hall et al., 1991). In contrast, there have also been some reports of a positive modulation of c-jun mRNA accumulation by CAMP.A delayed response was observed in 3T3-Ll preadipocytes (Cornelius et al., 19911, and a rapid induction was reported in PC12 cells (Wu et al., 1989), HL60 leukemia cells (Nakamura et al., 1990), and rat granulosa cells (Ness and Kasson, 1992). At variance with observations in canine thyrocytes, c-jun was reported to be precociously induced during the TSH and cAMP stimulation of proliferation in the thyroid FRTL6 cell line (Colletta and Cirafi, 1992). Unlike the case in c-fos, the c-jun promoter does not contain a CRE-like element (Angel et al., 1988).Nevertheless, in transfection experiments in 3T3 cells (Lamph et al., 19901, CREB can bind to the AP-1 site of c-jun and block the TPA-induced expression of a n exogenous c-jun promoter. This repression by CREB can be alleviated by its phosphorylation by PKA (Lamph et al., 1990). It remains unclear why PKA activation has opposite effects on the transcription from endogenous (Mechta et al., 1989) and transfected c-jun promoters in 3T3 cells. JunB and JunD proteins display strong homology in their C-terminal region to the DNA binding domain of c-Jun. They can form dimers with Fos proteins and recognize similar DNA sequences. However, JunB might inhibit the trans-activating and transforming properties
GROWTH MODULATION BY cAMP
103
of c-Jun (Chiu et al., 19891, and it could be involved in negative regulation of c-jun expression. Like c-fos, the transcription ofjunB appears to be induced by cAMP in all of the systems, irrespective of the effect of cAMP on growth. junB mRNA is rapidly and transiently accumulated in response to cAMP in BALB/c-3T3 cells (Mechta et al., 19891, 3T3L1 preadipocytes (Cornelius et al., 19911, porcine Leydig cells (Hall et al., 19911, adrenocortical cells (Viard et al., 19921, rat Sertoli cells (Hamil et al., 19941, and canine thyroid cells (Pirson and Dumont, 1994). The mouse j u d promoter contains a previously undescribed inverted repeat, which is necessary for cAMP induction, confers cAMP inducibility to heterologous promoter, and binds a 110-kDa protein (de Groot et al., 1991).junD is also observed to be positively regulated by cAMP in BALB/c-3T3 cells (Mechta et al., 1989) and canine thyrocytes (Reuse et al., 19911, but in some cells it seems to act as an inhibitor of proliferation and transformation (Pfarr et al., 1994). At present, the definitive influence on late gene expression of the different combinations of Fos and J u n proteins, the activities of which are also modulated by phosphorylation, is especially difficult to evaluate. For instance, as recently discussed (Tratner et al., 19921, PKA can be involved, directly or indirectly, in both activating and repressing Fos activity, either at the transcriptional level or in phosphorylation of the Fos protein involved in the down-regulation, or even by allowing its nuclear translocation (Roux et al., 1990). As detailed in Section II,B, an inhibition of c-myc expression is associated with the inhibition of growth by cAMP in several systems. It is thus interesting that, conversely, c-myc mRNA accumulation is rapidly induced in systems in which cAMP stimulates cell proliferation. In FRTL5 thyroid cells TSH, forskolin, and cAMP analogs trigger a marked increase in c-myc mRNA levels (Dere et al., 1985; Tramontano et al., 1986; Isozaki and Kohn, 1987). In canine (Reuse et al., 1986, 1990) and human thyrocytes (M. Taton, unpublished observations) the kinetics of CAMP-dependent c-myc expression, with a n early increase followed by a n abrupt decrease, are explained by a stimulation followed by a n active mechanism of termination of the myc response that is triggered by cAMP and depends on protein synthesis (Reuse et al., 1990). By contrast, in porcine thyrocytes TSH and cAMP are not mitogenic and do not enhance c-myc gene expression (Heldin and Westermark, 1988). cAMP triggers c-myc expression and DNA synthesis in H4IIE hepatoma cells (Squint0 et al., 1989). Although their mitogenic activity in uitro is controversial, gonadotropins, through CAMP,induce c-myc expression in ovarian granulosa cells (Delidow et al., 1990) and porcine Leydig cells (Hall et al., 1991). c-myc is potently induced by
104
PIERRE P. ROGER et al.
cAMP in Swiss 3T3 cells (Yamashita et al., 1986; Mehmet et al., 1988). In BALBc-3T3 cells, in which cAMP levels positively modulate the response to growth factor (Olashaw et al., 1984) and potentiate competence formation (Wharton et al., 19821, cholera toxin also potentiates the EGF effect on c-myc mRNA accumulation (Ran et al., 1986). Aside from the exception of the c-myc induction by cAMP in PC12 cells (Greenberg et al., 1985) and BC3H1 muscle cells (Hu and Olson, 19881, the c-myc protooncogene thus appears as the only “immediateearly gene” whose expression in response to cAMP is predictive of the effect of cAMP on growth in different normal cells. The CAMPdependent induction of c-myc could be necessary in the stimulation of proliferation by CAMP,while the CAMP-dependent inhibition of c-myc might suffice to explain a growth-inhibitory effect of CAMP. The mechanism involved in the dual effects of cAMP on c-myc mRNA transcription and/or stability is unknown. In canine thyrocytes inhibitors of protein synthesis potentiate the early induction by cAMP of c-myc mRNA accumulation, but prevent the late CAMP-dependent down-regulation (Reuse et al., 1990). However, the induction of c-myc expression is not sufficient to provoke CAMP-mediated induction of DNA synthesis, since it is too transient to explain the continuous requirement for high cAMP levels during progression into GI phase in canine thyrocytes (Roger et al., 1987a), and overexpression of transfected c-myc in thyroid cell lines does not diminish TSH dependence for growth (Fusco et al., 1987). b. Other Events. Although generally considered to be constitutively expressed, the CREB gene promoter contains three CRE elements, which might explain its up-regulation by FSH and cAMP in Sertoli cells (Meyer et al., 1993). LRF-1, a transcription factor induced early during the possibly CAMP-dependent liver regeneration, complexes with J u n proteins and activates CRE-containing promoters (Hsu et al., 1992). Other events associated with mitogenic stimulation include the induction of ODC activity and polyamine biosynthesis in canine thyrocytes (Mockel et al., 1980) in response to TSH or various cAMP enhancers, or in bronchial epithelial cells in response to a P-adrenergic comitogenic stimulation (Willey et al., 1985). The ODC mRNA accumulates 4 h after stimulation of FRTL5 thyroid cells by TSH or forskolin (Colletta and Cirafi, 1992). The promoter of the ODC gene contains a CRE (Fitzgerald and Flanagan, 1989). TSH, through CAMP, also transcriptionally induces the expression of hydroxymethylglutarylCoA reductase, a key enzyme of the synthesis of many isoprenoids required for cell proliferation (Grieco et al., 1990). Statin, a nuclear protein reported to specifically identify quiescent Go cells, rapidly dis-
GROWTH MODULATION BY cAMP
105
appears in rat thyroid gland stimulated by TSH (Bayer et al., 1992). The expression of a statin-related gene is also down-regulated in the rat parotid gland in response to the mitogenic stimulation by isoproterenol (Ann et al., 1991). An increase in ras protooncogene mRNA content was reported during TSH-induced G, phase progression in FRTL rat thyroid cells (Dere et al., 1986). In canine and human thyrocytes TSH, through CAMP,stimulates the synthesis of several proteins prior to stimulation of DNA synthesis, including PCNA/cyclin (Baptist et al., 1993; Lamy et al., 1989, 19901, the auxiliary protein required for DNA polymerase 6 activity. CDK2 is subjected to nuclear translocation and phosphorylation just before DNA synthesis initiation, and cyclin A and p34cdc2 progressively accumulate during S and G, phases in canine thyrocytes stimulated by TSH (Baptist et al., 1995). Therefore, distal to protein kinase activation, cAMP induces several of the well-known pleiotypic biochemical markers of cell progression to DNA replication. Thus far, almost all of the data concerning the biochemistry of cell cycle control by cAMP were obtained in highly differentiated thyroid epithelial cells and in the Swiss 3T3 fibroblast-like cell line. They confirm the kinetic data revealing an action at the major control points of G, phase progression. However, the biochemical description of the mitogenic cAMP pathways remains very sketchy. cAMP might also positively control the cell cycle or modulates the activity of other growth-promoting substances by quite diverse and perhaps unknown mechanisms.
E. CAMP-DEPENDENT AND -INDEPENDENT MITOGENICPATHWAYS In the late 1980s there was great interest in the realization that growth factors, acting through their tyrosine kinase receptors or through the activation of phospholipase C, separately trigger early events assumed to be important in mitogenesis (Chambard et al., 1987; Rozengurt, 1989). Thus, cell proliferation can be regulated by distinct mitogenic pathways. However, it is also manifest that these pathways interact and converge very early in the long path preceding the commitment for DNA replication (Fig. 1). Hence, it is often expected that the progression into the prereplicative phase induced by different growth factors involves a necessary sequence of events, which are common, since they are obligatory for the execution of the cell multiplication program. As in the previous sections, cAMP could act only as an accelerator of each of the tyrosine kinase or phospholipase C mitogenic pathways, or be able to activate its own mitogenic cascade.
106
PIERRE P. ROGER et al
1. Cross-signaling between CAMP-Dependent and -Independent Mitogenic Stimulations There are many possibilities of cross-signaling between the cAMP cascade and other mitogenic pathways. Among the proteins synthesized in response to CAMP,a certain number could function as relays in other mitogenic cascades. In bone cells PTH and cAMP induce the synthesis and secretion of IGF-I, a growth factor for these cells (McCarthy et al., 1990). FSH and hCG have the same effect in the testes (Naville et al., 1990) and in ovarian granulosa cells (Hsu and Hammond, 1987). Growth factor receptors are also induced by CAMP. For example, in porcine thyrocytes (in sharp contrast to what is found in canine, rat, and human cells) CAMP,on its own, is unable to trigger DNA synthesis or c-myc oncogene expression (Gartner et al., 1985; Heldin and Westermark, 1988). However, it provokes a n increase in the availability of EGF receptors and potentiates the mitogenic effect of EGF (Westermark et al., 1986). The promoter of the EGF receptor gene contains a CAMP-responsive enhancer (Hudson et al., 1990). IGF-I receptors (Adashi et al., 1986) and FGF receptors (Shikone et al., 1992) are induced by FSH and cAMP in rat granulosa cells. cAMP induces PDGF receptor mRNA and protein, as well as the PDGF response in rat Schwann cells (Weinmaster and Lemke, 1990). Low concentrations of gonadotropin (hCG) and 8-Br-CAMP enhance estrogen receptor mRNA levels in a Leydig cell line (Ree et al., 1990). The potentiation by cAMP of competence acquisition in BALB/c-3T3 cells is perhaps due to the fact that EGF can activate the phospholipase C cascade in the presence of high cellular CAMP,but not in its absence (Olashaw and Pledger, 1988). cAMP also potentiates the stimulation of phospholipase C by vasopressin in rat hepatocytes (Pittner and Fain, 1989). These effects are not due to the phosphorylation of phospholipase Cy by PKA (Olashaw et al., 1990). A recent report claimed that TSH up-regulates the expression of the “mitogenic” G i a l in human thyroid cells (Selzer et al., 1993). Moreover, agents enhancing cAMP levels stimulate the translocation of PKC to the nuclei of B lymphocytes (Cambier et al., 19871, where they could modulate gene expression. TSH and cAMP were reported to potentiate the IGF-I-dependent phosphorylation of proteins on tyrosine in FRTL5 thyroid cells (Takahashi et al., 1991). CAMP-dependent phosphorylations of p60src (Roth et al., 1983) or EGF receptor tyrosine kinase (Ghosh-Dastidar and Fox, 1984), or activation of tyrosine phosphatase (Brautigan and Pinault, 19911, provides other possibilities of modulation. Conversely, cAMP may also contribute to the growth response in-
GROWTH MODULATION BY cAMP
107
duced by PKC activators in Swiss 3T3 cells, since these potentiate the adenylate cyclase activation (Rozengurt et al., 19871, likely through direct phosphorylation of type I1 adenylate cyclase by PKC (Yoshimasa et al., 1987; Yoshimura and Cooper, 1993). In the same cell line the potent mitogenic activity of PDGF and bombesin is also partly mediated by the synthesis of PGE,, which, in turn, increases the cellular cAMP levels (Rozengurt et al., 1983b; Mehmet et al., 1990a). In a few systems EGF activates adenylate cyclase or potentiates its activation by other factors (Ball et al., 1990; Nair et al., 1990; Magnaldo et al., 1989a). In rat H4IIE hepatoma cells (Squint0 et al., 19891, primary rat hepatocytes (Skouteris and Kaser, 19911, and the rat mammary gland (Lavandero et al., 1990) this mechanism may, in part, explain the mitogenic activity of EGF. EGF stimulation of adenylate cyclase could be through G, activation (Nair et al., 1990) or synthesis of prostaglandins (Skouteris and Kaser, 1991). FGF also potentiates the stimulation of adenylate cyclase in CCL39 fibroblasts (Magnaldo et al., 1989a1, but displays the opposite effect in BALB/c-3T3 cells (Logan and Logan, 1991). Many examples of the interdependence of Ca2+ and cAMP also exist, with calcium activating adenylate cyclase (Whitfield et al., 1987) or cAMP inducing Ca2+ influx (Rasmussen and Barrett, 1984; Penner et al., 19881, or calcium decreasing cAMP levels by activation of the Ca2+/calmodulin-dependent phosphodiesterase (Dumont et al., 1984). Moreover, CAMP,via PKA activation, enhances the responsiveness to IP, of intracellular Ca2+ pools in guinea pig hepatocytes (Burgess et al., 19911, which could be mediated by the phosphorylation of the IP, receptor by PKA (Ferris et al., 1991). The transcription factor CREB is activated by phosphorylation not only by PKA but also by Ca"/ calmodulin-dependent protein kinase (Dash et al., 1991; Sheng et al., 19911, and the induction of c-fos transcription by Ca2+ is mediated by the CRE in its promoter (Sheng et al., 1990). It is now increasingly evident that CREB and CREM are activated by phosphorylation not only by PKA but also by several other kinases in response to growth factors (Ginty et al., 1994; de Groot et al., 1994). The signaling cascades of PKC and PKA also interact on the control of gene promoter activity in a n extremely complex manner. AP-2 transcription activators are responsive to both cAMP and PKC inducers (Imagawa et al., 1987). Moreover, PKC phosphorylates in uitro and stimulates the dimerization of CREB (Yamamoto et al., 1988). On the other hand, the Fos/Jun AP-1 factor induced by phorbol esters is negatively controlled by a n IP, factor, which itself is inactivated by CAMPmediated phosphorylation (Auwerx and Sassone-Corsi, 1991, 1992).
108
PIERRE P. ROGER et al.
Although Fos/Jun and ATF/CREB were initially thought to interact preferentially with different regulatory elements (the AP-1/TRE and ATF/CRE sites, respectively), several members of both transcription factor families form cross-family heterodimers with distinguishable DNA binding specificities (Hai and Curran, 1991; Lamb and McKnight, 1991). For example, while the dimer Fos/Jun preferentially binds to AP-1 sites, the CREBP2/Jun dimer binds to the CRE (Ivashkiv et al., 1990). Changes in the expression of CREBP2, c-fos, and c-jun by altering the ratio of CREBP2lc-Jun to c-Fos-c-Jun complexes, would thus affect the relative expression of PKA- and PKC-responsive genes. Moreover, the similarity between AP-1 and CRE sequences may itself involve an interplay in transcriptional regulation and “cross-signaling” between PKA and PKC pathways. Thus, c-jun efficiently transactivates CRE sequences and c-Fos and c-Jun efficiently bind and cooperate in activating CRE promoter elements (Sassone-Corsi et al., 1990; Ryseck and Bravo, 1991).Conversely, CREB binds to an AP-1 site in the c-jun promoter (Lamph et al., 1990). An even more complex situation has been reported for the CAMP-dependent regulation of the proenkephalin promoter at the CRE-2 site, which is mediated by binding of JunD but inhibited by JunB (Kobierski et al., 1991). These examples provide support for combinatorial models of gene regulation, whereby protein-protein interactions, which alter the DNA binding specificity of protein complexes, can expand the flexibility of cellular transcriptional responses (Ivashkiv et al., 1990).However, these different studies performed in vitro or using cotransfection models should be interpreted cautiously. It remains to be established that such interplays of transcription factors and regulatory elements are also functioning on the regulation of endogenous genes at the concentrations of transcription factors normally found in intact cells. The extremely various possibilities of cross-signaling between the different levels of the few major signal transduction pathways and their combinations are essentially cell type specific. Their diversity could greatly contribute to providing the cell type specificity of proliferation controls. 2. cAMP Can Induce Its Own Mitogenic Pathway Which is Quite Distinct from the CAMP-Independent Ones At variance with these various possibilities of positive interaction and intervention of cAMP in tyrosine kinase and phospholipase C pathways, a defect in cAMP phosphodiesterase, causing elevated cellular cAMP levels in a variant of MDCK cells, abolishes the growth requirement for PGE, but does not affect the mitogenic dependence on
GROWTH MODULATION BY cAMP
109
EGF (Taub et al., 1983, 1984). Other indications for the coexistence of self-sufficient CAMP-dependent and -independent stimulations of proliferation have been obtained in keratinocytes (Tongand Marcelo, 1983), mammary epithelial cells (Imagawa et al., 19881, and differentiated epithelial thyroid cells and Swiss 3T3 fibroblast-like cells. Comparative studies of the mitogenic activation through both types of mechanisms were performed only in these latter systems. In primarily cultured canine thyrocytes, which are quiescent in an insulin-supplemented serum-free medium, DNA replication is induced not only by TSH, via cAMP and PKA activation, but also by EGF and phorbol esters (Roger and Dumont, 1984;Roger et al., 1986,198713).The stimulation of DNA synthesis by EGF requires the comitogenic action of high insulin concentrations that activate IGF-I receptors, while very low insulin concentrations are sufficient to permit the growth response to TSH (Roger et al., 1987b).In the presence of insulin, EGF synergizes with cAMP enhancers to induce DNA synthesis, but its effects are not additive to those of phorbol esters (Roger et al., 1986,198713).This was a first indication that EGF and phorbol ester act through a somewhat common mechanism that is different from the mode of action of cAMP in thyrocyte mitogenesis. TSH and cAMP enhancers at mitogenic concentrations do not activate the PI cascade (Graff et al., 1987; Berman et al., 1987; Raspe et al., 1992)or protein tyrosine phosphorylation (Contor et al., 1988; F. Lamy, unpublished observations). The activation of the PI-calcium cascade by acetylcholine decreases cAMP levels (Dumont et al., 1984). Phorbol esters that are potent activators of PKC do not increase cAMP levels (Mockel et al., 1987; Roger et al., 1991). EGF, which activates protein tyrosine phosphorylation, does not modify cAMP levels and does not activate the PI cascade (Raspe et al., 1992). Thus, despite the great number of possible interactions between different second messenger systems, the primary effects of mitogenic agents acting on the three pathways are distinct at the level of intracellular signals in canine thyrocytes. In the FRTL5 rat thyroid cell line phorbol esters and IL-1 are also mitogenic without increasing cAMP levels, in contrast to TSH (Lombardi et al., 1988; Mine et al., 1987). Similarly, in the Swiss 3T3 cell line increasing cellular cAMP levels neither activates PKC nor mobilizes calcium, releases arachidonic acid (as a result of phospholipase A, activation), or increases Na+ influx (Paris and Rozengurt, 1982) and cellular pH (Hesketh et al., 1988), whereas mitogens that activate tyrosine kinase receptors (EGF)or PKC do not increase basal CAMP levels (Rozengurt and Mendoza, 1985).None of these events is thus obligatory for mitogenic activation. In these cells inhibition of the CAMP-PKA pathway by genetic expression of a mu-
110
PIERRE P. ROGER et al
tated subunit of PKA or yeast CAMP phosphodiesterase affects the CAMP-dependent mitogenic signaling, but not the activity of CAMPindependent growth factors (N. Huang et al., 1994). Rozengurt and Mendoza (1985) also suggested that in Swiss 3T3 cells the CAMPdependent pathway is clearly separate from other mitogenic pathways with which it acts synergistically. This conclusion was extended in canine thyrocytes by the analysis of the phosphorylation of intact cell proteins separated by two-dimensional gel electrophoresis, which gives a high-resolution picture of the activation state of protein kinase-dependent regulatory networks (Contor et al., 1988). As in fibroblasts (Cooper et al., 1984; Kohno, 19851, EGF rapidly induces the phosphorylation of five proteins, including the phosphorylation on serine of the 28-kDa heat-shock protein (L. Contor, unpublished observations) and the phosphorylation of the two related 42- and 44-kDa MAP kinases on tyrosine, threonine, and/or serine residues (Lamy et al., 1993). Identical activating phosphorylations of MAP kinases are a common response to all mitogenic factors examined so far by several groups in fibroblasts (Cooper et al., 1984; Nakamura et al., 1983; Rossomando et al., 1989). They are also phosphorylated during MPF activation of meiosis in Xenopus oocytes (Cooper, 1989; Lohka et al., 1987; Posoda and Cooper, 1992). MAP kinase substrates include microtubule-associated protein 11, the S6 kinase I1 (RSK), and other proteins on serine and threonine. S6 kinase 11, in turn, via the phosphorylation of the ribosomal protein S6, could play a major role in quantitative and qualitative changes in protein synthesis associated with cell cycle progression (Rossomando et al., 1989; Thomas, 1992). It also phosphorylates lamin C, which may play a role in the nuclear breakdown at mitosis (Ward and Kirschner, 1990). Moreover, MAP kinases are translocated to the nucleus (Lamy et al., 1993; Chen et al., 1992; Lenormand et al., 1993) and phosphorylate two serine residues in the N-terminal A 1 transactivation domain of c-Jun and a serine in the N-terminal domain of c-Myc, which positively regulates their transacting activities (Pulver et al., 1991; Alvarez et al., 1991).Activation of MAP kinases by phosphorylation is therefore assumed to play major roles in cell cycle control (Lenormand et al., 1993). Phorbol esters via PKC activation also cause MAP kinase phosphorylation and nuclear translocation in canine thyrocytes (Lamy et al., 19831, although they do not activate the EGF receptor kinase (Davis and Czech, 1985; F. Lamy, unpublished observations). In addition, they induce the phosphorylation of another set of proteins that could be related to the acute functional and morphological effects specific to PKC activation (Contor et al., 1988). This strongly suggested that tyrosine kinase- and
GROWTH MODULATION BY cAMP
111
PKC-dependent mitogenic pathways converge early on in a cascade of protein kinases and common MAP kinase phosphorylations. In sharp contrast, TSH, the most potent mitogen in canine thyrocytes, induces, via CAMP,the phosphorylation of a completely different set of proteins that includes neither the 42- and 44-kDa MAP kinases, the 28-kDa heat-shock protein, nor any phosphorylation on tyrosine residues (Contor et al., 1988; Lamy et al., 1993; F. Lamy, unpublished observations). Moreover, MAP kinases are not translocated to the nucleus in response to cAMP (Lamy et al., 1993). None of these events is thus necessary in the CAMP-dependent mitogenic pathway. Our data, based on 32P incorporation in NaOH-washed twodimensional gels and Western blots with antiphosphotyrosine and anti-MAP kinase antibodies from primary canine thyrocytes, differ from a recent report of a stimulation by TSH of tyrosine phosphorylations in the FRTL5 rat thyroid cell line (Takahashi et al., 1991). It was recently claimed that the activation of ras could be involved in the CAMP-dependent mitogenic pathway, since microinjection of a rasdominant negative mutant inhibits, in part, the induction of DNA synthesis by TSH and cAMP in the WRT rat thyroid cell line (Kupperman et al., 1993). We believe that this inhibition instead bears on the comitogenic (permissive) insulin-IGF-I signaling pathway of such cells (Brandi et al., 19871, which could even be activated through autocrine mechanisms. In canine thyrocytes the complete absence of MAP kinase phosphorylation in the cAMP pathway rules out any involvement of ras, since it is activated upstream of MAP kinases in signaling cascades. In Swiss 3T3 cells Rozengurt and collaborators also did not find any overlap between markers of PKC and PKA activation during mitogenic stimulation via these different pathways (Escribano and Rozengurt, 1988). The expression of c-fos and c-jun, encoding components of the AP-1 transcription factor, as well as c-myc is also generally considered a common necessary response in all of the mitogenic activation processes. Indeed, c-fos and c-myc mRNAs rapidly and transiently accumulate in response to both CAMP-dependent mitogens, as well as CAMP-independent factors such as EGF, phorbol esters, and serum. This was observed in Swiss 3T3 cells (Tsuda et al., 1986; Mehmet et al., 1988; Yamashita et al., 19861, FRTL5 rat thyroid cells (Tramontano et al., 1986; Isozaki and Kohn, 19871,primary canine thyrocytes (Reuse et al., 1986, 19901, rat hepatoma cells (Squint0 et al., 19891, and porcine Leydig cells (Hall et al., 1991).However, in Swiss 3T3 cells, while c-myc is similarly induced by cAMP and PKC activators, the c-fos response to cAMP is only 5-10% of c-fos mRNA levels reached after stimulation
112
PIERRE P. ROGER et al.
by PKC (Mehmet et al., 1988). In this system a large c-fos induction is not necessary in the CAMP-mediated mitogenesis. In comparison to PKC activators, the c-fos response to cAMP is also weaker in canine thyrocytes (Reuse et al., 1990; M. Baptist, unpublished observations), porcine Leydig cells (Hall et al., 19911, and adrenocortical cells (Viard et al., 1992). The c-fos gene transcription in response to CAMP-dependent or -independent mitogens involves independent mechanisms. (1)cAMP and EGF are synergistic on c-fos mRNA accumulation in several systems (Reuse et al., 1990; Hall et al., 1991; Mehmet et al., 1990b). (2) c-fos inductions in response to a number of CAMP-independent ligands (serum, TPA, EGF, FGF, NGF, and PDGF) converge at the SRE site of the c-fos promoter and involve the phosphorylation of the of the transcription factor P62TCF by MAP kinases (Gille et al., 1992), which are not activated in the cAMP cascade (Lamy et al., 1993).Mutations in the SRE abolish c-fos induction by these factors (for a review see Herschman, 1991), but not c-fos induction by agents that elevate cAMP or Ca2+, which act on the CRE at position -60. (3) The stimulation of c-fos transcription by PKC and tyrosine kinase activators is rapidly repressed by a retrocontrol mechanism involving the SRE and the c-Fos protein. On the contrary, more stable kinetics of c-fos expression were observed in response to cAMP (Bravo et al., 198713; Reuse et al., 1990), and c-Fos is unable to down-regulate the CAMP-dependent transcription of c-fos (Foulkes et al., 1991b). (4) The CREM protein specifically down-regulates the induction of c-fos mediated by the CRE but not the SRE (Foulkes et al., 1991b).Thus, the mechanisms involved in both the induction and the down-regulation of c-fos transcription in response to CAMP-dependent and -independent ligands are completely distinct. This suggests that CAMP-dependent or -independent mitogenic pathways separately control events that could be necessary for growth stimulation. In canine thyrocytes TPA and EGF increase the mRNA levels of c-jun andjunD. The effects are especially well observed in the presence of cycloheximide. Increasing cellular concentrations of cAMP by TSH or forskolin also stimulate junD expression, but inhibit c-jun expression (Reuse et al., 1991). Similarly, the TPA or EGF stimulation of c-jun expression is inhibited by cAMP (Reuse et al., 19911, as in fibroblasts in which cAMP inhibits proliferation (Mechta et al., 1989). The inhibition by cAMP of c-jun expression was confirmed in WRT rat thyroid cells (Tominaga et al., 1994). CAMP,unlike PKC activators, also fails to induce c-jun in porcine Leydig cells (Hall et al., 1991) and
GROWTH MODULATION BY cAMP
113
in ovine and bovine adrenocortical cells (Viard et al., 1992).In contrast to what has been postulated in fibroblasts, the expression of c-jun is not universally correlated with the stimulation of cell proliferation. It is not necessary in the CAMP-dependent mitogenic pathway, and thus represents another of the very different responses elicited by the cAMP cascade as compared to the other mitogenic pathways in canine thyrocytes. c-Jun is a substrate of MAP kinases that are activated in the EGF and PKC cascades, but it is not phosphorylated by PKA (Baker et al., 1992). In the CAMP-dependent mitogenic pathway in canine thyrocytes, it is tempting to connect the absence of effect on c-jun mRNA and presumably on c-Jun phosphorylation, both involved in AP-1 factor activity. This suggests a poor AP-1 activity in the cAMP pathway, and indeed there is some evidence that AP-1 activity is even repressed by cAMP in WRT thyroid cells (Tominaga et al., 1994).CREB and AP-1 activities depend on a common coactivator, CBP, the intracellular level of which might well be limiting (Arias et al., 1994). This may help to explain the specificity of, or even the competition between, the transcriptional responses to cAMP or CAMP-independentmitogens, despite the multiple possibilities of positive cross-signaling at this level (see Section III,E,l). Considering the opposite effects of cAMP and EGF or TPA on differentiation in thyroid cells (Roger et al., 1985, 19861, it is interesting that high cellular AP-1 activity can inhibit epithelial cell differentiation (Jehn et al., 1992) and that c-Jun alone could also function as a repressor of differentiation, as exemplified by its direct interaction with MyoD, inhibiting myogenic differentiation (Bengal et al., 1992). In primary canine thyrocytes the kinetics of c-myc mRNA accumulation are very different in response to either TSH or forskolin, or in response to EGF or TPA, again suggesting different mechanisms of control. c-myc mRNA levels are still enhanced over basal levels 9 h after EGF or TPA stimulation. By contrast, after the cAMP stimulation c-myc expression is biphasic, with an enhancement at 1h followed by a rapid down-regulation (Reuse et al., 1990). The effects of EGF and cAMP on c-myc are additive at 1h, in correlation with additive effects on DNA synthesis, but at 3 h cAMP even markedly inhibits the stimulation of c-myc expression by EGF (Reuse et al., 1990).The early stimulatory effects of cAMP and EGF are independent of protein synthesis, but the secondary inhibitory effect of cAMP is blocked by the protein synthesis inhibitor cycloheximide (Reuse et al., 1990). We hypothesize that the biphasic character of the cAMP action on c-myc is related to
114
PIERRE P. ROGER et al.
the opposite requirements of the stimulation of both proliferation and differentiation expression by CAMP, whereas EGF and TPA induce mitogenesis but inhibit differentiation. The expression of MAX (the c-Myc heterodimeric partner) is also subjected to some modulation in canine thyrocytes. EGF and TPA, but not TSH, produce a delayed increase in max mRNA content, which could modify the cell sensitivity to c-Myc as a transcription factor (Pirson et al., 1994). Uncoupling of c-fos and c-jun expression and transient c-myc expression in the cAMP cascade implies that the pattern of late gene expression could be different in the CAMP-dependent and -independent mitogenic pathways. Indeed, the patterns of protein synthesis (as analyzed by two-dimensional gel electrophoresis) induced by activation of canine thyrocytes by TSH and CAMP,on the one hand, and EGF, phorbol esters, and serum, on the other, are entirely different during G, phase progression (Lamy et al., 1986, 1989). The transient induction of the synthesis of a n 80-kDa protein in mid-G, constitutes a marker of G, phase progression stimulated by CAMP-independent mechanisms (EGF, serum, and phorbol esters), but not of the activation by cAMP (Lamy et al., 1986, 1989). The synthesis and nuclear accumulation of PCNA/cyclin, the auxiliary protein required for DNA polymerase 6 activity (Prelich et al., 1987; Bravo et al., 1987a), are already stimulated by TSH via cAMP in the second part of the prereplicative phase, but they are detected only just before the onset of DNA synthesis phase in response to CAMP-independent mitogenic treatments (Lamy et al., 1989; Baptist et al., 1993). Kinetic experiments suggested that the accumulation of PCNA to threshold levels could be a rate-limiting event for the initiation of DNA replication in the case of CAMPindependent mitogenic stimulation, but not in the case of TSH (Baptist et al., 1993). Nevertheless, as expected, CAMP-dependent and -independent mitogenic pathways converge before G1-S transition a t the late stage of proteins that control the cell cycle machinery. Convergence includes common changes of subcellular localization and phosphorylation of CDK2 and cdc2, and common induction of cyclin A and cdc2 expression (Baptist et al., 1995). Altogether, these studies of discrete biochemical events suggest that the CAMP-dependent and -independent mitogenic pathways may remain partly separated throughout the prereplicative phase. Kinetic experiments on the onset of DNA synthesis in canine thyrocytes also support this conclusion (Roger et al., 1987a,b). Although a late elevation in cAMP is still required for the commitment of DNA synthesis of cells already in late G, in response to cAMP (Fig. 2A), it is without any
115
GROWTH MODULATION BY cAMP
B
f
1: I
. lo
0
% r
X
z q 10
z 0
52
a: 0 0
zw 5 z 0
FI k
H 0
8
16
24
HOURS WITH FORSKOLIN
36
48
0
12
24 36 48 TIME (HOURS)
60
72
FIG. 2. (A) Time dependence of commitment to DNA synthesis in quiescent canine thyrocytes exposed to forskolin. Canine thyrocytes cultured in a serum-free medium supplemented with insulin were given forskolin (10-6M ) and [3Hlthymidine. At the time intervals indicated on the abscissa, cultures were washed free of forskolin. The cumulative labeling index was determined after continuous labeling for 48 h. (Inset)The kinetics of the accumulation of [3Hlthymidine-labeled nuclei after a continuous incubation with forskolin (10-6 M ) are shown for comparison. The experiment shows that in the CAMP-dependent mitogenic pathway, late rate-limiting events for the onset of DNA synthesis are dependent on continuous activation of adenylate cyclase. cAMP acts on late prereplicative events. (From Roger et al., 1987a.) (B) Kinetics of the cooperation of EGF and forskolin (FORSK) a t suboptimal concentrations on the stimulation of DNA synthesis in quiescent canine thyrocytes. Canine thyrocytes cultured in a serum-free medium supplemented with insulin received the following additions: (3, 1.6 F M forskolin a t time 0; 0 , 1 . 6 ng/ml EGF at time 0; and 0 , 1 . 6 W Mforskolin and 1.6 ngiml EGF added together a t time 0, or 1.6 (IM forskolin added a t 24 h to cells incubated since time 0 with EGF, or 1.6 ngiml EGF added a t 24 h to cells incubated since time 0 with forskolin. Kinetics of the accumulation of PHIthymidine into acid-insoluble material were determined as done by Roger et al. (1987b). When either forskolin or EGF is added 24 h after the other factor, their synergy on DNA synthesis is delayed by a time (17 h) that corresponds exactly to the duration of the prereplicative phase. The same 17-h lag time is thus necessary for the effect of each factor alone and for the establishment of its synergy with the other factor. Even though forskolin acts on late prereplicative events in the CAMP-dependent mitogenic pathway (A), it cannot interact with the late prereplica(continued)
116
PIERRE P. ROGER et al.
10 h Forsk
12
24
36
40
TIME (HOURS) tive development controlled by EGF without having executed its own mitogenic sequence. (From Roger et al., 198713) (C) Influence of a forskolin pretreatment on the kinetics of DNA synthesis induced by EGF in canine thyrocytes. Quiescent canine thyrocytes in a serum-free medium supplemented with insulin were pretreated (*, x ) or not (0,O) with 1 p M forskolin (Forsk)for 10 h, before stimulation at time 0 with EGF in the absence of forskolin (*, 0).Kinetics of the accumulation of [SHIthymidine into acidinsoluble material were determined as done by Roger et al. (198713).The preincubation of cells for 10 h with forskolin [a treatment too short to induce DNA synthesis by itself (A)] does not reduce the lag time for initiation of DNA synthesis stimulated by EGF. Thus, a 10-h progression into prereplicative phase controlled by cAMP is not helpful in the EGFdependent mitogenic cascade. (D) Phenomenological model of cooperation of distinct CAMP-dependent and -independent mitogenic cascades in canine thyrocytes. The triggering of S phase (S) in dog thyrocytes could require the induction of necessary early events such as c-fos and c-myc expression and the execution of a sequence of regulatory events. The sequence induced by TSH vis cAMP is partially distinct during the major part of the prereplicative phase from the sequence common to EGF and TPA actions. Examples of X early events involved in the EGFiTPA sequence but not in CAMPdependent mitogenesis are the activation of MAP kinases and the induction of c-jun.
117
GROWTH MODULATION BY cAMP
D
TS H
1
GAMP
0
20h
Examples of common steps are the early induction of c-myc and the induction of c-fos and jurzD. cAMP controls its sequence a t both early and late stages [Roger et al. (1987a); (A)]. Some cooperation between events of the separate CAMP-dependent and -independent sequences is demonstrated by their synergy, but it is possible only when both sequences have progressed to similar stages, as suggested by the kinetics of the synergism of EGF and forskolin on DNA synthesis [Roger et al. (1987b); (B)]. For example, the stimulation of the synthesis of a 80-kDa protein by EGF, which is potentiated by TSH (Lamy et al., 1989), has the characteristics of a Y event. The stimulation ofthe synthesis of PCNA in GI by TSH, which is potentiated by EGF (Lamy et al., 19891, has the characteristics of a C event. The coexistence of these partially separate mitogenic sequences, together with the fact that TSH only very transiently induces c-myc expression, could explain how TSH could stimulate growth and differentiation as compatible programs, a t the difference of EGF and TPA which stimulate growth while inhibiting differentiation. Such a working model integrates the presently known data on the phenomenology and kinetics of canine thyrocytes proliferation. It should help in the design of experiments to define the intermediate steps of both proliferation cascades.
rapid potentiating effect on the rate of entry into the DNA synthesis phase of cells already in late GI in response t o EGF (Fig. 2B). Late GI cells are therefore different, depending on whether they receive mitogenic stimulation by EGF or by CAMP.In fact, the experiment in Fig. 2B is not compatible with models in which early events triggered by either cAMP or EGF synergistically interact with later rate-limiting events dependent on the other factor. For example, a rapid CAMPdependent phosphorylation could not increase the activity of a tardily induced protein, on which would depend the rate of DNA synthesis in
118
PIERRE P. ROGER et al
the mitogenic sequence triggered by EGF. Furthermore, the CAMPdependent progression into prereplicative phase induced by a 10-h pretreatment with forskolin is without any influence on the kinetics of DNA synthesis induction by EGF (Fig. 2C). CAMP-dependent mitogenic events (even those that are also induced by EGF, such as c-fos and c-myc expressions) are not helpful in the sequence stimulated by EGF; they do not correspond to the late rate-limiting events of the EGF-induced prereplicative development. Therefore, distinct ordered sequences of mitogenic events stimulated by either EGF or cAMP should be self-sufficient and, in parallel, lead to commitment for DNA replication (Fig. 2D). In the rat thyroid gland in uiuo, an irreversible desensitization mechanism limits the TSH-dependent growth, but not the growth induced by tissue wounding (Wynford-Thomaset al., 1983; Smith et al., 1987). In canine thyrocytes, during the CAMP-independent cell division, a dominant repressor factor is generated that specifically inhibits the TSH (CAMP)-dependent mitogenic pathway but not the growth response to EGF (Roger et al., 1992). Similarly, TGF-p strongly inhibits the CAMP-dependent proliferation of these cells, but weakly affects the CAMP-independent mitogenesis elicited by various factors. (F. De Poortere and P. P. Roger, unpublished observations). This raises the concept that growth suppressor mechanisms could specifically affect the CAMP-dependentproliferation. These results also differentiate the cAMP and the growth factor mitogenic cascades.
IV. RELATIONSHIP BETWEEN GROWTH AND DIFFERENTIATION CONTROLS BY cAMP cAMP plays a major mediator role in the action of many hormones that stimulate or induce specialized functions related to differentiation in their target cells (Friedman, 1976). Such a situation is very frequent, as exemplified by the stimulation of melanogenesis in melanocytes, of steroidogenesis in adrenocortical cells or ovarian granulosa cells, and of thyroid hormone synthesis and secretion in thyroid follicular cells. In all of these cases, cAMP is a positive modulator of the function and “expression” of differentiation by acting via various mechanisms on the tissue-specific transcription of genes encoding proteins responsible for specialized functions. It is by far less clear that cAMP could also affect the process of “determination,” that is, the irreversible commitment of undifferentiated cells to become specialized cells. cAMP directs the differentiation of F9 embryonal carci-
GROWTH MODULATION BY cAMP
119
noma cells toward a parietal endoderm phenotype (Strickland, 1981)) and it participates in the induction of the adipose conversion of 3T3L1 fibroblasts. However, in most cases cAMP seems to be unable to change the tissue specificity of the expression of genes, but it has a very widespread role in maintaining the full expression of differentiated properties conferred during determination. According to a n old but still current view, growth and differentiation are often considered mutually exclusive processes in cell life. Terminal differentiation processes involve a n irreversible cessation of growth. In addition, several examples of initially highly differentiated cell types present a temporal loss of differentiation expression during the proliferative phase. This was reported very early for iris pigmented cells (Doljanski, 1930) and more recently for retinal pigment cells (Rodesch, 19731, hepatocytes (Leffert et al., 1978; T. Y. Nakamura et al., 1984; Nakamura and Ichihara, 19851,hepatoma cells (T. Nakamura et al., 19841, thyroid medullary carcinoma cells (Berger et al., 1984; de Bustros et al., 1986))ovarian granulosa cells (Epstein-Almog and Orly, 19851, and prostatic epithelial cells (Chevalier et al., 1981). In many tumor cell lines in which cAMP inhibits cell proliferation, it stimulates the expression of differentiation, thus restoring a n apparent normal phenotype (Pastan and Johnson, 1974; discussed more thoroughly in Sections V and VI). Conversely, growth factors (e.g., EGF) acting through tyrosine kinase receptors or tumor promoters acting through PKC are potent inhibitors of differentiation expression in a variety of systems, including mammary epithelial cells (Taketani and Oka, 19831, granulosa cells (Ascoli, 19811, Leydig cells (Welsh et al., 19841, adrenocortical cells (Simonian et al., 19821, keratinocytes, and thyroid epithelial cells (Roger et al., 1985, 1986; Bachrach et al., 1985). However, in the hematopoietic system hormones such as erythropoietin and CSF can induce both proliferation and differentiation. Hormones that stimulate growth through cAMP in normal specialized cells also often stimulate the expression of their differentiation. A nonexhaustive list of examples includes hepatocytes (Miyazaki et al., 1992; van Roon et al., 19891, Schwann cells (Sobue et al., 19861, keratinocytes (Tong and Marcelo, 19831, melanocytes (Halaban, 1988; Abdel-Malek et al., 19921, pituitary somatotrophs (Billestrup et al., 19861, MDCK epithelial cells (Lever, 1979; Taub et al., 19831, and thyroid epithelial cells (Roger and Dumont, 1984; Roger et al., 198813).The apparent paradox between the antagonism of growth and differentiation and the stimulation of both processes by cAMP has been addressed in canine thyroid epithelial cells in primary culture. While EGF and TPA stimulate proliferation and inhibit differentiation (in-
120
PIERRE P. ROGER et al.
cluding thyroglobulin gene expression), TSH, through CAMP, induces both processes (Roger and Dumont, 1984; Roger et al., 1985, 1986, 198813). Study of the induction of thyroglobulin and thyroperoxidase mRNA accumulations by in situ hybridization demonstrates that they are fully compatible with cell cycle progression when they are promoted by CAMP.Thus, cAMP can induce mitosis and differentiation expression at the same time, in the same cells (Pohl et al., 1990, 1993). Both the differentiation induced by TSH and cAMP and the dedifferentiation induced by EGF or phorbol esters are independent of the effects of these agents and cascades on proliferation, as they are obtained in the absence of insulin (which is required for proliferation) or in cells at confluence that do not proliferate anymore (Roger and Dumont, 1984). A molecular mechanism of the dissociation of these two types of effects, even when mediated by the same transcription factor, is suggested by the case of MyoD, in which separate domains of the protein are involved in the induction of differentiation and in the inhibition of proliferation (Crescenzi et al., 1990). Suppression of differentiation expression in thyrocytes therefore is not a consequence of cell cycling itself, but could be related to the activation of some mitogenic signaling cascades. Considering the evidence of the coexistence of largely separated controls of cell cycle progression in thyrocytes, it appears quite plausible that the CAMPdependent mitogenic pathway, by opposition to the CAMP-independent ones activated by EGF and TPA, might have characteristics that make it compatible with differentiation expression (Pohl et al., 1990; Dumont et aZ., 19891, such as a poor AP-1 activity, as discussed in Section III,E,2. Another clue to such a possibility is the observation that the c-myc response to EGF and TPA in canine thyrocytes is a relatively stable process, while its response to TSH and cAMP is very transient, due to a specific CAMP-dependent inhibitory mechanism (Reuse et al., 1990). A more stable induction of c-myc by cAMP is associated with the inhibition of the differentiation of the mouse muscle cell line BC3Hl (Hu and Olson, 1988). c-myc protooncogene expression prevents differentiation in most systems (reviewed by Dang, 19911, in addition to appearing to be a prerequisite for growth stimulation. Therefore, the rapid termination of c-myc induction in canine thyrocytes stimulated by cAMP might be necessary in order not to interfere with the maintenance of differentiation expression that occurs together with growth stimulation (Reuse et al., 1990; Pohl et aZ., 1990). At variance with the case of thyrocytes, the synthesis of myelin proteins and late differentiation responses in postnatal rat Schwann cells are induced by cAMP only in the fraction of nonproliferating cells
GROWTH MODULATION BY cAMP
121
or in conditions that do not allow cell division (Morgan et al., 1991). Stimulation of proliferation by cAMP and CAMP-dependent induction of late differentiation markers seem to be incompatible in this system. However, in Schwann cells, contrary to what we have shown in thyrocytes, it is not clear that cAMP can promote proliferation by acting on a separate cascade. The mitogenic effect of cAMP is dependent on other factors, such as PDGF, and cAMP induces the appearance of PDGF receptors and PDGF responsiveness in Schwann cells (Weinmaster and Lemke, 1990). Differentiation processes generally include modifications of growth characteristics. Various models of cell differentiation consider the concept of two kinds of cell cycle, either preserving or changing differentiation phenotypes (Lajtha, 1979; Yamada, 1989). The adipose conversion of 3T3-Ll fibroblasts is often considered a differentiation process that needs postconfluence cell cycle progression (Kuri-Harcush and Marsh-Moreno, 1983; Schmidt et al., 1990).According to Schmidt et al. (19901, the promoting effects of CAMP on postconfluent mitoses and on differentiation cannot be dissociated and are both potentiated by EGF together with IGF-I. However, the mitogenic effects of FGF and PDGF, which are not potentiated by CAMP,are not associated with adipose conversion (Schmidt et al., 1990). The CAMP-dependent and -independent cell cycles could thus differently modify the differentiated phenotype of 3T3-Ll cells, only the former inducing the adipose conversion. Also, in the canine thyrocyte experimental model CAMP-dependent and -independent cell cycles are not equivalent for the maintenance of the cell phenotype. While the TSH (CAMP)-dependentdivision of thyrocytes preserves their responsiveness to both TSH (CAMP)and EGF mitogenic pathways, cells that have divided in response to CAMPindependent stimuli (EGF and serum) lose mitogenic sensitivity to TSH and cAMP but retain sensitivity to EGF (Roger et al., 1992). Cell fusion experiments suggest that the extinction of the CAMP-dependent mitogenic pathway is due to the induction during the CAMP-independent cell cycle of a specific diffusible intracellular inhibitor (Roger et al., 1992). Differentiated functions are generally suppressed in heterokaryons or somatic hybrids formed by the fusion of a cell expressing this function with a nonexpressing cell (Harris, 1990). The TSH (CAMP)dependent mitogenic pathway, which is repressed in cell fusion experiments (unlike the EGF mitogenic pathway) (Roger et al., 19921, could thus be analyzed as a differentiated trait of thyrocytes. Together with differentiation markers, the TSH (CAMP)growth pathway is also specifically lost in somatic hybrids of FRTL5 thyroid cells and BRL rat
122
PIERRE P. ROGER et al.
liver cells (Veneziani et al., 1990) or in hybrids of FRTL5 and undifferentiated FRT thyroid cells (Zurzolo et al., 1991). The CAMP-dependent mitogenic pathwayk) is a tissue-specific characteristic of a set of cell types, including many types of differentiated epithelial secretory cells. “It is possible that secretory cells whose secretory functions are positively regulated by cAMP may also develop under the positive control of CAMP.. . . The [CAMP-dependent mitogenic] response . . . may also represent a characteristic of differentiating systems, that is that high levels of cAMP are required for differentiative transition. Thus proliferation of cells undergoing differentiation may have a different regulatory mechanism than the various cell cycles normally studied in cell cultures” (Filosa et al., 1975). This prophetic speculation about embryonic development of pancreatic cells is very well supported by recent investigations suggesting that the CAMPdependent growth stimulation and differentiation are intimately related processes in certain cell lineages, such as thyrocytes (see above) and somatotrophs. Overexpression of a dominant negative CREB in the pituitary of transgenic mice leads to ablation of the somatotroph lineage and dwarfism (Struthers et al., 1991). The Snell dwarf mutations, which cause both deficiency in GH synthesis and pituitary hypoplasia (Li et al., 1990), reside in GHF-1. GHF-1 is a tissue-specific CAMP-dependentgene under the control of CREB. Inhibition of GHF-1 synthesis with antisense oligonucleotides leads not only to a marked decrease in GH expression, but also to a marked inhibition of proliferation of somatotrophic cell lines (Castrillo et al., 1991).
V. A ROLE FOR
CYTOSKELETON CHANGES IN OF GROWTH BY CAMP?
CONTROL
There is still a continuously increasing body of evidence indicating that cell growth and differentiation are especially dependent on cell shape, direct cell-cell interactions, and mechanical as well as chemical properties of cell substratum (reviewed by Bissel and Barcellos-Hoff, 1987; Ben-Ze’ev, 1989; Ingber and Folkman, 19891, as is well illustrated by the anchorage dependence of proliferation of normal fibroblasts, which is relieved in cancerous or transformed cells. Extracellular matrix proteins influence morphogenesis, differentiation, and growth control characteristics (Juliano and Haskill, 1993). Their receptors (integrins) are directly coupled to the cytoskeleton through talin and vinculin (Carraway and Carothers Carraway, 1989). The cytoskeleton, providing a protein continuum between the cell
GROWTH MODULATION BY cAMP
123
membrane and the nucleus (Georgatos et al., 1987), could also convey various signals affecting gene expression. Different, but nonexclusive, hypotheses are envisaged to explain how the complete cell structure (or at least the cytoskeleton as a whole) may modulate the expression of genes involved in morphogenesis, differentiation, and growth control: (1) as most mRNAs are associated with the cytoskeleton, any change in the cytoskeleton organization could affect the translability of mRNA and macromolecular synthesis (Farmer et al., 1983); (2) since there is evidence for a connection between transcribed genes and the nuclear matrix, changes in the interaction of the cytoskeleton with the nuclear lamina could cause exposure of particular genes and sequestration of others (Bissell and Barcellos-Hoff, 1987; Ashall et al., 1988; Cook, 1989); (3) several protein kinases [including some A and C kinases (Miller et al., 1982; Papadopoulos and Hall, 1989) and growth factor receptors (Payrastre et al., 199211 and calmodulin are associated with the cytoskeleton, and reorganization of some cytoskeleton elements could be necessary for their translocation (e.g., toward the nucleus in order to phosphorylate transcription factors); and (4)cell adhesion or integrin clustering, as well as mitogenic neuropeptides and PKC activators, induce the tyrosine phosphorylation and activation of the p125fak kinase (Juliano and Haskill, 1993; Zachary and Rozengurt, 1992; Guan and Shalloway, 1992). The neuropeptide effect is independent of Ca2+ or PKC, but requires the actin cytoskeleton integrity (Sinnett-Smith et al., 1993). Cell adhesion could also affect other signaling cascades [Ca2+, inositol lipids (McNamee et al., 19931, and cAMP (Juliano and Haskill, 199311. Since the demonstration of a direct correlation between the degree of cell spreading or change in cell shape and the growth rate by Folkman and Moscona (1978), circumstantial evidence is accumulating that suggests a major role of the reciprocal relationship between cell shape and the cytoskeleton in the control of growth and its alterations associated with cell transformation. Thus, (1)the main identified substrates of oncogenic tyrosine kinases are components of adhesion plaques and cytoskeleton, such as integrins, vinculin, talin, and calpactin (Sefton et al., 1981; Pasquale et al., 1986; Glenney, 1986; Volberg et al., 19921, and growth factors and tumor promoters often produce rapid changes in cell morphology and cytoskeleton organization that resemble those associated with cell transformation (Bockus and Stiles, 1984; Herman and Pledger, 1985; Werth and Pastan, 1984; Nishida and Gotoh, 1992). (2) Various genes encoding extracellular matrix or cytoskeleton proteins are rapidly induced together with protooncogenes upon mitogenic stimulation in fibroblasts (Rollins and Stiles,
124
PIERRE P. ROGER et a1
1989). (3) On the other hand, one of the most prominent markers of cell transformation by a number of oncogenic viruses is the repression of the synthesis of some tropomyosin isoforms with high avidity for actin (Hendricks and Weintraub, 1984; Matsumura and YamashiroMatsumura, 1985; Cooper et al., 1985; Galloway et al., 1990). The repression of these tropomyosins is associated with the in uiuo tumorigenicity developed by some fibroblasts after transfection with a p-actin gene bearing a mutation frequently found in tumors induced by chemical carcinogens (Leavitt et al., 1987). p-actin could thus have the characteristics of a protooncogene activated by mutation. (4) While EGF and insulin disorganize microtubules and vimentin filaments during progression of BALBc-3T3 fibroblasts toward S phase (Bockus and Stiles, 19841, the disruption of the microtubule system by various microtubule poisons is sufficient to activate DNA synthesis in the absence of serum (Crossin and Carney, 1981) or to potentiate the DNA synthesis stimulation by serum of growth factors (reviewed by Otto, 1982). Microtubule inhibitors such as colchicine induce fos and myc protooncogene expression (Miura et al., 1987) and MAP kinase phosphorylation (Shinohara-Gotoh et al., 1991). How important is this conspicuous but still somewhat puzzling framework for the understanding of the role of cAMP in growth controls? cAMP profoundly affects cell morphology and the organization of the three main cytoskeleton systems in various ways in different cells. Hence, the early findings of cAMP effects on growth were already considered t o be intimately related to the CAMP-dependent morphological alterations (Pastan et al., 1975; Puck, 1987; Lockwood et al., 1982). Various drugs increasing cAMP and cAMP analogs restrict growth and restore “normal” fibroblastic morphology in spontaneously transformed CHO cells, src gene-transformed vole fibroblasts, and H-rastransformed NIH 3T3 cells (Puck, 1987; Meek, 1982; Lockwood et al., 1987). Although some of these drugs seem to produce this so-called “reverse transformation” partly as a result of their CAMP-independent side effects (Rajaraman and Faulkner, 1984), the involvement of CAMP-dependent protein kinases is well established by analysis of CHO mutants defective in PKA (Singh et al., 1981).The morphological changes follow an expansion of the cytoplasmic microtubule network, the assembly of microfilament bundles, and a redistribution of myosin into these bundles (Lockwood et al., 1982; Meek, 1982; Puck, 1984; Osborn and Weber, 19841, as well as rapid restoration of a wellorganized vimentin network (Chan et al., 1989).The CAMP-dependent reverse transformation also involves an increase in fibronectin depos-
GROWTH MODULATION BY cAMP
125
its (Puck, 1987). Puck proposed that the major but specific changes in the exposure or sequestration of genes associated with the reverse transformation and growth arrest are mediated by the CAMP-dependent changes in cytoskeleton organization (Ashall et al., 1988; Puck, 19871, since this gene exposure reaction is prevented by microtubule or microfilaments inhibitors such as colcemid and cytochalasin B. The CAMP-dependent reverse transformation is not general and is not specific to the cell type or the transforming factor. In src-transformed CHO cells CAMP,on the contrary, potentiates the tumorigenic action of the oncogene (Roth et al., 1983). On the other hand, the growth arrest provoked by cAMP in adrenal tumor cells is associated with a completely different morphological change (Hall et al., 1979). These cells rapidly retract and round up as a consequence of the CAMP-induced disruption of the microfilament bundles (stress fibers). This very rapid and dramatic morphological change also occurs in a variety of normal cells, including epithelial cells such as thyrocytes (Westermark and Porter, 1982; Tramontano et al., 1982; Nielsen et al., 1985; Roger et al., 19891, melanocytes (Preston et al., 1987), granulosa cells (Ben-Ze’evet al., 19891, Sertoli cells (Spruill et al., 1981), or some fibroblasts (Aubin et al., 1983; Lamb et al., 1988) and arterial muscle cells (Chaldakov et al., 1989). It presents obvious similarities with cytoskeleton changes associated with cell transformation (reviewed by Ben-Ze’ev, 1985) or provoked by PKC-activating tumor promoters, including the fact that this acute microfilament disorganization precedes decreases in the synthesis of actin (Cheitlin and Ramachandran, 1981; Passareiro et al., 1985) and of the high-molecular-weight “transformation-sensitive” tropomyosin isoforms (Roger et al., 1989; Ben-Ze’ev et al., 1989). The mechanisms involved in the early reorganization of cytoskeleton networks induced by cAMP are not well understood. The cAMP effect on actin cable disruption is mediated by the subunit of PKA (Roger et al., 1988a;Lamb et al., 1988),which, in uitro or in intact cells, phosphorylates the myosin light-chain kinase (Lamb et al., 19881, hence reducing its activity (Nishikawa et al., 1984). As the interaction of actin and myosin involved in microfilament assembly seems to be dependent mainly on the phosphorylation of the myosin light chain, it is plausible that the decrease in myosin light-chain phosphorylation observed in chick embryo fibroblasts microinjected with PKA C subunit (Lamb et al., 1988) or in thyrocytes treated with cAMP agonists (Ikeda et al., 1986; Contor et al., 1988) destabilizes microfilaments and provokes the disruption of stress fiber (Lamb et al., 1988; Roger et al., 1989; Baorto et al., 1992). However, a decrease in myosin light-chain phosphorylation has also been involved in the induction of stress fiber
126
PIERRE P. ROGER et al.
formation during the CAMP-dependent reverse transformation of CHO cells (Lockwood et al., 1982). A CAMP-dependent dephosphorylation of ADF/destrin might also be involved in the actin microfilament breakdown (Baorto et al., 1992; Saito et al., 1994). Similar confusion arises from comparison of the analysis of the CAMP-dependent phosphorylation of vimentin in different systems. In uitro, the phosphorylation of vimentin by the CAMP-dependent protein kinase disassembles the vimentin filaments (Inagaki et al., 1987).This phosphorylation, which resembles that occurring at mitosis, is associated with the collapse of vimentin filaments (Lamb et al., 1989). In Swiss 3T3 cells it was associated with CAMP-dependent mitogenesis (Escribano and Rozengurt, 1988). It is puzzling, however, that during the reverse transformation of CHO cells by CAMP,a phosphorylation of vimentin also precedes its reassembly in a well-expanded filamentous network (Chan et al., 1989). The CAMP-dependent disruption of actin-containing filaments was observed not only in cells that proliferate in response to CAMP, such as thyrocytes (Roger et al., 1989) or melanocytes (Preston et al., 19871, but also in unresponsive cells or in cells whose growth is inhibited by cAMP (Hall et al., 1979; Hornsby, 1985; Chaldakov et al., 1989). However, as cAMP may affect cell proliferation by acting on different intracellular targets in different systems, it cannot be excluded that some CAMP-induced cytoskeleton changes would be involved in the proliferative response in only some systems. Cytoskeleton changes could also be instrumental in some, but not all, of the mitogenic pathways. In Swiss 3T3 cells the mitogenic effect of cAMP does not involve changes in the organization of microtubules, although antimicrotubule drugs are mitogenic in these cells (Wang and Rozengurt, 1983). The mitogenesis of these cells in response to growth factors acting via CAMP-independentmechanisms is well correlated with increased motility (dependent on cytoskeletal changes), whereas CAMP-dependent growth-promoting treatments, on the contrary, inhibit cell migration (O’Neill et al., 1985). Similarly, canine thyrocytes that round up in response to TSH and cAMP have poor motility compared to cells treated with serum, EGF, or TPA (P. P. Roger, unpublished observations). In fact, the mitogenic effect of TSH via cAMP is similarly observed both in thyrocytes cultured in monolayer and as follicles in suspension (Roger et al., 1983; WynfordThomas et al., 1987). In human thyrocytes the stimulation of DNA replication and synthesis of PCNA/cyclin by TSH and cAMP occurs independently of culture configuration (monolayers or dense cell aggregates), although many other changes in gene expression induced
GROWTH MODULATION BY cAMP
127
by TSH are mimicked by culturing unstimulated cells as dense aggregates instead of monolayers (Roger et al., 1988b; Lamy et al., 1990). These observations do not favor a determinant role of cell shape and the cytoskeleton in the positive control of proliferation by CAMP, at variance with the possible major involvement of cytoskeleton changes in the CAMP-dependent restoration of restricted growth characteristics in some transformed cells. The positive effects of cAMP in some normal cells and its negative effect in some transformed cells might be exerted at quite different levels and may represent unrelated phenomena. cAMP stimulates growth by an action on key regulatory events of cell cycle progression, while in some transformed cells cAMP may affect proliferation characteristics by altering cytostructural changes involved in their mitogenic cascades. Clearly, despite 20 years of suspicion of an involvement of the cytoskeleton in growth control by CAMP, we are still looking forward to more direct evidence of this. Experimentation is difficult, because the significance of the message delivered by the cAMP system, including the activation of protein kinases and the effects of protein phosphorylation, is partly dependent on the complexity of the whole cytostructure (itself being affected by CAMP), which is lost in broken subcellular preparations. VI. cAMP AND
THE
GROWTH OF CANCERCELLS
A. NEGATIVE MODULATION Early reviews emphasizing a negative influence of cAMP on cell proliferation bear mostly on tumor-derived cell lines (Pastan et al., 1975). As pointed out in the more recent review by Boynton and Whitfield (1983) and in this chapter, cAMP is, on the contrary, a positive modulator for the multiplication of many normal differentiated cells. This might suggest that a change in the cAMP effect on growth is associated with cancerous transformation in many instances. The best examples are the inhibition of proliferation by cAMP in various melanoma cell lines (Wong and Pawelek, 1973; Slominski et al., 1989; Niles and Loewy, 19911, breast cancer cell lines (Cho-Chung et al., 1981, 1983; Fentiman et al., 1984; Iwasaki et al., 1983; Fontana et al., 1987; but see Kung et al., 1983; Welsch and De Hoog, 1983; Sheffield and Welsh, 19851, and Schwann-like cells from neurofibromas (Sobue et al., 19851, whereas cAMP is unequivocally demonstrated to be a mitogenic factor for their normal counterparts. Pregnancy-dependent
128
PIERRE P. ROGER et a1
mouse mammary tumors maintain the cAMP dependence of growth, but at the later ovarian-independent stage of progression, the growth is inhibited by cAMP (Imagawa et al., 19921. In the case of human melanoma cells, the sensitivity to growth inhibition by CAMP was closely correlated with the metastatic activity in different variants of the same parental cell line (Ormerod and Hart, 1989). Furthermore, transformation by Ha-ras oncogenes markedly increases the sensitivity to growth inhibition by cAMP in NIH 3T3 cell lines (Lockwood, et al., 1987; Davies et al., 19891, and cAMP specifically blocks the proliferation of rat 3T3 cells after their transformation by polyoma virus (Kamech et al., 1987). TSH via cAMP inhibits the growth of metastatic transformed cells that spontaneously arise from the TSH (CAMP)-dependentFRTL cloned rat thyroid cell line (Endo et al., 1990). Other recent examples of human cancer cell lines inhibited by CAMP include medullary thyroid carcinoma (de Bustros et al., 1986), salivary gland adenocarcinoma (Azuma et al., 19881, renal adenocarcinoma (Kinoshita et al., 1985), the gastric carcinoma cell line Kato I11 (Nakamura et al., 19891, HT29 colon adenocarcinoma cells (Garnet et al., 19921, small-cell lung carcinoma (Francis et al., 19831,and colon cancer cells (Tagliaferri et al., 1988a). cAMP mediates, in part, the antiproliferative effect of interferon a in PC3 prostate carcinoma cells (Okutani et al., 19911,but not in an astrocytoma cell line (Hubbel et al., 1991). Nevertheless, in the latter system cAMP can mediate the direct growth-inhibitory effect of double-stranded RNA (Hubbel et al., 1991). The apparent inversion of proliferative response in cancer cells is not unique to the CAMPcontrol of growth. Tumor-promoting phorbol esters, through PKC activation, in fact inhibit the growth of many tumor cells (Gescher, 19851, and the proliferation of several carcinoma cells is inhibited by EGF. In this case the EGF inhibition of growth has been correlated with the overexpression of EGF receptors (Gill and Lazar, 1981; Chen and Lin, 19931. EGF and TPA were reported to exert their inhibitory effects on the G2-mitosis transition (Kaszkin et al., 1992), as is the case for the CAMP-dependent inhibition of proliferation in many tumor cell lines (Friedman, 1976). Depending on the cell system and the putative factor responsible for cancerous transformation, different mechanisms have been proposed to explain how cAMP might specifically inhibit the growth of cancer cells. The hypothesis implicating the cytoskeleton in the “CAMPdependent reverse transformation” has been discussed in Section V. Another hypothesis relies on alterations in the permeability of the gap junctions that permit direct communication between adjacent cells by conducting low-molecular-weight cytoplasmic molecules, including
GROWTH MODULATION BY cAMP
129
small signal molecules. This junctional communication is profoundly altered in many transformed cells and cells from various cancers (reviewed by Sheridan and Atkinson, 19851, which may contribute to their autonomy. cAMP is a positive modulator of the junctional communication (Sheridan and Atkinson, 1985; Munari-Silem et al., 19901, even in neoplastic cells, in which this was correlated with growth inhibition by cAMP (Mehta et al., 1986; Murray and Taylor, 1988). Phosphorylation of the gap junction protein connexin 32 by PKA, as observed in hepatocytes (Saez et al., 19901, or CAMP-dependent transcription of connexin 43 (Saez et al., 1993) provides clues to the mechanism, as connexins are found to have tumor-suppressing activity. Quite interestingly, Mehta et al. (1986)observed that CAMP-increasingdrugs inhibit the proliferation of chemically and virally transformed 10 T 1/2 cells, only when promoting their communication through gap junctions with normal 10 T 1/2 cells. They suggested that cAMP promotes the transfer of small growth-inhibitory molecules from normal cells to cancerous ones. Thus, cAMP can inhibit the growth of cancer cells by restoring normal junction communication properties with normal cells. Also, in normal 3T3-Ll cells Shiba et al. (1989) found a correlation between CAMP-inhibitedgrowth stimulation by serum and CAMPinduced enhancement of gap-junctional communication. While MSH, through CAMP, stimulates the growth of normal melanocytes, it inhibits the proliferation of Cloudman S91 mouse melanoma cells. However, Pawelek et al. (1975) isolated variants of these cells that present a cAMP dependence of growth. Examination of these variants revealed that their type I CAMP-dependent protein kinases required a higher concentration of cAMP for their activation (Pawelek, 1979). It seemed that these cells required an optimum level of CAMP for proliferation. If the intracellular levels of CAMP are above or below this optimum, the cell cycle is lengthened, but this optimum may be modified by mutations of PKAs that alter their sensitivity to cAMP (Pawelek et al., 1983). This leads to the quite different hypothesis that the direction of the cAMP effect on growth is dependent on the characteristics of PKAs. A variation of this “kinase hypothesis’’was first proposed by Russell and collaborators (Byus et al., 1977) and is still receiving considerable support from a few groups, among which Cho-Chung’s (1989) is the most active. While type I PKA could be responsible for positive effects of cAMP on growth in some systems (see Section III,D,l; Livesey and Martin, 1988; Van Sande et al., 19891, type I1 kinase might mediate the inhibitory effects. Thus, the growth of several cell lines derived from hormone-dependent metastatic breast cancers, including MCF7 and
130
PIERRE P. ROGER et al.
T47D cells, is inhibited by cholera toxin, forskolin, and cAMP analogs (Cho-Chung et al., 1981, 1983; Fontana et al., 1987; Katsaros et al., 1988) and hormones increasing cAMP levels, such as calcitonin (Ng et al., 1983) and PGE (Iwasaki et al., 1983). Calcitonin was reported to activate only type I1 kinase (Ng et al., 19831, and site-specific cAMP analogs that preferentially activate type I1 kinase are the most potent growth inhibitors (Tagliaferri et al., 1988a). Furthermore, the growth arrest parallels an increase in type I1 PKA activity (Cho-Chung et al., 1981, 1983) and concentration (Ogreid et al., 19871, together with a decrease in type I PKA (Katsaros et al., 1988). A parallelism between RI/RII mRNA ratios and growth is also found in transformed versus control cells and in response to the inhibitor 8-C1-CAMP (Ciardiello et al., 1990). Cho-Chung (1974; Cho-Chung et al., 1983) also reported evidence that dibutyryl-CAMP and cholera toxin are potent inhibitors of the growth of hormone-dependent DMBA-induced mammary tumors in the rat in uiuo. Interestingly, autonomously growing rat mammary tumors that fail to regress after dibutyryl-CAMP treatment differ from CAMP-sensitive tumors by displaying another subtype of PKA I1 (Ogreid et al., 1987). Cho-Chung’s group generalized their findings to other tumor systems. The type I1 kinase-specific cAMP analog 8-C1-CAMP also inhibits the growth of human colony cancer cell lines (Katsaros et al., 1987; Tagliaferri et al., 1988a) and the clonogenic growth of leukemic blast progenitors (Pinto et al., 1992). In the LS174T colon cancer cell this inhibition involves the nuclear translocation of RII cAMP receptor protein and both an increase in RII and a decrease in RI gene transcription (Ally et al., 1988). The transformation of NIH 3T3 fibroblasts by Harvey murine sarcoma virus is reversed by cAMP analogs that predominantly activate PKA I1 (Tagliaferri et al., 1985). As in mammary tumor cells, the growth inhibition is accompanied by an increase in RII cAMP receptor protein and a decrease in RI protein (Tagliaferri et al., 1988b). Inhibition of RI and induction of RII gene expression precede the inhibition of growth and induction of megakaryocytic differentiation by 8-C1-CAMP in the K562 human leukemic cell (Tortora et al., 1989). A similar shift between the expression of RI and RII PKA subunits was also reported during the inhibition of growth of human lung carcinoma cells by 8-C1-CAMP in athymic mice (Ally et al., 1989). Similar effects of 8-Cl-cAMP, including growth inhibition (with no accumulation of cells at a specific cell cycle phase), a decrease in RI binding activity, and nuclear translocation of RII, have also been reported for some gastric carcinoma cell lines (Takanashi et al., 1991l.In HL60 leukemia cells antisense oligonucleotides against RIIp of PKA
GROWTH MODULATION BY cAMP
131
relieve the antiproliferative differentiating effects of cAMP analogs (Tortora et al., 19901, while antisense oligonucleotides against RIa induce an increase in RIIp mRNA, growth inhibition, and monocytic differentiation (Tortora et al., 19911, thus bypassing the effects of exogenous cAMP analogs. Similar findings were reported for different human cancer cell lines (Yokozaki et al., 1993). 8-Cl-cAMP, as a specific activator of PKA I1 and inducer of the RII nuclear translocation and transcription, was thus proposed as a potentially therapeutic agent in a wide spectrum of cancers (Katsaros et al., 1987; Cho-Chung, 1990; Pinto et al., 1992). The data of Cho-Chung and colleagues are obviously of important therapeutic significance. However, they are not undisputed and their interpretation recently raised major controversies. (1) Other groups even reported that cholera toxin and cAMP analogs stimulate the growth of MCF7 and T47D mammary carcinoma cells in uitro and in uiuo (Kung et al., 1983; Sheffield and Welsh, 1985) and promote rather than inhibit, the growth of MNU-induced rat mammary carcinomas (Welsch and De Hoog, 1983; Shefield and Welsch, 1988). (2) On the other hand, it is puzzling that the growth-inhibitory effects of 8-C1cAMP are not associated with an accumulation of cells at a specific phase of the cell cycle (Tagliaferri et al., 1988a; Takanashi et al., 1991; Pepe et al., 1991). In HL60 cells this is in sharp contrast with the accumulation of cells in G,/G, induced by N6-benzyl-CAMP (Pepe et al., 1991). Astonishingly, in HL60 cells and colon carcinoma cells, Sp and Rp phosphorothioate derivatives of 8-C1-CAMP (Sp. 8-C1-CAMPS and Rp 8-Cl-cAMPs)-agonist and antagonist, respectively, of PKAs-are both inhibiting growth (Yokozaki et al., 1992).The growth inhibition by Rp 8-C1-CAMP should be independent of PKA activation (Yokozaki et al., 1992). Van Lookeren Campagne et al. (19911, Langeveld et al. (19921, and Lange-Carter et al. (1993) ascribed the very strong inhibiting properties of 8-C1-CAMP on several cancer cell lines to the especially cytostatic properties of its metabolite &Cl-adenosine, which is formed in the culture medium [see also the published correspondence between Y. S. Cho-Chung and M. M. Van Lookeren Campagne and colleagues on this subject (Cancer Res. 51, 6206-6208, 199111. The growth-inhibitory effects of 8-Cl-CAMP are completely prevented by adenosine deaminase and even by the phosphodiesterase inhibitor 3-isobutyl-l-methylxanthine (by inhibiting the conversion of 8-C1-CAMP into 8-C1-adenosine) (Van Lookeren Campagne et al., 1991; Langeveld et al., 1992; Lange-Carter et al., 1993).Even the downregulation of the RIa and C subunits is mediated by 8-C1-adenosine through a CAMP-independent mechanism in normal and neoplastic
132
PIERRE P. ROGER et al.
epithelial cells (Lange-Carter et al., 1993). (3) Finally, Otten and McKnight (1989) demonstrated, using ras-transformed NIH 3T3 cells, that overexpression of the type I1 R subunit of PKA, which leads to elimination of type I holoenzyme, does not reverse the transformed phenotype. This indicates that the 8-C1-CAMP-induced reverse transformation reported by Tagliaferri et al. (1988b) in this system could not be caused by the shift in PKA isozyme expression. Such a shift could be simply a general consequence of the activation of PKAs (McKnight et al., 19881, which could also occur in cells whose growth is stimulated by CAMP.In Section III,D,l we discussed the possible significance of a similar shift between PKAs I and I1 during the CAMPdependent proliferation of thyroid cells (Breton et al., 1989; Roger et al., 1991). A clear means by which cAMP could specifically inhibit the growth of transformed or cancer cells is by interacting with the expression of the transforming genes (oncogenes).Indeed, Cho-Chung and collaborators also reported that cAMP analogs that activate type I1 PKA inhibit the synthesis of the p2lras protein (either cellular or viral) in NIH 3T3 cells transfected with Harvey murine sarcoma virus DNA (Tagliaferri et al., 1985) and in mammary carcinomas and MCF7 cells, restoring to a normal level their high cellular p21H-1-a~ expression (Huang and ChoChung, 1984; Tagliaferri et al., 1988a). Similarly, Azuma et al. (1988) observed that dibutyryl-CAMP suppresses p21ras expression and cell proliferation in a human salivary gland adenocarcinoma cell line. However, these observations were not confirmed in other systems. The CAMP-induced reverse transformation in ras-transformed BALB/c3T3 cells (Ridgway et al., 1988) and the induction of differentiation by cAMP in ras-transformed MDCK cells (Wu and Lin, 1990) occur without changed p2lv-ras levels. Whether ras overexpression is directly responsible for the formation of these various CAMP-sensitive tumors and how cAMP could negatively control both cellular and viral ras oncogene expressions remain important questions. Also, it is not known whether the p2lras function is altered by its CAMP-dependent phosphorylation in intact cells (Saikumar et al., 1988). The fact that dibutyryl-CAMP partially restores PDGF-dependent signaling events in ras-transformed NIH 3T3 cells (Olinger et al., 1989) is, however, supportive of the suggestion that cAMP can antagonize the ras function. A mechanism was recently provided by the demonstration that, in several cell types, including the H-ras-transformed NIH 3T3 cells (Chen and Iyengar, 1994), cAMP inhibits the raf-l-MAP kinase cascade activated by ras (see Section 11,B). Other human tumor cell lines, including glioma and osteosarcoma
GROWTH MODULATION BY cAMP
133
cell lines, express the c-sis oncogene encoding the PDGF, which may contribute to carcinogenesis by autocrine or paracrine mechanisms. Indeed, various cAMP enhancers and analogs block the transcription of this protooncogene in these cells (Harsh et al., 1989). Conversely, cAMP can also arrest growth by inducing the secretion of autocrine growth inhibitors, such as TGF-P, in the PC3 prostate carcinoma cell line (Bang et al., 1992). Besides the evidence that cAMP may specifically inhibit the growth of various cancer cells, it is obvious that cells whose growth is normally inhibited by cAMP (fibroblasts and lymphocytes) may maintain this negative control after neoplastic transformation. This is especially true for various leukemic cells in which cAMP inhibits growth and induces differentiation. This reverse transformation process seems to involve the inhibition of c-myc expression also observed in normal lymphocytes (Slungaard et al., 1987; Tortora et al., 1989). Whatever cellular mechanisms are implicated in the inhibition of growth of cancer cells by CAMP,the potentiality of using it as a therapeutic means to control the progression of tumors in uiuo has been considered very attractive by several authors. The B. pertussis invasive adenylate cyclase (Slungaard et al., 1983), cholera toxin (Cho-Chung et al., 1983; Lanotte et al., 19861, and 8-C1-CAMP (Cho-Chung, 1990; Pinto et al., 1992) have thus been proposed as tools to manipulate the cAMP cascade in tumor cells in uiuo. This enthusiasm should be moderated when considering the following caveats: (1) cAMP is a ubiquitous second messenger involved in almost all of the vital functions of the organism, which could be perturbed as cAMP is systematically increased: (2) the mechanisms involved in the cytostatic and cytotoxic properties of 8-C1-CAMP and its metabolite, 8-Cl-adenosine, are still poorly understood and are less specific than was initially claimed; (3) by limiting the proliferation of lymphocytes, cAMP enhancers and 8-C1-CAMP are potential immunosuppressive factors and thus could favor neoplastic development; and (4)as discussed in Section VI,C, CAMP,by stimulating the growth of some tumor cells, could act as a tumor promoter. B. ESCAPE FROM NEGATIVE MODULATION The hypothesis that cAMP is a ubiquitous inhibitory factor involved in normal growth control prompted a series of studies based on the notion that cancer could result from defects in the CAMP-PKA cascade (Pastan and Johnson, 1974; Chlapowski et al., 1975). Some recent studies still attempt to support this view, which contrasts sharply with
134
PIERRE P. ROGER et al.
that discussed in the previous section. Tumor cells lose hormone receptors coupled to adenylate cyclase, such as PGE receptors during the progression of rat mammary tumors (Abou-Issa and Minton, 1986) and in a culture of tumorigenic rat urothelium (Chlapowski and Nemecek, 19851, or P-adrenergic receptors in another rat urothelium cancer cell line (Chlapowski and Nemecek, 1985)and in NIH 3T3 cells upon transformation by N-ras (Davies et al., 1989).Moreover, a decrease in adenylate cyclase activity was observed in ras-transformed NIH 3T3 cells (Tarpley et al., 1986; Davies et al., 19891, in ras-transformed FRTL5 rat thyroid cells (Colletta et al., 19881, and in some rat urothelium tumor cells (Chlapowski and Nemecek, 1985).It was proposed as a marker to differentiate colon cancer from benign cells in culture (Nelson and Holian, 1988). The GTP binding t o the adenylate cyclase-activating subunit Gsa is defective in mouse lung tumors (Droms et al., 19871, leading to decreased hormone responsiveness (Droms et al., 1989). Whether tumor progression depends on such defects in cAMP metabolism remains doubtful, as they could be secondary or could reflect the general loss of differentiation linked t o transformation. In fact, Chlapowski and Nemecek (1985) did not observe any correlation between in vitro growth rate or tumorigenicity and the various abnormalities in cAMP metabolism (ranging from a strong reduction to an excess of adenylate cyclase activities) they found in urothelium tumor cells. The ras-transformed NIH 3T3 cells with reduced adenylate cyclase activity are paradoxically more sensitive to growth inhibition by cAMP (Davies et al., 1989). Similar reservations should be made for the interpretation of studies reporting alterations in PKA responses to cAMP in tumor cells. In mouse lung urethane-induced tumors the high-affinity binding of cAMP to the RII PKA subunit is strongly reduced, not because of a structural alteration of RII but possibly due to a modified conformational state or interaction with other cytosolic molecules (Butley et al., 1984, 1985). In other lung tumor cells lower PKA I expression as well as deficiencies in PKC activity were reported (Nicks et al., 1989). The decrease in PKA I expression is due to a decrease in RI mRNA content, likely at the transcriptional level (Lange-Carter et al., 1990; LangeCarter and Malkinson, 1991). Whether such modifications are causal in the tumorigenic process is not known. More significant could be the study by Hiwasa et al. (19871, who reported that some carcinogeninitiated clones (i.e., growing in soft agar with TPA) derived from BALB/c-3T3 fibroblasts are resistant to CAMP,while clones selected for cAMP resistance after carcinogen treatment behave like initiated
GROWTH MODULATION BY cAMP
135
clones in a two-stage carcinogenesis model. In both categories of clones, the cAMP resistance of growth is correlated with defects in PKA activities (Hiwasa et al., 1987). However, the alterations in PKA activities are only quantitative and might depend on several factors. C. cAMP AS
A
TUMOR PROMOTER
Since the activation of the cAMP cascade stimulates the growth of many differentiated cells, it is likely that cAMP could also stimulate the proliferation and thus the clonal expansion of some cells initiated by oncogene activation. Again, the thyroid gland provides the best illustration of this assumption. The classical experimental protocols to induce thyroid tumorigenesis all involve chronic stimulation by TSH (Christov and Raichev, 1972; Dumont et al., 1980). Thyroid follicular cell malignancies (mostly follicular carcinomas) frequently develop during TSH-dependent goitrogenesis (Konig et al., 1981; Williams, 1990), and TSH suppression by treatment with thyroid hormones delays the growth of many thyroid tumors (Clark, 1981).Follicular carcinomas likely progress from follicular adenomas, many of which could be caused by the mutational activation of ras oncogenes (Lemoine et al., 1989). Human thyroid follicular adenoma cells often retain the TSH dependence of growth in uitro (Williams et al., 1988) and in explants in nude mice (Dralle, 1989). Adenylate cyclase sensitivity to TSH is even greater in some follicular carcinomas (Clark and Gerend, 1985) and the proliferation effects of TSH on normal human thyrocytes are mediated by cAMP (Roger et al., 1988b). Therefore, the promotion and progression of many follicular tumors of the thyroid gland seem to be dependent on the mitogenic effects that TSH exerts via intracellular cAMP elevation. By contrast, papillary carcinomas of the thyroid gland are frequently associated with a high-iodine diet and thus could have progressed despite low circulating TSH levels (Konig et d., 1981; Williams, 1990). The activation of ret (Grieco et d., 19901, TRK (Bongarzone et al., 19891, and MET (DiRenzo et al., 1991) tyrosine kinase oncogenes, but not ras oncogenes (Lemoine et al., 19881, is often found in this second thyroid tumor type. Overexpression of the erb-B2 protooncogene was also reported in papillary carcinomas (Aasland et aZ., 1988). Thus, the two main thyroid malignancies not only seem to have been determined by different hormonal promoting environments, but also are associated with different oncogenes. We speculate that the activation of the distinct CAMP-dependent or -independent mitogenic
136
PIERRE P. ROGER et al
pathways (as discussed in Section 111,E,2) could have promoted-and selected-the clonal growth of cells initiated by the activation of different oncogenes, thus leading to different tumor phenotypes. Whether a similar tumor promoter role could be ascribed to cAMP and CAMP-enhancing factors in other differentiated tissues, the growth of which is also positively controlled via CAMP,remains poorly documented and awaits further studies. The proliferation of some mammary tumor cell lines (Taylor-Papadimitriou et al., 1980; Imagawa et al., 1992) and of a rat bladder carcinoma cell line (Boyer and Thiery, 1993) is stimulated by cAMP in synergy with other factors. There has been recent interest in finding VIP receptors coupled to adenylate cyclase in several tumor types (Ruellan et al., 1986; Siperstein et al., 1988; Gespach et al., 1988; Yu et al., 1992; Moody et al., 1993), as this neurotransmitter was proved to be mitogenic via cAMP in some normal cells in culture (Haegerstrand et al., 1989; Pincus et al., 1990a; Zurier et al., 19881, as well as in Lewis lung carcinoma cells, a mouse mammary carcinoma cell line (Scholar and Sudhor, 19911, and a human colon carcinoma cell line (Yu et al., 1992). It was hypothesized that VIP could serve as an autocrine growth factor in neuroblastoma (O’Dorisio et al., 1992). A VIP antagonist inhibits the accumulation of cAMP and growth of non-small-cell lung cancer in uitro and in uiuo (Moody et al., 1993). Gastrin, possibly via CAMP, is a potent promoter of the growth of a xenotransplantable human gastric carcinoma (Ochiai et al., 1985; Sumiyoshi et al., 1984; Yasui et al., 1986; but see Takanashi et al., 19911, but cholera toxin has no tumor-promoting activity in mouse skin, although it is a potent inducer of epidermal hyperplasia (Kuroki et al., 1986).
D. ONCOGENES RELATEDTO THE cAMP SIGNALING CASCADE Until 1987 there was no report of an oncogene product related to the cAMP signaling cascade (Gottesman and Fleischmann, 1986). This probably justified the relative lack of interest in the CAMP-dependent mitogenic cascade, as compared to tyrosine kinase-dependent mitogenic pathways that involve the cellular homologs of many oncogenes. Hyperactivation of the cAMP signaling cascade should lead to the generation of hyperfunctioning tumors in many systems of differentiated endocrine cells (Dumont et al., 1989).Our first biochemical analysis of the cAMP pathway in hypersecreting thyroid adenomas (“hot nodules”) did not demonstrate a major constitutive activation of the cAMP cascade (Van Sande et al., 1980, 1988). However, recently we
GROWTH MODULATION BY cAMP
137
have found new somatic mutations of the TSH receptor conferring constitutive activation of adenylate cyclase but not of phospholipase C in such adenomas (Parma et al., 1993). In late 1987 Vallar et al. reported that many pituitary adenomas that hypersecrete GH do carry autonomously active G, protein and adenylate cyclase. Extracts of metastatic B16 melanoma cells also confer increased adenylate cyclase sensitivity in reconstituted ,949 cyccell membranes (Lester et al., 1987). Inasmuch as the secretion and growth of somatotrophic cells were known to be cAMP dependent (Billestrup et al., 19861, the alteration of G, was assumed to be the direct cause of both high secretory activity and autonomous growth of the tumors in which it occurs. As a, also activates some Ca2+ channels, an additional role of Ca2+ in the pathologies caused by its constitutive activation cannot, however, be completely excluded (Hamilton et al., 1991). Further analysis of the pituitary tumors led to the demonstration of three somatic mutations of the G, protein a-chain at codons 201 and 227 (ArgzOlCys,ArgZOlHis, and Gln227Arg), which cause constitutive activation of a, by inhibiting its GTPase activity (Landis et al., 1989).Interestingly, the importance of a, GlnZz7mutations for a, activity had already been suggested by deliberate site-directed mutational analysis (Masters et al., 1989; Graziano and Gilman, 1989) and by its analogy with the Glu61 mutation, which inhibits GTPase activity of p2lras. On the other hand, Argzol is precisely the site of a,that is ADPribosylated by cholera toxin, also resulting in GTPase inhibition and block of Gs inactivation (Freissmuth and Gilman, 1989).These activating mutations of G,a were predicted t o convert G, into a dominant oncogene [anticipatively called gsp (Landis et al., 198911 in cells programmed to proliferate in response to CAMP.This prediction was confirmed in some cases of autonomously functioning thyroid adenoma (hot nodule) (Lyons et al., 1990; Suarez et al., 1991; O'Sullivan et al., 19911, in 50% of somatotroph adenomas, and in other hypophyseal tumors (Lyons et al., 1990; Landis et al., 1990; Spada et al., 19901, but not in a large panel of tumors including melanomas, ovarian, and adrenal cortical tumors (Lyons et al., 1990). However, ArgZolCys and ArgzolHis mutations of a, were recently demonstrated by Weinstein et al. (1991) and Scheidinger et al. (1992) as the probable cause of McCune-Albright syndrome, a disease characterized by polyostotic fibrous dysplasia, caf6 au lait pigmentation of the skin, sexual precocity, and hyperfunction of multiple endocrine glands. The somatic mutation should appear at an early postzygotic stage in order to explain its occurrence in multiple tissues and the
138
PIERRE P. ROGER et al.
mosaic distribution of cells bearing or not bearing the dominant mutation. In McCune- Albright syndrome hyperplasia and formation of functional nodules and adenomas associated with a, mutations are not restricted to thyroid cells and somatotrophs, but are also found in the ovaries, resulting in “autonomous” follicle maturation, and in the adrenal cortex (Weinstein et al., 19911, two tissues for which the growthstimulatory role of cAMP has been questioned and whose tumors were associated with mutational activation of ai2 rather than a, (Lyons et al., 1990). Thymic hyperplasia, gastrointestinal adenomatous polyps, and Leydig cells hyperplasia in the testes were also observed (Weinstein et al., 1991). The transfection of constitutively active G,a with the Glu227Leu mutation increases the mitogenic responsiveness of Swiss 3T3 cells (Zachary et al., 1990). Heterotypic expression of adenylate cyclase-controlling receptors could, depending on the role of cAMP in the proliferation of the cells involved, lead to growth or atrophy. A remarkable model for this has been provided by the adenosine A2 receptor. This newly cloned receptor, which positively controls adenylate cyclase, is in many cells physiologically constitutive (Maenhaut et al., 1990). This is, at least in part, due to the fact that the adenosine normally produced by these cells is sufficient to activate the receptor. When expressed in canine thyroid cells by microinjection of mRNA, this receptor induces constitutive proliferation (Maenhaut et al., 1990). When expressed in transgenic mice with a thyroid-specific thyroglobulin promoter its cDNA induces an hyperfunctioning goiter and thyrotoxicosis (Ledent et al., 1992). There are also some clues that overexpression of P-adrenergic receptors could be associated with thyroid neoplasia. In some cases of autonomous thyroid adenomas, the cAMP response to norepinephrine was found to be enhanced compared to that of normal tissue (Van Sande et al., 1980). In the FRTL5 thyroid cell line, which normally does not possess P-adrenergic receptors, introduction of a P,-adrenergic receptor by transfection (Hen et al., 1989) or infection with a retroviral construct (Tsuzaki et al., 1991) confers an isoproterenol-sensitive growth. Interestingly, in the latter case the overexpression of P,-receptors (105receptors per cell) also markedly increases the basal cAMP levels and causes an autonomous (i.e., in the absence of TSH or isoproterenol) proliferation (Tsuzaki et al., 1991).A spontaneously transformed FRTL5-derived cell line that is malignant in nude mice also has elevated basal cAMP levels and has acquired P,-adrenergic receptors (Endo et al., 1990). P,-Adrenergic receptor mRNA was preliminarily reported to be overexpressed in some neoplastic human thyroid tissues (Ling et al., 1992). Heterotypic expression of GIP receptors in the adrenal gland and the
GROWTH MODULATION BY cAMP
139
consequent activation of the cAMP cascade in these cells after each meal has now been shown to be responsible for some Cushing’s syndromes with adrenal hyperplasia and adenoma (Lacroix et al.,1993). Many other adenylate cyclase-activating receptors were previously found to be ectopically expressed in adrenocortical tumors, including receptors to TSH, LH, FSH, glucagon, vasopressin, and adrenaline (reviewed by Lacroix et al., 1993). As was first shown by Lefkowitz’s group, G protein-coupled receptors with seven transmembrane-spanning domains can also be activated by mutations and thus constitutively turned on (Lefkowitz et al.,1993).In tissues growing in response to CAMP,constitutive activation of adenylate cyclase by such mutations should cause tumors. Indeed, somatic Ala623 and Asp619 mutations of TSH receptors cause TSH-independent activation of adenylate cyclase, but not phospholipase C, and hyperfunctioning thyroid adenomas (Parma et al., 1993). Other adenylate cyclase-activating mutations of LH receptors of Leydig cells were found to be responsible for male precocious puberty, a disorder caused by Leydig cell hyperfunction and hyperplasia (Shenker et al., 1993). These observations suggest that G protein-coupled receptors that activate adenylate cyclase may behave as protooncogenes. In the yeast S. cerevisiae, the growth of which is dependent on CAMP, any mutation activating the CAMP-PKA signaling cascade leads to deregulated proliferation and escape from nutrient control (Dumont et al.,1989). By analogy, it could be inferred that other mutations overactivating this pathway will prove to be responsible for tumor formation in tissues in which growth is CAMP-dependent. For example, a mutation of the C subunit of PKA disrupts inhibition by the R subunit without altering substrate recognition in yeast and leads to autonomous growth (Levin et al.,1988; Levin and Zoller, 1990).This mutation affects the interaction between R and C subunits in a region that is well conserved, structurally and functionally, in mammals (Levin and Zoller, 1990). The homologous mouse mutant Thrl97Ala also remains partially active in the presence of excess R subunits (Orellana and McKnight, 1992). Even greater unregulated activities of the mouse C subunit are observed in the Hiss7Glu and Trpl96Arg mutants (Orellana and McKnight, 1992). A mutation of the mouse RII subunit was also described (ArgArg92,93AlaAla),which does not alter the holoenzyme formation but confers full enzymatic activity even in the absence of cAMP (Wang et al., 1991). Such mutations likely lead to functional tumors of thyroid cells or somatotrophs. Inactivation of elements negatively controlling the cAMP cascade should have the same result. Alteration of cAMP phosphodiesterase
140
PIERRE P. ROGER et al.
renders MDCK cells independent of the CAMP-dependent growth stimulation by glucagon or PGE, (Taub et al., 1983). cAMP phosphodiesterase could thus be a potential target for recessive oncogenic mutations in differentiated epithelial cells. In the thyroid gland cold adenomas (i.e., adenomas that do not take up iodide) have been found to have defects in iodide trapping and iodide oxidation (DemeesterMirkine et al., 1984). Considering that iodide through an oxidized intermediate inhibits the TSH-CAMP cascade, such adenomas could result from the relief of this negative control (Maenhaut et al., 1991). In cells in which cAMP is a signal for growth, any negative element in the cAMP cascade can thus be considered as a potential antioncogene, whose inactivating mutation could cause tumorigenesis; any positive element or gene in the cascade could be a potential oncogene when a mutation causes constitutive activation. While many undifferentiated tumors are associated with the activation or overexpression of oncogenes related to the tyrosine kinase and Ca2+ phospholipid pathways, further studies in differentiated hyperfunctioning adenomas should define the extent to which these and other alterations of the cAMP system account for some of these benign tumors. A second G-protein oncogene, g i 2 , results from somatic point mutations in the gene for the a-subunit of Gi2 (Lyons et al., 1990). gip2 is found in endocrine tumors of the adrenal cortex and the ovary (Lyons et al., 1990), and it induces neoplastic transformation of rat-1 fibroblasts (Pace et al., 1991) and, to a lesser extent, NIH 3T3 cells (Hermouet et al., 1991). Unlike gsp, gip2 inhibits cAMP accumulation (Wong et al., 19911, which was proposed to exert a mitogenic influence in some fibroblasts (Seuwen et al., 1988; Van Corven et al., 1989). However, it remains doubtful that this mechanism could explain the oncogenic effects of Gi2 mutations, since Gi2 can be coupled t o other cascades (Letterio et al., 1986; Zachary et al., 1991; Gupta et al., 1992). In particular, dissociation of Gi2 can activate phospholipase C by its py-subunits and ai2 might lead to p2lras activation independently of adenylate cyclase inhibition or phospholipase C stimulation (Van Corven et al., 1993). The fact that pertussis toxin inhibits the growth effects of gip2 in rat-1 cells (Pace et al., 19911, but not its inhibitory effects on adenylate cyclase, also suggests that its mitogenic effects are not due to this inhibition (Wong et al., 1991). VII. CONCLUSIONS AND PERSPECTIVES
A main drawback of previous analyses of the role of cAMP in proliferation has been the tendency of authors to define a “universal role”
GROWTH MODULATION BY cAMP
141
for the intracellular signal in cell proliferation. No such generalization is possible. It is quite evident that cAMP is a negative regulator of proliferation in some cells, is a positive regulator in others, and has a biphasic effect in a third category. Cells of the latter type can evolve clones in which only the stimulatory effects are seen (S91 melanoma cells) (Pawelek et al., 1975). On the other hand, cells positively regulated by the cAMP cascade (FRTL5 cells) can produce clones in which cAMP has only a negative effect (Endo et al., 1990). It is therefore apparent that animal cells possess in their genes both positive and negative CAMP-dependent proliferation cascades and that, depending on the program (differentiation) of the particular cell type, either one or both-or parts of both intervening at different steps of the mitogenic pathway-may coexist. Depending on this program, activation of the cAMP cascade will lead to the triggering of proliferation, proliferation arrest, or synergism with other factors eliciting either of these opposite outcomes. It is obvious that the CAMP-dependent machinery to induce growth must be very sophisticated: cAMP as a sufficient activator of mitogenesis must support all stages of the prereplicative phase. On the other hand, the machinery for negative control needs to operate on only one step (preferably Go-G, transition) of the mitogenic activation process. To add a supplementary level of diversity, the growthinhibitory effects of cAMP in normal cells such as fibroblasts or lymphocytes may use different strategies. Interaction with the signaling cascade of mitogens and/or inhibition of cell cycle progression at different restriction points, ranging from an inhibition of the raf-l-MAP kinase cascade and c-myc expression at the Go-G, transition to a block at the S-phase commitment point in late G,, have been convincingly reported. These mechanisms should be distinguished from much less specific inhibitory effects of CAMP,as they are observed in many cancer cells: they require cAMP clamping at levels never reached in physiology, involve abnormal blocks of the cell cycle in S or G, phase, involve neutralization of the effects of activated oncogenes, or are secondary to effects on cytoskeleton organization, cell-cell communication, or differentiation. The growth stimulation by cAMP is also produced by quite different mechanisms. As we have shown in thyroid cells, cAMP may trigger its own mitogenic cascade, but positive interactions with the signaling cascades of other mitogens are common. The effects of cAMP on proliferation may also be indirect, as in the case of the adrenal gland and ACTH, in which the cAMP cascade in viuo stimulates function, proliferation, and differentiation expression, but the effect on proliferation is not reproduced in uitro, as it is secondary t o the activation of an
142
PIERRE P. ROGER et al.
autocrine loop. The various strategies of cAMP growth stimulation are not exclusive and may cooperate in one system. In Schwann cells cAMP seems to exert a direct mitogenic influence together with induction of PDGF receptors (Weinmaster and Lemke, 1990) and inhibition of the production of an autocrine negative factor (Muir et al., 1990). The role cAMP plays in one type of cell depends on its position in the regulatory network. Vertebrate cells contain genes that can complement yeast cells defective in cell cycle control. The general machinery of the cell cycle is therefore probably remarkably conserved through evolution. That CAMP-dependent protein kinases are not involved in this basic machinery is suggested by the normal proliferation behavior of S49 kinase-deficient mutants (Coffin0 et al., 19751, although type I1 CAMP-dependent protein kinase was found to be associated with p34cdc2 kinase (Tournier et al., 1991). However, the physiological controls that operate on this machinery are specific for each cell type and may even evolve during cell cycle progression, as exemplified by the cell cycle-dependent coupling of the calcitonin receptor to different G proteins (Chakraborty et al., 1991). They must therefore themselves control the core machinery. Between these levels operate the different intracellular regulatory cascades. However, although many extracellular influences operate on any cell, the number of transducing cascades is remarkably small and their role varies in the different cells. It was therefore to be expected that each cascade, depending on its role on the general physiology of the cell, would either stimulate or inhibit cell proliferation. As we have shown, this is obviously the case for the cAMP cascade. The very interesting problem this raises is how the same cascade could act positively in some cells and negatively in others and what determines the wide diversity of the mechanisms of these positive and negative modulations in various systems. The solution to this problem could lie, in part, in the black box that links protein phosphorylation and protooncogene expression, and in the multiple cell type-dependent interactions between the major signaling cascades. An interesting example of how this could be achieved is the expression of isoenzymes with the same function but different regulation characteristics. For example, isozyme p of PKC inhibits cAMP accumulation, while isozyme y enhances it (Gusovsky and Gutkind, 1991). Another clue derives from the finding that a splicing event generates a switch of CREM function from an antagonist to an activator during spermatogenesis (Foulkes et al., 1992). Similar developmental switches by alternative splicing (Foulkes et al., 1991a) or alternative usage of initiation codons in mRNA (Delmas et al., 1992) could explain the differ-
GROWTH MODULATION BY cAMP
143
ential regulation (positive or negative) of the c-myc, ODC, and perhaps cyclin D1 genes by cAMP in some cells stimulated o r inhibited by CAMP, whereas the expression of fos and j u n protooncogenes is induced or inhibited by cAMP irrespective of the final effect on growth. To answer such questions, it will be necessary to study a few welldefined systems in depth to obtain coherent answers. Most often, in the cells in which it is a positive control on proliferation, cAMP is also a positive signal for function, and conversely, a similar conclusion can be drawn for the role of the PIP,-Ca2+ cascade: in the cells in which this cascade activates function, it also stimulates growth, for example, in lymphocytes stimulated by IL-4, in hepatocytes, and in smooth muscle. This makes physiological sense, as a tissue stimulated repeatedly would develop its capacity to respond to stimulation, that is, would grow. In adult tissues growth is the longterm adaptation to functional stimulation. In the thyroid and somatotroph paradigms and probably in many other cell types in which function is controlled by CAMP,the CAMP-dependent stimulation of cell proliferation appears as a differentiation characteristic. It has been suggested recently that in somatotrophs the lineage-specific transcription factor GHF-1 induced by cAMP is necessary not only for CAMP-dependent transcription of the GH gene, but also for lineagespecific cell proliferation (Castrillo et al., 1991). As a specialized function the CAMP-dependent mitogenic pathway in thyroid cells possesses various odd characteristics that make it unexpectedly dissimilar from the more convergent CAMP-independent growth-stimulatory mechanisms. Unlike the tyrosine kinase and PKC growth-signaling pathways, the cAMP mitogenic cascade of thyrocytes does not utilize the phosphorylation of proteins on tyrosine, the activation of MAP kinases, or the induction of c-jun. Kinetic experiments indicate that the events associated with the prereplicative development supported by cAMP are not helpful in the CAMP-independent mitogenic cascades or vice versa. It is recognized, sometimes with difficulty, that different cell types may utilize different genes (Baserga, 1990) in their regulation of cell cycle progression. Comparison of CAMP-dependent and -independent modes of cell cycle control in thyroid cells suggests that, in a given cell, different strategies utilizing different genes and sequences of regulatory events may coexist, separately leading to cell division. Far-reaching consequences ensue, opening new research perspectives. For instance, unique characteristics of the CAMP-dependent pathway in thyrocytes might explain how CAMPdependent growth and differentiation can be compatible. CAMPdependent and -independent cell division might also have different
144
PIERRE P. ROGER et al.
consequences on the cell phenotype, which might be important in differentiation processes. Intracellular inhibitors of cell cycle progression might well be instrumental on only some particular mitogenic pathways. They might be involved in pathway-specific growth desensitization mechanisms and senescence-like processes, and may constitute “conditional antioncogenes.” While perturbations of the tyrosine kinase and phospholipase C-PKC mitogenic pathways are often found in dedifferentiated tumors, hyperactivation of the cAMP signaling cascade (the gsp oncogene and the TSH receptor) is now well documented in several examples of hyperfunctioning tumors. It might now be fruitful to search for the genes and proteins specifically involved in the cAMP mitogenic cascade and for new oncogenic mutations that might occur at various levels of this cascade. NOTEADDED IN PROOF The promoter of the gene encoding cyclin A, a pivotal regulatory protein involved in the S phase, is inducible by activation of the cAMP signaling pathway in human fibroblasts (Desdouets et al., 1995). This is mediated via a CRE and cell-cycle-regulated phosphorylation of CREB and CREMT and transient disappearance of the inducible CAMPearly repressor (ICER) (Desdouets et al., 1995). This is reminiscent of evidence of CREB involvement in the mitogenic activation (Wollberg et al., 1994) and in the induction of the PCNA gene (D. Huang et al., 1994) in T lymphocytes. Thus, even in systems in which an overall negative effect of cAMP on cell cycle was shown, intermediates of the CAMP-signaling cascade might be positively involved in the sequential expression of cell cycle regulators in response to PKA activation or the activation of other kinases by growth factors (Ginty et. al., 1994; de Groot et al., 1994). ACKNOWLEDGMENTS We thank Stein 0. Doskeland for helpful suggestions and D. Leemans for secretarial assistance. This work was supported by the Belgian Programme on Interuniversity Poles of Attraction initiated by the Belgian State, Prime Minister’s Office, Federal Service for Science, Technology and Culture (scientific responsibility is assumed by the authors), the Fonds National de la Recherche Scientifique, Fonds de la Recherche Scientifique Medicale, Caisse GBneral d’Epargne et de Retraite, Euratom, Televie and Association contre le Cancer. P.P.R. is a Research Associate of the Fonds National de la Recherche Scientifique (Belgium). REFERENCES Aasland, R., Lillehaug, J. R., Male, R., Josendal, O., Varhaug, J. E., and Kleppe, K. (1988). Expression of oncogenes in thyroid tumours: Coexpression of c-erbB2ineu and c-erbB. Br. J. Cancer 576, 358-363. Abate, C., Marshak, D. R., and Curran, T. (1991). Fos is phosphorylated by ~34~11~2, CAMP-dependent protein kinase and protein kinase C at multiple sites clustered within regulatory regions. Oncogene 6,2179-2185. Abdel-Malek, Z., Swope, V. B., Paleas, J.,Krug, K., and Nordland, J. J. (1992).Mitogenic, melanogenic, and cAMP responses of cultured neonatal human melanocytes to commonly used mitogens. J . Cell. Physiol. 150, 416-425.
GROWTH MODULATION BY cAMP
145
Abou-Issa, H., and Minton, J. P. (1986). Loss of prostaglandin E receptors during progression of rat mammary tumors from hormonal dependence to autonomy. J . Natl. Cancer Znst. 76, 101-106. Adams, S., Bacskai, B. J., Harootunian, A. T., Mahaut-Smith, M., Sammak, P. J., Taylor, S. S., and Tsien, R. Y. (1993). Imaging of cAMP signals and A-kinase translocation in single living cells. Adu. Second Messenger Phosphoprotein Res. 28, 167-170. Adams, S. R., Bacskai, B. J., Taylor, S. S., and Tsien, R. Y. (1991). Fluorescence ratio imaging of cyclic AMP in single cells. Nature 349, 694-697. Adashi, E. Y.,Resnick, C. E., Svoboda, M. E., and Van Wyck, J. J. (1986). Folliclestimulating hormone enhances somatomedin C binding to cultured rat granulosa cells. Evidence for cAMP dependence. J . Biol. Chem. 261, 3923-3926. Albert, D. A., Nodzenski, E., Cruz, G. M., Kuchibholtla, J., and Kowalski, J . (1991). Deoxyadenosine- and cyclic AMP-induced cell cycle arrest and cytotoxicity. Ezp. Cell Res. 197, 75-81. Ally, S., Tortora, G., Clair, T., Grieco, D., Merlo, G., Katsaros, D., Ogreid, D., Doskeland, S. O., Jahnsen, T., and Cho-Chung, Y. S. (1988). Selective modulation of protein kinase isoenzymes by the site-selective analog 8-chloroadenosine 3',5'-cyclic monophosphate provides a biological means for control of human colon cancer cell growth. Proc. Natl. Acad. Sci. U.S.A.85, 6319-6322. Ally, S., Clair, T., Katsaros, D., Tortora, G., Yokazaki, H., Finch, R. A,, Avery, T. L., and Cho-Chung, Y. S. (1989). Inhibition of growth and modulation of gene expression in human lung carcinoma in athymic mice by site-selective 8-C1-cyclic adenosine monophosphate. Cancer Res. 49, 5650-5655. Alvarez, E., Northwood, I. C., Gonzalez, F. A., Latour, D. A., Seth, A,, Abate, C., Curran, T., and Davis, R. J. (1991). Pro-Leu-Ser/Thr-Pro is a consensus primary sequence for substrate protein phosphorylation. Characterization of the phosphorylation of c-myc and c-jun proteins by an epidermal growth factor receptor threonine 669 protein kinase. J . Biol. Chem. 266, 15277-15285. Andersson, K. B., Tasken, K., and Blomhoff, H. K. (1994). Cyclic AMP downregulates c-myc expression by inhibition of transcript initiation in human B-precursor Reh cells. FEBS Lett. 337, 71-76. Angel, P., and Karin, M. (1991). The role of Jun, Fos and the AP-1 complex in cell proliferation and transformation. Biochim. Biophys. Acta. 1072, 129-157. Angel, P., Allegretto, E. A., Okino, S.,Hattori, K., Boyle, W. J.,Hunter, T., and Karin, M. (1988). Oncogene jun encodes a sequence-specific trans-activator similar to AP-1. Nature 332, 166-171. Ann, D. K., Wechsler, A., Lin, H. H., and Wang, E. (1991). Isoproterenol downregulation of statin-related gene expression in the rat parotid gland. J . Cell Sci. 100, 641-647. Arias, J., Alberts, A. S.,Brindle, P., Claret, F. X., Smeal, T., Karin, M., Feramisco, J.,and Montminy, M. (1994). Activation of cAMP and mitogen responsive genes relies on a common nuclear factor. Nature 370, 226-229. Ascoli, M. (1981). Regulation of gonadotropin receptors and gonadotropin responses in a clonal strain of Leydig tumor cells by epidermal growth factor. J. Biol. Chem. 256, 179-183. Ashall, F., Sullivan, N., and Puck, T. T. (1988). Specificity of the CAMP-induced gene exposure reaction in CHO cells. Proc. Natl. Acad.Sci. U.S.A.85, 3908-3912. Aubin, J . E., Alders, E., and Heeksche, J. N. B. (1983). A primary role for microfilaments, but not microtubules, in hormone-induced cytoplasmic retraction. Exp. Cell Res. 143, 439-450. Auwerx, J., and Sassone-Corsi, P. (1991). IP-1: A dominant inhibitor of fos-jun whose activity is modulated by phosphorylation. Cell 64, 983-993.
146
PIERRE P. ROGER et al
Auwerx, J., and Sassone-Corsi, P. (1992). Ap-1 (Fos-Jun) regulation by IP-1: Effect of signal transduction pathways and cell growth. Oncogene 7, 2271-2280. Auwerx, J.,Staels, B., and Sassone-Corsi, P. (1990). Coupled and uncoupled induction of fos and jun transcription by different second messengers in cells of hematopoietic origin. Nucleic Acids Res. 18, 221-228. Averill, L. E., Stein, R. L., and Kammer, G. M. (1988). Control of human T-lymphocyte interleukin-2 production by a CAMP-dependent pathway. Cell. Zmmunol. 115, 8899. Azuma, M., Yoshida, H., Kawamata, H., Yanagawa, T., Furumoto, N., and Sato, M. (1988). Cellular proliferation and ras oncogene of p21 21,000 expression in relation to the intracellular cyclic adenosine 3’5’-monophosphate levels of a human salivary gland adenocarcinoma cell line in culture. Cancer Res. 48,2898-2903. Bachrach, L. K., Eggo, M. C., Mak, W. W., and Burrow, G. N. (1985). Phorbol esters stimulate growth and inhibit differentiation in cultured thyroid cells. Endocrinology 116,1603-1609. Baker, S. J., Kerppola, T. K., Luk, D., Vandenberg, M. T., Marshak, D. R., Curran, T., and Abate, C. (1992).J u n is phosphorylated by several protein kinases a t the same sites that are modified in serum-stimulated fibroblasts. Mol. Cell. Biol. 12, 4694-4705. Ball, R. L., Tanner, K. D., and Carpenter, G. (1990).Epidermal growth factor potentiates cyclic AMP accumulation in A-431 cells. J . Biol. Chem. 265, 12836-12845. Bang, Y. J., Kim, S. J., Danielpour, D., O’Reilly, M. A,, Kim, K. Y., Myers, C. E., and Trepel, J. B. (1992). Cyclic AMP induces transforming growth factor Pa gene expression and growth arrest in the human adrenogen-independent prostate carcinoma cell line PC-3. Proc. Natl. Acad. Sci. U.S.A. 89, 3556-3560. Baorto, D. M., Mellado, W., and Shelanski, M. L. (1992). Astrocyte process growth induction by actin breakdown. J . Cell Biol. 117, 357-367. Baptist, M., Dumont, J . E., and Roger, P. P. (1993). Demonstration of cell kinetics in thyroid primary culture by immunostaining of proliferating cell nuclear antigen: Differences in cyclic AMP-dependent and -independent mitogenic stimulations. J . Cell Sci. 105, 69-80. Baptist, M., Lamy, F., Gannon, J., Hunt, T., Dumont, J. E., and Roger, P. P. (1995). Expression and subcellular localization of CDK2 and cdc2 kinases and their common partner cyclin A in thyroid epithelial cells: Comparison of cyclic AMPdependent and -independent cell cycles. J . Cell. Physiol., in press. Barinaga, M., Bilezikjian, L. M., Vale, W. W., Rosenfeld, M. G., and Evans, R. M. (1985). Independent effects of growth hormone releasing factor on growth hormone release and gene transcription. Nature 314, 279-281. Barka, T., Gubits, R. M., and van der Noen, H. M. (1986). Beta-adrenergic stimulation of c-fos gene expression in the mouse submandibular gland. Mol. Cell. Biol. 6, 29842989. Bartik, M. M., Brooks, W. H., and Roszman, T. L. (1993). Modulation of T cell proliferation by stimulation of the P-adrenergic receptor: Lack of correlation between inhibition of T cell proliferation and CAMPaccumulation. Cell. Zmmunol. 148, 408-421. Baserga, R. (1985). “The Biology of Cell Reproduction.” Harvard University Press, Cambridge, Massachusetts. Baserga, R. (1990). The cell cycle: Myths and realities. Cancer Res. 50, 6769-6771. Bayer, I., Mitmaker, B., Gordon, P. H., and Wang, E. (1992).Modulation of nuclear statin expression in rat thyroid follicle cell following administration of thyroid stimulating hormone. J . Cell. Physiol. 150, 276-282. Beckner, S. K., and Farrar, W. L. (1986). Interleukin 2 modulation of adenylate cyclase. J . Biol. Chem. 261, 3043-3047.
GROWTH MODULATION BY cAMP
147
Beebe, S. J., Blackmore, P. F., Chrisman, T. D., and Corbin, J. D. (1988). Use of synergistic pairs of site-selective cAMP analogs in intact cells. Methods Enzymol. 159, 118139. Bengal, E.,Ransone, L., Scharfmann, R., Dwarki, V. J., Tapscott, S. J., Weintraub, H., and Verma, I. M. (1992). Functional antagonism between c-jun and MyoD proteins: A direct physical association. Cell 68, 507-519. Ben-Ze’ev, A. (1985). The cytoskeleton in cancer cells. Biochirn. Biophys. Actu 780, 197212. Ben-Ze’ev, A. (1989). Cell shape and cell contacts: Molecular approaches to cytoskeleton expression. In “Cell Shape: Determinants, Regulation, and Regulatory Role” (W. D. Stein and F. Bronner, eds.), pp. 95-119. Academic Press, San Diego. Ben-Ze’ev, A., Baum, G., and Amsterdam, A.(1989). Regulation of tropomyosin expression in the maturing ovary and in primary granulosa cell cultures. Deu. Biol. 135, 191-201. Berger, C. L., de Buostros, A., Roos, B. A., Leong, S. S., Mendelsohn, G., Gesell, M. S., and Baylin, S. B. (1984). Human medullary thyroid carcinoma in culture provides a model relating growth dynamics, endocrine cell differentiation, and tumor progression. J . Clin. Endocrinol. Metab. 59, 338-343. Berkowitz, L.A., Raibowol, K. T., and Gilman, M. Z. (1989).Multiple sequence elements of a single functional class are required for cyclic AMP responsiveness of the mouse c-fos promoter. Mol. Cell. Biol. 9, 4272-4281. Berman, M. I., Thomas, C. G., and Nayfeh, S. N. (1987). Stimulation of inositol phosphate formation in FRTL-5 rat thyroid cells by catecholamines and its relationship to changes in 45Ca2+ eMux and cyclic AMP accumulation. Mol. Cell. Endocrcnol. 54, 151-163. Berridge, M. J. (1975). Control of cell division: A unifying hypothesis. J . Cyclic Nucleotide Res. 1, 305-320. Berridge, M.J. (1987). Inositol lipids and cell proliferation. Biochim. Biophys. Actu 907, 33-45. Bilezikjian, L. M., and Vale, W. W. (1983). Glucocorticoids inhibit corticotropin-releasing factor-induced production of adenosine 3’,5‘-monophosphate in cultured anterior pituitary cells. Endocrinology 113, 1726-1731. Billestrup, N., Swanson, L. W., and Vale, W. W. (1986). Growth hormone-releasing factor stimulates proliferation of somatotrophs in vitro. Proc. Nutl. Acad. Sci. U.S.A. 83, 6854-6857. Billestrup, N., Mitchell, R. L., Vale, W., and Verma, I. M. (1987). Growth hormonereleasing factor induces c-fos expression in cultured primary pituitary cells. Mol. Endocrinol. 1, 300-305. Bissell, M. J.,and Barcellos-Hoff, M. H. (1987). The influence of extracellular matrix on gene expression: Is structure the message? J . Cell Sci. 8, 327-343. Blackwood, E. M., and Eisenman, R. N. (1991). Max: A helix-loop-helix zipper protein that forms a sequence-specific DNA-binding complex with Myc. Science 251,12111217. Bleg, M. A,, Simon, J. C., Estevez, A. G., Jimenez de Asua, L., and Baranao, J. L. (1992). Effect of follicle-stimulating hormone on insulin-like growth factor-I-stimulated rat granulosa cell deoxyribonucleic acid synthesis. Endocrinology 131, 1223-1229. Blomhoff, H. K., Smeland, E. B., Beiske, K., Blomhoff, R., Rund, E., Bjoro, T.,PfeiferOhlsson, S., Watt, R., Fundernd, S., Godal, T., and Ohlsson, R. (1987). Cyclic AMPmediated suppression of normal and neoplastic B cell proliferation is associated with regulation of myc and Ha-ras protooncogenes. J . Cell. Physiol. 131, 426-433. Blomhoff, H.K.,Blomhoff, R., Stokke, T., de Lange Davies, C., Brevik, K., Smeland,
148
PIERRE P. ROGER et al.
E. B., Fundernd, S., and Godal, T. (1988).CAMP-mediated growth inhibition of a B-lymphoid precursor cell line reh is associated with an early transient delay in GJM, followed by an accumulation of cells in G,. J . Cell Physiol. 137, 583-587. Bockus, B. J., and Stiles, C. D. (1984). Regulation of cytoskeletal architecture by platelet-derived growth factor, insulin and epidermal growth factor. Exp. Cell Res. 153, 186-197. Bomboy, J. D., and Salmon, W. D. (1980).Effects of cyclic nucleotides on deoxyribonucleic acid synthesis in hypophysectomized rat cartilage: Stimulation of thymidine incorporation and potentiation of the action of somatomedin by analogs of adenosine 3’,5’-monophosphate or a cyclic nucleotide phosphodiesterase inhibitor. Endocrinology 107,626-632. Bongarzone, I., Pierotti, M. A., Monzini, N., Mondellini, P., Manenti, G., Donghi, R., Pilotti, S., Grieco, M., Santoro, M., Fusco, A., Vecchio, G., and Della Porta, G. (1989). High frequency of activation of tyrosine kinase oncogenes in human papillary thyroid carcinoma. Oncogene 4, 1457-1462. Bornslaeger, E. A., Mattei, P., and Schultz, R. M. (1986). Involvement of CAMPdependent protein kinase and protein phosphorylation in regulation of mouse oocyte maturation. Deu. Biol. 114,453-462. Bourne, H. R., Coffino, P., and Tomkins, G. M. (1975).Somatic genetic analysis of cyclic AMP action: Characterization of unresponsive mutants. J . Cell. Physiol. 85, 611620. Boutillier, A. L., Barthel, F., Roberts, J. L., and LaeMer, J. P. (1992).P-Adrenergic stimulation of cFOS via protein kinase A is mediated by cAMP regulatory element binding protein (CREBI-dependent and tissue-specific CREB-independent mechanisms in corticotrope cells. J . Biol. Chem. 267, 23520-23526. Boutjdir, M., Mery, P. F., Hanf, R., Shrier, A., and Fischmeister, R.(1991).High affinity forskolin inhibition of L-type CaZ+ current in cardiac cells. Mol. Pharmacol. 38, 758-765. Boyer, B., and Thiery, J. P. (1993).Cyclic AMP distinguishes between two functions of acidic FGF in a rat bladder carcinoma cell line. J. Cell Biol. 120,767-776. Boynton, A. L.,and Whitfield, J. F. (1983).The role of cyclic AMP in cell proliferation: A critical assessment of the evidence. Adu. Cyclic Nucleotide Res. 15, 193-294. Boynton, A. L.,Kleine, L. P., and Whitfield, J. F. (1985).In “Control of Animal Cell Proliferation” (A. L. Boynton and H. L. Leffert, eds.), pp. 121-129. Academic Press, Orlando. Brandi, M. L., Rotella, C. M., Mavilia, C., Franceschelli, F., Tanini, A., and Toccafondi, R. (1987).Insulin stimulates cell growth of a new strain of differentiated rat thyroid cells. Mol. Cell. Endocrinol. 54,91-103. Brautigan, D. L.,and Pinault, F. M. (1991).Activation of membrane protein-tyrosine phosphatase involving CAMP- and Caz+/phospholipid-dependent protein kinases. Proc. Natl. Acad. Sci. U.S.A.88, 6696-6700. Bravo, R., Frank, R., Blundell, P. A., and McDonald-Bravo, H. (1987a).CycliniPCNA is the auxiliary protein of DNA polymerase-delta. Nature 326, 515-517. Bravo, R., Neuberg, M., Burckhardt, J., Almendral, J., Wallich, R., and Muller, R. (1987b).Involvement of common and cell type-specific pathways in c-fos gene control: Stable induction of cAMP in macrophages. Cell 48,251-260. Breton, M.F.,Roger, P. P., Omri, B., Dumont, J. E., and Pavlovic-Hournac, M. (1989). Thyrotropin but not epidermal growth factor down-regulates the isozyme I (PKa I) of cyclic AMP-dependent protein kinases in dog thyroid cells in primary cultures. Mol. Cell. Endocrinol. 61,49-55. Brooks, R. F., Howard, M., Leake, D. S., and Riddle, P. N. (1990).Failure of platelet-
GROWTH MODULATION BY cAMP
149
derived growth factor plus insulin to stimulate sustained proliferation of Swiss 3T3 cells. J . Cell Sci. 97, 71-78. Brown, J . E., and Adamson, J. W. (1977).Modulation of in vitro erythropoiesis: Enhancement of erythroid colony growth by cyclic nucleotides. Cell Tissue Kinet. 10,289298. Browne, C. L.,Bird, M. L., and Bower, W. (1987).Effect of inhibition of the catalytic activity of cyclic AMP-dependent protein kinase on mitosis in P t K l cells. Cell Motil. Cytoskel. 7, 248-257. Browne, C. L.,Bower, W. A., Palazzo, R. E., and Rebhun, L. I. (1990).Inhibition of mitosis in fertilized sea urchin eggs by inhibition of the cyclic AMP-dependent protein kinase. Exp. Cell. Res. 188,122-128. Burch, R. M., White, M. F., and Connor, J. R. (1989).Interleukin 1 stimulates prostaglandin synthesis and cyclic AMP accumulation in Swiss 3T3 fibroblasts: Interactions between two second messenger systems. J . Cell. Physiol. 139,29-33. Burgess, G. M., Bird, G. S. J., Obie, J. F., and Putney, J . W. (1991).The mechanism for synergism between phospholipase C- and adenylylcyclase-linked hormones in liver. J . Biol. Chem. 266,4772-4781. Burton, F. H., Hasel, K. W., Bloom, F. E., Sutcliffe, J. G. (1991).Pituitary hyperplasia and gigantism in mice caused by a cholera toxin transgene. Nature 350, 74-77. Butley, M. S.,Beer, D. G., and Malkinson, A. M. (1984).Functional changes in the regulatory subunit of the type I1 cyclic adenosine 3’:5’-monophosphate-dependent protein kinase isozyme during normal and neoplastic lung development. Cancer Res. 44,2689-2697. Butley, M. S., Stoner, G. D., Beer, D. S., Mason, R. J., and Malkinson, A. M. (1985). Changes in cyclic adenosine 3’:5’-monophosphate-dependentprotein kinases during the progression of urethan-induced mouse lung tumors. Cancer Res. 45, 36773685. Bybee, A,, and Tuffery, A. R. (1988).The short-term effects of a single injection of isoproterenol on proliferation in the submandibular gland, parotid gland and oesophagus in vivo. Cell Tissue Kinet. 21, 133-141. Bybee A,, and Tuffery, A. R. (1989).Rapid proliferative response of rat thyroid gland to a single injection of TSH in vivo. J . Endocrinol. 121, 27-30. Byus, C. V.,Klimpel, G. R., Lucas, D. O., and Russell, D. H. (1977).Type I and type I1 cyclic AMP-dependent protein kinase as opposite effectors of lymphocyte mitogenesis. Nature 268, 63-64. Cambier, J . C., Newell, M. K., Justement, L. B., McGuire, J. C., Leach, K. L., and Chen, Z.Z.(1987).Ia binding ligands and cAMP stimulate nuclear translocation of PKC in B lymphocytes. Nature 327, 629-632. Campisi, J., and Pardee, A. B. (1984).Post-transcriptional control of the onset of DNA synthesis by a n insulin-like growth factor. Mol. Cell. Biol. 4, 1807-1814. Campos-Gonzalez, R., and Glenny, J. R. (1992).Tyrosine phosphorylation of mitogenactivated protein kinase in cells with tyrosine kinase negative epidermal growth factor receptors. J . Biol. Chem. 267, 14535-14538. Cantley, L. C., Auger, K. R., Carpenter, G., Duckworth, B., Graziani, A,, Kapeller, R., and Soltoff, S. (1991).Oncogenes and signal transduction. Cell 64, 281-302. Carraway, K.L.,and Carothers Carraway, C. A. (1989).Membrane-cytoskeleton interactions in animal cells. Biochim. Biophys. Acta 988,147-171. Carroll, J., Whittingham, D. G., and Wood, M. J . (1991).Effect of dibutyryl cyclic adenosine monophosphate on granulosa cell proliferation, oocyte growth and meiotic maturation in isolated mouse primary ovarian follicles cultured in collagen gels. J. Reprod. Fertil. 92, 197-207.
150
PIERRE P. ROGER et al.
Carter, R., Cosenza, S. C., Perra, A., Lipson, K., Soprano, D. R., and Soprano, K. J. (1991). A potential role for c-jun in cell cycle progression through late G, and S. Oncogene 6, 229-235. Castrillo, J. L., Theill, L. E., and Karin, M. (1991). Function of the homeodomain protein GHFl in pituitary cell proliferation. Science 253, 197-199. Chakraborty, M., Chatterjee, D., Kellokumpu, S., Rasmussen, H., and Baron, R. (1991). Cell cycle-dependent coupling of the calcitonin receptor to different G proteins. Science 251, 1078-1082. Chaldakov, G. N., Nabika, T., Nara, Y., and Yamori, Y. (1989). Cyclic AMP- and cytochalasin B-induced arborization in cultured aortic smooth muscle cells: Its cytopharmacological characterization. Cell Tissue Res. 255, 435-442. Chambard, J. C., Paris, S., L’Allemain, G., and Pouyssbgur, J. (1987). Two growth factor signalling pathways in fibroblasts distinguished by pertussis toxin. Nature 236, 800-803. Chan, D., Goate, A., and Puck, T. T. (1989). Involvement of vimentin in the reverse transformation reaction. Proc. Natl. Acad. Sci. U.S.A. 86, 2747-2751. Cheitlin, R., and Ramachandran, J. (1981).Regulation of actin in rat adrenocrotical cells by corticotropin. J . Biol. Chem. 256, 3156-3158. Chen, J., and Iyengar, R. (1994). Suppression of ras-induced transformation of NIH 3T3 by activated Galphas. Science 263, 1278-1281. Chen, J. K., and Lin, S. S. (1993). Stimulation or inhibition of A431 cell growth by EGF is directly correlated with receptor tyrosine kinase concentration but not with PLCgamma activity. Life Sci. 53, 635-642. Chen, R. H., Sarnecki, C., and Blenis, J. (1992). Nuclear localization and regulation of erk- and rsk-encoded protein kinases. Mol. Cell. Biol. 22, 915-927. Chevalier, S., Bleau, G., Roberts, K. D., and Chapdelaine, A. (1981). Proliferation and differentiation of canine prostatic epithelial cells in culture. Mol. Cell. Endocrinol. 24, 195-208. Chiu, R., Angel, P., and Karin, M. (1989).Jun-B differs in its biological properties from, and is a negative regulator of, c-Jun. Cell 59, 979-986. Chlapowski, F. J., and Nemecek, G. M. (1985). Aberrant cyclic adenosine 3’:5’-monophosphate metabolism in cultures of tumorigenic rat urothelium. Cancer Res. 45, 122-127. Chlapowski, F. J., Kelly, L. A., and Butcher, R. W. (1975).Cyclic nucleotides in cultured cells. Adu. Cyclic Nucleotide Res. 6, 245-338. Cho-Chung, Y. S. (1974). In vivo inhibition of tumor growth by cyclic adenosine 3 ’ 3 ‘ monophosphate derivatives. Cancer Res. 34, 3492-3496. Cho-Chung, Y. S. (1989). Commentary: Site-selective 8-chloro-cyclic adenosine 3’,5‘monophosphate as a biologic modulator of cancer: Restoration of normal growth control mechanisms. J . Natl. Cancer Znst. 81, 982-987. Cho-Chung, Y. S. (1990). Role of cyclic AMP receptor proteins in growth, differentiation, and suppression of malignancy: New approaches t o therapy. Cancer Res. 50, 70937100. Cho-Chung, Y. S., Clair, T., Bodwin, J. S., and Berghoffer, B. (1981). Growth arrest and morphological change of human breast cancer cells by dibutyryl CAMP and L-arginine. Science 214, 77-79. Cho-Chung, Y. S., Clair, T., Shepheard, C., and Berghoffer, B. (1983). Arrest of hormonedependent mammary cancer growth in vivo and in vitro by cholera toxin. Cancer Res. 43, 1473-1476. Choi, P. M., Tchou-Wong, K. M., and Weinstein, I. B. (1990). Overexpression of protein
GROWTH MODULATION BY cAMP
151
kinase C in HT29 colon cancer cells causes growth inhibition and tumor suppression. Mol. Cell. Biol. 10, 4650-4657. Chouaib, S., Welte, K., Mertelsmann, R., and Dupont, B. (1985).Prostaglandin E, acts a t two distinct pathways of T lymphocyte activation: Inhibition of interleukin 2 production and down-regulation of transferrin receptor expression. J . Immunol. 135, 1172-1179. Chouaib, S., Robb, R. J., Welte, K., and Dupont, B. (1987). Analysis of prostaglandin E, effect on T lymphocyte activation. J . Clin. Inuest. 80, 333-340. Christoffersen, T., and Bronstad, 0.0. (1980). A commentary on the role of cyclic nucleotides in cell growth and malignancy. Prog. Pharmacol. 4, 117-135. Christov, K., and Raichev, R. (1972). Experimental thyroid carcinogenesis. Curr. Top. Pathol. 56, 79-114. Ciardiello, F., Tortora, G . K., Clair, T., Ally, S., Salomon, D. S., and Cho-Chung, Y. S. (1990). 8-Chloro-CAMP inhibits transforming growth factor a transformation of mammary epithelial cells by restoration of the normal mRNA patterns for CAMPdependent protein kinase regulatory subunit isoforms which show disruption upon transformation. J. Biol. Chem. 265, 1016-1020. Clark, 0. H. (1981).TSH suppression in the management of thyroid nodules and thyroid cancer. World J . Surg. 5, 39-47. Clark, 0. H., and Gerend, P. L. (1985). Thyrotropin regulation of adenylate cyclase activity in human thyroid neoplasms. Surgery 97, 539-546. Clark, J. G., Kostal, K. M., and Marino, B. A. (1982). Modulation of collagen production following bleomycin-induced pulmonary fibrosis in hamsters. Presence of a factor in lung that increases fibroblast prostaglandin E, and cAMP and suppresses fibroblast proliferation and collagen production. J . B i d . Chem. 257, 8098-8105. Clark, J. G., Kostal, K. M., and Marino, B. A. (1983). Bleomycin-induced pulmonary fibrosis in hamsters. An alveolar macrophage product increases fibroblast prostaglandin E, and cyclic adenosine monophosphate and suppresses fibroblast proliferation and collagen production. J. Clin. Znuest. 72, 2082-2091. Cocks, B. G . , Vairo, G., Bodrug, S. E., and Hamilton, J. A. (1992). Suppression of growth factor-induced CYLl cyclin gene expression by antiproliferative agents. J . B i d . Chem. 267, 12307-12310. Coffino, P., Gray, J. W., and Tomkins, G. M. (1975). Cyclic AMP, a nonessential regulator of the cell cycle. Proc. Natl. Acad. Sci. U.S.A. 72, 878-882. Colletta, G., and Cirafi, A. M. (1992). TSH is able to induce cell cycle-related gene expression in rat thyroid cell. Biochem. Biophys. Res. Commun. 183, 265-272. Colletta, G., Cirafi, A. M., and Vecchio, G. (1986). Induction of the c-fos oncogene by thyrotropic hormone in rat thyroid eells in culture. Science 233, 458-460. Colletta, G., Corda, D., Schettini, G., Cirafise, A. M., Kohn, L. D., and Consiglio, E. (1988). Adenylate cyclase activity of v-ras-k transformed rat epithelial thyroid cells. FEBS Lett. 228, 37-41. Connolly, B. J., Willits, P. B., Warrington, B. H., and Murray, K. J. (1992). 8-(4Chloropheny1)thio-cyclicAMP is a potent inhibitor of the cyclic GMP-specific phosphodiesterase (PDVE Va). Biochem. Pharmacol. 44, 2303-2306. Contor, L., Lamy, F., Lecocq, R., Roger, P. P., and Dumont, J. E. (1988). Differential protein phosphorylation in the induction of thyroid cell proliferation by thyrotropin, epidermal growth factor or phorbol ester. Mol. Cell. Biol. 8, 2494-2503. Cook, P. R. (1989). The nucleoskeleton and the topology of transcription. Eur. J . Biochem. 185,487-501.
152
PIERRE P. ROGER et al
Cook, S. J., and McCormick, F. (1993). Inhibition by cAMP of Ras-dependent activation of Raf. Science 262, 1069-1072. Cooper, J. A. (1989). Related proteins are phosphorylated a t tyrosine in response to mitogenic stimuli and a t meiosis. Mol. Cell. Biol. 9, 3143-4147. Cooper, S. (1979). A unifying model for the G, period in prokaryotes and eukaryotes. Nature 280, 17-19. Cooper, H. L., Feuerstein, N., Noda, M., and Bassin, R. H. (1985). Suppression of tropomyosin synthesis, a common biochemical feature of oncogenesis by structurally diverse retroviral oncogenes. Mol. Cell. Biol. 5, 972-983. Cooper, J. A., Sefton, B. M., and Hunter, T. (1984). Diverse mitogenic agents induce the phosphorylation of two related 42,000-dalton proteins on tyrosine in quiescent chick cells. Mol. Cell. Biol. 4, 30-37. Cornelius, P., Marbowe, M., Call, K., and Pekala, P. H. (1991). Regulation of glucose transport as well as glucose transporter and immediate early gene expression in 3T3-Ll preadipocytes by 8-bromo-CAMP. J . Cell Physiol. 146, 298-308. Coulson, R., and Harrington, W. W. (1979). Renal metabolism of N6,02’-dibutyryl adenosine 3’,5’-monophosphate. Am. J . Physiol. 237, F75-F84. Cowlen, M. S., and Eling, T. E. (1992). Modulation of c-jun and jun-B messenger RNA and inhibition of DNA synthesis by prostaglandin E, in Syrian hamster embryo cells. Cancer Res. 52, 6912-6916. Crescenzi, M., Fleming, T. P., Lassar, A. B., Weintraub, M., and Aaronson, S. A. (1990). MyoD induces growth arrest independent of differentiation in normal and transformed cells. Proc. Natl. Acad. Sci. U.S.A. 87,8442-8446. Crews, C. M., and Erikson, R. C. (1993). Extracellular signals and reversible protein phosphorylation: What to mek of it all. Cell 74, 215-217. Crossin, K. L., and Carney, D. H. (1981). Evidence that microtubule depolymerization early in the cell cycle is sufficient to initiate DNA synthesis. Cell 23, 61-71. Daar, I., Yew, N., and Vande Woude, G . F. (1993). Inhibition of mos-induced oocyte maturation by protein kinase A. J . Cell Biol. 120, 1197-1202. Damante, G., and Rapoport, B. (1988). TSH stimulates the activity of the c-fos promoter in FRTL-5 rat thyroid cells. Mol. Cell. Endocrinol. 58, 279-282. Dang, C. V. (1991). c-myc oncoprotein function. Biochim. Biophys. Acta 1072, 103-113. Daniel, C. W., Silberstein, G. B., and Strickland, P. (1984). Reinitiation of growth in senescent mouse mammary epithelium in response to cholera toxin. Science 224, 1245-1247. Dash, P. K., Karl, K. A,, Colicos, M. A,, Prywes, R., and Kandel, E. R. (1991). cAMP response element-binding protein is activated by Ca2+/calmodulin- as well as CAMP-dependent protein kinase. Proc. Natl. Acud. Sci. U.S.A. 88, 5061-5065. David, C. N., and Campbell, R. D. (1972). Cell cycle kinetics and development of Hydra aftenuatu. I. Epithelial cells. J . Cell Sci. 11, 557-568. Davies, S. A,, Houslay, M. D., and Wakelam, M. J. 0. (1989). The effects of p21N-ras expression in NIH-3T3 cells upon cyclic AMP metabolism. Biochin. Biophys. Acta 1013, 173-179. Davis, R. J. (1993). The mitogen-activated protein kinase signal transduction pathway. J . Biol. Chem. 268, 14553-14556. Davis, J. B., and Stroobant, B. (1990). Platelet-derived growth factors and fibroblast growth factors are mitogens for rat Schwann cells. J . Cell Biol. 110, 1353-1360. Davis, R. J., and Czech, M. P. (1985). Tumor-promoting phorbol diesters cause the phosphorylation of epidermal growth factor receptors in normal human fibroblasts a t threonine-654. Proc. Natl. Acad. Sci. U.S.A. 82, 1974-1978. Davison, P. M., and Karasek, M. A. (1981). Human dermal microvascular endothelial
GROWTH MODULATION BY cAMP
153
cells in vitro: Effect of cyclic AMP on cellular morphology and proliferation rate. J . Cell. Physiol. 106,253-258. Dean, N. M., and Boynton, A. L. (1990).Angiotensin I1 causes phosphatidylinositol turnover and increases 1,2-diacylglycerol mass but is not mitogenic in rat liver T51B cells. Biochem. J . 269, 347-352. de Asua, L. J., O’Farrell, M. K., Clingan, D., and Rudland, P. S. (1977).Temporal sequence of hormonal interactions during the prereplicative phase of quiescent cultured 3T3 fibroblasts. Proc. Natl. Acad. Sci. U.S.A. 74, 3845-3849. de Bustros, A., Baylin, S. B., Levine, M. A., and Nelkin, B. D. (1986).Cyclic AMP and phorbol esters separately induce growth inhibition, calcitonin secretion, and calcitonin gene transcription in cultured human medullary thyroid carcinoma. J . Biol. Chern. 261,8036-8041. Decker, S. J. (1993).Transmembrane signaling by epidermal growth factor receptors lacking autophosphorylation sites. J. Biol. Chern. 268, 9176-9179. De Felici, M., Dolci, S., and Pesce, M. (1993).Proliferation of mouse primordial germ cells in vitro: A key role for CAMP.Dev. Biol. 157,277-280. de Groot, R., Auwerx, J., Karperien, M., Staels, B., and Kruijer, W. (1991).Activation of junB by PKC and PKA signal transduction through a novel cis-acting element. Nucleic Acids Res. 19,775-781. de Groot, R. P., Ballou, L. M., and Sassone-Corsi, P. (1994).Positive regulation of the CAMP-responsive activator CREM by the p70 S6 kinase: An alternative route to mitogen-induced gene expression. Cell 79,81-91. Delidow, B. C., White, B. A,, and Peluso, J. J. (1990).Gonadotropin induction of c-fos and c-myc expression and deoxyribonucleic acid synthesis in rat granulosa cells. Endocrinology 126,2302-2306. Delmas, V., Laoide, B. M., Masquilier, D., de Groot, R. P., Foulkes, N. S., and SassoneCorsi, P. (1992).Alternative usage of initiation codons in mRNA encoding the CAMP-responsive-element modulator generates regulators with opposite functions. Proc. Natl. Acad. Sci. U.S.A. 89,4226-4230. De Luca, M., Siegrist, W., Bondanza, S., Mathor, M., Cancedda, R., and Eberle, A. M. (1993).A melanocyte stimulating hormone (aMSH) stimulates normal human melanocyte growth by binding to high-affinity receptors. J . Cell Sci. 105, 10791084. Demeester-Mirkine, N.,Van Sande, J., Dor, P., Heimann, R., Cochaux, P., and Dumont, J. E. (1984).Iodide organification defect in a cold thyroid nodule: Absence of iodide effect on cyclic AMP accumulation. Clin. Endocrinol. 20, 473-479. De Pinho, R. A., Schreiber-Agus, N., and Alt, F. W. (1991).Myc family oncogenes in the development of normal and neoplastic cells. Adv. Cancer Res. 57, 1-46. Dere, W. H., and Rapoport, B. (1986).Control of growth in cultured rat thyroid cells. Mol. Cell. Endocrinol. 44, 195-199. Dere, W. H., Hirayu, H., and Rapoport, B. (1985).TSH and cAMP enhance expression of the myc proto-oncogene in cultured thyroid cells. Endocrinology 117,2249-2251. Dere, W.H., Hirayu, H., and Rapoport, B. (1986).Thyrotropin and cyclic AMP regulation of ras proto-oncogene expression in cultured thyroid cells. FEBS Lett. 196,305-308. Desdouets, C., Matesic, G., Molina, C. A., Foulkes, N. S., Sassone-Corsi, P., Brechot, C., and Sobczak-ThBpot,J. (1995).Cell cycle regulation of cyclin A gene expression by the cyclic AMP-responsive transcription factors CREB and CREM. Mol. Cell. Biol. 15,3301-3309. DiRenzo, M. F., Narsimhan, R. P., Olivero, M., Bretti, S., Giordano, S., Medico, E., Caglia, P., Zara, P., and Comoglio, P. M. (1991).Expression ofthe MetiHGF receptor in normal and neoplastic human tissues. Oncogene 6,1997-2003.
154
PIERRE P. ROGER et a1
Doljanski, L. (1930). Sur le rapport entre la proliferation et I’activite pigmentogene dans les cultures d’epithelium de l’iris. C. R. Seances Soc. Biol. Ses Fil. 105, 343-345. Dorrington, J., Chuma, A. V., and Bendell, J. J. (1988).Transforming growth factor p and follicle-stimulating hormone promote rat granulosa cell proliferation. Endocrinology 123,353-359. Doskeland, S. O., Boe, R., Bruland, T., Vintermyr, 0. K., Jastorff, B., and Lanotte, M. (1991). Criteria used to judge that a cellular response is mediated by CAMP.I n “Methodological Surveys in Biochemistry and Analysis” (E. Reid, ed.), pp. 103-114. Doskeland, S. O., Maronde, E., and Gjertsen, B. T. (1993). The genetic subtypes of CAMPdependent protein kinase-functionally different or redundant? Biochim. Biophys. Actu 1178,249-258. Dos Reis, G. A,, Nobrega, A. F., and Paes de Carvalho, R. (1986).Purinergic modulation of T-lymphocyte activation: Differential susceptibility of distinct activation steps and correlation with intracellular 3’,5’-cyclic adenosine monophosphate accumulation. Cell Immunol. 101, 213-231. Draetta, G., Beach, D., and Moran, E. (1988). Synthesis of p34, the mammalian homolog of the yeast cdc2+/CDC28 protein kinase, is stimulated during adenovirus-induced proliferation of primary baby rat kidney cells. Oncogene 2, 553-557. Dralle, H. (1989). In vivo effects of thyroid-stimulating hormone on human thyroid tissue in nude mice. I n “Growth Regulation of Thyroid Gland and Thyroid Tumors” (P. E. Goreztki and H. D. Rijher, eds.), pp. 109-122. Karger, Basel. Droms, K. A,, Haley, B. E., and Malkinson, A. M. (1987). Decreased incorporation of the photoaffinity probe 8N3-[gamma-32Pl-GTP into a 45 kD protein in lung tumors. Biochem. Biophys. Res. Commun. 144, 591-597. Droms, K. A,, Haley, B. E., Smith, G. J., and Malkinson, A. M. (1989). Decreased 8N3[y-”P]GTP photolabeling of G p in tumorigenic lung epithelial cell lines: Association with decreased hormone responsiveness and loss of contact-inhibited growth. Exp. Cell Res. 182, 330-339. Dumont, J . E. (1971).The action of thyrotropin on thyroid metabolism. Vitam.Horm. 29, 287-412. Dumont, J . E., Malone, J . F., and Van Herle, A. J. (1980). “Irradiation and Thyroid Disease: Dosimetric, Clinical and Carcinogenic Aspects.” Office for Official Publications of the European Communities, Luxembourg. Dumont, J . E., Takeuchi, A., Lamy, F., Gervy-Decoster, P., Cochaux, P., Roger, P. P., Van Sande, J., Lecocq, R., and Mockel, J. (1981). Thyroid control: An example of a complex cell regulation network. Adu. Cyclic Nucleotide Res. 14, 479-489. Dumont, J . E., Miot, F., Erneux, C., Couchie, D., Cochaux, P., Gervy-Decoster, C., Van Sande, J., and Wells, J. N. (1984). Negative regulation of cyclic AMP levels by activating of cyclic nucleotide phosphodiesterases: The example of the dog thyroid. Adu. Cyclic Nucleotide Protein Phosphorylation Res. 16, 325-336. Dumont, J . E., Jauniaux, J . C., and Roger, P. P. (1989). The cyclic AMP-mediated stimulation of cell proliferation. Trends Biochem. Sci. 14, 67-71. Ealey, P. A., Ahene, C. A., Emmerson, J. M., and Marshall, N. J . (1987). Forskolin and thyrotropin stimulation of rat FRTL-5 thyroid cell growth: The role of cyclic AMP. J.Endocrinol. 114, 199-205. Ebsworth, N. M., Taylor-Papadimitriou, J., and Rozengurt, E. (1984). Cyclic AMP does not mediate inhibition of DNA synthesis by interferon in mouse Swiss 3T3 cells. J . Cell Physiol. 120, 146-150. Eccleston, P. A. (1992). Regulation of Schwann cell proliferation: Mechanisms involved in peripheral nerve development. Exp. Cell Res. 199, 1-9. Eccleston, P. A,, Jessen, K. R., and Missky, R. (1987). Control of peripheral glial cell
GROWTH MODULATION BY cAMP
155
proliferation: A comparison of the division rates of enteric glial and Schwann cells and their response to mitogens. Deu. Biol. 124, 409-417. Eisinger, M., and Marko, 0. (1982). Selective proliferation of normal human melanocytes in vitro in the presence of phorbol ester and cholera toxin. Proc. Natl. Acad.. Sci. U.S.A. 79, 2018-2022. Eldredge, E. R., Korf, G. M., Christensen, T. A,, Connolly, D. C., Getz, M. J., and Maihle, N. J. (1994). Activation of c-fos gene expression by a kinase-deficient epidermal growth factor receptor. Mol. Cell. Biol. 14, 7527-7534. Endo, T., Shimura, H., Saito, T., and Onaya, T. (1990). Cloning of malignantly transformed rat thyroid (FRTL) cells with thyrotropin receptors and their growth inhibition by 3’,5’-cyclic adenosine monophosphate. Endocrinology 126, 1492-1497. Epstein-Almog, R., and Orly, J. (1985). Inhibition of hormone-induced steroidogenesis during cell proliferation in serum-free cultures of rat granulosa cells. Endocrinology 116,2103-2112. Erneux, C., Miot, F., and Dumont, J. E. (1985).The control mechanisms of cyclic nucleotide phosphodiesterase activities: Regulation potential of cAMP catabolism. Horm. Cell. Regul. 9, 169-183. Erneux, C., Van Sande, J., Jastorff, B., and Dumont, J. E. (1986). Modulation of cyclic AMP action in the dog thyroid by its agonist and antagonist Sp- and Rp-adenosine 3’,5‘-monophosphorothioate. Biochem. J . 234, 193-197. Escribano, J., and Rozengurt, E. (1988).Cyclic AMP increasing agents rapidly stimulate vimentin phosphorylation in quiescent cultures of Swiss 3T3 cells. J. Cell. Physiol. 137, 223-234. Espinoza, B., and Wharton, W. (1986). Effects of cholera toxin and isobutylmethylxanthine on growth of human fibroblasts. A m . J . Physiol. 251, C238-C246. Ethier, S. P., Kudla, A., and Cundiff, K. C. (1987). Influence of hormone and growth factor interactions on the proliferative potential of normal rat mammary epithelial cells in vitro. J . Cell. Physiol. 132, 161-167. Ethier, S. P., Van de Velde, R., and Cundiff, K. C. (1989).cAMP levels in proliferating rat mammary epithelial cells in vitro and in vivo. Exp. Cell Res. 182, 653-658. Fantozzi, D. A,, Taylor, S. S., Howard, P. W., Maurer, R. A,, Feranmisco, J. R., and Meinkoth, J. L. (1992). Effect of the thermostable protein kinase inhibitor on intracellular localization of the catalytic subunit of CAMP-dependent protein kinase. J . Biol. Chem. 267,16824-16828. Farmer, S. R., Wan, K. M., Ben-Ze’ev, A., and Penman, S. (1983). Regulation of actin mRNA levels and translation responds to changes in cell configuration. Mol. Cell. Biol. 3, 182-189. Farrar, W. L., and Harel-Bellan, A. (1989). Myeloid growth factodd regulation of ornithine decarboxylase: Effects of antiproliferative signals interferon-gamma and CAMP.Blood 73, 1468-1475. Farrar, W. L., Evans, S. W., Rapp, U. R., and Cleveland, J. L. (1987). Effects of antiproliferative cyclic AMP on interleukin 2-stimulated gene expression. J . Zmmunol. 139, 2075-2080. Farrar, W. L., Vinocour, M., Cleveland, J. L., and Harel-Bellan, A. (1988). Regulation of ornithine decarboxylase activity by IL-2 and cyclic AMP. J . Zmmunol. 141, 967971. Fentiman, I. S., Duhig, T., Griffiths, A. B., and Taylor-Papadimitriou, J. (1984). Cyclic AMP inhibits the growth of human breast cancer cells in defined medium. Mol. Biol. Med. 2, 81-88. Fernandez, A,, Mery, J., Vandromme, M., Basset, M., Cavadore, J. C., and Lamb, N. J. C. (1991). Effective intracellular inhibition of the CAMP-dependent protein kinase by
156
PIERRE P.ROGER et al.
microinjection of a modified form of the specific inhibitor peptide PKI in living fibroblasts. Exp. Cell Res. 195, 468-477. Ferris, C. D., Cameron, A. M., Bredt, D. S., Huganir, R. L., and Snyder, S. H. (1991). Inositol 1,4,5-trisphosphate receptor is phosphorylated by cyclic AMP-dependent protein kinase a t serines 1755 and 1589.Biochim. Biophys. Res. Commun. 175, 192198. Feuerstein, N., Lindsberg, M., Tung, L., Francis, M., and Mond, J. J. (1991). Identification of a prominent 85-kDa CAMP-dependent phosphoprotein associated with late GI phase in mitogen-stimulated B lymphocytes. J. Biol. Chem. 266, 4746-4751. Filosa, S., Pictet, R., and Rutter, W. J. (1975). Positive control of cyclic AMP on mesenchymal factor controlled DNA synthesis in embryonic pancreas. Nature 257,702705. Fine, A., and Goldstein, R. H. (1987). The effect of PGE, on the activation of quiescent lung fibroblasts. Prostaglandins 33, 903-913. Fisch, T. M., Prywes, R., Simon, M. C., and Roeder, R. G. (1989). Multiple sequence elements in the c-fos promoter mediate induction by CAMP.Genes Deu. 3, 198-211. Fitzgerald, M. C., and Flanagan, M. A. (1989). Characterization and sequence analysis of the human ornithine decarboxylase gene. DNA 8, 623-634. Folkman, J., and Moscona, A. (1978). Role of cell shape in growth control. Nature 273, 345-349. Fontana, J. A., Miksis, G., Miranda, D. M., and Durham, J. P. (1987). Inhibition of human mammary carcinoma cell proliferation by retinoids and intracellular CAMPelevating compounds. J . Natl. Cancer Znst. 78, 1107-1112. Forsberg, K., Paulsson, Y., and Westermark, B. (1988). Effect on platelet-derived growth factor-induced mitogenesis of double-stranded RNA: Evidence for a n autocrine growth inhibition mediated by interferon-beta. J . Cell. Physiol. 136, 266-272. Foster, J. W., and Kinney, D. M. (1985). ADP-ribosylating microbial toxins. Crit. Reu. Microbiol. 11, 273-298. Foti, D., Damante, G., and Rapoport, B. (1990). Studies on the role of c-fos in TSHstimulated thyroid cell proliferation. Cell. Mol. Biol. 36, 363-373. Foulkes, N. S., Borrelli, E., and Sassone-Corsi, P. (1991a). CREM gene: Use of alternative DNA-binding domains generates multiple antagonists of CAMP-induced transcription. Cell 64, 739-749. Foulkes, N. S., Laoide, B. M., Schlotter, F., and Sassone-Corsi, P. (1991b).Transcriptional antagonist CAMP-responsive element modulator (CREM) down-regulates c-fos CAMP-induced expression. Proc. Natl. Acad. Sci. U.S.A.88, 5448-5452. Foulkes, N. S., Mellstrom, B., Benusiglio, E., and Sassone-Corsi, P. (1992). Developmental switch of CREM function during spermatogenesis: From antagonist to activator. Nature 355, 80-84. Francis, J., Thompson, R., Bernal, S. D., Luk, G. D., and Baylin, S. B. (1983). Effects of dibutyryl cyclic adenosine 3’:5’-monophosphate on the growth of cultured human small-cell lung carcinoma and the specific cellular activity of L-dopa decarboxylase. Cancer Res. 43, 639-645. Freissmuth, M., and Gilman, A. G. (1989). Mutations of G,a designed to alter the reactivity of the protein with bacterial toxins. J . Biol. Chem. 264, 21907-21914. Freter, R. R., and Schultz, R. M. (1984). Regulation of murine oocyte meiosis: Evidence for a gonadotropin-induced CAMP-dependent reduction in a maturation inhibitor. J . Cell Biol. 98, 1119-1128. Friedman, D. L. (1976). Role of cyclic nucleotides in cell growth and differentiation. Physiol. Rev. 56, 652-708. Friedman, D. L., Claus, T. H., Pilkis, S. J., and Pine, G. E. (1981). Hormonal regulation of
GROWTH MODULATION BY cAMP
157
DNA synthesis in primary cultures of adult rat hepatocytes: Action of glucagon. Exp. Cell. Res. 135, 283-290. Friedrich, B., Cantrell, D. A., and Gulberg, M. (1989). Increased cyclic AMP levels block interleukin 2-induced protein kinase C substrate phosphorylation but not the mitogenic response. Eur. J. Zmmunol. 19, 1111-1116. Fukumoto, Y., Kawahara, Y., Kariya, K. I., Araki, S. I., Fukuzaki, H., and Takai, Y. (1988). Independent inhibition of DNA synthesis by protein kinase C, cyclic AMP and interferon olip in rabbit aortic smooth muscle cells. Biochem. Biophys. Res. Commun. 157,337-345. Fusco, A., Berlinghieri, M. T., Di Fiore, P. P., Portella, G., Grieco, M., and Vecchio, G. (1987). One- and two-step transformations of rat thyroid epithelial cells by retroviral oncogenes. Mol. Cell. Biol. 7, 3365-3370. Gabellini, N., Facci, L., Milani, D., Negro, A., Callegaro, L., Skaper, S. D., and Leon, A. (1991). Differences in induction of c-fos transcription by cholera toxin-derived cyclic AMP and Ca2+ signals in astrocytes and 3T3 fibroblasts. Exp. Cell. Res. 194, 210217. Galloway, P. G., Likavec, M. J., and Perry, G. (1990). Tropomyosin isoform expression in normal and neoplastic astrocytes. Lab. Invest. 62, 163-170. Garnet, L., Murat, J. C., Remaury, A., Remesi, C., Valet, P., Paris, H., and DenisPouxviel, C. (1992). Vasoactive intestinal peptide and forskolin regulate proliferation of the HT29 human colon adenocarcinoma cell line. J . Cell. Physiol. 150, 501509. Gartner, R., Greil, W., Demharter, P., and Horn, K. (1985). Involvement of cyclic AMP, iodide and metabolites of arachidonic acid in the regulation of cell proliferation of isolated porcine thyroid follicles. Mol. Cell. Endocrinol. 42, 145-155. Geahlen, R. L., and Krebs, E. G. (1980). Regulatory subunit of the type I CAMPdependent protein kinase as a n inhibitor and substrate of the cGMP-dependent protein kinase. J . Biol. Chem. 255, 1164-1169. Georgatos, S. D., Weber, K., Geisler, N., and Blobel, G. (1987). Binding of two desmin derivatives to the plasma membrane and the nuclear envelope of avian erythrocytes: Evidence for a conserved site-specificity in intermediate filament-membrane interactions. Proc. Natl. Acad. Sci. U.S.A. 84, 6780-6784. Gescher, A. (1985). Antiproliferative properties of phorbol ester tumour promoters. Biochem. Pharmacol. 34, 2587-2592. Gespach, C., Bawab, W., de Cremoux, P., and Calvo, F. (1988). Pharmacology, molecular identification and functional characteristics of vasoactive intestinal peptide receptors in human breast cancer cells. Cancer Res. 48, 5079-5083. Gerard, C., Lefort, A., Christophe, D., Libert, F., Van Sande, J., Dumont, J. E., and Vassart, G. (1989). Control of thyroperoxidase and thyroglobulin transcription by CAMP: Evidence for distinct regulatory mechanisms. Mol. Endocrinol. 3, 21102118. Ghosh-Dastidar, P., and Fox, C. F. (1984). CAMP-dependent protein kinase stimulates epidermal growth factor-dependent phosphorylation of epidermal growth factor receptors. J.Biol. Chem. 259, 3864-3869. Gil, J., Higgins, T., and Rozengurt, E. (1991). Mastoparan, a novel mitogen for Swiss 3T3 cells, stimulates pertussis toxin-sensitive arachidonic acid release without inositol phosphate accumulation. J.Cell Biol. 113, 943-950. Gill, G. N., and Lazar, C. S. (1981). Increased phosphotyrosine content and inhibition of proliferation in EGF-treated A431 cells. Nature 293, 305-307. Gill, G. N., Ill, C. R., and Watson, R. K. (1980). Kinetics of cAMP inhibition of DNA synthesis in bovine adrenocortical cells. Mol. Cell. Endocrinol. 19, 113-122.
158
PIERRE P. ROGER et a1
Gille, H., Sharrocks, A. D., and Shaw, P. E. (1992). Phosphorylation of transcription factor p62TCF by MAP kinase stimulates ternary complex formation a t c-fos promoter. Nature 358, 414-417. Ginty, D. D., Bonni, A,, and Greenberg, M. E. (1994). Nerve growth factor activates a rasdependent protein kinase that stimulates c-fos transcription via phosphorylation of CREB. Cell 77, 713-725. Glenney, J. (1986).Phospholipid-dependent Ca2+ binding by the 36-kDa tyrosine kinase substrate (calpactin) and its 33-kDa core. J.Biol. Chem. 261, 7247-7252. Goldherg, Y., Glineur, C., Gesquiere, J. C., Ricouart, A., Sap, J., Vennstrom, B., and Ghysdael, J. (1988). Activation of protein kinase C or CAMP-dependent protein kinase increases phosphorylation of the c-erbA-encoded thyroid hormone receptor and of the v-erbA-encoded protein. EMBO J. 7, 2425-2433. Gonzalez, G. A., and Montminy, M. R. (1989). Cyclic AMP stimulates somatostatin gene transcription by phosphorylation of CREB a t serine 133. Cell 59, 675-680. Gottesman, M. M., and Fleischmann, R. D. (1986). The role of CAMP in regulating tumour cell growth. Cancer Suru. 5, 291-308. Goulding, M. D., and Ralph, R. K. (1989). Cyclic-AMP-induced c-fos expression and its relevance to differentiation of a transformed mast cell line. Biochim. Biophys. Acta 1007, 99-108. Graff, I., Mockel, J., Laurent, E., Erneux, C., and Dumont, J. E. (1987).Carbachol and sodium fluoride but not TSH, stimulate the generation of inositol phosphate in the dog thyroid. FEBS Lett. 210, 204-210. Graves, L. M., Bornfeldt, K. E., Raines, E. W., Potts, B. C., Macdonalds, S. G., Ross, R., and Krebs, E. G. (1993). Protein kinase A antagonizes platelet-derived growth factor-induced signaling by mitogen-activated protein kinase in human arterial smooth muscle cells. Proc. Natl. Acad. Sci. U.S.A. 90, 10300-10304. Graziano, M. P., and Gilman, A. G. (1989). Synthesis in Escherichia coli of GTPasedeficient mutants of G p . J. Biol. Chem. 264, 15475-15482. Greenberg, M. E., Greene, L. A,, and Ziff, E. B. (1985). Nerve growth factor and epidermal growth factor induce rapid transient changes in proto-oncogene transcription in PC12 cells. J . Biol. Chem. 260, 14101-14110. Gressens, P., Hill, J. M.., Gozes, I., Fridkin, M., and Brenneman, D. E. (1993). Growth factor function of vasoactive intestinal peptide in whole cultured mouse embryos. Nature 362, 155-158. Grieco, D., Beg, Z. H., Romano, A., Bifulco, M. and Aloj, S. M. (1990). Cell cycle progrescoenzyme A reductase are regulated by thysion and 3-hydroxy-3-methylglutaryl rotropin in FRTL-5 rat thyroid cells. J. Biol. Chem. 265, 19343-19350. Grieco, D., Avvedimento, E. V., and Gotesman, M. E. (1994).A role for CAMP-dependent protein kinase in early embryonic divisions. Proc. Natl. Acad. Sci. U S A 91, 98969900. Grieco, M., Santoro, M., Berlingieri, M. T. Melillo, R. M., Donghi, R., Bongarzone, I., Pierotti, M. A,, Della Porta, G., Fusco, A., and Vecchio, G. (1990). PTC is a novel rearranged form of the ret proto-oncogene and is frequently detected in vivo in human thyroid papillary carcinomas. Cell 60, 557-563. Grinstein, S., Smith, J. D., Rowatt, C., and Dixon, S. J. (1987).Mechanism of activation of lymphocyte Na+/H+ exchange by concanavalin A. J. Biol. Chem. 262, 1527715284. Grove, R. I., and Pratt, R. M. (1984). Influence of epidermal growth factor and cyclic AMP on growth and differentiation of palatal epithelial cells in culture. Deu. Biol. 106,427-437.
GROWTH MODULATION BY CAMP
159
Grove, J. R., Price, D. J., Goodman, H. M., and Avruch, J. (1987).Recombinant fragment of protein kinase inhibitor blocks cyclic AMP-dependent gene transcription. Science 238,530-533. Guan, J. L., and Shalloway, D. (1992). Regulation of focal adhesion-associated protein tyrosine kinase by both cellular adhesion and oncogenic transformation. Nature 358,690-692. Gupta, S. K., Diez, E., Heasley, L. E., Osawa, S., and Johnson, G. L. (1990).A G protein mutant that inhibits thrombin and purinergic receptor activation of phospholipase A,. Science 249, 662-666. Gupta, S. K., Gallego, C., and Johnson, G. L. (1992).Mitogenic pathways regulated by G protein oncogenes. Mol. Biol. Cell. 3, 123-128. Gusovsky, F., and Gutkind, J . S. (1991).Selective effects of activation of protein kinase C isozymes on cyclic AMP accumulation. Mol. Pharmacol. 39, 124-129. Haegerstrand, A,, Jonzon, B., Dalsgaard, C. J., and Nilsson, J . (1989).Vasoactive intestinal polypeptide stimulates cell proliferation and adenylate cyclase activity of cultured human keratinocytes. Proc. Natl. Acad. Sci. U.S.A. 86, 5993-5996. Hafner, S., Adler, H. S., Mischak, H., Janosch, P., Heidecker, G., Wolfman, A,, Pippig, S., Lohse, M., Ueffing, M., and Kolch, W. (1994). Mechanism of inhibition of Raf-1 by protein kinase A. Mol. Cell. Biol. 14, 6696-6703. Hagiwara, M., Brindle, P., Harootunian, A., Armstrong, R., Rivier, J., Vale, W., Tsien, R., and Montminy, M. R. (1993). Coupling of hormonal stimulation and transcription via the cyclic AMP-responsive factor CREB is rate limited by nuclear entry of protein kinase A. Mol. Cell. Biol. 13, 4852-4859. Hai, T., and Curran, T. (1991). Cross-family dimerization of transcription factors FosIJun and ATFICREB alters DNA binding specificity. Proc. Natl. Acad. Sca. U.S.A. 88, 3720-3724. Halaban, R. (1988). Responses of cultured melanocytes to defined growth factors. Pigment Cell Res. 1, 18-26. Halaban, R., Langdon R., Birchall, N., Cuono, C., Baird, A,, Scott, G., Moellmann, G., and McGuire, J . (1988).Basic fibroblast growth factor from human keratinocytes is a natural mitogen for melanocytes. J . Cell Biol. 107, 1611-1619. Hall, A. (1991).Hitting the mark. Trends Biochem. Sci. 16,42-43. Hall, P. F., Charponnier, C., Nakamura, M., and Gabbiani, G. (1979). The role of microfilaments in the response of adrenal tumor cells to adrenocorticotropic hormone. J . Biol. Chem. 254, 9080-9088. Hall, S. H., Berthelon, M. C., Avallet, O., and Saez, J. M. (1991). Regulation of c-fos c-jun, jun-B, and c-myc messenger ribonucleic acids by gonadotropin and growth factors in cultured pig Leydig cell. Endocrinology 129, 1243-1249. Hamil, K. G., Conti, M., Shimasaki, S., and Hall, S. H. (1994). Follicle-stimulating hormone regulation of AP-1: Inhibition of c-jun and stimulation ofjun-B gene transcription in the rat Sertoli cell. Mol. Cell. Endocrinol. 99, 269-277. Hamilton, S. L., Codina, J., Howkes, M. J., Yatani, A., Sawada, T., Strickland, F. M., FroChner, S. C . , Spiegel, A. M., Toro, L., Stefani, E., Birnbaumer, L., and Brown, A. M. (1991). Evidence for direct interaction of G,a with the Ca2+ channel of skeletal muscle. J . Biol. Chem. 266, 19528-19535. Hammond, J. M., and English, H. E’. (1987). Regulation of deoxyribonucleic acid synthesis in cultured porcine granulosa cells by growth factors and hormones. Endocrinology 120, 1039-1046. Hammond, S. L., Ham, R. G., and Stampfer, M. R. (1984). Serum-free growth of human mammary epithelial cells: Rapid clonal growth in defined medium and extended
160
PIERRE P. ROGER et a1
serial passage with pituitary extract. Proc. Natl. Acad. Sci. U.S.A. 81, 54355439. Hanski, E. (1989). Invasive adenylate cyclase toxin of Bordetella pertussis. Trends Biochem. Sci. 14,459-463. Hargrove, J . L., and Granner, D. K. (1982).Inhibition of hepatoma cell growth by analog of adenosine and cyclic AMP and the influence of enzymes in mammalian sera. J . Cell. Physiol. 111, 232-238. Harris, H. (1990). The role of differentiation in the suppression of malignancy. J . Cell Sci. 97, 5-10. Harsh, G. R., Kavanaugh, W. M., Starksen, N. F., and Williams, L. T. (1989).Cyclic AMP blocks expression of the c-sis gene in tumor cells. Oncogene Res. 4, 65-73. Hartig, E., Loncarevic, I. F., Biischer, M., Herrlich, P., and Rahmsdorf, H. J. (1991). A new CAMP response element in the transcribed region of the human c-fos gene. Nucleic Acids Res. 19, 4153-4159. Heldin, N. E., and Westermark, B. (1988). Epidermal growth factor, but not thyrotropin, stimulates the expression of c-fos and c-myc messenger ribonucleic acid in porcine thyroid follicle cells in primary culture. Endocrinology 122, 1042-1046. Heldin, N. E., Paulsson, Y., Forsberg, K., Heldin, C. H., and Westermark, B. (1989). Induction of cyclic AMP synthesis by forskolin is followed by a reduction in the expression of c-myc messenger RNA and inhibition of 3H-thymidine incorporation in human fibroblasts. J. Cell. Physiol. 138, 17-23. Hen, R., Axel, R., and Obici, S. (1989).Activation of the p,-adrenergic receptor promotes growth and differentiation in thyroid cells. Proc. Natl. Acad. Sci. U.S.A. 86, 47854788. Hendricks, M., and Weintraub, H. (1984). Multiple tropomyosin polypeptides in chicken embryo fibroblasts: Differential repression of transcription by Rous sarcoma virus transformation. Mol. Cell. Biol. 4, 1823-1833. Herlyn, M., Mancianti, M. L., Jambrosic, S., Bolen, J. B., and Koprowski, H. (1988). Regulatory factors that determine growth and phenotype of normal human melanocytes. Exp. Cell Res. 179, 322-331. Herman, B., and Pledger, W. J. (1985). Platelet-derived growth factor-induced alterations in vinculin and actin distribution in BALBlc-3T3 cells. J . Cell Biol. 100, 1031-1040.
Hermouet, S., Merendino, J. J., Gutkind, J. S., and Spiegel, A. M. (1991). Activating and inactivating mutations of the a subunit of G,2 protein have opposite effects on proliferation of NIH 3T3 cells. Proc. Natl. Acad. Sci.U.S.A.88, 10455-10459. Hermouet, S., de Mazancourt, P., and Spiegel, A. M. (1993). Mitogenic effect of pertussis toxin-sensitive G-protein a subunits: The mitogenic action of a,2 in NIH 3T3 cells is mimicked by a,l, but not q 3 . Cell. Signal. 5, 215-225. Herschman, H. R. (1991).Primary response genes induced by growth factors and tumor promoters. Annu. Rev. Biochern. 60, 281-319. Hesketh, T. R., Morris, J. D. H., Moore, J. P., and Metcalfe, J. C. (1988). Caz+ and pH responses to sequential additions of mitogens in single 3T3 fibroblasts: Correlations with DNA synthesis. J . Biol. Chem. 263, 11879-11886. Hidaka, H., Watanabe, M., and Kobayashi, R. (1991). Properties and use of H-series compounds as protein kinase inhibitors. Methods Enzymol. 201, 328-339. Hirobe, T. (1992). Basic fibroblast growth factor stimulates the sustained proliferation of mouse epidermal melanoblasts in a serum-free medium in the presence of dibutyryl cyclic AMP and keratinocytes. Development 114, 435-445. Hiwasa, T., Tanigawara, C., and Sakiyama, S. (1987). Defects of cyclic adenosine 3':5'-
GROWTH MODULATION BY cAMP
161
monophosphate-dependent protein kinases in initiated clones derived from BALBic 3T3 mouse fibroblasts. Cancer Res. 47, 953-959. Hoffman, M. K. (1988). The requirement for high intracellular cyclic adenosine monophosphate concentrations distinguishes two pathways of B cell activation induced with lymphokines and antibody to immunoglobulin. J . Immunol. 140, 580-582. Hofmann, B., Nishanian, P., Nguyen, T., Insixiengmay, P., and Fahey, J . L. (1993). Human immunodeficiency virus proteins induce the inhibitory cAMPiprotein kinase A pathway in normal lymphocytes. Proc. Natl. Acad. Sci. U S A . 90, 66766680. Hohmann, P., DenHaese, G., and Green, R. S. (1993). Mitotic CDC2 kinase is negatively regulated by CAMP-dependent protein kinase in mouse fibroblast cell free extracts. Cell Prolif 26, 195-204. Hoke, H., Torjesen, P., Blomhoff, H. K., Rund, E., Fundernd, S., and Smeland, E. B. (1988).Cyclic AMP has the ability to influence multiple events during B cell stimulation. Eur. J. Immunol. 18, 1359-1366. Homa, S. T., Webster, S. D., and Russel, R. K. (1991).Phospholipid turnover and ultrastructural correlates during spontaneous germinal vesicle breakdown of the bovine oocyte: Effects of a cyclic AMP phosphodiesterase inhibitor. Deu. B i d . 146,461-472. Hordijk, P. L., Verlaan, I., Jalink, K., Van Corven, E. J., and Moolenaar, W. H. (1994). cAMP abrogates the p2lras-mitogen-activated protein kinase pathway in fibroblasts. J. Biol.Chem. 269, 3534-3538. Hornsby, P. J . (1985).Regulation of adrenocortical cell proliferation in culture. Endocr. Res. 10, 259-281. Hoshi, T., Garber, S. S., and Aldrich, R. W. (1988). Effect of forskolin on voltage-gated K+ channels is independent of adenylate cyclase activation. Science 240, 16521655. Howe, P. H., Bascom, C. C., Cunningham, M. R., and Leof, E. B. (1989). Regulation of transforming growth factor PI action by multiple transducing pathways: Evidence for both G protein-dependent and -independent signaling. Cancer Res. 49, 60246031. Hsie, A. W., Kawashima, K., O’Neill, J. P., and Schroder, C. H. (1975). Possible role of adenosine cyclic 3’:5’monophosphate phosphodiesterase in morphological transformation of Chinese hamster ovary cells mediated by N6, 02’-dibutyryl adenosine 3’,5’-monophosphate. J. B i d . Chem. 250, 984-989. Hsu, C. J., and Hammond, J. M. (1987). Gonadotropins and estradiol stimulate immunoreactive insulin-like growth factor-I production by porcine granulosa cells in vitro. Endocrinology 120, 198-207. Hsu, J. C., Bravo, R., and Taub, R. (1992). Interactions among LRF-1, junB, c-Jun, and c-Fos define a regulatory program in the GI phase of liver regeneration. Mol. Cell. Biol. 12, 4654-4665. Hu, J . S., and Olson, E. N. (1988). Regulation of differentiation of the BC3H1 muscle cell line through CAMP-dependent and -independent pathways. J . Biol. Chem. 263, 19670-19677. Huang, F. L., and Cho-Chung, Y. S. (1984). Hormone-regulated expression of cellular rasH oncogene in mammary carcinomas in rats. Biochem. Biophys. Res. Commun. 123, 141-147. Huang, D., Shipman-Appasamy, P. M., Orten, D. J., Hinrichs, S. H., and Prystowsky, M. B. (1994). Promoter activity of the proliferating-cell nuclear antigen gene is associated with inducible CRE-binding proteins in interleukin 2-stimulated T lymphocytes. Mol. Cell. Biol. 14, 4233-4243.
162
PIERRE P. ROGER et al
Huang, N., Wang, D., and Heppel, L. A. (1994). Role of adenosine 3’5‘-monophosphatedependent protein kinase and CAMPlevels in ATP-dependent mitogenesis in Swiss 3T3 cells. J . Biol. Chem. 269, 548-555. Hubbel, H. R., Boyer, J. E., Roane, P., and Burch, R. M. (1991). Cyclic AMP mediates the direct antiproliferative action of mismatched double-stranded RNA. Proc. Natl. Acad. Sci. U.S.A. 88, 906-910. Huddart, H., and Syson, A. J. (1975). The effect of caffeine on calcium efflux and calcium translocation in skeletal and visceral muscle. J. Exp. Biol. 63, 131-142. Hudson, L. G., Thompson, K. L., Xu, J., and Gill, G. N. (1990). Identification and characterization of a regulated promoter element in the epidermal growth factor receptor gene. Proc. Natl. Acad. Sci. U.S.A.87, 7536-7540. Hultgardh-Nilsson, A., Nilsson, J., Jonzon, B., and Dalsgaard, C. J . (1988). Growthinhibitory properties of vasoactive intestinal polypeptide. Regul. Pept. 22,267-274. Hultgardh-Nilsson, A., Querol-Ferrer, V., Jonzon, B., Krondahl, U., and Nilsson, J. (1994). Cyclic AMP, early response gene expression, and DNA synthesis in rat smooth muscle cells. Exp. Cell. Res. 214, 297-302. Hunter, T. (1989).Zn “Oncogenes and the Molecular Origins of Cancer” (R. A. Weinberg, ed.), pp. 147-173. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Hunter, T. (1991). Cooperation between oncogenes. Cell 64, 249-270. Ikeda, M., Deery, W. J., Nielsen, T. B., Ferdows, M. S., and Field, J. B. (1986). Dephosphorylation of 19K and 21K polypeptides in response to thyroid-stimulating hormone in cultured thyroid cells. Endocrinology 119, 591-599. Imagawa, M., Chiu, R., and Karin, M. (1987). Transcription factor AP-2 mediates induction by two different signal-transduction pathways: Protein kinase C and CAMP. Cell 51, 251-260. Imagawa, W., Bandyopadhyay, G. K., Wallace, D., and Nandi, S. (1988).Growth stimulation by PGE, and EGF activates cyclic AMP-dependent and -independent pathways in primary cultures of mouse mammary epithelial cells. J . Cell Physiol. 135, 509515. Imagawa, W., Bandyopadhyay, G. K., Garcia, M., Matsuzawa, A., and Nandi, S. (1992). Pregnancy-dependent to ovarian-independent progression in mammary tumors delineated in primary culture: Changes in signal transduction, growth factor regulation, and matrix interaction. Cancer Res. 52, 6531-6538. Inagaki, M., Nishi, Y., Nishizawa, K., Matsuyama, M., and Sato, C. (1987). Site-specific phosphorylation induces disassembly of vimentin filaments in vitro. Nature 328, 649-652. Ingber, D. E., and Folkman, J . (1989).Tension and compression as basic determinants of cell form and function: Utilization of a cellular tensegrity mechanism. In “Cell Shape: Determinants, Regulation, and Regulatory Role” (W. D. Stein and F. Bronner, eds.), pp. 3-31. Academic Press, San Diego. Isozaki, O., and Kohn, L. D. (1987). Control of c-fos and c-myc protooncogene induction in r a t thyroid cells in culture. Mol. Endocrinol. 1, 839-848. Ivashkiv, L. B., Liou, H. C., Kara, C. J., Lamph, W. W., Verma, I. M., and Glimcher, L. H. (1990). mXBP/CRE-BP2 and c-jun form a complex which binds to the cyclic AMP, but not to the 12-O-tetradecanoylphorbol-13-acetate,response element. Mol. Cell. Biol. 10, 1609-1621. Iwasaki, Y., Iwasaki, J., and Freake, H. C. (1983). Growth inhibition of human breast cancer cells induced by calcitonin. Biochem. Biophys. Res. Cornrnun. 110,235-242. Iwaz, J., Lafont, S., and Revillard, J . P. (1986). Elevation of cyclic 3’5’ adenosine monophosphate levels by cholera toxin inhibits the generation of interleukin 2 activity. Cell. Zmmunol. 103, 455-461.
GROWTH MODULATION BY CAMP
163
Jackowski, S., Rettenmier, C. W., and Rock, C. 0. (1990). Prostaglandin E, inhibition of growth in a colony-stimulating factor 1-dependent macrophage cell line. J . Biol. Chem. 265,6611-6616. Janet, T., Labourdette, G., Sensenbrenner, M., and Pettmann, B. (1992). Mitogenic growth factors regulate differentially early gene mRNA expression: A study on two clones of 3T3 fibroblasts. Exp. Cell Res. 198, 305-314. Jehn, B., Costello, E., Marti, A,, Keon, N., Deane, R., Li, F., Friis, R. R., Burri, P. H., Martin, F., and Jaggi, R. (1992). Overexpression of Mos, Ras, Src, and Fos inhibits mouse mammary epithelial cell differentiation. Mol. Cell. Biol. 12, 3890-3902. Jiang, H., Shalb, J . B., and Corbin, J . D. (1992). Cross-activation: Overriding cAMP/cGMP selectivities of protein kinases in tissues. Biochem. Cell. Biol. 70, 1283-1289. Jin, S., Hornicek, F. J., Neylan, D., Zakarija, M., and McKenzie, J. M. (1986). Evidence that adenosine 3’,5’-monophosphate mediates stimulation of thyroid growth in FRTL-5 cells. Endocrinology 119, 802-810. Jonzon, B., Nilsson, J., and Fredholm, B. B. (1985). Adenosine receptor-mediated changes in cyclic AMP production and DNA synthesis in cultured arterial smooth muscle cells. J . Cell. Physiol. 124, 451-456. Joost, M. G., and Steinfelder, H. J. (1987). Forskolin inhibits insulin-stimulated glucose transport in rat adipose cells by a direct interaction with the glucose transporter. Mol. Pharmacol. 31, 279-283. Juliano, R. L., and Haskill, S. (1993).Signal transduction from the extracellular matrix. J . Cell Biol. 120, 577-585. Jung, M., Kern, F. G., Jorgensen, T. J., McLeskey, S. W., Owen, C. B., and Dritschibo, A. (1994). Fibroblast growth factor-4 enhanced G, arrest and cell survival following ionizing radiation. Cancer Res. 54, 5194-5197. Jurgensen, S. R., Chock, P. B., Taylor, S., Vandenheede, J . R., and Merlevede, W. (1985). Inhibition of the Mg(II).ATP-dependent phosphoprotein phosphatase by the regulatory subunit of CAMP-dependent protein kinase. Proc. Natl. Acad. Sci. U.S.A. 82, 7565-7569. Kaczmarek, L., and Kaminska, B. (1989). Molecular biology of cell activation. Exp. Cell Res. 183, 24-35. Kamech, N., Seif, R., and Pantaloni, D. (1987). Cyclic AMP specifically blocks proliferation of rat 3T3 cells transformed by polymavirus. J . Virol. 61, 1546-1551. Kammer, G. M. (1988). The adenylate cyclase-CAMP-protein kinase A pathway and regulation of the immune response. Immunol. Today 9, 222-228. Kaszkin, M., Richards, J., and Kinzel, V. (1992). Proposed role of phosphatidic acid in the extracellular control of the transition from G, phase to mitosis exerted by epidermal growth factor in A431 cells. Cancer Res. 52, 5627-5634. Kato, J . Y., Matsuoka, M., Polyak, K., Massaque, J., and Sherr, C. J. (1994). Cyclic AMPinduced G, phase arrest mediated by a n inhibitor (p27klpl) of cyclin-dependent kinase 4 activation. Cell 79, 487-496. Katsaros, D., Tortora, G., Tagliaferri, P., Clair, T., Ally, S., Neckers, L., Robins, R. K., and Cho-Chung, Y. S. (1987). Site-selective cyclic AMP analogs provide a new approach in the control of cancer cell growth. FEBS Lett. 223, 97-103. Katsaros, D., Ally, S., and Cho-Chung, Y. S. (1988). Site-selective cyclic AMP analogues are antagonistic to estrogen stimulation of growth and proto-oncogene expression in human breast-cancer cells. Int. J. Cancer 41, 863-867. Kaupp, U. B. (1991). The cyclic nucleotide-gated channels of vertebrate photoreceptors and olfactory epithelium. Trends Neurosci. 14, 150-157. Kelvin, D. J., Simard, G., Tai, H. H., Yamaguchi, T. P., and Connolly, J . A. (1989).Growth
164
PIERRE P. ROGER et al.
factors, signaling pathways, and the regulation of proliferation and differentiation in BC3H1 muscle cells. I. A pertussis toxin-sensitive pathway is involved. J. Cell Biol. 108, 159-167. Kempski, O., Wroblewska, B., and Spatz, M. (1987). Effects of forskolin on growth and morphology of cultured glial and cerebrovascular endothelial and smooth muscle cells. Int. J . Deu. Neurosci. 5, 435-445. Keryer, G., Rios, R. M., Landmark, B. F., Skalhegg, B., Lohmann, S. M., and Bomens, M. (1993). A high-affinity binding protein for the regulatory subunit of CAMPdependent protein kinase I1 in the centrosome of human cells. Exp. Cell Res. 204, 230-240. Kessin, R. H., Fleischmann, R. D., Gottesman, M. M., Jastorff, B., and Van Lookeren Campagne, M. M. (1992). Use of the yeast low-Km CAMP-phosphodiesterase gene to control cyclic AMP levels in mammalian cells. Adu. Second Messenger Phosphoprotein Res. 25, 13-27. Khatra, B. S., Printz, R., Cobb, C. E., and Corbin, J . D. (1985). Regulatory subunit of CAMP-dependent protein kinase inhibits phosphoprotein phosphatase. Biochern. Biophys. Res. Commun. 130, 567-573. Kim, D. K., Nau, G. J., Lancki, D. W., Dawson, G., and Fitch, F. W. (1988).Cholera toxin discriminates between murine T lymphocyte proliferation stimulated by activators of protein kinase C and proliferation stimulated by IL-2. J . Immunol. 141, 34293437. Kinoshita, Y., Fukase, M., Takenaka, M., Nakada, M., Miyuchi, A., and Fujita, T. (1985). Calcitonin stimulation of cyclic adenosine 3’:5’-monophosphate production with growth inhibition in human renal adenocarcinoma cell lines. Cancer Res. 45,48904894. Kleine, L. P., and Whitfield, J . F. (1987). Serum-activated T51B rat liver cells transiently accumulate cyclic AMP-dependent protein kinases on their surfaces during the GI phase. J . Cell. Physiol. 132, 354-358. Kobierski, L. A,, Chu, H. M., Tan, Y., and Comb, M. J . (1991). CAMP-dependent regulation of proenkephalin by junD and junB: Positive and negative effects of AP-1 proteins. Proc. Natl. Acad. Sci. U.S.A 88, 10222-10226. Koff, A,, Cross, F., Fisher, A., Schumacher, J., Leguellec, K., Philippe, M., and Roberts, J . M. (1991). Human cyclin E, a new cyclin that interacts with two members of the CDC2 gene family. Cell 66, 1217-1228. Kohno, M. (1985). Diverse mitogenic agents induce rapid phosphorylation of a common set of cellular proteins at tyrosine in quiescent mammalian cells. J. Biol. Chem. 260, 1771-1779. Koike, T., Iwamoto, M., Shimazu, A., Nakashima, K., Suzuki, F., and Kato, Y. (1990). Potent mitogenic effects of parathyroid hormone (PTH) on embryonic chick and rabbit chondrocytes. Differential effects of age on growth, proteoglycan, and cyclic AMP responses of chondrocytes to PTH. J. Clin. Inuest. 85, 626-631. Kolb, J . P., Abadie, A,, Paul-Eugene, N., and Dugas, B. (1993). Intracellular signaling events associated with the induction of DNA synthesis in human B lymphocytes. Cell. Immunol. 146, 131-146. Konig, M. P., Ruchti, C., Studer, H., Berchtold, R., and Rijsler, H. (1981). Advances in thyroid neoplasia. Field Educational Italia 177-188. Kovary, K., and Bravo, R. (1991). The jun and fos protein families are both required for cell cycle progression in fibroblasts. Mol. Cell. Biol. 11, 4466-4472. Kovary, K., and Bravo, R. (1992). Existence of different FoslJun complexes during the Go-to-G, transition and during exponential growth in mouse fibroblasts: Differential role of Fos proteins. Mol. Cell. Biol. 12, 5015-5023.
GROWTH MODULATION BY cAMP
165
Kung, W., Roos, W., and Eppenberger, V. (1983). Growth stimulation of human breast cancer MCF-7 cells by dibutyryl cyclic AMP. Cell Biol. Int. Rep. 7, 345-351. Kupperman, E., Wen, W., and Meinkoth, J. L. (1993). Inhibition of thyrotropinstimulated DNA synthesis by microinjection of inhibitors of cellular ras and cyclic AMP-dependent protein kinase. Mol. Cell. Biol. 13, 4477-4484. Kuri-Harcush, W., and Marsh-Moreno, M. (1983). DNA synthesis and cell division related to adipose differentiation of 3T3 cells. J . Cell. Physiol. 114, 39-44. Kuroki, T. (1981). Induction by cholera toxin of synchronous divisions in vivo in the epidermis resulting in hyperplasia. Proc. Natl. Acad. Sci. U.S.A. 78, 6958-6962. Kuroki, T., Chida, K., Munakata, K., and Murakami, Y. (1986). Cholera toxin, a potent inducer of epidermal hyperplasia but with no tumor promoting activity in mouse skin carcinogenesis. Biochem. Biophys. Res. Commun. 137,486-492. Lacroix, A., Bolte, E., Tremblay, J., and Hamet, P. (1993). Syndrome de Cushing induit par le GIP: Expression clinique d’un recepteur ectopique. MedecineiSciences 9,706715. Lajtha, L. G. (1979). Stem cell concepts. Differentiation 14, 23-34. Lamb, P., and McKnight, S. L. (1991).Diversity and specificity in transcriptional regulation: The benefits of heterotypic dimerization. Trends Biochem. Sci. 16, 417-422. Lamb, N. J. C., Fernandez, A,, Conti, M. A,, Adelstein, R. Glass, D. B., Welsh, W. J., and Feramisco, J. R. (1988). Regulation of actin microfilament integrity in living nonmuscle cells by the CAMP-dependent protein kinase and the myosin light chain kinase. J. Cell Biol. 106, 1955-1964. Lamb, N. J. C., Fernandez, A,, Feramisco, J. R., and Welc, W. J. (1989). Modulation of vimentin containing intermediate filament distribution and phosphorylation in living fibroblasts by the CAMP-dependentprotein kinase. J. Cell Biol. 108,2409-2422. Lamb, N. J. C., Cavadore, J. C., Labbe, J. C., Maurer, R. A., and Fernandez, A. (1991). Inhibition of CAMP-dependent protein kinase plays a key role in the induction of mitosis and nuclear envelope breakdown in mammalian cells. EMBO J . 10, 15231533. Lamph, W. W., Dwarki, V. J., Ofir, R., Montminy, M., and Verma, I. M. (1990). Negative and positive regulation by transcription factor cAMP response element-binding protein is modulated by phosphorylation. Proc. Natl. Acad. Sci. U S A . 87,4320-4324. Lamy, F., Roger, P. P., Lecocq, R., and Dumont, J. E. (1986). Differential protein synthesis in the induction of thyroid cell proliferation by thyrotropin, epidermal growth factor or serum. Eur. J . Biochem. 155, 265-272. Lamy, F., Roger, P. P., Lecocq, R., and Dumont, J. E. (1989). Protein synthesis during induction of DNA replication in thyroid epithelial cells: Evidence for late markers of distinct mitogenic pathways. J . Cell. Physiol. 138, 568-576. Lamy, F., Taton, M., Dumont, J. E., and Roger, P. P. (1990). Control of protein synthesis by thyrotropin and epidermal growth factor in human thyrocytes: Role of morphological changes. Mol. Cell. Endocrinol. 73, 195-209. Lamy, F., Wilkin, F., Baptist, M., Posada, J., Roger, P. P., and Dumont, J. E. (1993). Phosphorylation of mitogen-activated protein kinases is involved in the epidermal growth factor and phorbol ester, but not in the thyrotropin/cAMP, thyroid mitogenic pathway. J . Biol. Chem. 268, 8398-8401. Landis, C. A., Masters, S. B., Spada, A., Pace, A. M., Bourne, H. R., and Vallar, L. (1989). GTPase inhibiting mutations activate the a chain of G , and stimulate adenylyl cyclase in human pituitary tumours. Nature 340, 692-696. Landis, C. A., Harsh, G., Lyons, J.,Davis, R. L., McCormick, F., and Bourne, H. R. (1990). Clinical characteristics of acromegalic patients whose pituitary tumors contain mutant G, protein. J . Clin. Endocrinol. Metab. 71, 1416-1420.
166
PIERRE P. ROGER et al.
Landschulz, W. H., Johnson, P. H., and McKnight, S. L. (1988). The leucine zipper: A hypothetical structure common to a new class of DNA binding proteins. Science 240, 1759-1764. Lange-Carter, C. A., and Malkinson, A. M. (1991). Altered regulation of mRNA levels encoding the type I isoenzyme of CAMP-dependent protein kinase in neoplastic mouse lung epithelial cells. J. Biol. Chem. 226, 22529-22536. Lange-Carter, C. A., Fossli, T., Jahnsen, T., and Malkinson, A. M. (1990). Decreased expression of the type I isozyme of CAMP-dependent protein kinase in tumor cell lines of lung epithelial origin. J. Biol. Chem. 265, 7814-7818. Lange-Carter, C. A,, Vuillequez, J . J., and Malkinson, A. M. (1993). 8-Chloroadenosine mediates 8-chloro-cyclic AMP-induced down-regulation of cyclic AMP-dependent protein kinase in normal and neoplastic mouse lung epithelial cells by a cyclic AMPindependent mechanism. Cancer Res. 53, 393-400. Langeveld, C. H., Jongenelen, C. A. M., Heimans, J . J., and Stoof, J . C. (1992). Growth inhibition of human glioma cells induced by 8-chloroadenosine, a n active metabolite of 8-chloro cyclic adenosine 3'5'-monophosphate. Cancer Res. 52, 3994-3999. Lanotte, M., Hermouet, S., Gombaud-Saintonge, G., and Dobo, I. (1986). On growth regulation of the rat promyelocytic leukemia (BNML): Growth inhibition and eradication of clonogenic cells by cholera toxin. Leuk. Res. 10, 1319-1326. Lanotte, M., Riviere, J. B., Hermouet, S., Houge, G., Vintermyr, 0. K., Gjertsen, B. T., and Doskeland, S. 0. (1991). Programmed cell death (apoptosis) is induced rapidly and with positive cooperativity by activation of cyclic adenosine monophosphatekinase I in a myeloid leukemia cell line. J . Cell. Physiol. 146, 73-80. Laugier, C., Courion, C., Pageaux, J. F., Fanidi, A., Dumas, M. Y., Sandoz, D., Nemoz, G., Prigent, A. F., and Pacheco, H. (1988). Effect of estrogen on adenosine 3',5', cyclic monophosphate in quail oviduct: Possible involvement in estradiol-activated growth. Endocrinology 122, 158-164. Laurenza, A,, McHugh-Sutkowski, E., and Seamon, K. B. (1989). Forskolin: A specific stimulator of adenylyl cyclase or a diterpene with multiple sites of action? Trends Pharmacol. Sci. 10, 442-447. Lavandero, S., Ocaranza, M. P., and Sapag-Hagar, M. (1990). Epidermal growth factor increases cyclic AMP levels in pregnant rat mammary gland explants. Res. Commun. Chem. Pathol. Pharmacol. 69, 317-323. Lazo, P. S., Dorfman, K., Noguchi, T., Mattei, M. G., and Bravo, R. (1992).Structure and mapping of the fosB gene. FosB downregulates the activity of the fosB promoter. Nucleic Acids Res. 20, 343-350. Leavitt, J., Ng, S. Y., Varma, M., Latter, G., Burbeck, S., Gunning, P., and Kedes, L. (1987). Expression of transfected mutant P-actin genes: Alterations of cell morphology and evidence for autoregulation in actin pools. Mol. Cell. Biol. 7, 2467-2476. Ledent, C., Dumont, J. E., Vasart, G., and Parmentier, M. (1992). Thyroid expression of a n A2 adenosine receptor transgene induces hyperplasia and hyperthyroidism. EMBO J . 11,537-542. Leffert, H. T., Moran, T., Sell, H., Skelly, K., Ibsen, M., Mueller, M., and Arias, I. (1978). Growth state-dependent phenotypes of adult hepatocytes in primary monolayer culture. Proc. Natl. Acad. Sci. U.S.A. 75, 1834-1838. Lefkowitz, R. J., Cotecchia, S., Sarnama, P., and Costa, T. (1993). Constitutive activity of receptors coupled to guanine nucleotide regulatory proteins. Trends Pharmacol. Sci. 14,303-307. Leitman, D. C., Fiscus, R. R., and Murad, F. (1986). Forskolin, phosphodiesterase inhibitors, and cyclic AMP analogs inhibit proliferation of cultured bovine aortic endothelial cells. J. Cell. Physiol. 127, 237-243.
GROWTH MODULATION BY CAMP
167
Lemoine, N. R., Mayall, E. S., Wyllie, F. S., Farr, C. S., Hughes, D., Padua, R. A., Thurston, V., Williams, E. D., and Wynford-Thomas, D. (1988). Activated ras oncogenes in human thyroid cancers. Cancer Res. 48, 4459-4463. Lemoine, N. R., Mayall, E. S., Wyllie, F. S., Williams, E. D., Goyns, M., Stringer, B., and Wynford-Thomas, D. (1989). High frequency of ras oncogene activation in all stages of human thyroid tumorigenesis. Oncogene 4, 159-164. Lenormand, P., Sardet, C., Pages, G., L’Allemain, G., Brunet, A., and Pouyssegur, J. (1993). Growth factors induce nuclear translocation of MAP kinases ( ~ 4 and 2 ~ p44mapkj but not of their activator MAP kinase kinase (p45mapk) in fibroblasts. J. Cell Biol. 122, 1079-1088. Leof, E. B., Wharton, W., O’Keefe, E., and Pledger, W. J. (1982). Elevated intracellular concentrations of cyclic AMP inhibited serum-stimulated, density-arrested BALBIc-3T3 cells in mid GI. J. Cell. Biochem. 19, 93-103. Lester, B. R., Stadel, J. M., Buscarino, C., Sheppard, S., Greig, R. G., and Poste, G. (1987). Reconstitution of the G, protein from B16 melanoma clones of high and low experimental metastatic potential into S49 cyc-membranes. Biochem. Biophys. Res. Commun. 147,443-451. Letterio, J. J., Coughlin, S. R., and Williams, L. T. (1986). Pertussis toxin-sensitive pathway in the stimulation of c-myc expression and DNA synthesis by bombesin. Science 234, 1117-1119. Lever, J. E. (1979). Inducers of mammalian cell differentiation stimulate dome formation in a differentiated kidney epithelial cell line (MDCK). Proc. Natl. Acad. Sci. U.S.A. 76, 1323-1327. Levin, L. R., and Zoller, M. J. (1990). Association of catalytic and regulatory subunits of cyclic AMP-dependent protein kinase requires a negatively charged side group a t a conserved threonine. Mol. Cell. Biol. 10, 1066-1075. Levin, L. R., Kuret, J.,Johnson, K. E., Powers, S., Cameron, S., Michaeli, T., Wigler, M., and Zoller, M. J. (1988). A mutation in the catalytic subunit of CAMP-dependent protein kinase that disrupts regulation. Science 240, 68-70. Levine, M. A. (1991). The McCune-Albright syndrome. The whys and wherefores of abnormal signal transduction. N . Engl. J . Med. 325, 1738-1740. Lewinski, A. (1980). Effect of dibutyryl cyclic adenosine 3’5’-monophosphate on mitotic activity in the thyroid of hypophysectomized rats. Cell Tissue Res. 211, 349352. Li, S., Grenshaw, E. B., Rawson, E. J.,Simmons, D. M., Swanson, G. W. and Rosenfeld, M. G. (1990). Dwarf locus mutants lacking three pituitary cell types result from mutations in the POU-domain gene pit-1. Nature 347, 528-533. Libert, F., Parmentier, M., Lefort, A., Dinsart, C., Van Sande, J. Maenhaut, C., Simons, M. J., Dumont, J. E., and Vassart, G. (1989). Selective amplification and cloning of four new members of the G protein-coupled receptor family. Science 244, 16771682. Ling, L., Haraguchi, K., Ohta, K., Endo, T., and Onaya, T. (1992). P,-Adrenergic receptor mRNA is overexpressed in neoplastic human thyroid tissues. Endocrinology 130 547-549. Lingk, D. S., Chan, M. A., and Gelfand, E. W. (1990). Increased cyclic adenosine monophosphate levels block progression but not initiation of human T cell proliferation. J . Immunol. 145,449-455. Lippincott-Schwartz, J.,Glickman, J.,Donaldson, J. G., Robbins, J.,Kreis, T. E., Seamon, K. B., Sheetz, M. P., and Klausner, R. D. (1991). Forskolin inhibits and reverses the effects of brefeldin A on Golgi morphology by a CAMP-independent mechanism. J. Cell B i d . 112, 567-577.
~
~
~
168
PIERRE P. ROGER et al
Livesey, S. A,, and Martin, T. J. (1988). Selective activation of the CAMP-dependent protein kinase isoenzymes. Methods Enzymol. 159, 105-118. Livesey, S.A., Ng. K. W., Collies, G., Kubota, M., Steiner, A. L., and Martin, T. J. (1985). Effect of retinoic acid on cellular content and human parathyroid hormone activation of cyclic adenosine 3’:5’-monophosphate-dependentprotein kinase isoenzymes in clonal rat osteogenic sarcoma cells. Cancer Res. 45, 5734-5740. Trivette, D. D., and Pendergast, M. (1982). Molecular events in CAMPLockwood, A. H., mediated reverse transformation. Cold Spring Harbor Symp. Quant. Biol. 46, 909919. Lockwood, A. H., Murphy, S. K., Borislow, S., Lazarus, A., and Pendergast, M. (1987). Cellular signal transduction and the reversal of malignancy. J. Cell. Biochem. 33, 237-255. Logan, A,, and Logan, S. (1991). Studies on the mechanisms of signalling and inhibition by pertussis toxin of fibroblast growth factor-stimulated mitogenesis in Balbic 3T3 cells. Cell. Signal. 3, 215-223. Lohka, M. J.,Kyes, J. L., and Maller, J. L. (1987). Metaphase protein phosphorylation in Xenopus laevis eggs. Mol. Cell. Biol. 7, 760-768. Lohmann, S. M., and Walter, U. (1984). Regulation of the cellular and subcellular concentrations and distribution of cyclic nucleotide-dependent protein kinases. Adu. Cyclic Nucleotide Protein Phosphorylation Res. 18, 63-117. Lombardi, A., Veneziani, B. M., Tramontano, D., and Ingbar, S. H. (1988). Independent and interactive effects of tetradecanoyl phorbol acetate on growth and differentiated functions of FRTL5 cells. Endocrinology 123, 1544-1552. Lyons, J.,Landis, C. A,, Harsh, L., Vallar, K., Grunewald, H., Feichtinger, H., Duh, Q. Y., Clark, D. H., Kawasaki, E., Bourne, H. R., and McCormick, F. (1990). Two G protein oncogenes in human endocrine tumors. Science 249,655-659. Maca, R. D. (1984). The effects of cyclic nucleotides on the proliferation of cultured human T-lymphocytes. Immunopharmacology 8, 53-60. Maenhaut, C., Van Sande, J., Libert, F., Abramowicz, M., Parmentier, M., Vanderhaeghen, J. J.,Dumont, J. E., Vassart, G., and Schiffmann, S. (1990). RDC8 codes for a n adenosine A2 receptor with physiological constitutive activity. Biochem. Biophys. Res. Commun. 173, 1169-1178. Maenhaut, C., Roger, P. P., Reuse, S., and Dumont, J. E. (1991). Activation of the cyclic AMP cascade as a n oncogenic mechanism: The thyroid example. Biochimie 73,2936. Magnaldo, I., Pouyssegur, J., and Paris. S. (1989a). Fibroblast growth factor potentiates receptor- and nonreceptor-mediated stimulation of adenylate cyclase in hamster fibroblasts. Cell. Signal. 1, 507-517. Magnaldo, I., Pouyssbgur, J., and Paris, S. (1989b). Cyclic AMP inhibits mitogen-induced DNA synthesis in hamster fibroblasts, regardless of the signaling pathway involved. FEBS Lett. 245, 65-69. Maller, J. L. (1987). Mitogenic signaling and protein phosphorylation in Xenopus oocytes. J . Cyclic Nucleotide Protein Phosphorylation Res. 11, 543-555. Margolis, B., Zilberstein, A,, Franks, C., Falder, S., Kremer, S., Ullrich, A., Rhee, S. G., Skorecki, K., and Schlessinger, S. (1990). Effect of phospholipase C-gamma overexpression on PDGF-induced second messengers and mitogenesis. Science 248, 607610. Martin, T. F. J., and Kowalchyk, J. A. (1981). Growth inhibition by adenosine 3’,5‘monophosphate derivatives does not require 3’,5’ phosphodiester linkage. Science 213,1120-1122.
GROWTH MODULATION BY cAMP
169
Martin, T. F. J., and Ronning, S. A. (1981). Multiple mechanisms of growth inhibition by cyclic AMP derivatives in rat GH1 pituitary cells: Isolation of an adenylate cyclasedeficient variant. J . Cell Physiol. 190, 289-297. Mary, D., Aussel, C., Ferrua, B., and Fehlmann, M. (1987). Regulation of interleukin 2 synthesis by cAMP in human T cells. J. Zmmunol. 139, 1179-1184. Masters, S. B., Miller, R. T., Chi, M. H., Chang, F. H., Beiderman, B., Lopez, N. G., and Bourne, H. R. (1989). Mutations in the GTP-binding site of G p alter stimulation of adenylyl cyclase. J . Biol. Chem. 264, 15467-15474. Matsukawa, T., and Bertram, J. S. (1988).Augmentation of postconfluence growth arrest of 10T1/2 fibroblasts by endogenous cyclic adenosine 3’:5’-monophosphate. Cancer Res. 48, 1874-1881. Matsumura, F., and Yamashiro-Matsumura, S. (1985). Purification and characterization of multiple isoforms of tropomyosin from rat cultured cells. J. Biol. Chem. 260, 13851-13859. Mayer, T. C. (1982). The control of embryonic pigment cell proliferation in culture by cyclic AMP. Deu. Biol. 94, 509-514. McCarthy, T. L., Centrella, M., and Canalis, E. (1990). Cyclic AMP induces insulin-like growth factor I synthesis in osteoblast-enriched cultures. J . Biol. Chem. 265,1535315356. McCloskey, M. A. (1988). Cholera toxin potentiates IgE-coupled inositol phospholipid hydrolysis and mediator secretion by RBL-2H3. Proc. Natl. Acad. Sci. U.S.A. 85, 7260-7264. McClung, J. K., and Kletzien, R. F. (1984). Analysis of BHK cell growth kinetics after microinjection of catalytic subunit of cyclic AMP-dependent protein kinase. Mol. Cell. Biol. 4, 1079-1085. McCormick, F. (1993). How receptors turn ras on. Nature 363, 15-16. McDonald, B. R., Gallagher, J. A., and Russel, R. G. G. (1986). Parathyroid hormone stimulates the proliferation of cells derived from human bone. Endocrinology 118, 2445-2449. McGowan, J. A,, Strain, A. J., and Bucher, N. L. (1981). DNA synthesis in primary cultures of adult rat hepatocytes in a defined medium: Effects of epidermal growth factor, insulin, glucagon, and cyclic-AMP. J . Cell Physiol. 108, 353-363. McKeehan, W. L., Adams, P. S., and Rosser, M. P. (1984). Direct mitogenic effects of insulin, epidermal growth factor, glucocorticoid, cholera toxin, unknown pituitary factors and possibly prolactin, but not androgen, on normal rat prostate epithelial cells in serum-free, primary cell culture. Cancer Res. 44, 1998-2010. McKnight, G. S., Cadd, G. G., Clegg, C. M., Otten, A. D., and Correll, L. A. (1988). Expression of wild-type and mutant subunits of the CAMP-dependent protein kinase. Cold Spring Harbor Symp. Quant. Biol. 53, 111-119. McNamee, H. P., Ingber, D. E., and Schwartz, M. A. (1993). Adhesion to fibronectin stimulates inositol lipid synthesis and enhances PDGF-induced inositol lipid breakdown. J . Cell Biol. 121, 673-678. Mechta, F., Piette, J., Hirai, S. I., and Yaniv, M. (1989). Stimulation of protein kinase C or protein kinase A mediated signal transduction pathways shows three modes of response among serum inducible genes. New Biol. 1, 297-304. Meek, W. D. (1982). Morphological evidence for cyclic AMP-induced reverse transformation in vole cells infected with avian sarcoma virus. Mol. Cell. Biol. 2, 863-874. Mehmet, H., Sinnett-Smith, J., Moore, J. P., Evan, G. I., and Rozengurt, E. (1988). Differential induction of c-fos and c-myc by cyclic AMP in Swiss 3T3 cells: Significance for the mitogenic response. Oncogene Res. 3, 281-286.
170
PIERRE P. ROGER et al
Mehmet, H., Millar, J. B. A., Lehmann, W., Higgins, T., and Rozengurt, E. (1990a). Bombesin stimulation of c-fos expression and mitogenesis in Swiss 3T3 cells: The role of prostaglandin E,-mediated cyclic AMP accumulation. Exp. Cell Res. 190, 265-270. Mehmet, H., Morris, C., and Rozengurt, E. (1990b). Multiple synergistic signal transduction pathways regulate c-fos expression in Swiss 3T3 cells: The role of cyclic AMP. Cell Growth Differ. 1, 293-298. Mehta, P. P., Bertram, J. S., and Loewenstein, W. R. (1986). Growth inhibition of transformed cells correlates with their junctional communication with normal cells. Cell 44,187-196. Meinkoth, J . L., Ji, Y., Taylor, S. S., and Feramisco, J. R. (1990). Dynamics of the distribution of cyclic AMP-dependent protein kinase in living cells. Proc. Natl. Acad. Sci. U.S.A. 87, 9595-9599. Meinkoth, J. L., Goldsmith, P. K., Spiegel, A. M., Feramisco, J. R., and Burrow, G. N. (1992). Inhibition of thyrotropin-induced DNA synthesis in thyroid follicular cells by microinjection of an antibody to the stimulatory G protein of adenylate cyclase, G,. J . Biol. Chem. 267, 13239-13245. Menapace, L., Armato, U., and Whitfield, J. F. (1987). The effects of corticotrophin (ACTH1-24), cyclic AMP and TPA (12-0-tetradecanoylphorbol-13-acetate) on DNA replication and proliferation of primary rabbit adrenocortical cells in a synthetic medium. Biochem. Biophys. Res. Commun. 148, 1295-1303. Mesiano, S., Mellon, S. H., Gospadorowicz, D., Di Blasio, A. M., and Jaffe, R. B. (1991). Basic fibroblast growth factor expression is regulated by corticotropin in the human fetal adrenal: A model for adrenal growth regulation. Proc. Natl. Acad. Sci. U.S.A. 88, 5428-5432. Meyer, T. E., Waeber, G., Lin, J., Beckmann, W., and Habener, J. F. (1993).The promoter of the gene encoding 3’,5’-cyclic adenosine monophosphate (CAMP)response element binding protein contains CAMPresponse elements: Evidence for positive autoregulation of gene transcription. Endocrinology 132, 770-780. Miller, P., Walter, U., Theurkauf, W. E., Vallee, R. B., and De Camilli, P. (1982).Frozen tissue sections a s an experimental system to reveal specific binding sites for the regulatory subunit of type I1 CAMP-dependentprotein kinase in neurons. Proc. Natl. Acad. Sci. U.S.A. 79, 5562-5566. Mine, M., Tramontano, D., Chin, W. W., and Ingbar, S. H. (1987). Interleukin-1 stimulates thyroid cell growth and increases the concentration of the c-myc protooncogene mRNA in thyroid follicular cells in culture. Endocrinology 120, 12121214. Miot, F., Erneux, C., Wells, J . N., and Dumont, J. E. (1984). The effects of alkylated xanthines on cyclic AMP accumulation in dog thyroid slices exposed to carbamylcholine. Mol. Pharmucol. 25, 261-266. Miura, M., Minomiya-Tsuji, J., Tsuji, Y., Ishibashi, S., and Ide, T. (1987). Colchicine activates cell cycle-dependent genes in growth-arrested rat 3Y1 cells. Exp. Cell Res. 173,294-298. Miyazaki, M., Wahid, S., Bai, L., and Namba, M. (1992). Effects of intracellular cyclic AMP and cyclic GMP levels on DNA synthesis of young-adult rat hepatocytes in primary culture. Ezp. Cell Res. 200,404-409. Mockel, J., Decaux, G., Unger, J., and Dumont, J. E. (1980). In vitro regulation of ornithine decarboxylase in dog thyroid slices. Endocrinology 107, 2069-2075. Mockel, J., Van Sande, J., Decoster, C., and Dumont, J. E. (1987). Tumor promoters as probes of protein kinase C in dog thyroid cell: Inhibition of the primary effects of carbamylcholine and reproduction of some distal effects. Metabolism 36, 137-143.
GROWTH MODULATION BY CAMP
171
Molloy, S., and Little, M. (1992). p34cdc2 kinase-mediated release of lamins from nuclear ghosts is inhibited by CAMP-dependentprotein kinase. Exp. Cell Res. 201,494-499. Monniaux, D. (1987). Short-term effects of FSH in vitro on granulosa cells of individual sheep follicles. J.Reprod. Fertil. 79, 505-515. Moody, T. W., Zia, F., Draoui, M., Brenneman, D. E., Fridkin, M., Davidson, A,, and Gozes, I. (1993). A vasoactive intestinal peptide antagonist inhibits non-small cell lung cancer growth. Proc. Natl. Acad. Sci. U.S.A.90, 4345-4349. Morgan, L., Jessen, K. R., and Mirsky, R. (1991).The effects of CAMP on differentiation of cultured Schwann cells: Progression from a n early phenotype (04+)to a myelin phenotype (Po', GFAP-N-CAM-, NGF-receptor- ) depends on growth inhibition. J . Cell Biol. 112, 457-467. Morris, G. F., and Mathews, M. B. (1989).Regulation of proliferating cell nuclear antigen during the cell cycle. J.Biol. Chem. 264, 13856-13864. Moss, J., and Vaughan, M. (1988).ADP-ribosylation of guanyl nucleotide-binding regulatory proteins by bacterial toxins. Adv. Enzymol. 61, 303-379. Muir, D., Varon, S., and Manthorpe, M. (1990). Schwann cell proliferation in vitro is under negative autocrine control. J. Cell Biol. 111, 2663-2671. Munari-Silem, Y., Mesnil, M., Selmi, S., Vernier-Valentin, R., Rabilloud, R., and Rousset, B. (1990). Cell-cell interactions in the process of differentiation of thyroid epithelial cells into follicles: A study by microinjection and fluorescence microscopy in in vitro reconstituted thyroid follicles. J. Cell. Physiol. 145, 414-427. Muraguchi, A,, Miyazaki, K., Kehrl, J. H., and Fauci, A. S. (1984).Inhibition ofhuman B cell activation by diterpine forskolin: Interference with B cell growth factor-induced GI to S transition of the B cell cycle. J.Zmmunol. 133, 1283-1287. Murakami, Y., Satake, M., Yamaguchi-Iwai, Y., Sakai, M., Muramatsu, M., and Ito, Y. (1991).The nuclear protooncogenes c-jun and c-fos as regulators of DNA replication. Proc. Natl. Acad. Sci. U.S.A.88, 3947-3951. Murray, S. A,, and Taylor, F. (1988). Dibutyryl cyclic AMP modulation of gap junctions in SW-13 human adrenal cortical tumor cells. A m . J. Anat. 181, 141-148. Musgrave, I. F., Genieser, H. G., Maronde, E., and Seifert, R. (1993).Preparations of Rpcyclic adenosine 3'3'-phosphorothioate (Rp-CAMPS)can contain biologically active amounts of adenosine. FEBS Lett. 318, 227-230. Muller, R., Mumberg, D., and Lucibello, F. C. (1993). Signals and genes in the control of cell-cycle progression. Biochirn. Biophys. Actu 1155, 151-179. Nagata, Y., Rosen, 0. M., Makman, M. H., and Bloom, B. R. (1984).Biochemical analysis of mutants of a macrophage cell line resistant to the growth-inhibitory activity of interferon. J . Cell Biol. 98, 1342-1347. Nair, B. G., Parikh, B., Milligan, G., and Patel, T. B. (1990). G,a mediates epidermal growth factor-elicited stimulation of rat cardiac adenylate cyclase. J . B i d . Chem. 265,21317-21322. Nakaki, T., Nakayama, M., Yamamoto, S., and Kato, R. (1986).Alpha,-adrenergic stimulation and beta,-adrenergic inhibition of DNA synthesis in vascular smooth muscle cells. Mol. Pharmacol. 37, 30-36. Nakamura, T., and Ichihara, A. (1985). Control of growth and expression of differentiated functions of mature hepatocytes in primary culture. Cell. Struct. Funct. 10, 116. Nakamura, A., Chiba, T., Yamatani, T., Yamaguchi, A,, Inui, T., Morishita, T., Kadowaki, S., and Fujita, T. (1989). Prostaglandin E, and F,, inhibit growth of human gastric carcinoma cell line Kato I11 with simultaneous stimulation of cyclic AMP production. Life Sci. 44, 75-80. Nakamura, K. D., Martinez, R., and Weber, M. J. (1983). Tyrosine phosphorylation of
172
PIERRE P. ROGER et a!.
specific proteins after mitogen stimulation of chicken embryo fibroblasts. Mol. Cell. Biol. 3, 380-390. Nakamura, T., Nakayama, Y., Teramoto, H., Nawa, K., and Ichihara, A. (1984). Loss of reciprocal modulations of growth and liver function of hepatoma cells in culture by contact with cells or cell membranes. Proc. Natl. Acad. Sci. U.S.A.81, 6398-6402. Nakamura, T., Datta, R., Sherman, M. L., and Kufe, D. (1990). Regulation of c-jun gene expression by CAMP in HL-60 myeloid leukemia cells. J . Biol. Chem. 265, 2201122015. Nakamura, T. Y., Nakayama, Y., and Ishihara, A. (1984). Reciprocal modulation of growth and liver functions of mature rat hepatocytes in primary culture by an extract of hepatic plasma membranes. J . Biol. Chem. 259, 8056-8058. Narumiya, S., Hirata, M., Nanba, T., Nikaido, T., Taniguchi, Y., Tagaya, Y., Okada, M., Mitsuya, H., and Yodoi, J. (1987). Activation of interleukin-2 receptor gene by forskolin and cyclic AMP analogs. Biochem. Biophys. Res. Commun. 143, 753-760. Nasmyth, K. A. (1990). FAR-reaching discoveries about the regulation of START. Cell 63, 1117-1120. Naville, D., Chatelain, P. G., Avallet, O., and Saez, J. M. (1990). Control of production of insulin-like growth factor I by pig Leydig and Sertoli cells cultured alone or together. Cell-cell interactions. Mol. Cell. Endocrinol. 70, 217-224. Ndubuka, C., Li, Y., and Rubin, C. S. (1993). Expression of a kinase anchor protein 75 depletes type I1 CAMP-dependentprotein kinases from the cytoplasm and sequesters the kinases in a particulate pool. J . Biol. Chem. 268, 7621-7624. Nelson, R. L., and Holian, 0. (1988). Adenylate cyclase activity and cyclic adenosine monophosphate levels in colon cancer lines and dermal fibroblasts and the effects of cholera toxin and epidermal growth factor. J . Surg. Oncol. 38, 108-112. Ness, J. M., and Kasson, B. G. (1992). Gonadotropin regulation of c-fos and c-jun messenger ribonucleic acids in cultured rat granulosa cells. Mol. Cell. Endocrinol. 90, 1725. Ng, K. W., Livesey, A., Larkin, R. G., and Martin, T. J. (1983). Calcitonin effects on growth and on selective activation of type I1 isoenzyme of cyclic adenosine 3’:5‘monophosphate-dependent protein kinase in T 47D human breast cancer cells. Cancer Res. 43, 794-800. Nicks, K. M., Droms, K. A,, Fossli, T., Smith, G. J.,and Malkinson, A. M. (1989). Altered function of protein kinase C and cyclic adenosine monophosphate-dependent protein kinase in a cell line derived from a mouse lung papillary tumor. Cancer Res. 49, 5191-5198. Nielsen, T. B., Ferdows, M. S., Brinkley, B. K., and Field, J. B. (1985). Morphological and biochemical responses of cultured thyroid cells to thyrotropin. Endocrinology 116, 788-797. Nigg, E. A,, Schafer, G., Hilz, H., and Eppenberger, H. M. (1985). Cyclic-AMPdependent protein kinase type I1 is associated with the Golgi complex and with centrosomes. Cell 41, 1039-1051. Niles, R. M., and Loewy, B. (1991). B16 mouse melanoma cells selected for resistance to cyclic AMP-mediated growth inhibition are cross-resistant to retinoic acid-induced growth inhibition. J . Cell. Physiol. 147, 176-181. Niles, R. M., Ludwig, K. W., and Makarski, J. S. (1979). Differential growth inhibition in two human carcinoma cell lines by cyclic adenosine 5’-monophosphate analogs. J . Natl. Cancer Znst. 63, 909-911. Nilsson, J., and Olsson, A. G. (1984). Prostaglandin E, inhibits DNA synthesis in arterial smooth muscle cells stimulated with platelet-derived growth factor. Atherosclerosis 53. 77-82.
GROWTH MODULATION BY cAMP
173
Nishi, N., Matuo, Y., Nakamoto, T., and Wada, F. (1988). Proliferation of epithelial cells derived from rat dorsolateral prostate in serum-free primary cell culture and their response to androgen. I n Vitro Cell. Deu. Biol. 24, 778-786. Nishida, E., and Gotoh, Y. (1992). Mitogen-activated protein kinase and cytoskeleton in mitogenic signal transduction. Znt. Rev. Cytol. 138, 211-238. Nishikawa, M., Lanerolle, P., Lincoln, T. M., and Adelstein, R. S. (1984).Phosphorylation of mammalian myosin light chain kinases by the catalytic subunit of cyclic AMPdependent protein kinase and by cyclic GMP-dependent protein kinase. J. Biol. Chem. 259, 8429-8436. Novak, T. J., and Rottenberg, E. V. (1990). CAMPinhibits induction of interleukin 2 but not of interleukin 4 in T cells. Proc. Natl. Acad. Sci. U.S.A. 87, 9353-9357. Novogrodsky, A., Patya, M., and Stenzel, K. H. (1983). Agents that increase cellular cAMP inhibit production of interleukin-2, but not its activity. Biochem. Biophys. Res. Commun. 114, 93-98. Nurse, P. (1990). Universal control mechanism regulating onset of M-phase. Nature 344, 503-508. Ochiai, A,, Yasui, W., and Tahara, E. (1985). Growth-promoting effect of gastrin on human gastric carcinoma cell line TMK-1. Jpn. J . Cancer Res. 76, 1064-1071. O’Connor, P. M., Ferris, D. K., Pagano, M., Draetta, G., Pines, J.,Hunter, T., Longo, D. L., and Kohn, K. W. (1993). G , delay induced by nitrogen mustard in human cells affects cyclin AiCDK2 and cyclin BUcdc2-kinase complexes differently. J . Biol. Chem. 268,8298-8308. O’Dorisio, M. S., Fleshman, D. J., Qualman, S. J., and ODorisio, T. M. (1992). Vasoactive intestinal peptide: Autocrine growth factor in neuroblastoma. Regul. Pept. 37,213226. Ofir, R., Dwarki, V. J., Rashid, D., and Verma, I. M. (1990). Phosphorylation of the C terminus of Fos protein is required for transcriptional transrepression of the c-fos promoter. Nature 348, 80-82. Ogreid, D., Cho-Chung, Y. S., Ekanger, R., Vintermyr, O., Haavik, J., and Doskeland, S. 0. (1987). Characterization of the cyclic adenosine 3’:5‘-monophosphate effector system in hormone-dependent and hormone-independent rat mammary carcinomas. Cancer Res. 47, 2576-2582. Okamoto, Y., Kikuchi, E., Matsumoto, M., and Nakano, H. (1993). Dibutyryl cyclic AMP enhances the down-regulation of RB protein during G, phase in the proliferating primary rat hepatocytes, but inhibits their entries into S phase and RB’s phosphorylation. Biochem. Biophys. Res. Commun. 195, 84-89. O’Keefe, E. J., and Pledger, W. J. (1983). A model of cell cycle control: Sequential events regulated by growth factors. Mol. Cell. Endocrinol. 31, 167-186. Okutani, T., Nishi, N., Kayawa, Y., Takasuga, H., Takenaka, I., Usui, T., and Wada, F. (1991).Role of cyclic AMP and polypeptide growth regulators in growth inhibition by interferon in PC-3 cells. Prostate 18, 73-80. Olashaw, N. E., and Pledger, W. J. (1988). Epidermal growth factor stimulates formation of inositol phosphates in BALBlcI3T3 cells pretreated with cholera toxin and isobutylmethylxanthine. J . Biol. Chem. 263, 1111-1114. Olashaw, N. E., Leof, E. B., O’Keefe, E. J., and Pledger, W. J. (1984). Differential sensitivity of fibroblasts to epidermal growth factor is related to cyclic AMP concentration. J . Cell. Physiol. 118, 291-297. Olashaw, N. E., Rhee, S. G., and Pledger, W. J. (1990). Cyclic AMP agonists induce the phosphorylation of phospholipase C-gamma and of a 76 kDa protein co-precipitated by anti-(phospholipase C-gamma) monoclonal antibodies in BALBlc-3T3 cells. Biochem. J. 272,297-303.
174
PIERRE P. ROGER et al.
Olinger, P. L., Benjamin, C. W., Gorman, R. R., and Connor, J. A. (1989). Cyclic AMP can partially restore platelet-derived growth factor-stimulated prostaglandin E, biosynthesis, and calcium mobilization in EJ-ras-transformed NIH-3T3 cells. J . Cell. Physiol. 139, 335-345. O’Neill, C., Riddle, P., and Rozengurt, E. (1985). Stimulating the proliferation of quiescent 3T3 fibroblasts by peptide growth factors or by agents which elevate cellular cyclic AMP level has opposite effects on motility. Exp. Cell Res. 156, 65-78. O’Neill, J. P., Schroder, C. H., and Hsie, A. W. (1975). Hydrolysis of butyryl derivatives of adenosine cyclic 3’5’-monophosphate by Chinese hamster ovary cell extracts and characterization of the products. J . Biol. Chem. 250, 990-995. Orekhov, A. N., Tertov, V. V., Hudryashov, S. A., Khashimov, K. A., and Smirnov, V. N. (1986). Primary culture of human aortic intima cells as a model for testing antiatherosclerotic drugs. Arteriosclerosis 60, 101-110. Orellana, S. A,, and McKnight, G. S. (1992). Mutations in the catalytic subunit of CAMPdependent protein kinase result in unregulated biological activity. Proc. Natl. Acad. Sci. U.S.A 89, 4726-4730. Ormerod, E. J., and Hart, I. R. (1989). Different growth responses to agents which elevate cAMP in human melanoma cell lines of high and low experimental metastatic capacity. Clin. Exp. Metastasis 7, 85-95. Osawa, S., and Johnson, G. L. (1991). A dominant negative G alpha s mutant is rescued by secondary mutation of the alpha chain amino terminus. J . Biol. Chem. 266, 4673-4676. Osborn, M., and Weber, K. (1984). Actin paracrystal induction by forskolin and by dbcAMP in CHO cells. Exp. Cell Res. 150, 408-418. O’Sullivan, C., Barton, C. M., Staddon, S. L., Brown, C. L., and Lemoine, N. R. (1991). Activating point mutations of the gsp oncogene in human thyroid adenomas. Mol. Carcinog. 4, 345-349. Otten, A. D., and McKnight, G. S. (1989). Overexpression of the type I1 regulatory subunit of the CAMP-dependent protein kinase eliminates the type I holoenzyme in mouse cells. J. Biol. Chem. 264, 20255-20260. Otten, J., Johnson, G. S., and Pastan, I. (1972). Regulation of cell growth by cyclic adenosine 3’,5’-monophosphate. Effect of cell density and agents which alter cell growth on cyclic adenosine 3‘,5’-monophosphate levels in fibroblasts. J . Biol. Chem. 247, 7082-7087. Otto, A. M. (1982). Microtubules and the regulation of DNA synthesis in fibroblastic cells (a minireview). Cell Biol. Int. Rep. 6, 1-18. Owen, N. E. (1986). Effect of prostaglandin El on DNA synthesis in vascular smooth muscle cells. Am. J . Physiol. 250, C584-C588. Pace, A. M., Wong, Y. H., and Bourne, H. R. (1991). A mutant 01 subunit of Gi2 induces neoplastic transformation of rat-1 cells. Proc. Natl. Acad. Sci. U.S.A.88, 70317035. Papadopoulos, V., and Hall, P. F. (1989). Isolation and characterization of protein kinase C from Y-1 adrenal cell cytoskeleton. J . Cell Biol. 108, 553-567. Pardee, A. B. (1989). GI events and regulation of cell proliferation. Science 246, 603608. Paris, S., and Rozengurt, E. (1982). Cyclic AMP stimulation of Na-K pump activity in quiescent Swiss 3T3 cells. J . Cell. Physiol. 112, 273-280. Park, D. J.,Min, H. K., and Rhee, S. G. (1992).Inhibition of CD3-linked phospholipase C by phorbol ester and by cAMP is associated with decreased phosphotyrosine and increased phosphoserine contents of PLC-yl. J . Biol. Chem. 267, 1496-1501.
GROWTH MODULATION BY CAMP
175
Parma, J., Duprez, L., Van Sande, J., Cochaux, P., Gervy, C., Mockel, J., Dumont, J . E., and Vassart, G. (1993). Somatic mutations in the thyrotropin receptor gene cause hyperfunctioning thyroid adenomas. Nature 365, 649-651. Pasquale, E. B., Maher, P. A., and Singer, S. J. (1986). Talin is phosphorylated on tyrosine in chicken embryo fibroblasts transformed by Rous sarcoma virus. Proc. Natl. Acad. Sci. U.S.A. 83, 5507-5511. Passareiro, H., Roger, P. P., Lamy, F., Lecocq, R., Dumont, J. E., and Nunez, J . (1985). Thyrotropin modifies the synthesis of actin and other proteins during thyroid cell culture. Eur. J. Biochem. 147,263-272. Pastan, I., and Johnson, G. S. (1974). Cyclic AMP and the transformation of fibroblasts. Adu. Cancer Res. 19, 303-329. Pastan, I., Johnson, G. S., and Anderson, W. B. (1975). Role of cyclic nucleotides in growth control. Annu. Rev. Biochem. 44,491-522. Pawelek, J. M. (1979). Evidence suggesting that a cyclic AMP-dependent protein kinase is a positive regulator of proliferation in Cloudman S91 melanoma cells. J . Cell. Physiol. 98, 619-625. Pawelek, J., Halaban, R., and Christie, G. (1975). Melanoma cells which require cyclic AMP for growth. Nature 258, 539-540. Pawelek, J., Emanuel, J., Kahn, R., Murray, M., and Fleischmann, R. (1983).Interactions between insulin and the cyclic AMP system of Cloudman S91 mouse melanoma cells. J. Cell Biochem. 21, 289-297. Payrastre, B., van Bergen en Henegouwen, P. M. P., Breton, M., den Hartigh, J. C., Plantavid, M., Verkleij, A. J., and Boonstra, J . (1992). Phosphoinositide kinase, diacylglycerol kinase, and phospholipase C activities associated to the cytoskeleton: Effect of epidermal growth factor. J. Cell Biol. 115, 121-128. Pedersen, T. (1972). In “Oogenesis” (J. D. Biggers and A. W. Schuetz, eds.), p. 361. University Park Press, Baltimore. Pelech, S. L., Sanghera, J. S., and Daya-Makin, M. (1990). Protein kinase cascades in meiotic and mitotic cell cycle control. Biochem. Cell. Biol. 68, 1297-1330. Penner, R., Matthews, G., and Neher, E. (1988). Regulation of calcium influx by second messengers in rat mast cells. Nature 334, 499-504. Pepe, S., Tortora, G., Noguchi, P. D., Marti, G. E., Washington, G. C., and Cho-Chung, Y. S. (1991). Effects of 8-chloroadenosine 3’,5’-monophosphate and N6-benzyl-cyclic adenosine 5’-monophosphate on cell cycle kinetics of HL-60 leukemia cells. Cancer Res. 51, 6263-6267. Pfarr, C. M., Mechta, F., Spyrou, G., Lallemand, D., Carillo, S., and Yaniv, M. (1994). Mouse JunD negatively regulates fibroblast growth and antagonizes transformation by ras. Cell 76, 747-760. Pillai, S., Bikle, D. D., Mancianti, M. L., Cline, P., and Hincenbergs, M. 11990).Calcium regulation of growth and differentiation of normal human keratinocytes: Modulation of differentiation competence by stages of growth and extracellular calcium. J . Cell. Physiol. 143, 294-302. Pincus, D. W., Di Cicco-Bloom, E. M., and Black, I. B. (1990a). Vasoactive intestinal peptide regulation of neuroblast mitosis and survival: Role of CAMP. Brain Res. 514, 355-357. Pincus, D. W., Di Cicco-Bloom, E. M., and Black, I. B. (1990b). Vasoactive intestinal peptide regulates mitosis, differentiation and survival of cultured sympathetic neuroblasts. Nature 343, 564-567. Pinto, A,, Aldinucci, D., Gattei, V., Zagonel, V., Tortora, G., Budillon A., and Cho-Chung, Y. S. (1992). Inhibition of the self-renewal capacity of blast progenitors from acute
176
PIERRE P. ROGER et al.
myeloblastic leukemia patients by site-selective 8-chloroadenosine 3’,5’-cyclic monophosphate. Proc. Natl. Acad. Sci. U.S.A.89, 8884-8888. Pirson, I., and Dumont, J. E. (1994). J u n B expression is regulated differently by three mitogenic pathways in thyrocytes. Exp. Cell Res. 214, 561-569. Pirson, I., Reuse, S., and Dumont, J . E. (1994).Regulation of the max gene expression by different mitogenic pathways in dog primary thyrocytes. Exp. Cell Res. 210, 33-38. Pisarev, M., Degroot, L. J., and Wilber, J. F. (1970). Cyclic-AMP production of goiter. Endocrinology 87, 339-342. Pittner, R. A,, and Fain, J . N. (1989). Exposure of cultured hepatocytes to cyclic AMP enhances the vasopressin-mediated stimulation of inositol phosphate production. Biochem. J . 257,455-460. Pledger, W. J., Stiles, C. D., Antoniades, H. N., and Scher, C. D. (1978). An ordered sequence of events is required before BALBIc-3T3 cells become committed to DNA synthesis. Proc. Natl. Acad. Sci. U.S.A. 75, 2839-2843. Pohl, V., Roger, P. P., Christophe, D., Pattyn, G., Vassart, G., and Dumont, J. E. (1990). Differentiation expression during proliferative activity induced through different pathways: In situ hybridization study of thyroglobulin gene expression in thyroid epithelial cells. J . Cell Biol. 111, 663-672. Pohl, V., Abramowicz, M., Vassart, G., Dumont, J . E., and Roger, P. P. (1993). Thyroperoxidase mRNA in quiescent and proliferating thyroid epithelial cells: Expression and subcellular localization studied by in situ hybridization. Eur. J . Cell Biol. 62, 94-104. Posada, J., and Cooper, J. A. (1992). Requirements for phosphorylation of MAP kinase during meiosis in Xenopus oocytes. Science 255, 212-215. Pouysskgur, J., and Seuwen, K. (1992).Transmembrane receptors and intracellular pathways that control cell proliferation. Annu. Reu. Physiol. 54, 195-210. Prelich, G., Tan, C. K., Kostura, M., Mathews, M. B., So, A. G., Downey, M., and Stillman, B. (1987). Functional identity of proliferating cell nuclear antigen and a DNA polymerase-delta auxiliary protein. Nature 326, 517-520. Prescott, D. M. (1987). Cell reproduction. Znt. Reu. Cytol. 100, 93-128. Presta, M., Maier, J. A. M., and Ragnotti, G. (1989). The mitogenic signaling pathway but not the plasminogen activator-inducing pathway of basic fibroblast growth factor is mediated through protein kinase C in fetal bovine aortic endothelial cells. J . Cell Biol. 109, 1877-1884. Preston, S. F., Volpi, M., Pearson, C. M., and Berlin, R. D. (1987).Regulation of cell shape in the Cloudman melanoma cell line. Proc. Natl. Acad. Sci. U.S.A.84, 5247-5251. Puck, T. T. (1984). Transformation and reverse transformation in mammalian cells. Adu. Viral Oncol. 4, 197-215. Puck, T. T. (1987). Genetic regulation of growth control: Role of cyclic AMP and cell cytoskeleton. Somatic Cell Mol. Genet. 13, 451-457. Pulver, B. J., Kyriakis, J . M., Avruch, J., Nikolakaki, E., and Woodget, J. R. (1991). Phosphorylation of c-jun mediated by MAP kinases. Nature 353, 670-674. Qureshi, S. A,, Alexandropoulos, K., Joseph, C. K., Spangler, R., and Foster, D. A. (1991). Cholera toxin induces expression of the immediate-early response gene J E via a cyclic AMP-independent signaling pathway. Mol. Cell. Biol. 11, 102-107. Rabinovitch, A., Blondel, B., Murray, T., and Mintz, D. H. (1980). Cyclic adenosine-3’,5‘monophosphate stimulates islet B cell replication in neonatal rat pancreatic monolayer cultures. J . Clin. Znuest. 66, 1065-1071. Racowsky, C. (1984). Effect of forskolin on the spontaneous maturation and cyclic AMP content of rat oocyte-cumulus complexes. J . Reprod. Fertil. 72, 107-116.
GROWTH MODULATION BY cAMP
177
Raff, M. C., Hornby-Smith, A., and Brockes, J. P. (1978). Cyclic AMP as a mitogenic signal for cultured rat Schwann cells. Nature 273, 672-673. Rajaraman, R., and Faulkner, G. (1984). Reverse transformation of Chinese hamster ovary cells by methyl xanthines. Exp. Cell Res. 154, 342-356. Ran, W., Dean, M., Levine, R. A., Henkle, C., and Campisi, J. (1986). Induction of c-fos and c-myc mRNA by epidermal growth factor or calcium ionophore is cAMP dependent. Proc. Natl. Acad. Sci. U.S.A.83, 8216-8220. Rasmussen, H., and Barrett, P. Q. (1984). Calcium messenger system: An integrated view. Physiol. Rev. 64, 938-984. Rasp& E., Reuse, S., Roger, P. P., and Dumont, J . E. (1992). Lack of correlation between the activation of the Ca2+-phosphatidylinositol cascade and the regulation of DNA synthesis in the dog thyrocyte. Exp. Cell. Res. 198, 17-26. Ravid, A., Koren, R., Narinsky, R., Rotern, C., Novogrodsky, A., and Liberman, U. A. (1990). 1,25-Dihydroxyvitamin D, and agents that increase intracellular adenosine 3‘,5’-monophosphate synergistically inhibit the mitogenic stimulation of human lymphocytes. J. Clin. Endocrinol. Metab. 70, 1687-1692. Ray, A,, Sassone-Corsi, P., and Sehgal, P. B. (1989). A multiple cytokine- and second messenger-responsive element in the enhancer of the human interleukin-6 gene: Similarities with c-fos gene regulation. Mol. Cell. Biol. 9, 5537-5547. Rebhun, L. I. (1977). Cyclic nucleotides, calcium, and cell division. Znt. Rev. Cytol. 49, 154. Redmond, O., and Tuffery, A. R. (1981). Mitotic rate of rat thyroid follicular cells in vivo in response to a single injection of thyrotropin (TSH). Cell Tissue Kinet. 14, 625631. h e , A. H., Froysa, A,, Eskild, W., Jhansen, T., and Hansson, V. (1990). Biphasic regulation of the messenger ribonucleic acid coding for the estrogen receptor by cyclic adenosine 3’:5’-monophosphate in tumor Leydig cells. Cancer Res. 50, 1528-1531. Rehfuss, R. P., Walton, K. M., Loriaux, M. M., and Goodman, R. H. (1991). The CAMPregulated enhancer-binding protein ATF-1 activates transcription in response to CAMP-dependent protein kinase A. J . Biol. Chem. 266, 18431-18434. Reuse, S., Roger, P. P., Vassart, G., and Dumont, J. E. (1986). Enhancement of c-myc mRNA concentration in dog thyrocytes initiating DNA synthesis in response to thyrotropin, forskolin, epidermal growth factor and phorbol myristate ester. Biochem. Biophys. Res. Commun. 141,1066-1076. Reuse, S . , Maenhaut, C., and Dumont, J. E. (1990). Regulation of protooncogenes c-fos and c-myc expressions by protein tyrosine kinase, protein kinase C, and cyclic AMP mitogenic pathways in dog primary thyrocytes: A positive and negative control by cyclic AMP on c-myc expression. Exp. Cell Res. 189, 33-40. Reuse, S., Pirson, I., and Dumont, J. E. (1991). Differential regulation of protooncogenes c-jun and jun D expressions by protein tyrosine kinase, protein kinase C, and cyclicAMP mitogenic pathways in dog primary thyrocytes: TSH and cyclic-AMP induce proliferation but downregulate c-jun expression. Exp. Cell Res. 196, 210-215. Richardson, J. M., Howard, P., Massa, J . S.,and Maurer, R. A. (1990). Posttranscriptional regulation of CAMP-dependent protein kinase activity by cAMP in GH3 pituitary tumor cells. Evidence for increased degradation of catalytic subunit in the presence of CAMP.J. Biol. Chem. 265, 13635-13640. Ridgway, A. A. G., De Vouge, M. W., and Mukherjee, B. B. (1988). Dibutyryl cAMP inhibits expression of transformation-related properties in Kirsten murine sarcoma virus transformed Balbic-3T3 cells despite continued presence of p2lv-Ki-ras. Biochem. Cell. Biol. 66, 54-65.
178
PIERRE P. ROGER et a1
Ridley, A. J., Davis, J . B. Stroobant, P., and Land, H. (1989). Transforming growth factors-p, and pz are mitogens for rat Schwann cells. J . Cell Biol. 109,3419-3424. Rime, H., Haccard, O., and Ozon, R. (1992). Activation of p34cdcZ kinase by cyclin is negatively regulated by cyclic AMP-dependent protein kinase in Xenopus oocytes. Deu. Biol. 151,105-110. Rixon, R. H., and Whitfield, J. F. (1985). The possible cyclic AMP-dependence of an early prereplicative event that determines mitosis in regenerating rat liver. J. Cell. Physi01. 124,397-402. Rock, C. O., Cleveland, J. L., and Jackowski, S. (1992). Macrophage growth arrest by cyclic AMP defines a distinct checkpoint in the mid-G, stage of the cell cycle and overrides constitutive c-myc expression. Mol. Cell. Biol. 12,2351-2358. Rodesch, F. (1973). Differentiation, contract inhibition and intercellular communication in retinal pigment cells. Exp. Cell Res. 76,55-62. Roger, P. P., and Dumont, J . E. (1984). Factors controlling proliferation and differentiation of canine thyroid cells cultured in reduced serum conditions: Effects of tyrotropin, cyclic AMP and growth factors. Mol. Cell. Endocrinol. 36,79-93. Roger, P. P., Hotimsky, A,, Moreau, C., and Dumont, J. E. (1982). Stimulation by thyrotropin, choleratoxin and dibutyryl cyclic AMP of the multiplication of differentiated thyroid cells in vitro. Mol. Cell. Endocrinol. 26, 165-176. Roger, P. P., Servais, P., and Dumont, J. E. (1983).Stimulation by thyrotropin and cyclic AMP of the proliferation of quiescent canine thyroid cells cultured in a defined medium containing insulin. FEBS Lett. 157,323-329. Roger, P. P., Van Heuverswijn, B., Lambert, C., Reuse, S., Vassart, G., and Dumont, J. E. (1985). Antagonistic effects of thyrotropin and epidermal growth factor on thyroglobulin mRNA level in cultured thyroid cells. Eur. J . Biochern. 152,239-245. Roger, P. P., Reuse, S., Servais, P., Van Heuverswijn, B., and Dumont, J. E. (1986). Stimulation of cell proliferation and inhibition of differentiation expression by tumor-promoting phorbol esters in dog thyroid cells in primary culture. Cancer Res. 46,898-906. Roger, P. P., Servais, P., and Dumont, J. E. (1987a). Regulation of dog thyroid epithelial cell cycle by forskolin, an adenylate cyclase activator. Exp. Cell Res. 172,282-292. Roger, P. P., Servais, P., and Dumont, J . E. (198713). Induction of DNA synthesis in dog thyrocytes in primary culture: Synergistic effects of thyrotropin and cyclic AMP with epidermal growth factor and insulin. J . Cell Physiol. 130,58-67. Roger, P. P., Rickaert, F., Huez, G., Authelet, M., Hofmann, F., and Dumont, J . E. (1988a). Microinjection of catalytic subunit of cyclic AMP-dependent protein kinases triggers acute morphological changes in thyroid epithelial cells. FEBS Lett. 232,409-413. Roger, P. P., Taton, M., Van Sande, J., and Dumont, J. E. (1988b). Mitogenic effects of thyrotropin and adenosine 3',5'-monophosphate in differentiated normal human thyroid cells in vitro. J. Clin. Endocrinol. Metab. 66,1158-1165. Roger, P. P., Rickaert, F., Lamy, F., Authelet, M., and Dumont, J . E. (1989). Actin stress fiber disruption and tropomyosin isoform switching in normal thyroid epithelial cells stimulated by thyrotropin and phorbol esters. Ezp. Cell Res. 182,1-13. Roger, P. P., Breton, M. F., Dumont, J. E., and Pavlovic-Hournac, M. (1991). Epidermal growth factor and phorbol ester actions on the TSH induced down regulation of the isoenzyme I (PKAI) of cyclic AMP-dependent protein kinases in dog thyroid cell primary cultures. Cell. Signal. 3,145-151. Roger, P. P., Baptist, M., and Dumont, J. E. (1992). A mechanism generating heterogeneity in thyroid epithelial cells: Suppression of the thyrotropinlcAMP-dependent
GROWTH MODULATION BY cAMP
179
mitogenic pathway after cell division induced by CAMP-independent factors. J . Cell Biol. 117, 383-393. Rollins, B. J., and Stiles, C. D. (1989).Serum-inducible genes. Adu. Cancer Res. 53,l-32. Rossomando, A. J., Payne, D. M., Weber, M. J., and Sturgill, T. W. (1989). Evidence that pp42, a major tyrosine kinase target protein, is a mitogen-activated serinelthreonine protein kinase. Proc. Natl. Acad. Sci. U.S.A. 86, 6940-6943. Roth, C. W., Richert, N., Pastan, I., and Gottesman, M. M. (1983). Cyclic AMP treatment of Rous sarcoma virus-transformed Chinese hamster ovary cells increases phosphorylation of pp6Osrc and increases pp6@rc kinase activity. J. Biol. Chem. 258, 1076810773. Rothermel, J. D., Jastorff, B., and Bothello, L. H. 11984).Inhibition of glucagon-induced glycogenolysis in isolated rat hepatocytes by the Rp diastereomer of adenosine cyclic 3',5'-phosphorothioate. J . Biol. Chem. 259, 8151-8155. Roux, P., Blanchard, J. M., Fernandez, A., Lamb, N., Jeanteur, P., and Piechaczyk, M. (1990). Nuclear localization of c-Fos, but not v-Fos proteins, is controlled by extracellular signals. Cell 63, 341-351. Roy, S. K., and Greenwald, G. S. (1986). Effects of FSH and LH on incorporation of ["Hlthymidine into follicular DNA. J. Reprod. Fertil. 78, 201-209. Roy, S. K., and Greenwald, G. S. (1988). In vitro effects of follicle-stimulating hormone, luteinizing hormone, and prolactin on follicular deoxyribonucleic acid synthesis in the hamster. Endocrinology 122, 952-958. Roy, S. K., and Greenwald, G. S. (1989). Hormonal requirements for the growth and differentiation of hamster preantral follicles in long-term culture. J. Reprod. Fertil. 87, 103-114. Roy, S. K., and Greenwald, G. S. (1991).Mediation of follicle-stimulating hormone action on follicular deoxyribonucleic acid synthesis by epidermal growth factor. Endocrinology 129, 1903-1908. Rozengurt, E. (1982a). Adenosine receptor activation in quiescent Swiss 3T3 cells. Enhancement of CAMP levels, DNA synthesis and cell division. Exp. Cell Res. 139,7178. Rozengurt, E. (1982b). Synergistic stimulation of DNA synthesis by cyclic AMP derivatives and growth factors in mouse 3T3 cells. J. Cell. Physiol. 112, 243-250. Rozengurt, E. (1986). Early signals in the mitogenic response. Science 234, 161-166. Rozengurt, E. (1989). Signal transduction pathways in mitogenesis. Br. Med. Bull. 45, 515-528. Rozengurt, E., and Mendoza, S. A. (1985).Synergistic signals in mitogenesis: Role of ion fluxes, cyclic nucleotides and protein kinase C in Swiss 3T3 cells. J . Cell Sci., Suppl. 3,229-242. Rozengurt, E., Collins, M. K. L., and Keehan, M. (1983a). Platelet-derived growth factor elicits cyclic AMP accumulation in Swiss 3T3 cells: Role of prostaglandin production. J. Cell. Physiol. 116, 379-384. Rozengurt, E., Stroobant, P., Waterfield, M. D., Deuel, T. F., and Keehan, M. (1983133. Platelet-derived growth factor elicits cyclic AMP accumulation in Swiss 3T3 cells: Role of prostaglandin production. Cell 34, 265-272. Rozengurt, E., Murray, M., Zachary, I., and Collins, M. (1987). Protein kinase C activation enhances cAMP accumulation in Swiss 3T3 cells: Inhibition by pertussis toxin. Proc. Natl. Acad. Sci. U.S.A.84, 2282-2286. Ruellan, C., Scemama, J. L., Clerc, P., Fagot-Revurat, P., Clemente, F., and Ribet, A. (1986). VIP regulation of a human pancreatic cancer cell line: Capan-1. Peptides 7(Suppl. l),267-271.
180
PIERRE P. ROGER et al
Ryseck, R. P., and Bravo, R. (1991). c-jun, jun B, and jun D differ in their binding affinities to AP-1 and CRE consensus sequences: Effects of FOS proteins. Oncogene 6,533-542. Saez, J . C., Nairn, A. C., Czernik, A. J., Spray, D. C., Hertzberg, E. L., Greengard, P., and Bennett, M. V. L. (1990). Phosphorylation of connexin 32, a hepatocyte gap-junction protein, by CAMP-dependent protein kinase, protein kinase C and Ca%+ icalmodulindependent protein kinase 11. Eur. J . Biochem. 192,263-273. Saez, J . C., Berthoud, V. M., Moreno, A. P., and Spray, D. C. (1993). Multiplicity of controls in differentiated and undifferentiated cells and possible functional implications. Adu. Second Messenger Phosphoprotein Res. 27, 163-198. Saikumar, P., Ulsh, L. S., Clanton, D. J., Huang, K. P., and Shih, T. Y. (1988). Novel phosphorylation of c-ras p21 by protein kinases. Oncogene Res. 3, 213-222. Saito, T., Lamy, F., Roger, P. P., Lecocq, R., and Dumont, J . E. (1994). Characterization and identification as cofilin and destrin of two thyrotropin and phorbol ester regulated proteins in thyroid cells. Exp. Cell Res. 212, 49-61. Sandberg, M., Butt, E., Nolte, C., Fisher, L., Halbrugge, M., Beltman, J., Johnsen, T., Genieser, M. G., Jastorff, B. and Walter, U. (1991). Characterization of Sp-5,6dichloro-l-~-~-ribofuranosyl-benzimidazole-3‘,5’-monophosphorothioate (Sp-5,6-DClcBiMPS) as a potent and specific activator of cyclic-AMP-dependent protein kinase in cell extracts and intact cells. Biochem. J . 279, 521-527. Sassone-Corsi, P., Ransone, L. J., and Verma, I. M. (1990). Cross-talk in signal transduction: TPA-inducible factor juniAP-1 activates CAMP-responsive enhancer elements. Oncogene 5,427-431. Sato, E., and Koide, S. S. (1987). Biochemical transmitters regulating the arrest and resumption of meiosis in oocytes. Int. Rev. Cytol. 106, 1-33. Savion, N., Lui, G. M., Laherty, R., and Gospodarowicz, D. (1981). Factors controlling proliferation and progesterone production by bovine granulosa cells in serum-free medium. Endocrinology 109,409-420. Schimmer, B. P., Wong, M., O’Brien, D., and Schultz, P. (1986). Recovery of hormonal regulation in protein kinase defective adrenal cells through DNA-mediated gene transfer. J . Cell. Physiol. 126, 77-83. Schmidt, W., Poll-Jordan, G., and LoMer, G. (1990). Adipose conversion of 3T3-Ll cells in a serum-free culture system depends on epidermal growth factor, insulin-like growth factor I, corticosterone, and cyclic AMP. J. Biol. Chem. 265, 15489-15495. Scholar, E. M., and Sudhor, P. (1991). Stimulation of tumor cell growth by vasoactive intestinal peptide. Cancer Res. 67, 1561-1564. Schwindinger, W. F., Francomano, C. A., and Levine, M. A. (1992). Identification of a mutation in the gene encoding the a subunit of the stimulatory G protein of adenylyl cyclase in McCune-Albright syndrome. Proc. Natl. Acad. Sci. U S A 89, 5152-5156. Scott, J . D., and McCartney, S. (1994). Localization of A-kinase through anchoring proteins. Mol. Endocrinol. 8, 5-11. Scott, J. D., Stofko, R. E., McDonald, J. R., Comer, J. D., Vitalis, E. A., and Mangili, J. A. (1990). Q p e I1 regulatory subunit dimerization determines the subcellar localization of the CAMP-dependent protein kinase. J. Biol. Chem. 265, 21561-21566. Seamon, K. B., and Daly, J. W. (1986). Forskolin: Its biological and chemical properties. Adu. Cyclic Nucleotide Res. 20, 1-150. Sefton, B. M., Hunter T., Ball, E. H., and Singer, S. J . (1981). Vinculin: A cytoskeletal target of the transforming protein of Rous sarcoma virus. Cell 24, 165-174.
GROWTH MODULATION BY CAMP
181
Selva, E., Raden, D. L., and Davis, R. J. (1993). Mitogen-activated protein kinase stimulation by a tyrosine kinase-negative epidermal growth factor receptor. J . Biol. Chem. 268,2250-2254. Selzer, E., Wilfing, A,, Schiferer, A,, Hermann, M., Grubeck-Loebenstein, B., and Friessmuth, M. (1993). Stimulation of human thyroid growth via the inhibitory guanine nucleotide binding (G) protein G,: Constitutive expression of the G-protein cx subunit G,a-1 in autonomous adenoma. Proc. Natl. Acad. Sci. U.S.A.90, 16091613. Seuwen, K., Magnaldo, I., and Pouyssegur, J. (1988). Serotonin stimulates DNA synthesis in fibroblasts acting through 5-HTlB receptors coupled to a G,-protein. Nature 335, 254-256. Seuwen, K., Kahan, C., Hartmann, T., Pouyssegur, J. (1990a). Strong and persistent activation of inositol lipid breakdown induces early mitogenic events but not Go to S phase progression in hamster fibroblasts. Comparison of thrombin and carbachol action in cells expressing M1 muscarinic acetylcholine receptors. J . Biol. Chem. 265, 22292-22299. Seuwen, K., Magnaldo, I., Kobilka, B. K., Caron, M. G., Regan, J. W., Lefkowitz, R. J., and Pouyssegur, J. (1990b). Alpha 2-adrenergic agonists stimulate DNA synthesis in Chinese hamster lung fibroblasts transfected with a human alpha 2-adrenergic receptor gene. Cell Regul. 1, 445-451. Sewing, A,, and Muller, R. (1994). Protein kinase A phosphorylates cyclin D1 a t three distinct sites within the cyclin box and a t the C-terminus. Oncogene 9, 2733-2736. Sewing, A. S. B., Burger, C., Schalk, C., Lucibello, F. C., and Miiller, R. (1993). Human cyclin D1 encodes a labile nuclear protein whose synthesis is directly induced by growth factors and suppressed by cyclic AMP. J . Cell Sci. 104, 545-555. Shamay, A,, Pines, M., Waksman, M., and Gertler, A. (1990). Proliferation of bovine undifferentiated mammary epithelial cells in vitro is modulated by G-proteins. Mol. Cell. Endocrinol. 69, 217-226. Sheffield, L. G. (1991). Oligonucleotides antisense to catalytic subunit of cyclic AMPdependent protein kinase inhibit mouse mammary epithelial cell DNA synthesis. Exp. Cell Res. 192, 307-310. Sheffield, L. G., and Welsch, C. W. (1985). Cholera-toxin-enhanced growth of human breast cancer cell lines in vitro and in vivo: Interaction with estrogen. Znt. J . Cancer 36,479-483. Sheffield, L. G., and Welsch, C. W. (1988). Transplantation of human breast epithelia to mammary-gland-free fat-pads of athymic nude mice: Influence of mammotrophic hormones on growth of breast epithelia. Znt. J. Cancer 41, 713-719. Sheffield, L. G., Sinha, Y. N., and Welsh, C. W. (1985). Cholera toxin treatment increases in vitro growth and development of the mouse mammary gland. Endocrinology 117, 1864-1869. Sheng, M., McFadden, G., and Greenberg, M. E. (1990). Membrane depolarization and calcium induce c-fos transcription via phosphorylation of transcription factor CREB. Neuron 4,571-582. Sheng, M., Thompson, M. A,, and Greenberg, M. E. (1991). CREB: A Ca2+-regulated transcription factor phosphorylated by calmodulin-dependent kinases. Science 252, 1427-1430. Shenker, A,, Laue, L., Kosugi, S., Merendino, J. J., Minegushi, T., and Cutler, G. B. (1993). A constitutively activating mutation of the luteinizing hormone receptor in familial male precocious puberty. Nature 365, 652-654.
182
PIERRE P. ROGER et al
Shenker, A., Weinstein, L. S., Sweet, D. E., and Spiegel, A. M. (1994).An activating Gsa mutation is present in fibrous dysplasia of bone in the McCune-Albright syndrome. J. Clin. Endocrinol. Metab. 79, 750-755. Sheridan, J. D., and Atkinson, M. M. (1985). Physiological roles of permeable junctions: Some possibilities. Annu. Rev. Physiol. 47, 337-353. Shiba, Y., Sasaki, Y., Hirono, C., and Kanno, Y. (1989). Close relationship between modulation of serum-induced stimulation of DNA synthesis and changes in gapjunctional intercellular communication in quiescent 3T3-Ll cells caused by cyclic AMP and the tumor-promoting phorbol ester TPA. Exp. Cell Res. 185, 535-540. Shikone, T., Yamoto, M., and Nakano, R. (1992). Follicle-stimulating hormone induces functional receptors for basic fibroblast growth factor in rat granulosa cells. Endocrinology 131, 1063-1068. Shinohara-Gotoh, Y., Nishida, E., Hoshi, M., and Sakai, H. (1991). Activation of microtubule-associated protein kinase by microtubule disruption in quiescent rat 3Y1 cells. Exp. Cell Res. 193, 161-166. Shirakawa, F., Yamashita, U., Chedid, M., and Mizel, S. B. (1988). Cyclic AMP-An intracellular second messenger for interleukin 1. Proc. Natl. Acad. Sci. U.S.A. 85, 8201-8205. Silberstein, G. B., Strickland, P., Trumpbour, V., Coleman, S., and Daniel, C. W. (1984). In vivo, CAMP stimulates growth and morphogenesis of mouse mammary ducts. Proc. Natl. Acad. Sci. U.S.A.81, 4950-4954. Simkin, N. J., Jelinek, D. F., and Lipsky, P. E. (1987). Inhibition of human B cell responsiveness by prostaglandin E,. J. Zmmunol. 138, 1074-1081. Simonian, M. H., White, M. L., and Gill, G. N. (1982). Growth and function of cultured bovine adrenocortical cells in a serum-free defined medium. Endocrinology 111, 919-927. Singh, T. J., Roth, C., Gottesman, M. M., and Pastan, I. M. (1981). Characterization of cyclic AMP-resistant Chinese hamster ovary cell mutants lacking type I protein kinase. J . Biol. Chem. 256, 926-932. Sinnett-Smith, J., Zachary, I., Valverde, A. M., and Rozengurt, E. (1993). Bombesin stimulation of p125 focal adhesion kinase tyrosine phosphorylation. J . Biol. Chem. 268, 14261-14268. Siperstein, A. E., Miller, R. A., and Clark, 0. H. (1988).Stimulatory effect of vasoactive intestinal polypeptide on human normal and neoplastic thyroid tissue. Surgery 104, 985-991. Skalhegg, B. S., Landmark, B. F., Doskeland, S. O., Hansson, V., Lea, T., and Jahnsen, T. (1992). Cyclic AMP-dependent protein kinase type I mediates the inhibitory effects of 3’,5’-cyclic adenosine monophosphate on cell replication in human T lymphocytes. J. Biol. Chem. 267, 15707-15714. Skalhegg, B. S., Tasken, K., Hansson V., Huitfeldt, H. S., Jahnsen, T., and Lea, T. (1994). Location of CAMP-dependent protein kinase type I with the TCR-CD3 Complex. Science 263, 84-87. Skouteris, G. G., and Kaser, M. R. (1991). Prostaglandins E, and F,, mediate the increase in c-myc expression induced by EGF in primary rat hepatocyte cultures. Biochem. Biophys. Res. Commun. 178, 1240-1246. Slominski, A,, Moellmann, G., and Kuklinska, E. (1989). MSH inhibits growth in a line of amelanotic hamster melanoma cells and induces increases in cyclic AMP levels and tyrosinase activity without inducing melanogenesis. J. Cell. Sci. 92, 551559. Slungaard, A., Confer, D. L., Jacob, H. S., and Eaton, J . W. (1983).Antineoplastic effects
GROWTH MODULATION BY CAMP
183
of Bordetella pertussis adenylate cyclase. Trans. Assoc. A m . Ph,ysicians 96, 401405. Slungaard, A., Confer, D. L., and Schubach, W. H. (1987). Rapid transcriptional downregulation of c-myc expression during cyclic adenosine monophosphate-promoted differentiation of leukemic cells. J. Clin. Invest. 79, 1542-1547. Smets, L. A., and Van b o y , H. (1987). Mitogenic and antimitogenic effects of cholera toxin-mediated cyclic AMP levels in 3T3 cells. J. Cell. Physiol. 133, 395-399. Smith, L. D. (1989). The induction of oocyte maturation: Transmembrane signaling events and regulation of the cell cycle. Development 107, 685-699. Smith, P., Wynford-Thomas, D., Stringer, B. M. J., and Williams, E. D. (1986). Growth factor control of rat thyroid follicular cell proliferation. Endocrinology 119, 14391445. Smith, P., Williams, E. D., and Wynford-Thomas, E. D. (1987). In vitro demonstration of a TSH-specific growth desensitising mechanism in rat thyroid epithelium. Mol. Cell. Endocrinol. 51, 51-58. Smyth, M. J., Runnels, B., and Wharton, W. (1992). Cholera toxin potentiates TPAinduced mitogenesis and c-fos expression in BALBic-3T3-derived proadipocytes. J . Cell. Biochem. 50, 210-218. Soares, M. J., De, M., Pinal, C. S., and Hunt, J. S. (1989). Cyclic adenosine 3’,5’-monophosphate analogues modulate rat placental cell growth and differentiation. Biol. Reprod. 40, 435-447. Sobue, G., Sonnenfeld, K., Rubenstein, A. E., and Pleasure, D. (1985). Tissue culture studies of neurofibromatosis: Effects of axolemmal fragments and cyclic adenosine 3’,5’-monophosphateanalogues on proliferation of Schwann-like and fibroblast-like neurofibroma cells. Ann. Neurol. 18, 68-73. Sobue, G., Shuman, S., and Pleasure, D. (1986). Schwann cell responses to cyclic AMP: Proliferation, change in shape, and appearance of surface galactocerebroside. Brain Res. 362, 23-32. Somjen, D., Binderman, I., Schluter, K. D., Wingender, E., Mayer, H., and Kaye, A. M. (1990). Stimulation by defined parathyroid hormone fragments of cell proliferation in skeletal-derived call cultures. Biochem. J . 272, 781-785. Sontag, J. M., Thierse, D., Rouot, B., Aunis, D., and Bader, M. F. (1991). A pertussistoxin-sensitive protein controls exocytosis in chromafin cells a t a step distal to the generation of second messengers. Biochem. J . 274, 339-347. Spada, A,, Arosio, M., Bochicchio, D., Bazzoni, N., Vallar, L., Bassetti, M., and Faglia, G. (1990). Clinical, biochemical, and morphological correlates in patients bearing growth hormone-secreting pituitary tumors with or without constitutively active adenylyl cyclase. J. Clin. Endocrinol. Metub. 71, 1421-1426. Spangelo, B. L., Isakson, P. C., and McLeod, R. M. (1990). Production of interleukin-6 by anterior pituitary cells is stimulated by increased intracellular adenosine 3’,5’monophosphate and vasoactive intestinal peptide. Endocrinology 127,403-409. Spencer, C. A., and Groudine, M. (1991). Control of c-myc regulation in normal and neoplastic cells. Adu. Cancer Res. 56, 1-8. Spiegel, S., and Fishman, P. H. (1987). Gangliosides as bimodal regulators of cell growth. Proc. Natl. Acad. Sci. U.S.A.84, 141-145. Spiegel, S., Fishman, P. H., and Weber, R. J. (1985). Direct evidence that endogenous GM, ganglioside can mediate thymocyte proliferation. Science 230, 1285-1287. Spruill, W. A., White, M. G., Steiner, A. C., Tres, L. L., and Kiersenbaum, A. L. (1981). Temporal sequence of cell shape changes in cultured rat sertoli cells after experimental elevation of intracellular CAMP.Exp. Cell. Res. 131, 131-148.
184
PIERRE P. ROGER et al.
Squinto, S. P., Block, A. L., and Doucet, J. P. (1989). Epidermal growth factor induction of cellular proliferation and protooncogene expression in growth-arrested rat H4IIE hepatoma cells: Role of cyclic adenosine monophosphate. Mol. Endocrinol. 3, 433446. Stampfer, M. R. (!982). Cholera toxin stimulation of human mammary epithelial cells in culture. In Vitro 18, 531-537. Steinfelder, H. J., and Petho-Schramm, S. (1990). Methylxanthines inhibit glucose transport in rat adipocytes by two independent mechanisms. Biochem. Pharmacol. 40, 1154-1 157. Stephens, E. V., Kalinec, G., Brann, M. R., and Gutkind, J. S. (1993). Transforming G protein-coupled receptors transduce potent mitogenic signals in NIH 3T3 cells independent on cAMP inhibition or conventional protein kinase C. Oncogene 8, 19-26. Strickland, S. (1981). Mouse teratocarcinoma cells: Prospects for the study of embryogenesis and neoplasia. Cell 24, 277-278. Struthers, R. S., Vale, W. W., Arias, C., Souvchenko, P. E., and Montminy, M. R. (1991). Somatotroph hypoplasia and dwarfism in transgenic mice expressing a nonphosphorylatable CREB mutant. Nature 350,622-624. Studzinski, G. P., Shankavaram, U. T., Moore, D. C., and Reddy, P. V. (1991). Association of c-myc protein with enzymes of DNA replication in high molecular weight fractions from mammalian cells. J . Cell Physiol. 147, 412-419. Suarez, H. G., Villard, J. A,, Caillou, B., Schlumberger, M., Parmentier, C., and Monier, R. (1991). gsp mutations in human thyroid tumours. Oncogene 6,677-679. Sumiyoshi, H., Yasui, W., Ochiai, A., and Tahara, E. (1984). Effects of gastrin on tumor growth and cyclic nucleotide metabolism in xenotransplantable human gastric and colonic carcinomas in nude mice. Cancer Res. 44,4276-4280. Tagliaferri, P., Clair, T., De Bortoli, M. E., and Cho-Chung, Y. S. (1985). Two classes of cAMP analogs synergistically inhibit p21 ras protein synthesis and phenotypic transformation of NIHi3T3 cells transfected with Ha-MuSV DNA. Biochem. Biophys. Res. Commun. 130,1193-1200. Tagliaferri, P., Katsaros, D., Clair, T., Ally, S., Tortora, G., Neckers, L., Rubalcava, B., Parandoosh, Z., Chang, Y., Revankar, G. R., Crabtree, G., Robins, R. K., and ChoChung, Y. S. (1988a). Synergistic inhibition of growth of breast and colon human cancer cell lines by site-selective cyclic AMP analogues. Cancer Res. 48, 1642-1650. Tagliaferri, P., Katsaros, D., Clair, T., Neckers, L., Robins, R. K., and Cho-Chung, Y. S. (1988b). Reverse transformation of Harvey murine sarcoma virus-transformed NIH/3T3 cells by site-selective cyclic AMP analogs. J . Biol. Chem. 263, 409-416. Takahashi, S. I., Conti, M., Prokop, C., Van Wyk, J. J., and Earp, H. S. (1991). Thyrotropin and insulin-like growth factor I regulation of tyrosine phosphorylation in FRTL-5 cells. J. Biol. Chem. 266, 7834-7841. Takanashi, A., Yasui, W., Yoshida, K., Yokozaki, H., Saito, D., Abe, K., Urakami, K., Miki, K., and Tahara, E. (1991). Inhibitory effect of 8-chloro-cyclic adenosine 3 ' 3 ' monophosphate on cell growth of gastric carcinoma cell lines. Jpn. J . CuncerRes. 82, 325-331. Takeda, J., Adashi, K., Halpsin, K. M., Itami, S., Levine, V., and Woodyard, C. (1983). Forskolin activates adenylate cyclase activity and inhibits mitosis in in vitro in pig epidermis. J. Invest. Dermatol. 81, 236-240. Takeshita, T., Ohbo, K., Nakamura, M., Goto, Y., and Sugamura, K. (1990). Establishment of a cyclic adenosine monophosphate-dependent growing human T-cell line derived from an interleukin-2-dependent cell line. J . Cell. Physiol. 145, 238-243. Taketani, Y., and Oka, T. (1983). Tumor promoter 12-0-tetradecanoylphorbol13-acetate
GROWTH MODULATION BY CAMP
185
like epidermal growth factor, stimulates cell proliferation and inhibits differentiation of mouse mammary epithelial cells. Proc. Natl. Acad. Sci. U.S.A. 80, 16461649. Tarpley, W. G., Hopkins, N. K., and Gorman, R. R. (1986). Reduced hormone-stimulated adenylate cyclase activity in NIH-3T3 cells expressing the EJ human bladder ras oncogene. Proc. Natl. Acad. Sci. U.S.A.83, 3703-3707. Taub, M., Chuman, L., Saier, M. H., and Sato, G. (1979). Growth of Madin-Darby canine kidney epithelial cell (MDCK) line in hormone-supplemented, serum-free medium. Proc. Natl. Acad. Sci. U.S.A. 76, 3338-3342. Taub, M., Saier, M. H., Chuman, L., and Hiller, S. (1983). Loss of the PGEl requirement for MDCK cell growth associated with a defect in cyclic AMP phosphodiesterase. J . Cell. Physiol. 114, 153-161. Taub, M., Devis, P. E., and Grohol, S. M., (1984). PGE,-independent MDCK cells have elevated intracellular cyclic AMP but retain the growth stimulatory effects of glucagon and epidermal growth factor in serum-free medium. J. Cell. Physiol. 120, 1928. Taylor, S. S., Buechler, J. A., Slice, L. W., Knighton, D. K., Durgerian, S., Ringheim, G. E., Neitzel, J. J., Yonemoto, Y. M., Sowadski, J. M., and Dospmann, W. (1988). CAMP-dependent protein kinase: A framework for a diverse family of enzymes. Cold Spring Harbor Symp. Quant. Biol. 53, 121-130. Taylor-Papadimitriou, J., Purkis, P., and Fentiman, I. S. (1980). Cholera toxin and analogues of cyclic AMP stimulate the growth of cultured human mammary epithelial cells. J . Cell. Physiol. 102, 317-321. Tetsumoto, T., Takada, K., Amino, N., and Miyai, K. (1988). Regulation of proliferation by the cholera toxin B subunit in FRTL-5 cells may involve a mechanism independent from the modulation of membrane receptor function. Biochem. Biophys. Res. Commun. 157,605-610. Thomas, G. (1992). MAP kinase by any other name smells just as sweet. Cell 68, 3-6. Thompson, C. B., Challoner, P. B., Neiman, P. E., and Groudine, M. (1985). Levels of c-myc oncogene mRNA are invariant throughout the cell cycle. Nature 314, 363366. Thorensen, G. H., Sand, T. E., Refsnes, M., Dajani, 0. F., Guren, T. K., Gladhaug, I. P., Killi, A,, and Christoffersen, T. (1990). Dual effects of glucagon and cyclic AMP on DNA synthesis in cultured rat hepatocytes: Stimulatory regulation in early G, and inhibition shortly before the S phase entry. J. Cell. Physiol. 144, 523-530. Toda, T., Cameron, S., Sass, P., Zoller, M., and Wigler, M. (1987). Three different genes in S. cerevisiae encode the catalytic subunits of the CAMP-dependent protein kinase. Cell 50, 277-287. Tominaga, T., dela Cruz, J., Burrow, G. N., and Meinkoth, J. L. (1994). Divergent patterns of immediate early gene expression in response to thyroid-stimulating hormone and insulin-like growth factor I in Wistar rat thyrocytes. Endocrinology 135, 1212-1219. Tong, P. S., and Marcelo, C. L. (1983). Augmentation of keratinocyte differentiation by the epidermal mitogen, 8-bromo-CAMP. Exp. Cell Res. 149, 215-226. Tortora, G., Clair, T., Katsaros, D., Ally, S., Calamonici, O., Neckers, L., Tagliaferri, P., Jahnsen, T., Robins, R. K., and Cho-Chung, Y. S. (1989). Induction of megakaryocytic differentiation and modulation of protein kinase gene expression by siteselective CAMP analogs in K-562 human leukemic cells. Proc. Natl. Acad. Sci. U.S.A.86, 2849-2852. Tortora, G., Clair, T., and Cho-Chung, Y. S. (1990). An antisense oligodeoxynucleotide
186
PIERRE P. ROGER et a1
targeted against the type IIb regulatory subunit mRNA of protein kinase inhibits CAMP-induced differentiation in HL-60 leukemia cells without affecting phorbol ester effects. Proc. Natl. Acad. Sci. U.S.A. 87, 705-708. Tortora, G., Yokazaki, H., Pepe, S., Clair, T., and Cho-Chung, Y. S. (1991). Differentiation of HL-60 leukemia by type I regulatory subunit antisense oligodeoxynucleotide of CAMP-dependent protein kinase. Proc. Natl. Acad. Sci. U S A . 88, 2011-2015. Tortora, G., Pepe, S., Cirafici, A. M., Ciardiello, F., Porcellini, A., Clair, T., Colletta, G., Cho-Chung, Y. S., and Bianco, A. R. (1993). Thyroid-stimulating hormoneregulated growth and cell cycle distribution of thyroid cells involve type I isozyme of cyclic AMP-dependent protein kinase. Cell. Growth Differ. 4, 359-365. Tortora, G., Pepe, S., Bianco, C., Baldassarre, G., Budillon, A,, Clair, T.,Cho-Chung, Y. S., Bianco, A. R., and Ciardiello, F. (1994). The RIa subunit of protein kinase A controls serum dependency and entry into cell cycle of human mammary epithelial cells. Oncogene 9, 3233-3240. Tournier, S., Raynaud, F., Gerbaud, P., Lohmann, S. M., Doree, M., and Evain-Brion, D. (1991). Association of type I1 CAMP-dependent protein kinase with p34cdc2 protein kinase in human fibroblasts. J . Biol. Chem. 266, 19018-19022. Tramontano, D., Avivi, A., Ambesi-Impiombato, F. S., Barak, L., Geiger, B., and Schlessinger, J. (1982). Thyrotropin induces changes in the morphology and the organization of microfilament structures in cultured thyroid cells. Exp. Cell Res. 137, 269272. Tramontano, D., Chin, W. W., Moses, A. C., and Ingbar, S. H. (1986). Thyrotropin and dibutyryl cyclic AMP increase levels of c-myc and c-fos mRNAs in cultured rat thyroid cells. J.Biol. Chem. 261, 3919-3922. Tramontano, D., Moses, A. C., Veneziani, B. M., and Ingbar, S. H. (1988). Adenosine 3',5'-monophosphate mediates both the mitogenic effect of thyrotropin and its ability to amplify the response to insulin-like growth factor I in FRTL5 cells. Endocrinology 122, 127-132. Tratner, I., Ofir, R., and Verma, I. M. (1992). Alteration of a cyclic AMP-dependent protein kinase phosphorylation site in the c-Fos protein augments its transforming potential. Mol. Cell. Biol. 12, 998-1006. Tsang, B. K., Rixon, R. H., and Whitfield, J. F. (1980). A possible role for cyclic AMP in the initiation of DNA synthesis by isoproterenol-activated parotid gland cells. J. Cell. Physiol. 102, 19-26. Tseng, B. Y., Prussak, C. E., and Almazan, M. T. (1989). Primase p49 mRNA expression is serum stimulated but does not vary with the cell cycle. Mol. Cell. Biol. 9, 19401945. Tsuda, T., Hamamori, Y., Yamashita, T., Fukumoto, Y., and Takai, Y. (1986). Involvement of three intracellular messenger systems, protein kinase C, calcium ion and cyclic AMP, in the regulation of c-fos gene expression in Swiss 3T3 cells. FEBS Lett. 208,39-42. Tsuda, T., Fukumoto, Y., Hamamori, Y., Yamashita, T., and Takai, Y. (1987). Involvement of two intracellular messenger systems, protein kinase C and cyclic AMP, in the regulation of c-fos gene expression in human promyelocytic leukemia (HL-60) cells. J . Biochem. 102, 1579-1583. Tsuzaki, S., Cone, R. D., Frazier, A. L., and Moses, A. C. (1991). The interaction of signal transduction pathways in FRTLS thyroid follicular cells: Studies with stable expression of P,-adrenergic receptors. Endocrinology 128, 1359-1369. Vairo, G., Argyriou, S., Bordun, A. M., Whitty, G., and Hamilton, J. A. (1990). Inhibition
GROWTH MODULATION BY CAMP
187
of the signaling pathways for macrophage proliferation by cyclic AMP. J . Biol. Chem. 265,2692-2701. Valente, W. A., Vitti, P., Kohn, L. D., Brandi, M. L., Rotella, R., Toccafondi, R., Tramontano, D., Aloj, S. M., and Ambesi-Impiombato, F. S. (1983a). The relationship of growth and adenylate cyclase activity in cultured thyroid cells: Separate bioeffects of thyrotropin. Endocrinology 112, 71-79. Valente, W. A., Vitti, P., Rotella, C. M., Vaughan, M., Aloj, S. M., Grollman, E. F., Ambesi-Impiombato, F. S., and Kohn, L. D. (1983b). Antibodies that promote thyroid growth. A distinct population of thyroid-stimulating autoantibodies. N . Engl. J . Med. 309, 1028-1034. Vallar, L., Spada, A,, and Giannattasio, G. (1987). Altered G, and adenylate cyclase activity in human GH-secreting pituitary adenomas. Nature 330, 566-568. Van Corven, E. J., Groenink, A., Jalink, K., Eichholtz, T., and Moolenaar, W. H. (1989). Lysophosphatidate-induced cell proliferation: Identification and dissection of signaling pathways mediated by G proteins. Cell 59, 45-54. Van Corven, E. J., Hordijk, P. L., Medema, R. H., Bos, J. L., and Moolenaar, W. H. (1993). Pertussis toxin-sensitive activation of p21ras by G protein-coupled receptor agonists in fibroblasts. Proc. Natl. Acad. Sci. U.S.A. 90, 1257-1261. van der Plas, A,, Feyen, J . H. M., and Nijweide, P. J . (1985). Direct effect of parathyroid hormone on the proliferation of osteoblast-like cells; a possible involvement of cyclic AMP. Biochem. Biophys. Res. Commun. 129, 918-925. Van Lookeren Campagne, M. M., Villalba Diaz, F., Jastorff, B., and Kessin, R. H. (1991). 8-Chloroadenosine 3’,5’-monophosphate inhibits the growth of Chinese hamster ovary and m o b 4 cells through its adenosine metabolite. Cancer Res. 51,1600-1605. Van Meeteren, A., Loesberg, C., Van Rijn, J., and Van Wijk, R. (1982).The effects of 3 ‘ 3 ’ adenosine monophosphate on the proliferation of Reuben H35 rat hepatoma cells in vitro. I n Vitro 18, 891-899. van Roon, M. A., Eier, R., Charles, R., and Lamers, W. H. (1989). The initial accumulation of carbamoylphosphate synthetase in embryonic rat hepatocytes, and the cell cycle. Differentiation 41, 139-147. Van Sande, J., Mockel, J., Boeynaems, J. M., Dor, P., Andry, G., and Dumont, J . E. (1980). Regulation of cyclic nucleotide and prostaglandin formation in normal thyroid tissue and in autonomous nodules. J . Clin. Endocrinol. Metab. 50, 776-785. Van Sande, J., Lamy, F., Lecocq, R., Mirkine, N., Rocmans, P., Cochaux, P., Mockel, J., and Dumont, J . E. (1988). Pathogenesis of autonomous thyroid nodules: In vitro study of iodine and adenosine 3’,5’-monophosphatemetabolism. J. Clin. Endocrinol. Metab. 66, 570-579. Van Sande, J., Lefort, A., Beebe, S., Roger, P. P., Perret, J., Corbin, J., and Dumont, J. E. (1989). Pairs of cyclic AMP analogs that are specifically synergistic for type I and type I1 CAMP-dependent protein kinases, mimic thyrotropin effects on the function, differentiation expression and mitogenesis in dog thyroid cells. Eur. J . Biochem. 183, 699-708. Vasquez, A., Auffredou, M. T., Galanand, P., and Leca, G. (1991). Modulation of IL-2 and IL-4-dependent human B cell proliferation by cyclic AMP. J . Immunol. 146,42224227. Veneziani, B. M., Villone, G., Romano, R., Di Carlo, A., Garbi, C., and Tramontano, D. (1990). The tissue-specific pathways regulating cell proliferation are inherited independently in somatic hybrid between thyroid and liver cells. J.Cell B i d . 111,27032711.
188
PIERRE P. ROGER et al.
Vereb, G., Erdodi, F., Toth, B., and Bot, G. (1986). Regulatory subunit of type I1 CAMPdependent protein kinase as substrate and inhibitor of protein phosphatase-1 and -2A. FEBS Lett. 197, 139-142. Viard, I., Hall, S. H., Jaillard, C., Berthelon, M. C., and Saez, J. M. (1992).Regulation of c-fos, c-jun and jun-B messenger ribonucleic acids by angiotensin-I1 and corticotropin in ovine and bovine adrenocortical cells. Endocrinology 130, 1193-1200. Vintermyr, 0. K., Mellgren, G., Boe, R., and Doskeland, S. 0. (1989). Cyclic adenosine monophosphate acts synergistically with dexamethasone to inhibit the entrance of cultured adult rat hepatocytes into S-phase: With a note on the use of nucleolar and extranucleolar PHI-thymidine labelling patterns to determine rapid changes in the rate of onset of DNA replication. J . Cell. Physiol. 141, 371-382. Vintermyr, 0. K., Gjertsen, B. T., Lanotte, M., and Doskeland, S. 0. (1993b). Microinjected catalytic subunit of CAMP-dependent protein kinase induces apoptosis in myeloid leukemia (IPC-81) cells. Exp. Cell Res. 206, 157-161. Vintermyr, O., Boe, R., Bruland, T., Houge, G., and Doskeland, S. 0. (1993a). Elevated CAMP gives short-term inhibition and long-term stimulation of hepatocyte DNA replication: Roles of the CAMP-dependent protein kinase subunits. J . Cell. Physiol. 156, 160-170. Volberg, T., Zick, Y., Dror, R., Sabanay, L., Gidon, C., Levitzki, A., and Geiger, B. (1992). The effect of tyrosine-specific protein phosphorylation on the assembly of adherenstype junctions. EMBO J . 11, 1733-1742. Wagoner, P. K., and Pallotta, B. S. (1988). Modulation of acetylcholine receptor desensitization by forskolin is independent of CAMP.Science 240, 1655-1657. Walsh, D. A,, and Glass D. B. (1991). Utilization of the inhibitor protein of adenosine cyclic monophosphate-dependent protein kinase, and peptides derived from it, as tools to study adenosine cyclic monophosphate-mediated cellular processes. Methods Enzymol. 201,304-316. Walker, D. H., and Maller, J. L. (1991). Role of cyclin A in the dependence of mitosis on completion of DNA replication. Nature 354, 314-317. Walsh, D. A,, Angelos, K. L., Van Patten, S. M., Glass, D. B., and Garetto, L. P. (1990).I n “Peptides and Protein Phosphorylation” (B. E. Kemp, ed.), pp. 43-84. CRC Press, Boca Raton, Florida. Wang, Z. W., and Rozengurt, E. (1983). Interplay of cyclic AMP and microtubules in modulating the initiation of DNA synthesis in 3T3 cells. J . Cell Biol. 96,1743-1750. Wang, Y., Scott, J . D., McKnight, G. S., and Krebs, E. G. (1991). A constitutively active holoenzyme form of the CAMP-dependent protein kinase. Proc. Natl. Acad. Sci. U.S.A. 88, 2446-2450. Ward, G. E., and Kirschner, M. W. (1990). Identification of cell-cycle-regulated phosphorylation sites on nuclear lamin C. Cell 61, 561-577. Weber, W., and Hilz, H. (1986). CAMP-dependent protein kinases I and 11: Divergent turnover of subunits. Biochemistry 25, 5661-5667. Weidman, E. R., and Gill, G. N. (1977). Differential effects of ACTH o r 8-Br-CAMP on growth and replication in a functional adrenal tumor cell line. J . Cell Physiol. 90, 91-103. Weinmaster, G., and Lemke, G. (1990). Cell-specific cyclic AMP-mediated induction of the PDGF receptor. EMBO J . 9, 915-920. Weinstein, L. S., Shenker, A,, Gejman, P. V., Merino, M. J., Friedman, E., and Spiegel, A. M. (1991). Activating mutations of the stimulatory G protein in the McCuneAlbright syndrome. N . Engl. J. Med. 325, 1688-1695. Weisman, G. A,, Lustig, K. D., Lane, E., Huang, N., Belzer, I., and Friedberg, I. (1988).
GROWTH MODULATION BY cAMP
189
Growth inhibition of transformed mouse fibroblasts by adenine nucleotides occurs via generation of extracellular adenosine. J . Biol. Chem. 263, 12367-12372. Welsch, C. W., and De Hoog, J. V. (1983). Influence of cholera toxin on the growth and rat mammary carcinomas. Cancer development of N-methyl-N-nitrosourea-induced Lett. 20, 61-68. Welsh, T. H., Jones, P. B. C., and Hsueh, A. J. W. (1984). Phorbol ester inhibition of ovarian and testicular steroidogenesis in vitro. Cancer Res. 44,885-892. Werth, D. K., and Pastan, I. J. (1984). Vinculin phosphorylation in response to calcium and phorbol esters in intact cells. J . Biol. Chem. 259, 5264-5270. Westermark, B., and Heldin, C. H. (1985). Similar action of platelet-derived growth factor and epidermal growth factor in the prereplicative phase of human fibroblasts suggests a common intracellular pathway. J . Cell Physiol. 124, 43-48. Westermark, B., and Porter, K. R. (1982). Hormonally induced changes in the cytoskeleton of human thyroid cells in culture. J. Cell Biol. 94, 42-50. Westermark, K., Westermark, B., Karlsson, F. A., and Ericson, L. E. (1986). Location of epidermal growth factor receptors on porcine thyroid follicle cells and receptor regulation by thyrotropin. Endocrinology 118, 1040-1046. Wharton, W., Leof, E. B., Olashaw, N. E., Earp, H. S., and Pledger, W. J. (1982). Increases in cyclic AMP potentiate competence formation in BALBlc-3T3 cells. J . Cell. Physi01. 111, 201-206. Whitfield, J. F., Durkin, J. P., Franks, D. J., Kleine, L. P., Raptis, L., Rixon, R. H., Sikorska, M., and Walker, P. R. (1987). Calcium, cyclic AMP and protein kinase CPartners in mitogenesis. Cancer Metastasis Rev. 5, 205-250. Whitman, M., and Cantley, L. (1988). Phosphoinositide metabolism and the control of cell proliferation. Biochim.Biophys. Acta 948, 327-344. Willey, J. C., Laveck, M. A., McLendon, I. A., and Lechner, J. F. (1985). Relationship of ornithine decarboxylase activity and cAMP metabolism to proliferation of normal human bronchial epithelial cells. J. Cell. Physiol. 124, 207-212. Williams, E. D. (1990). TSH and thyroid cancer. Horn. Metab. Res. 23, 72-75. Williams, D. W., Wynford-Thomas, D., and Williams, E. D. (1988). Loss of dependence on IGF-1 for proliferation of human thyroid adenoma cells. Br. J . Cancer 57, 535539. Wilson, P. D., and Horster, M. F. (1983). Differential response to hormones of defined distal nephron epithelia in culture. Am. J. Physiol. 244, C166-Cl74. Wollberg, P., Soderqvist, H., and Nelson, B. D. (1994). Mitogen activation of human peripheral T lymphocytes induces the formation of new cyclic AMP response element-binding protein nuclear complexes. J. B i d . Chem. 269, 19719-19724. Woloshin, P. I., Walton, K. M., Rehfuss, R. P., Goodman, R. H., and Cone, R. D. (1992). 3’,5’-Cyclic adenosine monophosphate-regulated enhancer binding (CREB) activity is required for normal growth and differentiated phenotype in the FRTL5 thyroid follicular cell line. Mol. Endocrinol. 6, 1725-1733. Wong, G., and Pawelek, J. (1973). Control of phenotypic expression of cultured melanoma cells by melanocyte stimulating hormones. Nature 241, 213-215. Wong, Y. H., Federman, A,, Pace, A. M., Zachary, I., Evans, T., Pouyssegur, J., and Bourne, H. R. (1991). Mutant a subunits of Gi2 inhibit cyclic AMP accumulation. Nature 351, 63-65. Wu, Y. Y., and Lin, M. C. (1990). Induction of differentiation in v-Ha-ras-transformed MDCK cells by prostaglandin E, and 8-bromo-cyclic AMP is associated with a decrease in steady-state level of inositol 1,4,5-trisphosphate. Mol. Cell. Biol. 10,5767.
190
PIERRE P. ROGER et a1
Wu, B., Fodor, E. J . B., Edwards, R. H., and Rutter, W. J. (1989). Nerve growth factor induces the proto-oncogene c-jun in PC12 cells. J . Biol. Chem. 264, 9000-9003. Wu, J., Dent, P., Jelinek, T., Wolfman, A., Weber, M. J., and Sturgill, T. W. (1993). Inhibition of the EGF-activated MAP kinase signaling pathway by adenosine 3'3'monophosphate. Science 262, 1065-1068. Wynford-Thomas, D., Stringer, B. M. J., Harash, H. R., and Williams, E. D. (1983). Control of growth in the rat thyroid-An example of specific desensitization to trophic hormone stimulation. Enperientia 39, 421-423. Wynford-Thomas, D., Smith, P., and Williams, E. D. (1987). Proliferative response to cyclic AMP elevation of thyroid epithelium in suspension culture. Mol. Cell. Endocrinol. 51, 163-166. Yamada, T. (1989). Cell type expression mediated by cell cycle events, and signaled by mitogens and growth inhibitors. Znt. Reu. Cytol. 117, 215-255. Yamamoto, K. K., Gonzalez, G. A., Biggs, W. H., and Montminy, M. R. (1988). Phosphorylation-induced binding and transcriptional efficacy of nuclear factor CREB. Nature 334,494-498. Yamashita, M. (1988). Involvement of CAMPin initiating maturation of the brittle-star Amphipholis kochii oocytes: Induction of oocyte maturation by inhibitors of cyclic nucleotide phosphodiesterase and activators of adenylate cyclase. Deu. Biol. 125, 109-114. Yamashita, T., Tsuda, T., Hamamori, Y., and Takai, Y. (1986). Possible involvement of cyclic AMP and calcium ion in prostaglandin El-induced elevation of c-myc mRNA levels in Swiss 3T3 firbroblasts. J . Biol. Chem. 261, 16878-16882. Yang, J., Guzman, R., Richards, J., Imagawa, W., McCormick, K., and Nandi, S. (1980). Growth factor- and cyclic nucleotide-induced proliferation of normal and malignant mammary epithelial cells in primary culture. Endocrinology 107, 35-41. Yasui, W., Sumiyoshi, H., Ochiai, A., and Tahara, E. (1986). Cholecytokinin inhibition of tumor growth and gastrin-stimulated cyclic adenosine 3':5'-monophosphate metabolism in human gastric carcinoma in nude mice. Cancer Res. 46, 740-743. Yeh, C. K., Louis, J. M., and Kousvelari, E. E. (1988).Beta-adrenergic regulation of c-fos gene expression in an epithelial cell line. FEBS Lett. 240, 118-122. Yokozaki, H., Tortora, G., Pepe, S., Maronde, E., Genieser, M. G., Jastorff, B., and ChoChung, Y. S. (1992). Unhydrolyzable analogues of adenosine 3':5'-monophosphate demonstrating growth inhibition and differentiation in human cancer cells. Cancer Res. 52,2504-2508. Yokozaki, H., Budillon, A., Tortora, G., Meissner, S., Beaucage, S. L., Miki, K., and ChoChung, Y. S. (1993). An antisense oligodeoxynucleotide that depletes RIa subunit of cyclic AMP-dependent protein kinase induces growth inhibition in human cancer cells. Cancer Res. 53, 868-872. Yoshimasa, T., Sibley, D. R., Bouvier, M., Lefkowitz, R. J., and Caron, M. G. (1987). Cross-talk between cellular signalling pathways suggested by phorbol-esterinduced adenylate cyclase phosphorylation. Nature 327, 67-70. Yoshimura, M., and Cooper, D. M. F. (1993). Qpe-specific stimulation of adenylylcyclase by protein kinase C. J . Riol. Chem. 268, 4604-4607. Yoshimura, Y., and Tamura, T. (1991). Cyclic adenosine monophosphate promotes the proliferation of chicken granulosa cells in culture. Gen. Comp. Endocrinol. 84, 222227. Yu, D., Seitz, P. K., Selvanayagam, P., Rajaraman, S., Townsend, C. M., and Cooper, C. W. (1992). Effects of vasoactive intestinal peptide on adenosine 3',5'-monophos-
GROWTH MODULATION BY cAMP
191
phate, ornithine decarboxylase, and cell growth in a human colon cell line. Endocrinology 131, 1188-1194. Yun, K., Yamashita, S., Izumi, K., Yonemitsu, V., and Sugihara, H. (1986). Effects of forskolin on the morphology and function of the rat thyroid cell strain, FRTL-5: Comparison with the effects of thyrotropin. J.Endocrinol. 111, 397-405. ~ ) :of converZachary, I., and Rozengurt, E. (1992). Focal adhesion kinase ( ~ 1 2 . 5 ~A~point gence in the action of neuropeptides, integrins, and oncogenes. Cell 71, 891-894. Zachary, I., Masters, S. B., and Bourne, H. R. (1990).Increased mitogenic responsiveness of Swiss 3T3 cells expressing constitutively active GSa.Biochem. Biophys. Res. Cornmun. 168, 1184-1193. Zachary, I., Gil, J., Lehmann, W., Sinnett-Smith, J., and Rozengurt, E. (1991). Bombesin, vasopressin, and endothelin rapidly stimulate tyrosine phosphorylation in intact Swiss 3T3 cells. Proc. Natl. Acad. Sci. U.S.A. 88, 4577-4581. Zhang, Y., Lin, S. X., and Vilcek, J. (1988). Synthesis of interleukin 6 (interferon-beta 2iB cell stimulatory factor 2) in human fibroblasts is triggered by an increase in intracellular cyclic AMP. J.Biol. Chem. 263,6177-6182. Zorn, M., Maronde, E., Jastorff, B., and Richter-Landsberg, C. (1993). Differential effects of two structurally related NG-substituted cAMP analogues on C6 glioma cells. Eur. J. Cell. Biol. 60, 351-357. Zullo, J. N., Cochran, B. H., Huang, A. S., and Stiles, C. D. (1985). Platelet-derived growth factor and double-stranded ribonucleic acids stimulate expression of the same genes in 3T3 cells. Cell 43, 793-800. Zurier, R. B., Kozma, M., Sinnett-Smith, J., and Rozengurt, E. (1988). Vasoactive intestinal peptide synergistically stimulates DNA synthesis in mouse 3T3 cells: Role of CAMP,Ca2+,and protein kinase C. Enp. Cell. Res. 176, 155-161. Zurzolo, C., Gentile, R., Mascia, A., Garbi, C., Polistina, C., Aloj, L., Avvedimento, V. E., and Nitsch, L. (1991). The polarized epithelial phenotype is dominant in hybrids between polarized and unpolarized rat thyroid cell lines. J. Cell Sci. 98, 65-73.
This Page Intentionally Left Blank
VITAMINS AND HORMONES, VOL. 51
Regulation of G Protein-Coupled Receptors by Receptor Kinases and Arrestins RACHEL STERNE-MARR AND JEFFREY L. BENOVIC Department of Pharmacology Jefferson Cancer Cancer Thomas Jefferson University Philadelphia, Pennsylvania 19107
I. Introduction 11. GRKs A. Cloning B. Tissue Localization C. Membrane Localization D. Regulation E. Receptor Specificity F. Distinguishing Features of Members of the GRK Family 111. Arrestins A. Cloning B. Polypeptide Variants C. Tissue, Cellular, and Subcellular Localization D. Mechanism of Retinal Arrestin Binding to Rhodopsin E. Mechanism of Nonvisual Arrestin Binding to Receptors F. Receptor Specificity IV. Conclusions References
I. INTRODUCTION Signal transduction via seven-transmembrane domain or serpentine receptors, which is mediated by guanine nucleotide-binding (G) proteins, accounts for a significant fraction of all signaling in the body. Heterotrimeric G proteins modulate the activities of multiple effectors, including cGMP phosphodiesterase, adenylyl cyclase, phospholipases C and A,, and potassium and calcium ion channels, which alters the levels of second messenger molecules and ultimately leads to a variety of cell-specific events (Birnbaumer et al., 1990). One ubiquitous feature of signaling through G protein-coupled receptors (GPRs) and other cell surface receptors is the rapid loss of cellular sensitivity following presentation of a stimulus, a phenomenon alternatively referred to as desensitization, adaptation, deactivation, tolerance, tachyphylaxis, or quenching (Hausdorff et al., 1990; Dohlman et al., 193
Copyright 0 1995 by Academic Press, Inc. All rights of reproduction in any form reserved.
194
RACHEL STERNE-MARR AND JEFFREY L. BENOVIC
1991).Two receptor systems have served as useful models to dissect the mechanism of desensitization: visual transduction by the light-sensitive receptor rhodopsin in retinal rod cells and chemical transduction via the catecholamine-sensitive P2-adrenergic receptor (P,AR). Absorption of a photon of light by a molecule of rhodopsin residing in the disk membrane of rod outer segments results in the isomerization of the chromophore 1l-cis-retinal to the all-trans conformation, yielding the active form of rhodopsin, metarhodopsin I1 (MII) (Chabre and Deterre, 1989; Hargrave and McDowell, 1992). Activated rhodopsin catalyzes the exchange of GTP for GDP on the retinal-specific G protein, transducin (GT),and GTP-bound GT activates cGMP phosphodiesterase. The decrease in cGMP concentration closes cGMP-gated cation channels in the plasma membrane of rod cells, generating an electrical signal which is propagated along the visual pathway. In order to perceive continuous changes of light in the environment, the rod cell must recover from light activation; therefore, a rapid (highmillisecond-second time scale) “turnoff ” mechanism has evolved. Upon light activation MI1 becomes a substrate for a retinal-specific kinase, rhodopsin kinase (RK), which phosphorylates rhodopsin in a light-dependent fashion at multiple serine and threonine residues on the C-terminal tail of the receptor (Bownds et al., 1972; Kuhn and Dreyer, 1972; Frank et al., 1973; Wilden and Kuhn, 1982).Phosphorylation of MI1 lowers the affinity of the receptor for transducin while dramatically increasing its affinity for a 48-kDa protein called arrestin (also called S antigen for its immunogenic properties). The binding of arrestin to the light-activated and phosphorylated rhodopsin physically occludes G protein interaction with the receptor, thereby uncoupling the receptor from the G protein and “arresting” transmission of the visual signal (Wilden et al., 1986). An analogous system of desensitization by the tandem action of a kinase and an arrestin is operative in the hormonal regulation of adenylyl cyclase. Binding of catecholamine to the P2AR causes the “stimulatory” G protein, Gs, to activate adenylyl cyclase (Hausdorff et al., 1990; Dohlman et al., 1991). The newly synthesized CAMP activates CAMP-dependent protein kinase to phosphorylate various target proteins, which, in turn, leads to numerous physiological responses, such as increased heart rate and smooth muscle relaxation. This system is also rapidly desensitized within a period of seconds to minutes. A kinase that phosphorylates residues in the C-terminal tail of the P2AR in an agonist-dependent fashion was initially identified in S49 lymphoma cells and named P-adrenergic receptor kinase (PARK) (Benovic et al., 1986). A molecular search for an arrestin homologue resulted in
G PROTEIN-COUPLED RECEPTOR REGULATION
195
the identification of a bovine brain protein, called p-arrestin (parr), which could effectively uncouple PARK-phosphorylated p2AR from G, in uitro (Lohse et al., 1990). A t least three mechanisms for uncoupling G proteins from receptors are recognized: (1)homologous or agonist-specific desensitization requiring the kinase-arrestin duo as described above, (2) agonist-dependent sequestration of the receptor away from the plasma membrane into endocytic vesicles, and (3) heterologous or nonagonist-specific desensitization occurring in an arrestin-independent manner and requiring receptor phosphorylation by second messenger-dependent kinases such as the CAMP-dependent protein kinase and protein kinase C. The extent to which each of these three mechanisms is responsible for desensitization is cell type and receptor specific. For example, in rod cells rhodopsin is present not in the plasma membrane but in the disk membrane, and therefore endocytosis is unlikely to play a role in rhodopsin deactivation. Conversely, 95% of the thrombin receptors in HEL (and CHRF) cells are internalized within minutes of thrombin treatment, perhaps obviating the necessity for other desensitizing mechanisms (Brass, 1992; Hoxie et al., 1993). Since the model of homologous desensitization mediated by kinases and arrestins was elucidated in uitro with purified components (Benovic et al., 1987a; Lohse et al., 1992; Pitcher et al., 1992a1,it is important to ask whether this pathway occurs in uiuo. Clearly, in DrosophiZa a mutational analysis has verified that arrestins are required for the regulation of rhodopsin function (Dolph et al., 1992).Unfortunately, no vertebrate cell lines or animals that lack functional kinases or arrestins have been generated to enable an assessment of their role in uiuo. Nevertheless, several lines of experimentation support an in uiuo role for receptor kinases and arrestins. First, transfection of CHW cells with forms of the P2AR that lack serine and threonine residues normally present in the tail of the receptor abrogates both agonistdependent phosphorylation and rapid agonist-induced desensitization of the receptor (Bouvier et al., 1988). Second, expression of PARK or parr increases the agonist-stimulated desensitization of coexpressed p2AR in CHO cells (Pippig et al., 1993).Third, when a kinase-deficient form of PARK that retains its regulatory domains, a so-called dominantnegative mutant, is overexpressed in bronchial epithelial BEAS-2B cells, homologous desensitization of the P2AR is attenuated (Kong et al., 1994). In the last few years it has become clear that RK and PARK are members of a multigene family in which six mammalian members have been identified to date. Members of the mammalian family are
196
RACHEL STERNE-MARR AND JEFFREY L. BENOVIC
now referred to as G protein-coupled receptor kinases (GRKs), where RK is GRKl and PARK is GRK2. Likewise, visual arrestin and Parr are members of a family that now includes four mammalian homologs, three of the four being expressed as more than one polypeptide form. These observations raise questions of the receptor specificity of these proteins: Does receptor specificity occur in uiuo? Is it defined by cell type? Is it defined by unique interactions of GRKs and arrestins with receptors? Since G protein-coupled receptors are the targets of many drugs in clinical use today, defining the receptor specificity of GRKs and arrestins could have an enormous impact on pharmaceutical efficacy. In this chapter we focus on the cloning, tissue (and cell) localization, characterization, structure-function, and specificity studies of mammalian GRKs and arrestins. Since several similar reviews have recently been published (Hausdorff et al., 1990; Dohlman et al., 1991; Palczewski and Benovic, 1991; Inglese et al., 1993; Lefkowitz, 1993; Wilson and Applebury, 1993; T. Haga et al., 19941, our goal here is to provide complementary and updated information.
11. GRKs A. CLONING The first member of the GRK family to be cloned was bovine PARK. Oligonucleotides designed from the sequences of two bovine PARKderived peptides were used to probe a bovine brain cDNA library (Benovic et al., 1989). Cloning of the other five members took advantage of techniques that are reflective of the status of molecular biology in the early 1990s: low-stringency hybridization, polymerase chain reaction (PCR), rapid amplification of cDNA ends (RACE),and positional cloning in combination with exon amplification. PARK2 (GRK3) was cloned by hybridizing a bovine brain cDNA library with a fragment from the catalytic domain of PARK as a probe under low-stringency conditions (Benovic et al., 1991a). Like the cloning of PARK, initial isolation of the cDNA for RK utilized oligonucleotides based on the sequence of the purified bovine retinal protein (Lorenz et al., 1991). After probing two bovine retinal cDNA libraries, the 5’ RACE procedure (using retinal RNA) was required to identify cDNA sequences encoding the N terminus of RK. Serendipity played a role in identification of the fourth mammalian member of the family. In a search for genes that may be responsible for the dominant defect in Huntington’s
G PROTEIN-COUPLED RECEPTOR REGULATION
197
disease, IT11 (GRK4) was cloned using cDNA derived from the exon amplification procedure to probe a human frontal cortex cDNA library (Ambrose et al., 1992). Unlike other members of the GRK family, GRK4 may exist as polypeptide variants that differ by the absence or presence of 32 amino acids near the N terminus (Sallese et al., 1994). Variation near the C terminus of GRK4 may exist as well (Inglese et al., 1993). In 1988 Hanks (1988) compared 65 different members of the protein kinase superfamily and defined 11 catalytic “subdomains” that contain identical or conserved amino acids in these positions in each of the kinases. In a n attempt to isolate Drosophila GRKs, oligonucleotides designed from residues in subdomains VI and VIII and biased toward the PARK and PARK2 sequences were used in PCR amplification of fly retinal cDNA. PCR products that showed highest homology to bovine PARK were then used to probe Drosophila cDNA and genomic libraries to determine the structure of the full-length proteins. Two PARK homologs, designated GPRK-1 and GPRK-2, were identified in this manner (Cassill et al., 1991). In some subdomains the sequences of PARK, PARK2, RK, GPRK-1, and GPRK-2 are conserved among the five proteins but differ significantly from other kinases in the database. Taking advantage of this observation, degenerate oligonucleotides designed from subdomains I1 and VIII of the GRKs were used in PCR amplification of human heart cDNA, while subdomains I and VIII were used as primers to amplify sequences from bovine circumvillate papillae. Both primer pairs led to the identification of GRK5 (Kunapuli and Benovic, 1993; Premont et al., 1994). Finally, the catalytic domains of PARK and PARK2 were used as probes to screen a human heart cDNA library at low stringency. A partial clone for GRKG was obtained and subsequently used to isolate a full-length clone from a human fetal brain cDNA library (Benovic and Gomez, 1993). GRK5 and GRKG were also isolated from human neutrophil RNA using the sequences from subdomain VI common to GRKs and several other protein kinases for one forward primer, sequences unique to GRK subdomain I for a second forward primer, and sequences derived from GRK subdomain VIII for the reverse primer in PCR amplification (Haribabu and Snyderman, 1993). While the GRKs are most closely related to the protein kinase C and CAMP-dependent kinase families, three features distinguish the cloned GRKs from other mammalian kinases (Figs. 1 and 2). GRKs are serinehhreonine kinases with a central 263- to 266-amino-acid catalytic domain flanked by large amino (186- to 190-amino-acid) and carboxy (82- to 236-amino-acid) regulatory domains. [The catalytic domain is
198 HUMAN IIARK HUMAN O W 2 HUMAN GRKS HUMAN GRX6 HUMAN GRK4 BOVINE RK
RACHEL STERNE-MARR AND JEFFREY L. BENOVIC ..MADLEAVL ..MRDLEAVL MELENIV MELENIV MELENIV MDFGSLEW
... ...
...
00
--
ADVSYLHRME ADVSYLHRME ANNLLKIRE ANNLLKIRE ANSLLLKARQ ANSAFIAARG
-
0
KSFATPRIRh KSKATPW ..GGGGKRKG ..GGGGNRKG ..GGYCKKSG SFDASSGPAS
SKK.....IL SKR.....IV KSKKWKEILK KSKKWRQMLQ RSKKWKEILT RDRKYLARLK
-
0
-
LPEPSIRSVM LPEPSIRSVM FPHISQCEDL FPHISQCEEL LPPVSQCSEL LPPLSKCEAL 0
-
0
QKYLEDRGEV QKYLAERNEI RRTIDRDYCS RLSLERDYHS RHSIEKDYSS RESLDLGFEG
-
GYLLFRDFCL GFLLFKDFCL GRLLFRQFCE GRLLFREPCA GRRLFRQFCD GKRLFQQFLR
CQN1,RGDVFQ CESLRGDZFC HEYLRGFPFN HEYLSVAPFA HNYLRGEPFE LAHLGQAPFQ 0 0
UF:ESDKFTR kCOHKWELN 141T!54DtSY 1 9 3 KFMFSDKFTR FCOYKSVELN I'11TT4LfSV 1 9 3
...
- 0
NHLEERReLV NEINEAVPQV TR.PGLECYI TR.PELSRCV TK.PTLKRH1 TH.EQHGPAL
-
0
00
EFYEEIKKYE KFYEEIKEYE QFLDSVAEYE AFLDGVAEYE EFLDAVAEYE QLWKDIEDYD
93
TFEKIFSQKL TFDKIFNQKI LCDK...QPI LCER HA1 LCDK...QPI MCLE ...QPI
-
93
91 91 91 96
0-
""* ... HUMAN OARK HUMAX OARK2
HUMAN GRKS HUMAN GRK6 HUMAN G R K 4 BOVINE RK
KLETEEERVA RSRElFDSYl MKELUCSHP KLDIEEDRLC RSRQlYDAYl MKELLSCSHP .YTPnFY.GF Y G Y F . I M T K I I TPKSPWIAO ...~..~ .VTPDDKRKA CGRHVTQNFL SHTGPDLIPE .VMDEDRSD CGLSILDRFF N D K W L P E .TMDALPPQ XAQALRRAYL EPQAQLFCSF
- -
ISLSAIEHVC PSKQAVCHVQ VGO2LVSOTE VPRQLVTNCT IPPDWTECR LDAETVW
SHLGKKQVPP SHLSKKOVTS EKL.LOKPCK QRL.EQGPCK LGLKEENPSK AGAGD
~-~~ ~-~~ ~
~
~-
~~
.....
DLF3PYlEEl lLFQPYlEEl ELFSACADSV DLFQELTRLT KAFEECTRVA GLFQPLLRAV
-~
~~~
-0
~~~
~
EKLDSWkDil FLOWXILEHO . . .I'VIKN:FH3
DYLDSIYFNR FLQHKWLERQ .PVTKNTFRQ EYQESSYFSQ FLQWKWLERQ .PVTKNTFRH EFLDSLYFLR FLQWKWLEAQ .PMGEDWFLD -
4
0
t t i o # t i # t # n # i n i # w n i # i # i t t # #~ Y # # R # # Y Y # # # # # # I # # # # # # u # t w + i it # t # t # # # wi in n a n n t # n # # w # # N N # N
HUMAN dARX HUMAN PARK2 HUMAN G R K 5
HUMAN GRK6 HUMAN GRKl BOVINE RK
HRIIGRGGFG HRIIGRGGFG YRVLGKGGFG YRVLGKGGFG YRVLGKGGFG FRVLGRGGFG O--O-OOW
HUMAN
LIW
HUMAN tIARK2 HUMAN GRK5 HUMAN GRK6 HUMAN GRX4 BOVINE RK
HUMAN GRK6
HUMAN GRX4 BOVINE RK HUMAN UARK HUMAN PARK2 HUMAN GRKS HUMAN GRK6 HUMAN GRK4 BOVINE RK
KKRIKEIKQGE TLALNERIML KKRIKEIKQGE TLALNERIML KKRIKKRKGE SMALNEKQIL KKRIKKRKGE AMALNEKQIL KKRIKKRKGE AMALNEKRIL KKRLKKRKGY QGAMVEKKIL
-
000-0
0
SLVSTGDCPF SLVSTGDCPF EKV...NSQF EKV...NSRF EKV QSRF AKV...HSRF
...
0- 0- -0
-
0
IVCMSYAFHT IVCMTYAFHT WNLAYAYET WSUYAYET WSLAYAYET IVSLAYAFET
0 -0
-
00- 0
PDKLSFILDL PDKLCFILDL KDALCLVLTI KDALCLVLTL KDALCLVLTI KTDLCLVMTI 0-
-- -
MNGGDLHYHL MNGGDLHYHL MNGGDLKFHI MNGGOLKFHI MNGGDLKFHI MNGGDIRYHI 00000--4-
##II##NIN#N
.... SQHGVF
....SQHGVF YNM..GNPGF YHM..GQAGF YNL..GNPGF YNMEDNPGF C
# # + # # # # a # # # # # # t i # #i ## # t i t i n # # t# # # # : # # # ai # aa##n#### ###xaa###a
SEMMPXYAA SEKEMRFYAT EEERALFYIU PEIRIWYAA
PANILLDEHG PANILLDEHG PENlLLD3YG P5NILLDZYG PEV1LLD:IG PENVLLDDDG
HVRISDLGLA HARISDLGLA HIIISDLGLA H1IlSC:Z:A HlRlSCLGLA NVRISDLGLA
o-ooo-
-00000000
NRFWYRDLK NRFWYRDLK ENWYR3LX RCRIVYR3L% D E. O_ R A W Y L 4 ELCCGLPDLO RERIVYRDLK . QEPRAIFYTA QIVSGLEHLH QRNIIYRDLK
-w
EIILGLEHMH EIILGLEHVH ElLCGLEDLH LICCGLEDLH
-~
-
000
-
-00000 0
i i t w i ~ i +t # i w i n t i # t o i t i i u i HSPFRQHKTK DKH.EIDRHT LRlAVELPDS HSPFRQHKTK DKH.EIDRM LTVNVELPDT QSPFRGRKEK VKFZEV3PRV LETELVYSHK QSPFQQRKKK 1KRTEYERLV KEWEEYSES L ( P F I Y I Y F Y YI(YEBMORI .. .-.. KXDtEEYSEK RGPFRARGEK VENKELKQRV LEQAVTYPDK 00 0 0-
__
SFDELDTKGI SFDEEDTKGI OF..STVI(GV QF..STVKGV QF..SAVKGI AF..STVKGV 0
00-
HUMAN
UARKZ HUMAN GRKS HUMAN G R X 6
NRLEWRGEGE NRLEWRGEGE RPSQNNSKSS R..QDCCGNC
HUMAN @ARK HUMAN #ARK2
ANGL .NGL
HUMAN ,#ARK
GKMYAMKCLD GKMYAMKCLD GKMYACKRLE GKMYACKKLE GKHYACKKLQ GKLYACKKLN 00--0 0 0
KLLDSDQELY KLLDCDQELY NLDHTDDDFY ELEPTDQDFY YLDTMEDFY AFEKROTEFF 0
-
-
APQSLLTMZE SRQNLLTMEQ PSSKTSFNHH SDSEEELPTR
R&FPL.T:SE KMPL.V:SE SKFSTGSVS: QKFATGSVPI lUIFATGCVSI QEFASGTCPI 0
0
CDF..SKKKP HASVGTHGYM APEVLQKGVA CDF..SKKKP HASVGTHGYM APEVLQKGTA W ! ? E G D L . I RSSVGTVGYH APEVL.rCN3R VHV?EGOf.I KSRVGTV;YV APEW.KUL3 TE:?EZDR.V RZRVGTVGYM A?EW.NULK VELKAGQTKT KGYAGTPGFM APELL LGEE
-
-00 0-0 000.-
~~~~~~~~
00
-
MLLTKDMQI CLLCKCPAER HLLTKXPSKR ALLQKDPEKR ~~~
00
~~
-
LGCOEECME VKRHPFFiWn NFKRLIAGXL :G:RGSSAW WEHPLFKKL NFKRLGAGUL LGCRGEGIUG VKOHPVFK3I NFRRLEANW. LGFRDGSCDG LRTHPLFRDI SWRQLEAGML ~
0 00
~~
~
~
- - o
_____
289 289
283 283 284 286
YDSSADWFSL GCMLFKLLRG YDSSADWFSL GCMLFKLLRG
387 387
YSLSPCYWL1 '%C1IYPY!t6
381
G-:'"?YIA? YT'S?XWA: YTFSP3YWGL G: IYlVlOG YDFSVDYFAL GVTLYEMIRR
381 382 385
0
0 0.-
0 0
--
-0
i # i a # ~ n : n t a w t n i # : n ~ ~ t i i i i i#i #i i i t + # FSPELHSLLE GLLQRDVNRR LGCLGRGAQE VKESPFFRSL D W Q W L Q R Y PPPLIPPRGE VNAADAFDIG FSPELKSLLE GLLQRDVSXR LGCHGGGSQE VKEHSFFKGV DWQHVYLQKY PPPLIPPRGE VNAADAFDIG
FSEEAKSICK FSPQARSLCS FSEOAKSICR FSPASKDFCE
189 189
0
# i t # t i t i # tw # i t i # i i ## i # n t # i t # #a # # u + t n n # #
0
HUMAN dARK HUMAN BARK2 HUMAN GRK5
EVYGCRKRDT EVYGCRKADT EVCACQVRAT EVCACQVRAT EVCACQVRAT EVFACQMAT 00 -0 0
-
oooo-o-
188 188
3PYtVYDPRA EPPFKPDPQA EPPFCPDPHA TPPFVPDSRT 00
0
VYCKJVL2:I :YCKDVLC:L VYCKOVL3:E VYAKNIQDVG
-
-
186 486 481 481
482 485
0-
RWQQEVAEN F3TlliAETDR :.EARKKAKNK QLSHEEDYAL CKDCIMHGYM SKNGNPFLTC WQKRYFILtP
585
RWQQEVTEN YEAVNMTDK IE*RKRAIC4K QLGHEECYAL CKDCIMHGYM LKLSIPFLTQ WQRRYFYLF?
585 556 557
PYCNEMlETE PWQNEMYETE PUQNEDCLTM PWQEEMIETG 00 0
0
C......... C......... V......... V.........
..._.._.._ ...kKELNVF GPNGTLPP3L NR.IHPPEPP KKGLLORLFK ..__._..._ ...FQELNVf GLDGSVPPDL DWKGQPPAPP KKGLLQRLFS .......... ...PSEKEVE PKQC...... .......... .......... .......... ...FGDLNVW _ _ RPDGPEIPDDM KGVSGQEAAP SSKSGMCVLS
IQSVEETQIK ERKCLLLKIR GGKQFILQCD SDPELVQWKK ELRDAYREAQ QLVQRVPKMK NKPRSPWEL SKVPLVQRGS ILSVEETQIK DKKCILFRIK GGKQFVLQCE SDPEFVQWKK ELNETFKEAQ R L L W K F L NKPRSGTVEL PKPSLCHRNS INSNHVSSNS TGSS...... L.........
.......... .......... ..__..__.. .......... .......... .......... .......... .......... .......... ._.._._.__ .......... ........_...........
FIG. 1. Comparison of amino acid sequences of human PARK, PARK2, GRK5, GRK6, and GRK4 and bovine RK. The predicted sequences were aligned using the Pileup program [Wisconsin Genetics Computer Group (GCG)]. #, Residues that are part of the kinase catalytic domain as defined by Hanks and Quinn (1991); 0, identity among all six homologues; -, similar residues. For this analysis amino acids were deemed to be similar to other residues within six groups as follows: (a) S, T, and C; (b) D and E; (c) N and Q; (d) R, K, and H; (e) Y, W, and F; and (f) M, A, I, V, and L. The amino acids are numbered on the right-hand side of the sequence. The sequences for PARK, PARKB, GRK4, GRK5, RK, and GRK6 were obtained from the following sources: Benovic et al. (1991b), Parruti et al. (1993a), Ambrose et al. (19921, Kunapuli et al. (19931, Benovic and Gomez (1993). and Lorenz et al. (1991).
defined as described by Hanks and Quinn (199U.l With the exception of specific sequences in RK, PARK, and PARK2 (described below), neither the N- nor C-terminal regulatory .domains show significant
532 561
685 685 590 576
199
G PROTEIN-COUPLED RECEPTOR REGULATION
RK BARK
100
33.4(57.8)
33.566.1)
47.0(68.1)
47.266.5)
47.1(69.0)
100
83.7 (92.0)
36.8 (56.5)
37.0 (58.3)
38.6 (59.8)
100
37.0 (57.4)
37.8 (58.3)
36.3 (57.0)
100
68.7 (82.1) 67.6 (82.5)
BARK2 GRK4 GRK5 GRK6
100
70.1 (83.5)
100
FIG. 2. Comparison of amino acid homologies between the various GRKs. Amino acid sequences from human PARK, PARK2, GRK4, GRK5, and GRKG and bovine RK were compared in a pairwise fashion using the Gap program [Wisconsin Genetics Computer Group (GCG)]. The percentage of amino acid identity is given as well as the percentage of similarity (in parentheses). The references are as listed in the legend to Fig. 1.
homology to proteins in the database outside the GRK family. As mentioned above, the sequences of some GRK catalytic subdomains (I, 11, VI, VII, and VIII) are unique to the family. Most obvious in this regard is the substitution of leucine for phenylalanine in subdomain VII (DLG instead of DFG, which is found in all but two other protein kinases). The hallmark of the original members (RK and PARK) of the GRK family is their ability to specifically recognize and phosphorylate only the agonist (or lightbactivated form of the receptor. Indeed, all GRKs (except GRK4, which has yet to be heterologously expressed in a n active form), when expressed in COS-7 or Sf9 insect cells, specifically phosphorylate PzAR and/or rhodopsin in an agonist-dependent fashion (Lorenz et al., 1991; Benovic and Gomez, 1993; Kim et al., 1993b; Kunapuli and Benovic, 1993). B. TISSUELOCALIZATION l h o members of the GRK family appear to be expressed in a limited number of tissues (Table I). RK is predominantly expressed in the retina, where it localizes to both rod and cones by immunofluorescence, and in the pineal body to a lesser extent (Somers and Klein, 1984; Palczewski et al., 1993). While GRK4 was cloned from a brain library, its transcript is not easily detectable by Northern analysis in any tissue except testis (Ambrose et al., 1992). In contrast, PARK, PARKB, GRK5, and GRKG are ubiquitous. The patterns of tissue expression for PARK, PARKB, and GRKG are similar, with each homologue being most abundant in the brain, skeletal muscle, and hema-
TABLE I MOLECULAR PROPERTIES OF THE GRKsa Parameter
RK
PARK
PARK2
GRK4
Polypeptide mol. wt. Amino acids N-terminal Catalytic C-terminal Receptor sub&ratese
62,933
79,463
79,803
561 186 266 109 Rhodopsin, P2AR Acidic
688 190 263 235 P2AR, rhodopsin, m2 mAChR, m3 mAChR, SPR, azAAR, azcAR Acidic
532 186 264 826
Peptide substrates Tissue distribution
689 190 263 236 P2AR, rhodopsin, m2 mAChR, m3 mAChR, SPR, aZAAR, azcAR Acidic
Retina > pineal body
Ubiquitous, brain, hematopoietic, skeletal muscle
Testis
+++
+
Ubiquitous, brain, hematopoietic, olfactory, skeletal muscle
Autophosphorylation CovaIent modifications
Farnesylation
?
?
?
?
* brain
GRK5
GRK6
67,786
66,074
590 185 263 142 P A R , m2 mAChR, rhodopsin
576 185 263 128 P2AR, m2 mAChR, rhodopsin
Neutral > acidicd
Neutral > acidicd
Ubiquitous heart, lung, placenta, hematopoietic, retina
Ubiquitous, brain, hematopoietic, skeletal muscle
+
?
+++
9
9
9
+ Palmitoylation
Activators
Polycations
Inhibitors Sangivamycin Heparin sensitivity ? Chromosomal localizationf mRNA size (kb) 3.land 5.8
By-subunits, PS, PA, PG, PI, PE
py-subunits, PS, PA, PG, PI, PEe
?
Polyanions, PIP,
Polyanions, PIP,
? ?
llq13
22qll
4
8
+
Polycations, PC, PE, DAG, fatty acids Polyanions
Polycations
4p16.3
10q24-qter
2.5
3
5q35 13pter-q21 2.4 and 3
+
+++
Polyanions
++
aP2AR, p,-Adrenergic receptor; mAChR, muscarinic cholinergic receptor; SPR, substance P receptor; a,,AR, a,,-adrenergic receptor that maps to human chromosome 10; IX,~AR,a,,-adrenergic receptor that maps to human chromosome 2; PS, phosphatidylserine; PA, phosphatidic acid; PG, phosphatidylglycerol; PI, phosphatidylinositol; PE, phosphatidylethanolamine;PC, phosphatidylcholine; DAG, diacylglycerol; PIP,, phosphatidylinositol-4,5-bisphosphate. bPolymerase chain reaction analysis has suggested the presence of a n insert (the sequence of which has yet to be reported) in the C-terminal domain (Inglese et al., 1993). CReceptors that can act as phosphoacceptors in a light- or agonist-dependent fashion in uitro (irrespective of efficacy) or implicated as substrates from in viuo experiments. dAll peptides contained three arginine residues at the N terminus, which aids in the isolation of peptides on phosphocellulose paper. Peptides differed in that nearest neighbors of the phosphoacceptor serine were either neutral (alanine) or acidic (glutamic acid, aspartic acid, and phosphoserine). .The effects of phospholipids on PARK2 are presumed based on the similarity between @ARKand PARK2 i n the region thought to be involved in phospholipid binding. fSee the work of Benovic et al. (1991b), Calabrese et al. (1994b), Ambrose et al. (1992), Bullrich et al. (1995), and Haribabu and Snyderman (1993).
202
RACHEL STERNE-MARR AND JEFFREY L. BENOVIC
topoietic cells (Benovic et al., 1989, 1991a; Chuang et al., 1992; Benovic and Gomez, 1993; Parruti et al., 1993a). The tissues in which GRK5 is most prevalent are the heart, placenta, lung, retina, and hematopoietic cells (Kunapuli and Benovic, 1993; Premont et al., 1994; Sallese et al., 1994). Olfactory tissue expresses single isoforms of several proteins involved in signal transduction, that is, Golfand type I11 adenylyl cyclase. In line with this observation, PARK2 is apparently the only GRK expressed in olfactory epithelium (Dawson et al., 1993; Schleicher et al., 1993). With the generation of specific antibodies against each homologue it will be possible to quantitatively assess the relative amounts of each GRK in a given tissue or cell line. In situ hybridization and immunocytochemistry have been used to investigate PARK and PARK2 localization in the brain (Arriza et al., 1992). Both PARK and PARK2 are found extensively in many, but not all, regions of the brain. The patterns of abundance of the two proteins are similar but distinguishable. PARK is more abundant in most regions, but PARK2 was preferentially expressed in some. The expression of PARK and PARK2 does not precisely parallel the localization of the P2AR, consistent with the observation that other GPRs are substrates for these enzymes in uitro (see below). Among cell types in the brain, PARK and PARK2 are found predominantly in neurons. Both proteins are found in cell bodies, postsynaptic densities, and presynaptic axon terminals. C. MEMBRANE LOCALIZATION Although GRKs phosphorylate disk or plasma membrane receptors, RK and PARK are found in the soluble fraction following disruption of rod outer segments or cells, respectively. Upon agonist stimulation PARK activity has been reported to translocate from the cytosol to the membrane (Strasser et al., 1986; Mayor et al., 1987; Chuang et al., 1992). However, receptor activation alone may be insufficient to direct membrane translocation, since the GRKs appear to utilize distinct mechanisms to specify their subcellular localization. The C-terminal domain of GRKs plays a n important role in the membrane association of these kinases. The sequence of rhodopsin kinase terminates with -CVLS, which follows the consensus “CAAX box” (where C is cysteine, A is a small aliphatic residue, and X is a n uncharged amino acid), directing farnesylation (C15 isoprenylation) and carboxymethylation of the resulting farnesylated cysteine (Gibbs, 1991). Indeed, RK is farnesylated (Anant and Fung, 1992; Inglese et al., 1992a). Isoprenylation is neces-
G PROTEIN-COUPLED RECEPTOR REGULATION
203
sary for the membrane localization of one class of yeast mating pheromones; members of the ras superfamily, including some small G proteins involved in vesicular trafficking; the y-subunits of heterotrimeric G proteins; and the a-subunit of cGMP phosphodiesterase. Similar to small G proteins, but unlike the Ras oncoprotein, RK is not constitutively associated with the membrane and probably cycles on and off the membrane. In an in uitro assay RK associates with rod outer segments only in the presence of light (Inglese et al., 1992a,b). A mutant RK that lacks the isoprene moiety fails to translocate to the membrane upon light activation and also has diminished kinase activity. Conversely, a mutant that is modified with the more lipophilic C20 geranylgeranyl group is associated with rod outer segments independent of light activation and retains full kinase activity. Thus, isoprenylation not only directs RK to the membrane, but probably orients the kinase for optimal phosphorylation of its substrate. Unlike other GRKs, the ability of PARK and PARK2 to phosphorylate P2AR, M, muscarinic cholinergic receptor (m2 mAChR) and rhodopsin is significantly enhanced by brain Ply-subunits (Haga and Haga, 1990, 1992; Pitcher et al., 199213) and the Ply binding site of PARK has been mapped to residues 546-670 (Koch et al., 1993). This domain overlaps a recently described -100-amino acid domain (residues 553-656 of PARK), called the pleckstrin homology (PHI domain, found in cytoskeletal proteins as well as proteins involved in various signal transduction pathways (Musacchio et al., 1993; Gibson et al., 1994). While PH domains from several proteins are capable of binding Ply subunits, it is the C-terminal half of the PH domain and residues distal to the end of the recognized PH domain that are involved in the P/y interaction (Touhara et al., 1994). It has independently been proposed that residues 639-670 of PARK form a three-stranded coiled coil with polypeptide strands from the P- and y-subunits (Simonds et al., 1993), and this suggestion is consistent with the presence of a C-terminal a-helix, determined from the crystal and solution structures of PH domains (Macias et al., 1994; Yoon et al., 1994). It should be noted, however, that full-length PARK binds to ply-subunits with a Kd of -30 nM (Kim et al., 1993a), which is at least 10-fold tighter than the IC,,s (determined by phosphorylation inhibition or translocation inhibition assays) of fusion proteins bearing the entire C-terminal domain of PARK interacting with Ply-subunits (Koch et al., 1993; Touhara et al., 1994). Thus, it is possible that residues in the N-terminal and/or catalytic domain of PARK contribute to its high-affinity binding to p l y-subunits. The C-terminal domains of PARK and PARK2 are -120 amino acids
204
RACHEL STERNE-MARR AND JEFFREY L. BENOVIC
longer than that of RK (Fig. l),and truncation of the PH domain of PARK t o produce a RK-sized kinase abolishes Ply-inducible activity and even lowers the basal activity (Koch et al., 1993). Engineering a site for the addition of a geranylgeranyl group to the truncated PARK restores kinase activity up to -50% of the level observed with fulllength PARK in the presence of Ply-subunits. Since the Ply-subunits are constitutively associated with the membrane by virtue of geranylgeranylation of the y-subunit, Ply binding to PARK likely plays the same role in membrane localization as isoprenylation does for RK. Some PH domains also bind the membrane lipid phosphatidylinositol4,5-bisphosphate (PIP,), and residues in the N terminus of the PH domain are involved in this interaction (Harlan et al., 1994). In fact, recent studies have demonstrated that PARK can bind directly t o phospholipid vesicles containing either PIP, or phosphatidylserine (DebBurman et al., 199513). Interestingly, PIP, inhibits PARK activity severalfold, while PS, phosphatidylethanolamine,phosphatidylinositol, phosphatidic acid, and phosphatidylglycerol (but not phosphatidylcholine) all activate PARK -2-fold. Thus, these lipids may play a role in PARK localization and regulation. When cells from tissues or cells in culture are lysed in low or physiological salt concentrations by disruption utilizing sheer force, most of the immunoreactive PARK is found in the soluble fraction. In contrast, when cells are gently lysed in the presence of osmotic support, a substantial amount of PARK activity (as measured by the ability to phosphorylate rhodopsin) is associated with the membrane fraction (GarciaHiguera et al., 1994).In fact, sucrose gradient fractionation and immunoelectron microscopy studies indicate that 39-50% of the PARK is associated with the endoplasmic reticulum, 14-18% is associated with the plasma membrane, and 31-43% is associated with the cytoplasm. Furthermore, PARK can bind to microsomal membranes in a reversible and saturable manner. While the interaction of PARK with intracellular organelles is intriguing, the function of this association is currently unknown. Residues located in the C-terminal domain of GRK5 undergo autophosphorylation, which can be stimulated by the presence of various lipids, including phosphatidylcholine, phosphatidylethanolamine, diacylglycerol, myristic acid, and palmitic acid (Kunapuli et al., 1994a). Direct binding of GRK5 to phosphatidylcholine vesicles can be blocked by a fusion protein containing the last 102 amino acids of the kinase, implicating this domain in lipid binding. Unlike RK, which undergoes light-dependent association with rod outer segments in an in uitro translocation assay, GRK5 associates with rod outer segment membranes in the dark or light in the absence or presence of ATP (Premont
G PROTEIN-COUPLED RECEPTOR REGULATION
205
et al., 1994). Thus, it seems possible that in the cell GRK5 is constitutively localized at the membrane. GRKG is covalently modified by palmitoylation of one or more cysteine residues located 12-15 amino acids from the C terminus of the protein at positions 561, 562, and 565 (Stoffel et al., 1994). When GRKG is overexpressed in Sf9 insect cells, all of the palmitate-conjugated protein is localized to the membrane fraction, while the majority of GRKG remains soluble. It is not clear whether the presence of cytosolic GRKG is a result of saturation of the acylation machinery and whether a greater percentage of GRKG may be palmitoylated when the protein is expressed at endogenous levels. Thus, it is not yet clear in normal cells whether GRKG is constitutively associated with the membrane or whether its membrane localization is modified by receptor activation.
D. REGULATION One of the fascinating features of GRKs is their ability to specifically recognize the light- or agonist-activated form of the receptor. Rhodopsin is phosphorylated by RK at as many as seven sites located within a 10-amino-acid stretch of the C terminus of rhodopsin (Wilden et al., 1986). However, since the C-terminal tail of rhodopsin can serve as a substrate for proteases in the absence of light, it is not likely that activation recognition simply involves exposure of the substrate serines and threonines to the kinase. Instead, since activated rhodopsin promotes the phosphorylation of free peptides by RK, intracellular domains of the receptor must alter the kinase, thereby converting it from an inactive to an active enzyme (Fowles et al., 1988). A form of rhodopsin that lacks the C-terminal tail (G32Q-Rho)is sufficient to activate RK, suggesting that the substrate peptide is not involved in kinase activation (Palczewski et al., 1991b). Proteolytic clipping of the third intracellular loop abrogates activation, implicating this loop in RK activation, analogous t o the requirement of this loop for GT activation. Stimulation of peptide phosphorylation by light-activated G329Rho (G329-Rho") is a result of increasing the V, as well as lowering the K,. Peptides from the first, second, and third loops and the C-terminal tail had previously been shown, by stimulation or inhibition of kinase activity, to interact with RK (Palczewski et al., 1988, 1989a; Kelleher and Johnson, 1990). Furthermore, the wasp venom peptide mastoparan, which has been shown to mimic the receptor activation of G proteins, also stimulates peptide phosphorylation by RK (Palczewski et al., 1991b). An antipeptide antibody that recognizes residues 17-34 of RK blocks the ability of RK to phosphorylate rhodopsin. Since this
206
RACHEL STERNE-MARR AND JEFFREY L. BENOVIC
antibody has no effect on the catalytic activity, as judged by the ability of RK to phosphorylate exogenous peptides in the presence of antibody, these experiments suggest that the N terminus of RK may be a site of activation by rhodopsin (Palczewski et al., 1993). Light-activated rhodopsin (Rho”) and agonist-treated P2AR and m2 mAChR each activate PARK to phosphorylate exogenous peptides (Chen et al., 1993; K. Haga et al., 1994). The EC5,s for p2AR and rhodopsin are 12 nit4 and -1 FM,respectively, suggesting that PARK interacts with P2AR with a much higher affinity than it interacts with Rho, consistent with kinetic studies (Benovic et al., 1987b).Residues in the first intracellular loop and the N-terminal portion of the third intracellular loop of P2AR are likely candidates for involvement in kinase activation, since peptides encompassing these regions are potent inhibitors of receptor phosphorylation by PARK (Benovic et al., 1990). Peptides derived from the second and third loops and C-terminal tail of the m2 mAChR as well as mastoparan also stimulate PARK to phosphorylate the m2 mAChR and a substrate consisting of a portion of the third intracellular loop of m2 mAChR linked to glutathione-S-transferase (GST) (GST-3I-m2AChR) (K.Haga et al., 1994).Some peptides, especially those that lack acidic residues, are extremely poor substrates for PARK (Benovic et al., 1990; Onorato et al., 1991). However, activation by Rho* can increase the catalytic efficiency (VmaX/Km) of a poor substrate (RRRASAAASAA) as much as 200-fold (Chen et al., 1993). By comparison, the catalytic efficiency of a “good” substrate is increased only -10-fold. Thus, the acidic nature of the substrate is not essential for catalysis, and instead, it is possible that, once activated, the kinase may phosphorylate any serine or threonine in close proximity to the substrate binding pocket. G protein Ply-subunits increase the initial rate of PARK (and PARK21 phosphorylation of Rho, P,AR and muscarinic acetylcholine receptors 10-fold (Haga and Haga, 1992; Pitcher et al., 1992b; Kim et aZ., 1993a,b). Brain Ply-subunits are more effective than transducin P l y (Pitcher et al., 1992b), and the effect is observed with Ply-subunits derived from purified G,, Gi, and Go (Haga and Haga, 1992). As described above, because ply-subunits are membrane associated due to the isoprenylation of the y-subunit, interaction of PARK with the PI y-subunits may serve as a mechanism for membrane association of the kinase. Allowing PARK to diffuse in two, rather than three, dimensions to find the receptor in the membrane can account for a dramatic increase in initial rates of phosphorylation. However, Pl y-subunits do have a small but detectable capacity to increase the phosphorylation of exogenous peptides (Kim et al., 1993a). Furthermore, the stimulation
-
G PROTEIN-COUPLED RECEPTOR REGULATION
207
of PARK activity by activated receptor together with Ply-subunits is greater than the sum of the stimulation by either of these agents alone. This synergy has been demonstrated in five assay systems: direct binding of PARK to Ply and P2AR; agonist-treated P2AR and P l y stimulation of peptide phosphorylation; G329-Rho*and P l y stimulation of peptide phosphorylation; mastoparan and P l y stimulation of m2 mAChR phosphorylation; and mastoparan and P l y stimulation of GST-3I-m2AChR (Kim et al., 1993a; K. Haga et al., 1994). Therefore, it is clear that a ternary complex consisting of PARK, P l y , and activated receptor is formed, and the simultaneous interaction of PARK with receptor and Ply-subunits generates the maximally active kinase. RK and GRK5 undergo intramolecular autophosphorylation at multiple sites, while autophosphorylation of PARK, PARKB, and GRK6 is substoichiometric (Kelleher and Johnson, 1990; Buczylko et al., 1991; Kunapuli et al., 1994a; Loudon and Benovic, 1994). Autophosphorylation of RK occurs rapidly to yield three phosphates per mole and more slowly to achieve the maximal phosphorylation of four phosphates per mole (Kelleher and Johnson, 1990; Buczylko et al., 1991). The serines and threonine at positions 21, 488, and 489 have been identified as phosphoacceptors (Palczewski et al., 1992). Phosphorylated and unphosphorylated RKs do not differ appreciably in their ability to bind to (Buczylko et al., 1991) or utilize light-activated rhodopsin as substrate (Kelleher and Johnson, 1990). However, phosphorylated RK binds only weakly to phosphorylated and light-activated rhodopsin (Rho*-PI,while unphosphorylated RK demonstrates significant binding to Rho*-P (Buczylko et al., 1991). Based on these results, it has been suggested that autophosphorylation of RK may aid the dissociation of the kinase from the receptor following its phosphorylation. Phosphorylation of receptors by GRK5 requires prior autophosphorylation of S484 andlor T485, since substitution of alanine a t these two positions abolishes autophosphorylation as well as Rho* and &AR* phosphorylation (Kunapuli et al., 1994a). Interestingly, autophosphorylation is stimulated by several types of lipids, but not specifically by the activated receptor (Kunapuli et al., 1994a; Premont et al., 1994). Unphosphorylated GRK5 can bind to lipid, since the autophosphorylation-defective mutant effectively competes with the wild-type protein for binding to lipid vesicles. Light-activated rhodopsin stimulates the phosphorylation of exogenous peptides by GRK5, although the level of activation is low relative to Rho* stimulation of PARK. Although GRK5 phosphorylation of rhodopsin is strictly light dependent, significant phosphorylation of P2AR occurs in the presence of antagonist (Premont et al., 1994). Phospholipid-stimulated autophos-
208
RACHEL STERNE-MARR AND JEFFREY L. BENOVIC
phorylation of GRK5 may be analogous to P l y binding to PARK in that it provides a means of membrane localization as well as sufficient activation to allow phosphorylation of an antagonist-occupied receptor. While palmitoylation certainly plays a role in the membrane localization of GRKG, the lipid may also be required for receptor activation of the kinase. This is a prediction based on the assumption that the GRKG that has been characterized to date was purified from Sf9 insect cells, in which the majority of the expressed kinase is not palmitoylated. Relative to GRK5 and PARK, GRKG displays a poor capacity to phosphorylate receptors but a comparable ability to phosphorylate nonreceptor substrates, such as casein and phosvitin (Loudon and Benovic, 1994). E. RECEPTORSPECIFICITY Upon stimulation of intact cells with various agonists, soluble “PARK” activity diminishes, while membrane-associated kinase activity increases. In S49 lymphoma cells isoproterenol (a P-agonist) and prostaglandin El (PGE,) treatment promote an -80% decrease in soluble PARK activity, while increasing the membrane kinase activity -10-fold (Strasser et al., 1986). Somatostatin induces -50% diminution in soluble PARK activity in S49 lymphoma cells (Mayor et al., 1987). In mononuclear leukocytes platelet-activating factor (PAF) decreases soluble PARK activity -3-fold, while increasing kinase activity in the particulate fraction over 2-fold (Chuang et al., 1992). These results suggest that P-agonists, PGE,, somatostatin, and PAF induce translocation of PARK from the cytosol to the plasma membrane and imply that PARK plays a role in the desensitization of the cognate receptors. If so, PARK would be involved in the regulation of at least three types of GPRs: those coupled to the stimulation of adenylyl cyclase (PAR and PGE,), stimulation of phospholipase C (PAF),and inhibition of adenylyl cyclase (somatostatin). Interestingly, in DDT,-MF2 smooth muscle cells, in which a P-agonist increases membrane PARK activity, an a,AR agonist did not alter PARK activity (Strasser et al., 1986). In view of the current understanding of GRKs, these results must be put in perspective. Since the Ply-subunits from Gs, Gi, and Go all stimulate PARK activity (Haga and Haga, 1992), agonist stimulation may induce membrane association of PARK irrespective of receptor interaction. Thus, PARK translocation may be necessary but insufficient for receptor phosphorylation. Furthermore, the assays used to measure PARK activity (phosphorylation of P2AR and Rho) do not distinguish between GRKs. Thus, the “PARK” activity in these studies
G PROTEIN-COUPLED RECEPTOR REGULATION
209
could also have included PARK2, GRK5, GRK6, and possibly other activities. Direct interaction of GRKs with receptors following agonist treatment may be necessary to infer specificity in uiuo. Several combinations of receptor and GRK preparations have been used to assess the specificity of GRKs in uitro (summarized in Table 11). With the exception of rhodopsin, native receptors in crude membranes do not exist in sumcient quantity to allow their detection by kinase assay due to the substantial “background” generated by other endogenous membrane kinases and their substrates. Therefore, receptor purification from endogenous sources was initially used to assess phosphorylation of the P2AR (Benovic et al., 19861, m2 mAChR (Kwatra et al., 19891, and a,*AR (Benovic et al., 1 9 8 7 ~by ) PARK. Alternatively, overexpression of receptors in Sf9 insect cells followed by affinity chromatography has provided ample amounts of specific receptors: P2AR (Kim et al., 1993b), m2 mAChR (Richardson et al., 1993), and substance P receptor (Kwatra et al., 1993). Likewise, PARK, pARK2, GRK5, and GRKG have all been overexpressed in and purified from Sf9 cells (Kim et al., 1993b; Kunapuli et al., 199413; Loudon and Benovic, 1994). Using in uitro assays with purified P,AR, m2 mAChR, and GRKs, PARK and PARK2 (in the presence of Ply-subunits) are the best kinases, as determined by initial rate and extent of phosphorylation. In fact, no significant differences between PARK and PARK2 have been observed in uitro. GRK5 has similar or slightly less activity than PARK and pARK2, while GRKG has significantly less activity using these receptors as substrates. It must be noted, however, that GRKG may not be fully activated in these assays, because it is likely that the majority of the preparation is not palmitoylated (Stoffel et al., 1994). Likewise, GRK5 may require co- or post-translational modification or a cofactor for full activation. Enrichment of a plasma membrane fraction by sucrose gradient centrifugation (Pei et al., 1994) and stripping total membranes with 4 M urea (DebBurman et al., 1995a) are techniques that have recently been used to demonstrate the agonist-dependent phosphorylation of receptors in membranes from Sf9 cells overexpressing GPRs. These procedures have obviated the need to purify and reconstitute receptors into lipid vesicles. Using urea-treated membranes as substrates, the m2 and m3 mAChRs are both much better substrates for PARK and PARK2 than they are for GRK5 and GRKG (DebBurman et al., 1995a). With sucrose gradient-purified membranes aZcAR is better substrate for PARK and PARK2 than it is for GRK5, while PzAR is good substrate for PARK, pARK2, and GRK5 (Pei et al., 1994). One feature of the m2 mAChR and CX,~AR in membranes that distinguishes them
TABLE I1 GRK RECEPTOR SPECIFICITY Assay
Kinase preparation
Receptor preparation
I n vitro phosphorylation
Purified
Purified
I n vitro phosphorylation
Purified
Sf9 membranes from cells overexpressing receptor
Crude soluble fraction from Sf9 cells overexpressing kinase In vivo desensitization Injection of RNA into Xenopus oocytes (Ca2+ mobilization) In vivo desensitization Overexpression of kinasedefective PARK mutant (CAMPproduction) in BEAS-2B cells GRK antibody inhibi- Endogenous tion of desensitization in permeabilized olfactory epithelium
Zn vitro phosphorylation
Receptor
GRK preference
Rhodopsin PZAR %*AR m2 mAChR PzAR azcAR m2 mAChR m3 mAChR PAR rhodopsin
PARK-PARK~ZGRK~>GRK~O PARK- PARKB?GRKS>GRKG PARK pARK-pARKS>GRK5>GRKG PARK=PARKBzGRKB PARK = PARK2 > GRK5 PARK PARK2 + GRKS, GRKG PARK = PARK2 + GRKS, GRKG PARK GRKB > GRKG PARK > GRKS > GRKG
Injection of RNA into Xenopus oocytes Endogenous
Thrombin receptor PzAR > PGE, receptor*
PARK2 > PARK > RK
Endogenous
Citralva (olfactory receptor)
Purified
i=
OIt is likely that the majority of the GRKG used in these experiments was not palmitoylated, potentially resulting in lower levels of receptor phosphorylation (see text). *The “PGE, (prostaglandin E,) receptor” appears to be the EP2 prostanoid receptor (J. Regan, personal communication).
G PROTEIN-COUPLED RECEPTOR REGULATION
211
from purified and reconstituted receptors is that the stoichiometry of phosphorylation by PARK in uitro is similar to the observed level of receptor phosphorylation in whole cells following agonist treatment (Pei et al., 1994; DebBurman et al., 1995a). In contrast, purified and reconstituted receptors incorporate -2-fold more phosphate (relative to the in uiuo level) when assayed in uitro. By virtue of its preferential expression in olfactory epithelium relative to PARK, PARK2 appears to be specifically involved in the regulation of odorant receptors (Dawson et al., 1993; Schleicher et al., 1993). When added to permeabilized olfactory cilia, PARK2, but not PARK, antibodies block desensitization of odorant receptors coupled to the production of CAMP.In isolated olfactory cilia the simultaneous addition of a mixture of odorants results in the dose-dependent translocation of PARK2 immunoreactivity from the soluble to the particulate fraction (Boekhoff et al., 1994). Further, a GST fusion protein encoding the C terminus of PARK2, when added to permeabilized cilia, prolongs the odorant-induced stimulation of cAMP and blocks the phosphorylation of putative membrane receptors. These results suggest that upon odorant stimulation, PARK2 translocates to the ciliary membrane where receptors or other membrane proteins are phosphorylated, leading to the rapid diminution of cAMP production. Since Protein kinase A is also required for desensitization of the odorant response (Schleicher et al., 1993), the details of these events may be a permutation of the current models for desensitization of rhodopsin by RK and arrestin and PzAR by PARK and parr. PARK2 also appears to regulate the thrombin receptor in an in uiuo reconstitution system (Ishii et al., 1994). When RNA for the thrombin receptor is coinjected into Xenopus oocytes along with RNA encoding PARK, PARK2, or RK, only the PARK2 RNA significantly diminished thrombin receptor signaling, even though lysates prepared from oocytes injected with each of the GRKs could aptly phosphorylate P2AR or rhodopsin. Injection of mutated forms of PARK2 RNA, including deletions of subdomains I and I1 or site-directed mutagenesis of subdomain VII of PARK2, leaves thrombin receptor signaling unaltered. Likewise, substitution of serine and threonine residues in the C-terminal tail of the receptor with alanine residues allows the receptor to respond to thrombin but yields a receptor that is refractory to coinjection of PARK2 RNA. These results strongly implicate PARK2 in the regulation of the thrombin receptor, albeit in a n environment where this receptor is not naturally expressed. Another approach to assessing the in viuo specificity of GRKs is the use of dominant-negative mutations. An invariant lysine in subdo-
212
RACHEL STERNE-MARR AND JEFFREY L. BENOVIC
main I1 of the kinase family is required for phosphoryl transfer, and therefore mutation of this residue results in a kinase-deficient protein. Substitution of this invariant lysine (K220 in PARK) abolishes the kinase activity of PARK. Since the regulatory N- and C-terminal domains of PARK remain intact, this kinase-deficient mutant can effectively compete with the wild-type kinase for phosphorylation of activated P2AR in uitro (Kong et al., 1994). When this mutant is stably overexpressed in BE AS-2B cells [a bronchial epithelial cell line endogenously expressing P2AR and PGE, receptors that undergo homologous desensitization (Penn et al., 1994)], short-term desensitization of the PzAR is mitigated, allowing CAMPto accumulate 2- to 2.5-fold over wild-type cell levels (Kong et al., 1994). These results are the strongest evidence to date that PARK regulates the P2AR in uiuo. Since the desensitization of PGE, receptors is unaffected by the kinase-deficient mutant, PARK demonstrates a preference toward desensitization of the P2AR. However, it is possible that the overexpression of the kinase mutant also hampers the activities of other endogenous kinases. Therefore, it remains to be determined whether kinase-deficient versions of PARK2, GRK5, or GRKG demonstrate similar or unique specificities.
F. DISTINGUISHING FEATURESOF MEMBERS OF THE GRK FAMILY Several characteristics of members of the GRK family allow their distinction (Table I). By sequence homology PARK and PARK2 are members of a subfamily, while RK, GRK4, GRK5, and GRKG belong to a second subfamily. PARK and PARK2 bind to and are activated by G protein Ply-subunits, while the other GRKs are insensitive to Pi y-subunits (Haga and Haga, 1992; Pitcher et al., 1992b; Kim et al., 1993a; Kunapuli and Benovic, 1993; Loudon and Benovic, 1994). In contrast, the catalytic activities of RK (Palczewski et al., 1989a), GRK5 (Kunapuli et al., 1994b), and GRKG (Loudon and Benovic, 1994) are stimulated by the polycations spermine and spermidine, while PARK and PARK2 are insensitive to these agents. RK (Anant and Fung, 1992; Inglese et al., 1992b) and GRKG (Stoffel et al., 1994) are covalently modified by farnesylation and palmitoylation, respectively. Other characteristics do not seem t o correlate with membership in either subfamily. The sensitivity of GRKs to the polyanion heparin varies over at least three orders of magnitude, with the rank order of sensitivity being GRK5 > GRKG > PARK, PARK2 > RK (Palczewski et al., 1989a; Kim et al., 1993b; Kunapuli et al., 199413; Loudon and Benovic, 1994). PARK and RK phosphorylate acidic peptides in uitro (Benovic et al., 1990; Onorato et al., 19911, while GRK5 and GRKG prefer neutral
G PROTEIN-COUPLED RECEPTOR REGULATION
2 13
or basic peptides (Kunapuli et al., 1994b; Loudon and Benovic, 1994). Finally, autophosphorylation of RK and GRK5 occurs a t 2-3 mol of phosphate per mole of kinase, while autophosphorylation of PARK, pARK2, and GRK6 is substoichiometric (Palczewski et al., 1992; Kunapuli et al., 1994a). 111. ARRESTINS A. CLONING Visual arrestin was initially identified as a major protein that redistributed (along with RK) from the cytoplasm to the disk membrane following light activation of rod outer segments (Kuhn, 1978). Molecular cloning of visual arrestin utilized monoclonal and polyclonal antibodies directed against bovine arrestin to probe a bovine retinal cDNA expression library (Shinohara et al., 1987; Yamaki et al., 1987). Interestingly, arrestin cDNAs were represented in almost 1%of the plaques in the library. Evidence for a n arrestin homologue was initially suggested during the course of PARK purification, when it was found that crude preparations of the kinase could adequately uncouple P2AR from G,, while purified PARK did so to a much lesser extent (Benovic et al., 1987a). Moreover, purified retinal arrestin could increase the efficiency of P,AR/Gs uncoupling by PARK, demonstrating that a n arrestin could indeed specifically interact with the PARK-phosphorylated &AR. Using the visual arrestin cDNA as a probe under conditions of low stringency, a molecular search for a homologue resulted in the identification of a bovine brain protein, parr, which could uncouple P2AR/Gs with 20-fold greater efficacy than retinal arrestin (Lohse et al., 1990). Subsequently, the cDNAs for retinal arrestin alone, parr alone, or a mixture of retinal arrestin and parr were used as probes to isolate a third arrestin homologue. This homologue is variously called hTHY-ARRX (from human thyroid) (Rapoport et al., 1992), parr2 (from rat brain) (Attramadal et al., 19921, and arr3 (from bovine brain) (Sterne-Marr et al., 1993). A fourth homologue was isolated by preparing a retinal-enriched cDNA library (by depleting “generic” sequences with fibroblast cDNA) and using sequence analysis to screen for arrestin homologs. Because this homologue maps to the human X chromosome, it is called X-arrestin (Murakami et al., 1993). This fourth homologue (also called C-arrestin, cone arrestin, or CAR) was also identified in a PCR screen for arrestins utilizing a sense primer designed from the sequence LKHEDTN, which is conserved in all ar-
214
RACHEL STERNE-MARR AND JEFFREY L. BENOVIC
restins, and a vector-derived antisense primer (Craft et al., 1994). For the purpose of this chapter, we refer to the 48-kDa originally identified arrestin as visual or retinal arrestin (arr) (even though the fourth arrestin probably plays a role in the visual system as well); the first arrestin, which was shown to preferentially uncouple the P,AR/G, interaction, as parr (or arr2); the third arrestin homologue as arr3 (because this seems less inappropriate than hTHY-ARRX or parr2); and the fourth arrestin as arr4. The members of the arrestin family are very similar, with the residues corresponding to 16-349 of human retinal arrestin being 45% identical and 70% similar among all four homologues (Figs. 3 and 4). Each arrestin is most similar to parr with the identity varying from 56.5% to 78.3% and the similarity varying from 74.4% to 88.5%. Among the arrestins there are four variable regions that represent islands of disparity, designated V-I, V-11, V-I11 and V-IV (Fig. 3). In relation to retinal arrestin numbering, these include the N terminus to residue 15 (V-I),residues 97-104 (V-111, residues 364-372 (V-1111, and residue 385 to the C terminus (V-IV).While the C-terminal 30 amino acids of each arrestin are very acidic, there are three regions that are enriched in basic residues: the N terminus through residue 33 (B-I), residues 156- 180 (B-IT), and residues 237-242 (B-111). Straddling the V-IV region of retinal arrestin is the sequence ARHNLKDAGEA, which bears significant homology to the C terminus of the a-subunit of transducin (IKENLKDCGLF). Since this region of transducin and other G protein a-subunits has been implicated in receptor binding (for a review see Conklin and Bourne, 19931, it has been speculated that this region of arrestin might play a similar role. However, this homology is much weaker when arr is compared to other G protein a-subunits and when the other arrestins are compared to transducin. Furthermore, binding studies with truncation mutants of arr have not invoked a role for this region of arr in receptor binding (see below). Nevertheless, this homology is intriguing.
FIG. 3. Comparison of amino acid sequences of bovine parr and human arr3, arr4, and arr. The predicted sequences were aligned using the Pileup program [Wisconsin Genetics Computer Group (GCG)]. 0, Identity among all four homologues; -, similar residues. Similarity was determined as in the legend to Fig. 1. V-I, V-11,V-I11 and V-IV indicate regions of greatest variability between the four arrestins. B-I, B-11, and B-I11 designate three regions that are highly enriched in basic residues. The amino acids are numbered on the right-hand side of the sequence. The sequences were obtained from the following sources: Lohse et al. (1990), Rapoport et al. (1992), Murakami et al. (19931, and Yamaki et al. (1988).
215
G PROTEIN-COUPLED RECEPTOR REGULATION
P I
********** ***** v-1
bovine parr human a r r 3 human arrd human a r r
.......MGD K.GTRVFKKA SPNGKLTVYL GKRDFVDHID LVEPVDGWL ....... MGE KPGTRVFKKS SPNCKLTVYL GKRDFVDHLD KVDPVDGVVL .......... ..MSKVFKKT SSNGKLSIYL GKRDFVDHVD TVEPIDGWL MAASGKTSKS EPNHVIFKKI SRDKSVTIYL GNRDYIDHVS WQPVDGWL -0oo
--a0 0
0
oo--co-
42 43 38 50
00-00000
*** bovine parr human a r r 3 human a r r 4 human a r r
VDPEYLKERR VYVTLTCAFR YGREDLDVLG LTFRKDLFVA NVQSFPPAPE 92 VDPDYLKDRK VFVTLTCAFR YGREDLDVLG LSFRKDLFIA TYQAFPPVPN 93 VDPEYLKCRK LFVMLTCAFR YGRDDLEVIG LTFWLYVQ TLQWPAESS 88 VDPDLVKGKK VYVTLTCAFR YGQEDVDVIG LTFRRDLYFS RVQVYPPVGA 100
ooo- -0 --
--0 ooo333 00
-o--o-o
0-03-00--
0
0
v-11 DKK.PLTRLQ ERLIKXLGEH AYPFTFEIPP NLPCSVTLQP GPEDTGKACG PPR.PPTRLQ DRLLRKLGQH AHPFFFTIPQ NLPCSVTLQP GPEDTGKACG SPQGALTVLQ ERLLHKLGDN A Y P F T W NLPCSVTLQP GPEDAGKPCG AS..TPTKLQ ESLLKKLGSN TYPFLLTFPD YLPCSVMLQP APQDSGKSCG
I*****
bovine parr human a r r 3 human a r r 4 human a r r
000-0--oOo
bovine parr human a r r 3 human a r r 4 human arr
0-
0
0-
0
- 0
-o-o-oooo--0o
0---
- 0 0
0---ooo-m--00
0-0
0-000
.DDDIWEDF ARQFUKGMKD DKEEEEDGTG SPRLNDR TDDDIVFEDF ARLRLKGMKD DDYDDQLC.. SSEDIVIEEF TRKGEEESQK AVEAEGDEGS QDANLVFEEF ARHNLKDAGE AEEGKRDKND ADE....
....... .......
-00-0
0
288 289 285 298
338 334 335 346
0
................... ................
******** ********** ******* bovine W r r human a r r 3 human a r r 4 human a r r
00
v-111 * * ********** ********** ********** ASSDVAVELP FTLMHPKPKE E....PPHRE VPEHETPVDT NLIELDTN.. DVSVELP FVLMHPKPHD HIPLPRPQSA APETDVPVDT NLIEFDTNYA TASDVGVELP LVLIHPKPSH EA........ A TSSEVATEVP FRLMHPQPED PA........ KESI -0
238 239 235 248
-00-0
LDGKLKHEDT NLASSTLLRE GANREILGII VSYKVKVKLV VSRGGLLGDL LDGKLKHEDT NLASSTIVKE GANKEVLGIL VSYRVKVKLV VSRGG..... LDGKLKHEDT NLASSTIIRP GMDKELLGIL VSYKVRVNLM VSCGGILGDL LDGKIKHEDT NLASSTIIKE GIDRTVLGIL VSYQIKVKLT VS..GFLGEL
...
188 189 185 198
0000-
00--00000
ADICLFNTAQ YKCPVAMEEA DDTVAPSSTF CKVYTLTPFL ANNREKRGLA ADICLFSTAQ YKCPVAQLEQ DDQVSPSSTF CKVYTITPLL SDNREKRGLA TDWLYSLDK YTKTVFIQEF TETVAANSSF SQSFAVTPIL AASCQKRGLA ANWLYSSDY YVKPVAMEEA QEKVPPNSTL TKTLTLLPLL ANNRERRGIA
o300-00000---
bovine parr human a r r 3 human a r r l human a r r
0-
-oooo-oo
0 0---0-
-0-
bovine parr human arr3 human a r r 4 human a r r
00000000000000
8-111 QFLMSDKPLH LEASLDKEIY YHGEPISVNV HVTNNTNKTV KKIKISVRQY HFLMSDRSLH LEASLDKELY YHGEPLNVNV HVTNNSTKTV KKIKVSVRQY RFLLSAQPLQ LQAiQCXUWH YHGEPISVNV SINNCTNKVI KKIKISVDQI QFFMSDKPLH LAVSLNREIY FHGEPIPVTV TVTNNTEKTV KKIKACVEQV 0-0
bovine p a r r human a r r 3 human a r r 4 human a r r
-
8-11 VDYEVKAFCA ENL...EEKI HKRNSVRLVI RKVQYAPERP GPQPTAETTR VDFEIWCA KSL...EEKS HKRNSVRLVI RKVQFAPEKP GPQPSAETTP IDFEVKSFCA ENP...EETV SKRDYVRLW WQFAPPEA GPGPSAQTIR VDFEVKAFAT DSTDAEEDKI PKKSSVRYLI RSVQHAPLEM GPQPRAEATW -0-0-- 0
bovine parr human a r r 3 human a r r d human a r r
00
141 142 138 148
382 381 382 372
V-N 418 409 388 405
216
RACHEL STERNE-MARR AND JEFFREY L. BENOVIC
Darr arr
100
parr arr3 am4
&
57.9 (76.4) 56.5(74.3) 50.5(70.6)
100
78.3(88.5) 59.7(77.5) 100
56.5 (74.4) 100
FIG.4. Comparison of amino acid homologies between the various arrestins. Amino acid sequences from human arr, arr3, and arr4 and bovine parr were compared in a pairwise fashion using the Gap program [Wisconsin Genetics Computer Group (GCG)I. The percentage of amino acid identity is given as well as the percentage of similarity (in parentheses). The references are as listed in the legend to Fig. 3.
B. POLYPEPTIDE VARIANTS Three of the four arrestin homologues are expressed as polypeptide variants. Bovine retinal arrestin is expressed in three forms: the originally described 404-residue form; a 370-amino-acid form, called p44 (recently isolated from high-salt extracts of rod outer segment membranes), in which the last 35 residues of the 404-amino-acid form are replaced by a single alanine (Palczewski et al., 1994; Smith et al., 1994); and a 396-residue form that lacks the eight homologous amino acids encoded by exon 13 of the human gene (residues 338-345) (Yamaki et al., 1990; Parruti et al., 199313). The 404-residue form is - 10-fold more abundant than the 370-residue form in bovine retina (Palczewski et al., 1994). In addition to its expression in the retina, “visual” arrestin is present at much lower levels in the cerebellum and in leukocytes (Parruti et al., 1993b). While the retinal and cerebellar forms of visual arrestin have 404 amino acids, a small proportion of the leukocyte visual arrestin is the 396-residue form. Parr exists as two polypeptide variants: the initially described 418-amino-acid “long” form (ParrL) and a 410-amino-acid “short” form (ParrS) that lacks residues 334-341, which correspond to the analogous residues deleted in the 396-residue form of visual arrestin (Parruti et al., 1993b; SterneMarr et al., 1993). In general, ParrL is the predominant form in the brain, that is, cortex, cerebellum, striatum, pineal body, retina, and heart, but not in the pituitary gland; ParrS is the major form in peripheral tissues, that is, spleen, kidney, lung, liver, and pituitary, but not in the heart (Sterne-Marr et al., 1993). The majority of the arr3 is expressed as a 409-amino-acid protein, but some tissues, including brain, pituitary gland, pineal body, spleen, and lung, express a small amount of a 420-residue polypeptide that contains an insert of 11amino acids following residue 361 (Sterne-Marr et al., 1993).Unlike the three other
G PROTEIN-COUPLED RECEPTOR REGULATION
217
arrestin homologues, both forms of arr3 lack the eight amino acids which are present in arr4 and in the long forms of arr and parr. AND SUBCELLULAR LOCALIZATION C. TISSUE,CELLULAR,
By Northern analysis both arr (Lohse et al., 1990) and arr4 (Murakami et al., 1993) are expressed most abundantly in the retina and the pineal body (summarized in Table 111). However, the sensitivity of PCR analysis has enabled the detection of much lower levels of arr and arr4 in various other tissues. Thus, in addition to the expression of arr in the brain and in leukocytes, arr and p44 have been detected in heart, kidney, lung, and skeletal muscle when large amounts of poly(A)+ RNA are used for reverse transcription preceding PCR (Smith et al., 1994). Arr4 is found in the pituitary gland and the cerebral cortex when purified mRNA is used as the original template for PCR (Craft et al., 1994). From the combination of Northern, PCR, and immunoblotting analyses, parr and arr3 appear to be fairly ubiquitously expressed, with highest levels of expression in the brain, spleen, and prostate (Parruti et al., 1993b; Sterne-Marr et al., 1993).When the two polypeptides are compared directly in tissues using a monoclonal antibody that recognizes an epitope found in all arrestins, parr appears to be expressed at significantly higher levels than arr3 (Sterne-Marr et al., 1993). However, in olfactory epithelium, arr3 is the predominant arrestin isoform (Dawson et al., 1993). Further, arr3 is expressed at higher levels than parr in several cell lines (R. Sterne-Marr, unpublished observations). Immunocytochemistry, immunoelectron microscopy, and in situ hybridization have been used to study the cellular and subcellular distributions of members of the arrestin family. In dark-adapted retinas antibodies specific for the 404-residue visual arrestin label the inner and outer segments of the rod cells, somata, and synaptic layer (Smith et al., 1994). Upon light activation the signal from the inner segments, somata, and synapses decreases, while the outer segment signal is intensified. Interestingly, p44 labeling is found exclusively in the outer segments (consistent with its membrane localization) in dark-adapted retinas, and the C-terminal epitope is apparently masked upon light activation (Smith et al., 1994).Two groups have used in situ hybridization with nucleic acid probes to analyze the cellular localization of arr4 in the retina. Unfortunately, disparate observations resulted. Craft et al. (1994) detected specific labeling of cone cells, while Murakami et al. (1993) found a much broader distribution of arr4 in the rod inner and outer segments as well as the inner plexiform layer. Since the exact
TABLE I11 MOLECULAR PROPERTIES OF THE ARRESTINS Parameter
Pam
arr ~~
arr3
arr4
~~
Polypeptide variants (amino acids) Tissue distribution
404, 396, 370
418.410
420. 409
388
Retina > pineal body
Ubiquitous, brain, hematopoietic, prostate
Retina (Cone?)
mRNA size (kb) Chromosomal localization" Receptor affinity (K,) (nM) Rhodopsinb PzARC m2 mAChR. Uncoupling efficacy Receptor binding preference
1.5 2q37
7.5, 4.1, 1.3 llq13
Ubiquitous, brain, hematopoietic, prostate, olfactory cilia 1.7-2.4 17~13
30-50 2.1 7.2 Rhodopsin P pzAR Rhodopsin % P,AR m2 mAChR
0.14 0.48 PzAR rhodopsin m2 mAChR > pzAR rhodopsin
-
3
*
3
-
0.33 0.35 PzAR rhodopsin PzAR m2 mAChR rhodopsin
-
1.35 Xcen-q2l
?
2
? ? -
aSee the work of Calabrese et al. (1994a,b) and Murakami et al. (1993). bDetermined by stabilization of metarhodopsin I1 (Schleicher et al., 1989). CDetermined by Scatchard analysis using radiolabeled arrestins (Gurevich et al., 1993b, 1995).
G PROTEIN-COUPLED RECEPTOR REGULATION
219
sequence of the probe used in the former case is not clear, it is not possible to rationalize this disparity at this time. Immunocytochemistry was also used to compare parr and arr3 localization in the rat brain (Attramadal et al., 1992).Although the relative level of reactivity of each antibody is distinct, parr and arr3 antibodies label the cortex, olfactory bulb, hippocampus, cerebellum, and other neuronal structures. The neuronal labeling pattern roughly correlates with the reactivity levels detected by immunoblotting of lysates derived from various brain structures. Arr3 antibodies light up pyramidal cells of the cortex, while parr staining in the cortex is more diffuse. At the subcellular level both antibodies react with neuronal structures, most notably postsynaptic specializations as well as nonsynaptic plasma membrane, Golgi, and multivesicular bodies. Thus, while parr and arr3 are present in the soluble fraction following cell lysis, ultrastructural analyses show an association with membrane structures.
D. MECHANISM OF RETINAL ARRESTIN BINDING TO RHODOPSIN Studies to date addressing the mechanism of action of members of the arrestin family have focused primarily on the interaction of retinal arr with rhodopsin using a variety of techniques, including spectral studies, uncoupling assays, and direct binding assays (Wilden et al., 1986; Schleicher et al., 1989; Palczewski et al., 1991a,c; Gurevich and Benovic, 1992). Arr is a very abundant soluble rod outer segment protein that was discovered by its ability to bind to light-activated (Kuhn, 1978) and phosphorylated (Kuhn et al., 1984) disk membranes, inactivating the visual transduction cascade by effectively competing with transducin for the activated receptor (Wilden et al., 1986) and preventing receptor dephosphorylation (Palczewski et al., 1989b). When rhodopsin absorbs a photon, 11-&-retinal rapidly isomerizes, yielding an equilibrium between the spectrally distinguishable MI and MII. Interaction of transducin with MI1 not only initiates the visual cascade, but drives the equilibrium toward the MI1 tautomer (sometimes called the formation of “extra-MII”)(Stryer, 1986).Upon light activation arr also enhances the formation of MII, and this stabilization requires prephosphorylation of rhodopsin. Kinetic analysis indicates that the binding of arr to rhodopsin proceeds with a large Arrhenius activation energy, suggesting that arr undergoes a conformational change upon binding phosphorylated and light-activated rhodopsin (Schleicher et al., 1989). This conformational change is also detected as an increase in the sensitivity of arr to limited proteolysis (Palczewski et al., 1991a). The putative conformational change does not disrupt the secondary
220
RACHEL STERNE-MARR AND JEFFREY L. BENOVIC
structure of arr (as measured by circular dichroism), suggesting that rhodopsin-arr interactions replace intramolecular arrestin interactions (Palczewski et al., 1991~). Various truncated forms of arr have been used to investigate the binding-induced conformational change in arr and the interactions of arr with the light-activated and phosphorylated form of rhodopsin, leading to the assignment of functional domains of retinal arrestin. These studies take advantage of the ability to prepare the various functional forms of rhodopsin in uitro. Rod outer segment membranes purified from dark-adapted bovine retinas and stripped of arrestin and RK by urea treatment contain predominantly rhodopsin (>go% of the total protein). In the presence of light, rhodopsin can be phosphorylated to varying extents using purified RK or PARK (both phosphorylate the same C-terminal tail residues). After washing to remove the kinase, rhodopsin is again dark-adapted and regenerated with unbleached chromophore to produce “dark phosphorylated rhodopsin” (Rho-P). Upon light activation rhodopsin and Rho-P assume their active conformations, light-activated rhodopsin (Rho*) and light-activated phosphorylated (Rho*-P)rhodopsin, respectively. Full-length visual arrestin binds to Rho*-P with high affinity (-30-50 nM) and selectivity, since binding to Rho*-P is 10- to 12-fold higher than binding to Rho* or Rho-P (Bennett and Sitaramayya, 1988; Schleigher et al., 1989; Gurevich and Benovic, 1992). It should be emphasized that wild-type arrestin can clearly detect Rho* and Rho-P, since binding to these forms is greater than binding to Rho. Thus, selectivity refers to the ability of arr to preferentially bind to the light-activated and phosphorylated form of rhodopsin. Since the C-terminal 20% of each arrestin represents the most variable region among family members and contains some homology to G protein a-subunits, it has been suggested that this portion of the molecule might determine the specificity of binding to different receptors and therefore be directly involved in receptor binding. While arr is 100-fold (Lohse et al., 1992) more effective than parr in uncoupling GT-Rho*-P interaction, substitution of the last 59 residues of visual arrestin with the corresponding 78 amino acids of parr has little affect on its ability to bind Rho*-P (Gurevich et al., 1995). While a role in binding to the receptor cannot be ruled out, truncation mutagenesis studies suggest that the C terminus of arrestin plays a regulatory role, perhaps by interacting with other portions of arrestin, such as the N-terminal domain. Proteolytic fragments (residues 3-354 and other similar products) of arr not only bind Rho*-P, but do so with higher affinity than does the full-length arrestin (Palczewski et al., 1991a).In
G PROTEIN-COUPLED RECEPTOR REGULATION
221
vitro translation of functionally active arr has allowed a rapid systematic study of this phenomenon using various-length arr molecules. A form of arr that lacks the last 13 residues retains most of its binding to Rho*-P while also maintaining its selectivity. In contrast, a molecule that contains the first 365 (A366-404) residues can bind Rho*-P fairly well, but has lower selectivity: it binds Rho* and Rho-P to a greater extent than does full-length arrestin (Gurevich and Benovic, 1992). Therefore, the C terminus apparently maintains the rigidity of arr, preventing it from making strong contacts with rhodopsin unless the receptor is in the light-activated and phosphorylated form. Interestingly, the A366-404 mutant is similar in structure to the naturally occurring p44, which contains the first 369 residues and terminates with an alanine. Analogous to the 365-deletion mutant, p44 binds to Rho*-P, Rho*, and Rho-P equally well. However, using an uncoupling assay (an in uitro system measuring the ability of an arr to block transducin-mediated phosphodiesterase activity), p44 is more than 30-fold more potent than arr in competing with transducin for Rho*, while p44 is unable to compete with transducin for Rho*-P (Palczewski et al., 1994). Since p44 binds more avidly to the membrane than arr, this suggests that p44 may quench the transducin-mediated signal independently of phosphorylation. Further, since p44 binds Rho*-P but is not effective in inactivating the visual transduction cascade, this experiment underscores the importance of using a functional assay to measure arrestin activities. The product of the Drosophila arrl gene is quite similar in length to p44, yet mutations in this gene do not significantly alter visual transduction or desensitization (Dolph et al., 1992). Indirect evidence suggests that the regulatory role of the C terminus of arr is mediated by its interaction with the N terminus. Retinal arrestin is sensitive to certain polyanions, such as heparin and dextran sulfate, but not polyglutamic or polyaspartic acids (Palczewski et al., 1991). Not only does heparin prevent binding of arrestin to Rh*-P, as measured by direct binding (Gurevich et al., 1994) and extra MI1 stabilization studies (Palczewski et al., 1991), but it also mimics the phosphorylated receptor by inducing the conformational change in arr (Palczewski et al., 1991). Furthermore, heparin can also mimic the arr C terminus, since low levels confer selectivity to the A366-404 arr that otherwise lacks selectivity (Gurevich et al., 1994). The heparin binding site in arr maps in the amino half of the molecule, since a proteolytic fragment of arrestin containing residues 3-205 binds tightly to heparin-Sepharose (Palczewski et al., 1991) and a truncation mutant that terminates at residue 191 is exquisitely sensitive to
222
RACHEL STERNE-MARR AND JEFFREY L. BENOVIC
heparin (IC50,<1 pg/ml) (Gurevich et al., 1994). Since heparin binding domains consist of highly basic regions in other proteins, it is interesting that bovine retinal arr contains two such regions in this part of the molecule: residues 1-29 (B-I) and residues 162-181 (B-11).The N-terminal half (containing basic regions B-I and B-11) of parr (residues 1219) is 6-fold less sensitive to heparin than is the visual arr N terminus. However, a chimeric truncation mutant containing residues 1-47 of arrestin and 44-221 of parr has visual arrestin-like heparin sensitivity, while the sensitivity of the reverse chimera (residues 1-44 of parr and 48-187 of retinal arr) is more like that of parr (Gurevich et al., 1994). Therefore, residues 1-44 of visual arr appear to serve as the primary sites of heparin binding. Since one mechanism of heparin action is to mimic the C terminus of arrestin, it is plausible that the N and C termini interact to impart arr selectivity. Truncation mutants that terminate at residue 185 (A186-404) or 191 (A192-404) bind well to Rho-P and Rho*-P, suggesting that the N-terminal half of visual arrestin has the capability of recognizing phosphorylated rhodopsin (Gurevich and Benovic, 1993). While 8186404 binds Rho*-P and Rho-P fairly well, a mutant deleting 19 additional amino acids (Al68-404) has significantly lower ability to interact with phosphorylated forms of rhodopsin. Thus, the basic region B-I1 mentioned above is a likely candidate domain to interact with the acidic phosphorylated receptor. While the ability to recognize the phosphorylated receptor can be assigned to a discrete portion of arr, residues throughout the molecule seem to be involved in the discrimination of the activated form. It is unlikely that the C terminus plays a role in activation recognition, since p44 (residues 1-369 of arr) binds Rho* with higher affinity than does full-length arr (Palczewski et al., 1994). It is also clear that ionic and/or hydrogen bonding is involved in activation recognition as well as phosphorylation recognition, since salt (potassium acetate) inhibits binding of arrestin to Rho* and Rho-P, with IC,,s of 150 and 350 mM, respectively (Gurevich and Benovic, 1993). Interestingly, the binding of wild-type and A366-404 to Rho*-P, but not the Rho* or Rho-P forms, is actually stimulated by 200 mM salt, suggesting the involvement of hydrophobic interactions. Because binding to Rho*-P by the Al92-404 mutant is inhibited at 200 mM salt, the domain involved in hydrophobic interaction apparently resides between residues 192 and 365 (Gurevich and Benovic, 1993). Thus, the hydrophobic interaction requires the activation recognition, phosphorylation recognition, and hydrophobic interaction domains of arr to engage the activated and phosphorylated form of rhodopsin. Based on
G PROTEIN-COUPLED RECEPTOR REGULATION
223
these results and the observation that arrestin undergoes a conformational change upon binding to Rho*-P but not Rho* (Schleicher et al., 19891, it has been postulated that simultaneous occupation of the phosphorylation and activation domains allows arr to undergo a conformational rearrangement that exposes the hydrophobic interaction domain (Gurevich and Benovic, 1993). The interaction of these three domains of arr with light-activated and phosphorylated rhodopsin could explain the strong selectivity of retinal arr for Rho*-P. In this model the lack of selectivity of A366-404 arrestin is accounted for by its failure to maintain a rigid structure due to the lack of N- and C-terminal interaction, leading to the facile (light- or phosphorylationindependent) mobilization of the hydrophobic interaction domain. Similar to the interaction of receptors with G proteins and GRKs, arr interacts with multiple sites on rhodopsin (Krupnick et al., 1994). Synthetic peptides representing the first and third cytoplasmic loops of rhodopsin inhibit arr interaction, with IC,,s of 1000 and 34 FM, respectively. The region of arr that apparently interacts with the rhodopsin loops predominantly resides between residues 16 and 191. While the first and third cytoplasmic loop peptides of rhodopsin inhibit binding of arr to Rho*-P, these peptides actually stimulate binding of arr to Rho-P. These results suggest that the binding of arr with Rho*-P involves at least three domains of rhodopsin: the phosphorylated C-terminal tail as well as the first and third cytoplasmic loops.
E. MECHANISM OF NONVISUAL ARRESTINBINDING TO RECEPTORS To what extent does the model of retinal arr interaction with rhodopsin predict how the other arrestins interact with receptors? Similar t o arr, the polypeptide variants of parr and arr3 can be synthesized in a cell-free system as functionally active forms allowing the binding to rhodopsin, P2AR, and m2 mAChR to be studied (Sterne-Marr et al., 1993; Gurevich et al., 1994, 1995). As expected by the high homology between arrestins, the N-terminal halves of parr and arr3 retain the ability to recognize phosphorylated m2 mAChR (but not p2AR very well). Nonvisual arrestins bind to p2AR and m2 mAChR with very high affinity (0.06-0.6 nM), while visual arrestin binds to p2AR and m2 AChR with significantly lower affinity (2-7 nM)(Gurevich et al., 1995). parr and arr3 bind to p2AR and m2 mAChR with much lower selectivity than arr binds to rhodopsin, which can be attributed to (1)lower binding to the agonist-activated and phosphorylated form, R*-P, and (2) much higher binding to the antagonist-treated phosphorylated re-
224
RACHEL STERNE-MARR AND JEFFREY L. BENOVIC
ceptor, R-P (Gurevich et al., 1995). This results in only a 1.5-to 2-fold discrimination of the R-P versus R*-P forms. Unlike visual arr, these arrestins do not discriminate R and R* very well. While the C-terminal 51 and 34 residues of ParrL and arr3S, respectively, can play a regulatory role, the C-terminal truncations parrLA368-418 and arr3SA376409 show an -2-fold increase in the recognition of R-P with very little change in the ability to bind R*. While nonvisual arrestin binding to phosphorylated p,AR and m2 mAChR is resistant to 400 mil4 potassium acetate, the salt resistance does not require agonist activation. Thus, for parr and arr3 binding to the &AR and m2 mAChR, phosphorylation recognition seems more important than activation recognition (Gurevich et al., 1995). If hydrophobic interactions are involved in nonvisual arr interaction with P2AR and m2 mAChR, in contrast to arr-rhodopsin interaction, they can occur in the absence of agonist activation. Furthermore, since the N-terminal206 and 183 residues of parr and arr3, respectively, bind to P,AR and m2 mAChR at 400 mil4 salt, some hydrophobic interaction domains of these proteins appear to reside in the N-terminal half rather than solely in the C-terminal half, as in retinal arr. In summary, the mechanism of nonvisual arrestin binding to receptors appears t o be a variation of the theme described for retinal arrestin binding to rhodopsin. F. RECEPTOR SPECIFICITY The specificity of arr for rhodopsin has always been inferred from its specific localization and abundance in the retina (see Table 111). This assumption has been substantiated by the 100-fold preference of retinal arr over parr in the uncoupling of GT-rhodopsin interaction (Lohse et al., 1992). The recent identification of membrane-associated p44, which efficiently uncouples GT from Rh* (but not Rh*-P) forces the reevaluation of the in uiuo desensitization of rhodopsin (Palczewski et al., 1994). parr and arr3 are equally effective at uncoupling p2AR-Gs interaction, and both of these arrestins are -100-fold more potent than retinal arr (Attramadal et al., 1992). Binding studies demonstrate that while visual arrestin is highly specific for rhodopsin (it binds 10-fold more rhodopsin than PzAR or m2 mAChR), neither parrL, parrS, arr3S, nor arr3L demonstrates great specificity for rhodopsin, P2AR, or m2 mAChR (Gurevich et al., 1995). In addition, unlike visual arrestin and rhodopsin, neither parr nor arr3 colocalizes with P2AR at the cellular level. Therefore, it is not clear which of these arrestins plays a role in the desensitization of p,AR or other receptors in uiuo.
G PROTEIN-COUPLED RECEPTOR REGULATION
225
One exception is in olfactory epithelium, where arr3, but not parr, is present. Analogous t o the inhibition of olfactory receptor desensitization by PARK2 antibodies, addition of arr3-specific polyclonal antibodies to permeabilized olfactory cilia blocks desensitization of odorant receptors (Dawson et al., 1993). Because visual arr and parr display differential affinity and selectivity when binding to the various receptors, it is possible to use chimeric arrestins to investigate which portion of these molecules confers specificity. Consistent with their roles in maintaining selectivity, swapping N- or C-terminal domains of arr and parr does not dramatically alter specificity. Instead, the domains that determine specificity and are likely to be involved in direct binding to receptors appear to reside in the central residues 48-365 of arr and 45-367 of parr (Gurevich et al., 1995). One curious feature of studies with truncated and chimeric arrestins is that their behavior in binding to rhodopsin, p,AR and m2 mAChR appears to be more strongly influenced by the receptor than by the arrestin. For example, all arrestins bind to rhodopsin fairly well. Second, the major effect of truncation of -200 residues of parr or arr3 is increased binding to m2 mAChR-P but decreased binding to p2AR*-P and P2AR-P. Finally, interchange of parr domains into arr does not diminish arr binding to rhodopsin or arr selectivity, while substitution of arr domains into parr has opposing effects on binding to P2AR and m2 mAChR. However, parr and arr3 do demonstrate strong specificity in uncoupling p,AR/Gs versus rhodopsin/G, with IC5,s significantly higher (2-5 nM) than the Kd for binding (Attramadal et al., 1992). Therefore, the arrestin binding characteristics alone may not be a good measure of specificity (i.e., additional interactions may determine the uncoupling specificity). Arrestin binding t o different receptors may be determined by structural and perhaps yet to be described functional characteristics of the receptors themselves. The sizes of the intracellular third loop and the C-terminal tail are features that distinguish GPRs (Watson and Arkinstall, 1994). If these two intracellular domains are categorized as being small, medium, or large, then there are nine potential classes of receptors. Rhodopsin, which appears to be a “permissive” receptor with respect to arrestin binding, is representative of the largest class of receptors and has a short third intracellular loop and a medium-length tail. Other members of this class include the color opsins and the corticotropin, adenosine, angiotensin, bombesin, bradykinin, C5a, calcitonin, cannabinoid, endothelin, formyl peptide, follicle-stimulating hormone, glucagon, glutamate (R,, R,, and R4), histamine (H2),inter-
226
RACHEL STERNE-MARR AND JEFFREY L. BENOVIC
leukin 8, luteinizing hormone/human chorionic gonadotropin, melanocortin, neurokinin, neuropeptide Y 1, odorant, PAF, parathyroid, prostanoid (EP3 and thromboxane), secretin, somatostatin, tachykinin, thrombin, thyroid-stimulating hormone, and vasoactive intestinal polypeptide receptors. &AR is representative of a second class of receptors that have medium-sized third intracellular loops and mediumlength C-tails. Other members of this class include the pl- and &adrenergic, cholecystokinin, dopamine (D1 and D5), and thyrotropin receptors. A third class includes the muscarinic acetylcholine (ml, m2, (YZB, and aZc),dopamine (D2, D3, and and m4), the a,-adrenergic (aZA, D4), histamine (HI), serotonin (5HT-lA, 5HT-lB, 5HT-lD, 5HT-lE, 5HT-2A, 5HT-2B, and 5HT-20, and vasopressin receptors and are characterized by a large third intracellular loop and a short C-terminal tail, A fourth group contains the al-adrenergic receptors (alB,ale, and aID), which have a long third intracellular loop and a very long C-terminal tail. It will be interesting to determine whether the binding and uncoupling activities of the arrestins follow receptor class lines or whether some other mechanism of discrimination will emerge. IV. CONCLUSIONS Six mammalian GRKs and four arrestins have been identified by molecular cloning. Two major questions persist. (1) Does receptor specificity exist in uiuo? Because GPRs can activate multiple classes of G proteins (see Conklin and Bourne, 19931, it is also possible that these receptors may be regulated by more than one isoform of GRK or arrestin. (2) If receptor specificity does exist, what “rules” determine specificity? Four of these 10 proteins are expressed predominantly in a single tissue, that is, retina (RK, arr, and arr4) or testis (GRK4).While the role of RK and arr in regulating rhodopsin has been established by in uitro studies, what is the in uiuo role of the p44 form of arr? Our understanding of the regulation of cone opsins is less complete. Further, which GRKs and arrestins regulate other retinal receptors? The other GRKs and arrestins are fairly ubiquitous, although their levels of expression are undoubtedly quite variable. To date, in uitro assays have not demonstrated remarkable specificity of these ubiquitous GRKs and arrestins for receptors such as P,AR and m2 mAChR. However, the GRKs appear to show more distinguishable preferences when the receptors are overexpressed in membranes rather than purified and reconstituted in lipid vesicles. Perhaps binding of arrestins to such membranes may also show preferences that are not obvious from
G PROTEIN-COUPLED RECEPTOR REGULATION
227
experiments with purified preparations. If not, it is clear that binding and uncoupling assays must be used in parallel to analyze arrestin specificity. More importantly, in uiuo assays must also be developed. Most in uiuo assays used to date utilize overexpression of GPRs with GRKs or arrestins. This type of experiment yields information suggesting that a particular GRK or arrestin can regulate a GPR but does not prove that it does so in uiuo in the native environment. The use of dominant negative mutants is a n improvement; however, if overexpression is required in order to see an effect, the concern exists that the mutant is acting promiscuously. A more difficult but preferable approach would be to eliminate a GRK or arrestin by antisense oligonucleotides or by generation of transgenic animals and then assess the pharmacological effect. Indeed, it may be that even if receptor specificity exists, homologous desensitization of a given receptor in a particular cell may play a supporting role relative to other mechanisms of regulation such as heterologous desensitization and sequestration. Conversely, regulation by GRKs and arrestins may be the predominant mechanism of modulation of certain receptors in specific milieus. ACKNOWLEDGMENTS This work was supported in part by Grants GM-44944 and GM-47417from the National Institutes of Health (to J.L.B.) and by a grant from the American Heart Association, Southeastern Pennsylvania Affiliate (to R.S.-M.).J.L.B. is an established investigator of the American Heart Association. REFERENCES Ambrose, C., James, M., Barnes, G., Lin, C., Bates, G., Altherr, M., Duyao, M., Groot, N., Church, D., Wasmuth, J . J., Lehrach, H., Housman, D., Buckler, A., Gusells, J . F., and MacDonald, M. E. (1992). A novel G protein-coupled receptor kinase gene cloned Mol. Genet. 1, 697-703. from 4 ~ 1 6 . 3Hum. . Anant, J. S., and Fung, B. K. (1992). In vivo farnesylation of rat rhodopsin kinase. Biochem. Biophys. Res. Commun. 183,468-473. Arriza, J . L., Dawson, T. M., Simerly, R. B., Martin, L. J., Caron, M. G., Snyder, S. H., and Lefkowitz, R. J. (1992). The G-protein-coupled receptor kinases PARK1 and PARK2 are widely distributed a t synapses in rat brain. J. Neurosci. 12, 4045-4055. Attramadal, H., Arriza, J. L., Dawson, T. M.,. Codina, J., Kwatra, M. M., Snyder, S. H., Caron, M. G., and Lefkowitz, R. J. (1992). p-arrestin2, a novel member of the arrestin/p-arrestin gene family. J . Biol. Chem. 267, 17882-17890. Bennett, N., and Sitaramayya, A. (1988). Inactivation of photoexcited rhodopsin in retinal rods: The roles of rhodopsin kinase and 48-kDa protein (arrestin). Biochemistry 27, 1710-1715. Benovic, J . L., and Gomez, J. (1993). Molecular cloning and expression of GRK6: A new member of the G protein-coupled receptor kinase family. J . Biol. Chem. 268,1952119527. Benovic, J . L., Strasser, R. H., Caron, M. G., and Leflcowitz, R. J . (1986). P-Adrenergic
228
RACHEL STERNE-MARR AND JEFFREY L. BENOVIC
receptor kinase: Identification of a novel protein kinase that phosphorylates the agonist-occupied form of the receptor. Proc. Natl. Acad. Sci. U.S.A. 83, 2797-2801. Benovic, J. L., Kuhn, H., Weyand, I., Codina, J., Caron, M. G., and Lefkowitz, R. J . (1987a). Functional desensitization of the isolated P-adrenergic receptor by the P-adrenergic receptor kinase: Potential role of an analog of the retinal protein arrestin (48 kDa protein). Proc. Natl. Acad. Sci. U.S.A. 84, 8879-8882. Benovic, J . L., Mayor, F., Jr., Staniszewski, C., Lefkowitz, R. L., and Caron, M. G. (1987b). Purification and characterization of the P-adrenergic receptor kinase. J . Biol. Chem. 262,9026-9032. Benovic, J. L., Regan, J. W., Matsui, H., Mayor, F., Jr., Cotecchia, S., Leeb-Lundberg, Agonist-dependent phosphorylaL. M. F., Caron, M. G., and Lefkowitz, R. J . (1987~). tion of the a,-adrenergic receptor by the p-adrenergic receptor kinase. J . Biol. Chem. 262, 17251-17253. Benovic, J . L., De Blasi, A., Stone, W. C., Caron, M. G., and Lefkowitz, R. J . (1989). P-Adrenergic receptor kinase: Primary structure delineates a multigene family. Science 246, 235-240. Benovic, J . L., Onorato, J., Lohse, M. J., Dohlman, H. G., Staniszewski, C., Caron, M. G., and Lefkowitz, R. J. (1990). Synthetic peptides of the hamster P,-adrenoceptor as substrates and inhibitors of the P-adrenoceptor kinase. Br. J . Clin. Pharmacol. 30, 3s-12s. Benovic, J . L., Onorato, J. J., Arriza, J . L., Stone, W. C., Lohse, M., Jenkins, N. A,, Gilbert, D. J., Copeland, N. G., Caron, M. G., and Lefkowitz, R. J. (1991a). Cloning, expression, and chromosomal localization of P-adrenergic receptor kinase 2: A new member of the receptor kinase family. J. Biol. Chem. 266, 14939-14946. Benovic, J . L., Stone, W. C., Huebner, K., Croce, C., Caron, M. G., and Lefkowitz, R. J . (1991b). cDNA cloning and chromosomal localization of the human P-adrenergic receptor kinase. FEBS Lett. 283, 122-126. Birnbaumer, L., Abramowitz, J., and Brown, A. M. (1990). Receptor-effector coupling by G proteins. Biochim. Biophys. Acta 1031, 163-224. Boekhoff, I., Inglese, J., Schleicher, S., Koch, W. J., Lefkowitz, R. J., and Breer, H. (1994). Olfactory desensitization requires membrane targeting of receptor kinase mediated by py-subunits of heterotrimeric G proteins. J. Biol. Chem. 269, 37-40. Bouvier, M., Hausdorff, W. P., De Blasi, A., O'Dowd, B. F., Kobilka, B. K., Caron, M. G., and Lefkowitz, R. J. (1988).Removal of phosphorylation sites from the P,-adrenergic receptor delays onset of agonist-promoted desensitization. Nature 333, 370-373. Bownds, D., Dawes, J., Miller, J., and Stahlman, M. (1972). Phosphorylation of frog photoreceptor membranes induced by light. Nature 237, 125-127. Brass, L. F. (1992). Homologous desensitization of HEL cell thrombin receptors. J . Biol. Chem. 267,6044-6050. Buczylko, J., Gutmann, C., and Palczewski, K. (1991). Regulation of rhodopsin kinase by autophosphorylation. Proc. Natl. Acad. Sci. U.S.A. 88, 2568-2572. Bullrich, F., Druck, T., Kunapuli, P., Gomez, J., Gripp, K. W., Schlegelberger, B., Lasota, J., Aronson, M., Cannizzaro, L. A., Huebner, K., and Benovic, J. L. (1995). Chromosomal mapping of the genes encoding G protein-coupled receptor kinases GRK5 and GRK6. Cytogenet. Cell Genet. in press. Calabrese, G., Sallese, M., Stornaiuolo, A., Morizio, E., Palka, G., and De Blasi, A. (1994a). Assignment of the p-arrestin 1 gene (ARRB1) to human chromosome llq13. Genomics 23, 168-171. Calabrese, G., Sallese, M., Stornaiuolo, A., Stuppia, L., Palka, G., and De Blasi, A. (199413). Chromosome mapping of the human arrestin (SAG), p-arrestin 2 (ARRBS), and p-adrenergic receptor kinase 2 (ADRBK2) genes. Genomics 23, 286-288.
G PROTEIN-COUPLED RECEPTOR REGULATION
229
Cassill, J . A., Whitney, M., Joazeiro, C. A. P., Becker, A., and Zuker, C. S. (1991). Isolation of Drosophila genes encoding G protein-coupled receptor kinases. Proc. Natl. Acad. Sci. U.S.A. 88, 11067-11070. Chabre, M., and Deterre, P. (1989). Molecular mechanism of visual transduction. Eur. J . Biochem. 179, 255-266. Chen, C.-Y., Dion, S. D., Kim, C. M., and Benovic, J. L. (1993). P-Adrenergic receptor kinase: Agonist-dependent receptor binding promotes kinase activation. J . Biol. Chem. 268,7825-7831. Chuang, T. T., Sallese, M., Ambrosini, G., Parruti, G., and De Blasi, A. (1992). High expression of P-adrenergic receptor kinase in human peripheral blood leukocytes: Isoproterenol and platelet activating factor can induce kinase translocation. J . Biol. Chem. 267,6886-6892. Conklin, B. R., and Bourne, H. R. (1993). Structural elements of G, subunits that interact with G,,, receptors, and effectors. Cell 73, 631-641. Craft, C. M., Whitmore, D. H., and Wiechmann, A. F. (1994). Cone arrestin identified by targeting expression of a functional family. J. Biol. Chem. 269, 4613-4619. Dawson, T. M., Arriza, J . L., Jaworsky, D. E., Borisy, F. F., Attramadal, H., Lefkowitz, R. J., and Ronnett, G. V. (1993). P-Adrenergic receptor kinase-2 and p-arrestin2 as mediators of odorant-induced desensitization. Science 259, 825-829. DebBurman, S. K., Kunapuli, P., Benovic, J. L., and Hosey, M. M. (1995a). Agonistdependent phosphorylation of human muscarinic receptors in insect Sf9 cell membranes by G protein-coupled receptor kinases. Mol. Pharmacol. 47, 224-233. DebBurman, S. K., Ptasienski, J., Boetticher, L., Lomasney, J. W., Benovic, J . L., and Hosey, M. M. (1995b). Lipid-mediated regulation of G protein-coupled receptor kinases. J.Biol. Chem. 270,5742-5141. Dohlman, H. G., Thorner, J., Caron, M. G., and Lefkowitz, R. J. (1991). Model systems for the study of seven-transmembrane-segment receptors. Annu. Reu. Biochem. 60, 653-688. Dolph, P. J., Ranganathan, R., Colley, N. J., Hardy, R. W., Socolich, M., and Zuker, C. S. (1992). Arrestin function in inactivation of G protein-coupled receptor rhodopsin in uiuo. Science 260, 1910-1916. Fowles, C., Sharma, R., and Akhtar, M. (1988). Mechanistic studies on the phosphorylation of photoexcited rhodopsin. FEBS Lett. 238, 56-60. Frank, R. N., Cavanaugh, H. D., and Kenyon, K. R. (1973).Light-stimulated phosphorylation of bovine visual pigments by adenosine triphosphate. J.Biol. Chem. 248,596-609. Garcia-Higuera, I., Petronila, P., Murga, C., Egea, G., Bonay, P., Benovic, J . L., and Mayor, F., Jr. (1994). Association of the regulatory p-adrenergic receptor kinase with rat liver microsomal membranes. J. Biol. Chem. 269, 1348-1355. Gibbs, J. B. (1991). Ras C-terminal processing enzymes-New drug targets? Cell 65,l-4. Gibson, T. J., Hyvonen, M., Musacchio, A., and Saraste, M. (1994).PH domain: The first anniversary. Trends Biochem. Sci. 19, 349-353. Gurevich, V. V., and Benovic, J . L. (1992). Cell-free expression of visual arrestin: Truncation mutagenesis identifies multiple domains involved in rhodopsin interaction. J . Biol. Chem. 267, 21919-21923. Gurevich, V. V., and Benovic, J . L. (1993). Visual arrestin interaction with rhodopsin: Sequential multisite binding ensures strict selectivity toward light-activated phosphorylated rhodopsin. J.Biol. Chem. 268, 11628-11638. Gurevich, V. V., Chen, C.-Y., Kim, C. M., and Benovic, J . L. (1994). Visual arrestin binding to rhodopsin: Intramolecular interaction between the basic N terminus and acidic C terminus of arrestin may regulate binding selectivity. J.Biol. Chem. 269, an-8727.
230
RACHEL STERNE-MARR AND JEFFREY L. BENOVIC
Gurevich, V. V., Dion, S. B., Onorato, J. J., Ptasienski, J., Kim, C. M., Sterne-Marr, R., Hosey, M. M., and Benovic, J. L. (1995).Arrestin interactions with G protein-coupled receptors: Direct binding studies of wild-type, truncated, and chimeric arrestins with rhodopsin, P,-adrenergic, and m2 muscarinic cholinergic receptors. J . Biol. Chem. 270,720-731. Haga, K., and Haga, T. (1990).Dual regulation by G proteins of agonist-dependent phosphorylation of muscarinic acetylcholine receptors. FEBS Lett. 268, 43-47. Haga, K., and Haga, T. (1992).Activation by G protein beta gamma subunits of agonistor light-dependent phosphorylation of muscarinic acetylcholine receptors and rhodopsin. J. Biol. Chem. 267, 2222-2227. Haga, K., Kameyama, K., and Haga, T. (1994).Synergistic activation of a G proteincoupled receptor kinase by G protein Py subunits and mastoparan or related peptides. J . Biol. Chem. 269, 12594-12599. Haga, T.,Haga, K., and Kameyama, K. (1994).G protein-coupled receptor kinases. J . Neurochem. 63,400-412. Hanks, S. K., and Quinn, A.M. (1991).Protein kinase catalytic domain sequence database: Identification of conserved features of primary structure and classification of family members. Methods Enzymol. 200, 38-81. Hanks, S. K., Quinn, A. M., and Hunter, T. (1988).The protein kinase family: Conserved features and deduced phylogeny of the catalytic domains. Science 241, 42-52. Hargrave, P. A., and McDowell, J . H. (1992).Rhodopsin and phototransduction: A model system for G protein-linked receptors. FASEB J . 6,2323-2331. Haribabu, B., and Snyderman, R. (1993).Identification of additional members of human G protein-coupled receptor kinase family. Proc. Natl. Acad. Sci. U.S.A. 90,93989402. Harlan, J . E., Hajduk, P. J., Yoon, H. S., and Fesik, S. W. (1994).Pleckstrin homology Nature 371, 168-170. domains bind to phosphatidylinositol-4,5-bisphosphate. Hausdorff, W. P., Caron, M. G., and Lefkowitz, R. J. (1990).Turning off the signal: Desensitization of P-adrenergic receptor function. FASEB J. 4, 2881-2889. Hoxie, J . A.,Ahuja, M., Belmonte, E., Pizarro, S., Parton, R., and Brass, L. F. (1993). Internalization and recycling of activated thrombin receptors. J. Biol. Chem. 268, 13756-13763. Inglese, J., Glickman, J . F., Lorenz, W., Caron, M. G., and Lefkowitz, R. J. (1992a). Isoprenylation of a protein kinase: Requirement of farnesylationia-carboxylmethylation for full enzymatic activity of rhodopsin kinase. J . Biol. Chem. 267, 14221425. Inglese, J., Koch, W. J., Caron, M. G., and Lefkowitz, R. J. (199213).Isoprenylation in regulation of signal transduction by G-protein-coupled receptor kinases. Nature 359,147-150. Inglese, J., Freedman, N. J., Koch, W. J., and Lefkowitz, R. J. (1993).Structure and mechanism of the G protein-coupled receptor kinases. J . Biol. Chem. 268,2373523138. Ishii, K., Chen, J., Ishii, M., Koch, W. J., Freedman, N. J., Lefkowitz, R. J., and Coughlin, S. R. (1994).Inhibition of thrombin receptor signaling by a G-protein coupled receptor kinase: Functional specificity among G-protein coupled receptor kinases. J . Biol. Chem. 269, 1125-1130. Kelleher, D. J., and Johnson, G. L. (1990).Characterization of rhodopsin kinase purified from bovine rod outer segments. J. Biol. Chem. 265, 2632-2639. Kim, C.M., Dion, S. B., and Benovic, J . L. (1993a).Mechanism of p-adrenergic receptor kinase activation by G proteins. J . Biol. Chem. 268, 15412-15418.
G PROTEIN-COUPLED RECEPTOR REGULATION
23 1
Kim, C. M., Dion, S. B., Onorato, J. J., and Benovic, J. L. (199313). Expression and characterization of two p-adrenergic receptor kinase isoforms using the baculovirus expression system. Receptor 3, 39-55. Koch, W. J., Inglese, J., Stone, W. C., and Lefkowitz, R. J. (1993). The binding site for the py subunits of heterotrimeric G proteins on the P-adrenergic receptor kinase. J. Biol. Chem. 268,8256-8260. Kong, G., Penn, R., and Benovic, J . L. (1994). A p-adrenergic receptor kinase dominant negative mutant attenuates desensitization of the p,-adrenergic receptor. J. Biol. Chem. 269, 13084-13087. Krupnick, J. G., Gurevich, V. V., Schepers, T., Hamm, H. E., and Benovic, J. L. (1994). Arrestin-rhodopsin interaction: Multi-site binding delineated by peptide inhibition. J. Biol. Chem. 269,3226-3232. Kuhn, H. (1978). Light-dependent binding of rhodopsin kinase and other proteins to cattle photoreceptor membranes. Biochemistry 17,4389-4395. Kuhn, H., and Dreyer, W. J. (1972). Light dependent phosphorylation of rhodopsin by ATP. FEBS Lett. 20, 1-6. Kuhn, H., Hall, S. W., and Wilden, U. (1984). Light-induced binding of 48-kDa protein to photoreceptor membranes is highly enhanced by phosphorylation of rhodopsin. FEBS Lett. 176,473-478. Kunapuli, P., and Benovic, J . L. (1993). Cloning and expression of GRKS: A member of the G protein-coupled receptor kinase family. Proc. Natl. Acad. Sci. U.S.A. 90, 5588-5592. Kunapuli, P., Gurevich, V. V., and Benovic, J. L. (1994a). Phospholipid-stimulated autophosphorylation activates the G protein-coupled receptor kinase GRK5. J. Biol. Chem. 269,10209-10212. Kunapuli, P., Onorato, J. J., Hosey, M. M., and Benovic, J . L. (199413).Expression, purification, and characterization of the G protein-coupled receptor kinase GRK5. J. Biol. Chem. 269, 1099-1105. Kwatra, M. M., Benovic, J. L., Caron, M. G., Lefkowitz, R. J., and Hosey, M. M. (1989). Phosphorylation of chick heart muscarinic cholinergic receptors by the p-adrenergic receptor kinase. Biochemistry 28, 4543-4547. Kwatra, M. M., Schwinn, D. A., Schreurs, J., Blank, J. L., Kim, C. M., Benovic, J. L., Krause, J. E., Caron, M. G., and Lefkowitz, R. J. (1993). The substance P receptor, which couples to G,,,,, is a substrate of P-adrenergic receptor kinases 1 and 2. J. Biol. Chem. 268,9161-9164. Lefkowitz, R. J. (1993). G protein-coupled receptor kinases. Cell 74, 409-412. Lohse, M. J., Benovic, J. L., Caron, M. G., and Lefkowitz, R. J. (1990). P-Arrestin: A protein that regulates P-adrenergic receptor function. Science 248, 1547-1550. Lohse, M. J., Andexinger, S., Pitcher, J., Trukawinski, S., Codina, J., Faure, J.-P., Caron, M. G., and Lefkowitz, R. J . (1992). Receptor-specific desensitization with purified proteins: Kinase dependence and receptor specificity of p-arrestin and arrestin in the p,-adrenergic receptor and rhodopsin systems. J. Biol. Chem. 267, 85588564. Lorenz, W., Inglese, J., Palczewski, K., Onorato, J. J., Caron, M. G., and Lefkowitz, R. J . (1991). The receptor kinase family: Primary structure of rhodopsin kinase reveals similarities to the p-adrenergic receptor kinase. Proc. Natl. Acad. Sci. U.S.A. 88, 8715-8719. Loudon, R. P., and Benovic, J. L. (1994). Expression, purification, and characterization of the G protein-coupled receptor kinase GRK6. J. Biol. Chem. 269, 22691-22697. Macias, M. J., Musacchio, A., Ponstingl, H., Nilges, M., Saraste, M., and Oschkinat, H.
232
RACHEL STERNE-MARR AND JEFFREY L. BENOVIC
(1994). Structure of the pleckstrin homology domain from beta-spectrin. Nature 269, 675-677. Mayor, F., Jr., Benovic, J . L., Caron, M. G., and Lefkowitz, R. J. (1987). Somatostatin induces translocation of the p-adrenergic receptor kinase and desensitizes somatostatin receptors in S49 lymphoma cells. J. Biol. Chem. 262, 6468-6471. Murakami, A., Yajima, T., Sakuma, H., McLaren, M., and Inana, G. (1993). X-Arrestin: A new retinal arrestin mapping to the X chromosome. FEBS Lett. 334, 203-209. Musacchio, A., Gibson, T., Rice, P., Thompson, J., and Saraste, M. (1993). The PH domain: A common piece in the structural patchwork of signalling proteins. Trends Biochem. Sci. 18,343-348. Onorato, J. J., Palczewski, K., Regan, J. W., Caron, M. G., Lefkowitz, R. J., and Benovic, J . L. (1991). Role of acidic amino acids in peptide substrates of the P-adrenergic receptor kinase and rhodopsin kinase. Biochemistry 30, 5118-5125. Palczewski, K., and Benovic, J. L. (1991). G-protein-coupled receptor kinases. Trends Biochem. Sci. 16, 387-391. Palczewski, K., McDowell, J. H., and Hargrave, P. A. (1988). Rhodopsin kinase: Substrate specificity and factors that influence activity. Biochemistry 27, 2306-2313. Palczewski, K., Arendt, A., McDowell, J . H., and Hargrave, R. A. (1989a). Substrate recognition determinants for rhodopsin kinase: Studies with synthetic peptides, polyanions, and polycations. Biochemistry 28, 8764-8770. Palczewski, K., McDowell, J . H., Jakes, S., Ingebritsen, T. S., and Hargrave, P. A. (1989b). Regulation of rhodopsin dephosphorylation by arrestin. J . Biol. Chem. 264, 15770-15773. Palczewski, K., Ruczylko, J., Imami, N. R., McDowell, J. H., and Hargrave, P. A. (1991a). Role of the carboxy-terminal region of arrestin in binding to phosphorylated rhodopsin. J . Biol. Chem. 266, 15334-15339. Palczewski, K., Buczylko, J., Kaplan, M. W., Polans, A. S., and Crabb, J . W. (1991b). Mechanism of rhodopsin kinase activation. J . Biol. Chem. 266, 12949-12955. PhosphoryPalczewski, K., Pulvermuller, A., Buczylko, J., and Hofmann, K. P. (1991~). lated rhodopsin and heparin induce similar conformational changes in arrestin. J . Biol. Chem. 266,18649-18654. Palczewski, K., Buczylko, J., Van Hooser, P., Carr, S. A., Huddleston, M. J., and Crabb, J. W. (1992). Identification of the autophosphorylation sites in rhodopsin kinase. J. Biol. Chem. 267,18991-18998. Palczewski, K., Buczylko, J., Lebioda, L., Crabb, J. W., and Polans, A. S. (1993). Identification of the N-terminal region in rhodopsin kinase involved in its interaction with rhodopsin. J . Biol. Chem. 268,6004-6013. Palczewski, K., Buczylko, J., Ohguro, H., Annan, R. S., Carr, S. A,, Crabb, J . W., Kaplan, M. W., Johnson, R. S., and Walsh, K. A. (1994). Characterization of a truncated form of arrestin isolated from bovine rod outer segments. Protein Sci. 3, 319-329. Parruti, G., Ambrosini, G., Sallese, M., and De Blasi, A. (1993a). Molecular cloning, functional expression, and mRNA analysis of human beta-adrenergic receptor kinase 2. Biochem. Biophys. Res. Commun. 190,475-481. Parruti, G., Peracchia, F., Sallese, M., Ambrosini, G., Masini, M., Rotilio, D., and De Blasi, A. (1993b). Molecular analysis of human p-arrestin-1: Cloning, tissue distribution, and regulation of expression. Identification of two isoforms generated by alternative splicing. J. Biol. Chem. 268, 9753-9761. Pei, G., Tiberi, M., Caron, M. G., and Lefkowitz, R. J. (1994). An approach to the study of G-protein-coupled receptor kinases: An in uitro-purified membrane assay reveals differential receptor specificity and regulation by G,., subunits. Proc. Nutl. Acud. Sci. U.S.A. 91, 3633-3636.
G PROTEIN-COUPLED RECEPTOR REGULATION
233
Penn, R. B., Kelsen, S. G., and Benovic, J. L. (1994). Regulation of P-agonist- and prostaglandin E,-mediated adenylyl cyclase activity in human airway epithelial cells. A m . J. Respir. Cell Mol. Biol. 11, 496-505. Pippig, S., Andexinger, S., Daniel, K., Puzicha, M., Caron, M. G., Lefkowitz, R. J., and Lohse, M. J. (1993). Overexpression of p-arrestin and P-adrenergic - receptor kinase augment desensitization of P,-adrenergic receptor. J. Biol. Chem. 268, 32013298. Pitcher, J., Lohse, M. J., Codina, J., Caron, M. G., and Lefkowitz, R. J. (1992a). Desensitization of the isolated P,-adrenergic receptor by P-adrenergic receptor kinase, CAMP-dependent protein kinase, and protein kinase C occurs via distinct molecular mechanisms. Biochemistry 31, 3193-3197. Pitcher, J. A,, Inglese, J.,Higgins, J. B., Arriza, J. L., Casey, P. J., Kim, C., Benovic, J. L., Kwatra, M. M., Caron, M. G., and Lefkowitz, R. J. (199213). Role of Py subunits of G proteins in targeting the P-adrenergic receptor kinase to membrane-bound receptors. Science 257, 1264-1267. Premont, R. T., Koch, W. J., Inglese, J., and Lefkowitz, R. J. (1994). Identification, purification, and characterization of GRK5, a member of the family of G protein-coupled receptor kinases. J. Biol. Chem. 269, 6832-6841. Rapoport, B., Kaufman, K. D., and Chazenbalk, G. D. (1992). Cloning of a member of the arrestin family from a human thyroid cDNA library. Mol. Cell. Endocrinol. 84, R39-R43. Richardson, R. M., Kim, C., Benovic, J. L., and Hosey, M. M. (1993). Phosphorylation and desensitization of human m2 muscarinic cholinergic receptors by two isoforms of the p-adrenergic receptor kinase. J. Biol. Chem. 268, 13650-13656. Sallese, M., Lombardi, M. S., and De Blasi, A. (1994). Two isoforms of G protein-coupled receptor kinase 4 identified by molecular cloning. Biochem. Biophys. Res. Commun. 199,848-854. Schleicher, A., Kuhn, H., and Hofmann, K. P. (1989). Kinetics, binding constant, and activation energy of the 48-kDa protein-rhodopsin complex by extra-metarhodopsin 11. Biochemistry 28, 1770-1775. Schleicher, S., Boekhoff, I., Arriza, J., Lefkowitz, R. J., and Breer, H. (19931. A p-adrenergic kinase-like enzyme is involved in olfactory signal termination. Proc. Natl. Acad. Sci. U.S.A.90, 1420-1424. Shinohara, T., Dietzschold, B., Craft, C. M., Wistow, G., Early, J. J., Donoso, L. A,, Horwitz, J., and Tao, R. (1987). Primary and secondary structure of bovine retinal S antigen (48-kDa protein). Proc. Natl. Acad. Sci. U.S.A. 84, 6975-6979. Simonds, W. F., Manji, H. K., and Garritsen, A. (1993). G proteins and PARK: A new twist for the coiled coil. Trends Biochem. Sci. 18, 315-317. Smith, W. C., Milam, A. H., Dugger, D., Arendt, A., Hargrave, P. A., and Palczewski, K. (1994). A splice variant of arrestin: Molecular cloning and localization in bovine retina. J . Biol. Chem. 269, 15407-15410. Somers, R. L., and Klein, D. C. (1984). Rhodopsin kinase activity in the mammalian pineal and other tissues. Science 226, 182-184. Sterne-Marr, R., Gurevich, V. V., Goldsmith, P., Bodine, R. C., Sanders, C., Donoso, L. A,, and Benovic, J. L. (1993). Polypeptide variants of p-arrestin and arrestin3. J . Biol. Chem. 268, 15640-15648. Stoffel, R. H., Randall, R. R., Premont, R. T., Lefkowitz, R. J., and Inglese, J. (1994). Palmitoylation of G protein-coupled receptor kinase, GRK6: Lipid modification diversity in the GRK family. J. Biol. Chem. 269, 27791-27794. Strasser, R. H., Benovic, J. L., Caron, M. C., and Lefkowitz, R. J. (1986). p-Agonist- and prostaglandin El-induced translocation of the P-adrenergic receptor kinase: Evi-
234
RACHEL STERNE-MARR AND JEFFREY L. BENOVIC
dence that the kinase may act on multiple adenylate cyclase-coupled receptors. Proc. Natl. Acad. Sci. U.S.A.83, 6362-6366. Stryer, L. (1986). Cyclic GMP cascade of vision. Annu. Rev. Neurosci. 9, 87-119. Touhara, K., Inglese, J.,Pitcher, J. A., Shaw, G., and Lefkowitz, R. J. (1994).Binding of G protein py-subunits to plackstrin homology domains. J . Biol. Chern. 269, 1021710220. Watson, S., and Arkinstall, S. (1994). “The G-Protein Linked Receptor Factsbook,” Academic Press, San Diego. Wilden, U., and Kuhn, H. (1982). Light-dependent phosphorylation of rhodopsin: Number of phosphorylation sites. Biochemistry 21, 3014-3022. Wilden, U., Hall, S. W., and Kuhn, H. (1986).Phosphodiesterase activation by photoexcited rhodopsin is quenched when rhodopsin is phosphorylated and binds the intrinsic 48-kDa protein of rod outer segments. Proc. Natl. Acad. Sci. U.S.A. 83, 11741178. Wilson, C. J., and Applebury, M. L. (1993).Arresting G-protein coupled receptor activity. C u r . Biol. 3, 683-686. Yamaki, K., Takahashi, Y., Sakuragi, S., and Matsubara, K. (1987).Molecular cloning of the S-antigen cDNA from bovine retina. Biochem. Biophys. Res. Commun. 142,904910. Yamaki, K., Tsuda, M., and Shinohara, T. (1988).The sequence of human retinal S-antigen reveal similarities with a-transducin. FEBS Lett. 234, 39-43. Yamaki, K., Tsuda, M., Kikuchi, T., Chen, K.-H., Huang, K.-P., and Shinohara, T. (1990). Structural organization of the human S-antigen gene. J . Biol. Chem. 265, 2075720762. Yoon, H. S., Hajduk, P. J., Petros, A. M., Olejniczak, E. T., Meadows, R. P., and Fesik, S. W. (1994). Solution structure of a pleckstrin-homology domain. Nature 369, 672675.
VITAMINS AND HORMONES, VOL. 51
Vasopressin and Oxytocin: Molecular Biology and Evolution of the Peptide Hormones and Their Receptors EVITA MOHR", WOLFGANG MEYERHOF?, AND DIETMAR RICHTER* *Instit& fur Zellbiochemie und Klinische Neurobiologie Universitat Hamburg 20246 Hamburg, Germany TDeutsches lnstitut fur Erm%hrungsforschung Universitat Potsdnm 14558 Potsdum-Rehbriicke, Germany I. Introduction 11. VP and OT Gene Expression and Regulation A. Transgenic Animals B. Putative Regulatory Elements in the VP and OT Gene Promoters 111. VP and OT mRNA in Dendrites and Axons A. Dendritic Transcripts B. Axonally Localized Transcripts IV. Somatic Recombination between the VP and OT Genes in Hypothalamic Neurons V. Evolution of the Vertebrate VPIOT Gene Family VI. Nonapeptide Receptors A. Molecular Cloning B. Receptor Structure C. Tissue Distribution of Neurohypophysial Hormone Receptors and Their mRNAs D. Receptor Phylogeny References
I. INTRODUCTION The nonapeptides vasopressin (VP) and oxytocin (OT), the major secretory products of magnocellular neurons of the hypothalamoneurohypophysial system, are synthesized as parts of larger precursor proteins in different subsets of hypothalamic cells. They consist of a hydrophobic signal peptide at their N-terminal ends and the hormone moiety followed by the sequence Gly-Lys-Arg, the signal for hormone modification and precursor processing, respectively. Their major part is contributed by the carrier proteins, the neurophysins (NPs). The VP precursor only contains an additional component, the glycopeptide (GP). The biologically active hormones are secreted from the nerve 235
Copyright 0 1995 by Academic Press, Inc. All rights of reproduction in any form reserved.
236
EVITA MOHR et a1
terminals in the posterior pituitary gland into the systemic circulation subsequent to axonal transport in secretory vesicles and precursor processing (Richter, 1987). In the periphery VP and OT exert a variety of functions, including homeostasis of salt and water metabolism, blood pressure regulation (VP), and smooth muscle contraction during birth and lactation (OT) (Gainer and Wray, 1994; Mohr and Richter, 1994). VP and OT are also synthesized in hypothalamic parvocellular neurons of the paraventricular nucleus (PVN). These neurons project to the external zone of the median eminence and release their contents into the portal blood system, which is connected to the anterior pituitary gland. In this tissue VP acts in concert with corticotropin-releasing factor on the release of adrenocorticotropic hormone (Mohr and Richter, 1994). Both hormones are also synthesized in peripheral organs such as the adrenal gland, ovaries, uterus, and testes, where they are thought to act as paracrine regulators of body functions (Ivell, 1986). Their role as neurotransmitters and/or neuromodulators in the central nervous system is less well defined. It appears that centrally released VP and OT influence learning and behavior (Bohus et al., 1993; Kovacs and Versteeg, 1993). Even though it has been known for many years that the hormones exert their various biological functions via G protein-coupled receptors, considerable progress in elucidating their primary structures in mammalian and nonmammalian species has been made only in recent years. Two reviews have appeared recently dealing with the role of VP in the regulation of body functions (Mohr and Richter, 1994) and covering the broad field of the cellular and molecular biology of VP and OT (Gainer and Wray, 19941, respectively. This chapter concentrates on the molecular biology and evolution of VP and OT as well as their receptors. Considerable progress has been made with regard to peptide hormone gene expression and regulation using transgenic animals and promoter constructs coupled to heterologous reporter genes, one of the topics addressed here. Another interesting feature of vasopressinergic and oxytocinergic neurons to be considered is the notion that VP and OT mRNAs are not exclusively confined to the cell body. Rather, a subset of these transcripts is targeted to neuronal processes, that is, to axons and dendrites. Molecular biological as well as immunocytochemical evidence has been obtained showing that the closely related VP and OT genes are subject t o nonhomologous crossover events in individual magnocellular neurons, another topic of this review. Finally, we give some emphasis to evolutionary aspects of the VPIOT precursor family and their receptors.
VASOPRESSIN AND OXYTOCIN
237
AND REGULATION 11. V P AND OT GENEEXPRESSION
Characterization and definition of regulatory elements (cis-acting sequences) of the VP and OT genes responsible for cell-specific expression are largely hampered by the lack of suitable cell lines endogenously expressing the genes. Our current knowledge about regulatory aspects has been obtained by experimental designs using transgenic animals or by coupling putative regulatory elements to heterologous genes (reporter genes). After transfection of these constructs into suitable cell lines, the regulatory capacity of a DNA segment derived from the VP or OT gene can be investigated. Both approaches have been applied successfully to gain some insight into the cell-specific expression pattern.
ANIMALS A. TRANSGENIC 1. V P Gene A detailed analysis of bovine VP gene expression in transgenic mice (Ang et al., 1993) has revealed that VP gene expression is influenced by positive as well as negative regulatory elements and requires sequence elements located not only in the region upstream of the transcription initiation site but also downstream. When 1.25 kilobase pairs (kb) of proximal promoter sequences of the bovine VP gene were coupled to a heterologous reporter (Fig. la), transgene expression was observed in most peripheral and brain tissues. Transgene expression was limited to neuronal cells in the adrenal medulla and the brain when the bovine V P gene containing the same proximal promoter and, in addition, the complete structural gene sequences was introduced into the germ line of mice (Fig. lb). The most restricted pattern of transgene expression, however, was observed with a 13.4-kb genomic segment consisting of 9 kb of DNA upstream of the transcription start site, the complete structural gene, and 1.5 kb of DNA downstream of the poly(A) addition site (Fig. lc). In the central nervous system of two lines of mice, mRNA corresponding to the bovine transgene was restricted to the hypothalamus. One line also showed ectopic gene expression in the cortex and the hippocampus. Outside the brain several sites of gene expression have been detected. While all lines exhibited VP transgene expression in the ovaries, two lines showed corresponding mRNA in the anterior pituitary gland and one line did so in the kidneys. The latter tissues lack (Mohr et al., 1990) or are thought to
-
238
EVITA MOHR et al
a
CAT
b A
B
C
A
B
C
C
1 kb
FIG.1. Structural organization of constructs consisting of various parts of the bovine vasopressin (VP) gene used to generate transgenic mice (Ang et al., 1993).(a) Here, 1.25 kilobase pairs (kb) of proximal promoter sequences of the bovine VP gene (black line) were fused to a reporter gene (stippled box) encoding the bacterial chloramphenicol acetyltransferase (CAT).The construct gives rise to ubiquitous CAT expression in many cell types inside and outside the brain. (b) This construct, consisting of the same 1.25 kb of proximal promoter sequences in addition to the complete bovine VP structural gene (solid boxes A, B, and C denote the exons), is expressed in cells of neuronal origin. (c) The most restricted pattern of VP transgene expression is observed with a DNA fragment, 13.4 kb in size, spanning 9 kb of 5’ sequences upstream of the transcription initiation site, the complete structural gene, and 1.5 kb downstream of the poly(A) addition site.
lack VP transcripts. Expression in the adrenal gland, which normally harbors VP transcripts (Rehbein et al., 19861, was not evident. Transgene expression in the brain closely followed that of the endogenous mouse gene, that is, in magnocellular neurons of the PVN and the supraoptic nucleus (SON), accessory magnocellular neurons, and cells of the retrochiasmatic region. In contrast, bovine transcripts were absent in parvocellular neurons of the suprachiasmatic nucleus. A marked heterogeneity in the level of transgene expression has been observed, and in the PVN and SON fewer cells express the transgene than the endogenous gene, which might, for instance, be influenced by the site of integration of the bovine gene. Most importantly, like expression of the endogenous gene (Murphy and Carter, 1990),transgene expression is up-regulated during osmotic challenge. In summary, these data provide evidence that the proximal promoter sequence of the bovine V P gene does not contain any tissue-specific capacity, but contains positive regulatory elements active in many neuronal and nonneuronal tissues. Furthermore, negative regulation in peripheral tissues is mediated by a silencer element(s) located in the structural gene or 0.2 kb of DNA downstream of the poly(A) addition site. Restriction of gene expression in magnocellular neurons in the PVN and SON is mediated by a different silencer element(s) contributed by the largest 13.4-kb construct. This DNA fragment also contains regulatory elements necessary for the osmotically induced up-regulation of VP gene expression, which
VASOPRESSIN AND OXYTOCIN
239
has been shown t o occur at the transcriptional level in the rat (Murphy and Carter, 1990). Finally, the 13.4-kb DNA segment obviously harbors an ovary-specific enhancer, but it lacks sequences involved in the control of gene expression in the suprachiasmatic nucleus and the adrenal gland. Whether or not the VP gene of other mammalian species, such as humans, rats, and mice, is regulated in a similar manner remains to be seen. However, earlier studies have shown that a rat genomic DNA fragment that spans a region similar to that of the bovine VP gene, which confers transcriptional activation to most neuronal cell types, failed to be expressed in any tissue of transgenic animals (Younget al., 1990). Hence, it appears that cell type-specific VP gene expression is regulated differently, at least in the rat (see Section II,A,2.). 2 . OT Gene
Transgenic techniques have also been successfully applied to gain some insight into the regulation of rat OT gene expression (Young et al., 1990). Strikingly, introduction of the rat OT gene alone into mice never gave rise to a cell-specific expression pattern of the transgene. However, when the OT gene was coupled to the VP gene in a “minilocus,” thus mimicking the situation of the close proximity of the peptide hormone genes in the rat genome (Mohr et al., 1988a1, expression of the transgene proved t o be restricted to oxytocinergic neurons in the PVN and the SON of the mouse hypothalamus. This construct contained 1.63 kb of the OT gene with only 0.36 kb of proximal promoter sequences as well as the complete structural gene. The gene was coupled to a 3.55-kb Hind111 DNA fragment containing 1.4 kb of DNA upstream of the transcriptional start point of the VP gene in addition to the structural gene (Fig. 2). Ectopic OT gene expression in the cortex and the thalamus was observed in one cell line. Like the bovine VP transgene, the rat OT transgene was expressed at a lower rate than the endogenous counterpart and the corresponding mRNA was detected in only 90% of mouse oxytocinergic neurons. However, transgene-derived transcript levels rise in parallel with endogenous mouse OT mRNA in lactating animals. Rat OT expression in peripheral mouse tissues has not been observed. It should be emphasized that the rat VP gene was not expressed in any brain or peripheral mouse tissues. However, even though 3.55 kb of DNA spanning the rat VP gene are not sufficient to direct its own expression, this sequence obviously contains enhancer elements required for the cell-specific expression of the OT gene. In the rat genome these sequences are located at a distance of at least 10
240
EVITA MOHR et a1
.....................
Rat VP/OT Gene Locus
*.............*.m./A
B
(1)
C
C
C
-
3.55-kb VP Gene
B
A
i.63kb OT Gene
(2)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A
B
B C
A
5.2-kb VP/OT Ttansgene Mlnllocus
................................... A
*..=..-B
C
A
B
C
(3)
1 kb
FIG.2. Structural organization of a rat vasopressin/oxytocin (VPIOT) gene minilocus used to generate transgenic mice that express the OT but not the VP transgene in a cellspecific manner. (1)The genomic organization of the rat VPiOT genes is shown schematically. Both genes are located in close proximity on the same chromosome. They are oriented in opposite transcriptional directions and are separated by approximately 11 kilobase pairs (kb) of intergenic DNA (Mohr et al., 1988a). The VP gene is shown by dotted lines and solid boxes, which denote exons A, B, and C; the OT gene is shown by solid lines and stippled boxes, indicating exons A, B, and C. DNA fragments corresponding to 3.55 kb of the VP gene and 1.63 kb of the OT gene (2) were fused head to tail to generate a VPiOT gene minilocus [schematically shown in (3)l. This construct gives rise to a cell type-specific expression pattern of the OT gene, while no VP transgene transcripts have been detected in any mouse tissue (Young et al., 1990).
kb downstream of the OT gene (Mohr et al., 1988a). The data strongly support the current view of a rather complex nature and arrangement of regulatory elements residing within the peptide hormone genes.
B. PUTATIVE REGULATORY ELEMENTS IN THE VP AND OT GENEPROMOTERS In the rat VP and OT gene promoters several elements exist that resemble known transcription factor binding sites (Mohr et al., 1988b; Mohr and Richter, 19901, most of which are conserved among species such as humans, mice, rats, and cows. They have been functionally tested in heterologous systems by coupling either individual synthetic elements or various portions of promoter sequences containing putative transcription factor binding sites to reporter genes and by subsequent transfection into suitable cell lines. Earlier experiments performed with diencephalon-derived rat pri-
VASOPRESSIN AND OXYTOCIN
24 1
mary cultures have demonstrated a positive influence of cAMP on VP gene expression (Oeding et al., 19901, which could be mediated by a presumptive cAMP response element (CRE) located at position -227 or various putative activator protein (AP) 2 binding sites in the rat VP gene (Mohr and Richter, 1990). Similar sequences are found in the corresponding bovine and human genes. The implication of these elements in the CAMP-dependent up-regulation of transcription is further suggested by transient transfection studies. DNA fragments of the bovine and human VP genes containing elements resembling AP-2 binding sites or CREs were able to moderately elevate the transcriptional rate of reporter genes in heterologous systems (Verbeek et al., 1990; Pardy et al., 1992) upon appropriate stimulation. A negative influence of glucocorticoids on rat VP gene expression is obvious in parvocellular neurons of the PVN (Baldino et al., 1988). Whether this is mediated directly by a glucocorticoid response element (GRE), some of which also act as negative enhancers (Beato, 1989), located at position -662 (Mohr and Richter, 1990) remains to be seen. In a human small-cell lung carcinoma cell line with endogenous VP gene expression, glucocorticoid administration results in elevated rather than diminished VP mRNA levels. Moreover, the effect was seen only upon costimulation with analogs of CAMP, pointing toward an interaction of different transcription factors (Verbeek et al., 1991). Although the results suggest the existence of a GRE in the human VP gene sequence, data are not available. The OT gene promoter is characterized as containing several functional binding sites for steroid hormone (Mohr and Schmitz, 1991), retinoic acid (Richard and Zingg, 19911, and thyroid hormone receptors (Adan et al., 1992). The rat OT gene harbors two estrogen response elements (EREs) (Mohr and Schmitz, 19911, and OT mRNA levels rise in parallel with an increase in estrogens during pregnancy (Zingg and Lefebvre, 1988). The biological significance of these elements has long remained obscure, since oxytocinergic neurons that project to the posterior pituitary gland do not contain detectable amounts of estrogen receptor (Rhodes et al., 1981). Recently, however, expression of the rat OT gene has been shown to take place in epithelial cells of the myometrium of the rat uterus during late gestation (Lefebvre et al., 1992). An influence of estrogen on OT gene transcription in this tissue is certainly conceivable, even though this speculation requires experimental support. One functional ERE is also located in the promoter region of the human OT gene (Richard and Zingg, 1990; Mohr and Schmitz, 1991). In contrast, no such element has been defined in the bovine gene (Adan et al., 19911, indicating that transcriptional regula-
242
EVITA MOHR et al.
tion of the OT gene might be species specific, at least to a certain extent. The rat (as well as the bovine and human) OT gene also contains several TGACC motifs, which are integral parts of thyroid hormone response elements and which indeed confer responsiveness to thyroid hormone in heterologous gene constructs (Adan et al., 1992). Interestingly, two of these elements overlap with the distal ERE. Thyroid hormone may be a physiological regulator of hypothalamic OT gene expression. When experimental animals were treated with thyroid hormone, hypothalamic OT mRNA levels were moderately elevated. This was accompanied by slightly but significantly higher OT peptide levels in the posterior pituitary gland and in blood (Adan et al., 1992). Multiple hormonal regulation of the OT gene is further substantiated by the identification of functional retinoic acid response elements in the human OT gene promoter, some of which are likewise conserved among species (Richard and Zingg, 1991).Retinoids are important regulators of cellular differentiation during embryonic life. Their role in the adult organism is less clear, although retinoic acid receptors are detectable in a variety of tissues, including the brain, with highest concentrations in the hippocampus and the hypothalamus. 111. VP
AND
OT mRNA
IN
DENDRITES AND AXONS
In a variety of nonneuronal and neuronal cell types (Table I), some RNA species are not exclusively confined to the cell body but are specifically targeted to distinct subcellular domains (Steward and Banker, 1992; Steward, 1994). In neurons well-known examples of transcripts undergoing transport to dendrites include mRNAs encoding microtubule-associated protein 2 (Garner et al., 19881, the a-subunit of calcium/calmodulin-dependent protein kinase I1 (Burgin et al., 19901, brain-derived neurotrophic factor (Dugich-Djordjevic et al., 1992), and BC1 RNA, an RNA polymerase I11 transcript (Tiedge et al., 1991). Recently, a large number of mRNAs encoding different glutamate receptors, components of second messenger systems, and the translational control apparatus have been shown to be differentially distributed within individual neurite segments of rat hippocampal neurons in cell culture, supporting the notion that certain proteins are locally synthesized in dendrites (Miyashiro et al., 1994). Specific targeting has also been demonstrated in nonneuronal cells. In oligodendrocytes, for instance, myelin basic protein transcripts are localized to the distal processes of these cells (Brophy et al., 1993).Actin
243
VASOPRESSIN AND OXYTOCIN
TABLE I
PRESENCE OF RNAs IN EXTRASOMAL DOMAINS OF NEURONS~ Type of RNA Dendritic RNAs MAP 2 CaM kinase I1 BDNF BC1 VP OT Axonal RNAs VP OT Dynorphin NF-L BC 1 MAP tau ELH MIP NF
Organism/tissue
Reference
Rat ihippocampus Ratihippocampus Ratihippocampus Ratihippocampus Ratihypothalamus Ratihypothalamus
Garner et al. (1988) Burgin et al. (1990) Dugich-Djordjevic et al. (1992) Tiedge et al. (1991) Mohr et al. (1995) Mohr et al. (1995)
Ratihypothalamus Ratihypothalamus Rat/hypothalamus Ratihypothalamus Ratihypothalamus Raticortex Snailicaudodorsal cells Snaililight green cells Squid/stellate ganglion
Mohr et al. (1991) Mohr et al. (1991) Mohr and Richter (1992) Mohr and Richter (1992) Tiedge et al. (1993) Litman et al. (1993) Dirks et al. (1989) van Minnen (1994) Gluditta et al. (1991)
aIdentification of various RNA species in dendrites and axons of neuronal cell types in both vertebrates and invertebrates, respectively. BC1 RNA is a brain-specific polymerase I11 transcript. All other RNAs represent mRNAs. BDNF, Brain-derived neurotrophic factor; CaM kinase, calciumicalmodulin-dependent protein kinase; ELH, egglaying hormone; MAP, microtubule-associated protein; MIP, molluskan insulin-related peptide; NF, neurofilament; NF-L, low-molecular-weight neurofilament; OT, oxytocin; VP, vasopressin.
mRNA is preferentially located in lamellipodia of fibroblasts (Singer et al., 1989). Finally, a variety of different transcripts are transported to defined cellular domains in Xenopus and Drosophila during oogenesis and/or embryogenesis (Kislauskis and Singer, 1992). Thus, spacial restriction of mRNAs to discrete regions within the cell turns out to be a general phenomenon in many cell types and is thought to permit local translation, thereby establishing functionally different microdomains within a cell. In recent years biochemical, molecular biological, and histochemical analyses have unequivocally shown that in VP, OT, and some other RNA species, such as prodynorphin, lowmolecular-weight neurofilament protein-encoding transcripts and the RNA polymerase I11 transcript BC1 RNA are also located in extrasoma1 compartments of rat hypothalamic magnocellular neurons such as dendrites and/or axons (Bloch et al., 1990; Mohr et al., 1991; Mohr and Richter, 1992; Tiedge et al., 1993; Trembleau et al., 1994).
244
EVITA MOHR et al.
A. DENDRITIC TRANSCRIPTS Sorting of macromolecules within a neuron is a prerequisite to generating functionally different synapses in order to establish and maintain distinct sites of cell-cell communication. It is now well documented that some mRNAs are specifically targeted to dendrites, most likely to allow protein synthesis at appropriate sites (Steward and Banker, 1992). Dendrites are well equipped with all components required for translation (Steward and Levy, 19821, and, by using a special cell culture device, protein synthesis has indeed been demonstrated to occur in dendrites following detachment from the perikaryon (Torre and Steward, 1992). Both Northern blot analyses (Mohr et al., 1995) and in situ hybridization studies (Bloch et al., 1990; Trembleau et al., 1994) have shown that dendrites of hypothalamic magnocellular neurons contain peptide hormone encoding transcripts associated with dendritic rough endoplasmic reticulum (RER) (Trembleau et al., 1994). It may be assumed, therefore, that VP and OT precursors are also synthesized outside the cell body. Magnocellular neurons are remarkable in that VP and OT are not exclusively secreted into the systemic circulation from axons and nerve terminals in the posterior pituitary gland, but are also centrally released from dendrites (Morris et al., 1993). Even though the function of central VP and OT remains a matter of speculation, it is conceivable that central and peripheral release may be differentially regulated within the cell. This may be achieved, for example, by uncoupling precursor protein synthesis in the cell body (for peripheral release) from that in dendrites (for central release). There is evidence indicating that, even within the cell body, VP transcripts are not randomly distributed. Although VP mRNA is detectable in all parts of the cell enriched in RER, it appears to be concentrated in discrete areas along the external side of the RER membranes, while other parts are devoid of transcripts (Trembleau et al., 1994). This view is further supported by immunocytochemical work showing the presence of VP precursor protein in discrete areas of the RER, leaving other parts of the RER unstained by antibodies (Broadwell et al., 1979; Guldenaar and Pickering, 1988).
B. AXONALLY LOCALIZED TRANSCRIPTS According to a current view, the axon of a nerve cell is devoid of protein biosynthesis since it lacks organelles and factors required for translation (Lasek and Brady, 1981). Hence, any functional role for
VASOPRESSIN AND OXYTOCIN
245
transcripts located in this cellular domain is hard to envision, and it has been argued that this may be related to the high secretory activity magnocellular neurons. mRNAs may, for instance, stick to secretory vesicles and therefore unspecifically gain access to the axonal compartment (Gainer and Wray, 1994).In situ hybridization studies at the ultrastructural level, however, have not revealed any striking association of VP mRNA with secretory granules (Trembleau et al., 1994). Furthermore, osmotic challenge, which induces a 2- to 3-fold elevation of both peptide hormone mRNAs in the cell bodies, leads to a differential degree of OT (3-fold) versus VP (17-fold) mRNA accumulation after 7 days of salt loading (Mohr et al., 19911, contradicting the idea of unspecificity. In the cell bodies the osmotically induced increase in mRNA levels is accompanied by a concomitant poly(A) tail elongation of VP- and OT-encoding transcripts (Carrazana et al., 1988). This is quite in contrast to axonal mRNAs. These exhibit shorter poly(A) tails than their counterparts in the perikarya, even under normal conditions, and no elongation is observed following the application of an osmotic stimulus, which should be expected if unspecific targeting were the case (Mohr et al., 1991). Although we do not know many examples of neurons containing axonal mRNAs, this phenomenon is certainly not restricted to cells involved in neurosecretion. For instance, the mRNA encoding the microtubule-associated protein tau is also located in the somato-axonal compartment of rat whole-brain cultured neurons (Litman et al., 1993). Developmental studies performed with hippocampal neurons in cell culture have shown that some, but by no means all, poly(A) RNAs are targeted to axons of young neurons. Upon further maturation mRNAs are largely restricted to the somata and to dendrites. However, transcripts such as actin-, tubulin-, and growth-associated protein 43 mRNAs, which are confined to the cell body in mature neurons, are never found in axons (or dendrites), regardless of the developmental stage (Kleiman et al., 19941, indicating that mRNA targeting to axons of young neurons bears a component of specificity. The physiological significance remains to be investigated. Most likely, and in agreement with a central dogma ofbiology,the axonal V P and OT mRNAs are not translated. Although poly(A)-enriched RNA extracted from the nerve terminals is translatable into the VP and OT precursors in uitro, the majority of transcripts are apparently not associated with ribosomes in uiuo (Mohr et al., 1995). It cannot be ruled out that mRNAs destined to be degraded are targeted to the axonal domain for this purpose. For instance, there is evidence that the length of the poly(A) tract is positively correlated with translation-
246
EVITA MOHR et al
a1 efficiency (see Gallie, 1991, and references therein). mRNAs with relatively short poly(A) tails may therefore interfere with an optimal level of protein biosynthesis by competing with longer transcripts for ribosome binding. As a net result, protein synthesis might be less effective. To circumvent this problem, shorter molecules might be removed from the perikaryon and transferred to the axon for degradation. It may also be speculated that the axonal transcripts serve a different role in the control of translation. A variety of examples are known of mRNAs that are rendered translationally inactive by association with proteins to form ribonucleoprotein particles (Altmann and Trachsel, 1993; Cardinali et al., 1993; Goodwin et al., 1993; Jeffries et al., 1994; Minich et al., 1993; OstareckLederer et al., 19941, which represent a transcript storage pool to be reactivated on demand. In developing oocytes stored transcripts often exhibit shorter poly(A) tails, as do the axonal VP and OT mRNAs (Mohr et al., 1990), compared with their translationally active counterparts. Following exit from the cell nucleus, these RNAs are specifically deadenylated in the cytoplasm and, as a consequence, translationally dormant. Cytoplasmic readenylation leads to recruitment into the translationally active pool of mRNAs (Wickens, 1992). A similar mechanism might be operating in hypothalamic magnocellular neurons by inactivating and storing only a portion of VP and OT mRNAs in the axon. These could be transported back into the perikaryon and, following cytoplasmic poly(A) tail elongation, again serve as templates for translation. Such a mechanism might be a means to achieve efficient and fast replenishment of the depleted peptide hormone pools when, after release of the osmotic stimulus, VP gene expression has returned to control levels. Earlier work on OT gene expression in the bovine corpus luteum may indeed be indicative of translational control of OT transcripts in this tissue. OT mRNA levels are highest on day 3 of the estrus cycle and decline thereafter, while OT and NP immunoreactivity show a peak on days 8-13 of the cycle (Ivell et al., 1985; Ivell, 1986). Thus, translation apparently does not correlate with the highest levels of peptide hormone transcripts. Definite experimental proof for the control of VP and OT gene expression at the translational level is currently lacking. It is noteworthy, however, that axons of magnocellular neurons contain particles resembling ribonucleoprotein complexes. They increase in number in animals subjected t o an osmotic stimulus, a condition which is associated with a dramatic increase in the level of axonal VP and, to a lesser extent, OT mRNA (Mohr et al., 1993).
VASOPRESSIN AND OXYTOCIN
247
IV. SOMATIC RECOMBINATION BETWEEN THE VP AND OT GENES IN HYPOTHALAMIC NEURONS Hypothalamic neurons cease to divide during embryonic life (Altman and Bayer, 1978). Accordingly, their chromatin must be protected, by a currently unknown mechanism, from any possible damage during postnatal development and adulthood in order to maintain their appropriate phenotype. However, immunocytochemical data have indicated that magnocellular neurons of the VP-deficient Brattleboro (BB) rat may be subject to genomic alterations of the VP gene (Pow et al., 1992). The mutant animal is characterized by the absence of plasma VP, which is caused by a frameshift mutation in the second exon of the corresponding gene encoding the major portion of the NP carrier protein (Schmale and Richter, 1984). Although the gene is transcribed at an almost normal level, the mutant mRNA is only very inefficiently translated (Schmale et al., 1984). This is substantiated by undetectable amounts of VP and its associated NP. Moreover, due to the deletion of a single G nucleotide, the precursor contains an altered C-terminal end and thus lacks the glycopeptide moiety. Nevertheless, solitary neurons do exist in the BB rat that can be stained with anti-VP antibodies (Richards et al., 1985). It has been concluded from these observations that an event at the genomic level must have occurred that somehow restores the wild-type reading frame of the gene and the normal phenotype of the cell. Further investigations have revealed that these cells fall into two classes: The majority of VP-positive cells also show GPlike immunoreactivity, while a few others lack not only the GP but also the VP-associated NP. Instead, these neurons contain the NP carrier protein normally associated with the OT precursor (Pow et al., 19921, strongly suggesting a nonhomologous crossover event between the closely related peptide hormone genes schematically depicted in Fig. 3A. By applying reverse transcription (RT) of hypothalamic RNA in combination with polymerase chain reaction (PCR) amplification, hybrid transcripts encoding the N-terminal part of the VP precursor but the C-terminal part of the OT precursor have indeed been identified (Mohr et al., 1994). Strikingly, nonhomologous recombination was not restricted to vasopressinergic neurons of the BB rat, since hybrid transcripts encoding the OT precursor at the N terminus and the VP precursor at the C-terminal end have also been amplified using the same techniques (Fig. 3B). Somatic recombination events are therefore not correlated with the diseased state, a view that was further substantiated by the existence of both hybrid mRNA species in the wild-type rat
248
EVITA MOHR et al.
A r
.
ExonC
-
,
ExonB
. OT (*nr
-11 lrbp-18 kbp
v
Hybrid gene r
B
Hybrid transcript
4l
p
9'
number of clones sequenced wild-type BB rat
1 VPlOT
1
s+ - - s
1
5'
v V
2 VPlOT 3 VPlOT 4 VPlOT
5 OTIVP 6 OTIVP
7 OTlVP
r
ExmA
1
7
12
1
6
1
-
6 3
-5 - '5
-
3
2
, -
v
T
V
N P o *
. . . . v
v
V
T
-3'
V
N
V
T
* v
v
v
* v
T
*
5
m 3' * 3 '
' 5 5 ' i
5
v
'
m 3 '
.
s o l l - v % P r. 3~' q
FIG.3. (A) The organization of the vasopressin (VP; solid boxes) and oxytocin (OT; open boxes) genes in the rat genome is shown in the upper part. The genes are separated by approximately 11 kilobase pairs (kbp) of intergenic DNA and are oriented in opposite transcriptional directions, with their 3' ends facing each other (Mohr et al., 1988a). Below, a somatic nonhomologous crossover event involving exons B of the VP and OT genes is schematically demonstrated. This event would generate hybrid VPiOT as well as OTiVP genes. In vasopressinergic neurons only the hybrid VPiOT gene is likely to be expressed, because it is driven by the appropriate promoter. (B) The structural organization of various hybrid VPiOT and OTiVP transcripts is proposed based on data obtained by DNA sequence determination of reverse-transcribed polymerase chain reaction (PCR) products (for experimental details see Mohr et al., 1994). The number of PCR products subcloned and sequenced from wild-type and Brattleboro (BB) rats is indicated. VPencoding sequences are shown by black boxes, those encoding the OT precursor by open boxes. Open arrowheads, position of introns; closed arrowhead, position of a single GIC nucleotide difference in the otherwise identical sequences within the central part of exons B; *, a single G nucleotide deletion in the VP precursor gene of the BB rat; GP, glycopeptide; NPo, oxytocin-associated neurophysin; NPv, vasopressin-associated neurophysin; OT,oxytocin; S, signal peptide; VP, vasopressin.
VASOPRESSIN AND OXYTOCIN
249
(Mohr et al., 1994).A hot-spot region for recombination events seems to exist in the second exon of the peptide hormone genes, which are ideal candidates for nonhomologous crossover (Fig. 3B). First, their structural organization is very similar; both genes are composed of three exons (Schmale et al., 1983; Ivell and Richter, 1984). Second, they exhibit a high degree of sequence identity, particularly in the central part of the second exons. Finally, the genes are located in very close proximity in the rat genome and are oriented in opposite transcriptional directions, with their 3’ ends facing each other (Mohr et al., 1988a). RT-PCR techniques have also been used t o characterize events leading to a VP- and GP-positive phenotype (Evans et al., 1994). Restoration of the GP-encoding reading frame in the mutant VP gene of the BB rat is accomplished either by a 2-bp deletion, mainly at two different GAGAG sequence motifs within the second exon of the VP gene, or by insertion of one nucleotide residue at various positions. Again, these alterations do not appear to be restricted t o the BB rat. Immunocytochemical analyses with antibodies directed against synthetic peptides that should be encoded as a consequence of either GA deletion in the wild-type rat reveal the existence of corresponding mutant VP precursor proteins in solitary hypothalamic magnocellular neurons (Evans et al., 1994). None of these studies has unequivocally shown that the VP-positive phenotypes are indeed caused at the genomic level, but the combined immunocytochemical and molecular biological data strongly favor this idea. If true, this clearly indicates the possibility of DNA alterations and rearrangements in postmitotic neurons by mechanisms that must be defined. V. EVOLUTION OF THE VERTEBRATE VP/OT GENEFAMILY Within vertebrates-and even in invertebrates-hormones comprising members of the VP/OT family are widely distributed. Nonapeptide hormone precursors have now been characterized by molecular cloning from the most primitive vertebrate, the jawless cyclostome, up to humans, allowing insight into the evolution of the corresponding gene family. Arginine vasotocin (VT), the nonmammalian counterpart of VP, occurs throughout vertebrate evolution, from cyclostomes to birds. In mammals this peptide is replaced by either arginine or lysine VP. Peptides representing counterparts of mammalian OT, such as mesotocin (MT),isotocin (IT), glumitocin, and aspargtocin, are present in all
250
EVITA MOHR et al.
nonmammalian vertebrates except cyclostomes (for a review see Heierhorst et al., 1993). Although the absence of an OT-like peptide in Agnatha implies that the VP/OT gene family has evolved from an ancestral VT gene by duplication and subsequent diversification after the radiation of cyclostomes, this view might have to be revised, since an OT-like peptide has been identified in nerve tissue of cephalopods (Reich, 1992). Deduced from corresponding cDNA and/or genomic sequences, the following picture of precursor organization emerges. Compared with the mammalian VP precursor, the VT precursors of all nonmammalian vertebrates show a very similar organization. They contain a hydrophobic signal peptide at the N terminus, followed by the hormone moiety and the Gly-Lys-Arg motif. In cyclostomes (Heierhorst et al., 1992) and fish (Heierhorst et al., 1989) the NP carrier protein is C-terminally extended by approximately 30 amino acid residues that show some similarity to the mammalian GP sequence. However, a processing signal required for cleavage of the extended part is lacking in these precursors. A very similar organization is also apparent in amphibia. Here, a monobasic processing signal that could potentially serve as the processing signal for the C-terminal GP is present in the VT precursor (Nojiri et al., 1987). The structural organizations of the MT [amphibia (Nojiri et al., 198711 and IT precursors [fish (Heierhorst et al., 1989)J are also very similar t o that of the mammalian OT polyprotein, except that in the fish IT precursor the NP moiety is likewise C-terminally extended by approximately 30 amino acid residues. As in the corresponding VT precursor, a potential cleavage site is lacking. It appears likely that the extended part of the NP carrier protein of cyclostomes and fish has evolved into a separate GP in the VP and possibly the VT precursor of higher vertebrates, while it has been deleted in amphibian and mammalian MT and OT precursors, respectively, probably due to a lack of physiological function. The genomic organization of all genes belonging to the VPiOT family is very similar in mammalian and nonmammalian vertebrates, with one exception, the hagfish. In most cases, at least those that have been investigated to date, the hormone precursors are encoded by genes consisting of three exons and two intervening sequences at conserved positions (see Fig. 3 for an example). The hagfish contains an additional intron in the 5' region of its VT gene (see Heierhorst et al., 1992, and references therein). Another peculiarity is the extraordinarily large size of the third intron, spanning more than 14 kb. Sequence analysis has revealed the presence of an open reading frame, on the opposite DNA strand, with the capacity to encode a protein with striking similar-
VASOPRESSIN AND OXYTOCIN
251
ity t o the putative transposases of Tcl-like elements normally found in nematodes and Drosophila, indicating a broader distribution of these elements than anticipated (Heierhorst et al., 1992).
RECEPTORS VI. NONAPEPTIDE As outlined above, the mammalian neurohypophysial nonapeptide hormones, VP and OT, are involved in the mediation of a variety of physiological effects, such as salt and water balance, blood pressure regulation (VP), smooth muscle contraction during birth and lactation (OT), regulation of hormone secretion from the anterior pituitary gland, learning, and behavior (Bohus et al., 1993; Kovacs and Versteeg, 1993).The physiological roles of the nonmammalian nonapeptide homologues VT and IT/MT are less well defined. However, recent physiological and biochemical data, as well as peptide binding analyses and in sztu hybridization studies using prohormone oligonucleotide probes, point t o renal (Kloas and Hanke, 1992; Perrott et al., 1993), cardiovascular (Le Mevel et al., 19931, reproductive (Moore, 1992; de Kloet et al., 1993), metabolic (Moon and Mommsen, 19901, and hydroosmotic functions (Guibbolini and Lahlou, 1991, 1992),urine excretion (Bakos et al., 19921, and modulation of neurotransmission (Moons et al., 1989; Boyd, 1991). The different physiological effects of nonapeptides in mammals and lower vertebrates, together with binding studies using various nonapeptide analogs and the detection of diverse signal transduction pathways, suggested the existence of various nonapeptide receptor subtypes. Thus, OT receptors that are linked to phosphatidylinositol hydrolysis and mobilization of intracellular calcium exist in various tissues such as those of the reproductive organs, kidneys, and brain (reviewed in Mohr et al., 1992).VP has been shown to act on two classes of receptors, V1 and V2, that are coupled to the inositol trisphosphate/calcium and adenylate cyclase second messenger pathways, respectively (reviewed by Jard et al., 1987). the V1 receptors can be divided into two subtypes, termed V l a and Vlb receptors. While the latter is involved in the modulation of adrenocorticotropic hormone secretion from the anterior pituitary gland (Antoni, 1984; Jard et al., 19861, the former mediates blood vessel constriction and hepatic glycogenolysis (Mitchell et al., 1979).Final proof of the existence of neurohypophysial hormone receptor subtypes has recently been obtained by molecular cloning of V l a arginine VP (Morel et al., 1992; Thibonnier et al., 1994) and V2 arginine (Lolait et al., 1992; Birnbaumer et al., 1992; Barberis et al.,
252
EVITA MOHR et a1
1993) and lysine VP (Gorbulev et al., 1993), OT (Kimura et al., 1992, 1993; Gorbulev et al., 1993), VT and IT receptors (Mahlmann et al., 1994; Hausmann et al., manuscript submitted) and the molluskan conopressin receptors (van Kesteren et al., 1995). A. MOLECULAR CLONING Based on the observation that receptors can be functionally expressed and identified in Xenopus oocytes following microinjection of mRNA from appropriate tissues (Meyerhof et al., 1988; Morley et al., 19881, the V l a VP and OT receptors were the first neurohypophysial hormone receptors to be cloned (Morel et al., 1992; Kimura et al., 1992). The OT receptor cDNA was isolated by Xenopus oocyte expression cloning using a common electrophysiological assay (Kimura et al., 19921, while the V l a VP receptor was cloned by a variation of this method. Aequorin, a light-emitting calcium-binding protein, has been coinjected with cDNA-derived RNA preparations into oocytes, and VPmediated rises in cytosolic calcium levels were detected by liquid scintillation counting (Morel et al., 1992). The V2 VP receptor has consequently been isolated by PCR amplification of a kidney cDNA library with degenerate oligonucleotide primers deduced from the rat V l a receptor transmembrane domains I1 and VI (Lolait et al., 1992). Similar approaches have been used to isolate the pig V2 lysine VP (Gorbulev et al., 19931, the teleost fish VT and IT (Mahlmann et al., 1994; Hausmann et al., manuscript submitted), and molluskan conopressin receptors (van Kesteren et al., 1995). A completely different and independent approach was used to clone the human V2 receptor (Birnbaumer et al., 1992). Isolation of the gene encoding this receptor was achieved by stable transfection of mouse L cells, which exhibit a V2 receptor-negative phenotype using human genomic DNA. Mouse L-cell lines were subsequently investigated for their ability to stimulate adenylate cyclase following arginine VP application. DNA from a positive cell line was isolated and used to transfect mouse L cells to a V2-positive phenotype. After a third series of transfection, a genomic DNA library in phage A was constructed from a V2 receptor-positive cell line, and human genes were isolated by screening for the presence of human repetitive DNA elements. A 2.2kb BamHI DNA fragment was finally isolated and identified to contain the V2 receptor gene by its ability to confer VP responsiveness t o mouse L cells stably transfected with this DNA fragment. The fragment was used t o isolate a cDNA clone from a library that has been
VASOPRESSIN AND OXYTOCIN
253
prepared from a mouse L-cell line containing the human V2 receptor gene.
B. RECEPTOR STRUCTURE The deduced primary structures of the cloned neurohypophysial hormone receptors display lengths that vary between 371 and 435 amino acids for the rat or human V2 and the teleost VT receptors, respectively. The longer amino acid sequence of the VT receptor is primarily due to a 5-fold repetition of a dodecapeptide sequence of unknown significance at its C terminus (Mahlmann et al., 1994). Hydropathic analyses have revealed that the nonapeptide receptors possess seven regions of hydrophobic amino acids characteristic of G protein-coupled receptors with seven putative transmembrane domains. This structural organization has been suggested earlier based on biochemical and pharmacological data (see, e.g., the review by Mohr et al., 1992). The degree of sequence relationship observed among the neurohypophysial hormone receptors (at least 39% identity) clearly exceeds that of other G protein-coupled receptors (not exceeding 35%in the comparably well-conserved transmembrane regions), indicating that the neurohypophysial hormone receptors form a small subfamily. All nonapeptide receptors (Fig. 4) show highest homologies within their putative transmembrane regions, as observed in other subfamilies of G protein-coupled receptors (Probst et al., 1992). However, within extracellular loops 1 and 2, but not in loop 3, and in parts of the transmembrane regions that face the extracellular surface, certain amino acid residues are conserved among members of the neurohypophysial nonapeptide receptor subfamily but not among other G proteincoupled receptors (Fig. 4). It has been suggested that these residues might be involved in the formation of ligand binding sites (Sharif and Hanley, 1992). In fact, electrophysiological analyses in Xenopus oocytes of mutant VT and IT receptors indicated that the N-terminal extracellular domain and extracellular loops 1 and 2 decisively contribute to sensitivities of the receptors for nonapeptides, while extracellular loop 3 participates in peptide selectivity (Hausmann et al., manuscript submitted). These observations lend support from other neuropeptide receptors, in which it has also been shown that extracellular loops might well participate in ligand binding (reviewed by Coughlin, 1994; Baldwin, 1994). Direct proof for the involvement of extracellular domains in peptide hormone binding has been obtained by cross-linking experiments of a tritium-labeled photoreactive VP
254
EVITA MOHR e l al
hVlaR MDSMRLSAGP DAGPSGNSSP WWPLATGAGN TSREAEALGE GNGPP.. . . . . . . .RDVRNE ELAKLEIAVL CCVTR .............................. MGRIANQTTA SNDTD . . . . . . . . . PFGRNE EVAKMEITVL hOTR ........................ MEGALA ANWSAEAANA SAAPPGAEGN RTAGPPRRNE ALARVEVAVL hV2R
..........
hVlaR CCVTR hOTR hV2R hVlaR CCVTR hOTR hV2R
AVTFAVAVLG NSSVLLALHR SVTFFVAVIG NLSVLLAMHN CLILLLALSG NACVLLALRT SIVFVAVALS NGLVLAALAR
TPRKTSR..M TKKKSSR..M TRQKHSR..L RGRRGHWAPI
T1(3
VKHLQVFGMF VKHLQVLGMF VKYLQWGMF VKYLQMVGMY
hVlaR NVTKA C CCVTR NGSYV C hOTR DG..V C
ASAYMLWMT ASTYMMVMMT ASTYLLLLMS ASSYMILAMT
ADRYIAVCHP LDRYIAICHP LDRCLAICQP LDRHFAICRP
LKTLQQ.PAR RSRLMIAAAW LKTLQQ.PTQ RAYIMIGSTW LRSLRR.RTD R..LAVLATW MLAYRHGSGA HWNRPVLVAW Tld5 TFI PWG RA YVTWMTGGIF VAPWILGTC YGFICYNIWC HFI PWG RA YITWITVGIF LIPVIILMIC YGFICHSIWK VFI PWG KA YITWITLAVY IVPVIVLATC YGLISFKIWQ RT YVTWIALMVF VAPTLGIAAC QVLIFREIHA CFAi.2 Tld6 PCVSSVKSIS RAKIRTVKMT FVIVTAYIVC WAPFFIIQMW VSVSSVTIIS RAKLRTVKMT LVIVLAYIVC WAPFFIVQMW ARVSSVKLIS KAKIRTVKMT FIIVLAFIVC WTPFFFVQMW .........S AAVAKTVRMT L V I W W L C WAPFFLVQLW
$3
hV2R GGSGV hVlaR VAFQKGFLLA CCVTR GMIGK hOTR AAGDGGRVAL hV2R SPGEGAHV..
.....
hVlaR CCVTR hOTR hV2R hVlaR CCVTR hOTR hV2R
T1I7 ALLGSLNSCC ALLASLNSCC MLLASLNSCC MLLASLNSCT
NPWIYMFFSG NPWIYMLFSG NPWIYMLFTG NPWIYASFSS
-
HLFIRHLSLA D HLFIKHLSLA FFFMKHLSIA HVFIGHLCLA VLSFVLSTPQ LCSLLLSTPQ LGCLVASAPQ AFSLLLSLPQ
YFVFSMIEVN YFIFSLSEIQ VHIFSLREVA LFIFAQRNVE
NVRGKT. .AS RQSKGAEQAG NIKCKTMRGT RNTKD . . . . . NLRLKT..AA AAAAEAPEGA SLVPGPSERP GGRRRGRRTG
-
SVWDPMSWJT SVWDENFSWD SVWDAN...A AAWDPE ...A
ESENPTITIT DSENAAVTLS PKEASAFIIV PLEGAPFVLL
...
HLLQDCVQSF PCCQNMKEKF NKEDTDSMSR RQTFYSN NRSPTNSTGM HLLYDFLRCF PCCKKPRNML QKEDSDSSIR RNTLLTKLAA GRMTNDGFGS HLFHELVQRF LCC . . . . . . . . . . .SASYLK GRRLGETSAS KKSNSSSFVL SVSSE.LRSL LCCARGR... ....................... TPPSLGP
.............................
WKD...SPKS SKSIKFIPVS T ........... WRDPCNSRKS SQSIGLDCFC KSSQCLEHDC SRKSSQCIPL DCSRKSSQCI RKSSQ CMSKES SHRSSSQRSC SQPSTA............................ . . . . . . . . . . . . . . . . . . . . . . QDESCTTASS SLAKDTSS.. .......... ................................
FIG.4. Alignment of various neurohypophysial hormone receptors. The sequences of the human V l a arginine vasopressin [hVlaR (Thibonnier et al., 1994)1,teleostean vasotocin (ccVTR), human oxytocin [hOTR (Kimura et al., 1992)1, and human V2 arginine vasopressin [hVPR (Birnbaumer et al., 1992)l receptors are shown in single-letter code. Dashes denote gaps that have been introduced for maximal homology. The putative transmembrane regions (TM 1-7) are marked by lines above the sequences. Amino acid residues that are specifically conserved among neurohypophysial hormone receptors are boxed. Residues that have been shown to be involved in ligand binding in the bovine V2 receptor (Kojro et al., 1993) are underlined in the hV2R sequence.
agonist t o the bovine renal V2 VP receptor (Kojro et al., 1993). Following binding, cross-linking, and subsequent protease digestion of the ligand-occupied receptor, it has been shown that the radioactivity was associated with two amino acid residues in the second extracellular domain (the corresponding positions are underlined in the human V2 arginine receptor sequence in Fig. 41,suggesting that these residues participate in the constitution of the peptide hormone binding domain. At present, mutational analyses have not been described for VP and OT receptors. However, the importance of receptor domains or individual amino acid residues in receptor functioning can be inferred from
VASOPRESSIN AND OXYTOCIN
255
mutant receptors that have been characterized in patients suffering from congenital nephrogenic diabetes insipidus (CNDI),a rare X-linked recessive genetic disorder. The V2 receptor mediates, by stimulation of adenylate cyclase, the antidiuretic effect of VP (hence the alternative designation antidiuretic hormone) through an increase in water reabsorption across the apical membrane of the renal collecting duct cells (Skorecki et al., 1992). CNDI patients show renal insensitivity to VP. Recent studies have identified a variety of mutations in the V2 VP receptor genes in a number of patients (Pan et al., 1992; van den Ouweland et al., 1992; Davies, 1992; Bichet et al., 1992; Merendino et al., 1993; Holtzmann et al., 1993). Truncated receptors lacking various portions of the C terminus have been identified in addition to receptors exhibiting single amino acid substitutions. So far, only one case has been reported showing the functional consequences of a single amino acid substitution, namely, the exchange of Arg137 by His. This amino acid is located at the junction of the third transmembrane segment and the second intracellular loop and is highly conserved among G proteincoupled receptors (Probst et al., 1992). Mutational analyses performed on the p,-adrenergic receptor and rhodopsin have demonstrated that an arginine in this position is important for G protein interaction (Fraser et al., 1988; Franke et al., 1992). Consistent with these reports, the His137 mutant V2 VP receptor exhibits an unaltered agonist binding behavior but fails to stimulate adenylate cyclase (Rosenthal et al., 1993). These data not only explain the generation of the CNDI phenotype in patients carrying the His137mutation in the V2 VP receptor gene, but they underscore the importance of Arg137 for receptor coupling to the stimulatory G protein (GS).
C. TISSUEDISTRIBUTION OF NEUROHYPOPHYSIAL HORMONE RECEPTORSAND THEIRmRNAs The distribution of VP and OT receptors has been analyzed by binding studies and physiological assays in a number of tissues. While the distribution of V2 VP receptors in the collecting duct of the kidney is well established, their presence in the medullary thick ascending limb and the renal vascular system is still controversial, as is the presence of V l a receptors in the kidneys (Terada et al., 1993). The situation is even more complicated since it has been reported that OT receptors, as well as VP receptors that are different from the V2 and V l a subtypes and that are linked to calcium mobilization, are apparently present in the inner medullary collecting duct (Maeda et al., 1993). Cloning of the nonapeptide receptors has now provided the molecu-
256
EVITA MOHR et al.
lar tools necessary to analyze in detail the distribution of OT,V l a and V2 VP, and VT receptor mRNAs in various tissues. By using Northern blot hybridization, comparably high levels of Vla transcripts have been detected in the liver, while lower, but still clearly detectable, levels of this mRNA species have been observed in the kidneys (Morel et al., 19921, where, as anticipated, V2 receptor mRNA is abundant (Birnbaumer et al., 1992). The presence of corresponding mRNA in the liver and the pituitary gland suggests that, in part, the teleostean VT receptor is functionally equivalent to the mammalian V1-type VP receptors. However, the observation of VT receptor mRNA in the gills, lateral line, and swim bladder indicates specific functions for this receptor in fishes. Using in situ hybridization, Vla transcripts have been observed in medullary vascular elements, arcuate and interlobular arteries, short segments of the cortical distal tubule, and transitional epithelium and smooth muscle of the pelvic wall and the ureter, in agreement with VP agonist binding profiles (Tribollet et al., 1988). In contrast to the predominantly vascular localization of Vla transcripts, V2 mRNA was found primarily in cells of the collecting duct and nephron elements, which is in accordance with the known role of V2 receptors in the regulation of water reabsorption (Ostrowski et al., 1992, 1993). A differential distribution of V2 and V l a receptors along the rat nephron has also been observed using RT-PCR techniques on microdissected rat nephron segments (Terada et al., 1993).Moreover, in this study it was shown that, while V2 receptor mRNA levels decreased in the collecting ducts in response to dehydration, the amount of V l a mRNA did not change significantly. These data suggest different roles for both receptor subtypes in the modulation of renal function. The results further suggest, in combination with the observation of differential ontogenic expression patterns of Vla and V2 receptors in the liver and the kidneys, respectively, that their expression is not subject to common regulatory mechanisms (Ostrowski et al., 1993). In the rat central nervous system V l a but not V2 receptor transcripts have been detected. However, V l a mRNA is not exclusively confined to areas known to receive VP innervation (Ostrowski et al., 1992). For instance, the observed localization of Vla mRNA in the cerebellum, inferior olive, and arcuate nucleus that lack vasopressinergic input is less well understood. OT receptor mRNA has been localized by in situ hybridization in different regions of the olfactory bulb, cerebrum, diencephalon, mesencephalon, and brain stem (Yoshimura et al., 1993). The distribution of OT receptor mRNA agrees well with that of OT binding sites, suggesting that most OT receptors are located on somata and/or proximal dendrites. In the inferior olive,
VASOPRESSIN AND OXYTOCIN
257
however, intense OT binding has been reported to occur, but OT receptor mRNA was not detectable. A similar discrepancy has been detected in the posterior lobe of the pituitary gland (Yoshimura et al., 1993). In these regions the receptor molecules are presumably located far from the cell bodies on axons and/or distal dendrites (Yoshimura et al., 1993). An OT receptor cDNA fragment isolated from the ewe was used to investigate OT receptor mRNA levels in the sheep endometrium during the estrus cycle, which correlated well with the appearance and disappearance of OT binding sites in this tissue (Stewart et al., 1993). A more detailed study using a synthetic 45-mer oligodeoxynucleotide showed that both 0" binding and OT receptor mRNA prevalance occurred at estrus. Hence, the ovine endometrial OT receptor appears to be regulated primarily at the transcriptional, rather than the translational, level (Stevenson et al., 1994).
D. RECEPTOR PHYLOGENY A comparison of the mammalian and nonmammalian nonapeptide receptor sequences shows that the teleost fish VT receptor is most closely related to the mammalian Vla VP receptors (Fig. 5). Both resemble the OT receptors, while the V2 VP receptors are more dis-
hV2R rV2R pLV2R hOTR pOTR cclTR
/ I I ccVTR ICPR2
FIG.5. Sequence relationship of various neurohypophysial hormone receptors depicted in a dendrogram. The amino acid sequences of molluskan conopressin [ICPRl and ICPR2 (van Kesteren et al., 1995)],teleostean vasotocin [ccVTR (Mahlmann et al., 1994)l and isotocin [ccITR (Hausmann, et al., manuscript submitted)], rat and human V l a arginine vasopressin [rVlaR and hVlaR (Morel et al., 1992; Thibonnier et al., 1994)1, porcine [pOTR (Gorbulev et al., 1993)l and human oxytocin [hOTR (Kimura et al., 1992)1, rat [rV2R (Lolait et al., 1992)] and human V2 arginine vasopressin IhV2R (Birnbaumer et al., 1992; Barberis et al., 1993)], and porcine lysine vasopressin IpLV2R (Gorbulev et al., 1993)l receptors have been aligned for maximal homology using the Pileup algorithm (Genetics Computer Group, Madison, Wisconsin).
258
EVITA MOHR et al.
tantly related. Most likely, an ancestral neurohypophysial hormone receptor gene has been duplicated to give rise to two receptor subfamilies represented by the Vla and V2 VP receptors. The OT receptors evolved later during evolution from the V l a receptor subfamily, but before the separation of the VT and V l a VP receptors. A comparison of the human V l a VP, V2 VP, and OT receptor sequences, for example, reveals a divergence of Vla and V2 receptors of 64% and of V l a and OT receptors of approximately 46%. If one assumes a unit evolutionary period [defined as the time, in millions of years, required for fixation of 1%divergence (Wilson et al., 197713 of about 10, as observed for various globin genes (Efstratiadis et al., 1980; Perler et al., 1980; Knochel et al., 19861, the V l a and V2 receptor types evolved about 640 million years ago, well before the separation of cylostomes from other vertebrates (about 500 million years ago). This assumption predicts the presence of at least two VP-like receptor subtypes in cyclostomes. Thus, these nonapeptide receptor subtypes apparently evolved before the nonapeptide hormone family diverged into VP- and OT-like peptides. The OT receptors probably arose approximately 460 million years ago, before cartilaginous fish originated. Consequently, OT-like receptors should be present in all nonmammalian vertebrates, except cyclostomes, which coincides with the presence or absence of OT-like peptides in nonmammalian vertebrates (Acher, 1980). Surprisingly, the molluskan receptors ICPRl and ICPRB, which have been cloned from the pond snail Lymnaea stagnalis display only a low degree of sequence similarity, indicating a very early phylogenetic separation of their corresponding genes (van Kesteren et al., 1995). The sequence divergence of 71% between the molluskan ICPRB and the human V l a arginine vasopressin receptor sequences indicates that an ancestral conopressin receptor gene evolved before the separation of Vla- and V2-type receptor genes. This conclusion is consistent with the radiation of Protostomia and Deuterostomia in the Precambrian era. It further argues that the molluskan peptide cephalotocin (Reich, 1992) is likely to represent an unusual conopressin-like peptide that lost a positively charged amino acid residue in position 8 rather than an OTlike peptide (see Section V). REFERENCES Acher, R. (1980). Molecular evolution of biologically active polypeptides. Proc. R . Soc. London B 210,21-43. Adan, R. A. H., Walther, N., Cox, J. J., Ivell, R., and Burbach, J. P. H. (1991). Comparison of the estrogen responsiveness of the rat and bovine oxytocin gene promoters. Biochem. Biophys. Res. Commun. 175, 117-122.
VASOPRESSIN AND OXYTOCIN
259
Adan, R. A. H., Cox, J . J., van Kats, J. P., and Burbach, J . P. H. (1992).Thyroid hormone regulates the oxytocin gene. J. Biol. Chem. 267, 3771-3777. Altman, J., and Bayer, S. A. (1978). Development of the diencephalon in the rat. I. Autoradiographic study of the time and of origin and settling patterns of neurons of the hypothalamus. J. Comp. Neurol. 182, 945-972. Altmann, M., and Trachsel, H. (1993). Regulation of translation initiation and modulation of cellular physiology. Trends Biochem Sci. 18, 429-432. Ang, H.-L., Carter, D. A., and Murphy, D. (1993). Neuron-specific expression and physiological regulation of bovine vasopressin transgenes in mice. EMBO J . 12, 23972409. Antoni, F. (1984). Novel ligand specificity of pituitary vasopressin receptors in the rat. Neuroendocrinology 39, 186-188. Bakos, P., Shakhmatova, E. I., Ponec, J., Alexandrova, M., Lichardus, B., Lammek, B., Rekowski, P., and Kupyszewski, G. (1992). Hydroosmotic activities of argininevasopressin modified either in positions 1,2, and 4 or a t N-terminal extensions. Gen. Physiol. Biophys. 11, 359-376. Baldino, F., OKane, T. M., Fitzpatrick-McElligott, S., and Wolfson, B. (1988).Coordinate hormonal and synaptic regulation of vasopressin messenger RNA. Science 241,978981. Baldwin, J . M. (1994). Structure and function of receptors coupled to G-proteins. Curr. Opin. Cell Biol. 6, 180-190. Barheris, C., Seibold, A,, Ishido, M., Rosenthal, W., and Birnbaumer, M. (1993).Expression cloning of the human V2 vasopressin receptor. Regul. Pept. 45, 61-66. Beato, M. (1989). Gene regulation by steroid hormones. Cell 56, 335-344. Bichet, D. G., Hendy, G. N., Lonergan, M., Arthus, M.-F., Ligier, S., Pausova, Z., Kluge, R., Zingg, H., Saenger, P., Oppenheimer, E., Hirsch, D. J., Gilgenkrantz, S., Salles, J.-P., Oberle, I., Mandel, J.-L., Gregory, M. C., Fujiwara, T. M., Morgan, K., and Scriver, C. R. (1992). X-linked nephrogenic diabetes insipidus: From the ship Hopewell to RFLP studies. A m . J . Hum. Genet. 51, 1089-1102. Birnbaumer, M., Seibold, A,, Gilbert, S., Ishido, M., Barberis, C., Antaramian, A,, Brabet, P., and Rosenthal, W. (1992). Molecular cloning of the receptor for antidiuretic hormone. Nature 357, 333-335. Bloch, B., Guitteny, A. F., Normand, E., and Chouham, S. (1990). Presence of neuropeptide mRNAs in neuronal processes. Neurosci. Lett. 109, 259-264. Bohus, B., Borrell, J., Koolhaas, M., Nyakas, C., Buwalda, B., Compaan, J. C., and Roozendaal, B. (1993). The neurohypophysial peptides, learning, and memory processing. Ann. N.Y. Acad. Sci. 689, 285-299. Boyd, S. K. (1991). Effect of vasotocin on locomotor activity in bullfrogs varies with developmental stage and sex. Horm. Behau. 25, 57-69. Broadwell, R. D., Oliver, C., and Brighton, M. W. (1979). Localization of neurophysin within organelles associated with protein synthesis and packaging in the hypothalamoneurohypophysial system: An immunocytochemical study. Proc. Natl. Acad. Sci. U.S.A. 76,5999-6003. Brophy, P. J., Boccaccio, G. L., and Colman, D. R. (1993).The distribution of myelin basic protein mRNAs within myelinating oligodendrocytes. Trends NeuroSci. 16, 515521. Burgin, K. E., Waxham, M. N., Rickling, S., Westgate, S. A., Mobley, W. C., and Kelly, P. T. (1990). I n situ hybridization histochemistry of CaZ+/calmodulin-dependent protein kinase in developing rat brain. J . Neurosci. 10, 1788-1798.
260
EVITA MOHR et al.
Cardinali, B., di Christina, M., and Pierandrei-Amaldi, P. (1993).Interaction of proteins with the mRNA for ribosomal protein L1 in Xenopus: Structural characterization of in vivo complexes and identification of proteins that bind in vitro to its 5’ UTR. Nucleic Acids Res. 21, 2301-2308. Carrazana, E. J., Pasieka, K. B., and Majzoub, J. A. (1988). The vasopressin mRNA poly(A) tract is unusually long and increases during stimulation of vasopressin gene expression in uivo. Mol. Cell. Biol. 8, 2267-2274. Coughlin, S. R. (1994). Expanding horizons for receptors coupled to G-proteins: Diversity and disease. Curr. Opin. Cell Biol. 6, 191-197. Davies, K. (1992). Diabetes defect defined. Nature 359, 434. De Kloet, E. R., Elands, J., and Voorhuis, D. A. (1993). Implications of central neurohypophyseal hormone receptor-mediated action in timing of reproductive events: Evidence from novel observations on the effect of a vasotocin analogue on singing behaviour of the canary. Regul. Pept. 45, 85-89. Dirks, R. W., Raap, A. K., van Minnen, J., Vreugdenhil, E., Smit, A. B., and van der Ploeg, M. (1989). Detection of mRNA molecules for neuropeptide hormones in the pond snail Lymnaea stagnalis by radioactive and non-radioactive in situ hybridization: A model study for mRNA detection. J . Histochem. Cytochern. 37, 7-14. Dugich-Djordjevic, M. M., Tocco, G., Willoughby, D. A,, Najm, I., Pasinetti, G., Thompson, R. F., Baudry, M., Lapchak, P. A,, and Hefti, F. (1992).BDNF mRNA expression in the developing rat brain following kainic acid-induced seizure activity. Neuron 8, 1127-1 138. Efstratiadis, A,, Posakony, J. W., Maniatis, T., Lawn, R. M., OConnell, C., Spritz, R. A,, deRiel, J. K., Weissman, S. M., Slightom, J. L., Blechl, A. E., Smithies, O., Baralle, F. E., Shoulders, C. C., and Proudfoot, N. J. (1980). The structure and evolution of the human globin gene family. Cell 21, 653-668. Evans, D. A. P., van der Kleij, A. A. M., Sonnemans, M. A. F., Burbach, J. P. H., and van Leeuwen, F. W. (1994). Frameshift mutations a t two hotspots in vasopressin transcripts in post-mitotic neurons. Proc. Natl. Acad. Sci. U.S.A.91, 6059-6063. Franke, R. R., Koenig, B., Sakmar, T. P., Graham, R. M., and Khorana, H. G. (1992). Structure and function in rhodopsin. Studies of the interaction between the rhodopsin cytoplamsic domains and transducin. J . Biol. Chem. 267, 14767-14774. Fraser, C. M., Chung, F.-Z., Want, C.-D., and Venter, J. C. (1988). Site-directed mutagenesis of human p-adrenergic receptors: Substitution of aspartic acid 130 by asparagine produces a receptor with high-affinity binding that is uncoupled from adenylate cyclase. Proc. Natl. Acad. Sci. U.S.A.85, 5478-5482. Gainer, H., and Wray, S. (1994). Cellular and molecular biology of oxytocin and vasopressin. In “Physiology of Reproduction” (E. Knobil and J. D. Neill, eds.), pp. 10991129. Raven, New York. Gallie, D. R. (1991). The cap and poly(A) tail function synergistically to regulate mRNA translational efficiency. Genes Deu. 5, 2108-2116. Garner, C. C., Tucker, P. R., and Matus, A. (1988). Selective localization of messenger RNA for cytoskeletal protein MAP 2 in dendrites. Nature 336, 674-677. Giuditta, A,, Menichini, E., Perrone-Capano, E., Langella, C., Martin, R., Castigli, E., and Kaplan, B. B. (1991).Active polysomes in the axoplasm of the squid giant axon. J . Neurosci. 28, 18-22. Goodwin, E. B., Okkema, P. G., Evans, T. C., and Kimble, J. (1993).Translational regulation of tra-2 by its 3’ untranslated region controls sexual identity in C . elegans. Cell 75,329-339. Gorbulev, V., Buchner, H., Akhundova, A., and Fahrenholz, F. (1993).Molecular cloning
VASOPRESSIN AND OXYTOCIN
261
and functional expression of V2 [8-lysinel vasopressin and oxytocin receptors from a pig kidney cell line. Eur. J . Biochem. 215, 1-7. Guibbolini, M. E., and Lahlou, B. (1991). Evidence of a new type of neurohypophyseal hormone receptor in fish gill epithelium. A m . J . Physiol. 258, R3-R9. Guibbolini, M. E., and Lahlou, B. (1992). Gi protein mediates adenylate cyclase inhibition by neurohypophyseal hormones in fish gill. Peptides 13, 865-871. Guldenaar, S. E. F., and Pickering, B. T. (1988). Mutant vasopressin precursor in the endoplasmic reticulum of the Brattleboro rat. Ultrastructural evidence from individual “vasopressin” cells localized with the light microscope by use of a new goldisilver method for immunostain enhancement. Cell Tissue Res. 253, 671-676. Hausmann, H., Meyerhof, W., Zwiers, H., Lederis, K., and Richter, D. (1995).Teleost isotocin receptors: Structure, functional expression, mRNA distribution and phylogeny. FEBS Lett., submitted for publication. Heierhorst, J., Morley, S. D., Figueroa, J., Krentler, C., Lederis, K., and Richter, D. (1989). Vasotocin and isotocin precursors from the white sucker, Catostomus commersoni: Cloning and sequence analysis of the cDNAs. Proc. Natl. Acad. Sci. U.S.A. 86,5242-5246. Heierhorst, J., Lederis, K., and Richter, D. (1992). Presence of a member of the Tcl-like transposon family from nematodes and Drosophila within the vasotocin gene of a primitive vertebrate, the Pacific hagfish Eptatretus stouti. Proc. Natl. Acad. Sci. U.S.A. 89, 6798-6802. Heierhorst, J., Lederis, K., and Richter, D. (1993). Vasotocin neuropeptide precursors and genes of teleost and jawless fish. In “Biochemistry and Molecular Biology of Fishes” (P. W. Hochachka and T. P. Mommsen, eds.), Vol. 2, pp. 339-355. Elsevier, Amsterdam. Holtzmann, E. J., Harries, H. W., Kolakowski, L. F., Guay Woodford, L. M., Bothelho, B., and Ausiello, D. A. (1993). A molecular defect in the vasopressin V2-receptor gene causing nephrogenic diabetes insipidus. N . Engl. J . Med. 328, 1534-1537. Ivell, R. (1986). Biosynthesis of oxytocin in the brain and in peripheral organs. Curr. Top. Neuroendocrinol. 6, 1-18. Ivell, R., and Richter, D. (1984). Structure and comparison of the oxytocin and vasopressin genes from rat. Proc. Natl. Acad. Sci. U.S.A.81, 2006-2010. Ivell, R., Brackett, K. H., Fields, M. J., and Richter, D. (1985). Ovulation triggers oxytocin gene expression in the bovine ovary. FEBS Lett. 190,263-267. Jard, S., Gaillard, J. M. R., Guillon, G., Marie, J., Schoenenberg, P., Muller, A. F., Manning, M., and Sawyer, W. H. (1986). Vasopressin antagonists allow demonstration of a novel type of vasopressin receptors in the rat adenohypophysis. Mol. Pharmacol. 30, 171-177. Jard, S., Barberis, C., Audigier, S., and Tribollet, E. (1987). Neurohypophyseal hormone receptor systems in brain and periphery. Prog. Brain Res. 72, 173-187. Jeffries, H. B. J., Thomas, G., and Thomas, G. (1994). Elongation factor-la mRNA is selectively translated following mitogenic stimulation. J . Biol. Chem. 269, 43674372. Kimura, T., Tanizawa, O., Mori, K., Brownstein, M. J., and Okayama, H. (1992). Structure and expression of a human oxytocin receptor. Nature 356, 526-529. Kimura, T., Azuma, C., Takemura, M., Kikuchi, T., Kubota, Y., Ogita, K., Saji, F., and Tanizawa, 0. (1993). Molecular cloning of a human oxytocin receptor. Regul. Pept. 45,73-77. Kislauskis, E. H., and Singer, R. H. (1992). Determinants of mRNA localization. Curr. Opin. Cell Biol. 4, 975-978.
262
EVITA MOHR et a1
Kleiman, R., Banker, G., and Steward, 0. (1994). Development of subcellular mRNA compartmentation in hippocampal neurons in culture. J. Neurosci. 14, 1130-1 140. Kloas, W., and Hanke, W. (1992). Localization and quantification of nonapeptide binding sites in the kidney of Xenopus laeuis: Evidence for the existence of two different nonapeptide receptors. Gen. Comp. Endocrinol. 85, 71-78. Knochel, W., Korge, W., Basner, A,, and Meyerhof, W. (1986). Globin evolution in the genus Xenopus: Comparative analysis of cDNAs coding for adult globin polypeptides of Xenopus borealis and Xenopus tropicalis. J. Mol. Evol. 23, 211-223. Kojro, E., Eich, P., Gimpl, G., and Fahrenholz, F. (1993). Direct identification of an extracellular agonist binding site in the renal V2 vasopressin receptor. Biochemistry 32, 13537-13544. Kovacs, G. L., and Veersteeg, D. G. H. (1993). Neurohypophysial peptides and brain neurochemistry. Ann. N.Y. Acad. Sci. 689, 309-319. Lasek, R. J., and Brady, S. T. (1981). The axon: A prototype for studying expressional cytoplasm. Cold Spring Harbor Symp. Quant. Biol. 46, 113-184. Lefebvre, D. L., Giaid, A., Bennett, H., Lariviere, R., and Zingg, H. H. (1992). Oxytocin gene expression in rat uterus. Science 256, 1553-1554. Le Mevel, J. C., Pamantung, T. F., Mabin, D., and Vaudry, H. (1993). Effects of arginine vasotocin and related neuropeptides on blood pressure and heart rate in the conscious trout. Brain Res. 610, 82-89. Litman, P., Barg, J., Rindzoonski, L., and Ginzburg, I. (1993).Subcellular localization of tau mRNA in differentiating neuronal cell culture: Implication for neuronal plasticity. Neuron 10, 627-638. Lolait, S., OCarroll, A.-M., McBride, 0. W., Konig, M., Morel, A., and Brownstein, M. J. (1992). Cloning and characterization of a vasopressin V2 receptor: Chromosomal location of gene suggests link to hereditary nephrogenic diabetes insipidus. Nature 357,336-339. Maeda, Y., Han, J. S., Gibson, C. C., and Knepper, M. A. (1993).Vasopressin and oxytocin receptors coupled to Ca2+ mobilization in rat inner medullary collecting duct. A m . J . Physiol. 265, F15-F25. Mahlmann, S., Meyerhof, W., Hausmann, H., Heierhorst, J., Schonrock, C., Zwiers, H., Lederis, K., and Richter, D. (1994). Structure, function, phylogeny of [Arg8lvasotocin receptors from teleost fish and toad. Proc. Natl. Acad. Sci. U.S.A.91, 13421345. Merendino, J. J., Spiegel, A. M., Crawford, J. D., OCarroll, A.-M., Brownstein, M. J., and Lolait, S. J. (1993). A mutation in the vasopressin V2-receptor gene in a kindred with X-linked nephrogenic diabetes insipidus. N . Engl. J. Med. 328, 1358-1541. Meyerhof, W., Morley, S. D., Schwarz, J., and Richter, D. (1988). Receptors for neuropeptides are induced by exogenous polyA+ RNA in oocytes from Xenopus laevis. Proc. Natl. Acad. Sci. U.S.A.85, 714-717. Minich, W. B., Maidebura, I. P., and Ovchinnikov, L. P. (1993). Purification and characterization of the major 50-kDa repressor protein from cytoplasmic mRNP of rabbit reticulocytes. Eur. J . Biochem. 212, 633-638. Mitchell, R. H., Kirk, C. J., and Bikkah, M. M. (1979).Hormonal stimulation of phosphatidylinositol breakdown, with particular reference to the hepatic effects of vasopressin. Biochem. SOC.Trans. 7, 861-865. Miyashiro, K., Dichter, M., and Eberwine, J. (1994). On the nature and differential distribution of mRNAs in hippocampal neurites: Implications for neuronal functioning. Proc. Natl. Acad. Sci. U.S.A.91, 10800-10804. Mohr, E., and Richter, D. (1990). Sequence analysis of the promoter region of the rat vasopressin gene. FEBS Lett. 260, 305-308.
VASOPRESSIN AND OXYTOCIN
263
Mohr, E., and Richter, D. (1992). Diversity of mRNAs in the axonal compartment of peptidergic neurons in the rat. Eur. J. Neurosci. 4, 870-876. Mohr, E., and Richter, D. (1994). Vasopressin in the regulation of body functions. J. Hypertens. 12, 345-348. Mohr, E., and Schmitz, E. (1991). Functional characterization of estrogen and glucocorticoid responsive elements in the rat oxytocin gene. Mol. Bruin Res. 9, 293-298. Mohr, E., Schmitz, E., and Richter, D. (1988a). A single rat genomic DNA fragment encodes both the oxytocin and vasopressin genes separated by 11 kilobases and oriented in opposite transcriptional directions. Biochimze 70, 649-654. Mohr, E., Bahnsen, U., Kiessling, C., and Richter, D. (198813).Expression of the vasopressin and oxytocin genes occurs in mutually exclusive sets of hypothalmic neurons. FEBS Lett. 242, 144-148. Mohr, E., Zhou, A,, Thorn, N. A,, and Richter, D. (1990). Rats with physically disconnected hypothalamo-pituitary tracts no longer contain vasopressin-oxytocin gene transcripts in the posterior pituitary lobe. FEBS Lett. 263, 332-336. Mohr, E., Fehr, S., and Richter, D. (1991). Axonal transport of neuropeptide encoding mRNAs within the hypothalamo-hypophyseal tract of rats. EMBO J. 10, 24192424. Mohr, E., Meyerhof, W., and Richter, D. (1992). The hypothalamic hormone oxytocin: From gene expression to signal transduction. Rev. Physiol., Biochem. Phurmacol. 121,31-48. Mohr, E., Morris, J. F., and Richter, D. (1995).Differential subcellular mRNA targeting: Deletion of a single nucleotide prevents the transport to axons but not to dendrites of rat hypothalamic magnocellular neurons. Proc. Nutl. Acud. Sci. U S A . 92,43774381. Mohr, E., Terjung, D., Martin, R., and Richter, D. (1993). Vasopressin mRNA in the hypothalamo-hypophyseal tract. I n “Vasopressin” (P. Gross, D. Richter, and G. L. Robertson, eds.), pp. 119-129. Libbey Eurotext, Paris. Mohr, E., Peters, A,, Morris, J. F., and Richter, D. (1994). Somatic non-homologous crossing over between neuropeptide genes in rat hypothalamic neurons. Proc. Natl. Acad. Sci. U.S.A. 91, 11403-11407. Moon, T. W., and Mommsen, T. P. (1990).Vasoactive peptides and phenylephrine actions in isolated teleost hepatocytes. A m . J. Physiol. 259, E644-E649. Moons, L., Cambre, M., Batten, T. F., and Vandesande, F. (1989). Autoradiographic localization of binding sites for vasotocin in the brain and pituitary of the sea bass (Dicentrurchus lubrax).Neurosci. Lett. 100, 11-16. Moore, F. L. (1992). Evolutionary precedents for behavioral actions of oxytocin and vasopressin. Ann. N.Y. Acad. Sci. 652, 156-165. Morel, A,, O’Carroll, A. M., Brownstein, M. J., and Lolait, S. J. (1992). Molecular cloning and expression of the rat V l a arginine vasopressin receptor. Nature 356,523526. Morley, S. D., Meyerhof, W., Schwarz, J.,and Richter, D. (1988).Functional expression of the oxytocin receptor in Xenopus oocytes primed with mRNA from bovine endometrium. J. Mol. Endocrinol. 1, 77-81. Morris, J. F., Pow, D. V., Sokol, H. W., and Ward, A. (1993).Dendritic release of peptides from magnocellular neurons in normal rats, Brattleboro rats and mice with hereditary nephrogenic diabetes insipidus. I n “Vasopressin” (P. Gross, D. Richter, and G. L. Robertson, eds.), pp. 171-182. Libbey Eurotext, Paris. Murphy, D., and Carter, D. A. (1990). Vasopressin gene expression in the rodent hypothalamus: Transcriptional and posttranscriptional responses to physiological stimulation. Mol. Endocrinol. 4, 1051-1059.
264
EVITA MOHR et al.
Nojiri, H., Ishida, I., Miyashita, E., Sato, M., Urano, A., and Deguchi, T. (1987). Cloning and sequence analysis of cDNAs for neurohypophysial hormones vasotocin and mesotocin for the hypothalamus of toad, Bufo juponicus. Proc. Natl. Acad. Sci. U.S.A.84, 3043-3046. Oeding, P., Schilling, K., and Schmale, H. (1990). Vasopressin expression in cultured neurons is stimulated by cyclic AMP. J . Neuroendocrinol. 2, 859-865. Ostareck-Lederer, A., Ostareck, D. H., Standart, N., and Thiele, B. J. (1994). Translation of 15-lipoxygenase mRNA is inhibited by a protein that binds to a repeated sequence in the 3‘ untranslated region. EMBO J. 13, 1476-1481. Ostrowski, N. L., Lolait, S. L., Bradley, D. J., O’Carroll, A.-M., Brownstein, M. J., and Young, W. S. (1992). Distribution of V l a and V2 vasopressin receptor messenger ribonucleic acids in rat liver, kidney, pituitary and brain. Endocrinology 131, 533535. Ostrowski, N. L., Young, W. S., Knepper, M. A., and Lolait, S. L. (1993). Expression of vasopressin V l a and V2 receptor messenger ribonucleic acid in the liver and kidney of embryonic, developing and adult rats. Endocrinology 133, 1849-1859. Pan, Y., Metzenberg, A,, Das, S., and Gitscher, J. (1992). Mutations in the V2 vasopressin receptor gene (AVPR2)are associated with X-linked nephrogenic diabetes insipidus. Nature Genet. 2, 103-106. Pardy, K., Adan, R. A. H., Carter, D. A.,Seah, V., Burbach, J. P. H., and Murphy, D. (1992). The identification of a cis-acting element involved in cyclic 3’,5’-adenosine monophosphate regulation of bovine vasopressin gene expression. J . Biol. Chem. 267,21746-21752. Perler, F., Efstratiadis, A., Lomedico, P., Gilbert, W., Kolodner, R., and Dodgson, J. (1980). The evolution of genes: The chicken preproinsulin gene. Cell 20, 555-565. Perrot, M. N., Sainsbury, R. J., and Balment, R. J. (1993). Peptide hormone-stimulated second messenger production in the teleostean nephron. Gen. Comp. Endocrinol. 89, 387-395. Pow, D. V., Morris, J. F., and Ward, A. R. (1992). Immuno-electron microscopic evidence for two different types of partial somatic repair of the mutant Brattleboro vasopressin gene. Neuroscience 50, 503-512. Probst, W. C., Snyder, L. A., Schuster, D. I., Brosius, J., and Sealfon, S. C. (1992). Sequence alignment of the G-protein receptor superfamily. DNA Cell Biol. 11, 1-20, Rehbein, M., Hillers, M., Mohr, E., Ivell, R., Morley, S., Schmale, H., and Richter, D. (1986). The neurohypophyseal hormones vasopressin and oxytocin. Precursor structure, synthesis, and regulation. Biol. Chem. Hoppe-Seyler 367, 695-704. Reich, G.(1992). A new peptide of the oxytocinhasopressin family isolated from nerves of the cephalopod Octopus uulgaris. Neurosci. Lett. 134, 191-194. Rhodes, C. H., Morrell, J. J., and Pfaff, D. W. (1981). Distribution of estrogenconcentrating, neurophysin-containing magnocellular neurons in the r a t hypothalamus as demonstrated by a technique combining steroid autoradiography and immunohistology in the same tissue. Neuroendocrinology 33, 18-23. Richard, S., and Zingg, H. H. (1990). The human oxytocin gene promoter is regulated by estrogens. J. Biol. Chem. 265, 6098-6103. Richard, S.,and Zingg, H. H. (1991). Identification o f a retinoic acid response element in the human oxytocin promoter. J . Biol. Chem. 266, 21428-21433. Richards, S. J.,Morris, R. J.,and Raisman, G. (1985). Solitary magnocellular neurons in the homozygous Brattleboro rat have vasopressin and glycopeptide immunoreactivity. Neuroscience 16, 617-623. Richter, D. (1987). Biochemistry and biology of vasopressin, oxytocin, and their corre-
VASOPRESSIN AND OXYTOCIN
265
sponding neurophysins. In “The Peptides” (C. W. Smith, ed.), Vol. 8, pp. 41-75. Academic Press, Orlando, Florida. Rosenthal, W., Antaramian, A., Gilbert, S., and Birnbaumer, M. (1993). Nephrogenic diabetes insipidus. A V2 vasopressin receptor unable to stimulate adenylate cyclase. J . Biol. Chem. 268, 13030-13033. Schmale, H., and Richter, D. (1984). Single base deletion in the vasopressin gene is the cause of diabetes insipidus in Brattleboro rats. Nature 308, 705-709. Schmale, H., Heinsohn, S., and Richter, D. (1983). Structural organization of the rat gene for the arginine vasopressin-neurophysin precursor. EMBO J . 2, 763-767. Schmale, H., Ivell, R., Breindl, M., Darmer, D., and Richter, D. (1984). The mutant vasopressin gene from diabetes insipidus (Brattleboro) rats is transcribed but the message is not efficiently translated. EMBO J . 3, 3289-3293. Sharif, M., and Hanley, M. R. (1992). Stepping up the pressure. Nature 357, 279-280. Singer, R. H., Langevin, G. L., and Lawrence, J. B. (1989). Ultrastructural visualization of cytoskeletal mRNAs and their associated proteins using double-label in situ hybridization. J . Cell Biol. 108, 2343-2353. Skorecki, K. L., Brown, D., Ercolani, L., and Ausiello, D. A. (1992). Molecular mechanisms of vasopressin action in the kidney. In “Handbook of Physiology. section 8, Renal Physiology” Vol. 11, pp. 1185-1218. Oxford University Press, Oxford. Stevenson, K. R., Riley, P. R., Stewart, H. J., Flint, A. P., and Wathes, D. C. (1994). Localization of oxytocin receptor mRNA in the ovine uterus during the oestrus cycle and early pregnancy. J . Mol. Endocrinol. 12, 93-105. Steward, 0. (1994). Dendrites as compartments for macromolecular synthesis. Proc. Natl. Acad. Sci. U.S.A.91, 10766-10768. Steward, O., and Banker, G. A. (1992). Getting the message from the gene to the synapse: Sorting and intracellular transport of RNA in neurons. Trends Neurosci. 15, 180-186. Steward, O., and Levy, W. B. (1982).Preferential localization of polyribosomes under the base of dendritic spines in the granule cells of the dentate gyrus. Neuroscience 2, 284-291. Stewart, H. J., Stevenson, K. R., and Flint, A. P. (1993). Isolation and structure of a partial sheep oxytocin receptor cDNA and its use as a probe for Northern analysis of endometrial RNA. J . Mol. Endocrinol. 10, 359-361. Terada, Y., Tomita, K., Nonoguchi, H., Yang, T., and Marumo, F. (1993).Different localization and regulation of two types of vasopressin receptor messenger RNA in microdissected rat nephron segments using reverse transcription PCR. J . Clin. Znuest. 92, 2339-2345. Thibonnier, M., Auzan, C., Madhun, Z., Wilkins, P., Berti-Mattera, L., and Clauser, E. (1994).Molecular cloning, sequencing, and functional expression of a cDNA encoding the human V l a vasopressin receptor. J . B b l . Chem. 269, 3304-3310. Tiedge, H., Fremeau, R. F., Weinstock, P. H., Arancio, O., and Brosius, J. (1991).Dendritic location of neural BC1 RNA. Proc. Natl. Acad. Sci. U.S.A. 88, 2093-2097. Tiedge, H., Zhou, A,, Thorn, N. A., and Brosius, J. (1993). Transport of BC1 RNA in hypothalamo-neurohypophyseal axons. J . Neurosci. 13,4214-4219. Torre, E. R., and Steward, 0. (1992). Demonstration of local protein synthesis within dendrites using a new cell culture system that permits the isolation of living axons and dendrites from their cell bodies. J. Neurosci. 12, 762-772. Trembleau, A., Morales, M., and Bloom, F. E. (1994).Aggregation of vasopressin mRNA in a subset of axonal swellings in the median eminence and posterior pituitary: Light and electron microcopic evidence. J . Neurosci. 14, 39-53.
266
EVITA MOHR et al.
Tribollet, E., Barberis, C., Dreifuss, J.-J., and Jard, S. (1988). Autoradiographic localization of vasopressin and oxytocin binding sites in rat kidney. Kidney Int. 33,959-965. van den Ouweland, A. M., Dreesen, J. C. F. M., Verdijk, M., Knoers, N. V. A. M., Monnens, L. A. H., Rocchi, M., and van Oost, B. A. (1992). Mutations in the V2 vasopressin receptor gene (AVPR2) associated with nephrogenic diabetes insipidus. Nature Genet. 2, 99-102. van Kesteren, R. E., Tensen, C. P., Smit, A. B., van Minnen, J.,Kits, K. S., Meyerhof, W., Richter, D., van Heerikhuisen, H., Vreugdenhil, E., and Geraerts, W. P. M. (1995).A novel member of the vasopressin/oxytocin receptor family: Cloning, expression and functions of a Lymnaea conopressin receptor. Neuron, in press. van Minnen, J . (1994).Axonal localization of neuropeptide-encoding mRNA in identified neurons of the snail Lymnaea stagnalis. Cell Tissue Res. 276, 155-161. Verbeek, M. A. E., Adan, R. A. H., and Burbach, J. P. H. (1990). Vasopressin gene expression is stimulated by cyclic AMP in homologous and heterologous systems. FEBS Lett. 272, 89-93. Verbeek, M. A. E., Sutano, W., and Burbach, J. P. H. (1991). Regulation of vasopressin mRNA levels in small cell lung carcinoma cell line GLC-8: Interaction between glucocorticoids and second messengers. Mol. Endocrinol. 5, 795-801. Wickens, M. (1992). Forward, backward, how much, when: Mechanisms of poly(A) addition and removal and their role in early development. Semin. Deu. Biol. 3,399-412. Wilson, A. C., Carlson, S. S., and White, T. J. (1977). Biochemical evolution. Annu. Rev. Biochen. 46, 573-639. Yoshimura, R., Kiyama, H., Kimura, T., Araki, T., Maeno, H., Tanizawa, O., and Tomayama, M. (1993). Localization of oxytocin receptor ribonucleic acid in the ra t brain. Endocrinology 133, 1239-1246. Young, W. S., Reynolds, K., Shepard, E. A., Gainer, H., and Castel, M. (1990). Cellspecific expression of the rat oxytocin gene in transgenic mice. J. Neuroendocrinol. 2,917-925. Zingg, H. H., and Lefebvre, D. L. (1988). Oxytocin and vasopressin gene expression during gestation and lactation. Mol. Brain Res. 4, 1-6.
VITAMINS AND HORMONES, VOL. 51
Structure and Function of Estrogen Receptors M. G. PARKER Molecular Endocrinology Laboratory Imperial Cancer Research Fund Lincoln’s Inn Fields, London WC2A 3PX, England
I. Introduction 11. Physiological Responses to Estrogens 111. Intracellular Localization of Estrogen Receptors IV. Hormone Binding and Receptor Dimerization V. Target Gene Recognition VI. Transcriptional Activation VII. Specific Gene Transcription by the Estrogen Receptor VIII. Mechanism of Antiestrogen Action IX. Cross-coupling with Other Signaling Pathways References
I. INTRODUCTION Steroid hormones, retinoids, thyroid hormone, and other small lipophilic hormones elicit numerous effects in mammalian cells and tissues by binding to receptors that fbnction directly as ligand-dependent transcription factors. It is doubtful whether any tissue fails to respond in some way to at least one of these hormones, and in many cases the effects are profound. Thus, retinoids regulate the growth and development of many tissues and the sex steroids play important roles in reproductive tissues. It is well established from studies of individuals with defective androgen receptors that the development and growth of male sex accessory tissues are absolutely dependent on androgen action. Recent work analyzing mice with defective estrogen receptors has confirmed that estrogen action is essential for the growth of female sex accessory tissue but questions their importance in development, at least of the uterus (Lubahn et al., 1993). It remains to be established whether estrogens play a role in the development of other tissues, but in any event this work has emphasized the importance of estrogen action in the growth and normal function of the female reproductive tract. The primary role of estrogens in these processes is to regulate the expression of specific target genes, and the aim of this chapter is to review our current understanding of the molecular mechanism of action of estrogen receptors. 267
Copyright 0 1995 by Academic Press, Inc. All rights of reproduction in any form reserved.
268
M.G. PARKER C
AI0
4 1
D
E/F
4 4 4 4 4 4 2
3
4
5
6
7
8
lntrons Exons
FIG.1. Domain structure of nuclear receptors. The N-terminal domain comprises regions A and B and is encoded by exon 1, the DNA binding domain (DBD) comprises region C and is encoded by exons 2 and 3, and the ligand binding domain (LBD) comprises regions E and F and is encoded by exons 5-8.
The estrogen receptor is a member of a superfamily of nuclear receptors that includes receptors not only for steroid and thyroid hormones and certain vitamins but many other proteins, referred to as orphan receptors, whose ligand has yet to be identified. The steroid hormone receptors are encoded by unique genes that, in common with the majority of nuclear receptor genes, comprise eight coding exons. Transcription of the estrogen receptor gene is initiated from two promoters giving rise to RNA transcripts with alternate 5' ends (White et al., 1987; Keaveney et al., 1991), but nevertheless they both encode identical open reading frames. Members of the nuclear receptor family contain a highly conserved DNA binding domain and a moderately conserved C-terminal ligand binding domain, and most of them also have a third, N-terminal, domain, which is variable in length and sequence (Evans, 1988; Parker, 1993; Beato, 1989).The receptors have also been divided into six regions, A-F, based on their sequence homology with other receptors (Krust et al., 1986). Exon 1encodes the N-terminal domain (regions A and B); exons 2 and 3, the DNA binding domain (region C); exon 4,a so-called hinge region (region D); and exons 5-8, the hormone binding domain (regions E and F) (Fig. 1). Since the organization and sequence of nuclear receptors are conserved, it is assumed that they all function by similar mechanisms, and so, in addition to reviewing the structure and function of estrogen receptors, this chapter also compares estrogen action with that of other nuclear receptors.
11. PHYSIOLOGICAL RESPONSES TO ESTROGENS The first estrogen target genes to be studied in detail were those that encoded secretory proteins, since they were often expressed at high levels. The best-characterized estrogen-responsive genes are probably
STEROID RECEPTORS
269
the chicken ovalbumin gene and the Xenopus vitellogenin genes, and from their analysis it appears that the estrogen receptor can function by two distinct mechanisms. The regulation of vitellogenin gene expression involves binding of the estrogen receptor to relatively simple DNA binding sites (Wahli, 19881, whereas that of the ovalbumin gene may involve another transcription factor, activator protein 1 (AP-11, binding to a composite AP-1 site (Gaub et al., 1990). The first mechanism involves direct binding of the receptor to target sites, where it, itself, functions as a transcription factor. The second mechanism involves protein-protein interactions with the receptor modulating the activity of AP-1. It seems likely that estrogens regulate the transcription of target genes by one or the other, or possibly a combination, of these two types of mechanisms (Fig. 2). Estrogens stimulate the proliferation not only of normal cells in female sex accessory tissues, but also of a number of cancers, including breast, endometrial, and ovarian cancers. The mechanisms involved remain poorly understood, partly because estrogens regulate the expression of so many target genes implicated in cell proliferation that it is difficult to identify the crucial ones. For example, estrogens might function as mitogens indirectly by increasing the production of growth
+--b Simple Estrogen Response Element
AP-1 site FIG.2. Mechanisms of transcriptional activation by estrogen receptors. In the classical type of response, receptor (R) homodimers, occupied by ligand (L), bind to simple response elements to stimulate transcription. Alternatively, the transcriptional activity of Fos (F)/Jun (J)family members bound to their cognate binding site AP-1 can be modulated either positively or negatively by estrogen receptors.
270
M. G. PARKER
factors or possibly by reducing the secretion of growth-inhibitory factors (for reviews see Clarke et al., 1991; Roberts and Sporn, 1992). A number of groups studying breast cancer cell lines have obtained support for this concept by demonstrating that estrogens increase the expression of many types of growth factor, although only transforming growth factor a,(TGF-a) has been detected to any great extent in the majority of primary breast tumors (Gaub et al., 1990; Bates et al., 1988).Conversely,estrogens have been found to decrease the production of the inhibitory growth factor TGF-P in several estrogen receptorpositive breast cancer cell lines, suggesting that it, too, might be an important mediator of the action of nuclear receptors in cell growth (Knabbe et al., 1987).In addition to effects on growth factors, estrogens have also been shown to stimulate the expression of growth factor receptor genes, including the insulin-like growth factor (1GF)-I receptor and the epidermal growth factor (EGF) receptor (Stewart et al., 1990). Therefore, another effect of estrogens might be to increase the sensitivity of cells to growth factors produced either by tumor cells themselves or perhaps by surrounding stromal cells, thereby regulating cell proliferation by a paracrine mechanism (Clarke et al., 1991). Thus, according to these views, the major effects of steroid hormones and retinoids are indirect, stimulating cell growth by an autocrine or paracrine control mechanism. In support of the autocrine hypothesis, estrogen antagonists such as tamoxifen and the “pure” antiestrogen ICI 182780 are capable of reversing most, if not all, of the estrogen-inducible events described above. Thus, tamoxifen treatment prevents the increased expression of the growth factor TGF-a and stimulates the secretion of the growth inhibitor TGF-P, consistent with its effects on cell growth. However, it is doubtful that antiestrogens function simply by antagonizing the effects of estradiol, because they not only inhibit estrogen-induced cell growth but are also capable of blocking the mitogenic effects of growth factors such as EGF and IGF-I in the absence of estrogen (Berthois et al., 1989; Chalbos et ad., 1994). This observation indicates that antiestrogens elicit a growth-inhibitory response in their own right. This might be achieved by their effects on TGF-P production, but it might also involve the antagonism of other signaling pathways. In spite of the large body of evidence suggesting that estrogens stimulate cell growth by an autocrine or paracrine control mechanism, there are clearly direct effects of steroids on the cell cycle. It has been found that estrogens stimulate the recruitment of breast cancer cells into the cell cycle and shorten the length of their G, phase (Brunner et
STEROID RECEPTORS
271
al., 1989). On the other hand, estrogen antagonists reduce the proportion of cells in S phase, supporting the suggestion that steroid hormones, like other growth factors, act within the GI phase of the cell cycle t o regulate cell proliferation. Of particular importance is the observation that estrogen antagonists not only block the effect of estrogens but also proliferation that is induced by other growth factors, suggesting that the antiestrogen-receptor complex functions as a growth inhibitor in its own right (Berthois et al., 1989; Chalbos et al., 1994). Many different types of genes are implicated in the G, phase of the cell cycle, some of which are targets for estrogens. For example, the products of a number of immediate-early response genes, including members of the Fos and Myc families, which are required for cell cycle progression, are increased in MCF-7 breast cancer cells following estrogen treatment (Wilding et al., 1988; Dubik et al., 1987).The expression of these immediate-early response genes is transient, and it is assumed that they, in turn, regulate the transcription of downstream target genes involved in cell proliferation. One such group are the cyclins and the cyclin-dependent kinases. The D-type cyclins may function at G1 control points, and interestingly, the expression of cyclin D1 mRNA and protein is regulated by steroid hormones in breast cancer cells (Musgrove and Sutherland, 1994). It is often assumed that estrogens stimulate cell proliferation by directly regulating the transcription of one, or a few, of these genes. However, nuclear receptors are also capable of regulating the transcription of genes that lack response elements by modulating the activity of other transcription factors, such as AP-1, as described above. This could be particularly important in growth control, because AP-1 is implicated as a critical target for many signaling pathways that regulate cell differentiation, proliferation, and transformation (Angel and Karin, 1991; Vogt and Bos, 1990).In addition to its activation by phosphorylation, induced by a number of growth factors, the activity of AP-1 can be modulated by nuclear receptors. In some cases AP-1 activity is increased and in other cases it is decreased, the differential response depending on the composition of the AP-1 complex. In MCF-7 breast cancer cells growth factor-induced AP-1 activity is increased by estrogens and reduced by antiestrogens (Philips et al., 1993). Therefore, AP-1 might represent a critical target for the estrogen receptor and play a role in mediating both the mitogenic effects of estrogens and growth inhibition by antiestrogens. This might then explain how antiestrogens are capable of blocking the effect of certain growth factors as well as estradiol.
272
M. G. PARKER
111. INTRACELLULAR LOCALIZATION OF ESTROGEN RECEPTORS Estrogen receptors are located predominantly in the cell nucleus, irrespective of whether they are bound to estradiol. The major nuclear localization signals consist of a series of three basic regions located near the C-terminus of the DNA binding domain, which are presumably exposed even in the absence of estrogen (Picard et al., 1990; Ylikomi et al., 1992). Similarly, the majority of nuclear receptor family members are also found in the nucleus, irrespective of whether they are occupied by ligand; exceptions include the androgen and glucocorticoid receptors, whose nuclear uptake is dependent on hormone binding. While the estrogen receptor is a nuclear protein, recent studies indicate that, in fact, it undergoes a process of nucleocytoplasmic shuttling (Guiochon-Mantel et al., 1991; Dauvois et al., 1993).It appears that all classes of steroid receptors are constantly diffusing into the cytoplasm but are rapidly transported back into the cell nucleus in an energydependent step. Prior to estrogen binding the receptor exists as an inactive oligomeric complex that contains other proteins, including the heat-shock protein hsp90 (Catelli et al., 1985; Sanchez et al., 1985; Redeuilh et al., 1987). Other proteins that have been detected include hsp70, p59 (also called hsp561, a member of the immunophilin family, p54, p50, and p23, but it is not clear whether they are associated directly with the receptor or indirectly by means of their interaction with hsp90 (for reviews see Smith and Toft, 1993; Pratt, 1993). The hsp90 protein appears to be involved in the maintenance of the receptor in an inactive state in the absence of ligand and may also be important for folding of the receptor protein and/or transport across membranes (Smith and Toft, 1993), but the function of other associated proteins is unknown. Analysis of the deletion mutants of the estrogen receptor has shown that the main binding site for hsp90 is the hormone binding domain (Chambraud et al., 1990; Schlatter et al., 1992).Upon hormone binding the complex dissociates, a process referred to as activation or transformation, and the receptor is then capable of either binding directly to DNA as homodimers or interacting with other transcription factors by protein-protein interactions (Fig. 3).
IV. HORMONE BINDINGAND RECEPTOR DIMERIZATION The hormone binding domain, located in a C-terminal region of approximately 250 amino acids, has been characterized by analyzing the
STEROID RECEPTORS
Shultling
273
CYTOPLASM
NUCLEUS
ERE FIG. 3. Mechanism of estrogen receptor action. Estrogen receptors constantly shuttle between the cell nucleus and the cytoplasm, but under steady-state conditions they are predominantly nuclear. Upon hormone binding the receptor, complexed with the heatshock protein hsp90, dissociates to allow dimerization and high-affinity DNA binding to estrogen response elements (EREs). Transcription of genes by RNA polymerase (pol) I1 is initiated from sites approximately 30 bp downstream of a TATA box that binds TFIID and a number of basal transcription factors, including TFIIA, TFIIB, TFIIE, TFIIF, and TFIIH. The TFIID complex comprises a TATA box-binding protein (TBP) and TBPassociated factors (TAFs) and it is postulated that transcriptional activators, including receptors, stabilize the binding of these basal transcription factors either directly or indirectly by means of adapter proteins. R, receptor; L, ligand; ERE, estrogen response element.
activity either of proteolytic fragments of the native protein (Katzenellenbogen et al., 1987) or of recombinant proteins (Kumar et al., 1986; Fawell et al., 1989). The first approach to identifying important residues was to affinity-label the receptor with synthetic ligands. In the human estrogen receptor a cysteine at position 530 (equivalent to position 534 in the mouse receptor) was affinity-labeled with both an estrogen, ketononestrol aziridine, and an antiestrogen, tamoxifen aziridine, suggesting that it is in the vicinity of the ligand binding pocket (Harlow et al., 1989). This residue itself is unlikely to be involved in estrogen
274
M. G. PARKER
binding, however, since it can be mutated without affecting the ability of the receptor to bind estrogen (Fawell et al., 1990). That it is close t o the hormone binding site was supported by examining other mutant mouse receptors, which indicated that amino acids between residues 518 and 525 were required for estrogen binding activity (Danielian et al., 1993). Early biochemical studies suggesting that the estrogen receptor existed as a homodimer (Notides et al., 1981; Miller et al., 1985) were verified by subsequent experiments demonstrating that all steroid hormone receptors bound to response elements as homodimers (Tsai et al., 1988; Kumar and Chambon, 1988). Functional analysis of mutant receptors has indicated that the major dimerization interface is located in the hormone binding domain (Kumar and Chambon, 1988; Fawell et al., 1990a). However, there are sequences within the DNA binding domain that also form a dimer interface when the receptor is bound to DNA (Kumar and Chambon, 1988; Schwabe et al., 1993b), which is discussed later. Some of the residues in the hormone binding domain that are required for dimerization have been identified by functional analysis of mutant receptors and have been found to overlap with residues that are essential for estrogen binding (Fawell et al., 1990). It seems likely, therefore, that the hormone binding pocket is at or near the dimer interface and that hormone binding may result in stabilized dimerization, perhaps on the basis of hydrophobic shielding.
V. TARGET GENERECOGNITION The first target genes to be identified and characterized were those that encode vitellogenin, ovalbumin, and apo-very low-density lipoprotein in oviparous vertebrates. By searching for similar sequences in the 5' flanking regions of these genes, Walker et al., (1984) derived a putative sequence for an estrogen response element. The functional analysis of these and a number of other promoters, including those for rat prolactin (Maurer and Notides, 19871, calbindin D9k (L'Horset et al., 19941, human pS2 (Berry et al., 19891, and cathepsin D (Augereau et al., 19941, confirmed that an inverted repeat of the sequence A/ GGTCA separated by 3 bp was suficient to confer estrogen responsiveness. Such sites, which bind the receptor in the absence of other proteins, are called simple response elements. The binding afinity of receptors for sites varies widely, depending on their precise sequence and number. The majority of estrogen response elements that have been characterized to date contain imperfect inverted repeats that do
STEROID RECEPTORS
275
not function as well as the consensus sequence, but in some promoters there are several such sites that increase the estrogen response. Most remarkable is the discovery that most, if not all, binding sites for the entire family of nuclear receptors consist of different arrangements of half-sites derived from only two consensus sequences (Parker, 1993). The estrogen receptor binds cooperatively to an inverted repeat of the sequence GGTCA in which the half-sites are separated by 3 bp by forming a dimeric complex with DNA (Schwabe et al., 199313). This is the only spacing that allows cooperative binding, since other spacings cause steric interference so that only one DNA binding domain can bind. The DNA binding domain consists of two amphipathic ahelices arranged at about 90" to one another at their midpoints (Schwabe et al., 1993a). The first helix, derived from the first zinc finger motif, comprises the DNA recognition helix, and the second helix, derived from the second zinc finger motif, stabilizes the recognition helix. The dimer interface comprises amino acids located between the two helices and is postulated to be induced upon DNA binding. For a comprehensive review of DNA binding by the estrogen receptor, the reader is referred to the work of Schwabe et al. (199313). In addition to simple response elements consisting of inverted repeats that bind steroid receptors with relatively high affinity, there are likely to be many more elements, referred to as composite response elements (Diamond et al., 19901, which bind receptors only in association with another transcription factor, such as AP-1. To date, few of these composite response elements have been analyzed in detail and no structural information on the interaction of receptors with them is available.
VI. TRANSCRIPTIONAL ACTIVATION Two transcriptional activation functions (AFs) have been defined in the estrogen receptor, using transient transfection experiments (Kumar et al., 1987; Webster et al., 1988; Lees et al., 1989; Tora et al., 1989). The activity of AF-1, located in the N-terminal domain, is constitutive and can stimulate transcription in the absence of hormone when it is fused to a heterologous DNA binding domain; however, in the intact receptor estrogen binding is still required in order to promote DNA binding. AF-2, in the hormone binding domain, is markedly affected by the ligand bound and is still dependent on the binding of estrogen for full activity, even when it is fused to a heterologous DNA binding domain. The activities of AF-1 and AF-2 vary, depending on the responsive promoter and cell type, and in some cases both are required
276
M. G. PARKER
for full transcriptional stimulation (Tora et al., 1989). Inspection of their protein sequence indicates that these activation domains are not composed of a high proportion of acidic amino acids or glutamine residues found in some transcription factors. AF-1, however, is slightly proline rich, and a number of serine residues have been identified in the region as sites of phosphorylation (Ali et al., 1993; Aronica and Katzenellenbogen, 1993; Lahooti et al., 1994). Additionally, based on their ability to cooperate with or inhibit (squelch) other transcription factors when they are overexpressed, it appears that AF-1 and AF-2 are functionally distinct from each other and from acidic activation domains (Tora et al., 1989a; Tasset et al., 1990). Analogous AFs are likely to exist in the majority of nuclear receptors. Previous work has indicated that AF-2 comprises a number of dispersed elements throughout the hormone binding domain, brought together upon estrogen binding (Webster et al., 1989). We have shown that one essential element is located between residues 538 and 552 in the mouse estrogen receptor (Danielian et al., 1992), a region that is absolutely conserved in estrogen receptors from other species and highly conserved in other nuclear receptors. Point mutagenesis of this region showed that hydrophobic residues flanking a conserved glutamic acid are essential for transcriptional activation. Replacement of either pair of hydrophobic residues with alanines abolished transcriptional activation without significantly affecting steroid and DNA binding functions, indicating that these residues are specifically involved in transcriptional activation and in the interaction between AF-1 and AF-2. The negatively charged residues within the putative amphipathic a-helix are important for AF-2 activity, but not its ability to synergize with AF-1. The residues important for transcriptional activation by the mouse estrogen receptor were also shown to be essential for transcriptional activation by the mouse glucocorticoid receptor (Danielian et al., 1992)and the thyroid hormone receptor (Saatcioglu et al., 1993). Given that the putative amphipathic helix is conserved in nuclear receptors, it seems likely that they all contain a functionally equivalent AF-2 domain in which the putative amphipathic helix might interact directly with similar target proteins.
GENETRANSCRIPTION VII. SPECIFIC
BY THE
ESTROGEN RECEPTOR
The regulated transcription of protein coding genes by transcriptional activators involves the assembly of a preinitiation complex (Ptashne, 1988; Mitchell and Tjian, 1989). Numerous basal transcription factors
STEROID RECEPTORS
277
are required for transcriptional initiation by RNA polymerase I1 (Zawel and Reinberg, 1992), and it is believed that these are likely to be targets for the transcriptional activators, including nuclear receptors. Candidates include TFIIB, which has been shown to interact directly with the estrogen receptor (Ing et al., 1992) and the TATA boxbinding protein (TBP) (Sadovsky et al., 1995).On the other hand, since activation of transcription in uitro requires the TFIID complex, it has been proposed that activators might interact specifically with some of the TBP-associated factors (TAFs) (Pugh and Tjian, 1990; Goodrich et al., 1993; Hoey et al., 1993).However, alternative targets are suggested by so-called transcriptional interference or squelching experiments (Tasset et al., 1990; Martin et al., 1990). Since the acidic activation domain of VP16 does not interfere with transcriptional activation by AF-2 but does interfere with AF-1, it seems reasonable to suggest that AF-2, at least, contacts additional proteins, distinct from the basal transcription factors that are rate limiting. The target proteins for the conserved putative amphipathic a-helix are unlikely to be TFIIB or TBP, because they interact with these proteins in the absence of estrogen and the interaction is unaffected by disruptive mutations in AF-2. Recently, a number of candidate target proteins termed receptor-interacting proteins (RIPs) have been identified by far-Western blotting (Cavailles et al., 1994, 1995; Halachmi et al., 1994).Two proteins with molecular weights of 160,000 and 140,000 were shown t o contact the hormone binding domain of the estrogen receptor directly, but only in the presence of 17P-estradio1, not antiestrogens. Moreover, the interaction was abolished when point mutations were introduced into the putative amphipathic a-helix required for estrogen-dependent transcription (Danielian et al., 1992). Thus, in view of the correlation between these interactions and the ability of the hormone binding domain to stimulate transcription, these RIPs are candidates for a role in hormone-dependent transcriptional regulation. In view of the conservation of the amphipathic a-helix, it is conceivable that the RIPs may play a role in the transcriptional activity of the entire nuclear receptor family. ACTION VIII. MECHANISMOF ANTIESTROGEN Antiestrogens have been developed to inhibit the transcriptional activity of the estrogen receptor for a variety of clinical uses. Tamoxifen is the main therapeutic antiestrogen for treating advanced breast cancer and in adjuvant therapy following surgery. The patients showing
278
M. G . PARKER
the greatest benefits are those whose tumors are estrogen receptor positive. Estrogen receptors are easily detectable in at least 60% of primary tumors, and overviews of randomized worldwide clinical trials of tamoxifen therapy indicate that the 5- and 10-year mortality rates of breast cancer patients can be reduced by 20-25% (Bentley et al., 1992). It has been proposed that some patients with tumors that lack the estrogen receptor might also benefit from antiestrogen treatment, but this suggestion is complicated by the fact that receptor contents of less than 10 fmol/mg protein were scored as estrogen receptor negative. In some cases, however, low levels of receptor may still be sufficient to mediate an estrogen response that could be blocked by antiestrogens. The action of tamoxifen is far from straightforward, being a mixed agonist/antagonist. Thus, it is a fairly good antagonist in the mammary gland and effectively inhibits the growth of estrogenresponsive breast cancer cells, but it is an agonist in the endometrium and in bone. The binding site for tamoxifen overlaps with that of the
CYTOPLASM
1 AF-2 disrupted
1
Tamoxifen
A
AF-1
ERE FIG.4. Mechanism of action of antiestrogens with partial agonist activity. In the presence of the partial agonist tamoxifen, receptors dimerize and bind DNA with high affinity. AF-1 in the N-terminal domain is postulated to give rise to agonist activity, while the binding of tamoxifen to the hormone binding domain disrupts the function of AF-2, which accounts for its antagonistic activity.
279
STEROID RECEPTORS
estrogen binding site, so it acts as a competitive inhibitor of estrogen action. Transient transfection experiments suggest that tamoxifen binding allows the receptor to dimerize and bind to DNA with high affinity, but blocks transcriptional activity mediated by the hormone binding domain (Fig. 4). The agonist activity is thought to be derived primarily from AF-1, which appears to function even when tamoxifen is bound to the receptor (Dauvois and Parker, 1993; Berry et al., 1990). Thus, tamoxifen has the potential to act as an antagonist when the transcriptional activity of the receptor is mediated by AF-2, but as an agonist on promoters in which AF-1 is active or DNA binding affinity is weak. Alternative antiestrogens have also been developed with less agonist activity such as ICI 182780 which is reported to be devoid of estrogenic activity (Wakeling, 1992).It is currently being investigated as a potential second line endocrine treatment for tamoxifen-resistant patients. The effectiveness of ICI 182780 as an antiestrogen depends on an al-
CYTOPLASM
Dimerization Defect?
1 ERE
FIG.5. Mechanism of action of pure antiestrogens. The pure antiestrogen prevents nuclear uptake of the estrogen receptor so that during the process of nucleocytoplasmic shuttling the receptor accumulates in the cytoplasm and is degraded. Within the cell nucleus dimerization of the receptor might be disrupted and this would reduce its DNA binding activity. ERE, Estrogen response element.
280
M.G. PARKER
kylamine side chain with an optimum length of 16-18 carbon atoms at the 7a position in the B ring in the steroid. One of the major effects of ICI 182780 is to cause a decrease in the cellular content of receptor protein by markedly reducing its half-life (Gibson et al., 1991; Dauvois et al., 1992). The increased turnover is accompanied by a block in nucleocytoplasmic shuttling (Dauvois et al., 1993) (Fig. 5 ) but the molecular basis of ICI 182780 has yet to be elucidated. In common with other steroid receptors, the estrogen receptor seems to be constantly diffusing out of the cell nucleus, but is rapidly transported back into the nucleus in an energy-dependent step so that the receptor is predominantly nuclear under equilibrium conditions. In the presence of ICI 182780, nuclear reentry does not occur and the protein is degraded in the cytoplasm. An additional effect of ICI 182780 is that it might disrupt dimerization and, as a consequence, inhibit DNA binding. It seems likely that ICI 182780 binds to a site that is similar, if not identical, to that of estradiol, which we have shown overlaps with a region involved in receptor dimerization, so that the antiestrogen, by means of its 7a side chain, might sterically interfere with dimerization. This effect can be demonstrated in uitro,but has not been confirmed in uiuo. Finally, both tamoxifen and ICI 182780 modulate the response of AP-1 to growth factors by a mechanism that involves protein-protein interactions rather than direct DNA binding (Chalbos et al., 1994), but the details of this have not been elucidated.
IX. CROSS-COUPLING WITH OTHERSIGNALING PATHWAYS Recently, the transcriptional activity of the estrogen receptor has been reported to be modulated by a number of other signaling pathways involving different protein kinase cascades. In some cases the effect was estrogen independent, whereas in other cases the alternate pathway modulated the effect of estrogen by either enhancing or inhibiting the transcriptional activity of the receptor. One of the most striking observations is that the effects of estrogen on the mouse uterus could be reproduced by administering EGF to animals (IgnarTrowbridge et al., 1992). EGF was then shown t o stimulate the transcriptional activity of the estrogen receptor in uterine cells in the absence of estrogen and was blocked by treating cells with the pure antiestrogen ICI 164384, indicating that the effects were mediated by the estrogen receptor (Ignar-Trowbridge et al., 1993).EGF acts through several pathways, including protein kinase C and mitogen-activated protein (MAP) kinase, so it is possible that one of these directly phos-
STEROID RECEPTORS
281
phorylates the estrogen receptor. It is noteworthy that a serine in the N-terminal domain of estrogen receptors is a potential MAP kinase site and is phosphorylated after estrogen treatment in intact cells (Ali et al., 1993). Phosphorylation of this serine has been reported to modulate the activity of the receptor in response to estrogen, but it would obviously be interesting to investigate whether it plays a role in EGF responses. The possibility that protein kinase C is involved is difficult to assess, because it has been reported to either increase or antagonize the activity of the receptor, depending on the cell type and the target promotor (Tzukerman et al., 1991; Cho and Katzenellenbogen, 1993). Antagonism appears to be achieved by a combination of mechanisms, including down-regulation of estrogen receptor expression and transrepression of the receptor, itself. IGF-I has also been reported to stimulate the transcriptional activity of the estrogen receptor, but in this case the effect was dependent on the presence of estrogen (Aronica and Katzenellenbogen, 1993). Agents that increase intracellular CAMPlevels and protein kinase A activity also increase the transcriptional activity of the estrogen receptor. This effect was first demonstrated by treating cells with the neurotransmitter dopamine, acting through the dopamine D1 receptor, which was found to stimulate the transcriptional activity of several steroid hormone receptors in the absence of cognate hormone (Power et al., 1991). On the other hand, when protein kinase A was stimulated by treating cells with 8Br-cAMP7the effect was dependent on the presence of estradiol (Cho and Katzenellenbogen, 1993).The substrate for protein kinase A has yet to be identified, but since steroid receptors do not appear to contain a consensus protein kinase A phosphorylation site, it seems likely that the effects of this kinase on transcriptional activity are indirect. Finally, cross-coupling can also occur within the cell nucleus between receptors and other transcription factors whose activity is modulated by different signaling pathways, the best characterized of which is AP-1. Receptors and Fos/Jun family members seem to interact with one another in different ways, depending on the response element. On simple estrogen response elements the transcriptional activity of the estrogen receptor is inhibited by c-Jun and, to some extent, by JunB and by c-Fos (Doucas et al., 1991).The effect of c-Fos but not c-Jun can be overcome by overexpressing the receptor, suggesting that they inactivate the receptor by distinct mechanisms. It is possible that c-Fos interacts with the receptor and interferes with the function AF-1 and/ or AF-2, while c J u n may sequester a protein required by the estrogen receptor to mediate its transcriptional activity. Estrogen receptors also
282
M. G. PARKER
modulate the activity of c-Jun and c-Fos on AP-1 sites (Gaub et al., 1990; Philips et al., 19931, and almost certainly on so-called composite response elements. These elements only bind receptors in the presence of another transcription factor and have been well described only for glucocorticoid receptors, but it would be extremely surprising if they did not also exist for estrogen receptors. Thus, cross-couplingwith other signaling pathways almost certainly involves phosphorylation events. In certain cases the phosphorylation modulates the activity of the receptor itself, while in other cases the phosphorylation regulates the activity of a distinct transcription factor that, in turn, modulates the activity of the receptor. It is difficult now to assess the significance of all of the observations that have been made, but they certainly emphasize that estrogen receptors, and nuclear receptors in general, do not function in isolation but are coupled with many other signaling pathways in target cells. REFERENCES Ali, S., Metzger, D., Bornert, J.-M., and Chambon, P. (1993).Modulation of transcriptional activation by ligand-dependent phosphorylation of the human oestrogen receptor A/B region. EMBO J. 12, 1153-1160. Angel, P., and Karin, M. (1991). The role of Jun, Fos and the AP-1 complex in cellproliferation and transformation. Biochirn. Biophys. Actu 1072, 129-157. Aronica, S. M., and Katzenellenbogen, B. (1993). Stimulation of estrogen receptormediated transcription and alteration in the phosphorylation state of the rat uterine estrogen receptor by estrogen, cyclic adenosine monophosphate, and insulin-like growth factor-1. Mol. Endocrinol. 7, 743-752. Augereau, P., Miralles, F., Cavailles, V., Gaudelet, C., Parker, M., and Rochefort, H. (1994). Characterization of the proximal estrogen-responsive element of human cathepsin D gene. Mol. Endocrinol. 8, 693-703. Bates, S. E., Davidson, N. E., Valverius, E. M., Dicson, R. B., Freter, C. E., Tam, J . P., Kulow, J. E., Lippman, M. E., and Salomon, S. (1988). Expression of tranforming growth factor alpha and its messenger ribonucleic acid in human breast cancer, its regulation by estrogen and its possible functional significance. Mol. Endocrinol. 2, 543-545. Beato, M. (1989). Gene regulation by steroid hormones. Cell 56, 335-344. Bentley, A., Fentiman, I. S., Rubens, R. D., Cuzick, J., Crossley, E., Durrant, K., Harris, A,, Clarke, M., Collins, R., Godwin, J., Gray, R., Greaves, E., Harwood, C., Mead, G., Peto, R., and Wheatley, K. (1992). Systemic treatment of early breast cancer by hormonal, cytotoxic, or immune therapy: 133 randomised trials involving 3 1000 recurrences and 24000 deaths among 75000 women. Part 2. Lancet 339, 71-85. Berry, M., Nunez, A. M., and Chambon, P. (1989). Estrogen-responsive element of the human pS2 gene is a n imperfrectly palindromic sequence. Proc. Nutl. Acud. Sci. U.S.A. 86,1218-1222. Berry, M., Metzger, D., and Chambon, P. (1990).Role of the two activating domains of the oestrogen receptor in the cell-type and promoter-context dependent agonistic activity of the anti-oestrogen 4-hydroxytamoxifen. EMBO J. 9, 2811-2818.
STEROID RECEPTORS
283
Berthois, Y., Dong, X. F., and Martin, P. M. (1989). Regulation of epidermal growth factor-receptor by estrogen and antiestrogen in the human breast cancer cell line MCF7. Biochem. Biophys. Res. Commun. 159, 126-131. Brunner, N., Bronzert, D., Vindelov, L. L., Rygaard, K., Spang-Thomsen, M., and Lippman, M. E. (1989). Effect on growth and cell cycle kinetics of estradiol and tamoxifen on MCF-7 human breast cancer cells grown in vitro and in nude mice. Cancer Res. 49, 1515-1520. Catelli, M. G., Binart, N., Jung, T. I., Renoir, J . M., Baulieu, E. E., Feramisco, J. R., and Welch, W. J . (1985). The common 90-kd protein component of non-transformed “8s’ steroid receptors is a heat-shock protein. EMBO J. 4, 3131-3135. Cavailles, V., Dauvois, S., Danielian, P. S., and Parker, M. G. (1994). Interaction of proteins with transcriptionally active estrogen receptors. Proc. Natl. Acad. Sci. U.S.A. 91, 10009-10013. Cavailles, V., Dauvois, S., L’Horset, F., Lopez, G., Hoare, S., Kushner, P. J., and Parker, M. G. (1995). Nuclear factor RIP 140 modulates transcriptional activation by the estrogen receptor. EMBO J., in press. Chalbos, D., Philips, A,, and Rochefort, H. (1994). Genomic cross-talk between the estrogen receptor and growth factor regulatory pathways in estrogen target tissues. In “Seminars in Cancer Biology” (M. Parker, ed.), Vol. 5, pp. 361-368. Academic Press, San Diego. Chambraud, B., Berry, M., Redeuilh, G., Chambon, P., and Baulieu, E. E. (1990). Several regions of the human estrogen receptor are involved in the formation receptor-heat shock protein 90 complexes. J. Biol. Chem. 265,20686-20691. Cho, H., and Katzenellenbogen, B. S. (1993). Synergistic activation of estrogen receptormediated transcription by estradiol and protein kinase activators. Mol. Endocrinol. 7, 441-452. Clarke, R., Dickson, R. B., and Lippman, M. E. (1991). The role of steroid hormones and growth factors in the control of normal and malignant breast. I n “Nuclear Hormone Receptors” (M. Parker, ed.), pp. 297-319. Academic Press, San Diego. Danielian, P. S., White, R., Lees, J. A., and Parker, M. G. (1992). Identification of a conserved region required for hormone dependent transcriptional activation by steroid hormone receptors. EMBO J. 11, 1025-1033. Danielian, P. S., White, R., Hoare, S. A., Fawell, S. E., and Parker, M. G. (1993). Identification of residues in the estrogen receptor which confer differential sensitivity to estrogen and hydroxytamoxifen. Mol. Endocrinol. 7, 232-240. Dauvois, S., and Parker, M. G. (1993). Mechanism of action of hormone antagonists. In “Steroid Hormone Action” (M. G. Parker, ed.), pp. 161-185. IRL Press, Oxford. Dauvois, S., Danielian, P. S., White, R., and Parker, M. G. (1992). Antiestrogen ICI 164,384 reduces cellular estrogen receptor content by increasing its turnover. Proc. Natl. Acad. Sci. U.S.A. 89, 4037-4041. Dauvois, S., White, R., and Parker, M. G. (1993). The antiestrogen ICI 182780 disrupts estrogen receptor nucleocytoplasmic shuttling. J.Cell Sci. 106, 1377-1388. Diamond, M. I., Miner, J . N., Yoshinaga, S. K., and Yamamoto, K. R. (1990). Transcription factor interactions: Selectors of positive or negative regulation from a single DNA element. Science 249, 1266-1272. Doucas, V., Spyrou, G., and Yaniv, M. (1991). Unregulated expression of c-Jun or c-Fos proteins but not J u n D inhibits oestrogen receptor activity in human breast cancer derived cells. EMBO J. 10, 2237-2245. Dubik, D., Dembinski, T. C., and Shiu, R. P. C. (1987). Stimulation of c-rnyc oncogene expression associated with estrogen-induced proliferation of human breast cancer cells. Cancer Res. 47, 6517-6521.
284
M. G . PARKER
Evans, R. M. (1988). The steroid and thyroid hormone receptor superfamily. Science 240, 889-895. Fawell, S . E., Lees, J. A,, and Parker, M. G. (1989).A proposed consensus steroid-binding sequence-A reply. Mol. Endocrinol. 3, 1002-1005. Fawell, S. E., Lees, J. A., White, R., and Parker, M. G. (1990). Characterization and colocalization of steroid binding and dimerization activities in the mouse estrogen receptor. Cell 60, 953-962. Gibson, M. K., Nemmers, L. A., Beckman Jr., W. C., Davis, V. L., Curtis, S. W., and Korach, K. (1991) The mechanism of ICI 164,384 antiestrogenicity involves rapid loss of estrogen receptor in uterine tissue. Endocrinology 129, 2000-2010. Gaub, M.-P., Bellard, M., Scheuer, I., Chambon, P., and Sassone-Corsi, P. (1990).Activation of the ovalbumin gene by the estrogen receptor involves the Fos-Jun complex. Cell 63, 1267-1276. Goodrich, J . A,, Hoey, T., Thut, C. J., Admon, A,, and Tjian, R. (1993). Drosophila TAFl140 interacts with both a VP16 activation domain and the basal transcription factor TFIIB. Cell 75, 519-530. Guiochon-Mantel, A,, Lescop, P., Christin-Maitre, S., Loosfelt, H., Perrot-Applanat, M., and Milgrom, E. (1991). Nucleocytoplasmic shuttling of the progesterone receptor. EMBO J . 10, 3851-3859. Halachmi, S., Marden, E., Martin, G., MacKay, H., Abbondanza, C., and Brown, M. (1994). Estrogen receptor-associated proteins-Possible mediators of hormoneinduced transcription. Science 264, 1455-1458. Harlow, K. W., Smith, D. N., Katzenellenbogen, J . A,, Green, G. L., and Katzenellenbogen, B. S. (1989).Identification of cysteine 530 as the covalent attachment site of an affinity-labelling estrogen (ketononestrol aziridine) and antiestrogen (tamoxifen aziridine) in the human estrogen receptor. J . Biol. Chem. 164, 17476-17485. Hoey, T., Weinzieri, R. 0. J., Gill, G., Chen, J.-L., Dynlacht, B. D., and Tjian, R. (1993). Molecular cloning and functional analysis of Drosophila TAFllO reveal properties expected of coactivators. Cell 72, 247-260. Ignar-Trowbridge, D. M., Nelson, K. G., Bidwell, M. C., Curtis, S. W., Washburn, T. F., McLachlan, J. A,, and Korach, K. S. (1992). Coupling of dual signaling pathways: Epidermal growth factor action involves the estrogen receptor. Proc. Natl. Acad. Sci. U.S.A.89,4658-4662. Ignar-Trowbridge, L). M., Teng, C. T., Ross, K. A., Parker, M. G., Korach, K. S., and McLachlar (1993). Peptide growth factors elicit estrogen receptor dependent transcriptional activation of a n estrogen responsive element. Mol. Endocrinol. 7, 992998. Ing, N. H., Beekman, J. M., Tsai, S. Y., Tsai, M.-J., and OMalley, B. W. (1992). Members of the steroid hormone receptor superfamily interact with TFIIB (S300-11).J . Biol. Chem. 267, 17617-17623. Katzenellenbogen, B. S., Elliston, J. F., Monsma, F. J., Springer, P. A,, Ziegler, Y. S., and Greene, G. L. (1987). Structural analysis of covalently labeled estrogen receptors by limited proteolysis and monoclonal antibody reactivity. Biochemistry 26, 23642373. Keaveney, M., Klug, J., Dawson, M. T., Nestor, P. V., Neilan, J. G., Forde, R. C., and Gannon, F. (1991). Evidence for a previously unidentified upstream exon in the human oestrogen receptor gene. J. Mol. Endocrinol. 6, 111-115. Knabbe, C., Lippman, M. E., Wakefield, L. M., Flanders, K. C., Kasid, A., Derynck, R., and Dickson, R. B. (1987). Evidence that transforming growth factor-p is a hormonally regulated negative growth factor in human breast cancer cells. Cell 48, 417428.
STEROID RECEPTORS
285
Krust, A., Green, S., Argos, P., Kumar, V., Walter, P., Bornert, J. M., and Chambon, P. (1986). The chicken oestrogen receptor sequence: Homology with v-erbA and the human oestrogen and glucocorticoid receptors. EMBO J. 5, 891-897. Kumar, V., and Chambon, P. (1988). The estrogen receptor binds tightly to its responsive element as a ligand-induced homodimer. Cell 55, 145-156. Kumar, V., Green, S., Staub, A., and Chambon, P. (1986). Localisation of the oestradiolbinding and putative DNA-binding domains of the human oestrogen receptor. EMBO J. 5,2231-2236. Kumar, V., Green, S., Stack, G., Berry, M., Jin, J. R., and Chambon, P. (1987).Functional domains of the human estrogen receptor. Cell 51, 941-951. Lahooti, H., White, R., Danielian, P. S., and Parker, M. G. (1994). Characterization of ligand-dependent phosphorylation of the estrogen receptor. Mol. Endocrinol. 8,182188. Lees, J . A,, Fawell, S. E., and Parker, M. G. (1989). Identification of constitutive and steroid-dependent transactivation domains in the mouse oestrogen receptor. J . Steroid Biochem. 34, 33-39. L’Horset, F., Blin, C., Colnot, S., Lambert, M., Thomasset, M., and Perret, C. (1994). Calbindin-d9k gene expression in the uterus-Study of the 2 messenger-ribonucleicacid species and analysis of an imperfect estrogen-responsive element. Endocrinology 134, 11-18. Lubahn, D. B., Moyer, J . S., Golding, T. S., Couse, J. F., Korach, K. S., and Smithies, 0. (1993). Alteration of reproductive function but not prenatal sexual development after insertional disruption of the mouse estrogen receptor gene. Proc. Natl. Acad. Sci. U.S.A.90, 11162-11166. Martin, K. J., Lillie, J. W., and Green, M. R. (1990). Evidence for interaction of different eukaryotic transcriptional activators with distinct cellular targets. Nature 346, 147-152. Maurer, R. A,, and Notides, A. C. (1987). Identification of a n estrogen-responsive element from the 5’-flanking region of the rat prolactin gene. Mol. Cell. Biol. 7,42474254. Miller, M. A., Mullick, A., Greene, G. L., and Katzenellenbogen, B. S. (1985).Characterization of the subunit nature of nuclear estrogen receptors by chemical cross-linking and dense amino acid labeling. Endocrinology 117, 515-522. Mitchell, P. J., and Tjian, R. (1989). Transcriptional regulation in mammalian cells by sequence-specific DNA binding proteins. Science 245, 371-378. Musgrove, E. A., and Sutherland, R. L. (1994). Cell cycle control by steroid hormones. In “Seminars in Cancer Biology” (M. Parker, ed.), Vol. 5, pp. 381-389. Academic Press, San Diego. Notides, A. C., Lerner, N., and Hamilton, D. E. (1981). Positive cooperativity of the estrogen receptor. Proc. Natl. Acad. Sci. U.S.A.78, 4926-4930. Parker, M. G. (1993). Steroid and related receptors. Curr. Opin. Cell Biol. 5, 499-504. Philips, A., Chalbos, D., and Rochefort, H. (1993). Estradiol increases and anti-estrogens antagonize the growth factor-induced activator protein-1 activity in MCF7 breast cancer cells without affecting c-fos and c j u n synthesis. J. Biol. Chem. 268, 1410314108. Picard, D., Kumar, V., Chambon, P., and Yamamoto, K. R. (1990).Signal transduction by steroid hormones: Nuclear localization is differentially regulated in estrogen and glucocorticoid receptors. Cell Regul. 1, 291-299. Power, R. F., Mani, S. K., Codina, J., Conneely, 0. M., and O’Malley, B. W. (1991). Dopaminergic and ligand-independent activation of steroid hormone receptors. Science 254, 1636-1639.
286
M. G. PARKER
Pratt, W. B. (1993). Role of heat-shock proteins in steroid receptor function. In “Steroid Hormone Action” (M. G. Parker, ed.), pp. 64-93. IRL Press, Oxford. Ptashne, M. (1988). How eukaryotic transcriptional activators work. Nature 335, 683689. Pugh, B. F., and Tjian, R. (1990). Mechanism of transcriptional activation by Spl: Evidence for coactivators. Cell 61, 1187-1197. Redeuilh, G., M. B., S. C., and B. E.-E. (1987). Subunit composition of the molybdatestabilised “8-9 S” nontransformed estradiol receptor purified from calf uterus. J. Biol. Chem. 262,6969-6975. Roberts, A. B., and Sporn, M. B. (1992). Mechanistic interrelationships between two superfamilies: The steroid/retinod receptors and transforming growth factor-p. In “Cancer Surveys” (M. G. Parker, ed.), Vol. 14, pp. 205-220. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Saatcioglu, F., Bartunek, P., Deng, T., Zenke, M., and Karin, M. (1993). A conserved C-terminal sequence that is deleted in v-ErbA is essential for the biological activities of c-ErbA (the thyroid hormone receptor). Mol. Cell. Biol. 13, 3675-3685. Sadovsky, Y., Webb, P., Lopez, G., Baxter, J . D., Cavailles, V., Parker, M. G., and Kushner, P. J . (1994). Transcriptional activators differ in their response to overexpression of TBP. Mol. Cell. Biol. in press. Sanchez, E. R., Toft, D. O., Schlesinger, M. J., and Pratt, W. B. (1985). Evidence that the 90-kDa phosphoprotein associated with the untransformed L-cell glucocorticoid receptor is a murine heat shock protein. J . Biol. Chem. 260, 12398-12401. Schlatter, L. K., Howard, J . K., Parker, M. G., and Distelhorst, C. W. (1992). Comparison of the 90-kilodalton heat shock protein interaction with in vitro translated glucocorticoid and estrogen receptors. Mol. Endocrinol. 6, 132-140. Schwabe, J. W. R., Chapman, L., Finch, J. T., and Rhodes, D. (1993a). The crystalstructure of the estrogen-receptor DNA-binding DNA-How receptors discriminate between their response. Cell 75, 567-578. Schwabe, J . W. R., Chapman, L., Finch, J. T., Rhodes, D., and Neuhaus, D. (1993b). DNA recognition by the estrogen-receptor-From solution to the crystal. Structure 1, 187-204. Smith, D. F., and Toft, D. 0. (1993). Steroid-receptors and their associated proteins. Mol. Endocrinol. 7 , 4-11. Stewart, A. J., Johnson, M. D., May, F. E. B., and Westley, B. R. (1990). Role of insulinlike growth factors and the type I insulin-like growth factor receptor in the estrogenstimulated proliferation of human breast cancer cells. J. Biol. Chem. 265, 2117221178. Tasset, D., Tora, L., Fromental, C., Scheer, E., and Chambon, P. (1990). Distinct classes of transcriptional activating domains function by different mechanisms. Cell 62, 1177-1 187. Tora, L., White, J., Brou, C., Tasset, D., Webster, N., Scheer, E., and Chambon, P. (1989). The human estrogen receptor has two independent nonacidic transcriptional activation functions. Cell 59, 477-487. Tsai, S. Y., Carlstedt-Duke, J., Weigel, N. L., Dahlman, K., Gustafsson, J.-A., Tsai, M.-J., and O’Malley, B. W. (1988). Molecular interactions of steroid hormone receptor with its enhancer element Evidence for receptor dimer formation. Cell 55, 361-369. Tzukerman, M., Zhang, X.-K., and Pfahl, M. (1991). Inhibition of estrogen receptor A molecular activity by the tumor promoter 12-0-tetradecanoylphorbol-13-acetate: analysis. Mol. Endocrinol. 5, 1983-1992. Vogt, P. K., and Bos, T. J . (1990). Jun: Oncogene and transcription factor. Adu. Cancer Res. 55, 1-35.
STEROID RECEPTORS
287
Wahli, W. (1988). Evolution and expression of vitellogenin genes. Trends Genet. 4,227232. Wakeling, A. E. (1992). Steroid antagonists as nuclear receptor blockers. In “Cancer Surveys” (M. G. Parker, ed.), Vol. 14, pp. 71-85. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Walker, P., Germond, J. E., Brown, L. M., Givel, F., and Wahli, W. (1984). Sequence homologies in the region preceding the transcription initiation site of the liver estrogen-responsive vitellogenin and apo-VLDLII genes. Nucleic Acids Res. 12, 8611-8626. Webster, N. J . G., Green, S., Jin, J . R., and Chambon, P. (1988). The hormone-binding domains of the estrogen and glucocorticoid receptors contain an inducible transcription activation function. Cell 54, 199-207. Webster, N. J., Green, S., Tasset, D., Ponglikitmongkol, M., and Chambon, P. (1989). The transcriptional activation function located in the hormone-binding domain of the human oestrogen receptor is not encoded in a single exon. EMBO J. 8,1441-1446. White, R., Lees, J . A,, Needham, M., Ham, J., and Parker, M. (1987).Structural organization and expression of the mouse estrogen receptor. Mol. Endocrinol. 1, 735-744. Wilding, G., Lippman, M. E., and Gelmann, E. P. (1988). Effects of steroid hormones and peptide growth factors on proto-oncogene c-fos expression in human breast cancer cells. Cancer Res. 48, 802-805. Ylikomi, T., Bocquel, M. T., Berry, M., Gronemeyer, H., and Chambon, P. (1992). Cooperation of proto-signals for nuclear accumulation of estrogen and progesterone receptors. EMBO J. 11, 1-14. Zawel, L. and Reinberg, D. (1992). Advances in RNA polymerase I1 transcription. Curr. Opin. Cell Biol. 4, 488-495.
This Page Intentionally Left Blank
VITAMINS AND HORMONES, VOL. 51
Phosphorylation and Steroid Hormone Action
WENLONG BAI AND NANCY L. WEIGEL Department of Cell Biology Baylor College of Medicine Houston, Texas 77030
I. Introduction 11. Phosphorylation of Steroid Hormone Receptors A. Structure and Function of Steroid Hormone Receptors: An Overview B. Phosphorylation of Steroid Hormone Receptors 111. Regulation of Steroid Hormone Receptors by Phosphorylation A. In Vitro Studies B. Mutational Analysis IV. Interaction between Steroid Hormone Action and Signal Transduction Pathways A. Regulation of Receptor Activity by Modulators of Signal Transduction Pathways B. Activation of Signal Transduction Pathways by Steroids V. Summary: A Model of Steroid Hormone Action References
I. INTRODUCTION For many years steroid hormones and peptide hormones have been considered to act via distinctly different mechanisms. Peptide hormones, like many growth factors and neurotransmitters, act through membrane receptors. Through signal transduction pathways the transient signals generated by binding of peptide hormone to receptors at the cell surface are converted into long-term changes in gene expression in the nucleus. The signal transduction pathways are cascades of sequential phosphorylation/dephosphorylation reactions catalyzed by kinases and phosphatases. They often start with the activation of the membrane receptors, which are frequently kinases or phosphatases, and end with phosphorylation or dephosphorylation of specific growthregulated transcription factors. Obviously, phosphorylation plays a key role in peptide hormone action. In contrast to peptide hormones, steroid hormones act through intracellular receptors that are not associated with the cell membrane. It is believed that the lipophilic hormones pass through the cell membrane without the help of any membrane-associated protein, bind the receptors in the cell, and transform them into active transcription factors. In 289
Copyright B) 1995 by Academic Press, Inc. All rights of reproduction in any form reserved.
290
WENLONG BAI AND NANCY L. WEIGEL
this model the signal carried by steroid hormones is directly transduced into long-term changes in gene expression without a requirement for phosphorylation as an intermediate step. Studies in recent years have provided increasing evidence that the mechanism of steroid hormone action is not as distinctly different from that of peptide hormone action as was believed previously. In this chapter the phosphorylation of steroid hormone receptors is reviewed and those functional studies that have established the significance of phosphorylation and signal transduction processes in steroid hormone action are discussed.
11. PHOSPHORYLATION OF STEROID HORMONE RECEPTORS A. STRUCTURE AND FUNCTION OF STEROID HORMONE RECEPTORS: AN OVERVIEW Members of the steroid hormone receptor superfamily (Evans, 1988; Beato, 1989; O’Malley, 1990; Green and Chambon, 1988; O’Malley et al., 1969; Tsai and O’Malley, 1994) are ligand-regulated transcription factors. These include not only receptors for steroids (e.g., progestins, estrogens, androgens, and corticosteroids) and nonsteroid ligands (vitamin D,, thyroid hormones, and retinoids) but also orphan receptors (O’Malley and Conneely, 1992) whose ligands remain unidentified (e.g., Nur77) or whose ligand-binding property has been lost [e.g., thyroid receptor a2 (TRa2)I. Members of the superfamily regulate multiple biological processes ranging from development and morphogenesis to reproduction and behavior. Based on the mechanisms of action, the receptor superfamily can be further divided into group A (steroid hormone receptors) and group B (nonsteroid hormone receptors) (Tsai and O’Malley, 1994). Group A receptors form complexes with heat-shock proteins in the absence of hormone. After hormone binding the receptors dissociate from some of the heat-shock proteins and are transformed into complexes in which the receptors dimerize, bind DNA, and activate or repress the transcription of target genes. In the absence of hormone, group B receptors usually bind to DNA. Some members of group B receptors inhibit or silence the basal transcription of the target genes in the absence of hormone (Baniahmad et al., 1990; Damm et al., 1989). Hormone binding affects the function of the receptor by either enhancing the origi-
PHOSPHORYLATION AND STEROID HORMONE ACTION
291
nal silencing activity of the receptor or transforming the receptor into a transcriptional activator. In contrast to their functional diversity, steroid hormone receptors share a common structural organization, indicating that they are evolutionarily derived from a common ancestor. The receptors contain highly conserved DNA binding domains, a less conserved C-terminal hormone binding domain, and an N-terminal A/B region, which varies the most both in sequence and in length. Two transcriptional activation functions, AF-1 and AF-2, have been found in the A/B region and the hormone binding domain, respectively (Tora et al., 1989; Kumar et al., 1987; Webster et al., 1988; Lees et al., 1989). They can activate the transcription of target genes separately and/or synergistically in a promoter- and/or cell type-specific manner. Particularly, the function of the AF-1 in the A/B region of the receptors is highly cell type specific (Tora et al., 1989; Folkers et al., 1993).
B. PHOSPHORYLATION OF STEROID HORMONE RECEPTORS 1. Identification of Steroid Receptors as Phosphoproteins That steroid receptors are phosphoproteins was first suggested by metabolic studies showing that the hormone binding capacity of glucocorticoid receptor (GR) rises and falls with cellular ATP levels (Munck and Brinck-Johnson, 1968). Other studies showed that some steroid receptors, such as progesterone receptor (PR), exhibit a decreased mobility on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDSPAGE) in response to in uiuo hormone treatment and that this mobility shift can be reversed by the treatment of the receptor with phosphatases (Denner et al., 1990a; Horwitz et al., 1985; Sullivan et al., 1988). The conclusive evidence that receptors are phosphoproteins comes from the isolation of 32P-labeled receptors from cells and tissue minces treated with [32P]orthophosphate. Using this technique, steroid hormone receptors for progesterone (Denner et al., 1990a; Sheridan et al., 1989; Sullivan et al., 1988), glucocorticoid (Housley and Pratt, 1983; Mendel et al., 1987), estrogen (Migliaccio et al., 1984, 1986; Washburn et al., 19911, and androgen (van Laar et al., 1990; Kemppainen et al., 1992), the vitamin D receptor (Pike and Sleator, 19851, thyroid receptor (TR) and v-erbA (Goldberg et al., 19881, a retinoic acid receptor (FbchetteEgly et al., 19921, and orphan receptors such as NGFI-B (or Nur77) (Fahrner et al., 1990) were shown to be phosphoproteins. Phosphoamino acid analysis showed that the majority of the phosphorylation is on
292
WENLONG BAI AND NANCY L. WEIGEL
serine and threonine residues (Denner et al., 1990a; Washburn et al., 1991; Hoeck et al., 1989; Bodwell et al., 19911, and tyrosine phosphorylation has been reported only on estrogen receptors (ERs) (Migliaccio et al., 1986) and retinoic acid receptors (Rochette-Egly et al., 1992). 2. Identification of i n Viuo Phosphorylation Sites Most steroid receptors are phosphorylated in uiuo at multiple sites, suggesting that phosphorylation may play multiple roles in receptor function. In order to understand the role of phosphorylation at specific sites, the individual sites must first be identified. Two approaches have been used to identify phosphorylation sites. One can be viewed as biochemical and the other, molecular biological. The biochemical approach starts with the labeling of the cell or tissue mince with [32Plorthophosphate, followed by purification of the labeled receptor, typically using a specific antibody. The purified receptor is then digested with specific proteases such as trypsin and the resulting phosphopeptides are separated by one or more steps, typically using high-performance liquid chromatography. The site is then identified by direct sequencing of the phosphopeptide. Using this approach, multiple phosphorylation sites have been identified in the chicken PR (cPR) (Denner et al., 1990a) and the mouse GR (Bodwell et al., 1991). The molecular biological approach first localizes the phosphorylation to smaller regions by comparing the phosphorylation level of receptors with deletions of different regions. Then, specific sites are identified by mutating the potential site based on the consensus sequence of known kinases and by demonstrating a loss of phosphorylation. Both approaches have advantages and disadvantages. The biochemical approach relies on direct sequencing and, although the result is reliable, i t is time consuming and requires the use of large quantities of [32Plorthophosphate and reasonable expression levels of the receptor. The molecular biological approach avoids the use of such large quantities of 32P and does not require high expression levels of the receptor, but it is an indirect approach and mutation of an amino acid can sometimes change the structure of a protein such that phosphorylation of a site different from the mutated amino acid is altered. It has been reported (Hilliard et al., 1994) that the mutation of the major hormone-dependent phosphorylation site, Serzo5,in vitamin D3 receptor gave rise to a receptor that continued to be phosphorylated in a hormone-dependent manner on an alternative site. In the case of the human ER (hER), using a molecular biological approach, Ali et al.
PHOSPHORYLATION AND STEROID HORMONE ACTION
293
(1993) identified Serlls as the major hormone-dependent phosphorylation site in the receptor in COS-1 cells, but Le Goff et al. (1994) identified Serlls, as well as Serlo4and SerlO6,as the major hormone-regulated phosphorylation sites in the same cells. In contrast, using a biochemical approach, Arnold et al. (1994) identified Ser167 as the major hormoneinduced phosphorylation site in hER in MCF-7 cells. Whether the discrepancy is because of the different cells or different approaches used in the studies remains to be resolved. The phosphorylation sites identified in different steroid receptors have been described in previous reviews (Orti et al., 1992; Weigel, 1994). Here, the phosphorylation of cPR is used as an example to highlight some of the common features of steroid receptor phosphorylation. cPR is naturally expressed as two forms: cPRB and cPRA, which lacks the N-terminal 128 amino acids of cPRB. Four phosphorylation sites have been identified in cPR purified from the chicken oviduct (Denner et al., 1990a; Poletti and Weigel, 1993) as well as the receptors expressed in yeast (Poletti et al., 1993). The same pattern of phosphorylation was detected for both cPR, and cPRB. All four sites are in Ser-Pro motifs and they are the only Ser-Pro motifs in the receptors. Based on the numbering of cPRB, they are Ser211, Ser260, Ser367, and Ser530. Among them, Ser211and Ser260 are phosphorylated basally, but their phosphorylation is enhanced in response to hormone stimulation. Ser367 and Ser530 are phosphorylated primarily in response to hormone treatment. The kinaseb) that phosphorylates these sites remains unknown. Several features of cPR phosphorylation are common to the members of the steroid hormone receptor superfamily, especially the group A receptors. First, the phosphorylation sites can be basal, liganddependent, or DNA-dependent. Second, the majority of the phosphorylation sites identified so far are Ser-Pro motifs. Third, the majority of the sites are in the A/B region, and phosphorylation within hormone binding or DNA binding domains appears to be rare. 3. Kinases and Receptor Phosphorylation Steroid receptors have been reported previously to be kinases capable of autophosphorylation (Miller-Diener et al., 1985; Garcia et al., 19831, but the phosphorylation was the result of kinases copurifying with the receptor (Garcia et al., 1986; Perisic et al., 1987). Subsequent studies have shown that steroid receptors can be phosphorylated in uitro by several known kinases. In most cases it is unknown whether these kinases phosphorylate the receptors in uiuo.
294
WENLONG BAI AND NANCY L. WEIGEL
a. Mitogen-Actiuated Protein (MAP) Kinases and cdks. MAP kinases (Mulkhopadhyay et al., 1992; Boulton et al., 1991) and cdks (Elledge et al., 1992; Moreno and Nurse, 1990) are two classes of kinases, both of which contain a Ser-Pro motif in their consensus sequences for phosphorylation. Interestingly, the majority of the in uiuo phosphorylation sites identified in the group A receptors are Ser-Pro motifs. MAP kinases play important roles in signal transduction pathways downstream of ras and relay the signals generated by growth factors and peptide hormones from the membrane. The cdks are cyclindependent kinases which control the cell division cycle. Little is known about the kinases that phosphorylate Ser-Pro sites in cPR and hER. It was reported that a fragment of GR can be phosphorylated by a proline-directed kinase (Mazer et al., 1990).I n uitro, hPR was shown to be a very good substrate for cdk2 complexed with cyclin A (Zhang et al., 199413). Interestingly, GR was shown to be active only in specific phases of the cell cycle (Hsu et al., 1992).These data suggest that cell cycle-dependent kinases may regulate the activity of steroid receptors. b. Casein Kinase 11 (CKII). Like MAP kinase, CKII (Pinna, 1990; Tuazon and Traugh, 1991) is also an important component of signaling pathways that control the growth and division of the cells. In particular, CKII has been shown to phosphorylate, and in several cases regulate, the activity of a variety of nuclear proteins, mainly transcription factors and nuclear oncoproteins (Pinna, 1990; Tuazon and Traugh, 1991). Sites fitting the consensus sequence for CKII have been shown to be phosphorylated in uiuo in both group A and B receptors, including hPR (Zhang et al., 1994~1,hER (Arnold et al., 19941, GR (Bodwell et al., 1991), TR (Glineur et al., 19891, and the vitamin D receptor (Jurutka et al., 1993). Ser81, one of the in uiuo hPR-B-specific phosphorylation sites, has been shown to be specifically phosphorylated by CKII in uitro (Zhang et al., 1994~).Although the hPR contains 11 potential CKII sites, CKII phosphorylated purified hPR-B in uitro only at Sersl, suggesting that Sersl may be an authentic site for CKII in uiuo. In addition, Ser167 of hER (Arnold et al., 1994).one of the sites in TR (Glineur et al., 1989), and a site in the vitamin D receptor (Jurutka et al., 1993) have been reported to be phosphorylated in uitro by CKII. c. DNA-PK. DNA-PK (Carter and Anderson, 1991; Anderson and Lees-Miller, 1992) is a nuclear serinehhreonine kinase that requires double-stranded DNA for its activity and phosphorylates hsp90 and a number of DNA-binding proteins, such as Spl and the C-terminal domain of the large subunit of RNA polymerase 11.
PHOSPHORYLATION AND STEROID HORMONE ACTION
295
There is some evidence for DNA-dependent phosphorylation of steroid receptor. For example, hPR (Takimoto et al., 1992) does not undergo the second-round phosphorylation, which results in the reduced mobility in SDS-PAGE when a DNA binding-deficient mutant receptor is analyzed. Both cPR (Weigel et al., 1992) and hPR (Bagchi et al., 1992) are phosphorylated in a DNA-dependent manner during cell-free transcription in HeLa cell nuclear extracts. cPR is also phosphorylated in uitro by a DNA-PK highly purified from HeLa cell nuclear extracts (Weigel et al., 1992). In both cases the DNA-dependent phosphorylation resulted in reduced mobility of the receptor on SDS-PAGE. Whether DNA-PK phosphorylates PR in uiuo remains to be determined. d. Protein Kinase A (PKA) and Protein Kinase C (PKC). Both PKA and PKC are major mediators of signal transduction pathways that convert a transient signal generated on the cell surface into long-term changes in gene expression by modulating the activities of transcription factors (Karin, 1992; Karin and Smeal, 1992; Pennypacker et al., 1994). So far, no convincing evidence has been found for the direct phosphorylation of group A receptors by either PKA or PKC in uiuo. However, there is evidence that PKA phosphorylates TR. The phosphorylation of TRa increases about 10-fold after activation of PKA. A peptide of TRa which is phosphorylated in uiuo was phosphorylated by PKA in uitro (Goldberg et al., 1988). TR is also a substrate for PKC in uitro, and activators of PKC increase the phosphorylation of TRa (Goldberg et al., 1988). The phosphorylation and activity of the vitamin D receptor (Hsieh et al., 1991) are also increased by the activation of PKC. However, the direct phosphorylation of the vitamin D receptor at Ser51 in uitro inhibits the binding of the receptor to DNA (Hsieh et al., 1991). e. pp90rsk. The 90-kDa ribosomal protein kinase (Erikson, 1991), ppSOrsk, is a member of a cascade of growth-regulated protein kinases. Ser354 of Nur77 has been shown to be phosphorylated in uitro by pp90rsk (Davis et al., 1993). The phosphorylation of Ser354 in uiuo has been shown to be induced by 12-O-tetradecanoylphorbol-13-acetate (TPA) and inhibited by adrenocorticotropic hormone (ACTH) (Erikson, 1991). pp90rsk is activated by TPA (Davis and Lau, 19941, consistent with the possibility that pp90rsk might phosphorylate Nur77 a t Ser354 in uiuo. f. Other Kinases. A tyrosine kinase that copurifies with ER and is stimulated by Ca2+/camoddin has been described (Castoria et al., 1993) and has been shown to phosphorylate hER on Vr537. Two sites
296
WENLONG BAI AND NANCY L. WEIGEL
identified in uiuo in GR (Bodwell et al., 1991) do not contain consensus sequences for any of the kinases described above, suggesting the involvement of other kinases in steroid receptor phosphorylation.
111. REGULATION OF STEROID HORMONE RECEPTORS BY PHOSPHORYLATION Two approaches have been used to address the significance of steroid receptor phosphorylation. The first addresses the significance of the receptor phosphorylation by altering the overall level of the phosphorylation and assaying resulting activity in in uitro assays. The second approach addresses the significance of phosphorylation at specific sites by site-directed mutagenesis.
A. IN VITRO STUDIES 1. Hormone Binding While in uitro studies with GR and PR generated no evidence for the role of receptor phosphorylation in hormone binding, several studies have suggested the involvement of phosphorylation in the hormone binding of ER. Treatment of chicken ER in uitro with phosphatase or ATP alters the affinity of cER for its ligand (Raymoure et al., 1985, 1986; Dayani et al., 19901, resulting in the formation of three distinct forms. Studies with hER (Castoria et al., 1993) showed that treatment with a nuclear phosphatase results in the loss of hormone binding ability and that treatment with a Ca2+/calmodulin-stimulated kinase that phosphorylates tyrosine residues restores hormone binding activity. 2. DNA Binding The regulation of DNA binding by phosphorylation has been reported for a number of steroid receptors. It appears that phosphorylation, in general, enhances the DNA binding activity of group A receptors but may inhibit that of group B receptors. cPR isolated from hormone-treated oviduct minces has been shown to bind more tightly to DNA cellulose than the basally phosphorylated receptor, suggesting that additional phosphorylation enhances DNA binding (Denner et al., 1989). Treatment of T47D cells increases both the level of phosphorylation and the binding of hPR to specific DNA
PHOSPHORYLATION AND STEROID HORMONE ACTION
297
sequences, as judged by gel retardation assays (Beck et al., 1992). In contrast, DNase I footprinting and filter-binding assays detected no differences in the binding of the hyperphosphorylated and control rabbit PR (Bailly et al., 1986). The treatment of ER with potato acid phosphatase resulted in a decrease in the receptor's affinity for specific DNA sequences, as judged by a gel shift assay (Denton et al., 1991). In contrast, in uitro phosphorylation of the vitamin D receptor by PKC causes phosphorylation of Ser51 and reduces the binding of the receptor to DNA (Hsieh et al., 1991). There is also evidence that the DNA binding of TRa2 is negatively regulated by phosphorylation. The C terminus of TRa2 contains a cluster of nine serines and has been shown to be phosphorylated in uiuo when expressed in JEG-3 cells. Truncation of the receptor in this region or phosphatase treatment enhances the binding affinity of the receptor to a thyroid response element in uitro (Katz and Lazar, 1994). The phosphorylation of Ser354 in Nur77 by PKA in uitro resulted in a decreased afinity of the orphan receptor for DNA, as determined by a gel mobility shift assay (Hirata et al., 1993).ACTH treatment has been shown to cause the hypophosphorylation of Nur77, and only Nur77 isolated from ACTH-treated Y-1cells is capable of binding to its response element (Davis and Lau, 19941, indicating that the Ser354phosphorylation negatively regulates the DNA binding affinity under physiological conditions. 3. Protein-Protein Interaction Several studies suggest that phosphorylation affects the dimerization of the receptors. To date, no role of phosphorylation in interaction with other proteins has been reported. Coimmunoprecipitation experiments (DeMarzoet al., 1991)have shown that the hyperphosphorylated hPR-A, which has a slower migration rate in SDS-PAGE, binds more tightly to the hPR-B than does the basally phosphorylated hPR-A. Bhat et al. (1994) demonstrated that in uitro phosphorylation of hTRp expressed in Escherichia coli with crude HeLa cytosolic extract enhanced the heterodimerization of the receptor with retinoic acid X receptor (RXR),as judged by coimmunoprecipitation and gel shift assays. Dephosphorylation of the phosphorylated receptor with alkaline phosphatase led to loss of the ability of hTRPl to form a heterodimer with RXRp in either the absence or presence of the DNA. However, Sugawara et al. (1994) performed similar analyses and concluded that the phosphorylation increases the binding of the hTR homodimer to DNA but not the heterodimerization with RXR.
298
WENLONG BAI AND NANCY L. WEIGEL
B. MUTATIONAL ANALYSIS Mutational analysis of specific phosphorylation sites in a number of receptors, using reporter genes in transient cotransfection assays, has provided substantial evidence for the significance of phosphorylation in regulating the transcriptional activity of the receptors. This is an end point assay in that any defect in the receptor activation process will be reflected in the change in transcriptional activity of the receptor. 1. cPR The mutation of the first hormone-dependent phosphorylation site identified in cPR, Ser530 in the hinge region, to alanine (Bai et al., 1994) resulted in the reduced transcriptional activity of the receptor at low hormone concentrations but did not affect the maximal activity of the receptor at saturating levels of hormone, suggesting that the phosphorylation at Ser530 facilitates the response of the receptor to its ligand. The decreased sensitivity of the mutant receptor toward its ligand is not due to a decrease in hormone binding affinity, as judged by both in uztro and whole-cell hormone-binding assays, leading t o the hypothesis proposed by the investigators that Ser530 phosphorylation stabilizes the receptor in its active form, perhaps either by preventing the reassociation of the receptor with heat-shock proteins or by maintaining a conformation suitable for interaction with other transcription factors or components of the general transcriptional machinery. Although the hypothesis should be tested with further studies, the blocking of receptor reassociation with heat-shock proteins by phosphorylation is supported by both the timing and the position of the phosphorylation. First, the phosphorylation occurs in cytosol after hormone binding but before tight nuclear binding and DNA binding (Denner et al., 1990a; Poletti et al., 1993). This is concomitant with the dissociation of the heat-shock proteins from the receptor. Second, the hinge region has been shown to be involved in interaction with heatshock proteins (Carson-Jurica et al., 1990). In uitro, interaction with heat-shock protein is a reversible process and the complexes can be reconstituted (Smith et al., 1990). Hormone binding blocks the reconstitution. It will be interesting to find out whether Ser530 phosphorylation can block this reconstitution process. 2. hER LeGoff et al. (1994) have reported that the combined mutation of Serll8, Serl04, and SerlOG to alanine resulted in a 40% reduction in
PHOSPHORYLATION AND STEROID HORMONE ACTION
299
transcriptional activity for the mutant receptor in COS as well as 3T3 and CHO cells, while single-site mutation of any of them resulted in only a 15%reduction in transcriptional activity for the mutant receptor. However, a more extensive mutational analysis of Serll8 by Ali et al. (19931, using reporters containing promoters from genes that are responsive to estrogen stimulation under physiological conditions, has yielded substantially different results. The mutation of Serll8 to alanine caused, in both COS-1 and HeLa cells, a significant reduction in the transcriptional activity of the full-length hER. The reduction in activity for the mutant receptor ranged from 40% to 75%, depending on the cell type and the reporter gene used in the assay. The mutation of Serl18 to alanine in hER in which the hormone binding domain, including the AF-2 activity, has been deleted decreased the activity of the truncated receptor by 40% to 75% in COS-1 cells, which is similar to the degree of reduction observed for the full-length receptor. This suggests that Serllg phosphorylation modulates the AF-1 function of the receptor. However, in contrast to the results observed in HeLa and COS-1 cells, mutation of Sei-118 to alanine did not affect the activity of the receptor in chicken embryo fibroblast cells, indicating that Ser118 phosphorylation may not be important for transactivation by hER in these cells. 3. GR Mouse GR has been shown to be phosphorylated at seven residues clustered in the A/B region (Bodwell et al., 1991). Substitution of alanine or aspartic acid at single sites or all of the sites failed t o cause a significant reduction in the transcriptional activity of the receptor (Mason and Housley, 1993). However, as has been pointed out by the authors, only a single reporter gene containing the mouse mammary tumor virus promoter was examined.
4. TR Little is known about the significance of the phosphorylation in TR activation. However, mutational studies with two TR variants yielded interesting results. TRa2 is a naturally occurring variant of TR which does not bind thyroid hormone but competes with other TRs for DNA binding, thus inhibiting thyroid hormone action. The C terminus of TRa2 contains a cluster of nine serines and has been shown to be phosphorylated in uiuo when expressed in JEG-3 cells (Katz and Lazar, 1994).The simultaneous mutation of multiple serines in this region to alanine generated a mutant TRa2 which is a more potent inhibitor of thyroid hormone
300
WENLONG BAI AND NANCY L. WEIGEL
action than the wild type. These data suggest that the phosphorylation in the C terminus reduces the DNA binding affinity of TRa2 and attenuates the dominant negative effect of the receptor in uiuo. v-erbA is the ligand-independent mutant form of chicken c-erbAaencoded TR. v-erbA is an oncogenic protein that can inhibit the differentiation of erythroblasts and has been shown to be phosphorylated at Ser16’17 (Glineur et al., 1990). The phosphorylation is increased in uiuo after the activation of either PKA or PKC pathways, and treatment with H7, which can inhibit several kinases, including PKA and PKC, resulted in both a dose-dependent inhibition of the phosphorylation and the induction of terminal differentiation of the erythroblasts. The mutation of Ser16’17 to alanine blocked the ability of v-erbA to inhibit the differentiation of erythroblasts and to block the expression of erythrocyte-specific genes. This mutation did not affect either the nuclear localization or DNA binding of the protein. These data demonstrated that the phosphorylation of v-erbA is required for its transforming activity and is important for the transcriptional repression of at least some of the target genes in erythroid cells.
5. Vitamin D Receptors The mutational analysis of the major phosphorylation site, Ser208, a potential CKII site, by two groups yielded somewhat different results, but each of the studies raised an interesting point. Jurutka et al. (1993) reported that the mutation of Ser208 to glycine can block the enhanced transcriptional activity of the receptor caused by overexpression of CKII in COS-7 cells. This study raised an interesting suggestion that, although a phosphorylation might not be required for transactivation by the receptor, it may modulate the activity of the receptor under special cellular conditions. An independent study by Hilliard et al. (1994) reported that the mutation of the same CKII site to alanine, glutamic acid, or aspartic acid did not affect the transcriptional activity of the receptor. Under their conditions the alanine mutant protein continued to be phosphorylated in a hormone-dependent manner on an alternative site that remains to be identified. This raised the possibility that some of the phosphorylation in steroid receptors is so essential that the biological system has developed a redundant system during evolution. Considering the limitations of the transient transfection assays used in these studies, the significance of phosphorylation in regulating the receptor activity under physiological conditions may have been underestimated in some of the studies. First, the mutational studies are performed mostly in cells that contain no endogenous receptors.
PHOSPHORYLATION AND STEROID HORMONE ACTION
301
They are not the cells in which the receptor functions under physiological conditions. Many of the steroid receptors contains two transactivation functions, AF-1 and AF-2. The contribution of the two AFs, especially the AF-1 in the A/B region, to the overall transcriptional activity of the receptor varies in different types of cells. In some cells the synergism between the two is important, while in other cells only AF-2 is important. If the cells in which AF-1 has no function are used in the studies, the mutation of most of the phosphorylation sites in group A receptors is likely to yield negative results, because the majority of the sites identified s o far are in the AIB region of the receptors. Second, most of the reporters used are artificially constructed. The interaction between the steroid receptors and other transcription factors in the regulatory sequence of natural target genes for the receptor could be missing on these reporters. If a particular phosphorylation happens to be important for the protein-protein interactions occurring on the natural target gene, the analysis with the artificial reporter is likely to yield negative results. Third, transient transfections result in relatively high-level expression of the receptors in a small percentage of cells. If the phosphorylation is important for the affinity of the receptor for DNA or for other factors, it may not be reflected in the assay, because the high level of the receptor can mask the effect of phosphorylation. Fourth, the reporter genes used in most of the assays are not integrated into the chromosome. Thus, the effects of phosphorylation on the receptor in a natural target gene packaged into chromosome cannot be assayed. Finally, a particular phosphorylation could affect the activity of the receptor under special physiological or pathological conditions.
rv. INTERACTION AND
BETWEEN STEROID HORMONE ACTION SIGNAL TRANSDUCTION PATHWAYS
A. REGULATION OF RECEPTOR ACTIVITYBY MODULATORS OF SIGNAL TRANSDUCTION PATHWAYS Numerous studies have demonstrated that the treatment of cells with protein kinase activators, phosphatase inhibitors, growth factors, and neurotransmitters can modulate the activity of steroid hormone receptors. Although these treatments do not necessarily result in a direct change in the phosphorylation of the nuclear steroid receptor, these findings provide evidence for the involvement of the phosphorylation process in steroid hormone action.
302
WENLONG BAI AND NANCY L. WEIGEL
1. Ligand-Independent Activation
cPR is transcriptionally activated by 8-bromo-CAMP treatment of CV-1 cells (Denner et al., 1990b) that are cotransfected with a plasmid constitutively expressing the receptor and a reporter, PREtkCAT, which contains two progesterone response elements in front of a thymidine kinase promoter followed by the chloramphenicol acetyltransferase gene. The activation of the reporter gene is strictly receptor dependent and requires no hormone, and thus it has been termed ligand-independent activation. Subsequent studies from numerous laboratories with different receptors have yielded mixed results. Receptors for human estrogen (Ignar-Trowbridge et al., 1992; Aronica and Katzenellenbogen, 1991; Katzenellenbogen and Norman, 1990) and androgen (Culig et al., 19941, but not receptors for glucocorticoids (Rangarajan et al., 1992) and mineralocorticoids (Power et al., 1991b), showed ligand-independent activation. The ligand-independent activation of hPR has been reported by one group (Kazmi et al., 1993), but others have not detected ligand-independent activation for hPR (Beck et al., 1992; Nordeen et al., 1993). Besides kinase activators such as 8-bromo-cAMP, phosphatase inhibitors such as okadaic acid, Calyculin, and vanadate (Zhang et al., 1994a), growth factors such as epidermal growth factor (EGF) (IgnarTrowbridge et al., 1992, 1993; Zhang et al., 1994a; Culig et al., 19941, keratinocyte growth factor, and insulin-like growth factor I (Culig et al., 1994), and neurotransmitters such as dopamine (Power et al., 1991a,b) have been shown to cause ligand-independent activation of different receptors. Studies with EGF and dopamine are particularly interesting because of their physiological significance. The studies with EGF and vanadate provided the evidence for the involvement of tyrosine kinases and phosphatases in the ligand-independent activation process. The mechanism of ligand-independent activation is unknown at this stage. The ligand-independent activation caused by dopamine is mediated by the D1 type of dopamine receptor (Power et al., 1991b), which can activate signal transduction pathways mediated by both PKA and PKC. So far, no evidence exists for the involvement of direct phosphorylation of the receptor during the ligand-independent activation. cPR has been shown to be phosphorylated by PKA in uitro (Denner et al., 1990a). However, none of the four known phosphorylation sites has been shown to be phosphorylated by PKA, and the in vitro PKA site is Ser528,which has not been shown to be phosphorylated in viuo. Deletion of cPR in regions containing Ser530, Ser211, or Ser260 did not eliminate the dopamine-induced activation of the receptor (Poweret al., 1991b).A mutant hER with a point mutation at residue 400 can be activated by
PHOSPHORYLATION AND STEROID HORMONE ACTION
303
hormone but not by dopamine (Smith et al., 1993a). Similarly, a point mutation in the C terminus of cPR eliminated the dopamine-dependent activation of the receptor without affecting the hormone-dependent activity of the receptor (Power et al., 1991b). Whether these mutations change the structure of the receptor which prevents ligand-independent activation or blocks the phosphorylation of the receptors remains to be determined. 2. Effects of Modulators of Signal Transduction Pathways on the Ligand-Dependent Activation of Receptor As mentioned during the previous discussion, not all steroid receptors can be activated in a ligand-independent manner. However, the modulation of the ligand-dependent activation by these components appears to be much more common. Treatment of cells with various agents such as 8-bromo-cAMP, okadaic acid, and TPA has been shown to enhance the hormone-induced transactivation of almost all receptors examined (Beck et al., 1992; Culig et al., 1994; Rangarajan et al., 1992; Power et al., 1991b; Nordeen et al., 1993; Jones et al., 1994; Huggenvik et al., 1993). Most studies have shown that hPR is not activated by ligand-independent pathways. However, the studies with activators of kinase in the presence of agonist or antagonist yielded interesting results. The activation of hPR by R5020 is usually antagonized by the treatment of cells with RU486. However, in the presence of 8-bromo-cAMP, RU486 acts like an agonist that can activate the receptor to a level comparable with that induced by R5020 in T47D cells (Beck et al., 1993). This antagonist-to-agonist switch of RU486 has also been demonstrated for GR (Nordeen et al., 1993). In similar experiments ZK98299, a progesterone antagonist that does not induce the binding of hPR to DNA, failed to act as an agonist. The mutant ER that cannot be activated in a ligand-independent manner by dopamine can also be activated by the combined treatment with dopamine and the estrogen antagonist tamoxifen (Smith et al., 1993a,b). The antagonist-to-agonist switches of RU-486 and tamoxifen have important clinical implications, because hormonal therapy of breast cancer with tamoxifen might have opposite effects on breast cancer cell growth, depending on the state of the different signal transduction pathways in the cells. OF SIGNAL TRANSDUCTION PATHWAYS BY STEROIDS B. ACTIVATION
So far, a significant amount of evidence for the regulation of steroid receptor activity by growth factors and modulators of kinases and phosphatases has been reviewed. In this section the evidence that ste-
304
WENLONG BAI AND NANCY L. WEIGEL
roid hormones can activate second messenger-regulated signal transduction pathways is discussed. In the same study that first demonstrated the ligand-independent activation of cPR (Denner et al., 1990b), it was also shown that the hormone-stimulated transcriptional activity of the receptor was almost completely blocked by treating cells with PKA inhibitors, H8, or a peptide derived from protein kinase inhibitor (PKI). Subsequent studies showed that the hormone-dependent activation of hPR is also inhibited by H8, while general transcription is not affected by this treatment (Beck et al., 1992). These studies raise the question of whether the signal transduction pathways only modulate the activity of the receptor under certain conditions or whether it is actually required for the activation. Several studies have shown that steroids can activate a second messenger system in lower organisms such as yeast (Hasegawa et al., 1991; Motizuki and Tsurugl, 1992). Recently, it was reported (Aronica et al., 1994) that estrogenic hormones can increase the cAMP levels in target breast cancer and uterus cells in culture and the intact uterus in uiuo. The increase in intracellular cAMP is induced by physiological concentrations of estradiol and is due to enhanced membrane adenylate cyclase activity through an unknown mechanism that does not involve de nouo transcription or translation. The estrogen-stimulated increase in cAMP is sufficient to activate transcription from cAMP response element (CRE)-containing genes. The induction of adenylate cyclase activity by estrogen suggests a possible membrane action of steroid hormones. Several studies has provided evidence for the presence of possible binding sites for estrogen (Lieberherr et al., 1993; Pietras and Szego, 1979; Welshons et al., 1993), glucocorticoid (Gametchu et al., 1993; Joels and de Kloet, 19911, and progesterone (Tischkau and Ramirez, 1993; Sadler et al., 1985) in cell membranes. In the case of progesterone, the relatively highaffinity binding site in the cell membrane is clearly different from the intracellular receptor in terms of hormone binding selectivity and other properties (Tischkau and Ramirez, 1993; Sadler et al., 1985). In the case of glucocorticoids, a subpopulation of glucocorticoid receptor can be detected on plasma membranes of human leukemia cells by an antipeptide antibody against the intracellular GR (Gametchu et al., 1993). In addition, it has been shown that the binding of a steroidsteroid-binding protein (SBP) complex to the membrane receptor for SBP can increase the level of cAMP in the cell (Hryb et al., 1985; Fissore et al., 1994). Although the evidence for the activation of second messengers by
PHOSPHORYLATION AND STEROID HORMONE ACTION
305
steroids is preliminary, these studies potentially have a profound impact in terms of understanding the mechanism of steroid hormone action. Through the second messenger-regulated signal transduction pathways, steroids not only can enhance gene expression regulated by the nuclear receptors but can also cause a new pattern of gene expression through nonreceptor transcription factors such as CRE-binding protein.
V. SUMMARY: A MODELOF STEROID HORMONE ACTION Based on the data reviewed in this chapter, a new model of steroid hormone action is proposed and shown in Fig. 1. In this model the traditional two-step model of steroid hormone action is depicted in the lower part and indicated by thick arrows, to be
Ligand-independent
Protein Kinases
.).)
s *
Ltgand-dependent
Inactive
Active
JJ
FIG.1. A model of steroid hormone action. The traditonal ligand-dependent activation is depicted in the lower part and is indicated by thick arrows to distinguish from newly defined pathways and regulatory steps mediated through phosphorylation, which are marked with thinner arrows. The essential role of phosphorylation is manifested by the central position of kinases and phosphatases in the model. The steps that are not well defined are indicated by arrows with dashed lines and the components that are not well characterized are labeled with question marks. S, Steroid; P, a class of phosphorylation site; GF&NT, growth factor and neurotransmitter; SRE, steroid response element; CRE, CAMPresponse element; CREB, CRE-binding protein.
306
WENLONG BAI AND NANCY L. WEIGEL
distinguished from newly defined pathways and regulatory steps mediated through phosphorylation, which are marked with thinner arrows. The uncertain steps in this model are indicated by arrows with dashed lines, and the unidentified components are labeled with question marks. Different from the traditional view, the essential role of phosphorylation in steroid hormone action is manifested by the central position occupied by protein kinases and phosphatases in this model. In the ligand-dependent pathway the binding of steroids transforms the receptors from inactive forms into active forms. Then, the active receptors dimerize, bind steroid response elements and regulate the transcription of target genes. Through the kinases and phosphatases signals generated by growth factors and neurotransmitters in the cell surface can be used to modulate this ligand-dependent activation process. Under certain conditions, if the regulatory signal is strong enough, it can cause the ligand-independent activation of steroid receptors, as depicted in the upper part of the model. In addition, steroid can bind to membrane receptors and activate the second messenger-regulated signal transduction pathways. The possible membrane receptors remain to be identified, and thus are labeled with a question mark. Mediated through subsequent phosphorylation reactions, the signal generated from the membrane by steroids can be used to enhance the gene expression regulated by the classical nuclear hormone receptors. Alternatively, the signal can cause, through nonreceptor transcription factors such as CREB, a new pattern of gene expression by regulating the transcription of genes under the control of the CRE. In this model the inactive receptor is considered to contain only one class of phosphorylation site, which stands for the basal phosphorylation. After ligand-dependent activation the receptor phosphorylation is increased and the additional phosphorylation is represented by two additional sites in the active receptors, one standing for the hormonedependent phosphorylation and the other representing the DNA-dependent phosphorylation. Whether the additional phosphorylation occurs in the receptors during the ligand-independent activation remains to be determined, and thus is represented with two sites labeled with question marks. It also remains to be determined whether the ligandindependent activation or the regulation of ligand-dependent activation by EGF and neurotransmitters is the result of a change in the phosphorylation of the receptor itself or of a cofactor that interacts with the receptor. This uncertainty is also reflected in the model by the
PHOSPHORYLATION AND STEROID HORMONE ACTION
307
phosphorylation site placed between the receptor and the cofactor and labeled with a question mark. The increase in protein phosphorylation can be achieved through different mechanisms by either making the protein a better substrate for a kinase or by increasing the activity of a kinase. During liganddependent activation the increase in receptor phosphorylation is likely to be the result of a conformational change of the receptor induced by the ligand binding that makes the receptor a better substrate for existing kinases. However, during ligand-independent activation the increase in the phosphorylation of the receptor or the cofactor is likely through the activation of the kinases. In conclusion, studies of phosphorylation and steroid hormone action have made significant progress in the past few years. The question is no longer whether phosphorylation is important in steroid hormone action but what roles phosphorylation plays. REFERENCES Ali, S., Metzger, D., Bornert, J . M., and Chambon, P. (1993). Modulation of transcriptional activation by ligand-dependent phosphorylation of the human oestrogen receptor AiB region. EMBO J. 12, 1153-1160. Anderson, C. W., and Lees-Miller, S. P. (1992). The nuclear serineithreonine protein kinase DNA-PK. Crit. Rev.Eukaryotic Gene Expression 2, 283-314. Arnold, S. F., Obourn, J. D., Jaffe, H., and Notides, A. C. (1994). Serine 167 is the major estradiol-induced phosphorylation site on the human estrogen receptor. Mol. Endocrinol. 8, 1208-1214. Aronica, S. M., and Katzenellenbogen, B. S. (1991). Progesterone receptor regulation in uterine cells: Stimulation by estrogen, cyclic adenosine 3’,5’-monophosphate, and insulin-like growth factor I and suppression by antiestrogens and protein kinase inhibitors. Endocrinology 128, 2045-2052. Aronica, S. M., Kraus, W. L., and Katzenellenbogen, B. S. (1994). Estrogen action via the CAMP signaling pathway: Stimulation of adenylate cyclase and CAMP-regulated gene transcription. Proc. Natl. Acad. Sci. U.S.A.91, 8517-8521. Bagchi, M. K., Tsai, S. Y., Tsai, M.-J., and OMalley, B. W. (1992). Ligand and DNAdependent phosphorylation of human progesterone receptor in vitro. Proc. Natl. Acad. Sci. U.S.A.89, 2664-2668. Bai, W., Tullos, S., and Weigel, N. L. (1994). Phosphorylation of Ser530 facilitates hormone-dependent transcriptional activation of the chicken progesterone receptor. Mol. Endocrinol. 8, 1465-1473. Bailly, A,, LePage, C., Rauch, M., and Milgrom, E. (1986). Sequence-specific DNA binding of the progesterone receptor to the uteroglobin gene: Effects of hormone, antihormone and receptor phosphorylation. EMBO J. 5,3235-3241. Baniahmad, A,, Steiner, C., Kohne, A. C., and Renkawitz, R. (1990). Modular structure of a chicken lysozyme silencer: Involvement of a n unusual thyroid hormone receptor binding site. Cell 61, 505-514. Beato, M. (1989). Gene regulation by steroid hormones. Cell 56, 335-344.
308
WENLONG BAI AND NANCY L. WEIGEL
Beck, C. A,, Weigel, N. L., and Edwards, D. P. (1992). Effects of hormone and cellular modulators of protein phosphorylation on transcriptional activity, DNA binding, and phosphorylation of human progesterone receptors. Mol. Endocrinol. 6,607-620. Beck, C. A., Weigel, N. L., Moyer, M. L., Nordeen, S. K., and Edwards, D. P. (1993).The progestone antagonist RU486 acquires agonist activity upon stimulation of CAMP signaling pathways. Proc. Natl. Acad. Sci. U.S.A. 90, 4441-4445. Bhat, M. K., Ashizawa, K., and Cheng, S.-Y. (1994).Phosphorylation enhances the target gene sequence-dependent dimerization of thyroid hormone receptor with retinoid X receptor. Proc. Natl. Acad. Sci. U.S.A. 91, 7927-7931. Bodwell, J. E., Orti, E., Coull, J. M., Pappin, D. J., Swift, F., and Smith L. I. (1991). Identification of the phosphorylation sites in the mouse glucocorticoid receptor. J. Biol. Chem. 266,7549-7555. Boulton, T. G., Nye, S. H., Robbins, D. J.,Ip, N. Y., Radziejewska, E., Morgenbesser, S. D., DePinho, R. A., Panayotatos, N., Cobb, M. H., and Yancopoulos, G. D. (1991).ERKs: A family of protein-serinehhreonine kinases that are activated and tyrosine phosphorylated in response to insulin and NGF. Cell 65, 663-675. Carson-Jurica, M. A,, Lee, A. T., Dobson, A. D. W., Conneely, 0. M., Schrader, W. T., and O’Malley, B. W. (1990). Interaction of the chicken progesterone receptor with HSP-90. J . Steroid Biochem. 34, 1-14. Carter, T. H., and Anderson, C. W. (1991). The DNA-activated protein kinase, DNA-PK. Programm. Mol. Subcell. Biol. 12, 37-58. Castoria, G., Migliaccio, A., Green, S., DiDomenico, M., Chambon, P., and Auricchio, F. (1993). Properties of a purified estradiol-dependent calf uterus tyrosine kinase. Biochemistry 32, 1740-1750. Culig, Z., Hobisch, A., Cronauer, M. V., Radmayr, C., Trapman, J.,Hittmair, A., Bartsch, G., and Klocker, H. (1994).Androgen receptor activation in prostatic tumor cell lines by insulin-like growth factor-I, keratinocyte growth factor, and epidermal growth factor. Cancer Res. 54, 5474-5478. Damm, K., Thompson, C. C., and Evans, R. M. (1989). Protein encoded by v-erbA functions as a thyroid-hormone receptor antagonist. Nature 339, 593-597. Davis, I. J., and Lau, L. F. (1994). Endocrine and neurogenic regulation of the orphan nuclear receptors Nur77 and Nurr-1 in the adrenal glands. Mol. Cell. Biol. 14,34693483. Davis, I. J., Hazel, T. G., Chen, R.-H., Blenis, J., and Lau, L. F. (1993). Functional domains and phosphorylation of the orphan receptor Nur77. Mol. Endocrinol. 7, 953-964. Dayani, N., McNaught, R. W., Shenolikar, S., and Smith, R. G. (1990).Receptor interconversion model of hormone action. 2. Requirement of both kinase and phosphatase activities for conferring estrogen binding activity of estrogen receptor. Biochemistq 29,2691-2698. DeMarzo, A. M., Beck, C. A,, Onate, S. A,, and Edwards, D. P. (1991).Dimerization of mammalian progesterone receptors occurs in the absence of DNA and is related to the release of the 90-kDa heat shock protein. Proc. Natl. Acad. Sci. U.S.A. 88, 7276. Denner, L. A., Weigel, N. L., Schrader, W. T., and O’Malley, B. W. (1989). Hormonedependent regulation of chicken progesterone receptor deoxyribonucleic binding and phosphorylation. Endocrinology 125, 3051-3058. Denner, L. A,, Schrader, W. T., OMalley, B. W., and Weigel, N. L. (1990a). Hormonal regulation and identification of chicken progesterone receptor phosphorylation sites. J. Biol. Chem. 265. 16548-16555.
PHOSPHORYLATION AND STEROID HORMONE ACTION
309
Denner, L. A,, Weigel, N. L., Maxwell, B. L., Schrader, W. T., and O’Malley, B. W. (1990b). Regulation of progesterone receptor-mediated transcription by phosphorylation. Science 250, 1740-1743. Denton, R. R., Koszewski, N. J.,Carter, T. H., and Notides, A. C. (1991). Phosphorylation of the estrogen receptor increases its affinity for estrogen specific DNA response elements. Annu. Meet. Endocr. SOC.73rd, p. 312. Elledge, S. J., Richman, R., Hall, F. L., Williams, R. T., Lodgson, N., and Harper, J. W. (1992). CDK2 encodes a 33 kDa cyclin A-associated protein kinase and is expressed before CDC2 in the cell cycle. Proc. Natl. Acad. Sci. U.S.A. 89, 2907-2911. Erikson, R. L. (1991). Structure, expression, and regulation of protein kinases involved in the phosphorylation of ribosomal protein S6. J . Biol. Chem. 266, 6007-6010. Evans. R. M. (1988).The steroid and thyroid hormone receptor superfamily. Science 240, 889-895. Fahrner, T. J., Carroll, S. L., and Milbrandt, J. (1990). The NGFI-B protein, an inducible member of the thyroidisteroid receptor family, is rapidly modified posttranslationally. Mol. Cell. Biol. 10, 6454-6459. Fissore, F., Fortunati, N., Comba, A., Fazzri, A., Gaidano, G., Berta, L., and Frairia, R. (1994). The receptor-mediated action of sex steroid binding protein (SBP, SHBG): Accumulation of cAMP in MCF-7 cells under SBP and estradiol treatment. Steroids 59,661-667. Folkers, G. E., van der Leede, B. M., and van der Saag, P. T. (1993). The retinoic acid receptor-beta2 contains two separate cell-specific transactivation domains, at the N-terminus and in the ligand-binding domain. Mol. Endocrinol. 7, 616-627. Gametchu, B., Watson, C. S., and Wu, S. (1993). Use of receptor antibodies to demonstrate membrane glucocorticoid receptor in cells from human leukemic patients. FASEB J . 7, 1283-1292. Garcia, T., Tuohimaa, P., Mester, J., Buchou, T., Renoir, M., and Baulieu, E. E. (1983). Protein kinase activity of purified components of the chicken oviduct progesterone receptor. Biochem. Biophys. Res. Commun. 113, 960-966. Garcia, T., Buchou, T., Renoir, M., and Baulieu, E. E. (1986). A protein kinase copurified with chick oviduct progestone receptor. Biochemistry 25, 7937-7942. Glineur, C., Bailly, A., and Ghysdael, J. (1989). The c-erbA alpha-encoded thyroid hormone receptor is phosphorylated in its amino terminal domain by casein kinse 11. Oncogene 4, 1247-1254. Glineur, C., Zenke, M., Beng, H., and Ghysdael, J. (1990). Phosphorylation of the v-erbA protein is required for its function as an oncogene. Genes Deu. 4, 1663-1676. Goldberg, Y., Glineur, C., Gesquiere, J. C., Ricaurt, A., Sap, J., Vennstrom, B., and Ghysdael, J. (1988). Activation of protein kinase C or cAMP dependent protein kinase increases phosphorylation of c-erbA encoded thyroid hormone receptor and of the v-erbA encoded protein. EMBO J . 7, 2425-2433. Green, S., and Chambon, P. (1988). Nuclear receptors enhance our understanding of transcription regulation. Trends Genet. 4, 309-314. Hasegawa, S., Tanaka, S., Hishinuma, F., and Kurata, S.-I. (1991). Estrogen delays entry of the yeast Saccharomyces cereuisiae into meiosis. Cell Struct. Funct. 16, 195201. Hilliard, G. M., Cook, R. G., Weigel, N. L., and Pike, J. W. (1994). 1,25-Dihydroxyvitamin D3 modulates phosphorylation of serine 205 in the human vitamin D receptor: Sitedirected mutagenesis of this residue promotes alternative phosphorylation. Biochemistry 33,4300-4311. Hirata, Y., Kiuchi, K., Chen, H.-C., Milbrandt, J., and Guroff, G. (1993). The phospho-
310
WENLONG BAI AND NANCY L. WEIGEL
rylation and DNA binding of the DNA-binding domain of the orphan nuclear receptor NGFI-B. J . Biol. Chem. 268,24808-24812. Hoeck, W., Rusconi, S., and Groner, B. (1989). Hormone-dependent phosphorylation of the glucocorticoid receptor occurs mainly in the amino-terminal transactivation domain. J . Biol. Chem. 264, 14396-14402. Horwitz, K. B., Francis, M. D., and Weill, L. L. (1985). Hormone-dependent covalent modification and processing of human progesterone receptors in the nucleus. DNA 4, 451-460. Housley, P. R., and Pratt, W. B. (1983). Direct demonstration of glucocorticoid receptor phosphorylation by intact L-cells. J . Biol. Chem. 258, 4630-4635. Hryb, D. J., Khan, M. S., and Rosner, W. (1985).Testosterone-estradiol-bindingglobulin binds to human prostatic cell membranes. Biochem. Biophys. Res. Commun. 128, 432-440. Hsieh, J. C., Jurutka, P. W., Galligan, M. A,, Terpening, C. M., Haussler, C. A,, Samuels, D. S., Shimizu, Y., Shimizu, N., and Haussler, M. R. (1991). Human vitamin D receptor is selectively phosphorylated by protein kinase C on serine 51, a residue crucial for its transactivation function. Proc. Natl. Acad. Sci. U.S.A. 88,9315-9319. Hsu, S., Qi, M., and DeFranco, D. B. (1992). Cell cycle regulation of glucocorticoid receptor function. EMBO J . 11, 3457-3468. Huggenvik, J . I., Collard, M. W., Kim, Y.-W., and Sharma, R. P. (1993). Modification of the retinotic acid signaling pathway by the catalytic subunit of protein kinase A. Mol. Endocrinol. 7, 543-550. Ignar-Trowbridge, D. M., Nelson, K. G., Bidwell, M. C., Curtis, S. W., Washburn, T. F., Machlachlan, J . A., and Korach, K. S. (1992). Coupling of dual signaling pathways: Epidermal growth factor action involves the estrogen receptor. Proc. Natl. Acad. Sci. U.S.A. 89, 4658-4662. Ignar-Trowbridge, D. M., Teng, C. T., Ross, K. A., Parker, M. G., Korach, K. S., and McLachlan, J. A. (1993). Peptide growth factors elicit estrogen receptor-dependent trnscriptional activation of a n estrogen-responsive element. Mol. Endocrinol. 7, 992-998. Joels, M., and de Kloet, E. R. (1991). Effect of corticosteroid hormones on electrical activity in rat hippocampus. J. Steroid Biochem. Mol. Biol. 40, 83-86. Jones, K. E., Brubaker, J . H., and Chin, W. W. (1994). Evidence that phosphorylation events participate in thyroid hormone action. Endocrinology 134, 543-548. Jurutka, P. W., Hsieh, J. C., MacDonald, P. N., Terpening, C. M., Haussler, C. A., Haussler, M. R., and Whitfield, G. K. (1993).Phosphorylation of serine 208 in the human vitamin D receptor. The predominant amino acid phosphorylated by casein kinase 11, in vitro, and identification as a significant phosphorylation site in intact cells. J . Biol. Chem. 268, 6791-6799. Karin, M. (1992). Signal transduction from cell surface to nucleus in development and disease. FASEB J . 6, 2581-2590. Karin, M., and Smeal, T. (1992). Control of transcription factors by signal tranduction pathways: The beginning of the end. Trends Biochem. Sci. 17, 418-422. Katz, D., and Lazar, M. A. (1994). Phosphorylation of the thyroid hormone receptor variant TRa2 inhibits its DNA binding and dominant negative activity. Endocr. Soc. 584. Katzenellenbogen, B. S., and Norman, M. J. (1990). Multihormonal regulation of the progesterone receptor in MCF-7 human breast cancer cells: Interrelationships among insuliniinsulin-like growth factor-I, serum, and estrogen. Endocrinology 126,891-898. Kazmi, S. M. I. Visconti, V., Plante, R. K., Ishaque, A,, and Lau, C. (1993). Differential
PHOSPHORYLATION AND STEROID HORMONE ACTION
311
regulation of human progestone receptor A and B form-mediated trans-activation by phosphorylation. Endocrinology 144, 1230-1238. Kemppainen, J . A., Lane, W. V., Sar, M., and Wilson, E. M. (1992). Androgen receptor phosphorylation, turnover, nuclear transport and transcriptional activation. Specificity for steroid and antihormone. J . Biol. Chem.,267, 968-974. Kumar, V., Green, S., Stack, G., Berry, M., Jin, J., and Chambon, P. (1987). Functional domains of the human estrogen receptor. Cell 51, 941-951. Lees, J . A,, Fawell, S. E., and Parker, M. G. (1989). Identification of two transactivation domain in the mouse oestrogen receptor. Nucleic Acids Res. 17, 5477-5487. LeGoff, P., Montano, M. M., Schodin, D. J., and Katzenellenbogen, B. S. (1994). Phosphorylation of the human estrogen receptor. Identification of hormone-regulated sites and examination of their influence on transcriptional activity. J . Biol. Chem. 269, 4458-4466. Lieberherr, M., Grosse, B., Kachkache, M., and Balsan, S. (1993). Cell signaling and estrogens in female rat osteoblast: A possible involvement of unconventional nonnuclear receptors. J . Bone Miner. Res. 8, 1365-1376. Mason, S. A., and Housley, P. R. (1993). Site-directed mutagenesis of the phosphorylation sites in the mouse glucocorticoid receptor. J. Biol. Chem. 268, 21501-21504. Mazer, J . S., Garabedian, M., and Vulliet, P. R. (1990). Phosphorylation of a cloned glucocorticoid receptor fragment by a novel proline-directed protein kinase. FASEB J . 4, A2233. Mendel, D. B., Bodwell, J. E., and Munck, A. (1987).Activation of cytosolic glucocorticoid -receptor complexes in intact WEHI-7 cells does not dephosphorylate the steroidbinding protein. J . Biol. Chem. 262, 5641-5648. Migliaccio, A,, Rotondi, A., and Auricchio, F. (1984). Calmodulin stimulated phosphorylation of 17beta-estradiol receptor on tyrosine. Proc. Natl. Acad. Sci. U.S.A. 81, 5921-5925. Migliaccio, A., Rotondi, A,, and Auricchio, F. (1986).Estradiol receptor: Phosphorylation on tyrosine in uterus and interaction with anti-phosphotyrosine antibody. EMBO J . 5,2867-2872. Miller-Diener, A,, Schmidt, T. J., and Litwack, G. (1985). Protein-kinase activity associated with the purified rat hepatic glucocorticoid receptor. Proc. Natl. Acad. Sci. U.S.A. 82,4003-4007. Moreno, S., and Nurse, P. (1990). Substrates for p34cdc2 in vitro veritas? Cell 61, 549551. Motizuki, M., and Tsurugl, K. (1992). Effect of 17p-estradiol on the generation time of old cells of the yeast Saccharomyces cereuisiae. Biochem. Biophys. Res. Commun. 183,1191-1196. Mulkhopadhyay, N. K., Price, D. L., Kyriakis, J. M., Pelech, S., Sanghera, J., and Avruch, J. (1992). An array of insulin activated proline-directed serinelthreonine protein kinases phosphorylate p70 S6 kinase. J. Biol. Chem. 267, 3325-3335. Munck, A,, and Brinck-Johnson, T. (1968). Specific and non-specific physicochemical interactions of glucocorticoids and related steroids with rat thymus cells in vitro. J . Biol. Chem. 243,5556-5565. Nordeen, S. K., Bona, B. J., and Moyer, M. L. (1993). Latent agonist activity of the steroid antagonist, RU486, is unmasked in cells treated with activators of protein kinase A. Mol. Endocrinol. 7, 731-742. O’Malley, B. W. (1990). The steroid receptor superfamily: More excitement predicted for the future. Mol. Endocrinol. 4, 363-369. O’Malley, B. W., and Conneely, 0. M. (1992). Minireview: Orphan receptors: In search of a unifying hypothesis for activation. Mol. Endocrinol. 6, 1359-1361.
312
WENLONG BAI AND NANCY L. WEIGEL
O’Malley, B. W., McGuire, W. L., Kohler, P. O., and Koreman, S. G. (1969).Studies on the mechanism of steroid hormone regulation of synthesis of specific proteins. Recent Prog. Horm. Res. 25, 105-160. Orti, E., Bodwell, J. E., and Munck, A. (1992). Phosphorylation of steroid hormone receptors. Endocr. Rev. 13, 105-128. Pennypacker, K. R., Hong, J.-S., and McMillian, M. K. (1994). Pharmacological regulation of AP-1 transcription factor DNA binding activity. FASEB J . 8, 475-478. Perisic, O., Radojcic, M., and Kanazir, D. T. (1987). Protein kinase activity can be separated from the purified activated rat liver glucocorticoid receptor. J. Biol. Chem. 262, 11688-11691. Pietras, R. J., and Szego, C. M. (1979). Estrogen receptors in uterine plasma membrane. J . Steroid Biochem. 11, 1471-1483. Pike, W. J., and Sleator, N. M. (1985). Hormone-dependent phosphorylation of the 1,25 dihydroxyvitamin D3 receptor is generated through a hormone-dependent phosphorylation. Biochem. Biophys. Res. Commun. 131, 378-385. Pinna, L. A. (1990). Casein kinase 2: An “eminence grise” in cellular regulation? Biochim. Biophys. Acta 1054,267-284. Poletti, A., and Weigel, N. L. (1993). Identification of a hormone-dependent phosphorylation site adjacent to the DNA-binding domain of the chicken progesterone receptor. Mol. Endocrinol. 7,241-246. Poletti, A,, Conneely, 0. M., McDonnell, D. P., Schrader, W. T., O’Malley, B. W., and Weigel, N. L. (1993). Chicken progesterone receptor expressed in Saccharomyces cerevisiae is correctly phosphorylated a t all four Ser-Pro phosphorylation sites. Biochemistry 32, 9563-9569. Power, R. F., Lydon, J. P., Conneely, 0. M., and O’Malley, B. W. (1991a). Dopamine activation of an orphan member of the steroid receptor superfamily. Science 252, 1546-1548. Power, R. F., Mani, S. K., Codina, J., Conneely, 0. M., and O’Malley, B. W. (1991b). Dopaminergic and ligand-independent activation of steroid hormone receptors. Science 254, 1636-1639. Rangarajan, P. N., Umesono, K., and Evans, R. M. (1992). Modulation of glucocorticoid receptor function by protein kinase A. Mol. Endocrinol. 6, 1451-1457. Raymoure, W. J., McNaught, R. W., and Smith, R. G. (1985).Fkversible activation of nonsteroid binding oestrogen receptor. Nature 314, 745-747. Raymoure, W. J., McNaught, R. W., Greene, G. L., and Smith, R. G. (1986). Receptor interconversion model of hormone action. J . Biol. Chem. 261, 17018-17025. Rochette-Egly, C., Gaub, M., Lutz, Y., Ali, S., Scheuer, I., and Chambon, P. (1992). Fktinoic acid receptor-beta: Immunodetection and phosphorylation on tyrosine residues. Mol. Endocrinol. 6, 2197-2209. Sadler, S. E., Bower, M. A., and Maller, J. L. (1985). Studies of a plasma membrane steroid receptor in Xenopus oocytes using the synthetic progrestin RU486. J . Steroid Biochem. 22,419-426. Sheridan, P. L., Evans, R. M., and Horwitz, K. B. (1989). Phosphotryptic peptide analysis of human progesterone receptors. New phosphorylated sites formed in nuclei after hormone treatment. J. Biol. Chem. 264,6520-6528. Smith, C. L., Conneely, 0. M., and O’Malley, B. W. (1993a). Modulation of the ligandindependent activation of the human estrogen receptor by hormone and antihormone. Proc. Natl. Acad. Sci. U.S.A. 90, 6120-6124. Smith, C. L., Conneely, 0. M., and O’Malley, B. W. (199313). Estrogen receptor activation by ligand-dependent and ligand-independent pathways. In “Steroid Receptors in Health and Disease” (V. K. Moudgil, ed.), pp. 333-356. Birkhauser, Boston.
PHOSPHORYLATION AND STEROID HORMONE ACTION
313
Smith, D. F., Schowalter, D. B., Kost, S. L., and Toft, D. 0. (1990). Reconstitution of progesterone receptor with heat shock proteins. Mol. Endocrinol. 4, 1704-1711. Sugawara, A., Yen, P. M., Apriletti; J . W., Ribeiro, R. C. J., Sacks, D. B., Baxter, J . D., and Chin, W. W. (1994). Phosphorylation selectively increases triiodothyronine receptor homodimer binding to DNA. J . Biol. Chem. 269, 433-437. Sullivan, W. P., Madden, B., McCormick, D. J., and Toft, D. 0. (1988). Hormonedependent phosphorylation of the avian progesterone receptor. J . Biol. Chem. 263, 14717-14723. Takimoto, G. S., Tasset, D. M., Eppert, A. C., and Horwitz, K. B. (1992). Hormoneinduced progesterone receptor phosphorylation consists of sequential DNA-independent and DNA-dependent stages: Analysis with zinc finger mutants and the progesterone antagonist ZK98299. Proc. Natl. Acad. Sci. U.S.A. 89, 3050-3054. Tischkau, S. A., and Ramirez, V. D. (1993). A specific membrane binding protein for progesterone in rat brain: Sex differences and induction by estrogen. Proc. Natl. Acad. Sci. U.S.A. 90, 1285-1289. Tora, L., White, J. H., Brou, C., Tasset, D. M., Webster, N. J. G., Scheer, E., and Chambon, P. (1989). The human estrogen receptor has two independent nonacidic transcriptional activation functions. Cell 59, 477-487. Tsai, M.-J., and O’Malley, B. W. (1994). Molecular mechanisms of action of steroid/ thyroid receptor superfamily members. Annu. Rev. Biochem. 63, 451-486. Tuazon, P. T., and Traugh, J . A. (1991). Casein kinase I and 11-multipotential serine protein kinases: Structure, function and regulation. Adu. Second Messenger Res. 23, 123-164. van Laar, J. H., Bolt-de Vries, J., Trapman, J., and Brinkmann, A. 0. (1990). Androgen receptor heterogeneity and phosphorylation in human LNCaP cells. Bzochem. Biophys. Res. Commun. 166, 193-200. Washburn, T. F., Hocutt, A,, Brautigan, D. L., and Korach, K. S. (1991). Uterine estrogen receptor in vivo: Phosphorylation of nuclear specific forms on serine residues. Mol. Endocrinol. 5, 235-242. Webster, N. J. G., Green, S., Jin, J., and Chambon, P. (1988). The hormone-binding domains of the estrogen and glucocorticoid receptors contain an inducible transcription activation function. Cell 54, 199-207. Weigel, N. L. (1994). Receptor phosphorylation. In “Mechanism of Steroid Hormone Regulation of Gene Transcription” (M.-J. Tsai and B. W. O’Malley, eds.), pp. 93-110. Landes, Austin, Texas. Weigel, N. L., Carter, T. H., Schrader, W. T., and O’Malley, B. W. (1992). Chicken progestone receptor is phosphorylated by a DNA-dependent protein kinase during in vitro transcription assays. Mol. Endocrinol. 6, 8-14. Welshons, W. V., Grady, L. H., Judy, B. M., Jordan, V. C., and Preziosi, D. E. (1993). Subcellular compartmentalization of MCF-7 estrogen receptor synthesis and degradation. Mol. Cell. Endocrinol. 94, 183-194. Zhang, Y., Bai, W., Allgood, V. E., and Weigel, N. L. (1994a). Multiple signaling pathways activate the chicken progesterone receptor. Mol. Endocrinol. 8, 577-584. Zhang, Y., Beck, C. A., Hutchens, T. W., Clement, I. V., Harper, J. W., Edwards, D. P., and Weigel, N. L. (1994b). Identification of basal, hormone-dependent, and PR-B specific phosphorylation sites in human progesterone receptor (hPR). Endocr. SOC.630. Zhang, Y., Beck, C. A,, Poletti, A., Edwards, D. P., and Weigel, N. L. (1994~). Identification of phosphorylation sites unique to the B form of human progesterone receptor: I n vitro phosphorylation by casein kinase 11. J. Biol. Chem. 269, 31034-31040.
This Page Intentionally Left Blank
VITAMINS AND HORMONES, VOL. 51
Nucleocytoplasmic Shuttling of Steroid Receptors DONALD B. DEFRANCO, ANURADHA P. MADAN, YUTING TANG, UMA R. CHANDRAN, NIANXING XIAO, AND J U N YANG Department of Biological Sciences University of Pittsburgh Pittsburgh, Pennsylvania 15260
I. Introduction
11. Subcellular Localization of Steroid Receptors A. A Historical Perspective B. Nuclear Import of Steroid Receptors 111. Nucleocytoplasmic Shuttling of Steroid Receptors A. A Historical Perspective B. Direct Demonstration of Steroid Receptor Nucleocytoplasmic Shuttling C. Some Questions D. Some Possible Answers from a New Model E. Some Final Thoughts References
I. INTRODUCTION Ever since the discovery of soluble intracellular receptors for steroid hormones, there has been considerable debate over the precise mechanisms of intracellular transmission of steroid hormone signals, which ultimately results in specific alterations of gene activity. Although some early models entertained such notions as transfer of hormone from a cytoplasmic receptor to a distinct nuclear receptor, this was quickly dispelled as it was recognized that a single receptor species accounted for cytoplasmic and nuclear hormone binding activities (Gorski et al., 1968; Jensen et al., 1968). At that point attention was directed toward uncovering the mechanisms responsible for nuclear translocation of hormone-bound cytoplasmic receptors. Over the past 10 years a number of reports have claimed to “definitively” establish a pathway for subcellular trafficking of steroid receptors. Despite the increased sophistication of biochemical, cell biological, and molecular techniques to examine the subcellular localization of steroid receptors, conflicting models have been proposed. In this chapter we take advantage of some new insights concerning nucleocytoplasmic trafficking to elaborate a model of steroid receptor nucle315
Copyright %\ 1995 by Academic Press, Inc. All rights of reproduction In any form reserved.
316
DONALD B. DEFRANCO et al.
ar transport that should dispel any controversy concerning receptor localization. With the demonstration that steroid receptor transport through the nuclear pore is bidirectional (Guiochon-Mantel et al., 1991; Chandran and DeFranco, 1992; Dauvois et al., 1993; Madan and DeFranco, 1993), steroid receptors must no longer be considered statically confined to any subcellular compartment. Rather, the equilibrium subcellular distribution of steroid receptors in this dynamic trafficking pathway is determined by the relationship between nuclear import and export rates. As these overall rates of passage through the nuclear pore may be the summation of multiple distinct kinetic components (see Section III,D), there may be various levels at which the subcellular compartmentalization of steroid receptors is regulated.
11. SUBCELLULAR LOCALIZATION OF STEROID RECEPTORS A. A HISTORICAL PERSPECTIVE As early as 1968 it was recognized, using hormone binding as an assay, that steroid receptors could reside in both cytoplasmic and nuclear compartments (Gorski et al., 1968; Jensen et al., 1968). Accumulation of nuclear receptors was proposed to result from a two-step process in which cytoplasmic receptors, following hormone binding, underwent an “activation” step and then acquired the capacity to import into nuclei and bind DNA (for a recent review see Pratt, 1992). With the availability of antireceptor antibodies, direct visualization of receptor subcellular localization became feasible using various immunocytochemical methodologies. However, even with these useful reagents and sophisticated cell biological techniques, the controversy concerning the mechanism of intracellular trafficking of steroid receptors persisted. The predominant view was that unoccupied estrogen and progesterone receptors (ERs and PRs) localized within the nucleus (Welshonset al., 1984; Guiochon-Mantel et al., 19891,while unoccupied glucocorticoid receptors (GRs) localized within the cytoplasm (Picard and Yamamoto, 1987; Wikstrom et al., 1987; Qi et al., 1989; Cidlowski et al., 1990). However, there were a number of results that did not conform to this consensus view (Welshons et al., 1985; Brink et al., 19921, and it was argued that the use of different antibodies, fixation conditions, and cell and tissue types might account for seemingly contradictory results. There was no general agreement on the results that
NUCLEOCYTOPLASMIC SHUTTLING
317
depicted the most physiologically relevant localization of steroid receptors, or on the significance of this seemingly differential localization. OF STEROID RECEPTORS B. NUCLEAR IMPORT
1. General Considerations of Nuclear Import
Prior to detailing the most recent advances in steroid receptor nuclear transport, we briefly highlight some important general features of protein trafficking into the nucleus. Proteins that import into the nucleus pass through the nuclear pore complex (NPC), a large structure of approximately 106 MDa composed of at least 100 different proteins (Akey and Radermacher, 1993). The mechanism of nuclear import can be minimally divided into two steps, the first step being an energyindependent association of the karyophile with the NPC, and the second step, an energy-dependent translocation through the pore (Newmeyer and Forbes, 1988; Richardson et al., 1988). A small aqueous channel that exists within the NPC allows for the free diffusion of macromolecules below 40 kDa, but macromolecules that exceed the diffusion limit of the pore must traverse a central plug assembly in an active transport process (Akey and Radermacher, 1993). Size alone is not the sole determinant of the mode of translocation through the NPC, as some small proteins, such as histone H1, utilize an active transport mechanism to translocate through the NPC (Breeuwer and Goldfarb, 1990).
2. Nuclear Localization Signal Sequences Karyophilic proteins generally contain a sequence of amino acids, that is, a nuclear localization signal sequence (NLS) (Dingwall and Laskey, 19911, that serves as a recognition site for soluble receptor proteins (Adam et al., 1989; Yamasaki et al., 1989). In some cases proteins destined for nuclear import, which lack discernible NLSs, can be cotransported into nuclei presumedly via stable interactions with NLS-containing proteins (Zhao and Padmanabhan, 1988; Sommer et al., 1991). Although the initial prototype NLS identified within the simian virus 40 (SV40) large-tumor antigen (TAg) is comprised of a contiguous stretch of basic amino acid residues flanked by prolines (Kalderon et al., 1984; Lanford et al., 19861, most other NLSs are bipartite and in some cases comprise multiple short basic amino acid stretches that must cooperate to educe NLS activity (Robbins et al., 1991). Bipartite and contiguous NLSs do not appear to be distin-
318
DONALD B. DEFRANCO et al.
guished by their interactions with NLS-binding proteins (Stochaj et al., 1993) and probably utilize similar mechanisms to direct proteins to the NPC. 3. Steroid Receptor NLSs Through the analysis of chimeras possessing various segments of the GR rat linked to p-galactosidase, Picard and Yamamoto (1987) were the first to identify a steroid receptor NLS. In fact, their studies revealed the presence of two distinct NLSs within the GR: a hormoneindependent NLS, NL1, which is located near the GR DNA binding domain and resembles the prototype SV40 TAg NLS, and a hormonedependent NLS, NL2, which was not precisely defined but resides within the GR ligand binding domain (LBD). NLSs that resemble the GR NL1 have been identified within the PR (Guiochon-Mantel et al., 19891, ER (Picard et al., 1990b), and androgen receptor (AR) (Zhou et al., 19941, in each case localized near or within the receptor’s DNA binding domain. More detailed analysis of hormone-independent NLSs within steroid receptors (Cadepond et al., 1992; Ylikomi et al., 1992; Zhou et al., 1994) revealed that they comprised multiple noncontiguous clusters of basic amino acids, and as such resembled the nucleoplasmin bipartite NLS (Dingwall and Laskey, 1991) rather than the SV40 TAg NLS. We have recently identified three distinct components of the rat GR NL1 and have shown that clusters of basic amino acid residues within, and C-terminal to, the second zinc finger of the GR DNA binding domain contribute to NLS function (Fig. 1). These results confirm predictions of GR NL1 structure derived from the analysis of human GR deletion derivatives (Cadepond et al., 1992). The activity of various steroid receptor hormone-independent NLSs is suppressed, in cis, by the unoccupied GR LBD (Cadepond et al., 1992; Ylikomi et al., 1992). This repressive function is relieved upon the binding of hormone, most likely reflecting exposure of the NLS from a previously inaccessible site within a receptor heteromeric complex (Pratt, 1993). However, when isolated from NL1, the LBD of rat GR will target linked heterologous proteins to the nucleus in a hormonedependent manner (Picard and Yamamoto, 19871, suggesting that this domain could play some role in nuclear targeting of intact GR with a fully functional NLS. In contrast to the hormone-dependent NLS properties of the rat GR, the LBDs of the chicken PR (Ylikomi et al., 1992) and human ER (Ylikomi et al., 1992) AR (Zhou et al., 19941, and GR (Cadepond et al., 1992) do not appear to possess NLS activity on their own. However, in these cases the LBDs have the capacity, when bound
319
NUCLEOCYTOPLASMIC SHUTTLING
S A E
~ Y G
C
~
V 1
G
N
M
G ATA Q Q IG KI KKKT KRAEL
525
PNW FIG.1. The rat glucocorticoid receptor (GR) DNA binding domain and nuclear localization signal sequence (NLS). Amino acids 440-525 of the rat GR DNA-binding domain are shown in single-letter code, with the basic amino acids shown in outline which contribute to NLS activity within the three distinct proto-NLSs (i.e., pNL1, pNL2, and pNL3).
by hormone, to enhance the activity of linked hormone-independent NLSs (Cadepond et al., 1992; Ylikomi et al., 1992; Zhou et al., 1994). How do steroid receptor LBDs facilitate nuclear import? Steroid receptor LBDs do not possess readily identifiable stretches of basic amino acids, and thus probably do not function to cooperate with other basic stretches to generate a suitable substrate for soluble NLS receptor proteins. It seems more likely that the apparent NLS activity of steroid receptor LBDs is brought about by other proteins associated with this domain. There are examples of proteins that do not possess identifiable NLSs transporting to the nucleus in association with an NLS-containing protein (Zhao and Padmanabhan, 1988; Sommer et al., 1991). Such cotransport of steroid receptors deleted of their hormoneindependent NLSs has been implicated in at least two cases. For example, an exclusively cytoplasmic rabbit PR deleted of its NLS can accumulate within nuclei when coexpressed with an NLS-containing PR (Guiochon-Mantel et al., 1989). Likewise, human GR and chicken PR deleted of their functional NLSs can accumulate within nuclei when coexpressed with a derivative of the 90-kDa heat-shock protein (hsp90) to which an NLS has been linked (Kang et al., 1994). These results demonstrate that steroid receptors have the capacity to translocate through the NPC as a complex with other proteins [either as a receptor
320
DONALD B. DEFRANCO et al.
homodimer (Guiochon-Mantel et al., 1989) or as a receptor-hsp90 complex (Kang et al., 1994)1, but they do not establish whether such cotransport operates under normal physiological conditions. In summary, while it seems clear that the LBD of steroid receptors plays some role in nuclear transport, it remains to be established whether this domain possesses a bona fide novel NLS, or whether it contributes to nuclear transport strictly through its association with other proteins translocating through the NPC. 4. Mechanisms of Steroid Receptor Nuclear Import
I n uiuo studies have been instrumental in uncovering some important properties of nuclear import, such as the identification of NLSs (Kalderon et al., 1984; Lanford et al., 1986) and the ATP-dependent translocation through the NPC (Newmeyer and Forbes, 1988). However, with the development by Adam and Gerace and colleagues (Adam et al., 1990) of a competent in uitro nuclear transport system using digitonin-permeabilized cells, analysis of the mechanism of nuclear transport has been greatly facilitated. For example, cytoplasmic factors required for in uitro nuclear import, including the RaniTC4 monomeric G protein (Melchior et al., 1993; Moore and Blobel, 19931, have been identified. We have utilized the digitonin-permeabilized cell system to provide mechanistic details of GR nuclear import (Yang and DeFranco, 1994). I n uitro nuclear import of rat GR was shown to be hormone dependent and cold sensitive and to require ATP (Fig. 2) (Yang and DeFranco, 1994). In addition, analogous t o the nuclear import of nearly all NLS-containing proteins, efficient GR nuclear import in uitro was N-ethylmaleimide and wheat germ agglutinin sensitive (Fig. 2) and required GTP (data not shown). Our in uitro results recapitulate the hormone-dependent nuclear import of GR observed in many cell types and challenge the notion that cytoplasmic localization of unliganded GR in some systems reflects artifactual cell permeabilization and fixation conditions. Unliganded cytoplasmic GRs exist as a large heteromeric complex ~~
FIG. 2. Properties of hormone-dependent in vitro nuclear transport of the rat glucocorticoid receptor (GR). GR-containing cytosol from rat GrH2 hepatoma cells was added to permeabilized HeLa cells, and GR that transported into HeLa nuclei was detected by indirect immunofluorescence (for details see Yang and DeFranco, 1994).Import reactions were carried out under the following conditions: 30°C in the presence of dexamethasone and ATP (complete); and 30°C in the presence of N-ethylmaleimide (NEM), NEM and dithiothreitol (DTT), or wheat germ agglutinin (WGA). Arrows from nuclei incubated a t 0°C point out perinuclear staining.
Complete
+NEM, +DTT
-ATP
+WGA
+NEM
322
DONALD B. DEFRANCO et al
associated with a number of other proteins, including 90-, 70-, and 56kDa heat-shock proteins, hsp90, hsp70, and hsp56, respectively (Pratt, 1990; Smith and Toft, 1993). An hsp9O dimer, bound through the LBD of the GR, masks a number of receptor activities. The relief of this inactivation function upon hormone binding to the receptor is accompanied by the dissociation of hsp90 (Pratt, 1990; Smith and Toft, 1993). When sodium molybdate (Na2Mo0,) was included in GRcontaining cytosol to inhibit receptor activation, in uitro nuclear import of GR was completely inhibited (Yang and DeFranco, 1994). Since GR did not localize to the perinuclear regions in the presence of Na2Mo0, (Yang and DeFranco, 1994), it seems likely that activation of the receptor is required for the docking of GR to the NPC. In support of this notion, we observed perinuclear accumulation of activated GRs that were incubated with permeabilized cells at 0°C (Fig. 2). Translocation of the activated GRs to the nucleus was obtained upon shifting permeabilized cells from 0°C to 30°C (Yang and DeFranco, 1994).Thus, receptors held in a docking state at the NPC by a reduction in temperature are competent to complete their passage through the NPC upon temperature elevation. While the activation of GRs is accompanied by its release from a heterooligomeric complex and dissociation of hsp90, the fate of other GR-associated proteins is unclear. In some cells, hsp70 appears to remain bound to GR following its activation and tight association with nuclei (Sanchez et al., 19901, prompting speculation that hsp70 may influence some nuclear functions of the receptor or directly participate in the nuclear transport of GR. Since hsp70 has been shown to be required for in uitro nuclear import of nucleoplasmin (Shi and Thomas, 1992) and in uiuo nuclear import of SV40 TAg NLS conjugates (Imamoto et al., 19921, it was reasonable to expect that this chaperone might play an analogous role in GR nuclear transport. However, we were unable to demonstrate a requirement for hsp70 in GR nuclear import in uitro, using either intact GR with two NLSs, NL1 and NL2, or a GR derivative that possessed only the hormone-independent NLS, NL1 (Yang and DeFranco, 1994). Under identical experimental conditions SV40 TAg nuclear import was shown to require hsp70 (Yang and DeFranco, 1994). It still remains formally possible that hsp70 may participate in NL2 function, but it seems likely that other steroid receptors, whose hormone-independent NLS closely resembles the GR NL1, would also not require hsp7O function for nuclear import. What is the role of hsp70 in nuclear transport? In a recent model put forth by Goldfarb (1994), hsp70 was postulated to facilitate the disassembly of a nuclear translocation complex following its entry into the nucleus. This hsp70, along with the small GTPase, Ran, was hypothe-
NUCLEOCYTOPLASMIC SHUTTLING
323
sized to comprise part of the translocation complex, which could include the karyophilic protein and an NLS receptor (Goldfarb, 1994). While hsp70 can be cotransported into nuclei apparently through its association with an NLS conjugate (Okuno et al., 19931, it is unclear whether hsp7O cotransport is an essential step in the mechanism of protein translocation through the NPC. By virtue of its binding to NLS peptides (Imamoto et al., 1992), hsp70 may be stably associated with experimentally generated karyophiles with multiple linked NLSs, even during translocation through the NPC. However, hsp70 may only transiently associate with NLSs in their natural context and may be important for the appropriate presentation of these targeting signals to the nuclear import machinery but not participate directly in the translocation step (Dingwall and Laskey, 1992). Since hsp70 is required for the assembly of a GR heteromeric complex (Hutchinson et al., 1994), perhaps the lack of an hsp70 requirement for GR nuclear import may reflect the fact that the receptor, once assembled into a steroid-binding competent folding state, may likewise be folded into a nuclear transport competent state. As we have shown, uncovering of nuclear import activity of GR follows the dissociation of hsp90 (Yang and DeFranco, 1994). Since cytosolic GR used for in uitro nuclear transport had been assembled in uiuo into a heteromeric complex (Yang and DeFranco, 19941, additional unfolding of the NLS catalyzed by hsp70 may not be required for hormone-dependent GR nuclear transport. A GR deletion derivative purified from Escherichia coli, which includes only its DNA binding domain and NLS and does not form a heteromeric complex (Howard et al., 1990; Dalman et a,!., 1991), also does not require hsp70 for in uitro nuclear import (Yang and DeFranco, 1994). Thus, local conformation of the GR NLS, and not the global folding state of the receptor, may be critical for efficient nuclear transport of the receptor. The GR NL1 is closely associated with the receptor’s DNA binding domain (Fig. 11, which is compactly folded into a series of interacting a-helices (Luisi et al., 1991). The precise folding of this region of the receptor, perhaps with the assistance of hsp70, could preclude any subsequent requirement for hsp70 or any other chaperone for effective interaction of the receptor NLS with the nuclear import machinery.
SHUTTLING OF STEROID RECEPTORS 111. NUCLEOCYTOPLASMIC A. A HISTORICAL PERSPECTIVE During the time that cytoplasmic-to-nuclear transfer of steroid receptors was being debated, there were some claims that steroid recep-
324
DONALD B. DEFRANCO et al.
tor nuclear transport might, in fact, be bidirectional. Munck, Aronow, and co-workers (Ishii et al., 1972; Munck and Foley, 1976), upon their examination of hormone association and dissociation from GR, detected a cytoplasmic form of the receptor that appeared to derive from preexisting nuclear receptors. Additional support for the notion that GRs could export from nuclei and be reutilized has been provided by continuing biochemical studies (Orti et al., 19891, although alternative models for GR recycling have been proposed (Raaka and Samuels, 1983). For example, the hallmarks of one GR recycling model are an energy requirement for distinct steps in this pathway and the cyclic phosphorylation and dephosphorylation of GR that accompanies its procession through this cycle (Orti et al., 1989).
B. DIRECTDEMONSTRATION OF STEROID RECEPTOR NUCLEOCYTOPLASMIC SHUTTLING 1. Subcellular Localization of GRs by Indirect Immunofluorescence
For many years the idea that steroid receptor nuclear transport might be bidirectional received little attention, despite the continuing support provided by biochemical experiments from Munck's group (Orti et al., 1989). In recent years corroborating evidence for GR recycling has surprisingly come from indirect immunofluorescence (IIF) analysis of GR subcellular localization in oncogenically transformed cells (Qi et al., 1989). In v-mos-transformed rat NRK fibroblasts, hormone-bound GRs were found to be inefficiently retained within the nuclei and apparently redistributed to the cytoplasmic compartment (Qi et al., 1989). Analogous effects on GR nuclear retention were also observed in nontransformed NRK cells treated with okadaic acid, a specific inhibitor of protein phosphatase types 1 and 2A (DeFranco et al., 1991). It was in these later studies that the first direct visualization of GR nuclear export in individual cells was provided in IIF experiments examining the redistribution of nuclear receptors to the cytoplasm following hormone withdrawal (DeFranco et al., 1991). 2. Transient Heterokaryons Although various biochemical approaches have been used over the years to examine protein exchange between the nuclear and cytoplasmic compartments, the use of transient heterokaryons provides the optimum method for direct visualization of nucleocytoplasmic shuttling (Borer et al., 1989).In this procedure (Fig. 3) cells expressing a given nuclear antigen (donor cells) are fused to cells that do not
NUCLEOCYTOPLASMIC SHUTTLING
SR-Positive 8-Galactosidase-Negative DONOR NUCLEUS
325
SR-Negative IbGalactosidase-Positive RECIPENT NUCLEUS
1
PEG Fusion + Cycloheximide
HETEROKARYOB
FIG.3. A transient heterokaryon assay. A steroid receptor (SRbpositive P-galactosidase-negative cell, which provides the donor nucleus, is fused to a SR-negative P-galactosidase-positive cell, which provides the recipient nucleus, in the presence of polyethylene glycol (PEG) and cycloheximide. Donor and recipient nuclei in the heterokaryon are distinguished by differential staining characteristics, and SR within either nucleus is detected by indirect immunofluorescence. P-Galactosidase staining within the shared cytoplasm confirms cell fusion.
express that antigen (recipient cells). Redistribution of the antigen from donor cell nuclei to recipient cell nuclei is analyzed by IIF in the transient heterokaryons. A protein synthesis inhibitor is added prior to, and immediately after, fusion to block de nouo synthesis of the antigen and to ensure that antigen staining within recipient nuclei is due to its internuclear migration. We and others (Chandran and DeFranco, 1992; Dauvois et al., 1993; Guiochon-Mantel et al., 1991; Madan and DeFranco, 1993) have used this technique to unequivocally demonstrate nucleocytoplasmic shuttling of steroid receptors. Thus, receptors that accumulate within the nucleus are not statically confined to that compartment, but, rather, have the capacity to reversibly traverse the NPC (Fig. 4). With this discovery the controversy over the subcellular localization of unliganded steroid receptors becomes irrelevant. Given the reversible transport of steroid receptors through the NPC, the equilibrium sub-
326
DONALD B. DEFRANCO et al.
FIG.4. Nucleocytoplasmic shuttling of the rat glucocorticoid receptor (GR). Dexamethasone-treated GrH2 (GR containing) hepatoma cells were fused to GR-negative mouse NIH 3T3 fibroblasts using polyethylene glycol in the presence of cycloheximide. The right panel shows fixed cells stained with Hoechst dye to distinguish GrH2 nuclei (uniform staining; arrow) from NIH 3T3 nuclei (punctate staining; arrowhead). The left panel shows nuclear GR detected by indirect immunofluorescence in a n identical field of cells.
cellular distribution of steroid receptors must be governed by the relationship between the nuclear import (hi) and nuclear export (he)rates (Fig. 5). If the k, rate is much slower than the kirate, receptors would predominantly localize within the nuclear compartment, while if ki is much slower than k,, most receptors would be detected within the cytoplasm. We are now faced with the challenge of understanding how hi and k, are differentially regulated and identifying the cellular factors that influence these rates.
CYTOPLASM
N'UCLEUS
-
SR
Import rate limiting (Predominantly cytoplasmic) NEW
SR
$ki
SR
-
Export rate limiting (hedominantly nuclear)
FIG.5. Equilibrium subcellular distribution of steroid receptors (SRs). SR translocation through the nuclear pore complex (NPC) that is embedded within the nuclear envelope (NE). The relationship between relative rates of nuclear import (K,) and export (k,) influences the predominant subcellular localization of SR.
NUCLEOCYTOPLASMIC SHUTTLING
327
QUESTIONS C. SOME A number of questions come to mind when considering the bidirectional nuclear transport of steroid receptors. What role do hormones play in the nucleocytoplasmic shuttling of steroid receptors? Which step in this process, if any, requires ATP? Are there discrete signal sequences for nuclear import and export? While the transient heterokaryon assay seems amenable to experimental manipulations that should make these questions easily addressable, a clear consensus has yet to emerge concerning these fundamental aspects of steroid receptor nucleocytoplasmic shuttling. For example, PR and ER, which can accumulate within the nucleus of many cells when unliganded, do not require hormone binding for efficient nuclear export (GuiochonMantel et al., 1991; Chandran and DeFranco, 1992; Dauvois et al., 1993).In contrast, a GR derivative that possesses an additional heterologous NLS and accumulates within the nucleus when unliganded, is efficiently exported in the presence, but not the absence, of hormone (Madan and DeFranco, 1993). A heterologous TAg NLS-PR conjugate does not require hormone for efficient nuclear export (U. R. Chandran, unpublished observations, 19951, demonstrating that unliganded GR and PR have differential nuclear export capabilities when tested in analogous contexts. In one cell type in which unliganded wild-type GR had been shown to predominantly accumulate within the nucleus (Sanchez et al., 19901, efficient GR nuclear export also appeared to require hormone (A. P. Madan, unpublished observations, 1995). While there is little disagreement concerning the role of ATP in nuclear import, the role of ATP in nuclear export remains an unresolved issue not only for steroid receptors, but for other shuttling proteins as well (Nigg et al., 1991). Nuclear export of RNA and ribosomal subunits is a saturable energy-requiring process (Khanna-Gupta and Ware, 1989; Bataille et al., 19901, but recent analysis of an energy requirement for nuclear protein export has yielded conflicting results. Depletion of ATP, or incubation in the cold, can prevent export of both the human immunodeficiency virus type 1 (HIV-1) Rev protein from mammalian cell nuclei (Meyer and Malim, 1994) and various microinjected proteins from Xenopus oocyte nuclei (Schmidt-Zachmann et al., 1993). For steroid receptors, while it has been shown that GR becomes trapped within a biochemically unextractable nuclear compartment upon ATP depletion (Mendel et al., 19861, some, but not all, PR derivatives are excluded from the nucleus and apparently export without an energy requirement (Guiochon-Mantel et al., 1991). The suggestion
328
DONALD B. DEFRANCO et a1
that multiple pathways of nuclear export may exist provides a reasonable reconciliation of this controversy, but it is difficult to imagine why structurally related proteins such as PR and GR might utilize such fundamentally distinct pathways for nuclear export. Another source of disagreement concerns the existence of nuclear export signal sequences (NESs). In a number of cases, cytoplasmic proteins were shown to be retained within the nucleus if they were artificially introduced there by microinjection (Newmeyer et al., 1986; Guiochon-Mantel et al., 1994). These results imply that efficient nuclea r export requires the presence of a specific signal sequence that is not expected to be present in cytoplasmic proteins. However, efficient export of cytoplasmic pyruvate kinase was observed from microinjected Xenopus oocyte nuclei, leading to the conclusion that nuclear export does not require a specific NES (Schmidt-Zachmann et al., 1993). Most recently, specific signal sequences apparently required for efficient nuclear export have been identified, but they appear to be unrelated. Thus, while the NES of PR was shown to be identical to its NLS counterpart (Guiochon-Mantel et al., 19941, nuclear export of the HIV-1 Rev protein required a signal sequence that is distinct from its NLS (Meyer and Malim, 1994). Although independent studies have pointed out the likelihood that nuclear import and export share common mechanistic features (Featherstone et al., 19881, a n NLS did not appear to influence the export of pyruvate kinase protein microinjected into Xenopus oocyte nuclei (Schmidt-Zachmann et al., 1993). Despite these apparent controversies, none of the results obtained are inconsistent with the view that signal sequence-dependent versus -independent nuclear export pathways both operate but may be kinetically distinct. In the earliest models of GR nucleocytoplasmic shuttling, a case was made for a n involvement of receptor phosphorylation and dephosphorylation in this pathway (Orti et al., 1989). Support for the notion that GR nucleocytoplasmic shuttling is regulated by phosphorylation was provided by our group (DeFranco et al., 19911, although the precise role of receptor phosphorylation and dephosphorylation in this process seemed to differ somewhat from that proposed by Munck and associates (Orti et al., 1989). Since a GR deletion derivative that lacks nearly all putative phosphorylation sites is able to export from the nucleus (Madan and DeFranco, 1993), receptor phosphorylation is not obligatory for nucleocytoplasmic shuttling. Recent results from Munck and co-workers corroborate this conclusion (Hu et al., 1994). However, we cannot eliminate the possibility that, for wild-type GR, phosphorylation may still play a role in regulating the interactions of receptors
329
NUCLEOCYTOPLASMIC SHUTTLING
with nuclear and cytoplasmic components that influence nuclear import and export. D. SOME POSSIBLE ANSWERS FROM A NEW MODEL If all steroid receptors shuttle between the nuclear and cytoplasmic compartments, do they use the same mechanism for this transport process? The distinctions that we have apparently uncovered between GR and PR nuclear export (see Section II1,C) suggests that there may be some features of nucleocytoplasmic shuttling that differ for individual steroid receptors. We have devised a model that seeks to reconcile some conflicting views of steroid receptor nucleocytoplasmic shuttling. Irrespective of the precise mechanism used by steroid receptors to pass through the NPC, we hypothesize that the overall rate of either nuclear import or export is the summation of two kinetically distinct components: (1)an inherent rate of translocation through the NPC and (2) a rate of receptor release from anchoring complexes that exist in both the cytoplasm and the nucleus (Fig. 6). A number of groups have proposed that cytoplasmic heteromeric complexes between steroid receptors and members of the heat-shock and immunophilin families of proteins serve to anchor the receptors in the cytoplasm and influence nuclear import efficiency (Pratt, 1990; Smith and Toft, 1993), but our model has an additional feature of postulating the existence of nuclear anchoring complexes that restrict the nuclear export of steroid receptors. What evidence exists to support this model? This hypothesis is based primarily on the observed hormone dependence for the nuclear export of GR. Unliganded GR that accumulates within the nuclei of Chinese hamster ovary cells (Sanchez et al., 1990) or COS-1 cells, when supplied with an additional heterologous NLS (Madan and DeFranco,
e
S
R k, SR
SR
kr
SR-
NEm
FIG.6. Equilibrium subcellular distribution of steroid receptors (SRs). Influence of cytoplasmic (AC,) and nuclear (AC,) anchoring complexes. SR translocation through the nuclear pore complex (NPC) that is embedded within the nuclear envelope (NE). The overall rate of nuclear import or export is governed by the inherent rates of import ( k , ) and export (k,) and the rate of release (k,) from either AC, or AC,.
330
DONALD B. DEFRANCO et al.
1993), requires ligand binding for efficient export (A. P. Madan, unpublished observations, 1995). Thus, we postulate that, analogous to cytoplasmic anchoring complexes that restrict the nuclear import of GR (Pratt, 19921, GRs could be associated with nuclear anchoring complexes that restrict their nuclear export (Fig. 6). Since the receptor is apparently released from such complexes upon ligand binding, nuclear proteins that interact with the GR LBD, perhaps heat-shock proteins or immunophilins (Pratt, 1992), may be integral components of these anchoring complexes. Importantly, a GR deletion derivative that lacks its LBD efficiently exports from the nucleus (Madan and DeFranco, 19931, arguing that retention of GR within the nucleus is not significantly impacted by GR interactions with DNA. Rather, our model predicts that GR nuclear retention is governed more significantly by GRprotein interactions. If the association of unliganded GR with hsp90, or some other protein interacting with its LBD, is relatively stable within the nucleus, as it is in uztro (Beck et al., 19931, the rate-limiting step for GR nuclear export may be its release from such anchoring complexes. The nucleocytoplasmic shuttling of other steroid receptors could likewise be influenced by their association in nuclear anchoring complexes, while still accounting for the unique equilibrium subcellular distributions observed between different receptors. For example, PR that efficiently exports from the nucleus both when liganded and when unliganded (Guiochon-Mantel et al., 1991; Chandran and DeFranco, 1992) might be released from nuclear anchoring complexes much more readily than GR. In uitro experiments have shown not only that PR association with hsp90 dynamic, but that hormone binding does not influence the rate of hsp90 release from the receptor (Smith, 1993). Thus, if PR is associated with hsp90 in the nucleus, the kinetics of its release-and, as a direct result, the kinetics of its nuclear export-may not be significantly affected by ligand binding. In that case its subcellular distribution would be predominantly influenced by the relationship between inherent rates of nuclear import versus export. Since most studies support the view that nuclear export is slower than import (Madan and DeFranco, 1993; Schmidt-Zachmann et al., 1993), any receptor (e.g., PR) whose subcellular distribution is governed strictly by inherent rates of translocation through the NPC would be predicted to localize predominantly within the nucleus. While there is evidence both in favor (Kang et al., 1994) and against (Tuohimaa et al., 1993) the potential of steroid receptor interactions with heat-shock proteins in the nucleus, the validity of our nuclear anchoring model does not rely strictly on the existence of steroid
NUCLEOCYTOPLASMIC SHUTTLING
331
receptor-heat-shock protein interactions within the nucleus. Some other nuclear factors, which we believe must interact with the receptor’s LBD, could provide this function. As mentioned above, the most important aspect of the model is that protein-protein interactions, and not protein-nucleic acid interactions, govern the nuclear retention and thereby the nuclear export of steroid receptors. The recent demonstration of a role for hsp90 in facilitating the in uitro folding of the DNA binding domain of MyoD and other helix-loop-helix transcription factors (Shaknovich et al., 1992) raises new questions concerning a nuclear chaperoning activity for hsp90. It is well established both from in vitro (Bresnick et al., 1989) and in uiuo (Picard et al., 1990a) studies that hsp90 association with GR is required for hormone binding activity of the receptor. In uitro, the association of hsp9O with GR requires hsp70 and additional factors (Hutchinson et al., 1994). If we assume that hormone dissociation accompanies the nuclear export of GR, which was suggested by early kinetic studies of GR recycling by Munck and Holbrook (19841, the regeneration of GR competent to bind hormone would require its reassociation with hsp90. Since the in uitro reassociation of hsp90 with GR required hsp70 and other factors (Scherrer et al., 19901, perhaps the in uiuo reassociation of hsp90 with GR, which might occur in either the cytoplasm or the nucleus, likewise would require hsp70. This reformation of a GR heteromeric complex could represent the ATP-dependent step that has been postulated to be required for GR recycling and perhaps for nuclear export (Orti et al., 1989). In this case we would predict that GR in ATP-depleted cells would accumulate within the nuclei, since it may be unable to properly reassemble a heteromeric complex, which is required for hormone binding and perhaps for nuclear export. Since some PR derivatives, under some energy deprivation conditions, may export from the nucleus in the absence of hormone (Guiochon-Mantel et al., 1991), PR may differ from GR in this regard and could proceed through recycling and reutilization pathways without the obligatory assistance of hsp70. Could hsp70 play a more direct role in the nuclear export of steroid receptors? Although our in uitro experiments suggest that hsp70 is not involved in GR nuclear import (Yang and DeFranco, 19941, it remains to be determined whether hsp70 plays any role in steroid receptor nuclear export or anchoring (see Fig. 6). This hypothesis is not without precedence, as some hsp70 is found in the nucleus and has the capacity to export from that compartment (Mandell and Feldherr, 1990). Likewise, since hsp70 has also been shown to coimport into nuclei with an NLS-containing peptide (Okuno et al., 19931, it could also coexport
332
DONALD B. DEFRANCO et a1
from nuclei. Any presumed role for hsp70 in steroid receptor nuclear transport or anchoring must take into account the peptide binding and release activity of hsp70. In particular, since the release of bound peptide from hsp70 requires ATP hydrolysis (Milarski and Morimoto, 1989), any alterations in cellular ATP pools could have dramatic consequences for normally dynamic interactions of hsp70 with cellular proteins, such as steroid receptors. Since there is evidence suggesting that cytoplasmic and nuclear forms of steroid receptors may interact, even transiently with hsp7O (Kost et al., 1989; Sanchez et al., 1990), it is difficult to predict the exact status of steroid receptor-hsp70 interactions under conditions of ATP depletion. Clearly, all of the recent results concerning steroid receptor nuclear import and export, even if conflicting, point to the need for future experimentation that will more directly assess whether hsp90, hsp70, or some other chaperone has any role in steroid receptor passage in either direction through the NPC.
E. SOME FINALTHOUGHTS Many questions remain unanswered concerning the mechanism of steroid receptor nucleocytoplasmic shuttling. Do receptors utilize any specific cytoskeletal framework for intracellular trafficking? Despite the observation of some filamentous structure emanating from the NPC and the docking of proteins destined for import to the cytoplasmic face of the NPC (Richardson et al., 1988), definitive proof of the directed movement of material through the NPC via some type of filament system is lacking. There have been reports that the vitamin D receptor utilizes microtubules for its movement through the cytoplasm to the nucleus (Barsony et al., 19901, but it remains to be established whether this property can be extrapolated to steroid receptors or is unique to vitamin D receptors. In a model proposed by Pratt (19921, GRs are hypothesized to be anchored, via their association with heatshock proteins, to a cytoskeletal network and to track their way through the cell as part of a “transportosome” complex. Both in uiuo and in uitro nuclear import of GR proceeds in the presence of agents that disrupt either the actin- or tubulin-based cytoskeleton (Akner et al., 1992; K. Mt. Joy, J. Yang, and A. P. Madan, unpublished observations, 1999, suggesting that if GR moves through the cytoplasm on cytoskeletal tracks, it must utilize some component of the cytoskeleton that is resistant to these agents. There have been reports of protein (Meier and Blobel, 1992) and RNA (Spector, 1993) localization within fibrillar tracks with the nucleus, but the molecular identity of the framework that brings about this highly ordered alignment within the nucleus remains unknown.
NUCLEOCYTOPLASMIC SHUlTLING
333
The ability of steroid receptors to rapidly locate their target sites with the genome following their import into nuclei has been realized for some time, but has yet to be explained at the molecular level. Are there specific nuclear staging areas where steroid receptors accumulate? Both electron microscopic (Perrot-Applanat et al., 1986) and confocal microscopic (Martins et al., 1991) examination has revealed a nonrandom distribution of steroid receptors throughout the nucleus, but the functional significance of this distribution has yet to be determined. At early times following the import of GR into nuclei in uitro, a discrete speckled staining pattern was observed (Yang and DeFranco, 1994), hinting that high-affinity GR “acceptor” sites may exist within the nuclei. High-resolution analysis of hormone-regulated sites of transcription within individual nuclei, which is now possible (Jackson et al., 19931, should provide some clues as to the relationship between putative steroid receptor staging areas and active sites of transcription. Retinoic acid receptor (RAR) a also localizes within discrete subnuclear domains (Dyck et al., 1994; Weis et al., 1994).Disruption of the discrete nuclear staining pattern of this receptor, as observed with a promyelocyte-RAR OL fusion protein, leads to changes in cellular response to retinoic acid (Dyck et al., 1994; Weis et al., 19941, indicating that distinct subnuclear compartmentalization could have important functional consequences. In summary, our increased understanding of the mechanisms of intracellular trafficking of steroid receptors has helped to dispel many persistent controversies concerning steroid receptor subcellular localization. However, new questions have arisen that, even in early stages of investigations, have generated new controversies concerning steroid receptor movement through and within different subcellular compartments. We can look forward in future years to more in-depth molecular, biochemical, and cell biological investigation, the goals of which will be t o provide a detailed molecular description of the mechanisms of steroid receptor intracellular trafficking. REFERENCES Adam, S. A,, Lobl, T.J., Mitchell, M. A., and Gerace, L. (1989). Identification of specific binding proteins for a nuclear location sequence. Nature (London) 337, 276-279. Adam, S.A,, Sterne-Marr, R., and Gerace, L. (1990). Nuclear protein import in permeabilized mammalian cells requires soluble cytoplasmic factors. J. Cell Biol. 111, 807-816. Akey, C. W., and Radermacher, M. (1993). Architecture of the X e m p u s nuclear pore complex revealed by three-dimensional cryo-electron microscopy. J. Cell Biol. 122, 1-19. Akner, G., Mossberg, K., Sundqvist, K. G., Gustafsson, J.A,, and Wikstrom, A. C. (1992). Evidence for reversible, non-microtubule- and non-microfilament-dependent nu-
334
DONALD B. DEFRANCO et al
clear translocation of hsp90 after heat shock in human fibroblasts. Eur. J . Cell Biol. 58,356-364. Barsony, J., Pike, W. J., Deluca, H. F., and Marx, S. J. (1990). Immunocytology with microwave-fixed fibroblasts shows l a , 25-dihydroxyvitamin D,-dependent rapid and estrogen-dependent slow reorganization of vitamin D receptors. J . Cell Biol. 111,2385-2395. Bataille, N., Helser, T., and Fried, H. M. (1990). Cytoplasmic transport of ribosomal subunits microinjected into the Xenopus laevis oocyte nucleus: A generalized facilitated process. J . Cell B i d . 111, 1571-1582. Beck, C. A., Estes, P. A., Bona, B. J., Muro-Cacho, C. A., Nordeen, S. K., and Edwards, D. P. (1993). The steroid antagonist RU486 exerts different effects on the glucocorticoid and progesterone receptors. Endocrinology 133, 728-740. Borer, R. A., Lehner, C. F., Eppenberger, H. M., and Nigg, E. A. (1989). Major nucleolar proteins shuttle between nucleus and cytoplasm. Cell 56, 379-390. Breeuwer, M., and Goldfarb, D. S. (1990).Facilitated nuclear transport of histone H1 and other small nucleophilic proteins. Cell 60, 999-1008. Bresnick, E. H., Dalman, F. C., Sanchez, E. R., and Pratt, W. B. (1989). Evidence that the 90-kDa heat shock protein is necessary for the steroid binding conformation of the L cell glucocorticoid receptor. J . Biol. Chem. 264, 4992-4997. Brink, M., Humbrer, B. M., DeKloet, R., and Van Driel, R. (1992). The unliganded glucocorticoid receptor is localized in the nucleus, not in the cytoplasm. Endocrinology 130,3575-3581. Cadepond, F., Gasc, J . M., Delahaye, F., Jibard, N., Schweizer, G . G., Segard, M. I., Evans, R., and Baulieu, E. E. (1992). Hormonal regulation of the nuclear localization signals of the human glucocorticosteroid receptor. Exp. Cell Res. 201, 99-108. Chandran, U. R., and DeFranco, D. B. (1992). Internuclear migration of chicken progesterone receptor, but not simian virus-40 large tumor antigen, in transient heterokaryons. Mol. Endocrinol. 6, 837-844. Cidlowski, J. A,, Bellingham, D. L., Powell-Oliver, F. E., Lubahn, D. B., and Sar, M. (1990). Novel antipeptide antibodies to the human glucocorticoid receptor: Recognition of multiple distinct receptor forms in vitro and distinct localization of cytoplasmic and nuclear receptors. Mol. Endocrinol. 4, 1427-1437. Dalman, F. C., Scherrer, L. C., TayIor, L. P., Akil, H., and Pratt, W. B. (1991). Localization of the 90-kDa heat shock protein-binding site within the hormone-binding domain of the glucocorticoid receptor by peptide competition. J. Biol. Chem. 266, 3482-3490. Dauvois, S., White, R., and Parker, M. G. (1993). The antiestrogen ICI 182780 disrupts estrogen receptor nucleocytoplasmic shuttling. J. Cell Sci. 106, 1377-1388. DeFranco, D. B., Qi, M., Borror, K. C., Garabedian, M. J., and Brautigan, D. L. (1991). Protein phosphatase types 1 and/or 2A regulate nucleocytoplasmic shuttling of glucocorticoid receptors. Mol. Endocrinol. 5, 1215-1228. Dingwall, C., and Laskey, R. A. (1991). Nuclear targeting sequences-A consensus? Trends Bzochem. Sci. 18,478-481. Dingwall, C., and Laskey, R. A. (1992). The nuclear membrane. Science 258, 942-947. Dyck, J. A., Maul, G. G., Miller, W. H., Jr., Chen, J. D., Kakizuka, A., and Evans, R. M. (1994). A novel macromolecular structure is a target of the promyelocytP-retinoic acid receptor oncoprotein. Cell 76, 333-343. Featherstone, C., Darby, M. K., and Gerace, L. (1988). A monoclonal antibody against the nuclear pore complex inhibits nucleocytoplasmic transport of protein and RNA in uiuo. J. Cell Biol. 107, 1289-1297, Goldfarb, D. S. (1994). GTPase cycle for nuclear transport. Curr. Biol. 4, 57-60.
NUCLEOCYTOPLASMIC SHUTTLING
335
Gorski, J., Toft,D., Shyamala, G., Smith, D., and Notides, A. (1968). Hormone receptors: Studies on the interaction of estrogen with the uterus. Recent Prog. Horm. Res. 24, 45-80. Guiochon-Mantel, A., Loosfelt, H., Lescop, P., Sar, S., Atger, M., Perrot-Applanat, M., and Milgrom, E. (1989). Mechanisms of nuclear localization of the progesterone receptor: Evidence for interactions between monomers. Cell 57, 1147-1154. Guiochon-Mantel, A., Lescop, P., Christin-Maitre, S., Loosefelt, H., Perrot-Applanat, M., and Milgrom, E. (1991). Nucleocytoplasmic shuttling of the progesterone receptor. EMBO J. 10,3851-3859. Guiochon-Mantel, A., Delabre, K., Lescop, P., and Milgrom, E. (1994). Nuclear localization signals also mediate the outward movement of proteins from the nucleus. Proc. Natl. Acad. Sci. U S A . 91, 7179-7183. Howard, K. J., Holley, S. J.,Yamamoto, K. R., and Distelhorst, C. W. (1990). Mapping the HSP9O binding region of the glucocorticoid receptor. J. Biol. Chem. 265, 1192811935. Hu, L.-M., Bodwell, J.,Hu, J.-M., Orti, E., and Munck, A. (1994). Glucocorticoid receptors in ATP-depleted cells. Dephosphorylation, loss of hormone binding. Hsp9O dissociation and ATP-dependent cycling. J . Biol. Chem. 269, 6571-6577. Hutchinson, K. A., Dittmar, K. D., Czar, M. J., and Pratt, W. B. (1994). Proof that hsp70 is required for the assembly of the glucocorticoid receptor into a heterocomplex with hsp90. J. Biof. Chem. 269, 5043-5049. Imamoto, N., Mastsuoka, Y., Kurihara, T., Kohno, K., Miyagi, M., Sakiyama, F., Okada, Y., Tsunasawa, S., and Yoneda, Y. (1992). Antibodies against 70-kD heat shock cognate protein inhibit mediated nuclear import of karyophilic proteins. J. Cell Bwl. 119, 1047-1061. Ishii, D. N., Pratt, W. B., and Aronow, L. (1972). Steady-state level of the specific glucocorticoid binding component in mouse fibroblasts. Biochemistry 11, 3896-3904. Jackson, D. A., Hassan, A. B., Errington, R. J., and Cook, P. R. (1993). Visualization of focal sites of transcription within human nuclei. EMBO J. 12, 1059-1065. Jensen, E. V., Suzuki, T., Kawashima, T., Stumpf, W. E., Jungblut, P. W., and DeSombre, E. R. (1968). A two step mechanism for the interaction of estradiol with rat uterus. Proc. Natl. Acad. Sci. U.S.A.59, 632-638. Kalderon, D., Richardson, W. D., Markham, A. F., and Smith, A. E. (1984). Sequence requirements for nuclear location of simian virus 40 large-T antigen. Nature (London) 311.33-38. Kang, K. I., Devin, J., Cadepond, F., Jibard, N., Guiochon-Mantel, A,, Baulieu, E.-E., and Catelli, M.-G. (1994). In uiuo functional protein-protein interaction: Nuclear targeted hsp9O shifts cytoplasmic steroid receptor mutants into the nucleus. Proc. Natl. Acad. Sci. U S A . 91,340-344. Khanna-Gupta, A,, and Ware, V. C. (1989). Nucleocytoplasmic transport of ribosomes in a eukaryotic system: Is there a facilitated transport process? Proc. Natl. Acad. Sci. U S A . 06,1791-1795. Kost, S. L., Smith, D. F., Sullivan, W.P., Welch, W. J.,and TOR,D. 0. (1989). Binding of heat shock proteins to the avian progesterone receptor. Mol. Cell. Biol. 9, 38293838. Lanford, R. E., Kanda, P., and Kennedy, R. C. (1986). Induction of nuclear transport with a synthetic peptide homologous to the SV40 T antigen transport signal. Cell 46, 575-582. Luisi, B. F., Xu,W. X., Otwinowski, Z., Freedman, L. P., Yamamoto, K. R., and Sigler P. B. (1991). Crystallographic analysis of the interaction of the glucocorticoid receptor with DNA. Nature (London) 352,497-505.
336
DONALD B. DEFRANCO et al.
Madan, A. P., and DeFranco, D. B. (1993). Bidirectional transport of glucocorticoid receptors across the nuclear envelope. Proc. Nutl. Acad. Sci. U.S.A.90,3588-3592. Mandell, R. B., and Feldherr, C. M. (1990). Identification of two HSP70-related Xenopus oocyte proteins that are capable of recycling across the nuclear envelope. J . Cell Biol. 111, 1775-1783. Martins, V. R., Pratt, W. B., Terracio, L., Hirst, M. A,, Ringold, G. M., and Housley, P. R. (1991). Demonstration by confocal microscopy that unliganded overexpressed glucocorticoid receptors are distributed in a nonrandom manner throughout all planes of the nucleus. Mol. Endocrinol. 5, 217-225. Meier, U. T., and Blobel, G. (1992). Noppl40 shuttles on tracks between nucleolus and cytoplasm. Cell 70, 127-138. Melchior, F., Paschal, B., Evans, J., and Gerace, L. (1993). Inhibition of nuclear protein import by nonhydrolyzable analogues of GTP and identification of the small GTPase RanITC4 as an essential nuclear factor. J . Cell Biol. 123, 1649-1659. Mendel, D. B., Bodwell, J. E., and Munck, A. (1986). Glucocorticoid receptors lacking hormone-binding activity are bound in nuclei of ATP-depleted cells. Nature (London) 324,478-480. Meyer, B. E., and Malim, M. H. (1994). The HIV-1 Rev trans-activator shuttles between the nucleus and the cytoplasm. Genes Deu. 8, 1538-1547. Milarski, K. L., and Morimoto, R. I. (1989). Mutational analysis of the human hsp7O protein: Distinct domains for nucleolar localization and adenosine triphosphate binding. J. Cell Biol. 109, 1947-1962. Moore, M. S., and Blobel, G. (1993). The GTP-binding protein RaniTC4 is required for protein import into the nucleus. Nature (London) 365, 661-663. Munck, A., and Foley, R. (1976). Kinetics of glucocorticoid-receptor complexes in rat thymus cells. J . Steroid Biochem. 7, 1117-1122. Munck, A,, and Holbrook, N. J. (1984). Glucocorticoid-receptor complexes in rat thymus cells: Rapid kinetic behavior and a cyclic model. J. Biol. Chem. 259, 820-831. Newmeyer, D. D., and Forbes, D. J. (1988). Nuclear import can be separated into distinct steps in uitro: Nuclear pore binding and translocation. Cell 52, 641-653. Newmeyer, D. D., Lucocq, J. M., Burglin, T.R., and Roberts, E. M. D. (1986). Assembly in uitro of nuclei active in nuclear protein transport: ATP is required for nucleoplasmin accumulation. EMBO J . 5, 501-510. Nigg, E. A., Baeuerle, P. A,, and Luhrmann, R. (1991). Nuclear import-export: In search of signals and mechanisms. Cell 66, 15-22. Okuno, Y., Imamoto, N., and Yoneda, Y. (1993). 70-kDa heat-shock cognate protein colocalizes with karyophilic proteins into the nucleus during their transport in uitro. Exp. Cell Res. 206, 134-142. Orti, E., Mendel, D. B., Smith, L. I., Bodwell, J. E., and Munck, A. (1989). A dynamic model of glucocorticoid receptor phosphorylation and recycling in intact cells. J. Steroid Biochem. 34, 85-96. Perrot-Applanat, M., Groyer-Picard, M.-T., Logeat, F., and Milgrom, E. (1986). Ultrastructural localization of the progesterone receptor by an immunogold method: Effect of hormone administration. J. Cell B i d . 102, 1191-1199. Picard, D., and Yamamoto, K. R. (1987). Two signals mediate hormone-dependent nuclea r localization of the glucocorticoid receptor. EMBO J . 6, 3333-3340. Picard, D., Khursheed, B., Garabedian, M. J., Fortin, M. G., Lindquist, S., and Yamamoto, K. R. (1990a). Signal transduction by steroid receptors: Reduced levels of hsp90 compromise receptor action in uiuo. Nature (London) 348, 166-168. Picard, D., Kumar, V., Chambon, P., and Yamamoto, K. R. (1990b). Signal transduction
NUCLEOCYTOPLASMIC SHUTTLING
337
by steroid hormones: Nuclear localization is differentially regulated in estrogen and progesterone receptors. Cell Regul. 1, 291-299. Pratt, W. B. (1990). Interaction of hsp90 with steroid receptors: Organizing some diverse observations and presenting the newest concepts. Mol. Cell. Endocrinol. 74, C69C76. Pratt, W. B. (1992). Control of steroid receptor function and cytoplasmic-nuclear transport by heat shock proteins. BwEssays 14, 841-848. Pratt, W. B. (1993). The role of heat shock proteins in regulating the function, folding, and traflicking of the glucocorticoid receptor. J. B w l . Chem. 268, 21455-21458. Qi, M., Hamilton, B. J., and DeFranco, D. (1989). v-mos oncoproteins affect the nuclear retention and reutilization of glucocorticoid receptors. Mol. Endocrinol. 3, 12791288. Raaka, B. M., and Samuels, H. H. (1983). The glucocorticoid receptor in GH3 cells. J. Biol. Chem. 258,417-425. Richardson, W. D., Mills, A. D., Dilworth, S. M., Laskey, R. A., and Dingwall, C. (1988). Nuclear protein migration involves two steps: Rapid binding a t the nuclear envelope followed by slower translocation through nuclear pores. Cell 62, 655-664. Robbins, J., Dilworth, S. M., Laskey, R. A., and Dingwall, C. (1991). Two independent basic domains in nucleoplasmin nuclear targeting sequence: Identification of a class of bipartite nuclear targeting sequences. Cell 64, 615-623. Sanchez, E. R., Hirst, M., Scherrer, L. C., Tang, H. Y., Welsh, M. J., Harmon, J. M., Simons, S. S. J., Ringold, G. M., and Pratt, W. B. (1990). Hormone-free mouse glucocorticoid receptors overexpressed in Chinese hamster ovary cells are localized to the nucleus and are associated with both hsp70 and hsp9O. J. Biol. Chem. 265, 20123-20130. Scherrer, L. C., Dalman, F. C., Massa, E., Meshinchi, S.,and Pratt, W. B. (1990). Structural and functional reconstitution of the glucocorticoid receptor-hsp90 complex. J. B b l . Chem. 265,21397-21400. Schmidt-Zachmann, M. S., Dargemont, C., Kuhn, L. C., and Nigg, E. A. (1993). Nuclear export of proteins: The role of nuclear retention. Cell 74, 493-504. Shaknovich, R., Shue, G., and Kohtz, S. (1992). Conformational activation of a basic helix-loop-helix protein (MyoD1) by the C-terminal region of murine HSP9O (HSP84). Mol. Cell. Biol. 12, 5059-5068. Shi, Y., and Thomas, J. 0. (1992). The transport of proteins into the nucleus requires the 70-kilodalton heat shock protein or its cytosolic cognate. Mol. Cell Biol. 12, 21862192. Smith, D. F. (1993). Dynamics of heat shock protein 90-progesterone receptor binding and the disactivation loop model for steroid receptor complexes. Mol. Endocrinol. 7, 1418-1429. Smith, D. F., and Toft, D. 0. (1993). Steroid receptors and their associated proteins. Mol. Endocrinol. 7 , 4-11. Sommer, L., Hagenbuchle, O., Wellauer, P. K., and Strubin, M. (1991). Nuclear targeting of the transcription factor PTFl is mediated by a protein subunit that does not bind to the PTFl cognate sequence. Cell 67, 987-994. Spector, D. L. (1993). Macromolecular domains within the cell nucleus. Annu. Rev. Cell Biol. 9, 265-315. Stochaj, U.,Bossie, M. A., ven Zee, K., Whalen, A. M., and Silver, P. A. (1993). Analysis of conserved binding proteins for nuclear localization sequences. J. Cell sci. 104, 89-95. Tuohimaa, P., Pekki, A,, Blauer, M., Jonensuu, T., Vilja, P., and Ylikomi, T. (1993).
338
DONALD B. DEFRANCO et al
Nuclear progesterone receptor is mainly heat shock protein 90-free in uiuo. Proc. Natl. Acad. Sci. U.S.A. 90, 5848-5852. Weis, K., Rambaud, S., Lavau, C., Jansen, J., Carvalho, T., Carmo-Fonseca, M., Lamond, A,, and Dejean, A. (1994). Retinoic acid regulates aberrant nuclear localization of PML-RARa in acute promyelocytic leukemia cells. Cell 76, 345-356. Welshons, W., Lieberman, M. E., and Gorski, J . (1984). Nuclear localization of unoccupied estrogen receptors: Cytochalasin enucleation of GH3 cells. Nature (London) 307,747-749. Welshons, W. V., Krummel, B. M., and Gorski, J . (1985). Nuclear localization of unoccupied receptor for glucocorticoids, estrogens and progesterone in GH3 cells. Endocrinology 117, 2140-2147. Wikstrom, A.-C., Bakke, O., Okret, S., Bronnegard, M., and Gustafsson, J.-A. (1987). Intracellular localization of the glucocorticoid receptor: Evidence for cytoplasmic and nuclear localization. Endocrinology 120, 1232-1242. Yamasaki, L., Kanda, P., and Lanford, R. E. (1989). Identification of four nuclear transport signal-binding proteins that interact with diverse transport signals. Mol. Cell. Biol. 9, 3028-3036. Yang, J., and DeFranco, D. B. (1994). Differential roles of heat shock protein 70 in the in uitro nuclear import of glucocorticoid receptor and simian virus 40 large tumor antigen. Mol. Cell. Biol. 14, 5088-5098. Ylikomi, T., Bocquel, M. T., Berry, M., Gronemeyer, H., and Chambon, P. (1992). Cooperation of proto-signals for nuclear accumulation of estrogen and progesterone receptors. EMBO J . 11, 3681-3694. Zhao, L.-J., and Padmanabhan, R. (1988). Nuclear transport of adenovirus DNA polymerase is facilitated by interaction with preterminal protein. Cell 55, 1005-1015. Zhou, Z.-X., Sar, M., Simental, J. A,, Lane, M. V., and Wilson, E. M. (1994). A liganddependent bipartite nuclear targeting signal in the human androgen receptor. J . Biol. Chem. 269, 13115-13123.
VITAMINS AND HORMONES,VOL. 51
Transcriptional Regulation of the Genes Encoding the Cytochrome P-450 Steroid Hydroxylases KEITH L. PARKER* AND BERNARD P. SCHIMMERt *Departments of Medicine and Pharmacology and the Howard Hughes Medical Institute Duke University Medical Center Durham, North Carolina 27710 fBanting and Best Department of Medical Research and the Department of Pharmacology University of Toronto Toronto, Ontario, Canada M5G lL6
I. Introduction 11. Overview of Steroid Hormone Biosynthesis A. Pathways B. Regulation 111. Cell-Selective Expression A. SF-1 B. Other Regulators IV. Hormone-Regulated Expression A. CRE-Binding Protein B. Nuclear Receptor Proteins C. Pbxl D. Spl V. Perspectives and Future Directions VI. Summary Fkferences
I. INTRODUCTION Steroid hormones are essential regulators of such diverse processes as accommodation to stress, carbohydrate metabolism, fluid and electrolyte balance, and reproduction; therefore, understanding the mechanisms that regulate steroid hormone production represents an important objective of endocrinological research. Previous studies have indicated that steroid hormone biosynthesis is controlled at two levels: at the level of substrate mobilization, reflecting the acute control of steroid hormone biosynthesis, and at the level of gene transcription, reflecting a more long-term regulation of steroidogenesis. This review focuses on recent studies that provide intriguing insights into the mechanisms that regulate the expression of the genes encoding the steroidogenic cytochrome P-450 enzymes. In particular, we consider 339
Copyright I? 1995 by Academic Press, Inc All rights of reproduction in any farm reserved.
340
KEITH L. PARKER A N D BERNARD P. SCHIMMER
two distinct aspects of transcriptional regulation that govern the expression of the genes encoding the steroid hydroxylases: cell-selective expression that restricts the steroid hydroxylases to steroidogenic tissues and induction of these enzymes by trophic hormones. We recognize that it may be overly simplistic to separate these transcriptional regulators into those that determine cell-selective expression and those that mediate hormonal induction, since certain factors apparently contribute to both levels of regulated expression.
11. OVERVIEW OF STEROID HORMONE BIOSYNTHESIS A. PATHWAYS All endogenous steroids are derived from cholesterol and therefore have the cyclopentanoperhydrophenanthrenenucleus as part of their basic structure. Modifications to this basic nucleus generate specific ligands for the members of the nuclear receptor family of transcription factors, accounting for their specific actions in different target tissues. For example, C-21 (pregnane) steroids have progestational, glucocorticoid, or mineralocorticoid activities; C-19 (androstane) steroids have virilizing and anabolic activities; and C-18 aromatic steroids have estrogenic activities. The stepwise conversion of cholesterol t o steroid hormones requires the sequential action of a series of enzymes (Fig. 11, most of which belong to the cytochrome P-450 superfamily of mixed-function oxidases (Nebert and Gonzalez, 1987). These heme-containing P-450 enzymes use electrons derived from NADPH to activate molecular oxygen so that one oxygen atom is incorporated into substrate while the other oxygen atom is reduced to water. For mitochondria1 P-450 cytochromes electron transfer is carried out by adrenodoxin, a n ironsulfur protein, and adrenodoxin reductase, a flavoprotein. For P-450 cytochromes localized to the endoplasmic reticulum, only the flavoprotein cytochrome P-450 reductase is required for electron transfer. The cytochrome P-450 superfamily is well known for its ability to metabolize a diverse array of exogenous and endogenous substrates, and the ability of a relatively small group of genes to metabolize a n enormous number of compounds reflects the catalytic versatility of these cytochrome P-450 enzymes (Nebert, 1991; Nelson et al., 1993). For many of the steroidogenic P-450 enzymes, this catalytic versatility is reflected in their ability to hydroxylate steroids at more than one position and even to break carbon-carbon bonds.
341
CYTOCHROME P-450STEROID HYDROXYLASES
Adrenal Cortex Zona Glomerulosa
Zonae FasciculatalReticularis
Testis
Ovary FIG.1. Overview of biosynthetic pathways for steroid hormones. (Top) The pathways by which adrenal corticosteroids are derived from cholesterol in the zona glomerulosa and the zonae fasciculata/reticulais. (Bottom) The pathways by which sex steroids are produced in the gonads. P450,,, Cholesterol side chain cleavage enzyme; 3P-HSD, 3P-hydroxysteroid dehydrogenase; P450,,,, steroid 1711-hydroxylase; P450,,, steroid 21hydroxylase; P450,,,,, aldosterone synthase; P450,,,, steroid llp-hydroxylase; 17PHSD, 17P-hydroxysteroid dehydrogenase.
1. Adrenal Cortex The pathways of steroid hormone biosynthesis in the adrenal cortex are summarized in Fig. 1. Functionally and anatomically, the adrenal cortex is divided into two discrete compartments: the outer zona glomerulosa, which synthesizes mineralocorticoids, and the inner zonae fasciculata/reticularis,which synthesize glucocorticoids and androgens. Despite differences in steroid products and regulators, a number of steps in the synthesis of these two steroid classes in the two
342
KEITH L. PARKER AND BERNARD P. SCHIMMER
compartments are quite similar, and so the two pathways are described together here where possible. The initial and rate-limiting step in the biosynthesis of the adrenal steroid hormones is the conversion of cholesterol to pregnenolone. This reaction is catalyzed by the mitochondria1 cholesterol side chain cleavage enzyme (P-450,,,) and involves three distinct reactions (20ahydroxylation, 22'Rhydroxylation, and C-20,C-22lyase) that occur at a single active site in the P-450,,, molecule. P-450,,, is also expressed in other tissues capable of d e nouo steroid synthesis, such as ovary, testis, and placenta, but is generally not expressed in nonsteroidogenic tissues. Pregnenolone and 17a-hydroxypregnenolone are converted to progesterone and 17a-hydroxyprogesterone, respectively, by the microsoma1 enzyme 3P-hydroxysteroid dehydrogenase (3P-HSD).3P-HSD, one of the few non-cytochrome P-450 steroid-biosynthetic enzymes, is essential for the production of all major physiological steroid hormones. This enzyme carries out two reactions: the oxidation of 3P-hydroxyl groups and the isomerization of C-5,6 double bonds to generate 3-keto-A4 steroids. Unlike most of the other steroidogenic enzymes, 3p-HSD activity is widely distributed and is found not only in the primary steroidogenic tissues (adrenal cortex, gonad, and placenta) but also in breast, prostate, liver, and skin (Labrie et al., 1992). Recent studies have identified distinct isozymes of 3P-HSD that are differentially expressed in the various tissues, with one enzyme (type 11) apparently serving as the predominant, if not the only, source of activity in the primary steroidogenic tissues, as evidenced by the dramatic effects of inborn errors of the type I1 gene (Rheaume et al., 1992). The microsomal enzyme 17a-hydroxylase (P-45017,)is expressed in the inner zones of the adrenal cortex, where it serves as a branch point in the formation of both glucocorticoids and adrenal androgens. Interestingly, P-45OI7, is not expressed in the inner adrenal zones of rats or mice, so these rodents synthesize corticosterone as the major glucocorticoid and do not form adrenal androgens. P-45Ol,, is responsible for the hydroxylation of pregnenolone to 17a-hydroxypregnenolone and for the subsequent C-17,C-20 lyase reaction that converts 17a-hydroxypregnenolone to dehydroepiandrosterone. Site-directed mutagenesis (Kitamura et al., 1991) has been used to dissociate the hydroxylase activity of P-45OI7, from the lyase activity and to suggest a separation of these activities within the active site of the enzyme. The microsomal steroid 21-hydroxylase (P-4502,) is responsible for the conversion of 17a-hydroxyprogesterone to 1l-deoxycortisol in the inner glucocorticoid-producing zones. This same enzyme also partici-
CYTOCHROME P-450STEROID HYDROXYLASES
343
pates in mineralocorticoid biosynthesis in the zona glomerulosa, metabolizing 1l-deoxycorticosterone to corticosterone. The gene encoding P-4502, (CYP21) is expressed only in the adrenal cortex and thus carries out hydroxylations unique to the corticosteroids. CYP21 is localized to that portion of the genome encoding the class I11 genes of the major histocompatibility complex (MHC) in humans and mice and is duplicated such that a wild-type gene and a pseudogene are present in tandem array within the MHC. Mutations in CYP21 occur at relatively high frequency in the human population (15000 t o 1:10,000 for the severe form), frequently resulting from gene conversions between the wild-type gene and its neighboring pseudogene (Miller, 1994; White, 1994). The terminal reactions in the biosynthesis of both mineralocorticoids and glucocorticoids are carried out by two closely related mitochondrial enzymes: steroid 1lp-hydroxylase (P-450,,,), a product of the C Y P l l B l gene, and aldosterone synthase (P-450a1,,),a product of the CYPIIB2 gene. In the zona glomerulosa P-450ald,performs successive llp-hydroxylation, 18-hydroxylation, and 18-oxidation reactions to convert 1l-deoxycorticosterone to aldosterone. In the inner zones, by contrast, P-45Ol,, predominantly performs a single reaction converting 1l-deoxycortisol to cortisol. The expression of these functionally distinct C Y P l l B genes in the appropriate cortical zones plays an essential role in the gland's ability to regulate separately the biosynthesis of mineralocorticoids and of glucocorticoids. The bovine C Y P l l B genes appear t o be the exception, in that the P-450,,,. isoforms from this species contribute to both glucocorticoid and mineralocorticoid biosynthesis (Ogishima et al., 1989). 2. Testis
The predominant steroidogenic cells in the testis are the androgenproducing Leydig cells, which are located in the interstitial regions around the seminiferous tubules. Androgen biosynthesis requires many of the sam'e cytochrome P-450 steroid hydroxylases already discussed for adrenal steroid biosynthesis. Thus, P-450,,,, 3P-HSD, and P-45ol7, all play essential roles in androgen biosynthesis (Fig. 1). Following the conversions of 17a-hydroxypregnenolone to dehydroepiandrosterone and then to androstenedione, these steroids are metabolized by 17P-HSD to form the potent androgen testosterone. Although testosterone binds the androgen receptor with high affinity, certain components of androgen action are mediated by 5a- and 5P-dihydro derivatives of testosterone. For example, the normal virilization of the male external genital system is critically dependent on 5a-reductase
344
KEITH L. PARKER AND BERNARD P. SCHIMMER
activity, and a deficiency of the isozyme normally expressed in the developing male genital structures results in abnormal genital development with male pseudohermaphroditism (J.D. Wilson et al., 1993). The testis also exhibits aromatase (P-450a,,,) activity, produces detectable levels of P-450a,,, transcripts, and therefore synthesizes small amounts of estrogen (Bulun et al., 1994). The microsomal P-450a,,, catalyzes a series of reactions, including C-19 hydroxylation, removal of the C-19 hydroxymethyl group, and aromatization of the A ring of the steroid nucleus. There is uncertainty over the site of estrogen biosynthesis in the testis; Leydig cells, Sertoli cells, and germ cells have all been implicated (for a brief discussion see Nitta et al., 1992).
3. Ovary The basic functional unit of the ovary is the follicle, and the predominant hormones produced are estrogens and progesterone. Within the follicle there are two distinct steroidogenic cell types: an outer layer of theca interna cells surrounding an inner layer of granulosa cells. Like those of the testis, ovarian steroids are produced by the action of many of the same enzymes that make adrenal steroids (Fig. 1). One unique feature of ovarian steroidogenesis that distinguishes it from the adrenocortical cell is the obligatory interaction of two different cell types (Armstrong and Dorrington, 1977; Liu and Hsueh, 1986). The general organization is that thecal cells, predominantly under the control of luteinizing hormone (LH), secrete androgens such as androstenedione. The theca-derived androgens are then metabolized to estrogens by granulosa cells that express P-450,,,, under the control of folliclestimulating hormone (FSH). As follicles mature the granulosa cells acquire LH receptors, become responsive to LH, and produce both estrogens and progesterone under LH control. Ultimately, high levels of LH lead to follicular rupture and to luteinization, such that progesterone biosynthesis is maintained in the absence of ongoing trophic hormone stimulation. The preservation of progesterone biosynthesis by the corpus luteum is essential for the normal maintenance of pregnancy in several species, including humans (Solomon, 1994). 4. Placenta There is considerable interspecies variation in the relative role of the placenta in steroidogenesis. In humans and other primates placental steroidogenesis is a major source of circulating steroids in both the maternal and fetal blood, and these steroids are believed to play a major role in the maintenance of pregnancy and possibly also in par-
CYTOCHROME P-450STEROID HYDROXYLASES
345
turition. Important steroid products include progesterone, which is produced de nouo in the placenta by the actions of P-45OsCcand 3P-HSD, and estrogens, which are derived largely from fetal precursors by the action of aromatase. The primate placenta does not express P-45o1,,, and thus cannot produce androgens or estrogens de novo (Walsh et al., 1980). To date, the hormonal regulators of placental steroidogenesis have not been defined, although it is clear that elevations in CAMP levels can modulate the expression of P-45OsCcin JEG-3 human choriocarcinoma cells. As discussed in Section 111,B, 1,promoter analyses of the CYPllAl gene indicate that the regulatory elements that participate in the expression of P-45OsCcin placental JEG-3 cells are different from those that regulate the expression of P-45OsCcin adrenal cells (Moore et al., 1992; Guo et al., 19941, thereby suggesting that the expression of the steroidogenic enzymes is controlled differently in the two cell types.
5 . Brain Although the brain is not classically considered a steroidogenic organ, discrete regions can synthesize steroid hormones de nova The diverse actions attributed to these neurosteroids include stimulation or inhibition of chloride transport by the type A y-aminobutyric acid receptor, increases in gonadotropin-releasing hormone (GnRH) release from the hypothalamus, and enhancement of memory (Mellon, 1994). Consistent with the concept of de nouo steroid biosynthesis in the brain, a number of steroidogenic enzymes that are expressed in the primary steroidogenic organs are also expressed in the brain. Enzymes that have been implicated in neurosteroid biosynthesis include P-45OsCc,3P-HSD, P-4501,,, aromatase, and 5a-reductase (Mellon and Deschepper, 1993; Robel and Baulieu, 1994). P-450,, appears to be expressed predominantly in oligodendrocytes, whereas aromatase is present predominantly in neuronal cells. To date, there have been no reports on the regulation of the genes encoding these steroidogenic enzymes, in either neuronal or oligodendrocyte cell lines. 6 . Uterus Recent studies have shown that the rat uterus expresses high levels of P-45OsCctranscripts at early postimplantation stages (Schiff et al., 1993). These studies showed that the P-45OscC-expressingcells were of maternal origin and did not directly require any blastocyst contribution; based on these results, a role for locally produced steroids in the maintenance of pregnancy was proposed. As with the placenta and the
346
KEITH L. PARKER AND BERNARD P. SCHIMMER
brain, there is currently no information about hormonal regulators or transcriptional mechanisms that determine P-45OsCcexpression in the decidualized uterine cells.
B. REGULATION Steroid hormone production by the adrenal glands and the gonads is regulated by trophic protein hormones that stimulate steroidogenesis in their corresponding target tissues. These trophic hormones include adrenocorticotropic hormone (ACTH), which primarily regulates glucocorticoid biosynthesis; the gonadotropins FSH and LH, which regulate gonadal hormone biosynthesis; and angiotensin I1 (AII), which primarily regulates mineralocorticoid synthesis. These peptide hormones stimulate steroidogenesis by mobilizing substrate cholesterol and translocating this cholesterol across the mitochondrial membrane to the P-45OsCc, which is poised to carry out the first step in the steroidbiosynthetic pathway (Fig. 1).The translocation of cholesterol across the outer mitochondrial membrane to the inner mitochondrial cisternae, the rate-limiting step in this aspect of steroidogenesis, is under acute regulation by the trophic hormones. Considerable progress has been made in identifying potential mediators that deliver cholesterol to the P-45OsCc. For ACTH and the gonadal hormones the acute regulation of steroidogenesis is obligatorily dependent on the cAMP and CAMP-dependent protein kinase signaling cascade. One consequence of cAMP action is the rapid induction of a phosphoprotein that translocates to the mitochondria and becomes further processed in a manner that correlates with enhanced steroidogenesis. Originally identified on the basis of its migration on two-dimensional gels (Pon et al., 1986; Stocco and Kilgore, 1988), a cDNA encoding this protein (designated the steroidogenic acute regulatory protein, or StAR) has recently been isolated and characterized; in transfection experiments expression of this cDNA enhanced steroid production (Clark et al., 1994).Other proteins have been implicated in the acute response to ACTH, including the peripheral benzodiazepine receptor (Krueger and Papadopoulos, 1990; Garnier et al., 1994), steroidogenesis-activatorpeptide (Pedersen and Brownie, 19871, and sterol carrier protein (Pfeifer et al., 19931, and it is entirely plausible that interactions between these proposed mediators are needed to sustain the acute response to trophic hormones. Regulation of mineralocorticoid biosynthesis by AII, on the other hand, seems to involve Caz+-dependent signal transduction events that are less well defined (Barrett et al., 1989). As outlined above, the steroidogenic P-450cytochromes exhibit vary-
CYTOCHROME P-450 STEROID HYDROXYLASES
347
ing patterns of expression, particularly among the classical steroidogenic tissues (adrenal cortex, testis, ovary, and placenta). P-450,,, is found in all of the classical steroidogenic tissues. P-450,,,, likewise, is expressed in multiple steroidogenic tissues, but it is absent from the placenta and is variably expressed in the adrenal cortex, depending on in contrast, are expressed the species. p-45O2,,P-450,,,, and P-45OaldO, only in the adrenal cortex; in the adrenal glands of most species, P-450,,, and P-45oa1,, are further restricted to the zonae fasciculatah-eticularis or the zona glomerulosa, respectively. P-450ar0,, the most widely distributed steroidogenic P-450 enzyme, is found in the gonads, in the placenta, and in a wide range of nonsteroidogenic tissues (Simpson et al., 1994). These overlapping profiles of expression suggest that steroidogenesis is governed, in part, by factors that regulate the cell-selective expression of genes encoding the steroidogenic cytochrome P-450 enzymes. The levels of steroidogenic cytochrome P-450 enzymes within the steroidogenic tissues are also regulated by trophic protein hormones through transcriptional and posttranscriptional mechanisms (John et al., 1986; Boggaram et al., 19891, and thus reflect a more long-term regulation of steroidogenesis that is associated with maintenance of the functional and structural integrity of the various steroidogenic tissues. It is clear that CAMP and CAMP-dependent protein kinases mediate many of these trophic effects on gene expression.
111. CELL-SELECTIVE EXPRESSION Insights into the molecular mechanisms responsible for the cellselective expression of the steroidogenic cytochrome P-450 enzymes first came from the analysis of the mouse Cyp21 gene in transfection assays using the Y 1mouse adrenocortical tumor cell line as an expression system. These studies identified a proximal promoter region of Cyp21 that directed both constitutive and ACTH-induced gene expression in the mouse adrenal cell line but not in cell lines from other steroidogenic or nonsteroidogenic tissues. Within this promoter region multiple closely spaced cis-acting elements were shown to be required for gene expression (Rice et al., 1990a). Among these, variations of an AGGTCA sequence motif were shown to be major determinants of the cell-selective expression of Cyp21. Similar AGGTCA sequence motifs then were found to be essential components of cell-selective gene expression in both adrenocortical cells [CypllA (Rice et al., 1990b), Cypll b2 (Bogerd et al., 19901, and CYPll bl (Honda et al., 199011 and
348
KEITH L. PARKER AND BERNARD P. SCHIMMER SF.1 COUP-TF
-210
SF-1
SF-1 NGFI-8
-65
-140
SF-1
SF-1
-70
-40
CypllAl I
SF-1
Cypllb2 1
r
CREE
m -310
-55
FIG.2. Model of nuclear receptor binding sites within the promoter regions of three mouse steroid hydroxylases. The positions and sequences of selected promoter elements that regulate three mouse steroid hydroxylases via interactions with nuclear receptor proteins are indicated. Also shown is a binding site for the CAMP-responsive elementbinding protein (CREB). SF-1, Steroidogenic factor 1; COUP-TF, chicken ovalbumin upstream promoter-transcription factor; NGFI-B, nerve growth factor inducible-B.
gonadal cells [CYPIS (Fitzpatrick and Richards, 1993; Lynch et al., 1993) and C Y P l l A (Clemens et al., 199411, as reviewed by Parker and Schimmer (1993). DNase footprinting and gel mobility shift experiments demonstrated that these AGGTCA sequence motifs all interacted with a common nuclear DNA-binding protein, alternatively designated steroidogenic factor 1 (SF-1; Lala et al., 1992) or adrenal 4-binding protein (Ad4BP; Morohashi et al., 1992). A schematic summary of the known SF-l-responsive elements that regulate the mouse steroid hydroxylases is shown in Fig. 2. This specific DNA binding activity was found in steroidogenic tissues but not in a wide range of nonsteroidogenic cell types, providing the first clue that SF-1 was an important determinant in the cell-selective expression of these cytochrome P-450 enzymes.
A. SF-1 A definitive role for SF-1 in steroidogenesis was established following the cloning and sequencing of SF-1 cDNAs in two different laboratories using two different strategies. Inasmuch as the AGGTCA binding motif of SF-1 resembled the “half-site” motif for nuclear receptors of the estrogedthyroid hormone receptor family (Evans, 19881, Lala and colleagues reasoned that the DNA binding domains of SF-1 and
349
CYTOCHROME P-450STEROID HYDROXYLASES
members of the nuclear hormone receptor family might be homologous. Using the DNA binding domain of the retinoic acid X receptor p (RXRP)as a probe, these workers were able to isolate and identify an SF-1 cDNA by virtue of its steroidogenic cell-selective distribution (Lala et al., 1992). Morohashi, Omura, and colleagues used sequencespecific DNA affinity chromatography to purify the DNA-binding protein Ad4BP from bovine adrenal cortices, obtained a partial amino acid sequence, and screened a bovine adrenal cortex cDNA library with degenerate oligonucleotides corresponding to the peptide sequence (Honda et al., 1993). The nucleotide sequences of the SF-1 and Ad4BP cDNAs closely resembled that of a cDNA isolated from embryonal carcinoma cells, which was designated embryonal long terminal repeat (LTR)-binding protein (ELP) because of its ability to bind a negative regulatory element in retroviral LTRs (Tsukiyama et al., 1992). As shown in Fig. 3, isolation and characterization of genomic sequences encoding mouse SF-1ultimately showed that SF-1 and ELP
FIZ-F1 Box
TI
D
m
I
SF-1 ELP
Zn flngar FIG.3. Organization of the mouse F b - F l gene. (Top) Exon-intron organization of the mouse F b - F l gene, which encodes both SF-1 and embryonal long terminal repeatbinding protein (ELP). The positions of the initiator methionine and stop codons of both transcripts are indicated. Black areas indicate exons contained in both transcripts, white regions indicate SF-1-specific exons, and cross-hatched areas indicate ELP-specific exons. ITZ-Fl Box, an additional region implicated in DNA binding. (Bottom) Structural features of the ELP and SF-1 cDNAs, with putative structural motifs indicated according to the terminology of Laudet et al. (1992). Zn Finger, Zinc finger DNA binding domain; FTZ-F1 Box, the non-zinc finger region implicated in DNA binding; Ti, the conserved region of the ligand binding domain; D, putative dimerization domain not contained in the ELP transcript.
350
KEITH L. PARKER AND BERNARD P. SCHIMMER
transcripts arise from the same structural gene by alternative promoter usage and 3' splicing (Ikeda et al., 1993). Within both proteins are regions that match the sequences of functional domains of other members of the nuclear receptor family, including the zinc finger DNA binding domains and ligand binding domains associated with receptor activation. These proteins are classified as orphan members of the nuclear receptor gene family, since ligands responsible for activation of these transcription factors have not been identified (OMalley and Conneely, 1992). Additional notable features include a proline-rich region that may be involved in transcriptional activation, a potential phosphorylation consensus site for CAMP-dependent protein kinase, and binding specificity for two distinct sequence subsets: PyCAAGGPyC or PuPuAGGTCA. Transfection of nonsteroidogenic cell types with expression vectors encoding SF-1 enabled these cells to express the C Y P l l A and CYPllB genes, providing evidence for the functional importance of SF-1 in the cell-selective expression of the steroid hydroxylase genes (Morohashi et al., 1993). The mouse gene encoding SF-1 and ELP was designated Ftz-F1 because it most closely resembles a Drosophila nuclear receptor that also encodes two developmentally regulated nuclear receptors (Ueda et al., 1990; Lavorgna et al., 1991, 1993). By analogy with the developmentally regulated Drosophila Ftz-F1 gene, and in view of the known role of steroid hormones in mammalian embryonic development, it seemed reasonable to postulate a role for the mouse Ftz-Fl gene products during development. As summarized in Fig. 4, SF-1 transcripts were present in adrenal cells from the earliest time that the adrenal primordium is recognizable [embryonic day 10.5 (E10.511. Thereafter, as the primitive sympathoadrenal cells invade the adrenal gland to form the medulla, SF-1 expression is confined to the cortex, where steroidogenesis occurs. The early onset of SF-1 expression in the adrenal primordium precedes the onset of P-450,,, expression and is consistent with the essential role of SF-1in steroid hydroxylase gene expression and adrenocortical function (Ikeda et al., 1994). Male sexual differentiation requires the biosynthesis of two hormonal mediators: androgens, produced by Leydig cells in the fetal testis, and mullerian-inhibiting substance (MIS), produced by fetal Sertoli cells (reviewed by Josso, 1992).At the earliest stages of gonadal formation, testes and ovaries are histologically indistinguishable, and thus are termed the indifferent, or bipotential, gonad. Thereafter, the fetal testis becomes recognizable as it develops the testicular cords. At the earliest stage of gonadal organogenesis (E9.0-E9.5), faint expression of SF-1 was seen in all embryos, regardless of genetic sex. This
351
CYMCHROME P-450STEROID HYDROXYLASES
Tdl3
E9
El0
Ell
El2
El3
El4
El5
El6
El8
Pl
Adult
+
+
+
+
+
+
+
+
+
+
+
-
+
+
+
Indifferent shge
Sexual Differentiation
ovary
+
+
+
+
-
Adrenal
+
+
+
+
+
+
+
+
+
+
+
Dienaphalw
+1
+
+
+
+
+
+
+
9
+
+
-
+
+
+
+
+
+
Pituitary
FIG.4. Pattern of SF-1 expression during mouse embryonic development. In situ hybridization analyses with transcript-specific probes were used to study the profile of expression of both transcripts encoded by the Ft.-Fl gene. - and +, Absence or presence, respectively, of SF-1 transcripts a t the indicated times. Times are given as embryonic days, with noon of the day on which the coital plug was detected considered E0.5. 1SF-1 expression at this stage is seen in the neural plate. 2SF-1 expression a t this stage is localized to the ventromedial hypothalamic nucleus.
expression persisted throughout the indifferent gonad stage. At E12.5, coincident with formation of the testicular cords, SF-1 expression persists in testes but is extinguished in ovaries. This sexually dimorphic pattern suggests that SF-1 is very important in male sexual differentiation and that persistent expression of SF-1 in females might have deleterious effects. Moreover, SF-1 was expressed both in the interstitial region containing fetal Leydig cells and in the testicular cords that contain fetal Sertoli cells and primordial germ cells. An explanation for the expression of SF-1 within the testicular cords occurred when it was shown that SF-1 is expressed in Sertoli cells, where it regulates the expression of the MIS gene by binding a proximal promoter element (Shen et al., 1994).This finding thus implicates SF-1in roles that extend considerably beyond the regulation of androgen production. In contrast to SF-1, the functional significance of ELP has been more elusive. ELP transcripts have not been detected in postimplantation mouse embryos by in situ approaches (Ikeda et al., 1994). Similarly, the more sensitive reverse transcription-polymerase chain reaction (RT-PCR) assay did not detect ELP transcripts in mouse embryonal stem cells (D. S. Milstone, personal communication). These results suggest either that ELP does not play a major role during development or that its expression is confined to a precise stage of development that was missed in these studies. ELP transcripts were
352
KEITH L. PARKER AND BERNARD P. SCHIMMER
detected by RT-PCR in embryonic carcinoma cells and in the adrenal glands, ovaries, and testes of adult rats; however, ELP was unable to support transcription from SF-1 sites in functional assays, probably because of its weak ability to interact at SF-1 sites (Morohashi et al., 1994). Obligatory roles for SF-1 in adrenal and gonadal development and male sexual differentiation were established by gene knockout experiments (Luo et al., 1994). Targeted insertional mutagenesis was used to disrupt the Ftz-Fl gene in embryonic stem cells and to generate mice that were heterozygous for a mutant Ftz-Fl allele in which both SF-1 and ELP were inactivated. In crosses of these heterozygous mice, homozygous --I- offspring were born at the expected frequency of 25%, indicating that SF-1 is not essential for embryonic survival. All of the homozygous Ftz-Fl -disrupted animals died within 8 days of birth and had depressed corticosterone and elevated ACTH levels, consistent with a diagnosis of adrenocortical insufficiency as the cause of death. Surprisingly, all homozygous Ftz-Fl -disrupted animals lacked adrenal glands and gonads and had female internal genitalia (Fig. 5 ) . These results dramatically demonstrate the essential role of Ftz-Fl in adrenal and gonadal development and in male sexual differentiation. More recent studies suggest that the role of SF-1 in the reproductive axis extends considerably beyond its actions in the gonads. SF-1 is expressed in pituitary gonadotrophs and the Ftz-Fl -disrupted mice are selectively deficient in their expression of three distinct gonadotroph-specific genes: the P-subunit of FSH, the P-subunit of LH, and the receptor for GnRH (Ingraham et al., 1994). At least one of the genes essential for gonadotropin expression, the a-subunit of glycoprotein hormones, contains an SF-l-responsive promoter element, providing a possible mechanism for the absence of gonadotropin expression (Barnhart and Mellon, 1994; Ingraham et al., 1994). Treatment of the Ftz-F1 -disrupted mice with GnRH, however, induced pituitary expression of gonadotropins. This finding indicates that the gonadotroph lineage is present in these mice and implicates impaired GnRH delivery as a crucial defect in their gonadotroph function (Y. Ikeda, unpublished observations). The precise role of SF-1 within gonadotrophs thus remains to be defined. SF-1is also expressed in the hypothalamus in the ventromedial nucleus (Ikeda et al., 19951, a region that has been associated with the control of reproductive behavior (Pfaff et al., 1994). Although further studies are needed, the emerging evidence suggests an intimate link between SF-1 and the reproductive axis at all levels, in addition to its direct roles in steroidogenesis.
CYTUCHROME P-450 STEROID HYDROXYLASES
353
FIG.5 . The Ftz-F1 gene is essential for adrenal and gonadal development. The genitourinary tracts of newborn mouse pups from Ftz-FI -disrupted and control littermate mice were dissected and examined by photomicroscopy. (Left)Fb-FI-disrupted male, and kidney (K)but a n absent adrenal gland showing a n apparently normal oviduct (0) and gonad. (Right) Wild-type male littermate, with normal adrenal gland (A), kidney (K),and testis (T).
B. OTHERREGULATORS Although it is clearly a major determinant of steroid hydroxylase gene expression, SF-1 alone cannot fully account for the cell-selective expression of these genes. For example, additional regulators must be present to explain the restricted expression of CYP21 and C Y P l l B in the adrenal cortex or the expression of aromatase in the gonads but not the adrenal cortex, particularly since SF-1levels in these tissues are roughly comparable. Furthermore, elements that behave in a cellselective manner in expression studies using transfected cell lines may not adequately describe the in vivo situation. As shown by Milstone et al. (19921,the promoter regions identified as being important for the
354
KEITH L. PARKER A N D BERNARD P. SCHIMMER
cell-selective expression of the mouse Cyp21 gene in cultured adrenal cells were inadequate to direct expression of a reporter gene in transgenic mice; in fact, additional promoter sequences more than 5 kb upstream of the Cyp21 transcription initiation site and within the slp gene were required for adrenal cell-selective expression in this case. Inasmuch as the promoter regions of the steroid hydroxylase genes are complex and require a number of different elements for full promoter activity, it is reasonable to suggest that cell-selective expression of this gene family is determined by subsets of transcription factors that show overlapping patterns of cell-selective expression, with SF-1 serving as a common component. As discussed below, several additional candidates have been identified that may serve this function; however, it is worth briefly considering the possibility that cell-selective expression may, in some cases, be influenced by negative regulatory mechanisms. For example, a repressor element has been identified in the proximal promoter region of the human CYPl IAl gene and has been shown to function in JEG-3 human placental cytotrophoblasts but not in Y1 mouse adrenocortical tumor cells (Moore et al., 19921, and a negative element has been identified that suppresses bovine CYP21 promoter activity in luteal cells (Lauber et al., 1993). 1. Placenta-Specific Transcriptional Activator Miller and co-workers have identified a candidate cis-acting DNA element in the human C Y P l l A l gene that enhances the expression of reporter genes in a cell-selective manner when transfected into JEG-3 cells (Hum and Miller, 1994). The protein that interacts with this element is found in JEG-3 human choriocarcinoma cells but not in Y 1 or NCI-H295 adrenocortical tumor cells, suggesting that it is a key regulator of placenta-specific CYPI1AI expression. Preliminary studies indicate that this protein has a molecular mass of approximately 55 kDa and experiments are now under way to purify the protein and clone the cDNA that encodes it. 2. Adrenal-Specific Protein
A factor designated ASP (for adrenal-specific protein) has been shown to act at a CAMP-responsive element (CRE) within the proximal promoter regions of the human and bovine CYP21 genes to regulate gene expression in a CAMP-dependent manner. Initial studies suggested that ASP was expressed in cells of the adrenal cortex but not in other cell types, raising the possibility that this transcription factor acted in a cell-selective manner (Kagawa and Waterman, 1991, 1992). More recent studies, however, suggest that ASP negatively regulates
CYTOCHROME P-450STEROID HMROXYLASES
355
bovine P-450,,, expression (Momoi et al., 1992) and is not as exclusive to the adrenal cortex as was first thought (M. R. Waterman, personal communication). Therefore, the relative contribution of ASP to adrenal cell-selective gene expression has yet to be determined. EXPRESSION IV. HORMONE-REGULATED Studies on the molecular basis for hormone-regulated expression of the steroid hydroxylases contrast sharply with those on the cellselective expression of these genes. Whereas a single factor, SF-1, has emerged as a major determinant of the cell-selective expression of these genes within the primary steroidogenic tissues, we still lack a unifying hypothesis to account for their hormonal regulation. Even though most of the effects of the trophic hormones on gene expression are mediated by the cAMP second messenger system, a number of different transcriptional regulators seem to be required to bring about the coordinate induction of these genes in response to trophic hormones (for a recent review of this area, see Waterman, 1994).
A. CRE-BINDING PROTEIN Several groups have identified a palindromic sequence, the CRE, that is both necessary and sufficient for cAMP induction of known CAMP-induced genes such as somatostatin, phosphoenolpyruvate carboxykinase, and c-fos. The CRE interacts with a 43-kDa protein, the CRE-binding protein (CREB),which is a member of the leucine zipper family of transcriptional activators (See Brindle and Montminy, 1992; Meyer and Habener, 1993).In addition, recent studies have identified a coactivator for CREB, termed the CREB-binding protein (CBP), which is required for transcriptional activation in response to cAMP (Chrivia et al., 1993; Kwok et al., 1994). The involvement of CREB in the hormonal regulation of the steroid hydroxylases was first recognized in studies of the mouse C y p l l b2 gene encoding P-450ald0(Rice et al., 1989). This gene contains a sequence that closely resembles the classical CRE; this sequence conferred cAMP response to a heterologous promoter and interacted with a 43-kDa protein indistinguishable from CREB. Mutation of this sequence disrupted both constitutive and hormone-induced expression of P-450ald0, indicating its essential role. Like other genes regulated by CRE-CREB interactions (Roesler et al., 19871, induction of the C y p l l b2 gene was rapid, and increased transcript levels were detected
356
KEITH L. PARKER AND BERNARD P. SCHIMMER
within 1 h after stimulation (Domalik et al., 1991). Similar sequences are located at corresponding positions in the human genes encoding C Y P l l B l and C Y P l l B 2 (Mornet et al., 1989) and in the bovine gene for C Y P l I B I (Honda et al., 1990) and in a more distal part of the human C Y P l l A l gene (Watanabe et al., 1994); of these, the CRE in the bovine P-450,,, gene has been shown to be essential for CAMPinduced gene expression. Finally, recent studies with the rat aromatase 5’-flanking sequences show that CREB is important for the maximal hormonal induction of this gene in gonadal cells (Fitzpatrick and Richards, 1994). Classical CREs and CREBs, however, are not likely to fully account for the induction of most of the steroid hydroxylase genes. Most of the steroid hydroxylase genes do not show immediate responses to trophic hormones or CAMP, and are induced only after several hours of stimulation. Furthermore, in certain experimental systems inhibitors of protein synthesis such as cycloheximide inhibited transcriptional induction, suggesting that de nouo protein synthesis is required (John et al., 1986). These induction patterns are very different from those mediated by CREs and CREB. Finally, for many of the steroid hydroxylases, the 5’-flanking regions that participate in CAMP-regulated gene expression lack consensus sites for CREB binding. Based on these discrepancies, other elements have been examined for their potential roles in CAMP-dependent induction of the steroid hydroxylases.
RECEPTORPROTEINS B. NUCLEAR 1. SF-1 In addition to its role in steroidogenic cell-selective gene expression, SF-1 has also been implicated in the hormone- and CAMP-inducible expression of the steroid hydroxylases. As demonstrated by Parissenti et al. (1993), SF-1 sites require a functional CAMP-dependent protein kinase for promoter regulatory activity, and the SF-1 protein sequence contains a single consensus site for phosphorylation by CAMPdependent protein kinase (Honda et al., 1993), suggesting that phosphorylation of SF-1 by CAMP-dependent protein kinase is essential for its activity. SF-1 sites alone, when present as single copies, cannot support hormone- or CAMP-mediated gene transcription and require interactions with other promoter elements for this effect (Parissenti et al., 1993). For the C Y P l l B l and C y p l l b 2 genes SF-1 sites interact with classical CREs for CAMP-dependent induction of gene expression (Bogerd et al., 1990; Hashimoto et al., 1992). A similar interplay be-
CYTOCHROME P-450 STEROID HYDROXYLASES
357
tween SF-1 sites and CRE-like sequences has been observed in far upstream regions of the human CYP21 and C Y P l l A l genes (Watanabe et al., 1993, 1994). The more proximal promoter regions of Cyp21 and C Y P l l A l , however, contain SF-1 sites without CRE sequences, yet are still sufficient for hormone- and CAMP-stimulated gene transcription (Handler et al., 1988; Kagawa and Waterman, 1991). Therefore, in these promoter regions cAMP responsiveness presumably reflects a cooperation of SF-1 sites with other regulatory elements. As discussed below, candidates for these interactions include orphan members of the nuclear receptor family, such as nerve growth factor inducible-B (NGFI-B) and chicken ovalbumin upstream promoter transcription factor (COUP-TF), the homeodomain protein Pbxl, Spl, and a still unidentified protein that interacts with a regulatory element at -170 in the promoter of the mouse Cyp2l gene that requires a functional CAMP-dependent protein kinase for full promoter activity (Parissenti et al., 1993). 2. NGFI-B NGFI-B (also called nur77) is an orphan nuclear receptor that was isolated independently by several investigators searching for early response genes rapidly induced by growth factors (Milbrandt, 1988; Hazel et al., 1988; Ryseck et al., 1989).Subsequent studies established that NGFI-B, like SF-1,binds to half-site motifs that closely match the recognition sequence for a subset of SF-1 sites (Wilson et al., 1991). Moreover, although NGFI-B was isolated from rat PC12 pheochromocytoma cells treated with NGF, in situ hybridization analyses of NGFI-B expression in adrenal glands revealed high levels of NGFI-B in the adrenal cortex (T. E. Wilson et al., 1993). These observations suggested that NGFI-B contributes to steroid hydroxylase gene expression and prompted a series of experiments to test this hypothesis (T. E. Wilson et al., 1993). As determined by DNase I footprinting and gel mobility shift assays, recombinantly expressed NGFI-B interacted directly with the Cyp21 -65 element, one of several SF-1 sites in the promoter region of the gene encoding Cyp2l. Transfection of Y 1 cells with an expression vector encoding NGFI-B markedly increased promoter activity of the Cyp21 5’-flanking sequences in a manner dependent on the integrity of the -65 element. In addition, multiple copies of the -65 promoter element conferred ACTH and cAMP inducibility to a heterologous promoter. Finally, treatment of Y 1mouse adrenocortical tumor cells with ACTH rapidly induced NGFI-B expression with a peak effect within 1-2 h. Other studies extended the link between ACTH and NGFI-B action by showing that ACTH treatment of Y 1cells
358
KEITH L. PARKER AND BERNARD P. SCHIMMER
altered the phosphorylation and DNA binding of NGFI-B (Davis and Lau, 1994). Collectively, these results suggest that ACTH and cAMP stimulate the transcription of the Cyp21 gene by affecting the expression and phosphorylation of NGFI-B, which, in turn, cooperates with SF-1 through the -65 element to enhance gene expression. Whereas these studies have directly implicated NGFI-B in the regulation of the mouse Cyp21 gene, the role of NGFI-B in the expression of other steroid hydroxylases remains to be determined.
3. COUP-TF A series of indirect experiments implicate the orphan nuclear receptor, COUP-TF, in steroid hydroxylase gene expression. This nuclear receptor, originally characterized as a positive regulator of the ovalbumin gene (Wang, 19891, has subsequently been shown to inhibit transcriptional activation mediated by ligand-inducible nuclear receptors that form heterodimers with RXR (Kliewer et al., 1992; Tran et aZ., 1992; Cooney et al., 1992). Furthermore, COUP-TF activity seems to be regulated by cAMP and CAMP-dependent phosphorylation (Power et al., 1991). Based on effects of an anti-COUP-TF antiserum in gel mobility shift experiments, a n interaction of COUP-TF with the SF-1 site a t -210 in the P-4502, promoter has been postulated (Rice et aZ., 1991). To date, there are no functional data showing effects of COUPTF on steroid hydroxylase gene expression, although preliminary experiments with bovine CYPl7 suggest that COUP-TF and SF-1 compete for binding sites in a cAMP response sequence (CRS2) and that COUP-TF may antagonize SF-l-mediated transcriptional activation (Baake and Lund, 1995). Similarly, analyses of oxytocin gene expression in the bovine ovary suggest that SF-1 and COUP-TF compete for binding to DNA regulatory elements, with SF-1activating and COUPT F repressing promoter activity (Wehrenberg et al., 1994a,b).
C. Pbxl Studies by Waterman and colleagues identified a second region in the bovine CYPl7 promoter, designated CRS1, that was involved in CAMP-regulated gene expression. Purification and amino acid sequencing of transcriptionally active CRS1-binding proteins indicated that two isoforms of a widely expressed homeodomain protein, Pbxl, were present in the complex. Pbxl was originally isolated as a fusion protein with the helix-loop-helix DNA-binding protein E2A as the
CYTOCHROME P-450STEROID HYDROXYLASES
359
result of a t(1;19) chromosomal translocation found in certain patients with acute pre-B-cell leukemia (Kamps et al., 1990; Nourse et al., 1990). Although the fusion protein has oncogenic potential, the physiological role of Pbxl has yet to be determined. In transfection experiments in mouse Y 1cells, both Pbxl isoforms enhanced expression of a CRS1-containing promoter-reporter gene construct in a CAMPdependent manner (Kagawa et al., 1994). Interestingly, Pbxl has a consensus phosphorylation site for CAMP-dependent protein kinase, consistent with a possible role for this protein in the CAMP-dependent regulation of steroid hydroxylase gene expression. Additional studies are needed to reconcile the apparent adrenal-specific action of the CRSl sequence and the widespread expression of Pbxl and to demonstrate directly that CAMP-dependent phosphorylation of Pbxl is essential for CAMP-dependent gene expression in the context of the intact CYPl7 promoter. The Drosophila homologue of Pbx, extradenticle, has recently been shown to interact with homeodomain proteins to alter their DNA binding specificity (Chan et al., 1994; van Dijk and Murre, 19941, and it is attractive to propose that similar interactions with other transcriptional regulators contribute to transcriptional activation by Pbx. D. Spl Studies on the contributions of the proximal promoter regions of the human CYP21 and CYPl1A1 genes to CAMP-regulated gene expression identified GC-rich sequences in each promoter that resembled Spl binding sites and conferred cAMP responsiveness to reporter genes. In fact, these sites interacted with two separate proteins: the widely expressed transcription factor Spl (Briggs et al., 1986) and an adrenalspecific protein, ASP (see Section III,B,2). Mutational studies implicated ASP as the important determinant of CAMP-induced P-4502, gene expression. In contrast, CAMP-dependent regulation via the corresponding GC sequence of CYPllAl requires the Spl site, with the ASP site attenuating the response to cAMP (Momoi et al., 1992). Taken together, these results highlight the complexity of the hormonal regulation of the steroid hydroxylases and the problems in discerning a single paradigm that explains the coordinate regulation of these genes by hormones and CAMP.Furthermore, the strong possibility remains that requirements for gene expression defined using minimal promoter sequences in transfected cell lines may not apply in the context of the whole genome under more physiological conditions.
360
KEITH L. PARKER AND BERNARD P. SCHIMMER
V. PERSPECTIVES AND FUTURE DIRECTIONS Despite considerable progress in our understanding of the mechanisms that regulate the expression of the steroid hydroxylases, a number of major questions remain to be resolved. While it appears very likely that SF-1 coordinately regulates the expression of all of the steroidogenic cytochrome P-450s within the primary steroidogenic tissues, these conclusions are based exclusively on expression studies in isolated cells using relatively limited amounts of 5’-flanking sequences from each of the P-450 genes. Attempts to further address the role of SF-1 in steroidogenesis, using the Ftz-Fl -disrupted mouse model system, have shown that this gene plays a critical role in adrenal and gonadal development but have not resolved the physiological importance of SF-1 in P-450 gene expression. Targeted disruption of SF-1 in established differentiated cell lines or the generation of transgenic animals with conditionally defective SF-1 may provide alternative strategies to specifically address the role of SF-1 in P-450 gene expression. As noted in Section III,B, SF-1 alone is insufficient to account for the cell-selective expression of the steroidogenic P-450 enzymes, and the factors that determine hormone-regulated expression of these genes have yet to be identified. Therefore, other transcription factors, possibly acting in concert with SF-1, must also participate in the cellselective and hormonal regulation of the steroidogenic P-450 enzymes. In particular, the zone-specific expression of the P-450,,, isozymes within the adrenal cortex provides a striking example of exquisitely regulated gene expression. An essential role for the transcriptional regulation of this zonally restricted gene expression is convincingly demonstrated by patients with glucocorticoid-remediable aldosteronism, which results from ectopic expression of a protein with P-450,,,, activity in the inner zones under control of the CPYIIBI 5’-flanking DNA (Dluhy and Lifton, 1994).While considerable attention has been directed toward defining the mechanisms involved in CAMP-dependent regulation of the P-450 genes, we still understand very little about other signal transduction mechanisms that regulate these genes. In the adrenal zona glomerulosa the predominant regulator of steroidogenesis is AII, which is believed to modulate aldosterone production via changes in intracellular calcium (Barrett et al., 1992).Identifying the mechanisms by which A11 activates steroid hydroxylase gene expression will provide important insights into the regulation of mineralocorticoid production. Similarly, numerous other agents have been reported to affect gene
CYTOCHROME P-450 STEROID HYDROXYLASES
361
expression or promoter activity of the steroid hydroxylases, including activators of protein kinase C (Moore et al., 1990;Brentano et al., 1990; Baake and Lund, 1992;Lauber et al., 19931,transforming growth factor p (Rainey et al., 19901, cytokines (Xiong and Hales, 19931,and insulin-like growth factor I (Urban et al., 1994).Elucidation of the cisacting elements and transcriptional regulatory proteins by which these mediators regulate steroid hydroxylase gene expression will provide important insights into the interplay of various signal transduction pathways in regulating steroidogenesis. Although the studies reviewed here have focused on identifying and characterizing the role of transcriptional regulatory proteins in regulating the steroid hydroxylases, there are hints that epigenetic events may also regulate these steroid hydroxylases. One such epigenetic modification that has been shown to modulate gene expression is DNA methylation (reviewed by Tate and Bird, 1993;Eden and Cedar, 1994). In mouse Y 1 adrenocortical tumor cells the Cyp21 gene is highly methylated and is not expressed. Following replication of the gene in bacteria to remove epigenetic modifications, the Cyp21 gene was expressed and was hypomethylated when transfected into Y1 cells (Szyf et al., 1990).Upon prolonged growth of the transfected Y 1cells, Cyp21 methylation increased and Cyp21 expression was silenced; however, the silencing of gene expression temporally preceded the increase in methylation. Although a causal role has not been established, changes in the methylation and expression of P-450,,, are also seen in primary cultures of bovine adrenocortical cells (Hornsby et al., 1992). These results further suggest that changes in methylation status may affect steroid hydroxylase gene expression. The requirement for Ftz-FI in adrenal and gonadal development temporally precedes the onset of expression of the steroidogenic cytochrome P-450,and thus is unlikely to be a direct consequence of impaired steroidogenesis. Therefore, Ftz-Fl may regulate the expression of other genes that are important for steroidogenic organ development. Besides the steroid hydroxylases, previously identified SF-1responsive genes include MIS (Shen et al., 1994)and the a-subunit of glycoprotein hormones (Barnhart and Mellon, 1994;Ingraham et al., 1994);based on the link between SF-1 and steroidogenesis, other potential SF-l-responsive genes include the receptors for trophic hormones (ACTH, LH, and FSH), the StAR protein, and other gonadal peptides, such as activin and inhibin. Ultimately, the identification of additional genes that act downstream of Ftz-Fl in the complex cascades of adrenal and gonadal development may reveal why this gene is essential.
362
KEITH L. PARKER AND BERNARD P. SCHIMMER
It is known that nuclear receptor variants are found in a number of cancers arising in hormone-responsive tissues (Miksicek, 19941, and a causal role in oncogenesis has been implicated for the thyroid hormone receptor in avian erythroblastosis virus (Damm, 1993) and in acute promyelocytic leukemia associated with the reciprocal t(15;17) translocation involving the retinoic acid receptor-a gene (Kakizuka et al., 1991); it is therefore tempting to speculate that activating mutations or abnormal expression of SF-1may be causally related to the development of adrenocortical or gonadal carcinomas. Based on the phenotype of the Ftz-Fl-disrupted mice, one might also predict that SF-1 inactivation in humans would cause adrenal hypoplasia and hypogonadism. Although the most frequent form of congenital adrenal hypoplasia is X linked, ruling out FTZ-Fl as a candidate gene, there have been reports of rare autosomal-recessive variants (Kruger et al., 1993); intriguingly, these patients often also display hypogonadotropic hypogonadism (Kletter et al., 19911, closely matching the hypothalamic and pituitary defects seen in the Ftz-Fl-disrupted mice. It will be informative to ascertain whether any of these inborn errors of adrenal and gonadal development are caused by mutations in SF-1. It will also be important to see whether any of the multiple forms of gonadal dysgenesis or sex reversal (Grumbach and Conte, 1992) are associated with abnormalities in the structure or expression of SF-1.
VI. SUMMARY Steroid hormone biosynthesis requires the concerted action of a related group of cytochrome P-450 steroid hydroxylases. In recent years considerable effort has been directed toward defining the molecular basis for the cell-selective expression of these genes and their transcriptional regulation by trophic hormones. The orphan nuclear receptor SF-1, acting through a conserved element found in the proximal promoter regions of all steroid hydroxylase genes, seems t o be a major, but not exclusive, determinant of cell-selective gene expression. In contrast, the coordinate responses of the steroid hydroxylases to trophic hormones apparently involves an interplay of multiple proteins that collectively lead to a synchronous induction of gene expression. In some instances these interactions apparently involve transcription factors that also contribute to the cell-selective expression of these genes.
CYMCHROME P-450 STEROID HYDROXYLASES
363
REFERENCES Armstrong, D. T., and Dorrington, J. H. (1977). Estrogen biosynthesis in the ovaries and testes. Adv. Sex Horm. Res. 3, 217-258. Bakke, M., and Lund, J. (1992). A novel 3’,5’-cyclic adenosine monophosphateresponsive sequence in the bovine CYPl7 gene is a target of negative regulation by protein kinase C. Mol. E h c r i n o l . 6, 1323-1331. Bakke, M., and Lund, J. (1995). Mutually exclusive interactions of two nuclear receptors sequence determine activity of a cyclic adenosine 3’,5’-monophosphate-responsive in the bovine CYP17 gene. Mol. Endocrinol. 9,327-339. Barnhart, K. M., and Mellon, P. L. (1994). The orphan nuclear receptor, steroidogenic factor-1, regulates the glycoprotein hormone alpha-subunit gene in pituitary gonadotropes. Mol. Endocrinol. 8, 878-885. Barrett, P. Q., Bollag, W. B., kales, C. M., McCarthy, R. T., and Rasmussen, H. (1989). The role of calcium in angiotensin 11-mediated aldosterone secretion. Endocr. Rev. 10, 496-518. Bogerd, A., Rice, D. A., Franklin, A., Schimmer, B. P., and Parker, K. L. (1990). Identification and characterization of two upstream elements that regulate adrenocortical expression of steroid llp-hydroxylase. Mol. Endocrinol. 4, 845-850. Boggaram, V., John, M. E., Simpson, E. R., and Waterman, M. R. (1989).Effect of ACTH on the stability of mRNAs encoding bovine adrenocortical P45Oscc, P45011p, P45017a, P450c21, and adrenodoxin. Biochem. Biophys. Res. Comrnun. 160,12271232. Brentano, S. T., Picardo-Leonard, J., Mellon, S. H., Moore, C. C. D., and Miller, W. L. (1990). Tissue-specific, CAMP-induced, and phorbol ester repressed expression from the human P450c17 promoter in mouse cells. Mol. Endocrinol. 4, 1972-1979. Briggs, M. R., Kadonaga, J. T., Bell, S. P., and Tjian, R. (1986). Purification and biochemical characterization of the promoter-specific transcription factor, Spl. Science 23447-52. Brindle, P. K., and Montminy, M. R. (1992). The CREB family of transcription activators. C u m Opin. Genet. Dev. 2, 199-204. Bulun, S. E., Rosenthal, I. M., Brodie, A. M., Inkster, S. E., Zeller, W. P., DiGeorge, A. M., Frasier, S. D., Kilgore, M. W., and Simpson, E. R. (1994). Use of tissue-specific promoters in the regulation of aromatase cytochrome P450 gene expression in human testicular and ovarian sex cord tumors, as well as in normal fetal and adult gonads. J. Clin. E n d o c r i d . Metab. 78, 1616-1621. Chan, S. K., Jaffe,L., Capovilla, M., Botas, J., and Mann, R. S. (1994). The DNA binding specificity of Ultrabithorax is modulated by cooperative interactions with extradenticle, another homeoprotein. Cell 78, 603-615. Chrivia, J. C., Kwok, R. P., Lamb, N., Hagiwara, M., Montminy, M. R., and Goodman, R. H. (1993). Phosphorylated CREB binds specifically to the nuclear protein CBP. Nature 365,855-859. Clark, B. J.,Wells, J., King, S. R., and Stocco, D. J. (1994). The purification, cloning and expression of a novel LH-induced mitochondria1 protein in MA-10 mouse Leydig tumor cells: Characterization of the steroidogenic acute regulatory protein (StAR). J . Bwl. Chern. 229,28314-28322. Clemens, J.W., Lala, D. S., Parker, K. L., and Richards. J. S. (1994). Steroidogenic factor 1 binding and transcriptional activity of the cholesterol side chain cleavage promoter in rat granulosa cells. Endocrinology 134, 1499-1508.
364
KEITH L. PARKER AND BERNARD P. SCHIMMER
Cooney, A. J., Tsai, S. Y., O’Malley, B. W., and Tsai, M. J. (1992). Chicken ovalbumin upstream promoter transcription factor (COUP-TF) dimers bind to different GGTCA response elements, allowing COUP-TF to repress hormonal induction of the vitamin D3, thyroid hormone, and retinoic acid receptors. Mol. Cell. Biol. 12,41534163. Damm, K. (1993). ErbA: Tumor suppressor turned oncogene? FASEB J . 7,904-909. Davis, I. J., and Lau, L. F. (1994). Endocrine and neurogenic regulation of the orphan nuclear receptors Nur77 and Nurr-1 in the adrenal glands. Mol. Cell. Biol. 14,34693483. Dluhy, R. G., and Lifton, R. P. (1994). Glucocorticoid-remediable aldosteronism. Endocrinol. Metab. Clin. North A m . 23, 285-297. Domalik, L. J., Chaplin, D. D., Kirkman, M. S.,Wu, R. C., Liu, W., Howard, T. A,, Seldin, M. F., and Parker, K. L. (1991). Different isozymes of mouse llp-hydroxylase produce mineralocorticoids and glucocorticoids. Mol. Endocrinol. 5, 1853-1861. Eden, S., and Cedar, H. (1994). Role of DNA methylation in the regulation of transcription. Curr. Opin. Genet. Deu. 4, 255-259. Evans, R. M. (1988). The steroid and thyroid hormone receptor superfamily. Science 240, 889-895. Fitzpatrick, S. L., and Richards, J. S. (1993). Cis-acting elements of the rat aromatase promoter required for cyclic adenosine 3’,5’-monophosphate induction in ovarian granulosa cells and constitutive expression in R2C Leydig cells. Mol. Endocrinol. 7, 341-354. Fitzpatrick, S. L., and Richards, J. S. (1994). Identification of a cyclic adenosine 3’,5’monophosphate-response element in the rat promoter that is required for transcriptional activation in rat granulosa cells and R2C Leydig cells. Mol. Endocrinol. 8, 1309-1319. Gamier, M., Boujrad, N., Ogwuegbu, S. O., Hudson, J. R., Jr., and Papadopoulos, V. (1994). The polypeptide diazepam-binding inhibitor and a higher affinity mitochondrial peripheral-type benzodiazepine receptor sustain constitutive steroidogenesis in the R2C Leydig tumor cell line. J . Biol. Chem. 269, 22105-22112. Grumbach, M. M., and Conte, F. A. (1992). Disorders of sex differentiation. I n “Williams Textbook of Endocrinology” (J.D. Wilson and D. W. Foster, eds.), 8th ed., pp. 853951. Guo, I. C., Tsai, H. M., and Chung, B. C. (1994). Actions of two different CAMPresponsive sequences and a n enhancer of the human CYPllAl (P450scc) gene in adrenal Y 1 and placental JEG-3 cells. J . Biol. Chem. 269, 6362-6369. Handler, J. D., Schimmer, B. P., Flynn, T. R., Szyf, M., Seidman, J. G., and Parker, K. L. (1988). An enhancer element and a functional CAMP-dependentprotein kinase are required for expression of adrenocortical 21-hydroxylase. J . Biol. Chern. 263, 13068-13073. Hashimoto, T., Morohashi, K., Takayama, K., Honda, S., Wada, T., Handa, H., and Omura, T. (1992). Cooperative transcription activation between Adl, a CRE-like element, and other elements in the CYPllB gene promoter. J . Biochem. (Tokyo) 112, 573-575. Hazel, T. G., Nathans, D., and Lau, L. F. (1988). A gene inducible by serum growth factors encodes a member of the steroid and thyroid hormone receptor superfamily. Proc. Natl. Acad. Sci. U.S.A. 85, 8444-8448. Honda, S., Morohashi, K., and Omura, T. (1990). Novel CAMPregulatory elements in the promoter region of bovine P45011p gene. J. Biochern. (Tokyo) 108, 1042-1049.
CYTOCHROME P-450STEROID HYDROXYLASES
365
Honda, S.-I., Morohashi, K.-I., Nomura, M., Takeya, H., Kitajima, M., and Omura, T. (1993). Ad4BP regulating steroidogenic P-450 gene is a member of steroid hormone receptor superfamily. J.Biol. Chem. 288,7494-7502. Hornsby, P. J., Yang, L.. Raju, S. G., Maghsoudlou, S. S..Lala, D. S., and Nallaseth, F. S. (1992). Demethylation of specific sites in the 5'-flanking region of the CYP17 genes when bovine adrenocortical cells are placed in culture. DNA Cell Biol. 11,385-395. Hum, D. W., and Miller, W. L. (1994). Characterization of a placental specific transcriptional activator of the human gene for P450scc. Endocr. Soc. Abstr. p. 503a. Ikeda, Y., Lala, D. S., Luo, X., Kim, E., Moisan, M.-P., and Parker, K. L. (1993). Characterization of the mouse FTZ-Fl gene, which encodes a n essential regulator of steroid hydroxylase gene expression. Mol. Endocrinol. 7 , 852-860. Ikeda, Y., Luo, X.,Abbud, R., Nilson, J. H., and Parker, K. L. (1995). The nuclear receptor steroidogenic factor 1 is essential for the formation of the ventromedial hypothalamic nucleus. Mol. Endocrinol. 9, 478-486. Ikeda, Y.,Shen, W.-H., Ingraham, H. A., and Parker, K. L. (1994). Developmental expression of mouse steroidogenic factor 1, a n essential regulator of the steroid hydroxylases. Mol. Endocrinol. 8, 654-662. Ingraham, H. A,, Lala, D. S., Ikeda, Y., Luo, X., Shen, W.-H., Nachtigal, M. W., Abbud, R., Nilson, J. H., and Parker, K. L. (1994). The nuclear receptor steroidogenic factor 1 acts a t multiple levels of the reproductive axis. Genes Deu. 8, 2302-2312. John, M. E., John, M. C., Boggaram, V., Wimpson, E. R., and Waterman, M. R. (1986). Transcriptional regulation of steroid hydroxylase genes by corticotropin. Proc. Natl. Acad. Sci. U S A . 83, 4715-4719. Josso, N. (1992). Hormonal regulation of sexual differentiation. Semin. Perinatol. 16, 279-288. Kagawa, N., and Waterman, M. R. (1991). Evidence that an adrenal-specific nuclear protein regulates the cAMP responsiveness of the human CYPBlB (P450C21) gene. J . Biol.Chem. 266, 11199-11204. Kagawa, N., and Waterman, M. R. (1992). Purification and characterization of a transcription factor which appears to regulate cAMP responsiveness of the human CYP21B gene. J.Biol. Chem. 267,25213-25219. Kagawa, N., Ogo, A., Takahashi, Y., Iwamatsu, A., and Waterman, M. R. (1994). A CAMP-regulatory sequence (crsl) of cypl7 is a cellular target for the homeodomain protein pbx. J.Biol. Chem. 269, 18716-18719. Kakizuka, A., Miller, W. H., J r . , Umesono, K., Warrell. R. P., Jr., Frankel, S. R., Murty, V. V., Dmitrovsky, E., and Evans, R. M. (1991). Chromosomal translocation t(15;17) in human acute promyelocytic leukemia fuses RAR alpha with a novel putative transcription factor, PML. Cell 66,663-674. and Baltimore. D. (1990). A new homeobox gene Kamps, M. P., Murre, C. M., Sun,X.-H., contributes to the DNA binding domain of the t(1;19) translocation protein in pre-B ALL. Cell 60, 547-555. Kitamura, M., Bunko, E., and Dufau, M. (1991). Dissociation of hydroxylase and lyase activities by site-directed mutagenesis of the rat P45017a. Mol. Endocrinol. 5 , 1373-1380. Kletter, G. B., Gorski, J. L., and Kelch, R. P. (1991). Congenital adrenal hypoplasia and isolated gonadotropin deficiency. Trends Endocrinol. Metab. 2, 123-128. Kliewer, S. A,, Umesono, K., Heyman, R. A., Mangelsdorf, D. J., Dyck, J. A., and Evans, R. M. (1992). Retinoid X receptor-COUP-TF interactions modulate retinoic acid signaling. Proc. Natl. Acad. Sci. U.S.A.89, 1448-1452.
366
KEITH L. PARKER AND BERNARD P. SCHIMMER
Krueger, K. E., and Papadopoulos, V. (1990). Peripheral-type benzodiazepine receptors mediate translocation of cholesterol from outer to inner membranes in adrenocortical cells. J . Biol. Chem. 265, 15015-15022. Kruger, G., Mix, M., Pelz, L., and Dunker, H. (1993). Cytomegalic type of congenital adrenal hypoplasia due to autosomal recessive inheritance. Am. J . Med. Genet. 46, 475. Kwok, R. P., Lundblad, J. R., Chrivia, J. C., Richards, J. P., Bachinger, H. P., Brennan, R. G., Roberts, S. G., Green, M. R., and Goodman, R. H. (1994). Nuclear protein CBP is a coactivator for the transcription factor CREB. Nature 370,223-226. Labrie, F., Simard, J., Luu-The, V., Belanger, A., and Pelletier, G. (1992). Structure, function, and tissue-specific gene expression of 3 beta-hydroxysteroid dehydrogenase/5-ene-4-ene isomerase enzymes in classical and peripheral intracrine steroidogenic tissues. J. Steroid Biochem. Mol. Biol. 43, 805-826. Lala, D. S., Rice, D. A., and Parker, K. L. (1992). Steroidogenic factor 1,a key regulator of steroidogenic enzyme expression, is the mouse homolog of fushi tarazu factor 1. Mol. Endocrinol. 6, 1249-1258. Lauber, M. E., Picton, H. M., Begeot, M., Momoi, K., Waterman, M. R., and Simpson, E. R. (1993). Regulation of CYPllA gene expression in bovine ovarian granulosa cells in primary culture by CAMPand phorbol esters is conferred by a common cisacting element. Mol. Cell. Endocrinol. 94, 235-242. Laudet, V., Hanni, C., Coll, J., Catzeflis, F.,and Stehelin, D. (1992). Evolution of the nuclear receptor superfamily. EMBO J. 11, 1003-1013. Lavorgna, G., Ueda, H., Clos, J., and Wu, C. (1991). FTZ-F1,a steroid hormone receptor like protein implicated in the activation of fushi tarazu. Science 252, 848-851. Lavorgna, G., Karim, F.D., Thummel, C. S., and Wu, C. (1993).Potential role for a FTZF1 steroid receptor superfamily member in the control of Drosophila metamorphosis. Proc. Nutl. Acad. Sci. U.S.A. 90, 3004-3008. Liu, Y. X., and Hsueh, A. J. (1986). Synergism between granulosa and theca-interstitial cells in estrogen biosynthesis by gonadotropin-treated rat ovaries: Studies on the two-cell, two-gonadotropin hypothesis using steroid antisera. Biol. Reprod. 35,2736. Luo, X., Ikeda, Y., and Parker, K. L. (1994). A cell-specific nuclear receptor is essential for adrenal and gonadal development and sexual differentiation. Cell 77,481-490. Lynch, J. P., Lala, D. S., Peluso, J. J., Luo, W., Parker, K. L., and White, B. A. (1993). Steroidogenic factor 1, an orphan nuclear receptor, regulates the expression of the rat aromatase gene in gonadal tissue. Mol. Endocrinol. 7, 776-786. Mellon, S. H. (1994). Neurosteroids: Biochemistry, modes of action, and clinical relevance. J . Clin. Endocrinol. Metab. 78, 1003-1008. Mellon, S. H., and Deschepper, C. F. (1993). Neurosteroid biosynthesis: Genes for adrenal steroidogenic enzymes are expressed in the brain. Brain Res. 629, 283-292. Meyer, T. E., and Habener, J. F.(1993). Cyclic adenosine 3’,5’-monophosphate response element binding protein (CREB) and related transcription-activating deoxyribonucleic acid-binding proteins. Endocr. Rev. 14, 269-290. Miksicek, R. J. (1994). Steroid receptor variants and their potential role in cancer. Semin. Cancer Biol. 5,369-380. Milbrandt, J. (1988). Nerve growth factor induces a gene homologous to the glucocorticoid receptor gene. Neuron 1, 183-188. Miller, W. L. (1994). Clinical review: Genetics, diagnosis, and management of 21hydroxylase deficiency. J . Clin. Endocrinol. Metub. 78,241-246. Milstone, D. S., Shaw, S. K., Parker, K. L., Szyf, M., and Seidman, J. G. (1992). An
CYTOCHROME P-450 STEROID HYDROXYLASES
367
element regulating adrenal-specific steroid 21-hydroxylase expression is located within the Slp gene. J. Biol. Chem. 267. 21924-21927. Momoi, K., Waterman, M. R., Simpson, E. R., and Zanger, U. M. (1992). 3',5'-Cyclic adenosine monophophate-dependent transcription of the CYPllA (cholesterol side chain cleavage cytochrome P450) gene involves a DNA response element containing a putative binding site for transcription factor Spl. Mol. Endocrinol. 6, 16821690. Moore, C. C. D., Brentano, S. T., and Miller, W. L. (1990). Human P45Oscc gene transcription is induced by CAMP and repressed by 12-0-tetradecanoylphorbol-13-acetate and A23187 through independent cis-elements. Mol. Cell. Biol. 10, 6013-6023. Moore, C. C. D., Hum, D. W., and Miller, W. L. (1992). Identification of positive and negative placenta-specific basal elements and a cyclic adenosine 3',5'-monophosphate response element in the human gene for P45Oscc. Mol. Endocrinol. 6 , 20452058. Mornet, E., Dupont, J., Vitek, J., and White, P. C. (1989). Characterization of two genes encoding human steroid llp-hydroxylase (P45011p). J. Biol. Chem. 264, 2096120967. Morohashi, K., Honda, S., Inomata, Y.,Handa, H., and Omura, T. (1992). A common trans-acting factor, Ad4BP, to the promoters of steroidogenic P450s. J.Biol. Chem. 267, 17913-17919. Morohashi, K., Iida, H., Nomura, M., Hatano, O., Honda, S., Tsukiyama, T., Niwa, O., Hara, T., Takakusu, A., Shibata, Y., and Omura, T.(1994). Functional difference between Ad4BP and ELP, and their distributions in steroidogenic tissues. Mol. Endocrinol. 8, 643-653. Nebert, D. W. (1991). Proposed role of drug-metabolizing enzymes: Regulation of steady state levels of the ligands that effect growth, homeostasis, differentiation, and neuroendocrine functions. Mol. Endocrinol. 5, 1203-1214. Nebert, D. W., and Gonzalez, F. J. (1987). P450 genes. Structure, evolution, and regulation. Annu. Rev. Biochem.56,945-993. Nelson, D. R., Kamataki, T., Waxman, D. J., Guengerich, F. P., Estabrook, R. W., Feyereisen, R., Gonzalez, F. J., Coon, M. J.,Gunsalus, I. C., Gotoh, O., Okuda, K., and Nebert, D. W. (1993). The P450 superfamily: Update on new sequences, gene mapping, accession numbers, early trivial names of enzymes, and nomenclature. DNA Cell Bwl. 12, 1-51. Nitta, H., Bunick, D., Hess, R. A., Janulis, L., Newton, S. C., Millette, C. F., Osawa, Y., Shizuta, Y., Toda, K., and Bahr, J. M. (1992). Germ cells of the mouse testis express P450 aromatase. Endocrinology 132, 1396-1401. Nourse, J., Mellentin, J. D., Gallili, N., Wilkinson, J., Stanbridge, E., Smith, S. D., and Cleary, M. L. (1990). Chromosomal translocation of t(1;19) results in synthesis of a homeobox fusion mRNA that codes for a potential chimeric transcription factor. Cell 60,535-546. Ogishima, T., Mitani, F., and Ishimura, Y. (1989). Isolation of two distinct cytochromes P-450,, with aldosterone synthase activity from bovine adrenocortical mitochondria. J. Biochem. 105,497-499. O'Malley, B. W., and Conneely, 0. M. (1992). Orphan receptors: In search of a unifying hypothesis for activation. Mol. Endocrinol. 6 , 1359-1361. Parissenti, A. M., Parker, K. L., and Schimmer, B. P. (1993). Identification of promoter elements in the mouse 21-hydroxylase (Cyp2l) gene that require a functional cyclic adenosine 3',5'-monophosphate-dependentprotein kinase. Mol. Endocrinol. 7,283290.
368
KEITH L. PARKER AND BERNARD P. SCHIMMER
Parker, K. L., and Schimmer, B. P. (1993). Transcriptional regulation of the adrenal steroidogenic enzymes. Trends Endocrinol. Metab. 4, 46-50. Pedersen, R. C., and Brownie, A. C. (1987). Steroidogenesis-activatorpolypeptide isolated from a rat Leydig cell tumor. Science 236, 188-190. Pfaff, D. W., Schwartz-Giblin, S., McCarthy, M. M., and Kow, L.-M. (1994). Cellular and molecular mechanisms of female reproductive behaviors. I n “The Physiology of Reproduction” (E. Knobil and J. D. Neill, eds.), 2nd ed., pp. 107-220. Raven, New York. Pfeifer, S. M., Furth, E. E., Ohba, T., Chang, Y. J., Rennert, H., Sakuragi, N., Billheimer, J. T., and Strauss, J. F., I11 (1993). Sterol carrier protein 2: A role in steroid hormone synthesis? J . Steroid Biochem. Mol. Biol. 74, 167-172. Pon, L. A., Hartigan, J. A,, and OrmeJohnson, N. R. (1986). Acute ACTH regulation of adrenal corticosteroid biosynthesis. Rapid accumulation of a phosphoprotein. J . Biol.Chem. 261, 13309-13316. Power, R. F., Lydon, J. P., Conneely, 0. M., and O’Malley, B. W. (1991). Dopamine activation of a n orphan of the steroid receptor superfamily. Science 252, 1546-1548. Rainey, W. E., Naville, D., Saez, J. M., Carr, B. R., Byrd, W., Magness, R. R., and Mason, J. I. (1990). Transforming growth factor-beta inhibits steroid 17 alpha-hydroxylase cytochrome P-450 expression in ovine adrenocortical cells. Endocrinology 127, 1910-1915. Rheaume, E., Simard, J., Morel, Y., Mebarki, F., Zachmann, M., Forest, M. G., New, M. I., and Labrie, F. (1992). Congenital adrenal hyperplasia due to point mutations in the type I1 3 beta-hydroxysteroid dehydrogenase gene. Nature Genet. 1, 239-245. Rice, D. A,, Aitkens, L. D., Vandenbark, G. R., Mouw, A. R., Franklin, A., Schimmer, B. P., and Parker, K. L. (1989). A CAMP-responsive element regulates expression of the mouse steroid llp-hydroxylase gene. J. Biol. Chem. 264, 14011-14015. Rice, D. A,, Kronenberg, M. S., Mouw, A. R., Aitken, L. D., Franklin, A,, Schimmer, B. P., and Parker, K. L. (1990a). Multiple regulatory elements determine adrenocortical expression of steroid 21-hydroxylase. J . B i d . Chem. 265, 805243059. Rice, D. A,, Kirkman, M. S., Aitken, L. D., Mouw, A. R., Schimmer, B. P., and Parker, K. L. (1990b). Analysis of the promoter region of the gene encoding mouse cholesterol side-chain cleavage enzyme. J. Biol. Chem. 265, 11713-11720. Rice, D. A., Mouw, A. R., Bogerd, A. M., and Parker, K. L. (1991). A shared promoter element regulates the expression of three steroidogenic enzymes. Mol. Endocrinol. 5, 1552-1561. Robel, P., and Baulieu, E.-E. (1994). Neurosteroids: Biosynthesis and function. Trends Endocrinol. Metab. 5, 1-8. Roesler, W. J., Vandenbark, G . R., and Hansen, R. W. (1987). Cyclic AMP and the induction of eukaryotic gene transcription. J. Biol. Chem. 263, 9063-9066. Ryseck, R. P., Bravo, H. M., Mattei, M. G., Ruppert, S., and Bravo, R. (1989). Structure, mapping, and expression of a growth factor inducible gene encoding a putative nuclear hormonal binding receptor. EMBO J. 8, 3327-3335. Schiff, R., Arensburg, J.,Itin, A., Keshet, E., and Orly, J. (1993). Expression and cellular localization of uterine side-chain cleavage cytochrome P450 messenger ribonucleic acid during early pregnancy in mice. Endocrinology 133, 529-537. Shen, W.-H., Moore, C. C. D., Ikeda, Y., Parker, K. L., and Ingraham, H. A. (1994). Nuclear receptor steroidogenic factor 1regulates MIS gene expression: A link to the sex determination cascade. Cell 77, 651-661. Simpson, E. R., Mahendroo, M. S., Means, G. D., Kilgore, M. W., Hinshelwood, M. M., Graham-Lorence, S., Amarneh, B., Ito, Y., Fisher, C. R., Michael, M. D., Mendelson,
CYTUCHROME P-450 STEROID HYDROXYLASES
369
C. R., and Bulun, S. E. (1994). Aromatase cytochrome P450, the enzyme responsible for estrogen biosynthesis. Endocr. Rev. 16, 342-355. Solomon, S. (1994). The primate placenta as a n endocrine organ. In “The Physiology of Reproduction” (E. Knobil and J. D. Neill, eds.), 2nd ed., pp. 863-873. Raven, New York. Stocco, D. M., and Kilgore, M. W. (1988). Induction of mitochondria1 proteins in MA-10 Leydig tumour cells with human choriogonadotropin. Biochem. J. 249,95-103. Szyf, M., Milstone, D. S., Schimmer, B. P., Parker, K. L., and Seidman, J . G. (1990). Cismodification of the steroid 21-hydroxylase gene prevents its expression in the Y 1 mouse adrenocortical tumor cell line. Mol. Endocrinol. 4, 1144-1152. Tate, P. H., and Bird, A. P. (1993). Effects of DNA methylation on DNA-binding proteins and gene expression. Curr. Opin. Genet. Dev. 3, 226-231. Tran, P., Zhang, X.-K., Salbert, G., Hermann, T., Lehmann, J. M., and Pfahl, M. (1992). COUP orphan receptors are negative regulators of retinoic acid response pathways. Mol. Cell. Biol. 12, 4666-4676. Tsukiyama, T., Ueda, H., Hirose, S.,and Niwa, 0. (1992). Embryonal long terminal repeat-binding protein is a murine homolog of FTZ-Fl, a member of the steroid receptor superfamily. Mol. Cell. Biol. 12, 1286-1291. Ueda, H., Sonoda, S., Brown, J. L., Scott, M. P., and Wu, C. (1990). A sequence-specific DNA-binding protein that activates fushi tarazu segmentation gene expression. Genes Dev. 4, 624-635. Urban, R. J., Shupnik, M. A., and Bodenburg, Y. H. (1994). Insulin-like growth factor-I increases expression of the porcine P-450 cholesterol side chain cleavage gene through a GC-rich domain. J.B i d . Chem.269, 25761-25769. van Dijk, M. A., and Murre, C. (1994). Extradenticle raises the DNA binding specificity of homeotic selector gene products. Cell 78, 617-624. Walsh, S. W., Resko, J. A., Grumbach, M. M., and Novy, M. J. (1980). In utero evidence for a functional fetoplacental unit in rhesus monkeys. Biol. Reprod. 23, 264-270. Wang, L. H., Tsai, S. Y., Cook, R. G., Beattie, W. G., Tsai, M. J., and OMalley, B. W. (1989). COUP-transcription factor is a member of the steroid receptor superfamily. Nature 340. 163-166. Watanabe, N., Kitazume, M., Fujisawa, J., Yoshida, M., and Fujii-Kuriyama, Y. (1993). A novel CAMP-dependent regulatory region including a sequence like the CAMPresponsive element, far upstream of the human CYP21A2 gene. Eur. J. Biochem. 214,521-531. Watanabe, N., Inoue, H., and Fujii-Kuriyama, Y. (1994). Regulatory mechanisms of CAMP-dependent and cell-specific expression of human steroidogenic cytochrome P450,, ( C Y P l l A l ) gene. Eur. J.Biochem. 222,825-834. Waterman, M. R. (1994). Biochemical diversity of CAMP-dependent transcription of steroid hydroxylase genes in the adrenal cortex. J.Biol. Chem. 269, 27783-27786. Wehrenberg, U., Ivell, R., Jansen, M., von Goedecke, S., and Walther, N. (1994a). Two orphan receptors binding to a common site are involved in the regulation of oxytocin genes in the bovine ovary. Proc. Natl. Acad. Sci. U.S.A.91, 1440-1444. Wehrenberg, U., von Goedecke, S., Ivell, R., and Walther, N. (1994b). The orphan receptor SF-1 binds to the COUP-like element in the promoter of the actively transcribed oxytocin gene. J.Neuroendocrinol. 6, 1-4. White, P. C. (1994). Genetic diseases of steroid metabolism. Vitam. Horm. 49, 131-196. Wilson, J. D., Griffin, J. E., and Russell, D. W. (1993). Steroid 5 alpha-reductase 2 deficiency. Endmr. Rev. 14, 577-593. Wilson, T. E., Fahrner, T.J., Johnston, M., and Milbrandt, J. (1991). Identification of
370
KEITH L. PARKER AND BERNARD P. SCHIMMER
the DNA binding site of NGFI-B by genetic selection in yeast. Science 252, 12961300. Wilson, T. E., Mouw, A. R., Weaver, C. A., Milbrandt, J., and Parker, K. L. (1993). The orphan nuclear receptor NGFI-B regulates steroid 21-hydroxylase gene expression. Mol. Cell. Biol. 13,861-868. Xiong, Y., and Hales, D. B. (1993). The role of tumor necrosis factor-alpha in the regulation of mouse Leydig cell steroidogenesis. Endocrinology 132,2438-2444. Zanger, U.M., Kagawa, N., Lund, J., and Waterman, M. R. (1992). Distinct biochemical mechanisms for CAMP-dependenttranscription of CYP17 and CYP21. FASEB J . 6, 719-723.
VITAMINS AND HORMONES, VOL. 51
Stress and the Brain: A Paradoxical Role for Adrenal Steroids
BRUCE S. MCEWEN," DAVID ALBECK," HEATHER CAMERON," HELEN M. CHAO," ELIZABETH GOULD," NICOLAS HASTINGS," YASUKAZU KURODA," VICTORIA LUINE,? ANA MARIA MAGARINOS," CHRISTINA R. MCKITTRICK,* MILES ORCHINIK," CONSTANTINE PAVLIDES," PAUL VAHER,? YOSHIFUMI WATANABE," AND NANCY WEILAND" *Laboratory of Neuroendocrinology The Rockefeller University New York, New York 10021 *Department of Psychology Hunter College New York, New York 10021
1. 11. 111. IV. V.
VI. VII. VIII. IX. X.
Introduction Adrenal Steroids and Hippocampal Neuronal Atrophy Antidepressants Modify Stress-Induced Changes in the Hippocampus Paradoxical Effects of Adrenal Steroids in the Hippocampus A. Adrenal Steroids and Neuronal Birth and Death in the Dentate Gyms B. Adrenal Steroids and LTP Role of Adrenal Steroid Receptor Subtypes in Structural and Neurochemical Plasticity Other Actions of Adrenal Steroids Fklated to Neuronal Atrophy Stress Effects on Cognitive Performance in Rodents Deregulation of the HPA Axis in Depression and Other Disorders Effects of Exogenous Glucocorticoid Treatment on Cognitive Performance in Humans Conclusions References
I. INTRODUCTION Stress is common in everyday life and is blamed for many problems. There is mounting evidence that stressful experiences exacerbate disease processes (McEwen and Stellar, 1993). However, what is stressful for one person is not necessarily stressful for another. One reason for 371
Copyright ,O 1995 by Academic Press, Inc. All rights of reproduction in any form reserved.
372
BRUCE S. MCEWEN et al.
this is the interaction between brain function and behavior, since the brain is the organ that interprets what is “stressful” and is responsible for mounting physiological responses, which include activation of the sympathetic nervous system (fight or flight) and the hypothalamicpituitary-adrenal (HPA) axis. Circulating glucocorticoids play an important signaling role in coordinating a variety of physiological responses, ranging from containing immune and inflammatory processes to promoting the production of metabolizable energy in the form of carbohydrates and fat. The brain is a principal target organ for circulating glucocorticoids, which not only regulate the production and release of hypothalamic corticotropin-releasing factors and pituitary adrenocorticotropic hormone (ACTH) but also modulate neurotransmitter systems and regulate the structural plasticity and circuitry of key brain regions. One of the most important brain areas responding to circulating adrenal steroids is the hippocampal formation, and our laboratory has studied this brain structure since 1968, when we found that it contains high concentrations of adrenal steroid receptors (McEwen et al., 1968).The hippocampus is organized as shown in Fig. 1,with two major regions: Ammon’s horn and the dentate gyrus. The principal excitatory synaptic pathways include input to the granule cells of the dentate gyrus from the entorhinal cortex (via the perforant pathway). These, in turn, project via the mossy fibers to the CA3 pyramidal cells, which then
FIG.1. Diagram of hippocampal formation, showing the dentate gyrus (DG)and Amman’s horn (H) and depicting the three-cell circuit. pp, perforant path; ENT, entorhinal cortex.
STRESS AND THE BRAIN
373
project via the Schaffer collaterals to the CA1 pyramidal cells. Hippocampal excitability is regulated by the release of y = aminobutyric acid (GABA) from local neurons. The hippocampal formation is often described as the “cognitive” arm of the limbic system, since it plays an important role in spatial and declarative memory as well as in determining how expectations match with actual events (Eichenbaum and Otto, 1992). Hippocampal malfunction is implicated in schizophrenia and affective illness (Luchins, 1990; Axelson et al., 1993). Furthermore, glucocorticoids have a significant impact on this brain structure; for example, in Cushing’s syndrome and after exogenous glucocorticoid therapy for inflammatory disorders, declarative memory defects have been found (Martignoni et al., 1992; Wolkowitz et al., 1990a,b; Newcomber et al., 1994; Starkman et al., 1992), and verbal memory scores were decreased with elevated cortisol in subjects who also showed decreased hippocampal volume (Starkman et al., 1992). Moreover, the hippocampus is vulnerable to degenerative effects of stress and loss of neurons during aging as well as in response to ischemia, seizures, and head trauma (Sapolsky, 1992; Landfield, 1987; Sloviter, 1983; Lowenstein et al., 1994; Hsu and Buzsaki, 1993). Adrenal steroids represent only one of several neurochemical systems that mediate the delicate balance between hippocampal function and dysfunction. For example, the excitatory amino acid (EAA) transmitters play a key role in the destructive events mentioned above (Sapolsky, 19921, but they are also involved in synaptic plasticity (Woolley and McEwen, 1993), stabilization of neuronal populations (Gould and McEwen, 19931,long-term potentiation (LTP), and memory (Mondadori and Weizkrantz, 1993).Thus, it is very important to elucidate the mechanisms by which the hippocampus responds to these agents and to begin to discriminate between processes leading to adaptive responses and those leading to damage. Our laboratory has developed a model system for studying what may be the first link in the destructive cascade of events. Specifically, the CA3c region of the hippocampus (that part of CA3 closest to the dentate gyrus; see Ishizuka et al., 1990) shows atrophy of apical dendrites of pyramidal neurons as a result of repeated restraint stress or social subordination stress, and both adrenal steroids and EAAs are mediators of this atrophy, However, adrenal steroids are also implicated in other aspects of hippocampal function, including the turnover of neurons in the adult dentate gyrus and the modulation of excitability in the form of LTP. The purpose of this chapter is to discuss the role of adrenal steroids in the hippocampus and relate their effects to the actions of glucocorticoids
374
BRUCE S. MCEWEN et al.
on human memory and hippocampal volume and in relation to Cushing’s syndrome and depression. 11. ADRENAL STEROIDS AND HIPPOCAMPAL NEURONAL ATROPHY Aus der Muhlen and Ockenfels (1969) reported that ACTH or cortisone administration to mature guinea pigs caused neurons in the hippocampus and other forebrain regions to stain darkly and appear necrotic, as if undergoing atrophy. While the interpretation of these findings may be questioned in light of more recent findings regarding artifacts that can develop in fixing brains for histological analysis (Cammermeyer, 19781, they were instrumental in stimulating the work of Landfield and colleagues, which showed that aging in the rat results in some pyramidal neuron loss in the hippocampus and that this loss was retarded by adrenalectomy (ADX) in midlife (Landfield, 1987). Sapolsky’s group then demonstrated that 12 weeks of daily corticosterone (CORT) injections into young adult rats mimicked the pyramidal neuron loss seen in aging (Sapolsky et al., 1985).Sapolsky went on to demonstrate that EAAs play an important role in the cell loss by showing, first, that CORT exacerbates kainic acid-induced damage to hippocampus as well as ischemic damage, and, second, that glucocorticoids potentiate EAA killing of hippocampal neurons in culture (Sapolsky, 1992). In order to examine what actually happens to hippocampal neurons as a result of high levels of glucocorticoids, we have used the singlesection Golgi technique to demonstrate that after 21 days of daily CORT exposure, the apical dendritic tree of CA3c pyramidal neurons has undergone atrophy (Woolley et al., 1990). Moreover, this atrophy was prevented by the antiepileptic drug phenytoin, given prior t o CORT each day, thus implicating the release and actions of EAAs, since phenytoin blocks glutamate release and antagonizes t-type calcium channels that are activated during glutamate-induced excitation (Watanabe et al., 1992a) (Fig. 2). If CORT injections cause atrophy, would repeated stress do the same? Application of restraint stress (6 hlday for 21 days) produced the same pattern of atrophy as CORT treatment, namely, atrophy of the apical, but not the basal, dendrites of CA3c pyramidal neurons (Watanabe et al., 1992~)(Fig. 3). Such atrophy was also blocked by daily phenytoin administration, confirming that EAA release is involved (Watanabe et al., 1992a). Recent work has shown that NMDA receptor blockade is also effective in preventing stress-induced dendri-
STRESS AND THE BRAIN
It: c
m
control
375
cort cort + phenytoin
E B ._ O
O.OO*l
I
7
FIG.2. (A) Effects of daily corticosterone administration for 21 days on branching of the apical dendrites of CA3c pyramidal neurons in Ammon’s horn. Also shown is the protective effect of daily phenytoin administration. (B) Corticosterone (cort) treatment decreases thymus weight, and phenytoin does not protect against this effect. Data show the mean t SEM; asterisks indicate significant differences from control values a t P < 0.05. [From Watanabe et al. (1992a). Reprinted by permission.]
tic atrophy (Magarifios and McEwen, 1995). Thus, EAA release is involved in both CORT- and stress-induced atrophy of CA3c pyramidal neurons; however, this does not necessarily mean that CORT is involved in stress-induced atrophy. CORT is released in response to restraint stress (Watanabe et al., 1992c),but in much lower amounts than that which is given exogenously to cause atrophy (Watanabe et al., 1992~). In order to explore the role of endogenous CORT release in dendritic atrophy evoked by stress, we have used the steroid synthesis inhibitor
376
BRUCE S. MCEWEN et a1
A
B
C
FIG. 3. Daily restraint stress for 21 days causes atrophy of the apical dendrites of CA3c neurons. Camera lucida drawings show representative CA3c neurons under (A) control, (B) stress, and (C) stress plus daily phenytoin conditions. Note the decrease in the number of dendritic branches in (B) compared to (A) and (0.[From Watanabe et al. (1992a). Reprinted by permission.]
cyanoketone to reduce the magnitude of CORT secretion in response to restraint stress (Magarinos and McEwen, 1995). Cyanoketone attenuates the adrenal steroid stress response but allows basal CORT secretion to occur (Akana et al., 1983). We found that cyanoketone treatment blocked the atrophy of dendrites induced by daily restraint stress (Magarinos and McEwen, 1995). This result is consistent with a role for CORT secretion, but the story is not quite so simple, since repeated restraint stress for 21 days does not consistently cause adrenal hypertrophy; moreover, the CORT response to repeated restraint stress habituates over time, showing a progressively earlier shutoff during the 6-h restraint session (Watanabe et al., 1992~). Thus, the magnitude of CORT released in response to restraint stress is considerably smaller than that given in the exogenous CORT paradigm, and CORT probably plays a much smaller role in the atrophy produced by restraint stress. Nevertheless, it appears to play some role, and a possible link between CORT and EAA release is provided by the recent demonstration that ADX markedly reduces the magnitude of the EAA release evoked by restraint stress (Lowy et al., 1993).Future studies need to examine the role of cyanoketone in blocking EAA release during restraint stress, as further evidence for this link; and further work is needed to clarify the role of type I or I1 adrenal steroid receptors in stress-induced atrophy, using antagonists or agonists applied systemically during stress to rats with adrenocortical insufficiency.
STRESS AND THE BRAIN
377
Another important issue is the relationship between the atrophy of CA3c neurons induced by repeated COW treatment or by repeated restraint stress and the loss of pyramidal neurons that has been reported after both 12 weeks of CORT treatment and severe social stress (Sapolsky et al., 1985; Uno et al., 1989). Indeed, we have seen that chronic social stress in tree shrews causes atrophy of CA3c pyramidal neurons, much as restraint stress does in rats (A. M. Magariiios, E. Fuchs, G. Flugge, and B. S. McEwen, unpublished observations, 1995). However, the atrophy produced by stress in rats is reversible, within 7-14 days after the termination of stress (A. M. Magarifios and B. S. McEwen, unpublished observations, 1995).A possible link may be the fate of inhibitory interneurons that receive intense innervation from mossy fibers of dentate gyrus granule neurons and are especially vulnerable to a variety of insults (Hsu and Buzsaki, 1993). If some of these neurons were to die as a result of repeated restraint stress, there might be a cumulative effect, in which repeated bouts of stress might progressively deplete the dentate gyrus of the buffering action these inhibitory neurons appear to provide. Another possibility is that CORT or stress alters CA3 neuronal atrophy through regulation of GABAergic synaptic inhibition. In support of this notion, we have found that low levels of CORT alter mRNA levels for specific subunits of GABAA receptors in CA3 and the dentate gyrus of ADX rats (Orchinik et al., 1994). Ten-day treatment with stress levels of CORT has produced different effects on GABA, receptor subunit mRNA levels and receptor binding in hippocampal subregions, including CA3 (M. Orchinik, N. G. Weiland, and B. S. McEwen, unpublished observations, 1995).Therefore, it appears that CORT may alter the excitability of hippocampal neurons through regulation of GABAA receptor expression, but it remains to be seen whether the corticosteroid effects on neuronal morphology involve changes in the number or the pharmacological properties of GABA, receptors.
111. ANTIDEPRESSANTS MODIFY STRESS-INDUCED CHANGES IN THE HIPPOCAMPUS Repeated restraint stress in rats for 3 weeks causes changes in the hippocampal formation, including atrophy of the dendrites of CA3c pyramidal neurons as well as suppression of serotonin 1A (5-HT,,) receptor binding (Mendelson and McEwen, 1992; Watanabe et al., 1993; McKittrick et al., 1994) (Fig. 4). Because 5-HT is released by stressors (Kennet et al., 1985) and may play a role in the actions of stress on nerve cells, we investigated the actions of agents that facili-
BRUCE S. MCEWEN et al.
H Controls H Dominants Responsive Subordinates Nonresponsive Subordinates
T
CAI
CA2
CA3
CA4
DG-Supra
DG-lnfra
FIG.4. Chronic social stress suppresses 5-HT,, binding in the hippocampus. [3H]SHydroxy-DPAT binding to 5-HTl, receptors in hippocampal subfields and the dentate gyrus. Regions examined were CA1-CA4 of Ammon’s horn and the suprapyramidal (DG-Supra)and intrapyramidal (DG-Infra) of the dentate gyrus. Note that CA4 is what we refer to elsewhere as CA3c. Data are expressed as the group mean t SEM. Dominants and stress-responsive subordinates showed down-regulation in many regions, whereas the stress-nonresponsive subordinates did not show such reductions, except in the CA3 region. Asterisks indicate P < 0.05 versus controls. [From McKittrick et al. (1994). Reprinted by permission.]
tate or inhibit 5-HT reuptake. Tianeptine is known to enhance 5-HT uptake (Mennini et al., 19871, and we compared it with fluoxetine, an inhibitor of 5-HT reuptake (Beasley et al., 19921, as well as with desipramine, as an inhibitor of noradrenaline uptake. As shown in Fig. 5 , tianeptine treatment (10 mg/kg/day) prevented the stress-induced atrophy of the dendrites of CA3c pyramidal neurons (Watanabe et al., 1992b), whereas neither fluoxetine (10 mg/kg/day) nor desipramine (10 mg/kg/day) had any effect (Y.Watanabe, A. M. Magarinos, and B. S. McEwen, unpublished observations, 1994). Exogenous CORT treatment also produced atrophy of the dendrites of CA3c pyramidal neurons, and this effect was also blocked by tianeptine (15 mg/kg/day) treatment (Watanabe et al., 199213). Thus, the effect of tianeptine on CA3c morphology is not due to its reported effects to reduce CORT secretion (Delbende et al., 19911, but may instead be related to its reported effects to enhance the reuptake of 5-HT within the hippocampus (Whitton et al., 1991). Because CORT and
STRESS AND THE BRAIN
379
A
E -
B 2 2000
s ln
. Q r
t
sham
cort
cort+tianeptine
coti
corI+tianeptine
Isham
FIG.5. Tianeptine blocks stress-induced atrophy of the CA3c apical dendrites. Both (A) dendritic branch points and (B) dendritic length were reduced by 2 1 days’ stress, and this decrease was blocked by daily tianeptine treatment. Asterisks indicate significant differences from the other two groups a t P < 0.05. cort, Corticosterone. [From Watanabe et al. (1992b). Reprinted by permission.]
stress-induced atrophy of CA3c dendrites is also blocked by phenytoin (Watanabe et al., 1992a1, these results suggest that serotonin released by stress or CORT may interact pre- or postsynaptically with glutamate released by stress or CORT, and that the final common pathway may involve interactive effects between 5-HT and glutamate receptors on the dendrites of CA3c neurons innervated by mossy fibers from the dentate gyrus. There is evidence of interactions between 5-HT and NMDA receptors, indicating that 5-HT potentiates NMDA receptor binding as well as the activity of NMDA receptors and that it may do
380
BRUCE S. MCEWEN et al.
so via 5-HT2 receptors (Mennini and Miari, 1991; Rahmann and Neumann, 1993).
IV. PARADOXICAL EFFECTS OF ADRENAL STEROIDS IN THE HIPPOCAMPUS It is important to emphasize that adrenal steroids play other roles in hippocampal formation. Many of these actions seem almost paradoxical in light of the role of glucocorticoids in stress-induced neuronal atrophy. This section briefly summarizes some of these other effects. A. ADRENALSTEROIDS AND NEURONAL BIRTHAND DEATH IN THE DENTATE GYRUS In contrast to pyramidal neurons of Ammon’s horn, granule neurons of the adult dentate gyrus depend on adrenal steroids for their survival, and ADX of an adult rat causes increased granule neuron death (Gould and McEwen, 1993; Sloviter et al., 1989). Moreover, continuous birth of granule neurons occurs in the adult dentate gyrus, and neurogenesis also increases following ADX (Gould and McEwen, 1993). The first clue as to the magnitude of the ADX effect on neuronal death in the dentate gyrus was the finding that 3 months after ADX of adult rats, some rats showed almost a total loss of dentate gyrus granule neurons (Sloviter et al., 1989). This finding attracted considerable attention because it conflicted with the prevalent view of stress- and glucocorticoid-induced neuronal death in Ammon’s horn (see Section 11);it was also puzzling because only some ADX rats showed the loss of the entire dentate gyrus. We now have more information that allows us to better understand these events, based on the unique nature of adrenal steroid actions in the dentate gyrus. ADX has been shown to induce apoptotic death of dentate gyrus granule neurons within 3-7 days in virtually every rat from which all adrenal tissue has been removed (Gould et al., 1990). Loss of the entire dentate gyrus in only some rats may well be explained by an absence of accessory adrenal tissue in these rats; when not removed at the time of ADX, this tissue can supply enough adrenal steroids to prevent neuronal loss. Very low levels of adrenal steroids, sufficient to occupy type I adrenal steroid receptors, completely block dentate gyrus neuronal loss (Woolley et al., 1991) (Fig. 6). The role of type I and I1 adrenal steroid receptors is discussed in Section V. It should be noted that granule cell birth is also accelerated by ADX,
STRESS AND THE BRAIN
381
FIG. 6. Q p e I (mineralocorticoid) receptors protect granule neurons from cell death. Values represent the mean 5 SEM of the logarithm of the number of pyknotic cells/l06 mm2 in the granule cell layer of the dentate gyrus of adult rats that were shamadrenalectomized (SHAM; n = 5); adrenalectomized (ADX; n = 4); ADX and treated with the type I agonist aldosterone (ALDO; 1mg/h; n = 4); or ADX and treated with the type I1 receptor agonist RU28362 (10 mglh; n = 6). Asterisks indicate significant differences from SHAM and ALDO at P < 0.001. [From Woolley et al. (1991). Reprinted by permission.]
FIG.7. Location of labeled cells in the adult rat dentate gyrus a t (A) 1 h and (B)4 weeks after [3H]thymidine labeling in the hilus (h) and the granule cell layer (gcl). Solid circles represent labeled cells immunoreactive for neuron-specific enolase (NSE), while open circles represent NSE-nonimmunoreactive cells. Solid triangles represent glial fibrillary acidic protein-immunoreaetive thymidine-labeled cells. [From Cameron et al. (1993a). Reprinted by permission.1
382
BRUCE S. MCEWEN et al.
1hr
24hr
lwk
2wk
3wk
4wk
1hr
24hr
lwk
2wk
3wk
4wk
FIG.8. Newly L3HIthymidine-labeled cells do not express receptors for type I [mineralocorticoid receptor (MR)]or type I1 [glucocorticoid receptor (GR)]. (A) The percentages of MR-labeled cells are shown a t indicated times after 1"HIthymidine administration. Asterisks show significant differences from those a t 1 h and 24 h a t P < 0.05. (B) The percentages of GR-labeled cells are shown a t indicated times after PHIthymidine administration. The asterisk shows the significant difference from all other groups a t P < 0.05. [From Cameron et al. (199313). Reprinted by permission.]
and that the newly born neurons in the adult dentate gyrus arise in the hilus, close to the granule cell layer, and then migrate into the granule cell layer (Fig. 71, presumably along a vimentin-staining radial glial network that is also enhanced by ADX (Cameron et al., 1993). Specific antibodies to neuron-specific enolase have been used to show that the majority of newly born cells in the adult dentate gyrus are neurons. In
STRESS AND THE BRAIN
383
spite of the above-mentioned role for type I adrenal steroid receptors in suppressing neuronal turnover in the adult dentate gyrus, it should be noted that most neuroblasts labeled with PHlthymidine lack both type I and type 11 adrenal steroid receptors (Cameron et al., 19931, indicating that steroidal regulation occurs via messengers from an unidentified steroid-sensitive cell (Fig. 8). The possibility that other trophic factors may be involved is currently under investigation. One explanation for why the dentate gyrus makes new cells in adult life, as well as dispose of them, is so that it may process spatial information and related aspects of memory (Sherry et al., 1992). Birds that use space around them to hide and locate food, and voles as well as deer mice that traverse large distances to find mates, have larger hippocampal volumes than closely related species that do not; moreover, there are indications that hippocampal volume may change during the breeding season (Sherry et al., 1992; Galea et al., 1994). It remains to be determined whether it is the dentate gyrus that exhibits this plasticity. Another important question concerns the impairment of hippocampal function that may accompany adrenocortical insufficiency. Several reports have indicated that long-term ADX rats that have damage to the dentate gyrus similar to that described by Sloviter et al. (1989) show modest deficits in spatial memory (Armstrong et al., 1993; Conrad and Roy, 1993). Recent work has revealed that deficits in performance of an eight-arm radial maze task are evident in rats with ADX for approximately 70 days, without there being evidence of ongoing neuronal loss or degeneration in the dentate gyrus (Vaher et al., 1994). One possibility currently being tested (P. Vaher, unpublished observations) is that there are acute actions of adrenal steroids on brain function that are not manifested in terms of dentate neuronal turnover. These actions are potentially relevant to the acute actions of adrenal steroids on LTP, which are described in the following section. B. ADRENAL STEROIDS AND LTP A single burst of high-frequency stimulation to hippocampal afferents can immediately alter the responsiveness of neurons to further stimuli, an effect lasting from many hours to days. This type of plasticity, first described by Bliss and Lgmo (1973), is called LTP. A number of recent in vitro studies have demonstrated in the hippocampal CA1 field that acute stress produces an impairment in LTP or its close relative, primed-burst potentiation (Diamond et al., 1992; Pavlides et al., 1993). In these studies a negative correlation was also seen be-
384
BRUCE S. MCEWEN et al.
tween the degree of LTP-induced and plasma CORT levels. More recently, an inverted-U-shaped curve was found with regard to induction of plasticity in the CA1 field in relation to circulating CORT levels (Diamond et al., 1992).That is, both adrenocortical insufficiency and high levels of CORT had a detrimental effect on potentation, whereas optimal potentiation was obtained within an intermediate range (10-20 Fg/dl) of circulating CORT. Within the dentate gyrus an acute administration of CORT has also been shown to produce a decrement in LTP (Pavlides et al., 1993).More recent data, however, indicate that LTP can be modulated biphasically by adrenal steroids acting, respectively, via type I and type I1 receptors. In the dentate gyrus of ADX rats, LTP was found to be enhanced within 1h by adrenal steroids acting via type I receptors, whereas type I1 receptor stimulation was found to rapidly suppress LTP (Pavlides et al., 1995) (Fig. 9). These biphasic effects may help t o explain the biphasic dose-response curve with respect to induction of LTP in the CA1 region of Ammon’s horn, noted above, in which low levels of CORT at the nadir and the rising phase of daily adrenocortical activity facilitate LTP, whereas stress levels of CORT inhibit LTP (Diamond et al., 1992). Moreover, enhancement of LTP by the type I receptor agonist aldosterone (ALDO) administration lasts for at least 24 h, at which time it is still markedly higher than in ADX rats given only vehicle treatment before LTP induction (Pavlides et al., 1994) (Fig. 10). The CA1 field responds in a manner similar to that of the dentate gyrus to corticosteroids. This should not be surprising, for at least two reasons. First, it is likely that type I and I1 receptors coexist in pyramidal neurons and in dentate gyrus granule neurons (Herman et al., 1989). Second, in CA1 pyramidal neurons corticosteroids biphasically modulate excitability; that is, type I receptor stimulation facilitates excitability by disinhibiting a 5-HTIAreceptor input, whereas type I1 receptor stimulation inhibits excitability by suppressing a P-receptormediated noradrenergic input (Joels and DeKloet, 1992).It remains t o be seen whether biphasic actions of adrenal steroids on LTP may be a general feature of hippocampal formation. The precise mechanisms underlying the stress-induced impairments in LTP are not yet established. While CORT may be modulating LTP directly, via its action on hippocampal cells, the endogenous opiates and neurotransmitters such as glutamate and 5-HT may also be involved. Tail shock in a restraining tube for 30 min caused hippocampal slices from these rats to show significantly less LTP in CA1 pyramidal neurons (Foy et al., 1987). Although LTP was affected, paired-pulse facilitation was not altered, indicating that inhibitory mechanisms
Population Spike Control
3
r
Type1
Type11
.
Population EPSP
Group FIG. 9. Effects of adrenal steroid type I and I1 receptors on long-term potentiation (LTP) in the dentate gyrus of anesthetized rats. Following baseline recording of field potentials, animals were injected with either the type I or I1 agonistiantagonist (or a combination of the two) followed by high-frequency stimulation (HFS)of the perforant pathway and further recording. The effects of the HFS were measured and indicated on (A) population spike and (B) excitatory postsynaptic potential (EPSP) slope. Bar graphs represent the mean (tSEM) potentiation for each of the groups of animals included in the experiment. In comparison to adrenalectomized (ADX) controls, the type I adrenal steroid agonist aldosterone (ALDO) produced a significant enhancement in LTP as measured by the population spike. This enhancement was reversed by the type I antagonist RU28318. In contrast, the type I1 agonist RU28362 produced a significant reduction, which was reversible with preadministration ofthe type I1 antagonist RU38486. With respect to the slope of the EPSP, none of the type I or I1 agonistlantagonist groups was significantly different than the ADX group, although a trend similar to the spike was observed. The numbers in parentheses represent the number of animals in each group (**P < 0.001; *P < 0.05).The sham-operated animals were placed in two groups based on post hoc analysis of plasma corticosterone (CORT) levels and LTP results. Animals in the sham A group had levels lower than 45 pgidl and showed “normal” LTP, while those in the sham B group had plasma CORT values higher than this and showed either suppressed LTP or, in some cases, long-term depression. [From Pavlides et al. (1995). Reprinted by permission.]
386
BRUCE S. MCEWEN et al
:1 1 5'
0
ADX
*
ADX+Aldo
250.
HFSl
HFs3
3h
24h
Time of Testing
FIG. 10. Acute and long-term effects of adrenal steroid type I receptor activation on long-term potentiation (LTP) in the dentate gyrus of freely behaving adrenalectomized (ADX) rats. Following baseline (BL) recording a set of three high-frequency stimulations (HFSs) were applied, and changes in the size of the population spike were determined, in comparison to BL values. The first HFS produced significant LTP for both the ADX plus aldosterone (Aldo) and ADX plus vehicle animals, with higher increases seen after the third HFS. Although the ADX plus Aldo animals show somewhat higher LTP than the ADX plus vehicle animals, this difference was not significant. Thus, comparable LTP was obtained for both groups. However, while LTP decayed substantially by 3 h and returned to BL levels by 24 h in the ADX plus vehicle group, it persisted in the ADX plus Aldo animals not only a t 3 h but also a t 24 h following HFS. The asterisk indicates significant (P < 0.05) differences between ADX plus ALDO and ADX plus vehicle animals, a t the times indicated. [From Pavlides et al. (1994). Reprinted by permission. I
were not affected (Shors and Thompson, 1992). ADX blocked the stress effect, but this effect was related to the adrenal medulla, since CORT did not restore the effect, whereas adrenal demedullation mimicked the effect of total ADX (Shors et al., 1990a). Moreover, naltrexone, a n opiate antagonist, also blocked the effect of stress, suggesting that a n endogenous opioid mechanism may be involved (Shors et al., 1990b). The inhibitory effects of environmental stressors were found in other paradigms. An inhibitory effect on LTP was found after 1 week of training to escape low-intensity foot shock in a shuttle box and was more evident in the yoked control, which could not escape the shock (Shors et al., 1989). In another study exposure of rats to a novel environment resulted in a rapid and reversible impairment of plasticity in uiuo in the CA1 region (Diamond et al., 1994).
STRESS AND THE BRAIN
387
Recent studies have reported that plasticity in the hippocampus is accompanied by induction of a number of immediate-early genes (IEGs) (Worley et al., 1993; Abraham et al., 1991). In the hippocampus of unanesthetized freely behaving animals, it has been shown that IEGs, including the fos and jun families, ~$268, and krox, are activated following the induction of LTP. Interestingly, fos and zifZ68 have also been shown to increase in a number of brain areas, including the hippocampus, following acute stress (Schreiber et al., 1991; Watanabe et al., 199413). Activation of these IEGs, in turn, may regulate the expression of various late effector genes, for example, Fl/gap43, which could then produce structural changes that would be required for the longterm maintenance of plasticity. Because steroid hormone receptors are known to interact with IEGs via “composite response elements” on target genes (Miner and Yamamoto, 19911, it is conceivable that the actions of ALDO to facilitate LTP are accompanied by an increased expression of genes that participate in the establishment of new structural connections. Future studies will explore the structural and neurochemical changes that accompany the steroid enhancement of LTP, including possible interactions between glucocorticoids and induction of IEGs and changes in structural protein and neurotrophin gene expression.
V. ROLEOF ADRENAL STEROID RECEPTOR SUBTYPES IN STRUCTURAL AND NEUROCHEMICAL PLASTICITY It is evident from the distribution of mRNA and binding for type I and I1 adrenal steroid receptor subtypes that neurons in the dentate gyrus and Ammon’s horn most likely contain both types of receptors (Herman et al., 1989). Ongoing work in our laboratory has revealed other specific effects mediated by adrenal steroid receptor types in the hippocampus and other brain regions. In these studies we applied steroids via subcutaneously implanted Alzet minipumps (Alza Corp., Palo Alto, CA). The type I1 agonist RU28362 was administered in a dose (10 Fg/h) that suppresses almost to zero the corticotrophin releasing hormone (CRH) mRNA signal in the paraventricular nuclei; the type I agonist ALDO was given at a dose of 10 kg/h and had no effect on CRH mRNA levels in the paraventricular nuclei (Albeck et al., 1994) (Table I). In the hippocampus of the same animals, ALDO prevented the upregulation of 5-HTlA receptors that was produced by ADX, whereas RU28362 was without any effects (Kuroda et al., 1994). Other work in progress has revealed that ADX-induced decreases in dynorphin mRNA in the dentate gyrus and in kainate receptors in the stratum
388
BRUCE S. MCEWEN et a1
TABLE I SUMMARY OF T ~ P E I AND I1 RECEPTORMEDIATION OF GENEEXPRESSION I N HIPPOCAMPUS AND OTHERBRAIN REGIONS Gene product
Hippocampus
CRH mRNA Neuropeptide mRNA Dynorphin mRNA Serotonin receptors Kainate receptors
5Pe1 Type 1 5Pe1 Type 1
-
Paraventricular nucleus
Arcuate nucleus
Locus coeruleus
Type 11 -
-
-
TypeII
TypeII
-
-
-
-
-
-
-
Note. A dash indicates no information available.
lucidum of CA3 (representing the mossy fiber zone of terminals from dentate granule neurons) are reversed by ALDO but not by RU28362 treatment (Watanabe et al., 1994b). Furthermore, another study has revealed that type I receptors mediate a negative regulation by adrenal steroids of the expression of neuropeptide Y mRNA in the hilus of the dentate gyrus (Watanabe et al., 1994a). Thus, most of the known actions of adrenal steroids in the hippocampus appear to be mediated by type I receptors. Exceptions are the type II-mediated inhibition of LTP, noted above (Section IV,B), and the involvement of type I1 receptors in the excitotoxin-induced damage to hippocampal neurons in culture (Packan and Sapolsky, 1990).
VI. OTHERACTIONS OF ADRENAL STEROIDS RELATED TO NEURONAL ATROPHY Hippocampal neurons can also undergo increased dendritic branching and new synapse formation. Estrogens induce new synapses and spine formation on CA1 pyramidal neurons (Woolley and McEwen, 1992). Hibernation causes dendritic retraction of CA3 pyramidal neurons in ground squirrels, whereas bringing squirrels out of the hibernating state results in rapid dendrite expansion (Popov and Bocharova, 1992; Popov et al., 1992). EAAs and their ability to mobilize calcium ions may play an important role in dendritic plasticity, since, in cultured hippocampal neurons, low levels of glutamate induce neurite outgrowth, while intermediate levels lead to selective atrophy of dendrites, and high glutamate levels cause neuronal death (Mattson et al., 1988). Dendrite extension and spine formation may be promoted by d e
STRESS AND THE BRAIN
389
nouo synthesis of key cytoskeletal proteins via polyribosomes located near or in the base of spines (Steward et al., 1988). Assembly and disassembly of cytoskeletal proteins, promoted by phosphorylation and dephosphorylation, also play a role in the expansion and contraction of dendrites (Halpain and Greengard, 1990). The actions of calciumactivated neutral proteases may also be important for the physiological plasticity of dendrites (Nixon et al., 1986);moreover, the activation of calcium-dependent proteases is involved in damage associated with ischemic injury (Lee et al., 1991). Intracellular signaling mechanisms play a role in regulating plastic changes in the hippocampus, in part through calcium signaling pathways involving NMDA receptors and L-type calcium channels (Bading et al., 1993). IEGs are activated by a variety of stimuli (Morgan and Curran, 1989; Sheng and Greenberg, 19901, including stress (Shreiber et al., 19911, and involve intracellular calcium mobilization; IEGs also participate in the regulation of neurotrophin production by hippocampal neurons (Gall, 1992). Adrenal steroids are known to influence many of these processes, increasing calcium currents through both Nand L-type calcium channels via type I1 adrenal steroid receptors (Kerr et al., 1992) and facilitating the production of neurotrophins by hippocampal neurons in culture and in uiuo (Lindholm et al., 1992; Barbany and Persson, 1992), especially in conjunction with excitatory stimulation (Barbany and Persson, 1993). Stress is also reported to alter the expression of the neurotrophins brain-derived neurotrophic factor (BDNF) and neurotrophin 3 (NT-3) in the hippocampus (Smith et al., 1993). Very little is known about the adrenal steroid receptor types involved in these effects, or the localization of stress and adrenal steroid effects on neurotrophin expression within the hippocampus and their relationship to the conditions of repeated stress that bring about morphological changes. Moreover, hippocampal neurons also express receptors for BDNF and NT-3, which then act in an autocrine fashion (Cheng and Mattson, 1994). Recent evidence indicates that adrenal steroids exert some regulatory effects on mRNA levels for both BDNF and NT-3 in the hippocampus, with high doses of CORT reducing mRNA levels in a number of regions of the hippocampus; interestingly, the mRNAs for the receptors for these two neurotrophins, trk B and trk C, do not appear t o be regulated by adrenal steroids (Chao and McEwen, 1994) (Fig. l l a ) . Another growth factor that may play an important role in hippocampal neuronal survival is basic fibroblast growth factor (bFGF),which is present in hippocampal neurons and astrocytes (Woodward et al., 1992; Matsuyama et al., 1992). Receptors for basic fibroblast growth factor (bFGF) are expressed in hippocampal neurons (Asai et al.,1993);bFGF
€3 Sham
*
BDNF mRNA
;r,
0 ADX f3 ADX+lowCORT W HighCORT 1200
800
400
0
CAI
B
t
bFGF mRNA
DG
CA3
*
a
T,
0 Sham
0 ADX
E l 2000
-
ADX+lowCORT
W HighCORT
.a
1 3
4 1000
0
CAI
CA2
CA3
DG
Region
FIG. 11. (A) Expression of BDNF mRNA in the rat hippocampus. Levels of mRNA expression were analyzed in the CA1 and CA3 pyramidal cell layers and the granule cell layer of the dentate gyrus (DG). The asterisk indicates that the results of the adrenalectomized (ADX group) is significantly different from those of the high corticosterone (CORT) group (P < 0.05). [From Chao and McEwen (1994).1(B) Expression of basic FGF (bFGF) mRNA in the rat hippocampus in the CA1, CA2, and CA3 pyramidal cell layers and the DG. In CA2 ADX levels are lower than those of all other treatments (*P < 0.05). [From Chao and McEwen (1994) with permission.]
STRESS AND THE BRAIN
391
thus acts in an autocrine fashion, and hippocampal neuronal survival in culture is enhanced by this factor (Walicke et al., 1986; Needels and Cotman, 1988). Hippocampal damage increases the expression of bFGF (Kiyota et al., 1991; Gomez-Pinilla et al., 1992). Initial indications are that bFGF mRNA levels fall after ADX but only in the very small CA2 region; they do not change in the rest of Ammon's horn or in the dentate gyrus (Chao and McEwen, 1994) (Fig. llb). Initial results in our laboratory reveal a number of actions of adrenal steroids on neurotrophin mRNA levels in the hippocampus and their relationship to regulation of the mRNA for one structural protein, GAP43 (Chao and McEwen, 1994; Chao et al., 1992). ADX increases GAP43 mRNA levels in the hippocampus, which is consistent with an inhibitory effect of adrenal steroids on synaptic plasticity (Chao et al., 1992). However, this effect may be mediated by type I receptors and may have no direct relationship to stress effects, since GAP43 mRNA levels were not suppressed to any greater extent by high versus low doses of COR!l' (Chao and McEwen, 1994). As noted above, the mRNA levels for the neurotrophic growth factors BDNF and NT-3 both appeared t o be suppressed by high levels of CORT in regions of Ammon's horn and in the dentate gyrus, whereas there was no effect of adrenal steroid manipulations on mRNA levels for the neurotrophic factor receptors trk B and trk C or the bFGF receptor (Chao and McEwen, 1994).Taken together, these results reveal a complex picture of regulation, particularly in view of data indicating that stress also alters neurotrophin mRNA levels in the hippocampus by both adrenal steroid-dependent and independent mechanisms (Smith et al., 1993).
VII. STRESS EFFECTS ON COGNITIVE PERFORMANCE IN RODENTS If the actions of stress and adrenal steroids on the hippocampus are significant for the functions of this brain structure, then what about impairment of cognitive function associated with stress or elevated levels of adrenal steroids? In the rat model in which we have seen atrophy of dendrites of the CA3c pyramidal neurons, does repeated stress impair the performance of behaviors that depend on the hippocampus? A recent study has demonstrated an impairment of performance on an eight-arm radial maze in rats that received 21 days of restraint stress prior to maze training (Luine et al., 1993). As shown in Fig. 12, the stress effect is reversible, in that inhibition of the initial learning was evident when rats were trained and evaluated immediately aRer the end of stress but not 18 days later. This impairment
392
BRUCE S. MCEWEN et al. 8
T
-
CONTROL
D STRESS I
T T
7 W v)
0 0
I 0
6
5
Mistake Correct EXPERIMENT 1
T
II
Mistake Correct EXPERIMENT 2
FIG.12. Effect restraint stress on radial arm maze performance. The choice where the first mistake occurred (mistake) and the number of correct choices in the first eight visits (correct) are plotted for control (solid bars) and stressed (open bars) rats. Entries are the average t SEM for four trials. In Experiment 1, 13 control and 14 stressed rats were trained and evaluated immediately after restraint sessions. In Experiment 2, six control and nine stressed rats were trained and evaluated 18 days after the last restraint session. Data were analyzed by a two-way repeated-measures analysis of variance (ANOVA) (group x trials). In Experiment 1,there was a significant group effect (Fl,25 = 5.31;P < 0.03) and for correct (Fl,25= 3.90;P < 0.05)but no significant trials or interaction effects. In Experiment 2, there were no significant effects. [From Luine et al. (1993)with permission.]
was in the same direction, but not as great, as that found in aging rats; moreover, stress effects were prevented by prior treatment of rats with phenytoin or with tianeptine under the same conditions in which phenytoin was able to prevent the stress-induced atrophy of CA3c pyramidal neurons (Watanabe et d.,1992a; Luine et d., 1993). These findings are supported by data from studies on the human hippocampus and cognitive performance in relation to stress, glucocorticoid administration, and depression (see the next section). These studies are a good beginning, but must be complemented by other tests of cognitive function that allow for more immediate monitoring of the actions of stress on hippocampal function. Long-term stress also accelerates a number of biological markers of aging in rats, including increasing the excitability of CA1 pyramidal neurons via a calcium-dependent mechanism and causing a loss of hippocampal pyramidal neurons (Kerr et al., 1991). An important factor may be the enhancement by glucocorticoids of calcium currents in
STRESS AND THE BRAIN
393
the hippocampus (Kerr et al., 19921, in view of the key role of calcium ions along with excitatory amino acids in destructive as well as plastic processes in hippocampal neurons (see Sections I and 11).
VIII. DEREGULATION OF THE HPA AXIS IN DEPRESSION AND OTHER DISORDERS Under normal circumstances adrenal steroids help to contain responses to stress and maintain stability within the brain, as well as perform important signaling roles. For example, as we have seen, they can exert biphasic effects on LTP and regulate the balance between cell birth and cell death in the dentate gyrus, which we hypothesize may change according to the season. As long as steroid levels are themselves contained, the beneficial effects predominate; however, when adrenal corticosteroid levels are persistently elevated, due to a breakdown in negative feedback by adrenal steroids or a failure of neural containment of CRH and ACTH secretion, then a cascade of events begins in which neural and peripheral systems begin to show abnormalities resulting from excess exposure to corticosteroids (McEwen et al., 1992). Various types of pathology may then ensue. These include neuronal atrophy in the hippocampus and associated cognitive impairment (see Section 11); imbalances within the 5-HT system due to elevated S-HT,, and suppressed 5-HTIAreceptors (Graeff, 1993; Kuroda et al., 1992; Brindley and Rolland, 1989; Winokur et al., 1988), which may enhance anxiety and reduce the ability to suppress memories of aversive associations (Graeff, 1993); and elevated insulin and corticosteroids, which increase fat deposition and can increase atherosclerosis (Winokur et al., 1988). Some of these characteristics are associated with major depression (Winokur et al., 1988) and/or with animal models of this disorder (Graeff, 1993). Deregulation of the HPA axis associated with major depression is revealed by the dexamethasone (DEX) suppression test. Although this test most likely works at the pituitary level (Miller et al., 19921, the underlying deregulation is undoubtedly of central nervous system origin and reflects increased drive upon the CRH and arginine vasopressin systems of the hypothalamus (Gold et al., 1988) and constitutes a form of endogenously driven stress. Limbic brain structures such as the amygdala and the hippocampus participate in the activation of HPA activity (Jacobson and Sapolsky, 1991) as well as in cognitive function, which may become impaired, at least in part, as a result of
394
BRUCE S. MCEWEN et a1
the chronically elevated glucocorticoids (Martignoni et al., 1992; Wolkowitz et al., 1990a,b; Newcomber et al., 1994). One of the possible reasons for cognitive impairment associated with elevated adrenal steroids is structural plasticity of the hippocampus, a brain structure that is vulnerable to both stress and the wear and tear of aging (Sapolsky, 1992; Section 11). Recent studies using magnetic resonance imaging have indicated that elevated cortisol levels are associated with some shrinkage of the hippocampal formation and mild cognitive impairment (Axelson et al., 1993; Starkman et al., 1992). In the first study of Cushing’s syndrome, left plus right hippocampal volumes were negatively correlated with plasma cortisol levels; moreover, a smaller hippocampal volume was correlated with poorer scores for verbal paired-associate learning, verbal recall, and verbal recall corrected for full-scale I& (Starkman et al., 1992). In another study post-DEX cortisol blood levels at 23:OO were negatively correlated with left hippocampal volume; moreover, left and right hippocampal volume were negatively correlated with the number of hospitalizations for depression, although depressed patients as a group did not have a smaller hippocampal volume than controls (Axelson et al., 1993). IX. EFFECTS OF EXOGENOUS GLUCOCORTICOID TREATMENT ON COGNITIVE PERFORMANCE IN HUMANS There is other evidence of cognitive impairment associated with adrenal steroid treatment (Martignoni et al., 1992; Wolkowitz et al., 1990a,b; Newcomber et al., 1994). In the most recent of these studies, DEX was given for 4 consecutive days at 23:OO and cognitive testing was performed at 14:OO on study days 0, 1, 4, and 7 after the last treatment (Newcomber et al., 1994). There was decreased declarative memory recall on study days 4 and 7 after the last treatment in the DEX-treated subjects compared to placebo-treated subjects (Newcomber et al., 1994). No other cognitive measures were affected by DEX, indicating the specificity to a task that is known to involve hippocampal-temporal lobe function. Because the poorest performance was found on study day 4, along with the lowest levels of endogenous cortisol, it is impossible to say whether the cognitive deficit is associated with the action of DEX on the brain or the reduction in cortisol. Since glucocorticoids have biphasic effects on synaptic plasticity (Diamond et al., 1992; Pavlides et al., 1993, 1995; Section IV,B), which are thought of as neurophysiological models of information storage processes (Bliss and Collingridge, 1993), either explanation is plausible.
STRESS AND THE BRAIN
395
X. CONCLUSIONS We have seen that adrenal steroids affect the structure and function of the hippocampus in a variety of ways, and we have also seen that, in human subjects, there is evidence for cognitive impairment as well as hippocampal atrophy associated with altered levels of adrenal steroids. What are the possible mechanisms for these changes? As discussed here, the acute inhibition of LTP and primed-burst potentiation by high levels of glucocorticoids represents one mechanism by which this might occur, while the atrophy of rat hippocampal CA3c neurons after either repeated stress or repeated glucocorticoid treatment, along with the reported atrophy of the hippocampus of Cushing’s syndrome patients and depressed subjects, constitutes a long-term mechanism for cognitive impairment. Another aspect of the relationship between adrenal steroid levels and cognitive function is illustrated by studies of basal cortisol levels and cognitive deficits in human aging (Lupien et al., 1994). Aged patients followed up over a 4-year period who showed a significant increase in cortisol levels over the 4 years and had high basal cortisol levels in year 4 showed deficits on tasks measuring explicit memory as well as selective attention, compared to those with either decreasing cortisol levels over 4 years or increasing basal cortisol but moderate current cortisol levels. Thus, once again, the question of acute versus chronic effects of cortisol is unsettled as a primary cause, and both types of mechanisms described here may apply. However, this study evokes the model proposed by Sapolsky (19921, called the “glucocorticoid cascade hypothesis,” in which rising cortisol levels compromise the hippocampus by destroying neurons and contribute to cognitive impairment as well as reduce the inhibitory effect of the hippocampus on cortisol secretion. Consistent with this hypothesis is the fact that lesions of the fornix or the hippocampal formation-temporal lobe region of cynomolgus monkeys produced an elevation in cortisol secretion for at least 6 months postlesion (Sapolsky et al., 1991). A final point to be made about adrenal steroids, stress, and the hippocampus is that adrenal steroids do not act alone. Rather, we have seen that both EAAs and 5-HT release, possibly facilitated by circulating glucocorticoids, play a key role. In fact, the final common pathway for CA3c dendritic atrophy in rats treated either with or by restraint stress involves processes that are blocked by blocking glutamate release or actions using phenytoin or an NMDA antagonist, respectively, o r by facilitating 5-HT reuptake using tianeptine. The efficacy of these agents raises the attractive possibility of treating individuals-
396
BRUCE S. MCEWEN et al.
perhaps the depressed elderly-with agents such as phenytoin or tianeptine as a means of improving cognitive function. If such studies are carried out, it will be important to determine the degree to which hippocampal volume may be increased by such treatments and whether long-term treatment protects these individuals from dementia. ACKNOWLEDGMENTS Research support was provided by U.S. National Institutes of Health grant MH-41256, by the Health Foundation (New York), and by Servier (France). REFERENCES Abraham, W. C., Dragunow, M., and Tate, W. P. (1991). The role of immediate early genes in the stabilization of long-term potentiation. Mol. Neurobiol. 5, 297-314. Akana, S., Shinsaki, J., and Dallman, M. (1983). Drug-induced adrenal hypertrophy provides evidence for reset in the adrenocortical system. Endocrinology 113, 22322237. Albeck, D. S., Hasting, N. B., and McEwen, B. S. (1994). Effects of adrenalectomy and type I or type I1 glucocorticoid receptor activation on AVP and CRH mRNA in the rat hypothalamus. Mol. Brain Res. 26, 129-134. Armstrong, J. D., McIntyre, D. C., Neubort, S., and Sloviter, R. S. (1993). Learning and memory after adrenalectomy-induced hippocampal dentate granule cell degeneration in the rat. Hippocampus 3, 359-371. Asai, T., Wanaka, A., and Kato, H. (1993). Differential expression of two members of FGF receptor gene family, FGFR-1 and FGFR-2 mRNA, in the rat central nervous system. Mol. Brain Res. 17, 174-178. Aus der Muhlen, K., and Ockenfels, H. (1969). Morphologische veranderungen im diencephalon und telencephalon: Storungen des regelkreises adenohypophysenebennierenrinde. 2.Zellforsch. Mikrosk. Anat. 93, 126-141. Axelson, D., Doraiswamy, A. P., McDonald, W., Boyko, O., Qpler, L., Patterson, L., Nemeroff, C. B., Ellinwood, E. H. and Krishan, K. R. R. (1993). Hypercortisolemia and hippocampal changes in depression. Psychiatry Res. 47, 163-173. Bading, H., Ginty, D., and Greenburg, M. (1993). Regulation of gene expression in hippocampal neurons by distinct calcium signalling pathways. Science 260, 181-186. Barbany, G., and Persson, H. (1992).Regulation of neurotrophin mRNA expression in the rat brain by glucocorticoids. Eur. J. Neurosci. 4, 396-403. Barbany, G., and Persson, H. (1993). Adrenalectomy attenuates kainic acid-elicited increases of mRNA’s for neurotrophins and their receptors in the rat brain. Neuroscience 54, 909-922. Beasley, C., Masica, D., and Potvin, J. (1992). Fluoxetine: A review of receptor and functional effects and their clinical implications. Psychopharmacology 107, 1-10, Bliss, T.V. P., and Collingridge, G. L. (1993). A synaptic model of memory: Long-term potentiation in the hippocampus. Nature 361, 31-39. Bliss, T. V. P., and Lomo, T. (1973). Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path. J . Physiol. 232, 331-356. Brindley, D., and Rolland, Y. (1989). Possible connections between stress, diabetes, obesity, hypertension and altered lipoprotein metabolism that may result in atherosclerosis. Clin. Sci. 77. 453-461.
STRESS AND THE BRAIN
397
Cameron, H. A,, Woolley, C. S., and Gould, E. (1993). Adrenal steroid receptor irnmunoreactivity in cells born in the adult rat dentate gyrus. Brain Res. 611, 342-346. Cammerrneyer, J. (1978). Is the solitary dark neuron a manifestation of postmortem trauma to the brain inadequately fixed by perfusion? Histochemisty 56, 97-115. Chao, H., and McEwen, B. S.(1994). Glucocorticoids and the expression of mRNA's for neurotrophins, their receptors, and GAP43 in the rat hippocampus. Mol. Brain Res. 26,271-276. Chao, H., Spencer, R., Sakai, R., and McEwen, B. S. (1992). The expression of growthassociated protein, GAP-43, mRNA in the rat hippocampus in response to adrenalectomy and aging. Mol. Cell. Neurosci. 3, 529-535. Cheng, B., and Mattson, M. P. (1994). NT-3 and BNF protect CNS neurons against rnetabolic/excitotoxic insult. Bruin Res. 640, 56-67. Conrad, C. D., and Roy, E. J. (1993). Selective loss of hippocampal granule cells following adrenalectomy: Implications for spatial memory. J. Neurosci. 13, 2582-2590. Delbende, C., Contesse, V., Mocaer, E., Kamoun, A., and Vaudry, H. (1991). The novel antidepressant, tianeptine, reduces stress-evoked stimulation of the hypothalamopituitary-adrenal axis. Eur. J. Phnrmacol. 202, 391-396. Diamond, D. M., Bennet, M. C., Fleshner, M., and Rose, G. M. (1992). Inverted-U relationship between the level of peripheral corticosterone and the magnitude of hippocampal primed burst potentiation. Hippocampus 2, 421-430. Diamond, D. M., Fleshner, M., and Rose, G. M. (1994). Psychological stress repeatedly blocks hippocampal primed burst potentiation in behaving rats. Behnu. Brain Res. 62, 1-9. Eichenbaum, H., and Otto, T. (1992).The hippocampus-What does it do? Behau. Neural Biol. 57, 2-36. Foy, M. R., Stanton, M. E., Levine, S., and Thompson, R. F. (1987). Behavioral stress impairs long-term potentiation in rodent hippocampus. Behau. Neural Biol. 48,138149. Galea, L. A. M., Kavaliers, M., Ossenkopp, K. P., Innes, D., and Hargreaves, E. L. (1994). Sexually dimorphic spatial learning varies seasonally in two populations of deer mice. Bruin Res. 635, 18-26. Gall, C. (1992). Regulation of brain neurotrophin expression by physiological activity. Trends Pharmacol. Sci. 13,401-403. Gold, P., Goodwin, F., and Chrousos, G. (1988). Clinical and biochemical manifestations of depressions. N . Engl. J. Med. 319,348-353 (part 1) and 413-420 (part 2). Gomez-Pinilla, F., Lee, J., and Cotman, C. (1992). Basic FGF in adult rat brain: Cellular distribution and response to entorhinal lesion and fimbra-fornix transection. J . Neurosci. 12, 345-355. Gould, E., and McEwen, B. S. (1993).Neuronal birth and death. Curr. Opin. Neurobiol. 3, 676-682. Gould, E., Woolley, C., and McEwen, B. S. (1990). Short-term glucocorticoid manipulations affect neuronal morphology and survival in the adult dentate gyrus. Neuroscience 37, 367-375. Graeff, F. (1993). Role of 5-HT in defensive behavior and anxiety. Rev. Neurosci. 4, 181211. Halpain, S. and Greengard, P. (1990). Activation of NMDA receptors induces rapid dephosphorylation of the cytoskeletal protein MAPP. Neuron 5, 237-246. Herman, J., Patel, P., Akil, H., and Watson, S. J. (1989). Localization and regulation of glucocorticoid and mineralocorticoid receptor mRNAs in the hippocampal formation of the rat. Mol. Endocrinol. 3, 1886-1894.
398
BRUCE S. MCEWEN et al.
Hsu, M., and Buzsaki, G. (1993). Vulnerability of mossy fiber targets in the rat hippocampus to forebrain ischemia. J. Neurosci. 13, 3964-3979. Ishizuka, N., Weber, J., and Amaral, D. G. (1990). Organization of the intrahippocampal projections originating from CA3 pyramidal cells in the rat. J. Comp. Neurol. 295, 580-623. Jacobson, L., and Sapolsky, R. (1991). The role of the hippocampus in feedback regulation of the hypothalamic-pituitary-adrenocortical axis. Endocrinol. Rev. 12, 118134. Joels, M., and DeKloet, E. R. (1992). Control of neuronal excitability by corticosteroid hormones. Trends NeuroSci. 15,25-30. Kennet, G. A,, Dickinson, S. L., and Curzon, G. (1985). Enhancement of some 5-HT dependent behavioral responses following repeated immobilization in rats. Bruin Res. 330, 253-260. Kerr, S. D., Campbell, L. W., Applegate, M. D., Brodish, A., and Landfield, P. W. (1991). Chronic stress-induced acceleration of electrophysiologic and morphometric biomarkers of hippocampal aging. J . Neurosci. 11, 1316-1324. Kerr, S., Campbell, L., Thibault, O., and Landfield, P. (1992). Hippocampal glucocorticoid receptor activation enhances voltage-dependent Ca2+ conductances: Relevance to brain aging. Proc. Natl. Acad. Sci. U.S.A. 89,8527-8531. Kiyota, Y., Takami, K., and Iwane, M. (1991). Increase in basic fibroblast growth factorlike immunoreactivity in rat brain after forebrain ischemia. Bruin Res. 545, 322328. Kuroda, Y., Mikuni, M., Ogawa, T., and Takahashi, K. (19921. Effect of ACTH, adrenalectomy and the combination treatment on the density of 5-HT2 receptor binding sites in neocortex of rat forebrain and 5-HT2 receptor-mediated wet-dog shake behaviors. Psychopharmacology 108, 27-32. Kuroda, Y., Watanabe, Y., Albeck, D. S., Hastings, N. B., and McEwen, B. S. (1994). Effects of adrenalectomy and type I and type I1 adrenal steroid receptor activation on 5-HT1A and 5-HT2 receptor binding and 5-HT transporter mRNA expression in rat brain. Bruin Res. 647, 157-161. Landfield, P. (1987). Modulation of brain aging correlates by long-term alterations of adrenal steroids and neurally-active peptides. Prog. Bruin Res. 72, 279-300. Lee, K., Frank, S., and Vanderklish, P. (19911. Inhibition of proteolysis protects hippocampal neurons from ischemia. Proc. Natl. Acud. Sci. U.S.A.88, 7233-7237. Lindholm, D., Castren, E., and Hengerer, B. (1992). Differential regulation of nerve growth factor (NGF) synthesis in neurons and astrocytes by glucocorticoid hormones. Eur. J . Neurosci. 4, 404-410. Lowenstein, D., Thomas, M. J., Smith, D. H., and McIntosh, T. K. (1994). Selective vulnerability of dentate hilar neurons following traumatic brain injury: A potential mechanistic link between head trauma and disorders of the hippocampus. J. Neurosci. 12, 4846-4853. Lowy, M. T., Gault, L., and Yamamoto, B. K. (1993). Adrenalectomy attenuates stressinduced elevations in extracellular glutamate concentrations in the hippocampus. J . Neurochem. 61, 1957-1960. Luchins, D. G. (1990). A possible role of hippocampal dysfunction in schizophrenic symptomatology. Biol. Psychiatry 28, 87-91. Luine, V., Villegas, M., Luine, V. N., and McEwen, B. S. (1993). Repeated stress causes reversible impairments of spatial memory performance. Behav. Bruin Res. 639,167170. Lupien, S., Lecours, A. R., Lussier, I., Schwartz, G., Nair, N. P. V., and Meaney, M. J.
STRESS AND THE BRAIN
399
(1994). Basal cortisol levels and cognitive deficits in human aging. J . Neurosci. 14, 2893-2903. Magarifios, A., and McEwen, B. S. (1995). Stress-induced atrophy of apical dendrites of
hippocampal CA3c neurons: Role of glucocorticoid secretion and excitatory amino acid receptors. Neuroscience, in press. Martignoni, E., Costa, A., Sinforiani, E., Liuzzi, A., Chiodii, P., Mauri, M., Bono, G., and Nappi, G.'(1992). The brain as a target for adrenocortical steroids: Cognitive implications. Psychoneuroendocrinology 17, 343-354. Matsuyama, A,, Iwata, H., and Okumura, N. (1992). Localization of basic fibroblast growth factor-like immunoreactivity in the rat brain. Brain Res. 587, 49-65. Mattson, M., Dou, P., and Kater, S. (1988). Outgrowth-regulating actions of glutamate in isolated hippocampal pyramidal neurons. J. Neurosci. 8, 2087-2100. McEwen, B. S., and Stellar, E. (1993). Stress and the individual: Mechanisms leading t o disease. Arch. Intern. Med. 153, 2093-2101. McEwen, B. S., Weiss, J. M., and Schwartz, L. S. (1968). Selective retention of corticosterone by limbic structures in rat brain. Nature 220, 911-912. McEwen, B. S., Angulo, J., Cameron, H., Chao, H., Daniels, D., Gannon, M., Gould, E., Mendelson, S., Sakai, R., Spencer, R., and Woolley, C. (1992). Paradoxical effects of adrenal steroids on the brain: Protection versus degeneration. Biol. Psychiatry 31, 177-199.
McKittrick, C. R., Blanchard, D. C., Blanchard, R. J.,McEwen, B. S., and Sakai, R. R. (1995). Serotonin receptor binding in a colony model of chronic social stress. Bwl. Psychiatry 37, 383-393. Mendelson, S., and McEwen, B. S. (1992). Autoradiographic analyses of the effects of adrenalectomy and corticosterone on 5-HT,, and 5-HT,, receptors in the dorsal hippocampus and cortex of the rat. Neuroendocrinology 55,444-450. Mennini, T., and Miari, A. (1991). Modulation of 3H glutamate binding by serotonin in r a t hippocampus: An autoradiographic study. Life Sci.49, 283-292. Mennini, T., Mocaer, E., and Garattini, S. (1987). Tianeptine, a selective enhancer of serotonin uptake in rat brain. Naunyn-Schrniedeberg's Arch. Pharrnacol. 336,478482.
Miller, A. H., Spencer, R. L., Pulera, M., Kang, S., McEwen, B. S., and Stein, M. (1992). Adrenal steroid receptor activation in rat brain and pituitary following dexamethasone: Implications for the dexamethasone suppression test. Biol. Psychiatry 32,850869.
Miner, J. N., and Yamamoto, K. R. (1991). Regulatory crosstalk a t composite response elements. Trends Biochern. Sci. 16, 423-426. Mondadori, C., and Weizkrantz, L. (1993). NMDA receptor blockers facilitate and impair learning via difference mechanisms. Behw. Neural Biol. 60,205-210. Morgan, J., and Curran, T. (1989). Stimulus-transcription coupling in neurons: Role of cellular immediate-early genes. Trends Neurosci. 12, 459-461. Needels, D., and Cotman, C. (1988). Basic fibroblast growth factor increases the survival of rat dentate gyrus cells in culture. Abstr. Soc. Neurosci. 14, 363. Newcomber, J. S., Craft, S., Hershey, T., Askins, K., and Bardgett, M. E. (1994). Glucocorticoid-induced impairment in declarative memory performance in adult humans. J. Neurosci. 14, 2047-2053. Nixon, R., Quackenbush, R., and Vitto, A. (1986). Multiple calcium-activated neutral proteinases (CANP) in mouse retinal ganglion cell neurons: Specificities for endogenous neuronal substrates and comparison to purified brain CANP. J. Neurosci. 6, 1252-1263.
400
BRUCE S. MCEWEN et al.
Orchinik, M., Weiland, N. G., and McEwen, B. S. (1994). Adrenalectomy selectively alters GABA, receptor subunit mRNA levels in rat hippocampus. Mol. Cell. Neurosci. 5, 451-458. Packan, D. R., and Sapolsky, R. M. (1990). Glucocorticoid endangerment of the hippocampus: Tissue, steroid and receptor specificity. Neuroendocrinology 51, 613-618. Pavlides, C., Watanabe, Y., and McEwen, B. S. (1993). Effects of glucocorticoids on hippocampal long-term potentiation. Hippocumpus 3, 183-192. Pavlides, C., Kimura, A., Magarifios, A. M., and McEwen, B. S. (1994). Type I adrenal steroid receptors prolong hippocampal long-term potentiation. NeuroReport 5, 2673-2677. Pavlides, C., Watanabe, Y., Magarifios, A. M., and McEwen, B. S. (1995). Opposing role of type I and type I1 adrenal steroid receptors in hippocampal long-term potentiation. Neuroscience, in press. Popov, V., and Bocharova, L. (1992). Hibernation-induced structural changes in synatpic contacts between mossy fibers and hippocampal pyramidal neurons. Neuroscience 48,53-62. Popov, V., Bocharova, L., and Bragin, A. (1992). Repeated changes of dendritic morphology in the hippocampus of ground squirrels in the course of hibernation. Neuroscience 48, 45-51. Rahmann, S., and Neumann, R. S. (1993). Activation of 5-HT2 receptors facilitates depolarization of neocortical neurons by N-methyl-D-aspartate. Eur. J . Phurmacol. 231,347-354. Sapolsky, R. (1992). “Stress, the Aging Brain and the Mechanisms of Neuron Death.” Sapolsky, R., Krey, L., and McEwen, B. S. (1985). Prolonged glucocorticoid exposure reduces hippocampal neuron number: Implications for aging. J . Neurosci. 5, 12221227. Sapolsky, R. M., Zola-Morgan, S., and Squire, L. R. (1991). Inhibition of glucocorticoid secretion by the hippocampal formation in the primate. J . Neurosci. 11,3695-3704. Schreiber, S. S., Tocco, G., Shors, T. J., and Thompson, R. F. (1991). Activation of immediate early genes after acute stress. NeuroReport 2, 17-20. Sheng, M., and Greenberg, M. (1990). The regulation and function of c-fos and other immediate early genes in the nervous system. Neuron 4, 477-485. Sherry, D. F., Jacobs, L. F., and Gaulin, S. J . (1992). Spatial memory and adaptive specialization of the hippocampus. Trends Neurosci. 15, 298-303. Shors, T. J., and Thompson, R. F. (1992). Acute stress impairs (or induces) synaptic longterm potentiation (LTP) but does not affect paired-pulse facilitation in the stratum radiatum of the rat hippocampus. Synapse 11, 262-265. Shors, T. J., Seib, T. B., Levine, S., and Thompson, R. F. (1989). Inescapable versus escapable shock modulates long-term potentiation in the rat hippocampus. Science 244,224-226. Shors, T. J., Levine, S., and Thompson, R. F. (1990a). Effect of adrenalectomy and demedullation on the stress-induced impairment of long-term Potentiation. Neuroendocrinology 5 1, 70- 75. Shors, T. J., Levine, S., and Thompson, R. F. (1990b). Opioid antagonist eliminates the stress-induced impairment of long-term potentiation (LTP). Bruin Res. 506, 316318. Sloviter, R. (1983). “Epileptic” brain damage in rats induced by sustained electrical stimulation of the perforant path. I. Acute electrophysiological and light microscopic studies. Bruin Res. Bull. 10, 675-697. Sloviter, R. S., Valiquette, G., Abrams, G. M., Ronk, C., Sollas, A. I., Paul, L. A., and
STRESS AND THE BRAIN
40 1
Neubort, S. A. (1989). Selective loss of hippocampal granule cells in the mature rat brain after adrenalectomy. Science 243,535-538. Smith, M., Makino, S., Kvetnansky, R., and Post, R. (1993). Stress-induced changes in neurotrophic factor mRNA levels in rat hippocampus. Neuropsychophurmncology 9, 177. Starkman, M., Gebarski, S., Berent, S., and Schteingart, D. (1992). Hippocampal formation volume, memory dysfunction, and cortisol levels in patients with Cushing’s syndrome. Biol. Psychiatry 32, 756-765. Steward, O., Davis, L., Dotti, Phillips, L., Rao, A., and Banker, G. (1988). Protein synthesis and processing in cytoplasmic microdomains beneath postsynaptic sites on CNS neurons. Mol. Neurobiol2, 227-261. Uno, H., Ross, T., Else, J.,Suleman, M., and Sapolsky, R. (1989). Hippocampal damage associated with prolonged and fatal stress in primates. J . Neurosci. 9, 1705-1711. Vaher, P. R., Luine, V. N., Gould, E., and McEwen, B. S. (1994). Effects of adrenalectomy on spatial memory performance and dentate gyrus morphology. Brain Res. 656,7178. Walicke, P., Cowan, W., and Ueno, N. (1986). Fibroblast growth factor promotes survival of dissociated hippocampal neurons and enhances neurite extension. Proc. Natl. Acad. Sci. U.S.A. 83, 3012. Watanabe, Y.,Gould, E., Cameron, H., Daniels, D., and McEwen, B. S. (1992a). Phenytoin prevents stress- and corticosterone-induced atrophy of CA3 pyramidal neurons. Hippocampus 2,431-436. Watanabe, Y.,Gould, E., Cameron, H., Daniels, D., and McEwen, B. S. (1992b). Stress and antidepressant effects on hippocampus. Eur. J. Phnrmacol. 222, 157-162. Watanabe, Y., Gould, E., and McEwen, B. S. (1992~).Stress induces atrophy of apical dendrites of hippocampus CA3 pyramidal neurons. Brain Res. 588, 341-344. Watanabe, Y., Sakai, R., McEwen, B. S., and Mendelson, S. (1993). Stress and antidepressant effects on hippocampal and cortical 5-HT1A and 5-HT2 receptors and transport sites for serotonin. Brain Res. 615, 87-94. Watanabe, Y., Akabashi, A,, and McEwen, B. S. (1995). Adrenal steroid regulation of NPY mRNA: Differences between dentate hilus and locus coeruleus and arcuate nucleus. Mol. Brain Res. 28, 135-140. Stone, E., and McEwen, B. S. t1994b). Induction and habituation of c-FOS Watanabe, Y., and ZIFl268 by acute and repeated stressors. NeuroReport 5, 1321-1324. Watanabe, Y., Weiland, N. G., and McEwen, B. S. (1995). Effects of adrenal steroid manipulations and repeated restraint stress on dynorphin mRNA levels and excitatory amino acid receptor binding in hippocampus. Brain Res. submitted. Whitton, P., Sarna, G., O’Connell, M. T., and Curzon, G. (1991). The effect of the novel antidepressant tianeptine on the concentration of 5-hydroxytryptamine in rat hippocampal dialysates in vivo. Neurophurmacology 30, 1-4. Winokur, A., Maislin, G., Phillips, J., and Amsterdam, J. (1988). Insulin resistance after oral glucose tolerance testing in patients with major depression. Am. J.Psychiatry 145, 325-330. Wolkowitz, O., Reuss, V., and Weingartner, H. (1990a). Cognitive effects of corticosteroids. A n . J. Psychiatry 147, 1297-1303. Wolkowitz, O., Rubinow, D., and Doran, A. (1990b). Prednisone effects on neurochemistry and behavior: Preliminary findings. Arch. Gen. Psychiatry 47, 963-968. Woodward, W., Nishi, R., and Meshul, C. (1992). Nuclear and cytoplasmic localization of basic fibroblast growth factor in astrocytes and CA2 hippocampal neurons. J. Neuroscz. 12, 142-152.
402
KEITH L. PARKER AND BERNARD P. SCHIMMER
Woolley, C., and McEwen, B. S. (1992). Estradiol mediates fluctuation in hippocampal synapse density during the estrous cycle in the adult rat. J . Neurosci. 12, 25492554. Woolley, C., and McEwen, B. S. (1993). Estradiol regulates hippocampal dendritic spine density via an NMDA receptor dependent mechanism. Abstr. Soc. Neurosci. 19,379. Woolley, C., Gould, E., and McEwen, B. S. (1990). Exposure to excess glucocorticoids alters dendritic morphology of adult hippocampal pyramidal neurons. Bruin Res. 531,225-231. Woolley, C., Gould, E., Sakai, R., Spencer, R., and McEwen, B. S. (1991). Effects of aldosterone or RU28362 treatment on adrenalectomy-induced cell death in the dentate gyrus of the adult rat. Bruin Res. 554, 312-315. Worley, P. F., Bhat, R. V., Baraban, J. M., Erickson, C. A., McNaughton, B. L., and Barnes, C. A. (1993). Thresholds for synaptic activation of transcription factors in hippocampus-correlation with long-term enhancement. J . Neurosci. 13, 47764786.
VITAMINS AND HORMONES, VOL. 61
Retinoids and Mouse Embryonic Development T. MICHAEL UNDERHILL, LORI E. KOTCH, AND ELWOOD LINNEY Departments of Microbiology and Pharmacology Duke University Medical Center Durham, North Carolina 27710 I. Introduction 11. Retinoid Receptors
A. Background Information B. Functional Organization C. Factors That Influence Retinoid Receptor Activity D. Summary III. Retinoid-Binding Proteins A. A Family of Retinoid-Binding Proteins B. Transgenic Studies with Retinoid-Binding Protein Genes IV. Retinoid Receptor Expression Patterns and Genetic Analysis of Function A. RARs B. RXRs C. Summary V. Conclusions and Perspectives A. Retinoid Regulation of Important Development Target Genes B. Retinoid Receptor-Mediated Inhibition of AP-1 C. Future Directions References
I. INTRODUCTION Embryogenesis encompasses a very dynamic state of development characterized by precise temporal and spatial cell-tissue inductive interactions. The roles retinoids play during embryonic development are both basic and complex. This general area of research developed as a result of nutritional studies with animals which underscored the importance of vitamin A and its various metabolites in many biological processes. The nutritional studies at first examined the consequence of vitamin A-deficient diets on pregnant females and their fetuses. Pregnant females maintained on this diet produced offspring that had numerous congenital malformations (Hale, 1935;Warkany and Schraffenberger, 1946;Wilson and Warkany, 1948,1949;Wilson et al., 1953).However, a variety of congenital abnormalities were also 403
Copyright 1: 1995 by Academic Press, Inc. All rights of reproduction In any form reserved.
404
T.MICHAEL UNDERHILL et al.
observed when embryos were exposed in utero to a n excess of vitamin A (Cohlan, 1953; Kalter and Warkany, 1961). Kochhar (1967) reported that the acid form of vitamin A, retinoic acid (RA), was a more potent teratogen than vitamin A. This finding played a large role in the development of this area of research. Subsequent studies with rodents examined various aspects of retinoid teratology. The Shenefelt (1972) study underlined the variety of effects of RA on many different stages of development. RA was administered to pregnant golden hamsters a t various stages of embryonic development and the LD,, was determined for each stage. The effects of a standard dosage were then anatomically examined later in development and tabulated. These results demonstrated that RA-induced teratogenesis includes a very broad range of defects, which present themselves differentially and predictably, in accordance with the time of administration. A plethora of data further defining these susceptible developmental windows has since been collected and is discussed in a recent review (Linney and LaMantia, 1994). Two isomers of RA, all-trans RA (tretinoin, Retin-A@,RA) and 1 3 4 s RA (isotretinoin, Accutane R, and 13c RA), have been approved for therapeutic use in humans in the United States for several dermatological diseases. From the above rodent studies, one could have predicted the dramatic effects certain retinoids might have on human embryonic development. Unfortunately, however, and despite clear warning labels concerning potential teratogenic risk, since 1983 hundreds of women have been exposed to 13c RA during pregnancy. Many of these women bore offspring with defects similar to those described in the above experimental rodent work. Among the most prevalent malformations reported were central nervous system anomalies, cardiovascular defects, limb defects, and characteristic craniofacial dysmorphogenesis. This malformation complex associated with retinoid exposure has been termed RA embryopathy (Lammer et al., 1985). These clinical findings, in association with the experimental work, have prompted considerable interest in understanding the role of retinoids in embryonic development. Numerous components of the retinoid signaling pathway in vertebrates have now been identified. These include several molecules important in the metabolism and binding of retinoids and, probably most importantly, the retinoid-dependent transcription factors. With the identification of these intermediates, we are now in a position to understand how retinoids a t the molecular level influence cell behavior, thereby generating biological form.
RETINOIDS A N D MOUSE EMBRYOS
405
11. RETINOID RECEPTORS A. BACKGROUND INFORMATION Hyper- and hypovitaminosis A, as mentioned in Section I, leads to malformations in a number of anatomical structures, which can include the central nervous, skeletal, urogenital, cardiovascular, and immune systems. Further studies with RA, an active metabolite of vitamin A, have shown that RA is a potent mediator of these effects and that most mammalian tissues, at some point in development, are susceptible. From these studies and others examining the effect of RA on pattern formation, it has been inferred that RA may act as a morphogen. For a long time an operational explanation for the action of RA remained elusive. As discussed later, the cytoplasmic RA-binding proteins (CRABPs) were thought to be the primary mediator of these observed effects. It was not until 1987, when two independent groups reported on the isolation of a nuclear receptor for RA, that the picture became clearer. The first RA receptor (RAR), termed RARa, was isolated using probes based on DNA sequences from the highly conserved DNA binding domain found within members of the steroid hormone receptor superfamily (Giguere et al., 1987; Petkovich et at.,1987). Subsequently, two additional RARs have been isolated, termed RARp (Brand et al., 1988) and RARy (Zelent et al.,1989). Recently, a second group of RA-binding nuclear receptors has been identified, named the retinoid X receptors (RXRs) (Hamada et al., 1989; Mangelsdorf et al., 1990). As with the RARs, three RXRs have been isolated: a, p, and y (Hamada et al., 1989; Mangelsdorf et al., 1990, 1992; Yu et al., 1991; Leid et al., 1992b). The retinoid receptors have been grouped into the nuclear steroid hormone receptor superfamily because of their extensive structural similarity to members therein (Laudet et al., 1992). The RARs share the most homology with a subgroup of this family, which includes the thyroid hormone receptor (TR) and, t o a lesser extent, the vitamin D receptor (VDR) and certain receptors for which a ligand has not been identified (orphan receptors) (Laudet et al., 1992).The RXRs appear to be more closely related to the orphan receptor chicken ovalbumin upstream promoter transcription factor (COUP-TF) and Drosophila ultraspiracle gene (usp) (Oro et al.,1990; Laudet et at.,1992). The members of this subfamily share a common architecture, which has been divided into six prominent structural domains, A through F for the RARs and A through E for the RXRs. These domains, shown in Fig. 1,
406
T.MICHAEL UNDERHILL et al.
-A
t ransactivation
p.
B
C
D
E
F
-
P
es P,
DNA binding dimerization
nuclear localization ligand binding
dimerization
transactivation
FIG.1. Schematic representation of the functional domains of the retinoid receptors. A, or variable, domain contains putative phosphorylation sites and in conjunction with the B domain has ligand-independent transactivation properties (AF-1) and modulates activity of AF-2. The C domain contains a weak dimerization region which extends partly into the D domain. Two zinc fingers reside in this domain and bind DNA with sequence specificity. A nuclear localization signal is present within the D section. The E domain contains multiple overlapping activities involved in receptor dimerization (nine heptad repeats), ligand binding, and ligand-dependent transactivation (AF-2). The F domain is absent from RXRs and no function has been ascribed to it in RARs. N, nucleus; L, ligand; P, phosphate group.
contain regions important for ligand binding, DNA binding, transactivatiodrepression, protein-protein interactions, and nuclear localization. The RARs and RXRs exhibit a high degree of conservation of some domains and subdomains. Most notably, the RARs and RXRs share the highest degree of amino acid sequence homology in their DNA binding domains. However, they are clearly separated by the level of homology in their E domains, for which they share less than 40% homology. Overall, the C, D, and E domains are highly conserved within the individual receptor classes. The A/B and F domains are less well con-
RETINOIDS AND MOUSE EMBRYOS
40 7
served among receptors of the same class, but are evolutionarily conserved. The RARs and RXRs share a common gene structure. The primary difference in their respective gene structures is that the RXRs contain an intron within the first zinc finger coding sequence, while the RARs contain an intron between the first and second zinc finger coding sequences (Leid et al., 1992a; Liu and Linney, 1993; Nagata et al., 1994).
B. FUNCTIONAL ORGANIZATION 1. AIB Domain
All of the retinoid receptors contain a variable region at the NH, terminus, termed the A domain. Isoforms that differ in this region have been isolated for all of the RARs and RXRs (Gigu&reet al., 1990; Kastner et al., 1990; Leroy et al., 1991a; Zelent et al., 1991; Nagpal et al., 1992b; Liu and Linney, 1993; Nagata et al., 1994). These variable domains arise through a combination of alternative promoter usage andlor alternative splicing (Leroy et al., 1991a,b; Zelent et al., 1991; Nagpal et al., 1992b; Liu and Linney, 1993; Nagata et al., 1994). In the case of RARa, up to seven different isoforms have identified, RARal and RARa2 being the predominant isoforms (Leroy et al., 1991a). These two isoforms are expressed from different promoters (Leroy et aZ., 1991a,b). With respect to RARp, four different isoforms have been identified (Zelent et al., 1991; Nagpal et al., 1992b). RARPl and -3 share an upstream promoter, while RARP2 and -4 use a downstream promoter that is embedded within an intron of RARPl and -3 (Zelent et al., 1991; Nagpal et al., 1992b). Two isoforms have been isolated for RARy, 1and 2, which are expressed from different promoters (Gigukre et al., 1990; Kastner et al., 1990; Lehmann et al., 1991, 1992). Similarly, each of the RXRs are represented by at least two isoforms (Fleischhauer et al., 1992; Leid et al., 199213; Liu and Linney, 1993; Nagata et al., 1994). As is the case with the RARs, isoforms of RXRP and RXRy are expressed from different promoters (Liu and Linney, 1993; Nagata et al., 1994). Preliminary data from our laboratory suggest that the two identified isoforms of RXRa are also expressed from different promoters (T. M. Underhill, C. W. Hoopes, and E. Linney, unpublished observations, 1993). Hence, the A region distinguishes one receptor isoform from another. The B region has been defined at its 5' end by a splice acceptor site that joins it to the A domain sequences and is bounded on its 3' end by sequences in the C domain. Although the entire AIB region can comprise the first one fifth of
408
T.MICHAEL UNDERHILL et al.
the receptor protein, its exact role in receptor function is not entirely clear. In other steroid hormone receptors this region can function as a transactivation domain, now termed AF1 (Green and Chambon, 1988). Transient transfection experiments using reporter constructs containing various RA response elements (RAREs) have shown that in some instances the A domain and/or the A/B domain is dispensable for transactivation (Nagpal et al., 1992a). However, evidence has recently been obtained which suggests that the A/B domain is able to function independently as a transactivation domain (Folkers et al., 1993; Nagpal et al., 1993).Furthermore, overall receptor transactivation activity may be modulated, in part, by the A/B domain in conjunction with the context of the promoter (Nagpal et al., 1992a). The observations that the specific isoforms are so highly conserved between species and that this level of conservation extends to their expression patterns suggest an important role for the A/B domain in receptor function. One possibility is that this domain is important in modulating receptor function in a cell type- and ligand-dependent fashion. Evidence for this comes from observations that an isoform of RARa is phosphorylated during the course of RA-induced differentiation of F9 embryonal carcinoma (EC) cells (Gaub et al., 19921, and several RARp isoforms that exhibit distinct patterns of expression are differentially phosphorylated in response to ligand (Rochette-Egly et al., 1992). In this manner various signaling cascades could potentially feed into the retinoid receptor pathway to modify their function. This would certainly be consistent with the expression patterns of the receptors, whereby different isoforms could respond to or be modified by different resident cellular signaling systems. 2. C or D N A Binding Domain The C region comprises two classical zinc finger motifs (CI and CII), each of which contains four cysteines that coordinate zinc to mediate DNA binding of the receptor monomer to the canonical half-site A/GGG/TTCA (Umesono et al., 1988,1991; Umesono and Evans, 1989; Naar et al., 1991). A region within the CI finger, termed the P box, has been found to make direct contact with specific bases in the major groove (Luisi et al., 19911, and as such, appears to be the primary determinant of half-site sequence specificity (Truss and Beato, 1993; Glass, 1994). The retinoid receptors function primarily as dimers, and as such, RAREs contain two half-sites to presumably allow DNA binding of each receptor monomer (Glass, 1994).The spacing between these halfsites is very important in determining the nature of receptor binding
RETINOIDS AND MOUSE EMBRYOS
409
(Naar et al., 1991; Umesono et al., 1991; Forman et al., 1992). For instance, direct repeat (DR) half-sites spaced by 2 bp (DR2) (Smith et al., 1991; Durand et al., 1992) or 5 bp (DR5) (Yuet al., 1991; Leid et al., 1992b) favor binding of an RXR/RAR heterodimer. In contrast, homodimers of RXRs or RXR/RAR heterodimers can bind a DR1 site (Durand et al., 1992; Zhang et al., 1992a; Kurokawa et al., 1994).RARs do not appear to bind DNA as a homodimer efficiently because they lack the requisite homodimerization domain within the C region (Lee et al., 1993; Mader et al., 1993a). Spacing between the half-sites is therefore important in determining receptor binding specificity (Umesono et al., 1991). Receptor dimer spacing preferences are determined by several domains within the C region which have protein dimerization interfaces (Kurokawa et al., 1993; Pearlmann et al., 1993; Predki et al., 1994; Zechel et al., 1994a,b). These contribute significantly to the stability of the entire DNA-bound complex but appear to be weak in the absence of DNA, as reported for the glucocorticoid receptor (Hard et al., 19901, and they alone are not sufficient for dimerization in solution. The specific protein-protein contacts made between the two DNAbound receptors appear to be partly dependent on the spacing of the half-sites (See Fig. 2). In DR5 elements bound by an RXR/RAR heterodimer, the tip of the CI finger from the RAR contacts the D box of the CII finger from the RXR (Pearlmann et al., 1993; Zechel et al., 1994a). In contrast, on a DR2 element, it is the T box of the RAR that interacts with another portion of the CII finger from the RXR, which does not involve D box contacts (Zechel et al., 1994b). Homo- or heterodimeric receptors bound to DR1 sites use similar protein-protein interfaces to those used on DR2 sites (Zechel et al., 1994b). In this manner the D box contained within the amino portion of the CII finger is thought to be important in providing an interface for heterodimeric binding on DR5 sites. The rest of the CII finger and a region immediately adjacent to it, termed the T box, provide contacts for homo- and heterotypic binding on DR1 and DRl/DR2 sites, respectively. Additionally, experiments from several laboratories have shown that the receptor providing the CII interface must be located upstream in a DR configuration (Kurokawa et al., 1993; Pearlmann et al., 1993; Predki et al., 1994; Zechel et al., 1994a,b).The reason for this polarity is a result of the position of the CII interface on receptor-bound DNA relative to the rest of the molecule in these response element arrangements. This interface is in close proximity to that of an adjacent interface on a downstream receptor. Additionally, the strong C-terminal dimerization motif (contained within domain E) of the RAR is thought
410
T.MICHAEL UNDERHILL et al.
A
FIG.2. Pictorial depiction of the contacts made between heterodimers of RXR/RAR on (A) DR5 and (B) DR2 DNA binding sites. 5'-3' polarity is shown as left-right, with the RXR being bound to the site on the left. The interreceptor contacts made between the receptors are different on a DR5 site vs a DR2 site. See text for details. Briefly, on a DR5 HRE, the 5' positioned RXR makes contacts through the CII finger including the D box with the CI finger of the downstream RAR. In contrast, on a DR2 HRE, contacts are formed between the CII region excluding the D box of the upstream RXR and the T box of the downstream RAR. The E domain dimerization interface in the RAR is flexible and is able to .wive1 in order to associate with the RXR E region. T, T box; A, A box; Zn, zinc.
to rotate approximately 180" relative to the C domain. This would make this interface accessible for interaction with the dimerization domain of an upstream bound RXR (Kurokawa et al., 1993).Hence, the nature of these contacts in the C and E domains suggests a polarity in receptor binding, in which the RXR is generally positioned upstream. This has been confirmed by cross-linking studies (Kurokawa et al., 1993; Zechel et al., 1994b). Furthermore, this polarity is observed for RXR-TR DNA-bound complexes as well (Kurokawa et al., 1993;Zechel et al., 1994a,b). The one exception to this generalization is in experi-
RETINOIDS AND MOUSE EMBRYOS
411
ments conducted using a DR1 from the rat cytoplasmic retinol-binding protein I1 (CRBPII) promoter (Kurokawa et al., 1994). This response element is strongly activated by an RXRa homodimer (Zhang et al., 1992a) and inhibited by an RXR/RAR heterodimer (Mangelsdorf et al., 1991; Kliewer et al., 1992b; Underhill et al., 1994). In this case, using a combination of electrophoretic mobility shift assays and cross-linking analyses, it appeared that the RAR bound to the upstream half-site, as had been previously proposed by Zechel et al. (1994b). Most, if not all, of the DNA binding specificity of the receptors appears t o reside in the C domain (Kurokawa et al., 1993; Mader et al., 1993a; Pearlmann et al., 1993; Glass, 1994; Zechel et al., 1994a,b).The flexibility and position of the various C domain dimerization interfaces enable the receptors to bind cooperatively to a number of DR sites. In addition, receptor polarity exists on DRs with spacing of over 1. RXR/RAR heterodimers also bind and activate half-sites arranged as palindromes and inverted repeats, and RXR/RAR and RAR/TR heterodimers bind sites comprising everted repeats (Tini et al., 1993).The nature of the interactions used to establish these complexes may be different from those discussed above and may rely more heavily on the E domain dimerization domain (Zechel et al., 1994a). 3. DEF Domains: Activities Involved in Transactivation, Ligand Binding, and Receptor Dimerization The DEF domains contain many functions important t o receptor action, including a nuclear localization signal, ligand binding domain, protein dimerization region, and ligand-dependent transcriptional activation domain (AF2). The nuclear localization sequence appears to be part of the D domain as amino truncations that extend up to the D domain (DEF), and carboxyl truncations that include only the ABCD region are properly targeted to the nucleus (our unpublished observations, 1991; Espeseth et al., 1989; Pearlmann et al., 1993; Minucci et al., 1994). Hence, it appears that the D domain is at least minimally required for nuclear localization. The RXRs/RARs, unlike some members of the nuclear steroid hormone receptor family, appear to be in the nucleus in the absence of ligand (Dalman et al., 1991). Studies from our laboratory and others using constitutively activated receptors have shown that these chimeric receptors function to activate transcription from reporter constructs containing RAREs in the absence of ligand (Lipkin et al., 1992; Nagpal et al., 1993; Underhill et al., 1994). Furthermore, in the absence of ligand, P-galactosidase activity was found predominantly in the nuclei of cells that expressed a truncated RAR (ABCD) with a
412
T.MICHAEL UNDERHILL et a2
C-terminal fusion to the Escherichia coli P-galactosidase (Espeseth et al., 1989). The ligand-dependent transactivation activity (AF2) and proteinprotein dimerization region have been identified within the E domain (approximately 270 amino acids). As discussed later, RARs are able to utilize a number of retinoid ligands for transactivation, while RXRs are limited primarily to an isomer of all-trans and 9 4 s RAs (Allenby et al., 1993; Hall et al., 1993). AF2 is thought t o be the predominant transactivation domain and is located between amino acids 137 and 410 of RARP2 (Folkers et al., 1993). C-terminal deletions or point mutations in this region that retain dimerization ability abolish most if not all detectable transactivation activity in mammalian cells (Durand et al., 1992; Folkers et al., 1993; Nagpal et al., 1993; Tate et al., 1994; Zhang et al., 1994; Tairis et al., 1994).Furthermore, C-terminal truncations of RARa to amino acid 404 abolished transactivation activity and the ability to bind 9 4 s RA with high affinity; however, this receptor retained the ability to bind all-trans RA with high affinity (Tate et al., 1994). This suggested that the RARa ligand binding domains for all-trans and 9 4 s RAs were physically distinguishable and that ligand binding does not automatically confer transactivation activity on the receptor. Studies performed in yeast, however, showed that an RXR/RAR heterodimer could stimulate transcription from a DR5 RARE reporter construct in the absence of ligand (Heery et al., 1993), albeit to levels lower than that of the ligand-induced receptors. This level of constitutive activity of the native receptors has not been observed in mammalian cells and may reflect differences in the basal transcriptional apparatus between yeast and mammalian cells or the absence of specific competing or inhibitory pathways in yeast. Although ligand is required for transactivation, it is not, in most cases, necessary for dimerization (Nagpal et al., 1993). It is thought that the E domain provides the major protein-protein interface for dimerization of the RARs and RXRs in solution (Glass et al., 1990; Marks et al., 1992). Receptors first dimerize through contacts made between their respective E domains and are further stabilized by interprotein interactions made within the C domain and contacts with DNA (Glass, 1994). Several pieces of evidence from the analysis of dimerization in uitro and in uiuo have demonstrated the significance of the E domain in this important and requisite step in receptor function. Studies performed with truncated receptors containing only the DEF domain or parts thereof have found this region to be sufficient for efficient dimerization (Forman et al., 1988, 1989; Kliewer et al., 1992b; Nagaya and Jameson, 1993; Shen et al., 1993).In uiuo, our laboratory
RETINOIDS AND MOUSE EMBRYOS
413
and others have found that the expression of such DEF receptors function as very potent dominant-negative repressors through interference of normal receptor dimerization (Linney, 1992; Minucci et al., 1994). Furthermore, Zhang et al. (1994) have shown that separate but possibly overlapping regions within the E domain govern the homo- and heterodimerization of RXRs. Nine heptad repeats have been identified within the E region, which are highly conserved within this subfamily of nuclear hormone receptors (Forman et ctl., 1989; Forman and Samuels, 1990). The heptad repeats have been shown to be necessary for efficient dimerization between receptors and may contain amino acids important to these interactions (Forman et al., 1988,1989; Forman and Samuels, 1990; Au-Fliegner et al., 1993). THATINFLUENCE RETINOIDRECEPTOR ACTIVITY C. FACTORS 1. Ligand Availability As stated earlier, the retinoid receptors require ligand for transactivation in mammalian cells. The RARs bind all-trans and 9-cis RAs, with the exception of RARy (Allenby et al., 1994), with binding constants in the subnanomolar and 10-nM range, respectively (Allenby et al., 1993). Furthermore, the RARs were able to activate transcription when bound t o all-trans retinol (Repa et al., 1993) and a variety of other retinoids (Hall et al., 1993).In contrast, the RXRs appear to bind with high affinity and seem to be activated only by 9-cis RA, with a dissociation constant in the 1-t o 15-nM range (Heyman et al., 1992; Levin et al., 1992; Allegretto et al., 1993; Allenby et al., 1993). Ligand binding does not appear to be a prerequisite for binding DNA, although it may modify the stability of receptor-DNA-bound complexes (as discussed below; see Section II,C,2). Recently, it has been suggested that unliganded receptor bound to DNA is associated with a cellular inhibitory factor that prevents receptor-mediated transactivation (Casanova et ul., 1994).Addition of ligand relieves this repression by dissociating and/or inactivating this receptor complex. Ligand binding may also increase protein stability of the receptors, as has been shown for vitamin D-bound VDR (Santiso-Mere et al., 1993). These ligandassociated changes in receptor activity may result, in part, from receptor conformation changes that are induced by ligand binding, as has been shown for TRs (Toney et al., 1993) and RARs (Keidel et al., 1994). Furthermore, this may explain how ligand may stabilize and promote receptor-dependent transactivation. In some instances the presence of specific ligands has been found to
414
T. MICHAEL UNDERHILL et al.
stabilize some receptor pair combinations while destabilizing others (Cheskis and Freedman, 1994). This has been best shown for heterodimers of VDR or TR with an RXR, and for their respective homodimers, with the exception of TR (MacDonald et al., 1993; Cheskis and Freedman, 1994). As well, addition of 9-cis RA has been shown t o stimulate RXR homodimer formation (Zhang et al., 1992a; Medin et al., 19941, thereby reducing the level of RXR available t o participate in heterodimerization (Lehmann et al., 1993). Physiological ligand concentrations of approximately 10 nM do not correspond well with that required i n uitro to obtain significant levels of transactivation in transient transfections and/or de nouo gene induction. This could be due to a number of factors, one being that promoters usually contain a multitude of enhancer sites, such that high levels of activation are achieved only through the coordinated action of many distinct enhancer and transcription factors. Second, the level of retinoid receptor activity on any given response element varies. In some cases this activity may be low and difficult to detect, but still biologically relevant. Third, this raises the question of whether there may be so-called “hot spots for retinoid synthesis” i n uiuo (McCaffery and Drager, 1994). Presumably, the concentration of retinoids, such as RA, would be much higher in these regions, allowing the specific activation of receptors in a localized fashion. There is some evidence to support the existence of such sites in the embryo (Wagner et al., 1990; Chen et al., 1992) and that these sites may provide a means for cellular regionalization (Wagner et al., 1992; Colbert et al., 1993; LaMantia et al., 1993). Furthermore, small changes in the levels of ligand can have profound outcomes on development, suggesting that some regions of the embryo are exquisitively sensitive to the levels of RA and possibly small changes in retinoid receptor-mediated gene transcription. One example is the development of the chick wing, where addition of all-trans RA in concentrations of 10-fold above or higher than that normally present can cause mirror-image duplication of the digits of the limb (Summerbell, 1983; Tickle et al., 1982, 1989). Although the role of the receptors has primarily been thought of as one of a ligand-dependent transactivation factor, this should not imply that the receptors do not have a function in the absence of ligand. Because the RARs efficiently heterodimerize with the RXRs, an abundance of unliganded RARs could conceivably preclude RXRs from participating in other receptor-mediated pathways. Furthermore, unliganded TR has been shown to be actively involved in transcriptional activation and repression. Saatcioglu et al. (1993) identified a TR response element within the long terminal repeat of the b u s sarccoma
RETINOIDS AND MOUSE EMBRYOS
415
virus that was activated by TR in the absence of thyroid hormone (T3). In contrast, in the absence of ligand, many genes that contain TR response elements are inhibited directly by TR, partly through the inhibition of preinitiation complex formation (Fondell et al., 1993). Retinoid receptor-DNA-bound complexes could potentially function in a similar manner. Overall, the different affinities of the receptors for retinoids in combination with the distinct functions of ligand provide a mechanism for the selective activation of distinct retinoid pathways. 2. Receptor Dimerization The retinoid receptors, as mentioned previously, function primarily as dimers. The RARs function as a heterodimer in conjunction with the RXRs and to a lesser extent with TR (Glass et al., 1989; Tini et al., 1994). In contrast, the RXRs can function both as heterodimers and homodimers. These differences in dimerization ability have been attributed to differences in the C/D region (Lee et al., 1993) and are also dependent on the type of DNA binding site. RXRs, in addition to being a dimerization partner for the RARs, are able to dimerize with a number of members of the nuclear hormone receptor family, including VDR (Yu et al., 1991; Kliewer et al., 199213; Carlberg et al., 1993; MacDonald et al., 19931,TR (Yu et al., 1991; Bugge et al., 1992; Kliewer et al., 1992b; Leid et al., 1992b; Marks et al., 1992; Zhang et al., 1992131, COUP-TFI and -11 (Cooney et al., 1992; Kliewer et al., 1992a; Tran et al., 19921, peroxisome proliferator-activated receptor (PPAR) (Kliewer et al., 1992~; Gearing et al., 1993),other orphan receptors (Apfel et al., 19941, and yet unidentified proteins from different cell lines (Berrodin et al., 1992; Hallenbeck et al., 1992; Marks et al., 1992). Initial reports examining the binding efficiency of VDR, TR, and RARs to their cognate response elements using electrophoretic mobility shift assay demonstrated that addition of nuclear extracts from a number of different cell lines greatly facilitated receptor-dependent binding (Glass et al., 1990; Zhang et al., 1991; Ross et al., 1992a,b). Subsequently, this factor was identified by a number of groups and was found in many cases to be an RXR, usually a or p. In this manner RXRp was identified simultaneously by two groups as the cellular factor that greatly enhanced RAR binding to the RARE (DR5) of the RARp2 promoter (Yu et al., 1991; Leid et al., 1992b). It should also be noted that RXRs not only serve as dimerization partners but also participate directly in the ligand-dependent transactivation of target genes (Davis et al., 1994). RXRs have therefore been considered to be “promiscuous” partners (Bugge et al., 1992). RARs, on the other band, appear to function pri-
416
T.MICHAEL UNDERHILL et al.
marily through RXRs, suggesting that RAR transcriptional activity is dependent, to a great extent, on association with RXRs. It appears that only RXRs are capable of forming homodimers or at least functionally productive homodimers. The possibility that RARs can function as homodimers is unlikely. They do not efficiently form homodimers in solution (Glass et al., 1990; Kurokawa et al., 19931, nor do they contain a n identified homodimerization domain in their C domain (Mader et al., 1993a). As mentioned previously, on most DRtype response elements examined to date, the RARs required a n RXR partner for efficient binding and transactivation of RARE reporters, in both mammalian cells and yeast (Yu et al., 1991; Kliewer et al., 199213; Leid et al., 1992c; Marks et al., 1992; Zhang et al., 1992b; Allegretto et al., 1993; Hall et al., 1993; Heery et al., 1993, 1994). In yeast, however, RARs also stimulated transcription from a DR5 RARE to reasonable levels in the absence of an RXR partner (Allegretto et al., 1993; Hall et al., 1993; Heery et al., 1993), albeit high levels of RAR expression were required (Heery et al., 1993). Additionally, in some types of RA-responsive promyelocyctic (PML) leukemias, a translocation has occurred between the N terminus of the RARa gene and the PML gene, giving rise to a fusion receptor with altered properties. In this instance the addition of PML to RARa stimulated homodimerization of the RARs and altered their binding to DR elements (Perez et al., 1993). Hence, although the RARs do not appear to homodimerize efficiently in uitro, this has yet to be rigorously tested in uiuo and may be dependent, in part, on receptor expression levels and the type of RARE. Dimerization propensity and partner selection appear to be governed, in part, by ligand, especially for the RXRs. Addition of 9 4 s RA has been shown in uitro to enhance RXRa homodimerization and, to a certain extent, RXRp homodimerization as well (Zhang et al., 1992a; Medin et al., 1994). However, our own studies using ligand-independent receptors suggest that 9-cis RA is not absolutely required for the efficient activation of DR1 sites (Underhill et al., 1994). Recently, two additional studies examining RXR dimerization have found no evidence to support stimulation of RXR homodimerization by 9-cis RA. First, the addition of 9 4 s RA to a bacterially expressed RXRa DEF peptide did not stimulate dimerization (Cheng et al., 1994). Second, Chen et al. (1994) showed that the addition of 9-cis RA to RXRa expressed and purified from bacteria, Sf9 insect cells, or COS cells did not enhance homodimerization. The authors of the latter report suggest that the previous observations of 9-cis stimulation of RXR homodimerization may actually be a consequence of conformational
RETINOIDS AND MOUSE EMBRYOS
417
changes induced by receptor binding to 9-cis RA, which then stabilized the receptor. In yeast, activation of the apoAl or a DR1 response element by RXR homodimers occurred only in the presence of 9-cis RA; whether this is a consequence of the requirement for 9 4 s RA for transactivation or stimulation of RXR homodimerization is not known (Heery et al., 1994; Mak et al., 1994). Studies in yeast have complemented the results previously obtained in mammalian cells, supporting the important roles that RXRs play as a heterodimeric partner and possibly as a homodimer. As has become evident from the study of newly isolated response elements for other members of this steroid receptor subfamily, DNA binding is, t o a great extent, contingent on the composition of the dimer pair. Much of our current understanding of receptor-DNA complex formation has come from the analysis of a few select RAREs. Hence, the ability of RARs to function in other ways may await the discovery of different types of response elements. Finally, dimerization is important in that it provides a method by which numerous biological processes can be controlled efficiently through the combination of a minimal number of regulatory molecules (Leid et al., 1992a; Marks et al., 1992). 3. Inhibitors of Receptor Function
The retinoid receptors function as ligand-dependent transactivation factors. In order for the receptors to fulfill this role, they must presumably have access t o their cognate DNA binding site, which depends, in part, on the presence of the appropriate pai-tner and DNA binding site accessibility. Furthermore, for transactivation a suitable ligand needs to be available and bound, and finally, the receptor-DNA complex needs to communicate with the basal transcription apparatus either directly or indirectly. Hence, in this limited sense, receptor function could potentially be inhibited at any one of these steps. Evidence presently exists to support inhibition at most of these levels. As previously mentioned, the receptors may associate with a cellular inhibitory factor in the absence of ligand, which prevents communication with the basal transcriptional machinery (Casanova et al., 1994). The COUP-TFI and -11 gene products have been shown to interfere with retinoid signaling at a number of these steps (Cooney et al., 1992, 1993; Kliewer et al., 1992a; “ran et al., 1992). These factors compete directly for the RXR/RAR and RXR/RXR binding sites and have been found to bind a variety of DR response elements. Second, the COUPTFs are capable of forming heterodimers with RXRs, thereby limiting the free concentration of RXR available to participate in dimerization.
418
T. MICHAEL UNDERHILL et al.
Third, binding of COUP-TFs to retinoid response elements appears to silence the transcriptional machinery, causing a significant depression in the level of transcription. An inhibitory mechanism that involves receptor interaction with calreticulin has recently been described (Burns et al., 1994a). Ectopic expression of calreticulin in a neuronal cell line inhibited RA-induced differentiation (Dedhar et al., 1994). Calreticulin has been shown to directly interact i n uitro with various members of the steroid receptor superfamily (Burns et al., 1994b), including the RARs (Dedhar et al., 1994). Calreticulin is a 46-kDa Caz+-binding/storageprotein normally found in the endoplasmic reticulum. Calreticulin apparently binds to the receptors through a highly conserved KXFF[K/RlR peptide sequence contained within their DNA binding domains (Burns et al., 1994a). This peptide is similar to that of a peptide KLGFFKR contained within the cytoplasmic domain of integrin a-subunits, to which calreticulin binds (Burns et al., 1994a). Although overexpression of calreticulin appears to interfere with RA-mediated differentiation, and calreticulin interacts with the receptors in uitro, the exact way in which calreticulin inhibits receptor function in uiuo is unclear (Burns et al., 1994a). This stems from the localization of calreticulin within the endoplasmic reticulum of the cell. In order for calreticulin to interact with the receptors, it would need to be present in either the nucleus or the cytoplasm (Burns et al., 1994a). As a whole, these results suggest that retinoid signaling through RARs and RXRs can be intercepted at many different levels to selectively inhibit their activity. 4. Factors That Promote Retinoid Receptor-Mediated Transactivation Tissue culture cell lines exhibit a range of RA responsiveness to introduced reporter constructs containing RAREs. It has been previously shown that responsiveness is cell type dependent, with some cells, such as rat pituitary GH, cells, showing little or no activation of a pRARE (DR5) reporter construct in the presence of ligand and exogenous RARs (Davis and Lazar, 1993). We have found that a reporter plasmid containing approximately 1 kb of the rCRBPII promoter is very poorly activated in the presence of a cotransfected expression plasmid for RXRa in CV-1 and Chinese hamster ovary (CHO) cells, but is highly active in P19 EC cells (Underhill et al., 1994). Similarly, several other reporter plasmids constructed with either a fragment of the RARPZ promoter or part of the oxytocin promoter (T. M. Underhill, D. E. Cash, and E. Linney, unpublished observations, 1994) were examined and found to be poorly induced in CV-1 and CHO cells relative to that
RETINOIDS AND MOUSE EMBRYOS
419
observed in P19 EC cells (Underhill et al., 1994).We have been able to circumvent these poor inductions, in part, by using modified retinoid receptors that contain a heterologous activation domain. Furthermore, although these modified receptors functioned in the absence of retinoids, the addition of retinoids greatly stimulated their overall activity in P19 EC cells, in comparison to CV-1 and CHO cells, which showed no further stimulation in the presence of ligand. These differences in stimulation have been attributed to the expression of cell-specific factors that function in conjunction with the retinoid receptors to augment transcription. An E1A-like activity has been identified in some cell lines (Imperiale et al., 1984; La Thangue and Rigby, 1987), particularly EC cells, which appeared t o promote the RA responsiveness of cells to some RARE reporter constructs (Berkenstam et al., 1992; Kruyt et al., 1993; Stunnenberg, 1993). These results should not be surprising, as retinoids have very diverse effects on cells, presumably through the differential activation of downstream target genes. Consequently, the differences observed in RA responsiveness in uitro and in uiuo may reflect intrinsic cell differences in various receptor activation-dependent parameters, including coactivator expression.
5. Post-translational Modifications of the Receptors The RARs are differentially phosphorylated (Rochette-Egly et al., 1991, 1992; Gaub et al., 19921, and in some cases this phosphorylation is coupled with the addition of ligand. Phosphorylation appears to be isoform specific, as RARp2 expressed in COS cells was not further phosphorylated upon addition of RA, while RARPl and -3 were phosphorylated (Rochette-Egly et a1., 1992). RARP2 contains phosphotyrosine residues, while RARal and RARyl do not; however, RARp2 was only weakly phosphorylated in uitro with a CAMP-dependent kinase. Phosphorylation of the RARs appears to occur mainly in the NH,-terminal domain A/B region, which is consistent with the differences observed in phosphorylation of the receptors of the same type (Rochette-Egly et al., 1991). Although the RXRs may be phosphorylated and contain potential phosphorylation sites, this has yet to be reported. What effect phosphorylation has on overall receptor activity is not known. Protein phosphorylation is a common mechanism used to modify protein function, and as such, has the potential to be important in modifying many aspects of RAR and RXR function. For example, the activity of other steroid hormone receptors, such as the estrogen receptor, has been shown to be modified by phosphorylation of the A/B domain (Ali et al., 1993). Recently, phosphorylation of a closely related
420
T.MICHAEL UNDERHILL et al.
receptor, TR, dramatically enhanced its ability to heterodimerize with an RXR partner in uitro and in uiuo (Bhat et al., 1994).One possibility is that isoform-specific phosphorylation sites provide a portal by which other intracellular signaling cascades can modulate retinoid signaling. Hence, t o thoroughly investigate how this region figures in receptor function may require the analysis of receptor activity in various cellular backgrounds.
6. Configuration and Sequence of the DNA Binding Site The retinoid receptors bind to DNA that contains cognate binding sites in a direct, palindromic, inverted, and everted configuration. There are two main components to this DNA binding selectivity. The sequence of the half-site is typically A/GGG/TTCA, although a number of permutations have been identified. In almost all cases, however, at least one of the sites corresponds to the canonical half-site. The sequence of the half-sites, in addition to their polarity, may dictate differences in receptor dimer affinity. The second component is that of spacing of the half-sites. This appears to be important in specifying the type of receptor pair that binds to a DNA site; this was originally simplified as the 3-4-5 rule (Umesono et al., 1991) and was more recently extended t o a 1-2-3-4-5 rule (Mangelsdorf et al., 1994) (Fig. 3). Although a number of different receptor dimers and binding sites appear to fit this rule, there are exceptions, with some sites being more complex. In addition, both the sequence of the half-sites and flanking sequences influence the affinity of receptor binding (Nagpal et al., 1992a; Mader et al., 1993b; Predki et al., 1994).Given that the receptor half-site is composed of 6 bp and is approximately 50-70% AT in composition, it would then theoretically be distributed throughout the mammalian genome at a frequency of one in every 1024 nucleotides. This would give rise to 1million potential half-sites in the mammalian genome. As the receptors function as dimers and require appropriately placed half-sites for specific binding, the number of complete binding sites is dramatically reduced to about 1000 potential sites. A putative RARE has been recently identified in association with a pseudogene (Rudert and Gronemeyer, 19931, so it is not inconceivable that many nonfunctional binding sites exist within the mammalian genome. Additionally, a survey of the GenBank DNA sequence database for DR1 and DR5 sites has identified several perfect canonical sites within coding sequences. Hence, factors such as surrounding DNA sequences, response element context, chromatin structure, and accessibility may be important determinants in receptor-DNA complex formation. Recently, several groups have demonstrated that RXR/RAR heterodimers bind with specific polarity to DR5 and DR2 sites, with the RXR
RETINOIDS AND MOUSE EMBRYOS
ELEMENT
42 1
TRANSACTIVATION
DR1
+
DR 1
-
DR 1
+
DR1
-
DR2
+
DR3
+
DR4
+
DR5
+
ER8
+
FIG.3. Summary of hormone response elements bound by RARs and RXRs. RXRs are considered promiscuous as illustrated by the diversity of sites and partners to which they can interact with to form stable receptorlDNA complexes. RXRs are able to function as homodimers on DR1 HREs. RARs form homodimers inefficiently and rely on a partner such as RXR for DNA binding to HREs. The 5'-3' polarity is shown as leftright. Receptor binding polarity is as shown, with the exception of RXR/COUP and RXRlRAR on a n ER8. The transcriptional consequence of the addition of the preferred ligand of the 3' located receptor is noted in the transactivation column. This usually means, all-trans RA and %cis RA for RARs and 9-cis RA far RXRs.
binding the 5'-most half-site in a RARE with 5'-3' polarity, while the RAR binds to the downstream (3') site (Kurokawa et aZ.,1993; Mader et aZ.,1993a; Zechel et al., 1994b).Furthermore, the RARs appear to have a hinge region contained within the D domain that allows the C terminus of the receptor to rotate or swivel 180" so that the ligand binding dimerization domains can interact (Kurokawa et aZ., 1993). Receptor binding polarity has possible functional consequences, especially with respect to the orientation of the half-sites and ligand availability. In a series of elegant experiments by Kurokawa et al. (1994) using ligands specific to RARs and RXRs, it was found that the RAR partner of an
422
T. MICHAEL UNDERHILL et
a2
RXR/RAR heterodimer bound to the upstream half-site of a DR1 element from the rCRBPII. This reversed polarity prevented an RXRspecific ligand from binding to the downstream RXR, thereby inhibiting RXR-mediated activation of this promoter element. Overall, retinoid receptor RARE selection is dependent on a number of criteria, some of which have already been discussed. Of particular importance is the DNA binding domain and the E region dimerization domain. RXR/RARs efficiently form heterodimers in solution through strong interactions in their respective E domains. In this manner, the ligand binding/dimerization domain would be responsible for bringing the receptors together and therefore would dictate which receptors would dimerize. However, it is the interaction of the specific dimerization domains within the C domains on DNA that ultimately restricts response element selection. Both the DNA sequence and polarity of the DRs are important determinants in the formation of a productive receptor-DNA complex. D. SUMMARY Taken together, these many observations illustrate the inordinate complexity of the retinoid signaling pathway in vertebrate cells. Nonetheless, our current understanding of how the retinoid receptors function t o regulate gene transcription can be distilled to a few basic requirements and considerations (Fig. 4). RARs require a partner such as an RXR for efficient binding to many of the presently identified RAREs. Therefore, RAR function will be limited, t o a great extent, to the presence of an appropriate heterodimerization partner. The presence of competing nuclear hormone receptor pathways and inhibitory networks (e.g., COUP-TFs) needs to be considered. Such pathways have the potential to mute receptor-driven transactivation. Upon binding of the receptor complex to DNA, a suitable ligand is required to “activate” the complex. The nature of this activation can take many forms and is dependent on several parameters, including the specific ligand and its concentration; DNA sequence, spacing, context, and polarity of the half-sites; and the presence of coactivators and inhibitors. Collectively, all of these elements contribute to the spatiotemporal regulation of gene transcription by retinoids in uiuo. 111. RETINOID-BINDING PROTEINS An additional class of proteins that interact with retinoids includes both cellular binding proteins and proteins present in plasma. The
RETINOIDS AND MOUSE EMBRYOS
423
0-
-
mi I
A\-
CRBPl
FIG.4. Overview of the retinoid signaling pathway in mammalian cells. Ligand is made available to the receptors through complex metabolic pathways some of which are indicated and might involve one of all of the following molecules: CRABPI, CRABPII, CRBPI, to a lesser extent CRBPII, and possible FUR isomerases. With a net presence of the appropriate ligand, the receptors can stimulate or repress transcription from a variety of RAREs within target genes. Many factors control different aspects of this scenario some of which are shown and include: (i) the presence of other competing steroid hormone receptors and pathways (COUP); expression levels of retinoid receptors and isoforms thereof; phosphorylation state of receptors; DNA binding site accessibility; type, number, context, and polarity of response elements within promoter/enhancer regions; coactivator presence (CoA) and levels of inhibitory factors. BTA, Basal transcription apparatus.
CRBPs and CRABPs are low-molecular-weight molecules present in several tissues (for recent reviews see Ong et al., 1994;Soprano and Blaner, 1994).It was originally suggested that they may function as the steroid receptors in transducing signals to the nucleus (Ong and Chytil, 1978).However, after the discovery of the RARs (Giguere et al., 1987;Petkovich et al., 19871, it was clear that these molecules had different functions. Their specific functions are still argued, but there is a body of evidence that supports the possibility that they play different roles in the processing of retinoids. In this section we briefly review the binding proteins, and the reader is encouraged to examine
424
T.MICHAEL UNDERHILL et al.
recent reviews in this area (e.g., Giguere, 1994; Ong et al., 1994; Soprano and Blaner, 1994). OF RETINOID-BINDING PROTEINS A. A FAMILY
There are now at least two CRABPs and two CRBPs, which are the products of separate genes and are regulated in different ways. What was called CRABP is now CRABPI, with the newly identified CRABP has been named CRABPII. Similarly, the original CRBP is now CRBPI and the new gene product is named CRBPII. The CRABPs share considerable amino acid homology [78% amino acid sequence identity between human CRABPI and CRABPII (Astrom et d.,199111 as do the two CRBPs (56% amino acid sequence identity between rat CRBPI (Sundelin et al., 1985) and rat CRBPII (Li et al., 198611. In interpreting the past and present literature in this field, it is important to note these homologies, since sometimes authors still do not distinguish between the I and I1 forms of the binding proteins and some antisera generated against a particular binding protein may recognize the other member of its class because of amino acid homology. In addition to the cellular retinoid-binding proteins, there are extracellular binding proteins, whose specific roles during embryonic development have yet to be precisely determined. The plasma-binding protein is produced, to a large degree, in the adult liver; however, its mRNA is present in a variety of tissues (Soprano and Blaner, 19941, and it is also produced in the visceral yolk sac (Soprano et al., 1986) of embryonic rats. The epididymis RA-binding protein is a secreted androgen-dependent protein (Brooks et al., 1986; Ong and Chytil, 1988) that has been shown to bind all-trans and 9-cis RAs (Newcomer et aZ., 1993). The specific function of this protein has yet to be determined, although its ability to bind both all-trans and 9-cis RAs is intriguing. The expression of the cellular RA and retinol-binding proteins and their respective mRNAs has been examined by several laboratories (Ong, 1984; Crow and Ong, 1985; Li et al., 1986; Maden et al., 1988, 1990, 1991, 1992; Dolle et al., 1989, 1990; Perez-Castro et al., 1989; Dencker et al., 1990; Maden and Holder, 1991; Ruberte et al., 1991, 1992; Lyn and Giguere, 1994). A summary of this work can be found in recent reviews (e.g., Giguere, 1994; Linney and LaMantia, 1994; Ong et al., 1994). While the function of the CRABPs has yet to be accurately determined, their expression at developmental times and sites known to be particularly sensitive to RA as a teratogen, plus their distinct locations
RETINOIDS AND MOUSE EMBRYOS
425
in the developing nervous system, suggests that they may be involved in controlling the availability of RA to the RAR9. Given that they bind all-trans RA much more efficiently than 9-cis RA (Allenby et al., 19931, they may be selective “gatekeeperdYfor ligands to which the RARs, but not the RXRs, respond. Fiorella and Napoli (1991) have suggested that CRABPs may modulate RA levels by binding to RA and facilitating its catabolism. The possibility that this might have a biological effect on a differentiating system is suggested by studies (Boylan and Gudas, 1991) in which overexpression of CRABPI in EC cells, which can be induced to differentiate in the presence of RAYinhibited the RAinduced differentiation process. Since Espeseth et al. (1989) had shown that this process could be inhibited through overexpression of dominant-negative vectors directed toward the retinoid receptors, this CRABPI study suggests at least a contrived ability of CRABPI expression to inhibit a receptor-mediated event. Studies with the four cellular binding proteins suggest that three of the four genes are regulated by retinoids. While CRABPI is not RA inducible, CRABPII has been shown to be inducible with RA and to have sequences that resemble the RARE sequences through which activated RARs function (GiguBre et al., 1990; Astrom et al., 1991; Durand et al., 1992; MacGregor et al., 1992; Zhou et al., 1992). There is a suggestion that at least some of this RA inducibility might be at the level of mRNA stabilization and not just transcriptional induction (MacGregor et al., 1992). The CRBPs have both been shown to have RAinducible promoters. The mouse and rat CRBPI promoters have a defined response element (Smith et al., 1991). The mouse and rat CRBPII promoters both have RA response sequences (Mangelsdorf et al., 1991; Nagpal et al., 19931, but they are distinctly different. The rat CRBPII promoter has four or five DRs, each separated by a single base pairthis sequence appears to be activated via homodimers of RXRs (Mangelsdorf et al., 1991). The mCRBPI1 promoter does not have this series of repeats and has a more complex arrangement of binding sites (Nagpal et al., 1993). The differences could be significant with respect to what retinoid receptor combinations might regulate the transcription of the CRBPII gene in the two species. It should be underscored that these regulation studies are based on cell culture transfection or cotransfection studies in which a promoter fragment is coupled to a reporter gene such as chloramphenicol acetyltransferase (CAT) or firefly luciferase and transfected into a cell line, and then the reporter activity is measured in cells either untreated or treated with specific retinoic acids. In some cases expression vectors for individual RAR or RXRs are cotransfected into the cells.
426
T.MICHAEL UNDERHILL et al
Therefore, the results obtained are the result of contrived but controllable circumstances. This is underscored because the studies by several laboratories that have localized binding proteins or binding protein mRNAs during development do not necessarily result in colocalization of the RA-responsive binding protein gene products with known sources of RA in the embryo. B. TRANSGENIC STUDIES WITH RETINOID-BINDING PROTEIN GENES Transgenic studies have included the construction of transgenic lines that have the mouse CRABPI promoter driving the lac2 reporter gene (Wei et al., 1991), the ectopic expression of the CRABP I gene in the lens (Perez-Castro et al., 1993),the CRABPII promoter driving the lac2 reporter gene (V. Giguere, personal communication), the homologous recombination knockout of the CRABPI gene (Gorry et al., 1994) and that of the CRABPII gene (V. GiguBre, personal communication, 1994). Wei et al. (1991) used a 3233-bp promoter region to produce transgenic lines that would display ZacZ, presumably where the promoter functions in the embryo. They produced several independent transgenic lines and the expression patterns presented coincided, to a large degree, with independent studies on the localization of CRABPI in embryos. The same laboratory directed ectopic expression of CRABPI to the lens via the mouse a A crystallin promoter (Perez-Castro et al., 1993). They detected two different phenotypes in the transgenic mice: defective lens fiber differentiation and pancreatic tumorigenesis. The reason for the pancreatic tumors was not resolved, although the transgenic RNA was expressed in the pancreatic tumors. Gorry et al. (1994) mutated the CRABPI gene in mice through homologous recombination knockout of the gene via embryonic stem cells. They did not identify a phenotype in mice that had both copies of the gene mutated, resulting in their conclusion that CRABPI is a dispensable gene. When they challenged day 8.5 homozygous knockout embryos with a low dose of RA and compared these embryos at day 18.5 with normal treated embryos, they could detect no difference in observed phenotypes, suggesting that CRABPI does not have a unique role in teratogenic resistance or sensitivity at this stage of embryonic development. V. GiguBre and collaborators (personal communication, 1994) have constructed transgenic lines of mice that have the CRABPII promoter driving the lacZ reporter gene. One of the distinct regions of expression of the reporter gene is in a specific posterior domain during limb
RETINOIDS AND MOUSE EMBRYOS
427
bud development. When they knocked out the CRABPII gene in embryonic stem cells and produced mice with homozygous mutations in the gene, they observed a postaxial polydactyly in the forelimbs of the homozygous mutant. An additional postaxial digit is generally limited to a single forepaw of an individual animal. Lack of expression of CRABPII in the mutant forepaw might affect the retinoid signaling (perhaps increasing the local RA concentration), consistent with the possible role of CRABPs in modulating the concentrations of RA. To summarize this section, the critical findings that homologous knockouts of the CRABPI or CRABPII or both genes (Lampron et at., 1995)do not produce remarkable phenotypes in homozygous form directs one to the question of ultimate function of these two binding proteins. Furthermore, animals deficient in either binding protein exhibit no increased sensitivity to RA teratogenesis (Lampron et al., 1995). Lampron et al. (1995)suggested that the CRABPs may be important under conditions of malnutrition and hypovitaminosis-A whereby these intracellular binding proteins may facilitate the critical accumulation of retinoids in certain cells and tissues. Supportive information on CRBPI and CRBPII function await similar gene knockout studies.
IV. RETINOIDRECEPTOR EXPRESSION PATTERNS AND GENETIC ANALYSISOF FUNCTION As indicated by the extensive body of data, retinoids are likely to play a very crucial and dynamic role throughout vertebrate development. It has been well documented that extensive RA administration (as reviewed by Webster et al., 1986;Sulik et al., 1988;Gudas, 1994; Hofmann and Eichele, 1994;Linney and LaMantia, 1994;Nau et al., 1994)as well as vitamin A deprivation (Warkany and Schraffenberger, 1946;Wilson and Warkany, 1948,1949;Wilson et al., 1953)throughout development will produce a broad range of defects in a number of animal models, including the mouse. The pattern of effects varies considerably as a function of exposure time, clearly indicating differential temporospatial sensitivities among the different embryonic tissues. The effects of experimental retinoid manipulation become phenotypically more specific as development ensues and cells and tissues acquire a more differentiated state. Given the more generalized organization of the tissues in early conceptuses, experimental alterations of retinoid levels often produce more severe patterns of anomalies
428
T. MICHAEL UNDERHILL et al.
(Shenefelt, 1972), affecting progenitor cell types for many of the developing organ systems. In addition, these early cell populations apparently play a critical role in pattern formation (Rutledge et al., 1994). This section deals with the mouse model throughout these early stages of embryonic development. The dysmorphogenesis associated with experimentally altered retinoid availability to the embryo is often accompanied by changes in the domains of expression of developmental control genes, and the teratogenic effects of retinoid excess or deficiency are likely to be the result of anomalous activation/inactivation of RA-responsive genes by retinoid receptors in an altered temporospatial sequence (Giguere, 1994). The discovery of specific receptors for retinoids has greatly aided the process of trying to dissect apart the components of the retinoid signaling pathway. The three RARs, and RXRs, are organized at the 5' end of the gene with two major promoters, resulting in multiple isoforms of each gene with unique N-terminal protein sequences. Adjacent 3' sequences share significant homology within receptor subgroups (RAR and RXR) and encode the bulk of the receptor protein (Leid et al., 1992a).That each isoform is independently regulated by its unique promoter and that the N-terminal sequences of the isoforms have been highly conserved in evolution has led to the suggestion that each RAR and RXR isoform plays a unique role as a regulator of embryonic development and homeostasis (as reviewed by Mendelsohn et al., 1992; Chambon, 1994). A number of methods have been used in an attempt to dissect apart the retinoid signaling pathway, including transgenic approaches to determine where activated RARs exist during embryonic development (Rossant et al., 1991; Balkan et al., 1992). In this work a vector consisting of a PRARE (DR5) upstream of a minimal promoter coupled to a bacterial P-galactosidase gene was introduced into transgenic mice. Reporter gene expression represents the presence of appropriate activated retinoid receptors to the transgene. Expression was found in the closed portion of the neural tube, eye, frontonasal region, and limb bud in gestational day 8-10 embryos. Due to advances in in sztu hybridization techniques, systematic studies of receptor expression have been done and collectively support the idea of differential roles for each of the receptor subtypes throughout embryogenesis (Doll6 et al., 1989, 1990; Noji et al., 1989; OsumiYamashita et al., 1990; Ruberte et al., 1990; Mangelsdorf et al., 1992; Mendelsohn et al., 1992). While several of the RAR receptor subtypes show unique patterns of expression, others produce overlapping patterns of expression with the RXRs and retinoid-binding proteins CRBP and CRABP (Dolle et al., 1989, 1990; Ruberte et al., 1991).
RETINOIDS AND MOUSE EMBRYOS
429
A. RARs In early embryos RARa is ubiquitous, but shows a slight preference for mesodermal tissues (reviewed by GiguBre, 1994; Linney and LaMantia, 1994). RARp and RARy, on the other hand, show a more restrictive, and in many cases mutually exclusive, pattern of expression (Mangelsdorf et al., 1994). In neurulation stage embryos, for example, RARP transcripts are present in only fused regions of the neural tube, while RARy is present in the open unfused portions of the neural tube (Dolle et al., 1990).A t slightly later stages of development, gestational days 9-10, RARp is localized to the proximal part of the limb bud. Starting on gestational day 12-12.5, it is present in the interdigital mesenchyme and is thought to be involved in the regulation of apparent regression of these tissues (our unpublished observations, 1994; Dolle et al., 1989). During these later stages of limb morphogenesis, RARp has a more generalized expression throughout the mesenchyme of the developing limb bud, but is strongly excluded from the precartilaginous condensations, a region where RARy is abundantly expressed (Doll6 et al., 1990). RARp is also found in the epithelium of the inner ear and in endoderm-derived tissues (epithelia of the lungs, intestines, and genital tract), suggesting that RARp may play a role in the differentiation of epithelium during development (Doll6 et al., 1989, 1990). In the oldest embryos analyzed, RARp is found in the kidneys, heart, and muscle (as reviewed by Linney and LaMantia, 1994). Retinoids have deleterious effects on neural development and abundant expression of RARP has been found in the developing central nervous system, (Durston et al., 1989; Maden et al., 1991; Smith and Eichele, 1991; Wagner et al., 1992; Ruberte et ~ l .1993). , These findings lend support to the idea that RARp may act as a transducer of retinoid signal during normal development. Its pattern of expression, however, is quite dynamic throughout embryogenesis. For example, in the youngest embryos examined, RARp is expressed in the population of cells known as neural crest cells (Doll6 et al., 1990; Ruberte et al., 1990, 1991). It is this cell population that gives rise to the peripheral neural populations of the head and neck and the structural tissues of the face and skull (bones, cartilage, and connective tissue) (Noden, 1988). It also plays a very important role in neurulation, in that it provides much of the mesenchyme necessary to raise the neural folds so that they can become apposed in the midline. Teratogenic data related to excess administration of RA during these early stages of embryogenesis have shown that the neural crest-derived tissues are severely
430
T.MICHAEL UNDERHILL et al.
dysmorphic (Sulik et al., 1987, 1988, 19891, and a high incidence of exencephaly results (Yasuda et al., 1987). The developmental expression pattern of the RARp2 isoform was analyzed using a murine RARP2 promoter linked to a (3-galactosidase gene in transgenic mice (Mendelsohn et al., 1991; Reynolds et al., 1991). Expression of the reporter gene was comparable to that reported in the above in situ hybridization analyses (Dolle et al., 1989, 1990) and supported its proposed involvement in neural tissue formation. The generation of null mutations in mice by targeted mutagenesis via homologous recombination in ES cells has been used in order t o provide definitive evidence for the embryonic requirement of specific receptor subtypes. Given its specific pattern of RARp expression, it would be expected that a rather severe pattern of malformations would result if RARp receptors were rendered nonfunctional. Surprisingly, however, it has been shown that no apparent phenotype is present in RARP2 null mutant mice, although a significant decrease in fetal viability was noted (Chambon, 1994; Lohnes et al., 1994; Mendelsohn et al., 1994b). When RARp2 null mutant mice are intercrossed with null mutant mice for either a or y (all isoforms or specific isoforms) to produce double null mutant mice, however, a wide range of defects resulted, some of which are identical to those produced in vitamin A-deficient offspring (reviewed by Chambon, 1994). These defects include eye malformations, glandular defects, axial skeletal defects, respiratory tract defects, cardiac anomalies, and urogenital malformations (Lohnes et al., 1994; Mendelsohn et al., 1994a,b) (Table I). The differential effects observed in the double versus single null mutants indicate that there may be a potential redundancy in function among the different receptor subtypes and their isoforms. Like RARP, RAR-y shows a very distinct pattern of expression throughout embryogenesis, and in most cases are exclusive relative to RARP. In the earliest embryos examined, RARy was expressed in all three germ layers (Ruberte et al., 1990; Linney and LaMantia, 1994) and in the presomitic caudal region (Lohnes et al., 1994) of gestational day 8 mouse embryos. On gestational day 9 it was localized t o the frontonasal mesenchyme; pharyngeal arches, which contribute to the definitive frontal and lateral bones and cartilage of the face and neck; and scleratomes, which form the vertebrae (Kastner et al., 1990; Osumi-Yamashita et al., 1990; Ruberte et al., 1990; Lohnes et al., 1994). In the early limb buds RARy appears to be ubiquitously expressed throughout the mesenchyme (Dolle et al., 1990). As development proceeds it becomes localized to the precartilaginous condensations. In the central nervous system RARy transcripts were also found in several of
ANOMALIES I N RXR RA Ral Viability Eye malformation Heart defect (VSD) Outflow tract (ETA) Aortic arch Facial defects Brain defects Skeletal defects Cranial Axial Limb defects Glandular defects Esophageal anom. alies Tracheal cartilages Diaphragm Urinary system Genital system Lung Thymus, thyroid, parathyroid
V O
RA Ru
RA Rfi2
N O
RA Ry
N
V O
O
RAR al/fi2
RAR
s12h
s l h
+
b
dfi2
+
AND/OR
TABLE I RAR SINGLE AND DOUBLENULLMUTANTS~
RA Ruli P2.j-
RAR ally
-
s12h
b
O
RAR
O
uliyi a2+'-
a/y
RAR P2i.1
RX Ra
RXR uiR ARy
E
s12h
E
E
+
+
t
RAR
s l h t
b
+
+
0
+
o
+
+
-
-
0 -
+
-
0 -
0
-
-
-
+
+
o
+
+
o
o
o
o
+
+
+
0
o
+
o
o
o
o
o
o
+
+
+
0 0
0 0
0 0
0 0
0
-
0
-
0 -
+ + +
+ + +
o
0 0
0 0
0
-
-
t
+
+
+
o
o
-
-
-
-
-
t
-
-
-
-
-
+
0
+
-
-
+
+
-
-
-
-
0
0
0
-
-
-
t
+
+
-
-
+b
0 t
+
+b
t
0
0
0
0
0
o
t
o
+
-
+
0 0
0 0
0 0
0 0
0 0
0 0
0
-
o
o
o
o
o
t
+
+
0
0
o
o
o
t
+
+
-
+
0
0
0
0
0
o
o
t
+
0
o
t
o
o
o
+
t
+ + + + +
0
o
0 0
o
+
0
tb
0
0
+
+b
+
+ +
0
-
-
-
-
0
-
-
0
-
-
+b
b
E
0
O
o
V 0
0
+
0
+
Ra+/-
+
t
o
E
+
RARa
+ +
+
0
E
+
RXR
a+/-/ u i R A
0
o
o
V
RXR
t
o
o
RXR u/R ARu
0
o
0
RARy
RXR aiRA Ry+/-
a+'-/
0
o
o +
RXR
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
aV, viable; N, neonatal lethal; E, embryonic lethal; - no information available. 0,no apparent effects observed; +, effectsobserved. The references used to compile this table include Lohnes et a/.(1994), Mendelsohn et a/.(19941, Chambon (19941, and Kastener e t a / . (1994). bMinor defectilow frequency of effect.
432
T.MICHAEL UNDERHILL et a1
the neural crest craniofacial domains (Linney and LaMantia, 1994; Lohnes et al., 1994; Mangelsdorf et al., 1994). RARy is also highly expressed in differentiating squamous keratinizing epithelium (Osumi-Yamashita et al., 1990; Ruberte et al., 1990; Mangelsdorf et al., 1992). As in the case of RARP2 null mutant mice, RARy2 null mutant mice exhibit no apparent phenotype. Mutations directed toward abolishing expression of all RARy isoforms produced mice with only minor or no apparent alterations (Lohnes et al., 1993).Postnatal viability, however, was low, indicating that perhaps more sensitive methods of characterizing the phenotype are needed. Sporadic malformations in ocular gland epithelia and metaplasia of the seminal vesicles and prostrate were reported, as well as atlas-axis transformations of the cervical vertebrae (Chambon, 1993, 1994; Lohnes et al., 1993, 1994). All mutants had one common defect related to incomplete formation of the tracheal cartilages. Teratogenic administration of RA to gestational females often produces lumbosacral truncations in the offspring (Kessel, 1992). It was found that RARy null mutant mice are resistant to this particular teratogen-induced defect (Chambon, 1994; Lohnes et al., 19931, and it is presumed from this observation that RARy may be involved as a mediator in teratogen-induced dysmorphogenesis. A unique role, however, in normal morphogenesis remains questionable, since RARy null mutants do not present lumbosacral defects. Double null mutants deficient in RARy in conjunction with either RARa (all isoforms or specific isoforms) or RARP2 resulted in offspring with a broad spectrum of effects and range of severity (Lohnes et al., 1994; Mendelsohn et al., 1994a). The most severe malformations were associated with offspring of RARy and RARa (all isoforms) double mutants (RARa-/-/RARy-/-). In addition to being embryolethal, this double mutation produced offspring with a high incidence of exencephaly (presumably nonclosure of the hindbrain) and facial anomalies, including midfacial clefts, deficient maxillae, and severe eye malformations. Accompanying malformations in the skeletal tissues of the skull and the face were noted as well as a number of brain malformations, including failure of rostra1 interhemispheric commissures to cross the midline, agenesis of the abducens (cranial VI nerve) motor nucleus, and irregular telencephalic folding. Some of these mutants had irregular flexures of the spine (scoliosis),defects in the cervical vertebrae, and incomplete development of the ocular and salivary glands. In addition, these mutants also had severe malformations in the respiratory, cardiac, and urogenital systems (Table I). Given the expression patterns of RARy as described above, the lack of effect in
RETINOIDS AND MOUSE EMBRYOS
433
the skin and the epidermis of these mutants is somewhat surprising. While the RAh-/-IRARy-I- double null mutant produced the most severely affected offspring, analysis of other less severely affected double null mutants from a RARy and either RARa (specific isoforms) or RARp2 have provided some data that support the specific role of RARy in embryogenesis, especially in eye morphogenesis. For the most part, however, the overlap in phenotype from these particular crosses makes subtracting defects among the different crosses in order to determine function an almost impossible task, and lends support to the idea of functional redundancy among the various receptors. With the generation of additional gene knockout models and transgenic reintroduction of receptor genes into these mutants, sorting out receptor specificity and or redundancy will become more plausible. Homozygous null mutants of either RARal and RARa (all isoforms) produced no observable phenotype (Li et al., 1993; Lufkin et al., 1993). RARtrl null mutant fetuses remained viable, in parallel with the wild type, but RARa (all isoforms) null mutant fetuses had increased neonatal lethality and showed some growth deficiency. While R A h transcripts are ubiquitous throughout the embryo, data allowing comparison between the RARa (all isoforms) double null mutants and the RARa isoform-specific mutants suggest that RARa isoforms may have a unique function in the development several organ systems. For example, as described above, the most severe eye defect mutants generated to date were found in the RARaIRARy double mutants. However, in the mutant offspring in which only the RARal and RARa2 (one copy) isoforms of RARa were functionally deleted (in conjunction with all isoforms of RARy), the eyes were nearly normal. It would seem from these data that only one copy of RARa2 is essential to recover most of the events involved in early oculogenesis. Identical results were observed in relation to the limb malformations produced in this same set of double null mutants.
B. RXRs
RXRs, like RARs, are widely expressed in adult and embryonic organisms. While the available data on RXR expression are somewhat more limited than for the RARs, it appears that the RXRs show a more generalized pattern of expression throughout organogenesis, although transient increases in regional expression have been noted in relation to RXRy. In the earliest examined embryos RXRr has been found specifically in the somites and in Rathke’s pouch (Mangelsdorf et al., 1992, 1994). It is this structure, which extends from the oral cavity to
434
T. MICHAEL UNDERHILL et a1
the diencephalon, that will form the definitive tissues of the anterior pituitary. Expression patterns in older embryos (through gestational day 15) show that RXRy is localized in the pituitary gland throughout this period of development (Mangelsdorf et al., 1992). In addition, in older specimens RXRy is also localized in the spinal cord, skeleton, and neck musculature. RXRP is found in nearly all tissues throughout embryogenesis, although its expression is somewhat increased in the anterior portions of the spinal cord and the hindbrain (Mangelsdorf et al., 1992). Preliminary data from our laboratory have also shown that on gestational day 8 RXRP is increased in the anterior (forebrain and midbrain) region of the embryo neural tube. Additionally, a slight increase in expression within the peripheral limb bud mesenchyme was observed on gestational day 10. It appears that RXRa is also abundant throughout the early embryos, but is predominant in the visceral tissues, such as the liver, intestines, and kidneys, as well as in the skin and the epidermis (Mangelsdorf et al., 1990, 1992). Single null mutants derived from mutation of the RXRa gene show growth impairment as well as severe congenital cardiac malformations, which presumably cause the homozygotes to die in midgestation (Kastner et al., 1994; Sucov et al., 1994).The defect involves the anomalous differentiation of ventricular cardiomyocytes, rather than of the neural crest, which aids in septating the outflow tract of the heart. The same defect has been described in association with vitamin A deficiencies exclusively (i.e., not other nutritional deficiencies). Taken together, these data suggest an essential role for RXRa in the normal signaling pathway in ventricular morphogenesis, and that the pathway compromised by this mutation is retinoid specific. As discussed in the previous section, RXRs can bind as homodimers, but in most cases bind considerably more efficiently as heterodimers in uitro. In addition, RXR enhances the binding of other nuclear receptors (TR, VDR, and PPAR) (Giguere, 1994; Kastner et al., 1994).In conjunction with the relatively nonspecific transcript distribution of RXRa and RXRP, it is thought that RXRs may be pleiotropic in their effects, acting as heterodimers for a number of nuclear receptors in many biological processes (Kastner et al., 1994). In an attempt to provide some in uiuo data supporting the idea that RXR/RAR heterodimers act as transcriptional transducers of the RA signal during development, R X R d R A R a and RXRaIRARy compound null mutants were produced and characterized (Kastner et al., 1994). The results from this work were compared to those discussed above in the single RXRa single null mutant studies, in order to determine the possible functional signifi-
RETINOIDS AND MOUSE EMBRYOS
435
cance of the RXR in early development. A number of tissues were affected in the RXR/RAR mutants and in some cases were nonoverlapping with those defects described for the RXRa single null mutant (Table I). In relation to some of the ocular defects described, a synergistic effect was proposed, in that similar malformations were observed for both the RXRa and RXR/RAR mutants, but were more severe in the latter. In conjunction with the observed phenotypic overlap between the RXRa and RAR double null mutants described above, it was concluded from this work that many of the observed defects were likely to occur as a result of impaired RXR/RAR heterodimer formation. While the null mutant experiments point out the functional redundancy in the RARs, to date RXRa is the only RXR reported t o be knocked out. Because of the possible intersection of RXRs with VDR, TR, and PPAR signaling pathways, it would not be surprising to find rather unique phenotypes associated with knockouts of other RXRs or, in particular, with double RXR null mutant mice.
C. SUMMARY The current understanding of the developmental, teratological, and molecular details of retinoids in mouse embryogenesis provides a solid framework for further experimentation. Over the years considerable advances have been made in the in vitro identification and characterization of many of the components of the retinoid signaling pathway. Current work directed at trying to determine the in vitro interactions between the various components has greatly added to our understanding of the system, but at the same time has emphasized its complexity. While several potential downstream targets have been identified, determining which of them plays a primary (versus secondary or tertiary) role in the sequence of retinoid-associated developmental events is unclear at this time. In uiuo expansion of this work is now under way, using the embryo as a model system. A vast body of information has been collected concerning the characterization of expression of the various binding proteins, RARs, and/or RXRs at various stages throughout development. While there is some overlap between stage-specific sites of expression of RARs and/or RXRs and teratogenesis, it is becoming increasingly clear, especially from the available knockout data, that no single member of the retinoid signaling pathway dictates morphological outcome. Phenotypic overlap among the various null mutants generated to date is more the rule, rather than the exception. While these data suggest considerable functional redundancy among the receptors, specificity of
436
T. MICHAEL UNDERHILL et al
receptor function was indicated relative to several of the malformations (i.e., the limb and the eye). With the generation of additional gene knockout models and transgenic reincorporation of receptor genes into these mutants, sorting out receptor specificity and/or redundancy will become more plausible.
V. CONCLUSIONS AND PERSPECTIVES The retinoid signaling pathway plays a variety of roles in vertebrate cell biology, the basis of which is not entirely well understood. This is further compounded at the tissue and organism levels. This apparent diversity in retinoid receptor function is not so much the result of changes in the intrinsic functions of .the receptors; rather, it is a mere reflection of the different cellular theaters in which they perform. To understand the myriad of effects retinoids have on cell behavior, one needs to examine the specific cellular context in which they operate. This will involve, in part, the isolation of cell type-specific retinoid target genes and an examination of their function and regulation by retinoids. A number of genes having diverse roles in cell biology have been found to be regulated by retinoids and, in some cases, contain RAREs. Many of the RAREs isolated to date have come from the regulatory regions of either the retinoid receptors themselves or retinoid-binding proteins. RARa2, RARP2 and -P4, RARy2, and RXRyl (de The et al., 1990; Hoffmann et al., 1990; Sucov et al., 1990; Leroy et al., 1991b; Lehmann et al., 1992; Q. Liu and E. Linney, unpublished observations, 1994) appear to regulated, in part, by the retinoid receptors, as do CRABPII, CRBPI, and CRBPII (Mangelsdorf et al., 1991; Smith et al., 1991; Durand et al., 1992; Husmann et al., 1992; Nakshatri and Chambon, 1994). As can be expected from these results, the addition of retinoids to tissue culture cell lines or to embryos in many instances dramatically modifies expression of the retinoid receptors and binding proteins. In this respect the dynamic nature of receptor expression has made it more difficult to understand the role individual receptors may have in various biological and developmental processes. In an attempt to minimize the effects the presence of ligand has on changing receptor profiles and activating different retinoid signaling pathways, we are using retinoid receptors that do not require ligand for activation. In this manner we are able to introduce individual activated receptors into biological systems and address the consequence of their activation. This approach is currently being used to understand the role the
RETINOIDS AND MOUSE EMBRYOS
437
retinoid receptors may have in cell differentiation and development of the vertebrate limb.
REGULATION OF IMPORTANT A. RETINOID DEVELOPMENT TARGET GENES 1. Hox Gene Regulation by Retinoids Several genes important to various aspects of development appear to be regulated by RA. These include several of the Hox genes (Krumlauf, 1993; Mavillo, 1993; Langston and Gudas, 1994). The Hox gene family consists of more than 40 members, all of which contain a highly conserved 183-bp DNA sequence that encodes for the homeodomain that binds DNA with sequence specificity (Scott et al., 1989; Gehring, 1992). The mouse genome contains at least four Hox gene clusters. Each cluster is composed of nine or more members tandemly arranged along the DNA. Several reports have shown that the addition of retinoids to EC cells induced Hox gene expression. Furthermore, this appeared to be receptor mediated, as retinoids fail to induce Hox gene expression in the retinoid-resistant P19 EC derivative RAC65 (Pratt et al., 1993) and in P19 EC cells expressing a dominant negative RXRP (Minucci et al., 1994). The RAC65 cell line carries a truncated RARa receptor that functions as a dominant negative receptor (Pratt et al., 1990; Kruyt et al., 1992). Additionally, Boylan et al. (1994) have reported on an F9 cell line that was genetically modified to be deficient in RARy function in which H o d 1 expression was not induced upon RA treatment (Boylan et al., 1993).Hox genes located at the 3' end of the cluster were typically the first to be up-regulated after exposure to RA (Boncinelli et al., 1991; Simeone et al., 1991; Papalopulu et al., 1991; Moroni et al., 1993). The rest of the cluster was found to be sequentially expressed and started with the 3' genes and extended 5'. The extent of this progression is dependent on both the concentration of RA and the length of exposure (Papalopulu et al., 1991; Simeone et al., 1991; Moroni et al., 1993). These results have been extended to the embryo (Conlon and Rossant, 1992; Marshall et al., 1992). Marshall et al. (1992) have shown that RA treatment changes the "Hox code.'' Furthermore, treatment of embryos with all-trans RA has been shown to lead to the inappropriate expression of several of the Hoxb genes, leading to an anteriorization of Hox gene expression within the neural tube (Conlon and Rossant, 1992). Some of the defects observed in hyper- and hypovitaminosis A are thought to arise from altered expression of Hox genes (Kessel, 199213). In some instances there appears to
438
T. MICHAEL UNDERHILL et al.
be a correlation between the ectopic expression of Hox genes and retinoid-induced dysmorphogenesis. RAREs have been identified in three members of the Hox gene family: HoxAl (Langston and Gudas, 19921, HoxBl (Marshall et al., 19941, and HoxD4 (Popper1 and Featherstone, 1993) (Fig. 51, all of which are located near or at the 3’ end of the cluster. The mouse HoxBl gene was found to contain a DR2 that was bound eficiently by an RXR/RAR heterodimer (Marshall et al., 1994).This element was found to be highly conserved among fish, avian, and mammal species. Promoter elements that contained this RARE sequence from the puffer fish, chicken, or mouse were coupled to E. coli p-galactosidase and used to make transgenic mice. In all cases p-galactosidase activity was present in the neuroectoderm. Importantly, addition of RA modified this pattern of expression and shifted it anteriorly. Deletion of this element nullified expression in the neuroectoderm. This was the first report to directly
HoxA
HoxB
HoxC
HoxD Transcription Responsiveness to Retinoic Acid Low
4
High
FIG.5 . Schematic representation of the four mouse Hox gene clusters. Indicated are RAREs which have been identified within the flanking sequences of Hox genes that are stimulated by RA. The Hox genes are transcribed in a left-right orientation as shown. Expression of the individual Hox genes in response to RA occurs in a 3 ’ 4 ’direction. Hox genes located a t the 3‘ end of the cluster are expressed shortly after RA addition, whereas more 5’ located genes are sequentially activated in a time- and dose-dependent fashion.
RETINOIDS AND MOUSE EMBRYOS
439
demonstrate the importance of retinoid signaling in regulating Hox genes during vertebrate development. 2. Retinoids and the Transforming Growth Factor p Family Several members of the transforming growth factor p (TGF-p) family are important in morphogenesis and appear to be regulated by RA (Sporn and Roberts, 1991). Treatment of murine embryos in utero at the early neural plate stage with RA caused a reduced expression of TGF-p1 and TGF-P2 protein, with no apparent effect on TGF-p3 expression (Mahmood et al., 1992). %o other members of this family, bone morphogenetic proteins (BMP) 2 and 4 were found to be upregulated and down-regulated, respectively, in F9 EC cells exposed to RA (Rogers et al., 1992). In C3HlOT1/2 mesenchymal cells BMP-2 and -4 expression were down-regulated upon treatment with RA (Gazit et al., 1993).In ouo studies performed in the chick limb bud showed that placement of an RA-soaked bead in the anterior margin of the bud induced ectopic expression of BMP-2, while not having an appreciable effect on BMP-4 (Francis et al., 1994). We have found that the expression pattern of BMP-4 is modified in the limbs of mice that overexpress RARa (Cash et al., 1995). Efforts are currently under way to establish a direct connection between retinoid signaling and regulation of members of this family of developmentally important extracellular signaling molecules. 3. RA as a Morphogen RA, as mentioned previously, is a potent teratogen. Many of the observed defects are dependent on both the dose of RA and the gestational stage of the embryo at the time of treatment. The nature of these defects suggests that RA may have morphogenetic properties. Further evidence for this was provided by studies performed by Tickle et al. (1982) and Summerbell (1983) that showed that RA alone could mimic the effects observed on chick wing development of a region thought to release a morphogen. More recently, however, two other groups, using different methods, have shown that RA most likely acts by inducing the expression of a morphogen, rather than by acting as a morphogen itself (Noji et aZ., 1991; Wanek et al., 1991). A gene has recently been identified whose ectopic expression can induce the spectrum of effects on wing development observed with the addition of RAsoaked beads (Riddle et al., 1993; Chang et al., 1994). This gene, sonic hedgehog (shh), so called because of its homology to the Drosophila hedgehog gene, was thought to be the long-sought limb morphogen which controlled anteroposterior (AP) patterning of vertebrate limbs.
440
T. MICHAEL UNDERHILL et a2
More recent results suggest that shh alone does not control AP patterning and that RA still appears to play an important role in limb development, albeit most likely not as a morphogen. This has been partly demonstrated by the studies of Helms et al. (19941, who have shown that ectopic induction of shh alone is not sufficient to replace the complete activity of RA. 4. Cell Adhesion Molecules and Retinoids Cell aggregation is an integral part of many developmental pathways. For example, in the mammalian limb mesenchymal cells coalesce to form condensations (Ede, 1983; Solursh, 1983). Condensation occurs through a process of selective cell aggregation, mediated, in part, by cellular adhesion. These condensations are the forebears of what will eventually become bone. The morphological cellular changes associated with this condensation step appear to be mediated, at least in part, by the cellular adhesion molecules (CAM) neural CAM (NCAM) and N-cadherin (Widelitz et al., 1993; Oberlender and Tuan, 1994). Both of these gene products are expressed during limb bud outgrowth in mesenchymal condensations. Antibodies to either of these proteins diminishes cell condensation (Widelitz et al., 1993; Oberlender and Tuan, 1994). Both retinoids and Hox proteins have been implicated in regulating CAM expression. Homeodomain binding sites have been found in several CAM gene promoters (Hirsch et al., 1990; Jones et al., 1992, 1993; Perides et al., 1994) and, as mentioned previously, Hox genes are regulated by retinoids. In addition, retinoids induced N-CAM expression in tissue culture cells (Husmann et al., 1989). N-cadherin mRNA was recently shown to be induced upon differentiation of P19 EC cells with RA (Jonk et al., 1994).Hence, retinoids could be directly involved in regulating CAMS,or indirectly through regulation of other genes, such as the Hox genes, to coordinate temporospatial cell aggregation.
B. RETINOID RECEPTOR-MEDIATED INHIBITION OF AP-1 The retinoid receptors appear to have additional transcriptional roles that do not require DNA-binding. Ligand-activated retinoid receptors appear t o also affect cell proliferation through repression of the AP-1 complex. AP-1 is a heterodimer of c-fos and cjun which is important in cell proliferation (Angel and Karin, 1991).In many instances addition of retinoids to tissue culture cells causes a slowing or cessation of proliferation (Sporn and Roberts, 1993).Recently, expression of the receptors has been found to interfere with AP-1 function. Interestingly, the re-
RETINOIDS AND MOUSE EMBRYOS
441
ceptors appear to mediate this repression through DNA-dependent and -independent mechanisms (Pfahl, 1993). Fanjul et al. (1994) have recently shown, using synthetic retinoid ligands, that the transcriptional function of the receptors is distinct from their ability to inhibit AP-1 function. Several synthetic retinoids were synthesized and tested for their ability to stimulate transcription from a reporter plasmid and inhibit AP-1 function. Interestingly, some of these synthetic ligands were ineficient at enhancing receptor-dependent transcription, but still functioned as potent AP-1 antagonists. Surprisingly, this receptor function, as has also been shown by others, requires the presence of the DNA binding domain, as receptors devoid of this domain are ineffective at mediating Ap-1 repression. It has been proposed that the receptors may inhibit AP-1 activity through heterodimerization with members of the AP-1 complex (Schule et al., 1991; Yang Yen et al., 1991). The importance of retinoid receptors in regulating cell proliferation through antagonism of AP-1 in uivo has not yet been determined.
C. FUTUREDIRECTIONS Significant progress has been made in the last several years in understanding the role of retinoids in mammals. We now have a reasonable comprehension at the molecular level of how ligand-activated retinoid receptors mediate gene transcription. We are slowly beginning to unravel the role of retinoids and their receptors in cell biology and their interplay with other cellular signaling pathways. Information on the action of the retinoid receptors in development has been specifically enhanced by the creation of null alleles for some of the receptors and the effects that their absence, either individually or in combination, have on various developmental processes. Nonetheless, many important unresolved questions remain at each of these levels, some of which include: Do the receptors and isoforms thereof have different response element preferences? What other response site configurations are bound by the receptors? What other factors contribute to the formation of a receptor-DNA complex? Are there high- and lowaffinity RAREs, and if so, how are they regulated? Do the receptors exhibit a dimerization partner preference, that is, a heterodimeric code? Does ligand influence this partnership? Do RARs function with a partner other than an RXR? How does phosphorylation and dephosphorylation affect receptor function? Why are there three receptors of each class, when, from knockout studies, we find that one or two may be sufficient? How important are coactivators to receptor activity, and if important, do they govern cell-specific RA responsiveness? Do
442
T.MICHAEL UNDERHILL et al.
the retinoid receptors play a role in preimplantation mammalian embryonic development? Do the RXRs function as homodimers in uiuo, and is this important? How is ligand availability controlled and restricted within the cell and the embryo? In summary, development, in the simplest terms, is the controlled acquisition of new cellular properties in a temporally and spatially restricted manner. Cellular behavior is the result of numerous independent and intersecting regulatory pathways, in which some gene products play a more important role. The retinoid receptors appear to be such proteins. REFERENCES Ali, S., Metzger, D., Bornert, J.-M., and Chambon, P. (1993).Modulation of transcriptional activation by ligand-dependent phosphorylation of the human oestrogen receptor AIB region. EMBO J . 12, 1153-1160. Allegretto, E. A,, McClurg, M. R., Lazarchik, S. B., Clemm, D. L., Kerner, S. A., Elgort, M. G., Boehm, M. F., White, S. K., Pike, J . W., and Heyman, R. A. (1993). Transactivation properties of retinoic acid and retinoid X receptors in mammalian cells and yeast. J . Biol. Chem. 268,26625-26633. Allenby, G., Bocquel, M.-T., Saunders, M., Kazmer, S., Speck, J., Rosenberger, M., Lovey, A,, Kastner, P., Grippo, J. F., Chambon, P., and Levin, A. A. (1993). Retinoic acid receptors and retinoid X receptors: Interactions with endogenous retinoic acids. Proc. Natl. Acad. Sci. U.S.A. 90, 30-34. Allenby, G., Janocha, R., Kazmer, S., Speck, J., Grippo, J. F., and Levin, A. A. (1994). Binding of 9-cis-retinoic acid and all-trans-retinoic acid to retinoic acid receptors a , p, and y. J . Biol. Chem. 269, 16689-16695. Angel, P., and Karin, M. (1991). The role of Jun, Fos, and the AP-1 complex in cellproliferation and transformation. Biochim. Biophys. Acta 1072, 129-157. Apfel, R., Benbrook, D., Lernhardt, E., Ortiz, M. A., Salbert, G., and Pfahl, M. (1994). A novel orphan receptor specific for a subset of thyroid hormone-responsive elements and its interaction with the retinoidithyroid hormone receptor subfamily. Mol. Cell. Biol. 14, 7025-7035. Astrom, A, Tavakkol, A,, Pettersson, U., Cromie, M., Elder, J. T., and Voorhees, J . J . (1991). Molecular cloning of two human cellular retinoic acid-binding proteins (CRABP). Retinoic acid-induced expression of CRABP-I1 but not CRABP-I in adult human skin in vivo and in skin fibroblasts in vitro. J . Biol. Chem. 266, 1766217666. Au-Fliegner, M., Helmer, E., Casanova, J., Raaka, B. M., and Samuels, H. H. (1993).The conserved ninth C-terminal heptad in thyroid hormone and retinoic acid receptors mediates diverse responses by affecting heterodimer but not homodimer formation. Mol. Cell. Biol. 13, 5725-5737. Balkan, W., Colbert, M., Bock, C., and Linney, E. (1992). Transgenic indicator mice for studying activated retinoic acid receptors during development. Proc. Natl. Acad. Sci. U.S.A. 89, 3347-3351. Berkenstam, A., del Mar Vivanco Ruiz, M., Barettino, D., Horikoshi, M., and Stunnenberg, H. G. (1992). Cooperativity in transactivation between retinoic acid receptor and TFIID requires a n activity analogous to E1A. Cell 69, 401-412. Berrodin, T. J., Marks, M. S., Ozato, K., Linney, E., and Lazar, M. A. (1992). Hetero-
RETINOIDS AND MOUSE EMBRYOS
443
dimerization among thyroid hormone receptor, retinoic acid receptor, retinoid X receptor, chicken ovalbumin upstream promoter transcription factor, and an endogenous liver protein. Mol. Endocrinol. 6, 1468-1478. Bhat, M. K., Ashizawa, K., and Cheng, S.-Y. (1994).Phosphorylation enhances the target gene sequence-dependent dimerization of thyroid hormone receptor with retinoid X receptor. Proc. Natl. Acad. Sci. U S A . 91, 7927-7931. Boncinelli, E., Simeone, A,, Acampora, D., and Mavilio, F. (1991). HOX gene activation by retinoic acid. Trends Genet. Sci. 7, 329-334. Boylan, J. F., and Gudas, L. J. (1991). Overexpression of the cellular retinoic acid binding protein-I (CRABP-I) results in a reduction in differentiation-specific gene expression in F9 teratocarcinoma cells. J. Cell Biol. 112, 965-979. Boylan, J. F., Lohnes, D., Taneja, R., Chambon, P., and Gudas, L. J. (1993). Loss of retinoic acid receptor g function in F9 cells by gene disruption results in aberrant Hoxa-1 expression and differentiation upon retinoic acid treatment. Proc. Natl. Acad. Sci. U S A . 90, 9601-9605. Brand, N. J., Petkovich, M., Krust, M., Chambon, P., de Th6, H., Marchio, A., Tiollais, P., and Dejean, A. (1988).Identification of a second human retinoic acid receptor. Nature 332, 850-853. Brooks, D. E., Means, A. R., Wright, E. J.,Singh, S. P., and Tiver, K. K. (1986).Molecular cloning of the cDNA for two mqjor androgen-dependent secretory proteins of 18.5 kilodaltons synthesized by the rat epididymis. J. Biol. Chem. 26,4956-4961. Bugge, T. H., Pohl, J., Lonnoy, O., and Stunnenberg, H. G. (1992). RXRa, a promiscuous partner of retinoic acid and thyroid hormone receptors. EMBO J. 11, 1409-1418. Burns, K., Atkinson, E. A,, Bleackley, R. C., and Michalak, M. (1994a). Calreticulin: From Ca2+ binding to control of gene expression. TZCB 4, 152-154. Burns, K., Duggan, B., Atkinson, E. A,, Famulski, K. S., Nemer, M., Bleackley, R. C., and Michalak, M. (1994b). Modulation of gene expression by calreticulin binding to the glucocorticoid receptor. Nature 367, 476-480. Carlberg, C., Bendik, I., Wyss, A., Meier, E., Stunenbecker, L. J., Grippo, J. F., and Hunziker, W. (1993). Two nuclear signalling pathways for vitamin D. Nature 361, 657-660. Casanova, J., Helmer, E., Selmi-Ruby, S., Qi, J.-S., Au-Flieger, M., Desai-Yajnik, V., Koudinova, N., Y a m , F., Raaka, B. M., and Samuels, H. H. (1994). Functional evidence for ligand-dependent dissociation of thyroid hormone and retinoic acid receptors from a n inhibitory cellular factor. Mol. Cell. Biol. 14, 5756-5765. Cash, D. E., Underhill, T. M., and Linney, E. (1995). In preparation. Chambon, P. (1993). The molecular and genetic dissection of the retinoid signalling pathway. Gene 135,223-228. Chambon, P. (1994). The retinoid signaling pathway: Molecular and genetic analyses. Semin. Cell Biol. 5, 115-125. Chang, D. T., Lopez, A., von Kessler, D. P., Chiang, C., Simandl, B. K., Zhao, R., Seldin, M. F., Fallon, J. F., and Beachy, P. A. (1994). Products, genetic linkage and limb patterning activity of a murine hedgehog gene. Development 120, 3339-3353. Chen, Y., Huang L., Russo, A. F., and Solursh, M. (1992). Retinoic acid is enriched in Hensen’s node and is developmentally regulated in the early chicken embryo. P m . Natl. Acad. Sci. U S A . 89, 10056-10059. Chen, 2.-P., Shemshedini, L., Durand, B., Noy, N., Chambon, P., and Gmnemeyer, H. (1994). Pure and functionally homogeneous recombinant retinoid X receptor. J.Biol. Chem. 269,25770-25776. Cheng, L., Norris, A. W., Tate, B. F., Rosenberger, M., Grippo, J. F., and Li, E. (1994).
444
T. MICHAEL UNDERHILL et al.
Characterization of the ligand binding domain of human retinoid X receptor a expressed in Escherichia coli. J . Biol. Chem. 269, 18662-18667. Cheskis, B., and Freedman, L. P. (1994). Ligand modulates the conversion of DNA-bound vitamin D, receptor (VDR) homodimers into VDR-retinoid X receptor heterodimers. Mol. Cell. Biol. 14, 3329-3338. Cohlan, S. Q. (1953).Excessive intake of vitamin A as a cause of congenital anomalies in the rat. Science 117,535-536. Colbert, M. C., Linney, E., and LaMantia, A.3. (1993). Local sources of retinoic acid coincide with retinoid-mediated transgene activity during embryonic development. Proc. Natl. Acad. Sci. U.S.A. 90, 6572-6576. Conlon, R. A., and Rossant, J . (1992). Exogenous retinoic acid rapidly induces anterior ectopic expression of murine Hox-2 genes in viuo. Development 116, 357-368. Cooney, A. J., Leng, X., Tsai, S. Y., O’Malley, B. W., and Tsai, M.-J. (1993). Multiple mechanisms of chicken ovalbumin upstream promoter transcription factordependent repression of transactivation by the vitamin D, thyroid hormone, and retinoic acid receptors. J. Biol. Chem. 268, 4152-4160. Cooney, A. J., Tsai, S. Y., OMalley, B. W., and Tsai, M.-J. (1992). Chicken ovalbumin upstream promoter transcription factor (COUP-TF) dimers bind to different GTCA response elements, allowing COUP-TF to repress hormonal induction of the vitamin D3, thyroid hormone, and retinoic acid receptors. Mol. Cell. Biol. 12, 4153-4163. Crow, J. A,, and Ong, D. E. (1985). Cell-specific immunohistochemical localization of a cellular retinol-binding protein (type two) in the small intestine of rat. Proc. Natl. Acad. Sci. U.S.A.82, 4707-4711. Dalman, F. C., Sturzenbecker, L. J., Levin, A. A., Lucas, D. A,, Perdew, G. H., Petkovich, M., Chambon, P., Grippo, J . F., and Pratt, W. B. (1991).Retinoic acid receptor belongs to a subclass of nuclear receptors that do not form “docking” complexes with hsp90. Biochemistry 30, 5605-5608. Davis, K. D., Berrodin, T. J., Stelmach, J. E., Winkler, J. D., and Lazar, M. A. (1994). Endogenous retinoid X receptors can function as hormone receptors in pituitary cells. Mol. Cell. Biol. 14, 7105-7110. Davis, K. D., and Lazar, M. A. (1993). Induction of retinoic acid receptor+ by retinoic acid is cell specific. Endocrinology 132, 1469-1474. de The, H., del Mar Vivanco-Ruiz, M., Tiollais, P., Stunnenberg, H., and Dejean, A. (1990). Identification of a retinoic acid responsive element in the retinoic acid receptor p gene. Nature 343, 177-180. Dedhar, S., Rennie, P. S., Shago, M., Leung Hagesteijn, C.-Y., Yang, H., Filmus, J., Hawley, R. G., Bruchovsky, N., Cheng, H., Matusik, R. J., and Giguere, V. (1994). Inhibition of nuclear hormone receptor activity by calreticulin. Nature 367, 480483. Dencker, L., Annerwall, E., Busch, C., and Eriksson, U. (1990). Localization of specific retinoid-binding sites and expression of a cellular retinoic-acid-binding protein (CRABP) in the early mouse embryo. Development 110, 343-352. Dolle, P., Ruberte, E., Kastner, P., Petkovich, M., Stoner, C. M., Gudas, L. J., and Chambon, P. (1989). Differential expression of genes encoding a, p and y retinoic acid receptors and CRABP in the developing limbs of the mouse. Nature 342, 702-705. Dolle, P., Ruberte, E., Leroy, P., Morriss-Kay, G., and Chambon, P. (1990). Retinoic acid receptors and cellular retinoid binding proteins. Development 110, 1133-1151. Durand, B., Saunders, M., Leroy, P., Leid, M., and Chambon, P. (1992). All-trans and 9-cis retinoic acid induction of the CRABPII transcription is mediated by RAR-RXR heterodimers bound to DR1 and DR2 repeated motifs. Cell 71, 73-85.
RETINOIDS AND MOUSE EMBRYOS
445
Durston, A,, Timmermans, P., Hage, W., Hendriks, H., de Vries, N., Heideveid, M., and Nieuwkoop, P. (1989).Retinoic acid causes an antero-posterior transformation in the developing central nervous system. Nature 340,140-144. Ede, D. (1983).Cellular condensations and chondrogenesis. Zn “Cartilage: Development, Differentiation and Growth” (B. K. Hall, ed.), pp. 143-185. Academic Press, New York. Espeseth, A,, Murphy, S. M., and Linney, E. (1989).Retinoic acid receptor expression vector inhibits differentiation of F9 embryonal carcinoma cells. Genes Dev. 3,16471656. Fanjul, A., Dawson, M. I., Hobbs, P. D., Jong, L., Cameron, J. F., Harlev, E., Graupner, G., Lu, X.-P., and Pfahl, M. (1994).A new class of retinoids with selective inhibition of AP-1 inhibits proliferation. Nature 372, 107-111. Fawcett, D., Pasceri, P., Fraser, R., Colbert, M., Rossant, J., and Giguere, V. (1995). Postaxial polydactyly in forelimbs of CRABP-ZI mutant mice. Development 121, 671-679. Fiorella, P. D., and Napoli, J. L. (1991).Expression of cellular retinoic acid binding protein (CRABP) in Escherichia coli. Characterization and evidence the holoCRABP is a substrate in retinoic acid metabolism. J. Biol. Chern. 266,16572-16579. Fleischhauer, K.,Park, J. H., DiSanto, J. P., Marks, M., Ozato, K., and Yang, S. Y. (1992). Isolation of a full-length cDNA clone encoding a N-terminally variant form of the human retinoid X receptor j3. Nucleic Acids Res. 20, 1801. Folkers, G. E., van der Leede, B.J.M., and van der Saag, P. T.(1993).The retinoic acid receptor-j32 contains two separate cell-specific transactivation domains, at the N-terminus and in the ligand-binding domain. Mol. Endocrinol. 7 , 616-627. Fondell, J. D., Roy, A. L., and Roeder, R. G. (1993).Unliganded thyroid hormone receptor inhibits formation of a functional preinitiation complex: Implications for active repression. Genes Dev. 7 , 1400-1410. Forman, B. M., and Samuels, H. H. (1990).Dimerization among nuclear hormone receptors. New Biol. 2, 587-594. Forman, B. M., Casanova, J., Raaka, B. M., Ghysdael, J., and Samuels, H. H. (1992). Half-site spacing and orientation determines whether thyroid hormone and retinoic acid receptors and related factors bind to DNA response elements as monomers, homodimers, or heterodimers. Mol. Endocrinol. 6 , 429-442. Forman, B. M.,Yang C.-R., Au, M., Casanova, J., Ghysdael, J., and Samuels, H. H. (1989).A domain containing leucine zipper like motifs mediate novel in uivo interactions between the thyroid hormone and retinoic acid receptors. Mol. Endocrinol. 3, 1610-1626. Forman, B. M., Yang, C.-R., Casanova, F., and Samuels, H. H. (1988).c-erbA protooncogenes mediate thyroid hormone-dependent and independent regulation of the rat growth hormone and prolactin genes. Mol. Endocrinol. 2, 902-911. Francis, P. H., Richardson, M. K., Brickell, P. M., and Tickle, C. (1994).Bone morphogenetic proteins and a signalling pathway that controls patterning in the developing chick limb. Development 120, 209-218. Gaub, M.P., Rochette-Egly, C., Lutz, Y.,Ali, S., Matthes, H., Scheuer, I., and Chambon, P. (1992).Immunodetection of multiple species of retinoic acid receptor a:Evidence for phosphorylation. Exp. Cell Res. 201,335-346. Gazit, D., Ebner, R., Kahn, A. J., and Derynck, R. (1993).Modulation of expression and cell surface binding of members of the transforming growth factor-j3 superfamily during retinoic acid-induced osteoblastic differentiation of multipotential mesenchymal cells. Mol. Endocrinol. 7, 189-198.
446
T.MICHAEL UNDERHILL et al.
Gearing, K. L., Gottlicher, M., Teboul, M., Widmark, E., and Gustafsson, J.-A. (1993). Interaction of the peroxisome-proliferator-activated receptor and retinoid X receptor. Proc. Natl. Acad. Sci. U.S.A.90, 1440-1444. Gehring, W. J. (1992). The homeobox in perspective, Trends Biochem. Sci. 17, 277-280. Giguere, V. (1994). Retinoic acid receptors and cellular retinoid binding proteins: Complex interplay in retinoid signaling. Endocr. Rev. 15, 61-79. Giguere, V., Lyn, S., Yip, P., Siu, C. H., and Amin, S. (1990). Molecular cloning of a cDNA encoding a second cellular retinoic acid-binding protein. Proc. Natl. Acad. Sci. U.S.A. 87, 6233-6237. Giguere, V., Ong, E. S., Segui, P., and Evans, R. M. (1987). Identification of a receptor for the morphogen retinoic acid. Nature 330,624-629. Giguere, V., Shago, M., Zirngibl, R., Tate, P., Fbssant, J., and Varmuza, S. (1990). Identification of a novel isoform of the retinoic acid receptor y expressed in the mouse embryo. Mol. Cell. Biol. 10, 2335-2340. Glass, C. K. (1994). Differential recognition of target genes by nuclear receptor monomers, dimers, and heterodimers. Endocr. Rev. 15, 391-407. Glass, C. K., Devary, 0. V., and Fbsenfeld, M. G. (1990). Multiple cell type-specific proteins differentially regulate target sequence recognition by the a retinoic acid receptor. Cell 63, 729-738. Glass, C. K., Lipkin, S. M., Devary, 0. V., and Fbsenfeld, M. G. (1989). Positive and negative regulation of gene transcription by a retinoic acid-thyroid hormone receptor heterodimer. Cell 59, 697-708. Gorry, P., Lufiin, T., Dierich, A,, Rochette-Egly, C., Decimo, D., Dolle, P., Mark, M., Durand, B., and Chambon, P. (1994). The cellular retinoic acid binding protein I is dispensable. Proc. Natl. Acad. Sci. U.S.A. 91, 9032-9036. Green, S., and Chambon, P. (1988). Nuclear receptors enhance our understanding of transcription regulation. Trends Genet. 4, 309-314. Gudas, L. J. (1994). Retinoids and vertebrate development. J. Biol. Chem. 269, 1539915402. Hale, F. (1935). The relation of vitamin A to anophthalmos in pigs. A m . J . Ophthalmol. 18, 1087-1093. Hall, B. L., Smit-McBride, Z., and Privalsky, M. L. (1993). Reconstitution of retinoid X receptor function and combinatorial regulation of other nuclear hormone receptors in the yeast Saccharomyces cereuisiue. Proc. Natl. Acad. Sci. U.S.A. 90,6929-6933. Hallenbeck, P. L., Marks, M. S., Lippoldt, R. E., Ozato, K., and Nikodem, V. M. (1992). Heterodimerization of thyroid hormone (TH) receptor with H-2RIIBP (RXR p) enhances DNA binding and TH-dependent transcriptional activation. Proc. Natl. Acad. Sci. U.S.A. 89, 5572-5576. Hamada, K., Gleason, S. L., Levi, B.-Z., Hirschfield, S., Appella, E., and Ozato, K. (1989). H-BRIIBP, a member of the nuclear hormone receptor superfamily that binds to both the regulatory element of the major histocompatibility complex class I genes and the estrogen response element. Proc. Natl. Acad. Sci. U.S.A. 86, 82898293. Hard, T., Kellenbach, E., Boelens, R., Maler, B. A,, Dahlman, K., Freedman, L. P., Carlstedt-Duke, J., Yamamoto, K. R., Gustafsson, J. A., and Kaptein, R. (1990). Solution structure of the glucocorticoid receptor DNA-binding domain. Science 249, 157-160. Heery, D. M., Pierrat, B., Gronemeyer, H., Chambon, P., and Losson, R. (1994). Homoand heterodimers of the retinoid X receptor (RXR) activate transcription in yeast. Nucleic Acids Res. 22, 126-131. Heery, D. M., Zacharewski, T., Pierrat, B., Gronemeyer, H., Chambon, P., and Losson, R.
RETINOIDS AND MOUSE EMBRYOS
447
(1993). Efficient transactivation by retinoic acid receptors in yeast requires retinoid
X receptors. Proc. Natl. Acad. Sci. U S A . 90, 4281-4285. Helms, J., Thaller, C., and Eichele, G. (1994). Relationship between retinoic acid and sonic hedgehog, two polarizing signals in the chick wing bud. Deuelopment 120, 3267-3274.
Heyman, R. A.. Mangelsdorf, D. J., Dyck, J. A., Stein, R., Eichele, G., Evans, R. M., and Thaller, C. (1992). 9-cis retinoic acid is a high affinity ligand for the retinoid X receptor. Cell 68,397-406. Hirsch, M.-R., Gaugler, L., Deagostini-Bazin, H., Bally-Cuif, L., and Goridis, C. (1990). Identification of positive and negative regulatory elements governing cell-typespecific expression of the neural cell adhesion molecule gene. Mol. Cell. Bwl. 10, 1959-1968.
Hoffmann, B., Lehmann, J. M., Zhang, X.-K., Hermann, T., Husmann, M., Graubner, G., and Pfahl, M. (1990). A retinoic acid receptor-specific element controls the retinoic acid receptor+ promoter. Mol. Endmrinol. 4, 1727-1736. Hofmann, C., and Eichele, G. (1994). Retinoids in development. I n “The Retinoids: Biology, Chemistry and Medicine” (M. B. Sporn, A. B. Roberts, and D. S. Goodman, eds.), pp. 387-440. Raven, New York. Husmann, M., Gorgen, I., Weisgerber, C., and Bitter-Suermann, D. (1989). Up-regulation of embryonic NCAM in an EC cell line by retinoic acid. Deu. Biol. 136, 194-200. Husmann, M., Hoffmann, B., Stump, D. G., Chytil, F., and Pfahl, M. (1992). A retinoic acid response element from the rat CRBPI promoter is activated by an RARlRXR heterodimer. Biochern. Biophys. Res. Commun. 187, 1558-1564. Imperiale, M. J., Kao, H.-T., Feldman, L. T., Nevins, J. R., and Strickland, S. (1984). Common control of the heat shock gene and early adenovirus genes: Evidence of a cellular E1A-like activity. Mol. Cell. Bwl.4, 867-874. Jones, F. S., Holst, B. D., Minowa, O., De Robertis, E. M., and Edelman, G. M. (1993). Binding and transcriptional activation of the promoter for the neural cell adhesion molecule by HoxC6 ( H o d . 3 ) .Proc. Natl. Acad. Sci. U.S.A. 90, 6557-6561. Jones, F. S., Prediger, E. A,, Bittner, D. A., De Robertis, E. M., and Edelman, G. M. (1992). Cell adhesion molecules as targets for Hox genes: Neural cell adhesion molecule promoter activity is modulated by cotransfection with Hox-2.5 and 2.4. Proc. Natl. Acad. Sci. U.S.A. 89,2086-2090. Jonk, L. J., De Jonge, M. E. J., Vervarrt, J. M. A., Wissink, S., and Kruijer, W. (1994). Isolation and developmental expression of retinoic-acid-induced genes. Deu. Biol. 161,604-614.
Kalter, H., and Warkany, J. (1961). Experimental production of congenital malformations in strains of inbred mice by maternal treatment with hypervitaminosis. Am. J. Pathol. 38, 1-21. Kastner, P., Grondona, J. M., Mark, M., Gansmuller, A., LeMeur, M., Decimo, D., Vonesch, J.-L., Dolle, P., and Chambon, P. (1994). Genetic analysis of RXRn developmental function: Convergence of RXR and RAR signaling pathways in heart and eye morphogenesis. Cell 78, 987-1003. Kastner, P., Krust, A., Mendelsohn, C., Gamier, J. M., Zelent, A,, Leroy, P., Staub, A,, and Chambon, P. (1990). Murine isoforms of retinoic acid receptor y with specific patterns of expression. Proc. Natl. Acad. Sci. U S A . 87,2700-2704. Keidel, S., LeMotte, P., and Apfel, C. (1994). Different agonist- and antagonist-induced conformational changes in retinoic acid receptors analyzed by protease mapping. Mol. Cell. Biol. 14, 287-298. Kessel, M. (1992). Respecification of vertebral identities by retinoic acid. Deuelopment 115,487-501.
448
T. MICHAEL UNDERHILL et a1
Kliewer, S. A., Umensono, K., Heymann, R. A,, Mangelsdorf, D. J., Dyck, J. A,, and Evans, R. M. (1992a). Retinoid X receptor-COUP-TF interactions modulate retinoic acid signaling. Proc. Natl. Acad. Sci. U.S.A.89, 1448-1452. Kliewer, S. A., Umesono, K., Mangelsdorf, D. J., and Evans, R. M. (1992b). Retinoid X receptor interacts with nuclear receptors in retinoic acid, thyroid hormone and vitamin D, signalling. Nature 355, 446-449. Kliewer, S. A,, Umesono, K., Noonan, D. J., Heyman, R. A,, and Evans, R. M. (1992~). Convergence of 9-cis retinoic acid and peroxisome proliferator signaling pathways through heterodimer formation of their receptors. Nature 358, 771-774. Kochhar, D. M. (1967). Teratogenic activity of retinoic acid. Acta Pathol. Microbiol. Scand. 70,398-404. Krumlauf, R. (1993). Mouse Hox genetic functions. Curr. Opin. Genet. Deu. 3,621-625. Kruyt, F. A. E., Folkers, G. E., Walhout, A. J. M., van der Leede, B.J., and van der Saag, P. T. (1993). E1A functions as a coactivator of retinoic acid-dependent retinoic acid receptor-P2 promoter activation. Mol. Endocrinol. 7 , 604-615. Kruyt, F. A. E., van der Veer, L. J., Mader, S., van den Brink, C., Feijen, A., Jonk, L. J . C., Kruijer, W., and van der Saag, P. T. (1992). Retinoic acid resistance of the variant embryonal carcinoma cell line RAC65 is caused by expression of a truncated RARa. Differentiation 49, 27-37. Kurokawa, R., DiRenzo, J., Boehm, M., Sugarman, J., Gloss, B., Rosenfeld, M. G., Heyman, R. A., and Glass, C. K. (1994). Regulation of retinoid signalling by receptor polarity and allosteric control of ligand binding. Nature 371, 528-531. Kurokawa, R., Yu, V. C., Naar, A., Kyakumoto, S., Han, Z., Silverman, S., Rosenfeld, M. G., and Glass, C. K. (1993). Differential orientations of the DNA-binding domain and carboxy-terminal dimerization interface regulate binding site selection by nuclear receptor heterodimers. Genes Deu. 7 , 1423-1435. La Thangue, N. B., and Rigby, P. W. J. (1987). An adenovirus E1A-like transcription factor is regulated during the differentiation of murine embryonal carcinoma stem cells. Cell 49, 507-513. LaMantia, A.-S., Colbert, M. C., and Linney, E. (1993). Retinoic acid induction and regional differentiation prefigure olfactory pathway formation in the mammalian forebrain. Neuron 10, 1035-1048. Lammer, E. J., Chen, D. T., Hoar, R. M., Agnish, N. D., Benke, P. J., Braun, J . T., Curry, C. J., Fernhoff, P. M., Grix, A. W., Lott, I. T., Richard, J. M., and Sun, S. C. (1985). Retinoic acid embryopathy. N . Engl. J. Med. 313, 837-841. Lampron, C., Rochette-Egly, C., Gorry, P., DollB, P., Mark, M., Lufkin, T., LeMeur, M., and Chambon, P. (1995). Mice deficient in cellular retinoic acid binding protein I1 (CRABPII) o r in both CRABPI and CRABPII are essentially normal. Development 121,539-548. Langston, A. W., and Gudas, L. J. (1992). Identification of a retinoic acid responsive element 3' of the murine homeobox gene Hox-1.6 Mech. Deu. 38, 217-228. Langston, A. W., and Gudas, L. J . (1994). Retinoic acid and homeobox gene regulation. Curr. Opin. Genet. Dev. 4, 550-555. Laudet, V., Hanni, C., Coll, J., Catzeflis, F., and Stehelin, D. (1992). Evolution of the nuclear receptor gene superfamily. EMBO J. 11, 1003-1013. Lee, M. S., Kliewer, S. A., Provencal, J., Wright, P. E., and Evans, R. M. (1993).Structure of the retinoid X receptor a DNA binding domain: A helix required for homodimeric DNA binding. Science 260, 1117-1121. Lehmann, J., Zhang, X.-K., Graupner, G., Lee, M.-O., Hermann, T., Hoffmann, B., and Pfahl, M. (1993). Formation of retinoid X receptor homodimers leads to repression of
RETINOIDS AND MOUSE EMBRYOS
449
T, response: Hormonal crosstalk by ligand-induced squelching. Mol. Cell. Biol. 13, 7698-7707. Lehmann, J. M., Hoffmann, B., and Pfahl, M. (1991). Genomic organization of the retinoic acid receptor y gene. Nucleic Acids Res. 19, 573-578. Lehmann, J. M., Zhang, X.-K., and Pfahl, M. (1992). RARy2 expression is regulated through a retinoic acid response element embedded in Spl sites. Mol. Cell. Biol. 12, 2976-2985. Leid, M., Kastner, P., and Chambon, P. (1992a). Multiplicity generates diversity in the retinoic acid signalling pathways. Trends Biochem. Sci. 17, 427-433. h i d , M., Kastner, P., Lyons, R., Nakshatri, H., Saunders, M., Zacharewski, T., Chen, J.-Y., Staub, A., Garnier, J.-M., Mader, S.,and Chambon, P. (1992b). 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. h i d , M., Kastner, P., Lyons, R., Nakshatri, H., Saunders, M., Zacharewski, T., Chen, J.-Y., Staub, A., Garnier, J.-M., Mader, S.,and 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. Leroy, P., Krust, A,, Zelent, A., Mendelsohn, C., Garnier, J.-M., Kastner, P., Dierich, A,, and Chambon, P. (1991a). Multiple isoforms of the mouse retinoic acid receptor a are generated by alternative splicing and differential induction by retinoic acid. EMBO J . 10, 59-69. Leroy, P., Nakshatri, H., and Chambon, P. (1991b). Mouse retinoic acid receptor a 2 isoform is transcribed from a promoter that contains a retinoic acid response element. Proc. Natl. Acad. Sci. U S A . 88,10138-10142. Levin, A. A., Stunenbecker, L. J., Kazmer, S.,Bosakowski, T., Huselton, C., Allenby, G., Speck, J., Kratzeisen, C., Rosenberger, M., Lovey, A., and Grippo, J. F. (1992). 9-cis retinoic acid stereoisomer binds and activates the nuclear receptor RXRa. Nature 355,359-361. Li, E., Demmer, L. A., Sweetser, D. A., Ong, D. E., and Gordon, J. I. (1986). Rat cellular retinol-binding protein 11: Use of a cloned cDNA to define its primary structure, tissue-specific expression, and developmental regulation. Proc. Natl. Acad. Sci. U S A . 83,5779-5783. Li, E., Sucov, H. M., Lee, K.-F., Evans, R. M., and Jaenisch, R. (1993). Normal development and growth of mice carrying a targeted disruption of the a1 retinoic acid receptor gene. Proc. Natl. Acad. Sci. U S A . 90, 1590-1594. Linney, E. (1992). Retinoic acid receptors: Transcription factors modulating gene regulation, development and differentiation. In “Current Topics in Developmental Biology” (P. Ra, ed.), pp. 309-350. Academic Press, San Diego. Linney, E., and LaMantia, A . 3 (1994). Retinoic acid signaling during mouse development. Advances in Development BiologyP. L. Wassarman, ed.) in press. Lipkin, S. M., Nelson, C. A., Glass, C. K., and Rosenfeld, M. G. (1992). A negative retinoic acid response element in the rat oxytocin promoter restrich transcriptional stimulation by heterologous transactivation domains. Proc. Natl. Acad. Sci. U S A . 89, 1209-1213. Liu, Q., and Linney, E. (1993). The mouse retinoid-X receptor-y gene: Genomic organization and evidence for functional isoforms. Mol. Endocrinol. 7, 651-658. Lohnes, D., Kastner, P., Dierich, A., Mark, M., LeMeur, M., and Chambon, P. (1993). Function of retinoic acid receptor y in the mouse. Cell 73, 643-658. Lohnes, D., Mark, M., Mendelsohn, C., Dolle, P., Dierich, A,, Gorry, P., Gansmuller, A., and Chambon, P. (1994). Function of the retinoic acid receptors (RARS) during
450
T.MICHAEL UNDERHILL et a1
development (I) craniofacial and skeletal abnormalities in RAR double mutants. Development 120, 2723-2748. Lufiin, T., Lohnes, D., Mark, M., Dierich, A., Gorry, P., Gaub, M.-P., LeMeur, M., and Chambon, P. (1993). High postnatal lethality and testis degeneration in retinoic acid receptor ci mutant mice. Proc. Natl. Acad. Sci. U.S.A.90, 7225-7229. Luisi, B. F., Xu, W. X., Otwinowski, Z., Freedman, L. P., Yamamoto, K. R., and Sigler, P. B. (1991). Crystallographic analysis of interaction of the glucocorticoid receptor with DNA. Nature 352,497-505. Lyn, S., and Giguere, V. (1994). Localization of CRABP-I and CRABP-I1 mRNA in the early mouse embryo by whole-mount in situ hybridization: Implications for teratogenesis and neural development. Dev. Dyn. 199, 280-291. MacDonald, P. N., Dowd, D. R., Nakajima, S., Galligan, M. A., Reeder, M. C., Haussler, C. A., Ozato, K., and Haussler, M. R. (1993). Retinoid X receptors stimulate and 9-cis retinoic acid inhibits 1,25-dihydroxyvitamin D,-activated expression of the rat osteocalcin gene. Mol. Cell. Biol. 13, 5907-5917. MacGregor, T. M., Copeland, N. G., Jenkins, N. A., and Giguere, V. (1992). The murine gene for cellular retinoic acid-binding protein type 11. J. Biol. Chem. 267, 77777783. Maden, M., and Holder, N. (1991).The involvement of retinoic acid in the development of the vertebrate central nervous system. Development 2 (suppl.), 87-94. Maden, M., Horton, C., Graham, A,, Leonard, L., Pizzey, J., Siegenthaler, G., Lumsden, A., and Eriksson, U. (1992). Domains of cellular retinoic acid-binding protien I (CRABPI) expression in the hindbrain and neural crest of the mouse embryo. Mech. Deu. 37, 13-23. Maden, M., Hunt, P., Eriksson, U., Kurioiwa, A., Krumlauf, R., and Summerbell, D. (1991). Retinoic acid-binding protein, rhombomeres and the neural crest. Deuelopment 111,35-44. Maden, M., Ong, D. E., and Chytil, F. (1990). Retinoid-binding protein distribution in the developing mammalian nervous system. Development 109, 75-80. Maden, M., Ong, D. E., Summerbell, D., and Chytil, F. (1988). Spatial distribution of cellular protein binding to retinoic acid in the chick limb bud. Nature 335,733-735. Mader, S., Chen, J.-Y., Chen, Z., White, J., Chambon, P., and Gronemeyer, H. (1993a). The patterns of binding of RAR, RXR and TR homo- and heterodimers t o direct repeats are dictated by the binding specificities of the DNA binding domains. EMBO J . 12,5029-5041. Mader, S., Leroy, P., Chen, J.-Y., and P., C. (1993b). Multiple parameters control the selectivity of nuclear receptors for their response elements. J . Biol. Chem. 268,591600. Mahmood, R., Flanders, K. C., and Morriss-Kay, G. M. (1992). Interactions between retinoids and TGF Ps in mouse morphogenesis. Development 115, 67-74. Mak, P., Fuernkranz, H. A., Ge, R., and Karathanasis, S. K. (1994). Retinoid X receptor homodimers function as transcriptional activators in yeast. Gene 145, 129-133. Mangelsdorf, D. J., Borgmeyer, U., Heyman, R. A., Zhou, J . Y., Ong, E. S., Oro, A. E., Kakizuka, A., and Evans, R. M. (1992). Characterization of three RXR genes that mediate the action of 9-cis retinoic acid. Genes Dev. 6, 329-344. Mangelsdorf, D. J., Ong, E. S., Dyck, J. A., and Evans, R. M. (1990). Nuclear receptor that identifies a novel retinoic acid response pathway. Nature 345, 224-229. Mangelsdorf, D. J., Umesono, K., and Evans, R. M. (1994). The retinoid receptors. In “The Retinoids: Biology, Chemistry, and Medicine” (M. B. Sporn, A. B. Roberts, and D. S. Goodman, eds.), pp. 319-349. Raven, New York. Mangelsdorf, D. J., Umesono, K., Kliewer, S. A,, Borgmeyer, U., Ong, E. S., and Evans,
RETINOIDS AND MOUSE EMBRYOS
45 1
R. M. (1991). A direct repeat in the cellular retinol-binding protein type I1 gene confers differential regulation by RXR and RAR. Cell 66, 555-561. Marks, M. S., Hallenbeck, P. L., Nagata, T., Segars, J . H., Appella, E., Nikodem, V. M., and Ozato, K. (1992). H-2RIIBP (RXRP) heterodimerization provides a mechanism for combinatorial diversity in the regulation of retinoic acid and thyroid hormone responsive genes. EMBO J . 11, 1419-1435. Marshall, H., Nonchev, S., Sham, M. H., Muchamore, I., Lumsden, A., and Krumlauf, R. (1992). Retinoic acid alters hindbrain €€ox code and induces transformation of rhombomeres 213 into a 415 identity. Nature 360, 737--741. Marshall, H., Studer, M., Popperl, H., Aparicio, S., Kuroiwa, A., Brenner, S., and Krumlauf, R. (1994). A conserved retinoic acid response element required for early expression of the homeobox gene Hoxb-1. Nature 370, 567-571. Mavillo, F. (1993). Regulation of vertebrate homeobox.-containinggenes by morphogens. Eur. J . Biochem. 212,273-288. McCaffery, P., and Drager, U. C. (1994). Hot spots of retinoic acid synthesis in the developing spinal cord. Proc. Natl. Acad. Sci. U.S.A. 91, 7194-7197. Medin, J. A,, Minucci, S., Driggers, P. H., Lee, I. J., and Ozato, K. (1994). Quantitative increases in DNA binding affinity and positional effects determine 9-cis retinoic acid induced activation of the retinoid X receptor p homodimer. Mol. Cell. Endocrinol. 105, 27-35. Mendelsohn, C., Lohnes, D., Decimo, D., Lufkin, T., LeMeur, M., Chambon, P., and Mark, M. (1994a). Function of the retinoic acid receptors (RARs) during development (11) Multiple abnormalities at various stages of organogenesis in RAR double mutants. Development 120, 2749-2771. Mendelsohn, C., Mark, M., Dolle, P., Dierich, A., Gaub, M. P., Krust, A,, Lampron, C., and Chambon, P. (199413). Retinoic acid receptor B2 (RARB2) null mutant mice appear normal. Deu. Biol. 166, 246-258. Mendelsohn, C., Ruberte, E., and Chambon, P. (1992). Retinoid receptors in vertebrate limb development. Deu. Biol. 152, 50-61. Mendelsohn, C., Ruberte, E., LeMeur, M., Morriss-Kay, G., and Chambon, P. (1991). Developmental analysis of the retinoic acid-inducible RAR-pZ promoter in transgenic animals. Development 113, 723-734. Minucci, S., Zand, D. J., Dey, A., Marks, M. S., Nagata, T., Grippo, J. F., and Ozato, K. (1994). Dominant negative retinoid X receptor inhibits retinoic acid-responsive gene regulation in embryonal carcinoma cells. Mol. Cell. B i d . 14, 360-372. Moroni, M. C., Vigano, M. A,, and Mavilio, F. (1993). Regulation of the human HOXD4 gene by retinoids. Mech. Deu. 44, 139-154. Naar, A. M., Boutin, J.-M., Lipkin, S. M., Yu, V. C., Rolloway, J. M., Glass, C. K., and Rosenfeld, M. G. (1991). The orientation and spacing of core DNA-binding motifs dictate selective transcriptional responses to three nuclear receptors. Cell 65, 12671279. Nagata, T., Kanno, Y., Ozato, K., and Taketo, M. (199,Q).The mouse Rxrb gene encoding RXRp: Genomic organization and two mRNA isoforms generated by alternative splicing of transcripts initiated from CpG island promoters. Gene 142, 183-189. Nagaya, T., and Jameson, J. L. (1993). Thyroid hormone receptor dimerization is required for dominant negative inhibition by muta.tions that cause thyroid hormone resistance. J . Biol. Chem. 268, 15766-15771. Nagpal, S., Friant, S., Nakshatri, H., and Chambon, P. (1993). RARs and RXRs: Evidence for two autonomous transactivation functions (AF-1 and AF-2) and heterodimerization in uiuo. EMBO J. 12, 2349-2360. Nagpal, S., Saunders, M., Kastner, P., Durand, B., Nakshatri, H., and Chambon, P.
452
T. MICHAEL UNDERHILL et al.
(1992a). Promoter context- and response element-dependent specificity of the transcriptional activation and modulating functions of retinoic acid receptors. Cell 70, 1007-1019. Nagpal, S., Zelent, A,, and Chambon, P. (1992b). RAR-P4, a retinoic acid receptor isoform is generated from RAR-pZ by alternative splicing and usage of a CUG initiator codon. Proc. Natl. Acad. Sci. U.S.A. 89, 2718-2722. Nakshatri, H., and Chambon, P. (1994).The directly repeated RG(G/T)TCAmotifs of the rat and mouse cellular retinol-binding protein I1 genes are promiscuous binding sites for RAR, RXR, HNF-4, and ARP-1 homo- and heterodimers. J . Biol. Chem. 269, 890-902. Nau, H., Chahaud, I., Dencker, L., Lammer, E. J., and Scott, W. J. (1994). Teratogenicity of vitamin A and retinoids. I n “Vitamin A in Health and Disease” (R. Blomhoff, ed.), pp. 615-664. Dekker, New York. Newcomer, M. E., Pappas, R. S., and Ong, D. E. (1993). X-ray crystallographic identification of a protein-binding site for both all trans and 9-cis-retinoic acid. Proc. Natl. Acad. Sci. U.S.A. 90, 9223-9227. Noden, D. M. (1988). Interactions and fates of avian craniofacial mesenchyme. Development 103 (suppl.), 121-140. Noji, S., Nohno, T., Koyama, E., Muto, K., Ohyama, K., Aoki, Y., Tamura, K., Ohsugi, K., Ide, H., Taniguchi, S., and Saito, T. (1991). Retinoic acid induces polarizing activity but is unlikely to be a morphogen in the chick limb bud. Nature 350, 83-86. Noji, S., Yamaai, T., Koyama, E., Nohno, T., Fujimoto, W., Arafat, J., and Taniguchi, S. (1989). Expression of retinoic acid receptor genes in keratinizing front skin. FEBS Lett. 259, 86-90. Oberlender, S. A., and Tuan, R. S. (1994). Expression and functional involvement of N-cadherin in embryonic limb chondrogenesis. Development 120, 177-187. Ong, D. E. (1984). A novel retinol-binding protein from rat. J . B i d . Chem. 259, 14761482. Ong, D. E., and Chytil, F. (1978). Cellular retinoic acid-binding protein from rat testis. J . Biol. Chem. 253, 4551-4554. Ong, D. E., and Chytil, F. (1988). Presence of novel retinoic acid-binding proteins in the lumen of r a t epididymis. Arch. Biochem. Biophy. 267, 474-478. Ong, D. E., Newcomer, M. E., and Chytil, F. (1994). Cellular retinoid-binding proteins. I n “The Retinoids” (M. B. Sporn, A. B. Roberts, and D. S. Goodman, eds.), pp. 283317. Raven, New York. Oro, A. E., McKeown, M., and Evans, R. M. (1990). Relationship between the product of the Drosophila ultraspiracle locus and the vertebrate retinoid X receptor. Nature 347,298-301. Osumi-Yamashita, N., Noji, S., Noho, T., Koyama, E., Doi, H., Eto, K., and Taniguchi, S. (1990). Expression of retinoic acid receptor genes in neural crest derived cells during mouse facial development. FEBS Lett. 264, 71-74. Papalopulu, N., Lovell-Badge, R., and Krumlauf, R. (1991). The expression of murine Hox-2 genes is dependent on the differentiation pathway and displays a collinear sensitivity to retinoic acid in F9 cells and Xenopus embryos. Nucleic Acids Res. 19, 5497-5506. Pearlmann, T., Rangarajan, P. N., Umesono, K., and Evans, R. M. (1993). Determinants for selective RAR and TR recognition of direct repeat HREs. Genes Dev. 7, 14111422. Perez, A., Kastner, P., Sethi, S., Lutz, Y., Reibel, C., and Chambon, P. (1993). PMLRAR homodimers: Distinct DNA binding properties and heteromeric interactions with RXR. EMBO J . 12, 3171-3182.
RETINOIDS AND MOUSE EMBRYOS
453
Perez-Castro, A. V., Toth-Rogler, L. E., Wei, L.-N., and Nguyen-Huu, M. C. (1989). Spatial and temporal pattern of expression of the cellular retinoic acid-binding protein and the cellular retinol-binding protein during mouse embryogenesis. Proc. Natl. Acad. Sci. U S A . 86,8813-8817. Perez-Castro, A. V., "ran, V. T., and Nguyen-Huu, M. C. (1993). Defective lens fiber differentiation and pancreatic tumorigenesis caused by ectopic expression of the cellular retinoic acid-binding protein I. Development 119, 363-375. Perides, G., Safran, R. M., Downing, L. A,, and Charness, M. E. (1994). Regulation of neural cell adhesion molecule and L1 by the transforming growth factor-p superfamily. J. Biol. Chem. 269, 765-770. Petkovich, M., Brand, N., Krust, A., and Chambon, P. (1987). A human retinoic acid receptor which belongs to the family of nuclear receptors. Nature 330, 444-450. Pfahl, M. (1993). Nuclear receptor/AP-1 interaction. Endocr. Rev. 14, 651-658. Popperl, H., and Featherstone, M. S. (1993). Identification of a retinoic acid response element upstream of the murine HOX-42 gene. Mol. Cell. Bwl. 13, 257-265. Pratt, M. A. C., Kralova, J., and McBurney, M. W. (1990). A dominant negative mutation of the a retinoic acid receptor gene in a retinoic acid-nonresponsive embryonal carcinoma cell. Mol. Cell. Biol. 10, 6445-6453. Pratt, M. A. C., Langston, A. W., Gudas, L. J.,and McBurney, M. W. (1993). Retinoic acid fails to induce expression of Hox genes in differentiation-defective murine embryonal carcinoma cells carrying a mutant gene for a retinoic acid receptor. Differentiation 53, 105-113. Predki, P. F., Zamble, D., Sarkar, B., and GiguBre, V. (1994). Ordered binding of retinoic acid and retinoid-X receptors to asymmetric response elements involves determinants adjacent to the DNA-binding domain. Mol. E n d o c r i d . 8, 31-39. Repa, J. J.,Hanson, K. K., and Clagett-Dame, M. (1993). All-trans-retinol is a ligand for the retinoic acid receptors. Proc. Natl. Acad. Sci. U S A . 90, 7293-7297. Reynolds, K., Mezey, E., and Zimmer, A. (1991). Activity of the p-retinoic acid receptor promoter in transgenic mice. Mech. Dev. 36, 15-29. Riddle, R. D., Johnson, R. L., Laufer, E., and Tabin, C. (1993). Sonic hedgehog mediates the polarizing activity of the ZPA. Cell 75, 1401-1416. Rochette-Egly, C., Gaub, M.-P., Lutz, Y.,Ali, S., Scheuer, I., and Chambon, P. (1992). Retinoic acid receptor+: Immunodetection and phosphorylation on tyrosine residues. Mol. Endocrinol. 6, 2197-2209. Rochette-Egly, C., Lutz, Y.,Saunders, M., Scheuer, I., Gaub, M.-P., and Chambon, P. (1991). Retinoic acid receptor y: Specific immunodetection and phosphorylation. J. Cell Biol. 115, 535-545. Rogers, M. B., Rosen, V., Wozney, J. M., and Gudas, L. J. (1992). Bone morphogenetic proteins-2 and -4 are involved in the retinoic acid-induced differentiation of embryonal carcinoma cells. Mol. Bwl. Cell 3, 189-196. Ross, T. K., Moss, V. E., Prahl, J. M., and DeLuca, H. F. (1992a). A nuclear protein essential for binding of rat 1,25-dihydroxyvitamin D, receptor to its response elements.Proc. Natl. Acad. Sci. U . S A . 89,256-260. Ross, T.K., Prahl, J. M., Herzber, I. M., and DeLuca, H. F. (1992b). Baculovirus-mediated expression of retinoic acid receptor type y in cultured insect cells reveals a difference in specific DNA-binding behavior with the 1,25-dihydroxyvitamin D, receptor. Proc. Natl. Acad. Sci. U S A . 89,10282-10286. Rossant, J., Zirngibl, R., Cado, D., Shago, M., and GiguBre, V. (1991). Expression of a retinoic acid response element-hsplacZ transgene defines specific domains of transcriptional activity during mouse embryogenesis. Genes Dev. 5, 1333-1344. Ruberte, E., Dolle, P., Chambon, P., and Morriss-Kay, G. (1991). Retinoic acid receptors
454
T.MICHAEL UNDERHILL et a1
and cellular retinoid binding proteins 11. Their differential pattern of transcription during early morphogenesis in mouse embryos. Development 111,45-60. Ruberte, E., Dolle, P., Krust, A,, Zelent, A., Morriss-Kay, G., and Chambon, P. (1990). Specific spatial and temporal distribution of retinoic acid receptor y transcripts during mouse embryogenesis. Development 108,213-222. Ruberte, E., Friederich, V., Chambon, P., and Morriss-Kay, G. (1993). Retinoic acid receptors and cellular binding proteins. Development 118, 267-282. Ruberte, E., Friederich, V., Morriss-Kay, G., and Chambon, P. (1992).Differential distribution patterns of CRABP I and CRABP I1 transcripts during mouse embryogenesis. Development 115, 973-987. Rudert, F., and Gronemeyer, H. (1993). Retinoic acid-responsive elements with a highly repetitive structure isolated by immuno-selection from genomic DNA. J. Steroid Biochem. Mol. Biol. 46, 121-133. Rutledge, J. C., Shourbaji, A. G., Hughes, L. A,, Polifka, J. E., Cruz, Y. P., Bishop, J. B., and Generoso, W. M. (1994). Limb and lower-body duplications induced by retinoic acid in mice. Proc. Natl. Acad. Sci. U.S.A.91, 5436-5440. Saatcioglu, F., Deng, T., and Karin, M. (1993). A novel cis element mediating liganddependent activation by c-ErbA: Implications for hormonal regulation. Cell 75, 1095-1105. Santiso-Mere, D., Sone, T., Hilliard IV, G. M., Pike, J. W., and McDonnell, D. P. (1993). Positive regulation of the vitamin D receptor by its cognate ligand in heterologous expression systems. MoL. Endocrinol. 7 , 833-839. Schule, R., Rangarajan, P., Kliewer, S., Ransone, L. J., Bolado, J., Verma, I., and Evans, R. M. (1991). Retinoic acid is a negative regulator of AP-1 responsive genes. Proc. Natl. Acad. Sci. U.S.A.88, 6092-6096. Scott, M. P., Tamkun, J. W., and Hartzell, G. W., I11 (1989). The structure and function of the homeodomain. Biochim. Biophys. Acta 989, 25-48. Shen, S., van der Saag, P. T., and Kruijer, W. (1993). Dominant negative retinoic acid receptor p. Mech. Dev. 40, 177-189. Shenefelt, R. E. (1972). Morphogenesis of malformations in hamsters caused by retinoic acid: Relation to dose and stage a t treatment. Teratology 5, 103-118. Simeone, A., Acampora, D., Nigro, V., Faiella, A., D’Esposito, M., Stornaiuolo, A,, Mavilo, F., and Boncinelli, E. (1991). Differential regulation by retinoic acid of the homeobox genes of the four HOX loci in human embryonal carcinoma cells. Mech. Deu. 33, 215-228. Smith, W. C., Nakshatri, H., Leroy, P., Rees, J., and Chambon, P. (1991). A retinoic acid response element is present in the mouse cellular retinol binding protein I (mCRBPI) promoter. EMBO J . 10, 2223-2230. Smith, S. M., and Eichele, G. (1991).Temporal and regional differences in the expression patterns of distinct retinoic acid receptor-b transcripts in the chick embryo. Deuelopment 111, 245-252. Solursh, M. (1983). Cell-cell interactions and chondrogenesis. In “Cartilage: Development, Differentiation and Growth” (B. K. Hall, ed.), pp. 121-141. Academic Press, New York. Soprano, D. R., and Blaner, W. S. (1994). Plasma retinol-binding protein. In “The Retinoids” (M. B. Sporn, A. B. Roberts, and D. S. Goodman, eds.), pp. 257-281. Raven, New York. Soprano, D. R., Soprano, K. J., and Goodman, D. S. (1986). Retinol-binding protein and transthyretin mRNA levels in visceral yolk sac and liver during fetal development in the rat. Proc. Natl. Acad. Sci. U.S.A. 83, 7330-7334.
RETINOIDS AND MOUSE EMBRYOS
455
Sporn, M. B., and Roberts, A. B. (1991). Interactions of retinoids and transforming growth factor-f3 in regulation of cell differentiation and proliferation. Mol. Endocrinol. 5, 3-7. Sporn, M. B., and Roberts,A. B. (1993). Roles of retinoids in differentiation and carcinogenesis. Cancer Res. 43, 3034-3040. Stunnenberg, H. G. (1993). Mechanisms of transactivation by retinoic acid receptors. Bioessays 15, 309-315. Sucov, H. M., Dyson, E., Gumeringer, C. L., Price, J., Chien, K. R., and Evans, R. M. (1994). RXRa mutant mice establish a genetic basis for vitamin A signaling in heart morphogenesis. Genes Dev. 8, 1007-1018. Sucov, H. M., Murakami, K. K.,and Evans, R. M. (1990). Characterization of a n autoregulated response element in the mouse retinoic acid receptor type p gene. Proc. Natl. Acad. Sci. U S A . 87,5392-5396. Sulik, K.K., Cook, C. S., and Webster, W. S. (1988). Teratogens and craniofacial malformations: Relationships to cell death. Development 103 (suppl.), 213-232. Sulik, K. K., Johnston, M. C., Srniley, S.J., Speight. H. S., and Jarvis, B. E. (1987). Mandibulofacial dysostosis (Treacher Collins syndrome): A new proposal for its pathogenesis. A m . J. Med. Genet. 27, 359-372. Sulik, K. K., Smiley, S. J., Turvey, T. A., Speight, H. S., and Johnston, M. C. (1989). Pathogenesis of cleft palate and Treacher Collins, Nager, and Miller syndromes. Cleft Palate Craniofac.J. 26, 209-216. Summerbell, D. (1983). The effect of local application of retinoic acid to the anterior margin of the developing chick limb. J. Embryol. Erp. Morphol. 78, 269-289. Sundelin, J.,Anundi, H., Tragardh, L., Eriksson, U., Lind, P., Ronne, H., Peterson, P. A,, and Rask, L. (1985). The primary structure of rat liver cellular retinol-binding protein. J . Biol. Chem. 260, 6494-6499. Tairis, N., Gabriel, J. L., Gyda, M., 111, Soprano, K. J., and Soprano, D. R. (1994). Arg”69 and Lys220 of retinoic acid receptor+ are important for the binding of retinoic acid. J . Biol. Chem. 269,19516-19522. Tate, B. F., Allenby, G., Janocha, R., Kazmer, S., Speck, J.,Stunenbecker, L. J., Abarzua, P., Levin, A. A,, and Grippo, J. F. (1994). Distinct binding determinants for 9-cis retinoic acid are located within AF-2 ofretinoic acid receptor a. Mol. Cell. Biol. 14, 2323-2330. Tickle, C., Alberts, B. M., Wolpert, L., and Lee, J. (1982). Local application of retinoic acid to the limb bud mimics the pattern of chick wing development. Nature 296, 564-565. Tickle, C . , Crawley, A., and Farrar, J. (1989).Retinoic acid application to chick wing buds leads to a dose-dependent reorganization of the apical ectodermal ridge that is mediated by the mesenchyme. Development 106, 691-705. Tini, M., Tsui, L.-C., and Giguhre, V. (1994). Heterodimeric interaction of the retinoic acid and thyroid hormone receptors in transcriptional regulation of the yF-crystallin everted retinoic acid response element. Mol. Endocrinol. 8, 14941505. Tini, M., Otulakowski, G., Breitman, M. L., Tsui, L.-C., and Gigukre, V. (1993). An everted repeat mediates retinoic acid induction of the yF-crystallin gene: Evidence of a direct role for retinoids in lens development. Genes Deu. 7, 295-307. Toney, J. H., Wu, L., Summerfield, A. E., Sanyal, G., Forman, B. M., Zhu, J., and Samuels, H. H. (1993). Conformational changes in chicken thyroid hormone receptor a1 induced by binding to ligand or to DNA. Biochemistry 32.2-6. Tran, P., Zhang, X.-K., Slabert, G., Hermann, T., Lehmann, J . M., and Pfahl, M. (1992).
456
T.MICHAEL UNDERHILL et a2
COUP orphan receptors are negative regulators of retinoic acid response pathways. Mol. Cell. Biol. 12, 4666-4676. Truss, M., and Beato, M. (1993). Steroid hormone receptors: Interaction with deoxyribonucleic acid and transcription factors. Endocr. Rev. 14, 459-479. Umesono, K., and Evans, R. M. (1989). Determinants of target gene specificity for steroid/thyroid hormone receptors. Cell 57. Umesono, K., Giguere, V., Glass, C. K., Rosenfeld, G. M., and Evans, R. M. (1988). Retinoic acid and thyroid hormone induce gene expression through a common response element. Nature 336, 262-265. Umesono, K., Murakami, K. K., Thompson, C. C., and Evans, R. M. (1991). Direct repeats as selective response elements for the thyroid hormone, retinoic acid and vitamin D, receptors. Cell 65, 1255-1266. Underhill, T. M., Cash, D. E., and Linney, E. (1994).Constitutively active retinoid receptors exhibit interfamily and intrafamily promoter specificity. Mol. Endocrinol. 8, 274-285. Wagner, M., Han, B., and Jessell, T. M. (1992). Regional differences in retinoid release from embryonic neural tissue detected by an in vitro reporter assay. Development 116,55-66. Wagner, M., Thaller, C., Jessell, T., and Eichele, G. (1990). Polarizing activity and retinoid synthesis in the floor plate of the neural tube. Nature 345, 819-822. Wanek, N., Gardiner, D. M., Muneoka, K., and Bryant, S. V. (1991). Conversion by retinoic acid of anterior cells into ZPA cells in the chick wing bud. Nature 350, 8183. Warkany, J., and Schraffenberger, S. (1946). Congenital malformations induced in rats by maternal vitamin A deficiency. I. Defects of the eye. Arch. Ophthalmol. 35, 150169. Webster, W. S., Johnston, M. C., Lammer, E. J., and Sulik, K. K. (1986). Isotretinoin embryopathy and the cranial neural crest: An in vivo and in vitro study. J. Craniofm. Gen. Dev. Biol. 6, 211-222. Wei, L.-N., Chen, G.J., Chu, Y.-S., Tsao, J.-L., and Nguyen-Huu, M. C. (1991). A 3kb sequence from the mouse cellular retinoic-acid-binding protein gene upstream region mediates spatial and temporal LacZ expression in transgenic mouse embryos. Development 112, 847-854. Widelitz, R. B., Jiang, T.-X., Murray, B. A,, and Choung, C.-M. (1993). Adhesion molecules in skeletogenesis: 11. Neural cell adhesion molecules mediate precartilaginous mesenchymal condensations and enhance chondrogenesis. J. Cell. Physiol. 156, 399-41 1. Wilson, J . G., and Warkany, J. (1948).Malformations in the genito-urinary tract induced by maternal vitamin A deficiency in the rat. A m . J . Anat. 83, 357-407. Wilson, J. G., and Warkany, J . (1949). Aortic-arch and cardiac anomalies in the offspring of vitamin A deficient rats. A m . J . Anat. 85, 113-155. Wilson, J . G., Roth, C. B., and Warkany, J. (1953). An analysis of the syndrome of malformations induced by maternal vitamin A deficiency. Effects of restoration of vitamin A a t various times during gestation. A m . J . Anat. 92, 189-217. Yang Yen, H.-F., Zhang, X.-F., Graupner, G., Tzukerman, M., Karin, M., and Pfahl, M. (1991). Antagonism between the retinoic acid receptors an AP-1: Implications for tumor production and inflammation. New Biol. 3, 1216-1219. Yasuda, Y., Konishi, H., Hihara, T., and Tanimura, T. (1987). Developmental anomalies induced by all trans retinoic acid in fetal mice: 11.Induction of abnormal epithelium. Teratology 35, 355-366.
RETINOIDS AND MOUSE EMBRYOS
457
Yu,V. C., Delsert, C., Andersen, B., Holloway, J. M., Devary, 0. V.,Naar, A. M., Kim, S. Y.,Boutin, J.-M., Glass, C. K., and Rosenfeld, M. G. (1991). RXRp: A coregulator that enhances binding of retinoic acid, thyroid hormone, and vitamin D receptors to their cognate response elements. Cell 67, 1251-1266. Zechel, C., Shen, X.-Q., Chambon, P., and Gronemeyer, H. (1994a). Dimerization interfaces formed between the DNA binding domains determine the cooperative binding of RXR/RAR and RXRlTR heterodimers to DR5 and DR4 elements. EMBO J. 13, 1414-1424. Chen, Z.-P., Chambon, P., and Gronemeyer, H. Zechel, C., Shen, X.-Q., Chen, J.-Y., (1994b). The dimerization interfaces formed between the DNA binding domains of RXR, RAR and TR determine the binding specificity and polarity of the full-length receptors to direct repeats. EMBO J. 13, 1425-1433. Zelent, A,, Krust, M., Petkovich, M., Kastner, P., and Chambon, P. (1989). Cloning of murine a and p retinoic acid receptors and a novel receptor y predominantly expressed in skin. Nature 339, 714-717. Zelent, A,, Mendelsohn, C., Kastner, P., Krust, A,, Gamier, J.-M., Ruffenach, F., Leroy, P., and Chambon, P. (1991). Differentially expressed isoforms of the mouse retinoic acid receptor p are generated by usage of two promoters and alternative splicing. EMBO J . 10,71-81. Zhang, X.-K., Salbert, G., Lee, M. O., and Pfahl, M. (1994). Mutations that alter ligandinduced switches and dimerization activities in the retinoid X receptor. Mol. Cell. Biol. 14, 4311-4323. Zhang, X., Lehmann, J., Hoffmann, B., Dawson, M. I., Cameron, J., Graupner, G., Hermann, T., D a n , P., and Pfahl, M. (1992a). Hornodimer formation of retinoid X receptor induced by 9-cis retinoic acid. Nature 358, 587-591. Zhang, X.-K., Hoffmann, B., "ran,P. B.-V., Graupner, G., and Pfahl, M. (1992b). Retinoid X receptor is a n auxillary protein for thyroid hormone and retinoic acid receptors. Nature 355, 44-446. Zhang, X.-K., "ran, P. B.-V., and Pfahl, M. (1991). DNA binding and dimerization determinants for thyroid hormone receptor a and its interaction with a nuclear protein. Mol. Endocrinol. 5, 1909-1920. Zhou, L., Otulakowski, G., Pang, J., Munroe, D. G., Capetola, R. J., and Lau, C. (1992). Quantitation of human cellular retinoic acid-binding protein I1 (CRABP-11) RNA from cultured human skin fibroblast cells and human biopsies treated with retinoic acid. Nucleic Acids Res. 20, 6215-6220.
This Page Intentionally Left Blank
Subject Index
A
other actions related to neuronal atrophy, 388-391 P-Adrenergic receptor, overexpression, cancer cell growth, 138 P-Adrenergic receptor kinase, 194 activation, 206 amino acid sequences, 198-199 cloning, 196-199 expression, 195 in olfactory epithelium, 211 phosphorylation by and brain Ply-subunits, 203-204 pleckstrin homology domain, 203-204 receptor specificity, 208-212 thrombin receptor regulation, 211 Adrenocortical cells, CAMP-mediated positive growth control, 89 Anterior pituitary gland, gene transcription, 37 Antidepressants, modification of stressinduced changes in hippocampus, 377-380 Antiestrogens, mechanism of action, 277-280 Arginine vasotocin, 249-250 P-Arrestin, expression, 195 Arrestins, 213-226 amino acid sequences, 214-215 cloning, 213-216 C-terminal, 220-222 homologues, 213-214 molecular properties, 218 nonvisual, binding to receptors, 223224 polypeptide variants, 216-217 receptor specificity, 224-226 retinal, binding to rhodopsin, 219223 tissue, cellular, and subcellular localization, 217-219
Actin-containing filaments, CAMPdependent disruption, 125-126 Activating transcription 1 gene transcription, 6-8 oncogenic forms, 42-43 Activation functions, transcriptional, estrogen receptors, 275-276 Activator protein 1 estrogen effects, 271 retinoid receptor-mediated inhibition, 440-441 Adenosine A2 receptor, cancer cell growth, 138 Adenosine triphosphate, in nucleocytoplasmic shuttling, 327 Adenylate cyclase-controlling receptors, heterotypic expression, 138-139 Adenylate cyclase toxins, cyclic AMP elevation, 69 Adrenal cortex, steroid hormone biosynthetic pathway, 341-343 Adrenal gland-specific protein, cytochrome P-450 steroid hydroxylases expression, 354-355 Adrenal steroid receptor, subtypes, role in structural and neurochemical plasticity, 387-388 Adrenal steroids, 371-396 actions on neurotrophin mRNA levels, 39 1 cognitive function, 395 hippocampal formation response, 372373 hippocampal neuronal atrophy, 374377 long-term potentiation, 383-387 neuronal birth and death in dentate gyrus, 380-383 459
460
INDEX
Axon, localized transcripts, vasopressin and oxytocin, 244-246
Basic fibroblast growth factor, role in hippocampal neuronal survival, 389, 391 BDNF, mRNA, expression in hippocampus, 389-390 B lymphocytes, mitogenic activation, inhibition by cyclic AMP, 80 Bone cells, CAMP-mediated positive growth control, 88 Brain, see also Adrenal steroids gene transcription, 37-39 steroid hormone biosynthetic pathways, 345 8-Bromo-cAMP, steroid hormone receptors activation, 302
C Calcium, interdependence with CAMP, 107 Calreticulin, retinoid receptors function inhibition, 418 CAMP, 1-46 functions, 61 growth control, cytoskeleton change role, 122-127 cell shape, 123 mitogenic pathways, 126-127 reverse transformation, 125 growth stimulation by, mechanisms, 141-142 interdependence with calcium, 107 negative control of cell cycle progression, 73-83 cholera toxin and forskolin, 80-81 early work, 73-75 fibroblast, 75-77 immune system, 78-79 inhibition of G,-mitosis transition, 81-82 prophase block of meiosis in oocytes, 81 vascular endothelial and smooth muscle cells, 77-78
negative modulation of cancer cell growth, 127-133 8-CI-cAMP, 131-132 escape from, 133-135 hypotheses, 128-130 oncogene expression inhibition, 132133 type I1 PKA, 129-131 positive control of cell cycle progression, 83-118, 143 adrenocortical cells, 89 bone cells, 88 cross-signaling between mitogenic stimulations, 106-108 gene expression, 99-105 induction of its own mitogenic pathway, 108-118 mammary epithelial cells, 87-88 melanocytes, 90 mitogenic pathways, 105-118 nonepithelial continuous cell lines, 91-92 ovarian follicular granulosa cells, 88-89 peripheral nervous system, 90-91 positive regulation, 93-95 protein kinase activation, 95-98 protein phosphorylation, 98-99 somatotrophs, 87 synergism with other mitogenic factors, 92-93 thymocytes, 91 thyrocytes, 85-87 probes, 66-72 adenylate cyclase toxins, 69 cholera toxin, 67 cyclic AMP analogs, 69-71 forskolin, 67 genetic tools, 71-72 microinjection of purified protein kinase A, 71 pertussis toxin, 67-68 relationship between growth and differentiation controls, 118-122 roles in cell proliferation, 141-142 signaling cascade, oncogenes related to, 136-140 signal transduction pathway dependence on, 2-5 as tumor promoter, 135-136
INDEX
CAMPcascade, role, 71 assessment tools, 68 CAMP-mediated positive growth control, 85 CAMP-PKA signaling cascade, cancer cell growth and, 139-140 cAMP response element-binding protein alternatively spliced exons, 45 anterior pituitary, 37 autoregulation gene expression, 31-32 network, 40-42 genes, 21-31 alternative exon splicing, 27-28 exons encoding functionally distinct domains, 22-25 expression, in testes, 33-36 transcription, 6-8, 45 inactive and transrepressor isoforms, 25-26 oncogenic forms, 42-43 other transactivational domains, 17-21 phosphorylation, 13-14 possible role in memory, 39-40 Q1 and QZ regions, 18 role in brain, 37-39 structure, 12-13 unphosphorylated, gene expression repression, 28-31 cAMP response element modulator alternatively spliced exons, 45 autoregulation gene expression, 31-32 network, 40-42 exon-deleted repressor isoform, 31 genes, 21-31 alternative exon splicing, 27-28 exons encoding functionally distinct domains, 22-25 expression, in testes, 36-37 transcription, 6-8 oncogenic forms, 42-43 repressor isoforms, 25-27 role in brain, 37-39 cAMP response elements, gene transcription, 8-12 Cancer cells, growth, cyclic AMP and, 122-140 negative modulation, 127-133 escape from, 133-135
461
oncogenes related to cyclic AMP cascade, 136-140 as tumor promoter, 135-136 Casein kinase 11, steroid hormone receptor phosphorylation, 294 cdks, steroid hormone receptor phosphorylation, 294 Cell adhesion molecules, retinoids and, 440 Cell cycle controls, 63-66 GI phase, estrogen effects, 271 Cell lines continuous early work, 73-75 nonepithelial, CAMP-mediated positive growth control, 91-92 muscle, nonfusing, inhibition by cholera toxin and forskolin, 80-81 Chicken ovalbumin upstream promoter transcription factors cytochrome P-450steroid hydroxylases expression, 358 retinoid receptors function inhibition, 417-418 Cholera toxin cyclic AMP induction, 67 inhibition of nonfusing muscle cell lines, 80-81 Circadian rhythms, regulation, 38-39 Cloning arrestins, 213-216 G protein-coupled receptor kinases, 196-199 nonapeptide receptors, 252-253 Cognitive function, adrenal steroids, 395 Cognitive performance exogenous glucocorticoid treatment effects, 394 stress effects, 391-393 Cortisone, hippocampal neuronal atrophy, 374-377 CREB-binding protein, structure, 19-20 Cytochrome P-450steroid hydroxylases cell-selective expression, 347-355 adrenal gland-specific protein, 354355 other regulators, 353-355 placenta-specific transcriptional activator, 354
462
INDEX
SF-1, 348-353 expression, future directions, 360-362 hormone-regulated expression, 355359 COUP-TF, 358 CRE-binding protein, 355-356 NGFI-B, 357-358 Pbxl, 358-359 SF-1, 356-357 Spl, 359 steroidogenic, 346-347 Cytoskeleton, role of changes in growth control by cyclic AMP, 122-127
D Dendrite, transcripts, vasopressin and oxytocin, 244 Dentate gyrus, neuronal birth and death, adrenal steroids effects, 380-383 Depression, HPA axis deregulation, 393394 Desensitization, homologous, model, 194195 Dexamethasone suppression test, 393394 Differentiation control by cyclic AMP, 118-122 models, 121 Dimerization, retinoid receptors, 412 activity and, 415-417 DNA binding domain, estrogen receptors, 275 binding site, retinoid receptors, configuration and sequence, 420422 binding steroid hormone receptors, regulation by phosphorylation, 296-297 DNA-PK, steroid hormone receptor phosphorylation, 294-295 Dopamine, steroid hormone receptors activation, 302-303
E Embryogenesis, vitamin A-deficient diets, 403-404
Embryonal long terminal repeat-binding protein, functional significance, 351-352 Epidermal growth factor, cross-coupling with other signaling pathways, 280281 Estrogen autocrine hypothesis, 270 physiological responses to, 268-271 Estrogen receptors, 267-282 antiestrogen action mechanism, 277-280 cross-coupling with other signaling pathways, 280-282 DNA-binding domain, 275 hormone binding and receptor dimerization, 272-274 human, phosphorylation sites, 298-299 intracellular localization, 272-273 mechanism of action, 272-273 specific gene transcription, 276-277 target gene recognition, 274-275 transcriptional activation, 275-276 mechanisms, 269 Eukaryotic cells, cell cycle, 63-66 Excitatory amino acid transmitters, roles, 373
F Fibroblast growth factor, see Basic fibroblast growth factor Fibroblasts, growth inhibition by cyclic AMP, 75-77 Forskolin cyclic AMP pulse induction, 67 inhibition of nonfusing muscle cell lines, 80-81
G Genes c-fos down-regulation by exon-deleted repressor isoform, 31 transcription control, 100-102 c-jun down-regulation by exon-deleted repressor isoform, 31 transcription control, 102-103 c-myc, transcription control, 103-104
463
INDEX
CREB and CREM, autoregulation of expression, 31-32 expression, CAMP-mediated positive growth control, 99-105 Ftz-Fl, 349-350,352,361-362 Hox, regulation by retinoids, 437-439 recognition, estrogen receptors, 274275 retinoid-binding proteins, transgenic studies, 426-427 Gene transcription, 1-46 activating transcription 1, 6-8 anterior pituitary gland, 37 autoregulation of expression of CREB and CREM genes, 31-32 brain, 37-39 CAMP-dependent, CREB-mediated, 3-5 cAMP response element-binding protein, 6-8, 45 autoregulation network, 40-42 cAMP response element modulator, 68
autoregulation network, 40-42 CAMPresponse elements, 8-12 future directions, 43-46 homodimer and heterodimer combinations, 43-44 hypothalamus, 37-39 pineal gland, 37-39 regulation by signal transduction pathways, 3-4 testes, 33-37 transcriptional transactivation, gene transcription CREB domains, 17-21 interactions between signaling pathways, 15-16 kinases, 12-16 phosphatases, 17 Glucocorticoid receptor activation, 322 DNA-binding domain and nuclear localization signal sequence, 318320 nucleocytoplasmic shuttling, 325-326, 328 phosphorylation sites, 299 subcellular localization, indirect immunofluorescence, 324 Glucocorticoids
effect on vasopressin gene expression, 241 exogeneous treatment, effects on cognitive performance, 394 impact on hippocampal formation, 373 G protein, ply-subunits, 206-207 G protein-coupled receptor kinases amino acid sequences, 198-199 characteristics of members, 212-213 cloning, 196-199 intramolecular autophosphorylation, 207 membrane localization, 202-205 molecular properties, 200-201 phosphorylation of receptors, 207-208 receptor specificity, 208-212 regulation, 205-208 tissue localization, 199-202 G protein-coupled receptors, see also Arrestins as protooncogenes, 139 Growth, modulation by CAMP,see cAMP Growth factors estrogen effects on, 270 membrane receptors, 60 roles, 65 Growth-promoting factors, cyclic AMP mediation, vertebrate cells, 84
H Heterokaryons, transient, 324-326 Hippocampal formation response to adrenal steroids, 372-373 roles, 373 Hippocampus BDNF mRNA expression, 389-390 gene expression, adrenal steroids receptor mediation, 387-388 neuronal atrophy, adrenal steroids, 374-377 neuronal survival, basic fibroblast growth factor role, 389,391 plastic changes, regulation and intracellular signaling mechanisms, 389 stress-induced changes, antidepressant effect, 377-380 Hormone binding, steroid hormone
464
INDEX
receptors, regulation by phosphorylation, 296 Hormone response elements, bound by RARs and RXRs, 420-421 Hormones, growth stimulation, 119-120 HPA axis, deregulation, in depression, 393-394 Hypothalamus gene transcription, 37-39 neurons, somatic recombination between vasopressin and oxytocin genes, 247-249
I ICI 182780, antiestrogen action, 279-280 Immune system, cyclic AMP effects, 7880 Interferon, effect on cyclic AMP levels, 80 Intracellular signals, role, 61-63 Isotocin, 249-250
K Kinases, transcriptional transactivation mechanisms, 12-16
L Ligand, availability and retinoid receptors activity, 413-415 Long terminal repeat-binding protein, embryonal, functional significance, 351-352 Long-term potentiation, adrenal steroids, 383-387
Mammary epithelial cells, CAMPmediated positive growth control, 87-88
McCune-Albright syndrome, 137-138 Meiosis, in oocytes, CAMP-dependent prophase block, 81 Melanocytes, CAMP-mediated positive growth control, 90 Memory, possible CREB role, 39-40
Mesotocin, 249-250 Metarhodopsin 11, phosphorylation, 194 Microvascular endothelial cells, growth inhibition by cyclic AMP, 77 Mitogen-activated protein kinases phosphorylation, 110- 111 steroid hormone receptor phosphorylation, 294 Mitogenic factors, synergism with cyclic AMP, 92-93 Mitogenic pathways CAMP-dependent, 122 in thyroid, 143-144 CAMP-dependent and -independent activation state of protein kinasedependent regulatory networks, 110 c-fos and c-jun expression, 111-113 c-myc RNA levels, 113-114 cross-signaling, 106-108 late gene expression pattern, 114 MAP kinase phosphorylation, 110111 time dependence of commitment to DNA synthesis, 114-117 Mitogenic signals, CAMP-independent, transfer to nucleus, 61-62 Mitosis, transition, CAMP-dependent inhibition, 81-82 Morphogen, retinoids as, 439-440 Muscle cell line, nonfusing, inhibition by cholera toxin and forskolin. 80-81
N Nerve growth factor inducible-B, cytochrome P-450 steroid hydroxylases expression, 357-358 Neurochemical plasticity, adrenal steroid receptor subtypes, role, 387-388 Neurohypophysial hormone receptors sequence relationship, 257-258 tissue distribution, 255-257 Neuronal atrophy, adrenal steroids actions, 388-391 Nonapeptide receptors, 251-258 homologies, 253-254 molecular cloning, 252-253 phylogeny, 257-258 physiological roles, 251
INDEX
structure, 253-255 tissue distribution, 255-257 Nuclear import, steroid receptors, 317323 mechanisms, 320-323 nuclear localization signal sequence, 317-320 Nuclear receptors, domain structure, 268 Nucleocytoplasmic shuttling, see Steroid hormone receptors
0 Oncogenes gip2, 140 related to CAMPsignaling cascade, 136-140 Oncoproteins, CREB, CREM, and ATF-1, 42-43 Oocytes, meiosis, cyclic AMP-dependent prophase block, 81 Ovarian follicular granulosa cells, CAMPmediated positive growth control, 88-89 Ovary, steroid hormone biosynthetic pathways, 344 Oxytocin, 235-258, see also Nonapeptide receptors gene expression and regulation gene promoters, putative regulatory elements, 241-242 transgenic animals, 239-240 gene family, 249-251 mRNA in dendrites and axons, 242246 axonally localized transcripts, 244246 dendritic transcripts, 244 extrasomal domains of neurons, 242-243 somatic recombination with vasopressin genes in hypothalamic neurons, 247-249
P Peripheral nervous system, CAMPmediated positive growth control, 90-91
465
Pertussis toxin, prolonged cyclic AMP elevation, 67-68 Phorbol esters, tumor gmwth inhibition, 128 Phosphatases, transcriptional transactivation mechanisms, 17 Phosphoproteins, steroid receptors as, 291-292 Phosphorylation, 289-307 MAP kinase, 110-111 protein, CAMP-mediated positive growth control, 98-99 rhodopsin, 205 steroid hormone receptors, 291-296 Pineal gland circadian rhythm generation, 38-39 gene transcription, 37-39 Placenta, steroid hormone biosynthetic pathways, 344-345 Placenta-specific transcriptional activator, cytochrome P-450 steroid hydroxylases expression, 354 Plasticity, neurochemical, adrenal steroid receptor subtypes, role, 387-388 Polymerase 11, basal transcriptional complex, coupling to CREB transactivation, 19-21 Posttranslational modifications, retinoid receptors, 419-420 pp90=8*, steroid hormone receptor phosphorylation, 295 Prereplicative phase, concept, 65 Progesterone receptor chicken ligand-independent activation, 302-303 phosphorylation sites, 298 human, activation, 303 Prophase block, cyclic AMP-dependent, meiosis in oocytes, 81 Protein kinase activation, CAMP-mediated positive growth control, 95-98 mitogen-activated, see Mitogenactivated protein kinases Protein kinase A phosphorylation by, 4-5, 12-13 signaling cascade, 107-108 steroid hormone receptor phosphorylation, 295 Protein kinase C signaling cascade, 107-108
466
INDEX
steroid hormone receptor phosphorylation, 295 Proteins adapter, CREB transactivation, coupling to basal polymerase I1 transcriptional complex, 19-21 A kinase-anchoring, 4 bZIP, 6-8 criterion for establishing families, 9-10 cross-family heterodimers, 10-12 differential DNA-binding, 9-10 CRE-binding, cytochrome P-450 steroid hydroxylases expression, 355-356 heat-shock, 322-323 interactions with steroid hormone receptors, 330-331 karyophilic, nuclear localization signal sequences, 317-320 myelin, cyclic AMP proliferation stimulation, 120-121 Pbxl, cytochrome P-450 steroid hydroxylases expression, 358-359 phosphorylation, CAMP-mediated positive growth control, 98-99 -protein interaction, steroid hormone receptors, regulation by phosphorylation, 297 Protein-tyrosine kinase, pathways, 61 Protooncogenes, nuclear, expression, CAMP-mediated positive growth control, 99-104
R Ras-MAPK pathway, 14-16 Retinoic acid, 404 isomers, therapeutic use in humans, 404 teratogenesis, 404 Retinoid-binding proteins, 422-427 AP-1 inhibition and, 440-441 cellular, 424-425 cytoplasmic, 405 expression patterns and genetic analysis of function, 427-436 anomolies in single and double null mutants, 430-431 RARa, 433 RARP, 429-430
RAR-y, 430,432-433 RARs, 429-433 RXRs, 433-435 genes, transgenic studies, 426-427 Retinoid receptors AIB domain, 407-408 background, 405-407 C or DNA binding domain, 408-411 DEF domains, 411-413 dimerization, 412 activity and, 415-417 factors affecting activity dimerization, 415-417 DNA binding site, configuration and sequence, 420-422 inhibitors of function, 417-418 ligand availability, 413-415 mediation of transactivation, 418419 posttranslational modifications, 419420 function, 408-409 functional domains, 405-406 heterodimers, 409-410 nuclear localization sequence, 411 protein-protein contacts, 409 transactivation, 412 Retinoids, 403-442 cell adhesion molecules and, 440 future directions, 441-442 Hox gene regulation, 437-439 a s morphogen, 439-440 transforming growth factor p family regulation, 439 Retinoid X receptors, 405 dimerization propensity, 416-417 expression patterns and genetic analysis of function, 433-435 a s heterodimers, 434-435 homodimers, 416 Reverse transformation, CAMPdependent, 125 Rhodopsin activated, 194 arrestins specificity, 224-225 binding to retinal arrestin, 219-223 phosphorylation, 205 Rhodopsin kinase C-terminal domain, 202-203 intramolecular autophosphorylation, 207
467
INDEX
RNA, messenger BDNF, expression in hippocampus, 389-390 in dendrites and axons, vasopressin and oxytocin, 242-246 neurohypophysial hormone receptors, 255-257
S Serotonin, suppression by chronic stress, 377-380 Sexual differentiation, male, 350-351 Signal transduction pathways activation by steroids, 303-305 CAMP-dependent, 2-5 gene transcription regulation, 3-4 modulators, steroid hormone receptors activity regulation, 301-303 Somatotrophs, CAMP-mediated positive growth control, 87 Steroid hormone receptors equilibrium subcellular distribution, 326,329 heat-shock protein interactions, 33033 1 nucleocytoplasmic shuttling, 323-333 controversies, 327-329 glucocorticoid receptor subcellular localization, 324 historical perspective, 323-324 new model, 329-332 transient heterokaryons, 324-326 phosphorylation identification as phosphoproteins, 291-292 in viuo phosphorylation site identification, 292-293 kinases and, 293-296 regulation of activity by modulators of signal transduction pathways, 301-303 regulation by phosphorylation in vitro studies, 296-297 mutational analysis, 298-301 signal transduction pathway activation, 303-305 structure and function, 290-291 subcellplar localization, 316-323 historical perspective, 316-317
mechanisms of nuclear import, 320323 nuclear import, 317-323 nuclear localization signal sequences, 317-320 subcellular trafficking pathway, 315316 Steroid hormones, 289-307 action, 289-290 biosynthesis, 340-347 adrenal cortex, 341-343 brain, 345 wary, 344 pathways, 340-346 placenta, 344-345 regulation, 346-347 testis, 343-344 uterus, 345-346 model of action, 305-307 Steroidogenic factor 1, role in cytochrome P-450steroid hydroxylases expression, 348-353, 356-357, 360 Stress, see also Adrenal steroids chronic, CA3c pyrimidal neuronal atrophy, 377 effects on cognitive performance, 391-393 long-term potentiation, 383-387 hippocampal neuronal atrophy, 374377 induced, changes in hippocampus, an tidepressant effect, 377-380 interpretation by brain, 371-372 Structural plasticity, adrenal steroids receptor subtypes, role, 387-388
T Tamoxifen, antiestrogen action, 277-279 Testis gene transcription, 33-37 steroid hormone biosynthetic pathways, 343-344 Thymocytes, CAMP-mediated positive growth control, 91 Thyrocytes CAMP-mediated positive growth control, 85-87 suppression of differentiation expression, 120
468
INDEX
Thyroid receptors, phosphorylation sites, 299-300 Thyroid stimulating hormone, CAMPmediated positive growth control, 85-87 Tianeptine, effect on stress-induced atrophy, 378-379 T lymphocytes, mitogenic activation, inhibition by cyclic AMP, 78-80 Transactivation retinoid receptor-mediated, 418-419 retinoid receptors, 412 Transcription, specific gene, estrogen receptors, 276-277 Transcriptional activation, estrogen receptors, 275-276 Transcription factors CAMP-responsive, 6-8 formation dimers, 44 S p l , cytochrome P-450 steroid hydroxylases expression, 359 Transforming growth factor p family, regulation by retinoids, 439 Tropomyosins, repression, 124 Tumors, cyclic AMP as promoter, 135136
U Uterus, steroid hormone biosynthetic pathways, 345-346
ISBN 0-22-709852-8 90065
V Vascular smooth muscle cells, growth inhibition by cyclic AMP, 78 Vasopressin, 235-258, see also Nonapeptide receptors gene expression and regulation gene promoters, putative regulatory elements, 240-241 transgenic animals, 237-239 gene family, vertebrate, 249-251 mRNA in dendrites and axons, 242246 axonally localized transcripts, 244246 dendritic transcripts, 244 extrasomal domains of neurons, 242-243 somatic recombination with oxytocin genes in hypothalamic neurons, 247-249 Vimentin, CAMP-dependent phosphorylation, 126 Vitamin D receptors, phosphorylation sites, 300-301