International Review of Cytology, Volume 77

INTERNATIONAL

REVIEW OF CYTOLOGY VOLUME 77

ADVISORY EDITORS DONALD G.MURPHY H. W. BEAMS ROBERT G.E. MURRAY HOWARD A. BERN RICHARD NOVICK GARY G.BORISY ANDREAS OKSCHE PIET BORST MURIEL J. ORD BHARAT B. CHATTOO STANLEY COHEN VLADIMIR R. PANTIC RENE COUTEAUX W. J. PEACOCK DARRYL C. REANNEY MARIE A. DIBERARDINO LIONEL I. REBHUN CHARLES J. FLICKINGER JEAN-PAUL REVEL OLUF GAMBORG M. NELLY GOLARZ DE BOURNE JOAN SMITH-SONNEBORN WILFRED STEIN YUKIO HIRAMOTO HEWSON SWIFT YUKINORI HIROTA DENNIS L. TAYLOR K. KUROSUMI TADASHI UTAKOJI GIUSEPPE MILLONIG ARNOLD MITTELMAN ROY WIDDUS ALEXANDER L. YUDIN AUDREY MUGGLETON-HARRIS

INTERNATIONAL

Review of Cytology EDITED BY

G. H. BOURNE

J. F. DANIELLI

St. George’s University School of Medicine St. Georges, Granada West Indies

Worcester Polytechnic Institute Worcester, Massachusetts

ASSISTANT EDITOR K. W. JEON Department of Zoology University of Tennessee Knoxville, Tennessee

VOLUME77 1982

ACADEMIC PRESS

A Subsidiary of Harcourt Brace Jovanovich, Publishers

New York London Pans San Diego San Francisco Siio Paulo Sydney Tokyo Toronto

COPYRIGHT @ 1982, BY ACADEMIC PRESS, INC. ALL 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.

111 Fifth Avenue, New York. New York 10003

United Kingdom Edition published by ACADEMIC PRESS, INC. ( L O N D O N ) LTD. 24/28 Oval Road, London NW17DX

LIBFURY OF CONORESS CATALOG CARDNUMBER:52-5203 ISBN 0-12-364477-1 PRINTED IN THE UNITED STATES OF AMERICA

82 83 84 85

9 8 7 6 5 4 3 2 1

Contents CONTRIBUTORS . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

ix

Calcium-Binding Proteins and the Molecular Basis of Calcium Action LINDAJ . VAN ELDIK.JOSEPHG . ZENDEGUI. DANIELR . MARSHAK. A N D D. MARTINWATERSON I. I1 . Ill . IV . V. VI .

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . Proteins Containing y-Carboxyglutamic Acid . . . . . . . . . . . . . Concanavalin A and Calcium-Binding Lectins . . . . . . . . . . . . . Calcium-Binding Hydrolytic Enzymes . . . . . . . . . . . . . . . . . Calcium-Modulated Proteins . . . . . . . . . . . . . . . . . . . . . Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

i 4

13 18

25 46 48

Genetic Predisposition to Cancer in Man: In Vitro Studies LEVYKOPELOVICH

I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . I1 . The Experimental System . . . . . . . . . . . . . . . . . . . . . . 111. Studies on ACR Cell Cultures . . . . . . . . . . . . . . . . . . . . IV. Cell Culture Studies on Autosomal Dominant Syndromes (Other Than ACR) and Chromosome Instability Syndromes . . . . . . . . . . . . . V . On the Question of Tumor Promotion . . . . . . . . . . . . . . . . . VI . Genetic Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . VII . Is Genetic Predisposition to Cancer an Autosomal Dominant Trait . . . . VIII . On the Question on Cancer Prognosis and Cancer Control . . . . . . . IX. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

63 65 66

73 75 78

79 81 83 84

Membrane Flow via the Golgi Apparatus of Higher Plant Cells DAVIDG . ROBINSON A N D UDO KRISTEN I . Introduction . . . . . . . . . . . . . . . I1 GA Structure and Biochemistry . . . . . 111. Sites of Synthesis . . . . . . . . . . . . . IV . Secretion Kinetics and Membrane Turnover

.

V

............. . . . . . . . . . . . . . . ............. . . . . . . . . . . . . . .

89 91

102 108

vi

CONTENTS

V. Means and Ends . . . . . . . . . . . . . . VI . Membrane Recycling and the Golgi Apparatus VII Concluding Remarks . . . . . . . . . . . . References . . . . . . . . . . . . . . . . .

.

. . . .

. . . .

. . . .

......... . . . . . . . . . . ......... .........

112 114 119 120

Cell Membranes in Sponges

.

WERNER E G . MULLER

I. I1. 111. IV . V. VI . VII . VIII . IX . X.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cell Types in Sponges . . . . . . . . . . . . . . . . . . . . . . . . Intercellular Matrix . . . . . . . . . . . . . . . . . . . . . . . . . Morphology of Cell Contact . . . . . . . . . . . . . . . . . . . . . Primary Cellular Recognition. . . . . . . . . . . . . . . . . . . . . Secondary Cellular Recognition . . . . . . . . . . . . . . . . . Cell Movement and "Sorting Out" . . . . . . . . . . . . . . . . Cell Interactions in the Immune Response . . . . . . . . . . . . . Recognition of Symbionts . . . . . . . . . . . . . . . . . . . . . . Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

129 130

132 135

. . . . . .

137 145 161 168

173 17.5 176

Plant Movements in the Space Environment DAVIDG . HEATHCOTE

I. I1. Ill . IV .

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 Characteristics of the Space Environment . . . . . . . . . . . . . . . 184 Experimental Constraints Imposed by Space Flight . . . . . . . . . . . 189 The Study of Plant Movements in the Space Environment . . . . . . . . 190 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203

Chloroplasts and Chloroplast DNA of Acetabularia mediterrunea: Facts and Hypotheses ANGELALUTTKEA N D SILVANO BONOTTO 1. Introduction

I1. I11. IV . V.

............................

Chloroplast Morphology and Ultrastructure . . . . . . . . . . . . Chloroplast DNA . . . . . . . . . . . . . . . . . . . . . . . . . . Chloroplast Gene Products . . . . . . . . . . . . . . . . . . . . . . Summary and Outlook . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. .

205 207 218 235 237 238

vii

CONTENTS

Structure and Cytochemistry of the Chemical Synapses STEPHAN MANALOVAND WLADIMIR OVTSCHAROFF

I . Introduction . . . . . I1 . Presynaptic Part . . . 111. Synaptic Cleft . . . . IV . Postsynaptic Part . . . V . Concluding Comments References . . . . . .

....................... ....................... ....................... . . . . . . . . . . . . . . . . . . . . . . . .......................

243 244 270 272 277 278

. . . . . . . . . . . . . . . . . . . . . . .

INDEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CONTENTS OF RECENT VOLUMES A N D SUPPLEMENTS . . . . . . . . . . . . .

.

285 289

This Page Intentionally Left Blank

Contributors Numbers in parentheses indicate the pages on which the authors’ contributions begin.

SILVANO BONOTTO(205), Department of Radiobiology, Nuclear Research Centre, 2400 Mol, Belgium

DAVIDG . HEATHCOTE (183), Department of Plant Science, University College, Cardiff CFI IXL, Wales LEVYKOPELOVICH (63), Memorial Sloan-Kettering Cancer Center and Cornell University Graduate School of Medical Sciences, New York, New York 10021

UDO KRISTEN(89), Institut f u r Allgemeine Botanik, Universitat Hamburg, 2000 Hamburg 52, Federal Republic of Germany ANGELALUTTKE(205), Institute for Developmental Physiology, University of Cologne, Cologne, Federal Republic of Germany STEPHANMANALOV(243), Regeneration Research Laboratory, Bulgarian Academy of Sciences, Sofia, Bulgaria DANIELR. MARSHAK (l), Laboratory of Cellular and Molecular Physiology, Howard Hughes Medical Institute, and Department of Pharmacology, Vanderbilt University School of Medicine, Nashville, Tennessee 37232 WERNERE. G . MULLER(129), Institut fur Physiologische Chemie, Abteilung “Angewandte Molekularbiologie,” Universitat, Duesbergweg, 6500 Mainz, Federal Republic of Germany WLADIMIR OVTSCHAROFF (243), Department of Anatomy, Histology and Embryology, Medical Academy, Sofia, Bulgaria DAVIDG . ROBINSON (89), Abteilung Cytologie des Pflanzenphysiologischen Instituts der Universitat Gottingen, 3400 Gottingen, Federal Republic of Germany ix

X

LIST OF CONTRIBUTORS

LINDA J. VAN ELDIK(l), Laboratory of Cellular and Molecular Phy.sio1ogy, Howard Hughes Medical Institute, and Department of Pharmacology, Vanderbilt University School of Medicine, Nashville, Tennessee 37232 D. MARTINWATTERSON (l),Laboratory of Cellular and Molecular Physiology, Howard Hughes Medical Institute, and Department of Pharmacology, Vanderbilt University School of Medicine, Nashville, Tennessee 37232

JOSEPH G. ZENDEGUI (l), Laboratory of Cellular and Molecular Physiology, Howard Hughes Medical Institute, and Department of Pharmacology, Vanderbilt University School of Medicine, Nashville, Tennessee 37232

INTERNATIONAL REVIEW OF CYTOLOGY. VOL. 77

Calcium-Binding Proteins and the Molecular Basis of Calcium Action LINDAJ. VANELDIK,JOSEPH G. ZENDECUI,DANIELR. MARSHAK, AND D. MARTINWAITERSON Laboratory of Cellular and Molecular Physiology, Howard Hughes Medical Institute, and Department of Pharmacology, Vanderbilt University School of Medicine, Nashville, Tennessee I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Proteins Containing y-Carboxyglutamic Acid . . . . . . . . . . . . . . . . . .

A. Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Synthesis of y-Carboxyglutamic Ac C. Calcium-Binding Characteristics of D. Plasma Proteins Containing y-Carboxyglutamic Acid. . . . . . . . E. Bone Proteins Containing y-Carboxyglutamic Acid . . . . . . . . . F. Other Proteins Containing y-Carhoxyglutamic Acid . . . . . . . . . 111. Concanavalin A and Calcium-Binding Lectins . . . . . . . . . . . . . . . . . IV. Calcium-Binding Hydrolytic Enzymes. ....................... A. Phospholipase A 2 . . . . . . . . ........................ B. Calcium and the Active Site.. . . . . . . . . . . . . . . . . . . . . . . . . . . C. Calcium and Micellar Substrates. . . . . . . . . . .... D. Other Calcium-Binding, Hydrolytic Enzymes . . . . . . . . . . . . . . V. Calcium-Modulated Proteins. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Pdrvalbumin and EF Hand Calcium-Binding Structures . . . . . . B. Calmodulin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Vitamin-D-Dependent Protein and Multiple Calcium-Binding Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. SlOO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ......

............................. ..........

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1

4 5 6 7 9 10 13 13 18 19 20 23 24 25 25 28 39 42 46 48

I. Introduction Calcium is required for the maintenance of optimal growth and functioning of most living organisms. For example, calcium appears to be involved in the mechanism of action of many hormones, drugs, and toxins, regulation of secretion and contraction, digestion of food and absorption of nutrients, formation and maintenance of bone and teeth, regulation of blood clotting and wound healing, functioning of various enzymes and receptors, and regulation of photochemical events in plants and animals. While the molecular mechanisms by which calcium I Copyright 8 1982 by Academic Press. Inc.

2

LINDA J. VAN ELDIK ET AL.

mediates these processes probably involves interactions with a variety of molecules, a large body of evidence has demonstrated the direct involvement of proteins in the molecular mechanisms of calcium action. Further, the recent elucidations of the types of protein structures used in several different mechanisms of calciumlprotein interactions have allowed some correlation of calciumbinding protein structures with calcium-dependent activity. In this review we discuss selected precedents that demonstrate some of the ways in which calcium is bound to a protein and how calcium is involved in protein function. The proteins used as examples are, in general, ones which appear to have a physiological requirement for calcium and ones for which extensive molecular and atomic detail is available (e.g., X-ray crystal structures and amino acid sequences). Four classes of calcium-binding proteins that reflect different uses of calcium in protein function and different ways in which proteins bind calcium ions are reviewed. These four classes are (1) proteins containing ycarboxyglutamic acid, (2) calcium-binding lectins with emphasis on concanavalin A, (3) calcium-binding hydrolytic enzymes with emphasis on phospholipase A,, and (4) calcium-modulated proteins with emphasis on a relatively ubiquitous protein (calmodulin), a tissue-specific protein (S lOO), and a hormoneinduced protein (vitamin D-dependent calcium-binding protein). Proteins which contain y-carboxyglutamic acid are discussed as an example of proteins which bind calcium mainly by bidentate chelation. It is possible that these and other proteins which bind calcium with dissociation constants in the millimolar range, such as the calcium-binding phosphoproteins, may be a class of calcium-binding proteins that exert their calcium-dependent activities extracellularly. The most extensively characterized proteins that contain y-carboxyglutamic acid are the proteins involved in the regulation of blood clotting. The main function of calcium in these proteins appears to be the promotion of protein interaction with extracellular supramolecular structures or surfaces. Variations of this calcium-binding theme may be used by other proteins when interface reactions are the main function of calcium. The role of calcium in the calcium-binding lectin concanavalin A appears to be protein stabilization. In contrast to the simple chelation type of binding found in the y-carboxyglutamic acid-containing proteins, multiple protein ligands are used to bind calcium in lectins. The protein ligands are mostly the oxygens from amino acid side chains, but one oxygen is contributed by a carbonyl group from the peptide backbone. The polyhedron that is formed by wrapping the peptide chain around the calcium has a common edge with an adjoining polyhedron that forms a transition metal-binding site. In thermolysin, a calcium-binding enzyme, there is a pair of calcium-binding sites reminiscent of the calcium:manganese paired-ion sites in concanavalin A. The paired calcium sites in thennolysin appear to enhance the thermal stability of the enzyme. Paired cation binding sites

CALCIUM-BINDING PROTEINS

3

may be a structure used by a protein when calcium is functioning to ’‘lock’’ a particular conformation that enhances protein stability or activity. Phospholipase A, and other hydrolytic enzymes are discussed as examples of the involvement of calcium in substrate binding and enzymic catalysis. The coordination of calcium at the active site of phospholipase A, is by six ligands that form an octahedron. In contrast to concanavalin A and the calcium-modulated proteins, phospholipase A, uses mostly oxygens from peptide carbonyl groups as protein ligands for the active site calcium. At least one side chain oxygen, the carboxylate oxygen of an aspartic acid residue, is also used. Proposed mechanisms for the role of the active site calcium in enzyme function include stabilization of a tetrahedral intermediate generated during conversion of substrate to product. Lower affinity (K,, = 20 mM) calcium-binding sites are also found in phospholipase A,. These latter sites are away from the active site and are important in the conversion of zymogen to enzyme. The lower affinity calcium-binding sites are thought to be in the NH,-terminal portion of the enzyme, a region that is involved in the interaction between the protein and micellar substrates. Such lower affinity calcium-binding sites that appear to stimulate interface reactions remind one of the role of calcium in promoting the interaction of y-carboxyglutamic acid-containing proteins with membranes and phospholipids. Thus, phospholipase A, provides a precedent for a single protein having multiple uses of calcium and possible multiple calcium-binding structures, as well as being a well-characterized system in which ca1cium:protein:lipid interactions are involved in enzyme function. Parvalbumin, calmodulin, vitamin D-dependent calcium-binding protein, and brain S 100 are discussed as selected examples of calcium-modulated proteins. Calcium-modulated proteins reversibly bind calcium with dissociation constants in the nanomolar to micromolar range under physiological conditions. The tertiary structure and activity of these proteins are thought to be modulated by the reversible formation and dissociation of a calcium-protein complex. Because most of these proteins are intracellular and have dissociation constants that span the range of intracellular free calcium concentrations, they are postulated to be the major signal transducers of biological calcium signals. Two types of calciumbinding structures, both of which utilize an a-he1ix:loop:a-helix arrangement of the peptide chain, have been found in this class of proteins. In both of these structures the calcium-binding residues are located in the loop connecting the two a-helices. One type of calcium-binding structure found in calcium-modulated proteins is referred to as an EF hand structure. This name is derived from the calciumbinding structure in parvalbumin that is formed by the E-helix, the F-helix, and the peptide loop connecting these two a-helices. The other type of calciumbinding structure has been found recently in one of the mammalian vitamin D-

4

LINDA 1. VAN ELDlK ET AL.

dependent calcium-binding proteins. This latter structure is similar in general features to the EF hand structure. Both structures are octahedra, but the EF hand structure uses mostly oxygens from amino acid side chains as ligands whereas the vitamin D-dependent protein structure utilizes mostly carbonyl oxygens from the peptide backbone. As a result, the spacing of calcium ligands in the amino acid sequence of the vitamin D-dependent protein is different from that found in an EF hand type of structure. Interestingly, the vitamin D-dependent protein has a second calcium-binding site that utilizes an EF hand structure. One proposal for the role of these two different types of structures in the same protein is that the EF hand structure might be involved in cellular responses to calcium signals while the other site might be occupied by calcium in the resting cell and would, therefore, be mainly involved in the stabilization of protein structure or conformation. The amino acid sequences of two S 100 polypeptides have regions that are homologous with the vitamin D-dependent protein and might also have both of these calcium-binding structures. Calmodulin has four regions of amino acid sequence that appear to only have the sequence requirements for formation of an EF hand structure. Thus, all calcium-modulated proteins examined to date appear to have amino acid sequences that are capable of forming EF hand types of structures but may, in addition, have other calcium-binding structures. The division of calcium-binding proteins into classes based upon limited structural and functional information is a subjective decision. In fact, the entire process of selecting material and writing a review is a highly personal endeavor. A number of calcium-binding proteins are not discussed in this article, and entire articles have dealt with some proteins which are only briefly mentioned here. We hope that by a brief discussion of selected types of calcium-binding proteins we might provide the interested biologist with a limited insight into the complexities as well as the generalities that are emerging in our knowledge of calcium-binding proteins and the molecular basis of calcium action in cellular processes.

11. Proteins Containing y-Carboxyglutamic Acid

Extensive investigations have been done on a class of calcium-binding proteins containing the novel amino acid 3-amino- 1,1,3-propane-tricarboxylicacid, commonly called 4-carboxyglutamic or y-carboxyglutamic acid (Gla). Gla occurs as a vitamin K-dependent, posttranslational modification of glutamic acid residues. This modified amino acid is present in at least seven plasma proteins and several tissue-associated proteins. The presence of Gla in the plasma proteins apparently confers a number of physical and functional properties including the ability to bind calcium with a dissociation constant in the millimolar range and the ability to bind phospholipid in a calcium-dependent manner. The function of . Gla in tissue-associatc3d.pFotek.kless clear but may be related to their ability to

CALCIUM-BINDING PROTEINS

5

bind calcium salts as suggested by their presence in extracellular calcium deposits. Prothrombin is the prototypical Gla-containing plasma protein, especially with regard to structural characteristics and calcium- and phospholipid-binding properties. Likewise, BGP (or bone Gla-containing protein) has been more extensively investigated than other tissue-associated proteins containing Gla. The discussion which follows will, therefore, concentrate on these proteins as examplary of their respective subclass of Gla-containing proteins. A. BACKGROUND The discovery of Gla was a result of Henrick Dam’s observation (1935a,b) that vitamin K-deficient chickens displayed bleeding abnormalities due to a decrease in the amount of functional prothrombin in their plasma. Reduced levels of the clotting factors prothrombin, Factor VII, Factor IX,and Factor X were also reported in humans and cows following treatment with vitamin K antagonists such as dicumerol (Owens et al., 195 I ; Aggeler et al., 1952; Briggs et al., 1952; Hougie et al., 1957; Stenflo, 1970). An abnormal form of prothrombin was postulated by Hemker et al. (1963) and later shown by Stenflo ( 1 970) to be present in plasma of the dicumerol-treated animals. The lack of activity of the abnormal prothrombin was then demonstrated (Nelsestuen and Suttie, 1972; Stenflo and Ganrot, 1972) to be due to its inability to bind calcium and phospholipid. Subsequently, the vitamin K-dependent structure responsible for these activities was shown (Magnusson et al., 1974; Nelsestuen et al., 1974; Stenflo et al., 1974) to be y-carboxyglutamic acid. More extensive accounts of the history of vitamin K and vitamin K-dependent proteins can be found in several reviews (Almquist, 1975; Stenflo, 1976; Suttie, 1977, 1978, 1980a,b). The demonstration of y-carboxyglutamic acid in proteins was hampered for years because under conditions of acid hydrolysis used in amino acid analysis Gla is quantitatively decarboxylated to glutamic acid (Hauschka et al., 1975; Hauschka, 1977). Gla was originally identified (Stenflo ef al., 1974) by proton nuclear magnetic resonance spectroscopy and mass spectrometry. Release of stable Gla from peptide linkage can be achieved by alkaline hydrolysis (Hauschka et al., 1975; Hauschka, 1977). The amino acid can then be quantitatively determined using an amino acid analyzer (Hauschka, 1977; Tabor and Tabor, 1977). A number of compounds elute from analyzer columns close to Gla, so the identification of the presumptive Gla peak should be confirmed by its disappearance upon acid hydrolysis (Hauschka, 1977). Several other chemical and chromatographic methods have been described for the detection and quantitation of Gla (Howard and Nelsestuen, 1974, 1975; Zytkovicz and Nelsestuen, 1975; Fernlund and Stenflo, 1980b; Grundberg et al., 1980; Low et al., 1980; Petersen et al., 1980), but definitive identification of Gla is probably best accom-

6

LINDA J. VAN ELDIK ET AL

plished by mass spectrometry (Magnusson et al., 1974; Morris et al., 1976; Carr et al., 1981).

B. SYNTHESIS OF y-CARBOXYGLUTAMIC ACID The synthesis of y-carboxyglutamic acid and the metabolism of vitamin K have been described in several reviews and the proceedings of a recent symposium (Suttie and Jackson, 1977; Bell, 1978; Stenflo, 1978; Olson and Suttie, 1978; Suttie, 1978, 1980a,b,c; Suttie et al., 1980b) and are only summarized in this article. Three forms of vitamin K enter the liver vitamin K cycle: the quinone, the hydroquinone, and the 2,3-epoxide. The quinone and hydroquinone can be interconverted by a microsomal bound NAD(P)H-linked reductase. Vitamin KH, (the hydroquinone form) can be converted to the 2,3-epoxide form by a microsomal internal monooxygenase. The epoxide is in turn reduced to the quinone hy a warfarin-sensitive enzyme, 2,3-epoxide reductase. The reduced vitamin K is an essential cofactor for the vitamin K-dependent carboxylase which carboxylates protein-bound glutamyl residues to y-carboxyglutamyl residues. Subcellular localization studies have shown that the carboxylase activity is associated with rough endoplasmic reticulum and appears to be an integral membrane protein (Carlisle and Suttie, 1980). Most attempts to purify the carboxylase have not been successful although de Metz et al. (1981) have recently reported some progress. The molecular mechanism of the carboxylation and the role of vitamin K in the reaction are not understood. However, evidence suggests that the epoxidation of the vitamin is in some way linked to the carboxylation event (Carlisle and Suttie, 1980; Suttie et al., 1980a; Larson et al., 1981). Standard assays for vitamin K-dependent carboxylase activity have generally used precursors of vitamin K-dependent clotting factors isolated from vitamin Kdeficient rat liver microsomes as substrates. Soute et al. (1981) have prepared a peptide substrate for vitamin K-dependent carboxylase by limited proteolysis of bovine descarboxyprothrombin. This peptide, called Fragment-Su, is composed of amino acids 13-29 of descarboxyprothrombin and has an apparent K, in the micromolar range, approximately 100-fold lower than intact descarboxyprothrombin. The difficulty in isolating precursors and the problems associated with manipulating the concentration of an endogenous substrate have spurred the search for a simple, synthetic peptide to act as substrate (Suttie et al., 1979). A number of low-molecular-weight peptides have been synthesized for use as carboxylase substrates (Suttie et d.,1979; Rich et al., 1981). A pentapeptide, which consists of residues 5-9 of the uncarboxylated bovine prothrombin precursor (viz. Phe-Leu-Glu-Glu-Val), is the prototype peptide used as substrate for the carboxylase (Suttie et al., 1979) and has an apparent K, in the millimolar rmge (Soute et al., 1981).

7

CALCIUM-BINDING PROTEINS

Synthetic peptides have also been used in comparative studies of carboxylase substrate specificity (Houser et a l . , 1977; Esnouf et al., 1978; Suttie ef a l . , 1980b). The vitamin K-dependent carboxylase seems to prefer but not require adjacent glutamyl residues (Rich et al., 1981). It is not clear whether the primary sequence around the carboxylated residue is critical for efficient carboxylation (Houser et al., 1977; Suttie et al., 1979; Rich et a l . , 1981) although a hydrophobic environment appears to be most effective. Rich et al. (1981) have reported that peptides which have a high potential for helicity are the best substrates, suggesting that tertiary structure may also be of importance. OF PROTHROMBIN C. CALCIUM-BINDING CHARACTERISTICS

The calcium-binding and calcium-dependent phospholipid-binding characteristics of most of the Gla-containing plasma proteins appear to be quantitatively similar (Nelsestuen, 1976; Nelsestuen et a l . , 1978). The number of calciumbinding sites of these proteins has been estimated to be 6 (Prendergast and Mann, 1977; Bajaj et al., 1975; Nelsestuen et al., 1975) to 10 (Henriksen and Jackson, 1975) for prothrombin and 20 for Factor X (Henriksen and Jackson, 1975; Lindquist and Hemker, 1978). The similarities in calcium- and phospholipidbinding characteristics are not unexpected in light of the extensive sequence homology observed in the amino termini of the Gla-containing plasma proteins (see Fig. 1). Most studies of the kinetics of calcium-binding and calcium-dependent phospholipid binding have used prothrombin or the isolated prothrombin fragment-I (residues 1-156 of prothrombin) as an investigative model. For this reason, structural and calcium-binding characteristics of prothrombin will be considered here as a prototype of other Gla-containing plasma proteins. The complete amino acid sequence of bovine prothrombin has been determined (Magnusson et al., 1975) and two preliminary reports on the crystal structure have been presented (Aschaffenburg et al., 1977; Olsson et al., 1980). However, the complete three-dimensional structure of the molecule has not been determined. Prothrombin is composed of 582 amino acids including 10 Gla 10

20

30

40

Prothrombin

A-N-K-G-F-L-E'ELV-R-K-G-N-L-E'R-EIC-L-E'-E~P-C-S-R-E'E'A-R-E'A-L-E'S-L-S-~-T-D-A-F-W-A-

Factor X

A-N-S-. -F-L-EIE'V-K-N-G-N-L-E~R-E'C-L-E'EIA-C-S-L-E~E~A-R-E~V-F-ELD-A-E'-q-T-o-E~F-W-s-

Protein C

A-N-s-.

-F-L-E'-E~L-R-P-G-N-V-E'-R-E'-C-S-E'-E~V-C-E~F-~-~-A-R-~-

I-F-Q-N-T-E'-~-T-M-A-F-W-S-

FIG. 1. Amino terminal sequences of bovine prothrombin (Magnusson et a / . . 1979, Factor X light chain (Enfield er al., 1975). and protein C (Fernlund and Stenflo, 1980a). Dots in the sequence denote the presence of a gap introduced for the purpose of alignment. E' denotes y-carboxyglutamic acid. Modified from and numbered according to Davie (1980). The single letter code for the amino acids is as follows: A=Ala, D=Asp, E=GIu, F=Phe, G=Gly, H=His, I=lle, K=Lys, L=Leu, M=Met, N=Asn, P=Pro. Q=Gln, R=Arg, S=Ser, T=Thr, V=Val, W=Trp, Y=Tyr.

8

LINDA .I.VAN ELDIK ET AL.

residues and three complex carbohydrate chains. During activation of prothrombin to thrombin the NH,-terminal peptide consisting of 156 residues is cleaved (Suttie and Jackson, 1977) (see Fig. 2). This peptide, known as fragment-1, contains all 10 of the Gla residues (Magnusson et al., 1974) and two of the carbohydrate chains (Nelsestuen and Suttie, 1972; Magnusson et al., 1975) of the intact prothrombin. Removal of the majority of sugar residues from prothrombin apparently does not alter its calcium-binding activity (Nelsestuen and Suttie, 1972). Several lines of evidence indicate that prothrombin binds to phospholipid dispersions in the presence of calcium (Papahadjopoulos and Hanahan, 1964; Esnouf and Jobin, 1965; Barton and Hanahan, 1969; Nelsestuen, 1978) and that activation of prothrombin to thrombin is accelerated in the protein:calcium:phospholipidcomplex (Milstone, 1965). The phospholipid binding ability of prothrombin has been shown to be restricted to the fragment-1 region of prothrombin (Gitel et al., 1973; Benson et al., 1973) and acarboxyprothrombin does not bind phospholipid (Esmon er al., 1975). The exact chemical nature of the phospholipid mixture does not appear to be critical (Bangham, 1961; Papahadjopoulos et ul., 1962) but optimal binding has been shown to occur between prothrombin-calcium complex and phospholipids with a specific negative surface charge density (Papahadjopoulos et al., 1962; Dombrose er al., 1979). The 10 Gla residues in prothrombin fragment-I appear to be required for maximum efficiency of calcium and phospholipid binding (Esmon et al., 1975). When glutamyl residues are substituted for Gla residues both calcium and phospholipid binding are decreased (Esnouf and Prowse, 1977; Friedman et al., 1977). Tertiary structure of the protein may contribute to the spatial proximity of Gla residues and thus influence the strength of calcium binding at a particular binding site. However, data obtained (Marsh er al., 1980, 1982) with small

-F-2

coon

-

I

F- I

H, N C-H I F"1

n-c-cooH I

coon y-Cor boxyglutomic Acid

P- 2

P- I

A

I

t

:

Hl"" Glo Reqm

\

CoOn

LS-SJ

B

A

-L s - S J

Ihr

FIG 2. Structure of y-carboxyglutamic acid (Gla) and a diagrammatic representation cif the prothrombin molecule. Proteolysis of prothrombin by thrombin and factor X will cleave prothrombin into the specific large peptides shown: fragment-I (F-I), fragment-2 (F-2). prethrombin-I IP-1), prethrombin-2 (P-2). and thrombin (thr). For details of the activation of prothrombin and thrombin see Suttie and Jackson (1977). Modified from Suttie (1980a).

CALCIUM-BINDING PROTEINS

9

synthetic peptides suggest that the dianion contributed by Gla is sufficient to chelate calcium with an affinity close to that of the native protein. Calcium binding to prothrombin fragment- 1 has been described as consistent with both positive cooperative (Nelsestuen, 1976; Prendergast and Mann, 1977) and multiclass kinetics (Benson and Hanahan, 1975; Jackson, 1980). It has been suggested (Brenkle ef af., 1980; Jackson, 1980) that these data are not in conflict but that the complex calcium-binding behavior of isolated prothrombin fragment- I depends upon the protein concentration at which the binding studies are done. Prothrombin fragment- I undergoes a calcium-dependent self-association (Jackson et al., 1979; Brenkle et al., 1980; Jackson, 1980) and this protein-protein interaction may be responsible for the cooperative calcium-binding kinetics observed. However, self-association may not occur in the intact prothrombin molecule (Nelsestuen et al., 1980b) so the protein-protein interactions may not be physiologically relevant. Several lines of evidence suggest that a conformational change occurs in both isolated prothrombin fragment- 1 (Nelsestuen, 1976; Prendergast and Mann, 1977; Blanchard et al., 1980; Carlisle et af., 1980; Nelsestuen er a l . , 1980a,b) and in intact prothrombin (Bloom and Mann, 1978; Benarous et al., 1980; Blanchard et af., 1980; Nelsestuen er al., 1980b) upon binding of calcium and that this change may involve an increase in helicity (Bloom and Mann, 1978; Nelsestuen et af., 1980a). The conformational change occurs in the presence of millimolar concentrations of either calcium, Mg2 , or Mn2+ and has been measured by a number of techniques including fluorescence quenching (Prendergast and Mann, 1977) and circular dichroism (Nelsestuen, 1976; Bloom and Mann, 1978). However, fluorescence quenching occurs prior to fragment-1:phospholipid binding and metals other than calcium, e.g., Mg2+, can cause fluorescence quenching without bestowing phospholipid binding properties on the fragment-1:metal ion complex (Nelsestuen, 1976). In summary, it appears that calcium binding by prothrombin is mainly due to a chelation of the cation by the two anionic carboxyl oxygens contributed by the Gla residues. Available evidence suggests that a conformational change occurs following binding of metal ions to prothrombin fragment-1 (Prendergast et al., 1980). Whereas other metal ions may bind and induce a conformational change in protein structure, phospholipid binding will occur only in the presence of calcium. It has been proposed (Nelsesteun, 1976, 1978; Marsh et al., 1980) that this calcium-dependent conformational change is the rate-limiting step in phospholipid binding to prothrombin. +

PROTEINS CONTAINING 7-CARBOXYGLUTAMIC ACID D. PLASMA Seven plasma proteins containing y-carboxyglutamic acid have been described: prothrombin (Downing er af., 1975; Magnusson er al., 1975), factor X

10

LINDA J . VAN ELDIK ET AL.

(Aronson et al., 1969; Nelsestuen et al., 1974; Di Scipio et al., 1977), factor IX (Osterud and Flengstrud, 1975; Di Scipio et al., 1977; Davie, 1980), factor VII (Enfield et al., 1975), protein C (Fernlund et al., 1978), protein S (Di Scipio et al., 1977; Di Scipio and Davie, 1979; Stenflo, 1980), and protein Z (Petersen et al., 1980). Like prothrombin, the other six proteins contain 10-12 residues of Gla per mol of protein. Figure 1 shows the extensive homology that exists in the NH,-termini of three of these proteins. In addition to the structural similarities among these proteins, they appear to have similar calcium- and phospholipidbinding characteristics (Nelsestuen et al., 1978) and all are thought to be directly or indirectly involved in the clotting cascade. The amino acid sequence of bovine factor X consists of one heavy chain and one light chain linked by disulfide bonds. The light chain contains 140 amino acids including 12 Gla residues while the heavy chain contains 307 residues with two carbohydrate moieties but no y-carboxyglutamic acid (Enfield et al., 1975; Titani et al., 1975). Bovine factor IX is a single polypeptide chain of 416 amino acids with a total of 12 Gla residues, all in the amino terminal portion of the molecule (Davie, 1980). While evidence suggests that Factor VII would contain Gla at positions homologous to other plasma coagulation factors, its low concentration in the plasma has prevented a definitive sequence location of the Gla residues (Suttie, 1980a). Protein C is composed of 2 polypeptide chains linked by disulfide bonds (Femlund et al., 1978; Fernlund and Stenflo, 1980a). A light chain of 155 amino acids has 11 Gla residues and extensive sequence homology with the light chain of Factor X (Fernlund et al., 1978). The heavy chain consists of 258 amino acids (Fernlund and Stenflo, 1980a). The limited sequence data available for protein S suggest homology with other vitamin K-dependent plasma proteins (Di Scipio et al., 1977; Di Scipio and Davie, 1979; Stenflo, 1980). A seventh plasma protein containing y-carboxyglutamic acid, protein Z, has also been described (Mattock and Esnouf, 1973; Prowse and Esnouf, 1977; Petersen et a f . , 1980) but the amino acid sequence has not been reported. Some structural differences between bovine and human plasma proteins have been described. For a discussion of some of these differences the interested reader is directed to Di Scipio et al. (1977). E. BONEPROTEINS CONTAINING Y-CARBOXYGLUTAMIC ACID An abundant bone protein has been isolated and shown to contain y-carboxyglutamic acid (Price et al., 1976a; Hauschka and Gallop, 1977). The presence of this y-carboxyglutamic acid-containing protein was first demonstrated (Hauschka et al., 1975) in chick bone and similar proteins have been isolated from bovine (Price et a f . , 1976a), swordfish (Price et a f . , 1977), rat (Price et a f . , 1980a), and human (Poser et a l . , 1980) bone. These proteins have been called osteocalcin (Hauschka and Gallop, 1977) and bone Gla-containing protein

11

CALCIUM-BINDING PROTEINS

(BGP) (Price et al., 1976a). It is possible that these proteins represent speciesspecific forms of the same protein. For the purposes of simplicity, we will refer to these proteins as BGP throughout this article. BGP binds weakly to calcium ion (Kd = 3 mM) (Price et al., 1977) and strongly to hydroxyapatite (Price, 1980). The protein appears to comprise about 15%of noncollagenous protein of bone (Hauschka et al., 1975) and has been shown to be vitamin K-dependent (Hauschka and Reid, 1978). The amino acid sequences of the bovine (Price et al., 1976b, 1977), swordfish (Price et al., 1977), human (Poser et al., 1980), and chicken (Carr et al.. 1981) proteins have been reported. Comparison of the amino acid sequences (Fig. 3) of the bovine, swordfish, and chicken proteins shows extensive homology (Price et al., 1977; Carr er al., 1981). However, none of the bone proteins appears to have structural homology with the known vitamin K-dependent plasma proteins. The bovine and chicken bone proteins are 49 and 50 amino acids in length, respectively. Both proteins contain y-carboxyglutamic acid residues at positions 17, 21, and 24, and have a single disulfide between cysteine residues located at positions 23 and 29. In addition, the bovine protein has a residue of 4-hydroxyproline at position 9 (Price et al., 1976b) that is not found in the chicken protein (Carr et af., 1981). The region around the three y-carboxyglutamic acid and two cysteine residues is highly conserved and the spacial relationship of the Gla residues may be important to calcium- and hydroxyapatite-binding interactions (Poser and Price, 1979; Price, 1980; Price et al., 1980b; Carr et al., 1981). Thermal decarboxylation of the Gla residues greatly reduces affinity of the protein for calcium. Likewise, decarboxylation and reduction of the disulfide bond reduce both the affinity of the protein for hydroxyapatite and its ability to retard precipitation of calcium phosphate salts from supersaturated solutions. Chicken BGP has been synthesized in v i m in bone culture and a vitamin Kdependent carboxylase has been demonstrated in embryonic chick bone (Lian 1

20

10

Bovine

.-Y-L-D-H-W-L-G-A-P'-A-P-Y-P-D-P-L-E'-P-K-R-F-V-C-~-L-N-P-D-C-D-

Chicken

H-Y-A-0-D-S-G-V-A-G-A-P-

Swordfish

.-.- - .-A-T-R-A-G-D-L-T-P-L-O-L-E'-S-L-R-€-V-C-~-L-N-V-A-C-D-

Bovine

E-L-A-0-H-I-G-F-0-E-A-Y-.-R-R-F-Y-G-P-V

Chicken

E-L-A-D-E-L-G-F-0-E-A-Y-0-R-R-F-Y-G-P-V

Swordfish

E-M-A-0-T-A-G-I-V-A-A-Y-.-I-A-Y-Y-G-P-I-Q-F

30

.-P-N-P-I-f-A-Q-R-F-V-C-E'-L-S-P-D-C-N-

40

FIG. 3. Comparison of the primary structures of BGP from bovine (Price et al., 1976b), chicken (Camet al., 1981). and swordfish (Price et al.. 1980b) bone. The numbering system is according to the bovine sequence. The single letter code for the amino acids is as described in the legend to Fig. 1 . E' denotes y-carboxyglutamic acid and P' denotes 4-hydroxyproline.

12

LINDA J . VAN ELDIK ET AL.

and Friedman, 1978). While the cell type responsible for the synthesis of I3GP has not been unequivocally determined, immunoreactive BGP is found only in clones of osteosarcoma that have an osteoblast-like phenotype (Price, 1980; Nishimoto and Price, 1980). In addition, the presence of 4-hydroxyprolir~ein bovine BGP (Price er al., 1976b) indicates that the protein has been modified by prolyl hydroxylase, a “marker” enzyme used to distinguish osteoblasts from osteoclasts (Nishimoto and Price, 1980). The function of the Gla-containing proteins of bone is not known. The presence of y-carboxyglutamic acid appears to correlate with the degree of mineralization (Lian er al., 1980) suggesting that Gla-containing proteins may be involved in the mineralization process. However, developmental studies do not support this hypothesis, since BGP appears in bone after hydroxyapatite (Hauschka, 1980; Price, 1980; Price et al., 1980a) and low levels of BGI’ are present during early stages of mineralization (Price e? al., 1980a, 1981). Furthermore, treatment of developing animals with warfarin (a vitamin K antagonist) causes a sharp reduction in BGP but no gross defect in bone strength, morphology, histology, or protein content (Price et al., 1980a). Reddy and Suttie (1980) have reported that warfarin treatment reduced the amount of calcium in fetal rat bone and in the enamel of adult rat incisors without altering the phosphate content. These studies and the ability of BGP to control precipitation of calcium phosphate salts suggest that BGP may play a structural role after initial mineral deposition such as regulation of the transition from fetal amorphous calcium phosphate to hydroxyapatite crystals of adult bone (Price, 1980). An informational role has also been suggested for BGP (Price, 1980; Prict: and Baukol, 1980; Price et al., 1980a). This idea is supported by evidence that synthesis of BGP in rat osteosarcoma cells is increased by 1,25-dihydroxyvitamin D,, a compound associated with inhibition of bone matrix formation (Price and Baukol, 1980). Furthermore, BGP has been detected in serum by radioimmunoassay (Price and Nishimoto, 1980). While the protein is normally present at low levels in humans, analysis of plasma from patients with bone disease involving an increased turnover of the bone matrix (e.g., Paget’s disease or metastatic cancer) has shown BGP levels to be above normal (Price and Nishimoto, 1980). BGP may, therefore, be a useful diagnostic tool for such osteopathic diseases. Other proteins which contain Gla have been isolated from skeletal tissue. A higher molecular weight, Gla-containing protein has been detected in low amounts in both chicken (Lian and Heroux, 1980) and rat (Price, 1980; Price et al., 1981) bone matrix. It is not clear if these matrix-associated proteins represent a precursor of BGP (Price et al., 1980a). Nishimoto and Price (1980) have reported an intracellular Gla-containing protein of approximately 9OOO daltons (BGP = 5800 daltons) in rat osteosarcoma cells having an osteoblast-like phenotype. This protein cross-reacts with antisera directed against rat BGP suggest-

CALCIUM-BINDING PROTEINS

13

ing that it may represent a precursor to BGP. Lian er al. (1980) have recently examined several tissues of the elasmobranch species and reported that some of the noncalcified cartilage of shark contained more Gla per thousand glutamic acid residues than did some calcified cartilage. The nature of these Gla-containing proteins is not clear but their function may involve something other than regulation of mineralization.

F. OTHERPROTEINS CONTAINING y-CARBOXYGLUTAMIC ACID A number of other y-carboxyglutamic acid-containing proteins have been described. Several y-carboxyglutamic acid-containing proteins from various tissues have been reported in association with ectopic calcifications. Lian et al. (1976, 1977) have described Gla-containing proteins that are associated with calcium-containing renal calculi. A y-carboxyglutamic acid-containing protein has also been isolated from bovine kidney cortex (Griep and Friedman, 1980), chicken kidney microsomes (Traverso et al., 1980), and rat and rabbit renal tissue (Hauschka et al., 1976). These proteins appear to be distinct from other Gla-containing proteins by molecular weight and amino acid composition. In addition, Gla-containing proteins have been found in association with atherosclerotic plaque (Lian et al., 1976; Levy et al., 1979) and calcified exudates associated with scleroderma and dermatomyositis (Lian et al., 1976). Gla has been reported in association with purified ribosomes of both eukaryotic and prokaryotic cells (Van Buskirk and Kirsch, 1978a,b; Olson et al., 1978; Scheinbuks, 1980; Van Buskirk er al., 1980). Chorioallantoic membrane of chick embryos has been reported (Tuan and Scott, 1977; Tuan et al., 1978a,b; Tuan, 1980)to contain a high-molecular-weight, vitamin K-dependent protein which is involved in calcium transport to the embryo. However, others have been unable to demonstrate Gla in this tissue (Lian er al., 1980). A placental Gla-containing protein has also been described which may have a calcium transport function (Friedman et al., 1979).

111. Concanavalin A and Calcium-Binding Lectins Lectins are proteins that have no known enzymatic activity but exhibit numerous biological activities that are related to their ability to bind carbohydrates in the presence of divalent cations. Plant and animal lectins have specific binding sites for carbohydrates and thereby interact with specific cells, cell fractions, or glycoproteins. Proposals for the role of lectins in cell function include conferment of resistance to pathogens, promotion of symbiosis, involvement in cell recognition, and organization of supramolecular structures in eukaryotic cells (Liener, 1976; Sequeria, 1978; Goldstein and Hayes, 1978; Bangerth, 1979;

14

LINDA J . VAN ELDlK ET AL.

Barondes, 1981). A recent review (Barondes, 1981) discusses in detail the possible endogenous cellular functions of lectins. The majority of reports have been on exogenous functions, especially the effects of plant lectins (also called phytohemagglutinins) on animal cells. Plant lectins can induce alterations in animal cells including agglutination, cell proliferation, metabolism, and membrane mobility. Little is known about the role of plant lectins in plant cell function although a number of theories and models have been proposed (Liener, 1976; Goldstein and Hayes, 1978; Barondes, 1981). However, plant lectins arc the most extensively characterized lectins in terms of chemical and physical properties and, more relevant to this article, the plant proteins are the best characterized calcium-binding lectins. Calcium-binding studies have been done on only a few lectins probably due to the technical difficulties involved (Grimaldi and Sykes, 1975; Brown et a!., 1977; Cardin et a l . , 1979, 1981). For the lectins which have been examined, the estimated dissociation constant for calcium at 25°C is in the range of l o p 3 to 10 -5M.The proposed roles for calcium and the postulated calcium-binding sites are largely based on homology with conoanavalin A (Con A). Con A is the only calcium-binding lectin for which a complete amino acid sequence and a threedimensional crystallographic structure are available (Becker et al., 1975, 1976; Reeke ef al., 1975, 1978). It is a tetramer composed of four identical polypeptide chains of 237 residues each (Fig. 4). The tetramer structure of Con A appears to be required for in vitro agglutination or precipitation activity but the dimer is still capable of binding low-molecular-weight saccharide derivatives. The tetramer has 222 symmetry and each monomer can be schematically visualized a:s an ellipsoid dome with a narrow, flat base. Each monomer has two divalent cationbinding sites and a carbohydrate-binding site which are closely spaced near the apex of the dome. Metal ions contribute to the structural stability of Con A and are required for its in vitro activities. In the presence of divalent cations, Con A preferentially binds to a-D-mannopyranosyl, a-D-glucopyranosyl, and a-r>-Nacetylglucosaminyl residues. The saccharide-binding site has not been defined in as much atomic detail as the metal-binding sites, but has been localized to within 10-14 8, of the metal-binding sites (Becker et al., 1976). The residues that may participate in saccharide binding are found in close proximity in the threedimensional structure but are not clustered in the amino acid sequence. These include residues 14-16, 97-99, 168-169, 207-208, 224-228 and 235-237. In contrast to the saccharide-binding site, most of the residues involved in metal binding are clustered in the NH,-terminus of the amino acid sequence (Fig. 4). The divalent cations are bound in two adjoining octahedral sites that are 4.5 8, apart and have a common edge (see Fig. 5 ) . The ligands involved in binding of divalent cation at the first site, called S I , are the side chains of glutamic acid-8, aspartic acid-10, aspartic acid-19, histidine-24, and two water molecules. One of the water molecules appears to be hydrogen bonded to the carbonyl oxygen of

10 C. . n . l f o d i

0160

1.

1.b.

c. em.lfod.

v.

f.b

I. s"lln.ri,

v.

60

10

10

~-A-K-Y-N-I(-q-D-C-K-V-C-T-A-R-f-f-Y-N-S-V-D-~-R-L-S-A-V-V-S-l-P-N-A-D-A-T-S-V-S-Y-D0110

8180

01

~ - ~ - ~ - ~ - N - L - ~ - ~ - ~ - I - ~ ~ A ~ ~ - V - A - I - ~ - ~ - M - A - T - I - N - V - L - ~ - V - IL-T-C-Y-I-L-S-F-V-L-L-Y-P-N

I20 I30 140 I so I60 I-T-R-~-T-O-A-L-I-I-I(-P-~-9-P-S-K~D-~-~-D-L-l-L-~-C-O-A-T-T-C-T-N-C-N-L-E-L-T-R-V-S-S-

om 5-1

II T-D-L-I-T-S-P-~-1-P-K-P-~-P-O-~-P-N-L-I-P-q-C-C-C-Y-T-T-K-F-K-.-L-T-L-T-K-A-V-KI10 820 1130

am

61

110

820

030

T-t-T-T-S-P-S-l-T-K-P-S-P-D-9-q-~-N-L-I-?-q-C-D-C-T-T-C-K-~-C-.-L-I-L-T-~-V-S-K-

S-K-8

a30 ..Ll".

?. I . C l N

s-a

am 5-1-1-Q

210

2 30

220

-I-P-*-~-c-I-A-F-~-

4 '

C. *n*lforP1*

D-I-.

v.

1.-

eiio N - C - I - J - V - A - D - C - P - T - ~ - P - l - A - ? - V - O - T - K ~ p ~ g ~ ~ . ~ ~ c ~ c ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ y

L.

CU1*MrlO

Y-C-T-I-V-A-O-C-P-T-P-?-I-A-P-V-O-T-K-P-~-I-C-C-~-~-.-.-L-C-V-~-l-N-~-K-

(la0

a90

6100

Boo

IW

6100

3109

FIG.4. Comparison of the amino acid sequences of plant lectins. The amino acid sequences of lectins from C. ensifomis and V.faba (Cunningham er al., 1979), L. culinan's (Foriers er a / . . 1981). V. sativa (Gebauer et al.. 1981). and P. sativum (Richardson er al., 1978) are aligned by a circular permutation to demonstrate maximum homology. Asterisks (*) denote amino acid residues involved in metal binding at the S2 site (the calcium-binding site), while pluses (+) denote residues unique to divalent cation binding at the S1 site (the Mn2+ or transition metal-binding site). The S I site also uses Asp10 and Asp19 for metal binding. Lines above residues indicate those amino acids believed to be involved in saccharide binding. The single letter code for the amino acids is as described in the legend to Fig. 1.

16

LINDA J . VAN ELDIK ET AL.

i

'C

C.

FIG. 5 . The metal binding sites of concanavalin A. The ligands involved in binding divalent cation at the S, site (the Mn2+ or transition metal site) are the side chains of Glu-8, Asp-10, A:jp-19, His-24, and two water molecules. The S2 site (calcium-binding site) utilizes the side chains of Asp-10, Asn-14, Asp-19, the carbonyl oxygen of Tyr-12, and two water molecules as metal ligands. Asp-I0 and Asp-I9 are shared by the two metal-binding sites resulting in a common edge for the two metal-binding sites. Reproduced from Becker er al. ( 1975).

valine-32 and the side chain hydroxyl of serine-34 (Becker et al., 1976). The first site (Sl) binds Mn2 but a variety of metals can substitute for Mn2 at the S1 site (Liener, 1976). These include cadmium (Cd2+), terbium (Tb3+), lanthanum (La3+), zinc (Zn2+), cobalt (Co2+), nickel (Ni2+), and lead (Ph2+). The second site, S2, will bind calcium, Cd2+, and Mn2+ but not barium (Ba2+), fermium (Fm3+), or most transition metals (Shoham et al., 1973; Becker et al., 1975). The calcium-binding site (S2) utilizes the side chains of aspartic acid-10, asparagine-14, aspartic acid-19, the carbonyl oxygen of tyrosine-12, and two water molecules as metal ligands (Becker et al., 1975). The water molecules might be hydrogen bonded to the side chain carboxyl of aspartic acid-208 and the carbonyl oxygen of arginine-228. Both of the side chain oxygens of aspartic acid-10 are involved in metal binding and, as shown schematically in Fig. 5 , aspartic acids 10 and 19 are shared by the two metal-binding sites. Becker et al. (1975) have described the metal-binding sites as "a binuclear complex composed of two polyhedra sharing a common edge." Reeke et al. (1978) have proposed a mechanism for the interaction of the two metal-binding sites and the saccharide-binding site. The mechanism is based on comparative crystallographic data from their own laboratory (Reeke et al., 1978) and spectroscopy studies by others (Grimaldi and Sykes, 1975; Brown et al., 1977). When the crystal structures of the metal-saturated protein (native) and the +

+

17

CALCIUM-BINDING PROTEINS

demetallized protein (apoprotein) are compared, three of the residues (glutamic acid-8, aspartic acid-10, and histidine-24) in the S1 site are in similar positions whereas three residues (tyrosine-12, asparagine-14, and aspartic acid- 19) in the S2 site are in different positions in the two crystal structures. These data suggest that the S1 site (transition metal site) is mostly preformed in the apoprotein and that conformational changes in the protein occur upon binding calcium in the S2 site. In the proposed mechanism the conformational changes would be initiated by the interaction of aspartic acid-19 with the transition metal in S l followed by stabilization of the conformation by the interaction of aspartic acid-19 with calcium in S2. Aspartic acid-19 is one of two aspartyl residues (aspartic acid-10 is the other) that interact with both metal ions (see Fig. 5). The stabilization of the conformation with both calcium and Mn2+ bound would result in the saccharide-binding residues being brought into juxtaposition. There are other cation-binding plant lectins that are functionally similar to and have some sequence homology with Con A but differ in their subunit structure (Foriers et al., 1978, 1981; Richardson et al., 1978; Cunninghamn et al., 1979; Hemperly et al., 1979; Gebauer et al., 1981). These lectins, which include those from Lens culinaris, Pisum sativum, Vicia sativa, and Viciafaba,are composed of two types of polypeptide chains termed cu (M,= 6,000) and p (M,= 15,000) and have a subunit structure of (cup),. Partial or complete amino acid sequences of several of these proteins have been determined (see Fig. 4). The two-subunit lectins have a very high degree of sequence homology with each other and all are similar to the Con A amino acid sequence. Cunningham and co-workers (Cunningham et a f . , 1979; Hemperly et al., 1979) have determined the amino acid sequence of the a and p chains of favin and have demonstrated that the two chains are related to Con A by a circular permutation. As shown in Fig. 4, the NH,-terminal sequences of the p chains are homologous to residues 120-237 in the COOH-terminal half of Con A, and the COOH-terminal half of the p chains sequences are homologous to residues 1-69 of the Con A sequence. The cu chains are homologous to residues 70-1 19 in the middle of the Con A sequence. The metal binding residues are identical in the lectins shown in Fig. 3 except for the presence of a phenylalanine in V .faba and L. culinaris lectins in place of tyrosine-12 in Con A. However, the carbonyl oxygen is used at this position for calcium binding in Con A, so the change would not be expected to alter the metal coordination in the V. faba and L. culinaris lectins. The proposed saccharide-binding residues are not as highly conserved as the proposed metal-binding residues. This may explain differences in affinities and specificities among the lectins. Thus, the a and p chains contain the structural domains of Con A if the two amino acid sequences are compared by a circular permutation. It is possible that other lectins of the a p type may also be related to Con A or other single chain lectins by a circular permutation of the .

..

- .

18

LINDA J . V A N ELDlK ET A L .

amino acid sequences. Models which invoke gene duplication, gene rearrangement, and posttranscriptional processing have been proposed (Cunningham et al., 1979; Foriers et al., 1981; Gebauer et al., 1981) in order to explain the evolution and biosynthesis of these two classes of structurally and functionally homologous lectins. In summary, Con A is a calcium-binding lectin that has been characterized in detail and serves as a standard of comparison among lectins. The coordination of calcium in Con A and the proposed conformational or configurational changes that occur upon calcium binding are quite distinct from those of parvalbumin, phospholipase, and other calcium-binding proteins. The various in virro activities of lectins and the possible endogenous roles of lectins in cell function have been discussed in more detail in other reviews (Liener, 1976; Sequeria, 1978; Goldstein and Hayes, 1978; Bangerth, 1979; Barondes, 1981). Clearly, determination of the subcellular distribution of calcium-binding lectins and elucidation of the endogenous activities of these lectins would enhance our understanding of how calcium is involved in lectin function. Any postulated roles for lectins in the mechanism of calcium action must consider the extensive amount of physical and chemical data that are available. In the Con A apoprotein the apparent affinity (dissociation constant in the range of 10V3 M ) for calcium is low (Brown et af., 1977; Cardin et af., 1981). If the S1 site is preformed and a transition metal is bound, then the S2 site appears to have an increased affinity for calcium (Brown et af., 1977). However, in this more stable ternary complex the off rate for Mn2 and calcium is days at 25°C. Thus, it is not likely that Con A or similar lectins would be involved in rapid stimulus-response-relaxation cycles. It is more likely that if calcium-binding lectins are involved in the molecular basis of calcium action on cellular processes, it would be in the mediation of chronic or long-term effects. +

IV. Calcium-Binding Hydrolytic Enzymes A variety of enzymes which catalyze the hydrolysis of ester, phosphodiester, and peptide bonds bind calcium ions. This class of calcium-binding proteins includes many hydrolytic enzymes, although mechanistic and crystallographic data are available only for phospholipase A,, staphylococcal nuclease, thermolysin, trypsin, and chymotrypsin. Thus, hydrolytic enzymes have three uses for calcium. First, calcium can stabilize an intermediate in the active site as in phospholipase A2 and staphylococcal nuclease. Second, calcium can stabilize the enzyme at high temperature as in thermolysin and, to a lesser extent, in trypsin. Third, calcium can take part in the activation of a zymogen as in trypsin and phospholipase. The calcium-binding sites of phospholipase A, have many features characteristic of calcium-binding hydrolytic enzymes. Therefore, in this

19

CALCIUM-BINDING PROTEINS

article phospholipase A, will serve as a model for calcium-binding hydrolytic enzymes. A. PHOSPHOLIPASE A2 Phospholipase A2 catalyzes the specific hydrolysis of the fatty acid ester bonds at the C, position of I ,2-diacyl sn-phosphoglycerides (L-phospholipids) (van Deenen and de Haas, 1964). Both secreted and intracellular forms of phospholipase A2 activity have been described (Van den Bosch, 1980). The actions of the intracellular phospholipases appear to be important in cellular physiology and metabolism because the activities respond to hormonally induced changes in cyclic nucleotide and calcium ion levels (for a review see Van den Bosch, 1980). However, since the secreted forms of phospholipase A2 have been characterized in greater structural and functional detail, this article will focus on secreted phospholipase A,. Phospholipase A, has been isolated from mammalian pancreas (Drenth et al., 1976; Evenberg et al., 1977; Fleer et al., 1978), snake venoms (Botes and Viljoen, 1974; Joubert, 1975a,b; Heinrikson et al., 1977; Hanahan er al., 1980; Joubert and Taljaard, 1980; Kondo et al., 1981a,b; Lind and Eaker, 1981; Tsai et al., 1981), and bee venom (Shipolini et al.. 1971). These enzymes are single polypeptides of 118 to 129 amino acids, although some of the snake venom enzymes have been shown to dimerize under certain conditions (Keith et al., 1981). The amino acid sequences of phospholipases A, are homologous (see Fig. 6), with particular conservation of the amino acid residues thought to be involved in calcium binding and in forming the active site (Dijkstra et al., 1981a,b). Long and Penny ( 1957) have shown that phospholipase A, from snake venom has an absolute requirement for calcium ions and can be inhibited by ethylenediaminetetraacetic acid (EDTA). Other metal cations such as Na , K , Ba2 , Sr2 , Mg2 , and Cd2 could not replace the calcium-dependent activity, and Cu2+ and Zn2+ inhibited the basal activity. The porcine pancreatic enzyme is also highly specific for calcium (Pieterson et al., 1974a). Ba2+ and Sr2 bind to pancreatic phospholipase and competitively inhibit the calciumdependent activity, while Mn2 + ,Mg2 +,and Co2 do not bind to the enzyme or affect the activity (Pieterson et al., 1974a). Using equilibrium gel filtration on the porcine pancreatic enzyme, Pieterson et al. (1974a) have found that the dissociation constant for calcium varies from 12 mM at pH 5.0 to 0.25 mM at pH 8. In a later study using equilibrium dialysis, Slotboom et al. (1978) discovered that porcine pancreatic phospholipase A, binds 2 mol of calcium per rnol of protein at high pH (7.5-8) with dissociation constants of 0.1 and 3 mM at pH 8. The high-affinity calcium-binding site is involved with the catalytic site of the enzyme (Dijkstra et al., 1981a,b) and the low-affinity calcium-binding site is near the site of interaction with lipid surfaces (Slotboom er al., 1978). +

+

+

+

+

+

+

+

20

LINDA J . VAN ELDIK ET AL.

Bovine pancreas Equine pancreas P o r c i n e pancreas

_ B l_ tiS

&a&*

_~_

Crocalus adazanteus ~A a j a nelanclcura.

DE-I

Bovine pancreas

E q u i n e pancreas

Porcine pancreas B i t i s gabonica

Crotalus_ adazanteus ~ _

_

A a j a rnelanaleuca. DE-I

FIG. 6. The amino acid sequences of phospholipases A*. Sequences are shown for bovine pancreas (Fleer er al.. 1978), equine pancreas (Evenberg er al., 1977), porcine pancreas (Drenth er al., 1976). Bitis gabonica venom (Botes and Viljoen, 1974). Croralus adamanteus venom (Heinrikson er al., 1974). and Naja melanoleuca venom, form DE-I (Joubert, 1975a.b). The sequences have been aligned according to Kondo er al. (1981b). The amino acids marked are present in the cleft containing the active site, and starred (*) amino acids also coordinate the calcium ion (Dijkstra et al., 1981a.b). The single letter code for the amino acids is as described in the legend to Fig. 1.

B . CALCIUM AND

THE

ACTIVESITE

Calcium is coordinated near the active site by six ligands that form an octahedron (Dijkstra el al., 1981b, and Fig. 7). The carboxylate oxygen of aspartic acid-49 is one ligand (Fleer er al., 1981), and the peptide carbonyl oxygens of tyrosine-28, glycine-30, and glycine-32 are three other ligands (Verheij et al.,

CALCIUM-BINDING PROTEINS

21

1980). The fifth ligand is a water molecule, and the sixth may be both a second water molecule and the second carboxylate oxygen of aspartic acid-49 (Dijkstra et af., 1981b). Chemical modification studies (Fleer et af., 1981) have shown that aspartic acid-49 is essential for calcium binding. It has been proposed (Dijkstra et af., 198la; Verheij et af., 1980) that in the presence of phospholipid substrates, a phosphoryl oxygen replaces a water molecule in calcium coordination. However, a direct interaction between the phosphate and the calcium ion has not been demonstrated. Vensel and Kantrowitz (1980) have suggested that an arginyl residue binds the substrate phosphate but does not coordinate calcium. In a number of studies, changes in the spectral properties of phospholipase A, upon calcium binding implicated a histidine residue (Pieterson et af., 1974a; Roberts et af., 1977; Wells, 1974b). However, chemical modification (Volwerk et af., 1974), NMR (Aguiar et af., 1979), and X-ray crystallography (Dijkstra et af., 1981b) have demonstrated that histidine-48 has an essential role at the active site but is not directly involved in calcium binding. The structure of the active site of phospholipase A, has been defined by the elucidation of the three-dimensional structure of the bovine enzyme at 1.7 A resolution (Dijkstra et af.,198lb). The amino acid residues lining the interior of the active site have also been defined: phenylalanine-5, isoleucine-9, phenylalanine-22, alanine-102, alanine- 103, phenylalanine-106, and the cysteine-29:cysteine-45disulfide bridge. The scissile bond of the L-phospholipid is required for substrate to fit the active site (Dijkstra et af., 1981a,b). A mechanism for the hydrolysis of phospholipids by phospholipase A, has

Tyr 28 FIG. 7. A schematic representation of the ligands that coordinate calcium in bovine pancreas phospholipase A2 in the absence of substrate. Water molecules are numbered according to Verheij er al. (1980). The second carboxylate oxygen of Asp-49 assists water molecule 111 in ligating calcium (Dijkstra er at., 1981b). The substrate phosphate oxygen displaces water molecule I1 (Verheij et al., 1980).

22

LINDA J . VAN ELDIK ET AL.

been proposed (Verheij et af., 1980). The mechanism may be the same for pancreatic and venom phospholipase since the active site amino acids are conserved. A schematic diagram of the proposed mechanism is shown in Fig. 8. The hydrolysis of the fatty acid ester by phospholipase A, is analogous to the h:ydrolysis of peptide bonds by serine proteases (Verheij et al., 1980; Dijkstra ef al., 1981a). In phospholipase A,, however, a water molecule replaces the serine hydroxyl oxygen as the nucleophile, and no covalent intermediate is formed between the enzyme and the substrate. Instead, aspartic acid-99 forms a hydrogen bond with N-T of histidine-48 causing N-" to be somewhat more nucleophilic. N-" of histidine-48 in turn ionizes a water molecule, and the hydroxyl oxygen attacks the carbonyl carbon of the.fatty acid ester. Calcium stabilizes the

I

I

PRODUCTS

FIG. 8. A model of the proposed catalytic mechanism of phospholipase A2 (modified from Verheij el al., 1980). A description of the catalytic steps appears in the text. Dashed lines to the calcium ion represent noncovalent interactions and do not indicate stereochemistry. Bonds to the 2position of the phospholipid are drawn to indicate stereochemical configuration.

CALCIUM-BINDING PROTEINS

23

ensuing tetrahedral intermediate. Verheij et al. (1980) have speculated that the NH group of glycine-30 also helps to stabilize the intermediate. AND MICELLAR SUBSTRATES C. CALCIUM

Phospholipase A, from mammalian pancreas has much higher activity with phospholipid substrates organized into micelles than with monomeric phospholipids (Pieterson et al., 1974b). Calcium is required for the interaction between the enzyme and micellar lipid (van Dam-Mieras et al., 1975; Slotboom et al., 1978). The low-affinity calcium-binding site is postulated to be involved in micellar binding and to be located near the NH,-terminus of the enzyme (Slotboom el al., 1978). Two lines of evidence have indicated that the NH,-terminus of the enzyme is necessary for micellar binding. First, studies using chemical modification (Slotboom and de Haas, 1975; Slotboom er al., 1977) have directly demonstrated that the NH,-terminal L-alanine with an a-NH3+ group is necessary for micellar binding. Second, the pancreatic enzymes are secreted as zymogens which cannot utilize micellar substrates (Pieterson et al., 1974b; van Dam-Mieras et al., 1975). The zymogen is activated by the trypsin-catalyzed removal of an NH,terminal heptapeptide (de Haas et al., 1968a,b; Drenth et al., 1976) exposing the NH,-terminal alanine and leading to micellar binding. Thus, the activation of the zymogen and the property of micellar binding are dependent on proteolytic processing of the NH,-terminus. As noted above, the NH,-terminus is part of a postulated “interface recognition site” (Pieterson et al., 1974b) which is present on the active enzyme but not on the zymogen. Physical studies of the enzyme confirm and extend this postulate. First, spectroscopic measurements have implicated tyrosine-69 in the binding of the enzyme to a lipid/water interface (Meyer er al., 1979a,b). Second, the three-dimensional structure of phospholipase contains a ring of amino acids surrounding the entrance to the active site and this ring structure has been proposed to be involved in micellar binding (Dijkstra et al., 1981a,b). This region of the enzyme structure includes the NH,-terminal tripeptide and tyrosine-69, as well as asparagine-6, leucines-19 and 20, asparagines-23 and 24, leucine-31, and threonine-70. A number of mechanisms have been invoked for the interaction of phospholipase A, with a lipid/water interface (Wells, 1974a; Tinker and Wei, 1979; Deems et al., 1975; Tinker et al., 1980; Verheij et al., 1981; Menashe et al., 1981). These mechanisms have been proposed to explain the kinetic data for phospholipase A, with substrates such as phospholipid micelles, monolayer films, and dispersions. Dennis (1975) introduced the use of mixed micelles of phospholipid and detergents to measure kinetic parameters of phospholipase A,. This technique allows the substrate concentration to be lowered below the critical micellar concentration while retaining a micellar surface. The exact nature of the

24

LINDA J . VAN ELDIK ET AL.

interaction of phospholipase A, with micelles and other lipid/water interfaces has not been elucidated. In this regard, phospholipase may become a model for soluble enzymes which act at membrane surfaces. As a calcium-binding enzyme, phospholipase A, is a useful example with which to compare other calciumbinding enzymes.

D. OTHERCALCIUM-BINDING, HYDROLYTIC ENZYMES The extracellular nuclease of Stuphy/ococ.cus uureus (ribonucleate, tleoxyribonucleate 3'-nucleotide hydrolase EC 3. I .4.7) has a pH optimum of 9- 10 and requires millimolar concentrations of calcium ions (Cotton and Hazen, 1971). Staphylococcal nuclease is a single polypeptide composed of 149 amino acid residues and binds 1 rnol of calcium ions per mol of protein (Cotton et a / ., 1979). Studies on the three-dimensional structure of the enzyme (Cotton el a / . , 1979) in conjunction with chemical modification data (Anfinsen et a / . , 1971) have elucidated some characteristics of the active site and of the calcium-binding site. The calcium site is octahedral, employs oxygen-containing side chains of amino acids and peptide carbonyl groups, and contains at least one water niolecule. The calcium ion is coordinated by the side chains of aspartic acids-19, 21, 40 and glutamic acid-43, the carbonyl oxygen of threonine-41 and a sixth ligand, possibly a water molecule (Cotton et a / . , 1979). Calcium coordination in staphylococcal nuclease is similar to that in the high-affinity phospholipase A2 calcium-binding site. In addition, the active site of both enyzmes is proximal to the calcium site. It is possible that the essential calcium in staphylococcal nuclease stabilizts an intermediate of the phosphodiester hydrolysis reaction in a fashion similar to that which has been postulated for phospholipase A, (Cotton e t a / . , 1979). However, such a role for calcium in staphylococcal nuclease has not been unequivocally demonstrated. The phospholipid substrate for phospholipase A, and the nucleotide substrate for nuclease both have phosphate groups. Interestingly, the guanidinium group of arginine-87 in staphylococcal nuclease and that of an unidentified arginine in phospholipase A, have been implicated in binding, the phosphate portion of the substrate (Vensel and Kantrowitz, 1980). Thus, tlhese two enzymes of vastly different origin and activity have striking similarities in their use of calcium ion at the active site. The proteolytic enzymes thermolysin, trypsin, and chymotrypsin also bind calcium (Matthews et a/.. 1974; Bode and Schwager, 1975b; Cohen et a / . . 1981). In thermolysin, 4 mol of calcium ions are bound per mol of prcltein (Matthews et al., 1974; Voordouw and Roche, 1974). The calcium ions are not required for enzymatic activity, but are required for the thermal stability of the enzyme. Two of the bound calcium ions are coordinated at a double site reminiscent of the calcium/Mn2+ double site of concanavalin A (see Section 111). As Matthews et a / . (1974) have pointed out, this analogy is not complete because the conformation of the protein in the double metal binding site of concanavalin

CALCIUM-BINDING PROTEINS

25

A seems to be quite different from that of thermolysin. Thermolysin contains Zn2+ at its active site. This divalent metal cation may act like the Zn2+ in carboxypeptidase A (Quiocho and Lipscomb, 1971) and the calcium in phospholipase A, to stabilize a tetrahedral intermediate formed from nucleophilic attack on the substrate. Trypsinogen binds two calcium ions, one with Kd = 0.2 mM and one with Kd > 10 mM (Delaage et al., 1968). The lower affinity calcium-binding site of trypsinogen is at aspartic acids-4 and 5 , adjacent to a peptide bond (lysine-6:isoleucine-7) which is cleaved during conversion of trypsinogen to trypsin (Abita and Lazdunski, 1969). Calcium ions activate the conversion of trypsinogen to trypsin by accelerating the autocatalytic hydrolysis of this lysinylisoleucine peptide bond and by protecting the enzyme from autodegradation at other susceptible bonds (Abita et al., 1969). The calcium-dependent regulation of trypsinogen activation occurs at 2&100 mM calcium (Keil, 1971) and involves the lower affinity site of the zymogen. The low-affinity calciumbinding site of phospholipase A, (Slotboom et af., 1978) may play a similar role in the activation of zymogen. However, the calcium-binding site is lost with the removal of the NH,-terminal hexapeptide of trypsinogen but is retained in active phospholipase A,. Trypsin binds a single calcium ion with Kd = 0.16 mM (Epstein et al., 1974). As determined by X-ray crystallography (Bode and Schwager, 1975a.b; Fehlhammer et al., 1977), the structure of the calcium-binding site in trypsin is the same as the higher affinity site (Kd = 0.2 mM) of trypsinogen. The calcium ion is coordinated by an octahedral arrangement of six ligands: a carboxylate oxygen from glutamic acids-70 and 80, the peptide carbonyl oxygens of asparagine-72 and valine-75, and two water molecules. One of the water molecules forms a hydrogen bond with the carboxylate moiety of glutamic acid-77 (Bode and Schwager, 1975a). This site will bind Mn2+ and various lanthanides, but not Mg2 or Sr2 (Epstein et al., 1974). The calcium ion at this site probably stabilizes the enzyme as do the ions in thermolysin (Sipos and Merkel, 1970). Calcium is not involved in the mechanism of catalysis in trypsin (Kraut, 1977). Based on sequence homology (Bode and Schwager, 1975a), physical and chemical properties (Delaage et al., 1968), and X-ray crystallographic structure (Birktoft and Blow, 1972; Cohen etal., 1981), chymotrypsin is thought to have a calcium-binding site similar to that of trypsin. However, proteolytic activation of chymotrypsinogen is not regulated by calcium. +

+

V. Calcium-Modulated Proteins A. PARVALBUMIN AND EF HANDCALCIUM-BINDING STRUCTURES Parvalbumin is a low-molecular-weight, acidic, calcium-binding protein found in high concentrations in muscle. The type 3 parvalbumin from carp

26

LINDA J . VAN ELDlK ET AL

muscle (also referred to as carp 4.25 due lo its p/ = 4.25) has been studied in greatest detail. The crystal structure has been determined and refined at I .9 A resolution (Moews and Kretsinger, 1975). The function of this protein is not known and the details of its structure and calcium binding have been discussed in a recent reveiw (Kretsinger, 1980a). However, popular theories of the evolution, structure, and function of calcium-modulated proteins are based on the structure and calcium-binding properties of parvalbumin. Therefore, we have summarized information about parvalbumin that forms the basis of these models of calciumbinding proteins. Carp parvalbumin binds 2 mol of calcium with a dissociation constant of approximately lo-' M (Pechere, 1977). The dissociation constant for Mg2+ M and that for Na+ is approximately 10W2M (Potter et al., 1977) is (Grandjean et al., 1977). The calcium-binding properties and the essential features of the polypeptide chain structures of other parvalbumins are assumed to be indistinguishable from that of the carp 4.25 protein (Kretsinger, 1980a). The parvalbumin structure consists of six a-helical regions denoted by letters PL-F. Calcium ions are bound by the peptide regions between helices C and D and between helices E and F. The structure formed by the E and F helices and the connecting calcium-binding loop (E-he1ix:loop:F-helix) is called an EF hand. The CD and EF hands each form octahedral calcium-binding structures that are related to each other by an approximate 2-fold axis. The EF octahedral structure can be envisioned schematically by a hand model as shown in Fig. 9. The E-helix is represented by the extended forefinger, the F-helix by the extended thumb, and the calcium-binding loop by the clenched middle finger. The knuckle of the middle finger represents the sharp bend at glycine-95 (Fig. 10). The six ligands of the octahedra are mostly side chain oxygen atoms of the protein (aspartic acid-90, aspartic acid-92, aspartic acid-94, and glutamic acid-101). A peptide carbonyl oxygen is the ligand at the -Y vertex (lysine-96) and water, possilbly bonded to glycine-98, is the oxygen ligand at the -X position of the EF loop. 'The approximate spatial relationship between the CD and EF hands can be thought of as two right hands in an antiparallel arrangement and with the palmar surfaces facing each other. In this arrangement, the palmar surfaces of the two thumbs and forefingers would face the hydrophobic core of the protein. The Kretsinger hypothesis or EF hand model of calcium-modulated proteins is based on the calcium-binding EF hand of parvalbumin. The details of this maldel as well as associated postulates have been discussed by Kretsinger in a recent review (Kretsinger, 1980a) and will not be reiterated here. The essence of the Kretsinger hypothesis is that the targets of calcium acting as a second messenger are proteins which contain EF hand structures. This structure is depicted schematically in Fig. 9 and discussed above. Each model EF hand domain is composed of 31 amino acid residues. Using the recommended (Kretsinger, 1980b) numbering scheme, the amino acid sequence of the EF hand region of carp

27

CALCIUM-BINDING PROTEINS

FIG.9. Model of the paired CD and EF hands. As discussed in the text, an octahedral calciumbinding structure that utilizes oxygen ligands of a calcium ion is formed by a he1ix:loop:helix arrangement of the polypeptide chain. The vertices of the octahedron are indicated by X, Y , 2, -Y, -X, -Z. The two a-helices are represented by the extended forefinger and the thumb. The clenched middle finger represents the loop that contains the oxygen ligands of the calcium ion. The palmar surfaces of the forefingers and the thumbs face the hydrophobic core of the protein. Figure 10 gives the amino acid sequence of the EF hand region of parvalbumin. Reproduced from Kretsinger (1980b).

parvalbumin (residues 80-108) is shown in Fig. 10. The residues (denoted by n) at positions 2, 5, 6, 9 and 22, 25, 26, 29 are on the inner face of the a-helices E and F, respectively. These are the surfaces which face the hydrophobic core of the protein in Fig. 10. The sequence positions of the calcium liganding structures are 10, 12, 14, 16, 18, and 21. These correspond, respectively, to the octahedral vertices X, Y, Z , -Y, -X, and -Z shown in Fig. 9. In parvalbumin and in the EF n n

n

n X

V

2

-V

-X

12

14

16

18

-2 n

n

n

n

Xodel

Position

0

2

4

6

8

10

20

22

24

26

28

Number

A-E-I-K-A-F-L-K-A-G-D-S-D-G-D-G-K-I-G-V-D-E-F-A-A-L-V-K-A Sequence Residue Number

80

90

100

108

Fic. 10. The amino acid sequence of the EF hand region of carp parvalbumin. Based on the Xray crystal structure of carp parvalbumin (Moews and Kretsinger, 1975). the amino acid sequence of the EF hand region (residues 8&108) is shown. The EF hand model position numbers are from Kretsinger (1980b). The residues denoted by n are on the inner face of the a-helices E and F. The octahedral vertices X, Y, Z, -Y,-X, and -Z are the calcium-liganding residues. The single letter code for the amino acids is as described in the legend to Fig. I .

30

28

LINDA I. VAN ELDlK ET AL.

hand model the oxygen atom at the -Y vertex is from the peptide carbonyl, so no restraint is placed on the type of side chain that can occupy this position. At model position 18 (vertex -X), there is a glycine residue in parvalbumin and a water molecule is the ligand at the -X vertex. Thus, in the model this position is occupied by a residue with an oxygen-containing side chain, which would be used for calcium liganding, or by a glycine side chain with water supplying the liganding oxygen. Finally, there is no residue at position 29 in the amino acid sequence of parvalbumin because the sequence ends at residue 108 (position 28). In the EF hand model this position would be neutral or hydrophobic residue. Based on the EF hand model, Kretsinger and Bany (1975) have predicted a three-dimensional structure for troponin C, the calcium-binding subunit of skeletal muscle troponin. Troponin C isolated from a variety of species has been shown to have 3 4 putative EF hand structures, and the number of EF hands equals the moles of calcium bound per mole of protein (for a more extensive discussion of troponin C and predicted structures, see Kretsinger, 1980a). Based on the extensive amino acid sequence homology between troponin C and calmodulin (Watterson et al., 1980), a predicted structure for calmodulin can be inferred (Kretsinger, 1980b). It is not clear from these models of multiple EF hand proteins how the pairs of hands interact with each other. X-Ray crystallographic studies of these and other calcium-modulated proteins should allow further refinement of this model. As discussed throughout Section V, all of the amino acid sequences of calcium-modulated proteins that have been elucidated have a putative EF hand. However, two facts should be noted: (1) only low-molecular-weight, acidic, calcium-modulatedproteins have been characterized to date, and (2) the vitamin D-dependent calcium-binding protein (see Section V,E) has only one EF hand type of structure but binds 2 mol of calcium. The second calcium ion is bound by a he1ix:loop:helix peptide structure that is similar to, but distinct from an EF hand structure. The vitamin D-dependent calcium-binding protein structure indicates that intracellular calcium-modulated proteins may utilize other structures for calcium-binding in addition to EF hands. B. CALMODULIN Calmodulin is the name proposed (Cheung, 1980) for a well-characterized protein that binds 4 mol of calcium per mol of protein, is found in most eukaryotic cells, has an amino acid sequence that varies little among eukaryotes, and has multiple, calcium-dependent activities (for a recent comprehensive review see Klee er al., 1980). Calmodulin’s ubiquitous distribution and highly conserved structure suggest that it may be playing a fundamental role in the mechanism of calcium action in cell function. Its ability to bind calcium ions with dissociation constants in the micromolar range and activate a variety of enzymes in a calcium-

CALCIUM-BINDING PROTEINS

29

dependent manner suggests that calmodulin may be a pleiotropic mediator of calcium-dependent cellular processes. The phosphodiesterase activator activity of this unique chemical structure was first described independently by Cheung (1970) and Kakiuchi and co-workers (1970). Based on more recent reports (Seamon, 1980), it seems that calmodulin was also isolated by Moore (1965) as part of the S 100 protein fraction of brain. However, no activity other than calcium binding was ascribed to this heterogeneous Sl00 protein fraction (see Section V,D). Teo and Wang (1973) later showed that the phosphodiesterase activator was a calcium-binding protein. Chemically homogeneous preparations of calmodulin were not available until the middle of the 1970s when Watterson et al. (1976) showed that the homogeneous bovine brain troponin C-like protein was probably the same protein as the phosphodiesterase activator protein, and when Klee (1977) purified the porcine brain phosphodiesterase activator protein to chemical homogeneity. Based on the observation that a single protein had multiple activities, many investigators realized that they and others might be studying the same protein. Thus, in the earlier literature there are multiple names for the same molecule: calcium-dependent regulator, modulator protein, troponin C-like protein, activator protein. As will be summarized in this section, several approaches to the study of calmodulin’s role in cell function are being used with a variety of biological systems. These approaches can be grouped into five arbitrary classes: (1) pharmacology, (2) immunochemistry, (3) biochemistry, (4) mutant analysis/ pathophysiology , and ( 5 ) quantitative cellular and molecular biology. The major pharmacological tools used in studies of calmodulin and cell function are phenothiazine and naphthalene sulfonic acid derivatives. Immunochemical methods have allowed the establishment of specific radioimmunoassays for calmodulin and the localization of calmodulin in various cell types by immunocytochemical procedures. Biochemical studies of the calcium-dependent activation of enzymes by calmodulin have helped clarify many controversial points concerning calmodulin regulation, have defined what is thermodynamically and kinetically allowed in models of calmodulin regulation, and have suggested new approaches to the study of calmodulin regulation. Analysis of mutant and pathophysiological systems has demonstrated that calmodulin biosynthesis is quantitatively regulated and has provided well-characterized biological systems for the further study of the role of calmodulin in cell function. Quantitative cellular and molecular biological studies have provided insight into the physiological importance of some of calmodulin’s in vitro activities and have provided the necessary tools and information for future biological studies. 1. Comparative Biochemistry Vertebrate calmodulin is an acidic, 148-residue, calcium-binding protein that contains no tryptophan, cysteine, phosphate, or carbohydrate (Fig. 11). It has

30

LINDA J. VAN ELDIK ET AL

Verte brn t o Invertebrate Plant Protozoan

Vertebrate

-

50-

*

*

*

t o *

70

-

q-N-P-T-E-A-E-~-q-D-M-I-N-E-V-PA-D.C-N-G-T-I-D-F-P-E-~-L-T-M-M-A-R-~-M-K-

Invertebrate Plant Protozoan

Vertebrate Invertebrate Plant

I

..

Y

-

Protozoan

120Vertebrate

*130*

*

*

120

-

-

E-KrL-T-D-E-B-V-o-E-~-R-E-Si-N-I-PC-D-G-q-V-~Y-E-~-F-V-~M-M-T-A-~ "

K-s-

Plant

P V

K-V-

Protozoan

I

Invertebrate

E

H

-

I

FIG. I I Comparison of the amino acid sequences of calmodulins from four sources. This figure is a composite of the amino acid sequences for calniodulins isolated from four phylogenetically diverse sources: vertebrate (Watterson er a / . . 1980) with the amidation state of residues 24 and 135 corrected (Watterson, unpublished); invertebrate (Jamieson ef a / ., 1980; Takagi er a / ., 1980); plant (Iverson and Watterson, in preparation); and protozoan (Yazawa CI a / . , 1981). A solid line for a sequence indicates that, except for the deletions listed below, no clear differences between that protein and the vertebrate protein have been demonstrated. Asterisks above selected amino acids indicate proposed calcium-liganding residues according to the model of Kretsinger ( 1980b). Lines above other residues indicate those amino acids whose side chains are on the interior face of the ahelices of the Kretsinger model. Apparent deletions are found in two calmodulins. The sequence of calmodulin from the invertebrate R. renifbrrnis (Jamieson e t a / . . 1980) has a deletion at position 3 while that of M. senile (Takagi er a / . , 1980) does not. The protozoan (T.pyriformis) calmodulin sequence has a deletion at position 146. The single letter code for the amino acids is as described in the legend to Fig. I , and K' =trimethyllysine.

CALCIUM-BINDING PROTEINS

31

four structural domains that are similar to each other in amino acid sequence and similar to the four domains of skeletal muscle troponin C. Comparison of the complete amino acid sequences of calmodulin and troponin C unequivocally demonstrates that these proteins form a subclass of closely related calciummodulated proteins. Each of the calmodulin domains contains a model EF hand calcium-binding structure. The demonstration of which residues are actually involved in calcium binding will depend on the data from crystallographic studies which are in progress (Cook et a / . , 1980; Kretsinger et a l . , 1980). Several preliminary reports (Perry et al., 1979; Kuznicki et al.. 1981) have suggested that functional domains of calmodulin involved in protein binding can be isolated by limited chemical or enzymatic cleavage. However, complete chemical characterization of these fragments is lacking and the best activity was less than I % of that of the native protein. Recently (Fok et al., 1981; Van Eldik and Fok, 1981; Van Eldik and Watterson, 1981) a major immunoreactive domain in vertebrate calmodulin has been demonstrated and one immunoreactive site within this domain has been synthesized. This immunoreactive domain (residues 137-143) is contained in the fourth structural domain and is the only functional domain in calmodulin that has been unequivocally demonstrated. Although postulations are plentiful, it has not been shown how calmodulin structural domains are related to calmodulin function. Bovine brain calmodulin was the first chemically homogeneous calmodulin available and was the first calmodulin for which a partial or complete amino acid sequence was elucidated. It has, therefore, served as a standard of comparison for other calmodulins and its amino acid sequence is that shown in Fig. 11. Amino acid sequence studies of vertebrate calmodulins from a variety of other mammalian sources (Dedman et al., 1978a; Grand and Perry, 1978; Grand et al., 1981; Kasai et al., 1980) indicate that mammalian calmodulins may be identical in amino acid sequence. There are two preliminary reports of nearly complete amino acid sequences determined for invertebrate calmodulins: Renilla reniformis (Jamieson et al., 1980) and Metridium senile (Takagi et al., 1980). R . reniformis calmodulin has an apparent deletion of the glutamine at position 3 while M . senile calmodulin does not (see Fig. 11); in all other respects, the two invertebrate sequences are indistinguishable from each other. The differences between vertebrate and invertebrate calmodulin are shown in Fig. 11. Excluding the deletion in R. reniformis calmodulin, there are six differences between vertebrate and invertebrate calmodulin. Three of the amino acid sequence differences are in amide assignments and all six differences could be accounted for by single nucleotide changes in the codons. A protozoan calmodulin has been purified from Tefrahymena pyriformis (Jamieson et al., 1979; Kumagai et al., 1980; Yazawa et al., 1981) and its amino acid sequence (Yazawa et al., 1981) is shown in Fig. 11. Like the invertebrate calmodulins, the majority of the sequence differences are found in domains 3 and 4.

32

LINDA J . VAN ELDIK ET AL.

There are only 13 differences between Tetrahyrnena and bovine brain calmodulin. These few differences are interesting since Tetrahyrnena guanylate cyclase is activated by Tetrahyrnena calmodulin but not by vertebrate or invertebrate calmodulins (Kakiuchi et d., 198 1). However, Tetrahyrnena calmod.ulin will activate vertebrate phosphodiesterase (Jamieson et af., 1979; Kakiuchi et al., 1981), indicating that there may be species specificity in some calmodulinenzyme interactions. The only amino acid sequence available for a plant calmodulin is a nearly complete sequence for spinach calmodulin (Iverson and Watterson, 198 1; Iverson et af., 1981). The established differences between spinach and bovine hrain calmodulins are shown in Fig. 1 1. There may be additional differences found as these studies are completed but it is obvious from inspection that calmodulin is highly conserved throughout vertebrates, invertebrates, protozoans, and higher plants. One interesting difference in plant calmodulins is the apparent presence of a glutamine residue at position 96. This relative position is occupied by a glycine in vertebrate, invertebrate, and protozoan calmodulins and in many other calcium-modulated proteins. It is not clear from the EF hand model of calciummodulated proteins how this might affect, if at all, the calcium-binding prciperties and activities of this calmodulin. Limited sequence analysis of barley calmodulin has also demonstrated the presence of a glutamine at this position (Schleicher and Watterson, 1981). The region containing the single histidine and trimethyllysine residues is identical among all calmodulins for which sequence data have been reported. Although there is only one plant calmodulin sequence available, several plant calmodulins have been purified and characterized in detail (Anderson er al., 1980; Iverson et al., 1981; Watterson er af., 1980; Schleicher er af., 1981). All of these plant calmodulins will quantitatively activate bovine brain phosphodiesterase and have a number of other functional as well as chemical similarities to vertebrate calmodulins (Anderson et af., 1980; Iverson and Watterson, 1981; Iverson et af., 1981; Schleicher et af., 1982). Several recent reports (Bazari and Clarke, 1981; Grand et al., 1980; Van Eldik et af., 1980) of' calmodulin-like proteins in plants have appeared. It is not clear whether these are calmodulins, modified calmodulins, or proteins that are structurally and functionally homologous to calmodulin. More detailed biochemical characterizations should help clarify the relationships among these proteins. 2 . Biochemical Activities While calmodulin has not been shown to possess any enzymatic activity, numerous calcium-dependent effector activities have been attributed to calmodulin. These in vitro activities include enzyme activity stimulation, proteinbinding activity, drug- and dye-binding activity, and associations with cytoskeletal structures. While some of the in vitro functions of calmodulin are appealing in terms of models of cell function, it is still not known for many of these activities

CALCIUM-BINDING PROTEINS

33

whether they reflect physiological roles for calmodulin. A recent review (Klee et af., 1980) has discussed the data and biochemical basis for many of these calmodulin activities. We have listed some of these activities below and have tried to summarize the approaches, with their strengths and limitations, used to study calmodulin activities. We have also tried to summarize the approaches being used to study the possible roles of calmodulin in cell function. It should be kept in mind that the very things that make calmodulin an exciting molecule to study (ubiquitous distribution among eukaryotes, highly conserved structure and function, multiple activities, and similarity to tissue-specific calcium-modulated proteins) also make unequivocal interpretation of biological studies difficult. The enzyme activator activities of calmodulin include the calcium-dependent stimulation of a cyclic nucleotide phosphodiesterase (Cheung , 1970; Kakiuchi and Yamazaki, 1970), an adenylate cyclase (Brostrom et al., 1975; Cheung et al.. 1975; Toscano et af., 1979), myosin light chain kinase (Dabrowska and Hartshorne, 1978; Hathaway and Adelstein, 1979). skeletal muscle phosphorylase kinase (Cohen et al., 1978; Shenolikar et af., 1979), invertebrate (Epel et af.,1981) and plant (Anderson et af., 1980) NAD kinases, several different ATPases (Blum et af.,1980; Dieter and Marme, 1981; Gopinath and Vincenzi, 1977; Hogaboom and Fedan, 1981; Jarrett and Penniston, 1977), various protein kinase activities (DeLorenzo et al., 1979; Kennedy and Greengard, 1981), a phospholipase activity (Wong and Cheung, 1979), a guanylate cyclase (Nagao et af.,1979), and glycogen synthase kinase (Payne and Soderling, 1980). Most of these activities have been demonstrated by one or more of the following approaches: first, addition of calcium and calmodulin with resultant stimulation of a process or activity; second, removal of an endogenous calcium-binding subunit and subsequent demonstration that calcium and calmodulin will stimulate the measured activity; or, third, purification to homogeneity of a calcium-sensitive enzyme activity and direct demonstration that calmodulin is the calcium-binding subunit. Many of the calmodulin regulated enzymes have not been purified to homogeneity. It is especially not clear if the regulation of some of the enzymes by calmodulin is a direct effect on the enzyme. Relatedly, it is not known if calmodulin is the endogenous calcium regulatory protein for all of these activities. The ability to demonstrate calcium-dependent calmodulin stimulation of an enzyme activity does not necessarily reflect a physiological function of calmodulin. For example, it has been demonstrated that calmodulin and troponin C are members of a class of structurally and functionally related proteins. Experiments with reconstituted actomyosin ATPase activity have shown (Amphlett et af., 1976; Dedman et al., 1977) that in the reconstitution assay calmodulin will quantitatively substitute for troponin C, the physiological calcium-sensitizing factor of actomyosin ATPase. Thus, it is possible that some of the other in vitro activities attributed to calmodulin may also be examples of calmodulin’s ability to substitute for an endogenous calcium-binding protein.

34

LINDA J. VAN ELDIK ET AL.

Some in vitro calmodulin activities clearly reflect physiological roles for calmodulin. Two examples are the regulation of myosin light chain kinase (Conti and Adelstein, 1981; Hathaway et a f . , 1981) and muscle phosphorylase kinase (Grand et af., 1981; Pichard et al., 1981; Shenolikar e t a l . , 1979). In both cases the holoenzyme was purified to chemical and enzymatic homogeneity, then the calcium-binding subunit was directly shown to be calmodulin. Current biological studies of these enzymes are concerned with defining the quantitative importance of the calcium+almodulin regulation under a variety of physiological and disease states. Little information is available on the species and tissue distributions of calmodulin regulated enzyme activities. Among other technical problems, these investigations are hindered by the variable presence of calmodulin-binding proteins that can block calmodulin activation' of an enzyme, thereby masking the presence or amount of activity. The availability of well-characterized preparations of calmodulin-regulated enzymes and antisera directed against these proteins would greatly facilitate such studies. The distribution of calmodulin-regulated enzymes among various tissues of the same organisp has been examined for only a couple of activities (Cheung et al., 1978; Guerriero and Means, 1981; Kakiuchi et al., 1978). Within the limitations of the methods used, it seems that some activities such as the cyclic nucleotide phosphodiesterase and the adenylate cyclase are readily detected only in certain tissues or cell types, but others, siich as the myosin light chain kinase activity, have been found in most tissues examined. Initial reports of phylogenetic studies of calmodulin-stimulated enzyme activities suggest species specificity in some calmodulin activities. The Tetrahymena guanylate cyclase activity is stimulated by Tetrahymena calmodulin but not by vertebrate or invertebrate calmodulin (Kakiuchi et al., 198l). Plant calmodulins were not tested. NAD kinase was first shown (Anderson and Cormier, 1978; Muto and Miyachi, 1977) in plant tissue extracts to be calciumdependent and stimulatable by calmodulin. This activity has recently been described (Epel et a f . , 1981) in an invertebrate tissue homogenate. Attempts to demonstrate a calmodulin-stimulated NAD kinase in vertebrate tissues have not been successful (M. J . Cormier, personal communication). In addition to purified calmodulin-regulated enzymes, calmodulin has been shown to bind to a variety of proteins in a calcium-dependent or independent manner. Some examples of these calmodulin-binding proteins are histone (Wolff er al., 1981), myelin basic protein (Grand and Perry, 1980), microvillus protein (Glenney and Weber, 1980), troponin I (Amphlett et al., 1976; Dedman et uf., 1977), gap junction protein (Hertzberg et u f . , 19811, spectrin (Sobue er u f . , 1980, 1981a), caldesmon (Sobue et al., 1981b), inhibitor protein (Sharma et uf., 1978), and calcineurin (Klee et al., 1979) or modulator binding protein (Wallace et al., 1978; Wang and Desai, 1977; Sharma et a f . , 1979). The ability of various proteins to interact with calmodulin has usually been demonstrated by calcium-

CALCIUM-BINDING PROTEINS

35

dependent interaction with calmodulin-Sepharose conjugates or by the binding of iodinated calmodulin to proteins before or after separation by polyacrylamide gel electrophoresis. These methods and their limitations have recently been discussed (Schleicher et af., 1982) and will not be reiterated here. As with the calmodulin-regulated enzymes, the demonstration of a calcium-dependent complex formation does not necessarily demonstrate a physiologically relevant activity. Related to calmodulin’s ability to bind to proteins and enzymes is its ability to affect the assembly and disassembly of supramolecular structures such as the cytoskeleton. It has been shown (Marcum et al., 1978; Job et af.,1981) that calmodulin will stimulate microtubule disassembly under certain in vitro conditions. In addition, studies using immunocytochemical procedures have demonstrated calmodulin localization associated with microtubules of the mitotic apparatus (see Section V,B,3). These observations have suggested that calmodulin may be involved in calcium-dependent effects on the cytoskeleton. However, the regulation of microtuble disassembly by calcium and calmodulin appears to be a complex process. First, different microtubule preparations show differences in sensitivity to calcium, depending on ionic strength, temperature, tubulin concentration, and presence of microtubule-associated proteins (Berkowitz and Wolff, 1981; Rebhun et al., 1980, Schliwa et al., 1981). Second, calmodulin has no effect on some microtuble preparations, such as those isolated from sea urchin spindles (Rebhun et af., 1980) or from unfertilized sea urchin eggs (Nishida and Kumagai, 1980). Further, in microtubule preparations that are sensitive to calmodulin, high concentrations of calmodulin relative to tubulin are usually required for in vitro disruption of microtubules (Marcum et af., 1978; Nishida et al., 1979). Third, the targets of calmodulin acting as a regulatory protein in microtubule disassembly have not been defined. Calmodulin has been reported to bind to various components of the microtubule system, including tubulin (Kumagai and Nishida, 1979; Kumagai et al., 1980) and microtuble-associated proteins (Jemiolo et al., 1980; Sobue et al., 1981~).However, it has been reported that the direct interaction between tubulin and calmodulin is weak (Perry et al., 1980; Jemiolo et al., 1980; Rebhun et af., 1980), troponin C can be as effective as calmodulin (Marcum et af., 1978), and microtubuleassociated proteins can affect the depolymerization activity of calmodulin (Jemiolo et al., 1980). Thus, the molecular basis of the calcium-dependent regulation of microtubule disassembly is not defined and the physiological significance of calmodulin regulation of microtubules must await further investigations in a variety of biological systems. Calmodulin has also been shown to bind various drugs and dyes, especially phenothiazine and naphthalene sulfonic acid derivatives, in a calcium-dependent manner (Hidaka et al., 1979, 1980; LaPorte et al., 1980; Levin and Weiss, 1979). Phenothiazines are heterocyclic compounds that are used phar-

36

LINDA J . VAN ELDIK ET AL

macologically as antipsychotics, antiemetics, and tranquilizers (Baldessarini, 1980). A number of phenothiazine derivatives bind to calmodulin in a calciumdependent manner and block calmodulin stimulation of several enzymes (Weiss et al., 1980). Levin and Weiss reported (1979) that the affinities of various phenothiazine drugs for calmodulin correlated with their pharmacological activity as neuroleptic agents and implied that calmodulin might be involved in the mechanism of drug action. However, the affinity of all classes of neuroleptics for calmodulin does not correlate with pharmacological potency (Norman el al., 1979). Most importantly, other calcium-modulated proteins, such as troporiin C and S 100, interact with phenothiazine derivatives in a calcium-dependent inanner (Marshak er al., 1981) and phenothiazines may block lipid activation of enyzmes (Mori er al., 1980). Thus, the pharmacological significance of phenothiazine binding to calmodulin is not established and experiments using phenothiazines as anti-calmodulin drugs must be interpreted with caution. Naphthalene sulfonic acid derivatives such as W-7 [N-(6-aminohexyl)-5-chloro- 1 -naphthalenesulfonamide] have been used pharmacologically as smooth muscle reiaxants (Hidaka et al., 1979) and have been shown (Hidaka et al., 1980) to block the calcium-dependent activation of enzymes by calmodulin. The literature on these compounds is more recent than that on phenothiazines, so specificity studies and quantitative binding studies analogous to those recently reported for phenothiazines have not been done. Until these studies have been completed, the same caveats concerning phenothiazine usage in biological studies should be applied to naphthalene sulfonic acid derivatives and any other drugs or dyes which are claimed to be specific for calmodulin.

3. Subcellular Localization In examining the role of calmodulin in cell function, it is necessary to determine the subcellular localization of calmodulin. The localization of calmodulin in various cells and tissues has been studied by three approaches: differential centrifugation studies of homogenates, biochemical cytology, and immunocytochemistry. The majority of the localization studies have been done on brain tissue where freely soluble calmodulin is found in high levels. Differential centrifugation studies (Egrie et al., 1977; Kakiuchi et al., 1978) of brain homogenates have demonstrated that the greatest percentage of calmodulin is in the soluble fraction, but that a portion of calmodulin is also localized in the particulate fractions. The particulate fraction localization data have been supported by the direct demonstration of the presence of calmodulin in a supramolecular structure, the postsynaptic density (Grab ei al., 1979). The localization of calmodulin in other systems, such as kidney and heart tissue (Kakiuchi et d., 1978), liver cells (Smoake et al., 1974), C6 glial cells (Brostrom and Wolff, 1974), and platelets (Smoake et al., 1974) has been determined by assaying

CALCIUM-BINDING PROTEINS

37

differential centrifugation fractions for phosphodiesterase activator activity. These later studies have confirmed that calmodulin-like activity is distributed between soluble and particulate fractions, with the majority of the calmodulin activity in the soluble fractions in these cell types. There are several experimental limitations to the interpretation of differential centrifugation studies of calmodulin activity. In these studies no attempt was made to define homogenization conditions which give maximal cell disruption while maintaining organelle integrity, and no detailed data using marker enzyme assays and quantitation of recovery were presented. It has been shown that the presence of calcium in the homogenization medium (Teshima and Kakiuchi, 1978; Watterson et al., 1976), the cell density (Evain et a f . , 1979), phosphorylation (Evain et a l . , 1979), or proteolysis (Evain et a f . . 1979) may influence the distribution of calmodulin between soluble and particulate fractions. In addition, caution is necessary in the interpretation of studies where the presence of calrnodulin is determined solely by phosphodiesterase activator activity since other proteins may also activate phosphodiesterase. For example, there is a recent report (Liang et a l . , 1981) of the isolation of a membrane-bound protein from Lirnufus amoebocytes which activated phosphodiesterase in a calcium-dependent manner. The protein was shown to be different from calmodulin by several criteria, including amino acid composition, molecular weight estimates, and immunoreactivity. Quantitative biochemical cytology studies of calmodulin have been done in one biological system (Van Eldik, 1982; Schleicher et al., 1982). Conditions were defined for maximal cell disruption of chicken embryo fibroblasts with maintenance of organelle integrity. The homogenization mixture was subjected to differential centrifugation to yield four fractions: N (nuclear), ML (mitochondrial, lysosomal), P (microsomal), and S (soluble). Recovery of protein in each fraction was assessed, marker enzyme assays were performed, and electron microscopic characterization of each fraction was done. The presence of calmodulin in each fraction was assayed by activation of phosphodiesterase, radioimmunoassay, and gel electrophoresis. These data demonstrated that under the homogenization and fractionation conditions used, the majority of calmodulin (77-93%) was present in the soluble fraction. However, a small, reproducible amount of calmodulin was found in each of the three particulate fractions (N, ML, and P), and preliminary studies (Schleicher et al., 1982; Van Eldik, 1982) indicated the presence of calmodulin-binding proteins in these fractions. The association of calmodulin with particulate fractions may represent an artifact of the experimental conditions employed, analogous to the possible nonspecific association of calmodulin with brain mitochondria during immunocytochemical processing (Wood et al., 1980) or redistribution of proteins during homogenization or processing (Taylor, 1981; Scheele er a l . , 1978). However, it is clearly

38

LINDA I . VAN ELDIK ET AL.

possible that the presence of calmodulin and calmodulin-binding proteins associated with the particulate fractions may be physiologically relevant. Immunocytochemical localization data have in general confirmed the available biochemical cytology studies. However, the precise distribution of calmodulin appears to vary with the cell type, cell cycle, or other conditions that affect the cell, such as hormonal stimulation (Welsh et al., 1979; Harper et af., 1980). It bas been reported (Dedman et af., 1978b) that calmodulin localization is coincident with stress fibers and is diffuse throughout the cytoplasm during interphase. Based on this localization of calmodulin it has been suggested (Welsh et al., 1979) that calmodulin may interact with actin in the mitotic apparatus. However, other investigators have not observed calmodulin localization coincident with stress fibers (Andersen et a l . , 1978), no direct actin-calmodulin interaction has ever been demonstrated, and microfilament inhibitors such as cytochalasin R do not affect calmodulin localization (Welsh et al., 1979). Therefore, the physiological importance of stress fiber localization of calmodulin is not known. In mitotic cells, the calmodulin immunofluorescence is localized to the mitotic apparatus and there is overlap of tubulin staining patterns with calmodulin localization during certain stages of the cell cycle (Welsh et al., 1979; De Mey et al., 1980). These data suggest that calmodulin may play some role in calciumdependent effects on the microtubules of the mitotic apparatus. There have also been reports of immunofluorescence localization of calmodulin in plasma membranes and nucleus (Harper et al., 1980), and immunocytochemical studies using electron microscopy (Lin et al., 1980) have demonstrated calmodulin localization in a variety of intracellular membranous structures, including nuclei, membranes of the Golgi apparatus, endoplasmic reticulum, plasma membrane, mitochondria, and postsynaptic membranes. The limitations of immunocytochemical studies such as the ones described above are (1) nonspecific staining can occur due to sticking of antigen or antibody to cellular structures, (2) "masking" of the immunoreactive site inay prevent the antibody from interacting with the antigen, (3) other proteins inay cross-react with the antibodies, and (4)redistribution of antigen may occur during fixation and processing. Within the limitations of immunocytochemistry, differential centrifugation studies, and biochemical cytology, it appears that calmodulin can be found both freely soluble and in association with a variety of particulate structures. Although the majority of the calmodulin in many cells appears to be readily soluble, the presence of even small amounts of calmodulin associated with intracellular structures may be physiologically important. Investigations on the subcellular localization of calmodulin, calmodulin-binding proteins, and calmodulin-regulated enzymes in a variety of cell types may provide further insight into the physiological significance of these multiple calmodulin localizations.

CALCIUM-BINDING PROTEINS

39

C. VITAMIND-DEPENDENT PROTEINAND MULTIPLE CALCIUM-BINDING STRUCTURES An important contribution to the understanding of the regulation and control of calcium homeostasis has been the elucidation of the vitamin D endocrine system (for a recent review see DeLuca, 1981). Vitamin D, acquired from dietary sources or from an ultraviolet-dependent conversion of 7-dehydrocholesterol in skin, is hydroxylated in the liver to 25-hydroxyvitamin D,. The 25-hydroxyvitamin D, is further hydroxylated in the kidney to produce the hormonally active form of the vitamin, 1,25-dihydroxyvitarnin D,. This active metabolite is then transported to its target tissues where it acts either directly or in combination with hormones such as parathyroid hormone to stimulate calcium transport, mobilization, and reabsorption. One of the major biological effects of vitamin D is the enhancement of intestinal calcium absorption. The exact mechanism by which vitamin D elicits the increase in calcium absorption is not known; however, one response to vitamin D administration is the synthesis of a calciumbinding protein (CaBP). The presence of intestinal CaBP was demonstrated in 1966 by Wasserman and Taylor when they found that administration of vitamin D to rachitic chicks resulted in an increase in a protein that bound calcium with high affinity (Kd = 10-s-10-6 M).The protein was subsequently purified from chick intestinal mucosa and its amino acid composition and physical characteristics determined (Wasserman et a l . , 1968; Ingersoll and Wasserman, 1971; Bredderman and Wasserman, 1974). Avian intestinal CaBP is a heat-stable, acidic protein (p/ = 4 . 2 4 . 3 ) of approximately 28,000 MW. It appears to have four high-affinity calcium-binding sites (Kd = 2 X M )and 20 to 30 low-affinity sites (Kd = 10-3 MI. Since the discovery of avian intestinal CaBP, vitamin D-dependent calciumbinding proteins have been detected in several other species and tissues (for review see Norman, 1979a,b; Siege1 et af., 1980; Taylor, 1980a). These CaBPs appear to be of two general types: proteins with molecular weights of approximately 28,000 (type I), and proteins with molecular weights of approximately 10,000(type 11). Type I CaBPs (M,= 28,000) have been found in both avian and mammalian species, whereas type 11 CaBPs (M,= 10,000) have so far been detected only in mammalian species (Wasserman and Feher, 1977). The mammalian type 11 CaBPs are similar to type I avian CaBP in that they are acidic proteins (p/ = 4.7) which bind calcium with high affinity. Circular dichroism studies (Fullmer and Wasserman, 1973; Donington et al., 1978; Birdsall et al., 1979) have shown that the type I1 CaBPs contain approximately 3WO% ahelicity; however, X-ray diffraction data (Szebenyi et d . . 1981) indicate that the helix content is approximately 63%. In contrast to the type I CaBPs, the mam-

40

LINDA J . VAN ELDIK ET AL.

malian intestinal CaBPs have molecular weights of 9,000-10,000 and possess M). In addition, these only 2 high-affinity calcium-binding sites (Kd = type I1 CaBPs lack half-cystine, histidine, methionine, and tryptophan, whereas the type I avian CaBPs contain all these amino acid residues. No amino acid sequence data for type 1 CaBP have been reported. However, type I1 CaBPs from porcine and bovine intestine have been purified to homogeneity (Fullmer and Wasserman, 1973; Hitchman et al., 1973; Dorrington e t a l . , 1974), their amino acid sequences determined (Hofmann et al., 1979; Fullmer and Wasserman, 1980, 1981), and the crystal structure of the bovine protein elucidated (Moffat et al., 1975; Jones et al., 1980; Szebenyi e t a l . , 1981). Figure 12 shows the amino acid sequences of the porcine and bovine intestinal CaBPs. Although the complete amino acid sequence has been reported only for the bovine and porcine intestinal CaBPs, preliminary chemical characterization has been done on CaBP from other species (Fullmer and Wasserman, 1975; Bruns et al., 1977, 1978; Ueng and Bronner, 1979; Oldham et al., 1980; Gleason and Lankford, 198 1). These data suggest that the structure of type I1 CaBPs is highly conserved in mammalian species. The porcine and bovine intestinal CaBPs have an acetylated NH,-terminus, possess 87% amino acid sequence identity, and contain two different types of calcium-binding structures. The amino acid residues shown to be involved in calcium binding are marked in Fig. 12 with an asterisk. Crystallographic studies on bovine intestinal CaBP have demonstrated (Szebenyi et al., 1981) the presence of one calcium-binding loop corresponding to an EF hand structure and a second structure distinct from an EF hand. The protein contains four a-helices, designated I-IV. Two of the helices (111 and IV) and the calcium-binding loop *

*

*

*

Bovine CaBP

10 20 30 Ac-S-A-K-K-S-P-E-E-L-K-G-I-F-E-K-V-A-A-K-E-G-D-P-N-u-L-S-K-E-E-L-K-L-L-L-a-T-~-F-P-

Porcine CaBP

-Q-A-S

1

q-I-A--

* 50 Bovine CaBP Porcine CaBP

40

*

* 60

*

*

* 70

S-L-L-K-G-P-S-T-L-O-E-L-F-E-E-L-D-K-N-G-D-G-E-V-S-F-E-E-F-q-V-L-V-K-K-I-S-(~ R-0-0

N

FIG. 12. Comparison of the amino acid sequences of vitamin D-dependent calcium-binding protein. The amino acid sequences of bovine (Fullmer and Wasserman, 1980; 1981) and porcine (Hofmann er al., 1979) intestinal calcium-binding protein are shown. Only those amino acid residues of the porcine sequence that differ from the bovine sequence are shown. Based on X-ray crystallographic studies (Szebenyi et al.. 1981) of the bovine minor A component (residues 4-78), those residues involved in calcium binding are denoted by asterisks (*). The single letter code for the amino acids is as described in the legend to Fig. 1.

CALCIUM-BINDING PROTEINS

41

connecting them form a calcium-binding structure similar to the EF hand structure of parvalbumin. The calcium is coordinated by oxygens from the side chains of aspartic acid-57, asparagine-59, aspartic acid-6 I , serine-65, and glutamic acid-68, and the carbonyl of glutamic acid-63. The other calcium-binding structure is formed by helices I and I1 and a connecting loop which is two residues longer than the Ill-IV loop. The calcium ligands are the main carbonyls of alanine- 18, glutamic acid-20, aspartic acid-22, and glutamine-25, and the side chain oxygens of glutaniic acid-30. The Ill-IV site calcium appears to be more readily removed than the 1-11 site calcium. Based on this observation, Szebenyi et ul. (1981) have suggested that the 1-11 may be a structural site that stays saturated with calcium under all physiological states, whereas the Ill-IV site is a candidate for a regulatory site which could bind and release calcium under physiological conditions. Because no function other than vitamin D-dependent calcium-binding activity is known for these proteins, immunochemical procedures are important in determining the presence and amount of CaBP in various tissues. Antibodies against avian type I and mammalian type I1 CaBPs have been prepared and radioimmunoassays and radial immunodiffusion assays have been developed (Murray et al., 1974; Arnold et al., 1975a,b; Marche et al., 1977; Christakos et al., 1979; Christakos and Norman, 1980a). By these immunochemical criteria, material that cross-reacts with CaBP has been detected in several species and tissues. It appears that type 1 and type 11 CaBPs are immunologically distinct. It is interesting to note that antibodies produced against chick intestinal CaBP (type 1) will cross-react with certain mammalian tissues (e.g., kidney and brain), whereas antibodies produced against mammalian intestinal CaBPs (type 11) appear to be species specific (Murray et a l . , 1974; Fullmer and Wasserman, 1975; Wasserman and Feher, 1977). For example, even though the porcine and bovine intestinal CaBPs share 87% amino acid sequence identity, they d o not cross-react immunologically. There is also a report that a CaBP isolated from rat skin does not cross-react with rat renal or intestinal CaBP (Laouari et a l ., 1980; Pavlovitch et a / . . 1980). Since most of these CaBPs have been detected only by their qualitative calcium-binding ability and/or by immunological criteria, it is not known whether they are structurally similar to the well-defined bovine and porcine intestinal CaBPs. In addition, until the amino acid sequence of avian CaBP is elucidated, it is unclear what the structural relationships are among types I and I1 CaBPs. There are preliminary reports of other vitamin D-dependent calcium-binding proteins in the intestine, including a calcium-dependent ATPase, p-nitrophenyl phosphatase, alkaline phosphatase, and calcium-binding membrane proteins (Haussler et u / . , 1970; Holdsworth, 1970; Melancon and DeLuca, 1970; Kowarski and Schachter, 1973, 1975, 1980; Moriuchi and DeLuca, 1976; Wil-

42

LINDA J . VAW ELIXK ET AL

son and Lawson, 1977; Miller et al., 1979; Alpers et a / . , 1980). However, no detailed structural characterizations of these proteins have been reported so lheir relationship to intestinal CaBP is not known. There have been conflicting reports on the cellular and subcellular localization of type I and 11 CaBP in the intestine (for review see Taylor, 1980a). The two extreme localization patterns are ( I ) CaBP present in goblet cells and the mucous coat of the absorptive cell surface, but not in the absorptive cell cytoplasm, and (2) CaBP present in the absorptive cell cytoplasm, but not in goblet cells. The apparent reason for these conflicting localizations has recently been explained by Taylor (1980b, 1981) and Jande et a/. (1980) who showed that the cellular distribution of avian CaBP can be affected by the fixation conditions used. They suggested that the presence of CaBP in the goblet cell and absorptive cell surface may be a result of artifactual redistribution during fixation and processing, and that the true localization of CaBP is in association with the absorptive cells. The role of CaBP in intestinal calcium absorption is unclear. It has been shown (Corradino et a / . , 1976) that calcium transport can be stimulated in an in 1,itr-o organ culture of embryonic chick intestine by the exogenous addition of CaBP. It has also been demonstrated that vitamin D induces synthesis of CaBP on intestinal polysomes (Emtage et a / . , 1974a,b; Spencer el al., 1976; Christakos and Norman, 1980b), and enhances CaBP-specific mRNA activity (Spencer e f al., 1978; Charles et a/., 1981). However, the synthesis of CaBP in response to vitamin D does not always correlate temporally with the increase in intestinal calcium absorption. Therefore, a direct involvement of CaBP in the regulation of calcium absorption remains to be established.

D. SIOO

S 100 refers to a heterogeneous fraction of low-molecular-weight, acidic proteins from brain that have calcium-binding activity. Moore and McGregor (1'365) discovered S 100 when fractions from ion exchange chromatography of soluble tissue extracts were analyzed by starch gel electrophoresis. Rat liver and brain extracts showed similar two-dimensional patterns except for a low-molecularweight, acidic protein fraction that appeared only in the brain preparation. Moore (1965) developed a procedure to isolate this fraction, which he termed S I00 to signify its partial solubility in saturated ammonium sulfate at neutral pH. Since this first report, a number of methods for the isolation of SlOO have been described (Combos et al., 1966; McEwen and Hyden, 1966; Uyernura et a/., 1971; Dannies and Levine, 1971a; Stewart, 1972; Mahadik et a / . , 1979). These involve routine techniques of protein chemistry such as salt fractionations, column chromatography, and, in some cases, preparative gel electrophoresis. Purified polypeptide components of the SlOO fraction have been obtained by more extensive use of these same procedures (Isobe et a l., 1977, 1981) or by the

43

CALCIUM-BINDING PROTEINS

combined use of salt fractionation, column chromatography, and affinity-based adsorption chromatography (Marshak et a l . , 1981). Isobe and co-workers (Isobe and Okuyama, 1981a) have elucidated the amino acid sequence of two SlOO components, termed SlOOa and Sloop (see Fig. 13). The SlOOa protein has 38 amino acid sequence differences from Sloop, 23 of which are compatible with single nucleotide substitutions, and S lOOa contains a tryptophan which is lacking in SlOOp. Both SlOOa and Sloop contain the sequence requirements for the potential formation of an EF hand calcium-binding loop and a calcium-binding loop similar to the second calcium-binding site of the vitamin D-dependent protein (Section V,C). The production of crystals suitable for X-ray diffraction studies has been reported (Kretsinger et a l . , 1980). However, structural data that would demonstrate the presence or absence of these two types of calcium binding-structures are not available yet. The function of S 100 protein fraction is not known, but a few in vitro activities have been reported. Apparent calcium-dependent activities include interaction with membranes (Donato and Michetti, 1981; Calissano and Bangham, 1971), binding to immobilized phenothiazines (Marshak er al., 1981), and stimulation of nucleolar RNA polymerase activity (Michetti et al., 1976). The only calciumdependent activity that has been examined with purified polypeptides is that of immobilized phenothiazine binding (Marshak et al., 1981). Because of a lack of easily assayable activities the presence of S 100 is usually monitored by immunoreactivity with heterologous antisera prepared against a heterogeneous protein mixture (Moore and Perez, 1966). These heterologous antisera react with more than one of the polypeptides in the SlOO mixture (Isobe et al., 1977). Recently, monospecific antisera prepared against purified polypeptide components have become available (Marshak and Van Eldik, unpublished), but no biological or immunochemical studies have been reported with these antisera. . * * 20 30 40 50 x-G-S-E-L-E-T-A-M-E-T-L-l-N-V-F-H-A-H-S-G-K-E-G-D-K-V-K-L-5-K-K-~-L-K-E-L-L-~-l-E-L-5-G-F-L-D-~-~-K-D1

SIOOa

siooa

10

A~-~-E-L-E-K-A-Y-Y-A-L-I-O-Y-F-H-~-Y-S-G-R-E-G-D-K-H-K-L-K-K-5-E-L-K-E-L-I-N-N-E-L-5-H-F-L-E-E-I-K-E-

*

60

.

t

*

t

70

ao

51000

A-D-A-V-D-K-Y-M-K-E-L-O-E-D-G-O-G-E-~-D-F-~-E-Y-V-Y-L-Y-A-A-L-T-Y-A-C-N-N-F-F-W-E-N-S

51003

~-E-V-V-D-K-Y-M-E-T-L-D-S-D-G-D-G-E-C-D-F-~-E-F-M-A-F-V-A-M-l-T-T-A-C-H-E-F-F-.-E-H-E

90

FIG. 13. Amino acid sequences of SlOOa and Sloop. The amino acid sequences are from lsobe and Okuyama (1981a). Dots in the sequence denote the presence of a gap introduced for the purpose of alignment. Asterisks (*) denote amino acid residues postulated to be involved in calcium binding based on homology with the vitamin D-dependent calcium-binding protein (see Fig. 12). The single letter code for the amino acids is as described in the legend to Fig. I , and x represents an unidentified blocking agent.

44

LINDA J . VAN ELDlK ET AL

Most studies of S 100 have utilized or monitored a heterogeneous mixture of proteins. As a result, a large body of confusing literature has evolved. Three guidelines are useful in evaluation of this literature. First, SlOO is a protein fraction and is not a well-defined chemical entity. The primary structures of two polypeptides found in most S 100 preparations have been determined (Isobe and Okuyama, 1981a) but it is not clear what percentage of the total SlOO fraction is represented by each of these structures. Variable amounts of calmodulin are also present in some SlOO preparations (Seamon, 1980). Second, since there is no known function for calmodulin-depleted S 100, its protein components should not be considered subunits of an enzymatic, structural, or regulatory protein complex. Third, most antisera to SlOO are not monospecific and do not define a single structure. Further, some antisera might react with calmodulin or calmodulin fragments present as contaminants. With these experimental limitations in mind, we have summarized information on the properties, localization, and possible functions of S 100. Calissano et a/. (1969, 1971) have used the fluorescence change observed upon addition or removal of calcium to estimate the number ( n = 2 ) and affinity (Kdapproximately 0.65 mM) of calcium-binding sites in S 100 fractions. However, using equilibrium dialysis and doing binding studies in the presence of I< , Calissano et a/. (1969) determined that S 100 had 6-7 calcium-binding sites per 24,000 molecular weight oligomer and estimated the dissociation constant for calcium to be 30-60 pM. Other studies (Calissano et a / . , 1969; Dannies and Levine, 1971b; Kessler et a / . , 1968; Starostina et a/., 1981; Mahadik et a / . , 1979) have shown that the physical, chemical, and immunological properties vary with the amount of calcium or sulfhydral reagents added to SlOO preparations. Among these properties that are altered are the number and relative mobilities of polypeptides detected by analytical gel electrophoresis (Calissano et a / ., 1969), the absorption of polarized light (Kessler et af.,1968) and reactivity with SlOO antisera (Dannies and Levine, 1971b). Based on the amino acid sequence homologies among S 100a,Sloop, vitamin D-dependent calcium-binding protein, and parvalbumin, Moffat and co-workers (Szenbenyi et a l . , 1981) have predicted that Sloop contains two calcium-binding structures, one EF hand type of structure and an additional he1ix:loop:helix calcium-binding structure similar to that found in vitamin D-dependent calcium-binding protein. Direct calcium binding and crystallographic studies of homogeneous SlOOa and Sloop should clarify this relationship between protein structure and calcium-binding activity. The cellular and subcellular localizations of S 100 have been investigated by immunocytochemical techniques and by direct isolation. Immunoreactive S I00 is restricted to nervous tissue and, within brain, more Sl00 can be isolated from white matter than from gray matter (Cicero e t a / . , 1970b; Gombos et a/., 1966). Most of the evidence from cellular localization studies indicates that S lCl0 is primarily localized in glial cells. For example, SlOO has been found in bulkprepared glia but not neurons (Bock, 1978), and extracts of isolated glia give +

CALCIUM-BINDING PROTEINS

45

immunoprecipitates with anti4 100 serum while extracts of neurons show no reaction (Hyden and McEwen, 1966). However, other cellular localizations of SlOO have been reported. By the fluorescent antibody sandwich technique, Sl00 has been localized in glial cytoplasm and in neuronal nuclei (Michetti er al., 1974). Immunofluorescence studies (Moore et al., 1977) have demonstrated SlOO labeling in the astrocytes and oligodendroctyes of rat and chick brains, as well as in certain groups of chick neurons at early developmental stages. SlOO has also been found in rat brain astrocytes using the peroxidase-labeled antibody technique (Bock, 1978; Ludwin et al., 1976). Gombos et al. (1966) have measured the amount of SlOO in sequential slices of bovine brain by quantitative densitometry of electrophoretograms and have determined that oligodendrocytes are richer in SlOO than astrocytes. These authors observed a rostro-caudal gradient of increasing S 100 as well as an increase in total S 100 in the direction of the spinal cord from the cerebral cortex. The specific areas richest in Sl00 were the optical tract and the cerebral pedunculi. Studies on the subcellular localization of SlOO have suggested that nuclei may be one intracellular location of S100. Michetti et al. (1974) have found SlOO in isolated nuclei from rabbit brain cortex. Some of this SlOO was associated with nucleoli (Michetti and Donato, 1981). In these studies nuclear SlOO comprised less than 0.6% of total S100, with the remaining 99% in the cytosol, and the isolated nuclei were shown to be permeable to S 100 in the presence of calcium. However, small amounts of SlOO in the nucleus may be significant. Michetti et al. ( 1976) have reported that SlOO stimulates nucleolar RNA polymerase activity in chick embryo brain nuclei. Perumal and Rapport (1978) have reported that a nuclear protein kinase activity which specifically phosphorylates S 100 is found in rat brain, and not in liver. Moore (1965) originally found no phosphorous in native bovine SIOO. SlOO may also associate with plasma membranes of neurons. It has been found (Donato, 1978; Donato et al., 1975) that 1251-labeledSlOO binds to synaptosoma1 fractions in a calcium-dependent manner. Monovalent cations, high pH, and heat will dissociate SlOO and synaptosomes. Donato (1978) has postulated that synapses contain receptors for SlOO and that these receptors show negative cooperativity. Haglid er al. ( 1974) have used immunocytochemical localization at the electron microscope level to detect labeling of postsynaptic densities with peroxidase-labeled antibody. However, Grab et al. (1979) have shown that isolated postsynaptic densities contain calmodulin, and Wood et al. (1980) have used antiserum against calmodulin to localize calmodulin in post synaptic densities. It is possible that some of the localization of SlOO to synaptic and other structures may be an artifact of contaminating calmodulin in S 100 preparations used as the antigen. Most SlOO antisera have not been tested for cross-reactivity with calmodulin. Hyden and Ronnback (1978, 1979) have employed Sepharose beads conjugated to SlOO antiserum and metal strips coated with SlOO antiserum to demonstrate binding of neurons to the antiserum. There have been other

46

LINDA 1. VAN ELDIK ET AL.

reports (Donato, 1978; Haglid and Stavrou, 1973) that SlOO associates with plasma membranes and that membrane-associated S 100 can be extracted with pentanol. Related to a possible membrane localization of S100, Calissano and Bangham ( 197 I ) have demonstrated a calcium-dependent induction of 86Rb efflux from liposomes by SIOO. This induction was dependent on protein, was inhibited by EDTA, and was most effective in phosphatidylserine liposorries. It has been postulated (Calissano and Bangham, 1971) that SlOO acts as a calciumdependent ionophore for 86Rb. Results of other phenomenological studies suggest that the levels of S 100 may change during tissue regeneration (Perez and Moore, 1968), cell proliferation (Gysin et al., 1980), development (Cicero er al., 1970a), and learning (Hyden and Lange, 1970; Hyden and Ronnback, 1979). To date, no function for SlOO has been proven although cellular and siubcellular localization studies and phenomenological studies of the S 100 fraction suggest potential functions. The availability of chemically homogeneous S 100 polypeptides and well-defined antisera against these structures should allow a more unambiguous interpretation of future studies on the localization and biosynthesis of Sl00 as well as facilitate the search for possible biochemical activities.

IV. Overview Four different classes of calcium-binding proteins which are involved in a variety of biological functions have been used as examples of how proteins are involved in the molecular basis of calcium action. The proteins containing ycarboxyglutamic acid provide an example of how a calcium-binding protein is involved in interface reactions and 1ipid:protein interactions. Concanavalin A and the calcium-binding lectins are examples of how calcium is involved in pr0tein:carbohydrate interactions and are proteins well-designed for potentially mediating chronic effects of calcium in development and differentiation. The calcium-binding hydrolytic enzymes are one of the few cases where calcium is directly involved in enzymic catalysis. Our detailed knowledge of phospholipase A, also provides another precedent for how calcium, proteins, and lipids interact in mediating the biological effects of calcium. Finally, the calcium-modulated proteins are excellent examples of how proteins mediate the acute, regulatory effects of calcium. For some classes of calcium-binding proteins, unifying hypotheses that attempt to relate calcium-binding structures to calcium-binding activity have emerged in recent years. These models have been a significant addition to our understanding of how calcium is involved in biological function. However, it is not possible yet to examine the structure of a protein and predict exactly what kind of calcium-binding activities it will exhibit. Relatedly, it is not possible to predict from the known calcium-binding properties of a protein what types of

CALCIUM-BINDING PROTEINS

47

calcium-binding structures it will possess. It is clear, however, that wrapping the polypeptide chain around the calcium ion to form a polyhedral site results in a higher affinity binding than that from a simple posttranslational modification, such as carboxylation or phosphorylation of an amino acid side chain. It is probable that more ion-binding specificity also results from the use of a polyhedron. Within the group of proteins that utilize a polyhedral structure, it is not clear whether affinity and selectivity are affected by the preponderant use of oxygen-containing side chains or carbonyl oxygens as calcium ligands. These questions must be addressed as our knowledge of molecular details increases and new models proposed and existing models refined. An interesting trend that has emerged as more calcium-binding proteins have been studied in detail is the number of calcium-binding proteins that are not enzymes and, within calcium-binding enzymes, the comparatively few cases of direct involvement of calcium in the enzyme-catalyzed reaction. Many of calcium’s chronic and acute regulatory functions appear to be through nonenzymatic effector proteins or effector protein domains on an enzyme. In this type of regulation calcium would bind to a calcium-binding structure that is distinct from an enzyme active site. The conformational or configurational difference between the apoprotein structure and the ca1cium:protein structure would be the difference in activity states. It is important to note that the difference in activity can be between no activity and measureable activity, or between two distinct activities. For example, a calcium-binding protein kinase with no calcium bound might utilize one substrate, but with calcium bound might utilize a different substrate. Thus, depending on the substrate being used, increasing calcium concentrations may appear to stimulate or inhibit activity. Because many calcium-binding proteins are not enzymes or are enzymes in which calcium is not directly involved in the enzyme reaction, binding assays, effector activity assays, and immunoassays are important in investigations of the role of these proteins in cellular and organismic functions. Multiple assays, e.g., effector activity and immunoreactivity, are often done in biological studies of such calcium-binding proteins. This is especially important when the synthesis of the protein can be affected by the physiological state of the organism. Biological systems in which the expression of the calcium-binding protein can be varied have been invaluable in past investigations and carefully designed experiments using these systems should continue to reveal information about how the various calcium-binding proteins are involved in biological function. Prospects for future research directions in calcium regulation require a thorough knowledge of the biological and chemical precedents which exist. Models for the role of calcium and calcium-binding proteins in cell function cannot ignore the extensive thermodynamic, kinetic, and structural data which are available. Comparative biochemistry continues to provide understanding of the possible molecular mechanisms of calcium regulation as well as defining in chemical

48

LINDA J. VAN ELDIK ET A L

terms the various targets of calcium in biological systems. The manner in which calcium-binding proteins interact with other macromolecules, lipids, and carbohydrates has received only limited study in the past, but these studies indicai:e the biological importance of such interactions. The molecular genetics of calciumbinding proteins and the regulation of gene expression should also be a fruitful area for future research. Knowledge of how calcium and other low-molecularweight regulators control the synthesis and degradation of calcium-binding proteins should provide insight into the homeostatic mechanisms operating on the calcium-binding proteins themselves. A number of different calcium-binding proteins exist both inside and outside cells and are often present only under certain conditions or in limited types of cells. Elucidating the integration of the various calcium-binding proteins and the processes they affect is fundamental to our understanding of calcium's role in the homeostasis of living organisms.

ACKNOWLEDGMENTS We acknowledge our many colleagues for providing information useful in the preparation of this article. We also thank W. Burgess. J . Corbin, and J. Suttie for helpful criticism. Finally, we thank C. Coogan and J. Morris for their assistance in preparation of the manuscript.

REFERENCES Abita. J . P . . and Lazdunski, M . (1969). Biochem. Biophys. Res. Commun. 35, 707-712. Abita, J . P., Delaage. M., Lazdunski, M., and Savrda, J. (1969). Eur. J. Biochem. 8, 314-322. Aggeler. P. M., White, S. G.. Glendening, M . B., Page, E. W.. Leake. T . B.. and Bates. G . (1952). Proc. Soc. Exp. Biol. Med. 79, 692-694. Aguiar, A.. de Haas, G.H . , Janse. E. H. J. M . , Slotboom, A . J., and Williams, R . J. P. (1979). Eur. J. Eiochcm. 100, 51 1-518. Almquist, H . J . (1975). Am. J . Clin. Nutr. 28, 6 5 6 6 5 9 . Alpers, D. H., Grimme, N., Smith, R . , and Avioli, L. V. (1980). Gastroenterologj 79,259-264. Amphlett. G . W., Vanaman. T. C . , and Perry, S. V. (1976). FEES Lett. 72, 163-168. Andersen, B.. Osborn, M . , and Webcr. K. (1978). C'vtobiologie 17, 354-364. Anderson. J . M., and Cormier, M. J. ( 1978). Biochern. Eiophys. Rt..s. Cornnun. 84, 595-602. Anderson. J. M . , Charbonneau. H., Jones, H. P., McCann, R . 0 . . and Cormier. M . J. (1980). Biochemistrv 19, 3 1 13-3 120. Anfinsen. C . B.. Cuatrecasas, P., and Tdniuchi, H. (1971). In "The Enzymes" (P.D. Boyer, ed.), pp. 177-204. Academic Press, New York. Arnold. B. M., Kuttner, M . , Swarninathan. R . , Cane, A. D . , Hitchman, A. J. W., Harrison, J. E., and Murray. T . M. (1975a). Can. J . Phvsiol. Pharmacol. 53, 1129-1 134. Arnold, B. M . , Kuttner, M . . Willis, D. M., Hitchman. A . J. W . , Harrison, J. E.. and Murray, 'T.M. (1975b). Can. J . Phvsiol. Pharmacol. 53, 1135-1 140. Aronson. D. L., Mustafa, A. J . , and Mushinski. J . F. (1969). Biochim. Biophvs. Acra 188, 25-30.

CALCIUM-BINDING PROTEINS

49

Aschaffenburg, R., Blake, C. C. F., Bumdge, J. M., and Esnouf, M. P. (1977). J . Mol. Biol. 114, 575-579. Bajaj, S . P., Butowski, R. J., and Mann, K. G. (1975). J . Biol. Chem. 250, 2150-2156. Baldessarini, R. J . (1980). I n “The Pharmacological Basis of Therapeutics” (A. G. Gilman, L. S. Goodman, and A. Gilman, eds.), pp. 395418. Macmillan, New York. Bangerth, F. (1979). Annu. Rev. Phyroparhol. 17, 97-122. Bangham, A. D. (1961). Nature (London) 192, 1197-1 198. Barondes, S. H . (1981). Annu. Rev. Biochem. 50, 207-231. Barton, P. G., and Hanahan, D. J. (1969). Biochim. Biophys. Acra 187, 319-327. Bazari, W. L., and Clarke, M. (1981). J. Biol. Chem. 256, 3598-3603. Becker, J . W., Reeke, G. N., Wang, J. L., Cunningham, B. A , , and Edelman, G. M. (1975). J. B i d . Chem. 250, 1513-1524. Becker, J. W., Reeke, G. N., Cunningham, B. A,, and Edelman. G. M. (1976). Nature (London) 259, 406-409. Bell, R . G . (1978). Fed. Proc. Fed. Am. SOC. Exp. Biol. 37, 259S2604. Benarous, R., Gacon, G., Rabiet, M.-J., and Labie, D. (1980). I n “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (J. W. Suttie, ed.). pp. 49-53. Univ. Park Press, Baltimore, Maryland. Benson, B. J., and Hanahan, D. D. (1975). Biochemistry 19, 3265-3277. Benson, B. J . , Kisiel, W . , and Hanahan, D. J. (1973). Biochim. Biophys. Acra 329, 81-87. Berkowitz, S. A., and Wolff, J. (1981). J . Biol. Chem. 256, 1121611223. Birdsall, W. J . , Levine, B. A,, Williams, R. J. P., Fullmer, C. S., and Wasserman, R. H. (1979). Biochem SOC. Trans. 7, 702-703. Birktoft, J . J., and Blow, D. M. (1972). J . Mol. Biol. 68, 187-240. Blanchard. R . A., Furie, B., and Furie, B. C. (1980). I n “Vitamin K Metabolism and Vitamin KDependent Proteins” (J. W. Suttie, ed.), pp. 6 6 7 1 . Univ. Park Press, Baltimore, Maryland. Bloom, J. W., and Mann, K . G. (1978). Biochemistry 17, 4430-4438. Blum, J . J . , Hayes, A., Jamieson, G. A., and Vanaman, T. C. (1980). J. Cell B i d . 87, 386397. Bock, E. (1978). J . Neurochem. 30, 7-14. Bode, W., and Schwager, P. (1975a). FEES Lert. 56, 139-143. Bode, W . , and Schwager, P. (1975b). J. Mol. Biol. 98, 693-717. Botes, D. P., and Viljoen, C. C. (1974). J . Biol. Chem. 249, 3827-3835. Bredderman, P. J . , and Wasserman, R. H. (1974). Biochemisrry 13, 1687-1694. Brenkle, G. M., Carlisle, T. L., and Jackson, C. M. (1980). In “Calcium-Binding Proteins: Structure and Function” (F. L. Siegel, E. Carafoli, R. H. Kretsinger, D. H. MacLennan, and R. H. Wasserman, eds.), pp. 425432. Elsevier, Amsterdam. Briggs, R., Douglas, A. S.. MacFarlane, R. G., Dacie, J. V., Pitney, W. R., Merskey, C., and O’Brien, J. R. (1952). Br. Med. J . 2, 1378-1382. Brostrom, C. O . , and Wolff, D. J . (1974). Arch. Biochem. Biophvs. 165, 715-727. Brostrom. C. 0..Huang, Y.-C.. Breckenridge, B. M., and Wolff. D. J. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 64-68. Brown, R. D., 111, Brewer, C. F.. and Koenig, D. H. (1977). Biochemistry 16, 3383-3896. Bruns, M. E. H., Fliesher, E. B., and Avioli, L. V. (1977). J. Biol. Chem. 252, 4145-4150. Bruns, M. E. H.. Fausto, A., and Avioli, L. V. (1978). J. B i d . Chem. 253, 31863190. Calissano, P., and Bangham, A. D. (1971). Biochem. Biophys. Res. Commun. 43, 50&509. Calissano, P.. Moore, B. W., and Friesen, A. (1969). Biochemistry 8, 4318-4326. Cardin, A. D., Behnke, W. D., and Mandel, F. (1979). J. Biol. Chem. 254, 8877-8884. Cardin, A. D., Behnke, W. D., and Mandel, F. (1981). J. Biol. Chem. 256, 37-40. Carlisle, T. L., and Suttie, J. W. (1980). Biochemistry 19, 1161-1 167.

50

LINDA J. VAN ELDIK ET AL.

Cadisle, T. L., Moita, T., and Jackson, C. M. (1980). I n “Vitamin K Metabolism and Vitamin KDependent Proteins” (J. W. Suttie, ed.), pp. 58-62. Univ. Park Press, Baltimore, Maryland. Carr, S. A., Hauschka, P. V., and Bieman, K. (1981). 1.Eiol. Chem. 256, 9944-9950. Charles, M. A., Martial, J., Zolock, D., Monissey, R., and Baxter, J. D. (1981). Culcif. Tissue Int. 33, 15-18. Cheung, W. Y. (1970). Eiochem. Eiophys. Res. Cornmun. 38, 533-538. Cheung, W. Y. (1980). Science 207, 19-27. Cheung, W. Y., Bradham, L. S., Lynch, T. J., Lin. Y. M., and Tallant, E. A. (1975). Eiochem. Eiophys. Res. Commun. 66, 1055-1062. Cheung, W. Y., Lynch, T. J., and Wallace, R. W. (1978). Adv. Cyclic Nucleotide Res. 9,233-251. Christakos, S . , and Norman, A. W. (1980a). Methods Enzymol. 67, 500-503. Christakos, S., and Norman, A. W. (1980b). Arch. Eiochem. Eiophys. 203, 809-815. Christakos, S., Friedlander, E. J., Frandsen, B. R., and Norman, A. W. (1979). Endocrinology 104, 1495-1 503. Cicero, T. J., Cowan, W. M., and Moore, B. W. (1970a). Bruin Res. 24, 1-10, Cicero, T. J., Cowan, W. M., Moore, B. W., and Suntzeff, V. (1970b). Bruin Res. 18, 25-34. Cohen, G. H., Silverton, E. W., and Davies, D. R. (1981). J . Mol. Eiol. 148, 44-79, Cohen, P., Burchell, A., Foulkes, J. G., Cohen, P. T. W., Vanaman, T.C., and Naim, A. C. (1978). FEES Lett. 92, 287-293. Conti, M. A., and Adelstein. R. S. (1981). J . Eiol. Chem. 256, 3178-3181. Cook, W. J., Dedman, J. R., Means, A. R., and Bugg, C. E. (1980). J. Eiol. Chem. 255, 8152-8153. Corradino, R. A., Fullmer, C. S . , and Wasserman, R. H. (1976). Arch. Eiochem. Eiophys. 174, 738-743. Cotton, F. A., and Hazen, E. E.,Jr. (1971). I n “The Enzymes’’ (P. D. Boyer, ed.), pp. 15:C176. Academic Press, New York. Cotton, F. A., Hazen, E. E., Jr., and Legg, M. J. (1979). Proc. Nurl. Acad. Sci. U.S.A. 76, 2551-2555. Cunningham, B. A., Hemperly, J. J., Hopp, T. P., and Edelman, G. M. (1979). Proc. Nutl. Acad. Sci. U.S.A. 76, 3218-3222. Dabrowska, R., and Hartshome, D. J. (1978). Eiochem. Eiophys. Res. Commun. 85, 1352-1359. Dam, H. (1935a). Nature (London) 135, 652-653. Dam, H. (1935b). Eiochem. J . 29, 1273-1285. Dannies, P. S . , and Levine, L. (1971a). J . Biol. Chem. 246,62766283. Dannies, P. S., and Levine, L. (1971b). J . Eiol. Chem. 246, 62846287. Davie, E. W. (1980). In “Vitamin K Metabolism and Vitamin K-Dependent Proteins’’ (J. W. Suttie, ed.), pp. 3-7, Univ. Park Press, Baltimore, Maryland. Dedman, J. R.,Potter, J. D., and Means, A. R. (1977). J. Eiol. Chem. 252, 2437-2440. Dedman, J. R., Jackson, R. L., Schreiber, W. E., and Means, A. R. (1978a). J . Eiol. Chem. 253, 343-346. Dedman, J. R., Welsh, M. J., and Means, A. R. (1978b). J . Eiol. Chem. 253, 7515-7521. Deems, R. A., Eaton, B. R., and Dennis, E. A. (1975). J. Eiol. Chem. 250, 9013-9020. de Haas, G. H., Postema, N. M., Nieuwenhuizen, W., and Van Deenen, L. L. M. (1968a). Biochim. Eiophys. Actu 159, 1034 17. de Haas, G. H., Postema, N. M., Nieuwenhuizen, W., and Van Deenen, L. L. M. (1968b). Biochim Eiophys. Actu 159, 118-129. Delaage, M., Abita, J. P., and Lazdunski, M. (1968). Eur. J . Eiochem. 5 , 285-293. DeLorenzo, R.J., Freedman, S. D., Yohe, W. B., and Maurer, S . C. (1979). Proc. Nurl. Acud. Sci. U.S.A. 76, 1838-1842. DeLuca, H. F. (1981). Annu. Rev. Physiol. 43, 199-209.

CALCIUM-BINDING PROTEINS

51

de Metz, M.. Vermeer, C., Soute, B. A. M., and Hemker, H. C. (1981). J. Biol. Chem. 256, 10843-10846.

De Mey, I . , Moeremans, M., Geuens, G . , Nuydens, R., Van Belle, H., and De Brabander. M. (1980). I n “Microtubules and Microtubule Inhibitors” (M. De Brabander and J. De Mey, eds.), pp. 227-241. Elsevier, Amsterdam. Dennis, E. A. (1975). I n “Enzymes of Lipid Metabolism” (S. Gatt, L. Freysz, and P. Mandel. eds.), pp. 165-176. Plenum, New York. Dieter, P., and Marme, D. (1981). FEBS Letr. 125, 245-248. Dijkstra, B. W., Drenth, J., and Kalk. K. H. (1981a). Nature (London) 289, 604606. Dijkstra, B. W., Kalk, K. H., Hol, W. G. J., and Drenth. J. (1981b). J. Mol. Biol. 147, 97-123. Di Scipio, R. G., and Davie, E. W. (1979). Biochemistry 18, 899-904. Di Scipio, R. G.. Hermoden. M. A.. Yates, S. G., and Davie, E. W. (1977). Biochemistry 16, 698-706.

Dombrose, F. A., Gitel, S . N., Zawalich, K., and Jackson, C. M. (1979). J. Biol. Chem. 254, 5027-5040.

Donato, R. (1978). J. Neurochem. 30, 1105-1 I 1 I . Donato, R., and Michetti, F. (1981). J. Neurochem. 36, 1698-1705. Donato, R., Michetti, F., and Miani, N. (1975). Brain Res. 98, 561-573. Dorrington, K. J.. Hui, A., Hoffman, T., Hitchman, A. J., and Harrison, J. E. (1974). J. Biol. Chem. 249, 199-204. Dorrington, K. J., Kells. D. I. C., Hitchman, A. J. W.. Harrison, J. E., and Hofmann, T. (1978). Can. J . Biochem. 56, 492-499. Downing, M. R., Butowski. R. J.. Clark, M. M., and Mann, K. G. (1975). J . Biol. Chem. 250, 8897-8906.

Drenth, J., Enzing, C . M., Kalk, K. H., and Vessies, J. C. A. (1976). Narure (London) 264, 373-377.

Egrie, J. C.. Campbell, J. A., Flangas, A. L., and Siegel, F. L. (1977). J. Neurochem. 28, 1207-12 13.

Emtage, J. S., Lawson, D. E. M., and Kodicek, E. (1974a). Biochem. J . 140, 239-247. Emtage, J. S., Lawson, D. E. M.. and Kodicek, E. (1974b). Biochem. J. 144, 339-346. Enfield, D. L., Ericsson, L. H., Walsh, K. A., Neurath, H.. and Titani, K. (1975). Proc. Nail. Arad. Sci. U.S.A. 72, 16-19. Epel, D., Patton, C.. Wallace, R. W., and Cheung, W. Y. (1981). Cell 23, 543-549. Epstein, M., Levitzki, A., and Reuben, J. (1974). Biochemistry 13, 1777-1782. Esmon. C. T., Suttie. J. W., and Jackson, C. M. (1975). J. Biol. Chem. 250, 40954099. Esnouf, M. P., and Jobin, F. (1965). Thromb. Diarh. Haemorrh. (Suppl.) 17, 103-110. Esnouf, M. P., and Prowse, C. V. (1977). Biochim. Biophys. Acra 490, 471-476. Esnouf, M. P., Green, M. R.,Allen, H., Hill, 0.. Irvine, G., and Walter, S. J. (1978). Biochem. J. 174, 345-348.

Evain. D., Klee, C., and Anderson. W. B. (1979). Proc. Natl. Acad. Sci. U.S.A. 76, 3962-3966. Evenberg, A.. Meyer, H., Gaastra, W., Verheij, H. M., and de Haas, G. H. (1977). J. Biol. Chem. 252, 1189-1196.

Fehlhammer, H., Bode, W., and Huber, R. (1977). J. Mu/. B i d . 111, 415-438. Fernlund, P.. Stenflo, S., and Tufvesson. A. (1978). Proc. Narl. Acad. Sci. U.S.A. 75,5889-5892. Fernlund, P., and Stenflo, J. (1980a). In “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (J. w . Suttie, ed.), pp. 84-88. Univ. Park Press, Baltimore, Maryland. Fernlund, P., and Stenflo, J. (1980b). I n “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (J. W. Suttie, ed.), pp. 161-165. Univ. Park Press, Baltimore, Maryland. Fleer. E. A. M.. Verheij, H. M., and de Haas, G. H. (1978). Eur. J. Biochem. 82, 261-269. Fleer. E. A. M., Verheij, H. M.. and de Haas, G. H. (1981). Eur. J. Biochem. 113, 283-288.

52

LINDA J. VAN ELDIK ET AL.

Fok, K. F., Van Eldik, L. J., Erickson, B. W., and Watterson, D. M. (1981). In “Peptides: Synthesis, Structure, Function” (D. H. Rich and E. Gross, eds.), pp. 561-564. Pierce Chem. Co., Rockford, Illinois. Foriers, A., DeNeve, R., Kanarek, L.. and Strosberg, A. D. (1978). Proc. Nail. Acad. Sci. U.S.A. 75, 1136-1 139. Foriers, A., Lebrun, E., Van Rapenbusch, R., deNeve, R., and Strosberg, A. D. (1981). J . B i d . Chem. 256, 5550-5560. Friedman, P. A., Rosenberg, R. D., Hauschka, P. V., and Fitz-James, A. (1977). Biochim. Biophys. Acta 494, 271-276. Friedman, P. A., Hauschka, P. V., Shia, M. A., and Wallace, J. K. (1979). Biochim. Biophys. Aria 583, 261-265. Fullmer, C. S., and Wasserman, R. H. (1973). Biochim. Biophys. Aria 317, 172-186. Fullmer. C. S . , and Wasserman, R. H. (1975). Biochim. Biophys. Acta 393, 134-142. Fullmer, C. S., and Wasserman, R. H. (1980). In “Calcium-Binding Proteins: Structure and Function” (F. L. Siegel, E. Carafoli, R. H. Kretsinger, D. H. MacLennan, and R. H. Wasserman, eds.), pp. 363-370. Elsevier, Amsterdam. Fullmer, C. S . , and Wasserman, R. H. (1981). J. B i d . Chem. 256, 5669-5674. Gebauer, G . , Schiltz, E.. and Rudiger, H. (1981). Eur. J. Biochem. 113, 319-325. Gitel, S. N., Owen, W. G.. Esmon, C. T., and Jackson, C. M. (1973). Proc. Nail. Acati. Sci. U.S.A. 70, 1344-1348. Glenson, W. A., and Lankford, G. L. (1981). Anal. Biorhem. 116, 256-263. Glenney, J . R . , and Weber, K. (1980). J. B i d . Chem. 255, 10551-10554. Goldstein, 1. J., and Hayes, C. E. (1978). Adv. Carbohydr. Chem. Biochem. 35, 127-340. Combos, G., Vincendon, G., Tardy, J., and Mandel, P. (1966). C . R. Acad. Sci. (Paris) Ser. 13263, 1533-1535. Gopinath, R. M., and Vincenzi, F. F. (1977). Biochem. Biophys. Res. Commun. 77, 1203--1209. Grab, D. J . , Berzins, K., Cohen, R.. and Siekevitz, P. (1979). J. B i d . Chem. 254, 869@-8696. Grand, R. J. A., and Perry, S. V. (1978). FEBS Lett. 92, 137-142. Grand, R. J. A., and Perry, S. V. (1980). Biochemistry 19, 227-240. Grand, R . J . A., Nairn, A. C., and Perry, S. V. (1980). Biochem. J. 185, 755-760. Grand, R. J . A., Shenolikar, S., and Cohen, P. (1981). Eur. J. Biochem. 113, 359-367. Grandjean, L., Laszlo, P., and Gerday, C. (1977). FEBS Lett. 81, 376. Griep, A. E., and Friedman, P. A. (1980). In “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (J.W. Suttie, ed.), pp. 307-310. Univ. Park Press, Baltimore, Maryland. Grimaldi, J. J., and Sykes, B. D. (1975). J. Eiol. Chem. 250, 1618-1624. Grundberg, C. M., Lian, J. B., and Gallop, P. M. (1980). In “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (J. W. Suttie, ed.), pp. 153-156. Univ. Park Press, Baltimore, Maryland. Guerriero, V., and Means, A. R. (1981). Fed. Proc. Fed. Am. Sor. Exp. B i d . 40, 1738. Gysin, R . , Moore, B. W., Proffitt, R. T., Deuel, T. F., Caldwell, K., and Glaser, L. (1980). J. Biol. Chem. 255, 1515-1520. Haglid, K. G., and Stavrou, D. (1973). J. Neurochem. 20, 1522-1532. Haglid, K.. Hamberger, A,, Hansson, H.-A., Hyden, H., Persson, L., and Ronnback, L. (1974). Nature (London) 251, 532-534. Hanahan, D. J . , Joseph, M., and Morales, R. (1980). Biochim. Biophys. Acta 619, 640-649. Harper, J. F., Cheung, W. Y.,Wallace, R. W., Huing, H.-L., Levine, S. N., and Steiner, A. L. (1980). Proc. Natl. Acad. Sci. U.S.A. 77, 366-370. Hathaway, D. R., and Adelstein, R. S. (1979). Proc. Nad. Acad. Sci. U.S.A. 76, 1653-1657. Hathaway, D. R . , Eaton, C. R.. and Adelstein, R. S. (1981). Nature (London) 291, 252-254. Hauschka, P. V. (1977). Anal. Biochem. 80, 212-223.

CALCIUM-BINDING PROTEINS

53

Hauschka. P. V. (1980). In “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (J. W. Suttie, ed.), pp. 227-236. Univ. Park Press, Baltimore, Maryland. Hauschka, P. V., and Gallop, P. M. (1977). In “Calcium-Binding Proteins and Calcium Function” (R.H. Wasserman, R. A. Corradino. E. Carafoli, R. H. Kretsinger, D. H. MacLennan. and F. L. Siegel, eds.), pp. 338-347. Elsevier, Amsterdam. Hauschka, P. V., and Reid, M. L. (1978). J. Biol. Chem. 253, 9063-9068. Hauschka. P. V., Lian, J. B., and Gallop, P. M. (1975). Proc. Nut/. Acad. Sci. U.S.A. 72, 392553929, Hauschka, P. V., Friedman. P. A., Traverson, H. P., and Gallop, P. M. (1976). Biochem. Biophys. Res. Commun. 71, 1207-1213. Haussler, M. R., Nagode, L. A,, and Rasmussen, H. (1970). Nafure (London) 228, 1199-1201. Heinrikson. R. L., Krueger, E. T . , and Keim, P. S . (1977). J. B i d . Chem. 252, 4913-4921. Hemker, H. C., Veltkamp, J. J., Hensen, A., and Loeliger, E. A. (1963). Nature (London) 200, 589-590. Hemperly, J. J., Hopp, T . P., Becker, J. W., and Cunningham, B. A. (1979). J. B i d . Chem. 254, 6803-68 10. Henriksen, R. A., and Jackson, C. M. (1975). Arch. Biochem. Biophys. 170, 149-159. Hertzberg, E., Gilula. N. B.. Watterson. D. M., and Van Eldik, L. J. (1982). In preparation. Hidaka, H.. Yamaki, T., Totsuka, T . , and Asano, M. (1979). Mol. Phurmacol. 15, 49-59. Hidaka, H., Yamaki, T., Naka, M., Tanaka, T . , Hayashi, H., and Kobayashi, R. (1980). M o l . Pharmucol. 17, 66-72. Hitchman, A. J. W., Ken. M. K., and Harrison, J. E. (1973). Arch. Biochem. Biophvs. 155, 221-222. Hofmann. T., Kawakami, M., Hitchman. A. J. W., Harrison, J. E., and Dorrington, K. J. (1979). Can J. Biochem. 57, 737-748. Hogaboom, G. K., and Fedan, J. S. (1981). Biochem. Biophys. Res. Commun. 99, 737-744. Holdsworth, E. S. (1970). J. Membr. Biol. 3, 43-53. Hougie, C . , Barrow, E. M . , and Graham, J. B. (1957). J. Clin. Invest. 36, 485-495. Houser, R. M., Carey, D. J., Dus, K. M., Marshall, G. R.,and Olson, R. E. (1977). FEBS Len. 75, 226-230. Howard, J. B., and Nelsestuen, G. L. (1974). Biochrm. Biophys. Res. Commun. 59, 757763. Howard, J. B.. and Nelsestuen, G. L. (1975). Proc. Nafl. Acad. Sci. U.S.A. 72, 1281-1285. Hyden. H., and Lange, P. W. (1970). Proc. Nut/. Acad. Sci. U.S.A. 67, 1959-1966. Hyden, H., and McEwen, B. S. (1966). Proc. Natl. Acad. Sci. U.S.A. 55, 354-358. Hyden, H.. and Ronnback, L. (1978). J. Neurobiol. 9, 489-492. Hyden, H., and Ronnback, L. (1979). Behav. Neurol. Biol. 25, 371-379. lngersoll, R. J., and Wasserman, R. H. (1971). J. B i d . Chem. 246, 2808-2814. Isobe, T., and Okuyama, T. (1981a). Eur. J. Biochem. 116, 79-86. Isobe, T.. and Okuyama, T. (1981b). J. Neurochem. 37, 522-524. Isobe, T . , Nakajima, T . , and Okuyama, T. (1977). Biochim. Biophys. Acra 494, 222-232. Isobe, T., Ishioka, N.. and Okuyama, T. (1981). Eur. J. Biochem. 115, 469-474. Iverson, D. B., and Watterson, D. M. (1982). In preparation. Iverson, D. B., Schleicher. M.. Zendegui, J . G., and Watterson, D. M. (1981). Fed. Proc. Fed. Am. Soc. Exp. B i d . 40, 1738. Jackson, C. M. (1980). In “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (J. W. Suttie, ed.), pp. 16-27. Univ. Park Press, Baltimore, Maryland. Jackson, C. M . , Peng, C.-W. Brenckle, G. M., Jonas, A., and Stenflo, J. (1979). J. Biol. Chem. 254, 5020-5026.

54

LINDA J. VAN ELDIK ET AL.

Jamieson, G. A., Vanaman, T. C., and Blum, J. J. (1979). Proc. Narl. Acud. Sci. U.S.A. 76, 6471-75. Jamieson, G. A., Bronson, D. D., Schachat, F. H., and Vanaman, T. C. (1980). Ann. N.Y. Acud. Sci. 356, 1-13. Jande, S . S . , Tolnai, S., and Lawson, D. E. (1980). In “Calcium-Binding Proteins: Structure and Function” (F.L. Siegel, E. Carafoli, R. H. Kretsinger, D. H. MacLennan, and R. H. Wasserman, eds.), pp. 409-41 1. Elsevier, Amsterdam. Jarrett, H. W., and Penniston, J. T. (1977). Biochem. Biophys. Res. Commun. 77, 1210-1216. Jemiolo, D. K., Burgess, W. H., Rebhun, L. I., and Kretsinger, R. H. (1980). J. Cell Biol. 87, 248a. Job, D., Fischer, E. H., and Margolis, R. L. (1981). Proc. Nurl. Acud. Sci. U.S.A. 78,4679-4682. Jones, A. I. S., Szenbenyi, D. M., and Moffat, K. (1980). In “Calcium-Binding Proteins: Structure and Function” (F. L. Siegel, E. Carafoli, R. H. Kretsinger, D. H. MacLennan, and R. H. Wassennan, eds.), pp. 413-416. Elsevier, Amsterdam. Joubert, F. J. (1975a). Biochim. Biophys. Acra 379, 329-344. Joubert, F. J. (1975b). Biochim. Biophys. Acra 379, 345-359. Joubert, F. J., and Taljaard, N:(1980). Eur. J. Biochem. 112, 493499. Kakiuchi, S . , and Yamazaki, R. (1970). Biochem. Biophys. Res. Commun. 41, 1104-1 110. Kakiuchi, S., Yamazaki, R., and Nakajima, H. (1970). Proc. Jpn. Acud. 46, 587-592. Kakiuchi, S . , Yamazaki, R., Teshima, Y.,Uenishi, K., Yasuda, S., Kashiba, A,, Sobue, K., Ohshima, M., and Nakajima, T. (1978). Adv. Cyclic Nucleoride Res. 9, 253-263. Kakiuchi, S., Sobue, K., Yamazaki, R., Nagao, S., Umeki, S . , Nozawa, Y., Yazawa, kl., and Yagi, K. (1981). J. Biol. Chem. 256, 19-22. Kasai, H., Kato, Y., Isobe, T., Kawasaki, H., and Okuyama, T. (1980). Biomed. Res. 1, 248-264. Keil, B. (1971). In “The Enzymes” (P. D. Boyer, ed.), pp. 25G-275. Academic Press, New York. Keith, C., Feldman, D. S., Dagenello, S . , Glick, J., Ward, K. B., Jones, E. 0.. and Sigler, P. B. (1981). J. Biol. Chem. 256, 8602-8607. Kennedy, M. B., and Greengard, P. (1981). Proc. Narl. Acud. Sci. U.S.A. 78, 1293-1297. Kessler, D., Levine, L., and Fasman, G. (1968). Biochemistry 7, 758-764. Klee, C. B. (1977). Biochemistry 16, 1017-1024. Klee, C. B., Crouch, T. H., and Krinks, M. H. (1979). Proc. Narl. Acud. Sci. U.S.A. 76, 6270-6273. Klee, C . B., Crouch, T. H., and Richman, P. G. (1980). Annu. Rev. Biochem. 49,489-515. Kondo, K., Toda, H., and Narita, K. (1981a). J. Biochem. 89, 29-36. Kondo, K., Toda, H., and Narita, K. (1981b). J. Biochem. 89, 37-47. Kowarski, S . , and Schachter, D. (1973). J . Clin. Invest. 52, 2765-2773. Kowarski, S., and Schachter, D. (1975). Am. J. Physiol. 229, 1198-1204. Kowarski, S . , and Schachter, D. (1980). J . Biol. Chem. 255, 10834-10840. Kraut, J. (1977). Annu. Rev. Biochem. 46, 331-358. Kretsinger, R. H. (1980a). CRC Crir. Rev. Biochem. 8, 119-174. Kretsinger, R. H. (1980b). Ann. N.Y. Acud. Sci. 356, 14-19. Kretsinger, R. H., and Barry, C. D. (1975). Biochim. Biophys. Acra 405,40. Kretsinger, R. H., Rudnick, S . E., Sneden, D. A,. and Schatz, V. B. (1980). J. Biol. Chem. 255, 8154-8 156. Kumagai, H., and Nishida, E. (1979). J . Biochem. 85, 1267-1274. Kumagai, H., Nishida, E., Ishiguro, K., and Murofushi, H. (1980). J. Biochem. 87, 667-670. Kuznicki, J., Grabarek. Z., Brzeska, H., Drabikowski, W., and Cohen, P. (1981). FEES Len. 130, 14 1- 145.

Laouari, D., Pavlovitch, H.. Deceneux, G., and Balsan, S. (1980). FEBS Len. 111, 285-289.

CALCIUM-BINDING PROTEINS

55

LaPorte, D. C.. Wierman. B. M.. and Storm, D. R. (1980). Biochemistry 19, 3814-3819. Larson. A . E., Friedman, P. A., and Suttie. J . W. (1981 1. J. Biol. Chem. 256, 11032-1 1035. Levin, R. M.. and Weiss, B. (1979). J. Pharmurol. Exp. Ther. 208, 454-459. Levy, R. J . , Lian, J. B.. and Gallop, P. (1979). Biochem. Biophys. Reu. Commun. 91, 4 1 4 9 . Lian. J . B., and Friedman, P. A. (1978). J. B i d . Chem. 253, 6623-6626. Lian. J . B., and Heroux, K. M. (1980). I n “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (J. W. Suttie, ed.), pp. 245-254. Univ. Park Press, Baltimore, Maryland. Lian, J. B.. Skinner, M . , Glimcher, M. J., and Gallop. P. M. (1976). Biorhem. Biophys. Res. Commun. 71, 1207-1213. Lian, J. B., Prien, E. L., Jr., Glimcher. M. J.. and Gallop, P. M. (1977). J . Clin. Invest. 59, 1151-1157. Lian, J. B.. Levy, R. J., Levy, J. T., and Friedman, P. A. (1980). I n “Calcium-Binding Proteins: Structure and Function” (F. L. Siegel, E. Carafoli, R. H. Kretsinger, D. H. MacLennan, and.R. H. Wasserman, eds.), pp. 449-460. Elsevier, Amsterdam. Liang, S.-M., Liang, C.-M., and Liu, T.-Y. (198 I). J. B i d . Chem. 256, 49684972. Liener, I. E. (1976). Annu. Rev. PIanr fhysiol. 27, 291-319. Lin, C.-T.. Dedman. J. R., Brinkley, B. R., and Means, A. R. (1980). J . Cell B i d . 8 5 , 4 7 3 4 8 0 . Lind, P., and Eaker, D. (1981). Toxicon 19, 11-24, Lindquist, M. 1 . . and Hemker, H. C. (1978). Biochim. Biophvs. Acta 533, 318-326. Long, C.. and Penny, I. F. (1957). Biorhem. J . 65, 382-389. and Kirsch, W. M. (1980). I n “Vitamin K Metabolism and Vitamin Low, M., Van Buskirk, J. .I., K-Dependent Proteins” (J. W. Suttie, ed.), pp. 150-152. Univ. Park Press, Baltimore, Maryland. Ludwin, S . K., Kosek, J. C., and Eng, L. F. ( 1976). J. Comp. Neurol. 165, 197-208. McEwen. B. S.. and Hyden, H. (1966). J. Neurochem. 13, 823-833. Magnusson, S . , Sottrup-Jensen, L.. Petersen, T. E., Moms, H. R., and Dell, A. ( 1974). FEBS Lett. 49, 189-193. Magnusson, S., Petersen, T. E., Sottrup-Jensen, L., and Clays, H. (1975). In “Proteases and Biological Control’’ (E. Reich, D. B. Rifkin, and E. Shaw, eds.), Vol. 2, pp. 123-149. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. . Mahadik. S. P., Korenovsky, A., and Rapport, M. M. (1979). J. Neurochem. 33, 763-771. Marche. P., Pradelles, P., Gros, C., and Thomasset, M. (1977). Biochem. Biophys. Res. Commun. 76, 1020-1026. Marcum, J . M., Dedman. J . R.. Brinkley, B. R., and Means, A. R. (1978). Proc. Narl. Acad. Sci. U.S.A. 75, 3711-3775. Marsh, H. C., Boggs, N. T . , 111.. Robertson, P., Jr., Sarasua, M. M., Scott, M. E., Ten Kortenaar. P. B. W., Helpern. J. A., Pedersen, L. G.. Koehler, K. A., and Hiskey, R. G. (1980). I n “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (J. W. Suttie, ed.), pp. 137-149. Univ. Park Press, Baltimore, Maryland. Marsh. H. C.. Sarasua, M. M., Madar, D. A., Boggs, N. T.. 111.. Pedersen, L. G., Hiskey, R. G.. and Koehler. K. A. (1982). In “Proceedings of the Seventh American Peptide Symposium” ( D . H. Rich, ed.), pp. 277-286. Pierce Chemical Co.. Rockford, Illinois. Marshak, D. R., Watterson, D. M., and Van Eldik, L: J. (1981). froc. Narl. Acad. Sci. U.S.A. 78, 6793-6797. Matthews. B. W . , Weaver, L. H., and Kester, W. R. (1974). J. Biol. Chem. 249, 8030-8044. Mattock, P., and Esnouf, M. P. (1973). Nature (London) New Biol. 242, 90-92. Melancon, M. J., Jr., and DeLuca. H. F. (1970). Biochemistry 9, 1658-1664. Menashe, M., Lichtenberg, D., Gutierrez-Merino, C., and Biltonen, R. L. (1981). J. Biol. Chem. 256, 45414543.

56

LINDA J. VAN ELDIK ET AL.

Meyer, H.. Verhoef, H., Hendriks, F. F. A., Slotboom, A. I . , and de Haas, G. H. (1979a). Biochemistry 18, 3582-3588. Meyer, H., Puijk, W. C., Dijkman, R., Foda-Van der Hoorn, M. M. E. L., Slotboom, A. J . , and de Haas, G. H. (1979b). Biochemisrry 18, 3589-3597. Michetti, F.. and Donato, R. (1981). J. Neurochem. 36, 1706-1711. Michetti, F., Miani, N., DeRenzis, G.,Caniglia. A.. and Comer, S. (1974). J. Neurochem. 22, 239-244. Michetti, F., DeRenzis, G., Donato, R., and Miani, N. (1976). Brain Res. 105, 372-375. Miller, A., Ueng, T.-H., and Bronner, F. (1979). FEBS Lerr. 103, 319-322. Milstone, J. H. (1965). J. Gen. Physiol. 38, 757-769. Moews, P. C., and Kretsinger, R. H. (1975). J. Mol. Biol. 91, 201-228. Moffat. K., Fullmer, C. S., and Wasserman, R. H. (1975). J. Mol. Biol. 97, 661-664. Moore. B. W. (1965). Biochem. Biophys. Res. Commun. 19, 739-744. Moore, B. W., and McGregor, D. (1965). J. Biol. Chem. 240, 1647-1653. Moore, B. W., and Perez, V. J. (1966). J. Immunol. 96, 100&1005. Moore. B. W., Cimino. M., and Hartman. B. K. (1977). Proc. Inr. Soc. Neurochem. 6 , 31. Mori, T., Takai, Y., Minakuchi, R., Yu, B., and Nishizuka, Y. (1980). J. Biol. Chenr. 255, 8378-8380. Moriuchi, S., and DeLuca, H. F. (1976). Arch. Biochem. Biophys. 174, 367-372. Moms, H. R., Dell, A., Petersen, T. E., Sottrup-Jensen, L., and Magnusson, S. (1976). Biochem. J. 153, 663-679. Murray, T. M.. Arnold, B. M., Tam, W. H., Hitchman, A. J. W., and Harrison, J . E. (1974). Metabolism 23, 829-837. Muto, S., and Miyachi, S . (1977). Plant Physiol. 59, 55-60. Nagao, S., Suzuki, Y., Watanabe. Y., and Nozawa, Y.(1979). Biochem. Biophys. Res. Commun. 90, 261-268. Nelsestuen. G. L. (1976). J. Biol. Chem. 251, 5648-5656. Nelsestuen, G . L. (1978). Fed. Proc. Fed. Am. Soc. Exp. Biol. 37, 2621-2625. Nelsestuen, G. L., and Suttie, J . W. (1972). J. Biol. Chem. 247, 81768182. Nelsestuen, G.L., Zytkovicz, T.H., and Howard, 1. B. (1974)’. J. Biol. Chem. 249, 6347-6350. Nelsestuen, G.L., Broderius, M., Zytkovicz, T. H., and Howard, J . B. (1975). Biochem. Biophys. Res. Commun. 65, 233-240. Nelsestuen, G. L., Kisiel, W., and Di Scipio, R. G. (1978). Biochemistry 17, 2134-2138. Nelsestuen, G. L., Resnick, R. M., Kim, C. S . , and Pletcher, C. H. (1980a). In “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (1. W. Suttie, ed.), pp. 28-38. Univ. Park Press, Baltimore, Maryland. Nelsestuen, G. L., Resnick, R. M., Wei, G. J . , Pletcher, C. H., and Bloomfield, V. A. (1980b). I n “Calcium-Binding Proteins: Structure and Function” (F. L. Siegel, E. Carafoli, R. H. Kretsinger, D. H. MacLennan, and R. H. Wasserman, eds.), pp. 433-440. Elsevier, Amsterdam. Nishida, E., and Kumagai, H. (1980). J. Biochem. 87, 143-151. Nishida, E., Kumagai, H., Ohtsuki, K., and Sakai, H. (1979). J. Biochem. 85, 1257-12665. Nishimoto, S. K.. and Price, P. A. (1980). J. Biol. Chem. 255, 6579-6583. Norman, A. W. (1979a). “Vitamin D. The Calcium Homeostatic Steroid Hormone.’’ Academic Press, New York. Norman, A. W. (1979b). Am:J. Med. 67, 989-998. Norman, J . A., Dmmmond, A. H.. and Moser, P. (1979). Mol. Pharmacol. 16, 1089-1091. Oldham, S. B., Mitnick, S. A., and Coburn, J . W. (1980). J. Biol. Chem. 255, 5789-5794. Olson, R. E., and Suttie, J. W. (1978). Viram. Horm. 35, 59-108. Olson, R. E.,Houser, R. M., Searcy, M. T.. Gardner, E. J., Scheinbuks, J., Subba Rao, 0.N., Jones, J. P., and Hall, A. L. (1978). Fed. Proc. Fed. Am. SOC.Exp. Biol. 37, 2610-2614.

CALCIUM-BINDING PROTEINS

57

E., Magnusson, S., and Sottmp-Jensen. L. (1980). In “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (J. W. Suttie, ed.), pp. 8-12. Univ. Park Press. Baltimore, Maryland. Ostemd, B., and Flengstrud, R. (1975). Biochem. J. 145.46-74. Owens, C. A., Magath, T. B.. and Bollman. J. L. (195 I j. Am. J . Phvsiol. 166, 1-1 I . Papahadjopoulos, D., and Hanahan, D. J. (1964). Biochim. Biophys. Acfa 90, 436-439. Papahadjopoulos, D., Hougie, C., and Hanahan, D. J. (1962). Proc. SOC. Exp. Biol. Med 111, 4 12-416. Pavlovitch, J . H., Laouari, D., Didierjean. L., Saurat, J. H., and Balsan, S. (1980). In “CalciumBinding Proteins: Structure and Function” (F. L. Siegel, E. Carafoli, R. H. Kretsinger, D. H. MacLennan, and R. H.Wasserman, eds.), pp. 417-419. Elsevier, Amsterdam. Payne, M. E., and Soderling, T. R. (1980). J. &of. Chem. 255, 80544056. Pechere, J.-F. (1977).In “Calcium-Binding Proteins and Calcium Function” (R. H. Wasserman, R. Corradino, E. Carafoli, R. H. Kretsinger, D. MacLennan, and F. Siegel, eds.), p. 213. Elsevier, Amsterdam. Perez, V . J . , and Moore, B. W. (1968). J. Neurochem. 15, 971-977. Perry, G., Brinkley, B. R., and Bryan, J. (1980). 1.Cei/ Biol. 87, 253a. Perry, S. V.. Grand. R. J. A., Nairn, A. C., Vanaman, T. C., and Wall, C. W. (1979). Biochem. Soc. Trans. 7, 619-622. Perumal, A. S., and Rapport, M. M. (1978). Life Sci. 22, 803-808. Petersen, T. E., Thogersen, H. C., Magnusson, S.,and Sottrup-Jensen, L. (1980). In “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (J. W. Suttie. ed.), pp. 171-174. Univ. Park Press, Baltimore, Maryland. Pichard, A. L.. Daegelen-Proux, D., Alexander, Y..and Dreyfus, J. C. (1981). Biochim. Biophys. Acfa 657, 84-93. Pieterson, W. A.. Volwerk, J. J.. and de Haas, G. H. (1974a). Biochemistry 13, 1439-1445. Pieterson, W. A., Vidal, J. C., Volwerk, J. J.. and de Haas, G. H. (1974b). Biochemistry 13, 1455- 1460. Poser. J. W., and Price, P. A. (1979). J. Biol. Chem. 254, 431-436. Poser, J. W., Esch, F. S . , Ling, N. C., and Price, P. A. ( 1980). J. Biol. Chem. 255, 8685-8691. Potter. J. D.. Johnson, J . D., Dedman, J. R . . Schreiber, W. E., Mandel, F., Jackson, R. L., and Means, A. R. (1977). In “Calcium-Binding Proteins and Calcium Function” (R. H. Wasserman, R. Corradino, E. Carafoli, R. H. Kretsinger, D. MacLennan, and F. Siegel, eds.), p. 239. Elsevier, Amsterdam. Prendergast, F. G . . and Mann, K. G. ( 1977). J. B i d . Chem. 252, 840-850. Prendergast, F. G., Bloom, J., Downing, M. R., and Mann, K. G. (1980). In “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (J. W. Suttie, ed.). pp. 39-48. Univ. Park Press, Baltimore, Maryland. Price, P. A. (1980). In “Calcium Binding Proteins: Structure and Function” (F. L. Siegel, E. Carafoli, R. H. Kretsinger, D. H. MacLennan, and R. H. Wasserman, eds.), pp. 4 4 4 8 . Elsevier, Amsterdam, Price. P. A., and Baukol, S. A. (1980). J. Biol. Chem. 255, 11660-1 1663. Price, P. A., and Nishimoto, S. K. (1980). Proc. Nurl. Acad. Sci. U.S.A. 77, 2234-2238. Price, P. A., Otsuka. A. S . , Poser, J. W., Kristaponis. J., and Raman, N. (1976a). Proc. Natl. Acad. Sci. U . S . A . 73, 147-1451, Price, P. A., Poser. J. W., and Raman, N. (1976b). Proc. Natl. Acad. Sci. U.S.A. 73,3374-3375. Price, P. A., Otsuka, A . S., and Poser, J. W. (1977). I n “Calcium Binding Proteins and Calcium Function” (R. H.Wasserman, R. Corradino, E. Carafoli, R. H. Kretsinger, D. MacLennan. and F. Siegel, eds.), pp. 333-337. Elsevier, Amsterdam. Price, P. A., Lothringer, J. W., and Nishimoto, S. K. (1980a). J. Biol. Chem. 255, 2938-2942. Olsson, G., Lindquist. 0..Petersen, T.

58

LINDA J. VAN ELDIK ET AL.

Price, P. A., Epstein, D. J., Lothringer, J. W., Nishimoto, S. K., Poser, 1. W., and Williamson, M. K. (1980b). In “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (J. W. Suttie, ed.), pp. 219-226. Univ. Park Press, Baltimore, Maryland. Price, P. A., hthringer, J. W., Baukol, S. A., and Reddi, A. H. (1981). J . B i d . Chem. 256, 3781-3784. Prowse, C. P., and Esnouf, M. P. (1977). Biochem. Soc. Trans. 5, 255-256. Quiocho, F. A., and Lipscomb, W. N. (1971). Adv. Pro?. Chem. 25, 1-78. Rebhun, L. I., Jemiolo, D. K., Keller, T. C. S., Burgess, W. H., and Kretsinger, R. H. (1980). In “Microtubules and Microtubule Inhibitors” (M. DeBrabander and J. DeMey, eds.), pp. 243-252. Elsevier, Amsterdam. Reddy, S., and Suttie, J. W. (1980). I n “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (J. w. Suttie, ed.), pp. 255-258. Univ. Park Press, Baltimore, Maryland. Reeke, G. N., Becker, J. W., and Edelman, G. M. (1975). J . Biol. Chem. 250, 1525-1547. Reeke, G. N., Becker, J . W. and Edelman, G. M. (1978). Proc. Nurl. Acad. Sci. U.S.A. 75, 22862290. Rich, D. H., Lehrman. S. R., Kawai, M., Goodman, H. L., and Suttie, J. W. (1981). J . Med. Chem. 24, 7 0 6 7 1 1. Richardson, C., Behnke, W. D., Freisheim, J. H., and Blumenthal, K . M. (1978). Biochim. Biophys. Acru 537, 310-319. Roberts, M.F., Deems, R. A., and Dennis, E. A. (1977). J . Biol. Chem. 252, 601 1-6017. Scheele, G. A., Palade, G. E., and Tartakoff, A. M. (1978). J . CellBiol. 78, 110-130. Scheinbuks, J. (1980). I n “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (J. W. Suttie, ed.), pp. 279-285. Univ. Park Press, Baltimore, Maryland. Schleicher, M., Iverson, D. B., Van Eldik, L. J., and Watterson, D. M. (1981). In “The Cytoskeleton in Plant Growth and Development” (C. Lloyd, ed.). Academic Press, New York. Schliwa, M., Euteneuer, U., Bulinski, J . C., and Izant, J. G. (1981). Proc. Narl. Acad. Sci. U.S.A. 78, 1037-1041. Seamon, K. B. (1980). Biochemisrry 19, 210-215. Sequeria, L. (1978). Annu. Rev. Phyropurhol. 16, 453-481. Sharma, R. K., Desai, R., Thompson, T. R., and Wang, J. H. (1978). Can J. Biochern. 56, 598-604. Sharma, R. K., Desai, R., Waisman, D. M., and Wang, I. H. (1979). J . B i d . Chem. 254, 42764282. Shenolikar, S . , Cohen, P. T. W., Cohen, P., Nairn, A. C., and Perry, S. V. (1979). Eur. J . Biochem. 100, 329-337. Shipolini, R. A., Callewaert, G. L., Cottrell, R. C., Doonan, S., Verron, C . A,, and Banks, B. E. C. (1971). Eur. J. Biochem. 20, 459-468. Shoham, M.,Kalb, A. J., and Pecht, I. (1973). Biochemisrry 12, 1914-1917. Siegel, F. L., Carafoli, E., Kretsinger, R. H., MacLennan, D. H., and Wasserman, R. H. (19x0). In “Calcium-Binding Proteins: Structure and Function” (F. L. Siegel, E. Carafoli, R. H. Kretsinger, D. H. MacLennan, and R. H. Wasserman, eds.), pp. 357-422. Elsevier, Amsterdam. Sipos, T., and Merkel, J. R. (1970). Biochemisrry 9, 2766-2775. Slotboom, A. J., and de Haas, G . H. (1975). Biochemistry 14, 5394-5399. Slotboom, A. J., Van Dam-Mieras, M. C . E., and de Haas, G. H. (1977). J. Biol. Chem. 252, 2948-295 I . Slotboom, A. J., Jansen, E. H. J. M., Vlijm, H., Pattus, F., Soares de Araujo, P., and de Haas, G. H. (1978). Biochemistry 17, 45934600. Smoake, J. A., Song, S.-Y., and Cheung, W. Y.(1974). Biochim. Biophys. Acru 341, 402-41 I . Sobue, K., Fujita, M., Muramoto, Y.,and Kakiuchi, S. (1980). Biochem. In?. I, 561-566. Sobue, K., Muramoto, Y.,Fujita, M., and Kakiuchi, S.,(1981a). Biochem. Biophys. Res. Cornmun. 100, 1063-1070.

CALCIUM-BINDING PROTEINS

59

Sobue, K., Muramoto, Y., Fujita, M., and Kakiuchi, S. (1981b). Proc. Natl. Acad. Sci. U.S . A . 78, 5652-5655. Sobue, K., Fujita, M., Muramoto, Y., and Kakiuchi, S. ( 1 9 8 1 ~ )FEBS . Lett. 132, 137-140. Soute, B. A. M., Vermeer, C . , De Metz. M., Hemker, H. C., and Lijnen, H. R. (1981). Biochim. Biophys. Acta 676, 101-107. Spencer, R. M., Charman, P., Emtage, J. S., and Lawson, D. E. M. (1976). Eur. J. Biochem. 71, 399-409. Spencer, R., Charman, M., and Lawson, D. E. M. (1978). Biochem. J. 175, 1089-1094. Starostina, M. V., Malup, T. K., and Sviridov, S. M. (1981). J . Neurochem. 36, 1904-1915. Stenflo, J. (1970). Acta Chem. Scand. 24, 3762-3763. Stenflo, J. (1976). Trends Biochem. Sci. 1, 256258. Stenflo, J. (1978). Adv. Enzymol. 46, 1-31. Stenflo, I. (1980). In “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (J. W. Suttie. ed.), pp. 102-105. Univ. Park Press, Baltimore, Maryland. Stenflo, J., and Ganrot, P. (1972). J. Biol. Chem. 247, 81W8166. Stenflo. J.. Femlund, P., Egan, W., and Roepstorff, P. (1974). Proc. Narl. Acad. Sci. U.S.A. 71, 2730-2733. Stewart, J. S. (1972). Biochim. Biophys. Acta 263, 178-192. Suttie, J. W. (1977). In “Advances in Nutritional Research” (H. H. Draper, ed.), Vol. 1 . Plenum, New York. Suttie, J. W. (1978). In “The Fat-Soluble Vitamins, Handbook of Lipid Research” (H. F. De Luca, ed.), Vol. 2, pp. 21 1-277. Plenum, New York. Suttie, J. W. (1980a). CRC Crit. Rev. Biochem. 8, 191-222. Suttie, J. W. (1980b). Fed. Proc. Fed. Am. SOC. Exp. B i d . 39, 2730-2735. Suttie, J. W., ed. (1980~).“Vitamin K Metabolism and Vitamin K-Dependent Proteins.” Univ. Park Press, Baltimore, Maryland. Suttie, J. W., and Jackson, C. M. (1977). Physiol. Rev. 57, 1-70. Suttie, J. W., Lehrman, S. R., Geweke, L. 0.. Hageman, J. M., and Rich, D. H.(1979). Biochern. Biophys. Res. Cornmun. 86, 500-505. Suttie, J. W., Geweke, L. O., Martin, S . L., and Willingham, A. L. (1980a). FEBS Lett. 109, 267-270. Suttie, J. W., Larson, A . E., McTigue, J. J., Kawai, M., and Rich, D. H. (1980b). In “CalciumBinding Proteins: Structure and Function” (F. L. Siegel, E. Carafoli, R. H. Kretsinger, D. H. MacLennan, and R. H. Wasserman, eds.), pp. 461-468. Elsevier, Amsterdam. Szebenyi, D. M. E., Obendorf, S. K., and Moffat, K. (1982). Nature (London) in press. Tabor, H., and Tabor, C. W. (1977). Anal. Biochem. 78, 554-556. Takagi, T., Nemoto, T., and Konishi, K. (1980). Biochem. Biophys. Res. Commun. 96, 377-381. Taylor, A. N. (1980a).I n “Vitamin D. Molecular Biology and Clinical Nutrition” (A. W. Norman, ed.). pp. 321-351, 691-694. Dekker, New York. Taylor, A. N. (1980b). I n “Calcium-Binding Proteins: Structure and Function” (F. L. Siegel, E. Carafoli, R. H. Kretsinger, D. H. MacLennan, and R. H. Wasserman, eds.), pp. 393400. Elsevier, Amsterdam. Taylor, A. N. (1981). J. Histochem. Cytochem. 29, 65-73. Teo, T. S., and Wang, J. H. (1973). J . Biol. Chem. 248, 5950-5955. Teshima, Y., and Kakiuchi, S. (1978). J . Cyclic Nucleotide Res. 4, 21g-231. Tinker, D. O., and Wei, J. (1979). Can J . Biochem. 57, 97-106. Tinker, D. 0.. Low, R., and Lucassen, M. (1980). Can. J . Biochem. 58, 898-912. Titani, K., Fujikawa, K., Enfield, D. L., Ericsson, L. H., Walsh, K. A., and Neurath, H. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 3082-3086. Toscano, W. A., Westcott, K. R., LaPorte, D. C., and Storm, D. R. (1979). Proc. Natl. Acad. Sci. U.S.A. 76, 5582-5586.

60

LINDA 1 . VAN ELDIK ET AL

Traverso, H. P., Hauschka, P. V., and Gallop, P. M. (1980). I n “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (J. W. Suttie, ed.), pp. 303-306. Univ. Park Press, Baltimore, Maryland. Tsai, I.-H., Wu, S.-H., and Lo, T.-B. (1981). Toxicology 19, 141-152. Tuan, R. S. (1980). I n “Vitamin K Metabolism and Vitamin K-Dependent Proteins” (J. W. Suttie, ed.), pp. 294-298. Univ. Park Press, Baltimore, Maryland. Tuan, R. S., and Scott, W. A. (1977). Proc. Narl. Acad. Sci. V.S.A. 74, 19461949. Tuan, R. S . , Scott, W. A., and Cohn, Z. A. (1978a). J . Cell Biol. 77, 743-751. Tuan, R. S., Scott, W. A., and Cohn, Z. A. (1978b). J. Cell Biol. 77, 752-761. Ueng, T.-H., and Bronner, F. (1979). Arch. Biochem. Biophys. 197, 205-217. Uyemura, K., Vincendon, G., Combos, G., and Mandel, P. (1971). J . Neurochem. 18, 429-438. Van Buskirk, J. J., and Kirsch, W. M. (1978a). Biochem. Biophys. Res. Commun. 80, 1033-1038. Van Buskirk, J. J., and Kirsch, W. M. (1978b). Biochem. Biophys. Res. Commun. 80, 1329-1331. Van Buskirk, J. J., Low, M., and Kirsch, W. (1980). I n “Vitamin K Metabolism and Vitamin KDependent Proteins” (J. W. Suttie, ed.), pp. 274-278. Univ. Park Press, Baltimore, Maryland. Van Dam-Mieras, M. C. E., Slotboom, A. J., Pieterson, W. A,, and de Haas, G. H. (1975). Biochemistry 14, 5387-5393. Van Deenen, L. L. M., and de Haas, G. H. (1964). Adv. Lipid Res. 2, 167-234. Van den Bosch, H. (1980). Biochim. Biophys. Acru 604, 191-246. Van Eldik, L. J. (1982). In preparation. Van Eldik, L. J., and Fok, K. F. (1981). Fed. Proc. Fed. Am. SOC. Exp. Biol. 40, 1740. Van Eldik, L. J., and Watterson, D. M. (1981). J. Biol. Chem. 256, 4205-4210. Van Eldik, L. J., Piperno, G., and Watterson, D. M. (1980). Proc. Nurl. Acud. Sci. U.S..4. 77, 47794783.

Vensel, L. A., and Kantrowitz, E. R. (1980). J . Biol. Chem. 255, 73067310. Verheij, H. M., Volwerk, J. J., Jansen, E. H. J. M., Puyk. W. C., Dijkstra, B. W., Drenth, 1.. and de Haas, G. H. (1980). Biochemistry 19, 743-750. Verheij, H. M., Egmond, M. R., and de Haas, G. H. (1981). Biochemisrry 20, 94-99. Volwerk, J. J., Pieterson, W. A,, and de Haas, G. H. (1974). Biochemistry 13, 14461454. Voordouw, G., and Roche, R. S. (1974). Biochemistry 13, 5017-5021. Wallace, R. W., Lynch, T. J., Tallant, E. A., and Cheung, W. Y. (1978). J. Biol. Chem. 254, 377-382.

Wang, J. H., and Desai, R. (1977). J. B i d . Chem. 252, 4175-4184. Wasserman, R. H., and Feher, J. J. (1977). I n “Calcium-Binding Proteins and Calcium Function” (R. H. Wasserman, R. A. Corradino, E. Carafoli, R. H. Kretsinger, D. H. MacLennan, F. L. Siegel, eds.). pp, 293-302, Elsevier, Amsterdam. Wasserman, R. H., and Taylor, A. N. (1966). Science 152, 791-793. Wasserman, R. H., Corradino, R. A., and Taylor, A. N. (1968). J . Biol. Chem. 243, 3978-3986. Watterson, D. M.,Harrelson. W. G., Keller, P. M., Sharief, F., and Vanaman, T. C. (1976). J. Biol. Chem. 251, 4501-4513. Watterson, D. M., Sharief, F., and Vanaman, T. C. (1980). J. Biol. Chem. 255, 962-975. Weiss, B., Prozialeck, W., Cirnino, M., Bamette, M. S., and Wallace, T.L. (1980). Ann. N.Y. Acud. Sci. 356, 319-345. Wells, M. A. (1974a). Biochemistry 13, 2248-2257. Wells, M. A. (1974b). Biochemistry 13, 2265-2268. Welsh, M. J., Dedman, J. R., Brinkley, B. R., and Means, A. R. (1979). J. Cell Biol. 81,624-634. Wilson, P. W., and Lawson, D. E. M. (1977). Biochim. Biophys. Acru 497, 805-811. Wolff, D. I., Ross, J. M.. Thompson, P. N., Brostrom, M. A., and Brostrom, C. 0. (1981). J. Biol. Chem. 256, 1846-1860. Wong, P. Y. K., and Cheung, W. Y. (1979). Biochem. Biophys. Res. Commun. 90, 473-480.

CALCIUM-BINDING PROTEINS

61

Wood, J . G . , Wallace, R . W., Whitaker, J . N . , and Cheung, W. Y. (1980). J . CellEiol. 8 4 , 6 6 7 6 . Yazawa, M . , Yagi. K . , Toda, H . , Kondo, K., Narita, K . , Yamazaki, R . , Sobue, K., Kakiuchi, S . , Nagao, S . , and Nozawa, Y. (1981). Eiochem. Eiophys. Res. Commun. 99, 1051-1057. Zytkovicz, T. H . , and Nelsestuen, G.L. (1975). J . B i d . Chem. 250, 29662972.

This Page Intentionally Left Blank

INTERNATIONAL REVIEW OF CYTOLOGY. VOL.77

Genetic Predisposition to Cancer in Man: In Vitro Studies LEVYKOPELOVICH Memorial Sloan-Kettering Cancer Center and Cornell University Graduate School of Medical Sciences, New York, New York ................................ I . Introduction . . . . . . . ......................... II. 111. Studies on ACR Cell Cultures . . , . , . , , . . . . . . . . . . . . . . . . . . . . . . A. Serum Requirements and Growth Properties . . . . . . . . . . . . . . . B. lntra- and Intercellular Cytoskeletal Matrices .. C. Membrane-Associated Parameters. . . . . . . . . . . . . . . . . . . . . . . . D. Metabolic Alterations. . . . . . . . . , . . . . . . , . . . . . . . . . . . . . . . . . E. Differential Susceptibility to Transformation by Oncogenic Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. F. Cancer-Related Antigens . , . . . . . . . . . . . . . . . . . G. Summary of Tissue Culture Studies , . . . , . . , . . . . . . . . . . . . . . IV. Cell Culture Studies on Autosomal Dominant Syndromes (Other Than ACR) and Chromosome Instability Syndromes . . . . . . . . . . . . V. On the Question of Tumor Promotion,. . . . . . . . . . . . . . . . . . . . . . . ............ VI. Genetic Mechanisms . . . . . . . . . . . VII. Is Genetic Predisposition to Cancer an Autosomal Dominant VIII. On the Question of Cancer Prognosis and Cancer Control. . .......................... IX. Conclusion . . . . . . . . References ..............................................

63 65 66 66 61 67 68 68 70 70

73 75 78 19 81

83 84

1. Introduction

Human cancer is presumably effected through an interaction of genetic and environmental factors. The occurrence of cancer in bona fide genetic syndromes (Fraumeni, 1977; Knudson, 1981; Kopelovich, 1980a; Lynch, 1976; Mulvihill et al., 1977; Peto, 1980), in familial clusters (Anderson and Romsdahl, 1977; Fraumeni, 1977; Lynch et af., 1977; Peto, 1980), and in a significant number of first and second degree relatives in sporadic forms of cancer (Schneider, 1981) suggests, however, that genetic predisposition to neoplasia is far more important than heretofore realized. In this context, pedigree analysis provides a model for the elucidation of carcinogenic mechanisms and for the ascertainment of constitutional markers associated with a genetic predisposition to cancer. The establishment of constitutional markers should play a major role in the development of 63 Coovnrht 0 1982 by Academic Presi. fnc.

64

LEVY KOPELOVICH

means for identifying individuals predisposed to cancer. Lacking means for karyotypic identification of gene carriers, we must depend on the analysis of somatic cells through biological characterization and biological probes. Exceptions include the Philadelphia chromosome of chronic myeloid leukemia (Marinello et ul., 1981; Rowley, 1980), the balanced 3 to 8 translocation in hereditary renal cell carcinoma (Cohen et al.. 1979), the rare case of retinoblastoma (Knudson, 1981; Malenbaum et al., 1981; Strong, 1981), and possibly the Gardner Syndrome (Gardner et al., 1981). Recent developments in somatic cell genetics (Ruddle, 1981; Siniscalco, 1979), viral genetics (Huebner and Todaro, 1969; Collett et al., 1979; Oppermann et al., 1979; Hayward et al., 1981; Temin, 1980), and eukaryotii: cell differentiation (Brown, 1981; Sachs, 1980) make our interpretation of cancer predisposition more tangible. A consensus is that the transformed phenotype is expressed as an autosomal dominant trait (Comings, 1973; Shkolnick and Sachs, 1978; Stanbridge and Wilkinson, 1978; Wiener et al., 197 I ) , and that the nialignant phenotype is presumably expressed in the autosomal recessive (Comings, 1973; Kinsella and Radman, 1978; Ohno, 1971; Stanbridge and Wilkinson, 1978) or in the codominant mode. If, as has been suggested, the dominant cancer trait is expressed through a class of tissue differentiation genes (Comings, 1973; Knudson, 198 I), the elucidation of their mechanism of action could explain, in part, the tissue and organ specificity seen in human cancer. For example, most human tumors are of epithelial origin (Cairns, 1981). In general, differentiation genes might bear some resemblance in tissues of epithelioid origin, yet apparently they are distinctly unique for each organ, representing presumably different alleles and involving a variety of susceptibility mechanism(s). The variability of expression may also be due, in part, to a variety of epigenetic control mechanisms. Along similar lines, we have shown that embryo and foreskin cells may represent a state in which the differentiation process is presumably not yet complete and which is phenotypically similar to the pattern we have come to identify with transformation in vitro. Apparently, cells at this stage of development are genotypically more sensitive to an insult by carcinogens than are normal adult fibroblasts. Thus, studies about the oncogenic potential of normal diploid human cells should be qualified with respect to their state of development and genetic constitution (Kopelovich, submitted). Several recent reviews dealing with genetic (Marks, 1981; Ponder, 1980; Smets, 1980; Schimke, 1978) and cytogenetic (Rowley, 1980; Sandberg, 1980) factors in the causation of human cancer have been published. This review is based on studies with cultured skin fibroblasts derived from humans who are genetically predisposed to cancer. However, we shall emphasize primarily our work with heritable adenomatosis of the colon and rectum

GENETIC PREDISPOSITION TO CANCER IN MAN

65

(ACR), since it constitues the major portion of published material on this subject. We discuss the evidence and make certain deductions about mechanisms associated with cancer predisposition and cancer control in man.

11. The Experimental System

Studies concerning cancer biology in vitro require that three major requirements be met. The first requirement concerns the selection of a relevant model system to study this disease in humans. The second requirement concerns the histologic and anatomic proximity (identity) of the diseased tissue to the normal tissue of origin. The third requirement concerns the preservation of the biologic profile of a tissue in vitro, reflecting its qualities as they might presumably occur in vivo. Our approach to the first requirement has been to study an inherited syndrome of cancer, adenomatosis of the colon and rectum (ACR), an autosomal dominant mutation (Alm and Licznerski, 1973; Gardner and Richards, 1953; McConnel, 1980; Morson and Bussey, 1971). The trait is expressed regularly through consecutive generations in obligatory heterozygotes and carcinomas of the large bowel arise in virtually all untreated cases. Our approach to the second requirement has been to study the in vitro properties of cells obtained from cutaneous biopsies of patients and normal individuals rather than of cells grown from biopsy material of tissues at risk. By using only cultured skin fibroblasts (SF), we were able to examine a single cell type under reproducible conditions for presumptive differences between ACR-positive and ACR-negative individuals. Our approach to the third requirement is answered, in part, through our ability to compare cutaneous cells from both normal and ACR individuals under identical conditions in vitro. It is also borne out by experimental findings about the occurrence of abnormal phenotypic expressions in SF of ACR individuals which segregate as would be anticipated from an autosomal dominant trait. These abnormalities are found in ACR patients who are certain to develop cancer, in cancer cells derived from spontaneously occurring human tumors, and in cells treated in vitro with chemical and physical agents. They also appear to correlate well with current knowledge about the sociopathology of tumor cells in vivo (Kopelovich, I980a; Rubin, 1980; Tooze, 1979; Watson, 1979). In these studies no differences have ever been observed between tissues obtained from ACR patients and Gardner variant patients (Kopelovich, 1980a). The terminology of basic concepts used in this review will be as follows: A transformed cell phenotype or a preneoplastic state are used to connote the occurrence in cells of abnormal phenotypic expressions in vitro. The terms

66

LEVY KOPELOVICH

initiation and cancer predisposition will be used to describe the transformed phenotype when a discussion about mechanisms and in vivo genetic susceptibility, respectively, is indicated. The terms neoplasia and tumor will be qualified where necessary, as benign or malignant. All other synonyms for cancer are used to mean a tumor which is malignant by the usual criteria in vivo. The terms autosomal dominant and autosomal recessive cancer traits are used to describe susceptibility alleles conceived through statistical analyses of genetic models.

111. Studies on ACR Cell Cultures

A. SERUMREQUIREMENTS AND GROWTH PROPERTIES Normal fibroblastic cells in v i m cease to divide under conditions of serum deprivation and when extensive cell-cell contact occurs. These cells must also be provided with a solid substrate on which to anchor and spread. Physical and chemical agents and oncogenic viruses may cause the loss of sensitivity to one or all of these variables (Tooze, 1979; Watson, 1979). Occasionally, cell populations that lack serum sensitivity but retain an anchorage requirement have been shown to be nontumorigenic (Tooze, 1979; Watson, 1979). SF from ACR gene carriers but not from normal individuals have partially lost serum- and density-sensitive growth control in culture (Kopelovich, I977a; Kopelovich et a f . , 1979b; Pfeffer et al.. 1976). Their cloning efficiency is considerably higher, and their clonal morphology tighter than those of normal cells (Kopelovich et a l . , 1979b). We have also been able to show that anchorage sensitivity of ACR cells is not absolute and that growth in agar does occur spontaneously, albeit at low frequency (Kopelovich, 1980b, 1982). Apparently, the loss of sensitivity to one regulatory constraint in this cell system is coupled to the loss of sensitivity to all other growth controls. Peehl and Stanbridge (1981) have recently reported that normal foreskin l'ibroblasts will grow without anchorage in the presence of 20% fetal calf seruni and 10 Kglml of hydrocortisone. We have extended this observation to normal embryo cells and normal adult skin fibroblasts. For example, normal adult human skin fibroblasts appeared to proliferate in agar containing hydrocortisone considerably better than did ACR cells, while no growth of normal cells was noted in the absence of hydrocortisone (Kopelovich, in preparation). We feel, therefore, that the question of growth without anchorage of normal and transformed human cells should be reexamined in the light of these data. The apparent lack of sensitivity of ACR SF to hydrocortisone may be tightly linked to mechanisms associated with initiation and promotion for this form of cancer. It will be used to further distinguish these cells from normal SF.

GENETIC PREDISPOSITION TO CANCER IN MAN

67

B . INTRA-AND INTERCELLULARCYTOSKELETAL MATRICES Two major biologic correlates of in vitro malignancy are deformed actincontaining cables and fibronectin infrastructure (Ali et al., 1977; Goldman et al., 1976; Hynes and Destree, 1978). In the cytoplasm of well-spread cultured normal cells, actin is organized into a network of cables that run the length of the cell just inside the adherent cell membrane (Goldman et al., 1976). Fibronectin represents a principal component which in association with other extracellular proteins is responsible for the formation of organized cell adhesions (Review by Goldman et al., 1976). A partial to complete disappearance of these organized matrices occurs in fibroblasts that have become tumorigenic as a result of chemical or viral transformation (review by Goldman et al., 1976). Using indirect irnmunofluorescence we have found a disruption of actin organization in cultured SF from ACR gene carriers (Kopelovich et a ( . , 1977, 1980). However, these cells were normal with regard to other cytoskeletal structures such as microtubules and intermediate filaments (Kopelovich and Fusenig, unpublished), The possibility that ACR individuals possess high amounts of actin depolymerizing proteins (Norberg et al., 1979; Schliwa, 1981) should be examined. Alternatively this phenomenon may reflect increased sensitivity of the actin-containing cables under in vitro conditions. ACR cells were indistinguishable from normal cells in their ability to express fibronectin, indicating that alterations involving fibronectin are further along the transformation process (Renard and Kopelovich, in preparation). However, ACR cells did not respond to the addition of exogenous fibronectin with a normalization of their morphology and reorganization of their actin-containing cables as has been reported for some transformed cells in culture (Hynes and Destree, 1978). Conceivably, a constitutive alteration in the pattern of actin-containing cables renders the attachment plaques insensitive to the addition of exogenous fibronectin. Preliminary experiments using cis-hydroxyproline (Kopelovich, unpublished) and electron microscopy (S. Rogers and E. Gardner, personal communication) suggest that the final assembly of extracellular collagen matrices (tertiary structures) might also be defective in ACR cells. C. MEMBRANE-ASSOCIATED PARAMETERS Plasminogen activator is considered to be a biochemical correlate of in virro malignancy (Reich et ul., 1975), although several other proteases have also been implicated (Troll et al., 1975). Elevated levels of intra- and extracellular plasminogen-dependent protease were found in ACR cells but not in normal cells (Kopelovich, 1977a; Chopan and Kopelovich, 1981b). In a related study we have been able to show increased agglutination of ACR cells in the presence of concanavalin A. This was demonstrated both in suspension (Chopan and

68

LEVY KOPELOVICH

Kopelovich, 198lb) and in glutarylaldehyde-treated,derivatized, concanavalin A matrices (Braun and Kopelovich, unpublished observations). In addition, there was an increased uptake of 2-deoxyglucose by ACR cells (Chopan and Kopelovich, 1981b). The results may suggest that these membrane-associated alterations are related, in part, to the abnormal pattern of actin-containing cables seen in ACR cells. D. METABOLIC ALTERATIONS 1 . Cholesterol Feedback Regulation

Cholesterol is essential for normal growth and survival of mammalian cells (Brown and Goldstein, 1974). It has been proposed that defective feedback control of cholesterol biosynthesis observed in premalignant and malignant tumors may be specifically related to oncogenic transformation (Siperstein and Fagan, 1964). However, evidence from cell culture suggests no intrinsic link between malignancy and the loss of feedback control of sterol biosynthesis (Bierne and Watson, 1976; Watson, 1973). The possibility that the accumulation of undegraded cholesterol in the feces of ACR individuals (Reddy et al., 1976) is a consequence of defective feedback control by cells lining the gastrointestinal tract was investigated. Since the ACR trait appears to be a systemic disorder, we assumed that cultured SF from afflicted individuals would present a metabolic pattern similar to that of epithelial cells. The results showed that SF from ACR cell phenotypes are presumably normal with regard to the regulation of 3-hydroxy-3-methylglutarylcoenzyme A reductase (Kopelovich, 1977a, 1978a), indicating no correlation between the loss of feedback regulation for cholesterol biosynthesis and oncogenic potential of these cells in culture. The high level of undegraded cholesterol in ACR individuals may reflect a specific interaction of host cells with the bacterial flora in the gastrointestinal tract. 2 . Mitochondria An altered potential across the mitochondria1 membrane has been recently identified in ACR cells (Kopelovich, Melamed, and Darzynkiewicz, unpublished observations). This phenotypic aberration coupled with inefficient respiration and increased lactate production, if confirmed in a large number of experiments, may represent the most significant contribution of a metabolic pathway which largely defines the predisposed (initiated) state.

E. DIFFERENTIAL SUSCEPTIBILITY TO TRANSFORMATION BY ONCOGENIC AGENTS Chemical transformation of normal and ACR cells was carried out through treatment with various levels of N-methyl-A”-nitro-N-nitrosoguanidine

GENETIC PREDISPOSITION TO CANCER IN MAN

69

(MNNG). The ACR cells underwent morphological transformation, grew to higher saturation density, formed large aggregates in liquid growth medium above an agar base, and formed colonies in soft agar. These cells have also become resistent to a rechallenge with MNNG; they showed a prolonged life span in culture and a marked increase in cell ploidy compared with untreated cells (Rhim et al., 1980). On the other hand, normal cells have senesced under similar condition of exposure to MNNG. Studies on irradiation-induced cell survival and transformation were carried out concurrently. The ultimate goal was to determine the frequency of transformation per viable cell. Cell survival analysis for y- and UV-irradiated ACR cells showed a single cell survival curve similar to that of normal, age-matched, human cells (Kopelovich et al., 1981). The sensitivity of normal and ACR cells to irradiation was further analyzed through the ability of SV40-infected cells to repair defective SV40 function. It was determined by T-antigen display, following exposure of the SV40 virions before infection of the host cells, to X ray or UV irradiation. In these experiments as well, both cell types were equally competent in restoring irradiation-induced T-antigen display (Kopelovich, in preparation). ACR cells, but not their age-matched normal SF, have been transformed by yirradiation as determined by focus formation and growth in agar. They also showed prolonged survival in culture following irradiation. Studies are currently in progress to inoculate both MNNG-treated and y-irradiated cells in the anterior chamber of the eye of a nude mouse. Virally induced cell transformation has been used to study variations in susceptibility of human mutant cells to neoplasia (Aaronson and Todaro, 1975; Klement et af., 1971; Todaro and Martin, 1967). Infection of SF with the Kirsten murine sarcoma virus (KiMSV) showed that cell cultures derived from ACR individuals and a fraction of the clinically asymptomatic ACR progeny were considerably more susceptible to transformation by KiMSV than were normal subjects (Pfeffer and Kopelovich, 1977). The increased susceptibility was probably due to the transformation process and not to a type C virus replication step. In a separate study, we have shown that ACR cells were also more susceptible to an SV40-induced T-antigen display and transformation (Kopelovich and Sirlin, 1980). The virally transformed cells have become anchorage-independent and, in the case of KiMSV, formed transient nodules subcutaneously in athymic mice. They did not, however, acquire an infinite life span in culture (Kopelovich and Sirlin, 1980; Pfeffer and Kopelovich, 1977). The susceptibility of ACR cells to transformation by KiMSV and SV40 appears to segregate within the ACR progeny in a fashion identical to our observations about the occurrence of abnormal phenotypic expressions in these cells. It suggests that the ACR trait and its mode of inheritence is also responsible for the increased sensitivity to transformation by the viral probes. The apparent correla-

70

LEVY KOPELOVICH

tion between SV40 and KiMSV suggests that genetic information residing within ACR cells renders them more susceptible. In this respect, the oncogene postulate (Huebner and Todaro, 1969) and the DNA provirus hypothesis (Temin, 1980) are of interest. We are currently searching for putative gene products which may play a role in the malignant transformation ofthe ACR cell (Collet ef ( I / . , 1979; Deleo ct u / . , 1979; Ellis et ul., 1981 ; Linzer and Levine, 1979; Oppermaiin et a/., 1979). Results about the cellular content of the sarc gene product of avian sarcoma virus in normal and ACR cells, and in TPA-treated cells (e.g., Section V ) showed no correlation with disease or treatment (Goldberg and Kopelovich, unpublished).

F. CANCER-RELATED ANTIGENS The appearance of embryo-specific proteins in tumor cells suggests an association of cancer-related antigens with malignancy (Alexander, 1972; Coggins and Anderson, 1974; Stonehill and Bendich, 1970). The occurrence of embryo-specific proteins in SF from ACR individuals during the preneoplastic state and following transformation by KiMSV and SV40 was investigated. The results show that cancer-related antigens are expressed in the virally transformed cells, but not in mock-infected SF (Kopelovich, 1980a). Apparently, the expression of embryo-specific proteins in ACR cells is not associated with the preneoplastic state, but is a consequence of viral or chemical transformation. It is of considerable interest that infection of these cells by an RNA oncogenic virus elicited the synthesis of fetal-like antigens, whereas that by the DNA oncogenic virus affected the appearance of placental-like antigens (Kopelovich, 1980a). How the type of viral nucleic acid and its mode of replication might affect the synthesis of specific host cell neoantigens in the course of a virally induced cell transformation remains to be established. In this regard, TPA-treated ACR cells have become positive with respect to a fetal-like antigen (Kopelovich, 1980b, 1982). Mock-infected cells and both KiMSV and S'V40 transformed normal or ACR cells were negative with regard to human choriogondotropin, a2-microglobulin, carcinoembryonic antigen, and a-fetoprotein.

G . SUMMARY OF TISSUE CULTURE STUDIES Phenotypic expressions occurring in vitro in early passage human cell cultures (Table I) apparently reflect biologic properties occurring in situ (Kopelovich, 1980a; Rubin, 1980). Therefore, the study of SF derived from ACR genotypes provides a unique system for analysis of the oncogenic process. These abnormal phenotypic expressions represent an early event which presumably is tightly coupled to tumor progression, but nevertheless is insufficient to effect the final

GENETIC PREDISPOSITION TO CANCER IN MAN

71

TABLE I A PHENOTYPIC PROFILE OF A N INITIATED ACR CELL A. Growth parameters I . Growth in nutrient-deprived environment 2. Loss of contact inhibition 3. Formation of cell aggregates 4. Increased cloning efficiency 5 . A partial loss of anchorage sensitivity 6. Decreased sensitivity to hydrocortisone-induced cell proliferation in agar (Kopelovich. unpublished) B. Cytoskeletal structures I . Defective actin-containing cables (a-actin) 2. Defective myosin-associated cables (a-meroniyosin) (Kopelovich, unpublished data) 3. Normal organization of microtubules (a-tubulin) (Kopelovich, unpublished data) 4. Normal organization of intermediate filaments (a-vinmentin) (Kopelovich, unpublished data) C. Membrane-associated parameters I . Increased Con A agglutination (in suspension and onto derivatized matrices) 2. Increased intra- and extracellular levels of plasminogen activator 3. Normal expression of LETS (Renard and Kopelovich, in preparation) 4. Defective polymerization of collagen (E. M./hydroxy and cis-hydroxyproline) (R. Scott and E. Gardner, personal communication; Kopelovich, unpublished data) 5 . Proteoglycans (not done) D. Increased sensitivity to further transformation by oncongenic viruses I . KiMSV (associated with an expression of a human fetal-like antigen) 2. SV40 (associated with an expression of a human placenta-like antigen) E. Increased sensitivity to further transformation by chemical agents I . TPA (associated, presumably. with induced alteration followed by selection, and characterized by a decreased toxicity of ACR cells. and the occurrence of a human fetal-like antigen) 2. MNNG (associated. presumably. with an induced alteration followed by selection toward more resistant variants) F. Miscellaneous I . Normal cholesterol feedback regulation (HMG CoA reductase) 2. Apparently nomial radiosensitivity (X ray. U V ) 3. Altered membrane potential across mitochondria1 membrane

oncogenic event. A second event appears to be necessary for the malignant transformation of these cells. This event, however, is extremely rare at the cell level even in the target organ, and most initiated cells in vivo remain nontumorigenic throughout. Perturbation of these cells with oncogenic agents induced the following changes: In general, there was an amplification in the expression of the phenotypic aberrations occurring in the initiated state (Table 11) including the formation of dense, ridge-like, foci which secrected large amounts of plasminogen activator. However, cells isolated from these foci, or from agar-growing colonies

TABLE I1 ABNORMALS Y S T t M l C

M A N l t t S T A T I O N S A S S O C l A T t D W I T H L O S S O t R t G U L A T O R Y C O N T R O L M t C H A N l S M S A N D B l O C H t M l C A L A L T t R A T l O N S IN

HEREDITARY

A D t N O M A T O S l S OF T H t C O L O N A N D R t n U M " Human phenol) pet

Phenol\plL exprewons I n cultured huniAn A i n fihmhlast\ Ahmrnial growth paramc-

ten

Increased pmteolyIIC

Dccrcawd actin cahlcs

Anchorage indcpcndcncc

acliwty

Increawd agglutination hy Icctin,

Increased 2-devxyglucore uptale

Altered po tential acre+

mcmbraner

Incrcared wxeptihility 10 twnsforniation by oncogenic agent,

Aupmentation of

phenotypic CX-

prewon\ arrociated with the initiated \tale

Transformationinduced dense. ndgclike. foci (TPA: carcinogens. y-irradialion) den% ra-

Ernhryo spccafic proteins

Increased cell ploidy

fetal IKiMSV. TPA ) placental (SV4Ol

Ability 10 fomi palpahle hut transient nodules (virus) and neoplasia (TPA )

dial foci (vim$)

T The initiated phase imultiply \teps presumahly due

to a single mutation)

The neoplastic phase (multiple steps prcsuniahly due lo one or ninre niutatmm) I . Clinically rymplomatic

t

t

+ -

+ -

-

-

+

+

+

t

t

ND ND -

ND ND -

+

t

-

-

t

+

+

+

+

t

t

t

t

+

t

t

+

-

-

-

-

-

-

Z Clinically asymptomatic prowny a. Positive h. Negative

3. Normalr

-

-

Venical arrow indicates the tmnsformlng cvenl: in thi5 system. transfomiing agents were KiMSV. SV40. TPA. MNNC. and y-irndiation. The efficacy 01 \irm tran\formation of cell, fmm normal individuals was considerably l e s than that from ACR individuals. but a11 Irmsformed cell\ gave rise lo the vdnic phenotypic exprewon\. I n the ca\c of the chemical and physical agents. no transformation of nomial SF has k e n seen. The iliiiii2! z,jr,p:Gr,z;ic piGginy hi kir, 5"k"%<&d iiliG p,hiiiii an: 3iga;,ii aiioiding ;G i l U i e A p G , j K r , t * l fir,&,g. I'

GENETIC PREDISPOSITION TO CANCER IN MAN

73

did not breed true, nor did they acquire the ability to grow in culture indefinitely. Other major alterations included a considerable increase in cell ploidy, the expression of neoantigens, the ability of virus-treated cells to form palpalbe, albeit transient nodules, upon injection subcutaneously in athymic mice, and the ability of TPA-treated cells to grow in vivo upon injection intraoccularly in athymic mice (e.g., Section V). These changes might be considered late events (postinitiation) in spontaneously occurring human tumors as well. The apparent susceptibility of ACR cells to further transformation by oncogenic viruses (both KiMSV and SV40), chemical, and physical agents indicates that genetic information residing within these cells, probably in the form of a relatively limited and specific number of DNA sequences associated with the ACR mutation, renders them more sensitive to these three distinct classes of carcinogens. Interestingly, the chromosomal alterations and the increased sensitivity of ACR cells to further transformation by oncogenic agents, including viruses (Gardner et d., I98 1 ; Hori et ul., 1980; Miyaki et ul., 1980a). were not associated with increased toxicity to cell proliferation. Thus, if cell toxicity is any indication about repair mechanisms, these results suggest no immediate relationship between repair and susceptibility to transformation in our cell system. Similar reports about the ACR cell system have been recently published by 1980; Miyaki et a/., several laboratories (Gardner e/ a / ., 198 I ; Hori ef d., 1980a,b; Rasheed and Gardner, 198I ). A structural heteromorphism has been detected in the number 2 chromosome of ACR patients (Gardner et al.. 1981). One member of the No. 2 pair in ACR patients is irregular in the region of 14.2 to 21 .O. Whether this defect is inborn or congenital remains to be seen. In addition, hyperploid cells have been found in ACR patients by several laboratories (Danes, 1977; Delhanty et ul., 1980; Gardner et ul., 198 1). However, the nature of these cells and the mechanisms involved remain unclear. The suggestion was made that the occasional occurrence of very high levels of tetraploid cells in certain patients is a manifestation of the tendency for clone formation (Delhanty and Davis, 1981).

IV. Cell Culture Studies on Autosomal Dominant Syndromes (Other Than ACR) and Chromosome Instability Syndromes The growth properties and radiosensitivity of skin fibroblasts derived from hereditary retinoblastoma (RB) patients have been investigated by several laboratories. SF from bilateral RB patients were found to grow in a serum-deprived environment, and to form colonies at higher efficiency than cells taken from normal individuals (Phillips et al., 1979). SF from RB patients were also found to be uniformly more sensitive to X-irradiation in some cases (Nove et ul.,

74

LEVY KOPELOVICH

1981), while other laboratories have not been able to confirm this observation for all RB patients (Phillips et al., 1979). The discrepancy in these results may be explained through a recent study showing that the unusual sensitivity of RH cells to y-irradiation correlates with the D deletion on chromosome 13; specifically, a locus on 13q14 (Nove et al., 1981; Weichselbaum et al., 1980). In this reg,ard, it has been demonstrated that the locus for the human polymorphic esterase D and the D deletion both lie in the middle of band 13q14. This close linkage cl.11 ows esterase D to be a useful clinical marker for I3q-RB (Sparkes er al., 1980). To date, no other studies have been reported on cultured SF from RB patients or any other autosomal dominant syndromes (Knudson, 1980). Such studies would contribute considerably to our understanding of susceptibility mechanisms in human cancer. The autosomal recessive syndromes Fanconi’s anemia (FA), ataxia telangiectasia (AT), Bloom’s syndrome (BS), and xeroderma pigmentosum (XP) all show an increased cancer incidence in the homozygote gene carriers (Cairns, 198I ; Hecht and McCaw, 1977; German, 1972). While XP has been also included in this group (e.g., Section VII), it actually represents several complementation groups which do not show an increase in spontaneous chromosome aberrations, nor in SCE, although pseudodiploid clones have been described in cultured XP cells (Ponder, 1980). Cultured FA cells show a spontaneous increase in chromatid breaks and chromosomal rearrangements (German, 1972; Sasaki and Tonomura, 1973). They also show increased sensitivity to bifunctional alkylating agents and carcinogens (Auerbach and Wollman, 1976, 1978; Sasaki and Tonomura, 1973). Cultured BS cells show an increase in SCE occurring spontaneously often associated with the quadriradial configuration which arises by equal exchange of chromatid fragments generally between homologous chromosomes (Chaganti et al., 1974). AT cells grown in culture show spontaneous chromosome breakage (Chen tst af., 1978; Ponder, 1980). Pseudoclones of circulating AT lymphocytes involving a translocation at chromosome 14, without apparent loss of genetic material, are often observed (Review by Ponder, 1980). AT cells are unusually sensitive to yirradiation (Higurashi and Conen, 1973; Weichselbaum et al., 1980). They also show an increased sensitivity to chemical agents (Auerbach and Wollman, 1979; Ponder, 1980). Cultured XP cells are, in general, sensitive to UV-induced damage and to a variety of chemical mutagens and carcinogens (review by Ponder, 1980). In a recently published study it was shown that in two of the syndromes (FA and AT), chronic exposure to a low concentration of diepoxybutane induced extensive chromosome damage without reduction in cell viability. In contrast, exposure to the same concentration of this carcinogen had no such clastogenic effect in BS, XP, or normal fibroblasts (Auerbach and Wollman, 1979). These results suggest a genetically determined separation of the clastogenic effect from

GENETIC PREDISPOSITION TO CANCER IN MAN

75

the cytotoxic effect in sensitive cells. In this respect, SF from three of these syndromes (FA, AT, XP) have been shown to be deficient in repair of damage to DNA by environmental agents (Setlow, 1978). It is not clear, however, whether the occurrence of SCE in BS cells, spontaneous or induced, is associated with error-prone repair processes. Several conflicting studies about the nature of mechanisms concerned with SCE have been reported (Rudiger et al., 1980; Shiraishi et al., 198 I ; Tice et al., 1978). It was recently suggested that a factor found in cells of several species is capable of correcting the defect responsible for the high SCE frequencies in BS cells. The absence of such a factor in BS cells leads to the high SCE frequencies (Shiraishi et a/.. 1981). Interestingly, an anomalous intracellular distribution of topoisomerase activity was found in placental extracts of an FA individual (Wunder et al., 1981). The significance of this observation with respect to DNA repair processes at the chromosome and DNA levels is not clear at present. Another metabolic abnormality in FA fibroblasts is related to an increased sensitivity to killing by the purine analog 6-mercaptopurine (6-MP). It involves the conversion of 6-MP to the nucleotide level (Frazelle et al., 1981). One of the early observations on FA cells was the demonstration by Todaro and collaborators on the increased sensitivity of these cells to transformation by SV40 (Dosik et al., 1967). It is of interest that both FA and ACR (Kopelovich and Sirlin, 1980) cells were equally affected by this well-defined viral probe. Obviously, it is important to know whether differences between various forms of heritable cancer caused by a double dose (FA) and those caused by a single dose (ACR) of mutant genes provide any clues about the primary lesion. It is conceivable that transformation by SV40 is facilitated through different factors in the two cell types. For example, the possibility that SV40 functions primarily as an initiator, or as promoter, or both could be at least partially resolved through the use of these mutant cells. In our opinion, one cell type is primarily initiated (ACR) and the other is primarily promoted (FA) (e.g., Section VII). Kinetic studies on cultured SF from FA individuals showed decreased growth rate and increased generation times compared to normal SF (Weksberg et d., 1979). The occurrence of phenotypic abnormalities in cultured cells of FA, AT, BS, and XP gene carriers, both homozygotes and heterozygotes, provides a method for the clinical diagnosis of these patients (review by Auerbach and Wollman, 1979; Ponder, 1980; Wunder et al., 1981).

V. On the Question of Tumor Promotion Malignant transformation is a multiphase process apparently caused by carcinogens and subject to the influence of promoters (Berenblum, 1979; Foulds,

76

LEVY KOPELOVICH

1969; Heidlberger, 1975; Miller, 1978). A potent class of tumor-promoting agents are the naturally occurring phorbol esters (Berenblum, 1979; Hecker, 1971; Van Duuren, 1976), such as 12-0-tetradecanoyl phorbol 13-acetate (TPA). Through the use of phorbol esters, a two-stage process of malignant transformation has been demonstrated in vivo and more recently in cell culture systems (review by Blumberg, 1980, 1981; Weinstein et al., 1980). TPA effected further transformation of initiated ACR cells. However, the phenotypic profile of ACR cells chronically exposed to TPA although presenting a change toward a more transformed phenotype (e.g., growth in agar) was, in large measure, neither stable, nor uniform during consecutive passages, or for a given cell strain, during different periods of TPA application (Kopelovich, 1978b; Kopelovich and Bias, 1979; Kopelovich et al., 1979a). Previous attempts in our laboratory to inoculate TPA-treated ACR cells subcutaneously in the nude mouse have failed to yield any tumors. Recently, the inoculation of these cells from 10 different individuals into the anterior chamber of the eye of a nude mouse gave rise to a moderately differentiated fibrosarcoma which is characterized by uniformly appearing, highly basophilic, fibroblast-like cells. These cells, however, did not necessarily grow in agar, nor did they acquire an infinite life span in vitro (Kopelovich, 1982, Kopelovich er al., 1979a). Indeed, not all cells obtained from spontaneously occurring human tumors appear to grow indefinitely in culture (Smets, 1980). At present, we are trying to isolate the cells growing in the anterior chamber and establish their human origin by a karyotypic analysis. TPA has been shown to enhance the stable transformation of murine cells and more recently of human foreskin fibroblasts previously exposed to a carcinogen (review by Blumberg, 1980, 1981; Kopelovich, 1982; Weinstein et al., 1980). Thus, our results may indicate that the ACR mutation is a complete one for malignancy, representing an initiated state (Kopelovich, 1980a), and that the chronic application of TPA, in support of the two-stage “Berenblum-hypothesis” (Berenblum, 1979) can precipitate the final oncogenic event. Alternatively, the enhancement of agar-insensitive colonies and the growth of cells in vivo may represent an intermediary state, similar perhaps to the TPA-induced papillomas in the mouse skin model (review by Blumberg, 1980, 1981; Slaga et al., 1978), or to the clinical appearance of polyps in the colon; these may or may not regress spontaneously upon withdrawal of the insulting factor. The latter would suggest that an additional mutation(s) is necessary for the malignant transformation of ACR cells with certain promoters acting during all phases of oncogenesis to increase the probability of expression of the malignant phenotype. The ability to understand reversibility and adaptation (acquired resistance) in relation to inflammation, hyperplasia (transient and sustained), and promotion in TPA-sensitive animals (review by Slaga et al., 1978) should provide insight about the effects of TPA in human cells in vitro.

GENETIC PREDISPOSITION TO CANCER IN MAN

77

It has been suggested that promotion by TPA can be dissected into several stages although their nature is not well understood (Slaga et al., 1980). Based on our experience, we propose that the first stage involves the induction of a new variant (Kopelovich, 1982), and that the second stage(s) involves a further amplification of the transformed phenotype leading eventually to malignancy (see below our discussion of the chromosomal instability syndromes). These stages can now be monitored in vitro through a kinetic analysis of the bimodal doseresponse to TPA in human cells (Gansler and Kopelovich, 1981; Kopelovich and Gardner, 1982). We have recently found that the following TPA analogs, in diminishing order of toxicity, showed a bimodel dose-response: PDD, PDBu, PDB, and PDA. However, they were all considerably less toxic than TPA. On the other hand, mezerine exhibited the usual dose-response, characteristic of a drug with a single mode of action, although its toxicity greatly exceeded that shown for 4-0MeTPA. A bimodal dose-response to TPA has recently been observed in cloned ACR skin fibroblasts (Kopelovich, in preparation). This further supports our contention that a single cell population of the same lineage possesses at least two receptor sites to TPA (Kopelovich, 1982). Following our observation, Dr. M. Eisinger of this Institute found a bimodel dose-response to TPA in human keratinocytes, but only the stimulatory portion of the curve for melanoctyes (personal communication). If our assumption about the potential role of these two types of receptors is correct, the expected incidence of melanoma should be relatively high under conditions of sufficient exposure to potential promoters and/or carcinogens. Indeed, the incidence of melanotic foci increases considerably with age. This mechanism may also account for the cafe au lait spots in neurofibromatosis patients. The elucidation of the role of the two types of receptors to TPA in human cells and the search for a putative promoter-like factor(s) (Todaro et al., 1978; Weinstein et a/., 1980) occurring endogenously, fashioned perhaps after the morphines/endorphins system (Snyder, 1977) would be of great interest. Promotion is conceivably the single most important step(s) in the genesis of human cancer. Various hypotheses have been advanced to explain the promoting action of TPA. Thus, mechanisms involving sequential phenotypic alterations in the response of the same cell type to TPA, presumably leading to increased probability for the expression of the malignant phenotype, have been indicated (Weinstein et al., 1980; Yotti et al., 1979). Genomic effect due to TPA have also been proposed. For example: (a) aberrant mitotic segregation event(s) facilitated by a TPA-induced SCE, which might lead to the cosegregation of recessive chromosomal lesions (Kinsella and Radman, 1978), (b) a TPA-mediated enhancement in the frequency of specific locus mutations following insult with a carcinogen(s) (Trosko er al., 1977), (c) a TPA-induced aneuploidy concomitant

78

LEVY KOPELOVICH

with the expression of neoantigens (Kopelovich, 1982; Kopelovich r't al., 1979a), and (d) selective gene amplification (Varshavsky , 1981). Partial evidence that promotion by TPA does not proceed through mechanisms involving either genetic recombination or the altered suppression of newly mutated alleles has also been presented (Thompson et al., 1980). We speculate that both initiators and promotors, in that order, can be mutagenic. A TPA-induced aneuploidy , associated with aberrant chromosomal segregation and gene amplification, if stable, may conceivably be consistent with the expression of a second mutation(s). In this respect, we have recently noted that TPA inhibited the expression (about 99%) of SV40-induced T-antigen display in ACR cells. Virus adsorption was not effected by this treatment. TPA has also been shown to induce the expression of persisting viral genomes (Zur Hausen et al., 1979). In addition, TPA effected the reexpression of malignancy in (human X mouse) hybrids whose tumorigenic phenotype was stably suppressed (Chopan and Kopelovich, 1981a,b,c). The complex nature of changes occurring in eukaryotic chromosomes (e.g., transposable genes) has been recently documented (Abelson and Butz, 1980), and it may account not only for the prorrioting (Kopelovich, 1982), but for the initiating mutatioil as well (Cairns, 1981). Clearly, the proper monitoring of subtle mutations and mutation frequencies in human cells due to TPA will be of great significance.

VI. Genetic Mechanisms

In order to identify the molecular events that effect a change from nomial to malignant cells, we have to understand the nature of the genetic determinants which are associated with transformation in vitro and tumorigenicity in vivo. It is generally believed that all forms of cancer are due to heritable and permanent changes in the cell genome (Ames et al., 1975; Knudson and Strong, 1973). A model that suggests the occurrence of two major mutations in the genesis of cancer has been proposed (Knudson and Strong, 1972). A view that considers tumor cells an expression of a particular state of differentiation rather than a genetic variant has also been considered (Mintz and Illmensee, 1975). Along similar lines recent results involving radiation-induced transformation of C3H/ 10TY2 cells suggested that malignant conversion in this system may have been due to epigenetic mechanisms (Kennedy et al., 1980). Conceivably, both genetic and epigenetic mechanisms might be associated with the initiation and maintenance of the malignant state. Based on our experience with the ACR model, we contend that transformil t'Ion in vitro, which we equate with cancer predisposition in vivo, is due to an autosomal dominant mutation for most forms of cancer, although a pleiotropic effect by ancillary genes (polygenic effect) could not be discounted. An underly-

GENETIC PREDISPOSITION TO CANCER IN MAN

79

ing assumption in these studies is that the site of this mutation in nonhereditary cancers is similar to that in hereditary cancers for the same type of cancer. The nature of the second mutation giving rise to a cancer cell from a transformed cell is not known, but it might apparently involve a transition from the heterozygous to the homozygous state (see above). The homozygous state can be achieved through a recessive mutation (by an addition or a deletion) or a codominant mutation. Implicit in either mechanism is the possibility of a gene dosage effect. A position effect has also been considered (Rowley, 1980). Whether this transition occurs at a specific locus in the homologous chromosome remains to be established. Alternatively, it may represent a new double dose mutation, occurring on a different locus of the original chromosome pair, or on a different chromosome. Theoretical considerations in support of the concept that malignancy is expressed as a recessive trait have been made by Comings (1973). In general, increase in cell ploidy and chromosomal aberrations, spontaneous or induced, have been used as evidence for this concept (Comings, 1973; Kinsella and Radman, 1978; Kopelovich, 1980b; Ohno, 1971). It is of interest that TPA alone has been shown to cause similar effects (Kopelovich et al., 1979a). Experiments on cell hybridization (Chopan and Koplovich 1981a,b,c; Croche, 1980; Sabin, 1981) and on virus-transformed cells (Temin, 1980) appear to support this contention, although different results with cell hybrids have also been reported (Croche, 1980; Sabin, 1981). Along these lines, we have recently shown that a cell fusion between two transformed parents of the same lineage [human (ACR) X mouse (A9)], one of which was malignant (A9), could nevertheless lead to a nontumorigenic cell phenotype. Although the cell hybrids were essentially nontumorigenic, their transformed phenotype closely resembled the malignant A9 parent cell (Chopan and Kopelovich, 1981b). Thus, the human parent cell effected a genetic alteration consistent with the suppression of malignancy, while it had no apparent effect on the transformed phenotype of A9. Moreover, genetic recombination through cell fusion was expressed not only at the level of the hybrid cell itself, but also in its ability to elaborate a tumor-suppressing factor, both in vivo and in vitro, presumably a gene product which was not previously expressed by either parent cell (Chopan and Kopelovich, 1981~).

VII. Is Genetic Predisposition to Cancer an Autosomal Dominant Trait We believe that autosomal dominant traits which predispose individuals to cancer truely reflect genetic information [a specific mutation(s)] directly related to this disease. Stated differently, genetic predisposition to cancer in the autosoma1 dominant syndromes at the cellular level is associated with cancer initiation. In contrast, autosomal recessive traits, often regarded as cancer syndromes

80

LEVY KOPELOVICH

(Cairns, 1981; Ponder, 1980), reflect the state of the genome in cells of gene carriers. They are not directly linked to cancer. The high degree of DNA instability in these syndromes, spontaneous or induced, might produce a multitude of deletions and recombinational events, thereby increasing the probability of a specific mutation(s) leading to malignant expression. Malignant conversion in the later would, therefore, represent a different pathway from the clinical phenotype generally associated with these gene carriers (Mulvihill et al., 1977). Our contentions is that genetic predisposition to cancer in the recessive traits at the cellular level is presumably associated with the first phase of cancer promotion, i.e.. the induction of a new varianr (see above), with a certain probability that initiation might follow. In our view, xeroderma pigmentosum as well, represents a genetic aberration consistent with cancer promotion. Other mechanisms have also been provided for this disease (Bridges, 1981; Cairns, 1981). In this caanection, a reassessment of cancer predisposition of Fanconi anemia heterozygotes showed no overall excess of cancers or cancer deaths for any age or sex category of blood relatives, and no unusual number of cancers among the obligate heterozygotes (Swift et al., 1980). Obviously, larger samples of patients will be necessary to further ascertain these differences. Compared with the dominant syndromes, recessive traits would be anticipated to have the following characteristics: (1) a stronger impact by environmental factors, (2) a lower cancer incidence in the gene carriers, and (3) a shorter tumor latency. An autosomal dominant pattern has also been recognized in a large number of familial aggregates predisposed to various forms of cancer. These comprise a large segment of all cancers reported in the United States and elsewhere (Anderson and Romsdahl, 1977; Lynch et al., 1977). We further believe that a strong genetic component exists in the sporadic forms of cancer, the expression of which would be consistent with a dominant trait, albeit with only partial penetrance. A recent study at the Memorial Sloan-Kettering Cancer Center on over 1350 patients representing a variety of primary cancers demonstrated that a familial tendency to develop cancer exists in the general population of cancer patients (Schneider, 1981). This was not, however, related to the age of onset in the proband. Indeed, a late age of onset (about 60 years) has been associated with the Muir’s syndrome (Anderson, 1980), an autosomal dominant trait, in which essentially complete penetrance has been observed and in which multiple cancer forms occur in most affected individuals, colon cancer in particular. It is conceivable, therefore, that the basic genetic mechanisms which underlie a predisposition to cancer are essentially similar qualitatively in the autosomal dominant cancer syndromes, the familial cancer syndromes, and the sporadic forms of cancer. Quantitative differences could occur due to a gene dosage effect at the primary locus and at ancillary loci which modulate the expression of malignancy.

GENETIC PREDISPOSITION TO CANCER IN MAN

81

To reiterate, we feel that autosomal domnant traits which predispose to cancer represent an appropriate model system in which to study cancer-specific mutations (genes), whereas autosomal recessive traits should lend themselves to studies about certain stages of cancer promotion. It is also conceivable that the occasional occurrence of spontaneous polyploid cells in certain ACR individuals (Delhanty and Davis, 1981) may represent an example of an initiated cell in situ which has also undergone the first stage of tumor promotion (e.g., Section V).

VIII. On the Question of Cancer Prognosis and Cancer Control We propose that ACR gene carriers within recognized ACR clusters can be diagnosed at present through our tests on skin fibroblasts, with sufficient certainty to warrant immediate action. Since in humans the ultimate and consequently the only objective criteria of disease are the clinical manifestations of cancer, we recommend that individuals diagnosed by us as potential carriers should be advised of the possibility and should be monitored for the earliest clinical evidence. At that time, a rational approach with regard to therapeutic modalities can be immediately implemented. We further suggest that this approach is costeffective and should be adopted as public policy in the United States and elsewhere. Concurrently, attempts will be made to increase the number and specificity of tests used for cancer predisposition. Furthermore, based on our present knowledge about the ACR risk-profile, it seems that the time has come for a major undertaking to screen for persons who are likely to be at an increased risk of cancer, perhaps through walk-in clinics. The use of all available tests together, repeated skin biopsies from the same individual, and a large sample may help define confidence limits for determining the probable risk of cancer in any population with a certain value. In addition, the specificity of these tests with regard to various forms of malignancy should be ascertained in a large number of cancer syndromes. For example, persons from colon cancer-prone families without adenomatous polyposis showed no disturbance of actin-containing cables in skin fibroblasts (Kopelovich et ul., 19801, but can be distinguished through the use of a tumor promotor (Kopelovich 1981). Along similar lines, increased in virro tetraploidy has been demonstrated only for high-risk tissues from Gardner patients, but not for non-Gardner individuals with adenomatous polyps (Danes, 1977). In this respect, SF derived from individuals with the Huntington syndrome, and autosomal dominant trait, were similar to normal SF with respect to growth parameters and susceptibility to KiMSV (Goetz et a / . . 1981; Miller and Rasheed, 1981). To further facilitate the identification of high-risk individuals, fibroblastic cells residing in close proximity to a potential tumor site from well-characterized pedigrees, in which the target organ can be largely anticipated, should be used.

82

LEVY KOPELOVICH

This is based on the fact that fibroblastic cells growing at a location adjaceni: to a primary tumor strongly express the transformed phenotype (Kopelovich, unpublished; Azzarone et a/., 1976; Smith et al., 1976). This effect may be reminiscent of the ectopic, hormone-like, growth regulators produced by either normal or transformed cells (Chopan and Kopelovich, 1981~;Goldenberg and Pavia, 1981; Todaro et al., 1978). Another approach which might facilitate an effective screening program would be to find additional and meaningful probes to discern individuals at risk. Our experience suggests that the use of exogenous, well-defined, probes (virus, tumor promoters, others) amplifies genetic differences existing in cells considerably more than would be anticipated frorn the measurement of abnormal phenotypic expressions in unperturbed cells. For example, through the use of KiMSV a difference of 100- to 1000-fold has been observed between ACR and normal cells, while the determination of growth parameters occurring spontaneously showed a difference of about ?-fold (Kopelovich, 1980a). This result conforms with our current knowledge aboiit the regulation and expression of genetic determinants in biological systems. A likely explanation for the increased susceptibility of ACR cells to KiMSV would be a gene dosgae effect possibly in conjunction with an intracellular promoter-like factor (Harrison and Auersperg, 1981: Kopelovich, 1982). The identification of “risk profiles” might provide invaluable information about cancer prognosis and cancer control. Implicit in this statement arc: two important concepts we have developed in the ACR cell system: (a) that like in ACR, skin fibroblasts (or other readily accessible specimens, e.g., lymphocytes, urine, serum etc.) of cancer-prone individuals will show phenotypic expressions that are associated with a predisposition to cancer, and (b) that like in ACR, most of the cells although initiated, may nevertheless remain normal through a. long latent period and that, indeed, only a very small fraction of these cells in the target organ will become tumorigenic. In the first instance, we contend that individuals predisposed to cancer will manifest this lability systemically, albeit to a varying degree. In the second instance, we contend that tumor promotion, and not tumor initiation, is the critical period in the development of neoplasia. It follows then that once individuals at risk (predisposed) have been identified, intervention at several levels can ensue: dietary, psychological, genetic, occupational, and most importantly chemotherapeutic measures. The use of drugs, assuming the model of molecular events associated with promotion is correct (Kopelovich, 1982), should be designed to stabilize DNA (nuclear, mitochondrial) directly or indirectly (Troll, 1981). The use of vitamin A and closely related analogs, although controversial (Meyskens, 1981; Schroder and Black, 1980; Sporn and Newton, 1979), is an example of chemical intervention at the phenotypic level. Presumably, both genotropic and phenotropic drugs with ii high therapeutic index could be used to inhibit cancer promotion in man. Thus, intervention at the level of promotion may, in reality, be the most

GENETIC PREDlSPOSlTlON TO CANCER 1N MAN

83

important phase in cancer control of individuals at risk, including sporadic cases of cancer. This approach if successful, i.e., an accepted rate of proper diagnosis of individuals at risk and an accepted quality of prophylaxis applied to individuals at risk, would probably prove to be highly effective in controlling costs as well.

IX. Conclusion We propose that skin fibroblasts derived from normal-appearing biopsies of ACR gene carriers exist in an initiated state due to a dominant mutation. Based on our studies with the ACR cell system, we further suggest that, while an initiated state is essential to cancer development, not all initiated cells necessarily develop into cancerous cells. The genetic make-up of an initiated cell has been established through a linkage between the abnormal phenotypic markers and the pedigree profiles and through cell hybridization, including initial analysis of gene products. We believe that it is consistent with an autosomal dominant trait. In contrast, cells from patients who are homozygous for the chromosomal breakage syndromes, including xeroderma pigmentosum, represent an experiment of nature which, presumably, underlies factors associated with cancer promotion in man. We have demonstrated that ACR cells can be differentially transformed by oncogenic agents and that they can proliferate in vitro after exposure to a tumor promoter (TPA). This simple experimental model provides a novel system for the study of tumor promotion in vitro. The apparent susceptibility of ACR cells to further transformation by oncogenic viruses and chemical and physical agents indicates that genetic information residing within these cells, probably in the form of a relatively limited and specific number of DNA sequences associated with the ACR mutation, renders them more sensitive to these three distinct classes of carcinogens. We further suggest that through the use of TPA, various stages associated with cancer development in humans, i.e., initiation through promotion and progression, can be identified in virro. Attempts to apply these results in vivo are currently in progress. We submit that, through our tests on skin fibroblasts, ACR gene carriers within recognized ACR clusters can be diagnosed at present with sufficient certainty to warrant immediate action. In addition, it seems that the time has arrived for a major undertaking to screen for persons who are likely to be at an increased risk of cancer, perhaps through walk-in clinics. An underlying assumption in these studies is that predisposition to cancer, in general, is associated with an autosomal dominant trait in obligatory heterozygote gene carriers.

84

LEVY KOPELOVICH ACKNOWLEDGMENTS

We thank Ms. P. Monaghan, Ms. R. Vuolo, and Ms. T . Shapiro for excellent technical assistance. This work was supported by grants CA-19529 and CA-21623 from the National Large Bowel Cancer Project, National Cancer Institute. and National Cancer Institute grant CA-08748.

REFERENCES Aaronson. S . , and Todaro, G. (1975). Nature (London) 225, 458-459. Abelson, J., and Butz, E. (1980). Science 209, 137-1438. Alexander, P. (1972). Narure (London) 235, 137-141. Ali, U. I . , Mautner, V., Lanza, R., and Hynes, R. 0. (1977). Cell 11, 115-126. Alm, T., and Licznerski, G. (1973). In “Clinics in Gastroenterology” (E. McConnel, ed.), Vul. 2, pp. 577-601. Saunders, Philadelphia, Pennsylvania. Ames, B. N., McCann, J., and Yamasaki, B. (1975). Murat. Res. 31, 347-364. Anderson, D. E. (1980). Cancer 45, 1103-1107. Anderson, D. E.,and Romsdahl, M. M. (1977). In “Genetics of Human Cancer” (J. J. Mulvihill, R. W. Miller, and J. F. Fraumeni, Jr., eds.), pp. 357-362. Raven, New York. Auerbach, A. D., and Wolman, S . R. (1976). Nature (London) 261, 494-496. Auerbach. A. D., and Wolman, S. R. (1978). Nature (London) 271, 69-71. Auerbach, A. D., and Wolman, S . R. (1979). Cancer Genet. Cytogener. 1, 21-28. Azzarone, B., Pedulla, D., and Romanzi, C. A. (1976). Nature (London) 262, 74-76. Berenblum, 1. (1979). In “Carcinogens: Identification and Mechanisms of Action” (A. C. Griffin and C. R. Shaw, eds.), pp. 25-36. Raven, New York. Bieme, 0. R., and Watson, J. A. (1976). Proc. Nail. Acad. Sci. U.S.A. 73, 2735-2739. Blumberg, P. M. (1980, 1981). In virro studies on the mode of action of phorbol esters, potent tumor promoters. CRC Criri. Rev. Toxicol. Parts I and 11, 153-197, 19LL234. Bridges, B. (1981). Carcinogenesis 2, 471-472. Brown, D. D. (1981). Science 211, 667-674. Brown, M. S., and Goldstein, J. L. (1974). J . B i d . Chem. 249, 73067314. Cairns, J. (1981). Nature (London) 289, 353-357. Chaganti, R. S . K., Schonberg, S., and German, J. (1974). Proc. Narl. Acad. Sci. U.S.,1. 71, 4508-45 12. Chen, P. C., Lavin, M. F., Kidson, C., and Moss, D. (1978). Nature (London) 274, 484486. Chopan, M., and Kopelovich, L. (1981a). Exp. Cell Biol. 79, 78-90. Chopan, M., and Kopelovich, L. (1981b). Exp. CeflEiol. 49, 132-141. Chopan, M., and Kopelovich, L. (1981~).Oncology 38, 240-242. Coggins, J. H., and Anderson, N. G. (1974). Adv. Cancer Res. 19, 105-165. Cohen, A. J., Li, R. P., Berg, S., Marchetto. D. J., Tsai, S., Jacobs, S . C., and Brown, R. S. (1979). New Engl. J . Med. 201, 592-595. Collett, M.. Erikson, E., Purchio, A. F., Brugge, J. S., and Erikson, R. L. (1979). Proc. Narl. Acad. Sci. U.S.A. 76, 3159-3163. Comings, D. E. (1973). Proc. Nail. Acad. Sci. U.S.A. 70, 33243328. Croche, C. M. (1980). Eiochim. Biophys. Acia 605, 41 1-430. Danes, B. S. (1977). Cancer 36, 2327-2333. Deleo. A. B., Jay, G., Appella, E., Dubois, G. C., Law, L. W.. and Old, L. J. (1979). Proc. Nail. Acad. Sci. U.S.A. 76, 2420-2424. Delhanty, 1. D., and Davis, M. B. (1981). Inr. Congr. Hum. Genet. 6th. Jerusalem p. 254.

GENETIC PREDISPOSITION TO CANCER IN MAN

85

Delhanty, J. D., Pritchard, M. B., Bussey, H. J., and Morson, B. C. (1980). Lancet 21, 1365. Dosik, H., Hsu, L. Y., Todaro, G. J., Lee, S. L., Hirschhorn, K., Selirio, E. A., and Alter, A. A. (1969). Blood 14, 1008-1014. Ellis, R. W., DeFeo, D., Shih, T. Y., Gonda, M. A., Young, H. A., Tsuchida, N., Lowy, D. R., and Scolnick, E. M. (1981). Nature (London) 292, 506511. Foulds, L. ( 1969). “Current Concepts of Neoplastic Transformation in vitro. Neoplastic Transformation.” Academic Press, New York. Fraumeni, I. F., Jr. (1977). In “Genetics of Human Cancer” (I. J. Mulvihill, R. W. Miller, and J. F. Fraumeni, Jr., eds.), pp. 223-235. Raven, New York. Frazelle, J. H., Harris, J. S., and Swift, M. (1981). Mutation Res. 80, 373-379. Gander, T., and Kopelovich, L. (1981). Cancer Lett. 13, 315-325. Gardner. E., and Richards, R. (1953). Am. J. Hum. Genet. 5, 139-148. Gardner, E. J . , Rogers, S. W.. and Woodward, S. T. (1982). Cancer. in press. German, J. (1972). Prog. Med. Genet. 8, 61-101. Goetz, I. E., Roberts, E., and Warren, I. (1981). Am. J . Genet. 33, 187-196. Goldenberg, D. M., and Pavia, R. A. (1981). Science 212, 65-67. Goldman, R.. Pollard, T.. and Rosenbaum, J. (1976). “Cell Motility,” Vol. 1-3, pp. 1-1373. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Harrison, J., and Auersperg, N. (1981). Science 213, 218-219. Hayward, W. S . , Neel, B. G., and Astrin, S. M. (1981). Nature (London) 290, 475480. Hecht, F., and McCaw, B. K. (1977). In “Genetic of Human Cancer” (J. J. Mulvihill, R. W. Miller, and J. F. Fraumeni, IT.. eds.), pp. 105-125. Raven, New York. Hecker, E. (1971). Methods Cancer Res. 6 , 43-84. Heidelberger, C. (1975). Annu. Rev. Eiochem. 44, 7S121. Higurashi. M., and Conen, P. E. (1973). Cancer 32, 380-383. Hori, T., Murata, M., and Utsunomiya, J. (1980). G A ” 71, 628-636. Huebner, R. J . , and Todaro, G. J. (1969). Proc. Natl. Acad. Sci. U.S.A. 64, 1087-1094. Hynes, R. 0..and Destree, A. T. (1978). Cell 15, 875-886. Kennedy, A. R., Fox, M., Murphy, G., and Little, I. B. ( 1980). Proc. Natl. Acad. Sci. U.S.A. 77, 7262-7266. Kinsella, A. K., and Radman, M. (1978). Proc. Natl. Acad. Sci. U.S.A. 75, 6149-6153. Klement, V . , Friedman, M., McAllister, R., Nelson-Rees, W., and Huebner, R. (1971). J . Nurl. Cancer Inst. 47, 65-73. Knudson, A. G. (1980). ICRDB 197, 1-143. Knudson, A. G. (1981). In “Genes, Chromosome and Neoplasia” (F. F. Arrighi and P. N. Rao eds.), pp. 453-462. Raven, New York. Knudson, A. G., and Strong, L. C. (1972). Am. J. Hum. Genet. 24, 514-532. Kopelovich, L. (1977a). Cuncer 40,2534-2541. Kopelovich, L. (1977b). In “Workshop on Cancer Invasion and Metastasis Biologic Mechanisms and Therapy” (S. Day, ed.), pp. 375-387. Raven, New York. Kopelovich, L. (1978a). Monogr. Hum. Genet. 10, 154-168. Kopelovich. L. (1978b). Cold Spring Harbor Symp. p. 22. Kopelovich, L. (1980a). In “Progress in Cancer Research” (S. I. Winawer, P. Sherlock, and D. Schottenfeld, eds.), pp. 91-108. Raven, New York. Kopelovich, L. ( 1980b). In “Carcinogenesis: Fundamental Mechanisms and Environmental Effects” (B. Pullman, P. 0. P. Tso, and B. Gelboin, eds.), pp. 409-416. Reidel, New York. Kopelovich, L. (1981). Cancer Lett 12, 67-74. Kopelovich, L. ( 1982). In “Cocarcinogenesis and Biological Effects of Tumor Promoters” (E. Hecker et ul., eds.). Raven, New York, in press. Kopelovich, L., and Bias, N. (1979). Exp. CellBiol. 48, 207-217.

86

LEVY KOPELOVICH

Kopelovich, L., and Gardner, E. (1982). Cancer (in press). Kopelovich, L., and Sirlin, S. (1980). Cancer 45, 1108-1 I 1 I . Kopelovich, L., Conlon, S., and Pollack, R. (1977). Proc. Natl. Acad. Sci. U.S.A. 74, 3019-3022. Kopelovich, L., Bias, N., and Helson, L. (1979a). Nature (London) 282, 619-621. Kopelovich, L., Lipkin, M., Blattner, W. A., Fraumeni, J. F., Jr., Lynch, H. J., and Pollack, R. E. (1980). Int. J . Cancer, 26, 301-307. Kopelovich, L., Drozdoff, V., and Zeitz, L. (1981). Proc. Am. Assoc. Cancer Res. 21, 192. Kopelovich, L., Pfeffer, L., and Bias, N. (1979b). Cancer 43, 218-223. Linzer, H. I . , and Levine, A. J. (1979). Cell 17, 43-52. Lynch, H. T. (1976). “Cancer Genetics,” pp. 1-639. Thomas, Springfield, Illinois. Lynch, H. T., Harris, R. E., Lynch, P. M., Guirgis, H. A., Lynch, J. P., and Bardawil, W. A. (1977). Cancer 40, 1845-1849. McConnel, R. B. (1980). In “Progress in Cancer Research” ( S . J. Winawer, P. Sherlock, ,md D. Schottenfeld, eds.), pp. 69-72. Raven, New York. Malenbaum, G. B., Gilbert, F., Nichols, W. W., Hill, R., Shields, J., and Meadows, A. T. (1981). Cancer Genet. Cytogenetic. 3, 243-250. Marinello, M. J . , Morita, M., and Sandberg, A. A. (1981). Cancer Genet. Cytogenet. 3, 227-237. Marks, P. A. (1981). Blood 58, 415-419. Meyskens, F. L. (1981). Life Sci. 25, 2323-2327. Miller, C. A., and Rasheed. S. (1981). Am. J . Hum. Genet. 33, 197-202. Miller, E. C. (1978). Cancer Res. 38, 1479-1496. Mintz, B., and Illmensee, K. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 3585-3589. Miyaki, M., Akamatsu, N., Rokutanda, M., Ono,T., Yodhikura, H., Sasaki, M., Tonomura, A., and Utsunomiya, J. (1980a). G A ” 71, 797-803. Miyaki, M., Akamatsu, N., Hirono, U., Suzuki, K., Rokutanda, M., Ono, T., Sasaki, M. S., Tonomum, A., and Utsunomiya, J. (1980b). G A ” 25, 115-126. Morson, B., and Bussey, H. (1971). In “Current Problems in Surgery,” pp. 1-50, Yearbook Medical Publ. Chicago, Illinois. Mulvihill, J. J., Miller, R. W.. and Fraumeni, J. F., Jr. (1977). “Progress in Cancer Research and Therapy,” Vol 3, pp. 1-519. Raven, New York. Norberg, R., Thorstensson, R., Utter, G., and Fagraeus, A. (1979). Eur. J . Biochem. 100,575-585. Nove, J., Nichols, W. W., Weichselbaum, R. R., and Little, J. B. (1981). Murat. Res. 84, 157-167. Ohno, S. (1971). Physiol. Rev. 51, 496526. Oppermann, H., Levinson, A. D., Varmus, H. E., Levintow, L., and Bishop, J. M. (1979). Proc. Natl. Acad. Sci. U.S.A. 76, 1804-1808. Peehl, D. M., and Stanbridge, E. J. (1981). Proc. Natl. Acad. Sci. U.S.A. 78, 3053-3057. Peto, J. (1980). Banbury Rep. 4, 203-213. Pfeffer, L., and Kopelovich, L. (1977). Cell 10, 313-320. Pfeffer, L., Lipkin, M., Stutman, O., and Kopelovich, L. (1976). J. Cell. Physiol. 80, 29-38. Phillips, R. A,, Kosakowka, A., and Gallie, B. L. (1979). Proc. Am. Assoc. Cancer Res. 20, 266. Ponder, B. A. J. (1980). Biochim. Biophys. Acta 605, 369-410. Rasheed, S., and Gardner, M. (1981). J. Natl. Cancer Inst. 66, 4 H 9 . Reddy, B. S., Mastromarino, A., Gustafson, C., Lipkin, M., and Wynder, B. L. (1976). Cancer 38, 1694- 1698. Reich, E., Rifkind, D. B., and Shaw, E. (1975). “Proteases and Biological Controls,” pp. 1--1080. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Rhim, J. S., Huebner, R. J., Arenstein, P., and Kopelovich, L. (1980). Int. J. Cancer 26,565-569. Rowley, J. D. (1980). Cancer Genet. Cytogenet. 2, 175-198. Rubin, H. (1980). J . Natl. Cancer Inst. 64, 995-1000. Ruddle, F. H. (1981). Nature (London) 294, 115-120.

GENETIC PREDISPOSITION TO CANCER IN MAN

87

Rudiger, H. W.. Bartram, C. R., Harder, W., and Passarge, E. (1980). Am. J. Hum. Genet. 32, 150- 158. Sabin, A. B. (1981). Proc. Nutl. Acud. Sci. U.S.A. 78, 7129-7133. Sachs, L. (1980). J . Nutl. Cancer Inst. 65, 675-679. Sandberg, A. (1980). “The Chromosomes in Human Cancer and Leukemia,” pp. 1-748. Elsevier, Amsterdam. Sasaki, M. S., and Tonumura, S. (1973). Murut. Res. 20, 291-293. Schimke, R. N. (1978). “Genetic and Cancer in Man,” pp. 1-108. Churchill. London. Schliwa, M. (1981). Cell 25, 587-590. Schneider, N. (1981). Ph. D. thesis, Cornell University Graduate School of Medical Sciences. Schroder, E. W.. and Black, P. H. (1980). J . Nutl. Cancer Inst. 65, 671-673. Setlow, R. B. (1978). Nature (London) 271, 713-717. Shiraishi, Y., Matsui, S. I., and Sandberg, A. (1981). Science 212, 820-822. Shkolnik, T., and Sachs, L. (1978). Exp. Cell Res. 113, 197-204. Siniscalco. M. (1979). In “Approaches to Human Linkage Progress in Medical Genetics” (A. G. Steinberg, A. G. Beam, A. G. Molulsky, and B. Childs, eds.), New Series, Vol. 3, pp. 221-307. Saunders, New York. Siperstein, M. D., and Fagan, V. M. (1964). Cancer Res. 24, I 108-1 1 15. Slaga, T. J., Sivak, A,, and Boutwell, R. K. (1978). “Mechanism of Tumor Promotion and Cocarcinogenesis.” Vol. 2, pp. 1-715. Raven, New York Slaga, T. J . , Fischer, S. M., Nelson, K., and Gleason, G. L. (1980). Proc. Nurl. Acud. Sci. U.S.A. 77, 36593668. Smets, L. A. (1980). Biochim. Biophys. Act0 605, 93-1 11. Smith, H. S., Owens, R. B., Hiller, A. J., Nelson-Rees, W. A,, and Johnston, J. 0. (1976). Int. J . Cancer 17, 219-237. Snyder, H. (1977). Sci. Am. 236, 44-66. Sparkes, R. S., Sparkes, M. C., Wilson, M. G., Towner, J. W., Benedict, W., Murphree, A. L., and Yunis, J. J. (1980). Science 208, 1042-1044. Sporn, M. B., and Newton, 0. L. (1979). Fed. Proc. Fed. Am. SOC. Exp. Biol. 38, 2528-2534. Stanbridge, E. J., and Wilkinson, J. (1978). Proc. Nutl. Acad. Sci. U.S.A. 75, 1466-1469. Stonehill, E., and Bendich, A. (1970). Nature (London) 228, 370-372. Strong, L. C. (1981). Science 213, 1501-1503. Swift, M.. Caldwell, R. J., and Chase, C. (1980). J . Nut/. Cancer Insr. 65, 863-867. Temin, H. M. (1980). Symp. Quunr. B i d . 44, 1-8. Thompson, L. H., Baker, R. M., Carrano, A. V . , and Brookman, K. W. (1980). Cancer Res. 40, 3245-325 I . Tice, R., Windler, G., and Rary, J. M. (1978). Nature (London) 273, 538-540. Todaro, G.. and Martin, G. (1967). Exp. Biol. Med. 124, 1232-1236. Todaro, G. J., Delarco, J. E., and Cohen, S. (1978). Nuture (London) 264, 26-31. Tooze, J. (1979). “The Molecular Biology of Tumor Viruses,” pp. 1-743. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Troll, W. (1981). In “Cancer Achievements Challenges, and Prospects for the 1980’s” (J. H. Burchenall and H. F. Oettgen, eds.), Vol. 1, pp. 549555. Grune & Stratton, New York. Troll, W., Rossman, T.,Katz, J., Levitz, M., and Sugimura, T. (1975). “Proteases in Tumor Promotion and Hormone Action in Proteases and Biological Controls” (E. Reich, D. B. Rifkin, and E. Shaw, eds.), pp. 977-987. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Trosko, J. E., Cheng, C. C., Yotti, L. P., and Chu, E. H. Y. (1977). Cancer Res. 37, 188-193. Van Duuren, B. L. (1976). In “Chemical Carcinogens” (C. E. Searle, ed.), Monogr. 173, pp. 24-51. American Chemical Society, Washington, D.C.

88

LEVY KOPELOVICH

Varshavsky, A. (1981). Cell 25, 561-572. Watson, J . (1979). "The Transformed Cell," pp, 1-121. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Watson, I. A. (1973). "Tumor Lipids: Biochemistry and Metabolism," pp. 34-52. American Oil Chemists' Society. Champaign, Illinois. Weichselbaum, R. R., Nove, J . , and Little, J. B. (1980). Cancer 40, 9 2 6 9 2 5 . Weinstein, I. B., Mufson, A., Lee, L. S . , Fisher, P. B., Laskin, J. Horowitz, A. D., and Ivanovic, V. (1980). I n "Fundamental Mechanisms and Environmental Effects" (B. Pullman, P. 0. P. Tso, and H. Gelboin. eds.). Reidel, Amsterdam. Weksberg, R., Buchwald, M.,Sargent, P., Thompson, M. W., and Siminovich, L. (1979). .I. Cell. Physiol. 101, 31 1-323. Wiener, F., Klein, G., and Harris, H. (1971). J . Cell Sci. 8, 681-692. Wunder, E., Burghardt, V., Lang. B., and Hamilton, L. (1981). Hum. Genet. 58, 149-15.5. Yotti, L. P., Chang, C. C., and Troski, J. (1979). Science 206, 1089-1091. Zur Hausen, H., Bomkamm. G. W., Schmidt, R., and Hecker, E. (1979). Proc. Narl. Acad. Sci. U.S.A. 76, 782-785.

INTERNATIONAL REVIEW OF CYTOLOGY. VOL. 77

Membrane Flow via the Golgi Apparatus of Higher Plant Cells DAVIDG. ROBINSON*A N D

UDO K R I S T E N t

*Abteilung Cjtologie des Pflunzenph!t.siologiJc'he,i Instiruts der Universitiit Giittingen, Giirtingen. Federal Republic. c.f Germany. und tlnstitut fur Allgemeine Botunik. Universitiit Humburg, Humburg, Federol Republic of Germany

I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. GA Structure and Biochemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. General Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

111.

IV. V. VI.

VII.

B. Polarity.. . . . . . . . . . . . . . . . . . . . . . . . .............. C. Relationship to the E R . . . . . . . . . . . . . .............. D. Relationship lo the PM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sites of Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A . Secretional Material.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Endomembranes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Secretion Kinetics and Membrane Turnover . . . . . . . . . . . . . . . . . . . Means and Ends . . . . . . . . . . . . . . . . . . . . . ............... Membrane Recycling and the Golgi Appar ............... A . Necessity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Evidence forilndications of in Higher Plants . . . . . . . . . . . . . . . C. Possible Agents of Recycling . . . . . . . . . . . . . . . . . . . . . . . . . . . Concluding Remarks . . . . . . ............................ References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

89 91 91

92 95 98 102 102 105 108 112 114 114

I IS I17 119 120

I. Introduction Membrane flow sensu strictu is defined as a transfer of biomembranes from one compartment of the endomembrane system to another (Franke et a / . , 1971). Bennet (1956) was one of the first to discern the relation between membrane flow and the intracellular transport of products. Today it is well established that the Golgi apparatus (GA) plays a significant role in membrane flow processes, since a large part of the intracellular transport of macromolecules takes place via dictyosomes. Therefore the membranes of the GA may be considered as packing material as well as transportation vehicles for the transfer of polysaccharides and proteins from their sites of synthesis to the plasma membrane (PM) (Chrispeels, 1976; Mollenhauer and M o d , 1980). As a consequence, these transfer processes are thought to constitute a membrane flow beginning at the GA or, when starting at the endoplasmic reticulum (ER), via the GA. This is embodied in the

90

DAVID G . ROBINSON AND UDO KRISTEN

“Endomembrane Concept” as proposed by MorrC and co-workers (MorrC et al., 1971a,b, 1974, 1979; Morr6, 1980; MorrC and Mollenhauer, 1974; MorrC and Van der Woude, 1974; MorrC and Ovtracht, 1977). This concept, thought to be valid for all eukaryotes, integrates the ER and the GA into a functional unit, in which the dictyosomes are envisaged as being the main station of membrane transformation from an ER-like to a PM-like membrane type (see also Whaley et al., 1972; Whaley, 1975). Unfortunately despite their eloquence and appeal the various presentations of this concept contain interpretations of electron micrographs which are of an equivocal nature. There are in addition generalin.ations which, based on the paucity of information in some areas, are pure assumptions. This is particularly the case for higher plants where the only investigated system from both electron microscopic and biochemical viewpoints is essentially the maize root cap cell-a very specialized and untypical cell. Superficially the Endomembrane Concept finds support in a number of investigations on protein-secreting animal gland cells (e.g., Palade, 1975; Rothman, 1975; Jamieson and Palade, 1977; Hand and Oliver, 1977; Oron and Bdolah, 1978), and also from studies on the synthesis and intracellular transport of PM glycoproteins (Rothman et al., 1980; Bergmann et al., 1981; Croze and MorrC, 1981). Certainly in these cases a transfer ER+GA+PM for the substance in question can be regarded as a fact. Unclear is whether the externally directed transport is associated with a bulk or selective (MorrC et af., 1979) flow of membrane or whether a series of membrane “shuttles” between the organelles exist (Meldolesi, 1974a; Palade, 1975). Membrane transfer from the ER to the PM via the GA can be divided into three different phases, the first of which begins at the ER and ends at the forming faces of the dictyosomes. The second phase comprises the flow through the cisternal stack of the dictyosomes, and the third phase extends from the maturing faces of the dictyosomes to the PM. Whereas the latter phase has been confirmed for higher plant cells (Northcote and Pickett-Heaps, 1966; MorrC et al., 1967; Dauwalder and Whaley, 1974; Paul1 and Jones, 1975, 1976; Bowles and Northcote, 1976), the existence of the first and the second phase has not been unequivocally demonstrated in these cells. Only in the cells of some lower plants have there been structural indications for a membrane transfer from the ER or from the nuclear envelope to the GA (Bouck, 1965; Ueda, 1966; Falk, 1967; Massalski and Leedale, 1969; Evans and Christie, 1970; Bracker et af., 1971; Evans et al., 1974; Markey and Wilce, 1975; Pellegrini and Vogt, 1976; Ueda and Noguchi, 1976; Francisco and Roth, 1977; Gunning and Steer, 1975; Menge and Kiermayer, 1977; Dauwalder et af., 1980; Edgar, 1980; Oliveira ct af., 1980). For this reason doubts have arisen as to the general validity of the Endomembrane Concept, as proposed by MorrC and co-workers (Robinson, 1977, 1980a; Robinson and Ray, 1977; Kristen, 1980a).

MEMBRANE FLOW VIA THE GOLGl APPARATUS

91

Robinson ( I980b) has shown through serial sectioning that transition vesicles or direct contacts between the ER and dictyosomes appear, at the most, very infrequently in polysaccharide-secreting cells of higher plants. On the other hand, Kristen ( I980b) has observed frequently appearing connections between the ER and dictyosomes in the primarily protein-secreting ligular cells of fsoetes lacustris. Because of this difference, Kristen ( 1980b, 1982) therefore suggested a restriction of the Endomembrane Concept to the intracellular transport of secretory proteins. Even Mollenhauer and Morre (1980) now appear to be adopting a new standpoint when they say that “no unequivocal evidence is available to show product or membrane transfer between endoplasmic reticulum and GA in plants. This article represents a critical appraisal of the facts with respect to higher plant cells together with an unbiased evaluation of the pertinent information from animal systems in an attempt to bring together the conflicting viewpoints hitherto expressed in the literature. Although other routes which bypass the GA have been suggested (Schnepf, 19694 they will not be dealt with in this article. ”

11. GA Structure and Biochemistry

A. GENERAL REMARKS From the viewpoint of an electron microscopist higher plant dictyosomes very often appear as distinct organelles. They are not as closely associated with the endoplasmic reticulum as is often the case in animal cells (for details see Whaley, 1975). One of the first attempts at depicting a plant dictyosome in three dimensions was that of Drawert and Mix (1961/1962) based upon observations on the giant dictyosomes typical of the desmid Micrusterias rotata. A more detailed model showing cisternae with fenestrated and tubular peripheral regions and, in addition, intercisternal elements in the form of parallel tubular or fibrillar structures was published a few years later by Mollenhauer and Morr6 (1966a). Although peripheral fenestration has often been demonstrated by both negative staining (the “smear technique,” Mollenhauer and Morre, 1966b. 1976a) and by thin sectioning (Menge and Kiermayer, 1977; Kristen, 1980b; Robinson, 1980b) it is not yet clear whether this feature is associated with different types of dictyosomes or is due to different developmental stages. The basic features of higher plant dictyosomes are well known (see for example reviews by Schnepf, 1969b; Sievers. 1973; Fineran, 1973; Whaley, 1975; Mollenhauer and Morre. 1980) so that it is necessary here to dwell on only those aspects which are of pertinence to membrane flow.

92

DAVID C. ROBINSON A N D U D O KRISTEN

B . POLARITY A dictyosome can be considered as a structure constantly persisting in a stage between dissipation and restoration, since i t s cisternal stack is permanently transfluxed in a defined direction by a stream of different macromolecules. This stream results in a structural and biochemical polarity of the dictyosornal stnck in that cisternae at one pole (forming face) differ from those at the opposite pole (maturing face) (Figs. la and b and 2 ) . It must, however, be mentioned that the various indicators of polarity which are dealt with below are not always, in their entirety, present. Sometimes dictyosomes may be seen which show only mme, but not all, of these indicators.

I . Structural Structural indicators of polarity are the intercisternal spacing between adjacent cisternae, the thickness and stainability of the cisternal membrane. the freezefracture appearance of the cisternal membrane, and the size and content of attached secretory vesicles. The intercisternal spacing sequentially increases from the forming to the rnaturing face while the width of the cisternae decreases in the same direction (Amelunxen and Gronau, 1966; Mollenhauer and Morrk, 1975, 1978, 1980; Kristen, 1978; Ryser, 1979). Mollenhauer and Morre (1978) have determined the mean spacing between GA cisternae of different higher plant tissues in the range of 6.6-9.7 nm at the forming face and 9.7-18.0 nm at the maturing face. This increase of internal spacing seems to be correlated with the presence of intercisternal elements. These, probably filamentous, structures, which occupy a1 part

FIG. I . Median sections through dictyosomes from suspension cultured Acrrpseirdop/uturlrrs ( a ) and from the mucilage-secreting cells of the ovary of Apreniu cordfolia (b). Forming and secr:tory faces are designated FF and SF, respectively. ( a ) X76,000; (b) XSS.000; bar = 0.2 pm. (Micrographs courtesy of D. C. Robinson and U. Kristen.)

MEMBRANE FLOW VIA THE GOLGI APPARATUS GA

93

PM

I

0 On

-4

Intercisternal Width Cisternal Width Normal Staining

-

--

P l A l C A Staining Pol yracch Staining NDPase Activity -Osmium

impregnation

FIG.2. Diagrammatic representation of the various parameters contributing to polarity in the higher plant dictyosome. Gradients are marked accordingly as arrowheads.

of the intercisternal space between the flattened regions of adjacent cisternae, additionally indicate the structural polarity of the dictyosomes, since they predominantly occur in the maturing moiety of the dictyosomes (Mollenhauer, 1965; Turner and Whaley, 1965; Mollenhauer and Morrt?, 1966a, 1972, 1975, 1978, 1980; Cunningham et al., 1966; Mollenhauer et al., 1973; Kristen, 1978). There are some observations that may indicate a functional relationship between the intercisternal elements and vesicle production at the maturing face. Usually these elements are associated only with the extremely flattened central regions of the cisternae. Therefore, Kristen (1978) has postulated that they effect the compression of these regions in order to restrict vesicle formation to the marginal zone of GA cisternae. Thus the secretional material should remain concentrated within the vesicles. In outer root cap cells of maize, however, the intercisternal fibers are mainly associated with elongated secretion vesicles during their formative stages. Obviously in this case the intercisternal elements do not appear to be directly involved in holding cisternae together (Mollenhauer et al., 1973). Instead, Mollenhauer and Morrt (1975) have suggested that these fibers may aid in organizing this special type of vesicle. From the GA in onion stem and soybean hypocotyl cells data have been presented indicating that membrane thickness sequentially increases from the forming to the maturing face in the range from 5.3 to 8.8 nm (Morrt, 1977). These values lie between those for the ER (5.3 nm) and the plasma membrane

94

DAVID G . ROBlNSON A N D UDO KRlSTEN

(9.3nm). The progressive change in membrane thickness from the ER through the cisternal stack of a dictyosome to the plasma membrane is accompanied by an increase in the staining intensity of the membranes when using glutaraldehyde-osmium tetroxide fixation. As another indicator of membrane differentiation within the dictyosomcs the number and distribution of membrane-integrated particles revealed after fi-eezefracturing has been used. Vian (1974) has shown a progressive increase in number of such particles from the ER across the G A to the plasma membrane in root tip cells of Pisum sarivum. Staehelin and Kiermayer (1970) who have proposed a correlation between the density of particle packing and membrane staining intensity and thickness, also demonstrated an increase of particle numbers from the forming to the maturing face of dictyosomes in the alga, Micrasterias denticdata. Furthermore Volkmann ( 1981 ) has demonstrated a decrease in particle number from the center to the margin of the cisternae for dictyosomes from root cap cells of Lepidium sarivun). In many slime-secreting higher plant glands a clear polarity of the dictyosomes is given by the formation of the secretory GA vesicles. These vesicles preferentially bud off at two or three cisternae of only one face, which, therefore, definitely can be determined as the maturing face (Schnepf, 1968; Schnepf and Busch, 1976; Kristen, 1974, 1976, 1978; Dexheimer, 1976, 1978, 1981). In some instances the whole outermost cisternae at the maturing face swell to become a large G A vesicle (Schnepf and Busch, 1976; Kristen, 1978). Dictyosomes in tendril cells of Cucurbira maxinza have also been shown to prclduce vesicles preferentially at one face during secretion of cell wall material (Amelunxen et a / ., 1976). In the case of maize root cap cell dictyosomes a progression in size and stainability of content for the hypertrophied GA vesicles has been claimed (Mollenhauer and Whaley, 1963; Mollenhauer and Morre, 1980). Unfortunately, as Robinson ( 1980b) has shown, the three-dimensional arrangement of GA vesicles is such that a true estimate of vesicle size from single sectialns is not possible and that an oblique section can give rise to the illusion that a progression in vesicle size across the dictyosome stack exists.

2. Chemical a. In Situ. A variety of cytochemical tests indicate the polar nature of the dictyosome. On the one hand impregnation with osmium leads to a staining of the ER and the forming but not the maturing face (Dauwalder and Whaley, 1973; Poux, 1973; Dexheimer, 1981); on the other hand treatment with phosphotungstate chromate stains preferentially the PM, G A vesicles and mature cisternae of dictyosomes (Roland, 1969, 1973; Roland and Sandoz, 1969; Roland and C’ian, 1971). Staining for polysaccharides (Thiery’s test-Rougier, 197 1; Dexheiimer, 1981) and for nucleoside diphosphatase activity (Dauwalder et a / ., 1969; Zaar and Schnepf, 1969; Dexheimer 1978, 1981) both show a localization of the

MEMBRANE FLOW VIA THE GOLGI APPARATUS

95

reaction in GA vesicles and maturing face of the dictyosome. A progression in staining intensity from forming to maturing face can also be seen. b. In Vitro. Elements of the GA can be recognized in density gradients of cell homogenates with the help of two enzymes: inosine diphosphatase (1DPase) and glucan synthetase I (GSI) (Quail, 1979). Although a multisite localization of IDPase activity has been known for some time for some animal cells (WattiauxDe Comninck and Wattiaux, 1969; Tulkens et a / . , 1974), only recently have there been reports that this might also be so for plant cells. Thus for maize coleoptiles (Hendriks, 1978; M'Voula-Tsieri et a / ., 1981), maize roots (Robinson et a / . , 1982), and bean hypocotyls (Bowles and Kauss, 1976) considerable activity associated with ER fractions has been demonstrated. Nevertheless for a number of other higher plant tissues, e.g., pea stems (Ray er a / . , 1969), carrot phloem parenchyma (Gardiner and Chrispeels, 19751, onion stems (Powell and Brew, 1974; Morre et d., 1977), and roots (Klohs and Goff, 1980), only one peak of activity exists and this is coincident with that for GSI (Shore and MacLachlan, 1975). The GSI distribution tends, however, to be narrower than that for the IDPase activity. If the fractions containing these activities are layered onto renografin or metrizamide gradients, a separation of these two activities can be obtained through rate zonal centrifugation (Fig. 3). This result is taken as demonstrating that GSI is associated only with the cisternal portion of the dictyosomes, whereas IDPase is present in both GA vesicles (which have approximately the same isopycnic density as dictyosomes, Ray et a/., 1976) and cisternae. In support of this is the fact that a chase-out phenomena is associated with the two IDPase peaks: radiolabel from a [ ''C]sucrose pulse appears first in the coincident GSI/lDPase peak and then in the IDPase peak which does not exhibit GSI activity (Ewers, Robinson, and Taiz, unpublished observations). In general the localization of specific glycosyl transferase activities, including GSI, on maturing or forming face cisternae through gradient analysis, has not been investigated. One might speculate, for example, that cisternae of the forming face contain arabinosyl transferase activity since some cell wall proteins (see Section III,B, I ,b) are coupled to polysaccharide through ara-hypro linkages. Clearly this can only be investigated when it becomes possible to subfractionate a higher plant dictyosomal fraction as has been done by Bergeron et a/. (1978) for the GA of liver cells. C. RELATIONSHIP TO

THE

ER

In the preceding section dictyosomes have been described as polar structures. A prerequisite for the maintenance of this feature is a supply of membrane material which has to be transformed during its passage through the dictyosomal stack. As mentioned in the introduction, it has not yet been satisfactorily shown for higher plant cells from which source this material is derived. Nearly all

96

DAVID G . ROBINSON AND UDO KRISTEN

n

'/.

A

S u c r o s e or R s n o g r o l i n c x p r r a s c d 0 s r q u i v o l r n l -1. 5 u c r o s c

3. Distribution profiles of inosine diphosphatase (IDPase) and glucan synthetase I (GSI) activities from pea stem homogenates after (A) isopycnic centrifugation on linear sucrose gradients and ( B ) rate-zonal (velocity) centrifugation on renografin gradients. (A) Data of D. G . Robinson; (B) data courtesy of M. Ewers and L. Taiz. FIG.

previous review articles concerning the GA and/or membrane flow in plant!; have been very uncritical in this respect giving the impression that plant and animal dictyosomes are similar in terms of their relationship to the ER. Of course, there are ultrastructural indications of membrane transfer from the ER to the GA in higher plant cells, but they are not as numerous and prominent as these frequently demonstrated in protein-secreting animal tissue. A close structural relationship between the ER and the GA after glutaraldehyde-osmium tetroxide fixation has been recorded for example in developing sieve tube cells of some ferns (Evert, 1976; Evert and Eichhorn, 1974; Fisher and Evert, 1979), in pollen tubes of Impatiens (van Went, 1978), in the nectary cells of Ascr'epias curusavica (Schnepf and Christ, 1980), in the ligules of the water fern Isoetes lacustris Kristen, 1980b), and in the secretory trichomes of Psychofria hac-

MEMBRANE FLOW VIA THE GOLGI APPARATUS

97

teriophila (Dexheimer, 1981). Both of the latter tissues exhibit transition vesicles between ER and dictyosomes or even direct membrane connections, and both are probably involved in the formation and secretion of proteins as judged by analyses of their secretional material (Dexheimer, 1981; Kristen eta/., 1982). Similarly in cases where protein is stored rather than secreted, e.g., in seeds or sieve elements, there have been authors (Dieckert and Dieckert, 1972; Chrispeels et a/., 1979; Fisher and Evert, 1979; Bergfeld er al., 1980) who have implicated the GA in the transport from the ER to the storage body. But there are also sufficient examples for an intracellular protein transport which bypasses the GA (Yo0 and Chrispeels, 1980; Weber and Neumann, 1981; Kristen and Biedermann, 1981). The majority of higher plant dictyosomes are involved in cell wall synthesis and as such the amounts of protein secreted are relatively small. In agreement with this fact is the absence of an accumulation of transition vesicles between ER and GA (Robinson, 1980b). Although it is not possible in the electron microscope to visualize the occasional fusion of single vesicles, a forming face consisting of a cistemae with a number of simultaneously fusing transition vesicles in its central region is not a general feature of the higher plant dictyosome. As Robinson (1980b) has emphasized such a feature can be proven only through serial sectioning since single oblique sections, which cut through the tubular periphery of a forming face cistema, can give the illusion of an accumulation and fusion of vesicles. Regions of direct association between ER and GA have been claimed for maize root tips after cold treatment (Mollenhauer et a / . , 1975). These associations were suggested to be more transient in untreated roots. Similar claims for an ER-GA transition for cortical cells of Ricinus, Pisum sutivum. and Phaseolus vulgaris have been made, particularly at those regions where the ER changes from a sheetlike to a tubular form (Morre and Ovtracht, 1977; Mollenhauer and Morre, 1976a). The ER tubules then appear to be connected with the peripheral edges of the dictyosomes. Unfortunately many of the claims for direct ER-GA connections have been made on permanganate-fixed material where the question of artifact cannot be ruled out. Thus in a reinvestigation of the cold-effect on maize root tips (Robinson, 1981a; Robinson e t a / . . 1982) direct ER-GA connections could not be confirmed. One should also not forget the fact that in geotropism experiments on roots (Sievers and Volkmann, 1977; Volkmann and Sievers, 1979) a stratification of organelles occurs in statocytes whereby the ER and dictyosomes lie in different positions of the cell. If there really was a regular connection between these two elements of the endomembrane system it should also be maintained in such experiments. Thus whereas tubular ER-GA connections are quite common in animal cells, e.g., in liver (Morre et a/., 1971b; Ovtracht et al., 1970). their existence in higher plant cells must be described as, at best. ephemeral.

98

DAVID G . ROBINSON AND UDO KRISTEN

D. RELATIONSHIP TO THE PM The fusion with, and incorporation into, the PM of GA-derived secretory vesicles is the culminating feature of granulocrine secretion as defined by Schnepf (1969a). This step is implicated as an obvious consequence in numerous EM studies on cells where the GA is described as hypersecretory. Such is the case for slime and mucilage secretion in both root cap (see Rougier, 1982, for review) and a variety of gland cells (e.g., Mollenhauer, 1967; Schnepf, 1968; Kristen, 1974, 1976; Schnepf and Busch, 1976; Trachtenberg and Fahn, 1981). It is also inherent in studies on gland cells of carnivorous plants (Schnepf, 1961a; Dexheimer, 1972) and in some cases of nectar secretion (e.g., Benner and Schnepf, 1975; Fahn and Benouaiche, 1979). With respect to cell wall production a GA-PM vesicle transfer is suggested in many studies on cell plate formation (Whaley and Mollenhauer, 1963: Whaley et al., 1966; Frey-Wyssling et al., 1964; Mollenhauer and Mollenhauer, 1978) and on tip growth in pollen tubes or root hairs (Sievers, 1963, 1964; Sassen, 1964; Rosen et al., 1964; Morri and Van der Woude, 1974). Further electron microscopic evidence for a GA-PM relationship comes from studies involving coated vesicles (see Newcomb, 1980, for review). Because of their clathrin coat they can be easily recognized both in thin section and by negative staining. Thus in studies on protoplasts (Doohan and Palevitz, 1980; Van der Valk and Fowke, 1981) their incorporation in the membranes of the PM and of maturing GA cisternae can be clearly seen (Fig. 4). A transfer GA-PM is suggested in these studies and is supported by earlier observations on cell ]plate formation (Hepler and Newcomb, 1967; Franke and Herth, 1974). But since examples for coated vesicle mediated endocytosis (see review by Pearse and Bretscher, 1981) and exocytosis (e.g., Franke et al., 1976; Rothman et ai., 1980) in animal cells are well documented one must be careful about vectorial interpretations in electron microscopic studies from plant cells. More conclusive in terms of the vectorial question are pulse-chase and inhibitor studies. In the former the loss of radiolabel from the GA and its arrival a t the cell exterior as a result of chase-out is usually taken as demonstrating a vesicle transfer from GA to PM. Such a transfer has been demonstrated using both autoradiographic (Juniper and Roberts, 1966; Northcote and Pickett-Heaps, 1966; Paul1 and Jones, 1975) and fractionation techniques (Bowles and Northcote, 1974; Robinson and Cummins, 1976; Robinson er ul., 1976; Kawasaki1 and Sato, 1979; Kawasaki, 1981). The action of inhibitors may be conveniently divided up into those affecting secretory vesicle release from the GA, those interfering with vesicle transport, and those inhibiting fusion with the PM. That vesicle release from the GA in animal cells is energy-dependent is well known from the classic studies of Palade and co-workers (Jamieson and Palade, 1968). Their technique of inhibitor ap-

MEMBRANE FLOW VIA THE GOLGI APPARATUS

99

FIG.4. Negatively stained (with uranyl acetate) plasma membrane from protoplasts isolated from suspension cultured tobacco cells. Numerous coated vesicles (cv) as well as microtubules (mt) are visible. (a) X23.800; (b) X 161,000; bar = 0.5 p,m (a) and 0.05 pm(b). (Figures courtesy of L. C. Fowke.)

plication during the chase-out in pulse-chase studies has been used by Robinson and Ray (1977) and Robinson (1980a) on pea stem segments with similar results: secretory molecules synthesized prior to inhibitor application accumulate in the GA, attached vesicles swell and cistemae elongate and often curve (see also Section V1). Induction and reversibility of these effects are relatively fast, a matter of 3WO minutes. Both microtubules and microfilaments have been suggested at various times (see Chrispeels, 1976) as providing a framework for the directional control over GA+PM vesicle migration. Whereas the production of a cell plate through the fusion of GA-derived vesicles can be blocked by treatment with tubulin-binding substances, e.g., colchicine, these agents do not seem to be inhibitory toward exocytosis in higher plants where secretion is nonpolar. As far as the

100

DAVID G . ROBINSON AND UDO KRISTEN

cytochalasins are concerned, whose mode of action as a microfilament inhibitor has now been confirmed (Brown and Spudich, 1981), there are conflicting reports as to their physiological effects on exocytosis in higher plant cells (compare Chrispeels, 1972 with Robinson et al., 1976). Despite this discrepancy there are now a number of papers in which the cytological effects of these substances are clearly depicted (Mollenhauer and Morrk, 1976; Pope et al., 1979; Picton and Steer, 1981; Volkmann and Czaja, 1981; Kristen and Lockhausen, 1982). Common to all of these reports is the accumulation of vesicles in the immediate vicinity of the GA (see Fig. 5). These effects are reversible by washing out the drug. In contrast to other workers Volkmann and Czaja (1981) claim for Lepidiurn root cells that prolonged treatment with cytochalasin B leads to a destruction of the GA. Presumably in this case cytochalasin B has an indirect effect on energy metabolism (see Section V1,B). Inhibitor studies are now also beginning to shed light on the ultimate act of exocytosis, namely, the fusion of GA vesicles with the PM. From these it is now clear that, as is the case with animal cells (Douglas, 1974; Williams, !l980), Ca2+ ions play an important role in this event. Addition of Ca2+ to the growth medium of suspension cultivated cells leads to an increase in secretion as measured by release into the medium of cell wall polysaccharides (Morris and Northcote, 1977) or enzymes (Sticher er al., 1981). The Ca2+ selective ionophore A 23187 stops pollen tube growth and seems to cause an increase in the number of GA-derived vesicles which fuse with the PM (Reiss and Herth, 1979). The vesicle content does not become incorporated into the body of the wall but appears to collect in clumps in the periplasmic space. A similar effect has been observed with pea stem segments treated with this ionophore in the presence of Ca2+ (Griffing and Ray, 1982). These latter authors have furthermore shown that the increase in cytosolic Ca2+ levels, as a result of such a treatment, leads to a premature vesicle fusion with the PM. In contrast when Ca2+ levels are reduced by adding the nonpermeating chelator EGTA together with A 23 187 vesicle fusion is inhibited and results in an accumulation of GA vesicles in the cytoplasm (see also Robinson, 1981b). Confirmation of the importance of Ca2+ for GA vesicle-PM fusion has been given by the in vitro fusion experiments of Baydoun and Northcote (1980). Transfer of radioactivity from GA fractions to PM fractions from maize root tips is Ca2+-dependent but is unaffected by microfilament or microtubule inhibitors. The role of monovalent ions in exocytosis in pea stem cells has also been investigated by Griffing (1981). To do this the K + selective protonophore nigericin was used. At high concentrations (10 fl)oxidative phosphorylation can be uncoupled and effects on secretion and GA structure similar to those reported by Robinson and Ray (1977) and Robinson (1981b) are obtained. At I fl)nigericin does not affect GA structure but does lower concentrations (I lead to an accumulation of GA vesicles within the cytoplasm. By suitable experi-

MEMBRANE FLOW VIA THE GOLGI APPARATUS

101

FIG. 5 . Accumulation of secretory vesicles in the vicinity of the GA in ovary gland cells of Apfenia cordifolio as a result of cytochalasin B treatment. X9OOO. Bar = 2 km. (Micrograph courtesy of J. Lockhausen and U. Kristen.)

102

DAVID G. ROBINSON AND UDO KRISTEN

mentation involving alteration of the external pH, Griffing has concluded that nigericin probably acts by altering the intravesicular pH. Whereas some of the factors which control exocytosis in higher plant cells are now gradually being recognized, the molecular mechanisms involved in the fusion process have hardly been worked on. Although it is generally supposed from studies on animal cells that exclusively lipid bilayers free of protein constituents are involved in membrane fusion (Orci et al., 1977; Lemay et al., 1977), and that they have a similar lipid composition to one another when they fuse (Ahkong et al., 1975), the only experiments carried out on a higher plant system (Baydoun and Northcote, 1980) indicate in contrast the importance of membrane proteins in this process. Irrespective of the physiological characteristics of GA vesicle release and transport and fusion with the PM, an important question for botanists is whether the enzyme(s) responsible for the synthesis of cellulose are transported along this route. In addition to the GA-localized GSI there is a second glucan synthetase (GSII), measured at high (d) substrate concentrations (Van der Woude et al., 1974) and which is localized at the plasma membrane. Given that the GSII measured in vitro really does represent the cellulose synthetase measured in vivo (see Robinson and Quader, 1981, for a review) it has been proposed (Shore and MacLachlan, 1975) that GA vesicles do carry cellulose synthetase in a zymogenic form. The recent freeze fracture observations of Giddings ef al. (1980) on the green alga Micrasterias provide visual support for this suggestion. Rosettes of particles are characteristic of the PM in this organism and are clearly involved in microfibril synthesis. Similar rosettes are found in the membrane of secretory vesicles. It is possible that this might be true for higher plant cells since similar rosette complexes on the PM have also been recorded (Mueller and Brown, 1980) for maize root cells.

111. Sites of Synthesis

A. SECRETIONAL MATERIAL 1. Polysacchuride

From the cytological viewpoint there are two questions dealing with the localization of the enzymes responsible for the production of nucleoside diphosphate sugar pools and the localization of the glycosyl transferases. As regards the former the general opinion is that the enzymes in question are located in the cytosol (Neufeld et al., 1957; Dalessandro and Northcote 1977a,b). Nevertheless there are reports that, for example, some epimerases are membrane-bound (Feingold et al., 1960; Panayotatos and Villemez, 1973) although a differentiation between ER and GA is not given. There can be no question that the GA is the principal site for glycosyl trans-

MEMBRANE FLOW VIA THE GOLGI APPARATUS

103

ferase reactions in the cell (Lennarz, 1978; Franz and Haap, 1980). Although the products have not always been adequately defined the production of lipid-like compounds (steryl glycosides or polyprenol mono/diphosphate glycosides), presumably as intermediates, has been frequently documented. Thus arabinosyl (Gardiner and Chrispeels, 1975), fucosyl (Green and Northcote, 1979a; James and Jones, 1979), galactosyl (Ray, 1980), glucosyl (Bowles et al., 1977; Helsper et al., 1977; Lercher and Wojciechowski, 1976; Ray et al., 1969; Van der Woude et al., 19741, mannosyl (Lehle et al., 1978), and xylosyl (Ray, 1980) transferase activities have been demonstrated in vitro for GA-rich fractions isolated from a variety of higher plants actively involved in cell wall deposition. In vivo incorporation studies with radioactive sugars (Bowles and Northcote, 1976; Harris and Northcote 1971; Jilka et al., 1972; Paul1 and Jones, 1976; Ray et a l . , 1976) support these results in so far as labeled polysaccharides, as judged by solubility properties, gel filtration, and hydrolysis, are to be found in GA-rich fractions. More important for the subject of membrane flow in higher plant cells is the role of the ER in polysaccharide biosynthesis. In animal systems where glycoprotein biosynthesis has been extensively investigated it is clear that the formation of the core or G-oligosaccharide portion is a property of the ER (Czichi and Lennarz 1977; Lodish and Rothman, 1978) and that phosphorylated polyprenols act as lipid intermediates in the glycosylating reactions (Sturgess et al., 1978). Characteristic of the core oligosaccharides are mannose and N-acetylglucosamine (Sharon and Lis, I98 1). Lipid-linked transferase activities for both of these sugars have been localized in ER-rich fractions from pea cotyledons (Nagahashi and Beevers, 1978) and castor bean endosperm (Marriott and Tanner, 1979; Mellor et al., 1980). Although dictyosomes have been demonstrated in germinating pea cotyledons (Mollenhauer and Totten, 1971) and in castor bean endosperm at later stages of germination (Mollenhauer and Totten, 1970) their quantitative presence at the time of harvesting for the in v i m tests is unclear. Furthermore these ER-bound activities relate more to intracellular protein segregation (pea, reserve protein synthesis; castor bean, glyoxysomal enzyme synthesis) than to granulocrine secretion. Nevertheless ER-located polyprenyl phosphate-dependent mannosyl and glucosyl transfer reactions have been demonstrated in higher plant tissues where the involvement of dictyosomes in cell wall polymer secretion is undisputed (Diirr et a l . , 1979; Lehle et al., 1978; Hopp et a l . , 1979). In addition, in the case of the maize root cap, where the slime polysaccharide appears to be assembled as a glycoprotein (Green and Northcote, 19781, a lipid-dependent fucosyl transferase has also been demonstrated for ER fractions (Green and Northcote, 1979a). Although in animal cells a fucosyl transferase of this type has not been demonstrated (Molnar, 1976), one which transfers directly and terminally to an oligosaccharide is characteristic of maize root dictyosome fractions. A similar

I04

DAVID 0.ROBINSON AND LIDO KRISTEN

fucosyl transferase, though nonterminal in action, has also been demonstrated for maize root dictyosomes by Green and Northcote (1979b). prompting these authors to postulate the initial synthesis of slime polysaccharides in the ER. and their elaboration in the GA after transfer to this organelle. This is backed up, in the case of maize root tips, by analysis of the polymers synthesized after in vivo labeling with [ 14C]glucose(Bowles and Northcote, 1976): ER fraction polymers had an average MW of less than 4000 compared to the dictyosome fraction with values exceeding 40,000. In contrast, for pea stem tissue, where slime is not secreted, no indication for the participation of the ER in polysaccharide synthesis from in vivo labeling experiments could be obtained (Ray et al., 1976; Robinson et al., 1976).

2 . Protein Proteins may be synthesized on bound or free polysomes and their insertion into a membranous organelle can be described as co- or posttranslational, respectively (for a recent review see Waksman et a f . , 1980). Although the mechanism of insertion and passage through a membrane is undecided (Engelman and Steitz, 1981) the best examples for posttranslation are precursor forms of enzymes for chloroplasts and mitochondria which are synthesized outside of these organelles. Secretory proteins, on the other hand, are generally considered to be synthesized by ER-bound polysomes (Blobel et al., 1979). Several cell wall proteins of higher plants are characterized by appreciable amounts of the imino acid hydroxyproline (Lamport, 1980). Hydroxyproline is also characteristic of collagen and the synthesis and secretion of this latter molecule is well understood (Fessler and Fessler, 1978). Since hydroxyproline is nontranslateable, labeled proline is applied in vivo to tissues. In the case of collagen, the sites of incorporation and posttranslational modifications such as hydroxylation and glycosylation have been clearly demonstrated (Harwood et al., 1974; Olsen et a l . , 1975; Peterkovsky and Assad, 1976) and confirm an ER+GA transport. For plants the situation has, until very recently, not been at all clear. Although it has been adequately demonstrated that the protein is linked to polysaccharide through arabinose (Lamport and Miller, 197I ) and galactose (Lamport et al., 1973) residues and that the GA is involved in glycosylation (Gardiner and Chrispeels, 1975; Kawasaki and Sato, 1979), there are conflicting reports (Sadava and Chrispeels, 1971; Tanaka et a f . , 1980) as to the location of the enzyme prolyl hydroxylase. Thus Chrispeels (1976) has been very cautious in review articles on plant ER function about the question of hydroxyproline-rich proteins and has even gone so far as to omit all reference to such proteins in recent reviews (Chrispeels, 1977, 1980). That the ER is indeed involved in cell wall protein synthesis has recently been indicated in the laboratory of one of us (Wienecke et al., 1982). Hydrol,ysis of ER polysomes, polysome-free ER vesicle, and GA fractions from carrot tissues

MEMBRANE FLOW VIA THE GOLGI APPARATUS

105

has clearly shown the participation of the ER in the synthesis of hydroxyprolinecontaining glycoproteins (see Table I). Because newly synthesized proline-containing polypeptides are continually being inserted into the ER lumen the hydroxylation values for the ER in this table are lower than those for the GA. In contrast to the results of Chrispeels (1969) and Gardiner and Chrispeels (1975) who have claimed that proline given to aged carrot discs is at least 70% hydroxylated, a maximum degree of hydroxylation under 30% was obtained. This agrees with the more recent values given by Varner and Burton (1980) who have demonstrated that the degree of proline hydroxylation is dependent upon the presence of carrier proline and the method of tissue aging. Final confirmation of the ER as the intracellular site of synthesis of this cell wall protein awaits a reinvestigation of the problem of the localization of the key enzyme prolyl hydroxylase. Enzymes, for example, invertase and peroxidase, are also present in the cell wall. Whether they are also synthesized on the ER and exported via the GA is at the moment not known. B. ENDOMEMBRANES

1. Membrane Proteins Membrane proteins may be synthesized on either free or ER-bound polysomes (see Kreibich et a f . , 1980; Lodish et al., 1981, for recent reviews.) According to TABLE I DtCRtt OF PROLINE HYDROXYLATION IN FRACTIONS ISOLATED FROM CARROT DISCS INCUBATED FOR I HOUR WITH ['4c]PROLINE

Fraction Without carriefl Cytosol'' ER' GAc With carrier" cyt0s0I ER1 G A'

Percentage hydroxylation" in total nonfiltratable protein

4.I 11.0

19. I

11.1 17.9 28.0

' I Determined by hydrolysis and chromatographic separation of hydroxyproline and proline. Carrier = 50 pmole [lT]proline per 3 g carrot discs. ' Cytosol = nonrnembranous portion in 12-278 fractions, ER = membrane pellet in 12-27% fractions, GA = membrane pellet in 28-37% fractions from a linear sucrose isopycnic density gradient prepared with low (0.I mM) Mg2+ concentrations.

106

DAVID G . ROBINSON AND UDO KRISTEN

Lodish and Small (1975) and Shore and Tata (1977) ectoproteins and transmembrane proteins are principally hydrophobic and are synthesized on ERbound polysomes. Endoproteins, on the other hand, are considered as being synthesized on free polysomes and inserted into the cytoplasmic face of the membrane. As Elder and Morr6 (1976) have shown for rat liver certain GA proteins are synthesized on polysomes which are not ER bound but, instead, are loosely associated with the GA. Nevertheless the majority of GA membrane proteins are most likely to be of ER origin, particularly when one considers GA biogenesis and not just GA maintenance in cells engaged in granulocrine secretion. Having said this one must hasten to add that, as far as plants are concerned, with the exception of the observation that GA-associated polysomes are occasionally to be seen (Franke et al., 1972; Mollenhauer and Morr6, 1974), there is no information available on the synthesis of ER or GA membrane proteins. One can only speculate on the basis of experiments carried out on animal systems. 2. Membrane Lipids In contrast to the situation with the protein component of endomembranes a considerable amount of information is available concerning membrane lipid synthesis in plants. We can conveniently divide the discussion on membrane: lipid synthesis into four areas: (1) fatty acid (FA) biosynthesis, (2) phospholipid (PL) biosynthesis, (3) sterol biosynthesis, and (4) phospholipid exchange mechanisms. a. FA Biosynthesis. Although in animal cells mitochondria possess elongation enzymes and elements of the mitochondria1 fraction have desaturases, the major steps in FA synthesis, carried out by the FA synthetase complex, are localized in the cytoplasm (Vagelos, 1974). In contrast to this, recent research on plants has clearly demonstrated that FA biosynthesis is a characteristic of plastids and proplastids (Vick and Beevers, 1978; Ohlrogge et al., 1979; Roughan al., 1980). Free fatty acids, usually palmitic and oleic, accumulate and are transported out of the plastid with the help of acyl carrier protein. b. PL Biosynthesis. In common with endomembranes from animal tissues the ER, GA, and PM from plants possess large amounts of phosphatidylcholine (PC), phosphatidylethanolamine (PE), and phosphatidylinositol (PI) (Hitchcock and Nichols, 1971; McMurray and Magee, 1972). In contrast to animal cells sphingomyelin is absent from plant endomembranes (Morr15, 1977). The starting point for the synthesis of phospholipids is essentially phosphatidic acid. Fatty acids are converted to to acyl-CoA thioesters, acyl transferases then convert them into phosphatidic acid. The enzymes involved in phosphatidic acid synthesis are localized in microsomal fractions in both animals (Hendry and Possmayer, 1974) and plants (Vick and Beevers, 1977). In addition mitochondria may also possess these enzymes (Monroy et al., 1973). Two routes elf

MEMBRANE FLOW VIA THE GOLGI APPARATUS

107

for the synthesis of PC have been established from early work on animal systems. These are the nucleotide pathway (Wilgram and Kennedy, 1963) and the stepwise methylation of PE (Wilson et al., 1960). Both have now been shown to be operative in plants (Morre et al., 1970; MorrC, 1975), with the former pathway predominating. With the exception of the phosphorylation of choline by choline kinase all the reactions are membrane bound (see Morrk, 1975, for review). The ER has been clearly identified (Hoch and Hartmann, 1981), particularly in the case of the germinating castor bean (Beevers, 1975), but in systems where the GA is more apparent and well separated from other cell components, the enzymes are to be found here too (Morrk et al., 1970; Montague and Ray, 1977). Indeed on a specific activity basis the GA is more important than the ER in these systems. Additional enzymes involved in the synthesis of PI, PE, and PS have also been demonstrated for ER fractions from castor bean (Moore et al., 1973), and in the case of PE for Golgi fractions as well (Montague and Ray, 1977). These two papers, however, show that some other endomembrane components in addition to the ER and GA may be involved in phospholipid (PC) biosynthesis. c. Sterol Biosynthesis. The enzymatic steps leading to the synthesis of squalene are soluble (Staby etal., 1973). The conversion of squalene to choles-

%SUCROSE

FIG. 6. Distribution profiles of lipid radioactivity after in virro transfer experiments with in vivo labeled (from methyl [14C]cholinechloride) “ER” fractions (ER.) and unlabeled “GA” fractions from pea stem segments. Fractions from isopycnic gradients were isolated and incubated as indicated at 37°C for I hour in the presence (A) or absence (B) of the cytosol. The transfer was stopped through addition of glutaraldehyde (end concentration 0.5%) and the mixture recentrifuged isopycnically. Radioactivity in fractions having a density higher than 29% sucrose ( I . 125 g cm - 9 represents lipid transferred from ER to GA. (Unpublished results of M. Eberle and D. G. Robinson.)

DAVID G.ROBINSON AND UDO KRISTEN

108

terol involves many steps all of which are localized on microsomal membranes. This is so for plants (Hartmann ef al., 1973; Hartmann-Bouillon et al., 1979) as well as for animals (Scallen et al., 1974). d. Phospholipid Exchange Mechanisms. Mitochondria synthesize negligible amounts of PC, PE, and PI (Mom5, 1975). These are supplied by the endomembranes via cytoplasmic proteins known as phospholipid exchang: proteins (PLEPs). They were first demonstrated in and purified from animal cells (Wirtz and Silversmit, 1969; Wirtz etal., 1972), but have now been detected and isolated from plant tissues (Kader, 1977; Tanaka and Yamada, 1979). The exchange is bidirectional and can easily be demonstrated in vitro. However, it has hitherto been restricted to mixtures of microsomal fractions with either mitochondria, microbodies (Boussange et al., 1980), or plastids (Julienne et al., 1980). From the point of view of ER-GA relationships it would be most inleresting to see if phospholipid exchanges between components of a microsomal fraction can be demonstrated. Experiments along this line have been started in the laboratory of one of us and appear successful (see Fig. 6), but their interpretation is made difficult because of the contamination of GA fractions with other non-ER microsomal components, e.g., plasma membrane.

IV. Secretion Kinetics and Membrane Turnover

In discussing the velocity of granulocrine secretion one must be careful to distinguish between the various types of data available. Essentially there are three types: (1) measurements based on the use of content markers; (2) measurements based on membrane markers; and (3) calculations derived from a morphometric analysis of thin sections. Only if a bulk-flow of membrane accompanies the transport of secretory protein or polysaccharide can the values from (1) and (2) approximate one another. In animal cells where appropriate measurements are available (Franke et al., 1971; Meldolesi, 1974a; Castle et a l . , 1975; Wallach et al., 1975) this is clearly not the case. As Franke and Kartenbeck (1976) have stated: ‘‘When the kinetics of amino acids incorporated into membrane proteins are compared with those of the secretory products contained in the specific membranes, one notes that the major part of the radioactivity contained in membrane protein does not parallel the kinetics of typical secretory proteins but rather reveals a much lower turnover. Only a relatively small portion of the membrane protein of rER, GA and NE seems to parallel the kinetics of the vectorial flow of the secretory product. Whereas the turnover of membrane proteins has been extensively investigated in animal systems (see Hubbard, 1978, for a short review), there is to our knowledge no corresponding literature for plant endomembranes. With respect to the lipids of plant endomembranes the only pertinent study to date is ”

MEMBRANE FLOW VIA THE GOLGI APPARATUS

109

that of M o d (1970) which does in fact indicate that turnover is very slow (in the order of days). Clearly a selective transfer of membrane must instead occur and this is inherent in the studies of Rothman, Morrk, and others (Lodish and Rothman, 1979; Bergmann et al., 1981; Croze and Morre, 1981; Fries and Rothman, 1981; Morre, 1981; Rothman, 1981) on vesicular stomatitis virus G-protein and H-2 alloantigens. The protein portion of these is synthesized by ER-bound polysomes and the molecules remain anchored in the membrane during the various addition and removal reactions which occur in the ER and GA before their arrival at the cell surface. These are the best known membrane markers and measurements of their secretion kinetics give values of 10-15 and 40-60 minutes for the ER+GA and GA+PM steps, respectively. Studies with membrane markers have not yet been undertaken with higher plant cells but there are some studies (see Table 11) in which the kinetics of secretion based on a content marker have been estimated. In those studies in which sugar precursors have been used two sources of error are involved: (1) the incorrect recognition and differentiation of secretory (GA) vesicles from ER vesicles in gradient preparations (see Robinson, 1977), and (2) the possibility that incorporation occurs directly into secretory (GA) vesicles or into GA vesicles which become detached from the maturing face cisternae upon homogenization (see M o d et al., 1979). Despite these possible pitfalls displacement times of 2 4 minutes for both GA and secretory (GA) vesicles compartments have been calculated. In one case (pea stem segments: Robinson et al., 1976) incorporation into the ER was so low that secretion kinetics for compartments prior to the GA could not be given; in the other case (maize root tips) a sufficient amount of lowmolecular-weight polymeric material is synthesized in ER fractions (Bowles and Northcote, 1976) and a displacement time of 7 minutes for this compartment is given. Table I1 also includes a few examples of displacement times for animal systems. Whereas a value somewhat less than 10 minutes for the ER appears to be common for both animal and plant systems, displacement times appear to be much shorter for the GA in plants as against animals. The same is even apparent for the secretory (GA) vesicle stage. Whether these differences are true ones or artifactual (see above) is not clear. Particularly for plant cells, where the dictyosome is much more delineated from other membranes, there are a number of estimations of the turnover of membrane in the GA (see Table 111). Inherent in these calculations which are based on the morphometry of thin sections, is that a bulk flow of membrane through the GA occurs, i.e., cisternae are shunted through the dictyosome from the forming to the maturing face in order to replenish membrane lost from the latter. One must be careful, however, not to oversimplify matters. There are indeed cases (e.g., scale-producing algae and some root cap cells) where the

TABLE I1 KINETICS OF GRANULOCRINE SECRETION BASEDON CONTENT MARKERS Displacement time (min)

References

I c

0

A. Higher plants Chrispeels (1970); Gardiner and Chrispeels (1975) Bowles and Northcote (1 974) Robinson et al. (1976)

B. Animals (a selection) Jamieson and Palade (1967) Neutra and Leblond ( I 966) Bergeron et al. (1978) Kimber (1981)

Material

Method

Isotope

For ER

For GA

For secretory (GA) vesicles

Carrot root discs

Fractionation

[ 14CIProline

4

4

6 8

Maize mot tips Pea stem segments

Fractionation Fractionation

[ 14C]Glucose [ 14C]Glucose

7

-

2-3 3

2-3 3 4

Guinea pig: pancreas exocrine Rat: goblet cells

Autoradiography

[ I4ClLeucine

7

20

w

Autoradiography

['HJGlucose

-

15-35

60

Fractionation Autoradiography

[14C/3H]Leucine r3H]Leucine

2-4 2 4

6-1

3-8

6-7

3-8

Rat: liver Schistocera: egg shell

This is an example of regulated discharge (i.e., secretory vesicle accumulation).

TABLE Ill MEMBRANE DISPLACEMENT I N THE GOLGIAPPARATUS Renewal times (min)

References

-I

Material

Schnepf and Koch (1966) Neutra and Leblond (1968) Brown (1969) Eisinger and Ray (1972) Heinrich (1973) Bowles and Northcote (1974)

Vacuolaria virescens-water secretion Rat colon-mucin secretion Pleurochrysis scherfelii-scale secretion Pisum s a r i v u ~ e l wall l secretion Monardu fistulosa--’ ’water” secretion Zeo mays-slime secretion, cell wall secretion

Williams (1974) Mom6 et al. (1979) Picton and Steer (1981) Kristen and Lockhausen (1982)

Hymenomonas carterae-scale secretion Rat liver-protein secretion Tradescanria virginiana-cell wall secretion Aptenia cordifilia-slime secretion

Estimations not based on morphometry.

For individual cistemae

For entire dictyosomes

I 2-40 1-2 1-20

8-10 Not given 2&30

1

0.3° 2Sa 6 650

3.1 7.3

5-10 8

Not given Not given 3040 15-20 15- 18.5

66

112

DAVID G. ROBINSON AND UDO KRISTEN

F a . 7. Median section through a maize root tip dictyosome. The terminal cistema at the secretory face (SF) is seen leaving the stack. Arrows point to coated vesicles. FF, Forming face. X I10,OOO; bar = 0.I pm. (Micrograph courtesy of D. G. Robinson.)

entire cisternae at the maturing face is released as a vesicle. The majority of higher plant dictyosomes, however, tend to give off vesicles at the peripheral part of a cistema. In such cases the nonvesiculating central portions of the maturest cisterna degenerate or separate from the cisternal stack and are often still seen in the vicinity of the dictyosome before they finally disappear (Fig. 7) (Mollenhauer, 1971). It is not always clear whether this important difference in the events which take place at the maturing face of the dictyosome is taken into consideration in morphometric derivations. Nevertheless, apart from the extremely low values given by Bowles and Northcote (1974) membrane flow through the GA, as indicated by the renewal times for entire dictyosomes, lies for both plants and animals in the same order of magnitude, i.e., between 10 minutes and 1 hour.

V. Means and Ends

In the preceding sections we have attempted to give an unbiased account of the higher plant GA. It should, therefore, now be clear that, although there are similarities to the animal GA there are also significant differences. These: differences we believe are a consequence of the relative amounts of protein to polysaccharide in the secretion product. Since protein destined for export to the

113

MEMBRANE FLOW VIA THE GOLGI APPARATUS

cell exterior appears to be synthesized exclusively on the ER the major difference is reflected in the degree of vesicle transport between the ER and the GA. Two extremes are possible and these are diagrammatically presented in Fig. 8. In Type I, which represents essentially the classical secretion-associated membrane flow in animal cells, protein is the principal secretory component and, as a result, the ER-GA vesicle traffic is intense and the fusion of such vesicles at the forming face of the dictyosome is clearly recognizable. On the other hand, when there is relatively little protein, or better said few peptide linkages, there will be correspondingly few ER+GA transition vesicles. This situation is represented by Type I1 in Fig. 8 . These two types do not exactly follow taxonomic lines. Thus although the majority of higher plant cells which secrete cell wall material or slime, i.e., principally polysaccharide, appoximate Type 11, Type I is not restricted to animal cells. We have already mentioned that there are numerous, clear, examples for ER-XiA relationships in lower plants. Nevertheless we know of no animal cells which might be represented by Type 11. Vesicular protein/glycoprotein ER-GA transport involves consideration of a number of parameters: vesicle size and number, the packing of the secretory molecules in the vesicles, and also the frequency with which they are released and fuse. For animal cells coated vesicles have been shown capable of being the agents of ER-GA transport (Rothman et af., 1980)but for higher plant cells the identification, let alone characterization, of ER transition vesicles is lacking. In ER

TP

DP

SP 8 1

*

PM

_.._._________

FIG. 8. ER-dictyosome relationships expressed in terms of the nature of the secretory product. The relative amounts of protein in the secretory product are indicated as stippled sectors in the “composition circles.” Two extremes are visualized: Type I with a high proportion of protein and therefore a distinct vesicle traffic between ER and GA and Type I1 with mainly a polysaccharide secretion and therefore a much reduced ER-GA vesicle transfer. TP, secretory phase; DP, dictyosoma1 phase; SP. secretory phase.

114

DAVID G. ROBINSON AND UDO KRISTEN

animal cells the pathway for ER-GA transport is short. Because of cytoplasmic streaming this pathway is impossible to estimate for many higher plant cells. Under such circumstances it is a matter of conjecture as to how an ER-released vesicle might reach a dictyosome. One thing is certain, however, their an-ival at the forming face is much more unobtrusive than for animal cells and may well occur at the tubular periphery of the forming face cistema. With the exception that some animal cells store their GA vesicles, the contents of which are released to the cell exterior only in the presence of a secretagogue (Tartakoff and Vassalli, 1978), and whereas plant cells are continous in their mode of secretion, there appears to be no significant difference between the two cell types in terms of the transfer step GA-PM. There are indeed similarities in physiology and kinetics between the two (see Sections II,D and IV). We hiive no reason therefore for differentiating this stage in Fig. 8. If, however, the loss of membrane from the GA in the form of vesicles is the same in both types the vesicle input is not and needs to be counterbalanced in the case of Type 11 otherwise the dictyosome as an organelle will gradually disappear. Nonvesicular modes of membrane replenishment seem an inescapable consequence of this fact. In Section III,B the membrane lipid synthesizing capacity of the higher plant GA has already been mentioned together with the possibility of the participation of phospholipid exchange proteins which might transfer lipid from the ER to the GA. Another possibility which we will consider in the following section is that of the insertion of membrane material which has been recycled from the PM.

VI. Membrane Recycling and the Golgi Apparatus A. NECESSITY The recycling of membrane from the PM is a necessary consequence of granulocrine secretion. The amount of membrane reaching the PM and the amount retrieved depends on two factors: the intensity of secretion and the growth status of the cell. Thus for cells which are nongrowing and whose principal physiological function is secretion, e.g., gland cells, membrane recycling must be more intense than in, for example, tip-growing cells. The amounts of membrane which must be retrieved may be quite spectular. In the case of the pancreatic acinar cell Meldolesi (1974b) has calculated that at least 15 times the amount of membrane in the PM reaches and must be removed from the PM each day. Brown et al. (1970) have estimated that as a result of scale exocytosis in the alga Pleurochrysis an area of membrane equal to the PM itself is delivered there every 1-2 hours (i.e., about 1% per minute). Values of the same order of magnitude have also been given for gland cells of Mimulus tingilii (3% per minute; Schnepf and Busch, 1976), for the alga Vucouluria

MEMBRANE FLOW VIA THE GOLGI APPARATUS

I15

viresrens (5% per minute; Schnepf and Koch, 1966) and for the ovary gland cell of Aptenia cordifolia (9% per minute; Kristen and Lockhausen, 1982). Even more extreme values are to be found in the literature: for example, one can estimate that in the “fluid” secreting hydathodes of Monarda fistulosa (Heinrich, 1973) the PM receives 84% of its surface area in new membrane per minute. Finally Gunning and Steer (1975) using data of Schnepf (1961b) on the insectivorous plant Drosophyflurn have come to the astronomic value of 500% per minute. Clearly since in none of these cases does the cell in question enlarge at the same rate the extra membrane which is continually being added must be recycled back into the cell.

B. EVIDENCE FORANDICATIONS OF

IN

HIGHERPLANTS

There is no direct proof for membrane recycling in higher plant cells. Nevertheless there are some observations which might be regarded as indirectly indicating it. These are all inhibitor experiments and usually result in the production of a cup-shaped dictyosome (compare Fig. 9a and b). Table IV demonstrates how frequent this observation has been over the last 20 years or so. The factor common to all the various treatments, which ellicit this effect, is a direct or indirect action on energy metabolism. This is confirmed in those cases where measurements of ATP have been carried out (Griffing, 1981; Glas and Robinson, 1982). With respect to these inhibitor treatments two points have to be mentioned.

FIG.9. Dictyosomes of maize root cap cells. (a) Control, (b) from mots treated with 0.25 mM cycloheximide for 9 hours at 25°C. The lack of hypertrophied secretory vesicles and the membrane enlargement in the dictyosome are immediately apparent. (a,b) X33.000; bar = 0.25 pm. (Micrographs courtesy of D. G. Robinson.)

116

DAVID G. ROBINSON AND UDO KRISTEN

TABLE I V OVERVIEW OF THE VARIOUS AGENTSUSED FOR THE INDUCTION I N HIGHER PLANTS Agent A23 I87 Antipyrine Amiprophosmethyl Arsenic 6-Azd uracil CO atmosphere Cold temperature

Coffein CHCI3 Cycloheximide Dianemycin DNP EGTA Na-fluorescein Fluphenazine High temperature

High pressure Hz02, H2S KCN

Monensin Nigericin Valinomycin Water deficit X-537A

Plant (cell/organ type) Zea mavs (root cap cells) Allium eepa (root tips) Acer pseudoplatanus (Suspn. cultures) Zea mays (root cap cells) Drosophvllum lusiianicum (gland cells) Allium cepa (root tips) Allium cepa (root tips); Elodea canadensis (leaves) Allium cepa (root tips); Elodea canadensis (leaves) Zea mays (mot cap cells) Zea mavs (root cap cells) Allium cepa (root tips); Elodea canadensis (leaves) Zea mays (root cap cells) Zea mays (root cap cells) Pisum sativum (shoot segment) Zea mays (root cap cells) Lepidium sarivum (root hairs) Zea mays (root cap cells) Allum cepa (root tips); Elodea canadensis (leaves) Drosophyllum lusitanicum (gland cells) Allium cepa (root tips); Elodea canadensis (leaves) Allium cepa (root tips); Elodea canadensis (leaves) Allium cepa (root tips); Elodea canadensis (leaves) Drosophyllum lusitanicum (gland cells) Euphorbia pulcherima Allium cepa (root capcells); Zea mays (root capcells) Pisum sativum (shoot segments) Zea mays (root cap cells) Zea m v s (root cap cells) Pisum sativum (shoot segments) Zen mavs (root cap cells) Drosophyllum lusitanicum (gland cells) Lilium longiflorum (pollen tubes) Zea mays (root cap cells)

OF

CUP-SHAPED DICTYOSOMES

References Robinson (1981b) Deysson and Benbadis (1972) Glas and Robinson (1982) Robinson ( I98 I b) Schnepf ( 1963) Hall and Witkus (1964) Wrischer (1960, 1965) Wrischer (1960, 1965) Mollenhauer et a/. (1975); Robinson (1981a) Robinson ( 198 I b) Wrischer (1960, 1965) Robinson ( I98 I a) Robinson ( I 98 I b) Robinson and Ray (1977) Robinson ( 198I b) Kristen (1972) Robinson ( I98 I b) Wrischer (1960, 1965) Schnepf (1961a) Wrischer (1960. 1965) Wrischer (1960, 1965) Wrischer (1960, 1965) Schnepf (1963) Schnepf ( 1964) Morri and Mollenhauer (1964) Robinson and Ray (1977) Robinson (1981b) Robinson ( I98 I b) Griffing (I98 I ) Robinson (1981b) Schnepf ( I96 I ) Reiss and Herth (1980) Robinson ( 1981b)

MEMBRANE FLOW VIA THE GOLGI APPARATUS

117

First, the GA is not alone in being affected: the ER is also present in increased amounts (Robinson, 198la). Second, the induction of more endomembranes in higher plant cells is also a phenomenon which occurs naturally, e.g., during dormancy breaking of seeds etc. (see Bouvier-Durand et al., 1981, for references). The classic work of Jamieson and Palade (1968, 1971), which has been confirmed and extended for several other animal types (Tartakoff and Vasalli, 1978), has shown that the processes of budding-off and fusion of transport vesicles is energy-dependent. Thus under conditions of energy depletion secretion can be blocked and the effect is reversible. This has also been confirmed biochemically for higher plant cells (Robinson and Ray, 1977). If the only way that the GA were to receive membrane in order to replenish that lost by the release of secretory vesicles was through the fusion of ER vesicles at the forming face such inhibitor experiments should not result in the enlargement and curlingup of the dictyosome. This being so it seems unavoidable that nonvesicular modes of membrane transfer are involved in the maintenance of membrane equilibrium in the GA. Presumably they are relatively independent of the energy status of the cell so that intussuscpetion of membrane material continues when vesicle fusion is no longer possible. Since the release of vesicles at the maturing face is also prevented under these conditions the cisternae will “grow.” Of course it is possible that this membrane material originates from other membranes but it is attractive to think that it comes mainly from the PM and represents recycling. Recycling of membrane subunits from the PM is not the only form of recycling which is conceivable for the GA in higher plant cells. A significant observation in this regard is that of Mollenhauer (1971). As already mentioned (see Section IV), many higher plant dictyosomes are characterized by the degeneration of the central portion of the maturing face after secretory vesicle release at the periphery. The fate of this membrane material, as Mollenahuer has suggested, could well be a reinsertion in nonvesicular form at the forming face. Indeed he has calculated that the amount of membrane material which is lost from the GA in this way approximates that which reaches the PM as secretory vesicles.

c.

POSSIBLE AGENTSOF RECYCLING

Figure 10 is a diagrammatic representation of the various possibilities for recycling in higher plant cells. Essentially there are two forms: recycling in the form of membrane subunits (after Hokin, 1968) or, as originally suggested by Palade (1959) as endocytotic vesicles. Depending on the type of GA (dictyosomes with total cisternal release as a vesicle as against those showing peripheral vesicle release) the pool of membrane subunits depicted as being fed from the

DAVID G . ROBINSON AND UDO KRISTEN

118

A

B

FIG.10. Pathways and possibilities of secretion-associated membrane flow and recycling in higher plant cells. In (A) a dictyosome of the type characteristic for some slime-secreting cells is depicted. Here the entire maturing face cistema is released as a secretory vesicle. In (B), the more usual case, the central portion of the maturing face cistema degenerates and may be directly recycled back into the forming face cisternae either in vesicle form or in “membrane subunit form.” Recycling from the plasma membrane is presented in the form of endocytotic (coated) vesicles as well as through the release of membrane subunits which may constitute membrane subunit pools (MSP).ER and GA can be equally considered recipients for recycled membrane in these forms.

PM will be additionally supplied by the degeneration of the central portion of maturing face cisternae. 1. Membrane Subunits There are several cases in the animal literature whereby cell surface located radiolabel is internalized (e.g., Terris and Steiner, 1975; Carpenter and Cohen, 1976). In these two cases the labeled PM macromolecules appear to be considerably degraded so that a direct reinsertion into endomembranes does not seem possible. Nevertheless unpublished results of Reuter (personal communication) on rat liver indicate that at least one fucose-containing glycoprotein is released from the PM and finds its way through the cytosol back to the GA. As far as higher plants are concerned there is even less information available. One of us has recently tried to obtain evidence for the existence of a pool of membrane subunits in maize root tip cells (Robinson, 1981a; Robinson ez al., 1982). The idea behind the experiments is simple: if, for the induction of cupshaped dictyosomes (see above), cytosolic pools of membrane subunits are drawn upon, it should be possible to demonstrate a reciprocal relationship between protein in membrane form and protein as membrane subunits in the cytosol. As it turned out considerable increases in membrane protein could be measured but the expected decline in cytosolic protein was not observed; in fact

MEMBRANE FLOW VIA THE GOLGI APPARATUS

119

the latter increased as well. Unfortunately that which was designated as the cytosolic protein fraction, namely, the 100,OOO g supernatant, can be subjected to considerable contamination. In particular the protein component of the root cap slime which is retained in the various secretory organelles as a result of the inhibitor may be released during homogenization and can falsify the picture considerably. The need to separate membrane from secretory protein in such experiments is apparent, but perhaps simpler than extensive experimentation in this regard would be to concentrate on the lipid portion of the membrane since lipids, in the majority of the cells under consideration here, are not secreted.

2 . Endocytotic Vesicles Membrane retrieval through pinocytosis has been shown for a number of animal cells. In some cases the recipient organelle is a lysosome or the GERL (Golgi apparatus+mdoplasmic reticulum-lysosome) complex (see Farquhar, 1978; Morri et al., 1979, for summaries) but in the recent studies of Herzog and Reggio (1980) it is the GA itself. In these studies electron-dense substances such as ferritin or blue dextran are added to the cell exterior and these are internalized during pinocytosis. It is clear that such studies on higher plants can only be carried out with protoplasts; but to our knowledge this has not yet been done. Earlier pinocytotic studies on protoplasts have involved larger particles than ferritin, e.g., viruses (Otsuki et al., 1972) and have not been very conclusive with respect to membrane recycling. As has already been mentioned (Section II,D) coated vesicles have been shown to be involved in endocytotic processes in animal cells. After their beautiful demonstration and their large numbers in higher plant protoplast (Doohan and Palevitz, 1980; Van der Valk and Fowke, 1981) one is eagerly awaiting the answer to the question of whether they are here involved in endocytotic or exocytotic processes or both.

VII. Concluding Remarks The desire to unify data on secretion-associated organelles from a variety of eukaryotic sources led to the birth of the endomembrane concept (Morri et al., 1971b; Morri and Mollenhauer, 1974). There is no doubt that a flow and differentiation of membrane from the ER to the PM via the GA is very appealing, yet there are certain aspects which have not gone without criticism. Farquhar (1978) has, for example, pointed out the equivocal nature of much of the biochemical evidence in the animal literature upon which the concept has been based. The situation with respect to membrane flow in animal cells is, as we now know, much more complicated than that originally contemplated. A bulk or mass-flow

120

DAVID G. ROBINSON AND UDO KRISTEN

from the ER to the PM has given way to more complicated mechanisms of selective transfer (MorrC et al., 1979; Morrk, 1980) and two-way vesicle traffic between organelles (Rothman, 1981). As far as plants, particularly higher plants, are concerned, there is no corresponding data available. Despite the lack of research in this field we have attempted to show that there are significant, though not basic, differences between the organelles involved in secretion in plant and animal cells. The major difference is a relative one: in accordance with the much smaller amounts of protein which are secreted, transport relationships between ER and GA are not of the same magnitude or type as they are in animal cells. In addition to {his the higher plant GA is structurally a much more polar organelle than is the case with animal cells (Mollenhauer and MorrC, 1978). We hope that in future these considerations of the plant GA will not be overlooked.

ACKNOWLEDGMENTS We thank A. Braeutigam for preparing the drawings, C. Adami and H. Freundt for helping with the photography, and, last but not least, G. Heidenreich for typing the manuscript. A word of thanks is also due to Larry Fowke and Lincoln Taiz for generously providing some of the figures. We also acknowledge useful exchanges of opinion with a number of colleagues, in particular W. Herth, H. Quader, E. Schnepf, P. Sitte, and W. Tanner. Our own research which has been mentioned in this article has been supported by the Volkswagen Stiftung and the Deutsche Forschungsgemein!ichaft to whom we are most grateful.

REFERENCES Ahkong. Q . W., Fisher, D., Tampion. W.. and Lucy, J. A. (1975). Nature (London) 253, 104-195. Amelunxen, F., and Gronau, G. (1966). 2. PJlanzenphysiol. 55, 327-336. Amelunxen, F., Tio Tiang Nio, and Spiess, E. (1976). Cytobiologie 13, 233-250. Baydoun, E. A.-H. and Northcote. D. H. (1980). J . Cell Sci. 45, 169-186. Beevers, H. (1975). In “Advances in the Chemistry and Biochemistry of Plant Lipids” (T. Galliard and E. 1. Mercer, eds.), Vol. 9. pp. 287-299. Academic Press, New York. Benner. U., and Schnepf, E. (1975). Protoplusmu 85, 337-349. ~. Cytol. 2, 99-103. Bennet. H. S. (1956). J. B i r ~ p h Biochern. Bergeron, J. J. M., Borts. D., and Cruz, J. (1978). J. Cell B i d . 76, 87-97. Bergfeld, R.. Kiihnl, T.. and Schopfer, P. (1980). Pluntu 148, 146-156. Bergmann. I . E., Tokuyasu, K. T., and Singer, S. I. (1981). Proc. Nut/. Acud. Sri. U.S.A. 78, 1746-1 750. Blobel, G., Walter, P., Chang, C. N., Goldman, B. M., Frickson, A. H.,and Lingappa, V. R. (1979). S y n p . Soc. Biol. 33, 9-36. Bouck. G. B. (1965). J . Cell B i d . 26, 523-537. Boussange, J., Douady, D., and Kader, J . C. (1980). Plant Phyiol. 65, 355-358. Bouvier-Durand. M., Dereuddre, J., and Come, D. (1981). Plunru 151, 6-14,

MEMBRANE FLOW VIA THE GOLGI APPARATUS

121

Bowles, D. J.. and Kauss, H. (1976). Eiochim. Eiophvs. Acta 443, 36CL374. Bowles, D. J., and Northcote, D. H. (1974). Eiochem. J. 142, 139-144. Bowles, D. J.. and Northcote. D. H. (1976). Plantu 128, 101-106. Bowles, D. J.. Lehle, L., and Kauss, H. (1977). Pluntu 134, 177-181. Bracker, C. E.. Grove, S. N., Heintz, C. E., and M o d , D. J. (1971). Cytobiotogie 4, 1-8. Brown. R. M . . Jr. (1969). J. CellEiol. 41, 109-123. Brown, R. M.. Jr., Franke, W. W., Kleinig, H., Falk, H., and Sitte. P. (1970). J. Cell Eiol. 45, 246-27 I . Brown, S. S . , and Spudich, J. A. (1981). J. Cell Eiol. 88, 487-491. Carpenter. G . , and Cohen. S. (1976). 1. Cell Eiol. 71, 159-171. Castle, J. D., Jamieson. J. D., and Palade, G. E. (1975). J. Cell Eiol. 64, 182-210. Chrispeels. M. J. (1969). Plunt Physiol. 44, 1187-1 193. Chrispeels, M. J . (1970). Eiochem Eiophvs. Res. Commun. 39, 732-737. Chrispeels, M. J. (1972). Pluntu 108, 283-287. Chrispeels, M. J. (1976). Annu. Rev. Plunt Physiol. 27, 19-38. Chrispeels, M. 1. (1977). In “Jnternational Cell Biology” (B. B. Brinkley and K. R. Porter, eds.). pp. 284-292. Rockefeller Univ. Press, New York. Chrispeels, M. J. (1980). In “The Biochemistry of Plants” (N. E. Tolbert, ed.), Vol. I , pp. 389-412. Academic Press, New York. Chrispeels, I . M., Bollini. R., and Hams, N. (1979). Eer. Dtsch. Eot. Ges. 92, 535-551. Croze. E. M., and Morre, D. J. (1981). Proc. Nut/. Acad. Sci. U.S.A. 78, 1547-1551. Cunningham. W. P.. Morre, D. J., and Mollenhauer, H. H. (1966). J. Cell Biol. 28, 169-179. Czichi, U.,and Lennarz, W. J. (1977). J. Eiol. Chem. 252, 7901-7904. Dalessandro. G.. and Northcote. D. H. (1977a). Eiochem. J. 162, 267-279. Dalessandro, G.. and Northcote, D. H. (1977b). Eiochern. J. 162, 281-288. Dauwalder, M., and Whaley. W. G. (1973). J. Ulrrusrruct. Res. 45, 279-296. Dauwalder, M.. and Whaley, W. G. (1974). J. Cell Sci. 14, 11-27. Dauwalder, M., Whaley, W. G..and Kephart, J. E. (1969). J. Cell Sci. 4, 455497. Dauwalder, M., Whaley, W. G., and Starr, R . C. (1980). J . Ultrustrurt. Res. 70, 318-335. Deysson. G . , and Benbadis, M.-C. (1972). J. Microsc. 13, 207-216. Dexheimer. J. (1972). C. R. Acad. Sci. Paris Sci. D 275, 1983-1986. Dexheimer. J. (1976). C.vtobiologie 13, 307-321. Dexheimer. J. (1978). Z. Pjlunzenphysiot. 86, 189-201, Dexheimer. J . (1981). 2.Pjlunzenphysiot. 101, 333-346. Dieckert. J. W.. and Dieckert. M. C. (1972). In “Symposium: Seed Proteins” ( G . E. Inglett, ed.), pp. 52-85. Avi Publ., Westport, Connecticut. Doohan. W. E., and Palevitz. B. A. (1980). Pluntu 149, 389-401. Douglas. W. W. (1974). Eiochem. Soc. Symp. 39, 1-28. Drawert, H., and Mix, M. (1961/1962). Sitzh. Ges. Eefird. Ges. Nuturwiss. Murhurg pp. 83/84. Durr. M., Bailey. D. S., and MacLachlan (1979). Eur. J. Biochem. 97, 445-453. Edgar. L. A. (1980). J. Phycol. 16, 62-72. Eisinger. W.. and Ray, P. M. (1972). Plunt Physiol. 49, (Suppl.), 2. Elder. J. H., and Morr6. D. J. (1976). J. Biol. Chern. 251, 5054-5068. Engelnian. D. M., and Steitz, T. A . (1981). Cell 23, 41 1-422. Evans. L. V., and Christie, A. 0. (1970). Ann. Boi. 34, 451-466. Evans. L. V.. Callow, M. E., Perciva.. E., and Fareed, V. (1974). J. Cell Sci. 16, 1-21. Evert. R . F. (1976). Isr. J. Eor. 25, 101-126. Evert. R. F.. and Eichhorn, S. E. (1974). Pluntu 119, 319-334. Fahn, A.. and Benouaiche, P. (1979). Ann. Bot. 74, 85-93. Falk, H. (1967). Arch. Mikrobiol. 58, 222-227.

122

DAVID G. ROBINSON AND UDO KRISTEN

Farquhar, M.G. (1978). In “Transport of Macromolecules in Cellular Systems” (S. C. Silverstein, ed.), pp. 341-362. Dahlem Konferenzen. Abakon, Berlin. Feingold, D. S., Neufeld, E. F., and Hassid, W. Z. (1960). J. Biol. Chem. 235, 910-913 Fessler, J. H., and Fessler, L. 1. (1978). Annu. Rev. Biochem. 47, 129-162. Fineran, B. A. (1973). Cyrobiologie 8, 175-193. Fisher, D. G., and Evert, R. F. (1979). Ann. Bor. 43, 255-258. Francisco, A., and Roth, L. E. (1977). Cyrobiologie 14, 191-206. Franke, W. W., and Herth, W. (1974). Exp. Cell Res. 89, 447-451. Franke. W. W., and Kartenbeck, J. (1976). In “Progress in Differentiation Research” (N. MiillerBtrat er ul., eds.), pp. 213-243. North-Holland Publ., Amsterdam. Franke, W. W., M o d , D. J., Deumling, B., Cheetham, R. D., Kartenbeck, J., Jarasch. E. D., and Zentgraf, H. W. (1971). Z. Nururforsch. 26b, 1031-1039. Franke. W. W., Kartenback, J., Krien, S., Van der Woude, W. J., Scheer, U., and M o d , D. J . (1972). Z . Zelljorsch. 132, 365-380. Franke, W. W., Luder, M. R., Kartenbeck, J., Zerban, H., and Keenan, T. W. (1976). J . Ceil B i d . 69, 173-195. Franz, G., and H d , D. (1980). Prog. Bor. 42, 80-95. Frey-Wyssling, A., Mpez-Siies, J. F., and Muhlethaler, K. (1964). J. Ulfrusrrucr. Res. 10, 422-432. Fries, E., and Rothman, J . E. (1981). J. Cell Biol. 90, 697-704. Gardiner, M., and Chrispeels, M. J. (1975). Plant Physiol. 55, 536541. Giddings, T. H., Brower, D. L.. and Staehelin, L. A. (1980). J . Cell Biol. 84, 327-339. Glas, R., and Robinson, D. G. (1982). Protoplusmu (in press). Green, J. R., and Northcote, D. H. (1978). Biochem. J . 170, 599-608. Green, J. R., and Northcote, D. H. (l979a). J. Cell Sci. 40,235-244. Green, J. R., and Northcote, D. H. (1979b). Biochem. J . 178, 661-671. Griffing, L. R. (1981). Ph.D. Dissertation, Stanford University. Griffing, L. R., and Ray, P. M. (1982). Plunru (submitted). Gunning, B. E. S., and Steer, M. W. (1975). “Ultrastmcture and the Biology of Plant Cells.” Arnold, London. Hall, W. T.. and Witkus, E. R. (1964). Exp. Cell Res. 36, 494-501. Hand, A. R., and Oliver, C. (1977). J. Cell Biol. 74, 399413. Hams, P. J . , and Northcote, D. H. (1971). Biochim. Biophys. Acra 237, 5 6 6 4 . Hartmann, M. A., Feme, M., Gigot, C., Brandt, R., and Benveniste, P. (1973). Physiol. Veg. 11, 209-230. Hartmann-Bouillon, M. A., Benveniste, P., and Roland, J.-C. (1979). B i d . Cell. 35, 184-194. Hanvood, R., Grant, M. E., and Jackson, D. S. (1974). Biochem. J. 144, 123-130. Heinrich, G. (1973). Protoplusmu 77, 271-278. Helsper, J. P. F. G., Veerkamp, J . H., and Sassen, M. M. A. (1977). PIanfu 133, 303-308. Hendriks, T. (1978). Plant Sci. Lerr. 11, 261-274. Hendry, A. T.,and Possmayer, F. (1974). Biochim. Biophys. Acru 369, 156172. Hepler, P. K., and Newcomb, E. H. (1967). J. Ultrusrruct. Res. 19, 498-513. Herzog, V., and Reggio, H. (1980). Eur. J. Cell Biol. 21, 141-150. Hitchcock, C., and Nichols, B. W. (1971). “Plant Lipid Biochemistry.” Academic Press, New York. Hoch, K., and Hartmann, E. (1981). Plunt Sci. Lerr. 21, 38S396. Hokin. L. E. (1968). Inr. Rev. Cyrol. 23, 187-208. Hopp, H. E., Romero, P.. and Pont Lezica, R. (1979). Plunr Cell Physiol. 20, 1063-1069. Hubbard, A. L. (1978). I n “Transport of Macromolecules in Cellular Systems” (S. C. Silverstein, ed.), pp. 363-390. Dahlem Konferenzen, Abakon, Berlin.

MEMBRANE FLOW VIA THE GOLGl APPARATUS

123

James, D. W.. and Jones, R. L. (1979). Planr Physiol. 64, 914-918. Jamieson, J. D., and Palade, G. E. (1967). J. Cell Biol. 34, 577-596. Jamieson, J. D., and Palade. G. E. (1968). J. Cell Eiol. 39, 589-603. Jamieson, J. D., and Palade, G. E. (1971). J . Cell Biol. 48, 503-522. Jamieson, F. D., and Palade, G. E. (1977). In “International Cell Biology” (B. R. Brinkley and K. R. Porter, eds.), pp. 308-317. Rockefeller Univ. Press, New York. Jilka, R., Brown, 0.. and Nordin, P. (1972). Arch. Biochem. Biophys. 152, 702-71 I . Julienne, M., Douady, D., Dubacq, J.-P. Tremolieres, A., Drapier, D., Grosbois, M., Mazliak, P., and Kader, J.-C. (1980). In “Biogenesis and Function of Plant Lipids” (P. Mazliak, P. Benvensite, C. Costes, and R. Douce, eds.), pp. 81-84. Elsevier, Amsterdam. Juniper, M. B. E., and Roberts, R. M. (1966). J . R. Microsc. SOC. 85, 63-72. Kader, J. C. (1977). In “Dynamic Aspects of Cell Surface Organisation” ( G . Poste and G. L. Nicholson, eds.). pp. 127-204. Elsevier, Amsterdam. Kawasaki, S. (1981). Plant Cell Physiol. 22, 431-442. Kawasaki, S., and Sato, S. (1979). Bot. Mug. (Tokyo) 92, 305-314. Kimber, S. J. (1981). J . Cell Sci. 50, 225-243. Klohs, W. D., and Goff, C. W. (1980). Eur. J. Cell Eiol. 20, 228-233. Kreibich, G., Bar-Nun. S . , Czacho-Graham, M., Mok, W., Nack, E., Okada, Y.,Rosenfeld, M. G.. and Sabatini, D. D. (1980). In “Biological Chemistry of Organelle Formation” (Th. Biicher. W. Sebald, and H. Weiss, eds.), pp. 47-164. Springer-Verlag. Berlin and New York. Kristen, U. (1972). Planta 106, 159-167. Kristen, U . (1974). Cvrobiologie 9, 36-44. Kristen, U. (1976). Protoplasma 89, 221-233. Kristen, U . (1978). Plunta 138, 29-33. Kristen, U . (1980a). Ber. Dtsch. Eor. Ges. 93, 587-594. Kristen, U . (1980b). Eur. J . Cell Eiol. 23, 16-21. Kristen, U. (1982). Bull. Soc. Bor. Fr. Acrual. Bot. (in press). Kristen, U . . and Biedermann, M. (1981). Ann. Bor. 48, 655-663. Kristen, U . , Liebezeit, G . , and Biedermann. M. (1982). Ann. Eor. 49, 569-584. Kristen, U.,and Lockhausen, J. (1982). Eur. J . Cell Biol., submitted. Lamport, D. T. A. (1980). In “The Biochemistry of Plants” (J. Preiss, ed.), Vol. 3, pp. 501-541. Academic Press, New York. Lamport, D. T. A.. and Miller, D. H. (1971). Planr Physiol. 48, 454-456. Lamport, D. T. A,, Katona, L.. and Roerig, S. (1973). Biochem. J . 133, 125-131. Lehle, L., Bowles, D. J., and Tanner, W. (1978). Planr Sci. Lett. 11, 27-34. Lemay, P . , Collyn-D’Hooghe, M., and Torpier, G. (1977). Exp. Cell Res. 109, 79-86. Lennarz, W. J. (1978). In “Transport of Macromolecules in Cellular Systems” (S. C. Silverstein, ed.), pp. 315-330. Dahlem Konferenzen. Abakon, Berlin. Lercher, M., and Wojciechowski, Z. A. (1976). Plant Sci. Lett. 7 , 337-340. Lodish, H . . and Small, B. (1975). J. CellBiol. 65, 51-64. Lodish. H. F . , and Rothman, J. E. (1978). Sci. Am. 240, ( I ) , 38-53. Lodish, H. F . , Braell, W. A., Schwartz, A. L., Strous, G. J. A. M., andzilberstein, A. (1981).Int. Rev. Cvtol. Suppl. 12, 247-307. McMurray, W. C . , and Magee, W. L. ( 1972). Annu. Rev. Biochern. 41, 129-160. Markey, D. R.. and Wilce, R. T. (1975). Proroplasma 85, 219-241. Marriott, K. M., and Tanner, W. (1979). Planr Physiol. 64, 445-449. Massalski, A., and Leedale, G. F. (1969). Br. Phytal. J . 4, 159-180. Meldolesi. J. (1974a). Philos. Trans R. SOC. London 268, 39-53. Meldolesi, J. (1974b). Adv. Cytopharmacol. 2, 71-84. Mellor. R . B., Roberts, L. M., and Lord, J. M. (1980). J . Exp. Bor. 31, 993-1003.

1 24

DAVID G. ROBINSON AND UDO KRISTEN

Menge, U.,and Kiermayer, 0. (1977). Mikroskopie 33, 168-176. Mollenhauer, H. H. ( 1965). J. Cell B i d . 24, 504-5 I I . Mollenhauer, H. H. (1967). Protoplasma 63, 353-362. Mollenhauer, H. H. (1971). J . CellBiol. 49, 212-214. Mollenhduer, H. H., and Mollenhauer. B. A. (1978). Planta 138, 113-1 18. Mollenhauer, H. H., and Morre, D. J. (1966a). Annu. Rev. Plant Physiol. 17, 27-46. Mollenhauer, H. H., and Morri, D. J. (1966b). J. Cell Bid. 29, 373-376. Mollenhauer, H. H., and M o d , D. J. (1972). What’sNew Planr Phvsiol. 4, 1-4. Mollenhauer, H. H. and Morre. D. J. (1974). Protoplasma 79, 333-336. Mollenhauer, H. H., and Morre, D. J. (1975). J. Cell Sci. 19, 231-237. Mollenhauer, H. H., and Mom& D. J. (1976a). Cyfobiologie 13, 297-306. Mollenhauer, H. H., and Morre, D. J. (1976b). Protoplasma 87, 39-48. Mollenhduer, H. H., and Morre, D. J. (1978). J . Cell Sci. 32, 357-362. Mollenhauer, H. H., and M o d , D. H. (1980).In “The Biochemistry of Plants” (P. K. Sturnpf and Conn. E. E., eds.), Vol. I . pp. 437-488. Academic Press, New York. Mollenhauer, H. H., and Totten, C. (1970). Plant Physiol. 46, 794-799. Mollenhauer, H. H., and Totten, C. (1971). J. Cell B i d . 48, 387-394. Mollenhauer. H. H., and Whaley. W. G. (1963). J. Cell B i d . 17, 222-225. Mollenhauer. H. H., Morre, D. J., and Totten. C. (1973). Protoplasma 78, 443459. Mollenhduer, H. H.. Morrt, D. J., and Van der Woude, W. J. (1975). Mikroskopie 31, 257-272. Molnar, J. (1976). In “The Enzymes of Biological Membranes. Vol. 2. Biosynthesis of Cell Components” (A. Martonose, ed.), pp. 385-419. Wiley, New York. Monroy, G., Kelker, H. C., and Pullman. M. E. (1973). J. Biol. Chem. 248, 2845-2852. Montdgue, M. J.. and Ray, P. M. ( 1977). Plant Physiol. 59, 225-230. Moore. T. S. (1976). Plant Phvsiol. 57, 383-386. Moore. T. S., Lord, J . M., Kagawa, T., and Beevers, H. (1973). Plant Phvsiol. 52, 5CL53. M o d , D. 3. (1970). Plant Phvsiol. 45, 791-799. Morre. D. 1. (1975).Annu. Rev. Plant. Phvsiol. 26, 441-481. Morre. D. J. (1977). In “International Cell Biology” (B. R. Brinkley and K. R. Porter. eds.), pp. 293-303. Rockefeller Univ. Press, New York. Morre. D. J. (1980). I n “Cell Compartmentation and Metabolic Channeling” (L. Nover, F. Lynen, and K. Mothes, eds.), pp. 47-61. Fischer, Jena, and Elsevier, Amsterdam. Morrk, D. J. (1981). In “International Cell Biology 198CL1981” (H. G. Schweiger, ed.), pp. 622-632. Springer Verlag. Berlin and New York. Morre. D. J . , and Mollenhauer, H. H. (1974). In “Dynamic Aspects of Plant Ultrastructure” (A. W. Robards, ed.), pp. 84-137. McGraw-Hill. New York. Morre. D. J., and Ovtracht, L. (1977). Int. Rev. Cvtol. Suppl. 5 , 61-188. M o d , D. J., and Van der Woude, W. J. (1974). I n “Macromolecules Regulating Growth and Development” (E. D. Hay and T. J. Papaconstantinou, eds.). pp. 81-1 1 I . Academic Press, New York. MorrC. D. J., Jones, D. D., and Mollenhauer, H. H. (1967). Planta 74, 286-301. Morre, D. J., Nyquist, S.. and Rivera. E. (1970). Plant Physiol. 45, 800-804. Morre. D. J., Mollenhauer. H. H., and Bracker, C. E. (1971a). I n “Results and Problems in Cell Differentiation. 11. Origin and Continuity of Cell Organelles” (J. Reinert and H. Ursprung, eds.), pp. 82-126, Springer Verlag. Berlin and New York. Morre. D. J., Franke, W. W., Deumling, B., Nyquist, S. E., and Overtracht, L. (1971b). Biomembranes 2, 95-104. Morre, D. J.. Mahley, R. W.. Bennet. B. D., and Lequire. V. S. (1971~).Absrr. Am. Soc. CellBiol. p. 199. M o d , D. J., Keenan, T. W., and Huang, C. M. (1974). Adv. Cvtopharmarol. 2, 107-125.

MEMBRANE FLOW VIA THE GOLGl APPARATUS

125

M o d , D. J., Lembi, C . A., and Van der Woude, W. J. (1977). Cytobiologie 16, 72-81. Morrk, D. J., Kartenbeck, J., and Franke, W. W. (1979). Biochim. Biophys. Acta 559, 72-152. Moms, M. R., and Northcote, D. H. (1977). Biochem. J . 166, 603-618. Mueller, S. C., and Brown, R. M., Jr. (1980). J. Cell Biol. 84, 315-326. M’Voula-Tsieri, M., Hartmann-Bouillon, M. A., and Benveniste, P. (1981). Plant Sci. Lett. 20, 379-386. Nagahashi, I., and Beevers, L. (1978). Plant Physiol. 61, 451-459. Neufeld, E. F.. Ginsburg, V., Putman, E. W., Fanshier, D., and Hassid, W. Z. (1957). Arch. Biochem. Biophys. 69, 602-616. Neutra, M., and Leblond, C. P. (1966). J . Cell Biol. 30, 119-136. Newcomb, E. H. (1980). In “Coated Vesicles” (C. D. Ockleford, and A. Whyte, eds.), pp. 55-68. Cambridge Univ. Press, London and New York. Northcote, D. H., and Pickett-Heaps, J. D. (1966). Biochem. J . 98, 159-167. Ohlrogge, J. B., Kuhn, D. N., and Stumpf, P. K. (1979). Proc. Natl. Acad. Sci. U.S.A. 76, I1941 198. Oliviera, L., Walker, D. C., and Bisalputra, T. (1980). Protoplasma 104, 1-15. Olsen, B. R., Berg, R. A., Kishida, Y., and Prockop, D. J. (1975). J . Cell Biol. 64, 340-355. Orci, L., Perrelet, A., and Friend, D. S. (1977). J . Cell BioI. 75, 23-30. Oron, U . , and Bdolah. A. (1978). J . Cell Biol. 78, 488-502. Otsuki, Y . , Takebe, I., Honda, Y., and Matsui, C. (1972). Virology 49, 188-194. Ovtracht, L., M o d , D. J., and Nyquist, S. E. (1970). Microsc. Electron. 3, 81. Palade, G. E. (1959). In “Subcellular Particles” (T. Hayashi, ed.), pp. 64-83. Ronald, New York. Palade, G. E. (1975). Science 189, 347-358. Panayotatos, N., and Villemez, C. L. (1973). Biochem. J . 133, 263-271. Paull. R. E . , and Jones, R. L. (1975). Planta 127, 97-110. Paull, R. E., and Jones, R. L. (1976). Plant Physiol. 57, 249-256. Pearse, B. M. F., and Bretscher, M. S. (1981). Annu. Rev. Biochem. 50, 85-101. Pellegrini. L., and Vogt, G. (1976). Ann. Sci. Nut. Bor. 17, 333-344. Peterkovsky, B., and Assad. R. (1976). J. Biol. Chem. 251, 4770-4777. Picton, J. M . , and Steer, M. W. (1981). J . CellSci. 49, 261-272. Pope. D. G., Thorpe, J. R.. Al-Azzdwi, M.J., and Hall, J. L. (1979). Planta 144, 373-383. Poux, N. (1973). C. R. Acad. Sci. Paris. Ser. D 276, 2163-2166. Powell, J. R., and Brew, K. (1974). Biochem. J . 142, 203-209. Quail, P. H. (1979). Annu. Rev. Plant Physiol. 30, 4 2 5 4 8 4 . Ray, P. M . (1980). Biochim. Biophvs. Acra 629, 431444. Ray, P. M., Shininger, T. L.,and Ray, M . M. (1969). Proc. Narl. Acad. Sci. U.S.A. 64,605-612. Ray, P. M., Eisinger, W. R., and Robinson, D. G. (1976). Ber. Dtsch. Bot. Ges. 89, 121-146. Reiss, H. D., and Herth, W. (1979). Planta 145, 225-232. Reiss, H . D . , and Herth, W. (1980). Planta 147, 295-301. Robinson, D. 0 . ( 1977). Adv. Bot. Res. 5 , 89-15 I . Robinson, D. G . (1980a). In “Cell Compartmentation and Metabolic Channeling” (L. Nover, F. Lynen, and K. Mothes, eds.). Fischer, Jena, and Elsevier. Amsterdam. Robinson, D. G. (1980b). Eur. J . Cell Biol. 23, 22-36. Robinson, D. G . (1981a). In “Cell Walls ’81” (D.G. Robinson and H. Quader. eds.), pp. 4>52. Wissenschaftliche Verlagsges., Stuttgart. Robinson, D. G. (1981b). Eur. J . Cell Biol. 23, 267-272. Robinson, D. G., Miiller, S., Freundt. H., and Hoppe, H.-H. (1982). In preparation. Robinson, D. G., and Cummins, W. R. (1976). Protoplasma 90, 369-379. Robinson, D. G., and Quader, H. (1981). J. Theor. Biol. 92, 483-495. Robinson, D. G . , and Ray, P. M. (1977). Cytobiologie 15, 65-77.

126

DAVID G. ROBINSON AND UDO KRISTEN

Robinson, D. G., Eisinger, W. R., and Ray, P. M. (1976). Ber. Drsch. Bor. Ges. 89, 147-161. Roland, J.-C. (1969). C. R. Acud. Sci. Paris. Ser. D 269, 939-942. Roland, J.-C. (1973). Inr. Rev. Cyrol. 36, 45-92. Roland, J.-C., and Sandoz, D. (1969). 1.Microsc. 8, 263-268. Roland, J.-C., and Wan, 9. (1971). Proroplusmu 73, 121-137. Rosen, W. G., Gawlik, S. R., Dashek, W. V., and Siegesmund, K. A. (1964). Am. J . Bof. 51, 61-71. Rothman, L. E. (1981). Science 213, 1212-1219. Rothman, J. E., Bruszty-Pettegrew, H., and Fine, R. E. (1980). J. Cell Biol. 86, 162-171. Rothman, S. S. (1975). Science 190, 745-753. Roughan, P. G., Holland, R., and Slack, C. R. (1980). Biochem. J. 188, 17-24. Rougier, M. (1971). J . Microsc. 10, 67-82. Rougier, M. (1982). In “Encyclopedia of Plant Physiology (F. Loweus and W. Tanner, eds.). Springer Verlag, Berlin and New York, in press. Ryser, U. (1979). Protoplusma 98, 223-239. Sadava, D., and Chrispeels, M.’J. (1971). Biochim. Biophys. Acru 227, 278-287. Sassen, M. M. A. (1964). Acru Bor. Neerl. 13, 173-181. Scallen, T. J., Srikantaiah, M. V., Seetharam, B., Hansbury, E., and Gamey, K. L. (1974). Fed. Proc. Fed. Am. SOC. Exp. Biol. 33, 1733-1746. Schnepf, E. (1961a). Flora 151, 73-87. Schnepf, E. (1961b). Z. Nuturforsch. 16b, 605-610. Schnepf, E. (1963). Flora 153, 23-48. Schnepf, E. (1964). Protoplasma 58, 193-219. Schnepf, E. (1968). Plunra 79, 22-34. Schnepf, E. (1969a). Ber. Drsch. Bot. Ges. 82, 407-413. Schnepf, E. (1969b). Proroplusrnurologie VIII 8, 8-22. Schnepf, E.. and Busch, J. (1976). Z. Pflunzenphysiol. 79, 62-71. Schnepf, E., and Christ, P. (1980). Protoplusmu 105, 135-148. Schnepf, E., and Koch, W. (1966). Arch. Mikrobiol. 54, 229-236. Sharon, N., and Lis, H. (1981). Chem. Eng. News 59, 21-44. Shore, G., and MacLachlan, G. A. (1975). J. Cell Biol. 64, 557-571. Shore, G. G., and Tata, J. R. (1977). Biochim. Biophys. Acta 472, 197-236. Sievers, A. (1963). Profoplusmu 56, 188-192. Sievers. A. (1964). Ber. Drsch. Bor. Ces. 77, 388-390. Sievers, A. (1973). In “Grundlagen dercytologie” (G. C. Hirsch, H. Ruska, and P. Sitte, eds.), pp. 28 1-296. Fischer, Jena. Sievers, A., and Volkmann, D. (1977). Proc. R . SOC. London Ser. B. 199, 525-536. Staby, G. L., Hackett, W. P., and De Hertogh, A. A. (1973). Plunr Physiol. 52, 416-421. Staehelin, L. A,, and Kiermayer. 0. (1970). J. CeN Sci. 7, 787-792. Sticher, L., Penel, C., and Greppin, H. (1981). J. Cell Sci. 48, 345-354. Sturgess, J., Moscarello, M., and Schacter, H. (1978). Curr. Top. Membr. Trunsp. 11, 16-105. Tanaka, M.,Shibata, H.,and Uchida, T. (1980). Biochim. Biophys. Acru 616, 188-198. Tanaka, T., and Yamada, M. (1979). Plant Cell Physiol. 20, 533-542. Tartakoff, S., and Vasalli, P. (1978). J. Cell Biol. 79, 694-707. Terris, S.. and Steiner, D. F. (1975). J. B i d . Chem. 250, 8389-8398. Trachtenberg, S.. and Fahn, A. (1981). Bot. Guz. 142, 206-213. Tulkens, P., Beaufay, H., and Trouet, A. (1974). J . Cell Biol. 63, 383401. Turner, F. R.. and Whaley. W. G. (1965). Science 147, 1303-1304. Ueda, K. (1966). Cyrologiu 31, 461-472. Ueda, K., and Noguchi, T. (1976). Protoplusmu 87, 145-162.

MEMBRANE FLOW VIA THE GOLGl APPARATUS

127

Vagelos. P. R. (1974). In “MTP International Review of Science, Biochemistry of Lipids I. Vol. 4.” (T.W. Goodwing, ed.), pp. 100-140. Butterworth, London. Van der Valk, P., and Fowke, L. C. (1981). Can. J . Bor. 59, 1307-1313. Van der Woude. W. J., Lembi, C. A., M o d , D. J.. Kindinger, J. I . . and Ordin, L. (1974). Plunr Physiol. 54, 333-340. Vamer, J. W., and Burton, J. E. (1980). Plant Phvsiol. 66, 1044-1047. Vian, B. (1974). C. R. Acad. Sci. Paris. Sci. Ser. C 278, 1483-1486. Vick, B., and Beevers, H. (1977). Plant Physiol. 59, 495463. Vick, B., and Beevers, H. (1978). Plunr Physiol. 62, 173-178. Volkmann, D. (1981). Planru 151, 180-188. Volkmann, D., and Czaja, A. W. P. (1981). Exp. Cell Res. 135, 22S236. Volkmann, D., and Severs, A. (1979). In “Encyclopedia of Plant Physiology” (W. Haupt, and M. E. Feinleib, eds.), Vol. 7, pp. 573-600. Springer-Verlag, Berlin and New York. Waksman, A., Hubert, P.,Crimel, G., Rendon, A,, and Burgun, C. (1980). Biochim. Biophys. Acru 604, 249-296. Wallach, D.. Kirshner, N.. and Schramm, M. (1975). Biochim. Biophys. Acru 375, 87-105. Wattiaux-De Comninck, S.,and Wattiaus, R. (1969). Biochim. Biophys. Acru 183, I1&128. Weber, E., and Neumann, D. (1980). Biochem. Physiol. Pflanzen 175, 279-306. Went, van, J. L. (1978). Soc. Bor. Fr., Actual. Bot. 1-2, 149-153. Whaley, W. G. (1975). Cell Biol. Monogr. 2, 1-190. Whaley, W. G., and Mollenhauer. H. H. (1963). J. Cell Biol. 17, 216-221. Whaley, W. G . , Dauwalder, M., and Kephart, J. E. (1966). J. Ulrrastruct. Res. 15, 169-180. Whaley, W. G . , Dauwalder, M., and Kephart, J. E. (1972). Science 175, 596-599. Wienecke, K., Glas, R., and Robinson, D. G. (1982). Planra (in press). Wilgram, G. F., and Kennedy, E. P. (1963). J. Biol. Chem. 238, 2615-2619. Williams, D. C. (1974). Calcif. Tissue Res. 16, 227-237. Williams. J. A. (1980). Am. J. Physiol. 238, G 269-279. Wilson, J. D., Gibson, K. D., and Udenfriend, S. (1960). J. B i d . Chem. 235, 3213-3217. Wirtz, K. W. A., and Silversmit. D. B. (1969). Biochim. Biophys. Acra 193, 105-116. Wirtz, K. W. A., Kamp, H. H., and Van Deenen, L. L. M. (1972). Biochim. Biophys. Acra 274, 606-617. Wrischer, M. (1960). Nuturwissenschajien 47, 522-523. Wrischer. M. (1965). Proroplasmu 60,355400. Yoo, B. Y . , and Chrispeels, M. J. (1980). Proroplasma 103, 201-204. Zaar, K., and Schnepf, E. (1969). Planru 88, 224-232.

This Page Intentionally Left Blank

INTERNATIONAL REVIEW OF CYTOLOGY. VOL. 77

Cell Membranes in Sponges WERNERE. G. MULLER lnstitur fur Physiologische Chemie, Abteilung “Angewandte Molekularbiologie, ” Universitat, Duesbergweg, Mainz. Federal Republic of Germany

...................

129

111. Intercellular Matrix . . . . . . . . . . ................... IV. Morphology of Cell Contact.. ........................ V. Primary Cellular Recognition . . . . . . . . . ................ A. Primary Aggregation . . . . . . . . . . . . ............ B . Species Specificity. . . . . . . . . . . . . . . ............

132

........ A. Secondary Aggregation . . . . . . . . . . . . . . B . Species Specificity ........................... C. Formation of a Reconstituted Organism. ........

145 155 158

.......

161

I . Introduction .....................

IX. Recognition of Symbionts. ...................... X. Concluding Comments . . . . . . ...................

137 143

175

........................

I. Introduction Sponges are the first and oldest multicellular animals; marine species of the order Triaxonida are known since the Proterozoic (lo00 million years), while the first entire fresh water species, belonging to the family Spongillidae, is from the Tertiary (50 million years) (Orlov, 1971; Miiller et al.. 1982a). The cells of the sponges do not form true tissues or differentiated organs; nevertheless the different species are characterized by a more (e.g., Hyalonematidae) or less (e.g., Clionidae) typical form. During evolution from unicellular to multicellular organisms, sponges have developed a number of recognition systems localized on the cell periphery or on the para- and extracellular matrix. These systems are prerequisites for the establishment and stabilization of the functional arrangement of cells in the organism. A single sponge specimen appears to be a stable cell population organized into an external epithelium (wall to the surrounding milieu), internal epithelia (walls of the aquiferous canals), and mesohyl (space between the two epithelial forms). Cell coordination is not realized by a nervous system but takes place according to I29 Copyright 0 1982 by Academic Press. Inc. All rights of reproduction in any form reserved. ISBN 0-12-364477-1

130

WERNER E. G. MULLER

the following three pathways (Pavans de Ceccatty, 1974): (1) a nondirected diffusion into the extracellular spaces (extracellular fluid coordination pathway); (2) transmission through stable alignments of epithelial or mesohyl cells (fixed tissue coordination pathway); and (3) cellular transmission by the intermediary of a temporary network set up by means of transitory contacts between mesohyl cells (mobile cellular coordination pathway). The mesohyl in particular contains besides skeletal elements a variety of “wandering cells” belonging primarily to the archeocytes (John et al., 1971). Therefore, sponge cells must be and are indeed provided with well-developed biochemical mechanisms to enable them to contact inhibition of movement (Abercrombie and Heaysman, 1954), to contact promotion (Curtis, 1973a), and to contact guidance (Weiss, 1941). In addition, one might postulate that sponge cells receive and respond to chemical gradients; until now no experimental data have been presented, indicating the existence of a chemotactic and of a chemokinetic system in sponges. Because of these functions of sponge cells together with their pronounced characteristics of plasticity of their differentiation potency (Diaz, 1979) it is conceivable that sponges are provided with one or more especially efficient recognition systems, localized on their cell membranes. These possible recognition systems have been studied extensively on the cellular, subcellular, and biochemical level, starting with the pioneering works of the marine biologists Grant (1829, Bowerbank (1858), Vosmaer (1887), and Wilson (1912). In addition to studies of membrane-mediated recognition systems, membrane research in sponges has a second major subject, namely, elucidation of the composition and the function of the intercellular matrix (survey: Garrone, 1978). Other aspects of sponge cell membranes, e.g., architecture and topography, chemical composition and transport systems, have been only poorly studied and are not subjects of this article. In the first two sections of this article the different cell types in sponges and the intercellular matrix will be discussed only briefly, while in subsequent sections the mechanisms of different cell recognition systems are treated.

11. Cell Types in Sponges One characteristic of a sponge individual is that it consists of two phases: first, a living phase composed of different cell types, and second, a nonliving phase which contains the matrix material, synthesized by the cells. Due to this peculiarity some sponge cell types can move around within the mesoglial environment. The different somatic cell types can be classified according to their functions into the following groups (Bergquist, 1978; Borojevic et al., 1968). Cells which line surfaces. The pinacoderm, which separates the sponge mesohyl from the environment (external surface, basal attachment lamina, and all

CELL MEMBRANES IN SPONGES

131

exhalant and inhalant canals), is composed of pinacocytes. Because of structural and functional differences they are subdivided in exopinacocytes (lining the external surface), endopinacocytes (surrounding the canals), and basopinacocytes (forming the basal attachment lamina). The overall shape, especially of the exopinacocytes and the basopinacocytes, is a “T” with flattened overlapping extensions in the plane of the surface, while the cell body with the nucleus is embedded in the mesohyl. The porocytes, a further cell type, derive from the exo- or the endopinacoderm and enclose a pore. Provided with these morphologic characteristics, the porocyte can function as one component of the inhalant system, connecting the choanoderm with the external aqueous milieu. Porocytes are contractile. The choanocytes are flagellated cells and are surrounded by a collar of cytoplasmic microvilli. Cells which secrete the skeleton. Collencytes and lophocytes secrete fibrillar collagen. While the collencytes are rather immobile cells, the lophocytes are very mobile and synthesize collagen as they move. The two other major components of the sponge skeleton, spongin and the spicules, are produced by the spongocytes or the sclerocytes, respectively. Cells which cause contraction. The cells which have contractile functions are the fusiform myocytes. They are often grouped around oscules and major canals, the two predominant contractile organization units in sponges. Cells with inclusions. These mobile mesohyl cells are characterized according to the form and number of their granules or vesicles. The spherulous cells contain a large number of large round vacuoles while the microgranular cells are filled with small dense granules. The very mobile gray cells contain large oval to spherical, basophilic granules and small glycogen-containing granules. The globoferous cells are distinguished by having a large globule, consisting of parallel rows of cylindrical rods. The rhabdiferous cells are filled with rod-like inclusions. Totipotent amoeboid cells. The archaeocytes are amoeboid, very mobile cells and capable of being the origin of any other type of cell. Some evidence has even presented (Diaz, 1979) which indicates that at least some differentiated cell types (e.g., pinacocytes and choanocytes) still have the capacity to dedifferentiate and redifferentiate into any other cell type via the archaeocyte cell stage. In contrast to those eukaryotes which are organized into organs, the sponge body consists of only two relatively stable cell layers, the outer and the inner epithelial layer. Between these two walls is the mesohyl, which contains highly mobile cells. Some of these cells are apparently finally differentiated while others have retained their differentiation capacity and thereby can change their function. The cells are embedded in a mesoglial environment which provides the space for migration and displacement. Hence, in order to establish and to maintain a functional organism, the sponges must be equipped with precisely acting and tuned coordination principles (Pavans de Ceccatty, 1974). The fixed coordi-

132

WERNER E. G. MULLER

nating system requires cells which stay in permanent connection with each other and in addition a stable guideline (e.g., collagen fibers) for their orientation. Cells belonging to this system are found in the mesoglial contractile tissue and in the ephithelial layer. In the second, mobile coordination system cells must be involved which have retained their migrating capacity and which are able to form transitory contacts. These properties imply that these cells are provided with mechanisms to expose and to recognize differential cellular affinities. In other words, the membranes especially of these wandering cells must be equipped with cell-cell recognition systems which are not static but adaptable to different and changeable environmental conditions.

111. Intercellular Matrix

The intercellular matrix, the nonliving phase of the sponges, is composed of a fibrous organic network, of a nonstructural ground substance of a glycoproteic nature and of inorganic skeletal components. In this section, the organic matrix will be discussed with respect to its interaction with the cell membrane. The first type of fibrils, the fibrillar collagen (= spongin A), discovered in sponges by Gross et ul. (1956), was studied in detail later by Garrone and Pavans de Ceccatty (review: Garrone, 1978). The cells which synthesize collagen have been identified as lophocytes (Ankel and Wintermann-Kilian, 1952). The fibrils are of undefinable length and have a diameter ranging from 20 to 30 nm Two main morphological subtypes are distinguished; first, the smooth fibrils, occurring in the Demospongiae Tetractinomorpha, with a poorly defined but regular cross-banding of a periodicity of around 22 nm, and second, the rough fibrils of the Demospongiae Ceractinomorpha, with a periodicity of 22-25 nm. The fibrils are often surrounded with glycoproteins which appear either as a light ring around the fibrils or as fine filaments which interconnect the fibrils (Garrone et ul., 1975). The fine filaments linked to the collagen fibrils from Geodiu cydonium are very long (1.5 km) (Zahn and Miiller, unpublished) (Fig. IG). Some of the fine filaments are loosly associated with the fibrous complex of the aggregation factor. The second kind of fibers occurring in sponges are the skeletal spongin Fibers (= spongin B) which have been classified according to Garrone (1978): first, as spiculated fibers, which surround the endogenous inorganic skeleton of the animal; second, as the macroscopically visible ramified fibers, which are characteristic of horny sponges; and third, as the material which forms the shell of the gemmules. The different varieties of the spongin fibers have their collagenous nature in common (Junqua et ul., 1974). Although fibrillar, their molecular organization is different and only occasionally periodic traverse banding is observed. The spongin fibers are synthesized in a well-defined cell type, the spongocytes.

FIG. I . Reaggregation of chemically dissociated cells from Geodiu rydoniurn. (A) Single cells. ~ 2 0 0 (.B ) Primary aggregates. X 6 5 . (C) Secondary aggregates. X 15. (D)Reconstituted organism. x 15. (E) Electron micrograph of the native aggregation factor; preparation shadowed with platinum. ~ 7 0 , 0 0 0 .(F) Electron micrograph of the core structures of the aggregation factor; preparation shadowed with platinum. X55.000. ( G ) Smooth collagen fibrils, associated with long fine filaments from G . cvdonium; stained with uranyl acetate and lead citrate. X35.000. (Photo courtesy of R. K. Zahn.)

134

WERNER E. G. MULLER

Besides the collagenous fibrils, noncollagenous components contribute to the organic intercellular material. These glycoproteic substances in the intercellular space cannot be visualized electron microscopically due to their lack of electron density. A characterization of the glycoproteins has recently been carried out by Junqua er ul. (1981). Based on differences in the sugar contents two types are distinguished: glycoconjugates I, containing about 6% carbohydrate, show a high affinity for concanavalin A and/or wheat germ agglutinin, and glycoconjugates 11, containing 30% carbohydrate and some sulfate, are not bound hy the two lectins. The major sugar components which contribute to the carbohydrate moiety are glucuronic acid, galactose, fucose, mannose, N-acetyl glucosamine, and arabinose. Sponge cells are surrounded by characteristic coats (Garrone, 1978), which can be either specialized (hair-like appendages, protrusions of the bottom of the choanocyte collar, and the cover of the microvilli of the choanocyte collar) or unspecialized. The latter ones are also termed diffuse coats. They are not directly visible but can be visualized after special cytochemical treatments (Rambourg, 197I ) . The glycopeptides present in these coatings contain D-galactose, Lfucose, glucose, mannose, rhamnose, arabinose, xylose, hexosamines , and glucuronic acid (MacLennan, 1969). Sialic acid does not seem to be present in the cell coat (MacLennan, 1970, 1974), even though the material seems to be accessible to the enzymic action of the sialidase (Garrone et al., 1971). Little experimental work has been performed to elucidate the possible function of the extracellular matrix during cellular coordination in sponges. Cells from vertebrates are known to adhere to collagenous substratum (e.g., Michalopoulos and Pitot, 1975). Preliminary experiments with Heteronemu erecta revealed that cells from this sponge species attach firmly to homologous collagen which has been spread as a film on plastic plates (unpublished result). One is not allowed to conclude from this in virro study that native collagen is in vivo a favored substratum for cellular coordination. Nevertheless, results from Weissenfels ( 1978) indicate that the Ephyduria fluviutifis skeleton, composed of siliceous spicules cemented into a regular arrangement by the collagenous spongin, can be repeatedly used from germinating sponges. This result might imply that the organized skeleton of the sponge is used as a morphogenetic route. The question of which regulating factors modulate the activity of the lophocytes, spongocytes, and spherulous cells (Bretting et a l . , 1982) during the synthesis of the highly ordered skeleton is still unsolved; the observed regularity cannot be explained simply by a contact genesis of collagen secretion (Levi, 1960). More obvious is the relation of diffuse cell coats to cell recognition and cell interaction. Since the discovery that the glycoproteins of the extracellular matrix and of the cell coats show species-specific characteristics (Junqua et af.,1979, interest in sponge glycoconjugates has been greatly enhanced. The assumption

CELL MEMBRANES IN SPONGES

135

that the diffuse cell coat-associated glycoconjugates are involved in cell membrane recognition (Garrone, 1978) has been partially substantiated by recent findings (Muller et al., 1982b). The high-molecular-weight aggregation factor, isolated from G . cydoniurn, has been found to be embedded in the diffuse cell coating material. At the present state of knowledge the glycoconjugates of the diffuse cell coats may not be looked at as primary components in the cell-cell and cell-matrix interactions. It seems to be more likely that the glycoconjugates exhibit major functions in the control of the cell surface-bound aggregation recognition molecules and the intercellular aggregation factor.

IV. Morphology of Cell Contact Based on electron microscopic studies the following four types of cell contact have been recognized (Farquhar and Palade, 1963; Brightman and Palay, 1963; Goodenough and Revel, 1970; Curtis 1973a): first, the zonula adherens (= belt desmosome), in which the cells are separated by a gap of 60-300 A.Second, the zonula occludens (= tight junction), in which the cells come at least within 20 A of each other. Third, the gap junction (= nexus), in which the cell surfaces are connected by honeycomb structures permitting passage of certain molecules between their cytoplasms through channels. Fourth, macula adherens (= desmosome), which is characterized by the presence of electron-dense fibrous material in the intercellular space (about 200 A) between two cell membranes; the cytoplasm beneath the plasmalemmae appears as a dense fibrillar material. It is a general feature of cell adhesion that during cell-cell contact formation three or even four distinct types of junction with different rates of adhesion are involved. It is assumed (Garrod and Nicol, 1981) that the different junctional mechanisms contribute to cell adhesion at different times after the establishment of contact; the initial contact occurs by adherens junctions followed by gap junctions, tight junctions, and desmosomes. It is an open question as to whether the four types of ultrastructurally recognizable junctions also have four corresponding different types of aggregation receptors and aggregation factors. Sponge epithelia have two peculiar features: first, they do not contain a basal lamina (basement membrane) on their inner surface, and second, the cell junctions generally show little specialization (Garrone, 1978). Bagby (1970) reported that classical desmosomes do not occur between pinacocytes of Microciona proliferu. Using pinacocytes from E . Jluviatilis (Feige, 1969; Pottu-Boumendil , 1975) contact areas have been described which appear as typical desmosomes. These junctions appear as seam welds holding the two cell membranes together. The intercellular space varies between 100 and 150 the length of the desmosomes is approximately 2000 In addition the typical massive structure of microfilaments (diameter of filaments: 50-100 A) within the cytoplasm of the

A.

A;

I36

WERNER E. G . MULLER

two cells close to the contact is present. Based on the electron micrographs available it is not possible to judge whether an electron-dense fibrous material in the intercellular space exists. In analogy to the function of desmosomes in epithelia of higher organisms, the junctions of this type in the pinacoderrn of sponges seem to be purely mechanical. Tight junctions have been described for Hippospongia communis between pinacocytes (Pavans de Ceccatty et ul., 1970) and between choanocytes of Microciona cells (Revel and Goodenough, 1970). The apposed membranes remain separated by a 0-20 A space. No evidence has been presented demonstrating an exchange of intracellular material between two adjacent cells without permeating into the intercellular spaces. Therefore these junctions should not be classified as gap junctions, if the definition of Pappas (1975) is applied. Some mesohyl cells from E . fluviutilis (De Vos, 1974) and from Sycon ciliarum (Ledger, 1975) are temporarily aligned by specialized septate junctions (= septate desmosomes). Septate junctions are found between sclerocytes, which are involved in spicule secretion. The junctions show the typical ladderlike appearance with a membrane separation of 150-200 A. They are often irregularly spaced with a minimum center-to-center separation of about 200 A. The central portion of each septum is wider than the parts adjacent to the plasma membrane. Because this junction type is not found in epithelia, Ledger (1975) assumed septate junctions are not involved in a mechanical stabilization of the cellular contacts, but rather in the establishment of an occluding barrier in this highly specialized microenvironment. In the extracellular milieu of the sclerocytes steep ionic gradients (especially for Ca2 ) exist which have to be maintained for a controlled and efficient calcite deposition. No differences between the cell surface in intact sponges and the plasma membrane of cells disaggregated in Ca2 -free medium have been observed (Revel and Goodenough, 1970). Using the fresh water sponge Spongillu carteri. a variety of cell junctions have been visualized during the early phases of reaggregation of dissociated cells (Kartha and Mookerjee, 1979). The behavior of the cell membrane has been attributed to a sequence of changes which the cells undergo during the process of morphogenesis. Tight junctions are the initial mode of contact with some indication of formation of desmosomes. The pattern of contact varies from close apposition of membranes over a large area and joining of reciprocal, undulatory or dovetailing surfaces to filipodial contact. Summarizing, the structures of most junctions between sponge cells are of a primitive, unspecialized type and have an intercellular space between 300 and lo00 A. One might ask, whether this primitive feature of sponges is due to a lack in genetic potential of sponges or to a functional reason. The latter possibility seems to be more likely because in contrast to tissue cells from higher invertebrates, sponge cells are characterized by a higher motility. Even the cells in the sponge epithelia retain the capacity for a permanent sorting out and reanange+

+

CELL MEMBRANES IN SPONGES

137

ment. These properties stipulate mechanisms for a rapid loosening and a subsequent reestablishment of cellular contacts. Therefore, it might be of advantage for sponge cells not to be provided with highly specialized junctions, which might result in a strong intercellular adhesiveness but instead to be outfitted with a powerful, reversible cell membrane-localized molecular aggregation system. In general, two types of forces cause intercellular junctions: first, physical interactions, as for example charge patterns, which might be sufficient in some instances to control the strength of adhesion (Steinberg, 1970; Curtis, 1973b), and second, cell membrane components or cell membrane-associated elements, which bridge the gap between neighboring cells. The studies to elucidate cell surface components involved in recognition of sponge cells started already in 1825 (Grant). The results contributed to and influenced, in a dominant manner, our present view about biochemical aspects of intercellular adhesion in multicellular organisms.

V. Primary Cellular Recognition A. PRIMARY AGGREGATION Marine and fresh water sponges can be dissociated either mechanically by squeezing them through a 10-100 pm mesh (Wilson, 1907; Miiller, 191 1; Brien, 1936) or chemically, by dissociation in the presence of Ca2+- and Mg2+-free (CMF) sea water (Moscona, 1963; Humphreys, 1963) which contains in some cases proteolytic enzymes (Muller and Zahn, 1973). With both techniques single cell suspensions are obtained (Fig. IA). After washing, in order to remove the soluble aggregation factor (see next section), the cells form small aggregates with diameters ranging from 60 to 200 pm (Wilson, 1912; Humphreys, 1963; Muller and Zahn, 1973) (Fig. 1B). These relatively small aggregates were termed “primary aggregates” (Muller and Zahn, 1973) to emphasize that the biological and biochemical mechanisms underlying the formation of these aggregates are clearly to be distinguished from those of the “secondary aggregates.” We define primary aggregation as an event which occurs in the absence of any detectable, soluble aggregation factors. Cells from all marine sponges,. hitherto studied, form primary aggregates; Calcarea: Homocoela: Clarhrina coriacea (Muller, unpublished); Calcarea: Heterocoela: Sycon raphanus (Huxley, 1912; Miiller, unpublished); Tetraxonidae: Homosclerophora: Ancorina cerebrum (Miiller, unpublished) and C.cydonium (Muller and Zahn, 1973); Tetraxonidae: Astromonaxonellina: Cliona celata (Humphreys, 1963) and Cfiona vastifica (Muller, unpublished); Comacuspongida: Poikilorhabdina: M . prolifera (Wilson, 1912; Humphreys, 1963). The morphological appearance of the primary aggregates depends upon the assay method used for the reaggregation experi-

I38

WERNER E. G. MULLER

ments. If the reaggregation is performed in rotating flasks (rotation-mediated aggregation), the cells in the small clumps are arranged in a three-dimensional organization pattern (Humphreys, 1963); however if reaggregation occurs in roller tubes (roller-mediated aggregation), the primary aggregates show a typical two-dimensional organization pattern (Miiller and Zahn, 1973). Irrespective of the method used, the cells in the primary aggregates are in the initial state randomly mixed (Humphreys, 1963). The physical and chemical parameters of primary aggregation are known to a very limited extent and were found to be the same in the three most thoroughly studied systems: M. prolifera, Halichondria panicea, and G . cydonium (Moscona, 1963; Humphreys, 1963; Miiller and Zahn, 1973; Miiller er’ al., 1974). Aggregation requires the divalent cation calcium at a concentration of at least 18 mM; at 3.8 mM the degree of aggregation is reduced to 50%. Strontium and barium can replace calcium while beryllium does not support aggregation. The conflicting results about a possible influence of magnesium on cell aggregation in the Microciona and Geodia system can most likely be attributed to the different assays used in the studies; some evidence is available that this cation is at least necessary for cell migration (Galtsoff, 1925; Humphreys, 1963). Mechanically dissociated Microciona cells (Humphreys, 1963) and chemically dissociated Geodia cells (Miiller et al., 1978a) reaggregate equally well at incubation temperatures between 0 and 40°C. Cell aggregation shows a pH optimum around 8.2; below pH 6.5 and above pH 9.5 primary aggregation is aboliished. Our knowledge about the membrane components involved in the formation of primary aggregates is very limited. Early experiments by Moscona (1963, 1968) revealed that formalin-killed cells retain their ability to form primary aggregates with a diameter of 10-50 pm. Preincubation experiments with a series of proteases and carbohydrate-hydrolyzing enzymes (Miiller el d.,1974) failed to hinder the cells from forming primary aggregates. The first evidence for the existence of a “factor” involved in primary aggregation was presented by Humphreys (1963). In the Microciona system he could show that chemically dissociated cells do not form primary aggregates at 5”C, while mechanically dissociated cells adhere very rapidly. The conclusion from these results was the assumption that chemically dissociated cells have lost a membrane factor, involved in primary aggregation. Further support for the existence of such a factor came from immunological studies. Spiegel ( 1954) prepared antisera against intact sponge cells from M. prolifera and C . celata. With these antisera, reaggregation of homologous cells is reversibly inhibited; heterologous antiserum is not inhibitory. The conclusion of Spiegel, that these antisera block the aggregation-promoting function of a molecule present on the cell surface, is supported by recent studies (Conrad and Miiller, unpublished) (Table I). Antiserum against a fractionated extract from the sponge Geodia cydonium, containing molecules with a molecular weight larger than 17,000, was found to inhibit the formation of

CELL MEMBRANES IN SPONGES

139

TABLE I ANTISERUMAGAINST THE SOLUBLE EXTRACT FROM Geodia cvdonium ON PRIMARY AGGREGATION'I

EFFECTOF

AN

Size of the primary aggregates

Antiserum dilution X I

x4

x t X

h

Control serum Without serum

28 36 57 62 73 68

f

5

*4

2 7 f 10

39 49 78 85

* 12

100

? 10

93

(1 Geodia cydonium cubes were extracted with CMF-sea water, containing 20 mM EDTA (4°C; 3 hours). The cell-free supernatant was fractionated by gel filtration on Sephadex G-50; the fractions eluting between a VJV, value of I .O and I .4 were pooled and injected into rabbits. Antiserd were prepared (Conrad er a / . , I98 I j and 100-pIsamples were assayed in the standard aggregation assay (Miiller and Zahn, 1973) using chemically dissociated cells. The size of the aggregates was determined after 10 minutes (Miiller et al., 1979a). Controls were performed with serum from an unimmunized animal and in the absence of serum.

primary aggregates strongly. Taking these pieces of evidence together, it is conceivable that a specific molecule, which might be solubilized from the cell membrane, is involved in the formation of primary aggregates. Comprehensive studies by us and Humphreys (1963) demonstrated that dissociated cells from species, belonging to the following two classes (and subclasses): Calcarea (Homocoela and Heterocoela) and Demospongia (Homosclerophorida, Tetractinomorpha, and Ceratinomorpha), form well-pronounced primary aggregates (see above); unfortunately no data about aggregation phenomena of Hexactinellida species have been published. However it is important to note that Calcarea have the ability to only form primary aggregates (unpublished result) and hence do not contain a soluble aggregation factor. In contrast, the Demospongia are provided with the mechanisms for the formation of primary and secondary aggregates. From these data one might conclude that the primary aggregation mechanism is the initial and prerequisite event for the cell-cell recognition in sponges. This assumption could be supported by recent findings (Miiller, unpublished) which revealed that one sponge species belonging to the Demospongia does not contain a soluble aggregation factor under certain physiological conditions and consequently cells from these organisms have the ability to form only primary aggregates. The sponge C. celutu (Astromonaxonellina: Clionidae), one of the classic

I40

WERNER E. G. MULLER

sponges, used for cell aggregation studies (Moscona, 1965; Humphreys, 1963) shows a pronounced polymorphism and occurs in the following three stages (Vosmaer, 193I ) . In the mature a-stage this species inhabits limestone rooks and occasionally shells (= boring form) (Fig. 2A); in the P-stage the sponge overgrows the surface of the substratum (= transition form); and in the y-stage the sponge is massive and lives entirely on the substratum (= crust form) (Fig. 2B). The conditions which favor one of these stages are not known. The studies of Humphreys (1963) and Jumblatt et al. (1980) were performed with the crust form from which a soluble aggregation factor has been isolated and its function during secondary aggregation has been studied. We collected the sponge in its typical biotope (Hartman, 1957) in the low tide zone (boring form) or by scuba diving in the sand and mollusk shell zone (crust form) in a depth of 25-35 m near Rovinj (Yugoslavia). We compared the cell-cell recognition system(s) of cells obtained from the boring form with that of the cells from the crust form. The experiments revealed that mechanically dissociated cells from both forms assemble to primary aggregates (Fig. 2C and D). Compared to results obtained with cells from other species only a very few cells (less than 5%) were determined not to be included in the primary aggregates. The conditions necessary for the formation of primary aggregates were the same as those described above for G . cydonium and M. prolifera. Using the established procedures for the isolation of the soluble aggregation factor, treatment of the tissue with CMF sea water (Humphreys, 1970a), CMF-EDTA sea water (Ca2+- and Mg2+-free sea water containing 20 mM EDTA) (Muller and Zahn, 1973), or CMF sea water, containing 2 mM lithium diiodosalicylate (Muller et al., 1979b) an aggregation factor could be isolated only in a soluble state from the crust form (Table 11). In the absence of any extracts, the cells obtained by a chemical dissociation procedure form in the presence of Mg2 and Ca2 in the sea water primary aggregates with a diameter between I10 and 120 pm. Addition of a crude extract from the boring form has no influence on the size of the aggregates. However in the presence of crude extract from the crust form or even a purified aggregation factor from the same tissue, the single cells form large secondary aggregates (Fig. 2E). In order to rule out that a possible inhibitor of the aggregation factor, which might be present in the crude extract from the boring form, abolishes its activity, the extract was fractionated through a Sephacryl S- lo00 column (Fig. 3). No aggregatiori-promoting activity could be detected in any eluted fraction. In comparison, after fractionation of the crude extract from the crust form, the aggregation factor could be recovered within the characteristic VJV, range (Miiller and Zahn, 1973; Henkart et al., 1973) between 1.00 and 1.15; the specific activity was determined to be 1.7 X lo6 AU/mg protein. Also of possible interest are different elution patterns of the extracts (Fig. 3). From these results it must be concluded that the cellular system of the boring form of C. celata does not have the ability to form secondary aggregates but they do not prove rigorously the absence of a +

+

FIG. 2. Aggregation mechanisms in Cliona celafa. (A) C. celafa. boring form. X I .5. (B)C. celafa. crust form. X0.7. (C) Primary aggregates from cells of the boring form. X60. (D)Primary aggregates from cells of the crust form. x70. (E) Secondary aggregates from cells of the crust form after incubation with the soluble aggregation factor. X 2 0 . (F) Aggregation factor from the crust form stained with uranyl acetate. X50.000.

TABLE I1 A G G R ~ G A T I OPOT€ N N C ' l t S Ob SlNGLt CtLLS PROM C / l U t l U (BOTH BORINGFORMA V D CRUSTFORM)"

Single cells from

Size of the aggregates ( p i )

Addition

Boring form

Crust form

ll'/UtU

None Crude extract boring form Crude extract crust form Purified aggregation factor

110 2 15 115 t 15 2120 t 270

None Crude extract boring form Crude extract crust form Purified aggregation factor

120 110 2270 1800

1750

-t

200

2 15 2 15

t 280 t 200

The sponges were dissociated in CMF-EDTA sea water and the cells were incubated under standard conditions (Miiller and Zahn, 1973) in sea water, in the presence of crude extracts both from the boridg form and from the crust form, or in the presence of purified aggregation factor (see Fig. 3); the final concentrations of the crude extracts were 50 pg/ml and of the aggregation factor 1.5 X lo4 aggregation units/ml. The size of the aggregates was determined after 60 minutes.

7," ,

!

.

n

0

0

30

LO

50

60

70

80

90

F r a c t i o n number

FIG. 3. Fractionation of the crude extracts from Clioria celaia (boring form and crust form) by gel filtration on Sephacryl S-IOOO.The sponge tissues were extracted with CMF-EDTA SLI water (Miiller and Zahn, 1973); the cell-free extracts had a protein content of I . 3 mg/ml (boring form) or 2.7 mg/ml (crust form), respectively. Protein was determined according to Lowry er a / . ( I 95 I ). In separate experiments, 5 ml of the two crude extracts was applied onto a Sephacryl S- 1000 ccllumn (2 x 28 cm) and eluted with CMF sea water. Fractions of I ml were collected and aliquots were removed for analysis of aggregation activity (using "homologous cells") and of optical density. Abscissa above: VJV,, value (Determann, 1969). Elution profile of crude extracts from boring form ((c--.) and from crust form (0--0). Aggregation-promoting activity. given in aggregation no units (AU; Miiller and Zahn, 1973). in the fractions from the crust form extract (X-.--X); activity was detected in the fractions from the boring form extract (the zero-line is not ins:ned).

143

CELL MEMBRANES IN SPONGES

soluble aggregation factor. It could well be that such a factor is present also in the boring form but that the corresponding aggregation receptor which is also required in the two component system (see next section) is absent. However on the basis of the results obtained from cross-reactions (Table II), using single cells from the boring form and crude extract from the crust form, which result in the formation of secondary aggregates, an absence of a membrane-bound aggregation receptor can be excluded. In other words, the cells from the boring form are competent to undergo both primary and secondary aggregation in v i m . while in vivo only the first principle is realized and ‘obviously sufficient to maintain the functional organization of the sponge. One task for the future will be the solution of the question of whether reaggregation of single cells from sponges can also be mediated exclusively by the secondary aggregation mechanism. To answer this question, tools are necessary to block primary aggregation selectively. These might hopefully be available after the elucidation of the molecular events, occurring during the primary aggregation process. In summary, in view of the presented data it seems to be justified to propose in this article for the first time a distinct aggregation principle for sponges, underlying primary aggregation, which is basically different from that which provokes secondary aggregation. The latter process is mediated by the soluble, highmolecular-weight aggregation factor (see next section). The existence of two aggregation principles, which also differ on a molecular level from each other, is not restricted to sponges but has been proven in higher eukaryotic systems: first, a Ca2 -dependent mechanism (Takeichi, 1977) and second, a Ca2 -independent mechanism (Brackenbury et al., 1977; Thiery et a l . , 1977; Takeichi et al., 1979; Urushihara et al., 1979; Urushihara and Takeichi, 1980). +

+

B. SPECIES SPECIFICITY Since the publication of Wilson (1907), the marine sponges are the classic model to investigate differential cellular affinities. It is now well documented that sponge cells reaggregate species specifically (Wilson, 1907; Galtsoff, 1923, 1925; Moscona, 1963; Humphreys, 1963; Miiller and Zahn, 1973; Burger et al., 1978; Muller er af., 1980; Conrad et al., 1981), while cells from other models form tissue-specific complexes (Frazier and Glaser, 1979; No11 et al., 1979). However several other authors question the basic and general applicability of species-specific aggregation (Curtis, 1962, 1970a.b; MacLennan, 1970; Sara et al., 1966). The cause for the conflicting results published can be attributed to the different dissociation and reaggregation conditions used for the studies. Now, knowing that the Demospongiae are provided with two aggregation systems, we have to reconsider the conclusions drawn in earlier publications under the aspects of species specificity of primary and secondary aggregation. This operational distinction fails to some extent in those experiments using cells which heavily synthesize the soluble aggregation factor.

144

WERNER E. G. MULLER

In the experiments described by Curtis (1962, 1970a,b) single cells were used which had been freed from the soluble aggregation factor. This means that the author started with cells in his studies which had the potency to forni only primary aggregates. By measuring the collision efficiency of aggregating cells, as an experimental test for specific adhesion, Curtis (1970a,b) came to the clearcut conclusion that no species specificity of cell adhesion exists in sponges. For these experiments cells from the two species Haliclona occulata and H . ptznicea were used immediately after their chemical dissociation; the determinations were run for 1 hour. Under these conditions, Curtis has unequivocally studied the specificity of cell adhesion during primary aggregation and from his experiments one must conclude that this process is highly unspecific. The degree of unspecificity seems to be high because sponge cells even reaggregate with cells from the anthozoan Anemonia sulcata (Sara et al., 1966) during primary aggregation. In earlier experiments, Curtis ( 1962) monitored aggregation of chemically dissociated cells for a period of 20 hours covering the time necessary for secondary aggregation. Depending on the pairs of species used he could distinguish between the following four types of aggregates. First, complete separation of the reaggregates of each species occurs when cells of Microciona sanguinea, Suberites ficus, and H . panicea taken in pairs, are started reaggregating simultaneously. Second, chain aggregates: cells aggregate species specifically in small aggregates but each aggregate adheres to each other. This type is produced when, e.g., M . sanguinea cells (rapidly aggregating species) are added to S. ficus cells (slowly aggregating species) after the latter have started to reaggregate. 'Third, concentric coating: cells from one species aggregate around a core aggregate from a second species. M . sanguinea cells were found to form the coat if they are added to S. ficus cells when the latter are in the process of reaggregation. Fourth, intermingling: complete intermingling of cells results when M . sanguinea cells are added to S. ficus cells after the latter have completed reaggregation. From these experiments the author hypothesized that the species-specific separation is controlled by a mechanism involving a species-specific rate of aggregation. In view of the present knowledge the aggregates, analyzed in this study, are the results of the two consecutive aggregation mechanisms. This would mean that the observed partially up to totally species-specific separation of cell clumps is due to the soluble aggregation factor which could be synthesized during the observation period of 20 hours. This assumption has been supported in an elegant way by John et al. (1971). Using chemically dissociated cells of Ophlitaspongia seriara and H . panicea they found that monospecific aggregation in bispecific cell mixtures is initiated only when archaeocytes aggregate with homologous mucoid cells. The authors propose that these two cell types interact "in some way" during the reaggregation process of 16 hours to confer specificity on the aggregate. Heterologous cells are assumed to recognize this transfer as nonself. Again, also from these experiments it is concluded that aggregation is initially a random process. The formation of species-specific aggregates is the result of a

CELL MEMBRANES IN SPONGES

I45

secondary process which depends upon the metabolic activity of the cells. A first, strong evidence for the mechanism causing the species-specific sorting-out came from Sara (1956). He found that during an observation period of 48 hours only cells from sponges belonging to the Demospongiae have the ability to form species-specific aggregates; in contrast, bispecific cell mixtures of Leucosolenia botryoides and Leucosolenia complicutu (belonging to the Calcareae) are able to form only bispecific aggregates. The lack of a species-specific sorting out mechanism in cell suspensions of the latter species might be attributed to the nonexistence of a secondary cellular recognition system as mentioned earlier. The results reported in this section are in good agreement with the postulate of Moscona (1956, 1962a, 1963), implying that reaggregation occurs in two stages: initial nonspecific contacts, mediated by the primary cellular recognition system, followed by sorting out species specifically. The latter process, mediated by the secondary cellular recognition system, has been elucidated in sponges in a representative and exemplary way.

VI. Secondary Cellular Recognition A.

SECONDARY AGGREGATION

With the exception of the Calcareae, all marine sponge species are provided in addition to the primary cellular recognition system with a secondary cellular recognition mechanism (see previous section). This means that from these organisms an easily extractable intercellular factor can be separated. This material was first termed extracellular material (Moscona, 1962b) which presumably acted by bridging between cells; later it was named aggregation factor (Humphreys, I965a). The first direct evidence for the existence of a soluble aggregation factor came from the experiments of Galtsoff ( 1929). He succeeded in the isolation of a water-soluble substance which is responsible for agglutination of dissociated Renieru cinera cells. This result was later confirmed and extended by Moscona (1963) and Humphreys (1963). The purification of the aggregation factor turned out to be difficult (Humphreys, 1965a; Margoliash et al., 1965); this task was achieved in 1973 using the siliceous sponge G. cydonium (Miiller and Zahn, 1973) and the Cornacuspongida Microciona parthena (Henkart et a[. , 1973; Cauldwell et al.. 1973). These two aggregation factors were the first purified adhesion ligands ever described and are still the best characterized factors thus far investigated. The basic features of the Geodia and of the Microciona factor are identical; however, in view of their particular physical and chemical characteristics, they must be considered as evolutionary variants. The aggregation factor from M . prolifera (or parthenu) is obtained in a soluble form after treatment of the sponge tissue with CMF sea water (Humphreys, 1963; Moscona, 1963). From the crude supernatant, the purified aggregation factor is

I46

WERNER E. G . MULLER

obtained by differential centrifugation and gel chromatography (Henkart et al., 1973) or by fractionated calcium precipitation followed by gel chromatography (Jumblatt et al., 1980). The total purification of the factor from the whole sponge is several 1000-fold. The active aggregation factor has an intrinsic viscosity of about 500 cm3/g, a sedimentation constant of around 70 S, and an approximate molecular weight of 20 X lo6 (Cauldwell et al., 1973). Chemically, the factor is a glycoprotein composed of 47.1 % protein, 26.0% neutral hexose, 10.4%iironic acid, 5.8% galactosamine, and 5.5% glucosamine; the buoyant density is 1 46 gl cm3. The amino acid composition shows a high dicarboxylic amino acid content, relatively high values of glycine, alanine, and other aliphatic amino acids, and the presence of cysteine. This overall composition is typical for structural glycoproteins from sponges and different from collagen and elastin (Junqua et al., 1975). In negatively stained electron micrographs the Microciona factor appears as “sunbursts” with a circular center 800 A in diameter and 15 or 16 arms 1100 A in length radiating from the circle; the diameter of the fibers is approximately 45 A (Henkart et al., 1973; Humphreys et al., 1977). The Microciona factor is resistant to 1% sodium dodecyl sulfate and to hyaluronidase (Henkart et al., 1973; Cauldwell et a l ., 1973). However the aggregation factor is degraded in the presence of I mM dithiothreitol and 0.5 mg/ ml proteinase K to two glycoproteinaceous chains of 65,000 and 6000 daltons (Humphreys et al., 1977). Simultaneous heat, urea, and EDTA treatment at elevated pH produces fragments of about 200,000 daltons (Burger et al., 198 I). One approach has been published to determine that site of the Microciona aggregation factor which might be directly involved in mediating the secondary aggregation process (Turner and Burger, 1973a). Basing on chemical, enzymic, and competition experiments, the authors came to the conclusion that glucuronic acid residues present in the aggregation factor are important for its activity. The interrelation between the Microciona factor and Ca2 is also of functional importance. The aggregation factor is present in its monomeric form in the presence of 1 mM Ca2+. At concentrations of 20 mM Ca2+ and above, the factor converts reversibly to the gel state (Henkart et al., 1973). At 10 mM Ca2 (the concentration which is present in sea water) a gel is formed only at high factor concentrations. The ability to form a gel is destroyed after denaturation of the factor. It is remarkable that the aggregation factor is degraded easily in the presence of 1 mM EDTA or I mM EGTA, indicating that Ca2+ is necessary for the stability of the factor (Cauldwell et al., 1973); when Ca2+ is removed from the factor, its molecular weight gradually drops from 20 X lo6 to 2 X lo5. The aggregation factor apparently shows two types of Ca2 binding sites: one, with a high affinity and a number of 1000-1300 sitedfactor molecule and another class of weaker binding sites numbering several more thousand (Cauldwell et al., 1973). The aggregation factor is solubilized from G . cydoniurn tissue by use of CMF sea water, containing 20 mM EDTA (Miiller and Zahn, 1973). Omission of +

+

+

CELL MEMBRANES IN SPONGES

I47

EDTA results in a lower yield of solubilized factor. The purification procedure employs ammonium sulfate fractionation, acid precipitation, Sepharose 4B gel filtration and carboxymethyl-cellulose (Miiller and Zahn, 1973; Miiller et al., 1974). The purified material was characterized by sucrose gradient equilibrium centrifugation (density = I .3 1 g/cm3) and zonal density gradient sedimentation (90 S ; Miiller et al., 1979b). The overall purification from the CMF-EDTA supernatant ( = crude extract) is about 500-fold. The factor, with respect to its biological activity, is stable for more than 8 years, if stored in CMF-EDTA sea water at -20°C. In contrast to the Microciona factor, the Geodiu aggregation factor consists of more than 74% protein (Zahn er al., 1976). Based on the amino acid composition, the aggregation factor has to be classified to the structural gl ycoproteins. Examinations of the Geodia factor with the electron microscope reveal some distinct differences from the Microciona factor. While up to now, the sunburst structure could be visualized only in the purified aggregation factor preparation from Microcionu (Henkart et al., 1973), it is now known for the Geodia factor that the sunbursts are only the core structure of a “native” ball-like particle (Miiller et al., 1974, 1978b). The native particles appear as spheres with a concave cup structure (Fig. IE). The center of the structure is electron dense after platinum shadowing. The circular contour length was determined to be 2300-4000 A and the diameter is 740-1260 A. After negative staining with uranyl acetate, honeycomb-like protrusions are seen. The core structure of the aggregation factor is obtained after treatment with Nonidet (Miiller et al., 1978b). The core structures (Fig. IF) appear as “sunbursts” with a circular center of circumference of 3530 t 85 A and 25 ? 8 radiating arms. The length of the filaments has been found to be 610 t 120 A. The thickness of both the central ring and the filaments is 230 170 A (Zahn er al., 1976). The Geodia factor is resistant to 100 mM EDTA, carbohydrate hydrolyzing enzymes, deoxyribonuclease, and ribonuclease (Miiller and Zahn, 1973). The activity of the aggregation factor is destroyed by proteases; among the exopeptidases tested, carboxypeptidase B exerts the strongest inhibitory influence. Similar results have also been obtained with a crude aggregation factor from Huliclona variabifis (Gasid and Galanti, 1966). It is important to note that in contrast to the M. prolifera aggregation factor (Turner and Burger, 1973a), both the Geodia and the Huliclona aggregation factor are not inactivated by P-glucuronidase. As with the Microciona factor, the Geodia particle is destroyed after incubation in CMF artificial sea water of pH 10. The high-molecular-weight Geodia aggregation factor is provided with the following three functional subunits. First, the low-molecular-weight aggregation factor (Miiller er ul., 1974). This molecule is obtained in a soluble form after treatment of the high-molecular-weight complex with nonionic detergents. It consists of 91% protein and 6% carbohydrate. The sedimentation constant was determined to be 2.7 S, corresponding to a molecular weight of approximately

*

148

WERNER E. G. MULLER

20,000. Digestion experiments with carboxypeptidase B revealed that L-lysine and L-arginine are the terminal amino acids. Interesting is the fact that after removal of these amino acids the aggregation promoting activity of the factor is lost. This finding is an indication that L-lysine and/or L-arginine are functional groups of the aggregation factor causing secondary aggregation. Second, the glucuronosyltransferase (Muller et a l . , 1978d). This transferase uses the deglucuronylated aggregation receptor (that molecule on the cell surface which interacts with the aggregation factor during reaggregation) from the same species as acceptor with the high affinity ( K , value) of 216 M ; the K , value with respect to UDP-glucuronic acid is I . 1 IM. Third, the galactosyltransferase (Muller et al., 1978d). This enzyme transfers UDP-galactose ( K , = 0.1 mM) with a K , of 0.28 nM preferentially to the antiaggregation receptor (a cell membrane bound molecule, interfering with the binding between the aggregation factor and the aggregation factor). These two enzymes could not be solubilized in an active state from the high-molecular-weight complex. Two attempts were undertaken to elucidate the origin of the high-molecularweight aggregation factor from Geodia (Muller et al., 1981b, 1982b). We studied the formation of the aggregation factor by electron microscopic methods. Cross sections through Geodia tissue revealed that particles with a diameter between 360 and 950 A can be detected in less than 5% of sponge cells. Two types of cells are distinguishable. First round-shaped cells which contain large nuclei, surrounded by two nuclear membranes; the cytoplasm is rich in mitochondria. From comparable electron microscopic studies these cells are classified as archaeocytes. Second, amoeboid cells with slender branching pseudopodes packed with granules, spherules, and excretory inclusions. These cells are usually termed “spherulous cells” and are likely to be mucoid cells. The aggregation factor is closely associated with the double nuclear membrane in the round-shaped cells while the factor is exclusively found in a free state in the cytoplasm of the amoeboid cells. We assume that the archaeocytes are “embryonic” cells which can differentiate into any other cell type. Therefore, we postulate that the aggregation factor is formed in the round-shaped cells, arid that it is stored in the mucoid cells. Then, in analogy to the synthesis of herpes viruses, we assume that the aggregation factor is synthesized between the inner and outer nuclear membrane. From the photomicrographs, we assume that the factor originates from the inner lamellae of the nuclear double membrane. The particles are round-shaped with a diameter of 360-480 A and are filled with electron-dense material; in no case could a translucent or a semitranslucent center be detected. At present, we suggest that the aggregation factor is directly released into the perinculear space. This assumption is supported by the finding that the aggregation factor is stored in the cytoplasm without any visible contact to membrane structures, and is not released directly into the extracellular space. Before the aggregation factor is released by the cell, it associates with the cell

CELL MEMBRANES IN SPONGES

149

membrane and is then set free by a process of “reverse phagocytosis,” exocytosis. During this process the aggregation complex reaches its ultimate size (diameter 900-1 250 A). These data suggest that the high-molecular-weight aggregation factor is assembled stepwise as follows: first, in the nuclear double layer, second, during the transport into the cytoplasma, and third, during the release into the extracellular space. As the extracellular aggregation factor, the intracellular factor (entire particle) appears in its native form as a round-shaped sphere with a diameter of 900 A. The sedimentation coefficient was determined by sucrose gradient centrifugation and a value of 42 S was obtained. This figure is considerably lower than the constant for the extracellular factor (90 S). The intracellular entire structure could be disintegrated by NonideVEDTA yielding a “low-molecular-weight” protein species and a high-molecular-weight core. The core structures of the intracellular particles are not uniform; besides the “sunburst” structure (approximately 10% of the total number of particles), linear backbone structures (25%), and circles lacking the radiating arms (65%)can be observed. The contour length of the plain circles is identical to the length of the circular center of the sunburst structures. This suggests further that the aggregation factor is assembled stepwise intracellularly. By comparing the protein banding patterns obtained after separation of the soluble protein species from the extracellular and intracellular aggregation factors a close homology between the intracellular and the extracellular complex can be deduced. With one exception, all other protein bands are present in both factor preparations. The protein, being absent in the intracellular factor, was identified as the low-molecular-weight aggregation factor. This indicates that the aggregation factor becomes associated with the high-molecular-weight complex during its release from the intracellular to the extracellular space. Large fibrous complexes could be visualized electron microscopically in soluble, high-molecular-weight fractions (= aggregation factor) from such sponge species whose cells can form secondary aggregates in the presence of this fraction. Basically two different types of complexes are distinguished (Humphreys er al., 1977; Muller er ul., 1982d): first, circular structures and second, rod-like structures. The Geodiu and the Microciona aggregation factors belong to the first group. Up to now from more than 15 different marine sponges circular polymers have been isolated and visualized. The common feature of these structures is the circular backbone of a circumference ranging from 2500 A (M. prolifera) to 5000 A (Verongia aerophoba). Only from two sponge species (V. aerophoba and H . cornrnunis) plain circles have been isolated, all the other structures are provided with 8-25 radiating arms. Rod-like structures were detected in aggregation factor preparations from 6 species. As one example the complex from C. celara (crust form) is shown in Fig. 2F. The length of the backbone fiber varies between 2800 and 4800 A. Some of the rod-like structures are without (e.g., Terhyu lyncurium), others with side arms (e.g., H . occulutu, C . celura). It is

150

WERNER E. C. MULLER

interesting that these two structural forms of the sponge aggregation factor (circles and rods) have been highly conserved during an evolution period of 1000 million years. From the cellular and biochemical studies the existence of a specific molecule (= aggregation factor), involved in secondary aggregation of sponge cells, seems to be well documented. As a further rigorous approach to prove the occurrence of this external ligand, specific antibodies against the purified Geodia aggregation factor were produced by Conrad et al. (1981). They have shown that this antibody specifically inhibits the secondary aggregation process of Geodia cells in the presence of the soluble aggregation factor. The specific recognition during the secondary aggregation process, mediated by the soluble extracellular aggregation factor, also requires the existence of a specific component on the cell surface. Such a two-component system for a specific surface guided reassociation of sponge cells was worked out for M . prolifera (Weinbaum and Burger, 1973). The authors succeeded in the removal of an aggregation receptor (“baseplate”) from the cell surface by hypotonic shock. This aggregation receptor has been partially purified by differential centrifugation and gel chromatography (Burger and Jumblatt, 1977). The receptor has an apparent molecular weight of about 50,000, is stable to 5 mM EDTA, freezing, lyophilization, and pHs of 3-12 but is inactivated by heat (60°C lor 10 minutes). Further chemical or physical properties of the Microciona aggregation receptor are not known. Special attention has been focused on the nature of the aggregation receptor-aggregation factor interaction rather than aggregation receptor-cell interaction (Burger et al., 1978; Jumblatt et al., 1980). To demonstrate direct interaction between the Microciona aggregation receptor and the Microciona aggregation factor, the receptor was coupled to agarose beads. These precoated beads aggregated very rapidly in the presence of the homologous aggregation factor. Using this interesting system the authors established that the binding of the aggregation factor to the aggregation receptor is Ca2 -independent. In a second step, which occurs in the presence of C a 2 + , linkages between two aggregation factor molecules on adjacent cells are formed. This working model implies the existence of functionally univalent aggregation factor molecules provided with at least two different binding sites; one site carrying presumably specific saccharide residues (glucuronic acid) which recognize the aggregation receptor and a further one for a second aggregation factor molecule via (:a2+. However, in a recent contribution by Leith (1979), using mechanically dissociated cells from M . prolifera and Hafichondria bowerbanki, still carrying the aggregation factor on the cell surface, the conclusion was drawn that the intercellular aggregation factor is a symmetrical unit. Cell adhesion in G. cydoniurn also involves a bicomponent system consisting of an aggregation factor, the “bridging” molecule, and the aggregation receptor. +

CELL MEMBRANES IN SPONGES

15 1

The Geodia receptor was solubilized by extraction of cell membranes with trichloroacetic acid (MacLennan, 1970) and purified extensively by acetone fractionation, ion-exchange chromatography, gel chromatography, and cesium chloride isopycnic centrifugation (Muller et al., 1976a). The aggregation receptor is a glycoprotein (81% neutral carbohydrate and 7.5% protein) and has a molecular weight of 15,300 or 18,000, depending on the determination method used (sucrose sedimentation or gel filtration). The interaction between the aggregation receptor and the aggregation factor has been studied by competition experiments. These revealed that after preincubation of the aggregation factor with the solubilized aggregation receptor and a subsequent incubation with cells, containing cell membrane bound aggregation receptor molecules, only small cell aggregates are formed. From this indirect evidence we propose a working model, which is shown in Fig. 4.This model implies an at least bivalent aggregation factor; this assumption must be substantiated by future binding studies, similar to those performed in the Microciona model. No detailed knowledge about the role of Ca2+ ions and about the interaction between the aggregation receptor and the cell membrane is available in the Geodia system. The aggregation receptor was found to be only loosly bound to the cell surface; it can be removed by EDTA or by trypsin digestion (Muller et al., 1976a). After treatment with these agents, the cells have lost 70% of their

Cell

FIG. 4 Working model for the aggregation factor (AFtaggregation receptor (R) mediated cell recognition in Geodia cydonium. The aggregation factor bridges the space between two cells either in its complex form bound to the sunburst particles (abbreviated as C; sedimentation,coefficient: 90 S) (shown above) or in its particle-free form (sedimentation coefficient: 2.7 S) (below). The role of CA2+ ions (Ca) is not yet understood. (From Muller et al.. 1978h.)

I52

WERNER E. G. MULLER

capacity to form secondary aggregates in the presence of the aggregation factor. Special efforts were made to determine the binding site of the aggregation receptor for the aggregation factor (Vaith et al., 1979a). The chemical analysis of the receptor revealed that the three sugars, D-galactose, D-glucose, and D-glucuronic acid, account for about 85% of the total carbohydrate. Digestion experiments with P-glucuronidase using dissociated Geodia cells lacking the aggregation factor but containing cell surface bound aggregation receptor molecules, revealed that these cells have lost most of their aggregation potency. The assumption that D-glucuronic acid moieties of the aggregation receptor are recognition sites for the aggregation factor was confirmed by analytical studies (Muller et al., 1979a) showing that D-glucuronic acid is terminally located and, after its removal, the aggregation receptor is biologically inactive. It is interesting to find that the basic polymer poly-L-lysine strongly reacts with the aggregation receptor (Vaith et al., 1979a), especially in light of the above mentioned result suggesting lysine to be one functional group of the aggregation factor. Polylysine itself does not cause secondary aggregation (Miiller et al., 1978d). Therefore we are intrigued by the hypothesis that lysine (bound to the aggregation factor) and glucuronic acid (functional unit on the aggregation receptor) are major components in factor-receptor recognition; this model is of course prone to revision if further information becomes available. After binding to the aggregation receptor, the aggregation factor is not internalized, but remains on the cell surface from which it can be recovered in an active state (Miiller et a / . , 1978a). Further information that glucuronic acid moieties of the Geodia aggregation receptor have an important function in the factor-mediated cell recognition came from the observation that a cell surface bound P-glucuronidase controls the functional activity of this receptor. The P-glucuronidase is cell membrane associated (Miiller et a f . , 1979a) and has a molecular weight of about 25,000. The pH optimum is around 5.0. The enzyme activity is independent on the presence of Mg2 , Ca2 , and EDTA. For optimal enzyme activity 200 mM NaCl has to be added to the reaction. This enzyme was shown to hydrolyze terminal Dglucuronic acid units from the aggregation receptor. After enzymic deglucuronylation the aggregation was found to be biologically inactive. The deglucuronylated aggregation receptor can, under certain physiological conditions, be enzymatically reglucuronylated by the aggregation factor-linked glucuronyltransferase. Thus in the sponge (G. cydoniurn) model, used for this study, the glucuronyltransferase could be clearly localized in the extracellular space. In addition, and in contrast to the Roseman theory (Roseman, 1970), the aggregation factor can be separated from glycosyltransferases which indicates that the glycosyltransferases might only modulate the aggregation potency of the cells, but that they do not ultimately provoke adhesiveness between two cells. The latter process is, at least in the Geodia system, a result of an interaction +

+

CELL MEMBRANES IN SPONGES

153

between the aggregation factor present in the intracellular space and the aggregation receptor bound to the cell surface. From these experimental findings we postulate a working hypothesis to explain cell aggregation and cell separation on the level of an interaction between glycosidases and glucuronidases in the following series (Miiller et af., 1979a; Fig. 5): (1) activation of the aggregation receptor by its enzymic glucuronylation; (2) adhesive recognition of two cells, mediated by the aggregation factor and the glucuronylated aggregation receptor; (3) inactivation of the aggregation receptor by its deglucuronylation with the membrane-associated P-glucuronidase; and (4)

UOP-0

@ Inoctwe

AR

AR i oggrrgolion rrcrplor AF = aggrrplion

w w l , AR

bclor

CPP i c~rculor prolrid particlr

Glu A 3 gluuronu ocid

0 = glucuoruc

acid

FIG. 5 . Proposed molecular mechanisms for cell aggregation and cell separation (in the Geodia cydonium system) via activation or inactivation of the aggregation receptor. CPP, Circular proteid

particle (sunburst structure) carrying the aggregation factor. (From Miiller and Miiller, 1980.)

154

WERNER E. G . MULLER

cell separation due to the loss of the recognition site (glucuronic acid) of the aggregation receptor for the aggregation factor. Steps (3) and (4) are consequences of an “activation” of the membrane-associated P-glucuronidase, which is inactive under the pH and ion conditions of sea water. No data are available which give direct experimental evidence for a pH or ion change of the sponge cell membrane during the process of cell separation. However, from previous studies (Miiller et al., 1978d), it is known that the aggregation factor from Geodia alters the cell surface charge after its binding to homologous cells. Due to the fact that an alteration of the cell surface charge is also caused by a pH change at the cell surface, it is quite likely that pH fluctuations govern the activity of both the glycosidase and the glycosyltransferase. At present we assume that the described enzymic activation and inactivation processes of the aggregation receptor might control the initial events of cell aggregation and in addition play a role in the subsequent event, the “sorting out,” which occurs during the formation of a functional sponge from the secondary aggregates. The model of the adhesive mechanism first developed for G . cydonium (Miiller et al., 1979a) implies that first, two nonenzymic molecules (aggregation factor and aggregation receptor) mediate cell-cell recognition, and second, the functional activities of these molecules are controlled by cell surface or intercellular glycosidases and glycosyltransferases. Increasing evidence has accumulated that molecular interaction in adhesion of cells from higher organisms (e.g., embryonic chick neural retina cells, Balsarno and Lilien, 1980; hamster fibroblasts, Rauvala and Hakomori, 1981) is due to the same or to a similar mechanism. In vertebrate cells the mobility of the cell membrane receptors, involved in cell adhesion, seems to be an important parameter for their function (Lilien et al., 1978). However, in sponges, no evidence is available indicating that receptor mobility is a prerequisite for secondary aggregation (Jumblatt et al., 1977). Vertebrate cells are provided with both intrinsic membrane factors (e.g., CAM protein, Brackenbury et al., 1977; cognin, Hausman and Moscona, 1975) and extrinsic “bridging” factors (galaptins, Nowak et al., 1976). In addition, most of the vertebrate cells have the cell surface glycoprotein, fibronectin (Yamada and Olden, 1978), which is thought to be involved in cell-cell adhesion as well. Two groups have recently studied the possible role of fibronectin in sponge cells. Labat-Robert et al. (1979), who used the sponge T . lyncurium, detected fibronectin on the cell surface by indirect immunofluorescence methods. In addition, antifibronectin serum was found to inhibit reaggregation of dissociated cells. In contrast to this working group, we were not able (Conrad et a f . ,1982) to isolated fibronectin from G . cydonium. Also immunological data failed to prove the existence of fibronectin. Furthermore human fibronectin and antifibronectin serum was shown to exert no influence on adhesion of Geodia cells both in the absence and in the presence of the soluble aggregation factor.

CELL MEMBRANES IN SPONGES

155

B. SPECIES SPECIFICITY Since early biochemical studies (Moscona, 1963; Humphreys, 1963) it is well established that species-specific reaggregation of sponge cells is due to action of the species-specific aggregation factor, provided that the aggregation factor is sufficiently pure (Muller et al.. 1980). At present, two categories of specificity are distinguished (Humphreys, 1970b): first, partial species specificity and second, complete species specificity. Partially species specific are those aggregation factors which make heterologous cells adhere, but the aggregates are less compact than those produced by equal amounts of factor on homologous cells. These aggregates do not reconstitute to functional sponges. Cells from combinations of the two sponge species M . prolifera and H . oculata exhibit incomplete specificity. First it was demonstrated (Humphreys, 1970b) that the aggregation factors from the two species have no effect on the cells of the opposite species. Even in mixed cell suspensions the aggregation factor causes aggregation of the cells from the homologous species only. In a later study (Turner and Burger, 1973b) it was demonstrated that this species selectivity is due to only a quantitative and not a qualitative difference in the reactivity of the cell surface with the aggregation factor (Turner and Burger, 1973b). Microciona cells were reported to aggregate with both aggregation factor preparations although the Haliclona aggregation factor was less than 50% as effective as the Microciona aggregation factor on a protein basis. Using a final protein concentration of 50 pg/ml, both factors mediate aggregation of Microciona cells to the same extent. However, only Microciona cells, aggregated in the presence of the homologous aggregation factor, have the capacity to reform a complete functional sponge (Humphreys, 1970a). The aggregation factor preparations used for these studies were far from pure. They were obtained by differential centrifugation (Humphreys, 1963; Turner and Burger, 1973a) and presumably contained lectins. These carbohydrate binding molecules were reported (Muller et af., 1980) to cause cell agglutination in heterologous species. Therefore the results should be reconsidered in the light of a possible interference of the aggregation factor mediated aggregation with the lectin caused cell agglutination. More recent binding experiments with purified and radioiodinated Microciona aggregation factor already revealed (Jumblatt et al., 1980) that it is bound to Microciona cells at a 15-fold higher amount than to H . panicea cells or to C . celata cells. The crucial experiment needed to understand the specificity of this aggregation factor in molecular terms is still lacking: binding of homologous aggregation factor and subsequent determination of the number of the cell surface localized “binding sites” for the heterologous factor. Complete species-specific aggregation factors are those which recognize immediately heterologous cells and refuse to aggregate with them. The aggregates

WERNER E. G.MULLER

156

formed reconstitute to functional sponges (Humphreys, 1970b; Van de Vyver, 1978). These two prerequisites were fulfilled in the combination of G . cydonium and Suberires domuncula (Muller et al., 1978f,g). From these two sponge species both the aggregation factors and the aggregation receptors were isolated and purified. In the absence of any factor, cells from these species form only primary aggregates with a diameter less than 100 Frn (Fig. 6 ) . After addition of the factor to homologous cells, they respond and form secondary aggregates. The reciprocal experiments, also shown in Fig. 6, revealed no response of Geodia cells to Suberires aggregation factor and formation of cell clumps if Suberites cells were incubated with Geodia aggregation factor. However the latter effect is due to lectin contaminations present in the Geodia factor preparation used for this study. After removal of the lectin by hemabsorption (Conrad et al., 1981) the Geodia factor reacts species specifically even in a 10-fold excess of aggregation factor if based on protein content. This means that these two sponge factors recognize even in the initial phase heterologous cells and refuse to aggregate with them.

m

r

1

-.-.-.

I

0 L

al

0

L

I

0

-

:

60

.

120

Aggrega!ion time (min) FIG. 6. Time course of Geodiu cvdoriium and Suberites domuncula aggregation in the absence or presence of the homologous or heterologous aggregation factor. The aggregation assay war conducted at 20°C in 3 ml total volume of sea water containing 7 . 5 X lo7 cells. Where indicated, Ceodia aggregation factor (4 X 10" AU) or Suberites aggregation factor (I0 AU) was added to the a!+says. (From Muller et a / . . 1978a.)

CELL MEMBRANES IN SPONGES

157

With this knowledge it was possible to prove that the species-specific cell aggregation in these two sponge systems (Geodia and Suberites) is basically due to an interaction between the aggregation factor and the aggregation receptor (Muller et al., 1978g). We succeeded in obtaining aggregation receptor depleted cells by a combined treatment with trypsin and hypotonic solutions; these cells have lost the ability to form secondary aggregates even in the presence of the homologous aggregation factor (Fig. 7; assay I ) . During the procedure the cells do not lose their viability. However after preincubation of Geodia cells with Geodia aggregation receptor and after subsequent incubation with homologous factor, the cells respond and form secondary aggregates (Fig. 7; assay 4).These two experiments indicate that the Geodia cells can be reversibly charged with homologous receptor. It is not surprising that Geodia cells, charged with Geodia aggregation receptor, do not form secondary aggregates in the presence of the soluble, species-specific Suberites aggregation factor (Fig. 7; assay 3). In the crucial experiment (Fig. 7; assay 2), receptor-depleted Geodia cells were preincubated with Suberites aggregation receptor and subsequently incubated with Suberites aggregation factor. After this procedure, the Geodia cells respond with the heterologous Suberites aggregation factor. This means that Geodia cells which have been charged with Suberites aggregation receptor behave like Sub-

-E 2

-

I

VI

aJ

0

F ol

8 w-

0

0

a Aggregation time( min I FIG.7. Interaction of the aggregation receptor from Suberites domuncula with Geodia cvdonium cells in the presence of the heterologous or the homologous aggregation factor. For these experiments, aggregation receptor-depleted Geodia cells were used. The assays (see legend to Fig. 6) were assay composed as follows: assay I , Geodia cells and Suberites aggregation factor (50 AU) (O---O); 2, Geodia cells, Suberites aggregation receptor (400receptor units = RU; Miiller et al.. 1978g) and Suherires aggregation factor (50 AU) (U assay ) 3, ; Geodia cells, Geodia aggregation receptor (4 RU) and Suherites aggregation factor (50 AU) ( X . . . . X ) ; assay 4, Geodia cells, Geodia aggregaIf aggregation tion receptor (4 RU). and Geodia aggregation factor ( 2 X 104 AU) (A--.-A). receptor was added to the assays, this component was first preincubated (I0 minutes; 20°C) with the cells and subsequently the aggregation factor was added to the suspension. (From Miiller et a / . , 1978g.)

158

WERNER E. G. MULLER

erires cells in the presence of Suberires aggregation factor. From this study it seems to be obvious that species-specific aggregation is due to a qualitative difference in the reactivity of the aggregation factor with the aggregation receptor. However with respect to binding, no pronounced specificity exists between aggregation receptors and sponge cell membranes in heterologous systems. The possibility to transfer aggregation receptor molecules from one species to another is not restricted to the sponge system described, but exists also in other systems, e.g., in sea urchins (Noll et al., 1979). In an elegant study McClay (1974) demonstrated that a radioactively labeled aggregation factor from Tedania ignis preferentially binds to homologous cells, while only very little binds to heterologous cells (Haliclona viridis and Hornaxinella rudis). Based on the additional fact that the Tedania aggregation factor mediates aggregation only in the homologous system and is inactive toward cells from Halifona and Homaxinelfa, he concludes that Tedania cells contain specific receptors for the homologous aggregation factor. From a further pair of sponges (Thoosa istriacu and Thoosa mollis) completely species-specific aggregation factors have been isolated (Muller er al., 1979e). C. FORMATION OF A RECONSTITUTED ORGANISM Studies to understand morphogenesis of sponge cell aggregates to reconstituted functional organisms have been performed both with fresh water and marine sponges, although mainly on cellular level. Huxley (1912) already described the basic cellular events leading from cell aggregates to functional individuals. After mechanical dissociation cells from the marine sponge S. raphanus form aggregates in which the cells are arranged at random. In a second stage the cells sort out into their respective “categories” and occupy their proper positions. After 5 days the cells which will form the dermal layer have migrated to the exterior, and are united to form a continuous epithelium. After about 10 days production of spicules begins. This time sequence has been repeatedly described for other marine sponges as well, e.g., for M .prolifera (Hurnphreys, 1963) and for G . cydonium (Muller and Zahn, 1973). The first studies which described the reorganization of aggregates from fresh water sponges to functional organisms were by Muller (191 1) and Brien (1936). Using E. fluviatilis, Van de Vyver and colleagues (Van de Vyver and Buscema, 1977; De Sutter and Van de Vyver, 1977, 1979; Buscema and Van de Vyver, 1978) studied the function of the three cell types, pinacocytes, choanocytes. and archaeocytes, during the reconstitution process. Pinacoderms and choanocyte chambers were observed to be reconstituted by a progressive cell gathering. A typical pinacoderm surrounds the aggregates after 6 hours. Choanocyte chambers, which are small and irregular in young aggregates, progressively grow by adjunction of other choanocytes; they reach an almost normal size also after 6

CELL MEMBRANES IN SPONGES

159

hours. During the reconstitution period, archaeocytes actively phagocyte cellular debris, dying, and vital cells. It is noteworthy that only aggregates which contain archaeocytes undergo a process of reconstitution, indicating that these cells are totipotent with respect to their differentiation potential (Buscema et a/., 1980). Reorganization of functional sponges from cell aggregates stipulates that the cell surfaces are provided with receptors and factors which allow a differential cell movement in a nondirected and directed way. While the molecular mechanisms underlying random cell movement are discussed in the following section, nothing is known about the nature of the target orientated cell movement. It is still not known whether this type of movement is guided by gradients of small or macromolecular morphogenetic substances, or by the extracellular skeleton of proteinaceous and glycoproteinaceous nature. D. INFLUENCE ON CELLMETABOLISM Dissociated cells from species belonging to the Demospongiae form in the absence of the aggregation factor only primary aggregates, which do not have any morphogenetic potential and die after several hours. Only secondary aggregates can reconstitute to functional sponges. Because all evidence available indicates that there is only one molecule, the aggregation factor which initiates secondary aggregation, it is reasonable to assume that this aggregation factor exhibits a second function in the aggregates, that is the stimulation of cell metabolism. The only recognition molecule found to interact with aggregation factor is the aggregation receptor. Because of its relatively low molecular weight (18,000 for the G . cydoniurn receptor and 50,000 for the M. prolijera molecule; Muller et a / ., 1976a; Weinbaum and Burger, 1973) the aggregation receptor may belong to the peripheral membrane proteins. Assuming that this macromolecule functions also as a receptor for the “activating” stimuli emanating from the aggregation factor, we have to ask for integral membrane proteins or glycolipids which transmit the signal from the outer to the inner cell membrane. This processing system like the subsequent ones is not known. What we do know is that a series of metabolic systems exist which respond to the binding of the aggregation factor to the cell membrane. At present, we have no information as to which of the discussed response systems is the central one. The secondary aggregation process is independent on temperature ( 5 or 22°C) (Humphreys, 1970a) and protein synthesis (Humphreys, 1965b) and can even proceed with glutaraldehyde fixed cells (Jumblatt e t a / ., 1977). Alterations of the metabolic activities in cells, assembled in secondary aggregates, occur only in viable cells and under physiological temperatures ( 15-20°C). As one parameter for the overall activation of cell metabolism, the oxygen uptake in secondary aggregates was measured (Muller and Zahn, 1973). The aggregation factor mediated cell-cell contact causes an almost immediate in-

160

WERNER E. G. MULLER

crease of oxygen uptake of the cells. Two hours after addition of the factor to the cells the oxygen uptake intensifies from 31.2 to 37.8 pl O2/hour/lO7cells. As one main parameter for cell differentiation and cell proliferation, occurring in secondary aggregates (Humphreys, 1963), the alterations of programmed syntheses in secondary aggregates as a function of time were determined (Muller et al., 1976b, 1978a). Isolated G. cydonium cells as well as primary aggregates have no capacity to proliferate. However, after addition of the soluble aggregation factor to these cells a high mitotic activity is observed 8 hours after beginning of the secondary aggregation process. The increase of the mitotic activity is a result of an activation of DNA, RNA, and protein synthesis. Isolated Geodia cells do not show any change in the rate of macromolecular syntheses (as measured by precursor incorporation) in assays containing no aggregation factor. In response to the binding of the aggregation factor to the cells, DNA, RNA, and protein syntheses already increase 4 hours after the beginning of the secondary aggregation phase. The increase of the mitotic activity of the cells in these aggregates can be blocked by inhibitors of DNA, RNA, and protein synthesis. Also interesting is the finding that the DNA-replicating enzyme, DNA polymerase a, is virtually absent in primary aggregates, but is found in a high specific activity in secondary aggregates (Muller er al., 1978a). These findings may indicate that single cells or cells in the primary aggregates are in the Go (or G , ) phase and enter the proliferating pool via the G, phase after addition of the aggregation factor (Fig. 8). Great efforts are currently underway to elucidate the metabolic processes linking the events which occur during interactions of extracellular molecules

I

I

I

aggregation factor

transmitt .................

omc

I..

I

II I

CELL MEMBRANES IN SPONGES

161

with cell surface receptors and the metabolic machinery of the cytoplasm and nucleus. Understanding this transmission machinery is a basic prerequisite for understanding of cellular differentiation and function, e.g., understanding the biochemical processes during immune response and growth control. As a first parameter in the G. cydonium system the alterations on the level of cyclic AMP and cyclic GMP metabolism have been determined (Muller et al., 1978h). It was found that the cyclic AMP level decreases dramatically immediately before the beginning of DNA synthesis, while the cyclic GMP level increases, resulting in a decrease of the molar ratio of cyclic AMP to cyclic GMP from 3.14 to 0.15. From other biological models it is known (Rudland et al., 1974) that such an alteration always occurs at the beginning of S phase. In order to elucidate the factors which might cause these changes of cyclic nucleotide levels, both the activities of adenylate cyclase and guanylate cyclase, and the properties of the cyclic nucleotide receptors have been investigated (Muller e? al., 19788). However, no significant changes in response to the binding of the aggregation factor to the cell membrane were measured. Further experiments should now determine the role of phosphodiesterases and of Ca2 during the initial events of secondary aggregation. The question must be answered whether the Geodia aggregation receptor has the capability to translocate Ca2+ across the cell membrane, e.g., as a Ca2+ ionophore. +

VII. Cell Movement and “Sorting Out” The cytological aspect of cell movement, occurring during primary and secondary aggregation, and the subsequent reconstitution to a functional sponge have been described in the previous sections. It is conceivable that especially during the reconstitution process the cells must be provided with the following abilities (Curtis, 1973a): contact inhibition of movement when two locomoting cells make contact, chemotactic orientation, and oriented movement on a substrate. Using a more general terminology, those cell movements observed in an aggregate which result in a reconstituted organism may be termed “sorting out” movement (Curtis, 1967). For several systems it has been established that sponges consist of cells exhibiting different aggregation properties. The differences in their aggregative capacities can be either cell type specific (De Sutter and Van de Vyver, 1977; Leith, 1979) or transitory (Muller et af., 1979e). In an approach to elucidate the existing differences on the molecular level (Muller e? al., 1979e), the total cell population of C . cydonium was separated into aggregation-deficient cells and aggregation-susceptible cells. Aggregation-susceptible cells form secondary aggregates in the presence of the soluble aggregation factor (Fig. 9, assay I), while aggregation-deficient cells do not respond to the aggregation factor (Fig. 9, assay

162

WERNER E. G. MULLER

2). These two classes of cells are obviously interrelated; by screening experiments it was discovered that the deficient cells can be converted to susceptible cells after preincubation with P-galactosidase (Fig. 9, assay 4).After this result we assumed that a hitherto unknown macromolecule exists on the cell surface of aggregation-deficient cells which prevents an aggregation factor mediated cell aggregation. This molecule was subsequently isolated and purified; it was termed antiaggregation receptor. The antiaggregation receptor from G. cydoniurn was solubilized from membrane fractions with lithium diiodosalicylate and was obtained in a homogeneous form after heat treatment, acetone extraction, and differential affinity gel chromatography (Miiller et al., 1979e). This molecule was characterized as a glycoprotein (54% neutral carbohydrate; 3.2% hexuronic acid; 7.9% lipid; 17.8% protein); its molecular weight is approximately 180,000 and its buoyant density is 1.43 g/cm3. Enzyme digestion experiments revealed that D-galactose is the major component of the functional site of the antiaggregation receptor. This -000,2

Aggregation time (min)

Fic. 9. Interrelation between aggregation-deficient and aggregation-susceptible Gevdiu cydvnium cells. The assay conditions are as described in the legend to Fig. 6. Assay 1 contained aggregation-susceptible cells and aggregation factor (5.7 x 104 AU) ( G O ) ;assay 2 contained aggregation-deficient cells and aggregation factor (5.7 X 104 AU) ( X . . . . X ) ; assay 3 contained aggregation-susceptible cells previously preincubated with antiaggregation receptor (addition of 45 antiaggregation receptor units and incubated for 10 minutes at 27°C followed by washing procedures with sea water) and aggregation factor (5.7 x 104 AU) (A--.-A);assay 4 contained aggregationdeficient cells previously incubated with P-galactosidase (0.09 units of the enzyme from bovine liver; incubation for 120 minutes at 27°C followed by washing procedures with sea water) and aggregation factor (5.7 X 104 AU) (0-0). (From Miiller et al.. 1979c.)

CELL MEMBRANES IN SPONGES

163

macromolecule binds to the aggregation-susceptible cells and converts them to aggregation-deficient ones (Fig. 9, assay 3). These data led us to assume that the antiaggregation receptor determines the capability of the cells to react with the aggregation factor and thus controls the aggregation factor-mediated cell-cell aggregation. The soluble anti-aggregation receptor interferes with the process of secondary aggregation which is based on the complex formation between the aggregation factor and the aggregation receptor (Miiller et a l . , 1979e) (Fig. 10). The standard kinetics shows that in the presence of the aggregation factor, single cells form aggregates with a diameter of 900 Fm after incubation for 60 minutes. After 120 minutes the aggregates grow to clumps of 2200 pm. Addition of soluble antiaggregation receptor stops, after a lag phase of about 30 minutes, the aggregation process (arrow 1 in Fig. 10) and even causes a disaggregation to smaller aggregates; while in the absence of this macromolecule, the aggregates have a size of around 2200 pm (after 3 hours), the aggregates in the test assay supplemented with antiaggregation receptor 60 minutes after the beginning of the aggregation process show a diameter of only 350 Fm.The biological activity of the antiaggregation receptor, which is the inhibition as well as the destruction of the aggregates, can be abolished by addition of P-galactosidase (arrow 2 in Fig. 10). After a lag of 75 minutes, the diameter of the aggregates increases again and after 5 hours sizes are reached which are almost the same as those determined in the

lncubat ton t irne (hrs.)

FIG. 10. Influence of the antiaggregation receptor on the integrity of the secondary aggregates from Geodiu rvdonium. In the control assay ( 0 - 0 ) dissociated cells were incubated in the presence of the aggregation factor. In the test assay (0-0) soluble antiaggregation receptor was added (arrow I ) and incubated further for 75 minutes; 135 minutes after the beginning of the experiment (arrow 2) P-galactosidase was added. (From Miiller er al.. 1979c.)

164

WERNER E. G. MULLER

control assay. From these findings we conclude that Geodia cells are reversibly converted from the aggregation-deficient to the -susceptible state also under physiological conditions, basing on a controlled interaction of the aggregation factor with the antiaggregation receptor. The antiaggregation receptor from G. cydonium was found to bind to the cell surfaces and to the aggregation factor in the absence of Ca2 and Mg2 (Miiller et al., 1979e). To elucidate that the binding is restricted to the aggregation factor only, affinity column chromatographic studies were performed (Fig. 11). Using an antiaggregation receptor-Sepharose matrix (Fig. 11a), the aggregation factor was determined to be bound to it in Ca2+- and Mg2+-free sea water. The +

+

++ + ++

??t.

A

B

I ,b

0

FIG. 11. Elution pattern of Geodia cydonium aggregation factor using column chromatography with the following two column matrices: antiaggregation receptor coupled covalently to Sepharose (a) or aggregation receptor coupled to Sepharose (b). Antiaggregation receptor (15 mg) was coupled to 10 ml of Sepharose 4B with divinyl sulfone (Young and Leon, 1978);7.3 mg aggregation receptor was linked to 10 ml of CNBr-activated Sepharose 4B according to the instructions supplied by the manufacturer. Columns of 0.7 X 15 cm were prepared with these matrices and fractions of 0.7 ml were collected. Purified aggregation factor ( I .5 mi) (Miiller and Zahn,1973). containing 17.2 X lo5 AU (0.12 rng/ml), was applied to each column. (a) The antiaggregation receptor-Sepharose column was equilibrated with CMF sea water, loaded with the aggregation factor, and eluted with buffer A (100 mM phosphate, pH 6.5, 500 mM NaCI) and buffer B (buffer A supplemented with 100 mM lactose). (b) The aggregation receptor-Sepharose column was loaded with the same amount of aggregation factor and eluted with buffers A and B. The relative amount of sunburst particles, which have been visualized electron microscopically, is indicated: + few particles; + +, many particles.

CELL MEMBRANES IN SPONGES

165

aggregation factor can be eluted from the column using a pH 6.5 buffer, containing lactose (recovery: 78%). However using an aggregation receptor-Sepharose matrix, the aggregation factor did not bind in the absence of Ca2+ and Mg2+ (Fig. 1 1 b); it was recovered in the first eluate (yield: 92%). In a control study it was established that in the presence of Ca2 the aggregation factor binds to the immobilized aggregation receptor, indicating that the receptor activity was not destroyed during the coupling procedure. From these data we conclude that the antiaggregation receptor exclusively binds to the aggregation factor under conditions which are optimal for its activity in the cellular system. The complex formation between the antiaggregation receptor and the aggregation factor is relatively stabile; it is not destroyed during elution through Sepharose 6B. Binding studies were performed with the antiaggregation receptor and the aggregation factor, using concentration ratios of these two macromolecules which cause a 50% inhibition of the aggregation factor. The stoichiometric calculation revealed that under these conditions approximately two antiaggregation receptor molecules bind to one aggregation factor molecule. Theoretically we have two experimental alternatives to study the mechanism which controls the activity of the antiaggregation receptor. First, we may assume that the antiaggregation receptor remains during all stages in an active state; then we have to investigate the stage-dependent dynamics of this macromolecule. Second, we may hypothesize that the antiaggregation receptor stays on the cell membrane in close connection with the aggregation receptor (and consequently also in special connection with the aggregation factor) during the different transitional states of the cells; then we have to search for possible differences in the activity of the antiaggregation receptor. We have studied the latter possibility in greater detail. Based on the fact that the functionally active component of the Geodia antiaggregation receptor is a terminally located (3-linked D-galactose (Miiller et al., 1979e), Vaith et al. (1979b) screened for a D-galactose-specific lectin in Geodia. He succeeded in the detection and purification of a lectin from the extracellular material, which also contains the soluble aggregation factor. This lectin shows a high specificity for galactans which have P-linked D-galactose as terminal nonreducing sugar units. In a recent study (Bretting et al., 1981) some more physical and chemical characteristics of this lectin have been described. Because of its chemical composition ((3-linked D-galactose termini), it was not surprising that the lectin also precipitates with the antiaggregation receptor (Miiller et al., 1979d). Very interesting was the finding that this lectin has the potency to convert Geodia aggregation-deficient cells to aggregation-susceptible cells (Table 111). The different aggregation potencies of aggregation-susceptible cells and aggregation-deficient cells (Table 111; assays a, b and f, g) in the presence of the aggregation factor are lost if the aggregation-deficient cells are pretreated with lectin (assay i); in other words, the lectin, which by itself has no aggrega+

166

WERNER E. G. MULLER

TABLE 111 CONVERSION OF AGGREGATION-DEFICIENT CELLS TO AGCREGATtON-SUSCEPrlBLE CELLS BY WITH THE Geodia cydonium LECTIN“

1NCUHATlON

Pretreatment of the cells

Assay

Cell fraction

Lectin

Aggregation-susceptible cells

-

+ + -

Aggregation-deficient cells

-

-

+ + +

Antiaggregation receptor

Incubation with aggregation factor

Diameter of aggregates (pm) 80 t 15 2100 2 400 8 0 t 15 2000 2 400 90 t 20 80 t 15 120 t 60 80 t 15 1950 t 400 260 t 55

As indicated, aggregation-susceptible or aggregation-deficient Geodia cells were, in one senes, pretreated with lectin (10 p g of protein) or antiaggregation receptor (0.7 p,g of neutral carbohydrates), or with both components for 30 minutes at 20°C and, in a second series, assayed directly (-) in the reaggregation assay. Aggregation factor (where indicated) was added at a concentration of 1. I X 104 AU/ml of assay; incubation was for 180 minutes. (From Miiler er at., 1979d.) (+)

tion-promoting activity in both cell fractions (assay c and h), transforms deficient cells to susceptible cells. To test the hypothesis that the antiaggregation receptor is the target molecule for the lectin, competition experiments were performed. They revealed (assay k) that after addition of the soluble antiaggregation receptor to the lectin during the preincubation of aggregation-deficient cells the aggregation potency of these cells is only slightly enhanced. These results are taken as a strong indirect evidence that during the preincubation period a competition between the soluble aggregation factor and the lectin for the antiaggregation receptor occurs. In additional studies it was established that the process of lectincaused conversion of aggregation-deficient cells to aggregation-susceptible cells is completely reversible. From the analytical data of the lectin (demonstrating its D-galactose specificity) as well as from competition studies performed in the biological assay (using the soluble, active antiaggregation receptor and the lectin in experiments containing either aggregation-deficient or aggregation-susceptible cells) at least strong indirect evidence is obtained which indicates that the lectin “inactivates” the function of the antiaggregation receptor, resulting in a loss of its recognition site for the aggregation factor-aggregation receptor complex (Fig. 12). The ‘‘inactivation” of the antiaggregation receptor by the lectin occurs most likely by

167

CELL MEMBRANES IN SPONGES

(masking of O A R by l a t i n I

AR

w e active AR

= agqrgotton

nceplor

OAR :onti-aggrqallon

recepior

CPP = cn~ular prateid particle -4achw

OAR

4

0 xglucuronic acid

b: galactose

1

FIG. 12. A tentative model for the biological role of the Geodia cydonium lectin during the reaggregation of homologous cells. CPP, Sunburst structure carrying the aggregation factor. (From Mdller and Miiller, 1980.)

reversible masking of the terminal galactose moiety of the antiaggregation receptor; this event is followed by an aggregation factor-mediated cell-cell recognition. Hence, the galactose-specific lectin, which does not promote cell aggregation, prevents the active antiaggregation receptor from dissaggregating cell clumps. The conclusion that the lectins are involved in the control of cell adhesion is not restricted to sponges. With respect to this process, the definite role of some lectins has been established in slime molds (Barondes and Rosen, 1976); a

168

WERNER E. G. MULLER

comparable function in vertebrates is very likely (Kitamura, 1980). Nevertheless, there are a series of open questions left. The static model, presented in Fig. 12, does not explain, e.g., the origin of the lectin and the mechanism by which this molecule is guided to the antiaggregation receptor. Recently it was reported that the Geodia lectin is synthesized in mucoid or spherulous cells (Muller et al., 1981a; Bretting et al., 1982). From immunological and biological studies it was concluded (Muller et al., 1981a) that the lectin synthesized in mucoid cells is secreted and subsequently transferred to the cell surface of other cell types. The presence of the lectin on the cell surface was proven qualitatively by rosette formation in the adhesion assay. Quantitative studies revealed that a total of 9.1 X lo6 lectin molecules are present per mucoid cell; we do not know which percentage is localized on the cell surface. As one consequence of the binding of the lectin to cell surfaces, a conversion of aggregation-deficient choanocytes or archaeocytes to aggregation-susceptible ones was observed. In this context it is interesting that Springer et af. (1980) reported independently that the lectinproducing cells from the slime mold Dicryostelium purpureum also contain 5 X lo6 lectin (= purpurin) molecules; only about 2% of the lectin was found to be localized on the cell surface where it is bound to a receptor. While in these two studies the existence of an aggregation-controlling lectin on the cell surface has been demonstrated, Bretting et al. (1981) could not find any evidence that in Axinella polypoides the homologous lectin is cell surface bound. Using indirect immunohistochemical methods, they located the lectin in the vesicles of the spherulous cells and on the surface of spongin fibers. Due to the close association of the spherulous cells with spongin fibers, an involvement of lectin in the production of these fibers is suggested. The observation that two macromolecules (antiaggregation receptor and lectin; at least one of which is cell surface bound) are most likely involved in the “sorting out” process occurring in secondary sponge aggregates, is only a first step. The working model presented might be changed or altered in the future, e.g., after elucidation of the physiological functions of the cell surface bound pgalactosidase (Miiller et al.,1979a) and the aggregation factor associated galactosyltransferase (Miiller et al., 1978d) during the cell-cell interaction.

VIII. Cell Interactions in the Immune Response During the phylogenetic evolution from the unicellular to the multicellular organisms two essential mechanisms have developed: first, a tissue-specific recognition system localized on the cell surface and second, a host defense mechanism against foreign invaders. These two recognition mechanisms enable a multicellular organism to distinguish “self” from “nonself,” and to initiate cellular and molecular reactions which lead to the destruction of neoplasms and of invading microbes. While in protozoa incompatibility may be characterized as

CELL MEMBRANES IN SPONGES

169

intracellular and enzymatic (Cooper, I976), the metazoan incompatibility reactions are evoked at the cell surface. Vaillant observed in 1869 that sponge cells are provided with transplantation immunopotentialities. Since then, rejection of grafts by immune mechanisms in sponges has been thoroughly studied both on cellular (Moscona, 1968; Van de Vyver, 1970; Curtis, 1979; Hildemann et a l . , 1979b; Muller et a l . . 1982c) and on molecular levels (Van de Vyver, 1971; Miiller et a l . , 1976c, 1981~;Evans and Curtis, 1978). Graft rejection experiments are usually thought to be one of the clearest tests of individuality in higher organisms (Curtis, 1979). However, graft rejection not only occurs between individual sponges of different species but may happen also between individual specimens of the same species (Van de Vyver, 1970). As a consequence of this fact, a clear terminological distinction between individual and specimen is required (Weissenfels and Striegler, 1979). Some marine sponges (e.g., T . lyncurium, Chondrosia reniformis) and many members of the Spongillidae, the fresh water sponges, form asexual reproductive bodies (gemmules) in response to alterations of the environmental conditions. In autumn, fresh water sponges usually form large numbers of gemmules and then disintegrate. This means that one individual sponge body “fragments” into many viable entities which will hatch under suitable environmental conditions. Young sponges, developed from gemmules and derived from the same mother animal, merge into each other and form a functional, integrated sponge body (Van de Vyver, 1970; Weissenfels and Striegler, 1979); older sponges, however, with already developed exhalant canals do not fuse in the same system. In contrast, sponges hatched from gemmules of mother animals, collected in different ponds, do not merge. Based on these facts Weissenfels and Striegler (1979) define an individual of a fresh water sponge as the sum of specimens which may live spacially separated or in a fused state, from which disintegration (gemmulation) into many viable, functional specimens may occur. According to this definition the terms “individual” and “strain” (Van de Vyver, 1975) are identical. Graft rejection experiments revealed that some of the sponge species examined show intraspecific incompatibility (allograft rejection), e.g., T . lyncurium (Paris, 1957), Callyspongia diffusa (Hildemann et al., 1979b), and E . ji’uviatilis (Van de Vyver, 1971) reject allografts, while S. domuncula (Paris, 1957), M. prolifera (Moscona, 1968), and G. cydonium (Muller et a / . , 1982~)do not show any noticeable evidence of separation of graft tissue from host. It seems to me that sponge species which form gemmules reject allogeneic grafts, while those which do not produce these asexual bodies accept allografts. All sponge species investigated reject xenografts. It has even been suggested that some species are provided with an immune memory (Hildemann et a f . , 1979a). With respect to their immunopotentiality, the sponges are already highly developed; they are outfitted with the dual immune protection system: cellular immunity and humoral immunity. The fresh water sponge E . jluviatis has been used widely to elucidate al-

170

WERNER E. G . MULLER

logeneic immunocornpetence on cellular and subcellular levels. Van de Vyver (1970) succeeded in the cultivation of eight different E . flfluviatilis lines, which show intraspecific incompatibility. These lines were originally termed “strains”; in the light of recent findings (Weissenfels and Striegler, 1979) they may better be called individuals. When two sponge individuals are brought into contact, they do not fuse together. Along their zone of contact, they form a nonrnerging front which consists of the face-to-face contact of the respective pinacoderms (Van de Vyver and De Vos, 1978). In response to the contact: of the two individuals, an extracellular collagen-like material is secreted at the barrier. In contrast to histoincompatibility reactions with marine sponges, no cyrotoxic reactions occur in the fresh water sponge allogeneic systems. In an elegant study on cellular level it was established that the three purified Ephydutia cell fractions, archaeocytes, choanocytes, and pinacocytes, share the same allogeneic recognition properties; they form individual specific aggregates (De Sutter and Van de Vyver, 1979). However only pinacocytes are involved in the rejection process. Thus it appears, that all three cell types recognize the antigen (= allogeneic cells) but only the pinacocytes have the immunopotentiality to “eliminate” the antigen. Again using the E . fluviatilis system, Van de Vyver (Van de Vyver, 1971; Curtis and Van de Vyver, 1971) demonstrated that sponges are not only provided with cellular but also with a humoral immune system. When cells from two individuals were mixed individual specific aggregates were formed. In addition, individual specific aggregation factors with opposite influences on the adhesion of cells from homologous or heterologous individuals are produced. In other words, the factor enhaces only the mutual adhesiveness of homologous cells while it reduces the adhesiveness of cells from another individual. The highly discriminating function of this factor is observed only in the presence of (:a2+. This means that (in contrast to marine sponges) fresh water sponges contain an aggregation factor which controls adhesion both in a positive and a negative fashion depending on the origin of the cells. This aggregation factor is inactive at low temperature (4°C); at room temperature it promotes cell clustering even in the absence of Ca2 . However cell aggregates formed in the absence of Ca2 never develop into functional sponges. The Ephydatia factor is thermolabile and has a molecular weight of approximately 50,000 (Van de Vyver, 1975). It has been speculated that the concentration of the factor is highest within the sponge body, and minimal at the periphery (Van de Vyver, 1975) resulting in a lower adhesiveness of cells at the periphery. When two sponges come into contact, the concentration of the factor will rise in the contact region with the following consequence. When two sponge bodies, derived from the same individual, make contact the aggregation-promoting activity of the factor is utilized and the two bodies fuse; on the other hand when two sponges from different individuals come into contact the aggregation-inhibitory activity of the factor comes into function. +

+

CELL MEMBRANES IN SPONGES

171

The inhibitory effect of the factor on cell adhesiveness appears to be reversible. The sponge cells react to allogeneic signals not only in sorting out but also in phagocytosis (Van de Vyver and Buscema, 1977; De Sutter and Van de Vyver, 1979). In aggregates from autogeneic cells phagocytosis is limited, however, in allogeneic “aggregates,” phagocytosis attains gigantic proportions. This would mean that in essence the Ephydatia “aggregation factor” can also act as an antibody-like macromolecule. The producing cells have the capacity to recognize surface features of cells that are not normal constituents of the respective individual and respond with the synthesis of the factor. This factor prevents invasion of allogeneic cells by reducing their adhesiveness. In addition, it might be speculated that the factor stimulates phagocytosis resulting in an elimination of allogeneic cells. A recent report indicates (Van de Vyver, 1980) that Ephydatia is not provided with an inducible immunologic memory. As a rule, marine sponges are thought to accept allografts (Burger et al., 1978; Curtis, 1978; Miiller et a f . , 1981~);only gemmule-producing sponges, like T. lyncurium (Paris, 1961), C.diflusa (Hildemann et al., 1979a), and A. pofypoides (Van de Vyver, 1980) fail to accept allografts. The histological aspect of xenograft rejection has been studied in some detail with the following two pairs: Hymeniacidon perleve X Amphilectus fucorum (Evans and Curtis, 1978) and G . cydonium X Geodia rovinjensis (Muller et a f . , 1981~).In these studies, grafts were taken from donors by punching out a tissue mass of a diameter of 1.5 cm and a length of 3.0-6.0 cm (Fig. 13A). The grafts were then inserted into holes of the same or another species; the holes of the host had the same dimensions as the grafts. In the two Geodia species no allograft rejection was observed (Fig. 13B). The process of xenograft rejection in this system can be subdivided into the following three events: (1) fusion of approximately 60% of the contact areas during the first 2 days after grafting; (2) necrotic alteration of the graft, which was completed after 5 days (Fig. 13C) (in most cases the contact zone of the host with the graft was found to necrose as well); and (3) rejection of the xenogeneic graft and formation of a fissure between the transplant and the host; this phase started 3 to 5 days after grafting. Very interesting is the finding that the fusion zone of grafts in the H . perfeve and A. fucorum system is characterized by archaeocyte infiltration (Evans and Curtis, 1978; Evans et a f . , 1980). It is suggested that these cells are intimately involved in the grafting response, in particular in the tissue reorganization and necrosis and perhaps in the secretion of cytotoxic material, In contrast to fresh water sponges, marine sponges produce in response to grafts an inhibitory aggregation factor which is not identical with the aggregation factor (Muller et a / . , 1976c, 1981c; Evans and Curtis, 1978). However, this inhibitory aggregation factor influences the activity of the aggregation factor (which might be called in this context promotory aggregation factor). The existence of the inhibitory aggregation factor has been demonstrated for H . viridis

172

WERNER E. G . MULLER

FIG. 13. Graft experiments with Geodia cydonium and Geodia rovinjensis. (A) Insertion of a G. rovinjensis graft into a hole, punched into G . rydonium. X0.5. (B)Fusion of an allograft. X I . (C) Rejection of a xenograft from G . rovinjensis in G . rydonium. x I . (D)Cytotoxic reaction in G . rydonium caused by the inhibitory aggregation factor, isolated from xenografts of G . rovinjensis. x 2 . (From Miiller e r a / . , 1982c.)

(McClay, 1974), A. polypoides and Crambe crambe (Van de Vyver, 1975), Ircinia muscarum (Muller et al., 1976c), A . fucorum (Evans and Curtis, 1978), and G . cydonium (Muller et al., 1981~).Most is known about the factors from G . cydonium and G . rovinjensis. The inhibitory aggregation factor was found to be induced in suspensions of allogeneic single cells and in xenografts from these two Geodia species. After incubation of dissociated Geodia cells in sea water for 3 hours the cells release this factor into the culture medium. The synthesis is temperature dependent and can be blocked by tunicamycin. The identical molecule can be isolated from xenografts after a grafting period of 3 days (the period after which necrotic reactions start in the system). This means that the synthesis of the inhibitory aggregation factor is most likely controlled by cell-cell contact. In the homologous (allogeneic and isogeneic) system the inhibitory aggregation factor is synthesized when the cells do not have close cellular contact, that is, in

CELL MEMBRANES IN SPONGES

173

the absence of the homologous aggregation factor. It is not synthesized in allografts and not in a functional individual. In the heterologous system (in xenografts), the inhibitory aggregation factor is synthesized after a contact period with the host of 3 days during which some xenogeneic cells have been in close contact. The inhibitory aggregation factor from G. cydonium has been purified from the culture medium of primary aggregates (in the absence of the aggregation factor) (Muller et a l . , 1981~).It was obtained in a homogeneous state after heat treatment, ion-exchange chromatography, and gel filtration. The molecular weight is around 26,000. The relatively heat-stable inhibitory aggregation factor shows an ultraviolet absorption spectrum with a maximum at 258 nm and a minimum at 237 nm; this might hint at the presence of a higher amount of phenylalanine in the molecule. It consists of 50% neutral carbohydrate, 40% protein, and 10% hexuronic acid. The activity of the factor is destroyed by trypsin and a mixed glycosidase preparation from Turbo cornutus. In in vitro studies it was shown that the inhibitory aggregation factor affects the aggregation factor-mediated reaggregation of single cells by extension of the lag phase preceding the aggregation process; the endpoint of the reaggregation process is not changed. Kinetic studies revealed that the inhibitory aggregation factor reduces the process of aggregation by competing with the aggregation receptor for the binding sites at the aggregation factor. The binding of the inhibitory aggregation factor from Geodia to a heterologous aggregation factor is apparently reversible, ) obviously in an irrewhile that from I. muscarum (Muller et al., 1 9 7 6 ~binds versible manner. In contrast to the inhibitory activity of the aggregation factor from Ephydatia, the factor from G. rovinjensis causes necrotic alterations after injection into the endosome of G. cydoniurn (Fig. 13D). The reported exciting observations of transplantation studies in sponges indicate that these animals are provided with a well-developed tissue immunopotential. The sensor and the effector in this system are localized on the cell membrane. In the absence of the cell-cell contact or after contact with xenogeneic tissue, the cells produce the inhibitory aggregation factor. This factor affects the activity of the aggregation factor and perhaps, in addition, causes cytotoxic reactions in the xenogeneic tissue. Some observations even indicate that the marine sponges have an immunological memory (Evans et al., 1980).

IX. Recognition of Symbionts Most of the sponges live in a most spectacular association with bacteria and other unicellular organisms (review: Sara and Vacelet, 1973), e.g., some Verongia species contain in the mesohyl a bacteria population that accounts for 38%

174

WERNER E. G. MULLER

of the tissue volume, which is almost as much as the mesohyl volume itself (41%) and twice as much as the actual cellular volume (21%) (Bertrand and Vacelet, 1971). The bacteria in Verongia were determined to belong to the genera Pseudomonas and Aeromonas. In some sponge species bacteria are also found intracellularly in vacuoles and even within the nucleus (Vacelet, 1970). Since it is known that matrix bacteria are transferred into the larvae (Vacelet, 1975) or into the gemmules (Kilian, 1964), at least in some species young sponges begin life with an established bacterial flora. Garrone (1978) showed that in the mesohyl of Chondrosia nuciila bacteria are often surrounded by a granular zone which separates them from the collagen fibrils. This granular zone is rich in carbohydrate. It is assumed (Garrone, 1978; Bergquist, 1978) that the association between sponge cells and bacteria is a mutual symbiosis, where the sponge can phagocytose the bacteria and the bacteria are provided with a medium which enhances their growth. The biochemical basis of this symbiosis-like association is known to some extent. First, the bacteria show a great uptake rate of radioactively labeled amino acids which is in most cases higher than that determined in the neighboring sponge cells (Wilkinson and Garrone, 1980). Second, we demonstrated (Kurelec et al., 1977) the presence of the y-glutamyl cycle in sponges. This amino acid transport system enables the sponges to accumulate dissolved free amino acids from the surrounding milieu. Third, bacteria introduce carbohydrate material, some in the form of glycoproteic assemblages, into the sponge tissues (Heath, 1971). From these observations it seems to be very likely that sponge cells deliver amino acids to the bacteria which subsequently provide the host cells with diverse sugars. In addition, bacteria might be involved in a controlled fragmentation of collagen bundles, a process which is one prerequisite for the growth of a sponge (Garrone, 1975). It is reasonable to assume that the symbiotic association between bacteria and sponges is also based on a tuned cell-cell relationship. Initial experimental observations indicate that lectins are most likely the basis for this symbiosis (Muller et af, 1981d; Bretting et a f . , 1981). From the marine sponge H . panicea the bacterium P seudomonas insofira has been isolated; the bacteria account for less than 1% of total tissue volume. These bacteria were found (Muller et al., 1981d) to grow in vitro only in the presence of a homologous galacturonic acidspecific lectin which is synthesized by the sponge cells. This lectin is not utilized during growth of the Pseudomonas strain. Binding studies revealed that the lectin contains two different binding sites, one site for receptors on the sponge cells and a second one for membrane components of the bacteria. From these data we conclude that sponges are provided with at least two recognition systems for xenogeneic cells: one system which prevents invasion of cells from other species and a second one, which supports a symbiotic relationship to bacteria.

CELL MEMBRANES IN SPONGES

175

X. Concluding Comments Studies to elucidate the morphology and function of cell membranes in sponge are of importance for the understanding of this own biological phylum. Furthermore, they are even more important for the knowledge of the physical and especially biochemical mechanisms of cell adhesion in general. This is because the sponges are viewed as a suitable model system for investigations on basic biochemical aspects of intercellular recognition specificity. Sponges are the most simple multicellular organisms and consist of two main classes of cells, first, the germ cells, and second, the archaeocytes and their derivatives. It is known that the archaeocytes are “embryonic” cells, which can differentiate into any other tissue cell type; these “differentiated” functional cells still seem to have the potency to dedifferentiate and then to redifferentiate into another cell type via the archaeocyte stage. Most of the sponge cells are characterized by a comparably high motility. These “wandering” cells are not positioned in a static manner in the functional organism but are arranged in a highly flexible and dynamic way in the sponge tissue units. The potency of the cells to be motile can be realized in the sponge organism by the following three peculiarities: first, the junctions between the cells show only little specialization, second, the distance between cells are unusually high (between 300 and 1000 A), and third, the cells are embedded in an intercellular matrix, consisting of a fluid phase (carbohydraterich compounds) and a solid phase (collagenous fibers and fibrils, spicules). From the biological and biochemical point of view, the sponges are suitable systems for cell membrane studies, because they can be dissociated into single cells which rapidly aggregate into reconstructed functional organisms in simple, defined salt solutions. For the understanding of the molecular mechanisms underlying the cell specific recognition in a multicellular organism, biological systems consisting of only a few somatic cell types are preferable. Over 90% of the sponge cells were found to belong to four cell types: archaeocytes, choanocytes, pinacocytes, and spherulous cells; these cell types are not finally determinated but have retained some dedifferentiation potential. In such a model, cell- and tissue-specific recognition systems of only a few or even only one biochemical recognition principle might be realized. For practical reasons, the sponges offer a great advantage because the molecular components involved in sponge cell recognition can be isolated from an almost homogeneous starting material which is available in large quantities. Furthermore, sponge cells are morphogenetically characterized by their intercellular specificity. Knowing this background it is not surprising that in 1907 Wilson demonstrated differential cellular affinities in mixtures of mechanically dissociated marine sponges. Later sponges have been used for isolation and purification of the first soluble aggregation factor. Sponges also served again as a model system from

176

WERNER E. G. MULLER

which an aggregation receptor was identified and purified. While the :species specificity of some aggregation factor-aggregation receptor combinations is well documented, all the factors and receptors which have been isolated from other models are only tissue-specific. Using the two component system from sponges, aggregation factor and aggregation receptor, it was possible to demonstrate that they are nonenzymic molecules which provoke cell membrane interaction; enzyme systems on the other hand are involved in a controlled intercellular adhesion based on an activation or inactivation of the factor and receptor. Also important was the finding that adhesion and deadhesion for sorting out, for recognition, and for positioning are mediated by special macromolecules (antiaggregation receptor and a lectin). A new, very exciting era began with the tiiscovery that sponges are already provided with transplantation immunopotentialities. These animals seem to be outfitted with the dual immune protection systems known from vertebrates and higher invertebrates. Clearly sponge cell membranes are simple but not primitive: simple with respect to dynamic cellular structures involved in the formation of cell contact. However, the nature and the function of the molecular factors in cell-cell interaction are complex and integrated. Perhaps in future work, sponges might provide an excellent opportunity to study the influence of cell adhesion molecules on the regulation of genetic activity and on sponge development.

ACKNOWLEDGMENTS Work supported by a grant from the Stiftung Volkswagenwerk (38199;W.E.G. M.). The author is especially indebted to his colleagues of the two organizations (Center for Marine Research, Rovinj, Yugoslavia and the Society for Protection of Nature of Israel, Ofira, Israel) for their interest in and their support of the work.

REFERENCES Abercrombie, M., and Heaysman. I. E. M . (1954). Exp. Cell Res. 13, 276291. Ankel, W. E., and Wintermann-Kilian, G . (1952). Z. Narurj‘orsch. 76, 475481. Bagby, R. M. (1970). 2. ZeUforsch. Mikrosk. Anar. 105, 579-594. Balsamo, J . , and Lilien, J. (1980). Biochemistry 19, 2479-2484. Barondes, S. H., and Rosen, S . D. (1976). In “Neuronal Recognition” ( S . H. Barondes, ed.). pp. 331-356. Plenum, New York. Bergquist, P. R. (1978). “Sponges.” Hutchinson, London. Bertrand, J. C., and Vacelet, J. (1971). C.R. Hebd. Seances Acad. Sci. Paris 273, 638-611. Borojevic, R., Fry, W. G., Jones, W. C., Levi, C., Rasmont, R.,Sara, M., and Vacelet, J. (1968). Bull. Mus. Narl. Hisr. Nar. 39, 1224-1235. Bowerbank, J. (1858). R. SOC. Philos. Trans. 148. 279-332.

CELL MEMBRANES IN SPONGES

177

Brackenbury, R., Thiery, J. P., Rutishauser, U., and Edelman, G. M. (1977). J. Biol. Chem. 252, 6835-6840. Bretting, H., Phillips, S. G., Klumpart, H. J., and Kabat, E. A. (1981). J. Immunol. 127, 1652-1658. Bretting, H.. Jacobs, G., Donadey, C., and Vacelet, J. (1982). In “Lectins-Biology, Biochemistry, Clinical Biochemistry” (T. C. Bog-Hansen, ed.), Vol. 2. De Gruyter, Berlin, in press. Brien, P. (1936). Arch. Biol. 48, 185-268. Brightman, M. W., and Palay, S. L. (1963). J. CellBiol. 19, 415439. Burger, M. M., and Jumblatt, J. (1977). I n “Cell and Tissue Interactions” (J. W. Lash and M. M. Burger, eds.), pp. 155-172. Raven, New York. Burger, M. M., Burkart, W., Weinbaum, G., and Jumblatt, J. (1978). In “Cell-Cell Recognition” (A. S. G. Curtis, ed.), pp. 1-24. Society for Experimental Biology Symposium, Cambridge. Burger, M. M., Jumblatt, J., Misevic, G., and Burkart, W. (1981). J. Supramol. Struct. Cell. Biochem. Suppl. 5 , 243. Buscema, M., and Van de Vyver, G. (1978). In ”Biologie des Spongiaires” (C. Levi and N. BouryEsnault, eds.), pp. 225-23 I . Centre National de la Recherche Scientifique, Paris. Buscema, M., De Sutter, D., and Van de Vyver, G. (1980). Wilhelm R o u Arch. 188, 45-53. Cauldwell, C. B., Henkart, P., and Humphreys. T. (1973). Biochemistry 12, 3051-3055. Conrad, J., Zahn, R. K., Kurelec, B., Uhlenbruck, G., and Miiller, W. E. G. (1981). J . Supramol. Srrucr., 17, 1-9. Conrad, I . , Diehl-Seifert, B., Zahn,R. K., Uhlenbruck, G., Zimmerman, E., and Muller, W. E. G. (1 982). Submitted. Cooper, E. L. (1976). “Comparative Immunology,” p. 69. Prentice-Hall, New York. Curtis, A. S. G. (1962). Nature (London) 1%. 245-248. Curtis, A. S. G. (1967). “The Cell Surface: Its Molecular Role in Morphogenesis.” Logos, London. Curtis, A. S. G. (1970a). Nurure (London) 226, 260-261. Curtis, A. S. G. (1970b). J. Embryol. Exp. Morphol. 23, 253-272. Curtis, A. S. G. (1973a). In “Cell Biology in Medicine” (E. E. Bittar, ed.), pp. 441-471. Wiley, New York. Curtis, A. S. G. (1973b). I n “Progress in Biophysics” (J. A. V. Butler and D. Noble, eds.). Vol. 27, pp. 371-386. Pergamon, Oxford. Curtis, A. S. G. (1978). “Biologie des Spongiaires” (C. Levi and N. Boury-Esnault, eds.), pp. 205-209. Centre National de la Recherche Scientifique, Paris. Curtis, A. S. G. (1979). “Biology and Systematics of Colonial Organisms’’ (G. Larwood and B. R. Rosen, eds.), pp. 3 9 4 8 . Academic Press, New York. Curtis, A. S. G., and Van de Vyver, G. (1 971). J . Embryol. Exp. Morphol. 26. 295-3 12. De Sutter, D., and Van de Vyver, G. (1977). Wilhelm Rowr Arch. 181, 151-161. De Sutter, D., and Van de Vyver, G. (1979). Dev. Comp. Immunol. 3, 389-397. De Vos, L. (1974). “Etude ultrastructurale de la formation et de I’tclosion des gemmules d’ Ephydatia fluviatilis.” Doctoral thesis, Brussels. Determann, H. (1969). “Gel Chromatography.” Springer-Verlag, Berlin and New York. Diaz, J . P. (1979). “Variations, Diffkrentiations et Fonctions des Catkgories Cellulaires de la Demosponge d’Eaux Saumatres, Suberites Massa, Nardo, au Cows du Cycle Biologique Annuel et dans des Conditions Exptkirnentales.” Doctoral thesis, Montpellier. Evans, C. W., and Curtis, A. S. G. (1978). I n “Biologie des Spongiaires” (C. Levi and N. BouryEsnault, eds.), pp. 21 1-215. Centre National de la Recherche Scientifique, Paris. Evans, C. W., Kerr, J., and Curtis, A. S. G. (1980). “Phylogeny of Immunological Memory” (M. J. Manning, ed.), pp. 27-34. Elsevier. Amsterdam. Farquhar, M. G., and Palade, G. (1963). J. Cell Biol. 17, 375412. Feige, W. (1969). 2001.Jahrb. Anat. 86, 177-237.

178

WERNER E. G . MULLER

Frazier, W., and Glaser, L. (1979). Annu. Rev. Biochem. 48, 491-523. Galtsoff, P. S. (1923). Biol. Bull. 45, 153-161. Galtsoff, P. S. (1925). J . Exp. Zool. 42, 183-222. Galtsoff, P. S. (1929). Biol. Bull. Mar. Biol. Lab. Woods Hole 57, 250-260. Garrod, D.R., and Nicol. A. (1981). Biol. Rev. 56, 199-242. Gmone, R. (1975). In “Protides of the Biological Fluids; 22nd Colloquium” (H. Peeters, ed.), pp. 59-63. Pergamon, Oxford. Gmone, R. (1978). “Phylogenesis of Connective Tissue.” Karger, Basel. Garrone. R., Thiney, Y.,and Pavans de Ceccatty, M. (1971). Experienria 27, 1324-1326. Gmone, R., Huc, A., and Junqua, S. (1975). J. Ulrrasrruct. Res. 52, 261-275. Gasic. G. J . . and Galanti, N. L. (1966). Science 151,203-205. Goodenough, D.A., and Revel, J. P. (1970). J. Cell Biol. 45, 272-290. Grant, R. (1825). Edinburgh Philos. J . 13,94-107. Gross, J . , Sokal, Z., and Roughvie, M. (1956). J. Hisrochem. Cytochern. 4, 227-246. Hartman, W. D. (1957). Evolution 11, 294-297. Hausman, R. E., and Moscona, A. A. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 916920. Heath, E. C. (1971). Rev. Biochem. 40,29-56. Henkart, P.,Humphreys, S., and Humphreys, T. (1973). Biochemistry 12,3045-3049. Hildemann, W. H., Bigger, C. H., and Johnston, I. S. (1979a). Transplanr Proc. 11, 11361142. Hildemann, W. H., Johnson, 1. S., and Jokiel, P. L. (1979b). Science 204, 420-422. Humphreys, S., Humphreys, T . , and Sano, J. (1977). J . Supramol. Srruct. 7, 339-351. Humphreys, T. (1963). Dev. Biol. 8, 27-47. Humphreys, T. (1965a). Exp. Cell Res. 40,539-543. Humphreys, T. (1965b).J. Exp. Zool. 160, 235-240. Humphreys, T. (1970a). Symp. Zool. SOC. London 25, 325-334. Humphreys, T. (l970b). Transplant. Proc. 2, 194-198. Huxley, J. S. (1912). Philos. Trans. R. SOC.London Ser. B 202, 165-189. John, H. A., C a m p , M. S., Mackenzie, A. M., and Kemp, R. B. (1971). N a m e (London) New Biol. 230, 126-128. Jumblatt, I. E., Weinbaum. G . , Turner, R., Ballmer, K., and Burger, M. M. (1977). In “Surface Membrane Receptors” (K. Bradshaw, U. Frazier, R. Merrel, D. Gottlieb, and R. Hogue-Angeletti, eds.), pp. 73-86. Plenum, New York. Jumblatt, J . E., Schlup, V., and Burger, M. M. (1980). Biochemisfry 19, 1038-1042. Junqua, S.,Robert, L., G m o n e , R., Pavans de Ceccatty, M., and Vacelet, J. (1974). Connect. Tissue Res. 2, 193-203. Junqua, S.. Fayolle, J., and Robert, L. (1975). Comp. Biochem. Physiol. 50 B, 305-309. Junqua, S . , Lemmonier, M., and Robert, L. (1981). Comp. Biochem. Physiol.. 69B, 445-453. Kartha, S.,and Mookerje, S. (1979). Mikroskopie 35, 213-220. Kilian, E. F. (1964). Zool. Beitr. 10,85-199. Kitamura, K. (1980). J. Embryol. Exp. Morphol. 59, 59-69. Kurelec, B . , Rijavec, M., Miiller. W. E. G., Britvic, S., and Zahn, R. K. (1977).Rapp. Comm. inr. Mer Medit. 24(8), 67-71. Labat-Robert, J., Pavans de Ceccatty, M., Robert, L., Auger, C., Lethias, C., and Garrone, R. (1979). Proc. Int. Symp. Glycoconjugares. 5rh. pp. 431-432. Ledger, P. W. (1975). Tissue Cell 7, 13-18. Leith, A. (1979). Biol. Bull. 156,212-223. Levi, C. (1960). Cah. Biol. Mar. 1, 353-358. Lilien, I., Hermolin, I.. and Lipke, P. (1978). In “Specificity of Embryological Interactions” (D. R. G m o d , ed.), pp. 132-155. Chapman & Hall, London.

CELL MEMBRANES IN SPONGES

179

Lowry, 0. H.. Rosebrough. N. J., Farr. A. L., and Randall, R. J. (1951). J. Eiol. Chem. 193, 265-275. McClay, D. R. (1974). J. Exp. Zool. 188, 89-102. MacLennan, A. (1969). J. Exp. Zool. 172, 253-266. MacLennan, A. P. (1970). S m p . Zool. Soc. London 25, 299-324. MacLennan, A. (1974). Arch Eiol. 85, 53-90. Margoliash, E., Schenck, 1. R . , Hargie, M. P., Burokas, S., Richter, W. R., Barlow, G. H., and Moscona, A. A. (1965). Eiochem. Biophvs. Res. Commun. 20. 383-388. Michalopoulos. G., and Pitot. H. C. (1975). Exp. Cell Res. 94, 7G78. Moscona, A. A. (1956). Proc. SOC. Exp. Eiol. 92, 410. Moscona. A. A. (1962a). Inr. Rev. Exp. Parhol. 1, 371428. Moscona, A. A. (1962b). J. Cell. Comp. Physiol. 60 (Suppl. I ) , 65-80. Moscona, A . A. (1963). Proc. Narl. Acad. Sci. U.S.A. 49, 742-747. Moscona, A. A. (1965). In “The Living Cell” (D. Kennedy, ed.), pp. 213-221. Freeman, San Francisco, California. Moscona. A. A. (1968). Dev. Eiol. 18, 250-277. Miiller, K. (191 I ) . Zool. Anr. 37. 82-1 13. Miiller, W. E. G., and Zahn, R. K. (1973). Exp. Cell Res. 80, 95-104. Miiller, W. E. G., and Miiller, 1. (1980). Mol. Cell. Eiochern. 29, 131-143. Miiller, W. E. G., Miiller, I . , and Zahn, R . K. (1974). Experientia 30, 899-902. Miiller, W. E. G., Miiller, I., Zahn. R. K.. and Kurelec. B. (1976a). J. Cell Sci. 21, 227-241. Miiller, W. E. G., Miiller. I., and Zahn, R. K. (1976b). Biochim. Eiophys. Acra 418, 217-225. Miiller. W. E. G., Miiller, I., Kurelec, B., and Zahn, R. K. (1976~).Exp. Cell Res. 98, 31-40. Miiller. W. E. G., Miiller. I . , and Zahn, R. K. (1978a). Res. Mol. Eiol. 8. 1-87. Miiller, W. E. G.. Beyer, R., Pondeljak, V., Miiller, I . , and Zahn, R. K. (1978b). Tissue Cell 10, I9 I- 199. Miiller, W. E. G., Zahn, R. K., Kurelec, B., Uhlenbruck, G.. Vaith, P., and Miiller, 1. (1978~). Hoppe-Sevlers 2. Physiol. Chem. 359, 529-537. Miiller, W. E. G., Zahn, R. K., Kurelec, B., and Miiller, I. (1978d).Exp. CellRes. 113,409-414. Miiller, W. E. G., Zahn, R. K., Kurelec, B., and Miiller, 1. (1978e). Wilhelm Roux Arch. 184, 29-40. Miiller, W. E. G . , Miiller, 1.. Pondeljak, V., Kurelec, B., and Zahn, R. K. (19780. Differentiation 10.45-53. Miiller, W. E. G., Zahn, R. K., Kurelec, B., and Miiller. I. (19788). Differenriation 10, 55-60, Miiller, W. E. G., Miiller, I., Zahn, R. K.,and Kurelec, B. ( I978h). Cell Tissue Kiner. 11, 23-32. Miiller, W . E. G.. Zahn. R. K.. Kurelec, B., Miiller, I., Uhlenbruck, G., and Vaith, P. (1979a).J. Eiol. Chem. 254, 128W1287. Miiller, W. E. G., Zahn, R. K., Arendes, J., Kurelec, B.. Steffen, R., and Zahn, R. K. (1979b). Biochim. Biophys. Acra. 551, 363-367. Miiller, W. E. G . . Zahn, R. K., Kurelec, B., Miiller, I., Vaith, P., and Uhlenbruck, G. (1979~). Eur. J. Biochem. 97, 585-591. Miiller, W. E. G., Kurelec, B., Zahn, R. K., Miiller, I . , Vaith, P., and Uhlenbruck, G. (1979d). J. Eiol. Chem. 254. 7479-7481. Miiller, W. E. G., Zahn, R. K.. Rijavec, M., BritviC, S., Kurelec, B., and Miiller, 1. (1979e). Eiochem. Svst. Ecol. 7, 49-55. Miiller, W. E. G., Zahn, R. K., Kurelec, B., Miiller, I., Vaith. P., and Uhlenbruck, G. (1980).Int. J . Eiol. Macromol. 2, 297-301. Miiller, W. E. G., Zahn, R. K.,Muller, I., Kurelec, B., Uhlenbruck, G., and Vaith, P. (1981a). Eur. J. Cell Eiol. 24, 2&35.

180

WERNER E. G. MULLER

Miiller, W. E. G., Zahn, R. K., Bernd, A., Dawes, K., Miiller, I., Uhlenbruck, G . , and Dorn, A. (1981b). In “Cellular Recognition” (L. Glaser, W. Frazier, and D. Gottlieb, eds.). Liss, New York. Miiller, W. E. G., Bernd, A., Zahn, R. K., Kurelec, B., Dawes, K., Miiller, I., and Uhlenbruck, G. (1981~).Eur. J . Biochem. 116, 573-579. Miiller, W. E. G., Zahn, R. K., Kurelec, B., Lucu, C., Miiller, I., and Uhlenbruck, G . (1981d). J . Bacteriol. 145, 548-558. Miiller, W. E. G., Zahn, R. K., and Maidhof, A. (1982a). Senckenbergiana Lethuea. in press. Miiller, W. E. G., Zahn, R. K., Zahn, G., Rijavec, M., Batel, R., Kurelec, B.. Dorn. A., and Miiller, 1. (1982b). Thalassia Jugosl. in press. Miiller, W. E. G., Miiller, I., Zahn, R. K., Kurelec, B., Batel, R., and Uhlenbruck, G. (1982~) Senckenbergiana Biol., in press. Miiller, W. E. G., Conrad, J., Pondeljak, V., Steffen, R., and Zahn, R. K. (19824). Tissue Cell 14, 2 19-223. Noll, H., Matranga, V.. Cascino, D., and Vittorelli, L. (1979). Proc. Narl. Acad. Sci. U.S.A. 76, 288-292. Nowak, T. P., Haywood, P. L., and Barondes, S. H. (1976). Biochem. Biophys. Res. Commun. 68, 650-657. Orlov, Y. A. (1971). “Fundamentals of Paleontology,” Vol. 11, pp. 5-129. Israel Program for Scientific Translations, Jerusalem. Pappas, G . D. (1975). In “Cell Membranes” (G. Weissmann, ed.), pp. 87-94. HP Publ., New York. Paris, J . (1957). C.R. Acad. Sci. 245, 578-580. Paris, J. (1961). Vie Milieu Suppl. 11, 1-74. Pavans de Ceccatty, M. (1974). Am. Zool. 14, 895-903. Pavans de Ceccatty, M., Thiney, Y.,and Garrone, R. (1970). Symp. Zool. SOC. London 25, 449-466. Pottu-Boumendil, J. (1975). Thesis No. 41 I , Universitk Claude Bernard, Lyon. Rambourg, A. (1971). Int. Rev. Cytol. 31, 57-114. Rauvala, H., and Hakornori, S. (1981). J . Cell Biol. 88, 149-159. Revel, J. P., and Goodenough, D. A. (1970). In “Chemistry and Molecular Biology of the Intercellular Matrix” (E.A. Balazs, ed.), Vol. 3, pp. 1361-1380. Academic Press, New York. Roseman, S. (1970). Chem. Phys. Lipids 54. 217-221. Rudland, P. S., Seeley, M., and Seifert, W. (1974). Nature (London) 251, 417-419. Sara, M. (1956). Boll. Zool. (Torino) 23, 113-1 19. Sara, M., and Vacelet, J. (1973). In “Trait6 de Zoologie” (P. P. Grassk, ed.), pp. 462-576. Masson, Paris. Sara, M., Liaci, L., and Melone, N. (1966). Nature (London) 210, 1168-1 169. Spiegel, M. (1954). Biol. Bull. 107, 130-155. Springer, W. R., Haywood, P. L., and Barondes, S. H. (1980). J . Cell Biol. 87, 682-690. Steinberg, M. S. (1970). J. Exp. Zool. 173, 395-434. Takeichi, M. (1977). J. Cell Biol. 75, 464-474. Takeichi, M., Ozaki, H. S.,Tokunaga, K., and Okada, T. S. (1979). Dev. Biol. 70, 195-205. Thiery, J. P., Brackenbury, R., Rutishauser, U . , and Edelman, G. M. (1977). J. Biol. Chem. 252, 6841-6845. Turner, R. S., and Burger, M. M. (1973a). Nature (London) 244, 509-510. Turner, R . S., and Burger, M. M. (1973b). Rev. Physiol. 68, 122-155. Urushihara. H., and Takeichi, M. (1980). Cell 20, 363-371. Urushihara, H., Ozaki, H. S., and Takeichi, M. (1979). Dev. Biol. 70, 206-216. Vacelet, J. (1970). J. Microsc. 9. 333-346.

CELL MEMBRANES IN SPONGES

181

Vacelet, J. (1975). J. Microsc. Biol. Cell. 23, 271-288. Vaillant, L. ( 1869). C.R. Acud. Sci. 68, 88. Vaith, P., Miiller, W. E. G., and Uhlenbruck, G. (1979a). Dev. Comp. Immunol. 3, 257-259. Vaith, P., Uhlenbruck, G., Miiller, W. E. G., and Holz, G. (1979b). Dev. Comp. Immunol. 3, 399-416. Van de Vyver. G. (1970). Ann. Embryol. Morph. 3, 251-262. Van de Vyver, G. (1971). Ann. Embryol. Morph. 4, 373-381. Van de Vyver, G. (1975). In “Current Topics in Developmental Biology” (A. A. Moscona and A. Monroy, eds.), pp. 123-140. Academic Press, New York. Van de Vyver, G. (1978).In “Biologie des Spongiaires” (C. Levi and N. Boury-Esnault, eds.), pp. 195-204. Centre National de la Recherche Scientifique, Paris. Van de Vyver, G. (1980). In “Phylogeny of Immunological Memory” (M. J. Manning, ed.), pp. 15-26. Elsevier, Amsterdam. Van de Vyver, G., and Buscema, M. (1977). In “Developmental Immunbiology” (J. B. Solomon and J . D. Horton, eds.), pp. 3-8. Elsevier, Amsterdam. Van de Vyver, G., and De Vos, L. (1978). In “Biologie des Spongiaires” (C. Levi and N. BouryEsnault, eds.), pp. 233-237. Centre National de la Recherche Scientifique, Paris. Vosmaer, G. C . J. (1887). In “Klassen und Ordnungen” (H. Bronn, ed.), Vol. 11, pp. 369-496. Geest & Portig, Leipzig. Vosmaer, G. C. J. (1931). Acru Zoo/. 1, (3). 1. Weinbaum, G., and Burger, M. M. (1973). Nature (London) 244, 510-512. Weiss, P. (1941). Growrh Suppl. 5 , 163-203. Weissenfels, N. (1978). Zoo/. Juhrb. Anur. 99, 21 1-223. Weissenfels, N. and Striegler, B. (1979). Zoomorphologie 92, 49-63. Wilkinson, C., and Garrone, R. (1980). In “Nutrition in the Lower Metazoa” (D.C. Smith and Y. Tiffon, eds.), pp. 157-161. Pergamon, Oxford. Wilson, H. V. (1907). J. Exp. Zool. 5 , 245-258. Wilson, H . V. (1912). Bull. Bur. Fish. 30, 3-30. Yamada, K. M., and Olden, K. (1978). Nurure (London) 275, 17%184. Young, N. M., and Leon, M. A. (1978). Curbohydrure Res. 66, 299-302. Zahn, R. K.. Miiller, W. E. G., Geisert, M., Reinmiiller, J., Michaelis, M., Pondeljak, V., and Beyer, R. (1976). Cell Differ. 5 . 12%137.

This Page Intentionally Left Blank

INTERNATIONAL REVIEW OF CYTOLOGY, VOL. 77

Plant Movements in the Space Environment DAVIDG . HEATHCOTE Depcirtrnent o] Plant Science, Universiu College. Cardiff* Wales I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Characteristics of the Space Environment . . . . . . . . . . . . . . . . . . . . . A . Microgravity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

B . Environmental Periodicities. . . . . . . . . . . C. Radiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Other Environmental Factors. . . . . . . . . . . . . . . . . . . . . . . . . . . . 111. Experimental Constraints Imposed by Space Flight . . . . . . . . . . . . . IV. The Study of Plant Movements in the Space Environment . . . . . . . A. Plant Movements Which Can Be Meaningfully Studied under Space Flight Conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Flight Hardware for the Study of

............................. nts .......................... D. Current Projects for Space Flight E. Future Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I 83 I84 184 I86 I87 I87 I89 I90 I90 I94 I96 I99 202 203

I. Introduction The last quarter century has seen the gradual evolution of space flight technology from its early beginnings with Sputnik, to a current state of development in which manned orbital space stations, such as the Russian Salyut vehicles flown as part of the Cosmos program, have permitted prolonged experimentation in weightless conditions, and the advent of the American reusable Shuttle Orbiter System, which should provide experimenters with cheaper access to orbital flight than has hitherto been possible. Biologists have welcomed the past opportunities to utilize both manned and unmanned flights for basic biological investigations, which to date have been, of necessity, relatively simple in nature. Hopefully, more ambitious projects will be undertaken in the future using the European Spacelab payload on shuttle flights, starting in 1983, and future developments of the Salyut space stations. The interest of biologists in space flight opportunities can be divided into the two areas of “pure” and “applied” biology. The applied aspects, which have been the predominant interest to date, are concerned with the improvement of man’s functional abilities in space, including studies of physiological changes in man and “model” animal systems following exposure to the orbital environI83 Copyright 0 1982 by Academic h e i s . Inc. All rights of reproduction in any form reserved. ISBN 0-12-364477-1

184

DAVID G. HEATHCOTE

ment, and investigations of the effects of the spacecraft radiation environment on plant, animal, and bacterial cells and tissues. Pure biological research. takes advantage of a number of truely unique environmental conditions provided by space flight to investigate fundamental questions of animal and plant physiology. This utilization of space flight as a research tool for fundamental life science research is in its infancy. Astronomy, earth resources, geophysics, meteorology, etc. are many years ahead of biological sciences in the exploitation of the potential of orbital flight for pure research, and in the necessary technological expertise to complete such investigations successfully. In the inevitable competition for space flight opportunities and funding with the more established space user community, the biologist all too often appears naive and simple-minded in his approach to his more experienced colleagues. If the opportunities being offered by the availability of space flights for fundamental biological research ire not matched by logically argued, and technologically advanced proposals from the biological community, there is a real danger that pure life science research in space will be permanently relegated to an inferior position relative to the more established space science disciplines. The opportunities are there, but they must be earned. The last 10 years have seen a lull in American space flights, largely due to delays in the deployment of the Shuttle system. During this period, the USSR has undertaken an extensive series of manned flights, with the apparent aim of establishing a permanently-manned space station in earth orbit (Oberg, 198I). In this activity the Russians have surpassed the American program in tenns of total manned flights (49 to 32) and total man-days in orbit (2124 to 942). During these flights, Russian cosmonauts have undertaken many experiments with plants, but, unfortunately, few details of the majority of these experiments are available in the West, and no information is available about future experiments on USSR craft. For this reason, this review is, of necessity, biased toward discussion of experiments and planned projects of Western origin.

11. Characteristics of the Space Environment

A. MICROGRAVITY Probably the most important feature of the environment encountered during space flight as far as the investigation of plant movement is concerned is the very low residual mass acceleration force experienced by organisms within the spacecraft. This feature is sometimes referred to as weightlessness or zero-gravity; however, the preferred description should be “microgravity” since accelerative forces cannot be entirely eliminated in earth orbit, small g) residual forces must always be expected. All living organisms have evolved in a constant I g environment, and many plant movements are regulated, or affected by,

PLANT MOVEMENTS IN THE SPACE ENVIRONMENT

185

physiological responses to the direction of this constant field. Since the speed of response of plants to gravity is relatively slow compared with similar responses in animals, plant physiologists have for many years attempted to simulate low g levels with devices known as clinostats. Essentially these devices rotate the specimens about a horizontal axis at a rate which is fast enough to compensate gravity, the time-averaged force acting equally in all directions about the rotation axis, but too slowly to impose significant centrifugal acceleration forces. This technique, however, permits only an incomplete simulation of the microgravity condition, as the rotation can give rise to developmental artifacts (Zimmerman, 1927; Clifford, 1979), and can be shown to fail to remove completely g-magnitude-dependent responses in simulations or in actual space flight (Brown et al., 1976). Model experiments by Huisinga (1968) have shown that under clinostat rotation, sedimenting particles may precess around the periphery of a cell, possibly producing effective stimulation of georeceptors in this process. The plant physiologist wishing to investigate gravity-dependent, or influenced, processes is therefore limited to imperfect clinostat simulations of near zero gravity, or is forced toward more expensive and complex means of achieving microgravity environments for his experiments. True zero gravity cannot be easily obtained; exposure of test subjects to low, microgravity forces can be made by various means for durations of exposure ranging from seconds to months (see Fig. 1). These means include drop towers, aircraft in parabolic flight, ballistic rockets,

3

107 -

0

;lo6 9 a

g

-Month -Week

10'

'-Day

lo4

-

0

"1 h t e l l i tes

Manned

........................ Sat elli t es

-Hour

-

Sprelab nominal mission

CL

E 103 z

10

(RaketsJ

IMinute ACCELERATIVE G -LEVEL

FIG. I . A representation of the accelerative g levels and durations of exposure attainable by various means. Redrawn with modifications from Malmkjac et a / . (1981).

186

DAVID G. HEATHCOTE

and manned and unmanned spacecraft. Only the latter two space flight techniques can provide sufficient duration of exposure to low gravity to allow meaningful experiments on plant movements to be undertaken. Even in orbital flight, true zero gravity is unattainable; the low gravity environment is downgraded by various factors, such as firing of reaction control systems, crew activities (in manned flights), and various orbital accelerations, most importantly atmo:spheric drag (Malmkjac et af., 1981). Table I gives estimates for the Shuttle/Spacelab system of the g levels contributed by various factors, which may be ta.ken as representative for large, manned spacecraft. RCS firings, being of short duration, and crew activities, which characteristically result in periodic g loads (peak 1 Hz), are unlikely i q produce significant responses from the gravity detection systems of plants. The anticipated continuously present accelerations from atmospheric drag (3 X g maximum) are probably below the limit of detectability by plant geosensors.

B. ENVIRONMENTAL PERIODICITIES Many plant movements are known to have a circadian component--periodicities of close to 24 hours under constant conditions of temperature and lighting. These rhythms can be entrained. or synchronized, using various timing cues from the environment. Most investigators have concluded that such circadian rhythms are endogenous, and do not rely on periodic input from the environment for their functioning. However, one group of workers believes that they have evidence that circadian oscillations do require periodic timing information provided by environmental variables that are not controlled in the normal laboratory, such as magnetic fields, cosmic rap activity etc. (Brown, 1960; Brown et af., 1955). These "subtle, pervasive, geophysical periodicities" can be shown to have 24-hour periodicities, based on the rotation of the earth. In low earth orbit, any such periodicity would be lost, or replaced by new timings based on the orbital period of the spacecraft. Thus any persistance of circadian rhythms TABLE 1 LEVELSPKODUCEI> BY SOURCESI N THE SHIJTTL~SPACELAB SYSTEM"

ESTIMATEDRESIDUAL

ACCELERATION

VARlOLlS

~

Source

On orbit nianeuvers Crew activities Atmospheric drag" Gravity gradienth

Mode - 4 x 10-2 0 - 6 X 10-4 10-9 - 3 x 10-5 2 x 10-8 - 2 x 10-6 10-4

Data from Malmejac rt ul. (1981). Function of orbital altitude.

Step function 0.1-3 Hz Continuous Continuous

187

PLANT MOVEMENTS IN THE SPACE ENVIRONMENT

under orbital conditions could be taken as a refutation of Brown's theories of exogenous components in the mechanism of these rhythms, provided that sufficient care was taken to exclude timing signals reaching the experimental organisms originating from regular24 hourschedulingofcrew activities.

C. RADIATION The radiation environment of an orbiting spacecraft is different qualitatively and quantitatively from that of a ground laboratory. The background radiation will consist of protons-from cosmic rays (ca. lo4 protons cm sr- I d - I ) , and the magnetosphere (104-10s protons c m P 2 sr-I d - I ) in a typical orbit. In addition, heavy particle fluxes as detailed in Table I1 are to be expected. The total radiation environment is of interest to biologists concerned with radiation damage, mutation rates, etc., and many experiments have been carried on spacecraft in the past to investigate the impact of this environment on living material. However, there is no known or postulated mechanism whereby these radiation fluxes could interact with the physiology of plant movements, and therefore further discussion of this characteristic of the space environment is not appropriate here. -

D. OTHERENVIRONMENTAL FACTORS The space flight environment provides opportunities to use the unique conditions characteristic of orbital flight for the investigation of the physiology of plant movements. However, the use of spacecraft systems imposes other environmental conditions on the specimens and the equipment to be used for the investigation which are not the conditions of primary interest for the experiment. The possible influence of these factors on the plant material must be recognized, and, if possible, allowed for in the experimental data analysis. In general, these environmental factors are undesirable and detrimental to the conduct of the experiment, but they are a necessary and unavoidable consequence of the use of TABLE 11 TYPICAL COSMIC R A Y HEAVYN U C L ~Fl.UXES I PER SQUARE CENTIMETIR P t R STEHADIAN IN A N ORBIT INCLINED A l - 57"" ENCOUNTtKtiD DIIRINC A 7-DAY MISSION

400

80

20

20

10-1

('Only particles with an energy greater than 100 MeV nucleon- I are included.

2 x 10-3

188

DAVID G . HEATHCOTE

space flight as an experimental tool. The main factors which must concern the investigators in any space flight experiment are as follows. I . Atmospheric Composition

Manned spacecraft normally maintain a cabin atmosphere which is close to ground-level composition and pressures. It is likely that the slight differences in composition will be without influence on plant growth and development, except for the possible effects of carbon dioxide and the plant growth regulator, ethene. Carbon dioxide levels within the spacecraft cabin are maintained by lithium hydroxide or similar absorbent materials. In spacelab and shuttle flights, cannisters of absorbent are exchanged when exhausted, or on a fixed schedule. This results in periodic fluctuations of CO, partial pressure, but the levels will not be allowed to exceed 8 mbar CO, (Anonymous, 1979a). At these levels adverse effects on plant growth and physiology are not to be expected. Other trace gases may be present in the cabin atmosphere, arising from outgasing of materials, and emanations from the crew, experimental animals, and plants. For experiments with plants, the contaminant of most importance is the gas ethene (ethylene). Ethene is a potent regulator of plant growth, effective at concentrations as low as one part per billion (Abeles, 1973). It is a naturally occurring inhibitor of plant extension growth, and plays an important controlling role in several developmental processes in plants. It is actively synthesized by plant tissues, especially under stress conditions. Since the containers for growing plant specimens used in space flight are often of small volume, and since free gas exchange between the plant containers and the cabin atmosphere is ofttm discouraged, it is likely that ethene, synthesized by the specimens, may rise in concentration within the experimental containment to a level at which it has serious effects on the growth of the specimens. Therefore, steps may have to be taken to avoid this problem, for example by providing ethene absorbent materials within the containment. 2 . Launch and Reentry Stresses During the launch and reentry phases of a space mission, any living material on board will experience accelerations and acoustic and vibrational streses of considerable magnitude. These conditions may modify the responses of the specimens to any subsequent experimental stimuli. Previous experience with hydrated plant material, however, suggests that gross morphological or developmental effects are not to be expected from the levels of acceleration, noise, and vibration encountered during launch or recovery phases of a flight (Johnson and Tibbitts, 1968; Lyon, 1971). The launch accelerative forces predicted for the Shuttle/Spacelab system are not in excess of 3 g, well within the known ability of mature and seedling plants to survive. For the same spacecraft system, acoustic noise from rocket motor ignition and aerodynamic sources experienced during

PLANT MOVEMENTS IN THE SPACE ENVIRONMENT

189

launch at some point within the Spacelab structure is estimated to be the equivalent of an overall sound pressure level of 196 dB referred to 20 pN m-* lasting for 6 seconds. Random vibrations during launch may be as high as 6.5 g rms for specimens mounted in Spacelab racks (Anonymous, 1979b). These launch stresses, although clearly survivable by hydrated seed or seedlings, may well be sufficiently severe to produce physiological effects in the experimental subjects which may interact with the experimental treatments given in orbit. The period of recovery from such stresses will vary from species to species, and experience with one species in space flight conditions cannot necessarily be extrapolated to another. Therefore it is important that any flight experiment which involves exposure of plants in a hydrated condition to the stresses of the launch phase should include appropriate ground controls which can be used to separate the effects produced by the launch stresses from those due to the desired experimental variable. Clearly, if the experimental plant material can be grown in orbit from dry seed or other propagules, the importance of exposure to the stresses of launch are minimized, since dormant material is not likely to be capable of sensing or responding to such stresses. While there is no a priori reason to suppose that a seed hydrated under microgravity conditions will not germinate normally, and several species have in fact been successfully germinated in space during the Russian Cosmos program, appropriate in-orbit hydration and seed holding techniques for a particular experimental species must be devised and demonstrated in space before such techniques can be used routinely for experimental purposes. There are, for example, unconfirmed reports of at least one Cosmos experiment in which unexpectedly poor germination resulted following on-orbit hydration. This failure has been ascribed to excessive capillary “wicking” of water, in the absence of the gravitational field, leading to waterlogging of the seed, anaerobiosis, or other deleterious consequences leading to the poor germination.

111. Experimental Constraints Imposed by Space Flight

Any space flight experiment does not exist in isolation; it is part of a complex mission program which will typically consist of tens of experiments. For example, the first dedicated life sciences Spacelab mission, SL 4,is planned to consist of some 25 experiments sharing a 7-day flight. All these experiments compete for the resources available during the mission. Within the mission context, therefore, individual experiments are constrained, most importantly in the total number of individual specimens that can be accommodated, and the amount and timing of crew involvement with experiment-specific manipulations. In the present context, plant experiments are unlikely to be given high priority in comparison with experiments on human and animal physiology. These and other prob-

190

DAVID G. HEATHCOTE

lems associated with the inherent difficulty of performing experiments in an alien environment have tended to result in experiments which are relatively simple, often with rather minimal replication. Other problems arise with experimental manipulations which, although quite routine on earth, are in fact dependent on gravity for their success. In orbit, methods have to be devised to overcome these problems. For life science experiments, one crucial general problem is that of fluid handling, and the behavior of liquids in weightlessness. These questions become important, for example, when considering the hydration of dry seed in orbit, or for chemical fixation of' specimens during flight. Previous experiments (e.g., Gray and Edwards, 1971) have used fixation with simple formalin-acetic acid-alcohol formulations. For many possible investigations in plant movement physiology, and other areas of biological interest, better quality fixation, suitable for use with the electron microscope, is desirable, if not essential. In the plant movement field, much interest is focused on the ultrastructure of geoperceptive cells, in which lipoprotein membranes are thought to be implicated in the gravity sensing function. There are several, nontrivial problems that need to be overcome to allow sufficient quality fixation to be accomplished in flight to address such questions as are posed in these and similar physiological systems: (a) How to ensure efficient specimen/ fixative contact under zero g conditions? (b) To devise means of accomplishing complex fixation protocols, involving several fixation and buffer washing stages necessary to achieve the required standard of preservation. (c) To ensure complete and demonstrable containment of all toxic fixation chemicals within the experimental area. (d) To devise means of avoiding degradation of fixatives during storage in the prelaunch and early flight stages. In Spacelab operations, fixatives will probably need to be stowed on board Spacelab itself for a minimum of 10 days before launch, although it may be possible to store fixatives temporarily in the Shuttle lockers, to which later access is possible, some few hours before launch. Of these problems, perhaps the most complex is the requirement for methods of performing multistage fixation protocols in orbit. It is at least possible that semiautomatic means may have to be devised to avoid excessive demands on the time of the crew members being made by experiments with a requirement for high-quality fixation.

IV. The Study of Plant Movements in the Space Environment A. PLANTMOVEMENTS WHICHCAN BE MEANINGFULLY STUDIED U~DER

SPACEFLIGHT CONDITIONS In view of the complexity and expense of space flight operations, it is clear that only experiments which have direct space relevance will be likely to be selected for flight. Such experiments are those in which the unique characteris-

PLANT MOVEMENTS IN THE SPACE ENVIRONMENT

191

tics of the space environment can be used as tools to probe physiological effects and processes which can only be studied under these conditions. It is most unlikely that experiments will be flown to study problems that could equally well be studied in a ground laboratory. The various classes of plant movements which appear at present to be “space relevant” are considered below. As our understanding of the conditions of space flight and the responses of plants to this unique environment improves, it is possible that other types of movement may need to be added to the list. I . Taxes

Various freely motile plants exhibit directional movements, brought about by means of flagellae etc., in response to stimuli such as light or chemical gradients. In suspension culture, sedimentation of individuals under the force of gravity form a background against which such directed response movements must occur. Apart from this sedimentation effect, there is no evidence of a gravity-dependent mechanism at work in controlling or modifying taxes, and consideration of the normal “tumbling” motion of the motile algae makes it unlikely that any gdirection-dependent system can be operational in these organisms. However, many phototaxes (motile responses to light or light gradient stimuli) are known to have circadian components (Sweeney, 1979; Hader, 1979). It should therefore be possible to use the taxes of suitable green algae to devise tests of the dependence of circadian rhythmicity on subtle 24 hour geophysical periodicities.

2 . Tropisms Many higher plant organs-roots, shoots, stolons, etc.-respond to gravity and light stimulations by growth movements brought about by differential growth, resulting in bending movements directed toward, away from, or at a fixed angle to the direction of the stimulus. These responses are termed tropic responses and can be qualified by the prefix geo- or gravi- to denote gravitydirected movements, or photo- to denote responses to directional light stimulations. The geotropic response is a clear example of a plant physiological process which is known to be gravity dependent, and which is therefore amenable to investigation under microgravity. Geotropic response consists of a perception phase, in which changes in the orientation of the plant organ with respect to the gravity vector is detected, and in some way a future response committed, followed by a latent phase in which it is presumed that control signals, often thought to be hormonal in nature, pass between the site of detection and the responding region. A complex response phase follows in which differential growth of opposite flanks of the organ results in curvatures directed such that the original deviation from the desired relationship with respect to the gravity vector is corrected. The perception phase is now known to be completed in many plants within ca. 30 seconds of the onset of stimulation (Johnsson and Pickard, 1979;

192

DAVID G . HEATHCOTE

Heathcote, 1981). During this period, it is widely thought that aggregations of starch grains (amyloplasts) sediment within specialized cells under the influence of gravity and in some way interact with other components of the cell to trigger the ultimate bending response. The specialized cells, or statocytes, are located in the root cap tissues in the root system, and in the starch sheath in stem tissues. The amyloplasts in these cells sediment freely under gravity, unlike normal storage starch, and are known as statoliths, a terminology which implies their status as geoperceptive structures, although at present this widely presumed function has not been unequivocally proved. Several theories about how the statocyte-statolith system might function to allow the direction of the g vector to be sensed exist. The system may be thought to sense the new distribution of the statoliths after sedimentation has occurred, or the sliding of the statolith along cytoplasmic structures during sedimentation. Alternatively the cells may detect the change in the pressure exerted on the original site of sedimentation as the statoliths sediment away following reorientation (Volkmann and Sievers, 1979). There are experimental results which appkar to support each of these models, but it has not yet been possible to obtain definitive proof using earth-based experiments. Certainly the availability of a relatively gravity-free environment should allow investigators to design experiments to test these hypotheses, and so help to resolve this most fundamental question of geoperception in plants. The use of low-speed certrifuges on space craft will allow exploration of the threshold stimulus required for geotropic response in a more soundly based manner than is possible on the ground. At first sight, phototropic responses are not obviously gravity dependent, or likely to be modified significantly by the other unique features of the space environment. However, as soon as a plant responds phototropically by bending in a light field, it consequently and unavoidably initiates a geotropic response, since its orientation with respect to gravity changes. Thus it has always been impossible to observe a “pure” phototropic response on earth (Wilkins, 1977). The use of an orbiting laboratory may permit the first observation of phototropic response unaffected by gravitational interference. During the development of a tropic curvature, the pattern of differential growth is complex, finally resulting in a straightening of initially curved, distal portions of the responding organ, while more proximal portions remain strongly curved (Firn and Digby, 1979). Similar straightening phenomena have also been observed to take place during nutations (see Section V,A,4; Heathcote, 1965). The straightening responses have been termed “autotropism,” implying a response of the plant to tissue level stresses, or some other internal feature characteristic of a curved organ. It has been suggested alternatively that autotropic responses are gravity dependent, despite evidence to the contrary. Clearly, observation of autotropic behavior under microgravity conditions would finally eliminate any possibility of g dependence in this response.

PLANT MOVEMENTS IN THE SPACE ENVIRONMENT

193

3. Nustic Movements The angle which lateral organs-secondary roots, branches, and leaves, etc.make to the primary axis of the plant is determined by external factors, such as illumination, and internal factors relating to developmental status, etc. Movements of these organs leading to changes in the angle relative to the main axis are known as nastic responses, and can be induced by changes in illumination level, and other factors. The movements are brought about by differential growth or turgor pressure changes, usually near the point of insertion of the organ to the main axis. Clinostat studies have shown that the simulated zero gravity produced during rotation leads to changes in the angle between the leaves and the main axis-an epinastic response. The question of whether this response would be found in true microgravity conditions in orbit is one of the very few experimental investigations relating to plant movements that have been undertaken, forming part of the payload of Biosatellite I1 (see Section V,C). 4. Nutations A11 apically growing plant organs exhibit oscillatory movements during active growth, with the tip of the organ following an elliptical or near-circular trajectory about the mean growth direction. These movements characteristically have a period length of I to 5 hours duration, and an amplitude which varies from a few millimeters to 10 or more centimeters in vigorously growing climbing plants. There have long been two rival schools of thought concerning the underlying physiological basis of nutational movements. The earliest theory ascribed nutation to the functioning of an endogenous, autonomous growth control oscillator system. This concept was first proposed by Dutrochet (1843) and supported by Darwin (1880), and more recently by Arnal(l953, 1959), Heathcote (1968), and Heathcote and Aston (1970). This theory views nutation as separate and distinguishable from geotropic responses, although both processes act via modulations of the growth rates within the moving organ, and therefore will interact during their expression. The alternative school of thought (Gradmann, 1926; Johnsson and Israelsson, 1968) interprets nutational movement as a feedback oscillation of the gravity-sensing geotropic growth control mechanisms of the plant. A full discussion of the arguments in favor of each standpoint was provided by Johnsson and Heathcote (1973), who suggested that a definitive experiment to distinguish between the two theories could be performed in a microgravity environment during a space flight. An experiment and the required flight hardware to achieve this aim has been developed by A. H. Brown and his collaborators at Philadelphia for flight on the joint NASAIESA Spacelab demonstration mission, SL 1 , which is scheduled for launch in 1983. This experiment will look for nutational movements which persist after the removal of seedlings from a centrifugal 1 g field to microgravity conditions. Prolonged persistence of the rhyth-

194

DAVID G. HEATHCOTE

mic movement will indicate an endogenous origin as opposed to a geotropic feedback mechanism. Whatever the outcome of this experiment with a seedling epicotyl (Helianthus), it would be of interest to repeat the observations for the movements of a tendril-bearing climbing plant, in which nutation is quantitatively and possibly qualitatively different from that exhibited by noncliaibing plants. €3. SOMEEXAMPLES OF CURRENT FLIGHT HARDWARE FOR

THE

STUDYOF

PLANTMOVEMENTS The underlying philosophy of the Shuttle/Spacelab system is that, whenever possible, hardware should be reusable to maximize the potential scientific return on the development and flight-qualification program. Once flown on a NASA mission, equipment will be retained by NASA as Life Sciences Laboratory Equipment, an inventory of items which may be requested by future users (Schachter and Tyler, 1979). Many biological experiments require common facilities; in the case of research into plant movements such requirements are likely to include one or more of the following facilities: (a) A temperaturecontrolled incubator for the support of plant growth in space; (b) controlled conditions of illumination or constant darkness; (c) facilities for periodic recording of the positions of moving plant organs, typically by photographic or similar means; (d) facilities which allow fixation of whole or part specimens; (e) facilities to expose plant specimens to a 1 g centrifugal acceleration. This latter, rather paradoxical, requirement is needed to serve to orient plant growth during preexperimental development, to act as a 1 g control to separate true microgravity effects from those resulting from exposure to other spacecraft environment features such as launch stresses, cabin atmosphere, etc., and in some experiments to provide a mass acceleration stimulation. The following brief account gives an outline of some hardware items currently fabricated or projected for use in the Shuttle program. Unfortunately, although Russian investment in life sciences research in space has been considerable over the past decades almost certainly exceeding in cost even the peak United States efforts, the few details available of such space flight apparatus as “oasis,” “biofiksator,” and their 1 g plant centrifuge preclude any useful discussion at the present time. 1. Plant Growth Unit

The plant growth unit (PGU) was developed for the experiments of J . Cowles and W. Scheld of the University of Houston on plant lignification processes under space flight conditions. Initial design and development of the PGU was undertaken by Lockheed Missles and Space Co. (Maine et al.. 1979). Subsequent development and construction of the flight model has taken place at NASA Ames Research Center. The unit provides illumination and temperature control

PLANT MOVEMENTS IN THE SPACE ENVIRONMENT

195

on a dayhight cycle to six hermetically sealed chambers, each of which is able to accommodate 16 seedlings in two rows at a plant spacing of 2 cm. Septa fitted to ports in the base of each chamber permit gas purging and sampling proceedures. The design specification calls for a day temperature of 78 ? 1°F and a night temperature of 74 .t 1°F. Lighting is provided by fluorescent tubes giving a uniform flux density in the central four chambers of 85-90 kmole m - z sec- I in the photosynthetically active spectral range. Owing to losses in the end caps of the tubes, the flux density in the end two chambers of the PGU is said to be of the sec - I . Temperature and light status sensors are proorder of 55 Fmole m vided, and the data collected on an internal digital tape recorder at 15-minute intervals throughout the flight. In use, the PGU will replace one of the standard Shuttle mid deck lockers, and is intended to operate autonomously throughout the flight. A limited amount of data is available in real time, on a small panel display, for periodic functional checks. There is no facility for automatic photographic recording of the plant growth during flight, but individual chambers could be removed for photography under manual control. ~

2. The HEFLEX Hardware This equipment was developed at the University City Science Center and the Department of Biology of the University of Pennsylvania at Philadelphia by A . H . Brown and co-workers. The apparatus will support experiments, on the nutation of Helianthus seedlings, selected for the first Spacelab mission in 1983. Within a Spacelab rack are located a temperature-controlled I g rotor chamber, a “dark box” designed for infrared sensitive video imaging of plants under weightless conditions, and associated equipment for display and time lapse recording of the video data, and a control microcomputer. Additional space in the rack provides storage facilities. The rotor chambers and the dark box area will be temperature controlled at a set point above the ambient cabin temperature to within 1°C. Plant specimens are held in cylindrical modules which can be attached to fittings on the rotors or in the dark box. The rotor chamber contains two independently controlled rotors, each with a capacity of eight modules, which are capable of exposing the plants to centrifugal forces between 0 and 1 g. For the experiment with Helianrhus the rotors will provide a I g field for the orientation of seedling growth. Each module is light-tight except for IR transmitting windows, in the end and side walls, used for imaging purposes. The rotor chamber is equipped with IR-sensitive video cameras which can view the development of the seedlings on the rotors to allow the crew to make selections of suitable plants for experimental use. A small video monitor screen is built into the rack to facilitate the selection process. A similar video camera in the dark box permits time lapse records of the movements of plants under zero g conditions to be made. Although this apparatus was designed specifically for one investigation, it also has the potential to be used for a variety of experiments in various

196

DAVID G . HEATHCOTE

physiological disciplines. The design is modular, and, for example, the rotor system alone could be flown to provide 1 g control conditions.

3 . Biorack Biorack is a project of the European Space Agency which, at the time of writing, has not yet received formal funding for completion. A decision on this program is expected within a few weeks. The Biorack concept is a multiuser complex configured to fill a Spacelab half-rack consisting of several common facilities. The proposed design incorporates two temperature-controlled incubators, a refrigerated compartment, a glove box for specimen manipulations, and the necessary control electronics. The incubators are of two types, capable of maintaining set points to 0.5"C within the ranges 18-30 and 3040°C. Each incubator can accomodate 32 small specimen containers with a volume of approximately 50 ml, 4 larger containers (250 mi), and a small I g control centrifuge which can carry eight of the small containers. The refrigerated compartment has a temperature set point range of 48°C. The glove box unit is designed to provide positive containment of specimens and fixatives during specimen handling procedures, and is also intended to be fitted with a microscope and camera system. No provision for photography within the incubators is possible, and any essential illumination must be experimenter provided within the experiment containers. The small size of these containers precludes the use of other than very small seedlings or microorganisms in biorack experiments. 4. General Purpose Work Station

This item has been developed by NASA as an item of Life Sciences Laboratory Equipment, which, it is forseen, will have potential uses on many future Spacelab flights. It is accommodated in a double-width Spacelab rack, and provides a working surface large enough to support a variety of experiments, including microscopy and photography. The facility also provides control and containment of liquids and chemical vapors within the work station, and prevents the release into the spacecraft of materials in use within it. It is forseen that specimens will be transferred to the GPWS from their experimental locations for fixation and other manipulations.

C. PAST SPACEFLIGHTEXPERIMENTS There have been very few space flight experiments carried out to date which have a direct bearing on the study of plant movement physiology. Biosatellite 11, an unmanned satellite, was launched in 1967 and exposed several biological experiments to weightless conditions for a duration of 44 hours, rather shorter than the planned duration of 70 hours, because of storm conditions prevailing at the recovery site. In one of the experiments in the payload, wheat seedlings,

PLANT MOVEMENTS IN THE SPACE ENVIRONMENT

I97

supported in special holders, were supplied with water from the endosperm end, allowing growth and development of the seedling organs to take place freely in a moist atmosphere. The seeds were placed in the holders 15 hours before launch, and therefore experienced the launch ,g stresses in a hydrated condition. Vibrated and nonvibrated, erect and clinostat controls were carried out on the ground for comparison with the flight specimens. There were no differences in the growth of primary roots or coleoptiles, but a slight, statistically significant difference in the mean length of the lateral roots between both erect and flight speciments, which were both slightly shorter than the clinostat controls (Lyon, 1971; Gray and Edwards, 1971; Gordon, 1973). The orientation of the seedling organs in threedimensional space was significantly different in erect and flight specimens, and the clinostat controls were similar to the flight experiment in this respect, indicating that for this nastic, or tropic response, the clinostat appeared to be an adequate model of microgravity conditions. However, the experiment gave no information regarding the kinetics of the response, since only end points were recorded. Histological studies of specimens fixed with an FAA spray during flight, and later stained using the PAS reaction by Gray and Edwards (1971), showed that the amyloplast statoliths of the presumed geoperceptive cells in the rootcap and certain cells of the coleoptile were randomly distributed within the cells of the flight specimens, in marked contrast to the sedimentation observed to the bottom of these cells in erect plants. The flight pattern of statoloth distribution was mimicked in the clinostat rotated specimens. This sedimentation behavior is consistent with the postulated function of these cells as gravity receptors, but makes no real contribution to the understanding of the perception mechanism. Study of the ultrastructural details of the statocyte tissues was not attempted, and the short duration of the flight might not have provided sufficient exposure to microgravity to bring about marked changes within the putative receptor cells. In an experiment flown on a Soyuz mission in 1973, pea seeds of two varieties were allowed to complete the process of germination in orbit (Merkys er al., 1975). The seeds were presoaked for 4 hours and allowed to germinate on the ground at 20°C for about 2 I hours, before being planted in a sterile container and transferred to 4°C to arrest development of the seedlings during the transport and launch operations. On orbit, germination was allowed to proceed for 24 or 48 hours by warming the seedlings to a suitable temperature (23°C). After the completion of the experiment, the seedlings were fixed in Carnoy’s fixative for later examination. It is not entirely clear from the literature cited, but it is most probable that this fixation procedure took place on the ground, after the recovery phase of the mission, rather than during orbital flight. A 20-25% greater root length was reported in the flight specimens compared with the controls, reflecting statistically significant increases in the length of cells in the extension zone and the root cap of the flight plants. The presence of statolith amyloplasts within

198

DAVID G.HEATHCOTE

the root cap tissues was noted, showing that the development of these organelles was not affected by the conditions experienced during flight. This finding would be of greater interest to gravitational physiologists, had the seed been germinated entirely in microgravity conditions in space, since the exposure to 1 g for at least 24 hours before flight, and the launch forces might well provide a necessary gravitational stimulus for the initiation of statolith development. No mention is made of the distribution of statoliths within the cells of the root cap; since in all probability seedlings experienced 1 g exposure postrecovery, before fixation, during which statolith reorientation would be expected, such data would be without real value. Three Russian flights in the Cosmos series have studied the growth and development of the geotropically sensitive fruiting body of the fungus Polyporus brumalis under prolonged weightlessness. Flights were made on board the Isatellite Cosmos 690 (20.5 days), and the space stations Salyut 5 (17 days) and Salyut 6 (20 days). Fruiting bodies developed under weightless conditions, both in the dark and light, although the development of caps was light requiring, as on earth. The experimental fruit bodies were markedly distorted, showing random bends, twists, and spirals. In some cases the fruit body assumed an unusual, flattened form. In all cases the fruiting body structure contrasted markedly with the normal, erect, tubular pileus structure found under 1 g growth conditions. (Zharikova et al., 1977; Kasatkina et al., 1980). It seems that the gravitational field is essential for the normal control of fruit body morphology in this fungus. This points to a fundamental difference in the control of morphology between the higher fungi and seed plants; the rather limited experience we have with higher plants suggests that a seedling organ will grow in microgravity conditions in a direction determined by the orientation of the seed or plant in its holder, but with no marked disturbance of its normal morphology in contrast to the fungal response to weightlessness. Two space flight experiments have attempted to investigate phototropic responses in orbit. The American Skylab Student Project experiment ED 62 attempted to photograph the growth of rice seedlings in a directional light field. The seedlings were grown in completely agar-filled containers, and the reported lack of phototropic response may be due to this factor, or to other unsatisfactory aspects of the experiment (Anonymous, 1974). In complete contrast to this reported lack of phototropic response, the Russian Cosmonaut, Klimuk (Klimuk and Baburina, 1976) reports that “plants in weightlessness react to light much more sensitively and much more rapidly than on Earth” (NASA translation). However, the scientific observations on which this assertion is based have not yet been made available in the West so that reasoned assessments of the data cannot be made at present. Lateral branches and leaves of higher plants normally take up an orientation which has a fixed angular relationship to the gravity vector. Such behavior is

PLANT MOVEMENTS IN THE SPACE ENVIRONMENT

199

termed plagiotropic. Under certain conditions, the angle between a plagiotropic organ and the orthogeotropic axis to which it is attached may be influenced, bringing about nastic curvature. Increases in the angle, resulting from epinastic curvatures, have been observed to occur as a result of clinostat rotation. The hypothesis that similar epinastic curvatures would result from exposure to microgravity was tested in the “pepper plant experiment” flown on Biosatellite I1 (Johnson and Tibbitts, 1968). In this experiment four 28-day-old plants of Capsicum were placed in orbit in a capsule designed to allow photography of top and side views of each plant at 10-minute intervals throughout the flight, and for 20 hours postrecovery. As predicted by the experimenters’ hypothesis, all flight plants exhibited marked epinastic curvatures of the leaf petioles, which developed rapidly following orbital insertion, and reaching a fairly constant maximum value at about 20 hours into the flight. The mean angular displacement of the leaves after 20 hours with reference to the initial orientation at launch was of the order of 50”, but plant to plant variability was high. A series of clinostat and stationary erect control experiments, and a partial simulation of the launch accelerations and vibrations, were carried out on the ground. The launch stress simulations did not influence the curvature produced on the clinostat (Johnson and Tibbitts, 197 1). The original experimenters concluded from the experimental data that the behavior of plants in weightlessness and on the clinostat was “very similar,” but the death of the principal investigator, S. P. Johnson, prevented full analysis of their data. Brown et al. (1974) undertook a complete reanalysis of the film record from the pepper plant experiment, and were able to show that “the average kinetics of onset and recovery from epinasty, as well as the maximum excursion in mean petiole angle which was obtained, were not the same for plants exposed to weightlessness and for plants rotated on an earth based clinostat.” Initial curvature rates on the clinostat were significantly greater than in flight, but the maximum response was less, all differences being significant at the 1 % level. The flight plants also took significantly longer to recover from the epinastic curvature after restoration of the 1 g field after reentry. This reassessment of the original findings of the experiment therefore concluded that the behavior of clinostated plants did not correspond to the movements observed in orbital flight, and that therefore the clinostat did not provide an adequate model of microgravity as found during space flight.

D. CURRENT PROJECTS FOR SPACE FLIGHT The first flight of the European Space Agency Spacelab in 1983 will start an era in which more ambitious studies of plant movements can occur. The Spacelab, carried into orbit by the Shuttle, will provide pressurized working and equipment space for a crew of up to seven individuals. On biological missions it is likely that at least one of the crew members will have expertise in an appropri-

200

DAVID G . HEATHCOTE

ate biological or medical discipline. The availability of working biologists in the orbital laboratory will clearly be of great advantage to experimenters who rely on the crew to perform complex manipulations on their specimens. The first Spacelab flight will include the most advanced equipment yet devised for experiments on plant movements in space, the HEFLEX hardware (see Section IV,B,2). This equipment will support an experiment by A. H. Brown and D. K. Chapman which will provide a definitive test of the hypothesis that nutational movements arise as a result of geotropic feedback oscillations. Sunflower seeds will be planted in modules on the ground before launch, and during the flight. Once in orbit the modules will be transferred to the temperature-controlled rotors to allow orientation of the early stages of seedling growth to take place in this 1 g field. At various times during the mission, subsets of the modules will be selected by the responsible crew member using the rotor chamber cameras and the monitor screen. The selected modules will be transferred to the dark box area, where, under weightless conditions, time-lapse video records of the seedlings will be made. Later analysis of the video record will reveal whether nutational oscillations persist under microgravity-a critical test of the geotropic feedback hypothesis. Beyond this first Spacelab flight, planned reflights of the system should provide further opportunities for plant movement research. A mission funded by Germany is planned for 1985 (D I), which may include the ESA Biorack and other life science experiments in its payload. NASA is planning a series of Spacelab missions dedicated to research in the life sciences. The first of these dedicated missions, designated as SL 4, is currently scheduled for flight in late 1985. Future life science flights are tentatively planned at two yearly intervals following SL 4, but the frequency of missions is dependent on budgetary considerations. The tentatively selected payload for the SL 4 mission includes two experiments on plant movements which plan to reuse the HEFLEX hardware developed for SL 1. One, a joint US/Norwegian experiment will investigate the threslihold stimulus required for geotropic reaction using modified HEFLEX rotors to give stimulations of various durations and at various levels of acceleration between 0 and 1 g. The experimenters involved, A. H. Brown (Philadelphia) and A. Johnsson (Trondheim, Norway) will use Avena coleoptiles for this test of the reciprocity rule and the limits of geosensitivity of the geotropic response mechanism. The second SL 4 experiment is the result of cooperation between Dr. Brown and D. G. Heathcote (Cardiff, UK) and is intended to explore the reactions of seedling wheat coleoptiles following lateral stimulations with pulses of blue light. This experiment is expected to provide insight into the nature of the phototropic reaction in the absence of the complicating gravity force, to see whether phototropic stimulations result in the initiation of nutational oscillations, and to observe any autotropic straightening reactions that may occur under the wcightless conditions.

PLANT MOVEMENTS IN THE SPACE ENVIRONMENT

20 1

As mentioned earlier, the European Biorack has not been, as yet, formally approved, and selection of a tentative payload of experiments for Biorack will not be made until the early part of 1982. However, it is possible to give some pointers to the direction of European interest in the plant movement field, that may form part of the payload of Biorack or similar payload elements on the D 1 mission. There is a general interest shown by European centers in investigations of the fine structure of the presumedly geoperceptive statocyte cells of higher plants. For reasons of accesibility and speed of development the root cap statocytes are the preferred material for most of these groups. For these studies it will be necessary to arrange for germination to be started by hydration during orbital flight. The investigations will also demand the resolution of cellular components at the ultrastructural level, concentrating initially at least on the distribution of endoplasmic reticulum within the statocytes (Volkmann and Sievers, 1979). The feasibility of these investigations therefore depends critically on the ability to achieve high-quality fixation under weightless conditions. Such critical preservation of higher plant tissues has never been attempted in orbital flight as far as this author is aware, and the problems to be overcome to satisfy the requirement are by no means trivial. However, it is likely that a successful outcome of this type of experiment would be of great interest to gravitational physiologists, and should add to our current, lamentably poor, understanding of the mechanism of geoperception in plants. A different approach to the problem of geoperception has been proposed by D. Osborne and B. Juniper (Oxford, UK). They suggest that a simple series of centrifugal manipulations of the statoliths within a geoperceptive organ could demonstrate whether the pressure of statoliths against the lateral wall of the statocyte, or the mere presence of statoliths close to this wall is necessary for the initiation of geotropic response. They suggest that a brief, laterally directed centrifugal 1 g field should move the statoliths into contact with the lateral walls of the statocyte. Prolonged centrifugation would, in addition, cause these statoliths to exert a pressure equivalent to the pressure that would be produced in horizontally orientated statocytes under a normal gravitational field upon the wall, and any overlying sensitive structures. Analysis of the curvatures produced as a result of these treatments should allow conclusions to be drawn as to the validity of these models of geoperception. However, the success of this approach depends on the statoliths remaining near the lateral wall following a brief centrifugal treatment. Recent reports of active cyclosis within statocyte cells, which can move the amyloplasts through the cytoplasm (Clifford and Barclay, 1980; Heathcote, 198l ) , may suggest that prolonged residence of the statoliths near the lateral wall is not assured by short centrifugation treatments. A satisfactory experiment based on these manipulations would therefore need to include a microscopic examination of specimens fixed in flight to confirm that the centrifugal treatments did result in the predicted distribution of amyloplasts within the

202

DAVID G . HEATHCOTE

statocytes. There is no doubt that the availability of a microgravity laboratory will encourage plant physiologists to use various treatment protocols in tests of the many varied hypotheses of geoperception. Osborne and Juniper’s suggested manipulations of the statoliths are, no doubt, only the first ideas in what may turn out to be a fruitful field of investigation. Proposals have also been made for experiments to study the persistence of circadian rhythmicity in the time-cue-free environment of orbital flight. D. Mergenhagen and 1. Voelker (Hamburg, West Germany) have proposed an instrumented container for Biorack which would automatically monitor the phototactic circadian rhythm in the motile alga, Chlamydomonas reinhardii. Within the rather limited volume available in a standard large Biorack container, it should be feasible to monitor accumulations of the alga in a light beam in up to 12 small cuvettes, thus providing a reasonable degree of replication (Mergenhagen , personal communication). This proposed experiment would observe the persistence or otherwise of the free-running circadian rhythms of this species in an environment that was unequivocally devoid of 24 hour eivironmental periodicities which might serve as timing cues for the biological rhythm. E. FUTUREPERSPECTIVES There are many aspects of research in plant movements, especially those concerned with gravity perception, or with processes in which the gravitational force produces interactions, where the availability of opportunities for the use of space flight should prove invaluable. However, these rather limited opportunities are not now, and are unlikely to become, easily available to the biological community in the near future. Considerable ingenuity is required to devise suitable experiments and the necessary experimental hardware, and a sustained effort on the part of the experimenters, currently over a period of at least 7 years, is necessary to bring an individual experimental idea to fruition. The various constraints imposed on the experiment by safety requirements and the physical limitations involved in a space mission are daunting, but it is clear that they must be overcome, if the unique space environment is to fulfill its potential as a investigative tool for the biological sciences. Perhaps the biologist should learn from the experience of other disciplines more traditionally concerned wilh the utilization of space, for example, by setting up organized user communities which could define experimental goals, and discuss and devise means of achieving them, rather than relying on uncoordinated individual efforts as is largely the case at present. User groups are set up by NASA, but only afrer proposals have been formulated by the individual investigators, and at present, the European investigators are even less organized to speak and plan as a united group with common interests in the utilization of space flight opportunities. Such interest group cooperation should result in better designed experimental investigations,

PLANT MOVEMENTS IN THE SPACE ENVIRONMENT

203

employing more advanced technologies appropriate to the relatively high cost of the projects. In the past, too many biological experiments in space have produced rather inconclusive results, not necessarily as a fault of the experimenters, but more a reflection of the constraints imposed by the system and the real difficulties inherent in the exploitation of space. The wider biological community, largely in ignorance of these problems, has found it rather easy in the past to criticize these early attempts to perform biological experiments in space, contributing to a somewhat antagonistic attitude of many senior scientists to the very concept of pure biological research in the microgravity environment. Those of us who have an interest in problems of plant physiology which can only, or best, be addressed by the utilization of the unique opportunities afforded by access to space have a duty to ensure that proposed experiments are demonstrably “space relevant,” as technically sophisticated as required to ensure a high chance of a successful conclusion, and well supported by appropriate ground-based controls. Otherwise, the prejudices of the scientific community may be reinforced, making it even more difficult to undertake worthwhile experiments in the future. Biologists can certainly pose relevant questions for space exploitation programs; the technologies to address them are available. All that is needed is patience, care, perseverance, and a modicum of dollars to ensure successful achievement of relevant and theoretically important goals. The next decade shows promise of producing fundamental advances in our understanding of the physiology of plant movements that would not have been possible but for the advancement of man’s space technologies from the stage of exploration to that of exploitation of the space environment for scientific research.

REFERENCES

Abeles. F. B. (1973). “Ethylene in Plant Biology.” Academic Press. New York. Anonymous (1974). “Skylab Student Project Report,” pp. 4 6 5 1 . NASA TM X 64866. Anonymous (1979a). “Spacelab Payload Accommodation Handbook.” ESA SLP/2104-3. Anonymous (l979b). “Spacelab Users Manual,” p. 104. ESA DP/ST(79)3. Amal, C. (1953). Ann. Llnir.. Sarav. Naturw. Sri. 2 , 92-105. Arnal. C . (1959). Congr. Soc. Savanres. 8th. pp. 461466. Brown, A. H.. Chapman, D. K., and Liu. S. W. W. (1974). Eioscience 24, 518-520. Brown, A. H.. Dahl. A. 0.. and Chapman, D. K. (1976). Plant Physiol. 58, 127-130. Brown, F. A. (1960). Cold Spring Harbor Svrnp. Quant. B i d . 25, 57. Brown, F. A,. Webb, H. M., Bennett, M. F., and Sandeen. M. I. (1955). B i d . Bull. Wood’sHole 109. 238-242. Clifford. P. E. (1979). Z. Pflanzenphysiol. 95. 465469. Clifford, P. E., and Barclay. G. F. (1980). Planr Cell Environ. 3, 381-386. Darwin, C. (1880). “The Movements and Habits of Climbing Plants.” Murray. London. Dutrochet, R. (1843). C. R. Hebd. Seances Acad. Sci. Paris 17, 989-1008.

204

DAVID G. HEATHCOTE

Firn, R. D., and Digby, J. (1979). Plant Cell Environ. 2, 149-154. Gordon, S. A. (1973). COSPAR Life Sci. Space Res. 11, 155-162. Gradmann, H. (1926). Jahrb Wiss. Bor. 65, 224278. Gray, S. W., and Edwards, B. F. (1971). In “The Experiments of Biosatellite 11” (J. F. Sanders, ed.), pp. 123-165. NASA SP-204. Hader, D.-P. (1979). Encycl. Plant Physiol. New Ser. 7 , 268-309. Heathcote, D. G. (1965). Narure (London) 208, 909-910. Heathcote, D. G. (1968). Ph.D. Thesis, University of Wales. Heathcote, D. G. (1981). Plant Cell Environ. 4, 131-140. Heathcote, D. G., and Aston, T. J. (1970). J . Exp. Bor. 21, 997-1002. Huisinga, B. (1968). Acra Bor. Neerl. 17, 117-125. Johnson, S. P., and Tibbitts, T. W. (1968). Bioscience 18, 655-661. Johnson, S. P., and Tibbitts, T. W. (1971). In “The Experiments of Biosatellite 11” (J. F. Sanders, ed.), p. 247. NASA SP-204. Johnsson. A., and Heathcote, D. G. (1973). Z. Pflanzenphysiol. 70, 371-405. Johnsson, A., and Israelsson, D. (1968). Physiol. Planr. 21, 282-291. Johnsson, A., and Pickard, B. G. (1979). Physiol. Plant. 45, 315-319. Kasatkina, T. B., Zharikova, G. G., Rubin, A. B., Palmbakh, L. R., Vaulina, E. N., and Mashinsky, A. L. (1980). COSPAR Life Sci. Space Res. 18, 205-21 I . Klimuk, P. I., and Baburina. Ye. B. (1976). Zdorov’ye 4, 6-7. (NASA Transl. TT F-17438.) Lyon, C. J. (1971). In “The Experiments of Biosatellite 11” (J. F. Sanders, ed.), pp. 167-188. NASA SP-204. Maine, R. B., Wagner, P. A., Olcott, T. M . , and Luce, R. S. (1979). Proc. Intersoc. Conf. Environ. System, 9rh, San Francisco ASME Publ. 79-ENAS-19, 1-1 1. Malmkjac, Y., Bewersdorff, A . , Da Riva, I., and Napolitano, L. G. (1981). “Challenges and Prospectives of Microgravity Research in Space,” pp. 1-77. ESA BR-05. Merkys, A. J., Mashinsky, A. L., Laurinavchius, R. S., Nechitailo, G. S., Yaroshius, A. V., and Izipak, E. A. (1975). COSPAR Life Sci. Space Res. 13, 53-57. Oberg, J. (1981). New Sci. 92, 17-19. Schachter, P., and Tyler, J. ( I 979). “Life Sciences Laboratory Equipment (LSLE) Descriptions, October 19, 1979.” NASA, JSC 16254. Sweeney, B. M. (1979). Encycl. Plant Physiol. New Ser. 7. 71-93. Volkmann, D., and Sievers, A. (1979). Encycl. Plant Physiol. New Ser. 7, 573-600. Wilkins, M. B. (1977). Proc. R . SOC. London Ser. B 199, 513-524. Zharikova, G. G., Rubin, A. B., and Nemchinov, A. V. (1977). COSPAR Life Sci. Space Res. 15. 29 1-294. Zimmermann, W. (1927). Jahrb. Wiss. Bor. 66, 631-677.

INTERNATIONAL REVIEW OF CYTOLOGY, VOL 71

Chloroplasts and Chloroplast DNA of Acetabularia mediterranea: Facts and Hypotheses1 ANGELALUTTKE* A N D

SILVANO

BoNoTTot

*Institute f o r Developmental Physiology, Universiry of Cologne, Cologne, Federal Republic of Germany, and ?Department of Radiobiology, Nuclear Research Centre. Mol, Belgium 1. Introduction ....................................... 11. Chloroplast Morphology and Ultrastructure. . A. Chloroplasts in Situ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

B. Morphology and Properties of Isolated Chloroplasts. . . . . . . . .

....................... ....................... 111. Chloroplast DNA. . ............... .... A. Buoyant Density an ontent.. ...................... ....... B. DNA Content and Distribution . . . . . . . . C. DNA Replication.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. DNA Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. Physical Properties ........ IV. Chloroplast Gene Produ ..................... V. Summary and Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....................................... References

205 207 207 21 I 213 214 218 218 220 225 225 234 235 237 238

I. Introduction Reviews are generally written at times when established results in a certain field start stabilizing formulated hypotheses or evoke new concepts. Our decision for writing this article arose from the opinion that gaps in our knowledge will not be filled without knowing about them, hypotheses will not be proved or disproved without formulating them. Our knowledge about the plastome of the chloroplasts of the giant marine alga Acetabularia mediterranea2 (Fig. 1) is still in its infancy, but we believe that future and more rapid progress can be made by compiling and presenting the available data. 'Dedicated to the memory of our colleague and friend Professor Dr. Heinz Clauss. Wnder the International Code of Botanical Nomenclature, the correct name of Acetabularia mediterranea Lamouroux is now Acetabularia acetubulum (L.) Silva (Silva. 1952). As the name mediterranea has been used in all the published work referred to in this article, it has been retained to avoid confusion. 205 Copyright Q 1982 by Academic Press. Inc. All right3 of reproduction in any form reserved. ISBN 0-12-364477-1

206

ANGELA LUTTKE AND SILVANO BONOTTO

FIG. 1. Developmental stages of A . medirerranea. (A) Vegetative cell about 6 weeks after the outgrowth of the zygote; (B) cell with cap primordium; (C) cell with young cap; (D) cell with mature cap shortly before meiosis; (E) cell after meosis and subsequent mitoses with cysts in the cap rays. For detailed information on the biological cycle of Acetabularia see the early studies by Haminerling (1953) and Schulze (1939) and the more recent investigations by Crawley (1970) and Koop (1979).

Our intention is ( 1 ) to recall morphological and ultrastructural features of the organelles for the biochemically oriented investigator; (2) to lead the electron microscopist’s attention to biochemical aspects, whose solution need her/his techniques; (3) to encourage biologists interested in the molecular biology of chloroplasts to put their hands onto the plastids of Acetabufaria; (4)but also to point to difficulties genuine to Acetabufaria and its chloroplasts. This list explains why some literature on the cell’s cytology and numerous papers on chloroplast morphology and ultrastructure are quoted and several micrographs on this aspect are included. We think the heterogeneity of the chloroplasts is widely overlooked in many biochemical investigations and therefore we refer to it not only in the section on the plastids’ morphology and ultrastructure (Section II), but also in the section on the chloroplast DNA (Section 111). At the end some readers might feel uneasy since too many questions remain unanswered. For us it seems important to clearly state what we do not as well as what we do know. Provided with such background the investigator will be better able to evaluate where research on the plastome of Acetabularia stands and which aspects need to be reinvestigated or have not yet been tackled at all.

CHLOROPLAST DNA OF A . MEDITERRANEA

207

11. Chloroplast Morphology and Ultrastructure A. CHLOROPLASTS in Situ

In the cylindrical Acetabularia cells the numerous chloroplasts (several million), together with other cell organelles are embedded in a thin layer of cytoplasm lining the cell wall and enclosing a large central vacuole (Fig. 2). The vacuole having a pH between 2.5 and 3.5 (Crawley, 1963) makes up about 90% of the total cell volume, and the rather thick cell wall about 4%; thus only 6% is comprised by the cytoplasm. The cytoplasm forms a dynamic network of strands packed with chloroplasts, which are oriented in parallel to the long axis of the cell (Fig. 3A and B). Within the strands filaments also running lengthwise to the cell axis are visible (Puiseux-Dao, 1979; Puiseux-Dao et ul., 1979; Koop and Kiermeyer, 1979). Upon glutaraldehyde fixation the cytoplasm slightly retracts from the cell wall and can then be easily isolated and viewed under higher magnification (Fig. 4). A knobby appearance of the filaments lying across the long axis of the plastids and in parallel is discernible. Their width measures 3 6 7 1 nm, while the knobs being about one-third to one-half wider than the corresponding filaments are 7 1-107 nm in diameter. Filaments connecting the chloroplasts can also be seen in the hairs (Fig. 5A and B; Piques et a / . . 1979) and later in cap-bearing cells in the cap cytoplasm. Fragments of filaments

FIG. 2. Part of a cross-section through the stalk. The arrows indicate the cell wall. the arrowheads the cytoplasmic layer lining the wall, and V the central vacuole.

FIG.3. Optical section through the stalk (A.B). Note the numerous strands oriented in a longitudinal direction to the cell's long axis (0).which are particularly visible with focus settin[: on the upper level (B). Nomarski interference optics.

FIG.4. Isolated cytoplasm viewed from the tonoplast facing side. Numerous strands with knobby appearance (arrowheads) and chloroplasts (arrows) are oriented in a longitudinal direction to the cell's long axis (0). Nomarski interference optics.

CHLOROPLAST DNA OF A. MEDlTERRAh'EA

209

FIG. 5 . Longitudinal oriented strands in the hairs (A,B). The strands (arrowheads) running along the long axis of the chloroplasts on the plasmalemma facing side are very visible at higher magnification (B). Nomarski interference optics.

210

ANGELA LUTTKE AND SILVANO BONOTTO

remain attached to the chloroplasts when the cell is cut and the cytoplasm allowed to flow out freely (Fig. 6). Indirect evidence that the filaments consist of a complicated microtubular/ microfilament system comes from studies on the interference of cytochalasin B and colchicin with cytoplasmic streaming (Koop and Kiermeyer, 1980a,b). Most recently direct proof was provided for the presence of actin in Acetabularia cells by immunofluorescence microscopy (Dazy et al., 1982). A detailed ultr.astructural investigation of these structures is still lacking. Investigations on the spatial distribution of the plastids in the stalk by light and electron microscopy (Shephard, 1965a; Puiseux-Dao and Dazy, 1970; Boloukhkre-Presburg, 1972; Hoursiangou-Neubrun and Puiseux-Dao, 1974) or by buoyant density determination (Liittke et al., 1976; Liittke and Rahmsdorf, 1979) have shown an increase in size from the apex toward the base of the cell concomitant with an increase of the starch content. Environmental conditions like enucleation influence the apicobasal gradient in that a different degree of starch accumulation or consumption causes a changing proportion of small and large plastids (Vettermann, 1973; Hoursiangou-Neubrun and Puiseux-Dao, 1974; Schmid and Clauss, 1977; Puiseux-Dao et al., 1978; Hoursiangou-Neubrun er al., 1979; Luttke and Rahmsdorf, 1979). Recent investigations on cytoplasmic streaming by microcinematography and

Fic. 6 . Cytoplasm released from the cell. Strands (arrowheads) keep attached to the chloroplasts. Nomarski interference optics.

CHLOROPLAST DNA OF A . MEDITERRANEA

21 1

on chloroplast differentiation under various culture conditions suggest a close relation between both processes (Puiseux-Dao, 1979; Puiseux-Dao et ul., 1978, 1979; Hoursiangou-Neubrun et ul., I98 I ) .

B. MORPHOLOGY A N D PROPERTIES OF ISOLATEDCHLOROPLASTS Tremendous size differences of the chloroplasts in A . rnediterruneu have been pointed out as long as 50 years ago (Mangenot and Nardi, 1931). The careful isolation of the organelles in formaldehyde containing growth medium (Shephard, 196%) allows light microscopic observations of the plastids having practically an unaltered morphology to the in situ ones. The plastids’ gross morphology is rather simple. They are ovoid to sausage-shaped in the stalk (Fig. 6 ) and more rod-shaped in the hairs of the whorls (Fig. 5). The characteristic storage material starch is clearly visible in the form of prominent granules. They are far more abundant in the plastids of the basal cell part than the apical one (see Section 11,A). The variability in size found in the chloroplast population of a single cell must be mainly attributed to differences of the plastids’ long axis. It ranges from about I to 12 p,m, while the length of the short axis oscillates around 3 p,m (Fig. 7). From the size histogram it can be reasonably concluded that the size variation is due to growth of the organelles rather than fusion. If large plastids would arise from smaller ones by random fusion, a similar variation of the short and long axis must be expected. Fusion would also imply that the process exclusively takes place at the shorter axis of the plastids. Though the orientation of the plastids in situ gives a higher probability for contact zones

/I

L O N G AXIS

n

SHORT A X I S

140

,120 100 u)

u 80

f

60 40

8 20 SIZE C L A S S

FIG. 7. Size histograms for the long and thc short axes of chloroplasts isolated from A . rnediterruneu at different developmental stages. (A) Chloroplasts from 6-10 mm cells: ( B ) chloroplasts from 25 mm cells; (C) chloroplasts from caps of 5 mm cap-bearing cells; (D)chloroplasts from stalks of 5 mm cap-bearing cells.

212

ANGELA LUTT’KE AND SILVANO BONOTT’O

between the short axes, chloroplast streaming, on the other hand, appears to allow contact along the long axes also. Hence, the additional, yet unproven assumption of different chloroplast envelope properties along the axes needs to be made. In addition to the uneven spatial distribution of the chloroplasts in the stalk (see Section II,A) different proportions of small and large chloroplasts are present at the various stages of development (Fig. 7; Shephard, 1965a). This appears to be linked to the mode and speed of chloroplast division throughout the cell cycle (see Section 11,D). The cytology of the Acetabularia cell (see Section II,A) and the size differences of the chloroplasts create fundamental experimental difficulties the unexperienced investigator will undoubtedly be faced with, when starting to isolate the organelles. Hence, a few practical remarks might be useful, for understanding our critical notions in later sections. In common isolation procedures the chloroplasts are pelleted at about 1000 g and are then generally subjected to further rounds of centrifugation for purification. Acetabularia chloroplasts tend to form clumps, which are difficult to disjoin. Even in a “homogeneous” suspension, as judged by eye, microdroplets become visible upon microscopic observation. They consist of a thin cytoplasm layer enclosing two or more plastids and quite often mitochondria (Bidwell, 1972). It appears that the close spatial association of chloroplasts and mitochondria in situ (Fig. 8) is naturally kept in the “cytoplasts” as pointed out by several authors (for example D’Emilio et a / . , 1979, Fig. 3 ; Astaurova et al., 1979, Fig. 2). Shephard (1970) and Shephard and Levin (1972) proposed a flow diagramm for the isolation of Acetabularia chloroplasts, which includes three centrifugation steps of the plastid suspension through a buffered Ficoll phase at 350 g and 700 g, respectively. In addition the authors suggest passing the chloroplast suspension through a membrane filter of appropriate pore size. Cytoplasmic droplets were effectively sheared by using a filter with a pore diameter of 5 pm (Shephard and Levin, 1972, Fig. 2; Shephard and Bidwell, 1973, Fig. 6a and b). With this procedure, however, most of the chloroplasts larger than 5 pm might be lost. The cautious notion that pelleting of small plastids through the Ficoll phase is not warranted at 750 g appears necessary. We found that even after an increase of the speed of 1000 g a considerable percentage of plastids is retained in the supernatant and can only be spun down at 2500 g (Table I). Since the chloroplast size differences also appear to reflect differences in biosynthetic capacities, it can be speculated that many experiments are done with a nonrepresentative plastid population and biosyntheses might be over- or underestimated. A unique property of isolated Acetabularia chloroplasts is their stability over several hours. In contrast to higher plant plastids, they are not disrupted under hypotonic conditions nor fragmented by 1% Triton X-100, although the chlorophylls are completely solubilized in the presznce of this detergent concentra-

213

CHLOROPLAST DNA OF A . MEDITERRANEA

FIG.8. Close spatial association of chloroplasts and mitochondria in situ. CY. Cytoplasm; CP, chloroplasts; M, mitochondria; V , vacuole.

tion (Goffeau, 1969). The longevity of the isolated chloroplasts had been successfully exploited in longtime experiments (see Section IV).

C. CHLOROPLAST ULTRASTRUCTURE Despite the variability of the plastids’ size a relatively simple basic ultrastructure was revealed by electron microscopy. Peripheral thylakoids run in parallel to the double-membraned envelope, while others descending from the opposite TABLE I RECOVERYOF CHLOROPLASTS AT DIFFERENT CENTRIFUGAL FORCESISOLATED APICALAND BASALFRAGMENTS

FROM

WHOLECELLSA N D

Fragment type Whole cells Number of cells forces Percentage chlorophyll

230 1000

79

2500 21

Apical 280 lo00 2500 87 13

Basal 2x0 lo00

96

Apical 600

2500 4

1000

68

2500 32

Basal 600 1000 2500 89 II

214

ANGELA LUTTKE AND SILVANO BONOTTO

poles of the plastid cross the stroma in a diagonal direction. In larger chloroplasts the diagonal thylakoids always separate the two or more starch grains from each other (Figs. 9 and 10; Puiseux-Dao, 1970, plate 8; Puiseux-Dao and Dazy, 1970, Fig. lb; Boloukhkre, 1972, plate Ha). From the geometrical pattern of the thylakoid arrangement it was concluded that larger plastids consist of two or more identical entities, which eventually will separate along the plane of the merging diagonal lamellae (Puiseux-Dao. 1970; Puiseux-Dao and Dazy, 1970). Though Acetabularia chloroplasts resemble higher plant plastids at low magnification, the piles of stacked thylakoids had been termed “pseudograna” (Fig. I I ) , since they seem in sections to be formed by appressed and sometimes invaginated “stroma” thylakoids (Van Gansen and Boloukhkre-Presburg, 1965). In negatively stained thylakoid preparations one often sees discs forming short chains, which are attached to longer thylakoids (Hoursiangou-Neuhrun et al., 1977, P1. 111 and IV). The discs are more abundant in apical chloroplasts. The plastids’ ultrastructure is susceptible to environmental conditions, experimental manipulations, and different types of inhibitors. Under unfavorable growth conditions like pure seawater or red light, or after enucleation the chloroplasts contain huge starch grains all along the stalk (Puiseux-Dao et al., 1978; Vettermann, 1973; Vanden Driessche and Hars, 1972, 1973; Hoursiangou-Neubrun et al., 1979). In cells grown in the presence of actinomycin D or puromycin the number of thylakoids is reduced and they appear disorganized (Boloukh2rePresburg, 1965, 1966). In cells treated with the insecticide lindane the starch grains in subapical chloroplasts assume an unusual appearance (Borghi t?t al., 1973). On purpose we exclude in this section the findings on the presence and localization of the plastids’ DNA. A discussion on this matter will be given in Section III,D together with the observations made by light microscopy. D. CHLOROPLAST DIVISION Chloroplast division takes place by rupture of the chloroplast envelope at the site where the diagonal thylakoids are merging (Puiseux-Dao, 1970, Fig. 38; Fig. 9). Several lines of evidence suggest a higher division rate in the apical cell part than in the basal one: (1) on the average plastids in the apex are smaller than in the base; (2) large plastids in the base become smaller when the apical cell part is cut off; (3) plastids in enucleate apical fragments divide in contrast to plastids in enucleate basal fragments (Shephard, 1965a; Puiseux-Dao and Dazy, 1970; Clauss ef al., 1970). This interpretation is also supported by calculations for the doubling time of the cells’ total plastid population in comparison to the population of the apical and middle cell part. For the former the generation time is a little longer than 1 week (Shephard, 1965a). while for the latter it is 5 days (Puiseux-Dao, 1968). A schematic presentation for the positive correlation be-

CHLOROPLAST DNA OF A. MEDITERRANEA

215

FIG. 9. Chloroplast with diagonal lamellae (DL) separating two basic plastidal units. (Courtesy of Dr Simone Puiseux-Dao, University of Paris, VII.)

tween chloroplast division and cell growth (Clauss et a / ., 1970) is given in Fig. 12. While the rate of chloroplast division decreases in the apicobasal direction during most of the vegetative cell cycle, a different picture emerges for cells with mature caps entering the generative phase. In addition to active chloroplast division in the cap cytoplasm the large chloroplasts in the basal cell part bud off

216

ANGELA LUTTKE AND SILVANO BONOTTO

FIG. 10. Chloroplast with diagonal larnellae (DL) separating the starch (ST)-containing part from the stroma-rich part of the organelle.

FIG. I I .

Piles of appressed thylakoids forming pseudograna (boxed in).

WHOLE

T

FRAGMENTS

CMOAOPLAST

NllMBER

CELL

/ WREASE

Y

m

TI

cAL

-------

BASAL

BASAL

------INCREASE

-------

FIG. 12. Schematic representation for the positive correlation between growth and chloroplast division in A. mediferruneu. The chloroplast number increases in growing enucleate apical fragments; the chloroplast number stays constant in nongrowing enucleate basal fragments, but increases in growing nucleate basal fragments.

218

ANGELA LUTTKE AND SILVANO BONOTTO

small portions, which migrate into the cap cytoplasm. The rest of the plastids degenerates (Boloukhkre, 1970). Since chloroplast division also takes place in enucleate cells, apical cell fragments, and cytoplasts, the physical presence of the nucleus and newly synthesized transcription products is not necessary for the division process (Shephard, 1965b; Clauss et al., 1970; Vanden Driessche er a f . , 1973a,b). Though this finding seemingly favors the idea of chloroplast autonomy in respect to division, the nuclear DNA coded information stored in the cytoplasm (for review see Hammerling, 1953; Brachet, 1968) points to a certain degree of dependency. Chloroplast division is affected both by light quantity and quality. The division ceases in the dark (Shephard, 1965a) and in red light (Schmid and Clauss, 1974, 1975). After transfer of cells grown in red light for 3 weeks into blue light, a burst of synchronous divisions is induced after a short lag phase (Schmid and Clauss, 1975). The division rate undergoes a diurnal change in algae maintained in a light/ dark cycle of 12: 12 hours with a peak frequency during the second half of the light cycle (Puiseux-Dao and Gilbert, 1967; Vanden Driessche and Hars, 1973). The control of chloroplast division in relation to the physiological activities of the organelles was reviewed by Vanden Driessche (1973). 111. Chloroplast DNA

A. BUOYANT DENSITYAND GC CONTENT Stormy discussions centered around the buoyant density of chloroplast DNA from A . mediterranea in the late 1960s and the beginning of the 1970s. The early confusion cannot better be described than by Kirk’s sentence “will the real chloroplast DNA please stand up” (Kirk, 1971). A compilation of all published densities ascribed to cpDNA can be found in the review on the biology of Acetabularia by Bonotto et al. (1976). It is now generally accepted, that cpDNA of A . mediterranea has a buoyant density of 1.702-1.704 g/cm3 in neutral CsCl gradients (Fig. 13) corresponding to a GC content of 4 3 4 5 % (Green et a f . , 1967, 1970; Heilporn and Limbosch, 1971; Green, 1972; Bonotto et a f . , 1975). An average GC content of 43 and 42% is also calculated from the melting point (T,) in lXSSC and 0.1 XSSC, respectively (Green et a f . , 1977). Despite the homogeneous peak in neutral CsCl gradients considerable intramolecular heterogeneity is signalized by the plot of the first derivative of the melting curve (Green et al., 1977). Intramolecular heterogeneity is quite common for cpDNA of algae in contrast to higher plant cpDNA. For example, cpDNA of Euglena gracilis Z giving a homogeneous peak in neutral CsCl gradients by analytical

CHLOROPLAST DNA OF A . MEDITERRMEA

219

I

FIG. 13. Chloroplast DNA profile from A . medirerrunea after banding in a neutral CsCl gradient. The main peak at p = 1.703 corresponds to chloroplast DNA, the two additional peaks at p = 1.710 and p = 1.718 are not identified. Note that the buoyant density of the two minor peaks is probably underestimated in this graph. The reference DNA at p = I .73 I is from Micrococcus Ivsodeikricus.

centrifugation has five distinct thermal transition temperatures (Slavik and Hershberger, 1976). It seems reasonable to see an intimate relation between the greater diversity in plastome size and intramolecular heterogeneity found among algae in comparison to higher plants. A particularly intriguing question for Acetabularia is, whether there is a difference in the intramolecular heterogeneity between cpDNA of apical and basal plastids. A first attempt was made by comparing the UV profiles of cpDNA isolated from apical, middle, and basal plastids (Bonotto, unpublished). Though the spectra do not indicate any differences, other types of analysis like melting or analytical centrifugation of sheared DNA could change the picture. Despite the good agreement in published buoyant densities during the last years some cautious remarks need be made: ( I ) bacterial contamination was and is a major problem in the maintenance of large Acetabularia culture; (2) the use of axenic cultures is highly recommended; the less reliable alternative is the incubation with antibiotics for several days prior to chloroplast isolation; (3) the continuous or repeated use of antibiotics over a prolonged time might generate resistant bacterial strains; (4) contamination with nuclear DNA, a major source in other plant species with close buoyant densities for nuclear and chloroplast DNA, can easily be avoided by cutting off the rhizoid; and (5) this procedure also eliminates the main source of contamination.

220

ANGELA LUTTKE AND SILVANO BONOTTO

B. DNA CONTENT AND DISTRIBUTION Ample evidence for appreciable amounts of DNA in Acetabularia chloroplasts comes from micrographs of plastids stained with fluorescent dyes (Figs. 14 and 15) or of thin-sectioned plastids (see below). Unfortunately, only a few determinations were made by direct biochemical methods. In Table 11 the existing data for the DNA content per chloroplast are summarized and for comparison the values for the well-characterized cpDNA of Euglena gracilis (Nigon and Heizmann, 1978), Chlamydomonas rheinhardii (Bastia er a l . , 197 I), and Spinacia oleracea (Scott and Possingham, 1980). For convenience, all values are recalculated into daltons, where it was necessary. Except for the early low estimate by Gibor and Izawa (1963) the average DNA content of Acetabularia chloroplasts being in the range of 10’ daltons is in good agreement with that of the other species listed. For the decrease in the rate of chloroplast division from the apex to the base of the cell (see Section L D ) , the heterogeneous distribution of cpDNA (see below) and cpDNA replication (see Section 111,C) determinations for the different cell regions would be most desirable. This information bears even more value under the aspect that not all chloroplasts of A . medirerranea contain DNA (Table 111). On the other hand, the amounts vary up to 30-fold among the plastids with DNA (Coleman, 1979a,b). A lack of cpDNA in A . medirerranea and A . cliftonii in 65-80% of the total plastid population was first demonstrated by Woodcock and Bogorad ( I 970), using fluorescence staining, [3H]actinomycin D binding, and electron microscopy. Recently, these findings were confirmed and extended to the species A . calyculus, A . crenulata, and the related Dasycladacea Batophora oerstedii (Coleman, 1979a,b). Two questions arose by these findings: (1) is there an even spatial distribution of chloroplasts with DNA along the stalk in vegetative cells, and (2) is there a TABLE 11 DNA CONTENTPER CHLOROPLAST Species A . rnediterranea A . mediterranea A . medirerraneu A . clifonii A . ucerubulum (=rnedirerranea) Euglena Rracilis Chlamgdomonas rheinhardii (veget.

cpDNA content (daltons) 6.0 x 107

0.3-3.0 X 1.4 X 4.7 x u p to 4 x 6.0 X

10’ 10’

10’ 10’ 10’

10.0 x 10’

Reference Gibor and lzawa (196.3) Green er a / . ( 1970) Green et al. ( I 977) Green et 01. ( 1977) Mazza et a / . ( 1979) Nigon and Heizmann 1978) Bastia et 01. ( 197 I )

cells) Spinach oleracea

3-19.2

X

10’

Scott and Possingham ( 1980)

CHLOROPLAST DNA OF A. MEDITERRANEA

22 1

FIG. 14. Chloroplasts isolated from A. medirerruneu and stained with the DNA-specific fluorochrome DAPI. (A) Preparation viewed with a filter combination for the DNAIDAPI fluorescence; ( B ) the same preparation showing the autofluorescence of the chlorophyll. Glutaraldehyde fixation.

222

ANGELA LUTT'KE AND SILVANO BONOTT'O

FIG.15. Single enlarged chloroplasts from A. medirerranea stained with the DNA-specific fluorochrome DAPI. (A, B) Plastids from the apical cell part with long continuous DNA stretches; the white line indicates the outer shape of the plastids: (C) plastid from the basal cell part with an acentrically located DNA area; (D) the same underexposed plastid shows that the interspersed unstained areas within the nucleoid become visible.

223

CHLOROPLAST DNA OF A. MEDITERRANEA TABLE 111 DNA/DAPI FLUORESCENCE I N CHLOROPLASTS OF A. mediterruneu AT VARIOUS DEVELOPMENTAL STAGES"

Cell stage

Nucleus stage

Germlings 6-10 mm cells 25 mm cells Cells with 2 mm cap Cells with 5 mm cap Developing cysts Gametes

2n 2n 2n 2n 2n In In

Number counted

Number with DNA

Percentage with DNA

-

189 82 132 59 I -

IO O h

294 I74 265 805

-

64 47 50 73 98" IO O b

Modified from Liittke and Bonotto (1981b). All chloroplasts in germlings and gametes contain DNA. Data from Coleman (1979a).

regulatory mechanism to ensure the transmission of the chlorplasts' DNA into the daughter generation (Crawley, 1966, 1970).

I . The number of chloroplasts with DNA decreases from the apical to the basal part of the cell (Table IV; Liittke et al., 1980; Liittke, 1981). This finding is in good agreement with the electron microscopic observation that in basal chloroplasts nucleoids are less frequently visible (Puiseux-Dao and Dazy, 1970). 2. The data on the number of chloroplasts with DNA in cells at various stages of development show a tremendous change (Table 111). On the basis of the mode of chloroplast division during the various stages of the cells' development (Puiseux-Dao and Dazy, 1970; Boloukhere, 1970) and the acentrically located nucleoid in either one half in large plastids (Liittke, 1981), we propose a scheme for the distribution of cpDNA (Fig. 16). It also explains the unique situation of

TABLE IV DAPI FLUORESCENCE I N CHLOROPLASTS OF CELLFRAGMENTS OF A. mediterraneau

Fragment typeb

Number counted

Number with DNA

Percentage with DNA

Apical Subapical Middle Basal

I90 210 227 249

121 97 80 89

64 46 35 36

Data from Liittke (1981) and Liittke and Bonotto (1981a). Five millimeter fragments from 25 mm cells without cap.

224

ANGELA LUTTKE AND SILVANO BONOTTO

rtomal thylakoidr 7 r

NA

diagonal thylakoids

FIG. 16. Comparative representation of the chloroplast cycle (A) and the cell cycle (B).Sequence 1, 2 of the chloroplast cycle with equal distribution of the newly formed DNA replicas onto the daughter plastids takes place at the cell stages A, B and predominantly in the growing tip of cells in stage C, D, E; in cells of stages C, D. E sequence 3 , 4 leading to DNA-deficient chloroplasts takes place in the proportionately increasing middle and basal cell part; in cells at stage F, G sequence I , 2 takes place in the cap, while sequence 5 , 6 with budding off of a DNA containing small portion leaving a degenerating part takes place in the basal cell part. Not drawn to scale.

the simultaneous presence of plastids with and without DNA within one cell under normal culture conditions (Luttke and Bonotto, 1980). Combining the results on the DNA content per plastid and the number of plastids with discernible DNA it is obvious that the actual amount in a single plastid might exceed the average value several fold or be much lower.

CHLOROPLAST DNA OF A. MEDITERRANEA

225

C. DNA REPLICATION cpDNA replication in both nucleate and enucleate cells was first strongly suggested by the incorporation of tritiated thymidine and subsequent autoradiographic analysis (De Vitry, 1965; Shephard, 1965b). Thymidine label found in enucleate cells 1 week postenucleation (Shephard, 1965b) evoked the question of whether the incorporation indicated a real net synthesis in the absence of the nucleus. The increase of the DNA content up to 15 days after the removal of the nucleus, indeed, demonstrated that net DNA synthesis took place (Heilporn-Pohl and Brachet, 1966; Heilporn and Limbosch, 1970). Since there is no doubling of the DNA content either in nucleate or in enucleate cells during the time needed for a doubling of the cells’ total chloroplast population (see Section II,D), chloroplast division and DNA replication are not tightly coupled. This conclusion does not exclude, however, that there is no relation between both processes at all (see Section 11,B). In Section 11,D it was indicated that the rate of chloroplast division decreases from the apex toward the base in vegetative cells prior to cap formation. Results on the incorporation of [3H]methylthymidine and analysis by EM autoradiography suggest a similar apicobasal gradient for DNA replication (HoursiangouNeubrun, 1979; Luttke et al., 1980; Hoursiangou-Neubrun et al., 1981, and in preparation). The number of plastids with radiolabel in the apex is about twice as high as the number in the basal cell part. This finding alone might only support our recent results on the decrease of the number of chloroplasts with DNA in the apicobasal direction (see Section 111,B).Stronger support for a more intensive cpDNA synthesis in apical chloroplasts comes from the fact that the number of silver grains over apical organelles is higher than over basal ones (HoursiangouNeubrun et al., 1981, and in preparation). To date no investigations have been devoted to the mode of cpDNA replication. For example, the exceptional large size of the plastome (see Section 1II.D) suggests the question of whether ( I ) replication starts at one or several initiation points, ( 2 ) it is continuous throughout the plastome or (3) there is some kind of amplification of specific sequences, or (4)membrane attachment sites play a role for replication in a fashion similar to bacteria and animal cells (see for example the model proposed by Pardoll et af., 1980). Recent EM investigations by Mazza et al. (1980) have suggested that the minicircles, which appear to undergo DNA replication following the rolling circle model, might result from a gene amplification mechanism. D. DNA MORPHOLOGY Three configurations of the DNA can be distinguished in plastids stained with the fluorochrome DAPI. Small ovoid plastids appear to be completely filled with

226

ANGELA LUTTKE AND SILVANO BONOTTO

DNA and no separate nucleoids are visible (Fig. 14). In long rod-shaped chloroplasts the DNA often stretches throughout the entire organelle, and the less intensely fluorescent parts might indicate a later separation (Fig. 15A arid B). Both small ovoid plastids and the long ones are more abundant in the uppermost tip of the cell and the whorls than in the middle and basal region (see Section 11,B). In plastids of the latter cell parts the DNA is typically localized in either one half of the organelle, adjacent and often partially looping around a starch grain. The close-up demonstrates the mass of DNA being interspersed by unstained areas (Fig. 15C and D). A pleasant situation is met, when light and electron microscopic observations not only agree but complement each uther. The central location of the nucleoid in small ovoid organelles with a yet poorly developed thylakoid system was first demonstrated by Woodcock and Bogorad (1970, Fig. 17) in serial sections. Indicated in these micrographs but more convincingly shown by Puiseux-Dao el al. (1967, Plate I), Werz (1965, Fig. 3), and Boloukhkre (1972, Plate IV, Fig. a) are clusters of ribosomes at the border of the DNA areas and interspersed ones (Figs. 17 and 18). For the specificity of the fluorochrome DAPI for DNA we believe the ribosome clusters correspond to the unstained regions visualiz.ed by light microscopy (Fig. 15). It is interesting to note, that a close association between ribosomes and DNA fibers was also observed in other algal chloroplasts, for example in Ochromonas danica having a ringlike nucleoid (Gibbs et a f . , 1974, Fig. 3). Also in agreement are the light and electron microscopic observations that in large plastids from whole cells the DNA is distinctly localized in only one part of the organelle adjacent to a starch grain (for example Boloukhere, 1972, Plate 11, Fig. c). This finding matches our observation that the majority of plastids display only one DNA area in flourescently stained preparations. All published micrographs and our investigations speak against a scattered distribution of the DNA throughout Acerabuluria chloroplasts and therefore contradict Gibbs’ generalization that in green algae, like in higher plant plastids, the nucleoids exist as several separate units (Gibbs, 1970). It appears, however, that at least in some cases the DNA area is composed of several nucleoids (Fig. 17; Werz, 1965, Fig. 3; Boloukhkre, 1972, Plate VII, Fig. 3). An intriguing question is whether and, if it is so, how transcription is linked to the spatial arrangement of DNA and clustered ribosomes. Does it facilitate transcription or does the packaging prevent access? Puiseux-Dao ( 1970, Plate 8, Fig. 2) once compared the dense granular masses within the DNA area with nucleoli (Figs. 9 and 18). A similar nuleolus-like structure had been described for the chloroplast nucleoid of the dinoflagellate Scrippsielfa sweeneyae (Bibby and Dodge, 1974, Figs. 4-6). For the many prokaryotic-like features of chloroplasts and the similar gross appearance of the plastid nucleoid and the bacterial nucleoid (compare Ryter and Chang, 1975, Plate 11, a with Werz, 1965, Fig. 3)

CHLOROPLAST DNA OF A. MEDITERRANEA

227

Fic. 17. (A) DNA fibrils (arrows) in the stroma adjacent to a starch grain (ST);(B)disappearance of DNA fibrils after DNase treatment. (Courtesy of Dr Simone Puiseux-Dao, University of Paris VII.)

228

ANGELA LUTTKE AND SILVANO BONOTTO

FIG. 18. DNA fibrils (arrows) with centrally located nucleolus-like structure (boxed in)

CHLOROPLAST DNA OF A . MEDITERRANEA

229

we prefer the comparison to be made between those two. We think that some of the many successful approaches in this field to elucidate the relation between the three-dimensional structures, gene expression, and replication should be adopted to advance our knowledge on the Acetabularia chloroplast nucleoid (for review see Pettijohn and Worcel, 1976; Pettijohn and Carlson, 1979). Except for the intraorganellar localization very little can be learned about the DNA ultrastructure from thin sections. Tremendous progress in information on the ultrastructure and molecular organization of DNA was brought about after Kleinschmidt (1968) introduced a technique to visualize the molecules in the electron microscope. When chloroplasts of A . mediterranea are subjected to osmotic shock and spread out into a cytochrome monolayer large masses of DNA are released (Fig. 19) (Green and Burton, 1970; Green et al., 1970; Mazza et a f . , 1977). Careful processing in such a way results in many supercoiled portions suggesting a relatively low degree of degradation. The common presence of supercoiled regions not only in spread preparations of cpDNA but also of mitochondrial and bacterial DNA makes the assumption likely of it being an intermediate configuration between completely unfolded DNA caused by spreading and the packaged state in situ. Ample evidence that this assumption holds for the bacterial nucleoid is provided by Kavenoff and Bowen ( 1976) for E . coli. When membrane-free folded nucleoids are isolated and spread they unfold into many supercoiled loops held together by an electron-dense core. Though the original osmotic shock technique of Kleinschmidt and since then the modified versions are most useful for obtaining insight into the structural organization of the nucleoid and potential attachment sites, in most cases the entangled molecules on the grid do not allow length measurements of single molecules. This, however, is the information needed to get a rough idea about the genetic informational content. Hence, the problem is generally approached with purified DNA. A synopsis of the techniques used for the isolation and purification of Acetabularia cpDNA and the configurations after spreading is given in Table V . The important findings are that (1) no circular molecules (falling into the size of 40-50 pm) had been found like those found for other algae or higher plants (for review see Kirk and Tilney-Bassett, 1978); (2) except for one early investigation the lengths of the linear molecules by far exceed the commonly found average value of 40 pm; (3) the variable length of the linear molecules suggests random breakage of the DNA during the isolation and processing for electron microscopy; (4) for the lack of a mean length in fragment size a tandem repetition of a basic unit appears unlikely. It seems advised that to answer the question of whether the molecules are circular or not gentler methods need be applied. For example, the technique of Cairns (1962, 1963) might be suitable. The visualization of the labeled DNA by autoradiography on microscope slides was the first proof for the circularity of the I .2 mm E. coli genome.

230

ANGELA LUTTKE AND SILVANO BONOTTO

FIG. 19. Chloroplast DNA of A . mediterranea released by osmotic shock from a chloroplast of the cell apex. (Courtesy of Dr Antonio Mazza, International Institute of Genetics and Biophysics, Naples.)

This technique has been adopted for the visualization of DNA molecules from Euglena chloroplasts (Ishida er al., 1970). Successful application has been made of the osmotic shock technique in, looking for ultrastructural differences of the DNA in chloroplasts from the apical, middle, and basal part (Mazza et al., 1977). Though unambiguous measurements of entangled molecules are rather difficult, the data show a decrease in length from the apex toward the base of the cell (Table VI).

23 1

CHLOROPLAST DNA OF A . M E D I T E R W E A TABLE V LENGTHMEASUREMENTS ON CHLOROPLAST DNA Length ( I*m)

4 5- I000

Configuration

OF

A . mediterranea

Reference

Method

Linear Displays

Enzyme digestion Osmotic shock

419

Displays, supercoils

Osmotic shock

2 1-80

Displays, supercoils, linear Linear Linear Displays Displays, linear Circular

Osmotic shock, phenol extract ion Lysis with sarkosyl Molecular sieving Osmotic shock Osmotic shock, lysis/ CsCl extraction, lysisl molecular sieving Lysis with sarkosyl Osmotic shock, lysis/ CsCl extraction, lysis/ molecular sieving

48- 130 16200 3G2000 30- 100 0.75-2.8 4.3 0.1-1.5

Circular Circular

Werz and Kellner (1968) Woodcock and Bogorad ( 1970) Green and Burton ( 1970) Green et a / . (1970) Green (1972) Bonotto er a / . (1975) Mazza el a / . (1977) Mazza el a / . (1979)

Green (1976, 1977) Mazza et a / . (1980)

In addition to the differences in length, differences were also observed for the DNA ultrastructure between apical and basal plastids (compare Figs. 19 and 20) in that “spirals” were more abundant in basal organelles (Mazza et af., 1977, Figs. 5-7). Though no definite interpretation can be made, it should be recalled that in thin sections and in fluorescently stained plastids the DNA appears more dispersed in the stroma in apical plastids than in basal ones. It is tempting to speculate that the “spirals” indicate a more compact packaging of the DNA. A third point of importance is the occurrence of “bushy” structures being

LENGTHDISTRIBUTION

OF

TABLE VI APICAL., MIDDLE, VF A . mediierruneoli

DNA

IN

AND

BASALCHLVRVPLASTS

Length range Chloroplast type

%

I*m

Molecular weight

Apical

90

900-2000 I00 500-700 3G200

1.8-4.0 x 10’ 2.0 x 108 1.G1.4 X 10” 0.6.4.0 x 108

10

Middle

70 30

Data from Mazza et d.( I 979)

232

ANGELA LUTTKE AND SILVANO BONOTTO

FIG.20. Chloroplast DNA of A. mediterranea released by osmotic shock from a chloroplast of the cell base. (Courtesy of Dr Antonio Mazza, International Institute of Genetics and Biophysics, Naples.)

CHLOROPLAST DNA OF A . MEDITERRANEA

233

more frequent in apical plastids and interpreted as transcribed RNA (Mazza et a l . , 1977, Figs. 2 and 4). Since the small ovoid plastids in the apex with a still poorly developed thylakoid system are likely to be active in transcription, the authors’ view seems to fit the physiological situation of a higher protein synthesis in the apex (D’Emilio et a f . , 1979). Similar bush-like material attached to the bacterial nucleoid in spread preparations is also thought to represent nascent RNA chains (Kavenoff and Bowen, 1976). A new attempt for the visualization of the DNA ultrastructure was recently made by Mazza et al. (1979) by spreading purified DNA according to Rattner et al. ( 1975) and using the low-molecular-weight benzyldimethylalkylammonium chloride (BAC) as support film (Vollenweider et al.. 1975). This method is a modified version of the Miller and Beatty technique (1969) for the identification of chromatin fibers in the electron microscope. The micrographs provide a novel appearance for cpDNA. With either positive or negative staining large networks of fibers are visible, which depart from ringlike structures of 150 nm in diameter (Mazza er a l . , 1979, Figs. 8-10). Both the fibers and the ring-like structures display a globular substructure, the globules measuring 100-140 A.Two lines of evidence suggest that what we are looking at represents some kind of “secondary structure”: ( 1) pure double-stranded DNA spread on a BAC film has a width of 50-60 A after rotary shadowing, while double-stranded DNA spread on a cytochrome c film and rotary shadowed has a width of 150-200 (for review see Fisher and Williams, 1979). The globules of cpDNA of Acetabularia are twice as wide as normally found in BAC preparations, though no rotary shadowing is employed; ( 2 ) the globules appear to be spaced along threads of smaller width. At present any further interpretation of these structures appears premature particularly since a method known to preserve DNA/protein complexes was applied to CsCl purified DNA, which has presumably lost most of its associated proteins. Nonetheless we are convinced that proteins in association with cpDNA are a prerequisite for keeping a complicated supramolecular structure. Promising in this respect are the recent results of Dron et a f . (1979) on the presence of proteins in the nucleoid of Chlarnydornonas chloroplasts. In order to understand the meaning of the DNA network with a globular substructure from Acetabufaria chloroplasts further analyses are necessary. Puzzling is the finding of minicircles in the chloroplast preparations of A . rnedirerranea (Mazza et a f . , 1979, 1980). In contrast to the minicircles found in chloroplast preparation of A . clijitonii (Green, 1976, 1977) those of A . rnediterranea are smaller and of the same buoyant density as main band DNA. No functions can be ascribed to this minor DNA species and more questions are opened up than can presently be answered. For example: (1) are the minicircles merely artifacts introduced by the DNA isolation procedure; ( 2 ) are they “episomes” carrying cistrons for antibiotic resistance (Green, 1976; Mazza et a l . ,

234

ANGELA LUTTKE AND SILVANO BONOTTO

1980); and (3) do they have any genetic function at all. It is noteworthy that minicircles were also found in cpDNA preparations of Nicotiana tabacum and Phaseolus (Uphadhyaya and Grun, 1975) and in total cell DNA preparations of E. gracilis (Nass and Ben-Shaul, 1972). In neither case is their function clear.

E. PHYSICALPROPERTIES Investigations on the physical properties of cpDNA of Acetabularia are almost nonexistent. Heilporn and Limbosch (1971) pointed out that the incomplete renaturation of heat-denatured probes (it was compared to the banding in neutral CsCl gradients) suggests a large kinetic complexity. A large kinetic complexity was indeed later calculated from the renaturation rate constants (Green et al., 1977). With an average of I .9 +- 0.4 X lo9 daltons the kinetic complexity of the Acetabularia plastome is of the same order of magnitude as the kinetic cornplexity of the E. cofi chromosome with 2.5 X loy daltons. In comparison with the values for cpDNA of other algae or higher plants (for review see Bedbrook and Kolodner, 1979) the kinetic complexity of the Acetabularia plastome is almost 10 times larger. On the basis of the DNA content per plastid and the kinetic complexity, the copy number per chloroplast amounts to 1.3 on the average (Green et a / . , 1977). Taking into account that about half of the plastids do not contain DNA during the vegetative stage prior to cap formation (see Section III,B) the actual copy number in a given single plastid might be 2- or 3-fold higher. Yet, in comparison to the plastome number calculated for other algae and higher plants (Kirk and Tilney-Bassett, 1978), it appears extremely low. A large kinetic complexity and a low copy number per chloroplast ha.s also been found for A . clifonii (Padmanabhan and Green, 1978). Three plastomes per chloroplast result from the average kinetic complexity of 1.52 0.26 X lo9 daltons and a DNA content of 7.9 X 10-15g. As common interpretation for the high level of polyploidy in chloroplasts is that it is a protective means for survival in case of otherwise lethal mutations. It seems though that nature approached the same problem in different ways. In Acetabularia the mere presence of several million chloroplasts per cell prevents the expression of a mutation occurring in one or a few plastids. In other words, the “defensive” function of several copies in one plastid is taken over by the plastids themselves. A large kinetic complexity is certainly what is to be expected from the exceptionally large size of the DNA molecules. The fact that no other investigations on the physical properties exist must be attributed to the problems involved in culturing large enough quantities under axenic conditions for biochemical work. A reasonable suggestion for biochemists is, therefore, to consider suitable electron microscopic techniques, which allow work on a microscale.

*

CHLOROPLAST DNA OF A. M E D I T E R W E A

235

IV. Chloroplast Gene Products Application of refined electron microscopic and biochemical techniques like agarose gel electrophoresis, restriction enzyme analysis, and heteroduplex formation led to tremendous progress in respect to the physical mapping of the plastome, gene localization, and arrangement (for review see Bedbrook and Kolodner, 1979). Our knowledge in all these areas for the cpDNA of Acerabuluria is far behind that for other species. To date all our knowledge about the informational content of the plastome comes from in vivo and in virro studies of the chloroplasts. Hence, even carefully performed experiments give only circumstantial evidence for cpDNA genes. The measurement of end-products, often the only feasible experimental approach, leaves doubts as to whether the total synthetic sequence of a multistep process depends on cpDNA-encoded enzymes (see for example Green, 1980). On the other hand, for their longevity and integrity isolated Acetabufaria chloroplasts are particularly well-suited for in organufo studies and therefore can better enrich our knowledge on the real biosynthetic potential than plastids of many other sources. Evidence that the DNA of the Acetabularia chloroplasts contains the genes for the chloroplastic rRNA and tRNA came from incorporation studies of various RNA precursors into whole and enucleate cells (Janowski, 1965; Schweiger er al, 1967; Schweiger, 1970). In this context it is noteworthy that at least part of the chloroplast ribosomal proteins are coded by the nuclear genome (Kloppstech and Schweiger, 1973a,b). Berger (1967a,b) has shown that both RNA species are also synthesized by isolated chloroplasts. Since the incorporation of the RNA precursors is inhibited by darkness, actinomycin C, rifamycin, and DNase, and the results from experiments with sterile and nonsterile cultures were essentially the same, her controls firmly rule out the syntheses being due to bacterial and/or nuclear contamination. Isolated chloroplasts of Acerabularia incorporate 14C02 in comparable amounts to whole cells (Shephard et al.. 1968; Bidwell et al., 1969; Shephard and Bidwell, 1973). The labeling pattern for the alcohol-soluble products of photosynthesis is essentially the same for the chloroplasts and the cells. Also, the incorporation of I4CO2for 6-8 hours leads to an appreciable label in the chloroplast proteins. Again the pattern of the labeled amino acids from the hydrolyzed proteins of isolated chloroplasts is very similar to that of whole cells (Shephard and Levin, 1972). The same experimental approach of long-term photosynthesis in organufo was used to demonstrate that all the plastids’ pigments are synthesized from I4CO2 (Moore and Shephard, 1977). Chloroplast autonomy for pigment synthesis was further evidenced in that chloroplasts isolated from cells after 65 days of enuclea-

236

ANGELA LUTTKE AND SILVANO BONOTTO

tion synthesize pigments at normal rates and with the same composition as control cells (Moore and Shephard, 1978). The authors argument against a cytoplasmic supply of enzymes involved in the pigment synthetic pathways via stable mRNA of nuclear origin is that enucleate cells in stationary phase were used. Since by that time most of the cytosolic enzymes are declining it is concluded that the enzymes are transcribed from the plastid DNA. It is further argued that the modulation of the pigment content and composition in response to various light regimens in enucleate cells being indistinguishable from that in nucleate cells is also evidence for cpDNA as the site for enzyme transcription. Hoursiangou-Neubrun et al. (1979) noted an increase in the chlorophyll a/b ratio 3 weeks postenucleation. Together with other chloroplast characters the change was taken as an indication for an early chloroplast aging and therefore demonstrating nuclear-dependent chloroplast differentiation. According to Moore and Shephard (1978) it would merely reflect the relief of some kind of nuclear control on otherwise autonomous interplastidal regulations. Isolated chloroplasts also synthesize the plastoquinones A, 9, and C in equal rates to whole cells from 14C0, (Tschismadia and Moore, 1979). No statement on the site of transcription of the enzymes is made. From the whole set of experiments on pigment synthesis and content of chloroplasts derived from enucleate cells and the fact that part of the chloroplast ribosomal proteins are coded for by the nuclear genome (Kloppstech and Schweiger, 1973a.b) it must be concluded that chloroplast ribosomes are stable over several weeks. A recent study on the effect of chloramphenicol, rifampicin, and cycloheximide on the synthesis of ribulose- I ,5-diphosphate carboxylase (RuDPCase) in whole and in enucleate fragments indicated that the enzymes in Acetabularia depend on both nuclear and chloroplast genomes, like in other plant cells (Ibragimova et al., 1979). The strong increase of monogalactolipids in enucleate cells favors chloroplast autonomy for this thylakoid component (Hoursiangou-Neubrun er al., 1979). It is long been known that isolated chloroplasts of Acetabularia can utilize amino acids for the synthesis of both soluble and structural proteins (Goffeau and Brachet, 1965; Goffeau, 1969). Green has shown in a recent study (Green, 1980) that the incorporation is completely light dependent, sensitive to chloramphenicol and lincomycin, but insensitive to cycloheximide. These controls firmly rule out the incorporation being due to bacterial or cytoplasmic: contamination. The real aim of this investigation was to get insight into the coding capacity of the large plastome. Under the aspect that the large size implies a large coding capacity the results are most striking in that both the pattern arid the number of labeled polypeptides resemble those of higher plants. Hence the original idea of a large coding potential is not supported by these recent findings.

CHLOROPLAST DNA OF A. M E D I T E R W E A

237

They also contradict the large kinetic complexity (see Section 111,D) if the molecular organization of the plastome is thought about in conventional terms. An encouraging result of Green’s experiments is the finding that the apoprotein of CP I (the P-700 containing reaction center of photosystem I) is made on chloroplast ribosomes. To the best of our knowledge it is the first report unequivocally showing the synthesis of a specific chloroplast gene product in Aretabuluria.

V. Summary and Outlook Our writing of this article was governed by two basic ideas. Our purpose was to encourage more investigators to have a close look at that exceptional plastome. By presenting the currently available information including the characteristic features genuine to the organism and also pointing to the problems and open questions we think decision-making for any of the numerous unsettled topics will be easier for those not familiar with Acetabularia. We also thought it the right time to propose a simple model for the molecular organization of the DNA, which takes into account the somewhat controversial results accumulated. We are aware of the highly hypothetical character, but we are convinced that looking at it in the right context it gives a good working hypothesis with which to start. In either case we shall be equally glad about any proof or disproof, since even if only disproof results, we think this work was worth doing. We shall summarize what we believe are the most important results leading to the proposal outlined below: 1. cpDNA of Acetabularia at the level of fluorescence light microscopy and electron microscopy in thin sections is visible in plastids containing DNA in distinct regions of the organelle as a compact structure in close spatial relation to packages of ribosomes. 2. The DNA spread into a protein monofilm exists of exceptionally long molecules or complex networks. The length of the molecules by far exceeds that of other species and suggests a large genetic informational content. 3. The intramolecular heterogeneous DNA molecules have a high kinetic complexity. This also suggests that the plastome, which is present in a few copies only, contains a high genetic informational content. 4. The pattern and the number of labeled polypeptides from protein synthesis in organulo argue against a greater coding capacity than found for other species.

The simplest explanation, which appears compatible with all the listed results, is schematically outlined in Fig. 21: (1) the plastome consists of several se-

238

ANGELA LUTTKE AND SILVANO BONO’M’O

n-1-3

J

Comparative simplified representation of the conventionally found plastome and the hypothetical plastome of A. mediterranea. Transcribed gene clusters are indicated by straight lines, nontranscribed regions by wavy lines. n designates the ploidy level. (A) “Normal” plastome of about 40 pm circumference present in ICk100 copies per chloroplast. (B) Hypothetical Acetabularia plastome with an equal set of transcribing regions as in (A), separated by long nontranscribed regions. The plastome is present in 1-3 copies per chloroplast. FIG.

21,

quences of active genes with a genetic informational content similar to that found in other species; (2) each set should fit into a length of about 40 pm; (3) the plastome also contains “silent” nucleotide sequences of variable length, which are interspersed between the active sequences. These stretches could represent (a) sequences of too low a GC content to code for any protein; (b) sequences with lost genetic function; or (c) sequences with stabilizing functions for the supramolecular topology in situ. We hope some readers feel encouraged to prove whether further experimental results meet the theoretical predictions of this proposal.

ACKNOWLEDGMENTS

We thank Dr. S. Puiseux-Dao for critical reading and suggestions to the manuscript, Dr. A.-C. Dazy for the information on the presence of actin in Acetabularia cells, and Mrs. Chris HenielaersGielen for typing the manuscript. Our thanks are also due to Dr. Simone Puiseux-Dao for providing us with Figs. 9 and 17 and to Dr. Antonio Mazza with Figs. 19 and 20.

REFERENCES Astaurova, 0. B., Afanasova. L. A., Salamakha, 0. V., and Yazykov, A. A. (1979) In “Developmental Biology of Acetabularia” (S. Bonotto, V. Kefeli, and s. Puiseux-Dao, eds.), pp. 25 1-268. Elsevier, Amsterdam. Bastia, D., Chiang, K.-S., Swift, H.. and Siersma, P. (1971). Proc. Narl. Acad. Sci. U.S.A. 68, 1157-1 161.

Bedbrook, J. R., and Kolodner, R. (1979). Annu. Rev. Planr Physiol. 30,593-620. Berger, S. (1967a). Dissertation, University Koln.

CHLOROPLAST DNA OF A. MEDITERRANEA

239

Berger, S. (1967b). Protoplasma 64, 13-25. Bibby, B. T.. and Dodge, I . D. (1974). J. Ultrasrruct. Res. 48, 153-161. Bidwell, R. G. S. (1972). Nature (London) 237, 169. Bidwell, R. G. S., Levin. W. B., and Shephard. D. C. (1969). Plant Physiol. 44. 946954. Boloukhere, M. (1970). In “Biology of Acetabularia” (J. Brachet and S. Bonotto, eds.), pp. 145-155. Academic Press, New York. Boloukhere-Presburg. M. (1965). J. Microsc. 4, 363-372. Boloukhere-Presburg, M. (1966). J. Microsc. 5 , 619-628. Boloukhere-Presburg, M. (1972). J. Microsc. 13, 401-416. Bonotto, S., Lurquin. P. F., Baeyens, W., Charles, P., Hoursiangou-Neubrun. D., Mazza, A , . Tramontano, G., and Felluga, B. (1975). Protoplasma 83, 172-173. Bonotto, S., Lurquin, P. F., and Mazza, A. (1976). Adv. Mar. Biol. 14, 123-250. Borghi, H., Puiseux-Dao, S., Bonotto, S . , and Hoursiangou-Neubrun.D. (1973). Protoplasma 78, 99-1 12. Brachet, J. (1968). Curr. Top. Dev. Biol. 3, 1-36. Cairns, J. (1962). J. Mol. Biol. 4, 407-409. Cairns, J. ( 1963). J. Mol. Biol. 6. 208-21 3. Clauss, H., Liittke, A , , Hellmann, F., und Reinert, J. (1970). Protoplasma 69, 31S329. Coleman, A. (1979a). J. Cell Biol. 82, 299-305. Coleman, A. (1979b). J. Cell Biol. 83, 355a. Crawley, J. C. W. (1963). Exp. Cell Res. 32, 368-378. Crawley, J. C. W. (1966). Planta 69, 365-376. Crawley, J. C. W. (1970). In “Biology of Acetabularia” (J. Brachet and S. Bonotto. eds.). pp. 73-83. Academic Press, New York. Dazy, A., Hoursiangou-Neubrun, D., and Sauron, M. E. (1982). Biol. Cell (in press). D’Emilio, M. A., Hoursiangou-Neubrun. D., Baugnet-Mahieu, L., Gilles, J., Nuyts, G., Bossus. A., Mazza, A., and Bonotto, S. (1979). In “Developmental Biology of Acetabularia” ( S . Bonotto, V. Kefeli, and S. Puiseux-Dao, eds.), pp. 269-282. Elsevier, Amsterdam. De Vitry, F. (1965). Bull. SOC. Chim. Biol. 47, 1325-1351. Dron, M.. Robreau, G., and Legal, Y. (1979). Exp. Cell Res. 119, 301-305. Fisher, H. W., and Williams, R. C. (1979). Annu. Rev. Biochem. 48, 649-679. Gibbs, S. P. (1970). Ann. N. Y. Acad. Sci. 175, 454-473. Gibbs, S. P., Cheng, D.. and Slankis, T. (1974). J. Cell Sci. 16, 557-577. Gibor, A,, and Izawa, M. (1963). Proc. Natl. Acad. Sci. U.S.A. 50, 1 164-1 169. Goffeau, A. (1969). Biochim. Biophys. Acta 174, 3 6 3 5 0 . Goffeau, A,, and Brachet. J. (1965). Biochim. Biophys. Acta 95, 302-313. Green, B. R. (1972). Protoplasma 75, 478. Green, B. R. (1976). Biochim. Biophys. Acta 447, 156-166. Green, B. R. (1977). “Colloques Internationaux C.N.R.S. No. 261-Acides Nucleiques et Synthese des Proteines Chez les VCgttaux (L. Bogorad and J. H. Weil, eds.), pp. 133-136. Editions du centre National de la Recherche Scientifique, Paris. Green, B. R. (1980). Biochim. Biophys. Aria 609, 107-120. Green, B. R., and Burton, H. (1970). Science 168, 981-982. Green, B. R., Heilporn, V., Limbosch, S., Boloukhere, M.,and Brachet, J. (1967). Proc. Natl. Acad. Sci. U.S.A. 58, 1351-1358. Green, B. R.. Burton, H., Heilporn, V., and Limbosch, S . (1970).I n “Biology ofAcetabularia” (J. Brachet and S . Bonotto, eds.), pp. 35-59. Academic Press, New York. Green, B. R., Muir, B. L., and Padmanabhan U. (1977). In “Progress in Acetabularia Research” (C. L. F. Woodcock, ed.), pp. 107-122. Academic Press, New York. Hammerling, J. (1953). Ihr. Rev. Cytof. 2, 475498.

240

ANGELA LU’lTKE AND SILVANO BONOTTO

Heilporn, V.. and Limbosch. S . (1970). In “Biology ofdcetabularia” (J. Brachet and S. Honotto, eds.), pp. 61-72. Academic Press, New York. Heilporn, V., and Limbosch, S . (1971). Eur. J. Biochem. 22, 573-579. Heilporn-Pohl, V., and Brachet, J . (1966). Biochini. Biophys. Actu 119, 429-431. Hoursiangou-Neubrun, D. (1979). Thesis, University of Paris VII. Hoursiangou-Neubrun, D., and Puiseux-Dao, S. (1974). Plant Science Lert. 2, 209-219. Hoursiangou-Neubrun, D., Dubacq, J . P.. and Puiseux-Dao, S. (1977). In “Progress in Acerabularia Research” ( C . L. F. Woodcock, ed.), pp. 175-194. Academic Press, New York. Hoursiangou-Neubrun, D., Dubacq, J.-P., Bonotto, S., and Puiseux-Dao, S. (1979a). In “Developmental Biology of Acetabularia” (S. Bonotto, V. Kefeli, and S. Puiseux-Dao, ed!;.), pp. 131-139, Elsevier, Amsterdam. Hoursiangou-Neubrun, D., Puiseux-Dao, S . , Bonotto, S . , Dubacq, I . P., and Luttke, A. (1979b). B i d . Cell. 35, 52a. Hoursiangou-Neubrun, D.. Bonotto, S., Liittke. A,. and Puiseux-Dao, S. (1981). P r o t o p l a s m 105. 35fF357. Ibragimova. G. B., Alijev, K. A,, and Machmadbekova, L. M. (1979). In “Developmental Biology of Acetabularia” (S. Bonotto, V. Kefeli, and S . Puiseux-Dao, eds.), pp. 255-258. Elsevier, Amsterdam. Janowski, M. (1965). Biochim. Biophys. Acta 103, 399-408. Ishida, M. R., Ikushima, T., Matsubara, T., Mizuma, N., and Kikuchi. T. (1970). Ann. Rep. Res. Reacr. Inst. Kvoto Univ. 3, 180-183. Kavenoff, R., and Bowen, B. C . (1976). Chrumosnma 59, 8P-101. Kirk, J . T. 0. (1971). In “Autonomy and Biogenesis of Mitochondria and Chloroplasts” (N. K. Boardman, A. W. Linnane. and R. M. Smillie, eds.), pp. 267-276. North-Holland Publ., Amsterdam. Kirk, J. T. 0..and Tilney-Bassett, R. A. E. (1978). “The Plastids,” 2nd Ed. Elsevier/NorthHolland, Amsterdam. Kleinschmidt, A . K. (1968). Methods Enzymol. 12. 361-377. Kloppstech, K., and Schweiger, H. G. (1973a). Exp. CellRes. 80, 63-68. Kloppstech, K., and Schweiger, H. G. (1973b). Exp. Cell Res. 80, 69-78. Koop, H.-U. (1979). Protoplasma 100. 353-366. Koop, H.-U., and Kiermeyer, 0. (1979). Eur. J. Cell Biol. 20, 123. Koop, H.-U., and Kiermeyer, 0. (1980a). Protoplasma 102, 295-306. Koop. H.-U., and Kiermeyer, 0. (1980b). Eur. J. Cell Biol. 22, 355. Luttke, A. (1981). Exp. Cell Res. 131, 483488. Luttke, A., and Bonotto, S. (1980). “European Meeting on Molecular Genetics and Biology of Unicellular Algae,” Book of Abstracts. p. 23. Liege, Belgium. Luttke, A., and Bonotto, S. (1981a). Protoplasma 105, 358-359. Luttke, A , , and Bonotto, S. (1981b). Planta. 153, 536-542. Luttke, A., Rahmsdorf. U.,and Schmid, R. (1976). Z. Narurforsch. 31c, 108-110. Luttke, A., and Rahmsdorf, U. (1979). I n “Developmental Biology of Acetabulariu” (S. Bonotto, V. Kefeli, and S. Puiseux-Dao, eds.), pp. 169-177. Elsevier, Amsterdam. Liittke, A., Hoursiangou-Neubrun, D., Puiseux-Dao, S., and Bonotto, S. (1980). Eur. J. CrllBiol. 22, 127. Mangenot, G., and Nardi, R. (1931). Travaux Cryptogam. 11, 459-463. Mazza, A., Bonotto, S . , and Felluga, B. (1977). In “Progress in Acetabularia Research” (C. L. F. Woodcock. ed.), pp. 123-136. Academic Press, New York. Mazza, A., Bonotto. S., Felluga, B.. Casale. A., and Sassone Corsi, P. (1979). In “Developmental Biology of Acetabularia” (S. Bonotto, V. Kefeli, and S. Puiseux-Dao, eds.), pp. 115-130. Elsevier, Amsterdam.

CHLOROPLAST DNA OF A. MEDITERRANEA

24 1

Mazza, A,, Casale, A,, Sassone-Corsi, P., and Bonotto, S . (1980). Biochem. Biophys. Res. Commun. 93, 668-674. Miller, 0. C., and Beatty, B. R. (1969). Science 164, 955-957. Moore, F. D., and Shephard, D. C. (1977). Protoplasma 92, 167-175. Moore, F. D., and Shephard, D. C. (1978). Proroplasma 94, 1-17. Nass, M. M. K., and Ben-Shoul. Y. (1972). Biochim. Biophys. Acta 272, 130-136. Naumova. L. P., Pressman, E. K., and Sandakchiev, L. S . (1976). Plant Science Lett. 6, 231-235. Nigon, V., and Heizmann, P. (1978). Int. Rev. Cytol. 53, 21 1-290. Padmanabhan, U.,and Green, B. R. (1978). Biochim. Biophys. Acra 521, 67-73. Paques, M., Bonotto, S., and Sironval, C. (1979). In “Developmental Biology ofAcetabularia” (S. Bonotto, V. Kefeli, and S. Puiseux-Dao, eds.), pp. 179-182. Elsevier, Amsterdam. Pardoll, D. M., Vogelstein, B., and Coffey, D. S. (1980). Cell 19, 527-536. Pettijohn, D. E., and Carlson, J. D. (1979). Cell Biol. 2, 1-57. Pettijohn, D. E., and Worcel, A. (1976). C.R.C. Crir. Rev. Biochem. 4, 175-202. Puiseux-Dao, S. (1968). C.R. Acad. Sci. Paris 266, 1382-1384. Puiseux-Dao, S. (1970). “Acetabularia and Cell Biology,” pp. 1-161. Logos Press, London. Puiseux-Dao, S. (1979). B i d . Cell. 34,83-90. Puiseux-Dao, S., and Dazy, A.-C. (1970). In “Biology of Acetabularia” (1. Brachet and S . Bonotto, eds.), pp. I 1 1-122. Academic Press, New York. Puiseux-Dao, S., and Gilbert, A.-M. (1967). C.R. Acad. Sci. Paris 265, 87C873. Puiseux-Dao, S., Gibello, D., and Hoursiangou-Neubmn, D. (1967). C.R. Acad. Sci. Paris 265, 406408. Puiseux-Dao, S., Hoursiangou-Neubrun, D., and Dubacq, J. P. (1978). In “Chloroplast Development” ( G . Akoyunoglou and J. H. Argyroudi-Akoyunoglou, eds.), pp. 345-352. Elsevier, Amsterdam. Puiseux-Dao, S., Hoursiangou-Neubmn, D., and Dubacq, J. P. (1979).In “Developmental Biology of Acetabularia” (S. Bonotto, V. Kefeli, and S. Puiseux-Dao, eds.), pp. 141-153. Elsevier. Amsterdam. Rattner, J. B., Branch, A., and Hamkalo, B. A. (1975). Chromosoma 52, 329-338. Ryter, A., and Chang, A. (1975). J. Mol. B i d . 98, 797-810. Schmid, R., and Clauss, H. (1974). Protoplasma 82, 283-287. Schmid, R., and Clauss, H. (1975). Protoplasma 85, 315-325. Schmid, R., and Clauss, H. (1977). In “Progress in Acetabularia Research” (C. L. F. Woodcock, ed.), pp. 255-269. Academic Press, New York. Schulze, K. L. (1939). Arch. Protistenkunde 92, 179-225. Schweiger, H. G . (1970). Symp. Soc. Exp. Bor. 24, 327-344. Schweiger, H. G., Dillard, W. L., Gibor, A., and Berger, S. (1967). Protoplasma 64, 1-12. Scott, N. S., and Possingham, J. V. (1980). J . Exp. Bot. 31, 1081-1092. Shephard, D. C. (1965a). Exp. Cell Res. 37, 93-1 10. Shephard, D. C. (1965b). Biochim. Biophys. Acta 108, 635-643. Shephard, D. C. (1970). In “Biology of Acetabularia” (J. Brachet and S. Bonotto, eds.), pp. 195-212. Academic Press, New York. Shephard, D. C., and Bidwell, R. G. S. (1973). Protoplasma 76, 289-307. Shephard, D., and Levin, W. B. (1972). J. Cell B i d . 54, 279-294. Shephard, D. C., Levin, W. B., and Bidwell, R. G . S. (1968). Biochem. Biophys. Res. Commun. 32, 413-420. Silva. P. C. (1952). Univ. Cd$. Pub/. Bor. 25, 241-323. Slavik, N. S., and Hershberger, C. L. (1976). J . Mol. Biol. 103, 563-581. Tschismadia, J., and Moore, F. D. (1979). In “Developmental Biology of Acetabularia” (S. Bonotto, V. Kefeli, and S. Puiseux-Dao, eds.), pp. 183-194. Elsevier, Amsterdam.

242

ANGELA LUTTKE AND SILVANO BONOTTO

Upadhyaya, K. C., and Grun, P. (1975). I n “Proc. Symp. Struct. And Funct. Aspects of Chromosomes,” pp. 47-61, Bhabha Atomic Research Centre, Bombay. Vanden Driessche, T. (1973). Sub. Cell Biochem. 2, 33-67. Vanden Driessche. T., and Hars, R. (1972). J . Microsc. 15, 91-98. Vanden Driessche, T., and Hars, R. (1973). Arch. Biol. (Brwrelles) 84, 539-551. Vanden Driessche, T., Hars, R.. Hellin, J., and Boloukhtk, M. (1973a). J. Ultrusrrucr. Res. 42, 479-490. Vanden Driessche, T., Hellin, J., and Hars, R. (1973b). Protoplasma 76, 465472. Van Gansen, P.,and Boloukhhre-Presburg. M. (1965). J . Microsc. 4, 347-362. Vettermann, W. (1973). Protoplusmu 76. 261-278. Vollenweider, H. J., Sogo, J. M..and Koller, Th. (1975). Proc. Nut/. Acud. Sci. U.S.A. 72, 83-87. Wen, G.(1965). Brookhaven Symp. Biol. 18, 185-203. Wen, G., and Kellner, G.(1968). J . Ulrrusrrucr. Res. 24, 109-115. Woodcock, C. L. F.. and Bogorad, L. (1970). J . Cell Biol. 44, 361-375.

INTERNATIONAL REVIEW OF CYTOLOGY. VOL. 77

Structure and Cytochemistry of the Chemical Synapses STEPHAN

MANOLOV*AND

WLADlMIR OVTSCHAROFFt

*Regenerotion Research Laboratory, Bulgarian Academy of Sciences. and ?Department of Anatomy, Histology and Embryology, Medical Academy, Sofia. Bulgaria I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Presynaptic Part A. The Membrane of Presynaptic Element . . . . . . . . . . . . . . . . . . .

B . Synaptic Vesicles. . . . . . . . C. Coated and Other Vesicles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Origin of the Synaptic Vesicles ......................... E. Other Organelles in the Axondl Ending ........... F. Cytoplasm of the Axonal Ending ........................ G. Presynaptic Density . . . . . . ........... ... H . Transmitter Release . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111. Synaptic Cleft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Postsynaptic Part.. . . . . A. Postsynaptic Membr ................................ B . Postsynaptic Thickening . . . . . . . . C. Subsynaptic Structures. ................................ V. Concluding Comments ......................... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

243 244 244 248 262 262 265 267 267 268 270 272 273 275 276 277 278

I. Introduction The synapses present morphologically and functionally specialized contacts for an excitatory and inhibitory action between the nerve cells as well as between neurons and some nonneuronal cells. The introduction of the electron microscopy was the most significant step in the investigation of the synaptic structure. Substantial data on the functional mechanisms of the synapses were provided by applying some other methods as well as subcellular fractionation, intracellular microelectrodes, intracellular markers, etc. Cytochemistry and autoradiography present a link between the morphological studies on one hand and the biochemical and physiological ones on the other. They proved useful in localizing the sites of many enzymes as well as receptors on the synaptic membranes and contributed a great deal to the study of the macromolecules building up the “synaptic apparatus” and the regions where the transmitters are synthesized and/or stored. In this article we illustrate some 243 Cnnvrinhl 0 IYX2 hv Academic Prcsi. Inc.

244

STEPHAN MANOLOV AND WLADIMIR OVTSCHAROFF

morphological and cytochemical data of the chemical synapses which are discussed in the light of current concepts in synaptology. The great majority of the interneuronal synapses as well as the neuromuscular junctions in vertebrates are chemical synapses. Structurally they have three basic components: the presynaptic part, synaptic cleft, and postsynaptic part.

11. Presynaptic Part

In most cases the presynaptic part is presented by terminal dilatations of the axon (bouton terminal) or by dilatations down the amylinated axon (bouton en passant) or by an extension of the myelinated axon in the region of the node of Ranvier (nodal synapse). Occasionally the presynaptic part is a dendrite or a neuronal perikaryon. Although some authors consider the size of the axonal ending of essential importance, one must bear in mind that substantial errors can be made as only a small peripheral part of the axonal ending can be seen in the ultrathin section, hence misjudging of the size is very probable. In such cases serial section,, must be used and statistics also help. In the cytoplasm of the presynaptic part many cell organelles and inclusions are to be found: synaptic vesicles, sacs of the smooth endoplasmic reticulum, mitochondria, microtubules, neurofilaments, glycogen particles, and others. A. THEMEMBRANE OF PRESYNAPTIC ELEMENT The membrane of the axonal ending appears to be a direct extension of the axolemma and is 6-8 nm thick. As with the central synapses it is covered with processes of the glial cells. The axolemma of the neuromuscular junction is covered by Schwann cell processes. Occasionally a dendrite or a perikaryon is apposed to the membrane of the axonal ending of central synapses without forming synaptic contact. Furthermore the glial sheat of the axonal ending can be interrupted and another axonal ending would come in contact with the membrane of the ending, which sometimes forms an axoaxonal synapse. There are instances when two neighboring terminal endings can form desmosome-like contact. By means of subcellular fractionation considerable data were accumulated about the synaptosomal membrane. According to Morgan et al. (1973) these membranes have an apparently simple protein pattern. The lipids which constitute these membranes are cholesterol, phospholipids, and gangliosides (Breckenridge et al., 1972). The phospholipids are regarded as particularly important for the synaptic function. It is accepted that phosphatidylinositol is located on the cytoplasmic face of the synaptosomal membranes and can influ-

THE CHEMICAL SYNAPSES

245

ence their fluidity (Hawthorne and Pickard, 1979). The 5'-nucleotidase, AChE' (Whittaker, 1972), Na+-K -ATPase (Daniel and Guth, 1975), and neuraminic acid (Morgan et al., 1973) are used as markers of the synaptosomal membranes. The synaptosomal membrane fraction includes presynaptic and postsynaptic membranes. Using cytochemical methods the enzyme equipment of the membrane of the axonal ending and the postsynaptic element can be established (Manolov and Ovtscharoff, 1979a). By conventional electron microscopy, as well as fixation with KMnO,, the cell membranes of the presynaptic part and the presynaptic membrane, in particular, seem entirely identical morphologically (Ovtscharoff, 1979a). On the basis of cytochernical studies (Manolov, 1976; Ovtscharoff and Manolov, 1977; Ovtscharoff, 1979a) of the presynaptic axonal part individual membrane segments can be distinguished according to their cytochemical characteristics: axonal membrane, membrane of the axonal ending, and presynaptic membrane. A point of some controversy in our scheme is the discrimination of the axolemma from the membrane of the axon terminal. This would not have been a problem if there were well-shaped axon terminals but discrimination of these two membranes is very difficult because of badly shaped endings and boutons en passant. The presence of alkaline phosphatase, AChE, and ATPase was detected on the axolemma (Ovtscharoff, 1973, 1979a; Manolov and Davidoff, 1976; Ovtscharoff and Manolov, 1977; Manolov and Ovtscharoff, 1979a). On the bouton membrane in the rat cerebral cortex one more enzyme, TPPase, was detected. The occurrence of this enzyme on the axonal and the bouton membranes is in agreement with the hypothesis of Fox and Duppel(1975) which advances the idea that thiamine diphosphate and thiamine triphosphate are the active components of thiamine on nerve cell membranes. The presynaptic membrane can easily be distinguished from the rest of the axon terminal membrane by conventional electron microscopy as well as by freeze-etch technique, which shows a lot of tiny dimples corresponding to VASs (Sandri et al., 1977). Originally they were described as synaptopores (Pfenninger et al., 1971). It is assumed that these are the regions where the synaptic vesicles attach to the presynaptic membrane. The VASs arrangement in hexagons is in accordance with the concept about the presynaptic vesicular grid (Pfenninger et +

'Abbreviations: AC, adenylate cyclase; ACh, acetylcholine; AChE, acetylcholinesterase; ADP, adenosine diphosphate; AMP. adenosine monophosphate; ATP, adenosine triphosphate; ATPase, adenosine triphosphatase; Bi. bismuth iodide impregnation; BIUL, bismuth iodide impregnation and staining with uranyl acetate and lead citrate; ChAT, choline acetyltransferase; CAMP, cyclic adenosine monophosphate; cGMP, cyclic guanosine monophosphate; EDTA, ethylenediaminotetradcetic acid: E-PTA, ethanolic phosphotungstic acid; GABA, y-aminobutyric acid; GTP, guanosine triphosphate; MAO, monoamine oxidase; PDE. cyclic 3',5'-nucleotide phosphodiesterase; PNS, peripheral nervous system; TPPdse, thiamine pyrophosphatase; UTP, uridine triphosphate; VASs, vesicle attachment sites; ZIO, zinc iodide-osmium tetroxide staining.

246

STEPHAN MANOLOV AND WLADIMIR OVTSCHAROFF

al., 1969). On the corresponding E-face appropriate small, crater-like openings are lined up. The VASs are found on the presynaptic membrane of the neuromuscular junctions. Their number depends upon the synaptic activity (Heuser er a l . , 1974). Streit et al. (1972) had earlier found differences in the number of synaptopores in the spinal cord of anesthetized and unanesthetized rats. While at the central synapses VASs are arranged in hexagons, at the motor endplate they are located on both sites of the bars, which correspond to the dense projections of the central synapses (Sandri et al., 1977). On the P-face of this presynaptic membrane of the neuromuscular junction on both sites of the bars pairs of 10-nm particles are localized. Correspondingly shallow grooves and pits are found on the complementary E-face. No special arrangement of the intramembrane particles in the central synapses, similar to that in the neuromuscular junctions, has been observed (Fig. 1). Two types of intramembrane particles on the presynaptic membrane are established: large (10-19 nm) and small (4-10 nm). The large particles are localized in the proximity of VASs while the small ones are distributed evenly. It is assumed that the large particles participate in the control of calcium flux and may represent calcium ionophores (Venzin er al., 1977). Landis er al. (1974) have observed that at excitatory synapses there is subtle prominence of large particles on the presynaptic membrane while at inhibitory synapses the distribution is indistinguishible from the nonsynaptic region. In such cases the possibilities for artificial aggregation of intramembranous particles under the influence of the cryoprotectant glycerol must be accounted for (Morel et al., 1980). On the basis of cytochemical investigations of the presynaptic membrane of central synapses and neuromuscular junction a number of enzymes were established which are related to the synaptic transmission: AChE, ATPase, AC, and on its inner surface PDE (Manolov, 1976; Manolov and Ovtscharoff, 1979a) as well. Concerning this localization of ATPase it must be taken into consideration that in synaptic membranes this enzyme is combined with the active transport of K + and Na+ as well as with active transport of a number of molecules such as the mediators, including some amino acids and choline (Ovtscharoff, 1979a). Deposition of the reaction product, when demonstrating the catechol-stimulated AC on the presynaptic membrane, was also found by Markov and Diinova ( 1975), Hervonen and Rechardt ( 1976), and Rechardt and Harkonen ( 1977). This localization of the enzyme shows that it participates in the presynaptic function (Goldstein et al., 1975) in a way which is different from the well-known hypothesis about the postsynaptic localization of AC (Greengard and Kebabbian, 1974). Studying the cytochemical localization of this enzyme we (Manolov and Ovtscharoff, 1979a) applied both the methods of Reik et al. (1970) and Hervonen and Rechardt (1976). The latter one gives a small amount of reaction product. Probably this is due to the fact that in the incubation medium of Hervonen and Rechardt (1976) there are tetramisol and ouabain, potent inhibitors

Uneven distribution of intramembrane particles on the presynaptic membrane. Rat cereFIG. I . bral cortex. FIG. 2. Large clear synaptic vesicles. Anterior horn of frog spinal cord. X 80,000.

248

STEPHAN MANOLOV AND WLADlMlR OVTSCHAROFF

of the alkaline phosphatase and Na+-Kf-ATPase. Bearing in mind these data it is possible that part of the reaction product obtained by the method of Reik er al. ( 1970) is the result of the action of these two enzymes. Adding the activator (0.1 M imidasol) of PDE to the second incubation medium revealed granules of reaction product on the cytoplasmic surface of the presynaptic membrane (Ovtscharoff, 1979a). This enzyme localization is partially negated by addition of the inhibitor of the enzyme-theophyline (Florendo et al., 197 1). Moreover, Pardos and Lentz (1976) have described PDE activity on the presynaptic membrane of the neuromuscular junction of a newt. They admit that such localization of the enzyme confirms the proposition that CAMP affects the cholinergic transmission and more specifically in the neuromuscular junctions (Goldberg and Singer, 1969). An alkaline phosphatase (Sugimura and Mizutani, 1979) and 5 ’ nucleotidase (Bernstein et al., 1978) are found on the presynaptic membrane of central synapses. On the basis of pharmacological, biochemical, and physiological investigations it was assumed that the peripheral as well as the central synapses have receptors on the presynaptic membrane which can regulate the discharge of the transmitters from the terminal boutons as well as to modulate their synthesis (Roth et al., 1975). These presynaptic receptors or “autoreceptors” execute the feedback during synaptic transmission, and take part in important regulatory mechanisms (Manolov and Ovtscharoff, 1979a). It is possible that they are localized down the entire neuronal surface (Carlsson, 1975). According to Bloom ( 1975) the presynaptic @-receptors,presynaptic muscarinic receptors, and presynaptic prostaglandin receptors at aminergic nerve endings can participate in the control of the amount of mediators released and thus effect negative feedback. B. SYNAPTIC VESICLES The synaptic vesicles are one of the most significant structures of the: presynaptic element. It is generally admitted that they are a storage site for the neurotransmitters. From a morphological point of view the synaptic vesicles are classified into two large groups: clear vesicles and dense core vesicles. It is highly probable that both groups are heterogeneous within themselves. 1. Clear Vesicles The size of this type of vesicle is about 40-50 nm and they are surrounded with a membrane 6-9 nm thick. Synaptic vesicles of axonal endings of the same type have a more or less identical size and on this basis different types of axonal endings can be distinguished. Moreover it must always be taken into consideration that the size of the clear synaptic vesicles can be modified as a result of functional and pathological changes in the synapses (De Robertis, 1958; Cragg, 1969; CuCnod et al., 1970; Kawana et al.. 1971; Cheresharov er al., 1978).

THE CHEMICAL SYNAPSES

249

Investigating the synaptogenesis in rat cerebral cortex, Ovtscharoff ( 1979a) observed clear synaptic vesicles of 80-120 nm in animals which had not reached maturity. Laramendi et a / . (1967) have found that the diameter of the vesicles decreases with age. The above data show that classification of the axonal endings on the basis of the size of the clear vesicles may be erroneous. We have observed large clear vesicles of about 90-160 nm (Fig. 2) in the axonal endings of carp and frog spinal cord (Ovtscharoff, 1979a; Manolov and Ovtscharoff, 1979a). Clear vesicles of similar size were found in the terminal boutons of the nerve system of invertebrates (Westfall et a/., 1971; Hemandes-Nicaise, 1973). It is probable that this type of vesicle is characteristic for invertebrates and lower vertebrates. In mammals synaptic vesicles of that size in the earlier stages of the synaptogenesis in the rat sensorimotor cortex were observed (Ovtscharoff, I979a). West and Del Cero ( 1976) have similar observations on the synaptic formation in the molecular layer of rat cerebellar cortex. These results suggest that clear large synaptic vesicles are apparently not mature or they indicate a more primitive nervous system. Nevertheless, it must be pointed out that the clear synaptic vesicles of mammals are a more homogeneous population in respect to size in comparison to the lower vertebrates and the invertebrates. The size of the synaptic vesicles has never been a subject of much controversy, while the shape of the vesicles attracted the attention of morphologists in the last 10-17 years. Two basic shapes have been observed: spherical and flattened or elongated. The quantitative amount between them in one bouton is different. Some times neighboring boutons one of which contains spherical vesicles and the other one flattened vesicles can be observed. Uchizono (1965) suggested that the terminal boutons containing flattened vesicles are inhibitory and those containing spherical ones are excitatory. Other authors assumed that the flattening of clear vesicles was caused by aldehyde fixation (Walberg, 1965; Lund and Westrum, 1966; Ceccarelli and Pensa, 1968). Manolov (1967~)describes flattened vesicles in rat neuromuscular junction after glutaraldehyde fixation. These data annul Uchizono’s hypothesis at least for the case of mammal neuromuscular junctions. We observed boutons with flattened vesicles in snake cerebral cortex (Fig. 3) after osmium fixation (Manolov and Ovtscharoff, 1979a). Valdivia (1971) proved that the basic factor for flattening of the vesicles is the osmolarity of the buffer used during the treatment of the material. Korneliussen (1972a) points out the importance of the change in the molarity of the fixing buffer for the flattening of vesicles. Studying unfixed tissue by freeze-etching, only spherical vesicles were found in the axonal endings (Sotelo, 1971; Akert et al., 1972). Hence the natural shape of the synaptic vesicles appears to be spherical (Pfenninger, 1973) and the in vitro flattening of the spherical vesicles may be due to the specific sensitivity of the membrane of some types of vesicles. There are data showing that with age vesicle elongation increases (Laramendi et al., 1967). A similar process may be taking place during degeneration of the synapses (Jones, 1975). Some authors have found an increase of the amount of flattened vesicles as a

FIG.3. Flattened synaptic vesicles after osmium fixation. Snake cerebral cortex. X95,OOO. FIG.4. TPPase activity on the clear synaptic vesicles. Rat cerebral cortex. X400,OOO.

THE CHEMICAL SYNAPSES

25 1

result of electrostimulation (Birks, 197 1 ; Korneliussen, 1972b; Korneliussen et al., 1972). It is possible that stimulation lowers the osmotic pressure in the vesicle interior by reducing the amount of mediator in them and renders them more sensitive to the flattening action of the fixative (Manolov and Ovtscharoff, 1979a). Studying the changes taking place in spinal cord axodendritic synapses of rats which were exposed to prolonged locomotor activity we found an increased number of axonal endings with flattened vesicles. Whether the mediator exerts any influence upon the vesicle deformation is a problem which remains to be clarified (Matus and Dennison, 1971). The amount of synaptic vesicles in the axonal endings varies widely. There are boutons with a few vesicles as well as boutons which are entirely filled up with vesicles. As in most of the cases, the synaptic vesicles are in the proximity of the “active zone”; it must be admitted that in such instances the direction of sectioning may be of considerable importance to the number of synaptic vesicles. The last may depend on age as well as on the functional activity of the synapses. During synaptogenesis we found that in the initial stages of their formation the synapses have a few synaptic vesicles. These results are similar to the results of other authors, who have followed synaptogenesis (Jones and Revel, 1970; Jones, 1975; Konig et al., 1975). As for the fluctuation in the number of synaptic vesicles which occur during different types of stimulations, the results obtained are contradictory: without changes, there is an increase as well as reduction of their number (Hubbard and Kwanbunbumpen, 1968; Jones and Kwanbunbumpen, 1968, 1970; Siegesmund et al., 1969; Cragg, 1969; Birks, 1971; Heuser, 1971; Jones and Bradford, 1971a). Our studies showed that long-term motor activity in rats increased the number of boutons of the spinal cord axodendritic synapses with a small number of synaptic vesicles in contrast to those containing many vesicles. Numerous studies show a reduction of the amount of synaptic vesicles after stimulation with black widow spider venom and a-bungarotoxin (Chen and Lee, 1970; Landon et al.. 1980). Couteaux and PCcotDechavassine (1973) point out that the intensity of the vesicles depends to a great degree on the method of fixation and the following treatment of the material. Some authors assume that the number of synaptic vesicles can be a characteristic feature of a particular type of axonal ending originating from a certain region in the CNS (Rinvik and Grofova, 1974). On serial sections we have found that one bouton can acquire a different aspect of reference to number, type, size, and shape of vesicles (Ovtscharoff, unpublished data). The components of the vesicle membrane of the clear synaptic vesicles more specifically of the cholinergic ones are mainly obtained from studies of the synaptic vesicle membrane of the electric organ of Torpedo murmoruta (Breer ef al., 1978) and Narcine brusiliensis (Wagner et al., 1978). It was found that the cho1esterol:phospholipid ratio is 0.42: 1 (Morris, 1980). Integral and peripheral proteins are found in this lipid bilayer (Singer and Nicholson, 1972). Actin

252

STEPHAN MANOLOV AND WLADIMIR OVTSCHAROFF

(Tashiro and Stadler, 1978; Wagner and Kelly, 1979) and Ca2+-Mg2 +-ATPase (Breer et al., 1977) were also detected in the vesicle membrane. Studying the localization of ATPase in the synapses of the rat cerebral cortex we found that some reaction product was deposited on the membrane of the clear synaptic vesicles (Ovtscharoff and Manolov, 1977). It is possible that in this case the enzyme acts as a pumping mechanism and is involved in the translocation of protons (Winkler and Westhead, 1980). In our opinion there is a second possibility as well: one part of the vesicles in the axonal ending might be ordinary pinocytotic vesicles, which membrane originates from the membrane of boutons since they have a well-manifested enzymic activity (Ovtscharoff and Manolov, 1977). ATPase could be used as a marker of the vesicle heterogeneity and consequently as a confirmation of this assumption. We would like to point out that in the nerve endings of many capsulated receptors a great number of vesicles morphologically similar to the synaptic vesicles are found which can be taken as further proof of our idea (Chouchkov and Ovtscharoff, 1980). We have also observed TPPase activity (Fig. 4) on the membrane of the clear synaptic vesicles (Ovtscharoff and Manolov, 1977). Seijo and De Lorez Arnaiz (1970), applying microchemical analysis, found TTPase in the synaptosomal fraction and Pellegrino de lrraldi and De Lorez Amaiz (1970) found it in the vesicle fraction of the proximal part of the intermittent sciatic nerve. The specificity of the cytochemical reaction is an important problem as there are data showing discrepancies between the biochemical and the histochemical results concerning the characteristics of the enzyme (lijima, 1975). Some observations show that this enzyme is identical with the nonspecific nucleoside diphosphatase, which is associated with several membrane fractions of the CNS (Barchi and Braun, 1972). One should be careful when interpreting the results of experiments in which the substrate used was cocarboxylase or when nucleoside diphosphates were applied. In cases in which cocarboxylase was substituted with ATP wse also detected deposition of the reaction product on the vesicle membrane though in a lesser amount in comparison to the original method (Ovtscharoff and Manolov, 1977). These facts can be explained in the following way: it is possible that we have detected ATPase by not working at optimal conditions-Mn2+ is applied as an activator instead of Mg2+ or that TPPase or nucleoside diphosphates has reacted with a nonspecific substrate. Griffith and Bondareff (1973), Knyihlr et al. ( 1973), and Csillik et ul. ( 1974) have also observed TPPase on the membrane of the clear synaptic vesicles as well. Winkler (1977) assumes that the chromaffin granule membrane originates from the Golgi cisternae. Recently Sugimura and Mizutani ( 1979) demonstrated alkaline phosphatase on the rat synaptic vesicle membrane. Biochemically Ca2 -activated alkaline phosphatase has been confirmed as an integral protein of the vesicle membranes isolated from bovine cerebral cortex (Zisapel and Haklai, 1980). Pardos and Lentz (1976) have cytochemically demonstrated PDE 011 the synaptic vesicle membrane from neu+

THE CHEMICAL SYNAPSES

253

romuscular junctions of newts. This enzyme was detected on synaptic vesicle membrane from rat cerebral cortex only after the enzyme activator (0.1 M imidazol) was added to the incubation medium (Ovtscharoff, 1979a). PDE activity was established biochemically in the vesicles from the mouse CNS (Johnson et al.. 1973). An immunocytochemical technique was used by McLaughlin et al. (1974) to detect glutamate decarboxylase, an enzyme participating in the synthesis of the inhibitory transmitter GABA on the clear synaptic vesicle membrane from rat cerebellum. The cytochemical demonstration of a number of enzymes on the outer surface of the clear synaptic vesicle membrane is in agreement with the results of Wagner and Kelly (1979) who find that most of the basic proteins in the cholinergic vesicles isolated by Narcine brasifiensis are located on the external surface of the vesicles. On this surface tubulin has been identified as an integral protein (Zisapel et ul., 1980). By ruthenium red staining Ovtscharoff (1975a,c) has detected the presence of acidic groups on the synaptic vesicle surface. Digestion with trypsin negated the reaction whereas the action of hyaluronidase had a weaker effect. Thus it may be assumed that in the vesicle membrane there are acidic mucopolysaccharides (Castejon and Castejon, 1976) and acidic proteins as well as some glycoproteins (Richter-Landsberg et al., 1977) and glycolipids (Manolov and Ovtscharoff, 1979a). Khsa et al. (1970) introduced a cytochemical method for the demonstration of ChAT. Examining the localization of this enzyme in rat CNS we found deposition of reaction product on the external surface of the membrane of the round clear synaptic vesicle membranes as well as on the flattened ones (Ovtscharoff, 1977, 1979~).Hence it can be expected that the flattened vesicles may have a cholinergic nature. Feigenson and Barmet (1977) have reported deposition of reaction product on the vesicle membranes in the neuromuscular junction of newts by means of other cytochemical techniques for the demonstration of this enzyme. What holds the attention here is the fact that by biochemical methods ChAT has not been found in the vesicle fraction so far and it is thought of as a cytoplasmic enzyme (Fonnum, 1966). Cytochemical methods confirm this (Kisa et al., 1970; Kasa, 197 1; Ovtscharoff, 1977, 197912). As the enzyme is positively charged (Fonnum, 1973) its presence on some membrane (Ovtscharoff, 1977) may be due to the ionic attraction (Fonnum, 1973). This idea is rather plausible for the synaptic vesicle membrane where the presence of a great number of acidic groups has been cytochemically proved (Ovtscharoff, 1975a). Since the rat brain enzyme consists of three molecular forms with isoelectric points of 7.5, 7.7, and 8.4 (Malthe-Sorenssen and Fonnum, 1972), it is probable that one or more of these enzyme forms are electrostatically bound to the synaptic vesicle membrane and the other, or the others, could represent its cytochemical form. It is well known that Ca2+ is absolutely necessary for chemical synaptic transmission (Katz and Miledi, 1965, 1967; Grinell, 1966; Hackett, 1976). Dur-

254

STEPHAN MANOLOV AND WLADIMIR OVTSCHAROFF

ing the depolarization of the presynaptic part calcium enters into it. Though the exact function of calcium in these processes remains obscure it is admitted that divalent cations facilitate the fusion between the vesicle membrane and the presynaptic membrane (Hubbard, 1971 ) . Oschman and Wall (1972) identified calcium binding sites with calcium-containing fixatives. With the help of X-ray analysis it was established that the electron-dense granules on cell membranes are the regions where calcium is accumulated (Hillman and Llinas, 1974; Oschman et al., 1974). There are data which show the presence of calcium binding sites on the membrane of cholinergic synaptic vesicles (Politoff er al., 1974; Boyne et al., 1974). Our results reveal that some boutons of the CNS arid rat neuromuscular junctions contain synaptic vesicles on which membrane electrondense particles are observed after calcium-glutaraldehyde fixation (Ovtschiiroff, 1979a,b). Addition of EDTA to the fixative leads to the disappearance of the granules on the vesicle membrane as a result of the formation of chelates (Politoff et al., 1974; Boyne eral., 1974; Ovtscharoff, 1979b). This suggests that the dense material of the synaptic vesicles is mainly built up of calcium. Applying a freeze-etch technique Akert et al. (1977) found individual granules of about 8-1 3 nm on the P-face of the vesicle membrane. These authors report the granules to be identical to the calcium-binding sites on synaptic vesicles. The assumption that different cations can fuse with the clear synaptic vesicles (Miledi, 1964; Bloom and Barmett, 1967; Kokko and Barmett, 1971) as well as the fact that Mg2 and Co2 can block the synaptic transmission (Del Castillo and Katz, 1954; Weakly, 1973) led us to study the affinity of the clear synaptic vesicles toward different metalic cations (Ovtscharoff, 1979a, 198I). We used aqueous and ethanolic solutions of 19 different salts. The results showed that the clear synaptic vesicle membrane in some terminal boutons can bind monovalent (TI+), divalent (Fe2+, Ni2+, Co2+, Zn2 , Sr2+, Cd2+, Ba2+, Pb2+), trivalent (Cr3+, Fe3+, Sn3+), and tetravalent (TI4+) metalic cations (Fig. 5). We found that this property of the vesicle membrane is best represented with monovalent thallium, whereas Kokko and Barmett (1971) stress that it is best manifested with divalent cations (Pb2 , Cd2 ,Zn2 ). We did a series of controls in order to study the cytochemical characteristics of the metal-binding sites on the clear synaptic vesicle membrane (Ovtscharoff, 1979a, 1981). Blockage of sulfhydryl groups with iodoacetic acid or N-ethylmaleimide greatly decreased the number and the size of the dense granules on the vesicle membrane. As is known -SH combined with heavy metal cations give insoluble mercaptides. Kokko and Barmett (1971) presume that the precipitate on the clear vesicles is caused by CoA. However there are over 100 enzymes whose activity is determined by the sulfhydryl groups and the enzyme presence on this membrane is demonstrated in a biochemical and cytochemical way (Manolov and Ovtscharoff, 1979a). hdethylation also decreases the amount of synaptic vesicles with metal-binding :sites. Hence one can conclude that carboxyl and sulfate groups can also bring about the +

+

+

+

+

+

FIG.5 . Electron-dense spots on the synaptic vesicle membrane. Ethanolic solution of Cr(CH3COO)3.Rat cerebral cortex. X 138,000. FIG.6. Electron-dense material in the clear synaptic vesicles after treatment with acetic acid and postfixation with Os04. Rat cerebral cortex. x 189,000. FIG. 7. Electron-dense particles in the synaptic vesicles of rat neuromuscular junction after treatment with acetic acid and postfixation with Os04. x 162,000.

256

STEPHAN MANOLOV AND WLADIMIR OVTSCHAROFF

binding with heavy metals. The results of the digestion with proteolytic enzymes point out that the binding of the metalic cations is realized by protein and glycoprotein structures on the vesicle membrane (Ovtscharoff, 1979a, 198I). It is accepted that protein macromolecules have calcium-binding properties (Benshalom and Flock, 1977). We think it probable that the calcium-binding sites and the metal-binding sites are one and the same points which can generally be named metal-binding sites on the synaptic vesicles. It is possible that the symptoms of intoxication with heavy metals are due not only to the inhibitory action of the heavy metals upon some enzymic systems but may also be the result of the competition of the former for C a 2 + , in this way making the transmitter release almost impossible. Presently we lack cytochemical methods for the demonstration of mediators in the clear synaptic vesicles. Biochemical investigations show the presence of transmitter in these vesicles (Mclntosh, 1963; Marchbanks, 1969). There is an excellent source for isolation of clear cholinergic synaptic vesicles-the electric organ of the electric rays, family Torpedinidae (Sheridan et al., 1966). Marchbancks and Israel (1972) point out the presence of stable bound ACh on the interior of the cholinergic vesicles and loosely bound ACh on their external surface. In the electric organ of Torpedo marmorata Dowdall and Zimmermann ( 1974) have detected the presence of two types of ACh as well. One of them is characterized by a high ACh/ATP molar ratio, while the other type has a lower ACh/ATP ratio. Employing [3H]choline it has also been shown that some dear synaptic vesicles contain ACh (Marchbancks and Israel, 1971). The results of biochemical and autoradiographic studies suggest the presence of amino acid transmitters such as GABA, glycine, taurine, D-glutamic acid, a-aspartic acid, proline, and some others in the clear synaptic vesicles (Matus and Dennison, 1971; Roberts and Hammerschlag, 1972; Orrego, 1979). The detection of these amino acids in the axonal endings by autoradiographic methods and especially assigning a neurotransmitter role to them must be viewed with caution. The above mentioned amino acids may be part of the vesicle membrane, the ve.sicle core, or the axoplasm, so their presence in the synaptic vesicles or in the boutons is not sufficient evidence as to their neurotransmitter nature (Bloom, 1972; Manolov and Ovtscharoff, 1979a). The presence of nucleotides and in the first place of ATP in the clear synaptic vesicles has been pointed out many times (Whittaker and Zimmermann, 1974; Dowdall et a f . , 1974), so this nucleotide is now accepted as a second marker for the synaptic vesicles (Dowdall and Zimmennann, 1974). In cholinergic vesicles the presence of ADP and GTP was also shown (Wagner et al., 1978; Zimmermann, 1978). Whittaker et a f . (1974) found a low-molecular-weight acidic protein (MW 10,000) in osmotically ruptured cholinergic synaptic vesicles also called vesiculin. They assume that the vesiculin in the cholinergic system is an analog of

THE CHEMICAL SYNAPSES

257

chromogranin A (Smith and Winkler, 1967) or neurophysin C (Rauch et a / . , 1969) which are thought to participate in the packing of adrenaline in the chromaffin vesicles and vasopressin and oxytocin in the secretory granules in the neurohypophysis. It is assumed that ACh is packed as “acetylcholine vesiculinate” (Whittaker et a l . , 1974). Subsequently it was established that vesiculin is actually a glycosaminoglycan(Stadler and Whittaker, 1978). Cytochemically we showed the presence of proteinaceous material in the clear synaptic vesicles from various parts of the CNS and from the rat neuromuscular junction as well (Ovtscharoff, 1976a, 1978). The method employed is as follows: after glutaraldehyde fixation the tissue is incubated in acid and postfixed with OsO,. As a result electron-dense granules of about 7-15 nm appear in the interior of the clear synaptic vesicles (Figs. 6 and 7). In our opinion the presence of electron-dense granules in the clear vesicles is due to the precipitation and immobilization of some proteinaceous substance in these vesicles which is visualized after OsO, treatment. It is known that osmium tetroxide reacts with a number of amino acids, peptides, and proteins during tissue fixation (Bahr, 1954). We assume that visualized material is proteinaceous because after digestion with proteolytic enzymes the electron-dense particles practically disappear (Ovtscharoff, 1976a, 1978) whereas treatment with hyaluronidase and neuraminidase or lipid extraction has a weak effect (Ovtscharoff, 1979a). Electron-dense material is also found in clear synaptic vesicles after glutaraldehyde fixation and postfixation with OsO, at 60°C (Bloom and Aghajanian, 1968a). The appearance of electrondense material in this case is apparently due to the denaturation of proteins in the vesicles as a result of the calorific effect, which creates better conditions for interaction of some amino acids with OsO, (Hayat, 1970). Akert and Sandri (1968) adapted the ZIO method (Maillet, 1962) for cytochemical investigations. However in subsequent studies it was established that by this method both cholinergic and adrenergic vesicles can be impregnated (Kawana et a l . , 1969; Pellegrino de lrraldi, 1974). It was demonstrated that reaction product is deposited on other cell structures as well (Ovtscharoff, 1976b). Later it was found that this procedure visualized material of a lipid nature (Ostendorf et a!., 1971). Our results obtained after digestion of the tissue with trypsin and succeeded by lipid extraction support the above suggestion (Ovtscharoff, 1976b). The investigations of Reinecke and Walter (1978) and Pellegrino de lrraldi and Cardoni ( 1979) show that Z10 demonstrates intravesicular protein containing sulfhydryl groups. The results from X-ray powder microanalysis suggest that zinc osmate binds with high affinity calcium sites (Gilloteaux and Naud, 1979). Schmidt et al. (1980) have found that cholinergic synaptic vesicles from the electric organ of T. marmorata contain ACh, ATP, and metal cations (Na+, K + , Mg2+, Ca2+). At present it is not clear to what extent the positive charges of metal cations and ACh are counterbalanced by the negative charges of

25 8

STEPHAN MANOLOV AND WLADIMIR OVTSCHAROFF

vesiculin and ATP (Stadler and Whittaker, 1978). Stimulation of the electric organ leads to an increase of the level of the metal cations (Schmidt et ul., 1980). 2. Dense Core Vesicles According to their size the dense core vesicles are divided into two types: small and large ones; the border value between the two classes appears at 70 nm. The size of the small vesicles is about 30-60 nm and the large ones are usually above 80 nm. The dense core of these vesicles covers about half or a little more than half their diameter. Vesicles containing electron-dense material have been described for the first time by Palay (1955) in the neurohypophysis of rats.

3 . Small Dense Core Vesicles This type of vesicle was shown for the first time by De Robertis and Pellegrino de lrraldi (1961) in rat pineal. It should be noted, however, that it was Milofski (1957) who first described these vesicles without having them documented. There is convincing evidence which shows that the small dense core vesicles in the PNS (Iversen, 1967; Tranzer et af., 1969) as well as in the CNS (Hokfelt, 1968) represent a major monoamine store. In this instance what holds the attention is the nature of the clear synaptic vesicles which are collected in one and the same terminal bouton with the small dense core vesicles. It is assumed thi3t the former are the result of bad preservation of their dense core during fixation (Tranzer and Thoenen, 1967a; Tranzer et af., 1969). The best results in visualizing this type of vesicle were obtained by chromaffin reaction (Wood and Barmett, 1964) in rat pineal (Ovtscharoff, 1979a). Unfortunately this method does not result in the same quality when applied to central synapses. In order to treat a terminal bouton as an adrenergic one there must be at least one small dense core vesicle found in it. However, in studying the synaptic organization of a definite area of the CNS such a finding has to be proved at several occasions at least, because in some cases the small dense core vesicles may actually be a part of a large dense core vesicle correspondingly sectioned. There is also a phenomenon that has not been explained up to now: the presence of small dense core vesicles in the axon terminal which have a confirmed cholinergic nature (Jones, 1975; Manolov, 1976). The membrane of the adrenal chromaffin granules serving as a store of catecholamines (Hall, 1972) contains 30% of the total granule protein and almost all of the lipid. The proteins detected on the granule membrane are dopamine-@ hydroxylase, Mg2 -ATPase, cytochrome b-561, L-actinin, actin, phosphatidylinositol kinase, and synaptin (Winkler and Westhead, 1980). Cytochemically we found ATPase on the small dense core vesicle membrane (Ovtscharoff and Manolov, 1977). Proteins are found as membrane intercalated particles (9 nm size) on both faces of the vesicle membrane using the freeze-etch +

THE CHEMICAL SYNAPSES

259

technique (Eagles er al., 1977). The major glucosaminoglycan on the granule membrane is chondroitin sulfate, while heparan sulfate is found in a lesser amount (Winkler and Westhead, 1980). The major constituents of this membrane are cholesterol and phospholipids. The presence of neurotransmitters (dopamine, noradrenaline, adrenaline, serotonin) in the small dense core vesicles has been proved with cytochemical methods. Bloom (1972) thinks that the qualities of the cytochemical reaction for demonstration of the dense core vesicles follow a descending line: KMnO,, glutaraldehyde-dichromate-OsO,, glutaraldehyde-OsO,, and OsO,. Hokfelt (1971) considers that the fixation with KMnO, is advantageous because the speed of the reaction between monoamines and KMnO, is very much greater than that of the reaction between amines and OsO,. Using KMnO, fixation a greater number of dense core vesicles in the boutons of rat frontal cortex was established than with glutaraldehyde-0s0, fixation (Ovtscharoff, 1979a). Pellegrino de Irraldi el al. ( 1 97 1) state that it is possible to differentiate between catecholamine and indolamine-containing vesicles cytochemically employing different fixation methods combined with pharmacological treatment. The relative number of the small dense core vesicles diminishes after treating them with reserpine (Hokfelt, 1966) and metaraminol (Bondareff and Gordon, 1966). The small dense core vesicles which have been preliminarily depleted with reserpine reacquire their dense core after application of exogenous ‘‘true” transmitter (Bondareff, 1966). Application of “fals” transmitter 5-OH-dopamine leads to similar results (Tranzer and Thoenen, 1976b). Inhibiting of noradrenaline synthesis and subsequent electrical stimulation bring about the disappearance of the dense core of the small dense core vesicles (Hokfelt, 1967). On the other hand M A 0 inhibitors increase the possibility of demonstrating cytochemically the small dense core vesicles (Hokfelt, 1971). The cytochemistry of the biogenic monoamines has a great advantage-the presence of a magnificent light microscopic Falck-Hillarp technique for demonstration of these transmitters (Falck et a l . , 1962). We established an exceptionally high catecholamine and indolamine content including high density of fluorescent dots in the rostra1 part of the mesencephalon of turtles (Ovtscharoff, 1972). In the chromaffin granules the presence of ATP in a molar catecholamine/ATP ratio 4-4.5 has been established. However this ratio is not the same for all the chromaffin granules (Winkler and Westhead, 1980). It is generally accepted that ATP serves to neutralize the positive charges of four norepinephrine molecules (Hall, 1972). In addition to catecholamines and ATP the chromaffin vesicles also contain proteins (dopamine-P-hydroxylase, chromogranin A), enkephalin, ADP, AMP, GTP, UTP, Ca2+, Mg2+, ascorbinic acid, sialic acid, sulfated hexosamines, and glucuronic acid (Winkler and Westhead, 1980). The chief cells of the carotid and aortic bodies contain dense core vesicles

260

STEPHAN MANOLOV AND WLADIMIR OVTSCHAROFF

which are 60-140 nm in diameter and contain mainly dopamine. Like the clear vesicles these aminergic vesicles possess calcium-binding sites (Hess, 1977; Hansen and Smith, 1979).

4 . Large Dense Core Vesicles This type of vesicle has been described for the first time by Grillo and Palay (1962). Such vesicles have been found in the peripheral and central adrenergic axon terminals (Van Orden et al.. 1966; Bloom, 1972) as well as in the cholinergic boutons (Taxi, 1965). The large dense core vesicles have been observed in the nervous system of lower vertebrates and invertebrates (Bloom, 1972; Ovtscharoff, 1979a). Their size is in excess of 80 nm (Fig. 8). In the axonal endings of the CNS the large dense core vesicles usually do not exceed 5% of the vesicle population. Matsushima ef al. (1979) have found a marked diurnal rhythm in the number of large dense core vesicles in the sympathetic nerve fibers of the mouse pineal. They differ in their cytochemical characteristics from the small dense core vesicles as they are ZIO-negative (Kawana et al., 1969). Nevertheless there are data showing that in analogy to the small dense core vesicles the large dense core vesicles accumulate biogenic monoamines (Richards.and Tranzer, 1970). Probably these vesicles represent an additional storage site for these transmitter substances and participate in their metabolism. It must also be pointed out that the large dense core vesicles possess dopamine-Phydroxylase and chromogranin which confirms the above suggestion (Smith, 1971). On their membrane ATPase has been demonstrated cytochemically (Ovtscharoff and Manolov, 1977). But it appears that their dense core is not affected by drugs known to deplete amines (Bondareff, 1965). Their dense core can be visualized rather well with methods used for the demonstration of acidic and basic groups (Pfenninger er af., 1969; Ovtscharoff, 1975a). As digestion with proteolytic enzymes leads to negation of this center it is accepted that it has a proteinaceous nature (Pfenninger, 1973). Recently with immunocytochemical tecnniques in the large dense cote vesicles from different regions of the CNS and the PNS a number of peptides and hormones have been demonstrated: substance P, vasoactive intestinal polypeptide, somatostatin, enkephalin, thyrotropin-releasing hormone, angiotensin, vasopressin, and oxytocin (Larson, 1977; Pickel et al., 1979; Buijs and Swaab, 1979; Kemali and de Santis, 1980). In these cases it can be admitted that all this refers to peptidergic stmctures and synapses. It is difficult to decide whether these vesicles which have a dense core of a marked density and reticular appearance represent a different vesicle type (dense and semidense types of large dense core vesicles-Bloom and Aghajanian, 1968c) or they may be the result of a fixation failure. These two types of dense core vesicles have been observed in the synapses of the invertebrates and in the lower vertebrates (Ovtscharoff, 1979a). One is tempted to make a speculative assumption that there are two types of transmitters.

THE CHEMICAL SYNAPSES

FIG. 8. FIG. 9.

Large dense core vesicles. Frog optic tectum. ~80,000. Double synaptic vesicle. Frog optic tectum. X90.000.

26 1

262

STEPHAN MANOLOV AND WLADIMIR OVTSCHAROFF

C. COATEDAND OTHERVESICLES In the axonal endings of the central and the peripheral synapses there is a special type of vesicle called coated vesicles. Gray (1961) described them first and called them complex vesicles. The coat of these vesicles is formed of 12 pentagons and a varied number of hexagons with sides of an equal length i:Kanaseki and Kadota, 1969). Their coat has a proteinaceous nature and contains the protein clathrin (Jones, 1978). These vesicles attracted much attention after Kanaseki and Kadota (1969) obtained a fraction of coated vesicles and synaptic vesicles. Sometimes the central part of a vesicle is missing and only the coat remains (Jones, 1975). Gray (1972) described the so-called reticulosomes which are globular bodies similar to the coated vesicles but without a vesicle in inside. It is assumed that the coat of the coated vesicles participates in the formation of the dense projections (Gray and Willis, 1970; Jones, 1975). Their relation to synaptogenesis has been repeatedly pointed out (see, e.g., Altman, 1971; Ovtscharoff, 1979a) and it is accepted that they originate from the Golgi c,ysternae and the sacs of the smooth endoplasmic reticulum, leading their way to the periphery and carrying material for newly forming synapses. It is also thought that the coated vesicles participate in the endocytotic processes as well as in the calcium uptake in the axonal endings (Jones, 1978). Gray (1972) who first described the coated vesicles questioned their existence assuming that they might be artifactual structures. In axonal endings a certain type of vesicle called plasmalemmal vesicles was described (Palade and Bruns, 1968). These vesicles are best demonstrated with macromolecular markers-peroxidase and ferritin (Brightman, 1967; Zacks and Saito, 1969). In the neuromuscular junctions Korneliussen ( 1972a) observed vesicle profiles with a double membrane-double vesicles. We observed similar vesicles (Fig. 9) in the central synapses of frog (Manolov and Ovtscharoff, 1979a). D. ORIGINOF

THE

SYNAPTIC VESICLES

There are two basic groups of hypotheses about the origin of the synaptic vesicles. According to one of them the synaptic vesicles are produced in the neuronal perikarya and from there they are transported to the axonal endings. There are two structures with which the origin of the synaptic vesicles is associated: the Golgi apparatus and the endoplasmic reticulum (Palay, 1958). Aci:ually there is a cytochemical resemblance between the Golgi vesicles and the synaptic vesicles (Ovtscharoff, 1976b, 1979a). The other basic group of hypotheses assumes their local origin in the axonal endings. However, there is no agreement between the supporters of the local origin hypothesis. There is hardly a structure which had not been considered as the place where these vesicles might be

THE CHEMICAL SYNAPSES

263

formed. Pellegrino de Irraldi and De Robertis (1968) suggested that the vesicles are formed by budding from the microtubules. But this concept cannot account for the constituents and the structure of the vesicular membrane and the wall of the microtubules. Some authors consider that the synaptic vesicles are formed by budding from the smooth endoplasmic reticulum in the axonal ending (Korneliussen, 1972a; Pfenninger, 1973; Droz et al., 1975). In support of this hypothesis some cytochemical similarities can be pointed out between the two types of organelles discussed above. The endoplasmic sacs as well as the synaptic vesicles (except the large dense core vesicles) are 210-positive (Ovtscharoff, 1976b). The membrane of both structures shows TPPase and ATPase activity (Griffith and Bondareff, 1973; Ovtscharoff and Manolov, 1977). Examining the developing sympathetic fibers in rat pineal, Machado (1971) established that clear synaptic vesicles are formed from the cysternae of the smooth endoplasmic reticulum. Subsequently electron-dense material accumulates in these vesicles and they are transformed into dense core vesicles. Smith (1971) suggests that the large dense core vesicles discharge their content in the intercellular space whereupon coated vesicles start budding from them and then turn into clear synaptic vesicles. Some observations show that part of the vesicles are formed from the axonal membrane or via coated vesicles which are also formed from this membrane and, losing their coat, turn into synaptic vesicles with smooth walls (Kanaseki and Kadota, 1969; Gray and Willis, 1970). There are some biochemical data that the vesicle membrane and axon terminal membrane differ in their chemical constitution (Whittaker, 1966; Breckenridge et al., 1972) as the former contains more lipid (Morgan er a / . , 1973). Heuser and Reese (1973) proposed one of the most attractive hypotheses about the formation of the synaptic vesicles in the terminal boutons of the neuromuscularjunction and named it the recycling hypothesis. The advantage of this hypothesis is that it explains not only the origin of the vesicles but also their further fate. The hypothesis states that during transmitter release the synaptic vesicles fuse with the presynaptic membrane which dilates and gives rise to coated vesicles which lose their coat and fuse with a sac of the smooth reticulum which results in the formation of a cysternae, which in turn gives birth to the synaptic vesicles. All these processes are involved in the so-called exo-endocytotic cycle. However these authors do not exclude the possibility of an additional production of synaptic vesicles in the neuronal perikarya and their transport along the axons to the terminal boutons. Studying the ultrastructure of carp optic tectum we observed a special type of smooth endoplasmic reticulum (Fig. 10) in some neurons which was mainly composed of tubules which were 1 pm long and 50-80 nm wide (Fig. 11) and filled up with electron-dense material (Manolov and Ovtscharoff, 1979b). Some of these tubules were dichotomically branched. Large and dense core vesicles were observed in their vicinity (Manolov and Ovtscharoff, 1980). These structure were established in the neuronal perikarya, the dendrites, and the myelinated

264

STEPHAN MANOLOV AND WLADIMIR OVTSCHAROFF

FIG. 10. Endoplasmic reticulum with electron-dense contents. Carp optic tectum. X48,OOO.

FIG.I I . Long endoplasmic tubule filled with electron-dense material and dense core vesicles. Carp optic tectum. X92,oaO.

THE CHEMICAL SYNAPSES

265

and amyelinated axons (Fig. 12). Our results could have been treated as confirmation of the fact that both types of dense core vesicles are formed in the neuronal perikarya, the dendrites, and the axons. Besides there is more evidence which indicates a similar origin of the dense core synaptic vesicles: (1) Hokfelt (1968) demonstrated dense core vesicles in the perikarya in the cervical superior ganglion; and (2) the dense core vesicles accumulate above the site of a nerve lesion (Van Breemen et a l . , 1958). However our results do not exclude the possibility of local production (in the axonal endings) of the dense core vesicles. Actually the problem is not where these vesicles originate but rather where their production is greater, in the perikarya or in the boutons (Machado, 1971). It is probable that the synaptic vesicles are formed in the neuronal perikarya and subsequently transported along the axons and in the axonal endings where they acquire their enzymatic equipment and their chemical composition and are loaded with transmitters (Manolov and Ovtscharoff, 1979b). E. OTHERORGANELLES IN THE AXONAL ENDING 1. Smooth Endoplasmic Reticulum In the axonal endings single sacs of the smooth endoplasmic reticulum are frequently observed among the synaptic vesicles. Using impregnation with heavy metals Droz et al. (1975) established that in the terminal boutons the smooth reticulum forms a subsurface “primary net” which leads to a deeper construction of thinner sacs and tubules and a “secondary net,” and the synaptic vesicles are formed from the end of the sacs. On the membranes of the smooth endoplasmic reticulum TPPase (Griffith and Bondareff, 1973; Ovtscharoff and Manolov, 1977), ATPase (Ovtscharoff and Manolov, 1977; Manolov and Ovtscharoff, 1979a), and PDE (Florendo et al., 1971) have been detected.

2 . Microtubules and Neurofilaments The microtubules have a considerable length and a diameter of about 20-26 nm. They are long hollow cylinders with a 6-nm wall (Le Beux and Willemot, 1975). They are composed of protein subunits-tubulin (Matus et al., 1975; Jones, 1975). It is admitted that they participate in the axonal transport which can be inhibited after treatment with the plant alkaloids colchicine and vinblastine. Some authors consider that part of the microtubules are connected with dense projections and serve as a leading path for the synaptic vesicles (Gray, 1976). By ruthenium red staining we established deposition of electron-dense material on the external surface and in the inside of the microtubules. The neurofilaments are also built up on the participle of the hollow cylinder. Their diameter is about 7.5-10 nm (Pappas and Waxman, 1972; Le Beux and Willemot, 1975) and their wall is about 3 nm thick. Each neurofilament represents a spiral filament built up of the protein filarin. In the axonal endings actin-like microfilaments have been established. Their diameter is about 4-6 nm (Metuzals and Mushynski, 1974).

FIG. 12. Dendrite with electron dense endoplasmic reticulum and dense core vesicles. Carp optic tectum. ~ 3 4 , 0 0 0 . FIG. 13. Synaptic apparatus demonstrated after treatment with hydrochloric acid and subsequently staining with uranyl acetate. Rat cerebral cortex. X 189,000. FIG. 14. Synaptic apparatus after glutaraldehyde fixation and staining with uranyl acetate at pH 5. Lizard cerebral cortex. X 189,000. FIG. 15. Fusion of synaptic vesicles with presynaptic membrane. Axoaxonal synapse from frog optic tectum. X 135,000.

THE CHEMICAL SYNAPSES

267

F. CYTOPLASM OF THE AXONAL ENDING Gray ( I 972) assumes that the cytoplasm of the axonal endings is represented by a fine protein skeleton-the cytonet-which appears as polygonal cytonet or strandet cytonet depending on the applied technique. Lactate dehydrogenase is used as a marker of the soluble cytoplasmic subfraction of the synaptosomes (Whittaker, 1972). In the cytoplasm of the axonal endings ChAT was demonstrated cytochemically (Kasa et al., 1970; Kisa, 1970; Ovtscharoff, 1977). Knyihar and Csillik (1977) assume that acid phosphatase might be a cytoplasmic enzyme, as they have observed deposition of reaction product when visualizing this enzyme in rat substantia gelatinosa. G. PRESYNAPTIC DENSITY Electron-dense material was observed on the presynaptic membrane as early as the first electron microscopic investigations of the synapses (Palade and Palay, 1954; De Robertis and Bennett, 1954). Gray (1963) established structures called dense projections arranged in hexagons in material fixed with OsO, and impregnated with phosphotungstic acid. This author suggests that these formations may play the role of a guiding “device” for the synaptic vesicles to lead them to the presynaptic membrane. In a material treated in the routine electron microscopic method, dense projections are rarely observed and they are only partly presented. These structures are a consistent finding in material treated with E-PTA, B1, BIUL methods, staining with uranyl acetate at pH 5, or staining with tannic acid (Pfenninger, 1973; Ovtscharoff, 1975a,b; Wood and Cohen, 1980). Good results were obtained when demonstrating dense projections (Fig. 13) in glutaraldehydefixed tissue treated with hydrochloric or acetic acid and subsequently stained with uranyl acetate (Ovtscharoff, 1979a). Some authors established peripheral and central zones in the dense projections (Jones, 1978). Studying the organization of these presynaptic structures Akert et al. (1972) established that around each dense projection there are clear spaces in the form of hexagons which have approximately the size of the synaptic vesicles. They named the entire complex, located on the cytoplasmic surface of the presynaptic membrane of central synapses, a presynaptic vesicular grid. Akert et al. (1972) admit that this structure would be an ideal means for distributing the mediator in relation to the postsynaptic receptors. However the presynaptic vesicular grids are not constructed similarly in all cases. It might be assumed that this grid is a dynamic structure built up of contractile proteins and can temporarily “switch on” and “switch off“ the synapse. Another probable interpretation of the “more finely” and “more loosely” constructed vesicular grids is that the inhibitory synapses are equipped with smaller synaptic vesicles which allow them to invade the “finer” grid and reach the presynaptic membrane and the excitatory synapses; larger vesicles need a coarser grid (Akert et af., 1972). It is accepted that the

268

STEPHAN MANOLOV AND WLADIMIR OVTSCHAROFF

vesicular grid is disc shaped or slightly oval. Other forms are annulate or horseshoe-shaped, and in some cases they are huge and with bizarre forms (Vrensen er al., 1980). The occurrence of these complex contacts became more frequent with the increase of age and experience (Greenough et al., 1978). Applying various methods (digestion with proteolytic enzymes, aminoacetylation, carboxymethylation), it was established that the presynaptic density is composed of proteinaceous material (Bloom and Aghajanian, 1968b; Pfenninger, 1973; Ovtscharoff, 1975a,b). Many results indicate the presence: of a great deal of reactive basic and acidic groups. Pfenninger (1973) assumed that in the comparison to the synaptic cleft material the dense projections and the postsynaptic thickening contain the large part of the noncarboxyl acidic groups (sulfate and phosphate residues). After impregnation with E-PTA and staining with uranyl acetate at pH 5 of nonosmicated material from the CNS of lower vertebrates (carp, frog, lizard) we observed dense projections (Fig. 14) whose form and size do not differ from those of mammals (Manolov and Ovtscharoff, 1979a). Similar formations have been observed on the cytoplasmic surface of the presynaptic membrane in the nervous system of many invertebrates (Jones, 1978; Manolov and Ovtscharoff, 1979a). On the cytoplasmic surface of the presynaptic membrane of neuromuscular junctions electron-dense material around which the synaptic vesicles lie in groups is observed usually opposite the subjuncrional folds (Manolov, 1970). In fact this is an electron-dense bar on both sides of which there are vesicle attachment sites. H. TRANSMITTER RELEASE After Fatt and Katz (1952) demonstrated that ACh is released in discrete quanta1 packages and the ultrastructure of the synaptic vesicles was established (Palade and Palay, 1954; De Robertis and Bennett, 1954), Del Castilo and Katz ( 1956) proposed that these vesicles contain packets of transmitter and se:crete them in quanta. Thus the created vesicle hypothesis equalizes vesicles and quantum. Subsequently a considerable amount of evidence was accumulated in support of this hypothesis (Jones, 1975). The concept for the transmitters release from the synaptic vesicles by means of exocytosis (Fig. 15) brings up two basic problems-is the exocytosis reversible or is it irreversible. The former lies in the basis of the recycling hypothesis (Heuser and Reese, 1973). At reversible exocytosis the synaptic vesicle binds to the vesicle attachment site and releases the transmitter through the newly made canal in the synaptic cleft. The results of Berl er al. (1973) might be good evidence for this type of exocytosis. According to their concept the neurin of the presynaptic membrane (having actin-like groperties) and the stenin of the vesicular membrane (with myosin-like properties) bind together and bring about the release of the transmitter in the synaptic cleft. Studying the geometry of the vesicular grid in rat sensorimotor cortex we

THE CHEMICAL SYNAPSES

269

found that the ratio of the number of the dense projections to the number of the vesicle holes which theoretically must be equal to the VASs is approximately 1:3 (Ovtscharoff, 1979a). Vrenzen er af. (1980) find that this ratio fluctuates. But if we take into account that in our case we had a size grid which conformed to their medium size (19 dense projections: 54 VASs) we would have almost the same ratio, 1:3. We have calculated that the medium area of the vesicular grid is 0.086 pm2, 6 0 . 0 1 pm2, vesicular hole area 0.0017 pm2, 8-0.0002 pm2, and the area of the basis of the dense projections is 0.0045 pm2, 6 0 . 0 0 0 9 pm2. The area of the synaptic vesicle membrane of terminal boutons from the same region of the cortex is 0.005-0.008 pm2. From these data it is easy to calculate that in our case the presynaptic grids have approximately 20 to 30 holes. The biggest grids we found had 90-100 holes (Ovtscharoff, 1979a). In this case the fusion of 10-20 synaptic vesicles with the presynaptic membrane on the principle of irreversible exocytosis would lead to total substitution of the latter. But if we keep in mind the structure of the presynaptic grid, the fusion of only one vesicle with the presynaptic membrane would cause the exclusion of four holes, hence the simultaneous fusion of five to eight vesicles with the presynaptic membrane would block the entire presynaptic membrane. Now the question is what will happen to the VASs of the three neighboring holes which will be substituted with the synaptic vesicle membrane. Freeze-fracture data somewhat suggest that actually these are permanently differentiated structures in the CNS though during stimulation their number in the neuromuscular junctions increases (Heuser and Reese, 1977). Our studies show that the central synapses function under a considerably more economical regime compared to the 200-300 quanta (vesicles) necessary at the neuromuscular junctions. They also indicate that in the presynaptic membrane of rat neuromuscularjunction more than 3000 vesicles can be simultaneously incorporated (Ovtscharoff, 1979a). We also consider that the time for the realization of the irreversible exocytosis in the central synapses would be in accordance with the complex transformation of the presynaptic membrane under the grid (see also Jones, 1975). We think that the release of the transmitter in central synapses takes place on the principle of reversible exocytosis or the so-called transient vesicle-plasmalemma interaction. However there are data contradictory to the vesicle hypothesis (Dunant er al., 1971). They support the membrane control theory. This theory states that the transmitter which appears in the synaptic cleft comes from the cytoplasmic pool and the presynaptic membrane has a gating mechanism. However the evidence confirming this mechanism is rather scarce. Firm support for the membrane control theory is the fact that during stimulation the cytoplasmic ACh is released preferentially (Dunant and Israel, 1979). The following results also support the membrane control theory: during stimulation the endocytosis increases which is demonstrated with extracellular markers. However exocytosis is not absolutely proved (Dunant and Israel, 1979). A number of facts can be pointed out in

270

STEPHAN MANOLOV AND WLADIMIR OVTSCHAROFF

defense of the vesicle theory: the presynaptic vesicular grid, the presence of VASs, and the fluctuation of their number under experimental conditioris, the change in the number of vesicles during stimulation, etc. (Zimmerman, 1979). However it should be borne in mind that the change in number of the vesicles may not always correspond to the amount of transmitter released.

111. Synaptic Cleft

Conventional electron microscopy occasionally reveals filamentous structures in the synaptic cleft, which connect the presynaptic and the postsynaptic membranes (Van der Loos, 1963; De Robertis, 1964; Gray, 1966). Fine single or double lines running between the two membranes can also be observed (Pappas and Waxman, 1972). The distance between the presynaptic and the postsynaptic membranes at the interneuronal synapses is 15-30 nm. The synaptic cleft of the neuromuscular junctions differs from that of the interneuronal ones basically by its being wider and by the presence of basal lamina in it. The cytochemical studies showed that the synaptic cleft is far from being empty as the application of the conventional electron microscopic method might have suggested. In order to study the structure and the cytochemistry of the material in the synaptic cleft we used various methods (Ovtscharoff, 1979a). The visualization of this material depends on the method applied: the synaptic cleft appears empty after KMnO, or OsO, fixation. Fixation with the above fixatives followed by staining with uranyl acetate and lead citrate lead to the appearance of tiny granules and bars, which form single or double lines situated closer to the presynaptic membrani:. We assume that the appearance of this line or double-layered lamina is due to the visualization of acidic groups (by the uranyl cations) which are located 1.n this zone of the synaptic cleft (Ovtscharoff, 1975b). The method for demonstrating proteinaceous material in the clear synaptic vesicles (Ovtscharoff, 1978) as well as the reaction with colloidal thorium showed larger electron-dense granules situated more closely to the postsynaptic membrane (Ovtscharoff, 1979a). After impregnation with E-PTA and BI a relatively homogeneous material with nioderate electron density was established and sometimes in the middle of the synaptic cleft there is a thin electron translucent gap (Pfenninger, 1973). Staining with uranyl acetate at pH 5 and ruthenium red (Fig. 16), treatment of nonosmicated tissue with acid and staining with uranyl acetate and lead citrate (Ovtscharoff, 1975a,b,c, 1979a) lead to the appearance of a fairly homogeneous material with greater electron density in comparison to the former two methods. In these cases we could sometimes distinguish a thin electron lucent gap about the middle of the synaptic cleft a little closer to the presynaptic membrane. The material in the synaptic cleft has the highest electron density after BIUL. On the basis of these results the following conclusions about the distribution of chemical groups in the

FIG. 16. Electron-dense material in the synaptic cleft after ruthenium red staining. Rat cerebral cortex. ~ 3 2 4 , 0 0 0 . FIG. 17. Demonstration of ATPase in the synaptic cleft. Rat cerebral cortex. ~216,000. FIG. 18. Demonstration of PDE on the postsynaptic thickening. Rat cerebral cortex. x 162,000. FIG. 19. Localization of ATPase on the postsynaptic density. Rat cerebral cortex. x270,OOO.

272

STEPHAN MANOLOV AND WLADIMIR OVTSCHAROFF

synaptic cleft could be drawn: basic groups are distributed comparatively even (apparently less of them in the middle) on the synaptic cleft but they seem to be less in number compared to the acidic groups. The acidic groups appear to be more reactive in the medium zone of the synaptic cleft. The line they form is best visualized by the saccharated iron oxide method (Pappas and Purpura, 1966). The central zone is osmiophilic (Pfenninger, 1973). We are not certain as ito the identity of the osmiofile material because we have not observed it in material not additionally stained with uranyl acetate (Manolov and Ovtscharoff, 1979a). On the basis of these results three zones, or layers, can be distinguished in the synaptic cleft: ( I ) presynaptic, which is actually the cell coat of the presyriaptic membrane; (2) postsynaptic, the cell coat of the postsynaptic membrane; arid (3) the middle zone or layer, in which adhesion between pre- and postsynaptic membranes occurs. According to Pfenninger (1973) the mechanism of adhesion is of a polyionic nature rather than based on coordination complexes involving Ca2 and carboxyl groups. As to the chemical composition of the first two zones they should be considered in connection with the structure and the composition of the pre- and postsynaptic membranes and their external coat. The structure and the composition of these membranes play a role in the intercellular recognition during synaptogenesis. In such a way must be treated the establishment of AChE (Manolov, 1976; Westrum and Broderson, 1976; Manolov and Davidoff, 1976) and ATPase (Ovtscharoff and Manolov, 1977) in the synaptic cleft (Fig. 17). Probably this is due to diffusion of the reaction product as these enzymes are actually components of the pre- and postsynaptic membranes. Some authors, however, do not reject absolutely the possibility of the presence of AChE in the synaptic cleft (Manolov and Davidoff, 1976). With cytochemical techniques carbohydrates have been detected in the synaptic cleft: aqueous phosphotungstic acid (Pease, 1966), periodic acid-silver methenamine (Rambourg and Leblond, 1967), phosphotungstic acid-chromic acid (Rambourg, 19691, ruthenium red (Bondareff, 1967; Ovtscharoff, 1975a), and alcian blue (Castejon and Castejh, 1976; Ovtscharoff, 1979a). The presence of neuraminic acid in this part of the synaptic apparatus is shown with neuraminidase digestion (Pfenninger, 1973; Ovtscharoff, 1975b). Our studies with P-glucuronidase and sulfatase have established the presence of D-glucuronides and esters of sulfuric acid in the synaptic cleft (Ovtscharoff, 1979a). The presence of proteins and glycoproteins in the cleft has been confirmed over and over again (Bloom and Aghajanian, 1968b; Barrantes and Lund, 1970; Pfenninger, 1973; Ovtscharoff, 1975a,b). +

1V. Postsynaptic Part The structure, the chemical composition, and the enzyme equipment of the postsynaptic membrane, the postsynaptic thickening, and the subsynaptic

THE CHEMICAL SYNAPSES

273

organelles are of great importance for the recognition of nerve impulses from the presynaptic part as well as for the integration of synaptic information. A. POSTSYNAPTICMEMBRANE The receptors located on the postsynaptic membrane are of decisive importance to the execution of the synaptic transmission. They interact with transmitter released in the synaptic cleft which results in depolarization or hyperpolarization of the postsynaptic membrane. The receptors are integral membrane proteins or glycoproteins which can be separated from the membrane only by detergent extraction. When this membrane is stained negatively, toroidal particles of about 6-9 nm, built up of five or six subunits of about 2 4 nm each arranged in a circle around a central pit with a diameter of 1.5-2 nm, can be observed under the electron microscope (Fambrough, 1979). On freeze fracture replicas 10-nm particles in clusters, covering an area of the postsynaptic membrane which has various forms-macular, annular, or anastomotic, are observed (Korte and Rosenbluth, 1980). On the postsynaptic membrane of frog neuromuscular junction the density of the ACh receptors is about 26,0005 6000 pm2 (MathewsBellinger and Salpeter, 1979). Salpeter and Eldefrawi (1973) assume that 25% of the surface of the postsynaptic membrane is covered with ACh receptors which are actually glycoproteins. The presence of a-methyl-D-mannopyranoside residues, N-acetyl-D-glucosamine, D-galactose, and N-acetyl-D-galactosamine in them can be demonstrated with some lectins (Kelly et af., 1976). The carbohydrate moieties apparently are unique entities determining cell recognition during synaptogenesis. This function can also be played by glycosyltransferases on the neuronal membrane bound to their naturally occurring substrate on the membrane of the axonal ending (Jacob and Lentz, 1979). It must be taken into account that the receptors on the postsynaptic membrane are mobile, and their number and arrangement are affected during ontogenesis, after denervation, during blockage of synaptic transmission, during stimulation, etc. (Edwards, 1979; Fambrough, 1979). The number of the receptors out of the synaptic zone strongly diminishes. After establishing that a component (a-bungarotoxin) of the poison of the snake Bungarus mufticinctus combines irreversibly with ACh receptors (Chang and Lee, 1963), a-bungarotoxin is successfully applied to study the localization of these receptors (De Robertis, 1975). Applying '311-labeledabungarotoxin and I 251-labeleda-bungarotoxin it was found that this component is deposited on the postsynaptic membrane (Lee et al., 1967). Peroxidase-labeled a-bungarotoxin has also been used for this purpose in cytochemical studies (Daniels and Vogel, 1975). ACh reacts with two types of receptors: nicotinic, which can be stimulated with nicotine and blocked with a-tubocurarine, and muscarinic, which are activated with muscarine and can be blocked by atropine. According to the principle of Dale every synapse discharges one type of transmit-

274

STEPHAN MANOLOV AND WLADlMlR OVTSCHAROFF

ter. Depending on the mediator which each neuron releases in all endings of its axon, it can be excitatory or inhibitory. hence the presynaptic part plays a decisive role. Ahlquist (1948) launched the hypothesis about the existence of (Y and (3 receptors on the cell membrane, the former being accepted as excitatory and the latter as inhibitory. In this case there is a possibility for the same transmitter (adrenaline or noradrenaline) to exert excitatory or inhibitory action depending on the receptor equipment of the postsynaptic membrane. Serotoninergic receptors, glutamate receptors, GABA receptors, histamine receptors, and glycinic receptors have also been identified (De Robertis, 1975; Petkov, 1977). It was suggested that *AChE and ACh receptors are identical (Nachmansohn, 1959) or that they are parts of the same molecule (Changeux et al., 1969). Autoradiographically it was shown that these are two different macromolecules. Many authors established the presence AChE on the postsynaptic membrane cytochemically (Manolov, 1976; Manolov and Davidoff, 19761. Immunocytochemically Tsuji (1 977) localized this enzyme on the postsynaptic membrane in the electric organ of Electrophorus electricus and T. marmorata. By biochemical methods it was established that AChE is a membrane-bound enzyme (Hoskin, 1972). On the postsynaptic membrane ATPase activity which can be inhibited by ouabain and NiCI, was localized cytochemically (Ovtscharoff and Manolov, 1977). By means of a cytochemical method Sugimura and Mizutani ( 1979) have established alkaline phosphatase on the same membrane. They assume that this enzyme may be of importance in synaptic transmission. AC activity on the postsynaptic membrane was observed by Markov and Dimova (1975) and Hervonen and Rechardt (1976). Lemay and Jarett (1975) point out that the basic cytochemical method cannot be used for visualization of the enzyme activity as it inhibits this activity in the applied lead concentration. We have established deposition of the reaction product on the postsynaptic membrane in rat frontal cortex when demonstrating the enzyme according to Reik et al. (1970) with a reduced concentration of Pb(NO:,)2 (2 mM). We observed similar localization in the rat striatum after preliminary preincubation with 3 m M adenosine and 3 mM dopamine (Manolov and Ovtscharoff, 1979a). This finding is in agreement with the hypothesis which states that AC is found on the postsynaptic membrane where it participates in the formation of CAMP(Greengard et al., 1972). It is assumed that in some synapses of the CNS and in the autonomic ganglia this membrane has receptors which interact with biogenic amines, histamine, adenosine, and other adenine nucleotides (Manolov and Ovtscharoff, 1979a). It is possible that in these cases the interaction between the regulatory agent with the receptors activates the receptors bound to AC allosterically (Cuatrecasas, 1975). The product of the interaction between AC and AMP is considered as a second messenger (Sutherland et al.. 1968) in central and in peripheral synapses (Greengard and Kebabian, 1974). There are data which show that the action of some of the neurotransmitters is

THE CHEMICAL SYNAPSES

275

affected through cAMP and others by cGMP (Greengard, 1975). For example, it is assumed that cAMP is produced as a result of the interaction between noradrenaline and the P-adrenergic receptors and acts as a second messenger (Robison et al., 1971) while cGMP is the second messenger produced by the action of ACh upon muscarinic cholinergic receptors (Horn and McAfee, 1977). B. POSTSYNAPTIC THICKENING The density on the cytoplasmic surface of the postsynaptic membrane was already established by the first authors who studied the ultrastructure of synapses (De Robertis, 1956, 1958; Palay, 1956, 1958). This structure has had various names: postsynaptic organelle (Van der Loos, 1963), postsynaptic web (De Robertis, 1964), and postsynaptic differentiation (Sotelo, 1968), but it is generally known as postsynaptic density or postsynaptic thickening. Usually the deep seated part of the postsynaptic density has uneven contours and single or numerous filaments invade the cytoplasm to a different extent. Gray (1959) accepted this thickening as a basis critenum in distinguishing two types of synapses. Colonnier (1968) also described two types synapses: type I-“asymmetric”with thick thickening, and type 11-‘ ‘symmetrical”-with thinner postsynaptic density. In fact the chemical synapse is functionally and structurally asymmetrical. Nevertheless these terms are used by some authors when describing the synaptic organization of some areas in the CNS (Heuser and Reese, 1977). The angle of the section might be important to the thickness of this density. The postsynaptic density is round or elongated and its diameter varies between 0.2 and 2 p,m. It has been established that some large postsynaptic densities resemble a disc with a large perforation or hole at its center (Peters and Kaisermann-Abramof, 1969; Cohen and Siekevitz, 1978). Bearing in mind the shape of the active zone of the presynaptic part there might be analogous complex forms (Vrensen et al., 1980) on the postsynaptic thickening. Considering this preposition and the results of Cohen and Siekevitz (1978) it might be assumed that in some cases synapses with two or more active zones observed on electronograms actually have one active zone but of a complex form. On the basis of the methods applied for the demonstration of the synaptic apparatus we assume that the postsynaptic density is built of two or sometimes three zones which we term ( I ) the dense zone, lying right under the postsynaptic membrane; (2) the reticular zone; and (3) the deep postsynaptic net (Ovtscharoff, 1979a). The first zone proved to be most resistent one to digestion with proteolytic enzymes. Blomberg et al. (1977) have observed that isolated postsynaptic densities of dog cerebral cortex contain two types of filaments of different thickness (3-5 nm filaments and other filaments of 10 nm) built of neurofilament-like protein. These authors suggest that these filaments can interact through calcium cations. Cytochemical methods proved that the postsynaptic density has

276

STEPHAN MANOLOV AND WLADIMIR OVTSCHAROFF

a proteinaceous nature (Bloom and Aghajanian, 1968b; Pfenninger, 1973). It is possible that the third zone of this thickening, the deep postsynaptic net, which sometimes includes subsynaptic differentiations, can actually be precipitated hyaloplasmic proteins (Gray, 1975). On the basis of our data we admitted that the postsynaptic density has more phosphate groups in comparison to the other components of the synaptic apparatus (the synaptic cleft material and the dense projections). Le Beux and Willemot (1975) have found actin-like filaments in this thickening. By immunocytochemical methods the presence of S- 100 protein (Haglid et af., 1976) and tubulin (Matus et al.. 1975) was demonstrated in the postsynaptic density. With cytochemical techniques PDE (Fig. 18) was localized in the postsynaptic thickening (Florendo et af., 1971; Adinolfi and Schmidt, 1974; Sugimura and Mizutani, 1978; Manolov and Ovtscharoff, 1979a). This enzyme participates in the regulation of the level of CAMP. Studying the fine structural localization of ATPase (Fig. 19) in rat cerebral cortex we established enzyme activity in the postsynaptic density (Manolov and Ovtscharoff, 1979a). We succeeded in inhibiting this enzyme activity only after the addition of 25 mM NiCl,. Nothing definite can be said about this localization of the enzyme. ATPase may regulate the level of ATP in the postsynaptic density reaching the postsynaptic membrane to be turned from AC into CAMP. STRUCTURES C. SUBSYNAPTIC The subsynaptic structures are found in close proximity to the postsynaptic membrane or thickening. In the dendritic spine of rat occipital cortex Gray ( 1959) established a characteristic structure which he named the spine apparatus. It is built up of two or more sacs among which there are plaques of dense material. This apparatus was observed in the dendritic trunk (Gray and Guillery, 1963) as well as in the neuronal perikarya next to the axosomatic synapses (Pappas and Waxman, 1972). Nothing is known about the functional significance of the spine apparatus. Hamlyn ( 1962) assumed that it might be connected with the processes of learning and memory. However this supposition needs considerable experimental support. Rosenbluth ( 1962) has observed flattened sacs with slightly dilated endings lying close to the postsynaptic membrane. We designated them as subsurface cysternae. They are 1 4 km long and 5-12 nm wide. Sometimes ribosomes are attached to the internal side of their membrane. The distance between the external membrane of the cysternal and the neuronal membrane is about 10-15 nm. The endings of the subsurface cysternae are sometimes linked with the smooth or the rough endoplasmic reticulum. It is assumed that these subsurface cysternae may be an ion reservoir and in this way change the electric properties of the neuronal membrane (Rosenbluth, 1962). Taxi ( I 961) described a characteristic subsynaptic structure in the sympathetic

THE CHEMICAL SYNAPSES

277

ganglia of amphibia. This subsynaptic dense layer is disc-like or oval 25-50 nm thick and with a diameter of not more than 0.5 pm. This author described the dense layers as consisting of two parts, the deeper set layer which can also be presented by two parts (Taxi, 1967). Milhaud and Pappas (1966) paid attention to the so-called subjunctional dense bodies and established their hexagonal or rather triangular arrangement. They are rounded and have a diameter of about 20-45 nm. Their number varies from 2 to 12. Tracing the changes in the anterior horn axodendritic synapses of rats submitted to long-term locomotor activity we established that 8% of the synapses of the control animals contain subjunctional bodies. These structures, however, have not been observed in immobilized as well as in trained rats (Cheresharov et al., 1978). An analogy exists between the subjunctional dense bodies and the subsynaptic dense layer (Manolov, 1967a; Pappas and Waxman, 1972). Bodian (1966) and Manolov (1967b) have observed a formation of coated vesicles on the postsynaptic membrane. Waxman and Pappas (1969) admit that these coated vesicles are directly connected to the synaptic transmission.

V. Concluding Comments The chemical synapses possess characteristic structural organization, chemical contents, enzyme equipment, and a certain plasticity. The most typical presynaptic structures-the synaptic vesicles-are highly organized organelles. The question remains as to whether the size, the form, and the number of the synaptic vesicles are somehow related to the kind of mediator or transmitter contents in them. In any case these three criteria are changed at definite functional conditions as well as during synaptogenesis. The other organelles and inclusions in the axonal endings probably have no specific role in the synaptic transmission since they are not found in all cases in the presynaptic element. An exception in this sense may be the sacs of the smooth endoplasmic reticulum. Both cell membranes, which realize the direct conduction of bioelectrical information, the presynaptic and postsynaptic membranes, are highly organized membrane systems, with corresponding structure, cytochemical characteristics, and enzyme equipment. The definite arrangement of the acidic and basic groups on the presynaptic and postsynaptic elements probably plays a determining role in the formation of the synaptic contacts during synaptogenesis. The goals of future studies in synaptology will be to examine in detail the chemical contents of the synaptic structures, to establish new transmitter substances and their metabolism, to elucidate further the molecular mechanism of the synaptic transmission, and to study the presynaptic and postsynaptic receptors.

278

STEPHAN MANOLOV AND WLADIMIR OVTSCHAROFF REFERENCES

Adinolfi. A. M., and Schmidt. S . Y. (1974). Brain Res. 76. 21. Ahlquist, R. F. (1948). Am. J . Physiol. 153, 586. Akert, K., and Sandri, C. (1968). Brain Res. 7, 286. Akert, K., Pfenninger, K., Sandri, C., and Moor, H. (1972). In “Structure and Function of Synapses” (E. D. Pappas and D. P. Purpura, eds.), pp. 67-86. Raven, New York. Akert, K., Sandri, C., CuCnod, M., and Moor, H. (1977). Neurosci. Lerr. 5 , 253. Altman, J. (1971). Brain Res. 30, 311. Bahr, G. F. (1954). Exp. Cell Res. 7, 457. Barchi, R. L., and Braun, P. E. (1972). J. Neurochem. 19, 1039. Barrantes, F. J., and Lund, G. G. (1970). Brain Res. 23, 305. Benshalom, G., and Flock, A. (1977). Bruin Res. 121. 173. Bed, S., Puszkin, S., and Nicklas, W. J. (1973). Science 179, 441. Bemstein, H.-G., Weib, I., and Luppa, H. (1978). Hisrochemisrry 55, 261. Birks, R. J. (1971). Proc. Physiol. SOC. 216, 26. Blomberg, F., Cohen, R. S., and Siekevitz, Ph. (1977). J. Cell Biol. 74, 204. Bloom, F. E.(1972). In “Neurotransmitters, Res. Publ. Assoc. for Nervous and Mental Diseases” (I. J. Kopin, ed.), pp. 25-57. Williams & Wilkins, Baltimore, Maryland. Bloom, F. E. (1975). In “Pre- and Postsynaptic Receptors” (E. Usdin and W. E. Bunney, eds.), pp. 67-87. Dekker, New York. Bloom, F. E., and Aghajanian, G. K. (1968a). Experienria 24, 1225. Bloom, F. E., and Aghajanian, G. K. (1968b). J . Ulrrasrrucr. Res. 22, 361. Bloom, F. E., and Aghajanian, G. K. (1968~).J. Phrmacol. Exp. Ther. 159, 261. Bloom, F. E.,and Barmett, R. J. (1967). Ann. N . Y. Acad. Sci. 144, 626. Bodian, D. (1966). Bull, Johns Hopkins Hosp. 119, 129. Bondareff, W. (1965). Z . Zellforsch. 67, 211. Bondareff, W. (1966). Exp. Neurol. 16, 131. Bondareff, W. (1967). Anar. Rec. 157, 527. Bondareff, W., and Gordon, B. (1966). J . Pharmacol. Exp. Ther. 153, 42. Boyne, A. F., Bohan, T. P., and Williams, T. H. (1974). 1. Cell Biol. 63, 780. Breckenridge, W. D., Combos, G., and Morgan, J. G. (1972). Biochim. Biophys. Acra 266, 695. Breer. H., Moms, S. J., and Whittaker, V. P. (1977). Eur. J . Biochem. 80, 313. Breer, H., Morris, S. J., and Whittaker, V. P. (1978). Eur. J. Biochem. 87, 453. Brightman, M. W. (1967). In “Brain Barrier Systems” (A. Lajtha and D. H. Ford, eds.), pp. 19-37. Elsevier, Amsterdam. Buijs, R. M., and Swaab, D. F. (1979). Cell Tissue Res. 204, 355. Carlsson, A. (1975). In “Pre- and Postsynaptic Receptors” (E. Usdin and W. E. Bunney, eds.), pp. 49-65. Dekker, New York. Castejbn, H. V., and Castejbn, 0. J. (1976). Acra Hisrochem. 55, 300. Castel, M., and Dellmann, H.-D. (1980). Cell Tissue Res. 210. 205. Ceccarelli, B., and Pensa, P. (1968). Sperimentale 118, 197. Chang, C. C., and Lee, C. Y. (1963). Arch. Inr. Pharmacodyn. ?‘her. 144, 241. Changeux, J. P., Podleski, T. R., and Meunier, J. C. (1969). J . Gen. Physiol. 54, 225. Chen, J.. and Lee, C. Y. (1970). Virchows Arch. Abr. B, Zellparhol. 6 , 3 18. Cheresharov, L., Ovtscharoff, W., and Manolov, S . (1978). J. Neural Trunsm. 42, 9. Chouchkov, C., and Ovtscharoff, W. (1980). Acra Hisrochem. 66, 160. Cohen, R., and Siekevitz, P. (1978). J . Cell Biol. 78, 36. Colonnier, M. (1968). Brain Res. 9, 268. Couteaux, R., and Pkcot-Dechavassine, M. (1973). Arch. lral. B i d . 111, 231.

THE CHEMICAL SYNAPSES

279

Cragg. B. G. (1969). Brain Res. 13, 53. Csillik, B.. Knyihar, E., Laszlo, J., and Boncz, J. (1974). Brain Res. 70, 179. Cuatrecasas, P. (1975). In “Advances in Cyclic Nucleotide Research” (G. J. Drummond, P. Greengard. and G. A. Robison, eds.), Vol. 5 , pp. 79-104. Raven, New York. Cuenod, M., Sandri, C., and Akert, K. (1970). J. Cell Sci. 6, 6 0 5 . Daniel, A.. and Guth, L. (1975). Exp. Neurol. 47, 181. Daniels, M. P.,and Vogel, 2. (1975). Narure (London) 254, 339. Del Castillo, J . , and Katz, B. (1954). J . Physiol. (London) 124, 553. Del Castillo, J . . and Katz, B. (1956). Prog. Biophys. 6, 121. De Robertis, E. (1956). J. Biophys. Biochem. Cyrol. 2, 503. De Robertis, E. D. P. (1958). Exp. Cell Res. 5. 347. De Robertis, E. D. P. (1964). “Histophysiology of Synapses and Neurosecretion.” Pergamon, Oxford. De Robertis, E. (1975). “Synaptic Receptors. Isolation and Molecular Biology.” Dekker, New York. De Robertis. E . , and Bennett, H. S. (1954). Fed. Proc. Fed. Am. Soc. Exp. Biol. 13. 35. De Robertis, E., and Pellegrino de Iraldi, A. (1961). J. Biophys. Biochem. Cytol. 10, 361. Dowdall, M. J . , and Zimmermann, H. (1974). Brain Res. 71, 160. Dowdall, M. J . , Boyne, A. F., and Whittaker, V. P. (1974). Biochem. J . 140, I . Droz, B., Rambourg, A,, and Koenig, H. L. (1975). Brain Res. 93, 1. Dunant, Y., and Israel, M. (1979). T I “ . 130. Dunant, Y., Gautron, J., Israel, M., Lesbats, B., and Manaranche, R. (1971). Experientia 27, 4. Eagles, P. A. M.,Johnson, L. N., Van Horn, C., and Bullivant, S. (1977). Neuroscience 2, 153. Edwards, C. (1979). Neuroscience 4, 565. Falck, B., Hillarp, N. A., Thieme, G., and Torp, A. (1962). J. Hisrochem. Cyrochem. 10, 348. Fambrough, D. M. (1979). Physiol. Rev. 59, 165. Fatt, P., and Katz, B. (1952). J. Phvsiol. (London) 117, 109. Feigenson, M. E., and Bannett, R. J . (1977). Brain Res. 119, 155. Florendo, N. T., Bannett, R. J., and Greengard, P. (1971). Science 173, 745. Fonnum. F. (1966). Biochem. Pharmacol. 15, 1641. Fonnum, F. (1973). Brain Res. 62, 497. Fox, J . M . , and Duppel, W. (1975). Brain Res. 89, 287. Gilloteaux, J . . and Naud, J. (1979). Histochemistry 63, 227. Goldberg, A. L., and Singer, J . J . (1969). Proc. Narl. Acad. Sci. U.S.A. 64, 134. Goldstein. M . , Ebstein, B., Bronaugh, R. L., and Roberge. C. (1975). In “Chemical Tools in Catecholamine Research” (0. Alrngren, A. Carlson, and J. Engel, eds.), Vol. 11, pp. 257-264. North Holland Publ., Amsterdam. Gray, E. G. (1959). J . Anar. 93. 420. Gray, E. G. (1961). J. Anat. 95, 345. Gray, E. G . (1963). J . Anar. 97, 101. Gray, E. G. (1966). Int. Rev. Gen. Exp. Zool. 2. 139. Gray, E. G. (1972). J. Neurocytol. 1, 363. Gray, E. G. (1975). J . Neurocyrol. 4, 315. Gray, E. G . (1976). Prog. Brain Res. 45, 207. Gray, E. G., and Guillery, R. W. (1973). J. Anar. 97, 389. Gray, E. G., and Willis, R. A. (1970). Brain Res. 24, 149. Greengard, P. (1975). In ”Advances in Cyclic Nucleotide Research” (G. 1. Drummond, P. Greengard, and G. A. Robison, eds.), pp. 585-601. Raven, New York. Greengard. P., and Kebabian, J. W. (1974). Fed. Proc. Fed. Am. Soc. Exp. Biol. 33, 1059.

280

STEPHAN MANOLOV AND WLADIMIR OVTSCHAROFF

Greengard, P., Mc Afee, D. A.. and Kebabian, J. W.(1972). I n “Advances in Cyclic Nucleotide Research” (P. Greengard and G. A. Robison, eds.), Vol. I, pp. 337-355. Raven, New York. Greenough. W. T., West, R. W., and De Voogd, T. J. (1978). Science 202, 1096. Griffth, D. L., and Bondareff, W. (1973). Am. J. Anar. 136, 549. Grillo, M. A., and Palay, S. L. (1962). Proc. Inr. Congr. Electron Microsc., Srh, pp. U-I. Grinnell, A. D. (1966). J. Physiol. (London) 182, 612. Hackett, J. T. (1976). Brain Res. 114, 35. Haglid, K. G . , Hamberger, A., Hansson, H. A,, Huden, H., Persson, L., and Ronnback, L. (1976). J. Neurosci. Res. 2, 175. Hall, Z. W. (1972). In “Structure and Function of Synapses” ( G . D. Pappas and D. P. Purpura, eds.), pp. 161-171. Raven, New York. Hamlyn, L. M. (1962). J . Anar. 96, 112. Hansen, J. T., and Smith, N. K. R. (1979). Cell Tissue Res. 199, 145. Hawthorne, J. N., and Pickard, M. R. (1979). J . Neurochem. 32, 5 . Hayat, M. A. (1970). “Principles and Techniques of Electron Microscopy,” Vol. 1. Van NostrandReinhold, Princeton, New Jersey. Hernandez-Nicaise, M. L. (1973). J . Neurocyrol. 2, 249. Hervonen, H., and Rechardt, L. (1976). Histochemistry 48, 43. Hess, A. (1977). Brain Res. 138, 555. Heuser, I . E. (1971). Proc. Annu. Meer. Am. SOC. Cell Eiol., Ilrh, p. 124. Heuser, J. E. (1976). I n “Motor Innervation of Muscle” (S. Thesleff, ed.), pp. 51-1 15. Academic Press, New York. Heuser, J. E., and Reese, T. S. (1973). J . Cell Biol. 57, 315. Heuser, J. E., and Reese, T. S. (1977). I n “Handbook of Physiology” (J. M. Brookhart and V. B. Mountcastle, eds.), pp. 261-294. American Physiological Society, Bethesda. Maryland. Heuser, J. E., Reese, T. S., and Landis, D. M. D. (1974). J. Neurocyrol. 3, 109. Hillman, D. E., and Llinas, R. (1974). J . CeNEiol. 61, 146. Hokfelt, T. (1966). Experientia 22, 56. Hokfelt, T. (1967). Acra Physiol. Scand. 69, 125. Hokfelt, T. (1968). 2. Zellforsch. 91, 1. Hokfelt, T. (1971). Prog. Brain Res. 34, 213. Horn, J. P., and McAfee, D. A. (1977). J. Physiol. (LondonJ 269, 753. Hoskin, F. C. G. (1972). I n “Basic Neumhemistry” (R. W. Albers, G. J. Siegel, R. Katzman, and B. W. Agranoff, eds.), pp. 105-130. Little, Brown, Boston, Massachusetts. Hubbard, I. I. (1971). Ann. N.Y. Acad. Sci. 182, 131. Hubbard, J. I., and Kwanbunbumpen, S. (1968). J . Physiol. (London) 194, 407. Iijima, K. (1975). Hisrochemisfry 46, 61. Iversen, L. L. (1967). “The Uptake and Storage of Noradrenaline in Sympathetic Nerves.” Cambridge Univer. Press, London and New York. Jacob, M., and Lentz, T. L. (1979). J. Cell Biol. 82, 195. Johnson, G. A., Boukma, S. J., Lahti, R. A., and Mathews, J . (1973). J . Neurochem. 20, 1387. Jones, D. G. (1975). “Synapses and Synaptosomes. Morphological Aspects.” Chapman & Hall, London. Jones, D. G. (1978). Adv. Anar. Embryol. Cell Eiol. 55. Jones, D. G., and Bradford, H. F. (1971). Brain Res. 28, 491. Jones, D. G., and Revell, E. (1970). Z. Zellforsch. 111, 195. Jones, S. F., and Kwanbunbumpen, S. (1968). Life Sci. 7, 1251. Jones, S. F., and Kwanbunbumpen, S. (1970). J . Physiol. (London) 207, 31. Kanaseki, T., and Kadota, K. (1969). J . Cell Eiol. 42, 202. Kbsa, P. (1971). Prog. Brain Res. 34, 337.

THE CHEMICAL SYNAPSES

28 I

Kasa, P., Mann, S. P., and Hebb, C. (1970). Nature (London) 226, 812. Katz, B., and Miledi, R. (1965). Proc. R. Soc. Ser. B. 161, 491. Katz, B.. and Miledi, R. (1967). J. Physiol. (London) 189, 535. Kawana, E.. Akert. K.. and Sandri, C. (1969). Bruin Res. 16, 325. Kawana, E., Akert. K., and Bruppacher, H. (1971). J . Comp. Neurol. 142, 297. Kelly, P., Cotman, C. W., Gentry, Ch., and Nicolson, G. L. (1976). J. Cell Biol. 71, 487. Kemali, M., and DeSantis, A. (1980). Exp. Bruin Res. 40, 119. Knyihiir, E., and Csillik, B. (1977). J. Neural Trunsm. 40,227. Knyihir, E., Liiszl6, J., and Csillik, B. (1973). Neurobiology 3, 327. Kokko, A., and Barmett, R. J. (1971). f r o g . Bruin Res. 34, 319. Konig, N., Roch, G., and Marty, R. (1975). Anut. Embryol. 148, 73. Korneliussen, H.(1972a). Z. Zellforsch. 130, 28. Korneliussen, H. (1972b). J. Neurocytol. 1, 279. Korneliussen, H., Barstad, J. A., and Lilleheil, G. (1972). Experientiu 28, 1055. Korte, G. E., and Rosenbluth, J. (1980). J. Comp. Neurol. 193, 689. Landis, D., Reese, T., and Raviola, E. (1974). J. Comp. Neurol. 155, 67. Landon, D. N., Westgaard, R. H., Macdermot, J., and Thompson, E. J. (1980). Bruin Res. 202, 1. Larramendi, L. M. H., Fickenscher, L., and Lemkey-Johnston. N. (1967). Science 156, 967. Larsson, L.-L. (1977). Histochemistry 54, 173. Le Beux, Y. J., and Willemot, J. (1975). Cell Tissue Res. 160, 37. Lee, C. Y., Tseng, L. F., and Chin, T. H. (1967). Nature (London) 215, 1177. Lemay, A., and Jarett, L. (1975). J . Cell Biol. 65, 39. Lund, R. D., and Westrum, L. E. (1966). J. Physiol. (London) 185, 7. Machado, A. B. M. (1971). f r o g . BruinRes. 34, 171. Mclntosh, F. C. (1963). Can. J . Biochem. Physiol. 41, 2555. McLaughlin, B. J., Wood, J. G., Saito, K., Borber, R., Vaughn, J. E., Roberts, E., and Jong-Yen, Wa. (1974). Bruin Res. 76, 377. Maillet, M. (1962). Trub. Inst. Cujul Invest. Biol. 54, I . Malthe-Sorenssen, D., and Fonnum, F. (1972). Biochem. J. 127, 229. Manolov, S. (1967a). C . R. Acud. Bulg. Sci. 20, 617. Manolov, S. (1967b). Bull. Assoc. Anut. 138, 838. Manolov, S. (1967~).C . R. Acad. Bulg. Sci. 20, 493. Manolov, S. (1970). Bull. Assoc. Anut. 149, 848. Manolov, S. (1976). “Morphology of the Neuro-Muscular Synapses. Norm, Degeneration and Regeneration.” Publishing House of the Bulgarian Academy of Sciences, Sofia. Manolov, S . , and Davidoff, M. (1976). “Histochemistry of Cholinesterasesin the Nervous Tissue.” Publishing House of the Bulgarian Academy of Sciences, Sofia. Manolov, S., and Ovtscharoff. W. (1979a). “Ultrastructure of the Synapses in the Central Nervous System.” Publishing House of the Bulgarian Academy of Sciences, Sofia. Manolov, S . , and Ovtscharoff, W. (1979b). In “Contemporary Problems of Neuromorphology” ( G . Galabov and S. Manolov, eds.), Vol. 4-5, pp. 7-25. Publishing House of the Bulgarian Academy of Sciences, Sofia. Manolov, S . , and Ovtscharoff, W. (1980). In “Electron Microscopy” (P. Brederoo and W. de Priester, eds.), Vol. 2. pp. 96-97. Published by the 7th European Congress on Electron Microscopy Foundation, Leiden. Marchbanks, R. M. (1969). In “Cellular Dymanics of the Neuron” (S. H. Barondes, ed.), J. S. C. B. Symposia Vol. 8. pp. 115-135. Academic Press, New York. Marchbanks, R. M., and Israel, M. (1971). J. Neurochem. 18, 439. Marchbanks, R. M., and Israel, M. (1972). Biochem. J. 129, 1049. Markov, D. V., and Dimova, R. N. (1975). C. R. Acud. Bulg. Sci. 28, 565.

282

STEPHAN MANOLOV AND WLADIMIR OVTSCHAROFF

Matsushima, S., Morisawa, Y.,and Mukai, S. (1979). J. Neural Transm. 45, 63. Matthews-Bellinger, J., and Salpeter, M. M. (1979). J . Physiol. (London) 279, 197. Matus, A. J., and Dennison, M. E. (1971). Brain Res. 32, 195. Matus, A. I., Walters, B. B., and Mughal, S . (1975). J. Neurocytol. 4, 733. Metuzals, J., and Mushynski, W. E. (1974). J. Cell Biol. 61, 701. Miledi, R. (1964). Narure (London) 204, 293. Milhaud, M., and Pappas, G. D. (1966). Brain Res. 3, 158. Milofsky, A. (1957). Anar. Rec. 127, 435. Morel, N., Manaranche, R., Gulik-Krzywicki, T., and Israel, M. (1980). J . Ulrrasrruct. Res. 70, 347. Morgan, J. G . , Zanetta, J. P., Breckenridge, W. C., Vincendon, G.,and Gombos, G.(1973). Brain Res. 62, 405. Moms, S. 1. (1980). Neuroscience 5, 1509. Nachmansohn, D. (1959). “Chemical and Molecular Basis of Nerve Activity.” Academic Press, New York. Orrego, F. (1979). Neuroscience 4, 1037. Oschmann, J. L., and Wall, B. ‘J. (1972). J. Cell Biol. 55, 58. Oschmann, J. L., Hall, T. A., Peters, P., and Wall, B. J. (1974). J. Cell Biol. 61, 156. Ostendorf, M. L., Niedorf, H. R., and Blumcke, S. (1971). Z . Zellforsch. 121, 358. Ovtscharoff, W. (1972). Histochemie 29, 240. Ovtscharoff, W. (1973). Histochemie 37, 93. Ovtscharoff, W. (1975a). Histochemistry 44, 185. Ovtscharoff, W. (1975b). Histochemistry 45, 61. Ovtscharoff, W. (197%). Acra Hisrochem. 52, 117. Ovtscharoff, W. (1976a). Experienria 32, 263. Ovtscharoff, W. (1976b). Med.-Biol. Probl. 4, 41. Ovtscharoff, W. (1977). Acra Hisrochem. 60,41. Ovtscharoff, W. (1978). Hisrochemistry 55, 341. Ovtscharoff, W. (1979a). “Structure, Histochemistry, Plasticity and Ontogenesis of the Synapses in the Central Nervous System.” Dr. Sc. Dissertation, Sofia. Ovtscharoff, W. (1979b). C. R. Acad. Bulg. Sci. 32, 545. Ovtscharoff, W. (1979~).Prog. Brain Res. 49, 491. Ovtscharoff, W. (1982). Acra Morphol. (in press). Ovtscharoff, W., and Manolov, S. (1977). Acfa Hisrochem. 60. 329. Palade, G. E., and Bruns, R. R. (1968). J. Cell Biol. 37, 633. Palade, G. E.,and Palay, S. L. (1954). Anar. Rec. 118, 335. Palay, S. L. (1955). Anar. Rec. 121, 348. Palay, S. L. (1956). J. Biophys. Biochem. Cytol. 2, 193. Palay, S. L. (1958). Exp. Cell Res. 5 , 275. Pappas, G. D., and Purpura, D. P. (1966). Nature (London) 210, 1391. Pappas, G.D., and Waxman, S . G. (1972). In “Structure and Function of Synapses” (G.D. Pappas and D. P. Purpura, eds.), pp. 1-43. Raven, New York. Pardos, G.J., and Lentz, T. L. (1976). Bruin Res. 107, 355. Pease, D. C. (1966). J . Ulfrastrucr. Res. 15, 555. Pellegrino de Iraldi, A. (1974). Brain Res. 66, 227. Pellegrino de Iraldi, A., and Cardoni, R. (1979). Cell Tissue Res. 200, 91. Pellegrino de Iraldi, A., and De Lores Arnaiz, G.R. (1970). J. Neurochem. 17, 1601. Pellegrino de Iraldi, A., and De Robertis, E. (1968). Z. Zellforsch. 87, 330. Pellegrino de Iraldi, A., Gueudet, R., and Saburo, A. M. (1971). Prog. Brain Res. 34, 161. Peters, A., and Kaisermann-Abramoff, J. R. (1969). Z. Zellforsch. 100, 487.

THE CHEMICAL SYNAPSES

283

Petkov, V. (1977). “Mechanisms of Synaptic Transmission.” Publishing House of the Bulgarian Academy of Science, Sofia. Pfenninger, K. H. (1973). Prog. Hisrochem. Cvrochem. 5. Pfenninger, K. H., Sandri, C., Akert, K., and Engster, C. H. (1969). Bruin Res. 12, 10. Pfenninger, K. H., Akert, K., Moor, H., and Sandri, C. (1971). Philos. Trans. R . Soc. Ser. B (London) 261, 387. Pickel, V. M., Joh. T. H., Reis, D. J., Leeman, S. E., and Miller, R. J. (1979). Brain Res. 160. 387. Politoff, A. L., Rose, S., and Pappas, G. D. (1974). J. Cell Biol. 61, 818. Rambourg. A. (1969). J. Microsc. 8, 325. Rambourg, A , , and Leblond, C. P. (1967). J . Cell Biol. 32, 27. Rauch, R., Hollenberg, M. D.. and Hope, D. B. (1969). Biochem. J . 115, 473. Rechardt, L., and Harkonen, M. (1977). Histochemistry 51, 113. Reik, J., Petzold, G. L . , Higgins, P., Greengard, J. A,. and Barrnett. R. J. (1970). Science 168, 382. Reinecke, M., and Walther, C. (1978). J. Cell Biol. 78, 839. Richards, J. G., and Tranzer, J. P. (1970). Bruin Res. 17, 463. Richter-Landsberg, C . , Neuhoft, V., and Wachneldt, T. V. (1977). J. Neurosci. Res. 3, 103. Rinvik, E., and Grofova, 1. (1974). A n d . Embryol. 146, 57. Roberts. E.. and Hammerschlag, R. (1972). In “Basic Neurochemistry” (R. W. Albers, G. J . Siegel, R. Katzman, and B. W. Agranoff, eds.), pp. 131-165. Little, Brown, Boston, Massachusetts. Robison, G. A., Butcher, R. W., and Sutherland, E. W. (1971). “Cyclic AMP.” Academic Press, New York. Rosenbluth, J. (1962). J. CeN Biol. 13, 405. Roth, R. H., Walters, J. R., Murrin, L. C., and Morgenroth, V. H. (1975). In “Pre- and Postsynaptic Receptors” (E. Usdin and W. E. Bunney, eds.), pp. 5-48. Dekker, New York. Salpeter, M. M., and Eldefrawi, A. E. (1973). J. Historhem. Cytochem. 21, 769. Sandri. C., Van Buren, J. M., and Akert, K. (1977). Prog. Bruin Res. 46. Schmidt. R., Zimmermann, H., and Whittaker, V. P. (1980). Neuroscience 5, 625. Seijo, L.. and De Lores Arnaiz, G. D. (1970). Biochim. Biophys. Acru 211, 595. Sheridan, M. N., Whittaker, V. P., and Israel, M. (1966). 2. Zellforsch. 74, 291. Siegesmund, K. A., Sances, A., and Larson, S. J. (1969). J . Neurol. Sci. 9 , 89. Singer. S. J.. and Nicholson, G. L. (1972). Science 175, 720. Smith, A. D. (1971 ). Philos. Trans. R. Soc. London Ser. B. Biol. Sci. 261, 363. Smith, A. D.,and Winkler, H. (1967). Biochem. J. 103, 483. Sotelo, C. (1968). Exp. Bruin Res. 6, 294. Sotelo, C. (1971). In “Chemistry and Brain Development. Advances in Experimental Medicine and Biology” (R. Paoletti and A. N. Davison, eds.), Vol. 13, pp. 239-280. Plenum, New York. Stadler, H., and Whittaker, V. P. (1978). Brain Res. 153, 408. Streit, P., Akert, K., Sandri, C., Livingston, R. B., and Moor, H. (1972). Bruin Res. 48, 11. Sugirnura, K., and Mizutani, A. (1978). Hisrochemisrry 55, 97. Sugirnura, K.,and Mizutani, A. (1979). Histochemistry 61, 123. Sutherland, E. W., Robison, G. A., and Butcher, R. (1968). Circulation 37, 279. Tashiro, T.. and Stadler, H. (1978). Eur. J. Biochem. 90, 479. Taxi, J . (1961). C. R . Acad. Sci. (Paris) 252, 1174. Taxi, J. (1965). Ann. Soc. Nut. Zool. 12. 7, 413. Taxi, J. (1967). In “The Neuron” (H. Hyden, ed.), pp. 221-254. Elsevier, Amsterdam. Tranzer, J. P., and Thoenen, H. (1967a). Experienria 23, 743. Tranzer. J. P., and Thoenen, H. (1967b). Experienria 23, 743.

284

STEPHAN MANOLOV AND WLADIMIR OVTSCHAROFF

Tranzer, J. P., Thoenen, H.. Snipes, R. L., and Richards, J. G. (1969). Prog. Brain Res. 31, 33. Tsuji, S. (1978). J. Neurocytol. 7, 381. Uchizono, K. (1965). Narure (London) 207, 642. Valdivia, 0. (1971). J. Comp. Neurol. 142, 257. Van Breeman, V. L., Anderson, E., and Reger, J . F. (1958). Exp. Cell Res. 5. 153. Van der Loos, H. (1963). Z. Zellforsch. 60,815. Van Orden, L. S., Bloom, F. E., Barmett, R. J., and Giarman, N. J. (1966). J. Pharmacol. Exp. Ther. 154, 185. Venzin, M., Sandri, C . , Akert, K., and Wyss, U. (1977). Brain Res. 130, 393. Vrensen, G., Cardozo, J. N., Miiller, L., and Van der Want, J. (1980). Brain Res. 184, 23. Wagner, J . A., and Kelly, R. B. (1979). Proc. Natl. Acad. Sci. U.S.A. 76, 4126. Wagner, J . A., Carlson, S. S . , and Kelly, R. B. (1978). Biochemistry 17, 1199. Walberg, F. (1965). 1. Ultrastruct. Res. 12, 237. Waxman, S. G., and Pappas, G. D. (1969). Brain Res. 14, 240. Weakly. J. N. (1973). J. Physiol. (London)234, 597. West, M. J . , and Del Cero. M. (1976). J. Comp. Neurol. 165. 137. Westfall. J. A.. Yamataka, S., and Enos, P. D. (1971). J. CellBiol. 51, 318. Westrum, L. E. and Broderson, S . H. (1976). J. Neurocytol. 5, 551. Whittaker, V. P. (1966). In “Regulation of Metabolic Processes in Mitochondria” (J. M. Tager, S. Papa, E. Quagliariello, and E. C. Slater, eds.), pp. 1-27. Elsevier, Amsterdam. Whittaker. V. P. (1972). I n “Structure and Function of Synapses” ( G . D. Pappas and D. P. Purpura. eds.), pp. 87-100. Raven, New York. Whittaker. V. P., and Zimmermann. (1974). In “Synaptic Transmission and Neuronal Interaction” (M. V. L. Bennett, ed.), pp. 217-238. Raven, New York. Whittaker, V. P., Dowdall, M. J., Dowe, G. H. C., Facino, R. M., and Scotto, J. (1974). Brain Res. 75, 115. Winkler, H. (1977). Neuroscience 2, 657. Winkler. H., and Westhead, E. (1980). Neuroscience 5, 1803. Wood, J. G . , and Barmett, R. J. (1964). J. Hisrochem. Cyrochem. 12, 197. Wood, M. R.. and Cohen, M. J. (1980). Brain Res. 194, 613. Zacks, S. J . , and Saito, A. (1969). J. Hisrochem. Cytochem. 17, 161. Zimmermann, H. (1978). Neuroscience 3, 827. Zimmermann, H. (1979). Neuroscience 4, 1773. Zisapel, N., and Haklai, R. (1980). Neuroscience 5. 2297. Zisapel, N., Levi, M., and Gozes, J. (1980). J. Neurochem. 34, 26.

Index

A

Acetabularia mediterranea chloroplast DNA buoyant density and GC content, 218-219 content and distribution, 220-224 morphology, 225-234 physical properites, 234 replication, 225 chloroplast gene products, 235-237 chloroplast morphology division, 214-218 isolated chloroplasts, 21 1-213 in siru. 207-21 1 ultrastructure, 2 13-2 14

B Bone, proteins containing ycarboxyglutamic acid, 10- 13

L

Calcium active site and, 20-23 micellar substrates and, 23-24 Calmodulin, calcium and, 28-38 Cancer genetic predisposition in man as autosomic dominant trait, 79-8 I

cell culture studies on autosomal dominant syndromes and chromosomes instability syndromes, 73-75 experimental system, 65-66 mechanisms, 78-79 question of prognosis and control, 81-83 question of tumor promotion, 75-78 studies on ACR cell cultures cancer-related antigens, 70 differential susceptibility to transformation by oncogenic agents, 68-70 intra- and intercellular cytochemical structures, 67 membrane-associated parameters. 67-68 metabolic alterations, 68 serum requirements and growth properties, 66 summary of tissue culture studies, 70-73 y-Carboxyglutamic acid, synthesis of, 6-7 Chloroplast, of A. medirerranea division, 214-218 gene products, 235-237 isolated chloroplasts, 21 1-213 in siru, 207-21 I ultrastructure, 21 3-2 14 Chloroplast DNA, of A. mediterranea buoyant density and GC content, 218-219 content and distribution, 220-224 morphology, 225-234 physical properties, 234 replication, 225

285

286

INDEX

Concanavalin A, calcium binding and, 13-18

E

Endoplasmic reticulum, relationship to GA, 95-97 Enzymes, calcium-binding, 18- 19 active site and, 20-23 micellar substrates and, 23-24 other enzymes, 24-25 phospholipase A2. 19-20

C Golgi apparatus means and ends, 112-114 membrane recycling and evidence for/indications of in higher plants, 115-117 necessity, 114-1 15 possible agents of recycling, 117-1 19 secretion kinetics and membrane turnover, 108-112 sites of synthesis endomembranes. 105- 108 secretional material, 102-105 structure and biochemistry general remarks, 91 polarity, 92-95 relationships to ER, 95-97 to PM, 98-102

P

Parvalbumin, EF hand calcium-binding structures and, 25-28 Phospholipase A*, calcium and, 19-20 Plant movements, study in space environment current projects, 199-202 examples of current flight hardware for study, 194-196 future perspectives, 202-203 movements meaningfully studied, 190- 194 past experiments, 196- 199 Plasma, proteins containing y-carboxyglutamic acid, 9- 10 Plasma membrane, relationship to GA, 98-102 Protein( s) calcium-modulated calmodulin, 28-38 pawalbumin and EF hand calcium-binding StNCtUES, 25-28 SIOO, 42-46 vitamin-D-dependent protein and multiple calcium-binding structures, 39-42 containing y-carboxyglutamic acid, 4-5 background, 5-6 bone proteins, 10-13 calcium-binding characteristics of prothrombin, 7-9 other proteins, 13 plasma proteins, 9- 10 synthesis of y-carboxyglutamic acid, 6-7 vitamin-D-dependent, multiple calcium-binding structures and. 39-42 Prothrombin, calcium-binding characteristics of, 7-9

L

Lectins, calcium-binding, 13-18

M

Membrane recycling, Golgi apparatus and evidence for/indications of in higher plants 115-117 necessity, 114-1 15 possible agents of, 117- 119

S

SIOO, calcium and, 42-46 Space environment characteristics of environmental periodicity, 186- 187 microgravity, 184- 186 other factors, 187-189 radiation, 187 study of plant movements in current projects, 199-202

INDEX examples of current flight hardware for study, 194-196 future perspectives, 202-203 movements meaningfully studied, 190-194 past experiments, 196-199 Space flight, experimental constraints imposed by, 189- 190 Sponges cell interactions in immune response, 168-173 cell movement and sorting out. 161-168 cell types in, 130-132 intercellular matrix of, 132-135 morphology of cell contact, 135-137 primary cellular recognition primary aggregation, 137- 139 species specificity, 139-145 recognition of symbionts, 173- 174 secondary cellular recognition

287

formation of reconstituted organism, 158-159 influence on cell metabolism, 159- 16I secondary aggregation, 145- 154 species specificity, 155- 158 Synapses postsynaptic part, 272-273 membrane, 273-275 postsynaptic thickening, 275-276 subsynaptic structures, 276-277 presynaptic part coated and other vesicles, 262 cytoplasm of axonal ending, 267 membrane of, 244-248 origin of synaptic vesicles, 262-265 other organelles in axonal ending, 265-267 presynaptic density, 267-268 synaptic vesicles, 248-261 transmitter release, 268-270 synaptic cleft, 270-272

This Page Intentionally Left Blank

Contents of Recent Volumes and Supplements Biochemistry and Metabolism of Basement Membranes-NIcHoLAs A. KEFALIDES, C. CLARK ROBERTALPER,A N D CHARLES Transfer RNA-like Structure in Viral GenoThe Effects of Chemicals and Radiations meS-TIMOTHY c. HALL within the Cell: An Ultrastructural and Cytoplasmic and Cell Surface DeoxyribonuMicrurgical Study Using Amoeba proteus cleic Acids with Consideraton of Their Origin-BEVAN L. REIDA N D ALEXAN- as a Single-Cell Model-M. J. ORD Growth, Reproduction, and Differentiation DER J. CHARLSON Biochemistry of the Mitotic Spindlein Acanthamoeba -THOMAS J. BYERS SUBJECT INDEX PETZELT CHRISTIAN Alternatives to Classical Mitosis in Hemopoietic Tissues of VertebrateS-vIBEKE E. ENGELEBERT Volume 62 Fluidity of Cell Membranes-Current ConAND cepts and Trends-M. SHINITZKY Calcification in Plants- ALLAN PENTECOST P. HENKART Cellular Microinjection by Cell Fusion: Macrophage-Lymphocyte Interactions in Technique and Applications in Biology Immune Induction-MARC FELDMANN, and Medicine-MITsuRu FURUSAWA ALANROSENTHAL, A N D PETERE m Cytology, Physiology, and Biochemistry of lmmunohistochemistry of Luteinizing HorG e r m i n a t i o n of F e r n S p o r e s - V . mone-Releasing Hormone-Producing RAGHAVAN Neurons of the Vertebrates-JuLrEN Immunocytochemical Localization of the BARRY Vertebrate Cyclic Nonapeptide NeurohyCell Reparation of Non-DNA Injury-V. pophyseal Hormones and NeurophyYA. ALEXANDROV sins-K. DIERICKX Ultrastructure of the Carotid Body in the Recent Progress in the Morphology, HistoMammals- A L A I N VERNA chemistry, Biochemistry, and Physiology The Cytology and Cytochemistry of the of Developing and Maturing Mammalian WOO1 Fol~icle-DONALD F. G. ORWIN Testis-SARDUL s. GURAYA SUBJECT INDEX Transitional Cells of Hemopoietic Tissues: Origin, Structure, and Development Potential-JOSEPH M. YOFFEY Volume 61 Human Chromosomal Heteromorphisms: Nature and Clinical Significance--RAM The Association of DNA and RNA with S. VERMAA N D HARVEY DOSIK Membranes-MARY PATMOVER SUBJECT INDEX Electron Cytochemical Stains Based on Metal Chelation-DAVID E.ALLEN A N D DOUGLASD. PERRIN Cell Electrophoresis-THOMAS G . PRET- Volume 63 LOW, I1 A N D THERESA P. PRETLOW The Wall of the Growing Plant Cell: Its Physarum polycephalum: A Review of a Model System Using a Structure-FuncThree-Dimensional OrganiZatiOn-JEANROLANDA N D BRIGITTE VIAN tion Approach-EUGENE M.GOODMAN CLAUDE Volume 60

289

290

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS

Microtubules in Cultured Cells: Indirect Im- Structural Aspects of Brain Barriers, with munofluorescent Staining with Tubulin Special Reference to the Permeability of Antibody-B. BRINKLEY, S. FISTEL,J. the Cerebral Endothelium and Choroidal Epithelium-B. VAN DEURS M. MARCUM, AND R. L. PARDUE Septate and Scalariform Junctions in Ar- Immunochemistry of Cytoplasmic Contractthropods-ClcILE N O I R O T - T I M O T ~ E ile Proteins-UTE GR~SCHEL-STEWART AND CHARLES NOIROT The Ultrastructural Visualization of Nuclee lar and Extranucleolar RNA Synthesis The Cytology of salivary GhdS-CARLIN A. PINKSTAFF and Distribution-S. FAKANAND E. PuDevelopment of the Vertebrate ComeaVION Cytological Mechanisms of Calcium CarELIZABETH D. HAY Scanning Electron Microscopy of the Pri- bonate Excavation by Boring Spongesmate SpeTm-KENNETH G . COULD SHIRLEY A. POMPONI Cortical Granules of Mammalian EggsNeuromuscular Disorders with Abnormal BELAJ. GULYAS Muscle Mitochondria-Z. KAMIENIECKA SUBJECT INDEX A N D H. SCHMALBRUCH SUBJECT INDEX

Volume 64

Volume 66

Variant Mitoses in Lower Eukaryotes: Indicators of the Evolution of Mitosis-I. BRENTHEATH The Centriolar COmpkX-sCOTT P. PETERSON AND MICHAEL W. BERNS The Structural Organization of Mammalian Retinal Disc Membrane-J. OLIVE The Roles of Transport and Phosphorylation in Nutrient Uptake in Cultural Animal CdlS-ROBERT M. WOHLHUETER AND PETERG. W. PLAGEMANN The Contractile Apparatus of Smooth Muscle-J. VICTOR SMALL AND APOLINARY SOBIESZEK Cytophysiology of the Adrenal Zona GloIIlerUlOSa-GASTONE G . NUSSDORFER

Dynein: The Mechanochemical Coupling Adenosine Triphosphatase of Microtubule-Based Sliding Filament MechaniSmS-FRED D. WARNER AND DAVIDR. MITCHELL Structure and Function of Phycobilisomes: Light Harvesting Pigment Complexes in Red and Blue-Green Agae-ELisAeETH GANTT Structural Correlates of Gap Junction Permeation-CAMILLo PERACCHIA The Kinetics and Metabolism of the Cells of Hibernating Animals during Hibernation-S. G. K O L A E V AL, . I. K R A MAROVA, E. N. ILYASOVA, AND F. E. ILYASOV

CELLSIM: Cell Cycle Simulation Made Easy-CHARLES E. DONAGHEY The Formation of Axonal Sprouts in Organ Culture and Their Relationship to SproutVolume 65 ing in Vivo-I. R. DIJCEAND P. KEEN When Sperm Meets Egg: Biochemical Cell Surface Glycosyltransferase ActiviMechanisms of Gamete InteractiontieS-MICHAEL PIERCE, EVAA. TURLEY, BENNETT M. SHAPIROAND E. M. EDDY Perisinusoided Stellate Cells (Fat-Storing A N D STEPHEN ROTH Cells, Interstitial Cells, Lipocytes), Their The Transport of Steroid Hormones into Related Structure in and around the Liver Animal Cells-ELEONORA P. GIORCI

SUBJECT INDEX

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS Sinusoids, and Vitamin A-Storing Cells in Extrahepatic OrganS-KENJIRO WAKE SUBJECT INDEX

291

Differentiation of MSH-,ACTH-, Endorphin-, and LPH-Containing Cells in the Hypophysis during Embryonic and Fetal DeVelOpment-JEAN-PAUL DUPOUY

Volume 67 Membrane Circulation in Neurons and Photoreceptors: Some Unresolved IssuesAND ARTHUR M. MERERICHOLTZMAN

Cell Death: The Significance of Apoptosis-A. H. WYLLIE,J. F. R. KERR,A N D A. R. CURRIE INDEX

CURIO

Ultrastructure of Invertebrate Chemo-, Thermo-, and Hygroreceptors and Its Volume 69 Functional Significance-HELMUT ALTThe Structures and Functions of the MycoNER A N D LINDEPRILLINGER plasma Membrane-D. B. ARCHER Calcium Transport System: A Comparative Metabolic Cooperation between Cells-M. Study in Different Ceh-ANNE GODL. HOOPERAND J . H. SUBAK-SHARPE FRAIND-DE BECKERA N D T H ~ O P H I L E The Kinetoplast as a Cell Organelle-V. D. GODFRAIND KALLINIKOVA The Ultrastructure of Skeletogenesis in HerChloroplast DNA Replication in Chlamymatypic Corals-IAN S. JOHNSTON domonas reinhardtii-STEPHEN JAY Protein Turnover in Muscle Cells as VisualKELLERAND CHING H O ized by Autoradiography-J. P. DANucleus-Associated Organelles in FungiDOUNE I. BRENTHEATH Identified Serotonin Neurons-NEVILLE Regulation of the Cell Cycle in Eukaryotic N. OSBORNE A N D VOLKER NEUHOFF CCIk-ROSALIND M. YANISHEVSKY AND Nuclear Proteins in Programming Cell CyGRETCHEN H. STEIN cles-M. V. NARASIMHA RAo The Relationship of in Vitro Studies to in SUBJECT NDEX V i v o H u m a n Aging-EDWARD L . SCHNIEDER A N D JAMESR. SMITH Cell Replacement in Epidermis (KeratoVolume 68 poiesis) via Discrete Units of Proliferation-C. S. POTTEN Moisture Content as a Controlling Factor in INDEX Seed Development and Germination-C. A. ADAMSAND R. W. RINNE Applications of Protoplasts to the Study of Plant CellS-LARRY c. FOWKEA N D Volume 70 OLUFL. GAMBORG Control of Membrane Morphogenesis in Cycling e Noncycling Cell Transitions in Tissue Aging, Immunological SurveilBacteriophage-GREGORY J. BREWER lance, Transformation, and Tumor Scanning Electron Microscopy of IntracelGroWth-sEYMouR GELFANT IUkM StrUCtUreS-KEIICHI TANAKA The Relevance of the State of Growth and The Differentiated State of Normal and Malignant Cells or How to Define a “NorTransformation of Cells to Their Patterns mal” Cell in CUkUR-MINA J. BISSELL of Metabolite Uptake-RUTH KOREN On the Nature of Oncogenic Transformation Intracellular Source of BioluminescenceOf CellS-GERALD L. CHAN M. SWEENEY BEATRICE

292

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS

A N D P. Morphological and Biochemical Aspects of DNA Repair-A. R. LEHMANN Adhesiveness and Dissociation of Cancer KARRAN Celk-HIDEO HAYASHIA N D YASUJI Insulin Binding and Glucose TransportISHIMARU RUSSELLHILF, LAURIEK. %RGE, A N D The Cells of the Gastric MUCOS~-HERBERT ROGERJ. GAY Cell Interactions and the Control of DevelF. HELANDER Ultrastructure and Biology of Female Gaopment in Myxobacteria Populationsmetophyte in Flowering Plants-R. N. DAVIDWHITE KAPIL A N D A. K. BHATNAGAR Ultrastructure, Chemistry, and Function of INDEX the Bacterial Wall-T. J. BEVERIDGF:

INDEX

Volume 71

Volume 73

Integration of Oncogenic Viruses in Mam- Protoplasts of Eukaryotic Algae-MARTHA D. BERLINER malian Cells-CARLO M. CROCE Mitochondrial Genetics of Paramecium au- Polytene Chromosomes of Plants-WALTER NAGL relia-G. H. BEALEAND A. TAIT Histone Gene Expression: Hybrid Cells and Endosperm It s Morphology, Ultrastructure, and Histochemistry-S. P. BHATOrganisms Establish Complex ControlsNAGAR AND VEENASAWHNEY PHILIPHOHMANN Gene Expression and Cell Cycle Regula- The Role of Phosphorylated Dolichols in tiOn-sTEVEN J. HOCHHAUSER, JANET Membrane Glycoprotein Biosynthesis: Relation to Cholesterol BiosynthesisA N D GARYS. STEIN L. STEIN, JOANTUGENDHAFT MILLS A N D ANThe Diptera as a Model System in Cell and THONY M. ADAMANY Molecular Biology-ELENA c.ZEGARELMechanisms of Intralysosomal Degradation LI-SCHMIDT A N D REBAGOODMAN with Special Reference to AutophagocyComments on the Use of Laser Doppler tosis and Heterophagocytosis of Cell OrTechniques in Cell Electrophoresis: Re; ganekS-HANS GLAUMANN, JANL. E. ply to Pretlow and Pretlow’s ReviewERICSSON, AND LOUIS hlARZELLA JOELH. KAPLAN AND E. E. UZGIIUS Comments on the Use of Laser Doppler Membrane Ultrastructure in Urinary TuHUMbules-Letlo ORCI, FABIENNE Techniques as Represented by Kaplan BERT,DENNISBROWN,AND ALAINPERand Uzgiris: Reply t o Kaplan and

-

G.APRETLOW I1 AND UZ~~~~S-THOM S

THERESA P. PRETLOW INDEX

RELET

Tight Junctions in Arthropod TissuesNANCYJ. LANE Genetics and Aging in PrOtOZOa-JOAN SMITH-SONNEBORN INDEX

Volume 72

Volume 74 Microtubule-Membrane Interactions in Cilia and Flagella-WILLIAM L. DENTLER The Plasma Membrane as a Regulatory Site in Growth and Differentiation of NeuroThe Chloroplast Endoplasmic Reticulum: DE I-AAT blastoma Cells-sIeGFIUED Structure, Function, and Evolutionary AND PAULT. V A N DER SAAG Significance-SARAH P. GIBBS

w.

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS

293

Mechanisms That Regulate the Structural Organization and Expression of Viral Genes in Adenovirus- Transformed Cells-S. J. and Functional Architecture of Cell SurfWeS-JANET M. OLIVER A N D RICHARD FLINT Highly Repeated Sequences in Mammalian D. BERLIN Genomes-hhxlME F. SINGER Genome Activity and Gene Expression in Avian Erythroid Celk-hRLEN G . GA- Moderately Repetitive DNA in EvolutionROBERTA. BOUCHARD SARYAN Morphological and Cytological Aspects of Structural Attributes of Membranous OrAlgal CakifCatiOn-MICHAEL A. BOROganelles in Bacteria-CHARLEs C. REMWITZKA

SEN

Naturally Occurring Neuron Death and Its Separated Anterior Pituitary Cells and Their Response to Hypophysiotropic HorRegulation by Developing Neural Pathmones-CARL DENEF,Luc SWENNEN, Ways-TIMOTHY J. CUNNINGHAM ANDRIES The Brown Fat Cell-JAN NEDERGAARD A N D MARIA What Is the Role of Naturally Produced A N D OLOVLINDBERG Electric Current in Vertebrate RegeneraINDEX B. t i o n a n d Healing?-RIcHARD BORGENS Metabolism of Ethylene by Plants-JOHN Volume 75 A. HALL DODDSA N D MICHAEL Mitochondrial Nuclei-TsuNEYosHl Ku- INDEX ROIWA

Slime Mold Lectins-JAMES R. BARTLES, WILLIAMA. FRAZIER,A N D STEVEND. ROSEN Lectin-Resistant Cell Surface Variants of Eukaryotic Cells-EvE BARAKBRILES Cell Division: Key to Cellular Morphogenesis in the Fission Yeast, Schizosaccharomyces-BYRON F. JOHNSON,GODEB. CALLWA, BONGY. Yoo, MICHAELZuKER,A N D IANJ. MCDONALD Microinjection of Fluorescently Labeled Proteins into Living Cells, with Emphasis on Cytoskeletal PrOteinS-THOMAS E. KREISA N D WALTERBIRCHMEIER Evolutionary Aspects of Cell Differentiation-R. A. FLICKINGER Structure and Function of Postovulatory Follicles (Corpora Lutea) in the Ovaries of Nonmammalian V e r t e b r a t e s K. SAIDAPUR SRINIVAS INDEX

Volume 76

Cytological Hybridization to Mammalian Chromosomes-ANN S. HENDERSON

Supplement 1 0 Merentiated Cells in Aging Research

Do Diploid Fibroblasts in Culture Age?EUGENEBELL,Lours MAREK,STEPHAN I E S H E R ,C H A R L O T T M E ERRILL, DONALDLEVINSTONE, AND IANYOUNG Urinary Tract Epithelial Cells Cultured from Human Urine-J. S. FELIXA N D J. W. LITTLEFIELD The Role of Terminal Differentiation in the Finite Culture Lifetime of the Human Epid e r m a l Keratinocyte-JAMES G. RHEINWALD Long-Term Lymphoid Cell CulturesGEORGE F. SMITH,PARVIN JUSTICE, HENRI FRISCHER, LEE KIN CHU, A N D JAMESKROC Type I1 Alveolar Pneumonocytes in VitroWILLIAMH. J. DOUGLAS,JAMESA. MCATEER,JAMESR. SMITH, AND WALTER R. BRAUNSCHWEIGER Cultured Vascular Endothelial Cells as a Model System for the Study of Cellular Senescence-ELLIOT M. LEVINEA N D STEPHENM.MUELLER

294

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS

Vascular Smooth Muscle Cells for Studies Control of Morphogenesis by Inherent and Exogenously Applied Factors in Thin Cell of Cellular Aging in Vitro; an ExaminaLayers-K. TRANTHANH VAN tion of Changes in Structural Cell LipELAINE Androgenetic Haploids-INDRA K. VASIL ids-oLcA 0.BLUMENFELD, SCHWARTZ, VERONICA M. HEARN,AND Isolation, Characterization, and Utilization of Mutant Cell Lines in Higher PlantsMARIEJ. KRANEWOL Chondrocytes in Aging ResearchPAL MALIGA EDWARD J. MILLERAND STEFFENGAY SUBJECT INDEX Growth and Differentiation of Isolated Calvarium C e l l s in a S e r u m - F r e e Medium-JAMES K. BURKS A N D WIL- Supplement 1 1 B Perspectives in Plant Cell and Tissue Culture LIAM A. PECK Studies of Aging in Cultured Nervous System Tissue-DONALD H. SILBERBERCIsolation and culture of ProtopIasts--INAND SEUNG u. KIM DRA K. VASILAND VIMLAVASIL Aging of Adrenocortical Cells in CultureProtoplast Fusion and Somatic HybridizaPETER J. HORNSBY, MICHAELH.SIMONI- tion-OTTO SCHIEDER A N D I N D R ~K. AN,AND GORDONN. GILL VASIL Thyroid Cells in Culture-FRANCESCO s. Genetic Moditication of Plant Cells Through AMBESI-IMPIOMBATO AND HAYDENG. Uptake of Foreign DNA-C. I. KADO COON A N D A. KLEINHOFS Permanent Teratocarcinoma-Derived Cell Nitrogen Fixation and Plant Tissue CulLines Stabilized by Transformation with tUTe-KENNETH L. GILESAND INDKA K. VASIL SV40 and SV4OtsA Mutant VirusesWARRENMALTZMAN, DANIELI. H. Preservation of Germplasm-LyNDsr:Y A. BROWN,ANGELIKA WITHERS LINZER,FLORENCE K. TERESKY,MAURICEROSENSTRAUS,Intraovarian and in Vitro Pollination-M. AND ARNOLD J. LEVINE ZENKTELER Nonreplicating Cultures of Frog Gastric Tu- Endosperm Culture-B. M. JOHRI,P. S. bular CellS-GERTRUDE H. BLUMEN- SRIVASTAVA, AND A. P. RASTE THAL AND DINKAR K. KASBEKAR The Formation of Secondary Metabolites in Plant Tissue and Cell Cultures-H. SUBJECT INDEX MHM Embryo Culture-V. RAGHAVAN The Future-GEoRG MELCHERS Supplement 11A Perspectives in Plant Cell SUBJECT INDEX and Tissue Culture Supplement 12: Membrane Reaearch Cell Proliferation and Growth in Callus CulClassic O r i g h and Current Concepts tures-M. M. YEOMAN AND E. FORCHE Cell Proliferation and Growth in Suspension Membrane Events Associated with the Generation of a BhStOCySt-MARTIN H. Cultures-P. J. KING CytoditTerentiation-RrCHARD PHILLIPS JOHNSON Organogenesis in Vitro: Structural, Physio- Structural and Functional Evidence of logical, and Biochemical AspectsCooperativity between Membranes and TREVORA. THOWE Cell Wall in BaCteria-MANFRED E. BAYER Chromosomal Variation in Plant Tissues in PIant Cell Surface Structure and RecogniCulture-M. W. BAYLISS tion Phenomena with Reference to SymbiClonal Propagation-INDru K. VASILAND VIMLAVASIL S. REISERT OSS-PATRICIA

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS

295

Membranes and Cell Movement: Interac- Agrobacteriurn tumefaciens in Agriculture and Research-FAwzr EL-FIKI A N D tions of Membranes with the Proteins of L. GILES the Cytoskeleton-JAMES A. WEATHER- KENNETH Suppression of, and Recovery from, the NeBEE oplastic State-ROBERT TURGEON Electrophysiology of Cells and Organelles: Studies with Optical Potentiometric Indi- Plasmid Studies in Crown Gall TumorigeneSiS-STEPHEN L. DELLAPORTA AND CatOrS-JEFFREY c. FREEDMAN AND RICKL. PESANO PHILIPC. LARIS Synthesis and Assembly of Membrane and The Position of Agrobacterium rhizoOrganelle Proteins-HARVEY F. LODISH, genes-JESSE M. JAYNESAND GARYA. STROBEL W I L L I A MA. B R A E L L ,A L A N L . SCHWARTZ, GER J. A. M. STROUS,AND Recognition in Rhizobium-Legume Symbioses-TERRENCE L. GRAHAM ASHERZILBERSTEIN The Importance of Adequate Fixation in The Rhizobium Bacteroid State-W. D. SUTTON,C. E. PANKHURST, AND A. S. Preservation of Membrane UltrastrucCRAIG tUR-RONALD B. LUFTIGAND PAUL N. Exchange of Metabolites and Energy beMCMILLAN tween Legume and Rhizobium -JOHN Liposomes- As Artificial Organelles, ToIMSANDE pochemical Matrices, and Therapeutic The Genetics of R h i z o b i u m - A D A ~ CaITkr SyStelTlS-PETER NICHOLLS AND ANDREWW.B. JOHNKONDOROSI Drug and Chemical Effects on Membrane STON Transport-WILLrm 0. BERNDT Indigenous Plasmids of Rhizobium- J. INDEX D ~ N A R P. I ~ BOISTARD, , FRANCINE CASSE-DELBART, A. G. ATHERLY,J. 0. BERRY,AND P. RUSSELL Nodule Morphogenesis and Differentiation- WILLIAMNEWCOMB Mutants of Rhizobium That Are Altered in Supplement 13: Biolosy of the Rhizobkeae Legume Interaction and Nitrogen Fixation-L. D. KUYKENDALL The Taxonomy of the RhizobiaceaeThe Significance and Application of Rhize GERALD H. ELKAN bium in Agriculture-HAROLD L. PETERBiology of Agrobacterium tumefaciens: SON AND THOMAS E.LQYNACHAN Plant Interactions-L. W. MOOREAND INDEX D. A. COOKSEY

This Page Intentionally Left Blank