International Review of Cytology, Volume 85

INTERNATIONAL REVIEW OF CYTOLOGY VOLUME85 ADVISORY EDITORS H. W. BEAMS HOWARD A. BERN GARY G. BORISY PIET BORST BHAR...

Author: Geoffrey H. Bourne | James F. Danielli

9 downloads 1095 Views 18MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

INTERNATIONAL

REVIEW OF CYTOLOGY VOLUME85

ADVISORY EDITORS H. W. BEAMS HOWARD A. BERN GARY G. BORISY PIET BORST BHARAT B. CHATTOO STANLEY COHEN RENE COUTEAUX MARIE A. DIBERARDINO CHARLES J. FLICKINGER OLUF GAMBORG M. NELLY GOLARZ DE BOURNE YUKIO HIRAMOTO YUKINORI HIROTA K. KUROSUMI GIUSEPPE MILLONIG ARNOLD MITTELMAN AUDREY MUGGLETON-HARRIS ALEXANDER

DONALD G. MURPHY ROBERT G. E. MURRAY RICHARD NOVICK ANDREAS OKSCHE MURIEL J. ORD VLADIMIR R. PANTIC W. J. PEACOCK DARRYL C. REANNEY LIONEL I. REBHUN JEAN-PAUL REVEL JOAN SMITH-SONNEBORN WILFRED STEIN HEWSON SWIFT K. TANAKA DENNIS L. TAYLOR TADASHI UTAKOJI ROY WIDDUS YUDIN

INTERNATIONAL

Review of Cytology EDITED BY

G . H. BOURNE

J. F. DANIELLI

St. George's University School of Medicine St. George's, Grenada

Danielli Associates Worcester, Massachusetts

West indies

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

VOLUME85 1983

ACADEMIC PRESS

A Subsidiary of Harcour: Brace Jovanovich, Publishers

New York London Paris San Diego San Francisco Sfio Paulo Sydney Tokyo Toronto

COPYRIGHT @ 1983, BY ACADEMIC PRESS, INC. ALL RIGHTS RESERVED. NO PART O F 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. (LONDON) LTD. 24/28 Oval Road, London NWI

7DX

LIBRARY OF CONGRESS CATALOG CARDNUMBER:52 -5203

I S B N 0-12-364485-2 PRINTED IN THE UNITED STATES OF AMERICA 83848586

9 8 7 6 5 4 3 2 1

Contents CO~RIBUTORS ...........................................................

ix

Receptors for Insulin and CCK in the Acinar Pancreas: Relationship to Hormone Action IRA

I. I1. 111. IV . V. V1.

D. GOLDFINE AND JOHNA. WILLIAMS

Introduction . . . . . . . . . . . ................................ Isolated Pancreatic Acini Cholecystokinin . . . . . . . . . . . . . ........................... Insulin ............... Interactions between Insulin and CCK .................................... Summary and Conclusions .............................. References . . . . . . . . . . . . . . ..........................

1

2 4 19

29 34 35

The Involvement of the Intracellular Redox State and pH in the Metabolic Control of Stimulus-Response Coupling ZYGMUND ROTH. NAOMICHAYEN. AND SHABTAY DIKSTEIN I. I1. 111. IV .

Introduction ......................................................... Relationships between the Intracellular Redox State and Agonist Action . . . . . . . . The Involvement of Intracellular pH in Stimulus-Response Coupling . . . . . . . . . . Epilogue ............................................................ References ..........................................................

39 42 55 51

58

Regulation of DNA Synthesis in Cultured Rat Hepatoma Cells ROELAND VAN WIJK I. I1. I11. IV . V. VI . VII . VIII .

Introduction ......................................................... Cell Cycle of Hepatoma Cells .................................. Factors Influencing DNA Synthesis ...................................... Protein Synthesis and Initiation of DNA Synthesis ................. Polyamine Synthesis .................................................. Chromatin Structure and Its Changes during the Cell Cycle.. System of Second Messengers or Intracellular Regulators .................... Concluding Remarks .................................................. References .............................. ........................

V

63 65 12

79 82 86 96 98 99

vi

CONTENTS

Somatic Cell Genetics and Gene Mapping FA-TENKAO Introduction ......................................................... ............. Historical View of Gene Mapping . . . . . . . . . . . . . . . Development of Somatic Cell Genetics ................................... Use of Somatic Cell Genetics in Gene Mapping. ........................... Combined Use of Somatic Cell Genetics and Recombinant DNA Technology in Gene Mapping.. ................................................... VI. Prospect of Constructing a Complete Human Gene Map ..................... VII. Gene Mapping in Other Species Using Somatic Cell Genetics. . . . . . . . . . . . . . . . . .............. VIII. Conclusions ................... References .......................................................... I. 11. 111. IV . V.

109 110 112 116 126 135 138 139 140

Tubulin Isotypes and the Multigene Tubulin Families N. J. COWANAND L. DUDLEY I. 11. 111. 1V. V.

Introduction ......................................................... Number and Complexity of Tubulin Genes.. .............................. Functional and Nonfunctional Tubulin Genes .............................. Tubulin Gene Expression .............................................. Genetic Complexity and Functional Diversity .............................. References ..........................................................

147 148 157 167 171 172

The Ultrastructure of Plastids in Roots JEANM. WHATLEY I. Introduction

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

11. General Features

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

Nongreen Roots.. . . . . . . . . . . . ............. Green Roots.. ....................................................... ........................ The Greening of Roots ............... Greening and Plastid Division.. ......................................... Plastids in Sieve Elements.. ..... ........................ Geotropism and Plastids in Root Plastid Pigments and Responses ............. X. Some Nonphotosynthetic Functions of Plastids in Roots ..................... XI. Conclusions ......................................................... References ..........................................................

111. IV. V. VI. VII . v111. IX.

180 188 196 202 206 212 214 2 16 2 17

CONTENTS

vii

The Confined Function Model of the Golgi Complex: Center for Ordered Processing of Biosynthetic Products of the Rough Endoplasmic Reticulum ALANM . TARTAKOFF 1. Introduction ......................................................... II . The Confined Function Model of the Golgi Complex ........................ Ill . Covalent Modifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV . Noncovalent Modifications ............................................. V . Consequences of Processing for the Golgi Complex and the Cell as a Whole . . . . References ..........................................................

221 222 231 245 246 248

Problems in Water Relations of Plants and Cells PAULJ . KRAMER

I. II . III . IV . V. VI . VII .

...................... Introduction . . . . . . . . . Terminology . . . . . . . . . ............... Cell Structure and Water Relations ....................... Water Movement in Plants ............................................. Injury from Water Deficits .......... ........................ Adaptations Increasing ...................... Summary .................... ...................... References ....................................................

253 255 258 266 271 277 281 282

Phagocyte-Pathogenic Microbe Interactions ANTOINETTE RYTERAND CHANTAL DE CHASTELLIER

I. I1 . Ill . IV . V.

Introduction .................................... Adhesion . . . . . . . . . . .............................. Ingestion .................................... Microbicidal Activity . . .......................... Phagosome-Lysosome ....................

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

287 288 293 296 302 316 318 319

INDEX.................................................................... CONTENTSOF RECENTVOLUMESAND SUPPLEMENTS ..............................

329 335

v1.

VII . References

This Page Intentionally Left Blank

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

NAOMICHAYEN(39),Unit of Cell Pharmacology, School of Pharmacy, Hebrew University, Jerusalem, Israel N. J . COWAN(147),Department of Biochemistry, New York University School of Medicine, New York, New York 10016

CHANTAL DE CHASTELLIER (287),Unite' de Microscopie Electronique, De'partement de Biologie Mole'culaire, Institut Pasteur, 75724 Paris Cedex 15, France SHABTAYDIKSTEIN(39),Unit of Cell Pharmacology, School of Pharmacy, Hebrew University, Jerusalem, Israel L. DUDLEY(147),Department of Biochemistry, New York University School of Medicine, New York, New York 10016

IRA D. GOLDFINE (l), Cell Biology Laboratory, Harold Brunn Institute, and Department of Medicine, Mount Zion Hospital and Medical Center, San Francisco, California 94120, and Departments of Medicine and Physiology, University of California, San Francisco, California 94120

FA-TENKAO (109),Eleanor Roosevelt Institute for Cancer Research, and Department of Biochemistry, Biophysics and Genetics, University of Colorado Health Sciences Center, Denver, Colorado 80262 PAULJ. KRAMER(253), Department of Botany, Duke University, Durham, North Carolina 27706 ZYGMUNDROTH~(39),Unit of Cell Pharmacology, School of Pharmacy, Hebrew University, Jerusalem, Israel

ANTOINETTE RYTER (287),Unite'de Microscopie Electronique, De'partement de Biologie Mole'culaire, Institut Pasteur, 75724 Paris Cedex 15, France 'Present address: Johnson Research Foundation, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104. ix

X

CONTRIBUTORS

ALANM. TARTAKOFF (221), Department of Pathology, University of Geneva School of Medicine, Centre Mkdical Universitaire, CH-I211 Geneva 4,Switzerland ROELANDVAN WIJK(63), Department of Molecular Cell Biology, State University, 3584 CH Utrecht, The Netherlands JEAN M . WHATLEY(175), Botany School, Oxford University, Oxjord OX1 3RA, England JOHN A. WILLIAMS( l ) , Cell Biology Laboratory, Harold Brunn Institute, and Department of Medicine, Mount Zion Hospital and Medical Center, San Francisco, California 94120, and Departments of Physiology and Medicine, University of California, San Francisco, California 94120

INTERNATIONAL

REVIEW OF CYTOLOGY VOLUME85

This Page Intentionally Left Blank

INTERNATIONAL REVIEW OF CYTOLOGY, VOL. 85

Receptors for Insulin and CCK in the Acinar Pancreas: Relationship to Hormone Action IRAD. GOLDFINE AND JOHNA. WILLIAMS Cell Biology Laboratoty, Harold Brunn Institute, and Department of Medicine, Mount Zion Hospital and Medical Center, San Francisco, California, and Departments of Medicine and Physiology, University of California, San Francisco, California 1. Introduction .............................................. 11. Isolated Pancreatic Acini for the Study of Exocrine Function . . . . . . 111. Cholecystokinin . . . . A. Background .......................................... B. Effects on Acinar Cells . . . .

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

1

2 4 4 4 10 16 19 19

B . Effects on Acinar Cells ................................. C. Characteristics of Insulin Receptors ....................... D. Autoradiographs of '25I-Labeled Insulin . . . . . . . . . . . . . . . . . . . V. Interactions between Insulin and CCK ......................... VI. Summary and Conclusions .................................. References ...............................................

21 24 26 29 34 35

I. Introduction The pancreatic acinar cell has been a model for studying the various ultrastructural events involved in the synthesis, packaging, and release of secretory proteins. Secretory proteins are synthesized on membrane-bound ribosomes, sequestered within the lumen of the endoplasmic reticulum (ER), transported through the Golgi, packaged into granules, moved to the cell membrane, and then released by exocytosis (Case, 1978; Palade, 1975; Scheele, 1980). Subsequently, this model has been used (with some cell-specific modifications) to explain the secretory processes of a large number of endocrine, exocrine, and blood cells. In contrast, the biochemical events involved in the control of pancreatic secretion are less well understood. I n vivo, the major regulation of the exocrine pancreas is via polypeptide hormones and cholinergic neurons. Two recent developments for studying acini in v i m have contributed to an understanding of the regulation of pancreatic acinar cell function. First, hormone-sensitive preparaI Copyright 0 1983 by Academic Ress. Inc. All rights of reproduction in any form reserved. ISBN 0-12-364485-2

2

IRA D. GOLDFINE AND JOHN A. WILLIAMS

tions of pancreatic acini have been developed. Second, biologically active, radiolabeled hormones have been prepared. This article will survey the recent studies camed out in our laboratory and other laboratories to probe the receptors and mechanism of action of two major polypeptide hormone regulators of the exocrine pancreas, cholecystokinin (CCK), and insulin.

11. Isolated Pancreatic Acini for the Study of Exocrine Function Amsterdam and Jamieson (1972) were the first to devise a procedure for preparing isolated pancreatic acinar cells which employed digestion of the pancreas with collagenase and chymotrypsin, chelation of divalent cations with EDTA, and mechanical shearing. Similar preparations have now been used for studies of acinar cell hormone receptors, ion fluxes, and cyclic nucleotide levels (Christophe et al., 1976a,b; Gardner et al., 1975; Kondo and Schulz, 1976; Williams, 1977; Williams et al., 1976). The ability of secretagogues to induce 45Ca2+ efflux in isolated cells suggests that both hormone receptors and the initial steps in stimulus-secretion coupling are intact. In most investigations, however, the measurement of enzyme secretion (a distal event) by isolated acinar cells has been difficult. For example, amylase release from the perfused rat pancreas is increased 8- to 20-fold by both acetylcholine and CCK (Kanno, 1972), whereas it is increased only 2-fold or less from isolated rat acinar cells (Kondo and Schulz, 1976).

FIG. 1. Light micrograph of isolated mouse pancreatic acini.

3

INSULIN AND CCK IN THE ACINAR PANCREAS

T

10-12

10.11

I

1

1

10.10

10.8

10-8

I

10.7

I

I

10.6

h b w e (MI

Fro.2. Dose-response relationship for amylase release from isolated rat pancreatic acini induced by CCK and its analogs. In each case, basal release was subtracted and secretagogue-induced release was calculated as percentage of maximal release. All analogs induced a similar maximal release. (From Williams era!., 1981.)

When the dissociation procedure is modified to produce isolated acini, a considerably improved secretory response is observed (Peikin et al., 1978; Schultz et al., 1980; Williams ef af., 1978). Isolated acini are prepared in a manner similar to that for isolated cells but without the calcium chelation step necessary to break junctional complexes (Amsterdam and Jamieson, 1974). They consist of groups of acinar (and occasionally centroacinar) cells arranged around an intact lumen (Fig. 1). Ultrastructural evaluation reveals that the tight junctions connecting adjacent acinar cells are maintained along with the microvilli and their underlying microfilament network located at the apical border of the acinar cell (Schultz er al., 1980; Williams et al., 1978). Since isolated acini can be studied as a homogeneous suspension and have their basolateral plasma membrane exposed to the incubation medium, they possess all the advantages of the isolated cells. Moreover, it is possible to measure and correlate enzyme release with other cellular functions, including occupancy of membrane receptors (Williams, 1980). Isolated pancreatic acini have proven especially useful in evaluating the complex dose-response curves for secretagogues such as the neurotransmitter ACh and the hormone CCK (Fig. 2). Isolated acini have also been utilized to study secretagogue inhibitors, including atropine for the cho-

4

IRA D. GOLDFINE AND JOHN A. WILLIAMS

linergic agents and dibutyryl cyclic GMP for CCK (Peitkin et al., 1979; Williams et af., 1978).

111. Cholecystokinin

A. BACKGROUND Cholecystokinin (CCK) which was originally isolated from the porcine intestine based on its ability to stimulate both gallbladder contraction and pancreatic secretion, is a straight chain 33 amino acid peptide (CCK,,) with an amidated Cterminus (Jorpes and Mutt, 1973; Mutt and Jorpes, 1971). A prominent feature of CCK is the presence of a sulfated tyrosine at position 27 (7 residues from the C-terminus) (Table I). The C-terminal octapeptide (CCK,) possesses a high degree of biological activity and is, in fact, more potent than CCK,, (Mutt and Jorpes, 1968; Ondetti et af.,1970a,b). The unsulfated octapeptide, however, has only 1/150 of the activity of the sulfated form (Gardner et af., 1975; Ondetti et af., 1970a; Williams et af., 1981). The C-terminal tetrapeptide appears to contain all the biological activity of CCK although it is 30,000-fold weaker than the octapeptide whereas the C-terminal tripeptide or the deamidated tetrapeptide has no activity (Morley er af., 1965; Rajh et af., 1980; Sankaran et al., 1981a). CCK is also structurally similar to the frog skin decapeptide, caerulein, which acts like CCK, (Anastasi er al., 1968). Another notable feature of the C-terminal portion of CCK is its homology to the similar portion of the gastrin molecule. The C-terminal pentapeptides of both molecules are identical, and both CCK and some forms of gastrin contain a sulfated tyrosine, although not in the identical position (Table I). Gastrin, however, has only weak effects on acinar cells (Fig. 2). Separate CCK and gastrinlike molecules are only found in higher animals, such as reptiles, birds, and mammals (Larsson and Rehfeld, 1977), but not lower animals. The basic biological activity of CCK is contained in the C-terminal tetrapeptide amide (which is shared with gastrin) whereas the additional amino acids are essential to increase specificity for pancreas and gallbladder. B. EFFECTS ON ACINAR CELLS Cholecystokinin has a number of actions on pancreatic acinar cells that have been demonstrated both in vivo and in virro (Table 11). Injection of CCK in vivo leads to the release of zymogen granule contents into the acinar lumen. In some species, especially rodents, CCK also stimulates the production of a C1 --rich pancreatic juice such that the secreted enzymes pass via the pancreatic ducts to the intestine (Case, 1978). In sbme other species, such as the cat, fluid secretion

TABLE I AMINOACID SEQUENCE OF CHOLECYSTOKININ AND GASTRIN Cholecystokinin33

Lys -Ala -Pro -Ser -Gly -Arg -Val Ser -Met -Ile -Lys -Asn -Leu -Glu -Ser

503h

Leu -Asp -Pro -Ser -His -Arg -1le -Ser -Asp -Arg -Asp -Tyr -Met -Gly -Trp -Met -Asp -Phe -NH2 Gastrinl7

Glp -Gly -Pro -Trp -Met -Glu -Glu -Glu -Glu -Glu -Ala -Tyro -Gly -Trp -Met -Asp -Phe -NH2

PGastrin exists in two forms, I (nonsulfated tyrosine) and I1 (sulfated tyrosine). Glp, pyroglutamic acid.

6

IRA D. GOLDFINE AND JOHN A. WILLIAMS

EFFECTS

OF

TABLE rr CHOLECYSTOKININ ON PANCREATIC ACINARCELLS

1. Zymogen synthesis and secretion 2. Secretion of CI--rich pancreatic juice 3. Pancreatic hypertrophy and hyperplasia 4. Increased glucose and amino acid utilization and oxygen consumption

is bicarbonate rich and requires stimulation by the hormone secretin. CCK in vivo also accelerates the synthesis of digestive enzymes, indicating coordinated stimulation of zymogen synthesis and secretion (Case, 1979; Webster el af., 1977). More detailed studies suggest that CCK may regulate the synthesis of specific pancreatic zymogens (Dagorn and Mongeau, 1977). Chronic stimulation with either CCK or its analogs for 5- 15 days also induces pancreatic hypertrophy and hyperplasia due to an increase in structural proteins and nucleic acids as well as zymogen (Mainz et af.. 1973; Solomon et af., 1978). In v i m studies of the effects of CCK and the mechanisms involved have largely been carried out with isolated pancreatic acini. Secretion by isolated pancreatic acini is usually quantitated by measuring either the amount of a specific, easily measured digestive enzyme such as amylase (Fig. 2) or by pulse labeling the newly synthesized zymogen with radioactive amino acid. Enzyme secretion is an event that is clearly separated from enzyme synthesis, as secretion can take place even when protein synthesis is inhibited (Jamieson and Palade, 1971; Otsuki and Williams, 1982b). The predominant view is that all pancreatic zymogens are secreted in parallel by exocytosis and in proportion to their pancreatic content, implying a single control process (Case, 1978; Palade, 1975). An alternative view is that the secretion of digestive enzymes is controlled individually and that enzymes leave the acinar cell by a nonexocytotic mechanism (Rothman, 1975, 1980). Although it has not been possible in isolated acinar cells to directly measure fluid production, a direct effect of CCK on this function is indicated by the actions of either CCK or its analogs to increase both radiosodium uptake (Putney et af., 1980) and the turnover of the transport enzyme Na+ -K+ -ATPase ( S . R. Hootman and J. A. Williams, unpublished data). In addition to its effects on secretion, the synthetic and metabolic effects of CCK have also been studied in vitro. Recently, studies with isolated pancreatic acini have shown a direct effect of low concentrations of CCK on stimulation of acinar protein synthesis whereas higher concentrations of hormone bring about inhibition (Fig. 3) (Korc et al., 1981a). CCK stimulates the oxidation of glucose, alanine, and leucine by fragments of mouse pancreas (Danielsson and Sehlin, 1974). Oxygen consumption is increased (Dickman and Morrill, 1957), presumably reflecting increased energy turnover as both secretion and synthesis require

7

INSULIN AND CCK IN THE ACINAR PANCREAS

2501

1,

1

I

I

I

lo-ii 3.,o.ii

I

lo.io

I

I

lo.e

3.10.io

CCKg (MI

FIG.3. Effect of CCKs and insulin on [3H]leucine incorporation in isolated pancreatic acini from diabetic rats. Insulin was added at 0.17 fl.Each value is the mean SE from four experiments. (From Korc er al., 1981a.)

*

ATP. CCK also stimulates glucose transport by isolated pancreatic acini as determined by using both 2-deoxyglucose (Fig. 4) and 3-0-methylglucose. In contrast, CCK inhibits the uptake of the nonmetabolized amino acid a-aminoisobutyric acid (AIB); this inhibition is thought to be due to a reduction in the electrochemical gradient for Na+ (Iwamoto and Williams, 1980). Of interest is the difference in shape of the various dose-response curves for CCK acting on isolated pancreatic acini. The curves for enzyme secretion and protein synthesis are biphasic, whereas those for stimulation of glucose transport and the inhibition of AIB uptake are monophasic and require higher concentrations of CCK. Since CCK binds initially to receptors localized on the basolateral plasma membrane and then rapidly initiates zymogen release at the luminal membrane, it has long been apparent that that action of CCK must be mediated by a second messenger. Although early work focused on a possible role for cyclic AMP, it is now clear that this nucleotide is not involved. Under physiological conditions, CCK does not increase pancreatic cyclic AMP content and neither exogenous derivatives of cyclic AMP nor phosphodiesterase inhibitors mimic the action of CCK (Case, 1978). CCK does bring about an increase in cyclic GMP, but the rise in cyclic GMP appears secondary to a rise in intracellular Ca2 (Christophe et al., 1976b). In contrast to the results with cyclic AMP and cyclic GMP, considerable evidence exists for the role of cytoplasmic Ca2 as the intracellular +

+

8

IRA D. GOLDFINE AND JOHN A. WILLIAMS

9GI

'

0

a

L1llll

0.1

' '

*

1

' ' * l*ll.l

10

'

' * ,*J

100

CCK (nM)

FIG.4. Effect of CCK on uptake of [3H]2-deoxyglucose (2-DG) by isolated mouse pancreatic acini. Each value is the mean SE from four experiments. (From Sankaran er a/., 1982.)

*

mediator of CCK (Case, 1978; Schulz, 1980; Williams, 1980). The major points are that (1) CCK increases the movement of Ca2+ into and out of acinar cells, (2) removal of extracellular Ca2+ either reduces or abolishes the action of CCK, and (3) the action of CCK can be mimicked by artificial introduction of Ca2+ into acinar cells by means of calcium ionophores, particularly A23187. A recent report using Ca2 -sensitive microelectrodes has determined that the concentration of ionized Ca2+ in unstimulated pancreatic cells is 3 X lo-' M and that Ca2 increases upon stimulation with acetylcholine (O'Doherty and Stark, 1982). Since CCK is known to act similarly to acetylcholine (although via distinct receptors), it seems likely that CCK will have a similar effect. Some controversy exists over the source of Ca2+ as both entry from the extracellular fluid and release from intracellular stores has been proposed (Schulz, 1980; Williams, 1980). It seems clear, however, that release from intracellular stores is the predominant event since enzyme secretion can take place for several minutes in the complete absence of extracellular Ca2+, even in the presence of extracellular chelators such as EGTA (Scheele and Haymovits, 1980; Williams, 1980). It is not completely clear which organelle(s) releases Ca2+ in response to stimulation with CCK; mitochondria, plasma membrane, and endoplasmic reticulum have been proposed (Chandler and Williams, 1978; Dormer and Williams, 1981; Schulz er al., 1980). The nature of the signal from CCK receptors located on the plasma membrane to the intracellular Ca2+ stores is also a matter for further investigation. Little is known about the mechanism by which the rise in cytoplasmic Ca2 brings about the increase in amylase release. This process is energy dependent since inhibitors which lower cellular ATP levels block the action of Ca2+ on +

+

+

INSULIN AND CCK IN THE ACINAR PANCREAS

9

secretion (Williams and Lee, 1974). Calmodulin, a calcium receptor protein, is present in pancreas and may be involved in secretion (Vandermeers et al., 1977). In other cell types calmodulin, after binding Ca2 , activates a number of Ca2 activated protein kinases. In support of such an effect in acini, it has recently been shown that CCK, as well as both the cholinergic analogs and Ca2+ ionophore A23187, alter the phosphorylation of at least five proteins (Burnham and Williams, 1982). In this study the secretagogues increased the phosphorylation of a M, = 32,000 particulate protein and M, = 16,000 and 23,000 soluble proteins. The agents also caused the dephosphorylation of M, = 21,000 and 20,500 soluble proteins (Fig. 5 ) . The time course of phosphorylation, its +

+

FIG. 5. Autoradiographs of soluble and particulate fractions from mouse acini prelabeled with 32P for I hour and then incubated for 5 minutes with no additions (A), 3 pM carbachol (B). 3 pM ionophore A23187 ( C ) , or 300 pM CCKB (D). (From Burnham and Williams, 1982.)

10

IRA D. GOLDFINE AND JOHN A. WILLIAMS

dose-response, and the reversibility of these changes were consistent with their being involved in the mediation of the acute biological effects of CCK. In addition, Ca2+-activated kinases are present in pancreatic cytosol. Wrenn et al. (1981) described a phospholipid-dependent, Ca2 -activated kinase in the 30,000 g supernatant prepared from rat pancreas. Using isolated mouse pancreatic acini, we have found separate phospholipid-requiringand calmodulin-requiringCa2 activated kinases with different specificities for both endogenous and exogenous proteins (Burnham and Williams, unpublished data). +

+

C. CHARACTERIZATION OF CCK RECEPTORS In order to directly study the initial steps in pancreatic stimulus-secretion coupling that are initiated by CCK, we have employed isolated pancreatic acini to study CCK receptors. For these studies, we prepared radioiodinated CCK by its conjugation to 1251-labeledBolton-Hunter (BH) reagent (Sankaran et al., 1979). This procedure avoids the use of oxidizing conditions that destroy the biological activity of CCK, presumably by oxidation of essential methionine residues. The lZ51-labeled BH-CCK conjugate is of high specific activity (2 15-260 pCi/ pg) and retains full biological and immunological activity (Fig. 6) (Sankaran et al., 1979). Using this ligand, we were able to study hormone binding to acini at 37°C under the same conditions employed to study amylase release (Sankaran et af., 1980). Studies using isolated rat pancreatic acini demonstrated that the binding of 1251-labeledCCK was rapid, reversible, and linearly related to the number of

4r 1210-

a6-

4-

2-

OL

4/

0

3010.11

I

1

10.10

3010.10

I

10.9

I

3010.9

CCK (M)

FIG. 6. Comparative biological activities of CCK and 12sI-labeled BH-CCK in the rat amylase release assay. (From Sankaran er al., 1979.)

11

INSULIN AND CCK IN THE ACINAR PANCREAS Secretin

1 PP

0

200

400

CCK BOUNDlf m o l /

600

800

mg protein)

Fic. 7. (A) Competitive inhibition of 125I-labeled CCK binding in isolated rat pancreatic acini by CCK and the lack of effect of other hormones. (B) Scatchard plot of the specific binding of CCK to rat pancreatic acini. These data can be interpreted as indicating either the presence of two orders of binding sites or negative cooperativity. (From Sankaran et al., 1980.)

acini. Addition of nonradioactive CCK (10- ' I - lo-' M) competitively inhibited the binding of 1251-labeledCCK whereas other nonrelated hormones, such as secretin, pancreatic polypeptide, insulin, and glucagon, were without effect (Fig. 7A). The biphasic competitive inhibition curve suggesting two classes of CCK binding sites on acini was supported when the data were replotted as a Scatchard plot (Fig. 7B). These saturable components had affinities (Kds) of 70 pM and 20 nM. That the radioactivity bound to acini remained as intact CCK was confirmed

12

IRA D. GOLDFINE AND JOHN A. WILLIAMS

by extraction and gel filtration (Sankaran ef al., 1980). When analogs of CCK of known varying potency were studied there was an excellent correlation between the ability of the analogs to compete for CCK binding and to stimulate amylase release (Table 111). For both parameters, the potency ratios were caerulein> CCK,>CCK,,> desulfated caerulein>gastrin I>CCK,. A similar correlation between binding of CCK, stimulation of calcium efflux, formation of cyclic GMP,and release of amylase has also been reported for acini prepared from guinea pig pancreas (Jensen et al., 1980). As the progressive stimulation of amylase release by CCK correlated well with the occupancy of the high affinity CCK binding sites, we concluded that the high affinity CCK binding sites most likely are the receptors mediating the stimulation of amylase secretion by CCK. To study further the relationship of CCK receptor occupancy to biological effects, we then studied mouse pancreatic acini since mouse acini are more sensitive to CCK and in this species CCK stimulates not only amylase release but also increases glucose transport and inhibits AIB uptake. Binding of 1251-labeled CCK to mouse acini also produces a curvilinear Scatchard plot which is best interpreted as indicating two orders of binding sites with Kds of 25 pM and 2 nM, respectively (Sankaran et al., 1982). These affinity values are about 3-fold greater than those from the earlier study on rat acini; in agreement with these findings, mouse acini are more sensitive to CCK than rat acini in terms of the concentration of CCK required for maximal amylase release. Since both rat and mouse acini degrade 12SI-labeledCCK in a time- and concentration-dependent manner, we were concerned that this degradation might alter the binding parameters. However, when either the binding data were corrected for hormone degradation or the degradation was blocked by the addition of bacitracin, similar binding parameters were calculated (Sankaran et al., 1982). As shown in Fig. 8, while the curve for stimulation of both glucose transport and inhibition of AIB uptake was similar to the curve for occupancy of a single class of binding sites, TABLE 111 COMPARISON OF POTENCY VALUESOP CCK

Analog Caerulein CCKB CCK33 Desulfated caerulein Gastrin I

Inhibition of receptor binding 6.25

3.18 1 .oo 0.039 0.004

AND

ITS A N A L ~ G S ~ Stimulation of amylase secretion 13.5 5.41 1 .OO 0.042 0.002

“The ability of the polypeptides to one half-maximally inhibit receptor binding and to one half-maximallystimulate amylase secretion relative to CCK33is shown (from Sankaran et al., 1980).

13

INSULIN AND CCK IN THE ACINAR PANCREAS

(012

(0-11

(010

(0.8

(0-8

$07

CCK (M)

FIG.8. Concentration dependence of CCK-stimulated amylase secretion, glucose and AIB transport, and the occupation of high and low affinity CCK receptor sites in mouse acini. (Modified from Sankaran era!., 1982.)

the relation between receptor occupancy and amylase release is clearly complex. Qualitatively, the amylase release dose-response could be explained if the high affinity receptor stimulated and the low affinity receptor inhibited amylase release. To test this hypothesis we constructed a model of CCK-simulated amylase release (Fig. 9). In this model we incorporated a provision that the high affinity IRnl

+ [CCN

KdH

KdH

--

+ [CCK]

Amylase release

[(CCK),*R,*(CCK)*R,I

-C

KdH [R,]

>+

K ~ H [CCK*R,I + [ C C K l d [(CCK)2*RH]-

KdL [(CCK).RLl

Glucose and A I0 transport

FIG.9. Schematic model for production of various biological effects by CCK acting through two discrete receptors of high affinity (RH) and low affinity (RL). respectively. High affinity receptors bind two molecules of CCK per receptor molecule with the binding affinity of the second reduced by the cooperativity coefficient C*RL.The low affinity receptors bind only one molecule of CCK, and mediate the effects of CCK on glucose and AIB transport. Amylase release is a function of the high affinity receptor complex (CCK)z*RH.This complex is inactivated, however, by (CCK)*RLreceptor complex. This inactivation is shown here by combination of [(CCK).RL] and [ ( c c K ) y R ~to ] yield an inactive form [(CCK)yRH.(CCK).RL]. (From Sankaran er 01.. 1982.)

14

IRA D. GOLDFINE AND JOHN A. WILLIAMS

CCK receptor sequentially binds two molecules of CCK with modest cooperative interaction. The inclusion of a cooperative interaction was added to account for the observation that in the mouse (Sankaran et al., 1982), rat (Sankaran er al., 1980), and guinea pig acini (Jensen et al., 1980), the addition of nonradioactive CCK accelerates the dissociation of labeled CCK, a finding originally observed for insulin receptors and explained as indicating negative cooperative interactions (De Meyts et al., 1973). In the present model, the occupied high affinity receptor activates amylase release whereas the occupied low affinity receptor, either directly or by generation of a product, inactivates the high affinity receptor complex. By use of the experimentally derived binding constants and appropriate 100

80

-A -

6040

-

20 .-

0.01

0.1

10

100

10

100

CCK (nM)

0.01

0.1

1

CCK (nM)

FIG. 10. Correlation of CCK receptor occupancy to CCK-induced changes in amylase release (A) and 2-DG and AIB transport (B). Data for biological effects are plotted as percentage of maximal effect. In A, the line through the data points is the second power of [(CCK)yRH] complex computed from the model shown in Fig. 9 using experimentally derived values for KdHand KdL, a cooperativity coefficient C of 2, and a Ki (the constant for inactivation of the occupied high affinity receptor) of 50 pM. In B the dotted line is occupancy of the low affinity receptor (CCK).RL computed from the experimentally derived value for KdL and the lines through the data points are computed from an expression quantitating the effect of spare receptors. (From Sankaran et al., 1982.)

INSULIN AND CCK IN THE ACINAR PANCREAS

15

choices for the cooperativity coefficient C and the inactivation constant, Ki, it was possible to closely fit the experimentally derived amylase release doseresponse curve (Fig. 10A). While regulation of sugar and amino acid transport followed a dose-response consistent with the occupancy of a single order of binding sites, there was a slight difference in sensitivity for the two functions. By developing an equation for the effects of spare receptors or, alternatively, spare second messenger, the data for these parameters could also be fit (Fig. IOB). Thus this model is consistent with the concept that the major role of the high affinity receptor is to stimulate amylase release and the major role of the low affinity receptor is to inactivate the effects of the high affinity receptor. Moreover, another role of the low affinity receptor is to regulate glucose and amino acid transport. Furthermore, the low affinity receptor may also mediate the inhibition of the synthesis of pancreatic protein that is observed at high con-

FIG. 1 1 . Autoradiogram of a SDS-polyacrylamide gel of mouse pancreatic plasma membranes crosslinked to l2Wabeled BH-CCK with dissuccinimidyl suberate. Lane A: lZWabeled BH-CCK alone; Lane B: 12Wabeled BH-CCK plus dithiothreitol; Lane C: 12SI-labeled BH-CCK plus unlabeled CCKs; Lane D: I2Wabeled BH-CCK plus dithiothreitol plus unlabeled CCKs.

16

IRA D. GOLDPINE AND JOHN A. WILLIAMS

centrations of CCK (Fig. 4). It should be noted, however, that this type of deduction cannot tell us whether the two classes of CCK receptors are either distinct molecular entities or whether they are two states of the same receptor. To approach the molecular nature of the CCK receptor, we have studied the CCK receptor present in purified pancreatic plasma membranes. These CCK receptors show an analog specificity parallel to those of acini, but Scatchard plots reveal only a single class of binding sites with an affinity of 1.8 nM (Sakamoto er al., 1983). The receptor appears to be protein in nature as pretreatment of the membranes with trypsin abolishes binding of 1251-labeledBH-CCK (Sakamoto et al., 1983). To further characterize the receptor, we covalently crosslinked it to its receptors with dissucinimidyl suberate. After the plasma membranes were solubilized, reduced, and subjected to polyacrylamide gel electrophoresis, a major labeled protein was observed with M, of about 85,000 (Fig. 11). Several minor bands were present which may have represented partially degraded receptor. The appearance of both major and minor bands were eliminated by the inclusion of 100 nM CCK, during the 1251-labeledCCK binding. These data suggest that the pancreatic plasma membrane CCK receptor is a protein and has a molecular weight of around 80,000. Two other groups have also reached similar conclusions concerning the size of the reduced CCK receptor (Rosenzweig t?r al., 1982; Svoboda et al., 1982). It is clear that further studies of this type should lead to a more physical model of the CCK receptor based on its physicochemical as well as binding properties.

D. AUTORADIOGRAPHS OF I2’1-LABELED CCK Recently it has been possible to study the location of hormones in target cells with quantitative electron microscope autoradiographs. Such studies have been carried out for the polypeptide hormones insulin, growth hormone, glucagon, and parathyroid hormone (Barazzone et al., 1980a,b; Bergeron et al., 1979; Goldfine ef al., 1978, 1981a; Nordquist and Palmieri, 1974). In general, these hormones initially bind to the plasma membrane and then become internalized and associated with specific organelles. While the location over the plasma membrane suggests that this organelle is a site of hormone action, localization over intracellular organelles has also raised the possibility of intracellular loci of hormone action and/or degradation (Goldfine et al., 1981a; Gorden et al., 1980; Pastan and Willingham, 1981). However, no autoradiographic studies of cholecystokinin had previously been carried out. Accordingly, we used quantitative electron microscopic autoradiographyof 1251-labeledBH-CCK in isolated mouse pancreatic acini to localize the initial site of interaction and the subsequent fate of this ligand (Williams et al., 1982b). Acini were incubated with 0.5 nM 1251-labeledBH-CCK for 2 and 30 minutes and the distribution of 1251 grains over the various cellular organelles was deter-

17

INSULIN AND CCK IN THE ACINAR PANCREAS TABLE 1V OBSERVED AND THEORETICAL GRAINDISTRIBUTION I N ACINIINCUBATEDWITH '251-LABELED BH-CCK'*b Grain distribution (% total) Observed Organelle

2 minutes

30 minutes

Theoretical

Plasma membrane Mitochondria Zymogen granules Golgi Multivesicular bodies Endoplasmic reticulum Nucleus Smooth vesicles

64.4c 2.4 2.4 0.5 0 29.1

31.Ic 2.9 4. I 4.2 2.6' 49.9 3.8 1.3

9. I 5.9 4.2 3.9 0 73.6 4.8 0.2

1.1

0

"From Williams et al. (1982b). bPreparation of selections for autoradiography . Acini were incubated in Hepes Ringer buffer with 0.5 nM '*SI-labeled BH-CCK and the pellets were washed and fixed in glutaraldehyde and paraformaldehyde. The fixed acini were postfixed in unbuffered I % osmium tetroxide containing 1.5% potassium ferrocyanide (Kamovsky, 1971). dehydrated, and embedded in Epon 812 epoxy resin. Gold sections were coated with a monolayer of Ilford L-4 emulsion. The organelle distribution of the grains was determined using a circle analysis (Goldfine er al., 1981b; Williams, 1969). A transparent overlay, having a circle with a radius equal to the resolution half-distance (0.085 pm), was placed over the center of each grain and the underlying organelles recorded. When grains were localized over two or more organelles, credit was equally divided. Any circle touching the plasma membrane, however, was scored solely as plasma membrane. The theoretical random distribution of grains over the organelles in the same micrographs was then determined using a plastic overlay with randomly placed circles of the same radius (Williams, 1969). Three hundred and seventy observed grains at 2 minutes, 425 observed grains at 30 minutes, and 1557 theoretical grains were analyzed. cHormone concentration (observedltheoretical grains > I .5).

mined. In acini incubated 2 min, 64% of the observed grains were localized over the plasma membrane (Table IV, Fig. 12); all grains were over the basolateral plasma membrane and no grains were observed over the luminal plasma membrane. The plasma membrane over which the grains were localized did not have areas of specialization such as coated pits. The remainder of the grains were located over the endoplasmic reticulum, particularly near the basolateral plasma membrane. The micrographs were then analyzed by the random circle technique to obtain a theoretical grain distribution and this theoretical distribution was then compared to the observed distribution (Williams, 1969). With this analysis at 2 minutes, seven times as many grains were found over the plasma membrane than would be expected on a random basis; at this time, no other cellular organelle

18

IRA D. GOLDFINE AND JOHN A. WILLIAMS

FIG. 12. Electron microscope autoradiographs of pancreatic acini incubated with 1251-labeled

BH-CCK for 2 or 30 minutes showing localization of silver grains over typical organelles. Final magnification X40,OOO. (From Williams er nl., 1982b.)

showed a similar concentration of hormone. After a 30-minute incubation of acini with *251-labeledBH-CCK, the percentage of grains over the plasma membrane had fallen to 31% of total, but this value was 4 times that expected on a random basis (Table IV). At 30 minutes, the percentage of grains over other cellular organelles, especially the endoplasmic reticulum, had increased, but concentration of CCK was observed only over multivesicular bodies. To study the amount of CCK that had fixed to acinar cells, the acini were incubated for either 2 or 30 minutes with 0.5 nM 1251-labeledBH-CCK and then washed, fixed, and processed for electron microscopy. After fixation, postfixation and dehydration in ethanol, 84% of the radioactivity bound during a 2minute incubation remained with the acini. After a 30-minute incubation, the

19

INSULIN AND CCK IN THE ACINAR PANCREAS

TABLE V RADIOACTIVITY REMAINING AS INTACTCCK VERSUS PERCENTAGE REMAINING WITH TISSUE AFTER PROCESSING FOR ELECTRONMICROSCOPY lZSI-labeled BH-CCK Radioactivity (Q total)

Incubation (minutes)

Remaining intact

2 30

90 65

* 2 (3) * 1 (5)

Remaining with processed tissue 84 f 2 (3) 78 f 3 ( 6 )

a1251-labeledBH-CCK, 0.5 nM. was incubated with pancreatic acini for 2 and 30 minutes, and the amount of radioactivity bound to acini was measured. For each experiment, one portion of the preparation was processed for electron microscopy and the amount of radioactivity remaining with the tissue calculated. Another portion was extracted and the 1251 radioactivity analyzed by gel filtration to determine the percentage of radioactivity remaining as intact IZ5I-labeled BH-CCK. All values are the mean f SE for the number of experiments shown in parentheses (from Williams era/., 1982b).

percentage of bound radioactivity fixing to acini had decreased to 78% (Table V). After a 2-minute incubation, the fraction of radioactivity fixing to acini correlated well with the fraction remaining as intact CCK (Table V). After 30 minutes, slightly more radioactivity fixed to the acini than remained intact, as determined by gel filtration (Table V). This autoradiograph study, demonstrating the concentration of CCK grains over the plasma membrane, is in concert with the concept that this organelle is the major site of CCK action. In particular, the effects of CCK to both increase Ca2+ uptake into acini and to stimulate the release of Ca2+ from the plasma membrane most likely are a direct result of CCK binding to this organelle and reflect the very rapid effects of the hormone. However, we and others have shown that CCK also mobilizes Ca2 from the endoplasmic reticulum and other intracellular organelles (Dormer and Williams, 1981). Thus, it is possible that the presence of CCK in the cell interior could reflect the regulation of the endoplamic reticulum and other organelles by CCK. This observation could explain how CCK exerts long-term effects on pancreatic acini, such as those on protein synthesis (Korc et al., 1981a). The exact site in the endoplasmic reticulum where CCK acts is, however, unknown. +

IV. Insulin A. BACKGROUND The mammalian pancreas contains clusters of heterogenous endocrine cells, the islets of Langerhans, whose function is to secrete insulin, glucagon, and

20

IRA D. GOLDFINE AND JOHN A. WILLIAMS

other peptides into the blood. These islets are unevenly distributed throughout the pancreatic exocrine tissue that secretes both digestive enzymes and bicarbonate-rich fluid into the intestine via the pancreatic duct. This unique anatomical arrangement is also found in birds, but is either less extensively developed or is missing altogether in lower vertebrates where endocrine cells are clustered as a separate gland. In the most primitive vertebrates, these particular endocrine cells are found only in the mucosa of the gastrointestinal tract (Van Noorden and Falkmer, 1980). The reasons for the evolutionary migration of endocrine cells from the mtestine into numerous islets are unclear. It has been suggested, however, that the formation of islets may allow for hormonal interaction between the endocrine cells, while the dispersal of these islets within the pancreas allows for an interaction between islet and acinar cells. The possibility that hormone-secreting islet cells may interact with their neighboring exocrine cells and form an islet-acinar axis has thus given rise to the hypothesis that islet cell hormones may regulate the exocrine function of the pancreas (Henderson, 1969). There is histological evidence for the existence of a capillary portal circulation within the exocrine pancreas (Fujita and Murakami, 1973). Measurements of blood flow have shown that about 80% of the pancreatic arterial blood flows directly to the exocrine pancreas while 20% flows to the islets (Lifson et al., 1980). Venous blood from the islets then enters capillaries surrounding the acinar cells prior to returning to the heart via the systemic veins. This portal system thus carries hormone-rich blood from the islets directly to the acinar cells (Fig. 13), creating a milieu in which the levels of islet hormones are considerably higher than levels found in the peripheral circulation. These high hormone levels within the pancreas could exert trophic effects on the acinar cells, and thus explain the longstanding observation that acinar cells

FIG. 13. Schematic representationof intrapancreatic blood flow. Intralobular arterial blood supply is depicted by solid arrows. The arterial vessel entering directly into an islet drains via multiple branches into capillaries surrounding the neighboring acini. Venous drainage is depicted by open arrows.

INSULIN AND CCK IN THE ACINAR PANCREAS

21

adjacent to islets are both larger and richer in zymogen granules than other acini. Histochemical staining for digestive enzymes reveals these larger acini as dark rings or “haloes” surrounding the paler islets. The experimental basis for the existence of an islet-acinar axis is strongest for insulin, and was initially deduced from dietary studies. Grossman (Grossman et al., 1942), along with other investigators, showed that a glucose-rich diet was effective in inducing an increase in pancreatic amylase content. Similar results for both enhanced protein synthesis and raised amylase content are observed following the refeeding of starved animals. B. EFFECTSON ACINARCELLS 1. In Vivo Studies

Diabetes mellitus in man and animals may be associated with disturbances in pancreatic exocrine function, including decreased secretion of bicarbonate-rich fluid and decreased enzyme output. Although clinically overt pancreatic insufficiency is not a common clinical manifestation of diabetes, several studies have demonstrated that many insulin-dependent diabetic patients have abnormal pancreatic responses to secretin and cholecystokinin, the hormones that regulate the secretion of pancreatic fluid and enzymes (Chey et al., 1964; Domschke er al., 1975). Pathologically, the pancreas of diabetics may be small and the exocrine tissue shows fatty degeneration and fibrosis. Animals that develop spontaneous diabetes also have similar abnormalities (Balk et al., 1975). Three observations support the hypothesis that the deficiency of insulin is responsible for these abnormalities. First, pancreatic exocrine dysfunction is found only in insulinopenic diabetes (Peters er al., 1966). Second, the extent to which pancreatic exocrine function is preserved in insulin-dependent diabetic patients correlates directly with the persistence of f3 cell secretory activity, measured by C-peptide levels (Frier et d., 1978). This peptide is secreted with insulin and its plasma level can be used, therefore, as an index of endogenous insulin secretion. Third, in experimental animals, the induction of diabetes by p cell toxins, such as alloxan or streptozotocin, results in a time-dependent fall in pancreatic amylase levels that can be reversed by the in vivo administration of insulin (Soling and Unger, 1972). We have also confirmed that streptozotocin-induced diabetes is associated with a fall in pancreatic amylase levels, and that this decrease can be prevented by the in vivo administration of insulin. Moreover, we have shown that the marked fall in pancreatic amylase levels was associated with only a mild decrease in parotid gland amylase levels (Fig. 14), indicating that the synthesis of pancreatic amylase is considerably more sensitive to insulin than parotid gland amylase. Although parotid gland amylase levels did respond to the in vivo administration of insulin, the response was not of the same magnitude as that

22

IRA D. GOLDFINE AND JOHN A. WILLIAMS

[23 Poncrros

Control

Diabetic

Diobstic

+

Insulin

FIG. 14. Effects of streptozotocin-induced diabetes and insulin administration on rat pancreas and parotid amylase levels. Streptozotocin (75 pg/kg) was injected into the tail vein. Insulin was administered by regular twice-daily intraperitoneal injection (6 U/day) for 8 days, beginning 2 days after the injection of streptozotocin.

zooo-

-z

1000-

m

0-

Insulin I r e a t i d (days)

FIG. 15. Increase in amylase mRNA in pancreas of diabetic rats treated with insulin. mRNA was quantitated by use of a cDNA probe (plotted from data in Korc et a!., 1981~).

23

INSULIN AND CCK IN THE ACINAR PANCREAS

observed in the pancreas. Recently studies with diabetic rats have indicated that these animals have very low levels of mRNA for amylase, and that the mRNA levels are restored by insulin administration (Korc et al., 1981c) (Fig. 15). 2. In Vitro Studies Earlier studies employing pancreatic fragments either have failed to detect an effect of insulin on the pancreas (Couture et al., 1972) or have found that very high concentrations (>10 p M insulin) were needed (Danielsson and Sehlin, 1974). Recently, employing isolated pancreatic acini from streptozotocin-induced diabetic rats and mice, we have found that insulin, at concentrations of 1 nM or less, stimulated both sugar transport and protein sythesis (Korc et al., 1981b; Williams, 1980) (Fig. 16). Moreover, in isolated acini from such animals, insulin increases net amylase content (Korc et al., 1981b). The use of acini from diabetic animals eliminates endogenous insulin produced by the few islet cells that are present in the isolated acinar preparation.

g

120

N

f

loo ~

'

0

"""I p ' ' """I 1010 100

' '

"""I 108

' '

' '

"""I

107

Insulin (M)

I+/ ' ' ""'I

1010

' ' ' ""'I

' ' ' ""'I

100

108

' '

' ""'I

'

'

107

Insulin (M)

FIG. 16. Concentration dependence for insulin stimulation of [3H]2-DG uptake (A) and incorporation of [3H]leucine into protein (B) by diabetic mouse pancreatic acini. All values are the mean -C SE of 4 or 5 experiments. (From Williams et al., 1982a.)

24

IRA D. GOLDFINE AND JOHN A. WILLIAMS

FIG.17. Autoradiographs of soluble (A) and particular (B) fractions from acini incubated with (+) or without (-) 0.5 phf insulin for 30 minutes. Arrows indicate bands altered by insulin. (From

Bumham and Williams, 1982.)

Insulin is believed to carry out certain functions by phosphorylation and dephosphorylation reactions. When insulin was added to isolated mouse pancreatic acini, it increased the phosphorylation of three proteins with M, of 16,000, 23,000, and 32,500, respectively (Burnham and Williams, 1982) (Fig. 17).

C. CHARACTERISTICS OF INSULIN RECEPTORS While earlier studies reported effects of insulin on the exocrine pancreas in vivo, it had been previously difficult to study effects of insulin on the exocrine pancreas in vifro. Accordingly, to investigate both insulin receptors and insulin action in the pancreas, we studied isolated pancreatic acini. Insulin receptors were studied using biologically active porcine '251-labeled insulin prepared by the stoichiometric chloramine-T method (Goldfine and Smith, 1976). When

t!

ij\

; E

!L

.-

-2 Y

tn

.\;

0.01

N l-

w

0 0

10

20

30

25 I -Insulin bound (fmol/mg protein)

FIG. 18. Scatchard plot of specific binding of 1251-labeledinsulin to mouse pancreatic acini.

FIG. 19. Autoradiogram of an SDS-polyacrylamide gel of mouse pancreatic plasma membranes crosslinked to 12Tlabeled insulin with dissuccinimidyl suberate. 12sl-labeledinsulin alone (-), 125Ilabeled insulin + 10 p.M unlabeled insulin (+).

26

IRA D. GOLDFINE AND JOHN A. WILLIAMS

mouse acini were incubated with 167 pM 1251-labeledinsulin at 37"C, binding was one-half maximal after 10 minutes of incubation and maximal after 30 minutes of incubation. Scatchard plots of insulin binding, under conditions where insulin degradation was minimal, were linear and could be resolved into a single high affinity site with a Kd of approximately 1.0 nM (Fig. 18). Subsequently, we carried out experiments with pancreatic acini obtained from rats. L251-labeledinsulin binding to receptors in acini from normal and diabetic rats was similar to that seen with acini from normal mice. We then investigated the effects of insulin on [3H]leucineincorporation into protein (Jefferson, 1980) and correlated this biological function with insulin binding. In acini from diabetic rats, the effect of insulin was one-half maximal at 0.6 nM and maximal at 5 nM. In these acini from diabetic rats, [3H]leucineincorporation by insulin correlated closely with occupancy of high affinity insulin binding sites (Sankaran et af., 1981b). In liver and other tissues, the insulin receptor is believed to be an oligomeric protein with an M,of 350,000 which upon reduction is separated into two pairs of identical subunits of 130,000 and 90,000 (Czech, 1981). To study the pancreas insulin receptor, we first bound L251-labeledinsulin to purified pancreatic membranes and then crosslinked it with dissuccinimidyl suberate., After solubilization, reduction, polyacrylamide gel electrophoresis, and autoradiography, two subunits were seen with M,of 130,000 and 90,000 (Fig. 19).

D. AUTORADIOGRAPHS OF 251-LABELEDINSULIN Insulin regulates cellular functions in the exocrine pancreas (Kanno and Saito, 1976; Korc et al., 1978, 1981b; Saito et al., 1980). In turn, the pancreas degrades insulin (Sankaran et al., 1981b). However, the cellular sites for these processes are unknown. Recently, several electron microscope autoradiographic studies of liver and lymphocytes (Bergeron et af., 1979; Carpentier et al., 1979; Goldfine et al., 1978; Gorden et al., 1978; Renston et af.,1980) have indicated that insulin initially binds to receptors on the cell surface, enters the interior of cells, and then interacts with several intracellular structures (Carpentier et af., 1978; Goldfine and Smith, 1976; Goldfine et af., 1977; Posner et af., 1978). These findings have suggested the possibility, therefore, that insulin may either act on or be degraded in the cell interior (Goldfine, 1977; Steiner, 1976). In order to better understand the relationship between the intracellulardistribution of insulin and its action and/or degradation, we have quantitatively examined electron microscope autoradiographs of isolated pancreatic acini. These acini were incubated for 3 and 30 minutes with 1251-labeledinsulin in order to determine the initial site of insulin-acinar cell interaction and whether the hormone was translocated into the cell interior. Studies were first undertaken to determine whether the distribution of 1251-labeledinsulin grains among the vari-

27

INSULIN AND CCK IN THE ACINAR PANCREAS

ous cellular organelles was a random process. Using a circle with a radius of onehalf distance, the organelles underlying the grains were noted (Table VI). Next, using the same micrographs and circles of the same radius, a theoretical random grain distribution was determined (Williams, 1969). Chi-square analyses of the observed versus the theoretical grain distribution patterns indicated that at 3 and 30 minutes the observed grain distributions were nonrandom (P
3 minutes

30 minutes

Theoretical

Plasma membrane Rough endoplasmic reticulum 100 nm vesicles Zymogen granules Golgi Mitochondria Nuclei Multivesicular bodies and autophagic vacules Other

45.w 29.7 I 1 .9c 2.4 2.4c 5.8 0.8 1.4

19.7 37.5 15.& 10.0 7.7c 3.8 2.2 2.3

8.5 50.4 5.8 7.7 1.1 12.5 7.0

0.6

1.2

5.9

1.o

OFrom Goldfine er al. (1981b). this study, 915 theoretical grains and 3 16 observed grains at 3 minutes and 921 observed grains at 30 minutes were analyzed. =Hormone concentration (observed/theoreticaI grain > 1 S).

28

IRA D. GOLDFINE AND JOHN A. WILLIAMS

FIG.20. Electron microscope autoradiographs of pancreatic acini incubated with 1251-labeled insulin for 2 or 30 minutes showing localization of silver grains over typical organelles. Final magnification X40,OOO. (From Goldfine er al., 1981b.)

a random process. Compared with a theoretical random grain distribution, there was a relative concentration of lZ5I-labeledinsulin grains in vesicles with an average diameter of 100 nm at both 3 and 30 minutes of incubation. Presumably these vesicles originated from the cell surface. Prior studies in hepatocytes and fibroblasts suggested that insulin can be internalized from the cell surface via endocytotic vesicles (Bergeron et al., 1979; Carpentier et al., 1978, 1979; Gorden et al., 1978; Renston et al., 1980; Schlessinger et al., 1978). In an in vivo study with rat liver, we injected both lZ5I-labeledinsulin and horseradish peroxidase (a marker of endocytosis) into the portal vein and concomitantly localized the peroxidase by histochemical staining and the 1z51-labeledinsulin by autoradiography (Renston et al., 1980). We were able to calculate that approximately 30% of the cellular 125I-labeled insulin grains were present in 100 nm

INSULIN AND CCK IN THE ACINAR PANCREAS

29

endocytotic vesicles. These vesicles were then observed either in the Golgi area or fused with the bile canalicular membrane (Renston et al., 1980). Two other electron microscope autoradiographic studies of I2’1-labeled insulin in liver have also suggested that a fraction of the internalized 1251-labeledinsulin is present in small vesicles (Bergeron et al., 1979; Carpentier et al., 1979). In contrast to the studies in liver and pancreas, an in vitro electron microscopic study of 1251labeled insulin entry into human cultured lymphocytes demonstrated internalization but did not detect I2%labeled insulin grains in vesicular structures (Goldfine et al., 1977). These findings suggest, therefore, that vesicular transport is one mechanism whereby insulin is internalized into certain cell types, but in addition there may be other mechanisms through which the hormone is internalized. Grains representing 1251-labeledinsulin were also concentrated in the Golgi of pancreatic acini. We and others have observed that 1251-labeledinsulin grains accumulate in the Golgi of hepatocytes in vivo (Bergeron et al., 1979; Carpentier et a [ . , 1979; Renston et al., 1980). In pancreatic acini, one function of the Golgi is to concentrate and package newly synthesized proteins (Palade, 1975). In pancreatic acini isolated from diabetic rats, we find that although incubation with insulin increases the synthesis of amylase, but does not directly stimulate amylase release from the cell. However, incubation of these acini with insulin does potentiate amylase release evoked by cholecystokinin (Otsuki and Williams, 1983). These data, therefore, raise the possibility that the Golgi could be an intracellular regulatory site of insulin action. Alternatively, the interaction of insulin with the Golgi may represent either a step in a pathway of insulin degradation (Renston et al., 1980) or the recycling of insulin and its receptor to the cell surface.

V. Interactions between Insulin and CCK There is evidence that insulin and CCK act in concert on the exocrine pancreas. In the perfused rat pancreas, both exogenous insulin and endogenous insulin (released in response to glucose) potentiate the effect of CCK to release zymogen and juice (Saito er al., 1980). To further evaluate the influence of insulin on the regulation of secretion by CCK, we studied isolated pancreatic acini prepared from streptozotocin-induced diabetic rats (Otsuki and Williams, 1982a). In these studies, the content of two digestive enzymes, amylase and ribonuclease, the effect of secretagogues on their release, and the intracellular transport of newly synthesized proteins were examined. Two defects were observed in diabetes (Ostuki and Williams, 1982a). First, acini from diabetic rats contained less than 1% of the normal content of amylase, but 50% of the normal content of ribonuclease (Figs. 21 and 22). Furthermore, reduced amounts of both enzymes were secreted by diabetic acini in response to both CCK and the cho-

30

IRA D. GOLDFINE AND JOHN A. WILLIAMS 20000

A

10000

Normal -+*

t g!

1000

h

500

0 10.12

10.10

10-8

CCKg (MI

0

10-12

10.10

10.8

CCK8 (M)

FIG.21. Concentration dependence of amylase release stimulated by CCK8 from normal and streptozotocin-induceddiabetic rat acini. Amylase release over 30 minutes is plotted as a function of the concentration of CCKs in the medium both relative to acinar protein (A) and as a percentage of total amylase activity (B). (From Otsuki and Williams, 1982a.)

linergic agent, carbamylcholine. Second, the sensitivity to CCK but not to cholinergic agents was altered in acini from diabetic rats (Figs. 21 and 22). In acini from normal rats, the effect of CCK was maximal at a concentration of 100 pM; higher concentrations led to a submaximal enzyme release. In acini from diabetic rats, the dose-response curve was similarly shaped but shifted 3-fold toward higher concentrations of CCK. Also the effect of CCK on the mobilization of cellular Ca2 was selectively shifted toward higher hormone concentrations. In contrast to these results with CCK, the dose-response curve for carbamylcholine was unaltered by diabetes (Otsuki and Williams, 1982a). Control experiments were carried out to confirm that the effects observed were due to insulin deficiency and not due to a direct toxic effect of streptozotocin on +

'"1

A

0

(0.1210.11 1010 10.9 10.1)

CCKa (MI

a 10.1z 10.11 10-10 1 0 0 1 0 4

CCKa (MI

FIG. 22. Concentration dependence of ribonuclease release stimulated by CCKa from normal and streptozotocin-induced diabetic rat acini. Ribonuclease release over 30 minutes is plotted as a function of the concentration of CCKg in the medium relative to the acinar protein concentration (A) and as a percentage of the total ribonuclease (B). (From Otsuki and Williams, 1982a.)

INSULIN AND CCK IN THE ACINAR PANCREAS

31

acinar cells. First, streptozotocin treatment had no effect on acini when measured 24 hours after administration. Second, alloxan, another p cell toxin, induced similar changes in acinar enzyme content and secretory responses. Third, the administration of exogenous insulin to diabetic rats returned the content of pancreatic amylase and the secretory response to CCK towards normal (Fig. 23). Moreover the altered pancreatic function in diabetes did not appear due to malnutrition since 48 hours of fasting, although inducing a significant weight loss, did not mimic the effects of diabetes. Since pancreatic amylase content in diabetic rats is very low and levels can be restored to normal by replacement doses of insulin, we studied the effect of insulin in vitro. We found that treatment of these acini with insulin in v i m did not directly increase amylase secretion, but after a 2-hour preincubation period insulin potentiated the secretory effects of CCK and its analog caerulein (Otsuki and Williams, 1983; Williams et al., 1981) (Fig. 24). These studies demonstrate, therefore, that two major abnormalities in pancreatic exocrine secretion exist in the diabetic rat due to insulin lack. First, the content of certain digestive enzymes is markedly altered, leading to an altered amount of zymogen secretion. Second, the sensitivity to the secretagogue CCK is selectively reduced, most likely related to a defect in receptor-activated transmembrane signaling. These findings also raise the possibility that certain gas-

.--. Normal

-Diabetic

+ ljlsulin

CCK8

(M)

FIG. 23. Concentration dependence of amylase release stimulated by CCKs from normal and insulin-treated diabetic rat acini. Amylase release over 30 minutes expressed as a percentage of the total amylase activity initially present in the acini, is plotted as a function of the added concentration of CCKs. (From Otsuki and Williams, 1982a.)

32

IRA D. GOLDFINE AND JOHN A. WILLIAMS 25

-

25r

.m(hHlu*n

0 mhoul inWk

CCKe

(MI

FIG.24. Effect of a 2-hour pretreatment with 100 nM insulin on subsequent stimulation of amylase and ribonuclease release by CCKs. (From Otsuki and Williams, 1983.)

trointestinal abnormalities occurring in diabetes mellitus may be a direct result of insulinopenia. In addition, since both cholecystokinin and insulin stimulate protein synthesis in isolated acini, the effects of both were tested in acini from diabetic rats. The major effects of insulin were to enhance the action of CCK at all hormone concentrations (Korc et al., 1981a) (Fig. 3). In addition to synergistic effects on protein synthesis, both CCK and insulin stimulate glucose uptake into acinar cells (Table VII). Each hormone, however, appears to use a different intracellular mechanism. CCK acts via enhanced Ca2 mobilization whereas insulin has no effect on this function (Fig. 25). Since both insulin and CCK regulate protein phosphorylation in acini their effects were compared directly (Burnham and Williams, 1982). The effect of both insulin and CCK on the phosphorylation of 23,000 and 32,500 proteins was greater than the effect of either hormone alone (Fig. 26). In contrast, insulin had no effect on the dephosphorylation of the 2 1,000 and 20,500 proteins that were regulated by CCK. It is tempting to speculate, therefore, that the phosphoryliition of 23,000 and 32,500 proteins by insulin and CCK may be involved in their additive effects on glucose transport and protein synthesis, and that the dephosphorylation of the 21 ,OOO and 20,500 proteins may mediate CCK-stimulated zymogen release. +

33

lNSULIN AND CCK IN THE ACINAR PANCREAS

EFFECT

OF

TABLE VII INSULIN AND CCK O N [3H]2-DEOXYGLUCOSEUPTAKE AND INHIBITION BY DIBUTYRYL CYCLIC GMPO

BY

ACINI

[3H]2DG uptake (pYmg proteinl20 minutes)

Control Insulin CCK Insulin t CCK

-dbc GMP

+dbc GMP

3.0 f 0.3 6.0 f 0.6 8.2 0.6 10.3 f 0.8

3.3 f 0.3 6.3 2 0.3 3.6 ? 0.2 6.2 2 0.9

*

OConcentrations used were insulin, 1 pM: CCK8, 70 pM; and dbc GMP,I mM. Each value is the mean f SE of triplicate determinations (from Williams et al., I982a).

1

o

0

t

-

F 10

T

- r 20

l

30

40

Minutes

FIG. 25. Effect of insulin and CCK on 45Ca2+ efflux from prelabeled mouse pancreatic acini. (From Williams er a / . , 1982a.)

34

IRA D. GOLDFINE AND JOHN A. WILLIAMS

200

-

160

-

c

1

100 -

i B-

E

M)-

OL

32.5 K-P

23 K-S

260 -

11

CCh CCK I CCh CCK +I +I

1

CCh

CC

+I

+

21 K-S

C

!I

iln lM)

60

"

CCh CCK I CCh CCK +I +I

CChCCKICQICCK

+ I +I

FIG.26. Effect of insulin in the presence or absence of carbachol or CCKs on protein phosphorylation in acini from diabetic mice (Bumham and Williams, 1982).

VI. Summary and Conclusions These studies, therefore, allow a model of how CCK and insulin regulate the acinar pancreas in a coordinated manner (Fig. 27). CCK, after its secretion by gut cells, interacts with a specific receptor on the cell surface and then increases intracellular free Ca2+. Ca2+, in turn (1) interacts with the secretory granules leading to zymogen release, (2) stimulates protein synthesis, and (3) increases glucose transport. The model is supported on the finding of specific high affinity CCK receptors on acini and by the localization of CCK to the plasma membrane in EM autoradiographs. Insulin, secreted from the pancreatic islets, also interacts with a specific receptor on the cell surface. Either via a messenger generated by this reaction, or via insulin's subsequent direct interaction with intracellular organelles, such as the Golgi-endoplasmic reticulum, protein synthesis is initiated and glucose transport is increased. Then a series of events is initiated to increase cell growth,

INSULIN AND CCK IN THE ACINAR PANCREAS

35

FIG. 27. Schematic diagram of how insulin and CCK influence the exocrine pancreas. CCK, via Ca2+, acts to increase zymogen secretion directly. Insulin acting via an unknown mechanism increases protein synthesis to both change the acinar zymogen content and potentiate the actions of CCK. In addition, both hormones increase glucose transport.

amylase content, and sensitivity to CCK. These studies, therefore, indicate that the control of acinar cell function is a product of cooperative intrahormonal interactions.

ACKNOWLEDGMENTS Research reviewed here was supported by NIH Grants AM 21089, AM 26422, AM 26667, AM 29971, GM 19998. and the Elise Stern Haas Research Fund, Harold BNM Institute, Mount Zion Hospital and Medical Center.

REFERENCES Amsterdam, A.. and Jamieson, J. D. (1972). Proc. Natl. Acad. Sci. U.S.A. 69, 3028-3032. Amsterdam, A., and Jamieson, J. D. (1974). J . Cell Eiol. 63, 1037-1056. Anastasi, A., Erspamer, V., and Endean, R. (1%8). Arch. Eiochem. Biophys. 125, 57-68. Balk, M. W., Lang, C. M., White, W. J., and Munger, B. L. (1975). Lab. Invest. 32, 28-32. Barazzone, P., Gorden, P., Carpentier, J.-L., Orci, L., Freychet, P., and Canivet, B. (198Oa). J . Clin. Invest. 66, 1081-1093. Barazzone, P., Lesniak, M. A., Gorden, P., Van Obberghen. E.,Carpentier. J.-L., and h i , L. (1980b). J . Cell Eiol. 87, 360-369. Bergeron, J., Sikstrom, R., Hand, A., and Posner, B. I. (1979). J. Cell Eiol. 80, 427-443. Burnham, D. B., and Williams, J. A. (1982). J. Eiol. Chem. 257, 10523-10528. Carpentier, J.-L., Gorden. P., Amherdt, M., VanObberghen, E., Kahn, C. R. and Orci, L. (1978). J . Clin. Invest. 61, 1057-1070. Carpentier. J.-L., Gorden, P., Barazzone, P., Freychet, P., Le Cam, A., a n d h i , L. (1979). Proc. Natl. Acad. Sci. U S A . 76, 2803-2807. Case, R. M. (1978). Eiol. Rev. 53, 211-354.

36

IRA D. GOLDFINE AND JOHN A. WILLIAMS

Case, R. M. (1979).I n “Scientific Basic of Gastroenterology” (H.L. Duthie and K. G.Wornisley, eds.), pp. 163-198. Churchill Livingston, Edinburgh. Chandler, D. E., and Williams, J. A. (1978). J . Cell Biol. 76, 386-399. Chey, W. Y., Shay, H., and Shuman, C. R. (1964). Ann. In?. Med. 59, 812-815. Christophe, 1. P., Conlon, T. P., and Gardner. J. D. (1976a). J . Biol. Chem. 251, 4629-4634. Christophe, J. P., Frandsen, E. K., Conlon, T. P.,Krishna, G . , and Gardner, J. D. (1976b)J . Biol. Chem. 251,4640-4645. Couture, Y.. Dunningan, J., and Morisset, J. (1972). Scand. J . Gasrroenrerol. 7, 257-263. Czech, M. P. (1981). Am. J . Med. 70, 142-150. Dagorn, J.-C., and Mongeau, R. (1977). Biochim. Biophys. Acra 498, 76-82. Danielsson, A., and Sehlin, J. (1974). Acra Physiol. Scand. 91, 557-565. De Meyts, P., Roth, J., Neville, D. M., Jr., Gavin, J. R., and Lesniak, M. (1973). Biochem. Biophys. Res. Commun. 55, 154-161. Dickman, S. R.,and Momll, G. A. (1957). Am. J . Physiol. 190, 403-407. Domschke, W.,Tympner, F., Domschke, S., and Dernling, L. (1975). Digesr. Dis. 20, 309-315. Dormer, R. L., and Williams J. A. (1981). Am. J . Physiol. 240, G130-GI40. Frier, B. M., Faber, 0. K., Binder, C., and Elliott, H. L. (1978). Diaberologia 14, 301-306. Fujita, T., and Murakami, T. (1973). Arch. Hisrol. Jpn. 35, 255-267. Gardner, I. D., Conlon, T. P., Klaeveman, H. L., Adams, T. D., and Ondetti, M. A. (1975). J . Clin. Invest. 56, 366-375. Goldfine, I. D. (1977). Diaberes 26, 148-155. Goldfine, I. D., and Smith, G. J. (1976). Proc. Natl. Acad. Sci. U.S.A. 73, 1427-1431. Goldfine, I. D., Smith, G., Wong, K. Y.,and Jones, A. L. (1977). Proc. Narl. Acud. Sci. U.S.A. 74, 1368-1372. Goldfine, I. D., Jones, A. L., Hradek, G. T., Wong, K. Y., and Mooney, J. S.(1978). Scienct. 202, 760-763. Goldfine, 1. D., Kriz, B.M., Wong, K. Y., Jones, A. L., Renston, R.,and Hradek, G. T. (1981a). In “Receptor-Mediated Binding and Internalization of Toxins and Hormones” (J. L. Middlebrook, ed.),pp. 233-247. Academic Press, New York. Goldfine, I. D., Kriz, B. M., Wong, K, Y., Hradek, G . . Jones, A. L., and Williams, I. A. (1981b). Am. J. Physiol. 240, (36947% Gorden, P.. Carpentier. J.-L., Freychet, P., Le Cam, A., and Orci, L. (1978). Science u10, 782-785. Gorden, P., Carpentier, J.-L., Freychet, P., and Orci, L. (1980). J. Hisrochem. Cyrochem. 28, 811-817. Gmssman, M.I., Greengard, H.. and Ivy, A. C. (1942). Am. J. Physiol. 138, 676-682. Henderson, J. R. (1%9). Lancer 2, 469. Iwamoto, Y.,and Williams J. A. (1980). Am. J. Physiol. 238, G440-G444. Jamieson, J. D., and Palade, G. E. (1971). J. Cell Biol. 48, 503-522. Jefferson, L. S. (1980). Diabetes 29, 487-496. Jensen. R. T., Lemp, G. F., and Gardner, J. D. (1980). Proc. Nurl. Acud. Sci. U.S.A. 77, 2079-2083. Jorpes, J. E., and Mutt, V. (1973). In “Secretin, Cholecystokinin, Pancreozymin and Gastrin” (J. E. Jorpes and V. Mutt, eds.), pp. 1-179. Springer-Verlag. Berlin and New York. Kanno, T. (1972). J . Physiol. (Londoo)226, 353-371. Kanno, T. (1980). Biomed. Res. 1, (Suppl.), 10-16. Kanno, T.,and Saito, A. (1976). J . Physiol. (London)261, 505-521. Karnovsky, M. (1971). Ann. Meet. Am. SOC. Cell Biol., llrh, New Orleans p. 412 (Abstract). Kondo, S., and Schulz, I. (1976). Biochim. Biophys. Acra 419, 76-92. Korc, M., Sankaran, H.,Wong, K. Y.,Williams, J. A., and Goldfine, I. D. (1978). Biochem. Biophys. Res. Comnrun. 84, 293-299.

INSULIN AND CCK IN THE ACINAR PANCREAS

37

Korc, M., Williams, J. A., and Goldfine, I. D. (1979). J. Biol. Chem. 254, 7624-7629. Korc, M., Bailey, A,. and Williams J. A. (1981a). Am. J. Physiol. 241, G116-GI21. Korc. M., Iwamoto, Y.,Sankaran, H., Williams, J. A.. and Goldfine, 1. D. (1981b). Am. J. Physiol. 240, G 5 6 4 6 2 . Korc, M., Owerbach, D., Quinto, C., and Rutter, W. J. (1981~).Science 213, 351-353. Larsson, L.-I., and Rehfeld, J. F. (1977). Nature (London) 269, 335-338. Lifson, N., Kramlinger, K. G., Mayrand, R. R., and Lender, E. J. (1980). Gastroenterology 79, 466-473. Mainz, D. L., Black, 0.. and Webster, P. D. (1973). J. Clin. Invest. 52, 2300-2304. Morley, J. S., Tracy, J. H., and Gregory, R. A. (1965). Nature (London) 207, 1356-1359. Mutt, V., and Jorpes, J. E. (1968). Eur. J. Biochem. 6, 156-162. Mutt, V., and Jorpes, J. E. (1971). Biochem. J . 125, 57p-58p. Nordquist, R. E., and Palmieri, G. M. A. (1974). Endocrinology 95, 229-237. O’Doherty, J., and Stark, R. J. (1982). Am. J. Physiol. 242, G513-G521. Ondetti, M. A., Pluscec, J., Sabo, E. F., Sheehan, J. T., and Williams, N. (1970a). J. Am. Chem. SOC. 92, 195-199. Ondetti, M. A,, Rubin. B., Engel, S. L., Pluscec, J., and Sheehan, J. T. (1970b). Digest. Dis. Sci. 15, 149-156. Otsuki, M., and Williams, J. A. (1982a). C. Clin. Invest. 70, 148-156. Otsuki, M., and Williams, J. A. (1982b). Am. J. Physiol. 243, G2854290. Otsuki, M., and Williams, J. A. (1983). Diabetes 32, 241-246. Palade, G. (1975). Science 189, 347-358. Pastan, 1. H., and Willingham, M. C. (1981). Annu. Rev. Physiol. 43, 239-250. Peikin, S. R., Rottman, A. J., Batzri, S . , and Gardner, J. D. (1978). Am. J. Physiol. 235, E743-E749. Peitkin, S. R., Costenbader, C. L., and Gardner, J. D. (1979). J. Biol. Chem. 254, 5321-5327. Peters, N., Dick, A. P., Hales, C. N., Orrell, D. H.. and Sarner, M. (1966). Gur 7, 277-281. Posner, B. I., Josefsberg, Z., and Bergeron, J. J. M. (1978). J . Biol. Chem. 253, 4067-4073. Putney, J . W., Landis, C. A,, and Van De Walle, C. M. (1980). Pfiigers Arch. 385, 131-136. Rajh, H. M., Smyth, M.I., Renckens, B. A. M., Jansen. J. W. C. M., DePont, J. J. H. H. M., Bonting, S. L., Tesser, G. I., and Nivard, R. J. F. (1980).Biochim. Biophys. Acra 632,386-398. Renston, R. H., Maloney, D. G.,Jones, A. L., Hradek, G. T., Wong, K. Y.,and Goldfine, 1. D. (1980). Gastroenterology 78, 1373-1388. Rosenzweig, S. A., Miller, L. J., and Jamieson, J. D. (1982). Fed. Proc. Fed. Am. SOC.Exp. Biol. 41, 1183. Rothman, S. S. (1975). Science 190, 747-753. Rothman, S . S. (1980). Am. J. Physiol. 238, (33914402. Saito. A,, Williams J. A., and Kanno, T. (1980). J . Clin. Invest. 65, 777-782. Sakamoto, C., Williams, J. A., Wong, K. Y., and Goldfine, I. D. (1983). FEBS Lett. 151,63-66. Salpeter, M., Fertuck, H., and Salpeter, E. (1977). J. Cell B i d . 72, 161-173. Sankaran, H., Deveney, C. W., Goldfine, I. D., and Williams, J. A. (1979). J. Biol. Chem. 254, 9349-935 I . Sankaran, H..Goldfine, I. D., Deveney, C. W.. Wong, K. Y., and Williams, J. A. (1980). J. Biol. Chem. 255, 1849-1853. Sankaran. H., Bailey, A., and Williams, J. A. (1981a). Biochem. Biophys. Res. Commun. 103, 1356- 1362. Sankaran, H., Iwamoto, Y., Korc, M., Williams, J. A., and Goldfine, 1. D. (1981b). Am. J . Physiol. 240, G63-Ci68. Sankaran, H.,Goldfine, I. D., Bailey, A., Licko, V., and Williams, J. A. (1982). Am. J. Physiol. 242, G2504257. Scheele, G . A. (1980). Am. J . Physiol. 238, G4674477.

38

IRA D. GOLDFINE AND JOHN A. WILLIAMS

Scheele, G. A., and Haymovits, A. (1980). J. Biol. Chem. 255, 4918-4927. Schlessinger. J . , Shechter, Y.,Willingham, M., and Pastan, I. (1978). Proc. Nurl. Acud. Sci. U.S.A. 75, 2659-2663. Schultz, G. S . , Sarras, M. P., Gunther, G. R., Hull, B. E., Alicea, H.A., Gorelick, F. S., and Jamieson, J. D. (1980). Exp. Cell Res. 130, 49-62. Schulz, I. (1980). Am. J . Physiol. 239, G335-G347. Soling, H. D., and Unger, K. 0. (1972). Eur. J . Clin. Invest. 2, 199. Solomon, T. H., Petersen, H., Elashoff, J., and Grossman, M. I. (1978). Am. J . Physiol. 235, E7 14-E7 19. Steiner. D. F. (1976). Diubetes 26, 322-340. Svoboda, M.. Lambert, M., Furnelle, J., and Christophe, 1. (1982). Regul. Peprides 4, 163-172. Van Noorden. S., and Falkmer, S. (1980). Invesr. Cell Purhol. 3, 21-28. Vandermeers, A., Vandermeers-Piret, M. C., Rathe, J., Kutzner, R., Delforge, A., and Christophe, J. (1977). Eur. J . Biochem. 81, 377-386. Vigneri, R., Goldfine, I. D., Wong, K. Y.,Smith, G. J., and Pezzino, V. (1978). J. Biol. Chem. 253, 2098-2103. Webster, P. D., Black, 0.. Mainz. D., and Singh, M. (1977). Gastroenterology 73, 1434-1449. Williams, I. A. (1977). Cell Tissue Res. 179, 453-466. Williams, J. A. (1980). Am. J . Pbysiol. 238, G2694279. Williams, J . A., and Lee, M. (1974). Biochem. Biophys. Res. Commun. 60, 542-548. Williams, J . A., Cary, P., and Moffat, B. (1976). Am. J. Physiol. 231, 1562-1567. Williams, J . A., Korc, M., and Dormer. R. L. (1978). Am. J . Physiol. 235, E517-E524. Williams, J. A., Sankaran, H., Korc, M., and Coldfine, 1. D. (1981). Fed. Proc. Fed. Am Soc. Exp. Biol. 40, 2497-2502. Williams, J. A., Bailey, A., Preissler, M., and Goldfine, 1. D. (1982a). Diabetes 31, 674-682. Williams, J . A., Sankaran, H., Roach, E., and Goldfine, I. D. (1982b). Am. J . Physiol. 243, (32914296. Williams, M. A. (1969). Adv. Opt. Electron Microsc. pp. 219-272. Wrenn, R. W., Katoh. N., and Kuo, J. F. (1981). Biochim. Biophys. Actu 676, 266-269.

INTERNATIONAL REVIEW OF CYTOLOOY. VOL. 85

The Involvement of the Intracellular Redox State and pH in the Metabolic Control of Stimulus-Response Coupling ZYGMUND ROTH,~ NAOMICHAYEN,AND SHABTAYDIKSTEIN Unit of Cell Pharmacology, School of Pharmacy, Hebrew University, Jerusalem, Israel Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Stimulus-Response Coupling and the Redox State.. . . . . . . . . . B. The Relationship between Intracellular pH and the Redox State; Scope of the Review ................................... 11. Relationships between the Intracellular Redox State and Agonist Action ................................................... 1.

39 39 41 42 42 43 45

D. Redox Effects Influencing the Transducer Mechanisms, Effector Systems, and Sec Release . . . . . . . . . . . . . . . . .................... E. Redox Effects Influencing F. Redox Effects Influencing ATPases and Recovery . . . . . . . . . . . G. Conclusions and Speculations ............................ 111. The Involvement of Intracellular pH in Stimulus-Response Coupling. .. .......................... A. The Effect of Changes in pHi on Stimulus-Response Events . . B. The Effect of Stimulus-Response Phenomena on pH, . IV. Epilogue ................................................. References .........................

45 52 52 53 55 55 56 57 58

I. Introduction A. STIMULUS-RESPONSE COUPLING AND

THE

REDOX STATE

When a pharmacological agonist interacts with a receptor (target area), usually situated on the outer cell membrane, it sets in motion a chain of events known as the stimulus-response-recovery (SRR) cycle. The cycle starts with a “receptor mechanism”; it consists of those signaling processes that translate the recep‘Present address: Johnson Research Foundation, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104. 39 Copyright 0 1983 by Academic R e s s , Inc. All rights of reproduction in any form reserved. ISBN 0-12-364485-2

40

ZYGMUND ROTH ET AL.

tor-agonist interaction into a change in the intracellular concentration of the second messenger. Whereas formerly researchers only asked the questionWhat is the structure of the receptor? which represents a static approach, in the past decade they have begun asking-How does the receptor mechanism operate? demonstrating a more dynamic approach (Dikstein, 1973). For the purpose of this article, we shall note and briefly define the consecutive stages involved in the operation of the stimulus-response-recovery cycle. For a more detailed treatment the reader is referred to the articles by Robison (1980), Dumont (1980), and Berridge, 1980). Pharmacologists today discriminate between the following consecutive events in the operation of the receptor mechanism (Berridge, 1980). 1. Agonist-receptor interaction refers to the specific binding of an agonist at a “target area” of a receptor molecule, and the structural and functional changes induced thereby in the receptor molecule. 2. Transducer mechanism activation. Transducer mechanisms are activated by agonist-receptor interaction and stimulate or inhibit an effector system. The best known transducers are the GTP regulatory proteins; phospholipase C may also fulfil such a function (Putney, 1979). Agonist-induced changes in cell membrane electric properties should probably also be considered transducer mechanisms in some cases. 3. Effector systems are those which produce or control the cytoplasmic concentrations of second messengers. Examples are adenylate cyclase, guanylate cyclase, and calcium “channels” or “carriers. ” Phosphofructokinasemay also be such an effector (Clark and Patten, 1981).

Following the operation of the receptor mechanism, the cytoplasmic concentration of a second messenger is usually increased. Second messengers (in some cases referred to as “coupling agents”) are metabolites or ions which either directly or through complex formation elicit an intracellular response. Examples of second messengers are CAMP, cGMP, calcium ion (Ca2+), fructose diphosphate (Kirtley and McKay , 1977), prostaglandins, specific peptides, etc. The concept of stimulus-response coupling refers to the processes leading to the formation or release of the coupling agent, and is usually detected by the occurrence of a response. Response is the result of the activation of specific responding structures (e.g., contractile proteins, organelles). Finally, following the response, a recovery process occurs. This usually entails the breakdown or removal of the second messenger (or coupling agent) from the cytoplasm, with a return of the entire system to its original state. In this article we have not differentiated between the various kinds of response accord-

41

STIMULUS-RESPONSE COUPLING AND THE REDOX STATE

ing to their type (e.g., contraction, secretion, cell division, metabolic change, etc.). The possible involvement of the redox pyridine nucleotides in stimulus- response coupling was pointed out in an earlier review (Dikstein and Sulman, 1966). Even this early review suggested that the NADPH/NADP+ couple is more likely to be involved than the NADHINAD couple, since (1) there were some suggestions that -SH groups might play a role, and these are connected via the glutathione redox system to the NADPH/NADP+ couple; and (2) pentose phosphate shunt (hexose monophosphate shunt) activity seemed to be implicated, and the dehydrogenase enzymes of this shunt are dependent upon the NADPH/NADP+ couple (Dikstein, 1970, 1971). The present article will deal, in part, with the question: Can the cytoplasmic redox state, and the enzymes which control it, be involved in stimulus-response coupling? Where possible, we shall attempt to address the question of whether the redox state controlling enzymes and their products are themselves obligatory parts of a receptor mechanism (transducer, effector, second messenger), or whether they act in a regulatory manner by inhibiting or stimulating part of the receptor mechanism. Stated differently, is the redox state an essential link in the sequence of events between stimulus and response, or does it only modify some of those events? +

B. THERELATIONSHIP BETWEEN INTRACELLULAR pH AND STATE;SCOPEOF THE REVIEW

THE

REDOX

Pyridine nucleotide oxidation-reductionis accompanied by the consumption or release of protons, according to the equation: RHz

+ NAD(P)+

R

+ NAD(P)H + H +

Thus, in the absence of buffering, oxidation could lead to an increase in pH and reduction to a decrease in pH. Whether there is actually a significant change in intracellular pH (pH,) will depend upon the number of protons released/bound and upon the degree of pH buffering occurring in a given cell. Conversely, when the concentration of substrates and products of a dehydrogenase reaction are close to the equilibrium state, an increase in pH, would tend to favor pyridine nucleotide reduction whereas a decrease in pH, would favor oxidation, due to the effect of mass action. Thus, besides considering the involvement of the redox state, the present article will also consider the (possibly related) involvement of intracellular pH in stimulus-response coupling. The ATP:ADP system is also related both to the pyridine nucleotide redox state and pHi. Catabolic metabolism requires increase both of the ATP/ADP and of the NAD(P)H/NAD(P) ratios and the hydrolysis of ATP releases protons. It +

42

ZYGMUND ROTH ET AL.

is, however, outside the scope of this article to consider reports of the effects of the ATP:ADP couple on stimulus-response coupling.

11. Relationships between the Intracellular Redox State and

Agonist Action A. APPROACHES AND DEFINITIONS Much of the earlier literature documenting some connection between redox alterations and cell response noted the effect of exogenous oxidizing or reducing agents on the response size or duration following the application of some classical agonist. Only more recently have workers also attempted to measure naturally occurring changes in the redox state itself following agonist stimulation. Furthermore, while the early reports were often satisfied with the demonstration of any sort of effect by redox agents on the response, the realization of the importance of finding the specific stage of stimulus-response coupling which affects or is affected by a redox change has now become more widespread. In the present article we shall first deal with those reports in which the specific stage of the stimulus-response-recovery cycle affected by a redox change has not been defined, and then proceed to reports where there is evidence for a specific stage of involvement. It is hoped that this approach will allow some speculative conclusions to be made with regard to a general description of redox involvement in stimulus-response coupling. Emphasis will be given below to response systems activated by pharmacomechanic (chemopharmacodynamic)coupling rather than those occurring by electromechanic (electropharmacodynamic) coupling. Basically, the difference lies in the fact that electromechanic coupling occurs via a change in potential across the plasma membrane (e.g., an action potential), whereas pharmacomechanic coupling does not (Dikstein, 1973). It seems likely that such metabolic regulation as occurs will express itself on chemically coupled phenomena, although this is hardly an absolute rule. Some of the points dealt with below have also been raised in the review by h p p i et al. (1980), in which the emphasis is on electromechanically coupled tissues affected by redox agents. The term “redox state,” except where specifically limited, will be used for the purpose of the discussion below in a broad sense. It shall signify any of the following ratios: NAD(P)H/NAD(P) , R-OHIR-OOH, GSH/GSSG, as well as the ratio of -SH to -S-S- groups in enzymes, membrane proteins, or nonproteins. As will become clearer below, these ratios are usually interdependant. A “redox agent,” consequently, will herein be any chemical which has a significant ability to alter any of the above (redox state) ratios, and an -SH agent is one which can block the action of a redox agent by binding to or blocking -S-S- or +

STIMULUS-RESPONSE COUPLING AND THE R E W X STATE

43

-SH groups. Furthermore a “redox effect” shall be any condition or chemical which either alters one or more of the above ratios, or significantly changes the activity of one of the enzymes controlling such a ratio, e.g., the pentose shunt dehydrogenases (hexose monophosphate shunt dehydrogenases), glutathione reductase, or glutathione peroxidase.

B. GENERAL (OR UNDEFINED) INTERACTION BETWEEN REDOXSTATES AND AGONISTACTION Among the earliest reports in which there were shown to be clear effects of redox agents on classical pharmacologic models were those on isolated rabbit ileum (Laborit, 1965a) and on melanophore dispersion in hyla frog skin (Dikstein et al., 1965). The relaxing effect of adrenaline on rabbit ileum was antagonized by a reducing agent (aminoethylthiouronium) and stimulated by an oxidizing agent (methylene blue). In fact, methylene blue itself inhibited spontaneous contractions and the reducing agent stimulated such contractions. In Hyla frog skin, various oxidizing agents were shown to induce melanophore dispersion. Even ir. much more primitive cells, oxidizing agents were found to be excitatory: menadione, methylene blue, phenazine methosulfate, and 2pyridyldisulfide when added to Ca2 -containing medium all were found to induce rhythmic contractures in the protozoa Vorticelfa (Raad et al., 1974). In the protozoa Spirostomum, which does not depend on extracellular Ca2+ for contracture (mobilizing some intracellular Ca2 store), some of the same oxidizing agents were ineffective until microinjected (Hawkes and Dikstein, 1976) or until the external membrane was permeabilized by digitonin treatment (Maran and Dikstein, 1979). F’uppi et al. (1968), using frog ileum and frog heart, showed that oxidation (with methylene blue) caused acetylcholine (ACh) to act in a stimulatory manner, while reduction (with ascorbate) caused ACh to act in an inhibitory fashion; adrenaline, on the other hand, stimulated under oxidizing conditions and inhibited under reducing conditions. In fact, even without artificially altered redox states, different tissues (of the frog) seem to have characteristic, different normal redox states which may determine the direction of response to ACh: those in a relatively more reduced state (heart muscle) are inhibited by ACh (Puppi et al., 1976). The latter results might significantly aid our understanding of redox state involvement in stimulus-response coupling, provided that they could be verified by more direct means: in such studies the “redox state potential” was measured in the extracellular fluid and in the debris of damaged cells; such conditions may not exactly reflect the situation in intact cells. Quite a number of pharmacological responses and physiological processes in very different cell types have been shown to be sensitive to the GSH/GSSG +

+

44

ZYGMUND ROTH ET AL.

redox state. For a survey of some of the older reports of the involvement of GSH in stimulus-response coupling or as a regulatory agent, especially of ATPase activity, the reader is referred to an earlier review by one of us (Dikstein, 1971). The following are given as some examples of such glutathione dependent processes: (1) the feeding reflex of Hydra (Loomis, 1955; Lenhoff and Bovaird, 1961) and of the tick (Galun and Kindler, 1968); (2) Na+ transport in toad bladder is correlated with the protein -SH/-S-S- ratio, regardless of the agent applied (vasopressin, CAMP analogs, aminophylline) (Farah et al., 1969); (3) free fatty acid release from adipocytes can be stimulated by GSH (Yu and Triester, 1970); (4) microinjection or digitonin permeabilization of GSSG into the protozoa Spirostomum induces cyclic contracture (Hawkes and Dikstein, 1976; Maran and Dikstein, 1979); ( 5 ) sea urchin egg division can be stimulated by GSH (see further discussion below) (Nath and Rebhun, 1976): (6) insulin release has been linked to the GSH/GSSG ratio (see further discussion below; also, Ammon et al., 1977); (7) adenylate cyclase is inhibited by GSSG (see further discussion below) (Baba et al., 1978; Mukherjee and Lynn, 1979); and (8) oxidation of GSH in rat coronary arteries leads to increased coronary blood flow (Zimmer et al., 1981). In order to be able to claim that redox changes are an integral part of the mechanism of cell response to agonists it is not sufficient to demonstrate that artificial application of redox agents affects cell response. It is also necessary to measure naturally occurring changes in the redox state following agonist action. Some reports which emphasize this approach are the following: Laborit (1965b) reported that ACh inhibited pentose phosphate shunt dehydrogenase activity in rabbit ileum. Dibutyryl-CAMP was shown to inhibit 6-phosphogluconate dehydrogenase in fat cells and caused oxidation of glutathione (Bray, 1967). In perfused rat liver, thyroxine can alter the pyridine nucleotide redox state (Hassinen et al., 1970). An increase in extracellular K + ion (which leads to depolarization) causes an increase in the pyridine nucleotide redox potential in rat cerebral cortex (Bull and Cummins, 1973). Following ACh stimulation of frog ileum or heart muscle, the intracellular “redox state potential” rises, that is, there is a shift toward oxidation of the redox state (Puppi et al., 1976). Finally, catecholamines have been shown to slowly stimulate G6PD activity in rat heart by inducing G6PD synthesis (Zimmer and Ibel, 1979), and t-butyl hydroperoxide (2-BHP) application results in measurable GSH oxidation, decrease in NADPH/ NADP+ , and rapidly stimulated pentose phosphate shunt activity, together with increased coronary blood (Zimmer et al., 1981). Metabolic effects may influence the redox state of cells and thereby their response to agonists. For example, the supply of metabolizable sugars, or their relative absence, may affect pentose phosphate shunt activity and thereby the NADPH/NADP+ redox state. Such a consideration may explain the ability of

STIMULUS-RESPONSE COUPLING AND THE REDOX STATE

45

fructose to stimulate contraction of K -depolarized smooth muscle (Dikstein et al., 1966; Hamburger et al., 1967). From the above reports it would appear that, in many cases, a change in the redox state toward oxidation is correlated with stimulation, whereas reduction correlates with recovery or inhibition. One notable exception is the effect, reported by Puppi et al. (1976), indicating that the characteristic redox state of a tissue may determine whether the tissue becomes stimulated or inhibited by a particular agonist. This issue may be resolved if it turns out that such tissuespecific redox responses also occur in other species, but that, in most tissues, the “normal” resting redox state of cytoplasmic NADPH/NADP+ , GSHIGSSG, and -SHE-S is in equilibrium within a narrow range, in a relatively reduced state. Furthermore, the specific intracellular site at which various redox agents act could well determine the direction of the effect on the response. We shall now turn our attention to reports which have attempted to pinpoint the stage of the stimulus-response-recovery cycle affected by redox changes. +

C . REDOXEFFECTSINFLUENCING THE AGONIST-RECEPTOR INTERACTION

While there are reports which indicate that the reduction of disulfide groups inhibits (and reoxidation removes such inhibition from) the nicotinic receptor (Karlin and Bartels, 1966; Harvey and Dryden, 1974) and that a nonpenetrant -SH binding agent inhibits activation of an a-adrenergic receptor (Neering and Glover, 1979), it is not at all clear that such regulation occurs in vivo. In particular, it would first be necessary to show that intracellular changes in the redox state of NADPH or GSH are translated into redox changes by these or other agents in the extracellular space. Since we do not know of such a demonstration we shall not here deal further with reports which, though interesting, concentrate on redox interference with agonist-receptor interaction.

D. REDOXEFFECTSINFLUENCING TRANSDUCER MERCHANISMS, EFFECTOR FORMATION OR RELEASE SYSTEMS,AND SECONDMESSENGER 1. Adenylate Cyclase In recent years reports have appeared in the literature clearly linking redox changes with effects on well-defined effector systems. In particular, oxidation has been shown to inhibit adenylate cyclase (Baba et al., 1978; Mukherjee and Lynn, 1979; C. Mukherjee er al., 1980; Mukherjee and Mukherjee, 1980) and to activate guanylate cyclase (Murad et al., 1979). Adenylate cyclase from rat brain (Baba et al., 1978) and in adipocytes (Mukherjee and Lynn, 1979) was shown to be inhibited by GSSG as a result of the oxidation of key -SH groups of the enzymes, and stimulation of adipocytes

46

ZYGMUND ROTH ET AL.

by isoproterenol and glucagon could be correlated with the total content of nonprotein -SH groups (Mukherjee and Lynn, 1979). It was subsequently found that adenylate cyclases of pancreatic islets and of human leukocytes are also stimulated by reducing agents and inhibited by oxidizing agents (C. Mukherjee et al., 1980; Mukherjee and Mukherjee, 1980). The stimulation of leukocyte adenylate cyclase by either isoproterenol or by prostaglandins could also be inhibited by oxidizing agents. These findings suggest a sequential mechanism in which agonist-receptor interaction leads to prostaglandin synthesis which in turn activates adenyl cyclase, the latter being regulated by the GSH/GSSG redox state. In this context it should be noted that prostaglandin endoperoxide mtabolism also depends on the GSH/GSSG couple (Christ-Hagelhof and Nugteren, 1978). One somewhat contrary result concerning adenylate cyclase has also appeared: Low et al. (1978) reported that NADH inhibits adenylate cyclase in isolated plasma membrane preparations and that CAMPand NADH levels are inversely related in whole cells. If both NADH and GSSG inhibit adenylate cyclase in the same tissue, then this may be a demonstration of the different specificities of the NADH vs NADPH-dependent systems (see also Section I1,E). Besides reviewing the accumulating reports of the various redox effects regulating guanylate cyclase, Murad et al. (1979) offer an interesting suggestion concerning the function of cGMP. They propose that cGMP may be acting as a feedback inhibitor, via one or more redox reactions, of the cellular redox state since guanylate cyclase is a sensitive monitor of this redox state. 2. The Release and Mechanism of Action of Insulin The stimulation of insulin release from isolated pancreatic islets has been clearly linked to redox phenomena by the work of Ammon et al. (1977) and others (Watkins, 1972; Malaisse et al., 1978). It is a somewhat special case, compared to the other phenomena dealt with herein, in that activation of the “glucoreceptor” appears to be at the intracellular substrate level (on the glucose kinase molecule) and not at the level of the cell membrane (Ashcroft, 1980). It has been shown that glucose increases the NADPH/NADP+ ratio (Ammon et al., 1979). But in the presence of glucose, the oxidizing agent methylene blue inhibits an increase in NADPH/NADP+ and in insulin release. Diamide (which oxidizes GSH) and t-BHP (which causes the oxidation of GSH and NADPH) also inhibit glucose-stimulatedinsulin release and inhibit insulin release stimulated by p-chloromercuribenzoate, a nonpenetrant -SH blocking agent. The evidence clearly implicates pentose phosphate shunt activity, NADPH, GSH reductase, GSH, and membrane -SH groups in a mechanism whereby increased glucose leads to the reduction of -S-S- groups of the outer membrane, thereby making it more permeable to Ca2+ and/or K . The increase in Ca2 levels then induces insulin exocitosis (Ashcroft, 1980). By analogy with conventional receptor mechanisms, glucose-6-phosphate dehydrogenase (G6PD) and GSH reductase +

+

STIMULUS-RESPONSE COUPLING AND THE REDOX STATE

47

may be considered to act as transducing mechanism and effector system, and GSH behaves as a second messenger. The scheme described may not comprise the entire mechanism involved, but it is, apparently, an integral part of the glucose-insulin release, stimulus-response phenomenon. Not only has such glucose-stimulated insulin release been linked to redox phenomena but Mukherjee’s group has also suggested that insulin and fatty acids increase pentose phosphate shunt activity via a mechanism independent of the stimulation of glucose transport and operating via a plasma membrane-bound NADPH oxidase producing H,O, (Mukherjee and Lynn, 1977; S. P. Mukherjee et al., 1980). The H,O, produced by this insulin-induced increase in NADPH oxidase activity presumably leads to the formation of GSSG and NADP+ , the latter increasing G6PD activity. Perhaps even more important, the GSSG formed decreases adenylate cyclase activity of adipocytes (Mukherjee and Lynn, 1977) (see previous section), thereby offering an explanation of the mechanism of the action of insulin on adenylate cyclase. 3 . Sea Urchin Egg Development The stimulation of development of the sea urchin egg is a primitive but potentially very useful model for the study of stimulus-response phenomena. One can consider initiation of such development an agonist-receptor phenomenon because specific receptor proteins (on the vitelline layer around the egg) bind to complementary proteins on the acrosomal process of the sperm, thereby inducing molecular changes inside the cell. A good working knowledge of the temporal sequence of molecular events following stimulation has accumulated (for a review, see Epel, 1977). Some of these events, in sequential order, are a small Na+ influx into the cell; liberation of Ca2+ ions; a “cortical response”; H release and a major Na influx; conversion of NAD to NADP ; a rise in 0, consumption; a rise in pHi; and a “second response” including protein synthesis, activation of active transport, and cell division. One very interesting finding (with respect to redox involvement in stimulus-response coupling) is that, following fertilization, G6PD, the primary regulator of the cytoplasmic NADPH/NADP+ ratio, changes its subcellular location (Epel, 1979). Before fertilization, G6PD is associated with a particulate fraction but becomes solubilized following fertilization. Since this change occurs after the rise in intracellular pH, Epel suggests that G6PD solubilization is a direct result of the pHi increase. The mechanism by which Epel suggests that this may occur is one involving pH-sensitive changes in inhibitor proteins that bind G6PD and inhibit its activity. Since G6PD in most cells is found in the soluble cytoplasmic fraction, it may be that such G6PD activation is a mechanism peculiar only to stimulation of egg development; it remains to be seen if pH changes can influence G6PD location and activity as profoundly in other cells. It has been shown that methylxanthines inhibit egg division, not via phos+

+

+

+

48

ZYGMUND ROTH ET AL.

phodiesterase/cAMP but rather by a mechanism involving decreased GSH reductase activity and decreased Ca2+-activated ATPase activity (Nath and Rebhun, 1976). Also, GSH and diamide (the GSH specific oxidant) act as a stimulator and inhibitor, respectively, on egg development. Furthermore Li (1977) has shown that the GSH reductase of sea urchin egg can exist in a complex form which is less stable and less active than the dissociated form. The complex form can be rapidly dissociated into the active form by various proteinases, and only the complex form is strongly inhibited by the presence of NADPH. The following scheme can incorporate these various results: The rise in pH, following fertilization leads to solubilization and activation of G6PD and the production of NADPH. With NADPH inhibiting the complex, low-activity form of GSH reductase, dissociation (due to a specific proteinase) occurs and the enzyme becomes active, producing GSH. Apparently GSH is necessary for one or more of the processes involved in development, since stimulation or inhibition of development can be induced by GSH or GSSG, respectively. In the case of the stimulation of development of the sea urchin egg, one probably should not attempt to fit the successive molecular events into the standard receptor-transducer-effector-second messenger-response sequence. Each successive molecular event should probably be seen both as a “response” to the preceding event and as an “effector” or “second messenger” for the event following it. 4. Redox Effects of Agonists in Stomach Fundus Smooth Muscle

The fundus muscle is pharmacologically of interest since the source of Ca2 for contraction depends on the agonist applied: The depolarization caused by serotonin or K+ utilizes extracellular Ca2+, whereas acetylcholine (ACh) can mobilize both extracellular and intracellular Ca2 (Weinstcck and Weiss, 1979). The activity of fundus cell G6PD, following stimulation with various agonists, has been measured (N. Chayen, 1981). In order to introduce as little cell disruption as possible, the methods of cytochemistry were utilized, in which the biochemical activity of individual cells is measured under “almost in vlvo” conditions (J. Chayen, 1978). As can be seen from Table I, ACh caused a decrease in G6PD activity, whereas serotonin and KCl increased G6PD activity. In separate experiments, 6phosphogluconate dehydrogenase (6PGD) activity was unaltered by acetylcholine but was increased slightly by serotonin (from 5.720.1 to 8.2k0.2 within 2 minutes). The time-course of the effect of these agonists on G6PD is similar to the duration of contraction in intact fundus strips (as measured by conventional means) with ACh having a more prolonged effect than serotonin, while KCl exhibits a gradually increasing effect. Studies with isoprenaline were also conducted (Chayen and Dikstein, unpublished results), Applied directly to fundus sections at M ,isoprenaline +

+

49

STIMULUS-RESPONSE COUPLING AND THE REDOX STATE

TABLE I CHANGES (MEAN 2 SEM FOR AT LEASTTHREEANIMALS) I N G6PD ACTIVITY(MEANEXTINCTIONX 100) WITH TIMEA m E R TREATMENT WITH ACETYLCHOLINE OR SEROTONINa Activity at (minutes)

Maximum percentage change

Concentration Agent

(M)

T-8 medium alone

5

0

I

2.5

-

-

-

19.9 rt 1

-

9.3 12.9 17.7 19.2 30.2

-

( n = 24)

Acetylcholine

10-3

10-4 10-5

Serotonin KCI

10-4

10-2

17.8 18.0 19.9 16.9 20.0

f

1.1

f 1.0 2 0.9 f 2.2 f 2.0

27.0 ? 3.4

-

11.6 f 1.6 19.0 2 1.1 22.5 f 2.4 21.4 2 2.8

2 2.1 rt 1.5

-47.7 -35.5

-

rt

1.4

rt

0.9

+60

rt

2.5

+51

"From Chayen et al. (1983). bApart from these results, the activities recorded in the treated strips within each experiment were significantly different from those in the control strips at p = 0.05 or less.

increased G6PD activity by 33% while leaving that of 6PGD unaffected. When applied to strips of the fundus first, sections thereof exhibited an increase in G6PD activity of 68% without affecting that of 6PGD. Isoprenaline is a catecholamine receptor agonist whose action is the opposite of ACh, namely it causes fundus muscle relaxation. It is significant that their opposing physiological action is paralleled by their opposing effects on G6PD activity. It is unlikely that G6PD itself has separate recognition sites for all these agonists and K + depolarization. Rather, it is much more reasonable that the change in G6PD activity is mediated by receptor activation or the depolarization which can follow it. M )nor EGTA affected G6PD activity, the changes Since neither Ca2+ in enzyme activity induced by ACh, serotonin, and KCI are not related to the contraction or recovery stages of the stimulus-response-recovery cycle, but rather would appear to be initiated during stimulus-response coupling. Also, since it was found that the rate of reoxidation of NADPH generated by G6PD activity was not altered (N. Chayen et al., in preparation), it is possible that such modulation of G6PD activity would alter the NADPH/NADP+ ratio. It is tempting to speculate that the process of intracellular Ca2 release is associated with decreased pentose phosphate shunt activity and a decreased NADPH/NADP ratio (as suggested in Section 11,D,5), whereas extracellular Ca2+ entry is associated with an increase in these parameters. The latter also appears to pertain to Ca2+ resorption in the nephron when stimulated by parathyroid hormone (Chambers er al., 1978). Since serotonin and ACh receptors exhibit fundamentally different transducer mechanisms in various tissues (Bemdge, 1980), it is +

+

50

ZYGMUND ROTH ET AL.

possible that the increase and decrease of G6PD activity by these agonists may be a part of such separate transducer mechanisms.

5 . Interactions between the Redox State and the Cytoplasmic Ca2 Level One of the most common second messengers in pharmacological systems is the calcium ion (Ca2+),but in many cases the mechanism whereby its cytoplasmic level is increased following stimulation is poorly or not at all understood. While in electromechanically coupled stimulus-response phenomenon much is known about the mechanism for Ca2 release into the cytoplasm, the biochemical mechanism involved in pharmacomechanic coupling is only beginning to be elucidated (Berridger, 1980; Brading, 198 1). Evidence has accumulated which indicates that redox state alterations may be a mechanism whereby Ca2 release occurs. A major pool of intracellular Ca2+ is the mitochondrion. In the protozoa Vorticella and Spirostomum, oxidizing agents were found to induce rhythmic contractures, responses known to be coupled by Ca2 , with some evidence indicating mitochondria as the source of such Ca2+ (Dikstein and Mitelman, 1975; Hawkes and Dikstein, 1976). Spirostomum. in particular, depends exclusively on intracellular Ca2 for contracture (Ettienne and Dikstein, 1973; Hawkes and Dikstein, 1976). In isolated liver mitochondria, Lehninger c?t af. (1978) have shown that oxidation of pyridine nucleotides leads to massive Ca2 release and in isolated intact liver cells oxidizing metabolites promote Ca2 release (Krell et al., 1979). Also, oxidation of NADPH and of NADH by t-BHP, and the resulting Ca2 release, have been shown to be reversible, nondestructive phenomena (Lotscher et af., 1979) and, what is more, can lead to the hydrolysis of pyridine nucleotides (Lotscher et al., 1980). Other workers, however, have questioned the importance of such oxidationinduced Ca2+ release (Nicholls and Brand, 1980; Beatrice et al., 1980; Wolkowicz and McMillan-Wood, 1980). The criticism is, essentially, that such oxidation occurs with mitochondrial swelling and perhaps damage, and that this oxidation may be secondary to and distinguishable from other primary effects. It has been shown, however, at least for the liver mitochondrial Ca2+ release which is insensitive to ruthenium red (which blocks the membrane potentialdriven Ca2+-uptake carrier), that specific reduction of the NADPH/NAI)P couple from an initially physiological redox state inhibits Ca2+ release (Roth and Dikstein, 1980, 1982). Reduction of the NAD+/NADH couple alone was not effective in blocking ruthenium red-insensitive Ca2 release. Since ruthenium red-insensitive Ca2 release is a physiological (but normally masked) process, it was suggested that NADP+ :NADPH specific reactions (and not those of NAD+ :NADH) are part of a physiological mechanism regulating Ca2+ release/retention. This is also in agreement with the results of PrpiC and Bygrave +

+

+

+

+

+

+

+

+

+

+

STIMULUS-RESPONSE COUPLING AND THE REDOX STATE

51

(1980) who found that Ca2+ release from mitochondria due to oxidation is related to NADPH, not NADH, oxidation, and with the results of Sies et al. (1981), who found that Ca2+ efflux from whole liver cells is stimulated by drugs which specifically decrease the cytoplasmic NADPH/NADP ratio without affecting the NADH/NAD ratio. Finally, recent results by Baumhutter and Richter (1982) suggest that the mitochondrial membrane disruption observed by some workers (see above) following extensive oxidation was due to prolonged Ca2+ cycling. Even strong tBHP-induced oxidation did not cause a decrease in membrane potential when such Ca2 cycling was prevented by the presence of EGTA. It is suggested that, in vivo, active Ca2+ uptake/extrusion by the endoplasmic reticulum and cell membrane acts similarly to prevent Ca2+ cycling by mitochondria. It is, of course, not clear whether such Ca2+ release from mitochondria is just a “fine tuning” mechanism, say for regulating cytoplasmic metabolism, or if it can be triggered by agonist-receptor interaction. One report presents evidence that, indeed, a-adrenergic stimulation of liver cells leads to Ca2+ release from mitochondria (Blackmore et af., 1979). What are clearly lacking however, are the biochemical intermediate steps that could link agonist receptor activation with mitochondrial oxidation of NADPH and GSH to cause Ca2+ release. The following scenario may be suggested. It is not hard to imagine that agonist-receptor interaction could lead to the stimulation of an NADPH specific oxidase which forms H202 (Mukherjee and Lynn, 1977) or to the decrease of pentose phosphate shunt dehydrogenase activity (as first suggested by Laborit, 1965b). It was indeed found that the latter is the case for ACh stimulation of stomach fundus smooth muscle (Chayen, 1981; Table I), a phenomenon which depends on an intracellular source of Ca2+ (see Section 11,D,4). Decreased pentose phosphate shunt activity (acting as a transducer mechanism) could lead to a decrease of the cytoplasmic NADPH/NADP+ ratio, and this could be translated into an intramitochondrial NADPH/NADP decrease via such substrate couple shuttles as isocitrate/a-ketoglutarateor glutamate/a-ketoglutarate. Finally, the decreased mitochondrial NADPH/NADP redox state would induce the mitochondrial Ca2+ “gates” (the effector system) to release Ca2+ and thereby produce a response. It should be pointed out that thus far it is not completely clear that ACh-stimulated fundus G6PD inhibition actually leads to a significant change in the NADPH/NADP+ ratio, or that changing the cytoplasmic NADPH/NADP ratio alters the mitochondrial redox state. It thus remains to be demonstrated whether the ACh-stimulated decrease of G6PD activity in the fundus functions as a transducer mechanism or whether it simply acts in a regulatory capacity. As for Ca2+ which enters from outside the plasma membrane during stimulus-response, there is evidence to suggest that redox changes may also play a +

+

+

+

+

52

ZYGMUND ROTH ET AL.

role. In isolated rabbit ear arteries, p-chloromercuribenzoate (a nonpenetrant -SH blocking agent) causes an increase in Ca2+ permeability which induces contraction (Neering and Glover, 1979).

E. REDOXEFFECTSINFLUENCING RESPONSE Responding elements, whether contractile proteins, enzymes, or subcellular organelles, insofar as they are composed of complex proteins, contain -SH and -S-S- groups which may be sensitive to changes in the redox state. It is, however, not within the scope of this article to consider the effects of redox changes directly on contractile proteins. The interested reader is referred to the reviews of Hoffmann and Hamm (1978) and Tonomura (1973). It should be pointed out, however, that in order for reports of redox effects on contractile proteins to have pharmacological significance, it must in each case be shown that a given -SH/-S-S- change occurs physiologically and is the result of agonist stimulation. Apart from effects on contractile proteins, there have been few, if any, reports of direct redox effects on target enzymes mediated by receptor activation. (The effect of GSSG on adenylate cyclase must be considered as a regulatory effect on an effector system and not directly on the final response.) The kind of enzyme response that may be a candidate for direct redox control of activation is the recently reported stimulation of myocardial phosphofructokinase mediated by ciadrenergic receptors (Clark and Patten, 1981). Since phosphorylation of the enzyme is nor the mechanism whereby activation occurs (Patten et al., 198l), it is tempting to speculate that activation might occur via -S-S-/-SH redox changes. INFLUENCING ATPASESAND RECOVERY F. REDOXEFFECTS In an earlier review (Dikstein, 1971), one of us proposed the hypothesis, based on reports in the literature (Dick et af., 1969; Wilkers and Bader, 1969; Wins, 1969), that a major function of the GSH:GSSG system and related redox systems was to regulate cellular ATPases. Since then, additional reports have further confirmed this contention in a variety of cell models and isolated organelle preparations (Wald er al., 1972; Puppi et al., 1975; Fromm and Fuhro, 1978; Mushlin et al., 1978; Whikehart and Soppet, 1981). ATPases are inhibited by GSSG, and the GSH/GSSG ratio appears to be controlled by an intricate system of biochemical inhibition and stimulation (see Dikstein, 1971). The activity of glutathione reductase, the enzyme which reduces GSSG to GSH, is decreased by 6-phosphogluconate. Of the enzymes controlling the concentration of 6-phosphogluconate, G6PD activity is increased by GSSG and 6-phosphogluconate dehydrogenase activity is decreased by fruc-

STIMULUS-RESPONSE COUPLING AND THE REDOX STATE

53

tose 1,6-biphosphate. Phosphofructokinase activity, which regulates fructose 1,6-biphosphate production, is in turn increased by ADP. Thus, we have a situation in which phosphofructokinase and G6PD activity indirectly exert control over the GSH/GSSG ratio, and thereby, on ATPase activity. It is interesting to note that in cancer cells fructose 1,6-diphosphate levels are very high, while NADPH and GSH levels are low (Eigenbrodt and Glossman, 1980). The rate of plasma membrane Na -K -ATPase and Ca2 -ATPase pumping have a direct effect on the rate of repolarization and of Ca2+ removal from the cytoplasm; hence such ATPase activity will control the rate of recovery following cellular response. Consequently, it is quite possible that the GSSG concentration in the vicinity of plasma membrane ATPases may regulate the recovery phase of the stimulus-response-recovery cycle. It has been proposed that the release of various transmitters is controlled by Na+ -K+-ATPase activity in the plasma membrane (see the review of Vizi, 1978). Various agents that inhibit Na+-K+-ATPase, among them ouabain, Ca2 ,p-chloromercuribenzoate,and N-ethylmaleimide, also stimulate transmitter release. Obviously, such a system is a stimulus-response phenomenon, the secretion of transmitter being the response; ATPase may be acting as an effector system and could be regulated by the redox state of its -SH/-S-Sgroups. However, whether redox agents are here involved physiologically (as they can be experimentally) is still not clear. If the inhibition of plasma membrane Na -K -ATPase activity is strong enough in electromechanicallycoupled systems, a slow depolarization leading to Ca2+ influx and response would occur; in the latter case, ATPase inhibition could be considered a transducing mechanism causing cellular response. Similarly, stimulation or inhibition of Ca2 -ATPase of endoplasmic or sarcoplasmic reticulum in cells pharmacomechanically coupled by Ca2+ (Mushlin et af., 1978) could either affect the rate of recovery or induce cell response. While there are reports demonstrating that various agents inhibit or stimulate different ATPases (Schwartz et al., 1973, an attempt to demonstrate that the effect of an agonist on ATPase is mediated by redox changes has not, to our knowledge, been made. Separately, however, each of the two steps involved has been shown to occur, namely (1) agonists can change the redox state and (2) GSH:GSSG and other redox agents control ATPase activity (for references see above). An attempt to show the operation of such a two-step mechanism in a cellular model might prove a fruitful exercise. +

+

+

+

+

+

+

G. CONCLUSIONS AND SPECULATIONS

Changes in the cellular redox state certainly can affect events occurring during the stimulus- response-recovery cycle. Several basic questions must, however, be answered to determine the importance of such findings. First, are the numer-

54

ZYGMUND ROTH ET AL.

ous and different phenomena described in the reports discussed above only events artificially induced in the laboratory, or do they also have physiological significance involving the receptor mechanism? In at least a few of the cases mentioned, evidence suggests that changes in the redox state and/or redox agents do seem to be physiologically involved in some stage of the stimulus-response-recovery (SRR) cycle. Two further questions of importance are (1) whether a given change in the redox state is an obligatory link in the events occurring between agonist-receptor interaction and response, or whether such a change acts only as a modifier of one of the stages of the SRR cycle; and (2) which stage of the SRR cycle is affected. As pointed out in Section II,B, there is often an absence of experiments attempting to pinpoint the precise stage of the SRR cycle affected by redox changes. In practice, it is not always possible to devise experiments which can pinpoint the stage affected. We can, however, suggest that the following approaches (in the form of questions) be considered when analyzing a redox-sensitive phenomenon: 1. Can redox agents themselves elicit a response or do they only modify the response due to an agonist? 2. Following physiological stimulation, can changes in the activity of enzymes directly related to the redox state be determined (e.g., G6PD, 6PGDH, GSH peroxidase, GSH reductase, isocitrate dehydrogenase, glutamate dehydrogenase)? 3. Can changes be determined in the ratios of reduced vs oxidized metabolites which are substrates for enzymes that determine the redox state? 4. Does prolonged removal or block of reduced metabolites or their precursors (e.g., glucose, glucose 6-phosphate) inhibit, enhance, or perhaps even elicit the physiological response? 5 . If Ca2 is the coupling agent, is its source (for eliciting the response) intraor extracellular? 6. If the Ca2 source for coupling is extracellular, are permeable or impemeable redox or -SH agents effective? 7. If the Ca2+ source for coupling is intracellular, do agents which specifically activate or block certain cell organelles amplify or inhibit the redox-sensitive response? 8. Does elicitation of the response by direct introduction (if possible) of the coupling agent also elicit redox state changes? +

+

Let us finally consider the questions raised above (concerning the stage and obligatory vs modulatory involvement) with respect to some of the reports discussed in this article. In the case of the Hydra feeding response (Loomis, 1955) the natural redox agent, GSH, behaves as an agonist. In insulin release, redox systems appear to be acting as transducers and effectors, with GSH acting as a

STIMULUS-RESPONSE COUPLING AND THE REDOX STATE

55

second messenger (Ammon er al., 1979). The enzyme NADPH oxidase (Mukhejee and Lynn, 1977), stimulated by insulin, functions as a transducer for the adenylate cyclase effector. As for the cholinergically stimulated stomach fundus smooth muscle, it still must be shown whether a decrease in cytoplasmic G6PD activity is an obligatory part of a receptor mechanism leading to mitochondrial Ca2+ release (see Section II,D,4). In other cases, the function of redox state changes in the SRR cycle seems to be regulatory, not obligatory: regulation of adenylate cyclase activity by GSH/GSSG is an example thereof. The effect of the redox state on ATPases is, basically, a regulatory phenomenon, but, as pointed out earlier, sufficiently strong ATPase stimulation or inhibition might act as a transduction mechanism. In summary, a large number of reports demonstrate that redox state changes affect agonist-stimulated responses, and some show that such redox changes occur physiologically, but the precise stage of the SRR cycle involved is often not known. In a few cases there is evidence to suggest that redox agents and changes in the redox state may be obligatory links in stimulus-response coupling, but in many others, they appear to act as regulators of the receptor mechanism or of ATPases associated with the recovery stage.

111. The Involvement of Intracellular pH in Stimulus-Response Coupling As pointed out in Section I, the pyridine nucleotide (PN) redox state and intracellular pH (pH,) are related. Thus, PN oxidation may lead to pH, increase whereas PN reduction may lead to acidification. Conversely, a pH, increase tends to cause PN reduction and acidification favors PN oxidation, due to the effect of mass action. It is possible, therefore, that any effect ascribed to a change in the redox state of pyridine nucleotides may actually be due to a pH, change, and vice versa. To alleviate doubt about which of these two parameters is the final effector in a given model it is, thus, necessary to separately initiate pH, and redox changes and compare their respective effects on the response, if any. Ideally one would also wish to monitor both pH, and the PN redox state during any given stimulus-response event. A. THEEFFECTOF CHANGES IN pH,

ON

STIMULUS-RESPONSE EVENTS

While it is outside the scope of this article to consider the numerous effects of intracellular pH on cell functions (see the review of Roos and Boron, 1981), we do wish to consider various reports that directly relate to the effect of pH, on stimulus-response coupling. For example, increasing CO, pressures (which cause pH, decease) caused

56

ZYGMUND ROTH ET AL.

increasing inhibition of histamine-induced contraction in guinea pig intestine, and pure CO, inhibited ACh- and KC1-induced contraction (Halpern, 1956). In a somewhat inverted situation the effect of atropine as an ACh antagonist is increased in the presence of CO, and HC0,- (Bucher and Horiuchi, 1974). In cardiac muscle, hypercapnia leads to a fall in pHi and decreased cardiac function (Popham et al., 1980). In rat diaphragm skeletal muscle, removal of CO, leads to increased isotonic contraction height (Creese, 1950), as did addition of NH4+ in frog skeletal muscle (Washio and Mashima, 1963). In a smooth muscle, pH, increase led to increased carbamylcholine- and serotonin-induced contraction amplitude and pH, decrease inhibited such response (Roth and Dikstein, 1981). In more primitive tissues, it has been shown that the motility of sea urchin sperm can be initiated by pH, alkalinization, which is followed by CaZ+ influx (Lee et al., 1980). Also, microinjection of alkaline buffers into amoebae causes contracture (Heiple and Taylor, 1980), and we have observed the same phenomenon in microinjected Spirostomum (unpublished observations). The common finding of the various reports mentioned above may be summarized by the statement: Intracellular alkalinization acts in a stimulatory fashion whereas acidification acts in an inhibitory fashion. Little is known about the mechanism whereby such pHi changes affect stimulus-response events. While it has been shown that in vitro alkalinization leads to increased CaZ release from isolated sarcoplasmic reticulum (Nakamaru and Schwartz, 1972), and CO, or HCO, - releases Ca2 from sarcoplasmic reticulum in crustacean myofibrils (Lea and Ashley, 1981), such a mechanism does not appear to play an important role in an intact smooth muscle preparation (rat stomach fundus, Roth and Dikstein, 1981). In the latter case it was found that only the entry of extracellular CaZ+ is affected by pH, changes. +

+

B. THEEFFECTOF STIMULUS-RESPONSE PHENOMENA ON pHi While it appears clear that alterations of pH, change the response of many cells to agonists, we must again ask whether such changes in pH, occur physiologically and whether pH, change is an obligatory link in stimulus-response coupling. Such evidence as exists gives rather incomplete answers to these questions, except for a recent report demonstrating that pH, mediates insulinstimulated glycolysis (Fidelman ef al., 1982). In rat heart, there is an intramuscularpH oscillation which occurs independent of contraction but in phase with electrical stimulation (Dhalla et al., 1972); it is not clear, however, if such pH changes represent intracellular pH changes. In frog skeletal muscle, contraction occurs with a very rapid pH, acidification followed by a slower alkalinization when measured by rapid optical means (McDonald and Jobsis, 1976). In the same tissue, subliminal doses of caffeine and quinine (which release CaZ+), as well as K + depolarization, led to long-

STIMULUS-RESPONSE COUPLING AND THE REDOX STATE

57

term alkalinization (Connett, 1978). Both reports ascribe the alkalinization process to the recovery stage, since blockers of Ca2 release also block alkalinization. The initial acidification observed in frog skeletal muscle was also found, but over longer time intervals, in turtle bladder cells, where the Ca2+ releasing agents carbamylcholine and A-23187 led to a fall in pH, (Armda et al., 1981), and in cerebral cortex seizures induced with penicillin, where pH, fell from 7.10 to 6.94 (Tenny et al., 1978). The source of the acidification observed is either ascribed to ATP hydrolysis (in frog skeletal muscle) or has not been investigated. In two other models, stimulation is found to be associated with alkalinization. In frog gastric mucosa, stimulation (with histamine and theophylline) of mucosal acid secretion is accompanied by a rise in pH, (Ekblad, 1980). In the sea urchin egg, a rise in pH, follows stimulation (fertilization) but precedes cell division (Lop0 and Vacquier, 1977); Epel suggests that such an increase in pH, is a “primary effector” of various changes associated with cell division, among them being the solubilization and activation of glucose-6-phosphate dehydrogenase (Epel, 1979). The stimulation of glycolysis by insulin was recently shown to increase pH, in frog skeletal muscle, and decreasing pH, with CO, caused a decrease in glycolysis (Fidelman et al., 1982). Further evidence was also presented to support a molecular mechanism in which insulin activates a Na+:H+ exchange at the plasma membrane leading to pH, increase, with such alkalinization increasing phosphofructokinase activity to increase glycolytic flux. Such a mechanism implies that the activation of a Na+:H+ exchange “channel” occurs via insulin stimulation of a receptor site, perhaps also located on the plasma membrane. This model is the clearest demonstration, as far as we know, of pH, change as an integral part of stimulus-response coupling. Considering these various reports, it would appear that pH, changes do indeed occur physiologically during various stimulus-response-recovery cycle phenomena. In the case of insulin-stimulated glycolysis, and perhaps also for sea urchin egg stimulation, a good case can be made for an increase of pH, being an obligatory link in stimulus-response coupling. For the other cases cited, the nature of the pH, involvement is still unclear. Some of these pH, changes may be acting in a regulatory capacity or may be by-products of metabolic processes. In any case, the pH, changes occurring during the stimulus-response-recovery cycle should be monitored and their effect taken into account in the final elucidation of the sequence of events. +

.

IV. Epilogue This article suggests that the redox state and pH, are involved in the stimulus-response-recovery cycle. In the course of preparing it, we have become

58

ZYGMUND ROTH ET AL.

impressed with the need for future studies using intact cell models and relatively in vivo methods, such as those of cytochemistry, wherever possible. The explanation of pharmacological events in biochemical terms is the realm of biochemical or molecular pharmacology. In discussing some of the problems of receptor mechanisms (one of the central concerns of modern pharmacology) we have attempted to point out that studies in this field often lack evidence (not to mention proof) that (1) the biochemical event studied really occurs physiologically or, at least, in pharmacological situations, and (2) that the biochemical event is an obligatory step in the physiological or pharmacological process studied. With the use of proper cytochemical techniques, the interaction among cell organelles is preserved; even more important, the enzymes studied are in their natural environment with all their allosteric and other regulatory mechanisms in a relatively intact form. Such relatively nondestructive analysis is certain to result in more meaningful data than the destructive techniques (e.g., cell disruption and ultracentrifugation) usually employed. This is especially true in such highly organization-dependent parameters as redox potential and pH for which the destructive techniques almost certainly produce ambiguous, inconclusive data. It is hoped that the nondestructivetechniques will find greater favor and application among pharmacologists and biochemists.

ACKNOWLEM~MENTS Some of the original work quoted by the authors was supported by the Stiftung Volkswagenwerk Grant 1/34/722. >e authors wish to thank J. Chayen (Kennedy Institute for Rheumatology, Lmdon) for his help and constructive criticism in the preparation of this manuscript.

REFERENCES Ammon, H. P. T., Akhtar, M. S., Niklas, H.,and Hegner. D. (1977). Mol. Phurmacol. 13, 598-605.

Ammon, H. P. T., Hoppe, E., Akhtar, M. S., and Niklas, H.(1979). Diuberes 28, 593-599. Armda, J. A. L., Dytko, G., Lubansky, H.,Mola, R., Kleps, R., and Burt, C. T. (1981). Biochem. Biophys. Res. Commun. 102, 891-896. Ashcroft, S. J. H. (1980). Diuberologiu 18, 5-15. Baba, A., Lee, E., Matsuda, T., Kihara, T., and Iwata, H. (1978). Biochem. Biophyf. Rex Commun. 85, 1204-1209. Bull, R. J., and Cummins, J. T. (1973). J. Neurochem. 21, 923-937. Baumhiiter, S.. and Richter, C. (1982). FEBS Lerr. 148, 271-275. Beatrice, M. C., Palmer, 1. W..and Pfeiffer, D. R. (1980). J. Biol. Chem. 255, 8663-8671. Berridge, M. J. (1980). Trends Pharmucol. Sci. 1, 419-424. Blackmore, P. F., Dehaye, I . , and Exton, J. H. (1979). J. Biol. Chem. 254, 6945-6950. Brading, A. F. (1981). Trends Phurmacol. Sci. 2, 261-265.

STIMULUS-RESPONSE COUPLING AND THE REDOX STATE

59

Bray, G. A. (1967). Biochem. Biophys. Res. Commun. 28, 621-627. Bucher, K.. and Horiuchi, T. (1974). Agents Acfions 4, 410-412. Chambers, D. S . , Dunham, J., Zanelli, J. M., Parsons, J. A., Bitensky, L., and Chayen, J. (1978). Clin. Endocrinol. 9, 375-379. Chanutin, A., and Curnish, R. R. (1967). Arch. Biochem. 121, 96-108. Chayen, J. (1978). Inr. Rev. Cytol. 53, 333-396. Chayen, N. (1981). M. Sc. Pharm. thesis, Hebrew University, Jerusalem. Chayen, N., Roth, Z., Dikstein, S., Bitensky, L., and Chayen, J. (1983). In preparation. Christ-Hagelhof, E., and Nugteren, D. H. (1978). Hoppe-Seyler Z. Physiol. Chem. 359, 1040. Clark, M. G., and Patten, G. S . (1981). Nature (London) 292, 461-463. Connett, R. J. (1978). Am. J . Physiol. 234, ClIO-Cl14. Creese, R. (1950). J. Physiol. (London) 110, 450-457. Dhalla, N. S.,Yates, J. C., and Kleinberg, I. (1972). Can. J . Physiol. Phurmucol. 51, 234-238. Dick, D. A. T., Dick, E. G., and Tosteson, D. C. (1969). J . Cen. Physiol. 54, 123-133. Dikstein, S . (1970). Proc. Int. Congr. Phurmucol., 4th 3, 7-27. Dikstein, S . (1971). Nururwissenschufien 58, 439-443. Dikstein, S. (1973). I n “Fundamentals of Cell Pharmacology” (S. Dikstein, ed.), pp. 30-39. Thomas, Springfield, Illinois. Dikstein, S . , and Mitelman, S . (1975). Comp. Biochem. Physiol. 51C, 219-225. Dikstein, S., and Sulman, F. G. (1966). J . Theor. Biol. 12, 385-396. Dikstein, S., Tal, E., and Sulman, F. G. (1965). Biochem. Phurmucol. 14, 1151-1153. Dikstein, S., Hamburger, A. D., Tal, E., and Sulman, F. G. (1966). Isr. J . Chem. 4, 67-71. Dumont, J. E. (1980). Trends Phurmucol. Sci. 1, 219-222. Eigenbrodt, E., and Glossman, H. (1980). Trends Phurmucol. Sci. 1, 240-245. Ekblad, E. B. M. (1980). Biochim. Biophys. Acfu 632, 375-385. Epel, D. (1977). Sci. Am. 237, 128-138. Epel, D. (1979). In “Mechanisms of Cell Change” (J. Ebert and T. Okada, eds), pp. 17-35. Wiley, New York. Ettienne, E. M., and Dikstein, S . (1973). Nature (London) 250, 783-784. Farah, A., Yamodis, N. D., and Pessah, N. (1969). J . Phurmucol. Exp. Ther. 170, 132-144. Fidelman, M. L., Seeholzer, S . H., Walsh, K. B., and Moore, R. D. (1982). Am. J . Physiol. 242, C-87-93. Fromm, D., and Fuhro. R. (1978). Acfu Physiol. Scund. Special Suppl. pp. 91-100. Galun, R., and Kindler, S . H. (1968). J. Insecr Physiol. 14, 1409-1421. Halpern, 8. N. (1956). C. R. SOC. Biol. 150, 897-902. Hamburger, A. D., Dikstein, S., and Sulman, F. G. (1967). Arch. Int. Phurmucodyn. 167, 250-259. Harvey, A. L., and Dryden, W. F. (1974). Biochem. Phurmucol. 23, 3093-3096. Hassinen, J. E., Ylikhari, R. H., and Kahonen, M. T. (1970). Ann. Med. Exp. Biol. Fenn. 48, 176- 186. Hawkes, R. H., and Dikstein, S . (1976). Experientia 32, 1029-1031. Heiple, J. M., and Taylor, D. L. (1980). J. Cell Biol. 86, 885-890. Hofmann, K., and Hamm, R. (1978). Adv. Food Res. 24, 1-111. Karlin, A., and Bartels, E. (1966). Biochim. Biophys. Actu 126, 525-535. Kirtley, M. E., and McKay, M. (1977). Mol. Cell. Biochem. 18, 141-149. Krell, H., Baur, H., and Pfaff, E. (1979). Eur. J . Biochem. 101, 349-364. Laborit, H. (1965a). Agressologie 6, 129-138. Laborit, H. (1965b). “Les Rkgulations Mktaboliques.” Masson, Paris. Lea, T. J., and Ashley, C. C. (1981). J. Membr. Biol. 61, 115-125. Lee, H. C., Schuldiner, S . , Johnson, C., and Epel, D. (1980). J. Cell Biol. 87, 39a.

60

ZYGMUND ROTH ET AL.

Lehninger, A. L., Vercesi, A., and Bababunmi, E. A. (1978). Proc. Narl. Acad. Sci. U.S.A. 75, 1690-1694. Lenhoff, H. M., and Bovaird. J. (1961). Nature (London) 189,486-487. Li, 1. (1977). J . Biochem. 81, 673-680. Loomis, W. F. (1955). Ann. N.Y. Acud. Sci. 62, 209-228. h p o , A., and Vacquier, V. D. (1977). Nature (London) 269, 590-592. Liitscher, H.R., Winterhalter, K. H., Carafoli, E., and Richter, C. (1979). Proc. Narl. Acad. Sci. U.S.A. 76, 4340-4344. Liitscher, H. R., Winterhalter, K. H.,Carafoli, E., and Richter, C. (1980). J . Biol. Chern. 255, 9325-9330. Liiw, H.,Crane, F. L., Grebing, C., Tally, M., and Hall, K. (1978). FEES Leu. 91, 166-168. MacDonald, V. M., and Jobsis, F. F. (1976). J . Gen. Physiol. 68, 179-195. Malaisse, W.J., Hutton, J. C., Kawazu, S.,and Sener, A. (1978). Eur. J. Biochem. 87, I2 I - 130. Maran, M., and Dikstein, S. (1979). Microbios Lerr. 11, 55-60. Mukherjee, C., Lynn, W. S.,and Mukherjee, S. P. (1980). J . Cell Biol. 87, H-172a. Mukherjee, S. P., and Lynn, W. S. (1977). Arch. Biochem. Biophys. 184, 69-76. Mukherjee, S. P., and Lynn, W. S . (1979). Biochim. Biophys. Acra 568, 224-233. Mukherjee, S. P., and Mukherjee, C . (1980). J. Cell Biol. 87, H-172a. Mukherjee, S. P., Mukherjee, C., Yeager, M. D., and Lynn, W. S. (1980). Biochem. Biophys. Res. Commun. 94, 682-689. Murad, F.,Arnold, W. P., Mittal, C. K., and Braughler, J. M. (1979). Adv. Cyclic Nucleoride Res. 11, 175-204. Mushlin, P. S . . Rama Sastry, B. V., Boerth, R. C., Surber, M. J., and Landon, E. J. (1978). J . Phurmucol. Exp. Ther. 207, 331-339. Nakamaru, Y.,and Schwartz, A. (1972). J . Gen. Physiol. 59, 22-32. Nath, J., and Rebhun, L. I. (1976). J. Cell Biol. 68, 440-450. Neering, 1. R., and Glover, W. E. (1979). J . Phurmucol. Exp. Ther. 208, 335-339. Nicholls, D. G.,and Brand, M. D. (1980). Biochem. J. 188, 112-118. Patten, G. S., Filsell, 0.H., and Clark, H. G. (1981). FEES Len. 135, 219-222. Popham, P. A. B., Oldershaw, P. J . , and Cameron, I. R. (1980). Clin. Sci. 58, lop. PrpiC, V., and Bygrave, F. L. (1980). J. Biol. Chem. 255, 6193-6199. Puppi, A., Tigyi, A., and Szalay, L. (1968). Acta Biol. Hung. 19, 517-518. Puppi, A. Szalay, L., and Dely, M.(1975). Comp. Biochem. Physiol. 50C, 75-79. Puppi, A., Szalay, L., and Dely, M. (1976). Acra Eiochim. Eiophys. Hung. 11, 63-73. Puppi, A., Dely, M., and Riiger, P.(1980). Gen. Phurmucol. 11, 409-418. Putney, J. W.(1979). Phurmucol. Rev. 30, 209-245. Raad, F., Maran, M., and Dikstein, S. (1974). Comp. Gen. Phurmucol. 5 , 135-144. Robison, G. A. (1980). Trends Phurmucol. Sci. 1, V-XI. Roos, A., and Boron, W. F. (1981). Physiol. Rev. 61, 2%-434. Roth, Z., and Dikstein, S . (1980). Eur. J . Cell Biol. 22, 284. Roth, Z., and Dikstein, S. (1981). Isr. J . Med. Sci. 17, 1107. Roth, Z., and Dikstein, S . (1982). Eiochem. Eiophys. Res. Commun. 105, 991-997. Schwartz, A., Lindenmayer, G. E., and Allen, J. C. (1975). Phurmacol. Rev. 27, 1-134. Shapiro, H. M. (1972). J. Surg. Res. 3, 138-152. Sies, H.,Graf, P.,and Estrela, J. M. (1981). Proc. Natl. Acud. Sci. U.S.A. 78, 3358-3362. Tenny, R. T., Sharbrough, F. W., Anderson, R. E., and Sundt, T. M. (1978). Ann. Neurol. 8, 564-573. Tonomura, Y. (1973). "Muscle Proteins, Muscle Contraction and Cation Transport." Univ, Park Press, Baltimore, Maryland. Vizi, E.S . (1978). Neuroscience 3, 367-384.

STIMULUS-RESPONSE COUPLING AND THE REDOX STATE

61

Wald, C., Czaczkes. W., and Dikstein, S . (1972). Nuturwissenschufen 59, 316-317. Waley, S . G. (1959). Biochem. SOC. Symp. 17th. pp. 79-86. Washio, H., and Mashima, H. (1963). Jpn. J . Physiol. 13, 617-623. Watkins, D. T. (1972). Endocrinology 90, 272-275. Weinstock, M., and Weiss. C. (1979). Br. J . Pharmucol. 65, 593-599. Whikehart, D. R., and Soppet, D. R. (1981). Curr. Eye Res. 1, 451-455. Wilkes, A. B., and Bader, H. (1969). Fed. Proc. Fed. Am. SOC.Exp. Biol. 28, 548. Wins, P. (1969). Arch. Int. Physiol. Biochim. 77, 245-250. Wolkowicz, P. E., and McMillan-Wood, J. (1980). J. Biol. Chem. 255, 10348-10353. Yu,D. H., and Triester, S. (1970). Pharmacologist 12, 238-242. Zimmer, H.-G., and Ibel, H. (1979). FEBS Lett. 106. 335-337. Zimmer, H.-G., Biinger, R., Koschine, H., and Steinkopff, G. (1981). J. Mol. Cell. Curdiol. 13, 53 1-535.

This Page Intentionally Left Blank

INTERNATIONAL REVIEW OF CYTOLOGY, VOL. 85

Regulation of DNA Synthesis in Cultured Rat Hepatoma Cells ROELANDVAN WIJK Department of Molecular Cell Biology, State University, Utrecht, The Netherlands I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Cell Cycle of Hepatoma Cells. . . A. Cell Cycle Kinetics of HTC Cells . . . . . . . . . . . . . . . . . . . . . . . . B. Variation in GI Phase Duration C. Variation in G2 Phase Duration . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . D. Regulation of the Cell Cycle by Serum 111. Factors Influencing DNA Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Initiation of DNA Synthesis . . . . . . . . . . . . . . . . . . . . . B. Later Stages of the Cell Cycle: CAMP.. . . . . . . . . . . . . . . . . . . . IV. V. Polyamine Synthesis . VI . Chromatin Structure an .......... A. General.. . . . . . . . . . . . . . . . . . B. Nucleoskeletal Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . .. C. Histone Acetylation . . . D. Histone Phosphorylation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. ADP-Ribosylation of Histones . . . . . . . . . . . . . H. Replication and Nucleosome Formation. . . . . . . . . . . . . . . . . . . . VII. System of Second Messengers or lntracellular Regulators . . . . . . . . . VIII. Concluding Remarks . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

63 65 65 68 70 70 72 72 78 79 79 80 82 86 86 86 88 90 91 92 93 94 96 98 99

I. Introduction In a growing population of cells cells divide and undergo a new round of DNA synthesis each time cell mass doubles. In today’s biology our understanding of the regulation of initiation of DNA synthesis and cell division in relation to cellular growth is still one of the fundamental problems. It has become clear that knowledge of these regulatory mechanisms is not only crucial to solve specific problems in developmental biology and aging but also for tumor biology. The use of intact organs to study the molecular regulation of cellular growth 63 Copyright 0 1983 by Academic Press. Inc. All rights of reproduction in any form reserved.

ISBN 0-12-364485-2

64

ROELAND VAN WIJK

and initiation of DNA synthesis is limited, although the process of liver development and of liver regeneration has been studied extensively with respect to changes in growth, in initiation of DNA synthesis and division, and in their degree of expression of the various liver functions (Bucher and Malt, 1971; Short et al., 1973; Bucher and Swaffield, 1975). The development of a series of transplantable hepatomas (Moms, 1965) was an important step and offered greater advantage in the study of growth and regulation of DNA synthesis (Weber and Lea, 1967;Weber, 1966, 1968; Lea etal., 1966; Weinhouse, 1972). They have been maintained for a long time and many differentiated functions, characteristicfor mature liver, can be expressed in these tumor cells in a varying degree. Such transplantable hepatomas grow without major changes in their developmental programs which show a strong but not a complete resemblance to that occumng in certain phases of normal embryonic development (Walker and Potter, 1972). The following step, important for the study of growth regulation, was the availability of cultured hepatoma cell lines, mostly derived from (transplantable) tumors. These cell lines have retained the capacity to undergo growth and division as well as parts of their developmental programs reasonably well, in vitro (Wicks et al., 1973a; Volman, 1978; van Rijn et al., 1974; Schamhart et al., 1979). Early examples have been the establishment of the H4IIEC3 cell line from the transplantable Reuber H35 hepatoma (Pitot et al., 1964), the HTC cell line from the transplantable Moms 7288c hepatoma (Thompson et al., 1966), and the MH,C, cell line from the transplantable Moms 7795 hepatoma (Richardson et al., 1969). In cultured cells growth, DNA synthesis, and division can be studied in an experimentally controllable way. Basically, the study of the regulation of growth and initiation of DNA synthesis has been examined using five main approaches: (1) studies of individual mammalian cells to examine the kinetics of the crucial events, (2) the use of growth factors and hormones to analyze specific events in the cell cycle, (3) the use of temperature-sensitive mutants, (4) changes in physiological conditions, and (5) the use of specific inhibitors. It has attracted much attention that individual cells may divide considerably sooner or later than the average interdivisiontime. It is thus not surprising that a considerable amount of data is now available from kinetic analyses of cell cycle variability and of the initiation of DNA synthesis on a wide range of cell types. The ultimate aim of these studies, the demonstration of the critical signal(s) which happen to occur so variable in time even in so-called synchronized cells, is still unsolved although the kinetics of the signal system is now reasonably described. Data obtained from the other approaches are indispensable for the demonstration of the critical signal(s) for growth, for initiation of DNA synthesis, and the division. The necessity of appropriate external stimuli for growth and division enables us to study

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

65

causal relationships and the involvement of metabolic regulatory circuits in the initiation of growth, DNA synthesis, and cell division. There is some evidence that in vitro replication of mammalian cells occurs by orderly expression of sequences of signals and events involving DNA synthesis, doubling of cell mass, and cell division. The sequence of events which involves the “DNA-division cycle” may be partially separable from a second sequence that constitutes the growth cycle. The latter includes the main processes of protein and RNA synthesis which cause cell growth. The degree of separation of these two cycles in mammalian cells is not clear. The role of protein synthesis in initiation of DNA replication will be discussed. A few specific proteins have been shown to be involved in the initiation of DNA synthesis of hepatoma cells. Examples are ornithine decarboxylase and specific chromosomal proteins. It is reasonable to expect that chromosomal proteins and factors influencing modification of these proteins or the interactions between these proteins and DNA play a crucial role in initiation of DNA synthesis. Therefore I will also review the literature concerning the changes of chromatin in hepatoma cells, in relation to the stimulation of these cells into their division. I hope that the hypothetical model present may stimulate further research in this area.

11. Cell Cycle of Hepatoma Cells OF HTC CELLS A. CELLCYCLEKINETICS

A description of cell cycle kinetics in exponentially growing cultures is obtained most readily by analyzing the distribution of intermitotic times within a culture as determined by time lapse cinematography. Appropriate mathematical models had to be used and one of these will be described. For most of the studied cells the variability of cell cycle times is predominated by the time they spend in the G, interval (Baserga, 1965; Mendelsohn, 1965; Mueller, 1971; Prescott, 1968; Till et al., 1964). This does not imply that the remaining phases are of a fixed duration. As an example for the ear and skin epidermis, large variations of G , phase durations have been demonstrated (Bullough, 1963; Sherman et al.. 1961). But the dominant role of the G, interval and the detailed analysis of the variability of generation times have resulted in a model for the cell cycle consisting of one indeterminate phase from which cells proceed to one determinate phase at random with a constant specific rate. As first proposed by Bums and Tannock (1970), the indeterminate phase “Go” was part of G I , and the determinate phase consisted of s,,G,, M, and the remainder of G I . Smith and Martin (1973) proposed a similar model with the nomenclature of A state for the indeter-

66

ROELAND VAN WIJK

minate phase and a deterministic B phase, while the random but constant loss of cells from the A state to the B phase per unit time was termed the transition probability P. According to the original single random transition models, the a and p curves should yield linear and parallel lines with slopes according to the transition probability value P. The a curve is a semilogarithmic plot of the fraction of cells with intermitotic times larger than or equal to a certain time t, as a function oft. The p curve is the semilogarithmicplot of the fraction of siblings having a difference in intermitotic times larger than or equal to an indicated time t, as a function oft. The generality of this simple two parameter model may be open to question since considerable variations in the B phase were observed (Shield and Smith, 1977). This has resulted in a number of proposed modifications to accommodate for experimentally observed discrepancies with the predictions of the model (Brooks et al., 1980; Murphy et al., 1978; Svetina, 1977; van Zoelen et al., 1981). It has been argued that it is difficult to maintain constant environmental conditions, even in a simple tissue culture system with cells grown in monolayer. Critical components can be removed from the medium in a growing population of cells or cells themselves can produce components which can modify their own growth rates, either in a positive or a negative way. Another possibility of change of environmental conditions is when cells come into close contact with one another, showing the phenomenon of contact inhibition of growth (Levine et al., 1965), topo-inhibition of growth (Dulbecco, 1970), or density-dependentinhibition of growth (Stoker and Rubin, 1967). This might result in reduction of available cellular receptors for growth controlling agents (Holley et al., 1977) or alterations in diffusion boundaries at the cellular peripheries (Stoker, 1973). It is therefore of utmost importance to perform the kinetic analysis of cell cycle variability in exponentially growing populations of cells. In our cell cycle studies we have employed HTC cells which were grown in Eagle’s basal medium enriched 4-fold with vitamins and amino acids and buffered at pH 7.4. Fetal calf serum and calf serum were added to a final concentration of 5 and 1096, respectively. Two experimental designs have been used to study the intrapopulation variation in generation times by time lapse cinematography. The first type of analysis (van Wijk et al., 1977a) was performed on cells seeded at “high density” (Fig. 1). The advantage is that a large number of cells can be easily studied at the same time under more or less equal conditions, but only for a limited period of time. The second type of analysis (van Wijk and van de Poll, 1979) was performed on cells seeded at “low density” (Fig. 1). Seeding at a reasonable density leads to the formation of cell aggregates. At densities below a critical value this no longer occurs. At low cell densities a lag period has been observed before cells grow exponentially. The advantage of this procedure is that growth kinetics can be studied at increasing cell density with the recognition of family relationships. The time lapse record-

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

lo3

0

67

40 80 120 160 loo 140 180 320

Time (hr)

FIG. 1. Growth of HTC cells at 37°C after seeding at “high” and “low” densities: 4 x lo3 cells/crn2 (upper curve), lo3 cellslcm2 (middle curve), and 2.5 X 102 cells/cm* (lower curve). Conditions are described by van Wijk er al. (1977a,b) and by van Wijk and van de Poll (1979).

ings of HTC cells showed that division can occur as early as at the age of 13 hours, but the range extends to 22 hours. Despite these large differences in generation time, less difference was found between the generation times of sisters of a pair. The P-plots are reasonably linear for cells studied at “high density,” but tend to become nonlinear at lower densities (Fig. 2). The corresponding a-plots were only slightly changed. In fact, those a-plots show a significant initial downward curvature before linearity is attained. Apparently the distribution of intermitotic times in these hepatoma cultures resembles a normal more than an exponential distribution. The distribution of differences between intermitotic times for siblings obeys an exponential distribution only for cells grown in colony. Cells that have less contact with one another (no compact colony-type growth) show a significant altered slope of the p-curve. Probably

Age difference(hr1

FIG. 2. Age differences of HTC sister cells at time of mitosis in a high density culture or in an extreme sparse culture during clonal growth at third, fourth, and fifth generation after seeding. Conditions and procedures as described by van Wijk and van de Poll (1979).

68

ROELAND VAN WIJK

closely linked to this positive cell-cell influence is the tendency of hepatoma cells to aggregate. One of the striking features of many mammalian cells, including HTC cells, is a mother-daughter relationship (van Wijk and van de Poll, 1979). This result is confirmed for other mammalian cells (Sisken, 1963; Dawson er al., 1965; Collijn-D’Hooghe et al., 1977; Hemon et al., 1978; Miyamoto e? al., 1973) but is not a general phenomenon (Froese, 1964; Absher and Absher, 1976). One would expect a signal controlling proliferation rate to be transferred from the mother cell to the daughter cells. The degree of relationship between mother and daughter cell durations may be influenced by the transition of cells from a state (1) of loose or no colony growth type to a state (2) of strong-coupling or compact colony growth type. Cells which have become strongly coupled remain in this state. From the cells which are less coupled, part of them changes to the compact colony growth type. This has still to be proved and the underlying phenomenon (cell localization, etc.) is yet obscure.

B. VARIATION IN G,PHASEDURATION Data on the duration and variability of the G ,interval in randomly growing cultures can be obtained by the method of Sisken and Kinosita (1961). This method allows the autoradiographic detection of incorporation of [3H]TdR in individual cells whose previous mitotic histories had been recorded by time-lapse cinematography. For HTC hepatoma cells this technique showed large variations in the time required for cells to get through G,and into S phase. The minimal

1, ,id 1

5

,

,

,

10

15

20

25

r)

35

Age (hr)

FIG. 3. Proportion of HTC cells remaining in GIor in interphase 1 against age. A “high” density culture was analyzed according to van Wijk et al. (1977a).

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

69

hr FIG. 4. Nonequal labeled HTC sister cell pairs at various times after mitotic shake off and replating. Percentage labeled cells and percentage nonequal labeled sister pairs were determined according to van Wijk ef al. (1977a).

time for G I is 4-5 hours and the median time is 11 hours (van Wijk et al., 1977a). The variation in individual cell generation times is largely reflected by the variation of individual cells to proceed from mitosis to the initiation of their DNA replication. This was derived from studies with “high density” seeded cells (Fig. 3). Despite this variation there is a striking tendency with respect to the behavior in G I and S of sister cells. Sister cells enter and leave the S phase at approximately equal times, at least in “high density seeded” cultures (Van Wijk et al., 1977a). Data on the duration and variability of the G, interval in sparse cultures can be obtained by following the cell cycle of synchronized-cellpopulations by [3H]TdR

Age (hr)

Comparison of GIvariability in “sparse” and “high” density cultures of HTC cells. Sparse cultures were derived from exponentially growing “high density” cultures by mitotic shake off. FIG.

5.

70

ROELAND VAN WlJK

incorporation. HTC cells can be readily synchronized by selective detachment of mitotic cells without the use of interfering drugs. Sparse cultures were obtained after shaking off the mitotic cells from the “high density” culture, and seeding them at low density in conditioned medium. The cells divide and remain recognizable as sister cell pairs at low density and [3H]TdR incorporation and autoradiography could be performed. Under these conditions a similar minimum G I duration of 4-5 hours was found, as in the high density culture. However, variation in GIphase durations was increased, as well as the percentage of nonequally labeled sister pairs (Fig. 4). These data are in agreement with the density-dependent sister relationship for intermitotic times. Thus, in dense cultures the two sisters of a pair tend to influence one another specifically. In sparse cultures there is no such tendency and the rate of entrance into S phase follows the kinetics of a simple single transition in G I reasonably well (Fig. 5 ) . C. VARIATION IN G, PHASEDURATION The actual observed kinetics imply that in hepatoma HTC cells other portions in addition to the G, phase of the cell cycle can be regulated. Existing data from other cell types also showed that G, could be relatively variable (Guiguet et al., 1980). From FLM curves of HTC cells it is clear that the G, phase is highly variable (van Wijk et al., 1977a) although the establishment of the kinetics of progress of cells through the G, interval needs the use of additional techniques. The technique for determination of the average length and variability of G, phase is by (1) preparation of a marked early G, cell population in a nonsynchronized culture and (2) following this marked population through mitosis. This technique has been described for HTC cells seeded at high density (van Wijk et al., 1977a) and showed nearly linear a-plots for G, durations. Basically, the simplest cell cycle model for HTC hepatoma cells involves two separate regulated mechanisms, one leading to initiation of S phase and the other to the initiation of mitosis. The two signals might be determined by random transitions, but at least one is influenced by cellular relationship during exponential growth. At least two models are able to incorporate the finding that the variability in cycle times was due to both the variability of the G, phase and the GIphase: the growth controlled (GC) model (Alberghina, 1977; Alberghina et al., 1981) and the model of Valleron et al. (1981). It would be beyond the scope of this article to evaluate in more detail the various models.

D. REGULATION OF THE CELLCYCLEBY SERUM The rate of replication of hepatoma cells in tissue culture is modulated by serum. Serum concentrations below 1% result in growth arrest. The role of serum in the regulation of the G, and G, phases has been studied in Reuber €135

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

71

hepatoma cells. The attainment of a lower rate of cellular proliferation after a decrease of serum concentration could be achieved in one of two ways, either (1) by altering the rate of progress of cells through the regulable cell cycle phases G I and G, or (2) by altering the rate within one cell cycle interval, i.e., G I or G,. It appeared that serum concentrations below 1% preferentially arrested cells in G, (van Wijk et al., 1977b). Apparently, only the GI transitory event has an absolute requirement for serum. Removal of serum from an exponentially growing culture of Reuber H35 cells resulted in a G, population of cells and time studies showed only a minor delay of progression through the remaining part of the cell cycle once serum was removed. The compartment of the cell cycle which passed mitosis after serum depletion consisted of cells in G,, S, and cells present in the last 6-7 hours of GI (van Wijk er al., 1979a). Hence, the cells which are in the very early compartment of the G I interval, i.e., cells between mitosis and the time point 6-7 hours before their actual start of DNA synthesis remain arrested in G I . This critical serumdependent event at 6-7 hours before S was demonstrated also in a population of Reuber H35 cells synchronized by serum depletion and then restimulated to proliferate by fresh serum. It becomes of greater interest in a synchronized culture during the second cell cycle with an average G I duration of 3-4 hours (van Wijk et al., 1979a). The time of so-called commitment to DNA synthesis is 6-8 hours, independent of the length of the G I phase and must then occur during the G, period of the previous cell cycle. When the fraction of cells which is initiated to synthesize DNA after readdition of serum was followed in time a general pattern emerged. In the case of serum addition to quiescent hepatoma cells, few cells do not enter the S phase directly, but after a lag phase the labeling index increases abruptly. The duration of the lag phase is independent of serum or amino acids (van Wijk et al., 1979a; Koontz and Iwahashi, 1981; Streumer-Svobodovaet al., 1982). Other manipulations, however, alter the length of the lag phase after addition of serum in the same cell system, (1) keeping Reuber H35 cells for a longer period as a confluent monolayer without a change of medium, and (2) addition of protein synthetic inhibitor lenghtened the lag phase (Streumer-Svobodova et al., 1982). Under optimal conditions the minimal duration of the lag phase is 8 hours, approximately similar to the duration of the commitment period. A second characteristic of the pattern of cells stimulated to initiate DNA synthesis is the distribution of G I phase durations. Although in confluent cultures a wide distribution was observed (van Wijk et al., 1979a,b; Streumer-Svobodova et al., 1982) under optimal conditions of density the distribution of G I phase durations showed first-order kinetics (Koontz and Iwahashi, 1981). First-order kinetics have been observed in many other cell cultures after addition of serum. An example is the extensive experimental work with cultures of baby hamster kidney and Swiss 3T3 fibroblasts (Brooks, 1975; Jimenez de Asua et al., 1977;

72

ROELAND V A N WIJK

Mierzejewski and Rozengurt, 1976) although in these systems failures to observe first-order kinetics have been reported (Brooks, 1975). Two problems have to be solved before a more comprehensive model for the regulation of the hepatoma cell cycle can be presented. The first question involves the sister relationship for G, durations in quiescent Reuber H35 cells after addition of serum. It is tempting to speculate that the relationship in G I times for sister cells in a “dense seeded” culture became lost after deprivation of serum, thus resulting in first-order kinetics. The second question concerns the possible relationship between G, and G, durations. From the kinetic point of view our data would also be in accordance with the following hypothesis. The signal for initiation of S phase is build up starting at the end of the previous cell cycle. It needs the presence of serum for its formation. The signal reaches threshold levels according to first-order kinetics. Independently, the signal for mitosis is formed, also according to first-order kinetics. The resulting distribution of GIphase duration would then deviate from first order depending on the G , kinetics.

111. Factors Influencing DNA Synthesis

A. INITIATIONOF DNA SYNTHESIS The identification of the factors which can modulate growth rates is of major importance in the study of the regulation of growth and initiation of DNA synthesis. The analysis of serum has been disappointing in general. The complexity of serum has caused problems for the isolation of positive growth factors in that several different positive factors can be present in serum. In addition, there might also be components which increase the activity of positive factors but which have little or no effect alone in promoting cell multiplication. It has also been suggested that negative regulations exist in order to maintain a balanced control of cellular proliferation. Studies on the mechanism of action of factors is extremely difficult if one uses unpurified preparations. For instance, the changes suggested in hypothetical negative growth factor activity can be interpreted just as easily on the basis of a change of a positive regulator for cell proliferation. For the purpose of this article I decided to compare the growth-regulating activities found for hepatoma cells with that of the parental liver cells. This might be of interest because some hepatoma cell lines might have lost their responsiveness toward certain growth factors. 1. Hepatic Regenerative Stimulator Substances

The existence of a hepatic regenerative stimulator substance was reported by Labrecque (1980a,b). An extract of normal weanling or adult regenerating rat liver stimulates thymidine incorporation into liver DNA. ln vivo, it was organ

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

73

specific and species nonspecific. It retains organ specificity in experiments with cultured cells, stimulating normal adult hepatocytes in primary culture and HTC and MH,C, hepatoma cell lines. It does not stimulate various non-liver cells. 2. Serum Tripeptide Gly-His-Lys The isolation, purification, and characterization of a serum tripeptide which, in nanomolar concentrations, stimulates DNA, RNA, and protein synthesis in HTC cells was described by Pickart and Thaler (1973). The amino acid analysis suggested that the active factor was a tripeptide formed by glycine, histidine, and lysine. The synthetic glycyl-histidyl-lysine has biological properties (Pickart et al., 1973) similar to the native factor, which was shown to be Hglycyl-histidyl-lysine-OH (Schlessinger et al., 1977). A study on the effects of synthetic analogs of Gly-His-Lys on DNA synthesis in hepatoma cell cultures indicated that optimal activity appears to be associated with the histidyl-lysyl linkage (Pickart and Thaler, 1979). The tripeptide isolated is complexed with copper and iron ions, and acts synergistically with these transition metals to stimulate the growth and metabolism of hepatoma cells maintained in growthlimiting amounts of serum. This tripeptide may function as a chelator of transition metals. The results obtained by Pickart and Thaler (1980) suggest that these metals participate in cell surface-mediated mechanisms involved in cellular aggregation and monolayer formation. The metals in their ionic form appear to be toxic to cells. In addition, growth was not stimulated by Gly-His-Lys introduced separately from the metals. In contrast, Gly-His-Lys metal complexes supported active growth and DNA synthesis. Since these effects of Gly-His-Lys metal complexes on growth were not produced in cultures from which serum had been completely excluded, additional factors in serum must cooperate with chelated transition metals to promote cellular growth. 3. Epidermal Growth Factor (EGF) Epidermal growth factor (EGF) is capable of stimulating DNA synthesis and proliferation in rat AH66 hepatoma cells (Kaneko et al., 1979). Proliferation of AH66 cells in medium supplemented with 0.2% fetal calf serum was almost negligible. The presence of 5 ng/ml EGF resulted in maximal stimulation of cellular proliferation. Similar effects on cell proliferation were observed when the low medium was supplemented with insulin (50 ng/ml) but it is not clear whether the two hormones act additively on hepatoma cells. The effect on AH66 hepatoma cells resembles the stimulation of DNA synthesis by insulin and EGF in adult rat hepatocytes in primary culture (Nakamura et al., 1980; Leffert and Koch, 1982). Mature hepatocytes in primary culture are in a resting state and show neither DNA synthesis nor cell division. DNA synthesis was induced by insulin and EGF. Effect of EGF on proliferation kinetics of neonatal rat hepatocytes in primary culture were shown by Andreis and Armato (1981).

74

ROELAND VAN WIJK

4 . Insulin- and Insulin-like Activities Stimulation of DNA synthesis by insulin was observed in AH66 hepatoma cells (Kaneko et al., 1979). The necessary concentration of insulin in this experiment was comparable with the concentration of insulin affecting the stimulation of DNA synthesis of adult rat hepatocytes. Early studies by Leffert (1974) implicated the role of insulin in quiescent fetal rat hepatocyte growth control. A main conclusion from these studies was that only insulin and somatomedin detectably promoted significant initiation of DNA synthesis. The ability of insulin to act as a growth factor has been further studied in Reuber H35 hepatoma cells (Koontz and Iwahashi, 1981). It is shown that insulin at concentrations within the physiological range and in the absence of any serum factors is capable of triggering serum-depleted cells to traverse G, and enter S phase. The stimulatory activity on DNA synthesis was measured of fractionated polypeptides derived from extracted plasma protein Cohn fraction IV (Pickart and Thaler, 1980). It was compared with insulin-like activity on lipid synthesis and lactate production. Three peptide subfractions stimulate DNA synthesis in HTC cells at concentrations from 0.005 to 0.1 ng/ml. Authentic insulin at these concentrations had no effect on DNA synthesis in hepatoma cells. In the presence of EGF and glucagon, insulin can be replaced to differing degrees by any of three NSILAs including the insulin-like growth factors IGF-I, IGF-11, and somatomedin C to potentiate DNA synthesis in cultures of adult hepatocytes (Leffert and Koch, 1982).

5 . Glucagon The effect of glucagon can be discussed only in relation to the effect of insulin. Basically, glucagon is able to act either as a positive or a negative modulator of the initiation of DNA synthesis (Short et al., 1975; Price, 1976; Takatsuki er al., 1981). It has been demonstrated that at a high level of insulin, DNA synthesis in the regenerating liver was greatly delayed (Price, 1976). At this condition, there was a dose-related increase in the magnitude of DNA synthesis and an early peak response with glucagon supplementation. At the low level of insulin a minimal DNA synthetic response to partial hepatectomy was found in rats. A dose-related increase in DNA synthesis was found as glucagon supplementation was increased. A maximal DNA synthetic response can be reached, while increasing the glucagon level decreased the DNA synthetic response again (Price, 1976). Short et al. (1975) had shown that the glucagon in a stimulatory homione mixture could be completely replaced by Bt,cAMP. This evidence supported a positive role for CAMPin regulating DNA synthesis as was already suggested by MacManus et al. (1972, 1973). Comparative effects were found in cultured cells. Low concentrations of glucagon, when combined with insulin, partially inhibited the insulin-stimulated DNA synthesis in quiescent fetal rat hepatocytes (Leffert, 1974). In support of

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

75

the positive regulatory potential of cAMP in the pre-S phase it was shown that treatment of neonatal hepatocytes with exogenous cAMP stimulates their entry into DNA synthesis (Medoff and Parker, 1971; Armato er al., 1975). The studies on the hormonal regulation of DNA synthesis and the involvement of cAMP in liver cells have been further extended by Friedman et al. (1981) and Boynton er al. (1981). In hepatoma cells no comparable effects of glucagon and/or cAMP in initiation of DNA synthesis have been described. However, an additional effect of glucagon was found at the higher concentrations in liver, which seems to have its counterpart in hepatoma cells. Thus, higher concentrations of glucagon alone reduced the basal DNA synthetic rate. This effect could also be observed with exogenous Bt,cAMP, inhibiting both insulin stimulated and the basal DNA synthetic rate. L-Epinephrine, which, like glucagon, also acted via CAMP-mediated reactions, did not inhibit insulin-stimulated DNA synthesis at low physiological concentrations but showed at higher concentrations effects similar to those seen with high levels of cAMP (Leffert, 1974). Various lines of hepatoma cells have been tested for their sensitivity toward glucagon. There is evidence that the glucagon-sensitive site of adenylate cyclase has been altered (Wicks et al., 1973a; Wimalasena et al., 1980). Nevertheless in some hepatoma cells the secondary Bt,cAMP effect, the cAMP inhibition of the actual DNA synthesis, was well described (van Wijk et al., 1972, 1973). Since in this part, we only describe modulations of the initiation of DNA synthesis, the effect of cAMP on actual DNA synthesis will be discussed later (Section 111,B). 6. Glucocorticoids Examination of the effects of glucocorticoid hormones on DNA synthesis in cell lines of liver origin started with Raab and Webb (1969) and Loeb et al. (1973). The addition of hydrocortisone resulted, after a lag of 6-12 hours, in a progressive inhibition of DNA synthesis. In addition, the presence of the hormone resulted in inhibition of cell proliferation. In a study comparing cell lines derived from rat liver and hepatomas, a division could be made into two groups regarding the sensitivity toward glucocorticoid inhibition of cell growth (Taira and Terayama, 1978; Venetianer et al. 1980). The sensitive cells and insensitive cells seem to be different in the dexamethasone binding to nuclei in cells. Barnett et al. (1 979) reported the selection of dexamethasone-resistant(Reuber H35) rat hepatoma cells by exposure of Reuber H35 cultures to dexamethasone. This leads to a significant decrease in both rate of DNA synthesis and proliferation, but continuous culturing resulted in the slow selection of a resistant cell population. In this case the dexamethasone binding to nuclei was not determined but the resistance to dexamethasone-mediated growth inhibition was not accompanied by resistance to steroid-mediated induction of tyrosine aminotransferase. When Reuber H35 hepatoma cells were exposed to both progesterone and cor-

76

ROELAND VAN WIJK

tocosterone, progesterone could influence the inhibition of hepatoma cell DNA synthesis by corticosterone (Desser-Wiest, 1976). Progesterone is known to compete with glucocorticoids in binding to cytoplasmic receptor proteins. It is clear that additional experimental evidence has to be presented in order to elucidate the nature of the differential sensitivity for dexamethasone, especially its relationship to other growth factors. The effect of corticosteroid in the cell cycle has been investigated for Reuber H35 cells (Desser-Wiest, 1975). The conclusion drawn was that corticosteroneis able to block cells mainly in the G, phase of their cycle. 7. Liver Chalone Activity It has been proposed for a long time that mitotic activity in a number of tissues is regulated by endogenous negative feedback inhibitors, described originally as chalones (Bullough and Laurence, 1960). By definition a chalone is secreted by the same tissue on which it specifically acts; it is species nonspecific, inhibits mitotic activity both in vitro and in vivo, and is noncytotoxic (Maugh, 1972). Several experiments have been interpreted as showing the existence of a hepatic chalone. Stick and Florian (1958) observed that an intraperitoneal injection of rat plasma to a subtotally hepatectomized rat inhibits mitosis in the liver remnants of the recipient animal when the plasma donor has an intact liver but not when the donor is also hepatectomized. This humoral inhibitory fraction might originate from the liver. Saetren (1956) demonstrated that an intraperitoneal injection of a whole liver macerate decreases the mitotic index in the regenerating liver or the liver of young animals. Verly’s group purified the liver chalone activity from a homogenate of rabbit liver (Verly et al., 1971; Verly, 1973). The active fraction showed an effect on DNA synthesis in slices of rat liver and it has a molecular weight of 2000. Vinet and Verly (1976) isolated an additional factor from rabbit liver inhibiting DNA synthesis. This factor had a molecular weight of 40,000 and has been shown to inhibit DNA synthesis in Novikoff cells, in regenerating liver slices, and in cultivated normal hepatocytes (Vinet and Verly , 1976; Verly, 1976). Higuret et af. (1975) also obtained from normal rat liver homogenate an active protein fraction with molecular weight of 80,OOO inhibiting the DNA synthesis in liver cells in vitro and in vivo. In addition, they found that the perfusate of normal rat liver has an in vitro and in vivo inhibitory activity that can be localized in a fraction indicating an active molecule of -80,000 daltons (Molimard et af., 1975). It has been shown by Pietu er af. (1978) with immunological means that common inhibitors exist in the fraction isolated from the homogenate and the perfusate. Kuo and Yo0 (1977) demonstrated that rat liver extracts were also effective inhibitors of rat hepatoma 7777 cells in vitro. The factors from rat liver and hepatomas which inhibit thymidine incorporation into DNA of Novikoff hepatoma cells were found to cochromatograph with arginase. However, inhibition is not the result of this activity (Barra er af., 1979).

-

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

77

The determination of the phase of the cell cycle that was blocked by these substances was performed by Aujard et al. (1973). They have shown that the inhibition of rat liver extracts on cultured LF hepatoma cells is not due to a direct action of the liver extracts in S phase, but is the consequence of their action in G , phase. Simard et al. (1974) and Deschamps and Verly (1975) also showed that the inhibition of thymidine incorporation resulted chiefly from a block in the G,-S transition with little inhibitory effect on hepatocytes that were already in the S phase of the cell cycle. A G , effective inhibition of thymidine incorporation into rat liver DNA was also found for the factor purified from rat liver (Sekas and Cook, 1976). De Paermentier et al. (1979) presented interesting additional information on the effect of adult rat liver extract on Reuber H35 cells. Entering of these cells into S phase was prevented but the treated cells also showed some cellular differentiation. Extracts from adult rat liver after partial hepatectomy have no effect on cell multiplication and DNA synthesis and no morphological modifications are observed. In only a few studies the effect of liver cytosol preparations was tested on labeled amino acid incorporation into protein. These studies, however, are still difficult to interpret because of the presence of different inhibitory fractions with a different degree of specificity toward the types of macromolecular synthesis (Nilson, 1976; Barra et al., 1980). In conclusion, the study of the factors influencing initiation of DNA synthesis and growth has progressed, but interpretation of the results is very difficult and the degree of progress is disappointing. But it remains of primary interest for the study of the regulation of growth under physiological conditions. In this respect one must ultimately reach the goal of a classification of the factors according to two parameters: (1) on the basis of their specificity for the target tissue, and (2) on the basis of their source, i.e., the organ(s) of origin. These two parameters are closely linked in the case of the hepatic chalones (as negative regulator) or the hepatic regenerative stimulator substance (as positive regulator). Coupling of these parameters to a lesser extent can be found for growth factors like EGF and the insulin-like growth factors. They can be produced by the same or organs other than the target organ(s) and they have a broader, less cell-specific action. A third class includes the factors, mostly hormones, which show most of their effect on metabolism but have additional growth-modulator effects, either positive (insulin) or negative (corticosteroids) or are influenced by further interactions (glucagon). Before a study of the mechanism and action can be undertaken the factors have to be considered pure or nearly pure. They have to be carefully tested on the stimulation of the fraction of cells synthesizing DNA in a given time after addition of the test material. In a number of studies only TdR incorporation was used without any additive determination of the fraction of cells synthesizing DNA. In these cases effects on thymidine uptake or effects on intracellular thymidine metabolism can obscure the observed degree of growth modulation.

78

ROELAND V A N WIJK

In the study of the growth factors and modulators and their interactions one has to be aware of a few basic problems. They deal with the limitations of the test system. In many respects the method used to assay a factor involves (1) allowing the cells to attain a very low proliferative rate by lowering the serum concentration, and (2) testing the growth factor for its ability to increase the fraction of cells synthesizing DNA. But if additional components in serum (necessary for maintaining metabolism) are also lost, then the potential growth factor may be ineffective. The growth-stimulatingtripeptide might be an example. Finally, one has to take into account the loss of sensitivity for growth factors and modulators in some of the test systems. Hepatoma cells might respond differently from parental adult liver cells. Undoubtedly, it will be of advantage to use different systems, including hepatocytes and hepatoma cells. B. LATERSTAGESOF THE CELLCYCLE:cAMP The growth rate of Reuber H35 hepatoma cells has been shown to be inhibited by Bt,cAMP. The effect is concentration dependent and readily reversible (Van Wijk et al., 1972). Cyclic AMP itself was without any growth inhibitory effect. However, inhibition of proliferation of a selected clone (KRC) of Reuber H35 cells could also be obtained with cAMP at concentrations higher than M. The inhibition of the cell proliferation appears to be mediated by the action of cAMP itself as the effect is potentiated by the phosphodiesterase inhibitor 1methyl-3-isobutyl-xanthine (Van Meeteren et al., 1983). Careful analyses had to be performed, however, since cAMP can give rise to cytotoxicity caused from breakdown products by the serum (Hargroves and Granner, 1982; Van Meeteren er al., 1983). It is of interest that among other (cyclic) nucleotides tested only compounds capable of inducing tyrosine aminotransferaseare effective inhibitors of growth (van Wijk et al., 1972; Wicks er al., 1973a). Elevation of tyrosine aminotransferase activity by cAMP derivatives in rat liver and cultured hepatoma cells has been established as a process requiring protein synthesis (Barnett and Wicks, 1971; Snoek et al., 1981a,b, 1982). In extensive studies on the regulation of tyrosine aminotransferase by cyclic nucleotides it was demonstrated that enzyme induction is tightly coupled to protein kinase activation. Thus, a variety of cyclic nucleotide derivatives were tested for their ability to induce tyrosine aminotransferase in intact Reuber H35 cells and to activate CAMP-dependent histone H, protein kinase in the intact cell (Wimalasena and Wicks, 1979; Wimalasena et al., 1980; Wicks et al., 1975) and in vitro (Wagner et al., 1975). The results provide suggestive evidence that protein kinase mediates the effects of cAMP on tyrosine aminotransferase synthesis and growth regulation. The nature of the putative substrate(s) for protein kinase that might influence growth is unknown at present. The rate of protein synthesis was not inhibited during growth of Reuber H35 cells in the presence of BbcAMP. In contrast, DNA

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

79

synthesis was markedly inhibited shortly after the addition of Bt,cAMP (van Wijk et al., 1972, 1973; Wicks et al., 1973b) or cAMP (Van Meeteren et al., 1982). Although under conditions of growth the cyclic nucleotide inhibits DNA synthesis without any appreciable effect on the duration of G,, cells in progression through the G, phase could be strongly inhibited by cAMP after they were irradiated (Van Meeteren et al., 1983; van Rijn et al., 1983). Thus, one would expect logical candidate(s) for the kinase substrate on the level of chromatin. Potential sites of phosphorylation might include histone and nonhistone chromosomal proteins. Different hepatoma cell lines possess different patterns of growth regulation and enzyme regulation by Bt,cAMP. Bt,cAMP inhibited the growth rate and DNA synthesis of Reuber H35 cells and MH,C, cells but not that of HTC or RLC cells (Van Rijn et al., 1974). The reason that Bt,cAMP is only active in Reuber H35 cells and MH,C, cells can be explained on the basis of differences in ( I ) cyclic AMP metabolism or (2) protein kinase activities. The possibility that different metabolism of Bt,cAMP in HTC and RLC is responsible for the insensitivity of both cell lines was excluded by Bevers et al. (1976). A second possibility is the lack of protein kinase activity in insensitive cells. CAMPdependent protein kinases present in the cytosol fraction of rat liver and hepatoma cells have been characterized and the predominance of type I CAMPdependent protein kinase in rat liver and the predominance of type I1 CAMPdependent protein kinase in the hepatoma cell line HTC (Granner, 1972, 1974; Liu, 1980), Reuber H35, and MH,C, (Liu, 1980) were established. Apparently, a specific protein kinase not yet detected, the alteration or absence of kinase substrate, or the production of a modulating principle (see also Section IV) is involved in regulation of DNA synthesis by CAMP.

IV. Protein Synthesis and Initiation of DNA Synthesis A. EARLYSTIMULATION OF PROTEIN SYNTHESIS

Since in a growing population the cells divide and undergo a new round of DNA synthesis each time cell mass doubles, the control of the rate of proliferation and of protein production needs to be coordinated. Streumer-Svobodova et al. (1982) showed that serum addition to cultures of quiescent Reuber H35 cells leads to increase in protein synthesis preceding the initiation of DNA synthesis and progression of cells through the cell cycle. The stimulation of initiation of DNA synthesis is proportional to the stimulation of protein synthesis, as influenced by either serum or amino acid concentration. Only a few studies have been performed to examine the increase in rate of protein synthesis by addition of

80

ROELAND VAN WIJK

growth factors. Pickart and Thaler (1973) showed that the serum tripeptide Gly-His-Lys was able to stimulate DNA synthesis and protein synthesis in HTC cells maintained in growth-limiting amounts of serum. The role of the serum-stimulated protein synthesis in the entry of quiescent Reuber H35 cells into S phase has been studied in more detail because of the observation in hepatoma cells (van Wijk et al., 1979a) and many other cell lines (Todaro er al., 1965; Yoshikura and Hirokawa, 1968; Temin, 1971; Pardee, 1974; Brooks, 1976; Lindgren er al., 1975) that serum needs to be present only for a limited and specific period of time for commitment to DNA replication and proliferation. In accordance with the proposed role of protein synthesis the increased protein synthesis in Reuber H35 cells is important only during the first period of stimulation after serum readdition, as a signal for DNA replication and mitosis (Streumer-Svobodovaer al., 1982). Protein synthesis in the second (i.e., serum-insensitive)period can be partly inhibited without any effect on the progress of cells through their cycle. This was demonstrated by removal of serum as well as by the differential effect of low cycloheximide concentrations on different cell cycle phases. Protein synthesis did not increase continuously during the cell cycle following serum readdition. After its initial rapid increase the rate of protein synthesis increased steadily during the first 9 hours after the addition of serum. The rate of protein synthesis remains more or less constant during the second part of the cell cycle (van Wijk, unpublished results). Protein synthesis in hepatoma cells during progress through their cell cycle has also been studied for cells which were synchronized by selective detachment of mitotic cells. Such studies were performed with HTC cells (Martin el al., 1969a,b; Sellers and Granner, 1974; Emanuel and Gelehrter, 1975; van de Poll er al., 1979) or RLC cells (Sellers and Granner, 1974). In these cultures, rate of protein synthesis, expressed per milligram of cellular protein, remains constant during the cell cycle with the exception of a substantially lowered rate of protein synthesis during the stage of mitosis. Apparently the depletion of exponentially grown cells for serum and the subsequent stimulation of these quiescent cells with serum resulted in an unbalanced stimulation of protein synthesis. B. LEVELSREGULATING RATEOF PROTEIN SYNTHESIS The mechanism of the control of the rate of protein synthesis after serum readdition to quiescent cells is only partly understood. Streumer-Svobodova er al. (1982) demonstrated that the early increase in protein synthesis was initially actinomycin D independent and occurred at a posttranscriptional level. Basically, such increase might be due to (1) a relative increase of translation rate of the available messenger RNAs or (2) an increase in the number of ribosomes translating (the same amount of) messenger RNA at an unchanged translation rate. To discriminate between these possibilities the change of polysome size, the number of growing polypeptide chains, and the ribosomal transit time were determined

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

81

after stimulation of cells by serum (Streumer-Svobodovaet al., 1982). In quiescent cultures of Reuber H35 cells there is a large portion of monosomes and a small number of large polysomes. After 3 hours of stimulation of the cells by serum most of the ribosomes are mobilized into larger polysomes whereas the total amount of ribosomes had not changed. Moreover additional nascent chains were present in cells stimulated by serum. The ribosome transit time was found to be similar in quiescent and serum-stimulated cells. Thus, the messenger RNAs are translated by a higher number of ribosomes, whereas the rate of peptide elongation is unchanged. Most likely the first increase in rate of protein synthesis after serum stimulation is due to an increased probability of attachment of messenger RNA to ribosomes. In the second part of the cell cycle the increased rate of protein synthesis is maintained but partly due to the production of more ribosomes. Polysome size was decreased at that time and protein chain elongation rates are relatively unchanged (van Wijk, 1983). This can be explained by (1) decreased probability of initiation per messenger RNA or (2) altered content of ribosomal RNA relative to that of total mRNA. Faliks and Meyuhas (1982) studied the regulation of the messenger RNA for ribosomal proteins (rp) in rat liver cells that were stimulated to grow after partial hepatectomy. The content of the rp mRNA species relative to that of total mRNA increased significantly after growth stimulation. Since newly formed ribosomal proteins are essential for the processing of ribosomal precursor RNA, this might explain the early increased content of ribosomes relative to that of total messenger RNA. However, the extent of regulation of rp mRNA content varies somewhat in different cells under different physiological conditions (Geyer et al., 1982) and more data are required to clarify this issue for hepatoma cells. The rate of synthesis of ribosomal proteins relative to total cellular proteins has been determined in regenerating liver (Nabeshima and Ogata, 1980). Growth stimulation resulted in a significant increase in the rate of synthesis of ribosomal proteins, concomitant with an increase in the amounts of effective mRNAs for ribosomal proteins. But during the early steps of the transition of quiescence to growth ribosomal protein synthesis may be controlled by an alteration of the efficiency of translation of rp mRNA as suggested for mouse fibroblasts (Tushinski and Warner, 1982; Geyer et al., 1982). In summary, since with various cell types a number of workers have observed that the ribosome content of cells at quiescence is only half that in the exponential phase, the involvement of protein synthesis in growth after serum readdition might occur in two steps. The first step involves increased efficiency of translation of existing mRNA, resulting among others in increased production of ribosomal proteins. The increased ribosomal proteins are then involved in the processing of ribosomal precursor RNA and ribosome production. At a later stage the rate of protein synthesis is determined by the rate of the combined production of more ribosomes and messenger RNA. In hepatoma cells the mechanism which causes increased rate of initiation of

82

ROELAND VAN WIJK

protein synthesis is essentially unknown. Data obtained from other mammalian cell types suggest various mechanisms for the shift of 80 S monosomes into translating polysomes. 1. Regulation of eIF-2 activity by means of a recycling factor eRF. eIF-2 binds met-tRNA and GTP into a ternary complex whereafter it is assembled into a 40 S preinitiation complex. Regulation of the activity takes place by phosphorylation of eIF-2, which inactivates eIF-2 (Amesz et al., 1979; Goumans, 1981; Safer and Jagus, 1981; Voorma et al., 1983). The mechanism underlying the inactivation appears to be a very stable complex of phosphorylated eIF-2 with GDP which is formed upon formation of the 80 S initiation complex. Whereas GDP can exchange with GTP in a complex with nonphosphorylated eIF-2 in the presence of eRF, the exchange does not occur with phosphorylated eIF-2. This is an example of quantitative regulation of the production of protein. 2. Regulation of eIF-4B activity. eIF-4B plays an important role in messenger RNA binding together with eIF-4A and cap binding protein. The mode of action has been recently reviewed (Van Steeg, 1982). The latter protein recognizes the cap structure of eukaryotic mRNAs, i.e., m7G5’ppp5’X. Although the precise role of each factor is unknown one assumes that they all play a role in the melting of the secondary structure present in the leader sequence of the messenger. Inactivation of eIF-4B for instance should then favor the translation of messengers with a low content of secondary structure. From these messengers it is known that they have a low dependence on eIF-4A, eIF-4B, and eIF-4E cap binding protein, and ATP. ATP hydrolysis appears to be required for the melting of the secondary structure, a process similar to the unwinding of DNA during replication. From neuroblastoma cells grown under low serum conditions an inactivation of eIF-4B has been reported (H. van Steeg, personal communication). 3. Phosphorylation of ribosomal protein s6 appears to give the 40 S subunit a kinetic advantage in an initiation step. Phosphorylation of S , is a prerequisite for the shift of 80 S monosomes into polysomes (Thomas et al., 1980) and serum or growth factors stimulate the s6 phosphorylation as well as the shift in polyribosome profile. The mechanism is not understood since in all kinds of model assay systems this effect could not be repeated.

V. Polyamine Synthesis Polyamine synthesis is one of the molecular events that are stimulated during the G, period, and that has been attributed a role in the initiation of DNA replication. When HTC cells were stimulated to proliferate from a quiescent state, an increase in the ornithine decarboxylase activity appears to be one of the

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

83

very first changes taking place (McCann et al., 1977a, 1979a; Mamont et al., 1978a,b). The enzyme activity reaches a first peak in early GI and a second peak at the time of initiation of DNA synthesis. The increased ornithine decarboxylase activity is followed by intracellular accumulation of putrescine and spennidine, but not of spermine (Mamont et al., 1978a,b). The early increase in ornithine decarboxylase activity and polyamine synthesis following growth stimulation might be part of the change necessary for making a quiescent cell enter the cell cycle. The temporal correlation between the late G,-early S phase stimulation of ornithine decarboxylase activity and the time of onset of DNA replication suggest that late GI stimulation of polyamine synthesis is involved in the process of initiation of DNA synthesis. The obvious means of elucidating whether polyamines are absolutely required is to specifically deplete the cells of their polyamines with the use of specific inhibitors of the polyamine biosynthetic enzymes. In HTC cells stimulated to synthesize DNA, both a-methylomithine and a-difluoromethylornithineblock the usual increases of putrescine and spermidine. The results obtained with these inhibitors suggest that inhibition of polyamine accumulation does not affect the first traverse of the HTC cell cycle. This was confirmed by measurements of DNA synthesis and of the proportion of cells undergoing cell division (Mamont et al., 1976, 1978a). Once putrescine and spermidine depletion is achieved, within one generation, a striking decrease in the rate of DNA synthesis and cell multiplication occurs. Mamont et al. (1978b) isolated an a-methylornithineresistant (HMO,) cell line of HTC, under the selective pressure of a-methylornithine. The prolonged resistance of HMO, cells to the antiproliferativeeffect of a-rnethylornithine and a-difluoromethylornithine in comparision with parental HTC cells was explained by the marked increase in ornithine decarboxylase, putrescine, and spermidine levels. In the presence of the inhibitors an immediate decrease in putrescine and spermidine levels was observed but because of the increased initial levels extensive depletion of these polyamines was achieved at later times. This coincides with prolonged resistance to the effect of inhibitors. The inhibitory effects of a-methy lomithine and a-difluoromethylornithine on HTC cell growth were reversed by addition of polyamines. Mamont et al. (1978c, 1980) demonstrated that putrescine, spermidine, and spennine almost completely reversed the inhibition of cell proliferation brought about by incubation of HTC cells in a-difluoromethylornithine. Results in liver are, at least to some extent, comparable with those obtained for hepatoma cells. Thus, injection of 1,3-diaminopropane in rats after partial hepatectomy causes inhibition of DNA synthesis, represses ornithine decarboxylase, and prevents the increases in putrescine and spermidine concentrations that normally occur in the regenerating liver (Kato et al., 1978; Poso and Janne, 1976; Kallio et al., 1977; Wiegand and Pegg, 1978). Poso and Pegg (1982) studied the effect of a-difluoromethylornithine on polyamine and DNA synthesis in regenerating liver and demonstrated

84

ROELAND VAN WlJK

the reversal of inhibition of DNA synthesis by putrescine. Basically all experiments indicate that a normal polyamine complement is essential for the progression of cells through their life cycle. These studies also show that adult rat hepatocytes in vivo do not appear to contain sufficient polyamine to enter a cycle in the absence of the de novo polyamine synthesis. This is in contrast to the results obtained with HTC cells and with cultured fetal hepatocytes (Rupniak and Paul, 1978). The regulation of ornithine decarboxylase activity is very complex. It is dependent on intracellular and extracellular factors. With respect to the ornithine decarboxylase activity and changes of polyamine levels during G , phase, the regulation of the enzyme by polyamines might be of particular interest. Putrescine has been reported to dramatically inhibit the increase in ornithine decarboxylase in the rat liver after stimulation by growth hormone or partial hepatectomy (Schrock et al., 1970; Janne and Holtta, 1974). Evidence indicates that putrescine, spermidine, or spermine in Reuber H35 cells and rat liver promotes the synthesis of a heat-labile protein inhibitor to ornithine decarboxylase (Fong et al., 1976; Heller et al., 1976; Clark and Fuller, 1976; McCann et al., 1977a). This so-called ornithine decarboxylase antizyme is found after induction by polyamines and is bound tightly to ornithine decarboxylase. It exists also in cells, uninduced for the synthesis of ornithine decarboxylase antizyme, but in these cases it exists in a particle-bound form as identified in subcellular fractions of Reuber H35 cells and of rat liver (Heller et al., 1977a). Further studies on the relation of antizyme induction to ornithine decarboxylase activity regulation demonstrated that in Reuber H35 cells (Heller et al., 1977b, 1978) and HTC cells (McCann er al., 1979b) at least two mechanisms exist for regulation of ornithine decarboxylase by polyamines. Low concentrations of polyamines inhibit induction of ornithine decarboxylase without affecting the antizyme. The antizyme can be induced by high polyamine concentrations. It was further suggested that intracellular and exogenous polyamines regulate ornithine decarboxylase by these two different mechanisms via two different sites. A sensitive membrane-mediated site might respond to minute fluctuations of extracellular polyamine levels. A coarse site which responds to high polyamine levels may be intracellular, responding to the (larger) fluctuations of intracellular polyamines. Chen et al. (1976a) have presented some experimental evidence for the presence of membrane-associated sites that affect ornithine decarboxylase activity. Agents known to affect the membrane via the cytoskeleton (colchicine, cytochalasin, vinblastine) inhibit the induction of ornithine decarboxylase. Moreover, small increases of Na , K , or Mg2 in the growth medium eliminated the anticipated increase of ornithine decarboxylase activity that normally occurred upon the dilution of cells from the stationary phase. Recent studies with non-hepatoma cells further substantiate the notion that ornithine decarboxylase can be controlled at the plasma membrane level (Chen et al., 1982). The role of ion +

+

+

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

85

metabolism in relation to the induction of ornithine decarboxylase is very complicated. A role of calcium has been proposed in the induction of ornithine decarboxylase in rat liver HTC cells (Canellakis et al., 1981) (see Section VII). A further development in the study on regulation of ornithine decarboxylase by polyamines is the observation that acetylated polyamines, N-monoacetylputrescine and P-monoacetylspermidine activate ornithine decarboxylase of HTC cells (Canellakis, 1981). This is in sharp contrast to the suppressing effects of nonacetylated polyamines. The role of polyamine acetylation is not quite clear and two functions can be assumed to be likely (Blankenship and Walle, 1972). Since the acetylspermidine and monoacetylputrescine are actively excreted it is likely that polyamine acetylation participates in the polyamine turnover. Another possibility is the control of interaction between polyamines and molecular anionic sites. Positive changes from the polyamines are removed by acetylation. Acetylation increases the lipophilicity of the polyamine molecules and may facilitate their transport through lipid layers. It is also expected to diminish the polyamine functioning in the interaction between polynucleotide helices (Tsuboi, 1964; Bloomfield and Wilson, 1981) or the binding of polyamines to RNA (Cohen, 1971, 1978; Loftfield et al., 1981). Libly (1978) has shown that the histone- and polyamine-acetylating activities reside on the same molecule. Histone and polyamine acetylation have been suggested to be coupled events. This might be important for DNA structure and function but also in the temporal excretion of a molecule which can trigger events in neighboring cells. The question whether induction of polyamines does occur only as a specific response to the various growth factors must be answered negatively. It is evident that hormones with different mechanisms of action affect the induction of ornithine decarboxylase. The “trophic hormones” which interact with membrane receptors resulting in the activation of adenylate cyclase, cAMP synthesis, and activation of CAMP-dependent protein kinases are well known inducers. Thus, ornithine decarboxlyase activity was found to be increased by glucagon in liver (Holtta and Raina, 1973; Panko and Kenney, 1971), perfused liver (Mallete and Exton, 1973; Manen et al., 1977), or cultured liver cells (Byus et al., 1976a). The activity was also increased by cAMP analogs or phosphodiesteraseinhibitors in the hepatoma cells Reuber H35 (Byus et al., 1976b) and HTC (Canellakis and Theoharides, 1976). The steroid hormones are also able to cause ornithine decarboxylase induction. Dexamethasone has been reported to cause induction in liver (Panko and Kenney, 1971; Beck et al., 1972), HTC cells (Theoharides and Canellakis, 1975), and Reuber H35 cells (Byus et al., 1976b). Hormones with membrane receptors which are not coupled to adenylate cyclase activation but characterized by their migration to coated pits can also induce ornithine decarboxylase activity. As an example the induction by epidermal growth factor can be mentioned (Nakamura et al., 1980; Moriarity et al., 1980, 1981). From these data it is apparent that no tight correlation exists between the effect of hormones

86

ROELAND VAN WIJK

on growth (see Section 111) and the effect on ornithine decarboxylase induction. This is further demonstrated by Byus et al. (1976b). They found that insulin did not result in any increase in ornithine decarboxylase activity in H35 cells, while it stimulates the entry of quiescent cells into DNA synthesis without subsequent proliferation (Koontz and Iwahashi, 1981). However, insulin has been shown to stimulate ornithine decarboxylase activity in rat liver (Panko and Kenney, 1971; Mallete and Exton, 1973) and in primary cultured hepatocytes (Pariza et al., 1974; Nakamura et al., 1980). Apparently a second pathway independent of polyamines but dependent on insulin is in control of DNA synthesis in resting hepatoma cells.

VI. Chromatin Structure and Its Changes during the Cell Cycle A. GENERAL For the initiation (and continuation) of replication the chromatin must be in a decondensed state. A local change, at the level of the nucleosome structure, must occur during the assembly of histones into replicated chromatin. For transcription it is necessary to postulate that either histone-histone or histone-DNA interactions are at least temporarily interrupted. During replication even a greater change must occur. It is a complex process, involving association of both preexisting and newly synthesized chromatin proteins with daughter DNA duplexes. The different orders of chromatin structure are generated by interactions involving histones and subsets of nonhistone chromosomal proteins. In the regulation of chromatin structure a role has been proposed for the reversible chemical modifications of histones and nonhistone proteins. Chemical modifications by ADP ribosylation, methylation, acetylation, and phosphorylation have been described. Such chemical modifications could affect the contiguity of the nucleosomes, the degree of condensation, and therefore the capacity for transcription and replication of the DNA. In this context it is also significant that nuclear DNA is apparently attached to these nuclear proteins in the configuration of a matrix structure. This is not a fixed structure but attachment might be related to active DNA, i.e., DNA might be most active at the complexes anchored to the nuclear matrix. In order to understand the regulation of the initiation of DNA replication, we need to understand chromatin structure in cells during their progression from the resting to proliferative state.

B. NUCLEOSKELETAL MATRIX Recent work has postulated that DNA sequences involved in DNA synthesis were bound to supporting elements. Supporting elements would have a great

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

87

utility in eukaryotic cells where large amounts of DNA must be ordered spatially during replication such that daughter strands remain untangled and yet coupled during their later condensation and entry into mitosis. Although no preferential association of replicating DNA with the nuclear membrane was found by most investigators (Lewin, 1973) recent work has shown that structural elements exist in the nucleus in addition to the nuclear membrane. The underlying skeleton is often called the nuclear matrix and has been described for a variety of eukaryotic cells, among which is liver cells. The matrix is revealed by extraction of purified nuclei with detergents, hypotonic and low magnesium containing buffers, and high salt, and subsequent digestion with nucleases. The resultant structure retains the size and shape of the original liver nuclei and can be seen to contain a residual peripheral lamina with nuclear pore complexes, residual nucleoli, and an internal fibrogranular network (Berezney and Coffey, 1974, 1977; Berezney, 1979). In nuclei of dividing liver cells, approximately 1-2% of the total nuclear DNA remains anchored to the proteinaceous nucleoskeletal matrix. A special character of this DNA has been postulated since it became obvious that the DNA of replicating chromatin is preferentially anchored to the matrix (Berezney and Coffey, 1975, 1976; Berezney, 1977). Thus, the matrix attached DNA is initially enriched with in vivo pulse-labeled replicated DNA. The percentage of replicated DNA which is anchored to the matrix rapidly declines from approximately 50% after a 1-minute pulse to 5% after a 10-minute pulse (Berezney and Buchholtz, 1981). These data agree with the average rate of in vivo replication for movement of 3.2 kb of DNA per minute and the average fragment size of 1.6 kb of DNA anchored to the matrix (Lynch ef al., 1972). This result strongly implicates the matrix as the major intranuclear site for liver cell DNA replication. Two distinct structural models have been envisaged, illustrating this concept of matrix bound replicational complex (Berezney and Buchholtz, 1981). The fixed matrix model and an alternative model, termed the sliding matrix model, have been developed. These models differ with respect to the matrix bound replicational complexes in relation to the original matrix DNA attachment sites, but it is impossible at the moment to distinguish between these models. It must be kept in mind that the structure of DNA anchored to the proteinaceous nucleoskeletal matrix is not unique for chromatin in DNA replication. It is present throughout the cell cycle. The study of the proteins involved in this structure is difficult. We can distinguish between studies involving (1) selective proteins involved in structuring DNA and (2) modifications in proteins or protein structures associated with the attachment of DNA. Although no definite literature from liver or hepatoma cells is available on proteins involved in organization of DNA it is found that residual protein scaffolds can be isolated from nuclease digested interphase nuclei (Lebkowski and Laemmli, 1982b) and from the most condensed chromatin structure in metaphase chromosomes (Paulson and Laemmli, 1977) of HeLa cells. It is interesting that in the most condensed structure

88

ROELAND VAN WIJK

loop sizes are in a range which is comparable with the distribution of S phase replicon length and the average size of matrix attached replicating DNA loops (Huberman and Riggs, 1968). In the histone-depleted HeLa interphase nuclei evidence has been presented for two levels of folding of DNA. Stabilization of the folding was dependent on calcium and copper ions. Removal of the protein bound metal leads to partial DNA relaxation (Lebkowski and Laemmli, 1982a). The proteins associated with both types of histone-depleted nuclei are identified (Lebkowski and Laemmli, 1982b), as well as those associated with the scaffold structure of HeLa metaphase chromosomes (Lewis and Laemmli, 1982). Although this type of study has not yet been performed with hepatoma cells, some progress is made on (the changes in) proteins associated with the attachment of DNA. For liver cells it was shown that nonhistone polypeptides in the same molecular weight range as the major matrix polypeptides are major protein components associated with the attached DNA (Berezney, 1977). For these studies DNase I and micrococcal nuclease have been used extensively as probes. It has been shown that micrococcal nuclease preferentially releases active parts of chromatin of various systems, as was the case for DNase I. It is further shown that DNA of replicating chromatin is preferentially degraded by digestion of nuclei with exogenous DNase I or micrococcal nuclease (Berezney and Buchholtz, 1981). In the study of nuclear susceptibility it is found that the sites of increased susceptibility might be different in their molecular composition from the sites with less susceptibility. This is studied particularly for the hyperacetylated histones and phosphoproteins among the hepatoma chromosomal proteins. C. HISTONEACETYLATION Histone acetylation occurs on all the nucleosomal histones. The histones H,, H,A, and H, possess a metabolically stable amino-terminal acetate group. The four non-H, histones are all modified at internal lysine residues, a modification which is metabolically very active (Jackson er af., 1975). This provides the nucleosome with the possibility of reestablishing critical interactions shortly after they are broken. However, it cannot be established yet that this high turnover does occur on the same nucleosome. A substantial subclass of each acetylatable histone remains totally unacetylated (Cousens er al., 1979). In any case the modification is rapid and is similarly removed with great speed. The acetylation of histones is a complicated process. Thus, histone H, can be modified to a level of four acetate groups. Histone H, can be modified to the level of triacetylation (Jackson er af., 1976b). In HTC cells the steady-state level of modification amounts to about 40% of the total histone and this may be different from other cell types (Jackson er af., 1975). The rate of histone acetylation and deacetylation has also been examined in nuclei from fetal, adult, and two kinds of neoplastic rat hepatocytes. These results suggest that histone acetylation and deacetyla-

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

89

tion occur more actively in more transformed cells (Horiuchi et al., 1981). However, these data must be interpreted with caution. The degree of modification by acetylation, as determined from HTC cells, is much greater than might have been expected if it occurs only in the region of active genes. But the more extensive modification by three to four acetate groups per histone accounts for a much smaller fraction of the total histone. In normal growing cells the rate of deacetylation is exceedingly rapid and as a result the steady-state levels of triacetylated and tetraacetylated H, and triacetylated H, histones are very low. Further study on the two distinct populations showed that one population is characterized by rapid hyperacetylation and rapid removal of this modification. The second population is deacetylated at a slower rate and is hypermodified much less vigorously. Moreover there appears to be no interconversion of histone between these histone populations. It is suggested that there may be a discrete set of nucleosomes of the HTC chromatin which contains all of the cells most metabolically active histone (Covault and Chalkley, 1980; Garcea and Alberts, 1980). The acetylation and deacetylation of histones has been studied during the cell cycle by Moore et al. (1979). It is found that in G, and S phase cells the acetate associated with the non-H, histone is removed at a very rapid rate. During mitosis histone H, is still acetylated vigorously while the other histones are modified. From mitosis to G, phase, cells increase their rates of histone acetylation again, especially that of H, and H,B. In order to establish the role of hyperacetylation for the cells’ progression through their life cycle, the effect of butyrate was extensively investigated. Butyrate exerts a dramatic decrease in the acetate hydrolysis while acetylation continues initially quite vigorously (Cousens et al., 1979; Garcea and Alberts, 1980; Moore et al., 1979). In isolated nuclei from HTC cells it was found that butyrate effectively inhibits the endogenous deacylase activity (Perry et al., 1979). Moreover the in vifro inhibition of deacetylating enzyme activity shows the same fatty acid specificity as the in vivo inhibition of deacetylation. In this regard n-propionate and n-pentanoate are also quite effective (Cousins et al., 1979). The effect of butyrate on in vivo acetylation is caused by a direct effect on the deacetylase activity (Cousins et al., 1979). It was also tested whether there is a correlation between acetylation and DNase I susceptibility or susceptibility for micrococcal nuclease. Nuclei from butyratetreated cells exhibited a dramatic increased rate of digestion with DNase I (D. A. Nelson et al., 1978a,b). Micrococcal nuclease showed no such preference for chromatin containing hyperacetylated histones, but an effect of butyrate on the release of DNA fragments was found by others (Kitzis et al., 1980~).Butyrate administration to the cells inhibits their growth (Rubenstein et al., 1979; Sealy and Chalkley, 1978; van Wijk et al., 1981). The site of growth inhibition was studied by time lapse cinematography and [,H]thymidine incorporation studies (van Wijk et al., 1981). Evidence is presented that sodium butyrate affected the

90

ROELAND VAN WIJK

cell cycle at a point shortly after mitosis. There was no effect on the completion of the ongoing cell cycle. It is known that histone H, already deposited on the chromosome can be rapidly but only partially acetylated. Even for the highest butyrate concentration the amount of nonacetylated H, remains substantial, suggesting that part of the nucleosomes within the cell contains H, molecules which are essentially inaccessible to the acetylase. Incoming H, histone is 100%acetylated and does not return to the parental unmodified form in the presence of butyrate (Nelson et al., 1978a). Together with the data on the comparative rates of histone acetylation it can be speculated that the low degree of acetylation of some of the histones during mitosis is essential for the reordering of the divided chromatin into the state characteristic for a G I cell with subsequent potential to proliferate.

D. HISTONEPHOSPHORYLATION H I histones are particularly subject to phosphorylation reactions involving different regions of the HI molecule. The phosphorylation reactions could add great variety to HI histone interactions with other chromosomal proteins and with DNA. These interactions would be significant in events such as DNA synthesis and chromosome condensation. However, differences in the tryptic phosphopeptide pattern of HI histones were not found between interphase and mitotic HTC (Balhorn et al., 1975) and H35 cells (Hohmann, 1976) although they were clearly observed for CHO cells (Hohmann, 1976; Hohmann et al., 1976). Itientical HIhistone subfractions of two rat hepatoma cell lines (HTC and Reuber H35) are phosphorylated at different rates, and not at different sites (Hohmann, 1979). Phosphopeptide mapping studies have already shown that in interphase H I histone phosphorylation plays two distinct roles, one of which is cyclic .AMP dependent and one of which depends on the initiation of DNA synthesis (Lmgen and Hohmann, 1975). Such phosphorylation reactions involve different sites within the H I histone. This might be due to differences in the kinase to phosphatase action or reflect some changes in the conformation of the HI histone subfractions in the two cell lines. An intriguing observation is the demonstration of different H histone subfractions which could be chromatographically or electrophoretically resolved. The different HI histone subtypes are not expressed in the same relative quantities among tissues of a given animal. However, the direct comparison of rat liver hepatoma cell (H35, HTC, RLC) H I histones demonstrated that a cell line was capable of expressing the same variety of H I histones as the tissue of origin (Hohmann, 1980). It is therefore also possible that phosphorylation reactions, which clearly involve different sites within HI histones, may serve to change the conformation of a specific subset of H I histones. The data might also have implications for HI histone phosphorylation and the notion that each HI subtype

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

91

plays a crucial role in cell replication, that is, DNA synthesis and mitosis. The different H, histone subtypes might be responsible for the variation in repeat length of the chromatin depending upon the species and tissue of origin. Values of chromatin repeat length between 160 and 240 bp may be found in the literature and differences as large as 50 bp have been reported to exist between different tissue or developmental stages of the same organism (Thomas and Thompson, 1977; Weintraub, 1979; Morris, 1976; Compton et al., 1976; Spadafora et al., 1976). The nonheritability of a characteristic repeat length has been demonstrated by the analysis of repeat lenghts of somatic hybrids between parental cell lines characterized by different repeat lengths. Somatic hybrid clones isolated from fusion of H56 cells (subclone of H,, a dedifferentiated variant subclone of clone H4IIEC3 of the Reuber H35 rat hepatoma) and rat glial RGBA (subclone of C6) have repeat lengths intermediate between the parental clones (Sperling er al., 1980). This result is anticipated if it is assumed that repeat length is established via a common pool of substances, like the different histone H, types. The availability for study of numerous independent descendent clones of the welldifferentiated clone H4IIEC3 of Reuber H35 hepatoma permitted the determination of the repeat lengths in relation to the degree of differentiation as exposed by different liver functions. Evidence was presented supporting the notion that repeat length closely parallels the state of differentiation (Sperling and Weiss, 1980). It would be attractive to study further a possible relationship between the variation in repeat lengths, specific changes in the expression of histone H, subtypes, the expression of differentiated functions, and the transition from the quiescent to proliferative state. E. ADP-RIBOSYLATION OF HISTONES Liver and hepatoma cells contain an enzyme that incorporates the ADP-ribose moiety of NAD into proteins as either a monomer of ADP-ribose or a homopolymer of several units. Careful comparative studies of modification of total liver and hepatoma cell proteins by ADP-ribose transfer have shown that the monomeric and polymeric residues change independently. Subfractionation of the mono-(ADP-ribose)-protein conjugates on the basis of their NH,OH sensitivity resulted in two fractions (Adamietz et d., 1974; Adamietz and Hilz, 1976; Bredehorst er al., 1978; Wielckens et al., 1981). The level of the NH,OHsensitive conjugates exhibited an inverse relationship to cell proliferation. Study on the subcellular distribution of the (ADP-ribose),-protein conjugates indicates the presence of multiple mono-(ADP-ribose)-protein in nearly all compartments (Adamietz et al., 1981). The polymeric ADP-ribose residues of hepatoma cells are 18-48 time lower than the monomeric residues. Hepatoma cell nuclei contain the enzyme that incorporates ADP-ribose into nuclear proteins but the existence of multiple extranuclear mono(ADP-ribose) +

92

ROELAND VAN WIJK

transferase activities has been demonstrated (Adamietz et al., 1981; Cho-Chung er al., 1980; Boquez et al., 1980; Hofstetter et al., 1981). Interestingly poly(ADP-ribose)-protein conjugates appeared to relate to the nuclear ADPribose-transferase activity. In the nuclear protein fraction, the chemical modification of histones through ADP-ribosylation might be of interest because it could affect the contiguity of the nucleosomes. Among the histones, poly(ADP-ribose) is predominantly associated with the H, histone fractions (Smith and Stcrcken, 1975). Such a negatively charged homopolymer like poly(ADP-ribose) to the highly basic regions of the molecule might weaken the interaction of H, histones with DNA. In this respect it functions, just as phosphorylation, as a mechanism for extending chromatin. In the Novikoff hepatoma, an alteration in the distribution of nuclear ADPribose was found compared with liver. A marked reduction in ADP-ribosylation of the histone fraction was found in this hepatoma (Burzio and Koide, 1972). More extensive studies, however, support that no correlation of poly(ADPribose) polymerase activity and proliferation rate was found, but there may exist a correlation with the degree of differentiation of the hepatoma, as was also suggested for other chromatin markers (Perrella and Lea, 1978). The previous experiments did not rule out the possibility that changes in distribution of the monomer and polymer of ADP-ribose on histone do occur during progression of cells through their cell cycle. It is of interest that spermine causes a large stimulation in the ADP-ribosylation of nuclear proteins and a decreased incorporation into histones of hepatoma and liver cell nuclei (Tanigawa et al., 1977; Perrella and Lea, 1978). Spemine causes a redistribution of ADP-ribose moieties from nonhistone to the internucleosomal H, histones (Tanigawa et al., 1977). It is known that the levels of polyamines increase during cell proliferation (see Section V). Thus, the conipetition with histones might weaken the forces involved in maintaining DNA in its condensed form. A similar effect was induced by divalent cations like Mg*+, but spermine produces the same effects at concentrations 2 orders of magnitude lower (Tabor, 1962). Moreover, spermine has been shown to displace Mg2+ from rat hepatic cell nuclear structure (N. F.Nelson er al., 1978; Brown et al., 1975). The structural perturbation in chromatin and the alteration of stereospecificity of nuclear poly(ADP-ribose) polymerase might no longer allow the enzyme to bind nucleosomal core protein. The question to what extent this modification preferentially occurs in the active part of chromatin is not yet answered.

F. METHYLATION OF HISTONES The role of protein methylation in Chromatin structure and proliferation of hepatoma cells is not clear. When the nuclei isolated from Novikoff hepatomas

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

93

were incubated in virro with S-adenosyl-~-[~H-methyl]methionine and the nuclear proteins were subsequently fractionated, they incorporated radioactivity at a higher rate than liver nuclei and this increase was mainly due to methylation of H, and H, histones (Kim et al., 1980). The enzyme which methylates the Eamino group of lysine residues of histone (protein methylase 11; S-adenosylmethionine; protein-lysine-N-methyltransferase: EC 2.1.1.43) was increased in hepatomas compared to liver (Paik er al., 1972, 1975). Its occurrence during the cell cycle has not yet been studied in hepatoma cells but was already studied in regenerating liver (Tidwell et al., 1968). The methylation process is active late during the cell cycle and might be necessary for chromatin to condense prior to mitosis (Lee et al., 1973). G. NONHISTONECHROMOSOMAL PROTEINS AND THEIRPHOSPHORYLATION

Strong suggestions do exist that several specific high mobility group (HMG) proteins can shuttle between nucleus and cytoplasma, in response to the need of the nucleus as helix-destabilizing proteins in DNA replication. The HMG proteins are a class of nonhistone chromatin proteins that can be released from chromatin with 0.35 M NaCl and that are soluble in 2% TCA (Goodwin et af., 1973). The term was first applied to proteins from calf thymus, which have distinctive amino acid compositions with high contents of both acidic and basic amino acid residues. The HMG proteins isolated from cultured HTC cells, and called HN-1 and NH-2, are similar to the HMG-1 and HMG-2 from calf thymus chromatin in physical, chemical, or immunochemical properties (Bidney and Reeck, 1978). From HTC chromatin these proteins were isolated after release by NaCl and sequential chromatography on double-stranded DNA and singlestranded DNA columns. It suggested that HMG proteins are preferentially associated with transcribed DNA sequences. Consistent with these proteins being involved in DNA replications is their enrichment in chromatin of rapidly dividing HTC cells. HMG-1 and probably also HMG-2 were also purified from the cytoplasm of HTC cells (Isackson et af., 1979). Immunofluorescence studies have revealed that antibodies to HMG-1 stain both nucleus and the cytoplasm (Isackson et af., 1980). It was also reported that the localizations vary according to the phase of the cell life cycle. There is preferentially fluorescence in the nucleus during the S phase, but in the metaphase the chromosomes are devoid of fluorescence while the cytoplasm displays an intense fluorescence. These results lend support to the view that these HMG proteins shuttle between nucleus and cytoplasm in response to the need of the nucleus as helix-destabilizing proteins. It must be noticed that for some other mammalian cell types selective nuclease attack releases HMG proteins. Slight digestion of mouse myeloma nuclei with micrococcal nuclease caused release of the fraction of soluble nucleoproteins which was highly enriched in HMG-1 and HMG-2 (Jackson et af., 1979).

94

ROELAND VAN WIJK

In the period of helix destabilization the increased amounts of phosphorylated chromatin proteins are remarkable. The demonstration of phosphoproteins in the nonhistone fraction has been clearly demonstrated for HTC (Kitzis et af., 1980a) and Reuber H35 cells (Defer et af., 1979). The existence of nonhistone proteins in nucleosomes was controversial for some time because (1) most of these proteins are loosely bound and could be partly released during the nucleosome preparation and (2) these proteins tend to form aggregates and could behave like nucleosomes in separation techniques. The presence of nonhistone proteins in nucleosomes was first established for liver and it contained phosphorylated nonhistone proteins and protein kinases (Defer et al., 1978). This was subsequently established for HTC cells (Kitzis et al., 1980a). This work has raised the question of specificity of nonhistone proteins toward nucleosomes or of a preferential binding of some nonhistone proteins to the active part of chromatin. In order to answer these questions, chromatin was digested with micrococcal nuclease. When proteins from chromatin were analyzed two major bands were found itnd it was also suggested that phosphoproteins and protein kinases are associated with the active parts of chromatin (Kitzis et al., 1980b). A further argument arose from cells cultured in the presence of butyrate. As was indicated before (Section VI,C) butyrate modifies chromatin structures in such a way that a small part of it becomes more sensitive to the nuclear attack than in control cell: chromatin was digested at a higher rate until 1.5% of the DNA was rendered acid soluble (Kitzis et al., 1980b). The released phosphoproteins preferentially originated from hutyrate-treated cells. Butyrate also induced a modification in the localization of protein kinase in chromatin (Kitzis et al., 1980~).The preferential phosphorylation of chromatin nonhistone proteins was studied during the progression of cells through their cell cycle using serum-deprived and synchronized Reuber H35 cells (Defer et al., 1979). The radioactive pattern of protein bound 32Pof Go cells showed no radioactive peak. The radioactivity progressively increased until the cells reached S phase. When most of the cells were in the S phase the radioactivity strongly decreased. Chromatin protein kinase activities were found to increase in late G, and continued to increase in S phase (Defer et al., 1979). Apparently, in the period of helix destabilization the increased amount of chromatin phosphorylated proteins is substantial as is their more preferential localization in chromatin at nuclease-sensitive sites. H. REPLICATION AND NUCLEOSOME FORMATION A major point of regulation of DNA synthesis is the regulation of protein, in particular histone synthesis. This is of interest because of the temporal relationship between DNA replication and subsequent packaging into nucleosomes. The packaging is very rapid and apparently not a limiting step in the process of replication. Attempts have been made to retard the process of packaging and to

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

95

extend the life time of replicative DNA by treatment with protein synthesis inhibitors such as cycloheximide. DNA synthesis can continue in the presence of cycloheximide, albeit at a reduced rate. Upon such treatment an increased nuclease sensitivity is maintained for several hours after the DNA is synthesized (Weintraub, 1972; Seale and Simpson, 1975). Because it was observed that the newly replicated DNA has a 2-fold increased sensitivity to digestion by staphylococcal nuclease this prolonged nuclease sensitivity has been interpreted as arising from the absence of newly synthesized histones on the new DNA. However, the structure of the replication fork is not clear. Electron microscopic evidence suggested that one of the daughter strands is devoid of nucleosomes (Riley and Weintraub, 1979) but this could result from artificial but preferential dissociation (Jackson and Chalkley, 1981). Density gradients show that the replicated new DNA has a high protein content (Seale and Simpson, 1975). Such a differential dissociation on one of the daughter strands is not in agreement with a random distribution. It would agree with some selective association of newly made histones or other chromosomal proteins with newly replicated DNA. A random association was observed with experiments with isopycnic centrifugation of formaldehyde-fixed, density-labeled hepatoma chromatin (Jackson er af., 1976a). The problem was reinvestigated with MH-134SC hepatoma cells (Senshu er af., 1978; Senshu and Ohashi, 1979). A new method for isolating subcellular components after fixation of HTC cells with formaldehyde was then developed (Jackson and Chalkley, 198l), which could almost completely resolve newly replicated chromatin from preexisting material. Exploiting this method the result of Senshui er af. (1978) was affirmed, that newly synthesized histones H, and H, are deposited onto new DNA and stay in place for a significant time. In contrast new histone H, is deposited on old DNA and new histone H,A and H2B, while transiently bound to new DNA, are largerly associated with preexisting chromatin. In the presence of cycloheximide only histones H,, H2A, and H, are synthesized whereas the synthesis of the remaining two histones, H, and H2B, appears to be totally inhibited. Since the newly replicated DNA is associated with an essentially normal complement of histone the lower density than mature chromatin is due to the possibilities that either (1) the newly replicated chromosome has little extra protein and the nucleosome core histones are more closely spaced, or (2) there is normal spacing but extra protein is present. In view of current ideas of an association of the nuclear matrix with replicating DNA one might envisage the possibility of extra protein as well as of temporarily modification of the non-H, histone. Specific coupling of new histones to replicated DNA was also studied in cultures after interruption of DNA synthesis with hydroxyurea. In HTC cells no significant decrease was noted on total protein synthesis as a result of exposure to hydroxyurea, but histones behave in a different manner (Jackson and Chalkley, 1981; Senshu and Ohashi, 1979). Histone synthesis drops rapidly: histones H,A

96

ROELAND VAN WIJK

and H2B appear to be the most extensively inhibited and H, appears to be the least affected. However, the coupling between DNA and histone synthesis is not very tight and histone synthesis continues quite vigorously in HTC cells when DNA synthesis is inhibited (Nadeau et al., 1978). It might depend on the cell type or its average cell cycle duration. A study of the subcellular distribution of newly synthesized histones in MH-134SC hepatoma cells shows a more tight coupling but also in this cell type the abrupt shutdown of DNA synthesis would create a state of transient histone excess until the translation of histone message comes to a halt (Senshu and Ohashi, 1979). The interruption of DNA synthesis specifically blocks the association of H, and H, histones with chromatin and causes them to accumulate in the cytoplasm. It agrees with the report that new H, and H, were deposited exclusively on nascent DNA. Association of other histones may continue in so far as their messenger RNAs are available which would result in the deposition of new H2A and H,B on nonreplicating DNA.

VII. System of Second Messengers or Intracellular Regulators The concept of cyclic AMP as a second messenger for initiation of DNA synthesis of liver type cells was supported by various authors. It has been documented for liver stimulated DNA synthesis that a pre-S phase rise in cAMP occurs after hormone infusion (MacManus et af., 1973) or partial hepatectomy (Diamantstein and Ulmer, 1975; Ebina et af., 1975; MacManus et af., 1973; Sheppard and Prescott, 1972; Short et al., 1975). Maximum cyclic AMP concentrations occurring during G, were also observed for Novikoff hepatoma cells (Zeilig and Goldberg, 1977) and Reuber H35 cells (Van Wijk, 1983). Prevention of the pre-S accumulation of cyclic AMP in regenerating liver with adrenergic antagonists does prevent initiation of DNA synthesis (MacManus er af., 1973). Treatment of rat neonatal hepatocytes with exogenous cyclic AMP stimulates their entry into DNA synthesis (Armato et al., 1975). However, this was not observed for hepatoma cells (Brgnstad er af., 1978; Van Meeteren er al., 1983). Hepatoma cells may have lost the cAMP sensitive and positive acting control of initiation of DNA synthesis. In the regenerating liver cell it is currently speculated that the influx of calcium is an adequate stimulus (Rixon and Whitfield, 1976). After calcium depletion, early events such as increases in cyclic AMP, prostaglandin E, and polyamine occur in response to partial hepatectomy (Whitfield er af., 1976). Apparently calcium is required for the implementation of the action of (one of) these regulating substances. The loss of requirement for calcium is a consistent finding in neoplasm (Whitfield et al., 1976; MacManus er al., 1978). Although the trigger mechanism by which calcium and the cyclic nucleotide activate DNA synthesis is essentially unknown, it appeared that the hepatoma cells are characterized by a change in the sensitivity toward both calcium and cyclic AMP.

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

97

It has further been suggested that calmodulin/calcium induces DNA synthesis by prostaglandins (Boynton and Whitfield, 1981). The DNA synthesis response of calcium-deprived rat liver cells to calcium or calmodulin was blocked by indomethacin, an inhibitor of endoperoxide synthetase. This enzyme converts arachidonic acid into PGH,, the common precursor of prostaglandins and thromboxanes. Stimulation of DNA synthesis in primary cultures of neonatal rat liver by arachidonic acid and prostaglandins was demonstrated by Andreis et al. (1981). The effect of indomethacin has been studied in HTC cells. Indomethacin reversibly inhibited growth of HTC cells in the G I phase of the cell cycle (Bayer et al., 1979; Bayer and Beaven, 1979). Indomethacin suppressed specifically the increase in uptake of amino acids through the Na -dependent “A”-system and enhances their transport through the “L”-system. It is argued that indomethacin blocked the production of the carrier itself (Bayer et al., 1980). It has been mentioned before (Section IV) that control of initiation of DNA synthesis by hormones and amino acids was found in primary cultures of adult rat hepatocytes (Tomita et af., 1981) and Reuber H35 cells (Streumer-Svobodovaet al., 1982). The role of microtubules in insulin and glucagon stimulation of amino acid transport in isdated rat hepatocytes has been demonstrated by Prentki er al. (1981). The role of increased amino acid transport in initiation of DNA synthesis has been studied extensively in relation to the inhibition by colchicine of the initiation of DNA synthesis. Colchicine inhibited the initiation of DNA synthesis of hepatocytes in vivo. Hepatocytes showed a high sensitivity for colchicine shortly before the onset of DNA synthesis. The inhibitory action was not mediated by polyamine synthesis (Walker et al., 1977). On the other hand colchicine inhibits the induced amino acid uptake mechanism. Normally a biphasic increase of Na-dependent amino acid uptake occurred in the liver remnant of rats subjected to partial hepatectomy (Walker and Whitfield, 1977, 1978). The importance of increased amino acid transport for increased protein synthesis and initiation of DNA synthesis is without doubt. However, activation of other factors involved in increased initiation frequency in protein synthesis must be considered to play a role. Moreover, studies on protein phosphorylation through the cell cycle have revealed that both histone and nonhistone chromosomal proteins play a role in initiation of DNA synthesis. These various effects might also be under the regulatory control of calcium mediated by the ubiquitous calcium receptor calmodulin. This protein has been demonstrated to mediate either directly or indirectly the Ca2 regulation of cyclic nucleotide metabolism, protein phosphorylation, and microtubule assembly (Wang and Waissman, 1979; Cheung, 1980; Means and Dedman, 1980). In this respect it is interesting that (1) specific elevation of the intracellular calmodulin concentration occurs in G I , (2) a highly significant correlation between the intracellular calmodulin levels and progression of cells into S phase has been demonstrated, (3) a high calcium sensitivity was demonstrated in late G I , and (4) phosphorylation of histone HI and nonhistone chromosomal protein increases in late G, and early S (Chafouleas et al., +

+

98

ROELAND VAN WIJK

1982; previous section). Calmodulin directly stimulates a growing list of protein kinases but in addition, Wolff et al. (1981) suggested that calmodulin may also affect phosphatase activity. Thus, it appears that calmodulin may influence the degree of protein phosphorylation through regulation of both phosphatases and kinases.

VIII. Concluding Remarks In this article a survey is given of the studies regarding the regulation of DNA synthesis in hepatoma cells. In several cases a comparison is made with liver or cultured liver cells. A variety of external stimuli have been studied and it can be expected that they act via different systems for external signal reception. The cell has a choice out of a variety of mechanisms to transduce external signals and produce intracellular messengers. The plasma membrane plays an important role. For hepatoma cells the plasma membrane needs extensive study to clarify how it excerts its complicated role in growth regulation with the liver growth factors. In the second, or intracellular, messenger system, studies have been focused on the cyclic AMP system, the calcium-calmodulin system, the prostaglandins, and the microtubule system. It can be supposed that studies in this field of second messenger systems might be extremely complicated because of three main reasons. 1. The various second messenger systems might influence the crucial process of DNA restructuring and replication at several and different levels: the temporal formation of new protein and modification of new and existing nuclear proteins. 2. The various second messenger systems might influence each other as has been described for the role of calcium-cyclic AMP system or the cyclic AMP-prostaglandin systems etc. In my opinion the recent studies suggest that there is not one single sequence of events (i.e., calcium-prostaglandin-CAMP etc.) but parallel sequences which ultimately lead to the same aim namely DNA replication but show intercontrolling elements. 3. We have to be aware of the possibility that the second messenger systems might also influence the incoming signal and function as a feedback regulator.

It seems apparent that the transformed liver cell is not essentially different from the liver except the coupling between the various intracellular messenger systems or the coupling between the intracellular messenger system and the external signal reception. In hepatoma cells a hyperactive intracellular system is present which can hardly be influenced or activated from outside. Thus, deple-

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

99

tion for calcium or polyamines and addition of CAMPare beyond control. For an understanding of the intracellular, second messenger systems a comparative study of different hepatoma cell types might be advantageous. It is clear that the level of specific proteins must be altered for initiation of DNA synthesis to occur. A few of these specific proteins have been characterized. The regulation of their amount includes controls at the level of genome transcription, messenger RNA translation, and enzyme stability. Understanding of the differential gene expression during progression of cells through their cell cycle requires active cellfree protein-synthesizing systems derived from hepatoma cells as well as the further characterization of messenger RNAs. Finally, these proteins or their products may be supposed to cause structural changes at nuclear chromatin level specific for initiation of DNA synthesis. The occurrence of parallel pathways leading to different critical events, while any of these events can be basically rate limiting for initiation of DNA synthesis, can easily give rise to individual cells in a culture that are different in their ultimate limiting event. By way of immunolabeling techniques and other single cell methods it might be possible in the future to discriminate the cells for the presence of their various essential factors for initiation of DNA synthesis.

ACKNOWLEDGMENTS I thank my colleagues at the Department of Molecular Cell Biology (Utrecht) and the Departments of Zoology (Utrecht and Nijmegen): Drs. A. van Meeteren, D. H. J. Schamhart, H. 0. Voorma, W. L. M. Linnemans, W. Berendsen, W. Geilenkirchen, and R. van Haarlem for many fruitful discussions. This work was supported in part by the Koningin Wilhelmina Fonds (Netherlands Cancer Foundation), and the Foundation for Fundamental Biological Research (BION), which is subsidized by the Netherlands Organization for the Advancement of Pure Research (ZWO).

REFERENCES Absher, P. M., and Absher, R. G . (1976). Exp. Cell Res. 103, 247-255. Adamietz, P., and Hilz, H. (1976). Hoppe-Seylers Z. Physiol. Chem. 357, 527-534. Adamietz, P., Bredehorst, R., Oldekop, M., and Hilz, H. (1974). FEBS Lerr. 43, 318-322. Adamietz, P.,Wielckens, K., Bredehorst, R., Lengyel, H., and Hilz, H. (1981). Biochem. Biophys. Res. Commun. 101, 96-103. Alberghina, L. (1977). J . Theor. Biol. 69, 633-643. Alberghina, L., Mariani, L., and Martegani, E. (1981).In “Biomathematics and Cell Kinetics” (M. Rotenberg, ed.), pp. 295-309. Elsevier, Amsterdam. and Benne, R. (1979). Eur. J . Amesz, H., Goumans, H., Haubrich-Morree, T., Voorma, H. 0.. Biochem. 98, 513-520. Andreis, P. G., and Armato, V. (1981). Endocrinology 108, 1954-1964.

100

ROELAND VAN WIJK

Andreis, P. G., Whitfield, F. I., and Armato, U. (1981).Exp. Cell. Res. 134, 265-273. Armato, U., Andreis, P. G., and Draghi, E. (1975).Chem. Biol. Inreract. 11, 67-90. Aujard, C., Chany, E., and Frayssinet, C. (1973).Exp. Cell Res. 78, 476-478. Balhom, R., Jackson, V., Granner, D., and Chalkley, R. (1975).Biochemistry 14, 2504-2511. Bamett, C. A., and Wicks, W. D. (1971).J. Biol. Chem. 246, 7201-7206. Barnett, C. A., Bamhorst, M., Fooshee, C. M., and Saneto, R. P. (1979).I n Virro 15, 128-137. B m , R., Parsons, J., Koch, M. R., and Lea, M. A. (1979).Cancer Res. 39, 1655-1660. Barra, R., Parsons, J., and Lea. M. R. (1980). J. Cell Biol. 87, 2A. Baserga, R. (1965).Cancer Res. 25, 581-595. Bayer, B. M., and Beaven, M. A. (1979).Biochem. Phurmucol. 28, 441-443. Bayer, B. M., Kruth, H. S., Vaughan, M., and Beaven, M. A. (1979).J . Pharmacol. Exp. Ther. 210, 106-111. Bayer, B. M., Lo, T. M., and Beaven, M. A. (1980).J. Biol. Chem. 255, 8784-8790. Beck, W. T.,Bellantone, R. A., and Canellakis, E. S. (1972).Biochem. Biophys. Res. Commun. 48, 1649-1655. Berezney, R. (1977).Exp. Cell Res. 123, 41 1-414. Berezney, R. (1979). Cell Nucleus 7, 413. Berezney, R.,and Buchholtz, L. A. (1981).Exp. Cell Res. 132, 1-13. Berezney, R., and Coffey, D. S. (1974).Biochem. Biophys. Res. Commun. 60, 1410-1417. Berezney, R., and Coffey, D. S. (1975).Science 189, 291-293. Berezney, R., and Coffey, D. (1976).Adv. Enzyme Regul. 14, 63-110. Berezney, R., and Coffey, D. S. (1977).J. Cell Biol. 73, 616-637. Bevers, M. M., van Rijn, J., and Van Wijk, R. (1976).FEES Lert. 72, 275-278. Bidney, D. L., and Reeck, G. R. (1978).Biochem. Biophys. Res. Commun. 85, 1211-1218. Blankenship, J., and Walle, T. (1972).A&. Pofyumine Res. 2, 97- 110. Bloomfield, V. A,, and Wilson, R. W. (1981).I n “Polyamines in Biology and Medicine” (D. R. Moms and L. J. Marton, eds.). pp. 183-206. Dekker, New York. Boquez, P., Duflot, E.,and Godeau, F. (1980).Eur. J . Cell Biol. 22, 1 12. Boynton, A. L., and Whitfield, J. F. (1981).Adv. Cyclic Nucleotide Res. 14, 411-419. Boynton, A. L., Whitfield, J. F., MacManus, J. P., Armato, U., Tsang, B. K., and Jones, A. (1981).Exp. Cell Res. 135, 199-212. Bredehorst, R., Wielckens, K., Gastemann. A., Lengyel, H., Klapproth, K., and Hilz, H. (1978). Eur. J . Biochem. 92, 129-135. Brgnstad, G., Hafher, F., Baalsrud, A., Larsson, P., and Christoffersen, T.(1978).Adv. Cyclic Nucleoride Res. 9, 780. Brooks, R. F. (1975).J. Cell. Physiol. 86, 369-378. Brooks, R. F. (1976).Nature (London) 260, 248-250. Brooks, R. F., Bennett, D. C., and Smith, J. A. (1980). Cell 19,493-504. Brown, K. B., Nelson, N. F., and Brown, D. G. (1975).Biochem. J . 151, 505-512. Bucher, N. L.R., and Malt, R. A. (1971).“Regenerating of Liver and Kidney,” pp. 17-176. Little Brown, Boston, Massachusetts. Bucher, N. L. R., and Swaffield, M. N. (1975).Proc. Natl. Acad. Sci. U.S.A. 72, 1157-1160. Bullough, W. S. (1963).Nature (London) 199, 859-862. Bullough, W. S.,and Laurence, E. D. (1960).Proc. R. SOC. London [Biol] 151, 517-536. Bums, F. J., and Tannock, J. F. (1970).Cell Tissue Kine?. 3, 321-334. Bunio, L., and Koide, S. S. (1972).FEES Lert. 20, 29-32. Byus, C. V., Hayes, J. S., Brenda], K., and Russell, D. H. (1976a).Life Sci. 19, 329-336. Byus, C. V., Wicks, W. D., and Russell, D. H. (1976b).J. Cyclic Nucleoride Res. 2, 241250. Canellakis, Z.N. (1981).Biochem. Biophys. Res. Commun. 100,929-933.

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

101

Canellakis, Z. N., and Theoharides, T. C. (1976). J. Biol. Chem. 251, 4436-4441. Canellakis, Z. N., Theoharides, Th. C . , Bondy, Ph. K., and Triarhos, E. T. (1981). Life Sci. 29, 707-710.

Chafouleas, 1. G., Bolton, W. E., Hidaka, H., Boyd, A. E., and Means, A. R. (1982). Cell 28, 41-50.

Chen, K. Y.,Heller, J. S., and Canellakis, E. S . (1976a). Biochem. Biophys. Res. Commun. 68, 40 1 -408.

Chen, K. Y.,Heller, J. S., and Canellakis, E. S. (1976b). Biochem. Biophys. Res. Commun. 70, 2 12-2 19.

Chen, K. Y., Kyriakidis, D. A., and Canellakis, E. S. (1982). Biochim. Biophys. Acru 716, 72-78. Cheung, W. Y . (1980). Science 207, 19-27. Cho-Chung, Y. S., Clair, T., Bodwin, J. S . , and Hill, D. M. (1980). Biochem. Biophys. Res. Commun. 95, 1306-1313. Clark, J. L., and Fuller, J. L. (1976). Biochem. Biophys. Res. Commun. 73, 785-790. Cohen, S. S. (1971). ‘‘Introduction to the Polyamines.” Prentice Hall, New York. Cohen, S. S. (1978). Nature (London) 274, 209-210. Collijn-D’Hooghe, M., Valleron, A.-J., and Malaise, E. P. (1977). Exp. Cell Res. 106, 405-407. Compton, J. L., Bellard, M., and Chambon, P. (1976). Proc. Natl. Acud. Sci. U.S.A. 73, 4382-4386.

Cousens, L. L., Gallwitz, D., and Alberts, B. M. (1979). J . Biol. Chem. 254, 1716-1723. Covault, J., and Chalkley, R. (1980). J. Biol. Chem. 255, 91 10-9116. Dawson, K. B., Madoc Jones, H., and Field, E. 0. (1965). Exp. Cell Res. 38, 75-84. Defer, N., Kitzis, A., Levy, F.. Tichonicky, L., Sabatier, M.-M., and Kruh, J. (1978). Eur. J . Biochem. 88, 583-591. Defer, N., Tichonicky, L., Kruh, J., Zoutewelle, G., and Van Wijk, R. (1979). Biochimie 61, 855-860.

Deschamps, Y.,and Verly, W. G. (1975). Biomedicine 22, 195-208. Desser-Wiest, L. (1975). &err. 2. Onkol. 2, 56-59. Desser-Wiest, L. (1976). Horm. Metab. Res. 9, 97-98, Diamantstein, T., and Ulmerk A. (1975). Immunology 28, 121-125. Dulbecco, R. (1970). Nature (London) 227, 802-806. Ebina, Y., Iwai, H., Kukui, N., Ohtsuka, M., and Miura, Y. (1975). J. Biochem. 77, 641-645. Emanuel, J. R.,and Gelehner, Th.D. (1975). Biochem. Biophys. Res. Commun. 63, 825-831. Faliks, D., and Meyuhas, 0. (1982). Nucleic Acids Res. 10, 789-801. Fong, W., Heller, J., and Canellakis, E. S. (1976). Biochim. Biophys. Acru 428, 456-465. Friedman, D. L., Claus, T. H., Pilkis, S. J., and Pine, J. E. (1981). Exp. Cell Res. 135, 283-291. Froese, G. (1964). Exp. Cell Res. 35, 415-419. Garcea, R. L., and Alberts, B. M. (1980). J . Biol. Chem. 255, 11454-1 1463. Geyer. P. K., Meyuhas, O., Perry, R. P., and Johnson, L. F. (1982). Mol. Cell. Biol. 2,685-693. Goodwin. G. H., Sanders, C., and Johns, E. W. (1973). Eur. J. Biochem. 38, 14-19. Goumans, J. J. I. H. (1981). Doctoral Thesis, University of Utrecht. Granner, D. K. (1972). Biochem. Biophys. Res. Commun. 46, 1516-1522. Granner, D. K. (1974). Arch. Biochem. Biophys. 165, 359-368. Guiguet, M., Collijn-D’Hooghe, M., and Valleron, A.-J. (1980). Cell Tissue Kiner. 13, 689-694. Hargrove, J. L., and Granner, D. K. (1982). J. Cell. Physiol. 111, 232-239. Heller, J., Fong, W., and Canellakis, E. S. (1976). Proc. Narl. Acad. Sci. U.S.A. 73, 1858-1862. Heller, J. S., Kyriakidis, D., Fong, W. F.,and Canellakis, E. S. (1977a). Eur. J . Biochem. 31, 545-550.

Heller, J. S., Chen, K. Y.,Fong, W. F.,Kyriakidis, D., and Canellakis. E. S. (1977b). Fed. Proc. Fed. Am. SOC. Exp. Biol. 36, 799

102

ROELAND VAN WIJK

Heller, J. S . , Chen, K. Y., Kyriakidis, D. A., Fong, W. F., and Canellakis, E. S. (1978). J. Cell. Physiol. 96, 225-234. HCmon, D., Collijn-D’Hooghe, M., Valleron, A.-J., and Malaise, E. P. (1978). In “Biomathematics and Cell Kinetics” (A. J. Valleron and P. D. M. MacDonald, eds.), pp. 43-55. Elsevier, Amsterdam. Higuret, P., Chany, E.,Mousset, S., Gardes, M., and Frayssinet, C., (1975). C.R. Acud. Sci. (Paris) 281, 49-52. Hofstetter, W., Miihlebach, T., Liitschen, H., Winterhalter, K. H., and Richter, C. (1981). Eur. J. Biochem. 117, 361-367. Hohmann, P. (1976). J. Cell Biol. 70, 403A. Hohmann, P. (1979). J . Biol. Chem. 254, 9022-9029. Hohmann, P. (1980). Arch. Biochem. Biophys. 205, 198-209. Hohmann, P.,Tobey, R. A., and Gurley, 1. R. (1976). J. Biol. Chem. 251, 3685-3692. Holtta, E., and Raina, A. (1973). Acra Endocrinol. 73, 794-800. Holley, R. W., Armour, R., Baldwin, J. H., Brown, K. D., and Yeh, Y.-C. (1977). Proc. Nufl. Acud. Sci. U.S.A. 74, 5046-5050. Horiuchi, K., Fujimoto, D., Fukushima, M., and Kanai, K. (1981). Cuncer Res. 41, 1488-1491. Huberman, J. A., and Riggs, A. D. (1968). J . Mol. Biol. 32, 327-341. Isackson, P. J., Bidney, D. L., Fishback, J. L., and Reeck, J. R. (1979). Fed. Proc. Fed. Am. SOC. Exp. Biol. 38, 494. Isackson, P. J., Bidney, D. L., Reeck, J. R., Neihart, N. K., and Bustin, M. (1980). Biochemistry 19, 4466-4471. Jackson, V., and Chalkley. R. (1981). Cell 23, 121-134. Jackson, V., Shires, A., Chalkley, R., and Granner, D. (1975). J. Biol. Chem. 250, 4856-4863. Jackson, V., Granner, D., and Chalkley, R. (1976a). Proc. Nurl. Acud. Sci. U.S.A. 73,2266-2269. Jackson, V., Shires, A., Tamphaichitr, N., and Chalkley, R. (1976b). J . Mol. Biol. 104,471-483. Jackson, J. B., Pollock, J . M., and Rill, R. L. (1979). Biochemistry 18, 3739-3740. Janne, J., and Holltta, E. (1974). Biochem. Biophys. Res. Commun. 61, 449-456. Jimenez de Asua, L., O’Farrell, M. R., Bennett, D., Clingan, D., and Rudland, P.S . ( I 977). Nature (London) 265, 151-153. Kallio, A., Poso, H., and Janne, J. (1977). Biochim. Biophys. Acra 479, 345-353. Kaneko, Y., Ymai, Y., Matsuzaki, F., Endo, Y.,and Oda, T. (1979). Experienriu 35, 1660-1661. Kato, Y., Nishihara, H., Inoue, H., Gohda, E., and Takeda, Y. (1978). J. Biochem. 84, 1617- 1623. Kim, Y. S., Lou, M. F., Paik, M. K., Kim, S., and Paik, W. K. (1980). Curcinogenesis 1, 129- 134. Kitzis, A., Tichonicky, L., Defer, N., and Kmh, J. (1980a). Eur. J . Biochem. 111, 237--244. Kitzis, A., Tichonicky, L., Defer, N., and Kruh, J. (1980b). Biochem. Biophys. Res. Commun. 93, 833-841. Kitzis, A., Tichonicky, L., Defer, N., and Kmh, J. (1980~).Biochem. Biophys. Res. Commun. 97, 530-537. Koontz, J. W., and Iwahashi, M. (1981). Science 211, 947-949. Kuo, C. Y., and Yoo, T. J. (1977). J . Nurl. Cancer Inst. 59, 1691- 1695. Labrecque, D. R. (1980a). In Virro 16, 255. Labrecque, D. R. (1980b). Clin. Rex. 28, 279A. Langan, T. A., and Hohmann, P. (1975). In “Chromosomal Proteins and their Role in Regulation of Gene Expression” ( G . S. Stein and L. J. Kleinsmith, eds.), pp. 113-121. Academic Press, New York. Lea, M. A., Monis, H. P., and Weber, G. (1966). Cuncer Res. 26, 465-469. Lebkowski, J. S . , and Laemmli, U. K. (1982a). J . Mol. Biol. 156, 309-324. Lebkowski, J. S . , and Laernmli, U. K. (1982b). J. Mol. Biol. 156, 325-344.

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

103

Lee, H. W., Paik, W. K., and Borun, T. W. (1973). J. Biol. Chem. 248, 4194-4199. Leffert, H. L. (1974). J. Cell Biol. 62, 792-801. Leffert, H. L., and Koch, K. S . (1982). Cold Spring Harbor Conf. Cell Prolif. 9 . 597-613. Levine, E. M., Becker, Y., Boone, C. W., and Eagle, H. (1965). Proc. Narl. Acad. Sci. U.S.A. 53, 350-356. Lewin, B. M. (1973). “Gene Expression,” Vol. 2, pp. 60-64. Wiley, New York. Lewis, C. D., and Laemmli, U. K. (1982). CeN29, 171-181. Libby, P. R. (1978). J . Biol. Chem. 253, 233-237. Lindgren, A., Westermark. B., and Pontkn, J. (1975). Exp. Cell Res. 95, 311-319. Liu, A. Y.-C. (1980). J. Biol. Chem. 255, 4421-4429. Loeb, J. N.. Borek, C., and Yeung, L. L. (1973). Proc. Narl. Acad. Sci. U.S.A. 70, 3852-3856. Lottfield, R. B., Eigner, E. A., and Pastuszyn, A. (1981). I n “Polyamines in Biology and Medicine” (D. R. Moms and L. J. Marton, eds.), pp. 208-221. Dekker, New York. Lynch, W. E., Umeda, T., Uyeda, M., and Lieberman, I. (1972). Biochim. Biophys. Acra 287, 28-37. MacManus, J. P., Franks, D. J., Youdale, T., and Braceland, B. M. (1972). Biochem. Biophys. Res. Commun. 49, 1201-1207. MacManus, J. P., Braceland, B. M., Youdale, T., and Whitfield, J. F. (1973). J . Cell. Physiol. 82, 157-164. MacManus, J. P., Boynton, A. L., and Whitfield, J. F. (1978). Adv. Cyclic Nucleoride Res. 9 , 485-491. McCann, P. P., Tardif, C., Duchesne, M.-C., and Mamont, P. S . (1977a). Biochem. Biophys. Res. Commun. 76, 893-899. McCann, P. P., Tardit, C . , and Mamont, P. S . (1977b). Biochem. Biophys. Res. Commun. 75, 948-954. McCann, P. P., Homsperger, J.-M., and Seiler, N. (1979a). Neurochem. Res. 4 , 437-447. McCann, P. P., Tardif, C., Homsperger, J.-M., and Bohlen, P. (1979b). J. Cell. Physiol. 99, 183-190. Mallete, L. E., and Exton, J. H. (1973). Endocrinology 93, 640-644. Mamont, P. S., Bohlen, P., McCann, P. P., Bey, P., Schuler, F., and Tardif, C. (1976). Proc. Narl. Acad. Sci. U.S.A. 7 3 , 1626-1630. Mamont, P. S., Duchesne, M.-C., Grove, J., and Bey, P. (1978a). Biochem. Biophys. Res. Commun. 81, 58-66. Mamont, P. S., Duchesne, M.-C., Grove, J., andTardif, C. (1978b). Exp. CellRes. 115,387-393. Mamont, P. S . , Duchesne, M.-C., Joder-Ohlenbusch, A.-M., and Grove, J. (1978~).I n “EnzymeActivated Irreversible Inhibitors” (N. Seiler, M. J. Jung. and J. Koch-Weser, eds.), pp. 34-54. Elsevier, Amsterdam. Mamont, P. S . , Bey, P., and Koch-Weser, J. (1980). I n “Polyamines in Bio-Medical Research” (J. M. Gaugas, ed.), pp. 147-165. Wiley, New York. Manen, C. A., Hayes, J . S., Brendal, K., and Russell, D. H. (1977). Fed. Proc. Fed. Am. SOC.Exp. Biol. 36,408. Martin, D. W., Jr., Tomkins, G. M., and Bresler, M. A. (1969a). Proc. Narl. Acad. Sci. U.S.A. 63, 842-849. Martin, D., Jr., Tomkins, J. M., and Granner, D. (1969b). Proc. Natl. Acad. Sci. U.S.A. 62, 248-255. Maugh, T. H. (1972). Science 176, 1407-1408. Means, A. R., and Dedman, J. R. (1980). Narure (London) 285, 73-77. Medoff, J., and Parker, N. (1971). J. Cell Biol. 55, 173a. Mendelsohn, M. L. (1965). “Cellular Radiation Biology.” Williams & Wilkins, Baltimore, Maryland. Mierzejewski, K., and Rozengurt, E. (1976). Biochem. Biophys. Res. Commun. 73, 271-278.

104

ROELAND VAN WlJK

Mitchison, J. M. (1971).“The Biology of the Cell Cycle.” Cambridge Univ. Press, London and New York. Miyamoto, H.. Zeuthen, E., and Rasmussen, L. (1973).J. Cell Sci. 13, 879-888. Molimard, R., Pietru, G., Chany, E., Trincal, G., and Frayssinet, C. (1975). Biomedicine 23, 434-444. Moore, M., Jackson, V., Sealy, L., and Chalkley, R. (1979). Biochim. Biophys. Acra 561, 248-260. Moriarity, D. M., Disorbo, D. M., Litwack, G., Savage, C. R., and Marks, D. B. (1980).Fed. Proc. Fed. Am. SOC. Exp. Biol. 39, 1929. Moriarity, D. M., Disorbo, D. M., Litwack, G., and Savage, C. R., Jr. (1981).Proc. Narl. Acad. Sci. U.S.A. 78, 2752-2756. Moms, H. P. (1965).Adv. Cancer Res. 9, 227-302. Moms, N. R. (1976).Cell 9, 627-632. Mueller, J. (1971).I n “The Cell Cycle and Cancer” (R. Baserga, ed.), pp. 269-307. Decker., New York. Murphy, J. S., D’Alisa, R., Gershey, E. L., and Landsberger, F. R. (1978).Proc. Narl. Acad. Sci. U.S.A. 75, 4404-4407. Nabeshima, Y., and Ogata. K. (1980).Eur. J. Biochem. 107, 323-329. Nadeau, P., Oliver, D. R., and Chalkley, R. (1978).Biochemistry 17,4885-4893. Nakamura, T., Tomita, Y., and Ichihara, A. (1980).Cell Srrucr. Funcr. 5 , 412. Nelson, D. A.. Perry, M., Sealy, L., and Chalkley, R. (1978a)Biochem. Biophys. Res. Commun. 82, 1346-1353. Nelson, D. A,, Perry, W. M., and Chalkley, R. (1978b).Biochem. Biophys. Res. Commun. 82, 356-363. Nelson, N. F.,Brown, K. B., Fehlman, B. R., Stewart, G. P.. and Brown, D. G. (1978).Biochim. Biophys. Acra 517, 429-438. Nilson, J. (1976).Biochim. Biophys. Acra 418, 376-396. Paermentier, de, F., Barbason, H., and Bassleer, R. (1979).Biol. Cell. 34, 205-212. Paik, W. K., Lee, H. W., and Moms, H. P. (1972).Cancer Res. 32, 37-40. Paik, W. K., Kim, S., Ezirike, J., and Moms, H. P. (1975).Cancer Res. 35, 1159-1 163. Panko, W. B., and Kenney, F. T. (1971).Biochem. Biophys. Res. Commun. 43, 346-350. Pardee, A. B. (1974).Proc. Narl. Acad. Sci. U.S.A. 71, 1286-1290. Pariza, M. W., Becker, J. E.. Yager, J. D., Bonney, R. J., and Potter, V. R. (1974).In “Differentiation and Control of Malignancy of Tumor Cells” (W. Nakahara, T. Ono, and H. Sugimura, eds.), pp. 267-285. Univ. of Tokyo Press. Paulson, J. R., and Laemmli, U. K. (1977).Cell 12, 817-828. Perrella. F. W., and Lea, M. A. (1978).Cancer Res. 39, 1382-1389. Perry, M., Nelson, D., Moore, M., andchalkley, R. (1979).Biochim. Biophys. Acra561,517--528. Pickart, L., and Thaler, M. M. (1973).Nature (London) New Biol. 243, 85-87. Pickart. L.. and Thaler. M. M. (1979).FEES Len. 104, 119-122. Pickart, L., and Thaler. M. M. (1980).J . Cell. Physiol. 102, 129-139. Pickart. L.,Thalyer, L., and Thaler, M. M. (1973).Biochem. Biophys. Res. Commun. 54,562-566. Pietu, G., Chany, E., Molimard, R., and Frayssinet. C. (1978).Biomedicine 28, 169-175. Pitot, H.C., Peraino, C., Morse, P. A., and Potter, V. R. (1964).Narl. Cancer I m . Monogr. 13, 229-242. Poso, H., and Jiinne, J. (1976).Biochem. J . 158,485-488. Poso, H.,and Pegg, A. E. (1982).Biochim. Biophys. Acra 696, 179-187. Prentki, M., Crettaz, M., and Jeanrenaud, B. (1981).J . Biol. Chem. 256, 4336-4341. Prescott, D. M. (1968).Cancer Res. 28, 1815-1820. Price, J. B. (1976).Metabolism 25, 1427-1428.

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

105

Raab, K. H . , and Webb, T. E. (1969). Experienria 25, 1240-1243. Richardson, U. I . , Tashjian, A. H.. and Levine, L. (1969). J. Cell Biol. 40,236-247. Riley, D., and Weintraub, H. (1979). Proc. Narl. Acad. Sci. U.S.A. 76, 328-332. Rixon, R. H . , and Whitfield. I. F. (1976). J. Cell Physiol. 87, 147-155. Rubenstein, P . , Sealy, L., Marshall, S., and Chalkley, R. (1979). Nature (London) 280, 692-693. Rupniak, H. T., and Paul, D. (1978). J. Cell. Physiol. 96, 261-264. Russell, D. H . , and Byus, C. V. (1979). Adv. Biochem. Psychopharmacol. 15, 445-454. Saetren, H . (1956). Exp. Cell Res. 11, 229-232. Safer, B., and Jagus, R. (1981). Biochimie 63, 709-717. Schamhart, D. H. J.. van de Poll, K. W., and van Wijk, R. (1979). Cancer Res. 39, 1051-1055. Schlessinger, D. H., Pickart, L., and Thaler, M . M. (1977). Experientia 33, 324-325. Schrock, T. R., Oakman, N. J., and Bucher, N. L. R. (1970). Biochim. Biopys. Acta 204,564-577. Seale. R. L., and Simpson, R. T. (1975). J . Mol. Biol. 94, 479-501. Sealy, L., and Chalkley, R. (1978). Cell 14, 115-121. Sekas, G., and Cook, R. T. (1976). Exp. Cell Res. 102, 422-425. Sellers. L.. and Granner, D. (1974). J. Cell Biol. 60,337-345. Senshu, T., and Ohashi, M. (1979). J . Biochem. (Tokyo) 86, 1259-1267. Senshu, T., Fukuda, M., and Ohashi, M. (1978). J . Biochem. (Tokyo) 84, 985-988. Sheppard. J . R., and Prescott, D. M. (1972). Exp. CellRes. 75, 293-295. Sherman, F. G., Questler, H.. and Wimbler, D. R. (1961). Exp. CellRes. 24, 114-119. Shields, R., and Smith, J. A. (1977). J. Cell. Physiol. 91, 345-356. Short, J., Armstrong, N. B., Zemel, R., and Lieberman, 1. (1973). Biochem. Biophys. Res. Commun. 50,430-437. Short, J . , Tsukada, K., Rudert, W. A,, and Lieberman, 1. (1975). J. Biol. Chern. 250,3602-3606. Simard, A., Corneille, L., Deschamps, Y., and Verly, W. G. (1974). Proc. Natl. Acad. Sci. U.S.A. 71, 1763-1766. Sisken, J. E. (1963). In "Cinematography in Cell Biology" (G.G. Rose, ed.), p. 143. Academic Press, New York. Sisken, J . E., and Kinosita, R. (1961). J. Biophys. Biochem. Cytol. 9, 509-520. Smith, J. A., and Martin, L. (1973). Proc. Natl. Acad. Sci. U.S.A. 70, 1263-1267. Smith, J. A., and Stocken, L. A. (1975). Biochem. J. 147, 523-529. Snoek, G.T., Voorma, H. 0.. and van Wijk, R. (1981a). Biochim. Biophys. Acra 655, 107-1 12. Sn0ek.G. T . , vandePoll, K. W., Voorma, H.0.. andvan Wijk, R. (1981b).Eur. J. Biochem. 114, 27-3 1 . Snoek, G. T., Voorma, H. 0.. and van Wijk, R. (1982). Eur. J . Biochem. 123, 217-222. Spadafora, C., Bellard, M.,Compton, J. L., and Chambon, P. (1976). FEES Lett. 69, 281-285. Sperling, L., and Weiss, M. C. (1980). Proc. Narl. Acad. Sci. U.S.A. 77, 3412-3416. Sperling, L., Tardieu. A.. and Weiss, M. C. (1980). Proc. Natl. Acad. Sci. U.S.A. 77,2716-2720. Stich, H. F., and Florian, W. L. (1958). Can. J. Biochem. Physiol. 36, 855-866. Stoker, M. G. P. (1973). Nature (London) 246, 200-203. Stoker, M . G. P., and Rubin, H. (1967). Nature (London) 215, 171-172. Streumer-Svobodova, Z . , Wiegant, F. A. C., van Dongen, A. A. M. S . , and van Wijk, R. (1982). Biochimie 64, 41 1-418. Svetina, S. (1977). Cell Tissue Kiner. 10, 575-581. Tabor, H. (1962). Biochemistry 1,496-500. Taira, M., and Terayama, H. (1978). Biochim. Biophys. Acra 541, 45-58. Takatsuki, K., Fujiwara, K., Hayashi. S., Ota, Y., Torei, M., Mishiro, S., Ogato. I., Sakuma, A,, Oka, H., and Oda, T. (1981). Life Sci. 29, 2609-2617. Tanigawa, Y.,Kawamura, M., and Shimoyama. M . (1977). Biochem. Biophys. Res. Commun. 76, 406-4 12.

106

ROELAND VAN WIJK

Temin, H. M. (1971). J . Cell. Physiol. 78, 161-170. Theoharides, T. C., and Canellakis, Z. N. (1975). Nature (London) 255, 733-734. Thomas, G., Siegmann, M., Kubler, A. M., Gordon. J., and Jimenez de Asua, L. (1980). Cell 19, I015- 1023. Thomas, J. O., and Thompson, R. J. (1977). Cell 10, 633-640. Thompson, E. B . , Tomkins, G. M., and Curran, J. F. (1966). Proc. Nut/. Acud. Sci. U.S.A. 56, 296-303. Tidwell, T., Allfrey, V. G., and Mirsky, A. E. (1968). J . Biol. Chem. 243, 707-715. Till, J. E., McCullough, E. A.. and Siminovitch, L. (1964). Proc. Narl. Acud. Sci. U.S.A. 51, 29-36. Todaro, G. J., Lazer, J. K., and Green, H. (1965). J . Cell. Comp. Physiol. 66, 325-334. Tomita, Y., Nakamura, T., and Ichihara, A. (1981). Exp. Cell Res. 135, 363-373. Tsuboi, M . (1964). Bull. Chem. SOC. Jpn. 37, 1514-1522. Tushinski, R. J . , and Warner, I. R. (1982). J. Cell Physiol. 112, 128-135. Valleron, A.-J., Guiguet, M., and Puiseux-Dao, S. (1981). In “Biomathematics and Cell Kinetics” (M. Rotenberg. ed.), pp. 195-210, Elsevier, Amsterdam. Van Meeteren, A., Zoutewelle, G., and van Wijk, R. (1981). Cell Eiol. tnr. Rep. 5, 467. Van Meeteren, A., Loesberg, C., van Rijn, J., and van Wijk, R. (1983). In Vitro. in press. van de Poll, K. W., van Aken, J. H., and van Wijk, R. (1979). Cell Biol. Int. Rep. 3, 247-255. van Rijn, J . , Bevers, M. M., van Wijk, R., and Wicks, W. D. (1974). J. Cell Eiol. 60, 181-191. van Rijn, 1.. van den Berg, I., van Meeteren, A., and van Wijk, R. (1983). Radiation Res.. in press. van Steeg, H. (1982). Thesis, University of Utrecht. van Wijk, R. (1983). To be published. van Wijk, R., and van de Poll, K. W. (1979). Cell Tissue Kine!. 12, 659-663. van Wijk, R . , Wicks, W. D., and Clay, K. (1972). CuncerRes. 32, 1905-1911. van Wijk, R., Wicks, W. D., Bevers, M. M., andvanRijn, J . (1973). CuncerRes. 33, 1331-1338. van Wijk, R., van de Poll, K. W., Amesz, W. J. C., and Geilenkirchen, W. L. M. (1977a). Exp. Cell Res. 109, 371-379. van Wijk. R.. Koontz, J . W., and Wicks, W. D. (1977b). In Vitro 13, 162-163. van Wijk, R. Zoutewelle, G., and van de Poll, K. W. (1979a). Cell Biol. In!. Rep. 3, 447-452. van Wijk, R., Zoutewelle, G.,Defer, N., Tichonicky, L., and Kruh. J . (1979b). Eiochimie 61, 711-717. van Wijk, R., Tichonicky, L., and Kruh, J. (1981). In Vitro 17, 859-862. van Zoelen, E. J. J., van der Saag, P. T., and de Laat, S. W. (1981). Exp. Cell Res. 131,395-406. Venetianer, A., Pint&, Z., and Gal, A. (1980). Cyrogener. Cell Gener. 28, 280-284. Verly, W. G. (1973). Nutl. Cancer Inst. Monogr. 38, 175-184. Verly, W. G. (1976). In “Chalones” (J. C. Houck, ed.), pp. 401-428. American Elsevier, New York. Verly, W. G., Deschamps, Y., hshpathddam, J., and Desroisiers, M. (1971).Can. J. Biochem. 49, 1376-1383. Vinet, B., and Verly, W. 0. (1976). Eur. J. Cancer 12, 211-218. Volman, H. (1978). Virchows Arch. B Cell Puthol. 26, 249-259. Wagner, K., Roper, M. D., Leichtling, B. H., Wimalasena, I., and Wicks, W. D. (1975). 1.Biol. Chem. 250, 231-239. Walker, P. R., and Potter, V. R. (1972). Enzyme Regul. 10, 339-364. Walker, P. R . , and Whitfield, J. F. (1977). J. Cell Eiol. 75, 275a. Walker, P. R . , and Whitfield, J. F. (1978). Proc. Nutl. Acud. Sci. U.S.A. 75, 1394-1398. Walker, P. R., Boynton, A. L., and Whitfield, J . F. (1977). J . Cell. Physiol. 93, 89-98. Wang, J . H . , and Waissrnan, D.M. (1979). Curr. Top. CellRegul. 15, 47-107. Weber, G. (1966). Gann Monogr. 1, 151-178.

REGULATION OF DNA SYNTHESIS IN HEPATOMA CELLS

107

Weber, G. (1968). Natunvissenschafren 55, 418-429. Weber, G . , and Lea, M. (1967). Methods Cancer Res. 2, 523-578. Weinhouse, S. (1972). Cancer Res. 32, 2007-2016. Weintraub, H. (1973). Cold Spring Harbor Symp. Quant. Biol. 38, 247-256. Weintraub, H. (1979). Nucleic Acids Res. 5, 1179-1 188. Whitfield, J. F., Boynton, A. L., MacManus, J. P., Rixon, R. H., Walker, P. R., and Armato, U. (1976). In “Cyclic Nucleotides and the Regulation of Cell Growth” (M. Aou-Sabe, ed.), pp. 97-130. Dowden, Hutchison & Ross, New York. Wicks, W. D., van Wijk, R., and McKibbin, J. B. (1973a). Adv. Enzyme Regul. 11, 117-135. Wicks, W. D., van Wijk, R.. Clay, K., Bearg, C., Bevers, M. M., and van Rijn, I. (1973b). In “The Role of Cyclic Nucleotides in Carcinogenesis” (J. Schultz and H. G. Gratzner, eds.), pp. 103-124. Academic Press, New York. Wicks, W. D., Koontz, J.. and Wagner, K. (1975). J. Cyclic Nucleoride Res. 1, 49-58. Wiegand. L., and Pegg, A. E. (1978). Biochim. Biophys. Acta 517, 169-180. Wielckens, K., Bredehorst. K., Adamietz, P., and Hilz, H. (1981). Eur. J. Biochem. 117, 69-74. Wimalasena, J., and Wicks, W. D. (1979). Mol. Pharmacol. 16, 449-461. Wimalasena, J., Leichtling, B. H., Lewis, E. J.. Langan, T . A., and Wicks, W. D. (1930). Arch. Biochem. Biophys. 205, 595-605. Wolff, D. J., Ross, J. M., Thompson, P. N., Brostrom, M. A., and Brostrom, C. D. (1981). J. Biol. Chem. 256, 1846-1860. Yoshikura, H., and Hirokawa, Y. (1968). Exp. Cell Res. 52, 439-444. Zeilig, C. E., and Goldberg, N. D. (1977). Proc. Nail. Acad. Sci. U.S.A. 74, 1052-1056.

This Page Intentionally Left Blank

INTERNATIONAL REVIEW OF CYTOLOGY. VOL. 85

Somatic Cell Genetics and Gene Mapping FA-TENKAO Eleanor Roosevelt Institute for Cancer Research, and Department of Biochemistry, Biophysics and Genetics, University of Colorado Health Sciences Center, Denver, Colorado

.............. I. Introduction . . . Historical View ............... rrr. Development of Somatic Cell Genetics . . . , . . . . . . . . . . . . . . . . . . . . IV. Use of Somatic Cell Genetics in Gene Mapping. . A. Discovery of Preferential Loss of Human Chromosomes in Rodent-Human Cell Hybrids . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Establishment of Permanent Cell Cultures of Rodent-Human Cell Hybrids .......................................... C. Identification of Specific Human Chromosomes Retained in Cell Hybrids.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Synteny Analysis . E. Chromosomal Assignment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F. Regional Assignment. . . . . . . . V. Combined Use of Somatic Cell Genetics and Recombinant DNA Technology in Gene Mapping.. . . . . . . . . . . . . . . . . A. Use of Cloned Genes and Cell Hybrids for Ge B. Assignment of Random DNA Segments to Specific Human ......................... Chromosomes gments Isolated from Specific 11.

109 110 112 116

I I6 117 1 I8 120 121 123 126 127

130 131

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

Individual Chromosome E. Fine Structure Mapping F. In Situ Hybridization in VI. VI1. VIII.

ma1 Segments. . . . .

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

ap . . . . . . . . . .

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

132 133 134 135 138 139 140

I. Introduction The human genome is a complex structure, consisting of approximately 3 X lo9 base pairs (bp) per haploid genome distributed to 22 autosomes and one of the sex chromosomes. It has enough DNA to code for several million genes. However, various estimates have placed the actual number of structural genes I09 Copyright 0 1983 by Academic Press. Inc. All rights of repduction in any form resewed. ISBN 0-12-364485-2

110

FA-TEN KAO

(coding for the amino acid sequence of proteins) in the order of 50,000 to 100,OOO (McKusick and Ruddle, 1977). One estimate was based on chiasma frequency in the mouse in which the density has been estimated to be 20 structural genes per centimorgan, cM (cM is a genetic unit of the map distance equivalent to 1% recombination frequency between two genes). The human genome has about 3300 cM (Renwick, 1969), thus containing approximately 66,000 structural genes. The remaining DNA may code for genes with regulatory or structural functions. Furthermore, introns abundantly present in nearly all structural genes represent a significant portion of the noncoding DNA in the genome. Moreover, repetitive sequences like the Alu family and others whose functions remain to be elucidated constitute more than 30% of the human genome (Schmid and Deininger, 1975; Houck et al., 1979; Schmid and Jelinek, 1982). Thus, the estimated number of structural genes in the human genome may appear to be within reasonable expectation. About 1600 human genes have been confidently identified and about the same number of genes have been less confidently identified (McKusick, 1978, 1982). Among the genes that have been identified, more than 400 genes (plus about 150 DNA segments of unknown function) have now been mapped to all autosomes and the sex chromosomes (Human Gene Mapping 6, 1981). About 115 genes (plus about equal numbers of DNA segments) have been mapped to the X chromosome. Of particular significance is the assignment of autosomal genes by the somatic cell genetic method. Of the 345 autosomal assignments (including DNA segments), 67 were assigned by family study, 202 by cell hybrid study, 14 by both family and cell hybrid methods, and 62 by other methods (McKusick, 1982). The achievement of gene mapping by the somatic cell hybrid method clearly demonstrates the contribution made by a series of technological breakthroughs which have led to an entirely new approach to mapping genes of not only the human genome, but also the genomes of at least another 24 mammalian species including chimpanzee, gorilla, orangutan, rhesus, baboon, African green monkey, new world monkey, lemur, mouse, rat, deer mouse, golden hamster, Chinese hamster, Syrian hamster, rabbit, cat, American mink, dog, pig, Indian muntjak, cow, sheep, horse, and red kangaroo (Human Gene Mapping, 6, 1981). Although this article deals mainly with the application of somatic cell genetics to gene mapping in the human genome, the same methodologies have been applied equally well to mapping of genes in many other mammalian species where somatic cell genetics is applicable.

11. Historical View of Gene Mapping

Classical linkage studies in experimental animals and plants identify linkage groups in the genomes of respective species. In the species where cytogenetic analysis is practical, assignment of linkage groups to specific chromosomes can

SOMATIC CELL GENETICS AND GENE MAPPING

111

be made. In addition, informative crosses can be designed to estimate the distances among genes in terms of recombination frequency, i.e., the greater the distance between two genes, the higher the frequency of crossing over between them. Among mammalian species, the most complete gene map that has been constructed using this approach is that of the mouse (Green, 1968; Miller and Miller, 1972; Kozak and Ruddle, 1976). In humans, however, such linkage analysis had progressed at a much slower rate. The main reasons include (1) long generation time, (2) small family size, and (3) the most desirable and illuminating pedigree data are often not available. The first human gene assigned to a specific human chromosome is the color blindness gene. Following the pattern of inheritance of the color blindness in a number of families, Wilson concluded in 1911 that the gene must reside on the X chromosome (Wilson, 1911). The progress of gene assignment to specific autosomes by this method has been very slow. Most of the assigned genes have only been identified as either X-linked or autosomal. Between 1911 and 1967, less than 100 genes had been assigned to the X chromosome and very few genes had been assigned to autosomes. It should be pointed out, however, that the correct human chromosome number of 46 was not established until 1956 (Tjio and Levan, 1956), and the identification of each individual chromosome was not possible until 1970 (Caspersson er al., 1970). Thus, the recent accelerated progress in human gene mapping has coincided with the development of modem cytogenetics. The first autosomal gene assigned to a specific human chromosome was the Duffy blood group. Using a morphological marker “uncoiled” region on chromosome 1 , Donahue et al. (1968) assigned the Duffy blood group to this chromosome. Several other genes have been assigned to specific autosomes using a similar strategy, namely, by associating a gene with an identifiable chromosomal abnormality. Thus, the a-polypeptide of heptoglobin was assigned to chromosome 16 (Robson et al., 1969), and the major histocompatibility complex (HLA) to chromosome 6 (Lamm et al., 1974). However, the progress made by this method has been slow and only a few autosomal genes have been assigned in this way. Genes can be shown to be linked to each other by statistical analysis of the pedigree data derived from informative families. The most useful method has been the lod (log of the odds) score method (Haldane and Smith, 1947; Morton, 1955; Race and Sanger, 1975). This method measures the likelihood (odds) of linkage between two genes with a given recombination fraction theta (0). The “odds” is expressed as the ratio of the probability of the observed pedigree assuming that the two genes are linked (0 = 0), to the probability of observed pedigree assuming that the genes are not linked (0 = 0.50). The log,, of the likelihood ratio is the lod score. Lod scores of +3.0 or more are considered very strong evidence for linkage, while lod scores of -2.0 or less provide evidence against linkage. The “odds” can be tested for different recombination values (0

112

FA-TENKAO

= 0-0.50) against the probability of no linkage (0 = 0.50) and the recombination frequency giving the highest lod score will be used as the maximum likelihood estimate of linkage. Although linkage analysis will not assign genes directly to a particular chromosome, if one of the genes in a linkage group is assigned to a particular chromosome, all the other genes in the same group can be indirectly assigned to that chromosome. For example, AK1 (adenosine kinase-1) and the ABO blood group were shown to be linked by the pedigree method (Ropley et al., 1967). When AK1 was assigned to chromosome 9 by cell hybrid studies (Westerveld et al., 1976), the ABO blood group was thus assigned to the same chromosome. The breakthrough in human gene mapping came when an entirely new approach was adopted due to the development in somatic cell genetics and the successful fusion of somatic cells. Weiss and Green (1967) first observed in mouse-human cell hybrids that human chromosomes were preferentially lost. This finding marked the beginning of the rapid progress in gene mapping using somatic cell hybrids. This new approach can be used not only for establishing linkage among genes, but also for assigning genes to specific chromosomes, and in many cases to a specific region of the chromosome. Thus, expedient advances in human gene assignment ensued and, by now, more than 400 genes have been mapped to various human chromosomes. The power of gene mapping has also been extended to the fine structure level by combining the cell hybrid method and recombinant DNA technology. A detailed human gene map can assist in prenatal diagnosis by making more reliable predictions of the consequences of genes involved in chromosomal abnormalities detected by amniocentisis. Of equal or even greater importance, the availability of a detailed gene map may aid in the elucidation of the relationship between the geometric location of a gene and its function, and the identification of regulatory sequences and their relations to the function of structural genes. Such understanding is essential in unraveling the regulatory mechanisms underlying development and differentiation during normal and diseased states. Additional reviews on human gene mapping include the following articles: Ruddle and Creagan (1973, Ringertz and Savage (1976), McKusick and Ruddle (1977), Siniscalco (1979), Conneally and Rivas (1980), McKusick (1980, 1981), Ruddle (1982), and Puck and Kao (1982). In addition, six international conferences on human gene mapping had been held between 1973 and 1981, and reports for each conference have been published, including the most recent Oslo Conference held in 1981 (Human Gene Mapping 6, 1981).

111. Development of Somatic Cell Genetics

In 1955, a new approach to studying mammalian genetics was developed as a result of advances in techniques for plating and growth of single mammalian

SOMATIC CELL GENETICS AND GENE MAPPING

113

cells in vitro achieved by Puck and Marcus (1955). Thus, cloning of mutant cells to ensure genetic homogeneity became possible. Such advances paved the way for studying genetics in cultured somatic mammalian cells as effectively as in microbial systems (Puck, 1972). Subsequently, various mammalian cell mutants were isolated, including auxotrophic mutants (Kao and Puck, 1968), drug-resistant mutants (Chu and Malling, 1968), temperature-sensitive mutants (Thompson et al., 1970), UV-sensitive mutants (Stamato and Waldren, 1977), regulatory mutants (Sinensky, 1977, 1978), etc. Using the ability to support single cell growth in chemically defined synthetic media, Ham (1965) identified the specific chemical components and their critical concentrations used in the growth medium. During the titration process, a subclone of the Chinese hamster ovary (CHO) cells was found to require proline for growth (Ham, 1963). This mutant clone, pro-, was later shown to lack the enzyme converting glutamic acid to proline (Kao and Puck, 1967) and the human complementing gene for the CHO enzyme deficiency was assigned to human chromosome 10 (Jones, 1975). The chemically defined medium F12 developed by Ham (1965) was designed for maximal single cell growth of CHO cells. Using this medium and the procells in reconstruction experiments, a mutant isolation procedure was developed for isolating additional auxotrophic mutants in CHO cells (Puck and Kao, 1967; Kao and Puck, 1968). This method utilized the differential killing of normal cells containing 5-bromodeoxyuridine (BUdR) in the DNA by near-visible light, while mutant cells containing no BUdR under the prescribed conditions survived the treatment. Using this BUdR + visible light method, large numbers of auxotrophic mutants were isolated which lack the enzymes involved in biosynthesis of amino acids, purines, pyrimidines, etc. (Kao and Puck, 1975). These mutants have proved extremely useful not only in studying biosynthesis and metabolic regulation of various biochemical pathways (Patterson, 1980; Patterson et al., 1981), but also in serving as genetic markers in somatic cell genetic analysis with high resolving power (Puck, 1972). Complementation analysis of the mutants has identified at least 4 different genes for glycine biosynthesis (Kao ef al., 1969a), more than 10 genes involved in the purine biosynthesis (Patterson er al., 1981), and 3 genes in pyrimidine biosynthesis (Patterson, 1980). Since these mutants are recessive, they are extremely useful in retaining specific human chromosomes in cell hybrids for gene mapping and other chromosomal and fine structure studies (Kao er al., 1976; Jones et al., 1980; Law et al., 1982). The other highly useful mutant system is the HPRT (hypoxanthine guanine phosphoribosyltransferase) marker. Originally developed by Szybalska and Szybalski (1962), the HPRT- mutant lacks the enzyme HPRT which converts hypoxanthine or guanine to inosinic acid or GMP in the salvage pathway of purine biosynthesis. This marker affords both forward and reverse selection for

114

FA-TEN KAO

the loss (forward) and regain (reverse) of the marker in the cell. Thus, in the presence of the purine analog 8-azaguanine or 6-thioguanine, mutants lacking HPRT are drug resistant and survive. On the other hand, revertants possessing HPRT activity can be isolated in the HAT selective medium containing hypoxanthine, aminopterin, and thymidine (Littlefield, 1964). Using the HAT selection, it is also possible to isolate cell hybrids derived from fusions between HPRT- mutant and HPRT+ cells. The HPRT+ hybrids can be selected by 6thioguanine treatment. This system is particularly useful in a number of genetic operations to be described in later sections. Other selective systems possessing selective features like HPRT are APRT (adenine phosphoribosyltransferase) (Chasin, 1974) and TK (thymidine kinase) (Littlefield, 1964) markers. Another important development in somatic cell genetics which is crucial to gene mapping is karyotype analysis in various mammalian species using cells grown in culture. Either fibroblast cell cultures or stimulated lymphocyte cultures can be used for delineating the correct chromosome number and the chromosome identity of any animal species. Thus, the correct human chromosome number was proved to be 46 by taking cells from various tissues and growing them in culture to prepare metaphase chromosome spreads for accurate chromosome counts (Tjio and Puck, 1958). The Denver Classification System of human chromosomes was established in 1959 using karyotypes derived largely from cultured human cells as the first system to classify human chromosomes. Since then, more refined techniques have been developed for unequivocal identification of each individual human chromosome by its characteristic banding pattern (Caspersson er al., 1970). Additional development in high-resolution chromosome analysis (Yunis, 1976) utilized a synchronization technique to arrest human cells in S phase followed by release to collect more cells in late prophase, prometaphase, and early metaphase, all of which exhibit longer chromosomes than those in metaphase. By this procedure, identifiable bands can be increased from about 850 to more than 1000. This method can facilitate detection of small deletions and more precise determination of the breakpoints of chromosome translocations, thus increasing the resolution and accuracy of regional mapping of genes to small areas of the chromosome. In addition to the above developments, an important finding took place which greatly expanded the power of somatic cell genetic analysis and immensely facilitated gene mapping. In 1958, Okada (1958) first observed fusion of two different tumor cells infected with HVJ virus (hemagglutinating virus of Japan), a strain of mixovirus of the parainfluenza group very similar if not identical to Sendai virus. Later, Barski and his associates (Barski er al., 1960) reported hybrid cell formation after cocultivation of two different cell lines with characteristic marker chromosomes. However, the frequency of spontaneously occurring cell hybrids was low and the isolation of hybrid cells using this method would require an efficient selective system. Ephrussi soon confirmed Barski’s

SOMATIC CELL GENETICS AND GENE MAPPING

115

observation and recognized the importance and implications of cell fusion for genetic analysis in somatic cells (Sorieul and Ephrussi, 1961). He and his coworkers devised procedures for identifying and isolating hybrid cells for detailed characterization (Ephrussi, 1972). In 1964, Littlefield used the HAT selective system of Szybalski to isolate hybrids that occurred spontaneously at low frequencies (Littlefield, 1964). In this system, genetic complementation was exploited in which two mutants deficient in different genes, HPRT and TK, respectively, were mixed and cultured together. Hybrids possessing normal counterparts of these mutant genes survived in the HAT selective medium. The frequency of hybrids isolated under such per one parental cell. conditions was about In addition, half-selective systems were also developed in which only one parent possessed selective marker. One such procedure involved the use of human lymphocytes or leukocytes as the human parent while the rodent parent had a selective marker (Miggiano et al., 1968). After fusion, the unfused human cells could be removed by medium change. Another half-selective system employed was mixing in the fusions far more cells with selective marker than the cells without selective marker (Davidson and Ephrussi, 1970). After selection, the survivors were either hybrids or one of the parental cells without selective marker. The latter could be identified by cell morphology, cell size, cytogenetic analysis, or specific isozyme markers. The great impetus to the wide use of cell fusion as a research tool came when Harris and Watkins (1965) demonstrated high frequency of cell fusion mediated by inactivated Sendai virus. Cell fusion was achieved in many different cell types over a wide range of different animal species, including fibroblasts, macrophages, lymphocytes, and leukocytes, derived from mouse, Chinese hamster, rat, rabbit, and human (Harris, 1970). The procedure was immediately adopted by many laboratories and soon became an important technique for a large variety of cellular and genetic investigations (Ringertz and Savage, 1976). The outcome of cell fusion can be studied at two stages: heterokaryons and hybrid cells. Heterokaryons are fused cells containing nuclei derived from the two parental cells. Studies like the reactivation of chick erythrocytes (Hams, 1967), the role of the nucleolus in gene expression (Harris et al., 1969), the regulation of DNA synthesis (Rao and Johnson, 1970), and mitotic cell-induced premature chromosome condensation (Johnson and Rao, 1970) have been well demonstrated by heterokaryon analysis. When heterokaryons move into mitosis synchronously, the nuclear membrane disintegrates and metaphase chromosomes from the two nuclei coalesce. At the end of the first cell division after fusion, the fused cell contains a single nucleus with the combined genomes from the two parental cells. It is called a hybrid cell. Long-term culture of hybrid cells can then be established for further genetic analysis.

116

FA-TEN KAO

Cell hybrids have been highly useful in studying many important aspects of mammalian and human genetics, for example, the determination of dominance and recessiveness (Kao and Puck, 1972), the identification of genetic complementation groups among mutants with same or different phenotypes (Littlefield, 1964; Kao et al., 1969a,b), the elucidation of regulatory phenomena in gene expression particularly for tissue-specific functions (Davidson, 1974; Ringertz and Savage, 1976), and the formation of hybridoma for monoclonal antibody production (Kohler and Milstein, 1975). Cell fusion can also be accomplished by using polyethylene glycol (Pontecorvo, 1975). Perhaps one of the most valuable contributions of cell hybrids to human genetics has been in the area of gene mapping. This new approach has revolutionized the field which was previously studied mainly by family method. Using cell hybrids for gene mapping, impressive progress has already been made over the last decade. Further refined procedures and more useful cell hybrids are being developed which will make gene mapping even more efficient, especially in the combined use with the powerful recombinant DNA techniques. A ray of hope for an eventual construction of a complete human gene map is appearing on the horizon (Botstein et al., 1980), and a concerted effort from many interested laboratories is necessary to achieve this goal.

IV. Use of Somatic Cell Genetics in Gene Mapping A. DISCOVERY OF PREFERENTIAL LOSS OF HUMAN CHROMOSOMES IN RODENT-HUMAN CELLHYBRIDS When two rodent cell lines like mouse A9 and B82, or Chinese hamster cell mutants gly- and ade-, are fused, the resulting hybrids usually retain the combined chromosome number of the two parental cells. After long-term growth, the hybrids may lose about 10% of the chromosomes before they reach a stable modal chromosome number (Kao et al., 1969b; Littlefield, 1964). However, a totally unexpected result was found by Weiss and Green (1967) in which preferential and extensive loss of human chromosomes resulted in hybrids between a mouse cell line and a diploid, normal human cell culture. In these experiments, the mouse cell line B82 lacking thymidine kinase (TK) was used and hybrids were selected in HAT medium. The hybrids were found to retain the entire mouse genome but had lost nearly all human chromosomes except one. This chromosome was later identified to be human chromosome 17 (Migeon and Miller, 1968; Miller et al., 1971). Since the survival of these hybrids in HAT medium required the presence of thymidine kinase, the logical interpretation of these results was that human chromosome 17 carried a gene coding for thymidine kinase. This demonstration provided the first case of using cell hybrids for

SOMATIC CELL GENETICS AND GENE MAPPING

117

mapping a gene to a specific human chromosome. It should be pointed out that in this case, only one human chromosome was contained in the hybrid so that any human gene products found in the hybrid could be assigned to that chromosome. It was also found that different cell hybrids could contain different combinations of more than one human chromosome. Gene assignments can also be made in such hybrids by correlating the simultaneous presence or absence of a gene product with a specific human chromosome. Such hybrids with multiple human chromosomes have now been shown to be extremely useful as a general means for a quick and systematic way for gene mapping.

B . ESTABLISHMENT OF PERMANENT CELL CULTURES OF RODENT-HUMAN CELLHYBRIDS The most widely used selective system for establishing rodent-human cell hybrids has been the use of mouse mutant cell lines lacking an active HPRT gene (HPRT- ) and normal human fibroblasts or lymphocytes. Although the human parental cells used in these fusions generally possessed no specific selective markers, they could be eliminated either by washing (in case of lymphocytes) or by applying ouabain at a concentration M ) capable of killing human fibroblasts but leaving the hybrid cells unaffected (Baker, 1976; Law and Kao, 1978). Under selective conditions of HAT and ouabain, hybrid clones can be isolated about 2-3 weeks after fusion. Either inactivated Sendai virus or polyethylene glycol (Pontecorvo, 1975) can be used to promote cell fusion. Since the HPRT gene has been mapped to the human X chromosome, all the surviving hybrids must retain this chromosome. In addition, various other human chromosomes are also retained in different hybrids in a more or less random fashion for most of the chromosomes. After the initial loss of most human chromosomes in the hybrids, the remaining chromosomes tend to be stabilized, with only small losses over long periods in culture. Thus, the hybrids can be characterized for their human chromosome content at about 2-3 months after isolation. In addition to the HAT selective system, other selective techniques have also been used to establish useful cell hybrids. One'of these has been the use of auxotrophic mutant markers. Large numbers of such mutants have been produced in CHO-K1 cells (Kao and Puck, 1974; Patterson et al., 1981). It has been shown that human-Chinese hamster cell mutants also preferentially lose human chromosomes (Kao and Puck, 1970). The Chinese hamster mutant cells can be easily eliminated from the cell fusion population by growth in the nutritionally deficient medium designed for each specific auxotrophic mutant employed, for example, the glycine-requiring mutant can be eliminated by using glycine-free medium (Jones et af., 1972; Kao et af., 1976). Hybrids selected under such conditions will retain not only the specific human chromosome which comple-

118

FA-TEN KAO

ments the auxotrophy of the CHO-Kl mutant, but also other human chromosomes for general use in gene mapping (Jones et al., 1981; Kao et al., 1982). c . IDENTIFWATION OF SPECIFK HUMANCHROMOSOMES RETAINED IN CELLHYBRIDS Two major techniques, cytogenetic and isozyme analysis, have been used for identifying each of the 24 different human chromosomes in the hybrids. Other techniques, such as immunologic and nutritional analysis, can also be used for identifying certain specific human chromosomes, for example, the cell surface antigen marker A, for human chromosome 11 (Puck et al., 1971). Recently, various single-copy DNA segments from the human genome have also been used as markers for specific human chromosomes (Kao et al., 1982; De Martinville et al., 1982; Human Gene Mapping 6, 1981). Cytogenetic analysis using chromosome banding techniques developed in 1970 (Caspersson et af.,1970) has provided unequivocal identification of each human chromosome retained in hybrids. Furthermore, this analysis can also reveal the extent of heterogeneity of a specific human chromosome retained in the hybrids. By analyzing karyotypes of a large number of cells from each hybrid, the percentage of cells carrying a specific human chromosome can be determined. If the proportion of a chromosome retained in a particular hybrid is below a certain level (for example, 20%), the detection of a particular gene product of that chromosome will become questionable. Such hybrids should be excluded from mapping for that chromosome. The development of the alkaline Giemsa-11 differential staining method (Bobrow et al., 1972; Friend et al., 1976; Alhadeff et al., 1977) has also facilitated identification of human chromosomes in the rodent-human cell hybrids. Using this method, human chromosomes are stained light blue and rodent chromosomes like mouse and Chinese hamster are stained magenta. Even translocations between rodent and human chromosomes are discernible as to the species origin of the chromosome segments in the translocation (Friend et af., 1976; Kobutcher and Ruddle, 1979). Figure 1 presents metaphase chromosomes treated with either the Giemsa-11 procedure (Fig. 1A) or the trypsin banding technique (Fig. 1B) revealing the identity of human chromosme 14. The second widely used technique for identifying human chromosomes in the hybrid is isozyme analysis. Isozymes are enzymes catalyzing the same biochemical reaction but differing in certain physical or chemical properties (Marken and Moller, 1959). Isozymes are usually present among different species and a number of enzyme analyses can be used to distinguish them, such as pH optimum, thermostability, electrophoretic mobility, etc. The most commonly used technique involves the separation of isozymes by electrophoresis followed by simple histochemical staining to reveal the different location of the isozymes in

SOMATIC CELL GENETICS AND GENE MAPPING

119

FIG. 1. Metaphase cells of a human-CHO hybrid containing a complete set of the CHO genome plus a single human chromosome 14 as revealed by the Giemsa-1 I differential staining (A) and trypsin banding (B)techniques. This hybrid is positive for human NP (nucleoside phosphorylase), an isozyme marker for human chromosome 14. Arrows point to human chromosome 14.

120

FA-TEN KAO

H

1

Ch

2

+

3

I

4

+

FIG.2. Zymogram showing isozyme analysis for identifying the human gene product in the hybrids. Peptidase B is an isozyme marker for human chromosome 12. The positive hybrid containing human chromosome 12 exhibits both human (H)and Chinese hamster (Ch) isozymes (lane 3) and the negative hybrid containing no human chromosome 12 exhibits only the hamster isozyme (lane 4).

siru (Harris and Hopkinson, 1976). Figure 2 illustrates such an isozyme analysis.

Peptidase-B (PEPB) is an isozyme marker for human chromosome 12. The human enzyme (lane 1) and Chinese hamster enzyme (lane 2) migrated to tiifferent positions. The positive hybrid (lane 3) containing human chromosorne 12 exhibited both human and CHO isozymes and the negative hybrid (lane 4) containing no human chromosome 12 showed only CHO isozyme. This technique is simple enough to analyze large numbers of isozymes in many cell hybrids. When an isozyme marker is assigned to a specific human chromosome, the presence of this isozyme marker in the hybrid will indicate that at least a part of that chromosome is retained in the hybrid. Each human chromosome has at least one isozyme marker that can be assayed conveniently. .Many chromosomes have isozyme markers for each arm. Thus, the presence of isozyme markers near the distal part of the two arms of a chromosome strongly suggests the presence of an intact human chromosome in the hybrid.

D. SYNTENYANALYSIS In classical genetic analysis of experimental animals and plants, the progeny produced from informative crosses can be analyzed to determine linkage relationships among two or more genes. For each crossover event occurring in a meiotic cell, only 50% of the progeny will be the crossover type and the other 50% will be the wild type. This is because crossing over takes place between homologous chromosomes during the four-strand stage in meiosis and only two of the four strands (or chromatids) are engaged in crossing over. The other two strands remain intact. Thus, among the four meiotic products which give rise to

SOMATIC CELL GENETICS AND GENE MAPPING

121

offspring, only half will be recombinant and the other half will be parental type. In such an analysis, when crossing over occurs between two genes in 100% of the meiotic cells, only 50% of the offspring will be recovered as recombinant. This system sets the maximum observable recombinant frequency at 50%, indicating that the two genes are either far apart on the same chromosome, or on different arms of the same chromosome, or on different chromosomes, and are exhibiting independent segregation. Hence, if a recombination frequency between two genes is 2096, the crossing over event between the two genes has actually occurred in 40% of the meiotic cells. In addition, the linkage analysis is restricted to one arm only. No linkage measurements can be made across the centromere; that is, the two genes on opposite arms always behave as if they are on separate chromosomes. Now, let us turn to the linkage analysis using somatic cell hybrids. Instead of analyzing segregation of gene markers in the offspring, we are following the segregation of an entire chromosome in the cell hybrids. If two genes are on the same chromosome, they always segregate together. If they are on different chromosomes, they segregate more or less randomly. For this reason, Renwick (1971) coined a new term “synteny” (Greek: syn meaning together and taenia, ribbon) to express linkage relationships between two genes by cell hybrid analysis. Two genes are syntenic if they are on the same chromosome and segregate together (concordant segregation) in cell hybrids, regardless of their physical distance on the chromosome or on the same or opposite arms.

E. CHROMOSOMAL ASSIGNMENT A gene can be assigned to a specific human chromosome if it is shown to be syntenic to an isozyme marker of known chromosomal location, or if the gene segregates with a particular chromosome in the hybrids. Initially, only some hybrids have been characterized for all 24 different human chromosomes by either cytogenetic or isozyme techniques or both. A large number of hybrids have to be used to establish assignment to a particular chromosome, and also to rule out assignment to all other chromosomes. As more and more hybrids have been fully characterized in various laboratories (Owerbach et al., 1980a; Hobart et al., 1981; Slate et al., 1982; Kao et al., 1982; Hershfield and Francke, 1982; Lai er al., 1982; Heisterkamp er al., 1982; Swan et al., 19821, it is now possible to use a set of well-selected hybrids which have unique combinations of human chromosomes capable of distinguishing all 22 autosomes and the X and Y chromosomes. It has been calculated that the theoretical minimum number of hybrids required for such a set is five (Ruddle and Creagan, 1975). However, due to possible breakage and segregation of some human chromosomes in some hybrids, it is necessary to use additional hybrids to confirm the initial assignment established by the minimum hybrid set. Because such a set of hybrids has been

122

FA-TEN KAO

analyzed for human chromosome content for all 24 different chromosomes, the particular chromosome with the highest frequency of concordant segregation with a gene is most likely to be the chromosomal assignment of that gene. Occasional discordant hybrids can be analyzed for possible chromosome breakage. If they are shown to have identifiable terminal deletions, they should provide information on regional mapping of the gene. Efforts should also be directed to extensive characterization of large numbers of different enzymes or proteins in their physical, chemical, and immunological properties for use in gene mapping. Any gene product that can be distinguished between human and rodent species and is produced in the cell hybrids can, in principle, be mapped by somatic cell hybrid method. The development of twodimensional gel electrophoresis for separating proteins (O’Farrell, 1975) has added another useful analytical technique to the recognition and mapping of genes coding for these proteins (McConkey, 1980; Scoggin et al., 1981; Cox et al., 1981). Although using recombinant DNA techniques has increased enormously the mapping capacity for structural genes as will be described later, the mapping of genes with regulatory functions will largely depend on the analysis of gene expression and gene products in the cell hybrids. It is thus desirable to develop methods for inducing and maintaining tissue-specific functions in cell hybrids for mapping of such genes. Another approach to establishing useful hybrids for efficient and systematic gene mapping is to construct hybrids containing only a single human chromosome. For example, the CHO-K1 auxotrophic mutants have been fused with human cells and grown in the selective medium F12D (Kao and Puck, 1968); the surviving hybrids must retain the human chromosome which can provide the complementing gene for the auxotrophic deficiency. Specific hybrids have been isolated and shown to retain a complete CHO genome plus a single human chromosome 12 in the gly- A-human fusion (Kao et al., 1976; Law and Kao, 1978), and a single chromosome 21 in the ade-C-human fusion (Moore et al., 1977). Using similar procedures, additional single human chromosome hybrids have been established for chromosomes 8 (Jones et al., 1981), 9 (Jones el al., 1980), 14 (Lai et al., 1982), and the X (Migeon et al., 1981). It should be pointed out that in this method, the selective markers used in the fusion must be recessive in order to retain a specific human chromosome in the hybrids. The unique feature of such hybrids is that they can permanently retain the specific human chromosome as long as they are grown under selective conditions, for any hybrid losing the critical human chromosome will be eliminated from the culture. Thus, these hybrids can provide a homogeneous population of hybrid cells all possessing the same human chromosome. It is hoped that a complete set of hybrids will be established each containing a single, different human chromosome. Such hybrids will be useful not only for very efficient gene mapping with little ambiguity, but also for fine structure analysis of specific human chromosomes (Gusella et al., 1980; 1982; Law et al., 1982).

SOMATIC CELL GENETICS AND GENE MAPPING

123

F. REGIONAL ASSIGNMENT After a gene is assigned to a particular chromosome, the next step is to find its location within the chromosome. The exact location of a gene can be successively reduced to a small area on the chromosome as more regional mapping data are available. The term “SRO” (smallest region of overlap) has been adopted to indicate a consensus regional assignment of a gene by identifying the region that is present in all reported regional maps of that gene. The SRO will be further reduced as new and relevant data become available. One method of regional mapping is to use the gene dosage effect observed in patients or cells with deletions or duplications of specific segments of the chromosome. If the amount of the gene product correlates well with the number of gene copies, the gene can be assigned to that chromosome segment (FergusonSmith et a l . , 1976; Aitken and Ferguson-Smith, 1979; Rethore et al., 1976, 1977). Another method for regional mapping is to use cell hybrids established by fusing with human cells carrying defined translocations (Grzeschik et al., 1972; Ricciuti and Ruddle, 1973). The hybrids can be analyzed for the retention of the particular translocation chromosome but not the normal counterpart. Assay of the gene in this hybrid will reveal whether it is on the translocated piece of the chromosome, thus establishing regional assignment of that gene. If the translocation is reciprocal and each of the translocation chromosomes is retained in separate hybrids, the gene can be analyzed in both hybrids to confirm the presence of the gene in only one of the hybrids. The above methods rely largely on the availability of patients with relevant chromosomal abnormalities. A large collection of human cells with various chromosomal abnormalities is available in the Human Genetic Mutant Cell Repository, Institute for Medical Research, Camden, N.J. For a more systematic regional mapping, it appears desirable to use an empirical approach to generate more terminal deletions in cell hybrids with breakpoints in various positions of a specific human chromosome. A set of such deletion hybrids should be very useful in rapid regional mapping of genes assigned to that chromosome. Various chromosome-breaking agents such as X rays, y rays, BUdR + nearvisible light, UV, ethylmethanesufonate, ICR-191, and caffeine (Kao and Puck, 1969; Kao et al., 1977) have been used to induce chromosomal deletions. It should be pointed out that all these agents except caffeine produce both point mutations and chromosome breakage (Kao and Puck, 1969). Thus, caffeine may be used as an effective chromosome-breaking agent without causing significant mutations in the treated cells. Two procedures for introducing chromosome breakage have been used and their efficiencies compared. The first procedure involves treatment of human cells with chromosome-breaking agents followed by fusion with rodent mutant cells (Burgerhout et al., 1973, 1977; Goss and Harris, 1975, 1977a,b; Law and

124

FA-TEN KAO

Kao, 1978). Hybrids are selected and analyzed for deletions induced in various human chromosomes. In the second procedure, cell hybrids containing a single human chromosome are treated with chromosome-breakingagents and subclones are isolated and analyzed for deletions produced in that particular single human chromosome (Law and Kao, 1978). In these studies, the first method was shown to be more effective in producing more different kinds of deletions in a variety of human chromosomes, while the second method produced deletions in a specific human chromosome retained in the hybrids which contained no other human chromosomes, thus making cytogenetic analysis simpler and more reliable (Kao er al. 1977; Jones and Kao, 1978; Law and Kao, 1978, 1979). Construction of deletion hybrids using the second method has been accomplished with human chromosomes 11 (Jones and Kao, 1978) and 12 (Law and Kao, 1979, 1982). Selection for clones with deletions in a particular human chromosome can be greatly facilitated if the chromosome possesses a marker which can be used to select for the loss of a particular gene as a result of chromosome deletion. For chromosome 11, the species-specific cell surface antigen marker A, system has been successfully used for this purpose. Antisera prepared against human cells can kill only human cells in culture in the presence of complement, but have no effect on CHO cells (Oda and Puck, 1961; Puck et al., 1971). It was later found that the genetic determinants for the cell surface antigen complex A, reside on human chromosome 11 (Puck er al., 1971; Kao et al., 1976). Regional mapping has localized one of the genes, a,, to 1lpter-pl3 (Kao et al., 1977). Thus, any break occurring in the short arm of chromosome 11 between a, and the centromere will result in the loss of the a, marker and the subclones carrying such deletions can survive the antiserum treatment. Thus, AL-J, hybrid cells containing a single human chromosome 11 (Kao et al., 1976) were treated with various chromosome-breakingagents at low doses to avoid complex chromosomal translocations or interstitial deletions (Kao el al., 1977), followed by selection in the presence of anti-a, serum and complement. Surviving clones were isolated and found to possess various deletions with breakpoints between 1lp13 and the centromere (Kao er al, 1977; Jones and Kao, 1978). A set of such deletion hybrids has been established and used for a quick and systematic mapping of genes to various regions of chromosome 11, including LDHA (Jones and Kao, 1978), P-globin gene complex (Gusella er al., 1979), ACP2 (Jones and Kao, 1978), PBGD (Meisler et al., 1981), and cell surface antigens a3 (Kao er al., 1977) and a2 (Kao et al., 1977). In addition, 5 unique DNA sequences (Gusella er al., 1980) and 19 repetitive sequences (Gusella et al., 1982) have also been assigned to specific regions of chromosome 11 using these deletion hybrids. More cell surface antigen markers can be identified and mapped to various human chromosomes for use in isolating specific deletion

SOMATIC CELL GENETICS AND GENE MAPPING

125

hybrids. In addition, markers like HPRT, APRT, and TK, capable of both forward and reverse selection, are also useful for selecting deletion hybrids for chromosomes X, 16, and 17, respectively. For extending this approach to as many human chromosomes as possible, two procedures can be pursued: (1) Use of human cells with translocations between X and other human chromosomes to construct hybrids carrying the translocation chromosome including the HPRT gene from the X chromosome segment. The HPRT marker can then be used for selecting deletions in the arm to which the HPRT marker has been introduced. In the Human Genetic Mutant Cell Repository, translocations are available between X and more than 14 other human chromosomes. (2) Use of gene transfer techniques (Wigler et al., 1977, 1978, 1979) to introduce the HPRT gene into HPRT- human cells like Lesch-Nyhan cells or HT-1080 fibrosarcoma cells (Croce, 1976). The HPRT gene may be stably integrated into different human chromosomes in different human cells and then used as a selective marker for each specific human chromosome. The recently successful cloning of HPRT genes from mouse (Brennand et al., 1982) and human (Jolly et al., 1982) cells should greatly facilitate the transfection experiments. Another method for regional mapping without resorting to cytogenetic analysis was introduced by Goss and Harris (1975). It invokes a simple target theory which states that the greater the distance (or the target) between two syntenic genes, the greater is the probability for the two genes to be separated by a chromosome break event. The experimental procedure is essentially the same as the first method described above. Human cells were treated with various doses of y rays and then fused with the Chinese hamster HPRT- mutant. Independently occurring hybrids were isolated and various isozyme markers on the X chromosome and chromosome 1 were assayed (Goss and Harris, 1977a,b). The segregation data of these syntenic genes were analyzed by statistical methods following the probability estimates of the target theory. The linear order and the relative distances of these genes were determined for the human X chromosome (Goss and Hams, 1977a) and chromosome 1 (Goss and Hams, 1977b). This method was also extended to chromosome 12 (Law and Kao, 1978, 1979, 1982). When the regional maps constructed by the statistical analysis are compared with those prepared by cytogenetic analysis, the gene order is in agreement for the two maps, but the positions of the genes differ in some cases. It should be pointed out that the statistical regional map is based on chromosome breakage produced during interphase in which chromosomes are highly extended, whereas the cytogenetic regional map is constructed on metaphase chromosomes which are highly condensed. The exact correspondence between the states of chromosome condensation in relation to the physical gene maps remain to be elucidated. Because crossing over takes place when the homologous chromosomes are high-

126

FA-TEN KAO

ly extended, the statistical regional map appears to resemble more closely the linkage map constructed in animals and plants using the recombination frequency observed in the offspring. The statistical mapping method is particularly useful in determining the relative distance and gene order among very closely linked genes not resolvable by cytogentic techniques. Using the TK selective marker and the radiation-induced segregation frequency in the hybrids, Goss ( 1979) estimated the chromosomal interval between TK and GALK (galactokinase) to be about 0.04% of the human haploid genome, or 1.2 X lo6 bp, on the long arm of chromosome 17. By cytogenetic analysis, the syntenic genes TK and GALK were assigned to 17q2I q22, a region comprising about 1.2% of the haploid genome, or 3.6 X 107 bp, which is 30 times greater than the interval established by the statistical mapping technique.

V. Combined Use of Somatic Cell Genetics and Recombinant DNA Technology in Gene Mapping The remarkable advances in recombinant DNA technology during the last decade have revolutionized many areas of biology, particularly molecular genetics. A large number of genes have been cloned for detailed structural and functional studies. DNA sequences with regulatory functions, such as promoter and enhancer sequences, have been identified and characterized (Breathnach and Chambon, 1981). Of particular importance has been the finding of intervening sequences or introns within the genes (Jeffreys and Flavell, 1977; Leder et al., 1977), the recombination and elimination of DNA sequences of immunoglobulin genes during hematopoietic cell differentiation (Hozumi and Tonegawa, 1976; Bernard et al, 1978), the identification in the human genome of repetitive sequences interspersed with unique sequences (Schmid and Dehninger, 1975; Houck et af, 1979; Jelinek et af, 1980; Schmid and Jelinek, 1982), and the presence of multigene families (Heintz et al., 1981; Goeddel et af, 1981; Moore et al., 1982; Beaudet et al., 1982), pseudogenes (Bentley and Rabbitts, 1980; Denison et af. 1981; Little, 1982; Wilde et al., 1982), and oncogenes (Bishop, 1981; Weinberg, 1982; Cooper, 1982). These discoveries have generated profound insights into the understanding of the structure, organization, and function of genes and genomes in the eukaryotes. As a result, new concepts are emerging in elucidating regulatory mechanisms underlying development and differentiation in eukaryotic organisms including man. The impact of recombinant DNA on the progress of gene mapping is also overwhelming, particularly in combination with the powerful cellular approaches

SOMATIC CELL GENETICS AND GENE MAPPING

127

of somatic cell genetics. In the following are summarized important developments in gene mapping using these combined approaches.

A. USE OF CLONEDGENESAND CELLHYBRIDS FOR GENEMAPPING The first successful attempt at using molecular hybridization for gene mapping was carried out by Deisseroth et al. (1977, 1978) using cDNA probes of OL and p globin genes and Cot analysis in nucleic acid hybridization in solution with the genomic DNA from various cell hybrids. Specific annealing between the labeled cDNA and the genomic DNA sequences from human chromosomes retained in the hybrids provided evidence for the assignment of the OL and p globin genes to human chromosomes 16 (Deisseroth et al., 1977) and 11 (Deisseroth er al., 1978), respectively. In these experiments, the probe used in hybridization should not cross-hybridize with DNA of the other species in the hybrid genome. Later, filter hybridization using Southern blots (Southern, 1975) greatly improved the specificity and sensitivity of detecting the presence of specific gene sequences in the hybrid cell genomic DNA. Cross-hybridizing gene probes can also be used in these experiments. The method involves use of restriction enzymes to digest DNA from human and rodent parents and cell hybrids, agarose gel electrophoretic separation of digested DNA, transfer of DNA from gel to nitrocellulose filter, hybridization of the filter with the labeled gene probe, and autoradiographic detection of the duplex hybridization bands. In these procedures, the first step is to select a restriction enzyme that can cleave the genomic sequences containing the gene into fragments differing in size between human and rodent species involved. Because the DNA sequences of most genes are conserved to varying degrees among animal species, especially mammals, it is necessary to distinguish human sequences from rodent sequences by their fragment size patterns cleaved with a speGific restriction enzyme. The selected enzyme will be used to digest cell hybrid DNA to prepare blots for hybridization. Positive hybrids are those which contain not only rodent but also human gene sequences, while negative hybrids contain only rodent sequences. An example of such analysis is shown in Fig. 3. The hybridization results will be matched with the human chromosome content in the cell hybrids. The human chromosome with the highest concordance frequency will define the chromosomal assignment of the gene probe used. This procedure is essentially the same as that used for assigning an isozyme marker or other gene products using cell hybrid (Fig. 2). It should be pointed out that this method will map structural gene sequences in the human genome regardless of whether or not the gene is expressed in the cell hybrids. Thus, it is especially useful in mapping those genes coding for tissue-

128

FA-TEN KAO

Hindlll

EcoRl

FIG. 3. Autoradiogram demonstrating molecular hybridization between the human albumin gene probe and the total genomic DNA from the human (H) and the CHO (Ch) parental cells and their cell hybrids. In Hind111 digested DNA, a strong 3.5 kb band was present in the human DNA (lane I), while a 3.5 kb band was present in the CHO DNA (lane 2). In the positive hybrid ( + ) containing human chromosome 4, both 6.8 and 3.5 kb bands were present (lane 3), but the negative hybrid (-) containing no human chromosome 4 exhibited only the CHO 3.5 kb band (lane 4). Similar results were found in EcoRI digested DNA except that the fragment sizes were different: the human band being 12.5 kb (lane 5 ) and the CHO band 4.3 kb (lane 6).

specific functions which are usually not expressed in most hybrids using human fibroblasts or lymphocytes as the parent (Davidson, 1974). It should also be pointed out that the gene probes used in this method can be derived from any species provided that they possess enough homology to cross-hybridize with human sequences and yield unqiue digested fragments in the hybrids. Examples of structural genes mapped to various human chromosomes using cloned gene probes and cell hybrids are listed in Table I. This list is increasing rapidly as more and more gene probes become available. It is interesting to add that the recent application of gene transfer techniques has led to the identification and isolation of human genomic gene sequences homologous to a variety of viral oncogenes (Bishop, 1981; Weinberg, 1982; Cooper, 1982). Using gene mapping procedures similar to those described above, these human homologous genes have been assigned to specific human chromosomes (Table 11). From these assignments, it is clear that these genes are not clustered in the human genome, but scattered over different chromosomes.

129

SOMATIC CELL GENETICS AND GENE MAPPING

Gene mapping has also provided important leads in the molecular etiology of certain cancers. Burkitt lymphoma involves consistent translocations between chromosomes 8 and 14 (Klein, 1981). It has recently been shown that the piece translocated from 14 to 8 carries part of the DNA sequences of the variable region V, of the heavy chain of immunoglobulin gene which was previously mapped to chromosome 14 (Croce et al., 1979), while the remaining portion of V, and the constant region C, of the gene complex are still in the nontranslocated chromosome 14 (Dalla-Favera et af., 1982a). Even more interesting is the finding that the piece of chromosome 8 translocated to chromosome 14 carries the human gene c-myc homologous to the avian myelocytomatosis virus oncogene v-myc (Dalla-Favera et al., 1982b). Thus, the translocation of the cellular c-myc gene from chromosome 8 to the V, region on chromosome 14 may result TABLE I HUMAN CHROMOSOMAL ASSIGNMENT OF STRUCTURAL GENESUSING GENEPROBESI N MOLECULAR HYBRIDIZATION W I TH SOMATIC CELLHYBRIDS Human structural gene

Human chromosome assignment

a Globin

16

p Globin

I1

Growth hormone, growth hormone-like gene Chorionic somatomammotropin Insulin Interferon a

17

Interferon p Interferon y Prolactin Proopiocortin Collagen, Pro 2(I) a ,-Antitrypsin Albumin C. immunoglobulin light chain CA immunoglobin light chain Third component of complement (C3) Immunoglobulin heavy chain Chymotrypsinogen B

17

I1 9 9 12 6 2 7 14

4 2 22 19 14

16

Reference Deisseroth er a / . ( 1977) Deisseroth er a / . ( 1978); Scott er a / . (1979); Gusella ei a / . (1979); Lebo ei a / . (1979) Owerbach er a / . (1980a); George et a / . ( 198I ) Owerbach er a / . (1980a); George ei a/. (1981) Owerbach er a / . (1980b. 1981a) Owerback er a / . (1981b); Slate er a/. (1982) Owerbach e r a / . (1981b) Naylor er a / . (1982) Owerbach er a / . (1981~) Owerbach et a / . (1981d) Junien er a / . (1982) Lai er a / . ( 1982) Kao el a/. (1982) McBride er a / . (1982a) McBride et a / . ( 1982a) Whitehead er a / . (1982) Hobart er a/. ( I98 I ) Sakaguchi er a / . (1981)

130

FA-TEN KAO

TABLE I1 CHROMOSOMAL ASSIGNMENT OF HUMANHOMOLOGOUS GENESTO VIRAL ONCOGENES HYBRIDIZATION USING SOMATIC CELL HVBRlDS

Human homologous gene

Viral oncogene

c-sis c-fes

Simian sarcoma virus Feline sarcoma virus

c-myb c-mos c-ab/ c-src c-Ki-ras c-Ha-ras

Avian myeloblastosis virus Moloney murine sarcoma virus Abelson murine leukemia virus Rous sarcoma virus Kirsten murine sarcoma virus Harvey murine sarcoma virus

c-myc

Avian myelocytomatosis virus

Human chromosome assignment 22 15

6

8 9

20 12

II 8

BY

MOLECULAR

Reference Swan et a/. (1982) Dalla-Favera er al. (1982a); Heisterkamp et a / . (1982) Dalla-Favera ef a/. ( 1982a) Prakash er a/. (1982) Heisterkamp et al. ( I 982) Sakaguchi et a / . ( 1982) Sakaguchi et a / . ( I 982) De Martinville ef al. (1983); McBride el af. (1982b) Sakaguchi et a / . ( I 982); Dalla-Favera ef a / . (1982b)

in the activation of the c-myc gene and lead to the development of Burkitt lymphoma. In addition, other non-African forms of Burkitt lymphoma were found to involve translocations between 8 and 2, or 8 and 22 (Klein, 1981; Rowley, 1982). Gene mapping has assigned the kappa light chain of immunoglobulin gene to chromosome 2 (McBride ef af., 1982; Malcolm er af., 1982) and the lambda light chain to chromosome 22 (McBride er af., 1982; Erikson et af., 1981). Thus, it is possible that these forms of Burkitt lymphoma also involve translocation of the c-myc gene from chromosome 8 to the chromosomal sites where the immunoglobulin light chain gene complexes reside.

B. ASSIGNMENT OF RANDOM DNA SEGMENTS TO SPECIFIC HLJMAK CHROMOSOMES Maniatis and co-workers constructed the first human genomic DNA library using human fetal liver DNA (Lawn et af., 1978). This library consists of 7 X lo5 recombinant A phage each containing a human DNA segment of about 15-20 kilobases (kb) in length. Potentially, each segment can be used as a genetic marker for a specific region of the human genome since each segment is derived from a different site of the genome. Thus, a large number of DNA segments can be isolated, characterized, and assigned to specific chromosomes for various genetic analysis, particularly in mapping the human genome.

SOMATIC CELL GENETICS AND GENE MAPPING

131

Due to the abundance of interspersed repetitive sequences in the human genome (Schmid and Jelinek, 1982; Singer, 1982), the majority of the human inserts in the recombinant A phage possess both unique and repetitive sequences. In order to use unique sequences as genetic markers, it is necessary to remove the repetitive sequence portion from the insert and use only the unique sequence. Alternatively, recombinant phage containing only unique sequences can be identified and used directly for chromosomal assignment. It has been shown that about 1% of the recombinant phage in the human genomic library contains only unique sequences (Kao et al., 1982). These phage can be easily identified by their failure in colony hybridization (Benton and Davis, 1977) using human repetitive sequences as the probe. These human unique sequence segments can be assigned to specific human chromosomes using cell hybrids, as described above for assigning the cloned genes (Kao er al., 1982; De Martinville er al., 1982). This procedure is simple enough to provide large numbers of human unique DNA segments with known chromosomal location. When appropriate deletion hybrids are available, these DNA segments can also be localized regionally on the chromosome. C.

CLONING AND MAPPING OF DNA SEGMENTS ISOLATED HUMAN CHROMOSOMES

FROM SPECIFIC

In addition to isolating random DNA segments from human genomic library and subsequently assigning them to specific chromosomes, DNA segments can also be isolated from specific human chromosomes by molecular cloning. One method for cloning such segments is to use cell hybrids containing only a single human chromosome as the DNA source for cloning. Subsequently, phage containing human DNA segments can be distinguished from those containing rodent DNA by their differential hybridization to the species-specific middle repetitive DNA sequences (Gusella er af., 1980). Finally, the repetitive sequence portion in the human insert can be removed and the unique sequence can be used as a DNA marker for that human chromosome. Regional assignment for the DNA segment can be made using specific deletion hybrids if available. Another method for preparing DNA segments from a specific human chromosome is to use fractionated individual human chromosomes as the DNA source. Human metaphase chromosomes can be separated into as many as 20 fractions using a fluorescent activated cell sorter FACS-I1 (Young et al., 1981). DNA libraries have been constructed from chromosomes X (Davies er af., 1981), 21, and 22 (Krumlauf et al., 1982) using DNA from this source. The fractionated peaks containing individual chromosomes are not totally pure. For example, the fraction of the X chromosome contains about 10% of chromosomes 7 and 8 (Davies et al., 1981). In addition, chromosomes 9, 10, 11, and 12 cannot be fractionated into a single peak (Young er al., 1981), and some chromosomes

132

FA-TEN KAO

cannot always be separated completely, including chromosomes 1-2, 14- 15, 17-18. Despite these limitations, however, this method, when combined with single human chromosome hybrids, offers great promise for providing DNA libraries for most, if not all, human chromosomes. D. USE OF REPETITIVE SEQUENCES IN FINESTRUCTURE MAPPINGOF INDIVIDUAL CHROMOSOMES

The human genome consists of more than 30% of repetitive sequences interspersed with unique sequences throughout the genome (Schmid and Jelinek, 1982). Most of these repetitive sequences are families of highly repetitive sequences like the Alu family (Jelinek er al., 1980), with loJ or more copies in the genome. When an individual member of such sequences is used as a probe to hybridize to blots containing digested total human DNA fractionated on an agarose gel, a smear of hybridization bands results. Families of middle repetitive sequences possessing only several thousand copies or less are also found in the human genome (Adams et al., 1980; Schmeckpeper et al., 1981). Although these repetitive sequence probes still form a smear in the total human DNA blots, they have been found to exhibit distinct bands when hybridized to DNA from a single human chromosome (Law el al., 1982; Gusella et al., 1982). Such findings prompted use of species-specific, middle repetitive sequences as markers for multiple sites in the genome. Two repetitive sequence probes have been developed and used as multiple site markers on chromosomes 12 (Law et al., 1982) and 1I (Gusella et al., 1982), respectively. Using a cell hybrid 12A containing a single human chromosome 12 (Law and Kao, 1978), Law et al. (1982) cloned human DNA segments from chromosome 12. One repetitive sequence of 2.2 kb was isolated from a recornbinant phage containing both repetitive and unique human sequences. This 2.2 kb sequence was shown to have a few thousand copies in the human genome. When the 2.2 kb probe was hybridized to the DNA from 12A cell hybrid, multiple but distinct bands were formed. The 2.2 kb segment was further cleaved with PvuII to three subfragments of 1.2, 0.6, and 0.4 kb in length. When the subfragment probes were hybridized to 12A DNA, simpler band patterns resulted. Using these probes and various deletion hybrids of human chromosome 12, five bands were assigned to a specific region of chromosome 12 (Law et al.. 1982). The second repetitive sequence probe is a DNA segment A36Fc derived from the p globin gene complex on human chromosome 11 (Duncan et al., 1979). Gusella et al. (1982) used this probe to hybridize to the cell hybrid J1 containing only human chromosome 11 (Kao et al., 1976), and found that 24 bands were discernible. When the probe was hybridized to a series of deletion hybrids of chromosome 11 (Jones and Kao, 1978), 19 bands could be assigned to 11 regions of chromosome 11 (Gusella er af., 1982).

SOMATIC CELL GENETICS AND GENE MAPPING

133

These studies demonstrate the use of repetitive sequence probes to identify multiple sites of the chromosome in the human genome. The power of this approach can be increased by using additional restriction enzymes to generate different patterns of bands of the same chromosome. Furthermore, additional repetitive sequence probes can also be isolated and used for these purposes. Combining various repetitive sequence probes and different restriction enzymes, a reasonably detailed fine structure map of a chromosome can be established. These fine structure maps of specific chromosomes should be useful in detecting DNA sequence alterations in diseases with consistent chromosome abnormalities. Finally, cell hybrids containing a specific segment of a chromosome can be constructed. To expand the resolution of the fine structure map for this chromosomal segment, repetitive sequence probes with even higher copy numbers can be used to generate a more detailed map. Particular sites of interest on this segment can be isolated for sequence analysis. This approach appears to be feasible in achieving reasonably high resolution in fine structure mapping of specific segments of many human chromosomes. E. FINESTRUCTURE MAPPING OF SMALL CHROMOSOMAL SEGMENTS Cytogenetic analysis of human chromosomes has attained its current limits with high-resolution banding techniques (Yunis, 1976). About lo00 individual bands in the human genome can now be identified. Since the human genome of 3 X lo9 bp is equivalent to a genetic distance of about 3300 cM as measured by recombination frequency, each cM represents about lo6 bp. Thus, each cytogenetic band on the average consists of approximately 3 X lo6 bp or 3 cM. On the other end of the scale of gene mapping, the rapid DNA sequencing techniques (Sanger et al., 1977; Maxam and Gilbert, 1977) combined with chromosome walking may extend sequence analysis over a chromosomal region of 50-100 kb, or 0.05-0.1 cM. For regions larger than 100 kb, the present techniques are too laborious for routine uses. Thus, a significant gap in gene mapping exists between the cytogenetic resolution of the chromosome bands of 3 X lo6 bp or greater, and the molecular resolution of chromosomal regions of lo5 bp or less. New approaches and methodologies will have to be developed to fill this gap. One method to bridge the cytogenetic mapping and the sequence mapping involves construction of recombinant phage or plasmid libraries containing the entire DNA sequence of this intermediate size range of the chromosome ( lo5 to 3 X lo6 bp). For a chromosomal region of 3 X lo6 bp, it will require about 150 recombinant A phage with an average size of 20 kb human insert to represent the entire region. A DNA segment of 20 kb can be sequenced if necessary. Perhaps more difficult is the isolation of a particular chromosomal region for

134

FA-TEN KAO

constructing such a DNA library. Somatic cell genetic techniques can be used to introduce human chromosomal segments of this size range into recipient cells for DNA library construction. In these procedures, a selective marker is needed in the chromosomal region to be transferred. It has been shown that metaphase chromosome-mediated gene transfer (McBride and Ozer, 1973) can introduce human chromosome segments of varying sizes into recipient cells (Klobutcher and Ruddle, 1979). Although some segments are too small to be detectable cytogenetically, others are sufficiently large to be detected by Giemsa-1 1 differential staining technique. Some clones can be isolated containing only this piece of human material. Thus, human segments in these clones can be used for library construction using the method described above. If the chromosomal region contains no selective marker, another method can be used. This involves first construction of cell hybrids containing a single human chromosome of interest. Subclones with various terminal deletions of that chromosome can be prepared and a DNA library can be constructed from the subclone exhibiting the chromosomal deletion closest to, but still retaining, the region of interst, as demonstrated in the construction of a DNA library from the portion of human chromosome 1 1 including 1 lpter-q13 (Gusella et al., 1980). Subsequently, DNA segments from recombinant phage containing human DNA can be regionally mapped and those localized to the particular region of interest can be selected. These methods provide a feasible approach to isolating recombinant human DNA segments in a particular chromosomal region. If a disease is associated with a consistent chromosomal deletion, it should be possible to construct a fine structure map of that deletion region and to identify genes or regulatory sequences that may be responsible for the etiology of the disease.

F. In Situ HYBRIDIZATION IN GENEMAPPING The original development of techniques for in situ hybridization of DNA to metaphase chromosomes fixed OR microscopic slides (Pardue and Gall, 1970) has been extended from detecting locations of multiple gene copies arranged in tandem to detecting singly located unique genes (Harper et al., 1981;Malcolm et al., 1981). Although further technical improvements remain to be developed to make this method a routine laboratory procedure, this approach appears to offer great opportunity for a quick chromosomal and regional assignment of genes available in purified forms. Examples of genes mapped by this method are listed in Table 111. Due to the spread of the grains detected in autoradiographicpreparations, this method cannot provide mapping to a very small region on the chromosome, nor can it determine the order of closely linked genes with resolution higher than the current cytogenetic techniques. However, this approach does provide a general

135

SOMATIC CELL GENETICS AND GENE MAPPING TABLE 111 HUMAN STRUCTURAL GENEASSIGNMENTBY in Siru HYBRIDIZATION METHODUSING MOLECULAR PROBES

Human gene

Human chromosome assignment

Histone

7

a Globin

16

Insulin p Globin Placental lactogen-growth hormone gene cluster V, immunoglobin light chain Immunoglobulin heavy chain Interferon a Interferon p Interferon y

I1 11

17 2 14

9 9 12

Reference

Yu et a/. (1978); Chandler et a / . (1979) Gerhard et a / . ( I98 I ); Barton et al. (1982) Harper et a / . (1981) Malcolm et al. (1981) Harper et al. (1982) Malcolm er al. (1982) Kirsch et al. (1982) Trent et al. (1982) Trent et al. (1982) Trent et al. (1982)

method particularly useful for regional mapping without relying on the availability of specific chromosomal deletions or translocations. Improvements in resolution of the gene location on the chromosome have been made by using biotin-conjugated, instead of 3H-labeled, nucleotide probes followed by fluorescent or histochemical staining (Langer et d.,1981; LangerSafer et al., 1982). Sharp bands detected on the chromosomes revealing hybridization of the probe to specific genomic sequences have greatly reduced the area where the gene resides compared to that resolved by the radioactivity technique.

VI. Prospect of Constructing a Complete Human Gene Map The ultimate goal of gene mapping is to construct a complete map of all the genes that have been identified, their locations on the chromosomes and distances from the neighboring genes, the nucleotide sequences of these genes, the coding and intervening sequences as well as the regulatory and flanking sequences associated with the gene. Ideally, such a gene map can be constructed in terms of the number of base pairs as a measure of distance between two genes and between a gene and a cytogenetic landmark of the chromosome, such as centromere, telomere, secondary construction, etc. Such a detailed gene map is important in unraveling aspects of gene regulation in relation to its structure and organization. Furthermore, a complete gene map is essential in making reliable predictions in prenatal diagnosis, for the exact genes or DNA sequences involved in a chro-

136

FA-TEN KAO

mosomal deletion or translocation can be identified and their consequences evaluated. From a theoretical point of view, the complete nucleotide sequence of the human genome may be determined. However, such an endeavor will not be practical even for a fraction of the genome. For example, using the most efficient sequencing method now available, it will take more than 2000 man-years to sequence only the X chromosome, which is about 6% of the human genome. Moreover, even with the complete sequence of the human genome determined, specific genes will still have to be identified from the sequence data in order to make use of the sequence information. Sequencing of selected regions of the human genome is more practical with the present sequencing technology. Complete sequencing of the entire genome may have to await further breakthrough in the technology. In 1980, Botstein et af. proposed a scheme for constructing a complete linkage map of the human genome. In this scheme, DNA segment markers exhibiting restrictionfragment length polymorphisms(RFLPs) will be used. Such polymorphisms can be detected by hybridizing a DNA segment probe to blots containing genomic DNA from different individuals digested with various restriction enzymes. DNA sequence alterations involving changes in restriction enzyme recognition sites or deletions or additions of DNA within the restriction fragments can be detected by the change in hybridization band patterns. If a probe is found to exhibit different band patterns after testing in several individuals using several restriction enzymes, it is considered reasonably polymorphic for use in further linkage analysis. This probe will then be tested in pedigrees to confirm that the polymorphism is inherited in families in a Mendelian fashion. The probe can be used as a genetic marker to test linkage relationships with informative pedigrees in which human traits or diseases are segregating. If the polymorphic probe is found to segregate together with a specific human trait or a disease condition, linkage is indicated and further analysis using more pedigrees will be carried out to establish the closeness of the linkage. In the scheme of Botstein et af. (1980), it has been estimated that if a distance of 20 cM between two marker genes is optimal to detect linkage, a minimum of 150-200 DNA fragments exhibiting RFLPs and evenly spaced at a distance of 20 cM (or 2 X lo7 bp) will be required to establish linkage with any human gene (Botstein et af., 1980). However, because the DNA markers will fall randomly along the genome, considerably more markers will have to be isolated to cover the entire genome. Lange and Boehnke (1982) estimated that 1528 markers are required to cover 22 autosomes for detecting linkage with any loci at a minimum distance of 10 cM. If the marker isolation can proceed on individual chromosome basis, as demonstrated for chromosomes 11 (Gusella et al., 1980), 21 and 22 (Krumlauf et af., 1982), and the X (Davies et af., 1981), only 766 markers are

SOMATIC CELL GENETICS AND GENE MAPPING

137

required to cover 22 autosomes (Lange and Boehnke, 1982; Skolnick and White, 1982). This is because fewer redundant markers will be isolated from individual chromosome libraries than if the complete genome library is used. Further reduction in the number of markers can be achieved if individual libraries of chromosomal segments are used. This strategy is particularly important when large portions of the genome are well marked and only certain small regions remain to be filled, or when additional markers are desired to test linkage with a particular gene whose regional assignment is known. Somatic cell genetic techniques that can be used to dissect and isolate such small chromosomal regions for library construction and marker isolation will be particularly useful (Gusella et al., 1980; Klobutcher and Ruddle, 1979). Although difficulties in isolating sufficient numbers of DNA markers and in establishing linkage relationships in the pedigree analysis are anticipated, such a scheme provides for the first time a feasible approach for an eventual construction of a complete human linkage map. Kan and Dozy (1978) reported the first case of using the restriction fragment length polymorphism in the HpaI site about 5 kb from the p globin gene for prenatal diagnosis of sickle cell anemia, a mutant form of the p globin gene. However, if the distance between a mutant gene and the restriction polymorphic site is greater than 1 cM, it will not be very useful for prenatal diagnosis due to increased frequencies of crossing over between them. Perhaps the most desirable use of RFLPs is that reported by Geever ef al. (1981). A restriction enzyme recognition site DdeI was altered as a result of the nucleotide change inside the sickle cell anemia gene. In addition, another restriction site MstII was found to generate different fragment sizes in normal and sickle cell individuals (Wilson et al., 1982). Thus, the affected individual can be detected directly by the presence of the altered restriction fragment pattern. This direct correlation between a mutant gene and the altered restriction fragment pattern is particularly useful in prenatal diagnosis because no further linkage information is needed from the parents, and because no crossing over will occur to separate the mutant allele and the restriction site. In order to uncover more useful probes like this, it seems necessary to isolate and characterize the mutant genes for identifying possible changes in restriction sites or in flanking sequences, The first linkage relationship established between a random polymorphic DNA marker and an inherited disease was reported by Murray et al. (1982). A singlecopy DNA segment of 6.1 kb isolated from a human X chromosome library was shown to be polymorphic at a TaqI site and was regionally mapped to the middle part of the short arm of the X chromosome using cell hybrids containing various X chromosome deletions. When this probe was hybridized to families with Duchenne muscular dystrophy, linkage was indicated with a lod score + 1.766 at a recombination fraction of 0.10 (10 cM). Analysis of additional informative families is needed to confirm the linkage. This finding demonstrates the potential

I38

FA-TEN KAO

usefulness of DNA markers for establishing linkage without the elucidation of the biochemical nature of the disease.

VII. Gene Mapping in Other Species Using Somatic Cell Genetics Using well-designed matings and examining segregation in the offspring, linkage analysis has been carried out extremely successfully in laboratory animals such as Drosophilu and the mouse. Linkage studies in other mammalian species, however, has not progressed nearly as far as in the mouse. Use of somatic cell genetics in gene mapping has changed the outlook not only in human but also in other mammals. The linkage data accumulated in the mouse had established various linkage groups but made very few chromosomal assignments. Furthermore, the genes that have been extensively studied for linkage analysis in the mouse have been concerned mainly with morphological traits, and very few cellular and biochemical markers have been studied. Recently, the use of somatic cell hybrids for mapping genes in other species has begun when the loss of rodent chromosomes was also found in certain rodent-human cell hybrids (Jami et ul., 1971; Minna and Coon, 1974; Croce, 1976). The main difference between these and previous fusions was that a permanent human cell line was used instead of diploid human cells, and the rodent parent was normal cell cultures instead of permanent cell lines. This finding also demonstrates that it is the cell type, not the species. that determines the direction of chromosome loss in such hybrids. Thus, the wellestablished procedure for mapping human genes using cell hybrids can be applied equally well to mapping of genes in other species. Various interspecific cell hybrids have been established using permanent cell lines from mouse or Chinese hamster, and normal, diploid primary cell cultures from a wide vareity of mammalian species. Assignment of syntenic groups in these species has progressed at a rapid rate (Human Gene Mapping 6, 1981). Cytogenetic and isozyme techniques have been used for characterizing the chromosomes retained in the various hybrids. Comparatve gene mapping in different species revealed striking conservation of syntenic genes even among distantly related species. Significant similarities in gene maps were found between human and other primates, a fact already reflected by the homology in chromosome numbers and banding patterns. For example, chimpanzee has one more chromosome than human. Cytogenetic evidence shows that two of the chimpanzee’s chromosomes are actually equivalent to the two arms of human chromosome 2. Gene mapping has now furnished the equivalence in the gene content in these chromosomes. The human syntenic genes are frequently found to have homologous genes to be syntenic even in nonprimate species. For example, four genes on human

SOMATIC CELL GENETICS AND GENE MAPPING

139

chromosme 12, namely TPI, GAPD, LDHB, PEPB, are also syntenic in cat (O’Brien and Nash, 1982; Nash and O’Brien, 1982). Since not many data are available on regional mapping on genes in nonprimate species, the equivalence in gene order and distance cannot be assessed. On the other hand, several human syntenic genes were found to be asyntenic in other species, for example, LDHBPEPB are syntenic in human but asyntenic in pig, LDHA-ACP2 asyntenic in cat, MPI-PK asyntenic in sheep. The X chromosome has the highest degree of conservation and homology among all chromosomes in mammalian species. The mammals have about 6% of their haploid genome constituting the X chromosome. Although the same Xlinked genes have been found in most mammalian species, the gene order may not always be the same. For example, the gene order on the human X chromosome is centromere-GLA-HPRT, but in reverse order in the mouse X chromosome (Francke and Taggart, 1980). Comparative gene mapping has established extensive conservation of syntenic genes in diverse species and has elucidated important aspects of karyotypic evolution in mammals (Nash and O’Brien, 1982).

VIII. Conclusions Rapid progress in human gene mapping has achieved chromosomal and regional assignments of more than 400 genes. The combined use of somatic cell genetics and recombinant DNA technology has expanded the scope of mapping particularly for structural genes, and has also extended gene mapping to the fine structure level. Using restriction fragment length polymorphisms of specific DNA markers, a conceptual framework has been established which promises an eventual construction of a complete human linkage map. Such a complete gene map should aid in prenatal diagnosis in making more reliable predictions. The linkage relationships between a polymorphic DNA marker and an inherited disease can be established before the biochemical mechanism of the disease is known. More and more genes will be cloned, characterized, and sequenced. Fine structure analysis of the gene and the flanking sequences has increased our understanding of the regulatory processes underlying gene expression during development and differentiation. Once the regulatory mechanisms in normal individuals are delineated, insight will be gained into the developmental abnormalities in which malfunctioning in these processes are likely involved, and hopefully corrective measures can be devised for therapeutic purposes. Important progress has already been made in correlating gene location and its function in the possible etiology of certain forms of cancer (Dalla-Favera et af., 1982a,b). Eventually, manipulations in turning on and off of gene activities may offer

140

FA-TEN KAO

promises in molecular therapy of diseases caused by genetic disorders in somatic and germ cells. ,

1

ACKNOWLEDGMENTS This is contribution number 424 of the Eleanor Roosevelt Institute for ,Cancer Research. The work from the author’s laboratory was supported by NIH Grants GM26631 and HW2080. I wish to thank Drs. David Patterson and Martha Liao Law for critical reading of the manuscript.

REFERENCES Adams, I. W., Kaufman, R. E., Kretschmer, P. J., Harrison, M., and Nienhuis, A. W. (1980). Nucleic Acids Res. 8, 61 13-6128. Aitken, D. A., and Ferguson-Smith, M. A. (1979). Cyrogener. Cell Gener. 25, 131. Alhadeff, B’., Velivasakis, M., and Siniscalco, M. (1977). Cytogener. Cell Genet. 19, 236-239. Baker, R. M. (1976). In “Biogenesis and Turnover of Membrane Macromolecules” (J. S. Cook, ed.), pp. 93-103. Raven, New York. Barski, G., Sorieul, S., and Cornefert, F. (1960). C. R. Acad. Sci. 251, 1825-1827. Barton, P., Malcolm, S., Murphy, C., and Ferguson-Smith, M. A. (1982). J. Mol. B i d . 156, 269-278. Beaudet, A. L., Su, T. S., O’Brien, W. E., D’Eustachio, P., Barker, P. E., and Ruddle, F. H. (1982). Cell 30, 287-293. Bentley, D. L., and Rabbitts, T. H. (1980). Narure (London) 288, 730-733. Bpnton, W., and Davis, R. (1977). Science 196, 180-182. Bernard, 0..Hozumi. N., and Tonegawa, S. (1978). Cell 15, 1133-1144. Bishop, J. M. (1981). Cell 23, 5-6. Bobrow, M., Madan, K., and Pearson, P. L. (1972). Nature (London) New Biol. 238, 122-124. Botstein, D., White, R., Skolnick, M., and Davis, R. (1980). Am. J. Hum. Gener. 32, 314-331. Breathnach. R., and Chambon, P. (1981). Annu. Rev. Biochem. 50, 349-383. Brennand, J., Chinault, A. C., Konecki, D. S.,Melton, D. W., and Caskey, C. T. (1982). Proc. Narl. Acad. Sci. U.S.A. 79, 1950-1954. Burgerhout, W. G., Van Someren. H . , and Bootsma, D. (1973). Humangenerik 20, 159-162. Burgerhout, W. G., Leupe-De Smit, S., and Jongsma, A. P. M. (1977). Cyrogener. Cell Genet. 18, 267-283. Caspersson, T., Zech, L., and Johansson, C. (1970). Exp. Cell Res. 60,315-319. Chandler, M. E., Kedes, L. H., Cohn, R. H., and Yunis, J. J. (1979). Science 205, 908-910. Chasin, L. A. (1974). Cell 2, 37-41. Chu, E. H. Y., and Malling, H. V. (1968). Proc. Narl. Acad. Sci. U.S.A. 61, 1306-1312. Conneally, P. M., and Rivas, M. L. (1980). Adv. Hum. Genet. 10, 209~266. Cooper, G. M. (1982). Science 218, 801-806. Cox, D. R., Francke, U., and Epstein, C. J. (1981). Am. J. Hum Genet. 33, 495-512. Croce, C. M. (1976). Proc. Natl. Acad. Sci. U.S.A. 73, 3248-3252. Croce, C. M., Shander, M., Martinis, J., Cicurel, L., D’Ancona, G. G., Dolby, T. W., and Koprowski, H. (1979). Proc. Narl. Acad. Sci. U.S.A. 76, 3416-3419. Dalla-Favera, R., Franchini, G., Martinotti, S., Wong-Staal, F., Gallo, R. C., and Croce, C. M. (1982a). Proc. Narl. Acad. Sci. U.S.A. 79, 4714-4717.

SOMATIC CELL GENETICS AND GENE MAPPING

141

Dalla-Favera, R., Bregni, M., Erikson, J., Patterson, D., Gallo, R. C., and Croce. C. M. (1982b). Proc. Narl. Acad. Sci. U.S.A. 79, 7824-7827. Davidson, R. L. (1974). Annu. Rev. Genet. 8, 195-218. Davidson, R. L.. and Ephrussi. B. (1970). Exp. Cell Res. 61, 222-226. Davies, K. E., Young, B. D., Elks. R. G.. Hill, M. E., and Williamson, R. (1981). Narure (London) 293, 374-376. Deisseroth, A,, Nienhuis, A , , Turner, P., Velez, R.. Anderson. W. F., Ruddle, F.. Lawrence, J., Creagen, R., and Kucherlapati, R. (1977). Cell 12, 205-218. Deisseroth, A., Nienhuis, A., Lawrence, J., Giles, R.. Turner, P., and Ruddle, F. (1978). Proc. Narl. Acad. Sci. U.S.A. 75, 1456-1460. De Martinville, B., Wyman, A. R.. White, R., and Francke, U. (1982). Am. J . Hum. Genet. 34, 2 16-226. De Martinville, B., Fiacalone, J., Shih, C., Weinberg, R. A,, and Francke, U. (1983). Science 219, 498-501. Denison, R. R., Van Arsdell, S. W., Bernstein, L. B., and Meiner, A. M. (1981). Proc. Narl. Acad. Sci. U.S.A. 78, 810-814. Donahue, R. P., Bias, W. B.. Renwick, J. H., and McKusick, V. A. (1968). Proc. Narl. Acad. Sci. U.S.A. 61, 949-955. Duncan, C., Biro, P. A., Choudary. P. V.. Elder, J. T . , Wang. R. R. C., Forget, B. G., De Riel, J. K., and Weissman, S. M. (1979). Proc. Narl. Acad. Sci. U.S.A. 76, 5095-5099. Ephrussi, B. (1972). “Hybridization of Somatic Cells.” Princeton Univ. Press, Princeton, New Jersey. Erikson, J., Martinis, J., and Croce, C. M. (1981). Narure (London) 294, 173-175. Ferguson-Smith, M. A . , Aitken, D. A.. Turleau, C., and De Grouchy, J. (1976). Hum. Genet. 34, 35-43. Francke, U., and Taggart, R. T. (1980). Proc. Narl. Acad. Sci. U.S.A. 77, 3595-3599. Friend, K. K., Chen, S., and Ruddle, F. H. (1976). Somaric Cell Genet. 2, 183-188. Geever, R. F.. Wilson, L. B., Nallaseth, F. S . , Milner, P. F., Bittner, M., and Wilson, J. T. (1981). Proc. Narl. Acad. Sci. U.S.A. 78, 5081-5085. George, D. L., Phillips, J. A., 111. Francke, U., and Seeburg, P. H. (1981). Hum. Genet. 57, 138-141. Gerhard, D. S., Kawasaki, E. S . , Bancroft, F. C., and Szabo. P. (1981). Proc. Narl. Acad. Sci. U.S.A. 78, 3755-3759. Goeddel, D. V., Leung, D., Dull, T., Gross, M., Lawn, R., McCandiss, R., Seeburg. P., Ullrich. A., Yelverton, E., and Gray, P. W. (1981). Nature (London) 290, 401-408. Goss, S. (1979). Cytogenet. Cell Genet. 25, 161. Goss, S., and Harris, H. (1975). Nature (London) 255, 680-684. Goss, S . , and Harris, H. (1977a) J. Cell Sci. 25, 17-37. Goss, S., and Harris, H. (1977b). J . Cell Sci. 25, 39-58. Green, M. C . (1968). In “Biology of the Laboratory Mouse” (E. L. Green, ed.), pp. 87-150. Dover, New York. Grzeschik, K. H., Alderice, P. W., Grzeschik, A., Opitz, J. M., Miller, 0. J., and Siniscalco, M. (1972). Proc. Natl. Natl. Acad. Sci. U.S.A. 69, 69-73. Gusella, J., Varsanyi-Breiner, A., Kao, F. T., Jones, C., Puck, T. T., Keys, C.. Orkin, S.. and Housman, D. (1979). Proc. Narl. Acad. Sci. U.S.A. 76, 5239-5243. Gusella, J. F., Kays. C.. Varsanyi-Breiner, A., Kao, F. T., Jones, C., Puck, T. T., and Housman, D. (1980). Proc. Narl. Acad. Sci. U.S.A. 77, 2829-2833. Gusella, J. F., Jones, C., Kao, F. T., Housman, D., and Puck, T. T. (1982). Proc. Narl. Acad. Sci. U.S.A. 79, 7804-7808. Haldane. J. B. S., and Smith, C. A. B. (1947). Ann. Eugen. 14, 10-31.

142

FA-TEN KAO

Ham, R. G.(1963). Exp. Cell Res. 29, 515-526. Ham, R. G.(1965). Proc. Natl. Acad. Sci. U.S.A. 53, 288-293. Harper, M. E., Ullrich, A., and Saunders, G. F. (1981). Proc. Narl. Acad. Sci. U.S.A. 78, 458-4460, Hams, H. (1967). J. Cell Sci. 2, 23-32. Harris, H. (1970). “Cell Fusion.” Harvard Univ. Press, Cambridge, Massachusetts. Harris, H., and Hopkinson, D. A. (1976). “Handbook of Enzyme Electrophoresis in Human Genetics.” North-Holland Publ., Amsterdam. Harris, H., and Watkins. J. F. (1965). Nature (London) 205, 640-646. Harris, H., Sidebottom, E., Grace, D. M., and Bramwell. M. E. (1969). J. Cell Sci. 4, 499-526. Heintz, N.. Zernik, M.. and Roeder, R. G.(1981). Cell 24, 661-668. Heisterkamp, N., Groffen, J., Stephenson, J . R., Spurr, N. K., Goodfellow, P. N.. Soloman, E., Carritt, B., and Bodmer, W. F. (1982). Nature (London) 299, 747-749. Hershfield, M., and Francke, U. (1982). Science 216, 739-742. Hobart, M. J., Rabbitts, T. H., Goodfellow, P. N., Soloman, E., Chambers, S., Spurr. N.. and Povey, S. (1981). Ann. Hum. Genet. 45, 331-335. Houck, C. M., Rinehart, F. P., and Schmid, C. W. (1979). J . Mol. Biol 132, 289-306. Hozumi, N . , and Tonegawa, S . (1976). Proc. Narl. Acad. Sci. U.S.A. 73, 3628-3632. Human Gene Mapping 6 (1981). Cytogenef. Cell Genet. 32, 1-341. Jami. J., Grandchamp, S., and Ephrussi, B. (1971). C. R. Acad Sci. Paris 272, 323-326. Jeffreys, A. J., and Flavell, R. A. (1977). Cell 12, 1097-1 108. Jelinek, W. R., Tommey, T. P., Leinwand, L., Duncan, C. H., Biro, P. A., Choudary, P. V., Weissman, S. M.,Rubin, G. M., Houck, C. M., Deininger. P., and Schmid, C. W. (1980).Proc. Natl. Acad. S C ~U.S.A. . 77, 1398-1402. Johnson, R. T., and Rao, P. N. (1970). Nature (London) 226, 717-722. Jolly, D. J., Esty, A. C., Bernard, H. U.,and Friedmann, T. (1982). Proc. Narl. Acad. Sci. U S A . 79, 5038-5041. Jones, C. (1975). Somatic Cell Genet. 1, 345-354. Jones, C., and Kao, F. T. (1978). Hum. Genet. 45, 1-10, Jones, C., Wuthier, P., Kao, F. T., and Puck, T. T. (1972). J. Cell. Physiol. 80, 291-298. Jones, C., Kao, F. T., and Taylor, R. T. (1980). Cytogenet. Cell Genet. 28, 181-194. Jones, C., Patterson, D., and Kao, F. T. (1981). Somatic Cell. Genet. 7 , 399-409. Junien, C., Weil, D., Myers, 1. C.. Van Cong, N., Chu, M.L., Foubert. C., Gross,M. S., Prockop, D. J., Kaplan, J . C., and Ramirez, F. (1982). Am. J. Hum. Genet. 34, 381-387. Kan, Y. W., and Dozy, A. (1978). Proc. Natl Acad. Sci. U.S.A. 75, 5631-5635. Kao, F. T., and Puck, T. T. (1967). Genetics 55, 513-524. Kao, F. T., and Puck, T. T. (1968). Proc. Natl. Acad. Sci. U.S.A. 60, 1275-1281. Kao, F. T , and Puck, T. T. (1969). J. Cell. Physiol. 74, 245-258. Kao, F. T., and Puck, T. T. (1970). Nature (London) 228, 329-332. Kao, F. T., and Puck, T. T. (1972). J. Cell. Physiol. 80, 41-50. Kao, F. T., and Puck, T. T. (1974). Methods Cell Biol. 8, 23-29. Kao, F. T., and Puck, T. T. (1975). Genetics 79, 343-352. Kao, F. T., Chasin, L., and Puck, T. T. (1969a). Proc. Narl. Acad. Sci. U.S.A. 64, 1284-1291. Kao, F. T.,Johnson, R. T., and Puck, T. T. (1969b). Science 164, 312-314. Kao, F. T., Jones, C., and Puck, T. T. (1976). Proc. Natl. Acad. Sci. U.S.A. 73, 193-197. Kao, F. T., Jones, C., and Puck, T. T. (1977). Somatic Cell Genet. 3, 421-429. Kao, F. T., Hartz, J. A., Law, M. L., and Davidson, J. N. (1982). Proc. Natl. Acad. Sci. U.S.A. 79, 865-869. Kao, F. T., Hawkins, J. W., Law, M. L., and Dugaiczyk, A. (1982). Hum Genet. 62, 377-341. Kirsch, 1. R., Morton, C. C., Nakahara. K., and Leder, P. (1982). Science 216, 301-303.

SOMATIC CELL GENETICS AND GENE MAPPING

143

Klein, G. (1981). Nature (London) 294, 313-318. Klobutcher, L. A., and Ruddle, F. H. (1979). Nurure (London) 280,657-660. Kohler, G., and Milstein, C. (1975). Nature (London) 256, 495-497. Kozak, C. A , , and Ruddle, F. H. (1976). In Virro 12, 720-725. Krumlauf, R., Jeanpierre, M., and Young, B. D. (1982). Proc. Narl. Acud. Sci. U.S.A. 79, 297 1-2975. Lai, E. C . , Kao, F. T., Law, M. L., and Woo, S. L. C. (1983). Am. J. Hum. Genet. 35,385-392. Lamm, L. U.,Freidrich. U.. Peterson, G. B., Jorgensen, J., Nielsen, J., Therkelsen, A. J., and Kissemeyer-Bielsen, F. (1974). Hum. Hered. 24, 273-284. Lange, K., and Boehnke, M. (1982). Am. J. Hum. Gener. 34, 842-845. Langer, P. R.. Waldrop, A. A., and Ward, D. C. (1981). Proc. Nurl. Acad. Sci. U.S.A. 78, 6633-6637. Langer-Safer, P. R., Levine, M., and Ward, D. (1982). Proc. Narl. Acud. Sci. U.S.A. 79, 438 1-4385, Law, M. L., and Kao, F. T. (1978). Somaric Cell Genet. 4, 465-476. Law, M. L., and Kao, F. T. (1979). Cyrogener. Cell Genet. 24, 102- 114. Law, M. L., and Kao, F. T. (1982). J. Cell Sci. 53, 245-254. Law, M. L., Davidson, J. N., and Kao, F. T. (1982). Proc. Narl. Acad. Sci. U.S.A. 79,7390-7394. Lawn, R. M.. Fritsch, E. F., Parker, R. C., Blake, G.,and Maniatis, T . (1978). Cell 15, 1157- I 174. Lebo, R. V., Carrano, A. V., Burkhart-Schultz, K. J.. Dozy, A. M., Yu, L. C., and Kan, Y. W. (1979). Proc. Nurl. Acud. Sci U.S.A. 76, 5804-5808. Leder, P., Tilghman, S . M., Tiemeier, D. C., Polsky, D. C., Seidman, J. G.,Edgell, M. H., Enquist, L. W., Leder, A,, and Norman, B. (1977). Cold Spring Harbor Symp. Quunr. Biol. 42, 915-920. Little, P. F. R. (1982). Cell 23, 683-684. Littlefield, J. W. (1964). Science 145, 709-710. McBride, 0. W., and Ozer, H. L. (1973). Proc. Nurl. Acud. Sci. U.S.A. 70, 1258-1262. McBride, 0. W., Heiter, P. A., Hollis, G.F., Swan, D., Otey, M. C., and Leder, P. (1982a). J . Exp. Med. 155, 1480-1490. McBride, 0. W., Sean, D. C., Santose, E., Barbacid, M., Trenick, S . R., and Aaronson, S. A. (1982b). Nurure (London) 300,773-774. McConkey, E. H. (1980). Somuric Cell Genet. 6, 139-147. McKusick, V. A. (1978). “Mendelian Inheritance in Man,” 5th Ed. Johns Hopkins Univ. Press, Baltimore, Maryland. McKusick, V. A. (1980). J. Hered. 71, 370-391. McKusick, V. A. (1981). Birth Defects Orig. Arric. Ser. 17, 229-248. McKusick, V. A. (1982). Cyrogener. Cell Genet. 32, 7-23. McKusick, V. A., and Ruddle, F. H (1977). Science 196, 390-405. Malcolm, S., Barton, P., Murphy, C., and Ferguson-Smith, M. A. (1981). Ann. Hum. Gener. 45, 135- 141. Malcolm, S . , Barton, P . , Murphy, C., Ferguson-Smith, M. A,. Bentley, D. L., and Rabbitts, T. H. (1982). Proc. Nutl. Acud. Sci. U.S.A. 79, 4957-4961. Markert, C. L., and Moller, F. (1959). Proc. Narl. Acud. Sci. U.S.A. 45, 753-763. Maxam, A. M., and Gilbert, W. (1977). Proc. Nurl. Acud. Sci. U.S.A. 74, 560-564. Meisler, M. H., Wanner, L., Kao, F. T.,and Jones, C. (1981). Cyrogener. Cell Genet. 31, 124- 128. Migeon, B. R., and Miller, C. S. (1968). Science 162, 1005-1006. Migeon, B. R., Brown, T. R., Axelman, J., and Migeon, C. J. (1981). Proc. Narl. Acad. Sci. U.S.A. 78, 6339-6343.

144

FA-TEN KAO

Miggiano, V., Nabholz, M., and Bodmer, W. (1968). In “Hetero-Specific Genome Interaction” (V. Defendi, ed.), pp. 61-76. Wister Inst. Press, Philadelphia, Pennsylvania. Miller, D. A., and Miller, 0. J. (1972). Science 178, 949-955. Miller, 0. J.. Allderdice, P. W., and Miller, D. A. (1971). Science 173, 244-245. Minna, J. D., and Coon, H. G. (1974). Nature (London) 252, 401-404. Moore, D. D., Conkling, M. A., and Goodman, H. M. (1982). Cell 29, 285-286. Moore, E. E., Jones, C., Kao, F. T., and Oates, D. C. (1977). Am. J . Hum. Genet. 29, 389-396. Morton, N. E. (1955). Am. J. Hum. Genet. 7, 277-318. Mumy, J. M., Davies, K. E., Harper, P. S., Meredith, L., Mueller, C. R., and Williamson. R. (1982). Nature (London) 300, 69-71. Nash, W. G., and O’Brien, S. J. (1982). Proc. Nurl. Acud. Sci. U.S.A. 79, 6631-6635. Naylor, S . L., Sakaguchi, A. Y., Shows, T. B., Law, M. L., Goeddel, D. V., and Gray, P. W. (1983). J . Exp. Med. 57, 1020-1027. O’Brien, S. J., and Nash, W. G. (1982). Science 216, 257-265. Oda, M., and Puck, T. T.(1961). J . Exp. Med. 113, 599-610. O’Farrell, P. H. (1975). J . Eiol. Chem. 250, 4007-4021. Okada, Y. (1958). Eiken J . 1, 103-1 10. Owerbach, D., Bell, G. I., Rutter, W. J., and Shows, T. B. (1980a). Nature (London) 286, 82-84. Owerbach, D., Rutter, W. J., Martial, J. A., Baxter, J. D., and Shows, T. B. (l980b). Science 209, 289-292.

Owerbach, D., Bell, G. I., Rutter, W. J., Brown, J. A., and Show, T. B. (1981a). Diabetes 30, 267-270.

Owerbach, D.. Rutter, W. J., Shows, T. B., Gray, P., Goeddel, D. V., and Lawn, R. M. (1981b). Proc. Nurl. Acud. Sci. U.S.A. 78, 3123-3127. Owerbach, D., Rutter, W. J., Cooke, N. E., Martial, J. A,, and Shows, T. B. (1981~).Science 212, 815-816.

Owerbach, D., Rutter, W. J., Roberts, J. L., Whitfeld, J. S., Seeburg. P. H.,and Shows, T. B. (1981d). Somatic Cell Genet. 7, 359-369. Pardue, M. L., and Gall, J. G. (1970). Science 168, 1356-1358. Patterson, D. (1980). Somaric Cell Genet. 6, 101-114. Patterson, D., Craw, S.. and Jones, C. (1981). Proc. Nut/. Acud. Sci. U.S.A. 78, 405-409. Pontecorvo, G. (1975). Somaric Cell Genet. 4, 397-400. M a s h , K.. McBride, 0. W.. Swan, D. C.. Devare, S. G., Tronick, S. R., and Aaronson. S. A. (1982). Proc. Nurl. Acud. Sci. U.S.A. 79, 5210-5214. Puck, T. T. (1972). “The Mammalian Cell as a Microorganism: Genetic and Biochemical Studies In Vitro.” Holden-Day, San Francisco, California. Puck, T. T. and Kao, F. T. (1967). Proc. Nurl. Acud. Sci, U.S.A. 58, 1227-1234. Puck, T. T.. and Kao, F. T. (1982). Annu. Rev. Genet. 16, 225-271. Puck, T. T., and Marcus, P. I. (1955). Proc. Nurl. Acud. Sci. U.S.A. 41, 432-437. Puck, T. T., Wuthier, P., Jones, C., and Kao, F. T.(1971). Proc. Narf. Acud. Sci. U.S.A. 68, 3 102-3 106.

Race, R. R., and Sanger, R. (1975). “Blood Groups in Man,” 6th Ed. Blackwell, Oxford. Rao, P. N., and Johnson, R. T. (1970). Nurure (London) 225, 159-164. Renwick, J. H.(1969). Er. Med. Bull. 25, 65-73. Renwick, J. H. (1971). Annu. Rev. Genet. 5, 81-120. Rethore, M. O . , Junien, C., Malpuech, G., Baccichetti, C., Tenconi, C., Kaplan, J. C., de Romeuf, F., and Lejeune, J. (1976). Ann. Gener. 19, 140-142. Rethore, M. 0..Kaplan, J. C., Junien, C., and Lejeune, J. (1977). Hum. Genet. 36, 235-237. Ricciuti, F., and Ruddle, F. H.(1973). Nature (London) New Eiol. 241, 180-182. Ringertz, N. R., and Savage, R. E. (1976). “Cell Hybrids.” Academic Press, New York.

SOMATIC CELL GENETICS AND GENE MAPPING

145

Robson, E. B., Polani. P. E., Dart, S. J., Jacobs, P. A., and Renwick, J. H. (1969). Nature (London) 223, 1163-1 165. Ropley, S.. Robson, E. B., Harris, H., and Smith, S. M. (1967). Ann. Hum. Genet. 31, 237-242. Rowley, J. D. (1982). Science 216, 749-751. Ruddle, F. H. (1982). Nature (London) 294, 115-120. Ruddle, F. H . , and Creagan, R. P. (1975). Annu. Rev. Genet. 9, 407-486. Sakaguchi, A. Y . , Naylor, S. L., Quinto, C., Rutter, W. J., and Shows, T. B. (1981). Cytogenet. Cell Genet. 32, 313. Sakaguchi, A. Y.,Naylor, S . L., Weinberg, R. A., and Shows, T. B. (1982).Am. J. Hum. Genet. 34, 175A. Sanger, F., Nicklen, S., and Coulson, A. R. (1977). Proc. Natl. Acad. Sci. U.S.A. 74,5463-5467. Schmeckpeper, B. J., Huntington, F. W., and Smith, K. D. (1981). Nucleic Acids Res. 9, 1853-1872. Schmid, C. W., and Dehninger, P. (1975). Cell 6, 345-358. Schmid, C. W., and Jelinek, W. R. (1982). Science 216, 1056-1070. Scoggin, C. H., Gabrielson, E., Davidson, J. N., Jones, C., Patterson, D., and Puck. T. T. (1981). Somatic Cell Genet. 7, 389-398. Scott, A. F., Phillips, J. A., 111, and Migeon, B. R. (1979). Proc. Natl. Acad. Sci. U.S.A. 76, 4563-4565. Sinensky, M. (1977). Biochem. Biophys. Res. Commun. 78, 863-867. Sinensky, M. (1978). Proc. Narl. Acad. Sci. U.S.A. 75, 1247-1249. Singer, M. F. (1982). Cell 28, 433-434. Siniscalco, M. (1979). Prog. Med. Genet. 3, 221-307. Skolnick, M. H., and White, R. (1982). Cyrogenet. Cell Genet. 32, 58-67. Slate, D. L., D’Eustachio, P., Pravtcheva, D., Cunningham, A. C., Nagata, S., Weissman, C . , and Ruddle, F. H. (1982). J. Exp. Med. 155, 1019-1024. Sorieul, S., and Ephrussi, B. (1961). Nature (London) 190, 653-654. Southern, E. (1975). J. Mol. Biol. 98, 503-517. Stamato, T. D., and Waldren, C. A. (1977). Somatic Cell Genet. 3, 431-440. Swan, D. C . , McBride, 0. W., Robbins, K. C., Keithley, D. A., Reddy, E. P., and Aaronson, S. A. (1982). Proc. Narl. Acad. Sci. U . S . A . 79, 4691-4695. Szybalska, E. H., and Szybalski, W. (1962). Proc. Natl. Acad. Sci. U.S.A. 48, 2026-2034. Thompson, L. H . , Mankovitz. R., Baker, R. M., Till, J. E., Siminovitch. L., and Whitmore, G. F. (1970). Proc. Natl. Acad. Sci. U.S.A. 66, 377-384. Tjio, J . H., and Levan, A. (1956). Hereditas 42, 1-6. Tjio, J . H . , and Puck, T. T. (1958). Proc. Natl. Acad. Sci. U.S.A. 44, 1229-1231. Trent. J. M.. Olson, S., and Lawn, R. M. (1982). Proc. Natl. Acad. Sci. U . S . A . 79. 7809-7813. Weinberg, R. A. (1982). Cell 30, 3-4. Weiss, M. C . , and Green, H.(1967). Proc. Natl. Acad. Sci. U.S.A. 58, 1104-1 I 1 I . Westerveld, A., Jongsma, A. P. M.. Meera, Khan, P.. Van Someren, H.,and Bootsma, D. (1976). Proc. Natl. Acad. Sci. U . S . A . 73, 895-899. Whitehead, A. S., Solomon, E., Chambers, S . , Bodmer, W. F., Povery, S . , and Fey, G . (1982). Proc. Natl. Acad. Sci. U.S.A. 79, 5021-5025. Wigler, M., Silverstein, S., Lee, L. S.. Pellicer. A,, Cheng, Y.C., and Azel, R. (1977). Cell 11, 223-232. Wigler, M., Sweet, R., Sim, G. K., Wold, B., Pellicer, A., Lacy, D., Maniatis, T., Silverstein, S., and Azel, R. (1978). Cell 16, 777-885. Wigler, M., Pellicer. A . , Silverstein, S., and Axel, R. (1979). Cell 14, 725-731. Wilde, C. D., Chow, L. T., Wefald, F., andcowan, N. J . (1982).Proc. Narl. Acad. Sci, U.S.A. 79, 96-100.

146

FA-TEN KAO

Wilson, E. B. (191 1). Arch. Mikrosk. Anar. Enfwicklungsmech. 77, 249-211. Wilson, J. T., Milner, P. F., Summer, M. E., Nallaseth, F. S . , Fadel, H. E., Reindollar. I<. H., McDonough. P. G.,and Wilson, L. B. (1982). Proc. Narl. Acad. Sci. U.S.A. 79, 3628-3631. Young, B. D., Ferguson, Smith, M. A., Sillar, R . , and Boyd, E. (1981). Proc. Narl. Acad. Sci. U.S.A. 78, 1121-7731. Yu, L. C., Szabo, P., Borun, T . W . , and Prensky, W. (1978). Cold Spring Harbor Symp. Quanr. Biol. 42, 1101-1105. Yunis, J . J. (1976). Science 191, 1268-1210.

INTERNATIONAL REVIEW OF CYTOuXiY. VOL. 85

Tubulin Isotypes and the Multigene Tubulin Fami1ies N. J. COWANAND L. DUDLEY Department of Biochemistry, New York University School of Medicine, New York, New York I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Number and Complexity of Tubulin Genes

.......

A. Sequence Analysis of Tubulin B. Tubulin-Specific cDNA Probes. . ...... C. Tubulin Isotypes ............................... 111. Functional and Nonfunctional Tubulin Genes . . . . . A. Structure of Expressed Tubulin Genes . . . . . . . B. Tubulin Pseudogenes ........................... C. Dissection of the Multigene F IV. Tubulin Gene Expression ........................... V. Genetic Complexity and Functiona References ...........................................

147 148

148 156

I. Introduction Microtubules are long filamentous structures that are found ubiquitously in eukaryotic cells; and are intimately associated with many cellular functions. These functions include mitosis (as part of the mitotic spindle), cellular architecture (as part of the cytoskeleton), motility (as components of cilia and flagella), and intracellular transport and secretion. The major structural component of microtubules is tubulin, a dimer of a-and P-chains each of about 55,000daltons. In spite of the remarkable diversity of microtubule functions, many independent lines of evidence point to a close evolutionary conservation among tubulin proteins. For example, N-terminal sequence data indicate considerable homology between tubulins from sea urchin and chicken (Luduena and Woodward, 1973). Antisera raised against tubulin also cross-react across species boundaries (Osborne and Weber, 1977). Indeed, hybrid microtubules may be formed in v i m by copolymization of a-and P-tubulins from different species (Snyder and McIntosh, 1976). Finally, the evolutionary conservation of tubulin genes is dramatically shown at the nucleic acid level by the ability of cDNA probes constructed from chicken a- and P-tubulin mRNAs to cross-hybridize with genomic DNAs from a wide variety of eukaryotes under stringent conditions (Cleveland et al., 1980). 147 Copyright 0 1983 by Academic Press. Inc. All rights of reproduction in any form r e ~ e ~ e d . ISBN 0-12-364485-2

148

N. J. COWAN AND L. DUDLEY

Given the ubiquitous nature of microtubules, the evolutionary conservation of tubulin proteins is perhaps not surprising. Nonetheless, there has been speculation that different tubulin gene products contribute to functionally different microtubules. Indeed, the diversity of microtubule function, and the dynamic nature of at least some of these functions, suggests that microtubule assembly/ disassembly is carefully controlled. To what extent is this process dependent upon gene transcription and its regulation? A basic approach to this question requires an examination of tubulin genes themselves. This article describes our current knowledge of tubulin gene structure, evolution, and expression.

11. Number and Complexity of Tubulin Genes A. SEQUENCE ANALYSIS OF TUBULIN PROTEINS A great many reports exist describing heterogeneity in a- and P-tubulins; no attempt is made to list these here. With few exceptions, all are based on the observation of multiple bands or spots on one- or two-dimensional protein gels. This heterogeneity may well be a consequence of posttranslational changes, and thus, such an approach cannot satisfactorily address the question of the genetic complexity of tubulins. A more definitive approach would necessitate a direct examination either of protein sequences from purified microtubules, or of tubulin genes themselves and their transcription products. The former approach is restricted because information is necessarily confined to the expressed products in the tissue of origin, and only limited inferences may be drawn from amino acid sequence data about genetic structure and organization. Nonetheless, useful information has been generated from the complete amino acid sequence of porcine brain a-and P-tubulins (Kraus et al., 1981; Ponstingl et al., 1981). Comparison of the optimally aligned sequences from the two proteins shows a homology of 41%, strongly suggestive of a common ancestral gene encoding a-and P-chains. There are no disulfide bridges. In the case of a-tubulin, at least six positions (around amino acid 270) can each be occupied by one of two amino acids. By assigning the alternative sites to specific peptides, Ponstingl et al., concluded that four a-tubulins are expressed in porcine brain. A similar analysis of Ptubulin from the same source also revealed internal heterogeneity, indicating the existence of at least two expressed P-tubulin genes.

B. TUBULIN-SPECIHC cDNA PROBES The construction of cloned cDNA probes has proved invaluable in the analysis of the structure and expression of a broad array of eukaryotic genes, and in this regard tubulin is no exception. The first a-and p-tubulin-specific clones were

TUBULIN ISOTYPES AND FAMILIES

149

constructed from embryonic chick brain tissue, a rich source of microtubules, and an abundant source of translatable a-and P-tubulin mRNAs. Although these mRNAs are of almost indistinguishable size (as determined by their migration in gels under fully denaturing conditions), a partial resolution could be obtained under native conditions, a result presumably ascribable to differences in secondary structure (Cleveland er al., 1978). In any event, this material proved to be suitable for the successful construction of cloned cDNAs by established procedures. Plasmids containing about 75% and >90% of the sequences present in chicken a-and P-tubulin mRNAs were obtained and have served as the basis for much of our current knowledge of tubulin gene structure (Cleveland et al., 1980). In spite of the significant homology between a- and P-tubulins at the amino acid level, no cross-hybridization could be detected between the two probes. On the other hand, the chicken tubulin cDNA probes cross-reacted with genomic sequences from all eukaryotic species tested, under conditions of high stringency. Only yeast DNA fails to give a hybridization signal under these conditions, though yeast tubulin genes can be detected at lower stringency.

FIG. I . Cross-hybridization of Hind111 digests of DNA from yeast (track 1). sea urchin (track 2). chicken (track 3), and human DNA (track 4). (A) Chicken a-tubulin cDNA probe: (B)chicken ptubulin cDNA probe. Data from Cleveland el al. (1980).

1 5 Met Arg Glu Cys I l e ATG CGT GAG TGC ATC

Human

T

Rat Pig1

Hman e

10 20 30 Ser I l e H i s Val Gly Gln A l a Gly Val Gln I l e Gly Asn A l a Cys T r p Glu Leu T y r Cys Leu Glu H i s Gly I l e Gln P r o Asp Gly Gln TCC ATC CAC GTT GGC CAG GCT GGT GTC CAG ATT GGC M T GCC TGC TGG GAG CTC TAC TGC CTG G M CAC GGC ATC CAG CCC GAT GGC CAG C

C

Rat

T

T

lA

O

[Pig1

Human

40 50 60 Y e t P r o Ser Asp Lys T h r I l e Gly G l y Gly Asp Asp Ser Phe Asn Thr Phe Phe Ser Glu Thr Gly A l a Gly Lys H i s Val P r o Arg A l a ATG CCA ACT GAC M G ACC AT1 GGG GGA GGA GAT GAC TCC TfC M C ACC TTC TTC AGT GAG ACG GGC GCT GGC AAG CAC GTG CCC CGC GCT

Rat A

no

Chicken

G

A

data

C

C

G

G

C

C

C

T

C

CPigI Human

70 80 90 Val Phe Val Asp Leu G l u P r o Thr Val I l e Asp Glu Val A r g Thr Gly Thr T y r A r g Glu Leu Phe H i s Pro Glu G l n Leu I l e Thr Gly GTG TTT GTA GAC TTG GAA CCC ACA GTC AT1 GAT GAA GTT CGC ACT GGC ACC TAC CGC CAG CTC TTC CAC CCT GAG CAG CTC ATC ACA GGC

Rat

C

1

C

4

Chicken

C [Pi91

G

C

G

C

G

C

G

G

G

G

G

G

C

G

G

Human

100 110 120 Lys Glu Asp A l a A l a Asn Asn Tyr A l a Arg Gly H i s Tyr Thr I l e Gly Lys Glu I l e I l e Asp Leu V a l Leu Asp Arg I l e Arg Lys Leu M G G M GAT GCT GCC M T M C TAT GCC CW GGG CAC TAC ACC A T 1 GGC M G GAG ATC ATT GAC CTT GTG TTG CAC CGA ATT CGC AAG CTG

Rat

T

Chicken

G

G

C

C

C

C

G

C

C

G

A

c

C

c

[Pig1

Human

130 140 150 Ala Asp Gln Cys Thr Gly Leu Gln Gly Phe Leu Val Phe H i s Ser Phe Gly Gly Gly Thr Gly Ser Gly Phe Thr Ser Leu Leu Met Glu GCT GAC CAG TGC ACC CGT CTT CAG GGC TTC TTG G l T TTC CAC AGC TTT GGT GGG GGA ACT GGT TCT GGG TTC ACC TCC CTG CTC ATG GAA

Rat G

Chicken

G

C

C

G

Ser C Ser

G

CPi91

G

C

160 Hman Rat

G

C

C

C

C

C

C

G

170

G

G

G

G

180

A r g Leu Ser Val Asp Tyr Gly Lys Lys Ser Lys Leu Glu Phe Ser I l e T Y r Pro Ala Pro Gln Val Ser Thr A l a Val V a l Glu Pro Tyr CGC CTG TCA GTT GAT TAT GGC M G AAA TCC M G CTG GAG TTC TCC ATT TAC CCG GCA ccc CAG GTT TCC ACA GCT GTA GTT GAG CCC TAC

A G

Chicken

C

T G

C G

C C

C C

A

G

A

G

C

G

T

Arg C GT

G

G

C

G

G

[Pig1

Human

190 200 210 Asn Ser I l e Leu Thr Thr H l s Thr Thr Leu Glu His Ser ASP CYS A l a Phe Met Val Asp Asn Glu Ala I l e Tyr Asp I l e Cys Arg Arg M C TCC ATC CTC ACC ACC CAC ACC ACC CTG GAG C K TCT GAT TGT GcC TTC ATG GTA GAC MT GAG CCC ATC TAT GAC ATC TGT CGT AGA

Rat 1 Chicken

G

c

c

c

[Pig1

FIG.2. (see legend on p. 153).

G

C

C

C

G

Human Rat Chicken [Pis1

220 230 240 Asn Leu Asp I l e Glu Arg Pro Thr Tyr Thr Asn Leu Asn Arg Leu I l e Ser Gln I l e Val Ser Ser I l e Thr A l a Ser Leu Arg Phe Asp AAC CTC GAT ATC GAG CGC CCA ACC TAC ACT M C CTT M C CGC CTT A T 1 AGC CAG AT1 GTG TCC TCC ATC ACT GCT TCC CTG AGA TTT GAT GlY C T T A A G T G A G T A T C GlY A C A C C C G C C C C A G G A G G C G1Y

2M Human

260

270

Gly A l a Leu Asn Val Asp Leu Thr Glu Phe Gln Thr Asn Leu Val P r o T y r Pro Arg I l e H i s Phe Pro Leu A l a Thr T y r A l a P r o Val GW GCC CTG AAT GTT GAC CTG ACA GM TTC CAG ACC M C CTG GTG CCC TAC CCC CGC ATC CAC TTC CCT CTG GCC ACA TAT GCC CCT GTC

Rat

G G I

T

T

Chfcken T

c

c

G

G

T

T

Human

G

A T

G

G

G

/Ile

[Pig]

280 290 300 I l e Ser A l a Glu Lys Ala T y r H i s Gly Gln Leu Ser Val A l a Glu I l e Thr Asn A l a Cys Phe G l u P r o Ala Asn G l n Met Val Lys Cys ATC TCT GCT GAG A M GCC TAC CAT WA CAG CTT TCT GTA GCA GAG ATC ACC M T GCT TGC TTT GAG CCA GCC M C CAG ATG GTG A M TGT

Rat

C Chicken

G

C

G

G

G

G

T

C

G

TY r A

C

G

G

C

[Pig1

Human

310 320 330 Asp Pro Arg H i s Gly Lys T y r %t A l a Cys Cys Leu Leu Tyr A r g Gly Asp Val Val P r o Lys Asp Val Asn A l a A l a I l e A l a Thr I l e GAC CCT GGC CAT GGT MA TAC ATG GCT TGC TGC CTG TTG TAC CGT GGT GAC GTG GTT CCC M A GAT GTC MT GCT GCC ATT GCC ACC ATC

Rat T

C

C

Chicken

G C G [Pig1

C

C

G

G

C

C

G

G

C

C

C

Human

340 350 360 Lys Thr Lys Arg Thr I l e Gln Phe Val Asp T v Cys Pro Thr Gly Phe Lys Val Gly I l e Asn T y r G l n P r o P r o Thr Val Val P r o Gly AAA ACC AAG CGC ACG ATC CAG TTT GTG GAT TGG TGC CCC ACT GGC TTC AAG GTT CGC ATC AAC TAC CAG CCT CCC ACT GTG GTG CCT GGT

Rat

G

T

C

C

T

C

T

Chicken

G

C /Ser

[Pig1

Human

T

C

A

A

C

G

T

C

C

G

G1u

3 70 380 330 Gly Asp Leu A l a Lys Val Gln Arg A l a Val Cys Met Leu Ser Asn Thr Thr A l a I l e A l a Glu A l a T r p A l a Arg Leu Asp H i s Lys Phe GGA GAC CTG GCC AAG GTA CAG AGA GCT GTG TGC ATG CTG AGC M C ACC ACA GCC A T 1 GCT GAG GCC TGG CCT CGC CTG GAC CAC AAG TTT

Rat C

C

G

G

T

T

Chicken

C C

C

G

c

c

G

G

T

[Pig1

W VI

Human

400 410 420 Asp Leu Met T y r A l a Lys A r g Ala Phe Val H i s TrP TYr Val Gly Glu Gly Met Glu Glu Gly G l u Phe-Ser Glu A l a Arg Glu Asp Met GAC CTG ATG TAT GCC M G CGT GCC TTT GTT CAC TGG TAC GTG GGT GAG GCG ATG GAG GAA GGC GAG TTT TCA GAG GCC CGT GAA GAT ATG

Rat

T

G

C

G

A

C

C

G

G

C

T

G

C

Chicken

C

C

C

T

G

G

T

G

G

[Pig1

Human

430 440 450 Ala A l a Leu Glu Lys Asp T y r Glu G l u Val Gly Val Asp Ser Val Glu Gly Glu G l y Glu Glu Glu Gly Glu Glu T y r GCT GCC CTT GAG AAG GAT TAT GAG GAG GTT GGT GTG GAT TCT GTT GAA GGA GAG CGT GAG GAA GAA GGA GAG GAA TAC

Rat A

G

G

T

G

G

T

Chicken

G

C

G

G

G

G

G

CPigI

FIG. 2. Comparison of a-tublin sequences derived from human (Cowan et a / . . 1983). rat (Lemischka and Sharp, 1982). and chicken (Valenzuela er al., 1981) cDNA clones and the amino acid sequence of pig brain a-tubulin (Ponstingl et al.. 1981).

10 Hunan ( 0 - e l ) pig Chicken

Met Arg Glu I l c V a l Hls I l e

20

Gln Ala Gly Gln Cys f l y Asn Gln I l e Gly Ala LYS

30

Phe T r p

Glu Val I l e Ser A g Glu His GlY I l e

50 60 Ser Asp Leu Gln Leu Asp A r g I l e Ser Val Tyr Tyr Asn Glu A l a Thr G1y Gly L Y T~Y r Val Glu /Asn 0%' f i t

40 ASP pro Thr GlY Thr Tyr His Gly Asp

Ser

Ser

/Val

GlU

70 b a n (0-ell p i9 Chicken

g P

b a n (0-el) pig Chicken

e

Asn

Gly Asn

80

90

Pro A q Ala I l e Leu V a l ASP Leu Glu P r o Gly Thr #t Asp Ser V a l A r g Ser Gly Pro Phe Gly Gln

IlC

Phe A r p pro ASP

phe

100 110 120 V a l phe G1y Gln Ser Gly A l d Gly Asn Asn T r p A l d Lys Gly His Tyr Thr Glu Gly A l a Glu Leu V a l Asp Ser Val Leu ASP V a l Val

130

140

150

A r g LyS Glu A l a Glu Ser Cys Asp Cys Leu Gln Gly Phe Gln Leu Thr His Ser Leu Gly G l y Sly Thr Gly Ser Gly #t Gly Thr Leu Ser Ser

160

Hnan ( 0 - e l l

170

180

Leu I l e Ser Lys I l e A r g Glu Glu Tyr P r o Asp A r g I l e W e t Asn Thr Phe Ser V a l V a l Pro Ser Pro Lys V a l Ser Asp Thr V a l Val

p i9 Chicken

Wet

190

200

210

Glu Pro Tyr Asn A l a Thr Leu Ser Val H I S Gln Leu V a l Glu Asn Thr Asp Glu Thr Tyr Cys I l e Asp Asn Glu A l a Leu Tyr Asp I l e

Cyf Phe

220 A r s Thr Leu Am Leu Thr Thr Pro Thr T y r

2 30 Gly ASP Leu Asn His Leu V a l k r Gly Thr

a

A1 A1a

240 Met Glu Cys Yal Thr Thr Cys Leu

Ser Gly Ser Gly

Human (0-61) Pig Chicken H n a n (0-61) p i9 Chicken

250 2 60 270 A r g Phe P r o Gly Gln Leu Asn A l a Asp Leu A r g Lys Leu A l a Val Asn Met Val Pro Phe P r o A r g Leu His Phe Phe Met P r o Gly Phe

280 A l a P r o Leu Thr Ser A r g Gly Ser Gln Gln T y r A r g /A1 a

Ald

290 300 Leu Thr Val Pro Asp Leu Thr Gln Gln Val Phe Asp A l a Lys Asn Met Met G1u Met Gl u met Ser

310 320 330 A l a A l a Cys Asp Pro A r g ffls Gly A r g T y r Leu Thr Val Ala Ala Val Phe A r g Gly A r g Met Ser Met Lys Glu Val Asp Glu Gln Met Ile

340 350 360 Leu Asn Val Gln Asn Lys Asn Ser Ser T y r Phe Val Glu T r p I l e P r o Asn Asn Val Lys Thr A l a Val Cys Asp I l e P r o P r o Arg Gly

VI

Human (0-61) p i9 Chicken

LCU Lys

3 70 380 3 90 M t Ala Val Thr Phe I l e Gly Asn Ser Thr Ala H e Gln G l i i Leu Phe Lys Arg I l e Ser Glu Gln Phe Thr Ala Met Phe A r q

400 A r g LYS A l a Phe Leu ffls Trp Tyr Thr Gly

Glu

G l Y Met ASP Glu Met Glu

430 Glu T y r Gln Gln T y r Gln Asp A l a Thr Ala G1u G1u Glu G1u Asp Gln Gly Asp Gln Gly Gln Gly

Phe

410 420 Thr Glu Ala Glu Ser Asn Met Asn Asp Leu Val Ser

440 Asp Phe Gly Glu Clu Ala Glu Glu Glu Ala Glu Glu G1Y ASP Glu Glu 61Y ASP Glu Glu Glu Val

FIG. 3. Comparison of p-tubulin sequences from human, pig, and chicken. Data from Hall era!. (1983). Kraus et al. (1981). and Valenzuela et al. (1981). The C-terminal sequence encoded by a second human p-tubulin gene, SP, is included in the figure: note that the homology of this gene is greater compared to pig and chicken sequences than to the human (Dp-1) sequence (see text).

I56

N. J . COWAN AND L. DUDLEY

Interestingly, the number of bands detected in genomic Southern blots using the chicken cDNA probes varies considerably depending on the source of the DNA (Fig. 1). In the homologous case, i.e., chicken DNA, about four fragments are evident per restriction digest. Since probes derived from either the 3‘ or 5’ end of each cDNA clone also detected about four bands per restriction digest, this number must reflect the presence of four a- and four (3-tubulin genes in chickens. Thus, tubulins are encoded by a multigene family. By contrast with chicken, DNA from sea urchin, mouse, or humans yields a genomic Southern band-pattern of much greater complexity, with about 15 bands per restriction digest. The evolutionary significance of these large multigene families is discussed below. Since the construction of chicken a-and P-tubulin cDNA probes, clones from other sources have been obtained, either by traditional procedures using enriched mRNA fractions as in the case of rat (Lemischka er af., 1981; Ginzburg tit al., 1981), Chfamydomonas (Silflow and Rosenblum, 1981), and sea urchin (Alexandraki and Ruderman, 1981) or by screening cDNA libraries from a variety of sources (Cowan et af., 1983). In the case of a-tubulin, sequence analysis of cDNA clones has confinned the remarkable degree of conservation within the coding regions (Fig. 2): at the nucleic acid level the vast majority of differences between chicken, rat, and human P-tubulins are a consequence of ‘,silent” changes that do not affect the amino acid sequence. As might be expected, the rat and human sequences are more homologous to one another than either is to chicken. Curiously, the majority of third codon position differences between the avian and mammalian sequences are a result of G or C residues (in the chicken sequence) being substituted by A or T residues in the human or rat sequence. The G/C-rich nature of the chicken sequence, which is also a feature of the sequence of chicken P-tubulin genes, is of unknown significance. An unexpected result of the sequence analysis of a-tubulin cDNAs is the existence of a tyrosine codon immediately preceding the termination codon. An a-tubulin-specific enzyme which requires ATP and functions in the presence of protein synthesis inhibitors has been reported to be capable of the posttranslational addition of tyrosine to the protein (Raybin and Flavin, 1977). However, the existence of the C-terminal tyrosine codon in chicken, rat, and human atubulin cDNAs shows that the primary posttranslational event is the enzymatic removal, rather than the additon, of tyrosine (Valenzuela e t a f . ,1981; Lemischka et af., 1981; Cowan et al., 1983). The functional significance of this process is unknown. C. TUBULIN ISOTYPES In the case of P-tubulin, the picture is somewhat different. Here. the amino acid sequence deduced from a cDNA clone constructed from human fetal brain

TUBULIN ISOTYPES AND FAMILIES

157

mRNA shows less homology to pig and chicken P-tubulin sequences than the latter do to each other, with major divergence within the 15 carboxy terminal amino acids (Fig. 3). This surprising result contrasts with the expectation that the two mammalian species would be more similar to one another than to an avian species. On the other hand, an independently isolated, functionally expressed human genomic sequence (54,Cowan et a/., 1981 ; Hall et al., 1983) possesses very extensive homology with chicken and pig P-tubulins. Thus, human cells appear to contain at least two distinct P-tubulin isotypes. Of the three known P-tubulin sequences, two (chicken and human) were derived from nucleotide sequences of cDNA clones, whereas the third (pig) was determined by direct amino acid sequencing of purified P-tubulin. Brain was in all cases the tissue of origin. In view of the existence of two distinct human ptubulin isotypes, the detection of only limited heterogeneity in the porcine sequence is somewhat surprising. One possible explanation might be that the divergent isotype represented by the human fetal brain cDNA sequence is restricted in its expression to a defined developmental stage: this isotype could then be absent from nonembryonic porcine brain. Alternatively, it is conceivable that the purification procedure used to prepare the porcine tubulin-namely, from a post- 100,000g supernatant-could have selectively enriched for a limited subset of microtubules assembled from a distinct P-tubulin isotype. This raises the intriguing possibility that functionally different microtubules are indeed polymerized from distinctly different tubulin proteins, a hypothesis originally suggested by Fulton and Simpson (1976).

111. Functional and Nonfunctional Tubulin Genes

Are all sequences detected in genomic DNA by tubulin cDNA probes expressed? Two approaches have been fruitful in addressing this question. One has been to search for cloned cDNAs, and to determine how many kinds of sequence these clones represent. Since each cDNA clone represents the amplification of a single mRNA molecule, sequence differences between different clones entails the expression of different genes. Note that, because of the close evolutionary conservation within coding sequences, such differences would be expected to be largely confined either to the third condon position (within the coding sequence) so that the amino acid sequence would remain unchanged, or to the untranslated regions, where selective constraints appear to be less rigid. A second approach has been to examine the genes themselves, and to determine, by sequence analysis and/or electron microscopic heteroduplex experiments, and blot analysis using gene-specific probes, whether the structural requirements of expression are met.

158

N. J. COWAN AND L. DUDLEY

A. STRUCTURE OF EXPRESSED TUBULIN GENES

Stnictural data on tubulin genes from three vertebrate species-human, rat, and chicken-are currently available. The most notable distinction between chickens and mammalian species is that chickens possess only four genes each for a- and P-tubulin, whereas the corresponding number in mammals is far

FIG.4. (A) Structure of four (b-e) expressed chicken P-tubulin genes determined by electron microscopic heteroduplex analysis using a chicken P-tubulin cDNA probe. (B) Interpretative drawings of the gene structure derived from A. The structure of a human P-tubulin gene (a) is included for comparison. Data from Cowan et nl. (1981) and Lopata et nl. (1983).

159

TUBULIN ISOTYPES AND FAMILIES

B

*

5' I I I I I I

I 1

1 2

I

I l l

t i \ \ \ ,

Ill\ Ill/

\ \ \ \ \ \ \ \

I l l l

I I \ \

11 ll

I I I I

Ill1

I 1

An

,L

I

\

t

I

\\:

\ \

chicken PI

1

chicken 8 2

1

chicken 8 3

I

chicken 8 4

,

I

I l l I l l

I 1

1

I

'\

I1 I l

human 58

cc

1

1 1 1 1 1 I f I 1 I

I /

t

1 1

___--___---_---_---

#---

\ \ \ \

I I

I I

3'

..--

rn \ \ \ \ \ \ \ \

chicken 8 4 (alternate)

4

L

I

I

1

0

I

2

3

..

"

1

6

-I

7 kb

FIG. 4B.

greater, that is, 15-20. Sea urchins also possess a substantial number of both aand P-tubulin sequences (Alexandraki and Ruderman, 1981 ) . In chickens, all four P-tubulin sequences are apparently expressed: fragments derived from the 3' ends of each genomic sequence detect a uniquely distinguishable mRNA upon RNA blot analysis (Lopata et a l . , 1983). Electron microscopic heteroduplex analysis of the four P-tubulin genes reveals an overall similarity in structure, with 3 or 4 small intervening sequences clustered toward the 5' portion of the coding region (Fig. 4). Interestingly, the number and size of coding regions in the chicken P-tubulin genes corresponds exactly to the number and size of the coding regions in two expressed human P-tubulin genes. However, the size of the intervening sequences is highly variable. A complete sequence analysis of one of these genes has revealed some unusual features (Gwo-Shu Lee et al., 1983). First, the largest of three intervening sequences contains a member of the dispersed, middle-repetitive A h family elements. Second, the gene is transcribed to yield two cytoplasmic mRNAs of 1.8 and 2.6 kb, by readthrough of the polyadenylation signal AATAAA (Fig. 5). Finally, no TATA sequence appears at position -35 relative to the transcriptional initiation site, although this signal, commonly found at this location in many eukaryotic genes, does occur about 100 bp further upstream. It is intriguing to speculate that the unusual promoter sequence is somehow involved in the transcriptional regulation of this human ptubulin gene. Sequence and electron microscopic heteroduplex analyses have also been performed on an a-tubulin gene isolated from rat (Lemischka and Sharp, 1982) and humans (Wilde et a l . , 1982b). Although some of the coding blocks are equivalent in size between a-tubulin genes, there are some conspicuous differences. The tubulin genes of most species thus examined constitute dispersed multigene families. In Drosophila melanogaster, for example, in situ hybridization experiments have shown that a- and P-tubulin genes are located on different chromosomes (Sanchez et al., 1980; Mischke and Pardue, 1982) (Fig. 6); a

160

N. J. COWAN AND L. DUDLEY

FIG.5 . Two mRNAs are transcribed from a single functional P-tubulin gene by readthrough of a polyadenylation site. The structure of the expressed human P-tubulin gene, M40,is shown; probe A is a subclone covering a region from approximately the termination codon to the first poly(A) site; probe B spans sequences farther downstream. Two RNA blots prepared using HeLa polyiA)+ mRNA are shown in the figure, hybridized with either probe A or probe B. Data from Gwo-Sho Lee el a/. (1983).

similar situation exists in chickens (Cleveland et af., 1981a). It seems likely that a dispersed arrangement also exists in the higher vertebrates, where a significant proportion of the multigene family members are pseudogenes generated via an mRNA intermediate (see below). In contrast to the higher eukaryotes thus far examined, the unicellular eukaryote Chlamydomonas contains only two genes each each for a-and P-tublin

TUBULIN ISOTYPES AND FAMILIES

161

(Silflow and Rosenblum, 1981) all of which are probably functionally expressed (Brunke e? al., 1982a). In addition, a curious exception to the dispersed arrangement common to Drosophila, chickens, and mammals is that found in Trypanosmoma brucei (Thomashow ef al., 1983). Here, a- and P-tubulin genes are linked in a cluster of tandem repeats of about 13-17 copies per haploid genome of alternating a-and P-tubulin sequences (Fig. 7). This arrangement may facilitate coordinate expression of the a- and 9-tubulin subunits in a species where

FIG.6 . Chromosomal locatization of a-(A) and P-tubulin (B) genes on Drosophila polytene chromosomes using chicken cDNA probes. Data from Sanchez et al. (1980).

162

N. J. COWAN AND L. DUDLEY

FIG.7. Evidence for tandemly repeated tubulin genes in Trypanosome DNA. A consequence of this arrangement is that all enzymes that cleave once within the repeat should yield fragments of the same length. The figure shows Southern blots of Trypansome DNA cleaved with KpnI (track I), Hind111 (track 2). BarnHI (track 3), and EcoRl (track 4). The probes used were a Trypanosome atubulin cDNA (A) and the chicken P-tubulin cDNA, pT2 (Cleveland et al., 1980) (B). (C) is a Southern blot of trypanosome DNA from various stages of its life cycle cut withXhol(2-4) and XhoI plus Hind 111 ( 5 ) and probed with the a-probe. Lane I is a reconstruction of a single repeat haploid genome, used to approximate the actual number of repeating units. Data from Thomashow et al. (1983).

cycles of polymerization and deploymerization are major features of the cell cycle.

B . TUBULINPSEUDOGENES The small tubulin gene family in chickens contrasts dramatically with that in mammals: for example, in humans, for which the majority of a-and P-tubulinlike sequences have been isolated from genomic libraries (Cowan et al., 1981; Wilde et al., 1982b) restriction mapping and Southern blot analysis of these

163

TUBULIN ISOTYPES AND FAMILIES

sequences using cDNA probes revealed considerable variation among the clones. Sequence analysis of the chicken P-tubulin cDNA clone has shown the P-tubulin peptide to contain 445 amino acids (Valenzuela et al., 1981), although the two human P-tubulin sequences examined thus far contain only 444 (Hall et al., 1983). Assuming this figure to be typical of all P-tubulins, then about 1335 bp are required to encode the peptide. The difference between this figure and the size of human P-tubulin mRNA (about 1.8 kb) as measured in blotting experiments is ascribable to the 5' and 3' untranslated regions plus the 3' poly(A) tract typically associated with eukaryotic mRNAs. Thus, genomic sequences of I .6 kb or less, while capable of encoding a P-tubulin peptide, could not contain the intervening sequence(s) characteristic of the great majority of expressed eucaryotic genes. The nature of the short P-tubulin-specific genomic genes has recently been established by DNA sequence analysis. A typical case is shown in Fig. 8. As suggested by the Southern blot data, there are no intervening sequences, and there is a uniformly high degree of homology with the chicken cDNA probe throughout the coding region. However, the sequence is incapable of yielding a functional polypeptide, because (1) at amino acid position 230, the serine codon TCG appears as a termination codon, and (2) at amino acid position 270, the third base of the phenylalanine codon (TTC) is absent, resulting in a frameshift leading to a termination codon beginning at the second base of amino acid 299.

AATAAA

-

- An

mRNA

1

RTase

J.

Insertion

1 ------- +_-_--_-_--------_------Repair

-4

FIG. 8. Diagram showing sequence features of an intronless, poly(A)-containing human ptubulin pseudogene generated by reverse transcription of an mRNA molecule and integration of the denote mutations in the cDNA copy into a staggered host chromosomal break. This symbols coding region acquired following integration. Horizontal arrows indicate a short direct repeat that flanks the cDNA copy. Based on the sequence data of Wilde ef al. (1982a).

(v)

164

N . J . COWAN AND L. DUDLEY

This pseudogene, so called because of its genetic lesions, has two additional remarkable features. First, a short distance downstream from the poly(A) signal AATAAA, there appears a 17 bp uninterrupted tract of A-residues. Second, the sequence of 1 1 bp that directly abutts this oligo(A) tract is directly repeated 170 bp 5’ to the initiator ATG. These two features, and the absence of any intervening sequences, must be accounted for in any explanation of how this pseudogene was generated. The addition of A-residues to mRNA molecules is a posttranscriptionalevent; thus, the appearance of an intronless poly(A)-containingmolecule in the genome entails the reverse transcription of a polyadenylated mRNA. The known polymerases capable of generating cDNAs require primers for the synthesis of both plus and minus strands. How much priming might have occurred, or which enzyme is responsible for the generation of these molecules, is unknown: however, given that successful synthesis is probably extremely rare, the actual mechanism may itself involve unlikely priming events. In any case, the cDNA must have become integrated into the host germ line. If integration occurred at a staggered chromosomal break, then insertion of the cDNA and repair of the break would result in the generation of the observed flanking direct repeats. Was the generation of this intronless poly(A)-containingpseudogene a unique event, or is the reintegration of cDNA molecules into the genome a significant phenomenon in evolutionary history? Subsequent analysis of many short tubulinlike sequences in our laboratory has shown them to be a common feature of these multigene families. Thus far, five P-tubulin processed genes have been fully sequenced, and we may therefore conclude that while the reintegration of mRNA sequences is a rare event, nonetheless the cumulative effect is a significant one. Two points are noteworthy in this regard. First, the generation of processed pseudogenes depends on the expression of the parental genes in germ line cells, otherwise the consequences of reintegration would not be inherited. Second, the site at which integration may occur must be restricted to positions in the chromosome such that there is no interference with the expression of other genes that are essential to the survival and proper functioning of the cell. Does the flow of information from mRNA to DNA serve any useful purpose, or are processed pseudogenes merely evolutionary accidents? There is no reason to believe that these sequences are expressed, since the upstream signals commonly associated with eukaryotic promoters are absent. In addition, comparison of the sequence data obtained from several processed pseudogenes reveals that the number and variety of lesions extends from few (e.g., 46P, Wilde er al., 1982a) to many (e.g., 7p, Gwo-Shu Lee et al., 1983), and may include point mutations, insertions, or deletions. This lack of conservation suggests that, as unexpressed sequences, these regions of DNA are not subject to evolutionary pressure, and drift in a manner that reflects the average rate of mutation. Eventually, such sequences might become unrecognizable with respect to their

TUBULIN ISOTYPES AND FAMILIES

165

“progenitor” gene-melting, as it were, into the genomic background. In this regard, it is interesting to note that in Southern blot experiments performed with a P-tubulin cDNA probe, the number of identifiable bands rise as the stringency of washing is lowered. Presumably this observation reflects the hybridization of increasingly mismatched genomic sequences. Thus, the contribution of unexpressed pseudogenes to the overall complement of the multigene family is a very significant one. In spite of the divergence of sequences no longer under selective constraints, several gene-like features-an initiation codon, stretches of open reading frame, the poly(A) addition signal4ould well be retained for periods that traverse species divergence. It is not inconceivable that such units might be recruited, via transposition or other gene arrangements, for the construction of other functional sequences.

c. DISSECTION OF THE MULTIGENE FAMILIES The extent of sequence conservation within the coding regions of tubulin genes is such that intact cDNA probes from whatever source cannot be used to distinguish either between genes of different species or, indeed, within a single species. In order to investigate evolutionary relationships within a multigene family, recourse has therefore to be made to probes that lie outside the coding region. This approach has proved successful because, in contrast to coding sequences, the untranslated regions appear to be under no selective constraints: with the exception of a rat and human a-tubulin gene (Cowan et al.. 1983), the 3’ untranslated regions of all tubulin cDNAs sequenced thus far show no evidence of significant homology. Any sequence homology in this region at the level of genomic DNA may therefore be taken as indicative of a close evolutionary relationship. An example of the use of a 3’ untranslated region probe to dissect the ptubulin multigene family is as follows. A probe was constructed by subcloning the entire 3‘ noncoding sequences of a human f3-tubulin cDNA. When this probe was used in Southern blot analysis of human genomic DNA, a greatly simplified pattern was obtained relative to that observed with the intact cDNA probe (Fig. 9). About four bands are identifiable per restriction digest and, since the probe is short and contains no site for any of the restriction enzymes used, this number must reflect the number of genomic sequences bearing sequence relatedness. Screening of two human genomic libraries with the 3’ untranslated region subclone has resulted in the isolation of recombinant fragments that together account for almost all of the bands observed in the Southern blot. A full sequence analysis of these genes showed three to be processed pseudogenes, with features consistent with their generation via mRNA intermediates (Fig. 8). The fourth sequence contained three intervening sequences, of 1880, 275, and 303 bp. Within the coding region of this gene, the sequence is completely homologous

166

N. I. COWAN AND L. DUDLEY

FIG. 9. Dissection of the human p-tubulin gene family using a 3’ untranslated region fragment subcloned from a cDNA probe. Track I , EcoR1; track 2, BarnHI; track 3, BglII; track 4, Hindlll. (A) Inact cDNA probe; (B) 3’ untranslated region probe. Arrows denote bands in common between the two blots. Data from Gwo-Shu Lee er al. (1983).

with that of the cDNA clone. Thus, this multigene subfamily consists of one expressed gene and three processed pseudogenes, all bearing sequence homology in their 3’-noncoding regions. Because of this relationship, it seems reasonable to conclude that the three processed pseudogenes were generated, on indepen-

TUBULIN ISOTYPES AND FAMILIES

M40

210

140

167

70

FIG. 10. Evolutionary tree showing time of integration of three related intronless pseudogenes into the host germ line. Data from Gwo-Shu Lee e r a / . (1983).

dent occasions, by the reverse transcription of mRNAs transcribed from a single functional gene. Comparison of the sequence data from the three processed pseudogenes reveals some interesting features. The number of genetic lesions-point mutations, deletions, and insertions-with respect to the parental cDNA sequence shows significant variation. Since the pseudogenes are almost certainly not expressed, the extent of sequence drift from the parental prototype may be taken as a measure of the time elapsed since the original integration events (Fig. 10).

IV. Tubulin Gene Expression Since tubulins are universally expressed in eukaryotic cells, an important question is whether they are synthesized at constitutive rates, or whether their synthesis is influenced by functional demand. Recent evidence suggests that the level of unpolymerized tubulin itself modulates the level of tubulin mRNAs, and hence regulates the de novo synthesis of protein (Ben Ze’ev et al., 1979; Cleveland et al., 1981b). Thus, tubulin synthesis decreases rapidly in the presence of drugs that depolymerize microtubules, such as colchicine and nocodazole, whereas synthesis remains essentially unchanged in the presence of drugs such as taxol or vinblastine that maintain microtubules in polymerized form (Fig. 11). The decrease in tubulin synthesis upon drug-induced depolymerization is correlated with a corresponding decline in the level of cytoplasmic tubulin mRNAs using tubulin-specific cDNA probes. Since no decline was detected in parallel controls performed using an actin cDNA probe, this result implies a short tubulin mRNA half-life. Experiments using actinomycin D to inhibit de novo mRNA synthesis suggest that this is indeed the case, though measurement of mRNA half-life using this drug are notoriously misleading. In any event, it appears

168

N . J . COWAN AND L. DUDLEY

4 f

2 p:

a

I

2

3

4

5

6

TIME (hr)

FIG. 1 I . Kinetics of loss of a-tubulin, P-tubulin, and actin mRNAs in CHO cells treated with colchicine. Dishes of CHO cells were incubated with 10 p.44 colchicine. At various times, cytoplasmic RNA was prepared and aliquots analyzed for their tubulin and actin mRNA content using - -0, P-tubulin; A-- -A,actin. Data from Clevespecific cDNA probes. -0, a-tubulin; 0land er al. (1981b).

likely that the level of nonpolymerized tubulin modulates the level of tubulin mRNA, most likely by affecting the rate of transcription of tubulin genes. Inexplicably, however, the synthesis of tubulin proteins in chicken fibroblasts does not respond to microtubule depolymerizing drugs in the same manner as cell lines from mammalian species, though the mRNA half-life appears to be equally short. In addition to regulation in response to unpolymerized tubulin levels, tubulin genes are differentially expressed during development. In Drosophila mefanogasrer there are four a-tubulin genes, all located at different sites on the third chromosome (Kalfayan and Wensink, 198 1). Each gene yields characteristic mRNA levels at different stages of development as judged by RNA blot analyses using 3' subcloned probes (Kalfayan and Wensink, 1982) (Fig. 12). Some of these differences are sex-specific. Indeed, a genetic approach has been used to demonstrate the testis-specific expression of a P-tubulin gene (Kemphues et al., 1979) that has multiple functions in spermatogenesis (Kemphues et al., 1!382). Thus, the products of this testis-specific gene are not restricted to a single functional class. 3' untranslated regions probes have also proved valuable for the analysis of mRNAs transcribed from chicken and human tubulin genes. In chickens, where, like Drosophila, there are four genes each for a-and P-tubulins, a 3'

FIG. 12. Developmental pattern of the transcripts from the four a-tubulin genes of Drosophila. Poly(A) mRNA from different developmental stages and from male and female adults was analyzed by RNA blot transfer, using four (a-d) different 3'-end probes isolated from genomic sequences. Lanes marked Embs: poly(A)+ mRNA from 0-3.3-6.6-9. and 9-12 hour embryos. L, Larval mRNA; P,pupal mRNA; M,male mRNA; F, female mRNA; A, total adult (male and female) mRNA; r,p. molecular size markers. Data from Kalfayan and Wensink (1982). +

170

N. J. COWAN AND L. DUDLEY

I

Minutes after Flagellar Excision

FIG.13. P-Tubulin mRNA concentration in Chlamydomonas reinhardii cells during the induction and deinduction of tubulin synthesis triggered by flagellar excision. Data from Minami er a / . ( I98 1).

end fragment from each P-tubulin gene detects a distinct subset of mRNA species, some of them of unexpectedly large size, with a complex pattern of expression among cell lines of diverse origin. It seems likely that, in at least one case, two mRNAs are transcribed from a single gene. This is certainly the case in an expressed human P-tubulin gene: here, use of adjacent sucloned 3’ end probes clearly detects either two mRNAs of 1.8 and 2.6 kb (probe A, Fig. 5 ) or only the large mRNA (probe 8, Fig. 5 ) . Thus, the large mRNA is distinguished from the 1.8 kb species in that it possess an extended 3’ untranslated region, and is generated by readthrough of a poly(A) addition site. This phenomenon appears to be not uncommon among eukaryotic genes, and at least in the case of P-tubulins, does not seem to affect the translational efficiency. In sea urchins, the existence of several expressed a-and P-tubulin genes may be inferred from the heterogeneous products produced following translation in a cell-free system of mRNAs selected using specific cDNA probes. The levels of at least some of these mRNAs are modulated during development (Alexandraki and Ruderman, 1981). A number of other lower eukaryotic species including Chlamydomonas, Naegleria, Tetrahymena, and Polytonella and have also been used as model systems for the study of tubulin gene expression. An attractive feature of these organisms is that growth of cilia or flagella can be stimulated following their amuptation. In vivo labeling studies have shown that this deflagellation procedure results in the stimulation of tubulin synthesis (Guttman and Gorovsky, 1979; Brown and Rogers, 1978; Lefebvre et al., 1978a; Fulton and Kowit, 1975). This increased protein synthesis is accompanied by an enhanced level of translatable tubulin mRNAs (Fig. 13) (Marcaud and Hayes,

TUBULIN ISOTYPES AND FAMILIES

17 1

1979; Lai et a f . , 1979; Minami et al., 1981). Thus, the induction of tubulin synthesis following deflagellation appears to result from enhanced transcription of tubulin genes, and not from activation of a pool of sequestered mRNA. In Chlamydomonas, the a-tubulin synthesiszed in vivo following deflagellation or in vitro in a cell-free system does not comigrate with mature flagellar a-tubulin on two-dimensional gels (Lefebvre ef al., 1978b), suggesting that some posttranslational modification occurs before a-tubulin can assemble into the flagella. This modification appears to be coupled to flagellar assembly (Brunke et a f . , 1982a), and suggests a close association between the modifying enzymes and the site of axonemal microtubule assembly.

V. Genetic Complexity and Functional Diversity Given that all cells express tubulin genes, and that the proteins they encode are highly evolutionarily conserved, the extent of structural diversity among tubulin gene families from various species is rather remarkable. In Chlumvdumunas. Drosophila, and chickens, for example, most if not all the sequences appear to be functional, and the four a- and P-tubulin genes in chicken may represent the minimum cellular requirement in vertebrates. By contrast, the mammalian species thus far studied have acquired a significant number of homologus but nonfunctional sequences by a mechanism involving an RNA intermediate. In addition, there is variation in the chromosomal arrangement of tubulin genes: in Trypanosomes, they exist as tandem arrays;, in sea urchins, some genes within the same family are clustered, but in other species examined to date, the gene families are dispersed. In view of the broad diversity of microtubule function, the means whereby tubulin gene expression is regulated during development and the cell cycle poses intriguing questions. The observation that multicellular organisms contain several expressed a- and P-tubulin genes entails the genetic potential for subunit diversity. On the other hand, two observations raise doubt as to the contributions of genetic variability to the synthesis of different kinds of microtubule within a single cell. First, in Drosophifa, Kemphues et al. (1979) have described a structural mutation in a P-tubulin gene whose expression is testis-specific. The mutant gene product is not restricted to one type of microtubule: it is found in the meiotic spindle, the cytoplasmic microtubular network, and the axonemal microtubule of the sperm tail. Second, in Aspergillus, a heat-sensitive mutation in a single P-tubulin gene has been described that has a wide-ranging effect on many microtubule-mediated functions (Oakley and Morris, 1981). This result also suggests that tubulin synthesized from a single gene transcript can be incorporated into functionally different microtubules within a single cell. However, differences between tubulins that are functionally distinct may be only very slight.

172

N. J. COWAN AND L. DUDLEY

For example, sequence evidence obtained from two human (3-tubulin genes reveals a short region of significant divergence near the C-terminus, and it is tempting to speculate that these two isotypes defined on the basis of nucleic acid sequence data represent proteins that can coassemble into functionally distinct microtubules. In the absence of established functional distinctions between tubulin gene products, the role of multiple expressed genes might be quantitative rather than qualitative, with differential modulation of gene transcription in response to cellular demand during either differentiation or the cell cyle. In any event, the precise relationship between genetic complexity and microtuble function remains unclear.

REFERENCES Alexandraki, D., and Ruderman, J. V. (1981). Mol. Cell. Biol. 1, 1125-1 137. Ben Ze’ev, A., Fanner, S . R., and Penman, S . (1979). Cell 17, 319-325. Brown, D. L., and Rogers, K. (1978). Exp. Cell Res. 117. 313-324. Brunke, K. J., Collis, P. S . , and Weeks, D. P. (1982a). Narure (London) 297, 516-518. Brunke, K. J., Young, E. E., Buchbinder, U., and Weeks, D. P. (1982b). Nucleic Acid R t x 10, 1295- I3 10. Cleveland, D. W., Kirschner, M. W., and Cowan, N. J. (1978). Cell 15, 1021-1031. Cleveland, D. W., Lopata, M. A., McDonald, R. J., Cowan, N. J., Rutter, W. J., and Kirschner, M. W. (1980). Cell 20, 95-105. Cleveland, D. W., Hughes, S . H., Stubblefield, E., Kirschner, M. W., and Varmus, H.E. (1981a). J. Biol. Chem. 256, 3130-3134. Cleveland, D. W., Lopata, M. A., Sherline, P., and Kirschner, M. W. (1981b). Cell 25, 537-546. Cowan, J. J . , Wilde, C. D., Chow, L. T., and Wefald, F. C. (1981). Proc. Narl. Acad. Sci. U . S . A . 78, 4877-4881. Cowan, N. J., Dobner, P. R., Fuchs, E. V., and Cleveland, D. W. (1983). Mol. Cell. Biol. (in press). Fulton, C., and Kowit. J. D. (1975). Ann. N.Y. Acad. Sci. 253, 318-332. Fulton, C., and Simpson, P. A. (1976). In “Cell Motility” (R. Goldman, T. Pollard, and J. Rosenbaum, eds.), pp. 987-1005. Cold Spring Habor Laboratory, Cold Spring Harbor, New York. Ginzburg, I., Behar, L., Givol, D., and Littauer, U. Z . (1981). Nucleic Acids Res. 9, 2691-2697. Guttman, S . D., and Gorovsky. M. A. (1979). Cell 17, 307-317. Gwo-Shu Lee, M., Lewis, S . A. Wilde, C. D.. and Cowan, N. J. (1983). Cell, 33, 477-487. Hall, J. L., Dudley, L., Dobner. P. R., Lewis, S . A., and Cowan, N. J. (1983). Mol. Cell Biol. 3, 854-862. Kalfayan, L., and Wensink, P. C. (1981). Cell 24, 97-106. Kalfayan. L., and Wensink, P. C . (1982). Cell 29, 91-98. Kemphues, K. J., Raff, R. A. Kaufman, T. C., and Raff, E. C. (1979). Proc. Narl. Acad’. Sci. U.S.A. 76, 3993-3995. Kemphues, K. F., Raff, E. C. Raff, R. A., and Kaufman, T. C. (1980). Cell 21., 445-451. Kraus, E., Little, M., Kempf, T., Hofer-Warbinek, R., Ade, W., and Ponstingl. H. (1981). Proc. Natl. Acad. Sci. U.S.A. 78, 4156-4160. Lai, E.Y., Walsh, C., Wardell, D., and Fulton, C. (1979). Cell 17, 866-878.

TUBULIN ISOTYPES AND FAMILIES

173

Lefebvre, P. A.. Silflow, C. D., Wieben, E. D., and Rosenbaum, J. L. (1978a). Cell 20,469-477. Lefebvre, P. A. Nordstrom, S. A., Moulder, J. E., and Rosenbaum, J. L. (1978b). J . CellBiol. 78, 8-27. Lemischka, I. R., and Sharp, P. A. (1982). Nurure (London) 300, 330-335. Lemischka, 1. R., Farmer, S., Racaniello, V. R., and Sharp, P. A. (1981). J. Mol. Biol. 150, 101- 120. Lopata, M. A . , Havercroft, J. C., Chow, L. T., and Cleveland, D. W. (1983). Cell 32, 713-724. Luduena, R. F., and Woodward, D. 0. (1973). Proc. Nail. Acud. Sci. U.S.A. 70, 3594-3598. Marcaud, L., and Hayes, D. (1979). Eur. J. Biochem. 98, 267-273. Minami. S., Collis, P. S., Young, E. E., and Weeks, D. P. (1981). Cell 24, 89-95. Mischke, D., and Pardue, M. L. (1982). J . Mol. Biol. 156, 449-466. Oakley, 8 . R., and Moms, N. R. (1981). Cell 24, 837-45. Osbome, M., and Weber, K. (1977). Cell 12, 561-571. Ponstingl, H., Kraus, E., Little, M., and Kempf, T. (1981). Proc. Nafl. Acud. Sci. U.S.A. 78, 2757-2761. Raybin, D., and Flavin, M. (1977). J . CellBiol. 73, 492-504. Sanchez, F., Natzle, J. E., Cleveland, D. W., Kirschner, M. W., and McCarthy, B. J. (1980). Cell 22, 845-854. Silflow, C. D., and Rosenbaum, J. L. (1981). Cell 4, 81-88. Snyder, J. A., and Mclntosh, J. R. (1976). Annu. Rev. Biochem. 45, 699-720. Thomashow, L. S., Milhausen, M., Rutter, W. J.. and Agabian, N. (1983). Cell 32, 35-43. Valenzuela. P., Quiroga, M., Zaldivar, J., Rutter, W. J.. Kirschner, M. W., and Cleveland, D. W. (1981). Nature (London) 289, 650-655. Wilde, C. D., Crowther, C. E.,Cripe, T. P., Lee, M. G.-S., and Cowan, N. J. (1982a). Nurure (London) 297, 83-84. Wilde, C. D.. Chow, L. T., Wefald, F. C., and Cowan, N. J. (1982b). Proc. Nurl. Acud. Sci. U.S.A. 79, 96-100.

This Page Intentionally Left Blank

INTERNATIONAL REVIEW OF CYTOLOGY, VOL. 85

The Ultrastructure of Plastids in Roots JEAN M. WHATLEY Botany School, Oxford University. Oxford, England Introduction . . . . . , . , ......................... ... . ... General Features , . . . . . . , . . , . , . . . . . . . . . . . A . The Root.. . . . . . ............................ .. ..... B. Plastid Development . . . . . . . . . . . . . . . . . . ......................... 111. Nongreen Roots., . . . A. The Forms and Distribution of Plastids in Roots of Seedlings with Particular Reference to P haseolus vulgaris. . . . . . . . . . . . . B. Plastids in Radicles of Embryos.. . . . . . . . . . . . . . . . . ........................ IV. Green Roots.. . . . . . . . . . . . . . . A. A z o h pinnata-the Ultrast Size, and Numbers of Plastids in Different Cell Files . . . . . .................. B. Other Aquatic Species.. . . . . . . . . . . . . . . . . . . . . . . . C. Plastid Dedifferentiation and the Rhizophores of Selaginella

I.

11.

.

V.

VI.

VII. V111.

IX .

X.

XI.

175 176 176 178 180 180 187 188 188 193

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

193 196 196

cal Changes (Triticum vulgare. Secale cereale, a Hybrid Triticale, and Lens culinuris) . . . . . . . C. Ultrastructural Changes in Convolvulus arvensis . . D. Ultrastructural Changes in Daucus carota . . . . . . , . . . . . . . . . . . Greening and Plastid Division. . . . . . . . . . . . . . . . . . . . . . . A. The Ultrastructure of Dividing Plastids . . . . . . . . . B. Plastid Numbers and Sites of Plastid Division. . . . ................... C. The Plastid Genome.. . . . , . . . . , . . Plastids in Sieve Elements.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Geotropism and Plastids in Root Caps. . . . . . . . . . . . . . . . . . . . . . . . . A. Amyloplasts as Geoperceptive Organelles . . . . . . . . . . . B. Plastid Distribution in Cells of the C. Plastids in Root Caps of Some Lower Plastid Pigments and Responses to Light. . . A. Protochlorophyllide, Chlorophylls, and ,......... B. Blue Light and Greening.. . . . . . . . . . . . Some Nonphotosynthetic Functions of Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ............................ References . . . . . . . . . .

196 200 20 I 202 203 204 205 206 208 208 209 21 I 212 212 213 214 216 217

The A. B.

I. Introduction Most of our information about the ultrastructure, function, and development of plastids has been obtained from work carried out on the leaves of angiosperm I75 Copyright 0 1983 by Academic Press. Inc. All nghts of reprcduclion in any form reserved. ISBN 0 - 1 2 - 3 w n 5 - 2

176

JEAN M. WHATLEY

seedlings grown either in the dark and subsequently transferred to the light or grown normally with a diurnal cycle of light and darkness (Kirk and TilneyBassett, 1978). The observations made during these investigations have provided a basic model for what is generally accepted as the “normal” plastid. Though the plastids in other plant groups and the plastids in other plant organs (including roots) more or less fit this standard model, their structure and patterns of development nevertheless show subtle differences, the significance of which is far from clear. Though plastids in aerial roots and in the roots of aquatic species often develop into photosynthetically functional chloroplasts indistinguishable from those in leaves, the plastids in roots which penetrate the soil (i.e., most roots) normally lack chlorophyll. The lack of a photosynthetic apparatus in underground roots is not, however, necessarily just the result of growth in darkness, for these roots do not always become green when they are exposed to light. Even when greening does take place, not all cells respond. Fadeel (1962) has estimated that, in the light, proplastids develop into chloroplasts in only 5% of wheat root cells whereas they do so in 80% of leaf cells. When etiolated leaves are exposed to light the etioplasts lose their prolamellar bodies and are rapidly transformed into chloroplasts, sometimes within a matter of hours, but in roots of the same plant the plastids (which lack prolammellar bodies) may take days or even weeks to develop into chloroplasts, if they do so at all. Furthermore the intensity and the wavelengths of light required to promote greening in roots are often different from those required for the greening of leaves. Thus the plastids in nongreen roots clearly behave differently as well as differ in structure from those in nongreen leaves. Neither the structure nor the function of plastids in roots has been rigorously investigated. Ultrastructural studies have been few and then generally limited either to plastids in the apical meristem and the root cap in seedlings or to the changes in plastid structure which take place when the roots (usually the excised roots) become green on exposure to light. Both the small size of most root plastids and their small numbers in most root cells have hampered biochemical investigations using isolated plastids. Thus we have little detailed knowledge of the ways in which root plastids resemble or differ from the “normal” (leaf) plastid and we know even less about how plastids vary in different types of cell within the root or in cells at different distances from the apical meristem.

11. General Features A . THE ROOT A root is usually described as a naked axis of which the only superficial appendages are root hairs; lateral roots and, in some species, root buds both

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

177

originate in tissue deep within the root, most commonly in the pericycle. The apical meristem of a root is subterminal and bidirectional, contributing cells both upward to the root proper and downward to a unique protective structure, the root cap. The main functions of a root are anchorage and the absorption of water and solutes, but roots can also have other physiological functions, e.g., as storage, aerating or supporting organs. The varied morphology of roots reflects these different functions. A young root comprises (1) the epidermis-usually a single layer of cells some of which may be extended to form root hairs; (2) the cortex-concentric files of parenchymatous cells limited toward the interior of the root by the endoderrnis; and (3) the vascular system-a cylinder of parenchymatous cells, within which lie alternating radial strands of phloem and xylem, the system being limited toward the exterior by the pericycle. The epidermis and the cortex may be sloughed off in some older roots. Depending on the physiological function of the root, different types of specialized cell (e.g., sclerenchyma and secretory ducts) may be formed within any of the three main zones. Most dicotyledons and gymnosperms have a root system comprising the primary root and lateral roots which develop acropetally from it. In perennial species these roots can undergo secondary growth. Most monocotyledons have an ephemeral primary root and the functional root system is derived from stem-borne adventitious roots; these do not undergo secondary growth (Esau, 1965). Thus roots may contain many different types of cell and may follow many different patterns of growth. Presumably the plastids show corresponding if less obvious diversity, but, with the exception of some work on sieve elements, the plastids associated with the many specialized types of root cell have never been seriously investigated. It has been postulated that roots evolved from the leafless axes of early land plants, but how this took place is obscure. The nonvascular mosses and liverworts and the most primitive group of vascular plants (the Psilopsida) lack roots and the latter give no evidence in their embryogeny of ever having possessed them. In the Psilopsida the physiological functions of roots are carried out by underground stems and rhizoids (Foster and Gifford, 1974). In other primitive vascular plants the primary root which develops from the radicle is usually short lived; secondary roots commonly arise endogenously near the shoot apex. Among members of the Lycopsida, these roots subsequently branch dichotomously; no lateral roots are formed. In prostrate species of Lycopodiurn the endogenous secondary roots enter the soil more or less directly, but, in upright species, they grow downward through the stem cortex and only emerge near its base. In the Lycopsid genus, Selaginella, secondary roots do not arise endogenously, nor do they grow downward through the stem cortex. Instead cylindrical structures called rhizophores arise superficially in angle meristems at points of branching near the shoot apex. Occasionally these rhizophores give rise to leafy shoots but, usually, they produce no leaves and grow down toward the soil.

178

JEAN M. WHATLEY

A brief description of the status of rhizophores is included here because the ways in which their plastids develop may provide a clue to the behavior of plastids in roots of less primitive plants. It was formerly believed that the aerial rhizophores of Seluginellu produced no cap and, for this and other reasons, the rhizophores were not considered to be roots. Rather it was believed that true, dichotomously branching roots with root hairs and caps were initiated only at the distal ends of rhizophores when these, after growing aerially for several centimeters, at last came in contact with the soil. Recent anatomical, developmental, and hormonal investigations suggest, however, that rhizophores may indeed be roots able to form both root caps and root hairs and not just the root-bearing organs that their name implies (Grenville and Peterson, 1981; Webster, 1969; Webster and Jagels, 1977; Webster and Steeves, 1963, 1964, 1967; Wochok and Clayton, 1978; Wochok and Sussex, 1974). Though no cap is normally formed in rhizophores of some species of Seluginellu (e.g., S . rnurtensii), a cap does form at a very early stage of rhizophore development in other species (e.g., S. densu, S. kruussiunu, and S . wullucei). In the latter species the apical meristematic cell begins to cut off initials for the cap when the rhizophore is less than 1 mm in length. The first dichotomous branching of the apex which will lead to the formation of future roots can be identified when the rhizophores of S. kruussiunu and S . wullucei are about 2 mm long, though this dichotomy is not visible externally until the rhizophore is some 450 mm in length and is usually just about to enter the soil. It is the series of differential responses by plastids in rhizophores growing (1) above ground, when the contain chlorophyll; (2) below ground, when they lose their chlorophyll; and (3) in moist containers, when they again lose their chlorophyll, which are of particular interest with respect to the behavior of plastids in roots of higher vascular plants, and which will be discussed in some detail below.

B . PLASTIDDEVELOPMENT Ultrastructural investigations, particularly those on leaves, but also those on other plant organs, suggest that in differentiating cells, the development of proplastids into chloroplasts always follows the same basic pathway. Within that pathway five successive stages of development (Fig. 1) can be distinguished (Whatley, 1977). These successive stages can be identified either as a temporal sequence during synchronous development of plastids in tissue becoming green or as a spatial sequence which can be followed along the length of individual cell files from a meristem toward mature green tissue, e.g., from the basal meristem to the tip of a grass leaf (Whatley, 1978; Wellburn, 1982; Wellburn et uf., 1982). This basic pathway of chloroplast development is subject to temporary or permanent modification or blockage. The stage of plastid development at the time of modification or blockage and the nature of the plastid response depend on

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

179

J

FIG. I . Stages of plastid development. Eoplast (Stage I t a small more or less spherical plastid containing dense stroma and, usually, a small fragment of thylakoid membrane but no grana. Amyloplast (Stage 2 t r e s e m b l e s an eoplast but contains starch which is not a direct product of photosynthesis. Amoeboid (pleomorphic) plasrid (Stage 3 t w h e n starch is lost from the Stage 2 amyloplast the plastid loses its spherical shape and becomes pleomorphic. Extension of the thylakoid system begins. Pregranal plastid (Stage 4 t t h e plastid usually assumes the discoid shape typical of a mature chloroplast. The thylakoid system becomes much more extensive and appears perforated; bithylakoids (incipient grana) are formed but not true grana. Alternative forms of Stage 4 can be reached in plastids of plants grown either in the light or in darkness. Leaf plastids in plants of Phaseolus vulgaris grown in the light contain chlorophylls a and b in the same ratio as in mature chloroplasts and photosynthesis takes place. Leaf plastids in plants grown in the dark lack chlorophyll and contain crystalline prolamellar bodies in addition to their quite extensive thylakoid system. Mature chloroplast (Stage S t t h e thylakoids lose their performations and true grana are formed. Further development is quantitative and is associated with an increase in extent of the thylakoid system and in the number and depth of stacking of grana.

the type of cell, the plant organ, the species, the physiological state of the plant, and on a number of environmental factors. Such responses are diverse and produce many variations in plastid structure (Whatley, 1977; Klein. 1982). They can result in the synthesis of one normal chloroplast component but not of another, e.g., grana may be formed but not the interconnecting thylakoids or stroma lamellae. Alternatively blockages may result in the accumulation of a

180

JEAN M.WHATLEY

variety of precursor materials, the form which the accumulation body takes then depending on the nature of the blockage and the period during development at which it occurs. When angiosperms are grown in darkness, for example, the development of leaf proplastids is blocked at a time when some thylakoid extension has already taken place and precursor materials accumulate as a prolamellar body, but in roots, where the blockage apparently takes place slightly earlier, thylakoid extension is minimal and no prolamellar bodies are formed. If the diverse processes of chloroplast development become unbalanced some precursor materials may be produced in excess of requirements and so accumulate for a time before being incorporated into the developing system, e.g., phytoferritin sometimes accumulates in quite massive amounts in eoplasts but later disappears as development proceeds (Whatley, 1978). Though it has in the past been suggested that plastid development is a unidirectional, linear progression leading from proplastids to mature and then to senescent chloroplasts, more recent investigations have shown that this is not necessarily so. Just as proplastids can develop into mature chloroplasts, so can mature chloroplasts dedifferentiate into proplastids, though the latter progression is less common. Depending on the particular aspect of the process which requires emphasis, the basic changes which take place in plastid structure are therefore best represented either as a cycle (Whatley, 1978; Thomson and Whatley, 1980) or as a projection of that cycle representing successive waves of differentiation and dedifferentiation. Diversions from the basic pathway may take place at any point within the cycle. These diversions may be minor or may lead to the development of other distinctive plastid forms such as etioplasts or chromoplasts. These in turn may be further modified to one or another of the five basic stages of development and so join the plastid cycle once again. Senescence, the breakdown of plastid structure often associated with general cellular degeneration, may set in at any point within the plastid cycle; it is not a process restricted only to mature chloroplasts. 111. Nongreen Roots A. THEFORMSAND DISTRIBUTION OF PLASTIDS IN ROOTS OF SEEDLIKGS WITH PARTICULAR REFERENCE TO Phaseolus vulgaris

The basic pathway of plastid development has recently been identified in primary roots of seedlings of Phaseolus vulgaris where it follows two separate spatial sequences (Fig. 2). The first of these sequences extends from the subterminal meristem upward into the root proper and the second downward into the root cap (Whatley, 1983). Thus the bidirectional progress of cellular differentiation within the root is paralleled by the bidirectional progress of plastid develop-

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

lo-

-

Root pro+

-

(2.5mm) -

181

I

I

i

5-

-

c I

FIG. 2. The sequences of plastid development in a root of Phaseolus vulgaris. 1. Eoplast; 1 ', dedifferentiated plastid = eoplast; 2, amyloplast; 3, amoeboid (plemorphic) plastid; 4, pregranal plastid; +, direction of plastid differentiation;- - 9 , direction of plastid dedifferentiation?; 0 ,level of earliest sieve element plastids (Whatley, 1983).

ment. The basic pathway of plastid development in seedling roots of the grass, Zea mays, follows a similar bidirectional course (Whatley, unpublished), though the root cap of Zea, like that of many other monocotyledons, has separate initials. In roots of Phaseolus vulgaris cv. Canadian Wonder the Stage 1 eoplasts in cells at the tip of the root proper and in adjacent cap cells are spherical or rodshaped (Fig. 3a). The cells which contain spherical eoplasts may more or less coincide with the quiescent center, a feature first identified by Clowes in 1954, but the precise boundaries within the Phaseolus root have not been determined. The eoplasts are sacs containing stroma, phytoferritin (Fig. 4), a few ribosomes, occasional single thylakoid fragments, and aggregations of membranes or tubules (Whatley, 1983). More or less similar aggregations have previously been described from plastids at different stages of development in several organs, including roots, and these have been given a variety of alternative names, e.g., tubular complexes (Newcomb, 1967), thylakoid complexes (Whatley, 1978), and prothylakoid bodies (Wellburn, 1982). Though generally similar to each other in structure these aggregations, which I shall call thylakoid complexes, show some minor variations, viz. those in roots of Phaseofusvulgaris cv Canadi-

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

183

FIG. 4. The distribution of phytofenitin and thylakoid complexes in a root of Phaseolus vulgaris. Small circles, phytofenitin; large circles, thylakoid complexes (Whatley, 1983).

an Wonder are smaller and less well organized and have fewer and less welldeveloped tubules than those observed by Newcomb (1967) in roots of a different cultivar of the same species grown under different environmental conditions. The significance of such subtle variations in structure is not known. What is certain is that thylakoid complexes have some features in common with, but lack the regular paracrystalline organization of, the prolamellar bodies characteristic of leaf etioplasts. Furthermore the thylakoid complexes in root plastids seem to be formed at a slightly earlier stage of development (Fig. 4) than true prolamellar bodies, at a time when no significant thylakoid extension has taken place. ~

FIG. 3. (a) Stage I , possibly dividing, eoplast with a thylakoid complex (t) in the root meristem of Phaseolus vulgaris (Section 1 of the root proper). X24,OOO. (b) Stage 2 amyloplast in the central root cap of Phaseolus vulgaris (Section 3 of the root cap). X20,OOO. (c) Stage 3 pleomorphic plastid at the periphery of the root cap of Phaseolus vulgaris (Section 4 of the root cap): this plastid contains a membrane-bound body (m)as well as starch. X21 ,OOO. (d) All seven of rhe plastid profiles in this micrograph are part of the same Stage 3 pleomorphic plastid in the outer cortex of a root of Phaseolus vulgaris (Section 12 of the root proper). X 10,000. (e) Stage 4 discoid plastid from the outer cortex of a root of Phaseolus vulgaris (Root hair zone): the extent of the thylakoid system is greater in plastids in this part of the root tban in any other zone surveyed. X30,OOO. (0 Stage I ’ eoplast from the lateral root zone: the plastid contains phytoferritin but no thylakoid fragments. X27,OOO.

184

JEAN M. WHATLEY

..'. .:. .'.

. ,' . ...

.. ..

.. ., . ... *:*. : '

,

,

,

..:

..- ' . . . : 0.

'.

y,: .

:.

. . ....

. . .

.:.

. . . .. , . . .( .. . . . .9

. ,

.

.

'. . !

.

yy .,.

.:-:::. .: 1.'. *

.

'

..*

..'

FIG. 5 . The distribution of starch and membrane-bound bodies in a root of Phaseolus vulgaris. Small circles, starch; large circles, membrane-bound bodies (Whatley, 1983).

The onset of the second stage of plastid development is marked by the accumulation of starch (Figs. 3b and 5), a process which, in the root cap of Phaseolus, reaches its peak in the central cells where the amyloplasts are known to act as statoliths (Fig. 6). Toward the periphery of the cap, in cells which will soon be sloughed off, the plastids begin to lose their starch; a similar loss has also been reported for other species, e.g., Medicago sariva (Maitra and De, 1972)and tomato (Street er al., 1967). In Phaseolus the plastids in the peripheral cells become pleomorphic in shape, i.e., they are entering Stage 3, the maximum stage of development achieved within the cap. Plastids at the periphery of the cap have an electron-dense stroma, some starch, a very restricted system of single thylakoids, and membrane-bound bodies (Figs. 3c and 5 ) . The sometimes crystalline but more commonly granular contents of membrane-bound bodies in plastids of roots and other organs are probably proteinaceous, and may include ribulose 1,5-bisphosphatecarboxylase (Sprey and Lambert, 1977), though other possible components have been proposed, viz. lipids, phenolic compounds, peroxidase, and polyphenoloxidase (reviewed in Hurkman and Kennedy, 1977; Thomson and Whatley, 1980). The pattern of starch accumulation and degradation which forms part of the spatial sequence of plastid development within the root cap (Fig. 6) is displayed as a temporal sequence (Fig. 7) by plastids in the quiescent center of Zea rnays

185

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

Starc area Per lastid profi (urn21

0.8

0.6 0.4 0.2 \\

4 3 2 1

5 6

Section no. Root cap

8

11

---

,---- \---------14 Root hair Lateral m

2

\\

Section number Root proper

Lano

;pa0

FIG.6. The distribution of starch in a root of Phaseolus vulgaris. The number of starch grains was counted in successive transverse sections of a root. Starch grains per plastid profile in cells of the inner cortex(& stele (0); root cap The area of starch per plastid profile outer cortex (0); (pm2) in the outer cortex (0); inner cortex (A); root cap (0).

(a).

while a new cap begins to form after the original cap has been removed (Barlow and Grundwag, 1974). Within the root proper of Phaseolus vulgaris, the successive stages of plastid development are found at greater distances from the meristem than they are in the root cap, and, at Stage 2, less starch is accumulated (Whatley, 1983). It is with the beginning of starch accumulation that variations between plastids in different cell files can first be distinguished. Starch appears, for example, nearest to the meristem and in greatest quantity in cells of the outer cortex, and farthest from

,q No. of starch grains

80

40 20 Houn after decapping

FIG.7. Starch accumulation in plastids of the quiescent center following decapping of a root of Zea mays. Compare with root cap in Fig. 6. (From Barlow and Grundwag, 1974.)

186

JEAN M. WHATLEY

the meristem and in smallest quantity in cells within the stele (Figs. 5 and 6). Street er al. (1967) found somewhat similar patterns of differential distribution of starch within cultured roots of tomato. The pleomorphic Stage 3 plastids in cortical cells of the root proper are considerably larger and much more highly pleomorphic than the Stage 3 plastids at the periphery of the root cap or in young leaves (Fig. 3c and d). Many Stage 3 plastids in the cortical and epidermal cells contain membrane-bound bodies and the distribution of these at different levels within the root (Fig. 5 ) further illustrate the differential development of plastids in different cell files. However such differential development may in part reflect the fact that cells of the outer cortex are developmentally older than those of the inner cortex (Esau, 1965). In his detailed investigation of root tips of Phaseolus vulgaris, Newcomb (1967) found that the proteinaceous membrane-bound bodies were often connected to subunits of thylakoid complexes and that both of these subsidiary structures and their interconnections sometimes had similar contents. He therefore suggested that the two structures might be developmentally as well as physically linked. It has, inter alia, been proposed that the contents of membrane-bound bodies may later be used during the extension of the thylakoid system, and that both membrane-bound bodies and thylakoid complexes may represent storage deposits which form following temporary (perhaps sequential) blockages or slowing down of the normal pathway of plastid development (reviewed in Thomson and Whatley, 1980). Casodoro and Rascio (1977) followed the development of membrane-bound bodies in roots and shoots of Atropa belladonna. They suggested that vesicles at the periphery of plastids and thylakoid complexes at the center (both of which had electron-dense contents) were the precursors of electron-dense deposits which initially lacked a surrounding membrane. Casodoro and Rascio also observed that in the shoots, but not in the roots, the membranes which later surrounded these bodies were in direct contact with grana. Farther from the tip of the Phuseolus root the plastids lose their pleomorphic form and become discoid (Stage 4 of development-Figs. 2 and 3e). At the same time there seems to be some extension of the thylakoid system, but this is difficult to measure reliably, particularly in the amoeboid plastids. However, the thylakoid system in plastids of Phaseolus roots never approaches the extent or the characteristicessentially spiral conformation of that in Stage 4 chloroplasts or etioplasts in the leaves (Whatley et al., 1982). In addition the single thylakoids in root plastids of Phuseolus (but not in Zea) appear to be less highly perforated than those in leaf plastids and have fewer bithylakoids. In the primary roots of Phuseolus, the most extensive, though still restricted, thylakoid system so far observed is in plastids of cortical cells at the level of root hair maturation. Farther from the tip at the level of the lateral roots, plastids in the cortical cells are spherical sacs containing electron transparent stroma and phytoferritin. These plastids lack even the smallest fragments of thylakoid and

187

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

c

vl

\

-

Cap

Meristem

Tip of root proper zone

zone

Oirections of cellular differentiati;

FIG. 8. vulgaris.

Stages in plastid development in the cap and in the outer cortex of a root of Phaseolus

they contain no starch, membrane-bound bodies, or thylakoid complexes (Fig. 30. If the spatial sequence of plastid development in Phaseolus roots is a true reflection of an earlier temporal sequence, then, after reaching their maximum (albeit limited) state of differentiation (as represented by the Stage 4 plastids in the zone of mature root hairs), the plastids must have undergone dedifferentiation, losing whatever thylakoids they formerly possessed and reverting to a form of eoplast (Fig. 8), as represented by the plastids in the lateral root zone.

IN RADICLESOF EMBRYOS B. PLASTIDS

The roots of Phaseolus described above were of dark-grown seedlings 5-7 days old. At this time, the Stage 4 plastids in the root hair zone of the primary roots are very different in ultrastructure from the Stage 4 etioplasts in primary leaves. Several reports in the literature suggest that there are no such major differences between the dedifferentiated plastids in different organs within embryos at the start of germination. Plastids in radicles of developing and mature embryos have been described for several species of angiosperm including Phaseolus lunatus (Klein and Ben-Shaul, 1966; Klein and Pollock, 1968), Pisum sazivum (Bong Yul Yoo, 1970), Secale cereale (Hallam, 1972; Hallam et al., 1972; Sargent and Osborne, 1980), and Zea mays (Deltour and Bronchart, 1971). In ripe seeds, plastids in the radicle contain stroma and osmiophillic globules but few ribosomes and almost no thylakoids or vesicles; phytoferritin is often present and starch gains occasionally so. Plastids in the radicle therefore closely resemble those in other parts of the embryo, including shoot apices and cotyledons. Following germination the numbers of plastid ribosomes in the developing leaves

188

JEAN M. WHATLEY

increase greatly whether the plants are growing in the light or in the dark, but in the roots the number of plastid ribosomes remains low unless the roots are exposed to light. In the developing embryo of Secale cereale, four zones can be distinguished within the radicle+ap, miristem, procortex, and prostele (Hallam, 1972). At this time the cap of the young radicle is itself protected by an ephermeral coleorhiza. Within I hour of the start of germination starch becomes common in plastids of the coleorhiza (Stage 2) but not in those of the root (Sargent and Osborne, 1980). After 6 hours the coleorhiza plastids are swollen with starch and these amyloplasts also contain polyribosomes. At this time starch is still uncommon in radicle plastids and only monoribosomes are present. Only later, as the root tip breaks through the degenerating coleorhiza, does starch accumulate in the cap of the young root, and at this time the cap takes over its usual role of protecting the root proper.

IV. Green Roots

In roots which penetrate the soil, the normal pathway of chloroplast development is blocked; chlorophyll is not synthesized and thylakoid extension is limited. By contrast, aerial roots and roots of aquatic species often appear green and some plastids at least develop into chloroplasts in much the same way as they do in leaves. A. Azolla pinWtU-THE

ULTRASTRUCTURE, SIZE, AND

NUMBERS OF

PLASTIDSIN DIFFERENT CELLFILES The only detailed study which has been made of plastid development in complete roots which are normally green is one based on the water fern, Azolla pinnara (Whatley and Gunning, 1981). This investigation made use of a magnificent series of electron micrograph montages of median longitudinal sections of roots of different ages which had previously been assembled and studied by Gunning and his associates (Hardham and Gunning, 1977; Gunning, 1978; Gunning er al., 1978a-c). FIG.9. (a) Small, immature chloroplasts with true grana but with a limited thylakoid system from a cell close to the apical cell in the zone of formative divisions in Azolla pinnara (Root 3; see text). x 12,500. (b) Large, mature chloroplasts with a well-developed system of thylakoids and grana from a cell of the outer cortex in the upper part of Root 3 of Arollu pinnara. X 12,500. (c) Sieve element plastid in a root of Phaseolus vulgaris. X30,OOO. (d) The amyloplasts are concentrated toward the distal wall of this central cap cell of Zea mays (courtesy of Dr.C. R. Hawes). (e) Part of the large single plastid in a root cell of Isoeres lacusrris showing starch grains and many osmiophilic deposits. x5OOO.

190

JEAN M. WHATLEY

The major advantage in using this species of Azollu is that the single apical cell divides in a precise manner to establish a small root within which the lineages of individual cells can easily be determined. The apical cell produces a helical sequence of daughter cells which, in turn, undergo a series of longitudinal formative divisions giving rise to easily distinguished concentric files of cells. The first of the formative divisions gives rise to an outer cell and an inner cell. Following another longitudinal division, the outer cell gives rise to the two outermost cell files of the root, the dermatogen, toward the exterior, and the outer cortex, toward the interior. Longitudinal division of the inner cell produces one cell which is the precursor of both the inner cortex and the endodermis and another cell from which the initials of the pericycle and the inner stele are derived (Gunning et al., 1978~).Subsequent transverse proliferative divisions provide additional cells within each file segment. Gunning uses the term merophyte, for each daughter cell or its derivatives (Gunning et al., 1978a). A complete gyre of the root is made up of three merophytes of which two are visible in longitudinal section. The older the merophyte, the greater its distance from the apical cell and the higher the number it is given in the accompanying figures. In all the roots investigated, the single apical cell contained small pleomorphic plastids which already had some thylakoids and a few true grana, so the state of plastid development in the apical cell was already more “advanced” than in any cells within nongreen roots. In the youngest roots examined and in the stele and the zone of formative divisions in older roots (Fig. 9a), the plastids were more or less discoid in shape but the thylakoid system was scarcely more extensive than in the apical cells. The numbers of plastid profiles present in each cell section suggested that in these cells plastid division had roughly kept pace with cell

G

6

8 10 Gyres

12

FIG.10. The numbers of plastid profiles in cells belonging to different cell files within an older root (root 3) of Arollu pinnaru. 0, Dermatogen; B. outer cortex; 0 ,inner cortex; 0. endodermis; pericycle. (From Whatley and Gunning, 1981.)

A,

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

Fic. 1 I .

..

6

4

8 10 Gyres

191

12

Thylakoid development in plastids of different cell files within root 3 of Azollupinnura. outer cortex; 0 ,inner cortex; 0, endodermis; A,pencycle. (From Whatley and Gunning, 1981.)

0, Dermatogen;

division. Farther from the apex, however, in the outer cell files of the proliferative zone in older roots, the numbers of plastid profiles per cell section, the numbers of bands of thylakoids (Figs. 9b, 10, and 11), and the length of the plastids all increased progressively as the age of the merophyte increased. The apical cell in the Azolla root divides about 55 times (Gunning et al., 1978a). The number of plasmodesmata laid down in the walls between the apical cell and its successive daughter cells declines as the root ages; this suggests that the symplast of the apical cell becomes progressively more isolated from the rest of the root. At the same time, the first sites of xylem thickening, root hair formation, starch deposition, and, within each cell file, particular transverse

t

,

;

,

,

,

,

,

,

,

,

,

,

1 2 3 4 5 6 7 8 9101112 Gyres

FIG. 12. The numbers of plastid profiles in cells of the outer cortex in roots of different ages show acropetal drift. 0 , Root 2 (oldest); Root 4: A, Root 6; Root 8 (youngest). (From Whatley and Gunning, 1981.)

A.

+,

192

JEAN M. WHATLEY

divisions as well as the attainment by plastids of a particular size and state of development all take place progressively closer to the apical cell (Figs. 12 and 13), i.e., these features all show acropetal drift (Gunning, 1978; Whatley and Gunning, 1981). Thus in the green Azollu roots, chloroplast development can be seen as both a spatial sequence extending from the apical cell upward through the root and as a temporal sequence which is reflected by plastids at particular sites within roots of different ages. Though cells of the dermatogen and outer cortex have a common mother cell, their plastids show independent trends in development; cells of the inner cortex and the endodermis also have a common parentage but their plastids similarly follow distinctive pathways. However, cells of the outer and inner cortex, in spite of their different parentage, have plastids which begin to follow a similar course of development soon after these cell files have become established (Figs. 10 and 11). Other cell features, too, show patterns of development characteristic of particular cell files, e.g., the areas occupied by groundplasm and by vacuoles, which may reflect the extent of cytoplasmic protein synthesis, and the area occupied by chromocenters, which may reflect nuclear activity (Barlow er ul., 1982). In the small roots of Azollu pinnara, the differential development within different cell files of plastids and other cell features is easy to observe and the precision with which development appears to be controlled is impressive. We still do not know the basis of this control or how control is exerted but the state of plastid development in a particular cell is clearly related to (1) the age of the root, (2) the file in which the cell lies, and (3) the distance of the cell from the root apex.

8

7

6 5 4 3 2 1 Root no. Increasing age' FIG. 13. Acropetal drift. This is indicated by the several plastid criteria shown; at the same time, first the number of plastid profiles in the apical cell declines. 0 ,First site of starch deposition; achievement by plastids of a mean of 3.0 thylakoid bands; A,first achievement by plastids of a mean length of 1.5 pm; I, the number of plastid profiles per apical cell section. (From Whatley and Gunning, 1981.)

+,

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

193

B . OTHERAQUATICSPECIES The root tips of several aquatic angiosperms were examined by Kawamatu (1967), Mollenhauer (1967), and Wroblewski (1973). In Lemna, Nvmphoides, Hydrocharis, Trapa, Najas, and Hydrilla the roots appeared slightly green; in Eichornia they were colorless or purple (although the root caps are green). Plastids in the roots of Lemna and Trapa showed the most extensive development of a photosynthetic apparatus; plastids in the roots of other “green” species contained few grana. Kawamatu reported the presence of prolamellar bodies in plastids of some species, but it is not clear whether these were true prolamellar bodies or merely thylakoid complexes of some sort. In Najas roots development of the thylakoid system was preceded by the lining up of vesicles; grana were formed later but interconnecting thylakoids were poorly developed. In Lemna the plastids in different parts of the root were at different stages of development. At the same distance from the apex there were proplastids in the central pith, welldeveloped chloroplasts in the outer pith, and immature to mature chloroplasts in the cortex. The nongreen root proper of Eichornia had plastids which contained vesicles and starch but few thylakoids. These plastids resembled those found near the tips of most nongreen roots but they were very different in structure from the large well-developed chloroplasts present in adjacent cells in that portion of the Eichornia root cap which forms a sheath round the tip of the root proper (Mollenhauer, 1967). Mollenhauer also observed large, mature chloroplasts with many thylakoids and grana in the root caps of the duckweed, Spirodella, and of the epiphytic orchid, Cattleya. In roots of most species, plastid development appears to be bidirectional (Fig. 2), but, in the root of the aquatic fern, Azolla pinnata, there is no downward gradient extending into the cap (Whatley and Gunning, 1981). Instead there is, as in the ensheathing part of the Eichornia cap, a discontinuity in plastid structure. In Azolla the small, amoeboid chloroplasts with a restricted thylakoid system present in the youngest merophytes of the root proper, contrast with the large, fully mature chloroplasts with an extensive thylakoid system present in all cells of the adjacent two-layered root cap. This disjunction may reflect the way in which the root cap in Azolla is initiated. The first division of the apical cell is periclinal. It is from the two products of the daughter cell thus formed that all future cap cells are derived (Gunning er a l . , 1978~).Subsequently the apical cell contributes cells only to the root proper. Thus the root cap is, in effect, the oldest merophyte, and it is perhaps not unexpected that its plastids have reached an advanced state of development similar to that of the plastids in the oldest merophytes of the root proper. C. PLASTIDDEDIFFERENTIATION AND THE RHIZOPHORES OF Selaginella martensii The fact that in Eichornia part of the root cap is green but the adjacent cells in the root proper are nongreen emphasizes the importance of factors other than

194

JEAN M. WHATLEY

light in the control of greening. The rhizophores of the lower vascular plant, Selaginella martensii, may contribute to our understanding of this subject. Unlike the aerial rhizophores of some other species of Selaginella, those of S . martensii normally lack a cap and root hairs. These rhizophores are conspicuously green in the zone of cell maturation, i.e., over much of their length. A second conspicuous green zone extends from 0.2 to 1.4 mm behind the apical initial (Webster and Jagels, 1977). Between these two green zones, in the region of cell elongation, the rhizophore is white or scarcely green. The most obvious components of plastids in the epidermal and subepidermal files of the green tips are osmiophilic deposits, some of which are spherical and some irregular in shape. The plastids also contain a few thylakoids and structures with regularly arranged interconnected tubules. These thylakoid complexes and the thylakoids themselves are both strongly osmiophilic. Indeed the staining properties of the plastids in the electron micrographs published by Webster and Jagels appear similar to the temporary “reversed” staining images shown by thylakoids in immature leaf chloroplasts of some angiosperm species during the early stages of granal formation (e.g., Platt-Aloia and Thomson, 1977). In the cortical cells of the green rhizophore tips the plastids are large and apparently lobed. There may well be only one plastid in each cell. Plastids in cells of the outer cortex contain grana and have a more extensive thylakoid system than plastids in the epidermal and subepidermal cells, but these plastids, too, appear to have a “reversed” staining image. Some of these cortical cell plastids may contain true prolamellar bodies. In cells of the inner cortex, the chloroplasts contain many starch grains as well as a thylakoid system which is more extensive than but lacks the “reversed” staining image of that in plastids of the outer cortex. Webster and Jagels do not describe the ultrastructure of plastids in other parts of the aerial rhizophores. This is a pity as information about plastids in the adjacent white and upper green zones might provide an interesting insight into their behavior during later stages of cellular differentiation. The apparent scarcity of chlorophyll in the white zone may merely be a dilution effect resulting from the dispersal of an unchanged chloroplast complement within an elongating cell or, alternatively, it may reflect a true loss of chlorophyll during a phase of plastid dedifferentiation. Conversely the green of the upper part of the rhizophore may be the result of either (1) redifferentiationof nongreen plastids, or (2) chloroplast growth (perhaps accompanied by division) increasing the total volume of the chloroplast(s) already in each cell. When the aerial rhizophores of Selaginella martensii are kept in the light but placed in moist containers, the tips begin to lose their green color. At the same time root caps and root hairs begin to develop. In these apparently colorless tips, the plastids of epidermal and subepidermal cells cannot be distinguished from those in the same files of the green tips. However plastids in the cortical cells

195

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

become smaller and undergo major structural modification. Grana become swollen and the thylakoids which formerly linked them become inconspicuousor disappear. Later, the now isolated grana are reduced to only a few swollen compartments, though they retain their osmiophilic character. These dedifferentiated plastids now closely resemble those in the epidermal and subepidermal files. Plastids in the inner cortex contain less starch than those in the same cells in green rhizophores. Webster and Jagels state that the modifications in structure of the cortical cell plastids during their loss of chlorophyll are similar to those reported by Thomson (1966) for plastids in ripening oranges during the early stages of their dedifferentiation and transformation into chromoplasts, except that osmiophilic (carotenoid) globules do not accumulate in the rhizophore plastids. Whether the two pathways of dedifferentiation are identical or not, it is clear that in both the rhizophore tips and in the ripening oranges loss of the thylakoid system and of chlorophyll is promoted by factors other than the absence of light. Rhizophores of Sefagineflamarrensii which enter the soil also lose their green color, but they do so much more rapidly than those placed in moist containers. Webster and Jagels make an interesting comparison between their own observations and those of Cormack (1937) on roots of the angiosperm pondweed, Elodea canadensis. In both papers a correlation is made between the absence of chlorophyll and the presence of root hairs, conditions which were promoted in Elodea roots (grown in the light) by the introduction of ethylene (Table I), a hormone produced not only by plants but also by bacteria in the rhizosphere. In rhizophores of Selaginella marrensii which enter the soil, just as in those grown in moist containers, the plastids in the cortical cells become smaller and are modified in structure, but the patterns of modification are quite different under the two environmental conditions. In plastids of rhizophores which penetrate the soil the grana do not swell; structures which seem to be true prolamellar bodies appear within and connected to the thylakoid network. Subsequently some starch is retained but the grana disappear and the prolamellar bodies inTABLE I THEPRODUCTION OF CHLOROPHYLL AND HAIRSI N ROOTSGROWN UNDER DIFFERENT CONDITIONS Species

Growth conditions

S. marrensii Light + air (rhizophores) Light + moist containers Dark + soil Light + water E . canadensis Light + water + (roots) ethylene Dark + soil or water

Chlorophyll Root hairs

+ -

-

+ -

-

-

+ + + +

Source Webster and Jagels (1977)

Cormack (1937)

196

JEAN M.WHATLEY

crease in size. A reduced system of single thylakoids radiates from the prolamellar bodies, as it does in leaf etioplasts. This apparently dark-induced pattern of plastid dedifferentiation resembles that described by Cran and Possingham (1973) in ripening fruits of avocado, at a time when light penetration through the blackening skin is severely limited. Both the absence of light and other factors (possibly hormonal) can promote dedifferentiation of chloroplasts. However, each factor may promote its own particular form of dedifferentiation.

V. The Greening of Roots A. GENERAL OBSERVATIONS

In an investigation using light microscopy, Powell (1925) examined the distribution of chlorophyll in intact roots of seedlings of 16 angiosperm species grown for 14 days under continuous illumination. She found chlorophyll in roots of 13 out of the 15 species which grew successfully (Table 11). Longitudinal sections showed that in most species chlorophyll developed throughout the length of the root to within 10-20 mm of the tip, but in Triticum vulgare and Hordeurn vulgare it was limited to the uppermost 15-20 mm. Within the roots chlorophyll was not always present in all cell files and the particular files which contained or lacked chlorophyll were characteristic for the species. Indeed in Ranunculus fiearia the pattern of chlorophyll distribution differed in the two types of root investigated. In two species, Triticum vulgare and Ranunculus fiearia (fibrous roots), the chlorophyll was restricted to a single cortical cell file immediately outside the endodermis. In a similar investigation carried out many years later, Fadeel (1962) found that in roots of Triticum aestivum (T.vulgare) cv. Eroica and in Hordeum vulgare after 7 days exposure to light, chloroplasts were restricted to the two innermost files of the cortex, but that within those cells the chloroplasts were numerous. In roots of Linum usitatissimurn, however, chloroplasts were present in all cortical cells, but each cell contained only a few.

B. ULTRASTRUCTURE AND PHYSIOLOGICAL CHANGES (Triticum vulgare, Secale cereale, A HYBRIDTriticale, AND Lens culinaris) Salema ( 197 1) has investigated chloroplast development during greening of intact roots of the grasses, Triticum vulgare (including cv. ardito), Secale cereale (including cv G23 40) and a hybrid Triticale derived from the two cultivars mentioned. In Triticale the first sign of potential greening in proplastids in root meristematic cells was an increase in the number of vesicles derived from the

TABLE I1

THE DISTRIBUTTON OF CHLOROPHYLL w LIGHT-GROWN INTACTROOTSOF SOMEANGIOSPERM SEEDLINGS~

Species

Single Stele extraendodexmal Inner Middle Outer Pith Rays parenchyma Pericycle Endodermis cell file cortex cortex cortex

+

Acer pseudoplatanus

+

Aesculus hippocastanum R m x sp. Vicia foba Pisum sativum Vicia sativurn Zea mays Bellis perennis Helianthus annuus Hordeurn vulgare Ranunculus ficaria Scilla nutans Triticum vulgare Fagopyrum esculentum Ranunculusficaria Ricinus communis Alliurn cepa

+ + + +

Only in two rows of cells above protoxylem groups

+ +

+

+

+ + + + +

-t

acornpiled from Powell (1925).

Comments

+

+ + + +

Seven varieties tested: all gave same results Outer stele only

+ + Fibrous roots Thick contractile mots

Tuberous roots Experiment failed-no

growth

198

JEAN M. WHATLEY

inner plastid envelope. Salema reported that these vesicles assembled into “prolamellar bodies” after 4-5 days of exposure to light, but the low magnification of the published micrographs makes it difficult to determine whether or not these are prolamellar bodies in the now accepted sense. However micrographs in a more recent publication by Oliveira (1982) on roots of Secale suggest that they resemble the thylakoid complexes formed in the light by plastids in developing leaves rather than true prolamellar bodies. Nevertheless, in Triricale root plastids, the extending system of single, perforated thylakoids radiates from the periphery of these complexes as it does from prolamellar bodies in leaf etioplasts. During these early stages of plastid development in the light there is an increase in number of both the stromal and the membrane-bound plastid ribosomes (Oliveira, 1975). After a short time, grana are formed and by the twelfth day of illumination, the root chloroplasts are said to be indistinguishable from leaf chloroplasts. Plastid development during greening of roots of Secale cereale followed a similar course to that in the hybrid Triticale (Salema, 1971). However, in Triticum vulgare, the other parent, no plastids were observed which contained “prolamellar bodies.” Oliveira (1982) calls particular attention to the possibility of genetic control of prolamellar body formation which is suggested by these observations. Salema and Oliveira both consider that extension of the thylakoid

Days

FIG. 14. Chloroplast development during greening of mots of Secale cereule. (A) Hill reaction (02pmol mg-1 chl h-1); (B) chlorophyll content (Fg g - ’ dry wt.); (C) protochlorophyll content (pg g-1 dry wt.); (D) stromal ribosomes; (E) membrane-bound ribosomes. (From Oliveira, 1982.)

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

I

24

48

1

72

199

I I

Hours

FIG.15. Chloroplast development during greening of roots of Lens culinaris. (A) Hill reaction I chl h- 1); (B) chlorophyll content (pg g - I dry wt.); (C) RUBC-ase (COz fixed pmo1/25 roots h-1). (From Lance-Nougakde and Pilet, 1965; Nato and Deleens, 1975a.)

(02pmol mg-

system may follow either one of two distinct morphogenetic pathways, the first, as in Tritium, depending entirely on vesicles derived from the inner plastid envelope where synthesis of new membrane takes place, and the other, as in Secale and Triticale, at the expense of the prolamellar bodies. A more detailed investigation of greening in roots of Secale cereale, grown at 25°C and exposed to light of 20,000 lux, has been carried out by Oliveira (1982). This provides data on features such as chlorophyll synthesis and the Hill reaction as well as ultrastructural information (Fig. 14). Somewhat similar information is available for the greening roots of the legume, Lens culinaris (Fig. 15), also grown at 25°C but exposed to light at only 4500 lux (Caporali, 1959; LanceNougarkde and Pilet, 1965; Nato and Deleens, 1975a,b). An interesting comparison between these two species is thus possible. Chloroplast development and chlorophyll synthesis took place much more rapidly, and the Hill reaction began earlier and worked faster in Lens than in Secale (Figs. 14 and 15). However, in both species the relative timings of the biochemical activities measured and particular states of plastid development appear to be quite similar. The more rapid synthesis of chlorophyll in Lens may be the result of the lower light intensities to which the roots were exposed. This is perhaps borne out by the equally rapid synthesis of chlorophyll at low light intensity in the roots of another legume, Lupinus albus (Mesquita, 1971), and by

200

JEAN M. WHATLEY

an investigation on excised roots of Convolvulus arvensis which showed that the chlorophyll content during greening was inversely proportional to light intensities within the range 3000 to 30,000lux (Heltne and Bonnett, 1970).

C. ULTRASTRUCTURAL CHANGESIN Convolvulus arvensis A detailed description of the ultrastructure of the quiescent center and the root cap of Convolvulus arvensis has been given by Phillips and Torrey (1974a,b). The structure of the plastids is essentially similar to that already described for Phaseolus vulgaris and, in the root cap, the characteristic distal positioning of amyloplasts in the elongated cells of the central columella is clearly illustrated. In excised roots of Convolvulus, cultured at 3000 lux, the rate of chloroplast development in the root proper differs in different types of cell (Fig. 16). Heltne and Bonnett found that though cells were fully differentiated within 10 mm of the root tip, it was only at a distance of about 40 mm that chloroplasts with grana first appeared. Chloroplasts were then found only in the cortex, the rate of development of the chloroplasts being the same within all cortical cell files. The mature chloroplasts which developed in the light were similar in structure to leaf chloroplasts. Plastids in dark-grown roots had a dense stroma and contained starch and membrane-bound bodies but no phytoferritin, the accumulation of which, in this species, is apparently light induced. Neither prolamellar bodies nor thylakoid complexes were observed in plastids of either light- or dark-grown roots. In Convolvulus, as in many roots exposed to light, greening takes place first in

e + d a t

E

2

c

0) >

''cO b

p a + lA

60 80 100 150 200 m m from root tip FIG. 16. Chlorophyll content and stages of plastid development in different types of cell during greening of cultured mots of Convolvulus arvensis. (a) Proplastids with few thylakoids; (b) proplastids with many irregular thylakoid configurations; (c) phytoferritin present; (d) early stages of granal development (Stage 4); (e) mature chloroplasts with true grana (Stage 5). 0 ,Inner and outer xylem parenchyma; M, pericycle; 0, chlorophyll content (pglg fresh weight) of roots cortex; grown in the light at 3000 lux. (From Heltne and Bonnett, 1970.)

0

A,

20

LO

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

20 1

fully differentiated cells at some distance from the root tip and subsequently progresses in the direction of the apical meristem (Gautheret, 1932; LanceNougarkde and Pilet, 1965; Salema, 1971). A similar acropetal pattern of development is found in dark-grown leaves of Hordeum vufgarefollowing their transfer to light, when chloroplast maturation proceeds from the most fully differentiated etioplasts in the upper leaf toward the proplastids of the basal meristem (Robertson and Laetsch, 1974). D. ULTRASTRUCTURAL CHANGES IN Daucus carota Orange roots of the domesticated carrot, Daucus carota, contain chromoplasts rather than colorless proplastids, but these chromoplasts, too, are transformed into chloroplasts on exposure to light. An early description of the ultrastructural changes which took place during chromoplast development in carrot roots was given by Frey-Wyssling and Schwegler (1965). They found that, as the roots developed, starch accumulated in the eoplasts, but that the starch later disappeared as carotenoid crystals were formed and the roots changed color. FreyWyssling and Schwegler suggested that “vacuoles” formed within the plastids as the stroma broke down, and that carotenoid crystals later formed in these large “vacuoles.” Later, Ben-Shaul e? a f . (1968) carried out a more detailed survey using newer and better fixation techniques. They reported that, for the first 2 or 3 weeks of growth, the root remained colorless; the small proplastids were circular in section and contained some thylakoids. During the next few weeks the roots expanded and became yellow to orange in color; in the plastids the thylakoid system became reduced in extent and the remaining thylakoids “broke down” to become swollen vesicles. These thylakoids seemed thicker, more electron dense, and apparently “rigid. ” The electron transparent “spaces” within the thylakoid sacs often appeared angular. It was concluded from isolation, recrystallization, and electron diffraction pattern studies that the carotenoid crystals were formed within the swollen thylakoid sacs rather than in “vacuoles” derived from the stroma. This conclusion has been supported by subsequent investigations on chromoplast development in carrot roots and in other tissues. Ben-Shad and Klein (1965) have also measured the increase in content of the a and p carotenes in carrot roots during the 20 weeks following germination. They found that initially a carotene was synthesized most rapidly but after about 6 weeks synthesis of p carotene became more rapid and the final concentration of p carotene was more than twice that of ci carotene. The ultrastructural changes undergone by the chromoplasts when carrot roots are exposed to light were investigated by Gronegress (1971). He found that all the chromoplasts in cells of the cortex were rapidly transformed into chloroplasts, but that this transformation took place at different rates in different cell files. After 48 hours exposure to light (8000 lux) the plastids in cells of the

202

JEAN M. WHATLEY

innermost cortex retained much of their chromoplast character, but in cells of the outer cortex the plastids contained carotenoid crystals which had become smaller and had acquired occasional wide irregular grana but only a few single thylakoids. In the subepidermal parenchyma the chromoplasts had lost their carotenoid crystals and had become transformed into chloroplasts with a somewhat limited thylakoid system, which nevertheless contained true grana interconnected by stroma lamellae. The gradient across the carrot root represented by these differential states of chromoplast transformation is further illustrated by differences in the chlorophyll content in these three zones; innermost cortex, 0.5; outer cortex, 10.5; subepidermal parenchyma, 48.0 pg/g dry weight.

VI. Greening and Plastid Division The amount of chlorophyll in a plant organ depends not only on the quantity of chlorophyll present in each plastid but also on the number of plastids in each cell. In the mesophyll cells in expanding leaves of most plants grown in the light, plastid division initially seems to keep pace with cell division but later exceeds it, so that the plastid population in each cell increases to a number which tends to be characteristic of the species (Whatley, 1980); mature leaf palisade cells of Phuseolus vulgaris contain about 40 plastids, whereas those of Spinacea oleracea contain 200-250. In most leaves the total amount of chlorophyll usually increases most during this period of rapid chloroplast division and for a short time afterward. There are, however, a few species, mostly tropical, in which the leaves become visibly green only after leaf expansion is virtually complete. In Theobromacacao, for example, the expanding leaves do not look green although their chloroplasts contain a normal complement of thylakoids and grana. However, each cell contains only three small chloroplasts and it is for this reason that the total chlorophyll content of the leaf is low (Baker et al., 1975). In roots which have become green following exposure to light the chlorophyll content as measured per gram dry or fresh weight is significantly lower than it is in leaves (Fig. 17). Electron micrographs of root chloroplasts show that they often have similar numbers of grana to leaf chloroplasts, though the extent of the stromal thylakoids and the numbers of granal compartments are frequently less than in leaf plastids. Many roots may resemble young leaves of Theobroma in that the low chlorophyll content reflects a low chloroplast population as much as a scarcity of chlorophyll in the individual plastids. Certainly when the large, highly vacuolated cortical cells of nongreen Phaseolus vulgaris roots are examined in the electron microscope many grid squares must usually be scanned before a single plastid profile can be found. Though counts of plastid profiles in sections from the immediate root tip of Zea mays indicate that plastid division takes place within this zone, and indeed that the rate of plastid division within

1liIN t

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

203

3

8c

2

8

4

a -

b

1

C -

FIG. 17. A comparison between chloroplasts in green roots and leaves of 13 day old seedlings of Secale cereale. (a) Grana per cell; (b) number of compartments per granum; (c) photosynthetic efficiency (mg 0 2 mg- I chl.); (d) chlorophyll (mglg fresh weight); (e) total carotenoids. (From Oliveira, 1982.)

I

this selected area (which includes the quiescent center) is greater than the overall rate of cell division (Whatley, unpublished), we do not know how far upward through the root proper plastid division continues in dark-grown roots. (The presence in electron micrographs of constricted plastids is not adequate evidence that plastid division is being completed.) Light is known to promote plastid division in leaves. It is therefore probable that when roots are induced to green on exposure to light, plastid numbers increase [and indeed this has been shown by Heltne and Bonnett (1970) for Convolvulus arvensis], but neither the lag period likely to be involved nor the extent to which plastid division is promoted has ever been investigated. A. THEULTRASTRUCTUREOF DIVIDING PLASTIDS

It has long been assumed, and has now been established (Chaly et al., 1980), that mature chloroplasts in higher plants replicate following constriction. A similar division mechanism appears to operate in proplastids (Chaly and Possingham, 1980; Whatley, 1980). It is probable that, as division proceeds, the constriction between the two plastid halves becomes progressively narrower until, finally, the two daughter plastids become separated. During the final stages of division of mature chloroplasts, small osmiophilic deposits or plaques are commonly found associated with that part of the plastid envelope which lies within the constriction (Suzuki and Ueda, 1975; Leech er al., 1981). A recent survey of root apices of nine species of angiosperm and one species of conifer shows that dividing proplastids, too, have osmiophilic plaques and confirms that these plaques are part of an annulus which encircles the constriction (Chaly and Possingham, 1981). In most of the species examined the annuli in the dividing root proplastids were 15-40 nm wide but those in proplastids of tomato roots

204

JEAN M. WHATLEY

were 40-70 nm wide. The constrictions themselves varied in width from 50 to 400 nm within each sample, suggesting that the root proplastids do not divide in synchrony. In thin sections of mature leaf chloroplasts the constriction associated with division is usually central; in the root proplastids examined by Chaly and Possingham and in proplastids and immature chloroplasts in leaves of Phaseolus vulgaris (Whatley, 1980; Whatley ef al., 1982) the constriction often appears to be off center. It was suggested for Phaseolus that asymmetrical division might take place when polarized elongation growth of the plastid was out of phase with the rate of plastid division.

B. PLASTIDNUMBERS AND SITESOF PLASTID DIVISION Chaly and Possingham found that constricted plastids were most common in the root cap and common in the cortex and stele near the tip of the root proper. Following serial sectioning, no constrictions suggesting division were observed in plastids in epidermal cells, in amoeboid plastids, or in amyloplasts of the central root cap. In dark-grown roots of Phaseolus vulgaris constricted plastids have been seen in the apical meristem and throughout thecortex of the root up to the level of the root hair zone, but not in cortical cells of the lateral root zone or in the cells at the center of the root cap (Whatley, unpublished). Almost no information is available about the numbers of plastids in different types of cells in dark-grown roots. An early investigation of plastid numbers in the root cap and root tip of Zea mays was made from montages of electron micrographs (Clowes and Juniper, 1964; Juniper and Clowes, 1965) but no further work of this type seems to have been carried out. Using median longitudinal sections of resin embedded roots of Zea mays generously given to me by Dr. C. Hawes, I was able to prepare new montages of electron micrographs which I used to count the number of plastid profiles in cell sections within five adjacent files extending through the center of the root from 12 tiers above (i.e., toward the stele) to 12 tiers below the cap junction (Fig. 18). Counts were limited to these 24 tiers for, within them, the plastids were more or less spheroidal, or slightly rod-shaped or just beginning to become amoeboid, and so not subject to the complication of overestimation of plastid number which results when several profiles of the same highly pleomorphic plastid are included within one cell section. Furthermore, within the selected tiers, the cells of the root proper and of the upper part of the cap were of approximately the same size; cells of the central cap were larger but the distally concentrated plastids occupied a volume of cytoplasm not markedly different from that throughout which the plastids were dispersed in the other cells. It is therefore reasonable to assume that the average counts of plastid profiles per cell section within the five files provide a rough but direct correlation with the total number of plastids present in each cell (between two and three times the number of plastid profiles). The results obtained in this

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

205

No. plastid profiles

I

D :I C

'D

I

I D

:

4 Stages of development Plastid distribution

120

_______ ________ 1st -Cycle?

--;) 4-

- - - -- - - Basic plastid 4

12

8

8 4 Cell tiers

Cell tiers

COP

Stele

12 I

complement

survey are in general agreement with the earlier estimates of plastid numbers by Juniper and Clowes for root tips of the same species. Figure 18 suggests that plastid division takes place in both the cap initials and in the root proper. It also suggests that within the 12 tiers of the root proper included in the survey, the plastids undergo two cycles of division whereas plastids undergo only one division cycle within the 12 tiers of the cap. No division takes place among the displaced statoliths of the central cap, but both the counts of plastid profiles and the presence of constricted plastids in the electron micrographs point to plastid division taking place in cells of the quiescent center. However, none of the observations made here provides any information about comparative rates of plastid division in different parts of the root tip. Furthermore it should be noted that though Fig. 18 indicates that the average numbers of plastid profiles from the five central cell file sections show sharp increases which suggest cycles of plastid division, the plastid populations within these cells do not divide in synchrony. It is also apparent that the phasing of plastid division, the stage of plastid development, and the distribution of plastids within the cell are all features which are independent of each other.

C. THE PLASTIDGENOME Meristematic cells at the base of spinach leaves 2 cm long contain 10-15 plastids, each with about 200 copies of their genome (Scott and Possingham, 1980). These plastids divide but there is no significant synthesis of plastid DNA

206

JEAN M.WHATLEY

after the time when the young leaf cells contain approximately 25 chloroplasts which together have about 5000 genome copies. As these chloroplasts continue to divide the available DNA is partitioned between successive pairs of daughter plastids. By the time the leaves are 10 cm long each cell contains 150-200 chloroplasts each with about 30 copies of their genome. The DNA in root proplastids initially appears to be much less than in leaf proplastids. In the tips of spinach roots in cells containing about 10 proplastids, each plastid has been estimated to contain only about 10 genome copies. When spinach roots are fed with [3H]thymidine,one or two silver grains were found in each plastid profile within the cortex and stele after 60 minutes and three or more grains after 24 hours (Possingham er al., 1983). In these proplastids 75% of the grains were close to the plastid envelope; few were near the infrequent thylakoids. By contrast in mature leaf chloroplasts most of the silver grains are associated with the thylakoid system and few are seen near the plastid envelope. In the root cap, amyloplasts were labeled after 24 hours but not after 60 minutes. The silver grains in the cap amyloplasts were associated with small areas of stroma or with the few thylakoids but not with the starch grains or with the plastid envelope.

VII. Plastids in Sieve Elements Of the publications concerned with the ultrastmcture of plastids in roots, most restrict their observations to plastids in the cortex of the root proper: very few indeed consider plastids in other types of cell. Indeed only sieve elements and the root cap have received any significant attention. Plastids in mature sieve elements of angiosperms are diverse in structure and these plastids are unusual in that their structural differences are taxonomically distinctive (Behnke, 1972). Though plastids in all organs and in all types of cell generally follow the same basic pathway of development, the proplastids and mature chloroplasts in roots, for example, are usually not identical in structure to those in the leaves of the same plant. However, in any one species, sieve element plastids in all organs seem to develop in the same way from the eoplast stage add, at maturity, have the same distinctive characteristics. Behnke divides sieve element plastids into two main groups, S-type plastids, containing at maturity a sieve element starch which often appears coarsely granular (Fig. 9c), and P-type plastids, which may contain starch but, in addition, contain proteinaceous filaments or cuneate or crystalline inclusions that are not membrane bound. Both S-type and P-type plastids are found among dicotyledons and in gymnosperms (Evert, 1977); S-type plastids have not been reported in any monocotyledon. During their development, sieve element plastids become spheroidal in shape;

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

207

any thylakoids which were present disappear; the plastids accumulate the appropriate starch and proteinaceous deposits within the finely granular stroma; the stroma itself becomes progressively less electron dense until, by maturity, it seems to have disappeared; the plastid envelope becomes wavy in outline. Ultrastructural investigations of sieve element differentiation in roots which describe plastid development include those on the dicotyledons Gossypium hirsurum (Thorsch and Esau, 1981), Nicoriana rabaccum (Esau and Gill, 1972), and Phaseolus vulgaris (Esau and Gill, 1971) and on the monocotyledons Allium cepa (Esau and Gill, 1973), Lemna minor (Melaragno and Walsh, 1976), Zea mays (Walsh, 1980), and several palms (Parthasarathy, 1974a-c; Parthasarathy and Klotz, 1976). In roots, the sequential development of sieve elements and their plastids can be followed along the length of selected cell files. The crystalloid inclusions (subunit spacing 80A) appear in the sieve element plastids of Lemna minor roots at a very early stage of differentiation (Melaragno and Walsh, 1976). Indeed in Allium cepa the appearance of crystalline inclusions is the first visible indication that a cell will develop into a sieve element (Esau and Gill, 1973). Esau and her associates have reported on the distribution of plastids within individual sieve elements. In immature sieve elements of Gossypium hirsurum (Thorsch and Esau, 1981) the plastids are dispersed throughout the cytoplasm but in older cells the plastids move to the periphery of the cells often near the end walls; by contrast, mitochondria in these cells tend to move toward the lateral walls. Later some mitochondria, endoplasmic reticulum, and plastids appear to be aggregated around the disintegrating nucleus. In roots of Nicoriana rabaccum the distribution of sieve element plastids seems to be particularly precisely controlled (Esau and Gill, 1972). Even in young sieve elements most plastids lie near the end wall adjacent to the next older cell, i.e., the plastids occupy a site which is the mirror image of that occupied by the amyloplasts in the central cells of the root cap. This polarized distribution of sieve element plastids is very clearly illustrated in Figs. 1, 3, and 4 of the publication of Esau and Gill (1972). No such polarized distribution is shown by plastids in adjacent files of other types of cell in this species, nor does a polarized distribution appear to be shown by sieve element plastids in the other species investigated. In Allium cepa, for example, the published electron micrographs show sieve element plastids adjacent to the side walls and to both end walls (Esau and Gill, 1973). The means by which the distribution of plastids within cells is controlled is not known. Most investigations on plastids in the sieve elements of roots concentrate on the ways in which they develop or on their state when they reach maturity. Parthasarathy (1974~)observed that, in some palms, sieve elements could be long lived, that their mitochondria could persist for 3-6 years, depending on the species, and that plastids could remain for an even longer period. In old sieve elements, the plastids appear to have degenerated, the plastid envelope often

208

JEAN M.WHATLEY

appearing dilated and the stroma being electron transparent. The crystalline and other inclusions also broke down or disappeared. Parthasarathy points out that though some of these responses may be due to fixation injury, light microscopic evidence also suggests that the plastids do, indeed, degenerate.

VIII. Geotropism and Plastids in Root Caps It is more than 80 years since it was first suggested that plants perceive gravity by means of heavy sedimenting bodies, the statoliths. However the mechanism for geotropism is known only for the unicellular rhizoids of the green alga, Chara (Schroter er al., 1975). In Chara the statoliths are large vesicles containing barium sulfate crystals. These inert statoliths provide a gravity-dependent barrier to the movement of Golgi vesicles toward the growing tip of the rhizoid. The direction of growth of the rhizoid is controlled by the symmetry or asymmetry of release of the contents of the Golgi vesicles at the tip. In land plants many different organs, including roots, show geotropic sensitivity but the means by which geotropism is accomplished is uncertain, though it is assuredly different from that in Chara. The subject of geotropism in land plants has been discussed at length in the literature; only those aspects of it relating to plastids in the root cap will be considered here. A. AMYLOPLASTS AS GEOPERCEFTIVE ORGANELLES

The sensing organelles for geotropic response in land plants are generally believed to be amyloplasts and these are thought to function only in conjunction with other organelles. Displaced dictyosomes (McNitt and Shen-Miller, 1978), endoplasmic reticulum (Sievers and Volkmann, 1977; Hensel and Sievers, 1981; Olsen and Iversen, 1980), or endoplasmic reticulum and plasmodesmata operating together to form a multiple valve system (Juniper, 1976, 1977) are among the organelles which have been proposed to operate with the amyloplasts. The difficulty in finding out how the geotropic response is effected in land plants is accentuated by the short time (a few seconds) required for presentation to the stimulus and by the fact that the zone of perception can be many cells re:moved from the zone of response. In the roots of many land plants the main zone of gravity perception is the columella, a core of cells situated in the center of the root cap and acting together. This core of cells has been estimated to be 2000 in Zea muvs (Clowes, 1976), 250-600 in Pisum sativum (Olsen and Iversen, 1980), but only 96 in lateral roots of Nasturrium amphidium (Haberlandt, 1914); in lateral roots generally, geotropic sensitivity is low or absent. The sensitive cells contain more or less spheroidal amyloplasts (Stage 2 of development) (Figs. 2, 3b, 5 . 6, and 9d)

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

209

which are usually packed with starch grains. This starch is unusual in that it is highly resistant to digestion by artificial means (Audus, 1962); nevertheless there is rapid turnover of starch within these amyloplasts (Northcote and PickettHeaps, 1966). When starch is removed from the amyloplasts of the central cap of Lepidium sativum as a result of treatment with gibberellic acid and kinetin, the roots are no longer able to respond to geotropic stimulus (Iversen, 1969). Although plastids in peripheral cells of the root cap also commonly contain considerable amounts of starch, the quantity is less than in plastids in the central cap (Figs. 3b and c and 6) and the earlier capacity of these peripheral cells (formerly central cells) for geoperception has been lost. When the cap of a maize root is removed a new cap forms, following reactivation of mitotic activity in the quiescent center (Juniper et al., 1966). The eoplasts in the quiescent center undergo a temporal sequence of developmental changes, including starch accumulation and loss (Barlow and Grundwag, 1974; Barlow and Sargent, 1978), which parallel those in the spatial sequence of plastid development in the normal root cap (Figs. 6 and 7). In maize, starch accumulation begins within 3 hours of decapping, geotropic sensitivity is first apparent after 14 hours, but a new cap is not regenerated for 3 to 4 days (Barlow, 1974). Thus the presence of a morphologically identifiable cap is not a prerequisite of geotropic response in rootshdeed Haberlandt ( 1914) found that the starch-containing elongation zone of the cortex in roots of Viciafaba, Phaseolus multi’orus, and Lupinus albus all showed some geotropic sensitivity. B. PLASTIDDISTRIBUTION IN CELLSOF THE ROOT CAP In geotropically sensitive roots the amyloplasts in the central cells of the root cap are not randomly distributed but are generally concentrated toward the distal transverse walls (Juniper and French, 1970; Phillips and Torrey, 1974b). When roots are turned from the vertical to the horizontal, these amyloplasts are displaced and eventually become concentrated toward the former longitudinal wall which now forms the base of each cell. Both the Stage 1 eoplasts in the cap initials and the Stage 3 amoeboid plastids toward the periphery of the cap (Fig. 2) remain dispersed throughout their cells, as do the Stage 2 amyloplasts in the subapical cortex of the roots of most angiosperm species. Though the distal displacement of amyloplasts is a regular feature of geoperceptive cells, this displacement alone does not ensure a geotropic response. In Convolvulus arvensis (Tepfer and Bonnett, 1972) and in Zea mays (Shen-Miller, 1978), for example, the geotropic behavior of the roots is influenced by light (a phytochrome reaction). Roots show a positive orthogeotropic response only after exposure to red light even though the statoliths are distally displaced in roots grown both in the light and in the dark. Tischler ( 1905) investigated several angiosperm species with permanently

TABLE Ill

GEDTROPISM AND

Species Selaginella martensii (aerial rhizophore) Selaginelka martensii (aerial rhizophore in moist container) Sekaginella kraussiana (aerial rhizophore) Selaginelka kraussiana (underground root) Isoetes macrospora Equisem arvense Ophioglossum petiolarum %a., not applicable.

STARCH IN SOME LOWER VASCULAR h N T S

Single Cap Starch apical Cap Dichotomous Starch starch in Geotropic cell present branching in cap displaced cortex

Cortical starch displaced

Numberof plastids per cell

+ +

+ +

-

+

-

n.a.O

+

+

1

+

+

-

n.a.

+

?

1

+ + +

+ + + + +

+ + + + +

+ + +

-

n.a.

+

1

-

n.a.

+ +

?

1

+ + +

-

?

? ? ?

JO

+? +?

-

+ +

? ?

1

=

Principal souxe Webster and Jagels (1977) (Figs. 7, 10, and 11) Webster and Jagels (1977)

Grenville and Peterson (1981) (Fig. 4) Grenville and Peterson (1981) Peterson er al. (1979) Foster and Gifford (1974) Peterson and Brisson (1977)

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

21 1

ageotropic roots. These included some orchids and climbers with aerial roots, some aquatic species, including Eichorniu, and some species with subterranean roots. In these ageotropic species the cells of the central root cap either had plastids which lacked starch or had amyloplasts which were distributed throughout the cell. Tischler also found that in roots of Festucu ovinu and Pou spp. which were initially ageotropic, the appearance of starch in the caps coincided with the acquisition of geotropic sensitivity. In aroids the geotropic “nutritional” aerial roots were well provided with amyloplasts but the only slightly geotropic “grasping” roots had a reduced statolith system. The negatively geotropic “breathing” roots of some species had a normal statolith system. The distal concentration or displacement of amyloplasts within the central cells of the root cap is characteristic of many angiosperm species and of the few gymnosperm species which have recently been examined (Hestnes and Iversen, 1978; Peterson and Vermeer, 1980). However some lower vascular plants show some interesting variations from this pattern (Table 111).

C. PLASTIDS IN ROOTCAPSOF SOMELOWERVASCULARPLANTS Haberlandt (19 14) refers to the absence of starch from root caps of Seluginellu murrensii but to its presence in distally displaced amyloplasts in cells of the inner cortex of the root proper. More recent investigations have shown that some other species of Seluginellu also have root or rhizophore caps which lack starch (Table 111). The thickened, transverse, distal walls and osmiophilic globules have both been suggested as possible alternative sites but as earlier suggested by Haberlandt, the distally displaced amyloplasts of the cortex seem the more probable sites of geoperception. No species of Lycopodium seems to have been examined, but in another lycopod, Zsoeres macrosporu (Fig. 9e), the large, usually single plastid in each central cell of the cap contains starch. These amyloplasts are, however, amoeboid rather than spheroidal in shape and are positioned close to and often curving round the nucleus rather than adjacent to the distal transverse wall (Peterson et al., 1979). The investigation of Isoeres was not extended to the root proper and so information about the presence or otherwise of a possible geoperceptive starch zone in the cortex is not available. Among ferns, only the root cap of Ophioglossum petiolutum has been the subject of recent investigation (Peterson and Brisson, 1977). In Ophioglossum even the plastids in the single apical cell contain starch. Each plastid commonly contains a single, unusually elongated grain, a characteristic feature also of plastids in young fronds of the species. The root amyloplasts are distally displaced even in the proximal cells of the columella and in these cells also the plastids become amoeboid. One of the electron micrographs published by Peterson and Brisson suggests that toward the outer edge of the columella the plastids may contain a few grana, but this restricted thylakoid system, like the starch,

’

212

JEAN M. WHATLEY

appears to be lost toward the periphery of the cap and at the same time osmiophilic globules increase in number. These structural changes suggest plastid development in the root cap of Ophioglossum petiolatum proceeds not only to the mature Stage 5 chloroplast as it does in aquatic species like Eichornia (Mollenhauer, 1967) and Azolla pinnata (Whatley and Gunning, 1981), but that toward the outer edge of the cap the plastids may, uniquely, undergo dedifferentiation.

IX. Plastid Pigments and Responses to Light A. PROTOCHLOROPHYLLIDE, CHLOROPHYLLS, AND CAROTENOIDS

Using the characteristic red fluorescence as an indicator, Hejnowicz (1958) found protochlorophyllideto be present in the root meristems of all 16 species of plants (ranging from ferns to angiosperms) which he investigated. Just as Powell (1925) had found that during greening of seedling roots, the distribution of chlorophyll in different types of cell was characteristic of the species, so Hejnowicz found some species differences in the distribution of protochlorophyllide, though in most species the greatest red fluorescence response was shown by cells in the cortex. The state of protochlorophyllide in dark-grown roots seems to differ from that in etioplasts of dark-grown leaves (Bjom, 1976, 1980). In primary seedling roots of Zea mays the protochlorophyllideis mostly in the esterified form, though there is some unesterified protochlorophyllide at the extreme tip. In these roots the protochlorophyllide is slowly converted into chlorophyllide in red or blue light but, unless the intensity of the light is very low, the chlorophyllide is then destroyed (Bjom, 1963, 1976). In seedling roots of wheat and in the adventitious roots of maize (but not in the primary roots) synthesis of protochlorophyllide and chloroplast development are both triggered by light. The greening of intact roots generally takes place after a considerably longer lag period than it does in etiolated leaves. Once the roots are green, however, they are capable of normal photosynthesis. The low chlorophyll content of root plastids is reflected in the low rates of photosynthesis in roots (Fadeel, 1963; Oliveira, 1982). However, the photosynthetic assimilation number in roots is very high compared with that in green leaves but is similar to that found in yellow leaves of “golden” varieties of some species. The high photosynthetic efficiency in root plastids (Fig. 17) may also be increased by the presence of CO, concentrations in the intracellular spaces, which are significantly higher than those in leaves. These observations of course indicate that the complete photosynthetic apparatus has been synthesized in green root plastids. The chlorophyll a:b ratio appears to be similar to that in leaf plastids (Fadeel, 1962; Oliveira, 1982) but the carotenoids show some differences. Fadeel (1962) found that in

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

213

flax and wheat and Oliveira (1982) found that in rye (Fig. 17) the ratio of carotenoids to total chlorophyll was lower in leaf than in root plastids, the carotenoids being enriched with xanthophylls. Absorption spectra also showed peaks for lycopene and lycophyll in wheat roots but not in wheat leaves. Continuous illumination with white light of low intensity is the regime most commonly used to promote greening in roots. No evaluation seems to have been made of the effects on greening in intact roots by light of different wavelengths or of different intensities, though work of this type has been carried out on excised cultured roots of several species, inter alia, by Heltne and Bonnett (1970) and by Bajaj and McAllan (1969), but principally by Bjorn and by Richter and Dirks. B. BLUELIGHTAND GREENING An absolute requirement for blue light for chlorophyll synthesis and chloroplast development has been shown for excised roots of wheat,cucumber and pea (Bjorn, 1963, 1965, 1967, 1980; Bjorn and Odhelius, 1966; Dirks and Richter, 1975; Richter and Dirks, 1978). However, blue light alone does not bring about chloroplast development. Bjorn (1965, 1980) has suggested a scheme for chloroplast differentiation in roots which involves a series of reactions: (1) the first “red light” reaction brought about by either red or blue light, (2) a dark reaction, (3) a reaction requiring blue light alone, and (4) a second “red light” reaction, brought about by red light alone. The initial response to blue light appears to be the synthesis of specific types of plastid RNA; this response takes place without any lag period. Other slower responses to blue light include the synthesis of chlorophyll and of the chloroplast enzyme, D-glyceraldehyde 3-phosphate: NADP oxidoreductase, which take place only after a lag phase. From electron micrographs of Viciafaba, Dyer et al. (1971) calculated that the ratio of cytoplasmic to plastid ribosomes was low in both the roots and the shoot apices but high in both etiolated and green leaves. Though they give no absolute counts of plastid ribosomes per unit of stromal area it appears that even the most fully differentiated plastids in dark-grown roots usually contain many fewer ribosomes and have a much less extensive thylakoid system than do etioplasts in young leaves kept in the dark. When roots are exposed to light plastid ribsomes very quickly begin to increase in numbers (Fig. 14) but chlorophyll synthesis takes place only after a significant lag period (Oliveira, 1982). Plastids in dark-grown roots generally lack prolamellar bodies. Stetler (1973) examined dark-grown tobacco tissue cultures developed from the pith and equated an absence of prolamellar bodies with the failure (over a short term) of protochlorophyllide to be converted to chlorophyllide. Sundqvist et al. (1980) state that prolamellar bodies generally appear to be formed (as they may be during greening of some roots) only when a blue-mediated process, possibly protein synthesis, runs ahead of a red-mediated one such as the pro-

214

JEAN M. WHATLEY

tochlorophyllide conversion or a phytochrome-dependent reaction. However the results of other investigations (reviewed by Klein, 1982) show that the properties of prolamellar bodies are far from fully understood and their role in thylakoid development is by no means as certain as was formerly believed. The action spectra of the blue light effects in roots resemble the absorption spectra for various carotenoids and flavoproteins, and agree well with that for blue-light-requiring chlorophyll formation in glucose-bleached Chlorellu. It should, however, be noted that an absolute requirement for blue light during greening has only been shown for excised roots and other dark-adapted tissues cultured in media which, like the Chlorellu cultures, are rich in glucose or sucrose (Bjom, 1980). This response to blue light is reminiscent of the curious switch from carbohydrate to amino acid metabolism directed by blue light observed in many organisms (Voskresenskaya, 1972).

X. Some Nonphotosynthetic Functions of Plastids in Roots Little is known about the functions of plastids in roots and it is not the purpose of this article to consider them in any detail. Nevertheless it is perhaps appropriate to refer briefly to some of the few nonphotosynthetic functions which have been ascribed to root plastids, particularly as their isolation for biochemical investigation is now becoming more practicable and one can, perhaps, look forward to future research projects which compare their structure and function. Although it is some years since plastids were first isolated from roots (Thomson et al., 1972), most functional studies have, until recently, made use of' other methods. Autoradiography has been used by Northcote and Pickett-Heaps (1966) among others to trace the synthesis and transport by way of Golgi vesicles of polysaccharides in cells of the root cap. In the summary scheme proposed by Northcote and Pickett-Heaps, plastids have an obvious role as a site of starch storage and a source of the soluble pool of hexose phosphates in the cytoplasm. Plastids and Golgi vesicles have also been implicated in the formation and movement of phenolics, whose widespread occurrence and role, inter alia, in defense against pathogens, make them of particular interest. Mueller and Beckman have tried to determine the origin of phenolics in specialized xylem parenchyma cells in banana roots (1 974) and in endodermal cells in cotton roots (1976). In the endodermal cells of cotton, accumulation of phenols apparently takes place very rapidly (calculated from growth rate to occur within 30 minutes), and can first be detected in vacuoles of cells within 1 mm of the root tip. Mueller and Beckman had previously found a correlation in glandular hairs of tomato, between the accumulation of phenolics and the disappearance of starch from the plastids and they believe that their work on cotton roots shows a similar correlation. However, in the stele of banana roots, the large plastids in phenol-

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

215

accumulating cells have a homogeneous stroma without starch; they also lack thylakoids and invaginations of the plastid envelope, although these plastids do occasionally contain osmiophilic globules which the authors suggest may represent phenolic material. These plastids are closely surrounded by endoplasmic reticulum (ER), but the contents of the ER cisternae are not osmiophilic. The membrane-bound bodies of plastids in some other plant organs and in cell cultures (Gifford and Stewart, 1968; Israel et al., 1969) and osmiophilic deposits in plastids in the coralloid roots of the cycad, Cycas revoluta (Obukowicz et al., 1981) have also been proposed as sites of phenolic accumulation. However, at present no good ultrastructural or histochemical evidence exists which clearly implicates plastids in roots or any other organs of land plants as the site of origin of phenolic material. Nato and Deleens ( 1975a,b) have identified ribulose-bis-phosphate carboxylase (RUBP-carboxylase) inside the plastids of roots of Lens culinaris. Though the amount of RUBP-carboxylase present in dark-grown roots is small there is a significant increase when the roots are exposed to light. The increase in RUBPcarboxylase parallels the increase in chlorophyll synthesis and both processes have the same initial lag period (Fig. 15). When CO, is fixed via PEP-carboxylase, similar results are obtained in the light and in the dark. The amount of CO, fixed in light-grown roots by RUBP-carboxylase is similar to that fixed via PEPcarboxylase, though the latter is the main pathway of CO, fixation in dark-grown roots. Nato and Deleens confirmed (1975b) that the rates of synthesis by Lens roots of C , compounds is the same in the light as in the dark, but that the rates for C , compounds are markedly increased in light-grown roots. More recently there has been increased interest in the role of plastids in nitrite reduction in roots. Emes and Fowler (1979a,b), for example, using isolated amyloplasts from the apices of pea roots, have investigated the intracellular location of the enzymes associated with nitrate assimilation and have found that nitrite reductase, glutamate synthase, and glutamine synthetase are all located in plastids, though there is glutamine synthetase activity also in the cytoplasm. Emes and Fowler also found that all the enzymes of the pentose phosphate pathway were present in small amounts in plastids, though their main site was the cell cytoplasm. They have linked these and other observations together in a scheme which suggests how the pathways of carbohydrate oxidation and nitrate assimilation may interact within the plastids and with reactions taking place in the surrounding cytoplasm. Observations on the reduction of nitrite to ammonia in isolated amyloplasts from root tips of pea and wheat have been further extended to include examination of their responses to waterlogging or anaerobiosis. Dry et al. (198 1) found that under both these conditions nitrite is accumulated and not reduced; they propose that the inhibition of nitrite reduction results from an interruption in the supply of glucose 6-phosphate to the plastids, although ATP appears normally to regulate nitrite reduction. These few examples suggest

216

JEAN M.WHATLEY

that further biochemical work may well provide a much greater insight in the near future into the metabolic functions of plastids in roots.

XI. Conclusions Few land plants can synthesise chlorophyll in the dark. In evolutionary terms the last remnants of this earlier capacity seem to be found in the embryonic cotyledons of some gymnosperms (Kirk and Tilney-Bassett, 1978). Most roots grow below ground in darkness and lack chlorophyll. However, the absence of chlorophyll from these roots is not merely the result of absence of light. Plastid development in most roots (green, nongreen, or becoming green) seems to follow a bidirectional course equivalent to that of cell differentiation. The plastids in roots undergo the same basic sequence of developmental changes (Stages 1-5) as do the plastids in leaves and other organs, though development in roots is often brought to a halt before the plastids differentiate into mature (Stage 5 ) chloroplasts. Although the plastids in roots and, say, leaves follow the same basic pathway of development, the plastids in the two organs differ quantitatively from each other at every stage of development. The extent of the thylakoid system, for example, is usually considerably less in root plastids than in leaf plastids at the same developmental stage and under the same environmental conditions, and, when mature chloroplasts are formed in roots, they generally have fewer thylakoids linking the grana, though they may have as many grana as leaf chloroplasts (Azolla pinnata may be an exception here). In addition, the ancillary structures (containing precursor materials) which are formed in response to the temporary or permanent blockage of the developing thylakoid system differ in character and in the stages at which they develop in different organs. In dark-grown Phaseolus vulgaris seedlings, for example, development of the plastid thylakoid system is initially blocked earlier in the roots and hypocotyls than it is in the primary leaves and prolamellar bodies are formed later in the hypocotyls than in the leaves and not at all in the roots (Whatley, 1978, 1983). When dark-grown plants are transferred into the light, greening takes place much more slowly (if at all) in the roots than in the leaves and different intensities and possibly wavelengths of light are most effective. An early sign of future differences between roots and leaves in the precise patterns of development of their plastids is the smaller number of ribosomes in root eoplasts. The blue light effect shown for excised, cultured roots of several species (Bjorn, 1980) may be associated with reducing this disparity. When roots become green on exposure to light, mature chloroplasts develop in a much smaller proportion of cells than in the leaves; an extreme example of this, observed in the roots of several species, is the restriction of photosynthetically functional chloroplasts to

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

217

a single cell file of the inner cortex (Powell, 1925). In the rhizophores and roots of a Selaginella, loss of chlorophyll can be promoted either by growth in darkness or by growth in the light at high relative humidity (Webster and Jagels, 1977). However, in this Selaginella, each of these two factors promotes a different pattern of chloroplast dedifferentiation. Clearly, then, factors other than light are of major importance in controlling greening; these factors, and light, too, may all influence plastid ultrastructure in different ways, although the basic pattern of plastid development persists. The factors other than light appear to be of particular importance in controlling plastid development in roots, but, unfortunately, it is about these very factors that we have the least information.

ACKNOWLELXMENTS 1 would like to thank Drs. Barlow, Oliveira, Possingham, and Wellburn for sending me prepnnts of their recent publications; Dr. C. R. Hawes for providing me with sections of Zea roots and for an electron micrograph; and Miss T. Scaysbrook and Mr. I. D. A. Kerr for their help in printing the electron micrographs.

REFERENCES Audus, L. 1. (1962). Symp. Soc. Exp. Biol. 16, 197-226. Bajaj, Y. P. S., and McAllan, J. W. (1969). Physiol. Plant. 22, 25-28. Baker, N. R., Hardwick, K.. and Jones, P. (1975). New Phytol. 74, 513-518. Barlow, P. W . (1974). J . Exp. Bot. 25, 1137-1146. Barlow, P. W.. and Grundwag, M. (1974). 2. Pjlanzenphysiol. 73, 56-64. Barlow, P. W..and Sargent, J. A. (1978). Ann. Bot. 42, 791-799. Barlow, P. W . , Rost. T. L.. and Gunning, B. E. S. (1982). Protoplasma 112, 205-216. Behnke, H . D. (1972). Bot. Rev. 38, 155-197. Ben-Shaul. Y . , and Klein, S. (1965). Bot. Gaz. 126, 79-85. Ben-Shaul, Y . , Treffry, T.. and Klein, S. (1968). J. Microsc. 7, 265-274. Bjorn, L. 0. (1963). Physiol. Plant. 16, 142-150. Bjorn, L. 0. (1965). Physiol. Plant. 18, 1130-1 142. Bjorn, L. 0. (1967). Physiol. Plant. 20, 483-499. Bjorn, L. 0. (1976). Physiol. Plant. 37, 183-184. Bjorn, L. 0. (1980). I n “The Blue Light Syndrome” (H. Senger, ed.), pp. 455-464. SpnngerVerlag. Berlin and New York. and Odhelius, 1. (1966). Physiol. Plant 19, 60-62. Bjorn, L. 0.. Caporali, L. (1959). Ann. Sci. Nat. Ile Ser. Bot. 20, 215-247. Casodoro, G., and Rascio, N. (1977). J. Ultrasrruct. Res. 61, 186-192. Chaly, N . . and Possingham, J. V. (1981). Biol. Cell. 41, 203-210. Chaly, N., Possingham, J. V., and Thomson, W. W. (1980). J. Cell Sci. 46, 87-96. Clowes, F. A. L. (1954). New Phytol. 53, 108-1 16. Clowes, F. A. L. (1976). New Phytol. 77, 399-407. Clowes, F. A. L., and Juniper, B. E. (1964). J . Exp. Bor. 15, 622-630.

218

JEAN M. WHATLEY

Cormack, R. G. H. (1937). New Phytol. 36, 19-25. Cran, D. G., and Possingham, J . V. (1973). Ann. Eor. 37, 993-997. Deltour, R., and Bronchart, R. (1971). Planru 97, 197-207. Dirks, W., and Richter, G. (1975). Eiochem. Physiol. Pf7anz. 168, 157-166. Dry, I., Wallace, W., and Nicholas, D. J. D. (1981). PIantu 152, 234-238. Dyer, T. A., Miller, R. H., and Greenwood, A. D. (1971). J . Exp. Bot. 22, 125-136. Emes, M. I., and Fowler, M. W. (1979a). PIunru 144, 249-253. Emes, M. J., and Fowler, M. W. (1979b). Plunra 145, 287-292. Esau, K. (1965). “Plant Anatomy.” Wiley, New York. Esau, K., and Gill, R. H. (1971). J . Ulrrusrrucr. Res. 34, 144-158. Esau, K.,and Gill, R. H. (1972). J . Ulrrusrrucr. Res. 41, 160-175. Esau, K.,and Gill, R. H. (1973). J . Ulrrustruct. Res. 44, 310-328. Evert, R. F. (1977). Annu. Rev. Plant Physiol. 28, 199-222. Fadeel, A. A. (1962). Physiol. Plant. 15, 130-147. Fadeel, A. A. (1963). Physiol. Plant. 16, 870-888. Foster, A. S., and Gifford, E. M., Jr. (1974). “Comparative Morphology of Vascular Plants.” Freeman, San Francisco, California. Frey-Wyssling, A., and Schwegler, F. (1965). J. Ultrastrucr. Res. 13, 543-559. Gautheret, R. I. (1932). C. R. Acad. Sci. Paris D194, 1510-1513. Gifford, E. M., Jr., and Stewart, K. D. (1968). Am. J. Eot. 55, 269-279. Grenville, D. J., and Peterson, R. L. (1981). Eor. Gaz. 142,73-81. Grijnegress, P. (1971). Planru 98, 274-278. Gunning, B. E. S. (1978). Plunru 143, 181-190. Gunning, B. E. S., Hardham, A. R., and Huges, J. E. (1978a). Plunra 143, 145-160. Gunning, B. E. S., Hardham, A. R., and Hughes, J. E. (1978b). Planru 143, 161-179. Gunning, B. E. S . , Hughes, J. E., and Hardham, A. R. (1978~).PIunru 143, 121-144. Haberlandt, G. (1914). “Physiological Plant Anatomy.” MacMillan, New York. Hallam, N. D. (1972). PIunra 104, 157-166. Hallam, N. D., Roberts, B. E., and Osborne, D. J. (1972). Planru 105, 293-309. Hardham, A. R., and Gunning, B. E. S. (1977). Plunru 134, 201-203. Hejnowicz, Z.(1958). Physiol. Plant. 11, 878-888. Heltne, J., and Bonnett, H. T. (1970). PIuntu 92, 1-12. Hensel, W., and Sievers, A. (1981). PIantu 153, 303-307. Hestnes, A., and Iversen, T.-H. (1978). Physiol. Plant. 42, 406-414. Hurkman. W. J., and Kennedy, G. S . (1977). Am. J. Eot. 64, 86-95. Israel, H. W., Mapes, M. D., and Steward, F. C. (1969). Am. J . Eot. 56, 910-917. Iversen, T.-H. (1969). Physiol. Plant. 22, 1251-1262. Juniper, B. E. (1976). Annu. Rev. Plant Physiol. 27, 385-406. Juniper, B. E. (1977). Proc. R. SOC. London Ser. E 199, 537-550. Juniper, B. E., and Clowes, F. A. L. (1965). Nature (London) 208, 864-865. Juniper, B. E., and French, A. (1970). Plunra 95, 314-329. Juniper, B. E., Groves, S . , Landau-Schachar, B., and Audus, L. J. (1966). Nature (London) 209, 93-94. Kawamatu, S . (1967). Cyrologiu 32, 157-164. Kirk, J. T. O., and Tilney-Bassett, R. A. E. (1978). “The Plastids.” Elsevier, Amsterdam. Klein, S. (1982). I n “On the Origins of Chloroplasts” (J. A. Schiff, ed.), pp. 36-53. Elsevier, Amsterdam. Klein, S., and Ben-Shaul, Y. (1966). Can. J . Eot. 44, 331-340. Klein, S., and Pollock, B. M. (1968). Am. J . Ear. 55, 658-672. Lance-Nougadde, A., and Pilet, P. E. (1965). C. R. Hebd. Acad. Sci. Paris D 260, 2567-2570.

ULTRASTRUCTURE OF PLASTIDS IN ROOTS

219

Leech, R. M.. Thomson, W. W., and Platt-Aloia, K. A. (1981). New Phyrol. 87, 1-9. McNitt, R. E., and Shen-Miller, J. (1978). Plant Physiol. 61, 644-648. Maitra, S. C., and De, D. N. (19729. Qrobios 5, I 1 1-1 18. Melaragno, J. E., and Walsh, M. A. (1976). Am. J. Bor. 63, 1145-1157. Mesquita, J . F. (1971). Port. Acra Biol. Ser. A 12, 33-52. Mollenhauer, H. H. (1967). Am. J. Bor. 54, 1249-1259. Mueller, W. C., and Beckman. C. H. (1974). Physiol. Plunr Parhol. 4, 187-190. Mueller, W. C., and Beckman, C. H. (1976). Can. J . Bor. 54, 2074-2082. Nato, A., and Deleens, E. (1975a). Physiol. Planr. 34, 121-124. Nato, A., and Deleens, E. (1975b). Physiol. Planr. 34, 309-313. Newcomb, E. H. (1967). J. CellBiol. 33, 143-163. Northcote, D. H., and Pickett-Heaps. J. D. (1966). Biochem. J . 98, 159-167. Obukowicz, M., Schaller, M., and Kennedy, G . S. (1981). New Phyrol. 87, 751-759. Oliveira, L. (1975). J. Submicrosc. Cyrol. 7, 97-105. Oliveira, L. (1982). New Phvrol. 91, 263-275. Olsen. G. M., and Iversen, T.-H. (1980). Physiol. Planr. 50, 275-284. Parthasarathy, M. V. (1974a). Protoplasma 79, 59-91. Parthasarathy, M. V. (1974b). Protoplasma 79, 93-125. Parthasarathy, M. V. ( 1 9 7 4 ~ )Proroplasma . 79, 265-315. Parthasarathy, M. V., and Klotz, L. H. (1976). Wood Sci. Techno/. 10, 247-271. Peterson, R. L., and Brisson, J. D. (1977). Can. J. Bor. 55, 1861-1878. Peterson, R. L., and Vermeer, J. (1980). Am. J . Bor. 67, 815-823. Peterson, R. L., Scott, M. G., and Kott, L. (1979). Ann. Bor. 44, 739-744. Phillips, H. L., Jr., and Torrey, J. G. (1974a). Am. J . Bor. 61, 871-878. Phillips, H. L., Jr., and Torrey, J. G. (1974b). Am. J . Bor. 61, 879-887. Platt-Aloia, K. A,. and Thomson, W. W. (1977). New Phvrol. 78, 599-605. Possingham, J. V., Chaly, N., Robertson, M. A., and Cain, P. A. (1983). Biol. Cell 47, 205-21 1. Powell, D. (1925). Ann. Bor. 39, 503-513. Richter, G . , and Dirks, W. (1978). Phorochem. Phorobiol. 27, 155-160. Robertson, D., and Laetsch, W. M. (1974). Plant Physiol. 54, 148-159. Salema, R . (1971). An. Fac. Cienc. Port. 54, 1-9. Sargent, J. A , , and Osborne, D. J. (1980). Protoplusmu 104, 91-103. Schroter, K., Lauchli, A., and Sievers, A. (1975). PIanra 122, 213-225. Scott, N. S . , and Possingham, J. V. (1980). J . Exp. Bor. 31, 1081-1092. Shen-Miller, J. (1978). Plant Cell Physiol. 19, 445-452. Sievers, A., and Volkmann, D. (1977). Proc. R. Soc. London Ser: B 199, 525-536. Sprey, B . , and Lambert, C. (1977). Z.PJanzenphysiol. 83, 227-247. Stetler, D. A. (1973). Bor. Gaz. 134, 290-295. Street, H. E., Opik, H., and James. F. E. (1967). Phyromorphology, 17, 391-401. Sundqvist, C.. Bjorn, L. 0.. and Virgin, H . 1. (1980). I n “Results and Problems in Cell Differentiation” (J. Reinert, ed.), Vol. 10, pp. 201-223. Springer-Verlag, Berlin and New York. Suzuki, K.,and Ueda, R . (1975). Jpn. Bor. Mug. 88, 319-321. Tepfer, D. A., and Bonnett, H. T. (1972). Plantu 106, 31 1-324. Thomson, W. W. (1966). Bor. Gaz. 127, 133-139. Thomson, W. W., and Whatley, J. M. (1980). Annu. Rev. Plant Physiol. 31, 375-394. Thomson. W. W., Foster, P.. and Leech, R. M. (1972). Planr Physiol. 49, 270-273. Thorsch, I., and Esau, K. (1981). J. Ufrrasrrucr. Res. 75, 339-351. Tischler, G. (1905). Flora 94, 1-69. Voskresenskaya, N. P. (1972). Annu. Rev. Plant Physiol. 23, 219-234. Walsh, M. A. (1980). Ann. Bor. 46, 557-565.

220

JEAN M.WHATLEY

Webster, T. R. (1969).Can. J. Bor. 47, 717-722. Webster, T.R.. and Jagels, R. (1977).Can. J . Bor. 55, 2149-2158. Webster, T.R., and Steeves, T.A. (1963).Phytomorphology 13, 367-376. Webster, T. R., and Steeves, T.A. (1964).Can. J. Bor. 42, 1665-1676. Webster, T.R., and Steeves, T.A. (1967).Can. J. Bor. 45, 395-404. Wellbum, A. R. (1982).Inr. Rev. Cyrol. 80, 133-191. Wellbum. A. R., Robinson, D. C., and Wellbum, F. A. M. (1982).Planta 154, 259-265. Whatley, J. M. (1977).New Phyrol. 78, 407-420. Whatley, J. M.(1978).New Phyrol80, 489-502. Whatley, J. M. (1980).New Phyrol86, 1-16. Whatley, J. M. (1983).New Phyrol. 94, 381-389. Whatley, J. M.,and Gunning, B. E. S . (1981).New Phyrol. 89, 129-138. Whatley, J. M.,Hawes, C. R., Home, I. C., and Ken, J. D. A. (1982).New Phyrol. 90,619-629. Wochok, Z . S., and Clayton, D. L. (1978).Cyrobios 21, 103-Ill. Wochok, Z.S.,and Sussex, I. M. (1974).PIanr Physiol. 53, 738-741. Wroblewski, R. (1973).J. Submicrosc. Cyrol. 5, 97-105. Yoo. B. Y.(1970).J . CellBiol. 45, 158-171.

INTERNATIONAL REVIEW OF CYTOLOGY. VOL. 85

The Confined Function Model of the Golgi Complex: Center for Ordered Processing of Biosynthetic Products of the Rough Endoplasmic Reticulum ALAN M. TARTAKOFF Department of Pathology. University of Geneva School of Medicine, Centre Mkdical Universitaire, Geneva, Switzerland I. Introduction .................... II. The Confined Function M 111. Covalent Modifications . . . A. Proteolysis ......... B. Initiation and Elongation of Xylose-Linked Oligosaccharides . . C. Tailoring of Asparagine-Linked Oligosaccharides ............ D. Initiation and Processing of N-Acetylgalactosamine-Linked Oligosaccharides ....................................... E. Glycolipid Chain Elongation.. ........................... IV. Noncovalent Modifications . . . . . . . . . . V. Consequences of Processing for the Go1 the Cell as a Whole.. .............. References ...............................................

232 235 242 244

248

I. Introduction The organized and characteristic elements of the Golgi complex (GC)’ are the stacked smooth-surfaced cisternae which are found in the centrosphere of all eukaryotic cells. These cisternae, in conjunction with other associated smoothsurfaced membranes, are responsible for executing net unidirectional intracellular transport (ICT) from the rough endoplasmic reticulum (RER) toward more distally located structures, e.g., lysosomes and the cell surface (24, 35, 40, 47, 8 5 , 1 1 1, 134).The present article is not centered on this apparently most primary of Golgi functions, but rather on the broad range of accessory activities which occur during transport, the family of “posttranslational modifications.” T k s e ‘Abbreviations: endo H, endoglucosaminidase H; Gal, galactose; GalNAc, N-acetylgalactosamine; GC, Golgi complex; GDP, guanosine diphosphate; Glc, glucose; GlcN, glucosamine; GlcNAc, N-acetyl glucosamine; GlcUA, glucuronic acid; ICT, intracellular transport; IdUA, iduronic acid; Ig, immunoglobulin; Man; mannose; RER, rough endoplasmic reticulum; TPP, thiamine pyrophosphate; UDP, uridine diphosphate; VSV, vesicular stomatitis virus, Xyl, xylose.

22 1 Copyright 0 1983 by Academic Ress, Inc. All rights of reproduction in any form reserved. ISBN 0-12-364485-2

222

ALAN M. TARTAKOFF

events are, in all likelihood, not essential for the “primary” function of theGC yet they are crucial in allowing the cell to tailor its biosynthetic products for its own needs and the needs of the organism as a whole. In addition to modifying products of the RER,the GC may be involved in processing events as a result of its participation in other routes of vesicular traffic, e.g., centripetal traffic from the cell suface. At present, although the existence of such a return route is widely accepted, the only suggestions of processing are no more than fragmentary (23, 24, 47, 105). The vast body of structural and enzymological information which describes these processing event has been the subject of recent detailed reviews. In the present article no attempt is made to give a thorough exposition of this literature. Instead, the present article highlights those aspects of processing that may shed light on the detailed route and underlying mechanisms of transport. For example, posttranslational modifications may serve as a set of potential Golgi markers for describing the heterogeneity of the organelle. Furthermore, the results of such modifications (e.g., addition of a particular sugar) may provide circumstantial data indicating the itinerary followed by transported molecules. In this sense, these molecules may be considered a family of probes of the transport machinery. One is led to inquire which products are transported at what rate via which processing sites, or, if the structure of these molecules is altered, e.g., by genetic means, are they transported in the same fashion? Various nonequivalent criteria have been used to ascribe processing events to the GC-autoradiography , preparative or analytic subcellular fractionation, interruption by ICT inhibitors, delay in the impact of cycloheximide, etc. Each one, alone, provides necessarily incomplete information. The research covered here makes use of each of these approaches. In order to emphasize their nonequivalence and to avoid certain possible confusions, attention is paid to distinguishing between anatomic terms (“proximal,” “medial,” and “distal” cisternae) and kinetic terms (“early” vs “late”). Although the equivalences proximal = early and distal = late may be correct the needed supporting data for this notion are not yet at hand. Nevertheless, contradictions of this hypothesis are not known. 11. The Confined Function Model of the Golgi Complex

Given the body of anatomic and histochemical data which describes the structural heterogeneity of the stacked cisternae of the GC (24, 35, 134), it is plausible to suppose that selected posttranslational modifications are confined to distinct subcompartments of the GC. Such functional subcompartments might correspond to the relatively more proximal or distal cistemae, although lateral specialization might exist as well-electron microscopic observations suggest that vesicles may bud from the extremities of the cisternae and certain histo-

CONRNED FUNCTION MODEL OF THE GOLGI COMPLEX

223

chemical reaction products are not uniformly distributed along their length (24). Several conceptually quite different approaches have been used to discriminate among and ascribe function to distinct subcompartments:

I . The pharmacologic approach which makes use of monensin to interrupt ICT. As has been discussed elsewhere (134- 137), the carboxylic ionophore monensin causes partial equilibration of intra- and extracellular Na+ and K + , rapid (within seconds) dilation of all Golgi cisternae, and an impressive slowing of ICT. The agent is highly selective, at least after brief treatment ( 1 pJ4 monensin, 1 hour, 37°C): ATP levels and intracellular pH are little changed, many biosynthetic events continue [protein synthesis, lipid addition to glycoproteins (54) hyaluronic acid synthesis (84) etc.], and the effects of the ionophore are reversible.* Since in the presence of monensin newly synthesized proteins continue to exit from the rough endoplasmic reticulum (RER), and gain access to but do not traverse the GC, one may inquire whether given posttranslational modifications occur. If they do, they may be assigned to a relatively early subcompartment. If they do not, given appropriate control experiments, they may be assigned to relatively late sub compartment^.^ The first indications that the site of interruption of ICT by monensin lies only part way across the GC came from study of the progress of N-linked oligosaccharide maturation of Ig (139); however, a recent use of monensin should be mentioned since it adds independent support for this idea (37). When BHK cells are infected with Sindbis or Semliki Forest virus, viral budding is normally observed at the cell surface. If monensin is present, transport of the glycoproteins is interrupted in the GC (37, 54, 56) and viral nucleocapsids are found adhering to and budding across the membranes of dilated vacuoles (“ICBMs”) in the Golgi region. These nucleocapsids are presum*In the context of the later discussion, the study of hyaluronic acid biosynthesis is especially important since it strongly suggests that sugar nucleotide metabolism is little perturbed and that the effects of monensin on sugar incorporation (e.g., into glycoproteins) are directly related to intermption of ICT. Hyaluronic acid, since it lacks a protein core, would be free from such effects. 3The appropriate controls vary according to the event being studied. For example, in the case of Ig carbohydrate maturation although monensin blocks I3H]Gal and [3H]Fuc labeling of Ig, the two isotopes continue to be incorporated into other acid-insoluble material (139). In the case of the inhibition of acquisition of endo H resistence by transfenin in VSV-infected hepatoma cells, an internal control is provided by the G protein, which does become endo H resistant (133). Were posttranslation modifications due to enzymes with acid pH optima, inhibition by monensin might be due to alkalinization of the cistemal content resulting from H+-Na+ (or K + ) exchange. However, all oligosaccharide processing enzymes studied have roughly neutral pH optima. Furthermore, there are no indications that they are sensitive to relative N a + / K + levels. To date the investigators who use monensin unfortunately do not all conform to the rather short-term protocols first used. Comparison of data is therefore often difficult. The extent to which ICT is perturbed by monensin is thought to depend on the alterations of cytoplasmic ion levels which it provokes. The impact of the ionophore may therefore reflect the activity of the Na-K-ATPase of the cell under study.

224

ALAN M. TARTAKOFF

ably signaling the presence of and adhering to an overaccumulation of the viral glycoproteins. Since other nearby Golgi-derived elements are histochemically positive for thiamine pyrophosphatase (a histochemical marker of distal cisternae) but ICBMs are negative one can conclude that the viral glycoprotein has only partly traversed the Golgi stack. (The author’s interpretation of these data is in fact that the site of monensin arrest is within “medial” or “intermediary” cisternae since the ICBMs bear ricin binding sites.) Both the nucleocapsid-laden ICBMs and the nucleocapsid-free vacuoles can be partially separated from each other (101) (vide infra). The use of colchicine is more limited than that of monensin, since in a number of cell types ICT proceeds in the presence of the drug. Studies of the liver suggest that the site of arrest of ICT by colchicine may be later than the monensin site (103, 134). 2. The approach of analyric subcellular fractionation (20, 34a, 100, 101, 135, 147) in which microsomal fractions or subfractions are spread, for example, on continuous density gradients. The goal of this method is to be all inclusive rather than to enrich for selected markers which might prove nonrepresentativeof the whole. At present the cytologic origin(s) of such subfractions cannot be systematically identified.4 3. The in virro reconsrirurion approach (108, 112) in which an attempt is made to simulate selected steps of transport by mixing appropriate subcellular fractions or cell extracts. The read-out from such experiments may involve the monitoring of posttranslational modifications which are considered to be diagnostic of transport. This is a method which may provide proof of the adequacy of subfractionation procedures and lead toward elucidation of the mechanisms of transport. 4. The approach of cyrochernisrry (36, 109, 110, 138) making use of antisera which recognize processing enzymes or lectin conjugates. This method cannot indicate at which site processing enzymes are active, but it should locate the enzymes in a cytologic context. The lectin conjugates localize products of sugar transferase activity, with the reservation that it may prove difficult to establish whether a given sugar residue is borne by an N- or 0-linked or possibly lipid carrier. 4The presence of lipoprotein particles within the Golgi cisternae of the hepatocyte has major implications for liver Golgi subfractionation. Preparative Golgi fractions can be obtained free from smooth ER and plasma membrane because of the density perturbation which results from this buoyant content. For analytic isopycnic fractionation one would anticipate a dispersion of density in proportion to variability in the number of lipoprotein particles contained within Golgi fragments produced by homogenization. Moreover, the relative isopycnic density of different subregions of the GC would reflect the degree of maturation of the lipoproteins. If the lipid content-and as a result the density shift-increases monotonically during ICT then the relative densities of Golgi subfractions might correspond inversely to their position along the time axis of transport; however, there is no proof that such a monotonic relation applies.

CONFINED FUNCTION MODEL OF THE GOLGI COMPLEX

225

5. A possible means of localizing a Golgi component relative to selected processing enzymes is to inquire whether or not it has undergone proteolytic cleavage, terminal sugar addition to N-linked oligosaccharides, initiation or completion of any O-linked oligosaccharides etc. The degree to which it has undergone posttranslational modifications should reflect its ontogeny and site of residence. 6. A kinetic ordering of processing events can be obtained by performing, for example, an amino acid labeling pulse-chase experiment and inquiring as a function of time, when the product in question undergoes one or another maturation step. Distinct kinetics need not, of course, imply distinct compartmentalization. The resolution of this approach is limited by any dispersion of transport rates of the pulse-labeled molecules. The exploitation of these approaches has by no means been sufficiently extensive to allow assignment of the multitude of processing events to subregions of the GC. Moreover, the present lack of detailed knowledge of the GC is such that it is the analysis of numerous processing events which may actually serve to enumerate the number of subcompartments which exist. Judging from histochemical data (24, 35, 111, 134), at least three classes of cisternae can be distinguished: the most proximal (overstained by OsO,), medial cisternae (reactive for nicotinamide adenine dinucleotide phosphatase at pH 5 in certain cells) and the most distal (reactive for nucleoside diphosphatase = uridine diphosphatase = thiamine pyrophosphatase). In addition, the structurally somewhat removed tubules and cisternae of “GERL” are histochemically reactive for acid phosphatase. Present studies employing lectins to stain the Golgi cisternae of various cells have also distinguished two or three distinct staining regions along the proximal-to-distal axis (36, 37, 109, 138) (Table I). The idea of subcompartmentalizationof the GC is discussed here in the context of distinct varieties of posttranslational processing several of which are illustrated, in abbreviated form, in Figs. 1-4. To a certain extent, these events might all be found within a single cell; however, since the biosynthetic repertoire (e.g., of exported glycoconjugates) varies among cell types, there is every reason to anticipate significant enzymologic variability of the GC as well. In fact, it is clear that selected sugar transferases are highly enriched only in certain tissues. There also must exist a set of altogether constant Golgi components-those which are responsible for its characteristic structure and endow it with the ability to conduct transport. No such components have yet been identified. A secondary task in this mapping of Golgi functions is to position the processing paths one relative to the next. For example, is the compartment responsible for proteoglycan sulfation the same as that responsible for galactose addition to asparagine-linked oligosaccharides? The ideal systems to investigate to obtain such information are those in which a single polypeptide undergoes several

TABLE I SUB-GOLGIL~CALIZATION OF HISTOCHEMICAL MARKERS AND PROCESSING EVENTS Histochemical marker

Cell type

Localization

Reference

Overosmicated material Nicotinamide adenine dinucleotide phosphatase (pH 5 ) Nucleoside diphosphatase (UDPase,

Many cells Ameloblasts, spermatids

Proximal cisternae Medial cisternae

134 13a. 124a

Many cell types

Distal cisternae

134

Many cell types

GERL (and lysosomes)

134

BHK cells Myelomas (Ig , Ig - )

All cisternae (and RER) Proximal cistemae (and REW Medial and distal cistemae Distal cisternae Proximal cistemae (A+, A-) Distal cisternae (A+)

36 138

___

TPPiiSe) Acid phosphatase Lectin-binding sites Concanavalin A

+

Ricinus 120 Wheat germ agglutinin Helix lectin

Event Proteolysis Proalbumin cleavage

BHK cells Myelomas Goblet cells

36 138 109

A p p r o a c h e l l type

Observations

Reference

Liver SCF,U in vivo amino acid label

92, 102, 103

94

93

Proalbumin cleavage

Same. with colchicine

Proalbumin cleavage

Primary liver culture, biosynthetic

Proalbumin is enriched in %sternal” subfractions; albumin in “postcisternal” subfractions Overaccumulation of albumin and proalbumin, especially in “post-cisternal” fractions Oveiacc-urrrulationof albumin

Proalbumin cleavage

label T tax01 Same, T monensin ( I hour)

Overaccumulation of proalbumin

102, 103

S F V b glycoprotein cleavage Sindbis glycoprotein cleavage Corona virus glycoprotein cleavage Friend murine leukemia virus glycoprotein cleavage SFV glycoprotein cleavage Proopiomelanocortin cleavage

N

Xylose-linked chain elongation Repeating disaccharide synthesis

4 N

Sulfation Relative timing of sulfation and chain elongation Relative timing of chain elongation and acquisition of endo H resistance by N-linked units

Biosynthetic label of BHK cells T monensin (1.5 hours) Biosynthetic label of chick fibroblasts 7 monensin (3 hours) Biosynthetic label of 17Cll cells 7 monensin (16 hours) Biosynthetic label of Eveline cells T monensin ( 1 hour) Biosynthetic label of chick embryo fibroblasts T monensin (2 hours) Biosynthetic label of rat pituitary culture T monensin (2 hours) Biosynthetic label of chick chondrocytes T monensin (3 hours) Biosynthetic label of rat chondrosarcoma 7 monensin (7 hours) Biosynthetic label of human melanoma 7 monensin (18 hours) As above biosynthetic label, sizing of monomer, rat chondrosarcoma

Interruption of cleavage

37

Interruption of cleavage

54

Interruption of cleavage

88

Slowing of cleavage

127a

No effect on cleavage?

98

Interruption of cleavage

17

Inhibition of synthesis

90,91

Inhibition of synthesis

84

Inhibition of synthesis

12

Inhibition of synthesis

90,84, 12 64

35S04

Sulfation follows elongation, 5 minutes before discharge

Amino acid label, gel analysis human melanoma

Roughly simultaneous

12

(continued)

TABLE I (continued) Event Relative timing of chain elongation and GaNAc addition to 0-linked

Approach--cell type

Observations

Reference

Labeling with [3H]GlcN of rat chondrosarcoma

Roughly simultaneous

141

Biosynthetic labeling T rat chondrosarcoma monensin (6 hours)

No change from controls

84

Immunocytochemistry of HeLa cells

Restricted to distal (TPPase+) cisternae Gal transferase is lower density than M I C Same Coisopycnic Coisopycnic Slight differences Coisopycnic with MI

110

20

SCF of lymphoma, macrophage

Progressively decreasing

34a

Biosynthetic label of human fibroblasts 7 monensin Biosynthetic label of plasma cells, myeloma T monensin (1 hour) Biosynthetic label of hepatoma i monensin (1 hour)

Similar to controls

l00a

Intemption of terminal sugar addition Same

134, 135

UNtS

Hyaluronic acid synthesis

N

Asparagine-lied oligosaccharide addition Cytologic localization of p( 1-4) Gal transferase Relative density of Gal transferase and mannosidase I

m N

Relative density of tenninal sugar transferases Relative density of phosphotransferase and phosphodiesterase Relative density of phosphotransferase, diesterase, and Gal transferase Phosphotransferase action Terminal sugar addition to Ig Terminal sugar addition to transferrin

SCF of liver SCF of SCF of SCF of SCF of SCF of

CHO BHK liver lymphoma liver

100

101 10, 147

34a 100

133

Terminal sugar addition to viral glycoproteins

Terminal sugar addition to proopiomelanocortin Relative timing of lipid addition and acquisition of endo H resistance Relative density of site of lipid addition and Mannosidase I

Lipid addition

Biosynthetic label of Eveline cells infected with MuLV T monensin (2 hours) Biosynthetic label of BHK cells infected with Lacross virus T monensin (2.5 hours) Biosynthesis label of chick fibroblasts infected with SFV T monensin (2 hours) Biosynthetic label of chick tibroblasts infected with VSV 7 monensin (6 hours)

Same

127a

Partial block

79a

Partial block in oligosaccharide maturation

98

G becomes resistant to endo H

54

Biosynthetic label of hepatoma infected with VSV T monensin (2 hours) Biosynthetic label of pituitary culture T monensin (2 hours) Biosynthetic label of Vero cells infected with VSV

G becomes resistant to endo H

133

Proopiomelanocortin becomes resistant to endo H Palmitate addition precedes endo H resistance

17

SCF of pulse [3H]palmitate-labeled VSV-infected CHO cells

Site of G acylation is coisopycnic with mannosidase

20

SCF of pulse [3H]palmitate-labeled SFV infected BHK T monensin (4 hours) [3H]Palmitate label of Sindbis- or VSV-infected chick fibroblasts 7 monensin (3 hours)

As above, unless monensin has increased mannosidase density

101

Little or no inhibition

54

119

(continued)

TABLE I (conrinued) Event GalNAc-linked chain elongation Site of mucin Gal addition L

Relative timing of internal GalNAc and Nlinked peripheral GlcNAc addition Oligosaccharide elongation

Glycolipid elongation Conversion of glucosyl ceramide to complex glycosphingolipids Transport of gangliosides to cell surface aSCF, Subcellular fractionation. %FV, Semliki Forest virus. =MI,Mannosidase I.

Approach-cell type

Observations

Reference

Autoradiography of mucin-secreting cells of stomach (2 minute) [WIGal pulse) Biosynthetic label of chorionic gonadotropin-secreting cells [3H]GlcN label

Incorporation is over distal cisternae and vesicles

12

Similar rates of labeling in secreted hCG

41

Biosynthetic label 7 monensin (16 hours) 17Cll cells infected with coronavirus Biosynthetic label T monensin (6-9 hours) HEp2 cells infected with Herpes virus

Interruption of sugar addition

88

Inhibition of shift in glycoprotein gel mobility

S4a

Biosynthetic label of glioma [WIGal 7 monensin

Monensin blockes conversion

26a

[3H]Gal pulse-chase, evaluation of surface exposure 7 monensin

No change from controls

26a

CONFINED FUNCTION MODEL OF THE GOLGI COMPLEX

23 1

modifications. Possibly those hybrid structures which do not exist in nature might be constructed by appropriate gene splicing.

111. Covalent Modifications

A. PROTEOLYSIS Numerous secretory and membrane proteins are selectively cleaved during their passage through the GC to the cell surface. The best studied examples are proalbumin, the spike glycoproteins of many enveloped virus, and the family of precursors of polypeptide hormones [pro-insulin, -glucagon, -parathyroid hormone, -gastrin, -enkephalin ( 14), -opiomelanocortin (48), vasopressin-neurophysin precursor (74)]. Many of the data implicating the GC are in fact only indirect. The approaches used involved study of the effect of agents which block secretory protein exit from the RER (e.g., uncouplers) (130), correlation between the kinetics of ICT and cleavage, or, best of all, analysis of smooth microsomal or Golgi-enriched subcellular fractions [e.g., of hepatocytes (92, 102) or virally infected cells (68)]. In the case of Sindbis and Semliki Forest virus it has been suggested that cleavage occurs at the moment of arrival at the cell surface since extracellular antisera directed against their glycoproteins can block cleavage and since precursor species are not detected by surface-labeling procedures (8,15 1). However, the cytologic site of cleavage may depend on both the virus and the host cell-for several paramyxovirus, influenza virus and certain RNA tumor virus cleavage has been reported to occur intracellularly (68). For Rous sarcoma virus the suggested site of cleavage ranges from the GC to the virion itself (7a). Moreover, certain host cells do not cleave the viral glycoproteins and therefore may yield noninfectious virus. The covalent site of proteolysis is strikingly uniform (18, 32). In all the cases mentioned above, cleavage occurs adjacent to one to four basic amino acids and results in the release of lysine or arginine. Such amino acid excision and the generation, for example, of insulin from proinsulin, have been ascribed to the concerted action of both trypsin-like and carboxypeptidase B-like activities ( 18), however, none of the claims of attribution of these activities to identifiable enzymes is unequivocally accepted. Such protease activities have yet to be used as Golgi markers. An outstanding example of differential cleavage is the case of proopiomelanocortin. The same precursor is cleaved (always adjacent to basic amino acids) to yield different products by cells of the anterior or intermediary lobe of the pituitary (46, 48). Moreover, in other endocrine cells minor amounts of unorthodox or incomplete cleavage products may be detectable (18). The result is a

232

ALAN M. TARTAKOFF

variety of microheterogeneity reminiscent of the microheterogeneity of processed carbohydrate structures of glycoproteins. It has been suggested that the proteolysis of proinsulin to insulin may occur both within the GC and in secretion granules (18, 130). The experiments on which this idea is based involve study of isolated granule fractions derived from biosynthetically labeled tissue. These data suggest that protease(s) of the GC may be transported along with the secretory product. In the case of proopiomelanocortin processing, successive cleavage events may occur within different compartments-P-lipotropin and ACTH are generated during transport, while further conversion to P-endorphin occurs in granules (32a, 38). There are no indications that cleavage is a prerequisite for transport. In the case of proalbumin, proinsulin, and the precursors of viral proteins, altered polypeptides which resist cleavage (mutant species or the products of incorporation of amino acid analogs) have been shown to arrive at the cell surface (18,68). Amino acid labeling pulse-chase experiments show that monensin interrupts the cleavage of viral surface glycoproteins, proalbumin, and proopiomelanocortin. (Table I). Therefore, one can ascribe cleavage to a late subcompartnient. This conclusion is consistent with the liver Golgi subfractionation studies which show that proalbumin-albumin conversion occurs upon entry into ‘‘postcisternal” Golgi elements-those which carry the greatest load of lipoprotein (92, 103). Experiments on the influence of colchicine on N-linked oligosaccharide maturation of secretory glycoproteins show that it interrupts ICT at a very late site (102, 103). It is therefore not surprising that it causes an accumulation of postcisternal elements containing an excess of both albumin and proalbumin (102, 103). When taxol is used to stabilize microtubules cleavage to albumin proceeds, and discharge is much reduced (94). In the case of glycoproteins which undergo proteolytic cleavage prior to arrival at the cell surface there are insufficient data to generalize with respect to the relative order of the events of carbohydrate maturation and proteolysis. Nevertheless, pulse-chase studies of proopiomelanocortin biosynthesis employing monensin suggest that the acquisition of mature carbohydrate units precedes proteolysis (17). In this study the inhibition by monensin may reflect neutralization of intragranular pH since the granule protease has a pH optimum at pH 5 (69).

B . INITIATIONAND ELONGATION OF XYLOSE-LINKED OLIGOSACCHARIDES Chondroitin sulfate, dermatan sulfate, heparin, and heraran sulfate all contain the same internal tetrasaccharide structure (Xyl-Gal-Gal-GlcUA) linked to multiple serine residues of a protein core. The four corresponding monosaccharide transferases have been characterized and substantially purified, and a limited amount of information is available concerning the polypeptide recognized by the initiator xylosyl transferase ( 15).

CONFINED FUNCTION MODEL OF THE GOLGI COMPLEX

233

HS

1-O-q HA

FIG. 1 . Structural and biosynthetic relations among xylose-linked proteoglycans and haluronic acid (which lacks a protein core). Serine residues acquire a common tetrasaccharide by sequential monosaccharide addition and are subsequently elongated by addition of repeating disaccharidesagain from monosaccharide donors. For heparin, heparan sulfate, and dermatan sulfate a further epimerization of GIcUA to ldUA occurs at many sites. A single core protein may bear both chondroitin sulfate and keratan sulfate (cf. Figs. 2 and 3) as well as 0-GalNAc and N-GlcNAc-linked chains. For HS and DS some sulfation precedes epimerization. CS, Chondroitin sulfate; HS, heparan sulfate; DS, dermatan sulfate; HA, hyaluronic acid.

There are provisional indications that at least the xylosyl and first galactosyl transferases might be physically associated with each other-they are both recovered in immunoprecipitates prepared with monospecific anti-xylosyl transferase antisera ( 12 I). The literature is equivocal with respect to the cytologic site(s) of addition of these internal sugar residues. Attempts at subcellular fractionation of embryonic cartilage suggest that the four corresponding enzyme activities belong to both rough and smooth microsomal fractions; however, the enzyme assays employed are in several respects not ideal (51). Part of the fascination with proteoglycan biosynthesis comes from recognition of their immense size [beautifully visualized by electron microscopy (62)] relative to the dimensions of the machinery of ICT.A monomer of cartilage proteoglycan, for example, consists of a core protein (- 300 nm long when spread)

234

ALAN M. TARTAKOFF

which bears on the order of 80 lateral sulfated chains (internal tetrasaccharide and repeating sulfated GalNAc, GlcUA disaccharide) each of which is about 100 residues long. Recent experiments making use of antisera directed against protein determinants of the proteoglycan monomer indicate that there is a substantial intracellular pool of core protein lacking the repeating disaccharides (12, 65). In amino acid pulse-chase experiments, this pool, which is presumably in the RER, can be chased into mature proteoglycan with a half-time of 80 minutes, i.e., roughly equal to the half-time for the inhibition of proteoglycan sulfation by cycloheximide (64, 83). Autoradiographic data and experiments employing puromycin (33, 87, 140, 149) are clear in implicating the GC as the site of addition of the repeating disaccharides and their sulfation. Moreover, cell-free synthetic studies employing Golgi-rich fractions of a mastocytoma show (by analogy with other studies of disaccharide growth) that chain elongation is the result of alternate addition of the two component sugars from UDP-GalNAc and UDP-GlcUA (123). No information is available concerning determination of the length of the oligosaccharide chains, which are, in fact, not of altogether uniform length (25). Sulfation appears to occur after disaccharide elongation since 2-minute 35S0, pulse-labeled intracellular chains have already acquired their mature hydrodynamic size (64).The half-time for discharge of 35S0, pulse-labeled chains is only 5 minutes (63). Hence sulfation is a very late event. Several noncovalent assembly steps result in the ultimate extracellular product: monomeric units of derivatized core protein firmly associated with the hydrophobic “link” glycoprotein and hyaluronic acid. A set of ingenious competition experiments has shown that hyaluronic acid-which is synthesized by the same cells which synthesize the monomer-is bound only after discharge (63). Although intracellular monomer can bind the link protein after solubilization, it is not known whether this event or the association with collagen (121, 143) normally occurs prior to discharge. A word should be said about the biosynthesis of hyaluronic acid. It is at no time covalently bound to protein and cycloheximide does not interrupt its biosynthesis (80). Its repeating disaccharide (GlcNAc, GlcUA) can be assembled in vitro by incubating cell extracts with UDP-activated substrates (2). Growth is at the reducing end of the chain, which is transiently membrane bound, possibly via UDP (100b). Autoradiographic study of synovial cells proves that hyaluronic acid synthesis does indeed occur within the GC (5). The demonstration that its synthesis is not influenced by doses of monensin which block sulfation of chondroitin sulfate by the same cells is consistant with its not having a protein core-sugar addition can occur in situ without ongoing ICT (84). In order to assign its synthesis to a subcompartment it will be important to learn whether monensin blocks its secretion.

-

-

CONFINED FUNCTION MODEL OF THE GOLGI COMPLEX

235

The most complex of the proteoglycan biosynthetic pathways is that of heparin, which has been in large part elucidated with the help of mast cell tumors. Following the internal tetrasaccharide addition and elongation of the repeating (GlcNAc, GlcUA) disaccharide units, five further enzymes intervene (76, 77): (1) a deacylase which removes most acetyl groups from GlcNAc (this activity is low in cells synthesizing heparan sulfate and is stimulated by Nsulfation at other residues), (2) an N-sulfotransferase which acts on most of the sites exposed by ( I ) , (3) the epimerase which converts 70-90% of glucuronic acid residues of the polymer to iduronic acid, (4) an 0-sulfotransferase which acts at C-2 of many iduronic acid residues, and (5) an 0-sulfotransferase which acts at C-6 of glucosamine. There are no data indicating whether some or all of these activities are expressed in the same cytologic subcompartment in which disaccharide elongation occurs. Recent studies have indicated the existence of two classes of heparan sulfate. One is released from the cell (e.g., hepatocyte); the other is hydrophobic and membrane-associated. It is thought to have a hydophobic core protein which serves as anchor (66). Study of the biosynthesis of chondroitin sulfate is potentially of unique interest for defining the subcompartmentalizationof the GC since the same core protein which bears 80 xylose-linked repeating disaccharides also bears 12 N-linked complex oligosaccharides, 120 short 0-linked GalNAc-containing oligosaccharides, 60 single GalNAc residues, and (in many cases) keratan sulfate chains (44, 78, 79). Such multiple derivatization of a single polypeptide provides ideal material for establishing the relative sub-Golgi sites and rates of processing of such divergent structures. The presently available kinetic data shown that the acquisition of endoglycosaminidase H (endo H) resistance by the N-linked oligosaccharides of chondroitin sulfate is roughly coincident with the addition of the repeating disaccharides (12). Study of the incorporation of labeled glucosamine into the repeating disaccharides and the internal residues of the short 0linked units suggest that these two events occur simultaneously (141). Other data bearing on the sub-Golgi localization of proteoglycan synthesis come from use of monensin. Both the addition of repeating disaccharides and sulfation of chondroitin sulfate are markedly reduced in the presence of monensin though the pools of ATP and phosphoadenosine phosphosulfate are unchanged (12, 57, 90, 91). One can provisionally conclude that these events take place relatively late during transit through the GC. The limited amount of chondroitin sulfate which is secreted in the presence of monensin is undersulfated and may serve as a model for certain achondroplasias.

-

c.

-

-

TAILORING OF ASPARAGINE-LINKED OLIGOSACCHARIDES 1. Secretory and Membrane Glycoproteins ( 7 , 71, 113-115, 125, 127) With the realization that preassembled dolichol-linked oligosaccharides are cotranslationally added to selected asparagine residues, the further processing of

236

ALAN M. TARTAKOFF 4-4

0 P

4 A- m - 4 :

I

‘4

-4

4-4 4-4-4-+-+-+ 4 hA-00-4’ \+-+ -t lUnit /4

-

01.11 MI

\

4- 4

4 A-00-4’

‘4 \

4

21

’ /4\ A - W - + \

4

-

Till

HybrM(b1s)

TIV.TIII

FIG. 2. Structural and biosynthetic relations among asparagine-linked saccharides. The common precursor at the top left can be trimmed and further processed toward complex or hybrid structures, structures characteristic of lysosomal enzymes (“unblocked unit”), or keratan sulfate. As is extensively discussed elsewhere (71,113) along the “complex pathway” several options and control points exist. The activities of GlcNAc transferases (TI-TIV) are of key importance. TIII, for example. generates a “bisected” (bis) structure by addition of GlcNAc to the most internal (plinked) mannose and as a result blocks the activity of TII, TIV. fucosyl transferase, and mannosidase 11. TI action is a prerequisite for action of several further transferases. The structure bearing GlcNAc-phosphate, which can give rise to an unblocked unit, indicates only one of several possible phosphorylated mannose residues. The actual degree of concomitant mannose trimming is progressive but not uniform. Tr, Terminal sugar (fucose, Gal, sialyl) transferases; G , glucosidase; M, mannosidase.

CONFINED FUNCTION MODEL OF THE GOLGI COMPLEX

237

such moieties has been rapidly elucidated. The recent reviews of these investigations are too thorough to warrant an exposition of this now well-known sequence of steps which involves both sugar trimming and terminal sugar addition. The processing to a triantennary complex unit, e.g., of the G protein of vesicular stomatitis viruses (VSV) in infected CHO cells, involves nine enzymes and of these only the first two glucosidases and possibly the first mannosidase act at the level of the RER (71, 127, 135). The rest have been localized to preparative Golgi fractions. The p( 1-4)galactosyltransferaseresponsible for subterminal Gal addition is, at present, the best characterized Golgi enzyme. Not only have its enzymological properties been investigated in several tissues, but in addition first studies of its trans-membrane orientation (30) and biosynthesis (132) have been reported. In HeLa cells the transferase bears both 0-linked and endo H-resistant N-linked oligosaccharides (G. Strous, personal communication). Several laboratories have monitored attempts at sucrose gradient Golgi subfractionation with the use of this set of enzyme activities. The observations to date are fragmentary and the separations only partial. The first mannosidase has been reported to have an isopycnic density in sucrose gradients which is higher than that of galactosyltransferase [CHO cells (20) and liver (loo)]. Such a difference is not observed in BHK cells even when they are infected with Semliki Forest virus and treated with monensin so as to increase the density of medial Golgi elements (101). In the case of CHO cells the density differences are small but intriguing since the mannosidase activity is coisopycnic with those vesicles in which the G protein acquires covalently bound lipid (vide infra). Attempts to separate subfractions differentially enriched in the terminal sugar transferases (GlcNAc, Gal, Sialyl) by sucrose gradient analysis have failed (10, 147) or provided only slight differential enrichment (34a). Both affinity methods and countercurrent distribution studies do discriminate between liver Golgi subfractions differentially enriched in galactosyltransferase vs “ER-like” enzyme activities (49, 53). An outstanding only partly resolved question is what governs the particular path and degree of processing at a given glycosylation site. Among the key considerations is the repertoire of processing activities of a given tissue, possible competition between transferases of overlapping specicificity (7, 1 13), and the steric environment of the oligosaccharides. In the case of thyroglobulin (127), certain IgM myeloma proteins ( I ) and certain viral envelope glycoproteins the situation is particularly intriguing since different sites along a single polypeptide ungergo radically different degrees of processing. In this last-mentioned case, it has been observed that when Sindbis virus infects different host cells different sites are processed to different extents (125). Additional variables in terminal sugar addition concern (1) the degree to which the transferases may be present at the cell surface as well as in the GC, and (2)

238

ALAN

M.TARTAKOFF

the apparent intracellular pool size of secretory glycoproteins which have already accepted their terminal sugars. Both enzymologic (122) and, less equivocally, immunocytochemical studies (99) show that in selected cells galactosyltransferase may be exposed at the (apical) cell membrane. Presumably this is a reflection of some structural change in the enzyme or other determinative Golgi elements which abrogates the normal affinity of the enzyme for the GC. With respect to (2), comparative studies of immunoglobulin synthesis by several mouse myelomas indicate that the intracellular pool of Ig which can be labeled with terminal sugars may be either undetectably small or quite substantial (81, 97). These variations may parallel the degree of differentiation of the myeloma in question. A variable location of the terminal sugar transferases might explain such observations, i.e., in some cases they might act later along the path of ICT than in others. Several cytologic studies pertain to this pathway. An immunocytochemical study has localized p( 1-4) galactosyltransferaseto the distal (thiamine pyrophosphatase-positive) cisternae of the GC of HeLa cells (1 10). The presumed products of terminal sugar transferases have also been localized. Concanavalin A-gold conjugates (which interact primarily with immature asparagine-linked oligosaccharides) stain the cisternal space of the RER and all Golgi cisternae, whereas ricin 120 (which interacts with nonreducing galactosyl residues) stains only the more distal cisternae of BHK cells (36, 37). In a methodologically quite different study of IgM-secreting myeloma cells and an Ig-negative mutant, concanavalin A-peroxidase has been shown to stain the cisternal space of both the RER and proximal cisternae while a wheat germ agglutinin-peroxidase conjugate (specific for clustered sialic acid) stains the cisternal space of more distal cisternae and associated vesicles (138). Any such studies must be interpreted in light of what is known of the principle glycoproteins and glycolipids of the cells studied. Especially in the case of the IgM myeloma (and to a lesser extent with the BHK cells when they are infected with Semliki Forest virus) one has reason to consider much of the staining to be due to a single glycoprotein. Nevertheless, the study of the myeloma mutant suggests that the oligosaccharides of underlying Golgi components distribute in roughly the same way as the Ig oligosaccharides. Taken as a whole, the cytologic observations argue strongly that the direction of transport is indeed proximal-todistal and that the terminal sugars are acquired during passage across the stack of Golgi cisternae. These conclusions are also compatible with the available monensin data. In short-term experiments with IgM-secreting plasma cells and myeloma cells, monensin interrupts ICT within the GC at a site where the Ig heavy chains are still sensitive to endo H and have acquired neither galactose, fucose, nor sialic acid (134, 135). Terminal sugar addition to other glycoproteins continues and Ig

CONFINED FUNCTION MODEL OF THE COLCI COMPLEX

239

labeled with [3H]galactose or [3H]fucose is rapidly secreted (139). Similar results have been reported for the effect of monensin on maturation of the oligosaccharides of transfernin (133) and of the surface glycoproteins of Semliki Forest virus (37). In other cells and employing somewhat different protocols the major intra-Golgi site of accumulation of membrane proteins in the presence of monensin may be such as to have allowed somewhat further oligosaccharide maturation (Table I). Nevertheless, transport to the cell surface is not observed. There is a line of experimentation indicating the existence of two sequential subcompartmentstraversed by the G protein of VSV. The ultimate interpretation of the data remains to be determined, but is thought to pertain to the ability of the membranes of Golgi subfractions to undergo fusion with the membrane of a relatively late subcompartment. In the protocols employed (20, 30a, 11 1) VSVinfected mutant CHO cells (lacking GlcNAc transferase I) are pulse-labeled with [35S]methionineand chased for increasing intervals. At each time point a cell hornogenate is mixed with a wild-type homogenate, briefly incubated, and assayed for the proportion of 35S-labeled G (initially sensitive to endo H and perpetually so in the mutant CHO cell) which has acquired resistance to endo H. The in virro maturation is manifest only after a brief chase interval, suggesting that at later times G has gained access to a subcompartment which will no longer participate in such membrane-membrane fusion. Numerous yeast glycoproteins, for example, secretory invertase and acid phosphatase, bear both asparagine-linked and 0-linked polymannosyl oligosaccharides. The asparagine-linked units are donated from the same dolichol-linked intermediates as in mammalian cells, trimmed to a structure containing 2 GlcNAc and 8 Man, and elongated by massive further addition of mannose and phosphate (3). Studies of yeast temperature-sensitiveICT mutants which overaccumulate RER, Golgi-like cisternae, or vesicles suggest that the polymannosyl elongation of the N-linked oligosaccharides of invertase occurs within the GC (21). The phosphorylation of the secretory protein also occurs within this compartment ( 131). 2. Lysosomal Enzymes (71, 86, 124) In cultured fibroblasts, the proper intracellular targeting of lysosomal enzymes depends on the presence of mannose 6-phosphate on high mannose asparaginelinked oligosaccharides. The phosphorylated structure is the result of the posttranslational addition of one or several GlcNAc- 1-phosphate units to several oligosaccharides per polypeptide, followed by removal of much of the GlcNAc. Biosynthetic labeling studies show that the “unblocked” phosphate structure is achieved within 1 hour after a [3H]mannosepulse and that phosphoryl groups are ultimately lost, presumably upon entry into lysosomes (31). The enzymes responsible both for addition of the blocked phosphate and subsequent removal of

240

ALAN M. TARTAKOFF

the GlcNAc are both enriched in preparative liver Golgi fractions. In attempts at liver Golgi subfractionation, the two activities have been reported to be coisopycnic with mannosidase I, i.e., denser than galactosyltransferase (100). Sucrose gradient subfractionation of a mouse lymphoma and macrophage cell line has partially resolved the phosphotransferase, phosphodiesterase, and Gal transferase (34a). There is also a further suggestion that mannosidase I and the phosphotransferase are localized to the same Golgi subcompartment-in 3H-Man pulse-chase studies the degree of trimming to a Man, structure correlates with the progressive addition of blocked phosphates (34). Since lysosomal enzymes also bear complex oligosaccharides (34, 45, 124), they must pass via the terminal sugar transferases (i.e., anatomically distal cisternae), and indeed they can be visualized within all Golgi cisternae by histochemical methods. This route of passage is acutely manifest in the I-cell syndrome in which GlcNAc- 1-phosphotransferase activity is missing and many lysosomal enzymes are released from cells. Such molecules contain more sialic acid and are bound to insolubilized rich to a greater extent than in the wild type (124, 134a). A somewhat analogous situation is also observed for a CHO cell mutant in which lysosomal enzymes bear only complex oligosaccharides (7 1a). Also consistent with this line of reasoning is a report that the mannose-phosphate receptor-which is thought to mediate the ICT of newly synthesized lysosomal enzymes-bears sialic acid; however, the receptor molecules isolated may largely originate from the cell surface rather than from the GC (129). How does the cell determine whether or not to add the blocked phosphate to a given oligosaccharide? Two points should be made. (1) The signals have not been identified, but when the GlcNAc-phosphotransferase is presented with any of several high mannose secretory proteins or lysosomal enzymes, transfer is overwhelmingly more active toward the lysosomal enzymes (106, 144), and (2) the studies on processing of lysosomal enzymes once again indicate the existence of microheterogeneity: the degree of addition of blocked phosphate, of unblocking, and of addition of terminal sugars is variable. In yeast, where lysosomal enzyme oligosaccharides are not essential for targeting to the lysosome (1201, lysosomal carboxypeptidase Y is phosphorylated while in the RER (131). Its carbohydrate is processed in two stages while tranversing the GC-temperature-sensitive ICT mutants blocked at the level of the GC accumulate one of two endoglucosamindiase H-sensitive species (1 3 1). Data on the influence of monensin on lysosomal enzyme transport are suggestive. In I-cell fibroblasts monensin reduces hexosaminidase secretion ( 143b). In normal cells it may actually stimulate their release (100a). It is striking that chloroquine, which also raises intralysosomal pH, also causes release of newly synthesized lysosomal enzymes (124). This effect can be attributed to neutralization of lysosomal content by the ionophore and a consequent tying up of the

CONFINED FUNCTION MODEL OF THE GOLGI COMPLEX

24 1

mannose-phosphate receptor: however, the situation has not yet been fully analyzed. The secreted lysosomal enzymes are phosphorylated (100a). 3. Lipid Addition to Membrane Proteins (117, 118) The presence of covalently bound lipid is not restricted to proteins which traverse the secretory pathway; however, a number of plasma membrane glycoproteins have been shown to acquire fatty acids posttranslationally. The best studied cases are the envelope glycoproteins of Vesicular Stomatitis virus, Semliki Forest virus, and Sindbis virus, the proteolipid of myelin, the transfemn receptor, HLA antigens, and T200, a lymphocyte surface antigen. The covalent sites of fatty acid linkage are suspected to be cysteine or serine residues. The corresponding enzymes and activated fatty acid precursors have not been identified. Four lines of experimentation suggest that an early Golgi subcompartment is the site of palmitate addition in virally infected cells. (1) Pulse-chase studies employing labeled palmitic acid show that the half-time required by cycloheximide to shut off acylation of viral proteins is 15 minutes. Hence acylation is certainly not a cotranslational event ( I 19). (2) Within only 3 to 6 minutes after acylation these proteins become resistent to endo H (VSV) or undergo proteolytic cleavage (Sindbis), whereas when amino acids are used to label the same glycoproteins, the half-time for acquisition of endo H resistance is 20-30 minutes (1 19). (3) Subcellular fractionation of VSV-infected CHO cells pulse-labeled with [3H]palmitateshows that acylated G is coisopycnic with mannosidase I activity (20). In a closely related system, however, this isopycnic coincidence is eliminated when the density of "medial" Golgi elements containing the acylated viral proteins is increased by infecting the cells with Semliki Forest virus and then treating with monensin (101). In this situation mannosidose I is unperturbed and remains coisopycnic with galactosyl transferase. (4) The acylation of these glycoproteins is not blocked by monensin (54). It is curious that in yeast ICT mutants acylation has been assigned to the RER (M. Schlesinger, personal communication). Since the phosphorylation of yeast lysosomal enzymes also occurs in such RER-blocked mutants (131), but has been localized to preparative liver Golgi fractions (71), one wonders whether the conventional anatomic boundary between RER and GC corresponds to the mutant phenotypes. In higher organisms a partial enzymologic overlap has been noticed between the RER membranes and the GC (49, 53, 111) and electron microscopic lectin-binding studies (36, 135, 138) suggest that it is the anatomically proximal Golgi cisternae that resemble the RER. Moreover, as already mentioned, monensin experiments have assigned both acylation and lysosomal enzyme phosphorylation to an early subcompartment. Perhaps in the yeast ICT

242

ALAN M. TARTAKOFF

mutants in question the accumulated unit includes what is conventionally considered to be the proximal face of the GC as well as the RER. 4. Keratan Sulfate I

An additional modification of asparagine-linkedoligosaccharidesoccurs in the cornea where they may be processed to corneal keratan sulfate, a rather small (70 kilodalton) proteoglycan characterized by a repeating sulfated (GlcNAc, Gal) disaccharide. Since keratan sulfate I biosynthesis can be blocked by tunicamycin (43), these units appear to be the result of an alternate derivatization of the same trimmed oligosaccharide which can be processed to a sialic acid-containing complex oligosaccharide. Indeed, the linkage region is identical to that of the complex sites-only three mannose residues remain and these bear GlcNAc and Gal followed by the repeating disaccharide (89).

D. INITIATION AND PROCESSING OF N-ACTYLGALACTOSAMINE-LINKED OLIGOS ACCHARIDES Present data are largely in favor of the notion that the addition of the most internal GalNAc to serine or threonine takes place in the GC, long after polypeptide chain termination. The data derive from the distribution of GalNAc transferase activity in subcellular fractions of intestinal mucosa (61), and oviduct (42), from study of the impact of cycloheximide on GalNAc addition to mammary glycoproteins (146), and from the comparative evaluation of the kinetics of GalNAc and GlcNAc addition to O-linked and N-linked units, respectively (41). The result of glycosylation is most impressive for mucins-hundreds of oligosaccharides are added and the product may be greater than 80% carbohydrate (1 1). The enzymes responsible for chain elongation have been studied in sufficient detail as to rationalize the set of structures observed. It has been reported that neuraminic acid of mucins can be acetylated and that acetyl groups can be replaced by glycolyl groups after their addition to mucin ( 1 16). Two approaches localize mucin-specific transferase activities to Golgi subcompartments: (1) an autoradiographic study of [3H]Gal pulse-labeled mucus secreting cells which indicates that incorporation takes place over distal cisternae and associated vesicles (72), and (2) a cytochemical study of thin sections of goblet cells stained with lectin-gold conjugates to detect nonreducing terminal GalNAc (109). Proximal cisternae are stained and medial cisternae negative. Only in blood group A + material (where terminal GalNAc is present) arc: the distal cisternae and secretion granules also stained. Both studies are, therefore, consistent with the idea of progressive sugar addition during proximal-to-distal transport across the Golgi stack. A single investigation indicates that the subcellular site of incorporation of the

CONFINED FUNCTION MODEL OF THE GOLGl COMPLEX

243

S:Serlna(Thnonlm)

SIV

FIG. 3. Structural and biosynthetic relations among common GalNAc-linked oligosaccharides. The four basic core structures (I-IV) [cf. ( I l S ) ] are derivatized to generate the units found on membrane glycoproteins, hormones, mucins. and keratan sulfate. The blood group H determinant is Fuca( I-2)Gal-, the A determinant is GalNAc( I -3)[Fuca( I -2)]GalP-. The four structures indicated by the bracket "blood group antigens" are precursors of human antigens. Competition of transferases of overlapping specificities explains certain choices among the alternate biosynthetic paths. IgA, Immunoglobulin A; hCG, human chorionic gonadotropin. +

internal and peripheral 0-linked GalNAc is (kinetically) near the site of addition of peripheral GlcNAc to N-linked oligosaccharides (4I). Tissue culture cells synthesizing the beta subunit of human chorionic gonadotropin (which bears both 0-and N-linked oligosaccharides) are labeled with [3H]GlcN for periods up to 3 hours. Successive samples of secretion are collected and analyzed for 0-linked [3H]GalNAc, and N-linked peripheral and internal [3H]GlcNAc. The kinetics of labeling of the first two are rapid and similar. The internal GlcNAc is labeled only after a lag which is considered a measure of the time required for exit from the RER to the cell surface. The best characterized surface glycoprotein of the human red blood cell, glycophorin, bears 15 0-linked tetrasaccharides and one N-linked oligosac-

244

ALAN

M.TARTAKOFF

charide. The O-linked units, which begin with GalNAc, should serve as a model for step-by-step analysis of chain elongation, as should the O-linked oligosaccharides of the envelope glycoproteins of murine corona virus. The incomplete biosynthetic data suggest, paradoxically, that O-linked oligosaccharidesare added when the glycophorin message is translated in the presence of dog pancrease microsomes (55). The only published data on the corona virus show that very long-term monensin treatment blocks the addition of O-linked sugars (88). A similar observation has been made in the study of the influence of monensin on the biosynthesis of the O-linked oligosaccharides of Herpes virus glycoproteins (54a). Skeletal keratan sulfate (keratan sulfate 11) is produced by chrondrocytes of mature cartilage and contains the same repeating disaccharide as the N-linked variety (keratan sulfate I). In this case, the saccharide is linked via GalNAc to serine or threonine residues of the core protein. By analogy with data discussed above, one would expect both chain initiation and elongation to occur in the GC. Since the same protein core contains short mucin-like O-linked oligosaccharides and since the distribution of these units is not random there must be signals indicating which serine or threonine residues are to receive the repeating disaccharide chains (44,79).

E. GLYCOLIPID CHAINELONGATION The abundant and diverse classes of glycolipids have received little attention in the context of ICT, although glycolipids are enriched in plasma membrane and Golgi-rich fractions. Nevertheless, a unique elegant biosynthetic study of cultured neuroblastoma cells shows that several gangliosides are transported from an intracellular site (presumably the GC) to the cell surface with a half-time of 20 minutes (82). Following the addition of the most internal glucose or galactose to ceramide, a broad range of further processing pathways may be available, according to the tissue in question. The sugar transferases involved are thought to be largely distinct from the transferases involved in glycoprotein synthesis. Glycolipid sugar transferases responsible for synthesis of GM3, GM2, GMl, and GD1 have been found enriched in preparative Golgi fractions of rat liver, kidney, and bovine thyroid (22, 58, 96); however, it is striking that the activity responsible for the conversion of GM3 to GD3 is not especially enriched in the thyroid fractions (Fig. 4).Recent [3H]Gal pulse-chase studies of rat glioma cells show a precursor-product relation between glucosylceramide and more complex glycosphingolipids. This conversion is effectively blocked by monensin, suggesting that it involves transport from the early part of the GC (or RER) to a later region. In strict analogy with studies of secretion of [3H]Gal-labeled lg (139), [3H]Gal-labeled gangliosides are rapidly transported to the cell surface in the

245

CONFINED FUNCTION MODEL OF THE GOLGI COMPLEX

2

/2

OM2

OM1

\

c-

C-

OM3

1

C-

du-w Gma

c - H

C:Ceramide

C-

I'

J

C-

m

OD3

C-

L

OD2

-

c-

& c91,

FIG.4. Structural and biosynthetic relations among glycolipids. After glucose addition to ceramide several paths can be followed which result in the globo, muco, lacto, and ganglio series. Only some of the simplest structures are indicated. The ganglioside biosynthetic pathway is further detailed. It is surprising that enzyme ( I ) is not highly enriched in Golgi fractions. The conversion of GM3 to GM2 (enzyme 2) occurs with high efficiency in neural tissue. GM3 is the major glycosphingolipid of nonneural tissues.

presence of monensin (P. Fishman and H. Miller-Podraza, personal communication). Studies of the distribution of sulfotransferases involved in sulfatide synthesis have also localized these enzymes to preparative Golgi fractions of testes and kidney (27-29, 70). A single study demonstrates transport of sulfatides from GC to the cell surface (28). Since the methods for detailed analysis of glycolipids are now relatively streamlined and since the biological importance of glycolipids is increasingly apparent, further data on their ICT should be forthcoming.

IV. Noncovalent Modifications In the hepatocyte secretory lipoprotein particles are visible within the smooth endoplasmic reticulum and become conspicuous within the GC, especially after ethanol intoxication (1 28). The secreted particles contain noncovalently bound triglycerides, phospholipids, and free and esterified cholesterol. Recent studies of isolated Golgi fractions suggest that the phospholipids of the contained lipoproteins are in rapid equilibrium with the phospholipids of the membranes of

246

ALAN M. TARTAKOFF

Golgi cisternae (52). It therefore appears likely that it is within the GC that they acquire most of their lipids. Nevertheless, since lipoprotein particles recovered from rat and chicken liver Golgi fractions do not have the same lipid composition as in serum, further intra- or extracellular maturation events must be involved (4, 52). In murine macrophages the binding of apolipoprotein E to cholesterol occurs only after discharge from the cell (6). Secretory cells may be divided into two classes: the nonregulated class (e.g., plasma cells, fibroblasts) and the regulated class (e.g., pancreatic and pituitary cell types). Only members of the regulated class store their secretory products in granules. This content is released by exocytosis upon physiologic stimulation. The genesis of such secretory granules and lysosomes takes place either within distal Golgi cisternae or in vacuoles (e.g., “condensing vacuoles”) in their immediate vicinity, according to cell type and physiologic conditions (23, 24). Such impressive condensation of content is in many cases thought to reflect the electrostatic interaction between secretory products, for example, between secretory proteins and proteoglycans (24,77, 142). Presumably the site of visible condensation corresponds to the region of the GC within which the proteoglycan is synthesized (e.g., sulfated) or to a later region. As already mentioned, experiments employing monensin indicate that sulfation of chondroitin sulfate is a rather late Golgi event. The proteoglycans of secretory granules have been reported to be heparin, heparan, dermatan, and chondroitin sulfate (60, 104, 142, 150). A fundamental question pertaining to the organization of the GC should also be considered. A wealth of data demonstrates that the terminal sugar transferases, for example, are restricted to a subregion of the GC. Why do these proteins, synthesized in the RER, exit only as far as this portion of the GC and migrate, if at all, only much more slowly to the cell surface? Several explanations might be invoked: (1) the ionic content of the GC is unlike that the the RER and as a consequence a conformational change is induced in such membrane proteins which reduces their mobility-the possibility that the Golgi content may have unique ionic properties has recently been discussed (136), or (2) the protein encounters, within the membrane of Golgi cisternae, certain stable residents (proteins?) for which it has a high affinity, or (3) a posttranslational covalent modification executed within the GC is responsible. To date, no experimental means have been identified which cause mislocalization of such Golgi markers and no information is available to indicate whether one of the three proposed explanations might be correct.

V. Consequences of Processing for the Golgi Complex and the Cell as a Whole The point of departure for this article was the supposition that the primary function of the GC is to accomplish net unidirectional ICT. It is presumably

CONFINED FUNCTION MODEL OF THE GOLGI COMPLEX

241

through evolution that the GC has acquired an impressive array of other responsibilities. These activities are the subject of the preceding pages and are summarized in Table I. What is the impact of housing these processing events on the GC, and what are some of the dividends for the cell? 1. The GC has become such an exceedingly active center for sugar addition that the content of the cisternae must be a seething bath of activated sugar nucleotides. These species, to the extent that they have been investigated, gain access to the cisternal space by facilitated transport through specific sites (13, 16, 30, 126). This set of transporters may be considered as additional set of Golgi markers. 2. The GC must deal with the side products of these reactions-the mono- and dinucleosides which are produced by the transferases-especially since UDP inhibits GalNAc and Gal transferase hyaluronate synthetase (59,95, 100b). Both the UDP and GDP and to a lesser extent CMP are broken down within the cisternal space by nucleoside diphosphatase (9, 95). Furthermore, the end-products of hydrolysis must exit from the cisternae in order to function in repetitive cycles. Only a single report pertains to this second class of carriers (30). In the same vein, the ability of the GC to synthesize lactose (in the mammary gland) and the presence of both mannosidase and phosphodiesterase activities suggests that Golgi membranes are impermeable to disaccharides yet permeable to monosaccharides. Experimental observations on the mammary gland and myeloma cells suggest that this is the case (135, 145). 3. The ability to assemble proteoglycans may be considered a prerequisite for the existence of the regulated secretory cell phenotype. 4. The addition of both neutral and negatively charged sugars to intrisic glycoproteins and glycolipids of the GC obviously modifies the environment in which other Golgi activities occur. For example, the local membrane potential may be perturbed at sites where sialic acid has been added, and the presence of ceramide derivatives may result in membrane rigidification ( 148). 5 . The assembly of membrane-bound oligosaccharides clearly allows the cell to endow its surface with negative charge and an array of specific structural sites beyond those simply encoded in its polypeptides. 6. The generation of mannose phosphate groups, in many cells, is essential for targeting of lysosomal enzymes. Such considerations indicate that the existence of posttranslational processing events is an integral part of the cells repertoire of biosynthetic equipment. Given this realization, it would not be altogether suprising to learn that nucleoside diphosphatase, for example, had subsumed certain functions at the level or overall Golgi structrue and organization. Other proteins involved in more specialized processing events, e.g., select glycolipid sugar transferases, found only in the GC of certain cells, might not be expected to have acquired such dual responsibilities.

248

ALAN M. TARTAKOFF

Other common features of processing deserve comment, for example, the microheterogeneity which is commonplace among glycoproteins and may also be characteristicof proteolytic cleavage. No experimental studies have attempted to manipulate the degree of microheterogeneity, e .g., by reducing the temperature so as to slow transport through the GC. Nevertheless, in addition to the enzymologic considerations mentioned above (overlapping transferase specificities, etc.), one wonders whether dispersion in the rates of entry and exit and duration of residence within a given subcompartment may not be responsible. In this sense, the length of an oligosaccharide chain, e.g., of chondroitin sulfate, might be considered an indirect measure of the kinetics of transport. For the moment, knowledge of the sequences of posttranslational processing events can explain the final structures of many macromolecules which pass through the GC. This article has summarized the incomplete data now available which points to subcompartmentalizationof these events. It is to be anticipated that as the assignment of specific steps to subcompartments is made with greater certainty, this information will lead to an increased appreciation and understanding of the dynamics of membrane-membrane interactions which underly the operation of the organelle as a whole.

ACKNOWLEDGEMENTS Supported by Grant 3.059-1.81 of the Swiss National Science Foundation,

REFERENCES I . Anderson, D., and Grimes, E. (1982). J . Biol. Chem. 257, 14858-14864. 2. Appel, A., Horwitz, A., and Dorfman, A. (1979). J . Biol. Chem. 254, 12199-12203. 3. Ballou, C. (1976). Adv. Microbiof. Physiol. 14, 93-158. 4. Banerjee, D., and Redman, C. (1983). J . Cell Biol. 96, 651-650. 5 . Barland, P., Smith, C., and Hamerman, D. (1968). J. Cell Biol. 37, 13-26. 6. Basu, S., Goldstein, J., and Brown, M.(1983). Science 219, 871-873. 7. Beyer, T., Sadler, J., Rearick, J., Paulson, J., and Hill, R. (1981). Adv. Enzymol. 52,24-175. 7a. Bosch, J., Schwarz, R. Ziemiecki, A., and Friis, R . (1982). Virology 119, 122-132. 8 . Bracha, M.,and Schlesinger, M. (1976). Virology 74, 441-449. 9. Brandan, E., and Fleischer, B. (1982). Biochemistry 21,4640-4645. 10. Bretz, R., Bretz, H., and Palade, G. (1980). J. Cell Biol. 84, 87-101. 11. Briand, J., Andrews, S., Calik, E., Conway, N., and Young, J. (1981). J. Biol. Chem. 256, 12205-12207. 12. Bumol, T., and Reisfeld, R. (1982). Proc. Nurl. Acud. Sci. U.S.A. 79, 1245-1249. 13. Carey, D., Sommers, L., and Hirschberg, C. (1980). Cell 19, 597-605. 13a. Clermont, Y., Lalli, M., and Rambourg, A. (1981). Am. J. Anar. 201, 613-622. 14. Comb, M.,Seeburg, P., Adelman, J., Eiden, L., and Herbert, E. (1982). Nurure (London) 295, 663-606. 15. Coudron, C., Koerner, T., Jacobson, I., Roden, L., and Schwartz, N. (1980). Fed. Proc. 39, 1671.

CONFINED FUNCTION MODEL OF THE GOLGI COMPLEX

249

16. Creek, K., and M o d , D. (1981). Biochim. Biophys. Actu 643, 292-305. 17. Crine, P., and Dufour, L. (1982). Biochem. Biophys. Res. Commun. 109, 500-506. 18. Docherty, K., and Steiner, D. (1982). Annu. Rev. Physiol. 44, 625-638. 19. Dorfman, A. (1982). In “Cell Biology of the Extracellular Matrix” (E. Hay, Ed.). Plenum, New York. 20. Dunphy, W.,Fries, E., Urbani, L., and Rothman, J. (1981). Proc. Nurl. Acud. Sci. U.S.A. 78, 7453-7457. 21. Esmon, B., Novick, P., and Schekman, P. (1981). Cell 25, 452-460. 22. Eppler, C., M o d , D., and Keenan, T. (1980). Biochim. Biophys. Actu 619, 318-331. 23. Farquhar, M. (1978). I n “Transport of Macromolecules in Cellular Systems,” pp. 341-362. Dahlem Life Sciences Report 11. 24. Farquhar, M., and Palade, G. (1981). J . Cell Biol. 91, 7 7 ~ 1 0 3 s . 25. Fellini, S., Kimura, J., and Hascall, V. (1981). J . Cell Biol. 256, 7883-7889. 26. Fishman, P.. and Brady, R. (1976). Science 194, 906-915. 26a. Fishman, P., and Miller-Podraza, H. (1983). In press. 27. Fleischer, B. (1977). J Suprumol. Strucr. 7, 79-89. 28. Fleischer, B. (1978). I n “Cell Surface Carbohydrate Chemistry,” pp. 27-47. Academic Press, New York. 29. Fleischer, B., and Zambrano, F. (1974). J . Biol. Chem. 249, 5995-6003. 30. Fleischer, B., and Brandan, E. (1983).I n “Structural Carbohydrates in the Liver” (H. Popper, F. Gudat, E. Kottgen, and W.Rentter, Eds.). MTP Press, Lancaster, England. 30a. Fries, E., and Rothman, I. (1981). J . Cell Biol. 90, 697-704. 31. Gabel, C., Goldberg, D., and Kornfeld, S . (1982). J. Cell Biol. 95, 536-542. 32. Garten, W.,Bosch, F., Linder, D., Rott, R., and Klenk, H.-D. (1981). Virology 115, 361-374. 32a. Glembotski, C. (1981). J. Biol. Chem. 256, 7433-7439. 33. Godman, G., and Lane, N. (1964). J . Cell Biol. 21, 353-356. 34. Goldberg, D., and Kornfeld, S. (1981). J. Biol. Chem. 256, 13060-13067. 34a. Goldberg, D., and Kornfeld, S. (1983). J. Biol. Chem. 258, 3159-3165. 35. Goldfischer, S. (1982), J. Hisrochem. Cytochem. 30, 717-733. 35a. Green, J., Griffiths, G., Louvard, G., Quinn, P., and Warren,G. (1981). J. Mol. Biol. 152, 663-698. 36. Griffiths, G., Brands, R., Louvard, D., and Warren,G. (1983). J. Cell Biol. 95, 781-792. 37. Griffiths, G., Quinn, P., and Warren,G. (1983). J . Cell Biol. 96, 835-850. 38. Gumbiner, B., and Kelly, R. (1981). Proc. Nutl. Acud. Sci. U.S.A. 78, 318-322. 39. Hakamori, S. (1981). Annu. Rev. Biochem. 50, 733-764. 40. Hand, A., and Oliver, C. (1981). Methods Cell Biol. 23, 137-153. 41. Hanover, J., Eking, J., Mintz, G., and Lennarz, W. (1982). J. Biol. Chem. 257, 10172-10177. 42. Hanover, J., Lennarz, W.,and Young, J. (1980). J. Biol. Chem. 255, 6713-6718. 43. Hart, G., and Lennarz, W.(1978). J . Biol. Chem. 253, 5795-5801. 44. Hascall, V., and Ifascall, G.(1982). In “Cell Biology of the Extracellular Matrix” (E. Hay, Ed.). Plenum, New York. 45. Hasilik, A., and von Figura, K. (1981). Eur. J . Biochem. 121, 125-129. 46. Herbert, E., and Uhler, M. (1982). Cell 30, 1-2. 47. Herzog, V. (1980). I n “Biological Chemistry of Organelle Formation” (T. Biicherm W. Sebald, and H. Weiss, eds.), pp. 119-145. 48. Hinman, B.. and Herbert, E. (1980). Biochemistry 19, 5395-5402. 49. Hino, Y., Asano, A., and Sato, R. (1978). Biochem. J. (Tokyo) 83, 909-942. 50. Holzer, H., and Heinrich, P. (1980). Annu. Rev. Biochem. 49, 63-91. 51. Horwitz, A., and Dorfman, A. (1971). J. Cell Biol. 38, 358-368.

250

ALAN M. TARTAKOFF

52. Howell, K., and Palade, G. (1982). J . Cell Biol. 92, 833-845. 53. Ito, A,, and Palade, G. (1978). J. Cell Biol. 79, 590-597. 54. Johnson, D., and Schlesinger, M, (1980). Virology 103, 407-424. 54a. Johnson, D., and Spear, P. (1983). Cell 32, 987-997. 55. Jokinen, M.. Ulmanen, I., Andersson, L. Kairiainen. L.. and Gahmberg, C. (1981). Eur. J . Biochem. 114, 393-397. 56. K&&riainen,L., Hasimoto, K., Saraste. J., Virtanen, I., and Pentinnen, K. (1980). J . Cell Biol. 87, 783-791. 57. Kajiwara, T., and Tanzer, M. (1982). Arch. Biochem. Eiophys. 214, 51-55. 58. Keenan, T., Morri, D., and Basu, S . (1974). J . Biol. Chem. 249, 310-315. 59. Khatra, B., Hemes, D., and Brew, K. (1974). Eur. J. Biochem. 44,537-560. 60. Kiang, W.-L., KNsius, T.. Finne, I., Margolis, R., and Margolis, R. (1982). J . B i d . Chem. 257, 1651-1659. 61. Kim, Y., Perdomo, J., and Nordberg, J. (1971) J . Biol. Chem. 246, 5466-5476. 62. Kimura, J., Osdoby, P., Caplan, A., and Hascall, V. (1 978). J . Biol. Chem. 253,472 1-47 19. 63. Kimura, I . , Hardingham, T., Hascall, V., and Solursch. M. (1979). J . B i d . Chern. 254, 2600-2609. 64. Kimura, J., Caputo, C., and Hascall, V. (1981). J . Biol. Chem. 256, 4368-4376. 65. Kimura, J., Ihonam, E., Hascall, V . , Reiner, A,, and Poole, A. (1981). J. Biol. Chern. 256, 7890-7897. 66. Kjellen, L., Petterson, I . , and Ht%k, M. (1981). Proc. Narl. Acad. Sci. U.S.A. 78, 5372-5375. 67. Kleine, T. (1981). Inr. Rev. Connec. Tissue Res. 19, 27-98. 68. Klenk, H.-D., and Rott, R. (1980). Current Top. Microbiol. Immunol. 90, 19-48. 69. Loh, Y ., and Gaines, H. (1982). Proc. Nad. Acad. Sci. U.S.A. 79, 108-1 12. 70. Kornblatt, M., Knapp, A., Leinine, M., Schachter, H., and Murray, R. (1974). Can. J. Biochem. 52, 689. 71. Kornfeld, S. (1982) In “The Glycoconjugates 111,” pp. 3-23. Academic Press, New York. 71a. Krag, S., and Robbins, A. (1982). J. Bid. Chem. 257, 842443431, 72. Kramer, M. and Geuze, J. (1980). J . Hisrochem. Cyrochem. 28, 381-387. 73. Kuhn, N., and White, A. (1977). Biochem. J. 168, 423-433. 74. Land, H., Schutz, G., Schmale, and Richter, D. (1982). Nature (London) 295, 299-303. 75. Li, Y.-T., and Li, S.-C. (1982). Adv. Carbohydr. Chem. Biochem. 40,235-286. 76. Lindahl, V., HMk, M., Backstrom, G., Jacobsson, I., Riesenfeld, 1.. Malmstrom, A., Rodin, L., and Feingold, D. (1977). Fed. Proc. 36, 19-23. 77. Lindahl, V., and Hook, M. (1978). Annu. Rev. Biochem. 47, 385-417. 78. Lohmander, L., de Luca, S.. Hascall. V., Caputo, C., Kimura, J., and Heinegard, D. (1980). J . Biol. Chem. 255, 6084-6091. 79. de Luca, S., Lohmander, L., Nilsson, B., Hascall, V., and Caplan, A. (1980). J . Biol. Chem. 255, 6077-6073. 79a. Madoff, D., and Lenard, J. (1982). Cell 28, 821-829. 80. Mason, R., d’Annille, L., Kimura, J., and Hascall, V. (1982). Biochem. J. 207, 445-457. 81. Melchers, F. (1970). Biochem, J . 119, 765-772. 82. Miller-Podraza, H., and Fishman, P. (1982). Biochemistry 21, 3265-3270. 83. Mitchell, D.,and Hardingham. T. (1981). Biochem. J . 1%. 521-529. 84. Mitchell, D., and Hardingham, T. (1982). Biochem. J . 202, 249-254. 85. M o d , D., and Ovtracht, L. (1977). Inr. Rev. Cyrol. Suppl. 5, 61-188. 86. Neufeld, E., and Ashwell, G. (1980). “Biochemistry of Glycoproteins and Proteoglycans,” pp. 242-266. Plenum Press, New York. 87. Neutra, M., and Leblond, C. (1966). J . Cell B i d . 30, 137-150.

CONFINED FUNCTION MODEL OF THE GOLGI COMPLEX

25 1

88. Niemann, H.. Boscher, B., Evans, D., Rosing, M., Tamura, T., and Klenk, H.-D. (1982). EMBO J. 1, 1499-1504. 89. Nilsson, B., Nakazawa, K., Hassell, J., Newsome, D., and Hascall, V. (1983).J. Biol. Chem. 258, 6056-6063. 90. Nishimoto, S.. Kajiwara, T., Ledger, P., and Tanzer, M. (1982). J. Eiol. Chem. 257, 11712-1 1716. 91. Nishimoto, S . , Kajiwara, T., and Tanzer. M. (1982). J. Eiol. Chem. 257, 10558-10561. 92. Oda, K., Ikehara, Y., and Kato, K . (1978). Biochem. Eiophys. Res. Commun. 536, 97-105. 93. Oda. K.. and Kato, Y. (1982). Biochern. Biophys. Res. Commun. 105, 766-772. 94. Oda, K., and Ikehara, Y. (1982). Biochem. Biophys. Acra 107, 561-567. 95. Ohkubo, I., Ishibashi, T., Taniguchi, N., and Makita, A. (1980). Eur. J. Biochem. 112, 1 1 1-1 18.

96. Pacuszka, T., Duffard, R.. Nishimura, R.. Brady, R., and Fishman, P. (1978). J. Biol. Chem. 253, 5839-5846. 97. Parkhouse, R., and Melchers, F. (1971). Biochem. J. 125, 235-240. 98. Pesonen, M., and Kaariainen, L. (1982). J. Mol. Biol. 158, 213-230. 99. Pestalozzi, D., Hess, M., and Berger, E. (1982). J. Histochem. Cytochem. 30, 1146- 1152. 100. Pohlmann. R., Waheed, A., Hasilik, A., and von Figura, K. (1982). J. Biol. Chem. 257, 5323-5325. 100a. Pohlmann, R., Druger, S., Hasilik, A,, and von Figura. K. (1983). In press. 101. Quinn, P., Griffiths, G . . and Warren, G . (1983). J. Cell Biol. 96, 851-856. 102. Redman. C., Banerjee, D., Manning, C., Huang. C., and Green, K. (1978). J. Cell Biol. 77, 400-416. 103. Redman, C., Banerjee, S., and Yu,S. (1981). Methods CellBiol. 23, 231-245. 104. Reggio, H., and Palade, G. (1978). J . CellEiol. 77, 288-314. 105. Regoeczi. E., Chindemi, P., Debanne, M., and Charlwood, P. (1982). Proc. Nut/. Acad. Sci. U.S.A. 79, 2226-2230. 106. Reitman, M., and Kornfeld, S. (1981). J. Eiol. Chem. 256, 11977-11980. 107. RodCn. L. (1980). In “Biochemistry of Glycoproteins and Proteoglvcans,” pp. 267-371. Plenum, New York. 108. Rogers, G . , Gruenebaum, J., and Boime, I. (1982). J. Biol. Chem. 257, 4179-4186. 109. Roth, J. (personal communication). 110. Roth, J., and Berger, E. (1982). J. Cell Biol. 93, 223-229. 1 1 1 . Rothman, J. (1981). Science 213, 1212. 112. Rothman, J . , and Fries, E. (1981). J . Cell Biol. 89, 162-168. 113. Schachter, H., Narasimhan, S.. Gleeson, P.. Vella, G.,and Brockhausen, 1. (1982). Phil. Trans. R . Soc. London 8300, 145-159. 114. Schachter. H.. and Roseman, S. (1980). In “Biochemistry of Glycoproteins and Proteoglycans,” pp. 85-160. Plenum, New York. 115. Schachter. H., and Williams, D. (1982). In “Mucus in Health and Disease” (E. Chantler. J. Edler, and M. Elstein, eds.), Vol. 11, pp, 3-28. Plenum, New York. 116. Schauer, R. (1979). Proc. Int. Svmp. Gl.vcoconj. 4th 11, 597-612. 117. Schlesinger, M. (1981). Annu. Rev. Biochem. 50, 193-206. 118. Schmidt. M. (1983). Curr. Topics Microbiol. Immunol. 102, 101-130. 119. Schmidt, M., and Schlesinger, M. (1980). J. Eiol. Chem. 255, 3334-3339. 120. Schwaiger, H., Hasilik. A., von Figura, K., Wiemken, A., and Tanner, W. (1982). Eiochem. Biophys, Res. Commun. 104, 950-956. 121. Schwartz, N. (1975). FEES Lett. 49, 342-345. 121. Scott, J., Orford, C., and Hughes, E. (1981). Eiochem. J. 195, 573-581. 122. Schur, B. (1975). Biochim. Biophys. Acra 415, 473-512.

252

ALAN M. TARTAKOFF

123. Silbert, J., and Freilich, L. (1980). Biochem. J . 190, 307-313. 124. Sly, W., Fischer, H.. Gonzalez-Noriega, A., Grubb, J., and Natowicz, M. (1981). Merhods Cell Biol. 23, 191-214. 124a. Smith, C. (1981). J . Hisrochem. Cyrochem. 29, 822-836. 125. Snider, M.. Huffaker, T., Vouto, J.. and Robbins, P. (1982). Phil. Trans. R. SOC. London B300, 207-223. 126. Sommers, L., and Hirschberg, C. (1982). J . Biol. Chem. 257, 10811-10817. 127. Spiro, R., and Spiro, M. (1982). Phil. Trans. R. SOC. London B300, 9-20. 127a. Srinivas, R., Melsen, L., and Compans. R. (1982). J. Virol. 42, 1067-1075. 128. Stein, O., and Stein, Y. (1967). J . Cell Biol. 33, 319-339. 129. Steiner, A., and Rome, L. (1982). Arch. Biol. Biophys. 214, 681-687. 130. Steiner, D., Kemmler, W., Tager, H., and Petesson, I. (1974). Fed. Proc. 33, 2105-21 15. 131. Stevens, T., Esmon, B., and Schekman, R. (1982). Cell 30, 439-448. 132. Strous, G., and Berger, E. (1982). J. Biol. Chem. 257, 7623-7628. 133. Strous, G.. and Lodish, H. (1980). Cell 22, 709-717. 134. Tartakoff, A. (1980). Inr. Rev. Exp. Parhol. 22, 227-251. 135. Tartakoff, A. (1982). Phil. Trans. R. SOC. London B300, 173-184. 136. Tartakoff, A. (1983). Cell 32, 1026-1028. 137. Tartakoff, A. (1983). Methods Enzymol. 98, in press. 138. Tartakoff, A. (1983). J . Cell Biol. (in press). 139. Tartakoff, A., and Vassalli, P. (1979). J . Cell Biol. 83, 284-299. 140. Telser, A., Robinson, H., and Dorfman, A. (1965). Proc. Narl. Acad. Sci. U.S.A. 54, 9 12-9 19. 141. Thonar,E., Lohmander, S.. Kimura, J.. Fellini, S., Yanagashita, M., and Hascall, V. ( 1983). J . Biol. Chem. (in press). 142. Uvnas, B., and Aborg, C.-H. (1977). Acra Physiol. Scand. 99, 476-483. 143. Vertel, B., and Dorfman, A. (1979). Proc. Nurl. Acad. Sci. U.S.A. 76, 1261-1264. 143a. Vladutiu, G., and Rattazzi, M. (1975). Biochem. Biophys. Res. Commun. 67, 956-964. 144a. Vladutiu, G., and Rattazzi. M. (1980). Biochem. J . 192, 813-820. 144. Waheed, A., Hasilik, A., and von Figura, K . (1982). J. Biol. Chem. 257, 12322-12331. 147. Wibo, M., Thinks-Sempoux,D., Amar-Costesec, A., Beaufay, H., and Godelaine, D. (1981). J . Cell Biol. 89, 456-474. 148. Yamakawa, T., and Nagai, Y. (1978). Trends Biochem. Sci. 3, 128-131. 149. Young, R. (1973). J . Cell Biol. 57, 175-189. 150. Zanini, A., Giannattasio, G., Nussdorfer, G., Margolis. R.,Margolis, R., and Meldolesi, J. (1980). J. Cell Biol. 86, 260-279. 151. Ziemiecki, A., Garoff, H., and Simons, K. (1980). J . Gen. Virol. 50, I 1 1-123.

INTERNATIONAL REVIEW OF CYTOLOGY. VOL. 85

Problems in Water Relations of Plants and Cells PAULJ. KRAMER Department of Botany, Duke Universily, Durham, North Carolina lntrodution ......................... ......... A. The Roles of Water in Plant.. ........................... B. Problems. ....... 11. Terminology. .

253 254 254

........

255

......

257

...........

265

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

..........

266 266 267

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

271

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

278

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

281 282

1.

C.

Matric Water ......................

A. Cell Walls. ............................. B. Cytoplasm and Cytoplasmic Membranes . . . . . C. Vacuoles.. ................................... D. The Symplast Concept. . . . . .

A. Driving Forces.. . . . . . . . . . . . . . . B. Control of Cell Turgor. . . . . . . . . . C. Water Movement in Whole Plants . .

E. Water Movement Outside the Xylem.. . . . . F. Mathematical Models

VI.

B. Metabolic Processes. . . . . . . . Adaptations Increasing Tolerance o A. Postponement of Injury by De

References . . . . . . . .

I. Introduction The importance of water for plant growth is indicated by the fact that the distribution of land plants is more nearly controlled by the availability of water than by any other single environmental factor. Large areas of the earth’s surface 253 Copynght 6 1983 by Academic Press, Inc. All right5 of reproduction in any form reserved. ISBN 0-12-361485-2

254

PAUL J. KRAMER

where temperature would permit plant growth are deserts or semideserts because of lack of water. Even in temperate, humid climates crop yield is reduced more often by lack of water than by any other environmental factor (Boyer, 1982).The agricultural and ecological importance of water indicates its important role in the physiological processes of plants. A. THEROLESOF WATERIN PLANTS

The roles of water can be grouped in four categories: (1) a constituent, (2) a solvent, (3) a reactant, and (4) in the maintenance of turgidity. Water constitutes 80 to 90% of the fresh weight of most herbaceous plants and over 50% of the fresh weight of woody plants, and reduction of water content below some critical level usually results in breakdown of protoplasmic structure and death in all except a few “resurrection plants,” and some seeds and spores (Bewley, 1979; Gaff, 1980). Water is the solvent in which gases, mineral elements, and other solutes enter and leave cells, and it is a reactant in many important processes, including photosynthesis and hydrolytic reactions. Other important roles are the maintenance of the turgor essential for cell enlargement in plant growth, stomata1 opening, tropisms, and movements of leaves, flower petals, tendrils, and other specialized structures. Some of these phenomena require complex control of cell water relations. Turgor also is essential for the maintenance of the form of herbaceous leaves and stems, as shown by the wilting of water-stressed plants. Water deficits cause injury to plants from decreased cell enlargement, closure of stomata, damage to cell membranes, and disturbance of various enzyme-mediated metabolic processes. Plant water relations are dominated by cell water relations and this article will deal chiefly with the water relations of cells and tissues. Although Esau (1965, pp. 8-9) lists eight principal types of cells in plants we will deal chiefly with parenchyma cells and other thin-walled, vacuolated cells because they constitute the largest part of the tissue of herbaceous plants and most of the physiologically active tissue in woody plants. They form the absorbing surfaces in young roots, the evaporating surfaces in leaves, and the storage tissue of stems; and photosynthesis and other metabolic processes occur chiefly in parenchyma cells. Specialized cells such as guard cells and secretory cells present interesting problems in water relations, but they cannot be discussed in detail for lack of space.

B. PROBLEMS Among the interesting problems in cell and tissue water relation are the following: the Occurence of apparent negative turgor pressure; the nature of matric water; the relationship of turgor pressure and metabolism to cell enlargement; the distribution of water in cells; the measurement of cell and plant water status; the

WATER RELATIONS OF PLANTS AND CELLS

255

relative importance of the apoplastic and symplastic pathways for short distance water movement; mechanisms by which water deficits affect metabolic processes; and causes of differences in tolerance of dehydration.

11. Terminology

Before discussing the problems of cell water relations it will be helpful to define some of the terminology. For the purposes of this article a cell can be treated as an osmotic system consisting of a central vacuole surrounded by a layer of cytoplasm and by cytoplasmic membranes that are permeable to water, but relatively impermeable to solutes. This system, the protoplast, is enclosed in a relatively rigid wall that usually is permeable to water and solutes. Diagrams of a young cell and a mature parenchyma cell are shown in Fig. 1. A. WATERPOTENTIAL

The movement of water in and out of cells is controlled by differences in its chemical potential. The chemical potential or partial molal Gibbs free energy is a measure of the capacity of water to do work. The absolute free energy or potential of water is difficult to measure, but the difference in potential between pure, free water and water in cells and solutions can be measured and is termed the water potential, 9 or qw. The derivation of this term is discussed by Dainty (1976), Kramer et al. (1966), and Slatyer (1967, Chap. 1). In plant physiology potentials usually are expressed in units of pressure. In the S1 system pressure is expressed as pascals and lo5 pascals equal 1 bar or 0.1 MPa. As the water potential in cells and solutions is lower than that of free water, qW usually is a negative number. Thus it becomes lower or more negative as the concentration of Porenc hymo

cdl

Nucleus

Cell wll Nucleus Cytoplosrn

Cell woll Plosma mem brone Cytoplasm Vacuolar rnernbrone Vocuole

FIG.I . Diagrams of the meristematic cell and a mature, vacuolated parenchyma cell. The layer of cytoplasm in parenchyma cells often is thinner than shown in this diagram.

256

PAUL J . KRAMER

a solution or the water stress in cells increases, and a water potential of - 1 .O MPa indicates more stress than one of -0.5 MPa, just as a temperature of - 10°C is lower than one of -5°C. The various components of the water potential of cells and tissues can be summarized in the following equation: JIw

=

JIs

+

JIp

+

(1)

JIm

The term JI, refers to the cell or tissue water potential in MPas, and the other terms refer to the contributions made to , ) I by solutes (JI,, or sometimes JI,,J, by pressure (JI,), and by matric forces (Jim or JIJ. JI, and +,, are negative, JI, usually is positive. JI, depends on the concentration of solutes in the vacuole and JI, refers to water bound to surfaces and held in microcapillaries of cell walls. JI,, usually termed turgor pressure, results from inward diffusion of water. It varies from zero in flaccid cells to a pressure equal to the osmotic potential in turgid cells, as shown in Fig. 2. This equation can be written in other ways. However, the essential feature is that the cell or tissue water potential is equal to the difference between the turgor pressure which increases water potential and the osmotic and matric potentials which decrease it.

3 Ful I Turgor

Incipient Plasmolysi s Relative Cell Volume

FIG.2. A Hofler diagram showing the interrelationships among cell volume, osmotic potential, water potential, and pressure potential. The solid lines are for a highly expansible cell. The dashed lines are for a slightly expansible cell, line A representing pressure potential, line B cell water potential. Water potential and osmotic potential are negative. Pressure potential usually is positive, but if the protoplast adheres to the wall during plasmolysis the turgor pressure would hecome negative and VIEellwould be less than Ys. Redrawn from Kramer (1983).

WATER RELATIONS OF PLANTS AND CELLS

257

B. TURGORPRESSURE Turgor pressure is the pressure developed, in cells and tissues by inward The importance of turgor diffusion of water along gradients of decreasing qW. pressure in cell enlargement and the maintenance of form of young stems and leaves will be discussed later. Occasionally the water potential is lower than the measured osmotic potential, resulting in apparent negative turgor pressure. Examples are found in papers by Sionit et al. (1980), Slatyer (1957), and Wenkert (1980). This formerly was attributed to adhesion of cell walls to shrinking protoplasts (Crafts et al., 1949, pp. 83-85), but some investigators attribute it to dilution of the vacuolar sap by cell wall water, producing erroneously higher values for the osmotic potential (Tyree, 1976; Markhart et al., 1981). Although Tyree (1976) originally denied the existence of negative turgor, Tyree and Jarvis (1982) cited data suggesting that it might contribute up to 10% of the cell water potential, but their negative turgor seems to be what is usually termed a matric force. Acock and Grange (1981) argued that circumstantial evidence favors the existence of negative turgor pressure. Wenkert ( 1980) observed water potentials lower than the osmotic potentials measured on expressed sap, and found that the discrepancy could not be fully explained by dilution of vacuolar sap with cell wall water. C. MATRICWATER

This is often termed imbibed water. There is considerable uncertainty concerning the nature and importance of the matric term in Eq. (1). In soil, matric water usually is assumed to include that held by capillary forces in the pores among soil particles as well as that held on the surfaces of particles as water of hydration. Likewise, in cells it usually is assumed to include that held by capillary forces in pores in the cell walls and that bound on the surfaces of various hydrophilic cell components. Some of the water has been termed “bound water” and has been measured arbitrarily as the quantity unavailable to function as a solvent or that which could not be frozen at some temperature such as -20 or -25°C (Kramer, 1955). Tyree and Karamanos (1980) and Tyree and Jarvis (1982) argued that matric water should be restricted to that bound to cell surfaces by short range attractive forces operating over distances of about 2 molecules, and capillary water should be excluded because it can be treated as held by pressure. Acock and Grange (198 1) and Wilson ( 1 967) defended the broader use of the matric term, but Tyree and Jarvis (1982) questioned some of their assumptions. Dainty (1976) also discussed the difficulty in splitting the water potential into its components. Regardless of its exact definition, matric water must tend to come into equilibrium with the osmotically held water and the cell water potential measured by psychrometers must represent the combined matric and osmotic effects on the

258

PAUL J. KRAMER

vapor pressure of all the cell water. However, osmotic and matric potentials are not strictly additive. In experiments where sucrose solution was added to filter paper or cell wall material the total potential of the system did not equal the sum of the osmotic potential of the sucrose solution and the matric potential of the solid material (Boyer and Potter, 1973; Markhart et al., 1981). In parenchyma cells the fraction of water held by matric or imbibitional forces usually is small, but in thick walled tissue, in seeds, and in rneristematic tissue it may constitute a large fraction of the cell water. Acock (1975) and Wilson (1967) claimed that disregard of the matric term in the equation for cell water relations results in errors, but Passioura (1980) reviewed the literature on the status of matric water and recommended that the separate term for it be omitted. There is no reliable way to measure the matric potential separately from the osmotic potential so it has only a symbolic significance in Eq. (l), and the author believes that for practical purposes it can be omitted. This results in a simplified version of Eq. (1): The interrelationshipsbetween total cell water potential (JI,) and the osmotic and pressure terms of Eq. (1) are shown in Fig. 2 for a cell capable of considerable change in volume with change in water content. The dashed lines were added to show the changes for a cell with rigid walls that undergoes little change in volume with decreasing water potential.

111. Cell Structure and Water Relations

Even casual inspection reveals certain unique characteristics of plant cells that affect their functioning. For instance there appears to be more compartmentation of metabolism in plant than in animal cells (Dennis and Miernyk, 1982; Wagner, 1982). Most important with respect to water relations is the almost universal presence of relatively rigid walls, and the presence in parenchyma cells of large central vacuoles. A. CELLWALLS

Cell wall elasticity, permeability, and water content are of interest with respect to plant water relations. Cell expansion and turgor depends on cell wall elasticity, movement of water and solutes depends on wall permeability, and the water stored in cell walls must be taken into account in estimating the osmotic potential. When water diffuses into cell vacuoles the protoplasts tend to expand, somewhat as rubber balloons expand when air is blown into them. The resistance of

WATER RELATIONS OF PLANTS AND CELLS

259

the walls to expansion causes development of turgor pressure in cells, and when water diffuses out of cells the elasticity of their walls causes decrease in volume. Loss of too much water results in loss of turgor and wilting. Cell wall elasticity is expressed as the modulus of elasticity (historically Young’s modulus), E, which is the value relating change in pressure (Af)to change in relative volume (AvIv). The change in turgor pressure accompanying a change in volume depends on the modulus of elasticity, as shown in the following equation:

Af

= E (AvIv).

(3)

Thus the more elastic the cell wall the lower the modulus of elasticity, and it also is lower at low turgor pressures than at high turgor pressures. For example, it was reported in certain studies that E ranged from zero at zero turgor pressure to about 10 MPa at a turgor pressure of 0.5 MPa (Zimmermann and Steudle, 1980). Dainty ( 1976) gave values for E for leaves ranging from less than one to over 10 MPa, and considerably higher values for giant algal cells such as Cham and Nitella. The importance of the modulus of elasticity of cell walls in relation to cell water relations is discussed in detail by Dainty (1976), Tyree and Jarvis (1982), and Zimmermann (1978). If cell walls are relatively inelastic, i.e., have a high modulus of elasticity, a small decrease in water content produces a greater decrease in water potential and in turgor pressure than if they are more elastic. This is shown graphically in Fig. 2. Some investigators have stated that a small decrease in water content for a given decrease in water potential is associated with drought tolerance, but this is debatable. Some conflicting views and results are discussed by Sanchez-Diaz and Kramer (1971). They found that there was a smaller decrease in water content for a given decrease in water potential in sorghum than in corn, which is generally regarded as less drought tolerant than sorghum (Fig. 3). Elastic cell walls result in maintenance of turgor to a lower water content and this should be advantageous. For example, corn leaves lost more of their water than sorghum leaves before the stomata were fully closed (Sanchez-Diaz and Kramer, 1971). The ecological implications of variation in cell wall elasticity are discussed by Osmond et al. (1980, pp. 271-272) and Tyree and Jarvis (1982). Although the development of cell walls is outside of the scope of this article it has some effects on cell water relations. The chemistry of plant cell walls is discussed by Albersheim and by Karr in Bonner and Varner (1976). The primary walls of young cells consist of cellulose, varying amounts of hemicellulose and pectic compounds, and small amounts of protein. Such walls contain 70% or more of water by volume (Markhart et al., 1981) in the microcapillary spaces between cellulose fibrils and are highly permeable to water and solutes. They also shrink significantly when dehydrated, and study of killed, dehydrated tissues probably give a serious underestimate of the volume available for water storage and conduction. The effects of various methods of fixation on cell walls

260

PAUL J . KRAMER Leaf water potential ( MPa 1 -1.0 -1.2 -1.4

-0.2 -0.4 -0.6 -0.8

-1.6

FIG.3. Leaf water saturation deficit in percentage plotted over water potential in MPa for corn and sorghum. There is a much greater decrease in water content for a given decrease in water potential in corn than in sorghum. Redrawn from Sanchez-Diaz and Kramer (1971).

deserves more study. Berlyn (1969) reported that the walls of hydrated uacheids of Pinus resinosa contain about 25% of free space, but this is eliminated by dehydration in absolute alcohol. Walls of young cells are capable of both plastic and elastic expansion, but as they are thickened by deposition of secondary layers of cellulose and sometimes of lignin their expansibility decreases. Deposition of lignin and other materials in the spaces among the cellulose fibrils not only decreases wall elasticity, but also decreases the space available for water storage and movement. Cell walls in bark and some other tissues become suberized and impermeable to water, resulting in death of the protoplasts. There seems to be some uncertainty concerning the maximum size of molecules that can pass through cell walls. This is of interest in connection with the movement of enzymes and other large organic molecules in and through walls. Tepfer and Taylor (1981) stated that cell walls are permeable to proteins with molecular weights up to about 60,000, but Carpita (1982) claimed that the passage of molecules with a molecular weight greater than about 17,000 and a diameter greater than 4 nm is difficult. The former also concluded that dehdration reduces pore size, but the latter claims that it does not. As mentioned earlier, cell walls often shrink measurably when dehydrated, hence it would seem that pore size must be reduced. Apparently, more research is needed on the permeability of cell walls to large molecules. Pore size also is of interest with respect to retention of water in walls under the tension generated in transpiring plants. According to Slatyer (1967, p. 132) the interfibrillar spaces in cell walls range from 100 to 10 nm or smaller in diameter and pores 20 nm in diameter should drain at a water potential of -15.0 MPa which is much lower than that ordinarily developed in plants.

WATER RELATIONS OF PLANTS AND CELLS

26 1

The normal increase in cell wall thickness during the growing season can produce an apparent decrease in water content of leaves expressed as percentage of dry weight, although no decrease in absolute water content actually occurs (see Fig. 4). Wenkert (1980) reported that the ratio of fresh weight to dry weight of soybean leaves decreased from 6 at 20 days to about 3.3 at 80 days. It has been suggested that cell wall water might serve as a buffer against dehydration of the protoplasts (Gaff and Cam, 1961; Teoh et al., 1976). However, Slatyer (1967, p. 177) pointed out that the half-time for equilibration of wall and vacuolar water is about 10 seconds, so there could not be a significant reduction in cell wall or apoplastic water without an immediate reduction in vacuolar or symplastic water. There are other reasons for interest in the amount of water stored in cell walls. When plant tissue is killed the apoplastic water in the cell wall is mixed with the symplastic water, which is chiefly vacuolar sap, and if the former constitutes 10% or more of the total cell water it will produce a significant dilution error in the osmotic potential of the vacuolar sap (Markhart et al., 1981; Tyree, 1976). Boyer and Potter (1973) used a correction factor of 10%to correct thermocouple measurements of osmotic potential of sunflower leaves, and Wenkert (1980) reported that the osmotic potentials of expressed sap from greenhouse and field grown corn averaged 11 and 16% more dilute than measurements made by the pressure-volume method. However, the discrepancy in his measurements could not be explained completely by dilution of the vacuolar sap by apoplastic water. 700

260

- 240

Water at field weight

- 200 t #

1 May

I

Juna

1

I

I

June

July

Aug

'

120

Aug

17 26 7 20 IS S FIG. 4. Seasonal changes in leaf dry weight and weight of water per leaf and in water content as percentage of leaf dry weight, for pear. From data of Ackley (1954).

262

PAUL 1. KRAMER

There also is interest in the amount of water found in cell walls because it constitutes the pathway for movement of solutes in and out of cells and across tissues, especially in roots and leaves. B. CYTOPLASM AND CYTOPLASMIC MEMBRANES The chief role of cytoplasm with respect to cell water relations is to maintain the differentially permeable membranes within which are accumulated the solutes that make cells an osmotic apparatus. In mature parenchyma cells the cytoplasm forms a thin layer lining the walls, sometimes with strands extending across the vacuole, and it constitutes as little as 5% of the total cell volume. However, in meristematic tissue and seeds the vacuoles are small and the cytoplasm occupies most of the volume. The cytoplasm of parenchyma cells is a dilute gel containing 90 or 95% water bound on its protein framework, but in many mature seeds it becomes a rigid gel containing relatively little water. The various organelles in the cytoplasm, including the nucleus, chloroplasts, and mitochondria are enclosed by membranes and function as osmotic subunits. These are discussed in Bonner and Varner (1976). There also is a complex system of internal membranes known as the endoplasmic reticulum. As pointed out earlier, our chief interest is in the plasma membrane that covers the outer surface of the cytoplasm and the vacuolar membrane or tonoplast that encloses the vacuole. The structure and permeability of these membranes and the nature of the transport systems that move solutes across them are of great interest to plant physiologists, but a detailed discussion is beyond the scope of this article. Considerable information about plant cell membranes is given in Volume 2A, new series, of the “Encyclopedia of Plant Physiology.” It will suffice to point out that their functioning depends on maintenance of normal cell metabolism and that such environmental stresses as low and high temperature, high salinity, water deficits, deficient aeration, and various toxic substances such as herbicides, operate at least in part by affecting the structure and permeability of cell membranes. The effects of changes in cell permeability on the bulk water conductance of roots can have important effects on plant water relations. For example, nitrogen deficiency (Radin and Boyer, 1982) and low temperature (Kramer, 1983, Chap. 9) reduce the hydraulic conductance of roots and produce water stress in the shoots. The effect of low temperature on membrane lipids in roots was discussed by Markhart et al. (1980) and chilling effects on membranes in general are discussed in the monograph edited by Lyons et al. (1979). Effects of low temperature on membranes in shoots also is an important factor in chilling injury (McWilliam, 1983). Leakage through cell membranes under standardized conditions is often used as an indicator of injury to plant tissue (Leopold et al., 1981). The permeability or hydraulic conductance (15,) of plant cell membranes is of

WATER RELATIONS OF PLANTS AND CELLS

263

great interest. Measurements have been obtained from swelling and shrinking measurements, the rate of diffusion of isotopes, transcellular osmosis, and nuclear magnetic resonance. Dainty (1976) gives values ranging from l o p 4 to lo-* cm s - l bar- * and Stout et af. (1978) gave a value of 3 X l o p 2for cells of ivy bark. Most of these values are for cells with walls. It is not surprising that Dainty suggested that the values in the literature are questionable.

C. VACULOES The vacuoles of plant cells vary in size from the tiny structures in meristematic cells to the large central vacuoles occupying 50 to 90% of the volume of mature parenchyma cells. They once were regarded chiefly as passive storage places for materials not used in metabolism, but improved methods of isolating and studying vacuoles indicate that they play an important role in metabolism in addition to functioning as osmotic systems and storage compartments for ions and metabolites (Matile, 1978; Wagner, 1982). Vacuoles contain a wide range of substances including inorganic ions, sugars, enzymes, amino and other organic acids, amides, lipids, gums, mucilages, tannins, anthocyanins, and various kinds of crystals such as calcium oxalate. Some of these substances have important roles in metabolism, but others have no known function. Vacuoles are important centers for storage of intermediates in metabolic processes and for hydrolytic enzymes, including proteases, nucleases, exopeptidases, and phosphatases (Wagner, 1982). Kaiser et al. (1982) reported that within 15 minutes after supplying protoplasts from barley leaves with I4CO2, labeled carbon, chiefly in sucrose, was appearing in the vacuoles as rapidly as it was being fixed. Vacuoles are surrounded by membranes termed tonoplasts or vacuolar membranes. They retain differential permeability for some hours after separation from the cytoplasm and seem to be more durable than the plasma membrane. Some investigators report that the vacuolar membrane also is more permeable to water and solutes than the plasma membrane (Kiyosawa and Tazawa, 1977). This apparently differs among plants and Laties and his associates even claimed that the plasma membrane is so permeable to ions that cytoplasm could be regarded as a part of the apoplast (Torii and Laties, 1966). The vacuolar membrane is produced by the cytoplasm which also controls metabolic activities in the vacuoles (Matile, 1978). In this article we are concerned chiefly with vacuoles as osmotic systems that develop the turgor pressure essential for mechanical support of unlignified tissue, for certain movements of plant structures such as the leaves of sensitive plants, for opening of stomata, and for cell expansion. Loss of turgor pressure as a result of dehydration causes cessation of growth and wilting, and we will return to the relationship between turgor and cell expansion later. There has been some speculation concerning the reasons for the survival of

264

PAUL J. KRAMER

large vacuoles under selection pressure during the evolution of plants. FreyWyssling and Muhlethaler ( 1965) suggested that vacuoles are advantageous because it would be difficult for plants to provide enough protein to fill their cells with protoplasm. Wiebe (1978) proposed that vacuolated cells survived in the evolution of plants because they permit development of the maximum amount of root and leaf surface with the minimum use of carbohydrates and nitrogen. Because the concentration of the raw materials used in metabolism such as Co,, minerals, and nitrogen is low, plants benefit from large leaf surfaces exposed to the sun and air and large root surfaces in contact with the soil. The cost of building large roots and shoots from cells composed chiefly of protoplasm probably would severely limit plant size. However, most plant cells are composed of 70 to 90% water by volume and the protoplasm usually forms only a thin layer around the vacuoles and in turn typically is enclosed in thin walls, resulting in maximum size with minimum use of scarce, energy-expensive building materials. D. THESYMPLAST CONCEPT The water in plants often is treated as forming two systems, symplastic water occurring in the protoplasts, and apoplastic water occurring outside of the protoplasts in the cell walls and xylem elements. These terms originated with Munch (1930). Actually water forms a continuous system permeating both walls and protoplasts, and the distinction really applies to the solutes which are confined within the symplast by its boundary membranes, but can move freely through the apoplast by diffusion or mass flow. Apoplastic and symplastic water are discussed in detail by Tyree and Jarvis (1982). 1. Plasmodesmata

Plasmodesmata are complex tubules that extend through cell walls and provide the protoplasmic connections which unite the individual protoplasts into the more or less unified symplast. Their existence has been known for over a century (Tangl, 1879) and their general occurrence is well established. Their structure and functioning are discussed in detail in the monograph edited by Gunning and Robards (1976). Almost from the beginning it has been assumed that they function as pathways for intercellular transport of materials and stimuli (Pfeffer, 1897), although even today the evidence is largely circumstantial. Tyree (1970) claimed that they constitute the path of least resistance for water and salt movement across roots and Robards and Clarkson and other writers in Gunning and Robards (1976) seem to agree. Their role in transport will be discussed later.

E. DISTRIBUTION OF WATERIN CELLS The proportion of the total cell water occurring in various parts of cells varies widely, depending on the kind of tissue and the methods used to estimate it. In

265

WATER RELATIONS OF PLANTS AND CELLS

DISTRIBUTION OF WATER

Plant Eucalyptus leaves Rhododendron leaves Sunflower leaves Wheat leaves Potato leaves Soybean, immature leaves Soybean, mature leaves

IN

TABLE I CELLS AS PERCENTAGES

OF

TOTALWATER CONTENTO

Apoplastic or wall water

Symplastic or vacuolar water

40

(60) (68-75) (86-95) (70) (95) 84 70

25-32

5-14 30

5 (16) (30)

Reference Gaff and Carr (1961) Boyer (1 967) Boyer (1967) Campbell er al. (1979) Campbell er al. (1 979) Wenkert er al. (1978) Wenkert er al. (1978)

UThe values in parentheses are estimates derived from the differences between total water and the measured values. This does not account separately for the water in the cytoplasm, at least part of which is matrically bound and not available as a solvent.

meristematic tissue the walls are thin, the vacuoles are small, and most of the water is in the cytoplasm. As tissues mature the vacuoles enlarge and the walls grow thicker, changing the relative proportions of water in walls, cytoplasm, and vacuoles. For example, the symplastic water in soybean leaves decreased from 84 to 70% of the total as the leaves matured and the cell wall water increased proportionately. The increase in cell wall material in maturing leaves often results in an apparent decrease in leaf water content on a dry weight basis although the absolute amount of water undergoes little change (Ackley, 1954; Weatherley, 1950), as shown in Fig. 4. Some estimates of the distribution of water in leaf cells are given in Table I. F. MEASUREMENT OF PLANTWATERSTATUS The water status of cells usually is estimated by measurements made on masses of tissue such as leaves and represents the average condition of large numbers of cells, often of several types. The average bulk water potential frequently is measured on leaf disks in thermocouple psychrometers, but many measurements are made on leaves or twigs in pressure chambers. The methods are discussed by Kramer (1983, Chap. 12), Slavik (1974), and Turner (1981). The osmotic potential often is measured in a thermocouple psychrometer on the same tissue used to measure the water potential, after freezing it to eliminate the pressure component. This measures the combined matric and osmotic potentials, but the error usually is small. Sometimes osmotic potential is measured on sap expressed from crushed or frozen tissue. The turgor pressure is assumed to be the difference between the total water potential and the osmotic potential. However, as mentioned earlier, when plant tissue is killed the vacuolar sap is diluted by cell wall or apoplastic water, causing the osmotic potential to be somewhat higher

266

PAUL J. KRAMER

than it was in the living tissue. The magnitude of the error depends on the fraction of the total cell water occurring in the wall. This is reported to range from 5 to 40%, depending on the thickness of the walls. This error can be avoided by using the pressure-volume method of Tyree and Hammel(1972), but that method requires more time and tissue than the psychrometer method. Some direct measurements of turgor pressure are being made on single cells with micro pressure probes (Green and Stanton, 1967; Zimmermann, 1978).

IV. Water Movement in Plants Water movement can be conveniently considered at two levels, in cells and tissues such as algae, guard cells, and pulvini, and in whole plants. At both levels movement occurs along gradients of decreasing water potential, but the origin and control of the gradients is somewhat different in the two. A. DRIVINGFORCES In general water movement can be described by a simple equation, analogous to that for the movement of electricity: Water flux = differences in water potential (AWresistance ( r )

(4)

Differences in water potential develop either because of changes in concentration of solutes in cells, or in transpiring plants because of loss of water, or by a combination of the two. The resistances vary in magnitude, depending on the pathway. Water movement from roots to shoots occurs through dead xylem elements that offer relatively low resistance to water movement. However, water movement from root surfaces across the living cells of the cortex to the root xylem, and out of the leaf veins across living cells to the evaporating surfaces presents problems with respect to pathways and resistances.

B. CONTROLOF CELLTURGOR The water relations of cells depends on factors controlling the concentration of solutes in the vacuoles (qS), the permeability of cells to water, and cell turgor. The latter depends on the expansibility of the cell walls. The interrelations of qS, q,,, and qware shown in Fig. 2. When cells lose water and decrease in volume the concentration of solutes increases and the consequent decrease in 'Ps and 'Pw tends to cause entrance of water and correction of the loss of turgor. This constitutes a passive negative feedback control. In addition the concentration of solutes can be increased by active uptake of ions and other solutes and conversion of osmotically inactive organic compounds to osmotically active solutes.

WATER RELATIONS OF PLANTS AND CELLS

267

This is termed osmotic adjustment or osmoregulation (Hellebust, 1976; Turner and Jones, 1980). These changes occur rather slowly in water-stressed plants, often over a period of several days. The general theory of feedback control of cell turgor is discussed by Cram (1976). Other changes in cell turgor, such as those in stomata1 guard cells and in the pulvini of plants exhibiting nyctinastic movements, occur in seconds or minutes rather than in days. They usually seem to involve rapid transport of ions in response to a variety of stimuli operating through mechanisms too complex to be discussed here. Stomata1 mechanisms are discussed in Jarvis and Mansfield (1981), nyctinastic movements by Satter and Galston (1981), and the regulation of transport in general in Volume 2A of the “Encyclopedia of Plant Physiology.” The role of growth regulators, especially ABA, in the control of guard cell turgor was discussed by Mansfield and Davies and Milborrow in Paleg and Aspinall (1981) and in Jarvis and Mansfield (1981). Cell turgor also seems to be controlled by turgor pressure through its effects on cell permeability. In general, increase in turgor pressure causes a decrease in permeability to both water and ions. The importance of the effects of turgor pressure were discussed in detail in a review by Zimmermann (1978). The mechanism by which pressure controls permeability is not fully explained. It was originally suggested that pressure of the plasma membrane against the wall reduces pore size. Zimmermann suggested that as the cell membrane becomes thinner the distribution of electrical charges is changed, but this explanation has not been fully accepted for seed plants (Lucas and Alexander, 1981) and more research is needed.

C. WATERMOVEMENT IN WHOLEPLANTS The gradient in water potential that causes water movement through whole plants is produced in two ways. In very slowly transpiring plants root systems often function as osmometers because accumulation of solutes in the root xylem lowers its water potential below that of the substrate in which they are growing. Often sufficient water diffuses in to develop positive pressure (root pressure) in the xylem sap, resulting in guttation or exudation from wounds. As transpiration increases, evaporation of water produces tension in the leaves that is transmitted to the roots through the cohesive columns of sap in the xylem and lowers the water potential in the roots. The pull of the transpiration stream produces a much lower water potential than can be produced by the osmotic mechanism, and it is responsible for most of the water absorption by transpiring plants (Kramer, 1983, Chap. 8). A complete treatment of osmotic movement of water is complicated by the interaction existing between water and solute movement and requires the use of irreversible thermodynamics. Readers are referred to Slatyer (1967, pp. 162-171) and Fiscus (1975) for discussions of the mathematical treatment of this problem.

268

PAUL J. KRAMER

As transpiration increases there is gradual transition from diffusive osmotic flow to pressure or mass flow caused by tension or negative pressure in the xylem sap. A simplified equation for combined osmotic and pressure flow into roots follows: J, =

L, (A+, + aA+,)

Here J , is the volume of flow (cm3 cm-2 s - l ) , L, is the conductance or permeability in cm s - l MPa- I , A+, is the pressure potential difference between xylem and root surface in MPa, u is the reflection coefficient, and A& is the difference in osmotic potential. If A+, = 0, as in nontranspiring plant, all absorption is by osmosis, driven by A+,, but as transition from one driving force to the other occurs a plot of volume over driving force is curvilinear (Lopushinsky, 1964). This has been interpreted erroneously as indicating a change in root resistance, but the mathematical analyses of Fiscus (1975) and Dalton et al. (1975) explain the nonlinearity; also the apparent change in resistance, and apparent nonosmotic uptake of water. D. NONOSMOTICWATERMOVEMENT Thus far we have assumed that water movement in cells can be explained as occurring by the physical processes of diffusion or mass flow along gradients of water potential. However, from time to time claims are made that active transport of water occurs against water potential gradients by some mechanism dependent on metabolic energy. These claims are based on observations that the osmotic potential of the cell sap is higher (less negative) than that of the solution required to plasmolyze the cells, that the osmotic potential of the root pressure exudate is higher than that of the solution required to prevent root pressure exudation, that respiration inhibitors reduce water intake, and that auxin increases water intake. The most recent claim of active transport of water in roots was made by Russian investigators (Zholkevich et al., 1979) who attribute, it to pulsatory activity in root cells. The extensive literature was reviewed by Slatyer (1967, pp. 195-197) and by Kramer (1983, Chps. 2 and 8). These anomalous observations probably can be explained without recourse to an active transport mechanism. The discrepancies between cryoscopic and plasmolytic measurements often result from failure to take into account the leakage of plasmolyzing solutes into cells and roots because their reflection coefficient usually is less than one (Slatyer, 1966, 1967). Respiration inhibitors and oxygen deficiency probably reduce membrane permeability to water and auxin-induced increase in water uptake probably results from increased cell wall extensibility and increased permeability to water in the presence of auxin (Boyer and Wu, 1978). There also is apparent excretion or secretion of water from various glandular

WATER RELATIONS OF PLANTS AND CELLS

269

structures, discussed by Fahn (1979), Schnepf (1974), and others. In most instances it seems probable that solutes are excreted by active transport and water then diffuses out along the resulting gradient in water potential. Polar movement of water across membranes may sometimes be caused by electroosmosis, but more often results from unstirred layers and other experimental conditions (Dainty, 1963). Overall, it seems unlikely that there is any active transport of water in plants comparable to the active transport of solutes. E. WATERMOVEMENT OUTSIDE THE XYLEM As mentioned earlier, water movement from roots to shoots occurs in the xylem and presents no problems. However, the path of water movement from root surfaces to root xylem and from the xylem of leaf veins to the evaporating surfaces is less certain. Three possible pathways exist for water movement across a mass of cells; across the vacuoles, through the symplast, and through the walls. The relative amounts of water moving through the three pathways presumably is directly proportional to the relative transport capacities of the pathways and inversely proportional to their relative resistances. In 1928 Scott and Priestley pointed out that there is no reason why the soil solution cannot diffuse into roots along the cell walls, at least as far as the endodermis. Later Strugger (1949), Tanton and Crowdy (1970), and others argued from experiments with various kinds of tracers that cell walls might be an important pathway for the movement of water and solutes. This view was supported by calculations of Russell and Woolley (1961), Weatherley (1963), and Briggs (1967) who concluded that resistance to movement through cell walls is much lower than resistance to movement across the protoplasts of a mass of parenchyma cells. However, the endodermis usually presents a barrier to inward movement through the cell walls or apoplast in roots and this has resulted in consideration of the possibility that in roots water and solutes might move through the plasmodesmata that connect the cytoplasm of adjacent cells to form the symplast. This pathway eliminates the resistance to movement through several membranes existing if water moves from vacuole to vacuole, and it provides a path in the plasmodesmata through the endodermis where apoplastic movement is blocked by suberization of the radial walls. Most investigators have assumed that cell walls are much more permeable to water than the protoplasts. However, the data on wall and protoplast permeability are not very satisfactory. Newman (1976) compared data from various sources and concluded that wall permeability probably is lower than cytoplasmic permeability, as did Tyree and Yianoulis (1980). However, a year later Tyree et al. (1981) concluded that water follows both pathways. Circumstantial evidence that cell walls are more permeable to water than the cytoplasmic membranes is provided by observations that the half time for diffusion of labeled water into

270

PAUL J. KRAMER

living roots is several times greater than it is for dead roots (Kramer, 1983, pp. 137-138). Interest in the symplastic pathway has been increased by additional information concerning the high frequency of plasmodesmata in root cell walls, including those of the endodermis. The role of plasmodesmata is discussed in detail in Gunning and Robards (1976), where Robards and Clarkson present data indicating that plasmodesmata seem to provide an adequate pathway for transport of water and solutes across the endodermis of roots.

E. MATHEMATICAL MODELS Numerous attempts have been made to develop mathematical treatments or models of water movement in plants and in cells and tissues. Probably the oldest concept is the Ohm’s law analogy which developed over about three decades (Kramer 1983, Chap. 7), and is best known from papers by van den Honert (1948), Slatyer and Taylor (1960). and Cowan (1965). Philip (1966) published a critical review of this concept or model and introduced SPAC as an acronym for the soil-plant-atmosphere continuum. Richter (1973) also discussed problems in use of the SPAC model. Phillip (1958) made an early mathematical treatment of water movement at the cell level. Many others have followed, dealing with water movement through tissues (Dainty, 1963; Fiscus, 1975; Molz, 1976; Tyree et al., 1981) and with the dynamics of water movement in growing tissue (Molz and Boyer, 1978; Silk and Wagner, 1980). Molz and Ferrier (1982) reviewed various mathematical treatments of water movement and proposed a circuit analog model which they applied to movement of water through cells in series. Nevertheless, attempts to determine the most probable pathway for water movement across a mass of tissue have been unsuccessful. Models are no better than the data used in them and in this instance the data for hydraulic conductance (L,) of cytoplasmic membranes and walls are not sufficiently reliable to permit a clear choice between pathways. However, it appears probable to the writer that water moves across cells by all three pathways, the least moving across the vacuoles and the most through the cell walls, except at the endodermis where apoplastic movement usually is blocked by suberization of radial walls. G. CHANGES IN ROOTRESISTANCE One of the troublesome problems in plant water relations results from reports that root resistance tends to decrease with increasing rate of water flow. This is indicated by failure in some experiments of the leaf water potential to decrease as the rate of flow through the roots increases. Kaufman (1976) collected considerable data showing leaf water potential plotted over rate of transpiration and some of his data are shown in Fig. 5 . In some experiments the leaf water potential

WATER RELATIONS OF PLANTS AND CELLS

27 1

2 -1.5 L

s

"-

f -2.01 0

5

L

I

I

I

10

15

20

25

Transpiration Rate (Mug c r n s 2 s - ' )

Fic. 5 . Leaf water potential plotted over rate of transpiration for several species. There is no decrease in water potential with increasing transpiration for sesame or Tidestromia. nor in one experiment with sunflower, suggesting that an increase in root permeability occurred. See section on changes in root resistance for further discussion. Redrawn from Kaufmann (1976).

remained constant over a wide range of transpiration rates, suggesting that root resistance had decreased as water flow increased. There were wide variations in behavior of plants of the same species in different experiments, but there is no obvious explanation of how root resistance can be reduced by increasing the water flow. Fiscus (1975) suggested that some discrepancies are caused by the transition from osmotic to pressure flow as the rate of transpiration increases. Research by Bunce (1978) on soybeans suggested that some discrepancies occurred because the measurement time was too short for the leaf water potential to attain a new steady state after changing the rate of transpiration. Bunce also observed that low light and reduced leaf area resulted in lower leaf water potentials at a given rate of transpiration than high light and normal leaf area. Thus root resistance appears to be affected by the treatment received by the shoot. The problem of apparent changes in root resistance with changing rates of water flow deserves further study. The problem of diurnal variations in root resistance also needs further investigation (Parsons and Kramer, 1974).

V. Injury from Water Deficits At the cell level water deficits may be regarded as any water potential less than zero, or any condition less than fully turgid. However, cells are seldom fully turgid in growing plants (Molz and Boyer, 1978; Boyer and Wu, 1978; Cavalieri

272

PAUL J. KRAMER

and Boyer, 1982) and predawn water deficits of -0.2 or -0.3 MPa are common in well watered plants. Pieces of plant tissue floated on water show rapid absorption for a few hours, but aften continue to absorb water slowly for many hours because slow cell expansion continues after the major water deficit is eliminated (Barrs and Weatherley, 1962; Molz and Boyer, 1978). Water deficits affect every aspect of plant growth and development. Mild deficits cause decrease in water content, in turgor, and in water potential, resulting in decrease or cessation of cell enlargement, partial or complete closure of stomata, and wilting. Deficits also cause decrease or cessation of photosynthesis, disturbance of other metabolic processes, and finally death. The numerous effects of water deficits were discussed in detail by Hsiao (1973), by Slatyer (1967, Chap. 9), by Kramer (1983, Chap. 12), in Paleg and Aspinall (1981), and in numerous papers cited by those authors. Most of the injury caused by water deficits can be grouped in two categories, that caused by decrease in turgor and that caused by disorganization of cell structures such as membranes and organelles. The two most frequently cited effects of loss of turgor are cessation of cell enlargement and closure of stomata. More severe dehydration causes disorganization of cell membranes and disturbance of enzyme activity which affect important metabolic processes such as photosynthesis, respiration, and nitrogen metabolism. They will be discussed later. A. TURGOR-RELATED PROCESSES Among the common assumptions of plant physiology are (1) that water stress causes closure of stomata which in turn reduces photosynthesis, and (2) that cell expansion is directly dependent on turgor. Like so many other simple assumptions both of these are now being questioned and qualified. 1. Stornatal Opening Several investigators have observed simultaneous decrease in transpiration, photosynthesis, and stornatal conductance in water-stressed plants (see Fig. 6), and it generally is assumed that the reduction in photosynthesis is caused by closure of the stomata. There is no doubt that water stress causes stornatal closure, but there is increasing evidence that nonstomatal inhibition of photosynthesis can be as important or more important than stornatal inhibtion (Osmond ef al., 1980, pp. 354-364). Farquhar and Sharkey (1982) claimed that although stomatal closure in water-stressed plants reduces transpiration it only slightly limits photosynthesis because that process already is limited by the direct effect of water deficit on the photosynthetic machinery by the time stornatal closure occurs. In fact Wong et al. (1979) proposed that stornatal closure is caused by the decreased capacity to cany on photosynthesis. Thus, although the role of turgor

WATER RELATIONS OF PLANTS AND CELLS

273

e

7 30.v)

f 2 15-

2

f

CORN

’

$ 0 ‘ -0.4

L

I

-0.8

-1.2

L

I

-1.6

Leaf Wotor Potontiol (MPo)

FIG. 6 . Relationship between stomatal resistance and rates of photosynthesis and transpiration in corn. Redrawn from Boyer (1970).

in guard cell movement is not challenged the role of stomatal closure in reducing the photosynthesis of water-stressed plants is being questioned. Also, it was formerly assumed that the turgor of guard cells was controlled by bulk leaf turgor, but it is now known that the stomata of some plants close when exposed to dry air, independently of the bulk leaf turgor (Schulze et al., 1972). The complex effects on guard cell turgor of light, CO,, temperature, water, and growth regulators are discussed by Raschke (1975, 1979) and in the monograph edited by Jarvis and Mansfield (1981). 2. Cell Enlargement Another basic assumption of plant physiology is that water stress reduces cell enlargement because of reduction in turgor pressure ($p of Eq. 1). There is no doubt that some minimum degree of turgor pressure is necessary for cell enlargement (Cleland, 1971). This has been demonstrated repeatedly by many investigators and Hsiao (1973) and Turner and Begg (1978) give many examples of reduction in cell enlargement and vegetative growth caused by water deficits. However, there are reasons to question whether it is correct to assume that cell enlargement is controlled primarily by turgor. Burstrom (1971) and Cutler et a f . (1980a) pointed out that cell enlargement actually is caused by the diffusion of water into the vacuoles, and as soon as inward movement of water ceases cell enlargement ceases. Thus the driving force really is the difference in water potential between growing cells and their water source, and turgor pressure is merely a name for the pressure developed by the inward diffusion of water. Some readers may regard this as a semantic or philosophical question concerning which in a series of events is properly termed the “cause” of cell expansion. However, research on cell enlargement indicates that it is more closely related to $w than to *P.

274

PAUL J . KRAMER

This problem has been studied in detail in Boyer's laboratory and some of the results were presented by Cavalieri and Boyer (1982), who also cite much relevant literature. As shown in Fig. 7, they found the turgor pressure, (9,)was similar in growing and nongrowing regions of soybean hypocotyl, but the osmotic (9,)and water potentials (+,) were much lower in the growing region, in both water-stresed and unstressed hypocotyls. This supports the conclusion that gradients in water potential are the driving force for cell enlargement. The situation shown in Fig. 7 develops because growing regions are sinks for metabolites and the accumulation of solutes (osmotic adjustment) in growing cells lowers the osmotic potential, permitting the development of a low water potential which causes influx of water and cell enlargement. A continuous supply of solutes is necessary to maintain the gradient in in enlarging cells. In soybean hypocotyls this comes from the food stored in the cotyledons and in older plants from photosynthesis. The basic question in connection with cell enlargement is what makes growing regions strong sinks for metabolites. If we understood what controls partitioning of food first to growing regions and later to fruits and seeds we could modify the growth pattern of plants and increase economic yields. The size of the gradient in water potential depends on the hydraulic conduc-

0.004'

f n

-0.4 -05 -06 I I I 1 I 1 1 1 I I 1 1 1 0 1 2 3 4 5 6 7 8 9 10 11 12 I

Initial Distance From Cotyledons

(cm)

FIG.7. Water potential (qw), osmotic potential (*J, and pressure potential (qp) or turgor. and growth of germinating soybean hypocotyls 48 hours after transplanting to vermiculite with a qwof -0.01 MPa. Redrawn from Cavalieri and Boyer (1982).

WATER RELATIONS OF PLANTS AND CELLS

275

tance of the tissue (L,) and the extensibility of the cell walls (E), both of which are affected by the supply of auxin. According to Boyer and Wu (1978), cell enlargement in soybean hypocotyls is limited about equally by L, and cell wall extensibility. Differences in water potential of up to 0.5 MPa were observed between growing cells and substrate (see Fig. 7), indicating a low L,, for movement of water into growing tissue. The relationship between the pressure causing cell expansion and actual cell expansion is complex and varies with the kind, age, and past history of the tissue. Growth ceased at water potentials of -0.5, -0.75, -0.9, and -1.4 MPa, respectively, in elongating regions of stems, silks, leaves, and roots of corn (Westgate and Boyer, 1982). Growth apparently was maintained to different water potentials in various tissues by differences in solute accumulation, i.e., differences in osmotic adjustment. The investigators suggested that ability to grow at low water potentials depends at least in part on ability to maintain a water potential gradient between the growing tissue and the water source. Sharp and Davies ( 1979) also reported that root growth sometimes continues after shoot growth ceases because of osmotic adjustment in the roots. Musser et al. (1982) found that chilling to 10°C greatly reduced leaf growth of soybean although the turgor was much higher in chilled than in unchilled leaves. Thus high turgor does not guarantee cell enlargement. It is not surprising to find that there is not always a close correlation between cell turgor and expansion because cell expansion depends on two kinds of processes, physical and biochemical. The pressure necessary for expansion depends on the physical processes controlling inward diffusion of water, but plastic extension of cell walls also requires the biochemical processes involved in relaxation of bonding in the walls and synthesis and intussusception of new wall material (Roland and Vian, 1979). Inhibition of the biochemical components of the process by low temperature or water stress can therefore reduce or prevent cell enlargement just as effectively as reduction in water absorption. In fact Wenkert et al. (1978) concluded that under average field conditions, although midday loss of turgor may temporarily reduce growth, mean growth over long periods of time is limited primarily by metabolic processes rather than by intermittent lack of turgor pressure. 3 . Movements Caused by Growth and Turgor

Various kinds of movements in plants, either in response to stimuli such as touch, temperature, light, and gravity, or sometimes apparently independent of external stimuli, have been observed for centuries. They were studied extensively during the nineteenth century and Charles Darwin wrote about plant movements. They are mentioned here because they present interesting problems in cell water relations. Some movements involve differential growth, others result from rapid changes in turgor that require large changes in cell volume and rapid uptake

276

PAUL J. KRAMER

or loss of large amounts of water and solutes. If the volume changes are irreversible the response is regarded as growth, if reversible it is regarded as a turgor process, but sometimes it is difficult to distinguish between the two. Some phenomena, such as the opening and closing of tulip flowers, the nutational movements of elongating stem tips, movement of tendrils, and tropic movements are caused by unequal rates of growth on opposite sides of plant structures; others, such as the movements of stomata1 guard cells, the drooping of leaves and leaflets of Mimosa pudica when touched, the changes in leaf position between day and night (nyctinasty), and the rolling of grass leaves during wilting, clearly result from localized changes in cell turgor. In some instances the changes in turgor are restricted to specialized cells such as those of the pulvini at the bases of the leaflets and leaf petioles in Mimosa pudica and some other legumes, and the large thin walled cells along the main veins of grass leaves. The changes in turgor responsible for movements of guard cells and Mimosa leaflets are associated with changes in concentration of K + and other ions and energy provided by ATP. However, it is difficult to explain how the cells of pulvini can lose sufficient solutes and water to permit closure in seconds. It is even more diffult to explain how signals produced by a stimulus applied to the tip of a Mimosa leaflet is transmitted to the base of the petiole in seconds and causes changes in turgor to occur. Furthermore, nyctinastic movements often exhibit circadian rhythms and some continue for several days in plants kept in continuous darkness. The closure of the leaves of Venus flytrap (Dionea muscipula) is a particularly interesting example of movement in plants. When the sensitive hairs on the inner surface are touched the two halves of the leaf fold together within a second or two, and they reopen in a few hours or days. Closure generally has been attributed to increase in turgor of cells on the lower surface, reopening to growth of cells on the upper or inner surface. However, Williams and Bennett (1982) claim that closure results from rapid, irreversible cell enlargement on the lower surface of the leaves, made possible by release of H+ that cause acidification and increased plasticity of the cell walls. If true, this must represent one of the most rapid examples of growth on record. A full discussion of these phenomena lies outside the scope of this article, but they are mentioned because they are examples of complex control of cell water relations. Readers are referred to Volume 7 of the “Encyclopedia of Plant Physiology” and review articles such as those by Firn and Digby (1981) and Satter and Galston (1981) for detailed discussions of some of them. B. METABOLIC PROCESSES

The effects of water deficits on enzyme-mediated metabolic processes are very important. It often is stated that water stress causes an increase in activity of degradative enzymes, as evidenced by increases in soluble carbohydrates and

WATER RELATIONS OF PLANTS AND CELLS

277

amino acids (Todd, 1972). However, it is difficult to determine how much of this increase is caused by increased breakdown of polysaccharides and proteins and how much is caused by decreased utilization of simpler, soluble compounds in growth (Hsiao, 1973). Water stress can have multiple effects on physiological processes. For example, photosynthesis at the whole plant level is decreased by closure of stomata and reduction in leaf area, both of which are related to cell water status, but water deficit also reduces photosynthesis through its effects on enzymatic steps, electron transport, and chlorophyll content. Nonstomatal effects of water stress on photosynthesis are summarized by Osmond et al. (1980, pp. 354-364). Among the enzyme systems studied in relation to water deficits are those involved in phytosynthesis, respiration, chlorophyll synthesis, synthesis of growth regulators such as abscisic acid (ABA) and cytokinins, and various processes included in nitrogen metabolism. A much studied example is the accunulation of proline in water-stressed plants, which results from increased synthesis as well as because of decreased oxidation and decreased utilization in the synthesis of proteins. Synthesis of ABA is increased and that of cytokinin is decreased by water deficit. Thus the activity of some enzymes is decreased while that of others is increased by water deficits, and many of the changes in enzyme activity occur with relatively low water stresses (Hsiao, 1973). Although additional information concerning effects of water stress on enzyme activity continues to become available we do not yet fully understand how moderate stresses of 1 .O MPa or less produce such marked effects. Hsiao suggested that loss of water ( 1 ) reduces the chemical potential ($w), (2) decreases turgor pressure $p, (3) increases the concentration of solutes in cells ($s), (4) dehydrates macromolecules, and ( 5 ) changes spatial relationships in cell and organelle membranes as cell volume is reduced. After a detailed consideration of the possibilities he decided that changes in solute concentration and in spatial relationship in membranes are probably the most important ways in which water deficits affect metabolic processes. Dhindsa and Cleland (1975 ) also considered several factors that might cause the inhibition of protein synthesis observed in water-stressed tissue and concluded that decrease in membrane surface accompanying decrease in cell volume, is the most likely cause. The problem deserves more study. For more details readers are referred to Hsiao (1973), Kramer (1983, Chapters 12 and 13), Hanson and Hitz (1982), Paleg and Aspinall (1981), and the current literature.

VI. Adaptations Increasing Tolerance of Water Deficits There are wide differences in the ability of plants to tolerate droughts. Some of these differences result from adaptations at the whole plant level, such as deep, wide spreading root systems; some are at the organ level, such as heavily cu-

278

PAUL J. KRAMER

tinized leaves and good stomatal control of transpiration, while others such as osmotic adjustment and tolerance of dehydration operate at the cellular level. We are concerned chiefly with adaptations at the level of the cell that are effective either in postponing dehydration or increasing the tolerance of dehydration. Thus many other important adaptations are omitted from this paper. Some of them are discussed by Kramer (1983, Chap. 13), in Turner and Kramer (1980), and in Raper and Kramer ( 1983). A. POSTPONEMENT OF INJURY BY DEHYDRATION

Several kinds of adaptations occur that postpone dehydration. Most of these are morphological, such as extensive root systems, and the only ones closely related to the theme of this article are responsive stomata and osmotic adjustment. 1. Responsive Stomata Heavily cutinized leaves combined with responsive stomata that close rapidly when water stress begins to develop tend to reduce transpiration and postpone dehydration. Differences in stomatal control of transpiration seem to be related to differences in drought tolerance of mesophytes such as sorghum (Teare et a f . , 1973), monterey pine (Bennet and Rook, 1978), and poplar clones (Ceulemans et al., 1978). The stomata of some plants close in dry air, but the stomata of some desert plants do not respond to low humidity and the adaptive value of this characteristic is uncertain (Osmond et a f . , 1980, pp. 277-280). Cowan and Farquhar (1977) stated that stomata tend to adjust in a manner that keeps the internal CO, concentration and the ratio of transpiration to photosynthesis fairly constant over a wide range of stomatal openings. Such optimization requires complex control of the osmotic machinery of guard cells. It is generally assumed that prompt response of guard cells to water stress has the disadvantage of cutting off the supply of CO, and inhibiting photosynthesis. However, as mentioned earlier, some investigators claim that water stress inhibits photosynthesis first and this in turn causes stomatal closure (Farquhar and Sharkey, 1982). Ludlow (1980) discussed the ecological and physiological importance of the reaction of stomata to leaf water deficits and low humidity and stated that more research is needed. 2. Osmotic Adjustment This refers to a decrease in osmotic potential greater than can be explained by concentration of solutes caused by dehydration. For example, if the initial JI, is - 1.O MPa and the volume of vacuolar sap is reduced 10% by loss of water the new JI, will be - 1.1 MPa. However, accumulation of additional solutes might lower the JI, to - 1.3 or - 1.4. This would enable water uptake to continue,

WATER RELATIONS OF PLANTS AND CELLS

279

permitting cell enlargement to continue, stomates to remain open, and the photosynthetic apparatus to operate at a lower plant water potential than if it did not occur. Osmotic adjustment in roots is said to permit continued growth and water absorption after shoot growth has ceased (Sharp and Davies, 1979). The general importance of osmotic adjustment in growing tissues was mentioned in the section on cell enlargement. Osmotic adjustment is at least partly responsible for frequent observations that exposure to water stress decreases sensitivity to subsequent water stress, both with respect to leaf enlargement (Cutler et af., 1980b) and stomata1 closure. The importance of osmotic adjustment for plants growing in saline substrates was discussed in the monograph edited by Rains et al. (1980) and the general problem of osmotic adjustment is reviewed by Turner and Jones (1980) and by Jones et al. in Paleg and Aspinall (1981, pp. 25-30). B. METABOLIC ADAPTATIONS

In the section on injury from water stress it was pointed out that water deficits have many important effects on the physiology and biochemistry of plants. Some of these may either defer dehydration or increase the tolerance of it. For example, increase in abscisic acid may cause closure of stomata which would postpone dehydration, although its role is not entirely clear (Davies et af. in Jarvis and Mansfield, 1981; Milborrow, in Paleg and Aspinall, 1981). It is well established that proline accumulates in many kinds of water-stressed plants and betaine accumulates in plants of a few taxa, and attempts have been made to relate their accumulation to drought tolerance. However, this is difficult because their concentrations are too low to significantly reduce the +s of the vacuolar sap, although they might have some osmotic effect if they accumulate principally in the cytoplasm. It is well established that the accumulation of both betaine and proline occurs because of disturbance of nitrogen metabolism by water stress. For example, cell water deficits increase proline synthesis, inhibit its oxidation, and decrease its use in protein synthesis. Water stress also increases the synthesis of betaine and decreases its translocation and use (Hanson and Hitz, 1982). Thus any beneficial results of accumulation of betaine and proline probably can be regarded as fortuitous. Perhaps plants that accumlate these compounds can be regarded as less rather than more tolerant of dehydration. The possible roles of various compounds in drought tolerance are discussed by several writers in Paleg and Aspinall (1981) and by Stewart and Hanson (1980). Crassulacean acid metabolism (CAM) clearly results in postponement of dehydration. CAM refers to plants that have daytime closure of stomata and fix CO, into malic acid in darkness. This is decarboxylated in the light, releasing CO, that is refixed into carbohydrates by photosynthesis. Daytime closure of stomata combined with dark fixation of CO, greatly reduces water loss and postpones dehydration with a minimum decrease in production of dry matter.

280

PAUL J. KRAMER

The most important cultivated plant with CAM is pineapple. The C, carbon pathway in photosynthesis also tends to conserve water, but as Osmond er al. (1980, p. 355) point out, the C, carbon pathway confers no special tolerance after dehydration occurs, and there is a wide range of drought tolerance in both C, and C, plants. C. WATERSTORAGE

In desert succulents such as cacti and certain Euphorbias enough water is stored in the parenchyma cells of the stems to permit survival for months or even years. An example is the saguaro or giant cactus (Carnegia gigantea) which stores tons of water in its stems. The baobab tree (Adansonia digirara) of Africa also stores water in its enormous trunk. An appreciable amount of water is stored in the trunks of forest trees and Waring and Running (1978) suggested that this postpones development of water deficits. It might enable photosynthesis to continue longer each day if the deficit is replaced at night, but the benefit probably is minimal. Although some water is stored in crop plants it has no long-term effect on postponement of dehydration because daily transpiration often exceeds the total water content of herbaceous plants. D. TOLERANCE OF DEHYDRATION There are wide differences among species with respect to tolerance of dehydration, ranging from about 2.0 MPa in sunflower and corn to below - 10.0 MPa in various xerophytes, and to the air dry condition in some ferns and seed plants (Gaff, 1980). The reasons for these differences are not fully understood. Perhaps most puzzling is the fact that many crop plants suffer irreversible injury at about -2.0 to -3.0 MPa, but their seeds can be dehydrated to at least an air dry condition without injury (Adams and Rinne, 1980). The changes in fine structure occurring during the dehydration that accompanies seed maturation are similar to those occurring during dehydration of vegetative tissue (PoljakoffMayber in Paleg and Aspinall, 1981). However, they are not destructive in seeds, and dehydration even seems to be an essential prelude to germination of some seeds (Klein and Pollock, 1968). The reasons for the differences in tolerance of dehydration are not fully understood. Injury has been ascribed to mechanical disorganization of protoplasm, disorganization of cell membranes, protein denaturation, and accelerated gene mutations. In general terms injury probably can be attributed to physical changes in organelle and membrane structure. However, studies by electron microscopy show that the degree of disorganization varies widely. The very tolerant moss, Torrula ruralis, appears to retain its fine structure and capacity for physiological and biochemical activity because it resumes normal respiration and protein syn-

WATER RELATIONS OF PLANTS AND CELLS

28 1

thesis in about 30 minutes after rehydration, and photosynthesis after 2 hours. In some resurrection plants there is such severe disorganization of fine structure during dehydration that 1 or 2 days after rehydration are required for resumption of physiological activity. This problem is discussed in detail by Bewley (1979) and Gaff ( 1980). 1. Cell Size

The late W. S. Iljin observed that plants living in dry habitats and those subjected to water deficits have smaller cells than those living in moist habitats. He argued that such plants are more tolerant of dehydration because small cells shrink less than large cells when dehydrated and therefore suffer less injury (Iljin, 1957). More recently, Cutler et al. (1977) concluded that small cells than large cells. They suggested that should maintain turgor to a lower q,,, prestressing or hardening plants increases drought tolerance because it decreases cell size, and small cells should be able to maintain turgor to a lower water potential than larger cells. 2. Increasing Tolerance There is growing interest in the possibility of increasing tolerance of drought and dehydration by plant breeding. Much has been accomplished with conventional methods by hybridization and selection for deeper root systems, better control of transpiration, and earlier maturity (see papers in Raper and Kramer, 1983). More attention is now being given to selection for specific physiological and biochemical characteristics that should increase water use efficiency or tolerance of water stress. Examples are the work on tolerance of high salinity described in Rains er al. (1980), research on the role of betaine and proline cited by Hanson and Hitz (1982), and selection of cell lines tolerant of chilling cited by McWilliam (1983). There also is increasing interest in the use of cell cultures and even isolated protoplasts in screening for stress tolerance. However, Galun (1982) warned that protoplasts are not merely cells without walls, and Bressan er al. (1982) reported that tomato cells in culture become adapted to an osmotic stress fatal to mature tomato plants. In the long run, it seems reasonable to expect that cell cultures and recombinant DNA techniques will aid in selecting and producing plants with significantly increased tolerance of various kinds of stress. However, many difficulties remain to be overcome (Howell, 1982; Kado and Kleinhofs, 1980).

VII. Summary In conclusion, although a great deal is known about cell structure and cell water relations, the answers to some important questions remain in doubt. The

282

PAUL J. KRAMER

data on the hydraulic conductance of the apoplast and the symplast are icadequate to establish which is the principal pathway for water movement across masses of cells. We cannot distinguish clearly between water held by matric and by osmotic forces, and it is uncertain whether negative turgor pressure ever occurs in cells. The long held view that turgor pressure is the cause of cell enlargement is being questioned as evidence accumulatesthat cell enlargement in growing tissue is more closely related to water potential than to turgor pressure. Also, the assumption that stornatal closure is caused primarily by loss of bulk leaf turgor is being questioned. Although it is well established that many enzyme-mediated processes are perturbed by water deficits, it is not known exactly how this is brought about. Neither do we know why the cells of some kinds of plants and plant organs are irreversibly injured by relatively moderate dehydration whereas others survive air-drying or even more severe dehydration. Thus many important problems in the field of cell water relations remain in need of further investigation.

ACKNOWLEDGEMENTS The writer wishes to acknowledgethe helpful comments of John S. Boyer, Albert H.Markhart Ill, and James N. Siedow who read an early version of this article. The financial support of the National Science Foundation for 25 years of research on plant water relations also is gratefully acknowledged.

REFERENCES Ackley, W. B. (1954). Plant Physiol. 29, 445-448. Acock, B. (1975). Ausr. J . Plant Physiol. 2 , 253-264. Acock, B., and Grange, R. I. (1981). I n “Mathematics and Plant Physiology” (D. A. Rose and D. A. Charles-Edwards, eds.), pp. 29-47. Academic Press, New York. Adams, C. A., and Rinne, R. W. (1980). Inr. Rev. Cytol. 68, 1-8. Barrs, H. D., and Weatherley, P. E. (1962). Ausr. J . Eiol. Sci. 15, 413-428. Bennett, K. J., and Rook, D. A. (1978). Ausr. J. Plant Physiol. 5 , 231-238. Berlyn, G. P. (1969). Am. J . Bot. 56, 498-506. Bewley, J. D. (1979). Ann. Rev. Plant Physiol. 30, 195-238. Bonner, J., and Vamer, 1. E., eds. (1976). “Plant Biochemistry.” Academic Press, New York. Boyer, J. S.(1967). PIanr Physiol. 42, 213-217. Boyer, J. S. (1970). PIanr Physiol. 46, 236-239. Boyer, J. S . (1982). Science 218, 443-448. Boyer, J. S., and Potter, J. R. (1973). Plant Physiol. 51, 989-992. Boyer, J. S., and Wu, G. (1978). Planra 139, 227-237. Bressan, R. A., Handa, A. K., Handa, S., and Hasegawa, P. M. (1982). Plant Physiol. 70, 1303- 1309. Briggs, G. E. (1967). “Movement of Water in Plants.” Davis, Philadelphia, Pennsylvania. Bunce, J. A. (1978). J. Exp. Eor. 29, 595-601.

WATER RELATlONS OF PLANTS AND CELLS

283

Burstrom, H. G. (1971). Nature (London) 234, 488. Campbell, G. S., Papendick. R. I . , Rabie, E., and Shayo-Ngowi, A. J . (1979). Agron. J. 71,31-36. Carpita, N. C . (1982). Science 218, 813-814. Cavalieri, A. J . , and Boyer. J. S. (1982). Plant Physiol. 69, 492-496. Ceulemans, R . , Impens, I., Lemeur, R.. Moermans, R., and Samsuddin, Z. (1978). Oecol. Plant. 13, 139-146. Christiansen, M. N., ed. (1982). “Breeding PLants for Less Favorable Environments.” Wiley, New York. Cleland, R. (1971). Annu. Rev. Plant Physiol. 22, 197-223. Cowan, 1. R. (1965). J. Appl. Ecol. 2, 221-239. Cowan, 1. R., and Farquhar, C. G. (1977). Symp. Soc. Exp. Biol. 31, 471-505. Crafts, A. S . , Currier, H. B., and Stocking, C.R. (1949). “Water in the Physiology of the Plant.’’ Chronica Botanica, Waltham, Massachusetts. Cram, W. J. (1976). Encycl. Plant Phvsiol. New Ser. 2A, 284-316. Cutler, J . M., Rains, D. W., and Loomis, R. S. (1977). Physiol. Plant. 40,255-260. Cutler, J . M., Steponkus, P. L.. Wach, M. J., and Shahan, K. W. (1980a). Plant Physiol. 66, 147- 152. Cutler, J. M., Shahan, K. W., and Steponkus, P. L. (1980b). Crop Sci. 20, 307-310. Dainty, J. (1963). Protoplasma 57, 220-228. Dainty, J . (1976). Encycl. Plant Phvsiol. New Ser. 2A, 12-35. Dalton, F. N., Raats, P. A. C., and Gardner, W. R. (1975). Agron. J. 67, 334-339. Dennis, D. T., and Miernyk. J . A. (1982). Annu. Rev. Plant Physiol. 33, 27-50. Dhindsa, R . S . , and Cleland, R. E. (1975). PLanr Phvsiol. 55, 782-785. Esau, K. (1965). “Plant Anatomy.” Wiley, New York. Fahn, A. (1979). “Secretory Tissues in Plants.” Academic Press, New York. Farquhar, G. D.. and Sharkey, T. D. (1982). Annu. Rev. Plant Physiol. 33, 317-345. Firn, R. D., and Digby, J . (1980). Annu. Rev. Plant Physiol. 31, 131-148. Fiscus, E. L. (1975). Plan: Physiol. 55, 917-922. Frey-Wyssling, A., and Miihlethaler, K. (1965). “Ultrastmctural Plant Cytology.” American Elsevier, New York. Gaff, P. F. (1980). “In “Adaptation of Plants to Water and High Temperature Stress” (N. C . Turner and P. J. Kramer, eds.), pp. 207-230. Wiley. New York. Gaff, P. F., and Carr, D. J . (1961). Aust. J . Biol. Sci. 14, 299-31 1. Galun, E. (1981). Annu. Rev. Plant Phvsiol. 32, 237-266. Green, P. B., and Stanton, R. W. (1967). Science 155, 1675-1676. Gunning, B. E. S., and Robards, A. W . , eds. (1976). “lntercellular Communications in Plants: Studies on Plasmodesmata.” Springer-Verlag. Berlin and New York. Hanson, A. D., and Hitz, W. D. (1982). Annu. Rev. Plant Ph.vsio1. 33, 163-203. Hellebust, J . A. (1976). Annu. Rev. Plant Physiol. 27, 485-505. Howell, S . H. (1982). Annu. Rev. Plant Phvsiol. 33, 609-650. Hsiao. T. C. (1973). Annu. Rev. Plant Physiol. 24, 519-570. Iljin, W. S . (1957). Annu. Rev. Plant Physiol. 8, 257-274. Jarvis, P. G., and Mansfield, T. A.. eds. (1981). “Stomata1 Physiology.” Cambridge Univ. Press, London and New York. Kado, C. I., and Kleinhofs, A. (1980). Int. Rev. Cytol. Suppl. l I B , 47-80. Kaiser, G., Martiaia, E., and Wiemken. A. (1982). Z. Pflanzenphysiol. 107, 103-113. Kaufmann, M . R. (1976). In “Transport and Transfer Processes in Plants” (I. Wardlaw and J. Passioura, eds.), pp. 313-317. Academic Press, New York. Kiyosawa. K., and Tazawa, M. (1977). J . Membr. Biol. 37, 157-166. Klein, S . , and Pollock, 8 . M. (1968). Am. J. Bot. 55, 658-672.

284

PAUL J. KRAMER

Kramer, P. J. (1955). Encycl. Plant Physiol. 1, 223-242. Kramer, P. J. (1983). “Water Relations of Plants.” Academic Press, New York. Kramer, P. J., Knipling, E. B., and Miller, L. N. (1966). Science 153, 889-890. Leopold, A. C., Musgrave, M. E., and Williams, K. M. (1981). Plant Physiol. 68, 1222-1225. Lopushinsky, W. (1964). Plant Physiol. 39, 494-501. Lucas. W. J., and Alexander, W. 1. (1981). Plant Physiol. 68, 553-559. Ludlow, M. M. (1980). In “Adaptation of Plants to Water and High Temperature Stress” IN. C. Turner and P. J. Kramer, eds.), pp. 123-138. Wiley, New York. Lyons, J. M., Graham, D., and Raison, J. K. (1979). “Low Temperature Stress in Crop Plants: The Role of the Membrane.” Academic Press, New York. McWilliam, J. R. (1983). In “Crop Reactions to Water and Temperature Stress in Humid, Temperate Climates” (C. D. Raper, Jr. and P. J. Kramer, eds.), pp. 113-147. Westview Press, Boulder, Colorado. Markhart, A. H., 111, Peet, M. M., Sionit. N., and Kramer, P. J. (1980). Plant Cell Environ. 3, 435-441. Markhart, A. H., 111, Sionit, N., and Siedow, J. N. (1981). Can. J. Bor. 59, 1722-1725. Matile, P. (1978). Annu. Rev. Plant Physiol. 29, 193-213. Molz, F. J. (1976). J. Theor. Biol. 59, 277-292. Molz, F. J . , and Boyer, J. S. (1978). Plant Physiol. 62, 423-429. Molz, F. J . , and Ferrier, 1. M. (1982). Plant, Cell Environ. 5 , 191-206. Munch, E. (1930). “Die Stoffbewegung in der Ptlanze.” Fischer, Jena. Musser, R. L., Kramer, P. J., Naylor, A. W., and Thomas, S. A. (1982). Plant Physiol. 69(suppl. No. 4). 121. Newman, E. I. (1976). Philos. Trans. R. SOC.London Ser. B 273, 463-478. Osmond, C. B., 0. Bjorkman, and Anderson, D. J. (1980). “Physiological Processes in Plant Ecology.” Springer-Verlag, Berlin and New York. Paleg, L. G., and Aspinall, D., eds. (1981). “The Physiology and Biochemistry of Drought Resistance in Plants.” Academic Press, New York. Parsons, L. R., and Kramer, P. I. (1974). Physiol. Plant. 30, 19-23. Passioura, J. B. (1980). J. Exp. Eor. 31, 1161-1169. Pfeffer, W. (1897). “Ptlanzenphysiologie.” Engelmann, Leipzig. Philip, J. R. (1958). Plant Physiol. 33, 264-271. Philip, J. R. (1966). Annu. Rev. Plant Physiol. 17, 245-268. Radin. J. W., and Boyer, J. S. (1982). Plant Physiol. 69, 771-775. Rains, D. W., Valentine, R. C.. and Hollaender, A., eds. (1980). “Genetic Engineering of Osmoregulation.” Plenum, New York. Raper, C. D., Jr., and Kramer, P. J., eds. (1983). “Crop Reactions to Water and Temperature Stresses in Humid, Temperate Climates.” Westview Press, Boulder, Colorado. Raschke, K. (1975). Annu. Rev. Plan? Physiol. 26, 309-340. Raschke, K. (1979). Encyl. Plant Physiol. New Ser. 7 , 383-441. Richter, H. (1973). J. Exp. Eor. 24, 983-994. Rivera, C. M., and Penner, D. (1978). Weed Sci. 26, 647-650. Roland, J. C., and Vian, B. (1979). In?. Rev. Cytol. 61, 129-166. Russell, M. B., and Woolley. J. T. (1961). In “Growth in Living Systems” (M.X. Zamw, ed.). pp. 695-721. Basic Books, New York. Sanchez-Diaz, M. F., and Kramer, P. J. (1971). Plant Physiol. 48, 613-616. Satter, R. L., and Galston, A. W. (1981). Annu. Rev. Planr Physiol. 32, 83-110. Schnepf, E. (1974). I n “Dynamic Aspects of Plant Ultrastructure” (A. W. Robards, ed.1, pp. 331-357. McGraw-Hill, New York. Schulze, E.-D., Lange, 0. L., Buschbom, U., Kappen, L., and Evenari, M. (1972). PIantu 108, 259-270.

WATER RELATIONS OF PLANTS AND CELLS

285

Scott, L. I.. and Priestley, J. H. (1928). New Phyrol. 27, 125-141. Sharp, R. E., and Davies. W. J . (1979). PIunru 147, 43-49. Silk, W. K.. and Wagner, K. K. (1980). Plant Physiol. 66, 859-863. Sionit, N., Teare, 1. D., and Kramer, P. J. (1980). Physiol. Plunr. 50, 11-15. Slatyer, R. 0. (1957). Ausr. J. Biol. Sci. 10, 320-336. Slatyer, R. 0. (1966). Protoplusmu 62, 34-43. Slatyer, R. 0. (1967). “Plant Water Relationships.” Academic Press, New York. Slatyer, R. O., and Taylor, S. A. (1960). Nurure (London) 187, 922. Slavik, B. (1974). “Methods of Studying Plant Water Relations.” Springer-Verlag, Berlin and New York. Stewart, C. R.,and Hanson. A. D. (1980). In “Adaptation of Plants to Water and High Temperature Stress” (N. C. Turner and P. J. Kramer, eds.), pp. 173-189. Wiley, New York. Stout, D. G., Steponkus, P. L., and Colts, R. M. (1978). PIunr Physiol. 62, 636-641. Strugger, S.(1949). “Praktikum der Zell- und Gewebe-physiologieder Ptlanzen.” Springer-Verlag. Berlin and New York. Tang], E. (1879). Juhrb. Wiss. Bor. 12, 170-190. Tanton, T.W., and Crowdy, S. H. (1970). J. Exp. Bor. 21, 102-1 11. Teare, I. D., Kanemasu, E. T., Powers, W. L., and Jacobs, H. S. (1973). Agron. J . 65, 207-21 I. Teoh, T. S., Aylmore, L. A. G., and Quirk, J . P. (1976). Ausr. J. Biol. Sci. 20, 41-50. Tepfer, M., and Taylor, 1. E. P. (1981). Science 213, 761-763. Todd, G. W. (1972). In “Water Deficits and Plant Growth’’ (T. T. Kozlowski, ed.), Vol. 3, pp. 177-216. Academic Press, New York. Torii, K., and Laties, G. G. (1966). Plant Physiol. 41, 863-870. Turner, N. C. (1981). Plunr Soil 58, 339-366. Turner. N. C., and Begg, J. E. (1978). In “Plant Relations in Pastures” (J. R. Wilson, ed.), pp. 50-66. CSIRO, Melbourne. Turner, N. C., and Jones, M. M. (1980). In “Adaptation of Plents to Water and High Temperature Stress” (N. C. Turner and P. J . Kramer, eds.), pp. 85-103. Wiley, New York. Turner, N. C., and Kramer, P. J . (1980). “Adaptation of Plants to Water and High Temperature Stress.” Wiley, New York. Tyree, M. T. (1970). J. Theor. Biol. 26, 181-214. Tyree, M. T. (1976). Can. J. Bor. 54, 2738-2746. Tyree, M. T., and Hammel. H. T. (1972). J . Exp. Bor. 23, 267-282. Tyree, M. T., and Jarvis, P. G. (1982). Encycl. Plant Physiol. New Ser. 12B, 35-77. Tyree, M. T., and Karamanos, A. J. (1980). In “Plants and Their Atmospheric Environment” (J. Grace, E. D. Ford, and P. G. Jarvis, eds.), pp. 237-261. Blackwell, Oxford. Tyree, M. T., and Yianoulis. P. (1980). Ann. Bor. 46, 175-193. Tyree. M. T.. Cruiziat, P., Benis, M . , Gullo, M. A. L., and Salleo, S. (1981). Plunr, Cell Environ. 4, 309-317. van den Honert, T. H. (1948). Faruduy SOC.Discuss. (3). 146-153. Wagner, G. J. (1982).In “Cellular and Subcellular Localization in Plant Metabolism” (L. L. Creasy and G. Hrazdina, eds.), pp. 1-45. Plenum, New York. Waring, R. H., and Running, S. W. (1978). Plant, Cell Environ. 1, 131-140. Weatherley, P. E. (1950). New Phvrol. 49, 81-97. Weatherley, P. E. (1963). In “The Water Relations of Plants” (A. J . Rutter and F. H. Whitehead, eds.), pp. 85-100. Wiley, New York. Wenkert, W. (1980). Plan/ Physiol. 65, 614-617. Wenkert, W.,Lemon, E. R. and Sinclair, T. R. (1978). Ann. Bor. 42, 295-307. Westgate. M. E., and Boyer, J. S. (1982). Plunr Physiol. 69 (Suppl. to No. 4). 14. Wiebe, H. H. (1978). BioScience 28, 327-331. Williams, S. E., and Bennett, A. 9 . (1982). Science 218, 1120-1122.

286

PAUL J. KRAMER

Wilson, J. W. (1967). J . Biol. Sci. 20, 329-347. Wong, S. C., Cowan, I. R., and Farquhar, G . (1979). Narure (London) 282, 424. Zholkevich, V. N. Sinitsyna, Z. A., Peisakhzon, B . I., Abutolybov, V. F., and D’yachenko, 1. V. (1979). Sov. Plant Physiol. 26, 790-802. Zimmermann, U. (1978). Annu. Rev. Plant Physiol. 29, 121-148. Zimmermann, U . , and Steudle, E. (1980). In “Plant Membrane Transport” (R. M. Spanswick, W. J. Lucas, and J. Dainty, eds.), pp. 113-127. Elsevier, Amsterdam.

INTERNATIONAL REVIEW OF CYTOLOGY. VOL. 85

Phagocyte-Pathogenic Microbe Interactions ANTOINETTE RYTERAND CHANTAL DE CHASTELLIER Uniti de Microscopie Electronique, Dipartement de Biologie Moliculaire, Institut Pasteur, Paris Cedex, France I. Introduction .....................

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

288

111. Ingestion. .............................

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

VII. Conclusions ..................................... References . . . . . . . ...........................

296

319

I. Introduction Endocytosis is a widespread cellular function which allows the cell to ingest exogenous material. Pinocytosis, which occurs in most eukaryotic cells, corresponds to the uptake of very small particles such as ferritin, macromolecules, and low-molecular-weight solutes. Although it is difficult to visualize the initial step of pinocytosis, it probably arises by different mechanisms: fusion of membrane folds, or membrane invagination leading to the formation of depressions; in large amoebae the plasma membrane forms long channels (Chapman-Andresen, 1977; Stockem, 1970, 1977). Phagocytosis corresponds to the uptake of large particles such as microorganisms, latex beads, oil droplets, etc. Soon after particle adhesion, the cell membrane extends pseudopods around the particle, thus forming a sort of large depression termed the phagocytic cup. The latter gives rise to a phagosome upon membrane closure. As already described in excellent reviews (Holter, 1959; Jacques, 1969; North, 1970; Silverstein et al., 1977), pinosomes and phagosomes fuse with lysosomes and constitute secondary lysosomes in which the internalized material is digested by acid hydrolases. Phagocytosis is particularly well developed in Protozoa for which it represents the main feeding mechanism. The study of primitive pluricellular organisms in which digestive functions are ensured by a group of ameboid cells led Metchnikoff, 100 years ago, to parallel the behavior of these ameboid cells with that of 287 Copyright 0 1983 by Academic Press, Inc. All rights of reproduction in any form reserved. ISBN 0-12-364485-2

288

ANTOINETTE RYTER AND CHANTAL DE CHASTELLIER

cells found in inflammatory lesions of vertebrates. He discovered that specialized cells endowed with a high phagocytic activity play a crucial role in the host defense against microbe invasion. Although the concept of cellular immunity has been ackowledged for over a century, there has been a real explosion of inforrnation during the last decade on the origin, physiology, and properties of the phagocytes termed “professional phagocytes. The study of microbe phagocytosis showed that in most cases, microbes are killed and degraded either by blood phagocytes [polymorphonuclear (PMN) leukocytes], or by monocytes or macrophages. Some microbes (bacteria, fungi, protozoa), however, can either avoid phagocytosis and so invade the tissues, or survive and multiply inside macrophages. The ability of intracellular parasites to escape from the microbicidal activity of professional phagocytes confers to them pathogenic properties which in many cases create severe problems of public health. The study of pathogen-host cell interplay has given rise to a huge and complex amount of data. It is therefore out of the question to give a complete survey of this field in the present article and to mention all the literature. In the different sections, we shall first describe what is known of the sequential stages constituting microbe ingestion, killing, and digestion. We shall examine at the end of each section how pathogenic microbes can avoid or inhibit these different events, by giving the most characteristic examples. ”

11. Adhesion

The initial and obligatory event involved in the phagocytic process is the contact and adhesion of molecules or particles to the cell surface. The adhesion process was especially well studied for particles by using either latex beads or isolated cells such as bacteria or erythrocytes. Adhesion depends upon the surface properties of both the particle and the host cell. One of the factors which could play a role in cell adhesion is the surface electrostatic charges. Because vertebrate and more primitive cells such as amoebae bear a net negative charge given by glycoproteins and mucopolysaccharides of their glycocalyx (Braatz-Schade and Stockem, 1973; Brandt and Freeman, 1967a; Sherbert, 1978; Stockem, 1977; Weiss, 1969), one could expect that adhesion of positively charged molecules or particles would be easier and stronger than that of negatively charged ones on account of repulsive forces. This problem has been investigated either by using molecules or particles with different electric charges or by modifying the host cell surface charges. Despite the large amount of work on this topic, no clear conclusions can be drawn for several reasons: 1. According to Gingell and Vince (1982), the correlation between cell adhesion and cell surface electrostatic properties depends upon salt concentration.

PHAGOCYTE-PATHOGENIC MICROBE INTERACTIONS

289

2. A wide variety of particles (latex beads, oil droplets, bacteria, erythrocytes, chlorella, yeast) and phagocytes (ameboid cells, digestive cells of hydrae, slime mold, polymorphonuclear leukocytes, macrophages) have been used. 3. Experimental conditions differ from one study to another: modifications of surface charges were performed with different concentrations of polyanions or polycations. These charged molecules were applied either to particles or to phagocytes, before or while they were in contact. 4. Removal of negative charges from the host cell surface by neuraminidase could alter other cell properties. 5 . Finally, in most cases, the authors measured particle ingestion instead of adhesion. This may introduce some distortion because adhesion and ingestion are two distinct events that depend upon different parameters (Ito et af., 1981; Rabinovitch, 1967a). For instance, we observed that latex beads could saturate the cell surface of Dicryosfelium discoideum ameboid cells although few beads were ingested (unpublished results). Despite this confusing situation, it seems possible to conclude that cationic substances generally enhance adhesion (Beukers et af., 1980; Capo et af., 1981; Davier et af., 1981 ; Deierkauf et al., 1977; Kooistra et af., 1980; McNeil et al., 1981; Pruzanski and Saito, 1978; Westwood and Longstaff, 1976). As observed in electron microscopy (Capo et af.,1981), cationic substances favor the formation of large contact areas between the particle and the phagocyte. Similar results were obtained upon removal of negative charges mostly due to sialic acid residues from the erythrocyte surface by neuraminida: 2, prior to adhesion (Capo et af., 1981). However, electrostatic forces do not play the main role in particle adhesion because, in most cases, the prey and the phagocyte are both negatively charged (Beveridge, 1980; Brandt and Freeman, 1967a; Sherbert, 1978; Weiss, 1969). This is the case for primitive cells (Protozoa, slime molds) that feed upon bacteria, and for professional phagocytes (polymorphonuclear leukocytes, macrophages) that ingest prokaryotic and eukaryotic cells. The unspecific and specific receptor sites that have been discovered in recent years certainly play a more crucial role. The term unspecific binding sites is used for adhesion of latex, oil droplets, glutaraldehyde-fixed cells, and any misunderstood attachment (Benoliel et al., 1980; Rabinovitch, 1967b; Stossel, 1975; Vogel er al., 1980). In many cases, unspecific binding seems to be due to the hydrophobicity of particles or molecules that can promote physical forces leading to adhesion between particle and cell surface. From van Oss (1978) and Mudd et af. (1934) high interfacial tension between the particle and the surrounding medium but low interfacial tension against the phagocytic cells favor binding and also engulfment. Unfortunately, at the present time no theory can satisfactorily predict how alterations of any one of these properties will affect particle adhesion.

290

ANTOINETTE RYTER AND CHANTAL DE CHASTELLlER

In addition to hydrophobic forces, more specific binding sites have been found. Some of them behave as lectin receptors and recognize a specific sugar (glucose, mannose, galactose, etc.). They are located on the phagocyte cell surface (Glass et al., 1981; Stahl et al., 1978; Sung et al., 1983; Vogel et al., 1980; Warr, 1980; Weir and Ogmundsdottir, 1977). This was shown by the fact that uptake of bacteria or glycoproteins was inhibited when macrophages were previously treated by one of these sugars. Lectin-like receptors also exist on bacteria and yeast which no longer adhere to macrophages when they have been pretreated with mannose (Bar-Shavit et al., 1977; Sandin et al., 1982), or with N-acetyl-glucosamine (Levy, 1979). A similar situation was observed for the binding of yeast cells to Dictyostelium vegetative phase cells. When N-acetylglucosamine residues are blocked by wheat germ agglutinin or are missing from the cell surface of the phagocyte, yeast cells do not attach (Hellio and Ryter, 1980; Ryter and Hellio, 1980). Other binding sites seem to exist on pili (bacterial surface extensions). These proteinous thin filiaments have been shown to favor adhesion to varied kinds of cells (Beveridge, 1980; Pearce and Buchanan, 1980; Smith, 1977). Mono- and oligosaccharides containing a-D-mannose residues inhibit pili adhesion suggesting that these polysaccharidescorrespond to surface receptors for pili (Pearce and Buchanan , 1980). At last truly specific receptors were discovered some years ago on macrophages and leukocytes. Two receptors react specifically with the Fc portion of IgG molecules (Silverstein et af.,1977; Unkeless et al., 1981). One Fc receptor is trypsin, chymotrypsin, and pronase resistant and mediates the efficient binding and ingestion of IgG-antigen complexes (Arend and Mannik, 1972) and 1gGcoated particles (Griffin et al., 1975a; Mantovani et al., 1972). The other Fc receptor is trypsin-sensitive and binds to subclasses of IgG. Immunoglobulins of these subclasses are called cytophilic antibodies because they bind with high affinity to macrophage Fc receptors in the absence of antigens (Arend and Mannik, 1972). The other type of specific receptor found in professional phagocytes binds to C3, one of the complement proteins present in the serum. C3 is an inactive precursor molecule consisting of a heavy chain and a light chain joined by disulfide bridges (Silverstein et al., 1977). Specific proteases cleave a small fragment of the heavy chain, converting this inactive form into a molecule called C3b that specifically binds to C3 receptors (Reid and Porter, 1981). Both kinds of specific receptors are quite different and independent from nonspecific receptors (Griffin and Silverstein, 1974; Michl et al., 1976; Rabinovitch, 1967b). The respective role of Fc and C3b receptors in attachment or ingestion has been controvertial for many years. Some studies show that C3b receptors of both neutrophils and mononuclear phagocytes mediate attachment only whereas Fc receptors induce particle engulfment (Griffin, 1982; Hed, 1981; Hed and Stendahl, 1982). Others indicate that C3b also promotes ingestion (Muschel et al., 1977; Segerling et al., 1982; Shaw and Griffin, 1981). In a quite recent review,

PHAGOCYTE-PATHOGENIC MICROBE INTERACTIONS

29 1

Griffin (1982) discussed the difficulties linked to these studies and the possible reasons for these contradictory results. He finally proposed, as Ehlenberger and Nussenzweig (1977), that “the chief role of C3b and its receptors is to promote particle binding; the chief role of IgG and its receptors is to promote particle ingestion; the enhanced particle binding by C3b and its receptors serves to facilitate engagement of the cell’s phagocytic signal-generating Fc receptors by IgG.” Particle adhesion to cells bearing ligaad surface receptors is generally temperature and energy-independent (Benoliel et a l . , 1980; Griffin et a f . , 1975b) whereas unspecific binding may or may not depend upon these factors (Benoliel et al., 1980; Glynn, 1981; Michl et a f . , 1976; Rabinovitch, 1967b). Several authors have shown by electron microscopy that IgG-coated erythrocytes form tight and continuous contact areas with the host cell surface (Benoliel et al., 1980; Griffin et a f . , 1975b; Kaplan, 1977; Munthe-Kaas et al., 1976). This is not the case for C3 receptors which establish discontinuous contacts (Kaplan, 1977; Munthe-Kaas et al.. 1976). For unspecific or more specific binding sites, contacts differ according to the particle and the cell. In macrophages, glutaraldehyde-fixed erythrocytes and latex beads promote discontinuous binding areas whereas in Dictyostelium discoideum ameboid cells latex beads are tightly bound. In the latter cells, fixed yeast or bacteria are both loosely bound despite the presence of lectin-like receptors for yeast (Ryter and Hellio, 1980). This shows that no general rules can be established on the mode of attachment to specific or unspecific receptors. After describing the different factors intervening in particle adhesion, let us now see what pathways pathogens use to survive. Two kinds of pathogens must be distinguished: those able to multiply outside phagocytes or other cell types (bacteria and fungi), and the facultative or obligatory parasites that multiply inside phagocytes only. It is obvious that the first kind of pathogens must avoid adhesion and phagocytosis to survive whereas, for the second type, adhesion is a prerequisite for survival. Cell surface properties and environmental conditions favorable for their survival are thus quite different. We have seen that cell surface electrostatic charges do not seem to play the main role in adhesion because microbes and phagocytes are both negatively charged. Despite the repulsive forces thus generated, adhesion and ingestion generally take place very quickly. It is thus obvious that this factor does not intervene in the fate of both kinds of pathogens. In contrast, hydrophilic or hydrophobic properties of the microbe surface are much more important. The inhibitory effect of the hydrophilic bacterial surface was especially well studied in the case of smooth and rough strains of Escherichia cofi and Salmonella ophirnurium. The rough variants which are more hydrophobic are much better adsorbed and phagocytosed than the smooth hydrophilic variants that are more pathogenic (Nakano and Saito, 1970; Stendahl et al., 1973, 1981; Stjernstrom et al., 1977). A relationship between pathogenicity and the presence of hydrophilic

292

ANTOINETTE RYTER AND CHANTAL DE CHASTELLIER

capsules has also been shown for Bacteroides fragilis (Simon et al., 1982) and Streptococcus pneumoniae (Roberts, 1979). At last, slime produced in large amounts by Pseudomonas aeruginosa also seems to prevent adhesion and ingestion (Schwarzmann and Boring, 1971). In this case, however, it is not clear whether inhibition is related to the hydrophilic properties of the slime or (and) the presence of a glycoprotein which was shown to inhibit phagocytosis (Bishop et al., 1982). In conclusion, hydrophilic surface properties constitute one of the factors allowing several pathogenic bacteria to escape from adhesion. In contrast, it is not known whether adhesion and ingestion of facultative or obligatory parasites is favored by their surface hydrophobicity. The only clue suggesting that this factor can intervene is the presence of glycolipids on the cell wall of Mycobacterium species (Goren et al., 1980) which could confer strong hydrophobic properties. Obviously, the crucial factor in microbe adhesion is the presence of receptors on phagocytes or microbes. Lectin-like receptors are very efficient for adhesion, and their interaction with the polysaccharidic glycocalix of many bacteria and fungi has been well demonstrated for plant and animal cells (Costerton et al., 1981; Pistol, 1981; Smith, 1977). Although this type of linkage seems to play a role in pathogenicity (Costerton et al., 1981; Smith, 1977) its role in microbe uptake by professional phagocytes is still poorly documented. It was shown that lectin-like receptors implicated in bacterial pili adhesion do not necessarily enhance uptake of piliated bacteria by professional phagocytes. The relationship between the presence of pili and the rate of ingestion, particularly well investigated for Neisseria and E. coli, showed that pili generally enhance uptake of piliated E. coli strains (Blumenstock and Jann, 1982; Silveblatt et al., 1979) but reduce that of Neisseria gonorrhoeae (King and Swanson, 1978; Thomas et al., 1972; Witt et al., 1976). Attachment of intracellular parasites such as Trypanosoma cruzi (Alcantara and Brener, 1980; Nogueira and Cohn, 1976), Chlamydiapsittaci, and C . trachomatis (Byme and Moulder, 1978; Levy, 1979) seems to be mediated by a protease-sensitive component of the host cell surface, which has not yet been identified as a lectin. Fc and C3 receptors found on the professional phagocyte surface are of great importance in defense mechanisms. The numerous studies on the effects of opsonization illustrate the importance of these receptors for microbe adhesion and ingestion. The increased phagocytosis of opsonized microbes was demonstrated for a wide variety of bacteria, fungi, and Protozoa. We will mention only one conclusive example concerning Mycoplasma: these microbes can grow on the surface of cultured fibroblasts or macrophages and are immediately ingested upon addition of serum (Jones and Hirsch, 1971). This means that antibodies present in the serum of immune animals or man considerably decrease the probability that pathogenic bacteria escape from adhesion. However, the presence of a

PHAGOCYTE-PATHOGENIC MICROBE INTERACTIONS

293

capsule can suppress the opsonization effect. Capsules are usually less antigenic than cell wall components and can mask antigen-antibody complexes located in a deeper layer, thus impairing binding between IgG and Fc receptors (Horwitz, 1982; Simon er al., 1982; Stinson and van Oss, 1971; Wilkinson et al., 1979). This shows that capsulated bacteria have a greater chance of avoiding adhesion because of their hydrophilic surface and this masking phenomenon. Opsonization of obligatory parasites favors their adhesion and engulfment but may also induce their killing after phagocytosis because, as described in Section IV, the presence of IgG on the prey’s surface triggers the microbicidal mechanism of professional phagocytes. Opsonization therefore decreases the chance of survival of all kinds of pathogens. It is not excluded, in contrast, that C3 receptors are implicated in Leishmania attachment to macrophages (Mauel, 1980) without inducing its killing in the absence of IgG, and in Babesia adhesion to the erythrocyte surface (Chapman and Ward, 1977). Another way of escaping from adhesion for nonparasitic pathogens is the production of vaned toxic substances that impair professional phagocytes or inhibit chemotaxis and phagocytosis. These substances are produced by many bacterial species (Alouf, 1976; Ofek er al., 1972; Schwab, 1975). Some are located on the cell wall, for example, protein M of Streptococcus (Fox, 1976; Jones and Schwab, 1970) and protein A of Staphylococcus (Dossett er al., 1969; Iwata and Uchida, 1980; Musher et al., 1981). Other toxins are excreted by bacteria such as Streptococcus (Alouf, 1980; Bernheimer, 1974; Roberts, 1979; Weissmann er al., 1963). Pseudomonas aeruginosa (Nonoyama et al., 1979), Borderella perrusis (Utsumi er al., 1978), Serratia miscotorum (Saeki et al., 1974), and Corynebacrerium diphtheria (Pappenheimer, 1977). Even viruses and tumor cells (Fauve and Hevin, 1980) can produce substances that inhibit PMN leukocyte chemotaxis and suppress the normal inflammatory response.

111. Ingestion Ingestion corresponds to particle capture by membrane extensions that surround and finally fully enclose the particle in a phagocytic vacuole. The shape of these cell processes and their connection with the particle surface differ according to the nature of the particle and the kind of phagocyte. In some cases, the host plasma membrane forms thin filopodia that establish discrete contacts with particles (McNeil et al., 1981). In other cases, it forms thin lamellipodia that remain tightly apposed to the particle surface throughout the engulfment process (Silverstein et al., 1977; Stossel, 1976; Stossel and Hartwig, 1976). This mode of ingestion was observed for opsonized particles phagocytosed by macrophages

294

ANTOINETTE RYTER AND CHANTAL DE CHASTELLIER

and is due to their interaction with Fc receptors (Griffin et al., 1975b). Griffin et al. (1975b, 1976) proposed that the initial interaction of immune ligands with the particle generates a process that requires the continuous apposition of receptors to ligands until the particle is fully enclosed within the phagocytic vacuole. This mechanism, called the “zipper mechanism,” was not shown for all receptors (Kaplan, 1977; MacRae et a f . , 1980; Munthe-Kaas et a f . , 1976; Ryter and Hellio, 1980). Inversely, during latex uptake that is not related to specific receptors but rather to bead hydrophobicity tight contacts are established (personal observation). Ingestion is quite rapid; in favorable conditions it occurs within 10-30 seconds after the first contact between phagocyte and prey (Bowers, 1980; MacRae ti?al., 1980). It is temperature dependent since it does not take place below 18’C in professional phagocytes (Silverstein et a f . , 1977) and 16°C in Acanthamoeba castelfanii for which the optimal growth temperature is 30°C (Bowers, 1977). This threshold is probably lower for primitive phagocytes growing around 20°C in their natural environment. Ingestion also requires metabolic energy (Stossel, 1975) but the nature of the cellular organelle consuming this chemical energy is unknown. One likely candidate is the contractile apparatus. On thin section electron micrographs one observes immediately below the membrane of the phagocytic cup a thick zone from which all cytoplasmic organelles are excluded by an anastomosing meshwork of microfilaments. Immunofluorescence (Stendahl et al., 1980; Valerius et a f . , 1981) and the heavy meromyosin labeling technique applied to glycerinated cells demonstrated that this meshwork contains actin, actin-binding proteins, and myosin (Taylor and Condeelis, 1979). It is clear that the appearance of the network closely follows particle adhesion but the mechanism which triggers contractile protein mobilization is not yet elucidated. Griffin et af.(1976) proposed that in the case of IgG-coated particles, the ligand-receptor interaction generates a signal (maybe the release of actinbinding protein from the plasma membrane) that would initiate polymerization and aggregation of contractile proteins and lead to the extension of phagocytic pseudopods. Pseudopod extension would bring about further receptor-ligand interaction and this in turn would generate further contractile protein association. This hypothesis is certainly very attractive but cannot account for the many cases in which contacts between particle and phagocytic membrane are loose and discontinuous. From scanning electron microscopic observations made during the engulfment process, it appeared that contact areas were randomly located along pseudopods or filopods (MacRae et a f . , 1980; Saint-Guillain e? a f . , 1980; personal observation). This means that the signal triggering contractile protein mobilization can propagate even in the absence of a progressive and permanent contact between particle and phagocyte. The factors triggering the dissociation of the filament network once a particle

PHAGOCYTE-PATHOGENIC MICROBE INTERACTIONS

295

has been internalized also remain unknown. Experiments in which ingestion was slowed down by particle coating with Con A showed that the zone of organelle exclusion corresponding to the microfilament network disappeared from the bottom of the phagocytic cup before phagosome closure (de Chastellier and Ryter, 1982). This suggests that phagosome closure is not a prerequisite for local microfilament network dissociation. Different electron microscope techniques have revealed changes in the phagocytic membrane during the ingestion process. One of these modifications, that seems to be directly related to the extension of phagocytic pseudopods and the mobilization of contractile proteins, was observed in Dictyostelium discoideum. The use of Oschman and Wall’s technique (glutaraldehyde fixation in the presence of calcium) (Oschman and Wall, 1972) showed that calcium deposits were especially abundant along the inner face of the filopod and phagocytic cup plasma membrane (de Chastellier and Ryter, 1981, 1982), that is to say where the membrane was underlayered with a filament network. Slackened phagocytosis and chemotaxis experiments indicated that the formation of calcium deposits was related to the mobilization of contractile proteins (de Chastellier and Ryter, 1982) and not to their dissociation. These calcium deposits seem to result from a phosphatase activity but their exact meaning is not yet established. The most attractive hypothesis is that they correspond to calcium channels, more especially as many authors have shown that calcium promotes phagocytosis (Hartwig et al., 1980; Stossel, 1975) and is more abundant in actin filament-rich cellular regions (Taylor et al., 1980). Changes in membrane coat polysaccharides were also found during endocytosis in different cells. Sialic acid residues were no longer detectable in pinocytic pits in an epithelial cell (de Bruyn er al., 1978), and wheat germ agglutinin receptors (mainly N-acetyl-glycosamine residues) disappeared from the phagocytic membrane of Dictyostelium discoideum during yeast ingestion (Ryter and Hellio, 1980). Several years ago, Brandt and Freeman (1967b) observed that in the amoeba Chaos chaos the plasma membrane electric resistence considerably decreased prior to the formation of endocytic channels; this was accompanied by a doubling of the electron transparent inner side of the membrane. A spin label study of macrophage plasma membrane also suggested that a nonrandom clustering of lipids and surface proteins occurred during phagocytosis (Horvath et al., 1981). Another modification concerning membrane lipids was reported by Karnovsky and Wallach (1961) and Sastry and Hokin (1966). These authors observed an increased phosphorus incorporation into phosphatidylinosito1 and phosphatidic acid during phagocytosis, that seemed to be related to phospholipid degradation. It is not known whether this phenomenon is related to changes in membrane viscosity (Berlin and Fera, 1977) or membrane depolarization (Horvath et al., 1981), also associated with phagocytosis. Frustrated phagocytosis performed by spreading eosinophils on large nonphagocytosable sur-

296

ANTOINETTE RYTER AND CHANTAL DE CHASTELLIER

faces coated with anti-antibody complexes showed that two new proteins became accessible to iodination in the early attachment phase (Thorne et al., 1980) as if vertical movements of proteins had taken place during this process. Very similar observations have been made in macrophages during phagocytosis (Howard et a f . , 1982). The observation of freeze-fractured preparations during phagocytosis showed, however, no modifications in the number, size, or distribution of intramembrane particles (Favard-SCrCno et al., 1981). This suggests that the changes in lipids, cell coat polysaccharides, or proteins observed during endocytosis do not correspond to a major redistribution of proteins in phagocytic membranes. All these data indicate that the plasma membrane undergoes an important remodeling during endocytosis. The reason for these changes remains obscure. In particular, it is not known to what extent they actually deal with plasma membrane distortion occurring during the engulfment process and its interaction with contractile proteins. Interaction between pathogens and phagocyte during ingestion has not yet been studied. It is obvious that ingestion of bacteria is ensured by the phagocyte and not by the prey, because inert particles and dead microbes are phagocytosed as well as living bacteria. Some bacteria may, however, inhibit their own phagocytosis by excreting toxins. By changing the phagocyte membrane permeability, the latter could prevent or block contractile protein mobilization as suggested by Manjula and Fischetti (1980) for Streptococcus M protein. Penetration of obligatory parasites for which ingestion is an obligatory event does not seem to correspond to an active mechanism either (Baker and Liston, 1978; Mauel, 1980). It is only in the case of blood parasites, such as Pfasmodiunrand Babesia, that penetration into erythrocytes is apparently ensured by the parasite itself (Aikawa et al., 1978). This situation is however very peculiar because red blood cells are devoid of phagocytic activity.

IV. Microbicidal Activity Before describing the process leading to microbe killing, it is important to point out that microbicidal properties vary considerably with the kind of phagocyte. Polymorphonuclear (PMN) leukocytes, short living cells that constitute the first manifestation of host cellular defense against microbe invasion by the blood stream, exhibit strong and well-defined cytotoxic properties. Those of macrophages that are long living cells vary according to the kind of macrophage (peritoneal, alveolar, monocyte-derived macrophages differentiating in vitro, etc.), the animal species (Nguyen et al., 1982), and mostly upon their physiological state (Cohn, 1978; Karnovsky and Lazdins, 1978; North, 1978). The latter depends upon the stimulations macrophages have been submitted to. Macrophages from unstimulated animals are termed “resident,” those stimulated by

PHAGOCYTE-PATHOGENIC MICROBE INTERACTIONS

297

immunological mechanisms in vivo are called “activated,” and those which come from animals having received an injection of various substances (caseinate, glycogen, peptone, thioglycolate, etc.) (North, 1978) are termed “elicited.” Both activated and elicited macrophages display a pronounced ruffling of the plasma membrane, an increased capacity for adhering to and spreading onto a substratum, an enhanced capacity for phagocytosis, and a high number of lysosomes. They exhibit physiological properties such as high acid hydrolase activity, excretion of neutral proteinases, specific elastase and collagenase, plasminogen activator. They can also produce cytotoxic compounds that will be described below. All these biochemical properties are expressed to varied extents in PMN leukocytes and elicited or activated macrophages (Cohn, 1978; Karnovsky and Lazdins, 1978; Nathan and Root, 1977). They also vary according to the substances used for elicitation (Briles et al., 1981). It is thus not surprising that the large amount of data accumulated for the past 10 years on the professional phagocyte microbicidal response to microbe infection often appear contradictory or at least confusing. This is also due to the fact that microbicidal activity corresponds to a quite complex phenomenon which is not totally elucidated. Roughly, cytotoxic properties of PMN leukocytes and macrophages are of two orders: the production of toxic oxidative compounds and that of toxic cationic proteins. A. OXIDATIVE KILLING Many years ago, Baldridge and Gerard (1933) observed that phagocytosis in leukocytes and granulocytes was accompanied by a marked oxygen consumption that was shown to be insensitive to cyanide (Klebanoff, 1982). In the past 20 years there have been major advances in our understanding of this phenomenon and its role in the microbicidal and cytotoxic properties of phagocytes (Allen, 1979; Badwey and Karnovsky, 1980; Klebanoff, 1982; Klebanoff and Clark, 1978; Weiss and Lobuglio, 1982). It is now well established that the enhanced oxygen consumption is used for production of hydrogen peroxide and other toxic products such as superoxide anions (0,- ), hydroxyl radicals (-OH), and singlet oxygen (‘0,). This phenomenon has been termed “respiratory burst” or “oxidative metabolism. Superoxide anion is the primary product of the respiratory burst. It results from the activity of a plasma membrane NADH or NADPH oxidase: ”

NAD(P)H

+ 2 0 2 + 2 02- + NAD(P) + H +

Hydrogen peroxide is formed from two superoxide anions by a dismutation reaction, in which one radical is oxidized and the other reduced: 02-

+ 0 2 - + 2 H + + 0 2 + H202

This dismutation can occur spontaneously or be catalyzed by superoxide dismutase (Fridovich, 1975).

298

ANTOINETTE RYTER AND CHANTAL DE CHASTELLIER

The toxicity of H,O, is enhanced many fold by its transformation into a variety of highly toxic substances. As an example, the presence of peroxidase, a halide, and H,O, generates HOCI, OC1- , and CI,, of which concentration is pH dependent (Klebanoff, 1982). This catalysis is of special interest because peroxidases are present in high concentrations in certain phagocytes (Daems et af., 1979; van Furth et al., 1970; Nichols et af., 1971; Root and Stossel, 1974; Simmons and Karnovsky, 1973). The release of PMN leukocyte myeloperoxidase into phagosomes has been detected by cytochemistry (Briggs et af., 1975b). In addition, the microbicidal effect of this enzyme was demonstrated in a very convincing way by Locksley el af. (1982). Living toxoplasms, uncoated or coated with eosinophil peroxidase, were given to resident peritoneal macrophages which lack peroxidase granules. The peroxidase-coated toxoplasms were killed whereas uncoated ones survived. Recent observations made in patients with myeloperoxidase deficiency indicate however that this enzymatic reaction is not a major microbicidal mechanism of phagocytic cells (Thong, 1982). In the presence of superoxide anions and traces of metal, hydrogen peroxide can also lead to the production of singlet radicals and hydroxide radicals which are both very toxic because they react with a wide variety of cellular compounds. The interest for singlet oxygen involvement in phagocytosis began with the finding that in PMN leukocytes phagocytosis is associated with the emission of light (Allen, 1979). Subsequent studies indicated that chemiluminescence is a general property of stimulated phagocytes and is due to two general mechanisms, one involving peroxidase and the other the production of superoxide anions (Klebanoff, 1982). This property is now widely used to estimate quantitatively the phagocyte oxidative response to the endocytic process. Stimulation of oxidative metabolism is a very rapid process (20 to 50 seconds) (Root and Stossel, 1974) triggered by molecule or particle adhesion. This event corresponds to the induction of a plasma membrane enzyme, the NADH or NADPH oxidase. Cytochemical staining for this enzyme applied to PMN leukocytes showed that the final reaction product was located on the outer face of the plasma membrane (Briggs et af., 1975a) (Fig. 1). This does not prove however that NAD(P)H oxidase is exposed to this side of the membrane (Badwey et af., 1980; Karnovsky et al., 1981). After phagocytosis, a cerium precipitate resulting from the cytochemical reaction was found along the phagosomal membrane (Fig. 3). This indicates that the killing process takes place during and after phagocytosis. Oxidative products are also released into the extracellular medium and can impair nearby microbes. The generation of oxygen metabolites was first studied in PMN leukocytes. It was more difficult to detect in macrophages than in PMN leukocytes, first, because the reaction is generally weaker, and second, because it varies according to the physiological state of macrophages (Badwey and Karnovsky, 1980; Cohn,

Fic. I . Cytochemical demonstration of H202formation by the technique of Briggs er al. (1975a) in a PMN leukocyte incubated for 20 minutes in the presence of phorbolymyristate acetate. The dense deposit is visible along the plasma membrane and in some pinosomes (observations made by Ryter et al.. 1983b). FIG. 2. The same cytochemical reaction as in Fig. I , performed in elicited alveolar macrophages. In this phagocyte, the oxidative activity is much weaker than in PMN leukocytes. The dense precipitate is visible in pinosomes but is absent from the plasma membrane (arrows) (Ryter et al., 1983b).

FIG. 3. Cytochemical demonstration of H202 production during phagocytosis of zymosan (2)by PMN leukocytes. The dense precipitate is visible along the plasma membrane and phagosome membrane (courtesy of Karnovsky er al.. 1981).

300

ANTOINETTE RYTER AND CHANTAL DE CHASTELLIER

1978; Johnston et al., 1978; Kamovsky and Lazdins, 1978; Nathan and Root, 1977). NADH oxidase activity can, in fact, be visualized by cytochemistry only in activated or elicited macrophages (personal observation) (Fig. 2).

Although NAD(P)H oxidase seems to be the triggering site of the oxidative burst, the mechanism of induction is not yet understood. From transmembrane potential measurements, Miles et al. (1981) proposed that superoxide anion production is related to the membrane depolarization observed during phagocytosis. This potential change would precede superoxide production and act as a signal for the phagocytic response. On the other hand, Badwey and Karnovsky (1980) showed that the signal that triggers the oxidative burst is Ca or Mg dependent. However, the role of these cations remains obscure because calcium at certain concentrations can also inhibit NAD(P)H oxidase (Lew and Stossel, 1980). Only certain kinds of molecules stimulate NAD(P)H oxidase. Nonopsonized microbes generally have no effect whereas IgG and some other substances [phorbolmyristateacetate (PMA), ionophores] induce this activity (Figs. 1 and 2) (Badwey and Kamovsky, 1980). The stimulation of the oxygen burst by immunoglobulins suggests that Fc receptors could be implicated in this mechanism (Henrichs et al., 1982). Their role in microbe killing was especially well demonstrated for opsonized Toxoplasma (Wilson et al., 1980), Trypanosoma cruzi (Nathan el al., 1979), and Trypanosoma dionisii (Thorne et al., 1978). In many other studies, however, the role of IgG as a trigger of the killing process was less conclusive because the stimulating effect of IgG on phagocytosis, killing, or digestion was not studied independently. The possible role of sialic acid in the initiation of the oxidative metabolism was also studied but the results are contradictory. One study suggested that it is required for the stimulation of this process (Tsan and McIntyre, 1976) whereas others showed that it is its removal by neuraminidase that stimulates superoxide production (Henrichs et al., 1982; Mills et al., 1981). Despite the lack of knowledge on the triggering mechanism of the oxidative burst, it is now admitted that the production of oxidative compounds plays a crucial role in the microbicidal properties of professional phagocytes. This was demonstrated by the relationship between 0, and H,O, production and killing properties, and especially by the study of phagocytes isolated from patients with chronic granulomatous disease that have poor oxidative properties and low bactericidal activity (Klebanoff and Clark, 1978; Quie, 1972). It is obvious that microbes that can survive inside professional phagocytes use defense mechanisms against the oxidative burst. Three possibilities can be envisaged: (1) parasites inhibit or do not trigger NAD(P)H oxidase activity; (2) they are insensitive to oxidative products; or (3) their enzymatic equipment allows them to rapidly destroy toxic substances or prevent their transformation into more toxic ones.

PHAGOCYTE-PATHOGENIC MICROBE INTERACTIONS

30 1

As said above, triggering of the oxidative burst may be prevented by the absence of opsonins on the microbe surface. In addition, the chemical surface properties of the microbe certainly play a role in this process but there are no data on this topic. The second mode of resistance concerns the microbe insensitivity to toxic substances. Resistance to hydrogen peroxide and 0,- was reported for Mycobacteria (Jackett er al., 1978; Mitchison et al., 1963). These authors showed that virulent strains are more resistant to H,O, than avirulent ones. Jackett et al. (1978) concluded, however, that the loss of virulence in attenuated strains is not necessarily linked to a loss of resistance; other factors are probably involved in the mechanism of attenuation. Toxoplasma and Trypanosoma were also shown to be very resistant to hydrogen peroxide (Murray and Cohn, 1979; Nathan et al., 1979). The enzymatic content of microbes may also influence their survival. We have already mentioned that the presence of peroxidase inside phagosomes leads to microbe killing because peroxidase in the presence of hydrogen peroxide and halide gives rise to highly toxic products. Microbes rich in this enzyme are thus more easily killed. Conversely, catalase that degrades H,O, and thereby prevents production of more toxic substances seems to be responsible for the resistance of many pathogenic bacteria (Leijh ef al., 1980). Finally, several bacterial species (Escherichia coli, Proteus, P seudomonas, Salmonella, Klebsiella) are quite rich in superoxide dismutase (Britton et al., 1978). These authors suggested that this enzymatic activity could increase bacterial resistance to superoxide radicals by rapidly degrading them. However, this resistance mechanism does not appear to be very efficient becuase it is generally admitted that 0,- radicals are much less toxic than products resulting from their dismutation.

B. CATIONIC PROTEINS Another antimicrobial agent, ‘‘phagocytin,” was discovered by Hirsch ( 1956) in rabbit granulocytes. Later studies demonstrated that it was localized in lysosomes of PMN leukocytes and macrophages of several animals and man (Avila, 1979). This agent is composed of several proteins that are highly cationic because they are rich in basic amino acids, mainly arginine (Zeya and Spitznagel, 1966, 1968). Their bactericidal activities differ with the kind of microbe (bacteria or fungi), the bacterial species, and even the bacterial strain (Elsbach et al., 1979; Patterson et al., 1980; Weiss et al., 1982; Zeya and Spitznagel, 1966, 1968). Some are especially active against gram-positive bacteria (Odeberg and Olsson, 1975) whereas others specifically kill gram-negative bacteria (Weiss et al., 1978). Their visualization under the light microscope with the fast green stain (Spitznagel and Chi, 1963)or under the electron microscope (Weiss et al., 1976)

302

ANTOINETTE RYTER AND CHANTAL DE CHASTELLIER

showed that they bind to the bacterial surface. This attachment is certainly mediated by the outer membrane lipopolysaccharides of gram-negative bacteria or the gram-positive cell wall carbohydrates. Their specificity probably resides in the different nature of these cell surfaces. The main effect of cationic proteins is to change the membrane permeability (Beckerditeet al., 1974; Elsbach et al., 1979; Zeya and Spitznagel, 1966). In the case of the protein isolated from rabbit PMN, that specifically kills gram-negative bacteria, Elsbach et af. (1979) showed that it is associated with phospholipase A. Its bactericidal properties are conserved even after dissociation from this enzyme, thus indicating that the killing property belongs to the cationic protein only. Cationic proteins display microbicidal activity only after they have been released into phagosomes and therefore depend upon phagosome lysosome fusion. Their mode of action is therefore different from that of the oxidative mechanism which is triggered before engulfment and takes place in all phagosomes whatever the time at which lysosome fusion occurs. This also suggests that the resistance of intracellular parasites to cationic proteins or respiratory burst is due to quite different microbe properties or behavior. To our knowledge, such killing mechanisms have not been described in primitive phagocytes for which microbe phagocytosis represents the main feeding process. These phagocytes ingest large amounts of bacteria and probably display efficient microbicidal properties because ingested bacteria are quickly impaired and digested. The first clue to the existence of similar killing properties in these cells resides in the quite recent discovery of a protein isolated from Entanioeba histolyrica that alters the membrane permeability of E . coli (Young et af., 1982). Since its mode of action is analogous to that of cationic proteins it could deal with microbe killing.

V. Phagosome-Lysosome Fusion The fist evidence of fusion between phagosomes and lysosomes was obtained several years ago by light microscopic observations of PMN leukocytes and activated macrophages (Hirsch, 1962; Hirsch et a f . , 1968; Zucker-Franklin and Hirsch, 1964). Dark granules, identified as lysosomes, were no longer visible after phagocytosis or induced pinocytosis. Their disappearance was due to fusions with newly formed endosomes and was named “degranulation.” The term of phagosome-lysosome fusion corresponds to phagosome fusion with primary or secondary lysosomes. In fact, reported observations generally concern fusions with secondary lysosomes because their Occurrence is based on the presence inside phagosomes of fluorescent or electron-dense markers pinocytosed before particle ingestion.

PHAGOCYTE-PATHOGENIC MICROBE INTERACTIONS

303

Fusions generally occur soon after phagosome formation. Quantitative measurements of phagosome-lysosome fusion in Dictyosrelium discoideum show a “burst” of fusions within 15 minutes after particle addition (Favard-SCrCno et al., 1981). Fusions are less frequent afterward although phagosomes and lysosomes are still available. It is interesting to note that fusions observed in virro among endosomes and lysosomes isolated from Acanrhamoeba casrellanii also occur mostly during the first 10-15 minutes of incubation (Oates and Touster, 1976, 1978). Absence of fusion is therefore not due to exhaustion of lysosomes but to another factor. In professional phagocytes, fusions are also very rapid and may even take place during particle ingestion. In this case, the lysosomal content can be released into the extracellular medium (Nichols, 1982; Pryzwansky er al., 1979).

Although the mechanisms controlling fusion are unknown, in virro studies performed with natural or artificial membranes have given some information on the conditions in which membrane fusion takes place (Papahadjopoulos er al., 1977; Poste and Allison, 1973). It is now well established that fusion occurs when the phospholipid layers of both membranes are in contact, which implies that membrane proteins must be mobile and pushed apart (Volsky and Loyter, 1978). Protein motility and segregation were demonstrated by freeze-fracture preparations in which phagosome-lysosome fusion areas were smooth and devoid of intramembrane particles (Amherdt er al., 1978; Batz and Wunderlich, 1976; Bowers, 1980; Favard-Strho er al., 1981; Orci et al., 1977). It is thus not surprising that membrane protein immobilization by cross-linking agents such as concanavalin A (Con A) and polyanionic substances inhibits membrane fusion. As shown by the following examples contradictory results were obtained by different authors, indicating that inhibition of fusion also depends upon other factors. Edelson and Cohn (1974) and Storrie (1979) reported that Con A-containing vacuoles do not fuse with secondary lysosomes whereas Goldman and Raz ( 1975) concluded their fusion because these vacuoles contain acid phosphatase. The same discrepancy was found between morphological observations and cytochemistry in Dictyostelium discoideum having phagocytosed Con A-coated yeast. The phagosome membrane remained tightly apposed to the Con A-coated yeast surface for at least 1 hour whereas it immediately loosened around untreated yeast, thus suggesting that fusions with lysosomes were inhibited in the presence of Con A. Cytochemistry showed, however, that acid phosphatase had been discharged into both kinds of phagosomes (Figs. 4, 5, and 6). A likely explanation would be that Con A inhibits fusion with large secondary lysosomes but does not prevent fusion with tiny primary lysosomes. The importance of vesicle size in the fusion process is in good accord with previous in virro experiments (Poste and Allison, 1973; Wilschert et al., 1981).

304

ANTOINETTE RYTER AND CHANTAL DE CHASTELLIER

FIGS. 4 A N D 5 Con A-coated yeast-containing phagosome. Despite the tight apposition of the phagosome membrane to the yeast particle surface, suggesting that no lysosome fusions have occurred (Fig. 4). acid phosphatase activity is cytochemically revealed inside the phagosome. The dense deposit remains located in discrete areas at the periphery of the yeast particle (Fig. 5 ) (personal observation). FIG. 6. Untreated yeast-containing phagosome showing a large space between the yeast surface and phagosome membrane. Acid phosphatase activity is visible inside the yeast particle (personal observation).

Polyanionic substances also seem to inhibit lysosome fusion (Alexander, 1981; Geisow et al., 1980; Goren et al., 1976; Hart and Young, 1978; Kielian and Cohn, 1982). As for Con A, however, the lysosome size plays a role because Kielian et al. (1982) also observed that dextran sulfate vacuoles did not fuse with large phagosomes but fused with small pinosomes. Cyclic AMP has also been suspected of inhibiting lysosome fusion. This assumption comes from the fact that a relation could be established between the low frequency of fusions between Mycobacterium tuberculosis-containing phagosomes and lysosomes and the increased cellular CAMPconcentration induced

PHAGOCYTE-PATHOGENIC MICROBE INTERACTIONS

305

by phagocytosis of this bacillus (Lowrie et al., 1975). The absence of such a cAMP increase after phagocytosis of bacteria unable to inhibit lysosome fusion strengthened this hypothesis (Lowrie et al., 1975; Carrol et al., 1979). A direct effect of cAMP on lysosome fusion was however not found in the course of in vitro experiments. Rather, this nucleotide seemed to stimulate fusion between isolated secondary lysosomes and pinosomes (Oates and Touster, 1980). The discrepancy between in vivo and in vitro systems could be due to the fact that cAMP affects several cellular functions in vivo (Confer and Eaton, 1982). Inhibition of fusion observed after Mycobacrerium uptake could be produced by other metabolic perturbations. Finally amines and NH,Cl inhibit phagosome lysosome fusion (Gordon et al., 1980; Hart and Young, 1975; Young er al., 1981). However, the mechanism of this inhibition remains obscure because other amines such as chloroquine seem to stimulate fusion (Hart and Young, 1978, 1979). In conclusion, studies on inhibition of lysosome fusion are still confusing partly for technical reasons. As discussed by Pesanti (1978) and Goren et al. (1980), fluorescent markers must reach a high concentration inside phagosomes in order to be detected. This means that phagosome fusion with a small number of tiny fluorescent lysosomes will not be seen. The electron microscope seems to be a more sensitive tool although a given thin section represents only a small fraction of phagosome and the marker is not necessarily present in this phagosome portion. The comparison of the number of phagosomes containing either an electron-dense marker or a fluorescent marker indeed showed that the number of fusions observed under the electron microscope was higher than that found by fluorescence (Goren et al., 1980). The fact that phagocytes were often loaded with fusion inhibitors for 3-5 days before particle ingestion could also distort the results because the accumulation of these substances could impair the digestive functions of the cells. The use of matrix-bound indicator dyes or fluorescent molecules has shown that the lysosome pH is around 4.5 to 6 (Heiple and Taylor, 1982; Jacques and Bainton, 1978; Kielian and Cohn, 1982; Mandell, 1970; Reijngoud and Tager, 1977). Geisow et al. (1981) and Segal et al. (1981) observed a slight pH increase 2-3 minutes before its lowering. They proposed that this short period of alkaline pH favors cationic protein binding to microbes and thereby accelerates their killing. It is not yet known, however, whether lysosome fusion and acidification are two functionally independent events or not. The dissection of these processes will require techniques with an increased spatial and temporal resolution. Two mechanisms could be responsible for the acidification of lysosome pH. Either a membrane bound H+ trans-ATPase driving H+ into lysosomes (Reijngoud and Tager, 1977) or a Donnan equilibrium generated by the presence of glycoproteins with low isoelectric point (Goldman and Rottenberg, 1973). At

306

ANTOINETTE RYTER AND CHANTAL DE CHASTELLIER

present, none of these hypotheses has been definitively proved (Reijngoud, 1978) but recent observations made in different laboratories are in favor of the first one. The acidic environment constitutes optimal conditions for lysosomal enzyme activity. The latter comprises several hydrolases active against lipids, nucleic acids, complex lipids, polysaccharides, glycoproteins, or proteins. In theory, many of these enzymes should degrade the lysosomal membrane. Although several authors have tried to determine whether the plasma membrane was modified after its internalization and contact with lysosomal enzymes, this problem is still a matter of debate. This question is difficult to study first, because the phagosome membrane fraction is more or less contaminated by other cell membranes, and second, because it is difficult to distinguish the modifications that occur during the engulfment process from those resulting from lysosomal enzyme degradation. The large amount of data obtained up to now seem to indicate however that the phagosome membrane is very resistant and that its degradation is neglectible during its life span (Steinman et a f . , 1983). Let us now see how parasites escape from cationic protein killing and lysosomal hydrolase degradation. Several ways are used: (1) inhibition of phagosome-lysosome fusion; (2) resistance to lysosomal enzymes; and (3) escape from phagosome. 1. Inhibition of phagosome-lysosome fusion. This mode of resistance seems to be the most commonly used by bacteria (Goren, 1977). It was especially well studied in the case of Mycobacterium tuberculosis and related strains (Armstrong and d'Arcy Hart, 1975; Hart et a f . , 1972). At least four substances produced by M . tuberculosis under appropriate conditions are susceptible to inhibit phagosome-lysosome fusion: sulfatide, polyglutamic acid, CAMP,and ammonia. Sulfatide (2,3,6,6'-tetraacetyl-trehalose-2'-sulfate) is a strong anionic substance which corresponds to the major representative of a group of unusual sulfated glycolipids produced by this bacillus (Middlebrook et a f . , 1959). A correlation between the production of sulfatides and virulence was established. Goren et a f . (1974, 1976) showed that yeast-containing phagosomes did not fuse with secondary lysosomes loaded with fluorescent or electron-dense markers when macrophages had previously pinocytosed sulfatides. It still remains to be shown that the amount of sulfatides naturally produced by M . tuberculosis is sufficient to inhibit lysosome fusion. In addition, its precise location inside the bacterial cell wall has not been established. It is thus not known whether it is in contact with the phagosome membrane or not (Goren et al., 1980). Electron micrographs show that the virulent strains M . tuberculosis (Armstrong and d'Arcy Hart, 1975) and M . avium (personal observation) both inhibit lysosome fusion and are surrounded in phagosomes by an electron-transparent capsule-like zone (Fig. 7) that does not seem to exist in in vitro growing bacteria. This material, also found in M . fepraemurium(Hart et al., 1972), was isolated

FIGS.7 AND 8. Macrophage phagosomes containing bacteria of the pathogenic strain M. aviurn (Fig. 7) or the nonpathogenic strain M. aurum (Fig. 8) (DL dense lysosomal material). In Fig. 7 , bacteria are surrounded by an electron-transparent zone (arrows) which is not visible in Fig. 8 (Frihel e t a / . , 1983).

308

ANTOlNETfE RYTER AND CHANTAL DE CHASTELLIER

and shown to correspond to “mycoside,” a complex peptidoglycolipid usually located in the Mycobacterium cell wall (Draper and Rees, 1973). It is interesting that the avirulent strain M. aurum is generally not surrounded by this “capsule” and is quickly digested (personal observation; Fig. 8). However, its role in fusion inhibition remains questionable because M. lepraemurium-containing phagosomes fuse with lysosomes. Polyglutamic acid, the accumulation of which in secondary lysosomes was shown to inhibit fusion with phagosomes (Hart and Young, 1978), is naturally covalently linked to the cell wall of virulent strains of M. tuberculosis (Vilkas and Markavits, 1972). The role of this substance is, however, subject to controversy concerning the kind of isomer present in the Mycobacrerium cell wall and its amount, which is not related to bacterial virulence (Draper, 1981). In addition, polyglutamic acid is probably associated with peptidoglycan and therefore is deeply buried in the cell wall. It is thus certainly not in contact with the phagosome membrane. As mentioned above, cyclic AMP concentration was shown to increase inside macrophages after phagocytosis of M. tuberculosis and M. microri (Lowrie er af., 1975). Such an increase was not found after infection by M. fepraemirrium and Salmonella ryphimurium which do not inhibit lysosome fusion (Carrol er af., 1979). The assumption of an inhibitory effect of cAMP was strengthened by the fact that attenuated strains of M. tuberculosis, which inhibit phagosome-lysosome fusion for a limited time, produced only a temporary rise in cAMP (Lowrie et af., 1975). The participation of this cyclic nucleotide is not definitively established, however, because cAMP inhibits other cell processes such as phagocytosis, the oxidative process, the release of lysosomal enzymes, and the killing of intracellular microorganisms (Confer and Eaton, 1982; Draper, 1981). Ammonia is formed in excess in unbalanced physiological conditions by M. tuberculosis and, as other amines, seems to inhibit lysosome fusion (Hart and Young, 1978; Seglen and Reith, 1976). That such a production occurs in phagosomes where nutritional conditions are presumably quite different from those of in virro growth remains to be proved. Lysosome fusion inhibition was also clearly demonstrated several years ago for Toxopfusma gondii (Jones and Hirsch, 1972). The reason for this inhibition is unknown but these authors observed that phagosomes containing intact parasites were surrounded by several mitochondria and endoplasmic reticulum cisternae. These organelles could protect phagosomes against oncoming lysosomes. A relationship between virulence, viability, and lysosome fusion was observed for several fungi and bacteria. This is the case for Coccidioides immitis (Beaman and Holmberg, 1980), Nocazdia arreroides (Davis-Scibienski and Beaman, 1980), and Brucelfa suis (Oberti er af., 1981). The resistance of these pathogens has been much less intensively studied than that of Mycobacteria, and the factors responsible for lysosome inhibition are still unknown.

PHAGOCYTE-PATHOGENIC MICROBE INTERACTIONS

309

Rickettsiae and Chlamydiae also escape from lysosome fusion. The way in which Chlamydiae avoid lysosome fusion is still poorly documented (Todd and Storz, 1975; Wyrick and Brownridge, 1978) although the multiplication and development of these parasites have been intensively studied (Becker, 1978; Storz and Spears, 1977). Eissenberg and Wyrick (1981) observed that after double infection with E. coli and Chlamydiae, E. coli was quickly digested whereas Chlamydiae remained intact. They concluded that the factor inhibiting lysosome fusion does not act on the general phagocyte behavior but intervenes locally at the level of Chlamydia-containingphagosomes. It was also shown that the number of phagocytosed microbes plays a crucial role in their survival. When the multiplicity of infection is one Chlamydia per macrophage, phagosomelysosome fusion appears to be inhibited and microbe survival is very high. In contrast, a high multiplicity (100 bacteria per phagocyte) leads to a low survival and phagosome-lysosome fusions were frequently observed (Wyrick and Brownridge, 1978). The comparison of Chlamydia survival in normal PMN leukocytes or in those isolated from patients with myeloperoxidasedeficiency or granulomatousdisease showed that oxygen-dependent antimicrobicidal activity is not essential for Chlamydia killing (Yong et al., 1982). This confirms that phagosome-lysosome fusion is the crucial stage upon which survival or death depends. The mechanism of lysosome fusion inhibition developed by Rickettsia rsutsugamushi seems to depend upon the cytoplasmic environment of their phagosomes. Rikihisa and Ito (1979) observed that phagosomes in which microbes were intact were surrounded by a high number of glycogen particles (Fig. 13). This would isolate phagosomes from other cytoplasmic organelles, and more especially lysosomes, and allow Rickettsiae to dissolve the phagosome membrane and become free in the cytoplasm before phagosome-lysosome fusion. It must be pointed out, however, that inhibition of fusion is never total whatever the pathogen. One always finds a portion of phagosomes that has fused and contains degraded microbes. The number of fusions and microbe survival seem to depend upon many parameters dealing with the number of infecting microbes, and the state of activation of phagocytes. One of the factors which seems to promote phagosome-lysosome fusion is microbe opsonization. The presence of IgG on the microbe surface could mask surface exposed substances that normally inhibit lysosome fusion. In most studies, however, IgG stimulation of lysosome fusion could not be distinguished from IgG stimulation of the oxidative burst and microbe killing. It is therefore quite possible that the increased number of fusions also results from the increased amount of killed microbes that can no longer inhibit lysosome fusion. 2. Resistance to lysosomal hydrolases. This second mechanism of survival inside phagosomes seems to be less commonly used by pathogens. Only three microbes are known to display this property.

310

ANTOINETTE RYTER AND CHANTAL DE CHASTELLIER

The first one is Mycobacferium lepraemurium for which phagosome-lysosome fusion was observed in cultured macrophages and fibroblasts (Brown and Draper, 1970, 1976; Hart et al., 1972) and in vivo (Brown and Draper, 1976). This microbe seems to be especially resistant to cationic proteins and lysosomal enzymes. The presence of “mycoside” which surrounds microbes inside phagosomes (Draper, 1981) could be responsible for this property. A rather similar compound has been found in other Mycobacferium species. Cytochemical staining for acid phosphatase showed that lead precipitate remained located at the periphery of this transparent material in phagosomes containing the pathogenic strain M . avium; in the case of M. aurum, a nonpathogenic strain that does not form this capsule, the precipitate was in direct contact with bacteria (Frkhel et al., 1983; Figs. 9 and 10). This suggests that this material could act as a barrier to lysosomal enzyme diffusion. As in the case of fusion inhibition, resistance to lysosomal enzymes is never total. Generally, many M . leprae are killed during the first days of infection. The survivers can later multiply and invade the host cell.

FIGS.9 AND 10. Cytochemical demonstration of acid phosphatase in phagosomes containing M. aviurn (Fig. 9) or M. aurum (Fig. 10). While the dense precipitate is in contact with M.aurum and

tends to penetrate inside bacteria, it remains located at the periphery of the transparent zone surrounding M. nurium in Fig. 9, as if this material inhibited lysosomal enzyme diffusion (Frkhel er al., 1983).

PHAGOCYTE-PATHOGENIC MICROBE INTERACTIONS

31 1

The two other pathogens presenting a high resistance to hydrolases are Protozoa: Leishmaniae and Trypanosomas. Phagosome-lysosome fusion was detected after Leishmania phagocytosis by using fluorescent or electron-dense markers (Alexander and Vickerman, 1975; Chang and Dwyer, 1976). The electron-dense lysosomal content was also detected inside Leishmania-containing phagosomes of bone marrow-derived macrophages (Ryter e? al., 1983a) (Fig. 11) and staining for acid phosphatase confirmed the presence of this hydrolase in phagosomes (Alexander and Vickerman, 1975; Lewis and Peters, 1977; Ryter et al., 1983). In addition the membrane impermeability of this pathogen was illustrated by the fact that lead precipitate was strictly located in the lysosomal material surrounding the parasites but not in the parasites. After modification of the membrane permeability by killing of Leishmania inside phagosomes (Rabinovitch et af., 1982; Ryter et al., 1983a) lead precipitate was found inside the parasites. The reason for such a great impermeability is not understood. Contrary to Mycobacterium species that have complex cell walls, Leishmania is limited by

FIG. I I . Cytochemical staining for acid phosphatase in macrophages infected with Leishmania mexicana amazoniensis (L). The dense deposit is visible inside phagosomes but remains located in the dense lysosomal material (DL) surrounding the parasites. The latter remain intact and devoid of precipitate except in a lysosome of a Leishmania (L) (Ryter er al.. 1983a).

312

ANTOINETTE RYTER AND CHANTAL DE CHASTELLIER

its plasma membrane only and is not protected by a thick plysaccharide coat (personal observation). Trypanosoma cruzi trypomastigotes also momentarily resist hydrolases. ‘Their phagosomes fuse with lysosomes (Kress et al., 1977; Milder and Kloetzel, 1980; Nogueira and Cohn, 1976) and although part of them are destroyed (Nogueira and Cohn, 1976) some resist and escape later into the host cytoplasm. The resistance of these parasites varies according to the species (Liston and Baker, 1978; Thorne et al., 1979) and their development stage (Nogueira and Cohn, 1976; Thorne et al., 1979). It is not established whether these variations reflect varied hydrolase resistances or varied sensitivities to microbicidal products. 3. Escape from phagosomes. Early evidence that intracellular parasites can escape from macrophage phagosomes and multiply in direct contact with the host cytoplasm was given with Trypanosoma cruzi (Nogueira and Cohn, 1976). This phenomenon seems to occur by the progressive degradation of the phagosome membrane (Fig. 12). Forty-eight hours after infection, most parasites are in the

FIG. 12. Transverse section through an intracellular trypomastigote (T) of Trypanosomacruzi 90 minutes after macrophage infection. The parasite membrane (short arrow) has a typical unit membrane structure with underlying subpellicular microtubules. Another modified membrane is also seen around the parasite. It is thinner and has only a single leaflet (dashed arrow). This membrane corresponds to the phagosome membrane in the course of its degradation (see inset at higher magnification) (courtesy of Nogueira and Cohn, 1976).

FIG.13. Rickettsia tsutsugamushi escaping from a phagosome. Intact Rickettsia-containing phagosomes are generally located in glycogen-richcytoplasmic (upper right inset) areas. Some parasites have "dissolved" the phagosome membrane and are free in the cytoplasm (lower left inset) (courtesy of Rikihisa and Ito, 1979).

314

ANTOINETTE RYTER AND CHANTAL DE CHASTELLIER

cytoplasm and start to multiply. A rather similar process was observed with Rickettsia species. As already mentioned, Rickettsia tsutsugamushi, after avoiding lysosome fusion because of its location in glycogen regions, seems to dissolve the phagosome membrane (Rikihisa and Ito, 1979; Fig. 13). R . rickettsii can also become free in the cytoplasm and is later found inside the endoplasmic reticulum (ER) (Silverman and Wisseman, 1979) (Fig. 14). The way in which it reaches the ER remains obscure but it could occur by budding. Its penetration is accompanied by a considerable swelling of the cisternae leading after 120 hours to the appearance of a huge and unique vacuole. The host cell finally lyses,

FIG. 14. Chicken embryo fibroblast infected with R. rickcrtsii. The rickettsia are located in dilated rough endoplasmic reticulum (short arrows) and are membrane-bound (long arrows) (courtesy of Silverman and Wisseman, 1979).

PHAGOCYTE-PATHOGENIC MICROBE INTERACTIONS

315

liberating the microbes which invade neighbor cells. The development of Rickettsia orientalis (Higashi, 1962) and Scrub Typhus Rickettsia (Erwing et a l . , 1978) is slightly different. Parasites free in the cytoplasm extrude directly from the host cell surface by a budding process. It must be pointed out, however, that Rickettsia infection and multiplication have been studied in their normal environment, that is to say in nonprofessional phagocytes that do not possess developed microbicidal properties and a digestive apparatus. They probably survive less easily in PMN leukocytes or macrophages. A recent morphological study of Legionella pneumophila infection showed that these bacteria were killed by PMN leukocytes but survived in macrophages when the number of bacteria was low (Katz and Hashemi, 1982). Some hours after infection, bacteria were found in ER cisternae. As no pictures of phagocytosis could be obtained at low multiplicity, it is not yet known whether bacteria were first located in phagosomes and then reached the ER. Another defense mechanism used from inside the phagosomes of phagocytes by nonobligatory parasites is the production of toxins that rapidly impair and sometimes lyse the host cell. As already mentioned in Sections I1 and 111, many bacteria contain or excrete toxins. In some cases, they may immediately prevent microbe ingestion but some of them can also act later. For example, it has been shown that surface lipids of Corynebacterium ovis induced the swelling of ER, Golgi apparatus, and phagosomes leading finally to macrophage lysis (Hard, 1973). Although no toxin has yet been identified for Legionella pneumophila, it is possible that macrophage destruction observed at high bacterial multiplicity (Katz and Hashemi, 1982) is due to toxin production. Bordetella pertusis was shown to contain and excrete large amounts of adenyl-cyclase (Confer and Eaton, 1982). This enzyme is internalized by phagocytosis and catalyzes the formation of cyclic AMP, thereby disrupting normal phagocyte functions and probably inducing their further killing. This could explain why humans infected by this pathogenic bacterium are quite vulnerable to secondary infection (Confer and Eaton, 1982). The mechanism of membrane penetration used by diphtheria toxin is very interesting because it is pH dependent (van Heyningen, 1981; Sandvig and Olsnes, 1980). The toxin molecule is composed of two peptides: one binds to a membrane receptor and, at low pH, promotes membrane crossing by the other peptide. After phagocytosis of Corynebacterium diphtheria, the toxin could cross the phagosome membrane after pH lowering thereby destroying the phagocyte even after bacterium killing and digestion. Another interesting pH-dependent escape mechanism has been discovered for certain viruses (Helenius et a l . , 1980; Genchault er a l . , 1981). Virus particles that have been phagocytosed can escape from phagosomes after pH lowering. The viral envelope fuses with the phagosome membrane thus liberating viral nucleic acid into the cytoplasm.

316

ANTOINETTE RYTER AND CHANTAL DE CHASTELLIER

VI. Membrane Recycling During Endocytosis Endocytosis results in an extensive interiorization of the plasma membrane varying between 1 and 20 times the total cell surface per hour, depending upon cell type and culture conditions (Bowers and Olszewski, 1972; Bowers et al., 1981; Githens and Karnovsky, 1973;Hubbard and Cohn, 1975; Ryter and de Chastellier, 1977; Steinman et al., 1976, 1983; Stockem, 1973; Weisman and Korn, 1967). This raises two main problems: (1) what is the fate of the interiorized membrane and (2) how is the plasma membrane renewed? As described in the preceding section, the main fate of endocytic vesicles is to fuse with lysosomes and deliver their content to them. Although the surface area of incoming vesicles is much larger than the dimensions of the preexisting secondary lysosomes, the membrane area of the total intracellular compartment preserves a constant value throughout endocytosis in both macrophages and L cells (Steinman er al., 1976). Similar observations were made in ameboid cells (Bowers er al., 1981; Ryter and de Chastellier, 1977). The maintenance of the vacuolar compartment at a constant surface area implies a rapid reduction in vesicle size. Degradation of membrane components by lysosomal enzymes is, however, too slow to ensure this equilibrium (Hubbard and Cohn, 1975). The second problem concerns plasma membrane renewal. As shown in macrophages and ameboid cells, phagocytosis or pinocytosis does not induce a reduction of the cell surface area (Bowers er al., 198I ; Ryter and de Chastellier, 1977; Steinman ef al., 1976). This means that the plasma membrane must be replaced at a corresponding rate. De novo synthesis of membrane constituents is too slow to account for the rapid replacement of internalized membrane (Silverstein er al., 1977; Werb and Cohn, 1972). In addition, all studies on the degradation rate of plasma membrane constituents point to low values (for review see Hubbard, 1978). All these considerations led several authors to propose that plasma membrane internalized during pinocytosis or phagocytosis was recycled back to the cell surface. Indirect evidence for such a recycling came from varied morphological studies in amoebae, professional phagocytes, or other cell types, using endocytic markers, such as HRP, yeast particles, latex beads (Bowers er al., 1981; Ryter and de Chastellier, 1977; Steinman et al., 1976), or noncovalently linked markers to unspecified membrane components, such as cationized ferritin (Farquhar, 1978). Direct evidence for membrane recycling has only recently become available by studying the fate of plasma membrane antigens during pinocytosis (Schneider et af., 1979a,b; Tulkens et al., 1980), or that of radioactive markers, covalently bound to specific plasma membrane constituents, during pinocytosis (Burgert and Thilo, 1983; Thilo and Vogel, 1980) or phagocytosis (de Chastellier er al., 1983; Muller et al., 1980; Storrie et al., 1981). After internalization in the vacuolar compartment, the antigen or label reappears at the cell surface

PHAGOCYTE-PATHOGENIC MICROBE INTERACTIONS

317

(Burgert and Thilo, 1983; Muller et al., 1980, 1983; Schneider et al., 1979b; Storrie et al., 1981; Thilo and Vogel, 1980; Tulkens et al., 1980). As shown by the autoradiographic study of Muller et al. (1983), this phenomenon occurs rapidly since labeled phagolysosome membrane constituents reappeared at the cell surface within 5-10 minutes. The concept of recycling raises the question of the organization of this membrane flow. Most likely, the incoming pathway is constituted by the endocytic vesicles, but the cytological nature of the outgoing pathway is unknown. During pinocytosis the interiorized membrane could be returned to the cell surface in the form of small vacuoles, thus establishing a direct shuttle between cell surface and endosomal membranes (Muller er af., 1983; Steinman et al., 1976). These small vesicles could be issued by endosomal membranes that have not yet fused with lysosomes or by secondary lysosomes. The hypothesis of Duncan and Pratten (1977) is that the bulk of the endocytic vesicle membrane is recycled before it has even come into contact with lysosomes. Recent work from Burgert and Thilo (1983) in the macrophage cell line P388D, is in accord with this hypothesis. A similar membrane shuttle was observed during phagocytosis between the phagosome membrane and plasma membrane. In macrophages, the autoradiographic study of Muller et af. (1980) showed that the radioiodinated phagolysosome membrane proteins returned rapidly to the cell surface, most likely via small vesicles that bud from the phagolysosome and subsequently fuse with the plasma membrane. In ameboid cells, in which the endosomal compartment area is at least equal to that of the cell surface (Bowers et af., 1981; Ryter and de Chastellier, 1977), the mixing of membrane between incoming endocytic vesicles and preexisting ones seems predominant with respect to the direct shuttle between phagosome and plasma membrane (de Chastellier er af., 1983). An alternative pathway would include the Golgi elements, on the basis that certain pinocytosed substances appear in the Golgi complex (Farquhar, 1978; Thyberg, 1980; Thyberg et al., 1980). The exact pathway taken by the interiorized membrane between the cell surface and Golgi complex is, however, not clearly established. The most likely explanation is that endocytic vesicles fuse first with lysosomes and then, for membrane retrieval, with the Golgi complex, as suggested by Tulkens et al. (1980) and the recent work of Schwarz and Thilo (1983). This intense membrane traffic between plasma membrane and endosomal membrane shows that the phagosomal membrane is not a stable structure. A continous communication exists between the phagosome content and the extracellular medium: pinocytic vesicles bring extracellular substances to phagosomes and, inversely, the phagosome content is transported outside the cell. The latter process could play a role in the transport of microbe antigens and their exposure to the macrophage surface. At the present time, no data have been obtained on this membrane flow in

318

ANTOINETTE RYTER AND CHANTAL DE CHASTELLIER

professional phagocytes infected by intracellular pathogens. It is quite possible that the presence of living microbes inside phagosomes disturbs membrane flow especially with microbes that inhibit lysosome fusion. Although the loose or tight contact of the phagosome membrane with the internalized particle does not necessarily reflect its capacity to fuse with lysosomes (de Chastellier et al., 1983, see also Section V), it is striking that the membrane of M.avium-containing phagosomes remains in tight apposition with the electron-transparent zone throughout microbe multiplication (personal observation), as if no exchanges took place with other membrane compartments. Inversely, Leishmania mexicana amazonensis-containing phagosomes that fuse with lysosomes are transformed after 24-48 hours into huge parasitophorous vacuoles (Chang, 1980; Rabinovitch et a!., 1982). The size increase which is partly due to fusions with pinocytic vesicles (M. Rabinovitch, personal communication) could reflect an unbalanced membrane flow between phagosome and plasma membrane that could favor parasite multiplication. In conclusion, membrane flow could be an important phenomenon in host-invader interplay and represents an interesting new field of research.

VII. Conclusions This brief review shows that host-invader interplay is a complex phenomenon, that deals with different stages of phagocytosis or of the digestion process according to the pathogen. The best understood mechanism of pathogen escape is that used by bacteria able to grow outside cells. They succeed in avoiding adhesion and phagocytosis because of their peculiar surface properties. In contrast, mechanisms of resistance developed by intracellular pathogens are poorly understood. This is due first to the complexity of cellular processes implicated in engulfment, microbicidal properties, and lysosome fusion. Although great advances have been made in the past 10 years in our knowledge of these events, their exact mechanism and control are still far from being elucidated. Second, pathogen resistance is multifactorial and only exceptionally the property of a single determinant. Microbes must first inhibit or resist the oxidative burst and then neutralize the cationic proteins and lysosomal enzymes either by preventing lysosome fusion or by resisting lysosomal content. The study of these different processes is difficult because they form a cascade of interdependent events. It is obvious that the crucial stage of resistance concerns the oxidative burst because when microbes have been impaired by oxidative products, they can no longer inhibit phagosome-lysosome fusion and lose their resistance properties to lysosomal components. In addition, pathogen resistance to these different processes is seldom an all or

PHAGOCYTE-PATHOGENIC MICROBE INTERACTIONS

319

none phenomenon and varies according to microbe multiplicity, which complicates the kinetic study of their behavior. Finally, their resistance also depends upon the defense properties of the phagocyte which vary, for macrophages, with their state of activation and their origin. It is thus difficult to draw general conclusions from results obtained under different experimental conditions. We have focused our attention mostly on pathogen resistance toward professional phagocytes because these cells represent the first cellular defense of animals and they are the best equipped cells for microbe killing. It is obvious that when microbes have escaped from phagocyte vigilance, they can invade cells of other tissues with a greater chance of surviving and multiplying. Although in many cases, intracellular parasite multiplication finally leads to host cell destruction, the real goal of the pathogen is not to kill the host cell but to use its machinery to grow. This situation is similar to that of prokaryotic cells living in association with Protozoa, ameboid cells, algae, aquatic invertebrates, and insects. In several cases this cohabitation corresponds to a real symbosis in which the parasite confers specific functions to the host cell. This phenomenon is observed in phagocytic cells, for example, in a strain of Amoeba proreus that survives only when it is colonized by bacteria (Jeon and Jeon, 1976). The latter multiply inside phagosomes that do not fuse with lysosomes. It is interesting to note that the behavior of invading parasites varies with the host cell. As an example, when Chlorella (a Cyanophycea) are phagocytosed by the A. proteus strain cited above, phagosome-lysosome fusions occur and they are digested; when they are ingested by Paramecium or Hydra, phagosomes do not fuse with lysosomes and these organisms become symbionts (Karakashia and Karakashia, 1973; Karakashia and Rudzinska, 1981; Muscatine er al., 1975). These two examples suggest that symbiont-host cell interplay probably depends upon resistance or inhibition mechanisms similar to those observed in pathogen-professional phagocyte interaction. One of the most studied symbioses is that of Rhizobium which confers its nitrogen fixation properties to the host plant cell. This situation is, however, different from the previous ones, because plant cells do not display a phagocytic activity and bacterial penetration corresponds to a complex process (Dart, 1975). In any case, and whatever the cell type, parasitism is of special interest because it could represent a preliminary stage of intracellular adaptation that would have led to the total state of symbiosis displayed by chloroplasts and mitochondria, according to a very attractive hypothesis (Stanier, 1970).

REFERENCES Aikawa, M . , Miller, L. H . , Johnson, J . , and Rabbege, J. (1978). J . Cell B i d . 77, 72-82 Alcantara, A . , and Brener, Z. (1980). Exp. Parusirol. 50, 1-6.

320

ANTOINETTE RYTER AND CHANTAL DE CHASTELLIER

Alexander, J. (1981). Exp. Parasirol. 52, 261-270. Alexander, J., and Vickerman, K. (1975). J . Prororool. 22, 502-508. Allen, R. C. (1979).In “Lysosomes in Biology and Pathology” (J. T. Dingle, J. Jacques, and I. H. Shaw, eds.), Vol. 6, pp. 197-233. North-Holland Publ., Amsterdam. Alouf, J. E. (1976). In “The Specificity and Action of Animal, Bacterial and Plant Toxins” (P. Cuatrecasas, ed.), Series B, Vol I, pp. 221-270. Chapman & Hall, London. Alouf, J. E. (1980). Pharmacol. Ther. 11, 661-717. Amherdt, M., Baggiolini, M., Perrelet, A., and Orci, L. (1978). Lab. Invest. 39, 398-404. Arend, W. P., and Mannik, M. (1972). J. Exp. Med. 136, 514-531. Armstrong, J. A., and d’Arcy Hart, P. D. (1975). J. Exp. Med. 142, 1-16. Avila, J. L. (1979). In “Lysosomes in Biology and Pathology” (J. T. Dingle, J. Jacques, and 1. H. Shaw, eds.), Vol. 6, pp. 235-266. North-Holland Publ., Amsterdam. Badwey. J. A., and Karnovsky, M. L. (1980). Annu. Rev. Biochem. 49, 695-726. Badwey, J. A., Cumutte, J. T., Robinson, J. M., Lazdins, J. K., Briggs, R. T., Karnovsky, M. J., and Karnovsky, M. L. (1980). J. Cell. Physiol. 105, 541-551. Baker, J. R., and Liston, A. J. (1978). J . Gen. Microbiol. 104, 79-89. Baldridge, C. W., and Gerard, R. W. (1933). Am. J. Physiol. 103, 235-236. Bar-Shavit, Z., Ofek, I., Goldman, R., Mirelman, D.. and Sharon, N. (1977). Biochem. Biophys. Res. Commun. 78,455-460. Batz, W., and Wunderlich, F. (1976). Arch. Microbiol. 109, 215-220. Beaman, L., and Holmberg, C. A. (1980). Infect. Immun. 28, 594-600. Becker, Y. (1978). Bacreriol. Rev. 42, 274-306. Beckerdite, S., Mooney, C., Weiss, L., Franson, R., and Elsbach, P. (1974). J. Exp. Med. 140, 396-409. Benoliel, A. M., Capo, C., Bongrand, P., Ryter, A,, and Depieds, R. (1980). Immunology 41, 547-560. Berlin, R. D., and Fera, J. P. (1977). Proc. Narl. Acud. Sci. U.S.A. 74, 1072-1076. Bemheimer. A. W. (1974). Biochim. Biophys. Acra 344, 27-50. Beukers, H., Deierkauf, F. A. Blom. C. P., Deierkauf, M., Scheffers, C. C., and Rienezsma, J. C. (1980). Chem. Biol. Interact. 33, 91-100. Beveridge, T. J. (1980). Can. J . Microbiol. 26, 643-653. Bishop, 0..Orr, T.. and Bartell, P. F. (1982). Infecr. Immun. 37, 378-381. Blumenstock, E., and Jann, K. (1982). Infect. Immun. 35, 264-269. Bowers, B. (1977). Exp. Cell Res. 110, 409-417. Bowers, B. (1980). J . Cell Biol. 84, 246-260. Bowers, B., and Olszewski, T. E. (1972). J . Cell Biol. 53, 681-694. Bowers, B., Olszewski, T. E., and Hyde, J. (1981). J . Cell Biol. 88, 509-515. Braatz-Schade, K., and Stockem, W. (1973). Z. Zellforsch. 137, 97-109. Brandt, P. W., and Freeman, A. R. (1967a). J. Colloid Interface Sci. 25, 47-56. Brandt, P. W., and Freeman, A. R. (1967b). Science 155, 582-585. Briggs, R. T., Drath, D. B., Kamovsky, M. L., and Kamovsky, M. J. (1975a). J. Cell Biol. 67, 566-586. Briggs, R. T., Karnovsky, M. L., and Karnovsky, M. J. (1975b). J. Cell Biol. 64, 254-260. Briles, D. E., Lehmeyer, J., and Forman, C. (1981). Infect. Immun. 33, 380-388. Britton, L., Malinowski, D. P., and Fridovich, I. (1978). J. Bucreriol. 134, 229-236. Brown, C. A., and Draper, P. (1970). J . Parhol. 102, 21-26. Brown, I. N., and Draper, P. (1976). Infecr. Immun. 13, 1199-1204. Bruyn, de, P. P. H., Michelson, S., and Becker, R. P. (1978). J . Cell Biol. 78, 379-389. Burgert, H. G., and Thilo, L. (1983). Exp. Cell Res. 144, 127-142. Byme. G. I., and Moulder, J. W. (1978). Infect. Immun. 19, 598-606.

PHAGOCYTE-PATHOGENIC MICROBE INTERACTIONS

32 1

Capo, C., Bongrand, P., Benoliel, A. M., Ryter, A,, and Depieds, R. (1981). Ann. Immunol. 132 D, 165-173. Carrol, M. E. W., Jackett, P. S., Aber, V. R.. and Lowrie, D. B. (1979). J. Gen. Microbiol. 110, 42 1-429. Chang, K. G. (1980). Science 209, 1240-1242. Chang, K. P., and Dwyer, D. M. (1976). Science 193, 678-680. Chapman-Andersen, C. (1977). Physiol. Rev. 57, 371-385. Chapman, W. E., and Ward, P. A. (1977). Science 196, 67-70. Chastellier, de, C., and Ryter, A. (1981). Biol. Cell. 40, 109-118. Chastellier, de, C., and Ryter, A. (1982). Biol. Cell. 43, 121-128. Chastellier, de, C., Ryter, A., and Thilo, L. (1983). Eur. J . Cell Biol. 30, 233-243. Cohn, Z. A. (1978). J. Immunol. 121, 813-816. Confer, D. L., and Eaton, J. W. (1982). Science 217, 948-950. Costerton, J. W., Irvin, T., and Cheng, K. 1. (1981). Annu. Rev. Microbiol. 35, 299-324. Daems, W. T., Roos, D., van Berkel, Th. J. C., and van der Rhee, H. J. (1979). I n “Lysosomes in Biology and Pathology” (J. T. Dingle, P. J. Jacques, and I. H. Shaw, eds.). Vol. 6, pp. 463-514. North-Holland Publ., Amsterdam. Dart, P. J. (1975). In “The Development and Functions of Roots” (J. G. Torrey and D. T. Clarkson, eds.), pp. 467-506. Academic Press, New York. Davier, P. F., Rennke, H. G., and Cotran, R. S. (1981). J. Cell Sci. 49, 69-86. Davis-Scibienski, C., and Beaman, B. L. (1980). Infecr. Immun.28, 610-619. Deierkauf, F. A., Beukers, H., Deierkauf, M., and Riemzsma, J. C. (1977). J. Cell. Physiol. 92, 169- 175. Dossett, J. H., Kronvall, G., Williams, R. G., Jr., and Quie, P. G. (1969). J . Immunol. 103, 1405- 1410. Draper, P. (1981). In “Microbial Perturbation of Host Defense” (F. O’Grady and H. Smith, eds.), pp. 143-157. Academic Press, New York. Draper, P., and Rees, R. J. W. (1973). J. Gen. Microbiol. 77, 79-87. Duncan, R., and Pratten, M. K. (1977). J. Theor. Biol. 66, 727-735. Edelson, P. J., and Cohn, Z. A. (1974). J. Exp. Med. 140, 1364-1386. Ehlenberger, A. G., and Nussenzweig, V. (1977). J . Exp. Med. 145, 357-371. Eissenberg, L. G., and Wyrick, P. B. (1981). Infecr. Immun.32, 889-896. Elsbach, P., Weiss, J., Franson, R. C., Beckerdite-Quagliata, S., Schneider, A., and Harris, L. (1979). J. Biol. Chem. 254, 11OOO-11009. Erving, E. P., Takeuchi, A., Shirai, A., and Osterman, J. V. (1978). Infecr. Immun. 19, 1068- 1075. Farquhar, M. G. (1978). J . Cell Biol. 77, R35-R42. Fauve, R., and Hevin. B. (1980). In “The Host Invader Interplay” (H. van Bossche, ed.), pp. 285-292. Janssen Research Foundation. Elsevier, Amsterdam. Favard-Skkno, C., Ludosky, M. A., and Ryter, A. (1981). J . Cell Sci. 51, 63-84. Fox, E. N. (1976). Eacreriol. Rev. 38, 57-86. Frkhel, C., Ryter, A,, Rastogi, N., and David, H. L. (1983). In preparation. Fridovich, 1. (1975). Annu. Rev. Biochem. 44, 147-159. Furth, van, P., Hirsch. J. G.. and Fedorko, M. E. (1970). J . Exp. Med. 132, 794-812. Geisow, M. I., Beaven, G. H., d’Arcy Hart,P., and Young, M. R. (1980). Exp. Cell Res. 126, 159-165. Geisow, M. J., d’Arcy Hart,P., and Young, M. R. (1981). J . CellEiol. 89, 645-652. Genchault, J. L., Steffan, A. M., Bingen, A., and Kim, A. (1981). Virology 112, 375-384. Gingell, D., and Vince, S . (1982). J. Cell Sci. 54, 255-285. Githens, S., 111, and Karnovsky, M. L. (1973). J. Cell Biol. 58, 536-548.

322

ANTOINETTE RYTER AND CHANTAL DE CHASTELLIER

Glass, E., Stewart, J., and Weir, D. M. (1981). Immunology 44, 529-534. Glynn, P. J. (1981). Cyrobios 30, 153-166. Goldman, R.. and Raz, A. (1975). Exp. Cell Res. 96, 393-405. Goldman, R., and Rottenberg, H. (1973). FEES Len. 33, 233-238. Gordon, A. H., d’Arcy Hart, P., and Young, M. R. (1980). Narure (London) 286, 79-80. Goren, M. B. (1977). Annu. Rev. Microbiol. 31, 507-533. Goren, M. B., Brokl, 0.. and Schaefer, W . B. (1974). Infecr. Immun. 9, 142-149. Goren, M. B., d’Arcy Hart, P., Young, M. R.,and Armstrong, J. A. (1976). Proc. Narl. Acad. Sci. U.S.A. 73, 2510-2514. Goren, M. B., Swendsen, C. L., and Henson, J. (1980). I n “Mononuclear Phagocytes: Functional Aspects” (R. van Furth, ed.), pp. 999-1038. Nijhoff, The Hague. Griffin, F., Jr. (1982). In “Advances in Host Defense Mechanisms” (J. F. Gallin and A. S. Fauci, eds.), Vol. I, pp. 31-55. Raven, New York. Griffin, F. M., Jr., and Silverstein, S. C. (1974). J. Exp. Med. 139, 323-336. Griffin, F. M., Jr., Bianco, C., and Silverstein, S. C. (1975a). J . Exp. Med. 141, 1269-1277. Griffin, F. M., Jr., Griffin, J . A., bider, J. E., and Silverstein, S. C. (1975b). J. Exp. Med. 142, 1263- 1282.

Griffin, F. M., Jr., Griffin, J. A., and Silverstein, S. C. (1976). J . Exp. Med. 144, 788-4309, Hard, G. C. (1973) Infecr. Immun. 12, 1439-1449. Hart, d’Arcy, P.,and Young, M. R. (1975). Narure (London) 256, 47-49. Hart, d’Arcy, P., and Young, M. R. (1978). Exp. Cell Res. 114, 486-490. Hart, d’Arcy, P., and Young, M. R. (1979). Exp. CellRes. 118, 365-375. Hart, d Arcy, P., Armstrong, J. A., Brown, C. A,, and Draper, P. (1972). Infecr. Immun. 5, 803-807.

Hartwig, 1. H., Yin, H. L., and Stossel, T. P. (1980). In “Mononuclear Phagocytes: Functional Aspects” (R.van Furth, ed.), Part 11, pp. 971-996. Nijhoff, The Hague. Hed, J. (1981). Monogr. Allergy 17, 92-111. Hed, J., and Stendahl, 0. (1982). Immunology 45, 727-736. Heiple, J. M., and Taylor, D. L. (1982). J. Cell Biol. 94, 143-149. Helenius, A., Kartenbeck, J., Simons, K., and Fries, E. (1980). J. Cell Biol. 84, 404-420. Hellio, R., and Ryter, A. (1980). J. Cell Sci. 41, 89-104. Henrichs, P. A. J., van Erne-van der Tol, E., and Verhoef, J. (1982); Anronie Leeuwenhoek 48, 191- 192.

Heyningen, van, S. (1981). Nature (London) 292, 293-294. Higashi, N. (1962). I n “Recent Advances in Studies of Tsutsugamushi Disease in Japan” (T. Tamija, ed.), pp. 145-156. Medical Culture. Hirsch, 1. G. (1956). J . Exp. Med. 103, 589-611. Hirsch, J. G. (1962). J . Exp. Med. 116, 827-834. Hirsch, J . G., Fedorko, M. E., and Cohn, Z. A. (1968). J . Cell Biol. 38, 629-632. Holter, H. (1959). Inr. Rev. Cyrol. 8, 481-504. Horvath, L. I., Bagyinka, C. S., and Sandor, M. (1981). Period. Biol. 83, 143-145. Horwitz, M. A. (1982). Rev. Infect .Dis. 4, 104-123. Howard, F. D., Petty, H. R., and Mc Connell, H. M. (1982). J. Cell Biol. 92, 283-288. Hubbard, A. L. (1978). In “Transport of Macromolecules in Cellular Systems,” pp. 363-398. Dahlem-Konferenzen. Abakon, Berlin. Hubbard, A. L., and Cohn, Z. A. (1975). J . Cell Biol. 64, 461-479. Ito, T., Ueda, M. J., Okada, I. S., and Ohnishi, S. I. (1981). J . Cell Sci. 51, 189-201. Iwata, K., and Uchida, K. (1980). I n “The Host Invader Interplay” (H. van Bossche, ed.), pp. 319-325. lanssen Research Foundation. Elsevier, Amsterdam. Jackett, P. S., Aber, V. R., and Lowrie, D. B. (1978). J. Gen. Microbiol. 104, 37-45.

PHAGOCYTE-PATHOGENIC MICROBE INTERACTIONS

323

Jacques, P. J. (1969). In “Lysosomes in Biology and Pathology” (J. T. Dingle and H. B. Fell, eds.), Vol 2, pp. 395-420. North-Holland Publ., Amsterdam. Jacques, Y. V., and Bainton, D. F. (1978). Lab. Invest. 39, 179-185. Jeon, K. W., and Jeon, M. S. (1976). J . Cell. Physiol. 89, 337-344. Johnston, R. B., Godzik, C. A., and Cohn, 2 . A. (1978). J. Exp. Med. 148, 115-127. Jones, T. C.. and Hirsch, J. G . (1971). J . Exp. Med. 133, 231-259. Jones, T. C.. and Hirsch, J. G. (1972). J. Exp. Med. 136, 1173-1194. Jones, J. M., and Schwab. J. H. (1970). Infect. Immun. 1, 232-242. Kaplan, G. (1977). Scund. J. Immunol. 6, 797-807. Karakashian, M. W., and Karakashian, S. J. (1973). Exp. Cell Res. 81, 1 1 1-119. Karakashian, S. J., and Rudzinska, M. A. (1981). Exp. Cell Res. 131, 387-393. Kamovsky, M. L., and Lazdins, J. K. (1978). J . Immunol. 121, 809-813. Kamovsky, M. L., and Wallach, P. F. H. (1961). J. Biol. Chem. 236, 1895-1901. Karnovsky, M. J., Robinson, J. M., Briggs, R. T . , and Karnovsky, M. L. (1981).Histochem. J. 13, 1-22. Katz, S. M., and Hashemi, S. (1982). Lab. Invest. 46, 24-38. Kielian, M. C., and Cohn, Z. A. (1982). J . Cell Biol. 93, 875-882. Kielian, M. C.. Steinman, R. M., and Cohn, 2 . A. (1982). J . Cell Biol. 93, 866-874. King, G. J., and Swanson, J. (1978). Inject. Immun. 21, 575-584. Klebanoff, S. J. (1982). In “Advances in Host Defense Mechanisms” (J. 1. Gallin and A. S. Fauci, eds.), Vol. I, pp. 1 1 I- 162. Raven, New York. Klebanoff, S. J., and Clark, R. A. (1978). “The Neutrophil: Function and Clinical Disorders.” North-Holland h b l . , Amsterdam. Kooistra, T., Duursma, A. M., Bouma, J . H. W.. and Gurber, M. (1980). Biochim. Biophys. Acru 631, 439-450. Kress, Y., Tanowitz, H., Bloom, B., and Wittner, M. (1977). Exp. Parasitol. 41, 385-396. Leijh, P. C. J., van Zwet, Th.L., and van Furth, R. (1980). Infect. Immun. 30,421-426. Levy, N. J . (1979). Infect. Immun. 25, 946-953. Lew, P. D., and Stossel, T. P. (1980). J. Clin. Invest. 67, 1-9. Lewis, D. H., and Peters, W. (1977). Ann. Trop. Med. Parusitol. 71, 295-312. Liston, A. J . , and Baker, J. R. (1978). J. Gen. Microbiol. 107, 253-262. Locksley, R., Wilson, C. B., and Klebanoff, S. J. (1982). J. Clin. Invest. 69, 1099- I I 1 I . Lowrie, D. B., Jackett, P. S . , and Ratcliff. N. A. (1975). Nature (London) 254, 600-602. MacRae, E. K., Pryzwansky, K. B., Cooney, M. H., and Spitznagel, J. K. (1980). Cell Tissue Res. 209, 65-70. McNeil, P. L. (1981). J. CellSci. 49, 311-340. McNeil, P. L., Hohman, T. C., and Muscatine, L. (1981). J. Cell Sci. 52, 243-269. Mandell, G. L. (1970). Proc. Soc. Exp. Eiol. Med. 134, 447-449. Manjula. B. W., and Fischetti, V. A. (1980). J . Exp. Med. 151, 695-708. Mantovani, B. Rabinovitch, M., and Nussenzweig, V. (1972). J . Exp. Med. 135, 780-792. Mauel. J. (1980). In “The Host Invader Interplay” (H. van Bossche. ed.), pp. 165-177. Janssen Research Foundation. Elsevier, Amsterdam. Michl, J., Ohlbaum, D. J., and Silverstein, S. C. (1976). J. Exp. Med. 144, 1465-1483. Middlebrook, G., Coleman, C. M., and Schaefer. W. B. (1959). Proc. Nut/. Acad. Sci. U.S.A. 45, 1801- 1804. Milder, R., and Kloetzel, J. (1980). Parasitology 80, 139-145. Miles, P. R., Bowman, L., and Costranova, V. (1981). J . Cell. Phvsiol. 106, 109-117. Mills, E. L., Debets-Ossekopp, Y., Verbrugh. H. A , , and Verhoef, J. (1981). Infect. Immun. 32, 1200- 1205. Mitchison, D. A., Selkon, J. B., and Llyod, J. (1963). J. Pathol. Bacteriol. 86, 377-386.

324

ANTOINETTE RYTER AND CHANTAL DE CHASTELLIER

Mudd, S., McCutcheon, M., and Lucke, B. (1934). Physiol. Rev. 14, 210-275. Muller. W. A., Steinman, R. M., and Cohn, Z. A. (1980). J . Cell Biol. 86, 304-314. Muller, W. A., Steinman, R. M., and Cohn, Z. A. (1983). J. Cell Biol. %, 29-36. Munthe-Kaas. A. C., Kaplan, G., and Seljelid, R. (1976). Exp. Cell Res. 103, 201-212. Murray, H. W . , and Cohn, Z. A. (1979). J. Exp. Med. 150, 938-949. Muscatine, L., Cook, C. B., Pardy, R. L., and Pool, R. R. (1975). Symp. SOC.Exp. Biol. Med. 29, 175-203. Muschel, R. J., Rosen, N., and Bloom, B. R. (1977). J . Exp. Med. 145, 175-186. Muscher, D. M., Verbrugh, H. A., and Verhoef, J. (1981). J . Immunol. 127, 84-88. Nakano. M., and Saito, K. (1970). Jpn. J . Microbiol. 14, 451-461. Nathan, C. F., andRoot, R. K. (1977). J . Exp. Med. 146, 1648-1662. Nathan, C. E., Nogueira, N., Juangbhanich, C., Ellis, J., and Cohn, Z. A. (1979). J . Exp. Med. 149, 1056-1068. Nguyen, 9 . Y.,Peterson, P. K., Verbrugh, H. A., Quie, P. G . , and Hoidal, J. R. (1982). Infect. Immun. 36, 504-509. Nichols, B. A. (1982). Lab. Invest. 47, 235-246. Nichols, 9. A.. Bainton, D. F., and Farquhar, M. G. (1971). J. Cell Biol. 50, 498-515. Nogueira, N., and Cohn, Z. (1976). J . Exp. Med. 143, 1402-1420. Nonoyama, S., Kojo, H., Mine, Y., Nishida, M., Goto, S., and Kuwahara, S. (1979). Infect. Immun. 24, 399-403. North, R. J. (1970). Semin. Hemarol. 7, 162-171. North, R. J. (1978). J. Immunol. 121, 806-816. Oates, P. J., and Touster, 0. (1976). J . Cell Biol. 68, 319-338. Oates, P. J., and Touster, 0. (1978). J. Cell Biol. 79, 217-234. Oates, P. J., and Touster, 0. (1980).J . Cell Biol. 85, 804-810. Oberti, J., Caravano, R., and Roux, J. (1981). Ann. Insr. Pasreur Immunol. 132 D, 201-206. Odeberg. H.,and Olsson, I. (1975). J. Clin. Invest. 56, I 1 18-1 124. Ofek, I., Bergner-Rabinowitz, S., and Ginsberg, I. (1972). Infect. Immun. 6, 459-464. Orci, L., Perrelet, A., and Friend, D. C. (1977). J. Cell Biol. 75, 23-30. Oschman, J. L., and Wall, B. J. (1972). J . Cell Biol. 55, 58-73. Oss, van, C. J. (1978). Annu. Rev. Microbiol. 32, 19-39. Papahadjopoulos. W . , Vail, J., Newton, C., Nir, S., Jacobson, K.,Poste, G., and Lazo, R. (1977). Biochim. Biophys. Acra 465, 579-598. Pappenheimer, A. M. (1977). Annu. Rev. Biochem. 46, 69-94. Patterson-Delafield, J., Martinez, R. J., and Lehrer, R. I. (1980). Infect. Immun. 30, 180-192. Pearce, W. A., and Buchanan, T. M. (1980). I n “Bacterial Adherence” (E. H. Beachey, ed.), Series B, Vol. 6, pp. 291-340. Chapman & Hall, London. Pesanti, E. I. (1978). Infect. Immun. 20, 503-511. Pistel, T. G. (1981). Annu. Rev. Microbiol. 35, 85-1 12. Poste, G., and Allison, A. C. (1973). Biochim. Biophys. Acra 300, 421-465. Pruzanski, W., and Saito, S. (1978). Exp. CeNRes. 117, 1-13. Pryzwansky, K. B., MacRae, E. K.,Spitznagel, J. K., and Cooney, M. H. (1979). Cell 18, 1025- 1032. Quie, P. G. (1972). In “Phagocyte Mechanisms in Health and Disease” (R. C. William and H. D. Fudenberg, eds.), pp. 139-150. Intercontinental Medical, New York. Rabinovitch, M. (1967a). Exp. Cell Res. 46, 19-28. Rabinovitch, M. (1967b). Proc. SOC. Exp. Biol. Med. 124, 396-399. Rabinovitch, M., Dedet, J.-P., Ryter, A., Robineaux, R., Topper, G., and Brunel, E. (1982). J. Exp. Med. 155, 415-431. Reid, K. 9 . M., and Porter, R. R. (1981). Annu. Rev. Biochem. 50, 433-464.

PHAGOCYTE-PATHOGENIC MICROBE INTERACTIONS

325

Reijngoud, D. J. (1978). Trends Biochem. Sci. 3, 178-180. Reijngoud, D. J., and Tager, J. M. (1977). Biochim. Biophys. Acta 472, 419-449. Rikihisa, Y ., and Ito, S. (1979). J. Exp. Med. 150, 703-708. Roberts, R. B. (1979). In “Principles and Practice of Infectious Diseases” (G. L. Mandell, R. G. Douglas, and J. E. Bennett, eds.), pp. 1589-1616. Wiley, New York. Root, R. K., and Stossel, T. P. (1974). J. Clin. Invest. 53, 1207-1215. Ryter, A., and de Chastellier, C. (1977). J . Cell Biol. 75, 200-217. Ryter, A., and Hellio, R. (1980). J. CeN Sci. 41, 75-88. Ryter, A., Dedet, J.-P., and Rabinovitch, M. (1983a). Exp. Parasirol. 55, 233-242. Ryter, A., Lepoivre, M., and Petit, I. F. (1983b). In preparation. Saeki, K . , Endo, K., Tasaka, K., and Yamasuki, H. (1974). Jpn. J . Phurmacol. 24, 109-118. Saint-Gullain, M. L., Vray, B., Hoebeke, J., and Leloup, R. (1980). In “Scanning Electron Microscopy,” pp. 205-212. I1 Semin. Inc., AMF O’Hare, Chicago, Illinois. Sandin, R. L., Rogers, A. L., Patterson, R. J., and Benke, E. (1982). Infect. Immun. 35, 79-85. Sandvig, K., and Olsnes, S. (1980). J . Cell Biol. 87, 828-832. Sastry, P. S.,and Hokin, L. E. (1966). J . Biol. Chem. 241, 3354-3361. Schneider, Y. J., Tulkens, P., de Duve, C., and Trouet, A. (1979a). J. Cell Biol. 82, 449-465. Schneider, Y. J., Tulkens, P., de Duve, C., and Trouet, A. (1979b). J . Cell Biol. 82, 466-474. Schwab, J. H. (1975). Bacreriol. Rev. 39, 121-143. Schwarz. H., and Thilo, L. (1983). Eur. 1. Cell Biol. (in press). Schwarzmann, S., and Boring, J. R., 111. (1971). Infect Immun. 3, 762-767. Segal, A. W.. Geisow, M., Garcia, R., Harper, A., and Miller, R. (1981). Nature (London) 290, 406-409. Segerling, M., Hahn, H., and Opperkuch, W. (1982). Immunobiology 162, 39-45. Seglen, P. O., and Ruth, A. (1976). Exp. Cell Res. 100, 276-280. Shaw, D. R., and Griffin, F. M. (1981). Nature (London) 289, 409-41 I . Sherbert, G. V. (1978). In “The Biophysical Characterization of the Cell Surface,” pp. 36-145. Academic Press, New York. Silveblatt, F. J., Dreyer, J., and Schauer, S. (1979). Infect Immun. 24, 218-223. Silverman, D. I., and Wisseman, C. L. (1979). Infect. Immkn. 26, 714-727. Silverstein, S. C., Steinman, R. M., and Cohn, 2. A. (1977). Annu. Rev. Biochem. 46, 669-722. Simmons, S. R.,and Kamovsky, M. L. (1973). J . Exp. Med. 138, 44-63. Simon, G. L., Klempner, M. S., Kasper, D. L., and Gorbach, S. L. (1982). J. Infecr. Dis. 145, 72-77. Smith, H. (1977). Bacreriol. Rev. 41, 475-500. Spitznagel, J. K., and Chi, H. C. (1963). Am. J . Parhol. 43, 697-708. Stahl, P. D., Rodman, J. S., Miller, M., and Schlessinger, P. H. (1978). Proc. Narl. Acad. Sci. U.S.A. 75, 1399-1403. Stanier, R. Y. (1970). Symp. SOC. Gen. Microbiol.. 20th pp. 1-38. Steinman, R. M., Brodie, S. E., and Cohn, Z. A. (1976). J. Cell Biol. 68, 665-687. Steinman, R. M., Mellman, 1. S., Muller, W. A., and Cohn, Z. A. (1983). J . Cell Biol. 96, 1-28. Stendahl, 0.. Tagesson, C., and Edebo, M. (1973). Infect. Irnmun. 8, 36-41. Stendahl, 0. I . , Hartwig, J. H., Brotschi, E. A,, and Stossel, T. P. (1980). J . Cell Biol. 84, 215-224. Stendahl, 0. I., Dahlgren, C., Edebo, M., and Ohman, L. (1981). Monogr. Aflergy 17, 12-27. Stinson, M. W., and van Oss, C. J. (1971). J. Rericuloendothel. SOC. 9, 503-512. Stjernstrom, I . , Magnusson, K. E., Stendahl, 0..and Tagesson, C. (1977). Infect. Immun. 18, 26 1-265. Stockem, W. (1970). Mikroskopie 24, 332-344. Stockem, W. (1973). Z. Zellforsch. 136, 433-446.

326

ANTOINETTE RYTER AND CHANTAL DE CHASTELLIER

Stockem, W. (1977). In “Mammalian Cell Membranes” (G. A. Jamieson and D. M. Robinson, eds.), Vol 5, pp. 151-195. Butterworths, London. Storrie, B. A. (1979). J . Cell Biol. 83, LP1312. Stome, B. A., Dreesen, T. D., and Maurey, K. M. (1981). Mol. Cell. Biol. 1, 261-268. %OR, I . , and Spears, P. (1977). Curr. Top. Microbiol. Immunol. 76, 167-214. Stossel, T. P. (1975). Semin. Hemarol. 12, 83-116. Stossel, T. P. (1976). J. Rericuloendorhel. SOC. 19, 237-245. Stossel, T. P., and Hartwig, J. H. (1976). J . Cell Biol. 68, 602-619. Sung, S. S . I.. Nelson, R. S . , and Silverstein, S . C. (1983). J. Cell Biol. W, 160-166. Taylor, D. L., and Condeelis, I. (1979). Inr. Rev. Cyrol. 56, 57-142. Taylor, D. L., Blinks, J. R., and Reynolds, G. (1980). J. Cell Biol. 86, 599-607. Thilo, L., and Vogel, G. (1980). Proc. Natl. Acad. Sci. U.S.A. 77, 1015-1019. Thomas, D. W., Hill, J. C., and Tyeryar, F. J. (1972). Infect. Immun. 8, 98-104. Thong, Y. H. (1982). Med. Hypotheses 8, 249-254. Thorne, K. I. I., Svvennsen, R. J., and Franks, D. (1978). Infecr. Immun. 21, 798-805. Thorne, K. J. I., Glauert, A. M., Svvennsen, R. J., and Franks, D. (1979). Parasitology 79, 367-379. Thome, K. J. I., Oliver, R. C., and Glauert, A. M. (1980). J. Cell Sci. 42, 367-378. Thyberg, J. (1980). Eur. J. Cell Biol. 23, 95-103. Thyberg, J., Nilsson, J., and Hellgren, D. (1980). Eur. J. Cell Biol. 23, 85-94. Todd, W. J., and Storz, J. (1975). Infect. Immun. 12, 638-646. Tsan, M. F., and McIntyre, P. A. (1976). J . Exp. Med. 143, 1308-1316. Tulkens, P., Schneider, Y. J., and Trouet, A. (1980). In “Mononuclear Phagocytes: Functional Aspects” (R. van Furth. ed.), Part I, pp. 613-648. Nijhoff, The Hague. Unkeless, J. C., Fleit, H., and Mellman, I. S . (1981). Adv. Immunol. 31, 247-270. Utsumi, S., Sonoda, S., Imagawa, T., and Kanoh, M. (1978). Biken J . 21, 121- 135. Valerius, N. H., Stendahl, 0.. Hartwig, J. H., and Stossel, T. P. (1981). Cell 24, 195-202. Vilkas, E., and Markavits, J. (1972). C. R. Acad. Sci. Paris 275 C , 913-915. Vogel, G., Thilo, L., Schwarz, H., and Steinhart, R. (1980). J . Cell Biol. 86, 456-465. Volsky, D. J., and Loyter, A. (1978). Biochim. Biophys. Acra 514, 213-224. Warr, G. A. (1980). Biochem. Biophys. Res. Commun. 93, 737-745. Weir, D. M., and Ogmundsdottir, H. (1977). Clin. Exp. Immunol. 30, 323-329. Weisman, R. A., and Korn, E. D. (1967). Biochemistry 6, 485-497. Weiss, J., Franson, R. C., Schmeidler, K., and Elsbach, P. (1976). Biochim. Biophys. Acra 436, 154- 169. Weiss, J., Elsbach, P., Olsson, I . , and Odeberg, H. (1978). J . Biol. Chem. 253, 2664-2672. Weiss, J., Victor, M.. Stendahl, O., and Elsbach, P. (1982). J. Clin. Invest. 69, 959-970. Weiss, L. (1969). Inr. Rev. Cyrol. 26, 63-105. Weiss, S. I., and Lobuglio, A. T. (1982). Lab. Invest. 47, 5-18. Weissmann, G., Keiser, H., and Bernheimer, A. W. (1963). J . Exp. Med. 118, 205-222. Werb, Z., and Cohn, Z. A. (1972). J. Biol. Chem. 247, 2439-2446. Westwood, F. R., and Longstaff. E. (1976). Br. J . Cancer 33, 392-399. Wilkinson, B. J., Peterson, P. K.,and Quie, P. G.(1979). Infect. Immun. 23, 502-508. Wilschert, J., Diizgiines, N., and Papahadjopoulos, D. (1981). Biochemistry 20, 3126-3133. Wilson, C. B., van Tsai, V., and Remington, I. S. (1980). J . Exp. Med. 151, 328-334. Win, K., Veale, D. R., and Smith, H. (1976). J. Med. Microbiol. 9, 1-12. Wyrick, P. B., and Brownridge, E. A. (1978). Infecr. Immun. 19, 1054-1060. Yong, E. C., Klebanoff, S. J., and Kuo, C. C. (1982). Infecr. Immun. 37, 422-426. Young, I. D. E., Young, T. M., Lu, L. P., Unkeless, J. C., and Cohn,Z. A. (1982). J . Exp. Med. 156, 1677-1690.

PHAGOCYTE-PATHOGENIC MICROBE INTERACTIONS Young, M . R . , d'Arcy Hart, P., and Geisow, M. J . (1981). Exp. Cell Res. 135, 442-445. Zeya, H. I . , and Spitznagel, J . K . (1966). J . Bacterial. 91, 755-762. Zeya, H. I . , and Spitznagel, J. K . (1968). J . Exp. Med. 127, 927-941. Zucker-Franklin, D., and Hirsch, J . G. (1964). J . Exp. Med. 120, 569-576.

327

This Page Intentionally Left Blank

Index

A Adenosine triphosphatases, intracellular redox state and, 52-53 Adhesion, phagocyte-microbe interaction and, 288-293 Agonist-receptor interaction, intracellular redox state and, 45 Amyloplasts, as geoperceptive organelles, 208-209 Aquatic plants, plastids in roots of, 193 Azolla pinnata. cell files, ultrastructure, size and numbers of plastids in, 188-192 C

Cell(s), control of turgor, 266-267 Cell hybrids, use for gene mapping, 127-130 Cell walls, water relations and, 258-262 Cholecystokinin autoradiographs of 1251-labeledCCK, 16- 19 background, 4 characterization of CCK receptors, 10-16 effects on acinar cells, 4-10 interaction with insulin, 29-34 Chromosomal assignment, gene mapping and, 121-122 Chromosomes human discovery of preferential loss in rodenthuman cell hybrids, 116-1 17 identification of those retained in cell hybrids, 118-120 small segments, fine structure mapping of, 133- 134 Cloned genes, use for gene mapping, 127-130

Convolvulusarvensis. ultrastructural changes in greening roots, 200-201 Cytoplasm and cytoplasmic membranes, water relations and, 262-263

0 Daucus carora. ultrastructural changes in greening roots, 20 1-202 Dehydration postponement of injury by, 278-279 tolerance of, 280-281 Deoxyribonucleic acid, probes, tubulin specific, 148, 156 Deoxyribonucleic acid segments isolated from specific human chromosomes, cloning and mapping of, 131-132 random, assignment to specific human chromosomes, 130-131 Deoxyribonucleic acid synthesis chromatin structure and changes during cell cycle ADP-ribosylation of histones, 91-92 general, 86 histone acetylation, 88-90 histone phosphorylation, 90-9 I methylation of histones, 92-93 nonhistone chromosomal proteins and their phosphorylation, 93-94 nucleoskeletal matrix, 86-88 replication and nucleosome formation, 94-96 factors influencing initiation, 72-78 later stages of cell cycle: CAMP, 78-79

329

330

INDEX

polyamine synthesis and, 82-86 protein synthesis and initiation of early stimulation, 79-80 levels regulating rate of protein synthesis, 80-82 system of second messengers or intracellular regulators, 96-98

E Embryos, radicles, plastids in, 187-188 Endocytosis, membrane recycling and, 3 14-3 17

G Gene map, complete human, prospects of constructing, 135-138 Gene mapping historical view, 110-1 12 use of somatic cell genetics in chromosomal assignment, 121- 122 discovery of preferential loss of human chromosomes in rodent-human cell hybrids, 116-117 establishment of permanent cell cultures of rodent-human cell hybrids, 117-1 18 identification of specific human chromosomes retained in cell hybrids, 118-120 in other species, 138-139 regional assignment, 123- 126 synteny analysis, 120-121 Genetics somatic cell, development of, 112-116 Glycolipids, chain elongation, Golgi complex and, 244-245 Golgi complex confined function model of, 222-23 I consequences of processing, 246-248 covalent modifications and glycolipid chain elongation, 244-245 initiation and elongation of xylose-linked oligosaccharides, 232-235 initiation and processing of N-acetylgalactosamine-linked oligosaccharides, 242-244

proteolysis, 231-232 tailoring of asparagine-linked oli1:osaccharides, 235-242 noncovalent modifications and, 2458-246

H Hepatoma cells, cell cycle of kinetics of, 65-68 regulation by serum, 70-72 variations in G I phase duration, 68--70 variations in G2 phase duration, 70 Histones acetylation of, 88-90 ADP-ribosylation of, 91-92 methylation of, 92-93 phosphorylation of, 90-91 Hybridization, in situ. gene mapping and, 134- 135

I Ingestion, phagocyte-microbe interaction and, 293-296 Insulin autoradiographs of 1251-labeledinsulin, 26-29 background, 19-21 characteristics of insulin receptors, 24-26 effects on acinar cells, 21-24 interaction with CCK, 29-34 Isotypes, tubulin, 156-157

L Lens culinaris, ultrastructural and physiological changes in greening roots, 196-200 Lysosome, fusion with phagosome, 301-3 14

M Matric water, definition of, 257-258 Membrane, recycling during endocytosis, 314-317 Microbicidal activity, phagocyte-microbe interaction. 296-297

INDEX cationic proteins, 300-301 oxidative killing, 297-300

N Nucleoskeletal matrix, DNA synthesis and, 94-96 Nucleosome, formation, replication and, 94-96

0 Oligosaccharides initiation and elongation of xylose-linked, Golgi complex and, 232-235 initiation and processing of N-acetylgalactosamine-linked Golgi complex and, 242-244 tailoring of asparagine-linked, Golgi complex and, 235-242 Oxidative killing, microbicidal activity and, 297-300

P Pancreatic acini, isolation for study of exocrine function, 2-4 PH intracellular changes, stimulus-response events and, 55-56 effects of stimulus-response phenomena on pH, 56-57 relationship with redox state, 41-42 Phagocyte-pathogenic microbe interactions adhesion, 288-293 ingestion, 293-296 membrane recycling during endocytosis, 314-317 microbicidal activity, 296-297 cationic proteins, 300-301 oxidative killing, 297-300 phagosome-lysosome fusion, 301 -3 I4 Phagosome-lysosome fusion, phagocyte-microbe interaction and, 301-314 Phaseolus vulgaris. roots, forms and distribution of plastids in, 180-187

33 1

Plants measurement of water systems, 265-266 roles of water in, 254 water movement in changes in root resistance, 270-271 control of cell turgor, 266-267 driving forces, 266 mathematical models, 270 nonosmotic water movement, 268-269 outside the xylem, 269-270 in whole plants, 267-268 Plastid(s), in sieve elements, 206-208 Plastid division, greening and, 202-203 plastid genome, 205-206 plastid numbers and sites of plastid division, 204-205 ultrastructure of dividing plastids, 203-204 Plastid pigments, responses to light and blue light and greening, 213-214 protochlorophyllide, chlorophylls and carotenoids, 212-213 Plastids in root(s) general features plastid development, 178- 180 roots, 176-178 green roots Azolla pinnata-ultrastructure size and numbers of plastids in different cell files, 188-192 other aquatic species, 193 plastid dedifferentiation and rhizophores of Selaginella martensii. 193- 196 nongreen roots forms and distribution of plastids in seedlings with particular reference to Phaseolus vulgaris. 180- I87 plastids in radicles of embryos, 187-188 some nonphotosynthetic functions of, 2 14-2 I6 Plastids in root caps, geotropism and amyloplasts as geoperceptive organelles, 208-209 plastid distribution in cells of root cap, 209-2 I 1 plastids in root caps of some lower vascular plants, 21 1-212 Polyamine synthesis, DNA synthesis and, 82-86 Protein(s) cationic, microbicidal activity, 300-301

332

INDEX

nonhistone chromosomal, phosphorylation of, 93-94 tubulin, sequence analysis of, 148, 150-155 Protein synthesis, initiation of DNA synthesis and early stimulation of, 79-80 levels regulating rate of protein synthesis, 80-82 Proteolysis, Golgi complex and, 23 1-232

ultrastructural changes in Convolvulus arvensis, 200-20 I in Daucus carora, 201-202 in Triticum vulgare. Secale cereale, a hybrid Triticale and Lens culinaris, 196-200

S

Secale cereale, ultrastructural and physiological changes in greening roots, 196-200 Redox state Selaginella martensii, rhizophores, plastid difintracellular, agonist action and ferentiation and, 193-196 agonist-receptor interaction, 45 Sieve elements, plastids in, 206-208 approaches and definitions, 42-43 Somatic cell genetics ATPases and recovery, 52-53 combined use of recombinant DNA technolconclusions and speculations, 53-55 ogy in gene mapping, 126-127 general interactions, 43-45 assignment of random DNA segments to response and, 52 specific human chromosomes, transducer mechanism, effector system 130- 131 and second messenger formation or cloning and mapping of DNA segments isolated from specific human chrorelease, 45-52 stimulus-response coupling and, 39-41 mosomes, 131-132 Recombinant DNA technology fine structure mapping of small chromosouse in somatic cell genetics in gene mapmal segments, 133-134 ping, 126- I27 in situ hybridization and, 134-135 assignment of random DNA segments to use of cloned genes and cell hybrids, specific human chromosomes, 127-130 130- 131 use of repetitive sequences in fine struccloning and mapping of DNA segments ture mapping of individual chromosomes, 132-133 isolated from specific human chrodevelopment of, 112-116 mosomes, 13I- 132 fine structure mapping of small chromosouse in gene mapping mal segments, 133-134 chromosomal assignment, 121-122 discovery of preferential loss of human in situ hybridization and, 134- I35 use of cloned genes and cell hybrids, chromosomes in rodent-human cell 127-130 hybrids, 116-1 17 use of repetitive sequences in fine strucestablishment of permanent cell cultures of rodent-human cell hybrids, ture mapping of individual chromosomes, 132-133 117-118 Repetitive sequences, use in fine structure identification of specific human chromomapping of individual chromosomes, somes retained in cell hybrids, 132-133 118-120 Rodent-human cell, establishment of permaregional assignment, 123- 126 nent hybrid cell cultures, 117-118 synteny analysis, 120- 121 Roots Stimulus-response coupling, redox state and, changes in resistance, 270-271 39-41 greening of Symplast concept, water relations and, 264 general observations, 196 Synteny analysis, gene mapping and, 120- 12 I

R

333

INDEX

T

postponement of injury by dehydration, 278-279 Transducer mechanism, intracellular redox systolerance of dehydration, 280-28 1 tem and, 45-52 water storage, 280 Triricule. hybrid, ultrastructural and physinjury from deficits, 271-272 iological changes in greening roots, metabolic processes, 276-277 196-200 turgor-related processes. 272-276 Triricurn vulgure. ultrastructural and physmovement in plants iological changes in greening roots, changes in root resistance, 270-271 196-200 control of cell turgor, 266-267 Tubulin genes driving forces, 266 expression of, 167-171 mathematical models, 270 functional and nonfunctional, 157 nonosmotic water movement, 268-269 dissection of multigene families, 165- 167 outside the xylem, 269-270 structure of expressed genes, 158-162 in whole plants, 267-268 tubulin pseudogenes, 162-165 roles in plants, 254 genetic complexity and functional diversity, Water potential, definition of, 255-256 171 - 172 Water relations number and complexity cell structure and sequence analysis of tubulin proteins, cell walls, 258-262 148, 150-155 cytoplasm and cytoplasmic membranes, tubulin isotypes, 156- 157 262-263 tubulin-specific cDNA probes, 148, 156 distribution of water in cells, 264-265 Turgor pressure, definition of, 257 measurement of plant water status, 265-266 symplast concept, 264 v vacuoles, 263-264 in plants, problems and, 254-255 Vacuoles, water relations and, 263-264 terminology Vascular plants, lower, plastids in root caps matric water, 257-258 of. 21 1-212 turgor pressure, 257 water potential, 255-256 W Water adaptations increasing tolerance to deficits, 277-278 metabolic. 279-280

X Xylem, water movement outside, 269-270

This Page Intentionally Left Blank

Contents of Recent Volumes and Supplements Volume 70

Volume 72

*

Microtubule-Membrane Interactions in Cilia and Fkigella-wILLIAM L. DENTLER The Chloroplast Endoplasmic Reticulum: Structure, Function, and Evolutionary SignifiCanCe-sARAH P. GlBBS DNA Repair-A. R. LEHMANNA N D P. KARRAN Insulin Binding and Glucose Transport-RusSELL HILF, LAURIE K . SORGE, A N D ROGER J. GAY Cell Interactions and the Control of Development in Myxobacteria POpUlatiOnS-DAVlD WHITE Ultrastructure, Chemistry, and Function of the Bacterial Wall-T. J. BEVERIEGE

Cycling Noncycling Cell Transitions in Tissue Aging, Immunological Surveillance, Transformation, and Tumor GrowthSEYMOUR GELFANT The Differentiated State of Normal and Malignant Cells or How to Define a “Normal” Cell in Culture-MINA J. BISSELL On the Nature of Oncogenic Transformation of C e l l d E R A L D L. CHAN Morphological and Biochemical Aspects of Adhesiveness and Dissociation of Cancer C e l l s HIDEOHAYASHI AND YASUJI ISHIMARU The Cells of the Gastric MUCOSa-HERBERT F. HELANDER Ultrastructure and Biology of Female Gametophyte in Flowering Plants-R. N. KAPILA N D A. K. BHATNAGAR

INDEX

INDEX

Volume 73 Volume 71 Protoplasts Of Eukaryotic Algae-MARTHA D. BERLINER Integration of Oncogenic Viruses in Mammalian Polytene Chromosomes of Plants-WALTER CellS
PRETLOW INDEX

SONNEBORN INDEX

335

336

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS

Volume 74

Organization and Expression of Viral Genes in Adenovirus-Transformed Cells-S. J FLINT The Plasma Membrane as a Regulatory Site in Highly Repeated Sequences in Mammalian GeGrowth and Differentiation of Neunomes-MAXINE F. SINGER roblastoma CellS-SiEGFRiED W. DE LAAT Moderately Repetitive DNA in EvolutionA N D PAULT. V A N DER SAAC ROBERTA. BOUCHARD Mechanisms That Regulate the Structural and Structural Attributes of Membranous Organelles Functional Architecture of Cell Surfacesin BaCteriaAHARLES c . REMSEN JANETM. OLIVERAND RICHARD D. BERLIN Separated Anterior Pituitary Cells and Their ReGenome Activity and Gene Expression in Avian sponse to Hypophysiotropic H o r m o n e c Erythroid CellS-KARLEN G . GASARYAN CARLDENEF,Luc SWENNEN. A N D MARIA Morphological and Cytological Aspects of Algal ANDRIES Calcification-MICHAEL A. BOROWITZKAWhat Is the Role of Naturally Produced Electric Naturally Occurring Neuron Death and Its RegCurrent in Vertebrate Regeneration and Healulation by Developing Neural Pathwaysing?-RICHARD B. BORGENS TIMOTHY J . CUNNINGHAM Metabolism of Ethylene by Plants--JoHN The Brown Fat Cell-JAN NEDERGAARD AND A. HALL DODDSAND MICHAEL OLOVLINDBERG ININ:X INDEX

Volume 77 Volume 75 Mitochondria1 Nuclei-TsuNwosHi KUROIWA Slime Mold Lectins-JAMEs R. BARTLES, WILLIAMA. FRAZIER,AND STEVEND. ROSEN Lectin-Resistant Cell Surface Variants of Eukaryotic Ceils-EvE BARAKBRILES Cell Division: Key to Cellular Morphogenesis in the Fission Yeast, SchizosaccharomvcesBYRONF. JOHNSON, CODE B. CALLWA, BONGY.Y o o , MICHAEL ZUKER,A N D IAN J. MCDONALD Microinjection of Fluorescently Labeled Proteins into Living Cells, with Emphasis on Cytoskeletal Pl'Otein~THOMAS E. KREIS AND WALTERBIRCHMEIER Evolutionary Aspects of Cell DifferentiationR. A. FLICKINGER Structure and Function of Postovulatory Follicles (Corpora Lutea) in the Ovaries of Nonmammalian Vertebrates-SRiNivAs K. SAI-

Calcium-Binding Proteins and the Molecular Basis of Calcium Action-LINDA J . VANELDIK. JOSEPH G. ZENEDEGUI, DANIELR. MARSHAK, AND D. MARTINWATIXRSON Genetic Predisposition to Cancer in Man: In Virro Studies-LEVY KOPELOVICH Membrane Flow via the Golgi Apparatus of Higher Plant CellS-DAVlD G . ROBINSON A N D U r n KRISTEN Cell Membranes in Sponges-WERNER E. G . MULLER Plant Movements in the Space EnvironrnentDAVIDG . HEATHCOTE Chloroplasts and Chloroplast DNA of Acerabularia medirerranea: Facts and HypothesesANGELALUITKE AND SILVANO BONO-ITO Structure and Cytochemistry of the Chemical SynapSeS-STEPHEN MANALOV A N D WLADI M I R OVTSCHAROFF INDEX

DAPUR INDEX

Volume 76 Cytological Hybridization to Mammalian Chros. HENDERSON mOSOme-ANN

Volume 78 Bioenergetics and Kinetics of Microtubule and Actin Filament Assembly-DissassemblyTERRELL L. HILLAND MARCW. KIRSCHNER Regulation of the Cell Cycle by Somatomedins-How ARD ROTHSTEIN

CONTENTS O F RECENT VOLUMES AND SUPPLEMENTS Epidermal Growth Factor: Mechanisms of ActiOfl-MANJUSRI DAS Recent Progress in the Structure, Origin, Composition, and Function of Cortical Granules in Animal Egg-SARDUL s. GURAYA

337

Biological Interactions Taking Place at a LiquidSolid hterfaC+ALEXANDRE ROTHEN INDEX

INDEX

Volume 81 Volume 79

Oxidation of Carbon Monoxide by BacteriaYOUNGM. KIMAND GEORGED. HECEMAN The Formation, Structure, and Composition of Sensory Transduction in Bacterial Chemotaxthe Mammalian Kinetochore and Kii S q E R A L D L. HAZELBAUER A N D SHlGEAKl netochore Fiber-coNLY L. RIEDER HARAYAMA Motility during FertiliZatiOn-(kRALD SCHAT- The Functional Significance of Leader and TrailTEN er Sequences in Eukaryotic mRNAs-F. E. Functional Organization in the NucleusBARALLE RONALDHANCOCKAND TENI BOULIKAS The Fragile x ChromOSOme~RANT R. The Relation of Programmed Cell Death to DeSUTHERLAND velopment and Reproduction: Comparative Psoriasis versus Cancer: Adaptive versus Studies and an Attempt at ClassificationIatrogenic Human Proliferative DiseasesJACQUESBEAULATONAND RICHARDA. SEYMOUR GELFANT h K S H l N Cell Junctions in the Seminiferous Tubule and Cryofixation: A Tool in Biological Ultrastructhe Excurrent Duct of the Testes: Freezet U d Research-HELMUT PLATTNER AND Fracture Studies-TosHio NAGANO A N D Luis BACHMANN FUME SUZUKI Stress Protein Formation: Gene Expression and Geometrical Models for Cells in Tissues-HisEnvironmental Interaction with Evolutionary AO HONDA Significance-C. ADAMSAND R. W. RINNE Growth of Cultured Cells Using Collagen as INDEX Substrate-JASON YANG A N D s. NANDI INDEX

Volume 80 DNA Replication Fork Movement Rates in Mammalian Cells-LEON N. KAPP A N D ROBERTB. PAINTER Interaction of Virsues with Cell Surface Receptors-MARC TARDIEU,ROCHELLEL. EPSTEIN, AND HOWARD L. WEINER The Molecular Basis of Crown Gall lnductionW. P. ROBERTS The Molecular Cytology of Wheat-Rye Hybrids-R. APPELS Bioenergetic and Ultrastructural Changes Associated with Chloroplast Development-A. R. WELLBURN The Biosynthesis of Microbodies (Peroxisomes, GlyoxysomestH. KINDL Immunofluorescence Studies on Plant Cel1s-C. E. JEFFREE,M. M. YEOMAN,A N D D. C. KILPATRICK

Volume 82 The Exon:lntron Structure of Some Mitochondrial Genes and Its Relation to Mitochondria1 Evohtion-HENRY R. MAHLER Marine Food-Borne Dinoflagellate ToxinsDANIELG. BADEN Ultrastructure of the Dinoflagellate AmphieSm+LENtTA c . MORRILLA N D ALFREDR. LOEBLICH111 The Structure and Function of Annulate Lamellae: Porous Cytoplasmic and Intrsnuclear Membranes-RICHARD G . KESSEL Morphological Diversity among Members of the Gastrointestinal Microflora-DWAYNE C. SAVAGE INDEX

338

CONTENTS O F RECENT VOLUMES AND SUPPLEMENTS

Volume 83 Transposable Elements in Yeast-VALERIE MoROZ WILLIAMSON Techniques to Study Metabolic Changes at the Cellular and organ kVel-ROBERT R. DEFURIAAND MARYK. DYGERT Mitochondria1 Form and Function Relationhips in Vivo: Their Potential in Toxicology and Pathology-ROBERT A. SMITH A N D MURIEL 1. ORD Heterogeneity and Territorial Organization of the Nuclear Matrix and Related StructuresM. BOUTEILLE,D. BOUVIER,AND A. P. SEVE Changes in Membrane Properties Associated with Celhlar Aging-A. MACIEIRA-COELHO Retinal Pigment Epithelium: Proliferation and Differentiation during Development and Regeneration-OLcA G. STROEVAAND VICTOR 1. MITASHOV

Supplement 1 0 Differentiated Cells in Aging Research

Do Diploid Fibroblasts in Culture Age'?-EuGENE BELL, LOUIS MAREK. STEPHANIL SHER, CHARLOTTE MERRILL,DONALD A N D IAN YOUNG LEVINSTONE, Urinary Track Epithelial Cells Cultured from S. FELIX A N D J. W. Human Urine-J. LIlTLEFlELD The Role of Terminal Differentiation in the Finite Culture Lifetime of the Human Epidermal Keratinocyte-JAMES G. RHEINWALD Long-Term Lymphoid Cell CuItures-GEoRcx F. SMITH, PARVIN JUSTICE, HENRI LEE KIN CHU, A N D JAMES KRW FRISCHER, Type I1 Alveolar Pneumonocytes in \'ifr+ WII.I.IAMH. J . DOUGLAS. JAMES A . MCATEER.JAMES R. SMITH. A N D WALTER R. BRAUNSCHWEIGER Cultured Vascular Endothelial Cells as a Model System for the Study of Cellular SenesINDEX cence-ELLIOT M. LEVINEA N D STEPHEN M. MUELLER Vascular Smooth Muscle Cells for Studies of Volume 84 Cellular Aging in V i m : an Examination of Controls to Plastid Divsion-J. V. POSSINGHAM Changes in Structural Cell Lipids-oLcA 0. BLUMENFELD. ELAINESCHWARTZ.VERAND M. E. LAWRENCE ONICA M. HEARN. A N D M A R I E J . Morphology of Transcription at Cellular and KRANEWOL Molecular kVeIs-FRANClNE PUVIONChondrocytes in Aging Research-EDwMw I . DUTILLEUL MILLER A N D STEFFAN GAY An Assessment of the Evidence for the Role of Ribonucleoprotein Particles in the Maturation Growth and Differentiation of Isolated Calvarium Cells in a Serum-Free Mediumof Eukaryote mRNA-J. T. KNOWLER A. PECK JAMESK. BURKSA N D WILLIAM Degradative Plasmids-J. M. PEMBERTON Regulation of Microtubule and Actin Filament Studies of Aging in Cultured Nervous System H. SILBERBERGA N D TiSSue-DONALD Assembly-Disassembly by Associated Small SEUNGU. K I M and Large Molecules-TERRELL L. HILL Aging of Adrenocortical Cells in Culture-PEAND MARCw . KIRSCHNER TER J. HORNSBY, MICHAELH. SIMONIAN, Long-term Effects of Perinatal Exposure to Sex AND GORDONN. GILL Steroids and Diethylstilbestrol on the ReThyroid Cells in CUhre-FRANCiiSCO S . AMproductive System of Male MammalsBESI-IMPIOMBATOAND HAYDENG. COON YASUMASA ARAI. TAKAO MORI,YOSHIHIDE Permanent Teratocarcinoma-Derived Cell Lines SUZUKI,AND HOWARD A. BERN Stabilized by Transformation with SV40 and Cell Surface Receptors: Physical Chemistry and SVmSA Mutant VhSe+WARREN MALTZCellular Regulation-DOUGLAS LAUFFENMAN, DANIEL 1. H. LINZER,FLORENCE BURGER AND CHARLES DELISI BROWN,ANGELIKA K.TERESKY. MAURICE Kinetics of Inhibition of Transport Systems-R. AND ARNOLD J. LEVINE ROSENSTRAUS, M. KRUPKAAND R. DEVBS Nonreplicating Cultures of Frog Gastric Tubular INDEX

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS CellSAERTRUDE H. BLUMENTHALA N D DINKAR K. KASBEKAR

339

Embryo CU1tUre-V. RAGHAVAN The Future-J.koRc MELCHERS

INDEX

SUBJECT INDEX

Supplement IIA: Perspectives in Plant Cell and Tissue Culture

Supplement 12: Membrane Research: Classic Origins and Current Concepts

Cell Proliferation and Growth in Callus Cul- Membrane Events Associated with the GeneraM. YEOMAN A N D E. FORCHE tion Of a BlaStocySt-MARTIN H. JOHNSON tures-M. Cell Proliferation and Growth in Suspension Structural and Functional Evidence of Cooperativity between Membranes and Cell Wall in Cultures-P. J. KING Bacteria-MANFRED E. BAYER Cytodifferentiation-RICHARD PHILLIPS Organogenesis in Virro: Structural, Physiologi- Plant Cell Surface Structure and Recognition cal, and Biochemical A S ~ ~ C ~ S - T R E V A.O R Phenomena with Reference to SymbiosesPATRICIA S. REISERT THORPE Chromosomal Variation in Plant Tissues in Cul- Membranes and Cell Movement: Interactions of ture-M. W. B A Y L l s s Membranes with the Proteins of the Cytoskeleton-JAMES A. WEATHERBEE Clonal Propagation-INmA K. V A s r L A N D V l M L A VASIL Electrophysiology of Cells and Organelles: Control of Morphogenesis by Inherent and ExStudies with Optical Potentiometric Indicaogenously Applied Factors in Thin Cell AND PHILIP c . Iors-JEFPREY c. FREEDMAN Layers-K. TRANTHANH VAN LARIS Androgenetic HaplOidS-INDRA K. VASIL Synthesis and Assembly of Membrane and Isolation. Characterization, and Utilization of Organelle Proteins-HARVEY F. LODISH, Mutant Cell Lines in Higher Plants-PAL WILLIAM A. BRAELL,ALANL. SCHWARTZ. MALIGA GER J . A. M. STROUS, AND ASHER SUBJECT INDEX ZILBERSTEIN The Importance of Adequate Fixation in Preservation of Membrane UltrastructureSupplement 11B: Perspectives in Plant Cell RONALD B. Lumic A N D PAUL N. Mcand Tissue Culture MILLAN Liposomes-As Artificial Organelles, ToIsolation and Culture of Protoplasts-ImRn K. pochemical Matrices, and Therapeutic CarVASIL A N D VIMLAVASIL rier SyStemS--PETER NICHOLLS Protoplast Fusion and Somatic HybridizationDrug and Chemical Effects on Membrane TransOTTO SCHIEDER A N D INDRA K. VASIL POn-WILLIAM 0.BERNDT Genetic Modification of Plant Cells Through INDEX Uptake of Foreign DNA-<. I. KADO AND A. KLEINHOFS Nitrogen Fixation and Plant Tissue CultureSupplement 13: Biology of the Rhizobiaceae KENNETH L. GILESAND INDRA K. VASIL Preservation of Germplasm-LvNDsEv A. The Taxonomy of the Rhizobiaceae-JhWLD WITHERS H. ELKAN Intraovarian and in V i m Pollination-M. Biology of Agrobacrerium rumefaciens: Plant ZENKTELER Interactions-L. W. MOORE AND D. A. Endosperm Culture-B. M. JOHRI, P. S. S R i COOKSEY VASTAVA, A N D A. P. RASTE Agrobacrerium tumefaciens in Agriculture and The Formation of Secondary Metabolites in Research-FAwzi EL-FIKIAND KENNETHL. Plant Tissue and Cell Cultures-H. BOHM GILES

340

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS

Suppression of, and Recovery from, the Neo- Metabolic Interchange in Algae-Invertebrate plastic State-ROBERT TURGEON SymbiOSidLAYTON B. COOK Plasmid Studies in Crown Gall Tumorigenesis- The Molecular Biology of Rhizobium-Legume STEPHENL. DELLAPORTAAND RICK L. Symbiosis-DEsH PAL s. VERMA AND PESANO SHARON LONG The Position of Agrobacterium rhizogenesAnalysis of Possible Gene Transfer between an JESSE M. JAYNESAND GARYA. STROBEL Insect Host and Its Bacteria-like EndocytoRecognition in Rhizobium-Legume Symbionts-W. SCHWEMMLER biOSeS-TERRENCE L. GRAHAM Cellular and Molecular Mechanisms of Intracellular Symbiosis in Leishrnaniasis-K.-P. The Rhizobium Bacteroid State-W. D. SUTTON, C. E. PANKHURST. AND A. S. CRAIG CHANG Exchange of Metabolites and Energy between Symbiotic Interaction between Legionella pneuLegume and Rhizobium-JoHN IMSANDE mophila and Human Leukocytes-MARCUS The Genetics of Rhizobium-ADAM KONA. HORWITZ DOROSI AND ANDREW W. B. JOHNSTON Plastids-Past, Present, and Future-hN M. Indigenous Plasmids of Rhizobium-J. DEENWHATLEY ARIEE. P. BOISTARD,FRANCINE CASSE-INDEX DELBART,A. G. ATHERLY,J. 0. BERRY, A N D P. RUSSELL Nodules Morphogenesis and DifferentiationSupplement 15: Aspects of Cell Regulation WILLIAM NEWCOMB Mutants of Rhizobium That Are Altered in Cellular Factors Which Modulate Hormone Responses: Glucocorticoid Action in PerspecLegume Interaction and Nitrogen FixationL. D. KUYKENDALL tiVe-ROBERT w . HARRISON, 111 The Significance and Application of Rhizobium Regulation of Genetic Activity by Thyroid Hormones-A. ABDUKARIMOV in AgriCUltUre-HAROLD L. PETERSONAND The Partitioning of Cytoplasmic Organelles at THOMASE. LOYNACHAN Cell Division<. WILLIAM BIRKY.IR. INDEX Cell Cycle Mutants-WILLIAM L. WISSINGER AND RICHARD J . WANG Formation of Glyoxysomes-J. MICHAEL LORD Supplement 1 4 lntracellular Symbiosis AND LYNNE M. ROBERTS Some Eco-evolutionary Aspects of lntracellular Mitochondria, Cell Surface, and Carcinogenesis-D. WILKIE.l. H. EVAVS,V. Symbiosis-F. 1. R. TAYLOR EGILSSON,E. S. DIALA,AND D. COLLIER Integration of Bacterial Endosymbionts in Transforming Genes of Tumor Cek-RoBERT Amoebae-Kwmc W. JEON A. WEINBERG Perspective on Algal Endosymbionts in Larger Viral Carcinogenesis-FRED RAFT Foraminifera-JOHN 1. LEE The Biology of the Xenosome. an lntracellular The Origin of Viruses from Cells-R. E. F. MATTHEWS Symbiont-A. T. SOLW HECK- INDEX Endosymbionts of Euplores-KLAus MANN

Endonuclear Symbionts in Ciliates-HANs-DIETER

GORTZ