No title

INTERNATIONAL REVIEW OF CYTOLOGY VOLUME 80 ADVISORY EDITORS H. W. BEAMS HOWARD A. BERN GARY G. BORISY PIET BORST BHA...

Author: Author Unknown

3 downloads 229 Views 17MB 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 VOLUME 80

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

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 DENNIS L. TAYLOR TADASHI UTAKOJI ROY WIDDUS ALEXANDER L. YUDIN

'

INTERNATIONAL

Review of Cytology EDITED BY

G. H. BOURNE

J. F. DANIELLI

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

Worcester Polytechnic Institute Worcester, Massachusetts

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

VOLUME 80 1982

ACADEMIC PRESS

A Subsidiary of Harcourt Brace Jovanovich, Publishers

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

'/

COPYRIGHT @ 1982, BY ACADEMIC PRESS, INC. ALL RIGHTS RESERVED. NO PART OF THIS PUBLICATION MAY BE REPRODUCED O R 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 b y ACADEMIC PRESS, INC. ( L O N D O N ) LTD. 24/28 Oval Road, London NW1 7DX

LIBRARY OF CONGRESS CATALOG CARDNUMBER:5 2 - 5 2 0 3 ISBN 0-12-364480-0 PRINTED IN THE UNITED STATES OF AMERICA 82 83 84 85

9 8 76 5 4 3 2 1

Contents CONTRIBUTORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

ix

DNA Replication Fork Movement Rates in Mammalian Cells LEONN . KAPPA N D ROBERT B . PAINTER I. I1 . Ill . IV.

Introduction . . . . . . . . . . . . . . . Methodologies . . . . . . . . . . . . . . DNA Replication Fork Movement Rates . Concluding Remarks . . . . . . . . . . . References . . . . . . . . . . . . . . . .

. . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

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

1

2 9 22 23

Interaction of Viruses with Cell Surface Receptors MARCTARDIEU. ROCHELLE L . EPSTEIN.A N D HOWARD L . WEINER I . Definition of Viral Receptor Sites . . . . . . . . . . . . . . . . . . .

I1 . Biological Characteristics of Viral Attachment to Cells 111. Membrane Components Which Interact with Viruses .

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

IV. Viral Components Which Recognize Cellular Receptors . . . . . . . . . V. Virus-Receptor Interactions and Pathogenicity . . . . . . . . . . . . . VI . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

27 30 37 49

52 56 57

The Molecular Basis of Crown Gall Induction W. F? ROBERTS I . Introduction

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

63 64 65

I1 . The Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . .

111. IV. V. VI . VII . VIII . IX .

X. XI .

The Crown Gall Bacteria . . . . . . . . . . . . . . . . . . . . . . The Physiology of the Gall . . . . . . . . . . . . . . . . . . . . . . Involvement of Wounding in Gall Induction . . . . . . . . . . . . . The Opines . . . . . . . . . . . . . . . . . . . . . . . . . . . . Involvement of Plasmids in Tumor Induction . . . . . . . . . . . . Significance of Crown Gall Induction to Agrobucteriurn . . . . . . . . The Evolutionary Origin of Crown Gall . . . . . . . . . . . . . . . Agrobacteriurn and Genetic Engineering . . . . . . . . . . . . . . Future Work and Prospects . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . V

65

. .

. .

.

66 68 70 83 84 86 87 88

vi

CONTENTS

The Molecular Cytology of Wheat-Rye Hybrids R . APPELS

I. I1. I11. IV. V. VI . VII . VIII .

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Genetic Relationship between Rye and Wheat Chromosomes . . . . The Molecular Structure of Rye and Wheat Chromosomes . . . . . . . Translocations in Wheat-Rye Addition or Substitution Lines . . . . . . Polymorphisms in Regions of the Chromosomes Containing Repeated Sequence DNA . . . . . . . . . . . . . . . . . . . . . . The Biological Effects of Rye Chromosomes (or Rye Chromosome Fragments) in Wheat-Rye Hybrids: Specific Effects Related to Heterochromatin . . . . . . . . . . . . . . . . . . . . . . . . . . The Possible Origins of Polymorphism in Rye Heterochromatin . . . . . Prospects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

93 94 99 109 113

121 123 127 127

Bioenergetic and Ultrastructural Changes Associated with Chloroplast Development A . R . WELLBURN I. I1 . 111. IV. V. VI . VII. VIII .

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plastid Development in Different Systems . . . . . . . . . . . . . . . Semicrystalline Structures . . . . . . . . . . . . . . . . . . . . . . Storage Reserves and Mobilization during Plastid Development . . . . . Mitochondria and Respiration during Plastid Development . . . . . . . Transfer between Cell Compartments during Photomorphogenesis . . . . Biogenesis of Photochemical Activities . . . . . . . . . . . . . . . . Influence of Light and Hormones . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

134 134 139 145 149 156 169 174 179

The Biosynthesis of Microbodies (Peroxisomes. Glyoxysomes) H . KINDL I. I1. I11. IV V. VI .

.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . General Concepts of Organelle Biosynthesis . . . . . . . . . . . . . . Survey Obtained by in Vivo Studies . . . . . . . . . . . . . . . . . Single Steps of Assembly Studied in Vitro . . . . . . . . . . . . . . . Special Types of Cells . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

193 194 202 210 214 224 224

vii

CONTENTS

Immunofluorescence Studies on Plant Cells C. E . JEFFREE. M . M . YEOMAN.AND D . C . KILPATRICK I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . I1 . Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ill . Applications of Immunofluorescence Microscopy to Studies of Plant Cells and Tissues . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

246 262 263

Biological Interactions Taking Place at a Liquid-Solid Interface ALEXANDRE ROTHEN

I. I1 . I11 . IV.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . Immunologic Reactions . . . . . . . . . . . . . . . . . . . . . . . Experimental . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

INDEX. . . . . . . . . . . . . . . . . . . . . . . CONTENTS OF RECENTVOLUMES A N D SUPPLEMENTS . . .

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

267 269 270 302 303 305 311

This Page Intentionally Left Blank

Contributors

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

R. APPELS(93), Division of Plant Industry, CSIRO, Canberra ACT 2601, Australia ROCHELLE L. EPSTEIN (27), Department of Neuroscience, Children’s Hospital Medical Center, and Department of Medicine, Neurology and Infectious Disease Divisions, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts 02115

C. E . JEFFREE(231), Department of Botany, University of Edinburgh, Edinburgh EH9 3JH, Scotland LEONN. KAPP( I ) , Laboratory of Radiobiology and Environmental Health, University of California, San Francisco, California 94143

D. C . KILPATRICK (2311, Regional Blood Transfusion Service, Royal Infirmary, Edinburgh, Scotland H. KINDL(193), Biochemie (Fachbereich Chemie), Philipps- Universitat, 0-3550 Marburg, Federal Republic of Germany ROBERTB. PAINTER (l), Laboratory of Radiobiology and Environmental Health, University of California, San Francisco, California 94143

W. P. ROBERTS (63), Department of Microbiology, La Trobe University, Bundoora, Victoria 3083, Australia ALEXANDRE ROTHEN(2671, The Rockefeller University, New York, New York 10021 MARCTARDIEU‘ (27), Department of Neuroscience, Children’s Hospital Medical Center, and Department of Medicine, Neurology and Infectious Diseuse Divisions, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts 02115

‘Present address: lnserm U56, HBpital de Bicetre, Bicetre 94270, France. ix

X

CONTRIBUTORS

HOWARD L. WEINER (27), Department of Neuroscience, Children's Hospital Medical Center, and Department of Medicine, Neurology and Infectious Disease Divisions, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts 02115 A. R. WELLBURN (133), Department of Biological Sciences, University of Lancaster, Lancaster, England

M. M. YEOMAN (231), Department of Botany, University of Edinburgh, Edinburgh EH9 3JH, Scotland

INTERNATIONAL REVIEW OF CYTOLOGY. VOL. 80

DNA Replication Fork Movement Rates in Mammalian Cells LEONN. KAPP AND ROBERTB. PAINTER Laboraton, of Radiobiology and Environmental Health, University of California, San Francisco, California

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

......... Rates ...................... A. Fork Rates in Human Cells,. . . . . . . . . . . . . . . . . . . . . . . . . . . . B . Fork Rates in Nonhuman Mammalian Cells . . . . . . C. Fork Rates in Synchronized Cells.. ...................... D. Chemicals Affecting Fork Movement Rates.. . . . . . E. Fork Rates in Mutants and in Virro ...................... IV. Concluding Remarks ........................ References . . . . . . . . . ........................... 11. Methodologies.. 111. DNA Replication

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

1 2 9

9 15 15 18 21

22 23

I. Introduction The DNA replication rate is formed by two components: the number of actively operating replicating units (replicons), and the average linear rate at which the DNA replication fork moves along the parental DNA (DNA fork movement rate). The number of replicons in an average mammalian cell is about 100,000 (Painter et al., 1966; Okada, 1968), and the number of replicons active at any one time during S phase appears to be the main factor affecting changes in the DNA synthesis rate (Painter and Schaefer, 1971). DNA fork movement rates vary considerably from one cell type to another; published values range from 0.1 pdrninute in human cells to 2.5 pdminute in Chinese hamster ovary (CHO) cells. The majority of rates reported for human cells are in the range of 0.4 to 0.7 pdminute. The reasons for the reported variations are not completely known, but some factors are species, ploidy , whether normal or transformed, growth conditions (media, serum, etc.), and the experimental techniques used. Various aspects of DNA replication have been surveyed in recent reviews (Edenberg and Huberman, 1975; Painter, 1976 Sheinin and Humbert, 1978; Hand, 1978, 1979; De Pamphilis and Wassarman, 1980). In this article, we review DNA fork movement rates only in mammalian cells.

1 Copyright 0 1982 by Academic Press. Inc. All rights of reproduction in any form resewed. ISBN 0-12-364480-0

2

LEON N. KAPP AND ROBERT B. PAINTER

11. Methodologies

The first report on fork movement rates utilized DNA fiber autoradiography. Cairns (1966), extending a technique he first used with bacteria (Cairns, 1963), incubated HeLa cells for 45 or 180 minutes with [3H]thymidine. The cells were then lysed and the DNA was spread out and coated with a photographic emulsion. After a sufficiently long exposure time, the emulsion was developed, and the tracks produced by the radioactively labeled DNA were measured. Cairns found that a 45-minute labeling time produced labeled DNA lengths ranging from 10 to 30 pm, whereas the 180-minute labeling time produced lengths ranging from 50 to 100 pm. Thus, human DNA appeared to have replicated at a rate of 0.5 pdminute or less. Subsequently, other workers improved the DNA fiber autoradiography technique to obtain more precise results and more information about DNA replication. Huberman and Riggs (1968) first pretreated cells with fluorodeoxyuridine (FUdR) before labeling with [3H]thymidine. The FUdR depleted the cells of DNA thymine precursors, so when [3H]thymidine was added later it became the sole source of thymine for DNA replication. This led to an abrupt start of the labeled segment of DNA as well as a higher specific activity of thymidine, thus producing heavier grain tracks in the fiber autoradiograms. Another innovation used by Huberman and Riggs was to label cells first with [3H]thymidine with a high specific activity followed by [3H]thymidine with a low specific activity. This resulted in a high grain density track followed by a low grain density track. These variable grain density tracks allowed the findings that (1) replicons are arrayed tandemly (i.e., linearly or sequentially) along the DNA fiber, (2) many replicons on one fiber are of uniform size, and (3) DNA fork movement rates can be measured within each replicon. Huberman and Riggs reported that fork movement rates ranged from 0.5 to 2.5 pdminute in CHO cells. Recently, Yurov (1980) attempted to increase the information obtained by fiber autoradiography by isolating single labeled cells on microscope slides and then lysing the cell in situ. The resulting fiber autoradiograms then represented DNA from only one cell, thus allowing analysis of DNA replication in one nucleus. Examples of fiber autoradiograms are shown in Fig. 1. The advantages of DNA fiber autoradiography in examining DNA fork movement rates are that (1) single DNA chains are resolvable, (2) information can be obtained about replicon sizes, and (3) information can be obtained about relative times of replicon initiation. It should be noted that this technique was also used to show that DNA replication occurs bidirectionally (Huberman and Riggs, 1968). Limitations of this technique are: ( 1 ) Nonrepresentative sampling. Since the average mammalian cell contains approx. lo5 replicons and most autoradiographic data represent 150-300 tracks, only a very small percentage of the replicating DNA is represented. (2) Low resolution. The limit of resolution in

FIG. I . Autoradiograms from mouse L5 l78Y cells. (a) Thirty-minute pulse-labeling. Fibers show tandem arrays of replicons. (b) Labeling with high specific activity [3H]thymidine (hot pulse) for 30 minutes followed by labeling with low specific activity [3H]thymidine (warm pulse) for 30 minutes. The two dense unbroken tracks represent the origin regions of replicons that initiated operation after the beginning of the labeling period (i.e., a postpulse figure), whereas the gap represents the origin of a replicon that initiated operation before the beginning of the labeling period (i.e., a prepulse figure). The dense areas are the result of labeling by high specific activity [3H]thymidine, whereas the lighter trailing tracks represent the DNA labeled by low specific activity [3H]thymidine. The direction of replication was from the high to low specific activity areas. (c) A comet of tangled DNA. This may represent DNA from one cell or a portion of one cell. To measure the fiber lengths, it is necessary to find a region of the slide or comet where the tracks are sufficiently clear and separated. These autoradiographs were provided by S. Sawada, Kumamoto University, Kumamoto, Japan.

4

LEON N. KAPP AND ROBERT B. PAINTER

DNA fiber autoradiographs is reported as 1.5-5 pm (Huberman and Riggs, 1968; Ockey and Saffhill, 1976). For a cell with an average fork movement rate of 0.5 pndminute incubated for 30 minutes, the error of resolution in the resulting DNA fiber autoradiogram would be 10 to 30%. Longer pulse times could reduce this error, but when pulse times become too long, adjacent replicons fuse, making fork movement rates unmeasurable. (3) Coincidence counting problems. All tracks scored may not be from separate DNA fibers. Despite these shortcomings, DNA fiber autoradiography is the most commonly used technique for examining DNA fork movement rates in mammalian cells and, when carefully performed, gives consistent and valuable information. Several ultracentrifugal methods have been used to estimate DNA fork movement rates. These methods utilize the distribution of DNA along a gradient in an ultracentrifuge tube. The gradient can measure differences in molecular weight or differences in density, and these differences are used to calculate DNA fork movement rates. Lehmann and Ormerod (1970) calculated DNA fork movement rates by examining the sedimentation of labeled DNA on alkaline sucrose gradients. They incubated mouse L5 178Y cells with [3H]thymidinefor various lengths of time. The cells were next exposed to X rays to introduce random breaks into the DNA, lysed on the top of a 5-20% alkaline sucrose gradient to release single-stranded DNA, and then centrifuged. From the resulting distributions of the labeled DNA on the gradients, the average molecular weights were computed. By labeling the cells for different lengths of time and observing the change of the average molecular weights in the resulting gradients, the DNA replication fork movement rate was calculated. Similar approaches, utilizing the increase in DNA molecular weights as observed by sucrose gradient centrifugation, have been used by several other workers (Lanotte et al., 1977; Laughlin and Taylor, 1979) to calculate fork movement rates (Fig. 2). Painter and Schaefer (1969, 1971) utilized incorporation of [3H]thymidineand unlabeled bromodeoxyuridine (BUdR) in sequential pulse-labeling and CsCl equilibrium density gradients to measure DNA fork movement rates. With this technique, cells were first incubated with [3H]thymidinefor a short time (10 to 30 minutes), followed by incubation with BUdR for 1 to 2 hours. The DNA was then isolated and sheared, and one aliquot was analyzed by velocity sedimentation to determine the average molecular weight of the sheared fragments. A second aliquot was analyzed on CsCl equilibrium density gradients to determine the fraction of DNA that was distributed on the heavy side of the normal-density DNA, i.e., that which contained molecules that had 3H at one end and BUdR at the other. From this shift, DNA fork movement rates were calculated using published equations (Painter and Schaefer, 1971; Roti-Roti and Painter, 1977) (Fig. 3). More recently, Povirk and Painter (1976) pulse-labeled cells with [3H]BUdR

5

DNA REPLICATION FORK MOVEMENT RATES

0.6

0.3

24

12

S

l

o

b

2

0

Fraction number FIG. 2. (a-e) Alkaline velocity sedimentation profiles of pulse-labeled nascent DNA segments from CHO cells blocked by hydroxyurea at the beginning of S phase until 14 hours after division. The pulse times were (a) 4 minutes, (b) 8 minutes, (c) 12 minutes, (d) 16 minutes, and (e) 20 minutes. Each pulse was terminated by submerging the cultures in SSC (0.15 M sodium chloride, 0.015 M sodium citrate) at 0°C. The cells were lysed at 0°C with standard lysing solution containing 500 p,g/ ml proteinase K. The DNA from these lysates was denatured and then sedimented in alkaline NaI velocity gradients 9 hours after lysis. The gradients were centrifuged at 20,000 rpm for an 02t of 10' I rad2/second. The vertical arrows indicate the position of the center of the T7 marker DNA band. The T7 DNA included in these samples sedimented in each case as expected for a homogeneous population of phage-sized DNA molecules. Sedimentation was from left to right. Reprinted from Laughlin and Taylor (1979) with the permission of the authors and Springer-Verlag, Inc.

for short times, and then exposed them to several fluences of 313 nm light. Under the experimental conditions used, the only breaks induced in the DNA were caused by the action of the 3 13 nm source on the BUdR-substituted residues in the DNA. The cells were then lysed on the top of an alkaline sucrose gradient and centrifuged. The resulting DNA gradient profiles showed that increasing the

6

LEON N. KAPP AND ROBERT B . PAINTER

F R A C T I O N NUMBER

DNA REPLICATION FORK MOVEMENT RATES

7

exposure to 3 13 nm light caused an increased shift of the labeled DNA toward low molecular weights. This shift can be quantified and used to estimate the length of DNA labeled with BUdR. Dividing the length of labeled DNA by incubation time with [3H]BUdR yields the average fork displacement rate (Fig. 4). Other techniques or variations of some of the described techniques have also been used to calculate fork movement rates. For example, Planck and Mueller (1977a) labeled cells briefly with BUdR, and then with I3H]thymidine, and then exposed the labeled cells to 313 nm light. The light ruptured the DNA at the BUdR-labeled regions and released the "-labeled DNA. The [3H]thymidine incubation times were varied and the growth of labeled DNA chains was analyzed by velocity gradient sedimentation. Taylor ( 1968) used variable incubation times with [3H]BUdR and estimated the ratio of 3H in completely substituted DNA to that in partially substituted DNA as a function of incubation time. The rate of fork movement was estimated from the changes in this ratio. Gradient methods have the advantage of speed (most determinations take only 1 to 5 days) when compared to autoradiography, which often requires developing times of up to 1 year. In addition, gradient techniques also yield results that reflect an average of all the replicating DNA in the S phase cells. However, they give no information about distributions of fork movement rates within a cell. Several workers have directly compared various techniques for measuring fork movement rates. Laughlin and Taylor (1979) compared an alkaline velocity gradient method with fiber autoradiography in CHO cells. The alkaline gradient method gave an estimated fork movement rate of 0.5 to 0.6 pdminute. Fiber autoradiographic data from the same cells were in excellent agreement with the sedimentation results, indicating that two independent techniques can both result in the same findings. In another comparison of sedimentation and autoradiographic techniques, Richter and Hand (1979b) measured fork movement rates in monkey CV-1 cells. Fiber autoradiography gave a value of 0.56 pdminute, whereas the equilibrium density method described by Painter and Schaefer (1 971) gave a value of 0.36 to FIG. 3. CsCl equilibrium density gradient profiles of HeLa S3 DNA labeled for 30 minutes with [3H]thymidine and then for 2 hours with BUdR and sheared (upper panel) to produce number-average molecular weight DNA (E) of 1.3 X 107 and fraction of ['HIDNA at densities greater than normal ( F ) of 0.225 or sheared (lower panel) with ultrasound to produce B of 0.26 X lo7 and F of 0.026. A, 3H radioactivity; 0, I4Cradioactivity (adjusted). The ultrasound F value is considered the minimum possible and is primarily due to thymidine pool mixing of [3H]thymidine and BUdR. This is subtracted from the F value for the 12,000 rpm shearing to give F,,,. From B and F,,,, L , the average molecular weight of DNA labeled during the pulse with [3H]thymidine, can be estimated. Since about one-twelfth of the total DNA must be synthesized in 30 minutes (S period = 360 minutes), the total number of sites replicating DNA DNA molecular weight per ceW12 X L. Reprinted from Painter and Schaefer (1969) with permission of the authors and Academic Press, Inc.

8

LEON N. KAPP AND ROBERT B. PAINTER

BOTTOM

TOP

FRACTION

FIG. 4. Alkaline sucrose gradient profile of DNA from pulse-labeled E-11 human diploid cells that were exposed to various fluences of 313 nm light, lysed on a 5-20% alkaline sucrose gradient, and centrifuged. Cells were exposed to 0 (O), 60 (O), or 180 (0) seconds of 313 nm light. Such gradients allow calculations of parameters that can be used to estimate fork movement rates. Reprinted from Kapp and Paintep (1978) with permission of the Biophysical Society.

0.38 pdminute. Richter and Hand concluded that the equilibrium density method provided an objective measurement of fork movement rates, although it underestimated the actual rate. However, if these authors had used the equilibrium density gradient method exactly as described by Painter and Schaefer (1969, 1971) (one necessary step was omitted), the results from the two methods would have been in closer agreement. Kapp and Painter (1979) compared two sedimentation techniques: the equilibrium density method of Painter and Schaefer (1971) and the BUdR-313 nm photolysis method described by Povirk and Painter (1976). It was found that for asynchronous CHO cells, both techniques gave the same average value and range (about 1 pdminute). The same methods were used with synchronous CHO populations and resulted in the same conclusions: a constant rate of about 1 pm/ minute throughout S phase. Rates in synchronous HeLa cells were also measured using the BUdR-313 nm photolysis method and a 2- to 3-fold increase in rate from early to late S was found. This confirmed the earlier report of Painter and Schaefer (1971) for HeLa cells using the equilibrium density method. From this work, it appears that different sedimentation techniques using different approaches can be in excellent agreement. The work discussed above indicates that most of the techniques currently in use give comparable results. Therefore, it appears possible that the reported variations in fork movement rates actually are due to differences in the rates in

DNA REPLICATION FORK MOVEMENT RATES

9

different cell types or in the same cell type under different culture conditions, rather than to inappropriate techniques. To examine this question in more detail, Kapp and Painter (198 1) split a single cell culture into two equal subcultures and measured fork movement in them using the BUdR-313 nm photolysis method. All measurements were made in duplicate, portions of each cell culture being run in separate centrifuge rotors. The idea was to determine, when cell cultures were as identical as possible, how much variation in fork rate measurements was inherent in the experimental techniques. There was less than 10% variation between the most extreme values in this set of measurements. However, under normal experimental conditions, rates measured in any single cell type vary by about 30%. Again, these results suggest that the differences measured with different techniques are largely due to variations in cell types or culture conditions.

111. DNA Replication Fork Movement Rates

DNA fork movement rates have been examined in a wide variety of mammalian and human cells (Tables I and 11). The values shown in the tables are the average values reported in the papers. As mentioned above, large variations were seen around the mean values in most of the reports. Since not all means and variances were shown in the same manner, they are omitted in the tables for the sake of uniform presentation and comparison. In addition, some reports discussing fork movement rates presented comparisons but did not present data on actual fork movement rates, and thus are not listed here (Wickremasinghe and Hoffbrand, 1979; Giannelli et al., 1977; Hand, 1975a). For human cells (Table I) the population mean and standard deviation is 0.60 f .021 pdminute. For other mammalian cells (Table II), it is 0.75 f 0.33 p d minute. Considering the variety of cell types, culture techniques, and measurement techniques, this is a surprisingly narrow range. In contrast, fork rates in bacteria are reported to be up to 15 pdminute (Cairns, 1963), and workers examining amphibians have found relatively low values of 0.02 pdminute (Hyodo and Flickinger, 1973). A. FORKRATESIN HUMANCELLS

Kapp and Painter (198 1) measured fork movement rates in 20 human cell types using a single technique (BUdR-3 13 nm photolysis) and uniform culture conditions. The overall mean fork rate was 0.53 f 0.08 pdminute (population mean and population standard deviation). For individual cell types the means and standard deviations were much larger. For example, GM637, an SV40-transformed cell line, had an average fork movement rate of 0.75 ? 0.20 pdminute,

10

LEON N. KAPP AND ROBERT B. PAINTER TABLE 1 FORKMOVEMENT RATESIN HUMANCELLS ~~

Cell type

Technique

A. Diploid cells Skin fibroblasts Autoradiography Skin fibroblasts Autoradiography Skin fibroblasts Autoradiography Skin fibroblasts Au toradiography Skin fibroblasts Autoradiography Skin fibroblasts Autoradiography Skin fibroblasts Autoradiography Skin fibroblasts BUdR-3 13 nm photolysis Skin fibroblasts BUdR-3 I3 nm photolysis GM498 fibroblasts BUdR-3 13 nm photolysis Lymphocytes Autoradiography Lymphocytes Autoradiography Embryonic fibroblasts Autoradiography Embryonic fibroblasts Autoradiography Embryonic fibroblasts Autoradiography HuP, embryonic lung fibroblasts Autoradiography MRC-5 fetal lung fibroblasts, passage 16 Autoradiography MRC-5 fetal lung fibroblasts, passage 58 Autoradiography WI-38 fetal lung Equilibrium density gradient fibroblasts Fetal brain cells Autoradiography Fetal heart cells Au toradiography Fetal retinal epithelial cells Autoradiography Fetal fibroblasts Autoradiography Fetal fibroblasts Autoradiography Fetal hepatocytes Autoradiography B. Cells from donors with genetic diseasesb XP fibroblasts XP fibroblasts XP fibroblasts Bloom’s fibroblasts Bloom’s fibroblasts Bloom’s fibroblasts

Autoradiography BUdR-3 13 nm photolysis BUdR-3 13 nm photolysis Autoradiography Autoradiography Autoradiography

Rate (pdminute)

Reference

0.8 0.65 0.64 0.62 0.60 0.6 0.35O O.5la 0.5 0.42 0.6 0.41 0.70 0.6 0.6

Ockey (1979) Hand and German (1 977) Hand and German (1975) Hand ( 1977) Yurov (1980) Yurov (1977) Heenen and Galand (1980) Kapp and Painter (1981) Kapp and Painter (1978) Kapp et al. (1979a) Yurov (1977) Hand and German (1977) Yurov (1979a) Yurov and Liapunova (1977) Yurov (1979b)

0.67

Ockey and Saffhill (1976)

0.5

Petes et al. (1974)

0.38

Petes et al. (1974)

0.9P 0.6 0.6

Painter and Schaefer (1969) Yurov (1977) Yurov (1977)

0.4 0.6 0.6 0.6

Yurov (1977) Yurov (1 977) Yurov and Liapunova (1977) Yurov (1977)

0.6 0.52a 0.46 0.75 0.47 0.51

Yurov (1978) Kapp and Painter (1981) Kapp et al. (1979a) Ockey (1 979) Hand and German (1975) Hand and German (1 977) (continued)

11

DNA REPLICATION FORK MOVEMENT RATES TABLE I (Continued)

Cell type Bloom's lymphocytes Bloom's fibroblasts Fanconi's anemia fibroblasts Fanconi's anemia fibroblasts Fanconi's anemia fibroblasts A-T fibroblasts A-T fibroblasts A-T fibroblasts Trisomy-2 1 lymphocytes Trisomy-21 fibroblasts Trisomy-7 fibroblasts C. Transformed cells HeLa HeLa HeLa HeLa HeLa HeLa HeLa HeLa HeLa HeLa HeLa HeLa Chang AC297 1 KDT Basal cell nevus Basal cell carcinoma GM637 GM637 XPl2RO Fanconi's anemia (SV40-transformed)

Technique

Rate (pdminute)

Reference

Autoradiography BUdR-313 nm photolysis

0.24 0.26a

Hand and German (1977) Kapp and Painter (1981)

Autoradiography

0.67

Hand (1977)

Autoradiography

0.65

Hand and German (1975)

BUdR-313 nm photolysis Autoradiography Autoradiography BUdR-3 13 nm photolysis Autoradiography BUdR-313 nm photolysis Autoradiography

0.55" 0.79 0.65 0.50" 0.6 0.43 0.8

Kapp and Painter (1981) Ockey ( I 979) Hand (1977) Kapp and Painter (1981) Yurov (1978) Kapp et al. (1979b) Yurov (1978)

Autoradiography Autoradiography Autoradiography Autoradiography BUdR/sedimentation BUdR-313 nm photolysis BUdR-3 13 nm photolysis BUdR-313 nm photolysis BUdR-3 13 nm photolysis Equilibrium density gradient Equilibrium density gradient Equilibrium density gradient Autoradiography BUdR-3 13 nm photolysis BUdR-313 nm photolysis Autoradiography Autoradiography BUdR-313 nm photolysis BUdR-3 13 nm photolysis BUdR-313 nm photolysis

0.79 0.5 0.35 0.32 0.3a 0.9 0.65 0.55 0.42 1.6 0.95 0.8" 0.86 0.53 0.57 0.87 0.24 0.83 0.75 0.53

Ockey and Saffhill ( I 976) Cairns (1966) Edenberg (1976) Stimac et al. (1977) Planck and Mueller (1977a) Povirk and Painter (1976) Kapp and Painter (1978) Kapp and Painter (1981) Painter (1980) Gautschi and Kern (1973) Gautschi et a / . , (1973) Painter and Schaefer (1969) Ockey and Saffhill (1976) Kapp and Painter (1981) Kapp and Painter (1 98 1 ) Ockey (1979) Heenen and Galand (1980) Kapp er al. (l979a) Kapp and Painter (1981) Kapp er al. (1979a)

BUdR-313 nm photolysis

0.57

Kapp and Painter ( I 98 1 )

"These papers did not present a single average fork movement rate as shown here; an average was calculated from the published data. bXP, Xeroderma pigmentosum; A-T, ataxia telangiectasia.

12

LEON N. KAPP AND ROBERT B. PAINTER TABLE I1 FORKMOVEMENT RATESI N NONHUMAN MAMMALIAN CELLS

Cell type Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster Chinese hamster

ovary ovary ovary ovary ovary ovary ovary ovary ovary ovary ovary ovary ovary ovary ovary ovary ovary V79 V79 V79 V79 V79 B 1 IFAF B 1 1FAF lung lung

Technique Autoradiography Equilibrium density gradient BUdR-313 nm photolysis BUdR-3 13 nm photolysis BUdR-313 nm photolysis Equilibrium density gradient BUdR-3 13 nm photolysis Alkaline sucrose gradient Autoradiography Autoradiography Autoradiography Autoradiography Alkaline sedimentation Alkaline sucrose gradient Autoradiography Autoradiography Autoradiography Autoradiography Au toradiography Equilibrium density gradient Autoradiography Autoradiography Autoradiography Autoradiography Autoradiography Equilibrium density gradient

Rate (pdminute) 0.5-2.5 I .4a 1.3 1.1

I .o 1 .o 1 .o

0.98 0.88 0.85a 0.8 0.6a 0.6* 0.6 0.5 0.32 0.31 1.O" 0.93° 0.9 0.61 0.36 0.8 0.930 0.58 0.3°

Reference Huberman and Riggs (1968) Taylor (1968) Povirk and Painter (1976) Kapp and Painter (1978) Kapp and Painter (1981) Kapp and Painter (1979) Kapp and Painter (1979) Walters et al. (1976) Ockey and Saffhill (1976) Hori and Lark (1973) Lark et al. (1971) Laughlin and Taylor (1979) Laughlin and Taylor ( 1979) Kurek and Taylor (1977) Taylor (1977) Taylor and Hozier (1 976) Stimac et al. (1977) Dahle et al. (1978) Yurov ( I979a) Painter and Schaefer (1969) Dahle er al. (1980) Doniger (1 978) Yurov and Liapunova ( 1977) Yurov (1979a) Martin and Oppenheim (1977) Martin and Oppenheim (1977) (continued)

or a range of 54% around the mean value. Hydroxyurea-treated HeLa cells used for comparison had fork rates of only 30% of normal, and rodent cells (CHO and S-49 mouse lymphosarcoma cells) had rates of 1 .O and 1.1 pdminute, indicating that the technique used was sensitive enough to detect differences in fork rates in the range of interest (0.3 to 1.0 pdminute) (Fig. 5). Similar variation can also be seen in autoradiographicexperiments. For example, Probst el af. (1980) reported mean fork movement rates in synchronous cells ranging from 0.75 to 0.91 pdminute. The standard deviations for these values were 0.23 and 0.44 pdminute, or 31 and 48%, respectively. From these values and from an examination of reported data, the 30% or greater variation in fork movement rates again appears regardless of the technique used or cell lines examined.

13

DNA REPLICATION FORK MOVEMENT RATES TABLE I1 (Continued) Rate Cell type Mouse L cells Mouse L929 Mouse L929 Mouse L cells Mouse L929 Mouse L cells Mouse lymphoma L5 178Y Mouse L5178Y Mouse L5178Y Mouse S49 lymphosarcoma Primary mouse thymocytes Mouse Ehrlich Mouse Friend erythroleukemia Mouse 3T3 Mouse 3T3 (SV40-transformed) ordnii) Rat (0. Rabbit CBL Mole MI5 Shrew S3 Shrew S3/4 Monkey CV-I Monkey CV-I

Technique

( pndminute)

Reference

Equilibrium density gradient Autoradiography Autoradiography Autoradiography Au toradiography Au toradiography Alkaline sucrose gradient Autoradiography Equilibrium density gradient BUdR-3 13 nm photolysis Alkaline sucrose gradient Autoradiography

1.6 0.830 0.7O 0.55 0.53 0.41 0.7 0.54 0.04 I .o 0.6

Painter and Schaefer (1969) Hand and Tamm (1972) Hand and Tamm (1973) Jasny er a / . (1980) Hand and Tamm (1977) Stimac er a / . (1977) Lehmann and Ormerod (1 970) Watanabe (1974) Okada (1968) Kapp and Painter (1981) Lanotte er al. (1977) Probst et al. (1980)

Autoradiography BUdR-313 nm photolysis

0.43 0.2

Kundahl et al. (1981) Kapp et al. (1979a)

BUdR-313 nm photolysis Autoradiography Equilibrium density gradient Equilibrium density gradient Autoradiography Autoradiography Autoradiography Equilibrium density gradient

0.33

Kapp et a!. (1 979a) Hori and Lark (1974) Painter and Schaefer (1969) Ockey and Saffhill (1976) Ockey and Saffhill (1976) Ockey (1978) Richter and Hand (1979b) Richter and Hand (1979b)

1.1

1 .o 1 .o

0.95 0.87 1.13 0.56 0.37a

aThese papers did not present a single average fork movement rate as shown here; an average was calculated from the published data.

The slowest value reported for human cells is 0.1 pdminute (Yurov, 1980), and the highest value is 1.6 pdminute (Gautschi and Kern, 1973). Yurov (1980) reported that the average rate was 0.5 to 0.6 pdminute, in excellent agreement with other investigators, but that fiber autoradiographs from a preparation originating from a single cell displayed a range of values from 0.1 to 1.2 pdminute, indicating that the cellular average is made up largely of differing individual rates. There are no clear trends in fork movement rates that can be used to subclassify the human cells. Normal diploid fibroblasts yield a mean of 0.58 ? 0.15 pdminute; transformed cells (such as HeLa) have a mean of 0.66 ? 0.30 p d minute, and these two means are not significantly different. When transformed and normal cells are measured simultaneously, the same conclusion can also be

14

LEON N. KAPP AND ROBERT B. PAINTER

1,

r

IW

z

3s:

s

B 3t

V

DIPLOID HUMAN

RODENT

CELL TYPE FIG.5 . The mean rates of displacement of DNA replication forks are shown for each of the cell lineb used. The vertical bars represent the standard deviation for each mean. Data for each cell type were obtained from a minimum of six separate determinations. The horizontal solid line and dashed lines through the points represent the population mean and its standard deviation (0.53 2 0.08 pm/ minute). Reprinted from Kapp and Painter (1981) with permission of ElseviedNorth-Holland Biomedical Press.

drawn (Painter and Schaefer, 1969; Kapp and Painter, 1978; Ockey, 1979). Only Heenen and Galand (1980) have reported that transformed cells have slower rates than normal cells. However, their reported values (0.3 to 0.4 pdminute for normal and 0.22 to 0.26 pdminute for transformed cells) are different by approximately 30%, which again is in the range of variation seen for most fork movement rates.

DNA REPLICATION FORK MOVEMENT RATES

15

Cells from patients with the genetic disease, Bloom’s syndrome, are apparently the only human cells that have a slower than normal fork movement rate. This was first reported by Hand and German (1975), using fiber autoradiography, who found that Bloom’s syndrome fibroblasts and lymphocytes both displayed low fork movement rates when compared to normal fibroblasts and lymphocytes. (Bloom’s lymphocytes’ fork rates were 59% of normal and Bloom’s fibroblasts were 78% of normal.) Giannelli er al. (1977), using alkaline sucrose gradient analysis of replicating DNA, confirmed these results but did not measure fork movement rates directly. However, Ockey (1979), using fiber autoradiography, found that Bloom’s syndrome cells had almost normal fork movement rates except when measured within 24 hours of subculturing, when the Bloom’s syndrome cells’ rate was significantly slower. Most recently, Kapp (1982), using BUdR-313 nm photolysis, found that fork movement rates in Bloom’s syndrome lines were 55-65% of those in normal controls, and found no indication of the subculturing effect mentioned by Ockey (1979). Thus, the bulk of evidence suggests that Bloom’s syndrome cells have DNA replication fork movement rates that are slower than normal. B. FORKRATESIN NONHUMAN MAMMALIAN CELLS The individual values for nonhuman mammalian cells (Table 11) range from 0.04 pdminute for a mouse line (Okada, 1968) to 2.5 pdminute in Chinese hamster ovary cells (Huberman and Riggs, 1968). As with the reports for human cell lines, the average values are in a relatively narrow range, despite the variety of cell types examined and the range of techniques used. Average values (calculated from Table 11) for the various species measured are mouse, 0.71 0.36 pdminute; Chinese hamster, 0.79 & 0.31 pdminute; moukey, 0.47 0.13 pdminute; shrew, 1.13 pm/minute; and rabbit, 1 .O pm/ minute. These values suggest that, in culture, rodent cells have fork movement rates higher than human cells.

* *

C. FORKRATESIN SYNCHRONIZED CELLS Early observations in synchronized cells indicated that the rate of DNA synthesis was not uniform throughout S phase. This could be explained either by a variation in the number of operating replicons or by a variation in the fork movement rate in individual replicons at different times in S phase. In the reports published to date, the number of operating replicons appears to be the main factor regulating the rate of DNA synthesis. Variation in fork movement rates during S phase does not follow any universal pattern and appears to vary among the different cell lines used, regardless of the technique used (Table 111). In the earliest report, Huberman and Riggs (1968), using CHO cells partially

TABLE 111 FORKMOVEMENT RATESIN SYNCHRON~ZED CELLS Cell type

Synchrony technique

Fork movement technique

CHO

12 hour FUdR

Autoradiography

HeLa S3

Mitotic selection

Equilibrium density gradient

HeLa S3

Equilibrium density gradient

CHO

Thymidine double block Mitotic selection

Shrew

Mitotic selection

Autoradiography

Normal human fibroblasts

Mitotic selection

BUdR-3 13 nm photolysis

CHO CHO HeLa

Mitotic selection Mitotic selection Mitotic selection

BUdR-313 nm photolysis Equilibrium density gradient BUdR-313 nm photolysis

Monkey CV-I Monkey CV-1 Mouse ascites

Serum-induced S phase Sorting of normal population

Fiber autoradiography Equilibrium density gradient Autoradiography

L5 178Y

Thymidine and colcemid block

Fiber autoradiography

Autoradiography

Results No change in rates in early, middle or late S (52.5 pdminute) 2- to 3-fold increase from early to late S (O.&l. 1 pdminute) 2- to 3-fold increase from early to late S (0.4-1.1 pdminute) 2- to 3-fold increase from early to late S (0.2 or 0.3-0.6 or 0.7 pdminute) No change during S phase ( I .1-1.2 p d minute) Minimum rate in middle S (0.55 pdminute); maximum rate early and late (0.8 p d minute) Constant rate (- 1 pdminute) Constant rate (- 1 pdminute) 2- to 3-fold increase from early to late S (0.8-2.5 pdminute) Constant throughout S (0.34 pdminute) Constant throughout S (0.15 p n h i n u t e ) In vivo: constant (1.1-1.2 pdminute); in vifrot slight increase in late S (0.75-0.9 pdminute) 3-fold increase from early to late S (0.32-1 .O pdminute)

Reference Huberman and Riggs (1968) Painter and Schaefer (1971) Painter and Schaefer (1971) Housman and Huberman (1975) Ockey (1978) Kapp and Painter (1979)

Kapp and Painter (1979) Kapp and Painter (1979) Kapp and Painter (1979) Richter and Hand (1979a) Richter and Hand (1979a) Probst ef al. (1980)

S. Sawada and T. Enomoto (personal communication)

DNA REPLICATION FORK MOVEMENT RATES

17

synchronized with FUdR, found that fork movement rates varied from 0.5 to 2.5 prrdminute at each point in S phase where this was measured. In a later study, Painter and Schaefer (197 I ) , using an equilibrium density gradient technique with HeLa cells, found a 2- to 3-fold increase from early to late S phase. This was found whether the cells were synchronized by mitotic selection or a double thymidine block. Housman and Huberman (1975) then reported a 3-fold change in fork movement rates in mitotically selected CHO cells as measured by fiber autoradiography. Lowest rates were at the beginning of S phase, with a rapid increase in rate during the first 1 or 2 hours and a fairly constant rate for the remainder of S phase. In contrast to these results, however, other workers have reported constant fork movement rates in S phase. Ockey (1978), using autoradiography with mitotically selected shrew cells, found a constant rate of fork movement. Kapp and Painter (1979), using mitotically selected CHO cells, reported a constant rate of fork movement throughout the S phase, including the first hour of S. Kapp and Painter used two techniques-the equilibrium density gradient method of Painter and Schaefer (1971) and the BUdR-313 nm photolysis method of Povirk and Painter .(1 9 7 6 t a n d obtained identical results with both methods. However, using the BUdR-3 13 nm photolysis method with HeLa cells, a 3-fold increase in fork movement rates from early to late S phase was found, thus confirming the earlier report by Painter and Schaefer (1971) for the same cells using a different technique. Constant fork rates through S phase were also reported by Richter and Hand (1979a) with serum-induced synchrony in monkey CV-1 cells. This result was obtained using both fiber autoradiography and the equilibrium density gradient method of Painter and Schaefer (1971). Probst et al. (1980), using autoradiography, also found a constant fork movement rate through S in mouse ascites cells. They examined the cell cycle by sorting cells according to DNA content (and thus position in S phase). Using this approach, they tried to avoid all artifacts that might have resulted from an induced synchrony method. They also found a constant rate regardless of whether the cell population was labeled in vivo or in vitro. Recently, however, S. Sawada and T. Enomoto (personal communication), using fiber autoradiography, found a 3-fold increase in fork movement rates as mouse L5178Y cells progressed through S phase. Fork movement rates in synchronized cells therefore seem to follow one of two patterns, a constant rate throughout S phase or an increasing rate through S phase. However, to say that there are two patterns is probably an oversimplification. For example, HeLa cells (Painter and Schaefer, 1971; Kapp and Painter, 1979) and L5178Y cells (S. Sawada and T. Enomoto, personal communication) seem to have monotonically increasing rates of fork movement in S . However, CHO cells, when they do exhibit an increasing fork movement rate (Housman and Huberman, 1975), show most of the increase in early S phase. More uniform

18

LEON N. KAPP AND ROBERT B . PAINTER

findings may result if more investigators use similar patterns of sampling. Different cell lines may behave differently depending on their history, origin, and culture conditions, and this may partially explain the different results found for CHO cells by two different groups of investigators (Housman and Huberman, 1975; Kapp and Painter, 1979), and the variation in results for different lines of mouse cells (Probst et al., 1980; S . Sawada and T. Enomoto, personal communication). Regardless of whether the fork movement rates vary by a factor of 3 during S phase or are constant throughout, the variation in overall DNA synthesis during S phase is much larger than 3-fold and therefore must be primarily caused by the different numbers of operating replicons (Painter and Schaefer, 1971; Kapp et al., 1979b).

FORKMOVEMENT RATES D. CHEMICALS AFFECTING Although there have been a number of studies of the effects of drugs and other compounds on DNA synthesis, very few have explicitly examined the effects of these agents on DNA fork movement rates; the results of these studies are presented in Table IV. Hydroxyurea inhibits DNA synthesis by inhibiting ribonucleotide reductase and reducing the production of precursors of all four DNA bases (Young and Hodas, 1964). Inhibition of DNA synthesis by hydroxyurea depends on the concentration of the drug, but in the range 10W5 to 10 - 4 M , DNA fork movement rates are about 30 to 40% of those in untreated controls (Painter, 1980; Kapp and Painter, 1981; Gautschi er al., 1973). Inhibitors of protein synthesis also depress fork movement rates. Cycloheximide (0.1 to 1.0 kg/ml) slows fork movement rates to 25-35% of normal (Gautschi et al., 1973; Gautschi and Kern, 1973), and Stimac et al. (1977) M . Puromycin at 200 kg/ml reported rates of 55% of control at 1.8 X (Hand and Tamm, 1972) also inhibits fork movement rates to about 55% of control. Gautschi and his co-workers concluded that the depression in DNA synthesis in the presence of protein synthesis inhibition is completely accounted for by the depression in fork movement rates (Gautschi et al., 1973; Gautschi and Kern, 1973). However, Hand and Tamm (1972, 1973) and Hand (1975b) found that there must also be some inhibition of initiation in the presence of protein synthesis inhibition. Gautschi (1974) examined the effects of puromycin on DNA synthesis and found that its action was very similar to that of cycloheximide and that all of the depression of DNA synthesis could be accounted for by the slowing of fork movement rates. He concluded that fork movement was probably mediated through protein synthesis. Similar conclusions were also reached by Planck and Mueller (1977a). Stimac et al. (1977) examined the effects of a number of protein synthesis inhibitors on DNA synthesis using fiber autoradiography. They found that at

TABLE IV CHEMICALS AFFECTINGFORKMOVEMENT RATES

Cell type HeLa HeLa CHO Mouse CHO HeLa L929 HeLa HeLa HeLa HeLa Mouse Mouse CHO CHO HeLa HeLa Mouse Mouse

L

L929 L929

L929 L929

Chemical Cycloheximide (0.1 pg/ml) Cycloheximide ( I .O pg/ml) Cycloheximide (10 pglml) M) Cycloheximide (1.8 X Cycloheximide (1.8 X 10 - 4 M ) Cycloheximide (1.8 X 10-4M) M) Cycloheximide (1.8 x Hydroxyurea (10 - 5 M ) Hydroxyurea M) Hydroxyurea ( 10- 5 M) Hydroxyurea (0.1 mM) Puromycin (20 pglml) Puromycin (200 pg/ml) Puromycin (20 pg/ml) Puromycin (4.2 X M) 2,4-Dinitrophenol 2,4-Dinitrophenol Dichlorobenzimidazole riboside (60 pglml) Dichlorobenzimidazole riboside (90 pg/ml)

Technique Equilibrium density gradient Equilibrium density gradient Equilibrium density gradient Autoradiography Autoradiography Autoradiography Autoradiography Equilibrium density gradient Equilibrium density gradient BUdR-313 nm photolysis BUdR-313 nm photolysis Autoradiography Autoradiography Autoradiography Autoradiography Equilibrium density gradient Autoradiography Autoradiography Autoradiography

Fork movement rate (9% of control)

36 23 25 54 55

37 25 47 38 40 30 80 55 100

50 100

70 82 75

Reference Gautschi er al. (1973) Gautschi er al. (1973) Gautschi and Kern (1973) Stimac et al. (1977) Stimac et al. (1977) Stimac et al. (1977) Hand and Tamm (1973) Gautschi er a/. (1973) Gautschi er al. (1973) Kapp and Painter (1981) Painter ( 1980) Hand (1975b) Hand (1975b) Hori and Lark (1973) Hand and Tamm ( 1 973) Gautschi er al. (1973) Stimac er al. (1977) Hand and Tamm (1977) Hand and Tamm (1977)

20

LEON N. KAPP AND ROBERT B. PAINTER

early times after inhibition of protein synthesis, the decrease in fork movement rate could account for the decrease in overall DNA synthesis, and after later times the decrease in fork movement rates may still have been sufficient to explain the decrease in overall DNA synthesis during severe inhibition. However, if DNA synthesis was only moderately inhibited, the decrease in fork movement rate was not sufficient to account for the decrease in DNA synthesis, and under these conditions it appeared that initiation of new replicons must also have been inhibited. In marked contrast to the previously mentioned reports, Hori and Lark (1973) reported that puromycin had no effect on fork movement rates and therefore must act solely by blocking initiation of replicons. However, this is the only report that failed 'to observe an effect on fork movement rates by inhibition of protein synthesis. In reviewing the work that has been done, it appears that inhibition of protein synthesis rapidly inhibits DNA chain elongation and that the effect on initiation probably depends on the concentration of inhibitor and the length of treatment. Other inhibitors of protein synthesis, emetine and pactamycin, gave results similar to those described for puromycin (Stimac et al., 1977). Stimac er al. also claimed the same effect for 2,4-dinitrophenol, a general metabolic inhibitor. However, this was in contrast to Gautschi et al. (1973), who reported no effect of dinitrophenol on fork movement rates. The reported data of Stimac et al. (1977) do not completely support their contention that the effects of 2,4-dinitrophenol on overall DNA synthesis can be explained solely by its effect on rate of fork movement. Another agent examined in detail for its effects on fork movement rates is dichlorobenzimidazole riboside, which inhibits RNA synthesis. At concentrations that inhibit most heterogeneous nuclear RNA (hnRNA) synthesis, DNA synthesis is inhibited by 20 to 25% and this can be accounted for entirely by a reduced fork movement rate (Hand and Tamm, 1977). Methyl methanesulfonate was also reported to inhibit fork movement rates to 15 to 50% of normal, depending on the concentration used (Dahle et al., 1978). It was concluded that this drug produces lesions that can inhibit fork movement rates and, since the depression in fork movement rates does not account for the total depression of DNA synthesis, may also inhibit replicon initiation. However, there are other interpretations of the effect of this drug (Painter, 1977). Ultraviolet light strongly inhibits DNA synthesis in mammalian cells. Most reports have concluded that ultraviolet light exposure results in a lesion that blocks fork movement, but fork movement rates between lesions appear to be unaffected (Povirk and Painter, 1976; Edenberg, 1976; Dahle er al., 1980). Doniger (1978), however, reported that fork movement rates were depressed by ultraviolet light at 10 J/m2 but not at 5 J/m2. X Rays also affect the rate of DNA synthesis. However, low to moderate

DNA REPLICATION FORK MOVEMENT RATES

21

doses of X rays do not affect fork movement rates (Kapp and Painter, 1978), but primarily inhibit initiation of replicons (Watanabe, 1974; Makino and Okada, 1975; Dahle et al., 1979; Laughlin and Taylor, 1980). In summary, agents that directly or indirectly reduce the supply of DNA precursors slow the rate of fork displacement in mammalian cells. DNA-damaging agents, in contrast, probably do not affect the rate of fork displacement but inhibit DNA synthesis by blocking initiation of replicons or by completely blocking fork movement. E. FORKRATESIN MUTANTS AND in Vitro Mutants have been very important for DNA replication studies in bacteria, and many investigators have tried to obtain mutant mammalian cell lines with alterations in DNA synthesis pathways. Zannis-Hadjopoulos er al. (1980) studied DNA replication in a purine-auxotrophic mutant cell line derived from Chinese hamster V79 cells. When purines were removed from the medium, DNA synthesis slowed to approximately 35% of the rate in cells growing in purinesupplemented medium. By observing the effects of starving and refeeding the cells with purines, it was concluded that purine deprivation had a direct effect on fork movement. In addition, there appeared to be no effect on initiation of replicon operation at the level of individual replicons. A temperature-sensitive mutant, tsBN-2 of BHK-2 I cells, exhibits a depression of DNA synthesis at a nonpermissive temperature of 39.5"C. Examination of DNA synthesis in these cells (Eilen et al., 1980) indicated that the fork movement rate in the wild-type cells at 39.5"C was 0.74 to 0.86 pdminute, whereas the mutant had a fork movement rate of approximately 1 pdminute. Both wild-type and mutant cells have lower rates of fork movement at 33.5"C. Autoradiography indicated that the average distance between replication origins in the mutant increased from approximately 80 to 98 pm at the nonpermissive temperature of 39.5"C, whereas there was little effect on interorigin distance in wild-type cells. Thus, because the depression in DNA synthesis in the mutant BN-2 cells cannot be explained by slower fork movement, it appears to be due solely to fewer initiations at the nonpermissive temperature. Hyodo and Suzuki (1982) isolated a temperature-sensitivemutant from mouse FM3A cells that was found to have defective DNA replication at a nonpermissive temperature of 39.5"C. Cell cycle analysis indicated that most of the cells were arrested in S phase. DNA fiber autoradiography revealed that DNA fork movement rates were approximately 1.04 pdminute at the permissive temperature and 0.47 pdminute at the nonpermissive temperature. Thus, the reduced rate of DNA replication in these cells did result from a slow fork movement rate. Although many mutant cell lines have been isolated in the past few years, and some of them may represent mutations that affect DNA synthesis, most of them

22

LEON N. KAPP AND ROBERT B . PAINTER

have yet to be characterized or studied with respect to specific DNA synthesis parameters such as fork movement rate. Eventually, if such mutants are found and characterized, they should be of value as model systems for examining fork movement rates in mammalian cells. A number of in vitro systems have also been employed to examine DNA synthesis in mammalian cells, but very few workers have examined fork movement rates directly in these systems. Most investigators have reported that fork movement rates are retarded in such systems and range from as low as 10 to as high as 60% of the rate in vivo (Planck and Mueller, 1977b; Fraser and Huberman, 1977; Gautschi et al., 1977). In all cases, the depression in DNA synthesis rate was caused by a lowered fork movement rate (Planck and Mueller, 1977b; Fraser and Huberman, 1977; Gautschi ef a!., 1977; Hand and Gautschi, 1979; Griffiths and Carpenter, 1980) as well as premature chain termination (Hand and Gautschi, 1979) and lack of new initiations (Planck and Mueller, 1977b). These in vitru systems are incapable of maintaining DNA synthesis for more than a short time after their preparation (Hand and Gautschi, 1979; Fraser and Huberman, 1977). Such systems will become more useful when they are better characterized and standardized and are capable of supporting DNA synthesis at higher levels and for longer periods of time.

IV. Concluding Remarks DNA fork movement rates in most mammalian cells appear to fall into a range of approximately 0.2 to I pdminute. The average value for human cell lines is 0.60 k 0.21 pdminute, whereas the average value for nonhuman mammalian cells is 0.71 5 0.36 pdminute. In addition, most of the rates reported for human cell lines are in the range of approximately 0.4 to 0.7 pdminute. In contrast, bacterial fork movement rates are reported to be as high as 15 p d minute (Cairns, 1963) and amphibian rates are reported to be as low as 0.02 p d minute (Hyodo and Flickinger, 1973). There appears to be little variation among the various human cell lines examined. The rates are similar in diploid cells, cells from donors with genetic diseases, and in transformed cells. The only human cell type that appears to have a low fork movement rate is Bloom’s syndrome. Although a number of techniques have been used to measure fork movement rates in mammalian cells, most yield values in the same range and thus give comparable results. The variations reported for fork movement rates are probably due mainly to actual variations between the cells examined and, to a smaller extent, to differences caused by various experimental and cell culture conditions. Individual fork movement rates can vary by almost 10-fold within a single cell type. However, the average fork movement rate can either increase or remain

DNA REPLICATION FORK MOVEMENT RATES

23

constant throughout S phase, depending on the cell type examined. Because even the highest variation in fork movement rates observed in synchronized cells cannot account for the variation in cellular DNA synthesis rates, the main factor affecting the overall cellular rate of DNA synthesis must be the number of replicons operating at one time rather than the replication fork movement rate. There have been few studies examining the effect of chemicals or other agents on actual fork movement rate. In addition, there are still no clearly defined mutants in which an identifiable alteration or mutation can be correlated with an altered fork movement rate. In vitro systems in mammalian cells are still incapable of supporting a high level of DNA synthesis for more than a short period of time. Future work to examine the regulation and role of fork movement rates in mammalian cells will probably depend on developing new or improved methods, e.g., defined mutants, better in v i m systems, and new probes or techniques to measure fork movement rates.

ACKNOWLEDGMENT This work was supported by the U.S.Department of Energy.

REFERENCES Cairns, J. (1963). J. Mol. Biol. 6, 208-213. Cairns, J . (1966). J. Mol. Biol. 15, 372-373. Dahle, D. B., Griffiths, T. D., and Carpenter, J. G. (1978). Mol. Pharrnacol. 14, 278-289. Dahle, D., Griffiths, T. D., and Carpenter, J. G. (1979). Radiat. Res. 78, 542-549. Dahle, D., Griffiths, T. D., and Carpenter, J. G. (1980). Photochem. Phorobiol. 32, 157-165. DePamphilis, M . L., and Wassarman, P. M. (1980). Annu. Rev. Biochem. 49, 627-666. Doniger, J. (1978). J. Mol. Biol. 120, 433-446. Edenberg, H. J . (1976). Biophys. J . 16, 849-860. Edenberg, H. J . , and Huberman, J. A. (1975). Annu. Rev. Genet. 9, 245-284. Eilen, E., Hand, R., and Basilico, C. (1980). J . Cell. Physiol. 105, 259-266. Fraser, J . M . K., and Huberman, J. A. (1977). J. Mol. Biol. 117, 249-272. Gautschi, J . R. (1974). J. Mol. Biol. 84, 223-229. Gautschi, J. R., and Kern, R . M. (1973). Exp. Cell Res. 80, 15-26. Gautschi, J. R., Kern, R. M . , and Painter, R. B. (1973). J. Mol. Biol. 80, 393-403. Gautschi, J . R., Burkhalter, M., and Reinhard, P. (1977). Biochim. Biophys. Acta 474, 512-523. Giannelli, F., Benson, P. F., Pawsey, S. A,, and Polani, P. E. (1977). Narure (London) 265, 466-469. Griffiths, T. D., and Carpenter, J. G. (1980). Exp. Cell Res. 130, 470-473. Hand, R. (1975a). J. Cell Biol. 64, 89-97. Hand, R. (1975b). J . Cell Biol. 67, 761-773. Hand, R. (1977). Hum. Genet. 37, 55-64.

24

LEON N. KAPP AND ROBERT B. PAINTER

Hand, R. (1978). Cell 15, 317-325. Hand, R. (1979). In “Cell Biology; A Comprehensive Treatise” (D. M. Prescott and L. Goldstein, eds.), Vol. 2, pp. 389-437. Academic Press, New York. Hand, R., and Gautschi, I. R. (1979). J. Cell B i d . 82, 485-493. Hand, R., and German, J . (1975). Proc. Narl. Acad. Sci. U.S.A. 72, 758-762. Hand, R., and German, I. (1977). Hum. Genet. 38, 297-306. Hand, R., and Tamm, 1. (1972). Virology 47, 331-337. Hand, R., and Tamm, I . ( 1973). J . Cell B i d . 58, 4 10-4 18. Hand, R., and Tamm, I. (1977). Exp. Cell Res. 107, 343-354. Heenen, M., and Galand, P. (1980). Nature (London) 285, 265-267. Hori, T., and Lark, K. G. (1973). J . Mol. B i d . 77, 391-404. Hori, T., and Lark, K. G. (1974). J. Mol. Biol. 88, 221-232. Housman, D., and Huberman, J . A. (1975). J. Mol. Eiol. 94, 173-181. Huberman, J . A., and Riggs, A. D. (1968). J . Mol. Biol. 32, 327-341. Hyodo, M., and Flickinger, R. A. (1973). Eiochim. Biophys. Actu 299, 24-33. Hyodo, M., and Suzuki, K. (1982). Exp. Cell Res. 137, 31-38. Jasny, B. R.. Cohen, J . E., and Tamm, I. (1980). Chromosoma 79, 207-214. Kapp, L. N. (1982). Biochim. Biophys. Acta 696, 226-227. Kapp, L. N., and Painter, R. B. (1978). Biophys. J . 24, 739-748. Kapp, L. N., and Painter, R. B. (1979). Eiochim. Biophys. Acta 562, 222-230. Kapp, L. N., and Painter, R. B. (1981). Eiochim. Biophys. Acta 656, 36-39. Kapp, L. N., Park, S. D., and Cleaver, J. E. (1979a). Exp. Cell Res. 123, 375-378. Kapp, L. N., Millis, A. J. T.. and Pious, D. A. (1979b) In V i m 15, 669-672. Kundahl, E., Richman, R., and Flickinger, R. A. (1981). J. Cell. Physiol. 108, 291-298. Kurek, M. P., and Taylor, J. H. (1977). Exp. Cell Res. 104, 7-14. Lanotte, M., Moerman, C., and Panijel, J . (1977). Exp. Cell Res. 109, 191-200. Lark, K. G., Consigli, R., and Toliver, A. (1971). J. Mol. Biol. 58, 873-875. Laughlin, T. J., and Taylor, J. H. (1979). Chromosoma 75, 19-35. Laughlin, T. J., and Taylor, J. H. (1980). Radiut. Res. 83, 205-209. Lehmann, A. R., and Ormerod, M. G. (1970). Biochim. Biophys. Actu 204, 128-143. Makino, F., and Okada, S. (1975). Radiat. Res. 62, 37-51. Martin, R. G., and Oppenheim, A. (1977). Cell 11, 859-869. Ockey, C. H. (1978). Exp. Cell Res. 114, 446-45 I . Ockey, C. H. (1979). J. CellSci. 40, 125-144. Ockey, C. H., and Saffhill, R. (1976). Exp. Cell Res. 103, 361-373. Okada, S. (1968). Biophys. J . 8, 650-664. Painter, R. B. (1976). I n “Handbook of Genetics” (R. C. King, ed.), Vol. 5, pp. 169-186. Plenum, New York. Painter, R. B. (1977). M u m . Res. 42, 299-304. Painter, R. B. (1980). J . Mol. Biol. 143, 289-301. Painter, R. B., and Schaefer, A. W. (1969). J . Mol. B i d . 45, 467-479. Painter, R. B., and Schaefer, A. W. (1971). J. Mol. B i d . 58, 289-295. Painter, R. B., Jermany, D. A., and Rasmussen, R. E. (1966). J. Mol. Biol. 17, 47-56. Petes, T. D., Farber, R. A., Tarrant, G. M., and Holliday, R. (1974). Nature (London) 251, 434-436. Planck, S. R., and Mueller, G. C. (1977a). Biochemistry 16, 1808-1813. Planck, S. R., and Mueller, G. C. (1977b). Biochemistry 16, 2778-2782. Povirk, L. F., and Painter, R. B. (1976). Biophys. J. 16, 883-889. Probst, H., Blutters, R., and Fielitz, J . (1980). Exp. Cell Res. 130, 1-13. Richter, A,, and Hand, R. (1979a). Exp. Cell Res. 121, 363-371.

DNA REPLICATION FORK MOVEMENT RATES

25

Richter, A., and Hand, R. (1979b). J . Cell. Physiol. 101, 407-418. Roti-Roti, J. L., and Painter, R. B. (1977). J. Theor. B i d . 64, 681-696. Sheinin, R., and Humbert, J. (1978). Annu. Rev. Biochem. 47, 277-316. Stimac, E., Housman, D., and Huberman, J. A. (1977). J. Mol. B i d . 115, 485-51 I Taylor, J. H. (1968). J . Mol. B i d . 31, 579-594. Taylor, J . H. (1977). Chromosoma 62, 291-300. Taylor, J. H., and Hozier, J. C. (1976). Chromosoma 57, 341-350. Waiters, R. A,, Tobey, R. A,, and Hildebrand, C. E. (1976). Biochim. Biophys. Actu 447, 36-44. Watanabe, 1. (1974). Radiat. Res. 58, 541-556. Wickremasinghe, R. G., and Hoffbrand, A. V. (1979). Biochim. Biophys. Acta 563, 46-58. Young, C. W., and Hodas, S. (1964). Science 146, 1172-1174. Yurov, Y. B. (1977). Cell Differ. 6, 95-104. Yurov, Y. B. (1978). Hum. Genet. 43, 47-52. Yurov, Y. B. (1979a). Exp. Cell Res. 123, 369-374. Yurov, Y. B. (1979b). Chromosoma 74, 347-353. Yurov, Y. B. (1980). J . Mol. Eiol. 136, 339-342. Yurov, Y. B., and Liapunova, N. A. (1977). Chromosoma 60, 253-267. Zannis-Hadjopoulos, M., Taylor, M. W., and Hand, R. (1980). J. CellEiol. 85, 777-785.

This Page Intentionally Left Blank

INTERNATIONAL REVIEW OF CYTOLOGY, VOL. 80

Interaction of Viruses with Cell Surface Receptors MARCTARDIEU,~ ROCHELLE L. EPSTEIN,AND HOWARDL. WEINER Department of Neuroscience, Children's Hospital Medical Center, and Department of Medicine, Neurology and Infectious Disease Divisions, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachussets I. Definition of Viral Receptor Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Biological Characteristics of Viral Attachment to Cells . . . . . . . . . . A. Techniques Used to Study Viral Attachment B. Mathematical Analysis of Viral Binding . . . . . . . . . . . . . . . . . . 111. Membrane Components Which Interact with Viruses. . . . . . . . . . . . A. Density and Affinity of Viral Receptor . . . ....... .. omponents B. Relationship of Viral Receptors to Other S C. Definition of Cellular Receptor Units by Monoclonal Antibodies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Specificity of Cellular Receptor Sites for Viruses. . . E. In Vitro Manipulation of Cell Receptor Sites. . . . . . . F. Age Dependency of Viral Permissiveness . . . . . . . . . . . . . . . . . G. Genetic Control of Cell Receptor Site Expression. . . . . . . . . . . IV. Viral Components Which Recognize Cellular Receptors . . . . . . . . . A. Picomaviruses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Adenoviruses . . . . . . . . . . . . . . . ................ C. Reoviruses . . . . . . . . . . . . . . . . . . . . . . . . .

E. Retroviruses . . . ............... F. Coronaviruses. , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Virus-Receptor Interactions and Pathogenicity B. Role of Virus Attachment Proteins in Pathogenicity C. Induction of Cell-Specific Autoimmunity following Viral VI. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References

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

27 30 30 35 37 37 38 39 40 45 48 48 49 49 50 50 51 52 52 52 53 54 55 56 57

I. Definition of Viral Receptor Sites The definition of a receptor ultimately depends upon the structural and functional identification of a site that is specifically recognized by a ligand. The most rigorous characterizations of receptor-ligand interactions have been derived 'Present address: Inserm U56, HGpital de Bicetre, 94270 Bicetre, France. 27 Copyright 0 1982 by Academic Press. Inc. All rights of reproduction in any form reserved. ISBN 0-12-364480-0

28

MARC TARDIEU ET AL.

from binding studies of radiolabeled ligands in neuropharmacology and endocrinology. It has been more difficult to study the interactions of viral particles with the cell surface in this rigorous a fashion. Viral particles are several orders of magnitude larger than conventional ligands, making it impossible to perform many of the manipulations that are routinely done with hormones, for example. In addition, viruses contain repeating subunits and many copies of the viral attachment protein are present on the virion surface. Thus a single virion may interact with many receptor sites on the cell surface. Nonetheless, most of the principles of receptor-ligand interactions established for other ligands apply to viruses as well and this establishes a framework for the definition of viral receptors. At the simplest level, a viral receptor is the structure on the membrane surface of a cell to which virus binds prior to entering the cell. Often the presence or absence of a viral receptor determines whether or not the cell can be infected by virus. It is also possible that there are structures on the cell surface to which a virus binds but which do not serve as conduits for viral entry into the cell. Viral binding to such sites could have other important biologic consequences such as affecting cell function by triggering surface structures which affect cellular metabolism or being improtant for the generation of an immune response against the virus. In the broadest sense, then, a viral receptor is a structure on the cell surface to which a virus binds, the binding of which is of biological importance and can be measured in a biologically relevant way. The definition of viral recognition sites as “receptors” involves three major criteria which are derived from models of ligand-receptor interactions (Bennet, 1978). These include saturability, specificity, and competition. SaturnbiliQ. If virus interacts with the cell surface at discrete sites along the membrane, only a finite number of sites will be available for viral binding and high concentrations of virus should be able to fully occupy or “saturate” them. It should therefore be possible to increase the concentration of particles presented to the cell surface until no further binding results. This can be shown experimentally by determining viral binding as a function of increasing viral concentration. Saturation is demonstrated if a plot of the result is a hyperbolic curve (see Fig. IA). Specificity. Specificity is first demonstrated according to “biologic parameters.’’ In other words, viral binding is observed only to cells that the virus infects or to cells where viral binding induces some other biologically measurable response. The second measure of specificity relates to the binding assay itself. Even when virus binds to a biologically relevant cell, a certain proportion of viral binding is nonspecific and unrelated to specific viral receptors. For example, electrostatic forces result in some of the nonspecific adherence of particles to the cell surface. Nonspecific binding is contained in any binding curve but the nonspecific component is usually not saturable and therefore it is a

VIRUSES AND CELL SURFACE RECEPTORS

29

12 V-ImL7

,?‘?KF

Total Blnding

BOUND isaturable CPM x fO ’J

FIG. 1 . (A) Binding of 125I-labeled reovirus type I to L-cell fibroblasts. Each point represents the arithmetic mean of three separate determinations of uptake to 250,000 cellsisample equilibrated at 25°C for 60 minutes. The linear nonspecific binding was determined by mixing each concentration of labeled virus with an equal volume of high titer unlabeled virus containing 2.0 X 1013particlesiml. The saturable binding curve results from subtraction of the linear component from the total binding curve. The maximum uptake is the uptake value at saturation. and Kd is the virus concentration at the half saturation point. (9)Scatchard plot of ‘2SI-labeled reovirus bound to L cells. After subtracting the linear component, boundifree was calculated by dividing the corrected cpm by the viral concentration expressed in particles divided by Avogadro’s number (to result in a molar expression). Linear regression analysis of the data provides the slope to estimate K d .

linear function which can be measured, and subsequently subtracted from total binding to reveal “specific” binding (Fig. 1A). Competition. It should be possible to competitively inhibit specific binding using a second ligand which is known to bind to the same receptor. In neuropharmacology, for example, specific binding of acetylcholine to the muscarinic acetylcholine receptor is defined as that component of total binding which can be blocked by atropine. In viral systems, specific blockers of this type are usually unavailable and the only certain method for competition is to block the binding of

30

MARC TARDIEU ET AL.

radiolabeled virus with unlabeled virus or, if possible, the viral attachment protein. A variety of nomenclatures have been used to describe viral-receptor interactions, and depending on the context of the discussion several abbreviations will be used in this article. The protein by which the virus attaches to cells is the viral attachment protein (VAP). The “receptor” on the surface of the cell to which the virus binds is the cellular receptor site (CRS). Finally, it has been postulated that the CRS may be composed of multiple units and these units, which may bind a single VAP, are termed cellular receptor units (CRU) (Lonberg-Holm, 198 1).

11. Biological Characteristics of Viral Attachment to Cells

A. TECHNIQUES USED TO STUDYVIRALATTACHMENT A variety of approaches have been used to study the interaction of viral particles with cell surface receptors or reception sites. As described below, each approach has advantages and limitations. A rigorous study of viral-receptor interactions requires the use of more than one technique since different approaches provide complementary information about viral binding. 1. Infectivity Assays

These assays measure the number of infectious viral particles that either remain in the medium following viral binding to the cells or that have attached to cells. Attached virus may be detected by measuring infected cells or “infectious centers,” or by dissociation of receptor bound infectious virus, providing that receptor-mediated or cell-mediated modification (‘‘eclipse”) of the virus particles can be prevented or accounted for. The advantage of infectivity assays is that binding of very small amounts of virus can be detected and viral replication, a biologic function, is being measured. Infectivity assays have the following limitations: (1) Only binding of infectious viral particles can be measured. Most viral preparations have particle to plaque-forming unit (PFU) ratios of 10 to 100 and it is likely that noninfectious virus also binds to cellular receptor sites. (2) The measurement of infectious particles by plaque assay is time consuming, cumbersome, and statistically error-prone, making the generation of quantitative data difficult. (3) When infectious centers are being measured, cells which bear membrane receptors but which are unable to support viral replication cannot be detected. Attachment to such cells must be detected by measurement of virus removal from the media. (4) Factors which modify penetration and replication can affect results, since the data reflect only the end product of infectivity. In one series of experiments, investigators surmounted some of these limita-

31

VIRUSES AND CELL SURFACE RECEPTORS

tions. Infectivity assays were used to study receptor-determined host restriction for ecotropic and xenotropic murine leukemia viruses. In this approach, pseudotype virions were created which contained the envelope of one virus and the infectious genome of another. Pseudotype virions were generated by the mixed infection of VSV (for which the cells being studied were permissive) and various retroviruses (for which the cells were not permissive). Resultant virions that contained VSV envelopes were then inactivated by anti-VSV neutralizing antibodies. The remaining pseudotype virions (retrovirus envelope VSV genome) bound to the cell surface receptor for the retrovirus being studied and resulted in. VSV replication. Thus, VSV replication measured the presence of surface receptors for retroviruses on a variety of host cells, independent of their ability to replicate the virus (Besmer and Baltimore, 1977).

+

2 . Hemagglutination Hemagglutination assays depend on the ability of many viruses to agglutinate red blood cells. Although red cell receptors responsible for viral hemagglutination might not be the same as viral receptors on host cells, hemagglutination has provided important information about virus-receptor interactions. Of note is that hemagglutination depends upon lattice formation and is very dependent upon conditions (e.g., temperature). Thus, it does not measure “attachment” direct!y. Studies using hemagglutination have detected differences in neuraminidase sensitivity of erythrocyte receptors for various enveloped viruses. For example, the erythrocyte receptor for parainfluenza virus type 3 appears to be neuraminidase resistant, while receptors for influenza virus serotypes A and B (Hirst, 1950) are inactivated by neuramininidase. Adenovirus type 9 does not bind to neuraminidase-treated erythrocytes while types 2 and 7 do (Wadell, 1969; Boulanger et a l . , 1972). For some nonenveloped viruses, hemagglutination is totally resistant to neuraminidase (e.g., reovirus, Gomatos and Tamm, 1962). A modified version of hemagglutination is the rosetting technique. In this method, cells to be tested are added to aliquots of chronically infected cells. Cells which express viral receptors adhere to the surface of infected cells which express the the viral attachment protein (VAP). In order to be used, the rosetting technique requires that a sufficient number of VAPs be expressed on the surface of infected cells. Hemadsorption of erythrocytes to cultures of virus-infected cells is a variation of the rosetting technique. In this method rosettes are formed only if the erythroctyes have viral “receptors” on their surface. The technique of viral-induced agglutination has also been used for cells other than erythrocytes, specifically, lymphocytes. Woodruff and Woodruff ( 1972, 1974) have used the agglutination of T lymphocytes by viruses to define receptors for myxo- and paramyxoviruses on these cells (discussed in Section 111,D). These techniques also allow characterization of receptors by use of reagents

32

MARC TARDIEU ET AL.

which block agglutination (e.g., enzymes or antibodies) (Bankhurst et al., 1979). 3. Use of Fluorochromes Fluorescein- or rhodamine-labeled antiviral antibodies have been used to visualize the presence of virus bound to the surface of various cell types. For example, indirect immunofluorescence techniques were used to define B cells as the lymphocyte subset susceptible to EBV infection (Jondal and Klein, 1973). The major advantage of measuring virus binding by indirect immunofluorescence is that it affords a qualitative approach for the study of viral-receptor interactions. This is especially relevant for studying primary cell cultures which contain a mixture of cell types. For example, for the study of serotype-specific binding of reovirus to cells in the nervous system, we prepared cell suspensions from mouse brain which contained 75% ciliated ependymal cells; these cells are easily identified under phase microscopy by their cilia. Using indirect immunofluorescence, it was possible to demonstrate that reovirus type 1 (which has an in vivo affinity for ependymal cells) but not type 3 binds to these cells (see Fig. 2A and B, discussed in detail later). Similarly, indirect immunofluorescence was used to demonstrate that reovirus type 3 binds to a subset of murine and human lymphocytes (Fig. 2D) (Weiner et al., 1980a). “Co-capping” studies are an extension of the indirect immunofluorescence approach. These experiments take advantage of the property of physiologically active cells to modulate receptor sites bound by ligand to one pole of the cell (Fig. 2C). Using two different colored fluorochromes (one to mark bound virus and the other to label another cell surface structure) it is possible to determine if a cell surface component moves in association with the viral receptor when viral receptors are modulated to one pole of the cell. This approach demonstrated that EBV receptors on cultured cells were closely associated with complement receptors (Yefenof er al., 1976). Co-capping studies also demonstrated that reovirus receptors on murine lymphocytes are distinct from other surface antigens, such as C3 and Fc receptors (Epstein et al., 1981). Viral particles themselves may be directly fluoresceinated or rhodaminated. McGrath and colleagues have used a direct fluorescence technique to study the binding of MuLV virions to thymic lymphoma cells (McGrath et al., 1978; McGrath and Weissman, 1979). These studies allowed a quantitative as well as a qualitative measurement of binding since the amount of fluorochrome per viral particle could be compared with the total fluorescence per cell, as measured with a fluorescence activated cell sorter (FACS).

4 . Enrichment for Receptor-Positive Cells In an extension of the qualitative approach described above, it is possible to use the virus itself to select from a heterogeneous population those cells which

VIRUSES AND CELL SURFACE RECEPTORS

33

FIG. 2. Binding of reovirus to isolated ependymal cells, lymphocytes, and neurons demonstrated by fluorescence staining. (A and B) Unstained isolated ependymal cells examined by phase microscopy and the same fields seen by fluorescence microscopy showing bright labeling of the cells after incubation with reovirus type 1. Viral binding was demonstrated by indirect immunofluorescence with rabbit antibody to reovirus and with FITC-conjugated goat-anti-rabbit Ig. (C and D) Fluorescence microscopy showing labeling of lymphocytes with reovirus type 3 after capping of the receptors (C). (D) shows the appearance prior to capping. (E and F) Phase contrast photomicrograph of cultured neuronal cells (arrow) and of neuronal cells overlayed with reovirus type 3 and then labeled with FITC-labeled antireovirus antibody. Staining can be seen on the neuronal cell body surface and neuronal processes.

34

MARC TARDIEU ET AL.

bind the virus. For example, cells that have bound virus and then have been labeled with fluoresceinated antiviral antibody can be sorted on the FACS. Another method uses plate-adherence. We have developed the plate-adherence technique for the study of reovirus-receptor interactions. In this technique, cells are first incubated with virus and then with a rabbit antiviral antibody. These cells are then plated on a Petri dish previously coated with a goat-anti-rabbit immunoglobulin and separated into adherent and nonadherent populations. Using the plate-adherence technique, the percentage of reovirus type 3-positive murine splenic T lymphocytes was enriched from 2 1 % in the initial unseparated population to 88% in the plate-adherent population (Epstein et al., 1982).

5 . Radiolabels Radiolabeled virions permit the most quantitative measurement of viral binding, either by measuring loss of radioactivity from media, or more usually, by measuring attachment of labeled virus to cells. Because of impurities in labeled preparations (i.e., labeled particles that do not bind), these two approaches are not identical (Richter, 1976), and the uptake of labeled virus to cells is preferred because it is the more direct approach. Furthermore, measurement of uptake is more accurate. For example, 2% uptake can be more accurately measured than 2% loss from the medium (100 to 98%). For radiolabeled studies, many investigators have grown virus in the presence of 3H- or 14C-labeled amino acids (Lonberg-Holm, 1964; Lonberg-Holm and Whiteley, 1976; Fries and Helenius, 1979). These preparations had specific activities in the range of 1013 viral particles/@ (Lonberg-Holm, 1981), and have been used to study the time course of viral-receptor interactions, and competition for receptors by different viruses. The physical conditions which affect viral binding, such as ionic strength, pH, temperature, and cell concentration have also been studied (Lonberg-Holm and Whiteley, 1976; Lonberg-Holm and Philipson, 1974, 1980). Lactoperoxidase-catalyzed iodination labels viral surface components which contain tyrosine residues (Marchalonis et al., 197 1). Although it is more likely to result in a preparation containing inactivated virions (K. Lonberg-Holm and B. Korant, personal communication; Epstein et al., unpublished data), iodination of virus yields labeled virus with higher specific activities. With reovirus, it is possible to obtain specific activities of approximatively 10l2 particledyci. In general, the greater the specific activity of a ligand, the more accurate the quantitative measurements which can be made. Radiolabeled purified viral attachment proteins (i.e., retrovirus spike glycoproteins or adenovirus fibers) have also been used to study binding to cellular receptors (DeLarco and Todaro, 1976; Choppin et al., 1981). Finally, 1251-labeledprotein A has been used for the indirect study of surface

VIRUSES AND CELL SURFACE RECEPTORS

35

interactions of mouse mammary tumor viruses with mouse and rat cells. Virus bound to the cell surface was detected using monoclonal-antiviral antibodies that then bound labeled protein A (Altrock et a f . , 1981). B . MATHEMATICAL ANALYSIS OF VIRALBINDING The interaction of viral particles with receptors is dependent upon a number of physical conditions, including receptor affinity and density, viral concentration, temperature, pH, and ionic strength of the bathing medium. It is useful to study binding interactions using standard values for these variables and then varying one parameter at a time. Mathematical models for receptor interactions have been derived from equations for enzyme kinetics which were developed by Michaelis and Menten (Lehninger, 1975). This approach assumes reversible bimolecular binding, as represented below by Eq. ( l ) , a condition which is not necessarily true for viral binding to receptors. One approach which we have found useful for characterizing interactions of reovirus particles as ligands (L) with receptor sites ( R ) on cells has been to study binding under equilibrium conditions where a simple reversible bimolecular reaction holds:

where R*L represents viral particles bound to receptors (receptor-ligand complex), and k , and k - represent the forward or association and backward or dissociation rate constants for virus binding to receptors. The equilibrium dissociation constant for the reaction, K d , describes the relative concentrations of these reagents at equilibrium, or the ratio of the rate constants, and is represented by the equation:

,

Kd

=

k - , / k , = [Rl [LlI[RLl,

(2)

where the square brackets represent concentrations. For most binding interactions the concentration of [R] is unknown, and the cell number rather than the number of sites can be manipulated in experiments. With cell number held constant in an experiment, and viral concentration varied, an uptake curve of virus binding to receptors may be obtained (see Fig. 1A). The saturable hyperbolic binding curve in Fig. 1A can be derived from Eq. (2) above, and from Eq. (3), indicating that the total number of receptors, Rtota,,is the sum of the free (R) plus bound (RL) receptors:

+

(3) Rtota, = R RL Using Eq. (3) to eliminate [R] in Eq. ( 2 ) , one may express the number of bound virus receptors [RL] as a function of viral concentration [L] in terms of two

36

MARC TARDIEU ET AL.

parameters, the Kd value and Rtotal,the total number of available binding sites in the preparation:

The concentration at which [ L ] = Kd is a useful point on this curve, since it is the concentration of ligand at which 50% of the total receptor sites are occupied ([RL]= '/2 [R,,,,,]). Thus, Kd, the equilibrium dissociation constant, is a measure of the affinity of a receptor for its ligand, since a small Kd value indicates a tightly binding (high-affinity) ligand which saturates its receptor at a low ligand concentration (see Fig. IA). Hence, Kd is a standard measure used to compare the binding of ligands to a variety of receptors. It is important to note, however, that these equations express Kd in terms of the concentration of free ligand when equilibrium has been reached. At this point, the value of L , the concentration of free ligand, must be very close to the value of Ltota,,since as for R , the value of Ltotal = L LR. Therefore, binding studies are usually performed in a range where only small amounts of free ligand are removed from the solution, usually not more than 10- 15%, so that L can be approximated to equal Ltotal.Otherwise the free ligand concentration must be measured directly at equilibrium. The dissociation and association rate constants for binding, k - and k,, can also be estimated directly by measurement of cell-associated virus as a function of time. Rate equations fork, and k- in Eq. (1) can be derived from the same model of a bimolecular interaction, and can also provide estimates of Kd (see Bennett, 1978, for detail). Another approach to analyze the ligand-receptor interaction shown in Fig. 1A is to transform the data into a form which can be represented as a linear equation. One such form, the Scatchard plot shown in Fig. IB, plots [boundIfree ligand] against [bound ligand]. The purpose of this analysis is to allow the binding parameters to be directly estimated from the linear plot. Rearranging Eq. (4) to this form one obtains the expression:

+

,

,

If the resulting plot is a straight line, the x intercept is the maximum binding of the ligand at saturation [R,,,,,], and the slope is the negative reciprocal of the K d . Of concern, however, is that the manipulation of data for Scatchard analysis propagates errors in uptake measurements to both the x and y axis and changes the relative weighting of various regions of the binding curve. Since measurement of viral binding already has significant uncertainty, this magnification of errors further reduces the accuracy of binding data. Although it is used extensively to analyze ligand interactions in neuropharmacology and endocrinology, the Scatchard plot has not been used in the viral receptor field (see Incardona,

VIRUSES AND CELL SURFACE RECEPTORS

37

1981). Nonetheless, quantitative viral binding data can be expressed using Scatchard analysis (see Fig. 1B).

111. Membrane Components Which Interact with Viruses

A. DENSITYAND AFFINITY OF VIRALRECEPTORS Estimates of the density of cell membrane receptor sites for certain viruses have used either growth assays or radiolabeled binding techniques. Despite the variety of different cell types and viruses used, these estimates have been remarkably consistent, with values in the range of lo4 to los sites per cell. For instance, Lonberg-Holm and Philipson (1974) estimated 1 X lo4 receptor sites for poliovirus on HeLa cells, and found a similar value for adenovirus virions and other enteroviruses on several different permissive cell lines. Similarly, Birdwell and Strauss (1974) arrived at an estimate of lo5 sites per cell for binding of Sindbis virus. Recent studies from our laboratory using 251-labeledreovirions have provided estimates in this range for reovirus type 1 and 3 receptor sites on L-cell fibroblasts, and for reovirus type 3 receptors on murine lymphocytes (Epstein et al., unpublished). McClintock et af. (1980) arrived at a slightly higher estimate (1-5 x los siteskell) for EMC virus binding to HeLa cells. Studies using purified subviral binding components, e.g., the fiber protein of adenovirus, have usually led to higher estimates of receptor density than studies using whole virions. For example, Lonberg-Holm and Philipson (1974) demonstrated a 1 log increase in receptor site density when purified adenovirus fibers were used in place of virions (from lo4 to los sites per cell). Similarly, 5 x los sites per cell were estimated when DeLarco and Todaro (1 976) studied binding of the gp7 1 binding glycoprotein from Rauscher murine leukemia virus to NIH/3T3 cells. Such studies have supported the concept that viral receptor sites consist of multiple receptor “units” which can each bind an individual viral attachment protein, and that viruses can bind to the cell surface in a multivalent fashion. Few studies have quantitatively measured the affinity of viral binding to receptors in terms of the Kd of the equilibrium binding reaction (see Section 11,B). However, it has long been assumed that the virus-cell interaction is of very high affinity, both because of the rapid time course of viral binding and the difficulty in disrupting bound virus by physical means (Lonberg-Holm and Philipson, 1976). Although it has been suggested that in certain cases viral binding is virtually irreversible, some dissociation of bound virus probably occurs for most interactions (Lonberg-Holm, 1981). Rapid penetration after binding is one reason that dissociation has been difficult to measure. Initial studies from our laboratory have estimated very low Kd values for reovirus binding to L cells and

38

MARC TARDIEU ET AL.

lymphocytes, demonstrating extremely high-affinity binding sites for reovirus (Epstein et al., unpublished data). B. RELATIONSHIPOF VIRAL RECEPTORS TO OTHERSURFACE COMPONENTS It is unlikely that the structures on the cell surface which serve as viral receptors evolved merely for the purpose of virus binding. “Viral receptors,” most probably, serve other biological functions for the cell. It is known, for example, that bacteriophage receptors are components of transport systems for low-molecular-weight sugars (Hazelbauer, 1975). Influenza virus receptors on erythrocytes have been extensively characterized (Kathan et al., 1961). These are sialoglycoproteins (called glycophorins) which have multiple functions, including M or N blood group activity (Springer et al., 1966) and lectin binding activity (Jackson et al., 1973). Extensive biochemical characterization and purification of these receptors (see Burness, 1981, for review) and analogous studies on other types of cells have confirmed that sialic acid-containing glycoproteins are important structural features of myxo- and paramyxovirus reception sites. Receptors for the gp7 1 proteins of Rausher MuLV (previously described as the VAP) appear to be lipoproteins, since binding activity to fibroblast membranes is destroyed by protein-denaturing agents or treatments with chymotrypsin or phospholipase (Kalyanaraman et al., 1978). McGrath et al. (1978) have suggested that T-lymphoma cell surface receptors for MuLV might be identical to the T cell antigen-binding receptor. However, a recently isolated 190,000 dalton dimeric protein from thymocytes which retains Maloney MuLV binding activity differs in size from previously reported values for T-cell idiotype receptors (Binz and Wigzell, 1976; Schaffar-Deshayes et al., 1981). It also differs frcm immunoglobulin and Fc receptors which have also been suggested as candidates for the MuLV receptor. In other studies, co-capping experiments have shown a relationship (though not identity) between EBV receptors and receptors for complement components C3b and C3d (Yefenof et al., 1976). Helenius et al. (1978) reported that solubilized receptors for Semliki Forest virus (SFV) were enriched in HLA antigens but subsequent studies by Oldstone et al. (1980) showed binding and growth of SFV in cells lacking HLA antigens, indicating that HLA antigens were not biological receptors for SFV binding. More recently, acetylcholine receptors (AChR) have been proposed to function as rabies virus receptors on mouse muscle cells. This conclusion was based on the similar anatomic distributions of bound virus and acetylcholine antagonists as observed by fluorescence microscopy. In addition, blocking studies showed that pretreatment of cells with AChR blockers (a-bungarotoxin or dtubocurarin) decreased viral replication in susceptible cells (Lentz et al., 1982).

VIRUSES AND CELL SURFACE RECEPTORS

39

It is possible that more than one surface structure might function as a viral receptor under special different circumstances. Daughaday et a!. (1981) studied dengue virus binding to human macrophages and found that, although viral receptors were destroyed by trypsin, addition of specific nonneutralizing antibodies allowed penetration and replication of virus in the cell. They postulated that Fc receptors served as “nonneutralized viral receptors” in this instance by allowing binding of nonneutralized viral-antiviral antibody complexes to the cell via the Fc portion of the immunoglobulin molecule that was attached to the virus.

C. DEFINITION OF CELLULAR RECEPTORUNITS BY MONOCLONAL ANTIBODIES Monoclonal antibodies directed against viral receptors represent a new approach for the characterization of viral cellular receptor sites. The fine specificity of monoclonal antibodies should permit elucidation of the precise cell surface antigenic domains involved in viral binding. Although minimal data using this approach are currently available, it is timely to review some of the technical options and difficulties associated with this approach. One avenue is to screen a panel of monoclonal antibodies raised against the entire cell membrane, in hopes of isolating one which is specific for the viral receptor. These antibodies could be screened for based upon their ability to inhibit viral binding, viral growth, or hemagglutination. We have attempted this approach with reovirus. In order to find an antibody specific for the reovirus type 3 receptor on lymphocytes, a large panel of monoclonal antibodies against murine lymphocytes produced by Springer and colleagues (Springer, 1980) were tested for their ability to bind to lymphocytes which had been enriched for those bearing reovirus type 3 receptors (see Section 11,A). Monoclonal antibodies which showed significantly more binding to viral receptor-positive lymphocytes were then tested for their ability to inhibit the binding of 12%labeled reovirus type 3 to lymphocytes. Only one antibody was found which had some effect: it minimally reduced viral binding. In an analogous approach, Campell and Cords ( 1982) generated monoclonal antibodies against HeLa cells and have identified monoclonal antibodies which block binding of coxsackievirus but not poliovirus to HeLa cells. A second approach we have used takes advantage of the natural regulation of the immune response through the idiotype-antiidiotype network. To explain how this network might function during the normal immune response to a virus, assume that the virus has one antigenic determinant. That determinant will bind to B lymphocytes with an appropriate receptor or “idiotype” on the cell surface. The antiviral antibodies produced by the B lymphocytes bear idiotypic determinants that can bind the virus. These idiotypic determinants are themselves immunogenic and serve as antigens, so that antiidiotypic antibodies are made. These

40

MARC TARDIEU ET AL.

antiidiotypic antibodies do not bind virus, but may bind to the cell surface structure that originally bound the virus. Thus, in theory, they could be used as specific antireceptor antibodies. This approach was first used in endocrinology. Antiidiotypic antibodies were raised against antibodies to retinol binding protein and to insulin. These antiidiotypic antibodies were shown to recognize the cell surface receptor for either retinol binding protein or insulin. Furthermore, it was found that antiidiotypic antibodies made against insulin antibodies mimicked the biological effects of insulin on adipocytes (Sege and Peterson, 1978). We have attempted a similar approach to obtain an antibody against the reovirus type 3 receptor. In these experiments a polyclonal xenogeneic antiidiotype antiserum was made by injecting rabbits with hemagglutinin-specific mouse antireovirus type 3 antibodies (the viral hemagglutinin is the reovirus VAP) (Nepom et af., 1982). The antiidiotype antiserum was then absorbed with normal mouse immunoglobulins, and purified by using a monoclonal antibody to the neutralization site on the hemagglutinin of reovirus type 3. These purified antiidiotypic antibodies mimicked the virus in their binding patterns to various cell lines. Moreover, they bound to primary cultures of murine neuronal cells (which bind reovirus type 3) but did not bind to freshly prepared ependymal cells (which bind reovirus type 1 but not type 3). In addition, they appear to mimic the virus in its interaction with murine T lymphocytes (see Section V,A) (Nepom et al., in preparation; Tardieu et al., 1982). Thus, it appears that these antiidiotypic antibodies might recognize the CRS for reovirus type 3 on neurons and lymphocytes. Work is in progress to further characterize the properties of these antiidiotypic antibodies in terms of blocking viral growth and determining structures on the cell surface which they recognize. Using a similar approach, McGrath and Weissman, studying a spontaneous murine B cell lymphoma (BCLl ) recently produced a monoclonal antiidiotypic antibody against the BCLl-IgM (which binds BCLl -associated retrovirus) and demonstrated that the monoclonal antiidiotypic antibody blocks the binding of the retrovirus to BCLl cells (M. S. McGrath and I. L. Weissman, personal communication). D. SPECIFICITY OF CELLRECEPTOR SITESFOR VIRUSES

In some instances, structures on the plasma membrane which serve as viral receptors have specificity for a single virus, and in other instances, different viruses may share the same receptor. The concept of “viral receptor families” was introduced by Lonberg-Holm et al. (1976). In their experiments, they were able to block the binding of one virus to the cell surface by preincubating the cells with an unrelated virus (see Boulanger and Philipson, 1981, for review). Binding was measured either by infectivity (using UV-inactivated virus for blocking), by radiolabeled virus (using unlabeled virus for blocking), or by

VIRUSES AND CELL SURFACE RECEPTORS

41

immunofluorescence (using specific antiviral antibody which recognized the blocking virus). These studies established that HeLa cells have receptors for a variety of picornaviruses, and that some viruses share common receptors. Thus, the three poliovirus serotypes compete for a common receptor, which is distinct from the receptors that bind the six groups of Coxsackie B viruses (Crowell, 1966, 1976). Coxsackie viruses A13, A15, and A18 have a distinct receptor from the other two groups and Echoviruses and human rhinoviruses have separate receptors from the other picomaviruses (Crowell and Siak, 1978). Moreover, receptors were shown to be shared between viruses from different groups. Thus, the binding of adenovirus type 2 to HeLa cells was blocked by Coxsackie B3, and binding of human rhinovirus type 14 was blocked by Coxsackie virus A21 (Lonberg-Holm ef al., 1976). Confirmation of “receptor families” as defined in these experiments will ultimately depend upon biochemical characterization of the viral receptor sites. The serotype specificity of some viruses is also associated with serotypic differences in their receptors. For example, two serotypes of reovirus, types 1 and 3, differ in their ability to bind to primary cell cultures. Reovirus type 3 binds to neurons and lymphocytes whereas reovirus type I binds to ependymal cells (see below). In addition, binding experiments using 1251-labeledvirions suggest different binding patterns of the two reovirus serotypes to L cells even though both serotypes do bind to and grow in this continuous cell line. (Epstein ef al., unpublished). Specific receptors for the two serotypes of Herpes simplex virus have also been described (Vahlne et a l . , 1979). In these experiments, Herpes simplex type 1 (HSV 1) interfered with the adsorption of HSV 1 but not of HSV2 to human, monkey, and rabbit permanent cell lines. The adsorption rate was measured by assaying infective virus remaining in the medium or by measuring cell associated [3H]thymidine-labeled HSV. Adsorption profiles demonstrated that the monkey kidney cell line and the rabbit cornea cell line had more HSVl than HSV2 receptors, while HeLa cells expressed more receptors with affinity for type 2 than for type 1. Human embryonic lung cells and a cell line derived from a human carcinoma of the larynx showed equal amounts of HSV 1 and HSV2 receptors. Our experiments demonstrate that HSV 1 binds significantly more to murine ependymal cells than HSV2 (Tardieu and Weiner, 1982). A “viral interference” assay was used to study the specificity of cell surface receptors for retroviruses. Steck and Rubin (1966) first demonstrated retroviral interference by showing that chicken fibroblasts persistently infected by an avian retrovirus were not susceptible to superinfection by the same virus (Rubin, 1960, 1961). Later studies established that the interference resulted from a blockade of viral receptors by endogenously produced viruses (reviewed in Weiss, 1981). A similar approach has been used to study another group of retroviruses, murine leukemia viruses. Murine leukemia viruses (MuLV) are classified as ecotropic, xenotropic, or amphotropic depending on their ability to infect mouse

42

MARC TARDIEU ET AL.

cells, cells of other species, or cells of both mice and other species, respectively. Using the “cross-interference” approach, it was shown that cells infected with one ecotropic virus failed to bind a second ecotropic strain (R-MuLV), whereas cells infected with a xenotropic virus bound R-MuLV as well as uninfected cells (Hartley and Rowe, 1976; Besmer and Baltimore, 1977). Thus, murine ecotropic and xenotropic viruses appear to recognize different receptors on the murine cell surface, but various strains of ecotropic MuLV utilize the same receptors, since they are subject to cross-interference (Sarma et al., 1967); DeLarco and Todaro (1976) subsequently showed that infection of murine cells with various ecotropic viruses (S2CL3, AKR, R-MuLV, M-MuLV) prevented the binding of radiolabelled R-MuLV gp71 (the VAP of R-MuLV) to the surface of the infected cells. Interference studies using purified gp7 1 demonstrated that the ecotropic viruses used the same family of receptors despite marked differences in the antigenic properties of the viruses. It was then confirmed that the murine xenotropic, as well as amphotropic viruses, use a different family of receptors from murine ecotropic viruses, since they did not interfere with viral infectivity or gp7 1 binding. Similarly, a class of mouse mammary tumor viruses (MMTV) which share an antigenically similar surface glycoprotein gp52 (the VAP of C3H MMTV and GR MMTV) recognize a common cell surface receptor which is different from the surface receptor recognized by other MMTV which have an antigenically different gp52 (CH3 MMTV and RIII MMTV) (Altrock et al., 1981; Schochetman et al., 1979)). In these experiments, viral binding was studied by measuring binding of 1251-labeledprotein A to immune complexes composed of a C3H MMTV gp52 type-specific monoclonal antibody and receptor bound MMTV, or by directly measuring radiolabeled 3H C3H MMTV binding to the cells. Viruses which share class-specific gp52 determinants also share common surface antigen receptors involved in virus adsorption. Finally, the VAPs of type-C and type-D primate retroviruses recognize the same receptors (Moldow et al., 1979), a finding that might reflect a relationship between type-C and type-D VAPs as suggested by their immunological cross-reactivity (Stephenson et al., 1976; Devare er al., 1978). Most of the above studies were performed using continuous cell lines. It is important, however, to also study viral binding to cells which may be the target of viral infection in vivo. This approach has been most easily implemented using freshly isolated lymphocytes for the study of virus-receptor interaction with lymphocyte subpopulations. For example, Epstein-Ban virus selectively binds to human B lymphocytes, mouse adapted cytomegalovirus binds to murine B lymphocytes, and measles and murine leukemia virus bind to human T lymphocytes (Greaves, 1976). Woodruff and Woodruff (1974) have done a variety of studies on the binding of myxoviruses and paramyxoviruses to murine lymphocytes. They have found that Sendai virus, Newcastle disease virus (NDV),

VIRUSES AND CELL SURFACE RECEPTORS

43

influenza B virus and influenza A (H2N2) or (H3N2) virus agglutinate T lymphocytes in v i m , and, presumably, T lymphocytes have a receptor for these viruses. The receptors on T lymphocytes for the paramyxoviruses they studied (Sendai and NDV) differ from those for the myxoviruses (influenza). This was shown by differences in binding between these group of viruses according to temperature or in vitro treatment with fetuin, N-acetyl neuraminic acid, or periodate. In addition, after elution of NDV from lymphocytes, the lymphocytes are agglutinable by influenza virus but not by NDV or Sendai virus. The nine serotypes of influenza A virus they studied also demonstrated differences in their ability to bind T and B cells: five strains agglutinated T and B cells whereas four agglutinated only B lymphocytes. Thus lymphocyte receptors can distinguish among various serotypes of influenza A virus. Using an indirect immunofluorescence technique, we have found a receptor for reovirus type 3 on murine and human lymphocytes whereas only minimal binding was visualized for reovirus type 1 (Weiner et al., 1980a). These studies have recently been extended by quantitative studies using 1251-labeledvirions which show saturable binding of reovirus type 3 to lymphocytes and only minimal binding of reovirus type 1 with no saturable component (Epstein et al., in preparation). Thus, lymphocytes have different receptors for reovirus serotypes and only the receptor for reovirus 3 has a sufficiently high affinity for the binding to be characterized. More recently, the in v i m affinity of 3H-labeled mouse hepatitis virus 3 for macrophages and lymphocytes from both naturally resistant and susceptible mice was shown to be identical (Krystyniak and Dupuy, 1981). In addition, 3Hlabeled encephalomyocarditis virus bound to resident peritoneal macrophages. In contrast, unstimulated splenic lymphocytes did not have detectable numbers of EMC virus receptors, but these receptors could be induced on both T and B lymphocytes by mitogenic stimulation (Morishima et al., 1982). In addition to lymphocytes and macrophages, other cells of biological interest can be studied. We have a particular interest in viral receptors on nervous system tissue and techniques exist to obtain freshly isolated cells from the central nervous system, such as oligodendrocytes (Snyder et d . , 1980), astrocytes (Farooq and Norton, 1978), ependymal cells (Manthorpe et al., 1977), or in some instances neurons (Farooq and Norton, 1978). Freshly isolated cells can then be used to identify viruses which have an affinity for them. We have initiated this approach to study viral receptors on freshly isolated human and murine ependyma1 cells (Tardieu and Weiner, 1982). In these experiments, viral binding to the ependymal cells was demonstrated by indirect immunofluorescenceusing specific antiviral antiserum. Reovirus type 1 (which induces hydrocephalus in mice) bound to the surface of isolated human and murine ciliated ependymal cells whereas reovirus type 3 (which does not induce hydrocephalus in vivo) did not. The binding property of reovirus type 1 to ependymal cells was then mapped to

44

MARC TARDIEU ET AL.

the viral hemagglutinin (the VAP) with the use of single-segment recombinant clones between reovirus 1 and 3 (Weiner et al., 1980~).Clone 3.HA1, which contains nine genes from reovirus type 3 and one, the S1 gene, which encodes for the viral hemagglutinin, from type 1, bound to ependymal cells, whereas the reciprocal clone 1 .HA3 did not. In addition, mumps virus, measles virus, parainfluenza type 3, and Herpes simplex virus type 1 bound to murine ependymal cells, whereas Herpes simplex virus type 2 and poliovirus type 1 did not. (Further work on CNS viral receptors by McLaren and Holland is described later in this article.) Thus, it can be demonstrated that unrelated viruses may share a common receptor on the cell surface, and that viruses with the same VAP usually bind to the same receptor even if other parts of the virus are different (retroviruses, recombinant clones of reoviruses). On the other hand, different serotypes of a virus may (polioviruses, coxsackie A viruses) or may not share (reoviruses, Herpes simplex viruses, influenza viruses) a common cell surface receptor. Does a virus bind to different cells using identical or different cell receptor sites? This issue is particularly relevant since most studies of viral binding have utilized permanent cell lines. It is not known to what extent there is homology between surface receptors on the different cell lines to which a virus binds, or, more importantly, whether results obtained from binding studies using transformed cells can be generalized to in vivo virus-surface receptor interactions. Early investigations suggested that different structures served as viral receptors on different cell lines (Kodza and Junglebut, 1958; Sabin, 1959; Holland and McLarren, 1961). One experimental approach to address this issue is to compare two different permissive cell lines for such variables as number of receptors per cell, affinity of these receptors for virus, or susceptibility of receptors to inactivation by agents such as proteolytic enzymes. For example, Sindbis virus replicates in both mammalian and mosquito cell lines. Smith and Tignor (1980) studied the attachment of two Sindbis virus strains (avirulent or neurovirulent) to these cell lines both before and after enzyme treatment of the cells. Mammalian cellular receptors for the avirulent strain were sensitive to proteolytic cleavage while mosquito cells were insensitive to protease, phospholipase, and neuraminidase. The difference was less striking but still present for the neurovirulent strain. Reovirus type 3 binds both to freshly isolated lymphocytes and to L cells, a murine fibroblast cell line. Although binding studies using 1251-labeledvirus suggested similarities between receptors on these two types of cells, xenogeneic antiidiotypic antibodies raised against hemagglutinin-specific antireovirus type 3 antibodies (see Section III,C) bind to the same lymphocyte subpopulations as reovirus type 3 but do not bind to L cells (Nepom et al., in preparation; Tardieu et al., 1982). Assuming that this antiidiotypic antibody recognizes the CRS for reovirus type 3 on lymphocytes and neurons, it would appear that the receptor sites on lymphocytes and neurons express an antigenic determinant which is absent from the receptor site on L cells.

VIRUSES AND CELL SURFACE RECEPTORS

45

In an analogous approach, the attachment kinetics of radiolabeled encephalomyocarditis virus were compared on established murine and human cell lines (McClintock et a l . , 1980). The receptor for this virus on human cells had a higher affinity for virus than that on murine cells. In addition, the attachment of the virus to HeLa cells was temperature-independent over the range 0 to 40°C whereas attachment to murine cells progressively decreased with increasing temperature (associated with an increased rate of dissociation of virus). Thus, from these three models, it appears that there may be structural differences between receptor sites for a given virus on various cells. A definitive answer to this issue requires the isolation and characterization of the CRU for a particular virus and the determination of which component of the receptor the virus binds.

OF CELLRECEPTOR SITES E. In Vitro MANIPULATION

The attachment of a virus to a cellular receptor site is only the first step in a series of events (internalization, uncoating, replication, and assembly) that ultimately results in viral replication. The presence of a specific receptor on the cell surface is a necessary but not sufficient condition for viral replication. Thus, for nonpermissive cells, an important question is whether restriction occurs at the receptor or intracellular level. 1. Transfer of Epstein-Barr Virus (EBV) Receptors to Receptor-Negative

cells In vitro, EBV infection occurs only in human and some primate lymphocytes, and EBV receptors are present only on B lymphocytes (Jondal and Klein, 1973; Greaves 1976). To determine if the host-range restriction of EBV growth was receptor-mediated, EBV receptors from purified Raji cell membranes were transferred into the membranes of murine lymphocytes and cells from a human T cell line both of which were nonpermissive for the virus (Volsky et al., 1980). Transfer was accomplished using vesicles reconstituted from a mixture of purified Raji membranes and Sendai virus envelope proteins. Successful implantation of receptor-rich membranes into the membrane of the nonpermissive cells was demonstrated by monitoring the fate of radioiodinated donor membrane, and was confirmed by the detection of surface EBV receptors and complement C3 receptors (which are closely associated with EBV receptor) (Yefenof et al., 1976) on implanted cells. EBV receptors could be detected for 36 hours after implantation and radiolabeled EBV bound specifically to receptor-implanted cells. Furthermore, the implanted receptors were biologically functional, since virus penetration and replication were demonstrated in the normally resistant cells as measured by the expresssion of EBV early nuclear antigen and EBV capsid antigens.

46

MARC TARDIEU ET AL

2. Bypassing Receptor-Mediated Barriers to Infection Replication of a virus can occur in a cell lacking appropriate viral receptors if the barrier to infection at the cell surface is circumvented. This has been achieved with poliovirus by two methods: (1) direct inoculation of viral nucleic acid into the cytoplasm of a cell lacking poliovirus receptors and (2) physical entrapment of the virus into the cell by fusion of the cellular membrane with Sendai virus (which incorporate polioviruses bound nonspecifically to the cell surface) or with virus-containing liposomes (Enders et a l . , 1967; Wilson et al., 1977). 3. Binding of Polyoma and Sendai Virus to Specific Gangliosides Many studies have demonstrated the importance of sialic acids in the binding of Polyoma or Sendai virus (and other myxo- and paramyxoviruses) to the cell surface, and sialidase treatment of host cells can prevent viral infection with these viruses (Hirst, 1942; Klenk et a l . , 1955). These cellular receptors have been identified as glycoproteins with N-acetyl neuraminic acid as the terminal sugar in the carbohydrate side chains (Gottshalk, 1957). To further elucidate the role of sialic acid in cell permissiveness, binding of virus to isolated, highly purified gangliosides of defined structure was studied (Svennerholm and Fredman, 1980). Initially, binding to polystyrene Petri dishes coated with different gangliosides was studied (Holmgren et al., 1980), then host cells were made resistant to Sendai virus by removal of endogenous viral receptor with Vibrio cholerae sialidase (Markwell and Paulson, 1980). These receptor-negative cells were then incubated with individual purified gangliosides. Incubation of cells with gangliosides containing the sequence NeuAccu2, 3 GalPl , 3GalNAc fully restored susceptibility to Sendai virus infection. Furthermore, incubation with gangliosides with a sequence ending with two sialic acids in a NeuAca2, 8NeuAc linkage, rather than a single sialic acid, was 100 times more effective (Markwell et al., 1981). In an analogous way, susceptibilityto Polyoma virus infection was restored by implantation of the sequence NeuAca2, 3GalP1, 3GalNAc but not the sequence NeuAca2, 6GalP 1, 4GlcNAc even though the latter sequence contained a comparable amount of sialic acid (Fried et al., 1981). Thus, Sendai and Polyoma virus interact with specific ganglioside sequences and cell susceptibility to infection can be modified by implantation of different gangliosides into the cell membrane. 4. Inhibition of Receptor Binding Using Antireceptor Antiserum The attachment of enteroviruses to HeLa cells can be inhibited by heterologous antiserum raised against HeLa cells, suggesting that these antibodies in some way affect cell surface viral receptors (Quersin-Thiry, 1958; Axler and Crowell, 1968; Much and Zajac, 1973). This approach is limited, however, by the lack of fine specificity of the antiserum. Monoclonal antibodies directed

VIRUSES AND CELL SURFACE RECEPTORS

47

against the cell surface offer a more specific avenue for the generation of antireceptor antibody and recent work showing the blocking of Coxsackie virus binding to HeLa cells by monoclonal antibodies demonstrates that this should be a feasible approach (Campell and Cords, 1982). A similar approach has recently been described for monoclonal antibodies made against the thyrotropin receptor. These monoclonal antibodies competitively block binding of thyroid stimulating hormone (TSH) but are unable to stimulate adenyl cyclase activity as TSH does. This result suggests the existence of a second domain on the receptor which is associated with the stimulating activity (Yavin et al., 1981). 5. Modification of the Receptor Associated with Viral Binding It has recently been shown that binding of insulin can alter the conformation of insulin receptors on fat cells (Pilch and Czech, 1980). A similar question can be raised for viral receptors: can viral binding to the cell surface modify its own receptor? There are only a few investigations related to this issue. Following the incubation of human lymphoblastoid cells with Epstein-Barr virus (Hinuma et al., 1975), or of Ehrlich ascites tumor cells with mengovirus (Geschwender and Traub, 1979), modulation of cell-surface viral receptors (capping) was observed. This occurred after binding of virus alone. In our investigations using reovirus, capping of reovirus type 3 receptors on the surface of murine lymphocytes required cross-linking by antiviral antibody (Epstein et a l . , 1981). Levanon et al. (1977) have demonstrated that adsorption of infective encephalomyocarditis virus enhances fluidity of the plasma membrane. Binding of paramyxo- and orthomyxoviruses to the cell surface can result in destruction of the receptor itself. The hemagglutinin of paramyxoviruses binds to neuraminic acid-containing cell surface receptors and has neuraminidase activity which eliminates natural neuraminic acid-containing receptors from infected cells (reviewed by Choppin and Scheid, 1980). These two functions reside on two separate proteins in orthomyxoviruses. The role of neuraminidase activity, which is paradoxically present on the same viral protein that determines viral binding, is unclear. It has recently been shown that these two opposing activities can be regulated by environmental conditions such as chloride concentration and pH: high concentrations of halide ion enhance hemagglutinating activity and decrease elution from erythrocytes, while they inhibit neuraminidase activity (Merz et al., 1981). Studies with a specific chemical inhibitor of neuraminidase (Palese et al., 1974a; Palese and Compans, 1976) and temperature-sensitive mutants (Palese et al., 1974b) suggest a role for neuraminidase activity during the release of newly synthesized virus. Because of enhancement by low chloride ion concentration and an acidic pH, the neuraminidase activity is most prominently expressed intracellularly . In contrast, the ionic environment in the extracellular fluids favors virus attachment over receptor-destroying activity.

48

MARC TARDIEU ET AL.

F. AGE DEPENDENCY OF VIRALPERMISSIVENESS The age dependency of susceptibility to viral infection is relevant to the question of the role of receptors in the nonpermissiveness of a cell. In several instances, cells from adult animals are not as permissive for viral replication as cells from newborn animals, even though they appear to have the appropriate viral receptors. For example, allantoic sac cells from young chicken embryos are more permissive for influenza or vesicular stomatitis viruses than cells from older embryos. This age-dependent difference in permissiveness is due to a late intracellular event since viral attachment and penetration are the same in cells from young and old chicken embryos (White, 1959; Morahan and Grossberg, 1970). In our investigations, we have found that both isolated newborn and adult ependymal cells have receptors for reovirus type 1 even though ependymitis is observed more prominently in newborns (Tardieu and Weiner, 1982). In contrast, in studies of Coxsackie B5 infection of fibroblasts, Kunin (1962) reported that a slightly decreased ability to absorb the virus occurs in adult as compared to newborn cells, correlating with a decreased permissiveness of adult cells for viral replication. In this instance, the age-dependent reduction in permissiveness may be related to a change in receptor affinity for virus in older cells.

G. GENETIC CONTROL OF CELLRECEPTOR SITEEXPRESSION

In a few cases, a genetic basis for the expression of viral receptors on different cell types can be demonstrated. For example, cells from different mammalian species differ in their susceptibility to poliovirus, i.e., human cells are susceptible while murine cells are not. Somatic hybrids made between permissive (human) and resistant (rodent) cells (Belehradek and Barski, 1969; Wang et al., 1970) demonstrated that hybrids could be infected by poliovirus only when human chromosome 19 was present (Miller et al., 1974). Since viral replication does not require the presence of human genes once the viral nucleic acid has entered the cell (Holland et al., 1959; Wang et al., 1970), these experiments demonstrated that chromosome 19 carries the structural gene for the poliovirus receptor. On the other hand, permissiveness of cells for echo-7 and rhino-1A viruses could not be linked to the presence of a specific human chromosome (Miller et al., 1974). The genetic basis for the specificity of retrovirus cell surface receptors has recently been reviewed (Weiss, 1981). Utilizing interspecies somatic cell hybridization techniques, the gene encoding for the CRS for ecotropic MuLV on murine cells has been assigned to chromosome 5 (Oie et al., 1978; Ruddle et af., 1978; Marshall and Rapp, 1979) and the gene encoding the CRS for endogenous feline C-type virus (RDll4) on human cells assigned to human chromosome 19 (Schnitzer et al., 1980). It should be noted, however, that, for MuLV, the gene

VIRUSES AND CELL SURFACE RECEPTORS

49

expressed on chromosome 5 does not solely regulate leukemia virus replication. Another gene (Fv-I), located on murine chromosome 4, regulates the replication of the virus at a postpenetration step and may influence integration of proviral DNA into the host chromosome (Rowe and Sato, 1973; Gazdar e t a ] . , 1977). AS noted by Ruddle et a / . ( 1 978), additional host control is exerted at the level of differentiation, since bone marrow-derived but not thymus-derived lymphocytes are able to support replication of exogenous MuLV. The genetic regulation of receptor expression is, therefore, only one part of the genetic regulation of cell permissiveness for the virus.

IV. Viral Components Which Recognize Cellular Receptors Identification of the subviral components which are responsible for binding of viruses to cell surfaces has preceded structural understanding of the cellular receptors themselves. This section briefly summarizes current data concerning the viral attachment protein (VAP) of selected viruses. A . PICORNAVIRUSES

Picomaviruses are small nonenveloped viruses with icosahedral symmetry. The 22- to 30-nm capsid contains 60 copies of a “structural unit” consisting of four separate polypetides, VPI , 2, 3, and 4. Some picornaviruses contain a few copies of a precursor protein (VPO) which contains uncleaved VP2:VP4. It has been known for many years that shortly after poliovirus binds to cells, a fraction of the attached virus elutes (Halperen et a / ., 1964). This eluted virus has lost the polypeptide VP4 and is no longer infectious (Lonberg-Holm and Philipson, 1974). These experiments suggested that VP4 was the attachment protein. Further studies, however, provided contradictory evidence: ( I ) VP4 could not be labeled by techniques which label surface proteins such as ‘2sl-labeling using lactoperoxidase (e.g., Lonberg-Holm and Butterworth, 1976). (2) Naturally occuring empty capsids (top component) of some picornaviruses demonstrated identical binding characteristics as native virus, but were shown to lack VP4. (3) Antibodies to VP4 do not recognize the surface of native virions (Talbot et al., 1973). Other evidence suggests that VPl is the binding protein. VP1 is expressed on the surface of the capsid, and trypsin treatment of virions (which renders them incapable of binding to cells) appears to primarily cleave VPI , although some studies also have shown cleavage of VP3 (Boulanger, 1975; Boulanger and Lonberg-Holm, 1981). For Coxsackie virus B3 there is also evidence that VP2 is present at the capsid surface (Philipson el ul., 1973). Some investigations have suggested that no single protein functions as the viral attachment site, but that cooperative interactions among the viral proteins result in a unique conforma-

50

MARC TARDIEU ET AL.

tional state which allows binding. This view is supported by the demonstration of differences in antigenicity between native virions ( “D antigenic”) and inactive subviral particles ( ‘‘C antigenic”) for several enteroviruses and rhinoviruses. Native particles carry few of the antigenic determinants of inactive particles (Lonberg-Holm and Yin, 1973). Since the ability to attach to cells is irreversibly lost during the conversion from D- to C-antigenicity (Lonberg-Holm and Yin, 1973), this suggests that the conformation of capsid polypeptides may play a key role in the ability of the virion to attach to host cells, perhaps by regulating exposure of a polypeptide sequence carrying the determinants required for binding activity.

B. ADENOVIRUSES The adenoviruses (mammalian and avian) are larger and more complex than picornaviruses, and (for mammalian viruses) are classified according to speciesspecific hemagglutination properties. Adenovirus capsids are icosahedral and contain 252 capsomers. Most of these are called hexons because each has 6 neighbors. The 12 apical capsomers are surrounded by only 5, and are therefore called pentons. Each of these consists of a penton base and a 10- to 30-nm projection called a fiber. The fiber consists of three polypeptide chains. It has been shown that adenovirus binds to cells via determinants located in the terminal knob of the fiber. The fiber is also the hemagglutinin (Norrby et al., 1969). The ability to solubilize and purify the adenovirus fiber protein has led to the demonstration of serotypic differences in hemagglutinin among subgroups of adenoviruses. C. REOVIRUSES Reoviruses are nonenveloped viruses consisting of two concentric icosahedral capsid shells that surround a segmented double-stranded RNA genome. The outer capsid is composed of three polypeptides ( p l C , a3, and al) which are individually coded by three different viral genes. The a 1 polypeptide makes up I-2% of the outer capsid (24 copies per virion) and is located at the vertices of the icosahedral structure. The a1 polypeptide is the major determinant of reovirus interactions with cells. It is the viral hemagglutinin, elicits the formation of neutralizing antibody, and is responsible for development of delayed type hypersensitivity, generation of suppressor T cells, and generation of cytolytic T lymphocytes (Weiner and Fields, 1977; Weiner et al., 1980b; Greene and Weiner, 1980; Fontana and Weiner, 1980; Finberg et al., 1979). As described previously, it also determines the serotype specificity of viral tropism for different cells in the nervous system and the ultimate pattern of CNS virulence (Weiner et al., 1977, 1980~;Tardieu and Weiner, 1982). Tryptic peptide analysis of the

VIRUSES AND CELL SURFACE RECEPTORS

51

a1 polypeptide from reovirus serotypes 1 , 2 , and 3 has demonstrated both unique and common methionine- and tyrosine-containing peptides. This suggests that certain regions of the hemagglutinin have been conserved, while others have “drifted” with resultant serotypic changes (Gentsch and Fields, 1981). Monoclonal antibodies prepared against the viral hemagglutinin of reovirus type 3 have defined at least four antigenically different domains. One class of antibodies had neutralizing activity; a second class only hemagglutination inhibition (HI) activity. One monoclonal antibody had neutralizing and HI activity and a fourth class of monoclonal antibodies had no detectable neutralization or HI activity. These results suggest that marked functional specialization exists within regions of the reovirus type 3 hemagglutinin (Burstin et al., 1982). This separation of regions for hemagglutination and neutralization raises the possibility that, for other viruses as well, there might be separate viral determinants which bind to either red cells (hemagglutination) or host cells (infection). This suggests that RBC receptors and receptors on other cell types may not be homologous.

D. MYXOVIRUSES AND PARAMYXOVIRUSES For paramyxoviruses, the two glycoproteins which project from the viral surface have been isolated and purified (see Scheid, 1981, for review). The HN glycoprotein, which possesses both hemagglutinating and enzyme (neuraminidase) activity, is the viral receptor-binding protein and exists on the surface in a dimer configuration. It has been suggested that a single active site serves both functions, but this remains to be clarified (Scheid et al., 1972). The F glycoprotein is responsible for fusion activity (and thus hemolysis) and is involved in virus penetration into the cell (to be discussed in the next section). Morbilii viruses (measles, canine distemper virus) lack neuraminidase and their binding glycoproteins are designated H, rather then HN. The three serotypes of influenza virus comprise the myxovirus group and, unlike paramyxoviruses, influenza virions contain separate spikes for the hemagglutinin (HA) (present as a trimer) and the neuraminidase (NA) (present as a tetramer) (Schild, 1979). Influenza C virions differ as they have no neuraminidase. HA is the glycoprotein responsible for hemagglutination or adsorption to host cells and antigenic variations in this protein are largely responsible for periodic epidemics of influenza. Monoclonal antibodies raised against the HA of influenza A have identified three or four nonoverlapping antigenic domains on the protein (Wiley et al., 1981; Wilson et al., 1981). Direct correlation of these domains with functional differences have yet to be defined, but the structural definition of the HA has provided initial answers. The host-receptor binding site and antigenic determinants are located on a globular region which lies on top of a long fibrous coiled coil; the fusion activation peptide is located

52

MARC TARDIEU ET AL.

near the virus membrane end of the molecule (Laver and Valentine, 1969; Gerhard et al., 1981; Wiley et al., 1981; Wilson et al., 1981). E. RETROVIRUSES The viral attachment proteins of several retroviruses have been isolated and used in two types of experiments: (1) cross-interference experiments as discussed in Section III,D (reviewed by Weiss, 1981) and (2) experiments to define antigenic domains on this molecule with monoclonal antibodies. These latter experiments were performed using the major external glycoprotein (gp52) of mouse mammary tumor virus (MMTV). Two topographically distinct sites have been identified on gp.52. One site functions as a target for neutralization antibody and was defined by the observation that all monoclonal antibodies (MAb) which neutralized virus infectivity also competed for binding of a neutralizing MAb (used as a standard). The second site bound antibody but this binding had no effect on neutralization. This site was topographically distinct and its MAb could not compete for binding of a second neutralizing MAb (Massey and Schochetman, 1981a). It was further shown that the neutralizing site described above was not the receptor binding site but was adjacent to it as monoclonal antibodies were found which competed for the binding of the first neutralizing MAb but did not neutralize the virus. These antibodies functioned as blocking antibodies and protected virus particles from neutralization (Massey and Schochetman, 1981b).

F. CORONAVIRUSES Two glycoproteins are associated with the envelope of the A59 strain of mouse hepatitis virus (MHV): the E2 glycoprotein which makes up the peplomers of the virus and the El glycoprotein which is deeply embedded in the viral membrane (the portion of the El glycoprotein which protrudes from the viral membrane contains a small glycosilated portion) (Holmes et al., 1981). Monospecific antibodies directed against E2 glycoprotein prevent viral attachment. Virions lacking E2 (either because of growth in the presence of tunicamycin or treatment with bromelain) do not attach to the cell membrane. In addition, isolated E2 binds to the same receptor as intact virus since pretreatment of cells with unlabeled, concentrated MHV blocks the binding of raidolabeled E2. Thus, the E2 glycoprotein appears to be the virus attachment protein for the A59 strain of MHV (K. Holmes, personal communication; Holmes et al., 1981).

V. Virus-Receptor Interactions and Pathogenicity A major feature of certain viral infections is selective damage to specific tissues, and in some instances to specific cells within a tissue. The classic

VIRUSES AND CELL SURFACE RECEPTORS

53

example is poliovirus infection of anterior horn cells in the spinal cord. Other examples include selective infection of the limbic system by rabies virus, infection of ependymal cells by mumps virus (reviewed in Johnson, 1980), and infection of pancreatic beta cells by encephalomyocarditis virus or Coxsackie B4 virus (Craighead and McLane 1968; Boucher and Notkins, 1973; Yoon et al., 1980). It has long been postulated that the in vivo tropism of viruses is mediated in part by the presence or absence of specific receptor sites for viruses on the surface of the target cells (Holland, 1961).

A. ROLEOF CELLSURFACE RECEPTORS IN PATHOGENICITY The role of specific cellular receptors as determinants of cell tropism has been extensively studied (recently reviewed by Crowell and Landau, 1979; and Crowell et al., 1981). Initial studies with picornaviruses using organ minces and homogenates demonstrated a correlation between the presence of receptors on cells and the known in vivo tropism of poliovirus. Human and monkey CNS tissue and intestine were able to adsorb polioviruses whereas tissues from human lung, heart, and skin were not. The correlation was not absolute, however, since receptors were also detected on human liver, monkey heart, and skeletal muscles. Furthermore, poliovirus vaccine strains which did not induce cell damage were shown to bind to brain tissues (McLaren et al., 1959; Holland, 1961; Kunin and Jordan, 1961; LaPlaca, 1963; Harter and Choppin, 1965). A second line of evidence demonstrating a relationship between cell surface receptors and pathogenicity was the presence of a correlation between the grouping of viruses by receptor specificities and their classification according to subgroups which were derived from patterns of pathogenesis (see Section III,C and Lonberg-Holm et al., 1976). Finally, specific organ cultures have been used to show different growth specificities for picornaviruses. For example, some rhinoviruses multiply only in differentiated organ cultures of trachea (Hoorn and Tyrell, 1966); Coxsackie viruses A1 and A5 grow in differentiating primary fetal mouse muscles cultures but do not grow in nondifferentiating mouse cultures (Came and Crowell, 1964; Landau et al., 1972). In contrast, receptors for human enteroviruses exist on tissues which are not involved in their pathogenesis and in species other than their natural hosts (Holland, 1961; Kunin and Jordan, 1961; LaPlaca, 1963; Campbell, 1965). A genetic approach has been used to define the molecular basis for the different patterns of virulence and central nervous system cell tropism exhibited by reovirus serotypes 1 and 3. Using recombinant clones derived from crosses between reovirus types 1 and 3, it has been shown that the hemagglutinin of reovirus (encoded by the S1 gene) determines the central nervous system cell tropism of the reovirus serotypes (Weiner et al., 1977, 1980~).Reovirus type 3

54

MARC TARDIEU ET AL.

and clone 1.HA3 (containing nine genes from type I and the gene encoding the hemagglutinin from type 3) cause a fatal encephalitis in newborn mice with neuronal destruction but no ependymal cell damage whereas reovirus type 1 and clone 3.HAl (the reciprocal clone to clone 1.HA3) cause ependymal infection without neuronal damage. The affinity of the two serotypes for two different cell types in the nervous system appears to be due to the specific interaction of the viral hemagglutinin with the receptors on the surface of either ependymal cells or neuronal cells. These results have been confirmed in vitro by demonstrating that reovirus type 1 and clone 3.HA1 (but not reovirus type 3 and clone 1.HA3) bound to isolated human and murine ependymal cells (Fig. 2A and B) (Tardieu and Weiner, 1982). The reciprocal results have been shown on neural cells in culture (Dichter and Weiner, unpublished data; see Fig. 2E and F). The M variant of encephalomyocarditis virus (EMC) produces a diabetes-like syndrome in certain strains of mice by infecting and destroying pancreatic beta cells. Cultured pancreatic beta cells from mice resistant to EMC-induced diabetes are less able to absorb infectious EMC virus than beta cells from susceptible strains, suggesting that genetically determined differences in surface viral receptors on these cells may be one of the factors controlling susceptibility to the disease (Chairez et al., 1978). The presence of virus receptors on lymphocytes may correlate with the specific effect that some viruses may have on the immune response. T lymphocytes have a receptor for measles virus and measles infection is associated with a depression of tuberculin skin hypersensitivity, and a suppression of helper cell activity (McFarland, 1974). Reovirus type 3 binds primarily to the Ly2,3 subset of murine T lymphocytes (the suppressorkytotoxic subset) as well as to the human counterpart (T8+ cells). This binding is a property of the viral hemagglutinin (Epstein et al., 1982). Furthermore, in vitro, reovirus type 3 induces suppressor T cells capable of suppressing Con A proliferation (Fontana and Weiner, 1980). This, the generation of functionally active suppressor T cells in vitro by reovirus type 3 appears to be secondary to the interaction of the viral hemagglutinin with a specific receptor on the Ly2,3 subset of murine lymphocytes. B. ROLEOF VIRUS ATTACHMENT PROTEINS IN PATHOGENICITY The specificity of myxo- and paramyxoviruses for particular cell types depends both on the structural and functional activity of the viral surface glycoproteins and on the ability of the cells to cleave these proteins (reviewed by Choppin and Scheid, 1980). The interaction between a paramyxovirus and the cell surface is mediated by two glycoproteins projecting from the external surface of the virion: the hemagglutinin-neuraminidase (HN) and the fusion (F) glycoprotein. Binding to cellular neuraminic acid-containing receptors is a property of the HN

VIRUSES AND CELL SURFACE RECEPTORS

55

glycoprotein. Although this activity differs considerably according to the amount and position of neuraminic acid in the molecule (Markwell et al., 1981), the abundance of sialic acid in many biologic membranes limits the importance of the binding step as a determinant of cell specificity and tissue tropism. An active fusion (F) glycoprotein, however, is a requisite for infectivity and cell-to-cell spread of infection. The fusion glycoprotein acts during viral penetration, a step beyond adsorption. Virus infectivity requires cleavage of a precursor FO glycoprotein into two subunits (Fl, F2) and the host must provide the enzyme responsible for this cleavage (reviewed in Choppin and Scheid, 1980). Thus, host-dependent cleavage of FO is required for infectivity and therefore host range and tissue tropism of virus is determined by availability of the appropriate protease (Scheid and Choppin, 1975, 1976). It has recently been shown that neuraminidase (NANase) activity of the HN glycoprotein of mumps virus contributes to cytopathology . Although the HN glycoproteins of the six studied strains of mumps virus are similar in size and antigenic composition, each strain possesses a neuraminidase with distinct enzymatic properties. Strains with active NANase cause little cytopathology and no cell fusion on African green kidney cell lines, whereas infection with strains having less active NANase cause extensive cell fusion. Thus, viral NANase appears to contribute to full expression of the activity of the F protein and ultimately to cytopathology (Merz and Wolinsky, 1981). This extends the previous observation that influenza virions with less active NANase cause more cytopathology in tissue culture and were more pathogenic in viva than virions containing active NANase (Smith and Cohen, 1956; Choppin, 1963; Choppin and Tamm, 1964). Moreover, only strains of mumps virus with less active NANase were both neuroinvasive and neurovirulent (Wolinsky and Stroop, 1978; McCarthy et af., 1980). Thus, the pathogenicity of myxo- and paramyxovirus depends upon an interaction of viral glycoproteins and the cell surface at a step beyond viral adsorption. Cell specificity is determined by the availability of a protease on the surface of the cell to cleave one of the viral glycoproteins and allow viral penetration into the cell. There are few data concerning the role of viral receptor interactions in the pathogenicity of other enveloped viruses. C. INDUCTION OF CELL-SPECIFIC AUTOIMMUNITY FOLLOWING VIRAL INFECTION

Autiommune reactions against host tissue have been reported after certain viral infections. These include the production of autoantibodies against a variety of host antigens in experimental animals and man (DNA, lymphocytes, myelin) (reviewed in Onodera et af., 1981), and in man, the well-documented immunemediated damage to peripheral nerve myelin in infectious polyneuritis (Guillain-

56

MARC TARDIEU ET AL.

Barre Syndrome) which can occur 2-3 weeks after viral infections or following swine flu immunization (Schonberger et al., 1981). The mechanisms by which a viral infection can lead to an autoimmune response are not well understood, however, two possible mechanisms are relevant to the present review: (1) autoantibodies which appear following viral infection may recognize shared antigens between a viral protein and a determinant on the surface of the target cell; and (2) through the idiotypic-antiidiotypic network (described in Section II,C) antiidiotypic antibodies could be produced which recognize the viral receptor on the cell surface. Thus, the affinity of these autoantibodies for a particular cell would be identical to the tropism of the virus itself for the cell. To test these two hypotheses, we recently performed the following experiment: splenic lymphocytes from adult mice infected with purified reovirus (type 1 or 3) particles were fused with NS-1 myeloma cells. The resultant clones were then screened by radioimmunoassay for their ability to bind virus, T lymphocytes, brain, liver, and lung tissues. We found that (1) during the course of the normal immune response to reovirus, autoantibodies were generated which reacted with normal tissue, (2) monoclonal antibodies were generated which identified shared antigenic structures between viral determinants and normal tissue, and (3) some monoclonal antibodies appeared to have the same affinity for cells as the virus (putative antiidiotypic antibodies which recognize viral receptors) (Tardieu et al., 1982). In another group of experiments, performed by Onodera et u1. (1981), it was shown that mice infected with reovirus type 1 developed transient diabetes and a runting syndrome. Sera of infected mice contained autoantibodiesthat, by immunofluorescence, reacted with cytoplasmic antigens of the islets of Langerhans and anterior pituitary of uninfected mice, both target structures of the virus. The autoantibodies appeared to be directed against insulin or growth hormone. Since reovirus type 3 did not induce autoantibodies to growth hormone, using recombinant clones, it was possible to show that the ability to induce autoantibodies to growth hormone was a property of the viral hemagglutinin (Onodera et al., 1981).

VI. Conclusion Further progress in the study of virus-receptor interactions should occur in the following three areas: ( I ) the development of more sophisticated approaches for both quantitative (e.g., rigorous binding studies using radiolabeled virus with high specific activity) and qualitative (e.g., cell sorting techniques) measurements of viral interactions with the cell surface; (2) the production of monoclonal antibodies against cell receptors and against viral components. These reagents will lead to the isolation and biologic characterization of both the CRU and the

VIRUSES AND CELL SURFACE RECEPTORS

57

functional domains on the virus attachment protein; and (3) techniques for the isolation of single cell suspensions from organs such as brain. This approach will allow direct study of viral interactions with biologically relevant cells. The comparison of viral-receptor studies on permanent cell lines with studies utilizing freshly isolated cells is important since there is increasing evidence that different structures might serve as receptor sites for a virus on cells of different origins. The role of receptors in determining the in vivo affinity of certain nonenveloped viruses for specific cell types and thus determining viral pathogenicity is well established. It must be emphasized, however, that a cell is not permissive for a virus merely because it has a cell surface receptor to which the virus binds. Finally, although additional studies are needed, receptors may play a less important role in the pathogenicity of enveloped viruses than for nonenveloped viruses.

ACKNOWLEDGMENTS We want to thank Dr. K. Lonberg-Holm for critically reviewing the manuscript. MT is the recipient of a Lilly International Fellowship. This work was supported by NIH grant No. NSAI- 16998.

REFERENCES Altrock, B. W., Arthur, L. O., Massey, R. J., and Schochetman, G. (1981). Virology 109, 257. Axler, D. A,, and Crowell, R. L. (1968). J. Virol. 2, 813. Bankhurst, A. D., Maki, D., Sanchez, M., and McLaren, L. (1979). Infect. Irnrnun. 24, 65. Belehradek, J., Jr. and Barski, G. (1969). C.R. Acad. Sci. Paris 269, 672. Bennet, J . P., Jr. (1978). I n “Neurotransmitter Receptor Binding” (H. Yamamura, S. J. Enna, and M. J. Kuhar, eds.), pp. 57-90. Raven, New York. Besmer, P., and Baltimore, D. (1977). J . Virof. 21, 965. Binz, H.,and Wigzell, H. (1976). Cold Spring Harbor Symp. Quanr. Biol. 41, 275. Birdwell, C. R., and Strauss, J. H. (1974). J. Virol. 14, 672. Boucher, D. W., and Notkins, A. L. (1973). J. Exp. Med. 137, 1226. Boulanger, P. (1975). J. Virof. 16, 1678. Boulanger, P. and Lonberg-Holm, K. (1981). I n “Virus Receptors. Part 2: Animal Viruses’’ (K. Lonberg-Holm and L. Philipson, eds.), pp. 21-46. Chapman & Hall, London. Boulanger, P., and Philipson, L. (1981). I n “Virus Receptors. Part 2: Animal Viruses” (K. Lomberg-Holm and L. Philipson, eds.), pp. 119-139. Chapman & Hall, London. Boulanger, P., Houdret, N., Scharfman, A,, and Lamay, P. (1972). J. Gen. Virof. 16, 429. Bumess, A. T. H. (1981). In “Virus Receptors. Part 2: Animal Viruses” (K. Lonberg-Holm and L. Philipson, eds.), pp. 63-84. Chapman & Hall, London. Burstin, S . J., Spriggs, D.R., and Fields, B. N. (1982). Virology 117, 146. Came, P. E., and Crowell, R. L. (1964). Virology 23, 542.

58

MARC TARDIEU ET AL.

Campbell, C. H. (1965). J. Exp. Med. 121, 69. Campell, B. A., and Cords, C. E. (1982). Absrr. Annu. Meet. Am. SOC. Microbiol. p. 257. Chairez, R., Yoon, J. W., and Notkins, A. L. (1978). Virology 85, 606. Choppin, J., Shaffer-Deshayes, L., Debre, P., and Levy, J.-P. (1981). J . Immunof. 126, 2347. Choppin, P. W. (1963). Virology 21, 342. Choppin, P. W., and Scheid, A. (1980). Rev. Infect. Dis. 2, 40. Choppin, P. W., and Tamm, I. (1964). Cibu Found. Symp. C e f f .Biol. Myxovirus Infect. pp. 218-245. Craighead, J. E., and McLane, M. F. (1968). Science 162, 913. Crowell, R. L. (1966). J. Bucteriof. 91, 198. Crowell, R. L. (1976). In “Cell Membrane Receptors for Viruses, Antigens and Antibodies, Polypeptides, Hormones and Small Molecules” (R. F. Beers and E. G. Basset, eds.), pp. 179-202. Raven, New York. Crowell, R. L., and Landau, B. J. (1979). In “Receptors and Human Diseases” (A. G. Beam and P. W. Choppin, eds.), pp. 1-33. Josiah Macy Jr. Foundation, New York. Crowell, R. L., and Siak, 3.-S. (1978). In “Perspectives in Virology” (M. Pollard, ed.), Vol. X,pp. 39-53. Raven, New York. Crowell, R. L., Landau, B. J., and Siak, J . S . (1981). In ‘‘Virus Receptors. Part 2: Animal Viruses” (K.Lonberg-Holm and L. Philipson, eds.), pp. 171-184. Chapman & Hall, London. Daughaday, C. C., Brandt, W. E., McCown, J. M., and Russell, P. K. (1981). Infect. Immun. 32, 469. DeLarco, J. E., and Todaro, G. J. (1976). Cell 8, 365. Devare, S. G., Hanson, R. E., Jr., and Stephenson, J. R. (1978). J. Virol. 25, 797. Enders J. F., Holloway, A., and Grogan E. A. (1967). Proc. Nutl. Acud. Sci. U.S.A. 57, 637. Epstein, R. L., Powers, M. L., and Weiner, H. L. (1981). J . Immunof. 127, 1800. Epstein, R. L., Powers, M. L., Finberg, R., and Weiner, H. L. (1982). Fed. Proc. Fed. Am. SOC. Exp. Biof. 41, 568. Farooq, M., and Norton, W. T. (1978). J. Neurochem. 31, 887. Finberg, R., Weiner, H. L., Fields, B. N., Benacemaf, B., and Burakoff, S . J . (1979). Proc. Nurl. Acud. Sci. U.S.A. 76, 442. Fontana, A., and Weiner, H. L. (1980). J. Immunol. 125, 2660. Fried, H., Cahan, L. D., and Paulson, J. C. (1981). Virology 109, 188. Fries, E., and Helenius, A. (1979). Eur. J. Biochem. 97, 213. Gazdar, A. F., Oie, H., Lalley, P., Moss, W. W., Minna, J. D., and Francke, U. (1977). Cell 11, 949. Gentsch, J. R., and Fields, B. N. (1981). J. Virol. 38, 208. Gerhard, W., Yewdell, J., and Frankel, M. E. (1981). Nature (London) 290, 713. Geschwender, H. H., and Traub, P. (1979). J. Gen. Virol. 42, 439. Gomatos, P. J., and Tamm, 1. (1962). Virology 17, 455. Gottschalk, A. (1957). Biochim. Biophys. Acru 23, 645. Greaves, M. F. (1976). Scund. J. Immunol. 5 , 113. Greene, M. I., and Weiner, H. L. (1980). J. Immunof. 125, 283. Halperen, S., Eggers, H. J., and Tamm, I. (1964). Virology 23, 81. Harter, D. H., and Choppin, P. W. (1965). J. Immunol. 95, 730. Hartley, J. W., and Rowe, W. P. (1976). J . Virol. 19, 19. Hazelbauer, G. L. (1975). J. Bucteriol. 124, 119. Helenius, A., Morein, B., Fries, E., Simons, K., Robinson, P., Schirrwacher, V., Terhorst, C., and Strominger, J. L. (1978). Proc. Nurf. Acud. Sci. U.S.A. 75, 3846. Hinuma, Y., Suzuki, M., and Sairenji, T. (1975). Int. J. Cancer 15, 799. Hirst, G. K. (1942). J. Exp. Med. 75, 49.

VIRUSES AND CELL SURFACE RECEPTORS

59

Hirst, G. K. (1950). J . Exp. Med. 91, 177. Holland, J. J. (1961). Virology 15, 312. Holland, J. J., and McLaren, L. C. (1961). J . Exp. Med. 114, 161. Holland, J. J., McLaren, L. C., and Syverton, J. T. (1959). J . Exp. Med. 110, 65. Holmes, K. V., Doller, E. W., and Behnke, J. N. (1981). In “Biochemistry and Biology of Coronaviruses” (V. ter Meulen, S. Siddell, and H. Wege, eds.), pp. 133-142. Plenum, New York. Holmgren, J., Svennerholm, L., Elwing, H., Fredman, P., and Strannegard, 0. (1980). Proc. Natl. Acad. Sci. U.S.A. 77, 1947. Hoom, B., and Tyrrell, D. A. J. (1966). Arch. Ges. Virusforsch. 18, 210. Incardona. N. L. (1981). In Virus Receptors. Part 2: Animal Viruses” (K. Lonberg-Holm and L. Philipson, eds.), pp. 157-168. Chapman & Hall, London. Jackson, R. L., Segrest, J. P., Kahanes, I., and Marchesi. V. T. (1973). Biochemistry 12, 3131. Johnson, R. T. (1980). Brain 103, 447. Jondal, M., and Klein, G. (1973). J . Exp. Med. 138, 1365. Kalyanaraman, V. S . , Sarngadharan, M. G., and Gallo, R. C. (1978). J . Virol. 28(3), 686. Kathan, R. H., Winzler, R. J . , and Johnson, C. A. (1961). J . Exp. Med. 113, 37. Klenk, E., Faillard, H., and Lempfried, H. (1955). Z . Physiol. Chem. 301, 235. Kodza, H., and Jungeblut, C. W. (1958). J . Immunol. 81, 76. Krzystyniak, K., and Dupuy, J. M. (1981). J . Gen. Virol. 57, 53. Kunin, C. M. (1962). J . Immunol. 88, 556. Kunin, C. M., and Jordan W. S. (1961). Am. J . H y g . 73, 245. Landau, B. J., Crowell, R. L., Boclair, C. W., and Zajac, B. A. (1972). Proc. Soc. Exp. B i d . Med. 141, 753. La Placa, M. (1963). Nature (London) 199, 1211 . Laver, W. G., and Valentine, R. C. (1969). Virology 38, 105. Lehninger, A. L. (1975) In “Biochemistry. The Molecular Basis of Cell Structure and Function”, 2nd Ed., pp.183-216. Worth, New York. Lentz, T. L., Burrage, T. G., Smith A. L., Crick, J., and Tignor, G. H. (1982). Science 215, 182. Levanon, A . . Kohn, A,, and Inbar, M. (1977). J . Virol. 22, 353. Lonberg-Holm, K. (1964). In “Rapid Mixing and Sampling Techniques in Biochemistry” (B. Chance. R. H. Eisenhardt, Q. H. Gibson, and K. K. Lonberg-Holm, eds.), pp. 275-286. Academic Press, New York. Lonberg-Holm, K. (1981 ). In “Virus Receptors. Part 2: Animal Viruses’’ (K. Lonberg-Holm and L. Philipson, eds.), pp. 3-20. Chapman & Hall, London. Lonberg-Holm, K., and Butterworth, B. (1976). Virology 71, 207. Lonberg-Holm, K., and Philipson, L. (1974). In “Monographs in Virology” (J. L. Melnick, ed.), Vol. 9 , pp. 1-148. Karger, Basel. Lonberg-Holm, K., and Philipson, L. (1980). In “Cell Membranes and Viral Envelopes” (H. A. Blough and J. M . Tiffany, eds.), Vol. 2, pp. 789-848. Academic Press, New York. Lonberg-Holm, K., and Whiteley, N. M. (1976). J . Virol. 19, 857. Lonberg-Holm, K., and Yin, F. H. (1973). J . Virol. 12, 114. Lonberg-Holm, K., Crowell, R. L., and Philipson, L. (1976). Nature (London)259, 679. McCarthy, M., Jubelt, B., Fay. D. B., and Johnson, R . T. (1980). J . Med. Virol. 5, I . McClintock, P. R . , Billups. L. C., and Notkins, A . L. (1980). Virology 106, 261. McFarland, H. F. (1974). J . Immunol. 113, 1978. McGrath. M. S . , and Weissman, I. L. (1979). Cell 17, 65. McGrath, M. S . , Decleve, A , , Lieberman, M., Kaplan, H. S . , and Weissman, I. L. (1978). J . Virol. 28, 819. McLaren, L. C., Holland, J . J . , and Syverton, J. T. (1959). J . Exp. Med. 109, 475.

60

MARC TARDIEU ET AL.

Manthorpe, C. M., Wilkin G.P., and Wilson, J . E. (1977). Brain Res. 134, 407. Marchalonis, J . J . , Cone R. E., and Santer V. (1971). Biochem. J . 124, 921. Markwell, M. A. K., and Paulson, J. C. (1980). Proc. Narl. Acad. Sci. U.S.A. 77, 5693. Markwell, M. A. K., Svennerholm, L., and Paulson J. C. (1981). Proc. Natl. Acad. Sci. U.S.A. 78, 5406. Marshall, T. H., and Rapp, U. R. (1979). J. Virol. 29, 501. Massey, R. J., and Schochetman, G.(1981a). Virology 115, 20. Massey, R. J . , and Schochetman, G.(1981b). Science 213, 447. Merz, D. C., and Wolinski, J. S. (1981). Virology 114, 218. M e n , D. C., Prehm, P., Scheid, A,, and Choppin, P. W. (1981). Virology 112, 296. Miller, D. A,, Miller, 0. J., Dev, V. G., Hashmi, S., Tantravahi, R., Medrano, L., and Green, H. (1974). Cell 1, 167. Moldow, C. F., Kauffman, R. S., Devare, S. G.,and Stephenson, J. R. (1979). Virology 98, 373. Morahan, P. S., and Grossberg, S. E. (1970). J . Infecr. Dis. 121, 615. Morishima, T., McClintock, P. R., Billups, L. C., and Notkins, A. L. (1982). Virology 116, 605. Much, D. H., and Zajac, I. (1973). J. Gen. Virol. 21, 385. Nepom, J . T., Weiner, H. L., Dichter, M. A., Tardieu, M., Spriggs, D. R., Gramm, C. F., Powers, M. L., Fields, B. N., and Greene, M. I. (1982). J . Exp. Med. 155, 155. Nomby, E., Marusyk, H., and Hammerskjold, M. L. (1969). Virology 38, 477. Oie, H. K., Gazdar, A. F., Lalley, P. A., Russell, E. K., Minna, J. D., DeLarco, J., Todaro, G.J., and Francke, U. (1978). Nature (London) 274, 60. Oldstone, M. B. A,, Tishon, A,, Dutko, F. J., Ian, S., Kennedy, T., Holland, J J., and Lampert, P. W. (1980). J . Virol. 34, 256. Onodera, T., Toniolo, A., Ray, U. R., Jenson, A. B., Knazek, R. A., and Notkins, A. L. (1981).J. Exp. Med. 153, 1457. Palese, P., and Compans, R. W. (1976). J. Gen. Virol. 33, 159. Palese, P., Schulman, J . L., Bodo, G.,and Meindl, P. (1974a). Virology 59, 490. Palese, P., Tobita, K., Ueda, M., and Compans, R. W. (1974b). Virology 61, 397. Philipson, L., Beatrice, S. T., and Crowell, R. L. (1973). Virology 54, 69. Pilch, P. F., and Czech, M. P. (1980). Science 210, 1152. Quersin-Thiry, L. (1958). J. Immunol. 81, 253. Richter, J . M. (1976). Ph.D. Thesis, Univ. of Texas Southwestern Medical School at Dallas. Dallas, Texas. Rowe, W. P., and Sato, H. (1973). Science 180, 640. Rubin, H. (1960). Proc. Narl. Acad. Sci. U.S.A. 46, 1105. Rubin, H. (1961). Virology 13, 200. Ruddle, N. H., Conta, B. S., Leinwand, L., Kozak, C., Ruddle, F., Besmer, P., and Baltimore, D. (1978). J . Exp. Med. 148,451. Sabin, A. B. (1959). Br. Med. J . I, 663. Sarma, P. S . , Cheong, M., Hartley, J. W., and Huebner, R. J. (1967). Virology 33, 180. Schaffar-Deshayes, L., Choppin, J . , and Levy, J-P. (1981). J . Immunol. 126, 2352. Scheid, A. (1981) In “Viral Receptors. Part 2: Animal Viruses” (K. Lonberg-Holm and L. Philipson, eds.), pp. 47-62. Chapman & Hall, London. Scheid, A., and Choppin, P. W. (1975). In “Protease and Biological Control” (E.Reich, D. B. Rifkin, and E. Shaw, eds.), pp. 645-659. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Scheid, A., and Choppin, P. W. (1976). Virology 69, 265. Scheid, A., Caliguiri, L. A., Compans, R. W., and Choppin, P. W. (1972). Virology 50, 640. Schild, G.C. (1979). Br. Med. Bull. 35, 1-91. Schnitzer, T. J . , Weiss, R. A., Juricek, D. K., and Ruddle, F. H. (1980). J. Virol. 35, 575.

VIRUSES AND CELL SURFACE RECEPTORS

61

Schochetman, G . , Arthur. L. 0..Long, C. W., and Massey. R. J. (1979). J. Virol. 32, 1131. Schonberger, L. B., Hurwitz, E. S., Katona, P., Holman R. C., and Bregman, D. J. (1981). Ann. Neurol. 110, 105. Sege, K., and Peterson, A. S. (1978). Proc. Narl. Acud. Sci. U.S.A. 75, 2443. Smith, A. L., and Tignor, G. H. (1980). Arch. Virol. 66, 1 I . Smith, W., and Cohen, A. (1956). Br. J. Exp. Puthol. 37, 612. Snyder, D. S., Raine, C. S . , Farooq, M., and Norton, W. T. (1980). J. Neurochem. 34, 1614. Springer, G. F., Nagai, Y., and Tegtmeyer, H. (1966). Biochemistry 5, 3254. Springer, T. A. (1980) In “Monoclonal Antibodies. Hybridomas: A New Dimension in Biological Analyses” (R. H. Kennet, T. J. McKearn, and K. B. Bechtol, eds.), pp. 185-217. Plenum, New York. Steck, F. T., and Rubin, H.(1966). Virology 29, 628. Stephenson, J. R., Hino, S . , Garrett, E. W., and Aaronson, S. A.( 1976). Nurure (London)261,609. Svennerholrn, L., and Fredrnan, P. (1980). Biochirn. Biophys. Acru 617, 97. Talbot, P., Rowlands, D. J., Burroughs, J. N., Sangar, D. V., and Brown, F.(1973). J. Gen.Viro/. 19, 369. Tardieu, M., and Weiner, H. L. (1982). Science 215, 419. Tardieu, M., Neporn, J. T., Epstein, R. L.,Weiner, H. L., Noseworthy, J., Gentsch, J., Fields, B. N.,and Greene, M. 1. (1982). Fed. Proc. Fed. Am. Soc. Exp. B i d . 41, 959. Vahlne, A., Svennerholm, B., and Lycke, E. (1979). J. Gen. Virol. 44, 217. Volsky, D. J., Shapiro, I. M., and Klein, G . (1980). Proc. Nutl. Acud. Sci. U.S.A. 77, 5453. Wadell, G. (1969). Proc. Soc. Exp. Biol. Med. 132, 413. Wang, R., Pollack, R., Kusano, T., and Green, H. (1970) J. Virol. 5, 677. Weiner, H. L., and Fields, B. N. (1977). J. Exp. Med. 146, 1305. Weiner, H. L., Drayna, D., Averill, D. R., and Fields, B. N. (1977). Proc. Nurl. Acud. Sci. U.S.A. 74, 5744. Weiner, H. L., Auk, K. A , , and Fields, B. N. (1980a). J. Immunol. 124, 2143. Weiner, H. L., Greene, M. I., and Fields, B. N. (1980b). J. Immunol. 125, 278. ) Infect. Dis. 141(5), 609. Weiner, H. L., Powers, M. L., and Fields, B. N. ( 1 9 8 0 ~J. Weiss, R. A. (1981). In “Virus Receptors. Part 2: Animal Viruses” (K. Lonberg-Holm and L. Philipson, eds.), pp. 185-202, Chapman & Hall, London. White, D. 0. (1959). Virology 9, 680. Wiley, D. C., Wilson, I. A., and Skehal, J. J. (1981). Nature (London) 289, 373. Wilson, I. A., Skehal, J. J., and Wiley, D. C. (1981). Nature (London) 289, 366. Wilson, T., Papahadjopoulos, D., and Taber, R. (1977). Proc. Nutl. Acad. Sci. U.S.A. 74, 3471. Wolinski, J. S., and Stroop, W. G. (1978). Arch. Virol. 57, 355. Woodruff, J. F., and Woodruff, J. J. (1972). Cell. Immunol. 5, 296. Woodruff, J. F., and Woodruff, J. J. (1974). J . Immunol. 112, 2176. Yavin. E., Yavin, Z., Schneider, M. D., and Kohn, L. D. (1981). Proc. Nurl. Acad. Sci. U.S.A. 78, 3180. Yefenof, E., Klein, G , Jondal, M., and Oldstone, M. B. A. (1976). In?. J. Cancer 17, 693. Yoon, J. W., McClintock, P. R., Onodera, T., and Notkins, A. L. (1980). J. Exp. Med. 152, 878.

This Page Intentionally Left Blank

INTERNATIONAL REVIEW OF CYTOLOGY, VOL. 80

The Molecular Basis of Crown Gall Induction W. P. ROBERTS Department of Microbiology, La Trobe University, Bundoora, Victoria, Australia

1. Introduction . . . . ................................ 11. The Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

111. IV. V. VI . Vll.

VIII. IX.

X. XI.

The Crown Gall Bacteria ............................. The Physiology of the Gall.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Involvement of Wounding in Gall Induction . . The Opines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Involvement of Plasmids in Tumor Induction A. The Presence and Structure of Ti-Plasmi Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B . Products of the T-DNA . . . ............. C. Ongocenic-Related Functions Not on T-DNA . . . . . . . . . . . . . . D. Nononcogenic Functions on the Ti-Plasmid. . . . . . . . . . E. Ti-DNA Transfer and Integration into the Plant Cell . , F. Reversion to the Normal State . . . . . . . . . . . . . . . . . . . . , . . . . , Significance of Crown Gall Induction to Agrobacteriurn . . . . . . . . . The Evolutionary Origin of Crown Gall . . . . . . . . . . . . . . . . . . . . . . Agrobacteriurn and Genetic Engineering . . . . . . . . . . . . . . . . . . . . . . Future Work and Prospects . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ............. References . . . . . . . . . . . . . . . . . . . . . .

63 64 65 65 66 68 70 71 74

76 78 79 82 83 84 86 87 88

I. Introduction Although Smith and Townsend demonstrated in 1907 that the neoplastic plant disease, crown gall, was caused by a bacterium, there was little progress elucidating the mechanism of gall induction until White and Braun (1941, 1942) conclusively demonstrated that the bacteria initiated a stable transformation of the plant cell. Once the disease had been induced then the bacteria were no longer required for continued growth of the gall. These reports stimulated a great deal of research but little further progress was made in the understanding of this transformation until Van Larebeke et al. (1973, 1974) reported that crown gall bacteria carried a large plasmid that was essential for pathogenicity. Since this discovery, there has been many reports revealing the role of this plasmid in crown gall induction and the molecular basis of crown gall. 63 Copyright Q 1982 by Academic Press. Inc All rights of reproduction in any form reserved ISBN 0-12-364480-0

64

W.

P. ROBERTS

11. The Disease

Crown gall disease is characterized by the production of galls on plant roots. These galls often appear at the crown of the plant but may be anywhere on the root system. Occasionally galls appear on the aerial parts of the plant possibly due to bacteria being transferred from the soil. There are also reports that agrobacteria can move internally in the plant and give rise to tumors remote from the original infection point (see review by Lippincott and Lippincott, 1975). Experimentally, galls can be produced on all parts of a suitable host plant including the roots, stems, and leaves. The gall itself consists of undifferentiated cells which divide indefinitely to form a tumor. A gall may make up a very significant proportion of the dry weight of the plant and values up to 50 kg (Walker, 1969) gall weight have been recorded. As well as being a very significant drain on the photosynthetic capacity of the plant, a gall may also disrupt the phloem and xylem. This is particularly significant with galls that occur at the crown of the plant where the entire flow of nutrients up and down the plant may be affected. The overall result is that plants with crown gall grow poorly, give low yield, and may be very sensitive to stress. DeCleene and DeLey (1976) surveyed the available literature and carried out experiments to determine the host range of crown gall. They report that 643 species representing 331 genera of 91 families were susceptible to crown gall. This included gymnosperms and dictyledonous and monocotyledonous angiosperms. Only 7 out of 75 species of monocotyledonous plants tested have been reported as susceptible to crown gall and their own experiments gave only one possible positive response. Although only a few bacterial strains have been used to carry out these tests, it is clear that monocotyledonous plants are not generally susceptible to crown gall. The host list includes many economically important plants such as apricots, almonds, peaches, apples, pears, grape vines, and roses and the disease is very widely distributed (Hayward and Waterston, 1965). It is difficult to find data on the occurrence and economic loss caused by this disease but it is not uncommon to find complete orchards with every, or almost every, tree infected. The two common modes of infection are by contaminated material from the nursery or by planting into areas already infected. Many countries have legislation preventing the sale of material infected with crown gall. However, this only prevents the distribution of obviously diseased material not of material contaminated by Agrobacterium but lacking a visible gall. New (1972), for example, found that up to 40% of bundles of almond plants ready for distribution were contaminated with Agrobacterium. Pathogenic agrobacteria could not be isolated from soil not previosly used for stone fruit cultivation indicating that at least in the areas studied that it was an introduced disease. Until the introduction of a biological control method (New and Kerr, 1972; Kerr, 1972) there was no satisfactory control system for the disease. The biolog-

CROWN GALL INDUCTION

65

ical control method depends on the production of a bacteriocin by one Agrobacterium strain (strain K84) that inhibits a wide range of pathogenic agrobacteria. However, not all pathogenic strains are controllable (see, for example, Kerr and Panagopoulos, 1977). The biological control of crown gall has been reviewed by Moore and Warren ( 1979). 111. The Crown Gall Bacteria

Agrobacterium is a gram-negative organism with peritrichous flagella. Strains can be divided into a number of groups on the basis of a range of physiological tests (Keane et al., 1970; White, 1972; Kersters et a l . , 1973; Panagopoulos and Psallidas, 1973; Kerr and Panagopoulos, 1977; Sule, 1978). The organisms causing crown gall disease have been classified as Agrobacterium tumefaciens and nonpathogens as A. radiobacter. As nonpathogenic isolates can be converted to pathogenic forms by simple genetic manipulation, pathogenicity cannot be considered a valid species differentiation characteristic and it has been suggested that pathogenicity be reduced to a lower status (Keane et a l . , 1970). These authors also reported that the organism causing galls on Rubus spp., A . rubi, and the hairy root disease organism, A. rhizogenes were indistinguishable from A . tumefaciens by a range of biochemical tests. Hairy root disease (Riker et a / ., 1930) is characterized by the proliferation of roots at the site of infection and the molecular basis of the disease is similar to crown gall (White and Nester, 1980a,b; Petit e t a / . , 1981; Chilton et a l . , 1982). Holmes and Roberts ( 1981 ) carried out a taxonomic study of Agrobacterium strains. These strains could be divided into four clusters, three of which broadly corresponded to the physiological groups already described. The fourth group contained yellow-pigmented isolates which had not previously been recognized as related to Agrobacterium. These authors concluded that the clusters corresponded to species and without regard to pathogenicity assigned the names Agrobacterium tumefaciens, A . rhizogenes, and A . rubi to three of the groups. They suggested that within these species individual isolates be designated as saprophytic, tumorogenic, or rhizogenic. Determination of the taxonomic position of the yellow-pigmented isolates was not attempted as few strains had been included in the study. Agrobacterium is quite closely related to Rhizobium and there have been suggestions that the two genera be merged (Graham, 1964; Herberlein et a / . , 1967; DeLey, 1968; Moffett and Colwell, 1968; White, 1972).

IV. The Physiology of the Gall Much of the early work concentrated on the hypothesis that the gall was induced by the production of plant hormones by agrobacteria. Indeed agrobac-

66

W. P. ROBERTS

teria are able to synthesize plant hormones, at least in culture (Clark, 1974; Chapman et al., 1967; Lui and Kado, 1979). However, with the demonstration by White and Braun (1941, 1942) that gall tissue could be cultured free of bacteria on media containing no plant hormones or growth factors, it became clear that crown gall was a transformed plant tissue. Crown gall tissue can be cultured on standard plant tissue culture media without any hormones and will continue to grow in a stable manner indefinitely. Kado ( 1976) summarizes the physiological differences between normal plant tissue and crown gall tissue. Since this review, other reports have shown that crown gall cells may have qualitative and/or quantitative differences in cytokinins (Scott et al., 1980; Stuchbury, 1979; Peterston and Miller, 1977), adenyl cyclase, and cyclic AMP phosphodiesterase activity (Rutherford et al., 1977), cellular membranes (Cockerham and Lundeen, 1979; Phillips and Butcher, 1979), cation concentrations (Lentz et al., 1979; Radiosevich and Galsky, 1979), Feulgen DNA content (De Cleene et al., 1980), opines (Firmin and Fenwick, 1977, 1978; Kemp, 1977, 1978), putrescine biosynthesis (Speranza and Bagni, 1977), and ubiquinone formation (Ikeda et al., 1976). Although these studies clearly demonstrate that crown gall cells are physiologically different from normal plant cells, none of them indicates the molecular basis of cell transformation. However, the high levels of both cytokinins and auxins suggest that the biochemical basis of the gall is that the cells become independent of the normal hormonal control of the cell. The agrobacteria are only required during the initial gall induction phase, so they must somehow transform the plant cell at this time. Braun (1947) suggested that some principle which he called the Tumor Inducing Principle or T.I.P. was produced by the Agrobacterium and was responsible for causing plant cell transformation. Much of the reported work from this time has concentrated on the possible nature of the T.I.P. and its role in tumorogenesis. A range of substances including RNA, DNA, bateriophages, proteins, and lipopolysaccharides have been suggested as the T.I.P. (see review by Kado, 1976).

V. Involvement of Wounding in Gall Induction Although Riker and Berge (1935) demonstrated that wounding was an absolute requirement for gall induction the precise role of wounding has still not been determined. There are a number of possibilities that include (1) wounding exposes plant cell-Agrobacrerium recognition sites; (2) wounding alters the biochemical state of the plant cell to make it susceptible to gall induction; and (3) wounding creates a suitable environment for Agrobacterium to synthesize and transfer the T.I.P. The involvement of a plant cell-Agrobacterium recognition system in crown

CROWN GALL INDUCTION

67

gall induction was suggested by Lippincott and Lippincott (1 969a). They reported that certain avirulent strains of Agrobacteriurn when mixed with a virulent strain prior to inoculation could reduce the number of galls induced on Pinto bean leaves in a quantitative crown gall assay. Heat-inactivated virulent bacteria also gave a reduction in the number of galls induced. The kinetics of the system suggested that one avirulent bacterium could prevent gall induction by one virulent bacterium and they concluded that crown gall inducation only required one bacterium binding to a specific site. However, in their system approximately lo6 bacteria were needed to induce one gall but they concluded that this represented the probability of one bacterium binding to the correct site. Similar results were reported by Beiderbeck (1973), Glogowski and Galsky (1978), and Ohyama et al. (1979). Beiderbeck (1973) reported that cell wall extracts from virulent bacteria mixed with viable pathogens gave a reduction in gall formation. Whatley et al. (1976) and Matthysse et al. (1978) extended these findings when they reported that a lipopolysaccharide extract (LPS) from some Agrobacteriurn strains could reduce gall formation. LPS from strains that were found to be site-binding by Lippincott and Lippincott ( 1969a) inhibited gall induction while LPS from non-site-binding strains did not and they suggested that the LPS component of the bacterial cell wall was the bacterial site of the specific plant-bacteria binding system essential for gall induction. This LPS fraction was found to have a similar action on gametophore induction by Agrobacterium in moss (Whatley and Spiess, 1977). Both chromosomal and Ti-plasmid genes seem to be involved in this attachment process (Whately et al., 1978; Matthysse et al., 1978). The complementary experiment was reported by Lippincott et al. (1977). They found that cell wall extracts from plant cells when mixed with virulent agrobacteria could reduce the number of galls formed. Much of the inhibitory effect of the plant cell wall extracts was eliminated by pretreatment with LPS or dead virulent bacteria and they concluded that the site for attachment of agrobacteria was in the plant cell wall. Pueppke and Benny (1981) were able to isolate a soluble uronic acid-containing fraction from potato that inhibited gall induction and which they believed contained the bacterial attachment sites. The concept that there is a period of cell conditioning before the plant cell is sensitive to gall induction comes from the studies of Braun (1947, 1952) and Braun and Mandle (1948). This cell conditioning could take place at 25 and 32°C without the bacteria being present although gall induction occurred only below 28°C. Braun (1952, 1954) reported that plant cells were most susceptible to transformation just before the onset of the first cell division after wounding and suggested that the conditioning process may involve alterations in the permeability of the plant cell and so allow entry of the T.I.P. Lipetz (1965) extended this finding and reported that the time taken to the onset of the first cell division decreased with temperature as did the time to reach maximum sensitivity to gall

68

W. P. ROBERTS

induction. At all temperatures greatest sensitivity to induction was found just before the first cell division after wounding. These results suggested that the process of gall induction is closely related to cell division. However, cell division alone is not a satisfactory condition for gall induction as bacteria added to the surface of actively dividing callus tissue do not produce galls unless wounding occurs (Riker, 1923; Riker and Banfield, 1932). There are also reports of gall induction on Pinto bean leaves (Lippincott and Herberlein, 1965) and decapitated pea seedling (Kurkdjian et al., 1969) where gall induction occurs without wounding inducing cell division. Klein (1954) suggested that wounds were essential to release substances that were necessary for gall induction. He reported that washing the wound with sterile water prevented gall induction. In another report (Klein 1955), he showed that carrot discs of a variety which normally respond only weakly to pathogenic bacteria gave a much stronger response when treated with an extract from carrots which normally respond strongly. The possible importance of wound sap composition is emphasized by the studies of Lippincott and Lippincott (1966, 1969b) who showed that bacterial nutrition was a significant factor in the expression of virulence particularly when only small wounds were involved. However, there is one report (Therman and Kulipa-Ahvenniemi, 1971) which indicates that wound washing enhances tumorogenesis. These workers attributed this to a promotion of the first cell division after wounding. A similar promotion of the first cell division after wounding by washing has been reported by Lipetz (1967). Ohyama et al. (1979) report that agrobacteria in plant wounds are associated with fibrils. Matthysse et al. (1981) extended this work and concluded that these fibrils were composed of cellulose and were synthesized by the bacteria. The fibrils anchored bacteria to the plant cells and formed a matrix in which the bacteria became trapped. Fibrils were synthesized only in the presence of plant cells or when the media was supplemented with a soya bean extract and they concluded that the plant cells may release a substance that stimulates production of these fibrils.

VI. The Opines Considerable support for the idea that the T.I.P. was bacterial DNA which was incorporated into the plant cell was due to the discovery of a range of gall specific compounds (Table I, Fig. 1). These compounds which have been found only in crown gall tissue (Scott et al., 1979; Yang et al., 1980c) appear to be related to the Agrobacterium strain inducing the gall rather than to the plant species and Goldmann et al. (1968) suggested that these compounds were coded for by bacterial DNA incorporated into the plant cell during gall induction. One very significant feature was that the strain inducing a particular set of these

69

CROWN GALL INDUCTION TABLE I THE OPINES N2-( I -Carboxyethyl)-amino acid derivatives Octopine Menage and Morel (1964) Octopinic acid Menage and Morel (1965) Histopine Kemp (1977) Lysopine Biernann et a / . (1960) N2-( 1-3-Dicarboxypropyl)-aminoacid derivatives Nopaline Goldmann et ul. (1969) Nopalinic acid Firmin and Fenwick (1977) Other Agropine Coxon et a / . ( 1980); Tate et a / . (1982) Agrocinopines Ellis and Murphy (1981)

HN~C-NH-(CH2)3-CH-COOH

H 2 N-(CH2 13-

I I

H2 N

NH

I

H3C-CH-COOH

Hg C-CH-COOH

octopine

H2 N-(CH~ )

7-

H COOH NH

octopinic acid

4

' '

-

~ HC-C-CH2~ ~

I NH I H3C-CttCOOH

~ H-COOH ~ ~ H

I

H3C-CH-COOH

lysopine

histopine

H N > ~ - N ~ - t) ~ ~3 2 H2N NH

~

I

~

- H~N-(CH ~ ~ ~

CH-COOH ~

* 3 -YH~

I

I

HOOC-(CH2)2-CH-COOH

HOOC-(CH2)2-CH-COOH

nopaline

nopalinic acid

0

agropine

FIG. I .

Structures of the opines

compounds could utilize the same compounds for growth in culture (Petit et al., 1970; Bomhoff et al., 1976; Montoya et al., 1977; Firmin and Fenwick, 1978; Guyon et al., 1980; Ellis and Murphy, 1981). For example, strain T37 could induce nopaline and nopalinic acid in galls and utilize these compounds in

70

W. P. ROBERTS

TABLE I1 OPINESINDUCEDI N GALLSAND CATABOLIZED I N CULTURE BY Agrobacterium Agrocinopines Octopinic Nopalinic Octopine acid Histopine Lysopine Nopaline acid Agropine A B C D

culture. It does not utilize octopine or agropine while strain B6 induces these compounds in galls and utilizes them in culture. Using these criteria, pathogenic agrobacteria can be divided up into groups: the octopine group, the nopaline group, and the agropine group (Table 11). In addition, there are a number of “defective” nopaline strains that utilize nopaline in culture but induce galls containing no known opine (Petit et al., 1970; Sciaky et al., 1978). Most nopaline strains can utilize octopine when a transport system is induced by nopaline (Petit and Tempi, 1975; Klapwijk et al., 1977b). Kemp (1978) suggested that one enzyme coded for production of all the P-( 1carboxyethyl) derivatives in the gall and the properties of the purified enzyme support this conclusion (Hack and Kemp, 1977, 1980). Kemp (1978) also suggested that one enzyme synthesized nopaline and nopalinic acid. No data are available on the properties of these enzymes or the enzyme involved in agropine and agrocinopine synthesis in the gall. Gall specific amino acids were called opines (Tempi et al., 1977) but this definition was later extended to include all crown gall-specific compounds that are utilized by agrobacteria. Opines in gall tissue may reach high levels. Scott et al. (1979) report that nopaline may constitute 19 mg/g dry weight and agropine 35 mg/g dry weight. Octopine levels tend to be much lower at up to approximately 0.5 mg/g dry weight of gall.

VII. Involvement of Plasmids in Tumor Induction Despite the evidence of the opines, none of the early work attempting to demonstrate the presence of Agrobacterium DNA in crown gall cells could be reliably repeated (Kado, 1976). Van Larebeke et al. (1973, 1974) and Zaenen et al. (1974) first demonstrated

71

CROWN GALL INDUCTION

that plasmids were present in pathogenic agrobacteria. They found a good correlation between the presence of a plasmid and pathogenicity. Bacterial strains lacking a plasmid were nonpathogenic. They showed that the technique of producing avirulent agrobacteria by growth at 37°C (Hamilton and Fall, 1971) cured the bacteria of this plasmid. The reciprocal experiment was reported by Watson et al. (1975) and Van Larebeke (1975) who were able to transfer the plasmid back into cured strains using the in vivo transfer method of Kerr (1971). Nonpathogenic strains that received the plasmid became pathogenic. These experiments unequivocally established that this plasmid (that was named the Ti or tumor-inducing plasmid) was required for crown gall induction. The Ti-plasmid determines both the catabolism of opines in culture and the synthesis of opines in galls (Watson et al., 1975; Bomhoff et al., 1974; Montoya et al., 1977; Van Larebeke et al., 1975) and Ti-plasmids corresponding to all opine groups exist. These results suggested but by no means proved that the Ti-plasmid was the T.I.P. and that all or part of it was transferred to the plant cell during tumor induction where it coded for tumorogenicity and opine production. The simplest interpretation, that synthesis and breakdown were due to the activities of the same enzyme working in opposite directions in gall and bacteria, was eliminated by the reports that the enzyme systems from bacteria and gall were biochemically and genetically distinct (Bomhoff, 1974, cited in Montoya et al., 1977; Montoya et al., 1977; Petit et al., 1970; Sciaky et al., 1978). A. THE PRESENCE AND STRUCTURE OF TI-PLASMID DNA

IN THE

GALLCELL

The first convincing report of the presence of Ti-plasmid DNA in crown gall cells is that of Chilton et al. (1977). They used isolated restriction endonuclease DNA fragments of an octopine type Ti-plasmid to perform liquid-liquid hybridizations with DNA isolated from the crown gall cells and demonstrated that part of the Ti-plasmid was in the gall cells. This DNA became known as transfer or TDNA. Chilton et al. (1978) presented a restriction endonuclease map of an octopine Ti-plasmid that showed that the restriction fragments present in the gall cell were contiguous on the Ti-plasmid. This suggested that only one discrete region of the Ti-plasmid was stably incorporated into the plant cell. Merlo et al. (1980) and Thomashow (1980a,b) extended this study to a number of crown gall tumor lines induced by different octopine strains. They concluded that each tumor contains a core of T-DNA that is colinear with the Ti-plasmid, a given Tiplasmid does not always give rise to the same T-DNA complement, and that most, if not all the T-DNA, is integrated into plant DNA. In addition, these studies showed that the copy number of parts of the T-DNA varied between tumor lines and suggested that in some lines the T-DNA may be present as tandem repeats and integrated into repeated sequence plant DNA. Up to approx-

72

W . P. ROBERTS

imately 13 X lo6 daltons of Ti-piasmid DNA may be present in octopine type gall cells, depending on the tumor line (Thomashow et al., 1980a; Fig. 2 ) . The structure of the T-DNA in nopaline tumors has also been looked at in detail (Yang el al., 1980a; Zambryski et al., 1980; Lemmers er a!. , 1980; Yadav er al., 1980; Fig. 3). The T-DNA in nopaline tumor cells is larger, with approximately 18 X lo6 daltons of Ti-plasmid DNA (Chilton et al., 1980; Lemmers er al., 1980; Yang et al., 1980a) being present in tumor cells. Like the octopine TDNA it represents one continuous fragment from the Ti-plasmid and internal rearrangement does not take place after insertion into the cell. Each tumor line contains only a few T-DNA plant DNA junction fragments suggesting that the TDNA is inserted only once or a few times in the cell and does not “jump about.” Although there appears to be certain preferred ends of the T-DNA, these are not an absolute requirement as transposon mutagenesis of the ends will still allow tumorogenesis (Lemmers et al., 1980). Like the octopine T-DNA the nopaline T-DNA also seems to be inserted in repetitive sequence plant DNA. Unlike the octopine T-DNA, the nopaline T-DNA does not appear to vary in copy number along the T-DNA region. 1200

60

FIG. 2. A map of the B653 octopine plasmid. The map is calibrated in 10 megadalton segments with the zero site at an endonuclease Smal site common to both octopine and nopaline plasmids (Depicker e t a / . . 1978; Chilton et al., 1978). The extent of the T-DNA is indicated as well as areas known to be essential for pathogenicity (line within map) and areas homologous with the C58 nopaline plasmid (solid blocks). Also indicated are areas involved in determining gall morphology (Gm), octopine synthesis (Ocs), arginine catabolism (Arc), phage API exclusion (Ape), plasmid transfer (Tra), octopine catabolism (Occ), and agropine catabolism (Agc).

CROWN GALL INDUCTION

73

FIG. 3 . A map of the C58 nopaline plasmid. The map is calibrated in 10 megadalton segments with the zero site at an endonuclease Smal site common to both octopine and nopaline Ti plasmids (Depicker et a / . , 1978; Chilton e t a / . . 1978). The extent of the T-DNA is indicated as well as areas known to be essential for pathogenicity (line within map) and areas homologous with octopine plasmids (solid blocks). Also indicated are the areas involved in nopaline synthesis (Nos), arginine catabolism (Arc), nopaline catabolism (Noc), incompatability (Inc), indole acetic acid production (Iaa), phosphorylated sugar catabolism (Psc), Agrocin 84 sensitivity (Agrs), plasmid transfer (Tra), and phage API exclusion (Ape).

Octopine and nopaline plasmids show extensive areas of homology (Chilton et al., 1978; Depicker et al., 1978a,b; Drummond and Chilton, 1978; Hepburn and Hindley, 1979; Van Montagu eral., 1980; Figs. 2 and 3). The areas of homology include a substantial part of the T-DNA region and areas to the left of the T-DNA in both octopine and nopaline type plasmids. Mutagenesis of the T-DNA area common to both nopaline and octopine plasmids abolished pathogenecity or resulted in galls having altered morphology. Mutagenesis of common DNA regions outside the T-DNA also affects oncogenicity (Holsters et al., 1980; Garfinkel and Nester, 1980; DeGreve et al., 1981). This suggests that both octopine and nopaline tumors share the same mechanism of oncogenicity. However, it is interesting to note that White and Nester (1980b) found little homology between A. rhizogenes plasmids and the common T-DNA region of A . tumefaciens plasmids. A similar lack of homology was reported by Thomashow et al. (1980, 1981) for certain narrow host range plasmids. However, in both cases homology was detected at lower levels of stringency suggesting that common

74

W. P. ROBERTS

sequences were present but were perhaps arranged in a different manner to the “normal” Ti-plasmids. Although these studies showed that there were unique fragments formed at the ends of the T-region in plant cells indicating that the T-DNA was covalently linked to plant DNA none of them differentiated between chromosomal, mitochondrial, or chloroplast DNA. Chilton e r a / . (1980), Yadav e t a / . (1980), and Willmitzer et a / . (1980) isolated DNA from subcellular fractions of tumors. TDNA probes were used to hybridize to restriction digests of the isolated plant DNA. Both groups reported that hybridization occurred only with DNA isolated from plant cell nuclei and concluded that the T-DNA was covalently linked to plant nuclear DNA. This supports the report of Nuti et a/. (1980) who showed that Ti-plasmids hybridized to the Feulgen-positive DNA of tumor cells. At this stage it is difficult to generalize about the structure of the T-DNA in the tumor cell except to conclude that all tumors studied to date include an area of common T-DNA. It is presumably this area that is involved in maintaining the tumorous state. It is wise, however, to remember that all the tumors studied have been tobacco tumors induced by only a few strains of Agrobacterium. In most cases the tumor lines had been in culture for long periods so it is possible that the structure of the T-DNA in the cell may simply reflect the general gene rearragement that takes place in tissue culture (see, for example, DeCleene et a/. , 1980). B. PRODUCTS OF THE T-DNA 1. Transcription Drummond et al. (1977) showed that transcription takes place from the TDNA in tumor cells. They isolated pulse-labeled RNA from octopine tumors and normal tobacco tumor tissue and hybridized this to digested Ti-plasmid on cellulose nitrate. They obtained hybridization between the tumor RNA and a digestion fragment from the right-hand end of the T-DNA. This is the area thought to be involved in opine synthesis (see later). No hybridization to the common TDNA region was observed. Gurley et al. (1979) extended these findings when they reported transcription of T-DNA in three sunflower tumor lines incited by octopine strains of Agrobacterium. Their results showed that an area to the righthand end of the T-DNA was always strongly transcribed and the rest of the TDNA to a varying degree. Parts of the common DNA were transcribed strongly to not at all, depending on the particular tumor line. Yang er al. (1980a,b) looked at a nopaline tumor line of tobacco. They detected transcripts from both ends of the T-DNA but the intervening sequences were not transcribed at a level that they could detect. Willmitzer (1980) pulse-labeled RNA in highly purified nuclei of octopine tumors. This RNA was found to hybridize to restriction fragments covering most of the T-DNA in this tumor line. Bevan er al. (1981) and Murai and Kemp (1982) both report that there is extensive transcription of the T-DNA in the octopine tumors they studied. These studies confirm that the T-DNA is in

CROWN GALL INDUCTION

75

the plant cell nucleus and indicate that transcription can take place over most, if not all of the T-DNA. Gelvin et al. (198 1) studied the transcription of the T-DNA in the bacterium as well as in tumor cells. In the tumor cell they located specific areas of high level transcription at both ends of the T-DNA. In the bacterium slight transcription of the whole T-DNA occurred. This suggests that the T-DNA has eukaryote rather than prokaryote transcription signals and does not depend on specific insertion downstream of plant cell transcription signals. The conclusion from all these data is that the T-DNA may be transcribed to varying degrees in different tumor lines. The fact that transcription of the common DNA is detected in at least some tumors suggests that tumorogenesis involves the production of some product by the T-DNA rather than the simple integration of the T-DNA into a specific site. This will be considered in more detail in a later section. 2 . Translation McPherson er al. (1980) used RNA that hybridized to T-DNA in an in vivo translation system. RNA from one tumor line hydridizing to the right-hand end of the T-DNA produced proteins of 30,000 and 16,500 molecular weight. RNA from another tumor line lacked the smaller protein. One tumor line which did not produce octopine and in which the T-DNA region was altered in the right-hand end produced no proteins from RNA that hybridized to this region. The octopine synthase protein is approximately 38,000 (Hack and Kemp, 1980) which is significantly bigger than the protein detected in the in vitro traiislation system. McPherson et al. ( I 980) suggested that perhaps the enzyme is glycosylated thus accounting for the extra size. RNA hybridizing to the common T-DNA region from all tumors produced a protein of approximately 15,000 molecular weight. Because this region and presumably this protein is present in all tumors, it was suggested that the protein may be involved in tumorogenesis. The expression of cloned Ti fragments in E. coli minicells was studied by Schweitzer et al. (1980). Approximately five proteins with molecular weights ranging from 15,000to 45,000 were detected. Determination of the significance of these results will have to await studies looking for similar proteins in tumor cells but they do indicate the possible size and some of the properties of proteins that may be involved in tumorogenesis.

3 . Phenotypic Expression of the T-DNA The known functions of the T-DNA are opine synthesis and tumorogenesis. Several groups have attempted to map these functions more precisely using various mutagenesis mechanisms. Depicker et al. (1978a,b) found that the insertion of the plasmid RP4 into a nopaline type Ti-plasmid in the common T-DNA region abolished pathogenicity. Revertants that had lost the RP4 plamid regained pathogenicity.

76

W. P. ROBERTS

Koekman et al. (1 979) reported that a number of deletion mutants of octopine type plasmids affecting the T-DNA region were still virulent. These deletions involved approximately the right half of the T-DNA as defined by Thomashow et al. (1980a) but left approximately half of the DNA common to both octopine and nopaline Ti-plasmids. These deletion mutants were weakly virulent on Helianthus, Solanum, and Kalanchoe but avirulent on Nicotiana. None of them contained octopine. Their results suggested that opine syntheses per se was not the tumorogenic principle and that both ends of the T-DNA are not required for tumorogenicity. Ooms et al. (1980, 1981) used transpositional mutagenesis with Tn904 (Klapwijk et al., 1980) to obtain a series of mutants in an octopine Tiplasmid. They obtained a number of insertions into the T-DNA region which were virulent to weakly virulent but had an altered gall morphology. In some cases, galls similar to wild type galls could be induced by using mixtures of mutants or adding auxin during gall development. Garfinkel and Nester (1980) used Tn5 to generate a series of mutants of octopine plasmids. Like the mutants produced by Ooms et al. (1980) these were virulent to weakly virulent depending on the host and the tumors formed had an altered morphology compared to tumors from wild type strains. All these mutants produced octopine. Similar results were reported by DeGreve et al. (1981). Holsters et al. (1980) report a similar study on nopaline type plasmids using Tnl and Tn7 as mutagens. This allowed them to define the area which was essential for tumorogenicity. They also found mutations that affected the morphology of the tumor produced as well as a region to the right of the T-DNA which appeared to be involved in determining host range. The morphology of tumors formed by wild type Ti-plasmids depends on the particular plasmid/plant combination (Gresshoff et al., 1979) and even on the position of inoculation up the stem. This suggests that crown gall induction is not an all or none response and that there is some interaction between the host and the inducing bacteria. Deletions or transposon insertions into the right-hand end of the T-DNA of both octopine and nopaline plasmids abolish opine synthesis in the gall (Holsters et al., 1980; DeGreve et al., 1981) indicating that these functions are distinct from tumorogenicity. No information about the location of the genes for agropine and agrocinopine synthesis in galls is available. These functions are believed to be T-DNA encoded functions but as yet no reports indicate their position. The known functions of the T-DNA in the plant and their relative location on this DNA are summarized in Figs. 2 and 3.

C. ONCOGENIC-RELATED FUNCTIONS NOT ON T-DNA Although parts of the T-DNA are essential for tumor induction on plants, it is quite possible that other essential functions are not coded for by the T-region.

CROWN GALL INDUCTION

77

This may include functions involved in processing the T-DNA, its transfer to and incorporation into the plant cell DNA, survival in the wound, and bacterial attachment to the plant cell. So far, only bacterial attachment has been demonstrated to be at least partly a plasmid-coded function (Whatley et a/., 1978; Matthysse et al., 1978). Ooms et al. (1980, 1981) and Garfinkel and Nester (1980) observed that a number of transposon mutants outside the T-DNA of octopine plasmids abolished tumorogenicity. Similar results were reported by Holsters et al. (1980) for nopaline plasmids. They found at least two regions outside the T-DNA on the left-hand side of the plasmid which were essential for oncogenicity (Fig. 3). Lui et al. (198 1) also report that transposon mutagenesis into this area can abolish pathogenicity. This area appeared to be involved in the production of the plant hormone, indole acetic acid (IAA), by the bacteria and they suggested that 1AA production may be important in the initial stages of crown gall induction. All these mutations are in DNA that is common to both nopaline and octopine type plasmids again emphasizing that there probably is a common tumorogenic mechanism in all tumorogenic plasmids from Agrobacrerium. Loper and Kado (1979) and Thomashow et al. (1980~)report that the host range of Agrobacterium depends on the Ti-plasmid. Plasmids transferred from a narrow host range strain to a plasmidless strain gave a pathogen with a narrow host range and a plasmid transferred from a wide host range strain to a narrow host range strain cured of its Ti-plasmid gave a wide host range pathogen. Although these reports do not indicate where the genes for host range are located on the Ti-plasmid the mutation studies suggest that certain areas of the T-DNA may determine host range but this does not rule out the possibility that there are other areas around the Ti-plasmid involved in this function. Ooms et al. (1981) report that some transposon insertions into areas outside the T-region but in areas homologous in different Ti-plasmids gave galls with altered morphology. This is surprising as this DNA is not maintained in the gall so it is difficult to imagine how it could have a stable effect on the gall morphology. Perhaps this area is involved in transfer and processing of the T-DNA so that a mutation in this area leads to an altered T-DNA insert in the gall cell. The involvement of the plasmids in determining the morphology of the galls is further emphasized by the studies of Albinger and Beiderbeck (1977), Colak and Beiderbeck (1979), Costantino et al. (1980), and White and Nester (1980a,b) on induction of hairy root disease by A . rhizogenes. Although these studies clearly indicated that this disease is plasmid encoded, White and Nester (1980b) show that there is only slight homology between this plasmid and the common DNA of Ti-plasmids. However, it is clear that the disease is induced by the incorporation of plasmid DNA into the plant cell (Chilton et al., 1982) and that hairy root contains agropine (Petit et al., 1981). So the weak homology detected probably represents only rearrangement of the tumorogenic functions rather than a completely different mechanism.

78

W . P. ROBERTS

Several reports have shown that there is a plasmid-encoded temperature-sensitive phase in gall induction (Rogler, 1980, 1981). Rogler suggested that the temperature-sensitive step in gall induction was the temperature-sensitive replication mechanism for the Ti-plasmid (Hamilton and Fall, 1971) and that this replication mechanism was involved in tumorogenesis. Schilde-Rentschler et al. (1977) report the isolation of a number of temperature-sensitive mutants with respect to gall induction. It is possible that these are mutants in the temperaturesensititive replication mechanism of the Ti-plasmid but no data are available to confirm this hypothesis. D. NONONCOCENIC FUNCTIONS ON THE TI-PLASMID The known functions of both octopine and nopaline type plasmids are summarized in Figs. 2 and 3. The best studied functions are those for the catabolism of the opines. Both these functions map to the right of the T-DNA region in both octopine and nopaline type plasmids. Holsters et al. (1980) found two regions involved in nopaline catabolism. Mutations in one of these regions affected the catabolism of argine and ornithine. These compounds are thought to be breakdown products of some opines and catabolic genes for them are located on the Tiplasmid (Ellis er al., 1979). All this region may be involved with the breakdown products. The genes for the catabolism of another opine, agropine, are further around the plasmid. Agrocinopines (Ellis and Murphy, 1981) are thought to be the true substrates for the plasmid-encoded uptake system for agrocin 84 (Engler et al., 1975; Murphy and Roberts, 1979) and so this function should map to the same region. Although Ellis and Murphy (1981) showed that Tn7 insertion into the agrocin 84 sensitivity gene prevented catabolism of one of the agrocinopines, they did not show this was due to lack of uptake and a number of agrocin 84resistant strains could catabolize certain agrocinopines. Transfer functions for both plasmid types are shown. Despite suggestions that plasmid transfer and gall induction may share common biochemical events (Temp6 et al., 1977), Holster et al. (1980) were able to obtain mutants which were transfer negative and still oncogenic. Although this does not entirely rule out the possibility that there are common mechanisms, it is clear that a complete functional transfer mechanism is not an absolute requirement for tumorogenesis. However, the transfer functions of the Ti-plasmid are involved with crown gall in a unique way. Ken et al. (1977), Genetello et al. (1977), and Petit et al. (1978) report that conjugative transfer of the Ti-plasmid in Agrobacterium was induced only when either octopine or nopaline was present. For example, the presence of octopine in the mating system increased the frequency of transfer of an octopine Ti-plasmid from an undetectable level to a frequency of l o - ' transconjugants per recipient. The transfer frequency of nopaline type plasmids

79

CROWN GALL INDUCTION

was much lower at approximately ]OW4 transconjugants per recipient and despite the earlier report (Ken et al., 1976) does not seem to be induced by nopaline (Petit er al., 1978; Ellis and Kerr, 1981). Petit and Temp6 (1978) and Klapwijk et al. (1977a) both studied mutants affected in opine transport and catabolism. They found mutants could be obtained which were constitutive for opine catabolism and showed a high frequency of plasmid transfer in the absence of octopine and they concluded that both processes shared a common regulatory gene but had separate operons. Klapwijk and Schilperoort (1979) extended this work and reported that there was a common repressor system for the two functions. The repressor from a quite different compatible nontumorogenic octopine plasmid was functional in a strain containing a repressor-defective octopine plasmid; this suggests a common mechanism of regulation in both these plasmids. None of these studies on opine-induced transfer and catabolism of the opines has shown any obvious relationship between transfer ability and tumorogenicity. Other functions known or thought to be located on the Ti-plasmid are bacteriophage AP 1 exclusion (Schell, 1975) incompatability between Ti-plasmids (Hooykaas et al., 1980) and periplasmic proteins (Sonoki and Kado, 1978). Despite the functions of the Ti-plasmid detailed above, there is still a very substantial portion of the Ti-plasmid with no known functions. These plasmids are large and may constitute up to approximately 5% of the total DNA of the cell and it is unlikely that they would be so stably maintained in the cell if they did not have some adaptive advantage. This argument suggests that there may well be many functions still to be discovered on the Ti-plasmid. Some of the possibilities are survival and competition in the soil and the ability to utilize other substrates.

E. TI-DNA TRANSFER AND INTEGRATION INTO

THE

PLANTCELL

Almost nothing is known about the transfer mechanism of the Ti-plasmid to a plant cell. Possible mechanisms range from a highly specific transfer mechanism elaborated by the bacteria to a nonspecific release of DNA into the plant wound. Although the need for viable bacteria and apparent specific binding of Agrobacterium to plant cells suggest that a specific transfer mechanism is involved, Davey et al. (1980) report that Petunia protoplasts could be transformed with isolated Ti-plasmid. Transformed cells selected for hormone-independentgrowth were screened for lysopine dehydrogenase, octopine synthesis, and production of overgrowths when grafted onto normal plants. The results were consistent with the cells being transformed crown gall cells. Tests for the presence of T-DNA in the transformed tissue were not reproducibly positive but as the tissue had not been cloned it is possible that the tumors were a mixture of normal and transformed cells. Krens et al. (1982) reported reliable transformation of isolated

80

W.P. ROBERTS

tobacco protoplasts. The transformed nature of these cells was demonstrated by the detection of octopine and T-DNA in them. These experiments suggest that (1) Agrobacterium does not have to process the Ti-plasmid before it is transferred to the plant cell, (2) a highly developed transfer system involving the bacteria is not an absolute requirement, and (3) the Ti-plasmid alone is sufficient for tumorogenesis. In both the in vitro transformation experiments the efficiency of the process was very low requiring a high concentration of plasmid DNA to transform a few cells. This may be due to the lack of an Agrobacterium-mediated transfer and processing system for the T-DNA or it may simply reflect the problem of maintaining and delivering a fragile DNA molecule to the correct plant cell site in an environment that is likely to contain DNase and many other interfering compounds. It is possible that the major role of the bacteria is simply protecting the Ti-DNA and releasing it at a specific site on the plant cell to give a high level of transformation. This hypothesis can be at least partly investigated by refining the in vitro transformation techniques using methods such as liposomes to protect the DNA until it is in the plant cell. It will be interesting to see whether the efficiency of the process can be improved to the level achieved by viable agrobacteria. Very little is known about the mechanism or the precise site of integration of the T-DNA in the plant genome. Although the in vitro transformation experiments above suggest that the total Ti-plasmid enters the cell, it is not possible to eliminate the possibility that the transformation observed in these experiments was due only to a fragment of the Ti-plasmid present in the DNA samples. If this were the case then perhaps the low efficiency of transformation observed in these experiments is due to the low concentration of the required fragment in the Tiplasmid preparations. That the whole Ti-plasmid enters the plant cell is also suggested by the detection of considerable Ti-plasmid sequences in plant cells up to approximately 9 days after infection (Nuti et al., 1980). After this period no Ti-plasmid sequences were detected and it was suggested that the T-DNA was present below the level of detection by the methods used. Although this suggests that the Ti-plasmid is processed in the plant cell to give the T-DNA, it is difficult to evaluate the significance of this type of experiment. The major problem is deciding which cells are being transformed, given that crown gall induction can be a one cell process. On the available evidence then it is not possible to decide where processing of the Ti-plasmid to give the T-DNA occurs. Although it is possible that oncogenic Ti-plasmid functions not on the T-DNA region are involved in this processing, there are no experimental data supporting this suggestion. It may be significant that apparently no specific mechanism is required for the entry and integration of “foreign” DNA into cultured animal cells (Wigler et af.,1979; Hsiung et al., 1980). These reports indicate that prokaryote DNA

CROWN GALL INDUCTION

81

including bacteriophage +x 174 and pBR322 could enter and integrate into the DNA of cultured mouse cells. Plant cell protoplasts lack a complex cell wall and so, at least superficially, are like cultured animal cells, and it is possible that the in v i m transformation experiments with isolated Ti-plasmid involve a similar nonspecific uptake and integration system as has been demonstrated with animal cells. Possibly the wound is required only to remove enough cell wall to expose the cell membrane to allow uptake of released Ti-plasmid DNA. Although the site of integration of the T-DNA into plant DNA varied from tumor line to tumor line and between different strains of agrobacteria (Merlo et al., 1980; Chilton et al., 1980; Zambryski et a!. , 1980; Lemmers et al., 1980; Thomashow et al., 1980a) in all the cases studied, it was inserted in repetitive sequence plant DNA. However, as data were available only for a few tobacco galls, it is impossible to assess the significance of this finding. In tobacco repetitive DNA of some kind makes up approximately 75% of the total genome (Zimmerman and Goldberg, 1977) so it is quite possible that random insertion of T-DNA into the plant genome could give the reported results. There are two conflicting views of the role of repetitive sequence DNA is eukaryotes. One model (Davidson and Britten, 1980) suggests that this DNA is involved in gene regulation but the other suggests that much of it is simply “selfish” DNA making no specific contribution to the phenotype of the organism (Orgel and Crick, 1980; Doolittle and Sapienza, 1980). The Davidson-Britten model would suggest that the simple integration of foreign DNA in a critical regulatory area could cause crown gall. However, the evidence reviewed above showing that the T-DNA can integrate into a number of different sites, has a strongly conserved area essential for tumorogenesis, and is transcribed and translated in the gall suggests that the simple integration of the TDNA into a specific site is not the key event leading to tumorogenesis. If this were true, it would mean that integration of DNA into the plant genome may not induce a gall if the DNA lacks the tumorogenic functions. This hypothesis, if correct, has considerable practical importance for the use of Agrobacterium for genetic engineering in plants (see later section). Assuming that the T-DNA codes for some product, two of the possible mechanisms for tumorogenesis are (1) that the T-DNA codes for the production of plant hormones that are directly responsible for the establishment and maintenance of the tumorous state or (2) the T-DNA codes for products that either directly or indirectly deregulate the normal hormonal control of the cell. Lui and Kado (1979) suggested that IAA production coded for by the Tiplasmid is necessary for gall induction but as the information for this was in an area of the Ti-plasmid that does not integrate into the plant cell, this cannot be involved in tumor maintenance. There are also many studies (see section on physiology of gall) that show that crown gall cells have qualitative and quantitative differences in hormones compared to normal cells, however, none of these

82

W. P. ROBERTS

indicates if this is due to hormone production being directly coded for by T-DNA or that the T-DNA alters the level of hormones produced by the preexisting plant genes. From the available evidence, it is not possible to decide which of the suggested mechanisms of tumorogenesis is correct. F. REVERSIONTO THE NORMALSTATE Any model of crown gall induction has to take into account that tumors can sometimes revert to an apparently normal state and in some cases then revert back to the tumorous condition. Prior to the discovery of the Ti-plasmid and the role of the T-DNA in gall induction there were many reports of reversion of crown gall tissue to apparently normal plants (see review by Lippincott and Lippincott, 1975). However, because there was no definitive test for transformed cells, it is impossible to rule out the possibility that reversion in some of these studies was due to a few normal or habituated cells in the gall, rather than a genuine reversion of transformed cells. It is necessary to clone the gall line and check for opine production and the presence of the T-DNA, both before and after reversion of the tissue. The capacity to revert to normal varies between tumor lines and with the length of time the tumor has been in culture (Turgeon et al., 1976; Sacristan and Melchers, 1977). Braun and Wood (1976) and Turgeon et al. (1976) report regeneration of crown gall tissues into apparently normal plants. In their experiments teratoma shoots were grafted onto tobacco plants and some of the grafted shoots produced flowers and set seeds. The fact that this tissue was still tumorous was demonstrated by the fact that the apparently normal tissue produced nopaline and could grow in the absence of any added growth hormone. The F, generation from the seed produced completely normal plants which no longer produced nopaline or had the ability to grow autonomously in the absence of plant hormones. Yang et al. (1980a) looked at this system at the molecular level. The “normal” plant tissue that appeared on the teratomas grafted onto tobacco plants still retained the T-DNA but at a lower concentration. They were not able to decide from their experiments if the decrease in T-DNA was due to a decrease in copy number of all or some of the T-DNA or to a specific loss of a portion of the T-DNA. They were not able to detect the presence of T-DNA in the F, generation grown from the seed set by the grafted tumor lines and concluded that they were normal because of the lack of T-DNA. Lemmers et al. (1980) looked at the same set of regenerated tumor lines and substantially confirmed the work of Yang et al. (1980a). They report that all of the T-DNA was represented in regenerated plants but there appeared to be some rearrangement of the T-DNA ends. Like Yang et al. (1980a) they found no TDNA in the F, generation from seed of the regenerated plants.

CROWN GALL INDUCTION

83

It is not clear whether the grafting process and meoisis induce the reduction and subsequent loss of T-DNA or whether these events only selected for the cells in the tumor population that had reduced or lost the T-DNA. However, in this regard, O’Schieder (cited in Lemmers et al.. 1980) has reported that approximately 1% of the cells in one of the regenerating cell lines (BT37, Braun and Wood, 1976) is not transformed. The only other molecular-based study of regeneration into normal plants is that of Wullems et al. (1981). Shoots obtained from tumorous tissue and which contained opines were grafted onto healthy plants. Some produced male sterile aberrant flowers and were pollinated by normal pollen. Approximately 40% of the F, seedlings contained opines. They also contained T-DNA. Some of the plants did not contain opines but did contain T-DNA as shown by hybridization experiments. These results indicate that the T-DNA can survive through the process of meiosis.

VIII. Significance of Crown Gall Induction to Agrobacterium The aim of any pathogen is to obtain nutrition and/or a favorable environment from the host species and in this respect Agrobacterium is no exception. However, its method of achieving this aim is unique. The transfer and insertion of the T-DNA into the plant appears to be a specific genetic engineering system by the bacteria to induce the plant to produce opines. These opines are then used as carbon, nitrogen, and phosphorus sources by the bacteria, presumably conferring a selective advantage on those bacteria able to utilize these compounds. This process has been called “genetic colonization” by Schell et a f . ( 1979). Although there are many studies that show that Agrobacterium strains carrying Ti-plasmids can utilize opines in vitro, there are no studies that evaluate the significance of these compounds to the growth and survival of the bacteria in vivo. New and Kerr (1972) report that the ratio of pathogenic to nonpathogenic Agrobacterium was significantly greater around a gall than around a normal root: they suggested that the high number of pathogens around a gall was the cause of gall induction. However, if the opines were significant in enhancing survival and growth of Ti-carrying agrobacteria then high numbers should result from the presence of a gall although, presumably, a certain number of pathogenic bacteria is required to ensure that gall induction takes place. Opines are also involved in induction of Ti-plasmid transfer (Kerr et a f . , 1977; Genetello et a f . , 1977; Petit et al., 1978). This transfer system was first reported by Kerr (1969, 1971) who demonstrated that nonpathogenic agrobacteria acquired pathogenicity when swabbed on galls formed on tomato stems. Up to 75% of single colony isolates of the original nonpathogenic strain became pathogenic

84

W . P. ROBERTS

indicating that either the process was very efficient or that there was very strong selection for converted pathogens on the gall. The efficiency of this process under conditions approaching “natural” conditions suggests that it may be significant in the ecology of Agrobacteriurn. Perhaps the high level of pathogens around galled plants reflects both the selective advantage of opine catabolism and the conjugative spread of the Ti-plasmid to nonpathogens. Although there seems an obvious reason why agrobacteria should induce opine production, an answer to the question, why do they induce a gall is not as obvious. It is clear that opine production is not the tumorogenic principle and that strains inducing tumors containing no (known) opines have been found. It is conceivable that nontransformed plant tissues producing opines could be produced and indeed some of the revertant tumors are like this phenotypically. So it is not unreasonable to suggest that cell transformation and the resultant gall structure must have some adaptive advantage to agrobacteria other than simply opine production. One function of transformation may be to increase the number of cells producing opines and so the concentration of opine available to the bacteria. Another possibility is that the structure of the gall itself provides favorable environmental conditions for the growth of agrobacteria. The capacity for opine catabolism and agrocin 84 sensitivity is not always associated with the Ti-plasmid (Merlo and Nester, 1976; Kerr and Roberts, 1976; Hooykaas et al., 1979; Roberts, 1981) and in at least some strains that can catabolize both nopaline and octopine it is only the nopaline catabolic functions that are transferred with the Ti-plasmid suggesting that the octopine catabolism function is on the chromosome. The non-Ti-plasmid location of these functions may reflect the original genetic source of these functions but, more likely, simply reflects recombination between Ti-plasmids, other plasmids, and the Agrobacteriurn chromosome.

IX. The Evolutionary Origin of Crown Gall As crown gall induction is such a complex process involving the transfer and expression of a defined segment of bacterial DNA in a plant cell, it is difficult to think of a mechanism that would allow the slow development of the system by the mutation and selection of preexisting bacterial genes. The alternative origin is that the main elements of the process existed and, at some stage, combined to give a more or less working scheme. Perhaps the simplest hypothesis would be that the prototype crown gall organism grew around the roots of a plant which produced opines or opine-like substances. By the normal processes of selection and adaption the organism acquired the ability to utilize these compounds and possibly gained a competitive advan-

CROWN GALL INDUCTION

85

tage in this environment. At some stage the bacteria also acquired the genes from the plant to produce these compounds and the mechanism (if there is a specific mechanism) to insert these genes into a suitable host. Of course, the major question with this scheme is how could the bacteria acquire these genes from the plant? Two possible mechanisms are (1) a plant virus picked up plant DNA and subsequently established as a plasmid in the bacteria; and (2) plant DNA perhaps from decaying plants simply transformed into and perhaps recombined with preexisting plasmids or chromosomal DNA to make the prototype Ti-plasmid. On the available evidence, it is difficult to decide which of these mechanisms is more likely or indeed whether either mechanism is likely. Of course, they are not mutually exclusive. The current methods of DNA cloning effectively duplicate the sort of process that underlies both these suggested origins of the Tiplasmid and the mechanism of gall induction. Available techniques allow DNA from any origin to be cloned into either eukaryotic or prokaryotic organisms and with the right manipulations in some cases even be expressed. The soil and indeed most natural environments are constantly being exposed to DNA from many organisms and it is quite possible that there is much more genetic exchange between organisms than is in fact realized (Reanney, 1977). These suggested possible origins of crown gall would imply that ( I ) opines were originally plant products, (2) some of the T-DNA sequences may be present in normal plant DNA, and (3) some of the T-DNA may be homologous with plant viruses. There is some evidence that compounds structurally similar to some of the opines are found in plants. Durzan and Chalapa (1976) report the occurrence of a whole range of substituted guanidines including one compound in callus tissue and seedlings of Pinus banksiana that cochromatographed with octopine. It may be worthwhile looking for the presence of opines in nonhost plants for crown gall on the basis that these plants or their ancestors may have been the origin of these genes and so may have the mechanism to control them. If so they might be resistant to gall induction. It is interesting that Thomashow et al. (1980a) reported some slight hybridization between normal plant DNA and the T-DNA. This has not been confirmed by other workers but it is quite likely that if the TDNA did come originally from plant DNA that it may have diverged in sequence considerably and so detection of any hybridization between it and plant DNA may depend on the selection of the correct plant DNA and the use of low stringency hybridization conditions. There are no reports of homology between T-DNA and plant viruses. The problem with all these predictions is that if one of the mechanisms suggested above is correct these predictions may have only held true at the time the mechanism of crown gall was evolving and evolution since this time may have obliterated any trace of original conditions. Still they are quite testable experimentally and may be worth pursuing. Certainly, it would be surprising if evolu-

86

W.P. ROBERTS

tion has not tried the sort of experiments that many molecular biologists are now attempting. Perhaps crown gall is an experiment that was successful, at least for the bacteria.

X. Agrobacterium and Genetic Engineering A good deal of the interest in the crown gall area is due to the possibility of using the Ti plasmid system for the genetic engineering of plants. The type of experiments suggested is the insertion of desirable genes into the area of the T-DNA of a Ti-plasmid which is not essential for integration of this DNA into the plant cell using available in v i m recombinant DNA technology. This would then be used to induce crown gall tissue containing the new genes. Subsequent manipulations would recover a plant carrying the desirable characteristics in a form that would be agriculturally useful. At least the first steps of this process have been shown to be feasible. Hernalsteens et al. (1980) inserted transposon Tn7 into the area of the T-DNA coding for nopaline synthesis. Tn7 was subsequently maintained in tumor tissue induced by this modified Ti-plasmid. Tn7 has a molecular weight of 9.6 X lo6 so this approach has the potential to incorporate a substantial amount of DNA into the plant genome. Prokaryotic genes such as Tn7 are not likely to be expressed in eukaryotes as they lack the necessary eukaryotic promotor. One approach around this problem is to insert the required gene in the correct phase downstream from the opine promotor in the T-DNA. This should allow the expression of genes lacking eukaryotic promotors. Although the techniques are available to allow the insertion of foreign DNA into plants using this system, how a useful plant can be produced from this is still not obvious. One of the main problems is selecting a plant that has retained the added genes but is no longer tumorous. The work on reversion reviewed earlier suggests that this may not be impossible but that the resultant plants could well be abnormal both in genotype and phenotype. It is not clear whether this is only due to the tranformation process or is at least partly due to tissue culture allowing the development of chromosomal abnormalities, but this is an area that will require much more extensive study before the system can be routinely used for genetic engineering in plants. One possibility is the separation of the transformation process from the insertion process. If these are separate processes, then a reconstructed plasmid carrying only the insertion mechanism, the desired gene, and some selectable marker may well make an efficient system for incorporating new genes into plants. However, at the present time there are few selectable markers that could be used in such an approach.

CROWN GALL INDUCTION

87

One of the problems with using Agrobacterium for genetic engineering in plants is the host range of the disease. Although a range of economically important plants is affected some of the major food crops such as the cereals are not. This may limit utility of an Agrobacterium approach to genetic engineering unless this barrier can be overcome. If the limitation in host-range occurs at the level of survival of Agrobacterium and transfer to the plant then it may be possible to utilize in vitro transfer of the T-DNA by methods similar to those of Davey et al. (1980) and Krens et al. (1982). However, if it reflects a fundamental difference between hosts and nonhosts at the molecular level of T-DNA insertion and transformation then the limited host-range may be a fundamental barrier to the application of a genetic engineering system based on Agrobacterium.

XI. Future Work and Prospects Perhaps the area where least is known about crown gall induction is the process of transfer and integration of the T-DNA into the plant cell. This is an area which is very difficult to study as it involves the interaction of just a few bacteria with a few plant cells. Important questions in this area include: 1. Is the T-DNA involved in transfer and insertion into the plant DNA or is it only involved in tumorogenesis? 2. Does the complete Ti-plasmid enter the plant cell or only the T-DNA? 3. What is the function of the Ti-plasmid outside the T-DNA that is required for gall induction? These and other questions about crown gall induction may be answered by improvements in the in vitro transformation system (Davey et al., 1980; k e n s et al., 1982). Such a system combined with the use of in vitro modified Ti-plasmids and plasmid fragments may be crucial in determining the precise molecular mechanism of crown gall induction. Increases in the efficiency of the in vitro transformation systems may be possible using liposome incorporation of DNA and improvements in techniques for producing and handling protoplasts. There is also very little information on how the products of the T-DNA lead to cell transformation. This is a process that probably has direct bearing on gene regulation and development of plants. Crown gall offers a unique opportunity in that DNA involved in developmental changes can be isolated and manipulated conveniently and work in this area may significantly advance knowledge of developmental processes in plants. Other areas that may well bear further study are the ecology of Agrobacterium

88

W. P. ROBERTS

in relation to the opines, the significance of Ti and other Agrobacterium plasmids transfer in natural systems, and the basis of the host specificity of Agrobacterium.

REFERENCES Albinger, G., and Beiderbeck, R. (1977). Phytopathol. Z. 90, 306. Beiderbeck,R. (1973). Z. Narutforsch. 28, 198. Bevan, M. W., Bennet, D., and Chilton, M-D. (1981). In “Crown Gall-A Plant Cancer Caused by Agrobacterium tumefaciens. ” Kerr, Adelaide. Biemann, K., Lioret, C., Asselineau, J., Lederer, E.. and Polonski, J. (1960). Bull. Soc. Chim. 17, 979. Bomhoff, G. H., Klapwijk, P., Kester, H. C. M., and Schilperoort, R. A. (1976). Mol. Gen. Genet. 145, 177. Braun, A. C. (1947). Am. J . Bor. 34, 234. Braun, A. C. (1952). Growth 16, 65. Braun, A. C. (1954). Brookhaven Symp. Biol. 6, 115. Braun, A. C., and Mandle, R. J. (1948). Growth 12, 255. Braun, A. C., and Wood, H. N. (1976). Proc. Narl. Acad. Sci. U.S.A. 73, 496. and Zaerr, J. B. (1976). Nature (London) 262, 153. Chapman, R. W., Morris, R. 0.. Chilton, M-D., Drummond, M. H., Merlo, D. J., Sciaky, D., Montoya, A. L., Gordon, M. P., and Nester, E. W. (1977). Cell 11, 263. Chilton, M-D., Montoya, A. L., Merlo, D. J . , Drummond, M. H., Nutter, R., Gordon, M. P., and Nester, E. W. (1978). Plasmid 1, 254. Chilton, M-D., Saiki, R. K.,Yadav, N., Gordon, M. P., and Quetier, F. (1980). Proc. Narl. Acad. Sci. U.S.A. 77, 4060. Chilton, M-D., Tepfer, D., Petit, A., David, C., Casse-Delbart, F., and Tempe, J. (1981). Nature (London) 294, 432. Clark, A. G . (1974). Microbios l l A , 29. Cockerham, L. E., and Lundeen, C. U. (1979). Plant Physiol. 64, 543. Colak, O., and Beiderbeck, R. (1979). Phyroparhol. Z. 96, 268. Costantino, P., Hooykaas, P. J. J., den Dulk-Ras, H., and Schilperoort, R. A. (1980). Gene 11,79. Coxon, D. T., Davies, A. M. C . , Fenwick, G. R. Self, R., Firmin, J. L., and Lipkin, D. (1980). Tetrahedron Lett. 21, 495. Davidson, E. H., and Britten, R. J. (1980). Science 204, 1052. Davey, M. R., Cocking, E. C., Freeman, J., Pearce, N., and Tudor, 1. (1980). Plant Sci. Lett. 18, 307. DeCleene, M., and DeLey, J. (1976). Bor. Gaz. 42, 389. DeCleene, M., Van Mulders, R., and DeLey, J. (1980). Caryologiu 33, 151. DeGreve, H., Decraemer. H., Seurinck, J., Van Montagu, M., and Schell, J. (1981). Plasmid 6, 235. De Ley, J. (1968). Annu. Rev. Phytopathol. 6, 63. Depicker, A., Van Montagu, M., and Schell, J. (1978a). Narure (London) 275, 150. Depicker, A., Van Montagu, M., and Schell, J. (1978b). Arch. Inr. Physiol. Eiochim.. 86, 423. Doolittle, W. F., and Sapienza, C. (1980). Nature (London) 284, 601. Drummond, M. H., Gordon, M. P., Nester, E. W., and Chilton, M-D. (1977). Nature (London) 269, 535.

CROWN GALL INDUCTION

89

Drummond, M. H., and Chilton, M-D. (1978). J . Bacteriol. 136, 1178. Durzan, D. J., and Chalupa, V. (1976). Can. J . Bot. 54, 483. Ellis, J. G., and Kerr, A. (1981). In "Crown Gall-A Plant Cancer Caused by Agrobacterium tumefaciens." Kerr, Adelaide. Ellis, J. G., and Murphy, P. J. (1981). Mol. Gen. Genet. 181, 36. Ellis, J. G., Kerr, A., Tempe, J., and Petit, A. (1979). Mol. Gen. Genet. 173, 263. Engler, G., Holsters, M.,Van Montagu, M., and Schell, J. (1975). Mol. Gen. Genet. 138, 345. Firmin, J. L., and Fenwick, G. R. (1977). Phyrochemisrry 16, 761. Firmin, J. L., and Fenwick, G. R. (1978). Nature (London) 276, 842. Garfinkel, D. J., and Nester, E. W. (1980). J . Bacteriol. 144, 732. Gelvin, S. B., Gordon, M. P. Nester, E. W., and Aronson, A. 1. (1981). Plasmid 6, 17. Genetello, C., Van Larebeke, N., Holsters, M., De Picker, A,, Van Montagu, M., and Schell, J. (1977). Nature (London) 265, 563. Glogowski, W., and Galsky, A. G. (1978). Plant Physiol. 61, 1031. Goldmann, A., Tempi:, J., and Morel, G. (1968). C.R. Hebd. Seances Acad. Sci. Ser. D. 162,630. Goldmann, A., Thomas, D. W., and Morel, G. (1969). C.R. Acad. Sci. Paris 268, 852. Graham, P. H. (1964). J. Gen. Microbiol. 35, 5 I I . Gresshoff, P. M., Skotnicki, M. L., and Rolfe, B. G. (1979). J. Bacteriol. 137, 1020. Gurley, W. B., Kemp, J. D., Albert, M. J., Sutton, D. W., and Callis, J. (1979).Proc. Natl. Acad. Sci. U.S.A. 16, 2828. Guyon, P., Chilton, M-D., Petit, A., and TempC, J. (1980).Proc. Natl. Acad. Sci. U.S.A. 77,2693. Hack, E., and Kemp, J. D. (1977). Biochem. Biophys. Res. Commun. 78, 785. Hack, E. and Kemp, J. D. (1980). Plant Physiol. 65, 949. Hamilton, R. H., and Fall, M . 2. (1971). Experientia 27, 229. Hayward, A. C., and Waterston, J. M. (1965). C. M. I. Descriptions of Pathogenic Fungi and Bacteria No. 42. Hepburn, A. G., and Hindley, J. (1979). Mol. Gen. Genet. 169, 163. Herberlein, G . T., De Ley, J., and Tijtgat, R. (1967). J. Bacteriol. 94, 116. Hernalsteens, J. P., De Greve, H., Van Montagu, M., and Schell, J. (1978). Plasmid 1, 218. Hernalsteens, J. P., Van Vliet, F., De Beuckeleer, M., Depicker, A,, Engler, G., Lemmers, M., Holsters, M., Van Montagu, M.,and Schell, J. (1980). Nature (London) 287, 654. Holmes, B., and Roberts. P. (1981). J. Appl. Bacreriol. 50, 443. Holsters, M., Silvia, B., Van Vliet, F., Genetello, C., De Block, M., Dhaese, P., Depicker, A,, InzC, D., Engler, G., Villarroel, R., Van Montagu, M., and Schell, J. (1980). Plasmid 3, 212. Hookyaas, P. J. J., Roobol, C., and Schilperoort, R. A. (1979). J. Gen. Microbiol. 110, 99. Hooykaas, P. J. J., Den Dulk-Ras, H., Ooms, G., and Schilperoort, R. A. (1980). J. Bacteriol. 143, 1295. Hsiung, N., Warrick, H.,DeRiel, J. K., Tuan, D., Forget, B. G., Skoultchi, A,, and Kuckerlapati, R. (1980). Proc. Natl. Acad. Sci. U.S.A. 77, 4852. Ikeda, T., Matsumoto, T., and Noguchi, M. (1976). Phytochemistry 15, 568. Kado, C. I. (1976). Annu. Rev. Phytopathol. 14, 265. Keane, P. J., Kerr, A , , and New, P. B. (1970). Ausr. J . Biol. Sci. 23, 585. Kemp, J. D. (1977). Biochem. Biophys. Res. Commun. 74, 862. Kemp, J. D. (1978). Plant Physiol. 62, 26. Kersters, K., De Ley, J., Sneath, P. H. A,, and Sackin, M. (1973). J . Gen. Microbiol. 78, 227. Kerr, A. (1969). Nature (London) 233, 1175. Kerr, A. (1971). Physiol. Plant Pathol. 1, 241. Kerr, A. (1972). J. Appl. Bacteriol. 35, 493. Kerr, A,, and Panagopoulos, C. G. (1977). Phytopathol. 2. 90, 172. Kerr, A,, and Roberts, W. P. (1976). Physiol. Plant Pathol. 9, 502.

90

W. P. ROBERTS

Kerr, A,, Manigault, P., and TernpC, J. (1977). Nature (London) 265, 560. Klapwijk, P. M., and Schilperoort, R. A. (1979). J . Bacreriol. 139, 424. Klapwijk, P. M., Scheulderman, T., and Schilperoort, R . A. (1977a). J . Bacreriol. 136, 775. Klapwijk, P. M., Oudshoom, M., and Schilperoort, R. A. (1977b). J . Gen. Microbiol. 102, 1. Klapwijk, P. M., Van Breukeleen, J., Korevaar, K., Oorns, G., and Schilperoort, R. (1980). J . Bacteriol. 141, 129. Klein, R. M. (1954). Brookhaven Symp. Biol. 6, 97. Klein, R. M. (1955). Proc. Natl. Acad. Sci. U.S.A. 41, 271. Koekrnan, B. P., Ooms, G . , Klapwijk, P. M., and Schilperoort, R. A. (1979). Plasmid 2, 347. Krens, F., Molendijk, L., Wullems, G., and Schilperoort, R. A. (1982). Nature (London) 296, 72. Kurkdjian, A., Manigault, P., and Beardsley, R. (1969). Can. J . Bot. 47, 803. Lemrners, M., DeBeuckeleer, M., Holsters, M., Zambryski, P., Depicker, A , , Hernalsteens, 1. P., Van Montagu, M., and Schell, J. (1980). J. Mol. Biol. 144, 353. Lentz, C. B., Hodges, T. K., and Matthysse, A. G. (1979). Planta 146, 113. Lipetz, J. (1965). Science 149, 865. Lipetz, J. (1967). Ann. N . Y . Acad. Sci. 144, 320. Lippincott, J. A., and Herberlein, G. T. (1965). Am. J. Bor. 52, 396. Lippincott, B. B., and Lippincott, J. A. (1966). J. Bacteriol. 92, 937. Lippincott, B. B., and Lippincott, J. A. (1969a). J. Bacteriol. 97, 620. Lippincott, J. A,, and Lippincott, B. B. (1969b). J. Gen. Microbiol. 59, 57. Lippincott, J. A., and Lippincott, B. B. (1975). Annu. Rev. Microbiol. 29, 377. Lippincott, B. B., Whatley, M. H., and Lippincott, J. A. (1977). Plant Physiol. 59, 388. Loper, J. E., and Kado, C. I. (1979). J. Bacteriol. 139, 591. Lui, S-T., and Kado, C. 1. (1979). Biochem. Biophys. Res. Commun. 90, 171. Lui, S-T., Perry, K. L., Schandl, C. L., and Kado, C. I. (1981). In “Crown Gall-A Plant Cancer Caused by Agrobacterium tumefaciens. ” Kerr, Adelaide. Matthysse, A. G., Wyman, P. M., and Holmes, K. V. (1978). Infect. Immun. 22, 516. Matthysse, A. G., Holrnes, K. V., and Gurlitz, R. H. (1981). J. Bacteriol. 145, 583. McPherson, J. C., Nester, E. W. and Gordon, M. P. (1980). Proc. Natl. Acad. Sci. U . S . A . 77, 2666. Menage, A., and Morel, G. (1964). C.R. Acad. Sci. Paris 259, 4795. Menage, A., and Morel, G. (1965). C.R. SOC. Biol. 159, 561. Merlo, D. J., and Nester, E. W. (1976). J. Bacteriol. 129, 76. Merlo, D. J., Nutter, R. C., Montoya, A. L., Garfinkel, D. J., Dmmmond, M. H., Chilton, M-D., Gordon, M. P., and Nester, E: W. (1980). Mol. Gen. Genet. 177, 637. Moffett, M. L., and Colwell, R. R. (1968). J . Gen. Microbiol. 51, 285. Moore, L. W., and Warren, G. (1979). Annu. Rev. Phytopathol. 17, 163. Montoya, A. L., Chilton, M-D., Gordon, M. P., Sciaky, D., and Nester, E. W. (1977). J. Bacteriol. 129, 101. Murai, N., and Kemp, J. D. (1982). Proc. Natl. Acad. Sci. U.S.A. 78, 4344. Murphy, P. J., and Roberts, W. P. (1979). J. Gem Microbiol. 114, 207. New, P. B. (1972). PhD. Thesis, University of Adelaide. New, P. B., and Kerr, A. (1972). J. Appl. Bacteriol. 35, 279. Nuti, M. P., Ledeboer, A. M., Durante, M., Nuti-Ronchi, V., and Schilperoort, R. A. (1980). Plant Sci. Lett. 18, 1. Ohyama, K., Pelcher, L. E., Schaefer, A., and Fowke, L. C. (1979). Piant Physiol. 63, 382. Ooms, G., Klapwijk, P. M., Poulis, J. A,, and Schilperoort, R. A. (1980). J. Bacreriol. 144, 82. Ooms, G., Hooykaas, P. J. J., Moolmaar, G., and Schilperoort, R. A. (1981). Gene 14, 33. Orgel, L. E., and Crick, F. H. C. (1980). Nature (London) 284, 604. Panagopoulos, G. C., and Psallidas, P. G. (1973). J. Appl. Bacteriol. 36, 233.

CROWN GALL INDUCTION

91

Petit, A., Delhaye, S.. Tempe, J., and Morel, G. (1970). Physiol. Vegetule 8, 205. Petit, A . , and Tempe, J. (1975). C.R. Hebd. Seances Acud. Sci. Ser. D . 281, 467. Petit, A., and Tempi, J. (1978). Mol. Gen. Genet. 167, 147. Petit, A., Tempi, J., Kerr, A., Holsters, M., Van Montagu, M., and Schell, J. (1978). Nature (London) 271, 570. Petit, A., Tepfer, D., David, C., andTernpi, J. (1981). In “CrownGall-A Plant Cancercaused by Agrobucrerium tumefuciens. ” Kerr, Adelaide. Peterston, J. B., and Miller, C. 0. (1977). Plant Physiol. 59, 1026. Phillips, R., and Butcher, D. N. (1979). Phytochemistry 18, 791. Pueppke, S. G., and Benny, U. K. (1981). Physiol. Plant Pathol. 18, 169. Radosevich, J . , and Galsky, A. G. (1979). Plant Cell Phvsiol. 20, 185. Reanney, D. C. (1977). Brookhaven Symp. Biol. 29, 248. Riker, A. J. (1923). J. Agric. Res. 25, 119. Riker, A. J., and Banfield, W. M. (1932). Phvtopurhology 22, 167. Riker, A. J., and Berge, T. 0. (1935). Am. J. Cancer 25, 310. Agric. Riker, A. J., Banfield, W. M., Wright, W. H., Keitt, G. W., and Sagen, H. E. (1930). .I. Res. 41, 507. Roberts, W. P. In “Crown Gall-A Plant Cancer Caused by Agrobacterium tumefaciens. Sydney. Rogler, C . E. (1980). Proc. Nut/. Acad. Sci. U.S.A. 77, 2688. Rogler, C. E. (1981). Plant Physiol. 68, 5 . Rutherford, B., Jalovec, L., Nugent, D . , and Galsky, A. 0. (1977). Plant Physiol. 19, 1103. Sacristan, M. D., and Melchers, G. (1977) Mol. Gen. Genet. 152, I 1 I . Schell, J. (1975). In “Genetic Manipulations with Plant Material” (L. Ledoux, ed.), p. 163. Plenum, New York. Schell, J., Van Montagu, M., De Beuckeleer, M., De Block, M., Depicker, A,, De Wilde, M., Engler, G., Genetello, C., Hemalsteens, J. P., Holsters, M., Seurinck, J., Silva, B., Van Vliet, F., and Villavvoel, R. (1979). Proc. R. SOC. London B . 204, 251. Sciaky, D., Montoya, A. L., and Chilton, M-D. (1978). Plusmid 1, 238. Schilde-Rentschler, L., Gordon M. P., Saiki, R., and Melchers, G. (1977). Mol. Gen. Genet. 155, 235. Schweitzer, S., Blohrn, D., and Geider, K. (1980). Plasmid 4, 196. Scott, 1. M., Firmin, J. L., Butcher, D. N., Searle, L. M., Sogeke, A. K., Eagles, J., March, J. F., Self, R., and Fenwick, G. R. (1979). Mol. Gen. Genet. 176, 57. Scott, I. M., Morgan, R., and McGaw, B. A. (1980). Planru 149, 472. Smith, E. F., and Townsend, C. 0. (1907). Science 25, 671. Sonoki, S . , and Kado, C. I. (1978). Proc. Natl. Acud. Sci. U.S.A. 75, 3796. Speranza, A., and Bagni, N. (1977). Z . Pflanzenphysiol. 81, 226. Stuchbury, T., Palni, L. M., Morgan, R., and Waveing, P. F. (1979). Plantu 147, 97. Sule, S. (1978). J. Appl. Bacteriol44, 207. Tate, M. E., Ellis, J. G., Ken, A., Tempe, J., Murray, K. E., and Shaw, K. J. (1982). Carbohydrate Res.. submitted. Ternpi, J . , Petit, A., Holsters, M., Van Montagu, M., and Schell, J. (1977). Proc. Nut/. Acad. Sci. U.S.A. 74, 2848. Therman, E., and Kulipa-Ahvennierni, S. (1971). Physiol. Plant. 25, 178. Thomashow, M. F., Nutter, R., Montoya, A. L., Gordon, M. P. and Nester, E. W. (1980a). Cell 19, 729. Thornashow, M. F. , Nutter, R., Postle, K., Chilton, M-D., Blattner, F. R., Powell, A., Gordon, M. P., and Nester, E. W. (1980b). Proc. Nail. Acud. Sci. U.S.A. 77, 6448. Thomashow, M. F., Panagopoulos, C. G., Gordon, M. P., and Nester, E. W. (1980~).Nature (London) 283, 794.

92

W. P. ROBERTS

Thomashow, M. F., Knauf, U. C., and Nester, E. W. (1981). J . Bacreriol. 146, 486. Turgeon, R., Wood, H. N., and Braun, A. C. (1976). Proc. Natl. Acad. Sci. U.S.A. 73, 3562. Van Larebeke, N., Zaenen, I., Teuchy, H. and Schell, J. (1973). Arch. Inr. Physiol. Biochim. 81, 986. Van Larebeke, N., Engler, G . , Holsters, M., Van den Elsacker, S., Zaenen, I., Schilperoort, R. A,, and Schell, J . (1974). Nature (London) 252, 169. Van Larebeke, N., Gentello, Ch., Schell, J., Schilperoort, R. A., Hermans, A. K., Hernalsteens, J. P., and Van Montagu, M. (1975). Narure (London) 255, 742. Van Montagu, M., Holsters, M., Zambryski, P., Hemalsteens, J. P.,Depicker, A., De Beuckeleer, M., Engler, G., Lemmers, M., Willmitzer, L., and Schell, J. (1980). Proc. R. Sac. London B 210, 35 1 . Walker, J. C. (1969). “Plant Pathology.” McGraw-Hill, New York. Watson, B., Currier, T. C., Gordon, M. D., Chilton, M-D., and Nester, E. W. (1975).J . Bacteriol. 123, 255. Whatley, M. H., and Spiess, L. D. (1977). Plant Physiol. 60,765. Whatley, M. H., Bodwin, J. S., Lippincott, B. B., and Lippincott, J. A. (1976).Infecr. Immun. 13, 1080. Whatley, M. H., Margot, J. B., Schell, J., Lippincott, B. B., and Lippincott, J. A. (1978).J. Gen. Microbiol. 107, 395. White, F. F., and Nester, E. W. (1980a). J. Bacteriol. 141, 1134. White, F. F., and Nester, E. W. (1980b). J . Bacreriol. 144, 710. White, 0. L. (1972). J. Gen. Microbiol. 72, 565. White, P. R., and Braun, A. C. (1941). Science 94, 239. White, P. R., and Braun, A. C. (1942). Cancer Res. 2, 597. Wigler, M., Sweet, R., Sim, G. K., Wold, B., Pellicer, A., Lacy, E., Maniatis, T., Silverstein, S., and Axel, R. (1979). Cell 16, 777. Willmitzer, L., De Beuckeleer, M., Lemmers, M., Van Montagu, M., and Schell, J. (1980). Nature (London) 287, 359. Wullems, G. J., Molendijk, L., Ooms, G., and Schilperoort, R. A. (1982). Proc. Natl. Acad. Sci. U.S.A. 78, 4344. Yadav, N. S., Postle, K., Saiki, R. K., Thomashow, M. F., and Chilton, M-D. (1980). Nature (London) 287, 458. Yang, F., Montoya, A. L., Merlo, D. J., Drummond, M. H., Chilton, M-D., Nester, E. W., and Gordon, M. P. (1980a). Mol. Gen. Gener. 177, 707. Yang, F., McPherson, J. C., Gordon, M. P., and Nester, E. W. (1980b). Biochern. Biophys. Res. Commun. 92, 1273. Yang, F., Merlo, D. J., Gordon, M. P., and Nester, E. W. (1980~).Mol. Gen. Gener. 179, 223. Zaenen, I., Van Larebeke, N., Teuchy, H., Van Montagu, M., and Schell, J. (1974). J . Mol. Biol. 86, 109. Zambryski, P., Holsters, M., Kruger, K., Depicker, A., Schell, J., Van Montagu, M., and Goodman, H. M. (1980). Science 209, 1385. Zimmerman, J. L., and Goldberg, R. B. (1977). Chromosoma 59, 227.

INTERNATIONAL REVIEW OF CYTOLOGY, VOL. 80

The Molecular Cytology of Wheat-Rye Hybrids* R. APPELS Division of Plant Industry, CSIRO, Canberra, Australia

1. Introduction

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

11. The Genetic Relationship between Rye and Wheat Chromosomes

.

Ill. The Molecular Structure of Rye and Wheat Chromosomes . . . . . . . IV . Translocations in Wheat-Rye Addition or Substitution Lines . . . . . V. Polymorphisms in Regions of the Chromosomes Containing Repeated Sequence DNA .................................. VI . The Biological Effects of Rye Chromosomes (or Rye Chromosome Fragments) in Wheat-Rye Hybrids: Specific Effects Related to Heterochromatin ......................................... VII. The Possible Origins of Polymorphism in Rye Heterochromatin . . . VIII. Prospects ............................................... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

93 94 99 109 113

121 123 I27 127

I. Introduction Biologists have, for many years, combined the genome of one species with that of another species to produce a new combination of genomes interacting in the one cytoplasm. In animal cell biology this has been achieved mainly in tissue culture by the induced fusion of cells or using microinjection techniques. This type of manipulation provided opportunities for mapping chromosomes genetically (Ringertz and Savage, 1976), defining the biological prerequisites for expression of genes characteristic of a differentiated state (Ringertz and Savage, 1976; Willing etal., 1979; Allan and Harrison, 1980; Halaban et al., 1980), and analyzing the biochemistry of nuclei undergoing extensive structural change in a short period of time (Appels and Ringertz, 1975; Linder et al., 1981). In plants, interspecies combinations of genomes have also been achieved by cell fusion (Schieder and Vasil, 1980) as well as by natural means using cross-pollination followed by chromosome doubling with colchicine (Blakeslee and Avery, 1937). The manipulation of plant genomes has attained practical levels in providing commercially grown wheat with resistance to stem rust from Elytrigia pontica (Podp.) McGuire et Dvorak (= Agropyron elongaturn) (Host.) P.B. (Sr26), from *The term hybrid is used in a broad sense to include derivatives of primary wheat-rye hybrids such as triticales and wheat-rye addition and substitution lines. 93 Copyright 0 1982 by Academic Press. Inc. All rights of reproduction in any form reserved. ISBN 0-12-864480-0

94

R . APPELS

Secale cereale (Sr31), from Triticum timopheevi Zhuk ( S I ~ C as ) , well as from diploid wheats (reviewed by Knott and Dvorak, 1976). The development of wheat-rye hybrids into the new crop triticale ( X Triticosecale Wittmack) highlights the potential for combining the genomes of different species (for recent reviews see Gustafson, 1976, 1982). Attempts to form hybrids between wheat and other cereals such as rye date back to 1876 (for reviews see O’Mara, 1953; Briggle, 1969). Present-day cereal cytogenetics allows alien chromosomes to be introduced into wheat in an efficient manner (Sears, 1975) and it is the aim of this article to examine in detail how such studies have allowed new features of the wheat and rye genomes to be uncovered. Emphasis is placed on the analysis of rye chromosomes because the production of wheat-rye hybrids has been particularly significant in allowing the assignment of genetic and cytogentic markers to specific chromosomes.

11. The Genetic Relationship between Rye and Wheat Chromosomes An important use of wheat-rye substitution lines has been in the investigation of the genetic relationship between wheat and rye chromosomes. The 21 chromosomes of hexaploid wheat can be partitioned into seven homoeologous groupings (each containing three chromosomes) using nullisomic-tetrasomic tests (Sears, 1966). These groups are designated 1-7 for each of the A, B, and D genomes of hexaploid wheat. Homoeology is defined by the degree to which a given chromosome can overcome nullisomy for another chromosome and is thus a measure of the genetic similarity of chromosomes. Although the sensitivity of the assay depends on the severity of the nullisomic phenotype which is being rescued, the chromosomes of Secale cereale (rye) have been placed into the homoeologous groups defined in wheat (Sears, 1968; Gupta, 1971; Koller and Zeller, 1976). The extension of homoeologous chromosome relationships beyond hexaploid wheat requires the production of wheat-alien chromosome addition lines which can then be utilized for the production of substitution lines where an alien chromosome substitutes for a wheat chromosome (Sears, 1975). This procedure can generate artifacts, as discussed later, but has clearly shown that rye chromosomes are sufficiently closely related to wheat chromosomes to allow substitution to occur. This relationship is also observed when specific molecular markers such as allozymes (Hart, 1979), seed storage proteins (Shepherd, 1973; Lawrence and Shepherd, 1980), and genomic arrangement of the DNA sequences for 5 S, 18 S, and 26 S rRNA genes (Appels et al., 1980) are examined. The comparison of 1R and 1B is particularly striking since the relative chromosomal positions of the 5 S and 18 S-26 S rDNA regions have been retained. Figure 1 summarizes a number of markers which are located in the same approximate position in the respective wheat and rye chromosomes. Figure 1 also lists the genetic markers on rye

WHEAT-RYE HYBRIDS

95

chromosomes which do not show clearly defined homoeology with wheat; the extent of this list (see also Schlegel, 1982) indicates that homoeologous relationships have not been strictly maintained during the course of evolution, a point taken up later in this section. The differences in homoeologous genetic markers such as seed storage proteins and allozymes are detectable but are not so great as to prevent their expression, and the association of heterologous subunits of enzymes into active molecules in wheat-alien chromosome combinations (Hart, 1979). The 18 S-26 S rDNA and 5 S rDNA regions of 1R and 1B are similar in the size of the basic repeating unit as estimated using restriction enzyme sites located within the conserved gene sequences (Fig. 1). Spectacular differences are, however, observed when repeated sequences such as those found in heterochromatin (discussed later) or in the spacer regions of the 18 S-26 S rDNA and 5 S rDNA regions, are examined. The 5 S rDNA regions of rye and wheat are strikingly different when a restriction enzyme such as Hue111 is used to probe the region since the enzyme has cleavage sites located in the spacer region (Appels et al., 1980; Gerlach and Dyer, 1980). Even within wheat the two basic types of repeating unit of 5 S rDNA are highly diverged with respect to their spacer sequences (Gerlach and Dyer, 1980). The 18 S-26s rDNA spacer region of wheat has been mapped in detail (Appels and Dvorak, 1982). The most significant feature of the spacer region is the presence of 11-13 130-bp repeat units analogous to the repeated structure seen in animal spacer regions (Federoff, 1979; Rae et al., 1981; Kunz et al., 1981). The two major rDNA locations on 1B and 6B can be distinguished by the length of the spacer region but sequence divergence between these two chromosomes is very small compared to their divergence from the corresponding sequence in 1R. When the 130-bp repeated sequence is used to probe the rDNA region, high levels of divergence are observed between 1R and 1B (or 6B) (Fig. 2). The melting curves in Fig. 2 are the result of hybridizing a radioactive 130-bp sequence to either wheat (cytogenetic stocks of cv. Chinese Spring) or rye (cv. Imperial) DNA immobilized on nitrocellulose filters and determining the temperature at which radioactivity is lost from the filter. The difference observed in Fig. 2 is much greater than observed in the comparison of the entire genomes of wheat and rye by analyzing isolated DNA; on average, the genomes of rye and wheat have diverged to give a T,,, difference (in the heterologous hybridization reaction) of only 4°C (Bendich and McCarthy, 1970; Flavell et al., 1977). Since this is an average of conserved gene regions and diverged spacer regions this is not inconsistent with the large T , difference observed in Fig. 2. Translocations during the course of evolution tend to diffuse the homoeologous relationships with respect to the position of a particular locus on a chromosome. Secale cereale differs from the more primitive Secale montanum (which is considered to be more closely related to wheat, Stutz, 1972) in that at least three chromosomes have been involved in translocations (Price, 1955; Riley, 1955; Heemert and Sybenga, 1972). Consistent with this observation, the

96

R. APPELS

FIG. 1. Summary of homoeologous genetic markers on rye chromosomes. The data used were taken from Appels et al. (1980). Barber et al. (1968), Gerlach and Dyer (1981), Hart (1979). Irani and Bhatia (1972), Koller and Zeller (1976), Lawrence and Shepherd (1980), May (1974), Payne and Dyer (1976), Rao and Rao (19801, Sanchez-Monge et al. (1979), Shepherd (1973), and Zeller (1976). Phosphoglucose isomerase-1 has been located to the short armof 1R (R.Koebner, unpublished data). With respect to genetic markers on rye chromosomes the reader is also referred to Schlegel (1982) and the work of J. de Vries and J. Sybenga at the Department of Genetics, Agricultural University, Wageningen (in preparation). The pachytene karyotype of rye chromosomes is from

97

WHEAT-RYE HYBRIDS aminopeptidare- 1 glutamic oxaloacetic transaminare - 2 endopeptidare - 1

6R is:

6 - phoophogluconatedehydrogenase

:zc:$$::c :::1st :’fi:;:f ms: $$t;::u$*y*

: :: :I e: r; f?;:

7R

:J

:::::::::::

I:

I.: ::: :::)5?:W:: :: L:<: is: L l i I

acid phorphatase- 1

*...... a.-. , ~ ; : ~ ~ : : ~ : ~ : ~ : . ~ : : : ~ .. ~ : ~ :. ~ ~~ & ::;:= :* ::r:<:t:*:J‘*:a ? n : ’ ::

I:‘:::

::*

:

Rye (~rw or chromosomes not showing clearly defined hornowlogy

1R

Stem rust resistance powdery mildew resistance stripe rust resistance.

2R

LMW seed storage protein

-

3R 4R

Reduced tillering culm length

5R

Hairy neck esterase.

-

6R

-

7R

B chromosome

rw:

mm

trl

-:a*:r::s

no genetic markers arigned

::s.::xz% :::*:I:

:I

Lima-de-Faria (1952), with a “best-fit” of the chromosomes to current assignements of rye chromosomes (see legend to Fig. 6) being made by the author. The rye chromsomes 4R, 6R, and 7R do not have a simple homoeologous relationship to wheat chromsomes (see Fig. 3). The comparisons with respect to the allozymes and storage proteins refer to the entire homoeologous wheat group unless the wheat chromsome is specificed-in the latter case the comparison applies only to the wheat chromosome indicated. If the allozyme is placed over the centromere region it indicates no assignment has been made regarding its location on the long or short arm. Note that the differences between IR and 1B with respect to the sequence of the 5 S rRNA gene are at a similar level to those seen between 5 S rRNA genes in wheat per se (Gerlach and Dyer, 1980); the spacer regions of two wheat 5 S rRNA genes have been completely sequenced but not included in the diagram. The sequences (noncoding strands are shown) of the 5 S rRNA genes can be readily alligned if some small deletions are allowed for (see Gerlach and Dyer, 1980, for further discussion). The distribution of 18 S-26 S rDNA between homoeologous chromosomes is not simple in that in wheat they may be found on IA, IB, 6B, and 5D (Flavell and Smith, 1974). In rye they are found only on IR. The juxtapositioning of 5 S genes distal to the 18 S-26 S rDNA region is found only on 1R and 1B. The rDNA site on IA of wheat is only a major site in the variety of Capelle-Deprez (Flavell and Smith, 1974); in other varieties the IA site appears to have fewer rDNA cistrons than 5D (which has only 8-10% of the total rDNA) if any are present at all.

98

R. APPELS

f

loor

,L

/'

/ [

I

i

I II I

130 bp sequence (wheat)

- RYE

I

--- WHEAT

I

I

Ternp("C)

FIG.2. Cross-hybridization of a wheat rDNA spacer sequence to rye DNA. A component of the spacer region which is present 11-13 times (130 bp in length) was isolated and labeled with 32P (Appels and Dvorak, 1982). This radioactive probe was then hybridized to either wheat or rye DNA immbolized on nitro cellulose filters under standard conditions. The melting point of the hybrids formed was then determined (in 3 X SSC, 50% formamide).

chromosomes CR, DR, and FR (from the Imperial rye additions to Chinese Spring wheat, also referred to as 4R, 7R, and 6R, respectively) have been assigned mixed homoeologies (Koller and Zeller, 1976). Koller and Zeller (1976) suggest that DR of S. cereale includes segments from 4R, 6R, and 7R as

4R

4RL

CR=4R

v

7R

6RL

DR=7R

S.montanum

6R

6RS

7RL

FR=6R

S.cereale

FIG. 3 . Homoeologous relationship between S. cereale and S. monranum chromosomes. Diagram was redrawn from Koller and Zeller (1976). The relationship between the S. cereale chromosomes and its progenitor (very likely S. monranum) is partly based on the assignment of allozymes. The allozymes alcohol dehydrogenase and acid phosphatase are homoeologous group 4 markers in wheat but are located on 4R and 7R, respectively (Hart, 1978). The allozyme 6-phosphogluconate dehydrogenase is located on 4RL and 6RL (Rao and Rao, 1980) and shows homoeologous relationships with groups 7 and 6 of wheat, respectively (Hsam, Huber, and Zeller, submitted). The wheat homoeologous group 7 marker, an endopeptidase, is on 6RL while the wheat homoeologous group 6 marker, amino peptidase, is also on 6R (Hart, 1978).

WHEAT-RYE HYBRIDS

99

defined in S. montanum while CR consists of 4R and 7R, and FR consists of 6R and 7R. The representation in Fig. 3, taken from Koller and Zeller (1976), summarizes the data from the compensation shown by DR and CR (hereafter referred to as 7R and 4R, respectively) for various wheat homoeologous chromosomes as well as genetic markers. Although an approximation, the representation in Fig. 3 serves to illustrate the complexity of the genetic relationship of some rye chromosomes to those of wheat. The close genetic relationship between rye and wheat has allowed the successful development of triticales which have the genomic composition of AABBRR (cf. bread wheat AABBDD). The rye chromosomes from hexaploid triticale (AABBRR) can, during the course of a crossing program (for spring-type triticales), be substituted by D genome chromosomes to produce plants with the genome constitution AABBRWDD. Additional changes also appear to occur to rye chromosomes in a wheat background. But before discussing these in detail, however, the following sections outline the complexities which arise in analyzing the rye chromosomes in wheat-rye addition or substitution lines, the molecular nature of certain regions of chromosomes and their associated polymorphisms.

111. The Molecular Structure of Rye and Wheat Chromosomes

The genetic structure of the chromosomes discussed in the preceding section included loci which are classified as occurring in repeated sequence families (e.g., rDNA sequences, seed storage proteins) as well as those which are likely to be in the so-called single copy class of DNA sequences (e.g., allozymes). In rye and wheat only 10-20% of the genome can be assigned, biochemically, to the slowly renaturing (i.e., in the single copy sequence) category while much of the genome belongs to the repeated sequence category (Ranjekar et a l . , 1974, 1976; Flavell and Smith, 1976; Smith and Flavell, 1977). The kinetic analyses of genome organization have shown that on average repeated sequences are interspersed among unrelated sequences. The biological interpretation of the repeated sequence class of DNA is not clear, however (Moyzis et a l . , 1981), and this fact taken together with the problem of the analysis yielding only a genomic “average picture” has meant the kinetic analysis has not provided useful information about any specific region of the genome. The exception to this is the discovery of a very rapidly reannealing class of DNA sequence. In both rye and wheat this class of DNA constitutes 4-10% of the genome and although interpreted to be composed largely of sequences capable of “foldback” renaturation (Smith and Flavell, 1977; Flavell and Smith, 1976), this class also contains long tandem arrays of simple, repeated sequences (Appels et a l . , 1978). (Simple, repeated, sequences are capable of “foldback” renaturation as is the case with,

100

R. APPELS

for example, the Drosophifa mefanogaster 1.672 g/cm3 satellite which has a tandemly repeated many thousands of times.) The location of sequence $T:AX the very rapidly reannealing DNA sequences is largely within the cytologically defined heterochromatic regions of the chromosomes in both rye (Appels et al., 1978; Fig. 4a) and wheat (Gerlach and Peacock, 1980). The isolation of the very rapidly reannealing DNA sequences thus provided one of the first opportunities for a molecular description of a cytologically identifiable region of the cereal chromosomes. The other example is the description of the rDNA region associated with the secondary constrictions (Mohan and Flavell, 1974; Appels et al., 1980; Gerlach et a f . , 1980; Fig. 4b); as discussed later contemporary DNA cloning techniques are rapidly expanding the molecular description of specific regions of the cereal genomes. The distribution of very rapidly renaturing sequences correlates with the distribution of the major C-banded regions (Appels et a f . , 1978; Gerlach and Peacock, 1980; Jones and Flavell, 1982), but the nature of the contribution of DNA sequence composition to C-banding is not clear. The use of molecules with defined base specificity of binding, combined with counterstaining with a compound of complementary base specificity (Schweizer, 1981), does not reveal the terminal heterochromatic regions of the rye chromosomes (Schweizer, 1979). This lack of distinctive base composition (relative to the rest of the genome) of the heterochromatic regions is consistent with the fact that the very rapidly renaturing sequences fail to be resolved as buoyant density satellites in various types of cesium salt gradients (Appels et a l . , 1978). In wheat a major heterochromatic sequence can be isolated as a buoyant density satellite (Dennis et a f . , 1980). The genomic arrangement of the various forms of the sequence is, however, complex as evidenced by the fact that only a proportion of the sequence resolves as a satellite (Dennis et al., 1980) and digestion of total DNA with MbO I1 and probing with the sequence by hybridization gives a very complex pattern (R.Appels, unpublished). This may account for the lack of resolution of wheat chromosomal heterochromatin using base-specific fluorescent compounds (Schweizer, 1979). The rDNA region is revealed by base-specific fluorescent compounds (Schweizer, 1979; Fig. 5 ) ; consistent with this observation the region has a well-defined genomic structure (see Fig. I), and can be isolated as a buoyant density satellite in actinomycin-D/CsCl gradients (Hemleben and Grierson, FIG.4. In situ location of isolated sequences to rye chromosomes. (a) Rapidly renaturing (Cot = 0.01) sequences from rye DNA were used to provide a template for RNA polymerase to synthesize a [3H]cRNA probe. In situ hybridization locates the complementary sequences on the chromosomes. Unpublished photograph by C. May. (b) Ribosomal RNA labeled with 1251 was used to locate Complementary sequences by in situ hybridization. The inset shows the rye chromosome in the IR addition to wheat where double labeling with the rye-specific heterochromatic sequence probe [shown in (a)] mixed with the ribosomal RNA probe was used to characterize the chromosome. Taken from Appels et a / . (1980).

WHEAT-RYE HYBRIDS

101

102

R. APPELS

FIG. 5 . Fluorescent staining of rye chromosomes. Rye (S. cereale cv “Petkus Spring”) mitotic chromosomes were stained with chromomycin and counterstained with distamycin A and DAPI. Taken from Schweizer (1979). Photograph kindly supplied by D. Schweizer.

1978). Hoechst 33258 is the only fluorescent compound to date which preferentially stains rye heterochromatin (Sarma and Natarajan, 1973; Fig. 6). This compound appears to have a greater requirement with respect to the sequence of the DNA to which it will bind (at least three A.T pairs in the binding site, Muller and Gautier, 1975). Sequencing of the major heterochromatic sequence of rye

FIG.6. Fluorescent staining of rye chromosomes in wheat-rye hybrid. Mitotic chromosomes of a triticale were stained with Hoechst 33258. The fluorescent bright regions are the telomeric heterochromatin regions of rye chromosomes. Taken from Sarma and Natarajan (1973). Photograph kindly supplied by A. J. Natarajan.

WHEAT-RYE HYBRIDS

103

(Appels et al., 1981) has shown a predominance of three or more adjacent A.T pairs which would account for Hoechst 33258 preferentially staining these regions. The structure of heterochromatin which is detected by C-banding (Gill and Kimber, 1974; Verrna and Ress, 1974; Vosa, 1974; Fig. 7) apparently results from the repeated nature of DNA sequences present in this region of the chromosome. The DNA sequence per se does not appear to be an important variable and whatever the aspect of the chromatin which is detected (Sumner, 1980), it is transferred with the chromosomes when they are placed in a foreign cytoplasm. Although some cytological aspects of rye chromosomes appear to change upon incorporation into wheat-rye hybrids (e.g., suppression of the secondary constrictions, reviewed in Gustafson, 1982) heterochromatin is easily assayed in these instances by C-banding (reviewed in Gustafson, 1976). It is interesting that in wheat-rye hybrids, on average, the same chromosomal proteins must be bound to both wheat and rye chromosomes. Rye chromosomes in wheat-rye hybrids are thus very likely to have proteins bound to them which are different from the proteins bound in rye per se. Differences between wheat and rye H1 histones exist (LaRue and Pallotta, 1976; Spiker, 1976) and no doubt differences exist in other chromosomal proteins. Whether rye chromatin fine structure differs when rye chromosomes are in wheat-rye hybrids as compared to rye is an area of research just becoming amenable to analysis. The C-banding of cereal chromosomes has been extensively used in the analysis of rye chromosomes present in wheat-rye hybrids, particularly triticales (Gustafson, 1976). The extensive use of this technique has provided important data on the levels of polymorphism for this aspect of chromosome structure and is treated in Section V.

FIG. 7 . C-banding of rye chromosomes. C-banding of mitotic chromosomes (modified from Gill and Kimber, 1974). The chromosomes have been assigned their designations of IR etc. based on the most common C-banding karyotype observed in rye, as agreed upon at a workshop on rye genetics and cytology in March 1982, Wageningen (The Netherlands). The C-banding of B-chromosomes has also been described by Singh and Robbelen (1973, and is the same as that shown.

104

R. APPELS

The recovery of specific DNA sequences using the recombinant DNA techniques has been applied to rye and wheat and, in combination with the analysis of wheat-rye addition or substitution lines, has allowed a high resolution analysis of certain regions of rye chromosomes. The cloning of rye heterochromatic sequences was carried out, in parallel, in the CSIRO, Canberra (Appels et a l . , 1979, 1981) and in the Plant Breeding Institute, Cambridge (Bedbrook et a l . , 1980a). The approaches used are summarized in Fig. 8. The protocol used by Bedbrook et a l . , (1980a) to define the fraction of DNA of interest is specific to the Secale species used. The protocol developed by Appels et al. (198 1) utilizes the kinetic categories of the different DNA sequences discussed earlier to define the DNA of interest, and is a more general approach. Application to systems other than Secale, particularly to define species-specific sequences, appears to be very useful (R. Appels, unpublished observations). The use of synthetic linkers ligated onto the DNA segment of interest is at present the most efficient way of cloning the respective segment if standard restriction enzyme sites such as EcoRI, HindIII, BamHI, Pst, or Cla are absent. The major rye heterochromatic DNA sequence (recovered in the plasmid pSc 74, Bedbrook el a l . , 1980a; pSc 7235, Appels et a l . , 1981) as well as other, minor, sequences (e.g., pSc 7, pSc 33, pSc 119, Bedbrook et al., 1980a; pSc 1628, Appels et a l . , 1981) have been isolated. The distribution of the major heterochromatic DNA sequence is almost specific to the genus Secale (see Section VII for a further discussion) and has been used for a fine structure analysis of individual heterochromatic regions. In the large heterochromatic regions the sequence is distributed throughout the cytologically defined heterochromatin (Fig. 9). Analysis of the hybridization products formed between the cloned sequence and rye DNA or the DNA from wheat-rye chromosome substitution or addition lines shows that the melting points of the hybrids formed are indistinguishable from the homologous hybrid formed with the cloned sequence (Fig. 10). This observation taken together with data from the distributions of specific restriction enzyme cleavage sites within the sequence arrays of different chromosomes (Fig. 11) suggests that all the chromosomes of rye contain various blocks of sequences which are virtually identical. These sequences account for approximately lo5 kb of DNA (from a total genome of 8X lo6 kb of DNA, Bennett, 1972) and thus represent a significant proportion of the genome. One of the minor sequences (pSc 33, Bedbrook et al., 1980a) shows similar properties to the major heterochromatic sequence in that the structure of DNA containing this sequence is also the same an nonhomologous chromosomes. Wheat and barley heterochromatic DNA sequences show properties similar to those of the rye sequences with respect to chromosomal distribution. DNA segments from wheat and barley heterochromatin have been cloned by adding HindIII synthetic linkers to DNA isolated as a buoyant density satellite in cesium sulfate/silver nitrate gradients (Peacock et a l . , 1981). Using in situ hybridization the sequence located to nonhomologous chromosomes (Dennis et al., 1980); the

105

WHEAT-RYE HYBRIDS

2. Appels u4.11981 j

1. Bcdbrook gtd.11980a) Comparison of S. cereale and S. sylvestre defines a fraction of DNA undigested by Hind III which accounts for the difference in heterochromatic content between these species.

Heterochromatic sequences (in the heterogeneous, rapidly renaturing DNA fraction) are used,by hybridization to define a fraction of DNA which is not digested by Hind I l l and is highly enriched for the sequences.

t

JSize fractionation allows purification of heterochromatic DNA.

Size fractionation allows purification of heterochromatic DNA.

Hae III digestion and addition of synthetic Hind IU linker molecules allows cloning into the Hind III site of pBR322.

Generation of EcoRI cohesive ends (by digestion) allows cloning into the EcoRI site of pBR325.

DNA

FIG. 8. Summary of approaches used to clone heterochromatic sequences from rye. The irnportant difference between the two procedures is the initial identification of a DNA fraction containing the sequences of interst. This requires restriction enzyme digests of total DNA to be screened, to decide which restriction enzyme provides the most favorable purification of the required sequence(s). The Appels et al. (1981) procedure is useful in this respect since an initial, broad, fractionation of DNA is made on the basis of kinetic complexity of various sequences in the genome (i.e., the rate at which they renature). These classes of DNA are then used as probes to screen restriction enzyme digests of total DNA to define the restriction enzyme required for further purification of the respective sequences in native form (and thus ready for cloning). A further extension of this approach is the isolation of equivalent classes of sequences (on the basis of renaturation rate) from different species and using these as probes to screen a given set of clones-replica screening in this way can allow the identification of species specific sequences. The “Hind111 synthetic linkers” used by Bedbrook e r a / . (1980a) are a class of compounds now commercially available; the so-called linkers are 10-12 bp long and carry restriction enzyme cleavage sites which can be ligated onto segments of DNA.

106

R. APPELS

FIG.9. In situ location of the major heterochromatic sequence from rye. Pachytene chromosomes were prephotographed (left-hand panels) before the in siru hybridization procedure was canied out. The autoradiography of the material after in situ hybridization is shown in the right-hand panels, and demonstrate that the sequence analyzed is distributed throughout the entire block of heterochromatin. Taken, in part, from Appels er al. (1981).

107

WHEAT-RYE HYBRIDS

40

a0

60

TEMPERATURE

("C)

10. Melting curves of a clone of the major heterochromatin sequence from rye. The sequence was labeled with 3*P and hybridized to either rye DNA (-) or the cloned sequence itself (---), immobilized on nitrocellulose filters. Melting point determinations were then carried out in 3 X S S C , 50% formamide. The two curves are virtually indistinguishable, indicating that in rye DNA the sequence is a uniform distribution of the sequence which was cloned. Identical results are obtained when the DNA from individual rye chromosomes (in wheat-rye addition or substitution lines) are compared to the cloned sequences (Appels et al., 1981). FIG.

nature of the arrays on different chromosomes has not, to date, been examined at a molecular level. The study of cereal heterochromatin, and in particular the rye heterochromatin in wheat-rye substitution or addition lines, has thus shown that there must exist mechanisms which are capable of generating an identical array of sequences on nonhomologous chromosomes. Observations of this type have also been made in animal species (Peacock et al., 1978). It is however unlikely that within the cereals, or other species, particular tandem arrays of repeated sequences on nonhomologous chromosomes are necessarily identical. The arrays of 5 S rRNA genes and 18 S-26 S rRNA genes on nonhomologous chromosomes are good examples. In wheat, the spacer regions associated with the 5 S rRNA genes on chromosome 1B are quite different from those located in respective tandem arrays elsewhere in the genome even though the gene regions are identical (Gerlach and Dyer, 1980). Similarly the spacer regions of the 18 S-26 S regions on chromosomes lB, 6B, and 5D are readily distinguishable with respect to DNA sequence (Appels and Dvorak, 1982). In animals species similar observations have been made (for a review see Federoff, 1979). Apparently certain segments of a tandem array of sequences can accumulate changes more rapidly than other segments. The mechanism responsible for this differential change in various parts of a tandem array of sequences is presumably superimposed on a mechanism responsible for generating tandem arrays of sequences on nonhomologous chromosomes. Unfortunately no information, at a molecular level, describing these mechanisms is available.

109

WHEAT-RYE HYBRIDS

The long range arrangement of sequences in heterochromatin has not been studied in detail. It is known that in rye and wheat (as in animal species) individual repeated sequences can exist in tandem arrays of at least 30 kb within heterochromatic regions (Appels et a l . , 1978, 1981; Bedbrook et al. 1980a; Gerlach and Peacock, 1980; Jones and Flavell, 1982) but whether the various sequences in heterochromatin are present in extremely long, individual, tandem arrays, or interspersed in moderately long arrays among other sequences, is unknown. Bedbrook et al. (l980b) have shown heterochromatic sequences can occur linked to other (repeated) sequences. The latter possibility would be consistent with the observation that in vivo RNA synthesis occurs from heterochromatic regions in rye (R. Appels, unpublished observations) as well as from the heterochromatic regions of animal species (Varley et a l . , 1980; Diaz et a l . , I98 1). This fact taken together with the demonstration that well-defined genetic loci exist in Drosophila melanogaster heterochromatin (reviewed by Hilliker et a l . , 1980) indicates that these regions of the genome cannot be simply assigned as containing classes of DNA of no biological function. The data available suggest that our understanding of the heterochromatic regions in rye (as well as other species) is incomplete. Recent developments in isolating gene sequences from plants, especially cereals (Cocking et a l . , 1981; Burr and Burr, 1981; Brandt et a l . , 1982; Gerlach et a l . , 1982), indicate that the structure of several, well-defined regions of the genome will soon be understood. An important aspect of these studies relates to the analysis of flanking (and very likely repeated) sequences and whether any of the current observations on tandem repeat sequence arrays apply to these sequences.

IV. Translocations in Wheat-Rye Addition or Substitution Lines Schemes to produce wheat-rye addition and/or substitution lines can involve the rye chromosome existing as a univalent for several generations (see for example Fig. 12). Univalency often leads to misdivision of the chromosome at first and second anaphase of meiosis (Fig. 13). Telosomic fragments are more frequently produced at the second division of meiosis (Sears, 1952), and can ~~

FIG. 11. The analysis of heterochromatin regions from individual rye chromosomes. Wheat-rye addition (or substitution) line DNA was digested with a restriction enzyme (indicated) and the fragments separated by electrophoresis in agarose gels. Transfer of the fragments to nitrocellulose allowed the hybridization of rye-specific heterochromatic DNA probes. The numbers on the left-hand side of the autoradiograms are length markers (in kb). (a) The distribution of the “480”-bp sequence (from Bedbrook era/. , 1980a). Photograph kindly supplied by R. Flavell. (b) The distribution of the sequence in the plasmid pSc7235 (from Appels et al.. 1981). The “480”-bp sequence and the sequence in the plasmid pSc7235 have been shown to belong to the same family of sequences (Appels eta/.. 1981).

R. APPELS

110 Chinese Spring Aneuploid 2 0 ' : 2 0 " tt': 20"t 1 '

CS/Irnperial addition 21"t l " R

x Select

Selfing

20"t 1 ' W t 1 ' R

-

X

21°C l " R

Select 2 0 ' t 1' R Monosornic substitution

20"t 1 ' W t l " R

1

Select

Self ing

20'+ l " R Disornic substitution

Selfing

2

FIG. 12. Scheme for the development of wheat-rye substitution lines. Redrawn from Koller and Zeller (1976).

appear as micronuclei (see for example Crosby, 1957). The important consequence of misdivision is that different telosomic fragments can reunite to produce new combinations of chromosome arms; this leads to the formation of isochromosomes and translocations. Among wheat chromosomes Sears (1952) found several isochromosomes in a study analyzing the behavior of univalents in Triticum aestivum C.V.Chinese Spring. Sears (1973) furthermore specifically tested the possibility of obtaining spontaneous translocations through the union of telocentric chromosomes formed from chromosomes present as univalents. In his experiment Sears recovered a 6BL-5RL translocation from wheat carrying chromosomes 6B and 5R as univalants. A similar translocation was recovered when one of the 5R arms was telocentric. When rye chromosomes are transferred into wheat a number of spontaneous wheat-rye translocations often occur apparently as a result of telosome formation at some stage of the transfer (Table I). However, as indicated in Table I, not all translocations can be accounted for by the telosome mechanism. Although telosome production often appears to involve breakage at the centromere it may be merely too close to the centromere. In the latter case duplications and deficiencies could be generated, as a result, for genes located near the centromere. The spontaneous generation of a compound chromosome (= isochromosome) in Drosophila melanogaster has been argued to occur by a translocation mechanism involving breaks near to, but not at, the centromere (Gibson, 1977; D. G. Holm personal communication). Although the breaks occur in heterochromatin they nevertheless can affect genes since a detailed analysis of chromosome-2 heterochromatin has uncovered vital genes (Hilliker and Holm, 1975; Hilliker, 1976). The density of genes in this region of the chromosome, per unit of DNA length, is much less than in euchromatin. In terms of the genetic analysis of rye chromosomes, the occurrence of

WHEAT-RYE HYBRIDS

111

rye-rye translocations can be particularly confusing. Zeller ( 1977) demonstrated that Dakold rye chromosome I addition to Kharkov wheat (Evans and Jenkins, 1960) was composed of a lW3R translocation which could not be found in Dakold rye per se. In the King I1 rye additions to Holdfast wheat (Riley and Chapman, 1958) the VI addition has a rye-rye interchange involving chromo-

FIG. 13. Aberrant meiotic behavior of a rye univalent. Taken from Riley and Chapman (1958). The separation of the univalent at anaphase I tends to favor misdivision at the following meiotic division leading to the production of telosomic chromosomes. Alternatively, formation of telosomes can result from misdivision at the first meiotic division. Lima-deFaria (1949) related the telosome products misdivision to the symmetrical structure of the centromere region (as observed cytologically at pachytene. Fig. I ) and suggested that breakage occurred between the two small chromomeres present in this region. Photograph kindly supplied by R. Riley.

112

R. APPELS TABLE 1 WHEAT-RYEA N D RYERYE TRANSLOCATIONS RECOVERED I N T H E PRODUCTION MANIPULATION OF WHEAT-RYEHYBRIDS"

AND

Translocation

Reference

1RS.IBL IRS.IBL 1RS.lBL IBS.1RL 2AL.2RL 2RS. 2BLh 2BS.2RL 4A.2Rc 7BL.4RL 4ASRLd 5BS.5RLc 5DL.5RL 6BLSRL 6D.5RLc 6BS.6RL 4Aa.7RS lR.3R 4RL.4RL 5RL.5RL ?R.7R

Zeller (1973) Mettin et a / . (1973) Shepherd (1973) Lawrence and Shepherd (1980) Sears (1 972) May and Appels (1980) May and Appels (1980) Driscoll and Jensen (1964) Zeller and Koller (1981) Driscoll and Sears (1965) Sears (1967) Muramatsu (1968) Sears (1973) Sears (1 967) Tuleen, quoted in Zeller and Koller (1981) Zeller and Koller (1981) Zeller (1977) Koller and Zeller (1976) Sears (1967) Koller and Zeller (1976); T.E. Miller (personal communication)

OTranslocations between wheat and rye chromosomes have potential, practical, value in that a useful segment of rye chromatin may be introduced into wheat while minimizing the introduction of undesirable rye characters. Many wheat-rye and rye-rye translocations are being found as triticales are crossed to wheat in a number of breeding programs (R. Schlegel, personal communication; J . Pilch, personal communication; J. P. Gustafson, personal communication). These translocations have not yet been characterized to assign the origin of the respective wheat and rye arms involved. bHomozygous lethal as a substitution for 2B. CRadiation induced and not a Robertsonian type translocation. dThis translocation has been shown nof to be of a Robertsonian type translocation by Driscoll and Sears (1965).

some 7R (Koller and Zeller, 1976; T. E. Miller, personal communication) and certainly the rye chromosomes in additions Ill and VI cannot be found in King I1 rye per se (Singh and Robbelen, 1976) as judged from the C-banding pattern. Jones and Flavell (1982) recently investigated the King I1 rye additions to Holdfast wheat using the cloned heterochromatic sequences from rye. The studies confirmed the observations from C-banding. In Secale montanum additions to Kharkov wheat (Gustafson et a / ., 1976) one isolation of chromosome G was not recognizable in the original S. montunum, or the other addition lines analyzed. Not all such changes have been studied in detail but clearly most of them occurred subsequent to the introduction of the rye chromosomes into wheat. It is,

WHEAT-RYE HYBRIDS

113

furthermore, not unreasonable to suspect that the instability of univalents with the associated production of telosomes, is a major factor contributing to such changes. The production of wheat-Aegilops sharonensis (Miller et al., 1982), wheat-barley (Islam et al., 1981; C. C. Jan and J . Dvorak, personal communication), and wheat-Elytrigia elongata (Dvorak and Knott, 1974; Dvorak, 1981 ) addition lines has also generated chromosomes which are not recognizable in the respective A. sharonensis, barley or Elytrigia elongata parents. Similar changes occur in the wheat-Haynaldia villosa addition lines’ as judged from investigations using C-banding (A. Lukaszewski, E. R. Sears, and J. P. Gustafson, unpublished data). It is thus a general feature that introducing an alien chromosome into wheat can result in changes in the chromosome structure of the alien chromosome.

V. Polymorphisms in Regions of the Chromosomes Containing Repeated Sequence DNA The level of spontaneous nucleotide changes in DNA, based on allozyme analyses, is generally considered to be less than 10-5-10-6 per generation. Polymorphisms generated in this way would not be expected to have a detectable effect on the cytologically observed chromosome structure. Polymorphisms are however observable at the cytological level and these, to date, have involved regions containing repeated sequence DNA. In the ribosomal DNA region of rye chromosome IR, polymorphism for the amount of rDNA using nucleic acid hybridization techniques (Flavell and Smith, 1974; Miller et al . , 1980), and 5 S DNA (R. Appels, unpublished) have been observed. The spacer region shows extensive polymorphism within the Triticeae and has been analyzed in detail by Appels and Dvorak (1982). A result of critical significance is the finding that polymorphism for a particular length variant of the spacer region within a population of, for example, Triticum dicoccoides is found throughout most of the 2000-5000 units of rDNA present in the respective individual. There thus appears to be a corrective mechanism capable of “fixing” a given length of spacer repeated sequences within many rDNA units; since these stretches of genomic DNA amount to 18,000-45,000 kb the mechanism responsible for the correction and fixation of new variants is a significant factor affecting chromosome structure. Whether the rDNA locus is unique in this regard remains to be determined. Variations in the amount of heterochromatic DNA sequences as well as the types of sequences present have been found in analyzing the heterochromatic ‘The wheat-Haynaldiu villosa additions were originally isolated by Hyde (1953) and reisolated more recently by E. R. Sears.

114

R. APPELS

regions of chromosomes. The early analyses of Bose (1956, 1957, 1958) demonstrated the existence of polymorphism in the heterochromatic regions of the satellited chromosome (IR) of rye. Studies using C-banding to visualize heterochromatin have shown extensive quantitative variation in the amount of heterochromatin present on rye chromosomes. Species in the genus Secale show extensive variation relative to each other (Bennett et a l . , 1977) but almost similar levels of variation exists within Secale cereale (Darvey and Gustafson, 1975; Weimark, 1975; Tikhonovich and Fedeyeva, 1976; Lelley et a l . , 1978; Giraldez et al., 1979; Naranjo and Lacadena, 1980). The examples shown in Fig. 14a and b show C-band polymorphism in a Japanese variety of rye and the variety “Snoopy,” respectively. Most variation is of a quantitative nature as emphasized by some examples of homologous chromosomes alligned in Fig. 14c. Occasionally an apparently new C-band is present as observed for chromosome 2R where a prominent band is either present or absent in the middle of the long arm (J. P. Gustafson, personal communication; Fig 14c). Rye is not exceptional in the levels of polymorphisms observed in heterochromatin. The polymorphisms of C-band heterochromatin in barley have been studied in detail by Linde-Laursen (1978). Wheat also shows C-band polymorphism although this has not been extensively analyzed. Comparison of the Cbanding of B genome chromosomes in the varieties Hope, Timstein, Cheyenne, and Chinese Spring (J. Dvorak, personal communication) indicates that polymorphisms similar to those found in rye occur in these varieties. Animals show analogous polymorphisms (for a reviews see White, 1973; John, 1981). At a molecular level observations have been made concerning polymorphism of heterochromatin structure. The major (cloned) heterochromatic sequence of rye shows a marked difference in the level of hybridization to the telomeric regions of chromosome 3R, when rye varieties Imperial and Petkus are compared (Appels et a l . , 1981). Comparison of the varieties King 11, UC90, and Petkus with the different families of heterochromatic DNA sequences cloned by Bedbrook er al. (1980a) also shows marked quantitative differences between these varieties (Jones and Flavell, 1982). In wheat-rye hybrids several polymorphic rye chromosomes have been characterized using molecular probes. May and FIG. 14. Heterozygosity of C-banded chromosomes in rye varieties. (a) Chromosomes from a Japanese variety (JNK) of rye. The line (a winter type) has been kept in Spain with no special precautions to prevent crossing to other varieties. The photograph thus illustrates the heterozygosity possible in rye. Unpublished photograph kindly supplied by R . Giraldez. (b) Chromosome from the rye variety “Snoopy” (a spring type). Unpublished photograph kindly supplied by J. P. Gustafson. The right-hand photograph shows a translocation (4R-5R) which is present in some individuals (Lelley and Gustafson, 1979). This type of structural heterozygosity was noted earlier in some rye varieties (Miintzing and Prakken, 1941; Akdik and Miintzing, 1949). (c) Cut-out chromosomes from (a) and (b) to emphasize the C-band heterozygosity observed. The top row is from (a), the second from (b, left panel), and the third is from (b, right panel).

116

R. APPELS

Rye repeated sequence DNA probe

Wheat repeated sequence DNA probe

WHEAT-RYE HYBRIDS

117

Appels (1980) observed a polymorphism for chromosome 2R, using the in situ hybridization of a radioactive probe for rye heterochromatin, among progeny from a program to introduce 2R into comrnerical wheat varieties (Fig. 15a). Further characterization of this polymorphism showed that it was the result of a wheat-rye translocation; a radioactive probe for detecting wheat heterochromatic sequences demonstrated that the entire 2RS arm had been replaced by 2BS, as summarized in Fig. 15b. A polymorphic form of 2R (lacking terminal C-bands) was thought to be present in the triticale Rosner (Kaltsikes et al., 1977); Rosner hybrids with hexaploid wheat and diploid rye apparently form 15 and 7 bivalents, respectively, suggesting that a chromosome in Rosner is capable of pairing with both wheat and rye chromosomes. Detailed pairing data are not available, however, and the experiments discussed below indicate that Rosner has the wheat chromosome 2D substituting for the rye 2R chromosome. The hexaploid triticale Rosner is an early Canadian variety with a complex pedigree (Larter, 1976). The analysis of the Rosner “2R” by Appels et al. (1982) utilized both in situ hybridization and genetic homology to detect the presence or absence of 2R sequences. The radioactive [3H]cRNA probe used for in situ hybridization was prepared from the heterogeneous, rapidly renaturing, DNA fraction from the rye genome (see Fig. 4a) and was useful because it hybridized not only to the prominent, terminal heterochromatic regions but also to interstitial regions. No modified 2R was detected in Rosner by this assay; Fig. 16 shows the karyotypic analysis of the F l progeny from a Rosner X Siskiyou cross where the triticale Siskiyou is known to contain an easily identifiable 2R. The karyotype should thus have shown the presence of a chromosome hybridizing the radioactive probe but without an obvious karyotype partner if the 2R was present in a modified form. This was not found. The analysis of pollen mother cells confirmed that the Siskiyou 2R did not pair with any chromosome in Rosner. The presence of chromosome 2D in Rosner was determined genetically by demonstrating that the telosome 2DS always paired with a chromosome in Rosner. The examination of a number of triticales using the cloned, major, heterochromatic rye sequence shows extensive polymorphism for the amount of the sequence present in the rye chromosomes (Appels et al., 1982). A particularly interesting example of the variation is shown in Fig. 17. Comparing the karyo-

FIG.15. Identification of a wheat-rye translocation. (a) Complete mitotic chromosome spread showing an altered 2R chromosome; in situ hybridization with a heterogeneous probe for rye heterochromatic sequences was used. Only two rye chromosomes are present (taken from May and Appels, 1980), one of which (the lower one) has a modified appearance. (b) Karyotype of the rye chromosomes in (a) hybridized with either the rye repeated DNA sequence probe (the heterogeneous probe for rye heterochromatic sequences) or wheat repeated DNA sequence probe (a sequence detecting homology t o m ) . Derived from May and Appels (1978, 1980).This analysis demonstrates that a wheat-rye translocation is present and identifies the wheat arm as 2BS.

FIG. 16. Analysis of an F, progeny from a cross between two triticales, Rosner (lacking 2R) and Siskiyou (has all rye chromosomes). The probe used is that described in Fig. 4a. The metaphase spread in the top panel was karyotyped as shown so that pairs of chromosomes with maximum similarity were placed together. Taken from Appels er al. (1982).

WHEAT-RYE HYBRIDS

119

Fic. 17. Comparison of two triticales TT5 and T1006. The mitotic chromosomes were hybridized with either a heterogeneous, heterochromatic sequence probe (see Fig. 4a) on the top set of chromosomes or a probe synthesized from a clone of a major heterochromatic sequence, on the bottom set of chromosomes. The arrows indicate the chromosomes compared in the text. Taken from Appels et al. (1982).

types of two triticales TT5 and T1006 showed that the chromosome marked by an arrow in Fig. 17 was strikingly different in the quantity of the sequence present in the telomeric region of the short arm. In contrast the content of heterochromatic sequences, assayed with the heterogeneous probe synthesized from rapidly renaturing DNA, on the respective chromosomes did not differ to any marked degree. The polymorphism here is thus a case of differences in composition of sequences in the heterochromatic region. Since several different sequences are known to be present in heterochromatin (see Section 111) a heterogeneous probe would show positive hybridization even if two out of four sequences were missing from the region examined. The cloned probe is specific enough so that if its homologous sequence is absent, no hybridization is observed. Using the C-banding analysis of triticales and wheat-rye addition lines a great deal of polymorphism has been observed for the amount of heterochromatin present on rye chromosomes. Singh and Robbelen (1976) observed a polymorphism in the King I1 additions to Holdfast wheat for chromosome 6R, Merker (1975) reported a polymorphism for C-banding of chromosome 7R in the triticale DRIRA and Roupakais and Kaltsikes (1977) detected a polymorphism for 6R in the triticale Rosner. Within a group of 50 triticales Pilch (198 1) recorded polymorphic forms of chromosomes lR, 4R, 5R, 6R, and 7R; similar variation was observed in a group of 11 triticales analyzed by Sapra and Stewart (1980) as well

120

R. APPELS

as in the triticales examined by Darvey and Gustafson (1975). Extensive analysis of triticales by J . P. Gustafson uncovered the 2R variant in the triticale Siskiyou (discussed earlier), as well as variants of the other rye chromosomes and wheat-rye translocations (Gustafson, 1982; J. P. Gustafson personal communication). Many of the polymorphic forms of the rye chromosomes in triticales can be identified in varieties of rye such as “Snoopy” which are utilized in triticale breeding programs. Examples of such chromosomes are shown in Fig. 14c where variant forms of lR, 2R, 4R, 5R, 6R, and 7R closely resemble the respective variants observed in triticales. Thus many of the polymorphic forms of rye chromosomes present in triticales probably existed in the varieties of rye used to form the triticales. Chromosomal polymorphism in triticales (spring type) with respect to the substitution of rye chromosomes by wheat chromosomes has already been mentioned in the earlier discussion related to the triticale Rosner where rye 2R was substituted by wheat 2D. A similar event occurred in the CIMMYT triticale selection Armadillo (Gustafson and Zillinsky, 1973), which forms the basis of many commercial varieties of triticale. Additional rye chromosomes can also be replaced (Gustafson, 1982). These substitutions occur mainly when primary

FIG. 18. Wheat-rye translocation in progeny from a triticale X wheat cross. The C-banding technique was used to identify a 3RL/5BL translocation in the progeny of a cross between a triticale and wheat. Unpublished protograph kindly provided by A. Lukaszewski and I. P. Gustafson.

WHEAT-RYE HYBRIDS

121

hexaploid triticales are crossed to octaploid triticales or to wheat (for example, see Merker, 1976; Larter, 1976; Gustafson and Zillinsky, 1978). Certain (large) rye chromosomes such as 2R appear to be preferentially substituted (Gustafson and Zillinsky, 1978). In European triticales (winter types) the rye chromosome content is extremely constant (Lukaszewski and Apolinarska, 198 1; Pilch, 1981; A. Merker, personal communication) and all selections have seven pairs of rye chromosomes. Considering the fact that rye chromosomes often appear as univalents in triticales (Merker, 1975; Thomas and Kaltsikes, 1976; Roupakais et al., 1978) it is expected, from the behavior of univalents discussed in Section IV, that rye-wheat or rye-rye translocations should provide a source of apparent chromosomal polymorphism. It is surprising therefore that no such translocations have appeared in the published analyses of triticale varieties. Some translocations such as the 2RS-2BL translocation (Table I) which are homozygous lethal would not be expected to be recovered. It is possible that the unbalanced chromosomal constitution which results initially from translocations causes their loss at early stages of a triticale breeding program (Merker, 1975). Crossing triticales to wheat varieties has been reported to result in wheat-rye translocations (May and Appels, 1982; Lukaszewski and Gustafson, 1982; Fig. 18).

VI. The Biological Effects of Rye Chromosomes (or Rye Chromosome Fragments) in Wheat-Rye Hybrids: Specific Effects Related to Heterochromatin A major consequence of maintaining rye chromosomes in wheat-rye hybrids is that the rye chromosomes are transferred from an outbreeding species to a predominantly inbreeding species. The deleterious effects of inbreeding cultivated rye have been well documented and include aberrant exchanges (U-type exchanges), cytological meiotic instability (neocentromeric activity, reduced pairing), and poor fertility ( L a m , 1936/1937; Prakken and Muntzing, 1942; Rees, 1955; Jones, 1968; Giraldez and Lacadena, 1978). Lelley (1978) has pointed out that whether homozygosity of rye chromosomes in wheat-rye hybrids such as triticales contributes directly to their defects (reviewed by Scoles and Kaltsikes, 1974; Tsuchiya, 1974) has not been determined. Nevertheless considerable attention has been given to the possibility that the heterochromatic regions of rye chromosomes are directly related to the meiotic instability, reduced fertility, and seed shrivelling observed in triticales (Shkutina and Khvostova, 1971; Bennett, 1974, 1977; Merker, 1975; Kaltsikes and Roupakais 1975; Kaltsikes et al., 1975; Thomas and Katsikes, 1974; Gustafson and Bennett, 1982; Bennett and Gustafson, 1982). The heterochromatic regions in rye per se are late to finish replicating (Darlington and Hague, 1966; Lima de Faria and Jaworska, 1972; Ayonoadu and

122

R. APPELS

Rees, 1973) and although this has not yet been demonstrated to be the case in wheat-rye hybrids, the heterochromatic regions of rye chromosomes in triticale are involved in anaphase bridge formation in endosperm nuclei mitotic divisions (suggestive of late replication, Bennett, 1977). Persistence of this property of late replication in wheat-rye hybrids could generate a variety of undesirable side effects (Bennett, 1977). The fact that heterochromatic regions have biological effects within an organism has long been known for Drosophila melanogaster (for a recent review see Hilliker et al., 1980). A recent study systematically deleting X chromosome heterochromatin (Hilliker and Appels, 1982; Fig. 19) has elaborated the contribution this region makes to modulating levels of gene activity. The deletions shown in Fig. 19 were demonstrated to enhance the position effect variegation on the white locus and affect the dominance relationships, with respect to endoreduplication and activity, between the rDNA regions of the X and Y chromosomes (Hilliker and Appels, 1982). In rye the presence of B chromosomes (which carry a substantial amount of heterochromatin, see Fig. 6) has a relatively weak, but measurable, effect on phenotype (Miintzing, 1954; Ayonoadu and Rees, 1968). On the basis of the limited data available it is thus not unreasonable to consider that the large blocks of rye heterochromatin in wheat-rye hybrids may affect gene expression (Bennett, 1977; Gustafson and Bennett, 1982; Bennett and Gustafson, 1982). However, the effects may not necessarily be deleterious. The studies on B chromosomes have indicated they can have a selective value (Rees and Ayonoadu, 1973) and certainly in winter triticales there appears to have been a strong selection for a full complement of rye chromosomes with prominent heterochromatic blocks (see preceding section). The modulation of gene activity by heterochromatin appears to be complex and may work favorably in certain genotype/environment combinations while in other situations the effects may be detrimental. Unfavorable genetic interactions between rye chromosomes (or chromosome

types of su(f)deletions selected FIG. 19. Deletions of X chromosome heterochromatin. To examine the biological properties of heterochromatin Hilliker and Appels (1982) isolated a set of deletions, as indicated, from the X chromosome In(l)wmStb.The inversion in In(l)wm51bhas one break in the euchromatin near the white locus (w) and the other at the proximal boundary of the rDNA cluster (indicated as El).The heterochromatin which contains repeated sequence DNA (or DNA which has not yet been identified) not in the rDNA category is shown as B. Deletions for the su(f) locus were induced by y-irradiation and were characterized as extending to varying degrees into the heterochromatin, as indicated.

WHEAT-RYE HYBRIDS

123

fragments) and wheat chromosomes have been investigated as factors contributing to the properties of wheat-rye hybrids. The variable examined extensively is chromosome pairing and it is clear that in both tetraploid triticales (Roupakais et al., 1978) and in hexaploid triticales (A. Kiss, quoted in Scoles and Kaltsikes, 1974; Naranjo et al., 1979; Lelley and Larter, 1982), certain combinations of rye and wheat chromosomes are favored over others. This is in addition to the effects of the Ph system of wheat for suppressing homoeologous pairing (for review see Sears, 1976) which may be particularly sensitive to the heterozygosity of the rye chromosomes and result in their pairing failure (Riley and Miller, 1970; Jouve et al., 1980; Lelley, 1982). Different rye chromosome variants (for example 4R, 6R, and 7R) also have clear effects on endosperm development and kernel characteristics (Darvey, 1973; Gustafson and Bennett, 1982; Bennett and Gustafson, 1982) and if these rye variants existed prior to incorporation into triticale (as suggested in Section V) the effects may be the result of general genetic differences as well as differences in heterochromatin. In conclusion it can be said that on the basis of studies in Drosophila in particular it is reasonable to expect rye heterochromatin in wheat-rye hybrids to modulate gene activity in some manner. This has not, however, been clearly demonstrated. Genetic interactions not involving heterochromatin between wheat and rye chromosomes would appear to be demonstrably significant in determining the properties of wheat-rye hybrids.

VII. The Possible Origins of Polymorphism in Rye Heterochromatin In Section 111 it was suggested that, on the basis of the molecular analyses of rye heterochromatin, there must exist mechanisms for “generating an identical array of sequences on nonhomologous chromosomes,” within the genus Secale. It can be further argued that the results of such events are of relatively recent origin on an evolutionary time scale. Many of the available parameters such as those from morphological studies (Bell, 1965), nuclear DNA (Bendich and McCarthy, 1970; Flavell et al., 1977), and chloroplast/mitochondrial DNA studies (Vedel et al., 1980) indicate that rye and wheat are closely related. Within specific regions of chromosomes such as heterochromatin however rye and wheat are virtually unrelated. The tandem arrays of the major heterochromatic sequence (see Section 111) of S. cereale are present in S. montunum, but are difficult to detect in Triticum aestivum (Bedbrook et al., 1980a; Appels et a l . , 1981) as well as Secale silvestre and S . africanum (Bedbrook et al., 1980a; Jones and Flavell, 1982), and Aegilops species (Hutchinson et al., 1980). This suggests that the tandem arrays as found in rye may be of a relatively recent origin. The polymorphisms observed for the quantity and composition of heterochromatin in rye (see Section V) are consistent with this suggestion. Unequal sister chromatid exchange has long been considered a source of

124

R. APPELS

quantitative change in an array of repeated DNA sequences (Smith, 1973, 1976; Tartof, 1973). The technique of BUdR labeling clearly shows sister chromatid exchanges can be visualized in rye (Friebe, 1978) and that the heterochromatic regions are not excluded from this activity (Fig. 20). Although the high number of sister chromatid exchanges observed using BUdR in rye appears to be induced by the method of assay (Friebe et al., 1982), observations such as those in Fig. 20 suggest the presence of an enzymatic capability for catalyzing sister chromatid exchange in mitotic cells. The presence of recombinational enzyme activity in the plant mitotic cells has also been implicated from the recovery of recombinant cauliflower mosiac virus DNA molecules after infection of leaf cells (Hohn et al., 1981). The origin of twin spots of complementary dark-green and yellow phenotypes on the leaves of light green heterozygotes has been argued to be the result of somatic exchange events (and is inducable by certain mutagens) (Vig and Zimmerman, 1977). In principle therefore, it seems possible perhaps under a particular situation of stress, that high levels of sister chromatid exchange could occur, resulting in large fluctuations in the numbers of repetitive sequences in heterochromatin due to unequal exchanges. Special situations leading to dramatic changes in the genome have been discussed (McClintock, 1978) and it is conceivable that mechanisms involving the currently much discussed transposons (Cameron et al., 1979; Roeder et al.,

FIG.20. Sister chromatid exchange in rye. The BUdR labeling technique was used to detect sister chromatid exchanges. The arrows indicate points of sister chromatid exchange. Taken from Friebe (1978). Photograph kindly supplied by B. Friebe.

WHEAT-RYE HYBRIDS

125

FIG. 21. Location of meiotic crossover. Meiotic metaphase chromosomes were stained using the C-banding technique. Taken from Jones (1978). Photograph kindly supplied by G. Jones.

1980) and/or gene coversion (for a recent discussion see Egel, 1981) contribute to fluctuations in the number and type of sequences present in rye heterochromatin. Another source of change is occasional aberrations in mitotic divisions which result in anaphase bridges and the release of fragments of chromatin when the bridges break (Belling, 1925; Muntzing and Prakken, 1941; Gustafson et al., 1982). These are relatively rare events, however, and would presumably be required to have a selective advantage in order for the chromosome thus modified to be fixed in a given line of rye. Meiotic exchange involving C-banded heterochromatin in rye is not usually considered a possibility. Evidence for chiasmata within heterochromatin regions has not been obtained in studies analyzing crossing-over frequencies in rye cultivars heterozygous for certain blocks of C-band heterochromatin (Giraldez and Orellana, 1979; Orellana and Giraldez, 1982; R. Giraldez, personal communication). Heterozygosity for the heterochromatic regions does not appear to adversely affect the pairing between homologs at meiosis (Bose, 1956; Orellana and Giraldez, 1982). Jones (1978) demonstrated however that meiotic crossingover can occur very close to C-banded heterochromatin (Fig. 21) and it may be possible that repeated sequences at the boundaries of the C-banded heterochromatin undergo meiotic exchange. Appels and Peacock (1978) argued that the boundary between euchromatin and heterochromatin is not sharp and that techniques such as C-banding detected the high concentrations of repeated sequences. Unequal meiotic exchange within repeated sequences near the euchromatin-heterochromatin junction may therefore be a source of gradual, quantitative, variation in the repeated sequences and thus cytologically visible heterochromatin. In yeast, unequal meiotic crossing-over between chromatids has been shown to be a mechanism operating to maintain homogeneity within an

126

R. APPELS

rDNA cluster by removing an introduced Leu 2 gene (coding for p-isopropylmalate dehydrogenase) (Petes, 1980). In D. melanogaster the frequency of unequal crossovers in the rDNA region was estimated to be a minimum of 3 x 10-4 exchanges per gamete per generation (Frankham et al., 1980). The mechanism has therefore the potential for significantly affecting the numbers of a given repeated sequence in a chromosome. Aberrant meiotic recombination events such as U-type exchange events have been reported in inbred lines of rye (Jones, 1968; Giraldez and Lacadena, 1978) and can occur with a very low frequency in telomeric heterochromatin (Jones, 1978; Alamo, 1981). Rare events of this type in normal rye could provide a potential source of heterochrorhatin variation. The B chromosomes present in many organisms have been considered a possible source of variation in the heterochromatin of normal chromosomes by translocation events incorporating segments from the B chromosomes (White, 1973). Some populations of rye contain B chromosomes (Figs. 1 and 6) with C-banded heterochromatin (Fig. 6); the B chromosomes do not, however, contain easily detectable levels of heterochromatic DNA sequences (Appels et al., 1978; Jones and Flavell, 1982) and thus are unlikely to provide a source of this type of sequence. The mechanism(s) involved in rapidly spreading a repeated sequence to nonhomologous chromosomes (discussed earlier) may also contribute to the polymorphisms observed in heterochromatin. In somatic cells of rye the heterochromatic regions of nonhomologous chromosomes are often close to each other at interphase as judged cytologically from examination of nuclei after C-banding (Singh and Robbelen, 1975; Godin and Stack, 1975) or in situ hybridization with heterochromatic sequences (Appels et al., 1978). The new chromosomal associations induced by X-irradiation or mitomycin C (Natarajan and Ahnstrom, 1972) are consistent with the observed juxtaposition of telomeric heterochromatin regions. The potential therefore exists for some type of exchange event to occur in somatic cells. Meiotic exchange events at zygotene when telomeres are clustered (Carroll and Brown, 1976) could also lead to both quantitative and qualitative changes in the arrays of repeated sequences in nonhomologous chromosomes. Accepting that extensive polymorphism exists in rye per se it is not surprising that polymorphism in rye heterochromatin in wheat-rye hybrids is found. Considering the recent origin of triticale it seems likely that the variation in rye heterochromatin within triticales predates the synthesis of the triticales. Examples of heterozygosity such as those described for the 7R chromosome of triticale DRIRA (Merker, 1975) and 6R of Rosner (Roupakais and Kaltsikes, 1977) have been interpreted as developing after the formation of the triticale (for discussion see Gustafson, 1976). The argument in favor of this interpretation is however confounded by the complex origin of the triticales and the fact that under certain

WHEAT-RYE HYBRIDS

127

nursery conditions 10-40% outbreeding can occur (Kiss, 1970; Hughes et al., 1976).

VIII. Prospects The demonstration and analysis of rapid changes in the heterochromatic regions (and/or euchromatic regions) of rye chromosomes are exciting possibilities which may arise out of the study of wheat-rye hybrids. If most of the polymorphism observed in the rye chromosomes of triticales preexisted in rye cultivars used to synthesize these hybrids such changes may not be very common. Within S. cereale the C-banding pattern of chromosomes in different cells of the same plant is constant (R. Giraldez, personal communication) consistent with the evidence that rapid changes in heterochromatin may not be common. In contrast S. kuprijanovii does appear to show cell to cell variation in its C-banding pattern of chromosomes (Gustafson et al., 1982). Rapid changes in the rye heterochromatin of wheat-rye hybrids thus remains a possibility. The presumed changes, however, must be distinguished from possible abnormalities induced by colchicine doubling (O’Mara, 1953), wheat-rye (and rye-rye) translocations, as well as wheat-rye chromosome substitutions.

ACKNOWLEDGMENTS During the preparation of this article it was a priviledge for the author to work with Drs. J. Dvorak and C. E. May and thus gain new insights into the analysis of the Triticeae. The author is grateful to the following colleagues for having provided comments on the manuscript: J. Dvorak, R. Giraldez, J. P. Gustafson, A. J . Hilliker, H. Lorz, T. Lelley, C. E. May, and F. Zeller. Several of the photographs featured in the article have not been published and the author appreciated the opportunity to utilize these in the article; the source of the photographs (published or unpublished) is given in the legends. The author wishes to thank the IAC (Netherlands) for providing a fellowship which facilitated his attendance at a workshop organized by Dr. J. Sybenga (held in Wageningen, March 1982) on rye genetics and cytology; the workshop allowed the author to elaborate on many aspects of the data discussed in the manuscript.

REFERENCES Akdik, S., and Muntzing, A. (1949). Heraditas 35, 67-76. Alamo, I. (1981). Ph.D. dissertation, Departamento de GenCtica, Facultad de Biologia, Universidad, Complutense, Madrid, 1981. Allan, M., and Harrison, P. (1980). Cell 19, 437-447. Appels, R.,and Dvorak, J. (1982). Theor. Appl. Genet., in press. Appels, R., and Peacock, W. J. (1978). Inr. Rev. Cytol. Suppl. 8, 69-126.

128

R. APPELS

Appels, R., and Ringertz, N. R. (1975). Curr. Top. Dev. Biol. 9, 137-166. Appels. R., Driscoll, C., and Peacock, W. J. (1978). Chromosoma 70, 67-89. Appels, R., Dennis, E. S., and Peacock, W. J. (1979). Proc. Ausf. Biochem. Soc. 12, 70 (Abstr.). Appels, R., Gerlach, W. L., Dennis, E. S., Swift, H., and Peacock, W. J. (1980). Chromosoma 78, 293-3 1 1. Appels, R., Dennis, E. S . , Smyth, D. R., and Peacock, W. J. (198 1 ). Chromosoma 84, 265-277. Appels, R., Gustafson, J. P., and May, C. E. (1982). Theor. Appl. Genet. (in press). Ayonoadu, U. W., and Rees, H. (1968). Exp. Cell Res. 52, 284-290. Ayonoadu, U. W., and Rees, H. (1973). Heredity 30, 233-240. Barber, H. H., Driscoll, C. J., Long, P. M., and Vickery, R. S. (1968). Nature (London) 218, 450-452. Bedbrook, J. R., Jones, J., O’Dell, M., Thompson, R. D., and Flavell, R. B. (1980a). Cell 19, 545-560. Bedbrook, J. R., O’Dell, M., and Flavell, R. B. (1980b). Nature (London) 288, 133-137. Bell, G. D. H. (1965). “Essays in Crop Evolution” (J. Hutchinson, ed.), pp. 70-102. Cambridge Univ. Press, London and New York. Bendich, A. J., and McCanhy, B. J. (1970). Genetics 65, 567-573. Bennett, M. D. (1972). Proc. R. Soc. London Ser. B 181, 109-135. Bennett, M. D. (1974). Proc. Int. Triticale Symp. IDRC, Ottawa pp. 137-148. Bennett, M. D. (1977). Heredity 39, 41 1-419. Bennett, M. D., and Gustafson, J. P. (1982). Can. J. Genet. Cyrol. 24, 93-100. Bennett, M. D., Gustafson, J. P., and Smith, J. B. (1977). Chromosoma 61, 149-176. Blakeslee, A. F., and Avery, A. G. (1937). J . Hered. 28, 393-41 1. Bose, S. (1956). Hereditas 42, 263-292. Bose, S. (1957). Hereditas 43, 621-643. Bose, S. (1958). Hereditas 44, 257-279. Brandt, A., Ingverser, J., Cameron-Mills, V., Schmitt, J. M., Rasmussen, S. K., and Hopp, H. E. (1982). In “Cytogenetics of Crop Plants” (M. S. Swaninathan, P. K. Gupta, and U. Sinha, eds.), in press. Briggle, L. W. (1969). Crop Sci. 9, 197-202. Burr, B., and Burr, F. A. (1981). Generics 98, 143-156. Cameron, J. R., Loh, E. Y.,and Davis, R. W. (1979). Cell 16, 739-751. Carroll, D., and Brown, D. D. (1976). Cell 7, 477-486. Cocking, E. C., Davey, M. R., Pental, D., and Power, J. B. (1981). Nature (London) 293, 265-269. Crosby, A. R. (1957). Am. J. Bot. 44, 813-822. Darlington, C. D., and Hague, A. (1966). “Chromosomes Today” (C. D. Darlington and K. R. Lewis eds.), pp. 102-107. Oliver & Boyd, Edinburgh. Darvey, N. L. (1973). Proc. Inr. Wheat Gen. Symp., 4rh, pp. 155-159. Darvey, N. L., and Gustafson, J. P. (1975). Crop Sci. 15, 239-243. Dennis, E. S., Gerlach, W. L., and Peacock, W. J. (1980). Heredify 44, 349-366. Diaz, M. O., Barsacchi-Pilone, G., Mahon, L., and Gall, J. G. (1981). Cell 24, 649-659. Driscoll, C. J., and Jensen, N. F. (1964). Crop Sci. 4, 372-374. Driscoll, C. J., and Sears, E. R. (1965). Genetics 51, 439-443. Dvorak, J. (1980). Can. J. Genet. Cytol. 22, 237-259. Dvorak, J., and Knott, D. R. (1974). Can. J. Genet. Cyrol. 16, 399-417. Egel, R. (1981). Nature (London) 290, 191-192. Evans, 1. E., and Jenkins, B. C. (1960). Can. J. Genet. Cytol. 2, 205-215. Federoff, N. V. (1979). Cell 16, 697-710. Flavell, R. B., and Smith, D. B. (1974). Chromosoma 47, 327-334.

WHEAT-RYE HYBRIDS

129

Flavell, R. B., and Smith, D. B. (1976). Heredity 37, 231-252. Flavell, R. B., Rimpau, J . , and Smith, D. (1977). Chromosoma 63, 205-222. Frankham, R. D., Briscoe, A,, and Nurthen, R. K. (1980). Generics 95, 727-742. Friebe, B., (1978). Microsc. Acra 81, 159-165. Friebe, B., Linnert, G., and Vogel, R. (1982). Mutat. Res., in press. Gerlach, W. L., and Bedbrook, J. R. (1979). Nucleic Acids Res. 7, 1869-1885. Gerlach, W. L., and Dyer, T. A. (1980). Nucleic Acids Res. 8, 4851-4865. Gerlach, W. L., and Peacock, W. J. (1980). Heredity 44, 269-276. Gerlach, W. L., Miller, T. E., and Flavell, R. B. (1980). Theor. Appl. Genet. 58, 97-100. Gerlach, W. L., Pryor, A. J., Dennis, E. S., Ferl, R., Sachs, M. M., and Peacock, W. J. (1982). Proc. Natl. Acad. Sci. U.S.A. 79, 2981-2985. Gibson, W. G. (1977). Ph.D. dissertation, University of British Columbia, Canada. Gill, B. S., and Kimber, G. (1974). Proc. Natl. Acad. Sci. U . S . A . 71, 1247-1249. Giraldez, R., and Lacadena, J. R. (1978). Chromosoma 66, 193-204. Giraldez, R., and Orellana, J. (1979). Chromosoma 72, 377-385. Giraldez, R., Cermeno, M. C., and Orellana, J . (1979). Z. P’unzenzuecht. 83, 40-48. Godin, D. E., and Stack, S. M. (1975). Can. J . Genet. Cytol. 17, 269-273. Gupta, P. K. (1970). Phyton 14, 9-13. Gupta, P. K. (1971). Genetica 42, 199-213. Gustafson, J . P. (1976). Evol. Biol. 9, 107-135. Gustafson, J. P. (1982). In “Cytogenetics of Crop Plants” (M. S. Swaminathan, P. K. Gupta, and U. Sinha, eds.), in press. Gustafson, J . P., and Bennett, M. D. (1982). Can. J. Genet. Cytol. 24, 83-92. Gustafson, J. P., and Zillinsky, F. J. (1973). Proc. Wheat Genet. Symp. 4th pp. 225-232. Gustafson, J . P., and Zillinsky, F. J. (1978). Pruc. Wheat Genet. Symp. Sth pp. 1201-1207. Gustafson, J. P., Evans, L. E., and Josifek, K. (1976). Can. J. Genet. Cytol. 18, 339-343. Gustafson, J. P., Lukoszewski, A. J . , and Bennett, M. D. (1982). In preparation. Halaban, R., Nordlund, J . , Francke, V., Moellmann, G., and Eisenstadt, J. M. (1980). Somatic Cell Genet 6, 29-44. Hart, G. E. (1978). Cereal Res. Commun. 6, 123-133. Hart, G. E. (1979). Stadler Genet. Symp. 11, 9-29. Heemert, C., van, and Sybenga, J. (1972). Genetica 43, 387-393. Hemleben, V., and Grierson, D. (1978). Chromosoma 65, 353-358. Hilliker, A. J. (1976). Genetics 83, 765-782. Hilliker, A. J., and Appels, R. (1982). Chromosoma (in press). Hilliker, A. J., and Holm, D. G. (1975). Genetics 81, 705-721. Hilliker, A. J., Appels, R., and Schalet, A. (1980). Cell 21, 607-619. Hiraizumi, Y . (1981). Genetics 98, 105-114. Hohn, T.,Richard, K., and Lebeurier-G. (1981). Curr. Top. Microbiol. Immunol. 193-236. Hughes, J. L., Sapra, V. T., and Sharma, G. C. (1976). Cereal Res. Commun. 4, 355-361. Hutchinson, J., Chapman, V., and Miller, T. E. (1980). Heredity 45, 245-254. Hyde, B. B. (1953). Am. J . Bor. 40, 174-182. Irani, B. N., and Bhatia, C. R. (1972). Genetica 43, 195-200. Islam. A. K. M. R., Shepherd, K. W., and Sparrow, D. H. B. (1981). Heredity 46, 161-174. John, B. (1981). Chromosomes Today 7, 128-137. Jones, G. H. (1968). Mutar. Res. 5, 385-395. Jones, G. H. (1978). Chromosoma 66, 45-57. Jones, J . D. G., and Flavell, R. B. (1982). Submitted. Jouve, N., Diez, N., and Rodriquez, M. (1980). Theor. Appl. Genet. 57, 75-79. Kaltsikes, P. J., and Roupakais, D. G. (1975). Can. J . Bor. 53, 2068-2076.

130

R. APPELS

Kaltsikes, P. J., Roupakais, D. G., and Thomas, J. B. (1975). Can. J . Bot. 53, 2050-2067. Kaltsikes, P. J., Roupakias, D. G., and Thomas, J. B. (1977). Can. J. Genet. Cyrol. 19, 575 (Abstr.). Kiss, A. (1970). Wheat Inf. Serv. 31, 24-25. Knott, D. R., and Dvorik, J. (1976). Annu. Rev. Phytopathol. 211-235. Koller, 0 . L., and Zeller, F. J. (1976). Genet. Res. (Cambridge) 28, 177-188. Kunz, W., Peterson, G., Renkawitz-Pohl, R., Glatzer, K. H., and Schafer, M. (1981). Chromosoma 83, 145-158. Lamm, R. (1936/1937). Hereditas 22, 217-240. Larter, E. N. (1976). In “Evolution of Crop Plants” (N. W. Simmonds, ed.), pp. 117-120. Longman, London. LaRue, H., and Pallotta, D. (1976). Can. J . Biochem. 54, 765-771. Lawrence, G. J., and Shepherd, K. W. (1981). Theor. Appl. Genet. 59,25-31. Lelley, T. (1978). Int. Wheat Genet. Symp. 5th pp. 1213-1217. Lelley, T. (1982). Proc. Eucarpia, Wugeningen pp. 101-105. Lelley, T., and Gustafson, J. P. (1979). Z. Pflanzenzuecht. 82, 306-310. Lelley, T., and Larter, E. N. (1982). Can. J. Genet. Cyrol. 22, 1-6. Lelley, T., Josifek, K., and Kaltsikes, P. J. (1978). Can. J . Genet. Cyrol. 20, 307-312. Lima-de-Faria, A. (1949). Heredifas 35, 77-85. Lima-de-Faria, A. (1952). Chromosoma 5, 1-68. Lima-de-Faria, A , , and Jaworska, H. (1972). Heredifas 70, 39-58. Linde-Laursen, I. (1978). Heredifas 88, 55-64. Linder, S., Zuckerman, S. H., and Ringertz, N. R. (1981). Proc. Natl. Acad. Sci. U.S.A. 78, 6286-6289. Lukaszewski, A. J., and Apolinarska, B. (1981). Can. J. Genet. Cytol. 23, 281-285. Lukaszewski, A. J., and Gustafson, I . P. (1982). In preparation. McClintock, B. (1978). Stadler Symp. 10, 25-47. May. C. E. (1974). Ph.D. dissertation, School of Botany, University of NSW, Sydney, Australia. May. C. E., and Appels, R. (1978). Cereal Res. Commun. 6, 231-234. May, C. E., and Appels, R. (1980). Theor. Appl. Genet. 56, 17-23. May, C. E., and Appels, R. (1982). Can. J. Genet. Cytol. 24, 285-291. Merker, A. (1975). Hereditas 80, 41-52. Merker, A. (1976). Wheat In$ Sew. 41142, 44-48. Mettin, D., Bluthner, W. D., and Schlegel, G. (1973). Proc. Int. Wheat Genet. Symp., 4th pp. 179- 184. Miller, T. E., Gerlach, W. L., and Flavell, R. B. (1980). Heredity 45, 377-382. Miller, T. E., Hutchinson, J . , and Chapman, V. (1982). Theor. Appl. Genet. 61, 27-33. Mohan, J., and Flavell, R. B. (1974). Genetics 76, 33-44. Moyzis, R. K., Bonnet, J., Li, D. W., and Ts’o, P. 0. P. (1981). J. Mol. Biol. 153, 841-870. Muller, W., and Gautier, F. (1975). Eur. J. Biochem. 54, 385-394. Muntzing, A. (1954). Caryologia Suppl. 6, 282-301. Muntzing, A., and Prakken, R. (1941). Hereditas 27, 273-308. Muramatsu, M. (1968). Proc. Int. Congr. Genet., 12th 1, 180. Naranjo, T., and Lacadena, J . R. (1980). Chromosoma 81, 249-261. Naranjo, T., Lacadena, J. R., and Giraldez, R. (1979). Z. Pflanzenzuecht. 82, 289-305. Natarajan, A. T., and Ahnstrom, G. (1972). Symp. Schlob. ReinhartshuusenlRhein pp. 215-223. O’Mara, J. G. (1953). Bor. Rev. 19, 587-605. Orellana, J., and Giraldez, R. (1982). Chromosoma 84, 439-449. Peacock, W. J., Lohe, A., Gerlach, W. J., Dunsmuir, P., Dennis, E. S., and Appels, R. (1978). Cold Spring Harbor Symp. Quanr. Biol. 42, 1121-1 135.

WHEAT-RYE HYBRIDS

131

Peacock, W. J., Gerlach, W. L., and Dennis, E. S. (1981). “Wheat Science-Today and Tomorrow” (L. T. Evans and W. L. Peacock, eds.), pp. 4 1 4 0 . Cambridge Univ. Press, London and New York. Petes, T.D. (1980). Cell 19, 765-774. Pilch, J . (1981). Theor. Appl. Genet. 60, 145-149. Prakken, R., and Miintzing, A. (1942). Hereditas 28, 441-482. Price, S. (1955). Genetics 40, 651-667. Rae, P. M. M., Bamett, T., and Murtiff, V. L. (1981). Chromosoma 82, 637-655. Ranjekar, P. K., Lafontaine, J. G., and Pallotta, D. (1974). Chromosoma 48, 427-440. Ranjekar, P. K., Pallotta, D., and Lafontaine, J. G. (1976). Eiochim. Eiophys. Acra (Aust.) 425, 30-40. Rao, I. N., and Rao, M. V. P. (1980). Genet. Res. (Cambridge) 35, 309-312. Rees, H. (1955). Heredity 9, 93-116. Rees, H., and Ayonoadu, U. (1973). Theor. Appl. Genet. 43, 162-166. Riley, R. (1955). J. Agric. Sci. 46, 377-383. Riley, R., and Chapman, V. (1958). Heredity 12, 301-315. Riley, R., and Miller, T. (1970). Nature (London) 227, 82-83. Ringertz, N. R., and Savage, R. E. (1976). “Cell Hybrids.” Academic Press, New York. Roeder, S. G., Farabough, P. J., Chaleff, D. T., and Fink, G. R. (1980). Science 209, 1375-1380. Roupakais, D. G., and Kaltsikes, P. J. (1977). Can. J. Genet. Cyto/. 19,543-548. Roupakais, D. G., Kaltsikes, P. J., and Krowlow, K. D. (1978). Proc. Int. Wheat Gen. Symp., 5th pp. 1218-1227. Sanchez-Monge, R., Delibes, A., HernandCz-Lucas, C., Carbonero, P., and Garcia-Olmedo, F. (1979). Theor. Appl. Genet. 54, 61-63. Sapra, V., and Stewart, M. D. (1980). Euphytica 29, 497-509. Sarma, N. P., and Natarajan, A. T. (1973). Hereditas 74, 233-238. Schieder, O.,and Vasil, I. (1980). In?. Rev. Cytol. Suppl. IlB, 21-46. Schlegel, R. (1982). Proc. Eucarpia Sym. Rye Genet., in press. Schweizer, D. (1979). Proc. John Innes Symp., 4th pp. 61-72. Schweizer, D. (1981). Hum. Gent. 57, 1-14. Scoles, G., and Kaltsikes, P. J. (1974). 2. Pjlunzenzuecht. 73, 13-43. Sears, E. R. (1952). Chromosoma 4, 535-550. Sears, E. R. (1966). “Chromosome Manipulations and Plant Genetics” (R. Riley, and K. R. Lewis, eds.), pp. 29-45. Oliver & Boyd, Edinburgh. Sears, E. R. (1967). 2. Pjlanzenzuecht 57,4-25. Sears, E. R. (1968). Proc. Int. Wheat Genet. Symp., 3rd pp. 53-61. Sears, E. R. (1972). Stadler Genet. Symp. 4, 23-38. Sears, E. R. (1973). Genetics 74, S247 (Abstr.). Sears, E. R. (1975). “Handbook of Genetics” (R. C. King, ed.), Vol. 2. Plenum, New York. Sears, E. R. (1976). Annu. Rev. Genet. 10, 31-51. Shepherd, K. W. (1973). Proc. Int. Wheat Genet. Symp., 4th pp. 745-759. Shkutina, F. M., and Khvostova, V. V. (1971). Theor. Appl. Genet. 41, 109-119. Singh, R. J., and Robbelen, G. (1975). Z. Pjlanzenzuechr. 75, 270-285. Singh, R. J., and Robbelen, G. (1976). 2. Pjlanzenzuecht. 76, 11-18. Smith, D. B., and Flavell, R. B. (1977). Eiochim. Eiophys. Acta 474, 82-97. Smith, G. P. (1973). Cold Spring Harbor Symp. Quant. Eiol. 38, 507-514. Smith, G.P. (1976). Science 191, 528-535. Spiker, S. (1976). Nature (London) 259, 418-420. Stutz, H. C. (1972). Am. J. Eot. 59, 59-70. Sumner, A. T. (1980). J . Microsc. 119, 397-406.

132

R. APPELS

Tang, K. S., and Hart,G. E. (1975). Genet. Res. (Cambridge) 26, 187-201. Tartof, K. D. (1973). Cold Spring Harbor Symp. Quant. Biol. 38, 491-500. Thomas, J. B., and Kaltsikes, P. J. (1974). Proc. Natl. Acad. Sci. U.S.A. 71, 2787-2790. Thomas, J. B., and Kaltsikes, P. J. (1976). Can. J. Genet. Cytol. 18, 687-700. Tikhonovich, I. A., and Fedeyeva, T. S. (1976). Genetika 12, 5-14. Tsuchiya, T. (1974). In “Triticale: First Man-made Cereal” (C. C. Tsen, ed.), pp. 62-89. Amer. Assoc. Cereal Chemists, St. Paul, Minnesota. Varley, I. M., Macgregor, H.C., Nardi, I., Andrews, C., and Erba, H.P. (1980). Chromosoma 80, 289-307. Vedel, F., Lebacq, P., and Quetier, F. (1980). Theor. Appl. Genet. 58, 219-224. Verma, S. C., and Rees, H.(1974). Heredity 32, 1 18-122. Vig, B. K., and Zimmerman, F. K. (1977). Environ. Exp. Bot. 17, 113-120. Vosa, C. G. (1974). Heredig 33, 403-408. Weimark, A. (1975). Hereditas 79, 293-300. White, M. J. D. (1973). “Animal Cytology and Evolution,” 3rd Ed., pp. 308-309. Cambridge Univ. Press, London and New York. Willing, M. C., Nienhuis, A. W., and Anderson, W. F. (1979). Nature (London) 277, 534. Zeller, F. J. (1973). Proc. Int. Wheat Genet. Symp.. 4th pp. 209-221. Zeller, F. J. (1976). Wheat InJ Serv. 41-42, 48-52. Zeller, F. J. (1977). Can. J . Genet. Cyrol. 19, 745-748. Zeller, F. J., and Koller, 0. L. (1981). Theor. Appl. Genet. 59, 33-37.

INTERNATIONAL REVIEW OF CYTOLOGY, VOL XO

Bioenergetic and Ultrastructural Changes Associated with Chloroplast Development A. R. WELLBURN Department of Biological Sciences, University of Lancaster, Lancaster, England I. Introduction . . . . . 11. Plastid Developme

A. Algae . . . . . .

111.

IV.

V.

VI.

VII.

VIII.

. . . . . . . . . . . . . .. . . . . . . . . . . . . . . . nt Systems.. . . . . . . . . . . . . . .

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

134 i34 134 136 136 137 137 139 139 144 144 145 145 146 146 149 149

C. Greening Angiosperms ....................... D. Angiosperms Grown in Intermittent Light . . . . . . . . . . . . . . . . . E. Light-Grown Angiosperms . . . . Semicrystalline Structures. . . . . . . . . A. Prolamellar Bodies of Etioplasts . . . . . . . . . . . . , . . B. Semicrystalline Bodies in Algae.. . . . . . . . . . . . . . . . . . . . . . . . C. Prothylakoid Bodies in Light-Grown Tissue . . . . . . . , . . . . , . . Storage Reserves and Mobilization during Plastid Development.. . A. Algae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Light-Grown Angiosperms ........................ C. Dark-Grown and Greening sperms . . , . , . , . , . , . , . . . , . Mitochondria and Respiration during Plastid Development A. Plastid-Mitochondria1 Associations . . . . . . . . . . . . . . . . . . . . . . B. Mitochondrial Association in Algae . . . . . . . . . . . , . . . . 150 C. Algal Respiratory Participation . . . . . . . . . . . . . . . . . . . . . . . . . . 151 D. Higher Plant Mitochondrial Associations . . . . . . . . . . . . . . . . . . 152 E. Respiratory Participation during Greening . . . . . . . . . . . . . . . . . 153 F. Oxygen Consumption during 154 155 G. Dark Respiration after Maturity Transfer between Cell Compartments 156 A. The Plastid Envelopes during Gr 156 B. Evidence for Changes in Fluxes 158 C. Changes in Translocators dur 160 D. Nucleotide Pool Sizes during Development . . . . . . . . . . . . . . . . 164 166 E. An Overview of Changes in F l u x . . . . . . . . . . . . . . . . , . Biogenesis of Photochemical Activities . . . . . . . . . . . . . . . . . . . . . . . 169 A. ATP Formation in Developing Plastids . . . . . . . . . . . . . . . . . . . 169 B. Development of the Photosystems . _ . . . _ _ _ _ _ _ _ _ _ 171 ..... C. Appearance of Coupling during G g _ _ . _ . . . _ _ . _ . . . . 172 .. D. Biosynthetic Requirements during Development . . . . . . . . . . . . 172 Influence of Light and Hormones.. . . . . . . . . . . . . . . . . . . . . . . . . . . 174 A. Photo-Control of Chloroplast Development . . . . . . . . . . . . . . . . 174 B. Respiratory Enhancement and Reserve Mobilization by Light . 175 C. Light Effects on Transport and Bioenergetic Parameters . . . . . 176 178 D. Hormonal Effects on Plastid Development., . ...................................... 179 191 Note Added in Proof.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I33 Copyrighl 6 ) I Y X 2 hy Academic Prar. Inc. All right\ 01 rcproduction in any Iorm reserved

ISBN 0-12-364480-0

134

A. R. WELLBURN

I. Introduction Reviews of development within chloroplasts are numerous and comprehensive (Kirk, 1970; Rosinski and Rosen, 1972; Bishop, 1974; Anderson, 1975; Arntzen and Briantais, 1975; Thornber, 1975; Leech, 1976; Boardman, 1977a,b; Boardman and Anderson, 1978). It is not the intention of this article to recover these events but to consider a series of related aspects not covered as a whole before; concentrating on those energy and ultrastructural transformations that take place mainly outside developing thylakoids before photophosphorylation is possible. Attention has been paid to the energy sources and transformations, the transport mechanisms, and the involvement of respiration by mitochondria. A wide variety of experimental systems have been used to study chloroplast development with a range of different illumination conditions and temperatures. The morphology and storage reserves of each of the major systems are described including consideration of those semicrystalline structures, usually called prolamellar bodies (PLBs), so often found during plastid morphogenesis. During development large amounts of lipids, pigments, proteins, and nucleic acid are synthesized and transformed, requiring expenditure of energy. These biosynthetic demands are discussed in relation to ultrastructure, the appearance of photophosphorylation, and the various control mechanisms involved in the regulation of these events. 11. Plastid Development in Different Systems

A. ALGAE 1. Euglena Various aspects of chloroplast development in Euglena have been reviewed already (Buetow, 1968; Schiff, 1973, 1975, 1978; Schmidt and Lyman, 1976; Buetow et a l . , 1980; Nigon and Heizmann, 1978; Partier, 1981). When grown in the dark for several generations the organisms are colorless but still retain primitive plastids, often referred to as proplastids. Over the years the term “proplastid” has been used to describe a variety of different plastid types ranging from true proplastids to etioplasts and to large immature chloroplasts from greening tissues. To avoid confusion the classification of Whatley (1977) is recommended. The early plastids of dark-grown Euglena would then be called true proplastids or eoplasts. These eoplasts are generally irregular in shape (1-2 bm in diameter) with several structures similar to PLBs (see later) at the extremities of protrusions. Other inclusions include osmiophilic globules and small amounts of primitive lamellae mainly in the form of irregular tubules. Plastid ribosomes are usually absent (Osafune and Schiff, 1980). After 30 minutes of illumination the true proplastid or eoplast assumes a more globular appearance and membrane whorls

CHLOROPLAST DEVELOPMENT

135

are found as transient structures within the PLB-like complex (Osafune et al., 1980). Thylakoids are attached to the outside of these bodies. Between 1 and 2 hours these whorls disappear and are replaced by elongated tubules which in turn are lost leaving a cavity surrounded by the remains of the original irregular tubules. By 3 hours the plastids are approximately 3 pm in size and 5 pm after 8 hours, with five to six thylakoids running the whole length of the stroma. The thylakoids are paired well before 24 hours and development fully complete after 72 hours. This general pattern can be altered by the nutritional conditions, turning off the light, adding inhibitors, using different mutants, and many other influences. These have been reviewed by Nigon and Heizmann (1978). An alternative approach to greening dark-grown Euglena is to synchronize cell division by a repetitive 1ight:dark cycle, usually of 14:10 hours. In this case the lamellae appear to grow laterally (Cook et a l . , 1976) with chloroplasts having 6-8 thylakoid bonds both at the beginning and end of the light cycle. The use of synchronized systems has been reviewed by Buetow et ul. (1980). Chloroplast multiplication may precede cell division (Cook, 1966a; Orcival-Lafont et a / ., 1972) or is coincident with cytokineses (Boasson and Gibbs, 1973).

2 . Chlamydomonas As in the case of Euglena, phototrophically grown wild-type Chlamydomonas can be synchronized to a high degree with, for example, 12 hour light:12 hour dark cycles (see Buetow et a l . , 1980). Both cell size and chloroplast volume increase substantially during the light period. Chloroplast division takes place immediately after nuclear division has occurred and cytokinesis has been initiated by a cleavage furrow (Goodenough, 1970). The chloroplast then constricts in the plane of the cleavage furrow to form daughter plastids, a process which may be repeated two or more times. An alternative system of plastid development in Chlamydomonas exploits the use of mutants such as Y1 (Ohad, 1975) or Y2 (Hudock et al., 1964) which grow heterotrophically in the dark and green rapidly in the light. 3 . Chlorella Plastid development in Chlorellu is strongly dependent upon the nutritional conditions (Aoki and Hase, 1964; Shihara-Ishikawa and Hase, 1965). Chlorella may be bleached in the presence of glucose and regreened in a glucose-free medium (Oschio and Hase, 1969; Ochiai and Hase, 1970) or by nitrogen-sparse and nitrogen-rich media, respectively (Grimme and Porra, 1974; Meller and Harel, 1978). Various synchronized cultures have been obtained (see Buetow et al., 1980) and nuclear division followed immediately by plastid division is confined to the dark period. Mutants of Chlorella exist which grow heterotrophically and synthesize chlorophyll only in the light (Dubertret and Joliot, 1974; Galling, 1978; Herron and Mauzerall, 1972; Wild, 1978). The G-I form is

136

A . R. WELLBURN

particularly unusual in that chlorophyll synthesis is stopped at temperatures over 28°C but thylakoid development still continues. 4. Scenedesmus Alternating 1ight:dark cycles have been used to synchronize Scenedesmus (Das, 1973), cell division occurring in the dark and plastid division taking place just after nuclear division. A range of Scenedesmus mutants has been produced (Bishop, 1971) some of which grow heterotrophically in the dark and green in the light, notably 2-CA’ and C-6D. Dark-grown cells of 2-CA’ contain rather better developed eoplasts than Euglena, with semicrystalline bodies resembling PLBs and single lamellar layers with few perforations (Senger et a l . , 1974; Wellburn et a l . , 1980). Perforation of the membranes precedes doubling and, later, tripling of the thylakoid bands. Plastid development is faster in C-2A’ than in C-6D despite the fact that the latter starts with the advantage of having PSI activity. Membrane-bound ribosomes are a feature of the intermediate stages of development. B. LOWERPLANTSAND GYMNOSPERMS Studies of plastid development in lower vascular plants and gymnosperms are few. Undoubtedly they deserve more attention in the future. Whatley (1971) described a series of zones of plastid development in Equisetum with true proplastids of eoplasts rapidly forming mature chloroplasts which still contain PLBlike bodies, shortly followed by an increase in osmiophilic globules and, finally, more disrupted lamellar alignments. From a survey of the literature, Whatley (1977) pointed out the high number of species in these groups known to have plastid-endoplasmic reticulum associations of one kind or another by comparison to angiosperms. Gymnosperms are particularly interesting in that etiolated seedlings of Pinus have PLB-containing plastids which synthesize chlorophyll in the dark but have no photosynthetic capacity (Treffrey, 1970; Jeske and Senger, 1978; Michel-Wolvertz and Bronchart, 1974). They also contain regular PLBs which when isolated have similar amphipathic properties to internal plastid membranes from angiosperms (Treffrey, 1976). In the Gnetales, a taxonomic group with both gymnosperm and angiosperm features, Ephedra distachya appears to resemble the gymnosperms while Welwirschia rnirabilis behaves like an angiosperm (Jeske and Senger, 1978).

C. GREENING ANGIOSPERMS By far the largest number of studies of plastid development have used the greening of etiolated seedlings as a model of chloroplast morphogenesis. The

CHLOROPLAST DEVELOPMENT

137

commonest species employed have been Hordeum, Phaseolus, Avena, Vicia, Zea, Triticum, and Pisum, probably in the order listed. Numerous reviews on the subject have appeared and the most recent are those of Boardman (1977a,b) and Boardman and Anderson (1978). The ultrastructural changes in most species are similar, the main variations being in the relative speeds of plastid morphogenesis. Etioplasts (1-2 pm in diameter) containing highly crystalline PLBs and few unperforated stromal lamellae are rapidly phototransformed (i.e., photoreduction of protochlorophyllide occurs at the same time as the PLBs lose regular crystallinity). The lamellae extend and perforate to form granal initials which subsequently extend to form recognizable granal stacks (see reviews by Rosinski and Rosen, 1972; Gunning and Steer, 1975). Over this period polyribosome configurations are associated with the developing thylakoid systems and are less apparent when mature chloroplasts are fully formed.

D. ANGIOSPERMS GROWNIN INTERMITTENT LIGHT Under intermittent light, greening proceeds rather differently. The lag phase of chlorophyll is abolished (Madsen, 1963) and an intermediate agranal plastid between an etioplast and a chloroplast, devoid of chlorophyll b, called a protochloroplast may be isolated (Akoyunoglou el al., 1966; ArgyroudiAkoyunoglou and Akoyunoglou, 1970). The PLBs persist and only occasionally are two or three thylakoid stacks observed (Sironval et al., 1968; Bahl el al., 1977).

E. LIGHT-GROWN ANGIOSPERMS Plastid development studies have been carried out using both dicotyledonous and monocotyledonous light-grown plants. Whatley (1974,1977) has shown that in all cases the plastids of the primary leaves of light-grown Phaseolus vulgaris undergo the same sequence of structural changes and has identified seven common and distinct stages of development applicable to both dicotyledonous and monocotyledonous plants with optional diversions in several species. The first three proplastid stages she describes, namely, those of the eoplast, amyloplast, and amoeboid plastids, are quite unlike plastids in mature etiolated or greening tissue. The five micrographs in Fig. la to e show the five principal stages of lightgrown plastid development in Phaseolus as outlined by Whatley (1977). Figure la shows an eoplast free from internal membranes from the plumule of 1-day germinated beans and Fig. lb illustrates the dramatic accumulation of starch to form an amyloplast 1 day later. The plastid in Fig. lc from 3-day germinated beans exhibits an amoeboid appearance after the starch has been metabolized. The invagination of the inner envelope to form the early internal membrane

138

A. R. WELLBURN

FIG. 1. Development of plastids in light-grown Phaseolus. (a) Eoplast. X31,200. (b) Amyloplast packed with starch (S) granules. x 18,000. (c) Amoeboid plastid showing invaginations (arrowed) of the inner envelope. X40.500. (d) Immature plastid with typical prothylakoid body. X31.200. (e) Mature plastid also with invaginations (arrowed). x27,OOO. The solid bars represent a distance of I pm. All plates in this figure were kindly provided by Dr. J. M. Whatley, Oxford.

vesicles can clearly be observed (arrowed). An immature bean plastid (5 days from germination) is shown in Fig. Id. The irregular PLB-like structure still retains evidence of the pinocytotic vesicular association with the inner envelope. After 7 days of growth, mature bean chloroplasts are observed similar to that

CHLOROPLAST DEVELOPMENT

139

shown in Fig. le. Only a suspicion of pinocytotic vesicular association exists at either end of the plastid (arrowed) and starch grains similar to those shown are most likely formed by internal light-driven events. Monocotyledons differ developmentally from dicotyledonous plants by virtue of having a basal intercalary meristem (Esau, 1953). Leech and co-workers have undertaken a series of light-grown studies using both maize and wheat. Leese er al. (1971) first showed that different populations of plastids could be effectively prepared from segments taken along the length of maize seedlings. Previously Robertson and Laetsch (1974) had shown that the age of developing etiolated tissue has a considerable effect on the rate of chlorophyll formation during greening. Boffey er al. (1980), using light-grown wheat to avoid dimorphism, have extended this concept to show that the greening of mature etioplasts cannot be used as a model for the normal development of eoplasts into mature chloroplasts, as not only exposure to light, but also chronological age of the plastid, are important factors in development.

111. Semicrystalline Structures A. PROLAMELLAR BODIESOF ETIOPLASTS This semicrystalline structure is remarkable not only in appearance and phototransformation but also for the number of different names it had during early descriptions (see Gunning, 1965a) before the name prolamellar body was widely adopted. Full descriptions of the architecture of the PLB are to be found elsewhere (Gunning and Steer, 1975). Particles are observed in the stroma spaces of PLBs from a wide range of different genera when glutaraldehyde-osmium teroxide fixation is used (see Wellburn et a l . , 1977). These have been referred to as “ribosome-like’’ on the basis of size and staining characteristics (Gunning, 1965) and their possible role as shape-determining factors in PLB morphogenesis has been the subject of conjecture (Gunning and Jagoe, 1967). The application of freeze-fracture techniques to the study of PLBs is particularly difficult and the few reported investigations have served only to confirm the lattice dimensions of the PLBs (Bronchart, 1970a,b; Phung Nhu Hung ef al., 1970a; Ophir and Ben-Shaul, 1974; Ophir er al., 1974; Pyliotis and Goodchild, 1975). Simpson (1978) has managed to obtain better replicas which confirmed the model proposed by Gunning (1965a) and related the dispersal to the appearance of the early thylakoids with high particle densities and the later perforated lamellae with lower particle numbers. Nevertheless, replica evidence for particles within the lattice corresponding to the “ribosome-like” particles has not hitherto been described. Figure 2c shows a stereo-pair of micrographs which should be viewed at a distance of 12 cm using a commercial 3-D hand viewer

140

A . R . WELLBURN

CHLOROPLAST DEVELOPMENT

141

(e.g., Casella). These are reproduced from the unpublished work of Goodchild, Gunning, and Wellbum carried out in Canberra in 1976 using isolated Avena etioplasts. The detail of the PLB rpelica in Fig. 2c reveals central depressions within each of the lattice spaces which are made by the ribosome-like particles. On the upper left-hand side of the PLB proper there is an area which clearly shows the tetrahedrally branched tubular units, one of the most common of the different crystalline arrangements in regular tight PLBs (Gunning and Steer, 1975). On the upper right periphery of the PLB in Fig. 2c are two raised areas which show the surface detail of the prothylakoids. The high particle density may reflect the large number of CF, particles known to be associated with these early lamellae (Wellburn, 1977). There have been many attempts to isolate PLBs from preparations of etioplasts (Kahn, 1968a,b; Kahn et a l . , 1970; Treffrey, 1970; Lutz, 1975a,b; Bahl and MonCger, 1976; Bahl et a l . , 1976) but in none of these preparations was the retention of these particles within the isolated PLBs demonstrated. Wellburn et al. (1977) described the retention of these particles in isolated PLBs and were able to completely remove them with ribonuclease treatment. At the same time vesicles heavily studded with coupling factor particles (CF,) were observed to be associated with the periphery of the isolated PLBs. This observation was taken further by Wellburn (1977) using the CF, visualizing procedure of Oleszko and Moudrianakis (1974) to show that CF, was only attached to lamellar membranes of the PLB and not the PLB proper. Increasing length of illumination had the effect of progressively increasing the mean distance between the CF, of the lamellae. It may be that the amphipathic changes of etioplast membranes previously described by Treffrey (1974, 1975) using countercurrent studies may be due to this dual property. At present it is not known if there are CF, components in prothylakoids or how their numbers relate to those of the CF, . The early lamellae (now referred to as prothylakoids) readily form vesicles in vitro which are called prothylakoid vesicles (PTVs). Subsequently Hampp and Wellburn (1978) and Lutz (1978) independently developed methods for the separation of PTVs away from the PLBs. The early photochemical events connected with plastid development were associated only with the PTVs (Hampp and Wellburn, 1978; Wrischer, 1978; Wellburn and Hampp, 1979) while the PLB proper contains two saponins, avenacosid A and B (Lutz, 1980) previously described in crude PLB preparations by Kesselmeier and Budzikiewicz (1979). ~~

~ _ _ _ _ _ _

FIG. 2. (a and b) Immature plastids and mitochondria from light-grown Hordeurn. On the left (a) a membrane whorl (arrowed) separates the two organelles. X36.000. To the right (b) a mitochondrion is apparently enveloped by the plastid by the under-flow process discussed in the text. X32.000. The two freeze fracture stereo-pairs (c) were prepared by Dr. D. Goodchild (Canberra) from isolated Avena etioplasts (see text for viewing details). X45,OOO. The solid bars represent a distance of I pm.

142

A. R . WELLBURN

These saponins in association with the polypeptide (MW = 21,000) reaggregate in vitro to form tubules resembling native PLBs (Ruppel et al., 1978; Kesselmeier and Ruppel, 1979). Another polypeptide (MW = 36,000) has also been found as a major component to PLBs which disappears during greening (Murakami, unpublished observations). Upon extraction of this protein from isolated PLBs, disruption to the crystallinity occurs similar to that observed in situ during development. Purified NADPH:protochlorophyllide oxidoreductase from solubilized PLBs has also been found to have the same molecular weight of 36,000 (Ape1 et al., 1980). Studies of the PLB proper are hampered by unavoidable PTV contamination but Lutz and Manning ( 1980) concluded that most of the protochlorophyllide is located in the prothylakoids of etioplasts. Lutz et af. (198 1) have established that NADPH:protochlorophyllideoxidoreductase (Griffiths, 1978) is also associated with the prothylakoids showing the same distribution pattern previously established for CF,-ATPase (Wellburn, 1977; Hampp and Wellburn, 1978; Wellburn and Hampp, 1979). Recently Lutz (1981b) in a review of his work has listed some of the components of PLBs proper and the prothylakoids. He has also pointed to the possibility of diverted isoprenoid biosynthesis during etiolation after the farnesyl pyrophosphate (not farnesol) stage, previously defined by Wellburn (1978), as a possible explanation for the enhanced formation of the steroidal saponins in prolonged darkness. Two other activities have been associated with PLBs. Doll et al. (1976) demonstrated a Cn2+/Zn2 -containing superoxide dismutase in crude PLB preparations while Hampp and DeFilipis (1980) showed the presence of two proteases specific to Avena plastids which exhibited their highest specific activities with purified PLB preparations (i.e., some PTVs removed). This suggests that proteolytic breakdown of PLB polypeptides is an early event of phototransformation and special protection of the early photochemistry is necessary. Carotenoids have recently been shown to protect the mechanisms involved in phototransformation from photodestruction (Ryberg et a f . , 1981). Two other research interests closely connected with isolated etioplasts and prolamellar bodies are (1) the amount of phototransformation that is possible in vitro and (2) the appearance of regularly spaced parallel arrays of tubules often found during these studies. Ultrastructural changes in isolated illuminated etioplasts have often been described (Klein and Poljakoff-Mayber, 1961; Wellburn and Wellburn, 1971; Wrischer, 1973a,b; Kohn and Klein, 1976) showing that a structural transformation and partial dispersal of the PLBs occur when isolated etioplasts are illuminated in suspension. Changes outside the PLB proper are more difficult to define. Increases in primary thylakoid length and area of plastid profiles have been reported but apparent changes in the nature of the lamellae are uncertain. This appears mainly due to a question of definition rather than differences be+

CHLOROPLAST DEVELOPMENT

143

tween species and incubation conditions. Wellburn and Wellburn (1971, 1973a) have used the term “bi-thylakoid” to describe a perforation with slight initial overlap without extension (i.e., grana initials at Stage F of Wellburn and Wellbum, 1971) often found even in older unilluminated etioplasts but Wrischer (1973a) and Kohn and Klein (1976) have unfortunately taken this term to mean fully recognizable grana (i.e., like stage G of Wellburn and Wellburn, 1971). This condition is never reached however long the plastids are incubated. Perhaps “grana initials” is a better term for this stage of development. Whatley (1977) uses the less definitive term ‘‘incipient grana” for a similar stage in light-grown tissues. Arrays of regular tubules (25-30 nm outer diameter) have often been reported in situ after indifferent processing for electron microscopy (von Wettstein and Kahn, 1960; Ericksson et al., 1961; Henry, 1979) or after incubation of isolated etioplasts (Appleqvist et al., 1968; Kahn, 1968b; Wellburn and Wellburn, 1971, 1973a; Wrischer, 1973a; Kohn and Klein, 1976; Simpson, 1978). These tubules can be induced by a variety of incubation conditions: high or low temperatures or light intensities, blue light, darkness, or even poor fixation. They do not appear to be part of development as observed in in situ plastids from plants greened under normal conditions and well prepared for ultrastructural examination. In a later study, Wellburn et al. (1977) concluded that their formation was induced by a critical loss of ions during the incubation which in turn probably causes additional reaggregation of avenacosid A and B with the 21 kD PLB polypeptide (see earlier) within the disturbed plastids. These tubules differed in size from those discrete arrays of much smaller tubules (10-15 nm diameter) which are sometimes observed in situ in light-grown barley tissue (Wellburn et al., 1982) in close proximity of PLB-like structures. These small tubules are similar to those in dark-grown barley described by Sprey (1975), both of which resemble sections of those protrusions of the inner plastid envelope recently described by Newcornb and co-workers. Reports of dark-grown tissue without PLBs are rare. Stetler (1973) attributed the lack of PLBs in plastids from dark-grown tobacco tissue culture to a failure to photoconvert protochlorophyllide to chlorophyllide. Klein and Schiff ( 1972) reported that the eoplasts of 3- to 4-day-old dark-grown bean seedlings did not contain PLBs and concluded that conventional PLBs of etioplasts were not an obligate step in chloroplast development. Whatley (1974) from a more comprehensive survey showed that 3-day and older bean plastids did contain PLBs and that the frequency of plastids with PLB-like structures was much higher in a 12 hour light:12 hour dark cycle than if they were always kept in the dark. In the early stages of growth of etiolated oat seedlings larger proportions of perforated lamellae in relation to PLBs exist together with some starch grains (Liitz, 1981a). However, as the etiolated seedlings grow the PLBs become bigger and starch disappears entirely. When senescence takes place after 1 1 days the PLB numbers

144

A. R. WELLBURN

and sizes are not significantly reduced but the proportional amount of lamellar material is much smaller. B. SEMICRYSTALLINE BODIESIN ALGAE PLB-like structures have been described in all the greening algal systems previously outlined. Generally these structures are smaller and more irregular and amorphous in appearance by comparison to the conventional PLBs of etioplasts. Brief descriptions of PLB-like structures in Euglena are frequent (Schiff, 1970, 1978; Klein et a l . , 1972; Salvador et a l . , 1971; Schwartzbach et a f . , 1974; Nigon and Heinzmann, 1978) but recently Osafune and Schiff (1980) and Osafune et al. (1980) have described their structure, transformation, and association with membrane whorls in considerable detail. In the greening mutants of Scenedesmus (Senger et al., 1974; Wellburn et al., 1980), Chlamydomonas (Friedberg et al., 197 l), and nitrogen-deficient Chlorella (Pyliotis and Goodchild, 1975) the semicrystalline bodies do not have the regular appearance of PLBs and rapidly disperse on greening.

C. PROTHYLAKOID BODIESIN LIGHT-GROWN TISSUE Light-grown tissues may contain PLB-like structures especially in plastids of cells close to the meristem. They are always much smaller than etioplast PLBs and usually occur toward the edges of the plastid discs. Sometimes they have a regular structure as in light-grown maize (Rascio et a l . , 1976) or beans (Wrischer, 1966) but usually are amorphous in appearance as in sunflower (Walles, 1966), Equisetum (Whatley, 1971), beans (Whatley, 1977), and barley (Wellburn et a l . , 1982), very similar to the semicrystalline structures described in algae (see above). The name prothylakoid bodies (PTBs) is proposed to describe these structures in light-grown tissues and distinguish them from classical PLBs. An alternative name, tubular connections, has already been proposed by Platt-Aloia and Thomsom (1977) but unfortunately this term has the possibility of confusion with the additional artifactual tubular connections associated with detached PLBs or isolated etioplasts (see Section III,A) and consequently to be avoided. When attempts were made to prepare plastid membranes containing these by the PLB isolation method of Wellburn et al. (1977), no trace of crystalline-like bodies was found to be associated with the lamellae (Wellburn et al., 1982). This implies a high degree of instability of PTBs under conditions in which etioplast membranes readily retain PLBs (Wellburn et al., 1977). It is possible that PTBs are converted completely to PTVs in vitro. Vesicles were often found adhering to where PTBs ought to have been by comparision to the in situ appearance of the

CHLOROPLAST DEVELOPMENT

145

lamellae systems. If this is so then PTBs are probably wholely engaged in photochemical assembly and are unlike the PLBs which also contain additional material as described earlier.

IV. Storage Reserves and Mobilization during Plastid Development A. ALGAE The diversity of ultrastructural characteristics within the different algal groups, especially with respect to pyrenoid structure and appearance, has been delightfully outlined by Manton (1967). Many of the questions she posed to the experimentalists are still unanswered. The bulk of the effort to remedy this has still been devoted to the Euglenophyceae and Chlorophyceae. The major storage reserve in Euglena is paramylum which mainly exists as a (3- 1,3-glucan, amounts of which are controlled by many factors, principally carbon-nitrogen rations, pH, and phosphate levels (Freyssinet et al., 1972; Freyssinet, 1976a,b). Levels in dark-grown resting Euglena are usually high but are rapidly reduced upon illumination (Smilllie et al., 1963; Dwyer et al., 1970; Dwyer and Smilie, 1971; Freyssinet e t a l . , 1972; Schwartzbach etal., 1975). If illumination is stopped the paramylum degradation rate slows down and it is then reaccumulated (Freyssinet et al., 1972). Free sugars, principally galactose (Schantz et al., 1971) and glucose (Schwartzbach et al., 1975) together with enzymes of metabolism, p- 1,3-glucan hydrolase and phosphatase (Dwyer and Smillie, 1970), increase while paramylum is degraded upon illumination of darkgrown cells. Other storage materials appear to be excess RNA, which is broken down upon illumination (Freyssinet and Schiff, 1974; Nigon and Heizmann, 1978), and phosphate contained within polyphosphate bodies (Freyssinet et al., 1972) located in the cytoplasm. Osmiophilic globules are evident in micrographs of most plastids. In light-dark synchronized cultures of Euglena the pyrenoid is present over the first half of the light cycle (Konitz, 1965; Cook et al., 1976) and the content of the paramylum per cell increases 7-fold over whole light period (Cook, 1966b). During the dark period the paramylum is used up for general metabolism and cell division (Cook, 1963). Chlamydomonas plastids also have central pyrenoids which in light-grown cells have characteristic ensheathing starch grains. In dark-grown Y-I cells the pyrenoid is free from the starch sheath and large numbers of independent starch grains are scattered throughout the stroma (Ohad, 1975). During illumination these are rapidly reduced and the starch shells are fully reformed after 9 hours (Ohad, 1975). Osmiophilic globules are present in all types of plastid and there

146

A. R. WELLBURN

appear to be vesicles (containing material other than polyphosphate) within the cytoplasm. In Chlorella the single chloroplast of each autospore contains a small protein body which develops into the pyrenoid during the light (Wanka, 1975). There are also small starch grains within the chloroplasts of the autospores. In light-dark cycles the starch grains disappear during the last 6 hours of darkness as starch layers are formed around the pyrenoid. In the light the main accumulations of starch around the pyrenoids occur between 4 and 12 hours and thereafter are degraded, with increasing numbers and amounts of independent starch grains appearing from 4 hours onward. In Scenedesmus the pyrenoid does not divide but disappears during nuclear division (Bisalputra and Weier, 1964). Usually there is a starch shell surrounding the pyrenoid. On the outside of the plastid envelopes, toward the base of each pyrenoid, microbodies are often found, an association previously described by Gunning and Steer (1975). Vacuoles in the rest of the cell are filled with two types of material (Wellburn et al. , 1980). Those opaque to electrons are probably filled with polyphosphate (Nilshammer and Walles, 1974) while the electrontranslucent material located near the plasma membrane may be involved in pinocytotic mechanisms releasing and forming daughter cell wall material (Wellbum et al., 1980). Independent and very large starch grains abound in darkgrown cells (Senger et al., 1974). These are rapidly reduced in size and disappear during illumination. Osmiophilic globules are present in both Chlorella and Scenedesmus but are not numerous. In Eremosphaera, Holdsworth (1971) has shown pyrenoids to comprise 40% of the ribulose-l,5-bis-phosphate carboxylase (RuBPC) within the cell. This is in agreement with the observation that, in light-dark synchronized Euglena, pyrenoid bodies in the early light phase give way to high RuBPC activity in the second half of the light period (see Buetow et al., 1980). More detailed studies of algal pyrenoids are required. B . LIGHT-GROWN ANGIOSPERMS It has long been known that starch is synthesized and stored as granules in mature chloroplasts after photosynthesis (Sachs, 1862). Formation of starch from triose phosphate is favored by the activation of ADP-glucose pyrophosphorylase brought about by a combination of high 3-phosphoglyceric acid (PGA) and relatively low concentrations of inorganic phosphate (Preiss et al., 1967). Starch synthesis in green plants may also be promoted in the dark by glucose (Boehm, 1883a,b; Brown and Morris, 1893; Parkin, 1899; Tollenaar, 1925; Phillis and Mason, 1937; MacLachan and Porter, 1959; Chen-She et al., 1975) by either glycolytic formation of triose phosphate followed by import of triose phosphate or a slow penetration of hexose phosphate (Walker, 1976). The results of Mac-

147

CHLOROPLAST DEVELOPMENT

Lachan and Porter (1959) favor the latter. In either case activation of ADPglucose pyrophosphorylase would be required for subsequent starch synthesis. Proplastid development in light-grown plants involves an obligatory amyloplast stage (Whatley, 1977). Earlier eoplasts do not contain starch. During later amoeboid and immature stages the sizes and numbers of starch grains fall, only to rise again when photoreduction, photophosphorylation, and CO, fixation have been operative for some time. Figure 3 shows the starch to plastid ratios in plastids along the leaf blade measuring upward from the intercalary meristem in 7-day-old low-light-grown barley seedlings. The amyloplast stage containing imported starch reserves is clearly evident around 0.3-0.8 cm from the meristem while later accumulations around 2-4 cm are a consequence of local photosynthetic activity. The same figure also shows the dramatic increase in plastid volume upward along the blade as more mature plastids are formed. A major area of division occurs around 1 - 1.75 cm where amoeboid plastids are the majority. Prothylakoid bodies (PTBs) are encountered at all stages of plastid development. The frequency of observation falls during development because the increase in plastid volume reduces the chances of intersection. Levels of inorganic phosphate, PGA, and ATP along the blade of light-grown barley suggest that over the eoplast-amyloplast stages activation of ADP-glucose pyrophosphorylase, the regulatory enzyme of starch biosynthesis (Ghosh and Preiss, 1966), is favored by high PGA and ATP together with low levels of inorganic orthophosphate which are normally inhibitory (Fig. 4). Higher up the laminae the situation is reversed as amoeboid plastids turn into immature plastids. Only when the chloroplasts are "-. ...... ...2....

r-..B Ranges of plastid types 79 50-A 7:::: D

1

x

i I I

.'

/

I

Plaslid Volume

~

/

1

1

2

3

4

5

6

7

Centimeters up leaf from meristem

FIG. 3 . Changes in plastid volume, numbers of plastids with intersected PTBs, and ratios of starch to plastid on an area basis in low light-grown barley seedlings (500 lux for 16 hours plus 8 hours darkness for each of 7 days at 20°C) measuring upward from the intercalary meristem. Uppermost are the approximate ranges of the various plastid types (A, eoplasts; B, amyloplasts; C I , amoeboid plastids; Cz, immature chloroplasts; D, chloroplasts) using the classification of Whatley (1977) and the same nomenclature applicable to Fig. 5A-D.

148

A. R . WELLBURN

FIG.4. Changes in the levels of orthophosphate, ATP, and 3-phosphoglycerate (PGA) in the same experimental system as that of Fig. 3.For fuller experimental details see Wellburn et al. (1982). The curves in Figs. 3 , 4 , 6 , and 7 are plots of cubic splines (Ahlberg etal.. 1967) generated by means of a computer program (Parrot, 1981) written in BASIC.

fully formed and have been functional for some time are more favorable ATP and PGA to inorganic phosphate ratios reestablished allowing starch synthesis once more (see Note Added in Proof). Osmiophilic globules are present in most chloroplasts (Park and Pon, 1961) and appear to increase during later stages of maturity especially during senescence. They may be isolated (Bailey and Whyborn, 1963) and appear to contain increasing amounts of isoprenoids as aging progresses (Lichtenthaler, 1966, 1968; Wellburn and Hemming, 1967). Protein bodies are frequently found in light-grown plastids, especially after stress, inhibition, or infection (Buvat, 1959; Thomson et al., 1965, 1966; Cronshaw et al., 1966; Dolzmann and Ullrich, 1966; Lemoine, 1966; Price et al., 1966; Purcifull et al., 1966; Esau and Cronshaw, 1967; Newcomb, 1967; Shumway et al., 1967; Stetler and Laetsch, 1969; Cran and Possingham, 1972, 1973, 1974; DeGreef and Verbelen, 1973). Early suggestions that protein deposits are directly involved in the development of grana are probably wrong. They are most likely to be centers of crystallized RuBPC. When this occurs as it does sometimes within the lumen of the thylakoids, causing distortion, a membranebound appearance is given. Phytoferritin particles are not often observed in light-grown plastids although they have been observed in amyloplasts of Phaseolus seedlings germinated and grown in the light. They are much more pronounced in plastids of the hypocotyls of dark-grown plants (Whatley, 1977). It is probable that iron required for development is temporarily stored as nontoxic phytoferritin before it is required for cytochrome biosynthesis, etc., and only when this normal development is frustrated by darkness or inhibition are extensive accumulations observed (Hyde et al., 1963; Robards and Humpherson, 1967; Whatley, 1977).

CHLOROPLAST DEVELOPMENT

149

C . DARK-GROWN AND GREENING ANGIOSPERMS Starch grains are not always observed in etioplasts. In some species such as Avena they are extremely rare, occasionally present in barley and wheat etioplasts and more common in etioplasts of species with large seed reserves such as maize, peas, and beans. As a consequence senescence is observed earlier in dark-grown cereal tissues than in maize and legumes. Different rates of greening are observed not only between species but between dark-grown seedlings of the same species but of different ages (Robertson and Laetsch, 1974). Undoubtedly this is due to the different amounts of storage reserves available to the different plants or tissues to finance the greening process. Studies to demonstrate and substantiate this have not been undertaken. In dark-grown Avena, starch reserves are rare even at the 3-day-old stage (Liitz, 1981a) and much of the metabolism appears to be directed to the formation of large quantities of steroidal saponins which may be, in part, incorporated into the PLB proper (see earlier). This material is possibly broken down and rernetabolized in place of starch providing suitable precursors from sugars and sterol alike, and, in addition, releasing energy upon degradation. Protein bodies are also frequently observed in dark-grown Avena tissues (Gunning, 1965b; Gunning et al., 1968) and isolated oat etioplasts (Wellburn and Wellburn, 1971). The protein bodies normally take the form of whorl-like “stromacenters” which consist of a certain form of crystallized RuBPC (Steer et al., 1968), possibly just the large subunit. If isolated oat etioplasts are subjected to osmotic stress they contain regular crystals of RuBPC-like material as well as the normal stromacenters (Wellburn and Wellburn, 1971). During greening of etiolated oats stromacenters are much reduced and often disappear entirely. This may be due to sufficient small subunits being available and the increase in plastid volume reducing the internal concentration of RuBPC below a critical crystallization level. Osmiophilic globules are contained in all etioplasts and etiochloroplasts. They are mainly associated with or trapped within PLBs even after isolation (Wellburn et al., 1977). They appear to reduce in number and size during greening but this impression has not been fully investigated. V. Mitochondria and Respiration during Plastid Development A. PLASTID-MITOCHONDRIAL ASSOCIATIONS Claims of ultrastructural associations between organelles especially plastids and mitochondria are haunted with problems. The terminology used by many descriptions is confusing. A suggested terminology would be as follows: (1) The

150

A. R. WELLBURN

complete communication of mitochondria1 matrix space with plastid stroma and the inner and outer envelopes of both organelles shown only as two continuous enclosing boundaries, if it ever exists, would be calledfused. (2) Where matrix space and stroma remain distinct with their own inner envelopes, but having unified outer envelopes, would be deemed to be conjoined. This may be of two types. Close conjoining where large areas of inner envelopes are close to each other or distant conjoining where connections between the outer envelopes are tubular or have an endoplasmic reticulum-like appearance. (3) Where no clear joining of the outer envelopes is observed but the two organelles are clearly very close to each other they would be described as being appressed. Observations using the light microscope and the transmission electron microscope are not in accord. The former is three-dimensional in nature however thin the sections may be and the latter highly prone to the creation of artifacts at all stages of processing prior to examination. All electron microscopists, observational as well as experimental, should watch the cinefilms associated with the work described by Mersey and McCully (1978). Fixation processes for tissue preservation, even the best, are far from perfect, taking several minutes to complete, long enough for organelles to detach themselves one from another. Freeze-fracture studies have not revealed associations claimed by transmission microscopy. Consequently great caution should always be exerted when interpretations of continuities are made. B. MITOCHONDRIAL ASSOCIATION IN ALGAE Close appressions of plastids and mitochondria are often observed in all groups of algae. So common are they that they rarely receive comment. Particularly good examples of these appressions are to be found in those illustrations provided by Manton (1967), Bisalputra and Bisalputra (1969, 1970), and Cattolico et al. (1976). Appressed mitochondria and plastids are observed in Euglena (Schiff, 1970) but suggestions of fusion (Osafune et al., 1980) probably mean close conjoinings, using the suggested terminology from above. They are also interesting in that they take place in close proximity to the myelin-like membrane whorls which are such a feature of plastid development in Euglena. These membrane whorls are also observed to penetrate to the cores of microbodies which are in close proximity to the plastids (Osafune and Schiff, 1980; Osafune er a l . , 1980). Myelin-like membranes have been described before in a variety of organisms (Schiff, 1970; Bobak and Herich, 1978; Goff, 1979; Swanson and Floyd, 1979) including mutant higher plant plastids (Wellburn and Wellburn, 1979). Occasionally myelin-like membrane whorls occur in normal higher plant tissue. Figure 2a shows a whorl appressed between mitochondria and an immature plastid in 7-day-old barley grown normally in the light.

CHLOROPLAST DEVELOPMENT

151

Appressions of mitochondria and plastids have also been described in Chlamydomonas (Osafune et a l . , 1972) and Scenedesmus (Wellbum et al., 1980). In the latter, however, they were only observed in large numbers during illumination of dark-grown cells. When Scenedesmus was grown heterotrophically or phototrophically for some time these associations were not observed. C. ALGALRESPIRATORY PARTICIPATION Increased respiratory rates are measured when dark-grown Euglena are illuminated (Schiff, 1975; Michel, 1978) and 3-(3,Cdichlorophenyl)-l,l-dimethylurea (DCMU) fails to arrest plastid development (Schiff, 1975). A decrease in glycolytic activity during illumination has been reported (Dockerty and Merrett, 1979)although fermentative ATP formation in Euglena is probably very low (Schiff, 1975). Evidence of diversion of metabolites and reduction in the levels of mitochondrial proteins as development of dark-grown Euglena is completed has been provided (Bovamick et al., 1974;Davis and Merrett, 1974; Brown and Preston, 1975;Brunold and Schiff, 1976;Wolff and Schantz, 1978). At the same time mitochondrial sizes are reduced from 13 to 6% of total cell volume (Wolff and Schantz, 1978). Hormm and Schwartzbach (1980)have examined fumarase and succinate dehydrogenase activities over the whole period of illumination. During the first 8-12 hours these enzymes increase in specific activity and thereafter decrease. Their induction is directly controlled by a nonchloroplast-based photoreceptor and not a consequence of photoinduced carbohydrate breakdown. Interpretations of experiments using inhibitors or drugs should always be treated with caution. Problems due to lack of penetration (Schiff, 1975) or unusual ultrastructural changes such as extra large mitochondria1 formation (Lefort-Tran, 1975) may be encountered. Inhibitors of oxidative phosphorylation, such as dinitrophenol (Smillie et a l . , 1963) or antimycin A (Calvaryac et al., 1971), or of protein synthesis (chloramphenicol and streptomycin) prevent plastid development and stimulate mitochondrial enlargement in Euglena (BenShad and Ophir, 1970;Neumann and Parthier, 1973)at the same time. When glucose-bleached cells of Chlorella are illuminated there is an early increase in respiration followed by a later decline. Some inhibitors of cell respiration, such as arsenate, arsenite, and cyanide, suppressed plastid development while others (azide, sodium fluoride, amytal, and dinitrophenol) enhanced the rate of chloroplast morphogenesis (Matsuka and Hase, 1966). DCMU suppressed instead of enhanced the rate of greening in Chlorella, exactly the opposite of the response given by Euglena. These results clearly illustrate the problems associated with interpretations made of inhibitor effects upon whole organisms.

152

A. R. WELLBURN

During the illumination of the Scenedesmus mutant C-2A’, respiration rates increase dramatically over the first 2-4 hours of illumination and then fall equally quickly to rates lower than those to be found in dark-grown cells (Senger and Bishop, 1972). At the same time the rates of starch breakdown decline when peak respiratory activity is attained (Brinkman and Senger, 1978). In the same mutant fluoroacetate, arsenite, and DCMU have no effect on respiration but inhibit greening and photosynthesis. Azide and cyanide affect all three processes while dinitrophenol inhibits greening and photosynthesis but enhances respiration (Senger and Bishop, 1972). In light-grown Scenedesmus the tricarboxylic acid rates are the same in the light or the dark (Marsh er al., 1965) which contrasts with the observations using higher plants (see later). D. HIGHERPLANTMITOCHONDRIAL ASSOCIATIONS Experiments using phase cinephotomicrography with the light microscope have revealed streaming phenomena along the pleomorphic canalicular system in the cells of higher plants. While this is in progress, fusion events apparently take place between plastids and organellular structures indistinguishable from mitochondria (Wildman et a l . , 1962). Experiments using the transmission electron microscope have been undertaken to verify or refute these in vivo observations. The possibility of artifacts due to preparation for electron microscopy has already been mentioned. An early study by Vesk et a l . , (1965) provided evidence that both organelles undergo independent divisions in young tissues and, in addition, suggesting a possible mechanism of mitochondria flowing beneath the margin of more static plastids into pockets from which they are subsequently released. This would explain electron micrographs of mitochondria enclosed within plastids such as those shown by Montes and Bradshaw (1976) and that in Fig. 2b. Mitochondria-like protruberances from plastids with both envelopes clearly in continuity around the whole were also described (Vesk et a l . , 1965; Valanne and Valanne, 1972). With improvements in methods of preparation for electron microscopy, such as the use of glutaraldehyde as a fixative, reports of these apparent fusions ceased and were replaced by claims of distant conjoinings between mitochondria and plastids (Crotty and Ledbetter, 1973; Schotz and Diers, 1975). General experience has been that even conjoinings are extremely rare especially in mature tissues. Electron microscopists when observing an apparent conjoining and having access to a tilt facility are usually able to resolve and distinguish the outer envelopes of each organelle. The nearest we have come to conjoined plastids and mitochondria of the close type was to be found in meristemmatic tissues of barley, especially in tissue which is responsible for the formation of the white parts of the striped leaves of the nuclear gene-induced barley plastome mutant “albostrians” (Wellburn and Wellburn, 1979, 1980). In this mutant there is a total absence of plastid ribosomes at all stages of plastid

CHLOROPLAST DEVELOPMENT

153

development and a deficiency in the synthesis of many proteins such as CF, polypeptides (Borner et al., 1976, 1979). With the lack of ATP formation by photophosphorylation it might be anticipated that the closest of relationships exists between the parasitic plastids and the normal mitochondria which increased our chances of observation. In comparisons of the cinefilm of Wildman et al. (1962) with in vivo streaming as seen directly in, for example, isolated protoplasts, the former loses the threedimensional advantages of direct viewing. Mitochondria that would appear to fuse are often found to glide along a different plane and disappear as distinct entities below the plastids when the plane of focus is changed, exactly as suggested by Vesk et al. (1965) and Wildman et al. (1974). The latter suggest that an association occurs where starch formation takes place but only speculate on the possibility of fusion. The consensus of view at present would favor close appression as the normal type of association between mitochondria and plastids with close conjoining only a possibility in very early stages of development of illuminated dark-grown algae (Osafune et al., 1980) or light-grown (Wellburn and Wellburn, 1980), but not etiolated, higher plant tissue. The problems of interpretation of transmission electron micrographs have long been realized (Vesk et al., 1965). Clearly other techniques must be employed if the problems of mitochondrial-plastid associations are to be resolved. Freezefracture relicas promise to be of use provided that the fracture face details of the different envelopes are established and prove to be distinguishable. Alternatively, antibodies with heavy metal, radioactive, or fluorescent tags could be raised to the different surfaces of the various envelope membranes. The demonstration of flow (or nontransfer) of antibody markers from mitochondria to plastid and vice versa would be useful to assess the nature of any association.

E. RESPIRATORY PARTICIPATION DURING GREENING Measurement of respiration is much easier in etiolated and greening tissues than in either algae or mature higher plants. Oxygen uptake by dark-grown leaves after irradiation has been described many times (Gabrielson et al., 1961; De Greef et al., 1976; Selldtn and Selstam, 1976; Redlinger and McDaniel, 1977, 1978). This may be considerably increased by 6-amino-levulinic acid treatment (Carlsson and Sundqvist, 1979) but this increase is cyanide-insensitive and not related to mitochondria1 activity. In vivo fluorescence changes during the greening of Euglena in intermittent light have revealed that once the pool of phototransformable protochlorophyllide is exhausted, its reformation and synthesis of reductant appears to be dependent upon respiration (Michel, 1978). This reductant is now known to be NADPH (Griffiths, 1978) which is also required for the hydrogenation of the chlorophyll side-chain (Wellburn, 1968; Benz et al., 1980) to form phytylated chlorophyll a

154

A. R. WELLBURN

through the four stage esterification route (geranyl-geranyl pyrophosphate + chlorophyllide a + chlorophyll,, + chlorophyll,,,, + chlorophyll,,,, + chlorophyll,) elucidated by Schoch et al. (1977). Recently it has been shown that these isoprenoid hydrogenations are much more sensitive to anaerobiosis than phototransformation because the geranyl-geranyl ester of chlorophyllide accumulates under these conditions (Schoch et al., 1980). Greening occurs normally in a variety of atmospheres but there is a requirement for the presence of small amounts of oxygen (Smith and Young, 1956; Steer and Walker, 1965; Wolf and Kidd, 1973). Furthermore plastid development is completely inhibited by high levels of CO,, methane or ethylene, and partly by carbon monoxide. Experiments using respiratory inhibitors such as cyanide or azide (Wolf and Kidd, 1973) or uncouplers like dinitrophenol (Wolf and Kidd, 1973; DeGreef and Verbelen, 1977) demonstrate the necessity for the production of additional energy from respiration and oxidative phosphorylation to assist with plastid development. Isolated mitochondria from greening Avena laminae exhibited a pronounced increase in respiratory chain phosphorylation (Hampp, 1979; Hampp and Wellburn, 1980) when driven by externally supplied NADH through the rotenoneinsensitive pathway of plant mitochondria (see Palmer, 1976). The rates of oxidative phosphorylation were highest after 2 hours of illumination and declined thereafter to rates at 24 hours which were comparable to those in mitochondria from etiolated tissue (Hampp and Wellburn, 1980). Similar results are reproduced in Fig. 6 in relation to the rates of DAD-dependent light-driven ATP formation by isolated plastids from the same tissues. Higher levels of fumarase activity in mitochondria from 3-hour illuminated dark-grown oat seedlings than mitochondria from dark-grown, light-grown or 24-hour-illuminated laminae (Hampp, 1980) are in accordance with the overall respiratory changes previously observed. DURING LIGHT-GROWN DEVELOPMENT F. OXYGENCONSUMPTION

It is well known that germination of seeds is accompanied by a marked increase in mitochondrial activities (Cherry, 1963; Opik, 1965; Breidenbach et al., 1967a,b; Kolloffel, 1967; Nawa and Asahi, 1971; Solomos et al., 1972; Morohashi and Shimokoriyama, 1975). These early activities over the first 6 hours are due to increases in succinate or endogenous NADH oxidation within the cotyledons (Morahashi and Bewley, 1980). Malate and a-ketoglutarate oxidation increases only after 12 hours of inhibition suggesting that alternative electron transport or translocator pathways are functional over different periods of maturation. By contrast to studies of seed germination, investigations of respiration and mitochondrial behavior during early development in leaves grown in the light are

CHLOROPLAST DEVELOPMENT

155

rare. Primary leaves of Phaseolus show an exponential increase in the rate of respiration between 50 and 110 hours after soaking the seeds if expressed on a per leaf basis (WiesCkowski, 1969). If however the same results are expressed on per square centimeter of leaf tissue, the rate of respiration rises to a maximum value at 70 hours and then declines slowly therafter. Respiration rates in peas also expressed on a per leaf basis achieved a maximum 9 days after germination although levels of various mitochondria1 enzymes were already declining (Smillie, 1962). In older tissues the general experience is that respiration declines with increasing leaf age (Kidd er af., 1921). An unusual situation prevails in rice seedlings, germinated and grown in the light, if the water covering them is stagnant and becomes anaerobic (Kordan, 1976). Etiolation and poor root growth are induced. These may be reversed if oxygen is added into the surrounding water, a clear demonstration of the requirement of respiration for the greening process. G. DARKRESPIRATION AFTER MATURITY There has been much experimentation to determine the relative importance of “dark respiration” in mature green plant tissues in the light and the dark. This particular topic has been reviewed recently (Moore, 1982) and is only covered in outline here. In the past, difficulties and conflicting results have stemmed from problems of accurate measurement of dark respiration in the presence of photosynthesis and photorespiration. Glycolysis in algae and higher plants is immediately inhibited by illumination at the triose phosphate dehydrogenase step, probably by increases in the ATP/ ADP ratios (Kandler and Haberer-Lisenkrotter, 1963; Santarius and Heber, 1965; Mangat et af.,1974). At the same time, levels of oxidized NAD+ decline while those of NADP+ rise (Graham and Walker, 1962; Heber and Santarius, 1965; Olgren and Krogmann, 1965; Graham and Cooper, 1967). These changes in the ATP/ADP and reduced NAD(P)H/oxidized NAD(P)+ ratios are most likely initiated in the plastid and transmitted to the cytosol and/or the mitochondria by translocator-mediated shuttle mechanisms such as those proposed by Heber (1974) and Walker (1976). Label from I4CO, incorporated into photosynthetic products does not rapidly pass into the tricarboxylic cycle in the light but does so when illumination ceases. It has been suggested that the tricarboxylic acid cycle is inhibited in the dark. However, interconversion of tricarboxylic acid cycle intermediates does occur (Graham and Walker, 1962; Chapman and Graham, 1974b) which may mean that, while photosynthetically-derived 14C is diverted to sugar synthesis, the tricarboxylic acid cycle is maintained by endogenous unlabeled reserves (Graham, 1978). Using inhibitors such as malonate and fluoroacetate, Chapman and Graham ( 1974a) demonstrated that the tricarboxylic acid cycle functions in

156

A.

R. WELLBURN

the light at a rate comparable with that in the dark except for a brief initial inhibition on transition from darkness to light. Other inhibitor experiments using DCMU have shown that apparent photosynthesis may be strongly inhibited without effect upon dark respiration (Chapman and Graham, 1974a; Gnanam er al., 1974; Chevallier and Douce, 1976). This contrasts with the behavior of most inhibitors which interfere simultaneously with both processes (Gnanam et al., 1974), thereby invalidating any useful interpretation.

VI. Transfer between Cell Compartments during Photomorphogenesis A. THE PLASTIDENVELOPES DURING GREENING The importance of the envelopes during plastid biogenesis has already been stressed in a comprehensive review (Douce and Joyard, 1978). The polypeptide composition of the plastid envelope membranes differs obviously from that of the mitochondria1 membranes, the endoplasmic reticulum, and the internal thylakoids. By comparison to the latter, envelopes contain a higher preponderance of high-molecular-weight polypeptides, including several above 100,000 (Cobb and Wellburn, 1974; Pineau and Douce, 1974; Joy and Ellis, 1975; Sprey and Laetsch, 1975). Amounts, numbers, and sizes of thylakoid polypeptides fluctuate during greening (Lursson, 1971; Cobb and Wellburn, 1973) but a study of the major envelope polypeptides of etioplasts, 0-24 hour etiochloroplasts, and mature chloroplasts revealed a remarkable constancy in the overall pattern (Cobb and Wellburn, 1974); only the amounts of each increased in line with the increase in plastid volume as etiolated tissue is greened. When the lipid composition of envelopes from etioplasts, etiochloroplasts, and chloroplasts was determined, a similar constancy of distribution throughout greening emerged (Bahl et al., 1976). However, pulse-labeling studies of envelope lipids revealed that small but rapidly turned-over pools exist (Joyard et al., 1980). The plastid envelopes (probably mainly the inner) are also the site of considerable biosynthetic capability, particularly with respect to galactose metabolism (Douce, 1974; Van Besouw and Wintermans, 1978), acyl-CoA synthesis and transfer (Bertram and Heinz, 1976; Joyard and Douce, 1977; Roughan and Slack, 1977; Heinz er al., 1978; Joyard and Stumpf, 1980, 1981), ring methylation during isoprenoid biosynthesis (Sol1 et al., 1980a,b), as well as monoterpene (Carde et al., 1980) and carotenoid biosynthesis (Costes et al., 1979). All these activities have been identified using envelopes from mature chloroplasts. It is possible even more will be revealed in the future if studies are made on envelopes from developing (greening and light-grown) plastids which are in their most active biosynthetic phase of morphogenesis. Although promising indications have now been reported by Keegstra, the unequivocal demonstra-

CHLOROPLAST DEVELOPMENT

157

tion of a successful technique to separate inner plastid envelopes from outer envelopes is awaited. This will enable the accurate location of these capabilities and define the nature of the relationship of the inner envelope to the internal membranes. In mature chloroplasts there appears to be no connection between the intermembrane space of the envelopes and the thylakoid lumen, which account in part for the ability of chloroplasts to generate considerable pH gradients between the stroma and the thylakoid lumen. In greening and developing light-grown plastids, associations between the inner envelope and the internal membranes are often observed. It has been suggested that internal membranes are formed by continued invagination of the plastid envelope (Muhlethaler and Frey-Wyssling, 1959; Menke, 1960; Kirk and Tilney-Bassett, 1967) but it is more probable that the invaginations occur as discrete vesicles which then reassociate with the internal membranes. Electron micrographic evidence for this type of process during development is widespread (e.g., Douce and Joyard, 1978; Wellburn and Wellburn, 1979). Figure lc-e all show evidence of this process. Douce and Joyard ( 1978) believe that these lightly staining “in-pushings” (loco cizazo) represent rudimentary lamellar vesicles rich in galactolipids which are subsequently modified by additions once fully inside the stroma. Freeze-fracture studies of envelopes from mature plastids have revealed that particle sizes and distribution are quite different to those of thylakoid membranes (Sprey and Laetsch, 1978; Lefort-Tran er al., 1980). It would be useful to have similar comparative studies done on developing plastids to establish the nature of these temporary “in-pushings.” The recent report by Newcomb and co-workers of distinct villilike structures projecting from the inner envelope toward the stroma may suggest that the origins of the in-pushings may be more specific in location and have some relationship to the PTBs of developing light-grown plastids, as has been mentioned earlier. Another unexplained phenomenon of plastid envelopes are the close contact or fused areas between outer and inner plastid envelopes (Douce and Joyard, 1979; Lefort-Tran et al., 1980). These may account for the difficulties of detaching inner from outer plastid envelopes although mitochondria1 envelopes also have similar contact areas (Hackenbrock and Miller, 1975) and yet they are readily separable. Lefort-Tran el al., ( 1980) have shown, using freeze-fracture techniques, specific fracture face areas in Euglena envelopes which may correspond with these fused areas. At present there is no evidence to confirm a role for these areas but it has tempted Douce and Joyard (1978) to propose that they may be one of the mechanisms for phospholipid transfer. Alternatively, they may be implicated in the transport of proteins across the envelope membranes using signal peptides to recognize only these particular fused regions of the plastid envelopes and thereby allow nascent precursor proteins made on cytoribosomes to be guided directly through both envelopes into the stroma before a signalase or

158

A. R . WELLBURN

processing protease detaches the signal peptide. Such a mechanism would be an easier route to envisage for polypeptides destined for the plastid stroma rather than a two-stage crossing of discrete envelopes. It has been argued recently (Douce and Joyard, 1981) that there is little evidence to conclude that the endoplasmic reticulum (ER) and the outer plastid envelope are in continuum or that they have similar characteristics and have a common origin from the ER as has been suggested by Morrk (1975). They support their argument with numerous differences in composition (Douce and Joyard, 1979) and point out that the possibility of either a “passive corridor” (loco citato) of ER leading the polypeptide products of rough ER directly to the interenvelope space or that reverse pinocytotic mechanisms involving fusion of ER vesicles with the outer envelope have little structural evidence to support them. Most nuclear-coded plastid proteins are synthesized on free cytoribosomes anyway. Furthermore, Chua and Schmidt ( 1979) have demonstrated that protein synthesis and transport are independent events and that the latter is an active process across both envelopes. Clearly more definitive work which relates the structural to the synthetic aspects of this problem is required to resolve these matters and it may well be that a developing rather than a mature system may be a more suitable model to use in order to accomplish this.

B. EVIDENCE FOR CHANGES IN FLUXESBETWEEN COMPARTMENTS The first experiments to study differences in uptake of metabolites during greening were part of a larger study of terpenoid biosynthesis. These established the concept of compartmentalization of enzymes of isoprenoid synthesis both inside and outside plastids and the relative impermeability of mature plastid envelopes to mevalonic acid (MVA: Goodwin and Mercer, 1963; Goodwin, 1965; Threlfall and Griffiths, 1967; Whistance and Threlfall, 1967, 1968; Threlfall et al., 1968). However, in these studies, the period of exposure of etiolated tissue to [2-I4C]MVA, 14C02, and light were usually longer than 24 hours (Threlfall er al., 1967), sufficient time for more or less mature chloroplasts with reasonable C0,-fixation capabilities to form (Hampp and Wellburn, 1976~).An electron microscope autoradiographic study using the uptake of [2-3H]MVA of high specific activity revealed that after 1 hour of illumination of oat seedlings there was a preferential accumulation of label into the plastids but this was much reduced over long periods of illumination (Cockburn and Wellburn, 1974) and, in accordance with the previous observations of Goodwin and co-workers, there was no significant uptake of MVA by plastids in mature lightgrown seedlings. Recently a comparison has been made of the relative abilities of different potential precursors of isoprenoid and fatty acid biosynthesis to cross the chloroplast envelopes (Grumbach and Forn, 1980). Not only were CO, and acetate readily taken up but MVA was also incorporated into isoprenoids, imply-

CHLOROPLAST DEVELOPMENT

159

ing that the complete isoprenoid biosynthetic sequence exists from CO, through acetyl-CoA synthesis to mevalonate and beyond within chloroplasts. It appears that plastids can synthesize acetyl-CoA (Stumpf et al., 1967; Shah et al., 1969) and metabolize mevalonate further by phosphorylation (Rogers et al., 1966; Hill and Rogers, 1974; Cooke, 1977) although there has been some doubt about the latter (Gray and Kekwick, 1973). The presence of P-keto-thiolase, hydroxymethyl-glutaryl-CoA-synthetaseand hydroxy-methyl-glutaryl-CoA-reductaseis assumed and has yet to be verified. Considerable enhancement of the rate of amino acid conversion into protein occurs when etiolated plants are illuminated (Margulies and Parenti, 1968; Drumm and Margulies, 1970). The intracellular distribution of 3H-labeled tyrosine and leucine was also determined by electron microscope autoradiography over a range of illumination periods. The percentage of both labels in plastids, as compared to the remainder of the cell, varied with the period of illumination and from each other (Cockburn and Wellburn, 1974), with high concentrations during short illuminations and also after 8 hours of illumination. The variations in uptake of MVA and amino acids were ascribed to selective changes in the permeability of the plastid envelopes during the developmental process which differed according to the metabolite involved. Pulse-chase experiments with [35S]methionine for 1 hour before illumination of oat seedlings followed by cold methionine during greening showed that rapid protein interchanges occurred even in unilluminated tissue which increased markedly upon illumination (Cobb and Wellburn, 1976). The bulk of the label was progressively associated with the thylakoid membranes as greening progressed but a peak of envelope-associated label occurred after a 2 hour chase in the light. The radioactive profile of these envelope polypetides on SDS gels revealed that the distribution of label was not coincident with the major staining bands of the envelopes. This is explained by observations such as those of Dobberstein et a1 (1977) using Chlamydomonas which have been confirmed by Ellis and co-workers with peas (Ellis et al., 1977; Highfield and Ellis, 1978; Ellis and Barraclough, 1978). Cell-free translation systems synthesize a 200,000 MW popypeptide which is immunoprecipitated by antibodies raised against the small subunit of ribulose- 1,5 bisphosphate carboxylase which is of low molecular weight (13.5 to 16.5K). The additional polypeptide portion apparently serves as a signal protein to assist with recognition and transport across the envelope. A similar signal protein is attached to the chlorophyll a/b binding protein during synthesis outside the plastid which is also detached once inside (Ape1 and Kloppstech, 1978). Hampp and Schmidt (1976) first applied the technique of silicone oil centrifugal filtration, developed by Klingenberg and Pfaff (1967), to etioplasts and older (24-72 hour) etiochloroplasts. They showed that a change in envelope permeability to glutamate, a-keto-glutarate, succinate, citrate, malate, and glycine

160

A . R . WELLBURN

occurred during greening. Further investigation by Hampp and Wellburn (1976a) showed that the rates of uptake of such metabolites into plastids over the first 8 hours varied considerably. Oxaloacetate and succinate exhibited pronounced increases in rates of uptake after only 1 to 2 hours of greening. Others such as malate, glycine, and glutamate were progressively less permeable as development progressed while another group (alanine, aspartate, and 3-PGA) showed an initial decline in uptake rate after 1 hour but ready uptake over 2 to 4 hours of greening. Similar studies were employed to study the rates of uptake of acetate, mevalonate, and different regulator substances during greening of oat seedlings (Wellburn and Hampp, 1976a,b) and also over chloroplast-chromoplast transformations in ripening peppers (Schneider et af., 1977). Mitochondria isolated from greening etiolated oat seedlings also showed changes in the rates of uptake of metabolites using the silicone-oil centrifugal filtration technique. Oxaloacetate, a-keto-glutarate, and especially succinate most readily entered those mitochondria isolated from 2 hour-illuminated tissue while glutamate and glycine showed enhanced rates of uptake into mitochondria from tissue given longer (4-8 hours) periods of illumination (Hampp and Wellbum, 1976b). By contrast 6-amino-leavulinic acid did not readily penetrate mitochondria isolated from greening tissue. These observations indicated for the first time that not only are there synchronized changes in the flux of metabolites across plastid envelopes during greening but that there are also synchronized changes across mitochondrial membranes while plastid morphogenesis is in progress. This was substantiated by employing one of the additional possibilities of silicone-oil centrifugal filtration. It is possible to adjust the density of the intervening silicone-oil so that, for example, intact plastids pass through the oil but mitochondria remain above. By preloading mitochondria (isolated from laminae given different periods of illumination) with labeled succinate and incubating them with unlabeled plastids from the same source in an unlabeled medium, the ready transfer of label from mitochondria to plastid was demonstrated (Wellburn and Hampp, 1976c) by those organelles which had previously shown high rates of uptake of succinate (i.e., both mitochondria and plastids from 2-hour illuminated oat seedlings). Similar but earlier synchronized changes were also shown using oxaloacetate.

C. CHANGES IN TRANSLOCATORS DURING GREENING 1. The Phosphate Translocators

The properties of the plant mitochondrial phosphate translocator have been described elsewhere (Heldt, 1976b). Phosphate can enter mitochondria by different transport mechanisms which may be distinguished using inhibitors. It may be taken up either in exchange for OH- ions or cotransported with protons, phe-

CHLOROPLAST DEVELOPMENT

161

nomena which cannot be distinguished experimentally (Papa et a/., 1969; Chapell and Crofts, 1966; McGivan and Klingenberg, 1971; Meijer et a / . , 1970). Alternatively it may be exchanged for dicarboxylate anions. The latter transport is specifically inhibited by substrate analogs such as n-butyl malonate (Robinson and Williams, 1970; Palmieri et a / . , 1971) but both processes are inhibited by sulfhydryl-blocking reagents such as mersalyl (Meijer et a/., 1970). It has been assumed that the orthophosphate/OH - exchange served to neutralize the charge imbalance occurring from the uptake of ADP3- and the release of A T P - during oxidative phosphorylation (McGivan et a / . , 1971) while the orthophosphate/dicarboxylate exchange allows a net uptake of anionic substances at the expense of the mitochondria1 pH gradient (Klingenberg, 1970b; McGivan and Klingenberg, 1971). Consequently, transport of anions into mitochondria is dependent on a respiration-linked proton efflux, coupled to an orthophosphate/OH - exchange (phosphate translocator), followed by substrate/ orthophosphate exchange using the dicarboxylate translocators. Transport via the phosphate translocator of animal mitochondria shows very high rates of orthophosphate movement which are not rate-limiting for ATP formation. On the other hand, Day and Hanson (1977) showed for plant mitochondria that the rate of orthophosphate transport (reduced by inhibitors and uncouplers) is closely related to the amount of the substrate oxidation. Rapid oxidation of malate and succinate required high transport activities of both the phosphate and the dicarboxylate translocators of plant mitochondria. Using the technique of silicone-oil filtering centrifugation of organelles and the inhibitor-stop method, the kinetics of transport of inorganic phosphate across the inner mitochondrial membrane were tested in relation to different stages of greening (0 to 24 hours) of etiolated laminae of Avena by Hampp (1979). There was an increase in phosphate transport after 3 hours of greening, reaching values of V,,, (about 17 pmol mg protein - hr - I ) that are three times as high as those measured with mitochondria from etiolated tissue. After 24 hours of light the rates of phosphate transport decreased again to levels below those in mitochondria from etiolated tissues. Hampp (1979) also found that there was no change in affinity between inorganic phosphate and the binding sites of the transporting systems involved. The K, remained constant throughout the development study but that 10 times the concentrations of mersalyl were required at all stages of development to produce similar inhibition to that shown by animal mitochondria. Transport by the phosphate translocator of plastids is highly specific and is affected by the pH of the stroma (see Heldt, 1976a). It requires either orthophosphate or phosphate esterified to a three carbon molecule such as 3-PGA or DHAP. Neither pyrophosphate nor 2-PGA are alternatives. The mechanism is by a very strict counterexchange (Heldt et a/., 1975) and may be inhibited by chloromercuriphenylsulfonic acid. The rates of orthophosphate uptake by etioplasts and etiochloroplasts from

162

A. R. WELLBURN

dark-grown oats given different lengths of illumination (3-72 hours) have also been determined (Hampp, 1978). The rates of uptake were found to be high at all stages but with maximum values given by 3 hour etiochloroplasts. Concentration-dependence studies (Hampp, 1978) revealed hyperbolic saturation characteristics, just as in mitochondria from different developmental stages (Hampp, 1979), demonstrating substrate saturation of transport at all stages of plastid greening together with rises in the levels of phosphate in the stromal spaces especially at low concentrations. These orthophosphate accumulationswere most pronounced in 3 hour etiochloroplasts, less in etioplasts, and smallest in older etiochloroplasts (Hampp, 1978). Velocities of orthophosphate transport in plastids also increased between 0 and 3 hours of illumination but decreased strongly thereafter. This was not due to changes in affinity for orthophosphate (K,) during development. Furthermore, inhibition of orthophosphate transport by 3PGA was only shown after 3 hours of illumination (Hampp, 1978) which suggests that the early phosphate transport system is slightly different to that of the phosphate translocator of more mature plastids. 2. The Carboxylate Translocators Three carboxylate carriers catalyzing the counterexchange of anions are present in mitochondria (see Heldt, 1976b). The principal dicarboxylate translocator of plant mitochondria transports either orthophosphate or dicarboxylates such as malate, succinate, oxaloacetate, and even malonate. The subsiduary tricarboxylate carrier exchanges citrate, isocitrate, aconitate, or phosphoenolpyruvate for dicarboxylates and a third distinct a-ketoglutarate translocator exists which has a binding site for a-ketoglutarate or malate but which will also react with aspartate, glutamate, and oxaloacetate. All three are inhibited by butyl-malonate, phenyl-succinate, pentyl-malonate, and bathophenanthrolin (Quagliariello and Palmieri, 1972), all of which have no effect on the phosphate carrier. The behavior of these three carriers in mitochondria during plastid development has not been extensively studied. Strongly increased (10-fold) rates of uptake of succinate are observed in mitochondria from 2-hour-greened etiolated oat seedlings (Hampp and Wellburn, 1976b) by comparison to mitochondria from seedlings left in the dark or illuminated for much longer periods. Meanwhile, changes in the rates of uptake of oxaloacetate or a-keto-glutarate during greening were less spectacular. The plastid dicarboxylate translocator differs from that in mitochondria by its specificity only for C, and C, dicarboxylates (including a-keto-glutarate). Moreover, malonate is not transported nor is there a direct connection with the transport of phosphate or tricarboxylates (Heldt, 1976a). Hampp (1978) demonstrated that the affinity of etioplasts for malate was low but, when plastids were isolated from dark-grown oat seedlings given different light treatments, the K , values for malate were relatively constant throughout greening. Velocities of malate trans-

CHLOROPLAST DEVELOPMENT

163

port, however, were highest in etioplasts and declined thereafter. At all stages of development there was competition for the plastid dicarboxylate translocator between succinate, malate, and oxaloacetate. In accordance with higher rates of uptake of succinate (and oxaloacetate) reported earlier (Hampp and Wellburn, 1976a), competition between succinate and malate was highest after 3 hours of illumination. 3. The Adenine Nucleotide Translocators The properties of the plant mitochondria1 adenine nucleotide translocator have been reviewed by Heldt (1976b). It has many characteristics in common with the much studied animal counterpart (see Pfaff and Klingenberg, 1968). However, there is a difference in inhibitor response between animal and plant mitochondria. Adenine nucleotide transport in vertebrate mitochondria is specifically inhibited by atractyloside (Chappel and Crofts, 1965) and carboxyatractyloside (Pfaff et al., 1969; Vignais et a l . , 1973; Luciani, 1975). Similar effects have also been noted in yeast (Onishi et al., 1967). Plant mitochondria on the other hand are inhibited by very much higher concentrations of either compound (Jung and Hanson, 1973; Earnshaw and Hughes, 1976; Hampp and Wellburn, 1980) and mitochondria from Helianthus tuberosus tubers are claimed to be completely insensitive to atractyloside (Passam et al., 1973; Passam and Coleman, 1975). Estimates of the activity of the plastid adenine nucleotide translocator amount to no more than 5% of the other major plastid translocators (Heldt, 1969; Strotmann and Heldt, 1969; Heber and Santarius, 1970; Walker, 1976). As a consequence this carrier is not thought to be a major part of the mechanism of energy transfer from the plastid to the cytoplasm in the light (Heldt et al., 1971). Nevertheless it may have greater importance for transfer between cytoplasm and plastid for the mature plant in the dark at night (Heldt, 1969) or in low light conditions (Tarchevsky and Konjukhova, 1981). The chloroplast envelope also contains two other activities, the significance of which have yet to be fully established. An envelope-bound magnesium- or manganese-dependent nonlatent ATPase exists (Sabnis et al., 1970; Joy and Douce, 1974) which is lost when chloroplasts lose their envelopes. In addition, Murakami and Strotmann ( 1978) have demonstrated that adenylate kinase (ATP:AMP phosphotransferase), working both forward (2 ADP to ATP plus AMP) and backward, is associated with chloroplast envelopes. They may be alternative systems to control adenylate compartmentation between plastid and cytoplasm. There has been only one study of the relative importance of both adenine nucleotide translocators during greening. Studies using silicone-oil filtering centrifugation of isolated mitochondria and plastids from dark-grown and 1-48 hour illuminated etiolated seedlings showed an increase in the velocities of transport

164

A. R. WELLBURN

of ADP and ATP for both plastids and mitochondria after only a short period (1-3 hours) of illumination of Avena seedlings, lasting for slightly longer in plastids (Hampp and Wellburn, 1980); precisely the period during which pool sizes of ATP change most violently (Hampp, 1980) and while photophosphorylation develops (Wellburn and Hampp, 1979). It may well be that the plastid adenine nucleotide translocator is more important at this stage for energy transfer than at any other time thereafter. Substrate saturation with respect to ATP and ADP has been shown by both organelles at all stages of development but the mitochondrial translocator differs from that of plastids by exhibiting a different affinity to ADP as compared to ATP (Hampp and Wellburn, 1980). The rate of adenine exchange for both changes over the greening period. Rates at 0 to 24 hours were found to be lower than those after 3 hours of illumination, further evidence for a synchronized and coordinated change in translocators of both organelles to facilitate the transfer of energy to plastids from mitochondria during greening. D. NUCLEOTIDE POOLSIZESDURING DEVELOPMENT Attempts to define the energetic state of biological systems have employed several different parameters. One of the most popular is the adenylate energy charge (EC) ratio, which is defined as the ratio of “energy-rich” phosphate bonds in adenylic compounds (two in ATP and one in ADP) to the total amount of adenylates (Atkinson, 1968, 197 I ) . Another parameter is the phosphorylation potential (Kpho)whichis the free energy of hydrolysis of ATP. This takes into account the intracellular concentration of orthophosphate (Kraayenhof, 1969), so if this then alters, the physiological significance of the two parameters will differ. The dependence of respiratory activity on the ATP/ADP ratio has been recognized for some time (Klingenberg, 1970a,b; Lowry et al., 1971; Slater et al., 1973; David and Lumeng, 1975; Kuster et al., 1976) but oxidative phosphorylation is also dependent upon the levels of orthophosphate (Holian et al., 1977) and hence the use of the KPhoratio is probably more appropriate. A more comprehensive model to account for the regulation of mitochondrial oxidative phosphorylation which combines KPho with the NAD +/NADH ratio and the cellular respiratory rate has been proposed [Eq. (I)] by Erecinska and Wilson (1978). “AD+] [cyt. c2+12 [ATP12

‘Ox.

Phos. - [NADH] [cyt. c3+]* [ADPI* [PJ2

(1)

In algae and higher plants, endogenous levels of ATP, ADP, and orthophosphate are known to change considerably in dark to light and light to dark transitions (Kandler and Haberer-Lisenkrotter, 1963; Santarius and Heber, 1965; Pedersen et a l . , 1966; Bomsel and Pradet, 1967; Strotmann and Heldt, 1969; Lewenstein and Bachofen, 1972; Krause, 1971; Larsson e t a l . , 1978). Over short

165

CHLOROPLAST DEVELOPMENT

periods the increase in ATP in the light is at the expense of ADP thereby increasing the EC ratios. In the longer term, often after considerable oscillation, the original ratios are reestablished under the new conditions. Changes in orthophosphate, triose phosphates, and the redox state of NAD(P) /NAD(P)H may differ and as a consequence changes in KPhoor more comprehensive parameters [e.g., the KPhotratio of Eq. (2)] may more accurately reflect the homeostatic readjustments taking place in the cytoplasmic and plastidic compartments of green plants. +

KPhot.

=

[PGAI [ATP] [H +]“ADPHI [DHAP] [ADP] [P,][NADP ‘1

(2)

The considerable technical difficulties associated with the isolation of intact mitochondria and plastids from green algae have confined studies of cellular concentrations of metabolites and energy and their fluxes between the three compartments (plastid, mitochondria, and “cytosol”) to higher plants. These have been reviewed elsewhere (Heber, 1974; Heldt, 1976a,b; Strotmann and Murakami, 1976; Walker, 1976). Much is still to be learned about the various complex interrelationships that exist between the three different compartments in mature plants growing in the light. Not surprisingly therefore very little is known about the changes in these concentrations and fluxes in developing plants or even in mature plants during the night or at very low light levels. In intact greening tissues, EC ratios remain constant throughout the developmental process (Friederich and Mohr, 1965; Hampp and Wellburn, 1980). This is in spite of the various biosynthetic processes and changes in bioenergetic dependency that take place during greening. Phosphorylation potentials may be calculated during greening. This has revealed high potentials in etioplasts and etiolated tissues with a progressive fall as plastids and tissues mature (Hampp and Riehl, 1981). Recently an interesting development has taken place which promises considerable advance not only in understanding developmental changes but also the nature of the balances between cell compartments in mature plants in the light. By taking advantage of improved methods of protoplast isolation (Kanai and Edwards, 1973; Edwards et al., 1978; Hampp and Ziegler, 1980) and the technique of protoplast fragmentation using nylon mesh at the top of a silicone-oil centrifugation tube (Robinson and Walker, 1979), Hampp (1980) has devised a combined two-stage silicone-oil fractionation below the protoplast fragmentation in the same tube. As a consequence it is possible to rapidly separate plastids, mitochondria, and cytosol within seconds. By monitoring marker enzymes characteristic of each compartment, the high degree of purity of the different fractions obtained by this method has been verified (Hampp, 1980; Wirtz et al., 1980). The most interesting observation so far from these studies is that the ATP concentration in the cytosol fraction, expressed either per protoplast number or in

166

A. R. WELLBURN

terms of protein, is higher than levels either in plastids or mitochondria when the protoplasts are derived from either etiolated or light-grown or partially illuminated (2 or 24 hour) dark-grown Avena tissue (Hampp, 1980). This is despite the fact that the cytosol must include not only cytoplasmic but also vacuolar components. Consequently the true cytoplasmic ATP concentration may be even greater. It is especially high in the cytosol from protoplasts prepared from 2 hour illuminated etiolated tissue. The levels of ATP associated with mitochondria decrease during development if expressed on a per protoplast basis but remain constant if a protein basis is employed. Similar ATP levels in plastids were generally lower than mitochondria even in protoplasts from light-grown Avena but interestingly the highest concentrations of ATP were in the 2 hour illuminated etiochloroplasts (Hampp, 1980) where no significant photophosphorylation occurs but an extensive biosynthetic requirement exists (Wellburn and Hampp, 1979). Clearly application of the simultaneous triple separation technique of Hampp (1980) will yield much more information in the future. The determination of the other nucleotides will allow not only the calculation of the phosphorylation potential but other more useful parameters (e.g., KPhot)which will more accurately reflect the coordinated changes taking place between the compartments during plastid development.

E. AN OVERVIEW OF CHANGES IN FLUX With so many changes taking place between the mitochondria, developing plastids, and the cytoplasm, it is difficult to appreciate the overall pattern or, at least, to glean some understanding of the general principles involved. Figure 5A to D is an attempt to clarify this problem for the development of both greening and light-grown mature chloroplasts. Conceptually it is convenient to define four different stages (A to D) which cover the various stages of development. For greening monocotyledonous systems, A would refer to the etioplast, B and C would represent early (0.1 to 1.5 hour) and medium (1.5 to 4 hour) term etiochloroplasts, while stage D would cover older (4 hours and beyond) etiochloroplasts and chloroplasts. In cereals most of the changes encompassed within these schemes have been established by observation (see earlier). For greening dicotyledonous systems where no equivalent studies have been carried out the timings most likely require lengthening to take into account the slower rates of greening. Similarly it is difficult to relate precisely the different stages of light-grown plastid development to this scheme. It is suggested that Stage A would be analogous to the eoplast phase and B the amyloplast phase, while C would be equivalent to the amoeboid plus the early immature plastid phases, and D the later immature as well as mature chloroplast phases. Each diagram illustrates the 3 compartments involved (left to right:

CHLOROPLAST DEVELOPMENT

167

mitochondria, cytoplasm and plastid) with the inner and outer envelopes of each organelle shown by solid and dashed lined respectively. The main translocators of each inner envelope are symbolised by rings and the direction of fluxes by arrows. The increasing solidity and sizes of the arrows are intended to represent the relative natures of the fluxes involved. The same abbreviations of the phosphate and adenine nucleotide translocators (PT and AT respectively) are used for both organelles but CT (for carboxylate translocator) of the mitochondrion is used instead of DT (for dicarboxylate translocator of plastids) to signify the triple nature of these translocators in mitochondria as well as their different behaviour especially with respect to orthophosphate movement. In mitochondria, TCAC represents the tricarboxylic acid cycle and CO, with an arrow in plastids signifies CO, fixation. The sizes of the boxes of the C3.5 pools (sugars, organic acids and sugar phosphates) delineated by alternate dots and dashes are not meant to have any pool size connotations. 1, Stage A (Etioplasts and Eoplasts)

As shown in Fig. 5A, the first stage (A) is probably characterized by a strong dependency upon mitochondrial activity for which the mitochondrial phosphate

FIG. 5 . (A-D) Schematic representations of early changes in flux across the mitochondrial (left) and plastid (right) envelopes during chloroplast development. Diagram A represents the etioplast or eoplast phase while B portrays the conditions in the early etiochloroplast or amyloplast stages. Diagram C represents the intermediate etiochloroplast or amoeboid and early immature plastid phases while D exhibits the conditions in late etiochloroplasts or late immature and mature chloroplasts. For a fuller description refer to the text.

168

A.

R. WELLBURN

and, possibly, carboxylate translocators are required to maintain respiratory activity. This gives rise to ATP which is made available across the cytoplasm and to the plastid by means of the adenine nucleotide translocators of both organelles. Respiratory chain phosphorylation must be maintained by a pool of organic acids which are replenished from other sources. In the case of eoplasts and young etiolated tissue this probably means mobilized seed reserves but for older etiolated tissue of species with small seeds and little plastidic starch this must be at the expense of tissue reserves elsewhere such as the steroidal saponins.

2 . Stage B (Early Etiochloroplasts and Amyloplasts) Events thought to take place during stage B are illustrated in Fig. 5B. Further increases in respiratory activity are reflected in increased rates of transport by all mitochondrial translocators to allow more orthophosphate, ADP, and organic acids to enter and ATP to leave. This is then made available to the plastids by increased plastidic adenine nucleotide transport. At the same time enhanced rates of orthophosphate flux and possible sugar phosphate uptake from the cytoplasmic pool are achieved by increases in activity in the plastid phosphate translocator. Under certain conditions, such as in amyloplasts, this may be of such a rate as to allow significant accumulation of starch. Possibly the plastid dicarboxylate translocator also assumes some significance to allow the import of certain components (e.g., oxaloacetate and a-keto-glutarate) for biosynthetic purposes. Rates of supplementation of the cytoplasmic C3-5 pool by extracellular import must be increased to compensate for all these demands. In the case of the amyloplast this may be either from seed reserves or, alternatively, from photosynthate made available from older mature tissues elsewhere in the plant. 3 . Stage C (Intermediate Etiochloroplasts and Amoeboid Plastids) A reduction in respiration follows a decrease in the rates of exchange of orthophosphate, organic acids, and adenine nucleotides across the mitochondrial inner envelope (Fig. 5C). At the same time the breakdown of accumulated reserves, such as starch in the case of light-grown systems, releases potential metabolites and energy which are supplemented by the advent of photophosphorylation (see later) within the plastids thereby reducing the dependance upon imported ATP and orthophosphate. As a consequence the rates of exchange carried out by both phosphate and adenine nucleotide translocators decline. Nevertheless biosynthetic demands for short chain organics within the plastid are likely to be at their greatest and for this purpose the dicarboxylate translocator shows especially high rates of exchange, relying upon the C3-5 pool in the cytoplasm for certain precursors. Fortunately the mitochondrial demands upon this pool are lower reducing the necessity for additional extracellular import. Furthermore the mitochondrial carboxylate translocators may be responsible for the transport of certain key metabolites (e.g., oxaloacetate) out of the mitochondria to supplement the pool upon which the plastid draws.

CHLOROPLAST DEVELOPMENT

169

4. Stage D (Late Etiochloroplasts and Chloroplasts) The fluxes between chloroplasts and their surroundings shown in Fig. 5D are more familiar. The appearance of the ability to fix CO, using the Calvin-Benson-Bassham cycle, assisted by adequate supplies of photoreductant and ATP, causes an abrupt change in behavior of the plastid with respect to its surroundings. This change thought to be transmitted by reversals of translocator exchange as the organelle becomes a net exporter not just to the rest of the cell but allowing extracellular export as well. Biosynthetic demands within mature chloroplasts are probably much reduced and the dicarboxylate translocator reverts to its more normal function of allowing the dicarboxylate shuttle to export photoreductant. With ample photophosphorylation and the triose phosphate shuttle operating, the importance of the plastid adenine translocator disappears. By contrast mitochondria1 activities decline to a lower level with reduced rates of respiratory chain phosphorylation and exchange of adenine nucleotides, orthophosphate, and organic acids.

VII. Biogenesis of Photochemical Activities The formation of pigments, lipids, and the photochemical systems during chloroplast morphogenesis was the subject of an earlier discussion in this series 4 years ago (Treffrey, 1978). The intention here is to add only that additional information that has since appeared or was not covered then which directly relates to the bioenergetic aspirations of this article. A. ATP FORMATION IN DEVELOPING PLASTIDS The time of onset of photophosphorylation varies between species, upon the environmental grown conditions and also according to the mode of assay to elicit and assay the formation of ATP. Gyldenholm and Whatley (1968) pointed out the necessity of attaining a minimum structural complexity before photophosphorylation is possible. In greening Phaseolus plastids they found that phenazine methosulfate (PMS)-dependent light-driven ATP formation started between 5 and 10 hours after illumination of etiolated tissue while other methods of assay required 15 hours or longer. Greening barley plastids behaved in a similar manner (Phung Nhu Hung et al., 1970a; Remy et al., 1972) but wheat (Shen and Hung, 1964) and Y-1 Chlumydomonas (Eytan and Ohad, 1972) showed PMS-dependent light-driven ATP formation after only 1 to 2 hours. In contrast, other workers have shown that in beans (Oelze-Karow and Butler, 1971) and barley (PIesniEar and Bendall, 1971, 1973) this function starts immediately upon illumination. Usually in light-grown tissues rates of PMS-dependent light-driven ATP formation rise in young leaves and fall during aging (Smillie and Krotkov, 1959;

170

A. R. WELLBURN

Zaitlin and Jagendorf, 1960; Bourdu et al., 1965; Iordanov and Popov, 1967; Vecher et al., 1967; Iordanov, 1970). Sometimes a second maximum of activity occurs which coincides with the achievement of maximum leaf area (Heyes and Dale, 1971; Sestik et al., 1977). Although these results are useful in a comparative sense, the problem lies with the ability of PMS to generate artificially high rates of ATP formation. Witt et al. (1968) showed that reduced PMS is able to react directly with the electronaccepting site of PSI (P-700+) and can support cyclic electron flow around this site at extremely high rates with practically no light saturation. Other donors of electrons for PSI assay which carry protons may be used as alternatives to measure ATP formation. Reduced diaminodurene (DAD) and dichloroindophenol (DCIP) are highly suitable and require at least plastocyanin and possibly some of the cytochromes to close the cycle around PSI (see Trebst, 1974). Rates of ATP formation using these donors are much lower when they are used for assay during greening studies and they also show the appearance of ATP formation much later than previously indicated by PMS-driven assays. In the course of evaluating the photochemical properties of prothylakoid vesicles (PTVs) as compared to the PLBs proper and establishing that only the former were the sites of photochemical activity, a clear distinction was shown by PMS on the one hand and DAD- or DCIP-driven assays (Wellburn and Hampp, 1979). A time course of DAD-dependent light-driven ATP formation by plastids in relation to mitochondria1 oxidative phosphorylation is shown in Fig. 6 while Table I shows a summary of the times of appearance of the various functions in PTVs of different ages. These are slightly more advanced in time than the equivalent functions as determined in whole plastids. PMS-dependent lightdriven ATP formation appeared shortly after illumination (15 minutes) but for the DAD- or DCIP-driven equivalents to occur virtually all the other pho-

1

4

8 12 16 Hours of illumination

20

24

FIG. 6. Rates of ATP formation by isolated mitochondria and plastid thylakoid (a)preparations from different stages of greening (12 klux) of 8-day dark-grown oat seedlings (25°C). The procedures and assays were similar to those described by Hampp and Wellburn (1980) but with differing radiant light flux.

CHLOROPLAST DEVELOPMENT

171

TABLE I DEVELOPMENT OF PHOTOSYNTHETIC FUNCTIONS I N PROTHYLAKOID V E S I C L E SFROM ~.~ GREENING 8-DAY-OLDOAT SEEDINGS Time of appearance during development (hours) 0

0.25 0.5 1 2

3

4

8

Photosynthetic function‘ NADP-dependent oxidoreductase activity AscorbateIDAD + PS I + MV electron flow PMS-dependent photophosphorylation AscorbatemMPD 4 PS I + MV electron flow Ascorbate/DCIP + PS I + MV electron flow DPC -+ PS I1 -+ DMMIB electron flow NH; -induced uncoupling of ascorbatelDAD electron flow DAD-dependent proton pumping H 2 0 + PS I1 + DMMIB, DCIP, or femcyanide electron flow DCIP-dependent proton pumping DAD-dependent photophosphorylation HzO or DPC + PS I1 -+ PS I + MV or PNDA electron flow DCIP-dependent photophosphorylation Light-dependent C 0 2 fixationh Disappearance of DPC -+ PS I1 + DMMIB electron flow indicating completion of water-spitting capabilitya

aExcept unresolved inner membranes from 8-hour etiochloroplasts. bExcept intact 4-hour etiochloroplasts. CDAD, Diaminodurene; DBMIB, dibromothymoquinone; DCMU, 3-(3,4-dichlorophenyl)-l,1dimethylurea; DCIP, 2,6-dichlorophenolindophenol; DMMIB, 2,3-dimethyl-5,6-methyleneioxy-p-benzoquinone; DPC, 1,5-diphenylcarbazide; MV, rnethylviologen; PMS, phenazine methosulfate; PNDA, p-nitrosodimethylaniline;PS, photosystem; TMPD, N,N.N’,N’-tetramethylphenylenediamine.

tochemical functions have also to be present (2 hours). The early lack of response to uncouplers like NHZ ammonium ions indicates that an initial “leakiness” of protons only finally disappears after water-splitting is fully developed (Wellburn and Hampp, 1979). B. DEVELOPMENT OF THE PHOTOSYSTEMS

Using the oxygen electrode, activities associated with photosystem 2 (PS 11) are detected after a certain period of illumination which is usually longer than estimates of appearance of photosystem 1 (PS I) formation (Hiller and Boardman, 1971; Egneus et al., 1972; PlesniEar and Bendall, 1973; Boardman, 1977b). Nevertheless, when primary photochemical activities are measured spectrophotometrically by photooxidation of P-700 and cytochrome b-559 and the photoreduction of C-550 in intact leaves frozen to -196°C the reaction centers of both PS I and PS I1 appear within minutes of illumination with PS I1

172

A. R. WELLBURN

activities slightly ahead of those of PS I (Baker and Butler, 1976). This difference is taken to mean that a block exists in secondary electron transfer reactions perhaps on the water-splitting site of PS 11. Vernon and Shaw (1969) were the first to demonstrate that 1,5diphenylcarbazide (DPC) is only an electron donor for PS I1 when the water-splitting site is inactivated (e.g., by Tris). Accordingly, the transient appearance of rates of DPC-dependent activities compared to those involved with H,O splitting (Table I) would suggest that a step-bystep coupling of PS I1 activities and H,O-splitting occurs during greening. Starting from about 4 hours after the onset of illumination, PS 11-associated electron flow increases rapidly in prothylakoid vesicles (Wellburn and Hampp, 1979) due to a lag between pigment synthesis and the later formation of the H,O-splitting system. During this period PS I1 is readily accessible to DPC. With increasing time of development the rate of construction of the H,O-splitting site catches up with that of the pigment system. As shown in Table I, this process is finished after about 8 hours of illumination because of the inaccessibility of PS I1 to DPC. DPC as a donor for PS I1 was also used by Henningsen and Boardman ( 1973) for developmental studies, but these authors were not able to detect similar effects because of the use of Tris-washed organelles. It is not known if another developmental delay exists somewhere after the PS I1 reaction center and before plastoquinone.

C. APPEARANCE OF COUPLING DURING GREENING The degree of coupling of photophosphorylation to electron transport during plastid development has been measured by changes in the P/e$ ratios in greening beans (Howes and Stem, 1973) and barley (Phung Nhu Hung et al., 1970). It has also been determined in greening of intermittently illuminated wheat by comparing the ratios of electron transport during phosphorylation of ADP (State 3) to the rates after ADP utilization (State 4: Duysen et al., 1980). Tight coupling of electron flow with phosphorylation was not observed until quite late during chloroplast development in all cases. This delayed coupling is in accordance with the observations made using reduced DAD or DCIP as the donors to drive ATP formation (Wellburn and Hampp, 1979).

D. BIOSYNTHETIC REQUIREMENTS DURING DEVELOPMENT Although there are many reviews (see earlier) concerning the individual increase in amounts of different major plastid components such as lipids, pigments, nucleic acids, or proteins during greening, there have been no studies which directly and simultaneously compare the rates of accumulation by all these

CHLOROPLAST DEVELOPMENT

173

FIG.7. Rates of accumulation at successive intervals of time of carotenoids, lipids, chlorophyll, protein, and RNA during the greening of similar oat seedlings to those appertaining to Fig. 6 . RNA was isolated by a method based on that of Smillie and Krotkov (196) described by Gyldenholm (1968) and estimated using the orcinol procedure of Ogur and Rosen (1950). Total chlorophyll, carotenoids, and protein were estimated by the procedures of Arnon (1949), Hager and MeyerBertenrath (1 966), and Lowry er al. (195 1 ). Total lipids were estimated by weighing after extraction by the Bligh and Dyer (1959) procedure. Note the scale change required to encompass the protein accumulations.

different biosynthetic processes. By measuring the rate of accumulation (i.e., the rate of biosynthesis less the rate of turnover) rather than directly expressing the accumulated amount with time, it is possible to evaluate the likely bioenergetic demands as plastid development takes place. Figure 7 shows the rate of accumulation of total lipids, carotenoids, chlorophylls, and RNA, together with total protein on a much larger scale, during the greening of 8-day-old etiolated Avena seedlings at 25°C. When expressed in this manner several salient features emerge. First, the amount of biosynthetic “effort” diverted to protein biosynthesis far outweighs that devoted to the synthesis of everything else. Bioenergetic requirements for each type of biosynthesis differ both in terms of reductant and free energy input (i.e., how many phosphate bonds of ATP are hydrolyzed). Nevertheless, with its relatively high bioenergetic requirements, protein biosynthesis probably consumes well over 80% of the total available energy in relative terms at any time throughout plastid development. Significant increases in the rates of accumulation of either total lipid, carotenoid, or RNA only occur after the appearance of adequate photophosphorylation (cf. Fig. 7 with Fig. 6). The rates of overall chlorophyll accumulation are biphasic in nature. They show an additional early accumulation before the main phase which is most likely at the expense of imported or extraplastidic reserves. During early greening, there is also evidence that an early portion of protein biosynthesis precedes the appearance of photophosphorylationwhich is similarly dependent upon additional resources.

174

A.

R. WELLBURN

VIII. Influence of Light and Hormones A. PHOTO-CONTROL OF CHLOROPLAST DEVELOPMENT Various aspects of chloroplast development are directly or indirectly initiated, maintained or terminated by light. There have been a number of reviews of this particular topic, notably those of Mohr and Kasemir (1977), Schiff (1978), and Kasemir (1979). Consequently it is the intention in the following two subsections to cover only those aspects directly relevant to the mobilization of energy reserves, mitochondria1 participation, and transport processes rather than reiterate direct plastidic effects in detail. Two or three salient features require prior comment. For example, there appears to be sufficient evidence for up to three light-mediated receptors: (1) phytochrome, acting with a main peak responses at the red end of the spectrum, (2) a “red plus blue” light response (localized within the plastid) which may be associated with protochlorophyllide and/or phytochrome, and (3) a blue light absorbing photoeffector (occasionally called cryptochrome) which is probably a flavin rather than a carotenoid (Presti and Delbriick, 1978; Gressel, 1979). Whereas in higher plants light-induced plastid morphogenesis is largely under the regimen of phytochrome (Mohr, 1977), this red-sensitive photoreceptor is thought to be absent in algae (Holowinsky and Schiff, 1970; Boutin and Klein, 1972). In all algal plastid developmental systems, the blue light receptor is dominant (Sokawa and Hase, 1967; Senger and Bishop, 1972; Egan et al., 1975; Ohad and Drews, 1974) and the minor “red plus blue” receptor may be operative, as in Euglena (Egan and Schiff, 1974) or Scenedesmus (Brinkmann and Senger, 1978; Senger et al., 1980) or absent as in Chlumydornonas (Steup and Ssymank, 1978). Schiff (1978) has described an elegant model for overall control in Euglena based on the formation of inhibitory proteins during paramylum breakdown, negative control by blue light upon nuclear transcription, regulatory signals passing back from the plastid to the nucleus, and at least two sites within the plastid sensitive to red plus blue light. A conspicuous change in relative importance of the photoreceptors has occurred during the evolution of green plants. With some exceptions, in higher plants phytochrome becomes more predominant and the relative significance of the blue light photoreceptor is much reduced (Mohr, 1980). Interactions between blue and red light operate either through phytochrome alone or in association with a blue-sensitive receptor. These red plus blue light effects are more evident in lower plants and gymnosperms (Raghavan and DeMaggio, 1971; Kasemir et ul., 1980). Notable blue light-induced responses in higher plants have been detected in rhythms (Wagner, 1977), stomata1 opening (Mansfield and Meidner, 1966), root plastid development (Bjom et al., 1963; Bjom, 1980), “sun-leaf” effects (Lich-

CHLOROPLAST DEVELOPMENT

175

tenthaler ef al., 1980), lipid biosynthesis (Kleudgen and Lichtenthaler, 1974; Tevini, 1977), and increases in certain light-induced transcriptions (Gressel, 1978, 1980) and in nitrogenkarbon assimilation ratios (Voskresenskaya, 1972) as well as oxidative phosphorylation (Robinson and Wellburn, 1981). To what extent in each case phytochrome is functional at shorter wavelengths or, alternatively, an exclusively blue-sensitive photoreceptor is involved is not known. Earlier claims of PLB dispersal by blue light (Henningsen, 1967) have not been substantiated (Wellburn and Wellburn, 197313; Bradbeer er al., 1974). Studies to establish the intracellular localization of photoreceptors are equally fraught with problems. Fractions from Zea exhibiting blue light sensitivity were not specifically associated with either mitochondria or plastids (Hertel et al., 1972). Claims for phytochrome association are more wide-spread although in dark-grown plants the bulk of phytochrome is found in the cytoplasm and only binds to particulate fractions after irradiation (Coleman and Pratt, 1974). Studies using '251-labeled phytochrome have detected enhanced binding to mitochondria-rich fractions from Avena (Georgevich et al., 1977; Cede1 and Roux, 1980a) while de novo synthesis of phytochrome has been detected in plastid-containing, b d not mitochondria-enriched, fractions from phytochromedepleted oat seedlings (Grombein er al., 1978). B. RESPIRATORY ENHANCEMENT A N D RESERVE MOBILIZATION BY LIGHT When dark-grown algal cells are illuminated to induce greening, marked enhancement of respiration and degradation of polysaccharide occurs (Kowallik and Gaffron, 1966, 1967; Schmid and Schwarze, 1969; Senger and Bishop, 1972; Schiff, 1974; Brinkman and Senger, 1978; Kowallik and Schatzle, 1980). Inhibitor experiments have shown that partial control of light-enhanced respiration is exerted at the transcriptional level (Oh-hama and Senger, 1975; Watanabe er al., 1980). Schwartzbach et al. (1975) have postulated the existence of a regulatory protein which normally blocks paramylum breakdown. Blue light blocks the induction of this regulator thereby allowing paramylum breakdown. Experiments with added glucose have suggested that blue light mainly enhances respiration by virtue of provision of substrate from polysaccharide degradation (Brinkmann and Senger, 1980) but changes in membrane permeability may also be involved. Studies with light-grown higher plant plastid developmental systems are not sufficiently far advanced for there to have been any similar studies of spectral quality upon amyloplast transformations or mitochondria1 behavior. Increased respiration has been reported during germination (Woodstock and Toole, 1977) and this is under phytochrome control (Pecket and Al-Charchafchi, 1979). In greening tissues there are a number of reports of red light-sensitive changes in the rates of oxidation of NADH by mitochondria (Manabe and Furuya, 1974;

176

A. R . WELLBURN

Hampp and Wellburn, 1979; Cede1 and Roux, 198Ob). Induction-reversion experiments and treatment with continuous far-red light (to disengage phytochrome activity from photosynthesis) established that the changes in rates of respiration by isolated mitochondria from greening tissue given different light treatments are under phytochrome control (Hampp and Wellburn, 1979). Illumination with blue and red light or blue light alone upon intact leaves before mitochondrial isolation also enhances the rates of oxidative phosphorylation (Robinson and Wellburn 1981) but isolated mitochondria are completely insensitive (in vitro) to light stimulation. Quail and Briggs (1978) have shown that disruption of cellular integrity diminishes or destroys the phytochrome response so either the phytochrome exerts its affect upon mitochondria by means of events elsewhere in the cell or mitochondrial isolation damages potential receptive sites. C. LIGHTEFFECTSON TRANSPORT AND BIOENERGETIC PARAMETERS Red/far-red induction-reversion and decoupling far-red irradiations have established that phytochrome mediates changes in envelope permeability of both mitochondria and plastids from greening laminae (Hampp and Schmidt, 1977). Subsequent experiments showed that only the velocities of orthophosphate uptake by mitochondria were influenced by red light treatments (Hampp and Wellburn, 1979), the affinity values (&)remained unchanged. Red or far-red light modifies the levels of the adenine nucleotides (Sandmeier and Ivart, 1972; White and Pike, 1974; Friederich and Mohr, 1975; OelzeKarow and Mohr, 1978). By itself red light usually increases overall ATP levels although this can vary with time of sampling. Pool sizes also change and as a consequence total energy charge ratios are little altered throughout greening. TABLE I1 EFFECTSOF PHYTOHORMONES ON FUNCTIONS ASSOCIATED WITH PLASTID DEVELOPMENT ~~

Function Reserve mobilization Starch formation Starch hydrolysis a-Amylase activity

HormoneU

IAA ABA GA CK C2H4

Invertase activity

ABA GA CK CzH4

Change

Reference

Woiny er al. (1973) Huber and Sankhla (1974) See Yomo and Varner (1971) Clum (1967) Herrero and Hall ( 1 960) Cornforth er al. (1965) See Higgins and Jacobsen (1978) Rutherford and Bard (1971) See Abeles ( 1972) (continued)

CHLOROPLAST DEVELOPMENT

177

TABLE 11 (Continued) Function

HormoneB

Storage protein degradation

CK

Storage lipid degradation

GA

Respiration 0 2 consumption by tissues

IAA ABA C2H4

ABA GA

O 2 consumption in vitro

Oxidative phosphorylation (treatment before isolation) Oxidative phosphorylation (treatment after isolation) Respiratory control in v i m Altered TCA cycle activity

CK IAA GA IAA GA ABA CK GA ABA CK GA IAA C2H4

Nucleotide levels and transport ATP levels Cyclic AMP levels

Membrane phosphorylation Fluidity of membranes Orthophosphate uptake Plastid bioenergetics Photophosphorylation

Photophosphorylation NADPH levels

IAA IAA GA CK GA ABA

IAA GA or CK ABA GA CK

Change

+ + + + + + V 0 0

+ + -

0 0 0 0 0 V f

+ + + + +

Reference Jakubek and Mlodzianowski (1978) Mamott and Northcote (1975)

Bonner (1933); Key et a / . (1960) Hemberg (1978); Goldthwaite (1974) See Higgins and Jacobsen (1978) Huber and Sankla (1974) Pel’tek and Kalinin (1967); Voinilo ef a / . (1967) See Higgins and Jacobsen (1978) Key et a/. (1960) Frost and Wilson (1972) Sarkissan and McDaniel (1966) Robinson and Wellburn (1981) Robinson and Wellburn (1981) Moore and Miller (1972) Robinson and Wellburn (1981) Robinson and Wellburn (1981) Simkins and Street (1970) Frost and Wilson (1972) See Higgins and Jacobsen (1978) See Higgins and Jacobsen (1978)

-

Marre and Forti (1958) Salomon and Mascarenhas (1971) Tarantowicz-Marek and Kleczkowski (1978) Ralph et al. (1976) Wood and Paleg (1974) Hemberg (1978)

+ +

Tamas et a / . (1972) Yakushkina and Pushkina (1971)

0

+ -

Davies et a/. ( 1 979) Yakushkina and Pushkina ( 1971) Yakushkina and Pushkina (1971)

oIAA, Indole-acetic acid; CK, cytokinins or kinetin; GA, gibberellic acid(s); ABA, abscisic acid; increase; -, decrease; 0, no effect; V , increases and decreases depending on C2H4. ethylene; conditions.

+,

178

A. R. WELLBURN

Similarly, changes in the levels of nicotinamide nucleotides also occur but, as in the case of the adenine nucleotides, they are not thought to be direct links in any casual chain from phytochrome leading to the phenomenon of photomorphogenesis (Frosch e t al., 1974; Friederich and Mohr, 1975). On the other hand the rate of irradiation-enhanced phytochrome pelletability is strongly correlated with cellular amounts of ATP (Quail and Briggs, 1978); therefore phosphorylation capacity appears to be a necessary requirement for phytochrome rearrangement upon illumination.

D. HORMONAL EFFECTS ON PLASTIDDEVELOPMENT Reviews have appeared concerning the regulation of chloroplast development and respiration by plant hormones (Kirk, 1970; Higgins and Jacobsen, 1978). From these it is clear that, with the possible exception of nitrogen metabolism, studies of effects of plant hormones on basic processes such as photosynthesis, respiration, and lipid metabolism have been neglected by contrast to hormonal studies of gene expression. Consequently the understanding of hormonal effects on all aspects of plastid development is fragmentary and in need of systematic studies not least in the areas of reserve deposition and mobilization, respiration, transport, photophosphorylation, and significant changes in the levels of nucleotides throughout the cell. Table I1 indicates the salient observations with respect to different hormones for these processes available so far. In specific areas such as studies of a-amylase activity during germination or respiratory responses during ripening a voluminous literature exists while in others virtually no studies have been published. Contradictory reports are frequent and as a consequence no overall concept or trend with respect to plastid development is apparent. In higher plants it is likely that the gibberellic acids (GA) are the prime regulatory molecules with respect to storage reserve mobilization and possibly are also involved in respiratory enhancement. From comparisons of in situ (treatment before isolation) and in vitro (treatment after isolation) effects of GA upon mitochondria1 function (Robinson and Wellburn, 1981), it appears that either isolation damages hormone receptor sites on mitochondria or that receptor sites are located elsewhere in the cell and the mitochondria respond to secondary stimuli of some kind. These could be changes in general levels of nucleotides and substrates or they may be more specific. The case for the likely intervention of CAMP (adenosine-3’,5’-cyclic monophosphate) in this role during greening of dark grown Euglena has been discussed by Nigon and Heizmann (1978). Until recently the occurrence of CAMP, especially in higher plants, has been a subject of controversy (Keates, 1973; Amrhein, 1974, 1977; Lin, 1974; Hintermann and Parish, 1979) but unambiguous identification in higher plant tissue has recently been demonstrated (Newton et al., 1980; Janistyn, 1981; Johnson et al., 1981) using mass spec-

CHLOROPLAST DEVELOPMENT

179

trometry in conjunction with GLC. These techniques should now allow more meaningful examinations of the various function ascribed to CAMP(see Letham, 1978, for review) including those related to the various aspects of plastid development in both higher plants and algae.

ACKNOWLEDGMENTS I am indebted to Dr. Jean Whatley (Oxford) and Dr. David Goodchild (C.S.I.R.O., Canberra) for some of the micrographs and to all those co-workers with whom I have been privileged to collaborate at one time or another on various aspects of plastid development.

REFERENCES Abeles, F. B. (1972). Annu. Rev. Plant Physiol. 23, 259-292. Ahlberg, J. H., Nilson, E. H., and Walsh, J. L. (1967). “The Theory of Splines and Their Applications.” Academic Press, New York. Akoyunoglou, G., Argyroudi-Akoyunoglou, J. H., Michel-Wolwertz, M. R., and Sironval, C. (1966). Physiol. Plant. 19, 1101-1104. Amrhein, N. (1974) Plunta 118, 241-258. Amrhein, N. (1977). Annu. Rev. Plant Physiol. 28, 123-132. Anderson, J . M. (1975). Biochim. Biophys. Acru 416, 191-235. Aoki, S., and Hase, E. (1964). Plant Cell Physiol. 5, 473-484. Apel, K., and Kloppstech, K. (1978). Eur. J . Biochem. 85, 581-588. Apel, K., Santel, H.-J., Redlinger, T. E., and Falk, H. (1980). Eur. J . Biochem. 111, 251-258. Appelqvist, L. A., Stumpf, P. K., and von Wettstein, D. (1968). Plant Physiol. 43, 163-187. Argyroudi-Akoyunoglou, J. H., and Akoyunoglou, G. (1970). Plunr Physiol. 46, 247-249. Arnon, D. 1. (1949). Plunr Physiol. 24, 1-15. Arntzen, C. J., and Briantais, J. M. (1975). In “Bioenergetics of Photosynthesis” (Govindjee, ed.), pp. 51-1 13. Academic Press, New York. Atkinson, D. E. (1968). Biochemistry 11, 4030-4034. Atkinson, D. E. (1971). In “Metabolic Pathways” (H. J. Vogel, ed.), Vol. 5, pp. 1-21. Academic Press, New York. Bahl, J., and MonCger, R. (1976). C. R . Acad. Sci. 282, 177-180. Bahl, I., Franke, B., and MonCger, R . (1976). PIanra 129, 193-301. Bahl, J . , Franke, B., and MonCger, R. (1977). Biol. Cell. 30, 283-292. Bailey, J . L., and Whyborn, A. G. (1963). Biochim. Biophys. Actu 78, 163-174. Baker, N. R., and Butler, W. L. (1976). Plant Physiol. 58, 526-529. Ben-Shaul, Y., and Ophir, I. (1970). Can. J. Bot. 43, 129-136. Benz, J., Wolf, C., and Riidiger, W. (1980). Planr Sci. Lett. 19, 225-230. Bertram, M., and Heinz, E. (1976). Planfa 132, 161-168. Besouw, A,, van, and Wintermans, J. F. G. M. (1978). Biochim. Biophys. Actu 529, 44-53. Bisalputra, T., and Bisalputra, A. A. (1969). J. Ultrustruct. Res. 29, 151-170. Bisalputra, T., and Bisalputra, A. A. (1970). J. Ulrrustruct. Res. 32, 417-429. Bisalputra, T., and Weier, T. E. (1964). Am. J . Bot. 51, 881-892. Bishop, D. G. (1974). Phoruchem. Phorobiol. 20, 281-299.

180

A. R. WELLBURN

Bishop, N. I. (1971). In “Methods in Enzymology” (A. SanPietro, ed.), Vol. 23, Part A, pp. 130-143. Academic Press, New York. Bjorn, L. 0. (1980). In “The Blue Light Syndrome” (H. Senger, ed.), pp. 450-464. SpringerVerlag, Berlin and New York. Bjorn, L. O., Suzuki, Y., and Nilsson, J. (1963). Physiol. Planr. 16, 132-141. Bligh, E. G., and Dyer, W. I. (1959). Can. J. Biochem. Physiol. 37, 911-917. Boardman, N. K. (1977a). Annu. Rev. Plant Physiol. 28, 355-377. Boardman, N. K. (1977b). In “Encyclopaedia of Plant Physiology. New Series” (A. Trebst and M. Avron, eds.), Vol. 5, Photosynthesis I, pp. 583-600. Springer-Verlag, Berlin and New York. Boardman, N. K., and Anderson, J. M. (1978). In “Chloroplast Development” (G. Akoyunoglou and J. H. Argyroudi-Akoyunoglou, eds.), pp. 1-14. Elsevier, Amsterdam. Boasson, R., and Gibbs, S. P. (1973). Planta 115, 125-134. Bobrik, M., and Herich, R. (1978). Cyrologia 43, 665-670. Boehm, J. (1883a). Bot. 2. 41, 33-38. Boehm, J. (1883b). Bot. 2. 41, 49-54. Boffey, S. A., SelldCn, G., and Leech, R. M. (1980). Plant Physiol. 65, 680-684. Bomsel, J-L., and Pradet, A. (1967). Physiol. V i g . 5, 223-236. Bonner, J. (1933). J. Gen. Physiol. , 63-76. Borner, T., Schumann, B., and Hagemann, R. (1976). In “Genetics and Biogenesis of Chloroplasts and Mitochondria” (T. Bucher., W. Neupert., W. Sebald, and S. Wemer, eds.), pp. 41-48. North-Holland Publ., Amsterdam. Borner, T., Manteuffel, R., and Wellbum, A. R. (1979). Protoplasma 98, 153-161. Boutin, M. E., and Klein, R. M. (1972). Plant Physiol. 49, 656-657. Bourdu, R., Champigny, M.-L., Lefort, M., Maslow, M., and Moyse, A. (1965) Physiol. Vkg. 3, 355-392. Bovarnik, J. G., Schiff, J. A., Freedman, A., and Egan, J. M. (1974). J. Gen. Microbiol. 83, 63-71. Bradbeer, J. W., Gyldenholm, A. O., Smith, J. W., Rest, J., and Edge, H. J. W. (1974). New Phytol. 73, 281-290. Breidenbach, R. W . , Castelfranco, P., and Criddle, R. C. (1967a). Plant Physiol. 42, 1035-1041. Breidenbach, R. W., Castelfranco, P . , and Peterson, C. (1967b). Plant Physiol. 41, 803-809. Brinkmann, G., and Senger, H. (1978). In “Chloroplast Development” (G. Akoyunoglou and J. H. Argyroudi-Akoyunoglou, eds.), pp. 201-206. Elsevier, Amsterdam. Brinkmann, G., and Senger, H. (1980). In “The Blue Light Syndrome” (H. Senger, ed.), pp. 526-540. Springer-Verlag. Berlin and New York. Bronchart, R. (1970a). C. R. Acad. Sci. 270, 1789-1791. Bronchart, R. (1970b). Proc. Int. Congr. Electron Microsc., 7th, Grenoble 3, 187-188. Brown, G. E., and Preston, J. F. (1975). Arch. Microbiol. 104, 233-236. Brown, H. T., and Morris, G. H. (1893). J . Chem. SOC. , 604-677. Brunold, C., and Schiff, J. A. (1976). Plant Physiol. 57, 430-436. Buetow, D. E. (1968). “The Biology of Euglena.” Academic Press, New York. Buetow, D. E., Wurtz, E. A., and Gallagher, T. (1980). In “Nuclear Cytoplasmic lnteractions in the Cell Cycle,” pp. 9-55. Academic Press, New York. Buvat, R. (1959). C. R. Hebd. Skanc. Acad. Sci.. Paris 249, 289-292. Calvaryac, R., Leute, F. V., and Butow, R. A. (1971). Science 173, 252-254. Carde, J. P., Bernard-Dagan, C., and Gleizes, M. (1980). In “Biogenesis and Functions of Plant Lipids” (P. Mazliak, P. Benveniste, C. Castes, and R. Douce, eds.), pp. 441-444. Elsevier, Amsterdam. Carlsson, R., and Sundqvist, C. (1979). Physiol. Plant. 47, 105-1 11.

CHLOROPLAST DEVELOPMENT

181

Cattolico, R. A., Boothroyd, J. C . , and Biggs, S . P. (1976). Plant Physiol. 57, 497-503. Cedel, T. E., and Roux, S. J. (1980a). Plant Physiol. 66, 696-703. Cedel, T. E., and Roux, S. J. (1980b). Plant Physiol. 66, 704-709. Chappell, J. B., and Crofts, A. R. (1965). Eiochem. J . 95, 707-716. Chappell, J . B., and Crofts, A. R. (1966). In “Regulation of Metabolic Processes in Mitochondria” (J. M. Tager, S. Papa, E. Quagliariello, and E. C. Slater, eds.), Vol. 7, pp. 293-316. BBA Library, American Elsevier, New York. Chapman, D. E. A., and Graham, D. (1974a). Plantfhysiol. 53, 879-885. Chapman, D. E. A., and Graham, D. (1974b). Planr Physiol. 53, 886-892. Chen-She, S. H., Lewis. D. H., and Walker, D. A. (1975). New Phytol. 74, 381-390. Cherry, 1. H. (1963). Plant Physiol. 38, 440-446. Chevallier, D., and Douce, R. (1976). Plant Physiol. 57, 400-402. Chua, N. H., and Schmidt, G. (1979). J . Cell Eiol. 81, 461-483. Clum, H. H. (1967). Plant Physiol. 42, 568-572. Cobb, A. H., and Wellburn, A. R. (1973). Planra 114, 131-142. Cobb, A. H., and Wellbum, A. R. (1974). Planra 121, 273-282. Cobb, A. H., and Wellbum, A. R. (1976). Planta 129, 127-131. Cockburn, B. I., and Wellburn, A. R. (1974). J . Exp. Eot. 25, 36-49. Coleman, R. A,, and Pratt, L. H. (1974). Planta 121, 119-131, Cook, J. R. (1963). J . Protozool. 10, 436-444. Cook, J. R. (1966a). J . Cell Eiol. 29, 369-373. Cook, J. R. (1966b). In “Cell Synchrony : Studies in Biosynthetic Regulation (I. L. Cameron and G. M. Padilla, eds.), pp. 153-168. Academic Press, New York. Cook, J. R., Haggard, S. S . , and Harris, P. (1976). J . Prorozool. 23, 368-373. Cooke, R. J. (1977). New Phytol. 78, 91-94. Cornforth, J. W., Milborrow, B. V., Ryback, G . , and Wareing, P. F. (1965). Nature (London) 205, 1269- 1270. Costes, C., Burghoffer, C., Joyard, J., Block, M . , and Douce, R. (1979). FEESLetr. 103, 17-21. Cran, D. G . , and Possingham, J. V. (1972). Protoplasma 74, 345-353. Cran, D. G., and Possingham, J. V. (1973). Protoplasma 79, 197-213. Cran, D. G . , and Possingham, J . V. (1974). Ann. Eot. 38, 843-847. Cronshaw, J., Hoefert, L., and Esau, K . (1966). J . Cell Eiol. 31, 429-443. Crotty, W. J., and Ledbetter, M. C. (1973). Science 182, 839-841. Das, G. (1973). Can. J . Eot. 51, 113-120. Davies, W I., Mansfield, T. A,, and Wellburn, A. R. (1979). In “Plant Growth Substances 1979” (F. Skoog, ed.), pp. 242-253. Springer-Verlag. Berlin and New York. Davis, B., and Merrett, M. J. (1974). Plant Physiol. 53, 575-580. Davis, E. J., and Lumeng, L. (1975). J . Eiol. Chem. 250, 2275-2282. Day. D. A., and Hanson, J . B. (1977). Plant Physiol. 59, 630-635. DeGreef, J. A,, and Verbelen, J. P. (1973). Ann. Eor. 37, 593-596. DeGreef, J. A , , and Verbelen, J. P. (1977). Ann. Eor. 41, 1371-1373. DeGreef, J. A , , Caubergs, R., Verbelen, J. P., and Moereels, E. (1976). In “Light and Plant Development” (H. Smith, ed.), pp. 295-316. Butterworths, London. Dobberstein. B., Blobel, G . , and Chua, N. H. (1977). Proc. Nurl. Acad. Sci. U . S . A . 74, 1082- 1085. Dockerty, A,, and Merrett, J. M. (1979). Plant Physiol. 63, 468-473. Doll, S., Lutz, C., and Ruppel, H. G. (1976). Z . Pflanzenphysiol. 80, 166-176. Dolzmann, P., and Ullrich, H. (1966). Z. Pflanzenphysiol. 55, 165-180. Douce, R. (1974). Science 183, 852-853.

182

A. R. WELLBURN

Douce, R., and Joyard, J. (1976). Adv. Bot. Res. 7, 1-1 16. Douce, R., and Joyard, J. (1978). In “Chloroplast Development” (G. Akoyunoglou and J. H. Argyroudi-Akoyunoglou, eds.), pp. 283-296. Elsevier, Amsterdam. Douce, R., and Joyard, J. (1981). TrendsBiochem. Sci. 6, 237-239. Dmmm, H. E., and Margulies, M. M. (1970). Planr Physiol. 45, 23-33. Dubertret, G., and Joliot, P. (1974). Biochim. Biophys. Acra 357, 399-411. Duysen, M. E., Freeman, T. P., and Zabrocki, R. D. (1980). Plant Physiol. 65, 880-883. Dwyer, M. R., and Smillie, R. M. (1970). Biochim. Biophys. Acra 216, 392-401. Dwyer, M. R., and Smillie, R. M. (1971). Ausr. J. Biol. Sci. 24, 15-22. Dwyer, M. R., Smydzuk, J., and Smillie, R. M. (1970). Ausr. J. Biol. Sci. 35, 1005-1013. Earnshaw, M. J., and Hughes, E. A. (1976). Plant Sci. Lett. 6, 343-348. Edwards, G. E., Robinson, S. P., Tyler, N. J. C., and Walker, D. A. (1978). Plant Physiol. 62, 313-319. Egan, J. M., and Schiff, J. A. (1974). Plant Sci. Lert. 3, 101-105. Egan, J . M., Dorsky, D., and Schiff, J. A. (1975). Plant Physiol. 56, 318-323. Egneus, H., Reftel, S., and SelldCn, G. (1972). Physiol. Planr. 27, 48-55. Ellis, R. J., and Baraclough, R. (1978). In “Chloroplast Development” (G. Akoyunoglou and J. H. Argyroudi-Akoyunoglou, eds.), pp. 185- 194. Elsevier, Amsterdam. Ellis, R. J., Highfield, P. E., and Silverthome, J. (1977). Proc. Int. Congr. Phorosynth.. 4rh pp. 497-506. Erecinska, M., and Wilson, D. F. (1978). Trends Biochem. Sci. 3, 219-223. Eriksson, G., Kahn, A., Walles, B., and von Wettstein, D. (1961). Ber. Drsch. Bot. Ges. 74, 221-232. Esau, K. (1953). “Plant Anatomy”. Wiley, New York. Esau, K., and Cronshaw, J. (1967). J. Cell Biol. 33, 665-678. Eytan, G., and Ohad, I. (1972). J . Biol. Chem. 247, 1 12-12], Freyssinet, G. (1976a). PIanr Physiol. 57, 824-830. Freyssinet, G. (1976b). Plant Physiol. 57, 831-835. Freyssinet, G., and Schiff, J. A. (1974). Plant Physiol. 53, 543-554. Freyssinet, G., Heizmann, P., Verdier, G., Trabuchet, G . , and Nigon, V. (1972). Physiol. Vbg. 10, 421-442. Friedberg, I., Goldberg, I., and Ohad, 1. (1971). J. Cell Biol. 50, 268-275. Friederich, K. E., and Mohr, H. (1975). Phorochem. Phorobiol. 22, 49-53. Frosch, S . , Wagner, E., and Mohr, H. (1974). Z . Narurforsch. 29c, 392-398. Frost, R . , and Wilson, H. (1972). Phyron 29, 79-88 Gabrielsen, E. K., Madsen, A., and Vejlby, K. (1961). Physiol. Plant. 14, 98-1 10. Galling, G. (1978). In “Chloroplast Development” (G. Akoyunoglou and J. H. ArgyroudiAkoyunoglou, eds.), pp. 439-444. Elsevier, Amsterdam. Georgevich, G., Cedel, T. E., and Roux, S. J. (1977). Proc. Natl. Acad. Sci. U.S.A. 74, 4439-4443. Ghosh, H. P., and Preiss, J. (1966). J . Biol. Chem. 241, 4491-4504. Gnanam, A., Govindarajan, A. G., and Vivekanandan, M. (1974). In “Biomembranes: Architecture, Biogenesis, Bioenergetics and Differentiation” (L. Packer, ed.), pp. 353-367. Academic Press, New York. Goff, J. L. (1979). J . Phycol. 15, 87-100. Goldthwaite, J. (1974). Plant Physiol. 54, 339-403. Goodenough, U. W. (1970). J. Phycol. 6, 1-6. Goodwin, T. W. (1965). In “Biosynthetic Mechanisms in Higher Plants” (I. B. Pridham and T. Swain, eds.), pp. 57-71. Academic Press, New York. Goodwin, T. W., and Mercer, E. I. (1963). Biochem. SOC. Symp. 24, 37-56.

CHLOROPLAST DEVELOPMENT

183

Graham, D. (1978). Proc. Ausr. Biochem. SOC. 11, 46-47, Graham, D., and Cooper, J. E. (1967). Aust. J . Biol. Sci. 20, 319-327. Graham, D., and Walker, D. A. (1962). Biochem J. 82, 554-560. Gray, J. C., and Kekwick, R. G. 0. (1973). Biochem. J. 133, 335-347. Gressel, J. (1978). Photochem. Photobiol. 27, 167-169. Gressel, J. (1979). Photochem. Photobiol. 30, 749-754. Gressel, J. (1980. In “The Blue Light Syndrome” (H. Senger, ed.), pp. 133-153. Springer-Verlag, Berlin and New York. Griffiths, W. T. (1978). Biochem. J. 174,681-692. Grimme, L. H., and Porra, R. J. (1974). Arch. Microbiol. 99, 173-179. Grombein, S.,Rudiger, W., and Hampp, R. (1978). Plantu 141, 273-277. Grumbach, K. H., and Forn, B. (1980). Z. Naturforsch. 35c, 645-648. Gunning, B. E. S. (1965a). Protoplasma 60, 111-130. Gunning, B. E. S. (1965b). J. CellEiol. 24, 79-93. Gunning, B. E. S . , and Jagoe, M. P. (1967). In “Biochemistry of Chloroplasts” (T. W. Goodwin, ed.), Vol. 2, pp. 665-676. Academic Press, New York. Gunning, B. E. S., and Steer, M. W. (1975). “Ultrastructure and the Biology of Plant Cells.” Arnold, London. Gunning, B. E. S., Steer, M. W., and Cockrane, M. P. (1968). J. Cell Sci. 3, 445-456. Gyldenholm. A. 0. (1968). Hereditas 59, 142-168. Gyldenholm, A. O., and Whatley, F. R. (1968). New Phyrol. 67,461-468. Hackenbrock, C. R., and Miller, K. J. (1975). J . Cell Biol. 65, 615-630. Hager, A., and Meyer-Bertenrath, R. (1966). Planta 69, 198-217. Hampp, R. (1978). Plant Physiol. 62, 735-740. Hampp, R. (1979). Planta 144, 325-332. Hampp, R. (1980). Planra 150, 291-298. Hampp, R., and DeFilippis, L. F. (1980). Plant Physiol. 65, 663-668. Hampp, R.,and Riehl, M. (1981). Proc. fnt. Congt. Phorosynth. 5th 5 , 969-976. Hampp, R., and Schmidt, H. W. (1976). Planta. 129, 69-73. Hampp, R.,and Schmidt, H. W. (1977). Z. Pflanzenphysiol. 82, 68-77. Hampp, R., and Wellburn, A. R. (1976a). J. Exp. Bor. 27, 778-784. Hampp, R., and Wellburn, A. R. (1976b). PIanra 131, 21-26. Hampp, R., and Wellburn, A. R. (1976~).Z. Pflanzenphysiol. 79, 246-253. Hampp, R., and Wellburn, A. R. (1978). Ber. Drsch. Eot. Ges. 91, 551-561. Hampp, R., and Wellbum, A. R. (1979). Planra 147, 229-235. Hampp, R., and Wellburn, A. R. (1980). Z. Pjlanzenphysiol. 98, 289-303. Hampp, R., and Ziegler, H. (1980). Planta 147, 485-494. Heber, U. (1974). Annu. Rev. Plant Physiol. 25, 393-421. Heber, U., and Santarius, K. A. (1965). Biochim. Biophys. Acta 109, 390-408. Heber, U., and Santarius, K. A. (1970). Z. Naturforsch. 25b, 718-728. Heinz, E., Bertram, M., Joyard, J., and Douce, R. (1978). Z. Pjlanzenphysiol. 87, 325-331. Heldt, H.W. (1969).FEBSLett. 5, 11-14. Heldt, H. W. (1976a). In “Encyclopedia of Plant Physiology, New Series” (C. R. Stocking and U. Heber, eds.), Vol. 3, pp. 137-143. Springer-Verlag, Berlin and New York. Heldt, H. W. (1976b). In “Encyclopedia of Plant Physiology, New Series” (C. R. Stocking and U. Heber, eds.), Vol. 3, pp. 235-254. Springer-Verlag, Berlin and New York. Heldt, H. W., Sauer, F., and Rapley, L. (1971). Proc. f n t . Congr. Photos.vnth. Res., 2nd pp. 1345- 1355. Heldt, H. W., Fliege, R., Lehner, K., Milovancev, M., and Werdan, K. (1975). Proc. fnr. Congr. Photosynrh. Res., 3rd 1369-1379.

184

A. R. WELLBURN

Hemberg, T. (1978). Physiol. Plant. 43, 65-67. Henningsen. K. W. (1967). In “Biochemistry of Chloroplasts” (T. W. Goodwin, ed.), Vol. 2, pp. 453-457. Academic Press, New York. Henningsen, K. W., and Boardman, N. K. (1973). Plant Physiol. 51, I 1 17-1 126. Henry, E. W. (1979). Cyrologia 44, 727-738. Herrero, F . , and Hall, W. C. (1969). Physiol. Plant 13, 736-750. Herron, H. A. and Mauzerall, D. (1972). Plant Physiol. 50, 141-148. Hertel, R., Thompson, K. St., and Russo, V. (1972). Planfa 107, 325-340. Heyes, I. K., and Dale, J. E. (1971). New Phyrol. 70, 415-426. Higgins, T. J. V., and Jacobsen, J. V. (1978). I n “Phytohormones and Related Compounds-A Comprehensive Treatise” (D.S. Letham, P. B. Goodwin, and T. J. V. Higgins, eds.), Vol. 1 , pp. 467-5 14. Elsevier, Amsterdam. Highfield, P. E., and Ellis, R. J. (1978). Nature (London) 271, 420-424. Hill, H. M., and Rogers, L. I. (1974). Phyrochemistry 13, 763-777. Hiller, R. G., and Boardman, N. K. (1971). Biochim. Biophys. Acta 253, 449-458. Hintermann, R., and Parish, W. R. (1979). Planta 146, 495-461. Holdsworth, R. H. (1971). J . Cell Biol. 51, 499-513. Holian, A,, Wilson, D. F., and Owen, C. S. (1977). Arch. Biochem. Biophys. 181, 164-171. Holowinsky, A. W., and Schiff, J. A. (1970). Plant Physiol. 45, 339-347. Hormm, M. A., and Schwartzbach, S. D. (1980). Planta 149, 376-383. Howes, C. D., and Stem, A. I. (1973). Plant Physiol. 51, 386-390. Huber, W., and Sankhla, N. (1974). Z. Pfanzenphysiol. 71, 275-280. Hudock, G. A,, McLeod, G. C., Morakova-Kiely, J., and Levine, R. P. (1964). Plant Physiol. 39, 898-903. Hyde, B. B., Hodge, A. J., Kahn, A., and Bimstiel, M. L. (1963). J. Ultrastruct. Res. 9,248-258. Iordanov, 1. (1970). Dokl. Bolg, Akad. Nauk 23, 999-1002. Iordanov, I . , and Popov, K. (1967). Stud. Biophys. 5, 123-130. Jakubek, M., and Mlodzianowski, F. (1978). Acra SOC. Bot. Pol. 47, 319-324. Janistyn, B. (1981). 2. Naturforsch. 36c, 193-196. Jeske, C., and Senger, H. (1978). In “Chloroplast Development” (G. Akoyunoglou and J. H. Argyroudi-Akoyunoglou, eds.), pp. 475-480. Elsevier, Amsterdam. Johnson, L. P., MacLeod, J. K., Parker, C. W., and Letham, D. S. (1981). FEBS Lett. 124, 1 19- 121. Joy, K. W . , and Ellis, R. J. (1975). Biochim. Biophys. Acfa 378, 143-151. Joyard, J . , and Douce, R. (1974). FEES Lett. 51, 335-340. Joyard, J., and Douce, R. (1977). Biochim. Biophys. Acfa 486, 273-285. Joyard, J., and Stumpf, P. K. (1980). Plant Physiol. 65, 1039-1043. Joyard, J., and Stumpf, P. K. (1981). Plant Physiol. 67, 250-256. Joyard, J., Douce, R., Siebertz, H. P., and Heinz, E. (1980). Eur. J . Biochem. 108, 171-176. Jung, D. W., and Hanson, J. B. (1973). Biochim. Biophys. Acfa 325, 189-192. Kahn, A. (1968a). Plant Physiol. 43, 1769-1780. Kahn, A. (1968b). Plant Physiol. 43, 1781-1785. Kahn, A., Boardman, N. K., and Thorne, S. W. (1970). J. Mol. Biol. 48, 85-101. Kanai, R., and Edwards, G. E. (1973). Plant Physiol. 51, 1133-1137. Kandler, O., and Haberer-Liesenkrotter, I. (1963). Z. Narurforsch. 18b, 718-730. Kasemir, H. (1979). Cell B i d . Int. Rep. 3, 197-214. Kasemir, H., Schroder, S., and Steinhilber, A. (1980). In “Photoreceptors and Plant Development” (J. DeGreef, ed.), pp. 185-198. Antwerp Univ. Press, Antwerp. Keates, R. A. B. (1973). Nafure (London) 244, 355-357. Kesselmeier, J., and Budzikiewicz, H. (1979). Z. Pflanzenphysiol. 91, 333-344.

CHLOROPLAST DEVELOPMENT

185

Kesselmeier, J., and Ruppel, H. G. (1979). In “Advances in the Biochemistry and Physiology of Plant Lipids” (L. A. Appelqvist and C. Liljenberg, eds.), pp. 187-191. Elsevier, Amsterdam. Key, J. L., Hanson, 1. B., and Bils, R. F. (1960). Plant Physiol. 35, 177-183. Kidd, F., West, C., and Briggs, G. E. (1921). Proc. R. SOC. Ser. B 92, 368-384. Kirk, J . T. 0. (1970). Annu. Rev. Plant Physiol. 21, 11-42. Kirk, J. T. 0.. and They-Bassett, R. A. E. (1967). “The Plastids”. Freeman, San Francisco, California. Klein, S., and Poljakoff-Mayber, A. (1961). J. Eiochem. Biophys. Cyrol. 11, 433-440. Klein, S., and Schiff, J . A. (1972). Plunr Physiol. 49, 619-626. Klein, S., Schiff, J. A., and Holowinsky, A. W. (1972). Dev. Biol. 28, 253-273. Kleudgen, H. K., and Lichtenthaler, H. K. (1974). Z. Naturforsch. 29c, 142-146. Klingenberg, M. (1970a). Essays Biochem. 6, 119-159. Klingenberg, M. (1970b). FEES Lett. 6, 145-154. Klingenberg, M., and ffaff, E. (1967). Methods Enzymol. 10, 6 8 M 8 4 . Kohn, S., and Klein, S. (1976). Planra 132, 169-175. Kolloffel, C. (1967). Acia Bot. NePrl. 16, 111-122. Konitz, W. (1965). Planfa 66, 345-373. Kordan, H. A. (1976). Ann. Bot. 40, 347-350. Kowallick, W., and Gaffron, H. (1966). Planta 69, 92-95. Kowallick, W., and Gaffron, H. (1967). Nature (Londonj 215, 1038-1040. Kowallick, W.. and Schatzle, S. (1980). In “The Blue Light Syndrome” (H. Senger, ed.), pp. 344-360. Spinger-Verlag, Berlin and New York. Kraayenhof, R. (1969). Biochim. Biophys. Acra 180, 213-215. Krause, G. H. (1971). Z. Pflanzenphysiol. 65, 13-23. Kuster, U., Bohnensack, R., and Kunz, W. (1976). Biochim. Biophys. Acta 440, 391-402. Larsson, C.-M., Tillberg, J.-E., and HallmCn, G. (1978). Physiol. Plant. 44, 115-121. Leech, R. M. (1976). In “Perspectives in Experimental Biology” (N. Sunderland, ed.), Vol. 2, pp. 145-184. Pergamon, Oxford. Leese, B. M., Leech, R. M., and Thompson, W. W. (1971). Proc. Int. Congr. Phorosynth. Res., 2nd 3, 1485-1493. Lefort-Tran, M. (1975). Colloq. Int. C. N . R. S . 240, 297-301. Lefort-Tran, M., Pouphile, M., Freyssinet, G., and Pineau, B. (1980). J. Ulrrusirucr. Res. 73, 44-63. Leomine, Y. (1966). C. R. Hebd. Skunc. Acad. Sci., Paris 263, 105-108. Letham, D. S. (1978). In “Phytohormones and Related Compounds-A Comprehensive Treatise” (D. S. Letham, P. B. Goodwin, and T. J. V. Higgins, eds.), Vol. I , pp. 349-417. Elsevier, Amsterdam. Lewenstein, A,, and Bachofen, R. (1972). Biochim. Biophys. Acta 267, 80-85. Lichtenthaler, H. K. (1966). Ber. Drsch. Bot. Ges. 79, 82-88. Lichtenthaler, H. K. (1968). Endeavour 27, 144-149. Lichtenthaler, H. K., Buschmann, C., and Rahmsdorf, U. (1980). In “The Blue Light Syndrome” (H. Senger, ed.), pp. 485-494. Springer-Verlag, Berlin and New York. Lin, P. P.-C. (1974). In “Advances in Cyclic Nucleotide Research” (P. Greengard and A. G. Robinson, eds.), Vol. 4, pp. 439-461. Raven, New York. Lowry, 0. H., Rosebrough, N. J . , Farr, A. L., and Randall, R. J . (1951). J. Biol. Chem. 193, 265-275. Lowry, 0. H., Carter, J., Ward, J. B., and Glaser, L. (1971). J . Biol. Chem. 246, 6511-6521. Luciani, S. (1975). Life Sci. 10, 961-964. Liirssen, K. (1971). 2.Naturforsch. 25b, 1 1 13- I 119. Liitz, D. (1975a). Z. Pfanzenphysiol. 75, 346-359.

186

A. R. WELLBURN

Liitz, C. (1975b). Z. PjZanzenphysiol. 7 6 , 130-142. Liitz, C. (1978). In ‘‘Chloroplast Development” (G. Akoyunoglou and J. H. ArgyroudiAkoyunoglou, eds.), pp. 481-488. Elsevier, Amsterdam. Liitz, C. (1980). Z. Naturforsch. 35c, 519-521. Lutz, C. (1981a). Protoplastnu 108, 83-98. Liitz, C. (1981b). Protoplasma 108, 99-115. Liitz, C., and Klein, S. (1979). Z. PjZanzenphysiol. 95, 227-237. Lutz, C., and Manning, U. (1980). In “Photoreceptors and Plant Development” (J. DeGreef, ed.), pp. 229-236. Antwerp Univ. Press, Antwerp. Liitz, C., Roper, U., Beer, N. S., and Griffiths, W. T. (1981). Eur. J. Biochem. 118, 347-353. McGivan, J. D., and Klingenberg, M. (1971). Eur. J. Biochem. 20, 392-399. McGivan, J. D., Grebe, K., and Klingenberg, M. (1971). Biochem. Biophys. Res. Commun. 45, 1533-1541. MacLachan, G. A,, and Porter, H. K. (1959). Proc. R. SOC.Ser. B 150, 460-473. Madsen, A. (1963). Photochem. Photobiol. 2 , 93-100. Manabe, K., and Furuya, M. (1974). Plant Physiol. 53, 343-347. Mangat, B. S . , Levin, W. B., and Bidwell, R. G. S. (1974). Can J . Bot. 52, 673-681. Mansfield, T. A., and Meidner, H. (1966). J. Exp. Bot. 17,510-521. Manton, 1. (1967). In “Biochemistry of Chloroplasts” (T. W. Goodwin, ed.), Vol. 1, pp. 23-47. Academic Press, New York. Margulies, M. M., and Parenti, F. (1968). Plant Physiol. 43, 504-514. Marre., E., and Forti, G. (1958). Physiol. Plant. 11, 36-47. Maniott, K. M., and Northcote, D. H. (1975). Biochem. J . 148, 139-144. Marsh, H. V., Galmiche, J. M., and Gibbs, M. (1965). Plant Physiol. 40, 1013-1022. Matsuka, M., and Hase, E. (1966). Plant Cell Res. 7 , 149-162. Meijer, A. J . , Groot, G. S. P., and Tager, J. M. (1970). FEES Lett. 8, 41-44. Meller, E., and Harel, E. (1978). In “Chloroplast Development” (G. Akoyunoglou and J. H. Argyroudi-Akoyunoglou, eds.), pp. 51-57. Elsevier, Amsterdam. Menke, W. (1960). Z. Naturforsch. 15b, 479-482. Mersey, B., and McCully, M. E. (1978). J . Microsc. 114, 49-76. Michel, J.-M. (1978). Plantu 141, 45-50. Michel-Wolwertz, M. R., and Bronchart, R. (1974). Plant Sci. Lett. 2, 45-54. Mohr, H. (1977). Endeavour 1, 107-1 14. Mohr, H. (1980). In “The Blue Light Syndrome” (H. Senger, ed.), pp. 97-109. Springer-Verlag, Berlin and New York. Mohr, H., and Kasemir, H. (1977). In “Research in Photobiology” (A. Castellani, ed.), pp. 501-509. Plenum, New York. Montes, G., and Bradbeer, J. W. (1976). Plant Sci. Lett. 6 , 35-41. Moore, A. L. (1982). SOC.Exp. Biol. Symp. 18, in press. Moore, T. S., and Miller, C. 0. (1972). Plantfhysiol. 50, 594-598. Morohashi, Y., and Bewley, J. D. (1980). Plantfhysiol. 66, 70-73. Morohashi, Y., and Shimokoriyama, M. (1975). J. Exp. Bot. 26, 932-938. M o d , D. J. (1975). Annu. Rev. Plant Physiol. 26, 441-481. Miihlethaler, K., and Frey-Wyssling, A. (1959). J. Eiophys. Biochem. Cyrol. 6, 507-512. Murakami, S . , and Strotmann-H. (1978). Arch. Biochem. Biophys. 185, 30-38. Nawa, Y.,and Asahi, T. (1971). Plantfhysiol. 48, 671-674. Neumann, D., and Parthier, B . (1973). Exp. Cell Res. 81, 255-268. Newcomb, E. H. (1967). J . Cell Biol. 33, 143-163. Newton, R. P., Gibbs, N., Moyse, C. D., Wiebers, J. L., and Brown, E. G. (1980). Phytochemistry 19, 1909-1911.

CHLOROPLAST DEVELOPMENT

187

Nigon, V., and Heinzmann, P. (1978). Inr. Rev. Cvtol. 53, 21 1-290. Nilshammer, M., and Walles, B. (1974). Protoplasma 79, 317-332. Ochiai, S., and Hase, E. (1970). Planr Cell Physiol. 11, 663-673. Oelze-Karow, H-M., and Butler, W. L. (1971). Plant Physiol. 48, 621-625, Oelze-Karow, H-M., and Mohr, H. (1978). Photochem. Photobiol. 27, 255-258. Ogur, M., and Rosen, G. (1950. Arch. Biochem. Biophys. 25, 262-276. Ohad, 1. (1975). In “Membrane Biogenesis of Mitochondria, Chloroplasts and Bacteria” (A. Tzagoloff, ed.), pp. 279-350. Plenum, New York. Ohad, I., and Drews, G. (1974). Proc. Inr. Congr. Phorosynrh. Res., 3rd pp. 1907-1912. Oh-hama, T., and Senger, H. (1975). Plant Cell Physiol. 116, 395-405. Oleszko, S., and Moudrianakis, E. N. (1974). J. Cell B i d . 63, 936-948. Olgren, W. L., and Krogman, D. W. (1965). J . Biol Chem. 240, 4603-4608. Onishi, T., Kroger, A., Heldt, H. W., Pfaff, E., and Klingenberg, M. (1967). Eur. J. Biochem. 1, 301-311. Ophir, I., and Ben-Shaul, Y. (1974). Protoplasma SO, 109-127. Ophir, I., Talmon, A., and Ben-Shaul, Y. (1974). Proc. Int. Congr, Electron Microsc.. 8rh 2, 364-365. Opik, H. (1965). J. Exp. Bor. 16, 667-682. Orcival-Lafont, A. M., Pineau, B., Ledoight, G., and Calvayrac, R. (1972). Can. J . Bot. 50, 1503- 1508. Osafune, T., and Schiff, J. A . (1980). J . Ultrasrruct. Res. 73, 64-76. Osafune, T., Mihara, S., Hase, M., and Ohkuro, I. (1972). Plant Cell Physiol. 13, 981-986. Osafune, T., Klein, S., and Schiff, J. A. (1980). J . Ulrrasfruct. Res. 73, 77-90. Oshio, Y., and Hase, E. (1969). Plant Cell Physiol. 10, 51-59. Palmer, J. M. (1976). Annu. Rev. Plant Physiol. 27, 133-157. Palmieri, F., Prezioso, G., Quagliariello, E., and Klingenberg, M. (1971). Eur. J. Biochem. 22, 66-74. Papa, S., Lofrumento, N. E., Loglisci, M. and Quagliariello, E. (1969). Biochim. Biophys. Acta 189, 311-314. Park, R. B., and Pon, N. G. (1961). J . Mol. Biol. 3, 1-10. Parkin, J. (1899). Proc. R. SOC.Philos. Trans. Ser. B . 191, 35-74. Parrot, M. J. (1981). Personal Computer World 4 (12), 132-133. Parthier, B. (1981). In “Biochemistry and Physiology of Protozoa” (A. Lwoff, ed.), 2nd Ed., pp. 261-300. Academic Press, New York. Passam, H. C., and Coleman, J. 0. D. (1975). J. Exp. Bot. 26, 536-543. Passam, H. C., Souverijn, J. H. M., and Kemp, A. (1973). Biochim. Biophys. Acta 305, 88-94. Pecket, R. C., and Al-Charchafchi, F. (1979). J. Exp. Bot. 30, 839-842. Pedersen, T. A , , Kirk, M.,and Bassharn, J. A. (1966). Physiol. Plant. 19, 219-231. Pel’tek, C. M., and Kalinin, F. L. (1967). Ukr. Bor. Zh. 24, 28-36. Pfaff, E., and Klingenberg, M. (1968). Eur. J . Biochem. 6, 66-79. Pfaff, E., Heldt, H. W., and Klingenberg, M. (1969). Eur. J. Biochem. 10, 484-493. Phillis, E., and Mason, T. G. (1937). Ann. Bot. 1, 231-237. Phung Nhu Hung, S., Hoarau, A,, and Moyse, A. (1970a). Z. Pflanzenphysiol. 62, 245-258. Phung Nhu Hung, S., Lacourly, A,, and Sarda, C. (1970b). A. Pflanzenphysiol. 62, 1-16. Pineau, B., and Douce, R. (1974). Proc. Inr. Congr. Phorosyn. Res., 3rd 3, 1667-1673. Platt-Aloia, K. A,, and Thomson, W. W. (1977). New Phytol. 78, 599-605. PlesniEar, M., and Bendall, D. S. (1971). Proc. Int. Congr. Phorosynrh. Res.. 2nd 3, 2367-2373. PlesniEar, M., and Bendall, D. S. (1973). Biochem. J. 136, 803-812. Preiss, J., Ghosh, H. P., and Wittkop, J. (1967). In “Biochemistry of Chloroplasts” (T. W. Goodwin, ed.), Vol. 2, pp. 131-153. Acacemic Press, New York.

188

A. R. WELLBURN

Presti, D., and Delbriick, M. (1978). Plant. Cell Environ. 1, 81-100. Price, W. C., Martinez, A. P., and Warmke, H. E. (1966). J. Ultrastruct. Res. 14, 618-621. Purcifull, D. E., Edwardson, J. R., and Christie, R. G. (1966). Virology 29, 276-284. Pyliotis, N. A,, and Goodchild, D. J . (1975). Arch. Microbiol. 103, 259-270. Quagliariello, E., and Palmieri, F. (1972). In “Biochemistry and Biophysics of Mitochondria1 Membranes” (G. F. Assone, E. Carafoli, A. L. Lehninger, E. Quagliariello, and N. Siliprandi, eds.), pp. 659-680. Academic Press, New York. Quail, P. H., and Briggs, W. R. (1978). Plant Physiol. 62, 773-778. Raghavan, V., and DeMaggio, A. E. (1971). Plant Physiol. 48, 82-85. Ralph, R. K., Bullivant, S., and Wojcik, S. J. (1976). Biochim. Biophys. Acta 421, 319-327. Rascio, N., Orsenigo, M., and Arboit, D. (1976). Protoplasma 90, 253-263. Redlinger, T. E., and McDaniel, R. G. (1977). Plant Physiol. 60,452-455. Redlinger, T. E., and McDaniel, R. G. (1978). Planr Physiol. 61, 1006-1009. Rimy, R., Phung Nhu Hung, S., and Moyse, A. (1972). Physiol. Vkg. 10, 269-290. Robards, A. W., and Humpherson, P. G. (1967). Planta 76, 169-178. Robertson, D., and Laetsch, W. M. (1974). Plant Physiol. 54, 148-159. Robinson, B. H., and Williams, G. R. (1970). Biochim. Biophys, Acta, 216, 63-70. Robinson, D. C., and Wellburn, A. R. (1981). Biochem. Physiol. Pflanzen 176, 54-59. Robinson, S. P., and Walker, D. A. (1979). Arch. Biochem. Biophys. 196, 319-323. Rogers, L. J., Shah, S. P. J., and Goodwin, T. W. (1966). Biochem, J . 99, 381-388. Rosinski, J., and Rosen, W. G. (1972). Q. Rev. Biol. 47, 160-191. Roughan, P. G., and Slack, C. R. (1977). Biochem. J. 162, 457-459. Ruppel, H. G., Kesselmeir, J., and Lutz, C. (1978). Z. Pflanzenphysiol. 90, 101-110. Rutherford, P. P., and Bard, D. R. (1971). Phytochemistry 10, 1635-1638. Ryberg, H., Axelsson, L., Klockave, B., and Sandelius, A. S. (1981). Proc. Int. Congr. Phorosynth., 5th, 295-304. Sabnis, D. D., Gordon, M., and Galston, A. W. (1970). Plant Physiol. 45, 25-32. Sachs, J. (1862). Bot. Z. 20, 365-373. Salomon, D., and Mascarenhas, J . P. (1971). Life Sci. 10, 879-885. Salvador, G. F., Lefort-Tran, M., Nigon, V., and Jourdan, F. (1971). Exp. CellRes. 64,457-462. Sandmeier, M., and hart, J. (1972). Photochem. Photobiol. 16, 51-59. Santarius, K. A., and Heber, U. (1965). Biochim. Biophys. Acta 102, 39-54. Sarkissian, I. V., and McDaniel, R. G. (1966). Biochim. Biophys. Acta 128, 413-418. Schantz, R., Salaun, J. P., Schantz, M. L., and Duranton, H. (1971). C. R . Hebd. Skanc. Acad. Sci., Paris 273, 1795- 1798. Schiff, J. A. (1970). Symp. SOC. Exp. Biol. 24, 277-301. Schiff, J . A. (1973). Adv. Morphog. 10, 265-312. Schiff, J . A. (1975). Colloq. Int. C. N . R . S . 240, 79-93. Schiff, J . A. (1978). In “Chloroplast Development” (G. Akoyunoglou and J . H. ArgyroudiAkoyunoglou, eds.), pp. 747-767. Elsevier, Amsterdam. Schmidt, G.,and Schwdrze, P. (1969). Hoppe Seyfers Z . Phys. Chem. 350, 1513-1520. Schmidt, G., and Lyman, H. (1976). In “Genetics of Algae” (R. Lewin, ed.), pp. 257-299. Univ. of Calif. Press, Berkeley, California. Schneider, M. M., Hampp, R., and Zeigler, H. (1977). Plant Physiol. 60, 518-520. Schoch, S., Lempert, U., and Riidiger, G. (1977). Z. Pflanzenphysiol. 83, 427-436. Schoch, S., Hehlein, C., and Rudiger, W. (1980). Plant Physiol. 66, 576-579, Schotz, F., and Diers, L. (1975). Planra 124, 277-285. Schwartzbach, S. D., Freyssinet, G., and Schiff, J . A. (1974). Plant Physiol. 53, 533-542. Schwartzbach, S. D., Schiff, J . A., and Goldstein, N. H. (1975). Plant Physiol. 56, 313-317. Selldtn, G., and Selstam, E. (1976). Physiol. Plant. 37, 35-41.

CHLOROPLAST DEVELOPMENT

189

Senger, H., and Bishop, N. I. (1972). Plant Cell Physiol. 13, 633-649. Senger, H., Bishop, N. I.. Wehrmeyer, W., and Kulandaivelu, G. (1974). Proc. Int. Congr. Photosynrh. Res., 3rd pp. 1913-1923. Senger, H., Klein, O., Domemann, D., and Porra, R. T. (1980. In “The BlueLight Syndrome” (H. Senger, ed.), pp. 541-551. Springer-Verlag. Berlin and New York. Sestlk, Z., Zima, J., and Strnadovh, H. (1977). Photosynthetica 11, 282-290. Shah, S. P. J., Rogers, L. T., and Goodwin, T. W. (1969). Prog. Phorosynrh. Res. 2, 662-668. Shen, Y-K., and Hung, Y-C. (1964). Sci. Sin. 13, 1667-1674. Shihira-Ishikawa, I., and Hase, E. (1965). Plant Cell Physiol. 6, 101-110. Shumway, L. K., Weier, T. E., and Stocking, C . R. (1967). Planra 76, 182-189. Simpkins, J., and Street, H. E. (1970). J. Exp. Bor. 21, 170-185. Simpson, D. J. (1978). Carlsberg Res. Commun. 43, 145-170. Sironval, C., Bronchart, J. M. M., Bronchart, M., and Kuyper, Y. (1968). Bull. SOC.Fr. Physiol. V i g . 14, 195-225. Slater, E. C., Rosing, I., and Mol, A. (1973). Biochim. Biophys. Acra 292, 534-553. Smillie, R. M. (1962). Plant Physiol. 37, 716-721. Smillie, R. M., and Krotkov, G. (1959). Can. J. Bot. 37, 1217-1225. Smillie, R. M., and Krotkov, G. (1960). Can. J. Bor. 38, 31-49. Smillie, R. M., Evans, W. R., and Lyman, H. (1963). Brookhaven Symp. Biol. 16, 89-108. Smith, J. H. C., and Young, V. M. K. (1956). In “Radiation Biology” (A. Hollaender, ed.), Vol. 3, pp. 393-442. McGraw-Hill, New York. Sokawa, Y., and Hase, E. (1967). Plunr Cell Physiol. 8, 495-508. Soll, J., Douce, R., and Schultz, G. (1980a). FEBS Lett. 112, 243-264. Soll, J., Kemmerling, M., and Schultz, G. (1980b). Arch. Biochem. Biophys. 204, 544-550. Solomos, T., Malhotra, S. S., Prasad, S., Malholtra, S. K., and Spencer, M. (1972). Can. J . Biochem. 50, 725-737. Sprey, B. (1975). Protoplasma 84, 197-203. Sprey, B., and Laetsch, W. M. (1975). Z. Pflunzenphysiol. 75, 38-52. Sprey, B., and Laetsch, W. M. (1978). Z. Pflunzenphysiol. 87, 37-54. Steer, B. T., and Walker, D. A. (1965). Plant Physiol. 40,577-581. Steer, M. W., Gunning, B. E. S., Graham, T. A,, and Carr, D. J. (1968). PIanta 79, 254-267. Stetler, D. A. (1973). Bor. Gar. 134, 290-295. Stetler, D. A., and Laetsch, W. M. (1969). Ann. J. Bot. 56, 260-270. Steup, M., and Ssymank, V. (1978). Ber. Drsch. Bot. Ges. 91, 243-250. Strotmann, H., and Heldt, H. W. (1969). In “Progress in Photosynthesis Research” (H. Metzner, ed.), Vol. 3, pp. 1131-1140. Laupp, Tubingen. Strotmann, H., and Murakami, S. (1976). In “Encyclopedia of Plant Physiology, New Series” (C. R. Stocking and U. Heber, eds,), Vol. 3, pp. 398-416. Springer-Verlag. Berlin and New York. Stumpf, P. K., Brooks, I., Galliard, T., Hawke, 3. C., and Simoni, R. (1967). In “Biochemistry of Chloroplasts” (T. W. Goodwin, ed.), Vol. 2, pp. 213-239. Academic Press, New York. Swanson, J., and Floyd, G. L. (1979). Phycologia 18, 362-368. Tamas, I. A., Atkins, kl. D., Ware, S. M., and Bidwell, R. G. S. (1972). Can. J . Bot. 50, 1523-1527. Tarantowicz-Marek, E., and Kleczkowski, K. (1978). Planr Sci. Left. 13, 121-124. Tarchevsky, I. A,, and Konjukhova, T. M. (1981). Proc. Int. Congr. Phorosynth.. 5rh, pp. 731-736. Tevini, M. (1977). In “Lipids and Lipid Polymers in Higher Plants” (M. Tevini and H. K. Lichtenthaler, eds.), pp. 121-145. Springer-Verlag. Berlin and New York. Thomson, W. W., Dugger, W. M., and Palmer, R. L. (1965). Bor. Gar. 126, 66-72. Thomson, W. W., Dugger, W. M., and Palmer, R. L. (1966). Can. J . Bot. 44, 1677-1682.

190

A. R. WELLBURN

Thornber, J. P. (1975). Annu. Rev. Plant Physiol. 26, 127-158. Threlfall, D. R . , and Griffiths, W. T. (1967). In “Biochemistry of Chloroplasts” (T. W. Goodwin, ed.), Vol. 2, pp. 255-271. Academic Press, New York. Threlfall, D. R., Griffiths, W. T., and Goodwin, T. W. (1967). Biochem. J. 103, 831-851. Threlfall, D. R., Whistance, G. R., and Goodwin, T. W. (1968). Biochem. J . 108, 24-25. Tollenaar, D. (1925). “Omzettingen van Koolhydraten in het blad van Nicotiunu tubacum L.” Veenmann, Wageningen. Trebst, A. (1974). Annu. Rev. Plunt Physiol. 25, 423-458. Treffrey, T. (1970). Pluntu 91, 279-284. Treffrey, T. (1974). Eur. J . Biochem. 43, 349-351. Treffrey, T. (1975). Pluntu 126, 11-17. Treffrey, T. (1976). Plunt Sci. Lett. 6, 193-196. Treffrey, T. (1978). Int. Rev. Cytol. 52, 159-196. Valanne, N., and Valanne, T. (1972). Can. J . Bor. 50, 1835-1839. Vecher, A. S., Lebedeva, T. I., and Raitsina, G. 1. (1967). Dokl. Akud. Nuuk B . S . S. R. 11, 727-730. Vernon, L. P., and Shaw, E. R. (1969). Plunt Physiol. 44, 1645-1649. Vesk, M., Mercer, F. V., and Possingham, J. V. (1965). Aust. J. Bot. 13, 161-169. Vignais, P. V., Vignais, P. M., and Defaye, C. (1973). Biochemistry 12, 1508-1519. Vionilo, V. A., Deeva, V. P., and Mashtakov, S. M. (1967). Dokl. Akud. Nuuk. B. S . S . R. 11, 930-937. Voskresenskaya, N. P. (1972). Annu. Rev. Plant Physiol. 23, 219-234. Wagner, E. (1977). In “Integration of Activity in the Higher Plant” Soc. Exp. Biol. Symp., 31st pp. 33-72. Walker, D. A. (1976). In “Encyclopedia of Plant Physiology, New Series”, (C. R. Stocking and U. Heber, eds.), Vol. 3, pp. 85-136. Springer-Verlag, Berlin and New York. Walles, B. (1966). Hereditus 56, 131-136. Wanka, F. (1975). Colloq, In?. C. N . R. S . 240, 131-136. Watanabe, M . , Oh-hama, T., and Miyachi, S. (1980). In “The Blue Light Syndrome” (H.Senger, ed.), pp. 332-343. Springer-Verlag, Berlin and New York. Wellburn, A. R. (1968). Phytochernistry 7, 1523-1528. Wellburn, A. R. (1977). Pluntu 135, 191-198. Wellburn, A. R. (1978). In “Chloroplast Development” (G. Akoyunoglou and J. H. ArgyroudiAkoyunoglou, eds.), pp. 837-841. Elsevier, Amsterdam. Wellburn, A. R., and Hampp, R. (1976a). Biochem. J. 158, 231-233. Wellburn, A. R., and Hampp, R. (1976b). Plunru 131, 95-96. Wellburn, A. R . , and Hampp, R. (1976~).PIuntu 131, 17-20. Wellburn, A. R., and Hampp, R. (1979). Biochim. Biophys. Acru 547, 380-397. Wellburn, A. R., and Hemming, F. W. (1967). In “Biochemistry of Chloroplasts’’ (T. W. Goodwin, ed.), Vol. 1, pp. 174-180. Academic Press, New York. Wellburn, A. R., and Wellburn, F. A. M. (1973a). AnnBot. 37, 11-19. Wellburn, A. R. Quail, P. H., and Gunning, B. E. S. (1977). Pluntu 134, 45-52. Wellburn, A. R., Robinson, D. C., and Wellburn, F. A. M. (1982). Plunru, 154, 259-265. Wellburn, F. A. M., and Wellbum, A. R. (1971). J. CellSci. 9, 271-287. Wellburn, F. A. M., and Wellburn, A. R. (1973b). New Phyrol. 72, 55-60. Wellbum, F. A. M., and Wellburn, A. R. (1979). Pluntu 147, 178-179. Wellbum, F. A. M., and Wellburn, A. R. (1980). In “Endocytobiology” (W. Schwemmler and H. E. A. Schenk, eds.), pp. 839-848. De Gruyter, Berlin. Wellburn, F. A. M., Wellburn, A. R., and Senger, H. (1980). Protoplusmu 103, 35-54. Wettstein, D., von. (1959). Brookhaven Symp. Biol. 16, 123-160.

CHLOROPLAST DEVELOPMENT

191

Wettstein, D., von, and Kahn, A. (1960). Proc. Eur. Conf. Electron Microsc. D e l f t , 1051-1054. Whatley, J. M. (1971). New Phytol. 70, 1095-1102. Whatley, J. M. (1974). New Phytol. 73, 1097-1110. Whatley, J. M. (1977). New Phytol. 78, 407-420. Whistance, G . R., and Threlfall, D. R. (1967). Biochem. Biophys. Res. Commun. 28, 295-301. Whistance, G. E., and Threlfall, D. R. (1968). Biochem. J . 109, 482-483. White, J. M., and Pike, C. S. (1974). Plant Physiol. 53, 76-79. WiGckowski, S. (1969). Acta SOC. Bor. Pol. 38, 103-114. Wild, A. (1978). In “Chloroplast Development” (G. Akoyunoglou and J. H. ArgyroudiAkoyunoglou, eds.), pp. 553-538. Elsevier, Amsterdam. Wildman, S. G . , Hongladarom, T., and Honda, S. I. (1962). Science 138,434-436: See also 16-mm sound film “Organelles in Living Plant Cells,” Educational Film Sales and Rentals, University of California, Berkeley. Wildrnan, S. G., Jope, C., and Atchison, B. A. (1974). Plunr Physiol. 54, 231-237. Wirtz, H., Stitt, M.,and Heldt, H. W. (1980). Planr Physiol. 66, 187-193. Witt, H. T., Rurnberg, B., and Junge, W. (1968). In “Biochemie des Sauerstoffs” (B. Hess and H. J. Standinger, eds.), pp. 262-306. Springer-Verlag. Berlin. Wolf, F. T., and Kidd, G. H. (1973). Z. Pflanzenphysiol. 70, 115- 118. Wolff, B., and Schantz, R. (1978). In “Chloroplast Development” ( G . Akoyunoglou and J. H. Argyroudi-Akoyunoglou, eds.), pp. 245-250. Elsevier, Amsterdam. Wood, A., and Paleg, L. G. (1974). Aust. J . Planr Physiol. 1, 31-40. Woodstock, L. W., and Toole, V. K. (1977). Planr Cell Physiol. 18, 1-8. Woiny, A., Gwbidi, E., and Szweykowska, A. (1973). Protoplasma 79, 109-114. Wrischer, M. (1966). 2. Pflanzenphysiol. 55, 296-299. Wrischer, M. (1973a). Protoplasma 78, 291-303. Wrischer, M. (1973b). Protoplasma 78, 417-425. Wrischer, M. (1978). Protoplasma 97, 85-92. Yakushkina, N. I . , and Pushkina, G. P. (1971). Fiziol. Rust. 18, 898-903. Yomo, H., and Varner, J . E. (1971). Curr. Top. Dev. Biol. 6, 11 1- 144. Zaitlin, M., and Jagendorf, A. T. (1960). Virology 12, 477-486.

NOTE ADDEDI N PROOF. Unfortunately the last three values for 3-PGA on Fig. 4 have been incorrectly plotted. The correct values for 5.25, 6, and 7 cm up the laminae are 0.17,0.23,and 0.35 kmoles 3-PGA . mg protein - I , respectively, indicating that there is a later increase in triose phosphate content toward the tip of the barley primary leaf.

This Page Intentionally Left Blank

INTERNATIONAL REVIEW OF CYTOLOGY. VOL. 80

The Biosynthesis of Microbodies (Peroxisomes, Glyoxysomes) H. KINDL Biochemie (Fachbereich Chemie), Philipps-Universitat, Marburg, Federal Republic of Germany I. Introduction

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

11. General Concepts of Organelle Biosynthesis . . .

111.

IV.

V.

VI.

A. Pathway via ER and Subsequent Segregation of Vesicles. . . . . 9 . Import from Cytosol (Organelle Biosynthesis via Cytosolic Precursor Pools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ............ C. De Novo Synthesis o ............ D. Do Transient Forms Survey Obtained by in Vivo Studies ................... teins . . . . . . . . A. Expression of Genes 9. Site of Biosynthesis of Microbody Proteins.. . . . . . . . . . . . . . . C. Participation of the ER . . . . . . . . . . . . . . . . D. Modification and Aggregation. . . . . . . . . . . . . . . . . . . . . . . . . . . Single Steps of Assembly Stud A. Poly(A+)-mRNA Coding B. Synthesis on Free Polysomes ............... C. Products of in Virro Trans D. Import of Proteins into Microbodies.. .................... E. Chemical Modification and Changes in Hydrophobicity and Aggregation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Special Types of Cells .................... A. Animal Cell . . . . . . . . . . . . . . . . . . . . . . . . . B. Differentiated Microbodies in Plant Cells, Fungi, and Algae. . ..................................... Conclusion ................ References . . . . . . . . . . . . . . . . . . .

193 194 195 197 199 201 202 202 204 208

211 212 2 13 214 214 216 224 224

I. Introduction The biosynthesis of microbodies was of interest ever since these organelles were detected in distinctive cells such as kidney (Rhodin, 1954), liver (DeDuve and Baudhuin, 1966), cotyledons (Mollenhauer er al., 1966), and protozoa (Muller et af., 1968). The aspect of biosynthesis became especially intriguing when this process appeared to be one of the main events in the differentiation of a particular type of cells, e.g., the enhancement of microbodies in liver cells after feeding of hypolipidemic drugs (Lazarow and DeDuve, 1976; Moody and Reddy, 1976) or the drastic increase of microbodies in leaf cells after irradition (Tolbert, 197 1 , Gerhardt, 1978). Interest in the principles of microbody bioI93 Copyright 0 19R2 hy Academic Press. Inc. All right5 of reproduction in any lomi reserved. ISBN 0-12-364480-0

194

H. KINDL

synthesis was reinforced by the desire to test the concept of signal hypothesis (Goldman and Blobel, 1978) and by speculations on how glycoproteins are brought into various compartments (Neufeld, 198I ) . Microbodies can function in very different modes in a wide variety of cells. Although they participate in distinct metabolic pathways, there are enough common features to permit classifying them as an entity. In terms of morphology, microbodies are characterized as spherical or dumbbell-shaped organelles (Hruban and Rechcigel, 1969; Lazarow er al., 1980; Schopfer er al., 1976), delimited by a single membrane. The size can vary from 0.5 to 1.2 ym (Vigil, 1973; Hruban and Rechcigl, 1969). But microperoxisomes (Novikoff et al., 1973) with diameters of 0.2 ym were found in many cells, and on the other hand tubes of several micrometers of length must also be considered. The matrix of the organelles is usually granular and sometimes contains ordered structures (Frederick et al., 1975) which are labeled “crystalline.” These inclusions most likely consist of one or two dominating proteins and probably of membranous material. With respect to biochemical properties, microbodies are characterized by at least one oxidase reducing 0, to H,O,. This production of peroxide, and its employment by catalase-catalyzed reactions, led to the term peroxisome (DeDuve and Baudhuin, 1966). We use the terms microbody and peroxisomes as synonyms, but will specify liver peroxisomes, leaf peroxisomes, glyoxysomes, when specialized forms are meant. Another more general feature of microbodies is their relatively high equilibrium density when separated on sucrose gradients (Beevers, 1979; Kind1 and Kruse, 1982; Huang et al., 1982). Besides H,O, production and its use for oxidative steps, peroxisomes can carry out such diverse processes as .supplementing fatty acid oxidation in liver (Leighton et al., 1982) or proximal tubulus of kidney; heat production in brown fat (Cannon et al., 1982); conversion of fat into building material for gluconeogenesis in specialized cells of plants (Beevers, 1979), in some fungi, and algae; oxidation of methanol (van Dijken et al., 1982); oxidative steps in purine metabolism (Tolbert, 1982); or a major part of the photorespiration in green leaves (Tolbert, 1982). Despite the differences in the metabolic role of microbodies, we consider them here as organelles whose components are synthesized and assembled by the same principal processes. This unifying concept, applying to liver peroxisomes as well as to glyoxysomes, relevant to increases in the microbody space within a cell as well as to the transition from one microbody species to another, will be discussed in relation to the biosynthesis and assembly of other organelles.

11. General Concepts of Organelle Biosynthesis Ribosomes are not associated directly with the peroxisome membranes. RNA may be bound in some indirect way to plant microbodies that are at the stage of

BIOSYNTHESIS OF MICROBODIES

195

fat mobilization (glyoxysomes) (Gerhardt and Beevers, 1969). Despite conflicting reports (White and Brody, 1974; Ching, 1970; Osumi et al., 1978) it is very unprobable that microbodies contain DNA. Detailed studies with homogenates from plants demonstrated that traces of DNA possibly contaminating microbodies originated from mitochondria and plastids (Douglass et a l . , 1973). This was evidenced by studying the molecular properties of the contaminating DNA. According to these observations, we imply that microbodies perform neither transcription nor translation. Microbody biosynthesis can thus be regarded as being dependent solely on the nucleuskytosol system. A genetic interplay, as found with mitochondria and chloroplasts, has not been considered. We assume, therefore, that compounds destined for microbodies must be synthesized at one of the two possible sites for this sort of protein synthesis: at polysomes associated with the ER, or at free polysomes in the cytosol. In the former case, the newly synthesized protein directly enters the luminal site of the ER (cotranslationally) and occurs at no time in the cytosol. When peptides are formed on free polysomes, however, a definite pool of these precursors must be established before the peptides are taken up selectively and probably processed by the respective organelles (posttranslationally). A. PATHWAYVIA ER

AND

SUBSEQUENT SEGREGATION OF VESICLES

Protein biosynthesis via ER is linked with the production of peptide precursors possessing an extra sequence (signal). Peptides with signals are only transient entities which already lose their extra sequence at the site of protein synthesis. The signal sequence binds during the protein synthesis to the ER membrane, and this binding is supported by ribosome-ER interactions (Warren and Dobberstein, 1978; Blobel et a l . , 1979; Meyer and Dobberstein, 1980a,b). During biosynthesis, the growing protein is vectorially discharged into the luminal site of the ER (Davis and Tai, 1980). It seems that at least all transmembrane proteins of organelles are synthesized at the ER and inserted into membranes in this way. They may then be sequestered into the membranes of vesicles by which they are transported to the target organelle, without being exposed to the cytosol. The sorting out of the precursors destined for a certain organelle is done on the ER membrane or on the luminal site of the ER. In the case of soluble proteins to be transferred from the ER lumen to the matrix of an organelle, we must assume that some kind of interaction exists between the soluble protein and part of the ER membrane which is exploited as vehicle for the transport to the organelle. There, too, some sort of recognition should govern the fusion of the transport vehicle or transport tubulus with the growing organelle. This model may well apply to describe the formation of a properoxisome originating at the ER and later fused with other microbody species. The recognition of an oligosaccharide chain by a lectin-type selector is certainly one possible mechanism of sorting out. A soluble protein in the ER lumen

196

H.KINDL

can only be sorted if it shows affinity toward a membrane constituent and a mechanism already functions that enhances the local concentration of the leading membrane protein in the budding vesicle. Another kind of selector could function like clathrin. Clathrin, found in different sorts of cells, is a component of coated vesicles that are involved in the budding from the plasma membrane, in the process of adsorptive endocytosis. Furthermore, proteins that have to be exported were found to be associated with clathrin during their transfer from Golgi to the plasma membrane. Clathrin acts as receptor for uptake of extracellular compounds, by selecting or excluding compounds. If a kind of sorting, as found prior to the endocytosis mediated by clathrin, is also operative at the ER, one might envisage not only a sorting of luminal proteins, to be surrounded by the specialized portion of the ER membrane, but also an enveloping of cytosolic molecules after being selected by protein-protein interaction on the ER surface facing the cytosol. Clearly, the sidedness of the outcoming vesicles must change. This sort of transfer implies that the vesicles are directed to the correct organelle. The target, therefore, must be able to sort out the correct transport vesicle. Membranes have to be recycled in cases where the organelles do not increase in volume. Phospholipid exchange proteins could also maintain this transfer of membrane components. In other cases, an increase in membrane lipids could parallel the increase in proteins. We find examples of organelle biosynthesis that can be thought to be related to the assembly of microbodies in diverse cells. 1. Lysosomal Enzymes Lysosomal enzymes are secreted, not only into the organelle but also into the extracellular space. Export of lysosomal proteins is found predominantly during diseases. The example of cathepsin D shows that precursors of lysosomal proteins are cotranslationally segregated into the lumen of ER (Erickson and Blobel, 1979). When porcine spleen mRNA was translated, a M, 43,000 peptide was obtained in the absence of ER vesicles. In the presence of ER, as a glycosylation agent, a M, 46,000 was detected in the ER as precursor which was eventually converted into the M, 30,000 mature protein. The pathways of lysosomal proteins, and probably peroxisomal components, too, have subsequently to diverge from that of secretory proteins. 2. Protein Bodies and Vacuoles During the stage of seed development, reserve proteins are accumulated and deposited in a part of the vacuolar system, the protein bodies. Ultrastructural studies showed that in legumes (Yo0 and Chrispeels, 1980) part of the central vacuole is used for the deposition of storage proteins. Then new protein bodies arise from the central vacuole by pinching-off small entities of deposited protein being surrounded by tonoplast membranes. In a much later stage, during ger-

BIOSYNTHESIS OF MICROBODIES

197

mination and seedling growth, the reserve proteins are catabolized within the protein bodies which eventually become part of the lytic compartment and fuse to form the central vacuole (Van der Wilden et al., 1980). Thus, protein bodies and vacuoles can be considered as related organelles whose components should be synthesized by a common pathway. Apart from the possibility that these lytic organelles take up and digest other organelles (Herman et a / . , 1981), they acquire their proteins by a pathway very similar to formation and transport of secretory proteins. A feature of such a mechanism is that the protein is transported cotranslationally, and thus a signal sequence cleavable in the ER should be found. Several studies with constituents of protein bodies failed to demonstrate higher molecular weight peptides when in vitro translation was performed in the absence of microsoma1 membranes (Croy et a/., 1980a,b). But some of these studies may have been hampered by the complicated sequence of glycosylation and protein processing, both during translation in the ER and later in the protein bodies (Badenoch-Jones et a/.. 1981). Most other studies are in agreement that during translation at the rough ER a size modification results from the cleavage of a signal sequence (Nelson and Ryan, 1980). But posttranslational modifications also take place and the result can be that we find a picture difficult to survey because df the many modifications. Roberts and Lord (198 la) had to face such a situation when they studied the biosynthesis of Ricinus communis agglutinin. Probably, a similar combination of cotranslational processing and post-translational modifications, in the ER and later in the protein bodies, will also be encountered in the biosynthesis of glycosylated storage proteins of legumes. Whether glycoprotein biosynthesis also proceeds via ER in those cases in which reserve proteins are not stored in the protein bodies but rather, during seed germination, are degraded, is still an unsolved question (Kara and Kindl, I972b). All these modifications are to be taken into account when we discuss biosynthetic sequences leading to microbodies. B. IMPORT FROM CYTOSOL(ORGANELLE BIOSYNTHESIS VIA CYTOSOLIC PRECURSOR POOLS) Synthesis of organellar proteins on free cytosolic polysomes has been studied, with mitochondria as well as with chloroplasts. These examples (Chua and Schmidt, 1979), and new findings on the biosynthesis of components of the ER, are very suited to discussing and explaining mechanisms considered in the field of microbody biosynthesis. 1 . tmport and Assembly of Mitochondria1 Proteins In mitochondria we have to distinguish, in terms of processes obligatory during import, between several categories of proteins: components of the outer

198

H. KINDL

membrane, subunits of inner membrane protein complexes which consist of cytosol-made and mitochondrion-made components, inner membrane proteins imported from cytosol, and proteins of the matrix being synthesized in the cytosol. As examples of the first category, the biosyntheses of cytochrome c peroxidase and porin have been investigated. Both are formed at free cytosolic ribosomes. Extensive studies are being made aiming at the elucidation of assembly of the multicomponent enzymes cytochrome c oxidase and ATP synthase (Lewin et al., 1980). The cytoplasmically made subunits of ATPase and cytochrome c oxidase are made as larger precursors. Earlier reports about polyprotein precursors could not be confirmed (Lewin et a l . , 1980). As for a component of the intermembrane space, it has been established conclusively (Henning and Neupert, 1981) that apocytochrome c is synthesized at free polysomes with the same M, as the mature enzyme; it reaches the mitochondria as apoprotein, is bound by a specific receptor, and then deposited in the inner membrane as holocytochrome c. Hydrophilic holocytochrome c does not leave the mitochondria nor can it be imported as such by the mitochondria, it is present only at the site of its function. The ADP/ATP carrier (Zimmermann et a / . , 1979) as hydrophobic integral protein of the inner membrane is synthesized as polypeptide with the same molecular weight as the mature monomeric protein. Whereas the manner of import of cytochrome c may have analogies with the biosynthetic processes of catalase assembly in peroxisomes, the ADP/ATP-carrier and the hydrophobic proteins of glyoxysome membrane, e.g., malate synthase, can be compared in terms of changes in tertiary structure and solubility. Both proteins, although exhibiting their hydrophobic nature when associated with the organelle’s membrane, have to pass the hydrophilic environment of the cytosol when they get from the polysomes to the organelle. Citrate synthase destined for the matrix of mitochondria is synthesized on free polysomes as a precursor with a greater M, than the mature enzyme (Harmey and Neupert, 1979). Earlier reports that mitochondrial proteins might be synthesized on rough ER or by ribosomes bound to the outer mitochondrial membrane do not seem to be relevant with respect to microbody biosynthesis. Ribosomes associated with microbody membranes have not yet been found. A vectorial discharge of proteins from ribosomes directly into the organelle is, therefore, not likely.

2 . Import of Plastidic Proteins Integral membrane proteins facing the thylakoid lumen have to pass two membranes (of the envelope) and, more or less, a third one (thylakoid membrane). Plastocyanine, part of the P,, center, and the light harvesting protein are candidates for this sort of transfer processes. Chua et al. (1980) studying the

BIOSYNTHESIS OF MICROBODIES

199

import of thylakoidal proteins used intact chloroplasts and in vitro synthesized precursor proteins and showed that posttranslational processing is a general feature of the energy-dependent import. Work of Ellis (1981) contributed extensively to this field. While components of the inner membranes may not have their equivalent in the case of microbodies, the biosynthesis of stroma enzymes, possibly arranged at or near the thylakoid membranes, should bear great similarities to the formation and uptake of glyoxysomal matrix proteins. This holds true for the small subunit of ribulosebisphosphate carboxylase (Chua and Schmidt, 1979; Ellis 1981). 3 . Import at the ER Membrane Although protein synthesis can be performed on the ER by its “own” machinery, the bound polysomes, some constituents are nevertheless taken up from the cytosolic pool. The syntheses of the ER constituents cytochrome b, and NADH:cytochrome b, oxidoreductase occur on free polysomes (Okada er al., 1982). If NADH:cytochrome b, oxidoreductase indeed is also a component of the glyoxysomal membrane (Donaldson et a l . , 1981), we suppose that the transfer from cytosol to the microbody membrane can take place either directly or via uptake by ER membrane and subsequent segregation. In both cases, the biosynthesis of this component of the microbody membrane does not include a cotranslational transport. Similar considerations apply for cytochrome b,. These proteins being exposed on the cytoplasmic surface of the ER membrane are anchored to the membrane by their carboxy-terminal domains, which is opposite to the orientation of some other ER membrane proteins, e.g., cytochrome P-450 reductase with its active site also being accessible from the cytosol was found to be embedded in the membrnae by its amino-terminal sequence (Black et al., 1980). Biosynthetic studies revealing whether an intermediary pool of proteins manufactured in the cytosol and destined for microbodies exists at the ER would have great implication for the mechanism of microbody biosynthesis. They could be used as a model for the idea that outgrowing buds are the main site of the microbody assembly.

C. De NOVO SYNTHESIS OR FISSION AND FUSION

In many instances, we can assume that small amounts of microbody-like structures, proforms, or small but differentiated species are already present in cells before a certain stimulus turns on microbody biosynthesis. Thus, we anticipate that a constantly growing organelle can be propagated and can furnish “seeds” or single organelles for the new cells by fission and fusion (Rigatuso et al., 1970). A discussion of de novo synthesis of a microbody in a cell that does

200

H. KlNDL

not already contain microbody-like structures presupposes a detailed knowledge of the amphipathic character of all microbody membrane proteins and an understanding of self-assembly of membranes, topics beyond the scope of this article. For fission and fusion, the most likely processes occurring in propagation of microbodies, it is reasonable, but only incompletely proven (Legg and Wood, 1970; Yokota and Nagata, 1974), that transfer between organelles occurs by vesicles budding from one membrane and fusing with another. Newly synthesized protein may be transferred from the ER to microbodies or, equally likely, small proforms grow by uptaking cytosolic precursors and fusing then with other members of this family. If the hydrophobic nature of the appropriately folded protein only causes the insertion of the protein into membrane, an incorporation of the cytosolic protein is imaginable at the ER as well as at the outbudding ER or the microbody itself. But fission and fusion and traffic between the ER and microbodies would be salient features of the microbody biosynthesis when most of its proteins were synthesized at the ER. Fission and fusion, at the level of microbodies themselves, are necessary only during cell division or during pronounced proliferation of this organelle. In yeasts (Osumi et al., 1975) and in green algae (Floyd, 1972) fission of microbodies has been observed. Most impressive is fission during division of cells which are already distinguished by a microbody as dominating organelle (van Daijken et al., 1982). Polydispersity of peroxisomes, and smaller vesicles also containing cyanideinsensitive P-oxidation enzymes, was found in rat liver (Poole et d . , 1970; Flatmark et al., 1981, 1982) following treatment with peroxisomal proliferators. Normal rat livers contain only a single population of peroxisomes, whereas in liver homogenates from clofibrate-treated rats three classes of peroxisome population (12,000 S, 4,000 S, 1,000 S) were found. A model of properoxisomes already containing the P-oxidation enzymes, but not having reached the size of the main population of peroxisomes, was put forward. According to this the proorganelles increase by size and content when additional components are taken up from the cytosol. This conception could also be valid for catalase-containing structures observed in etiolated leaves (Gruber et al., 1973; Feierabend and Beevers, 1972a), in ripening seeds (Kind1 et al., 1980a), or plant suspension cultures (Kudielka et a / ., 1982) during intensified synthesis of organelles. External effectors, as hypolipidemic drugs in liver cells or light in plant cells, may serve a. a cause producing a definite effect on the extent of fission of already existing microbodies, thus providing sufficient starting points for a growth of the peroxisomal continuum (Lazarow et al., 1980). It remains an open question whether the concept of microperoxisomes described for almost all types of animal cells (Novikoff et al., 1973) should also be applied to plant cells in transition stages.

20 1

BIOSYNTHESIS OF MICROBODIES Stage

I

I

Transition stage

I

!Stimulus I

I

-

Stagen

I I I

I

Precursor pool

FIG. I . One-population model. Scheme illustrating the alterations in organelles and precursor pools during transition from a cell with microbodies at stage I to a cell characterized by microbodies at stage 11. The scheme includes one type of organelles only, namely, microbodies; all other changes in cell structure were not considered. 0,Typical components of microbodies at stage I; 0 ,enzymic inactive form; A, marker proteins of microbodies at stage 11. D, Degradation (turnover).

D. Do TRANSIENT FORMSOF MICROBODIES EXIST? We know of several physiological situations where the catalytic capacity of one type of microbodies is exchanged by a new set of enzymes likewise housed in microbodies. In such a case two modes can be envisaged on how one form of an organelle can be replaced by another. In the one-population model, only once is the entire organelle with its delimiting membrane formed. During the transition, the enzymatic activity of the first microbody species is turned off by some sort of degradation and the new set of components is transferred into the existing organelle. The two-population model implies that during the transition two independent populations of microbodies occur in the same cell. The amount of the former species decreases while a concomitant increase of the new organelle form takes place. The new microbody form should, therefore, be synthesized de novo. This model has, in its extreme form, far-reaching consequences, as the process of differentiation has to start from zero as far as the identity of organelles is

\

,/ I

Lytic compartment

that vesicles containing a few FIG. 2. Growth and turnover of microbodies. It is hypothesized .. membrane proteins are segregated from the ER and grow by uptake of cytosolic precursors. The illustration should emphasize that microbodies are part of a dynamic compartment and that all forms of microbodies, at any stage, are subjected to a constant turnover which included degradation by lytic organelles.

H.KINDL

202 Stage

1

l I

Transition

,

Stage

I

I

I I

IS

I I

I

FIG. 3. Two-population model. Two independent populations of microbodies are present during the transition stage. The new population grows as soon as a starting point is provided by selfassembly (S). Symbols as in Fig. 1.

concerned. There is no growing minivesicle that could be the target of transfer vesicles from the ER. This lack of starting points has some clear bearing upon the discussion of both the sorting out process in the ER and the selection carried out by soluble cytosolic precursors looking for their targets. Figures 1 and 3 outline the principal differences between the one-population hypothesis and the two-population hypothesis. The impression of a more static model, as given by the pronounced simplification of Fig. 1, has to be corrected in the way that every form of microbodies is, in addition to the processes shown in Fig. 1, in a steady state of organelle growth and degradation (Fig. 2). Orientation about the frequently occurring microbodies and brief, incomplete, information concerning their main constituents are given in Table I. 111. Survey Obtained by in Vivo Studies

First insights into the essential steps of organelle biosynthesis, i.e., the site of protein synthesis, the sequence of intermediary pools, and the form of final arrangement of proteins within the organelle, can be gained by in vivo studies. Surveying these processes by means of studies with intact cells, organs, or organisms permits us to design experiments which answer more specific questions. If these experiments are done under defined physiological conditions, the direction and intensity of the respective process under steady-state conditions can be revealed. A. EXPRESSION OF GENESCODINGFOR MICROBODY PROTEINS

Little is known about the arrangement of genes of microbody proteins and their transcription. The synthesis of glyoxysomal proteins seem to be repressed

203

BIOSYNTHESIS O F MICROBODIES

TABLE I EXAMPLES OF BIOCHEMICALLY CHARACTERIZED FORMSOF MICROBODIES A N D THEIRMAIN PROTEIN COMPONENTS Protein component

Prosthetic groups

Aggregation

FAD

Animal peroxisomesU Carnitine palmitoyl transferase Acyl-CoA oxidase Multifunctional protein Thiolase Catalase Urate oxidase

Heme FAD

Oligomer as micelle Heterooligomer Monomer Dimer Tetramer Octamer

Specialized peroxisomesb Alcohol oxidase Catalase

FMN Heme

Octamer Tetramer

Leaf peroxisomesc Glycollate oxidase Hydroxypyruvate reductase Malate dehydrogenase Serine-glyoxylate-aminotransferase Catalase Urate oxidase Glyoxysomesd Acyl-CoA oxidase Multifunctional protein Thiolase Malate synthase Citrate synthase Malate dehydrogenase Isocitrate lyase Catalase

-

FMN Pyridoxalphosphate

Tetramer, 16-mer Dimer Dimer Tetramer

Heme FAD

Tetramer Octamer

FAD

Heterooligomer Monomer Dimer Octamer, Trimer Dimer, Tetramer Dimer, Hexamer Tetramer Tetramer

-

-

-

Heme

UExample: liver peroxisomes. Metabolic capacities: fatty acid p-oxidation, peroxidative processes. bExample: peroxisomes from methanol-grown yeasts. Metabolic capacities: oxidation of methanol. EExamples: Microbodies in green leaves of plants, in algae under photoautotropic conditions. Metabolic capacities: photorespiration, urate degradation. dExamples: Microbodies in plants during degradation of storage lipids; fungi or algae grown on acetate; Tetruhyrnenu. Metabolic capacities: fatty acid p-oxidation, glyoxylate cycle.

by glucose. This was unequivocally demonstrated for fungi (Maxwell et al., 1975) and algae (Dunham and Thurston, 1978, 1980). Catalases have been shown to be absent from anaerobically grown cells of Saccharomyces cerevisiae

204

H. KINDL

and to be induced by oxygen (Zimniak et al., 1975). Oxygen controls levels of catalases by affecting heme formation (Ruis, 1979). Mutants have been obtained which allow the investigator to study these aspects in more detail. With Candida boidinii, induction of formation of peroxisomal alcohol oxidase was observed with methanol, whereas glucose or ethanol led to repression (Roggenkamp er al., 1975). In yeast (Saccharomyces cerevisiae), two genes for catalase have been found (for summary see Ruis, 1979), but none of the two isoenzymes may be located in microbodies. In the case of Zea mays, too, mutants are available; with them it could be shown that more than one gene is responsible for the expression of catalase activity. Control of catalase synthesis is lost by a switch in a regulatory gene (Tsaftaris and Scandalios, 1981). The gene product of Cat 1 is already formed during seed formation while Cat 2 gene directs the synthesis of the catalase that is found to increase up to the fourth day of germination. When Cat 2 gene is hit by mutation, only residual amounts of catalase, steming from Cat 1, are present during germination. On the other side, an overproduction of catalase was observed when Cat 2 was not more under the control of a distinct regulatory gene (Car). Whether catalase 1 and catalase 2 differ with respect to intracellular or intraorganellar location is not clear. It is highly probable that phytohormones modulate the gene expression, and of microbody proteins, by alteration of mRNA pools. Gibberellic acid (Gonzalez and Delsol, 1981), abscissic acid (Radin and Trelease, 1976; Choinski et al., 1981), and kinetine (Theimer et al., 1976; Naito er al., 1980; Lampugnani er al., 1980) were found to influence the level of enzyme activities. Conclusions drawn with respect to the level of translatable mRNA have been substantiated by in vitro studies (Martin and Northcote, 1982). That selective effectors affecting the gene expression of one glyoxysomal protein do not exert control on the synthesis of other glyoxysomal proteins, was shown by Khan and McFadden (1 979). They used itaconate which interfered with the developmental time course of isocitrate lyase, but had no effect on the development of malate synthase or catalase. B . SITEOF BIOSYNTHESIS OF MICROBODY PROTEINS As outlined in Sections II,A and B, it is highly probably that microbody enzymes are synthesized either at free polysomes or at polysomes bound to the ER. Accordingly, one should be able to detect precursors of the organellar proteins either in the cytosol or in the ER. It is then feasible to distinguish between these two alternatives by feeding radioactive amino acids to organisms under in vivo conditions. A subsequent cell fractionation followed by isolation of the respective protein or its precursor permits us to attribute the label either to the soluble, i.e., cytosolic fraction, or to the ER, membrane, or lumen. Investiga-

BIOSYNTHESIS OF MICROBODIES

205

tions of this kind are intrinsically hampered by the fact that during cell desintegration and cell fractionation a significant and sometimes varying portion of microbodies is destroyed, and matrix proteins of glyoxysomes or peroxisomes can thus be found in the soluble fraction, i.e., usually the supernatant of a density gradient. Such a complication is not encountered when proteins bound to the organelle's membrane are under investigation. After pulse labeling for a short period of time, rat liver was fractionated and the labeled microbody enzyme (catalase) was found only in the cytosol, not in the ER (Lazarow and DeDuve, 1973a,b). These studies have been extended (Lazarow er a/., 1982) and proved that labeled precursors of microbody proteins were absent from the ER. Albumin used as marker of the ER lumen was found highly labeled, but in microsomes only. With this marker, the extent of contamination of the supernatant by soluble proteins of the ER lumen can be determined (Lazarow et af., 1980). Redman er al., ( 1972) had already found that radioactive catalase appeared in the cytosol of liver cells 10 minutes after the rats received ['4C]leucine. At that time, these findings had to be seen in contrast to the work of Higashi and Peters (1963) who detected labeled catalase in the rough microsomal fraction. In plant cells, cytosolic precursor pools of microbody enzymes could be determined only recently (Kindl et al., 1980b). The time course of labeling in the cytosolic pool and later in the microbody pool was investigated for different stages of organelle differentiation (Koller and Kindl, 1980; Frevert et al., 1980). For three enzymes of the glyoxysomal matrix, catalase, isocitrate lyase (Frevert and Kindl, 1978), the multifunctional protein of the P-oxidation processing enoyl-CoA hydratase activity and 3-hydroxyacyl-CoA dehydrogenase activity (Frevert and Kindl, 1980), the sequence of pools cytosol + glyoxysomes was established. Soluble proteins, indicative of the ER lumen, were exploited for assessing how much ER constituents contaminate the fraction with soluble proteins. Like the preparation from liver, desintegration and fractionation of plant cells can also be performed without losing high proportions or the ER lumen (Sturm and Kindl, 1982). Analogous results, i.e., cytosolic precursor pools, have now been obtained for a series of different plant microbody enzymes, glyoxysomal and peroxisomal (Kindl, 1982). The pool of isocitrate lyase in the ER was shown by Roberts and Lord (198 Ib) to be negligible. With a membrane-bound protein of glyoxysomes, i.e., malate synthase, it was possible to demonstrate quantitatively that pulse-labeled microbody proteins are initially detectable only in the cytosol and can be chased into the organelle only after a lag period. After some controversy about this point (see Beevers, 1982; Kindl, 1982) the data are now accumulating that malate synthase, although a peripheral protein of the glyoxysomal membrane, is formed via cytosolic pools. Malate synthase is bound to the glyoxysome membrane as peripheral, and partly as integral protein (Kruse and Kindl, 1982), and does not lead to a

206

H . KINDL

contamination of cytosolic fractions, when leakage or organelles takes place during separation procedures. The criterion of Zilversmit et al. (1943) based on specific radioactivities is therefore fully applicable and allows a clear precursor-product correlation for the transfer of malate synthase from cytosolic pool into the glyoxysomal (Kindl, 1982). As the pool size of the precursor, i.e., methionine, is already very small, the form of the radioactivity profiles during the course of the experiment (up to 4 hours) is not altered markedly whether a pulse or a pulse-chase experiment is performed. With the exception of malate dehydrogenase, and probably citrate synthase (Desel er al., 1982; and for the mitochondrial form: Harmey and Neupert, 1979) and catalase, all cytosolic precursor forms seem to lack additional peptide sequences (Kindl, 1982). Malate dehydrogenase transferred from cytosol into the glyoxysomes undergoes an interesting processing which is paralleled by a change in M, (see Section IV,D).

C. PARTICIPATION OF THE ER Occasionally observed continuities between the ER and microbodies, comparison of the polypeptide profiles of microbody membranes and the ER, and the hypothesis that eventual glyoxysomal glycoproteins would have to be synthesized at the site of glycosylation, i.e., the ER, seem to argue for a participation of the ER in the microbody biosynthesis (DeDuve, 1973). That could imply that almost all components are already present at the ER when the microbody is budding from the ER. Alternatively, it suggests that a “minimum” membrane vesicle is furnished by the ER, the main portion of proteins being taken up from the cytosol. All three arguments cited above have been questioned. While some investigators emphasize that continuities between the ER and microbodies are found (Novikoff and Shin, 1964; Novikoff et al., 1973; Vigil, 1973; Goeckermann and Vigil, 1975; Wanner and Theimer, 19821, others state that “no clear openings between the ER and peroxisomes have been detected,” and others contradict the fact that such continuities can be observed, e.g., in liver cells (Yokota and Fahimi, 1980). Thorough investigations (Mollenhauer and Totten, 1970; Gruber et al., 1973) did not provide unequivocal proof of continuities. Equally controversial is the question of whether the membranes of the ER and microbodies resemble each other. There is no question that the enzymes of the phospholipid biosynthesis are located on the ER and that the lipid constituents of the two membranes do not differ qualitatively and quantitatively. It is highly likely that the same kind of NADH:cytochrome b, oxidoreductase is present at the ER, the outer mitochondrial membrane, and the glyoxysomal membrane (Donaldson et al., 1972, 1981). Therefore, it is conceivable that identical antigenic determinants can be found on the ER and the glyoxysomal membrane (Hock, 1974).

BIOSYNTHESIS OF MICROBODIES

207

More complex is the discussion of experiments which were performed with ER preparations not being extensively purified by at least a second gradient especially suited to separate the ER from other light membranes and membrane fragments. In addition, in many types of cells dominating proteins exist which stick to almost every membrane and are, therefore, found throughout the gradient. Hence, it was not surprising when almost identical polypeptide profiles were demonstrated for glyoxysomal membranes and microsomal membranes (Bowden and Lord, 1976a,b; Brown er al., 1976). Even the use of several methods of analysis of this pattern does not bring us further, if the membranes were not previously extensively purified. Investigations by others (Kruse and Kindl, 1982; Sturm and Kindl, 1982) do not show comparable protein patterns when collating the dominating bands of glyoxysomal membranes and the ER membranes. Even the comparison of fluorographies of newly synthesized glycoproteins did not reveal any striking homology. This is surprising because we must assume that core glycosylation at least for the N-glycosides takes place at the inner surface of the ER membrane. The reason may be that only a minor amount of the cell’s glycoproteins is destined for glyoxysomes; or we hypothesize that glycoproteins are also built at the surface of the ER that faces the cytosol. Extensively purified membranes from the ER and peroxisomes were analyzed electrophoretically (Fujiki et a t . , 1982), and the peptide patterns differing very greatly from each other confirmed that for liver cells it is not possible to demonstrate the close relationship of the two membranes that was proposed earlier. Until now most investigators failed to demonstrate the occurrence of glycoproteins in rat liver peroxisomes (see discussion in Kindl and Lazarow, 1982), the participation of the ER in the biosynthesis of peroxisomal proteins in this tissue is not mandatory. In mouse liver, however, catalase seems to be a glycoprotein with sialic acid residues (Masters, 1982). Other catalases, e.g., the ones of yeast (Ammerer et al., 1981), are not glycosylated. A detailed discussion of the localization of glyoxysomal enzymes in the ER is required in the case of malate snythase and citrate synthase. Since Gonzalez and Beevers (1976) described an extra form of malate synthase, an additional peak of malate synthase activity in the gradient, close or coinciding with the ER marker, and appearing during the time period of maximal glyoxysome biosynthesis, these forms got more and more the role of precursors of the respective glyoxysomal proteins. A series of papers (Bowden and Lord, 1976a; Lord and Bowden, 1978; Mellor ef a / . , 1978; Lord, 1978, 1980) described the synthesis of malate synthase at the ER, its glycosylation there, and its ultimate sequestration into glyoxysomes. The question of whether or not malate synthase is a glycoprotein has invoked controversies. While Lord (1980) and Riezman et al., (1980) consider it a protein with a distinct glycomoiety , others disagree on that (Bergner and Tanner, 1981; Kindl, 1982). Disagreement also exists as to the localization of the malate

208

H. KINDL

synthase in the ER. Koller and Kindl already showed in 1978 that the coincidence of profiles, malate synthase activity, and ER markers is only fortuitous, as this form of malate synthase exhibits a sedimentation behavior similar to ER vesicles. Detailed investigations (Koller and Kindl, 1980) corroborated the lack of any close relation between this form of malate synthase and the ER. The fact that the reversible aggregation of malate synthase depends on the surrounding medium and the difficulties arising by contamination with components and fragments originating from protein bodies (Kara and Kindl, 1982b) may render separations more difficult. Furthermore, the ‘‘microsomal’’ form of malate synthase was shown not to be associated with phospholipid, but rather to represent a highly aggregated homomer. In addition, this aggregated malate synthase behaving, upon sedimentation velocity centrifugation, like microsomes, was rigorously ruled out to function as direct precursor of glyoxysomal malate synthase (Kindl et af., 1980b; Kindl, 1982). Other enzymes were also excluded as being synthesized via a precursor pool in the ER (Kindl er al., 1980a,b; Frevert et al., 1980). Renewed reports that citrate synthase and malate synthase could be localized at the ER (Kagawa and Gonzalez, 1981 ;Gonzalez, 1982) are difficult to interpret since a survey of all forms of the respective enzyme was not provided but rather the main fractions were eliminated by differential centrifugation or prolonged velocity sedimentations. Although it is not excluded that small amounts of citrate synthase or malate synthase, both exhibiting rather alkaline isoelectric points, are associated with the ER, such an observation should not have a bearing on the pathway of glyoxysome biosynthesis. D. MODIFICATION AND AGGREGATION Some of the proteins that have been localized in the matrix of microbodies may not necessarily affect their enzyme activity there, but may function as catalysts only when they are aggregated to a defined state of oligomerization, or when they coaggregate with consecutive enzymes and channel their intermediates, or when they were associated selectively with single molecules of amphipathic lipids, or with the membrane. Therefore, for every microbody enzyme the mode of structural arrangement within the organelle has to be considered, and the process of transfer of the newly imported peptide in the correct spatial arrangement must be seen as part of the total process of biosynthesis and assembly. One of the intriguing examples revealing a complicated sequence of processes during the course of enzyme assembly is the biosynthesis of catalase. Lazarow and DeDuve (1973a,b) presented evidence, by in vivo investigations, that the apomonomer is the entity that is transferred into the organelle, and heme is acquired prior to oligomerization. Zimniak et al. (1975) also detected an apomer during the biosynthesis of catalase T in S. cerevisiae. Comparing the M, of the

BIOSYNTHESIS OF MlCROBODIES

209

two isoenzymes from S . cerevisiae and from Neurospora crassa (Jacob and Orme-Johnson, 1979) we realize remarkable differences. Whether enzymatically active catalase, a tetramer, is attributable also to the cytosol is still an unsolved question (Roels et al., 1982; Kindl et al., 1980a). When proteins lacking a prosthetic group were examined, enough data were provided to prove that functional active enzymes, oligomers in most instances, are also present in the cytosol. But it is very likely that the monomeric cytosolic forms are actual precursors of microbody proteins whereas oligomeric cytosolic forms originate from these precursors only when the precursors are piled up, the latter process facilitating modifications. The cytosolic pools of malate synthase and isocitrate lyase were analyzed by sedimentation velocity centrifugation. Monomeric and oligomeric forms of malate synthase were detected besides aggregates. The highest radioactivity after in vivo labeling was attributable to the monomeric form. If the labeling was longer than 30 minutes, the presence of the octamer of malate synthase was verified (Koller and Kindl, 1980). With short pulses of radioactive amino acids, the cytosolic monomers of isocitrate lyase and malate synthase were the exclusive products of de novo synthesis in vivo. The newly synthesized cytosolic monomers were found not to be capable of oligomerization or aggregation. On this basis, it is possible to distinguish between the precursor form and the modified form which can be a monomer or an oligomer. At present, it is not known how malate synthase and isocitrate lyase are processed. But both the change of capability to oligomerize or aggregate, and the fact that monomeric precursor and mature enzyme exhibit different affinities toward amphipathic lipids (Zimmermann and Neupert, 1980; Kruse and Kindl, 1980) are hints that refolding of the monomeric protein or minor chemical modifications must take place on the way from the polysomes to the fully assembled structure within the organelle. Both the homogeneous monomeric malate synthase prepared from glyoxysomes and the monomeric malate synthase obtained from the 100 S aggregated cytosolic form were quantitatively shifted into the 100 S aggregate. The newly synthesized cytosolic form, however, was not amenable to oligomerization (Kruse and Kindl, 1980). It is noteworthy that yeast malate synthase is a trimer (Durchschlag er al., 1981) in contrast to the octameric forms observed in plants. More significant changes in terms of size, and thus easier detectable differences upon electrophoresis, characterize the biosynthesis of microbody malate dehydrogenase. Gietl and Hock ( 1982) demonstrated by pulse-chase experiments that the precursor of the glyoxysomal malate dehydrogenase (subunit M,41,000) was processed to the significantly smaller form of mature glyoxysomal enzyme (subunit M, 33,000). In this respect, the mitochondria1 form and one of the cytosolic isoenzymes of malate dehydrogenase differ from the glyoxysomal form.

210

H . KINDL

In the case of catalase, too, forms of different molecular size were established (Lazarow et al., 1980, 1982; Kindl et al., 1980b); but the appearance of multiple forms is most likely caused by the susceptibility of catalases to proteolytic degradation. Mainferme and Wattiaux (1982) provided evidence that M,64,000 corresponds to the peroxisomal enzyme, while the M, 61,500 form originates only after the proteolytic modification by the lysosomal cathepsin B. The correct sequence of modifying steps, the chemical nature of the groups removed or added, and the sites where the modifying enzymes function have to be elucidated.

IV. Single Steps of Assembly Studied in Vitro Some of the many steps between gene activation, mRNA processing, translation, and import plus assembly can now be examined in vitru. Investigations with cell-free systems and purified components provide the opportunity to study the individual processes in great detail. Consecutive steps can be entirely separated or can be governed to proceed according to a selected sequence. A. POLY(A )-mRNA CODINGFOR MICROBODY PROTEINS +

Translatable mRNA directing the synthesis of precursors of microbody enzymes has been prepared from various sources, usually in the form of total poly(A +)--A. Earlier studies using inhibitors of RNA synthesis have already tried to decide whether the appearance of enzyme activities is suppressed by these inhibitors. While some investigators found such a relationship (Hock and Beevers, 1966; Radin and Trelease, 1976), others (Tester, 1976; Khan et al., 1979) came to conclusions diverging from the model of a dominating control of translation. Inhibitor studies so far produced conflicting claims, especially with respect to the question of from which stage of development onward is the level of translatable mRNA already established. These difficulties can only be overcome by an exact measurement of mRNA availability. Messengers from rat liver (Goldman and Blobel, 1978; Robbi and Lazarow, 1978), Chlurella fusca (Dunham and Thurston, 1980), cucumber (Weir et al., 1980; Becker et al., 1982; Kruse er al., 1981), watermelon (Walk and Hock, 1978; Hock and Gietl, 1982), castor bean endosperm (Roberts and Lord, 1981a,b; Martin and Northcote, 1982), and green leaves of Lens culinaris (Gerdes and Kindl, 1982) are being used. For cucumber cotyledons, the levels of mRNA for two glyoxysomal enzymes were found to rise and fall parallel with the development of glyoxysomes, but preceding the latter by about 1 day (Weir et al., 1980). A similar rise of mRNA levels for glycollate oxidase was observed in cotyledons being illuminated, and

BIOSYNTHESIS OF MICROBODIES

21 1

for catalase mRNA during greening of leaves (Gerdes and Kindl, 1982). The size of poly(A +)-mRNA was assessed for catalase mRNA (Gerdes and Kindl, 1982) with I7 S malate synthase and isocitrate lyase with 22 S each (Weir et al., 1980). B. SYNTHESIS ON FREEPOLYSOMES Goldman and Blobel (1978) showed that albumin, from which we know that it is eventually secreted, but not catalase or urate oxidase can be synthesized at bound polysomes. It turned out that all microbody enzymes so far investigated are produced at free polysomes. This is true for isocitrate lyase in N . crassa (Zimmermann and Neupert, 1980) or R . communis (Roberts and Lord, 1981b) or C . fusca (Dunham and Thurston, 1980).

C. PRODUCTS OF in Vitro TRANSLATIONS Several enzymes from liver peroxisomes, algae, fungi, and plant microbodies have been translated in vifro.Lysate from rabbit reticulocytes, the wheat germ 30 S system, ascites tumor cells, and cell free preparations from yeast (Hofbauer et al., 1982) have been successfully applied. Isocitrate lyase was isolated from translation mixtures directed by mRNA from acetate-grown C . fusca (Dunham and Thurston, 1980), acetate-grown N . crasm (Zimmermann and Neupert, 1980), cucumber cotyledons (Weir ef al., 1980, Kindl, 1982), and endosperm of R . communis (Roberts and Lord, 1981b). In all cases except Chlorella, a M, = 64,000 form of isocitrate lyase was obtained which did not differ in size from the mature organellar enzyme. The value of M, = 47,000 of Chlorella isocitrate lyase deviates drastically. Isocitrate lyase preparations are known to vary in the amount of proteolytic products (isocitrate lyase A and B; Weir et al., 1980), and we found that a very quick purification procedure including phenylmethylsulfonylfluoride in all early steps greatly diminishes the amount of smaller M, 62,000 form. But it is unlikely that the M, 47,000 protein is produced by a degradation of the M, 64,000 form. It is important to point out that the Chlorella isocitrate lyase could as yet not be attributed to microbodies . Malate synthase produced by in vitro translation was not distinguishable in the size of its subunits from the mature enzyme (Kruse et al., 1981); but other interpretations, with respect to its possible nature as glycoprotein, had also been put forward (Riezman et al., 1980). The M,of catalase as analyzed after in vifrotranslation does not seem to differ significantly from a form of the enzyme detectable also in vivo when proteolytic alterations can be avoided (Goldman and Blobel, 1978; Robbi and Lazarow, 1978; Ammerer et al., 1981). These data were further strengthened by in vitro incorporation of the f~rmyl-[~~S]methionyl group from f~rmyl-[~~S]methionyl-

H. KINDL

212

tRNApt (Lazarow et al., 1982). The formyl group prevents the hydrolysis of terminal amino acids by the function of amino peptidases present on eukaryotic ribosomes. In vitro translated catalase apomonomer (a) was compared with the peroxisomal catalase (b): after trypsin digestion and two-dimensional mapping, the individual peptides from both forms of catalase (a and b) comigrated. This implies that catalase does not undergo any change in its primary structure, neither during translation nor during import into the organelle. The same may apply for urate oxidase and other liver peroxisomal proteins (Goldman and Blobel, 1978; Lazarow et al., 1982). Malate dehydrogenase, prepared from glyoxysomes of germinating watermelon seedlings,is the example of a microbody enzyme by which posttranslational modification can best be studied (Walk and Hock, 1978; Hock and Gietl, 1982; Gietl and Hock, 1982). But here, too, a cotranslational modification could be excluded. The in vitro formed monomer M , 41,000 is indeed larger than the mature organellar protein (Mr, 33,000) but processing occurs when translation is turned off. All studies aimed at the demonstration of a cotranslational modification by simultaneous incubations of translation systems with microsomal preparations failed (Goldman and Blobel, 1978; Roberts and Lord, 1981b). Likewise, the translation product was not transferred into the microsomal vesicles and was not protected against proteolytic degradation.

D.

IMPORT OF PROTEINS INTO

MICROBODIES

Since in vitro translation of microbody enzymes has been successfully performed, it was also intended to demonstrate in vitro the subsequent steps of import into the organelle. The rationale was to separate in time the translation process and the import process. Under these conditions it can be decided whether or not the two processes have to take place at the same time, whether the import is cotranslational or posttranslational . In addition, import studies evidence the selectivity of uptake, with respect to the protein to be imported as well as the kind of organelle performing the import. Zimmermann and Neupert (1980) working with N . crassa succeeded in demonstrating the uptake of isocitrate-lyase into organelle preparation containing glyoxysomes. As a control, a corresponding preparation was used from glucosegrown cells lacking glyoxysomes. Malate synthase was imported posttranslationally into glyoxysomes from cucumber cotyledons (Kruse et al., 1981). Newly translated malate synthase, after the addition of translation inhibitors and in the absence of ribosomes, was added to a preparation containing glyoxysomes together with plastids, mitochondria, and ER vesicles. The process was shown to be quite selective and efficient, as only a small proportion remained in the medium and almost all of the malate synthase bound to organelles was attributable to glyoxysomes. Malate synthase did not change its subunit M,in a percepti-

BIOSYNTHESIS OF MICROBODIES

213

ble way. In contrast, malate dehydrogenase imported into the glyoxysomes by similar means (Hock and Gietl, 1982) exhibited a sequence of consecutive proteolytic steps which parallel the uptake. At present, there are no hints of how the proteins are selected and with respect to the driving force used for the import. The addition of ATP or GTP does not seem to enhance the rate of uptake. It is possible that there exist endogenous sources providing an energy gradient or even a temperature gradient (Hryb, 1981) created by the organelle itself. As examplified by the brown fat tissues or the heat liberated by the uncoupled fat oxidation by germinating seeds, it is feasible that microbodies produce enough heat to maintain a temperature gradient between the organelle and the cytosol.

E.

CHEMICAL MODIFICATION A N D CHANGES IN HYDROPHOBICITY A N D AGGREGATION

Tetrameric isocitrate lyase prepared from acetate-grown N . crassa was found to differ significantly from the monomeric precursor obtained by in vitro translation. Although the monomeric form was soluble, it could interact with Triton X-100 to form mixed micelles. The functional tetrameric form, however, did not show such a behavior (Zimmermann and Neupert, 1980). Malate synthase from cucumber cotyledons is capable of binding phospholipids. This is true for the mature glyoxysomal enzyme (octamer). But both the cytosolic monomeric form and the monomeric form translated in vitro are characterized by an exceedingly high hydrophobicity and capacity to bind phospholipids (Kruse and Kind], 1980; Kind], 1982). Octameric mature malate synthase or processed monomeric malate synthase obtained from cytoplasm can be shifted to the 100 S aggregate, while an in vitro translated but unprocessed form was not amendable to such a shift. Exemplified by the import of precursor malate synthase into glyoxysomes, a hypothetical model of organelle biosynthesis is illustrated schematically in Fig. 4. It takes into consideration that in several cases of microbody enzymes processing and oligomerization can also take place in the cytosol. The question remains open whether a minimal modification, probably within the peptide molecule, occurs or whether chemical modification does not take place at all. Incorporation of formyl-methionine, if it can also be established with the newly synthesized protein afer the import, seems to exclude any processing at the N-terminal site; the change in hydrophobicity, however, is a reason to believe that import is paralleled by minor modifications. Whether those modifications should be discernible by fingerprint techniques remains to be clarified. There are clues that a peptide can be embedded into a membrane or even pass a membrane just by refolding. A membrane-induced or phospholipid-induced refolding of peptides can be envisaged according to the model proposed by Wickner (1979; Ito er al., 1980).

214

H. KlNDL

Aggregate

Octamer

FIG.4. Hypothetical model of biosynthesis of a microbody protein by the example of malate synthase.

Different forms of oligomerization or aggregation are not restricted to malate synthase. Varying aggregation has been described for plant malate dehydrogenase (Walk and Hock, 1976; Walk et al., 1977; Koller and Kindl, 1977; Blackwood and Miflin, 1976; Longo et al., 1977). Wood et al. (1981) have ascertained similar properties by testing pig heart malate dehydrogenase being known to bind to membranes (Comte and Gautheron, 1978). The changeover to aggregation, when a hydrophobic protein molecule cannot bind to another amphipathic compound, has been observed in other cases (Ott and Brodbeck, 1978; Francesco and Brodbeck, 1981).

V. Special Types of Cells The pathway of microbody biosynthesis has so far been discussed with little reference to the kind of cell in which these organelles function. It is necessary to catch up on some of the properties, and modes of induction, of the tissues containing peroxisomes or glyoxysomes as prominent metabolic compartments. A. ANIMALCELLS With respect to microbody function, we anticipate differences in biochemical characteristics depending upon the function of the cells. We know of tissues that

BIOSYNTHESIS OF MICROBODIES

215

are characterized by lipid degradation (liver), or tissues with marked gluconeogenic activities (kidney, bladder), or intensive urate oxidation. I . Liver Cells Hepatocytes are unusual cells in that they release not only proteins into the blood but also secrete proteins into the bile by a system of endocytic vesicles. In both cases the exported proteins are protected against the degradative function of lysosomes. It is not clear whether the proteins to be secreted are being sorted immediately within the ER or whether newly synthesized proteins and glycoproteins migrate with other newly synthesized proteins in ER vesicles to the respective compartment and are then distinguished at these sites. If some of the peroxisomal proteins should also pass through sites at the ER, a high extent of differentiation is proposed for the region of sorting out. Despite my intention, mentioned at the beginning, to provide a unifying concept of microbody biosynthesis, several results question this point of view. Even with so closely related tissues, as rat liver and mouse liver, the as yet presented data seem to exclude each other. Marked variation in the multiplicity of catalase between the homologous tissues in different species has been observed. While no hints for its glycoprotein nature were found in rat liver, and catalases from other organisms do not seem to contain glycomoieties (Ammerer et ul., 1981), the mouse liver enzyme is described as containing sialic acid units (Master, 1982). In liver, the enzymes of the two sites involved in fatty acids degradation, namely mitochondria and peroxisomes, can be compared. In plant cells, however, mitochondrial @-oxidationdoes not play any role if glyoxysomes are present (Cooper and Beevers, 1969; Frevert and Kindl, 1980). Also in terms of biosynthesis it is noteworthy that the compounds of the fatty acid degradation system, which includes enzymes for transfer of unsaturated fatty acids (Dommes et al., 1981), bear little resemblance if we compare microbodies and mitochondria. As discussed in more detail (Frevert and Kindl, 1980; Tolbert, 198l), the enzymes for P-oxidation in microbodies (Osumi and Hashimoto, 1979; Frevert and Kindl, 1980) are arranged in a manner very similar to Escherichiu coli (Pawar and Schulz, 1981), but differ in many respects from the mitochondria1 system. 2 . Other Cells Cytochemical studies have revealed microbodies in a large number of cells. The appearance of catalase and oxidase-containing organelles closely associated with myelin-forming sheaths suggests that peroxisomes may play roles in membrane lipid metabolism of nervous tissues. The enzymes of diglyceride formation from dihydroxyacetone phosphate have recently also been detected in brain microperoxisomes (Hajra and Bishop, 1982). In terms of biosynthesis, these

216

H. KlNDL

microperoxisomes may, therefore, not be considered as proforms but rather as microbodies that differ from liver peroxisomes only in size. In the preputial gland, a massive proliferation of microbodies, but no membrane continuities between the ER and the microbodies was detectable (Gorgas, 1982). The appearance of microperoxisomes fully competent in fatty acid poxidation was triggered in the adrenal cortex by hormone (ACTH) (Russo and Black, 1982). A drastic defect in peroxisome assembly (Goldfischer, 1982) may be the reason for a cerebrohepatorenal disease (Zellweger’s syndrome). This assumption, which is underlying the observations that the most remarkable feature of this disease is the absence of hepatic and renal peroxisomes, was supported by the investigation of abnormal metabolism of cholesterol esters and bile acids. That may have some parallels with the peroxisomal location of the conversion of cholestanoic acid into cholic acid (Pedersen and Gustafsson, 1980). This process chemically resembles fatty acid p-oxidation. New dimensions for animal peroxisomes and for processes of differentiation may be found if reports concerning the occurrence of glyoxylate cycle enzymes will be further strengthened. Fetal tissues and the bladder of toads (Jones et al., 1982) have recently been found to contain low amounts of malate synthase and isocitrate lyase.

B. DIFFEREBTIATED MICROBODIES IN PLANT CELLS, FUNGI, AND ALGAE 1 . Glyoxysomes

The appearance of these organelles can be expected when cells possess pools of reserve lipids or are externally supplied with fatty acids or acetic aJid. In seeds, the storage pools can consist to a varying extent of lipids which are mobilized and prepared for gluconeogenesis essentially by catalysis of glyoxysoma1 enzymes. The storage tissue is an endosperm in the case of seeds of R . communis, a scutellum ( Z . mays) or cotyledons (Cucurbitaceae,Pisum sativum, Helianthus annuus). During early postgerminative development in the dark, glyoxysomal protein components undergo an increase and subsequent decline. This holds true for the enzymes of fatty acid p-oxidation and the enzyme of glyoxylate cycle (Breidenbach et al., 1968; Cooper and Beevers, 1969; Bieglmayer et al., 1973; Huang and Beevers, 1973; Schnarrenberger et al., 1980; Tchang et a f . , 1981). Glyoxysomal activities are also functioning in fern spores during the first stage of germination (Gemmrich, 1979; DeMaggio et al., 1980). Microbodies with enzymes for P-oxidation and glyoxylate cycle can be induced when Candida yeasts grown on sucrose are transferred to alkanes as Csource (Tanaka et al., 1982). Although electron microscopic studies demonstrate that these microbodies are frequently located near the ER, the microbodies are

BIOSYNTHESIS OF MICROBODIES

217

proposed to arise by division of preexisting organelles. As to the synthesis of catalase, a measure of microbody formation, a similar inhibitory effect by glucose was notices with alkane-utilizing Candida tropicalis as with S . cerevisiae. The latter yeast is one of the few examples where localizations of microbody enzymes have intensively been tried, but, apart from conflicting reports, not resolved (Ruis, 1979). Glyoxysomes from Aspergillus tamarii (Graves et al., 1976), Blastocladiella (Mills and Cantino, 1975), and from N . crassa (Kobr, 1969; Kobr and Vanderhaeghe, 1973) have been studied in detail. At present, Neurospora glyoxysomes are one of the few examples by which import of glyoxysomal precursors posttranslationally can be demonstrated (Zimmermann and Neupert, 1980; Desel et al., 1982). The aquatic fungus Enthophlyctis variabilis shows microbodies in developing Zoosporangia; they were found to have continuities with the ER (Powell, 1976). The observations suggested that glyoxysomes originate as local dilations from the tubular ER. The growth of many algae on acetate in the dark is dependent on the derepression of glyoxylate cycle enzymes. Their induction could be studied with synchronous cultures of Euglena gracilis (Woodward and Merrett, 1975). Chlorella isocitrate lyase, so far not assigned to organelles, was inducible, only in the dark, in the absence of glucose (Dunham and Thurston, 1978). In Tetrahymena pyriformis the glyoxylate cycle enzymes are present and functioning also in the presence of glucose in the medium (Blum, 1982). During germination glyoxysomes do not seem to be synthesized completely de novo. as small but significant amounts of microbodies are already present. They are already formed during seed ripening before the state of dry seed is reached (cf. Section V,B,3). 2. Leaf Peroxisomes Leaf peroxisomes represent a well-defined species of microbodies (Tolbert, 1971, 1981, 1982). They are appressed to chloroplasts and cooperate with them in photorespiration. Photorespiration can be at half the rate of photosynthesis or five times more active than mitochondria1 respiration. Peroxisomes participate in this cyclic pathway by contributing three enzymes: glycollate oxidase, a transaminase which converts glyoxylate into glycine and serine into P-hydroxypymvate, and P-hydroxypyruvate reductase (Rehfeld and Tolbert, 1972). Furthermore, leaf peroxisomes house catalase, malate dehydrogenase, urate oxidase, and the enzymes of fatty acid P-oxidation (Gerhardt, 1981). A comparable equipment with peroxisomal enzymes was found in some algae (Silverberg, 1975; Stabenau, 1976). The biosynthesis of leaf peroxisomes is under the control of phytochrome (Feierabend, 1975; Hong and Schopfer, 1981). According to Feierabend and

218

H. KINDL

Beevers ( 1 972a), the formation of microbodies and the differentiation of plastids in the light are independently controlled. The formation of microbody enzymes, i.e., hydroxypyruvate reductase and glycollate oxidase, is considerably promoted without greening by very short light exposure, or when the chlorophyll synthesis is prevented, by continuous light, in the presence of the inhibitor aminotriazol. It is remarkable that another kind of signal response chain, a longlived transmitter, is responsible for the connection of urate oxidase biosynthesis with the phytochrome system. Biosynthetical studies have been performed with greening leaves from Lens culinaris. A concomitant increase in catalase activity and in equilibrium density of microbodies was established (Ludwig and Kindl, 1976). Compared to other plant sources, the highest level of translatable mRNA coding for catalase was found in these greening leaves. Oligomeric glycollate oxidase labeled in vivo exhibited the same subunit M , as the in vitro translated precursor form (Behrends et al., 1982). 3 . Transition Forms of Microbodies In cotyledons containing fat reserves and capable of greening, an exchange of glyoxysomal activities for leaf peroxisomal function can be effected. This is a light-dependent process, or more precisely, the decrease of glyoxysomal activities is somewhat enhanced by light, in an unknown manner, whereas the increase of leaf peroxisomal activities is under the positive control of phytochrome (Hong and Schopfer, 1981). Several mechanisms have been envisaged for this change in microbody species. One extreme can be seen by the postulation of two independent populations of microbodies being present at this transition state (Fig. 3). Different rates of synthesis and turnover for each of these species could account for the observation that peroxisomal activities increase and glyoxysomal activities decrease. Kagawa and Beevers (1975) used for their studies a markedly prolonged dark period and found that the decline of glyoxysomal activities paralleled the decline of protein content in the microbody fraction. Upon subsequent illumination, the same increase of leaf peroxisomal activities was observed as was expected for a normal development, i.e., without a preceding depletion of glyoxysomes. One can therefore argue that peroxisome biosynthesis took place after illumination, despite the long period of heterotrophic growth paralleled by the marked decline of enzyme levels of glyoxylate cycle. It must have led to an extremely low level of glyoxysomes. Since the appearance of peroxisomal activities was not influenced by the special dark treatment, the peroxisome biosynthesis can be assumed to be entirely independent from preexisting glyoxysomes. These are the strongest arguments for the so-called two-population hypothesis. Most crucial is, in this context, the assessment of the amount of protein in the microbody fraction, as in the plants used here (cucurbitaceae) glyoxysomes tend to be contaminated with

BIOSYNTHESIS OF MICROBODIES

219

fragments from protein bodies (Kara and Kindl, 1982a). As the amount of proteins in the latter organelle is tremendously reduced during germination, the observed decline of protein in the glyoxysome fraction may just reflect the decrease in contamination by protein bodies. Most data available at present are rather in accord with the assumption that there is only one microbody population (Fig. 1). It is assumed (one-population hypothesis) that the content, and also the enzymic capacity, of the organelle is gradually changed without altering, not beyond the basic turnover, the principal structure. Theimer el al. (1975) failed to detect an alteration of the organelle’s equilibrium density when they fed D,O during the greening phase. Gerhardt (1978) did not observe an enhanced increase in catalase as would be expected when a new population of organelles was formed. Incorporation of labeled aminolevulinic acid into enzymatically active catalase was not changed during the transition state (Gerhardt and Betsche, 1976). This kind of experiment with precursors of the heme moiety only shows the synthesis of the tetrameric enzyme, but does not allow for any synthesis of the monomeric apoprotein. Koller and Kindl (1978b) revealed that the amount of malate synthase synthesized 10 hours after onset of illumination was almost the same as during the dark period. This demonstrated that the synthesis of glyoxysomal proteins is not abruptly turned off upon illumination. While this study with malate synthase was performed by immunoprecipitation and subsequent electrophoresis of the de n o w synthesized protein, isocitrate lyase formation was examined with the same purpose but by means of density labeling (Franzisket and Gerhardt, 1980). In a more detailed investigation including several glyoxysomal and leaf peroxisomal proteins, it was demonstrated (Kindl, 1982) that the rate of de n o w synthesis of catalase during the transition phase does not significantly change. Catalase is a constituent of both glyoxysomes and peroxisomes. While the biosynthesis of the other leaf peroxisomal proteins increases markedly, catalase synthesis does not parallel these courses. Catalase does, therefore, not behave as a component of a newly constructed organelle, but rather as part of an already existing compartment. Schopfer er al. (1976) have hypothesized that there should exist a transient organelle functioning simultaneously as glyoxysome and as leaf peroxisome. They came to this conclusion by an electron microscopic search which yielded a close correlation between the period of transition and the maximum in the appearance of microbodies appressed to both the lipid bodies and the chloroplasts. That means that these microbodies (glyoxyperoxisomes) function as glyoxysomes in contact with lipid bodies, and as leaf peroxisomes exchanging products with chloroplasts. As Schopfer postulated that these microbodies originate from cisternae of the ER, a high turnover of the organelles, by the lytic compartment,

220

H.KINDL

is required to account for the gradual exchange of one population of microbodies by another. The fact that investigators failed to separate the two populations of microbodies proposed, the biochemical work concerning the formation and uptake of protein components, and the histochemical studies support the concept that only one population is present during transition. But these data do not unequivocally prove this hypothesis by excluding all other conceptions. Another approach, differing from the former in principle, is necessary to provide, together with all former data, a final decision. The new conception simply implies that in the transition state the single microbody must contain both the enzymes of the original organelle and the enzymes of the organelles eventually resulting. This means, exemplified for the transition glyoxysome + peroxisome, that P-oxidation, glyoxylate cycle, and part of the glycollate pathway has to be functioning in one and the same organelle. Therefore, conversions from a substrate of the glyoxysomal pathway to the glycollate pathway and vice versa should be possible without leaving the organelle. If this is not the case and a substrate indeed has to pass from one organelle to another and can thus be diluted by the surrounding pool, one should be able to discern it on the basis of conversion rates. If the glyoxylate cycle and glycollate pathway (glycollate + glycine) are proceeding in two different organelles, only a very small proportion of an intermediate common to both pathways is transferred from one reaction sequence to the other. But if both pathways are located in the same organelle, the enzymes of both pathways can compete for a common intermediate. Figure 5 illustrates how a flux along the glyoxylate pathway can be diverted to the glycollate pathway and vice versa. This scheme allows for the conversion of isocitrate or palmitoyl-CoA into glycine and accounts for the incorporation of glycollate into malate. Both serine glyoxylate transaminase and malate synthase compete for the substrate glyoxylate. If K , and V,,, do not differ markedly, the formation of malate and glycine is a measure of the amount of transaminase and malate synthase present in the vicinity of the glyoxylate-producing enzyme. It has to be considered that glyoxysomes already contain low amounts of glycollate oxidase, and that serine glyoxylate transaminase becomes active after illumination only. Therefore, a minor flux of glycollate + malate is to be expected during darkness. In the light, when the activities of glycollate oxidase and the transaminase are increasing manyfold, a maintainance of malate formation should only be observed if glyoxylate has not been transferred to another organelle. Oxidation of glycollate (Chang and Huang, 1981) and of glycine (Arron and Edwards, 1980) has to be kept in mind, if our considerations comprise transport to other organelles. In contrast to this, the conversion of isocitrate into glycine cannot be detected in the dark, as the level of transaminase is too low. After illumination, a signifi-

22 1

BIOSYNTHESIS OF MICROBODIES

-

Patmitoy1 COA

1

Acetyl -CoA

(I)

(1)

I \ Citrate

V

Glyoxylate

1

*Glyoxylate

(1)

SGT

M a l a t e (1)

Glycine

(1)

FIG. 5 . Intersystem crossing between the glyoxylate cycle and glycollate pathway. The possibilities of conversions glycollate-, malate and isocitrate + glycine are brought out. The numbers in parentheses denote a possible labeling with I4C. MS. Malate synthase; SGT, serine glyoxylate transaminase.

cant proportion of C - 1 of isocitrate of palmitoyl-CoA applied should be attributable to glycine. Data obtained with isolated microbodies and radioactive substrates show (Table 11) transfers from glyoxylate cycle to glycollate pathway and vice versa. In accordance with the general concept for compartments and minicompartments (cf. Kindl, 1979), the intermediate glyoxylate offered in the surrounding medium or by other organelles has less access to the glyoxylate-converting enzymes than the glyoxylate molecules that are produced in proximity of malate synthase or transaminase. The data support the conception of transition forms, of glyoxyperoxisomes in this particular case. Transition forms of microbodies, as we are discussing here in detail, probably occur also in several other cases when changes in metabolic situations are paralleled by changes in organelle equipment. That should be true of fungal cells being supplied with glucose and then transferred to acetate or methanol. A very attractive example of transition was expected to be amenable to intensive investigations, namely, algae that are able to switch from photoautotropic growth to

H.KINDL

222

TABLE I1 TTRANSFORMATION OF GLYCOLLATE INTO MALATEAND OF ISOCITRATE (OR PALMITOYL-COA) INTO GLYCINE BY ISOLATED GLYOXYPEROXISOMES~

Substrate

Products

[ 1-14C]Glycollate

Malate Succinate Glycine Malate Succinate Glycine Malate Succinate Glycine

[ I -'4C]Isocitrate

[ I-14C]Palmitoyl-CoA

Amount produced by organelles from etiolated cotyledons

Amount produced by organelles from greening cotyledons

5.5 0.5

9.2 0.6

0.2

15.8

18.5

15.0

2. I < 0.1 15.2 18.0 < 0.1

3.2 7.2 17.3

19.5 5.8

oPartially purified, but intact microbodies were prepared either from etiolated cotyledons or from cotyledons of cucumber seedlings which were illuminated for 20 hours. They were incubated with a 5 JLM solution of the labeled substrate (20 pCi) simultaneously with 1 )LM glyoxylate for 5 minutes at 25°C. Subsequently, the educts, as well as several dicarboxylic acids and tricarboxylic acids, glyoxylate, glycine, and serine were isolated. Their radioactivity was determined and the yield was given as percentage of the radioactivity administered.

heterotrophic growth on acetate. Unfortunately, the isolation of intact microbodies posed a great problem. Microbodies from etiolated leaves gaining their new components at different rates and exhibiting increasing equlibrium densities upon short illuminations (Feierabend and Beevers, 197213) certainly also belong to the group of transient forms. The same question could be asked for peroxisomes in achlorophyllous tissue (Berger and Gerhardt, 1971). In the cotyledons discussed above, even more than one transition stage can be envisaged. The cells made early during seed ripening are not dividing during the following stages: late stages of ripening + dry seed + germinating seed + greening cotyledons. It has been shown that (1) during ripening of seeds the microbodies change their constituents, and that (2) at a late stage of greening the same cell again alters its metabolic capacity by microbodies acquiring new protein components (cf. Sections V,B,2 and 4). Thus, exactly the same cell houses, at different times, at least four different forms of microbodies (Kindl, 1982). In ripening cotton seeds (Choinski and Trelease, 1978) and cucumber seeds (Kindl et al., 1980; Frevert et al., 1980), a population of microbodies (peroxisomes) is present that is characterized by the occurrence of catalase and the multifunctional protein (enoyl-CoA hydratase/3-hydroxyacyl-CoA dehydrog-

BIOSYNTHESIS OF MICROBODIES

223

enase). The latter protein is indicative of peroxisomes and glyoxysomes (Frevert and Kindl, 1980). In contrast, at a later stage of seed ripening the microbodies drastically change their enzymatic capacity and become more or less glyoxysomes. According to Choinski et a f . (1981), cotton embryos excised from seeds 40 days after anthesis and cultured in vitro on a medium containing abscissic acid continued to increase the enzyme activities of P-oxidation, malate, synthase, catalase, and aspartate aminotransferase. With ripening cucumber seeds, de novo synthesis via cytosolic pools was demonstrated in vivo (Frevert et al., 1980) for several glyoxysomal proteins. The presence of these glyoxysome-like organelles at the latest stage of seed ripening and in the dry seed (Koller et af., 1979; Kbller and Kindl, 1979) could suffice that during germination the glyoxysomes do not have to be made from scratch. The most plausible model of microbody formation arises when we consider both de novo synthesis, or growing and fissioning microbodies, and the assembly of transition forms. It is then decisive which precursors are offered by the synthesis at the free polysomes and are thus provided in the cytosolic pool (cf. Fig. 1). Microbodies probably cannot select if several different, but appropriate precursors are at disposal. The problem then encountered is the mode of degradation of the components already present in the preceding form of microbodies but to be eliminated during the changeover to the successive form. If the general turnover is very fast, the exchange is only a question of a good functioning lytic compartment. If yet the turnover of microbodies is slow and the components of the former organelle have to be selectively degraded, we are lacking in a useful model. It is not unlikely that complexes or other quaternary structures not properly functioning anymore are the preferred targets of peptidases. Compatible with this hypothesis are the morphological data (Burke and Trelease, 1975; Thomas and Trelease, 1981) and the biochemical data concerning the rate of de novo synthesis of microbody proteins during the early transition state (Behrends er al., 1982; Kindl, 1982). 4. Other Microbodies Very specialized forms of microbodies have been described for methanolgrown yeasts (Hansenula polymorpha, C . boidinii). Upon shift from glucose medium to alcohol medium, instead of a few small microbodies one big organelle housing catalase and alcohol oxidase is decernible. There are hints that the assembly of the flavoprotein takes place within the peroxisomes. At a certain stage of growth, a microbody appears with a crystalline body indicating that alcohol oxidase apoprotein is present. If methanol-grown cells were transferred into a glucose-containing medium the synthesis of alcohol oxidase was instantaneously repressed and the large peroxisome was degraded (Bormann and Sahm, 1978). Another form of microbodies which seems only to contain catalase should be

224

H . KINDL

mentioned. While in many algae under photoautotropic conditions an oxidase, i.e., glycollate oxidase, is involved in the photorespiration, other algae do not use an oxidase at this step but rather a dehydrogenase (Stabenau, 1974b). In these instances, e.g., Chlorogonium elongatum (Stabenau, 1974a) or Dunaliella marina (Kruger and Kindl, 1980), the conversion of glycollate into glycine takes place in the mitochondria. Among the peroxisomes differing from liver peroxisomes and from leaf peroxisomes we find ureide synthesizing microbodies in the nodules of leguminosae (Hanks et a l . , 1981) and in a late form of leaf peroxisomes primarily involved in urate oxidation (Hong and Schopfer, 1981).

VI. Conclusion Most, or all, protein components of peroxisomes or glyoxysomes are synthesized at free ribosomes and are imported posttranslationally in the growing organelle. Assembly takes place within the organelle by acquiring prosthetic groups, processing, or oligomerization, respectively.

ACKNOWLEDGMENTS Financial support of the Deutsche Forschungsgemeinschaft and Fonds der Chemischen Industrie is gratefully acknowledged.

REFERENCES Ammerer, G.. Richter, K., Hartter, E.. and Ruis. H. (1981). Eur. J . Biorhem. 113, 327-331. Anon, G. P., and Edwards, G. E. (1980). Plant Sri. Lett. 18, 229-235. Badenoch-Jones, J., Spencer, D., Higgins, T. J. V., and Millard, A. (1981). Pluntu 153, 201-209. Becker, W. M., Riezman, H., Weir, E. M.,Titus, D. E., and Leaver, C. J. (1982). Ann. N . Y . Acud. Sci. 386, 329-349. Beevers, H. (1979). Annu. Rev. Plant Physiol. 30, 159-193. Beevers, H. (1982). Ann. N.Y. Acad. Sci. 386, 243-253. Behrends, W., Gerdes, H.-H., Kindl, H. (1982). Unpublished results. Berger, C., and Gerhardt, B. (1971). Planta 96, 326-338. Bergner, U . , and Tanner, W. (1981). FEBS Lett. 131, 68-72. Bieglmayer, C., Graf, J., and Ruis, H. (1973). Eur. J. Biochem. 37, 553-562. Black, S. D., French, J. S . , Williams, C. H., and Coon, M. J . (1980). Biochem. Biophys, Res. Commun. 84, 705-716. Blackwood, G. C., and Miflin, B. J. (1976). Plant Sri. Lett. 7 , 435-446. Blobel, G., Walter, P., Chang, C. N., Goldman, B. M., Erickson, A. H., and Lingappa. V. R. (1979). Symp. SOC. Exp. Biol. 33, 9-36.

BIOSYNTHESIS OF MICROBODIES

225

Blum. J. J. (1982). Ann. N . Y . Acud. Sci. 386, 217-227. Bormann, C., and Sahm, H. (1978). Arch. Microbiol. 117, 67-72. Bowden, L., and Lord, J. M. (1976a). Biochem. J . 154, 491-499. Bowden, L., and Lord, J. M. (1976h). Biochem. J . 154, 501-506. Breidenhach, R. W., Kahn, A.. and Beevers, H. (1968). Plant Phvsiol. 43, 705-713. Brown, R. H., Bowden, L., and Lord, J. M. (1976). Planta 130, 95-96. Burke, J. J., and Trelease, R. N. (1975). Plant Physiol. 56, 710-717. Cannon, B., Alexson, S., and Nedergaard, J. (1982). Ann. N . Y . Acad. Sci. 386, 40-58. Chang, C.-C., and Huang, A. H. C. (1981). Plant Physiol. 67, 1003-1006. Ching, T. M. (1970). Plant Physiol. 46, 475-482. Choinski, J. S., Jr., and Trelease, R. N. (1978). Plant Physiol. 62, 141-145. Choinski, J. S., Jr., Trelease, R. N., and Doman, D. C. (1981). Planta 152, 428-435. Chua, N.-H., and Schmidt, G. W. (1979). J. CellBiol. 81, 461-483. Chua, N.-H., Grossman. A. R., Bartlett, S. G., and Schmidt, G . W. (1980). In “Biological Chemistry of Organelle Formation” (T. Biicher, W. Sehald, and H. Weiss, eds.), pp. 113-118. Springer-Verlag, Berlin and New York. Comte, J., and Gautheron, D. C. (1978). Biochirnie 60, 1299-1305. Cooper, T. G . , and Beevers, H. (1969). J . Biol. Chem. 244, 3514-3520. Croy, R . R. D., Gatehouse, J. A., Evans, I. M., and Boulter, D. (1980a). Planta 148, 49-56. Croy, R . R. D., Gatehouse, J. A,, Evans, I. M., and Boulter, D. (1980h). Planra 148, 57-63. Davis, B. D., and Tai, P.-C. (1980). Nature (London) 283, 433-438. DeDuve, C. (1973). J . Histochem. Cytochem. 21, 941-948. DeDuve, C., and Baudhuin, P. (1966). Physiol. Rev. 46, 323-357. DeMaggio, A. E., Greene, C., and Stetler, D. (1980). Plant Physiol. 66, 922-924. Desel, H., Zimmermann, R., and Neupert, W. (1982). Ann. N.Y. Acud. Sci. 386, 377-389. van Dijken, J. P., Veenhuis, M., and Harder, W. (1982). Ann. N.Y. Acad. Sci. 386, 200-216. Dommes, V., Baumgart, C., and Kunau, W. H. (1981). J. Biol. Chem. 256, 8259-8262. Donaldson, R. P., Tolbert, N. E., and Schnarrenberger, C. (1972). Arch. Biochem. Biophys. 152, 199-2 15. Donaldson, R. P., Tully, R. E., Young, 0. A,, and Beevers, H. (1981). PlantPhysiol. 67, 21-25. Douglass, S. A., Criddle, R. S., and Breidenbach, R. W. (1973). Plant Physiol. 51, 902-906. Dunham, S. M.. and Thurston, C. F. (1978). Biochern. J . 176, 179-185. Dunham, S. M., and Thurston, C. F., (1980). Arch. Microbiol. 128, 98-104. Durchschlag, H., Biedermann, G., and Eggerer, H. (1981). Eur. J . Biochem. 114, 255-262. Ellis, R. J. (1981). Annu. Rev. PlantPhysiol. 32, 111-137. Erickson, A. H., and Blohel, G. (1979). J. B i d . Chem. 254, 11771-1 1774. Feierabend, J. (1975). Plunta 123, 63-77. Feierabend, J., and Beevers, H. (1972a). Plant Physiol. 49, 28-32. Feierabend, J., and Beevers, H. (1972b). Plant Physiol. 49, 33-39. Flatmark, T., Christiansen, E. N., and Kryvi, H. (1981). Eur. J. Cell Biol. 24, 62-69. Flatmark, T., Kryvi, H., and Christiansen, E. N. (1982). Ann. N . Y . Acad. Sci. 386, 456-459. Floyd, G. L., Stewart, K. D., and Mattox, K. R. (1972). J. Phycol. 8, 176-184. Francesco, C., and Brodbeck, U. (1981). Biochim. Biophys. Acfa 640, 359-364. Franzisket, U., and Gerhardt, B. (1980). Plant Physiol. 65, 1081-1084. Frederick. S. E., Gruher, P. J., and Newcomh, E. H. (1975). Protoplasma 84, 1-29. Frevert, J., and Kindl, H. (1978). Eur. J. Biochem. 92, 35-43. Frevert, J., and Kindl, H. (1980). Eur. J. Biochem. 107, 79-86. Frevert, J., Koller, W., and Kindl, H. (1980). Hoppe-Seyler’s 2. Physiol. Chem. 361, 1557-1565. Fujiki, Y., Fowler, S., Hubbard, A. L., and Lazarow, P. B. (1982). J . Cell Biol. 93, 103-110. Gemmrich, A. R. (1979). Phyrochemistry 18, 1143-1 146.

226

H. KINDL

Gerdes, H.-H., and Kindl, H. (1982). unpublished results. Gerhardt, B. (1978). “Microbodies/Peroxisomenpflanzlicher Zellen.” Springer-Verlag, Berlin and New York. Gerhardt, B. (1981). FEES Lerr. 126, 71-73. Gerhardt, B., and Beevers, H. (1969). Plant Physiol. 44, 1475-1477. Gerhardt, B., and Betsche, T. (1976). Eer. Dtsch. Eot. Ges. 89, 321-324. Gietl, C., and Hock, B. (1982). Plant Physiol.. in press. Goeckermann, J. A,, and Vigil, E. L. (1975). J. Hisrochem. Cytochem. 23, 957-973. Goldfischer, S. (1982). Ann. N.Y. Acad. Sci. 386, 526-529. Goldman, B. M., and Blobel, G. (1978). Proc. Natl. Acad. Sci. U.S.A. 75, 5066-6070. Gonzalez, E. (1982). Plant Physiol. 69, 83-87. Gonzalez, E., and Beevers, H. (1976). Plant Physiol. 57, 406-409. Gonzalez, E., and Delsol. M. A. (1981). Plant Physiol. 67, 550-554. Gorgas, K. (1982). Ann. N . Y . Acad Sci. 386, 519-522. Graves, L. B., Jr., Armentrout, V. N., and Maxwell, D. P. (1976). Planta 132, 143-148. Gruber, P. J., Becker, W. M., and Newcomb, E. M. (1973). J. Cell Eiol. 56, 500-518. Hajra, A. K., and Bishop, J. E. (1982). Ann N.Y. Acad. Sci. 386, 170-182. Hanks, J. F., Tolbert, N. E., and Schubert, K. R. (1981). Plant Physiol. 68, 65-69. Hanover, J. A., and Lennan, W. J. (1980). J . Eiol. Chem. 255, 3600-3604. Harmey, M. A., and Neupert, W. (1979). FEES Lett. 108, 385-389. Henning, B., and Neupert, W. (1981). Eur. J . Eiochem. 121, 203-212. Herman, E. M., Baumgartner, B., and Chrispeels, M. J . (1981). Eur. J . Cell Eiol. 24, 226-235. Higashi, T., and Peters, T. (1963). J . Eiol. Chem. 238, 3952-3954. Hock, B. (1974). Planfa 115, 271-280. Hock, B., and Beevers, H. (1966). Z. PjZanzenphysiol. 55, 405-414. Hock, B., and Gietl, C. (1982). Ann. N.Y. Acad. Sci. 386, 350-361. Hofbauer, R., Fessel, F., Hamilton, B., and Ruis, H. (1982). Eur. J . Eiochem. 122, 199-2036. Hong, Y.-N., and Schopfer, P. (1981). Planta 152, 325-335. Hruban, Z., and Rechcigl, M. (1969). In?. Rev. Cyrol. Suppl. I. Hryb, D. J . (1981). FEBS Lett. 128, 1-4. Huang, A. H. C., and Beevers, H. (1973). J . Cell Biol. 58, 379-389. Huang, A. H. C., Moore, T., and Trelease, R. N. (1982). “Microbodies.” Academic Press, New York. Ito, K., Date, T., and Wickner, W. (1980). J. Eiol. Chem. 255, 2123-2130. Jacob, G. S., and Orme-Johnson, W. H. (1979). Biochemistry 18, 2967-2975. Jones, R. G., Davis, W. L., and Goodman, D. B. P. (1982). Ann. N.Y. Acad. Sci. 386, 165-169. Kagawa, T., and Beevers, H. (1975). Arch. Eiochem. Biophys. 167, 45. Kagawa, T., and Gonzalez, E. (1981). Plant Physiol. 68, 845-850. Kara, U. A. K., and Kindl, H. (1982a). Eur. J. Eiochem. 121, 533-538. Kara, U. A. K., and Kindl, H. (1982b). Eur. J. Biochem. 121, 539-544. Khan, F. R., Saleemuddin, M., Siddiqi, M., and McFadden, B. A. (1979). J. Eiol. Chem. 254, 6938-6944. Khan, R. F., and McFadden, A. B. (1979). Planr Physiol. 64, 228. Kindl, H. (1979). In “Regulation of Secondary Product and Plant Hormon Metabolism” (H. Luckner and K. Schreiber, eds.), pp. 49-61. Pergamon, Oxford. Kindl, H. (1982). Ann. N.Y. Acad. Sci. 386, 314-328. Kindl, H., and Kruse, C. (1982). Methods Enzymol.. in press. Kindl, H., and Lazarow, P. B. (1982). Ann. N.Y. Acad. Sci. 386. Kindl, H., Schiefer, S., and Lijffler, H.-G. (1980a). Planra 148, 199-207. Kindl, H., Koller, W., and Frevert, J. (1980b). Hoppe-Seyler’s Z. Physiol. Chem. 361, 465-467.

BIOSYNTHESIS OF MICROBODIES

221

Kobr, M. J. (1969). Biochem. Biophys. Res. Commun. 37, 640-645. Kobr, M. J., and Vanderkaeghe, F. (1973). Experientia 29, 1221-1223. Koller, W., and Kindl, H. (1977). Arch. Biochem. Biophys. 181, 236-248. Koller, W., and Kindl, H. (1978a). FEBS Lett. 88, 83-86. Koller, W., and Kindl, H. (1978b). Z. Naturforsch. 33c, 962-968. Koller, W.,and Kindl, H. (1979). Z . Naturforsch. Me, 1237-1242. Koller, W., and Kindl, H. (1980). Hoppe-Seyler’s Z . Physiol. Chem. 361, 1437-1444. Koller, W., Frevert, J., and Kindl, H. (1979). Hoppe-Seyler’s Z . Physiol. Chem. 360, 167-176. Kriiger, U., and Kindl, H. (1980). Unpublished. Kruse, C., and Kindl, H. (1980). Eur. J . Cell Biol. Kruse, C., and Kindl, H. (1982). Ann. N.Y. Acad. Sci. 386, 499-501. Kruse, C., Frevert, J., and Kindl, H. (1981). FEBS Lett. 129, 36-38. Kudielka, R., Kock, H., and Theimer R. R. (1981). FEBS Lett. 136, 8-12. Lampugnani, M. G., Martellini, P., Servettas, O., and Longo, C. P. (1980). Plant Sci. Lett. 18, 351-358. Lazarow, P. B., and DeDuve, C. (1973a). J . Cell Biol. 59, 491-506. Lazarow, P. B., and DeDuve, C. (1973b). 1. Cell B i d . 59, 507-524. Lazarow, P. B., and DeDuve, C. (1976). Proc. Natl. Acad. Sci. U.S.A. 73, 1043-1046. Lazarow, P. B., Shio, H., and Robbi, M. (1980). In “Biological Chemistry of Organelle Formation” (T. Biicher, W. Sebald, and H. Weiss, eds.). Springer-Verlag, Berlin and New York. Lazarow, P. B.,Robbi, M., Fujiki, Y., and Wong, L. (1982). Ann. N.Y. Acud. Sci. 386,285-300. Legg, P. G., and Wood, R. L. (1970). J . Cell B i d . 45, 118-129. Leighton, F., Bronfman, M., Lazo, O., and Brandan, E. (1982). Ann. N.Y. Acad. Sci. 386,62-80. Lennarz, W. J., (1980). “Glycoproteins and Proteoglycans.” Plenum, New York. Lewin, A. S., Gregor, I., Mason, T. L., Nelson, N., and Schatz, G.(1980). Proc. Nut/. Acud. Sci. U.S.A. 77, 3998-4002. Longo, G. P., Bracci, C., Bucceri, C., Pedretti, M., and Longo, C. P. (1977). Plant Sci. Lett. 9, 381-390. Lord, J. M. (1978). J. Exp. Bot. 29, 13-23. Lord, J. M. (1980). Subcell, Biochem. 7, 171-211. Lord, J. M., and Bowden, L. (1978). Plant Physiol. 61, 266-270. Ludwig, B., and Kindl, H. (1976). Hoppe-Seyler’s Z. Physiol. Chem. 357, 393-399. Mainferme, F.,and Wattiaux, R. (1982). Ann N.Y. Acud. Sci. 386, 507-509. Martin, C., and Northcote, D. H. (1982). Planta 154, 174-183. Masters, C. J. (1982). Ann. N.Y. Acad. Sci. 386, 301-313. Maxwell, D. P., Maxwell, M. D., Hanssler, G.,Armentrout, V. N., Murray, G.M., and Hoch, H. C. (1975). Planra 124, 109-123. Mellor, R. B., Bowden, L., and Lord, J. M. (1978). FEBS Lett. 90, 275-278. Meyer, D. I., and Dobberstein, B. (1980a). J. Cell B i d . 87, 498-502. Meyer, D. I., and Dobberstein, B. (1980b). J . Cell B i d . 87, 503-508. Miemyk, J. A,, and Trelease, R. N. (1981). Plant Physiol. 67, 341-346. Miernyk, J. A., Trelease, R. N., and Choinsky, J. S. (1979). Plant Physiol. 63, 1068-1071. Mills, G.L., and Cantino, E. C. (1975). Biochem. Biophys. Res. Commun. 67, 1256-1263. Mollenhauer, H. H., and Totten, C. (1970). Plant Physiol. 46, 794-799. Mollenhauer, H. H., Morre, D. J., and Gelley, A. G. (1966). Protoplasma 62, 44-52. Moody, D. E., and Reddy, J. K. (1976). J. Cell. Biol. 71, 768-780. Miiller, M., Hoggs, J. F., and DeDuve, C. (1968). J. B i d . Chem. 243, 5385-5395. Naito, K., Ikeda, K., Saito, T., Suzuki, H., Yamada, M., and Kuraishi, S. (1980). Plant Cell Physiol. 21, 1419-1430. Nelson, E., and Ryan, C. A. (1980). Proc. Natl. Acad. Sci. U.S.A. 77, 1975-1979.

228

H. KINDL

Neufeld, E. F. (1981). In “Lysosomes and Lysosomal Storage Diseases” (J. W. Callahan and J. A. Lowden, eds.), pp. 115-1 19. Raven, New York. Novikoff, A. B., and Shin, W. Y. (1964). J. Microsc. (Paris) 3, 187-201. Novikoff, A. M., Novikoff, P. M.,Davis, C., and Quintana, N. (1973). Histochem. Cytochem. 21, 737-75 1. Okada, Y., Frey, A. B., Guenthner, T. M., Oesch, F., Sabatini, D. D., and Kreibich, G. (1982). Eur. J . Biochem. 122, 393-402. Osumi, T., and Hashimoto, T. (1979). Biochem. Biophys. Res. Commun. 83, 479-485. Osumi, M., Fukuzumi, F., Teranishi, Y., Tabaka, A., and Fukui, S. (1975). Arch. Microbiol. 103, 1-11. Osumi, M., Kazama, H., and Sato, S. (1978). FEBS Lett. 90, 309-312. Ott, P., and Brodbeck, U. (1978). Eur. J . Biochem. 88, 119-125. Pawar, S., and Schulz, H. (1981). J . Biol. Chem. 256, 3894-3905. Pedersen, J. I., and Gustafsson, J. (1980). FEBS Left. 121, 345-348. Poole, B., Higashi, T., and DeDuve, C. (1970). J . Cell Biol. 45, 408-415. Powell, M. J. (1976). Protoplusmu 89, 1-27. Radin, J. W., and Trelease, R. N. (1976). Plant Physiol. 57, 902-905. Redman, C. B., Grab, D. I., and Irukulla, R. (1972). Arch. Biochem. Biophys. 152, 496-501. Rehfeld, D. W., and Tolbert, N. E. (1972). J. Biol. Chem. 248, 4803-481 1. Rhodin, J. (1954). Thesis, Aktiebolaget Godvil, Stockholm. Riezman, W., Weir, E. M., Leaver, C. J., Titus, D. E., and Becker, W. M. (1980). Pluntfhysiol. 65, 40-46. Rigatuso, J. L., Legg, P. G. and Wood, R. L. (1970). J. Histochem. Cytochem. 18, 893-900. Robbi, M., and Lazarow, P. D. (1978). Proc. Nutl. Acad. Sci. U.S.A. 75, 4344-4348. Robbi, M., and Lazarow, P. D. (1982). J. Biol. Chem. 257, 964-970. Roberts, L. M., and Lord, J. M. (1981a). Eur. J. Biochem. 119, 31-41. Roberts, L. M., and Lord, J. M . (1981b). Eur. J. Biochern. 119, 43-39. Roggenkamp, R., Sahm, H., Hinkelman, W., and Wagner, F. (1975). Eur. J. Biochem. 59, 231-236. Roels, F., Geerts, A., and Goldfischer, S. (1982). Ann. N.Y. Acud. Sci. 386, 534-536. Ruis, H. (1979). Can. J. Biochem. 57, 1122-1 130. Russo, J. J., and Black, V. H. (1982). Ann. N.Y. Acud. Sci., in press. Schiefer, S., Teifel, W., and Kindl, H. (1976). Hoppe-Seyler’s 2. Physiol. Chem. 357, 163-175. Schnarrenberger,C., Zehler, H., and Fitting, H.-H. (1980). Hoppe-Seyler’sZ. Physiol. Chem. 361, 328-329. Schopfer, P., Bajracharya, D., Bergfeld, R., and Falk, H. (1976). Plunra 113, 73-80. Silverberg, B. A. (1975). Protoplusmu 83, 269-295. Stabenau, H. (1974a). Plant Physiol. 54, 921-924. Stabenau, H. (1974b). Pluntu 118, 35-42. Stabenau, H. (1976). Ber. Dtsch. Bor. Ges. 88, 469-471. Sturm, A. J., and Kindl, H. (1982). Unpublished. Tanaka, A., Osumi, M., and Fukui, S. (1982). Ann. N.Y. Acud. Sci. 386, 183-199. Tchang. F., Robert, D., and Mazliak, P. (1981). Physiol. Veg. 19, 9-22. Tester, C. F. (1976). Plant Sci. Lett. 6, 325-333. Theimer, R. R., Anding, G., and Schmid-Neuhaus, B. (1975). FEBS Lett. 57, 89-92. Theimer, R. R., Anding, G., and Matzner, P. (1976). Plunru 128, 41-47. Thomas, J., and Trelease, R. N. (1981). Protoplusma 108, 39-53. Tolbert, N. E. (1971). Annu. Rev. Plant Physiol. 22, 45-74. Tolbert, N. E. (1981). Annu. Rev. Biochem. 50, 133-157. Tolbert, N. E. (1982). Ann. N.Y. Acad. Sci. 386, 254-268.

BIOSYNTHESIS OF MICROBODIES

229

Tsaftaris, A. S., and Scandalios, J. G . (1981). Dijferenriurion 18, 133-140. Vigil, E. L. (1973). Subcell. Biochem. 2, 237-285. Walk, R. A., and Hock, B. (1976). Eur. J . Biochem. 71, 25-32. Walk, R. A,, and Hock, B. (1978). Biochem. Biophys. Res. Commun. 81, 636-643. Walk, R. A,, Michaeli, S., and Hock, B. (1977). Planfu 136, 21 1-220. Wanner, G . , and Theimer, R. R. (1982). Ann. N . Y . Acud. Sci. 386, 269-284. Warren, G., and Dobberstein, B. (1978). Nature (London) 273, 569-571. Weir, E. M., Riezman, H., Grienenberger, J.-M., Becker, W. M., and Leaver, C. J . (1980). Eur. J . Biochem. 112, 409-477. White, J. E., and Brody, M. (1974). FEBS Left. 40, 325-330. Wickner, W. (1979). Annu. Rev. Biochem. 48, 23-45. Van der Wilden, W., Herman, E. M., and Chrispeels, M. J. (1980). Proc. Nut/. Acud. Sci. U.S.A. 77, 428-432. Wood, D. C., Jurgensen, S. R., Geesin, J . C., and Harrison, H. (1981). J . B i d . Chem. 256, 2377-2383. Woodward, J., and Merrett, M. J . (1975). Eur. J . Biochem. 55, 555-559. Yokota, S., and Fahimi, H. D. (1980). Eur. J . Cell Biol. 22, 165. Yokota, S., and Nagata, T. (1974). Hisrochemistry 40, 165-174. Yoo, B. Y., and Chrispeels, M. J. (1980). Proloplusmu 103, 201-204. Zilversmit, D. B., Entenman, C., and Fishler, M. C. (1943). J . Gen. Physiol. 26, 325-331. Zimmermann, R., and Neupert, W. (1980). Eur. J . Biochem. 112, 225-333. Zimmermann, R., Paluch, U.,Sprinzl, M., and Neupert, W. (1979). Eur. J . Biochem. 99,247-252. Zimniak, P . , Hartter, E., and Ruis, H. (1975). FEBS Lett. 59, 300-304.

This Page Intentionally Left Blank

INTERNATIONAL REVIEW OF CYTOLOGY. VOL. KO

Immunofluorescence Studies on Plant Cells C. E. JEFFREE,M. M. YEOMAN,AND D. C. KILPATRICK" Department of Botany, University of Edinburgh, and *Regional Blood Transfusion Service, Royal Infirmary, Edinburgh, Scotland 1. Introduction . . . . ....................... 11. Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Preparation of Antibodies

B. Preparation of Plant Tissues for Immunofluorescence Microscopy . . . . . . . . . . . . . . . . . . . ................. C. Fluorescence Microscopy of Plant 111. Applications of Immunofluorescence Microscopy to Studies of Plant Cells and Tissues ............................ A. Localization of Enzymes in Plant Tissues . . . . . . . . . . . . . . . . . B. Localization of Lectins in Plant Tissues.. . . . . . . . . . . . . . . . . . C. Localization of Other Plant Proteins in Plant Tissues.. . . . . . . IV. Conclusions . . . . . . . . . .. References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

23 1 232 232 235 242 246 246 249 260 262 263

I. Introduction More than a decade has now elapsed since the first publications appeared on the application of immunofluorescence procedures to plant cells. Knox et al. (1970) localized pollen cell wall antigens by means of antibodies raised against pollen diffusates. This and the almost concurrent paper by Graham and Gunning (1970) on the localization of legumin and vicillin in the cells of bean cotyledons by immunofluorescent microscopy heralded the beginning of an interest in immunofluorescent techniques by plant scientists which has since been sustained. Prior to these pioneering papers other immunlogical procedures had been applied to plants. Lewis ( I 952) used serological methods to identify incompatibility substances in Oenothera organgereus, while Linskens (1960) used a variety of immunodiffusion techniques in his studies of incompatibility in Petunia. Nasrallah and Wallace ( 1967) also used immunodiffusion to detect antigens attributable to alleles at the S locus in the stigmas of cabbage, Brassica oleracea. Serological procedures have also been used extensively by chemotaxonomists since, and continue to be exploited (see Smith, 1976, for recent review). However, in this article, attention will be focused on techniques in which a specific antibody is attached by direct or indirect means to a fluorescent dye, and the fluorescent complex used to locate, and in some instances to quantify, the antigen to which the antibody has been raised. 23 I Copyright 0 19x2 by Academic Press, Inc All right\ of rcproduction in any form reserved. ISBN 0-12-364480-0

232

C. E. JEFFREE ET AL.

lmmunofluorescent techniques have now been applied to a variety of plant species, including algae, fungi, gymnosperms, and angiosperms. Perhaps it is not surprising that the majority of the antigens investigated have been proteins, although carbohydrates and other molecular species, such as plant secondary products, could be employed as antigens, either directly or after suitable treatment. Particular attention has been afforded to storage proteins, enzymes, lectins, and tubulin. The techniques applied to immunofluorescence microscopy are without exception based on those widely and successfully employed with microorganisms and animal cells in diagnostic pathology and medical research. The wealth of background experience in these fields is invaluable to the botanical application of immunofluorescence also (see reviews by Nairn, 1969; Kawamura, 1977). 11. Techniques

A. PREPARATION OF ANTIBODIES 1. Purification of Antigens

Antigen preparations used for raising antibodies may be crude mixtures of macromolecules, partially purified proteins or complex carbohydrates, or highly purified proteins or carbohydrates. Although, for some purposes, homogenates or crudely fractionated material might suffice, it is generally desirable to purify the antigen as highly as possible. Some antigens are much more immunogenic than others and it is not always possible to predict which protein(s) in a complex mixture will elicit the strongest response in a particular animal. Nor is it necessarily the most abundant protein in a mixture that elicits the greatest antibody production. An interesting illustration of this is provided by Muller and Gerisch (1978) who raised antibodies to a crude mixture of cellular membranes; the antibodies raised were mainly against a single glycoprotein of MW 80,000 that is quantitatively a minor component of the plasma membrane. In general, purified macromolecules are desirable as antigens since the results obtained are less likely to be ambiguous in interpretation. Even when a single homogeneous polypeptide or carbohydrate is used as antigen, unless it is extremely small, a heterogeneous group of antibodies is likely to be synthesized in response, each antibody specific for a single antigenic determinant. This is not normally a disadvantage, unless it is hoped to distinguish between two structurally similar antigens. Even cross-reactivity is not an insoluble problem, however, as the antiserum to one antigen can be absorbed out with the cross-reacting antigen to provide a specific antiserum. However, this may not be of much practical value if the two cross-reacting entities are very closely related as the final specific antibody preparation might well be of too low an activity to be useful.

IMMUNOFLUORESCENCE STUDIES ON PLANT CELLS

233

Antigens may be purified from mixtures by a combination of protein separation methods, bften involving a large number of steps, or by one or two fractionation steps followed by affinity chromatography. The former is much more time-consuming and laborious, and often does not result in an entirely pure product. Affinity chromatography should be used whenever possible. Even if complete purification is not achieved by the affinity step, it is often a simple matter to remove contaminating proteins by other chromatographic procedures such as gel filtration or ion-exchange. It should be stressed that “purified proteins” are rarely 100% pure or even 95% pure, whatever methods have been employed for their isolation. The usual criteria of homogeneity, such as electrophoretic evidence, do not guarantee against minor contaminants. Even an impurity constituting less than 1% of the preparation could be of major relevance to raising antibodies, especially if it were highly immunogenic and the major antigen only weakly so. The best antigens are polypeptides and carbohydrates, and generally they have to be of a size corresponding to a molecular weight of at least 2000 and preferably over 5O00. Sometimes antibodies to a low-moelcular-weight entity such as a lipid are required, and such haptens have to be coupled to a suitable carrier. A procedure successfully adopted in this laboratory employs paraminobenzoic acid, linked to the antigen (Beiser et al., 1960), as an intermediary between the antigen and human (or bovine) serum albumen (Langone et al., 1973). This method has enabled the production of antiserum against the plant secondary product capsaicin (M. Aitken, 1982, personal communication), and would appear to have a promising future in plant immunocytochemistry. 2. Ruising Antibodies Rabbits, goats, and sheep are perhaps the most suitable and widely used animals for the production of specific antisera. All produce a relatively large quantity of antiserum, are easy to inject and bleed, and are readily available. The rabbit is usually the laboratory animal of choice since it is cheaper and easier to accommodate, and generally yields enough antiserum for most laboratory purposes (Herbert, 1978). Using a rabbit, it is generally sufficient to inject 1-10 p.g of highly purified antigen at a time, but if the antigen is nontoxic considerably more may be injected on a single occasion. If a crude mixture of macromolecules is to be used, the equivalent of a milligram or more of protein should be injected each time. The antigen should be dissolved (suspended) in isotonic saline or buffered saline and sterilized by a passage through a Millipore filter. It is common to inject antigens in the presence of an adjuvant, a term used to describe a group of chemically unrelated substances including aluminium hydroxide and Bordetella pertussis antigens that have the ability to stimulate the immune response in a nonspecific way. The mycobacterial-based F r e u d ’ s complete adjuvant is com-

234

C. E. JEFFREE ET AL.

monly used to help elicit a primary response, and subsequent injections are usually given in Freund’s incomplete adjuvant, an emulsifying oil without the bacterial component. Immunization programs vary, but typically a rabbit would be injected subcutaneously at multiple sites on its neck or back with antigen in the presence of Freund’s complete adjuvant to elicit a primary response. The procedure may be repeated within the next 2 or 3 days. After 2 to 4 weeks, the rabbit is given a further set of multiple site injections using antigen emulsified in Freund’s incomplete adjuvant to elicit a secondary response. The animal may be bled from an ear vein about 10 days later, although it is not uncommon to give two to three booster injections at around 2 weekly intervals before bleeding. These methods are more comprehensively described by Clausen (1971), Garvey et al. (1977), and Herbert (1978). Mice can yield only small volumes of antisera by normal bleeding techniques. However, their yield can be enhanced by the technique of inducing peritoneal irritation with ascitic tumor cells. The ascitic tumor cells invade the peritoneum and local organs, causing discharge into the peritoneal cavity of relatively large quantities of ascites fluid which can be drawn off repeatedly, at intervals of a few days. The inoculum is prepared with adjuvant in the usual way, but injected peritoneally, with a subsequent subcutaneous booster. The introduction of ascites is timed to coincide with the rise in the antibody titer a few days later (Francki and Habili, 1972). 3 . Purification of Antibodies

The serum obtained from the blood of an immunized rabbit may be used on tissue sections without further treatment but sometimes this gives rise to a substantial degree of nonspecific, background staining. This problem can sometimes be overcome simply by diluting the antiserum. More usually, however, it is necessary or at least advantageous to partially purify the immunoglobulins. A crude immunoglobulin fraction can readily be obtained by precipitation using ethanol or ammonium sulfate (Deutsch, 1967) or by ion-exchange chromatography using DEAE-cellulose (Levy and Sober, 1960; Stansworth, 1960). Often a combination of salt fractionation and ion-exchange is employed. A more recent technique is to employ affinity chromatography using immobilized Protein A, a cell surface protein from Staphylococcus aureus having the property of binding to the Fc region of immunoglobin from human, rabbit, and several other mammalian sources (Goding, 1978). The above procedures all yield a reasonably pure immunoglobulin fraction, but nonspecific immunoglobulins copurify with specific antibodies. Better results in this respect can be obtained by affinity chromatography using immobilized antigen. Typically, the antiserum is applied to a column of antigen coupled to Sepharose beads; the column is then washed thoroughly with buffered saline; and finally the specific antibodies are eluted with sodium thiocyanate solution or a buffer of low pH. Examples include the

IMMUNOFLUORESCENCE STUDIES ON PLANT CELLS

235

preparation of specific antibodies to rabbit y-globulin in goats using SepharoseIgG columns and eluting with 0.1 M acetic acid (Kyte, 1974), and the preparation of rabbit anti-(propylhydroxylase) antibodies by immunoadsorption on agarose-enzyme followed by elution with 3 M sodium thiocyanate (Olsen ef al., 1973). An alternative procedure for obtianing specific antibodies that is gaining in prominence is the monoclonal antibody technique origianlly described by Kohler and Millstein (1975) and recently reviewed by Secher (1980). The principle of the technique is illustrated in Fig. 1. Briefly, antibody-producing cells are fused to neoplastic (myeloma) cells to produce a hybrid potentially capable of perpetual production of specific antibody. The potential of the technique is enormous and is likely to replace conventional antiserum production methods for commercially produced antibodies. However, the procedure is enormously expensive of labor and materials, and this is likely to make monoclonal antibodies inaccessible to the average user for the majority of routine immunochemical procedures. Checking the Purity of Antibodies. The double diffusion method of Ouchterlony (1958) is the most widely used test of antibody specificity on account of its simplicity. If purified and unfractionated antigen preparations placed in adjacent wells yield a single line of identity in reaction with the antiserum, it is usually taken as sufficient proof of specific antibody production to a purified antigen (see Fig. 6). However, it should be mentioned that multiple bands need not necessarily mean that the antigen used was impure, nor does a single line unequivocally indicate purity of antigen. The former could arise, for example, from a pure antigen with a tendency to aggregate; the latter might be observed, for example, when a separate (soluble) antigen-antibody complex is formed in antigen excess. The antibody titer may be estimated by the Bentonite flocculation test (Bozicevich er al., 1963). In this procedure the purified antibody is adsorbed on to the surface of Bentonite particles in aqueous suspension, and stained with methylene blue to aid their subsequent visualization. The adsorbed antibody preparation is placed in the wells of a microtiter plate, and tested by the addition of serial dilutions of the antigen. If the antibody is specific, and the antigen present in sufficient quantities, the suspended Bentonite will be cross-linked, and thus agglutinated, in much the same manner as in a hemagglutination assay of lectins. OF PLANTTISSUESFOR IMMUNOFLUORESCENCE B . PREPARATION MICROSCOPY

1. Fixation Labeled antibodies may be unable to penetrate intact, unfixed, cell and organelle membranes, and soluble antigens can be lost to the staining and wash-

236

a

C . E. JEFFREE ET AL. Malignant myeloma cells

Antigen

Mixed antibodies Selection

of hybrids

1

Mixed +hybrid population

Clone

Hybrid clone Monoclonal antibodies FIG. 1 . Schematic comparison of conventional and monoclonal methods of antibody production. In the conventional procedure of animal inoculated with a purified antigen carrying several antigenic determinants produces several lymphocyte lines each of which secretes an immunoglobulin specific for a single antigenic determinant. Immunoglobulins purified from the serum by, for example. immunoabsorption chromatography will be heterogeneous, and will recognize any molecules which have antigenic determinants in common with the original antigen. Monoclonal antibodies are produced by fusion of lymphocytes of the immunized animal with malignant myeloma cells. The resulting somatic hybrid cells have differing metabolic requirments from either of the parent cell lines, and can be selected and isolated by their survival on specially designed media. The hybrids are heterogeneous at this stage, and secrete a range of immunoglobulins. lndividual cells are selected and cloned. Each clone secretes a single immunoglobulin, specific for a single antigenic determinant.

IMMUNOFLUORESCENCE STUDIES ON PLANT CELLS

237

ing media during incubation. Cells and tissues must ordinarily be fixed before incubation with antisera. Certain soluble antigens, such as carbohydrates, may not respond to conventional fixatives, which are chiefly employed to immobilize the protein matrix of cells by cross-linking or coagulation of proteins. It should be noted that chemical fixatives may result in sufficient change in protein structure to alter their antigenic properties. For general plant ultrastructural studies, glutaraldehyde is usually regarded as the fixative of choice (see, e.g., O’Brien and McCully, 1981). Graham and Gunning (1970) and Barlow et a/. (1973) have used glutaraldehyde in immunocytochemical fixation with apparent success. For immunocytochemical purposes, however, glutaraldehyde has two major contraindications. First, when used at concentrations sufficient to ensure satisfactory fixation, major losses of antigenicity result (Kraehenbuhl et a f . , I980), especially with protein antigens. Second, glutaraldehyde fixation results in substantial yellow-green autofluorescence of tissues, masking the specific fluorescence of FITC. For any quantitative analysis of staining intensity glutaraldehyde must be excluded on these counts. Fulcher and Holland (197 1) also found acrolein to be detrimental to staining intensity in fungal hyphae (Ophiobolus graminis). Solvent fixatives, ethanol, or methanol are commonly used with animal tissues for fixing protein antigens. Hapner and Hapner (1978) used 95% ethanol on seeds and roots of Sanfoin (Onobrychis viciifolia). Solvent fixatives may overcome masking of antigens due to lipid membranes. However, we have found ethanol to opacify Datura seed tissues, resulting in excessive light scattering and resolution loss. Ten percent formalin or paraformaldehyde in phosphate buffer or PBS gives good results at the light microscope level in most plant tissues. Recent advances in immunocytochemical fixation include new permeabilization fixatives employing mixtures of ethyldimethylaminopropyl carbodiimide and glutaraldehyde, which give good ultrastructural preservation while retaining antigenicity of proteins, and rendering the cytoplasm matrix permeable to globulins (Willingham and Yamada, 1979; Willingham, 1980; Willingham et a f . , 1981). This technique enables incubation of intact cultured cells and tissues, and should be equally applicable to plant and animal tissue, perhaps permitting whole plant apices to be stained with fluorochrome or electron-opaque probes prior to embedding and sectioning. 2 . Embedding and Sectioning Animal tissues are typically prepared for immunofluorescence microscopy by freeze sectioning them with steel or long edged glass (Ralph) knives (Lindner and Richards, 1978; Bennett et a/., 1976). and by fixation in media which preserve antigenicity without sacrifice of fine structure. Although plant material can be cryosectioned successfully, many soft, vacuolate tissues disintegrate when cut thinly and protoplasts may be lost from the cut

238

C. E. JEFFREE ET AL.

cells. However, some authors claim sections as thin as 3 Fm (Knox, 1970; Jacobsen and Knox, 1973, barley aleurone cells) and 4 Fm (Vreeland, 1972, Fucus distichus) from amenable tissues. Resolution of the fluorochrome distribution is inevitably coarsened in thick sections. Furthermore penetration of cells by the relatively large antibody moelcules may be inhibited by intact membrane systems in thick sections, masking antigens. Although the simplicity of the cryostat method has resulted in its widespread use for plant immunofluorescence studies, techniques for embedding in waterpermeable resins, such as glycol methacrylate (GMA) offer substantial advantages in terms of improved resolution (Cole and Sykes, 1974; Bennett et a l . , 1976; Tippet and O’Brien, 1975; Fulcher and Holland, 1971; Carnegie et a l . , 1980). However, as we shall see later, the limited resolution of the light microscope imposes insurmountable constraints on the usefulness of the system for localization of molecules at the subcellular level. In the procedure used by Craig et al. (1979) pea seeds fixed in paraformaldehyde were both dehydrated and embedded directly in GMA at or below 0°C. the GMA was UV polymerized at - 10°C and 0.5- to 3-pm-thick sections were stained with TRITC using an indirect immunofluorescent procedure. Purification of the BMA was necessary to prevent nonspecific binding of TRITC to the polymerized plastic. The inherent simplicity of this method, and the quality of the results produced, should recommend it to botanists. The method is readily adaptable to a freeze-substitution approach to tissue fixation. Since ice is soluble in GMA at subzero temperatures the specimen need not be brought above 0°C until the incubation steps are carried out. Furthermore the same blocks could also be stained and sectioned for electron microscopy using appropriate electronopaque probes. 3. Staining Procedures for Immunojluorescence Microscopy The principles and precedures for immunofluorescence microscopy of plant tissues are essentially identical to those which were developed about 30 years ago, primarily in medical research laboratories, and which are now in widespread research and diagnostic use. Two major reviews, by Nairn (1968) and Kawamura (1977), are invaluable sources of background and data on this subject. Two principal approaches to immunofluorescence microscopy are commonly employed. In the “direct” or “fluorescent antibody” method (Fig. 2) the specific antibody is located by conjugating it directly with a suitable fluorescent dye (fluorochrome). The conjugated antibody is incubated with the specimen, and after washing to remove excess antiserum, the bound conjugate is visualized by excitation of the fluorochrome in light of an appropriate wavelength. The method is developed from procedures described by Coons et al. (194 1) and Coons and Kaplan ( 1950).

239

IMMUNOFLUORESCENCE STUDIES O N PLANT CELLS Direct method

Fluorochrome

Indirect method

Antibody

*-+-))

Prepared tissue

7 loantibody

x

Fluorescent antibody

Incubate 30 min, 2OoC

I I

Incubate 30 min,

I

I

vAntigen/antibady complex

*OoC

Wash

FIG. 2. Principles of the direct and indirect methods of immunofluorescent staining. In the direct or fluorescent antibody method a purified specific antibody is conjugated with a fluorochrome and the conjugate is incubated with a tissue specimen. The location of the antigen is revealed by fluorescence of the conjugate. Since only one antibody is used the method is highly specific. The indirect method employs an unconjugdted primary antibody which binds with the antigen. The location of the primary antibody/antigen complex is visualized by means of a fluorochrome-conjugated secondary antibody raised against the primary antibody in another animal species. The indirect method has the advantage over the direct method that a greater molar ratio of fluorochrome to antibody may be achieved without jeopardizing the specificity of the primary antibody. A disadvantage is that since immunoglobulins of two animal species are employed the scope for nonspecific staining may be increased.

240

C . E. JEFFREE

ET AL

The direct method has the main disadvantage that each antiserum must be separately conjugated with the fluorochrome. The specificity and specific activity of the antiserum can be markedly reduced by the process of conjugation with the fluorochrome, especially if the fluorochrome:protein ratio is high (Kawamura, 1977), and this can result in increased nonspecific staining. Nevertheless a carefully made conjugate will give greater specificity of staining than with indirect methods, which involve the separate and often differing nonspecific staining properties of two globulin preparations (Nairn, 1969; Kawamura, 1977). In the “indirect” or “sandwich” method, attributable to Weller and Coons (1954) (Fig. 2), tissue sections are incubated with an unmodified primary specific antibody, which may, for example, have been raised in a rabbit. After washing, to remove unbound immunoglobulins, the section is then incubated in a secondary labeled anti-rabbit antibody solution raised in, e.g., goat inoculated with purified rabbit immunoglobulins. The indirect method offers much greater sensitivity than the direct method, due to the fact that for each reactive site on an antigen available for binding of a primary antibody, perhaps 4 to 10 more are available to the secondary antibody, and the brightness of the fluorescent image is therefore increased. However, the importance of the control of nonspecific binding by the primary antibody is magnified by a corresponding amount. It is interesting to note that the limit of detection of molecules by direct immunofluorescence methods may be about 0.5 to 1 .O x lo4 molecules per cell, or about 1 .O X l o p 4 p,g/mm2 of a 5-mm-thick tissue section (Coons, 1956; Pressman et al., 1958). Recent developments in immunochemical staining procedures have resulted in some quite different approaches which may have promising application to plant immunochemistry, although little work in these areas has appeared to date. In one of these procedures labeled Protein A from S . aureus is used as a “secondary labeled antibody” in the indirect system, employing its ability to bind in a specific manner with the Fc portion of mammalian IgG molecules. Roth et al. (1 980) employed an FITC-Protein A-colloidal gold complex to produce a staining system suitable equally for light and electron microscopy, and permitting a correlative approach. In a more complex procedure, the ferritin bridge technique, described by Willingham (1980) and Willingham et al. (1971, 1981), unlabeled intermediate specific antibodies are used to link unconjugated ferritin label with the primary antibody (Fig. 3). Willingham er al. (1981) localized actin in culture fibroblasts with this procedure, which has the following incubation steps: (1) affinity-purified rabbit anti-(actin); (2) goat anti-(rabbit IgG); (3) affinity-purified rabbit anti-(horse ferritin); and (4) horse spleen ferritin. The ferritin bridge technique is claimed to be highly sensitive. primarily due to the exclusive use of unconjugated antibodies with high specific activity (Willingham, 1980; Willingham et al., 1981). Because of the large number of steps affinity purification is an essential step with each reagent in order to reduce

IMMUNOFLUORESCENCE STUDIES ON PLANT CELLS

24 1

FIG. 3. A diagram of the ferritinlfluorochrome bridge procedure. The steps are as follows. (a) Tissue is incubated with affinity purified IgG raised in rabbit. (b) Goat anti-(rabbit IgG) is used as a bridge between the primary antibody and a tertiary rabbit antiserum (c). raised against ferritin (for immunoelectron microscopy) or against FITC-conjugated mouse IgG (or another protein conjugated with FITC) for immunotluorescence. Finally, (d) the tissue is incubated with the appropriate labeling molecule. This method makes it possible to localize antigens without conjugating any of the specific antibodies used.

242

C. E. JEFFREE

ET AL.

nonspecific staining of tissues. Although ferritin is not suitable as a probe for light microscopy, there is in principle the possibility of using a fluorochrome attached as a ligand to a suitably antigenic protein molecule as the final label, and raising the tertiary antibody against that. A suitable labeled secondary antigen might be FITC conjugated with the immunoglobulins of a third mammal. The proposed fluorescent bridge technique would thus involve four incubation steps as follows: (1) affinity-purified rabbit anti-(X); (2) goat anti-(rabbit IgG); ( 3 ) affinity-purified rabbit anti-(FITC-conjugated mouse IgG); and (4) FITC-conjugated mouse IgG. The procedures are summarized in Fig. 3 . As a general principle, affinity purification of antibodies is a highly desirable step in immunofluorescence studies. Although crude homogenates and exudates containing multiple antigens have been employed in “tissue-typing”approaches to problems in plant taxonomy (Raff et a l . , 1979) no meaningful cytochemical data can be derived from such crude analysis, principally because, even if the composition of the antigen mixture were accurately known, the wide variation in antigenicity of different constituents will result in an unpredictable antibody specificity. The nature of the labeled cell constituents will therefore be in doubt. Techniques for the affinity purification of antisera are described by Cooper (1977, chapter 7) and Kawamura (1977, chapter 3 ) . C. FLUORESCENCE MICROSCOPY OF PLANTTISSUES 1. Basic Principles of Fluorescence Microscopy Most modem fluorescence microscopes are based on epiillumination principles described by Ploem (1967, 1971). In Ploem’s system, the excitation light, generally obtained from a high-pressure mercury vapor source, is filtered via solid glass or dichroic excitation filters, and is passed into the microscope objective lens via a dichroic mirror above the lens and at 45” to the lens axis (Fig. 4). The objective lens is thus utilized as a simple form of dark field consenser. Since the dichroic mirror is designed to transmit to the eyepiece light of wavelengths outside the range used for excitation the same mirror thus separates from the image-forming light any reflected excitation light, and the final image consists entirely of fluoresced wavelengths. Further filtration of the fluoresced wavelengths can be achieved by barrier filters if it is necessary to distinguish between autofluorescence and specific fluorescence at distinct wavelengths. Barrier filters can be of dyed glass, interference (dichroic), or made of dyed gelatin or optical resins (e.g., Kodak Wratten filters). The most commonly used fluorescent dyes for immunofluorescence are isothiocyanates of fluorescein (FITC) and members of the rhodamine group (RITC, TRITC), chosen for maximum separation between absorption and fluorescence wavelength maxima, and for high quantum efficiency. FITC has an absorption maximum at 495 nm and fluoresces in the

IMMUNOFLUORESCENCE STUDIES ON PLANT CELLS

243

Observed fluorescence

4 Microscope eyepiece

High pressure mercury vapor lamp

FIG.4. Diagram of the epiilluminating fluorescence microscope, the principle of which is attributable to Ploem (1967). showing how the fluoresced light is separated from wavelengths used for excitation by means of a 45" angled dichroic reflector.

yellow-green, about 500 nm. The rhodamines absorb in the blue-green (ca. 540 nm) and fluoresce in the orange and red, between about 570 and 600 nm. Problems can arise due to autofluorescence of plant tissues with either system (Fig. 5). Most biological materials autofluoresce slightly in the yellow-green when excited by blue or violet light, and this autofluorescence can be further provoked by glutaraldehyde fixation. Plant tissues are no exception, but while the autofluorescence from cellulosic cells may be of low intensity and may cause little difficulty, lignified tissues, and tissues with phenolic deposits fluoresce strongly, competing in intensity with specific fluorescence from FITC (Fig. 5).

244

C. E. JEFFREE ET AL.

FIG. 5 . Autofluorescence in plant tissue. (a) Paraformaldehyde-fixed section of a tomato stem, illuminated in blue-violet light (350-500 nm) using an FITC dichroic filter and viewed via yellow barrier filters GG9 and OG4 which transmit wavelengths > 500 nm. Bright red autofluorescence is present from chlorophyll in addition to yellow-green fluorescence from lignified cells. (b) The same field viewed via a Wratten 58 green barrier filter. The red autofluorescence from chlorophyll is eliminated. Green autofluorescence is visible from lignified cells and from cytoplasm. (c) Paraformaldehyde-fixed section of stem cortex of Canavalia erisiformis. illuminated in green light of about 540 nm using a dichroic filter designed for use with rhodamines. and viewed via a red barrier filter transmitting wavelengths > 590 nm. Bright red autofluorescence is emitted from chloroplasts, and illuminates cell walls by reflection. (d) The vascular region of the same section as in (c) shows red autofluorescence from lignin, in addition to that from chloroplasts.

Some authors claim to be able to distinguish visually between the pale greenish tissue autofluorescence and the distinctly yellowish specific fluorescence of FITC (Jacobsen and Knox, 1973; Hattersley el a/., 1977). For the purposes of measurement and photography, however, the two wavelengths are essentially indistinguishable, and this can be a major disadvantage. Illumination in both blue-violet light (FITC system) and in the green (rhodamine system) causes chloroplasts to fluoresce intensely in the red, at about 670 nm. This autofluorescence is readily filtered out of the FITC image by means of green barrier filters such as Wratten 58 or 61, (Fig. 5a and b) but the closeness of the chlorophyll autofluorescence to the orange-red specific fluorescence from rhodamine makes

IMMUNOFLUORESCENCE STUDIES ON PLANT CELLS

245

their separation by filtration difficult, and seriously impairs the usefulness of rhodamines in immunofluorescent studies of photosynthetic plant tissues. Hapner and Hapner (1978) concluded on the basis of studies of sanfoin (Onobrychis vicifolia) roots that plant tissue exhibits little or no autofluorescence when illuminated in green light. However, our studies on Daturu stramonium stems showed lignified and phenolic-rich tissues also autofluoresce strongly in green light (Jeffree and Yeoman, 1981). The only advantage rhodamines appear to have over FITC for work with plant tissues is their substantially greater resistance to fading in the excitation beam. Stilbene isothiocyanate may provide a useful alternative to FITC and rhodamines. Since it fluoresces in the blue with UV excitation, the red and green autofluorescence of plant tissues may be readily susceptible to barrier filtration. However, this dye has not been evaluated for use with plant tissue at the time of writing. 2 . Quantitation of Specific Fluorescence Quantitation of in situ antigens by immunochemical techniques is an attractive possibility, but one which present a number of very demanding technical difficulties. It has already been noted that aldehyde fixatives, especially glutaraldehyde and acrolein, can alter the nature of the antigen-antibody interaction. Yet the tissue and the antigen must clearly be fixed, not only to stabilize the cell matrix and eliminate turnover, but to immobilize the antigen within it. Loss of soluble antigens may occur during fixation steps which involve immersion of tissues in aqueous media. This problem may be alleviated by freeze substitution techniques, expecially for protein antigens but is not inevitably resolvable in this way because the specimen must pass into a solvent-based fix at some stage in the procedure. If the antigens survive fixation unchanged a number of physical difficulties can result in their underestimation. Since immunoglobulins are relatively large molecules their progress will be impeded by intact membrane systems, which may act as masks, preventing the labeling of enclosed antigens. Similarly, “steric” masking may occur in crystalline arrays of proteins, as are found in crystal-containing bodies [microbodies (Tolbert, 1971; Richardson, 1974) and cytosomes (Sitte, 1965)l. This type of difficulty may sometimes be resolved by use of fixatives of the types described by Willingham et al. (1981). Conversely antisera may be unable to distinguish between inactivated or stored enzymes, and active enzymes. Consequently the technique is inapplicable to the estimation of amounts of active enzyme in cells (see Section 111,A). Closely related enzymes or isozymes may have sequences which are recognized as antigenically similar by the antisera. In practice this is not commonly recorded (some of the isozymes of cellulase are, for example, antigenically distinct), but this problem has made it impossible for us to distinguish between the lectins in grafts between Solanaceous species in which the lectin proteins are highly conserved and show antigenic similarities (Kilpatrick et ul., 1980). Tech-

246

C. E. JEFFREE ET AL

niques for measurement of specific fluorescence have been described by Kawamura (1977) and Yamamoto et a l . , (1970). The technique is based on measurement of excitation light intensity, and the comparison of fluorescent intensity of the specimen with that of a known standard. It is a major disadvantage of fluorochrome probes in quantitative immunocytochemistry that the specific fluorescence fades rapidly on exposure to the excitation beam. Although TRITC is less rapidly faded than FITC, nonfluorescent probes such as peroxidases, which are visualized as nonfading colored reaction products by bright field microscopy, offer the potential of their straightforward quantiation by microdensitometry.

111. Applications of Immunofluorescence Microscopy to Studies of Plant Cells and Tissues A. LOCALIZATION OF ENZYMES IN PLANTTISSUES Only limited attention has been given to the localization of enzymes in plant cells using immunofluorescence procedures. So far no attempt has been made to measure the amount of enzymic protein present at a particular site. The quantitation of a particular catalytic protein in a cell or subcellular organelle is an attractive proposition but is extremely difficult, if not impossible, to achieve and the difficulties are further compounded by problems associated with the measurement of fluorescence. The major stumbling block is that different forms of an enzyme both inactive and active may well possess the same antigenic properties and therefore will respond similarly to the antibody conjugate. If this is true for a particular enzyme then it is impossible to differentiate between active and inactive forms. Certainly there appears to be little doubt that at least two of the isoenzymes of cellulase, the 9.5 and the 4.5 form, do not have common antigenic determinants (Koehler et a l . , 1981) so that it is at least theoretically possible that different isoenzymes of the same enzyme may be distinguished by immunological procedures. Still, of course, bearing in mind the insurmountable hurdle of not being able to dintinguish between active and inactive forms, it would therefore appear that at best the technique provides a means of locating catalytic proteins which may or may not be in an active form. Four enzymes have received some attention: a-amylase, urease, cellulase, and ribulose- 1,5-bisphosphate carboxylase. 1. Localization of a-Amylase Jacobsen and Knox (1973) examined in some detail the cytochemical localization and antigenicity of a-amylase in barley aleurone and showed that Group A and Group B a-amylases were immunologically identical. a-Amylase was localized in fresh and fixed thin sections (2 pm) of aleurone using a substrate film

IMMUNOFLUORESCENCE STUDIES ON PLANT CELLS

247

technique and immunofluorescence. They used an indirect immunofluorescence procedure with sections of fresh tissue which were either used directly, or fixed and then air dried. The antiamylase serum was raised in rabbits to a highly purified a-amylase from barley and then used in conjunction with a goat antirabbit globulin serum labeled with FITC. Controls were made by using preimmunization serum in place of the antiamylase serum. Using this technique the distribution of a-amylase within the aleurone was demonstrated. Most were associated with the periphery of the cells. The green autofluorescence due to the cell walls interferes with this method and reduces the resolution markedly. The authors claim that the specific fluorescence emitted from the a-amylase conjugate is an intense yellow color and can be easily distinguished from the bright green autofluorescence of the walls; however, such a difference cannot be seen from the black and white photographs. An additional problem pointed out by the authors is that fixation, however brief, in an aqueous fixative (in this case formaldehyde) will remove the enzyme and consequently result in reduced fluorescence. Following on this pioneer work by Jacobsen and Knox (1973), Jones and Chen (1976) employed a slightly different approach with improved techniques in an attempt to localize a-amylase in the cells of barley aleurone. They used freeze-substituted paraffin-embedded sections (4-5 and 8- 10 pm). Antisera were raised in rabbits to highly purified antigens and these were used in conjunction with a goat anti-rabbit IgG fraction conjugated to RITC (rhodamine isothiocyanate). Controls were prepared routinely with sections incubated with staining buffer only, RITC goat anti-rabbit IgG only, rabbit anti-a-amylase IgG immune serum only, and nonimmune rabbit IgG plus RITC conjugated to goat anti-rabbit IgG. The red autofluorescence using this procedure was low, and compared favorably with the intense green autofluorescence using FITC. 2. Localization of Urease Urease is an enzyme which catalyzes the hydrolysis of urea to ammonia and carbon dioxide, and was the first enzyme to be isolated in a crystalline form, by Sumner (1926). Seeds of jack bean, Canavafia ensiformis, are a rich source of urease. Murray and Knox (1977) investigated the distribution of this enzyme within the cells of cotyledons from germinating seeds, by means of the direct immunofluorescent method. Urease was localized in transverse free hand sections of cotyledons, either imbibed and unfixed, or fixed in 4% paraformaldehyde. The sections were incubated with saline containing no glycosides, methyl a-D-glucoside alone, salicin alone, or with both glycosides. These additional pretreatments are essential in this case because the cotyledonary cells contain concanavalin A and a p lectin, both of which are capable of binding to the carbohydrate moieties of immunoglobulins. This binding could be totally abolished by pretreatment of the tissue sections with the inhibitory sugars for the two lectins, namely, methyl a-D-glucoside for concanavalin A and salicin for p

248

C. E. JEFFREE ET AL

lectins with p-glycosyl specificity. This again stresses the point that the application of immunofluorescence techniques must be preceded by studies on the properties of the tissue under investigation. After pretreatment the sections were suspended in saline solution containing an FITC-labeled specific immunoglobulin fraction and glycosides corresponding to the pretreatments. All treatments were finally washed with their respective pretreatment solutions and mounted in saline for fluorescence microscopy. High titer antisera to purified antigen were raised in rabbits, purified and assayed for antibody content by microcomplement fixation. The purified immunoglobulin fraction was conjugated with FITC, freed from unreacted FITC, lyophilized, and stored at 4°C. Controls were made by using preimmunization serum in place of the antiurease serum. Using this technique the distribution of urease in sections of cotyledons from germinating seeds was localized at two sites, within the cytoplasm of storage parenchyma cells in spherical granules and within the intercellular spaces also in spherical granules. Although similar in size (ca. 3 pm in diameter) the latter was distinguished from the cytoplasmic granules by the presence of the p lectins. This investigation of Murray and Knox (1977) is a model for those wishing to investigate the distribution of an enzyme using an immunofluorescent procedure. 3. Localization of Cellulase Immunochemical procedures have been used to localize cellulolytic enzymes (p- 1,4-glucan, 4-glucanohydrolase) in the tissues of higher plants (Bal et a l . , 1976; Sexton et al., 1980; Durbin et a l . , 1982). In studies on auxin-induced growth in etiolated pea epicotyles Bal et al. (1976) have identified, purified, and characterized two different species of cellulase and showed that these enzymes are immunologically different. Only brief mention of this and the more recent work of Sexton et al. (1980, 1981a,b) on the study of leaf abscission is made here because Bal et al. (1976) used ferritin-conjugated antibodies to the two cellulases and the other authors employed a peroxidase-conjugated goat antirabbit immunoglobulin, the bound peroxidase being localized by the technique of Graham and Karnovsky ( 1966) using 3,3'-diaminobenzidine. These techniques are extremely important but outside the scope of a review on immunofluorescence. 4. Localization of Ribulose-l,5-bisphosphateCarboxylase. Hattersley et al. (1977) have used antibodies to wheat and spinach ribulose-1,5-bisphosphate carboxylase (RuP,Case) to locate the enzyme in leaf tissues of 40 C, and C, species, and one crassulacean acid metabolism (CAM) plant using an indirect immunofluorescence procedure. It is known that the antiserum raised to RUP,Case from one species cross-reacts with enzyme from a wide variety of other species (Gray and Kekwick, 1974; Gray and Wildman, 1976). The antiserum was raised in rabbits against a highly purified preparation

IMMUNOFLUORESCENCE STUDIES ON PLANT CELLS

249

of RuP,Case from leaves of Triticum aestivum (C3). Serum specificity was analyzed qualitatively by Ouchterlony double diffusion. The indirect ( “sandwich”) method was employed on hand cut sections from segments of young fully expanded kaves previously fixed in 70% ethanol. Rinsed section (10- I5 km thick) were transferred to neat or diluted antiserum. After an incubation of 1 hour the sections were rinsed again and transferred to the FITC-labeled sheep antirabbit IgG. Sections were incubated for 1 hour in the dark, rinsed in PBS, and mounted in 50% glycerol (aqueous) containing 1% (w/v) thymol. Slides were kept in the dark until observation. For each antiserum labeling test, two controls were performed: ( 1 ) as described above, using normal or preimmunization serum instead of the specific antiserum, and (2) fixed and sectioned but mounted directly with no further treatment (autofluorescence control). Hattersley er al. ( 1977) were unable to label sections from fresh unfixed tissue, or with glutaraldehyde-fixed material. They state that ethanol treatment was essential for the penetration of chloroplasts by the antibodies. Difficulty was encountered in recording results on black and white film because the images obtained often failed to distinguish between specific fluorescence and autofluorescence of chloroplasts. Such differences were obvious from color transparencies. Using this elegant approach Hattersley et al. (1977) were able to demonstrate that in C, and CAM species, specific fluorescence is associated with the chloroplasts of all leaf chloroenchymatous cells, while in species with “classic” C, leaf anatomy the enzyme is located almost exclusively in the “bundle sheath.”

B. LOCALIZATION OF LECTINS I N PLANTTISSUES Lectins occur widely and abundantly throughout the plant kingdom. Despite being extensively used as tools to bind to animal cells, and despite being increasingly studied in their own right in recent years, the function of plant lectins is still unknown. Attempts to establish the tissue and subcellular location of lectins might therefore be of great value in helping to understand their function. Although subcellular location cannot be used alone to infer a particular function, any suggested function must be consistent with the distribution of the lectin within the plant and within the cells, and such information can be used to exclude some suggestions. Much interest has been focused on the lectin from the cellular slime mould, Dictyostelium discoideum. This protein, named discoidin (Simpson et al., 1974), exhibits distinctive developmental kinetics (Rosan et al., 1973) and appears to be essential to the developmental progress of the organism (Ray er al., 1979). Chang er al. (1975) used an indirect immunofluorescence method to demonstrate the presence of discoidin on the surface of aggregating cells, and its apparent absence from cells still in the vegetative stage of the life cycle. Both living cells and cells fixed in situ by brief exposure to glutaraldehyde exhibited surface

250

C. E. JEFFREE ET AL.

fluorescence when treated with rabbit anti-discoidin antiserum followed by FITC-conjugated goat anti-rabbit IgG. Fixed cells had a diffuse distribution of fluorescence; living cells had small patches that formed a single cap after several hours. These observations were confirmed by performing parallel experiments using a ferritin-labeled second antibody. Similar experiments have also been performed on pallidin, the corresponding lectin from the related organism, Polysphondylium pallidum (Chang er al., 1977). Similar results were obtained. A major problem associated with studying lectin distribution in higher plants is that some well-characterized lectins can bind to the carbohydrate moiety of immunoglobulins. Concanavalin A, and PHA from Phaseofus vufgaris, for example, will form a number of precipitation bands after double diffusion against rabbit serum, of which one is due to interaction with IgG (Fig. 6c-f). Clarke et al. (1975) exploited this drawback to their advantage when investigating lectin localization in cotyledon sections of jack bean and red kidney bean. Sections were reacted with FITC-conjugated immunoglobulin in the presence or absence of high concentrations of the appropriate saccharide inhibitor for the lectin under study. By this means it was concluded that the lectins from both species are associated with cytoplasmic organelles with no appreciable lectin external to the plasmalemma. Perhaps the most extensive studies have been made on lectins from Solanaceous plants, particularly Datura stramonium, and these studies well illustrate the applicability of immunofluorescent techniques to the study of plant antigens. The Datura lectin does not form precipitation lines in double diffusion against rabbit serum (Fig. 6b), and is therefore less prone to the nonspecific binding of lectins with immunoglobulins which was exploited by Clarke et al. ( 1975), but nonspecific staining is noticeably further reduced when purified immunoglobulin fractions are used in staining procedures instead of the crude antiserum. Kilpatrick et al. (1978) used formaldehyde-fixed sections of Datura seeds and a purified immunoglobulin fraction from a rabbit inoculated with highly purified Datura seed lectin to study the subcellular distribution of the lectin. Using the indirect immunofluorescenceprocedure, with FITC-conjugated goat anti-rabbit IgG as the secondary antibody, virtually all the specific fluorescence was found in the plasmalemma and throughout the cytoplasm of cotyledonary cells (Figs. 7 and 8). The staining pattern in the cytoplasm was not uniform, but occurred in association with the membranes of all identifiable organelles including the nuclear membrane. However, fluorescence was absent from the cell walls. These studies have been expanded by Jeffree and Yeoman (1981), who investigated the intracellular and intercellular distribution of Datura lectin in seeds, stem, and callus. Specific fluorescence was observed in all cell types examined, always in association with membranous structures (Figs. 7 and 8). In stems, epidermis, endodermis, and phloem were the richest sources possibly indicating

IMMUNOFLUORESCENCE STUDIES ON PLANT CELLS

25 1

FIG. 6. (a-c) Testing specificity of antisera by Ouchterlony double diffusion tests. (a) Purified rabbit anti-(Datum strarnonium seed lectin) IgG in the center row of wells is tested against purified D. strarnoniurn lectin (bottom row) and a crude extract of potato proteins containing lectins (upper row). (b) FITC-conjugated goat anti-(rabit IgG) IgG in the center well tested against purified D. strarnoniurn lectin showing freedom of nonspecific binding of the lectin to goat immunoglobulins. (c) TRITC-conjugated goat anti-(rabbit IgG) IgG in the center well against purified concanavalin A, showing characteristic double lines of precipitation due to nonspecific binding. (d) Fluorescence micrograph showing the distribution of nonspecific binding of TRITC-conjugated goat anti-(rabbit IgG) IgG in cells of an imbibed cotyledon of Canavalia ensiforrnis. (e) Unstained cotyledon of C. ensiformis, illuminated in green light (TRITC system) showing low level of autofluorescence. (0 Transmitted light micrograph of the field shown in (e).

252

C. E. JEFFREE ET AL.

FIG. 7. Immunofluorescent localization of lectin in tissues of Dururu srrumoniurn stained with rabbit anti-(Daturu lectin) IgG and FITC-conjugated goat anti(rabbit IgG) IgG using the indirect method. (a) Specific fluorescence in cells of imbibed cotyledons is confined to the cytoplasm, and is absent from cell walls. Note empty cells (arrowed) which appear dark. (b) Control section of cotyledon treated with nonimmune I" antibody showing low levels of nonspecific staining and autofluorescence. (c) Pith cell from stem showing specific fluorescence associated with the plasmalernma (PL), nuclear envelope (NE), chloroplasts (C), and with fine granules ( G ) possibly due to staining of mitochondria or endoplasmic reticulum. (d) Pith cells stained with control nonimmune IgG. (e) Specific fluorescence from cytoplasm in a section of embryo. (f)Specific fluorescence from cells of stem callus.

IMMUNOFLUORESCENCE STUDIES ON PLANT CELLS

253

FIG. 8 . Immunofluorescence localization of lectin in tissues of Duturu strummiurn stem, stained with rabbit anti-(Datum lectin) IgG and FITC-conjugated goat anti-(rabbit IgG) IgG, using the indirect method. (a) Intense specific fluorescence from the epidermis (EP) compared with a less intensely fluorescent cortex (C). (b) Nonimmune control corresponding to (a) showing some autofluorescence from the cuticle (arrows). (c) Intense specific fluorescence from phloem bundles (PH). Note that cell walls are delimited by bright lines of fluorescence associated with the cell surface, but the walls are dark. (d) Intense specific fluorescence from the endodermis (EN). The vascular cambium (VC) is less intensely stained. Bright autofluorescence emanates from the xylem (X). (e) Specific fluorescence from the endodermal cells showing bright fluorescence from the plasmalemma (PL) than from plastids (P). (f) Bright field micrograph of the field shown in (e).

254

C . E. JEFFREE ET AL.

the involvement of lectins in transport or metabolism of carbohydrates and glycoproteins (Fig. 8). Serological studies have shown that the Datura lectin is one of a group of closely related glycoproteins occurring in Solanaceous plants and even distantly related Solanaceous lectins have antigenic determinants in common (Kilpatrick et al., 1980). This fact has enabled immunofluorescence studies to be carried out on tomato, pepper, potato, Nicandra physaloides, and AtrGpa belladonna (Fig. 9) using a single antiserum, raised against Datura seed lectin, and has demonstrated the presence of Darura seed “lectin-like” proteins in stem sections of these species. The intercellular and intracellular distribution of specific fluorescence in stems of these species corresponded closely with that in Datura stems, and the absence of specific stain in control sections indicated a high degree of specificity throughout (Figs. 9 and 10). An interesting result from this work was that despite the fact that the serological techniques of double diffusion and immunoprecipitationemployed by Kilpatrick et al. (1980) were in combination capable of identifying the Nicundra lectin as antigenically related to but distinct from the Datura lectin, the tissues of the two species were indistinguishable in sections of NicandralDatura heterografts stained with anti(Datura lectin) immunoglobulins. This suggests that the method can be markedly less discriminating than serological procedures when comparing molecules with common antigenic determinants. The successful use of antiserum against Datura lectin to locate lectins in other Solanaceous species, begged the question: To what extent is the system capable of recognizing lectins outside the Solanaceae, in more distantly related taxa? We have found that Chrysanthemum, Dianthus, and Phaseolus stem sections were also stained, apparently with high specificity, by indirect immunofluorescent procedures employing rabbit anti-(Datura seed lectin) as the primary antibody (Figs. 11 and 12). The intracellular distribution of the specific fluorescence was the same in these species as in Datura cells. The plasmalemma was labeled and subcellular organelles showed strong specific fluorescence (Figs. 1 la and b and 12a and b) associated with chloroplasts, amyloplasts, the nuclear envelope, and fine granules which were not readily identifiable by light microscopy, but which might be consistent with staining of the mitochondria or endoplasmic reticulum. The intercellular patterns of distribution of specific fluorescence were also broadly similar in these species to those observed in Solanaceae, generally showing greater intensity, consistent with greater lectin concentration, in the epidermis, endodermis (where present), phloem, and vascular cambium. In Dianthus the guard cells were notable sources (Fig. 12b). In red kidney beans, however, the cortex was the most intense source of fluorescence. Since the controls work (Figs. 1 lc and f and 12d), the occurrence of specific fluorescence in sections of these species can be taken as evidence that lectins in three further dicotyledonous families have some antigenic determinants in common with Datura lectin. Sections of maize stems on the other hand gave negative results (Fig. 12e and f),

IMMUNOFLUORESCENCE STUDIES ON PLANT CELLS

255

FIG.9. Cryostat sections of Atropa belladonna stems, stained with rabbit anti-(Darure lectin) IgG and FITC-conjugated goat anti-(rabbit IgG) IgG, using the indirect method. (a) A cell of the outer cortex showing specific fluorescence from granular material in the cytoplasm (G) and from the plasmalemma (PL). The cell walls are free of stain. (b) Intense specific fluorescence in the epidermis (EP) diminishes toward the inner cortex (CO). (c) Phloem bundles (PH) exhibit bright specific fluorescence. (d) Nonimmune control corresponding with (b), showing some autofluorescence from the cuticle but little from other tissue components. (e) Relative intensities of specific fluorescence emitted by different tissues are well illustrated by this low power micrograph. Epidermis (EP), cortex (CO), phloem (PH), pith (P), xylem (X).(f) Vascular region, showing bright specific fluorescence from the phloem (PH).

256

C. E. JEFFREE ET A L

FIG. 10. Stem sections of three Solanaceous species, stained with rabbit anti-(Datura lectin) IgG and FITC-conjugated goat anti-(rabbit IgG) IgG, using the indirect method. (a-d) Nicandra physaloides. (a) Pith cell showing specific fluorescence from the nucleus (N), plastids (P), the plasmalemma (PL),and fine granules ( G ) . Compare this distribution with that shown in the Datum pith cell (Fig. 7c). (b) Nonimmune control corresponding with (a). (c) Specific fluorescence from the epidermis is brighter than adjacent cortex. (d) Intense specific fluorescence from xylem parenchyma (XP) outshines that from the phloem (PH). (e) Solanum ruberosum. Pith cell showing specific fluorescence from the nucleus (N), plastids (P), plasmalemma (PL), and finely granular material ( G )(cf. Figs. 7c and 10a). (0 Lycopersicon esculentum. The epidermis (EP) and endodermis (EN) are the most intense sources of specific fluorescence. Note that the thickened comers of the collenchyma cell walls (CO) appear dark.

IMMUNOFLUORESCENCE STUDIES ON PLANT CELLS

257

FIG. I I . Stem sections of red kidney bean (Phaseolus vulgaris) and Chrysanthemum morijolium stained with rabbit anti-(Datura lectin) IgG and FITC-conjugated goat anti-(rabbit IgG) IgG using the indirect method. (a-c) Red kidney bean. (a) Specific fluorescence from a pith cell. (b) Intense specific fluorescence associated with the vascular cambium (VC). (c) Nonimmune control corresponding with (b), showing only autofluorescence from lignified cells. (d) Specific fluorescence from cortex chlorenchyma. Plastids (P) and plasmalemma (PL) are notable sources. The cell walls appear dark. (e) The most intense sources of specific fluorescence are the epidermis (EP). endodermis (EN), and vascular cambium (VC). The inner cortex (CO) is markedly less intense. (f) Nonimmune control corresponding with (e) showing only autofluorescence from the cuticle (arrow).

25 8

C. E. JEFFREE ET AL.

FIG. 12. Stem sections of Dianthus caryophyllus and Zea mays stems stained with rabbit anti-(Datura lectin) IgG and FITC-conjugated goat anti-(rabbit IgG) IgG. (a-d) Dianthus caryophyllus. (a) Specific fluorescence from cells of the cortex. (b) Intense specific fluorescence from guard cells (GC). (c) Specific fluorescence from the phloem (PH)is brighter than adjacent tissues.(d) Nonimmune control corresponding with (c), showing autofluorescence from lignified cells. (e,f) Zea mays. (e) Brightly autofluorescent stem section. (f)No increase in fluorescence is measurable after staining.

IMMUNOFLUORESCENCE STUDIES ON PLANT CELLS

259

suggesting that the lectins of this species of monocot may be antigenically unrelated to Datura lectin. The immunofluorescent system employed here was unable to distinguish between the antigenically distinct Datura and Nicandru lectins in heterografts of these species, and is therefore at a disadvantage compared with the more sensitive standard immunoprecipitation tests of antigenic identity. This restricts the usefulness of the technique for cytochemical studies involving structurally related molecules. On the other hand the very fact that lectins of species quite dissimilar phylogenetically from Datum are recognized by antibody raised against Datum lectin has provided an indication that these molecules have a highly conserved structure. TABLE I RELATIVE FLUORESCENT I N T E N s r T y OF TISSUESFROM STEMSECTIONS OF Datura stramonium A N D FOUR NON-SOLANACEOUS SPECIES STAINED WITH RABBIT ANTI-(Datura SfrUmotliUm SEED LECTIN)IgG A N D IgG) FITC-CONJUGATED GOATANTI-(RABBIT

Tissue Epidermis

Cortex

Endodermis

Phloem

Xylem vessels

Vascular cambium Pith

Guard cells

Antibodyu ADL NI SF ADL NI SF ADL NI SF ADL NI SF ADL NI SF ADL NI SF ADL NI SF ADL NI SF

Datura stramonium 156.5 2 28.3 t 128.2 56.6 C 16.1 C 40.4 188.0 C 16.4 C 171.6 185.0 f 24.5 f 160.5 227.4 f 163.9 f 63.5 85.6 12.9 72.7 54.9 15.3 39.6

2

15.0 1.0 5.3 0.5 9.2 1.0 1.2 0.7 15.4 6.8

3.9

t 0.4 t 5.9 C

-

0.3

Chrysanthemum morifolium 191.7 55.9 135.8 103.0 32.4 70.6 213.3 33.4 179.9 179.0 54.3 124.7 212.3 115.1 97.2

2 2

15.4 5.9

C

9.0 3.1

f f f

18.3 3.7

C 13.0 f

3.0

C 10.0 5

12.5

252.8 t 38.8 C 184.0 58.7 f 29.1 C 29.6 -

17.2 1.9 5.9 4.0

Dianthus caryophyllus

Phaseolus vulgaris

375.8 t 28.0 114.6 2 7.7 37.1 2 2.4 24.0 t 1.6 338.7 90.6 82.2 f 10.2 156.8 -C 9.9 35.6 C 2.7 13.1 f 0.5 46.6 143.7 132.3 f 12.4 18.9 -C 0.9 113.2 245.5 t 9.0 367.8 C 23.5 31.7 f 1.2 23.2 C 0.5 213.8 344.6 222.3 2 5.4 160.5 t 6.9 215.0 2 8.9 70.1 f 5.8 90.4 7.3

259.7 t 22.6 265.9 t 26.0 -6.2 44.5 f 3.0 50.6 f 4.8 -6. I -

160.2 t 143.7 C 16.5 177.6 2 219.3 f -41.7

-

-

-

-

491.2 f 49.8 19.9 C 0.5 47 1.3 37.2 f 3.2 38.1 C 3.2 26.8 2 1.1 11.6 f 0.6 10.4 26.5 300.1 2 28.0 99.4 ? 8.2 200.7

“ADL, 10 antibody = anti-(Datura lectin) IgG; NI, l o antibody specific fluorescence, calculated as ADL - NI.

Zea mavs

=

5.5 5.8 7.1 14.3

-

-

nonimmune rabbit IgG (control); SF.

260

C. E. JEFFREE ET AL.

The sensitive immunoprecipitation test applied by Kilpatrick et ul. (1980) who found the Nicandru lectin was not precipitated from solution by anti-(Duturu seed lectin) IgG over the concentration range used could be interpreted, if used in isolation as implying insignificant structural relationship between the two species. It is important to appreciate that the various immunological procedures available have different ranges of sensitivity, and there is a strong argument therefore for interpreting immunofluorescent studies in the light of data obtained from alternative serological techniques which have differing sensitivity. This also emphasizes the importance of raising antisera against highly purified and characterized antigens, since the interpretation of the results can become hopelessly complicated if the specificity of the antiserum cannot be accurately defined. In this study, the limited resolution of the immunofluorescent system has proved to be a major limiting factor to precise localization of molecules at the subcellular level. As noted previously it was impossible to confirm that localization of specific fluorescence at the surface of the cell was due to binding of antibody to the plasmalemma. Equally the small granular sources of fluorescence within cells could not be identified, although they were thought consistent with labeling of the mitochondria1 membranes or endoplasmic reticulum. The method is clearly more useful, in this type of study, for assessment of the distribution of antigens within tissues, or in relation to coarser structures within cells. In this study it has become clear that the autofluorescence of plant tissues can be a major handicap irrespective of the fluorescent dye used as a label. For this work, involving green plant tissues, we found, contrary to the observations of Hapner and Hapner ( 1978), that the rhodamine system resulted in unacceptable levels of autofluorescence from chlorophylls which could not readily be separated from the specific fluorescence. This chlorophyll autofluorescence is present in the FITC system also, but can be readily removed from the image by filtration. Lignin, polyphenolic materials, and cutin autofluoresce in the filtration excitation wavelengths used for rhodamine ITC and FITC, and can be occasionally troublesome, but the autofluorescence from nonlignified cells is low in both systems. Although about four times as sensitive to fading as TRITC (Nairn, 1969), FITC is therefore the preferred fluorescence marker (tracer) for most immunofluorescence work. Both of these shortcomings of the immunofluorescent system strongly argue in favor of the development of EM procedures applicable to plant tissues. c.

LOCALIZATION OF OTHER PLANT PROTEINS IN PLANT

TISSUES

As well as enzymes and lectins, immunofluorescence studies have been carried out on a diverse variety of other proteins in both protists and higher plants. The use of antibodies to complex carbohydrate fractions of Neurosporu crussu hyphal walls enabled differences in composition along the hyphae to be visu-

IMMUNOFLUORESCENCE STUDIES O N PLANT CELLS

26 1

alized by indirect immunofluorescence (Hunsley and Kay, 1976). Their main finding was that apical and subapical regions reacted more than the mature wall with antibodies raised against a glucan-peptide-galactosamine complex, and also with antibodies to unknown wall antigens distinct from those contained in the fractions used. Hauser et a f . (1975) used the direct immunofluorescent method to study myosin in the myxomycete, Physarum polycephalum. FITC-conjugated antibodies were used to demonstrate the nucleolar location of nuclear myosin, and to demonstrate a serological relationship between cellular myosins of lower eukaryotes. In higher plants, too, immunofluorescent techniques have been used to study the location of both single antigens and complex fractions. An example of the former is the use of FITC-labeled specific antibodies to legumin and to vicilin to study the appearance and distribution of those storage proteins in broad bean (Viciafaba) cotyledons (Graham and Gunning, 1970). In contrast, Barlow et al. (1973) used a crude mixture of soluble proteins as antigen to study the possible association of non-gluten-forming proteins and starch granules in the mature wheat endosperm cell. Glutaraldehyde-fixed endosperm sections were stained by indirect immunofluorescence; specific fluorescence was observed only around starch granules. The two examples given above illustrate the advantages of using highly purified proteins as antigens. The former yielded results from which precise information was obtained; in the latter case, the pattern of fluorescence could have been due to any one of many proteins, or mixtures in any combination, and did not exclude the possibility that the majority of proteins used in the antigen preparation might not be associated with starch granules. One area of research in which immunofluorescence has been a major tool is that of pollen-stigma interactions. Some of the earliest immunocytochemical work on plant material was carried out on pollen antigens (Hagman, 1964). Immunofluorescent techniques have been used to demonstrate the presence of pollen antigens on the outer surface of pollen grains (Knox et al., 1970; Knox and Heslop-Harrison, 1971a,b), suggesting a possible role of such antigens in male-female recognition at the stigma surface. This inference was given support by observations on the readiness of the antigens to duffuse away from the pollen grain and the demonstration of their presence on the stigma surface (Knox and Heslop-Harrison, 197 1 a,b; Knox, 1973). Antibodies have been raised to purified pollen glycoproteins from rye-grass (Knox et al., 1980). Indirect immunofluorescent techniques using these specific antisera have demonstrated the presence of glycoproteins in the wall cavities of the exine layer as well as in the cytoplasm, and the observations were confirmed by immunoferritin labeling (Knox et al., 1980). Little specific fluorescence was apparent in the intine layer. It should be stressed that immunofluorescent localization of diffusible antigens is possible only by using anhydrous processing and postembedding staining procedures.

262

C. E. JEFFREE ET AL.

IV. Conclusions Immunofluorescence techniques have enabled the localization at the tissue and subcellular level of a number of plant antigens, almost all proteins, which would have been difficult or impossible to localize by any alternative methods. There is no doubt that immunocytochemical methods have a promising future in botanical research. Properly employed they provide a unique system for positive identification, quantitation, and localization of macromolecules. For example, their use in the assay of secondary products in plant tissue culture is one obvious potential application. However, major pitfalls arise from the very power of immunocytochemical methods to disclose the presence of antigens present in plant tissues in minute quantities. It is essential, if the results are to be meaningful, that the purity of the antigen employed to raise antibodies should be beyond doubt. If there is contamination in the antigen preparation then not only will the antiserum raised to it be of dubious specificity, but it will be impossible to make an accurate assessment of its specificity, since the improperly purified antigen is the standard against which the results are judged. Purification of the IgG fraction is an important second step in improving the specificity of a good antiserum, and still further benefits are obtainable by affinity chromatography using the bound, purified antigen, to separate specific from nonspecific immunoglobulins. The logical extension of these considerations to the production of monoclonal antibodies is a desirable step, but the application of this method may be limited, for the forseeable future, by the high costs involved, to the study of a few antigens of high commercial significance. The localization of antigens in plants by immunofluorescent techniques is limited by the rather coarse resolution obtainable by fluorescence microscopy, and the problems of autofluorescence of plant tissues, although not always insuperable, frequently reduce confidence in the pattern of specific staining. There are no straightforward solutions to either of these problems. It seems probable for these reasons that immunoelectron microscopic methods, although scarcely yet available as routine techniques, and hardly at all yet applied to botanical problems, will achieve a much greater importance than at present for the localization of antigens at the subcellular level in plants, and in the long term will largely replace immunofluorescence microscopy for this purpose. Immunofluorescence microscopy nevertheless has a useful place in the range of available immunological techniques applicable to botanical problems-as a means of amplifying the evidence obtainable from serological procedures, as the primary means of localizing antigens at the level of tissues, and at low resolution, within cells. ACKNOWLEDGMENTS

The authors thank the Agricultural Research Council for financial support and Mrs. E. Raeburn for her work in preparing the manuscript.

IMMUNOFLUORESCENCE STUDIES ON PLANT CELLS

263

REFERENCES Bal. A. K., Verma, D. P. S . , Byrne, H.. and McLachlan. G . A. (1976). J. CellBiol. 69, 97-105. Barlow, K . K., Simmonds, D. H., and Kenrick, K. G., (1973). Experienria 29, 220-231. Beiser. S. M.. Burke, G. C., and Tanenbaum, S. W. (1960). J. Mol. Biol. 2, 125-132. Bennett, H. S., Wyrick. A . D.. Lee, S . W., and McNeil, J. H. (1976). Stain Techno/. 51, 71-97. Bozicevich, J.. Scott. H. A., and Vincent, M. M., (1963). Proc. Soc. Exp. B i d . 114, 794-798. Carnegie. J. A., McCully. M. E., and Robertson, H. A. (1980). J. Hisrochem. Cytochem. 28, 308-3 10. Chang. C. M., Reitherman. R. W., Rosen, S. D., and Barondes, S. H. (1975). Exp. Cell Res. 95, 136- 142. Clarke. A. E...Knox. R. B., and Jermyn, M. A. (1975). J . Cell Sci. 19, 157-167. Clausen, J. (1971 ). “Immunochemical Techniques for the Identification and Estimation of Macromolecules. ’’ North-Holland Publ., Amsterdam. Cole. M. B.. and Sykes. S. M. (1974). Stain Technol. 49, 387-400. Coons. A. H. (1956). Int. Rev. Cvtol. 5, 1-23. Coons. A. H., and Kaplan, M. H. (1950). J. Exp. Med. 91, 1-13 Coons. A. H., Creech, H. J., and Jones, R. N. ( 1941). Proc. Soc. Exp. B i d . 47, 200-202. Coons. A. H., Leduc, E. H.. and Connolly, J. M. (1955). J. Exp. Med. 102, 49-60. Cooper, T. G. (1977). “The Tools of Biochemistry.” Wiley, New York. Craig. S.. Goodchild. D. J.. and Millerd, A. (1979). J. Histoehem. Cvrochem. 27, 1312-1316. Deutsch, H. F. (1967). In “Methods in Immunology and Immunochemistry. I . ” (C. A. Williams and M. W. Chase, eds.). Academic Press. New York. Durbin, M. C., Sexton, R., and Lewis, L. N. (I98 I ). Plant Cell Environ. 4, 67-73. Francki. R. 1. B., and Habili, N. (1972). Virologv 48, 309-315. Fulcher, R. G., and Holland, A. A. (1971). Arch Mikrobiol. 75, 281-284. Garvey, J. S., Cremer, N. E.. and Sussdorf, D. H. (1977). “Methods in Immunology. A Laboratory Text for Instruction and Research. Benjamin. Reading, Massachusettes. Coding, J. W., (1978). J. Immunol. Merhods 20, 241-253. Graham, R. C., and Karnovsky, M. I . (1966). J . Histochem. Cvrochem. 14, 291-302. Graham, T. A.. and Gunning, B. E. S. (1970). Nature (London) 228, 81-82. Gray. J . C.. and Kekwick, R. G. 0. (1974). Eur. J . Biochem. 44, 481-489. Gray, J. C., and Wildman. S. G. (1976). Plant Sci. Lett. 6 , 91-96. Hagman, R. (1964). In “Pollen Physiology and Fertilization” (H. F. Linskens, ed.), pp. 244-250. North-Holland Publ., Amsterdam. Hapner, S. J.. and Hapner, K . D. (1978). J. Hisrochem. Cvtochem. 26, 478-482. Hattersley, P. W.. Watson, L., and Osmond, C. B. (1977). Aust. J. Plant Physiol. 4, 523-539. Hauser, M., Beinbrech, G., Groschel-Stewart, U., and .Jokusch, B. M. (1975). Exp. Cel/ Res. 95, 127. Herbert, W. I. (1978). In “Handbook of Experimental Immunology” (D. M. Weir, ed.). Blackwell. Oxford. Hunsley, D., and Kay, D. (1976). J. Gen. Microhiol. 95, 233-248. Jacobson, J. V., and Knox, R. B. (1973). Planra 112, 213-214. Jeffree, C. E., and Yeoman, M. M. (1981). New Phvtol. 87, 463-47 I . Jones, R. L.. and Ru Fang Chen (1976). J. Cell Sci. 20, 183-198. Kawamura. A. ( 1977). “Fluorescent Antibody Techniques and their Applications.” Univ. of Tokyo Press, Tokyo. Kilpatrick, D. C., Yeoman, M. M., and Could, A. R. (1979). Biochem J. 184, 215-219. Kilpatrick. D. C . , Jeffree, C . E., Lockhart, C. M., and Yeoman, M. M. (1980). FEBS Leu. 113, 129- 129. Knox, R. B. (1970). Stain Technol. 45, 265-272. ”

264

C. E. JEFFREE ET AL.

Knox, R. B. (1973). J . CellSci. 12, 421-443. Knox. R. B., and Clarke. A. E. (1978). In ”Electron Microscopy and Cytochemistry of Plant Cells” ( J . L. Hall, ed.). pp. 149-186. Elsevier, Amsterdam. Knox, R. B., and Heslop-Harrison, J. (1971a). Cyrobios 4, 49-54. Knox, R. B., and Heslop-Harrison, J. (1971b). J . CellSci. 9, 239-251. Knox, R. A , , Heslop-Harrison, J., and Reed, C. E. (1970). Narure (London) 225, 1066-1068. Knox, R. B., Vithanage, H. 1. M. V., and Howlett, B. J. (1980). Hisrochem. J. 12, 247-272. Kohler, G., and Millstein, C. (1975). Narure (London) 256, 495-497. Koehler, D. E., Lewis, L. N., Shannon, L. M.,and Durbin. M. L. (1981). Phytochemisrry 20, 409-4 I 2. Kraehenbuhl, J. P., Racine, L., and Griffiths, P. W. (1980). Hisrochem. J. 12, 317-332. Kyte. J. (1974). J. Biol. Chem. 249, 3652-3660. Langone, J. J., Gjika, H. B., and van Vunakis, H. (1973). Biochernisrp 12, 5025-5030. Levy, H. B., and Sober, H. A. (1960). Proc. Soc. Exp. Biol. Med. 103, 250-252. Lewis, D. (1952). Proc. Soc London Ser. B. 140, 127-135. Lindner, M., and Richards, P. (1978). Sci. Tools 25, 61-67. Linskens, H. F. (1960). Z. Bor. 48, 126-135. Muller, K., and Gerisch, G. (1978). Nature (London) 274, 445-449. Murray, D. R., and Knox, R. B. (1977). J. Cell Sci. 26, 9-18. Nairn, R. C. (1969). “Fluorescent Protein Tracing.” Livingstone, Edinburgh. Nasrallah, A. T., and Wallace, D. H. (1967). Heredity (London) 22, 519-527. Nothnagle, E. A., and Webb, W. W. (1979). J. Hisrochem. Cvrochern. 27, 1000-1002. O’Brien, T. P., and McCully, M. E. (1981). “The Study of Plant Structure: Principles and Selected Methods.” Termarcarphi, Melbourne. Olsen, B. R., Berg, R. A , , Kishida, Y., and Prockop, D. J. (1973). Science 182, 825-827. Ouchterlony, 0. (1958). Prog. Allergy 5 , 1-78. Ploem, J. S. (1967). Z. Wiss. Mikrosk. 68, 129-142. Ploem, J. S. (197 I). Ann. N.Y. Acad. Sci. 177, 414-429. Pressman, D., Yagi, Y.,and Hiramoto, R. (1958). fnr. Arch. Allergv 12, 125-136. Raff, J. W., Hutchinson, J. F., Knox, R. B., and Clarke, A. E. (1979). Differentiation 12, 179-186. Ray, J., Shinnick, T., and Lerner, R. (1979). Narure (London) 279, 215-221. Richardson, M. (1974). Sci. Prog. (Oxford) 61, 41-61. Rosen, S. D., Kafka, J. A., Simpson, D. L., and Barondes, S. H. (1973). Proc. Narl. Acad. Sci. U.S.A. 70, 2554-2557. Roth, J., Bendayan, M., and Orci, L. (1980). J . Hisrochem. Cyrochem. 28, 55-57. Secher, D. S. (1980). fmmunol. Today 1, 22-26. Sexton, R., Durbin, M. L., Lewis, L. N., and Thomson, W. W. (1980). Nature (London) 282, 873-874. Sexton, R., Durbin, M. L., and Lewis, L. N. (1981a). Plant Cell Environ. 4, 1-7. Sexton, R., Durbin, M. L., and Lewis, L. N. (1981b). Proroplasma 109, 335-347. Simpson, D. L., Rosen, S. D., and Barondes, S. H. (1974). Biochemisrry 13, 3487-3493. Sitte, P. (1965). “Bau und funktion der Pflanzenzelle.” Fischer, Stuttgart. Smith, P. M. (1976). “The Chemotaxonomy of Plants.” Arnold, London. Stansworth, D. R. (1960). Narure (London) 188, 156-157. Sumner, J. B. (1926). J . Biol. Chem. 69, 435-441. Tippett, J. T., and O’Brien, T. P. (1975). Lab. Praci. 24, 239-240. Tolbert, N. E. (1971). Annu. Rev. Plant Physiol. 22, 45-74. Vreeland (1972). J. Hisrochem. Cytochem. 20, 358-364. Weller, T. H., and Coons, A. H. (1954). Proc. Soc. Exp. Biol. 86, 789-794. Willingham, M. C. (1980). Hisrochem. J . 12, 419-434.

IMMUNOFLUORESCENCE STUDIES ON PLANT CELLS

265

Willingham, M . C., and Yamada, S. S . (1979). .I. Hisrochem. Cyrochem 27, 947-960. Willingham, M . C., Spicer, S. S . , and Graber, C. D. ( 197 I). Lab. Invesr. 25, 2 I 1-2 19. Willingham, M . C . , Yamada, S . S . , Davies, P. J. A , , Rutherford, A. V . , Gallo. M. G . , and Pastan, I . (1981). J . Hisrochem. Cytochem. 29, 17-37. Yamarnoto, A., Kawamura, A . , and Wada, K. (1970). I n “Standardisation in Immunofluorescence” (E. J. Holborow, ed.), pp. 113-115. Blackwell. Oxford.

This Page Intentionally Left Blank

INTERNATIONAL REVIEW OF CYTOLOGY. VOL. X0

Biological Interactions Taking Place at a Liquid-Solid Interface ALEXANDREROTHEN The Rockefeller University, New York. New York 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Immunologic Reactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111. Experimental . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

267 269 270 302 303

I. Introduction The beginning of systematic investigations on biochemical reactions occurring at a liquid-solid interface can be traced back to the historical work of Blodgett and Langmuir who, in the mid-l930s, made two important contributions (1-3). First they showed how monomolecular films formed at an air-water interface could be transferred to solid slides, and second they devised an ingenious optical method for determining the thickness of such films with an accuracy of a few Angstroms, after they had been transferred to reflecting surfaces. These contributions were undoubtedly the most important ones for the further development of the technique of surface films since the introduction of the spreading trough by Langmuir. At that time surface chemistry had come to a standstill and investigations were limited to layers located at an air-liquid interface. With the help of the new technique of Blodgett and Langmuir a large amount of experimenting proceeded in many laboratories with all kinds of materials capable of being spread and transferred [see, e.g., Rothen and Landsteiner (4~. In solutions molecules react individually when they are in close proximity. Furthermore Brownian motion limits the distance at which interactions occur. The possibility of testing the reactivity of an ensemble of molecules forming multimolecular or submolecular layers immobilized on solid slides opened up a vast new field of research. Under these conditions, as will be shown, cooperative phenomena take place between immobilized molecules and as a result with a considerable extension of the distance at which interactions occur. In a way the immobilized molecules react as a whole. The investigated reactions presented in this article are of two kinds: (1) In261 Copyrighl 0 19x2 hy Academic Press. Inc. All rights of reproduction in any form rcserved. ISBN 0-12-364480-0

268

ALEXANDRE ROTHEN

teractions of adsorbed layers of antigen, protein, or polysaccharide, with antibodies present in solutions into which the antigen-coated slides were immersed. (2) Proteolytic action of trypsin on adsorbed layers of susceptible substrate. All interactions between adsorbed antigen and antibody molecules as well as between substrate and trypsin were determined by measuring changes in film thickness following interaction. One of the main drawbacks of the Blodgett-Langmuir optical method for thickness determinations is the necessity of transferring a large number of long chain fatty acid layers onto the slide before depositing the film whose thickness is to be measured. As will be shown, the reactivity of a protein layer depends not only on the anchorage to which the layer is attached but also on the number of underlying layers located between the metallized slide and the anchorage. As a consequence the thickness of a layer whose reactivity is under investigation cannot be determined by the method of Blodgett and Langmuir. At the time they were carrying out their investigations, it was generally assumed that only the anchorage, that is the molecules or atoms to which the investigated layer was directly attached, was of importance and that layers situated under the anchorage played no part in the reactivity of the top layer. In spite of numerous results obtained in the last 30 years in the author’s laboratory, showing this assumption to be incorrect, many surface chemists still adhere to it. To facilitate progress in this area, a new method was necessary for determining film thickness within a fraction of an Angstrom that did not require the preliminary deposition of some 50 odd layers of barium stearate. An optical method developed in 1890 by P. Drude met the required conditions ( 5 , 6 ) . Drude had shown theoretically as well as experimentally that the ellipticity of polarized light reflected from a surface was sufficiently altered by the presence of a film a few Angstroms thick to allow the determination of its thickness within a fraction of 1 8, unit. Drude’s work was ignored for nearly 40 years until Tronstad in Norway in the early 1930s applied Drude’s theory to the determination of the thickness of oxide layers coating metallic surfaces. In 1935 he was able to determine the thickness of long chain fatty acids adsorbed on a mercury surface (7). His values, though slightly high, were of the right order of magnitude. Ten years later a simple visual instrument was developed and was called an “ellipsometer.” It permitted the determination of the thickness of molecular monolayers with an accuracy of a fraction of an Angstrom unit (8). The optics of the instrument were similar to one described by Tronstad and later used successfully by Winterbottom (9) and was based on Drude’s theory for the determination of the thickness of molecular layers. The half-shadow device of this instrument was a new contribution. Since then instrumentation has been very much improved with the introduction of the photomultiplier tube and of automation. Manually operated ellipsometers do not allow thickness measurements to be

INTERACTIONS AT A LIQUIWSOLID INTERFACE

269

made rapidly, but their accuracy is of the order of 0.1 A. The recent book by Asham and Bashara (10) gives a comprehensive survey of ellipsometry.

11. Immunologic Reactions It is currently assumed by immunologists that specific interaction between antigens and antibodies takes place in definite multiple proportions, that is in stoichiometric ratios, and that no covalent bonds are involved. The consensus of opinion is that specific forces result from the summation of weak nonspecific short-range forces such as van der Waals forces and hydrogen bonds. Forces become specific and strong enough to form an antigen-antibody complex when there is a sufficient number of these nonspecific contributions bringing extended areas of antigen and antibody into close proximity. Specificity results from the overall geometry of the reacting molecules which might be compared to pieces of a jigsaw puzzle fitting together. Comparison to a lock and key model has been used for three quarters of a century. The Belgian immunologist Jules Bordet believed that the antigen-antibody complex can occur in a continuously variable proportion of the two constituents. This view was discredited when an overwhelming abundance of evidence was collected in favor of so called stoichiometric ratios between antigen and antibody and the terms “valence” and “combining sites” are still in current use. Such was the situation in 1945 when I started working on the interaction of antibodies with films of antigen deposited on metallized slides. It quickly became evident that interactions at the liquid-solid interface were not stoichiometric and furthermore that antigen-antibody interaction could be observed across a membrane which prevented direct contact between antigen and antibody. This raised the question about the distance at which antigens could react in a specific way with antibodies. Thirty years ago there was no theory to explain long-range interaction. The opinion among physicists and chemists was that specific long-range forces were unlikely and that the results could best be explained by assuming diffusion through the membrane of detached antigen or antibody molecules, or both. Some opponents of the idea of long-range interaction published data which they thought proved their point (1 1-13). However, in 1943 London prophetically wrote that “there can be, however, legitimate doubt whether at present we are able to survey all consequences of quantum mechanics to such an extent with certainty to preclude the existence of specific macromolecular forces which could not be built up by the well known elementary atomic or molecular forces but would rather depend on properties of the molecule as a whole” (14). it should also be mentioned that Coulson and Davies (15) were able to calculate dispersion forces of long range. Their model, however, did not foresee any specificity.

270

ALEXANDRE ROTHEN

One had to wait until 1955 when Lifshitz and his collaborators showed in historic papers that van der Waals forces are long-range when the interaction of assemblies of molecules in condensed systems is considered instead of the interaction between individual molecules (16, 17). This was a most important step. For organic molecules the frequencies of the electromagnetic radiation involved would be located mostly in the infrared according to Ninham and Parsegian (18). Macroparameters such as dielectric constants and absorption bands enter into the theory. Depending on the geometry of the system the energy of interaction diminishes with the fourth or third power of the distance instead of the sixth power in the case of single molecules. Since the liquid-solid interface is a condensed system, any law valid for condensed systems should apply to the interaction between adsorbed antigen and antibodies. The specificity of Lifshitz forces is too broad to explain the stereospecificity of immunologic reactions. It might, however, be possible to develop further the model of Lifshitz to explain highly specific forces. As will appear from the experiments to be reviewed, the long-range order of the adsorbed molecules has a paramount influence on the range of interaction.

111. Experimental

All of the experiments described below involved the interaction of protein or polysaccharide molecules adsorbed on the surface of a slide with molecules of antibodies or enzymes in solution. The degree of interaction was measured by the thickness of the layer of antibody which was subsequently adsorbed. The thickness increased when antibodies were allowed to react with a layer of adsorbed antigen. On the other hand, when the antigen layer was first subjected to proteolytic action, the adsorbed thickness decreased and no antibodies were adsorbed on subsequent immersion of the slide in a solution of homologous antiserum. If the enzymatic action was incomplete, only a much diminished thin layer of antibodies was adsorbed. All thicknesses were measured optically with an ellipsometer. A description of the manual and recording ellipsometers is not given in this article and for details the reader is referred to “Physical Techniques in Biological Research” (D. H. Moore, ed.), Vol. IIA, chap. 6. Academic Press, New York, 1968 and also to references (19, 20). 1. Determinations of Antigen-Antibody Reactions and of Tryptic Action in the

Presence of an Intervening Membrane at a Liquid-Solid Interface The reactions were carried out using metallized glass slides for the adsorption of the antigen or the substrate susceptible to tryptic action. Glass slides, 12 X 0.6 x 0.15 cm, were coated by evaporation in vacuo with a metallic layer approx-

INTERACTIONS AT A LIQUIWSOLID INTERFACE

27 1

imately 4000 A thick. For the preparation of nickel-plated slides called “active” that permit the determination of antigen-antibody reactions with extremely dilute solutions, the slides were first glow discharged at l o p 3 mm of mercury with some aluminum sputtered on the glass. The evaporation of nickel was carried out very slowly at 10W5mm of mercury and a layer about 4000 A thick was obtained in 10 minutes. The evaporation was carried out in the presence of a magnetic field, I S , of a few thousand Gauss obtained with a bar magnet or an electromagnet.’ A much weaker field would be sufficient. The following symbols will be used throughout: I S means that the lines of force of the magnetic field were perpendicular to the slide, the metallized side facing the south magnetic pole. I N means that the lines of force were perpendicular to the slide and the metallized side faced the north magnetic pole. 11 stands for lines of force parallel to the slide. When no field was present during nickel evaporation, the slides were not “active,” or only moderately so, but could be activated by a magnetic field I S . The activation-inactivation is a reversible process brought about by I S and 11. Slides coated with T i 0 and TiO,, or metallized with Th, Cr, inconel, or nichrom have been used as well as stainless-steel slides and carbide gauges sufficiently plane to allow the adhesion of two gauges in optical contact. As long as the concentration of the antigen solution was greater than l o p 6 g/ ml, any metallic surface was satisfactory. When very dilute solutions of antigen were used, the nature of the physical state of the surface, that is the long-range order of the domains of the surface, were of paramount importance. 2. Antigen Deposition Antigens can be adsorbed directly on the slide by simply immersing the slide in an aqueous solution of the antigen, then rinsing and drying. When a relatively concentrated solution is used, containing 10W3g/ml, the adsorbed layer is nearly complete in a few seconds. One monomolecular layer is adsorbed; its thickness is = 35 A in the case of ovalbumin and bovine serum albumin. The forces involved at the liquid-solid interface are not strong enough to denature the proteins. In all the experiments described in this article, the aqueous antigen concentrations were lo-’ g/ml unless otherwise indicated. A surface film can also be spread on the surface of a Langmuir trough by simply depositing on the surface droplets of a very dilute solution. The total protein content of the droplets should not be large enough to form a film covering the whole area of the trough. The spread film should be under no compression. Its thickness is 6 to 7 A,corresponding to a fully extended polypeptide chain; the protein molecules have been denatured at ‘The slides were prepared under stringent conditions by Evaporated Metal Films C o p . , 701 Spencer Road. Ithaca, New York 14850.

272

ALEXANDRE ROTHEN

the air-water interface. If the film, after spreading, is compressed to 6 or 8 dynes, it can be transferred to a slide by immersion, and a second monolayer is transferred on emersion of the slide. The symbol ( & f ) stands for one double layer transferred by one immersion followed by one emersion. A great many double layers of ovalbumin can thus be transferred. In the case of bovine albumin two layers only can be transferred since on further immersion and emersion the layers slip back to the surface of the trough. If, however, the slide is first coated with a few layers of Ba stearate and then “conditioned” by treatment with a solution of uranyl acetate brought to a pH of 7.5, a great number of bovine albumin layers can be transferred. It is characteristic of transferred bovine albumin layers that the thickness of antibody subsequently adsorbed from homologous antibody is proportional to the number of subjacent layers of albumin, up to four or five double layers. Depending on the antiserum and the slide, 60 to 70 A can be adsorbed per double layer. In the case of ovalbumin the thickness adsorbed from homologous antiserum is nearly independent of the number of layers of ovalbumin transferred to the slide (21). In all the experiments to be described the antisera were diluted 1/10 in veronal buffer, 0.03 M ,pH 7.5, unless otherwise indicated.

3 . Formation and Porosity of Protective Formvar Membranes Formvar membranes can be formed directly on antigen-coated slides by smearing a solution of Formvar in ethylene dichloride of appropriate concentration for a few seconds. The slide is then kept in a vertical position to let the solution drain. The Formvar membrane can be formed also on a glass slide. If the slide is slowly immersed in water, the membrane becomes detached and floats on the surface where it can be transferred to an albumin-coated slide. Experiments were carried out to evaluate the porosity of a Formvar membrane in the dry state. Slides coated with three free stearic acid layers, transferred by Blodgett’s technique or coated by sublimation of free stearic acid in vacuo, were covered with preformed Formvar membranes of various thicknesses. The slides were heated to 30°C in a high vacuum Hg). The loss in thickness, due to stearic acid subliming through the membrane, was compared to the loss occurring in the absence of a protective membrane. A loss of 2 to 4 A occurred in 1 hour in the presence of a membrane 100 A thick protecting 49 A of stearic acid. Without a membrane the loss was 44 A. The mesh of a Formvar membrane is therefore a very tight one (22). The closely knit structure of Formvar membranes was demonstrated by another type of experiment (22). Protein layers submitted to a-particles from a polonium source (5 mCi/cm2) lost their capacity to combine with antibodies. The inactivation does not result from the direct hit of the a-particles but is brought about by the ions produced by a-particles in the atmosphere surrounding the slide. This was demonstrated by the fact that no appreciable inactivation took place when the bombardment was carried out in a high vacuum. The negative

INTERACTIONS AT A LIQUID-SOLID INTERFACE

273

ions are mostly responsible for the inactivation since, when the slide was kept at a positive potential of 300 V with respect to the polonium source, the inactivation was greatly enhanced. The inactivation increased with the diameter of the molecules or atoms of the atmosphere in the order H, < He < 0, = N, < CO, = A (23). After being submitted to 5 mCi from polonium for 2 hours, six monolayers of bovine albumin were able to adsorb a thickness of 40 A from an antiserum instead of 180 A in the absence of irradiation. If a Formvar membrane 100 A thick was present during irradiation, the thickness adsorbed after removal of the membrane was 80 A, an increase of 40 A. This clearly indicates that 100 A of Formvar offers an efficient barrier to the small ions of N, and 0,. When six up layers ( )6 of bovine albumin protected by a Formvar membrane 80 A thick were irradiated for 2 hours, they completely lost their capacity for reacting with the antiserum in the presence of the membrane. However, when the membrane was dissolved before antiserum treatment, 100 8, was adsorbed. Without irradiation, antibodies could be adsorbed through a membrane 80 A thick protecting six up layers. This definitely demonstrated that adsorption of antibodies did not take place by free diffusion through the membrane, since it was the condition of the albumin layers, irradiated or not irradiated, that determined whether or not antibodies could be adsorbed in the presence of a membrane. As will be shown later, the antibodies probably did not penetrate the membrane but remained immobilized on the upper surface. Monolayers of proteins were even more efficient than Formvar membranes in shielding antigen from radiation. For example, one monolayer of bovine albumin 6 A thick in less than 20 minutes of irradiation completely lost its ability to combine with homologous antibodies. It took 45 minutes to inactivate one double layer (12 A) and 135 minutes was insufficient to inactivate three double layers (36 A). In other words, a layer of protein 6 A thick offered a considerable obstacle to the diffusion of small ions of 0, and N,. 4. Adsorption of Antibody Molecules from an Antiserum on an AntigenCoated Slide with or without a Protective Membrane The antigen-coated slide is dipped into a small test tube containing as little as 0.5 ml of the diluted antiserum for 7 to 10 minutes. The serum is diluted in veronal buffer, pH 7.5. It is advisable to stir the tube during the adsorption, when the antiserum is highly diluted. A rinsing period of 10 to 15 seconds is adequate and has to be done carefully and always under the same conditions to obtain reproduceable results.

5 . Rates of Reaction between Antigen and Antibody at a Liquid-Solid Inte$ace Investigated with a Recording Ellipsometer The recording ellipsometer allows the determination of the rate of formation of a layer at a liquid-solid interface. As an illustration the rate of adsorption of

274

ALEXANDRE ROTHEN

antibodies from an anti-human y-globulin serum on a chromium-plated slide densely coated with human y-globulin will be described. The human y-globulin was adsorbed for 3 minutes from a veronal solution containing 10W3 g/ml of globulin. The adsorbed layer was = 35 thick. The cuvette of the ellipsometer in which the slide was immersed was filled with 0.5 ml of the diluted antiserum. An important phenomenon was observed that is very significant for the understanding of the mechanism of action between antigen and antibody carried out at a liquid-solid interface. If the antigen-coated slide was wet when immersed in the antiserum, the adsorption of antibody occurred at a much faster rate than when immersed dry, except at the very beginning. In Fig. 1, curve a shows the rate of the reaction when the slide was wet, curve b when the slide was dry, and curve c is a control in which the slide was coated with the heterologous antigen rabbit y-globulin and introduced wet into the antiserum. The curves are strikingly different. When the slides are introduced wet, the rate is slower at the beginning than a few minutes later since it takes time for the molecules to cross the thin layer of water wetting the slide. The large rate of adsorption shown by curve a, 1 or 2 minutes after the beginning, compared to curve b, is inexplicable on classical grounds assuming that the motion of the

60 -

A

40 -

C

0

4

8

12

16

20

24

28

32

36

40

Time in minutes

FIG. 1. Adsorption in Angstroms from an antiserum diluted 1/10 in veronal against human yglobulin on a slide coated with human y-globulin. Curve (a) slide introduced wet, curve (b) slide introduced dry, and curve (c) slide introduced wet into the antiserum. The slide for curve (c) was coated with a heterologous antigen, namely, rabbit globulin.

275

INTERACTIONS AT A LIQUIBSOLID INTERFACE

antibodies toward the slide is due to a concentration gradient only. Other forces must be assumed (24). It can be seen from curve c that no globulin was adsorbed in 1 to 2 minutes when the slide coated with a heterologous antigen was introduced wet into the antiserum. Antibodies are adsorbed faster on an antigencoated slide immersed wet rather than dry, probably because the thin layer of water adhering firmly to the slide acts as a barrier that slows the diffusion of nonspecific globulin while specific antibodies because of a long-range interaction with the antigenic layer are able to cross the water barrier and become immobilized on the slide. It should be remembered that the thin layer of water adjacent to the slide has a high viscosity. Palmer el al. showed years ago (25) that layers of water 20 to 50 pm thick immobilized between mica plates have a dielectric constant of 10 and 20, respectively, and that this water should be considered as liquid ice. Cope (26) recently stated on the basis of experimental evidence the cell water is structured. Winter (27) more than 20 years ago showed that near a solid surface water exhibits long-range order. It is therefore to be expected that a thin layer of water adjacent to the slide would offer an important TABLE I INFLUENCEOF THE CONDITION OF THE SLIDE,DRYOR WET WHENINTRODUCEDINTO THE THICKNESS ADSORBEDO Type polysacc haride concentration (g/ml)

{

{

{

{ { {

Time Condition adsorption (min) slide

111 1 0 - 3 11110-3 111 3.2 X 10-3

1

x 10-3 x 10-3 x 10-3 x 10-3

1

III 3.2 111 3.2 III 3.2 111 3.8 III 3.8 VIII 2.4 VIII 2.4 VIII 2.4 III 9 III 9 III 2 III 2 111 2 111 2 111 3.8 III 3.8 111 3.8

x 10-3

x 10-3 x 10-3 x 10-3 x 10-4 x 10-4

x 10-4 x 10-4 x 10-4 x 10-4 x 10-3 x 10-3 x 10-3

1 1

1 1 1 1

I I 1 2 2 2 2 2 2 2 2 2

Dry Wet Dry Wet Dry Wet Dry Wet Dry Wet Wet Dry Wet Dry Wet Dry Wet Dry Wet Wet

Type antiserum dilution I11 1/20 111 1/20 111 1/100 111 1/100 111 1/100 111 1/100 111 1/20

I11 VlII VIIl VIII

1/20 1/100 1/100 1/100

111 1/20 111 1/20 111 1/20

I11 1/20 111 1/20 111 111 111 111

1/20 1/20 1/20 1/1000

THE

Time adsorption (min) 10 min 10 min 10 min 10 min 60 min

60 min 10 min 10 min 10 min 10 min Overnight 10 min 10 min 4 min 4 min 10 min 10 min 10 min 10 min 10 min

ANTISERUM, ON

Thickness adsorbed in A from antiserum 70 I00 5 5 , 55

116, 117, 116 80, 80.7 166 96

i

1 1

224 51

103 150 56

101 34 51

38 84 78 278 34

OSystems polysaccharide from Type 111 or Type VIII pneumococcus and the corresponding antiserum.

1 1 1

276

ALEXANDRE ROTHEN

obstacle to the diffusion of molecules moving under a gradient of concentration. The adsorption of nonspecific globulins is negligible when the slide is wet but is quite appreciable when dry, the opposite of the situation with the immunologic reaction. Type 111 and Type VIII pneumococcus polysaccharides also showed increased rates of reaction when the slides were introduced wet into the antiserum solution. Table I shows the strong influence of the wet or dry condition of the slides on the thickness of antibody adsorbed from the antiserum. When no polysaccharide was present on the slide, the nonspecific thickness adsorbed was of the order of 30 A. There is a strong cross-reaction between the two polysaccharides, as shown in Table 11, that could still be detected with antigen solutions at concentrations of 10-l3 g/ml of polysaccharide if a weak electric current was used for antigen adsorption. When cross reactions were carried out with concentrated solutions of polysaccharide ( l o p 3g/ml), the interaction was strong enough to allow a larger adsorption from an immune serum on a wet slide rather than a dry one. The thickness adsorbed was larger on dry than on wet slides when dilute solutions of polysaccharide were used ( 10W6 g/ml). The experiments demonstrate that the forces involved in cross-reactions are weaker than those in homologous reactions. The thickness of the layer adsorbed from antiserum is very dependent on the pH of the polysaccharide solution. The maximum thickness of -- 200 A was obtained at pH 2.2, while at pH 7.00 only 50 A could be adsorbed from the serum. It appears that the condition necessary to obtain a faster rate on a slide immersed wet rather than dry is that the thin layer of water in contact with the antigenic TABLE I1 SPECIFIC ADSORFTIONOF ANTIBODIES AS A FUNCTION OF CONCENTRATIONS OF POLYSACCHARIDE SOLUTIONS~ Antiserum against polysaccharide from Type I11 pneumococcus Conc. poly Wml) 3 4 4 4

x x x x

10-3

10-6 10-7 10-8

Poly 111

Poly VIII

(8,)

(8,)

218 55 30

I80 43 27 23

-

Antiserum against polysaccharide from Type VIII pnzumococcus Conc. poly Wml) 3 2 4 8

Poly I11

Poly VIIl

(8,)

(8,)

x 10-3

89

x 10-4 x 10-6

35 35 24

158 92 36 23

x 10-7 1.6 x 10-1(’

-

aThe figures under the headings poly 111 and poly VIII, which refer to the type of polysaccharide used as antigen, are the thicknesses in 8, adsorbed from antisera against either Type 111 or Type VIII pneumococcus.

INTERACTIONS AT A LIQUIPSOLID INTERFACE

277

layer should be firmly bound; the solid surface should be strongly hydrophilic. In all the experiments mentioned above the antigen was adsorbed on a chromiumplated slide. A different result using other metals has not been ruled out.

6 . Immunoelectroadsorption Method The sensitivity of the immunoassay can be greatly increased by carrying out the adsorption of both antigen and antibody with the help of a weak electric current (28) which increases the concentration of the molecules near the slide, by electrophoresis. If the antigen is positively charged, the slide is connected to the negative pole of a constant power supply and the positive pole to a thin platinum wire immersed in solution next to the slide. If the antiserum is diluted in veronal buffer, pH 7.5, the antibody molecules are negatively charged and the slide is connected to the positive pole. In this case a certain amount of oxidation takes place that is detrimental to thickness determination by ellipsometry. Of all the metals tested for metallizing slides, chromium was the most suitable since, with a slide connected to the positive pole in a buffer solution 0.03 M at pH 7.5 and 0.3 mA for 1 minute, there was an increase in optical thickness of 9 based on calibration with Ba stearate layers. Thicknesses at 35 and 100 were observed with inconel and titanium, respectively. Nickel-coated slides were so damaged that they became useless. This is the reason why chromium-coated slides were chosen when slides had to be positively charged for the adsorption of antiserum. One of the greatest difficulties in using chromium-plated slides for an immunologic assay at high dilution is its reproducibility. Slides apparently prepared under identical conditions by evaporation of the metal under high vacuum vary greatly in their activity. The maximum dilution of the antigen solution at which one can still detect an immunologic reaction is a measure of the slide’s activity. In the case of chromium slides this minimum concentration may vary by five orders of magnitude, from 10- l 2 to lo-’ g/ml. This is also true for inconel and nichrom-plated slides. The surface of chromium slides was investigated by electron diffraction at the Bell Telephone Laboratories at 40 keV (29) and all slides that did not show a crystalline size of at least 25 A were inactive. However, some slides which did show a good pattern were inactive for an immunoelectroadsorption assay. Good crystallinity seems to be a necessary but not sufficient condition. When an active chromium slide is submitted for 1 or 2 minutes to a magnetic field of a few thousand Gauss with its lines of force parallel to the slide’s surface, its activity is cut down by five to six orders of magnitude. No protein film should be present on the slide for this inactivation by a magnetic field. With pneumococcus polysaccharides Types I11 and VIII and their antisera, immunologic reactions were still detectable with solutions containing as little as l o - ” g/ml of polysaccharide when the slide was active. After inactivation by a magnetic field a polysaccharide concentration in excess of lo-’ g/ml was necessary to observe a

278

ALEXANDRE ROTHEN

positive immunologic reaction. It is not too surprising that a magnetic field has an influence on the superficial chromium atoms in view of the paramagnetism of chromium. When the lines of force were perpendicular to the slide, the activity was sometimes slightly improved. It was observed that the electron diffraction pattern by reflection exhibited by the chromium surface was not affected when the slide was submitted to the field 11. Since electrons of 40 keV can penetrate to a depth of a few hundred Angstroms it follows that the magnetic field affects only the order of the superficial layer of the chromium coating. After the influence of a magnetic field on the activity of chromium-plated slides was appreciated, nickel-plated slides were investigated. Since nickel is ferromagnetic, it was expected that nickel slides would be even more sensitive to magnetic activation than chromium slides, and this proved to be correct. Active nickel slides are so sensitive that positive assays can be obtained with solutions containing as little as to g/ml of antigen without using a current for adsorbing the antiserum. A fully inactive nickel slide requires antigen concentrations of more than lo-’ g/ml. It is interesting to observe that inactive chromium slides can be rendered active by coating them prior to adsorption of the antigen with five layers of Ba stearate transferred from a Langmuir trough. Such slides have to be used within a few hours of preparation, but if kept in a water-saturated atmosphere they remain active overnight. Slides coated with five monolayers of Ba stearate become inactive when submitted to a high vacuum or when heated to 50°C for 1 minute. The long-range order of five monolayers of Ba stearate is extremely labile, and this proved to be true for all antigen-antibody systems tested: human growth hormone, bovine serum albumin, Schistosoma mansoni, polysaccharides from Type 111 and Type VIII pneumococcus, and their respective antisera.

7 . Nickel-Plated Slides As mentioned above, a magnetic field influences markedly the activity of a nickel-plated slide. The activity can be measured by the maximum dilution of either antigen or antibody with which one can detect an interaction. Maximum activity is obtained when the lines of force are I S to the slide, the metallized surface facing the south magnetic pole. A slide on which the nickel has been evaporated in the presence of a magnetic field keeps its activity for months with a daily periodic variation. 8. Activation-Inactivation An active nickel slide can easily be inactivated by a magnetic field as small as 200 Gauss with its lines of force parallel to the field ([I). Heating an inactive slide at 80°C for 1 minute inactivates it. An inactivated slide can be reactivated by a magnetic field I S , sometimes as small as a few hundred Gauss. Activation-inactivation is a reversible process. An active slide can be inactivated, 5

279

INTERACTIONS AT A LIQUI&SOLID INTERFACE TABLE 111 THICKNESS I N A OF THE LAYERADSORBED FROM AN ANTIBOVINE ALBUMIN SERUM ON A BOVINE ALBUMIN-COATED SLIDEAS A FUNCTION OF THE ORIENTATION OF THE SLIDEI N THE FIELD Orientation in field

II

IN

IS

45" N

45" s

Thickness adsorbed

42

59

68

45

59

(A, reactivated, and reinactivated by subjecting the slide to a magnetic field in turn 11, I S , 11. Freshly metallized slides are more readily affected by a magnetic field than slides that are a few months old. The degree of activation achieved by a magnetic field whose lines of force are I depends on whether the metallized surface is facing the south ( I S ) or the north ( I N ) magnetic pole. The activation is greater for the orientation I S than I N . As expected, when the slide is inclined at 45" to the lines of force, the activation is intermediate between positions I S and 1 , or I N and (I. The results have been summarized in Table I11 (30). The slides densely coated with bovine albumin were immersed for 7 minutes in the antiserum diluted 1/200. All experiments were carried out early in the morning. One disadvantage of nickel slides for performing immune electroadsorption assays is that the slides can never be positively charged since the nickel layer becomes disrupted under the influence of the current. The adsorption of the antigen has to be carried out with a current for all concentrations of < lop6g/ml, the slide being negatively charged. The specific adsorption from a homologous antiserum varies over the range of 30 to 40 A for positively charged chromium slides as well as for uncharged nickel ones, when the average thickness of the antigen layers is of the order of A. When the antiserum rather than the antigen is highly diluted, chromium slides permit the performance of assays at much smaller concentrations of the antiserum than do nickel ones, which cannot be positively charged. For example, using chromium slides densely coated with polysaccharide from Type Ill pneumococcus, an immunologic reaction could be detected with an antiserum down to dilutions of 1/400,000. A sensitivity of the same order of magnitude was obtained in the case of an immune serum against arbovirus (28) where the serum had to be diluted more than 1/100,000 before specific adsorption became undetectable. 9. Influence of Dilution of an Anti-Bovine Albumin Serum on Its Interaction with Bovine Albumin Densely Coating a Nickel Slide For this assay some slides were submitted to the magnetic field I S and therefore were active, whereas others were submitted to the field 11 and became

280

ALEXANDRE ROTHEN

inactive. All slides were then densely coated with one molecular layer of bovine albumin. The layer adsorbed was 35 A thick. After immersion in a diluted antiserum, a layer 200 8, thick (more than one molecule) was adsorbed. This thickness is independent of whether the slide is active or inactive, and decreases linearly with the decrease in concentration of the antiserum until a critical concentration of 0.013 is reached. From the critical concentration down, the thickness adsorbed on an active slide was greater than on an inactive one. For one particular serum the thickness adsorbed at the critical concentration was 120 A. For a concentration of 0.004, 90 A was adsorbed on an active slide and less than 10 A on an inactive one. The critical concentration of 0.013 is so sharply defined that it can be determined within a very narrow range of concentrations. For one particular antiserum the value of the critical concentration was 0.0128 k 0.0002. Results have been summarized in Fig. 2 (31). It is very significant that, depending on the antiserum and the slide, the amount adsorbed on an active or an inactive slide varies within wide limits, but the critical concentration remains the same. To compare a thickness adsorbed on a slide in the active state with that adsorbed in the inactive one, it is essential to use the same solution and the same slide and to do the experiments in quick succession. It is better to use the active slide first and then inactivate it because it is more difficult to reactivate completely an inactive slide than to inactivate an active one. -L

140

I20

In

f

60

r 0 .-

40

20

.002

.OM

,010

.014

.018

.022 ,026

,030

Concentrotion of antiserum

FIG.2. Thickness in Angstroms adsorbed from an antibovine serum albumin as a function of the concentration of the serum or slides subjected to the magnetic field either I S or It.

28 1

INTERACTIONS AT A LIQUID-SOLID INTERFACE

The same critical concentration was observed for the system ovalbumin-antiovalbumin. The slide was first thickly coated with ovalbumin. The thickness adsorbed from a diluted antiserum was = 5 1 A on an active as well as on an inactive slide, since the antiserum was not very potent. When the concentration was decreased from 0.013 to 0.010, the thickness adsorbed on the inactive slide dropped spectacularly from 51 to 31 A, whereas the thickness adsorbed on the active slide remained practically unchanged at 50 A. The critical concentration was located between 0.0 13 and 0.010, the value found for the bovine albumin system. A critical concentration of the same order of magnitude has been found for a system using apolactoferritin. When bovine albumin was adsorbed on three or five monolayers of Ba stearate transferred to active and inactive nickel slides, the critical concentration was unchanged and was as sharply defined as in experiments carried out directly on the nickel surface. For all concentrations greater then 0.01, stirring of the solutions (0.5 ml) had no appreciable influence on the thickness adsorbed. However, when the concentration was reduced to 0.005, stirring increased the thickness by 30%. Both curves in Fig. 2 were obtained with stirring. Table IV summarizes the results obtained when the adsorption was lengthened to 15 and 64 hours, using very dilute serum solutions. For a dilution of 1/1000, the thickness adsorbed on the active slide was twice that obtained on the inactive one. It is important to notice that the thickness adsorbed on the inactive slide at that dilution was more than twice the thickness adsorbed when the slide was coated with the heterologous antigen human y-globulin. In other words, even at high dilution an appreciable amount was specifically adsorbed on an inactive slide. The activity of a slide mainly influences the specific and very little the nonspecific adsorption from a homologous antiserum. There is little or no difference in the thickness adsorbed from the same antiserum on an active or an inactive slide coated with a heterologous antigen, as is evident from Table IV. This important result favors the view that the long-range order of the nickel domains necessary for a strong interaction between antigen and antibody does not play a role in nonspecific adsorption. The critical concentration is easily determined with an inactive slide because it is the concentration below which the adsorbed thickness is less than that observed with an active slide and it is also at the point where a break occurs in the curve of adsorbed thickness versus concentration. The difference between active and inactive slides results from different orientations of the nickel magnetic domains and their boundaries, allowing the antigen molecules to be subsequently adsorbed with different orientations in the long-range order. An important question is: why is there a critical concentration located within such a narrow range of antiserum concentration? The sharpness of the critical L-

282

ALEXANDRE ROTHEN

TABLE N ADSORPTION OF ANTIBODIES FROM A VERY DILUTEANTISERUM AGAINST BOVINE SERUM ALBUMIN THICKLY COATING ACTIVEAND INACTIVE NICKEL SLIDES

I

Slide Field Antigen Immersion time in antiserum (hours) Antiserum dilution Thickness adsorbed from antiserum

(A)

2

3

4

5

6

I

IS II IS II IS II IS Bovine Bovine Bovine Bovine Human albumin albumin Ovalbumin Ovalbumin albumin albumin y-globulin

8

II Human y-globulin

15

15

15

15

64

64

14

14

1/400

1/400

11400

1/400

1/1000

1/1000

1/1000

111000

100

66

19

13

13

36

16

16

concentration precludes an adsorption mechanism based on antibody molecules being adsorbed independently of each other to the antigen-coated slide by a simple diffusion process. Presumably because of cooperation, ordered assemblies of molecules interact as a whole. At the critical concentration the number of antibody molecules in 0.7 ml of diluted antiserum is roughly 1000 times greater than the number of adsorbed molecules. The sharpness of the critical concentration is probably a good indication of a phase transition. One might assume that the antibody molecules are not randomly distributed in the thin layer of the antiserum solution close to the interface on account of a long-range interaction with the antigen molecules orderly adsorbed. It would be this orderly phase of antibodies that would be suddenly disrupted or melted, so to speak, when the critical concentration is reached. The difficulty with this interpretation is that the same critical concentration is observed for different systems with antisera of different titers. This excludes the possibility that the critical concentration occurs at a definite concentration of antibodies. It might be that the stability of the orderly phase of antibodies is determined by the total protein content of the diluted antiserum. It would be important to determine whether the addition of serum proteins to the antiserum solutions would decrease the critical concentration. Unfortunately no data are available on this point. When the antiserum solution is relatively concentrated, the thickness of the adsorbed antibody layer is the same on active and inactive slides. Below the critical concentration the antigen-antibody interaction is not strong enough on inactive slides to allow the adsorption of as much antibody as occurs on active slides. The stronger the interaction the smaller the critical concentration. It is

INTERACTIONS AT A LIQUID-SOLID INTERFACE

283

most important to emphasize the effect shown in Table I11 showing that a slide I N is not as active as a slide 11. Therefore the critical concentration determined with a slide I N will be lower than with a slide 11. The critical concentration decreased from 0.0128 to 0.0100 when the active slide ( I S ) is submitted to the field ( I N ) instead of ([I). A field of attractive electromagnetic forces is probably present between the antigen molecules adsorbed with a definite long-range order and the antibody molecules situated in the thin layer of the antiserum solution adjacent to the slide. Two opposite forces would be at play concerning the stability of the assumed orderly assembly of antibodies. On the other hand, the concentration of proteins diminishes with the dilution of the antiserum. It is known that the dielectric constant is increased by the presence of proteins. Therefore, electrostatic forces of repulsion, occurring between antibody molecules of the same charge, increase with dilution tending to disrupt the ordered assemblies. The assemblies on an inactive slide melt at the critical concentration for which the forces stabilizing the assemblies equal those disrupting them. Below the critical concentration the electrostatic forces become stronger than the forces stabilizing the assemblies. This explanation for the sharpness of the critical concentration is of course based on many assumptions. Nevertheless it is most important to keep in mind two hard facts: (1) the lower the critical concentration the stronger the interaction, and (2) the sharpness of the critical concentration indicates a phase transition. It should be remembered that the phase considered here involves antibody molecules located in a thin layer of solution, perhaps 1000 A thick in the immediate neighborhood of the interface. 10. Influence of the Dilution of Antigen Solutions Used for Coating Nickel

Slides on the Interaction with Sera Diluted 1/10 in Verona1 Buffer, p H 7.5 Table V (32) summarizes the results obtained with bovine albumin and human y-globulin systems and their respective antisera. Active nickel slides were used and the slides were negatively charged for the adsorption of bovine albumin and human y-globulin for 2 minutes with a current of 0.3 mA. No current was used for the 7-minute adsorption from the diluted antisera. It is apparent from Table V that an active slide permits the detection of an to 10- l 3 g/ml of antigen immunologic reaction down to concentrations of solutions. The controls were obtained on the same slide when no antigen was present in the solution. For a concentration of 10- I 2 g/ml, 37 to 43 A were adsorbed from the immune serum to bovine albumin, compared to 28 when no bovine albumin was present on the slide or when the slide had been inactivated by the field 11. The concentration of the antigen solution must be greater than lo-’ g/ml to detect an immunologic reaction on an inactive slide. For all antigen concentrations smaller than ]OW7 g/ml no specific adsorption from an immune

284

ALEXANDRE ROTHEN

TABLE V SERUM OR A N ANTI-7-GLOBULIN SERUM THICKNESS ADSORBEDIN A FROM A N ANTIBOVINE ALBUMIN A FUNCTIONOF THE CONCENTRATION OF THE BOVINE ALBUMIN SOLUTION OR OF THE HUMAN yGLOBULIN SOLUTION USEDTO COATTHE NICKEL SLIDES Concentration bovine albumin 10-13 0 10-12 0 10-11 0 10-1" 0 (gW Active slides 35 29 37,43 29 44 30 53,46 36 46 31 Same slide inactivated 11 28 I N 22 30 Same slide reactivated I S 43 45 Concentration human 7-globulin (@I) Active slides Same slide inactivated

10-13 0 10-12 0 l o - " 46

11

41

47

33

55

0

IO-IO

10-8

0 lo-'

AS

0

66 37 77 49 59 38 36

43 39

0 10-9 0 10-8

0 10-7

0 10-7

0 10-6 0

- 47,47 38 50,54 38 5 7 3 8 38 41.43 35 41 34 45 50 35

38.36

L

43 J cuzent for deposition of antigen

NO

serum can be observed on an inactivated slide. In other words, the inactivation by the field cuts down the sensitivity of the assay by four to five orders of magnitude. It is important to note that the activation-inactivation is reversible. A slide activated by the field 11 can be reactivated by the field I S . When the concentration of the solution of bovine albumin was 10- g/ml, the slide in its active state could adsorb successively 53, 46, and 46 A from the antiserum; 36 and 31 A were adsorbed in the control experiment. After inactivation 30 A was adsorbed. Upon reactivation, 45 A was again adsorbed from the antiserum. For all concentrations of antigen greater than 10 g/ml, the adsorbed thickness from the antiserum is the same on active and inactive slides. In other words, when the concentration of the antigen solution is large, the long-range order of superficial nickel domains plays no role in the thickness of the adsorbed layer of antibodies. These results are analogous to those described previously where the thickness adsorbed from an antiserum was the same on active and inactive slides densely coated with antigen, provided that the concentration of the antiserum was greater than 0.013. The main difference is that the critical concentration in the latter case is sharply defined whereas the critical concentration determined as a function of ~

-

INTERACTIONS AT A LIQUIBSOLID INTERFACE

285

the antigen solution concentration can be determined only within one order of magnitude. The number of antibody molecules specifically adsorbed per molecule of bovine albumin present on a slide, as a function of the concentration of the albumin solutions, is presented in Table V1. These numbers are readily obtained from Table V and from the molecular weight of bovine albumin (= 65,000). It is assumed in the calculation that all the albumin present in the 0.5 ml of the very dilute solutions has been adsorbed. This is far from being the case, and the true ratios must be much larger. The results are striking; they demonstrate that when the antigen molecules are sparsely adsorbed with long-range order a very large number of antibody molecules can be specifically adsorbed. There is a complete breakdown of stoichiometric relations between antigen and antibody for all concentrations of albumin solutions < 10- g/ml. It may be no coincidence that this concentration, for which the antibody/antigen ratio is close to 1 , is the limiting concentration below which no immunologic reaction can be detected when using inactive slides. It is apparent from Table V that the same kinds of results were obtained with the system human y-globulin anti-human y-globulin. This is also true for the polysaccharides from Types 111 and VIII pneumococcus and their antisera, as can be seen from Table VII. Immunologic reactions were still detectable with a solution containing as little as 6 X 10-I6 g/ml of polysaccharide from Type VIII pneumococcus and an antiserum diluted 1/10 in veronal buffer. It also appears from the table that there is a strong cross-reaction between the polysaccharides from Types 111 and VIII pneumococcus. In these experiments the polysaccharides were adsorbed on chromium plated slides coated with five monloayers of Ba stearate. Slides so coated are very active provided that a weak current is used for the deposition of the antigen from very weak solutions. The sensitivity of the assay is cut down by = five orders of magnitude and the large antibody-antigen ratio is not observed once the longrange order of the stearate layers has been disrupted by heating the slide to 50°C

’

TABLE VI RATOOF ANTIBODIES TO ANTIGEN MOLECULES Concentration bovine albumin Wml)

Ratio antibody molecules to antigen

10-12 10-11

63,000 8,000 850 98

10-10 10 -9 lo-* 10-7

10

1.6

286

ALEXANDRE ROTHEN

TABLE VII THICKNESS OF ANTIBODY LAYERI N A ADSORBED ON CHROMIUM-PLATED SLIDESWITH THE HELPOF AN ELECTRIC CURRENTFOR THE DEPOSITION OF BOTHANTIGEN AND ANTI BODY^ Rabbit antiserum against polysaccharide from Type VIIl pneumococcus

Rabbit antiserum against polysaccharide from Type I11 pneumococcus Conc. poly (giml)

Poly 111

Poly VIII

Conc. poly

Poly I11

Poly VIII

(A,

(A)

WmU

(A)

(A)

3 x I0-'2 3 x 10-13 3 x 10-14 0

80 74,78 60 60

50 55,50 56

x x x x

98 94 I08

122 115

3 3 3 3 3

10-1" 10-11 10-13 10-'6 x 10-1*

77

120 86

75

77

aOne minute for both adsorption periods.

or, in the case of nickel slides, by submitting them to a field 11 or by heating to 80-90"C, as illustrated in Table V. Unordered adsorbed antigen molecules behave in the same way as randomly distributed antigen molecules in solution where the antibody-antigen ratio is of the order of one. 1 1. Influence of a Carrier in the Antigenic Solution on the Sensitivity of lmmunoelectroadsorption Assays It seems obvious that the presence of foreign proteins in the antigen solutions should decrease the sensitivity of the assay. When the antigen solution consisted of 1.7 X 10- g/ml of bovine serum albumin dissolved in 2% rabbit serum, it was remarkable to observe that 117 A was adsorbed in 4 minutes from a rabbit antiserum against bovine albumin and 82 A when no albumin was present in the 2% rabbit serum. In this particular case the ratio by weight of antigen to foreign proteins was -IOW7. Inconel slides were used for these experiments (33). 12. Applications of the Immunoelectroadsorption Method It might be thought that the simplest way to perform an immunoelectroadsorption assay, in order to differentiate homologous from heterologous antisera with respect to a given antigen, would be to coat the slides first with a layer of antigen and then to immerse the slides in the antisera to be tested. The thickest adsorbed layer would be given by the homologous antiserum. This procedure may lead to erroneous conclusions because the heterologous may have a larger globulin content than the homologous, and nonspecific adsorption from the heterologous serum may be larger than specific adsorption from the homologous one. This may be the pitfall that Williams and co-workers (34) have fallen into

INTERACTIONS AT A LIQUIPSOLID INTERFACE

287

when they reached the conclusion that the immunoelectroadsorptionmethod was unable to distinguish true immunologic reactions. The assay should be conducted by comparing the thickness adsorbed from one antiserum on two parts of a slide, one coated with a carrier plus antigen and the other coated with a carrier only, and not by comparing thicknesses adsorbed from different antisera on slides coated with the same antigen. The use of a carrier such as 2% human or rabbit serum is recommended when the antigen is a small fraction of the antigenic preparation. When the thickness adsorbed from an antiserum is the same on the two parts of the slide, one coated with antigen and carrier and the other with carrier only, it can be said that the serum is normal or heterologous. A carrier is unnecessary when very dilute solutions of pure antigen are used. The control slide is then first treated with the buffer used for dissolving the antigen. The immunoelectroadsorption method has been successfully applied to the identification of eight arthropod-borne viruses (28). Dengue virus of types 1, 2, 3, and 4 were also tested with their homologous antisera. In all cases the homologous reactions could be identified and weaker cross-reactions were observed with dengue type 2 antigens against antisera for type 3. The sera from five patients who had suffered from St. Louis encephalitis 1 or 2 months before blood samples were taken were assayed with this method. The presence of specific antibodies was readily established in all five patients. The sera had to be diluted 1/1000 before the immune reaction became undetectable. The method was also applied to the quantitative determination of human and bovine growth hormones. The limit of sensitivity was 0 . 2 X l o p 6 mg/ml (35). The appearance of antibodies against Friend’s virus was observed by immunoelectroassay as soon as 2 days after infection (36). This assay proved also to be very sensitive in the detection of leishmaniasis infection (37). Serodiagnosis of Schistosoma mansoni was possible by immunoelectroadsorption (38). It appears, however, that a cross-reaction takes place with other common helminths (19). There is no question but that immunoelectroadsorption could be applied to many other diseases for serodiagnosis. 13. Interactions through a Membrane One of the key experiments that led to the postulation of long-range forces in antigen-antibody reactions was the use of barrier membranes and the demonstration that specific immunologic forces were in fact transmitted through such membranes. A basic experiment was carried out as follows. Two nickel slides were used, one in which the nickel was active and the other rendered inactive by submitting the slide to a few thousand Gauss )I for 2 minutes. Formvar membranes of different thicknesses were formed on the slides after they had adsorbed a monomolecular layer of bovine albumin 30 to 35 8, thick. The bovine albumin was adsorbed from an aqueous solution containing

288

ALEXANDRE ROTHEN

g/ml and lasted 2 minutes. After being coated with Formvar membranes of varying thicknesses, the slides were dipped for 7 minutes in a diluted antiserum. Two curves have been drawn (Fig. 3), one obtained with the active nickel slide and the other with the inactive one. With all Formvar membranes thinner than 80 A, the thickness of the antibody layer was essentially the same on both types of slides. With membranes between 90 and 100 8, thick, there was a sharp drop in the thickness adsorbed on an inactive slide. No specific adsorption was observed when the membranes were thicker than 130 A. In contrast, the decrease in the amount of antibody adsorbed on an active slide was more gradual with the increase in membrane thickness, and specific adsorption was still observed with membranes over 200 A thick. These curves illustrate the influence of a long-range order of nickel domains on the range of interaction between antigen and antibody. Table VIII shows convincingly that the activation-inactivation process is reversible. The same slide was used for all assays. It was activated for 2 minutes (IS),and a monomolecular bovine albumin layer was adsorbed on its lower part on which a Formvar membrane 97 A thick was then deposited. A layer 83 A thick was adsorbed from a homologous antiserum. The used part of the slide was then cut off, the slide was inactivated ([I), and the whole cycle was repeated. Altogether four cycles were performed, the slide being alternately active and inactive. It can be seen from the table that the thickness adsorbed from the

0

40

80

120

160

200

Membrane thickness in A

FIG. 3. Thickness adsorbed from an antibovine albumin serum on a slide coated with bovine albumin protected by a Formvar membrane.

289

INTERACTIONS AT A LIQUID-SOLID INTERFACE

INFLUENCE REVERSIBLE

Condition of slide Thickness Formvar membrane (A) Thickness adsorbed from antiserum (A)

OF THE

TABLE VIII MAGNETIC FIELD DEPENDING O N ITS ORIENTATION, 1 OR WITH RESPECTTO THE SLIDE

11,

Activated IS

Inactivated

Reactivated IS

Reinactivated

II

97

94

106

I10

83

48

74

30

II

antiserum on an active slide was twice that adsorbed on an inactive one when the membrane thickness was in the range of 110 A. 14. Locution of Adsorbed Antibodies in the Presence of a Membrane Burrier Deposited upon the Antigen Layers When the antigen-antibody reaction through a membrane was first described (21) it became widely assumed that the reagents were making contact through holes in the membranes and that the postulation of long-range forces was unnecessary.* Since that time a great deal of work has been done which makes the assumption of reaction through holes totally inadequate. If free diffusion is not the explanation for reactions through a membrane, where are the immobilized antibodies located when reacting through a protective membrane? Have they, because of a “forced diffusion” process, penetrated the membrane and then become anchored to the antigen layer, or are they immobilized on the upper surface of the membrane because of long-range interaction with the antigen? The following experiments carried out in 1972 (32) indicate that the antibodies remain specifically immobilized on the upper surface of the membrane when homologous antigen molecules are situated underneath. Bovine albumin was adsorbed for 2 minutes on an active nickel slide from an aqueous solution. The layer adsorbed was 35 A thick and -100 A was subsequently adsorbed on the lower part of the slide from a diluted homologous immune serum. The used lower part of the slide was then cut off, a Formvar membrane 164 A thick was formed on the slide, and a layer 46 A thick was *As late as 1967, in the Translation Supplement of the Federation of American Societies for Experimental Biology (39). the Russian authors Deborin e t a / . referred to one of my papers (40). One editor of the journal took it upon himself to add the following footnote: “Rothen’s work has long been discredited due to minute holes in his membrane as revealed by microscopy, Ed.” In this connection, see the recent paper by Semka (41).

290

ALEXANDRE ROTHEN

TABLE IX OF ANTIBODY LAYERADSORBED ON TOPOF A FORMVAR MEMBRANE IN FOUR THICKNESS SUCCESSIVE CYCLES“ Membrane thickness

Thickness antibody layer

Cycle

(A)

(A)

1

I64 158 195 142

46 34 33 46

2 3 4

Thickness remaining after C2H4Cl2 treatment

I 5 13 0

OThe membrane is dissolved at -4O”C, and reformed after each cycle.

adsorbed from the antiserum. The slide was cooled to -40°C and rapidly washed with ethylene dichloride. This operation removed not only the membrane but also the antibodies since, as determined by the ellipsometer, the material remaining on the slide was only 7 A thicker than the original bovine albumin. The whole cycle was repeated four times and the results are summarized in Table IX. The total thickness in four successive operations in the presence of a membrane 142 to 195 A thick was 159 A, compared to 100 A, the maximum adsorbed directly on the bovine albumin layer without a membrane present. In another similar experiment a total of 237 A was adsorbed from the same antiserum after five successive adsorptions, the membrane being dissolved at -40°C after each adsorption. A control experiment was carried out by adsorbing an ovalbumin layer 34 A thick on an active slide. A layer only 4 A thick was adsorbed on this slide from the same antibovine albumin serum instead of 100 A when bovine albumin had been adsorbed instead of ovalbumin. In the presence of a Formvar membrane 94 to 131 A thick there was no adsorption from this antiserum. The adsorption shown in Table IX thus appears to be truly specific. Antibodies were not removed when the Formvar membranes were dissolved with ethylene dichloride at room temperature. At the higher temperature the rate of interaction antibody-antigen is fast enough to prevent antibodies from being washed off by ethylene dichloride. When antibody is adsorbed directly on bovine albumin monolayers, they cannot be removed with ethylene dichloride at any temperature; the binding antigen-antibody is too strong. 15. Strip Technique Another method named the “strip technique” has been devised to remove antibodies adherent on a slide in the presence of a Formvar barrier. It consists in pressing a piece of Scotch tape on the slide to be stripped. Upon removal of the tape all the layers, with the exception of the first one deposited, are stripped off.

INTERACTIONS AT A LIQUID-SOLID INTERFACE

29 1

The splitting takes place wherever binding is weakest. It may happen between the last layer deposited and the tape, in which case no stripping occurs. The “strip technique” may be considered to be performing the task of a “molecular microtome” (42, 43). An active nickel slide was coated with seven double layers of bovine albumin, transferred from a Langmuir trough, with a total thickness of 77 A.A layer 99 A thick was then adsorbed from an immune serum. This low value was due to the fact that, in order to transfer a number of double layers of bovine albumin, the film was compressed to 10 dynes, resulting in deformation of the layers. The used part of the slide was cut off and a Formvar membrane 250 A thick was transferred to the remaining part of the slide. It was then possible to adsorb 36 A from the same antiserum. After stripping 69 A remained on the slide, corresponding to the thickness of the albumin layers (77 A). Stripping had removed antibodies plus membrane. The slide was again treated with antiserum and a layer 90 A thick was adsorbed, roughly the same as that adsorbed prior to the deposition of the Formvar barrier. The bovine layers which permitted the adsorption of 36 A of antibody in the presence of the membrane were still able to adsorb 90 A after removal of the 36 A of antibody plus the membrane. When the antibodies were adsorbed directly on antigen layers it was impossible to remove them by stripping, the adhesion of antigen to the slide and of antibodies to the antigen being stronger than the adhesion of the strip to antibodies. The same experiment was performed with an inactivated (11 2 minutes) slide, metallized at the same time. The slide was also coated with seven transferred double layers of bovine albumin and it was able to adsorb 83 A from the same antiserum. After cutting off the used portion of the slide, a Formvar membrane 228 A thick was transferred to the slide, but no increase in thickness took place on exposure to antiserum. This demonstrates that a definite long-range order of antigen molecules is necessary for a long-range interaction. In another example eight double layers of bovine albumin were used and the upper part of the active slide was covered with 211 A of Formvar; 44 A was adsorbed from antiserum. After stripping, the ellipsometer reading indicated that all the antibody layer had been removed but that the membrane was still present. On treatment with the antiserum 54 A was adsorbed. 16. Tryptic Action through an Intervening Membrane The interaction between antigen and antibody through a barrier can also be shown between enzyme and substrate (protein layers separated from trypsin). In all likelihood the trypsin molecules remain on the surface of the barrier. In both cases the long-range order of the adsorbed molecules exerts a fundamental influence on the range of the interaction. Thicker Formvar membranes are needed to prevent tryptic action if the subjacent albumin layers are adsorbed with a definite long-range order imparted by the long-range order of the anchorage. The same

292

ALEXANDRE ROTHEN

kind of long-range order is needed to render a nickel slide “active” for immunologic as well as tryptic action. The magnetic field has to be I S to activate and 11 to inactivate a slide. Heating at 80°C inactivates an active slide for antigen-antibody reactions, but does not affect the activity necessary for tryptic action. Tryptic action occurs through a Formvar membrane 600 8, thick protecting six “up” layers of bovine albumin anchored to Ba stearate layers transferred to chromium slides. However, a Formvar membrane 130 to 200 8, thick, coated on its upper surface with a gold film 40 to 60 8, thick, full of lacunae, offers complete protection to six “up” underlying layers of bovine albumin. Tryptic action most probably takes place through electromagnetic radiation, not necessitating a direct contact between enzyme and substrate. This type of radiation is screened off by conducting surfaces, and this would explain the inhibition of tryptic action by a gilded membrane (44). Long-range interaction explains the different rates of tryptic hydrolysis in the presence or absence of a membrane when an increasing number of bovine albumin layers is located under the membrane. The experimental evidence is as follows. Chromium slides were coated over their whole length with three conditioned monolayers of the Ba salt of the acid with 23 carbon atoms. Unfolded bovine albumin layers were transferred to the slides from the surface of a Langmuir trough in a stepwise fashion, with nine monolayers at the bottom of the slide, six in the middle, and three at the top. The hydrolysis of the albumin layers by

INFLUENCE OF A

TABLE X FORMVAR MEMBRANE ON THE RATEOF HYDROLYSIS OF BOVINE ALBUMIN LAYERS~ Number of double layers of bovine albumin

Thickness in 8, of antibody layer adsorbed (no trypsin) Thickness in 8, of antibody layer adsorbed after trypsin action through 600 8, Formvar for 3 minutes and dissolution of the membrane Thickness in 8, of antibody layer adsorbed after trypsin treatment directly for 3 minutes

3

6

9

127

222

318

I27

60

40

70

112

161

OTransferred in a stepwise fashion on chromium slides uniformly coated with a conditioned anchorage of three monolayers of the Ba salt of the acid with 23 carbon atoms.

293

INTERACTIONS AT A LIQUIPSOLID INTERFACE

trypsin was carried out either directly with a very dilute solution of trypsin (0.001% in veronal, pH =7.5) or after the deposition of a Formvar membrane 600 thick, using 0.05% trypsin. The results have been summarized in Table X (42). In the absence of a membrane, the step coated with nine layers of bovine albumin was able to adsorb, after trypsin treatment, a thicker layer of antibody than the step coated with six layers, and the thickness adsorbed on the latter was greater than that adsorbed on the step with three layers. This is what could have been anticipated. On the other hand, the reverse was true when tryptic action took place in the presence of a membrane. The step with three double layers was not hydrolyzed at all. The step with nine double layers can be seen to have been almost completely hydrolyzed. The long-range interaction across the membrane increases with the number of underlying layers. 17. Tryptic Action on Bovine Albumin Anchored Directly on Nickel Slides Tryptic action through intervening Formvar membranes on subjacent bovine albumin layers directly attached to active nickel slides is extremely strong, much stronger than when bovine albumin is anchored to long-chain fatty acids. In other words, thicker membranes are necessary to prevent tryptic action. In the experiments summarized in Table XI (32) the albumin layers were transferred from a Langmuir trough onto either active or inactive slides. After the formation of the Formvar membrane a drop of trypsin solution was applied for 2 minutes on a limited area of the slide. The membrane was then dissolved in ethylene dichloride, the slide placed for 7 minutes in an antiserum solution, and the thickTABLE XI MEMBRANES ON BOVINE ALBUMIN LAYERS TRYITICACTIONTHROUGH FORMVAR TO NICKEL SLIDES TRANSFERRED

Slide

Slide condition

Bovine albumin

Formvar thickness

(A)

1

Active

(Jrh

198

2

Inactivated )I 2 minutes Inactivated IN 2 minutes Reactivated I S 2 minutes

(Irh

205

(it)'$

I30

(m3

I75

3

3

Trypsin treatment

Thickness in A adsorbed from antiserum after dissolution of membrane

Yes No Yes No Yes

14 65 15

No Yes No

61

65 58

20 39

294

ALEXANDREROTHEN

ness of the layer adsorbed was determined on two areas of the slide, one over which the trypsin drop had been located and the other one not submitted to trypsin. It clearly appears from the figures in the last column of Table XI that for active slides the thickness adsorbed from the antiserum is considerably less on the area above which trypsin had been located than on the area not treated with trypsin. For inactivated slides no enzymatic action took place through the same thickness of Formvar or even a thinner one. The thickness adsorbed from the immune serum is the same on the trypsin-treated area and on the nontreated one. Without a protective membrane tryptic action takes place very rapidly, whatever the concentration of bovine albumin solution used for coating the slide. With a solution whose concentration was 10- ‘ I g/ml, 15 8, only was adsorbed from an immune serum after trypsin treatment, whereas 50 8, was adsorbed on an area of the same slide not treated with trypsin. Two conditions are required to detect trypsin action through a thick Formvar membrane (-200 A). Nickelcoated slides must be active and the concentration of the albumin solution used to coat the slides must be greater than lo-’ g/ml. Depending on the potency of the immune serum, the thickness adsorbed on an antigen-coated slide not treated with trypsin may vary within wide limits. The values given were all obtained with the same antiserum in experiments performed within a few days.

18. Bovine Albumin Layers Anchored to Long-Chain Fatty Acid Layers Deposited on Active Nickel Slides The purpose of these experiments was to find out whether it was necessary to have the antigen in direct contact with the nickel surface in order to observe the orienting influence of the magnetic domains and their boundaries on the slides’ activity. It was found that no intimate contact between slide and protein layers was necessary to demonstrate whether a slide was active or inactive. This confirms the assumption that the magnetic field of the oriented nickel domains is the factor contributing to the orientation of the albumin layers. When the slides were coated with five monolayers of Ba stearate, trypsin action was stronger when the lines of force were parallel to the slides, and weaker when perpendicular. This is just the opposite of what occurs when no monolayers or three monolayers of stearate are present on the nickel slides. 19. Breath Figure Test

Langmuir and Schaefer (45) noticed that a “sensitive indication of the uniformity of an initial layer” (of Ba stearate) “may be obtained by cooling the back of the plate with running water and breathing on the front side so as to form a foglike deposit of minute water drops. Lack of uniformity is made apparent by this “breath figure.” This simple test proved very useful for detecting whether an assembly of protein layers protected by a membrane had been acted on by a proteolytic

INTERACTIONS AT A LIQUIWSOLID INTERFACE

295

enzyme. The test does not necessitate the removal of the membrane or the use of antiserum (42). After the trypsin drop has been washed off, water vapor, on breathing, condenses on the surface in uniform small droplets if no proteolytic action has taken place. If the protein layers located under the Formvar membrane have been hydrolyzed, the number of charges per unit area has increased and the droplets of condensed fog are much larger on the area previously occupied by the trypsin drop; this area appears shiny in contrast to the dull appearance of the rest of the slide. The mechanism of this phenomenon is probably akin to what occurs in a Wilson cloud chamber. There is excellent agreement between shininess and degree of hydrolysis, as testing with antiserum has shown. The greater the hydrolysis, the shinier the appearance of the condensed fog. This test is especially striking when there is a series of steps of barium salt of long-chain fatty acids under the protein layers and when tryptic action occurs in the presence of a protective Formvar membrane. A chromium-plated slide was coated step-wise with layers of barium stearate. There was a difference of two monolayers between each successive step, progressing from 13 monolayers at the bottom to one monolayer at the top. The deposition was carried out at pH 7.5, so the layers were a mixture of free acid and its barium salt. There was BaCl, in the solution filling the trough. The layers were conditioned with uranyl acetate. Three double layers of bovine serum albumin were transferred to the whole slide that was heated to 75°C for 10 minutes. Heating can also be performed before the protein deposition. A Formvar membrane 70 A thick was formed on the slide which was then treated with trypsin solution (0.1%, pH 7.5). After washing off the trypsin, one can observe the breath pattern seen in Fig. 4. The contours of two trypsin drops deposited on the slide are apparent. The trypsin drops on the left and on the right remained on

FIG.4. Tryptic action through a Formvar membrane 70 A thick on three double layers of bovine albumin deposited on I , 3, 5 , 7, 9, I I , and 13 monolayers of Ba stearate.

296

ALEXANDRE ROTHEN

the slide 3 and 1 minutes, respectively. The different numbers of monolayers forming the step ladder of the anchorage have been indicated. It is clearly apparent that steps 1 , 7, and 9 have been inactivated while steps 3 and 5 were protected 3 a little more than 5, as shown by the right drop. Quantitative results obtained when the ladder consisted of 1 , 3, 7, 11, 15, and 17 monolayers of the Ba salt of the straight chain acid with 23 carbon atoms are presented in Table XII. Four double layers of bovine albumin were deposited on top of the ladder. Those anchored to 7 and 17 layers were considerably hydrolyzed. A periodicity in the hydrolysis as a function of the number of underlying fatty acid layers is clearly apparent. It is essential, in order to obtain a clear-cut pattern in the hydrolysis of the protein layers, that a membrane be present on top of the bovine albumin layers during trypsin treatment. In its absence the pattern disappears almost completely, which is evidence that trypsin does not diffuse through the membrane. To obtain a clear-cut breath pattern, the following conditions have to be observed. Heating the slides at 75°C before or after deposition of the protein layer is necessary. The fatty acid layers are in general formed at pH 7.5-7.7. There is a loss of = 10 A in optical thickness when a slide coated with five monolayers of stearic acid is heated at 75°C for 5 minutes, from the sublimation of the free stearic acid molecules. The two-dimensional crystalline structure is not destroyed by this step. When the stearate layers are formed at pH 10, the layers consist most entirely of Ba stearate with no free stearic acid present. In this case, no marked difference is observed on heating upon trypsin treatment in the presence of a protective membrane, in the degree of hydrolysis of the protein layers as a function of the number of subjacent layers of Ba stearate. Removal of the free stearic acid by mild acid treatment can be substituted for heating in order to TABLE XI1 INFLUENCE OF THE NUMBEROF FATTY ACIDLAYERSI N THE ANCHORAGE UPON THE HYDROLYSIS BY TRYPSIN OF FOURDOUBLELAYERS OF BOVINE ALBUMIN IN THE PRESENCE OF A FORMVAR MEMBRANE 150 A THICK Number of monolayers of the Ba salt of the acid with 23 carbon atoms serving as anchorage (31 A per monolayer) Thickness in 8, of antibody layer adsorbed after trypsin treatment and dissolution of the membrane

1

79

3

I

11

15

17

128

61

110

135

29

INTERACTIONS AT A LIQUID-SOLID INTERFACE

297

produce a breath pattern which is less marked than the one shown in Fig. 4. The presence of lacunae in the anchorage constituted by the Ba stearate layers thus appears to be a necessary condition for the appearnace of the breath pattern. Heating a slide coated with steps of octadecylamine hydrochloride layers located beneath the protein layers also allows the appearance of a breath pattern similar to that shown in Fig. 4. It must be emphasized that, in order to bring about trypsin action on bovine albumin protected by a membrane, it is not necessary to use a solution of trypsin. A monomolecular layer of trypsin transferred to a strip of Scotch tape can be a source of tryptic action. The trypsincoated strip is prepared as follows. A slide is coated with a few layers of Ba stearate and then with a molecular layer of bovine albumin. Next, a monomolecular layer of trypsin is adsorbed by dipping the slide into a trypsin solution for only a few seconds, and this is important. After washing and drying, a Formvar strip is applied to the slide and, on stripping, the monomolecular layer of trypsin is transferred to the strip. The strip is then placed on a slide coated with bovine albumin layers protected by a membrane, with a thin layer of veronal buffer (pH 7.5, = 0.2 mm thick) between strip and slide. After removing the strip, washing, drying, and dissolving the membrane, it can be observed that the bovine albumin layers located under the strip have been inactivated. No inactivation takes place if the strip does not carry a molecular layer of trypsin. 20. Pattern Transfer

Using such a technique, a pattern of inactivation can be transferred from one slide to another as follows. A Formvar membrane 80 A thick was formed on a slide coated with steps of one, three, and five monolayers of conditioned Ba stearate plus four double layers of bovine albumin on top, then heated to 75°C. After 1 minute of trypsin treatment the membrane was dissolved in C,H,Cl,, and the slide stripped. The strip was placed for 6 minutes on a recipient slide uniformly coated with five conditioned Ba stearate monolayers plus four double layers of bovine albumin, with a thin layer of veronal preventing contact between strip and slide. After removal of the strip, washing and treating with an antiserum solution, it was seen that those protein layers of the recipient slide located under that part of the strip corresponding to step 3 of the donor slide were much less hydrolyzed than the layers located under the strip corresponding to either step 1 or 5, as shown in Table XIII. In other words, there was much less trypsin on that part of the strip that corresponded to step 3 of the donor slide than on step 1 or step 5. The lines of demarcation separating the three different areas of the recipient slide were practically as sharp as the corresponding lines of the donor slide. When an antigen-antibody interaction takes place through an intervening membrane, the immobilized antibody layer can be washed off when the membrane is dissolved at -4O"C, as has been described above. It was thought that an

298

ALEXANDRE ROTHEN

TABLE XI11 TRANSWRRED PATTERNOF INACTIVATION BY TRYPSIN THROUGH A MEMBRANE AS A FUNCTION OF THE NUMBER OF Ba STEARATE LAYERSIN THE ANCHORAGE OF THE DONORSLIDE Steps of donor slide Adsorbed thickness of antibodies in 8, on the recipient slide on areas corresponding to the step of donor slide

1

3

5

88

160

10

analogous phenomenon might be observed when an enzymatic action took place across a membrane, and that the enzyme, trypsin in this case, might be washed off by dissolving the membrane at -40°C. A representative experiment was performed as follows (46). A Formvar membrane 120 A thick was formed on top of an active nickelplated slide coated with one monomolecular layer of adsorbed bovine albumin. Three separate drops of trypsin solution (0.1%, pH 7.5) were deposited on the membrane for 15, 30, and 50 seconds, respectively. The membrane was then dissolved either at room temperature or at -40°C. The slide was cut into three pieces, each containing one of the three locations where the trypsin drops had been placed. The three pieces were positioned for 2 minutes on a recipient slide coated with a monomolecular layer of bovine albumin with a thin layer of veronal buffer (pH 7.5) between slide and cut pieces. The thickness of the veronal layer was determined by the weight of the cut piece and by capillary forces. Finally the slide was washed and treated for 7 minutes with antiserum (Table XIV). TABLE XIV INFLUENCE OF THE TEMPERATURE A T WHICH THE MEMBRANE IS DISSOLVED O N OF TRYPSIN REMAINING O N THE SLIDE

AMOUNT

(30 sec)

Trypsin (60 sec)

60

64

50

114

110

99

No trypsin

Trypsin (5 sec)

Thickness in 8, adsorbed on recipient slide from antiserum. Membrane dissolved at room temperature

140

Thickness in 8, adsorbed on recipient slide from antiserum. Membrane dissolved at -40°C

111

Trypsin

THE

INTERACTIONS AT A LIQUID-SOLID INTERFACE

299

The difference is striking between the thickness of the layer of antibodies adsorbed on a “recipient” slide when the membrane of the “donor” slide was dissolved at room temperature or at -40°C. If the trypsin molecules had diffused through the membrane of the donor slide to inactivate the protein layer at close range, then the thickness of the antibody layer adsorbed on the recipient slide ought to have been independent of the temperature at which the membrane was dissolved. The conclusion seems inescapable that the trypsin molecules that had inactivated the protein layer of the donor slide in the presence of the membrane had remained on top of it since they were washed away by ethylene dichloride when the dissolution was carried out at low temperature. When the membrane was dissolved at room temperature, the adsorption of trypsin on bovine albumin apparently was fast enough to immobilize the trypsin molecules on the slide before they could be washed off by ethylene dichloride. 21. Interaction between P O ~ Y - and D Poly-L-Lysine with Trypsin The following experiments also favor the view that trypsin (38, 45) does not diffuse through thin Formvar membranes. For example, poly-L-lysine (MW =300,000) is readily hydrolyzed by trypsin, whereas poly-D-lysine (same molecular weight) is a good inhibitor of trypsin. In solution poly-D-lysine combines with trypsin and the complex shows no tryptic activity. A film of poly-D-lysine 10 to 15 A thick was adsorbed from solution (0.05 M NaHCO,, pH 8.6) on a chromium-plated slide coated with Ba stearate (in general 5 monolayers) [ ( 4 )J. The thickness of the poly-D-lysine layers was independent of the number of stearate layers in the anchorage. When a monomolecular layer of trypsin was adsorbed on top, then the whole pile of layers could be stripped (except the one deposited first). The trypsin transferred to the tape was found to have been inactivated by the layer of poly-D-lysine, since the tape was unable to hydrolyze the bovine albumin coating a recipient slide. When the order of deposition of the poly-D-lysine and trypsin was reversed, no active trypsin was found in the stripped tape. When, however, a Formvar o r Ba stearate membrane intervened between the poly-D-lysine and the trypsin, the results obtained are summarized in Table XV. The condition necessary to inhibit tryptic action was that poly-D-lysine and trypsin had to be in contact. One can conclude from these results that trypsin did not diffuse through the membranes in significant amounts because, if it had, it would have been inactivated when coming in contact with poly-D-lysine. It was consistently observed that a tape stripped from a slide coated with Ba stearate, poly-L-lysine, and trypsin carried much less active trypsin than if a Formvar membrane had been present on the slide before trypsin treatment. If, in the presence of a membrane, trypsin molecules had diffused through it to interact at close range, after stripping the tryptic activity of the tape should be less than or at most equal to the activity recovered on tape stripped when trypsin was placed directly on poly-L-lysine.

300

ALEXANDREROTHEN TABLE XV INFLUENCE OF A MEMBRANE SEPARATING POLY-D-LYSINE FROM T R Y P S I N ~

Exp.

Trypsin (1 min)

Po~Y-Dlysine

Thickness Formvar

(A)

Trypsin (1 min)

Membrane dissolved

Stripped

Thickness adsorbed on recipient slide from antiserum 90 40 20 25

139 139 80 80

0 20

Thickness Ba stearate

(4th 7 8 9 10 II

+ + + +

15 70 20 70

Directly adsorbed 70

aChromium-plated recipient slides were coated with five conditioned monolayers of Ba stearate ( . J r )plus l] three double layers of bovine albumin [( jr)-J over the whole area and heated for 1 minute at 85°C. Donor slides were similarly coated with stearate and albumin layers. They were treated thereafter as indicated in the columns from left to right. [

This was not found, which is consistent with trypsin exerting its enzymatic action without crossing the membrane. Another method to detect trypsin consisted in using tritiated trypsin and counting in a flow-counter the p emission present on a tape stripped from the slide. The sample of tritiated trypsin had an activity of 1.3 mCi/mg, corresponding roughly to one tritium atom per molecule of trypsin. A piece of tape I3 X 4 mm stripped from a slide coated with fatty acid layers and one monomolecular layer of tritiated trypsin on top gave 1400 cpm. When a Formvar membrane 50 A thick was protecting a slide coated with fatty acid layers prior to the tritiated trypsin treatment, the count dropped to 80 cprn. The Formvar membrane had to be at least 200 thick to bring the count down to 80 when poly-L-lysine was present on the fatty acid layers beneath the membrane. In other words, a long-range interaction extended to at least 200 A between poly-L-lysine and tritiated trypsin.

INTERACTIONS AT A LIQUID-SOLID INTERFACE

30 1

22. Factors Influencing the Rate of Hydrolysis of Protein Layers by Trypsin in the Presence of a Protective Membrane Many factors influence the rate of hydrolysis of protein layers by trypsin only when they are protected by a membrane. This in itself is a strong argument against free diffusion of trypsin molecules through a membrane. For example, the rate of hydrolysis depends on the number and the mode of deposition of the protein layers. The greater the number of protein layers, the thicker the membrane necessary to prevent tryptic action. A thicker membrane is needed to protect against trypsin action layers transferred by emersion only than those transferred by immersion followed by emersion. When a large number of double layers of bovine albumin is located under a membrane, tryptic action will hydrolyze them into fragments small enough to diffuse out during trypsin treatment. With a similar number of ovalbumin layers located under the membrane, no such break-up is detectable after inactivation by trypsin and dissolution of the membrane. The ovalbumin no longer reacts with antiserum, but fragments do not diffuse out (42). A Formvar membrane 40 A thick was sufficient to protect one double layer of ovalbumin against tryptic action, whereas 300 A was needed to protect 10 double layers, and 500 A to protect one double layer plus 18 “up” layers [ & ( t ),J. The longer the fatty acid chain in the anchorage, the faster the hydrolysis, but only when a membrane is present. The rate increases when palmitate is replaced by stearate and this by lysergate. Another significant factor is the thickness of the Formvar membrane necessary to prevent the hydrolysis of albumin layers. That thickness increases with the size of the cation of the salt of the fatty acid layers of the anchorage in the order Ca < Sr < Ba, but only when the hydrolysis is carried out through a membrane. It has been shown in the section above on breath figure patterns that, when the anchorage coated slide is heated at 80°C for 5 minutes, the degree of hydrolysis through a membrane is very dependent on the number of fatty acid layers. Protein layers deposited on three monolayers of fatty acid are very resistant to tryptic action through a membrane. A Formvar membrane transferred to a slide before deposition of the anchorage layers plays a crucial role in the rate of hydrolysis. When this “bottom” Formvar membrane was between 50 and 120 A thick (it is difficult to transfer a Formvar film thinner than 50 A), the protein layers deposited on five monolayers of stearate and heated at 70°C were much less inactivated than those deposited on three layers. The minimum in inactivation was thus displaced from step with three to step with five stearate layers. However, when the thickness of the bottom Formvar membrane was increased to 140 A or more, protein layers on three fatty acid monolayers were less hydrolyzed than on five, similar to the result without a bottom Formvar membrane. The results have been condensed in Table XVI (47). After a 3-minute trypsin treatment the Formvar membranes were dissolved in ethylene dichloride before

302

ALEXANDRE ROTHEN TABLE XVI “BOTTOM” FORMVAR MEMBRANE O N THE INACTIVATION B Y TRYPSIN MEMBRANEOF FOURDOUBLE LAYERSOF BOVINE ALBUMIN DEPOSITED O N 1, 3, 5 , 7, OR 9 MONOLAYERS OF Ba STEARATE

INFLUENCE OF A THROUGH A

Slide 1 1

2 2

Thickness in 8, adsorbed from antiserum when anchorage consists of

Thickness of Formvar bottom membrane

Thickness of Formvar membrane

(A)

(A)

70

85 85

32 55

60 102

I44 83

30 2

14

115 115

78

137 118

86 86

65 65

24 24

Without membrane 127

Without membrane

I

53

3 5 7 monolayers of Ba stearate

9

5

immersing the slides in the antiserum. The transition from step 5 to step 3 for the least tryptic action occurred within a narrow range of thickness of the bottom membrane of about 10 A. When sudan black, a diazo dye with two diazo groups in the molecule, was dissolved in the bottom Formvar (conc. 0.1 to 0.4% in C,H,Cl,), the protein layers deposited on step 5 were always more inactivated than those on step 3, even with the thinnest bottom membrane. When less adsorbing dyes such as scarlet BFS and yellow CFS were substituted, the large difference in inactivation between steps 3 and 5 was greatly diminished, but 5 remained less inactivated than 3 for the thinnest bottom Formvar membranes.

IV. Conclusions It is considered that these investigations carried out over the last 30 years have brought to the fore new fundamental facts in biology as well as in physics. A definite long-range order of antigen molecules adsorbed on a slide permits a long-range interaction with antibody molecules, which does not require contacts between antigen and antibody. A specific field of force extending to hundreds of Angstroms results from cooperation between antigen molecules arranged in a definite order. Under these conditions stoichiometric rules break down completely. This is in contradiction to the current view of immunologists, whose credo is that only short-range forces are involved in antigen-antibody interaction, and that appropriate geometry of the interacting molecules is required to bring them into close contact. This view is reasonable when the reaction takes

INTERACTIONS AT A LIQUIWSOLID INTERFACE

303

place in solution between individual molecules, but is unable to explain immunologic reactions occurring at a liquid-solid interface. Only in the last few years has it been possible to demonstrate experimentally the fact that antibodies can be specifically immobilized on a membrane separating them from adsorbed underlying antigen molecules. The long-range order of the antigen molecule is imprinted by the long-range order of the surface of the solid phase serving as anchorage. In the case of nickel-plated slides, the long-range order results from the orientation of the magnetic domains by a magnetic field. When the anchorage consists of long-chain fatty acids, the layers transferred from a Langmuir trough exhibit a two-dimensional crystalline structure, that is a long-range order. Biologists so far have not considered the interaction of an ordered assembly of molecules reacting as a whole. It is not excluded that in vivo long-range interaction might arise from cooperation between molecules condensed with long-range order in view of the prevalence of interface at many levels of organization.

ACKNOWLEDGMENT The author is greatly indebted to Dr. George Hirst, whose numerous and wise suggestions contributed to a better presentation of the experiments summarized in this article.

REFERENCES I. 2. 3. 4. 5. 6.

7.

8. 9. 10.

II. 12. 13. 14. 15. 16. 17. 18. 19.

Blodgett, K . B. (1934). J. Am. Chem. Soc. 56, 495. Blodgett, K . B. (1934). J. Opt. Soc. Am. 24, 313. Blodgett, K . B., and Langmuir, 1. (1937). Phvs. Rev. 51, 964. Rothen, A,, and Landsteiner, K. (1942). J. Exp. Med. 76, 437. Drude, P. (1889). Ann. Phvs. 36, 865. Drude, P. (1890). Ann. Phys. 39, 481. Tronstad, L. (1935). Trans. Furaduy Soc. 31, 1151. Rothen, A. (1945). Rev. Sci. Inst. 16, 26. Winterbottom, A. B. (1955). Norske Videnskubus Selskub 41, 1. Asham, R. M., and Bashara, N. M. (1977). “Ellipsometry and Polarized Light.” NorthHolland F’ubl., Amsterdam. Singer, S. J. (1950). J . B i d . Chem. 182, 189. Trurnit, H. J. (1950). Science 112, 329. McGavin. S . . and Iball, 1. (1953). Trans Furuduv Soc. 49, 984. London, F. (1943). “Surface Chemistry,” monograph No. 21, p. 141. AAAS. Coulson. C. A.. and Davies, P. L. Personal communication. Dzyaloshipskii, I. E.. Lifshitz, E. M., and Pitaevskii, L. P. (1956). Adv. Phvs. 10, 295. Derjaguin, D. V., Abrikossova, 1. I . , and Lifshitz, E. M. (1956). Quart. Rev. 10, 295. Ninham, B. W.. and Parsegian, V. A. (1970). Biophys. J. 10, 646. Rothen, A. (1974). In “Progress in Surface and Membrane Science,” Vol. 8, pp. 81-118. Academic Press. New York.

304 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47.

ALEXANDRE ROTHEN

Rothen, A. (1957). Rev. Sci. Insr. 28, 283. Rothen, A. (1947). J. Biol. Chem. 168, 75. Rothen, A. (1950). Helv. Chim. Acta 33, 834. Rothen, A. (1948). J. Biol. Chem. 172, 841. Rothen, A., and Mathot, C. (1971). Helv. Chim.Acra 54, 1208. Palmer, L. S., Cunliffe, A., and Hough, J. M. (1962). Narure (London) 170, 796. Cope, F. W. (1977). Physiol. Chem. Phys. 9, 155. Winter, J. (1954). Ann. Insr. Henri Poincari 14, 1. Mathot, C., Rothen, A., and Casals, J. (1964). Narure (London) 202, 1181. Rothen, A,, Mathot, C., Kammlott, G. W., and Shay, M. (1971). Physiol. Chem. Phys. 3,66. Rothen, A. (1975). Physiol. Chem. Phys. 7, 133. Rothen, A., and Kincaid, M. (1974). Physiol. Chem. Phys. 6 , 417. Rothen, A. (1973). Physiol. Chem. Phys. 5 , 243. Rothen, A., and Mathot, C. (1969). Immunochemisrry 6, 241. Williams, J . , Stechschulter, D., and Salum, E. (1968). Am. J . Trop. Med. Hyg. 17, 568. Rothen, A., Mathot, C., and Thiele, E. (1969). Experienria 25, 420. Mathot, C., Rothen, A., and Scher, S. (1965). Nature (London) 207, 1263. Mathot, C., D’Alessandro. P. A., Scher, S., and Rothen, A. (1967). Am. J . Trop. Med. Hyg. 16, 443. Mathot, C., and Rothen, A. (1968). Am. J. Trop. Med. Hyg. 17, 756. Deborin, G. A., Tyurina, I. P., Torkhovskaya, T. I., and Oparin, A. 1. (1966). Fed. Proc. 25 (translation supplement) T863. Rothen, A. (1962). J. Coll. Chem. 17, 124. Semka, T. J. (1978). Physiol. Chem. Phys. 10, 89. Rothen, A. (1959). J . Phys. Chem. 63, 1929. Rothen, A. (1964). Biochim. Biophys. Acfa 88, 606. Rothen, A. (1948). J. Am. Chem. SOC. 70, 2732. Langmuir, I.. and Schaefer, V. J. (1937). J . Am. Chem. SOC. 59, 2400. Rothen, A. (1979). Physiol. Chem. Phys. 11, 481. Rothen, A. (1960). Helv. Chim. Acra 43, 1873.

Index

A Activation-inactivation, liquid-solid interface and, 278-279 Adenoviruses, cellular receptors and, 50 Age, dependency of viral permissiveness and, 48 Agrobacterium genetic engineering and, significance of crown gall induction to, 83-84 Algae microbodies in, 216-224 mitochondria1 association in, 151-152 plastid development in, 134- 136 storage reserves and mobilization during, 145-146 semicrystalline bodies, 144 semicrystalline bodies in, 144 Angiosperms greening plastid development in, 136-137 storage and reserve mobilization and, 149 grown in intermittent light, plastid development in, 137 light-grown plastid development in, 137- 139 storage and reserve mobilization and, 149 Antibodies absorbed, location in presence of membrane barrier, 289-290 absorption on antigen-coated slides, 273 to plant proteins, preparation for immunofluorescence studies, 232-235 Antigen, deposition at liquid-solid interface, 271-272

305

Antigen-antibody reaction dilution of antigens and, 283-286 dilution of antiserum and, 279-283 at liquid-solid interface, 270-271 rates of reaction, 273-277 ATP, formation in developing plastids, 169-17 I Autoimmunity, cell-specific, viral infection and, 55-56

B Bacteria, crown gall and, 65 Breath figure test, liquid-solid interface reactions and. 294-297

C Cells, synchronized, DNA replication fork movement rates in, 15-18 Chloroplast development biogenesis of photochemical activities appearance of coupling during greening, 172 ATP formation and, 169-171 biosynthetic requirements and, 172-173 development of photosystems and, 171-172 in different systems algae, 134-136 angiosperms in intermittent light, 137 greening angiosperms, 136- 137 light-grown angiosperms, 137-139 lower plants and gymnosperms, 136

306

INDEX

influence of light and hormones hormonal effects, 178- 179 light effects on transport and bioenergetic parameters, 176- I78 photo-control of, 174- 175 respiratory enhancement and reserve mobilization by light, 175-176 mitochondria and respiration during algal respiratory participation, I5 1-152 dark respiration after maturity, 155- I56 higher plant mitochondrial associations, 152-153 mitochondria1 association in algae, 150-151 oxygen consumption during light-grown development, 154- I55 plastid-mitochondria1 association, 149-150 respiratory participation during greening, 153-154 semicrystalline structures prolamellar bodies of etioplasts, 139- 144 prothylakoid bodies in light-grown tissue. 144- I45 semicrystalline bodies in algae, 144 storage reserves and mobilization during algae, 145-146 dark-grown and greening angiosperms, 149 light-grown angiosperms, 146- 148 transfer between cell compartments during photomorphogenesis changes in translocations during greening, 160-164 evidence for changes in fluxes between compartments, 158- 160 nucleotide pool sizes during development, 164-166 overview of changes in flux, 166-169 plastid envelope during greening, 156-158 Chromosomes, rye and wheat genetic relationships between, 94-99 molecular structure, 99- 109 Compartments, evidence for changes in fluxes between, 158-160 Coronaviruses, cellular receptors and, 52 Coupling, development during greening, 172 Crown gall bacteria and, 65 as disease, 64-65

evolutionary origin of, 84-86 induction, involvement of wounding in, 66-68 involvement of plasmids in induction, 70-7 1 nononcogenic functions of Ti-plasmid, 78-79 oncogenic-related functions not on TDNA, 76-78 presence and structure of Ti-plasmid DNA in gall cell, 71-74 products of T-DNA, 74-76 reversion to normal state, 82-83 Ti-DNA transfer and integration into plant cell, 79-82 opines and, 68-70 physiology of, 65-66 significance of induction to Agrobacteriurn, 83-84

D DNA, repeated sequence, polymorphism in regions of chromosomes of wheat-rye alloploids, 113-121 DNA replication fork, movement rates chemicals affecting, 18-21 in human cells, 9-15 methodologies, 2-9 in mutants and in v i m , 21-22 in nonhuman mammalian cells, 15 in synchronized cells, 15- I8

E Envelope, plastid, greening and, 156-158 Enzymes, plant, localization by immunofluoresence, 246-249 Etioplasts, prolamellar bodies of, 139- 144

F Fatty acid layers, bovine albumin layers anchored to, 294 Fluorescence microscopy, of plant tissue, 242-246 Formvar membranes, protective, formation and porosity of, 272-273 Fungi, microbodies in, 216-224

307

INDEX

Membranes components which interact with viruses Genes, coding for microbody proteins, exage dependency of viral permissiveness, pression of, 202-204 48 Genetic engineering, Agrobacterium and. definition of cellular response units by 86-87 monoclonal antibodies, 39-40 Genetics, control of cell viral receptor site density and affinity of viral receptor, expression and, 48-49 37-38 Gymnosperms, plastid development in, 136 genetic control of cell receptor site expression, 48-49 relationship of receptors to other surface H components, 38-39 Heterochromatin specificity of cellular receptor sites, biological effects of rye chromosomes in 40-45 wheat-rye alloploids, 121-123 in virro manipulation of cell receptor rye, possible origins of polymorphism in, sites, 45-47 123-127 interactions through, 287-289 Hormones, plastid development and, 178- 179 Microbodies Human cells, DNA replication fork movement proteins rates in, 9-15 expression of genes coding for, 202-204 poly(A+)-mRNA coding for, 210-21 I site of biosynthesis, 204-206 I synthesis on free polysornes, 21 I Immunoelectroadsorption, liquid-solid interface transient forms of, 201-202 and, 277-278 Mitochondria, association in higher plants, applications of, 286-287 150-151 influences of carrier on, 286 Monoclonal antibodies, definition of cellular Immunofluorescence studies, on plant cells viral receptor units by, 39-40 enzyme localization and, 246-249 Mutants, DNA replication fork movement rates lectin localization and, 249-250 in, 21-22 localization of other proteins, 260-261 Myxoviruses, cellular receptors and, 5 1-52 Immunologic reactions, liquid-solid interface and, 269-270 N

G

L Lectins, plant, localization by immunofluorescence, 249-260 Light chloroplast development and, 174- 175 respiratory enhancement and reserve mobilization by, 175- 176 Lower plants, plastid development in, 136 Liquid-solid interface, reactions occurring at experimental, 270-302 immunologic reactions, 269-270

Nucleotides, pool sizes during greening, 164- I66

0

Opines, crown gall and,68-70 Organelle biosynthesis general concepts, 194- 195 import from cytosol, 197-199 pathway via ER and subsequent segregation of vesicles, 195-197 de novo synthesis of fission and fusion, 199-20 1 transient forms of microbodies, 201-202 M single steps of assembly studied in vitro chemical modification and changes in hyMammalian cells, nonhuman, DNA replication drophobicity and aggregation, 21 3-214 fork movement rates in. 15

308

INDEX

import of proteins into microbodies, 2 12-21 3 poly(A +)-mRNA coding for microbody proteins, 210-21 1 products of in vitro translations, 211-212 synthesis on free polysomes, 21 I in special types of cells animal cells, 214-216 differentiated microbodies in plant cells, fungi and algae, 216-224 in vivo studies expression of genes coding for microbody proteins, 202-204 modification and aggregation, 208-2 I0 participation of ER, 206-208 site of biosynthesis of microbody proteins, 204-206 Oxygen, consumption during greening, 154- I55

P Paramyxoviruses, cellular receptors and, 51-52 Pattern transfer, liquid-solid interface reactions and, 297-299 Photosystems, development of, 171- 172 Picornaviruses, cellular receptors and, 49-50 Plant cells applications of immunofluorescence microscopy enzyme localization, 246-249 lectin localization, 249-260 localization of other proteins, 260-261 differentiated microbodies in. 216-224 techniques for immunofluorescence studies fluorescence microscopy, 242-246 preparation of antibody, 232-235 preparation of tissues, 235-242 Plasmids, involvement in induction of crown gall, 70-71 nononcogenic functions on Ti-plasmid, 78-79 oncogenic-related functions not on T-DNA, 76-78 presence and structure of Ti-plasmid DNA in gall cell, 71-74 products of T-DNA, 74-76 reversion to normal state, 82-83

Ti-DNA transfer and integration into plant cell, 79-82 Plastid-mitochondria1 association, chloroplast development and, 149-150 Pol ymorphisms in regions of wheat-rye alloploid chromosomes containing repeated sequence DNA, 113-121 Polymorphism, in rye heterochromatin, possible origins of, 123-127 F’rothylakoid bodies, in light-grown tissue, 144-145 Protein microbody expression of genes coding for, 202-204 poly(A+ )-mRNA coding for, 210-21 1 site of biosynthesis, 204-206 synthesis on free polysomes, 21 I plant, localization by immunofluorescence, 260-26 1 R Reoviruses, cellular receptors and, 50-51 Respiration dark, after maturity, 155-156 during greening, 153- 154 Retroviruses, cellular receptors and, 52 Rye, heterochromatin, possible origins of polymorphism in, 123-127 Rye chromosomes biological effects in wheat-rye alloploids, 121-123 wheat chromosomes and genetic relationships, 94-99 molecular structure, 99-109 S Slides, nickel-plated, liquid-solid interface and, 278 Strip technique, liquid-solid interface reactions and, 290-291 T T-DNA oncogenic-related functions not on, 76-78 products of, 74-76

309

INDEX Ti-plasmid nononcogenic functions of, 78-79 presence and structure of DNA in gall cell, 71-74 transfer and integration of DNA in plant cell, 79-82 Translocations changes during greening, 160- 164 in wheat-rye addition or substitution lines, 109-1 I3 Transport, light and, 176- I78 Trypsin factors influencing rate of hydrolysis of protein, hydrolysis in presence of protective membrane, 301-302 interaction with p o l y - ~and poly-L-lysine, 299-300 Tryptic action, at liquid-solid interface, 270-27 1, 29 1-294

v Viral attachment to cells, biological characteristics mathematical analysis of binding, 35-37 techniques used for study, 30-35 Viral receptor sites, definition of, 27-30 Virus(es) components which recognize cellular receptors adenoviruses, 50 coronaviruses, 52 myxoviruses and paramyxoviruses, 5 1-52 picornaviruses, 49-50 reoviruses, 50-5 1 retroviruses, 52

membrane components interacting with age dependency of viral permissiveness, 48 definition of cellular response units by monoclonal antibodies, 39-40 density and affinity of viral receptor, 37-38 genetic control of cell receptor site expression, 48-49. relationship of receptors to other surface components, 38-39 specificity of cellular receptor sites, 40-45 in vitro manipulation of cell receptor sites, 45-47 receptor interactions and pathogenicity, 52-53 induction of cell-specific autoimmunity, 55-56 role of cell surface receptors, 53-54 role of virus attachment proteins, 54-55

W Wheat chromosomes, rye chromosomes and genetic relationships, 94-99 molecular structure, 99- 109 Wheat-rye alloploids addition or substitution lines, translocations in, 109-113 biological effects of rye chromosomes in, 121-123 polymorphisms in regions of chromosomes containing repeated sequence DNA, 113-121 Wounding, crown gall induction and, 66-68

This Page Intentionally Left Blank

Contents of Recent Volumes and Supplements Volume 60

The Effects of Chemicals and Radiations within the Cell: An Ultrastructural and Micrurgical Study Using Amoeba proteus as a Single-Cell Model-M. J. ORD Growth, Reproduction, and Differentiation in Acanthamoeba -THOMAS J. BYERS

Transfer RNA-like Structure in Viral GenomeS-TIMOTHY c. HALL Cytoplasmic and Cell Surface Deoxyribonucleic Acids with Consideraton of Their Origin-BEVAN L. REID A N D ALEXAN- SUBJECT INDEX DER J. CHARLSON Biochemistry of the Mitotic SpindleVolume 62 CHRISTIAN PETZELT Alternatives to Classical Mitosis in HemoCalcification in Plants- ALLAN PENTECOST p i e t i c Tissues Of Velt€2br&S-vIBEKE E. ENGELEBERT Cellular Microinjection by Cell Fusion: Fluidity of Cell Membranes-Current ConTechnique and Applications in Biology and Medicine-MITsuRu FURUSAWA cepts and Trends-M. SHINITZKY AND Cytology, Physiology, and Biochemistry of P. HENKART G e r m i n a t i o n of F e r n Spore s-V . Macrophage-Lymphocyte Interactions in h"mne Induction-MARC FELDMANN, RAGHAVAN ALANROSENTHAL, AND PETERERE Immunocytochemical Localization of the Vertebrate Cyclic Nonapeptide NeurohyImmunohistochemistry of Luteinizing Horpophyseal Hormones and Neurophymone-Releasing Hormone-Producing sins-K. DIERICKX Neurons of the Vertebrates-JuLIEN Recent Progress in the Morphology, HistoBARRY chemistry, Biochemistry, and Physiology Cell Reparation of Non-DNA Injury-V. of Developing and Maturing Mammalian YA.ALEXANDROV Testis-SARDUL s. GURAYA Ultrastructure of the Carotid Body in the Transitional Cells of Hemopoietic Tissues: M ~ I I U ~ ~ ~ S - A LVERNA AIN Origin, Structure., and Development PoThe Cytology and Cytochemistry of the tential-JOSEPH M. YOFFEY WOO1 FoUicle-DONALD F. G. ORWIN Human Chromosomal Heteromorphisms: SUBJECT INDEX Nature and Clinical Significance-RAM S. VERMAAND HARVEY DOSIK Volume 61 SUBJECT INDEX The Association of DNA and RNA with Volume 63 Membranes-MARY PAT MOYER Electron Cytochemical Stains Based on Metal Chelation-DAVID E. ALLEN AND Physarum polycephalum: A Review of a Model System Using a Structure-FuncDOUGLAS D. PERRIN tion Approach-EUGENE M. GOODMAN Cell Electrophoresis-THOMAS G. PRETMicrotubules in Cultured Cells: Indirect ImLOW, I1 A N D THERESA P. PRETLOW munofluorescent Staining with Tubulin The Wall of the Growing Plant Cell: Its Antibody-B. BRINKLEY, S. FISTEL,J. Three-Dimensional Organization-JEANM. MARCUM, A N D R. L. PARDUE AND BRIGITTE VlAN CLAUDE ROLAND Biochemistry and Metabolism of Basement Septate and Scalariform Junctions in ArNOIROT-TIMOTHEE Membranes -NICHOLAS A. KEFALIDES, thropods-CEcILE NOIROT ROBERTALPER,A N D CHARLES C. CLARK A N D CHARLES 311

312

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS

The Cytology of Salivary Glands-CARLIN A. PINKSTAFF Development of the Vertebrate ComeaELIZABETH D. HAY Scanning Electron Microscopy of the Primate S~~IIII-KENNETH G. GOULD Cortical Granules of Mammalian EggsBELAJ. GULYAS

Cytological Mechanisms of Calcium Carbonate Excavation by Boring SpongesA. POMPONI SHIRLEY Neuromuscular Disorders with Abnormal Muscle Mitochondria-Z. KAMIENIECKA A N D H. SCHMALBRUCH SUBJECT INDEX

SUBJECT INDEX

Volume 66

Dynein: The Mechanochemical Coupling Adenosine Triphosphatase of Microtubule-Based Sliding Filament MechaVariant Mitoses in Lower Eukaryotes: Indinisms-FRED D. WARNER A N D DAVIDR. cators of the Evolution of Mitosis-I. MITCHELL BRENTHEATH The Centriolar Complex-SCOTT P. PETER- Structure and Function of Phycobilisomes: Light Harvesting Pigment Complexes in SON A N D MICHAEL W. BERNS Red and Blue-Green Algae-ELISABETH The Structural Organization of Mammalian GANTT Retinal Disc Membrane-J. OLIVE The Roles of Transport and Phosphorylation Structural Correlates of Gap Junction Permeation-CAMILLo PERACCHIA in Nutrient Uptake in Cultural Animal CellS-ROBERT M. WOHLHUETER AND The Kinetics and Metabolism of the Cells of Hibernating Animals during HibernaPETERG. W. F’LAGEMANN tion-S. G . K O L A E V AL, . I. K R A The Contractile Apparatus of Smooth Muscle-J. VICTORSMALLA N D APOLINARY MAROVA, E. N. ILYASOVA,AND F. E. ILYASOV SOBIESZEK Cytophysiology of the Adrenal Zona Glo- CELLSIM: Cell Cycle Simulation Made Easy-CHARLES E. DONAGHEY meNlOSa-GASTONE G . NUSSDOWER The Formation of Axonal Sprouts in Organ SUBJECT INDEX Culture and Their Relationship to Sprouting in Vivo-I. R. DUCEAND P. KEEN Volume 65 When Sperm Meets Egg: Biochemical Mechanisms of Gamete InteractionBENNETTM. SHAPIRO AND E. M. EDDY Cell Surface Glycosyltransferase ActivitieS-MICHAEL PIERCE,EVAA. TURLEY, Perisinusoided Stellate Cells (Fat-Storing AND STEPHEN ROTH Cells, Interstitial Cells, Lipocytes), Their The Transport of Steroid Hormones into Related Structure in and around the Liver h h d Cells-ELEONORA P. GIORGI Sinusoids, and Vitamin A-Storing Cells in Structural Aspects of Brain Barriers, with Extrahepatic Organs-KENJIRO WAKE Special Reference to the Permeability of SUBJECT INDEX the Cerebral Endothelium and Choroidal Epithelium-B. VAN DEURS Immunochemistry of Cytoplasmic ContractVolume 67 ile ProkinS-uTE GROSCHEL-STEWART The Ultrastructural Visualization of NucleoMembrane Circulation in Neurons and Pholar and Extranucleolar RNA Synthesis toreceptors: Some Unresolved Issuesand Distribution-S. FAKAN AND E. PuERICHOLTZMAN AND ARTHURM. MERVolume 64

VION

CURIO

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS

313

Ultrastructure of Invertebrate Chemo-, Volume 69 Thermo-, and Hygroreceptors and Its Functional SignifiCanCe-HELMUT ALT- The Structures and Functions of the MycoNER A N D LINDEPRILLINGER plasma Membrane-D. B. ARCHER Calcium Transport System: A Comparative Metabolic Cooperation between Cells-M. Study in Different CellS-ANNE GODL. HOOPERAND J. H. SUBAK-SHARPE FRAIND-DE BECKERA N D T HEOP HI L E The Kinetoplast as a Cell Organelle-V. D. GODFRAIND KALLINIKOVA The Ultrastructure of Skeletogenesis in Her- Chloroplast DNA Replication in Chlamymatypic Corals-IAN S. JOHNSTON domonas reinhardtii-STEPHEN JAY Protein Turnover in Muscle Cells as VisualKELLERAND CHINGHo ized by Autoradiography-J. P. DA- Nucleus-Associated Organelles in FungiDOUNE I. BRENTHEATH Identified Serotonin Neurons-NEVILLE Regulation of the Cell Cycle in Eukaryotic N. OSBORNE A N D VOLKER NEUHOFF Cells-ROSALIND M. YANISHEVSKY A N D Nuclear Proteins in Programming Cell CyGRETCHEN H. STEIN cles-M. V. NARASIMHA RAO The Relationship of in Vitro Studies to in SUBJECT INDEX V i v o H u m a n Aging-EDWARD L. SCHNIEDER A N D JAMESR. SMITH Cell Replacement in Epidermis (Keratopoiesis) via Discrete Units of Proliferation-C. S. POTTEN INDEX

Volume 68 Moisture Content as a Controlling Factor in Seed Development and Germination-C. A. ADAMSA N D R. W. RINNE Applications of Protoplasts to the Study of Plant CellS-LARRY c. FOWKEA N D OLUFL. GAMBORG Control of Membrane Morphogenesis in J. BREWER Bacteriophage -GREGORY Scanning Electron Microscopy of IntracelIUlW Structures-KEIIcHI TANAKA The Relevance of the State of Growth and Transformation of Cells to Their Patterns of Metabolite Uptake-RUTH KOREN Intracellular Source of BioluminescenceM. SWEENEY BEATRICE Differentiation of MSH-, ACTH-, Endorphin-, and LPH-Containing Cells in the Hypophysis during Embryonic and Fetal DeVelOpment-JEAN-PAUL DU-

Volume 70

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

Cycling zz Noncycling Cell Transitions in Tissue Aging, Immunological Surveillance, Transformation, and Tumor Growth-SEYMOUR GELFANT The Differentiated State of Normal and Malignant Cells or How to Define a "Normal" Cell h CultUre-MlNA J. BISSELL On the Nature of Oncogenic Transformation Of CellS-GERALD L. CHAN Morphological and Biochemical Aspects of Adhesiveness and Dissociation of Cancer Cells-HIDE0 HAYASHIA N D YASUJI ISHIMARU The Cells of the Gastric MUCOSa-HERBERT F. HELANDER Ultrastructure and Biology of Female Gametophyte in Flowering Plants-R. N. KAPIL A N D A. K. BHATNAGAR

INDEX

INDEX

POUY

3 14

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS

Polytene Chromosomes of Plants- WALTER NAGL Morphology, UltrastrucIntegration of Oncogenic Viruses in Mam- Endosperm-Its ture, and Histochemistry-S. P. BHATmalian Cells-CAmo M. CROCE NAGAR A N D VEENASAWHNEY Mitochondrial Genetics of Paramecium aurefia-G. H. BEALEAND A. TAIT The Role of Phosphorylated Dolichols in Membrane Glycoprotein Biosynthesis: Histone Gene Expression: Hybrid Cells and Relation to Cholesterol BiosynthesisOrganisms Establish Complex ControlsPHILIPHOHMANN JOAN TUGENDHAFT MILLS A N D ANGene Expression and Cell Cycle RegulaTHONY M. ADAMANY tiOn-STEVEN J. HOCHHAUSER, JANET Mechanisms of Intralysosomal Degradation L. STEIN,AND GARYS. STEIN with Special Reference to AutophagocyThe Diptera as a Model System in Cell and tosis and Heterophagocytosis of Cell OrMolecular Biology-ELENA c. ZEGAREL- ganelleS-HANS GLAUMANN, JANL. E. LI-SCHMIDT ERICSSON, AND REBAGOODMAN A N D LOUIS MARZELLA Comments on the Use of Laser Doppler Membrane Ultrastructure in Urinary TuTechniques in Cell Electrophoresis: Rebules-LELIo ORCI, FABIENNE HUMply to Pretlow and Pretlow’s ReviewBERT, DENNISBROWN, A N D ALAINPERJOELH. KAPLAN AND E. E. UZGIRIS RELET Comments on the Use of Laser Doppler Tight Junctions in Arthropod TissuesTechniques as Represented by Kaplan NANCYJ. LANE and Uzgiris: Reply t o Kaplan and Genetics and Aging in PrOtOZOa-JOAN U Z ~ ~ ~ ~ S - T H OG. M APRETLOW S I1 AND SMITH-SONNEBORN THERESA P. PRETLOW INDEX Volume 71

INDEX

Volume 72

Microtubule-Membrane Interactions in Cilia and Flagella-WILLIAM L. DENTLER The Chloroplast Endoplasmic Reticulum: Structure, Function, and Evolutionary Significance-SARAH P. GIBBS A N D P. DNA Repair-A. R. LEHMANN KARRAN Insulin Binding and Glucose TransportRUSSELLHILF, LAURIEK. SORGE,AND ROGERJ. GAY Cell Interactions and the Control of Development in Myxobacteria PopulationsDAVIDWHITE Ultrastructure, Chemistry, and Function of the Bacterial Wall-T. J. BEVERIDGE INDEX

Volume 73

h t o p h S t S Of Eukaryotic &%-MARTHA D. BERLINER

Volume 74

The Plasma Membrane as a Regulatory Site in Growth and Differentiation of Neuroblastoma Cells-SIEGFRIED w. DE LAAT AND PAULT. VAN DER SAAG Mechanisms That Regulate the Structural and Functional Architecture of Cell SurfaceS-JANET M. OLIVER AND RICHARD D. BERLIN Genome Activity and Gene Expression in Avian Erythroid Cells-KARLEN G. GASARYAN

Morphological and Cytological Aspects of Algal CdCificatiOn-MICHAEL A. BOROWITZKA

Naturally Occumng Neuron Death and Its Regulation by Developing Neural PathWays-TIMOTHY J. CUNNINGHAM The Brown Fat Cd-JAN NEDERGAARD AND OLOV LINDBERG INDEX

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS

Metabolism of Ethylene by Plants-JOHN DODDSA N D MICHAELA. HALL

Volume 75

Mitochondria1 Nuclei-TsuNEYosHr

3 15

Ku-

INDEX

ROIWA

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

Volume 76

Cytological Hybridization to Mammalian Chromosomes-ANN s. HENDERSON Organization and Expression of Viral Genes in Adenovirus-Transformed Cells-S. J. FLINT Highly Repeated Sequences in Mammalian Genomes-MAXIME F. SINGER Moderately Repetitive DNA in EvolutionROBERTA. BOUCHARD Structural Attributes of Membranous Organelles in BaCteria-cHARLES C. REM-

Volume 77

Calcium-Binding Proteins and the Molecular Basis of Calcium Action-LINDA J. VANELDIK, JOSEPHG. ZENDEGUI, DANIEL R. MARSHAK, A N D D. MARTIN WATTERSON

Genetic Predisposition to Cancer in Man: I n Vitro Studies-LEVY KOPELOVICH Membrane Flow via the Golgi Apparatus of Higher Plant CdS-DAVID G . ROBINSON A N D UDOKRISTEN Cell Membranes in Sponges-WERNER E. G . MULLER Plant Movements in the Space Environment-DAVlD G . HEATHCOTE Chloroplasts and Chloroplast DNA of Acerabularia mediterranea: Facts and Hyp o t h e s e s - A N G E L A L U T T K EA N D SILVANO BONOTTO Structure and Cytochemistry of the ChemiC d Synapses-STEPHEN MANALOV AND WLADIMIR OVTSCHAROFF INDEX

Volume 78

Bioenergetics and Kinetics of Microtubule and Actin Filament Assembly-Disassembly-TERRELL L. HILLA N D MARCw . KIRSCHNER Regulation of the Cell Cycle by SomatomeSEN dins-HOWARD ROTHSTEIN Separated Anterior Pituitary Cells and Their Epidermal Growth Factor: Mechanisms of Response to Hypophysiotropic Hormones-CARL DENEF,Luc SWENNEN, Action-M~~lusRlDAS Recent Progress in the Structure, Origin, A N D MARIAANDRIES Composition, and Function of Cortical What Is the Role of Naturally Produced Granules in Animal Egg-SARDUL S. Electric Current in Vertebrate RegeneraGURAYA tion a n d Healing?-RlcHARD B. INDEX BORCENS

316

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS

Vascular Smooth Muscle Cells for Studies of Cellular Aging in Vitro; an Examination of Changes in Structural Cell LipThe Formation, Structure, and Composition ids-OLGA 0.BLUMENFELD, ELAINE of the Mammalian Kinetochore and KineSCHWARTZ, VERONICA M. HEARN,A N D tochore Fiber-CONLY L. RIEDER MARIEJ. KRANEPOOL Motility during Fertilization-GERALD C h o n d r o c y t e s in Aging Re se a rc hSCHATTEN EDWARD J. MILLERAND STEFFENGAY Functional Organization in the NucleusRONALD HANCOCK AND TENIBOULIKAS Growth and Differentiation of Isolated C al v arium Ce lls in a S e r u m - F r e e The Relation of Programmed Cell Death to Medium-JAMES K. BURKSAND WILDevelopment and Reproduction: ComparLIAM A. PECK ative Studies and an Attempt at ClassifiStudies of Aging in Cultured Nervous Syscation-JACQUES BEAULATON AND RItem Tissue-DONALD H. SILBERBERG CHARD A. LOCKSHIN AND SEUNGU. KIM Cryofixation: A Tool in Biological UltraAging of Adrenocortical Cells in Culturestructural Research-HELMUT PLATTPETERJ. HORNSBY, MICHAEL H. SIMONINER AND LUISBACHMANN AN, AND GORDON N. GILL Stress Protein Formation: Gene Expression and Environmental Interaction with Evo- Thyroid Cells in Culture-FRANcEsco S. AMBESI-IMPIOMBATO A N D HAYDEN G. lutionary Significance-C. ADAMSA N D COON R. W. RINNE Permanent Teratocarcinoma-Derived Cell INDEX Lines Stabilized by Transformation with SV40 and SV40tsA Mutant VirusesWARRENMALTZMAN, DANIELI. H . LINZER,FLORENCE Supplement 10 Differentiated CeUs in BROWN,ANGELIKA Aging Research K. TERESKY, MAURICEROSENSTRAUS, AND ARNOLDJ. LEVINE Do Diploid Fibroblasts in Culture Age?Nonreplicating Cultures of Frog Gastric TuEUGENEBELL, LOUISMAREK,STEPHAbular Cells-GERTRUDE H. BLUMENN I E S H E R , CHARLOTTE M E R R I L L , THAL AND DINKAR K. KASBEKAR DONALDLEVINSTONE, AND IANYOUNG SUBJECT INDEX Urinary Tract Epithelial Cells Cultured from Human Urine-J. S. FELIXA N D J. W. LITTLEFIELD The Role of Terminal Differentiation in the Finite Culture Lifetime of the Human Epi- Supplement 11A Perspectives in Plant Cell d e r m a l Keratinocyte-JAMES G. and Tissue Culture RHEINWALD Cell Proliferation and Growth in Callus CulLong-Term Lymphoid Cell CulturesGEORGEF. SMITH,PARVINJUSTICE, tures-M. M. YEOMAN AND E. FORCHE HENRIFRISCHER, LEE KIN CHU, A N D Cell Proliferation and Growth in Suspension JAMESKROC Cultures-P. J. KING Type I1 Alveolar Pneumonocytes in VitroCytOd~erentiatiOn-RICHARD PHILLIPS WILLIAMH. J. DOUGLAS, JAMESA. Organogenesis in Vitro: Structural, PhysioMCATEER,JAMES R. SMITH,A N D WALlogical, and Biochemical AspectsTER R. BRAUNSCHWEIGER A. THORPE TREVOR Cultured Vascular Endothelial Cells as a Chromosomal Variation in Plant Tissues in Model System for the Study of Cellular Culture-M. W. BAYLISS Senescence-ELLIOT M. LEVINEA N D Clonal Propagation-INDm K. VASILA N D STEPHENM. MUELLER VIMLAVASIL Volume 79

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS Control of Morphogenesis by Inherent and Exogenously Applied Factors in Thin Cell Layers-K. TRANTHANHVAN Androgenetic Haploids-hDRA K. VASIL Isolation, Characterization, and Utilization of Mutant Cell Lines in Higher PlantsPALMALIGA

317

Membranes and Cell Movement: Interactions of Membranes with the Proteins of the Cytoskeleton-JAMES A. WEATHERBEE

Electrophysiology of Cells and Organelles: Studies with Optical Potentiometric Indicators-JEFFREY c. FREEDMAN AND SUBJECT INDEX PHILIPC. LARIS Synthesis and Assembly of Membrane and Organelle Proteins-HARVEY F. LODISH, W I L L I A MA . B R A E L L ,A L A N L . Supplement 1 1 B Perspectives in Plant Cell SCHWARTZ, GER J. A. M. STROUS, AND and Tissue Culture ASHERZILBERSTEIN The Importance of Adequate Fixation in Isolation and Culture of Protoplasts-INDRA K. VASILAND VIMLAVASIL Preservation of Membrane UltrastrucProtoplast Fusion and Somatic HybridizatUre-RONALD B. LUFTIGAND PAUL N. tion-O-rro SCHIEDERA N D INDRA K. MCMILLAN VASIL Liposomes-As Artificial Organelles, Topochemical Matrices, and Therapeutic Genetic Modification of Plant Cells Through Uptake of Foreign DNA-C. I. KADO Carrier SyStemS-PETER NICHOLLS Drug and Chemical Effects on Membrane A N D A. KLEINHOFS Nitrogen Fixation and Plant Tissue CulTransport-WILLIAM 0. BERNDT INDEX tUCe-KENNETH L. GILESA N D INDRA K. VASIL Preservation of Germplasm-LYNDsEY A. Supplement 13: Biology of the Rhimbiaceae WITHERS Intraovarian and in Vitro Pollination-M. ZENKTELER The Taxonomy of the RhizobiaceaeEndosperm Culture-B. M. JOHRI,P. S. GERALDH. ELKAN SRIVASTAVA, A N D A. P. RASTE Biology of Agrobacterium tumefaciens: The Formation of Secondary Metabolites in Plant Interactions-L. W. MOOREAND D. A. COOKSEY Plant Tissue and Cell Cultures-H. BOHM Agrobacterium tumefaciens in Agriculture Embryo Culture-V. RAGHAVAN and Research-FAwzi EL-FIKI A N D KENNETHL. GILES The Future-GEoRG MELCHERS SUBJECT INDEX Suppression of, and Recovery from, the Neoplastic State-ROBERT TURGEON Plasmid Studies in Crown Gall TumorigeneSiS-STEPHEN L. DELLAPORTA AND Supplement 12: Membrane Research: RICKL. PESANO Classic Origins and Current Concepts The Position of Agrobacterium rhizogeneS-JESSE M. JAYNESAND GARYA. Membrane Events Associated with the Generation of a Blastocyst-MARTIN H. STROBEL Recognition in Rhizobium --Legume SymJOHNSON Structural and Functional Evidence of bioses-TERRENCE L. GRAHAM The Rhizobium Bacteroid State-W. D. Cooperativity between Membranes and SUTTON,C. E. PANKHURST, AND A. S. Cell Wall in BaCteria-MANFRED E. BAYER CRAIG Plant Cell Surface Structure and Recogni- Exchange of Metabolites and Energy betion Phenomena with Reference to Symbitween Legume and Rhizobium -JOHN IMSANDE OSOS-PATRICIA s. RElSERT

318

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS

The Genetics of Rhizobium-ADAM Mutants ofRhizobium That Are Altered in KONDOROSI A N D ANDREW w. B. JOHN- Legume Interaction and Nitrogen FixaSTON tion-L. D. KUYKENDALL Indigenous Plasmids of Rhizobium-J. DE- The Significance and Application of RhizoENARIEE, P. BOISTARD,F R A N C I N E bium h AgTiCdture-HAROLD L. PETERCASSE-DELBART, A. G. ATHEIUY,J. 0. SON A N D THOMAS E. LOYNACHAN BERRY,A N D P. RUSSELL INDEX Nodules Morphogenesis and DifferentiatiOn-wILLIAM NEWCOMB

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

Recommend Documents