International Review of Cytology: A Survey of Cell Biology, Volume 79

INTERNATIONAL

REVIEW OF CYTOLOGY VOLUME 79

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 O L U F 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. P A N T I ~ 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 S t . George’s, Grenada West lndies

Worcester Polytechnic Institute Worcester, Massachusetts

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

VOLUME 79 1982

ACADEMIC PRESS

A Subsidiary of Harcourt Brace Jovanovich. Publishers

New York London Paris San Diego San Francisco S5o Paulo Sydney Tokyo Toronto

COPYRIGHT @ 1982,

UY

ACADEMIC PRESS,INC.

ALL RIGHTS RESERVED. N O PART O F THIS PUBLICATION MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM OR BY ANY MEANS, ELECTRONIC OR MECHANICAL, INCLUDING PIiOTOCOPY, 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

Utiited Kirrgdotu Editioti prtblished by ACADEMIC PRESS, INC. ( L O N D O N ) LTD. 24/28 Oval Road,

LIBRARY OF

London NWI 7DX

CONGRESS CATALOG CARD

NUMBER:52-5203

ISBN 0-1 2-364479-8 PRINTED IN THE UNITED STATES OF AMERlCA 82 83 84 x5

9 8 7 6 5 4 3 2 1

Contents LISTOF CONTRIRUTORS . . . . . . . . . . . . . . . . . . . . . . . . . . .

vii

The Formation. Structure. and Composition of the Mammalian Kinetochore and Kinetochore Fiber CONLY L . RIEDER Introduction . . . . . . . . . . . . . . . . . . . Kinetochore versus Centromere . . . . . . . . . Types of Kinetochores . . . . . . . . . . . . . . Mammalian Kinetochore Structure and Chemistry . V . The Kinetochore Organizer . . . . . . . . . . . . V1 . The Mammalian Kinetochore Fiber . . . . . . . . References . . . . . . . . . . . . . . . . . . . . I. II . III . IV .

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

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

1 2 6 8 27 29 53

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dynamics of Fertilization . . . . . . . . . . . . . . . . . . . . . . Detection of Cytoskeletal Elements during Fertilization . . . . . . . . . Effects of Motility Inhibitors . . . . . . . . . . . . . . . . . . . . . The Regulation of Motility at Fertilization . . . . . . . . . . . . . . . Consequences of Fertilization for Later Embryonic Development . . . . Motility during Fertilization and Its Regulation: A Model . . . . . . . . Prospectives and Conclusions . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

62 86 95 119 135 140 152 156

. . . . . . . . . .

Motility during Fertilization GERALD SCHATTEN I. I1 . III . I V. V. VI . VII . VIII .

60

Functional Organization in the Nucleus RONALDHANCOCK A N D TEN]BOULIKAS I. II. Ill . IV . V. VI .

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Nuclear Envelope . . . . . . . . . . . . . . . . . . . . . . . Chromatin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Topological Organization of Interphase DNA and Chromatin . . . . . . Transcription, Processing. and Packaging of RNA . . . . . . . . . . . Changes in Nuclear Organization during the Cell Cycle . . . . . . . . . V

165

166 174 180 188 198

vi

CONTENTS

VII . Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

206 206

The Relation of Programmed Cell Death to Development and Reproduction: Comparative Studies and an Attempt at Classification JACQUES BEAULATON A N D RICHARD A . LOCKSHIN I . Introduction . . . . . . . . . . . . . . . . . . . . I1. Modes of Cell Death . . . . . . . . . . . . . . . . 111. Means of Elimination of Degenerate Cells . . . . . . IV . Determination of Degeneration by Cellular Interaction References . . . . . . . . . . . . . . . . . . . . .

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

. . . . . . . .

215 216 223 227 231

Cryofixation: A Tool in Biological Ultrastructural Research HELMUTPLATTNER A N D Luis BACHMANN I. I1. 111. IV . V. VI . VII . VIII . IX . X. XI . XI1.

Scope and Goal of Cryofixation . . . . . . . . . . . Basic Methodical Aspects . . . . . . . . . . . . . . . Experimental Consequences . . . . . . . . . . . . . . Cryofixation Techniques-Advantages and Restrictions . Cooling Rates . . . . . . . . . . . . . . . . . . . . “Resolution” of Cryofixation . . . . . . . . . . . . . Ultrastructural Side-effects of Chemical Pretreatments . Tests for the Quality of Cryofixation . . . . . . . . . Combination of Cryofixation with Other Techniques . . Suspensions. Emulsions. and Solutions . . . . . . . . Nonaqueous Systems . . . . . . . . . . . . . . . . . Concluding Remarks . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . .

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

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

. . . . .

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

237 240 243 247 271 273 274 278 280 290 292 293 293

Stress Protein Formation: Gene Expression and Environmental Interaction with Evolutionary Significance C . ADAMSA N D R . W . R I N N E

I. I1. 111. IV . V.

Introduction . . . . . . . . . . . . . . . . . . . Response to Environmental Stresses . . . . . . . Molecular Events in Stress Protein Formation . . . Function of Stress Proteins . . . . . . . . . . . . Evolutionary Significance of Stress Proteins . . . . References . . . . . . . . . . . . . . . . . . . .

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

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

305 306 310 311 312 314

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

317 321

. . . . . . . . .

List of Contributors

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

C. ADAMS(305).U.S. Department of Agriculture, Agricultural Research Service, and Department of Agronomy, University of Illinois, Urbana, Illinois 61801 L U I SBACHMANN (237), Institute of Technical Chemistry, Technical University of Miinchen, Garching, Federal Republic of Germany JACQUES BEAULATON (215), Department de Zoologie, Universite de Clermont, Aubiere, France

TENIBOULIKAS ( 169, Swiss Institute for Experimental Cancer Research, EpalingeslLausanne, Switzerland RONALDHANCOCK (163, Swiss Institute for Experimental Cancer Research, EpalingeslLausanne, Switzerland RICHARD A. LOCKSHIN (215), Department of Biological Sciences, St. John’s University, Jamaica, New York 11439 HELMUTPLATTNER (237), Faculty of Biology, University of Konstanz, Konstanz, Federal Republic of Germany CONLYL. RIEDER ( I ) , New York State Department of Health, Division of Laboratories and Research, Albany, New York 12201 R. W. R I N N E(305), U.S. Department of Agriculture, Agricultural Research Service, and Department of Agronomy, University of Illinois, Urbuna, tllinois 61801 GERALDSCHATTEN (59), Department of Biological Science, Florida State University, Tallahassee, Florida 32306

vii

This Page Intentionally Left Blank

INTERNATIONAL REVIEW OF CYTOI.OGY, VOL. 79

The Formation, Structure, and Composition of the Mammalian Kinetochore and Kinetochore Fiber CONLYL. RIEDER New York Stute Department oj Health Division .fLaborntories und Research, Albany, New York I . Introduction.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I

11. Kinetochore versus Centromere . . . . . . . . . . . . . . . . . . . 111. Types of Kinetochores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

6

IV. Mammalian Kinetochore Structure and Chemistry B . Prophase . . . . . .

D. Metaphase . . . .

........

12

..........

22 27 29 30 31 39 53

V. The Kinetochore Organizer.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI. The Mammalian Kinetochore Fiber . . . . . . . . . . . . . . . . . . . . . . . . . .

A. General Properties during Metaphase and Early Anaphase . , . . B . Structure and Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Kinetochore Microtubule Formation . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I. Introduction The initiation of chromosome movement during cell division can be correlated with the formation of a fiber, composed primarily of microtubules (MTs) and associated proteins which connects each chromosome to the polar area of the spindle (e.g., Begg and Ellis, 1979a,b). The region on the chromosome where the MTs attach is referred to as the kinetochore and the MTs themselves are known as kinetochore MTs (K-MTs). In the past ultrastructural studies have provided us with a wealth of information with regards to the structure of kinetochores from a variety of organisms (see reviews of Luykx, 1970; Bajer and Mole-Bajer, 1972; Kubai, 1975; Fuge, 1977; Heath, 1979). However, until recently little was known about the formation and chemistry of this organelle and the mechanism by which it functions. The greatest single factor in slowing progress in this area has been the lack of success in obtaining bulk isolates of kinetochores for in virro analysis. Recently, alternate approaches, including the use of enzymes, immunoelectron, and electron microscopic cytochemical staining procedures, have been developed and have proven useful for dissecting the macromolecular organization of this organelle. Similar-

Copyright IG 1982 by Acadcniic Press. Inc. All nghh 01 reproduction in any lurm reserved ISBN 0-12-364479-8

2

CONLY L. RIEDER

ly, the perfection and routine use of correlative light and electron microscopic methods, of lysed cell systems augmented with polymerization competent MT protein (tubulin), and of potent but reversible inhibitors of MT assembly have also advanced our understanding concerning the composition and formation of the kinetochore fiber (K-Fiber). The purpose of this article is to summarize recent advances in understanding the formation, structure, and chemistry of the mammalian kinetochore and its associated fiber. Its namow emphasis reflects an almost complete lack of knowledge concerning the formation and chemistry of kinetochores in other types of organisms.

11. Kinetochore versus Centromere The terms “centromere” and “kinetochore” were coined by early light microscopists as synonyms to denote that region on the chromosome which becomes attached to the spindle during mitosis and meiosis (see reviews of Schrader, 1953; Ris and Witt, 1981). In most organisms this region corresponds to a narrow constriction on the chromosome known as the primary constriction (Fig. I ) . The primary constriction frequently contains a chromatin component, situated peripheral to the kinetochore, which fails to decondense after cell division. This “constitutive heterochromatin” (Brown, 1966) is similarly more resistant to treatments which disperse the remainder of the chromosome (e.g., Rattner et al., 1975, 1978; Roos, 1977; Brinkley et al., 1980; Ris and Witt, 1981). It can be detected with the light microscope (LM) after various staining procedures (see reviews of Commings et al., 1973; Commings, 1978). The DNA within this “pericentromeric” heterochromatin is generally replicated later in S phase than the bulk of the remaining chromosomal DNA and contains, in some organisms (e.g., calf, mouse, humans, flies), tandem arrays of highly repeated nucleotide sequences (i.e., satellite DNA-see Rae, 1972; John and Miklos, 1979). The heterochromatin of the primary constriction is appropriately referred to as pericentromeric or procentromeric due to its location peripheral to the kinetochore (centromere). It should be stressed that its exact relationship with the kinetochore as well as its function remain unknown. Some have suggested that it is involved in (1) chromatid adhesiveness until the onset of anaphase (Stubblefield, 1973; Vig, 198l ) , (2) synapsis and recombination during meiosis (Miklos and John, 1979), (3) karyotype evolution via translocations of the Robertsonian type (references in Yunis and Yasmineh, 1972), or (4) protecting the kinetochore from evolutionary changes (Yunis and Yasmineh, 1972). The proximity of the pericentromeric heterochromatin to the kinetochore has even led some investigators to speculate that the (mammalian) kinetochore is formed! in

THE MAMMALIAN KINETOCHORE

3

FIG. 1. Mouse L929 chromosomes isolated from a colcemid-arrested mitotic cell by detergent lysis. The kinetochores (K)appear as electron opaque plaques on the surface of the pericentromeric heterochromatin (H). See text for details. X 12,800. (Courtesy of J . B. Rattner.) FIG. 2. Human metaphase chromosomes stained by a Giemsa technique (C, staining) which reveals two identical dots, one on each side of the centromere. See text for details. (Courtesy of H. Eiberg by permission of Nature (London), 1974.)

part, from DNA continuous with the heterochromatin (e.g., Brinkley and Stubblefield, 1970; Yunis and Yasmineh, 1972; Hennig, 1973; Pepper and Brinkley, 1980). This is not an unreasonable assertion in light of the recent evidence suggesting the presence of DNA in the kinetochore (see Section IV). Yet heterochromatin (and satellite DNA) has not been detected in and may be absent from the primary constriction of a number of plants and animals including some

4

CONLY L. RIEDER

mammals (see Table Vl in John and Miklos, 1979). Thus the heterochromatized DNA of the pericentromeric heterochromatin should not be considered as a structural component of the kinetochore (e.g., Nicklas, 1971; Brenner et al., 1980): it does not appear to be present in many organisms and when it is present it is located subjacent to the region which binds spindle MTs. Over the past 25 years many electron microscopists have examined the area on the chromosome which becomes associated with spindle MTs. They often found distinct structures (e.g., single or multilayered disks) which they specifically referred to as kinetochores. However, such structural differentiations are not found in all stages of mitosis ( e . g . ,prophase). Furthermore, the K-MTs in some organisms appear to terminate directly on chromatin (i.e., a discrete structure cannot be distinguished, after conventional fixation and staining, from the chromosome-see Section 111; also Ris and Witt, 1981). The exact definition of the term “kinetochore” has therefore been ambiguous: should it be used with reference to a specific structure, a region on the chromosome, or both? To eliminate this confusion Ris and Witt (1981) suggested that “kinetochore be used in its original meaning, synonymous with centromere as the region on the chromosome that becomes attached to the spindle” (italics mine). Implicit in this definition is the notion that the kinetochore is a constant segment of the chromosome which (in mammals) differentiates into a well-defined structure during mitosis. It is certainly advantageous to use the term kinetochore with reference to a region on the chromosome which may or may not contain an obvious structural differentiation. Nevertheless, the notion that this region (the kinetochore) should continue to be used synonymously with “centromere” can, in itself, lead to confusion. In the current literature ‘‘centromere” is no longer equated solely with that region on the chromosome which associates with spindle MTs. For example, the term “centromeric heterochromatin” is used so often in the ultrastructural literature that many now erroneously equate the term centromere with the heterochromatin of the primary constriction (e.g., Fuge, 1977; Roos, 1977; Brinkley et al., 1980) even though this heterochromatin cannot be considered to be a part of the kinetochore (see above). This association leads to such statements as “the kinetochore is a specialized chromosomal structure situated on the surface of the centromere” (Alov and Lyubskii, 1977) or, “this observation indicates that the kinetochores are physically distinct from the centromere” (Brmner et a/., 1980). Obviously, in these instances the terms kinetochore and centromere are not used as synonyms. A similar situation is encountered in the current LM literature where centromere (and even kinetochore) is used to describe stained or unstained “dots” in the region of the primary constriction (Fig. 2 ) . It is clear that the centrclmere (kinetochore) can sometimes be seen with the LM under suitable circumstances (e.g., Lima-de-Faria, 1958; Hard and Allen, 1977; Clapham and Ostergren, 1978). On the other hand, in many cases one can argue, as Roos (1975) has, that

THE MAMMALIAN KINETOCHORE

5

these preferentially stained regions may in reality be areas of more densely packed heterochromatin, protein components unique to regions of this heterochromatin (e.g., Matsukuma and Utakoji, 1977), or even clear zones similar to those sometimes seen adjacent to the kinetochore in the electron microscope (e.g., Roos, 1975; Heneen, 1975a; Fig. 12 of Rieder and Borisy, 1981). For example, Marks (1975; see also Denton et al., 1977; Brat et al., 1979) considered his differentially stained dots in the primary constriction of Nigella chromosomes to be kinetochores (centromeres) even though these dots bridged “the space across each chromatid.” Brown and Loughman (1980) noted that Cbanding is often used to locate the centromere/kinetochore (whereas in reality it locates the pericentromeric heterochromatin) and they then developed a silver stain which “unequivocably stains the centromeres” of Indian muntjac (minature deer) chromosomes. Similarly, Alves and Jonasson (1978) developed a direct Giemsa technique which they used “for the detailed cytological study of the mouse kinetochore. However, they stress that their observations indicate that they are staining mouse satellite DNA which they then equate with centromeric heterochromatin, which they consider equivalent to the kinetochore (centromere). Finally, Moroi et al. (1981) attempted to localize, at the electron microscope (EM) level, antigen(s) which they had previously (Moroi et al., 1980) shown with the LM to specifically bind to the centromere of mammalian chromosomes. At the ultrastructural level, though, these “centromere” antigen(s) were found to be distributed over a large area corresponding the the whole of the primary constriction (see however Brenner et a/., 198 I ) , an observation which was interpreted by these authors to indicate the specificity of the antibody(s) to the centromere. It is clear from the few examples noted above, and there are many more, that the (mis)use of the term “centromere” has progressed to the point where many even define it as a synonym for the primary constriction (e.g., DuPraw, 1970; Stack, 1974; Fuge, 1977; Clapham and Ostergren, 1978; Bostock and Sumner, 1978; Holmquist and Dancis, 1979; Mclntosh, 1979). This being the case one can no longer argue that the terms kinetochore and cerltromere be used as synonyms, regardless of their original definition. It is no surprise that electron microscopists favor the term kinetochore since they are not referring to an area on the chromosome as broadly defined as the “centromere” currently is. To eliminate this confusion I suggest that the term kinetochore be used as defined by Ris and Witt (1981) to note, at the ultrastructural level, the precise region on the chromosome that becomes attached to spindle MTs. In mammalian cells this region differentiates into a trilaminar disk structure during mitosis and appears to contain unique components (see Section IV) not found in the adjacent heterochromatin or on the remainder of the chromosome. However, unlike Ris and Witt (1981) I suggest that the term centromere be used, as it is now often used by cytologists and geneticists, in a less precise manner to note the region on ”

6

CONLY L. RIEDER

the chromosome ( e . g . , the primary constriction, pericentromeric heterochromatin, etc. J with which the kinetochore is associated.

111. Types of Kinetochores

The chromosomes in some organisms (pafticularly among the monocotyledons and arthropods-see Schrader, 1953; Maeki, 1980) lack a primary constriction and the chromosomal spindle fibers appear at the LM level to terminate along the entire length of each chromatid. These features, combined with the fact that the separating sister chromatids remain uniformly parallel to one another during anaphase (i.e., they exhibit holokinetic movement-Bauer, 1952), prompted Schrader (1935) to suggest that these chromosomes possessed a “diffuse” or “nonlocalized” kinetochore. This view gained considerable experimental support by the finding that chromosome fragments, generated in these organisms by UV irradiation, continue to function in a normal manner throughout mitosis (see references in Schrader, 1953; Hughes-Schrader and Schrader, I96 I). Data concerning the ultrastructure of diffuse kinetochores are sparse and are based, for any one organism (except Luzula), on random nonserial sections. Nevertheless, it suggests the existence of at least two morphologically distinct types of diffuse kinetochores. In the mitosis of Rhodnius (Buck, 1967) iind Oncopelfus (Commings and Okada, 1972) the kinetic activity occurs along most or all of the chromosome length, and the kinetochore material is evenly tiistributed along its poleward edge in a loosely defined “plate” which characteristically stains lighter than the chromatin. This type of chromosome is generally referred to as holocentric (e.g., Braselton, 1971, 1981; Commings ;and Okada, 1972). On the other hand, the MTs associated with the meiotic kinetochores in Philaenus (Ris and Kubai, 1970), Lepidoptera (Maeki, 1981), iind Bombyx mori (Friedlander and Wharman, 1970) appear to terminate at multiple sites along the chromosome directly on the chromatin. This type of chromosome is often referred to as polycentric. A reconstruction of mitotic chromosomes in Luzula and Cyperus (Braselton, 1971) suggests that the kinetic activity in these organisms is also restricted to numerous discrete units along the chromosome since the kinetochore material appears to be distributed as multiple, light staining irregularly shaped packets, which are often recessed within the chromosome. However, Braselton’s (197 1, 1981) conclusion that Luzula chromosomes are polycentric has been questioned by Bokari and Godward (1980) who also reconstructed the kinetochores of metaphase Litzula chromosomes from serial sections and concluded that a single kinetochore extends continuously along most of the length of each chromatid (as in Rhodnius and Oncopeltus). These authors note that fixation and embedding induces a lateral side-by-side adherence of adjacent chromosomes in Luzula and that photomicrographs of sections through these

THE MAMMALIAN KINETOCHORE

7

fused chromosomes can easily be erroneously interpreted as photographs of single chromosomes with multiple kinetochores. Similarly, the polycentric condition reported to exist in the alga Spirogyru (Mughal and Godward, 1973) appears to have also been based on images of localized kinetochores on adjacent overlapping chromosomes. An LM analysis of fixed specimens suggests that the kinetic activity in those Hemiptera which possess diffuse kinetochores becomes restricted to the terminal region of the chromosome during anaphase of meiosis (Hughes-Schrader and Schrader, 1961). This modified behavior does not appear to involve an irreversible change in the distribution of the kinetochore material since fragments of meiotic chromosomes continue to divide normally (Hughes-Schrader and Schrader, 1961). Indeed, the available ultrastructural data on Hemiptera (Rhodnius and Oncopeltus) indicate that the meiotic chromosomes lack the plate-like kinetochore structure characteristically associated with mitotic chromosomes; the K-MTs appear to terminate instead along the body of the metaphase chromosome within the chromatin (i.e., these chromosomes are now polycentric). However, there is currently no ultrastructural evidence to indicate that these K-MTs become redistributed and/or are restricted, during early anaphase, to the terminal ends of the chromosomes (e.g., Buck, 1967). Rather, Commings and Okada ( 1972) have suggested that the change in kinetochore structure between mitosis allows for the terminalization of chiasmata. In this respect, the terminalization of chiasmata may lead to a restriction of anaphase kinetochore activity which is more apparent than real. The diffuse kinetochores of some organisms (e.g., Philaenus, Lepidoptera) appear to consist primarily of DNA and protein (i.e., chromatin) since, after conventional fixation and staining, the associated MTs appear to terminate directly on the chromatin without evidence of an additional structural component. On the other hand, the kinetochores on all of the holocentric chromosomes (including Luzula) contain additional material which stains lighter than the chromatin. In one case (Luzula; Braselton, 1980) this material has been shown to contain an RNase-sensitive component which can be selectively stained by a method (Bernhard, 1969) which preferentially stains ribonucleoprotein (RNP). To my knowledge there is no additional information regarding the cytochemistry of diffuse kinetochores. In contrast to the diffuse kinetochore, the kinetochore in most organisms is located on only a small segment of the chromosome (i.e., the primary constriction-see Section 11). During anaphase this segment leads the way poleward, bending the chromosome into the familiar “V” or “J” shape. Early investigators found that these “localized” kinetochores could, in some cases, be preferentially stained (see reviews of Schrader, 1953; Lima-de-Faria, 1958; Mazia, 1961). However, their small sizes (in most cases about the limit of resolution of the LM) prohibited a detailed analysis of their composition and structure.

8

CONLY L. RJEDER

Ultrastructural investigations have since revealed that the localized kinetochore varies considerably in structure according to the stage of division and the organism. In general it appears during metaphase as either (1) a single or multilayered disk on the surface of the chromosome (e.g., mammals, some algae, insects, slime moulds, protozoa, ciliates-see references in Luykx, 1970; Fuge, 1974, 1977; Heath, 1979; Bostock and Sumner, 19781, (2) a ball of illdefined material embedded in a more electron opaque chromatin cup (e.g., most plants including Haemanthus, Lilum,Alliuni-see references in Bajer and MoleBajer, 1972; Alov and Lyubskii, 1977), or featureless and difficult to differentiate from the chromatin (e.g., some insects, fungi, yeast, and protista-see references in Luykx, 1970; Kubai, 1975; Fuller, 1976; Heath, 1979). These latter types generally possess only one K-MT. The remainder of this article will focus primarily on the formation, structure, and composition of the mammalian kinetochore (and its associated fiber). For a more detailed description of kinetochore morphology, in a variety of nonmammalian cells, the reader should consult the reviews cited above. IV. Mammalian Kinetochore Structure and Chemistry A. INTERPHASE Discrete patches of material shown to be precursors of mitotic kinetochores are visible within the interphase nuclei of some plants (Church and Moens, 1976; Moens and Moens, 1981) and protozoa (e.g., Kubai, 1973; Ris and Kubai, 1974). Similar structures are not apparent in the interphase nuclei of mammals after conventional fixation and staining. It has recently been shown however that sera from patients with the CREST variant of progressive systemic sclerosis contain high titers of an antibody which binds specifically to the centromere region of chromosomes from mammals (Moroi et al., 1980; Tan et al., 1980; Fritzler and Kinsella, 1980), flies (Will et al., 1981), and probably to the same region in other types of organisms. This antibody has been demonstrated by indirect immunoelectron microscopy to be specific for antigen(s) associated with the mitotic kinetochore (Brenner et al., 1981; Fig. 3). More important in the context of this section is the fact that this antibody binds to discrete spherical patches, approximately 0.22 pm in diameter, within the nuclei of inrerphuse mammalian cells (Figs. 4 and 5 ) . The number of these interphase staining foci corresponds with the number of chromosomes within the cell. Since this antibody binds to mitotic kinetochores, as well as localized foci within interphase nuclei, at least some component(s) of the mitotic kinetochore remain associated with the interphase chromatin throughout the cell cycle. This observation prompted Pepper et al. (1980) to suggest “that a kinetochore organizer exists in the genome which retains some structural integrity in the decondensed chromatin of interphase nuclei” (see also Section V).

THE MAMMALIAN KINETOCHORE

9

FIG. 3. Electron micrograph of colcemid-treated PtK2 kinetochores stained with the antikinetochore serum by the immunoperoxidase method. Note the double-layered appearance of the kinetochore and the specificity of the staining reaction for this structure. X l6.000: inset. X32,OOO. (From Brenner et d.,198 1 .) FIG. 4. lmmunoperoxidase localization of the kinetochore antigen (Fig. 3) in an interphase PtK2 cell. The antigen is restricted to localized foci or “presumptive kinetochores.” See text for details. X12.800; inset, X42.300. (From Brenner cf a/., 1981.)

Brenner et al. ( 198 I ) refer to these interphase staining foci as “presumptive kinetochores” because of their lack of structural similarity to similarly stained regions on metaphase chromosomes. These investigators note that the prekinetochores in PtK, (male rat kangaroo) cells become visibly duplicated during late G , of the cell cycle, after the completion of DNA synthesis (Fig. 5). At present it remains an open question as to whether this antigen becomes associated with the newly replicated chromatin prior to, during, or after the condensation of

10

CONLY L. RIEDER

the later replicating pericentromeric heterochromatin. (In this respect the question of whether prekinetochores are formed, in part, from the condensation of the pericentromeric heterochromatin can be approached by determining whether similar structures are present in mammals which appear to lack this heterochromatin.) Information concerning the nature of this kinetochore antigen was obtained by Moroi et al. (1980) who first attempted to solubilize it with different reagents in hopes of later identifying the antigen-antibody complex. The lack of success of this method (see also Cox et a l . , 1980) forced them to pretreat periodiate/lysine/ paraformaldehyde-fixed RAMOS cells with a variety of enzymes and chemical reagents prior to reaction with the antiserum. Their results, assayed by an absence or significant decrease of immunofluorescent staining intensity, indicated that the antigen was probably a protein tightly bound to centromere DNA (see also Cox et al., 1980) since it was destroyed by DNase and trypsin but not by RNase (Fig. 6). Brenner er a / . (1981), using serum supplied by Moroi, further characterized this antigen by noting that the immunofluorescent staining intensity of the kinetochore was not diminished by initially preabsorbing the serum with tubulin, actin, or microtubule-associated proteins (MAPS). This was interpreted to indicate that the antigen was not a previously recognized or suspected component of the kinetochore (see Sections IV,C and D). The ability to specifically stain interphase prekinetochores allows one to approach the question of how these structures are arranged three dimensionally within the interphase nucleus. Moroi ef al. ( I98 1) found that the prekinetochores of human lymphoid and Chinese hamster cells were associated predominantly with either the surface of the nuclear envelope (NE) or with nucleoli. They concluded that the “centromere regions of the chromosomes in interphase are not randomly distributed within the nucleus.” At the same time a similar study by Brenner er a/. (198 1) on PtK, cells revealed a random distribution of prekinetochores with no consistent association with the NE. They did note that a few prekinetochores had an affinity for the nucleoli, but they considered this to be a manifestation of the proximity between the kinetochore and the nucleolar organizer on the sex chromosomes of PtK,. Although the results of these two studies differ, it is apparent from both that prekinetochores in mammalian cells are not clustered or polarized within the interphase nucleus as they are in Alliurn (Church and Moens, 1976). (This difference may be due, in part, to the stability of the chromocenter in those plant cells which have a long G , phase of the cell cycle.) The immunological studies described above should be considered at present as preliminary since the possibility exists that the serum used by these investigators contained a variety of antibodies-some of which bound to antigens found only on mitotic kinetochores and others which bound to totally different antigens unique to the so-called “presumptive kinetochores” of interphase cells. Even though these results need confirmation with a monoclonal antibody system, they

THE MAMMALIAN KINETOCHORE

11

FIG. 5 . (A and B) lnterphase presumptive kinetochores as seen by indirect immunofluorescence. These cells have also been double-stained with anticentrosome antisera to reveal the centrosome (arrow). (A) Early interphase cell with single presumptive kinetochores. (B) Late interphase cell with double presumptive kinetochores. See text for details. x832. (From Brenner et a / . . 1981.) FIG.6. (A-D) The effects of various enzymatic treatments on the indirect immunofluorescent staining of interphase presumptive kinetochores with the kinetochore antibody in RAMOS cells (human B lymphocyte cell line). x300. (From Moroi et a / . . 1980.) ( A ) Control. (B) Image after digestion with DNase I. Kinetochores no longer stain after this treatment. (C) Image after digestion with RNase A. No significant changes in the kinetochore staining pattern are visible. (D)Image after digestion with a combination of trypsin and 0.01% NaDOdS04. Kinetochores no longer stain after this treatment (however, digestion with trypsin alone does not remove the kinetochore antigen-not shown).

12

CONLY L. RIEDER

illustrate the potential of the immunological approach for analyzing the composition of the kinetochore. It can be expected that monoclonal antibodies to kinetochore components will be produced in the future. These will likely be used as probes for determining the molecular composition and location of components making up the kinetochore, as well as any cell cycle-related changes in the distribution of these components.

B. PROPHASE Kinetochores can be recognized, after conventional fixation and staining, in very early prophase PtK cells as soon as the chromosomes begin to condense. They appear as 0.6- to 0.8-pm-diameter spherical patches of a homogeneous fibrillar material located in the vicinity of the condensed pericentromeric heterochromatin (Heneen, 1975a). These kinetochores occur in pairs and frequently stain lighter than the heterochromatin, a feature which often facilitates their detection within the nucleus. There is good evidence that early prophase kinetochores arise, in part, from the smaller interphase prekinetochores described in the last section. As chromosome condensation continues through mid-late prophase, the replicated kinetochores can be seen to occupy positions on opposite sides of the forming primary constriction (Jokelainen, 1967; Brinkley and Stubblefield, 1966, 1979; Roos, 1973a; Heneen, 1975a; Figs. 7 and 8), an arrangement which is thought to play a role in their bipolar attachment to the early prometaphase spindle (see Mazia, 1961; Nicklas, 1971; Roos, 1973a,b). At this time they consist of a sphere (or ball) of fibrillar material, similar in size to and generally undistinguishable from those found in early prophase cells (Roos, 1973a; Heneen, 1975a). The kinetochore “ball” is embedded in a well-defined and denser staining “cup” (Fig. 7). This cup forms the apex of the primary constriction and is thought to be composed, in some organisms (see Section II), of the pericentromeric heterochromatin (Heneen, 1975a). In some cases the kinetochore material appears to be “in linear continuity with the chromatin of the two” chromosome arms (Heneen, 1975a). This continuity, though, may simply reflect an overlapping of the chromatin proximal to the kinetochore. It has been suggested that the kinetochore forms after the initiation of chrornosome condensation (Journey and Whaley, 1970; Roos, 1973a), or instead that its formation “is a parallel and not an ensuing process to chromosome condensation” (Heneen, 1975a). However, the recent discovery of prekinetochores in interphase cells (last section) indicates that the formation of what will be the (mitotic) kinetochore is initiated during interphase (i.e., after S), prior to the initiation of chromosome condensation. The “ball” of the prophase kinetochore appears to undergo little structural modification throughout prophase (Roos, 1973a; Heneen, 1975a). In addition to

THE MAMMALIAN KINETOCHORE

13

FIG.7. Conventionally fixed and stained thick section (0.25 pm) through a mid-prophase PtK, nucleus. Note the intimate association between the ends of the chromosomes and the NE. Grey patches near the middle of the nucleus are nucleolar material. The arrow notes the position of the kinetochore shown at a higher magnification in the inset. See text for details. ~ 7 3 5 0 ;inset, X25.600. (From Rieder, 1980.) FIG. 8. Thick section (0.25 pm) through a mid-prophase PtK, nucleus stained by the uranyl-EDTA-lead method. The most electron opaque components of the nucleus are the nucleolar material (large arrow) and the numerous kinetochores (arrowheads). At a higher magnification (inset) the electron opaque component of the kinetochore can be seen to be restricted to the ball, which sits in an electron translucent chromatin cup. See text for details. x4800; inset, X 19,500. (From Rieder. 1980.)

14

CONLY L. RIEDER

the antigen described in the previous section, it also contains a component which can be selectively stained by a method which preferentially stains RNP (Rieder, 1980; Fig. 8). The fact that this staining reaction is abolished by RNase digestion and cold perchloric acid extraction (Rieder, 1979a,b) supports the contention that the prophase kinetochore contains RNP. The antigenic component found to be associated with interphase and prophase kinetochores by Brenner et af. (198 1) is removed by DNase and trypsin but not by RNase (Fig. 6). Thus it is suspected to be a protein which is strongly bound to kinetochore-associated DNA. Unfortunately, RNA bound to DNA via a protein (i.e., RNP) would also be expected to be extracted by either DNase or trypsin but not necessarily by RNase alone (e:.g., if the RNA was complexed to proteins which make it inaccessible to RNase--see Pepper and Brinkley, 1980). Yet the different distribution of these two components, within the fully formed metaphase kinetochore (see Section IV ,D) indicates that they are not the same. As a rule MTs are not associated with prophase kinetochores in vivo or after lysis into tubulin-containing solutions which promote MT assembly on prometaphase kinetochores (Snyder and McIntosh, 1975; DeBrabander et af., 1980, 1981a). Thus the lack of MTs on prophase kinetochores is not due simply to the absence of tubulin within prophase nuclei. On the other hand the ability of the kinetochore to acquire MTs, which occurs immediately after the breakdown of the NE, appears to correlate with and may be the result of a structural change within the kinetochore (see next section). Kinetochore-directed chromosome movement does not occur in the mitotic prophase nucleus of higher plants and animals (see exception in McIntosh, 1979) although the chromosomes may be passively moved in response to nuclear rotation (e.g., Rieder and Bajer, 1977b). In fact, the positioning of kinetochores within late prophase PtK, nuclei appears to be determined primarily by the placement of a kinetochore on a particular chromosome (i.e., metacentric versus submetacentric), and by the fact that the arms of the condensing chromosomes in these cells are attached to the nuclear envelope (see Rieder, 1980). This is in contrast to the meiotic prophase nuclei of some spermatocytes (e.g., house cricket, earwig, and certain mantids) where the ends of the chromosomes, and sometimes even the kinetochores, may show movement directed toward and away from the centrosomes (references in Schrader, 1953; Rickards, 1975). This movement often results in a weak accumulation of bivalents around the late prophase centrosomes. Micromanipulation experiments on cricket spermatocytes (Richards, 1975) indicate that these movements occur in close association with the NE. Even though they are reversibly blocked by colcemid (but not lumicolcemid) they appear to occur in the absence of intranuclear MTs (Rickards, 1975). This has been interpreted to indicate that the force producer for this movement is associated with the centrosomes, which somehow influence chro-

THE MAMMALIAN KINETOCHORE

15

mosome behavior through or at the surface of the NE (see discussion in McIntosh, 1979). C. PROMETAPHASE The breakdown of the NE at the end of prophase initiates prometaphase, a stage which is characterized by the attachment of kinetochores to the forming spindle. This attachment correlates with a structural change in the kinetochore, with the acquisition of MTs by the kinetochore, and with the initiation of directed chromosome movement (reviews of Nicklas, 197 1 ; Bajer and Mole-Bajer, 1972; McIntosh, 1979). The “ball and cup” kinetochore of mammalian prophase chromosomes differentiates after NE breakdown, into a single dense-staining plate-like structure which is surrounded by a lighter staining fibrillar material known as the “corona” (Jokelainen, 1967; Figs. 9 and 10). Roos (1973a.b) observed that kinetochores closest to the site of initial NE breakdown were the first to differentiate, prompting him to suggest that this process is triggered by cytoplasmic factors diffusing into the rupturing nucleus. Some investigators (e.g., Fuge, 1974; Alov and Lyubskii, 1977) have ruled this unlikely due to a single but widely circulated early report (Brinkley and Stubblefield, 1970) describing the presence of kinetochore plates in late prophase nuclei. Neveretheless, a careful examination of this work (from which Fig. 9 is taken; see also Fig. 17 in Nicklas, 1971) indicates that these so called “prophase” cells are in reality early prometaphase cells in the process of NE breakdown. Similarly, Fuge (1977) feels that the differentiation of the kinetochore plate “can be assumed to coincide with the assembly of the first MTs in animals.” Yet this is clearly not the case since ( 1 ) a kinetochore plate is frequently present on early prometaphase kinetochores which have not yet acquired MTs (see Roos, 1973a; Rieder and Borisy, 1981; Fig. 9), and (2) a kinetochore plate is formed after the NE breaks down at temperatures (e.g., 6”C), or in the presence of drugs (e.g., colcemid), which inhibit spindle MT formation (Figs. 10-12). A likely possibility is that the formation of the kinetochore plate in mammalian cells is a prerequisite for the appearance of MTs at the kinetochore. Although one could argue that early prometaphase kinetochores that are attached to the spindle sometimes lack this plate, these kinetochores are, as a rule, stretched and may only appear to lack the plate due to deformation and/or an inappropriate plane of section (see Roos, 1973a,1976; e.g., Fig. 13). Furthermore, since the kinetochore plate is the eventual anchor point and termination site for K-MTs (see Sections IV,D and VI,B and C) it is likely that its formation in vivo preceeds the acquisition of MTs by the kinetochore. (It should be noted that the formation of a plate structure is not a necessary prerequisite for the function of many types of

16

CONLY L. RIEDER

FIG.9. Thin section from an early prometaphase Chinese hamster fibroblast. Note the light staining corona material associated with the well differentiated kinetochore ( K ) plates. Remnants of the nuclear envelope (NE) are visible in the vicinity of the chromosomes. x24.000. (From Brinkley and Stubblefield, 1970.) FIG. 10. (A and B) Thick sections (0.25 p,m) through kinetochores from PtKl cells which entered prometaphase at a temperature (6°C) which inhibited spindle MT formation. (A) Sister kinetochores from a cell held at 6°C and then fixed.2.5 hours after NE breakdown. Note that a loosely organized plate appears to be associated with the surface of each kinetochore. (B) A kinetochore from a cell fixed 6 minutes after the initiation of an 18°C recovery from 6°C. The outer kinetochore plate and its associated light staining corona material (arrow) is cleraly revealed in this section. ( A ) ~ 2 5 , 0 0 0 (B) ; ~ 2 7 , 5 0 0 .(From Rieder and Borisy, 1981.)

nonmammalian kinetochores. The ball and cup prophase kinetochore in Haemanthus, for example, acquires MTs during prometaphase without undergoing an obvious structural change-see Bajer and Mole-Bajer, 1972; Jensen, 1982.) The prophase kinetochore in mammalian cells has been shown to contain at least two components (and most likely also chromatin) including one (RNP)

THE MAMMALIAN KINETOCHORE

17

FIG. 1 1 . A thick section (0.25 pm) through kinetochores, stained by the uranyl-EDTA-lead method, in a PtK, cell which entered prometaphase in the presence of colcemid. The kinetochore (outer) plates and associated corona material remain electron opaque while the chromosomes appear bleached. See text for details. x23.700. FIG. 12. Thin section through sister kinetochores from a colcemid-treated PtK, cell. This cell was processed for the indirect localization of tubulin by the peroxidase method. Note the dense staining of the kinetochore outer plates but not of the corona material (cf. Fig. I I ) . X36.500. (From Pepper and Brinkley. 1977.)

which is preferentially stained by Bemhard’s ( 1969) treatment, and a second (protein) component which can be visualized by immunocytochemical means (see Section IV,B). Both of these components are associated, after NE breakdown, with the prometaphase kinetochore: one (RNP) is associated with the disk and corona while the other (protein) is located within the disk and on the chro-

CONLY L. RIEDER

FIG. 13. Thick section (0.25 p m ) from an early prometaphase RK,cell recovering from a 6°C cold shock. An analysis of serial sections indicated that only three kinetochores in this cell were attached to the forming spindle (i.e., those closest to the centrosomes) at the time of fixation and two are pictured here. Both are stretched toward the polar area (asterisk) and possess associated MTs. See text for details. X53,500. (From Rieder and Borisy, 1981.)

matin immediately subjacent to the disk (see below), These cytochemical results are corroborated by the observation that the formation of the kinetochore plate during early prometaphase can be correlated with a reduction of the finely fibtillar material associated with the prophase kinetochore (Jokelainen, 1967; Roos, 1973a; Fig. 10A). Thus, the kinetochore plate formed during early prorneraphase appears to be organized, in part, from components associated with the prophase kinetochore. The ultrastructure of the prometaphase kinetochore in cells treated with agents which inhibit spindle MT formation (e.g., cold, colcemid, vinblastine, etc.) has been thoroughly documented (e.g., Figs. 10- 12). After conventional fixation and staining these kinetochores appear as a single filamentous plate which is free of MTs and which sometimes shows continuity with the adjacent chromatin (Krishan, 1968; Journey and Whaley, 1970; Brinkley and Stubblefield, I 966,1970; Roos, 1973a; Alov and Lyubskii, 1977; Bielek, 1978a,b; Rieder, 1979a). This plate is approximately 25-40 nm thick, is variable in length (see below), and runs parallel to the surface of the chromosome. It is separated from the chromosome by a 25- to 40-nm-wide zone which stains less dense than either the chromosome or the plate. Looping out from the kinetochore plate is a fine fibrillar light staining material analogous to the corona first described by Jokelainen ( 1967). The corona material, which appears preferentially stained after Bemhard’s (1969) treatment (Fig. I I), extends from the surface of the kinetochore plate for a variable (i.e., up to 0.25 pm) distance (Brinkley and Stubblefield, 1970; Roos, 1973a; Bielek, 1978a,b).

THE MAMMALIAN KINETOCHORE

19

The kinetochore plate, formed in prometaphase cells under the influence of mitoclastic drugs (e.g., colcemid), varies in diameter according to the cell type. Luykx ( 1970; see also Krishan, 1968) noted that colcemid treated kinetochores are much larger than metaphase kinetochores. In PtK, (female rat kangaroo) cells the colcemid treated kinetochore plate is 0.5-0.9 pm in diameter compared to a range of 0.4-0.6 pm in untreated cells with fully differentiated metaphase kinetochores (Roos, 1973a). These observations have been confirmed by others (Bielek, 1978a,b; Rieder, 1979a; Brenner et a l . , 1981) and have been interpreted to indicate that the kinetochore plate, formed in the presence of colcemid, grows to an exaggerated size. Yet it appears that an increased plate size is a general feature of unattached and/or differentiating prometaphase kinetochores. Both Jokelainen ( 1967) and Roos (1973a) note, for example, that unattached early prometaphase kinetochore plates in untreated cells are larger than those found in metaphase cells. Jokelainen (1967) even stresses, in respect to this point, that it is important “to distinguish between differentiating and mature kinetochores.” Roos (1973a) attributed the progressive decrease in plate size during prometaphase to condensation of the whole structure and to the association of MTs with the kinetochore. Kinetochores which do not acquire MTs during prometaphase appear to be inhibited from fully differentiating or condensing into the smaller trilaminar structures seen in metaphase cells. This, and the continued condensation of the chromosome in the presence of mitoclastic drugs, may lead to an exaggerated condition in which the sister kinetochores appear to almost completely encircle the whole of the primary constriction (e.g., Journey and Whaley, 1970; Bielek, 1978a,b). In addition to a larger plate diameter, conventionally fixed and stained colcemid-treated kinetochores also appear to lack the inner plate which is characteristically associated with fully differentiated (trilaminar) metaphase kinetochores (Roos, 1973a; Pepper and Brinkley, 1977; Bielek, 1978a,b; Rieder, 1979a; see Section IV,D). Pepper and Brinkley (1977) suggested that the lack of a trilaminar organization in colcemid-treated kinetochores is partly due to the specific interaction of colcemid with kinetochore-associated tubulin (which is thought to be a component of the kinetochore-see below; Fig. 12 and Section 1V.D). However, unattached kinetochores in untreated prometaphase cells are larger than metaphase kinetochores (see above) and also lack the inner layer (Roos, 1973a; Rieder, unpublished). In addition, prophase kinetochores differentiate into a structure resembling colcemid-treated kinetochores when the N E is allowed to break down at a temperature which inhibits MT formation (Rieder and Borisy, 198I ) . Thus the structural differences between the colcemid-treated kinetochore and the fully differentiated trilaminar disk cannot be ascribed to the influence of the drug. Finally, unattached kinetochores may only appear to lack an inner plate since components of this structure are ill defined and difficult to distinguish from the underlying chromatin even in fully differentiated kinetochores (e.g., cf.Figs. 14 and 15). In support of this, Brenner et d . (1981)

FIG. 14. Thin section through the sister kinetochores of a metaphase PtK, chromosome. X-MTs terminate in the outer kinetochore plate which is clearly separated from the chromosome by a more electron translucent layer. Note that some MTs of this K-Fiber (arrows) do not terminate on the kinetochore but penetrate instead into the adjacent chromatin. X35,200. (From Mclntosh er al.. 1975a; by permission of Raven Press.) Fic. 15. Very early anaphase PtK, kinetochores. Note their trilaminar appearance and the distinct bundles of K-Fiber MTs. some of which bypass the kinetochore. X9000. (From Roos, 1973a.) R c . 16. Thick section (0.25 pm) through a chromosome from a mouse cell which was allowed to recover for 30 minutes from a colcemid treatment. The outer kinetochore disk on the left shows a well-developed fibrous corona (arrow). MTs are attached to the sister kinetochore (at right I which shows a reduction of corona material. X25.600. (From Ris and Witt, 1981.) FIG. 17. (A and B) Thick (0.25 pm; A) and thin (B) sections of metaphase PtK, kinetochores stained by the uranyl-EDTA-lead method. The electron opaque material is restricted to (or in part forms) the inner plate of the trilaminar structure while the outer plate appears similar in contrast to the chromosome (cf.Figs. 8 and 11). (A) x32.000; (B) X24.300. (From Rieder, 1979a.)

THE MAMMALIAN KINETOCHORE

21

have recently shown that colcemid-treated kinetochores possess a protein component which is closely associated with the chromosome subjacent to the outer plate, and that this component is not readily distinguishable from the chromosome after conventional fixation and staining (e.g., cf.Fig. 3 and 9). In an early report Brinkley and Stubblefield (1966, 1970) noted that the outer kinetochore plate of colcemid-treated Chinese hamster cells consisted of a pair of axial fibrils 5-8 nm in diameter. (Their kinetochore model, as discussed in these papers, has been retracted in Pepper and Brinkley, 1977, and will not be discussed here.) These fibrils were either distinctly separate or twisted together in a cohelical manner. This observation has since been confirmed by others (e.g., Krishan, 1968; Roos, 1977) and a similar filamentous substructure can be seen in the kinetochore of untreated Indian munjac chromosomes. In this organism the outer plate appears to contain two parallel or intertwined 9-nm fibers in longitudinal sections whereas in transverse sections these structures appear dot or tubelike (Commings and Okada, 1971). Ris and Witt (1981) have also observed similar images by stereomicroscopy in thick sections of Chinese hamster and mouse kinetochores. However, they interpret these filamentous structures not to be continuous parallel or coiled 10-nm fibrils but rather to be overlapping adjacent and parallel aligned 10-nm loops, the sides of which only appear to form continuous fibrils as a result of being cut in thin section. As noted above the colcemid-treated prometaphase kinetochore contains an RNase-sensitive component (Bielek, 1978a,b; Rieder, 1979a; Fig. 1 I ) and a distinctly different protein component which enables it to be selectively stained by immunocytochemical methods (Brenner et al., 1981; Fig. 3). indirect immunoelectron microscopy suggests that tubulin is also present within the outer plate of these kinetochores (Pepper and Brinkley, 1977; Fig. 12) and there is some evidence that this component promotes K-MT assembly (Pepper and Brinkley , 1979; see Section VI). Finally, the kinetochore plate in colcemid-treated cells is specifically decondensed by DNase I treatment (Pepper and Brinkley, 1980) a result which strongly suggests the presence of DNA in this structure (see also Section IV,D). The notion that the kinetochore plate of colcemid-treated cells contains DNA is strengthened by the observation of others that this plate frequently appears continuous with the adjacent chromatin. However, Rattner et al. (1978) noted that DNase I digestion of chromosomes, isolated from colcemid-treated mouse cells, resulted in the preferential digestion of the chromosome arms but left the pericentromeric heterochromatin and kinetochores intact. The reason for the obvious discrepancy between this study and that of Pepper and Brinkley (1980) is unclear. It is certainly possible that DNase I digested the kinetochore-associated DNA in the study of Rattner et al. (1978) without removing the additional kinetochore components described above. These digested kinetochore remnants could then easily have been mistaken for "intact" kinetochores since even intact

22

CONLY L. RIEDER

untreated kinetochores have an ill-defined morphology in whole mount chromosome preparations (e.g., see Moses and Counce, 1974; Rattner et al., 1975; Fig. 1). The structure and cytochemistry of the prometaphase kinetochore in cells not treated or fully recovered from treatment with mitoclastic drugs are not well documented. Unattached early prometaphase kinetochores (i.e., those which have not yet acquired MTs) in PtK, (Roos. 1973a,b; Rieder and Borisy, 1981) and in Chinese hamster (Brinkley and Stubblefield, 1970) cells appear structurally as either a fine fibrillar ball resembling prophase kinetochores or as a disk and associated corona material (e.g., Fig. 9) identical to that of kinetochores formed in the presence of colcemid. Attached kinetochores appear as a single disk surrounded by corona material (e.g., Fig. 28) or stretched and fibrillar without an obvious internal fine structure (Roos, 1973a; Alov and Lyubskii, 1977; Rieder and Borisy, 1981; Goldstein, 1981; Fig. 13). As previously noted this latter type of kinetochore may contain a plate which is obscured by distortion arising from stretching or by the plane of section. According to Roos (1973a; see also Goldstein, 1981) fully formed trilaminar kinetochores (see next section) are never seen in the early prometaphase of PtK, cells. D. METAPHASE

By metaphase the mammalian kinetochore is considered to be a fully differentiated structure composed of three distinct plates or disks which behave as a single unit (Jokelainen, 1967; Roos, 1973a). In general it consists of (Figs. 14, 15, and 17) (1) an outer plate, approximately 30-40 nm wide, which is analogous to the single plate seen in conventionally fixed and stained early prometaphase or colcemid-treated kinetochores; (2) a 15- to 35-nm-wide middle layer composed of a loosely organized fibrillar component; and (3) a 20- to 40nm-wide coarsely granular or fibrillar inner plate which is continuous with the subjacent chromatin. Alov and Lyubskii (1977; see also Lyubskii, 1974) report that the kinetochores of Chinese hamster cells undergo a morphological change during spindle formation, from a plate-like structure in early prometaphase to a loosened “fibrillospherical” or irregular flame shape by late prometaphase. These authors suggest that the plate-like kinetochore is “inactive” and that “the main mechanism of kinetochore activation during prometaphase is despiralization of the fibers of the kinetochore core and plate. It is apparent, however, that these investigators failed to consider that the “loosened” hemisphere or flame shape of the late prometaphase (i.e., active) kinetochore may result from,and nor be a cause of, kinetochore activity. Fuge ( 1977) has similarly suggested that the morphological changes noted to occur in the kinetochore during prometaphase in newts (MoleBajer et al., 1975) and in crane flies (Muller, 1972) “can obviously be attributed to pulling forces generated outside the kinetochore region. ”

”

THE MAMMALIAN KINETOCHORE

23

Kinetochore MTs terminate in the outer kinetochore plate (see Section VI, B), which by metaphase shows a significant reduction of associated corona material (Ris and Witt, 1981; cf.Figs I 1 and 17; Fig 16). The structural characteristics and staining properties of the outer plate have been interpreted by some to indicate a lack of chromatin in this structure (Commings and Okada, 1971; Roos, 1977; Fuge, 1977). On the other hand, the continuity of the outer plate with the adjacent chromatin has been cited by others (e.g., Brinkley and Stubblefield, 1970; Luykx, 1970; Rieder, 1979a) as evidence that this structure contains chromatin. More recently Pepper and Brinkley (1980) studied the effects of nuclease treatment on the structure of the PtKz kinetochore and noted that “DNAse 1 specifically decondensed the plate structure,” an observation which led them to conclude that “DNA is a major structural and functional component of the kinetochore.” At the same time Brinkley er a / . (1980) found that a brief hypotonic treatment can be used to reversibly disperse the chromosomes in mammalian cells into chromatin fibers. Although this treatment dissociated the kinetochore plate the K-MTs were stable to the treatment and appeared attached to chromatin fibers. This observation was interpreted to indicate that chromatin is a major component of the outer plate. A similar but more thorough i1,vestigation was conducted by Ris and Witt (1981) who found that the stability of the outer plate in Chinese hamster and mouse cells was influenced by ionic strength. At high KCI concentrations (in D20) the outer disk was present, while at low concentrations no disk was recognizable and K-MTs ended in a mass of chromatin fibers (Fig. 18; see also Peterson and Ris, 1976). These fibers had a diameter of 10 nm, compared to the 20-nm chromatin fibers making up the remainder of the chromosome, and stained positive for DNA by Aggarwal’s ( 1976) Feulgen-Pt thymine method. Additional stereo observations on whole mounts and thick sections led to the conclusion that the outer disk is formed from 10-nm chromatin fibers “arranged in parallel in one plane and which issue from the inner disk on the side and end blindly in a hairpin on the opposite edge of the disk.” The dispersion experiments of Brinkley er a / . (1980) and Ris and Witt (1981) are not as convincing for the presence of chromatin in the outer plate as these authors would lead one to believe since their results are also consistent with the notion that chromatin simply adhered to these MTs after the outer plate was destroyed. Nevertheless these results, along with the observation that the outer plate stains positive for DNA (Ris and Witt, 1981) and is specifically decondensed by DNase treatment (Pepper and Brinkley, 1980), provide strong support for the contention that the outer plate of the mammalian kinetochore contains DNP. Kinetochores can be selectively contrasted in siru (Esponda, 1978) and in whole mount chromosome preparations (Moses and Counce, 1974) with alcoholic PTA. an observation which was initially interpreted to indicate that they possess (basic) proteins not found on the remainder of the chromosome. Although the validity of this interpretation is questionable on the grounds that the

24

CONLY L. RIEDER

FIG. 18. Stereo photographs of a thick section (0.25 pm) from a Chinese hamster metnphase chromosome treated for 15 minutes with a hypotonic solution before fixation. No outer disk structure is present and the K-MTs end on what appears to be a mass of chromatin fibers. See text for details. x40.000. (From Ris and Witt, 1981.)

staining mechanism of alcoholic PTA is unknown, it is supported by a recent immunoelectron microscopic study which indicates that both the inner and outer kinetochore plates in metaphase PtK cells contain an uncharacterized protein component (Brenner et al., 1981). In addition to this antigen, there is also immunocytochemical evidence which suggests that the protein tubulin is associated with both plates of the metaphase kinetochore (Pepper and Brinkley, 1977; Fig. 19). This latter result is not surprising when one considers that MTs, which are composed of tubulin, are organized near (see Section V1,C) and become firmly attached to the kinetochore. Yet a similar study on PtK, by DeBrabander et al. (1979a) failed to detect tubulin in the metaphase kinetochore. Since the discrepancy between these two studies has yet to be explained the question of whether tubulin is an intrinsic structural component of the kinetochore remains to be resolved. As previously noted the outer disk of the metaphase kinetochore lacks the prominent fibrous corona found to be associated with the outer disk of kinetochores which are free of MTs. There is preliminary evidence to suggest that this component simply becomes redistributed as the kinetochore acquires M T s : the electron opaque material seen to be associated with the kinetochore plate and corona of colcemid-treated (prometaphase) kinetochores after Bernhard’s treatment (Fig. 11) becomes restricted to the inner layer of the attached trilaininar kinetochore (Fig. 17; see below). Recent studies on the mechanism of I(-MT formation (Section VI,C) indicate that these MTs are nucleated by and/or associate with the corona prior to becoming attached to the outer disk. This structure

,

THE MAMMALIAN KINETOCHORE

25

therefore appears to be an important, but for the most part overlooked, component for kinetochore function. Other than the preliminary evidence suggesting the presence of RNP (see Section IV,C), the composition of the corona is unknown although it does not appear to contain the 210,000 MW MAP (DeBrabander et a / . , 1981b), tubulin (cf. Figs. 11 and 12), or the unidentified antigenic protein (cf. Figs. 3 and 11). (In this respect the extent to which this component is preserved after cell lysis into physiological buffers, or even by the wide variety of fixation protocols currently in use, has yet to be addressed.) The middle layer of the metaphase kinetochore is characterized by a weak staining which reflects either a paucity of (preserved?) structural elements or the presence of additional components not stainable by uranyl and lead ions (Figs. 14-17). This layer is not an artifact produced by tension at the kinetochore since it is also present on unattached kinetochores. One cannot eliminate the possibility, however, that it is formed in part by the differential shrinkage of the chromosome in relation to the outer plate during dehydration and embedding for EM. At present nothing is known concerning its cytochemistry although Ris and

Fic;. 19. Thin section from a PtK, metaphase spindle processed for the indirect immunoperoxidase localization of tubulin. Note the stained kinetochores ( K ) on the unstained chromosomes. x8500. (From Pepper and Brinkley. 1977.)

26

CONLY L. RIEDER

Witt (1981) speculate that the loosely organized structural elements in this region represent 10-nm chromatin fibers. The inner plate of the metaphase kinetochore is intimately associated with and often difficult to distinguish from the chromosome (e.g., Fig. 14). Roos (1073a) noted that its formation appears to correlate with the acquisition of MTs by the kinetochore. As noted above a component, thought to contain RNP, is ibund within the corona material of unattached kinetochores and this component becomes restricted to the inner electron-dense plate after attachment (Rieder, 1979a.1980; cf.Figs. 1 1 and 17). The inner plate also appears to contain additional components not readily distinguishable from the chromosome after conventional fixation and staining procedures. For example, when metaphase cells are digested with RNase, extracted with cold-perchloric acid, and stained by Bernhard’s method a discrete inner kinetochore plate remains faintly visible (Rieder, 1979a). The decrease in contrast of this inner plate in response to Bernhard’s treatment suggests the presence of DNP. That the inner plate contains DNP is to be expected in light of its intimate association with the chromosome, and this is verified by the fact that this region stains positive for DNA after Aggarwal’s (1976) Feulgen reaction (Ris and Witt, 1981). As noted above:, the h e r plate may also contain tubulin (Fig. 19; see however DeBrabander er al., 1979a) and an uncharacterized (protein) component which is not tubulin (Brenner et al., 1981). Lyubskii et al. ( 1979) used eletron microscopic cytochemical methods to localize Ca-ATPase activity in the mitotic spindle of Chinese hamster cells. ‘They found what they considered to be a significant amount of Ca-ATPase activity associated with metaphase and anaphase kinetochores but not with early prometaphase or prometaphase-blocked (colcemid) kinetochores. This increased ATPase activity suggested to them that the kinetochore plays an active role in generating the force for chromosome movement (see also Mota, 1957; Margolis and Wilson, 1981). Although these results are potentially significant they are not adequately documented and will require further investigation. Sanger (1975) incubated mitotic PtKz cells in fluorescently labeled heavy meromyosin and noted “a small dot of fluorescence in the kinetochore region where the fibers meet the chromosome.” This suggested to him that actin was a component of the kinetochore. Similar studies using fluorescently tagged and purified heavy meromyosin (Herman and Pollard, 1978; Aubin e t a / . , 1979) or antibodies against actin (Cande et al., 1977) have since failed to corroboratt: this claim. Thus, at present, there is no convincing evidence to support the claim that actin is a component of the kinetochore (see also Section VI,B). In summary the mitotic kinetochore in mammals appears to be formed, in part, from components (DNA, RNA, protein) present at the primary constriction prior to NE breakdown. The formation of the outer disk and corona occurs during Iearly prometaphase after NE breakdown and before MTs become associated with the

THE MAMMALIAN KINETOCHORE

27

kinetochore. Some components of the inner layer are also associated with the chromosome prior to attachment but others become associated with this region during or after attachment. Based on what is currently known about the formation and cytochemistry of the kinetochore one can speculate that the differentiation of the prophase “ball” into the prometaphase disk is triggered by the breakdown of the NE which allows yet to be characterized cytoplasmic factors to associate with the kinetochore. These factors then interact with components already present at the kinetochore to promote disk formation. Further differentiation of this structure into a trilaminar disk most likely results from the continued condensation of the chromosome and from the acquisition of MTs by the kinetochore.

V. The Kinetochore Organizer The ultrastructural similarities between the mammalian kinetochore and other chromosomal structures formed as a result of genetic activity prompted Brinkley and Stubblefield (1970) to consider the kinetochore as a specialized “gene.” A similar theme was also espoused by Luykx (1970; see also DuPraw, 1968; Ruthman and Permantier, 1973) who felt that the kinetochore was formed partly as a product of genetic activity within the chromatin subjacent to the outer plate (i.e., within the inner plate). He viewed this chromatin as a “kinetochore organizer,’’ analogous in many respects to the nucleolar organizer (as it was understood at the time). Both Luykx ( 1970) and Brinkley and Stubblefield (1970) cited early cytological evidence suggesting the presence of genes in the kinetochore or kinetochore region (i.e., pericentromeric heterochromatin), and Luykx even calculated that there was sufficient DNA in this region to code for 1000 different proteins. The putative presence of RNP within the mitotic kinetochore does lend credence to this notion. Unfortunately, there is currently no compelling reason to suspect that the pericentromeric heterochromatin or the DNA associated with the kinetochore codes for proteins (e.g., Fuge, 1977), and this hypothesis must be considered, in the words of Bostock and Sumner (1978), “as entirely speculative.” On the other hand, Nicklas (1971) noted that the minimal function of the kinetochore is to bind spindle protein subunits (i.e., tubulin or MTs). Thus, if one assumes that “DNA is the basis of linear continuity in chromosomes and is ultimately responsible for the specificity of . . . the kinetochore” then “the primary role of kinetochoric DNA may be the binding of particular proteins.” Although Nicklas’ concept pertained to how the kinetochore functioned, it may be equally applicable for how the kinetochore forms. In regard to this, Holmquist and Dancis (1979) have suggested that the kinetochore DNA acts as a kinetochore organizer not by coding for plate proteins, but by organizing these proteins into a functional structure upon activation (i.e., after NE breakdown).

28

CONLY L. RIEDER

Similarly, any kinetochore-associated RNA may be transcribed at loci distal to the kinetochore and then become localized at this structure (much like the nucleolar organizer). The finding that an identical protein component is associated with both interphase prekinetochores and with the outer plate of mitotic kinetochores supports the concept of a kinetochore organizer as presented by Holmquist and Dancis (1979)-it is likely that the prekinetochores (see Section IV,A) arise from the binding of protein(s) to a specific segment of DNA (the kinetochore organizer) during or after S phase, and that this DNA-protein complex then interacts with additional cytoplasmic factors after NE breakdown to form the “backbone” of the mitotic kinetochore. Such a concept is also strengthened by the recent evidence (see Sections IV,C and D) which indicates that chromatin is a major component of the mitotic kinetochore and that this chromatin differs in its structure from that of typical chromatin (Ris and Witt, 1981). That the integrity of a kinetochore component(s) is necessary for the subsequent formation of a functional kinetochore on a newly replicated chromatid is implicit in the findings of Brenner er af. (1980). These investigators (see also McNeil and Berns, 1981) selectively destroyed the outer plate (as indicated by EM of similarly irradiated chromosomes) of both kinetochores of a prometaphase PtK, chromosome by laser microirradiation, and then followed the fate of the chromatids. After irradiation the replicated chromosome detached from the spindle and the chromatids separated but showed no directed movement for the duration of mitosis. When the cell next divided the two irradiated and now replicated chromosomes condensed but neither attached to the spindle or underwent poleward anaphase movement even though both separated normally into two chromatids during anaphase. Apparently in each case damage to the original kinetochore inhibited the subsequent formation of a functional kinetochore on the newly replicated chromatid. In this respect it is tempting to suggest that damage to the original kinetochore organizing DNA (i.e., within the laser-irradiated kinetochore) was passed on to the daughter chromatid during semiconservative DNA replication (e.g., see Mazia, 1977) and that this in turn prohibited the organization of functional kinetochores on both chromatids. A report by Hsu et al. (1975) deserves special consideration when discussing the concept of a kinetochore organizer. These authors argue that “In the process of centromeric or telomeric fusion or of fusion of a centromere with a telomere, centromeric inactivation may. occur, thus preserving both centromeres-one functional, the other latent-in the resultant translocation chromosome.“ The concept of a latent kinetochore organizer then explains how additional functional kinetochores arise during the reverse process of chromosomal fragmentation or fission. Holmquist and Dancis (1979) expanded on this theme and postulated that ‘‘a telocentric chromosome could have a terminal centromere (kinetochore plate overlapping the telomere) and a telomere that functions for terminal replication

THE MAMMALIAN KINETOCHORE

29

without having the kinetochore organizer as a terminal sequence.” Then, when telocentrics fuse the two kinetochore organizers would be brought close enough together so that the plates they organize overlap. This stable arrangement would explain why the centric region of Robertsonian metacentrics frequently appears doubled (and organizes twice as many MTs as rod chromosomes-see Section Vt,B) and why “a transverse break in the kinetochore region can produce two chromosome fragments both of which are capable of normal behavior during cell division” (Luykx, 1970). (This latter point was used by Luykx as evidence for many gene copies in the kinetochore region.) As previously noted, Nicklas ( 197 1 ) suggested that the kinetochore functions, during mitosis, by simply binding spindle subunits to its DNA, and he posed three questions for the future concerning the finer points of his hypothesis. These included: ( 1 ) where (in the mammalian kinetochore) is the DNA located, (2) how does this location relate to the binding sites of spindle subunits, and (3) is binding mediated by an adapter protein between DNA and tubulin/MTs? The current evidence indicates that kinetochore MTs are bound to sites on the outer kinetochore plate and that this plate is composed of DNA and associated proteins (but probably not RNA). In this respect, our understanding of the formation, structure, and chemical composition of the outer plate lends strong support to the notion of Nicklas (197 I ) , that the kinetochore is a specialized gene whose primary function involves protein binding. The question of whether the binding of MTs to DNA in the outer kinetochore plate is mediated by adapter protein(s) (3 above) remains unanswered. Yet it is reasonable to suspect that any DNA within this structure must be complexed to protein (to form a modified chromatin) and that some of these proteins are likely to be involved in binding MTs to the kinetochore. There is biochemical evidence which indicates that MAPS bind preferentially to satellite DNA sequences (Wiche et al., 1978) and that this mediates the binding of tubulin or MTs to the centromere region of the chromosome (Corces et al., 1978). However, this hypothesis for how MTs bind to the kinetochore lacks additional support since ( I ) there is no evidence to indicate that satellite DNA sequences are components of the kinetochore outer disk, (2) not all mammals contain detectable amounts of satellite DNA in their centromeres (see Section II), and (3) there is no evidence indicating that any one MAP is associated with the kinetochore although there is evidence to the contrary (DeBrabander et al., 1981b).

VI. The Mammalian Kinetochore Fiber One of the most important aspects of mitosis is the molecular mechanism by which a kinetochore on a chromosome attaches to and orients toward the pole of the forming spindle. Elegant chromosome micromanipulation experiments have clearly demonstrated that this attachment arises from the formation of a

30

CONLY L. RIEDER

birefringent fiber (i.e., a K-Fiber) which connects the kinetochore with the polar region (Begg and Ellis, 1979a,b) and that this fiber is more firmly anchored in the polar region than along the rest of its length. Since most of the K-Fiber birefringence arises from MTs within the fiber (Sato et ul., 1975), kinetochore attachment unquestionably involves the formation and/or accumulation of MTs at the kinetochore. The acquisition of MTs by the prometaphase kinetochore also appears to be a prerequisite for directed chromosome movement. The exact role K-MTs play in this process is, however, a subject of considerable controversy. It is unclear whether they directly generate and/or transmit the mitotic forces (as suggested by Inout, 1964; Dietz, 1972; Bajer, 1973; Margolis et al., 1978) or whether they act simply as a velocity governor to regulate the speed of movement produced by a mechanistically separate force generator (Forer, 1974; Forer et a / ., 1979; Nicklas, 1975,1977a, 1979; Sanger, 1977; Sanger and Sanger, 1979). Since the major goal of mitosis research is to understand how the kinetochore attaches to iind is transported on the spindle, elucidating the structure of the K-Fiber, and the mechanism by which the kinetochore acquires MTs, have always been important considerations. A N D EARLY ANAPHASE A. GENERAL PROPERTIES DURING METAPHASE

By micromanipulation, an early anaphase chromosome (in grasshopper or crane fly spermatocytes) can be displaced toward the near pole without becoming detached from that pole, yet it resists displacement away from that pole or beyond the spindle periphery (Nicklas, 1967; Begg and Ellis, 1979a,b). Since each chromosome is connected to a spindle pole by a K-Fiber, one can conclude from these observations that K-Fibers are flexible but not very extendable (i.e., they cannot be stretched). Furthermore, the fully formed K-Fiber is not {easily broken or detached from the kinetochore since a lateral displacement of the fiber (by micromanipulation) simply changes the position of and/or stretches the chromosome (Begg and Ellis, 1979a). Similar micromanipulation experiments have yet to be conducted on mammalian cells (due to their small size) but it is reasonable to expect that the salient conclusions drawn from grasshopper and crane fly spermatocytes are, for the most part, applicable to the K-Fibers in vertebrates. Additional support for the apparent strength of the K-Fiber arises from the observation that metaphase kinetochores in crane fly spermatocytes “cannot be displaced > 1 pm toward either spindle pole, even by a force which is sufficient to displace the entire spindle within the cell” (Begg and Ellis, 1979b). The resistance of K-Fibers to breakage by micromanipulation runs contrary to the earlier belief of Ostergren (see also Inout, 1964; Dietz, 1972) that K-Fibers are “tactoid” (i.e., liquid crystal) in nature and “permit the body of one chromo-

THE MAMMALIAN KINETOCHORE

31

some to pass through the traction fibers of another” (Ostergren et al.. 1960). Although the chromosome to pole connection can be disassembled and reassembled under certain conditions (e.g., during chromosome reorientation), Ostergren’s conclusion concerning the nature (or strength) of the K-Fiber was based on the difSu.se K-Fibers of Luzula and may not be applicable to other cell types. K-Fibers in living Haemanthus endosperm as seen by high resolution differential interference contrast microscopy appear to be covered with “beads” which move along the K-Fiber toward the pole “at approximately the velocity at which the chromosome will be transported” (Allen et al., 1969; also Hard and Allen, 1977). Although it is currently unknown whether the migration of such “particles and states” occurs within the K-Fibers of mammals, it does occur in the KFibers of other vertebrates (R. Hard, personal communication) and is likely to be a general transport feature of K-Fibers. This transport phenemenon may not be restricted only to K-Fibers: particles and akinetochoric chromosome fragments show a similar poleward migration when placed between the chromosomes and poles in plant (Ostergren et al., 1960) and animal (Nicklas and Koch, 1972) metaphase spindles. The nature of this force, which is responsible for the poleward movement of chromosomes and other bodies within the spindle, remains the most important and least understood question in the field of mitotic research. B. STRUCTURE A N D COMPOSITION An analysis of serial sections (e.g., Brinkley and Cartwright, 1971; Witt et d.,1981; Rieder, 1981) and stereo viewing of thick sections (McIntosh et al., 1975a,b) reveals that the fully formed metaphase mammalian K-Fiber is composed of K-MTs (i.e., those which terminate in the outer kinetochore disk), which intermingle with numerous polar and free MTs (see McIntosh et al., 1975a. for terminology) along the length of the fiber. A major assumption in many models of force production is that individual K-MTs run uninterrupted between the pole and the kinetochore. The validity of this assumption was recently questioned by Fuge (1977), who measured the lengths of individual KMTs in crane fly spermatocytes and concluded that “a direct continuity of KMTs between kinetochores and polar regions in higher eukaryotes . . . is not essential. Although this continuity may not be essential for force production, a reconstruction of metaphase K-Fibers in PtK, (Rieder, 1981) and in CHO (Witt et al., 198 I ) cells reveals that the majority of K-MTs in mammals do in fact run continuously between the kinetochore and the polar region (see also Jensen, 1982). Some of the polar and free MTs of the K-Fiber penetrate the chromosome very near the kinetochore (e.g., Jokelainen, 1967; Commings and Okada, 197 I ; Brinkley and Cartwright, 1971; Roos, 1973b; Fuge, 1977; Figs. 14 and 15). The presence of these MTs near the kinetochore, along with the difficulties encoun”

32

CONLY L. RIEDER

tered in determining MT endpoints with certainty in random longitudinal nonserial sections, may have led some early investigators (e.g., Jokelainen, 1967; Luykx, 1970; Commings and Okada, 1971) to conclude that K-MTs in mammalian cells terminate at different positions within the kinetochore. However, recent serial section studies convincingly demonstrate that K-MTs terminate directly within the outer plate of the mammalian kinetochore (Brinkley and Cartwright, 1971; Roos, 1973b, 1977; Pepper and Brinkley, 1977; Rieder, 1981). As a rule K-MTs, as well as other spindle MTs, show a center-to-center spacing of at least 50-60 nm in untreated cells (Brinkley and Cartwright, 1971; Rieder and Bajer, 1977a; review of Bajer and Mole-Bajer, 1972). This minimal spacing correlates with the presence of a “sleeve element” (Stebbings and Benett, 1975), which is seen to surround the walls of MTs when viewed in cross sections. Although this component usually appears electron translucent after conventional staining (hence the older and somewhat misleading terms of Mt “clear zone” or MT “halo”) it can be made visible with several cationic stains including lanthanum hydroxide, ruthenium red, alcian blue, and dialized iron (see references in Stebbings and Benett, 1975; Hyams and Stebbings, 1979). The fact that MTs stained to reveal the sleeve element have a total diameter similar to that of unstained freeze-etched MTs suggests that this structure contains components not normally made visible by uranyl and lead stains (Stebbings and Benett, 1975). There is some evidence to suggest that the sleeve element-MT association is sensitive to thermal disruption (i.e., it is labile), since K-Fiber MTs undergo a dramatic reduction in their minimal center-to-center spacing (i.e. 2 60 vs 30 nm) in response to heating (Rieder and Bajer, 1977a) and sometimes even to cooling (Rieder, 1981). Although the exact composition of the sleeve element and its function remains to be resolved, it is thought to contain high-molecularweight MAPS (Amos, 1977; Valee and Borisy, 1977; DeBrabander et al.. 1981b) and/or glycoproteins (Stebbings and Benett, 1975; Behnke, 1975). The number of MTs associated with a metaphase kinetochore differs conaiderably between different organisms (see reviews of Bajer and Mole-Bajer, 1972; Fuge, 1977), and to a lesser extent between chromosomes within the same cell. For example, the kinetochores in fetal rat cells possess between 4 and 6 MTs each (N=19; Jokelainen,l967), those in CHO between 9 and 16 MTs each (N=23; Witt et al., 1981), and those in PtK, between 21 and 41 MTs each (N=19; Brinkley and Cartwright, 1971; Mclntosh et al., 1975a). Thus, the kinetochores of different mammals may show up to a 10-fold difference in the number of associated MTs (e.g., fetal rat versus PtK,), whereas in a cell type from any one organism this difference is approximately 2-fold at the most. Fuge (1977; see also Jensen, 1982) conducted a brief survey of chromosome size versus the number of K-MTs in higher eukaryotes and concluded that “it seems probable that the number of K-MTs is directly correlated with chromo-

33

THE MAMMALIAN KINETOCHORE

some size.” Similarly, Heath (1979) noted a general correlation in lower eukaryotes between the number of MTs per kinetochore and the apparent size of the chromosome moved. Unfortunately, there appears to be no correlation between the size of a chromosome and the number of MTs associated with its kinetochore in those few higher eukaryotes where it has been carefully examined. For example, Moens (1979) recently studied the number of K-MTs associated with the chromosomes in three species of grasshoppers and noted that, although there were significant differences between species, there was clearly “no strict proportionality between chromosome size and K-MT numbers.” Similarly, Lin et a / . (1981) studied K-MT numbers during the first and second meiotic division in male Drosophila and found “that the tiny acrocentric chromosome number 4 contains no fewer kinetochore microtubules than the largest metacentric chromosome despite an approximate tenfold volume difference. A strict correlation between chromosome size and the number of K-MTs also appears to be lacking in PtK, (and probably in other mammals) since the number of MTs reported to be associated with kinetochores in PtK, ranges only 2-fold (see above) despite an approximate 5-fold difference in size (from Levan et a / . . 1966) between the largest and smallest chromosomes. If there is no strict proportionality between chromosome size and the number of K-MTs, then how is this number established‘?Moens (1979) noted that the most pronounced variation in K-MT numbers in grasshoppers is between species and between the telocentric and the Robertsonian fusion chromosomes (the kinetochores on Robertsonian fusion chromosomes were found to possess twice as many MTs as those on telocentrics). This suggested to him that “the evolutionary history of the species and its karyotype is a significant determinant of K-MT numbers. Although the evolutionary history of an organism may influence the number of MTs associated with a kinetochore, there has yet to be a molecular explanation for how this number is determined. In this respect, an analysis of cross sections of kinetochores from untreated (Roos, 1973a) and cold-treated (Rieder, 198 I ; Fig. 20A) metaphase PtK, cells suggests that the number of MTs associated with the kinetochore may be limited to the number of MTs, spaced approximately 60 nm apart (see above), which completely saturate the surface area of the outer disk (see also Figs. 12-14 in Jokelainen, 1967; Fig. 2a in Witt e t a / . . 1981). This impression appears to be validated when one calculates the approximate surface area of kinetochores from a variety of mammalian cells and compares this area to the average number of MTs associated with the kinetochores. The outer plate in fetal rat cells is about 0.25 km and it possesses approximately 6 MTs (Jokelainen, 1967); in CHO cells it averages about 0.35 pm and possesses approximately 14 MTs (Witt et al., 1981); while in PtK, it averages 0.5 krn and possesses an average of 25 MTs (Brinkley and Cartwright, 1971; Roos, 1973a). The ratio of the average number of K-MTs per kinetochore between these organ”

”

34

CONLY L. RIEDER

Fia. 20. (A and B) Cross sections, 0.25 pm thick, of different cold-stable K-Fibers in a metaphase PtK, cell. The section in (A)Justgrazes a kinetochore. Note that the MTs appear fairly evenly spaced and maintain a minimum center-to-center spacing of approximately 60 nm. The section in (B) is 1.0 p,m distal to the kinetochore. Note that the regularly packed but well-separated MTs are surrounded by an ill-defined matrix. X60.000. (Fig. 2OA from Rieder. 1981.)

isms ( 1:2.3:4.2) compares favorably with the ratio between the average surface area of their outer disks (this ratio, calculated from the average diameter and assuming a uniformly circular outer plate, is approximately 1 :2.1:4.4). This analysis indicates that the surface area of the kinetochore is important in determining the number of MTs which can associate with it. This concept provides a simple explanation as to why sister kinetochores on the same metaphase chromosome in PtK, possess similar numbers of MTs (McIntosh et a f . , 1975a; Section V), and why the number of MTs associated with a metaphase kinetochore does not change appreciably after lysis into buffers containing polymerization cornpetent tubulin (Snyder and McIntosh, 1975). In addition, Robertsonian fusion chromosomes in the grasshopper would be expected to carry twice as many KMTs as telocentrics since these chromosomes most likely possess two (Moens, 1978) closely associated kinetochores (and hence have twice the surface area). Finally, a correlation between kinetochore surface area and the number of KMTs may also exist in other types of cells. For example, the fact that the surface area of a Haemanthus kinetochore, which consists of a 0.5-0.6 pm bail, is much greater than the surface area of a 0.5-pm-diameter disk (as in PtK,) may explain why these kinetochores possess a corresponding greater number (i.e., up to 100) MTs . K-MTs are stable to many treatments which disrupt the majority of free and polar spindle MTs (see references in Rieder and Bajer, 1977a; Salmon and Begg, 1980; Rieder, 1981; Witt et a / . 1981; Figs. 18 and 20-22). This “differential stability” is not acquired during metaphase but arises during prometaphase as the kinetochore becomes attached to the forming spindle (Rieder, 1981; Fig.22). In all cases it can be correlated with a change in K-Fiber geometry; after treatment

THE MAMMALIAN KINETOCHORE

35

K-MTs appear clustered into discrete bundles of evenly spaced MTs (Fig. 20B) while untreated K-MTs show a less precise spacing and a greater divergence, especially distal to the kinetochore. As previously mentioned, K-MTs in untreated cells intermingle with numerous free and polar MTs. During cold treatment some of these MTs become incorporated into the bundles of cold-stable K-MTs (Rieder. 198 I ). Similar observations were made by Rieder and Bajer (1977a) with regard to the heat-

FIG. 21. Longitudinal thick section (0.25 pn) through a metaphase PtK, cell cooled to 6 4 ° C for 6 hours prior to fixation. Note the bundles of cold-stable K-Fiber MTs. See text for details. x7700. (From Rieder, 1981.) FIG. 22. (A-D) Sections I , 2. 4, and 5 of a serial series through a monooriented chromosome from a prometaphase PtK, cell cooled to 6-8°C for 4 hours. Cold-stable MTs are associated with the kinetochore pictured in (A) and ( B ) while the sister kinetochore (arrows in C and D) is free of associated MTs. See text for details. x21.100.

36

CONLY L. RIEDER

stable K-MTs of newt cells. The differential stability of K-Fiber MTs is therefore not due to the selective incorporation of a stabilizing factor into only K-MTs, or to an absence of “free ends” on these MTs (as suggested by Salmon et a/., 1976; Salmon and Begg, 1980; Kirschner, 1980), since free and polar MTs within the bundle are stable to the treatment. Rather, the observed cold stability of free and polar MTs within K-Fibers indicates that stability is conferred to a MT, regardless of its origin or termination site, as long as it is in the vicinity of or present within a K-Fiber. With this consideration in mind the most viable hypotheses for the increased stability of K-Fiber MTs. relative to other spindle MTs, include a greater degree of linkage between neighboring K-Fiber MTs (Brinkley and CartWright, 1975; Witt er al., 198I ) and/or to the association of additional proteins, found predominantly within K-Fibers, with K-Fiber MTs (e.g., Webb and Wilson, 1980; Margolis and Rauch, 1981; Job er a/., 1982). The MTs of the cold stable K-Fiber in PtK, appear to be embedded in an illdefined electron opaque matrix (Brinkley and Cartwright, 1975) which is particularly evident in thick sections of tannic acid-treated cells (Rieder, 1981; Fig. 20B). A similar material appears to permeate the K-Fiber in untrcured Odegonium after fixation and tannic acid treatment (Schibler and Pickett-Heaps, 1980), and it has been speculated that this material represents a second component in the traction apparatus of the K-Fiber. A careful examination of the micrographs presented in these studies (also Rieder, unpublished) reveals that the MTs are not really “embedded” in this material, but that it is associated with the wall ofeach MT. This is particularly evident in sections distal to the kinetochore in which the K-Fiber MTs are less regularly arranged (e.g., Fig. 10D of Schibler and Pickett-Heaps, 1980). The location of this material suggests that it represents components of the sleeve element (see above) which stain moderately electron opaque after the tannic acid treatment. In this respect, indirect immunofluorescent (Bulinski and Borisy , 1980) and immunoelectron microscopic (DeBrabander et al., 1981b) studies do indicate that MAPS (a suspected sleeve coniponent) are associated with the spindle, especially with K-Fiber MTs (Sherline and Schiavone, 1978). It has been suggested that these proteins form periodic sidearms along the surface of the K-Fiber MTs (e.g., Witt et d., 1981). Many investigators have noted the presence of arms (i.e., cross-bridges) which connect adjacent K-Fiber MTs (e.g., references in McIntosh et af., 1975a; Fuge, 1!)77; Witt et a f . , 1981), but it remains to be demonstrated that these structures are analogous to the MAPS detected at the LM level. Light and electron microscopic studies indicate that the K-Fiber contains other components in addition to MTs and MAPS. Welsh er a f . (1979) found calmodulin within the metaphase PtK, spindle using immunofluorescent methods and noted that it was associated primarily with K-Fibers (Fig. 23). This result has been confirmed at the EM level by DeMey er al. (1980), who noted that calmodulin in PtK, was distributed primarily within K-Fibers where the MTs “show a

THE MAMMALIAN KINETOCHORE

37

high density and/or extensive lateral interactions. Immunofluorescent methods also reveal that the sea urchin spindle contains an associated dynein-like component in siru (Mohri et al., 1976), and biochemical studies indicate that a similar component remains tightly associated with the isolated sea urchin mitotic apparatus (Pratt et a/., 1980). Although the exact location and identification of this ATPase are unknown, they are suspected to be associated with K-Fibers since anaphase chromosome movement in model (i.e., lysed or glycerinated) cells is inhibited both by antidynein (Sakai er al., 1976; Sakai, 1978) and by vanadate (an inhibitor of ciliary dynein-see Cande and Wolniak, 1978). There remains a considerable amount of controversy over whether actin is associated with K-Fibers. Sanger ( 1975) stained mitotic PtK, cells with fluorescent-labeled heavy meromyosin, and found that actin was present within the ”

FIG. 23. Indirect immunofluorescent localization of tubulin ( A ) and calmodulin (B) in metaphase PtK, cells cooled to 0°C for l hour. Both tubulin and calmodulin appear to be concentrated in the cold-stable kinetochore fibers (cf.Figs. 20-22). X 1600. (From Welsh rt a / . . 1979.) FIG. 24. A thin section through both kinetochores of a metaphase chromosome in Oedogonirrm. Numerous thin filaments appear to be associated with the kinetochore fiber. See text for details. X46.000. (Courtesy of M. M. Schibler.)

38

CONLY L. RIEDER

spindle and that it was “confined to the fibers that connect the chromosomes with the centriolar region.” Similar results were obtained by Cande et a / . ( 1977; see also Schloss e t a l . , 1977) using rabbit antibodies against actin. Yet, Herman and Pollard (1978) repeated Sanger’s original study with some modifications in technique, and found that the fluorescent pattern usually appeared to be spread diffusely within the spindle. Similarly, in a more detailed study Aubin ct nl. ( 1979; see also Herman et a/., 1980; Barak et al., 198 1 ) used fluorescent tagged heavy meromyosin and two different antibodies against actin and “found no evidence for increased accumulations of actin in the mitotic” PtK2 spindle. These conflicting LM results are similar to those obtained by EM. Actin filaments are frequently seen in the spindle region and K-Fibers after glycerination and stabilization with heavy meromyosin (e.g., Forer and Behnke, 1972; Gawadi, 1974; Forer and Jackson, 1975; Schloss et nl., 1977; Forer ct al., 1979). However, with few exceptions (e.g., Bajer and Mole-Bajer, 1969; Muller, 1972; Euteneuer e t a l . , 1977) actin-like filaments are not detected after conventional fixation and staining (see LaFountain, 1975; Fuge, 1981). F-actin, if it is present within K-Fibers, may be labile to the extensive osmication procedures generally used for the fixation of whole cells (Maupin-Szamier and Pollard, 1978). In fact, recently improved fixation procedures employing a tannic acid/glutaraldehyde mixture and brief osmium treatment have revealed actinlike filaments associated with Odegonium K-Fibers (Schibler and Pickett-Heaps, 1980; Fig. 24), but it remains to be demonstrated whether these filaments are actin and/or whether they are present in the K-Fibers of other types of cells (i.e., mammals). Hepler (1980; see also Porter and Machado, 1960) has recently shown that a membrane system is closely associated with the K-Fiber in barley cells, from prometaphase through mid-anaphase, and that this system extends from the pole to the point of chromosome attachment. The similarities between the membrane system associated with the spindle and that of the sarcoplasmic reticulum of muscle prompted him to speculate that the mitotic membranes “might contain a Ca-ATPase and be capable of alternately releasing and sequestering Ca++ , thus controlling ion concentrations even in local regions, for example along kinetochore fibers.” Moll and Paweletz (1981) and Paweletz and Finze (see Paweletz, 1981) found a similar but much less developed membrane system associated with the mitotic apparatus of HeLa cells, and also suggested that it was involved in Ca2+ regulation. In summary the K-Fiber is a multicomponent system composed of membranes and various proteins which are closely associated with the kinetochore and nonkinetochore MTs of the fiber. The number of MTs associated with a kinetochore in mammals is not proportional to the size of the chromosome, but it may be related to the surface area of the kinetochore. The increased stability of K-Fiber MTs, relative to the other spindle MTs, is probably due to the ability of these

THE MAMMALIAN KINETOCHORE

39

MTs to associate laterally with each other and from the association of additional (protein) components, found predominantly within the K-Fiber, with K-Fiber MTs. The possible association of a dynein-like ATPase with K-Fiber MTs will undoubtedly stimulate additional research in this area, especially in light of the capacity of dynein to generate movement between neighboring MTs. In addition, it will be of future importance to determine whether the MTs of the forming KFiber orient the membranes and additional protein components (which may be initially concentrated in the polar regions and then transported toward the chromosomes during K-Fiber formation), or whether these components themselves determine the directionality of MT formation. Finally, future improvements in fixation, immunocytochemistry, and microinjection techniques will undoubtedly increase our understanding of the composition of, and the dynamic changes which occur within, the K-Fiber during its formation and function.

C. KINETOCOHORE MICROTUBULE FORMATION 1 . The Centrosome-Kinetochore Interaction One of the most vivid demonstrations of K-Fiber formation is seen during the astral divisions of vertebrate epithelial cells (e.g., newt, frog, PtK) in monolayer tissue culture. These cells remain flat throughout mitosis which makes them the material of choice for high resolution correlative light and EM studies of animal cell mitosis. The attachment of kinetochores to the spindle has been best documented in the rat kangaroo lines PtK, and PtK, (Heneen, 1970; Brinkley and Stubblefield, 1970; Roos. 1973a,b, 1976; Rieder and Borisy, 1981). The initial behavior of a chromosome in this material is determined primarily by the position of its kinetochore relative to the spindle pole at the time of NE breakdown. As a rule, when prometaphase is initiated, chromosomes whose kinetochore region is in the vicinity of a pole orientfirst. The prevalen! behavior pattern of these chromosomes is to intimately associate with the nearest pole. “lf the kinetochore region is very close to the pole this association occurs without any measurable or even perceptible motion, but it is nevertheless recognizable by a poleward bend in the kinetochore region. Chromosomes whose kinetochore region is less close to a pole become associated with it as a result of a distinct poleward movement” (Roos, 1976). The movement of these kinetochores begins as a sudden jerk and the maximum poleward velocity is reached almost immediately (Roos, 1976; Mole-Bajer et al., 1975). An ultrastructural feature of these “centrophilic” (see Zirkle, 1970) or “poleassociated” (Roos, 1976) chromosomes is the presence of MTs only on the kinetochore closest to and facing the polar region (i.e., they are truly monooriented). The sister kinetochore, which sits less than 0.5 pm away and faces in the opposite direction, invariably possesses no MTs (Jokelainen, 1965; Roos, 1973a,1976; Mole-Bajer eta/., 1975; Rieder and Borisy, 1981; Figs. 16 and 2 2 ) .

40

CONLY L. RIEDER

Thus, there is little doubt that a centrophilic chromosome moves toward the centrosome at the beginning of prometaphase due to the initial formation of a fiber only on that kinetochore which faces the pole. (In this respect these chromosomes are analogous to anaphase chromosomes: in both cases movement to the proximal pole can be correlated with the presence of only one K-Fiber on the chromosome .) A similar “unitelic” monoorientation (Roos, 1976) of chromosomes can be induced in a variety of plant and animal cells by treatment with low concentrations of mitotic inhibitors. For example, when mammalain cells enter prometaphase under the influence of 0.06 kg/ml of colcemid all the chromosomes within the cell are transported to, and become positioned in a sphere around, the unseparated centrosomes (Brinkley et al., 1967; Brinkley and Stubblefield, 1970; Fig. 25). The chromosomes in these “chromosomal spheres” resemble naturally occurring centrophilic chromosomes (see above) in that the kinetochore facing the single polar area possesses MTs while its sister, which sits less than 0.5 Frn away and faces in the opposite direction, appears similar in structure but possesses no MTs (Brinkley et al., 1967; Brinkley and Stubblefield, 1970; see also McGill and Brinkley, 1975; Barham and Brinkley, 1976; Mazia et al., 1981 ; Fig. 26). The LM observation, that those chromosomes closest to the polar area al. the time of NE breakdown are the first to attach to the forming astral spindle (Roos. 1976; Izutsu et al., 1977; Rieder and Bowser, unpublished), has been confirmed at the ultrastructural level since those kinetochores closest to and facing the polar areas at the time of NE breakdown are the first to acquire MTs (Roos, 1973a,b; Paweletz, 1974; Mole-Bajer, 1975; Rieder and Borisy, 1981). This “proximity effect” (Rieder and Borisy, 1981) is also seen during the formation of chromosomal spheres (Rieder and Bowser, unpublished). It was attributed by Roos (1976) to the fact that NE breakdown is initiated in the area of the centrosomes. However, in a recent study Rieder and Borisy (1981) allowed the NE to breakdown in PtK, cells at a temperature (6°C) which inhibited spindle formation and then used correlative light and EM methods to study the formation of the spindle during recovery at 18°C. They found that those kinetochores closest to and facing the centrosomes were always the first to acquire MTs, and that these MTs were oriented toward the centrosome. Since the spindles in these cold-treated cells were allowed to form well after the breakdown of the NE, the observed proximity effect could not be attributed to influences arising from the asynchronous breakdown of the NE (as suggested by Roos, 1976). Rather, the data of Rieder and Borisy ( 198I ) clearly demonstrate that a kinetochore-centrosome interaction occurs during spindle formation which cannot be attributed to transient influences. The literature on the ultrastructure of early prometaphase in untreated cells with asters indicates that the first MTs associated with a kinetochore are invari-

THE MAMMALIAN KINETOCHORE

41

Fic. 25. This PtK, cell was treated with 0.05 p g h l of colcemid prior to NE breakdown in (A). The cell subsequently (B and C) formed a chromosomal sphere. Time in lower right hand comer = hours/minute. See text for details. X 1250. Fio. 26. Thin section through a chromosomal sphere in a PtK, cell similar to that pictured in Fig. 2SC. The chromosomes are grouped around the centrioles and only those kinetochores facing the centrosome area possess MTs. Some astral MTs (e.g.. A) can be seen to penetrate the sphere without terminating on a kinetochore. See text for details. X33.200. (From Bajer and Mole-Bajer. 1972.)

ably oriented toward a polar area, regardless of the orientation of the kinetochore plate (i.e., the fiber is not necessarily formed perpendicular to the long axis of the kinetochore-see Luykx, 1970; Tippit et a/., 1980; Rieder and Borisy, 1981). Similarly, as a micromanipulated chromosome reattaches to the grasshopper spermatocyte spindle, long MTs already oriented toward the polar region rapidly reappeared at the kinetochore (Nicklas er a/., 1979). DeBrabander et a/. (1980,1981a) also stress that bundles of K-MTs form preferentially, even during

42

CONLY L. RIEDER

recovery from Nocodazole treatment, between the centrosomes and kinetochores in PtK, cells. It can be concluded from these recent studies that K-Fiber formation is normally directed initially toward the polar regions and not toward random points within the cell. The proximity effect, the initial unitelic monoorientation of prometaphase chromosomes and the tendency of K-Fibers to form toward the polar regions regardless of the initial orientation of the kinetochore, clearly demonstrates that K-Fiber formation is initiated as a result of an interaction between the centrosomes (i.e., polar regions) and the kinetochores (see also Brinkley and Stubblefield, 1970). Any mechanism of K-Fiber formation must therefore explain the nature of this kinetochore-centrosome interaction. 2. Models of K-MT Formation An interaction between the polar area and a prometaphase kinetochore, which results in the accumulation of MTs at the kinetochore and which is consistent with the above observations, has been postulated to occur by the following mechanisms. a. An Interaction as a Result of a Transient Diffusion Gradient (Roos, 1976; see also discussion in Nicklas, 1977a). This model, which attemprs to explain the initial behavior of PtK, chromosomes during spindle formation, is based on correlative LM and EM observations of untreated cells. It assumes that MT subunits are concentrated in the polar regions during late prophase and that the initial breakdown of the NE in the polar areas allows for the diffusion of MT subunits into the nucleus “so that two gradients of decreasing concentration of subunits initially exist from the polar areas to the center of the nucleus.” ,4s a result of this gradient, K-MT formation (by nucleation from the kinetochores) would be directed toward a polar area and would occur first on those kinetochores closest to and facing a pole. The formation of a fiber on the kinetochore most directly fixing a pole would subsequently result in translocation of the chromosome to the pole because of the unidirectional nature of the applied force. b. An Interaction to Form a Polymerization Gradient (DeBrabander et a l . , 1979b, 1980). DeBrabander and co-workers used a PAP-immunocytochernical LM-EM approach to study spindle reformation in PtK, cells during reversion from Nocodazole treatment. They found short MTs associated with the kinetochore corona during the initial stages of recovery (see also Witt et al., 1980; Ris and Witt, 1981; Fig. 27). These short MTs then became attached to the kinetochore plate and elongated preferentially along the axis between the kinetochore and centrosome. DeBrabander et a f . speculate that the kinetochores and centrosomes act “at least partly, by creating a gradient in which tubulin assembly is favored” (DeBrabander et a l . , 1979b). In this manner the preferential nucleation and elongation of MTs would be expected to occur between two nucleating sites

THE MAMMALIAN KINETOCHORE

43

FIG. 27. Stereo photographs of a thick section (0.25 wrn) through a Chinese hamster metaphase chromosome from a cell allowed to recover for 30 minutes from colcemid treatment. Short MTs have reappeared and are mostly oriented parallel to the kinetochore disk. The short vertical MT at the arrow is entirely within the section and is clearly unattached. See text for details. X 3 5 , l O O . (From Ris and Witt. 1981.)

where the two gradients overlapped. K-Fiber formation would result from an interaction between the centrosome and kinetochore which would create “a localized environment which favors polymerization in a general environment which disfavors polymerization” (DeBrabander et al., I979b). c . An Inieruclion between the Kinetochore and Centrosome Nucleated MTs. (recently resurrected by Pickett-Heaps and Tippit, 1978; Tippit et al., 1980, for diatoms; Rieder and Borisy, 1981, for mammals). Rieder and Borisy (1981) used correlative LM and HVEM methods to investigate the initial stages of spindle formation in PtK, cells recovering from low temperature treatment. They found that the centrosomes generate MTs during the initial stages of recovery well before MTs were seen to be associated with the kinetochores. At a later stage of recovery, those kinetochores closest to and facing the poles were the first to show associated MTs, and these MTs were already oriented toward a centrosome. It was then postulated that the poles influence the order in which ki-

44

CONLY L. RIEDER

netochores acquire MTs because the kinetochores in these cells possess an intrinsic tendency to initially associate with MTs growing from the centrosome: (see also Paweletz, 1974). The subsequent poleward translocation of the chromosome could then be attributed to well-described (see review of Rebhun, 1972; Bajer and Mole-Bajer, 1972) but little understood transport properties of the asters. Each of the above models perdicts that the appearance of MTs at the kinetochore would be closely coupled with the poleward orientation of the chromosome, and that the centrosome influences the order in which kinetochores acquire MTs. [An additional mechanism, in which a kinetochore first nucleates MTs toward random points within the cell and then becomes oriented by the interaction of its associated MTs with other spindle MTs (e.g., Fig. 11B in McIntosh et af., 1975a) is not consistent with the ultrastructural observations described in Section VI,C, 1.] The experiment of Rieder and Borisy (1981) was designed to directly test the model proposed by Roos (1976). These investigators allowed the NE to breakdown in late prophase PtK, cells at a temperature (6°C) which inhibited MT formation and then warmed the cell, at a later time, to induce spindle formation. Under these conditions any transient gradient of diffusable substances formed at the onset of NE breakdown would no longer exist. The results obtained from this study were remarkably similar to those obtained by Roos (1 976) for the initial behavior of prometaphase chromosomes in untreated PtK, cells at 37°C. Since the predominant pathway, monoorientation followed by bipolar orientation and movement to the metaphase plate, was observed in both treated and untreated cells, the mitotic process in the cold-treated cells was considered to be fundamentally normal. It was therefore concluded that the nature of the kinetochore to pole interaction could not be accounted for by a transient diffusion gradient of tubulin resulting from the asynchronous breakdown of the NE (as suggested by Roos, 1976). DeBrabander et al. (1979b,1980) note that their own concept, in which the kinetochore to pole interaction is ascribed to overlapping pole and kinetochore gradients, is speculative since the mechanism by which the gradients form and their nature are not well defined. In addition, this hypothesis must make further assumptions (e.g., that the gradient forms only when a kinetochore faces a pole) in order to explain why the kinetochore facing away from the pole on a unitelically monooriented chromosome lacks MTs (see Rieder and Borisy, 198I ) . A recently revised version of this concept (DeBrabander et af., 198la) proposes that the kinetochores and not the centrosomes release or concentrate a factor which promotes tubulin self-assembly. In this version the preferential elongation of K-MTs toward the pole is envisioned to occur via a lateral interaction with centrosomal MTs. Although this latter version offers an explanation for the nature of the kinetochore-centrosome interaction, it makes no provision for how the proximity effect arises or for the ultrastructural characteristics of unitelically monooriented chromosomes.

THE MAMMALIAN KINETOCHORE

45

The hypotheses of Roos (1976) and DeBrabander et al. (1979b, 1980, 198I a) are both based on the notion that the kinetochore directs the formation of its associated MTs. Yet, as noted above, these concepts either fail to stand up to a direct test or are based on hypothetical and (therefore) difficult to prove gradients. On the other hand the last hypothesis described above (that the kinetochore initially functions to recruit polar nucleated MTs) is attractive from a theoretical point of view since it provides a simple explanation, without additional assumptions, for (1) the origin and nature of the centrosome-kinetochore interaction, (2) why the kinetochore facing away from the pole on a unitelically monooriented chromosome initially possesses no MTs, (3) the proximity effect, and (4) the tendency of K-MTs to form between the pole and the kinetochore. As noted by Schrader (1953) this concept is the oldest and at one time was the one most “firmly ensconsed in the biological mind.” Yet, although it is consistent with the ultrastructural data on the prometaphase in a wide variety of untreated cell types (e.g., Sakai, l968,1969a,b; Bajer and Mole-Bajer, 1969; Paweletz, 1974; Kubai, 1973,1975; Mole-Bajer, 1975; Ritter et a l . , 1978; Nicklas et a l . , 1979; LaFountain and Davidson, 1979; Tippit et al., 1980; Solari, 1980; Goldstein, 1981), it has lost its appeal in recent years in favor of the notion that the kinetochore nucleates its associated MTs. 3. Do Kinetochores Nucleate Their Associated MTs? The concept that the kinetochore nucleates its associated MTs ‘constitutes a basic building block upon which most models of mitosis are erected” and “is virtually unquestioned in nearly every paper or review of mitosis” (PickettHeaps and Tippit, 1978). Yet, as noted by Mazia (1977) this hypothesis was incorporated into theories of mitosis before there was any evidence to support it. It evolved initially from the work of Hughes-Schrader (1924,1942) who deduced, from LM observations of sectioned material, that the “chromosomal fibers” in many Coccidae and in Acroschismus formed without reference to the location of the future spindle poles. Later work suggested that her conclusion, that K-Fibers originate at the kinetochore (see discussion in Schrader, 1953), was valid for most types of organisms. For example, InouC (see review of 1964) found that the birefringence of the K-Fiber in plant and animal spindles is greatest near the kinetochore, an observation which was interpreted to indicate that the kinetochore acts as a nucleating center. Forer (see review of 1969) provided additional support for this view when he noted that an area of reduced birefringence, produced on a metaphase K-Fiber in crane fly spermatocytes by UV irradiation, moves poleward as the fiber regenerates. In this respect many investigators (see references in Luykx, 1970: McIntosh et a l . , 1975a; Witt et a l . , 1980) have reported that K-Fibers reform after disruption by biochemical or physical agents from the kinetochore toward the poles. Finally Dietz (1966) reported that meiosis proceeds in a normal fashion in mechanically flattened crane fly spermatocytes even when the poles (i.e., centrosomes) fail to separate,

46

CONLY L. RIEDER

an observation which was interpreted to indicate that centrosomes are not necessary for the formation of K-Fibers. [A similar conclusion was reached by Borisy ( 1978) from the laser microbeam work of Berns et al. (1977) which suggests that damage to a centrosome during prophase disrupts the formation of interpolar but not K-MTs.] These and similar studies (see additional citations in the above references) provide what appears at first to be rather convincing evidence that the kinetochore is responsible for the formation of its associated MTs. Yet the results of each of these studies are consistent with other interpretations. For example, Inout’s (1964) observation that K-Fiber birefringence is greatest near the kinetochore is more likely a manifestation of the fact that K-MTs are more highly organized (i.e., show a higher density and more pronounced packing) near the kinetochore. The initial interpretation of Forer’s (1969) result, and the results of all regrowth experiments, can be questioned on the grounds that these cells were already in mitosis and that K-Fiber regeneration occurred not by the nucleation of new MTs by the kinetochore but simply by the elongation of existing K-MTs (Witt et a l . , 1980; see also Pickett-Heaps and Tippit, 1978, for an additional interpretation of Forer’s result). Likewise, an examination of Hughes-Schrader’s ( 1924) work on coccids reveals that both the kinetochores and the polar areas in the meiotic spindles of these organisms are difuse, a fact which could easily lead to an erroneous interpretation as to how these kinetochores acquire MTs, especially when it is based solely on LM data from fixed cells. Similarly, her classic work on the formation of the compound meiotic inrranuclear spindle in Acroschismus oocytes, which has been cited by many (e.g., Nicklas, 1971; Snyder and McIntosh, 1975; Mclntosh et al. , 1975a; Gould and Borisy, 1978) as one of the most convincing demonstrations of kinetochore directed fiber formation, has yet to be reinvestigated at either the LM or EM level. Dietz’s (1966) work, which also lacks both confirmation and an ultrastructural verification, is open to the criticism that a microtubule organizing center (i.e., aggregrates of pericentriolar material) was still present in the polar areas of these spindles regardless of the position of the asters (e.g., see Szollosi et a / . , 1972). In this respect, Borisy’s ( 1978) interpretation of the laser microbeam data of Berns tat al. (1977) is not convincing since the original study mentions that functional pericentriolar material with associated MTs frequently “appears between the irradiated centriole and the chromosome material.” The isolation and purification of polymerization-competent MT protein (tubulin) provided a new approach for the study of how kinetochores acquire MTs. It was quickly demonstrated that mammalian kinetochores on isolated chromosomes (Telzer et al. , 1975; Gould and Borisy, 1978; Bergen et al. , 1980) or on chromosomes in situ (McGill and Brinkley, 1975; Snyder and McIntosh, 1975; Pepper and Brinkley, 1979) acquire MTs when exposed to exogenous tubulin. Yet the interpretation of most of these in virro results has been subjected to the

THE MAMMALIAN KINETOCHORE

47

criticism (e.g., see Nicklas, 1977b; Mazia, 1977; Pickett-Heaps and Tippit, 1978; Witt et ( I / . , 1980) that the kinetochores were attached to MTs prior to incubation in exogenous tubulin, and that those MTs formed in the presence of added tubulin were simply elongating K-MT fragments and not newly nucleated MTs. On the other hand, two of these in vitro studies deserve special consideration since one can be reasonably sure that colcemid was added prior to NE breakdown and in a concentration which was sufficient to completely block spindle formation. Synder and McIntosh (1975) treated prophase PtK, cells with colcemid and lysed these cells, at different intervals after NE breakdown, into solutions containing a high concentration of tubulin. They noted that chromosomes in cells lysed into tubulin long after NE breakdown showed an increase in the number and length of MTs associated with their kinetochores relative to those lysed immediately after NE breakdown (which showed only a few short tubule structures oriented perpendicular to the long axis of the kinetochore). This result was interpreted to indicate that the differentiation of the kinetochore during the prophase-prometaphase transition (see Section IV,C), controls “the apparent ability of the chromosome to initiate tubules or tubule fragments.” Could and Borisy (1978) examined the number of MTs assembled by chromosomes, isolated from CHO cells which entered mitosis under the influence of colcemid, and noted that the fraction of kinetochores that acquired MTs when exposed to tubulin depended on the tubulin concentration and the age of the chromosome preparation. Under the best of conditions (i.e., freshly isolated chromosomes augmented with a concentration of tubulin ur the threshold of self’initiution)98% of the chromosomes gave rise to MTs. Although the average number of MTs per kinetochore was only one-third to one-half that reported for metaphase CHO cells in vivo (i.e., 4 versus 10-14), 36% of the chromosomes had 10 or more MTs per kinetochore. These findings indicate that not all kinetochores on these chromosomes possess the same nucleating capacity, a result which could similarly reflect the state of kinetochore differentiation since the cell population entered prometaphase asynchronously under the influence of the drug. The tubulin augmentation experiments cited above clearly demonstrate that kinetochores can, under suitable conditions (e.g., high concentrations of exogenous tubulin), act as microtubule nucleating centers in situ and in vitro. Yet they provide no evidence that kinetochores function as nucleating sites in vivo. In this respect it has recently been shown that when mitotic CHO cells are allowed to recover from a prolonged colcemid treatment, “inexperienced” kinetochores (i.e., kinetochores which have never seen a MT) acquire MTs without facing a centrosome at a time when the MT nucleating capacity of the centrosome is greatly supressed (Witt et a/., 1980). It can be concluded from this study that kinetochores can direct the formation of MTs in vivo after release from prolonged colcemid treatment. Similar observations were reported by DeBrabander et t i / .

48

CONLY L. RIEDER

(1979b,1980, 1981a) who studied the formation of K-MTs in recovering RK, cells after the cells had entered mitosis in the presence of Nocodazole. It should be stressed that in both of these in vivo studies MTs appeared to form initially during recovery in the vicinity of the kinetochore plate and only later became attached to rhis structure (see below; Fig. 27). This prompted DeBrabander et al. ( 1980) to redefine kinetochore-directed “nucleation” as the “capacity to induce assembly in a more or less localized region within a general environment where assembly does not occur,” a definition which does “not necessarily imply the presence of seeds or templates” on kinetochores. This definition of nucleation differs substantially from that arising from the in vitro work (e.g., see Snyder and McIntosh, 1975; McGill and Brinkley, 1975; Pepper and Brinkley, 1979) which implies that MTs arise directly from the kinetochore disk. Thus, the process by which MTs are ‘ ‘nucleated’’ by kinetochores in vitro appears to differ from that occurring in vivo. The fact that kinetochores can be shown to nucleate MTs in vitro and in vivo during recovery from colcemid or Nocodazole treatment does not tell us to what extent they do so during the course of a normal mitosis since it by no means excludes the possibility that the primary function of the kinetochore is to attach to polar nucleated MTs. Although there is no direct experimental evidence to support this long-standing hypothesis, it is certainly consistent with the in vivo results of Witt et a f . (1980; see also DeBrabander er at., 1979b, 1980, 198la; Ris and Witt, 1981) which indicate that the kinetochore plate functions to “capture” preformed MTs. The light and electron microscopic observations described in the beginning of this section which demonstrate the existence of a kinetochore-pole interaction during K-MT formation, have yet to be satisfactorily explained by the kinetochore-directed nucleation of MTs without assuming the presence of difficult to demonstrate gradients. On the other hand they are readily explained without additional assumptions if the primary in vivo function of the kinetochore is to recruit polar nucleated MTs. It should be reiterated that these two concepts are not mutually exclusive and evidence supporting one by no means excludes the other. A proximity effect would be expected if kinetochores attach to polar MTs since polar MT density increases at progressively closer distances to the centrosomes (i.e., more MTs are available for recruitment closer to the pole). Unitelic monoorientation would arise from clearly documented but little understood aster transport properties (e.g., Rebhun, 1972; Bajer and Mole-Bajer, 1972; Rickards, 1975) and from the fact that the initial attachment and movement of a kinetochore toward the near pole would orient the sister kinetochore away from that pole (see Section IV,B) and block it from acquiring MTs. Syntelic malorientation (e.g., Fig. 28) would be expected to arise in those instances in which both kinetochores, or kinetochore equivalents (as in meiotic monoorientationsee Nicklas, 1961, 1967) on a chromosome attach simultaneously to MTs from

THE MAMMALlAN KINETOCHORE

49

FIG.28. Thin section through a maloriented R K , chromosome. Both kinetochores (K) are apparently attached to MTs running toward the near pole (direction of arrow). X 4 1 ,OOO. (From Roos, 1973a.)

the same pole. This mechanism also offers a straightforward explanation as to why the first MTs on prometaphase kinetochores “are often approximately parallel to the interpolar axis regardless of the orientation of the kinetochore itself” (Luykx, 1970), and how a single kinetochore may come to possess MTs which are oriented toward opposite poles (e.g., Luykx, 1970; Heneen, 1975; Lambert and Bajer, 1975; Nicklas et al.. 1979; Jensen, 1982). Finally, such a mechanism explains why early prometaphase kinetochores frequently appear stretched into a “flame” shape (see Section IV,C; Fig. 13). This would be expected if polar MTs, which need not necessarily terminate at the kinetochore, laterally associate with and produce a pole-directed force on the kinetochore. The numerous polar and free MTs of the forming spindle, which penetrate the chromosome very near the kinetochore, illustrate at the very least that an abundant supply of potentially recruitable MTs exists in the vicinity of the kinetochore throughout K-Fiber formation. The notion that the prometaphase kinetochore can associate with polar nucleated MTs also offers a simple explanation for the induction of chromosomal spheres by low concentrations of mitotic inhibitors (see above.) Brinkley and Stubblefield (1970) originally proposed that a low dose of colcemid preferentially inhibits the nucleated assembly of centrosomal MTs and also of those MTs which would normally form on the outer kinetochores without inhibiting the nucleation and growth of MTs on those kinetochores facing the single polar area. This hypothesis can be subjected to several criticisms, one of which is that the initial nucleation of MTs by the kinetochores (immediately after NE breakdown) would frequently be expected to occur well removed from the polar area, where

50

CONLY L. RIEDER

both kinetochores on a chromosome would certainly see the same environment (and either be equally inhibited or equally able to nucleate MTs). Moreover we have found that concentrations of colcemid which induce sphere formation, in cells entering prometaphase under the influence of the drug, do not destroy a large proportion of MTs associated with the centrosomes of late prophase cells (Rieder, unpublished). However a concentration of colcemid which destroys all or most centrosomal MTs in late prophase cells invariably inhibits the subsequent formation of a chromosomal sphere. These observations, together with the fact that only a few MTs are observed to penetrate through fully formed chromosomal spheres without terminating on a kinetochore (e.g., Bajer and Mole-Bajer, 1972; Fig. 26), support our interpretation that all of the kinetochores in these cells are inhibited from nucleating MTs, and that they are attaching to and being transported poleward by centrosomal MTs which were formed during interphase (prior to colcemid treatment) and not destroyed by the subsequent colcemid treatment. This hypothesis for chromosomal sphere formation is strengthened by recent results which indicate that substoichiometric concentrations of colcemid (or colchicine) interfere preferentially with MT nucleation and elongation without immediately destroying preexisting MTs (e.g., Oppenheim et al., 1973; Wilson et a / ., 1976; Margolis and Wilson, 1977; Jaeckell-Williams, 1978). It is also consistent with the recent data of Heidemann (1980) concerning the polarity of chromosomal sphere MTs (see below). MTs grown from kinetochores in vitro elongate at a rate which corresponds to the addition of subunits only at the plus (i.e., fast growing) end of the MT which is distal to the kinetochore (Summers and Kirschner, 1979; Bergen ef al., 1980). Similarly, MTs generated in vivo by centrosomes, basal bodies, melanophores, and heliozoan centroplasts (i.e., all the “true” MT nucleating centers examined to date) are oriented with their plus ends distal to the site of nucleation (Heidemann et al., 1980; Kirschner, 1980; Euteneuer and McIntosh, 1981a). One would therefore expect that the question regarding the origin of K-MTs in vivo could be resolved by simply determining their polarity relative to the polar MTs. A recruitment hypothesis predicts that K-MTs would possess the same polarity as the polar MTs within that half-spindle containing the K-Fiber while a nucleation hypothesis predicts that it would be the opposite. Euteneuer and McIntosh (1981b; see also Heidemann, 1980; Telzer and Haimo. 1981) have recently shown that the polarity of K-MTs in metaphase P t K , cells is the same as the polariry of the polar MTs in that half spindle (Fig. 29). This finding, which was totally unexpected in light of the conclusions drawn from the in vitro work (see above), can be interpreted in only two ways: either the prometaphase kinetochore functions primarily to recruit centrosome nucleated MTs or it initiates MT growth in vivo (but not in vitro) with the minus (i.e., slow growing) end of the MT distal to the site of nucleation. The recent in vivo experiments of Witt et a / . (1980) and DeBrabander el’ al.

THE MAMMALIAN KINETOCHORE

51

Fici. 29. Cross section of a cold-stable metaphase PtK, kinetochore fiber treated in such a way that the polarity of its MTs can be determined. This section is “looking” toward the pole. Note that the great majority of K-Fiber MTs have been decorated with hooks bending clockwise. This indicates that the fast growing ends of these MTs are proximal to the kinetochore (i.e., the same polarity as the centrosomal MTs in the half-spindle). X70.000. (Courtesy of U. Euteneuer.)

(1980,1981a) suggest that the second possibility may not be as unlikely as it first appears. As previously noted MTs “nucleated” in vivo by the kinetochore, during recovery from colcemid or Nocodazole treatment, are initially formed in the immediate vicinity of bur not in direct contacr wirh the kinetochore plate (Witt et al., 1980; DeBrabander er a l . , 1980,1981a; Ris and Witt, 1981). The subsequent attachment of these short corona nucleated MT fragments to the outer plate appears to precede the formation of K-MT bundles (i.e., the attachment of corona nucleated MTs to the outer disk is a separate characteristic of this microtubule organizing center). Thus, the polarity observed in vivo for K-MTs could result from the initial nucleation of these MTs in the immediate vicinity of the kinetochore followed by the attachment of the plus end of each into the outer plate. Further growth of the MT would then be expected to occur slowly at the distal end. On the other hand, corona nucleated MTs not yet attached to the kinetochore disk may still show a net elongation from growth at their plus ends

52

CONLY L. RIEDER

by the addition of subunits proximal to the kinetochore. Yet if this were the case one would expect the plus end to elongate past the kinetochore plate (unless the act of elongation somehow pushes the MT poleward). Furthermore, once these MTs attach to the outer disk growth would most likely be restricted to the Idistal (i.e., minus) end since “it is difficult to conceive how a structure could simultaneously be anchored and growing at the same point” (Borisy, 1978). [The fact that metaphase K-MTs are firmly anchored in the outer disk, along with the fact that these MTs show an increased stability to disruption by physical and chemical agents (see Section VI,B), raises doubts of whether the tubulin subunits within these MTs are in the sort of dynamic equilibrium envisioned by InouC (1964), or whether they even “treadmill” through K-MTs as postulated by Margolis t?t al. (1978). The problem boils down to whether MT subunits can be added to the KMTs of a metaphase K-Fiber at the kinetochore (e.g., see Fig. 8 of Margolis and Wilson, 1981). It may be presumed that this question will soon be resolved by microinjecting tubulin, labeled with ferritin, into metaphase cells.] If K-Fiber formation can occur by the attachment of corona-nucleated MTs to the outer plate it should also be able to occur by the attachment of polar nucleated MTs to this structure since it is highly unlikely that the outer plate can discriminate between the plus end of corona nucleated MTs and penetrating polar nrrcleated MTs. In this respect it should be remembered that the early prometaphase spindle is composed primarily (if not exclusively) of centrosome nucleated MTs and that in a particular half-spindle K-MTs have the same polarity as centrosomal MTs. Based on these facts I suggest that kinetochore attachment (and chromosome orientation) arises initially from the recruitment of polar nucleated MTs by the prometaphase kinetochore. The recruitment of one or a few polar .MTs immediately connects the kinetochore to the polar region and these MTs are already oriented toward the pole. Growth of the K-Fiber then most likely occurs by the lateral association of additional polar and/or corona-nucleated MTs with the first few MTs recruited by the kinetochore or with the kinetochore itself. The number of MTs associated with the kinetochore would be expected to progressively increase until its surface area is filled with MTs spaced approximately 50-60 nm apart (see Section VI,B). Neighboring MTs unable to attach to the “saturated” kinetochore would penetrate the chromosome in the vicinity 0 1 the kinetochore. In this scheme the formation of MTs by the kinetochore (corona) is envisioned to be of secondary importance, facilitating K-Fiber growth (by interacting laterally with previously bound centrosomal nucleated MTs), and/or ensuring that the kinetochore will attach to the spindle in the absence of recniitable centrosomal MTs (e.g., to allow amphi-orientation of a unitelically monooriented chromosome). In light of the importance of the formation of a kinetochore to pole attachment, such an additional mechanism for K-Fiber formation would certainly be selected for by evolution. The origin of K-MTs, as outlined above, is consistent with the morphological

THE MAMMALIAN KINETOCHORE

53

changes which occur within the astral spindle during prometaphase (i.e., diminution of asters concurrent with the growth of the spindle between the chromosomes and poles), with the structure of the K-Fiber, with the ultrastructural data on prometaphse in many types of cells, with the in vivo nucleation data, and with the in vivo polarity determinations. Moreover, it offers a simple explanation for various in vivo observations (e.g., the proximity effect, unitelic monoorientation) which have yet to be adequately explained by a mechanism of K-MT formation based solely on the nucleation of MTs by the kinetochore.

ACKNOWLEDGMENTS The author would like to thank Drs. H. Ris, D. Parsons, E. D. Salmon, S. Brenner, and Ms. S. Nowogrodzki for their valuable comments and careful reading of this manuscript. 1 am grateful to the following scientists for supplying copies of their unpublished or published micrographs: Drs. S. Brenner, H. Eiberg, U. Euteneuer, J. R. McIntosh. Y. Moroi, D. Pepper, J. B. Rattner, H. Ris. U-P. Roos. M. J. Schibler, and M. J. Welsh. During the preparation of this article my work was supported by HRI Grant 65027 and BRS Grant 37007 awarded by the New York State Department of Health and by a Biotechnological Resource Grant PHS RR 01219 awarded by the Division of Research and Resources, D.H.H.S., to support the N.Y.S.D.H. (at Albany) High Voltage Electron Microscope.

REFERENCES Aggarwal, S. K. (1976). J. Hisrochem. Cvtochem. 24, 984-992. Allen, R. D., Bajer, A.. and LaFountain, J. (1969). J. Cell Biol. 43, 4a. Alov, 1. A,, and Lyubskii, S. L. (1977). f n t . Rev. Cvrol. Suppl. 6 , 59-74. Alves, P., and Jonasson. J. (1978). J. CelISci. 32, 185-195. Amos, L. A. (1977). J. Cell Biol. 72, 642-654. Aubin. J. E., Weber. K.. and Osborn, M. (1979). Exp. Cell Res. 124, 93-109. Bajer, A. S. (1973). Cyobios 8, 249-281. Bajer. A. S., and Mole- Bajer, J. (1969). Chromosoma 27, 448-484. Bajer. A . S.. and Mole-Bajer. J.. (1972). fnt. Rev. Cvtol. Suppl. 3, 1-271. Barak. L. S.. Nothnagel, E. A,, DeMario, E. F., and Webb, W. W. (1981). Proc. Nail. Acad. Sci. U.S.A. 78, 3034-3038. Barham. S. S.. and Brinkley, 9. R. (1976). Cvrobios 15, 97-109. Bauer. H. (1952). Zoo/. Anz. Suppl. 17, 252. Begg, D. A., and Ellis, G.W. (1979a). J. Cell Biol. 82, 528-541. Begg. D. A., and Ellis, G. W. (1979b). J. Cell Biol. 82, 542-554. Behnke. 0. (1975). Cytobiologie 11, 366-381. Bergen, L., Kuriyama. R., and Borisy, G . G. (1980). J. Cell Biol. 84, 151-159. Bernhard, W. (1969). J. Ulfrastruct. Res. 27, 250-265. Berns. M. W., Rattner, J. B., Brenner. S., and Meredith, S . (1977). J. Cell Biol. 72, 351-367. Bielek, E. (1978a). Verh. Anat. Ges. 72, 193-198. Bielek. E. (1978b). Cv!obiologie 16, 480-484.

54

CONLY L. RIEDER

Bokhari. F. S . . and Godward, M. B. E. (1980). Chromosoma 79, 125-136. Borisy, G. G . (1978). J. Mol. Biol. 124, 565-570. Bostock, C. J., and Sumner, A. T. (1978). “The Eukaryotic Chromosome,” pp. 1-525. NorthHolland Publ., Amsterdam. Braselton. J. P. (1971 ). Chromosoma 36, 89-99. Braselton, J. P. (1980). Can. J . Genet. Cvtol. 22, 7-10. Braselton, J. P. (1981). Chromosoma 82, 143-151. Brat, S. V.. Verma. R. S. and Dosik, H. (1979). Stain Techno/. 54, 107-108. Brenner. S. L.. Liaw. L. H.. and Bems. M. W. (1980). Cell Biophvs. 2, 139-155. Brenner. S. L., Pepper. D., Berns, M. W., Tan, E. and Brinkley, B. R. (I98 I ) . J. Cell Biol. 91, 95-102. Brinkley. B. R.. and Cartwright, J. (1971). J. Cell Biol. 50, 416-431. Brinkley, B. R.. and Cartwright, J. (1975). Ann. N.Y. Arad. Sci. 253, 428-439. Brinkley, B. R., and Stubblefield, E. (1966). Chromosoma 19, 28-43. Brinkley, B. R., and Stubblefield. (1970). Adv. Cell B i d . 1, 119-185. Brinkley, B. R., Stubblefield. k . , and Hsu, T. C. (1967). J. Ultrastruct. Res. 19, 1-18. Brinkley, B. R., Cox, S. M., and Pepper, D. A. (1980). Cvtogenet. Cell Genet. 26, 165-174. Brown, P. A,, and Loughman, W. D. (1980). Cvtogenet. Cell Genet. 27, 123-128. Brown. S. W. (1966). Science 151, 417-425. Buck, R . (1967). J. Ultrastruct. Res. 18, 489-501. Bulinski, J. C., and Borisy, G. G. (1980). J . Cell Biol. 87, 792-801. Cande. W. A,, and Wolniak, S. M. (1978). J . Cell B i d . 79, 573-580. Cande. W. Z.. Lazarides, W. and Mclntosh, J. R. (1977). J. Cell Biol. 72, 552-567. Church, K., and Moens. P. B. (1976). Chromosoma 56, 249-263. Clapham. L., and Ostergren, G. (1978). Hefeditas 89, 89-106. Commings, D. E. (1978). Annu. Rev. Genet. 12, 25-46. Commings, D. E.. and Okada, T. A. (1971). E?cp. Cell Res. 67, 97-1 10. Commings, D. E., and Okada, T . A. (1972). Chromosoma 37, 177-192. Commings, D. E., Avelino, E., Okada, T. A,. and Wyandt. H. E. (1973). Exp. Cell Re:;. 77, 469-493. Corces. V. G., Salas, J.. Salas, M. L., and Avila, T. (1978). Eur. J . Biochem. 86, 473-479. Cox, J. V.. Schenk, E. A., Olmsted, J. B. (1980). J. Cell Biol. 87, 240a. DeBrabander, M., Geuens, G., DeMey, J.. and Joniau. M. (1979a). Biol. Cell. 34, 213--226. DeBrabander. M., DeMey. J., Geuens. G., and Joniau, M. (1979b). Electron Microsc. Soc. Am. Proc. 37, 10-13. DeBrabander, M., Geuens, G., Nuydens, R.. Willebrords, R., and DeMey, J. (1980). In “hlicrotubules and Microtubule Inhibitors 1980” (M. DeBrabander and J. DeMey, eds.), pp. 255--268. Elsevier. Amsterdam. DeBrabander, M., Geuens, G., DeMey, J.. and Joniau, M. (1981a). CellMotil. 1, 469-48L. DeBrabander, M., Bulinski, J. C.. Geuens, G., DeMey, .I., and Borisy, G. G. (1981b). J. Cell Biol. 91, 438-445. DeMey, J.. Moeremans, M. Geuens, G. Nuydens. R.. VanBelle, H., and DeBrabander, M. (1980). In “Microtubules and Microtubule Inhibitors 1980” (M. DeBrabander and J. DeMey. eds.), pp. 227-242. Elsevier, Amsterdam. Denton. T. E.,Brooke, W. R., and Howell, W. M. (1977). Stain Techno/. 52, 31 1-313. Dietz, R. (1966). Proc. Oxjord Chromosome Conf., 1st. 1964 1, 161-166. Dietz, R. (1972). Chromosoma 38, 11-76, DuPraw, E. J. (1968). “Cell and Molecular Biology.” Academic Press. New York. DuPraw, E. J. (1970). “DNA and Chromosomes.” Holt, New York. Eiberg. H. (1974). N a m e (London) 248, 5 5 .

THE MAMMALIAN KINETOCHORE

55

Esponda. P. (1978). E.rp. Cell Res. 114, 247-252. Euteneuer. U.. and Mclntosh. J. R. (1981a). Proc. Nut/. A i a d . Sci. U.S.A. 78, 372-376. Euteneuer. U.. and Mclntosh. J. R. (1981b). J. CcllEiol. 89, 338-345. Euteneuer. U.. Bereiter-Han. J.. and Schliwa. M. (1977). Cjrobiologie 15, 169-173. Forer. A. (1969). In “Handbook of Molecular Biology” ( A . Lima-de-Fdria, ed.). pp. 554-601. North-Holland Publ., Amsterdam. Forer. A. (1974). In “Cell Cycle Controls” ( G . M . Padilla, 1. L. Cameron. and A. M . Zimmerman. cds.). pp. 319-336. Academic Press. New York. Forer. A , . and Behnke. 0. (1972). Chromosomu 39, 145-173. Forer. A,. and Jackson. W. T. (1975). Cytohiologie 10, 217-226. Forcr. A , . Jackson. W. T.. and Enberg. A . (1979). J. Cell Sci. 37, 349-371. Friedlender. M.. and Wahrman. J. (1970). J. Cell Sci. 7, 65-89. Fritzler. M. J . . and Kinsella. T. D. (1980). Am. J. Med. 69, 520-526. Fuge, H. (1974). Protoplusmu 82, 289-320. Fuge. H. (1977). Int. Re).. C,vtol. Sidppl. 6, 1-58, Fuge. H. ( 19811, Eur. J . Cell B i d . 25, 90-94. Fuller. M . S. (1976). Inr. Rev. C y t ~ ~45, l . 113-151. Gawadi. N. (1974). Cvrohios 10, 17-35. Goldstein. L. S. B. (1981). Cdl25, 591-602. Could. R. R., and Borisy. G. G . (1978). Exp. Cell Res. 113, 369-374. Hard. R.. and Allen, R. D. (1977). J. Cell Sci. 27, 47-56. Heath, I. B. (1979). 1,v. Rev. Cytol. 64, 1-80, Heidemann, S. R . ( 1980). In “Microtubules and Microtubule Inhibitors 1980” ( M . DeBrabander and J . DeMey. eds.). pp. 341-355. Elsevier, Amsterdam. Heidemann. S. R . . Zieve. G. W . . Mclntosh, J. R. (1980). J. Cell B i d . 87, 152-159. Heneen. W . K. (1970). Chromosomu 29, 88-1 17. Heneen. W. K. (1975a). Hc,reditus 79, 209-220. Heneen. W . K. (1975b). E.rp. Cell Res. 91, 57-62. Hennig. W. (1973). I n t . Rev. C\wl. 36, 1-44. Hcpler. P. K. (19x0). J . Cell Biol. 86, 490-499. Herman. I. M.. and Pollard. T. D. (1978). E~rp.Cell Res. 114, 15-25. Herman. I. M.. Maupin, P.. and Pollard. T. D. (1980). J. Cell Eiol. 87, 224a. Holmquist. G . P.. and Dancis. B. (1979). Proc. Nut/. Acad. Sci. U.S.A. 76, 4566-4570. Hsu. T. C., Pathak. S.. and Chen. T. R. (1975). C,vtogenet. Cell Genet. 15, 41-49. Hughes-Schrader. S. (1924). J. Morphol. 39, 157-207. Hughes-Schrader, S. (1942). J. Morphol. 70, 261-299. Hughes-Schrader. S.. and Schrader. F. (1961). Chromosoma 12, 327-350. Hyams. J. S.. and Stebbings, H. (1979). In “Microtubules“ (J. S. Hyamsand K. Roberts. eds.). pp, 487-530. Academic Press, New York. Inoue, S. (1964). In “Primitive Motile Systems in Cell Biology” (R. D. Allen and N. Kamiya, eds.), pp. 549-598. Academic Press. New York. Izutsu. K., Sato, H.. Nakabayashi. H.. and Aoki. N. (1977). Ccll Struct. Funct. 2, 119-133. Jaeckel-Williams. R. (1978). J. Cell Sci. 34, 303-319. Jensen, C. G . ( 1982). J . Cell B i d . 92, 540-558. Job, D.. Rauch. C . T.. Fischer, E. H . , and Margolis. R. L. (1982). Eiochernistrv 21, 509-515. John. B.. and Miklos. G . L. (1979). In/. Rev. Cvtol. 58, 1-114. Jokelainen, P. T. (1965). J. Cell B i d . 27, 48a. Jokelainen, P. T. (1967). J. Uhrustruct. Res. 19, 19-44. Journey. L. J.. and Whaley, A. (1970). J. CeN Sci. 7, 49-54. Kirschner. M . W . (1980). J. Cell Eiol. 86, 330-334.

56

CONLY L. RIEDER

Krishan, A. (1968). J. Ultrustruct. Res. 23, 134-143. Kubai, D. F. (1973). J . Cell Sci. 13, 51 1-552. Kubai. D. F. (1975). h t . Rev. Cyrol. 43, 167-227. LdFountain, J. (1975). Eiosvsrems 7, 363-369. LaFountain, J., and Davidson, L. A. (1979). Chromosoma 75, 293-308. Lambert, A. M.. and Ba.jer, A. S. (1975). J. Microsc. B i d . Cell. 23, 181-194. Levan, A., Nichols, W., DeLuse, M., and Corieil, L. L. (1966). Chromosornu 18, 343-355. Lima-de-Faria. A. (1958). Int. Rev. Cytol. 7, 123-157. Lin, H-P.. Auk. J. G . . and Church, K. (1981). Chromosoma 83, 507-521. Luykx. P. (1970). In,. Rev. Cvtol. Suppl. 2, 1-173. Lyubskii, S . L. (1974). Bull. Exp. Biol. Med. (Moscow) 78, 80-83. Lyubskii, S . L.. Buchwalow. I . B.. and Raikhlin, N. T. (1979). Acfa Hisrochem. Cytochetn. 12, 1-6. McGill. M.,and Brinkley, B. R. (1975). J. Cell Eiol. 67, 189-199. Mclntosh, J. R. (1979). In “Microtubules” (K. Roberts and J. Hyams, eds.), pp. 382-441. Academic Press, New York. Mclntosh. J. R.. Cande, W. Z., and Snyder, J. D. (1975a). In “Molecules and Cell Movement” ( S . lnoue and R. E. Stephens, eds.) pp. 31-76. Raven, New York. Mclntosh, J. R., Cande, W. Z.. Snyder, J. and Vanderslice, K. ( 1975b). Ann. N . Y . Acud. Sci. 253, 407-427. McNeil. P. A,. and Berns. M. W. (1981). J. Cell B i d . 88, 543-553. Maeki. K. (1980). Proc. Jpn. Acad. Ser. B . 56, 152-156. Maeki, K. (1981). Proc. Jpn. Acad. Ser. 8. 57, 71-76. Margolis, R. L., and Rauch, C. T. (1981). Biochemistry 20, 4451-4458. Margolis, R. L., and Wilson, L. (1977). Proc. Natl. Acad. Sci. U.S.A. 74, 3466-3470. Margolis. R . L.. and Wilson, L. (1981). Nature (London) 293, 705-71 I . Margolis. R. L., Wilson. L. and Kiefer, B. (1978). Nature (London) 272, 450-452. Marks, G . E. (1975). J. Cell Sci. 18, 19-25. Matsukumd. S.. and Utakoji, T. (1977). Exp. Cell Res. 105, 217-222. Maupin-Szamier. P., and Pollard, T. D. (1978). J. CelIEio/. 77, 837-852. Mazia, D. (1961). In “The Cell” (J. Brachet and A. E. Mirsky. eds.), Vol. 3, pp. 77-412. Academic Press. New York. Mazia. D. (1977). In “Mitosis Facts and Questions” (M. Little. N. Paweletz. C. Petzelt, H. Ponstingl. D. Schroeter. and H-P. Zimmerman, eds.), pp. 196-213. Springer-Verlag, Berlin and New York. Mazia, D.. Paweletz. N., Sluder, G., and Finze, E. M. (1981). Proc. N u t / . Acad. Sci. U.S.A. 78, 377-38 1 . Miklos. G . L., and John, B. (1979). Am. J. Hum. Genet. 31, 264-280. Moens. P. 9. (1978). Chromosoma 67, 41-54. Moens, P. B. (1979). J. Cell Biol. 83, 556-561. Moens. P. B.. and Moens, T. (1981). J. Ultrastruct. Res. 75, 131-141. Mole-Bajer, J. (1975). Cvtobios 13, 117-140. Mole-Bajer, J., Bajer, A,, and Owczarzak, A . (1975). Cvtobios 13, 45-65. Mohri. H.. Mohri. T.. Mabuchi, I.. Yazaki, I . , Sakai, H.,and Ogawa, K. (1976). Dev. Growth Difer. 18, 391-398. Moll, E.. and Paweletz, N. (1980). Eur. J . Cell Eiol. 21, 280-287. Moroi. Y.. Peebles, C.. Fritzler, M. J., Steigenvald, J., and Tan, E. M. (1980). Proc. Notl. Acad. Sci. U.S.A. 77, 1627-1631. Moroi, Y., Hartman. A. L.. Nakane, P. K., and Tan, E. M. (1981). J. Cell B i d . 90, 254-259. Moses, M. J.. and Counce. S. J. ( 1974). J. Exp. Zoo/. 189, I 15- 120.

THE MAMMALIAN KINETOCHORE

57

Mota, M. ( 1957). Proc. In!. Genet. Svmp. Suppl. Cwol. pp. I 13- I 16. Mughal. S., and Godward, M. B. E. (1973). Chromosoma 44,213-229. Muller. W. (9172). Chromosoma 38, 139-172. Nicklas. R. B. (1961). Chromosoma 12, 97-1 15. Nicklas, R . B. (1967). Chromosomu 21, 17-50. Nicklas. R . B. (1971). I n “Advances in Cell Biology” (D. M. Prescott, L. Goldstein. and E. H. McConkey. eds.). pp. 225-297. Appelton, New York. Nicklas. R . B. (1975).I n “Molecules and Cell Movement” (S. lnoue and R. E. Stephens. eds.). pp. 97-1 17. Raven Press, New York. Nicklas. R. B. (1977a). I n “Mitosis Facts and Questions” (M. Little, N. Paweletz, C. Petzelt, H. Ponstingl. D. Schroeter, and H.-P. Zimmerman, eds.), pp. 150- 155. Springer-Verlag. Berlin and New York. Nicklas. R. B. (1977b). Philos. Trans. R. SOC. London. Ser. B 277, 267-276. Nicklas. R . B. (1979). Chromosoma 74, 1-37. Nicklas, R. B., and Koch. C. A. (1972). Chromosoma 39, 1-26. Nicklas, R. B.. Brinkley, B. R., Pepper, D. A,. Kubai, D., and Rickards. G. K. (1979). J. CellSri. 35, 87-104. Oppenheim, D. S.. Hauschka, and Mclntosh, J. R. (1973). Exp. Cell Res. 79, 95-105. Ostegren, G., Mole-Bajer. J., and Bajer, A. (1960). Ann. N.Y.Acad. Sri. 90, 381-406. Paweletz, N. ( 1974). Cvrohiologie 9, 368-390. Paweletz. N. (1981). CellBiol. Inr. Rep. 5 , 323-336. Pepper, D. A., and Brinkley, B. R. (1977). Chromosoma 60, 223-235. Pepper, D. A,. and Brinkley, B. R. (1979). J. Cell Biol. 82, 585-591. Pepper. D. A , , and Brinkley, B. R. (1980). Cell Motil. 1, 1-15. Pepper. D. A . , Brenner, S., Turner. D. S., Tan, E., and Brinkley, B. R. (1980). J. Cell Biol. 87, 240a. Peterson. J . B.. and Ris, H. (1976). J. CellSci. 22, 219-242. Pickett-Heaps, J. D., and Tippit, D. H. (1978). Cell 14, 455-467. Porter, K. R., and Machado. R. (1960). J. Biophvs. Biochem. Cytol. 7, 167-180. Pratt, M. M., Otter, T.. and Salmon, E. D. (1980). J. Cell Biol. 86, 738-745. Rae, P. M. (1972). In “Advances in Cell and Molecular Biology” (E. J. Duprdw, ed.), Vol. 2, pp. 109-180. Academic Press, New York. Rattner, J. B., Branch, A., and Hamkalo, B. A. (1975). Chromosoma 52, 329-338. Rattner, J. B.. Krystal, G., and Hamkalo. B. A. (1978). Chrornosoma 66, 259-268. Rebhun, L. I. (1972). Inr. Rev. Cytol. 32, 93-131. Rickards. G. K. (1975). Chromosoma 49, 407-455. Rieder. C. L. (1979a). J. Ulrrasrruct. Res. 66, 109-1 19. Rieder, C. L. (1979b). J. Cell Biol. 80, 1-9. Rieder. C. L. (1980). In “Microtubules and Microtubule Inhibitors 1980” M. DeBrabander and J. DeMey, eds.). pp. 31 1-324. North-Holland Publ., Amsterdam. Rieder. C. L. (1981). Chromosoma 84, 145-158. Rieder, C. L., and Bajer, A. S. (1977a). J. Cell Biol. 74, 717-725. Rieder, C. L., and Bajer, A. S. (1977b). Cvrobios 18, 201-234. Rieder, C. L., and Borisy, G. G. (1981). Chromosoma 82, 693-716. Ris, H., and Kubai, D. F. (1970). Annu. Rev. Genet. 4, 263-294. Ris, H., and Kubai. D. F. (1974). J. Cell Biol. 60, 702-720. Ris, H., and Witt, P. L. (1981). Chromosoma 82, 153-170. Ritter, H., Inoue. S. and Kubai, D. F. (1978). J. CellBiol. 77, 638-654. Roos, U.-P. (1973a). Chromosoma 41, 195-220. Roos, U.-P. (1973b). Chromosoma 40,43-82.

58

CONLY L. RIEDER

Roos. U.-P. (1975). Nature (London)254, 463. Roos, U.-P. (1976). Chromosoma 54, 363-385. Roos. U.-P. (1977). Cvtobiologie 16, 82-90. Ruthman, A.. and Pennantier, Y. (1973). Chrornosoma 41, 271-288. Sakai, A. (1968). Cyrologia 33, 318-330. Sakai. A. (1969a). Cvrologia 34, 57-70. Sakai. A. (1969b). Cvtologia 34, 593-604. Sakai. H. (1978). Inr. Rev. Cyrol. 55, 23-48. Sakai, H . , Mabuchi, I . , Shimoda, S., Kuriyama, R.. Ogawa, K.. and Mohri, H., (1976). Dev. Growth Differ. 18, 2 I 1-2 19. Salmon, E.D., and Begg, D. A. (1980). J. Cell Biol. 85, 853-865. Salmon, E. D., Goode, D., Maugel, T. K., and Boner, D. B. (1976). J . Cell tliol. 69, 443-454. Sanger, J. (1975). Proc. Natl. Acad. Sci. U.S.A. 72, 2451-2455. Sanger. J. (1977). In “Mitosis Facts and Questions” (M. Little, N. Paweletz. C. Petzelt, H. Ponstingl, D. Schroeter, and H.-P. Zimmerma.1, eds.), pp. 98-1 13. Springer-Verlag. Berlin and New York. Sanger, J., and Sanger. J. M. (1979). Methods Achiev. Exp. Pathol. 8, 110-142. Sato, H., Ellis, G. W.. and Inoue, S. (1975). J. CellBiol. 67, 501-517. Schibler, M. J., and Pickett-Heaps, J. D. (1980). Eur. J. Cell Biol. 22, 687-698. Schloss, J. A,. Milsted, A,, and Goldman, R. D. (1977). J. Cell Biol. 74, 794-815. Schrader. F. (1935). Cyrologia 6, 422-430. Schrader, F. (1953). “Mitosis, The Movement of Chromosomes in Cell Division” (L. C. Dunn, ed.). pp. 1-170. Columbia Univ. Press, New York. Sherline. P., and Schiavone, K. (1978). J. Cell Biol. 77, R9-RI2. Snyder, J. A., and Mclntosh. I. R. (1975). J. Cell Biol. 67, 744-760. Solari. A. J. (1980). Chromosoma 78, 239-255. Stack, S. M. (1974). Chromosoma 47, 361-378. Stebbings. H.. and Benett, C. E. (1975).In “Microtubules and Microtubule Inhibitors” (M. Borgers and M. DeBrabander, eds.). pp. 35-45. North-Holland Publ., Amsterdam. Stubblefield, E. (1973). Int. Rev. Cytol. 35, 1-60. Summers, K., and Kirschner, M. (1979). J. Cell Biol. 83, 205-217. Szollosi, D., Calarco, P., and Donahue. R. P. (1972). J. Cell Sci. 1 1 , 521-541. Tan, E. M.. Rodan, G. R., Garcia, I., Moroi, Y., Fritzler, M. J., and Peebles, C. (1980). Arthritis Rheum. 23, 617-625. Telzer, B. R., and Haimo, L. T.(1981). J. CellBiol. 89, 373-378. Telzer, B. R., Moses, M. I., and Rosenbaum, J. L. (1975). Proc. Natl. Acad. Sci. U S A . 72, 4023-4027. Tippit. D. H.. Pickett-Heaps, J. D., and Leslie, R. (1980). J. Cell Biol. 86, 402-416. Valee, R. B., and Borisy, G. G. (1977). J . Biol. Chem. 252, 377-382. Vig, B. K. (1981). Cvtogenet. Cell Genef. 31, 129-136. Webb, B. C., and Wilson, L. (1980). Biochemistrv 19, 1993-2001. Welsh, M. J., Dedman, J. R., Brinkley, B. R., and Means, A. R. (1979).J . CellBiol. 81,624-634. Wiche, G., Corces, V. G., and Avila, J. (1978). Nature (London) 273, 403-405. Will, H., Lakomek, H. J., and Bautz, E. K. R. (1981). Exp. CellRes. 134, 129-140. Wilson, L., Anderson, K., and Chin, D. (1976). Cold Spring Harbor Conf. Cell f r o / $ 3, 1051-1064.

Witt, P. L.. Ris, H., and Borisy, G. G. (1980). Chromosoma 81, 483-505. Witt, P. L.. Ris, H., and Borisy, G. G. (1981). Chromosoma 83, 523-540. Yunis, J. J., and Yasmineh, W. G. (1972). Adv. Cell Mol. Biol. 2, 1-44. Zirkle. R. E. (1970). Rudiat. Rex 41, 516-537.

INTERNATIONAL REVIEW OF CYTOLOGY. VOL. 79

Motility during Fertilization GERALDSCHAITEN Department of Biological Science. Florida State University, Tallahassee. Florida

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Overview of Fertilization .....................

B. Requirement for Mov ........... A. The Spermatozoon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. The Egg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111. Detection of Cytoskeletal A. Sperm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Dynamics of Fertilization

IV.

..................................... ibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Microfilament Inhibitors B. Microtubule Inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

C. Summary ........................ V. The Regulation of Motility at Fertilization . . . . . . . . . . . . . . . . . . . . A. The Onset of Fertilization.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Ionic Controls ........................ C. Cyclic Nucleotides. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Calmodulin . . . . . VI . Consequences of Fertilization for Later Embryonic Development . . A. The Centrioles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. The Establishment of the First Embryonic Axis . . . . . . . . . . . . C. Fate of the Sperm Tail ............. VII. Motility during Fertilization and Its Regulation: A Model. . . . . . . . A. Motility during Fertilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Regulation of Motility.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Mechanisms for Movement: Implications for Other lntracellular Translocations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII. Prospectives and Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Prospectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

60 60 61 62 62 64 86 87 88 95 95 101 117 1 I9 119 120 130 134

134 135 135 137

138 140 140 146 148 152

152 155 155 156

59 Copyright 0 IYX2 by Academic k s s . Inc. All righa uf reproduction in any form relerved. ISBN 0-12-364479-8

60

GERALD SCHAlTEN

1. Introduction

A. OVERVIEW OF FERTILIZATION The sperm and the unfertilized egg have single-minded purposes: namely, to fuse first their plasma membranes and then their respective nuclear envelopes. Should this fusion not occur within hours of spawning or ovulation, these cells are doomed. If, however, the sperm and egg are successfully joined together within an activated egg’s cytoplasm, the discontinuity of generations will have been bridged and the now fertilized egg will have the potential to develop into a new individual. The scope of this article will be to investigate and review the knowledge regarding the manner in which the sperm is able to move to the egg, the manner in which first contact between the sperm and egg surfaces occurs, the processes involved in the physical incorporation of the sperm into the egg’s cytoplasm, and the mechanism whereby the sperm and egg nuclei within the egg cytoplasm are able to locate one another, migrate together, and complete the fertilization process by fusing their nuclear membranes. Fertilization represents the extremes in cytoskeletal organization and the complexity of motile systems. On the one hand, the cytoskeleton of the sperm is reduced to that of two polarized fibers: the axoneme emanating from the base of the centriole and, following the acrosome reaction, the microfilaments of the acrosomal process extending in the opposite direction. In contrast, the egg at fertilization has a vast array of complex behaviors including the de novo assembly of a true cytoskeleton. The repertoires of both gametes include microtubulemediated and microfilament-mediated motions. all of which are required for the successful completion of fertilization, i.e., the merging of the maternal and paternal genomes. For these reasons, and since the motile apparatus of the sperm and egg are readily accessible to isolations and ionic manipulations, the study of motility and cytoskeletal rearrangements in gametes has played and will continue to play an important role in our understanding of the cytoskeleton and its means for transducing motion. In this article, the movements during fertilization, the systems responsible for each motion, and the regulation of the associated motility will be reviewed. The major focus will be to examine the means by which the sperm is incorporated into the egg and, once within the cytoplasm proper, the sperm and egg nuclei are united to achieve syngamy. The study of invertebrate gametes, especially those of echinoderms, has led to many of the pioneering discoveries in this field and will be reviewed in detail. The reviews by Bedford (1970), Austin (1968), Longo (1973), Gwatkin (1977), and Yanagimachi (1978, 1981) are recommended for readers interested in the state of mammalian fertilization.

61

MOTILITY DURING FERTILIZATION

B. REQUIREMENT FOR MOVEMENT DURING FERTILIZATION In order for fertilization to be successful several movements must occur (see Table I). Sperm must be transported near the egg surface, the sperm and egg must achieve close contact required to effect membrane fusion, and then the sperm must be drawn into the egg cytoplasm proper. Once within the egg cytoplasm the sperm and egg nuclei must locate one another and move together to establish the contact that will eventually result in the fusion of their nuclear envelopes, a complex example of membrane fusion involving the pair of nuclear membranes. On the basis of work using motility inhibitors, both microtubule-mediated and microfilament-mediated motions have been implicated as central for the completion of fertilization. Electron microscopy and indirect immunofluorescence microscopy have detected changing arrays of microtubules and microfilaments in TABLE I MOVEMENTS D U R I N G FERTILIZATION (L.variegatus

0:oo 3.36 seconds 12.85 seconds 20.00 seconds 35 seconds 50 seconds

AT

4.4-6.8 minutes

=

63)

A. Sperm incorporation Sperm+gg adhesion Onset of bioelectric responses Cessation of sperm tail motility Cortical granule discharge starts Duration Of egg cortical contraction (rate: ca. 5.9 pdsecond) Formation of fertilization cone (elongation rate: 2.6 pdminute; height: 6.7 Pm)

1.25 minutes

23°C; X

*

Gliding of sperm along egg cortex (average rate, 3.5 1.3 pdminute; average distance, 12.4 4.9 p m ) and resumption of sperm tail beating in egg cytoplasm B. Formation of sperm aster Assembly of microtubules to form sperm aster; average rate of movement of the astral center from the egg periphery. 4.9 2 I .7 pmiminute; average distance traversed, 14.3 5.5 p n

*

*

6.8-7.8 minutes

C. Migration of female pronucleus Movement of female pronucleus to center of sperm aster; average rate. 14.6 t 3.5 pdminute; average distance traversed, 19. I ? 7.0 p m

7.8-14.1 minutes

D. Pronuclear centration Movement of adjacent pronuclei to egg center; average rate, 2.6 2 0.9 p d minute; average distance traversed, 12.3 4.0 pm

*

E. Syngamy 14.7- 15.2 minutes Pronuclear fusion; male pronucleus coalesces into female pronucleus at an average rate of 14.2 2.6 pmhinute

*

62

GERALD SCHATTEN

each gamete throughout the fertilization process; the apparent changes in the cytoskeletal configurations correlate well with living studies of the actual movements during fertilization, leading to interpretive models elucidating these events. Biochemical characterizations of isolated components from both sperm and eggs verify the presence in large quantities of actin, tubulin, and their associated ATPases and regulatory components. Finally, the elucidation ol’ the ionic sequence involved in the acrosome reaction of the sperm and the program of activation of the egg permits informed speculation regarding the intracellular regulation of these motions during fertilization.

11. Dynamics of Fertilization As a model for studying motion, fertilization is unique for a variety of reasons. Perhaps foremost is the polarity of each gamete’s quest for survival; the nature of the sperm’s motility is to propel it through the suspending fluid first to the proximity of and then into actual contact with the egg. Contrasting with this cellular migration is the egg’s movements, which are entirely intracellular translocations. First the egg participates in sperm incorporation, i.e., the motion that brings the attached sperm from the egg exterior into its cytoplasm, and then the pronuclear migrations, i.e., the intracellular movements of the sperm and egg nuclei that result in syngamy. Simultaneous with these motions of the sperm and egg, dramatic surface alterations involved with the establishment of the block to polyspermy and metabolic activation occur. With the completion of pronuclear fusion the now fertilized egg is prepared to begin to undergo cleavages and the morphogenetic movements required during embryogenesis. A. THESPERMATOZOON

Although the primary focus of this review article is the movements during fertilization from the movement of gamete adhesion through syngamy and early development, it is germane to consider briefly the structure and motility of the sperm. The cytoskeleton of the sperm is polarized and segregated, and represents the two motile activities in this cell. The sperm is propelled through its suspending medium by the beating of its tail. a now classic model in the understanding of ciliary motion (reviewed by Gibbons, 1977, 1981; Haimo and Rosenbaum, 1981; Satir, 1974). At the other end of this cell a packet of monomeric actin is polymerized into microfilaments during the acrosome reaction, resulting in the extrusion of the acrosomal process (Tilney el al., 1973). This process establishes the initial contact between the gametes. The activation of the incorporated sperm nucleus has been reviewed by Poccia (1982).

MOTILITY DURING FERTILIZATION

63

1 . Sperm Tail M o t i l i ~

The understanding of the manner in which the 9 + 2 microtubules in the sperm axoneme can effect the movement required for swimming has advanced greatly in recent years. There is now little question that the sliding of adjacent axonemal doublets is transduced into the bending motion in the sperm tail. Satir ( 1968) first ingeniously proposed the model of microtubule sliding to account for ciliary bending on the basis of quantitative ultrastructural analysis. Summers and Gibbons (197 l ) , working with trypsin-treated sea urchin axonemes, have been able to demonstrate directly the sliding of adjacent microtubules when these partially digested axonemes are exposed to ATP. The magnesium-sensitive dynein cross bridges can be disassociated from and reassociated with the axonemes and excellent evidence is available indicating that dynein is the predominant ATPase involved in microtubule sliding during ciliary motion. The recent review by Warner and Mitchell (1980) on the biochemistry of dynein is recommended to interested readers. Though it is clear that in the isolated axonemes microtubule doublets are competent to slide past one another in the presence of the ATP, it is not understood how this lateral sliding motion of doublets is converted into the bending waves required for sperm swimming. Nexin and elastin-like spoke proteins are prominant candidates for the conversion of the lateral motion into bending waves. The understanding of the biochemistry of ciliary motion has important implications for human health (reviewed by Katz, 1981). For example Afzelius ( 1976) has described a class of infertile men who are sterile because of a lack of dynein arms in the sperm axoneme; their sperm are immotile. The force, frequency, and wave form of the beat of the sperm tail vary in relation to the physiological state of the sperm. In mammalian sperm, it has been shown by Garbers, Hoskins, and co-workers (Cassillas et a / ., 1980; Garbers et a / . , 1971, 1973a,b; Garbers and Hardman, 1975; Garbers and Kopf. 1980; Hoskins and Cassillas, 1974; Hoskins et a l . , 1974, 1975) that cyclic nucleotides, hormones, and egg-associated factors (diffusable glycoproteins, complex mucopolysaccharides, etc.) play crucial roles in regulating tail motility.

2. Acrosome Reaction When exposed to egg-associated factors, sperm display a reaction, the acrosome reaction (Dan, 1954), that both externalizes contents required to penetrate the extracellular investments of the egg and extends the acrosomal process, which establishes the first contacts between the sperm and egg plasma membranes. The diversity and roles of the sperm plasma membrane has recently been reviewed by Friend (1982). In mammalian sperm, an additional complexity is added, namely, the phenomenon of capacitation. Since each sperm can undergo only one acrosome reaction, which must occur within moments of gamete fusion, and since the mammalian sperm is subjected to a wide variety of environments of varying composition prior to reaching the proximity of the unfertilized

64

GERALD SCHATTEN

egg in the oviduct, mammaliam sperm are rendered incompetent to undergo an acrosome reaction by factors placed on the sperm surface in the epididymus; these factors are removed during the sperm’s passage through the female reproductive tract. The phenomenon is known as capacitation. Upon reaching the oviduct, mammalian sperm are competent to undergo acrosome reactions, having been capacitated. The acrosome reaction involves both the secretion of the acrosomal vesicle and the polymerization of actin in the periacrosomal region of the sperm head (Tilney et al., 1973; Tilney, 1978; Tilney and Kallenbach, 1979). These events are coupled in nature and can be triggered by a variety of ionophores, including A23187, X537a, and nigericin (Tilney et al., 1978; Schackman et al., 1978). In sea urchins the secretion of the acrosomal vesicle results in the externalization of a species-specific egg-binding protein, bindin (Vacquier and Moy, 1977), as well as the release of a tryspin-like protease, acrosin (Levine et al., 1978). The protease activity is likely to be responsible for the initial digestion of the vitelline layer covering the unfertilized egg, and bindin, capable of aggregating eggs in a species-specific fashion, may play a role in causing sperm to adhere to eggs of the same species. B. THE EGG I . Sperm Incorporation The response of the egg to the successful sperm has been studied for over 100 years. In the late 187Os, Fol(1877a,b) described the formation of the fertilization cone in eggs and oocytes of sea urchins and starfish. Seifritz (1926), Chambers ( 1933), and later Tyler (1965) relied on the advances in light microscopy to describe more accurately the formation and behavior of the fertilization cone: and response to the sperm. With the advent of electron microscopy, the ultrastructural features of sperm incorporation have been documented (Fig. 1-8). Longo and Anderson (1968), using sea urchin eggs, have captured the cytoplasmic upwelling around the successful sperm and have demonstrated the formation of microfilaments within these cones. Franklin (1965), using oocytes, has studied the formation of microfilaments in polyspermic fertilization cones and recently Tihey and Jaffe (1980) have studied the formation and reorganization of microfilaments within fertilization cones of eggs and oocytes. In that work, microfilaments within the fertilization cone have been beautifully prepared by the application of nicotine to the egg prior to fixation. Transmission electron microscopy of sperm attachment (Mann et a f . , 1976; Fig. 2A) documents the microfilaments comprising the acrosomal process as well as extracellular adherent material, likely the species-specific binding protein bindin (Vacquier and Moy, 1977). The surface features of sperm

MOTILITY DURING FERTILIZATION

65

FIG. I . Insemination observed by scanning electron microscopy. An early stage of insemination of an egg glued to a polylysine-coated slide. Only the tops and sides of the egg are available for sperm binding. S/rongv/ocen/ro/uspurpuratus. Bar, 10 pm. (Reprinted with permission from Schatten and Mazia, 1976b.)

incorporation in sea urchins have been studied by Schatten and Mazia (1976a,b), Schatten and Schatten (1980a), Tegner and Epel (1976), and Usui et al. (1980). The mammalian egg surface during incorporation has been studied by Shalgi and Phillips (1980a,b), Shalgi et 01. (1978). and Yanagimachi (1978) and Yanagimachi and Noda (1972). The similarity between mammalian and sea urchin fertilization is indeed striking. Scanning electron microscopy of sperm incorporation as viewed from the extracellular surface, i.e., the vitelline layer, presented by Schatten and Mazia ( 1976a), demonstrates the initial attachment of the acrosome-reacted sperm to the egg surface by the extruded acrosomal process (Fig. 2B). The successful sperm invariably attaches in a perpendicular fashion (Fig. 3A) whereupon mem-

66

GERALD SCHATTEN

FIG. 2. The acrosome reaction of the sperm. (A) Transmission electron microscopy. Microfilaments comprising the core of the extended acrosomal process are apparent as is amorphous extracellular material associated with the apical region of the acrosomal process. The chromatin of the sperm nucleus is the electron-opaque material at the top of this image. Lyrechinus pictus. Bar, 500 nm. (From Mann et al.. 1976; micrograph courtesy of D. S. Friend.) (B) Scanning electron microscopy. The acrosomal process observed with scanning electron microscopy appears as an elongated fiber and it establishes the initial contact with the surface of the unfertilized egg. S. purpuratux. Bar: 500 nm. (Reprinted with permission from Schatten and Mazia, 1976b.)

brane fusion is noted (Fig. 3B). The earliest stages of sperm-egg plasma fusion depict the erect sperm on the egg surface apparently attached by the adherent extracellular material. Small vesicles are noted at the fusion site (Fig. 4A-C). Microvilli adjacent to the successful sperm elongate and cluster around the sperm head to form the fertilization cone (Fig. 5 ) . The sperm during incorporation is obscured by the elevation of the fertilization coat (Fig. 6A and B).

MOTILITY DURING FERTILIZATION

67

FIG. 3. Sperm-egg attachment. (A) Scanning electron microscopy. The sperm attach perpendicularly to the vitelline sheet of the egg surface. The short. arrayed microvilli are characteristic of an unfertilized egg. The sperm always adhere to the egg surface by the apical tip of the sperm head. S. purpi~ratus.Bar: I pm. (Reprinted with permission from Schatten and Mazia. 1976b.) ( B ) Transmission electron microscopy. The attachment of the sperm to the egg surface is quite apparent and appears to involve the amorphous extracellular material. bindin. Bundles of cortical microfilaments are present in this image and are the result of our having activated this egg prior to insemination. L. picfirs. Bar, 500 nm. (From Mann c f a / . . 1976; micrograph courtesy of D. S. Friend.)

In addition to signaling the egg to begin the process of sperm incorporation, the successful fusion of the sperm with the egg initiates the cortical reaction (see next section), which results in the elevation and hardening of the fertilization coat. The elevation of the vitelline layer around the successful sperm, which then hardens into the electron-opaque fertilization coat, precludes direct observations with scanning electron microscopy at the plasma membrane surface of the events

68

GERALD SCHATTEN

FIG.4. Sperm-egg membrane fusion. (A) Transmission electron microscopy. The initial stage of membrane fusion involves a minor upwelling from the egg surface and the appearance of dense material at the fusion channel. Bindin, the amorphous material extruded with the acrosome reaction, appears to join the sperm to the egg. Note the small vesicles in the egg at the fusion site. L . picrus. Bar, 250 nm. (From Mann et a / . . 1976; micrograph courtesy of D. S. Friend.) (B) Transmission electron microscopy. The next phase of membrane fusion is an increase in the egg upwelling, which now takes the shape of a small cone. Now numerous vesicles and membrane trabeculae surround the site of fusion in the egg. The bindin material clearly connects the gametes. L. picrus. Bar, 250 nm. (Reprinted with permission from Mann er a / . , 1976; micrograph courtesy of D. S. Friend.) ( C ) Scanning electron microscopy. A later stage of membrane fusion. The membrane derived from the egg now surrounds the anterior portion of the spermhead. S. purpurutus. Bar, I Fm. (Reprinted with permission from Schatten and Mazia, 1976b.)

during sperrn incorporation (Fig. 6A and B). To overcome this steric problem, eggs can be denuded of their vitelline layers with disulfide-reducing agents such as dithiothreitol (Epel e? af., 1970) and then studied with scanning electron microscopy (Schatten and Schatten, 1980a). The basic ultrastructural features of

MOTILITY DURING FERTILIZATION

69

FIG. 5 . Scanning electron microscopy. The egg membrane continues to rise around the spermhead. Microvilli elongate around the spermatozoon as the membrane derived from the sperm appears slack and convoluted. S. purpururus. Bar, I pm. (Reprinted with permission from Schatten and Mazia, 1976b.3

sperm incorporation are analogous in these denuded eggs to hose observed in untreated controls. The initial contact again is via the extended acrosomal process; sperm incorporation starts with the localized elongation of microvilli around the sperm head (Fig. 7A). These microvilli elongate around the sperm head, the midpiece, and, surprisingly, the sperm tail, to engulf the entire spermatozoon and form into the fertilization cone (Fig. 7B), which at this stage is apparent in light microscopy. The sperm head and midpiece rotate at the egg surface to lie parallel with the egg cortex during incorporation (Fig. 8). Note the continuity of the sperm-egg plasma membrane in Fig. 8. Since scanning electron microscopy is limited to the study of cell surfaces, it is not possible to follow the sperm through the egg plasma membrane directly during incorporation. The only means by which the investigator could exploit the resolving power and depth of focus of the scanning electron microscope was to study cellular fractions or isolates. In the next paragraphs, advances in under-

10

GERALD SCHATTEN

FIG. 6. Scanning electron microscopy. (A) The head of the spermatozoon has almost completed its entry as the deformation on the surface in the region of the sperm has spread. The indentations in the tips of the papillae of the vitelline sheet around the entering spermatozoon may have resulted when the underlying microvilli were withdrawn. The 50-nm strands now radiate from the penetration site. S. purpurarus. Bar, I pm. (Reprinted with permission from Schatten and Mazia. 1976.) (B) Finally, the spermhead and midpiece are pulled from view, leaving only the spermtail protruding as the fertilization coat starts to elevate. S. purpurarus. Bar. I pm. (Reprinted with permission from Schatten and Mazia, 1976b.)

standing the surface events at the inner face of the egg surface during incorporation are reviewed by study of the cortical surface of isolated egg surfaces irnmediately following insemination. Vacquier (1975) developed an ingenious method for isolating the unfertilized egg surface by affixing eggs to cationic surfaces and shearing the tops of the eggs off in a zero calcium environment. These cortical “lawns” could be induced to undergo secretion in vitro by the addition of calcium ions. By a modification of this method and the development of an isolation medium that mimicked the intracellular environment of the unfertilized egg surface, Schatten and Mazia (1976a) were able to isolate entire unfertilized egg surfaces in suspension. When this egg surface isolation was performed 1 minute after insemination, the ultra-

MOTILITY DURING FERTILIZATION

71

FIG. 7. Scanning electron microscopy at the egg plasma membrane. (A) In these eggs, devoid of their vitelline layers, the activity of the egg surface in engulfing the sperm is clearly apparent. Microvilli have elongated, to 1.2 pm, to completely surround the successful sperm. These microvilli will continue to elongate to form the fertilization cone. L . variegarus. Bar, 1 pm. (Reprinted with permission from Schatten and Schatten. 1980a.) (B)The fertilization cone forms from these elongating microvilli, which surround the base of the fertilization cone, and which continue to engulf the sperm. Note the microvilli sourrounding the sperm tail. L. varicgarus. Bar, I pm. (Reprinted with permission from Schatten and Schatten, I980a.)

structural features of the sperm moving through the egg surface could be captured with scanning electron microscopy. In Fig. 9A, the phase contrast image of an unfertilized egg surface isolated in suspension is depicted. Figure 9B is a low magnification scanning electron micrograph of an egg surface isolated 1 minute after insemination in which the sperm is captured within the egg surface. In this

72

GERALD SCHATTEN

FIG. 8. Transmission electron microscopy. Following fusion, the sperm rotates during incorporation. In this image, the rotated sperm resides just beneath the egg plasma membrane. The sperm centriole, microtubules comprising the axoneme, and a cross section of the mitochondrion are depicted. L. picfus. Bar, 250 nm. (From Mann e t a / . . 1976; micrograph courtesy of D. S. Friend.)

scanning electron micrograph, the outer faces of the egg are apparent at the left and right. The egg cortex with its adherent cortical granules is apparent and toward the top of the image a patch devoid of cortical granules through which the sperm is entering is dmumented. The isolation medium employs a calcium-free environment; had calcium been present the exocytosis of the cortical granules would have continued to radiate from the site of sperm-egg fusion. In Fig. 10, a higher magnification image of the sperm during incorporation demonstrates an intricate array of fibrous netting. This netting is extractable with 0.6 M KI, and the extract contains an electrophoretic band comigrating with rabbit muscle actin

MOTILITY DURING FERTILIZATION

73

(see Section 111,B,3),inviting speculations about the role of microfilaments during sperm incorporation. In addition to the development of microfilamentous cortex, numerous small vesicles, arrayed circularly, are frequently observed at the site of sperm-egg fusion (Fig. 11). These vesicles appear to be derived from the sperm acrosome and could be involved in the initial triggering of egg activation. Recent advances in video microscopy using differential interference contrast optics has permitted the recording of the movements during fertilization in the living state (Schatten, 1981a,b). In Fig. 12, the entire sequence of fertilization is documented and the motions during sperm incorporation can be resolved. Unlike electron microscopy, which requires the study of fixed, and therefore static, specimens, video tape recording permits a relatively high degree of resolution in the living state in which the sequence of fertilization is directly observed rather than compiled from a sequence of static micrographs. In this figure, the initial stages of sperm incorporation are observed as involving first the attachment of

FIG. 9. The cortical view of insemination. (A) A phase-contrast micrograph of an isolated surface cortex. Although empty, these surfaces tend to maintain the spherical shape of the egg. The cortical granules can be observed as the small granules underlying the egg surface. S. purpurarus. Bar, 10 p.m. (Reprinted with permission from Schatten and Mazia, 1976b.3 (B)Scanning electron microscopy of a surface isolated shortly after fertilization. The cortical granules are discharged around the site of sperm entry. The outer surface can be observed at the right and left where the surface is folded over. S. purpurutus. Bar, 10 p.m. (Reprinted with permission from Schatten and Mazia. 1976b.)

74

GERALD SCHATTEN

Fic. 10. Scanning electron microscopy of the cortical surface during insemination. The intimate association between these cortical elements and the membraneless spermatozoon can be observed as the spermatozoon continues to rotate through the egg surface. This rotation appears to star: at the anterior end of the sperm. S. purpururus. Bar, 2 Fm. (Reprinted with permission from Schatren and Mazia, 1976b.)

the sperm by the acrosomal process to the egg surface. Following a varying time when the sperm gyrates about its attachment site, the sperm stands erect on the egg surface and the motility of the sperm tail ceases. Moments later, the elevation of the fertilization coat around the successful sperm occurs and the unsuccessful sperm attached to the egg surface are lifted from the plasma membrane by the elevation of this extracellular coat. The fertilization cone begins to form around the perpendicularly oriented and static spermatozoon. Shortly afterward, the sperm rotates 90" to lie parallel with the egg cortex and it then begins to undergo a lateral displacement along the egg cortex from the site of sperm-egg fusion. Concomitant with this lateral displacement the erratic beating of the sperm tail is observed, perhaps causing the displacement. The momentary arrest in tail beating and the resumption of this example of ciliary motility may be

MOTILITY DURING FERTILIZATION

75

FIG. 1 I . The vesicular coronet of fertilization. In these remarkable transmission electron and scanning electron micrographs, a circle of about 20 small membrane vesicles is demonstrated at the site of sperm-egg fusion, presumably derived from the sperm acrosomal membrane. (A-D. From Mann e r a / ., 1976; micrograph courtesy of D. S . Friend; E, reprinted with permission from Schatten and Mazia, 1976b.) (A) Transmission electron microscopy. The vesicles apparently budding off the sperm acrosome from this incorporated sperm. L . pictus. Bar, 500 nm. (B) Transmission electron microscopy. The vesicles now oriented in the acrosomal fossa. L. pictus. Bar, 500 nm. ( C )Transmission electron microscopy. A cross-section at the fusion site demonstrating numerous vesicles and trabeculae. L. picrus. Bar, 500 nm. (D) Transmission electron microscopy. A cross-section documenting the circular array of vesicles just below the fusing sperm. L. pictus. Bar, I pm. (E) Scanning electron microscopy. When it is compared with the transmission electron micrograph in (D) the significance of this circle of membrane vesicles becomes intriguing. Here, during monospenny, the circle is found at the center of a patch devoid of cortical granules. Speculations regarding the role of these vesicles in triggering egg activation and the cortical reaction seem attractive. S. purpuratus Bar. 2 pm.

76

GERALD SCHATTEN

FIG. 12. Movements during fertilization studied in living eggs. Time-lapse video microscopy of fertilization with water immersion, differential interference contrast optics. Sperm-egg attachment occurs in ( A ) at 1:36 (minutes:seconds). The sperm tail becomes immotile in ( B ) and a second later the fertilization coat (white arrow) elevates over the attached sperm (black arrow). The fertilization cone forms around and above the erect sperm in (F-I). The static sperm tail, which projects through the elevated fertilization coat, can be observed in (E-H).The displacement of the sperm nucleus (male pronucleus) within the egg cytoplasm occurs in (I-P); the sperm tail beats erratically at this stage. The sperm aster forms as the male pronucleus is moved centripetally (Q-U). In (V),the field has been shifted to include the sperm aster (large black arrow) and the female pronucleus. Fibers radiating from the sperm aster are denoted by black v's in (T-CC). The migration of the female

MOTILITY DURING FERTILIZATION

77

indicators of changing ionic conditions (see Section V,B,3). Following this rotation of the sperm along the egg cortex, the sperm is discharged into the egg cytoplasm proper with its midpiece directed toward the egg center (see Fig. 17A). Interestingly, a similar pattern of motile events during sperm incorporation appears to occur during mammalian and amphibian fertilization. In amphibians the sperm is only able to enter by a microvilli-rich site (Campanella, 1975; Elinson, 1980; Elinson and Manes, 1978; Picheral, 1977). In mammals, egg microvilli adjacent to the successful sperm elongate and cluster around it (Shalgi and Phillips, 1980a,b; Shalgi et al., 1978; Yanagimachi, 1978), and then the sperm glides and rotates at the egg cortex (Wolf and Armstrong, 1981). In summary then, the events during sperm incorporation, following the swimming of the sperm to the egg surface and the contact by the extruded acrosomal process of the sperm to the egg, involve first the formation of the fertilization cone around the erect and static sperm, and then the rotation and lateral displacement of the sperm along the egg cortex, which discharges it into the egg cytoplasm in a rotated position so that its centriole end faces toward the egg center. This latter point will be of importance when the significance of the centrioles contributed by the sperm during the pronuclear migrations (see Section ll,A,3) is considered and during later development (see Section VI).

2 . Corricul Restrucluring The egg surface at fertilization undergoes a complex series of modifications and alterations, which are crucial in converting the unfertilized egg to a fertilized zygote and which affect the manner in which the egg interacts with its environment. Various surface components have been isolated and characterized: the vitelline layer (Glabe and Vacquier. 1977). the plasma membrane with attached cortical granules (Detering et a / ., 1977), and the entire unfertilized cortex (Vacquier, 1975). The biochemical features of this surface restructuring have recently been reviewed by Shapiro and Eddy (1980). To understand the surface events at fertilization it is first essential to review the layers covering the unfertilized egg. Beneath the jelly coat, which likely plays an important role in triggering the acrosome reaction of the sperm, is the vitelline layer. The vitelline layer is draped over the plasma membrane of the unfertilized egg and conforms to its topography. Beneath the plasma membrane of the unfertilized egg and attached to it are a monolayer of about 1 k m secretory pronucleus to the center of the sperm aster occurs in (W-Z): the female pronucleus is distorted from a sphere to an ovoid during this migration. Pronuclear centration (AA-BB) occurs as the fibers of the sperm aster (black v’s) continue to elongate. Small particles (black triangles) appear on the nuclear surface in (BB); these particles may represent the centrioles since they are positioned along the presumptive axis for mitosis. Cleavage (DD) occurs parallel to the direction of pronuciear centration. L . variqatus. Bar. 10 pm. (Reprinted with permission from Schatten. 1981b.)

78

GERALD SCHAlTEN

FIG. 13. The unfertilized egg surface. In this isolated egg surface, observed by scanning electron microscopy, the various layers are depicted. At the bottom, the casts of microvilli forming papillae in the vitelline layer are apparent. The vitelline layer drapes over the egg plasma membrane and conforms to its topography. At the top of the image, viewing the inner cortical face of the egg surface, are numerous cortical granules, which appear attached to the plasma membrane by fibrous elements. Immediately following sperm-egg fusion, these cortical granules fuse their membranes with the plasma membrane and the resultant secretory event elevates the fertilization coat. 5’. purpuratus. Bar. 2 p n . (Reprinted with permission from Schatten, 1975.)

granules, the cortical granules (Fig. 13; Anderson, 1968; reviewed by Anderson, 1974; Schuel, 1978). At the moment of sperm-egg fusion, the cortical granules fuse their membranes with the plasma membrane in a wave-like motion starting at the site of sperm-egg fusion and radiate to encompass the entire egg surface. The fusion of the cortical granules with the plasma membrane, the cortical reaction, externalizes the contents of the cortical granules into the space between the vitelline layer and the plasma membrane, the perivitelline space. The contents of the cortical granules laminate the vitelline layer, elevating it and hardening it to form the fertilization coat (Chandler and Heuser, 1979, 1980, 1981; Foerder and Shapiro, 1977). The membrane added during the cortical reaction may well be resorbed by clathrin-coated vesicles (Fischer and Rebhun, 1981). The fertilization coat serves as a barrier between the developing embryo and the

MOTILITY DURING FERTILlZATION

79

external environment, protecting it from supernumerary sperm and bacterial infestation ( M . Daniels, unpublished results). In addition to the secretion of the cortical granules the cell surface undergoes a series of motile events in addition to that of sperm incorporation (discussed in the previous section). The egg cortex has recently been carefully reviewed by Vacquier (1981). The microvilli on the unfertilized egg are short, stubby, and well arrayed. Following fertilization, the microvilli adjacent to the sperm elongate to form the fertilization cone, and are essential for proper sperm incorporation. Additionally, the remaining microvilli on the egg surface undergo two bursts in microvillar elongation (Schroeder, 1979); the first burst occurs within 5 minutes of the sperm-egg fusion, associated with the addition of the cortical granule membrane to the plasma membrane, and then the second burst (Fig. 14) occurs prior to mitosis, perhaps resulting in the increase in surface area necessary for

Fic;. 14. Elongated microvilli, Following fertilization, the egg microvilli undergo microfilament-mediated bursts in microvillar elongation. In this scanning electron micrograph of an egg dcnudcd of all surface layers at 40 minutes following insemination. the elongated pattern of microvilli is striking when compared to the previous imagc of the unfertilized egg surface. L. \,arie,qutrts. Bar. I pm. (From Schatten e/ a / . . 1981a.)

80

GERALD SCHATTEN

cell division. Cytokinesis, of course, is another cyclical event mediated by cortical motility. Jacques Loeb (1913) was one of the earliest workers recognizing the crucial importance of the cortical rearrangement to the onset of development. He regarded the cortical reaction as a cytolytic process, an analogy that is not unfair considering the dramatic changes in physiognomy between the surfaces of unfertilized and fertilized eggs. The requirements for first the rapid (Jaffe, 1976) and then the permanent blocks to polyspermy (Vacquier er al., 1972a,b; reviewed by Dale and Monroy, 1981, and Wolf, 1981) are perhaps obvious reasons for this dramatic alteration of surface features as the unfertilized egg progresses to the fertilized state. Still unexplained surface changes, shown to be correlated with fertilization or artificial activation (Mazia et al., 1975b; Spiegel and Spiegel, 1977) and required for the proper completion of the first cell cycle (Schatten and Schatten, I98 I ) , are the changes in microvillar length. The changing configurations of the surface microvilli have been studied by scanning electron microscopy (Eddy and Shapiro, 1976; Mazia er al., 1975b; Spiegel and Spiegel, 1977) and the presence of microfilaments as the underlying substructure has been demonstrated by transmission electron microscopy (Longo and Anderson, 1968; Kidd et al., 1976; Burgess and Schroeder, 1977; Tilney and Jaffe, 1980). Interestingly, the sperm-induced cortical reaction appears insensitive to cytochalasin B, an inhibitor of microfilament assembly, which itself has been reported to cause limited exocytosis (Longo, I978a; Schatten and Schatten, 1980a). However, the elongation of the egg microvilli is sensitive to this inhibitor (Eddy and Shapiro, 1976; Longo, 1980; Schatten and Schatten, 1980a). Additionally, the progression of the fertilized egg through the first cell acycle requires the proper restructuring of the egg cortex, which itself is sensitive to cytochalasin B (Schatten and Schatten, 198I ) . In summary then, the surface modifications occurring during fertilization include sperm incorporation and then the elevation of the fertilization coat resulting from the cortical reaction. Two bursts of microvillar elongation occur during the first cell cycle, and changes in total surface area have been noted. It appears likely that one function of the alteration in microvillar length is to provide a means first to secure and later to resorb the membrane added by the cortical granules following insemination and required during development. 3. The Pronuclear Migrations For fertilization to be successful, one, and only one, sperm must be incorporated into the egg cytoplasm. Following penetration, the sperm and egg must be translocated through the egg cytoplasm into a proximity close enough to permit the fusion of their nuclear membranes. In this section, the migrations of the male pronucleus (sperm nucleus) and female pronucleus (egg nucleus) will be traced from the moment when the sperm leaves the egg surface following incorporation

MOTILITY DURING FERTILIZATION

81

until the pronuclei fuse. Syngamy completes the fertilization process. The terminology used throughout this article will be to refer to the unincorporated sperm nucleus as a sperm nucleus; the sperm nucleus within the egg cytoplasm will be referred to as the male pronucleus. The egg nucleus will be referred to as the female pronucleus. The documentation of the pronuclear migrations has been a difficult undertaking since most eggs are relatively opaque because of the presence of numerous yolk platelets and since the incorporated male pronucleus migrates centripetally from the surface where it is visible into the egg center where it is not. In recent years eggs have been gently compressed to study the pronuclear migrations with greater clarity. Modern sophistication in optics has increased the visibility of the pronuclei and their motile structures, and quite importantly the nearly transparent egg of the Gulf Coast sea urchin Lyrechinus vuriegutus has contributed greatly in the living documentation by virtue of its glass-like characteristics. The pronuclear movements at fertilization involve the formation of [he sperm aster, which is initially a radially symmetrical structure emanating from the sperm centrioles at the base of the rotated incorporated sperm midpiece. The formation of the sperm aster moves the male pronucleus centripetally at a rate of 4.9 pm/minute (Table I). Concomitant with this centrad motion the male pronucleus begins to undergo chromatin decondensation. When the rays of the sperm aster contact the surface of the female pronucleus, the next of the three pronuclear migrations occurs, i.e., the migration of the female pronucleus (Fig. 15). The movement of the female pronucleus to the center of the sperm aster is the swiftest and most dramatic of the pronuclear migrations occurring at a rate of 14.6 pndminute, often traversing half the diameter of the egg. The final movement of the now adjacent pronuclei is again dependent on the extension of the sperm astral rays, which push the pronuclei to the egg center. The fusion of the pronuclei typically occurs at the egg center shortly after the sperm aster has reached its maximal size. This final motion occurs at a rate of 2.6 pm/minute. The study of the movements during fertilization parallels the history o f niicroscopic developments and recording methods. In the last century (Wilson and Mathews, 1895; Wilson and Learning, 1895; reviewed by Wilson. I923 classical cytological preparations of fixed materials were studied extenkively. During the 1930s camera-lucida drawings (R. Chambers, 1933; E. Chambers, 1939) were employed. Cinematography has been used by Kuhl and Kuhl ( 1949), Allen (1958), and Rothschild (1956) with echinoderms and by Brackett (1970), Borghese and Cassini (1963), Kuhl and Friedrich-Freksa (1936), Lewis and Gregory (1929), Mulnard (1967), and Ries (1909) with mammals. With the advent of phase contrast microscopy, Dan (1950) studied sperm incorporation and recently differential interference contrast microscopy has been employed to study the movements of the nuclei (Schatten, 1979, 1981a,b; Mar, 1980; Hamaguchi and Hiramoto, 1980). With the recent advances in vidco tape record-

82

GERALD SCHATTEN

FIG. 15. The movement of the female pronucleus. Oil immersion, differential interference optics with the compression chamber. Time (minutes:seconds) lower right. Bar, 10 pm. (Reprinted with permission from Schatten. 1981a.) (A) Astral rays (black v's) emanate from the male pronucleus (black arrow) as the sperm aster begins to form. White bar = 10 pm. (B) Prior to the onsei of the migration of the female pronucleus (black triangle), astral rays (black v's) appear to interconnect the male pronucleus (black arrow) and the female pronucleus. (C) Rays appear to radiate from the female pronucleus (white v's) as well as from the male pronucleus (black v's). (D) Motion is initiated by an erratic movement of the female pronucleus (black triangle). (E) The fibers interconnecting the two pronuclei (black v's) are apparent. (F) A protuberance (double white arrow) on the surface of the female pronucleus forms as the female pronucleus begins its migrations. (G) the shape of the female pronucleus is distorted as movement occurs. (H) The fibers emanating from the male pronucleus (black v's) and projecting past the female pronucleus (white v's) are apparent. (I) The interconnecting fibers (black v's) are more prominent, as are the fibers projecting past the female pronucleus (white v's). (J) The female pronucleus retains the oblate form during this movement. Fibers (white v's) trailing from the female pronucleus can be observed. (K) The migration of the female pronucleus is nearing completion. (L) At the conclusion of the movement of the female pronucleus, it again assumes a spherical shape. (M) Pronuclei in contact. (N) Zygote nucleus; nucleolus (black arrow). (0)Nuclear breakdown at prophase initiates at the region of the zygote nucleus originally derived from the male pronucleus (white arrows). (P) Nuclear breakdown spreads from the disintegrated region (white arrows). (Q) Nuclear breakdown at prophase. Only a small portion of the nuclear envelope remains intact (white arrows). (R) Nuclear breakdown is complete. L . variegatus.

MOTILITY DURING FERTILIZATION

83

ing and contrast enhancement (reviewed by Allen et al., 1981a,b; InouC, 198 la), future studies of living cells will undoubtedly rely on the increased resolution, sensitivity, spectral capability, and versatility of a video recording through the microscope of living eggs and zygotes at fertilization. With the knowledge of the motions occurring at fertilization, the biochemical and ultrastructural nature of the motile apparatus and the manner for transducing force have been explored by transmission electron microscopy (Longo and Anderson, 1968; Mann et al., 1976; Harris, 1979), by indirect immunofluorescence microscopy with monospecific tubulin antibody (Harris et al., 1980a,b; Bestor and Schatten, 1981), and by isolation of the structures responsible for these movements (Schatten and Mazia, 1976a,b; Schatten and Schatten, 1979). The detection of microtubules and microfilaments during the movements at fertilization is considered in Section Ill.

4. Movements Leading to Cell Division Once fertilization has been successfully completed, signaled by the proper fusion of the male pronucleus with the female pronucleus, the task of the zygote is to initiate the cell divisions that will permit the morphogenetic motions leading to embryogenesis and later development. A full discussion of the theories concerning the process at mitosis and cytokinesis are beyond the scope of this article. Interested readers are directed to now classic (Mazia, 1961; Wilson, 1925) and more recent reviews (InouC, 1981b; Forer and Zimmerman, 1982). The events of cell division concern us insofar as the manner of fertilization has consequences on the characteristics of the resultant cleavage, in that fertilization triggers the onset of development which requires division and the morphogenetic movements, and since the motions at fertilization are likely to be a paradigm for intracellular translocations in general and the movements during mitosis and cytokinesis specifically. To appreciate the events leading to first division it is important to analyze the contribution of each gamete to the resultant zygote. Of course, each has a haploid genome that will be required for later development. The egg, containing a large pool of microfilament and microtubule precursors, does not appear to have any existing centrioles; a pair are contributed by the sperm at insemination. This point is underscored by the problems confronting a polyspermic egg, when instead of having a pair of centrioles contributed by one sperm and leading to a bipolar mitotic apparatus, each supernumerary sperm contributes a pair, leading to a multipolar mitotic apparatus and an abortive multiple cleavage attempt. Each contributes a pair of centrioles, all of which attempt to organize asters at the first division. The significance of the contributed sperm centrioles and the manner in which they can be manipulated are fully discussed in Section VI. The pair of sperm centrioles is directed toward the egg center when the sperm rotates during sperm incorporation. At first, a monopolar structure, the sperm

84

GERALD SCHATTEN

aster forms; it will play crucial roles in the pronuclear migrations. During the final stages of the growth of the sperm aster this monastral structure develops two focal points because of the separation, and perhaps even replication, of the pair of sperm-contributed centrioles. This separation of the pair of sperm centrioles occurs perpendicular to the direction of pronuclear centration and will be crucial in the establishment of the first embryonic axis. The sperm aster typically disassembles prior to syngamy and as the cell progresses through the first cell cycle a planar apparatus forms immediately prior to prophase. This structure has been referred to as the “streak,” the “interim apparatus,” or the “interphase asters” and is rather transitory in nature. The axis of the streak is typically perpendicular to the final motion of the sperm aster and is usually parallel with the mitotic axis. The streak distorts the swollen spherical zygote nucleus and typically disassembles to permit the zygote nucleus to reform its spherical shape immediately prior to nuclear envelope breakdown. Mitosis involves the well-known scheme of astral formation at the poles and chromosome condensation moving the paired chromosomes to the metaphase plate when first the anaphase movements of the chromosomes and later the separation of the poles occur. In addition to these cytoplasmic events during the first cell cycle, there are also cortical changes. Tracing the surface alterations from fertilization, microvilli form into the fertilization cone surrounding the successful sperm when the cortical reaction propagates over the egg surface. The remaining egg microvilli undergo their first burst of elongation within 5 minutes of insemination. Following this first burst of elongation, the net surface area is reduced, presumably because of the resorption of the added cortical granule membrane. Prior to prophase and around the streak stage the microvilli undergo a second burst of elongation. Following mitosis, the cell surface undergoes perhaps the most impressive cortical change, namely, first cleavage, when the contractile ring divides the fertilized egg in two.

5 . The Movements during Fertilization Before exploring the biochemical and ultrastructural components involved in the movements at fertilization, and their molecular and ionic regulators, the motions during fertilization are described in Table 11 and are depicted schematically in Fig. 38. The final schematic (Fig. 38) is based primarily on observations of living recordings rather than on the sequencing of fixed specimens compiled into what might appear to be a reasonable sequence. The rates and timing of fertilization are described in Table 11. The beating of the sperm tail propels the spermatizoon to the egg surface. Near or perhaps even at the egg surface the acrosome reaction occurs whereupon the acrosomal process is extruded from the apex of the sperm head. This process establishes the initial contact between the gametes by effectively “harpooning” the egg surface. The sperm, attached by its acrosomal process, continues to beat actively, resulting in gyration of the sperm

MOTILITY DURING FERTILIZATION TABLE 11 EFFECTSOF MICROFILAMENT INHIBI.I-OKS A.

DURING

FERIII.IZATION

Assembly inhibitors

Cytochalasin B ( 2 = 2 7 )

w

25

-

15 to 0 minutes

+ I minute

+ 2 minutes 25 phi (DTT-treated eggs) - IS to 0 minutes 10

w

- 1 5 minutes

Cytochalasin D (22) 10 - 15 to 0 minutes

Sperm incorporation blocked. aberrant egg cortex at +45 minutes Resorption of fertilization cone. aberrant egg cortex at +45 minutes Normal pronuclear migrations. syngamy. egg cortex. and mitosis; cytokinesis blocked Sperm incorporation blocked. aberrant egg cortex at +45 minutes Sperm incorporated. aberrant egg cortex at +45 minutes

w

+I

minute

10 phf (DTT-treated eggs) - I S to 0 minutes

5w - 15 minutes Cytochalasin E ( 5 ) 10 -15 to 0 minutes

Sperm incorporation blocked. aberrant egg cortex at +45 minutes Resorption of fertilization cone. aberrant egg cortex at +45 minutes Sperm incorporation blocked. aberrant egg cortex at +45 minutes Sperm incorporated. aberrant egg cortex at +45 minutes

w

+ I minute

+ 2 minutes 5P.M - 15 minutes

Sperm incorporation blocked. aberrant egg cortex at +45 minutes Resorption of fertilization cone. aberrant egg cortex at +45 minutes Normal pronuclear migrations, syngamy. egg cortex. and mitosis; cytokinesis blocked Sperm incorporated. aberrant egg cortex at +45 minutes B.

Phalloidin (9) ImM -60 minutes

Disassembly inhibitor

Retarded rate of sperm incorporation. prominent fertilization cone. cytokinesis arrested

85

86

GERALD SCHATTEN

about its attachment site on the egg surface. A varying time later, sperm incorporation occurs, characterized first by the sudden immobilization of the sperm tail with the sperm head and midpiece held in an erect and perpendicular fashion on the egg surface. The fertilization cone begins to form on the egg surface at the site where the sperm head is attached and the fertilization coat elevates over the attached sperm and propagates from that site of attachment to envelop the now fertilized egg. Unsuccessful sperm attached to the surface of the fertilization coat are physically removed from the egg surface by the elevation of the fertilization coat. Sperm incorporation involves first the formation of the fertilization cone around the stationary and erect sperm and then later the rotation and lateral displacement of the sperm head, midpiece, and tail along the egg cortex. Though the sperm tail is immotile at the earliest stages of sperm incorporation it begins to beat shortly later and continues to beat within the egg cytoplasm in an erratic fashion for the next few minutes. It should be noted that in virtually all recent studies the sperm tail has been found to be fully incorporated into the fertilized egg cytoplasm. To summarize then, the pronuclear migrations begin with the formation of the sperm aster emanating from the base of the sperm head and midpiece. The sperm aster first pushes the male pronucleus centripetally and, upon contact with the female pronucleus, it pulls the egg nucleus to the center of the sperm aster. The now contiguous pronuclei are moved to the center of the egg cytoplasm whereupon pronuclear fusion occurs. The remaining cytoskeletal alterations during the first cell cycle are characterized by another burst of microvillar elongation and by the formation and regression of the streak prior to the events at cell division.

111. Detection of Cytoskeletal Elements during Fertilization

By their very nature, eggs are not very amenable to the conventional means of detecting microtubules and microfilaments. Thin section, transmission electron microscopy of eggs is fraught with problems, all resulting from the relatively huge size of eggs. Fixation is quite slow and even in the best of cases the resultant micrographs do not always correlate well with phase contrast and tiifferential interference contrast images of living eggs or eggs processed for immunofluorescence microscopy. With thin sectioning, only a small slice of the huge egg can be visualized; it is indeed unlikely that the sperm aster will be sectioned in a plane in which both pronuclei are viewable. The problems of thin sectioning have been partially overcome by the use of the relatively small egg of Arbacia punctulata (Longo, 1973). In contrast, irnmunofluorescence microscopy with rnonospecific antibody, e.g., tubulin antibody (Wassarman and Fujiwara, 1978; Harris er ul., 1980a,b; Bestor and Schatten, 1981, 1982), is similarly troubled by the relatively slow diffusion of antibodies into and out of the fertilized egg and is additionally complicated by the hardened fertilization coat following insemina-

MOTILITY DURING FERTILIZATION

87

tion. However, with the advent of immunofluorescence microscopy for the study of global reorganization in cytoskeletal elements and thin section transmission electron microscopy for the confirmation of the ultrastructural features of each component the configurations of microtubules and microfilaments in the egg during fertilization can be accurately compiled. Future advances in preparative methods for electron microscopy, including rapid freezing and deep etching (Chandler and Heuser, 1981) as well as studies using thick section high-voltage transmission and scanning transmission electron microscopy, promise to reveal the finer details of the cytoskeleton during fertilization. A. SPERM The cytoskeletal arrangement of the sperm is of crucial importance in light of its partnership role at fertilization. The information concerning the regulation of the acrosome reaction and ciliary beating might also prove helpful in understanding the means by which the egg regulates its cytoskeletal rearrangements, and of course the contribution, directionality, and state of cytoskeletal elements contributed by the sperm at fertilization influence the manner in which sperm incorporation and the pronuclear migrations are liable to be effected. The 9+2 configuration of the sperm tail has long been a model system for the study of ciliary motion in general (see reviews by Gibbons, 1977, 1981; Satir, 1974). Though beyond the scope of this article, the sliding of adjacent microtubule pairs mediated by the ATPase dynein (Sale and Gibbons, 1979) results in the lateral translocation of one ciliary doublet past another. Presumably, elastic spoke proteins, referred to as nexin, convert this lateral movement into bending waves. The plasma membrane of the sperm plays a crucial role in regulating its motility (reviewed by Friend, 1982). Factors released by eggs (reviewed by Garbers and Kopf, 1980) have been shown to affect cyclic nucleotide levels within the sperm and result in alterations in beat frequency. Interestingly, Afzelius (1976) has identified a class of sterile men in which the dynein cross bridges are absent; the sperm are immotile. Gibbons and Gibbons (1979) have described the cessation of tail beating in the presence of high calcium, a point which will have a bearing in analyzing the sudden immobilization of the sperm at the onset of sperm incorporation. Tihey and co-workers (Tilney and Kallenbach, 1979; Tihey et a / . , 1973, 1978) have studied the polymerization of actin in the sperm head which forms the extruded acrosomal process. Somewhat analogous to the centriole in the sperm tail, the actomere (Tilney, 1978) appears responsible for nucleating the initial assembly of these microfilaments. During the acrosome reaction the secretion of the acrosomal vesicle and the polymerization of the acrosomal process occur almost simultaneously; indeed, recent video tapes by Tihey and InouC (1982) provide insights into the manner in which these events are coupled. It appears

88

GERALD SCHA'ITEN

that the initial polymerization of the acrosomal process pushes it through the acrosomal vesicle, which then coats the extending process with secreted contents; the driving force may well be the influx of water that hydrates the actin in the periacrosomal cap (Tilney and InouC, 1981). From the studies of Schackman et al. (19781, Tilney et al. (1978), and Cantino (1982), it appears that calcium ions and protons both play essential roles during the acrosomal reaction. It may be that calcium ions are responsible for the secretion of the acrosomal vesicle, whereas the change in intracellular pH is predominantly responsible for the polymerization of the microfilaments. B. EGG 1. Transmission Electron Microscopy Since the earliest descriptions of the movements at fertilization (reviewed by Wilson, 1925) the similarity between the sperm aster and asters of the mitotic apparatus have been noted; these morphological similarities are a direct result of their composition. Early transmission electron microscopy using polyspemically inseminated oocytes (Franklin, 1965) and the later investigation of monospermy (Longo and Anderson, 1968) clearly revealed microfilaments in the fertilization cone parallel with the entering spermatozoon. At later stages of insemination the microtubules of the sperm tail and microtubules forming the rays of the sperm aster were characterized. In the work by Longo and Anderson (1968), the distortion of the female pronuclear surface and the ultrastructural features of pronuclear fusion were noted. The ultrastructural features of in vitro fertilization in several mammals have been documented by Anderson et al. (1975), Brackett et al. (1980), Longo (1976), Longo and Anderson (1969), Yanagimachi and Noda (1970), and Zamboni (1972). The ultrastructural detection of microfilaments in eggs has undergone a resurgence in the last years as a result of the interest in nonmuscle motility and because isolation of the egg cortex is now routinely attainable. In the early work on this topic, microfilaments were found in the fertilization cone and recently Tilney and Jaffe ( 1980) have produced interesting images of microfilaments and microfilament bundles in fertilization cones of oocytes and eggs fixed in the presence of nicotine. In addition to the presence of microfilaments in the fertilization cone itself, microfilaments comprise the substructure of the egg microvilli following fertilization and form the basis of the egg cortex. Though few, if any, microfilaments are present in the short microvilli of unfertilized eggs, the microvillar elongation following the cortical reaction or artificial activation is dependent on the polymerization of actin into microfilaments (Kidd et al., 1976; Burgess and Schroeder, 1977; Chandler and Heuser, 1981; Eddy and Shapiro, 1976; Kidd, 1978; Mann et al., 1976; Schroeder, 1978a, b; Spudich and Spudich, 1979; Spudich and Amos, 1979). These microfilaments can be isolated

MOTILITY DURING FERTILIZATION

89

with the egg surface (Schatten and Mazia, 1976a, b; Kidd er al.. 1976) and Begg and Rebhun (1979) have reported the importance of the pH of the isolation media in the preservation of microfilaments in the egg cortex. Inhibitors of microfilarnent assembly (see Section IV,A) prevent the extension of the fertilization cone and the bursts of microvillar elongation. It appears then that insemination triggers a burst in the appearance of cortical microfilaments. Microtubules do not appear to be present in unfertilized eggs either. Following sperm incorporation, the sperm axoneme is found within the fertilization cone, and later rnicrotubules are found in the region of the sperm aster (Longo and Anderson, 1968). In comparison to those in micrographs of the asters of the mitotic apparatus at first division, the rnicrotubules comprising the sperm aster are sparse indeed; this sparseness might well be the result of problems of fixation or perhaps an indication that far fewer microtubules are present in the sperm aster, as supported by immunofluorescence evidence (Bestor and Schatten , 1981). As the sperm aster increases in size, the lengths of microtubules similarly increase and Harris (1979) has described a spiral basket of microtubules in the subsurface region oriented along the egg radii. This microtubule basket might represent remnants from the fully extended sperm aster or might be a novel anastral structure. Microtubules are found at three distinct times during the first cell cycle; the first appearance is in the sperm aster, later the streak stage is found to be composed of microtubules, and finally the mitotic apparatus displays numerous microtubules. Transmission electron microscopy confirms the presence of microtubules in the sperm aster, at the subsurface region of the interphase egg, in the streak, and in the mitotic apparatus.

2 . Immunojluorescence Microscopy Indirect immunofluorescence microscopy with monospecific antibodies represents the single greatest technical advance for the study of the cytoskeleton during the last decade (Brinkley ef a/., 1975; Fuller et al., 1975; Lazarides and Weber, 1974; reviewed by Brinkley et a / . , 1980). The advantages of this technique are the ability to detect the three-dimensional organization of cytoskeletal elements with a detection ability rivaling that of conventional electron microscopy. The technique is not without its disadvantages and cautions are in order regarding the specificity of stain and postfixation extractions of interesting rnaterial. Eggs and oocytes represent a special problem for immunofluorescence because of their large size, the abundant pool of unpolymerized precursor rnolecules, and the difficulties in fixing the cytoskeleton in an unmodified manner. Nevertheless, immunofluorescence microscopy with tubulin antibody has been performed on mammalian oocytes during meiosis (Wassarman and Fujiwara, 1978) and on eggs (Harris et af., 1980a,b; Bestor and Schatten, 1981, 1982) during fertilization and early development. As inferred from transmission electron microscopy, unfertilized eggs do not

90

GERALD SCHA’ITEN

appear to contain any tubulin-staining structures. Interestingly, though, the unfertilized egg nucleus does in many batches stain specifically with tubulin antibody, in a punctate pattern. Immediately following insemination the speim tail can be found within the egg cytoplasm often in a “U”-shaped configuration resulting from the rotation of the sperm during incorporation (Fig. 16). Within 5 minutes of insemination, microtubules begin to form at the base of the sperm axoneme (Fig. 16). This initially formed sperm aster moves the male pronucleus toward the egg center. At this stage the sperm aster is radially symmetrical and it increases in size, moving the male pronucleus toward the egg center. The sperm tail is visible throughout these stages. The microtubules comprising the sperm aster appear to contact the surface of the female pronucleus as the migration of the female pronucleus begins. Relatively few sperm astral microtubules contact the female pronucleus during its migration. Following the migration of the female pronucleus, the remaining microtubules comprising the sperm aster continue to elongate as the adjacent pronuclei undergo centration. At this stage the previously monastral structure develops two focal points, perhaps an indication of centriole separation. The sperm aster increases in size until the pronuclei reach the egg center (Fig. 16). Pronuclear fusion typically follows the disassembly of the sperm aster. Microtubules are next found to assemble during the streak stage into a planar bipolar structure that distorts the zygote nucleus (Fig. 17). The streak is a transient microtubule-containing structure that is typically lost prior to the formation of the mitotic apparatus. lmmunofluorescence and fluorescence microscopy of cytoskeletal components holds promise for accurate descriptions of the swift configurational changes during fertilization. Wang and Taylor ( 1979), studying microinjected fluorescently labeled actin, have described the presence of actin in the egg cortex and its behavior during cytokinesis. Hamaguchi and lwasa ( 1980) have successfully employed similar methods with fluorescently labeled calmodulin and with rather interesting results.

3. Biochemical Isolations The surface of unfertilized eggs was first isolated by Vacquier (1975) using a novel technique of affixing eggs to a cationic substrate and shearing the tops off these cells in a zero-calcium environment. Modifications of this method permitted the isolation of the entire surface in suspension (Detering rt a l., 1977; Schatten and Mazia, 1976a,b; Vacquier and Moy, 1980) and since then nunierous transmisson electron microscopy, scanning electron microscopy, irnmunofluorescence microscopy, and biochemical isolations and characterizations of protein content have been performed. Transmission electron microscopy of egg surfaces isolated after insemination have demonstrated the presence of microf ilaments in egg microvilli (Burgess and Schroeder, 1977; Kidd et al., 1976; Begg et

MOTILITY DURING FERTILIZATION

91

FIG. 16. Growth of sperm aster. (A) The sperm aster is moved into the cytoplasm of the egg, accompanied by the elongation of astral microtubules. Sperm nucleus is visible as an area from which microtubules are excluded (arrow); all microtubules appear to be organized around the sperm midpiece. A . pitnctulata. (B) Astral microtubules continue to proliferate and begin to approach the egg nucleus, visible at upper left (arrow). Microtubules are apparent in region between egg cortex and sperm nucleus. A . puncrulara. (C and D) Sperm asters at 7 and 10 minutes postinsemination, respectively, showing pronounced radial arrangement of microtubules in eggs of A. punctulata. Many of the fibers visible in these micrographs are of substantially lower intensity than the sperm axoneme, suggesting that microtubule bundles containing only a few microtubules are visible by immunofluorescence microscopy. Bars, 10 Fm. (A-C. Reprinted with permission from Bestor and Schatten, 1981; D, from unpublished work of Balczon and Schatten.)

al., 1978) and have provided information regarding the ionic controls for cortical microfilament assembly (see Section V,A,2). Initial biochemical evidence for the presence of gelating factors in eggs came from the work of Kane (1975, 1976), who was able to demonstrate the role of

92

GERALD SCHAITEN

FIG. 17. Centration and first cleavage in A. punciulaia. (A) and (B) show the expansion of the sperm aster that occurs during the centripetal motion of the pronuclei after they have been brought into apposition (at about 20 minutes postinsemination). In these examples, fusion of the pronuclei has not yet occurred and the male pronucleus is visible as a dark region within the sperm aster. (C) Expansion of the sperm aster results in the construction of a spherical shell of microtubules that confines the zygote nucleus to the center of the cell. In the egg at right, microtubules have begun to disassemble at their inner ends (25 minutes postinsemination). (D)A well-developed streak-stage cell at 40 minutes. Arrows indicate microtubules near the cortex of the cell; these may represent remnants of the sperm aster. (E) Metaphase mitotic apparatus; (I:)Anaphase mitotic apparatus. Bars, 10 pm. (A-D, Reprinted with permission from Bestor and Schatten, 1981; E and F, from unpublished results of Balczon and Schatten.)

actin polymerization in cytoplasmic extracts. In that and later work (Bryan and Kane, 1978; Kane, 1980) proteins in addition to actin have been isolated within these microfilament-containing gels and a 58,000-dalton protein termed fascin has been described, which appears responsible for bundling microfilaments together and for producing the striped pattern of these types of microfilaments.

MOTILITY DURING FERTILIZATION

93

Burgess and Schroeder (1977) described this striped pattern with a 12 nm periodicity, which has been confirmed by DeRosier et al. (1977) in reconstituted bundles and by Spudich and Amos (1979) using optical defraction and image reconstitution techniques. Otto et al. ( 1980) have utilized immunofluorescence microscopy and electrophoretic techniques to study actin and fascin content in isolated egg surfaces and have demonstrated the increase of actin and fascin association with the egg surface following fertilization. Their results are that about 35% of both proteins are redistributed from the cytoplasmic pool to the isolatable egg cortex following fertilization. In addition to the presence of actin (Burgess and Schroeder, 1977; Spudich and Spudich, 1979; Spudich and Amos, 1979: Mabuchi and Spudich, 1980; Otto et a l . , 1980) and fascin (Otto et al., 1980), egg myosin has been isolated from cytoplasmic extracts (Mabuchi, 1973, 1974; Kane, 1980). In the important work of Kane (1980) the identical cytoplasmic extract could be induced either to form structural protein cores similar to those found in egg microvilli or to undergo contraction in v i m in an amoeboid-like fashion: these tho events are behaviors displayed in regions of the egg surface at fertilization. The fertilization cone behaves in a pseudopod-like fashion (Tyler, 1965), and the elongation of the remaining microvilli probably is the result of the formation of these structural protein cores. It appears that monomeric actin predominates in the unfertilized egg cortex. Shortly after insemination the extension of the microvilli o c c m concomitant with the appearance of striped microfilaments resulting from first actin polymerization and then microfilament-fascin associations, and myosin could possibly play a role in the actual incorporation of the sperm. In contrast to the many successful uses of the isolated egg cortex, the isolation of the sperm aster has only recently been attempted (Schatten and Schatten, 1979). Utilizing its features in common with the mitotic apparatus, the methods for isolating the mitotic apparatus developed by Kane (1962), Mazia and Dan (1952), and Forer and Zimmerman (1974) have been employed. Consistent with its transmission electron microscopy and antitublin immunofluorescence staining pattern the sperm aster appears to be composed of far fewer microtubules than the mitotic apparatus and its isolation has been fraught with problems during centrifugal purification. Improvements in the ability to isolate microtubule-containing structures (Salmon and Segall, 1980) are showing promising results. Substantial quantities of tubulin have been isolated from unfertilized sea urchin eggs, and estimates have been made that 5% of the total egg protein is tubulin (Raff et af., 1971). In v i m tubulin assembly has been investigated by Suprenant et al. ( 1981). Additionally dynein, the ATPase active in microtubule sliding during ciliary motility, has been found in unfertilized eggs (Pratt, 1980), isolated egg cortices (Kobayashi er af., 1978), and isolated mitotic apparatus (InouC, 1976; Mohri et al., 1976; Pratt et al.. 1980). Recently Naruse and Sakai ( 198I ) have reported the existence of an egg cortical polysaccharide which both inhibits microtubule assembly and promotes microtubule disassembly.

94

GERALD SCHATTEN

4. Biophysical Evidence Regarding Cytoskeletal Alterations Changes in the biophysical properties of sea urchin eggs at fertilization have been reviewed by Mitchison (1956), Hiramoto (1970), and Yoneda (1973). Mechanical properties of the egg surface at fertilization have been examined by Mitchison and Swann (1955). Using the behavior of microinjected iron particles in an electromagnetic field, Hiramoto (1974) has demonstrated a sharp increase in cortical stiffness within the first 5 minutes after insemination, which then declines and finally gradually increases in correlation with changes in microvillar lengths. This biophysical evidence of changes in surface stiffness is cornplemented by studies of the refractive indices of regions of the egg determined by interference microscopy (Hiramoto et a l . , 1979), which demonstrates that the egg cortex has a higher refractive index than the more interior cytoplasm. Periodic changes in surface contractions have been described following fertilization or artificial activation by Yoneda et al. (1978), Hara er a / . (1980), and Schatten (1979). These studies underscore the dynamic nature of the cytoskeletal reorganizations detected by electron microscopy and biochemical analysis.

5 . Cyroskeletal Changes during Fertilization On the basis of biophysical, biochemical, and microscopic evidence, there clearly are dramatic changes in the configuration of microtubules and microfilaments throughout fertilization and, in concert with evidence demonstrating the selective inhibition of specific motions by motility inhibitors (see Section IV), it is clear that the cytoskeletal rearrangements are responsible for the proper sequence of movements at fertilization. It is similarly clear that the participation of both microfilaments and microtubules is essential. The presence of and any possible roles for intermediate filaments (reviewed by Lazarides, 1980) have yet to be explored. The unfertilized egg does not appear to have any assembled cytoskeletal elements. However, these may be artificially induced either by activation (see Section V,A) or by the use of drugs which stabilize microtubules (see Section IV ,B,2). Moments after sperm-egg fusion, microfilament assembly appears to be nucleated at the egg plasma membrane with the polarity of the microfilaments pointing toward the egg center. At first, these microfilaments apear to be randomly oriented. Because of the binding of fascin to these individual microfilaments, bundles appear as the core structure of each microvillus. The possibility of cytoplasmic contractions within the fertilization cone remains attractive, Following the initial formation of these microfilament bundles, a second burst of microfilament assembly elongates these microfilaments further and later the formation of the contractile ring effects cytokinesis. The first microtubules found in fertilized eggs are those of the sperm axoneme. The incorporated sperm tail retains its 9 + 2 configuration and remains attached to

MOTILITY DURING FERTILIZATION

95

one of the centrioles often throughout the first cell cycle (Fig. 37). Within 5 minutes after insemination, microtubule assembly is noted in the formation of the initially radial sperm aster. The aster develops an asymmetrical appearance concomitant with the migration of the female pronucleus and the remaining microtubules extend to their maximal length as the pronuclei move to the egg center. These microtubules appear to be predominantly organized around the pair of sperm centrioles. which by the time of syngamy have separated to opposing poles. The question of anastral microtubules remains and may be the result of species differences or specimen preparation. Following the loss of the microtubules of the sperm aster, microtubules again appear in the streak and then form the mitotic apparatus.

IV. Effects of Motility Inhibitors Studies with selective inhibitors of specific types of motility have increased our knowledge of the specific role of each sort of motility at every phase of fertilization (reviewed by Schatten, 1982). Foi. example, Zimmerman and Zimmerman ( 1967) were able to predict the involvement of microtubules in the pronuclear migrations by demonstrating that colcemid would block pronuclear fusion. Since so many events occur virtually simultaneously at the early moments of fertilization, it is particularly important that the inhibitory studies be critically examined to eliminate the possibility that indirect effects are being confused with the primary conclusion. For example, it has been reported that cytochalasin will at times induce cortical reactions; a premature elevation of the fertilization coat would itself prevent sperm incorporation. Numerous studies have discounted this effect but it is a good example demonstrating the various possible alternatives when working with inhibitors. A. MICROFILAMENT INHIBITORS 1 . Assemblv Inhibitors

The use of inhibitors of microfilament function to study the mechanisms responsible for fertilization has been investigated by a number of workers. Sanger and Sanger (1975) demonstrated the cytochalasin B insensitivity of the actin polymerization during the acrosome reaction in the sperm. Gould-Somero et ul. (1977), Longo (1979a. 19801, Byrd and Perry (19801, and Schatten and Schatten (1979, 1980a, 1981) present evidence that eggs treated with cytochalasin B are unable to incorporate the spermatozoon even though the sperm-induced egg activation occurred. Banzhaf et ul. (1980) have demonstrated the cortical sensitivity of fertilized eggs to cytochalasins and have shown the speed of their permeability. A problem with the use of cytochalasins on populations of eggs is

96

GERALD SCHATTEN

that a percentage of these eggs, which varies from batch to batch, undergoes cortical granule discharge merely because of the addition of the drug (Longo, 1978; Schatten and Schatten, 1980a). With recent video methods (Schatten and Schatten, 1981) it is possible to ensure that the observed egg has not undergone a premature cortical granule discharge, which would interfere with the subsequent fertilization attempt. Cytochalasins B, D, or E added before or at insemination prevent sperm incorporation and the formation of the fertilization cone; the sperm-induced secretion of the cortical granules with the accompanying elevation of the fertilization coat and egg activation appear to occur normally. The bioelectrical responses during fertilization occur normally during CD insemination (Hiilser and Schatten, 1980, 1982; Dale and de Santis, 1981) and a cytoplasmic bridge briefly connects the gametes (Fig. 18); these events are likely indicators that sperm-egg membrane fusion occurs in the presence of the cytochalasins which interfere with a process during fertilization following fusion, i.e., sperm incorporation. When these drugs are added within 1 minute after the observed incorporation of the sperm, the forming fertilization cone is rapidly resorbed, the gliding of the sperm along the egg cortex is terminated, but the pronuclear migrations, syngamy, and mitosis occur on schedule; cytokinesis is prevented. These results indicate that these inhibitors of microfilament assembly (I'anenbaum, 1978; Flanagan and Lin, 1980; Brenner and Kom, 1980) will interfere with the fertilization process during the stage when the sperm is translocated from the exterior of the egg to the cytoplasm. Since the actin polymerization in the acrosome of the sperm is not sensitive to cytochalasin B (Sanger and Sanger, 1975), and since sperm incubated in cytochalasin D containing egg water will fertilize untreated eggs, the likely site of action of the cytochalasins is the egg cortex and the forming fertilization cone. The cytochalasins do not interfere with the pronuclear migrations, while microtubule inhibitors do (Aronson, 1973;Zimmerman and Zimmerman, 1967; Schatten and Schatten, 1981; revieRed in Schatten, 1982); the primary role of the assembly of egg microfilaments must be sperm incorporation, i.e., the movement and discharge of the sperm from the exterior into the egg cytoplasm, and the restructuring of the fertilized egg cortex. The surface of the cytochalasin-treated eggs develops an aberrant appearance within an hour if these drugs are added either prior to or within 1 minute of sperm-egg fusion (Fig. 18). However, when they are added later than 2 minutes post sperm-egg fusion, the egg surface retains its normal appearance for at least the first two cell cycles. Banzhaf et ul. (1980). using the formation of cytoplasmic protrusions, have found a similar cortical sensitivity. These findings have some importance in light of the work by Schroeder (1979) which reports the changes in microvillar length and surface area in Strongylocentrorus purpiirutus following fertilization. In that species, the egg microvilli and the surface area are greatest within 2 minutes following insemination, and the microvillar length and

MOTILITY DURING FERTILIZATION

97

FIG. 18. The effects of the microfilament inhibitors during fertilization. Cytochalasin E (10 phi) 5 minutes prior to insemination. Cytochalasin does not affect the normal appearance of the unfertilized egg (A). Sperm-egg binding (B), fusion (C). and the sperm-induced elevation of the fertilization coat (C-F) occur normally. The fertilization cone does not form around the successful sperm (C. D).and instead of the fertilization coat’s elevating over the attached sperm. its elevation detaches the sperm (black arrow) from the egg surface (C-L). L. varieRufu.5. Bar. 20 pm. (Reprinted with permission from Schatten and Schatten, 1981.)

surface area returned to that of unfertilized eggs by 20 minutes later. It may well be that the surface distortions observed when the cytochalasins are added within 2 minutes of fusion result from an impaired ability of the egg to revert back to its surface area and microvillar length. The resorption of this excess in surface area, resulting from the addition of the cortical granule membrane to the egg plasma membrane during secretion, may require a microfilament-mediated process. The critical time for the cytochalasin influence, at about 2 minutes postfusion, corre-

98

GERALD SCHATTEN

lates with Schroeder’s maxima (1979). However, the tropical L . variqatus develop almost three times faster than S. purpurarus, and it may be that by 2 minutes postfusion in this species the majority of the resorption has already occurred. It has long been recognized that changes in the egg surface are correlated with the transition from an unfertilized state to a fertilized one (Loeb, 19 13; Hiramoto, 1974; Mazia et al., 1975b; Eddy and Shapiro, 1976; Schatten and Mazia, 1976a,b; Burgess and Schroeder. 1977), and this result supports that idea. It appears that the egg cortex is fundamentally altered following sperm-egg fusion as judged by the new appearance of assembled microfilaments (Burgess and Schroeder, 1977), alterations in the stiffness of the surface (Hiramoto. 1974), and changes in the overall surface area and microvillar lengths (Schroeder, 1979). The effects of the cytochalasins support the notion that microfilaments play an integral role in the maintenance of the egg cortex, and this finding reiterates the drastic reorganization of the egg cortex within minutes of sperm-egg fusion. The finding that the fertilization cone and sperm incorporation are block.ed by the cytochalasins when they are added before insemination (Fig. 19). and that the fertilization cone is rapidly resorbed when the cytochalasins are added subsequent to sperm-egg fusion, argues strongly for an important role of the fertilization cone during sperm incorporation. The TEM observation of microfilament bundles (Longo, 1980) supports this supposition. During cytochalasin fertilization, the successful sperm is separated from the egg by the elevation of the fertilization coat. The inability of these inhibited eggs to form fertilization cones raises questions concerning the actual function of the fertilization cone and demonstrates an additional role of the fertilization coat. Since the sperm is separated from the egg surface when a fertilization cone is not formed, it appears that an important role of the fertilization cone may be to anchor the sperm at the egg membrane as the fertilization coat elevates around it. The physical motions associated with the elevation of the fertilization coat are apparently capable of separating sperm which are not anchored by a fertilization cone from the egg plasma membrane. This removal of unanchored, supernumerary, sperm from the egg membrane may well be an important role of the fertilization coat elevation. DTT-treated eggs, unable to elevate a fertilization coat, though competent to undergo the cortical reaction, will not permit sperm incorporation when exposed to the cytochalasins. Therefore, the primary effect of the cytochalasins is to interfere with the formation of the fertilization cone. It appears then that microfilament assembly at the site of sperm-egg fusion is necessary to form the fertilization cone, which is essential to the movement of the sperm from the surface into the cytoplasm and additionally anchors the successful sperm so that it is not separated from the egg during the elevation of the fertilization coat. Recent biochemical investigations concerning the mode of action of the

MOTILITY DURING FERTILIZATION

99

FIG.19. The surface events during cytochalasin fertilization. Sperm binding occurs at the plasma membrane (white triangles). though the egg-mediated elongation of microvilli and the formation of the fertilization cone arc prevented by cytochalasin B. L . vuriegatrts. Bar, I p,m. (Reprinted with permission from Schatten and Schatten. 198 I ).

cytochalasins (Brenner and Kom, 1980; Flanagan and Lin, 1980) indicate that cytochalasin binds to actin monomers (G-actin) which are then incompetent to permit further assembly; in other words, cytochalasin prevents the G + F-actin conversion. The relative efficiency of the various cytochalasins in v i m matches the efficiencies of these same drugs in vivo, i.e., CD 2 CE > CB. The molecular basis of sperm incorporation could be viewed as involving first the assembly of the periacrosomal actin to form the acrosomal fiber (Tilney et al., 1973). Fusion between the plasma membranes of the acrosome-reacted sperm and the egg might initiate the polymerization of egg microfilaments around the successful sperm. The forming fertilization cone anchors the sperm so that it is not separated from the egg surface during the elevation of the fertilization coat. When the microfilaments of the fertilization cone and the egg cortex reach a threshold degree of stability, sliding between the acrosomal filaments and the fertilization cone and egg cortical filaments might occur; the egg cortical myosin (Mabuchi, 1973; Kane, 1980) might well be the active ATPase. Alternatively, the microfilaments in the fertilization cone could assist in sperm incor-

100

GERALD SCHATTEN

poration by undergoing treadmilling of monomers, or by supporting the orientation of the sperm while a force drives it into the cytoplasm. The addition of the cytochalasins prior to sperm-egg fusion prevents the assembly of egg microfilaments which is necessary for the formation of the fertilization cone, the actual movements required for sperm incorporation, and the restructuring of the fertilized egg cortex. The rapid resorption of the fertilization cone after the addition of the cytochalasin may indicate that an equilibrium between microfilament assembly and disassembly is occurring during the stage when the fertilization cone is extending. This apparent equilibrium between microfilament assembly and disassembly in the fertilization cone makcs this structure transient. During the first minutes after sperm-egg fusion, assembly is favored and the fertilization cone increases in size; a short time later, disassembly is favored and the cone is resorbed. The finding that the addition of the cytochalasins after sperm-egg fusion does not interfere with the later nuclear movements of fertilization may indicate that only a few microfilaments are necessary to discharge the sperm into the cytoplasm; the gliding of the sperm along the egg cortex is prevented, which might well indicate that this lateral displacement requires a microfilamentous cortex, The pronuclear migrations and syngamy are completely independent of microfilament assembly. 2. Disassembly Inhibitor In contrast to the cytochalasins, which interfere with microfilament assembly, phalloidin prevents microfilament disassembly (Wieland, 1977). Though the permeability of this drug appears limited (Wehland et al., 1977b, 1978). long incubations in high concentrations have some effect. Following an hour in 1 mM phalloidin, a rather large and persistent fertilization cone elevates after insemination. The phalloidin-treated fertilization cones display a great deal of internal activity. Particles added to the base of the cone flow to the apex from the central regions of the cone and then return to the base along the periphery. The cortical reaction in phalloidin-treated eggs appears retarded and the elevation of the fertilization coat is restricted. Furthermore the lateral displacement of the sperm during incorporation is reduced from an average distance of 12.4 to 10.8 pin in 1 mM phalloidin, and the rate of movement is reduced from an average of 3.5 pm/ minute in controls to 2.2 pdminute in phalloidin. The entry of the sperm, the formation of the sperm aster, and the accompanying pronuclear migrations all occur at the normal rate on schedule. The mitotic apparatus forms and the chromosomes undergo normal separation. However, the contractile ring does not form in these phalloidin-treated eggs. Interestingly, the reconstituted nuclei in this binucleate egg are rather close together. Implications of microfilament inhibitor studies on the mechanisms of cell division predominately relate to the contractile ring (Schroeder, 1978b). Concentrations of the cytochalasins which will permit sperm incorporation will pre-

MOTILITY DURING FERTILIZATION

101

vent cytokinesis, perhaps indicating a greater requirement for synchronous polymerization. Phalloidin-treated eggs are unable to organize a functional contractile ring, which might somehow be involved in the separation of the reconstituting nuclei. It appears that pole separation during mitosis does not occur in phalloidin-treated eggs, and it may well be that phalloidin, which might possibly rigidify the egg surface, might prevent normal separation of the asters. The observation that an impaired cytokinesis prevents nuclear separation is consistent with the experiments using microinjected myosin antibody at cleavage (Mabuchi and Okuno, 1977), in which it was shown that the daughter nuclei were unusually close together in arrested eggs. In summary, the cytochalasins added before sperm-egg fusion prevent sperm incorporation; if added later, they permit all the necessary migrations and syngamy (Table 111). The assembly of egg microfilaments appears necessary for the formation of the fertilization cone, which most probably is active in translocating the sperm from the egg surface to the cytoplasm, and in the restructuring of the egg cortex required during the first cell cycle. The fertilization cone also has the role of anchoring the sperm to the egg membrane so that it is not separated from the egg during the elevation of the fertilization coat. The alternation of motility due to microfilaments with that due to microtubules (see next section) in each gamete is necessary for the successful union of the nuclei during fertilization.

B. MICROTUBULE INHIBITORS Fertilization is an almost unique system in which to study the various modes of microtubule-mediated motility. The beating of the sperm tail, which propels the sperm to the egg surface and which may not be required for incorporation (Epel et a l . , 1977), is an example of the sliding of adjacent microtubules (Gibbons, 1981). The centripetal migration of the male pronucleus and the later centration of the adjacent pronuclei are systems of movement in which microtubules push the pronuclei away from a surface, in this case the egg cortex. Finally, the migration of the female pronucleus to the center of the sperm aster appears to require the pulling of microtubules; this is emphasized by the distortion of the typically spherical female pronucleus into an oblate spheroid during its migration. 1 . Assembly Inhibitors

Zimmerman and Zimmerman ( 1967) first demonstrated that colcemid would prevent pronuclear fusion presumably by interfering with normal microtubule functions; the independence of the initiation of DNA synthesis and pronuclear fusion was noted by these and other workers (Bucher and Mazia, 1960; Longo and Plunkett, 1973; Schatten et af., 1982). Aronson (1973) corroborated the colcemid finding and demonstrated that, while microtubules were normally re-

102

GERALD SCHATTEN

EFFECTS

OF

TABLE 111 MICROTUBULE INHIBITORS DURING FERTILIZATION A.

Colcemid (500 nm) 500 nM -15 minutes At insemination

Colchicine (100 pM) - 15 minutes At insemination Griseofulvin (50 pM) -15 minutes At insemination +5 minutes +7 minutes + 10 minutes 14 minutes

+

Maytansine (50 nM) - 15 minutes At insemination + 5 minutes 10 nM at -15 minutes

Assembly inhibitors

Normal incorporation; aster formation blocked Normal incorporation; sperm aster formation attempted and then regression

Normal incorporation; sperm aster formation Normal incorporation; aster formation attempted and then regression

Normal incorporation; unusually pronounced sperm gliding along cortex; aster formation blocked Normal incorporation; unusually pronounced sperm gliding along cortex; aster formation blocked Rapid resorption of sperm aster; migration of female pronucleus blocked Resorption of sperm aster; arrest of migrating female pronucleus Centration of pronuclei arrested Fusion of centered pronuclei arrested

Normal incorporation; pronounced sperm gliding sperm aster arrested Normal incorporation; sperm aster formation attempted and then regression Lag of 3 minutes prior to regression of sperm aster: pronuclear centration blocked Incomplete centration; diminutive sperm aster

Nocodazole (50 nM) -15 minutes At insemination

+5 minutes

Podophyllotoxin (50 nM) - 15 minutes At insemination

+ 10 minutes Vinblastine (I0 pM) - 15 minutes

Normal incorporation; sperm aster formation blocked Normal incorporation; sperm aster formation attempted and then regression Lag of 3 minutes prior to regression of sperm aster; pronuclear centration blocked

Normal incorporation; unusually pronounced sperm gliding along cortex; aster formation blocked Normal incorporation; sperm aster formation attempted and then regression Normal female pronuclear migration and centration; mitosis blocked

Normal incorporation; unusually pronounced sperm gliding along cortex; aster formation blocked

MOTILITY DURING FERTILIZATION

103

TABLE III-Continued

At insemination

+S minutes

Photochcrnical reversal of assembly inhibitors Colchicinc (100 p M ) - 5 minutes + 3 minutes + 15 minutcs

A. Assembly inhibitors (continued) Normal incorporation; unusually pronounced sperm gliding along cortex; aster formation blocked Rapid resorption of \perm aster; migration of female pronucleus blocked

Normal incorporation; pronuclear movements and development Normal incorporation: pronuclcar movements and development Pronuclear movements and syngamy arrested: two cytasters at mitosis associated with male pronucleus

B . Disassembly inhibitor Taxol (10 p M ) -60 niinutcs 30 llllllUtCS ~

- IS

minutes

At insemination

+IS minutes +45 minutes

De nova aster formation Nornial incorporation and cortical reaction; formation of huge persisting sperm aster; migration of female pronucleus inhibited Formation of huge persisting sperm aster after migration of female pronucleus Formation of huge persisting sperm aster after migration of female pronucleus Formation of enlarged monoaster; cleavage unsuccessful Formation of enlarged mitotic apparatus; cleavage unsuccessful

quired for bringing the pronuclei into close proximity, syngamy itself could be effected by centrifugal force. Schatten ( 1977) and Schatten et al. (1982), using another microtubule inhibitor, griseofulvin, could demonstrate that syngamy was prevented with this drug though DNA synthetic rates were rather unaffected (Fig. 24). The presence of microtubules in the sperm aster has been demonstrated by transmission electron microscopy (Longo and Anderson, 1968) and indirect immunofluorescence with tubulin antibody (Harris et a l . , 1980a,b; Bestor and Schatten, 198I ) . Several inhibitors of microtubule assembly, i.e., colcemid. colchicine, griseofulvin, maytansine, nocodazole, podophyllotoxin, and vinblastine, were tested for their effects on each phase during fertilization (Schatten and Schatten, 1981). This was performed by monitoring individual eggs with time-lapse video, differential interference microscopy, and adding the drug at a defined moment during fertilization. The readily permeant drugs griseofulvin and vinblastine, which displayed antimicrotubule effects within a minute of addition, had the advantage that each phase of the pronuclear migrations could be investigated for its dependence on assembling microtubules. Colcemid and colchicine, which can be photochemically inactivated with light at 366 nm (Aronson and InouC, 1970; Sluder, 1979), were employed to demonstrate that sperm

104

GERALD SCHATTEN

incorporation is completely independent of assembling microtubules (Sc hatten and Schatten, 1981). Inhibitors of microtubule assembly prevent the formation of the sperm aster, which is required for the centripetal migration of the male pronucleus, the movement of the female pronucleus to the male pronucleus, and the centration of the adjacent pronuclei (Fig. 20; Table 111). The sperm aster is unable to form in the presence of these inhibitors (Fig. 21). Sperm incorporation in the presence of these inhibitors occurs normally with the exception that the lateral displacement of the sperm along the egg cortex is often far more pronounced than usual (Table IV); the formation of the sperm aster may well be involved in the conclusion of sperm incorporation. Pronuclear formation, but not fusion, occurs on schedule as does the breakdown of the two haploid nuclei at prophase. Though no mitotic apparatus forms, cycles of cleavages attempts, interspersed with cycles of nuclear reconstitution and breakdown, are noted. In conjunction with the preceding section, it appears clear that sperm incorporation is mediated by the assembly of egg microfilaments whereas the pronuclear migrations are mediated by the assembly of the microtubules which comprise the sperm aster. Knowledge concerning the specific biochemical site of interference with microtubule assembly of the inhibitors (reviewed by Bryan, 1974; Wilson and Bryan, 1974; Snyder and McIntosh, 1976; Soifer, 1976; Dustin, 1978) is excellent for colcemid, colchicine, podophyJl&xin, and vinblastine, and quite good for griseofulvin (Roobol et al., 1976; Sloboda et al., 1976; Wehland et al., 1977a), maytansine (Ludueiia et al., 1979; Rebhun et al., 1979), and nocodazole (DeBrabander et al. 1976). Since the evidence is that these drugs prevent the assembly of microtubules, the stable fibers comprising the sperm axoneme remain unaffected. Though the sperm enters the egg normally in these treated eggs, the incorpo-

FIG. 20. The effects of microtubule inhibitors during fertilization. Podophyllotoxin (50 I#); 15 minutes prior to insemination. Though sperm incorporation (A-D) occurs, the pronuclear migrations are arrested. The male pronucleus remains at the location to which it was incorporated (E) until nuclear breakdown. Time (hours:minutes:seconds),lower right. L. variegatus. Bar, 20 pin. (Reprinted with permission from Schatten and Schatten, 1981 .)

105

MOTILITY DURING FERTILIZATION

FIG. 21. Inhibition of pronuclear movements and microtubule assembly by colchicine. (A) Sperm axoneme in eggs that had been pretreated for 15 minutes with M colchicine prior to insemination, fixed and stained for tubulin at 20 minutes postinsemination. Sperm centrioles (arrow) remain close to cortex. (B)Living egg treated as above. Arrow indicates male pronucleus. None of the pronuclear motions (other than sperm incorporation) occur in such eggs. L . rwiegarus. Bars, 10 pm. (Reprinted with permission from Bestor and Schatten. 1981.)

TABLE IV DISPLACEMENT OF SPERMHEADD ~ J R I N GINCORPORATION ~~

Control Microfilament inhibitors Phalloidin ( I mM) Cytochalasins (25 pM CB; 10 pM CD; 10 pM CE)

Microtubule inhibitors Colcemid (500 nM) Griseofulvin (50 pM) Maytansine (SO nM) Nocodazole (50 nM) Podophyllotoxin (SO nM) Vinblastine (10 pM) Average of microtubule inhibitors

Average distance displaced from site of sperm-egg fusion (pm)

Average time for displacement (minutes)

Average rate ( pm/ minute)

12.4 t 4.9

3.7 t 0.7

3.5 t 1.3

6.3 t 2.6 NA

2.2 t 1.3 NA

3.7 t 0.7 7.0 2 3.0 4.7 2 0.5 4.7 t 0.5 5.3 t 1.9 3.8 5 1 . 1

4.4 5.2 6.1 5.9 4.3 6.6

4.9 r 1.2

5.4 t 0.9

10.8 2 3.8 0; detached by fertilization coat elevation

15.8 36.5 30.4 29.4 23.5 26.7

C 6.8

t 15.1 t 11.8

8.7 9.8 t 13.4 2 2

27.1 t 7.0

5

2. I

2 0.1

t 2.4 t 1.7 5

0.9

t 3.2

106

GERALD SCHATTEN

rated sperm undergoes a far greater, longer, and faster lateral displacement along the egg cortex than is usually observed (see Table IV). This finding has implications both for the mechanism of sperm incorporation and for the method whereby incorporation is terminated and the sperm is discharged into the egg cytoplasm. Since the primary site of interference in these experiments is the assembly of microtubules, it appears reasonable that the new assembly of microtubules both terminates the gliding of the sperm along the egg cortex and simultaneously moves the male pronucleus centripetally. It may be that the incorporated spermatozoon will glide along the microfilaments of the egg cortex until the time when microtubules grow on the sperm centrioles to form the sperm aster. These microtubules then might interfere with the lateral displacement either by separating the interacting molecules of the sperm, say the acrosomal filaments, from those of the egg, e.g., the egg cortex, or alternatively the microtubules comprising the sperm aster might produce sufficient drag that, while attempts for translocations are being made, little movement can occur. The movement of the male pronucleus centripetally is precisely coupled with the growth of the microtubules which comprise the sperm aster. In the absence of microtubule assembly, the sperm nucleus undergoes chromatin decondensation at the region in which it was discharged into the cytoplasm following incorporation. Cycles of nuclear breakdown and reconstitution of the two haploid nuclei occur on schedule, and the male pronucleus often undergoes nuclear envelope breakdown in advance of the female pronucleus. An experiment underscoring the lack of any requirement for microtubule assembly during sperm incorporation is presented in Fig. 22. Here the egg was inseminated after treatment with 1 p M colcemid and the sperm was successfully incorporated. At 1 minute postinsemination the colcemid was photochemically inactivated with light at 366 nm and the movements for fertilization and later development occurred. This indicates that whereas microtubule assembly is necessary for the pronuclear migrations it is not required for sperm incorporation. The rapid regression of the sperm astral fibers following the introduction of either griseofulvin or vinblastine indicates that a turnover of tubulin monomers is occurring (Fig. 23). Were this not the case, then the introduction of these assembly inhibitors would be expected to prevent any further increase in size, but they would not be expected to result in the dissolution of the structure. To explain the finding that the sperm aster actually increases in size during this stage in normal fertilization, an equilibrium between microtubule assembly and disassembly must exist, with the net equation favoring assembly. The addition of these assembly inhibitors, which would not be expected to influence the rate of disassembly, results in the observed disappearance of the sperm aster. The migration of the female pronucleus to the center of the sperm aster, which is an example of microtubule pulling, is also inhibited by the permeable assembly inhibitors griseofulvin and vinblastine. This argues that an assembling sperm

MOTILITY DURING FERTILIZATION

107

FIG. 22. Photochemical inactivation of colcemid at 2.5 minutes post-sperm-egg fusion. Sperm incorporation occurs normally in the presence of I p,M CMD (B). In (C) the egg is irradiated with light at 366 nm for I minute. Following this photochemical inactivation of the CMD. the sperm aster forms (D) and the migration of the female pronucleus is observed (E-G). The later events during development. mitosis, and cytokinesis (I-K) occur normally except that the two planes for cleavage at second division are no longer parallel (L). Time (hours:minutes:seconds), upper left. L . vuriegarus. Bar, 20 Km. (Reprinted with permission from Schatten and Schatten, 1981.)

aster is necessary for the migration of the female pronucleus. These results indicate that both microtubule assembly and disassembly may be occurring during the time of the migration of the female pronucleus, and this motion might well constitute movement due to a dynamic equilibrium of microtubule assembly and disassembly (Inout and Sato, 1967; Margolis et al., 1978). Cytoplasmic particles undergo saltatory motions from the periphery of the sperm aster to the aster center and the direction of the saltatory movement is identical to that of the migration of the female pronucleus and the rates are somewhat similar. In addition to the dynamic equlibrium model, alternative theories for the movement of the female pronucleus, e.g., the sliding of the female pronucleus along the microtubules of the sperm aster, are at present equally attractive. The centration of the adjacent pronuclei, like the centripetal migration of the male pronucleus, requires the continuing assembly of sperm astral microtubules. At this stage microtubule assembly pushes the pronuclei from the egg surface. This pushing of the pronuclei from the egg cortex could possibly aid in syngamy by pushing the two pronuclei into close proximity. The necessity of close proximity, regardless of the source of the force bringing the pronuclei into contact,

108

FIG.

GERALD SCHATTEN

23. The effects of microtubule inhibition during the pronuclear migrations. Griseofulvin

(50 pM) 5 minutes post-sperm-egg fusion. The normal events during fertilization, i.e., incorporation

of the sperm (A), formation of the fertilization cone (B),growth of the sperm aster (C), occur prior to the addition of griseofulvin in (D). The rays of the sperm aster begin to shorten and the migration of the female pronucleus is arrested (E-G). The pronuclei remain at their respective locations until nuclear breakdown. Time (hours:minutes:seconds), lower right. L. variegatus. Bar, 20 pni. (Reprinted with permission from Schatten and Schatten. 1981 .)

for syngamy was elegantly demonstrated by Aronson (1973), who, using colcemid-treated eggs, could effect pronuclear fusion with external forces during centrifugation. The arrest of syngamy observed following the addition of low concentrations of griseofulvin or vinblastine could result from the incomplete growth of the sperm aster; the diminuitive aster might not have moved the two pronuclei sufficiently close together for the fusion of the nuclear membranes (Schatten er al., 1982). Though the normal configurations of microtubule assembly is arrested in all these treated eggs, it is interesting that they are able to undergo a number of cellular cycles on the normal schedule. These cycles would have to include the gliding of the sperm along the egg cortex and the beating of the sperm tail during the first 4 minutes following sperm-egg fusion, the onset and termination of DNA synthesis cycles (Fig. 24; Bucher and Mazia, 1960; Zimmerman and Zimmerman, 1967; Longo and Plunkett, 1973; Schatten, 1977; Schatten et al., 1982), the chromosome cycles of condensation and decondensation (hdazia,

MOTILITY DURING FERTILIZATION

109

1974), the cycles of nuclear envelope breakdown and reconstitution by fusion of the karyomeres, and cycles of cytokinesis attempts and relaxation (Schatten and Schatten, 1981 ) . Whether the trigger for cyclical cellular processes is altering levels in free ions, e.g., Ca2+ (Petzelt, 1972; Harris, 1978), or another regulatory system (rev. in Mazia, 1961; Mitchison, 1971), it can be stated that the cycles of microtubule assembly and disassembly observed in each cell cycle must depend on the master regulatory factors, and do not exert any control over the master triggers. 2 . Disassembly Inhibitors Taxol, a drug that has been shown to inhibit microtubule depolymerization, will induce aster formation in unfertilized eggs in the absence of any centrioles (Schatten et al., 1981b, 1982); taxol itself does not induce artificial activation. At fertilization, taxol does not affect sperm tail motility, sperm incorporation, or the cortical reaction. However, the cytoplasmic migrations of the sperm and egg nuclei are inhibited and syngamy does not occur. If it is added more than 15 minutes prior to insemination, the sperm aster will form, increase in size, and move the male pronucleus centripetally. Importantly, the migration of the egg nucleus to the center of the sperm aster does not occur, and furthermore the sperm aster remains as a large and permanent cytoplasmic structure. Taxol then prevents the normal movement of the pronuclei at fertilization, will stabilize the sperm aster, and will induce asters de novo in unfertilized eggs. Taxol has been shown to endow microtubules with an unusual stability in vitro (Schiff et al., 1980), in vivo in developing (Heidemann and Callas, 1980), differentiated (Mazurovsky et al. , 1981), and cultured (DeBrabander et al. , 1981; Schiff and Horwitz, 1980; Simone et al.. 1981) cells. The effects of taxol appear related to its direct binding of MAPS in promoting microtubule assembly (Herman and Albertini, 1981; Manfredi e t a / . . 1981; Vallee, 1981). The finding that taxol reduces the rate of microtubule “treadmilling” (Thompson et al. , 198I ) might well have important implications in understanding its effects on cellular movements proposed to be effected by microtubule treadmilling (Kirschner, 1980; Margolis et a / ., 1978) such as chromosome movement during mitosis (Inoue and Sato, 1967; Margolis and Wilson, 1981), or, in the case of fertilization, the migration of the female pronucleus. In unfertilized sea urchin eggs microtubules are rarely, if ever, observed. However, when these unfertilized eggs are exposed to 10 p,M taxol, numerous microtubule-containing asters assemble (Figs. 25 and 26). In Fig. 25A, it will be noted that within 15 minutes of exposure to taxol multiple punctate sites stain for tubulin. These sites enlarge at a half hour later (Fig. 25B) to form discrete astral structures throughout the unfertilized egg cytoplasm and persist for at least 2.5 hours (Fig. 25C).

110

GERALD SCHATTEN

h

7 0 . I ( W

E 0.

ie.

5.

e. Fic. 24. The onset of the first cell cycle in the absence of pronuclear fusion. Incorporation of [3H]thymidine into DNA of S. purpurarus. In control cells (A), the amount of DNA is roughly doubled each cycle. In cells incubated with I X 10-4M griseofulvin right after insemination (B), the rate of DNA synthesis in the second and following cycles is delayed and somewhat less compared to the control cells. In cells washed free of the drug after 150 minutes (C), DNA synthesis is resumed to the control level faster than the continuously treated cells. (Reprinted with permission from Schatten et a/., 1982.)

MOTILITY DURING FERTILIZATION

111

FIG. 25. The effects of taxol on unfertilized eggs. Antitubulin immunofluorescence microscopy of taxol-treated unfertilized eggs. Within I5 minutes of taxol addition, numerous punctate tubulincontaining structures appear throughout the egg cytoplasm (A). These structured increase in size by 30 minutes and appear as discrete asters (B), and persist for at least 2.5 hours ( C ) .Arbacia punctulata. Bars, 10 p n . (From Schatten e t a / . . 1982a. reprinted with permission.)

Transmission electron microscopy (Fig. 26) demonstrates that the tubulinstaining structures observed by immunofluorescence microscopy are microtubules. In unfertilized control eggs microtubules have not been observed by TEM nor detected by antitubulin immunofluorescencemicroscopy. Microtubules observed following taxol treatment are most prominent within discrete vesiclerich clear zones in the egg cytoplasm that are devoid of yolk platelets (Fig. 26A), which can be correlated with the antitubulin staining patterns. Though the microtubules tend to emanate radially from these zones, microtubules at oblique angles to the astral radii are frequently observed. Microtubule bundles emanate from a variety of centers, which can be the nuclear envelope (Fig. 26B), running paral-

112

GERALD SCHA’ITEN

FIG. 26. Transmission electron microscopy of taxol-treated unfertilized eggs. A survey of the taxol-treated unfertilized egg (A) confirms the presence of intact cortical granules and microtubulecontaining zones rich in membrane vesicles and devoid of yolk. Microtubule bundles emanate in unfertilized eggs from a variety of centers including the egg nuclear envelope (B), along annulate jamellae (C), electron-transparent regions (D), and osmiophilic focal points (E-F). At times a structure appearing partially like a centriole, with double and triple microtubule sets (E; arrows), is observed. A. punctulatu. Bars, (A) 10 pm; (B-D, F) 1 pm; (E)0.1 pm. (From Schatteri ef a / . , 1982a. reprinted with permission.)

lel with annulate lamellae (Fig. 26C) or electron transparent regions (Fig. 26D). Centrioles have never been observed in unfertilized taxol-treated eggs. However, the astral centers often contain osmiophilic granules from which the microtubules appear to extend (Fig. 26E and F): At times, a “centriole-like” structure is observed with double and triple microtubule sets (Fig. 26E, arrows). The effects of taxol in unfertilized eggs may be indicative of an equilibrium that, in the normal case, favors microtubule disassembly. Addition of taxol,

MOTILITY DURING FERTILIZATION

113

which seems to block the normal disassembly without affecting the rate of assembly, makes the slow but existing microtubule assembly detectable. Thus it becomes clear that the control of microtubule assembly in unfertilized eggs and at fertilization is not at the level of initiating assembly following sperm incorporation but rather at that of shifting the equilibrium from one that favors disassembly in the unfertilized egg to one which favors assembly as the sperm aster is formed. This method of regulation would permit a fine control of the rate and extent of microtubule elongation and resorption. The pattern of microtubule assembly in these taxol-treated unfertilized eggs is of interest. In this case dozens of foci appear and microtubules enlarge off of these sites. This configuration is quite different from that observed during fertilization when the sperm centrioles are the only assembly sites, during artificial activation when a single radial monaster forms around the centering egg nucleus (Moore, 1937; Paweletz and Mazia, 1979; Mar, 1980; Bestor and Schatten, 1981) or following parthenogenetic activation with heavy water when only a few (<25) larger asters form around the egg nucleus, and at the subcortical regions (Fig. 27; Bestor, Balczon, and Schatten, unpublished results). An important question yet to be resolved is whether the unfertilized egg possesses the ability to nucleate several different microtubular arrays that are utilized for varying purposes during development or if the taxol treatment resulting in the formation de novo of asters is the direct result of taxol itself.

FIG.27. The effects of heavy water (50% DzO) on unfertilized eggs. Antitubulin immunofluorescence microscopy. Several large asters form in these unfertilized eggs. Note that these asters tend to reside either at the cell surface or by the female pronucleus and display a different morphology from either the taxol-induced asters (Fig. 25) or the monasters in artificially activated eggs (Fig. 3 2 ) . Bars, 10 Fm. (From Bestor, 1981.)

114

GERALD SCHA’ITEN

The absence of any observable effects of taxol on sperm tail motility underscores the effects of taxol on assembling, not assembled, microtubules. The stable axonemal microtubules are apparently able to continue their normal dynein-mediated sliding in the presence of this drug, which presumably interferes with disassembly or “treadmilling.” In the presence of taxol, the initial formation of the sperm aster appears largely unaffected; however, all of the movements mediated by the sperm aster are altered. The formation, enlargement, and apparent motility of the taxol-treated sperm aster during fertilization is depicted in Fig. 28, of an egg treated for 30 minutes with 10 pA4 taxol and then inseminated. The sperm aster again forms on schedule (Fig. 28B, arrow) but enlarges to a much greater than normal extent. The female pronucleus is contacted (Fig. 28C) and, for the next half hour, the sperm aster distorts the female pronucleus as if slowly pulling it to the astral center (Fig. 28C-E); this distortion of the female pronucleus from a sphere to an oblate ovoid is characteristic of the egg nucleus during its migration, which typically occurs within 1 minute. The sperm aster continues to enlarge (Fig. 28F and G) as nuclear breakdown and reformation occur (Fig. 28H and I). Transmission electron microscopy of fertilized taxol-treated eggs at 15 minutes postinsemination demonstrates the abundance of microtubules and membrane vesicles within the sperm aster. In Fig. 29A, the relationship between the sperm astral microtubules and the female pronucleus is depicted. Microtubule bundles run parallel with, and ramify through, the egg nucleus. The microtubules of the sperm aster are not always aligned along a sperm astral radius (Fig. 29B, C, and D) and at times appear to emanate from focal points independent of the sperm centrioles: an osmiophilic source in Fig. 29E and annulate lamellae in Fig. 29F. The rates of the centripetal movement of the male pronucleus, the migration of the female pronucleus, and the centration of the pronuclei are all greatly reduced. These results indicate that the proper functioning of microtubules is affected by taxol, and, if our biochemical understanding of taxol is complete, then the conclusion is that microtubule disassembly is central to the motility functions of microtubules. An analogy between the movements at fertilization and those at mitosis can easily be drawn if the sperm aster is viewed as a half spindle. Instead of a set of chromosomes moving to one of the astral poles, here the female pronucleus is drawn to the center of the monopolar sperm aster. In this context, it is particularly interesting that the rate of the migration of the female pronucleus is so drastically reduced in taxol-treated eggs. The proposition that microtubult: disassembly mediates the migration of the egg nucleus to the center of the sperm aster appears warranted since taxol blocks the migration of the female pronucleus, leaving the female pronucleus at the periphery of the sperm aster and since immunofluorescence staining in control cells demonstrates the loss of the

MOTILITY DURING FERTILIZATION

115

FIG.28. The effects of preventing microtubule disassembly during fertilization. The behavior of the taxol-treated sperm aster during fertilization. The sperm aster forms on schedule after insemination of an egg previously treated for 30 mintues with 10 )wl.I taxol (B; arrow). By 9 minutes postinsemination, astral fibers have contacted, and appear to distort, the female pronucleus ( C ) .For the next half-hour (D-E) the female pronucleus is slowly pulled to the center of the sperm aster. The sperm aster persists during the normal time for mitosis (F, G) and it develops a thicker peripheral ring (H). The cycle of nuclear breakdown (F) and reconstitution (I) during division appear normal though syngamy and karyomere fusion are blocked. L . varirgarus. Time (hours:minutes:seconds), lower right. Bar, 10 km. (From Schatten e t a / . . 1982a. reprinted with permission.)

microtubules contacting the egg nucleus. Should microtubule disassembly be the driving force in the migration of the female pronucleus, it is not inconceivable that a similar mechanism could be responsible for the anaphase movements of the chromosomes during mitosis.

116

GERALD SCHATTEN

FIG. 29. Transmission electron microscopy of fertilized taxol-treated eggs at 15 minutes postinsemination demonstrates the formation of numerous microtubule bundles in association with, and ramifying through, the female pronuclear envelope (A), oriented primarily, though not exclusively, radial to the sperm aster (B, C, D)and, at times, emanating within the sperm aster from MTOCs other than the sperm centrioles (D-F). The abundance of membranes, and the lack of yolk platelets, within the taxol-stabilized sperm aster is noteworthy (D). A. puncruluru. Bars, I pm. (From Schatten et a/., 1982a. reprinted with permission.)

It appears then that taxol will induce the appearance of microtubules in unfertilized eggs, which is consistent with a model in which a normally undetectable amount of microtubule assembly occurs in the unfertilized egg but it is outweighed by microtubule disassembly. Taxol then, by blocking disassembly in the unfertilized eggs, permits the detection of this alleged slow rate microtubule assembly. It does not appear to affect the functioning of the stable microtubules of the sperm axoneme. During fertilization taxol does not appear to influence the rate of sperm incorporation or the formation of the fertilization cone, supporting

MOTILITY DURING FERTILIZATION

117

the ideas regarding the absence of egg microtubular involvement during sperm incorporation. The formation of the sperm aster, the centripetal migration of the male pronucleus, the migration of the female pronucleus, and pronuclear centration are all affected by taxol and syngamy is precluded in the presence of this drug; these findings confirm the involvement of microtubules in the pronuclear migrations and lead to speculations concerning the role of microtubule disassembly as a motive force during the migrations of the pronuclei. C. SUMMARY As a model for the study of cellular motility, fertilization is unexcelled. The beating of the sperm tail, which propels the sperm to the egg surface, is an example of the sliding of adjacent microtubules (Gibbons, 1977). The extrusion of the sperm acrosomal process requires the assembly of microfilaments (Tilney et al., 1973), as does the formation of the fertilization cone required during sperm incorporation (Gould-Somero et al., 1977; Longo, 1978, 1980; Byrd and Perry, 1980; Schatten and Schatten, i980a. 1981). The centripetal migration of the male pronucleus and the later centration of the adjacent pronuclei (reviewed by Schatten, 1982) are systems of movement in which microtubules appear to push the pronuclei away from a surface, in this case the egg cortex. Finally, the migration of the female pronucleus to the center of the sperm aster appears to require the pulling by microtubules; this is emphasized by the distortim of the typically spherical female pronucleus into an oblate spheroid during its migration. Current theories regarding the treadmilling of monomers through polymers (InouC and Sato, 1967; Wegner, 1976; Margolis er ul., 1978; Kirschner, 1980; Margolis and Wilson, 1981) may be evoked to describe the behavior of both the microfilamentous fertilization cone and the microtubular sperm aster. The evidence in support of these notions is derived from the phalloidin-fertilization cones, which take on the described fountain-like appearance, from the observations of saltatory flow (Rebhun, 1972) in the sperm asters of cytochalasin Dtreated eggs, from the rapid regression of the sperm astral fibers following the introduction of griseofulvin or vinblastine and the manner in which taxol-stabilized asters develop from the periphery and migrate to the center. If tubulin is polymerized at the periphery of the aster and treadmills to its center, attached particles could be translocated from the extremities of the sperm aster to the center. Interestingly, this manner of motility could well be involved in the migration of the female pronucleus to the center of the sperm aster. In light of recent studies concerning the effects of motility inhibitors on the fertilization process (reviewed in Schatten, 1982), a synthesis of the systems active in achieving syngamy can be compiled (summarized in Table V). The sliding of the microtubules of the sperm tail propels the sperm to the egg surface.

118

GERALD SCHATTEN TABLE V SUMMARY OF EFFECTSOF MOTILITY INHIBITORS~ Microt ubule inhibitors Microfilament inhibitors (Assembly) Cytochalasins

Sperm-egg attachment and fusion Cortical reaction, fertilization, coat elevation Fertilization cone formation Lateral displacement of sperm during incorporation Restructuring of fertilized egg cortex Formation of sperm aster Migration of female pronucleus Pronuclear centration SYngamY Formation of streak Mitosis Cytokinesis

<,- - , Event blocked; -,

+

+

(Assembly) Colcemid, griseofulvin. nocodazole. (Disassembly) maytansine. Phalloidin vinblastine. etc.

(Disassembly) Iaxol, DIO

+

+

+

-

__

++

+ +

+ +

__

_

++

+

__

+ + + + + + +

+

+ ++

+ + + + + +

__

event retarded;

__

__ __ __

+

__

__ __

__ __

_-

+

+

__

+, normal event; + +. event enhanced.

Near the egg surface (Aketa and Ohta, 1977), actin assembly in the sperm results in the extrusion of the acrosomal process (Tilney et af., 1973). Following attachment of the acrosome-reacted sperm to the egg surface, by the sperm protein bindin (Vacquier and Moy, 1977), plasma membrane fusion between the gametes occurs (Gage1 et al., 1979a,b). Microfilaments in the egg are assembled to form the fertilization cone which extend up and around the erect and stationary sperm (Longo and Anderson, 1968; Schatten and Mazia, 1976a,b; Schatteri and Schatten, 1980a). A short time later the sperm rotates and glides along the microfilaments of the egg cortex (Schatten, 1981b); this lateral displacement is most probably terminated by the assembly of microtubules on the sperm centrioles. The assembly of these microtubules forms the sperm aster (Bestor and Schatten, 1981; Harris er ul., 1980a,b), which first pushes the male pronucleus from the egg cortex into the cytoplasm. When these microtubules contact the female pronucleus, the migration of the female pronucleus occurs (Schatten, 1981a,b; Chambers, 1939). The discovery that the dynein inhibitor erythrohydroxy-nonyl adenine (EHNA) specifically inhibits the female pronuclear migration (Schatten et af., 1982~)implicates dynein ATPase activity in this motion.

MOTILITY DURING FERTILIZATION

119

The adjacent pronuclei are moved to the egg center by the continued growth of the sperm aster. The growth of the sperm aster during the last stages of centration may push the adjacent pronuclei to the point of membrane coalescence; the fusion of the nuclear membranes of the pronuclei constitutes the completion of fertilization.

V. The Regulation of Motility at Fertilization Eggs are ideal models for studying the regulation of motility. First, both microtubule- and microfilament-mediated motions occur and unlike those in many other systems, these motions are well separated both spatially and temporally. Second, in working particularly with sea urchin eggs, the quantity of cells and the near perfect synchrony permit mass isolation, fixation, or ionic studies of populations of cells; the synchrony in sea urchin fertilization is unrivaled. Also, in the sea urchin system the unfertilized egg is spawned without any discernible cytoskeleton and within 10 minutes of insemination a complex array of both microtubules and microfilaments is observed. This phenomenon permits the dissection of the events that comprise the formation of the motile apparatus independent of the studies attempting to elucidate the manner in which force is generated. Finally, the proper dissolution of these cytoskeletal elements is required for the normal progression through the first cell cycle, permitting studies regarding disassembly as well as assembly during motion. Finally, the literature regarding the ionic regulation of egg activation is very impressive indeed. Indirect as well as direct studies concerning fluxes in protons, calcium ions, sodium ions, and bioelectric potentials have all been fit into an orderly sequence, permitting an almost complete and certainly verifiable program of activation, which has been tested and accepted by most workers. This understanding of the program of activation now permits specific questions regarding the role of each ionic event at fertilization and its contribution in regulating the complex repertoire of movements to effect syngamy. A. THEONSET OF FERTILIZATION

Prior to considering in detail the regulation of motility during fertilization, it is germane to consider the moment at which the process of fertilization begins. Definition of that moment, of course, is fraught with problems-problems akin to the current legislative quest to define the moment at which life starts. Nevertheless, this section will explore the moment at which the sperm triggers the earliest detectable response in the egg. Several criteria could possibly be used to answer the question “When does the sperm awaken the egg?” For example, the engulfment of the sperm by egg microvilli would be a morphological criterion or the increased rate of protein synthesis following insemination would be a biochemical marker. However, the

120

GERALD SCHATTEN

metabolic awakening of the egg is rather slow, taking at least 3 minutes, and even sperm incorporation might not be the earliest of events during fertilization. To test this possibility we recently (Schatten and Hulser, 1982) used bioelectric (Chambers and de Armendi, 1979; deFelice and Dale, 1979; Hagiwara and Jaffe, 1979; Steinhardt et al., 1971; Taglietti, 1979) and video microscopic data to determine when the sperm triggers the earliest egg response. To do this we impaled an unfertilized egg with a microelectrode to record the bioelectric responses and simultaneously observed this very same egg with an insulated water immersion differential interference objective. Then, by the use of video signal mixing, the bioelectric responses were superimposed on the microscopic image of the impaled egg. With the aid of a slow-motion video analysis monitor, we determined that the successful sperm triggers the bioelectric responses of the egg well before any morphological changes occur. In other words, the successful sperm, while actively beating on the egg surface and indistinguishable from his doomed brethren, has already triggered the bioelectric responses at fertiliz,ation and initiated the fast block to polyspermy (Jaffe, 1976). At times the sperm would attach to the egg surface within 0.8 seconds of insemination and trigger the fertilization potential 720 msec (average 3.4 seconds) later. The spenn tail would not cease beating for another 12.9 seconds, which is the earliest morphological criterion of fertilization. It appears, then, that within milliseconds of sperm-egg binding, and well before sperm incorporation, the sperm is able to trigger the bioelectric responses of the egg. Whether this triggering is solely the result of sperm-egg attachment or indicates the moment of sperm-egg plasma membrane fusion is a matter of conjecture until techniques are developed that will permit accurate and swift time measurements on a living system with resolving power capable of visualizing or detecting fusing membranes. It is only after a dozen or more seconds that the cessation of tail beating, sperm incorporation, and the cortical reaction are noted. In consideration of the bioelectric responses at fertilization, it is noteworthy that neither microfilament (Dale and de Santis, 1981; Hulser and Schatten, 1980, 1982) nor microtubule inhibitors (Hulser and Schatten, 1980, 1982) significantly interfere with the membrane potential events.

B. IONIC CONTROLS 1. The Program of Activation

The program of activation (reviewed by Epel, 1977, 1978, 1980, and Nishioka, 1982) has been compiled as a result of nearly a century of work. Though details of this scheme are still under active investigation, the essential features are summarized in Fig. 30. Following the acrosome reaction the sperm has greatly elevated levels of cytoplasmic calcium ions [Ca2 ] and intracellular pH. It is in this stage that it fuses its membrane with the plasma membrane of the +

121

MOTILITY DURING FERTILIZATION

-

A23187

Sperm

Release of

r Increased

1

Na+free

//

7f

cytoplasmic pH

Exchange of external Na' for internal H'

Consequences of egg activation

-

1

i

I

DEVELOPMENT

N~ A ~ ~ +(e.g.~ protein + ~ synthesis, DNA synthesis, chromosome cycles, etc) pH 6.5

NH;

Procaine Nicotine FIG. 30. The ionic program of activation. This schematic diagram is a compilation of the program of metabolic activation and the sites at which it can be entered or terminated. During normal fertilization the sperm triggers a release of intracellular calcium that leads to an exchange of intracellular protons for extracellular sodium ions. This proton efflux results in the elevation of the cytoplasmic pH, which leads to many of the events triggered at fertilization. Artificial activation (double arrows in scheme) of unfertilized eggs is possible with divalent ionophores (A23187) or by treatments with isoosmotic nonelectrolyte solutions (not shown), which trigger the intracellular calcium release and the rest of activation, by direct manipulation of the cytoplasmic pH with alkaline ammonia, procaine. or nicotine. The sequence of activation may be blocked (double slashes) by transferring fertilized eggs into sodium-free media or by reducing the intracellular pH with a permeant weak acid such as sodium acetate at pH 6 . 5 . (From Schatten er d..1981b.)

egg and thereby triggers the onset of development. It is presently thought that the sperm enters the egg effectively as a calcium bomb (Schmidt et al., 1982; estimates of the intracellular calcium concentration in the acrosome-reacted sperm are over 10 mM) and perhaps as a pH-bomb (elevated internal pH). An alternative view is that the contribution of the sperm membrane to the egg membrane with its presumed calcium channels serves as an endogenous ionophore. The sudden and localized increase in cytoplasmic calcium at the site of sperm-egg fusion is sufficient to trigger the explosive discharge of adjacent cortical granules. This initial discharge stimulates a propagating wave of released calcium from intracellular stores, which is followed by the secretion of the cortical granules. The sites of the intracellularly sequestered calcium are still a matter of some debate; there is evidence implicating the cortical granules themselves as the sites of sequestration (Cardasis er a l . , 1978; Schatten and Hemmer, 1979). The calcium transient is concluded within 4 minutes of sperm-egg fusion, when the calcium is presumably resequestered. The intracellular release of calcium initiates a Na+-H+ exchange, which results in an increase in cytoplasmic pH. This alkalinization of the egg cytoplasm appears to be the pervasive ionic trigger for signaling the egg that it is now fertilized.

122

GERALD SCHAITEN

2. lntracellular pH as an Initiator of Motility during Fertilization The change in intracellular pH at fertilization appears to be a primary modulator of motility on the basis of studies with artificially activated eggs in which pH is manipulated to match that of fertilized eggs and in studies with normally fertilized eggs in which the normal alkalinization at fertilization is either suppressed or later reduced to the unfertilized value. In Fig. 3 1 , the fluorescence intensity of fluorescein, which is sensitive to pH at physiological ranges (Thomas et al., 1979). demonstrates the effects of fertilization and the ionic manipulation diagrammed in Fig. 30. In these studies, when the pH is permitted to rise to fertilized values, both microfilament bundles and microtubules are found to appear (Fig. 32), and when the values are prevented from reaching the fertilized values, neither cytoskeletal structure is found (Schatten et al., 1981a). The work by Mazia, Steinhardt, Epel, and co-workers (reviewed by Epel, 1977, 1978, 1980; Steinhardt and Winkler, 1979) using ammonia (Mazia and Ruby, 1974; Epel et al., 1974; Lop0 and Vacquier, 1977), procaine (Vacquier and Brandiff, 1975), nicotine (Johnson et al., 1976), and other weak bases has proven invaluable in elucidating the role of pH during egg activation. When these weak bases are applied at alkaline pHs, they rapidly diffuse across the plasma membrane and elevate the cytoplasmic pH (Shen and Steinhardt, 1978; Johnson and Epel, 1981) and thereby initiate egg activation as judged by protein synthesis, DNA synthesis, potassium conductance, polyadenylation of RNA, and cycles of chromosome condensation. Changes in cytoplasmic pH at fertilization have also been observed in amphibians (Lee and Steinhardt, 198 1; Nuccitelli et al., 1981; Webb and Nuccitelli, 1981). The first indications that artificial activation would result in cytoskeletal rearrangements came from the work of Mazia et al. (1975b), in which microvillar elongation was correlated with ammonia activation. Begg and Rebhun (1979), studying isolated egg cortices, have speculated that the change in intracellular pH results in the polymerization of microfilaments at the egg surface. Recent work (Begg et al., 1982; Carron and Longo, 1982) argues that microfilament bundling, but not strictly assembly, is under pH control. Figure 33A and B summarize the effects of ionic manipulation on the microvillar length of fertilized eggs. These results, based on SEM, do not discern flaccid microvilli, i.e., microvilli without microfilament core bundles. In 1939, Moore described the centration of the unfertilized female pronucleus in butyric acid-activated eggs, and this migration of the unfertilized egg nucleus during artificial activation has been recently confirmed (Mar, 1980; Bestor and Schatten, 1981). Paweletz and Mazia (1979) have elegantly documented the configuration of the microtubules in ammonia-activated eggs and can demonstrate, even in the absence of the introduction of the sperm centrioles, the nucleation of microtubules off osmiophilic foci. By the use of antitubulin nmmunofluorescence staining, Bestor and Schatten ( 1982) have recently described the

123

MOTILITY DURING FERTILIZATION

w

4

4c

7.6

choline seawater at 6min.

7.4

7.2 7.0

choline seawoter at Irnin.

0

2

4

6

8

10

12

14

16

18

TIME (minutes) FIG. 31. Modulation of pHi from 6.8 to 7.2. Water changes do not affect the emission level, showing that the dye is intracellular. Second and third panels show effects of weak acid (acetic) and weak base (NH,+) on pHi. Fourth panel shows that blocking Na+-H+ exchange reaction by substituting choline for Na+ blocks the increase in pHi. These data are in agreement with values obtained by intracellular pH recordings. pHi values on right axis were obtained with a pH microelectrode and agree with those determined by comparison of emission intensities at 435 and 495 nm. (From Bestor. 1981.)

I24

GERALD SCHATTEN

FIG.32. Effects of increased intracellular pH on unfertilized eggs: artificial activation. (A) A23187. With this divalent ionophore, a change in intracellular calcium and pH have been induced, and the egg nucleus undergoes centration and a monastral microtubular array forms. A . pimciularu. Bar, 10 pm. (From Bestor and Schatten, 1982.) (B) NH&l in calcium-free media. This activation

treatment results in a change in intracellular pH in the absence of a calcium release. Again the centration of the egg nucleus and a monastral microtubule shell are noted. A. punctulutu. Bar, 10 pm. (From Bestor and Schatten, 1982.)

sequence whereby a shell of microtubules forms following artificial activation. This subsurface radially oriented shell increases in diameter to push the female pronucleus toward the egg center. Finally, the shell condenses to form the “nuclear mitotic apparatus” of Paweletz and Mazia (1979) or apolar apparatus of Mazia et al. (1981). The effects of various ionic media on the velocity of centration is shown in Fig. 34. Grainger et al. (1979), Winkler et af. (1980). and Winkler and Steinhardt ( 1981) have recently demonstrated the reacidification of the fertilized egg

A

15 10 5

0 R . 2 . 4 . 6 .B~.BI.P1.41.61.82.~.2

IIICRDRE'IERS

MICROVILLAR LENGTH I1.2 e41

1

T

T

I .o

.8

E

a

.6 .4

.2 0

FIG. 33. Microvillar length and activation. Since microfilament assembly and fascin bundling are required for the elongation of the microvilli. microvillar length is an excellent marker to study the normal restructuring of the fertilized egg cortex and cytoskeleton. (A) A histogram of the length of 100 microvilli in unfertilized (stippled bars) and fertilized (solid bars) eggs at 40 minutes postinsemination. Note the increase in microvillar length. (B)The effects of manipulation of the program of activation on microvillar elongation. The average lengths of the microvilli of unfertilized and fertilized eggs are compared in the first two columns. When the normal rise in intracellular pH is blocked by the absence of sodium ions (third column) or with acidic acetate ions (fifth column) the microvilli remain at unfertilized egg length even after 40 minutes. These effects are reversible by the addition of sodium ions (fourth column) or the removal of acetate ions (sixth column). Unfertilized eggs activated with NH4CI, which results in an intracellular pH increase (seventh column) or A23 187, which results in both a release of intracellular calcium ions and an increased intracellular pH (eighth column), induce microvillar elongation and bundling to the length of normal fertilization. The ionophore activation is sensitive to the normal inhibitors of the program of activation (ninth and eleventh columns) and is recoverable (tenth and twelfth columns). It should be noted that flaccid microvilli might collapse during SEM processing and therefore these lengths are indicative of microvillar bundles.

126

GERALD SCHATTEN

1.0-

FIG. 34. Nuclear centration during artificial activation. The rates of the centrad migration of the egg nucleus during artificial activation demonstrates that when the intracellular pH is increased (with A23 I87 in the second column, NH4CI in the fourth column, calcium-free NH4CI in the fifth column, nicotine in the sixth column, and procaine in the seventh column) the movement of the egg nucleus ensues at a rapid pace. When the release of intracellular calcium alone is triggered (third column, with A23187 in Na+ -free media) the movement is not initiated.

cytoplasm and the concomitant decrease in protein synthesis. Carron and Longo (1980) have demonstrated the lack of the formation of the male pronucleus in such cases, and Schatten et al. (1981a) have demonstrated the inability of the sperm to be properly incorporated and of the pronuclear migrations to be initiated when the change in intracellular pH is blocked. In that work, using antitubulin immunofluorescence staining to observe microtubules, scanning and transmission electron microscopy to study microfilaments and time lapse video microscopy to observe directly motions in living cells, all signs of movement cease when intracellular pH is decreased and movement can be reinitiated when the pH is again elevated (see Tables VI and VIII). It appears then that the change in intracellular pH is one of the major regulatory factors at fertilization. The synergetic relations of pH, and cytoplasmic calcium ions are considered in the next section. Still unknown is the finer control whereby microfilaments and microtubules may function at one stage but not at another.

3 . Calcium Ions as Regulators The role of calcium ions in microtubule assembly is well established (Borisy and Olmsted, 1972; Weisenberg, 1972) and their transient release at fertilization makes them a likely candidate for a regulatory role (Mazia, 1937; Nakamura and

MOTILITY DURING FERTILIZATION

127

TABLE VI IONIC INHIBITION Ob M O V t M t N T S DLlRlNG

Na -free sea water at + 15 seconds + 2 0 seconds +30 seconds + I minute + 5 minutes t 10 minutes

FERTILIZATION

+

Na+-free sea water at 30 seconds to I minute, then recovery at + 5 minutes +90 minutes t 10 mM NHICI, pH 8.5 in Na+-free sea water at + 5 minutes t 15 minutes 10 mM Na acetate, pH 6.5 at I minute t 3 minutes

+

+

7 minutes

Movements cease Movements cease Movements cease Movements cease Normal divisions Normal divisions

Normal divisions: 5 minute delay Normal divisions; 90 minute delay

No division; movement attempts No division; movement attempts

Movements cease Movements cease Movements cease

10 mM Na acetale. pH 6.5 at 1-2.5 minutes, then

Sea water at + 5 minutes +30 minutes 10 mM NHdCI. pH 8.5 + 5 minutes Calcium-free sea water at 1 minute + 10 A23187

Normdl division: 5 minute delay Normal division: 30 minute delay Normal divisions

Normal divisions

Yasumasu, 1974). Microinjected calcium chelators block the sequence of activation at fertilization (Zucker and Steinhardt, 1978; Steinhardt et al., 1977; Gilkey et a / ., 1978; Ridgway et al., 1977). By preparing eggs for immunofluorescence staining in elevated calcium solutions, Bestor and Schatten (1981) have demonstrated that, whereas the incorporated sperm axoneme is not sensitive to calcium ions, the fibers of the sperm aster and the microtubules of the sperm aster, streak, and mitotic apparatus are. This result provides a clue concerning the sensitivity of these astral microtubules to calcium ions. However, artificial activation with the calcium ionophore A23187 (Steinhardt and Epel, 1974; Chambers et al., 1974) in sodium-free medium, in which a calcium release will occur normally though no change in intracellular pH can be detected, has failed to stimulate the

128

GERALD SCHATTEN

movements during artificial activation. A reasonable sequence would be that the change in intracellular pH and the calcium transient together stimulate the assembly and functioning of the egg cytoskeleton. The sperm tail ceases to beat in a manner already described by Gibbons and Gibbons (1979) in vitro. Following the calcium transient, at first the sperm tail resumes its beating, albeit erratically, and then the cytoplasmic calcium concentration declines to a level permitting microtubule assembly into the sperm aster. Perhaps there are fluctuations in free calcium concentrations in the now alkalinized egg cytoplasm, permitting first the assembly of the sperm aster and later its disassembly and then other cycles of streak formation and regression, followed by mitosis. During the period of elevated calcium ions, intracellular membrane fusion, i.e., fusion of the membranes of the male and female pronuclei might then be permitted. Perhaps even the alternation of microfilament-mediated and microtubule-mediated events is controlled strictly by fluctuations in free calcium in the alkalinized cytoplasm of the fertilized egg. It is germane to consider briefly the large body of hypotheses regarding the role of intracellular calcium ions as regulators of the events during cell division and during the cell cycle generally. The idea of fluctuations in cytoplasmic calcium content, produced by cyclical sequestration and release of internal calcium by membrane vesicles roughly analogous to the sarcoplasmic reticulum in muscle, has been described in detail by Petzelt (1972), Harris (1975), and Hepler (1980) and reviewed by Paweletz (1981). At mitosis, in consideration of the sensitivity in vitro of microtubule polymerization, the initial assembly of the mitotic apparatus would occur under a prevalence of low calcium concentration. As the chromosomes move during anaphase and as microtubule shortening apparently occurs, selective regions of the spindle would be exposed to slight increase in cytoplasmic calcium, likely the result of selective leakage of sequestered calcium from internal stores. Evidence in support of the notion that intracellular calcium is crucial for mitotic spindle fiber assembly comes from Kiehart (1981), who microinjected calcium or caffeine into the mitotic apparatus and demonstrated a transient loss in spindle birefringence, likely due to the depolymerization of adjacent microtubules. Harris ( 1975) has demonstrated by transmission electron microscopy the presence of abundant membrane vesicles in the regions surrounding the mitotic apparatus, and Schatten and Schatten (1977), using scanning electron microscopy, have demonstrated that these vesicles remain associated with the isolated mitotic apparatus. Silver et al. and Egiie and Nagle (1980) have shown that these vesicles found in association with the isolated mitotic apparatus can sequester exogenous radioactive calcium. Using the calcium-sensitive fluorescence chelate chlorotetracycline, Schatten and Schatten (1980b), Schatten et al. (1982b), Silver et al. (1980b), Sisken et al. (198 I), and Wolniak et al. (1980) have demonstrated the sequestration of endogenous calcium by membranes near the mitotic apparatus. Importantly, Wolniak e? al. (1980) have demonstrated cones of chlorotetracycline fluorescence, liable to be

MOTILITY DURING FERTILIZATION

129

sites of increased concentration of calcium stores, adjacent to microtubules predicted to disassemble during the anaphase movements. The studies in which calmodulin has been selectively localized to the chromosome-pole microtubules (Welsh et a / ., 1978), especially in light of the important discovery that calmodulin associated with microtubules renders the microtubules more sensitive to calcium ions (Marcum et a / . , 1978), invites further speculations that both regional increases in calcium and the specific association of calmodulin could well regulate the coordinated chromosome and pole movements during mitosis and perhaps during fertilization. For these studies, it is significant that the sperm aster is calcium labile and numerous membrane vesicles are found in association with it. It may well be that the decreased calcium concentration following the transient burst at sperm-egg fusion permits the assembly of microtubules on the sperm centrioles and that regional elevations in calcium concentration, potentiated by the specific binding of calmodulin to the microtubules that contact the female pronucleus, results in the selective disassembly of these microtubules and the migration of the female pronucleus, while the remaining microtubules continue to elongate. It appears likely that localized changes in ion concentration will be found to modulate the alterations in the cytoskeleton after its initial assembly.

4. The Cytoskeleton in Artificially Activated Eggs It is particularly instructive to analyze the cytoskeletal rearrangements during artificial activation. Clues concerning the natural regulation during fertilization can be easily inferred, and, furthermore, in the absence of a sperm, its direct contribution to the reorganization of the cytoskeleton following fertilization can be assessed. Microvilli of activated eggs elongate as shown by Mazia et al. (1975b). This microvillar elongation is likely to be the direct result of changes in intracellular [Ca2 ] and pH. Microtubules also form in artificially activated eggs (Paweletz and Mazia, 1978, 1979); however, in the absence of a sperm centriole, these microtubules do not develop into the sperm aster, but rather start as a subcortical disarray (Schatten et al. , 1981b; Bestor and Schatten, 1982). These microtubules elongate to form a radial shell that moves the female pronucleus toward the egg center, and finally this shell coalesces at the egg center to form the nuclear mitotic apparatus (Paweletz and Mazia, 1979). The appearance, using antitubulin immunofluorescence staining, of the microtubule arrays during artificial activation with ammonia and with the divalent ionophore A23 187 does not differ much, again underscoring the importance of pH and [Ca2+]as prime regulators for cytoskeletal formation. +

5 . Experimental Manipulation of pH during Fertilization There are primarily two sites at which the intracellular pH during normal fertilization can be manipulated (Fig. 30). Johnson et al. (1976) demonstrated the absence of the acid efflux when insemination occurred in choline-replaced

130

GERALD SCHAITEN

Na+-free sea water. Grainger et al. (1979) have succeeded in reacidifying the cytoplasm of fertilized eggs by the addition of 10 mM sodium acetate at pH 6.5. Perhaps it should also be mentioned that the addition of ammonium chloride also will increase the cytoplasmic pH of fertilized eggs though at present this method had not proven as successful as the former two. Begg and Rebhun (1979) analyzed the role of pH during the isolation of the egg cortex. Carron and Longo (1980) demonstrated that in sodium-free media the male pronucleus will not form properly, and, importantly, when sodium was added back, recovery could be observed. Begg et al. (1982) and Carron and Longo (1982) have demonstrated loose microfilament networks after treatments permitting the calcium transient, but blocking the change in pH. Schatten et al. (198 1b) have shown that microvillar extension, an indication of microfilament bundling, microtubule assembly, and the actual motions during normal fertilization will all be arrested in sodiumfree sea water or when the cytoplasmic pH is reduced to unfertilized values (Fig. 35A and B). Importantly, when the cytoplasmic pH is elevated by the addition of sodium ions, recovery is observed and the normal sequence of development ensues (Fig. 3%). Somewhat surprisingly, rather late stages, e.g., morula, blastula, can be arrested by the addition of acidified acetate and when recovery is permitted they are delayed in development by precisely the length of time during which their cytoplasmic pH was depressed (Henson and Schatten, unpublished results). In conclusion, then, cytoplasmic pH and calcium ions appear prime regulator ions of motility during fertilization.

C. CYCLICNUCLEOTIDES Castafieda and Tyler (1969) demonstrated an increase in adenyl cyclase activity following insemination, starting the speculation that cyclic nucleotides might play a role during egg activation. Yasumasu et al. (1973) extended this finding, demonstrating the cyclic fluctuations in cyclic AMP concentration correlated with each cell division and further demonstrating that caffeine, a potent phosphodiesterase inhibitor, would block cell division. Rebhun and co-workers (Amy and Rebhun, 1977; Nath and Rebhun, 1973, 1974) confirmed these findings and demonstrated that cyclic AMP analogs, e.g., dibutyryl cyclic AMP, would rapidly enter unfertilized eggs. However, they showed that cyclic nucleotide analogs would not affect the normal consequences of cell division, of which proper fertilization is a prerequisite. Utilizing the improved sensitivity of the radioimmunoassay, we have recently analyzed the concentrations of cyclic AMP (Fig. 36A) and cyclic GMP (Fig. 36B) during fertilization and early development (Cline and Schatten, 1983) and can demonstrate swift, dramatic, and reproducible fluctuations in both cyclic nucleotides throughout the fertilization sequence. In support of the work by

MOTILITY DURING FERTILIZATION

131

FIG. 35. Effects of decreased intracellular pH on the movements during fertilization: cessation of motility. (A). Sodium-free fertilization. When eggs, within 30 seconds of insemination. are transferred to sodium-free media. the normal Na+:H+ exchange, which results in the increased cytoplasmic pH. cannot occur. Even though the intracellular calcium release has occurred. in the absence of the change in intracellular pH. no movements occur and the cytoskeleton cannot form. In this image of an egg at 15 minutes postinsemination. the only microtubule staining structure is the incorporated sperm axoneme. A . puticlukuru. Bar. 10 p i . (From Schatten PI ul.. 1981a.) (B) Effects of acidic acetate. The intracellular pH can be reduced by the addition of sodium acetate at pH 6.5. which terminates the assembly of the cytoskeleton and all motions. Again the incorporated axoneme is the only microtubular structure present. A . punrtukuta. Bar, 10 pm. (From Schatten e r a / ., 1981a.) (C)Recovery. Upon the addition of sodium ions. the sodium:proton exchange occurs. Though development is delayed by about 15 minutes. these eggs can develop normally. This is an image of the first mitotic apparatus. A . punctularu. Bar. 10 pm. (From Schatten ct a / . . 1981a).

132

GERALD SCHATTEN

.

t

o\

I

~

,

,

o ~

Ip ;s,;;+

- Eggs

02

030

0

20

40

60

80

T I ME (minutes)

Post insemination

FIG.36. Cyclic nucleotides during fertilization. (A) Radioimmunoassays of cyclic AMP concentrations during fertilization. Fertilization triggers a swift doubling of the [CAMP].Fluctuations in [cyclic AMP] throughout devclopment are noted, with maxima occurring when microtubules are disassembling and microfilaments are assembling and with minima occurring when microtubules are assembling and microfilaments are disassembling. (€3)Radioimmunoassay of cyclic GMP concentrations during fertilization. As in the case of cyclic AMP, [cyclic GMP] fluctuates during fertilization, but apparently in the opposite direction. (C) Effects of isobutylmethylxanthine (IBMX). IBMX will prevent the normal completion of fertilization by blocking the disassembly of the sperm aster. The effects of IBMX may not be due solely to its role as a phosphodiesterase inhibitor since permeant cyclic AMP and cyclic GMP analogs do not prevent the normal movements at fertilization. Bar, 10 pm. 15. variegarus. (From Cline and Schatten, 1983.)

MOTILITY DURING FERTILIZATION

133

Rebhun and co-workers, we find that cyclic nucleotide analogs such as 8-bromo cyclic AMP and 8-bromo cyclic GMP as well as the dibutyryl analogs rapidly permeate these cells. However, these analogs themselves do not alter the motions during fertilization. Studies with phosphodiesterase inhibitors and adenyl cyclase activators are illuminating in evaluating a role for cyclic nucelotides as regulators during fertilization. Isobutylmethylxanthine (IBMX), a phosphodiesterase inhibitor, will result in a vast increase in cyclic AMP and cyclic GMP levels following fertilization. IBMX does not affect the unfertilized values much, an indication that phosphodiesterase activity is not the primary regulator of cyclic nucleotide concentration in unfertilized eggs, but in conformation of the original work of Casteiiada and Tyler (1969), adenyl cyclase activity is activated following fertilization. lBMX blocks fertilization at a site not dissimilar from that affected by taxol (Fig. 36C). Sperm incorporation, the formation of the fertilization cone, and the migration of the female pronucleus appear to occur normally. However, the sperm aster never disassembles (Fig. 36C) and continues to elongate to push the contiguous pronuclei past the egg center. These cells never undergo division. Caffeine and theophylline appear to have similar results. However, R,- I724/2, a phosphodiesterase inhibitor active at a site different from that of caffeine, increases cyclic AMP levels only slightly and does not disturb the motions at fertilization. Furthermore, cholera toxin, a potent adenyl cyclase activator, will increase the cyclic AMP concentration in unfertilized eggs, but does not interfere with the normal movements during fertilization. To summarize, then, the role of cyclic nucleotides as regulators during fertilization is still somewhat of an open question. The effects with the methyl xanthines may or may not involve cyclic nucleotides (indeed in sarcoplasmic reticulum it has been shown that caffeine interferes with calcium pumping). The fluctuations observed with radioimmunoassay techniques are indeed tantalizing, though the simplest explanation in light of the results with analogs and inhibitors other than methyl xanthines would tend to argue against a role of cyclic nucleotides in regulating motility. The times when cyclic nucleotides are elevated are correlated with increases in surface area, the first being during the time of the cortical reaction and then the first burst in microvillar elongation, and the second being the time of the second burst of microvillar elongation and then at each cytokinetic attempt; it may be that the newly exposed area results in an activation of adenyl cyclase. However, the times when the cyclic nucleotides are elevated also correlate with microfilament-mediated motions. The role of cyclic nucleotides as regulators during fertilization is particularly tantalizing in light of the many reports demonstrating phosphorylation of cytoskeletal accessory factors and the work of Keller et al. (1980) demonstrating protein kinase activities in cellular components at fertilization. This latter point is especially intriguing in light of the work from Browne et al. (1980). in which the cyclic GMP-dependent

134

GERALD SCHATTEN

protein kinase and a regulatory subunit of cyclic AMP protein kinase were localized along spindle fibers during mitosis. The possibility that the cytoskeleton somehow regulates cyclic nucleotide metabolism also exists. The future should resolve the question concerning the roles of cyclic nucleotides during fertilization and the mechanism by which their values are induced to undergo fluctuations.

D. CALMODULIN Calmodulin, also referred to as calcium-dependent regulatory (CDR) protein, has been shown to be a ubiquitous protein serving as an intermediary in nearly all calcium-triggered events (reviewed by Cheung, 1980; Means and Dedman, 1980). It has been isolated from eggs (Head e t a / . , 1979) and sperm (Garbers et al., 1980; Jones er al., 1980) and has been implicated by Epel et al. (1981) as a primary regulator in the activation of NAD kinase metabolism. Using immunofluoresence staining with calmodulin antiserum, Jones et al. (1978;) have localized calmodulin to the midpiece of mammalian sperm. Nishida and Kumagai (1980) have demonstrated, in vitro, the Ca2+ sensitivity of sea urchin tubulin during assembly to calmodulin. Hamaguchi and Iwasu (1980) have microinjected a fluorescent calmodulin derivative and have reported its localization near the sperm aster. Their data on calmodulin localization in the mitotic apparatus of blastomeres is convincing. Baker and Whitaker (1979) have shown that fluphenazine, a calmodulin inhibitor, will block the exocytosis of the cortical granules on isolated surface:. from unfertilized eggs. Steinhardt and Alderton (1981) have blocked cortical exocytosis on isolated surfaces with calmodulin antibody. This finding also implicates calmodulin activity in the calcium-mediated secretory event. Difficulties in permeation of calmodulin inhibitors have so far prevented any conclusive results on the effects of these compounds during the normal motions at fertilization.

E. OVERVIEW OF REGULATION It seems reasonable to conclude at this stage that protons, calcium, calmodulin, and perhaps cyclic nucleotides will all be involved in regulating the motile events at fertilization. Intracellular pH appears to be a major regulatory system in shifting the metabolism from that of an unfertilized egg into the fertilized state. Fluxes in calcium ions trigger the onset of the activation sequence and may also play a role in regulating the sort of motility active at each stage. For example, it is conceivable that during the initial alkalinization of the egg cytoplasm both microtubules assemble and microfilaments rigidify as bundles. However, the

MOTILITY DURING FERTILIZATION

135

prevailing calcium concentration might be too high for microtubules to form and consequently, only the microfilament-mediated motions are observed. Extracellular Mg2+ has been found to be required for fertilization and might well be necessary for sperm-egg adhesion and fusion (Sano and Mohri, 1976, 1977; Sano et al., 1980; Ludert and Schatten, unpublished results). It is quite clear from the work by Epel et al. (1981) that calmodulin plays a role during the fertilization sequence and indeed its isolation from unfertilized eggs has been convincingly demonstrated. The significance of the fluctuations in cyclic nucleotide concentration remains still elusive.

VI. Consequences of Fertilization for Later Embryonic Development A. THECENTRIOLES

I . Normal Fertilization The major contributions of the sperm to the egg at fertilization are the sperm nucleus, the ionic trigger for initiating development, and the pair of centrioles serving as the basal body for the sperm tail. Unfertilized eggs do not contain morphologically apparent centrioles (Dirksen, I96 1); for the fertilized egg to develop a normal bipolar mitotic apparatus, it must receive one pair, and only one pair, of sperm centrioles. A simple, but elegant, experiment highlighting this contribution of the sperm to the egg, in the absence of any complementary contribution by the female pronucleus, has been performed by fertilization of enucleated unfertilized eggs (Harvey, 1956). In this work, the enucleated merogon will normally incorporate the sperm, and a bipolar mitotic apparatus, of course with only the haploid sperm chromosomes oriented along the metaphase plate will develop (R. Balczon, unpublished results). During normal fertilization then, the sperm brings in a pair of centrioles, which organize a bipolar mitotic apparatus. Each supernumerary sperm during polyspermy contributes a pair, resulting in multipolar mitoses and aberrant cleavages.

2 . Parthenogenesis There are conditions in which asters and centrioles can be induced de novo and true parthenogenesis has been observed (Brandriff et al., 1975). Loeb (1913) described the two-step procedure for inducing parthenogenesis, the first step being an activating treatment and the second he referred to as a “corrective treatment.” In Loeb’s case, he activated with butyric acid and later induced asters by the use of hypertonic sea water. The discovery that heavy water (D,O) would stimulate the assembly of microtubules led modem workers to the conclusion that the stimulation of the microtubule assembly in unfertilized activated

136

GERALD SCHATTEN

eggs could result in the formation of asters and centrioles in unfertilized eggs (Miki-Nomura, 1977; Brachet, 1974; Mazia, 1977). It is particularly interesting in the work of Mazia that not only are first- and second-generation centrioles found in their characteristic perpendicular oriention, but indeed even the thirdgeneration centriolar anlage can be found in thin sections. While it is clear that heavy water and hypertonic sea water will induce centrioles de n o w , the simple conclusion that the artificial induction of microtubule assembly in unfertilized eggs will directly result in centriolegenesis is unwarranted. This statement is based on two different lines of evidence: the artificial induction of microtubule assembly during artificial activation and the formation of asters in taxol-treated unfertilized eggs. During artificial activation, e.g., ammonia (Paweletz and Mazia, 1979), procaine (Moy and Vacquier, 1977), or A23187 (Bestor and Schatten, 1982), microtubules are found to assemble in a radial array, which finally coalesces toward the egg center to form the"nuc1ear mitotic apparatus" (Paweletz and Mazia, 1978). On the basis of transmission electron microscopy and indirect immunofluorescence microscopy these fibers are unquestionably microtubules, but TEM fails to demonstrate any centrioles as microtubule organizing centers (MTOCs); instead only osmiophilic foci appear as the nucleation centers. When unfertilized eggs are treated with taxol, artificial activation is not noted as judged by the following criteria: DNA synthesis, chromosome cycles, microvillar elongation, nuclear breakdown, and the cortical reaction (see Section 1V,B,2). However, antitubulin immunofluorescence staining and transmission electron microscopy demonstrate the presence of numerous asters in these unfertilized eggs. The effect of taxol is probably to block the normal disassembly of microtubules, which might well far exceed the rate of assembly in unfertilized eggs; this rate difference would explain the paucity of microtubules in unfertilized eggs. Importantly, these asters are not nucleated by centrioles, but rather by osmiophilic foci, or at times, in association with annulate lamellae. These results during artificial activation and with taxol argue that the formation of centrioles is more complicated than simply the initiation of microtubule assembly. The contribution of nucleic acids in centriole formation (Berns et al., 1977; Heideman et et al., 1977; Went, 1977; Peterson and Berns, 1978) is likely to represent one of the additional synthetic events required for centriole formation. Still unresolved is the fate of the centrioles in the maturing oocyte, which appear to be lost in the fertilized egg. Furthermore, the procedure whereby centrioles can be induced de n o w during artificial activation followed by treatment with heavy water, but not during normal fertilization followed by treatment with heavy water, leads to the speculation that the sperm centrioles play a role in suppressing the formation of exogenous centrioles. The manner in which such centriolar dominance can be effected is mysterious.

137

MOTILITY DURING FERTILIZATION

ESIABLlSHMtNI Average angle between axis for first cleavage furrow and Site of sperm entry Movement of sperm aster prior to the migration of the female pronucleus Centration of contiguous pronuclei

TABLE VII W- FIKSIEMBRYONIC AXIS

Ccmpressed eggs

Spherical eggs

Total

13.5 t 12.5" ( N = l 3 )

32.6 2 24.3" (27)

27.9 t 24.6" (40)

24.5 2 27.0" (31)

20.4

6.5

?

4.4" (13)

5.8 t 4.5" (13)

7.9

-t

10.9" (14)

?

25.3" (44)

6.9 2 8.1" (27)

B. THEESTABLISHMENT OF THE FIRSTEMBRYONIC Axis The establishment of the first embryonic axis (Table VII) has been a problem of classic interest. To restate the problem in more modem terms, the question relates to the final positioning of the pair of sperm centrioles prior to the formation of the first mitotic apparatus. The question simply stated is, "Does the unfertilized egg have a predetermined axis for first division, to which the sperm centrioles migrate along a path specified within the egg, or is the egg axis completely unspecified and the movements during sperm incorporation and during the pronuclear migrations of importance in specifying the first axis'?'' Wilson and Mathews ( 1895), studying slightly compressed L . variegatus eggs, report that cleavage occurs within 15" of the radius passing through the sperm-egg fusion site. Horstadius (1928), using the pigment band of the unfertilized Paracentrotus lividus eggs and vital stains as markers, presents evidence that the random site of sperm-egg fusion is inconsequential to the first cleavage axis. Boveri (1901), studying P . lividus, diagrams a prominent micropyle at the pole distal from the pigment ring through which the sperm entry typically occurred; entry at this point might play a role in the establishment of the cleavage plane. The unfertilized egg of P . lividus, having a pigment band and therefore an observable polarity. must be considered different from that of L . variegatus, in which no polarity can be detected. It is noteworthy that the only other situation in which polarity in an unfertilized echinoderm egg can be achieved that results in the specification of the first cleavage plane is in the centrifugally stratified egg, classically of Arbacia puncrulata (Harvey, 1956). In these cases a polarity is centrifugally induced so that the pigment granules reside at the centrifugal pole and cleavage typically bisects this pigment band (Lyon, 1907; Morgan and Lyon, 1907; Morgan and Spooner, 1909; Motomura, 1936). The point of sperm entry is indeed variable in centrifuged A . puncrulata eggs and time lapse video microscopy (Schatten, unpublished results) confirms that cleavage bisects the pigment band.

138

GERALD SCHATTEN

The mitotic apparatus in these eggs always forms in the centripetal regions, which comprise the endoplasmic reticulum and the mitochondrial layers; apparently the densely packed yolk and pigment at the centrifugal pole will not support the formation of the mitotic apparatus. The steric restraint of the lower viscosity centripetal regions, i.e., the endoplasmic reticulum and mitochondrial layers, alters the positioning to the egg equator; the resultant cytokinesis then divides the egg through the pigment band. Though it is not possible to state with certainty that a similar mechanism is responsible for the predetermined polarity in the P. lividus egg, Lindahl and Oerstrom (1933) describe an increased cytoplasmic viscosity by the pigment ring in P . lividus that possibly could account for the orientation of the mitotic apparatus at first division. In L . variegatus, the first embryonic axis appears to be specified by the final positioning of the sperm-derived centrosomes. The centrosomes separate in opposite directions from each other immediately prior to syngamy. This antiparallel separation occurs at right angles to the direction of pronuclear centration. The direction of pronuclear centration is usually quite similar to the movement of the sperm aster, which, in the absence of an extremely greater lateral displacement of the sperm during incorporation, is generally near the egg radius passing through the sperm-egg fusion site. Czihak (1973) noted the importance of astral microtubules in the establishment of cleavage planes. It should be noted that vortical motions occur within the egg cytoplasm, perhaps resulting from or responsible for the spiral cortical microtubules (Harris et al., 1980b), which can shift the axis of division from that which might be predicted in light of the pronuclear migrations. Schroeder (1980) has correctly reported that cleavage does not always bisect the sperm entry site (see Table 111). However, Schatten (1981b) reports finding that cleavage occurs within 8" of the direction of pronuclear centration. It may be that the sperm centrioles separate at right angles to the direction of pronuclear centration. This movement would result in the positioning of the first mitotic apparatus perpendicular to the centration direction, with the resultant first cleavage plane parallel. The discrepancy between reports of Wilson and Mathews (1895) that cleavage occurs along the egg radius passing through the sperm entry site and the findings of Horstadius (1928, 1939) that the axis is variable can in part be explained as the result of compressing of the developing zygote. Spherical eggs, perhaps having a greater freedom for cytoplasmic rotations, have a larger variation between the radii of cleavage and the sperm entry site.

TAIL C. FATEOF THE SPERM A consideration of the fate of the sperm tail seems warranted in this section in

light of the misconception that the sperm tail is not incorporated normally during

MOTILITY DURING FERTILIZATION

I39

fertilization. This does not appear to be the case on the basis of transmission electron microscopy (Longo and Anderson, 1968), scanning electron microscopy (Schatten and Mazia, 1976a,b), time lapse video microscopy (Schatten, I98 I a,b), and indirect immunofluorescence microscopy with tubulin antibody (Bestor and Schatten, 1981). In light of the new discoveries concerning the primary cilium of many mammalian cells in culture (Tucker er al., 1979), it is hard to imagine any cilium more primary than that of the incorporated sperm axoneme. However, the axoneme does not seem to persist far beyond the first division. In Fig. 37, it has been detected by the utilization of a high-calcium fixative that disassembles the labile streak microtubules. It is interesting that it remains associated with one of the centrioles, presumably the original basal body in the sperm midpiece. The sperm tail may have an important function in propelling the sperm to the egg surface, but experiments by Epel er al. (1977), using immotile or tailless sperm, have reported successful incorporation, indicating a lack of a crucial role during normal fertilization. The discovery of the beating of the sperm tail in the egg cytoplasm might indicate a role for the tail in the movement of the sperm from the interface of the egg cortex into the egg cytoplasm, but this remains to be demonstrated. In summary, the fate of the sperm tail is not yet known, though it seems reasonable to predict that, during the cycles of microtubule assembly and disassembly, the axoneme itself appears eventually to be disassembled and could even contribute to the pool of tubulin required for the fdrmation of mitotic apparatus during cleavages and perhaps even as a precursor for cilia during blastula formation.

FIG.37. Persistance of sperm axoneme and calcium sensitivity of cytoplasmic microtubules. (A) Sperm aster stage (7 minutes) egg fixed in 90% methanol-108 0.01 M CaC12. Astral rays have disappeared but the axoneme is unaffected. (B) Streak-stage cell fixed as above; sperm axoneme is associated with one polc of the elongated streak-stage nucleus (arrow). A . punctulara is depicted here; similar results were obtained with gametes from L. variegotus. Bars, A, 10 pm; B . I pm. (Reprinted with permission from Bestor and Schatten, 1981.)

140

GERALD SCHATTEN

VII. Motility during Fertilization and Its Regulation: A Model A. MOTILITY DURING FERTILIZATION 1. Sperm Incorporation A compilation of the movements during fertilization, the responsible systems of motility, and their regulation is presented in this section. The acrosome reaction of the sperm must be considered the first event during fertilization (Aketa and Ohta, 1977). In sea urchins, it is triggered by diffusible factors from the egg that, presumably by binding to the sperm plasma membrane, induce the opening of ionic channels. Following the uptake of calcium ions and an efflux of protons, as well as possible changes in membrane potential, the secretion of the acrosomal vesicle and the extrusion of the acrosomal process occur. The important points to underscore concerning the sperm at this juncture are that the tail is actively beating, that the actin in the periacrosomal cap is polymerized into a presumably rigid fiber, that the cytoplasmic ionic content of the sperm is elevated in calcium ions and depleted in protons, and that the sperm carries a pair of centrioles as well as the haploid male genome. The sperm swimming to the egg surface undergo the acrosome reaction, which externalizes the species-specific protein bindin, which attaches the sperm to the egg surface. Within hundreds of milliseconds after sperm-egg attachment the bioelectric changes associated with fertilization are initiated (Schatten and Hulser, 1982). These bioelectric potentials have been implicated in the fast block to polyspermy (Jaffe, 1976) and appear triggered quite swiftly after sperm-egg attachment. It is still a matter of conjecture when sperm-egg plasma membrane fusion actually occurs; in the absence of resolving the necessary detail with the light microscope debates will prevail. However, if the bioelectric responses at fertilization are indeed triggered by the successful sperm at the moment of sperm-egg fusion, then sperm incorporation does not begin until 15 seconds later. As shown in Fig. 38, sperm incorporation begins following the gyrations of the attached sperm about their acrosomes and is characterized by the sudden immobilization of the sperm axoneme. The successful sperm is held erect and stationary on the egg surface presumably by the extension of the adjacent egg microvilli. These microvilli elongate and engulf the entire sperm and form the fertilization cone. Following the formation of the fertilization cone this sperm is incorporated into the cytoplasm in two stages. First it is pulled from the exterior into the subcortical region of the egg and then it is translocated laterally along the egg cortex to be discharged into the egg cytoplasm proper. During the final stages of sperm incorporation the sperm tail resumes its beatit.:, albeit erratically, within the egg cytoplasm.

MOTILITY DURING FERTILIZATION

14 I

2 . Pronucleur Motions The growth of the sperm aster is responsible both for the cytoplasmic movements of the sperm and egg nuclei and for terminating the lateral displacement of the sperm during incorporation. This latter point is demonstrated in Table 1V in which sperm incorporation is found to occur for a longer time and at a faster rate in eggs inseminated in the presence of microtubule inhibitors. The initial assembly of microtubules on the sperm centrioles pushes the male pronucleus toward the egg center. This appears to be simply a steric effect resulting from the elongation of these fibers. It is of interest, however, that the incorporated sperm axoneme ceases to beat once the sperm aster begins to assemble, an indication of a change in the intracellular ionic concentration. The sperm aster initially is a radially symmetric spherical structure. As such it would be expected to move the sperm nucleus toward the egg center regardless of its position within the spherical egg. When the microtubules of the sperm aster elongate to an extent that they are able to contact the nuclear surface of the female pronucleus, the swift migration of the female pronucleus occurs. During this migration the sperm aster develops an asymmetric appearance and indeed inhibitors of microtubule disassembly will block this motion. Speculations that the disassembly of the microtubules that interconnect the sperm and egg nuclei generate the motive force for this migration seem unlikely. The actual mechanism for this motion is not yet clear, while it is certain that blocking the disassembly of the sperm aster microtubules will block the normal motion of the female pronucleus. Other interpretations accounting for generation of the force are still attractive. To clarify this point, it appears clear that the normal disassembly is necessary for the female pronucleus to invade the sperm aster but that this disassembly is actually moving the egg nucleus is still the subject of conjecture. This point will be considered in Section VIII. Following the migration of the female pronucleus into the center of the sperm aster, the now adjacent pronuclei are moved to the egg center. This motion is caused by the elongation of the remaining majority of sperm astral microtubules; only those sperm astral microtubules important in pushing the adjacent pronuclei together into the egg center are those inserted into the egg cortex. 3. Cytoskeletal Changes Leading to Division

In addition to its de novo formation, the cytoskeleton undergoes further reconfigurations during the first cell cycle prior to cell division. In the case of microfilaments these involve two bursts of microvillar elongation; in the case of microtubules the sperm aster is lost and the interim apparatus or the “streak” forms followed by the assembly of the mitotic apparatus.

IONICREGULATION

TABLE VIlI MOTILITYDURING FERTILIZATION”

OF

Microfilament-mediated events

Microtubule-mediated events

Fertilization cone

Microvillar elongation

Microvillar bundling

Cytokinesis

Sperm aster fomiation

Nuclear migrations

Microtubule “shell”

+++

+++ +++ +++ +++

+++

+++

+++

NA

+++

+++ +++

+++

+

NA NA

+++ ++ ++

+++ +++

+

NA

++

+++

NA

NA

++

+++

NA

Mitosis

1. Changes in intracellular pH [pHJ and

intracellular [Ca2+ ] release Normal fertilization Fertilization, into Ca2+ -free sea water

(SW) 5 pM A23181 5 pM A23187 in Ca2+-free SW 2. pHi change only 10 mM NH4CI, pH 8 . 5 , 45 minutes 10 mM N h C I . pH 8.5 in CaZ+-free SW. 45 iriirruks 10 mM procaine, pH 8.0 10 mM nicotine, pH 8.5

+++ NA NA

NA iu’A NA NA

++ ++ + +

+++

+++

+

NA NA

++ +/-

NA

+/-

+++ +++ NA NA

NA NA

3. [Ca2+] only Fertilization, into Na+-free SW 5 phf A23187 in Na+-free SW Fertilization, + 10 mM Na acetate, pH 6.5 5 phf A23187. + 10 mM Na acetate, pH 6.5

+ +/NA

++ ++

++/-

++

-

-

NA

++

-

-

++

+++

+++

++ +/-

+++ +++

NA NA

-

-

-

-

-

-

NA

-

-

NA

-

-

-

NA

-

-

NA

+++

+++

+++

NA

+++

+++

+++

+++

+++

NA

+++

+++

-

-

NA

-

+++

+++

-

NA

+++

-

NA

+++

+++

-

NA

+++

-

NA

~

~

4. Recovery from blocked pH, change

Fertilization, + 10 mM NaAc, pH 6.5, then NH4CI, pH 8.5 Fertilization. into Na+-free SW, then

sw

fi

Fertilization, into Na+-free SW. then NHdCI, pH 8.5 5 phf A23187, +I0 mM NaAc, pH 6.5. then SW 5 phf A23187. into Na+-free SW. then SW (1

+ + +, Normal: + +, slightly reduced: +. reduced; + I - .

marginal: -, absent: NA. not applicable or expected

144

GERALD SCHATTEN

I

I

FIG. 38. The movements during fertilization. Sperm attach to the egg surface ( A ) and gyrate about their attachment sites for varying times prior to fusion (B). Following a rapid cortical contraction radiating from the fusion site. the fertilization coat elevates (D-F). Sperm incorporation is characterized by the formation of the fertilization cone around the erect and stationary sperm; the sperm tail is immotile at this stage (C-F). The sperm glides along the egg cortex during penetration ( G , H).The formation of the sperm aster moves the male pronucleus centripetally (1, J:l. The migration of the female pronucleus occurs when the fibers of the sperm aster interconnect the pronuclei (K,L). The adjacent pronuclei are moved to the egg center by the continuing elongalion of

MOTILITY DURING FERTILIZATION

145

The importance of the bursts in microvillar elongation caused by microfilament assembly are still somewhat obscure. Schroeder ( 1979) presents convincing arguments that the first burst in microvillar elongation is involved in taking up the membrane slack resulting from the exocytosis of the cortical granules. Banzhaf et al. (1980) and Schatten and Schatten (198 1) have demonstrated that prevention of the microvillar elongation by the microfilament inhibitors, the cytochalasins, will prevent the normal restructuring of the fertilized egg cortex. The second burst in microvillar elongation occurring shortly before prophase might well be involved in reorganizing the egg cortex in preparation for cytokinesis. It is of interest to speculate on the role of these microvilli during interphase. Observations of microvillar behavior using video enhancement techniques and water immersion objectives with high numerical apertures (G. Schatten, unpublished results) demonstrate the vigorous activity of the microvilli throughout the first cell cycle. Scanning electron micrographs of egg microvilli of course could not be expected to detail the motility of these cellular processes, and those static electron images leave viewers with a lack of the sense of activity. Video tape studies, at the limit of present-day optical detecting methods, demonstrate the incessant activity of these microfilament-containing processes and raises questions about the requirement for their motion. In effect, they appear to be constantly moving particles from the egg periphery toward the center and perhaps even setting up cytoplasmic eddy currents. It may well be that there is a swift turnover in membrane components or an absorption of material from the cell’s exterior and resultant movement to the egg center. In any case the egg cortex throughout the first cell cycle is under constant motion even during times when neither sperm incorporation nor cytokinesis is occurring, leaving questions regarding the function of this cortical churning. Analogous to the cyclical burst in microfilament activity, the microtubules also undergo cycles of assembly and disassembly. Immediately following the formation of the sperm aster and the centration of the pronuclei, the sperm aster typically disassembles. The cortical polysaccharide described by Naruse and Sakai (1981) is a likely candidate for promoting the disassembly of these microtubules. This cycle is followed by another cycle of microtubule assembly and disassembly, that of the “streak” or interim apparatus. Interestingly, the “streak” or interim apparatus assembles at roughly the time of the second burst in microvillar elongation. However, the function of this structure is as obscure as the sperm aster (M): the centrioles may separate during this motion. and the sperm aster appears to have two focal points. Syngamy typically occurs at the egg center after the disassembly of the sperm aster ( N ) . The streak forms around and distorts the zygote nucleus (0).The axis of the streak (0)is usually identical to the axis of the mitotic apparatus ( Q ) . The streak is disassembled prior to the nuclear breakdown at prophase (P). Cleavage is perpendicular to the axis of the mitotic apparatus and is usually parallel to the egg radius passing through the sperm entry site ( Q ) . (Reprinted with permission from Schatten. 1981b.)

146

GERALD SCHAlTEN

that of the second microvillar burst. The interim apparatus is a planar hipolar structure that might be involved in preparing the cell to assemble the mitotic apparatus, perhaps by moving the centrioles to their final position for mitosis, or perhaps by preparing the nucleus for division.

B. REGULATIONOF MOTILITY I . The Formation of the Cytoskeleton The assembly of the egg’s cytoskeleton is under the control of cytoplasmic pH and is likely regulated by fluctuations, and perhaps microenvironments, of cytoplasmic calcium ions. An important question that must ultimately be answered is whether intracellular pH regulates cytoskeletal formation in most all cells or whether it is unique to eggs at fertilization. The latter point is within the realm of possibility since the unfertilized egg has no cytoskeleton and it is only after sperm incorporation that a network of cytoplasmic microtubules and tnicrofilaments assembles. In all other cells there is an interphase cytoskeleton that undergoes rearrangements in preparation for the mitotic phase. It may well be that the factors that control the global assembly of a cytoskeleton are distinct from those that modulate the reconfiguration and reorganization of existing cytoskeletal components. The conclusion that the egg cytoskeleton is under the control of intracellular pH is described in full detail in Section V,A and is primarily based on two different series of experiments (Fig. 30). In one of the two protocols the intracellular pH changes normally occurring at fertilization are blocked by transference of the inseminated eggs following their intracellular calcium release to media that preclude the change. In these cases motility and the assembly of the cytoskeleton cease. In the other sequence of experiments unfertilized egl,JS are treated to raise artificially the intracellular pH, in which case movements of the egg nucleus and the formation of cytoskeletal components are observed. The simplest common denominator in these experiments is that, when the normal intracellular pH changes are blocked, no motion or cytoskeleton appears. When the pH change is induced either naturally or artificially, both motions arid the formation of the cytoskeleton are observed. These experiments highlight the importance of pH and [Ca2+]in regulating the egg cytoskeleton. 2 . Sperm Incorporation: Is the Sperm an H -Bomb and a Ca2+-Bomb? If, as stated in the previous paragraphs, elevations in intracellular pH and calcium ions are responsible for the assembly of the cytoskeleton, then how is it that the fertilization cone forms only at the region adjacent to the fused successful spermatozoon? A possible solution for this paradox is the notion that the sperm carries with it the triggers to elicit a transient and localized assembly of the egg microfilaments. This localized burst of microfilament assembly would result in the formation of the fertilization cone around the successful sperm and would be followed, after the program of activation results in the cytoplasmic pH and +

MOTILITY DURING FERTILIZATION

147

[Ca2 ] changes, by the global assembly of microfilaments observable as the first burst in microvillar elongation. This scheme seems plausible in light of the ionic studies performed on sperm during the acrosome reaction. In this body of work it has been demonstrated that an acid efflux occurs during the acrosome reaction and results in an alkalinization and increased [Ca2 1 of the sperm’s cytoplasm. It should also be pointed out that the sperm takes up considerable calcium during the acrosome reaction (Schackman et a/., 1978) so that when the sperm fuses with the egg it is effectively introducing both a high concentration of calcium ions and a depleted concentration of protons, i.e., a somewhat alkaline cytoplasm. The assembly of egg cortical actin, presumably with the polarity already specified, results first in the extension of microvilli and then later in the full growth of the fertilization cone. +

+

3. The Regularion of the Sperm Aster Analogous to the alleged Ca2 - and pH-sensitive assembly of microfilaments to form the fertilization cone, microtubule assembly on the incorporated sperm centrioles seems also to be under the identical [Ca2 ] and pH regulatory systems. This then implies that the program of activation, which permits the change in intracellular pH and [Ca2+] and the resultant extension of the fertilization cone and burst in microvillar length, would also result in the assembly of the microtubules to form the sperm aster. If this simplified scheme is entirely accurate, a problem in regulating the premature assembly of the sperm aster must be addressed for fertilization to occur successfully. The microfilaments in the fertilization cone must first assemble, the sperm must be incorporated, and then the aster must assemble to effect the pronuclear movements. Were microtubule and microfilament assembly occurring instantly following fusion and simultaneously in the fertilization cone, it is conceivable that the male pronucleus could be caught in a web of microtubules and microfilaments. Several different systems might be involved in preventing this occurrence. In the first case the radially projecting nature of microtubules growing in asters might move the male pronucleus out of any cortical web regardless of its density, which renders this issue not a problem. Second, it may well be that either the sperm centriole or the egg’s tubulin pool is momentarily incompetent to assemble microtubules and that there are regulatory factors in the egg cytoplasm preventing the premature assembly of microtubules. The most attractive hypothesis to regulate the motility, however, again relies on the role of the sperm both as a transient pH bomb and also as a calcium bomb, which triggers the sequence of egg activation. Finally the transient elevation in [Ca2+] might preclude premature microtubule polymerization. In this scheme the transient alkalinization and increased [Ca2+]of the egg cytoplasm adjacent to the site of fusion would result in a localized assembly of microfilaments to form the fertilization cone. During these first minutes the cone would extend around the successful sperm, and the sperm would be drawn into +

+

148

GERALD SCHATTEN

the egg cytoplasm. Simultaneously with the transient localized assembly of the microfilaments in the fertilization cone, the sperm discharges its cytoplasmic calcium into the egg cytoplasm, resulting in the propagative explosion of the sequestered egg calcium stores, which transiently increases the egg’s cytoplasmic calcium content. It may well be that this elevation in cytoplasmic calcium is responsible for the immotility of the sperm tail during the early stages of incorporation (Gibbons and Gibbons, 1979). Following the seconds required for the calcium release and the minutes required for the sodium:proton exchange, the entire egg cytoplasm would undergo its elevation in intracellular pH. Now the cytoplasmic conditions are favorable for microfilament assembly and bundling. As the cytoplasmic [Ca2 ] is reduced by resequestration, conditions favoring microtubule assembly would develop. In the cases of microfilaments they would assemble on all the microvilli; in the case of microtubules they would only assemble on the sperm centrioles. This hypothesis then predicts that the intracellular pH and [Ca2+] of the sperm is sufficient to trigger a transient assembly of microfilaments to form the fertilization cone. It also predicts that the sperm’s intracellular calcium is sufficient to trigger the release of the egg’s calcium stores, which in turn leads to a change in intracellular pH and then the global reorganization of the egg’s cytoskeleton. This transient [Ca2 ] and pH change resulting in the localized assembly of the fertilization cone followed by the global [Ca2+] and pH changes resulting in the organization of the entire cytoskeleton would be sufficient to account for the spatial and temporal events occurring during the first minutes of fertilization. Another problem of interest to be considered in the next section is the manner in which the sperm aster can simultaneously push the male pronucleus toward the egg’s center and pull the female pronucleus toward its center. This phenomenon could be explained on the basis of localized ionic conditions due to the intrinsic polarity of microtubules, and will be considered in Section VII,C,3. +

+

4. Secondary Modulators

It appears quite likely that in addition to the global changes in intracellular pH and [Ca2 ] and localized fluctuations in intracellular calcium, other modulators of motility may exist. The importance of localized regions of intracellular calcium concentrations during mitosis is now being established and it appears likely that the same system of regulation will be active during fertilization. The possibility of other modulators must be viewed as attractive. +

C. MECHANISMS FOR MOVEMENT: IMPLICATIONS FOR OTHER INTRACELLULAR TRANSLOCATIONS 1. Sperm Incorporation While it is clear that microfilament assembly in the egg is required for sperm incorporation, the manner in which the sperm is first pulled into the egg

MOTILITY DURING FERTILIZATION

I49

cytoplasm and later pushed along the egg cortex is still obscure. The polarity of the sperm acrosome process when decorated with heavy meromyosin is opposite to that of the egg microfilaments, which would preclude a simple model in which microfilament sliding, analogous to skeletal muscle contraction, would pull the sperm into the egg cytoplasm. In contrast, were sliding between the sperm acrosomal microfilaments and the egg microvillar microfilaments to occur it would in effect push the sperm off the egg surface. This sort of movement could assist in the final phase of sperm incorporation after the rotation of the sperm. More attractive hypotheses for understanding the mechanism generating the movement of the sperm during incorporation include the idea of microfilament treadmilling within the egg microvilli and especially within the fertilization cone. This treadmilling, a process not yet conclusively demonstrated to occur in vivo, is well documented in vitro both for microtubules and for microfilaments. It relies on the intrinsic polarity of both microfilaments and microtubules. Since these fibers have preferential ends for assembly and preferential ends for disassembly, there is a constant motion of monomers flowing through the polymer even in an apparently static fiber. The analogy to a conveyor belt represents a gross simplification but will underscore the point that in the absence of net increases or decreases in length a polymeric fiber such as a microtubule or a microfilament could, by adding monomers at one end and removing monomers at the other end, move structures associated with its surface. For treadmilling to account for sperm incorporation would imply that actin is assembling at the peripheral regions of the microvilli and disassembling at the cortical junction with the cytoplasm. The flow of monomers would either selectively move the sperm toward egg cytoplasm and perhaps establish more generalized current within the fertilization cone that would sweep particles toward the cytoplasm. This sort of behavior of the fertilization cone has been observed in phalloidin-treated eggs in which the fertilization cone appeared as a fountain sweeping particles within it toward the cytoplasm. Another possible factor that might drive the sperm from the exterior during sperm incorporation might be the accumulation of water during the acrosorne reaction. Tilney and Inout ( I 982) recently presented convincing evidence that the actin in the periacrosomal cap is quite dehydrated and during the acrosomal process takes up considerable water. This cap, the hydration of which appears to drive the extension of the acrosomal process, might still be in a slightly dehydrated state during the initial stages of incorporation and the anticipated flux of water driving from the exterior into the cytoplasm might also effectively flush the sperm into the egg cytoplasm. This notion is not completely unwarranted since the fertilization cone, at least during the first minute of its growth, is rather swollen and bulbous, later to be replaced by the more angular, constricted, and extended fertilization cone. The termination of sperm incorporation is mediated by the growth of the sperm aster on the sperm centrioles. The extension of the sperm astral microtubules

150

GERALD SCHATTEN

would well be expected to increase the drag of the male pronucleus, rendering it resistant to further lateral displacement along the egg cortex. The growth of the sperm aster might well also separate the sperm from the egg cortical microfilaments, should these fibers be involved in the lateral displacement of the sperm during the later stages of incorporation. Another motile comTonent that must be considered in the light of the movements during sperm incorporation is the incorporated sperm axoneme, which resumes its motility following the initial extension of the fertilization cone. The sperm tail is always incorporated and routinely beats in the egg cytoplasm. This erratic and arrhythmic beating might well be sufficient to push the sperm off the egg cortex into the egg cytoplasm proper. 2. The Centrad Migration of the Sperm Aster The migration of the sperm aster to the egg center appears to be the direct result of the assembly of microtubules on the sperm centrioles. The extension of these assembling fibers effectively pushes the sperm aster with the intercalated pronuclei away from the inner egg surface. Of course once the sperm aster is equidistantly separated from the inner faces of the egg's surface it arrives at the egg center. The effect, if any, of microtubule treadmilling during this centrad migration is difficult to assess. It is quite clear that in the absence of assembly or in the absence of disassembly the sperm aster will not arrive properly at the egg center.

3 . The Migration of the Egg Nucleus The swiftest and most dramatic of the movements during fertilization is the migration of the female pronucleus into the center of the sperm aster. While it is clear that assembling and disassembling microtubules are required for this motion to occur, the molecular mechanism involved in this system of motility is not yet clear. There are, however, two attractive hypotheses to account for this motion: dynein ratcheting and dynamic anchoraging-treadmilling. The dynein ratcheting theory rests on the experimental evidence involving the role of dynein as an ATPase during ciliary motility. As dynein arms are capable of ratcheting along microtubules to effect axonemal bending, in this model dynein arms are hypothesized to ratchet along the sperm astral microtubules. For this model to account for the female pronuclear migration, however, the dynein would have to be attached to the female nuclear surface. Interestingly, the dynein inhibitor EHNA specifically blocks this motion (Schatten et al., 1982~).Now when the sperm astral microtubules extend to the length that permits them to contact the egg nuclear surface, the dynein on the egg nuclear surface begins to ratchet along these microtubules dragging the female pronucleus to the center of the sperm aster. It is noteworthy that dynein has been found in sea urchin egg cytoplasm though its function has not yet been formulated.

MOTILITY DURING FERTILIZATION

151

The dynamic anchorage-treadmilling hypothesis requres that the microtubules of the sperm aster be treadmilling. The polarity for treadmilling would be such that monomeric tubulin is being added at the aster periphery and is being withdrawn at the centriolar region. During the initial stages of fertilization the net rate of assembly far outweighs the net rate of disassembly, providing the appearance simply of sperm astral growth, even though monomers are fluxing through each individual microtubule. When the female pronucleus is contacted by these treadmilling microtubules the nuclear surface is predicated to contain an anchorage system that can attach to specific monomers. In this model the attachment by the nucleus to specific tubulin monomers would pull the female pronucleus to the sperm aster as the monomers treadmill through the assembling microtubules. The corresponding attachment of the male pronuclear envelope to the sperm aster would be the force retaining the male pronucleus at the aster center during its extension phase. A rather attractive variation on this same model is the notion of the female pronuclear surface as a microtubule cap. If, as proposed in this model, the net assembly end of astral microtubules is at the astral periphery and the net disassembly end is at the aster center, this theory predicts that the migration of the female pronucleus is the direct result of the attachment of the growing sperm aster microtubules to the female pronuclear surface. However, in contrast to the preceding postulate, in this case the direct attachment of the astral microtubules to the envelope would effectively cap the growing end of the microtubules and prevent any further elongation. Since the net rate of disassembly is unaffected by this capping, the microtubules would appear only to shorten since all assembly is not precluded. This model is particularly attractive since it is found by antitubulin immunofluorescence microscopy that following the migration of the female pronucleus the sperm aster develops an asymmetric appearance, i.e., the microtubules that would have interconnected the pronuclei have disassembled; this would result in just such a configuration. 4. Implicarionsfor Other Systems

The formation and functioning of the cytoskeleton during fertilization might well prove a most instructive model for understanding motility in general. In this system, the cytoskeleton forms de novo, and, furthermore, both spatially and temporally separated, unidirectionally oriented microfilament- and microtubulemediated events occur. Analogies between the cortical microfilaments active during cytokinesis and the cortical microfilaments active during sperm incorporation could be drawn and indeed analogies between the microtubules of the sperm aster and those participating during mitosis can easily be understood. This latter analogy is particularly strong if mitosis is viewed as occurring in two half spindles. Instead of a set of chromosomes moving toward the astral poles, during fertilization the female pronucleus is drawn to the center of the model monopolar

152

GERALD SCHAITEN

sperm aster. In both the half spindle at mitosis and the sperm aster during fertilization centrioles serve as the nucleating centers; microtubules emanating from these centers might well attach to either kinetochores or the female pronuclear surface. The mode of attachment is not understood at all and might well be at the heart of the riddle regarding the mechanisms for the chromosome and pronuclear movements. In both cases the chromosomes or the female pronucleus migrate to the center of the aster, perhaps concomitant with the selective shortening of nuclear chromosome to pole fibers.

VIII. Prospectives and Conclusions A. PROSPECTIVES

In this section a number of problems, with only a few answers, are posed. The effort during the upcoming years will likely provide many of the solutions. 1. Totipolarity

The result of fertilization is quite clear, i.e., a zygote competent to begin cleavages. The manner in which this result is achieved is not as transparent. A problem of polarity and recognition exists since the sperm is able to enter anywhere on the egg surface but the fertilization cone will only form at the site adjacent to a successful sperm. Once the sperm is incorporated another problem of polarity appears, namely, the manner in which the sperm, which entered in a random fashion on the egg surface, can locate the eccentrically located female pronucleus. The answers to these problems probably rely on the predicted transient ionic controls for the formation of the fertilization cone and in the radial array of the sperm aster. Clearly a signal is emitted at the site of sperm-egg fusion, perhaps a transient change in intracellular [Ca2+] and pH, which results in the localized assembly of microfilaments resulting in the fertilization cone. In the case of polyspermy several cones will form, each at a fusion site. It is certainly reasonable to imagine that the entire egg surface is competent to form fertilization cones and that the sperm brings with it the catalyst for assembly. In the case of locating the eccentric female pronucleus, nature endowed the spherical sperm aster with an equipotent radial array of microtubules. If the migration of the female pronucleus involves dynein linkages, then whichever microtubules the female pronuclear-associated dynein arms contact would be competent to serve as the guiding tracks leading the female pronucleus to the male pronucleus. If, on the other hand, these microtubules are treadmilling and the female pronucleus caps the assembling end, permitting the resulting disassembly of the interconnecting microtubules, then all of these microtuhules

MOTILITY DURING FERTILIZATION

153

would be undergoing treadmilling and any of them would be candidates for capping and disassembly. This sort of an equipotent sperm aster negates all considerations regarding differences resulting from the sperm’s entering the egg surface adjacent to the female pronucleus or at a maximal distance or angle away from the female pronucleus. 2. Recognition and Regulation The examples of recognition and regulation in this system number close to a dozen. The sperm recognizes that the egg is near and undergoes the acrosome reaction. The sperm surface components recognize and first selectively adhere to the egg vitelline layer and later penetrate through this extracellular coat and preferentially attach to the egg plasma membrane. Following the adhesion between the plasma membranes, gamete fusion occurs. Another example of recognition occurs when the fertilization cone forms only at the fusion site though presumably the egg is competent to form the cone at any site. The orientation of the sperm must be recognized since it invariably rotates so that its centriolar end faces toward the egg cytoplasm and the lateral displacement of the sperm and the movements during sperm incorporation invariably result in the pulling of the sperm to the cytoplasm rather than the pushing of it to the exterior. The manner in which the sperm centrioles inform the egg that they are the prime and dominant MTOCs and that microtubule assembly should only occur on them, especially in light of the numerous other potential MTOCs existent within the unfertilized egg, is a topic for further inquiry. Whether the sperm astral microtubules are directly and specifically associated with the egg cortex or merely push off the inner face of the egg surface needs also to be elucidated. The manner in which sperm astral microtubules contact the female pronucleus, presumably attaching to it rather than to any of the other intracellular organelles, is mysterious. The procedures leading to the separation and later replication of the sperm centrioles are also unknown. Finally the ionic and molecular controls of this complex repertoire need to be carefully evaluated. While pH and [CaZ+J play dominant roles the existence of finer, subtle controls and indeed localized cytoplasmic gradients must be understood. This point underscores the relations between recognition and regulation since it is possible to have specific adhesion, e.g., between the sperm and egg plasma membrane, or alternatively intermediary regulatory systems, perhaps attaching the egg nuclear envelope to sperm astral microtubules and simultaneously eliciting a localized disassembly of astral microtubules could also result in a coordinated behavior. 3 . Polarity, Orientation, and Nucleation Centers

Questions concerning polarity and orientation of the egg cytoskeleton must be posed when we attempt to understand movements during artificial activation, especially in comparison with those occurring during normal fertilization. In the

154

GERALD SCHATTEN

case of artificial activation a single monaster forms, which slowly coalesces toward the egg center. In the case of normal fertilization the sperm aster i,'rows from the sperm centrioles eventually filling the egg cytoplasm with radially arrayed fibers. In both cases the final positioning of the pronuclei is at the egg center, but the manner in which they are translocated differs radically. In the case of normal fertilization, just focusing on the female pronucleus in relation lo the sperm aster, it is pulled to the center of the sperm aster. Later the sperm aster moves both to the egg center. During artificial activation the female pronucleus is pushed to the egg center. It seems reasonable to predict that in both cases microtubules have their net assembly end at the periphery and their net disassembly end at the egg center; this assumption is based on the centrad saltation of particles and the manner in which taxol-induced microtubules form at the periphery and migrate to the egg center. If this assumption is valid, then the sole difference in the two motions might well be the orientation of the egg nucleus in relation to the forming microtubules. During fertilization the aster is external to the female pronucleus, whereas during artificial activation the female pronucleus is within the developing monaster. If treadmilling is occurring and if the egg nuclear surface can cap the growing, but not the disassembling, end of these treadmilling microtubules, then during fertilization the attachment of the assembling end of the astral microtubules to the egg nuclear surface would permit it to be pulled to the egg center. In contrast during artificial activation the only possible association is between the disassembling end of the microtubule on the egg nuclear surface, which results in the pushing and effective repulsion of the egg nucleus from the disassembling end of these microtubules. The problem of potential nucleation centers in eggs is perplexing indeed. During fertilization the sperm centrioles, themselves elusive organelles, are the dominant MTOCs. The manner in which they convey their presence must be explored. During artificial activation a monastral microtubule-containing structure will form. However, if activation is followed by treatment with heavy water or hypertonic solutions, the centrioles will appear de novo and take on the appearance of geodesic domes. Taxol, even in unfertilized eggs, initiates a completely different pattern of microtubule assembly, raising further questions about the potential configuration of the egg's cytoskeleton. 4 . Anchorage and Capping It is clear that biological structures are transiently anchored to the cytoskeleton in this system. The fertilization cone anchors the successful sperm to the egg surface so that the elevation of the fertilization coat does not detach it from the egg. Also the egg nucleus is somehow anchored to the sperm aster and then drawn to the aster center. The manner in which a cytoskeletal element finds and then later anchors to a structure or organelle and the effects that this anchorage

MOTILITY DURING FERTILIZATION

155

might have, especially if the fiber is treadmilling, are liable to provide answers for a number of questions regarding intracellular motility.

5. Functions and Fates: Shape and Motion A number of perplexing structures appear and later disappear during fertilization and their role in development and their fate must be understood. Perhaps the first problem is the disappearance of the centrioles within the oocyte during oogenesis and, once posed, this question leads to others regarding the de novo induction of MTOCs and true centrioles. The fates of the sperm acrosomal process and the sperm axoneme have yet to be fully explored as do the functions of the “streak” or interim apparatus and the role of the microvillar burst in development. The temporal correlation between assembly of the streak and the second burst of microvillar elongation is intriguing. Questions regarding the active maintenance of a spherical shape in relation to the structural role of the egg cortex or the importance of the churning behavior elicited by the elongated microvilli must similarly be posed.

B . CONCLUSIONS This article attempts to highlight the importance of the cytoskeletal elements in both gametes during fertilization. It appears quite clear that the physical incorporation of the sperm and the restructuring of the fertilized egg cortex are mediated by microfilament assembly and that the microtubules that form the sperm aster are essential for the nuclear movements. The later cycles in microfilaments (microvillar burst, cytokinesis) and microtubules (formation of the streak mitosis) have also been reviewed. The importance of the program of activation, especially the change in intracellular pH and [Ca2 1 following sperm-egg fusion, have key roles in the formation of the egg’s cytoskeleton. +

C. SUMMARY Eggs at fertilization appear ripe for a multitude of questions regarding intracellular motility. The de n o w formation of a cytoskeleton, including questions concerning its polarity, nucleation, orientation, attachment sites, and regulation, are as yet largely unexplored. The dynamics of this system involve cycles of cytoskeletal assembly and disassembly, simultaneous pushing and pulling of structures mediated by both microfilaments and microtubules, and a complex repertoire of ionic and molecular regulators. In summary, multiple avenues for future experiments are available, which are predicted to be of great importance in understanding motility during fertilization specifically and cellular motility in general.

156

GERALD SCHATTEN ACKNOWLEIKMENTS

It is my pleasure to acknowledge formally the collaborators who have participated in various aspects of these studies, including Professors Daniel Mazia, Daniel Friend. Dieter Hiilser, David Nishioka, Neidhard Paweletz, Richard Steinhardt, and Win Sale, Drs. Tim Bestor, Paris Kidd, and Susan Mann. Ms. Christi Cline, Messrs. Ron Balczon, Manuel Daniels, John Henson, Juan Ludert. and Calvin Simerly, and especially Dr. Heide Schatten. The support of the author’s research reviewed in this article by the National Institutes of Health (research grant HD 12913; research career development award HD00363; electron microscope instrumentation grant RR1466) is gratefully acknowledged.

REFERENCES Afzelius, B. A. (1976). Science 193, 317-319. Aketa. K., and Ohta. J . (1977). Dev. Biol. 61, 366-372. Allen. R. D. (1958). I n ”The Chemical Basis of Development” (W. 0. McElroy and 9. Glass. eds.), pp. 17-73. John Hopkins Univ. Press, Baltimore, Maryland. Allen, R. D., Travis, J. L., Allen, N. S.. and Ylimaz, H. (1981a). CeIIMoril. 1, 275-289. Allen, R. D., Allen, N. S., and Travis, J. L. (1981b). Cell Moril. 1, 291-302. Amy. C. M., and Rebhun, L. I. (1977). Exp. Cell Res. 104, 399-409. Anderson, E. (1968).J . CellBiol. 37, 514-539. Anderson, E. (1974). Inr. Rev. Cvrol. Suppl. 4, 1-70. Anderson, E., Hoppe, P. C., Whitten, W. K., and Lee, 0. S. (1975). J . Ulrrasrrucr. Res. 50, 23 1-252. Aronson, 1. F. (1973). J . Cell Biol. 58, 126- 134. Aronson, J. F.. and Inoue. S. (1970). J . Cell Biol. 45, 470. Austin, C. R. (1968). “Ultrastructure of Fertilization.” Holt, New York. Baker, P. F., and Whitaker, M. J. (1979). J. Phvsiol. (London) 79, 5 5 . Banzhaf, W. C., Warren. R. H., and McClay. D. R. (1980). Dev. Biol. 80, 506-515. Bedford, J. M. (1970). I n “Mammalian Reproduction” (H. Gibbian and E. J . Plotz, ecls.). pp. 124- 182. Springer-Verlag, Berlin and New York. Begg, D. A., and Rebhun, L. 1. (1979). J . Cell Biol. 83, 241-248. Begg, D. A., Rodewald, R., and Rebhun, L. 1. (1978). J . Cell Biol. 79, 846-852. Begg, D. A , , Rebhun, L. I . , and Hyatt, H. (1982). J . Cell Biol. 93, 24-32. Berns, M. W., Rattner. 1. B., Brenner, S., and Meredith, S. (1977). J . Cell B i d . 72, 35-367. Bestor, T. H. (1981 ). Ph.D. thesis. Florida State University. Tallahassee. Bestor, T . H., and Schatten, G . (1981). Dev.Biol. 88, 80-91. Bestor, T . H., and Schatten, G. (1982). Exp. Cell Res.. in press. Borghese. E., and Cassini, A. (1963). In “Cinemicrography in Cell Biology” (G. G . Rose, ed.), pp. 201-227. Academic Press, New York. Borisy, G. G . , and Olmsted, J. 9. (1972). Science 177, 1196-1 197. Boveri, Th. (1901). Zoool. Jahrb. A h . Anar. Onrog. 14, 630-653. Brachet, J. (1974). In “Introduction to Molecular Embryology.” Heidelberg Science Library. Brackett. B. 0. (1970). Fertil. Steril. 21, 169-176. Brackett, 9. G., Oh, Y. K., Evans, J. G., and Donawick, W. J. (1980). Biol.Reprod. 23, 189-205. Brandriff, B., Hinegardner, R. T., and Steinhardt, R. S. (1975). J. Exp. Zoo/. 192, 13-23. Brenner, S. L., and Korn. E. D. (1980). J . Biol. Chem. 255, 841. Brinkley, 9. R., Fuller, G. M., and Highfield, D. P. (9175). Proc. Narl. Arad. Sci. U.S.A. 72, 498 1-4985. Brinkley. B., Fistel, R.. Marcum. I . , Pardue, R. (1980). Inr. Rev. Cvrol. 63, 59-95.

MOTILITY DURING FERTILIZATION

157

Browne, C. L.. Lockwood, A. H., Su, J.-L., Beavo, J. A,, and Steiner, A. L. (1980).J. CcllBiol. 87, 336-345. Bryan. J. (1974). BioScience 24, 701-71 I . Bryan. J.. and Kane. R. E. (1978). J. Mol. B i d . 125, 207-244. Bucher. N . L. R., and Mazia, D. ( 1960). J . Biophys. Biochem. Cviol. 7, 65 I . Burgess. D. R . . and Schroeder, T. E. (1977). J. Cell Biol. 74, 1032- 1037. Byrd. W.. and Perry G . (1980). Exp. Cell Res. 126, 333-342. Campanella, C. (1975). Biol. Reprod. 12, 439-447. Cantino. M. (1981). I n “Microprobe Analysis of Biological Systems” (T. E. Hutchison and A. P. Somlyo. eds.). pp. 65-82. Academic Press, New York. Cardasis, C. A,. Schuel, H., and Herman. L. (1978). J. CellSci. 31, 101-115. Carron. C. P., and Longo. F. J . (19x0). Dev. B i d . 79, 478-487. Carron. C. P., and Longo. F. J. (1982). Dev. Biol. 89, 128-137. Casillas. E. R.. Elder. C. M.. and Hoskins. D. D. (1980). J. Reprod. Fertil. 59, 297-302. Castaieda. M.. and Tyler, A. (1969). Biochcm. Bioph,u. Res. Conimun. 33, 782-786. Chambers. E. L. (1939). J. E.rp. Biol. 16, 409-424. Chambers. E. L.. and de Armendi. J. (1979). E.rp. Cell Res. 122, 203-218. Chambers. E. L.. Pressman, B. C.. and Rose. B. (1974). Biochem. Biophw. Res. Commrrn. 60, 126- 139. Chambers. R. (1933). J. E.rp. Biol. 10, 130-141. Chandler. D. E.. and Heuser. J . (1979). J. Cell Biol. 83, 91-108. Chandler, D. E.. and Heuser, J. (1980). J. Cell Biol. 84, 618-632. Chandler, D. E.. and Heuser, J. (1981). Dev. Biol. 82, 393-400. Cheung. W . Y. (1980). Science 207, 19-27. Cline, C., and Schatten. G. (1983). In preparation. Czihak. G. (1973). Exp. Cell Res. 83, 424-426. Dale, B., and Monroy. A . (1981). Gamete Res. 4, 151-169. Dale. B . . and de Santis. A. ( I98 I ). Dev. B i d . 83, 232-237. Dan, J . C. ( 1950). Biol. Bull. 99, 399-41 I . Dan, J. C. (1954). Biol. Bull. 107, 335-349. DeBrabander. M. J . , Van de Veire, R. M. L.. Aerts. F. E. M.. Borgers. M.. and Janssen. P. A . J. (1976). Cunwr Res. 36, 905. DeBrabander. M.. Geuens, G.. Nuydens, R.. Willebrords. R.. and DeMey. J . (1981). Proc. Nail. Acad. Sci. U.S.A. 78, 5608-5612. deFelice. L. J.. and Dale. B. (1979). Dev. B i d . 72, 327-341. DeRosier. D., Mandelkow. E.. Silliman. A,. Tilney. L., and Kane. R. E. (1977).J. Mol. B i ~ l 113, . 679-695. Detering. N. K.. Decker. G . L.. Schmell. E. D.. and Lennarz. W. J . (1977). J . Cell B i d . 75, 899-9 14. Dirksen. E. R . (1961). J. Biophw. Biochem. Cytol. 11, 244-247. Dustin, P. (1978). “Microtubules.” Berlin. Edds. K. T. (1977). Exp. Cell Res. 108, 452-456. Eddy. E. M.. and Shapiro. B. M. (1976).J. Cell Biol. 71, 35-48. Egrie. J . C., and Nagle. B. W. (1980). A m . N.Y. Acad. Sci. 356, 376-377. Elinson. R . P. (1980). I N “The Cell Surface” ( S . Subtelney and N . K. Wessells. eds.). pp. 169-183. Academic Press. New York. Elinson. R. P.. and Manes. M. E. (1978). Dev. Biol. 63, 67-75. Epel. D. (1977). Sci. Am. 237, 128-138. Epel, D. (1978). In “Current Topics in Developmental Biology” ( A . A . Moscona and A. Monroy. eds.). Vol. 12. pp. 186-246. Academic Press. New York. Epel. D. (1980). Anti. N. Y . Acad. Sci. 339, 74-85,

I58

GERALD SCHATTEN

Epel. D.. Weaver, A. M., and Mazia, D. (1970). Exp. Cell Res. 61, 64-68. Epel. D..Steinhardt, R. Humphreys, T..and Mazia, D. (1974). Dev. Biol. 40, 245-255. Epel, D.. Cross, N. L., and Epel, N. (1977). Dev. Growth Difler. 19, 15-21. Epel, D., Patton. C., Wallace, R. W.. and Cheung, W. Y. (1981). Cell 23, 543-549. Fischer, G. W., and Rebhun, L. I . (1981). J. Cell B i d . 91, 185a. Flanagan. M. D., and Lin, S. (1980). J. Biol. Chem. 255, 835. Foerder, C. A,, and Shapiro, B. M. (1977). Proc. Natl. Acad. Sci. U.S.A. 74, 4214-421s. Fol, H. (1877a). C. R. Hebd. Seances Acad. Sci. 85, 233-236. Fol, H. (1877b). Arch. Zoo/. Exp. Gen. 6, 180-192. Forer, A,, and Zimmerman, A. (1974). J . Cell Sci. 16, 481-497. Forer, A,, and Zimmerman, A. (1982). “Mitosis and Cytokinesis.” Academic Press, New York. Franklin. L. (1965). J. Cell B i d . 25, 81-100. Friend. D. S. (1982). J . Cell Eiol. 93, 243-250. Fuller, G. M., Brinkley, B. R., and Boughter, J. M . (1975). Science 187, 948-950. Gabel. C. A,. Eddy, E. M., and Shapiro, B. M. (1979a). Cell 18, 207-215. Gabel. C., Eddy, E., and Shapiro, B. (l979b). J. Cell Eiol. 82, 742-754. Garbers, D. L., and Hardman, J. G. (1975). Nature (London) 257, 677-678. Garbers, D. L., and Kopf, G. S. (1980). Adv. Cyclic Nucleotide Res. 13, 251-306. Garbers, D. L., Lust, W. D., First, N. L., and Lardy, H. A. ( I97 I ) . Biochemistry 10, 1825- I83 I . Garbers, D. L.. First, N. L.. Gorman, G. K., and Lardy, H. A. (1973a). Eiol. Reprod. 8,599-606. Garbers, D. L., First, N. L., and Lardy, H. A. (1973b). Biol. Reprod. 8, 589-598. Garbers. D. 0.. Hansbrough, J . R., Radany, E. W., Hyne, R. V.. and Kopf, G. S . (1980). J. Reprod. Fertil. 59, 377-381. Gibbons, B. H., and Gibbons, I. R. (1979). J . Cell Biol. 83, 806a. Gibbons, I. R. (1977). In “International Cell Biology” (B. R. Brinkley and K. R. Porter, eds.). p. 348. New York. Gibbons, 1. R. (1981). J. CellEiol. 91, 1075-1245. Gilkey, 1. C., Jaffe, L. F., Ridgway, E. B., and Reynolds, G. T. (1978). J . CellBiol. 76,448-466. Glabe, C. G., and V. D. Vacquier (1977). J. Cell B i d . 75, 410-421. Could-Somero, M., Holland, L., and Paul, M. (1977). Dev. B i d . 58, 11-22, Grainger, 1. L., Winkler, M. M., Shen, S. S., and Steinhardt, R. A. (1979). Dev. Biol. 68, 396-406. Gwatkin. R. B. L. (1977). “Fertilization Mechanisms in Man and Animals.’’ Plenum, New York. Hagiwara, S . , and Jaffe, L. (1979). Annu. Rev. Biophys. Bioeng. 8, 385-416. Haimo, L. T., and Rosenbaum, J. L. (1981). J. CellEiol. 91, 1255-1305. Hamaguchi. M. S., and Hiramoto, Y. (1980). Dev.. Growth Differ. 22, 517-530. Hamaguchi, M. S., and Iwasu, F. (1980). Biomed. Res. 1, 502-509. Hara, K., Tydeman, P., and Kirschner, M. (1980). Proc. Natl. Acad. Sci. U . S . A . 77, 402-466. Harris, P. (1975). Exp. Cell Res. 94, 409-425. Harris. P. (1978). In “Cell Cycle Regulation” (I. R. Jeter, Jr., I. L. Cameron, G. M. Padilla, and A. M. Zimmerman, eds.), pp. 75-104. Academic Press, New York. Harris, P. (1979). Dev. Eiol. 68, 525-532. Harris. P.. Osborn, M., and Weber, K. (1980a). J . Cell Eiol. 84, 668-679. Harris, P., Osborn, M.. and Weber, K. (1980b). Exp. Cell Res. 126, 19-28. Harvey, E. B. (1956). “The American Arbacia and Other Sea Urchins.” Princeton Univ. Press., Princeton, New Jersey. Head, J . F., Mader, S., and Kaniiner, B. (1979). J. Cell Biol. 80, 21 1-218. Heidemann, S. R., and Gallas, R. T. (1980). Dev. Biol. 80, 489-494. Heidemann, S. R., Sander G., and Kirscher, M. W. (1977). Cell 10, 337-350. Hepler, P. K. (1980). J. Cell Biol. 86:490-499.

MOTILITY DURING FERTILIZATION

159

Herman. B.. and Albertini. D. F. ( I98 I ). J. Cell B i d . 91, 338a. Hinegardner. R. T.. Rao. B., and Feldman. D. E. (1964). Exp. Cell Res. 36, 53-61. Hiramoto. Y. (1970). Biorhrologv 6, 201-234. Hirdmoto. Y . (1974). Exp. Cell Res. 89, 320-326. Hirdmoto, Y.. Shimoda. S.. and ShGji, Y. (1979). Drv. Growth Diaer. 21, 141-153. Horstadius, S. (1928). Acta Zool. 9, 1-191. Horstadius. S. (1939). Biol. Rev. 14, 132-179. Hoskins. D. D.. and Casillas, E. R. (1974). f f i “Advances in Sex Hormone Research” (R.L. Singhal and J. A. Thomas, eds.), pp. 283-324. Univ. Park Press, Baltimore. Maryland. Hoskins. D. D.. Stephens. D. T.. and Hall. M. L. (1974). J. Reprod. Frrtil. 37, 131-133. Hoskins. D. D.. Hall. M. L.. and Munsterman. D. (1975). Biol. Reprod. 12, 566-572. Hiilser, D.. and Schatten, G. (1980). Eur. J. Cell B i d . 22, 253a. Hiilser, D.. and Schatten, G. (1982). Gamete Rrs. 5 , 363-377. Inoui. S. (1976). Cold Spring Harbor Svmp. Quani. Biol. 3, 1317-1328. Inoue. S. (198la). 1.Cell B i d . 89, 346-356. h u e . S. (1981b). J. CellBiol. 91, 1315-1475. Inoue. S.. and Sato. H. ( 1967). J. Gen. Phvsiol. 50, 259-285. Jaffe. L. A. (1976). Nature (London) 261, 68-71. Johnson, C. H . , and Epel, D. (1981). J. CellBiol. 89, 284-291. Johnson. J.. Epel, D., and Paul, M. (1976). Nature (London) 262, 661-664. Jones. H. P.. Bradford, M . M.. McRorie, R. A , , and Cormier. M. J . (1978). Biuchem. Biuphy. Res. Commun. 82, 1264- 1272. Jones. H. P.. Lenz, R. W., Palevitz, B. A., and Cormier M. J . (1980). Proc. Natl. Acad. Sci. U.S.A. 77, 2772-2776. Kane, R. E. (1962). J. Cell Biol. 12, 47-56. Kane. R. E. (1975). J. Cell Biol. 66, 305-315. Kane. R. E. (1976). J. CellBiol. 71, 704-714. Kane, R. E. (1980). J. Cell E d . 86, 803-809. Katz, D. F. (1981). Res. Reprod. 13, 2 . Keller. C.. Gundersen. G.. and Shapiro, B. M. (1980). Dett. Biol. 74, 86-100. Kidd, P. (1978). J. Ultrastrrrct. Res. 64, 204-215. Kidd, P.. Schatten, G., Grdinger, J.. and Mazia, D. (1976). Biophys. J . 16, I17a. Kiehan. D. E. (1981). J. CellBiul. 88,604-617. Kirschner. M. W. (1980). J. Cell B i d . 86, 330-334. Kobayashi. Y..Ogawa, K.. and Mohri, H. (1978). Exp. Cell Res. 114, 285-292. Kojima. M. K . (1960). Ernbqologia 5, 1-7. Kuhl, W.. and Friedrich-Freksa, H. (1936). Verh. Dtsch. Zool. Ges. Leipzig. Suppl. 9, 187-195. Kuhl. W.. and Kuhl.. G. (1949). Zoo/. Jahrb. Abt. Anat. 70, 1-59. Lazarides. E. (1980). Nature (London) 283, 249-256. Lazarides, E.. and Weber, K. (1974). Proc. Nut/. Acad. Sci. U.S.A. 71, 2268-2272. Lee. S. C.. and Steinhardt. R. A . (1981). Dev. Biol. 85, 358-369. Levine. A. E.. Walsh. K. A,, and Fodor, E. J. B. (1978). Dev. Biol. 63, 299-306. Lewis, W. H., and Gregory. P. W. (1929). Science 69, 226-229. Lindahl. E.. and Oerstrbm, A. (1933). Protoplasma 17, 25-31. Loeb, J . ( 1913). In “Artificial Parthenogenesisand Fertilization.” Univ. of Chicago Press, Chicago, Illinois. Longo, F. J. (1973). Biol. Reprod. 9, 149-215. Longo, F. J. (1976). J. Cell B i d . 69, 539-547. Longo, F. (1978a). Dev. B i d . 62, 271-291. Longo, F. (1978b). Dev. Biol. 67, 249-265.

160

GERALD SCHATTEN

Longo, F. (1980). Dev. Biol. 74, 422-431. Longo, F., and Anderson, E. (1968). J . CellBiol. 39, 339-368. Longo. F. J., and Anderson, E. (1969). J. Ulrrastruct. Res. 29, 86-1 18. Longo. F. J., and Plunkett, W. (1973). Dev. B i d . 30, 56-67. Lopo. A., and Vacquier, V. D.(1977). Nature (London) 269, 590-592. Luduena, R. F.. Horowitz. P. M.. and Roach, M. C. (1979). J . Cell Biol. 83, 340a. Lyon. E. P. (1907). Arch. Enrw. Org. 23, 151-173. Mabuchi, I. (1973). J. Cell B i d . 59, 542. Mabuchi, I. (1974). J. Biochem. 76, 47-55. Mabuchi, I . , and Okuno, M. (1977). J . Cell Biol. 74, 251. Mabuchi. I . , and Spudich, J. A. (1980). J . Biochem. 87, 785-802. Manfredi, J. J.. Parness, J., and Horwitz, S . B. (1981). J . Cell B i d . 91, 330a. Mann, S.. Schatten. G., Steinhardt, R., and Friend, D. S. (1976). J. Cell Biol. 70, 1 IOa. Mar, H. (1980). Dev. Biol. 78, 1-14. Marcum. J . M.. Dedman. J., Brinkley, B. R., and Means, A. ( 1978). Proc. Natl. Acud. Sci. U.S.A. 75, 377 1-3775. Margolis, R. L., and Wilson, L. (1981). Nature (London) 293, 705-71 I . Margolis. R. L., Wilson, L., and Kiefer, B. (1978). Nature (London) 272, 450-452. Mazia, D. (1937). J. Cell. Comp. Phvsiol. 10, 291-304. Mazia. D. (1961). “The Cell” (J. Brachet and A. E. Mirsky, eds.), Vol. 3. p. 77. New York. Mazia, D. (1974). Proc. Natl. Acad. Sci. U.S.A. 71, 690-693. Mazia. D. (1977). I n “Mitosis: Facts and Questions” (M.Little, N. Paweletz, C. Petzelt, H. Ponstingl, D. Schroeter, and H.-P. Zimmerrnan, eds.), pp. 196-213. Springer-Verlag. Berlin and New York. Mazia, D.. and Dan, K . (1952). Proc. Natl. Acad. Sci. U.S.A. 38, 826-838. Mazia. D.. and Ruby, A. (1974). Exp. Cell Res. 85, 167-172. Mazia, D., Schatten, G . , andsale, W. (1975a). J. Cell Biol. 66, 198-200. Mazia, D., Schatten, G.. and Steinhardt, R. (1975bj. Proc. Nut/. Acad. Sci. U S A . 72,4469-4473. Mazia, D.. Paweletz. N.. Sluder, G., and Finze, E. H. (1981). Proc. Natl. Acad. Sci. U.S.A. 78, 377-38 I . Mazurovsky, E. B., Peterson, E. R., Crain, S . M., and Horwitz, S. B. (1981). Brain Rcs. 217, 392-398. Means, A. R., and Dedman, J . R. (1980). Nature (London) 285, 73-77. Miki-Noumura, T. (1977). J. Cell Sci. 24, 203-216. Mitchison, J . M. (1956). Q.J. Microsc. Sci. 97, 109-121. Mitchison, J . M. (1971). “The Biology of the Cell Cycle.” Cambridge Univ. Press, London and New York. Mitchison. J. M., and Swann. M. M. (1955). J . Exp. Biol. 32, 734-750. Mohri, H.. Mohri, T., Mabuchi. I., Yazaki, I . , Sakai, A,, and Ogawa, K. (1976). Dev. Growth Difer. 18, 391-398. Moore, A. R. (1937). Protoplasma 27, 544-551. Morgan, T. H., and Lyon, E. P. (1907). Arch. Enrw. Org. 24, 147-159. Morgan, T. H., and Spooner. G . B. (1909). Arch. Enfw. Org. 28, 104-1 17. Motomura. 1. (1936). T6hoku Imp. Utiiv. Sci. Rep. 4, 212-245. Mulnard. J. G.(1967). Arch. Biol. (LiPge) 78, 107-138. Nakamura, M., and Yasumasu, 1. (1974). J . Gen. Physiol. 63, 374-388. Naruse, H., and Sakai, H. (1981). J. Biochem. 90, 581-587. Nath. I . . and Rebhun, L. I. (1973). Exp. Cell Res. 77, 319-322. Nath. J . , and Rebhun, L. E. (1974). Biochim. Biophvs. Acta 370, 498-509. Nishida, E., and Kumagai, H. (1980). J. Biochem. 87, 143-151. Nishioka. D. (1982). J. Wash. Acad. Sci. 72, 1-11.

MOTILITY DURING FERTILIZATION

161

Nuccitelli. R., Webb, D. J . , Lagier, S. T., and Matson, G. B. (1981). Proc. Null. Acud. Sci. U.S.A. 78, 442 1-4425. Otto. J. J.. Kane. R. E.. and Bryan. J. (1980). Cell Motil. I, 31-40. Paul. M.. and Epel, D. ( 1975). Exp. Cell Res. 94, 1-6. Paweletz. N . (1981). Cell Biol. f n r . Rep. 5, 323-336. Paweletz, N . . and Mazia, D. (1978). In "Cell Reproduction: In Honor of Danial Mazia" (E. R . Dirksen. D. M. Prescott. and C. F. Fox, eds.). pp. 495-503. Academic Press, New York. Paweletz, N., and Mazia. D. (1979). Eur. J. Cell B i d . 20, 37-44. Peterson. S . P., and Berns. M. W. (1978). J. Cell Sci. 34, 289-301. Petzelt. C. (1972). Exp. Cell Res. 70, 333-339. Picherdl, B. (1977). J. Ultrustrucl. Res. 60, 181-202. Poccia, D. L. (1982). J. Wush. Acud. Sci. 72, 24-33. Pratt. M. M. (1980). Dev. Biol. 74, 364-378. Pratt. M. M.. Otter, T.. and Salmon. E. D. (1980). J. Cell B i d . 86, 738-745. Raff. R . A,. Greenhouse. G., Gross. K. W.. and Gross, P. R. (1971). J. Cell B i d . 50, 516-527. Rcbhun. L. I. (1972). f n r . Rev. C w l . 32, 93. Rebhun, L. I.. Schnaitman, T. C., Wang, R., and Mclvor, W. (1979). J. Cell Biol. 83, 350a. Ridgway. E. G., Gilkey, J. C., and Jaffe, L. F. (1977). Proc. Nut/. Acud. Sci. U.S.A. 74,623-627. Ries, J. (1909). Arch. Mikrosk. Anur. XX, 1-31. Roobol. A , . Gull, K., and Pognos. C. 1. ( 1976). FEBS Lett. 67, 248-25 I . Rothschild, L. (1956). "Fertilization." Methuen, London. Sale, W. S., and Gibbons, I. R. (1979). J. Cell Biol. 82, 291-298. Salmon. E. D.. and Segall, R. R. (1980). J. Cell B i d . 86, 355-365. Sanger. J. W.. and Sanger, J. M. ((1975). J. E.rp. Z o o / . 193, 441-447. Sano. K . , and Mohri. H. (1976). Science 197, 1339-1340. S a w , K.. and Mohri. H. (1977). Dev. Growtth Difler. 19, 275-281. Sano. K.. Usui, N., Ueki, K.. Mohri, T . , and Mohri. H. (1980). Dev. Growth Dijjer. 22,531-541. Satir, P. (196X). J. Cell B i d . 39, 77-94. Satir, P. (1974). Sci. Am. 231, 44-52. Schackman, R . W.. Eddy, E. M., and Shapiro, B. M. (1978). D w . B i d . 65, 483-495. Schatten. G. ( 1975). Ph.D. thesis. University of California, Berkeley. Schatten, G . (1979). J. Cell B i d . 83, 198. Schatten. G. (1981a). J. Morphol. 167, 231-247. Schatten, G. (1981b). Dev. Biol. 86, 426-437. Schatten. G. (1982). f n "Cellular Biodynamics: Mitosis and Cytokinesis" (A. M. Zimmerman and A. Forer, eds.). pp. 59-82. Academic Press, New York. Schatten. G.. and Hemmer, M. (1979). J. Cell Biol. 83, 199a. Schatten. G.. and Hiilser. D. (1982). Submitted. Schatten. G., and Mazia. D. (1976a). Exp. Cell Res. 98, 325-337. Schatten. G., and Mazia, D. (1976b). J. Suprumol. Strucr. 5, 343-369. Schatten, G., and Schatten, H. (1979). Scan. Elecrron Microsc. 111, 299-305. Schatten. G.. and Schatten, H. (1980b). Eur. J. Cell B i d . 22, 314. Schatten, G., and Schatten, H. (1981). Exp. Cell Res. 135, 31 1-330. Schatten. G.. Bestor, T.. Cline, C.. and Schatten, H. (1981a). J. Cell Biol. 91, 184a. Schatten. G., Schatten, H . , Bestor. T., and Balczon. R. (1981b). J. Cell B i d . 91, 185a. Schatten, G.. Schatten, H . , Bestor, T., and Balczon, R. (1982a). J. Cell B i d . 94, in press. Schatten, G., Schatten, H . , and Simerly. C. (1982b). Cell Biol. f n t . Rep., in press. Schatten. G.. Balczon. R . , Cline. C.. and Schatten, H. ( 1 9 8 2 ~ )J. . Cell Biol., in press. Schatten, H. (1977). Doctoral dissertation, University of Heidelberg. Schatten, H., and Schatten, G. (1977). J. Cell Biol. 75, 284. Schatten. H.. and Schatten. G. (1980a). Dev. B i d . 78, 435-449.

162

GERALD SCHATTEN

Schatten, H.. Petzelt, C., Mazia, D., and Schatten, G. (1982). Eur. J. Cell Biol. 27, 7 4 4 7 . Schiff, P. B., and Horwitz, S. B. (1980). Proc. Natl. Acad. Sci. U.S.A. 77, 1561-1565. Schiff, P. B., Fant, J., and Honvitz, S. B. (1979). Nature (London) 277, 665-667. Schmidt, T., Patton, C., and Epel, D. (1982). Dev. B i d . 90, 284-290. Schroeder, T. E. (1978a). Dev. Biol. 64, 342-346. Schroeder, T. (1978b). In “Cytochalasins: Biochemical and Cell Biological Aspects” (S. W. Tannenbaurn, ed.), p. 91, North-Holland Publ., Amsterdam. Schroeder, T. E. (1979). Dev. Biol. 70, 306-326. Schroeder, T. E. (1980). Dev. Biol. 79, 428-443. Schuel. H. (1978). Gamete Res. 1, 299-382. Seifritz, W. (1926). Protoplasma 1, 1-14. Shalgi, R., and Phillips, D. M. (1980a). Biol. Reprod. 23, 433-444. Shalgi, R., and Phillips, D. (1980b). J. Ultrastrucr. Res. 71, 154-161. Shalgi, R., Phillips, D. M., and Kraicer. P. F. (1978). Gamete Res. I, 27-37. Shapiro, B. M . , and Eddy, E. M. (1980). Int. Rev. Cvtol. 66, 257-302. Shen. S. S., and Steinhardt, R. A. (1978). Nature (London) 272, 253-254. Silver, R. B., Cole, R . D., and Cande, W. Z. (1980a). Cell 19, 505-516. Silver, R. B., Cole, R. D., and Cande. W. Z. (1980b). Eur. J. Cell Biol. 22, 315a. Simone, L. D., Brenner. S. L.. Wible, L. J.. Turner. D. S . . and Brinkley, B. R. (1981). J. CellBiol. 91, 337a. Sisken. J . E., Awesu, J. E., and Forer, A. (1981). J . CellBiol. 91, 315a. Sloboda, R. D., Malawistd, S. E., Vanblari, G . , Creasy, W. A , , and Rosenbaum, J . L. (1976). J. Cell Biol. 70, 290a. Sluder, G. (1979). J. Cell Biol. 80, 674. Snyder, J. A., and Mclntosh, J. R. (1976). Annu. Rev. Biochem. 45, 699. Soifer, D. (1976). Ann. N . Y. Acad. Sci. 253, I . Spiegel, E., and Spiegel, M. (1977). Exp. Cell Res. 109, 462-465. Spudich, A,, and Spudich, J. A. (1979). J . Cell Biol. 82, 212-226. Spudich, J. A., and Amos, L. A. (1979). J. Mol. Eiol. 129, 319-331. Steinhardt, R. A,, and Alderton, J. M. (1981). J. CellBiol. 91, 180a. Steinhardt, R. A,, and Epel, D. (1974). Proc. Natl. Acad. Sci. U.S.A. 71, 1915-1919. Steinhardt, R. A,. and Winkler. M. A. (1979). In “The Molecular Basis of Immune Cell Function” (J. G. Kaplan, ed.). pp. 174-186. Elsevier, Amsterdam. Steinhardt, R. A,, Lundin, L.. and Mazia. D. (1971). Proc. Natl. Acad. Sci. U.S.A. 68,2426-2430. Steinhardt, R. A,, Zucker, R., and Schatten, G. (1977). Dev. Biol. 58, 185-196. Summers. K. E.. and Gibbons, I. R. (1971). Proc. Natl. Acad. Sci. U.S.A. 68, 3092- 3096. Suprenant, F.. Bendex, P., and Rebhun, L. (1981). J. Cell Biol. 87, 296. Taglietti, V. (1979). Exp. Cell Res. 120, 448-451. Tanenbnum, S. W. ( 1978). “Cytochalasins: Biochemical and Cell Biological Aspects.” Elsevier, Amsterdam. Tegner. M.. and Epel, D. (1976). J. Exp. Zool. 197, 31-58. Thomas, I.. Buchsbaum, R., Zimniak, A,, and Racker, E. (1979). Biochernistrv 18, 2210-2218. Thompson, W. C., Purich, D. L., and Wilson, L. (1981). J. Cell Biol. 91, 329a. Tilney, L. G. (1978). J. Cell Biol. 77, 551-564. Tilney. L. G., and Inoue. I. (1982). J. Cell Biol. 93, 820-827. Tilney, L. G., and Jaffe, L. A. (1980). J. Cell Biol. 87, 771-782. Tilney, L. G., and Kallenbach, N. (1979). J. Cell Biol. 81, 608-623. Tilney. L. G.,Hatano, S.. Ishikawa, H., and Mooseker, M. S. (1973). J. CellBiol. 59, 109-126. Tilney, L. G.,Kiehart, D. P.,Sardet, C., and Tilney. M. (1978). J. Cell Biol. 77, 536-550. Tucker, R. W.. Pardee, A. B., and Fu-iiwara, K. (1979). Cell 17, 527-535.

MOTILITY DURING FERTILIZATION

163

Tyler. A. (1965).Am. Nut. 159,309-334. Usui. N..Sano, K., and Mohri, H. (1980).Dev. Growrh LXjfer. 22,461-473. Vacquier. V. D. (1975).Dev. B i d . 43,62-74. Vacquier, V. D. (1981).Dev. Biol. 84, 1-26. Vacquier. V. D . , and Brandiff, B. (9175).Dev. B i d . 47, 12-31. Vacquier, V. D.. and Moy, G . W . (1977).Proc. Natl. Acad. Sci. U.S.A. 74,2456-2460. Vacquier. V. D . , and Moy, G . W . (1980).Dev. Biol. 77, 178-190. Vacquier, V. D., Epel, D.. and Douglas, L. (1972a).Narure (London) 237,34-36. Vacquier, V. D.. Tegner, M . J., and Epel. D. (1972b).Nature (London) 240, 352-353. Vallee, R . B. (1981).J. Cell B i d . 91,326a. Wang. Y-L., and Taylor, D. L. (1979).J. Cell Biol. 81,672-681. Warner, F. D., and Mitchell. D. R . (1980).Inr. Rev. Cyrol. 66, 1-44. Wassarman, P. M., and Fujiwara, K . (1978).J. CellSci. 29, 171-188. Webb, D. J., and Nuccitelli, R. (1981).J. Cell B i d . 91,562-567. Wegner. A. (9176).J. Mol. Biol. 108, 139. Wehland. J., Herzog. W . , and Weber, K. (1977a).J. Mol. Biol. 111, 329. Wehland, J.. Osborn, M., and Weber, K. (1977b).Proc. Narl. Acad. Sci. U.S.A. 74,5613. Wehland. J., Stockem, W . , and Weber, K. (1978).Exp. Cell Res. 115, 451. Weisenberg. R. C. (1972).Science 177, 1104-1106. Welsh, M. I.. Dedman. I. R . , Brinkley. B. R., and Means. A . R . (1978).Proc. Narl. Acad. Sci. U.S.A. 75, 1867-1871.

Went, H . A . (1977).Exp. Cell Res. 108,63-73. Wieland, Th. (1977).Natumissenchufren 64,303. Wilson, E. B. (1925).“The Cell in Development and Heredity.” New York. Wilson, E. B., and Learning. E. (1895).“An Atlas of Fertilization and Karyokinesis.” Macmillan. New York. Wilson, E. B., and Mathiws, A. P. (1895).J. Morphol. 10, 319-342. Wilson. L., and Bryan, J . (1974).Adv. Cell. Mol. B i d . 3, 21-72. Winkler. M. M.. and Grainger, I. L. (1978).Nature (London) 300, 489-504. Winkler. M. M.. and Steinhardt, R. A . (1981).Dev. B i d . 84, 432-439. Winkler, M . M., Steinhardt. R. A., Grainger, J. L.. and Miming, L. (1980).Nature (London) 287,

558-560. Wolf, D. P. (1981).I n “Fertilization and Embryonic Development in V i m ” (L. Mastroianni, Jr. and J. D. Biggers, eds.). pp. 185-200.Plenum, New York. Wolf, D. P.. and Armstrong. P. B. (1981).Gamete Res. I, 39-46. Wolniak. S. M.. Hepler, P. K., and Jackson, W . T. (1980).J . Cell B i d . 87,23-32. Yanagimachi, R. (1978).Current Top. Dev. B i d . 12,83-105. Yanagimachi, R. (I98I ) . I n “Fertilization and Embryonic Development in Virro” (L. Mastroianni, Jr. and J. D. Biggers, eds.). pp. 82-182.Plenum, New York. Yanagimachi. R . , and Noda. Y.D. (1970).Am. J. Anar. 128,429-462. Yanagimachi. R.. and Noda. Y. D. (1972).Experientia 28,69-72. Yasumasu. I., Fujiwara. A . , and Ishida. K. (1973).Biochem. Biophvs. Res. Commun. 54,628-632. Yoneda, M. (1973).Adv. Biophvs. 4, 153-190. Yoneda, M . , Ikeda. M.. and Washitani. S. (1978).Dev. Growrh Difler. 20,329-336. Zamboni, L. (1972).I n “Biology of Mammalian Fertilization and Implantation” (K.S. Moghissi and E. S. E. Hafez. eds.). Thomas, Springfield, Illinois. Zimmerman. A. M.. and Zimmerman, S. (1967).J. Cell Biol. 34,483-488. Zucker. R. S.. and Steinhardt. R. A. (1978).Biochim. Biophys. Acru 541,459-466. Zucker, R. S.. Steinhardt. R. A , , and Winkler.‘M. M. (1978).Dev. B i d . 65, 285-295.

This Page Intentionally Left Blank

INTtRNATlONAL REVIEW OF CYTOLOGY. VOL. 7Y

Functional Organization in the Nucleus RONALDHANCOCKA N D TENIBOULIKAS Swiss Insriture ,for Experimental Crmcer Re.serrrck Epa fingeslLcntsnnrte~Switzrrfrind

........................................ lope . . . . . . . . . . . . . . . . . . . . A. The Membrane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. The Pore Elements,. . , , . . , . , , . . . . . . . . , . . . . . . . , . . . . . . . . C. The Lamina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Functional Activities of the Envelope.. . . , . . . . . . . . . . . . . . . . 111. Chromatin . . . . . . , . . . . , . . . . . . . . . . . . . . , . . . . . A. The Nucleosome . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . B. The Nucleosomal .......... C. Higher Order Folding of the Nucleosomal Fib IV. Topological Organization of Interphase DNA and Chromatin . . . . . .

.

.

.

.

.

.

.

.

.

.

I

.

.

.

.

.

.

165 i66 167 169 169 172 174 175 178

178 180 180

B. Structural Elements in Chromatin. . . . . . . . . . . . . . . . . . . . . . . . C. Topological Organization of DNA . . . . . . . V . Transcription, Processing, and Packaging of RNA . . . . . . . . . . . . . . B. Extranucleolar Transcription . . . . C. The Nucleolus and Nucleolar Tra

......... .. ,...............

A. G , and S Phases: DNA Synthesis and Chromatin Assembly . . B. Gz Phase and Mitosis . , . . , . . , . . . . .

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

References . , . . . . . . . . , . . , . . . . . . . . . . . .

183 I 8X 188 190 191 194 197 198 198 203 206 206

I. Introduction In this article we attempt to integrate current understanding of the major nuclear functions (replication of DNA, transcription of RNA, transport of RNA and other macromolecules, formation of mitotic chromosomes, etc.) into the framework of the ultrastructural organization of the nucleus, to provide a topological view of nuclear activities. We have emphasized some areas where newer ultrastructural cytochemical methods, such as the localization of specific proteins and nucleic acid sequences by immunocytochemical and in situ hybridization methods, could be fruitfully applied. We have attempted to present syntheses derived from what appears to us to be the most soundly established recent evidence, rather than to comprehensively discuss unresolved or controversial I65 Copyrighi (0 IYX? hy Academic Pres. Inc. All nghh of rcpmducliun In any l o r n reserved. ISBN O-I?-36447Y-X

166

RONALD HANCOCK AND TEN1 BOULIKAS

questions; for this reason, and also because of the breadth of the subject, we have undoubtedly sometimes leaned toward overselection and oversimplification.

11. The Nuclear Envelope

The term “nuclear envelope” appears to us the most appropriate for the complex structure which bounds the interphase nucleus, constructed of at least three ultrastructurally distinct elements: the nuclear membrane sensu strich, the pores, and the underlying lamina (Barton et al., 1971; reviewed by Franke, 1974; Franke er a / . , 1981b; Harris, 1978) (Figs. 1, 2, 3, 6 , and 10). The: close ultrastructural integration of these three components of the envelope probably reflects important functional interactions in vivo, and it has proved experimen-

FIG. 1 . A section through the interphase nucleus of a rat liver cell fixed in glutaraldehyde, embedded in Epon, and stained by the EDTA technique which bleaches chromatin (Bernhard, 1969). Structures discussed in the text include condensed chromatin (C), the nucleolus (N), perichromatin fibrils (arrowheads). perichromatin granules (small arrows), clusters of interchromatin granules (large arrows). Bar = I pm. (Courtesy of S. Fakan.)

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

I67

FIG. 2. The lamina (L). underlying the membrane element of the nuclear envelope. is unusually deep and well-visualized in neurons of the leech. and shows channels corresponding to the positions of the pores (P).Fixation in 2 9 glutaraldehyde-29 formaldehyde followed by 2% OsO,. Bar = 500 nm. (From Stelly el d.,1970.)

tally difficult to separate and characterize each individually. Most preparative techniques for “nuclear membranes” result in copurification of all three elements together (for example, Franke, 1974; Conner et al.. 1980). and it has not been possible by biochemical procedures to isolate the pore complex separately from the lamina. A. THEMEMBRANE The membrane component sensu strictu appears in EM sections as two layers of typical bilayer membrane of 5-7.5 nm thickness after fixation, whose appearance suggests that it may be folded back on itself at the periphery of the pores (for example, Fig. 3). The lipid content is relatively low in comparison with that of other cellular membranes; 6 0 4 5 % of the lipids are phospholipids, and the lipid pattern resembles that of the endoplasmic reticulum membranes except for a several-fold higher level of esterified cholesterol (Kashnig and Kasper, 1969; Kleinig, 1970; reviewed in Franke, 1974). These characteristics do not as yet appear to give any clues as to special structural or functional properties of the membrane element, but the cholesterol content is known to have important effects on the dynamic properties of membrane lipids. The application to the nuclear membrane of biophysical techniques used to study cell surface membranes (for example, Nicolau et al., 1978) may be rewarding. Most reports indicate that both layers of the membrane, and not only the outer layer, are solubilized by nonionic detergents such as Triton X- 100 and Nonidet P-40 under the conditions used in nuclear purification procedures, with extraction of essentially all the phospholipid components (for example, Barton er al., 1971; Aaronson and Blobel, 1974; Scheer et al.. 1976).

FIG. 3. Localization of lamin A in nuclei and in the nuclear envelope of rat liver by ultrdstructurd1 immunocytochemical methods. ( A ) Labeling of nuclei with antilamin A IgG by an indirect immunoperoxidase technique shows an exclusively perinuclear location of lamin A. which is not seen in regions where the lamina has been detached (lower large arrow). No structures in the nuclear interior, including nucleoli (small arrows), show reaction product even in nuclei which have been fractured open before the staining procedure (large arrows). Bar = 5 pm. (From Gerace e t a / ., 1978.) (B and C) Labeling of the nuclear envelope with antibody specific for lamins A and C by an indirect immunoferritin procedure, as seen in transverse ( B ) and tangential sections (C). INM and ONM. inner and outer nuclear membranes; arrowheads, pores. (From Gerace and Blobel. 1982.)

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

169

The proteins removed simultaneously probably represent at least a part of those of the membrane element; they differ extensively from those of the endoplasmic reticulum, and many are glycoproteins (Sieber-Blum and Burger, 1977; Richardson and Maddy, 1980b). The membrane element cannot yet be experimentally separated from the other envelope components, so it is not clear to what extent the permeability properties of the whole envelope reflect those of the membrane element (see Section 11,D).

B. THEPOREELEMENTS The pores are ultrastructurally complex structures embedded in and spanning the two membrane layers (Figs. 2, 3, 6, and lo), whose external diameter as seen by sectioning, freeze-fracturing, and negative staining is very close to 100 nm (Scheer et al., 1976; Krohne et a!., 1978; reviewed by Franke, 1974; Maul, 1977b). In some cases the pores may be visualized as linked together by a network of fibers (Scheer et a/., 1976) which appear to be derived from the lamina (Krohne, 1982). Isolation and characterization of the pore element separately from the lamina have been possible only in the case of amphibian oocyte nuclei, where the envelope is highly enriched in pores and may be isolated manually (Maul and Avdalovic, 1980; Krohne eta!., 1978, 1982). Pores isolated in this way contain one principal polypeptide of 68,000 MW, whose ultrastructural identification as a component of the pore element has been established (Stick and Krohne, 1982). Polypeptides of a similar molecular weight are found in nuclear envelope preparations from rat liver, and may therefore be universal components of the pores (Krohne el al., 1978), whose architecture appears to be identical in many widely different cell types (Franke, 1970, 1974). Intermediates in pore assembly have not been detected in ultrastructural studies; by analogy with the formation of the new nuclear envelope after mitosis (see Section VI,B), it appears possible that new pores may be inserted into the envelope essentially completely assembled. The annulate lamellae, structures found in the cytoplasm and nucleus which bear a close ultrastructural resemblance to pores and, like them, contain a nucleoside triphosphatase activity (Scheer and Franke, 1969; Franke, 1974; Maul, 1977a,b), could represent a reserve of pores for this purpose.

C. THELAMINA The lamina is visualized in electron microscopic sections as a structure of homogeneous texture lying between the membrane and the peripheral chromatin, whose thickness lies between 80 and 300 nm in different cell types, and which is interrupted at the pores (Figs. 2 , 3, 6, and 10)(Patrizi and Poger, 1967; Kalifat et al., 1967; Stelly et al.. 1970; Barton er al., 1971). Although the lamina is a

I70

RONALD HANCOCK AND TEN1 BOULIKAS

ubiquitous component of the envelope of somatic cell nuclei, it is absent from spermatocytes and spermatids of the chicken (Stick and Schwartz, 1982). Its relative visibility depends on the fixation procedure, and varies greatly in different cell types for reasons which are not clear; the optimum conditions for reaction with electron microscopic fixatives and stains may not yet have been established. Ultrastructural cytochemistry and the effects of enzymic digestions indicate a predominantly protein composition (Stelly et a/., 1970). The lamina is usually copurified together with the pore element in fractionation procedures; the predominant polypeptides in such preparations of lamina plus pores are the lamins, polypeptides in the range of 65,000-75,000 MW. These appear by immunological crossreactivity, peptide mapping, and two-dimensional gel analysis to be members of a structurally related family present in all species examined; there is evidence that lamin C may be a cleavage product of lamin A (reviewed by Shelton et a / . , 1981). The lamins have been shown by ultrastructural immunocytochemistry to be localized exclusively in the Iitmina element (Fig. 3) (Gerace et a/., 1978). Molecules of lamins A and B im in

Fio. 4. A section through a chromatin structure from a P815 mouse cell, fixed in glutaraldehyde and Os04. and stained with uranyl acetate and lead citrate. The membrane elements and the pores of the nuclear envelope are no longer visible. but the lamins are retained in these structures. which maintain the form of the chromatin in the interphase nucleus (Hancock et a/..1977: Hancuck and Hugues. 1982). Bar = I p n . (Courtesy of S . Fakan.)

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

171

!

FIG. 5 . The peripheral cortex, which represents the peripheral chromatin attached to elements probably derived from the lamina or pore-lamina structures. detached from the surface of chromatin structures of HeLa cells (see Fig. 4) by exposure to 20 mM Tris-HCI (pH 8) at 4°C (Bouvier. 1982). Chromatin fibers (arrows) are stretched between the detached cortex (C) and the remaining chromatin (CHM). Spread by the Miller procedure (for example, McKnight et a / . . 1978). Bar = I pn. (Courtesy of D. Bouvier.)

contact with each other in the nucleus in vivo. because they form homotypic oligomers upon crosslinking with reagents or by oxidation of -SH groups; lamin B appears to exist as a tetramer (reviewed in Shelton er a/., 1981). Both indirect and direct evidence is compatible with the idea that the lamina represents, or is a major component of, a skeletal element to which chromosomal DNA is attached in the interphase nucleus (see Sections IV,B and C ) . The lamins remain associated with the peripheral chromatin, which is apparently largely free of the pore element, in chromatin structures released by lysing cells with a nonionic detergent in very low ionic strength conditions (Fig. 4) (Hancock et a/. , 1977; Hancock abd Hugues, 1982). The lamina, together with the peripheral chromatin, may then be detached in the form of a nuclear cortex or shell (Fig. 5) (Hubert er a/., 1979; reviewed by Bouvier, 1982). The apparent relative resistance of the chromatin in this peripheral structure to unfolding suggests that it may be related to the chromatin which maintains a more compact conformation during nuclease digestion of nuclei and is highly enriched in satellite DNA (Horvath and Horz, 1981). and is compatible with the idea that the peripheral lamina-associated chromatin may possess a specific comformation. The nonhistone proteins associated with chromatin containing satellite DNA in some (for example, Hsieh and Brutlag, 1979; Levinger and Varshavsky, 1982) but not all

172

RONALD HANCOCK AND TEN1 BOULIKAS

cases (Mathew et al., 1981), and the regular phasing of nucleosomes on clertain satellite DNAs (Musich et al., 1982) could be implicated in such a specific conformation, which could be related to the orientation of peripheral chromatin approximately perpendicular to the nuclear surface (discussed in Olins and Olins, 1979) (Figs. 9 and 11). D. FUNCTIONAL ACTIVITIES OF THE ENVELOPE Very few enzymic activities have been rigourously demonstrated in the nuclear envelope. A glucose-6-phosphatase activity, whose functions are unknown, has been demonstrated within the cisternal space by ultrastructural cytochemical techniques (Fig. 6B) (Kartenbeck et al., 1973; Sikstrom et al., 1976). .4nucleoside triphosphatase activity (Fig. 6A) (Scheer and Franke, 1969; Vorbrodt and Maul, 1980) located in or close to the lamina has been extensively studied; its possible role in the transport of RNA out of the nucleus is discussed in Section V,D. The ultrastructural similarity of the membrane element of the nuclear envelope with other cellular membranes suggests that it offers a permeability barrier to small ions and macromolecules, but since it has not been experimentally separated from the pore element its intrinsic permeability properties have not yet been studied. It is not immediately apparent why a membrane element is necessary around the nucleus, since the pores appear to offer open channels of mac-

FIG.6. Localization of enzyme activities in elements of the nuclear envelope by ultrasiructural cytochemical procedures. (A) The distribution of Mg2+-activated ATPdse in the envelope of oocytes of the newt Triturus dpestris. The reaction product is localized in the perinuclear cisterna Ixtween the two layers of the membrane element, with a tendency to concentrate adjacent to the pores (From Scheer and Franke, 1969.) (B) The distribution of glucose-6-phosphatase in isolated rat liver nuclei. The reaction product fills the perinuclear space, with no preferential association with the pores; in tangential sections, it appears to be located on the intracisternal surface of both membrane layers. Bars = 500 nm. (From Kartenbeck el a / . . 1973.)

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

I73

romolecular diameter across the envelope. The space between the two layers of the membrane forms a discrete compartment, the perinuclear cisterna (Figs. 6A and lo), within which enzymatic activities have been localized, and the function of the membrane elements may be to isolate this compartment rather than to isolate the nuclear interior from the cytoplasm. The potential importance of the perinuclear cisterna has been somewhat neglected, and it may be noted that enzymic or other activities localized in this compartment may be lost from “purified” nuclei, since parts or all of the outer membrane layer are often removed during their preparation, even in the absence of detergents (for example, Richardson and Maddy, 1980a). It may not be without significance that the membrane is a double structure whose appearance in sections suggests that it may be folded back on itself at the pores. The polarity of each individual layer could thus be in the opposite sense, so that phenomena normally dependent on membrane assymetry could be suppressed when the two layers are considered together. It is striking that the envelope can be punctured during microinjection procedures without affecting the viability of the cell (Stacey, 1980). The envelope appears to be freely permeable to many monovalent inorganic ions, amino acids, dyes, nucleosides and nucleotides, and glycolytic intermediates (reviewed by Kohen el al.. 1971). Earlier observations showing higher concentrations of Na+ , K , and CI ions within the nucleus (Siebert et al., 1965) have not been confirmed by recent measurements made by X-ray microanalysis on sections of rapidly frozen cells, which do not show a higher concentration of these ions in the nucleus, and indicate that K is the preponderant nuclear cation and that migration of ions may occur during the isolation of nuclei (Jones et al., 1979; Pool et al., 1981). These findings may imply a reevaluation of the “physiological conditions” which should be used for in vitro studies of nuclear functions. In vivo there is extensive and highly specific vectorial transport of many macromolecules, including RNP particles and their protein components, histones, regulatory proteins, and free or receptor-bound hormones, through the nuclear envelope. The ultrastructure of the pores suggests that they provide channels of macromolecular diameter through the membrane. Our knowledge of their permeabilty properties is derived predominantly from studies of the envelope of the giant nuclei of amphibian oocytes, which is composed virtually completely of pores whose structure and dimensions are essentially identical to those of the pores of somatic cell nuclei (reviewed by Franke, 1974; Maul, 1977b), so that it is likely (although not established) that their permeability characteristics reflect the general properties of nuclear pores. The oocyte nuclear envelope shows a sieving effect for entering colloidal gold particles (Feldherr, 1965), dextran (Paine et al., 1975), and inulin (Horowitz and Moore, 1974) compatible with the existence of diffusion channels of about 4.5 nm effective diameter which can almost certainly be identified with the pores. Proteins of up +

~

+

174

RONALD HANCOCK AND TEN1 BOULIKAS

to 20,000 MW, including histones, enter the oocyte nucleus freely when microinjected into the cytoplasm, while ovalbumin (68,000) suffers some retardation (Bonner, 1975a,b); the entry of many endogenous cytoplasmic proteins into the oocyte nucleus does not appear to be limited by the nuclear envelope (Feldherr and Ogburn, 1980). The conformation of proteins also plays a role, since some cellular proteins of up to 130,000 MW are able to enter the oocyte niicleus (Bonner, 1975b). Factors other than size are clearly responsible for the selective reentry of nuclear and nucleolar proteins into the same compartment of the host cell when introduced into salivary gland cells (Kroeger et a / . , 1963), oocytes (Bonner, 1975b), cells of Amoeba proteus (Jelinek and Goldstein, 19731, and mouse-chicken heterokaryons (Jost et al., 1979) Nucleic acids microinjected into the cytoplasm can pass into the nucleus (reviewed by Stacey, 1980). The genomes of viruses which replicate in the nucleus pass through the nuclear envelope in both directions, either as intact viral particles (for example, parvoviruses: Mattern et a / ., 1966; Hummeler tit a / ., 1970) or as nucleoprotein core particles (for example, adenoviruses: Morgan et a / . , 1969; Chardonnet and Dales, 1972; reviewed by Heine, 1974). The exit of completed herpes viruses occurs by a budding process through the cisternal space (Stackpole, 1969; Smith, 1980). Receptor activity for a number of regulatory molecules has been demonstrated in the nuclear envelope. The specific nuclear binding of nerve growth factor is observed only when the membrane element is present (Yankner and Shooter, 1979), and its receptor may therefore lie in the cistemal space. High-affinity specific binding sites for insulin have been identified in the envelope of rat liver nuclei (Vigneri et a / . , 1978), but the resistance of binding to Triton X-100may indicate that in this case the membrane element is not involved; a similar conclusion may be drawn for the progesterone receptor (O’Malley et al., 1970). The transport of RNA out of the nucleus is discussed in Section V,D. At the onset of mitosis all the elements of the nuclear envelope are dissassembled into forms which are essentially undetectable at the ultrastructural level, and are reassembled around the two daughter chromosome complements after rriitosis (see Section VI,B).

111. Chromatin

Several comprehensive reviews of molecular aspects of chromatin, including the structure of nucleosomes (McGhee and Felsenfeld, 1980; Mirzabekov, 1980), the H1 histones (Hohmann. 1978), the HMG proteins (Goodwin and Johns, 1982), variant histones (Zweidler, 1980), modification of histones (lsenberg, 1979), the characteristics of transcribed chromatin (Mathis et al., 1980; Cartwright et a/., 1982), chromatin replication (DePamphilis and Wassarman,

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

175

1980), and general aspects (Igo-Kemenes et al., 1982) have been published recently. We have therefore limited this section predominantly to a discussion of some areas to which ultrastructural immunocytochemical or cytochemical methods could be fruitfully applied to supply valuable data. A. THE NUCLEOSOME 1. Structure

It has become apparent that the nucleosomal subunit of chromatin is not a static entity, but undergoes changes associated with transcription, replication, and formation of higher order structures, and which possesses an important degree of heterogeneity. The nucleosome has the form of a flat cylinder with dimensions in solution of approximately 1 I X I I x 5.5 nm, measured by neutron scattering (Suau et al., 1977; Finch et al., 1977). Studies of crystals of nucleosomes and of core histone octamers by X-ray diffraction have led to a model of the nucleosome as a discoidal particle composed of a protein core shaped as a left-handed helical spool with a dyad axis of symmetry, surrounded by 1.75 superhelical turns (146 base pairs) of DNA (Fig. 7A) (Klug er ul.. 1980; Finch et al., 1981). Individual nucleosomes may be visualised in sectioned nuclei (Fig. 9) (Olins er al.. 1980). 2. Heterogeneity of Nucleosomes Several factors are responsible for an important degree of compositional heterogeneity of nucleoiomes, which may reflect functional heterogeneity. The significance of the existence of nonallelic primary structure variants of the histones is still not clear. In the mouse, the variants of H 1 , H2, and H3 have been classified (Zweidler, 1980) into replication variants which occur in rapidly replicating cells, and replacement variants which may be incorporated into nucleosomes, possibly those associated with specific functions, in the absence of DNA replication (discussed in Russev and Hancock, 1981). The histones whose synthesis continues at a low level through the whole cell cycle may represent this class (Wu and Bonner, 1981). A third class of variants appears in the chromatin of spermatocytes during meiotic prophase. In chromatin or nucleosomes of Friend cells, IgG directed against variant 1 of H2A shows much higher binding than antibody against variant H2A.2, suggesting that the antigenic determinants of these two variants are differently exposed; a mononucleosome fraction which probably arises from transcribed chromatin is more reactive toward anti-H2A. I than is the remaining chromatin (Benezra et a / . , 1981). The replacement of H2A and H2B by their variants leads to significant differences in the properties of nucleosomes (Shaw et al.. 1981; Simpson, 1981). The changes in the spectrum of variants during development of the sea urchin and the chicken (for example, Newrock et a l . , 1978; Urban and Zweidler,

176

RONALD HANCOCK AND TEN1 BOULIKAS -

2

'I

1"

116

coo-

HISTONE HI

Hist one

C FIG.7. Models of the hierarchy of conformations into which the DNA is packed in chromatin. (A) In the nucleosome, the histone octamer forms a core on which two turns of DNA are wound. The

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

177

1980) could be associated with functionally significant modifications of nucleosomal properties, although as yet no direct experimental evidence bears on this point. A correlation has been made between the relative levels of a phosphorylated hydrophobic variant of H2A and of constitutive heterochromatin in cell lines from the deer mouse (Halleck and Gurley, 1980). However, it appears that in yeast cells neither H2B variant has a unique essential function, since frameshift mutations in either of the H2B variant genes are not lethal (Rykowski er al., 1981). Further heterogeneity of nucleosomes arises from the presence of 5-10% of the total H2A in the form of its ubiquitin derivative (Goldknopf and Busch, 1977; Albright et a/., 1979), whose ubiquitin moeity may be replaced independently of nucleosome assembly (Seale, 1981). In addition, subpopulations of all the histones contain poly(ADP-ribosyl) substituents, whose physiological function remains unknown (for example, Jump er al., 1979; Braeuer e f al.. 1981). Further subpopulations of nucleosomes, which may be associated with transcriptional activity, contain HMG proteins (reviewed by Goodwin and Johns, 1982). Secondary modification of histones by acetylation, phosphorylation, and methylation also contributes to nucleosome heterogeneity; for example, in most tissues of the mouse and chicken, 30 to 40% of the H3 and H4 molecules are thus modified (Urban and Zweidler, 1981; Zweidler and Urban, 1982). Many observations have suggested a correlation between histone acetylation and transcriptional activity of chromatin, but the relationship between these two events at the molecular level is not clear (Mathis et a / . , 1980). The application of ultrastructural cytochemical methods to the identification, in sectioned or spread chromatin, of nucleosomes containing these specific components or modifications could provide important information as to their functional significance. The spacing of nucleosomes on DNA is variable in a single cell type (for example Todd and Garrard, 1977; Samal et a / . . 1981) and shows temporal changes during chromatin replication (reviewed by DePamphilis and Wassarman, 1980) and during development (Brown, 1978; Savic er a/., 1981). numbered points correspond to the double-helical repeat of the DNA, and approximately to sites of maximal cutting frequency by DNase I. (The diameter of the DNA is underrepresented for clarity.) (From Klug et a / . . 1980.) ( B ) In the approximately 10-nm-diameter nucleosomal fiber, the central globular region of HI (residues 40 to I 16) seals the nucleosomal DNA at its entry and exit points, and the terminal regions bind to the internucleosomal linker DNA. (From Boulikas, 1979.) (C) The conformation of the approximately 34-nm-diameter higher order coiled conformation of chromatin in solution, constructed from low-angle X-ray scattering studies. The location of HI in this conformation (internal, external, or alternating) is not yet established (see Section 1lI.C). (From Notbohm CI a / . . 1979.)

178

RONALD HANCOCK AND TEN1 BOULIKAS

B. THENUCLEOSOMAL FIBER The major part of chromatin is organized in the form of a chain of H1containing nucleosomes or chromatosomes (Simpson, 1978b; Reudelhuber et al., 1980). Chromatin spread for electron microscopy shows a zig-zag appearance with apparently coincident entry and exit points of linker DNA, whereas in HI-depleted chromatin these points are more random and are predominantly on opposite sides of the nucleosome (Thoma et al., 1979; Thoma and Koller, 1981). These observations, together with evidence that the globular part of 141 is . in contact with the C-terminal region of H2A in chromatosomes and oligonucleosomes, has led to a model that HI seals nucleosomal DNA at its entry and exit points on the nucleosome (Fig. 7B) (Boulikas, 1979; Boulikas et al., 1980). This approximately 10-nm-diameter conformation appears as a row of closely spaced nucleosomes, and the core particles are probably arranged with their flat faces inclined at 20” to the fiber axis (Suau et al., 1979; McGhee et al., 1980) C. HIGHERORDERFOLDING OF THE NUCLEOSOMAL FIBER The nucleosomal fiber may be further folded into a coil with a hydrated diameter of about 34 nm (25-30 nm in sectioned or spread chromatin) and a central hole of about 10 nm diameter, and a linear compaction of DNA along the solenoid axis of about 40 (Finch and Klug, 1976; Suau et al., 1979). This conformation can be visualized by electron microscopy when chromatin in the zig-zag conformation is exposed to 20 mM NaCl or to 0.3-1 mM MgCI,, and is stabilized by HI (Fig. 8e and r) (Finch and Klug, 1976; Thoma et a/., 1979; Thoma and Koller, 1981; Murcia and Koller, 1981; Labhart and Koller, 1981). This conformation contains about six nucleosomes per turn (Finch and Mug, 1976; Suau et al., 1979; McGhee et al., 1980), whose flat faces are probably oriented approximately parallel to the solenoid axis (Fig. 7C) (McGhee e! al., 1980). This conformation is probably that which is visualized as 25- to 30-nmdiameter fibers in fixed chromatin after sectioning or spreading (Figs. 9 and 1 I ) ( O h and O h , 1979; Rattner and Hamkalo, 1979). Other observations suggest that this higher order conformation may be subdivided into oligonucleosomal elements ( “superbeads”: reviewed by Jorcano et al., 1980) which could correspond to the oligonucleosomal structures visualized in sectioned and in spread chromatin (Pruitt and Grainger, 1980; Zentgraf et al., 1980; Samuel et al., 1981). Short micrococcal nuclease digestion of nuclei or chromatin generates octanucleosomal units (Stratling et a l . , 1978; Butt et al., 1979); the oligomeric products of crosslinking H1 are compatible with the existence of units containing about 12 nucleosomes (Itkes et al., 1980). Histone HI plays a central role in forming and stabilizing these higher order conformations. If about 10%of the total H1 is removed from chromatin, regular

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

179

FIG. 8. (a-f) The influence of ionic strength on the conformation of soluble HI-containing chromatin (unfixed). The NaCl concentration was increased from 0 mM (left) to 10 mM (center) and to 75 mM (right); the chromatin was adsorbed to uncoated (upper) or to amylamine-treated carbon films (lower). Bar = 200 nm. (From Murcia and Koller. 1981.)

higher order structures are no longer formed upon increasing the ionic strength (Thoma and Koller, 1981). H1 shows a preference for binding to octanucleosomes in 40 mM NaCl (Renz er a l . , 1977), and when formation of higher order structure is induced in oligonucleosomes by exposure to 80 mM NaCI, the H I becomes masked to anti-H 1 antibody (Takahashi and Tashiro, 1979). The contact sites between H 1 and the core histones (other than the primary H2A contact) are more predominant in chromatin possessing higher order structure (Boulikas et a / ., 1980). It is not yet clear if H 1 molecules lie inside the 34-nm fiber. or are located alternately inside and outside (McGhee et al., 1980). The different subfractions of H 1 differ greatly in their ability to condense dinucleosomes (Liao

180

RONALD HANCOCK AND TEN1 BOULIKAS

FIG. 9. Stereo images of 25-nm-diameter chromatin fibers in the nucleus of a chicken erythrocyte. Close-packed nucleosomes can be visualized within the higher order structure in the bracketed area of the right image. At the periphery of the nucleus (above), the 25-nm fibers show a parallel alignment approximately perpendicular to the envelope. Bar = 100 nm. Tilt angle i: 10". (From Olins and Olins. 1979.)

and Cole, 1982); the changes in their relative amounts in nondividing cells (Pehrson and Cole, 1980), during embryonic development (Seale and Aronson, 1973), and in different organs (Seyedin and Cole, 1981) could therefore be reflected in fine differences of higher order structure.

IV. Topological Organization of Interphase DNA and Chromatin A. INTRANUCLEAR DISTRIBUTION OF CHROMATIN A considerable amount of evidence indicates that chromatin is organized in a precise and topologically defined manner within the interphase nucleus. Three-

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

181

dimensional reconstruction from serial sections shows that the number of nuclear envelope-associated areas of condensed chromatin in the newt erythrocyte is very close to the number of metaphase chromosomes, compatible with the notion that each represents an element derived from one mitotic chromosome which is conserved throughout interphase (Murray and Davies, 1979). Structures located in the centromeres of mitotic chromosomes can be identified during interphase in contact with the nuclear envelope in CHO cells, by ultrastructural immunocytochemical detection of their proteins (Moroi er a / . , 1981). The inactive X chromosome (the Barr or sex chromatin body) also remains during interphase in close contact with the lamina on the inner surface of the nuclear envelope (Fig. 10) (Schellens er a/., 1979). Heterochromatin regions containing satellite DNA, found in the centromeric regions of mitotic chromosomes, can also be identified at the nuclear periphery during interphase (Rae and Franke, 1972; see also Hecht er a / ., 1974). At mitotic prophase, chromosomes condense toward sites located on the nuclear envelope (Comings and okada, 1970), and there is some, but not yet rigorous, evidence that these sites may be distributed in a defined topological pattern (for example, Coll er a / . , 1980). Taken together, these observations suggest that structural elements originating in the centromeric region of each metaphase chromosome, together with some condensed chromatin (or all in the case of the inactive X chromosome), are retained as discrete structural entities during interphase, when they occupy sites in contact with the nuclear envelope.

FIG. 10. The condensed inactive X chromosome (sex chromatin or Barr body) in an interphase cell of the cat adrenal cortex. in close contact with the lamina (L) except in regions corresponding to the pores. Bar = 0.5 p m . (From Schellens CI ul.. 1979.)

I82

RONALD HANCOCK AND TEN1 BOULIKAS

Reconstruction from serial sections suggests that nucleoli are always in contact with the nuclear envelope during interphase (Bouregois et al., 1979). The close contact of polytene chromosomes, and of all meiotic chromosomes, to the inner surface of the nuclear envelope is well documented (Skaer and Whytock 1975; Moens, 1969); the lampbrush chromosomes of the amphibian oocyte appear to be an exception to this generalization. The attachment of chromatin fibers to an element of the inner surface of the nuclear envelope has been frequently documented, and is seen particularly clearly under some conditions of preparation for electron microscopy (Fig. 1 1 ) (Puvion-Dutilleul and Puvion, 1980) or when the chromatin is displaced within the nucleus by exposure to an electrical field (Skaer ef al., 1976). At the biochemical level, chromatin and DNA frequently remain attached to the inner surface of isolated nuclear envelopes; the pores, the lamina, and the membrane have all been proposed as possible attachment sites. Distinct compartments within the nucleus differ in their accessibility to ions and macromolecules, and may represent the regions of decondensed and condensed chromatin. Only about 30% of the space within nuclei of rat liver is

F a . 1 1 . Chromatin fibers (25 nm diameter) attached and oriented perpendicularly to an element in the nuclear periphery of a CVl cell; the membrane element of the envelope is no longer detectable. The cell was treated with Photoflo during fixation, then pelleted, embedded, and sectioned. Bar = 500 nm. (From Puvion-Dutilleul and Puvion, 1980.)

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

183

accessible to glycogen molecules of 1 I - I5 nm diameter; 40-50% is accessible to molecules of 4-5 nm diameter (a size class which includes DNase I and micrococcal nuclease). while about 30% is essentially inaccessible. An important consequence of this compartmentalization is that a number of enzymes, including RNA and DNA polymerases, would be excluded from all but the first compartment of the intranuclear space solely because of their size. This situation can also explain several aspects of the selective sensitivity of nuclear DNA to DNase 1 and to micrococcal nuclease (Burgoyne et a / . , 1978; Burgoyne and Skinner, 1979), and may be the origin of the apparent compartmentalization of nucleosomes which are accessible to histone acetylation systems (Cousens et a/., 1979).

B. STRUCTURAL ELEMENTS IN CHROMATIN The nature of the structure(s) responsible for the ordered topological organization of DNA and chromatin in the interphase nucleus is the subject of much current work. The structures which have been described fall in general into one of two morphological classes: they contain in all cases an element derived from the lamina of the nuclear envelope, but are distinguished by either the presence or the absence of additional structural elements extending throughout the nuclear interior. The most studied class is that termed the nuclear matrix, which contains extensive internal elements (for example, Berezney, 198 I ; Kaufmann et al., 1981; Zbarsky, 1981). The presence of these internal elements depends on the experimental protocol; they are not observed if RNA is removed from the initial nuclei by RNase treatment before dissociation of histones by exposure to solutions of high salt concentration (Fig. 12A and B) (Adolph, 1980; Kaufmann et al., 1981). RNase-resistant internal elements are formed if the formation of protein disulfide bonds is favored by storage of the experimental material, the use of certain detergents, or by the presence of oxidizing agents, but these elements may be removed by subsequent treatment with reducing agents (Fig. 13). When nuclei allowed to swell in EDTA are digested by DNase 11, only a peripheral shell remains, but after reincubation of the nuclei with Mg2+ ions followed by a conventional nuclear matrix isolation procedure, extensive internal elements, which resist subsequent extraction by EDTA, are visualised (Galcheva et a/.. 1982) (Fig. 12C and D). These observation suggest that contacts established between intranuclear elements in the presence of Mg2+ ions result in formation of an internal network, which is resistant to subsequent extraction in high salt concentrations and EDTA. The estrogen and androgen-binding functions of nuclear matrix preparations have been localized in the internal elements, because they may be removed by RNase digestion in the presence of a reducing agent (see references in Kaufmann et al., 1981).

184

RONALD HANCOCK AND TEN1 BOULIKAS

FIG. 12. The ultrastructural appearance of residual structures prepared from rat liver nuclei by different experimental protocols. (A and B) Freshly prepared nuclei were extracted with 1% Triton X-100in STM buffer [250 mM sucrose, 50 mM Tris-HCI (pH 7.4). 5 mM MgS04], and digested with DNase I in the presence (A) or absence (B)of RNase A. They were then extracted with 10 mM Tris-HCI (pH 7.4). 0.2 mM MgS04, followed by the same buffer containing 2 M NaCl (all buffers contained I mM phenylmethylsulfonyl fluoride). Without initial RNase digestion, the presence of residual nucleoli (B, arrow) is reflected in increased RNA and protein contents. (From Kaufniann ef a/., 1981, where the detailed experimental protocols are described.) (C and D) Nuclei were first allowed to swell in 5 mM EDTA, and then either digested with DNase I1 (C), or digested with DNase I and RNase in the presence of Mgz+ followed by extraction in 2 M NaCl (D) under conditions similar to (A) and (B).(From Galcheva el a/.. 1982.) Bars = 2 pm.

DNA replication and RNA transcription are included in the wide range of functions which have been ascribed to the nuclear matrix (references summarized in Kaufmann et af., 1981). In contrast, when chromatin is prepared from cells or nuclei by gentle lysis in solutions of very low ionic strength (for example, McKnight and Miller, 1979), DNA replication and RNA transcription are visualized by electron microscopy on regions of DNA having no evident relation with any structural elements (see following section). It must therefore be concluded that the internal elements of the nuclear matrix are either labile under these latter experimental conditions, a property not easy to reconcile with their postdated role as a rigid structural element of the nucleus, or that their formation is induced by the experimental conditions used for isolation of the nuclear matrix.

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

185

FIG. 13. The ultrastructural morphology of residual structures from nuclei is markedly altered if formation of intermolecular disulfide bonds occurs during the isolation procedure. Rat liver nuclei were incubated in ST M buffer (Fig. 12) containing 2 mM sodium tetrathionate (A and B) or 10 mM iodoacetamide (C and D). After washing, they were digested with DNase I and RNase A and extracted as described in Fig. 12. The residual structures from nuclei exposed to tetrathionate (A and B), but not from those exposed to the sulfhydryl-blocking agent iodacetamide, contain remnants of nucleoli (N) and an extensive internal network which is resistant to digestion by DNase I and RNase A . This internal network is removed (E,F, and G ) when the residual structures in (A) and ( B ) are treated with the disulfide reducing agent P-mercaptoethanol( 1%). (From Kaufmann et a / . . 1981 where the detailed experimental protocols are described.) Bars = 2 pm.

186

RONALD HANCOCK AND TEN1 BOULIKAS

We believe that the evidence available at present is not sufficient or rigorous enough to establish that the intranuclear network elements of nuclear matrix preparations exist in vivo, and critical analyses of this question have been made by Kaufmann et af. (198 1) and by Galcheva et al. (1982). The available results could be accommodated equally well if the internal elements in nuclear inatrix preparations result from aggregation of nascent, or other classes of, RNP particles induced by high salt concentration, by Mg2+ ions, or by protein-protein crosslinking through sulfhydryl groups (observed by Malcolm and Sommerville, 1974) or through other mechanisms (for example, StCvenin et al., 1979) during the experimental manipulations, forming an internal network which survives subsequent RNase digestion unless reducing agents are present. The regions of DNA where replication is occurring are prone to associate with other nuclear components during fractionation procedures (Fakan et al., 1972), and it is notoriously difficult to eliminate artifacts of this type experimentally; controls which show that exogenous RNA added before fractionation of nuclei does not associate with nuclear matrix preparations (Miller et af., 1978) are probably not appropriate because the first products of transcription are in the form of R N P (see Section V,B). Other classes of structures with which DNA is associated have been isolated using different experimental protocols. The folded interphase genome (Benyajati and Worcel, 1976) represents the interphase chromatin depleted of H1 and nonhistone proteins. Simultaneous cell lysis and dissociation of nucleosomes releases nucleoids (Cook and Brazell, 1977, 1978; McReady e f al., 1970); the chromosomal DNA is constrained as supercoiled loops in both of these structures. By dissociating histones from chromatin structures (Fig. 4) (Hancock et al., 1977) or interphase chromosomes, we have characterized a skeletal structure which constrains interphase DNA as loops whose mean measured length is 53 kbp (Fig. 14). This structure is resistant to RNases, suggesting that DNA--RNA interactions are not involved in the attachment of DNA, and contains polypeptides essentially derived from the lamina (Hancock and Hughes, 1982). This skeletal structure is able to bind DNA in a manner resistant to high salt concentrations, and which shows specificity for those sequences which were associated with it in vivo (Dessev and Hancock, 1982). Preliminary evidence suggests that the lamins are in contact with DNA in nuclei (Boulikas and Haiicock, unpublished), and since they exist as oligomers in vivo (see Section II,C) they are good candidates for a DNA binding structure; the lamins may be further representatives of a class of proteins similar to the Int protein of phage lambda (Hamilton et a/., 1981), which is able both to bind to DNA and also to form multimeric structures. The molecular features of the attachment of DNA to these different classes of skeletal structures represent an important area for future investigations. The

FUNCTIONAL ORGANIZATION IN T H E NUCLEUS

187

FIG. 14. The looped conformation of chromosomal DNA after dissociation of nucleosomes from chromatin structures of mouse cells (Hancock e t a / . , 1977).The DNA loops have a mean length of 53 kb. with a range of 10- 180 kb. and are attached to a skeletal element derived predominantly from the lamina. Spread in a cytochrome c monolayer onto 0.5 M ammonium acetate (Kleinschmidt, 1968). Bar = I p m . (From Hancock and Hugues. 1982.)

188

RONALD HANCOCK AND TEN1 BOULIKAS

possibility that skeletal proteins may be integrated covalently into DNA should not be ruled out a priori, in view of the existence of other such proteins in eukaryotic DNA (Werner and Petzelt, 1981), and of the evidence that topoisomerases reversibly integrate into DNA (reviewed by Cozzarelli, 1980). The DNA sequences close to, or at, the attachment sites have been reported in some cases to be highly repetitive (Jeppesen and Bankier, 1979; Matsumoto, 1981; Hancock and Hughes, 1982), but intermediate repetitive (Razin er al., 1970) and random sequences (Pardoll and Vogelstein, 1980; Bowen, 1981) have been identified in other preparations. The interpretation of such experiments 1s not simple, and has been carefully discussed by Jeppesen and Bankier (197911. ORGANIZATION OF DNA C. TOPOLOGICAL Indirect evidence has suggested for some time that interphase DNA is separated by regularly spaced topological constraints into independent supercoiled domains (for example, Ide et al., 1975; Benyajati and Worcel, 1976; Igo-Kemenes and Zachau, 1978; Hartwig, 1978). These domains may probably be identified with the DNA loops of 20-180 kbp length (mean 53 kbp), whose two ends are adjacent in the majority of cases, which we have visualized (Fig. 14) (Hancock and Hughes, 1982). Sites of RNA transcription (Fig. 15) and DNA replication are found predominantly on the loop regions of DNA, and show no preferential association with the skeletal element, and the location of coding sequences within the loops is an important question to be investigated. The observed length of these DNA loops is comparable with that of clusters of closely related genes; for example, the chicken ovalbumin and two adjacent related genes, and the mouse adult P-globin genes, cover regions of 35 and 30-40 kb, respectively (Royal et al., 1979; Jahn et al., 1980). The attachment of DNA loops to elements at the nuclear periphery (Fig. 16) suggests that the expression of the genes contained in a loop could be detennined or controled by cytoplasmic signals such as hormone receptors which operate at, or near to, the attachment sites and modify the higher order folding of a loop to allow its transcription (for example, Zuckerkandl, 1981). Each loop may be regarded as topologically equivalent to a circular DNA molecule, and this equivalence facilitates reflection on the topological aspects of chromosomal DNA replication (see Section VI,A).

V. Transcription, Processing, and Packaging of RNA Recent ultrastructural and biochemical observations show that processing and packaging of RNA commence while the growing RNA chain is still being transcribed on the chromatin template, and can be integrated into a general picture of

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

I89

FIG. 15. Putative nascent RNA transcripts (arrows) on the DNA of nucleosome-free interphase chromatin. The observed predominant localization of these structures on skeleton-distal regions of DNA loops is supported by the localization of radioactive pulse-labeled RNA. and also of RNA polymerase quantitated by titration with radioactive a-amanitin, on the DNA fragments detached from the skeletal structure by digestion with restriction enzymes. Preparation as in Fig. 14. Bars = 200 nm. (From Hancock and Hugues. 1982.)

190

RONALD HANCOCK AND TEN1 BOULIKAS

/ ‘ - 1

I

I INTERPHASE CHROMOSOME(d1.m.-

23~M)

I

NUCLEUS (dlam. 8 LM) ~

FIG. 16. A model of the topological organization of chromatin in the interphase nucleus (right). based on the measured length and organization of DNA loops (Section 1V.C. and Fig. 14) (left). The lengths of the DNA and chromatin loops are represented approximately to scale relative to the diameters of nucleosome-free chromatin structures or interphase chromosomes (left) and of nuclei (right), using approximate values of 5 both for the linear compaction of DNA in the nucleosomal fiber and for the further compaction of the nucleosomal chain into a higher order coiled confotmation (Finch and Klug, 1976; Sperling and Tardieu, 1976; Renz et a/., 1977); nucleosomes are no1 drawn to scale. (From Hancock and Hugues, 1982.)

the molecular processes involved (reviewed by Fakan and Puvion, 1980; Fuvion and Moyne, 1982).

A. TRANSCRIBING CHROMATIN The specific properties of chromatin undergoing transcription have recently been reviewed (Mathis ef al., 1980; Igo-Kemenes et al., 1982). Several r1:ports suggest that histone H 1 is absent, or present in relatively low levels, in transcribing chromatin (discussed in Gabrielli er af., 1981), and in view of the role of H1 in the formation of higher order structures (see Sections IIlB and C), it is plausible that its absence would allow the nucleosomal fiber to be unfolded for transcription. Subpopulations of nucleosomes which lack H1 (Goodwin et a / ., 1979; Kuehl er al., 1980; Jackson and Rill, 1981) and are selectively soluble under appropriate ionic conditions may be derived from transcriptionally active chromatin, but rearrangement of H I during experimental manipulations is difficult to rigorously exclude at present. The selective sensitivity to nucleases of transcribed chromatin within the nucleus probably reflects both the greater accessibility of the internucleosomal DNA in the unfolded conformation (see Section III,B), as well as the greater accessibility of the intranuclear compartment in which it is situated (see Section

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

191

1V.A). Nucleosome-free regions situated in defined locations relative to coding sequences may also play a role (McGhee et al., 1981). The possible specific localization and significance of HMG proteins (especially HMGs 14 and 17) in nucleosomes of transcribing chromatin have been reviewed (Mathis er al., 1980; Gabrielli et al., I98 1 ). It is not yet clear if or how variations in the internal conformation of the nucleosome core, which have been studied in vitro (for example, Dieterich et al., 1980; Burch and Martinson, 198I ) , occur in transcribed chromatin. The development and application of ultrastructural immunocytochemical methods could provide valuable information in this area (for example, McKnight et a l . , 1978).

B . EXTRANUCLEOLAR TRANSCRIPTION The first products of transcription are visualized in sections of the nucleus as the perichromatin fibrils (Figs. I and 17), and can be identified in spread nuclear contents as nascent RNA chains already associated with, and compacted by, proteins which may produce a regular arrangement of 20- to 24-nm particles (Malcolm and Sommerville, 1974) or be distributed irregularly but no,irandomly

FIG. 17. Nascent RNA in perichromatin fiber5 in the nucleus of a rat hepatocyte after 2 minutes labeling with [jHjuridine. stained by the EDTA procedure (Bernhard, 1969). Arrows. clusters of perichromatin fibrils. Bar = I p m . (Courtesy of S . Fakan.)

192

RONALD HANCOCK AND TEN1 BOULIKAS

(Fig. 18) (Beyer etal., 1980, 1981). Kinetic experiments are compatible with the idea that at least part of the RNA is then converted into structures visualized in sections as perichromatin granules of about 30-50 nm in diameter (Figs. 1 and 19) (Fakan and Puvion, 1980; Puvion and Moyne, 1982), which could represent perichromatin fibrils folded into a higher order conformation. A nuclear fraction identified as perichromatin granules has been isolated (Daskal, 1982). Ultrastructural evidence suggests that addition of poly(A) sequences to premessenger RNA occurs in the perichromatin region of the nucleus (Fakan el a l . , 1982). The RNP particles isolated from nuclei by biochemical procedures, into which premessenger RNA sequences are packaged, show close ultrastructural similarity with perichromatin fibrils (Fig. 20) (Jacob et al., 1982), and are visualized as a chain of monomeric subunits ("monoparticles") which appear to be heterogeneous in size (Fig. 20). The protein core of purified monoparticles (informofers: Samarina et al., 1968) shows two predominant polypeptide components of about 38,000 and 40,000 MW (for example, Quinlan et al., 1974) which appear to be common to many species. However, the nuclear RNP particles considered globally show a more complex spectrum of proteins (Preobrazhensky and Spirin, 1978; Pederson and Davis, 1980; reviewed by Jacob et a l . , 1982), some of which are restricted to discrete classes of particles (Maundrell and Scherrer, 1979). These proteins probably include those derived from snRNP particles (see below), which may all share the same group of seven polypeptides (Lemer and Steitz, 1979), and also the predominant polypeptide of 75,000-80,000 MW bound to the terminal poly(A) sequence of premessenger and messenger RNA. During passage of messenger RNA into the cytoplasm, this polypeptide remains attached to the poly(A) sequence (Quinlan et al., 1974, 1977; Schwartz and Damell, 1976; Roy et al., 1979), in contrast to the polypep-

FIG. 18. A nascent RNA molecule from a P815 mouse cell, with irregularly distributed associated proteins. Stained with JTA. Bar = 100 nm. (Courtesy of S . Fakan.)

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

193

FIG. 19. Perichromatin granules in the nucleus of a rat hepatocyte, sectioned by cryoultramicrotorny and stained with uranyl citrate. Bar = 100 nm. (Courtesy of E. Puvion and W . Bernhard.)

tides of the informofers which are replaced by new, specifically cytoplasmic proteins (reviewed by Preobrazhensky and Spirin, 1978) and appear to be recycled (Ivanova ef al., 1981). Additional proteins could be responsible for the formation of higher order structure in polyparticles. Functions identified in isolated 30-40 S particles include a poly(A)-synthesizing system and an endonuclease possibly related to a processing enzyme (Niessing and Sekeris, 1970, 1972); an RNA helix-destabilizing protein of 40,000 MW which binds to RNA in vitro is a major component of the informofer proteins of Artemia salina (Thomas et a / ., 1981 ). About one to two molecules of low-molecular-weight RNAs (snRNAs) per particle are base-paired to the pre-mRNA in RNP particles, probably at sites which are potential splice junctions (reviewed by Heinrich and Northemann, 1981; Lemer and Steitz, 1981; Zieve, 1981; Liautard e t a / . , 1981; Gallinaro et a / . , 1981; Reddy and Busch, 1982; Calvet and Pederson, 1981). Models of the structural organization of RNP particles, derived from those proposed by Georgiev (1974), have been proposed (Fig. 21) (Heinrich and Northemann, 1981). The functional significance of the third major class of RNP particles visualized

194

RONALD HANCOCK AND TEN1 BOULIKAS

FIG. 20. Ultrastructural comparison of perichromatin fibrils in siru in sectioned chromatin, and of purified hnRNP particles, from nuclei of rat brain. (A) In the nucleus the perichromatin fibrils (pf). first described by Monneron and Bernhard (1969), appear as linear arrays of granules. The sections were stained by the regressive EDTA method (Bernhard, 1969) which bleaches chromatin (chr) and preferentially stains RNP. (B) Isolated hnRNP particles have the same general ultrastructure, in particular with respect to the size and distribution of granules. Bar = 100 nm. (A and B from Devilliers et a / . , 1977.) (C) Individual hnRNP fibrils of 100 S, isolated and purified in sucrose gradients, were negatively stained with uranyl acetate. They appear as folded chains of granules of heterogeneous size, apparently randomly distributed along the fibril. (From StCvenin ei d.,1976.) Bar = 50 nm.

in sectioned nuclei, the interchromatin granules, has not been established. They possess some properties comparable to those of the snRNPs, including relatively high resistance to RNase, metabilic stability of their RNA component, and relative abundance, but their diameter of 15-25 nm (Fakan and Puvion, 1980) appears to be too great for individual snRNPs alone ( 17O,OOO-2OO,OOO MW: Liautard er al., 1981).

C. THENUCLEOLUSAND NUCLEOLAR TRANSCRIPTION The nucleolus includes the regions of the genome responsible for transcription of preribosomal RNA, together with their transcription products (Fig. 22); how-

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

I95

ever, in isolated nucleoli the rRNA genes represent only a small proportion of the total DNA (Bachellerie et al., 1977). The ultrastructure and functional organization of the nucleolus have been recently reviewed (Fakan and Puvion, 1980). Although the ribosomal DNA sequences are replicated during the S phase, the number of nucleoli per nucleus appears to remain constant during interphase (Stambrook, 1974; Lepoint and Bassleer, 1978). The number of nucleoli may vary from cell to cell in a population (Evans et al., 1974), and is usually less than the total number of nucleolar organizing regions (NORs) in the metaphase chromosomes, implying either that several NORs associate to form a single nucleolus, or alternatively that some NORs are repressed and do not function, a phenomenon which is common in hybrid cells (for example, Miller et al., 1976). The formation of a single nucleolus by association of multiple NORs is correlated in some cases with a close association of the mitotic chromosomes bearing NORs, and with interconnecting fibers between their NORs (discussed in Evans et al., 1974), and could be mediated by nucleolar skeletal structures such as those described by Todorov and Hadjiolov (1979) and Franke et al. (1981a). Ultrastructural evidence suggests that the nucleolus may be always in contact with the nuclear envelope (see Section 1V.A).

RNase

sensitive

n

i

Urea

ore-mRNA

sensitive NaCl

FIG.21. A model of the internal structural organization of the monomer nuclear RNP particles, derived from studies of their disassembly by NaCl, urea, and proteolytic enzymes. Basic proteins of 30,000-45.000 MW (triangles) ionically bound to the snRNA (Imw RNA) form a core particle; larger proteins of more than 45,000 MW (rectangles) are associated with ptemessenger RNA. (From Heinrich and Northemann, 1981 .)

196

RONALD HANCOCK AND TEN1 BOULIKAS

FIG.22. Internal structures of the nucleolus of a P815 mouse cell. showing the fibrillar (F) and granular ( G )components and the fibrillar centers (arrows). Fixed in glutaraldehyde and OsOd and stained with uranyl acetate and lead citrate. Bar = I pm. (Courtesy of S. Fakan.)

The fibrillar centers of the nucleolus are very probably derived from the nucleolar-organizing regions (NORs) of metaphase chromosomes, since both contain characteristic silver-binding proteins (Fig. 23) (reviewed by Goessens and Lepoint, 1979; Hernandez-Verdun er al., 1980). The general pattern of processing and packaging of ribosomal RNAs shows substantial overall similarity to that of messenger RNAs (see Section V,B), and the association of ribosomal proteins commences in a defined sequence while the preribosomal RNA molecule is being transcribed (Chooi and Leiby, 1981; Fujisawa er al., 1979). The first product of transcription, the RNP derivative of the 45 S rRNA precursor, is visualized at the ultrastructural level as the fibrillar component (Fig. 22), and molecular events in processing this precursor into 28 S and 18 S RNAs are reflected in the transformation of the fibrillar component into the granular component (Fakan and Puvion, 1980; Puvion-Dutilleul et al., 1981). The U3 snRNA, which can be isolated base-paired to the 28 S nucleolar RNA (Prestayko et al., 1970), may play a role in this processing analogous to that of the U 1 snRNA in premessenger RNA processing.

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

D. TRANSPORT OF RNA

OUT OF THE

197

NUCLEUS

The mechanisms responsible for the vectorial transport of RNP particles from their site of transcription toward the nuclear periphery are completely unexplored. Much circumstantial evidence suggests that RNA export from the nucleus occurs through the pores (see, for example, Franke and Scheer, 1974). The abundance of pores in the envelope varies in a manner which can best be correlated with the transcriptional activity of the nucleus (Maul et et a l . , 1980). Ultrastructural evidence is compatible with the outward passage of RNP particles through the pores, although it unfortunately cannot provide proof for this process

FIG. 23. Ultrastructural localization of nucleolar silver-staining proteins in human cells. In early interphase (a), these proteins occur in the fibrillar centers of the nucleolus (DF, dense fibrillar component). and in metaphase chromosomes (b) in the nucleolar organizer region. Bars = 500 nm (a) and 100 nm (b). (From Hernandez-Verdun er d..1980.)

198

RONALD HANCOCK AND TEN1 BOULIKAS

(for example, Stevens and Swift, 1966; Monneron and Bernhard, 1969). The acquisition of a different complement of proteins upon exit from the nucleus (see Section V,B) may be accompanied by changes in particle conformation. The possible localization of some RNA processing steps at the nuclear periphery has been discussed; three viral RNA transcripts which are possible processing intermediates are found specifically in a “paranuclear” fraction, which includes elements of the nuclear envelope, prepared by extracting nuclei of SV40 virusinfected cells with NP40 and deoxycholate (Villareal, 198I). However, the enzymes responsible for processing tRNA are not associated with the nuclear envelope in oocytes of Xenopus (DeRobertis et a f . , 1981), and at least some processing of premessenger and preribosomal RNAs occurs during transcription (see Sections V,B and C ) . A role in the processing and vectorial transport of RNA has been discussed for proteins coded by intron sequences, and attached to the nuclear envelope (Slonimski, 1980). Studies of the transport of labeled RNA out of the nucleus suggest that the nucleoside triphosphatase activity localized at, or close to, the lamina and the pores (Fig. 6A) (Yasuzumi and Tsubo, 1966; Chardonet and Dales, 1972; Kartenbeck et a f . , 1973; Vorbrodt and Maul, 1980) could be implicated in this process. The two functions show similar sensitivities to inhibitors, similar kinetic parameters, and a certain degree of stoichiometry (Agutter and Birchall, 1979; Clawson et al., 1980). It has been suggested that the membrane element can also modulate RNA transport by changing the state of openness of the pores (reviewed by Herlan et al., 1979, 1980), and cytoplasmic factors affecting release of RNA from nuclei have been described (Yannarell et af., 1976).

VI. Changes in Nuclear Organization during the Cell Cycle Many ultrastructural and biochemical changes during the different stages of the cycle have been extensively documented, but there has been very little integration of these two experimental approaches and the causal relationships involved have not been elucidated (see, for example the reviews by Tobey et al., 1974; Borun et a f . , 1974; Prescott, 1976a,b). Parallel ultrastructural and biochemical studies of mutants blocked at defined stages of the cycle could represent a fruitful approach to these questions (for example, Setterfield et af., 1978; Liskay and Prescott, 1978; Sheinin and Lewis, 1980; Yasuda e t a / . , 1981). AND CHROMATIN ASSEMBLY A. G I AND S PHASES:DNA SYNTHESIS

Our understanding of the events during the G , phase has not significantly progressed since Prescott ( I976b) observed that “the time-occupying events that precede DNA replication remain to be identified.” The existence of cell lines in

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

199

which a G I phase is not detectable suggests that these events can be compressed into a sequence much more rapid than that usually observed. Changes of the distribution and degree of decondensation of chromatin during G I have been detected by morphometric methods, by ultrastructural and staining methods, and by induction of premature chromosome condensation (Blondel and Tolmach, 1965; Moser et al., 1975; Kendall et al., 1977; Rao and Sunkara, 19781, and probably reflect changes in the molecular conformation of chromatin, for example the unfolding of chromatin loops, in preparation for DNA replication and transcription. Biochemical changes detected in G , have been reviewed (Gurley et al., 1978; McCarty and McCarty, 1978; Bhorjee, 1981). Five genes determining events during G , have been identified using mutants of CHO cells (Liskay and Prescott, 1978), and it has been suggested that intracellular cyclic nucleotide levels play a role in the switches occurring during this period (reviewed by Berridge, 1975; Dolby et a l . , 1981). Initiation of DNA replication occurs in the interior of the nucleus, as does ongoing replication which takes place close to the boundary of condensed chromatin regions (Huberman et al., 1973; Wise and Prescott, 1973; Fakan, 1978). The initiation of DNA synthesis appears to be under positive control by diffusible factors which are not cell specific, since it is induced when heterologous S phase nuclei are introduced into G I phase cells by fusion, and is synchronous in binucleate cells and in heterokaryons (reviewed by Rao and Sunkara, 1978). Nascent DNA is prefeFentially associated with the matrix elements of nuclear matrix preparations, leading to a model of movement of DNA through a fixed replication site (for example, Pardoll et al., 1980). In contrast, sites of DNA replication are not associated with structural elements in chromatin spread for electron microscopy (for example, McKnight and Miller, 1979) or in nucleosome-free chromatin structures (see Section IV,C) (Hancock and Hughes, 1982). The average time for replication of a single replicon is of the order of 20 minutes; the sequence of replication of groups each containing about 1000 replicons is temporally defined, but does not follow the topological sequence of DNA organization along the metaphase chromosome for reasons which are not yet understood (reviewed by Prescott, 1976b). 1 . Replication of DNA Loops

The independent topologically constrained loops in interphase DNA are of the same order of length as replicons, and it is plausible to propose that each represents a single replicon (see discussion in Hancock and Hughes, 1982). This organization of the very long chromosomal DNA molecule simplifies the topological problems involved in its replication, and suggests a model for the spatial and temporal organization of replication sequences. The coding sequences in chromosomal DNA which have been examined, including those of the globin, dihydrofolate reductase, and ribosomal RNA

200

RONALD HANCOCK AND TEN1 BOULIKAS

genes, are replicated during the early part of the S phase, and the greater lethality of incorporation of base analogs into DNA during this period suggests that this may be a general phenomenon (Kajiwara and Mueller, 1964; reviewed by Chang and Baserga, 1977; Furst et al., 1981; Hamlin and Bieder, 1981). Transcribed sequences appear to occur predominantly within DNA loops (Hancock and Hughes, 1982), suggesting that these regions contain coding sequences and would thus be early replicating. (However, in nuclear matrix preparations, coding sequences have been localized near the matrix element: Nelkin et al., 1980; Maundrell et al., 1981.) On the other hand, satellite or highly repetitive DNA sequences are found in general to replicate late in the S phase (reviewed by Brutlag, 1980). The immediately adjacent nonsatellite sequences are also late replicating in mouse 3T3 cells, and it has been suggested that complete replication of nonsatellite sequences is a prerequisite for satellite sequences to replicate (Dooley and Ozer, 1979). As discussed in Section IV,C, repetitive sequences have been found near, or at, the attachment sites of DNA to skeletal structures in several cases. Recent observation sindicate that in the mouse, the highly repetitive satellite sequences occur not only organized as long tandem repeated units, but are also interspersed throughout the genome (Stambrook, 1981). We assume in the following discussion that the observations that coding sequences are located in DNA loops, and late replicating repetitive sequences close to the sites of attachment to the skeletal element, may be generalized. The replication of a constrained DNA loop is comparable in a topological sense to that of the circular DNA of a virus such as SV40 (Fig. 24). This DNA replicates bidirectionally until the two parental strands remain interlocked by only a few interstrand twists (Fig. 24B) (Sundin and Varshavsky, 1980, 1981), and then pauses while they are unlocked (presumably by a topoisomerase) and the two daughter strands are completed and sealed (Fig. 24C). By topological analogy, the replication of a loop of chromosomal DNA would pause after replication of the coding sequences in the loops but before replication of the repetitive sequences associated with the skeletal element, with the parental strands still interlocked close to their attachment site (Fig. 24B). The close proximity of the progeny molecules at the last step of replication could be potentially favorable for duplication of the DNA sequences in a loop by processes topologically similar to those which form dimer DNA molecules from catenated molecules in phase lambda (Fig. 24D and E) (reviewed by Nash et al., 1977, 1980). Further, the formation of structures such as the Holliday intermediate, a probably step in chromosome recombination and sister chromatid exchanges (SCE; reviewed by Wolff, 1977), is topologically plausible at this stage. The occurrence of repetitive DNA sequences near, or at, the site of skeleton attachment (see Section IV,C) would allow recombination or duplication processes to occur productively even with recombination mechanisms of relatively low precision, and the higher frequency of sequence divergence near to the ends

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

t

20 I

t

FIG. 24. A comparison of steps in the replication of the circular DNA of SV40 virus (left), and possible topologically analogous steps in the replication of a loop of chromosomal DNA constrained by attachment to a skeletal element (right). In the nucleus, the parental DNAs are in the form of a nucleosomal chain. and the progeny DNAs are probably rapidly assembled into nucleosomes after replication (see Section V1.A). (A) Parental DNA molecules. (B) In [he terminal stage of replication of SV40 DNA. the two parental strands (heavy and light), which are relaxed by a topoisomerase as replication proceeds. remain locked by several interstrand twists (for clarity, only one is shown) before completion and sealing of the progeny chains (dashed) (Sundin and Varshavsky, 1980, 1981). A comparable conformation can be envisaged for a loop of chromosomal DNA. We assume that topoisomerase is located in the unreplicated part of the DNA molecule (an energetically favorable location: Champoux and Been, 1980) near. or at. the attachement site of the chromosomal DNA loop. ( C ) The normal sequence of unlocking of the parental strands. and completion and sealing of the progeny strands. allows complete separation of the progeny DNA molecules or loops (C). In SV40 DNA, sealing of the progeny strands before resolution of the interlocked parental strands. and/ or failure of normal separation of the completed progeny molecules (probably topoisomerase-mediated), leads to the formation of catenated dimers (D) which may be precursors of recombinant molecules (E) (Jaenisch and Levine, 1971). as they are in phage lambda (Nash el a/..1977, 1980). An analogous requencc of events in a loop of chromosomal DNA could lead to its duplication (D. E).

of blocks of repated DNA sequences (Brutlag, 1980) would be compatible with the occurrence of exchanges and duplications in these regions. The possible role of duplication in the generation of the highly repetitive sequences themselves has been pointed out frequently (for example, Musich et al., 1977) Replication of a topologically constrained DNA loop requires a site of rotation about the helix axis, in the form of a topoisomerase probably integrated covalently into the DNA (reviewed by Cozzarelli, 1980). We find attractive the idea that this topoisomerase activity could be located close to, or at, the site of attachment of a DNA loop to the skeletal element, where replication would terminate according to the model proposed here. The role of topoisomerases in catenation of DNA is established (for example, Hsieh and Brutlag, 19801, and their possible role in duplication and recombination of DNA sequences has been discussed (Nash er a/.. 1977, 1980; Ikeda el a / . , 1980; Cleaver, 1981). Topoisomerase activity has been detected in nuclear envelope preparations (Yoshida et a / . , 1977); topoisomerase molecules have not yet been reported

202

RONALD HANCOCK AND TEN1 BOULIKAS

integrated into eukaryotic cellular DNA, but are found associated with SV40 virus DNA (Hamelin and Yaniv, 1979). An integrated topoisomerase molecule could itself provide an attachment site for other skeletal proteins. A group of proteins which are very tightly bound to eukaryotic DNA do not appear to correspond to topoisomerase (Werner and Petzelt, 1981). Several eukaryotic topoisomerases require ATP, and thus could theoretically be capable of driving DNA unwinding during replication (Charnpoux and Been, 1980). Replication and transcription can occur simultaneously, or in a close temporal sequence, on the same region of DNA (McKnight and Miller, 1979; Wolgemuth and Hsu, 1981). The rotation of the whole DNA of each loop about the helix axis at a point at the base of the loop during replication, required by topological considerations, could also provide a mechanism for the progressive detachment of RNA transcripts. The recent identification in biological material of DNA with a left-handed helix, the Z conformation, may have a major influence on models of DNA organization and replication (Nordheim er al., 1981). For example, the unwinding of regions of B form DNA during replication could be transferred into regions which are able to adopt the Z conformation, such as sequences comtaining alternating purines and pyrimidines or rich in 5-methyl cytosine, thus reducing the extent of topoisomerase-mediated unwinding which is required. 2 . Assembly of New Chromatin Newly replicated DNA is rapidly assembled into nucleosomes, and interinediate structures in nucleosome assembly have been detected (reviewed by DePamphilis and Wassarmann, 1980; Jackson er al., 1981; Annunziato e? al., 1981). The major part of new histone synthesis is coupled to DNA replication, but an independent synthesis of some histone fractions, which continues through the whole cell cycle, has been detected (Wu and Bonner, 1981). Newly synthesized H4 molecules pass through an acetylation-deacetylation cycle, and are methylated and phosphorylated, during their incorporation into chromatin, perhaps reflecting passage through intermediates in the assembly process (Horida et a f . . 1975; Ruiz-Carrillo et al., 1975). During a period approximately equivalent to that required for complete replication of a replicon, the new histones synthesized in parallel with DNA do not form nucleosomes on new DNA, but rather at sites which are distant from the replication fork, whose precise location has not yet been established (reviewed by Murphy et al., 1980). The maintenance of overall regularity of nucleosome spacing would thus appear to require movement of histones or nucleosome cores onto new DNA during DNA replication. A further consequence of this process of assembly is that new histones must not necessarily transit to the sites of DNA replication in the interior of the nucleus, but could associate with DNA near or at the nuclear periphery, since the newly replicated DNA is furnished with preexist-

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

203

ing nucleosome cores. Kinetic studies are compatible with the idea that the complete assembly of the DNA of each replicon into mature chromatin may occtir only when its replication is completed (Murphy et al., 1980); the compact regions seen in interphase chromosomes induced to condense prematurely (reviewed by Rao and Sunkara, 1978) could represent such completed, packaged replicons.

B. THEG , PHASEA N D MITOSIS Profound structural changes occur in all of the major components of the interphase nucleus before and during mitosis; the molecular mechanisms of these changes, and the signals which control them, remain to be elucidated (reviewed by Prescott, 1976a.b). Synthesis of proteins is required during the G , phase up to about 10 minutes before the initiation of mitosis; these could include structural proteins of DNA skeletal elements which must be reorganized or replaced for assembly and separation of mitotic chromosomes, and it would be valuable to identify them and their cellular location. 1 . Structural Changes in the Nuclear Envelope

All three elements of the nuclear envelope lose their ultrastructural identity concomitantly with the onset of chromosome condensation, and the macromolecular elements of which they are constructed must be extensively disassembled. The lipid and protein components of the envelope appear to be conserved in the cell during mitosis, possibly in the form of the granules and vesicular material, and of pore remnants, seen around metaphase chromosomes (Chai et a f . , 1974; Maul, 1977a). The lamins from the lamina are dispersed into the cytoplasm (Fig. 25), and are later reutilized to form the new lamina of the daughter nuclei (Gerace and Blobel, 1980; Jost and Johnson, 1981). The disassembly of the lamins is accompanied by, and could be caused by, their phosphorylation (Gerace and Blobel, 1980); the envelope contains an endogenous protein kinase which can phosphorylate one of the lamins in vitro (Lam and Kasper, 1979). Depolymerization of the lamina of G , or interphase cells can be induced by fusion with mitotic HeLa cells, but the reciprocal deposition of lamina onto mitotic chromosomes cannot be detected (Jost, 1982). After mitosis, new nuclear envelope is assembled on the surface of the condensed telophase chromosomes (Maul, I977a). This assembly process is blocked if the DNA in the mitotic chromosomes has been crosslinked by psoralen (Peterson and Berns, 1978), suggesting that some element of the reassembling envelope, perhaps the lamins, may first recognize a feature of the DNA which requires strand separation, The macromolecular components of the envelopes of the daughter nuclei, including the pores, are essentially completely derived from the envelope of the parental nucleus; the formation of the new envelope is thus

204

RONALD HANCOCK AND TEN1 BOULIKAS

FIG. 25. Progressive disassembly of the lamina during the prophase-anaphase stages of mitosis (shown in the phase-contrast images b-f and h-I) in rat K22 cells, visualized by immunofluorescent staining with antilamin A serum. In metaphase (i) and anaphase (k) the lamin antigens have become diffusely localized. (From Gerace era/.. 1978.)

not sensitive to inhibitors of protein synthesis (Sieber-Blum and Burger, 1977; Maul, 1977a; Peterson and Berns, 1978; Conner et a l . , 1980; Jost and Johnson, 1981). 2. Structural Changes in Chromatin Both the basic nucleosomal fiber and the 25-nm chromatin fiber conformations are conserved in mitotic chromosomes (Compton er a l . , 1976; Adolph, 1980). The 25-nm fibers of metaphase chromosomes, which form microconvular struc-

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

205

tures as visualized by scanning electron microscopy, are relaxed to lateral loops of approximately 10-nm fibers upon chelation of divalent cations or when H I is removed (Daskal and Busch, 1978; Marsden and Laemmli, 1979). Several molecular changes accompany the conformational changes in chromatin at metaphase, but their causal relationships have not yet been established (reviewed by Simmons et a/., 1980). The level of phosphorylation of histone HI increases markedly, with addition of three to six phosphates per molecule on serine and threonine residues of virtually all H 1 molecules (reviewed by Gurley et a/., 1978; D' Anna et a/., 1981 ; Ajiro et a/., 198 1). This hyperphosphorylation may promote the formation of higher order conformations by inducing interaction between the central globular regions of neighboring H 1 molecules (reviewed by Matthews, 1977; Dolby et al.. 1979). Phosphorylation of HI has also been proposed to play a role in the compact structure of satellite DNA-containing heterochromatin (Billings et a/., 1979). Essentially all H3 molecules are also phosphorylated at metaphase. Histone H4 is deacetylated before mitosis (Gomez-Lira and Bode, 1981; Chahal et a / . , 1980); since its acetylation sites, and those of the other core histones, lie within the N-terminal region, deacetylation loosens this portion of H4, perhaps making it available for internucleosomal interactions, and modifies several properties of nucleosomes (Simpson, I978b; Bode et al., 1980). The level of methylation of the core histones is also higher in G, and mitosis (Borun et a/.. 1974). The modified H2A, protein A24, is not detected in mitotic chromosomes (Matsui et al., 19791, and its removal from those nucleosomes which contain A24 in interphase could modify their interactions. The permeability of the cell membrane to Naf and K + ions, and the intracellular K + ion activity, change markedly during mitosis (Boonstra et a/., I98 I ) ; these variations, together with the disassembly of the nuclear envelope, must result in major changes in the ionic environment of the chromatin, whose importance in initiating chromatin compaction cannot yet be evaluated. The possible role of changes in the intracellular Ca2+ ion concentration or distribution in the processes accompanying mitosis has been frequently discussed (for example, Henry et al., 1980). The phosphorylation of H3, which occurs before mitosis (see above), is a Ca2 -stimulated reaction in isolated nuclei (Whitlock et a / ., 1980). The intracellular level of cyclic AMP falls before mitosis (Friedman et a/., 1976). Some structural elements derived from mitotic chromosomes appear to be conserved in the interphase nucleus (see Section IV,A); these may contribute to the formation of the skeletal structure of mitotic chromosomes (Adolph et a/., 1977), whose origin is not yet established. Changes in the topological constraints on DNA, resulting from new conformations imposed by new, specifically mitotic skeletal elements, could also play a role in generating the metaphase chromosome conformation. +

206

RONALD HANCOCK AND TEN1 BOULIKAS

VII. Perspectives The rapid recent advances in understanding of the structure and functions of chromatin, the structural organization of genes, and transcription, have left many areas of nuclear structure and activity relatively unexplored. The development and improvement of methods, especially for ultrastructural immunocytochemical approaches using antibodies prepared by conventional or monoclonal techniques, or occurring in the sera of subjects with autoimmune diseases such as lupus, and their application to sectioned or spread preparations, could help to elucidate many of the unanswered questions concerning the localization of specific proteins discussed in this review. In a similar manner, the development of in situ hybridization methods should allow the detection and localization of specific DNA or RNA sequences using appropriate probes. The mechanisms responsible for vectorial transport of macromolecules both within the nucleus, and through the nuclear envelope, represent virtually unexplored areas. Characterization of structural elements responsible for the organization of DNA in interphase and in mitotic chromosomes, the location of gene and control sequences with respect to these elements, and the mechanisms of attachment of DNA, are among the challenging areas for future research. Understanding of the topological organization of DNA should provide many clues not only to mechanisms of replication and transcription and their control, and of chromatin assembly, but also to the molecular basis of gene duplication and elimination, chromosome breakage, and recombination within the cellular genome and with viral genomes.

ACKNOWLEDGMENTS

The preparation of this article was made possible by the help of many friends. especially S. pdkan. who supplied and allowed us to use published or unpublished illustrations taken from their work. or sent us unpublished manuscripts. We are specially grateful to P. Dubied, N. Iranpour, and L. Morand for help in preparing the manuscript. Work in our laboratory was supported by grants from the Swiss National Science Foundation and EMBO.

REFERENCES Aaronson. R. P., and Blobel, G. (1974). J . Cell B i d . 62, 746-754. Aaronson. R. P., and Blobel, G. (1975). Proc. Nut/. Acad. Sci. U.S.A. 72, 1007-1011. Adolph, K. W. (1980). J . Cell Sci. 42, 291-304. Adolph, K. W., Cheng, S . M.. and Laemmli. U. K. (1977). Cell 12, 805-816. Agutter, P. S.. and Birchall, K. (1979). Exp. Cell Res. 114, 453-459. Ajiro. K.. Borun, T. W., and Cohen, L. H. (1981). Biochemistry 20, 1445-1453.

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

207

Albright. S. C.. Nelson. P. P., and Garrard, W. T. (1979). J. Eiol. Chem. 254, 1065-1073. Annunziato. A. T.. Schindlcr. R. K.. Thomas, C. A.. Jr.. and Seale. R. L. (1981). J. Eiol. Cheni. 256, IIXXO-I 1886. Bachcllerie. J . P.. Nicoloso. M.. and Zalta. J. P. (1977). Eur. J. Eiochem. 79, 23-32. Barton. A. D.. Kisieleski. W. E.. Wassermann. F . . and Mackrvicius. F. (1971). Z. Zel/fi,r.wA. 115, 299-306. Benezra. R.. Blankstein. L. A,. Stollar. B. D.. and Levy. S. B. (1981). J. Eiol. Chem. 256, 6837-6841. Benyajati, C.. and Worcel. A. (1976). Cell 9, 393-407. Berezney. R. (1981).In "International Cell Biology" (H. G . Schweiger, ed.). pp. 214-224. Springer-Verlag. Berlin m d New York. Bcrnhard. W. (1969). J . Ultrcrstruc~.Res. 27, 250-265. Bcrridge. M. J . (1975). I n "Advances in Cyclic Nucleotidc Research" (P. Grecngard and G . A. Robinson. eds.). Vol. 6. pp. 2-98. Raven. Ncw York. Beyer. A. L.. Miller. 0. L.. and McKnight, S. L. (1980). Cell 20, 75-84. Bcycr. A. L.. Bouton. A. H., and Miller. 0 . L. (1981 ). Cell 26, 155-165. Bhorjee. J. S. (1981). P m r . Nail. Acud. Sci. U . S . A . 78, 6944-6948. Billing$. P. C.. Orf. J . W.. Palmer. D. K . , Talmage. D. A.. Pan, C. G . . and Blumenfeld. M. (1979). Nircleic Acids Res. 6, 2151-2164. Blondel. B.. and Tolmach. L. J. (1965). Ex/). Cell Rrs. 37, 497-501. Bode. J.. Henco. K . . and Wingender (1980). Eur. J. Eiochem. 110, Bonner. W. M. ( 197521).J. Cell B i d . 64, 42 1-430. Bonner. W. M. (1975b). J. Cell Eiol. 64, 431-437. Bonnstra. J.. Mummery. C. L., Tertoolen. L. G. J . , Van der Saag. P. T.. and De Laat. S. W. (1981). J. Cell. Phvs. 107, 75-83. Borun. T. W.. Paik. W. K., Lee. H. W., Pearson. D.. and Marks, D. (1974). "Control of Proliferation in Animal Cells" (B. Clarkson and R. Baserga. eds.). pp. 701-717. Cold Spring Harbor Lab. Cold Spring Harbor. New York. Boulikas. T. (1979). Ph.D. thesis. University of Texas. Dallas. Boulikas. T . , Wiseman. J . M.. and Garrard. W. T. (1980). Proc. Nail. Acad. Sci. U.S.A. 77, 127-131. Bourgeois. C . A , , Hernon. D.. and Bouteille, M. (1979). J . Ultrustruct. Res. 68, 328-340. Bouvier. D. (1982). Inf. Rev. Cytol. (in press). Bowen. B. C. (1981). Nudeic Acids Rrs. 9, 5093-5108. Braeuer. H. C.. Adamietz. P.. Nellessen, V.. and Hilz. H. (1981). Eur. J. Eiochem. 114, 63-68. Brown. 1. R . ( 1978). Biochem. Eiophvs. Res. Commun. 84, 285-292. Brutlag. D. L. (1980).Annu. Rev. Genet. 14, 121-144. Burch. J. B. E.. and Martinson. H. G. (1981). Nuclcic Acids Res. 9, 4367-4385. Burgoyne. L. A,. and Skinner, J. D. (1979). J. Cell Sci. 37, 85-96. Burgoyne. L. A,. Skinner. J. D.. and Marshall, A. (1978). J. Cell Sci. 31, 1 - 1 I . Butt. T. R.. Jump. D. B . , and Smulson, M. E. (1979). Proc. Narl. Acad. Sci. U . S . A . 76, 1628-1632. Calvet, J . P.. and Pedcrson, T. (1981). Cell 26, 363-370. Cartwright, 1. L.. Keene. M. A,, Howard, G. C., Abmayr, S. M., Fleischmann. G.. Lowenhaupt. K . . and Elgin, S. C. R. (1982). Crit. Rev. Eiochcm. 6, (in press). Chahal. S. S., Matthews, H. R., and Bradbury. E. M. (1980). Nature (London) 287, 76-79. Chai, L. S.. Weinfcld, H., and Sandberg, A. A. (1974). J . Nail. Cancer Inst. 53, 1033-1050. Champoux, J . J.. and Been. M. D. (1980). I n "Mechanistic Studiesof DNA Replication andGenetic Rccornbination" ( 6 . Alberts. ed.), pp. 809-815. Academic Press. New York. Chang. H. L., and Baserga, R. (1977). J . Cell. Phvsiol. 92, 333-344. Chardonnet. Y . , and Dales, S. (1972). Virologv 48, 342-350.

208

RONALD HANCOCK AND TEN1 BOULIKAS

Chooi, W. Y., and Leiby, K. R. (1981). Proc. Natl. Acad. Sci. U.S.A. 78, 4823-4827. Clawson, G. A,, James, J., Woo, C. H., Friend, D. S., Moody, D.. and Smuckler, E. A. (1980). Biochemistry 19, 2748-2756.

Cleaver, J. E. (1981). Exp. Cell Res. 136, 27-30. Coll. M. D.. Cuadras, C. M., and Egozcue, J. (1980). Genet. Res. 36, 219-234. Comings, D. E., and Okadd, T. A. (1970). Exp. Cell Res. 63, 471-473. Compton, J. L.. Hancock, R., Oudet. P., and Chambon. P. (1976). Eur. J. Biochem. 70.555-568. Conner, G. E., Noonan, N. E., and Noonan, K. D. (1980). Biochemistrv 19, 277-289. Cook. P. R., and Brazell, 1. A. (1977). Eur. J. Biochem. 74, 527-531. Cook. P. R., and Brazell, 1. A. (1978). Eur. J. Biochem. 84, 465-477. Cousens, L. S.. Gallwitz, D.. and Alberts, B. M. (1979). J . Biol. Chem. 254, 1716-1723. Cozzarelli, N. R. (1980). Science 207, 953-960. D'Anna. J. A., Gurley. L. R., and Becker. R. R. (1981). Biochemistr?, 20, 4501-4505. Daskal, Y.. and Busch, H. (1978). In "The Cell Nucleus" (H. Busch, ed.). Vol. IV, pp. 3-45. Academic Press, New York. Daskal, Y. (1982). In "The Cell Nucleus" (H.Busch, ed.). Vol. 8, pp. 117-137. Academic Press, New York. DePamphilis. M. L., and Wassarman, P. M. (1980). Annu. Rev. Biochem. 49, 627-666. DeRobertis, E. M., Black, P., and Nishikura, K. (1981). Cell 23, 89-93. Desai, L. S., and Foley. G . E. (1970). Biochem. J. 119, 165-170. Dessev. G.. and Hancock, R. (1982). In preparation. Devilliers, G., StCvenin, J., and Jacob, M. (1977). B i d . Cell. 28, 215-220. Dietreich. A. E., Eshaghpour, H., Crothers, D. M., and Cantor. C. R. (1980). NucleicAcids.Res. 8, 2475-2488.

Dolby. T. W., Ajiro, K., Borun, T. W., Gilmour, R. S., Zweidler, A.. Cohen, L.,Miller, I'.. and Nicolini, C. (1979). Biochemistry 18, 1333-1344. Dolby. T. W., Belmont, A,, Borun, T. W . , and Nicolini, C. (1981). J. Cell Biol. 89, 78-85. Dooley, D. C., and Ozer, H. L. (1979). J. Cell. Physiol. 98, 515-526. Evans, H.J., Buckland. R. A,, and Pardue, M. L. (1974). Chromosoma 48, 405-426. Fakan. S. (1978). In "The Cell Nucleus" (H. Busch. ed.), Vol. 5, pp. 3-53. Academic Press, New York. Fakdn. S., and Puvion, E. (1980). Int. Rev. Cytol. 65, 255-299. Fakan, S.. Turner, G. N.. Pagano. J. S., and Hancock, R. (1972). Proc. Natl. Acad. Sci. U.S.A. 69, 2300-2305.

Fakan. S., Puvion, E., and Spohr, G. (1982). In preparation. Feldherr, C . M. (1965). J. Cell Biol. 25, 43-53. Feldherr. C. M., and Ogburn, J . A. (1980). J. CellBiol. 87, 589-593. Finch, J. T., and Klug. A. (9176). Proc. Narl. Acad. Sci. U.S.A. 73, 1897-1901. Finch, J. T., Lutter. L. C.. Rhodes, D., Brown. R. S., Rushton, B., Levitt, M., and Klug, A. (1977). Nature (London) 269, 29-36. Finch, J. T., Brown, R. S.. Rhodes, D., Richmond, T., Rushton, B., Lutter, L. C.. and Klug, A. (1981). J. M d . Biol. 145, 757-769. Franke, W. W. (1970). Z. Zellforsch. 105, 405-429. Franke. W. W. (1974). Int. Rev. Cvtol. Suppl. 4, 72-236. Franke, W. W., and Scheer, U. (1974). Svmp. Soc. Exp. Biol. 28, 249-282. Franke, W. W.. Kleinschmidt, J. A., Spring, H., Krohne, G., Grund, C., Trendelenburg. M. F.. Stoehr, M.. and Scheer. U. (198la). J. Cell B i d . 90, 289-299. Franke, W. W., Scheer, U . , Krohne, G . , and Jarasch, E. (1981b). J. Cell B i d . 91, 39s-50s. Friedman, D. L., Johnson, R. A., Zeilig. C. E., Kurz, J. B., Kumar, K. V., andGray, P. (1976). In "Cyclic Nucleotides and the Regulation of Cell Growth" (M. Abou-SaE, ed.). pp. ti7-80. Dowden, Hutchinson & Ross, Stroudsburg, Pennsylvania.

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

209

Fujisawa. T.. Imai. K., Tanaka, Y., and Ogata, K. (1979). J. Biochem. (Tokvo)85, 277-286. Furst, A., Brown, E. H., Braunstein, J. D.. and Schildkraut, C. L. (1981). Proc. Nufl. Acud. Sri. U.S.A. 78, 1023-1027. Gabrielli, F., Hancock. R., and Faber, A. J. (1981). Eur. J. Biochem. 120, 363-369. Galcheva, Z., Petrov, P., and Dessev. G. (1982). Eur. J. Cell B i d . (in press). Gallinaro, H., Lazar, E., Jacob, M., Krol, A,, and Branlant, C. (1981). Mol. B i d . Rep. 7, 31-39. Georgiev, G. P. (1974). In “The Cell Nucleus” (H. Busch, ed.), Vol. 3, pp. 67-108. Academic Press, New York. Gerace, L., and Blobel, G. (1980). Cell 19, 277-287. Gerace. L.. and Blobel, G. (1982). In preparation. Gerace. L., Blum. A , , and Blobel, G. (1978). J. Cell B i d . 79, 546-566. Goessens, G., and Lepoint, A. (1979). B i d . Cell. 35, 211-220. Goldknopf. I. L., and Busch, H. (1977). Proc. Natl. Acud. Sci. U.S.A. 74, 864-868. Gomez-Lira, M. M.. and Bode, I . (1981). FEBS Lett. 127, 228-232. Goodwin, G. H.. and Johns, E. W. (1982). In “The Biochemistry and Modem Biology of the Cell Nucleus” (L. Hnilica, ed.), Vol. 2. CRC, Cleveland, Ohio (in press). Goodwin, G. H., Mathew, C. G. P., Wright, C. A , , Venkov, C. D., and Johns. E. W. (1979). Nuclc4c Acids Rex 7, 1815-1835. Gurley, L. R., Tobey. R. A,, Walters, R. A,. Hildebrand, C. E., Hohmann. P. G.. D’Anna, 1. A., Barham, S. S., and Deaven, L. L. (1978). In “Cell Cycle Regulation” (J. R. Jeter. I. L. Cameron, G. M. Padilla, and A. M. Zimmermann, eds.), pp. 37-60. Academic Press. New York. Halleck. M. S . . and Gurley, L. R. (1980). Exp. Cell Res. 125, 377-388. Hamelin C., and Yaniv. M. (1979). Nucleic Acids Res. 7, 629-637. Hamilton, D., Yuan. R.. and Kikuchi. Y. (1981). J. Mol. Biol. 152, 163-169. Hamlin. J . L., and Biedler, 1. L. (1981). J . Cell. P h w 107, 110-114. Hancock. R., and Hugues, M. E. (1982). B i d . Cell. (in press). Hancock, R . , Faber, A. J., and Fakan, S. (1977). In “Methods in Cell Biology” (D. M. Prescott, ed.). Vol. 15, pp. 127-147. Academic Press, New York. Harris. J. R. (1978). Biochim. Biophys. Actu 515, 55-104. Hartwig, M. (1978). Acfa Biol. Med. Germ. 37, 421-432. Hect. F., Wyandt, H. E . , and Magenis, R. E. H. (1974). In ”The Cell Nucleus” (H. Busch. ed.). Vol. 2, pp. 33-121. Academic Press, New York. Heine, U. I. (1974).I n “The Cell Nucleus” (H. Busch. ed.). Vol. 3, pp, 489-536. Academic Press, New York. Heinrich, P. C., and Northemann, W. (1981). Mol. Biol. Rep. 7, 15-24. Henry S . A., Weinfeld, H., and Sandberg, A. A. (1980). Exp. Cell Res. 125, 351-362. Herlan, G.. Giese, G., and Wunderlich, F. (1979). Exp. Cell Res. 118, 305-309. Herlan, G . , Giese, G.. and Wunderlich, F. (1980). Biochemist? 19, 3960-3966. Hernandez-Verdun, D., Hubert, J., Bourgeois, C. A,. and Bouteille, M. (1980). Chromosomu 79, 349-362. Hohmann, P. (1978). Suhcell. Biochem. 5, 87- 127. Honda. B. M., Candido. E. P. M., and Dixon, G. H. (1975). J. B i d . Chem. 250, 8686-8689. Horowitz, S. B.. and Moore, L. C. (1974). J. Cell Biol. 60, 405-415. Horvath, P., and Hiirz. W. (1981). FEBS Lett. 134, 25-28. Hsieh, T.. and Brullag, D. L. (1979). Proc. Nail. Acud. Sci. U.S.A. 76, 726-730. Hsieh. T.. and Brutlag. D. L. (1980). Ce/I21, 115-125. Huberman. J . A.. Tsai. A.. and Deich, R. A . (1973). Nature (London) 241, 32-36. Hubert. J., Bouvier, D.. and Bouteille. M. (1979). Biol. Cell. 26, 87-90. 87-90. Hummeler, K . . Toniassini, N., and Sokol, F. (1970). J. Virol. 6, 87-91. Ide. T.. Nakane, M.. Anzai, K . . and Andoh. T. ( 1975). Nature (London) 258, 445-447.

2 10

RONALD HANCOCK AND TEN1 BOULIKAS

Igo-Kemenes, T., and Zachau, H. G . (1978). ColdSpring HurborSvmp. Quant. Biol. 42, 109-1 18. Igo-Kemenes, T., Horz, W., and Zachau, H. G. (1982). Annu. Rev. Eiochem. (in press). Ikeda, H., Moriya, K., and Matsumoto. T. (1980). Cold Spring Harbor Svrnp. Quanr. Eiol. 45, 399-408. Isenberg. I. (1979). Annu. Rev. Eiochem. 48, 159-191. Itkes. A. V., Glotov, B. 0.. Nikolaev, L. G., Preem, S. R., and Severin. E. S. (1980). Nucleic Acids Res. 8, 507-527. Ivanova. E., Pironcheva, G., and Djondjurov, L. (1981). Eur. J. Eiochem. 113, 569-573. Jackson. J. B., and Rill. R. L. (1981). Birichemisrry. 20, 1042-1046. Jackson, V.. Marshall, S., and Chalkley, R. (1981). Nucleic Acids Res. 9, 4563-4581. Jacob, M., Devilliers, G., Fuchs. J-P.. Gallinaro, H., Gattoni, R., Judes, C., and Sthenin, J. (1982). In "The Cell Nucleus" (H. Busch, ed.), Vol. 8, pp. 194-246. Academic Press, New York. Jaenisch. R . , and Levine, A. (1971). Virology 44, 480-493. Jahn. C. L., Hutchinson, C. A., Phillips, S. J.. Weaver. S., Haigwood, N. L.. Voliva. C. F.. and Edgell. M. H. (1980). Cell 21, 159-168. Jelinek. W., and Goldstein, L. (1973). J. Cell. Physiol. 81, 181-198. Jeppesen. P. G. N.. and Bankier. A. T. (1979). Nucleic Acids Res. 7, 49-67. Jones, R . T . , Johnson, R. T., Gupta, B. L., and Hall, T. A. (1979). J. Cell Sci. 35, 67-85. Jorcano, J . L., Meyer, G., Day, L. A,, and Renz. M. ( 1980). Proc. Nut/. Acud. Sci. U.S.A. 77, 6443-6447. Jost, E. (1982). In preparation. Jost, E.. and Johnson, R. T. (1981). J. Cell Sci. 47, 25-53. Jost, E., d'Arcy, A., and Ely, S. (1979). J. Cell Sci. 37, 97-107. Jump, D. B., Butt, T. R., and Smulson, M. (1979). Biochemistry 18, 983-990. Kajiwara, K., and Mueller, G. C. (1964). Biochim. Eiophys. Acru 91, 486-493. Kalifat, S. R . , Bouteille, M., and Delarue. J. (1967). J. Microsc. 6, 1019-1026. Kartenbeck, J.. Jarosch, E. D., and Franke, W. W. (1973). Exp. Cell Res. 81, 175- 183. Kashnig. D. M., and Kasper, C. B. (1969). J. Biol. Chem. 244, 3786-3792. Kaufmann, S. H., Coffey, D. S., and Shaper. J. H. ( 1981). Exp. Cell Res. 132, 105-123. Kendall, F., Swenson, R., Borun. T.. Rowinski, J., and Nicolini, C. (1977). Science 196, 1106-1 109. Kleinig, H. (1970). J. Cell B i d . 46, 396-402. Kleinschniidt, A. K. (1968). In "Methods in Enzymology" (S. P. Colowick and N . 0. Kaplan. eds.), Vol. 12. pp. 361-377. Academic Press, New York. Klug. A., Rhodes, D., Smith, J., Finch, J . T., and Thomas, J. 0. (1980). Nulure (London) 287, 509-5 16. Kohen. E.. Siebert, G., and Kohen, C. (1971). Hoppe-Sevler's Z. P h y i o l . Chem. 352, 937--937. Kroeger, H., Jacob, J., and Sirlin, J. L. (1963). E.rp. Cell Res. 31, 416-423. Krohne. G.. Franke, W. W., and Scheer. U. (1978). Exp. Cell Res. 116, 85-102. Krohne, G. (1982). In preparation. Kuehl, L., Lyness, T., Dixon, G. H., and Levy-W. B. (1980). J. Eiol. Chem. 255, 1090-1095. Labhart. P.. and Koller, T. (1981). Eur. J . Cell Eiol. 24, 309-316. Lam. K. S . , and Kasper, C. B. ( 1979). Biochrmistrv 18, 307-3 I I . Lepoint, A,, and Bassleer, R . (1978). Virchows Arch. Ser. E 26, 267-273. Lerner, M. R., and Steitz, J. A. (1979). Proc. Nut/. .4cud. Sci. U.S.A. 76, 5495-5499. Lerner, M. R., and Steitz, J. A. (1981). Cell 25, 298-300. Levinger, L., and Varshavsky, A. (1982). Cell 28, 375-385. Liao. L. W., and Cole, R. D. (1981). J. Biol. Chem. 256, 10124-10128. Liautard, J. P., Widada. J. S., and Brunel, C. (1981). Mol. Eiol. Rep. 7, 41-45.

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

21 1

Liskay. R . M., and Prescott. D. M. (1978). In “Cell Reproduction” (E. R. Dirksen, D. M. Prescott, and C. F. Fox, eds.). pp. 115-125. Academic Press. New York. McCarty. K. S.. and McCarty, K. S.,Jr. (1978). In “Cell Cycle Regulation” (J. R. Jeter, 1. L. Cameron, G. M. Padilla, and A. M. Zirnmerman, eds.). pp. 9-35. Academic Press. New York. McGhee. J. D., and Felsenfeld, G. (1980). Annu. Rev. Eiochem. 49, 1 115- 1156. McGhee. J . D., Rau, D. C., Charney, E., and Felsenfeld. G . (1980). Cell 22, 87-96. McGhee, J. D., Wood. W. I.. Dolan. M., Engel. J. D.. and Felsenfeld. G. (1981). Cell27,45-55. McKnight, S. L., and Miller, 0. L. (1979). Cell 17, 551-563. McKnight, S. L., Bustin. M., and Miller, 0. J. (1978). ColdSpring HarborSvmp. Quant. Eiol. 42, 741-754. McReady, S. J.. Akrigg. A,. and Cook, P. R. (1979). J. Cell Sci. 39, 53-62. Malcolm. D. B., and Sommerville, J. (1974). Chrornosoma 48, 137-158. Marsden. M. P. F., and Laemmli. U. K. (1979). Cell 17, 849-858. Mathew, C. G. P.. Goodwin. G. H.. and Johns, E. W., (1979). Nucleic Acids Res. 6, 167-179. Mathew. C. G. P.. Goodwin, G. H.. Igo-Kemenes, T., and Johns, E. W. (1981). FEESLett. 125, 25-29. Mathis. D.. Oudet. P.. and Charnbon, P. (1980). f r o g . Nucleic Acid Res. Mol. Eiol. 24, 1-55. Matsui, S . I.. Seon, B. K.,and Sandberg, A. (1979).Proc. Natl. Acud. Sci. U.S.A. 76,6386-6390. Matsumoto. L. H. (1981). Nature (London) 294, 481-482. Mattern. C. F. T., Takernoto, K. K., and Daniel, W. A. (1966). Virolop 30, 252-256. Matthews, H. R. (1977). In “The Organisation and Expression of the Eukaryotic Genome” (E. M. Bradbury and K. Javaherian, eds.), pp. 67-80. Academic Press, New York. Maul, G. G . (1977a). J . Cell Eiol. 74, 492-500. Maul, G. G . (1977b). I n / . Rev. Cvtol. Suppl. 6. 76-186. Maul. G. G.. and Avdalovic, N. (1980). Exp. Cell Res. 130, 229-240. Maul, G. G . , Deaven, L. L., Freed, J . J . , Campbell, G. L. M..and Becak, W. (1980). Cvtogenet. Cell Genet. 26, 175- 190. Maundrell, K., and Scheyer, K. (1979). Eur. J. Eiochem. 99, 225-238. Maundrell, K.. Maxwell, E. S., Puvion, E., and Schemer, K. (1981).Exp. Cell Res. 136,435-445. Miller, 0. J . , Miller, D. A,. Dev. V. G . , Tantravahi, R.. and Croce, C. M. (1976). Proc. Natl. Acad. Sri. U.S.A. 73, 4531-4535. Miller, T. E., Huang. C.. and Pogo, A. 0. (1978). J. Cell Eiol. 76, 675-691. Mirzabekov. A. D. (1980). Q . Re,,. Eiophvs. 13, 255-295. Moens. P. B. (1969). Chrornosoma 28, 1-25. Monneron. A,. and Bernhard, W. (1969). J. Ultrastrucr. Res. 27, 266-288. Morgan. C.. Rosenkranz. H. S., and Mednis, B. (1969). J. Virol. 4, 777-796. Moroi, Y., Hartman, A. L., Nakane, P. K., and Tan, E. M. (1981). J. Cell Eiol. 90, 254-259. Moser, G . C.. Mueller, H.. and Robbins, E. (1975). Exp. Cell Res. 91, 73-78. Murcia, G., and Koller. T. (1981). Eiol. Cell. 40, 165-174. Murphy, R. F., Wallace, R. B.. and Bonner, J. (1980). Proc. Natl. Acud. Sci. U.S.A. 73, 5903-5907, Murray, A . B.. and Davies, H. G . (1979). J. Cell Sci. 35, 59-66. Musich, P. R.. Brown, F. L., and Maio, J. J. (1977). Cold Spring Harbor Sympt. Quant. Eiol. 42, 1147-1 160. Musich. P. R . , Brown, F. L., and Maio, J. J. (1982). Proc. Natl. Acad. Sci. U.S.A. 79, 118-122. Nash, H. A., Mizuuchi. K., Weisberg, R. A , , Kiluchi, Y., and Gellert, M. (9177). In “DNA Insertion Elements. Plasmids and Episomes” (A. 1. Bukhari, J. A. Shapiro, and S . L. Adhya. eds.). pp. 263-273. Cold Spring Harbor Lab., Cold Spring Harbor, New York. Nash, N. A., Mizuuchi. K., Enquist, L. W., and Weisberg. R . A. (1980). Cold Spring Harbor Svmp. Quanr. B i d . 45, 417-428.

212

RONALD HANCOCK AND TEN1 BOULIKAS

Nelkin, B. D., Pardoll, D. M., and Vogelstein, B. (1980. Nucleic Acids Res. 8, 5623-5633. Newrock, K. M.. Alfageme, C. R.. Nardi. R. V., and Cohen, L. H. (1978). Cold Spring Harbor Svmp. Quant. Eiol. 42, 42 1-43 I . Nicolau. C.. Hildenbrand, K., and Johnson, S. M. (1978). In “Virus-transformed Cell Membranes” (C. Nicolau, ed.). pp. I 11-183. Academic Press. New York. Niessing, J.. and Sekeris, C. E. (1970). Eiochim. Eiophvs. Actu 209, 484-492. Niessing, J., and Sekeris, C. E. (1972). FEES Let/. 22, 83-88. Nordheim, A., Pardue, M. L.. Lafer, E. M.. Mlller, A,. Stollar, B. D., and Rich, A. (9181). Nature (London) 294, 417-422. Notbohm. H., Hollandt, H., and Meissner, J. (1979). Int. J . Eiol. Marrornol. 1, 180-184. Olins, A. L., and Olins, D. E. (1979). J . Cell B i d . 81, 260-265. Olins. A. L., Olins, D. E., Zentgraf, H.. and Franke. W.W .(1980). J . Cell Eiol. 87, 833-836. O’Malley, B. W.,Sherman, M. R., and Toft, D. 0. (1970). Proc. Natl. Acad. Sci. U . S . A . 67, 50 1 -508. Paine. P. L., Moore, L. C.. and Horowitz, S . B. (1975). Nature (London) 254, 109-1 14. Pardoll. D. M.. and Vogelstein, B. (1980). Exp. Cell Res. 128, 466-470. Pardoll. D. M., Vogelstein, B., and Coffey, D. S. (1980). Cell 19, 527-536. Patrizi, G., and Poger, M. (1967). J. Ultrastruct. Res. 17, 127-136. Pederson, T.. and Davis. N. G. (1980). J. Cell Eiol. 87, 47-54. Pehrson, J., and Cole, R. D. (1980). Nature (London) 285, 43-44. Peterson, S. P., and Berns, M. W . (1978). J . Cell Sci 32, 197-213. Pool. T. B.. Cameron, 1. L., Smith. N. K. R., and Sparks, R. L. (1981). In “The Transformed Cell” (I. L. Cameron and T. B. Pool, eds.), pp. 398-420. Academic Press, New York. Preobrazhensky, A. A., and Spirin, A. S. (1978). frog. Nucleic Acid Res. Mol. B i d . 21, 1-38. Prescott. D. M. (9176a). Adv. Genet. 18, 100-177. Prescott, D. M. (1976b). “Reproduction of Eukaryotic Cells.” Academic Press, New York. Presrayko. A. W.,Tonato, M., and Busch. H. (1970). J. Mol. Eiol. 47, 505-515. Pruitt, S. C., and Grainger, R. M. (1980). Chromosoma 78, 257-274. Puvion-Dutilleul, F., and Puvion. E. (1980). J . Cell Sci. 42, 305-321. Puvion-Dutilleul, F., Laithier, M.. and Puvion E. (1981). Eur. J . Cell Eiol. 25, 233-241. Puvion, E., and Moyne, G . (1982). In “The Cell Nucleus” (H. Busch, ed.), Vol. 8, pp. 5‘3-1 15. Academic Press, New York. Quinlan, T. J., Billings, P. B., and Martin, T. E. (1974). Proc. Not/. Acad. Sci. U.S..4. 71, 2632-2636. Quinlan, T. J., Kinniburgh, A. J., and Martin, T. E. (1977). J. Eiol. Chem. 252. pp. 1156-1 161. Rae, P. M. M., and Franke, W.W.(9172). Cltromosoma 39, 443-456. Rao, N., and Sunkara. P. S. (1978). In “Cell Cycle Regulation” (J. R. Jeter, 1. L. Cameron, G . M. Padilla, and A. M. Zimmennann, eds.)., pp. 133-147. Academic Press, New York. Rattner, J. B., and Hamkalo, B. A. (1979). Chromosoma 81, 453-457. Razin, S. V.. Mantieva, V. L., and Georgiev, G. P. (1979). Nucleic Acids Res. 24, 1713--1735. Reddy, R.. and Busch, H. (1982). In “The Cell Nucleus” (H. Busch, ed.). Vol. 8, pp. 261-306. Academic Press. Renz, M.. Nehls, P., and Hozier, J. (1977). Proc. Natl. Acad. Sci. U.S.A. 74, 1879-1883. Reudelhuber, T. L., Boulikas, T., and Garrard, W.T. (1980). J . B i d . Chem. 255, 4511--4515. Richardson, J. C. W.,and Maddy, A. H. (I980a). J . Cell Sci. 43, 253-267. Richardson, J. C. W.,and Maddy, A. H. (1980b). J . Cell Sci. 43, 269-277. Roy, R. K., Lau, A. S., Nunro, H. N., Baliga, B. S., and Sarkar, S. (1979). Proc. Natl. Acad. Sci. U.S.A. 76, 1751-1755. Royal. A., Garapin. A., Cami, B., Perrin, F., Mandel, J. L., LeMeur, M., Brigigtgre, F., Cannon, F.. LePennec, 1. P., Chambon, P., and Kourilsky, P. (1979). Nature (London) 279, 125-132.

FUNCTIONAL ORGANIZATION IN THE NUCLEUS

213

Ruiz-Carrillo. A , . Wangh. L. J . , and Allfrey, V. G. (1975). Science 190, 117-128. Russev. G.. and Hancock. R. (1981). Nucleic Acids Re.\. 9, 4129-4137. Rykowski, M. C.. Wallis, J . W., Choe, J . , and Grunstein. M. (1981). Cell 25, 477-487. Samal. B.. Worcel, A., Louis, C., and Schedl. P. (1981). Cell 23, 401-409. Saniarina, 0. P.. Lukanidin, E. M., Molnar. I . . and Georgiev. G . P. (1968). J. Mol. Biol. 33, 25 1-263. Samuel. C.. Mackie. J., and Sommerville, J . (1981). Chromosoma 83, 481-492. Savic. A , . Richnian. P.. Williamson. P., and Poccia, D. (1981). Proc. Natl. Acud. Sci. U.S.A. 78, 3706-3710. Scheer. U.. and Franke. W. W. (1969).J . CellBiol. 42, 519-533. Scheer. U . . Kartenbeck, J.. Trendelenburg, M. F., Stadler, J.. and Frdnke. W. W. (1976). J. Cell B i d . 69, I - 18. Schellens. J. P. M.. James, I . . and Hoeben, K . A. (1979). B i d . Cell. 35, I I - 14. Schwartz. H.. and Darncll. J. E. (1976). J. M o l . Biol. 104, 833-851. Sealc. R. L. (1981). Nuclei(,Acids Rrs. 9, 3151-3158. Scale. R., and Aaronson, A. L. (1973). J. Mol. B i d . 75, 647-658. Setterfield, G.. Sheinin. R . , Dardick, I., Kiss, G . , and Dubsky, M. ( I 1978). J. Cell Biol. 77, 246-263. Seyedin. S . M.. and Cole, R . D. (1981). J. B i d . Chem. 256, 442-444. Shaw. B. R.. Cognctti. G . , Sholes, W. M., and Richards. R . G. (1981). Biochemistry 20, 497 I -4978. Sheinin. R.. and Lewis. P. N. (1980). Yomatic Cell Genet. 6, 225-239. Shelton, K. R.. Egle, P., and Cochran, D. L. (1981). In “The Nuclear Envelope and the Nuclear Matrix” (G. G. Maul, ed.). Wistar Symposium Series Vol. 2. Liss, New York (in press). Sieber-Blum, M., and Burger. M. M. (1977). Biochem. Biophys. Res. Commun. 74, 1-8. Siebert. G . , Langendorf. H.. Hannover. R . , Nitz-Lifzow. D.. Pressman. B. C., and Moore. C. (1965). Hoppe-Seylers Z. Physiol. Chem. 343, 101-1 15. Sikstrom. R . . Lanoix, J.. and Bergeron, J. J. M. (1976). Biochim. Biophjrs. Actu 448, 88-102. Simmons. T., Henry. S.. and Hodge, L. D. (1980). I n “Nuclear-Cytoplasmic Interactions in the Cell Cycle” (G. L. Shitson. ed.). pp. 57-103. Academic Press, New York. Simpson, R. T. (1978a). Cell 13, 691-699. Simpson, R. T. (1978b). Biochemisrry 17, 5524-5531. Simpson. R. T. (1981). Proc. Natl. Acad. Sci. U.S.A. 78, 6803-6807. Skaer. R . J.. and Whytock. S. (1975). J. Cell Sci. 19, 1-10. Skaer, R. J., Whytock. S.. and Emmines. J. P. (1976). J. Cell Sci. 21, 470-496. Slonimski. P. P. (1980). C. R. A i d . Sci. Paris 290, 331-333. Smith, J . D. (1980). Intervirology 13, 312-316. Sperling, L., and Tardieu. A. (1976). FEBS Lett. 64, 89-91. Stacey. D. W. (1980). In “Introduction of Macromolecules into Viable Mammalian Cells” (R. Baserga, C. Croce. and G. Rovera, eds.), pp. 125-134. Liss, New York. Stackpole. C. W. (1969). J. Virol. 4, 75-93. Stambrook, P. J., (1974). J. Mol. B i d . 82, 303-313. Stambrook, P. J. (1981). Biochemistry 20, 4393-4398. Steer, R. C., Wilson, M. J., and Ahmed, K. (1979). Biochem. Biophys. Res. Cummurc. 89, 1082- 1087. Stelly. N.. Stevens. B. J.. and Andre. J. (1970). J. Microsc. 9, 1015-1028. Stevenin, J., Gattoni, R., Devilliers, G., and Jacob, M. (1979). Eur. J . Biochem. 95, 593-606. Stevens, B. J.. and Swift, H. (1966). J. Cell Biol. 31, 55-77, Stick, R., and Krohne. G . (1982). Exp. Cell Res. 138, 319-330. Stick, R., and Schwartz, H. (1982). Cell D i r e r . (in press).

214

RONALD HANCOCK AND TEN1 BOULIKAS

Stratling, W. H., Miiller, U., and Zentgraf, H. (1978). Exp. Cell Res. 117, 301-31 I . Suau, P., Kneale, G. G., Braddock, G. W., Baldwin, J. P., and Bradbury, E. M. (1977). h’ucleic Acids Res. 4, 3769-3786. Suau, P., Bradbury, E. M., and Baldwin, J. P. (1979). Eur. J . Biochem. 97, 593-602. Sundin, 0.. and Varshavsky. A. (1980). Cell 21, 103-1 14. Sundin, O., and Varshavsky, A. (1981). Cell 25, 659-669. Takahashi, K., and Tashiro, Y. (1979). Eur. J . Biochern. 97, 353-360. Thoma, F., and Koller, T. (1981). J. Mol. Biol. 149, 709-733. Thoma. F., Koller. T., and Klug, A. (1979). J . Cell B i d . 83, 403-427. Thomas, J. O., Raziuddin, Sobota, A,. Boulik, M.. and Szer. W. (1981). Proc. Natl. Acad. Sci. U.S.A. 78, 2888-2892. Tobey, R. A,. Gurley. L. R., Hildebrand, C. E., Ratliff. R . L., and Walters. R. A. (1974). In “Control of Proliferation in Animal Cells” ( B . Clarkson and R. Basaga. eds.). pp. 665-679. Cold Spring Harbor Lab., Cold Spring Harbor. New York. Todd, R. D.. and Garrard. W. T. (1977). J. B i d . Chem. 252, 4729-4738. Todorov. I. T.. and Hadjiolov. A. A. (1979). Cell B i d . Int. Rep. 3, 753-757. Urban, M. K.. and Zweidler, A. (1982). Dev. B i d . (submitted). Vigneri, R., Goldfine, I . D., Wong. K . Y.. Smith, G. J., and Pezzin. V. (1978).J. Biol. Chem. 253, 2098-2103. Villarreal. L. (1981). Virology 113, 663-671. Vorbrodt, A,, and Maul. G. G. (1980). J. Hisrochem. Cvtochem. 28, 27-35. Werner, D.. and Petzelt. C. (1981). J. Mnl. Biol. 150, 297-302. Whitlock, J. P., Jr., Augustine, R., and Schulman. H. (1980). Nature (London) 287, 74-76. Wise. G. E., and Prescott, D. M. (1973). Proc. N d l . Acud. Sci. U.S.A. 70, 714-717. Wolff, S. (1977). Annu. Rev. Genet. 11, 183-201. Wolgemuth, D. J . , and Hsu M. T. (1981). J. Mol. Biol. 147, 247-268. Wu, R. S., and Bonner, W. M. (1981). Cell 27, 321-330. Yankner. B. C., and Schooter, E. M. (1979). Proc. Nurl. Acad. Sci. U.S.A. 76, 1269-1273. Yannarell. A,, Schumm, D.. and Webb, T. E. (1976). Biochem. J. 154, 379-385. Yasuda, H.. Matsumoto, Y.. Mita. S.. Marunouchi, T., and Yamada, M. (1981). Biochemis/rv 20, 44 14-441 9. Yasuzumi, G . , and Tsubo. J . (1966). E.rp. Cell Res. 43, 281292. Yoshida, S . , Ungers, G.. and Rosenbcrg, B. H. (1977). Nucleic Acids Res. 4, 223-228. Zbarsky, I. B. (1981). Mol. B i d . Rep. 7, 139-148. Zentgraf, H. W.. Muller. U . , and Franke, W. W. (1980). Eur. J. Cell B i d . 23, 171- 188. Zieve, G. W. (1981). Cell 25, 296-297. Zuckerkandl, E. (1981). Mol. Biol. Rep. 7, 149-158. Zweidler, A. (1980). In “Gene Families of Collagen and Other Structural Proteins” (D. J. Prockop and P. C. Champec, eds.), pp. 47-56. Elsevier, Amsterdam. Zweidler, A,, and Urban, M. K. (1982). Biochemistry (submitted).

INTERNATIONAL REVIEW OF CYTOLOGY. VOL 7Y

The Relation of Programmed Cell Death to Development and Reproduction: Comparative Studies and an Attempt at Classification JACQUES BEAULATON A N D RICHARD A . LOCKSHIN Drpartmrrit de Zoologie, Universitc. de Clrrmonr. Aubiere, France, and Department o] Biologicul Sciences, St. John’s University. Jamaica. New York I. Introduction . . . . . . . . . . . . . . . . . .................. 11. Modes of Cell Death, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....... A. Diffuse Degeneration. . . . . . . . . . . . . . . .

215 216 216

B. Autophagic Degeneration . . C. Degeneration by Primary or Precocious Pycnosis . 111. Means of Elimination of Degener A. Extrusion into the Extracellular Milieu. . . . . . . . . . . . . . . . . . . . B. Elimination by Phagocytes . . . . ................ I V . Determination of Degeneration by Cellular Interaction. . . . . . . . . . . A. Signals Derived from Cellular Interactions . . . . . . . . . . . . . . . . B. Control Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

225 226 227 227 229 23 I

I. Introduction Cell death is a widespread phenomenon. As first emphasized by Glucksmann (195 I ) , not only is it a major factor in the morphogenesis of mature amphibia and insects, it is also a major factor in the formation of animal and plant shape, the sculpting of extremities in vertebrates and invertebrates, the formation of divided leaves or sculpted insect wings, the fusion of the soft palate, the regulation of motor and sympathetic nerves, and the differentiation of reproductive organs. All of these involve massive, coordinated cell death. There are in addition numerous instances in which individual cells rather than large patches die, and these may represent part of the same phenomenon. In adult life, adjustment of the sexual organs to seasonal reproduction involves cell death, and several ancillary tissues likewise respond. Although few of these systems have actually been tested, those which have, have been shown to display a form of death known as programmed cell death-a developmental event, engendered by extracellular signals, in which the affected cell responds by a series of reactions which culminate, hours to days later, in the collapse of the cell. Recent experiments have led to a greater understanding of that programming, as is discussed in the final section. Since the story is just 215 Copyright 4 ’ IYX2 hy Academic I’rew. Inc. All nphh of ruproduction in any lomi re\erved. ISBN 0-12-364479-8

216

JACQUES BEAULATON AND RICHARD A. LOCKSHIN

beginning to unfold, it is appropriate first to review the occurrence and appearance of cell death. Our understanding of the mechanisms of which programmed cell death is brought about is, in comparison to the knowledge of other genetically controlled developmental processes, still rudimentary. We know that cell death occurs normally at different stages of morphogenesis, growth, and development of metazoans and multicellular plants; it may even occur in unicellular organisms, although identification is less certain. It is both banal and universal in metazoans, and it assures, in the diverse cellular populations of a healthy organism, a mechanism complementary to mitotic division (Ken er a / . , 1972). By programming this degeneration, the organism maintains the capacity to control the proliferation of cell populations. Thus, during embryonic development, cell death plays an essential regulatory role in morphogenesis (morpho- or histogenetic degeneration, according to Glucksmann, 1951 ; see also the reviews by Saunders, 1966, and Forsberg and Kallen, 1968) and in the disappearance or atrophy of various organs, members, or appendices (phylogenetic degeneration according to Glucksmann, 1951). The concept of programmed cell death is founded on the study of a major morphogenetic mechanism-remolding. From these models the idea was extended to other situations related to senescence and pathology (Lockshirr and Beaulaton, 1974). The concept assumes the initiation of a sequence of irreversible steps. During the early phases, judicious experimental intervention can reverse processes over which the animal would normally have no control [blocking protein synthesis (Weber, 1964; Lockshin, 1969) or transplanting tissue to another location-Saunders er al., 1962)l. In the following discussion we will examine not only the diverse modes of degeneration and of elimination of degradation products, but also the role of cellular interactions in the mechanisms of control of cell death.

11. Modes of Cell Death

Diverse types of cell degeneration have been described in examples ranging from ontogenesis through postembryonic development to reproduction. Without attempting to force the data, it would seem that one could, from cytocheinical and ultrastructural information, distinguish three principal modes of degeneration: diffuse degeneration, autophagic degeneration, and degeneration by pycnosis. A11 of these can be accompanied by cellular fragmentation. A. DIFFUSEDEGENERATION This mode of cellular degeneration (Figs. 1 and 2) is characterized first of all by a retraction of the plasma membrane which isolates the injured cell from

PROGRAMMED CELL DEATH

217

FIG. I . Diffuse degeneration of two hemocytoblasts (or prohemocyte-arrow) in the hemocytopoietic organ of the silkworm Antheraeu pernvi at prepupal stage. This photomounting at low magnification shows on the left the normal appearance of three hemocytoblasts (H). Note by comparison the loss of osmiophilia in the cytoplasm of degenerating hemocytoblasts and the swollen aspect of the nucleus (N). Basement membrane (B). X7500.

neighboring cells (Stephan-Dubois et a / ., 1974; Andries, 1975, 1976; Beaumont, 1977). The retraction of the cell membrane probably derives from a decrease in activity of ion pumps leading to a redistribution of ions and water on the two sides of plasma membranes and from changes in cytoskeleton associated with intracellular Ca levels controlling cell shape. These alterations are seen primarily in the membranes of the mitochondria, the endoplasmic reticulum near the nuclear envelope, and the Golgi apparatus, all of which develop an irregular structure accompanied by a dilation of the intermembrane spaces (Fig. 2 ) or the areas within the reticulum reflecting changes in ion and water transport and most probably loss of K ions (Trump et a/., 1981). There are also condensation of ribosomes or reduction in the proportional surface area which they occupy (Stephan-Dubois et a / . , 1974; Andries, 1975). Degeneration continues with a karyolysis (Figs. 1 and 2) or edema of the nucleus (Policard and Bessis, 1968)-

218

JACQUES BEAULATON AND RICHARD A. LOCKSHIN

FIG. 2. Diffuse degeneration of a granulocyte in an interstitial space of the hemocytopoietic organ in the silkworm Anrheraea pernvi at prepupal stage. Note the swelling of the cytoplasmic organelles: mitochondria (M).endoplasmic reticulum (E), nuclear envelope (arrows). Nucleus ( N ) in process of karyolysis. X96M).

that is, a loss of osmiophilia in the nucleus accompanied by the loss of basophilia and by the dispersion of chromatin to the periphery of the nucleus. This edematous swelling of the cytoplasmic organelles continues, terminating in a generalized vacuolization. In this type of cell death, the degradation never results from an autophagic activity. The origin of the hydrolases involved in this phenomenon is not clear, if indeed hydrolysis takes place within the cell rather than the cell losing its materials by osmotic lysis. This is probably the case in a type of insect hemocyte, the cystocyte or coagulocyte, pertaining to the granular cell class, which quickly disintegrates during plasma coagulation (Jones, 1975; Goffinet and Gregoire, 1975; Rowley and Ratcliffe, 1976; Rowley and Ratcliffe, 1976). According to some authors digestion of the materials is carried out by extralysosomal hydrolases (Fox, 1973b; Hourdry, 1974; Beaumont and Hourdry, 1976) while according to others hydrolases escaping by diffusion from primary and secondary lysosomes following lysosomal swelling associated with changes in membrane permeability digest the tissue constituents (Stephan-Dubois et a / ., 1974; Trump et a / . , 1981) or by lysosomal exocytosis of the killer T-lymphocytes during the lethal hit given to the target cells (Nicolas and Zagury, 1980). This type of involution named necrosis or cytolysis has been seen predomi-

PROGRAMMED CELL DEATH

219

nantly and dramatically in prenatal tissues such as cartilage cells (Schweichel and Merker, 1973) and in numerous tissues of metamorphosing tadpoles and insects, but also in different cells subjected to lethal injuries (anoxia, ischemia acting upon mitochondria1 oxidative phosphorylation, various environmental and physical changes, effect of membrane-damaging agents acting upon permeability, and transport processes). This type of involution has been seen primarily in ectodermal and endodermal cells, perhaps because these cells, exposed to the external milieu, are rapidly subject to osmotic damage if their energy pumps or membrane integrity are compromised.

B. AUTOPHAGIC DEGENERATION In this mode of degeneration (second type of necrosis according to Schweichel and Merker, 1973) cellular lysis results first of all in a massive autophagy (Figs. 3-5), ending any possibility of a reversal of the situation. Thus it is different from local autophagic manifestations which are more tightly linked to normal cytoplasmic turnover. As in the case of diffuse degeneration, the first alterations are recognized at the level of the plasma membrane. In epithelial cells, microvilli (Salzgeber and Weber, 1966) or junctional complexes (Fox, 1977) are lost. In other cells such as the granulosa cells of ovarian follicles undergoing atresia the first alterations are the blebbing of cytoplasm which is related to the loss of hormonal receptors (Peluso et a/. , 1980). Simultaneously, several organelles undergo a dilation: mitochrondria (Stephan-Dubois et a/., 1974; Scheib, 1977; Hourdry, I977), or endoplasmic reticulum (Hourdry, 1977) which is initially reversible during cell injury before the “point of no return” (Trump e t a / ., 1981) morphologically reflected by a high amplitude swelling (stage 4 according to these authors). Again, these manifestations suggest osmotic disequilibria at least within the cell organelles. The activity of the Golgi apparatus, often noted, appears to be related to the rapid development of primary lysosomes. Islands of cytoplasm are sequestered by membranes deriving from the Golgi or endoplasmic reticulum (Beaulaton, 1967; Locke and Sykes, 1975; Beaulaton and Lockshin, 1977) which form the envelope of the autophagic vacuoles. The process of dedifferentiation continues along a well-known sequence to the progressive degradation of these cytoplasmic isolates by the acquired acid hydrolases. The enzymes are acquired either directly from the sequestration envelope (Beaulaton, 1967; Hourdry, 1977) or by the incorporation of primary or secondary lysosomes (Hourdry, 1971a, 1977). If one considers the autophagic activity to be a cytoplasmic defense reaction on the part of the cell (De Duve and Wattiaux, 1966; Hourdry, 1971b; Beaumont and Hourdry, 1976; Beaumont, 1977; Hurle and Hinchliffe, 1978) deriving from a primary alteration of one or more organelles (Helminen and Ericsson, 1968; Scheib, 1977) then the death of

220

JACQUES BEAULATON AND RICHARD A. LOCKSHIN

FIG.3. Autophagic degeneration in an intersegmental muscle of the silkworm Antheruea p d v phemus 20 hours after emergence. Characterization of acid phosphatase activity. Low magnification of a fiber dramatically subdivided and invaded by numerous tracheoles (T). In the sarcoplasm, note the complete erosion of the myofilaments and the predominant organelles which are autophagic vacuoles demonstrating acid phosphatase activity (arrows) and liposomes (L). X9000.

the cell might arise as follows: above a certain threshold (such as a mitochondria1 population incapable of maintaining a sufficient level of energy in the cell) the sequence of defensive hydrolysis becomes irreversible and progresses to the death of the cell, passing through a phase of secondary pycnosis of the nucleus (Figs. 6 and 7-Peluso et al., 1980; Bowen, 1981; Hinchliffe, 1981; Lockshin, 1981 a,b; Trump et al., 1981 ; Wyllie, 198I). This ultimate phase of dedifferen-

PROGRAMMED CELL DEATH

22 1

FIGS.4 and 5 . Autophagic degeneration in the intersegmental muscles of the silkworm Anrheruea po/vphemus after treatment with chloroquine followed, 30 hours after emergence. by incubation for acid phosphatase activity. Figure 4.At low magnification, the fiber displays an accumulation of large autophagic vacuoles (large arrows) without cytochemically revealed acid phosphatase activity. These autophagic vacuoles are scattered among swollen mitochondria (M) and numerous vesicles derived from the endoplasmic reticulum, the Golgi apparatus, and other cytoplasmic organelles. The lysosomotropic drug chloroquine causes the retention of the autophagic vacuoles. Nucleus (N) with clumped chromatin. X9000. Figure 5. Peripheral region of the cytoplasm showing various stages of the autophagic vacuoles in which the newly formed sequestration envelope exhibits acid phosphatase activity (arrows). X 15,000.

222

JACQUES BEAULATON AND RICHARD A. LOCKSHIN

FIGS.6 and 7. Later stage in the autophagic degeneration of intersegmental muscles in the silkworm Anrherueu po/yphemus (41 hours after emergence). Figure 6. Residual fiber illustrating the extreme reduction of cytoplasm and the pycnotic nuclei (N). X9000. Figure 7. Dying cell which is highly reduced to a fragmented nucleus (N) surrounded by a cytoplasmic lamina. The view suggests an apoptotic mechanism of degradation. In a later phase these apoptotic bodies will be released into the hemolymph. x9000.

tiation results in cell death and appears to be closely related to the next mode of cell death. It may differ only quantitatively from metabolic regulation in the cell. According to some authors the lysosomal hydrolases can leak into the cytosol where they digest the cell structures (Jurand and Pavan, 1975; Bowen and Lewis, 1980). It is found in numerous glandular tissues, primarily those of ectodermal origin (Schweichel and Merker, 1973) but also in mesodermal cells (Pautou and Kieny, 1971; Kieny and Sengel, 1974) and in muscle tissues (Beaulaton and Lockshin, 1973, 1977; Lockshin, 1980; Cruz-Landim and Silva de Moraes, 1979; Moe er af., 1979; Johnson, 1980).

PROGRAMMED CELL DEATH

223

c. DEGENERATION BY PRIMARY OR PRECOCIOUS PYCNOSIS This other mode of degeneration, also described under the terms “apoptosis” and “shrinkage necrosis,” is apparently as common as the preceding ones, and can be characterized by a cytoplasmic and nuclear condensation (Manasek, 1969; Hammar and Mottet, 1971; Kerr, 1971; Kerr et al., 1972, 1974; Pannese et al., 1976). As in the two preceding types of cell death, the process frequently begins by an isolation of the affected cell from its neighbors. Nuclear alterations occur early and are identified by a condensation of the chromatin and of the nucleolus, both manifestations reflecting an inactivation of transcription mechanisms (Littau et al., 1964; La Pushin and de Harven, 1971; Lockshin et al., 1981). However, the first stage of chromatin clumping is considered reversible during cell injury (Trump et al., 1981). The metabolic activity of the cell is therefore probably much decreased. The condensed appearance of the cytoplasm perhaps results from a water leakage as well as a degradation of the endoplasmic reticulum, releasing a large percentage of the ribosomes, together with a disaggregation of polysomes and an accumulation of hyaloplasm-material such as proteins. Progress of the condensation mechanism can lead to a very compacted cytoplasm (Figs. 8 and 9). In a few types of cells, a crystallization of free ribosomes has been seen and also appears to reflect a cessation of protein synthesis (Bellairs, 1961; Mottet and Hammar, 1972; Morimoto et al., 1972; Ojeda and Hurle, 1975; Pannese er al., 1976). In this type of cell death, primarily embryological (the first type of necrosis described in prenatal tissues such as nervous system by Schweichel and Merker, 1973), failure of synthetic machinery appears to be a triggering force. As is discussed in the last section, failure of the synthetic machinery may also play a role in nonembryonic cell death.

111. Means of Elimination of Degenerate Cells

Degeneration of the cell can be followed by fragmentation of the cytoplasm, as is the case especially in pycnotic cell death. The process of fragmentation, otherwise known as clasmatosis (see review of Policard and Bessis, 1968) or apoptosis (Kerr et al., 1972, 1974; Wyllie et al., 1980), ends in the formation of buds or droplets derived from the cells. These clasmatic or apoptotic bodies, known to pathologists as Councilman bodies, minisegregants, caryolytic bodies, or sarcolytes (for discussion on nomenclature see Wyllie er a/., 1980) can contain both regions of apparently intact cytoplasm (Seligman et al., 1975), and fragments of nuclei which themselves contain highly condensed chromatin (Kerr et al., 1974). Fragments which contain nuclei are considered to be cases of karyorrhexis. In all the models studied so far, this type of degeneration appears to be totally independent of the lysosomal system (Wyllie et a[. , 1980).

224

JACQUES BEAULATON AND RICHARD A. LOCKSHIN

FIGS.8 and 9. Electron microscope autoradiograms of the male accessory glands in Drosophila mekunogasrer 5 days after emergence at end of 2 hours labeling with [3H]leucine. Figure 8. Low magnification illustrating a dying secondary secretory cell (SC 11) surrounded by normal primary secretory cells (SC I). Label is predominantly located over normal cells. The dying cell displays an extreme condensation of the residual cytoplasm (arrows) which is dispersed by a giant secretory vacuole containing tubule bundles in a clear matrix. Residual secretory granules (*). Nucleus (N) with condensed chromatin. X 3600. Figure 9. Higher magnification of the peripheral region illustrating condensation of the cytoplasm which is reduced to a very pronounced Condensation of v,:sicles (arrows). X 25.000. (Unpublished micrographs from a study by Claude Perrin-Waldemer.)

The cellular material which is in the process of being degraded or which has been degraded is removed from the sites of necrosis, either by in situ phagocytosis, or by extrusion into the extracellular milieu (see Beaumont and Hourdry, 1976; Beaumont, 1977). In the latter instance, extrusion can be followed by

PROGRAMMED CELL DEATH

225

phagocytosis by specialized cells if the extracellular medium is still within the body (i.e., not extrusion into intestinal lumen or outward from the epidermis). A. EXTRUSION INTO THE EXTRACELLULAR MILIEU Extrusion of degraded cells or cellular debris can be into external milieu in the case of epidermis or exocrine glands, or organs in direct communication with these regions. For other tissues, elimination is achieved into the internal milieu. 1 . Elimination into the External Milieu

This type of expulsion has been observed in the reorganization of body structures during metamorphosis of anuran amphibians, in the most exterior epidermal cells (Fox, 1972a-c, 1974), pancreatic canals and ducts (Beaumont, 1973), and the intestinal lumen (Hourdry and Dauca, 1977). In mammals, this transfer of cellular debris frequently occurs into the lumen of various glands such as mammary glands (review by Beaumont and Hourdry, 1976; Hourdry and Dauca, 1977). In insects, this type of expulsion is known, for example, in the lumen of the mesenteron in which degenerate cells are found after each molt, or only at metamorphosis (see review by Andries, 1976). In this case, however, an acid phosphatase activity is localized at the level of intercellular spaces in the region of degenerating cells, or at the level of pericellular spaces of the basal region (Bautz, 1975, 1979). According to some authors, this hydrolytic activity (or other hydrolytic activity not revealed by the cytochemical technique) could be involved in the process of rejection by facilitating the expulsion of the damaged cells (Mayer-Criel, 1971; Andries, 1976). The degradation of these cells can continue in the intestinal lumen where healthy newly formed cells of the intestinal epithelium provide the enzymes for their digestion (Andries, 1976). 2. Elimination into the Internal Milieu This mode of elimination is frequently encountered in invertebrates. Thus, in molluscs during sexual inversion, the myocytes of the penis, having attained a stage of subtotal lysis, are ejected into the intercellular spaces where cells presumably derived from a hematocytic (granulocytic) line undertake the phagocytosis of cellular fragments (Streiff and Le Gall, 1976). In insects, a similar type of elimination is seen in the hemocoele, in which apoptotic bodies of diverse origins are found during periods of pupal or imaginal molts. Similar bodies have also been described in Lepidoptera in the alary lacunae, where they are phagocytosed by macrophages (Lavenseau and Surleve-Bazeille, 1976). In the Dipteran Lucilia plaque-like bodies rich in microtubules, present in the hemolymph at imaginal emergence, derive not from hemocytes but from alary epidermal cells which have degenerated by apoptosis (Seligman et a l . , 1975).

226

JACQUES BEAULATON AND RICHARD A . LOCKSHIN

B. ELIMINATION B Y PHAGOCYTES In situ heterophagic degradation of cellular debris is a common occurrence, whether the phagocytes are specialized or not. Examples of degradation by specialized phagocytes include the dentary pulp of amphibians (Roux and Chibon, 1974), endocardiac tubes fused during the morphogenesis of the heart (Ojeda and Hurle, 1975; Hurle and Ojeda, 1979), uterine epithelium (Sandow et al., 1979), follicular atresia (Peluso et al., 1980), Mullerian ducts of male avian (Scheib, 1977), or mammalian (Price et a f . , 1977) embryos, and the tail epidermis of anurans (Kerr er al., 1974). Nonspecialized neighboring healthy cells also phagocytose dead cells or apoptotic fragments, for instance, in the tail epidermis of anurans (Kerr et al., 1974). It is not known by what mechanism the phagocytes recognize apoptotic bodies or degenerate cells. One can hypothesize an alteration-ne hesitates to call it denaturation-of the proteins of the plasma membrane or, more likely, a niodification of its properties by alteration of its glycoprotein coat. This alteration would permit phagocytes to distinguish it from neighboring healthy cells. Kecognition would provoke an endocytic reaction which would have as a primary purpose recycling of metabolites and destruction of possible pathogens. It is known that treatment of developing chick limb with the microfilament-disrupting agent cytochalasin B inhibits the endocytic activity of phagocytes but does not prevent programmed cell death (Kieny and Sengel, 1974). 1 . Phagocytic Activity by Specialized Macrophages

Among vertebrates, two types of phagocytes can take part in the elimination of necrotic material (Beaumont and Hourdry, 1976). These two types are true macrophages, derived from monocytes, which phagocytose cellular debris in, for example, larval anuran pancreas (Beaumont, 1977, or macrophages derived from the mesenchyme, such as those which eliminate the remnants of the internal gills (Beaumont and Hourdry, 1976), the epidermis, and the tail muscles (Weber, 1964; Kerr et al., 1974; Fox, 1972a,b,c, 1973a, 1977) of tadpoles, or the mesenchyme of embryonic avian limbs (Hammar and Mottet, 1971; Kieny and Sengel, 1974) or the necrotic epithelial septum remaining from the fusion of the palatal plates (Farbman, 1968). In insects, intervention, at various stages of cellular degeneration, of macrophages derived from hemocytes is frequently noted: in the epidermis (Whitten, 1968, 1969a,b; Lavenseau and Surleve-Bazeille, 1974; Bautz, 1975), or rnuscle fibers (Perez, 1910; Crossley, 1968), or extraembryonic epithelia (Petavy, 1976). Reduction of the number of phagocytes by irradiation of the hematopoetic organs or by transfusion of hemolymph slows elimination of larval cells from the abdominal epidermis of Calliphora larvae (Bautz, 1979). In the same manner, in polychaete annelids, eleocytes (coelomic cells) phagocytose the sarcolytes pro-

PROGRAMMED CELL DEATH

227

duced by the degeneration of muscle fibers (Clark and Clark, 1962; Dhainaut, 1966; Wissocq, 1976). 2. Phagocytic Activit?, by Nonspecialized Phagocvtes It is clearly established that cells neighboring degenerating cells can also phagocytose them; examples include digestion of apoptic bodies in diverse epithelia of larval anurans (Kerr ef al. , 1974; Beaumont and Hourdry, 1976; Beaumont, I977), and epithelia of metamorphosing insects (Fristrom, 1968); Giorgi and Deri, 1976; Andries, 1976, 1977). This phenomenon has also been seen in human embryonic ovary in which degenerate germinal cells are phagocytosed by adjacent follicular cells (Gondos, 1973). in young mouse blastocysts covered by the zona pellucida, in which phagocytosis of necrotic cells is effected by trophoblasts or by cells of the embyronic disk (El-Shershaby and Hinchliffe, 1974, 1975), in the developing limb (Dawd and Hinchliffe, 1971; Hinchliffe and Ede, 1973), in embryo heart where the healthy myocardial cells phagocytose apoptotic bodies (Hurle and Ojeda, 1979), during the closure of neural tube by the neuroepithelial cells (Schleuter, 1973), or during the formation of lens vesicles (Garcia-Porrero el a/., 1979). Similarly, in neural tissue gliocytes undertake the elimination by phagocytosis of necrotic neurons or their fragments (O’Connor and Wyttenbach. 1974; Pilar and Landmesser, 1976; Stocker et a/.. 1978).

IV. Determination of Degeneration by Cellular Interaction The several models of dedifferentiation and cellular involution studied to date indicate that the phenomenon is of major biological interest in metazoans, but that it occurs by several programmed mechanisms. In all cases, however, whether the target cells are found in an embryo, larva, or adult, the initiation of programmed cell death appears to be provoked by the reception of extrinsic signals. A . SIGNALS DERIVED FROM CELLULAR INTERACTIONS

The agents that provoke degeneration of cells competent to degenerate are in general hormonal or locally diffusing chemicals (morphogenetic factors). 1 . Hormonal Action

In male mammals, it has been demonstrated that the involution of the Mullerian duct requires the early and precisely timed action of an anti-Miillerian hormone secreted by the Sertoli cells (Josso, 1974, 1978; Blanchard and Josso, 1974). Similarly, the involution of mammary glands in males requires the action of androgens. The hormones act directly on the mesenchyme, again during a

228

JACQUES BEAULATON AND RICHARD A . LOCKSHIN

well-defined period in fetal development (Kratochwil, 1977a,b). The involution of the epithelial cells does not however result from a direct effect of the hormone (Kratochwil and Schwartz, 1976; Drews and Drews, 1977; Duemberger et al.. 1978). The mechanism in this instance is poorly understood, but is supposed to involve the activation of mesenchyme cells surrounding the mammary bud. These latter cells then provoke an intermediary step which appears to consist of secretion of an antimammary substance which has a cytolytic effect on the epithelial cells (Raynaud and Delost, 1977; Kratochwil, 1977a). Hormonal control is of course also prominent in the cataclysmic phenomena of amphibian metamorphosis, in which thyroid hormones initiate the degeneration of the intestinal epithelium (Hourdry, 1977a) and several tissues in the tail (Fox, 1977; Weber, 1977). In insects, eclosion hormone provokes the degeneration of the intersegrnental muscles of Lepidoptera shortly after the adult emerges (Truman and Riddiford, 1970; Truman, 1973; Lockshin and Beaulaton, 1974). Most importantly, ecdysone initiates destruction of larval tissues as well as the development of the imaginal tissues: involution of Dipteran salivary glands (Hendrikson and Clever, 1972), prothoracic glands in roaches (in the absence of protective factors secreted by the corpora cardiaca-Lanzrein, 1975). Injection of ecdysteroids into female Dipterans (Phormia reginu and Sarcophaga bulluta) has demonstrated that 20-hydroxy ecdysone can cause the degeneration of oocytes (Pappas and Fraenkel, 1978; Fraenkel and Hollowell, 1979). These results were not modified if juvenile hormone (JH) was injected at the same time as the ecdysone. Similar results were obtained by Deoras and Bhaskaran (1967), who were studying ovarian degeneration after implantation of Weissman’s ring glands into Musca. In Drosophila, the ovarian chambers involute spontaneously in normal conditions (Giorgi and Deri, 1976) but in this situation the control mechanisms have not been worked out. The absence of egg laying provokes the secretion of ovarian oostatic hormone (Adams et al., 1968) which acts either on the secretion of juvenile hormone by the corpora allata or indirectly on the cerebral neurosecretory system which regulates the activity of the allata cells (Adams el al., 1975; Fraenkel and Hollowell, 1979). Retention of ripe oocytes in the ovary thus brings about a blockage of the development of other oocytes and of vitellogenesis, because of an inactivation of the corpora allata. Juvenile hormone is essential for both oogenesis and vitellogenesis (Sakurai, 1977). If juvenile hormone is blocked, presumably the level of ecdysteroids is high enough to provoke the degeneration of the oocytes as in the experimental situation (Hodgetts er al., 1967). Degeneration of larval mesenteric epithelium (Radford and Misch, 1971) appears to be a secondary effect of ecdysone or ecdysterone. The hormone acts first on the differentiation of the imaginal epithelium when juvenile hormone is very low or absent (Andries, 1976, 1979). The blood level of JH influences the degeneration of other target cells, such as

PROGRAMMED CELL DEATH

229

flight muscle fibers (Unnithan and Nair, 1977). In many insects, particularly those which become inactive during oviposition, the flight muscles degenerate in the presence of JH: the Heteropteran Dysdercus (Edwards, 1970; Davis, 1975), the Coleopterans Ips (Borden and Slater, 1968; Unnithan and Nair, 1977), and Dendroctonus (Sahota, 1975). In contrast, in Leptinotarsa, which awakens for oviposition, a drop of JH caused by allatectomy causes the involution of the flight muscles (Stegwee et al., 1963). Among other invertebrates, a cerebral neurosecretion in the mollusc Crepidulu controls the regression of the muscle fibers of the penis at the time of sexual inversion (Streiff and Le Gall, 1976). 2. Morphogenetic Chemicals During the morphogenesis of metazoans, the beginning of dedifferentiation of target cells generally implies the existence of nonhormonal chemical signals which by diffusing from a specific site would initiate the process. Thus, reduction of the concentration of a morphogenetic (growth stimulating) substance below a particular threshold along a gradient appears to provoke death of competent cells (Ede, l976b). During development, programming of genetically controlled cell death relates to signals regarding the positions of cells. The effect of certain mutations, such as those affecting the skeleton of limbs of vertebrates (Hinchliffe and Ede, 1967), is manifested by alterations in the chronology of distribution of necrotic areas (Ede, 1976a,b). According to this hypothesis, genetic anomalies are reflected perhaps by modifications in the properties of the plasma membrane which perturb the diffusion of morphogenetic substances. This alteration causes an expansion or reduction of necrotic zones, according to the genes being expressed (Hinchliffe, 1977, 1981). In the situation of the posterior necrotic zone of the chick wing bud the “death clock” is irreversibly started at an early stage of development (Saunders et al., 1962), concomitant with a reduction in synthetic activity in the nucleus (Pollack and Fallon, 1976). The diminution in synthesis of DNA and RNA precedes by 6 hours the diminution in synthesis of proteins (Pollack and Fallon, 1974, 1976); death follows shortly after the loss of protein synthesis. Death appears to result from the blockage of the genome. Similar mechanisms have been proposed for death in thymocytes (Munck and Crabtree, 1981) and sympathetic nerves (LeviMontalcini and Aloe, 1981) although several other factors are now known to be involved. In thymocytes, death is triggered by a decrease in thymic growth factor (Munck and Crabtree, 1981). In nerves as well as in thymocytes, several pathways can lead to death (Levi-Montalcini and Aloe, 1981).

B. CONTROLMECHANISMS We do not at present know the molecular bases of the control of cell death. Presumably the cellular response to hormones involves shifts of inorganic ions such as divalent cations or cyclic nucleotides functioning as second messengers

230

JACQUES BEAULATON AND RICHARD A. LOCKSHIN

(McMahon and West, 1976; Truman and Schwartz, 1980; Lockshin, 1981a; Lockshin and Barata, unpublished). It has been demonstrated, for example, that cyclic adenosine monophosphate (CAMP) mimics ecdysone in the programmed death of alary epithelium in an insect (Seligman and Doy, 1973). Cyclic AMP is assumed for several reasons to be the initiator of cell death during the fusion of the palatal shelf in embryonic rodents (Greene and Pratt, 1979) and cyclic GMP is very high just before the involution of the intersegmental muscles of silkmoths (Schwartz and Truman, 1982). Levels of both cyclic GMP and cyclic AMP are high in muscles and salivary glands of Manducu sexru (Lepidoptera) before these tissues involute, and the levels fall rapidly at the beginning of involution (Lockshin, Barata, Ackerman, and Could, in preparation). The hypothesis of the intervention of regulatory proteins such as protein kinases (Weber, 1977) suggests the existence of regulatory mechanisms comparable to those which function on the proliferation and differentiation of cells (McMahon and West, 1976). This control may function either at the transcriptional level or at the translational level. In the instance of cells caused to degenerate by morphogenetic substances, the cells appear to interpret information relating to their position as compared to that necessary for the formation of other organs, as in the interdigital areas of limb buds (Whitten, 1969b). It has been suggested that the transfer of regulatory molecules is through low resistance junctions (“gap junctions”), the forniation and disappearance of which appears itself to be controlled in vitro by cyclic AMP or cyclic GMP (Sheridan, 1976). It is now well established that biochemical transformations such as a reduction of the rate of synthesis of DNA (Saunders and Fallon, 1967; Beaupain, 1979) or of SDH (Hammar and Mottet, 1971) precede the ultrastructural transformations which have been observed. (The latter hypothesis has been challenged: see Hinchliffe, 1981.) In this context, one must work out the relations of this sytem of intracellular communications to dedifferentiation, as is being done for differentiation (Gilula, 1977; Pitts, 1977). The importance of intercellular communications in the beginning of the process of degeneration should be emphasized, especially if one considers the situation in neurons. The lack of peripheral synapses at certain critical stages of embryogenesis has a strong negative effect on the differentiation of motor neurons (Price, 1974; Prestige, 1974; O’Connor and Wyttenbach, 1974; Gutmann, 1976; Sotelo, 1977). As many as twice as many neurons may be formed as eventually survive (Landmesser and Pilar, 1974; Pilar and Landmesser, 1976). In the same manner, immature neurons in the sympathetic ganglia that have not established synaptic connections with peripheral tissues rapidly degenerate (Levi-Montalcini and Calissano, 1979). Nevertheless, in other neural systems, neurons which have established functional synapses still degenerate (Landmesser and Pilar, 1976). This mode of degeneration seems to imply a competition between cells for postsynaptic sites (Landmesser and Pilar, 1976). In these latter systems, cell

PROGRAMMED CELL DEATH

23 1

death is suggested not to be programmed, in that it depends on peripheral connections or other external input (O’Connor and Wyttenbach, 1974; Price, 1974; Prestige, 1974) and retrograde transmission of the signal (Gutmann, 1976). However, as is described above, most instances of programmed cell death depend in their early stages on an external signal and thus are reversible until the process of death has actually begun; the importance is that they do respond to a signal by dying, and that, owing to the metabolic preparations undergone by the cell in preparation for its death, the signal is relatively distinct from the actual collapse of the cell (Lockshin, 1982). In the case of neurons the nature of the signal or lack of it is not actually known ( C r a g , 1970; Gutmann, 1976) but it appears that retrograde axonal transport (from axon to perikaryon) plays a role in releasing the process of cell death. Finally, it appears that we still lack information as to the molecular bases of programmed cell death. Traumatic death appears to result from the entry of calcium into restricted areas (Schanne et al., 1979; Trump et al., 198I ; Farber, 1981). Programmed cell death may ultimately lead to this situation, but by directions under the control of the organism (Lockshin, 1981). A surprising recent suggestion is that this control may take any of several routes (LeviMontalcini and Aloe, 1981; Munck and Crabtree, 1981; Wyllie, 1981). Thus any fragile component may falter. Rather like a spaceship erring in space, the primary mode of failure may result from the first system to collapse once the central metabolic or nutritional status has been compromised. Interpretation of the signal may thus be a metabolic shift of the cell so that it can no longer maintain net metabolic balance. Even if we have made enormous progress in identifying several steps in the process, we still must define how signals furnished by direct or indirect intercellular interactions are interpreted by target cells, and how they are then converted into a complete and ultimately irreversible reorientation of cellular activity.

ACKNOWLEDGMENT Much of the research reported here has been sponsored by grants to R.A.L. from the National Science Foundation. U.S.A.

REFERENCES Adarns, T. S . , Hintz, A. M . , and Pornonis, J . G . (1968). J . lnsecr Phvsiol. 14, 983-993. Adarns. T. S . , Grugel. S . . Jttycheriah, P. I . , Olstand. G . . and Caldwell, J. A. (1975). J . Inserr Phvsiol. 21, 1027-1043. Andries. J . G . (1975). J . Microsr. Biol. Cell. 24, 327-350.

232

JACQUES BEAULATON AND RICHARD A. LOCKSHIN

Andries. J. G. (1976). Bull. Soc. Zool. F r . 101 (Suppl. 5 ) . 1-7. Andries, J. G. (1977). Biol. Cell. 29, 203-308. Andries, J. G. (1979). Gen. Comp. Endorrinol. 39, 174-191. Bautz, A.-M. (1975). In!. J. Insect Morphol. Embryol. 4, 495-515. Bautz, A.-M. (1979). Arch. Zool. Exp. Gen. 120, 183-194. Beaulaton. J . (1967). J. Microsc. 6, 349-370. Beaulaton, I . , and Lockshin, R. A. (1973). C . R. Acad. Sci. (Paris) 277D, 1573-1576. Beaulaton, J., and Lockshin, R. A. (1977). J. Morphol. 154, 39-58. Beaumont, A. (1973). Bull. Sor. Zool. F r . 99, 134. Beaumont, A. (1977). Collog. I n f . C.N.R.S. (Paris) 266, 113-124. Beaumont, A., and Hourdry, J. (1976). Bull. Soc. Zool. Fr. 101, 89-94. Beaupain, R. (1979). Experientia 35, 1380-1381. Bellairs, R. (1961). J. Anat. 95, 54-60. Blanchard, M. G., and Josso, N. (1974). Pediarr. Res. 8, 968-971. Borden, I. H., and Slater, C. E. (1968). Z. Verol. Physiol. 61, 366-368. Bowen, I. D. (1981). In "Cell Death in Biology and Pathology" (1. D. Bowen and R. A. Lockshin. eds.), pp. 381-446. Chapman & Hall. London. Bowen, 1. D.. and Lewis, G. H. J. (1980). Hisrochemisrrv 65, 173-179. Clark, M. E., and Clark, R. B. (1962). Zool. Jahrb. Abt. A&. Zool. Physiol. 70, 24-90. Cragg, B. G. (1970). Brain Res. 23, 1-21. Crossley, A. C. (1968). J. Insert Physiol. 14, 1389-1407. Cruz-Landim, C., da, and Silva De Moraes, R. L. M. (1979). C.ylobios 24, 13-23. Davis, N. T. (1975). Ann. Entomol. Soc. Am. 68, 910-914. Dawd, D. S., and Hinchliffe, J. R. (1971). J. Embryol. Exp. Morphol. 26, 401-424. De Duve, C., and Wattiaux, R. (1966). Annu. Rev. Physiol. 28, 435-492. Deoras, G.S., and Bhaskaran, G. (1967). J. Univ. Bombay 25, 73-87. Dhainaut, A. (1966). C. R. Arad. Sci. Paris Ser. D. 262, 2740-2743. Drews, V.. and Drews. U. (1977). Cell 10, 401-404. Durnberger. H., Heuberger, B., Schwartz, P.. Wasner, G.,and Kratochwil, K. (1978). Canrer Res. 38, 4066-4070. Ede, D. A. (1976a). Collog. Int. C.N.R.S. (Paris) 266, 187- 191. Ede, D. A. (l976b). Cell Surface Rev. 1 , 495-543. Edwards, F. J. (1970). J. Insert Physiol. 16, 2027-2031. El-Shershaby. A. M., and Hinchliffe, J. R. (1974). J. Embpol. Exp. Morphol. 31, 643-654. El-Shershaby, A. M., and Hinchliffe, J. R. (1975). J. Embryo/. Exp. Morphol. 33, 106~'-1080. Farber, J. L. (1981). LifeSci. 29, 1289-1295. Farbman, A. 1. (1968). Dev. Biol. 18, 93-1 16. Forsberg, J. G., and Kallen. €3. (1968). Rev. Roum. Embryol. C.vtol. Ser Embrvol. 5 , 9-102, Fox, H. (1972a). Arch. Biol. 83, 373-394. Fox, H. (1972b). Arch. Biol. 83, 395-405. Fox. H. (1972~).Arch. Biol. 83, 407-417. Fox, H. (1973a).J. Embpol. Exp. Morphol. 30, 377-396. Fox, H. (1973b). Folia Morphol. 21, 109-112. Fox, H. (1974). J. Zool. 174, 217-235. Fox. H. ( 1977). Collog. Int. C.N.R.S. Paris 266, 93- 112. Fraenkel, G.. and Hollowell, M. (1979). J. Insect Physiol. 25, 305-310. Fristrom, D. (1968). J. Cell Biol. 39, 488-491. Garcia-Porrero, J. A , , Collado, J. A,, and Ojeda, J. L. (1979). Anat. Rec. 193, 791-804 Gilula, N. B. (1977). In "International Cell Biology" (B. R. Brinkley and K. R. Porter, eds.), pp. 61-69. Rockefeller Univ. Press, New York.

PROGRAMMED CELL DEATH

233

Giorgi, F., and Deri. P. (1976). J. Embryol. Exp. Morphol. 35, 521-533. Glucksmann, A. (1951). Biol. Rev. Cambridge Philos. Soc. 26, 59-86. Goffinet. 0.. and Gregroire, C. (1975). Arch. Inr. Phvsiol. Biochim. 83, 707-722. Condos, B. (1973). Z. Zellforsch. Mikrosk. Anat. 138, 23-30. Greene, R. M.. and Pratt, R. M. (1979). J. Histochem. Cvtochem. 27, 924-931. Gutmann, E. (1976). Annu. Rev. Physiol. 38, 177-216. Hammar. S . P., and Mottet, N. K . (1971). J. Cell Sci. 8, 229-251. Helminen, H. J.. and Ericsson, J. L. E. (1968). J. Ultrasrrucr. Res. 25, 228-239. Hendrikson, P. A., and Clever, U. (1972). J. lnsecr Phvsiol. 18, 1981-2004. Hinchliffe, J. R. (1977). Colloq. Inr. C.N.R.S. (Paris) 266, 173-185. Hinchliffe, J. R. (1981). I n “Cell Death in Biology and Pathology” (1. D. Bowen and R. A. Lockshin, eds.), pp. 35-78. Chapman & Hall, London. Hinchliffe, J. R., and Ede, D. A. (1967). J. Embryol. Exp. Morphol. 17, 385-404. Hinchliffe, J. R., and Ede, D. A. (1973). J. Embryol. Exp. Morphol. 30, 753-772. Hodgetts, R. B., Sage, B., and O’Connor, J. D. (1977). J. lnsecr Phvsiol. 60, 310-317. Hourdry, J. (1971a). Histochemie 26, 142-159. Hourdry, J. (1971b). J . Microsc. 10, 41-58. Hourdry, J. (1974). W. Roux Arch. Entwicklungsmech. Org. 174, 217-235. Hourdry, J. (1977). Colloq. Int. C.N.R.S. (Paris) 266, 123-136. Hourdry, J.. and Dauca, M. (1977). Inr. Rev. Cvrol. Suppl. 5, 337-385. Hurle, J.. and Hinchliffe, J. R. (1978). J . Embryol. Exp. Morphol. 43, 123-136. Hurle, J. M.. and Ojeda, J . L. (1979). J. Anar. 129, 427-439. Johnson, B. (1980). Tissue Cell 12, 529-539. Jones, J. C. (1975). I n “Invertebrate Immunity. Mechanisms of Invertebrate Vector-Parasite Relations” (K. Maramorosch, ed.), pp. 119-128. Academic Press, New York. Josso, N. (1974). Pediarr. Res. 8, 755-578. Josso, N. (1978). Recherche 9, 379-381. Jurand, A,, and Pavan, C. (1975). Cell Differ. 4, 219-236. Kerr, J. F. R. (1971). J. Parhol. 105, 13-20. Kerr, J. F. R., Wyllie, A . H.. and Cume, A. R. (1972). Br. J. Cancer 26, 239-257. Kerr. 1. F. R.. Harmon, B., and Searle, J. (1974). J. Cell Sci. 14, 571-585. Kieny, M., and Sengel. P. (1974). Ann. Biol. 13, 57-68. Kratochwil, K . (1977a). Colloq. Inr. C.N.R.S. (Paris) 266, 85-92. Kratochwil, K. (1977b). Dev. Biol. 61, 358-365. Kratochwil, K., and Schwartz, P. (1976). Proc. Narl. Acad. Sci. U.S.A. 73, 4041-4044. Landmesser, L., and Pilar, G . (1974). J . Phvsiol. (London) 241, 737-749. Landmesser, L.. and Pilar, G. (1976). J. Cell Biol. 68, 357-374. Lanzrein, B. (1975). J. Insect Phvsiol. 21, 367-389. La Pushin, R. W., and De Harven, E. (1971). J. Cell Biol. 50, 583-597. Lavenseau. L.. and Surleve-Bazeille, J. E. (1974). J. Microsc. 21, 189-192. Lavenseau, L.. and Surleve-Bazeille, J. E. (1976). Bull. Soc. Zoo/. Fr. 101, 69-74. Levi-Montalcini. R . , and Aloe, L. (1981). I n “Cell Death in Biology and Pathology” ( I . D. Bowen and R. A. Lockshin, eds.). pp. 295-328. Chapman & Hall, London. Levi-Montalcini. R.. and Calissano, P. (1979). Pour / a Sci. 22, 12-22. Littau, V., Allfrey. V., Frenster, J.. and Mirsky, A. (1964). Proc. Narl. Acad. Sci. U.S.A. 52, 93-10, Locke, M., and Sykes, A. K . (1975). Tissue Cell7, 143-158. Lockshin, R. A. (1969). J . Insect Physiol. 15, 1505-1516. Lockshin, R. A. (1980). In “Degradative Processes in Heart and Skeletal Muscle” (K. Wildenthal, ed.). pp. 225-254. Elsevier, Amsterdam.

234

JACQUES BEAULATON AND RICHARD A. LOCKSHIN

Lockshin, R. A. (1981). In “Cell Death in Biology and Pathology” (I. D. Bowen and R. A. Lockshin, eds.), pp. 79-122. Chapman & Hall, London. Lockshin, R. A. (1982). In “Cell Biology, C.R.C. Handbook Series on Aging” (R. C. Adelinan and V. J. Cristofalo, eds.). C.R.C., Boca Raton, Florida, in press. Lockshin, R. A,, and Beaulaton, J. (1974). Life Sci. 15, 1549-1565. Lockshin, R. A., Royston, M., Joesten, M., and Carter, T. (1981). In “Cell Death in Biology and Pathology” (1. D. Bowen and R . A. Lockshin, eds.). pp. 273-295. Chapman & Hall, London. McMahon, D. (1974). Science 185, 1012-1021. McMahon, D., and West, C. (1976). Cell Su$. Rev. 1, 449-493. Manasek, F. J . (1969). J. Embryol. Exp. Morphol. 21, 271-284. Mayer-Criel, C., de (1971). Arch. Biol. 82, 163-165. Moe, H., Thorball, N., and Winther Nielsen, H. (1979). Cell Tissue Res. 203, 339-354. Morimoto, T., Blobel, G., and Sabatini, D. D. (1972). J. Cell Biol. 52, 355-366. Mottet, N. K . , and Hammar, S . P. (1972). J. C e / / Sci. 11, 403-414. Munck. A,, and Crabtree, G. R. (1981). In “Cell Death in Biology and Pathology” (1. D. Bowen and R. A. Lockshin, eds.), pp. 329-362. Chapman & Hall, London. Nicolas, G . , and Zagury, D. (1980). Biol. Cell. 37, 231-234. O’Connor, T., and Wyttenbach, C. B. (1974). J. Cell Biol. 60, 448-459. Ojeda, I. L . , and Hurle, J. M. (1975). J . Embryol. Exp. Morphol. 33, 523-534. Pannese, E., Luciano, L., lurato, S . and Reale, E. (1976). Acta Neuropathol. 36, 209-220. Pappas, C., and Fraenkel, G. (1978). J. Insert Physiol. 24, 75-80. Pautou, M. P.. and Kieny, M. (1971). C.R. Acad. Sci. (Paris) 272D, 2025-2028. Peluso, J. I . , England-Charlesworth, C., Bolender, D. L., and Steger, R. W. (1980). Cell Tissue Res. 211, 105-1 15. Perez, C. (1910). Arch. Zool. E.rp. Gen. 4, 1-274. Petavy, G. (1976). Int. J. Insect Morphol. Embryo/. 5, 167-186. Pilar, G., and Landmesser, L. (1976). J. Cell Biol 68, 339-356. Pitts, J . D. (1977). In “International Cell Biology” (B. R. Brinkley and K. R. Porter, eds.), pp. 43-49. Rockefeller Univ. Press, New York. Policard, A,. and Bessis, M. (1968). “Elements de Pathologie Cellulaire,” pp. 1-285. Masson, Paris. Pollak, R. D., and Fallon, J. F. ( 1974). Exp. Cell Res. 86, 9- 14. Pollak, R. D., and Fallon, J. F. (1976). Exp. Cell Res. 100, 15-22. Prestige, M. C. (1974). Br. Med. Bull. 30, 107-1 1 I. Price, D. L. (1974). Ann. N.Y. Acad. Sci. 238, 255-263. Price. J. M.. Donahoe. P. K., Ito, Y., and Hendren, W. H. 111. (1977). Am. J . Anat. 149,353-375. Radford, S . V., and Misch, D. W. (1971). J. Cell Biol. 49, 702-71 I . Raynaud, A,, and Delost, P. (1977). Colloq. Int. C.N.R.S. (Paris) 266, 71-84. Roux, J. P., and Chibon. P. (1974). C. R. Acad. Sci. (Paris) Ser. D. 278, 1369-1372. Rowley. A. F., and Ratcliffe, N. A. (1976). Tissue Cell 8, 437-446. Sahota. T. S. (1975). J. Insect Physiol. 21, 471-478. Sakurai, H. (1977). J . Insect Physiol. 23, 1295-1302. Salzgeber. B., and Weber. R. (1966). J. Embryo/. Exp. Morphol. 15, 397-419. Sandow. B. A,. West, N. B., NOrmdn, R. L., and Brenner. R. M. (1979).Am. J. Anat. 156, 15-36. Saunders, J . W., Jr. (1966). Science 154, 604-612. Saunders, J . W., Jr., and Fallon, J. F. (1967). In “Major Problems in Developmental Biology” (M. Locke, ed.), pp. 289-314. Academic Press, New York. Saunders, J. W., Jr., Gasseling, M. T., and Saunders, L. C. (1962). Dev. Biol. 5, 147-1?8. Schanne, F. A. X.,Kane, A. B., Young, E. E., and Farber, J. L. (1979). Science 206, 700-702. Scheib, D. (1977). Colloq. Inr. C.N.R.S. (Paris) 266, 59-70.

PROGRAMMED CELL DEATH

235

Schlueter, G. (1973). Z. Anaf. Entwicklungsgesch. 141, 251-264. Schwartz, L. M . , and Truman. J. W. (1982). Science 215, 1420-1421. Schweichel, J . U., and Merker, H. J. (1973). Terafology. 7, 253-266. Seligman. I. M., and Doy, F. A. (1973). J. Insect Phvsiol. 19, 125-136. Seligman, I . M., Filshie, B. K., Doy, F. A.. and Crossley, A. C. (1975). Tissue Cell 7, 281-296. Sheridan, J . D. (1976). Cell SurJ Rev. 1, 409-447. Sotelo, C. (1977). In “International Cell Biology” ( B . R . Brinkley and K. R. Porter, eds.), pp. 83-92. Rockefeller Univ. Press, New York. Stegwee. D., Kimmel, E. C., De Boer, J. A., and Henstra. S. (1963). J. Cell B i d . 19, 519-529. Stephan-Dubois, F . . Lanot, R., and Bautz, A.-M. (1974). Ann. Biol. 13, 27-34. Stocker, R. F . , Edwards, J. S., and Truman, J . W. (1978). Cell Tissue Res. 191, 319-332. Streiff, W., and Le Gall, S. (1976). Colloq. fnt. C.N.R.S. (Paris) 251, 139-147. Truman, J . W. (1973). Am. Sci. 61, 700-706. Truman, J . W., and Riddiford, L. M. (1970). Science 167, 1624-1626. Truman, J . W., and Schwartz, L. (1980). In “Peptides: Integrators of Cell and Tissue Function” (F. E. Bloom, ed.), pp. 55-67. Raven, New York. Trump. B. F.. Berezesky, 1. K.. and Osornio-Vargas, A. R. (1981). I n “Cell Death in Biology and Pathology” (1. D. Bowen and R. A. Lockshin, eds.), pp. 209-242. Chapman & Hall, London. Unnithan, G. C . , and Nair, K. K. (1977). Cell Tissue Res. 185, 481-490. Weber, R. (1964). J. CelI Biol. 22, 481-487. Weber, R. (1977). Colloq. fnr. C.N.R.S. (PariA) 266, 137-146. Whitten, J . (1968). In “Metamorphosis. a Problem in Developmental Biology” (W. Etkin and L. I. Gilbert, eds.), pp. 43-105. North-Holland Publ.. Amsterdam. Whitten. J. (1969a). J. fnsect Phvsiol. 15, 763-778. Whitten, J . (1969b). Science 163, 1456-1457. Wissocq, J . C. (1976). Bulk Soc. Z o o / . Fr. 101 (Suppl. 5). 19-23. Wyllie. A. H. (1981 ). In “Cell Death in Biology and Pathology” (1. D. Bowen and R. A. Lockshin, eds.), pp. 9-34. Chapman & Hall, London. Wyllie, A . H., Kerr. J. F. R., and Currie, A. R. (1980). Inr. Rev. Cyrol. 68, 251-306.

This Page Intentionally Left Blank

INTERNATIONAL REVIEW OF CYTOLOGY, VOL. 79

Cryofixation: A Tool in Biological Ultrastructural Research’ HELMUTPLATTNER*

AND LUIS BACHMANNt

*Faculty of Biology, University of Konstanz, Konstanz. Federal Republic of Germany, and flnstitute of Technical Chemistry, Technical University (# Miinchen, Garching. Federal Republic of Germciny

I. Scope and Goal of Cryofixation 11. Basic Methodical Aspects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111. Experimental Consequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Cryofixation Techniques-Advantages and Restrictions. . . . . . . , . . A. Conventional Techniques. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Advanced Cryofixation Techniques ... V. Cooling Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI . “Resolution” of Cryofixation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII. Ultrastructural Side-effects of Chemical Pretreatments . . . . . . . . . . . VIII. Tests for the Quality of Cryofixation.. . . . . . . . . . . . . . . . . . . . . . . . A. Physical Tests B. Viability Tests C. General Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IX. Combination of Cryofixation with Other Techniques . . . . . . . . . . . . A. Cryofixation as a Tool in Cytochemistry . . . . . . . . . . . . . . . . . . B. Combination with Freeze-Fracturing .... C. Combinations with Surface Analyses . . . . . . . . . . . . . . . . . . . . . D. Combinations with Other Cytochemical Procedures . . . . . . . . . E. Combinations with Ultrathin Sectioning.. . . . . X . Suspensions, Emulsions, and Solutions. . . . . . . . . . . . . . . . . . . . . . . A . Size and Shape of Macromolecules . . . . . . . . . . . . . . . . . . . . . . B. Cryofixation of Heated Solutions . . . . . . . . . . . . . . . . . . . . . . . . C. Determination of Particle Weight by Counting . . . . . . . . . . . . . XI. Nonaqueous Systems .... XII. Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

237 240 243 247 248 250 27 I 273 274 278 278 278 279 280 280 280 286 288 289 290 29 I 292 292 292 293 293

I. Scope and Goal of Cryofixation From molecules to cells and organisms, nature itself shows us some strikingly successful examples of the preservation of biological materials by freezing. A recent analysis of mammoths, which were dug out from the Siberian ice, where ‘Dedicated to Professor Fritz Miller on the occasion of his seventieth birthday. 237 Copyright 0 1982 by Academic Press. Inc. All rights of repmduclion in any fom reserved. ISBN 0-12-364479-8

238

HELMUT PLATTNER A N D LUIS BACHMANN

they froze some 40,000 years ago, revealed that their serum albumin still displayed some of its antigenicity (Prager et al., 1980). Some of their red and white blood cells were allegedly also intact (see O’Toole, 1980). Whole organisms can also be very resistent to freezing damage due to the accumulation of polyols, glycerol (see Zachariassen, I979), glycoproteins, or other cryoprotective compounds (see also Franks, 1977). Most fascinating for an electron microscopist with a special interest in cryofixation are some “cryptobiotic” organisms, which (because they are almost fully dehydrated) can survive cooling close to absolute zero. They would also survive transfer into a vacuum and to some extent the bombardment with electrons for the purpose of observation in a scanning electron microscope (see Crowe and Cooper, 1971). In practice, cryofixation and freeze-storage of biological materials is an important problem in many regards: for the storage of living materials (e.g., latent cell cultures, sperm, egg and embryo conservation, organ transplantaion, etc.), for the storage and processing of proteins (e.g., maintainance of enzyme activity and antigenicity), and finally for the preservation of structural details (i.e., from macroscopic dimensions to the ultrastructural level). It must be clearly stated at the start that the problems of cryofixation in relation to electron microscopy are quite different from other aspects. As we shall discuss in more detail, there is no direct correlation between ultrastructural and functional preservation of biological materials. The goal of ultrastructural fixation methods is to preserve the momentary distribution of all components in a system. For molecular systems, i.e., solutions and suspensions, this implies the maintenance of the statistical distribution of all solute components in the solvent and, furthermore, the preservation of the size and shape of the macromolecules in the hydrated state. In nonideal solutions, e.g., micellar solutions or molecular aggregates, the association and orientation of interacting particles must be preserved. With cellular systems, the original localization of subcellular components, including ions and molecules (in the cytoplasm and in membranes), and the dynamic organelle and membrane inceractions should be instantaneously “fixed. This means high “spatial resolution” for the morphology and high “time resolution” for dynamic processes. Undoubtedly chemical fixation works too slowly to meet these requirements (see Zingsheim and Plattner, 1976), which at present can be fulfilled only hy the “physical procedure” of cryofixation. In the field of electron microscopy cryofixation obtained special actuality as soon as the freeze-fracturing technique was developed (Steere, 1957; Moor et al., 1961; see also the earlier work by Hall, 1950, and Meryman, 1950, 1957). At the beginning this new technique was appreciated as an invaluable control of the standard chemical fixation and ultrathin sectioning. It was also fascinating to look at “potentially viable” cells-or rather a metal replica thereof (Moor, 1964). Still, freeze-fracturing offers us the most direct control of the quality of cryofixation achieved with a certain freezing method, since, throughout the ”

CRYOFIXATION

239

procedure the specimen can be kept at a sufficiently low temperature. Secondary effects (such as recrystallization) can be easily avoided. Therefore, in this article we will frequently refer to freeze-fracture results as a test for the quality of cryofixation. Freeze-fracturing and etching is the most widely used “secondary technique” or follow-up procedure after cryofixation. For “merely ultrastructural” purposes many laboratories still use “standard” cryofixation techniques which involve chemical pretreatments (see Section VII). Materials are prefixed by aldehydes, and impregnated with antifreeze agents, mainly glycerol. Then the samples are plunged into a cooling fluid. The use of an antifreeze agent is necessary to avoid the formation of large ice crystals during freezing. Chemical fixation in turn is required to avoid structural artifacts introduced by the antifreeze. With this procedure the ultimate goal of cryofixation as a straightforward physical procedure is largely abandoned. Analytical techniques, especially on soluble compounds, would not be feasable. They depend on the development of more “advanced” freezing techniques. Several authors have cautioned for a long time that severe ultrastructural artifacts occur when chemical pretreatments are applied before cyrofixation. As we shall discuss in Section VII, ultrastructural perturbations might take place during fixation and/or during cryoprotection. They concern mainly changes of the ultrastructure of biomembranes; this is even more serious, since freezefracturing is the method of choice for the ultrastructural study of membranes. The main problem of cryofixation is the fact that the “fixed” structures are thermodynamically unstable at low temperatures (Figs. I and 2). Actually the better the fixation, the greater the instability of the structures after cryofixation. There is always the tendency of the water to form pure ice crystals and to “squeeze” the solute into the areas between the crystals (Fig. 3). The goal of cryofixation and of the secondary procedures is to keep these phase separation (“segregation”) phenomena as small as possible. The only way to overcome this problem during freezing is to speed up the cooling process to such an extent, that there is not enough time for the system to rearrange itself into a thermodynamically more stable state. Depending on the water content of the specimen, cooling rates of lo4 degrees/second and higher are required for the practical purposes of electron microscopy. Once the specimen is at liquid nitrogen (LN,) temperature (- I96”C), the activation energy for a fast enough rearrangement of the water and the solute is no longer available. The specimen is “stable” in practical terms. For cryofixation itself sufficient stability begins at temperatures below -80°C. In all cases, the “spatial resolution” (caused by the size of “segregation” zones) remains far beyond the limits preset by the performance of the electron microscope and sometimes even beyond those of replicating techniques. The “time resolution” (for dynamic processes) achieved with advanced cryofixation techniques is in the range of 0.1 to I msec and is just sufficient to catch rapid cellular processes (like different stages of exocytosis; see Plattner, 1981).

240

HELMUT PLATTNER AND LUIS BACHMANN

1000

2000

3000

4000

Pb r 3

FIG. I . Phase diagram of water at high pressure. The liquid-vapor equilibrium curvc: is not discernible on this scale. Ice I to V are different crystalline states (which are of little concern in this context). From Taylor (1960).

Meanwhile a range of “secondary techniques” other than freeze-fracturing have been developed for electron microscopy. Considerable progress was made in the past decade with ultracryotomy, combined with scanning transmission, and X-ray microanalysis, as well as low-temperature microscopy. The techniques can be applied in widely diversified combinations (Section IX). However, we are at the point where the applicability of all these new developments largely depends on the quality of cryofixation (see, e.g., Echlin, 1977; Shotton, 1980; Severs, 1980). Another field with increasing interest in advanced cryofixation is the electron microscopy of emulsions, suspensions, and macromolecular solutions (Section X). There are also more trivial reasons to use cryofixation methods, e.g., for rendering the material hard enough to be cut. This can then help, for instance, to make cells accessible to substrates, antibodies, or other ligands, etc., which can then be localized in the electron microscope. This implies that it will not always be necessary to apply the latest rage of a cryofixation method and we will, therefore, not totally neglect conventional techniques in this article. It is important to have several different cryofixation techniques at hand, if one considers the variety of biological materials and the broad spectrum of secondary techniques, whereby each has its special requirements. 11. Basic Methodical Aspects

In electron microscopy the goal of cryofixation is to transfer a hydrated system from the fluid to the amorphous solid state (vitrification). For pure water it was

24 1

CRYOFIXATION

20

1

/

Solution

t Tr

u

U 0

c

-*1 -80

0 100 x

Ice

+

Solid Glrcerol 100 K 0

Gbcerol Water

FIG. 2. Phase diagram of glycerol-water. E. Eutectic point at approximately 67% glycerol and -46°C (Lane. 1925). At the eutectic point a mixture of small glycerol and ice crystals is formed, if the system is in a thermodynamic equilibrium. Under nonequilibrium conditions, as in cryofixation, ice is formed in the presence of undercooled glycerol-water solution (dashed line according to Lusena, 1960). For more compiled data on glycerol-water see Umrath (1977).

FIG. 3. Freeze-etch replica from a 5% solution of glycerol in water as a test specimen for the size of phase separation zones in dependence of the cooling rate. (a) Standard freezing by dipping a I-pl sample (on a Bakers freeze-fracture specimen carrier. gold, diameter -3 mm) into melting Freon 12. (b) Rapid cryofixation by spray-freezing as small droplets (diameter 10 pm) in propane. At this magnification the fractured solution appears literally "vitrified." even after "etching" (which. in fact, is not possible due to the even distribution of glycerol, so that the relief seen represents the original fracture).

-

242

HELMUT PLA’ITNER AND LUIS BACHMANN

accepted for a long time that a truely amorphous state is attainable only by condensation from the vapor phase (Angel1 and Sare, 1970; McMillan and Los, 1965; Pryde and Jones, 1952; Venkatesh et al., 1974). Only recently Briiggeller and Mayer (1980) and Mayer and Briiggeller (1982) reported vitrification of liquid water by specimen jet freezing (Section IV). Their results were verified by Dubochet and McDowall (1981). If so, would it then be possible to vitrify biological objects using practical cooling rates between lo3 and los Wsecond? This depends mainly on the water content and the chemical properties of the solute. To some extent the problem of vitrification in electron microscopy is simply semantic. The prime concern is not whether the water is in a crystalline or amorphous state but a question of phase segregation due to ice crystal formation. Only when solute material is displaced further then the resolution limit of the following secondary procedure, do freezing artifacts become a practical problem. Obviously the quality of cryofixation, which is better than the resolution of secondary techniques, cannot be judged by electron microscopy. Electron spin resonance studies of frozen salt solutions are much more sensitive in this respect (Briiggeler and Mayer, 1980). With the best cryofixation methods and under the most favorable conditions (Fig. 3b) electron microscopy gives the impression of a totally homogeneous texture with only little background granularity. In most cases, however., the occurrence of phase separation is recognizable at the ultrastructural level as a meshwork of segregated solute along the grain boundaries (Fig. 3a). In conclusion the practical goal of cryofixation of biological systems is to minimize the unavoidable ice crystal formation. The term vitrification is frequently used by electron microscopists, when an ice crystal background relief in the range of only about 10-30 nm is achieved. Figure 2 shows that even in a simple glyceroVwater solution the components segregated during freezing vary their composition. With more complicated systems like cytoplasm, the situation is much more complex. Generally the concentrated solute is segregated from the solvent (water) in smaller or larger compartments (Fig. 3). It has already been mentioned that phase separation requires a transport of material and is therefore relatively slow. It can be suppressed by high cooling rates and by influencing nucleation. The phase diagram of pure water is represented in Fig. 1. It can be seen that the melting point is pressure-dependent. The lowest melting point (-22°C) is reached at about 2 kbar. At this pressure supercooling to -92°C is possible (Kanno et al., 1975). Under normal pressure supercooling is possible only to -40°C (Kanno et al., 1975; Stillinger, 1980). Freezing point depression and supercooling under a pressure of 2 kbar change the speed of crystal formation (nucleation rate) and of crystal growth. Riehle (1968a,b; Riehle and Hochli, 1973) made practical use of these phenomena in his “high-pressure cryofixation method.” The behavior of biological specimens is not equivalent to that of pure water.

CRYOFIXATION

243

The solute materials may as well act as additional nucleation sites (heterogeneous nucleation), while in other cases they prevent heterogeneous nucleation which is usually a limiting factor for supercooling (MacKenzie, 1977). The nucleation rate might indeed be a critical parameter in cryofixation. If fluid samples are split up into small compartments by vigorous stirring in paraffin oil before they are “standard” frozen (emulsification method: Buchheim, 1972a,b), the quality of cryofixation is considerably improved, although the cooling rate is not increased. Similarly, by spraying small droplets onto a cold metal surface (Williams, 1954) or into a cryogen medium (Bachmann and Schmitt, 197 I ) , heterogeneous nucleation is believed to be reduced. Briiggeller and Mayer ( 1 980) also achieved the best cryofixation with emulsified samples. However, up to now there are no quantitative data on nucleation rates and crystal growth during cryofixation at high cooling rates. Another reason to minimize the sample size is the increased heat transfer due to the larger surface/volume ratio. In practice (Section IV) this can be done by injecting (spraying, jetting) the sample as small droplets into the cryogen (Bachmann and Schmitt, 1971, 1973a; Bachmann et al.. 1972) or by using thin sandwich samples (Luyet et al., 1962; Gulik-Kryzwicki and Costello, 1977, 1978). The latter method can be eventually combined with the application of the cryogen as a jet shot onto the sample (Moor et a / . , 1976; Miiller et al., 1980a; Pscheid and Plattner, 1980; Pscheid et al., 1981; Knoll et al., 1982a,b). 111. Experimental Consequences

The crystal growth during rewarming (recrystallization) is also very different in pure water and in biological samples. Amorphous water begins to recrystallize at about - 130°C (Briiggeller and Mayer, 1980; McMillan and Los, 1965; Pryde and Jones, 1952; Yannas, 1968). Most biological materials display various effects (like cytolysis after thawing) related to recrystallization effects at considerably higher temperatures (-80 to -50°C: Mazur, 1960, 1977; Meryman, 1957; Moor, 1964; Nei, 1973; Sakai et al., 1968; Taylor, 1960). Naturally no precise recrystallization point can be indicated, because biological material is not a phase in the physicochemical sense. The different behavior is also due to the binding of some of the cellular water by the solute materials. For practical work these data lead to the following recommendations: First, the temperature range between +20 and -80°C should be passed as fast as possible; the cooling rate is especially critical in this temperature region. Second, all manipulations required subsequent to cryofixation should be done at a temperature below -80°C. The physical properties of cooling media and of support materials are of great importance for the cooling rates obtainable. Table I indicates the relevant parameters for those materials which are most widely used. Cryogens are either liquids or metals. The following aspects are important. (1) The coolant should have a

244

HELMUT PLATTNER A N D LUIS BACHMANN

very low temperature, since a large temperature gradient facilitates heat diffusion. (2) The thermal conductivity of the coolant should be high. (3) Its heat capacity per volume (specific heat) should also be high. (4) Liquid coolants should have a large temperature difference between freezing and boiling, so that no insulating gas layer can be formed in contact with the warm sample (Leidenfrost phenomenon). (5) Cryogenic fluids should have a low viscosity to facilitate the injection of the sample and convection within the cryogen. Some metals, like silver and copper, have been used for a long time as a heatsink for the cryofixation of tissues due to their high thermal conductivity and heat capacity. Eranko (1954) has proposed the freezing of tissues on cold metal blocks. The method was introduced into electron microscopy by VanHarreveld (VanHarreveld and Crowell, 1964; VanHarreveld et al., 1965, 1974) and further improved by Heuser and co-workers (1975, 1979). The sample is pushed onto a polished metal surface, which can be cooled down to any temperature. If liquid helium is used (see below), the high temperature gradient allows one to achieve maximal cooling rates. Up to lo5 degrees/second were measured in the outermost zone (Heuser et al., 1979), but the low thermal conductivity of water and ice (Table I) will reduce the cooling rate within the sample itself. As the heat conduction within the sample is a limiting factor, the use of an LN,-cooled metal block yields quite similar results (VanHarreveld et al., 1974), but at much lower costs and with much less experimental efforts. Metals are also used as a specimen support during freezing. If the dissipation of heat occurs through this metal support, its thermal conductivity must be high and its heat capacity low (see Table I and Figs. 4 and 5). In order to keep the mass of metal support as low as possible this requires sufficient mechanical strength of the metal used (especially when secondary techniques like freezefracturing are applied). In addition the support material must be compatible with the biological material and have no toxic side-effects. A compromise must be made between these different requirements. For instance, copper or silver would be more appropriate from the physical viewpoint, but gold might be more favorable because it has no toxic side-effects. Titanium, which has recently been recommended as a support material with very high mechanical strength, allows the use of very thin foils (Handley et al., 1981). Its thermal conductivity, however, is about 20 times lower than that of copper. In some cases (like cell monolayer cultures grown on plastic sheets) the sample is frozen in a way that the heat does not have to dissipate through the support (e.g., one-sided cryogen jet; Section IV). In these cases a plastic material with low heat conductivity is appropriate. Plastic supports can then also facilitate further processing, e.g., by freeze-substitution and ultrathin sectioning. Both freezing point depression and reduced nucleation rate can also be achieved with the use of cryoprotective agents (Rey, 1960). As already stated, however, any chemical treatment is a principal hazard and should be possibly avoided. “Permeating” and “nonpermeating” antifreezes are commonly dis-

00s-

Z'Z

30001 - "€'I

9S.O ' 3 , O

300 I 'Z 300 'Z'P

8'2 3,OS-

S ' f 3,001 -

00sPJPaaN

Z'Z '3,o I 1 '0 51'0 I '0 1'0

SI '0

(05+)SL (09+ )08 (OS+)S8 (82.+ )OZ I (9Z+)ZZ

Z'I €'I t'I f'l 9' I

Off

0s 09 0s 82

9z

Of OP 8Z + ZP 961 -

8SI 091 091 -

L81OIZ-

PZI u093tl

JZ uoaq aueiuados~ auedoq

ZN1

246

HELMUT PLA‘ITNER AND LUIS BACHMANN L

73 K

-1000 173K

0 273K

30°

TI

FIG.4. Temperature dependence of the thermal conductivity of different metals used in cryofixation. (The thermal conductivity of titanium, which was recently introduced into cryofixation work by Handley era/.. 1981, is on average about 10-20 times lower than for aluminum.) Data from “east (1976).

tinguished. Among the first group are dimethyl sulfoxide (DMSO) and, most frequently used, glycerol. In contrast to this common classification, glycerol does not permeate into all cells (see Merynian, 1971, 1974), although it apparently does so in other cells (Niederrneyer et a l . , 1977). Among the second group are polymers, which have attracted more attention only in the past few years, since low molecular antifreezes were recognized as producing considerable artifacts. Polyvinylpyrrolidone (PVP) is the most common one, although some others, like hydroxyethyl starch and dextrans, are also in use (Barnard, 1980; Franks et al., 1977; Mazur et al., 1970; McGann, 1978; Skaer et al., 1‘377, 1978, 1979; Williams and Harris, 1977; Wilson and Robards, 1980). Sucrose, which is present in rather high concentrations in fractions obtained by density gradient centrifugation, as well as some other gradient materials, also have to be considered as nonpenetrating or only slowly penetrating cryoprotectants (Fairant and Woolgar, 1970); therefore, it is easy to process such fractions by cryofixation.

247

CRYOFIXATION

Although some basic mechanisms of the cryoprotective action are known, some points need further clarification. The effect of nonpermeant cryoprotective agents tells us that the extracellular medium is somehow interrelated to intracellular processes during freezing. It is possible that they prevent heterogeneous nucleation (seeding), which would occur when an extracellular ice crystal would puncture the cell membrane and, thus, could act as a nucleation site for intracellular ice crystal formation (Mazur, 1960; Skaer er al., 1977, 1979). In this context it is interesting that glycerol acts as a cryoprotective agent even when it does not penetrate a certain cell type (Meryman, 1974). Normally, the binding of intra- and extracellular water would account for the protective effect. It is also remarkable that polymeric antifreezes, although used in concentrations similar to their low molecular counterparts, exert a similar protective effect despite their smaller freezing point depression capacity (Farrant and Woolgar, 1970; Luyet, 1 970). Given the notorious structural side-effects of cryoprotectants (material extraction, intramembranous changes, etc.; Section VII), one wants to keep its concentration as low as possible. In practice 10-30% of glycerol results in a sufficient freezing point depression, so that the time available for ice crystal formation is reduced (Luyet, 1970).

IV. Cryofixation Techniques-Advantages

and Restrictions

Table I1 summarizes the cryofixation techniques available now and the ways in which they were combined with other techniques. Figures 6-8 show examples of results obtained with different cryofixation techniques.

AU

160 -200 73

-150 123

-100 173

-50

ooc

223

273 K

FIG. 5 . Temperature dependence of the heat capacity of different metals used in cryofixation. According to data from Weast (1976).

248

HELMUT PLATTNER AND LUIS BACHMANN TABLE I1 CRYOFIXATION: STANDARD TECHNIQUE VERSUS ADVANCED PROCEDURES (WITHOUT CHEMICAL PRETREATMENTS)

Method Standard freezing

Spray and specimen jet Emulsification

Sandwich Sandwich-cryogen jet Cryogen jet Metal surface High pressure

Procedure Aldehyde fixation + antifreeze + immerse as block or pellet into liquid coolant Spray suspension as small droplets into liquid coolant Suspend sample by stirring in paraffin oil + immerse as pellet into liquid coolant Sandwich between thin support sheets Sandwich between support plates and shoot liquid coolant upon plate(s) Shoot liquid coolant upon sample Press sample onto cold metal surface Use of 2 kbar pressure for 0. I second and shoot liquid coolant upon sample

Principle Binding of cellular water

Increasing surface/volume ratio and cryogen convection, reduction of nucleation in small volume Reduction of heterogeneous nucleation in small volume Increasing surface/volume ratio Increasing surface/volume ratia and cryogen convection Increasing cryogen convection Increasing heat conductivity Changes in nucleation and crystal growth rates, increase of cryogen convection, no Leidenfrost phenomenon

A. CONVENTIONAL TECHNIQUES With conventional or standard cryofixation techniques, bulk specimens are used. About I ~1 of a solution, suspension, pellet, or tissue excisate is placed onto a massive metal holder (e.g., Balzers gold specimen holders of 50-60 mg mass). For freezing the sample is plunged into a coolant, mainly in melting Freon 12 or 22, propane, or melting nitrogen (“slush”). In practice, most samples are further processed by freeze-fracturing; such samples can be cut with a knife. Alternatively bulk specimens can be mounted or pipetted into a hollow cylinder (metal rivets or capillaries with prefurnished fracture sites); such samples are appropriate for fracturing. However, the cooling rates obtainable hardly exceed lo3 Usecond; concomitantly the cryofixation of specimens with the usual water content is poor (Fig. 6), unless the samples were pretreated by an antifreeze (Fig. 7). As soon as the noxious side-effects of gylcerol impregnation were realized, prefixation by aldehydes was introduced (Moor, 1966). It is obvious that aldehydes cannot stabilize structural components without interacting with them, and, thus, submitting them to certain changes, which then might show up as ultrastructural artifacts. Many laboratories then tried to reduce these side-effects by using a brief and “mild” fixation (i.e., low fixative concentration). This now represents the “standard” cryofixation technique for most electron microscopic work in cell biology. The following problems then became

250

HELMUT PLATTNER AND LUIS BACHMANN

FIG. 7 . Paramecium retraureliu cell frozen as in Fig. 6 but after prefixation with glutaraldehyde and impregnation with 206 glycerol. The background granularity is reduced to about 20 to 30 nm. For symbols see Fig. 8 . Unpublished micrograph. Bar = I p.m.

Nevertheless, this “standard” cryofixation technique is still largely used by electron microscopists, especially for “merely” ultrastructural analyses of bulk specimens (mainly by freeze-fracturing). In many cases, with the above restrictions, it can yield valuable ultrastructural information. However, it should be abandoned wherever possible, i.e., where advanced cryofixation techniques can be used and especially where further analyses require one to restrain from any chemical pretreatment. The hazard of morphological artifacts is discussed in more detail in Section VII.

B. ADVANCEDCRYOFIXATION TECHNIQUES The methods listed here allow the cryofixation of native, nonchemically pretreated materials of different kinds.

CRYOFIXATION

249

evident, step by step. (1) “Mild” fixation might induce even more artifact hazards, especially on the membrane level; first, by insufficient stabilization for subsequent antifreeze treatment and, second, by allowing for supravital artifacts within the short time period, where structures are not yet sufficiently stabilized. The latter type of artifact is clearly a major problem, e.g., in exocytosis research (Plattner, 1978). (2) Stronger glutaraldehyde fixation leads to denaturation of membrane proteins (Lenard and Singer, 1968). (3) Even relatively strong prefixation can often not suppress all structural alterations, again mainly on the membrane level, which might occur during subsequent treatment with the usual antifreeze. (4) The application of the double “chemical club,” fixative and cryoprotectant, before freezing leads far away from the original goal of cryofixation as outlined in the introductory section. Materials prepared by such methods are useless for most analytical procedures.

Fic. 6. Nonfixed, noncryoprotected Paramecium fetruureliu cell frozen by dipping as a pellet into Freon 12. Ice crystals, about 200-800 nm in size, distort intracellular details. ci, Cilium; pm, plasma membrane. For further details see Fig. 8. Unpublished micrograph. Bar = I pn.

CRYOFIXATION

25 1

1. Spray-Freezing and Specimen Jet-Freezing (Figs. 8-10) This method was already used some time ago by Williams (1954) for the rapid freezing of virus suspensions. It was considerably refined for freeze-fracture work by Bachmann and Schmitt (1971), so that it could then be used for the rapid freezing of macromolecular and micellar solutions (Bachmann and SchmittFumian, 1973a; Bachmann et al., 19811, cell organelles (Bachmann and Schmitt-Fumian, 1973b; Bachmann et a/.. 1972; Schuler et al., 1978), liposomes (Ververgaert et al., 1973a), and cell suspensions (Bachmann et al., 1972; Plattner et a/., 1972, 1973). The fluid sample is sprayed by a stieam of compressed air through a commercial retouching air brush (as used by graphic artists). The sample splits up into small droplets (e.g., 5-30 pm) which are blown into liquid propane as a coolant. A minimum pressure for spraying is required, not only to generate sufficiently small droplets but also to carry the specimen droplets fast enough through the

FIG.8. Nonfixed, noncryoprotected Pururneciirm ieiraurelia cell after spray-freezing. Segregation artifacts cannot be recognized and the ultrastructural details are no longer obscured by ice crystal damage. mit. Mitochondria; pm, plasma membrane; tr, trichocyst (secretory organelle). Unpublished micrograph from Plattner, Lefort-Tran. and Pouphile. Bar = I km.

252

HELMUT PLA’ITNER AND LUIS BACHMANN

FIG. 9. Freeze-fracture replica from a spray-frozen specimen which is divided into small droplets (diameter -I0 pm) within the butylbenzene “glue” (with a rippled appearance). From Bachmann and Schmitt (1971). Bar = 10 pm.

I compressed air I 2-15atm I

1

thermostat

sample O-ring apertureholder with aperture

Mb

C

d

e

-screw

cap

FIG. 10. Principle of spray-freezing. The sample is (a) sprayed as small droplets into melting propane, which is (b) subsequently removed. The droplets, containing cells, organelles, ot‘ molecules, can be processed by different secondary techniques. The variation presented here (c, d ) is for freeze-fracturing. A “glue” is added at -85°C (c), the paste of glue and spray droplets is transferred onto standard specimen holders (d), and solidified by dipping into LN2. (a-e) According to Bachmann and Schmitt (l973b). The detail (0 on the right is from Kutter er a/. (1976) and shows that the sample can also be injected into the cryomedium through an aperture (specimen jet-freezing). This method can also be used for relatively large cells (Plattner er al.. 1973; see Fig. 8 for an example). A similar device (operated under pressure up to 150 bar) was used also by Briiggeller and Maier (1980) for the vitrification of pure liquid water. When combined with a thermostat unit, this device also allows the specimen to be kept at a constant temperature until it hits the cryomedium.

CRYOFIXATION

253

cold gas phase above the cryogen bath. Larger objects, like suspensions of Paramecium cells (40 X 120 pm; Fig. 8) can be shot in a fine stream through a 50-p-large hole of an electron microscope aperture, fitted onto a glass tube (Bachmann et al., 1972; Plattner et al., 1973). This so-called “specimen jetfreezing” has also been applied for freezing solutions kept at elevated temperatures (Section X). Using specimen-jet freezing with a driving pressure up to 200 bar the best vitrification of pure water from the liquid state has been obtained so far (Briiggeller and Mayer, 1980). With cell suspensions this method requires very careful viability controls to ascertain whether or not the cells are mechanically damaged by the arising shearing forces; these controls must not include the freezing process (Section VIII). Limiting values for the maximal pressures applicable (e.g., 0.5-2 atm) can thus be found for the actual type of nozzle available and for the different types of cells to be processed. After cryofixation the propane is evaporated under reduced pressure. For freeze-fracturing, the spray droplets have to be mixed with a precooled “glue,” such as n-butylbenzene (melting point, mp = -88”C), toluene (mp = -95”C), or best ethylbenzene (mp = -95°C). It should be considered that the specimens might be impaired by recrystallization artifacts, when they remain for too long a time period at a temperature around -80°C. In a final step the glue, containing the frozen spray droplets, is solidified by immersion into LN,. Occasional melting and bubble formation upon insertion of specimens into the vacuum is due to the undercooling of the glue with considerable solution of nitrogen and/or traces of propane. This is avoided by the selection of toluene or ethylbenzene as a glue material. Without the application of a glue the specimens can be subjected to any other secondary technique. For cryosectioning and subsequent X-ray microanalysis Steinbrecht and Zierold ( 1982) have recently successfully used heptane (mp = -90.6”C) as a glue. For the spray-freezing technique a commercial device is available. A modification for processing small quantities of samples was proposed by Ververgaert er al. (1973a) and by Lickfeld et al. (1976) to cope with infectious materials. For those who are reluctant to use a potentially explosive cryogen, such as propane, we mention the article by Stephenson ( 1954) which tells how to avoid such problems. 2. Emulsification Technique Buchheim ( 1972a,b) observed an improved cryofixation of suspensions, like milk, by preparing an emulsion of the sample in parraffin oil via vigorous stirring. A small sample (about 1 pl) of emulsion is then pipetted onto a conventional specimen mount and frozen in one of the usual coolants. Results appear of lower quality than with the other advanced methods and the extraction of hydrophobic substances may also occur. When the application to liver tissue was

254

HELMUT PLATTNER A N D LUIS BACHMANN

attempted (Buchheim and Welsch, 1977), the effect of homogenization on cellular integrity was totally neglected. In our view the interesting aspect of this technique is the following: Since the cooling rates obtainable in a paraffin oil emulsion are necessarily even lower than in conventional techniques, in our opinion the improved cryofixation must be due to the lowered influerice of heterogeneous nucleation by dividing the sample into small compartments (Section 11). 3. Sandwich Freezing (Fig. 11) Very thin sample layers have already been used by Luyet er al. (1962) and by Rebhun and GagnC (1962) in order to improve the quality of cryofixation. When artificial lipid membranes were sandwiched between very thin copper supports, pushed into Freon and analyzed in parallel by low-temperature X-ray diffraction and freeze-fracturing, both types of analyses visualized the advantage of sandwich freezing over conventional freezing (Gulik-Krzywicki and Costello, 1977, 1978). The method also proved appropriate for subcellular fractions and small cells in suspension (Lefort-Tran et al., 1978). In some cases this simple method, which needs no expensive equipment, might allow for adequate cryofixation. It is important to use specimen holders (Cu or Au) of low mass (10 mg) of the type proposed by Gulik-Kryzywicki and Costello (1977, 1978) or Miiller and Pscheid (1978). For subsequent double fracturing the holders must have sufficient mechanical strength. Alternatively the cover piece can be removed before the thin sandwich sample is cut by carefully advancing the microtome knife of the freeze-fracture machine. The thickness of the sandwich can be selected by spacer rings. These can be most easily manufactured by punching out electron microscope grids, which have a well-defined thickness depending on mesh size (Pscheid er al., 1981): Values for grids from VeCo are between 12 pm (300 mesh) and 50 pm (slot grids); 30-pm spacers can be obtained from 75 mesh and 20 pm spacers from 100 or 150 mesh type grids. rc

suspension

Y

n

'

freezing

removal of "hat" and spacer p i e c e

AspPcime

specimen

I support

@

FIG. 1 I . Principle of sandwich-freezing. A small droplet of a suspension is sandwiched between two specimen supports separated by a spacer. Sandwiches can be frozen by dipping into a cryogen (Gulik-Krzywicki and Costello, 1977. 1978) or as in Fig. 12.

CRYOFIXATION

255

Up to now the possible side-effects of sandwiching, e.g., by contact with copper or by anoxia, were largely neglected. Our data obtained with cell monolayer sandwiches (Pscheid et u l . , 1981) indicate that viability can be remarkably impaired when the time period for assembling the sandwich is excessively long (Section VI11). In a more sophisticated set-up the sandwich method can be combined with cryogen jet-freezing (see below). 4. Cpogen Jet-Freezing

When a cryogen is shot onto the surface of a suitable bulk specimen its uppermost layer can be “vitrified” over a range of 5-10 km. For instance, Burstein and Maurice (1978) used a propane jet for freezing the surface of corneal tissue. Due to the impact of the cryogen jet, more delicate surface structures might be impaired.

5 . Sandwich Crvogen Jet-Freezing (Figs. 12 and 13) This method was introduced by Moor et al. (1976) and Muller et al. (1980a). Their device is commercially available. A double jet of cryogen (propane at LN, temperature) is synchronously shot through two opposite nozzles onto both sides of a sandwich sample. The method was applied to solutions, liposomes, organelles, and cell suspensions. Cryofixation is more critical in the case of cell monolayer cultures, if one wants to maintain the cellular arrangements as they have formed during growth on a physiological substrate. As discussed in more detail by Pfenninger (1976) the cryofixation methods applied up to now to cell monolayers involved aldehyde fixation, glycerol impregnation, and sometimes scraping off the cells from the support material. In a few instances, specimen carriers (carbon or collagen-

FIG. 12. Principle of freezing by a double cryogen jet according to Moor ct a/. ( 1976) and Miiller ct a/.(1980).

256

HELMUT PLATTNER AND LUIS BACHMANN

r

r

1

I

I I

I L

J

I

‘0

r-----

iII

I

I 1

FIG. 13. Flow chart of a combined thin-sandwich and cryogen jet-cryofixation method (Pscheid and Plattner, 1980; Pscheid e r a / . , 1981). As presented here the method is designed for moriolayer cell cultures grown on Thermanox plastic sheets. From these a small disc is punched out ( B ) and combined with a low mass object holder [manufactured with a punch (A) from 0. I-mm copper or 0.3mm gold sheets]. A spacer piece is put between object holder and Thermanox in order to protect the cells from being squeezed. Inverted tweezers with an appropriate tip piece (into which the sandwich can be fitted) may be used to facilitate the assembly of the sandwich (C, D).To maintain the cells undamaged the assembly should be performed within a minimum of time, e.g., -10 seconds. The sample is frozen by shooting a jet of liquid propane onto the copper piece (E). Excess of propane is removed by a brief jerk or by a bried flash of cold dry nitrogen gas (F). The sample is mounted under LN2 in a double fracture holder (G). In the case shown here the sample half adhering to the copper carrier can be replicated by heavy metal shadowing (see Fig. 17). the corresponding half adhering to

CRYOFIXATION

257

coated gold disks) were introduced into the cultures and the cells were then fixed and cryoprotected for conventional freezing. As an alternative, Pscheid and Plattner (1980) and Pscheid et al. (1981) proposed the following method for the cryofixation of cell monolayer cultures (Fig. 13), which avoids the application of unusual substrate materials, the detachment from the substrate, and any chemical pretreatment. Cells are grown on plastic sheets (Thermanox) as they usually are for tissue culturing. Samples are punched out with a special punching device, which guarantees that the cells are not squeezed (except for an outermost rim). The sample disk is removed from the culture with some medium still adhering. Then a spacer ring (see above) and finally a U-shaped specimen holder (Cu or Au) are put on the cell layer. This provides a uniform sample thickness, guarantees that the cells are always covered by culture fluid, and avoids any mechanical damage during further processing. A punching device for the production of the metal specimen holders is also commercially available. The assembly of the sandwich can be facilitated by the use of inverted “assembly tweezers,” so that the sandwich assembly requires only fractions of a minute; in that case the cells show no signs of damage (Section VIII). The sandwich is then frozen by shooting melting propane only onto the metal holder. The application of a one-sided cryogen jet is feasable because freezing must not start at the thermically insulated side of the Thermanox sheet (which could occur if a double jet set-up would not be extremely well synchronized). The cryogen jet comes from an LN,-cooled brass vessel which can be instantaneously pressurized by operating a compressed air gun (Fig. 14). The equipment for cryogen jet-freezing of thin sandwiches can be kept so simple that it can be easily manufactured in any workshop and still works very efficiently. If for some reason the sample has to be sandwiched between two copper sheets, the freezing device can very simply be adapted for a double jetting (Fig. 14). Thermocouple measurements have so far given little difference between a one-sided jet on a copper-plastic sandwich and a double jet on a double copper sandwich (Knoll et al., 1982a,b; Bachmann, unpublished results). The problem of synchronizing the two jets within a fraction of a millisecond for the metalhetal sandwich should not be overlooked. The application of propane as a coolant is sometimes handicapped by inadvertent freezing in LN,, while mixtures of propane and propylene or isopentane remain liquid (Jehl et al., 1981). Bearing in mind the low heat conductivity of Thermanox plastic sheets, we the Thermanox can be used for freeze-substitution fixation and ultrathin sectioning (see Fig. 23). This method requires only simple equipment yet gives higher cooling rates than the method presented in Fig. 12 and it is also well suitable for the investigation of solutions and suspensions up to the size of average cells (Knoll e t a / . . 1982a,b).

25 8

HELMUT PLAITNER AND LUIS BACHMANN

4

P

FIG. 14. Pressure vessel for freezing by a one-sided cryogen jet according to Fig. 13. From Knoll ef al. 1982b (see also Pscheid e? al.. 1981). ( I ) Teflon connecting piece for compressed air gun. (2) Lid. (3) Sealing ring (Teflon). (4)Cryogen (propane) container. ( 5 ) Support ring. (6) Cooling rod. (7)Nozzle for one-sided jet. (8) is added as an alternative nozzle for two-sided jet freezing (see text). Micrographs presented in Figs. 15-17 and 23 were prepared with this device according to Fig. 13.

extended the above method by Pscheid and Plattner (1980) and Pscheid et al. ( 1981 ) to other samples (Knoll et al. 1982a,b). Good cryofixation could thus also be achieved with solutions and suspensions of cells up to average size (Figs. 15- 17). Ultrastructural data are corroborated by thermocouple measurements (Table 111) by which a variety of parameters were optimized (Knoll et al., 1982a,b) and which allowed us to achieve very high cooling rates. These were even higher (especially in the critical upper temperature regions) than with a commercial double propane jet-freezing device (measured under identical conditions). On the other hand we could not observe the temperature fluctuations reported by Robards and Severs ( 1981). Thermanox is mechanically so stable that such sandwiches can be easily processed by double fracturing; a modified hinge-type specimen table is required for this purpose (Fig. 13). Furthermore, the fractured half on the Thermanox sheet can be processed by any other technique, such as freeze-substitution fixa-

CRYOFIXATION

259

tion (e.g., to localize the fracture plane through a tissue culture or within organelles), because the material is insensitive to organic solvents (like dehydration and embedding media) and can be easily processed to ultrathin sections by routine ultramicrotomy (Pscheid et al., 1981; Section IX). 6 . Freezing on Cold Metal Su$aces (Fig. 18)

The uppermost 5- 10 pm layer of a bulk sample can be excellently frozen by pushing it against a cold, smoothly polished copper or silver block. Eranko (1954) was the first to use the method at the light microscopic level; then Van Harreveld and Crowell (1 964) and VanHarreveld et al. (1965) refined the method for freeze-substitution fixation (i.e., fixation and dehydration at subzero temperatures) on the electron microscopic level. They used liquid helium as a

FIG. 15. Sarcoplasmic reticulum membrane fraction (kindly provided by Dr. D. Pette, University of Konstanz), sandwich-frozen by a one-sided propane jet according to Fig. 13. From Knoll er (I/. (1982b). Bar = 1 wm.

260

HELMUT PLATTNER AND LUIS BACHMANN

FIG.16. Asolectin liposomes (kindly provided by Dr.R. Benz, University of Konstanz) in dilute buffer, sandwich-frozen by a one-sided propane jet according to Fig. 13. Note the absence of membrane distortions and the even distribution of liposomes down to the smallest size (which would be most sensitive to segregations due to improper freezing). From Knoll era/. (1982b). Bar = I )rm.

coolant and a permanent stream of cold He gas over the metal block to avoid condensation of air. The device has a rod with the specimen fixed on its lower end so that one can press the sample in a vertical direction against the metal block. A similar set-up is used by Heuser (1977; Heuser et al., 1975, 1979) who combined the release of the specimen carrier with an electrical triggering of nerve-muscle preparations; this allows precise time sequence studies of dynamic processes, like exocytosis (Chandler, 1979; Chandler and Heuser, 1979, 1980; Heuser, 1977; Heuser et al. 1975, 1979). The absence of nonetchable cryoprotectants also permits deep etching studies for identifying coated pits (Heuser, 1980) or cytoskeletal elements (Heuser and Kirschner, 1980; Heuser and Salpeter, 1979).

CRYOFIXATION

26 1

This method was recently criticized for the unavoidable mechanical deformation and possible damage of sensitive structures (PintoDaSilva and Kachar, 1980). In addition one must consider the uncertainty of the environmental conditions of cells, e.g., when cultured cells are frozen without a defined layer of culture medium. Boyne ( 1979) and Sitte et al. (1 977) proposed some technical improvements to arrange the sample “bounce-free’’ and at a strictly right angle onto the metal block surface. A simplified version was designed by Coulter and Terracio (1977).

FIG. 17. Rat liver cell from a monolayer culture on Therrnanox (kindly provided by Dr. C. Schudt, University of Konstanz). Frozen as a copper-spacer-Therrnanox sandwich by a one-sided propane jet according to Fig. 13. No ice background granularity or membrane distortions are recognizable (Fig. 23 shows an aliquot after freeze-substitution fixation.) From Pscheid and Plattner (1980). Bar = 0.1 pm.

TABLE 111 COOLING RATESCOLLECTEDFROM LITERATURE DATAO Cooling rate [degrees/sec]

Cryogen I. Propane

Sample 1. Bare thermocouples Size ? 70 pm 0 25 pm 0

Mode of sample handling

Propel propel Dip

0+ -100°C

+20 --* -80°C

2,000 20,000 3,900 7,000

2,100 20,000

2,200 20,000

1.600 20.000

8,000

8,000

5,000

1,100 350 460

1,300 350 520

1,300 350 420

1.100 240 420

Bullivant (1965) Umrath (1977) Costello and Corless (1978)

26,200

I-sided jet

33,000

100 pm

0

Dip

52,000

0 0 360 pm 0 460 wn 0 2. Massive samples 3-mrn 0 solder sphere I .5-mrn 0 araldite Massive freeze-etch holder (Balzers Au planchet) As before

Dip 2-sided jet Dip Dip

Dip Dip propel

Reference

Rebhun ( 1972) Costello and Corless (1978) Pscheid and Plattner (unpublished results) Pscheid and Plattner (unpublished results) Fritzmann and Bachmann (unpublished results) VanVenetie er 01. (1981) VanVenetie er al. (1981) Schwabe and Terracio (1980) Barlow and Sleigh (1979)

0

100 prn

At -80°C

5,900 98,000 25,200

25 pm

100 pm

At -50°C

Dip

I .so0

As before

I -sided jet

2.100

0.7-mm 0 steel needle

Supersonic shot

28,200

15,700

Pscheid and Plattner (unpublished results) Pscheid and Plattner (unpublistid rrsultb)

into cryogen

-3,000b

Monroe er a/. ( 1968)

1 -mg gelatine

Propel

3-mm 0 sample

Dip

3. Small samples 13-pm sandwich between Balzers freeze-etch holders 20-pm sandwich between 0. I-mm Cu freeze-etch holders 75-pm sandwich between 0. I-mm Cu freeze-etch holders 20-pm sandwich between 0. I-mm Cu freeze-etch holders As before

IJ

w m

As before

50-pm sandwich between 0. I-mm Cu freeze-etch holder and Thermanox As before 11. lsopentane

?-sided jet I-sided jet

Dip

1,700 200

-8,000 5.500

200

200

8 .000

-8.000

7.900

500

-5,000 3.700

Fritzmann and Bachmann (unpublished results) Barlow and Sleigh ( 1979)

Miiller et ul. (1980b) Knoll et ol. (1982b)

-9.000

9.000

I-sided jet

8.300

8.900

7.100

Knoll et ul. (1982b)

Tangentional 2sided jet Dip I-sided jet

8,600

9.900

6.300

Knoll et ul. ( 1982b)

3.300

2.900 3.100

I .600

Knoll et ul. ( 1982b) Pscheid and Plattnei (unpublished results)

Dip

1,300

2,400

850

Pscheid and Plattnei (unpublished results)

I . Bare thermocouples 70 pm 0 160 k m 0 size '? 360 p m 0 460 pm 0

Propel Dip Propel Dip Dip

2. Massive samples I .S-mm 0 araldite

Dip

-9,000

1.700

-9.000

45.000 2.900 2.400 I.500 3.000

3.00

3.500

2.300

130

I30

200

I so

Robards and Severs (1981)

Costello and Corless (1978) Echlin (1978b) Rebhun (1972) Schwabe and Terracio ( 1980) Barlow and Sleigh (1979)

Umrath (1977) (continued)

TABLE IIl-Conrinued Cooling rate [degreeslsec]

Sample

Cryogen

2.8-mm 0 sample Massive tissue measured in I-mm depth 3-mm 0 sample 111. Freon 12

Mode of sample handling

Propel Dip

0-

-100°C

+20+ -80°C

2,400 7

8

At -50°C

7

At -80°C

5

Echlin (1978b) Wollenberger et a / . (1960)

Barlow and Sleigh (1979)

250

Dip

Reference

1 Bare thermocouples

Rebhun (1972) Echlin (1978b) Costello and Corless (1978) Fritzmann and Bachmann (unpublished results) Glover and Garvitch (1974) Glover and Garvitch (1974) Schwabe and Terracio (1980) Barlow and Sleigh (1979) Glover and Garvitch (1974) Glover and Garvitch (1974)

Size ? 160 pm 0 70 pm 0 100 pm 0

Propel Propel Propel Propel

pm 0 pm 0 pm 0

Propel Propel Dip Dip Propel Propel

4,500

Propel Propel Propel

840 500 940

Glover and Garvitch (1974) Glover and Garvitch (1974) Glover and Garvitch (1974)

Propel

750

Glover and Garvitch (1974)

190 240 360 460 430

pm 0 pm 0

900 pm

0

2. Massive samples 0.3-1 p1 sample 3-5 pl sample 0. I-mm Cu freeze-etch holder + 0.3-1 pl sample 0. I -mm Cu freeze-etch holder + 1-3 pI sample 3-mm 0 sample

Dip

51,200 47,000 22,000

3,900 2,300 2.500 1.200 280

3.000

2,700

370

1.900

Barlow and Sleigh (1979)

3. Small samples 0.1-mm Cu freeze-etch holder (-sample) I mg gelatine IV. Freon 22

I . Bare thermocouples Size ? 0.2-pm Cu + 0.5-pm Ni

25 pm 0 70 p m 0 360 p m 0 2. Massive samples 1.5-mm0 araldite Massive freeze-etch holder (Balzers Au planchet) surface 0.5 mm depth tissue 2-mm cubes I mm depth 0. I-mm Cu freeze-etch holder + 0.3-1 pl sample

1

3. Small samples 0. I-mm Cu freeze-etch holder (-sample) 10-pm tissue section 20-pm tissue section 40-pm tissue section Thin (?) Cu-sandwich 50-pm sandwich between 0. I-mm Cu freeze-etch holder and Thermanox

Propel

I .200

Propel

900

Propel Propel (Freon I2 or 22 ? ) Dip

Glover and Garvitch ( 1974) Fritzmann and Bachmann (unpublished results)

98,000

70.000

59.000

4.000 22.000

10.500

12.200

10,500

7.600

Rebhun (1972) Escaig et a / . (1977) Pscheid and Plattner (unpublished results) Costello and Corless (1978) Schwabe and Terracio (1980)

Propel Dip

66,ooo 2,400

Dip Propel

230 260

230 300

260 240

170 240

Umrath (1977) Costello and Corless (1978)

Dip Dip Dip Propel

170 200 180 840

210 I30 I10

I80 410 5 20

120 240 I70

VanVenrooij et a / . (1975) VanVenrooij et a / . (1975) VanVenrooij et a/.(1975) Glover and Garvitch (1974)

Propel

1,400

Propel

&pel Propel Propel

Dip

19.500 14.300 11,200 i5,noo 7,400 9,500 2,200 1,OOO 1,400

Glover and Garvitch (1974)

14,300 12,400 7,600

9.500 7,600 5,300

1.100

750

Escaig et a/.(1977) Escaig et al. (1977) Escaig er al. (1977) Costello and Corless (1978) Pscheid and Plattner (unpublished results)

(continued)

TABLE III-Co~itinrred Cooling rate [degreedsec]

Sample

Cryogen

Mode of sample handling

o-,

+20+ -80°C

-100°C

At -50°C

At -80°C

Reference

-

V. Liquid nitrogen (LNd

I . Bare thermocouples Size ? 160 pm 0 160 p m 0

250 pm 0 300 pm 0(?) 360 pm 0 460 pm 0 70 pm 0 I 0 0 pm 0

2. Massive samples 2 . 8 mm 0 I .5-mm 0 araldite 3-mm 0 sample 0.3-1 pl sample 0. I -mm Cu freeze-etch holder + 3-5 pl sample Massive tissue measured in 1 m m depth 3. Small samples 0. I-mm Cu freeze-etch holder (-sample)

Propel Dip Propel Dip Propel Dip Dip Propel Propel

Propel Dip Dip Propel Propel Dip

Propel

70

75

70

-.170

-170

350 25

350 40

240 40

15

40

25

80

I .000

5,000 440 1,900 I00 -170 -170 16,000 14,000 2,100

1.800 350 40 270 460

30

1,100

Rebhun (1972) Echlin (1978b) Echlin (1978b) MacKenzie (1969) Glover and Garvitch (1974) Schwabe and Terracio (1980) Barlow and Sleigh (1979) Costello and Corless (1978) Fritzmann and Bachmann (unpublished results)

Echlin (1978b) Umrath (1977) Barlow and Sleigh (1979) Glover and Garvitch (1974) Glover and Garvitch (1974) Wollenberger et a / . (1960)

Glover and Garvitch (1974)

1 mg gelatine

VI.

Melting nitrogen (LN2"slush")

25 wm 0

Dip

1.000

70 pm 0

Propel Propel

I .ooo

Dip Propel

2.500 2.300

Pscheid and Plattner (unpublished results) Costello and Corless (1978) Fritzmann and Bachmann (unpublished results) MacKenzie (1969) Glover and Garvitch (1974)

Propel Propel

1.400 900

Glover and Garvitch (1974) Glover and Garvitch (1974)

Dip

1.100

0

250 p m 0 300 pm 0 ( 9 ) 2. Massive samples 0.3-1 pI sample 3-5 pl sample 3. Small samples 50-pm sandwich between 0. I-mm Cu freeze-etch holder and Thermanox 1 mg gelatine

Helium I

VIII. Helium I1 IX.

LN2-cooled Al-covered pliers

Fritzmann and Bachmann (unpublished results)

75

1 . Bare thermocouples

100 pm

VII.

Propel

2. Massive samples 3-mm 0 solder sphere 2. Massive samples 3-mm 0 solder sphere 1. Bare thermocouples 400 pm 0

7.900

19,000

15.800

3.800

Propel

800

1,200

1,700

Pscheid and Plattner (unpublished results) Fritzmann and Bachmann (unpublished results)

190

Dip

I30

160

I20

120

Bullivant (1965)

Dip

250

300

210

210

Bullivant (1965)

Compress

20,000

Wollenberger et a/. (1960)

(conrinued)

TABLE III-Conrinued Cooling rate [degreesisec]

Cryogen

Sample

Mode of sample handling

o+ - 100°C

+20+ -80°C

At -50°C

At -80°C

Reference

2. Massive samples 0.7 mm thickness

X. Cold metal surface face LN2-cooled Ag HeI-cooled Cu

Surface of massive samples 1a. thermocouple 360 pm 0 Ib. Electrical capacitance measurements 1a. Thermocouple measurements (size ?) In 15 pm depth (2 msec) In 100 pm depth (8 msec) Ib. Electrical capacitance measurements In 10 Fm depth (2 msec) In 20 pm depth (4 msec) In 50 pm depth ( 12 msec)

Compress

630

690

810

430

Wollenberger et a / . (1960)

Dip Propel

9,200 71,000

Propel propel

23,000 7,500

Heuser et al. (1979) Heuser et 01. (1979)

50,000 25 ,000 8,30Ob

Heuser er al. (1979) Heuser er al. (1979) Heuser et al. (1979)

Schwabe and Terracio (1980) VanHarreveld er 01. (1974)

Propel

a Most data are not directly comparable because of a widely varying experimental set up. In spite of this limitation important trends can be recognized, e.g., when cooling rates are compared for different cryogens. A variety of such conclusions are discussed in Section V. As far as possible, values were derived from published curves (rather than from tables) in order to calculate cooling rates in different temperature regions. Dipping means an uncontrolled insertion of the sample into the coolant. Propel indicates a forced immersion by mechanical gadgets. Occasionally !he speed was measured (Fritzmann and Bachmann. unpublished rcsults, 5 m’ second; Costello and Corless, 1978, 0.5 &second; Glover and Garvitch, 1974, 2 &second). Indications for other temperature regions are set between the columns.

CRYOFIXATION

I

i

269

‘’

FIG. 18. Cold metal surface cryofixation method. The most essential piece is a polished metal block (d). fixed onto a tube (c) with a fitting piece (e). A detail of this is shown at the right side. The metal block (d) is cooled with LN2 (or liquid He; see text) contained in a Dewar (a). Evaporating N2 (He) escapes through tube (0 in the lid (b). To avoid 0 2 condensation on the metal surface, He gas streams through the tubing (g). The specimen is fixed on a rod (p); guiding system, switches etc. are not shown. According to VanHarreveld and Crowell (1964).

Freezing on a cold metal surface was also used by others, namely, in connection with freeze-fracturing (Dempsey and Bullivant, 1976a,b), freeze-substitution (see above), ultracryotomy (Christensen, 1971; Seveus, 1977), and freezedrying (Terracio et al., 1981). 7. High-pressure Freezing (Fig. 19) Elaborate, but not yet commercially available equipment, as developed by Riehle (1968a,b), is required for this technique. It was only very recently that the first micrographs from nonchemically pretreated biological specimens were published (Moor er a / ., 1980). The sample, which can be even a small piece of tissue (but only up to 0.2 mm diameter), is put into a tube which is inserted into a pressure vial. In order to minimize specimen damage the high pressure (2 kbar) is applied only for fractions of a second, while a stream of LN, is shot onto the sample vial. The cryofixation quality, analyzed by freeze-fracturing, does not seem to be totally comparable with that obtained with other advanced techniques (see Moor et al., 1980; Wolf et a f . , 1981). Nevertheless, this particular method deserves special interest due to its potential applicability to bulk specimens. However, previous tests with Euglena cells proved them to be rather sensitive to the high pressures required (Riehle and Hochli, 1973). As far as tissue samples are

270

HELMUT PLA'ITNER AND LUIS BACHMANN P

FIG. 19. Principle of high-pressure freezing. A specimen (SP)contained in a holder (A) with an inside diameter of -0.2 mm and a preformed fracture site is installed in a cooling nozzle (B) within a high-pressure cylinder (D). For a fraction of a second a pressure of 2 kbar is applied, before IAN*is shot onto the sample via the nozzle. According to Riehle and Hochli (1973).

concerned, the excision and handling must entail serious functional and ultrastructural problems and the size is limited to 0.2 mm. These problems must be subjected to careful control before the method can be generally applied in future work. 8. Other Cryofixation Methods Some biological samples have such a high surface-to-volume ratio thal one achieves sufficient cryofixation simply by immersing them into a cryogen bath. Examples are butterfly antennae (Steinbrecht, 1980; Steinbrecht and Zierold, 1982) and fungal hyphae (Howard and Aist, 1979, 1980). Other possible candidates would be isolated renal tubules, which Humbert et al. ( 1977) have demonstrated-albeit with conventional methods-to be amenable to freeze-fracturing. In some cases the handling of the samples during the subsequent processing steps raises a problem. Insect antennae, for instance, were fragmented after freezing and either subjected to freeze-substitution fixation, freeze-fracturing (Steinbrecht, 1980), or ultracryotomy (Steinbrecht and Zierold, 1982). For the last two alternatives the fragments were glued together with ethylbenzerie or heptane, respectively. When an electron microscope grid is dipped into a suspension or solution, these can be perfectly frozen by dipping into a usual cryogen for direct use in

CRYOFIXATION

27 1

low-temperature electron microscopy (Heide and Grund, 1974; Taylor and Glaeser, 1976; Dubochet and McDowall, 198 I ) . There are two reports in the literature on cryoballistic methods. A sample cylinder is cut out from the tissue with a hypodermic needle, which is shot through the sample into a cryogen bath (Monroe et al., 1968). Without referring to this work Chang et al. (1980) have introduced a similar method. Apparently there is very little experience on how useful such methods really would be. As an alternative cryomedium Fernandez-Moran (1960) tried to exploit the properties of superfluid helium I1 (2K). The heat transfer at the phase boundary, however, is rather poor (Mendelson, 1956). Indeed thermocouple measurements showed no improved cooling rates (Bullivant, 1965, 1970). The use of the rather expensive helium I1 seems, therefore, not justified (see Table 111). Several new gadgets which may turn out to be useful for sandwich- and metal surface-freezing were recently discovered by Escaig ( 1982).

V. Cooling Rates Cooling rates relevant to cryofixation have been reported by a large number of authors (Table 111). It is impossible to compare the published data on an absolute scale, since the experimental set-up was very different. This includes the size of the thermocouple, the size and the composition of model specimens, the dipping or propelling speed, and the time resolution of the registration unit. Frequently, when cooling rates are given, the temperature range for which they are valid is not indicated. The main value of these investigations is to have measured the relative cooling rates for different crfogens and for the various techniques by which they are applied. In addition they can give at least an estimate of the actual cooling rates obtained by advanced techniques. From the data compiled in Table 111 one can derive the following conclusions ( 1 ) In the range between +20 and -80°C the best cooling rates are obtained in propane close to LN, temperature, closely followed by other hydrocarbons and by the Freons. The cooling rates in melting N, (“nitrogen slush”) are definitely lower, although much better than in LN, (see also Fig. 20). (2) If the cryogen is shot onto the specimen or if the specimen is forcefully injected into the cryogen, the cooling rates can be enhanced by orders of magnitude. (3) In “standard” freezing (Section IV,B,I) cooling rates up to lo3 K/second are available. Advanced techniques using quenching fluids allow rates up to 104-10s K/second. Similar rates are obtained using cold metal surfaces (Section IV,B,6). Although the advantage of a forceful injection of samples was utilized quite long ago (Rebhun and Gagne, 1962; Zalokar, 1966), this was hardly accepted for practical work. Since this would be a very simple way of improving cooling

272

HELMUT PLATTNER AND LUIS BACHMANN

10

20

30

t [ms]

FIG.20. Cooling curves obtained with a bare thermocouple (chromel-alumel, -50 pm) dipped by hand into different cryogen fluids. Unpublished data from Pscheid and Plattner as well as from Fritzmann and Bachmann.

rates, attempts along these lines should be encouraged. In contrast, measurements by Schwabe and Terracio (1 980) indicated that the stirring of the cryogen did not exert an effect similar to specimen injection or cryogen jetting (although it appears possible that this could be so only for very simple reasons, such as the geometry of the sample and the stirring velocity etc.). Extremely forceful “stirring” with the use of a special centrifuge rotor can increase cooling rates by a factor of 10 (Barnakov, 1982). As already pointed out high cooling rates are not the only criterion for the selection of a technique. The reduction of heterogeneous nucleation is an alternative for some type of specimens (Section IV). The choice will also depend on the secondary technique. Since the actual cooling rate at a given point in a small specimen (diameter 1 mm) cannot be measured, freeze-fracture replicas must serve as indicators (Figs. 3, 6, and 8). In a few cases, rates within a bulk sample (frozen on a cold metal block) were determined by capacitance measurements (Heuser et af., 1979; VanHarreveld et al., 1974). Very high rates up to lo5 Wsecond were observed, but only for the outermost layer. Concomitantly, the ultrastructural preservation was optimal only over a thickness of several microns. In summary,

CRY OFIXATION

273

physical measurements and the electron microscopic evidence substantiate the predictions that cooling rates required for sufficient cryofixation of highly hydrated biological material must be in the 104-105 Khecond range (see e.g., Bullivant, 1970; Mazur, 1960; Moor, 1971). For biochemical work the most frequently used and quoted “fast” freezing method is the application of cryopliers, first used by Wollenberger et al. (1960). The obtainable cooling rates are low and can be checked in Table 111.

VI. “Resolution” of Cryofixation One should differentiate between the “spatial” and “time” resolution (for dynamic processes) of cryofixation. The spatial resolution obtainable with a cryofixation method would be equivalent to the size of the area within which solute material has been displaced (segregation zones). The size of these areas varies considerably, not only depending on the type of cryofixation used but also on the water content and on the depth within a sample (VanVenrooij et al., 1975). Evidently spatial resolution can be analyzed best upon freeze-fracturing, where the sample stays at a sufficiently low temperature. The best frozen regions might be devoid of any ice background, when visualized by heavy metal replication. The smallest ice crystals observable are 10 nm in size; they become especially visible after a short period of sublimation (“etching” at - 100°C). An ice crystal size of up to 30 to 50 nm is still considered as acceptable. The extent of such “useful zones” from the periphery, where freezing conditions are optimal, to deeper sample layers depends again on the size and geometry of the sample and on the freezing technique. The ice granularity should be as fine as possible, not only for ultrastructural work but also for elemental analyses. The spatial resolution provided by cryofixation is one of the limting parameters for the quantitative localization of soluble components by X-ray microprobe analysis. For most investigations, the material is frozen simply by immersion into a cryogen. It is striking that “nitrogen slush” was very much favored by workers in this field, although other quenching fluids are clearly superior (Fig. 20, Table 111). As discussed above, freezing by simple immersion of small tissue pieces into any cryogen entails large-scale phase separation and redistribution artifacts. At present, this drawback appears to be deliberately neglected due to the restricted resolution of microprobe analyses in the sense of composition-tostructure correlation. This is, however, valid only for bulk specimens and it is clear that the higher resolution obtainable with ultrathin frozen-hydrated or freeze-dried sections (see Chandler, 1977; Echlin and Galle 1975; Hall, 1979; Lechene and Warner, 1979; Somlyo and Silcox, 1979) would require the full potential of advanced cryofixation techniques, provided that segregation during processing and analysis is minimized. Sufficient cryofixation can be obtained only within a useful zone (as defined

-

274

HELMUT PLATTNER AND LUIS BACHMANN

above) of a 5- to 10-pm-thick surface layer, when bulk samples are dipped into a cryogen. When applied as a cryogen jet to tissue surfaces, propane produces a useful cryofixation zone of about the same thickness (Burstein and Maurice, 1978). This useful zone may be, if at all, only slightly broader (10-15 ym) when a bulk sample is pressed onto a cold metal block (Dempsey and Bullivant, 1976a,b; Heuser et al., 1975, 1979; VanHarreveld and Crowell, 1964; VanHarreveld et al., 1965, 1974). The depth of the vitrified zone is restricted by the poor heat conductivity of the frozen cells. Only the outermost surface zone (< I ym) might have experienced cooling rates of up to loJ Kkecond (Heuser et a / ., 1979). True vitrification seems to take place in this zone depending on the water content. A real alternative, applicable also to small tissue pieces (200 pm), could be provided only by a method which makes use of other effects. Influencing the nucleation and crystal growth rate ‘‘high-pressure freezing” could gain some importance (with the still existing drawbacks outlined above). Freeze-fracture micrographs which were recently obtained from small tissue samples by the high-pressure method (Moor et al., 1980; Wolf et al., 1981) display a sufficiently fine ice background. Cryoprotective agents also enhance the “spatial resolution,” but at the price of the interference of a chemical pretreatment (see Section VII). “Time resolution” is intimately connected with the cooling rate. Only fast freezing methods appear appropriate to catch fast processes. Among them are the ciliary beat (Barlow and Sleigh, 1979), intracellular movements, membrane fusion processes, and redistribution or reorientation of membrane components. The need of high time resolution is extensively documented for neurotransmitter release (Heuser, 1977; Heuser er al., 1975, 1979) and other exocytotic processes (Chandler and Heuser, 1979, 1980). These processes may last only milliseconds. The problem is mainly to pass as quickly as possible through the temperature range in which lipid phase transitions can occur. Therefore, the uppermost regions of the cooling curves (starting from a physiological or ambient temperature) are most relevant. Under these aspects the time resolution can be estimated to be between 0.1 and 1 msec with the best methods available. It is still a matter of debate whether this is sufficient to catch the most short-lived stages, e.g., of exocytotic membrane fusion (see Plattner, 1978, 1981; PintoDaSilva and Kachar, 1980). The problem of structural rearrangement is also important for work with membranes or with liposomes (see also Section X,B). This is another reason to include the temperature region above 0°C into the cooling rates (Section V).

-

VII. Ultrastructural Side-effects of Chemical Pretreatments Over one decade ago, the first critical reports were published on the ultrastructural artifacts that arose as a consequence of chemical pretreatments, routinely

CRYOFIXATION

275

used by most people before freezing according to standard methods. A few points were already mentioned in Section IV. A large number of papers along these lines appeared, mainly in combination with freeze-fracturing. This is indeed an important problem, although it should be clear by now that a variety of alternative advanced cryotechniques have been available over several years, for many problems, for many objects, and for the further processing by a variety of secondary techniques. Some types of ultrastructural artifacts, due to chemical pretreatments, were reviewed by Hudson et al. (1979) and Stolinski and Breathnach (1976) as well as in the volumes by Benedetti and Favard (1973), Rash and Hudson (1979), and Stolinski and Breathnach (1975). We selected a few reports for the following exemplary discussion of this topic. It was important that advanced techniques allowed us to make one step “behind the scene” and, thus, to pinpoint and to control such artifacts. This will appear even more important when we have recognized that it is totally impossible to foresee the type and extent of ultrastructural changes which might occur in a certain case. Bovine spermatozoa were freeze-fractured by Plattner ( 197 1 ) either after rapid freezing or after slow freezing (following glycerination, as for artificial insemination, i.e., with a high postthaw survival rate). The occurrence of “intramembranous segregation,” as it was called, of membrane-intercalated particles was considered as an artifact which showed up selectively after slow freezing (Plattner, 197I ) . Subsequently similar phenomena were frequently reported. In some cases they could then be interpreted as a lateral displacement of membraneintegrated proteins by lipid phase separation (Kachar et al., 1980; Kleemann and McConnell, 1974; Speth and Wunderlich, 1973; Verkleij et al., 1972; Wunderlich et a / . , 1973, 1974). Now we know that such effects may easily occur when biological materials are incubated in glycerol at +4”C (for minimizing autolytic changes), especially when no aldehyde fixation was performed before. Even this response, again, might depend on the physiological condition of the cells (Robinson et a / . , 1979). Even if one cannot apply any of the advanced cryofixation techniques, one should in all these cases perform at least some parallel control experiments by simple standard freezing without prefixation and/or antifreeze impregnation. A rearrangement of lipid components during phase separation was also directly visualized (Grant et al., 1974; Shimshick et al., 1973; Ververgaert et al., 1973a, and many others since then). Interestingly, we found that this thermotropic effect does not occur in all membranes (Schuler et al., 1978) and that the normal freezing method (i.e., plunging a 1-pl sample into a coolant) does not provoke any visible displacement of membrane-intercalated particles. In contrast, with some lipids, such rearrangement phenomena might take place during standard freezing, but they can be suppressed by fast freezing methods (Ververgaert et al., 1973a; VanVenetie et al., 1981).

276

HELMUT PLAITNER AND LUIS BACHMANN

It has been known for quite a while that the usual antifreeze agents, like glycerol, can be very toxic to some cells (Richter, 1968) and that they can also induce gross ultrastructural damage (Fineran, 1970). For a few cell types both aspects were analyzed in parallel by Plattner et al. (1972, 1973) with the use of spray-freezing as a control. More thorough analyses revealed that the response of glycerol also depends upon the functional state of the cells (Moms, 1976a,b). It has also been known for a long time that a prolonged incubation of tissue excisates in physiological solutions or the nonintentional incubation under anoxic conditions provokes gross ultrastructural defects (Plattner, 1970). Precisely for this reason, it is amazing that much of the freeze-fracture work still continued to rely on the transmitted methodology. At the membrane level the following disturbances were noted after glycerination (1) Some cells react by a rearrangement of membrane-intercalated particles (MIP) or even by a formation of totally MIP-free blisters; lymphocytes are a well known example of this (DeGroot and Leene, 1979; McIntyre er al., 1973, 1974). Interestingly, glycerination can even be used for the preparation of cell membrane subfractions which are devoid of membrane-integrated proteins (VanDen Burgh, 1977). (2) In some cases the MIP rearrangement in the cell membrane is dependent on the functional state of the cells (Eger and Rifkin, 1980). (3) With still other cell types the cell membrane shows no such reaction, whereas certain endomembranes, like vacuoles in yeast, suddenly display a paracrystalline MIP arrangement (Niedermeyer and Moor, 1976; Niedermeyer et al., 1977). (4) It has also been reported that glycerination would reduce the number of MIP in membranes (Pricam et al., 1977). (5) Like spermatozoa (see above), erythrocytes, frozen for maximal survival in clinical use, also show considerable rearrangements of MIPS (Allen and Weatherbee, 1980). This indicates the restricted value of postthaw survival rates for judging the quality of cryofixation (see also Section VIII). There is very little experience available for polymeric, nonpermeant antifreeze agents; some relevant points are discussed above and in Section IX,D. The formation of gross ultrastructural artifacts during antifreeze impregnation led to the recommendation (see Moor, 1966) of stabilizing the structures by a previous aldehyde fixation. For the freeze-fracturing of tissues this is still a common practice (Stolinski and Breathnach, 1975) because of the limited applicability of “advanced” cryofixation techniques. Formaldehyde is only rarely used (Furcht and Scott 1974, 1975; Scott et al., 1977). Glutaraldehyde is more generally applied, preferably at low concentrations, for a “mild” prefixation with the aim of making the cells tolerable to the subsequent antifreeze treatment. Meanwhile, the following side-effects of prefixation were recognized. (1‘) The fracture plane might change (Furcht and Scott 1975; Nermut and Ward, 1974; Scott et af., 1977). (2) The number of MIP, i.e., membrane-integrated particles (Furcht and Scott, 1975; Parish, 1975), or of surface particles (Wendelaar-Bonga

CRYOFIXATION

277

and Veenhuis, 1974) might be reduced. (3) The freeze-fracture appearance of tight junctions (compare Wade and Karnovski, 1974, and VanDeurs and Luft, 1979) and also that of gap junctions (Raviola et al., 1980) might be altered. (4) The partition coefficient (as defined by Satir and Satir, 1974) of MIP is also liable to changes, namely, in a specific manner with specific MIP populations (Lefort-Tran et al., 1978). (5) The notorious problems of visualizing the polysaccharide materials of the bacterial capsule or of the eukaryote glycocalyx (see Luft, 1976; Parsons and Subjek, 1972) might also be caused partly by a change of these materials during the usual procedures (Plattner et al., 1973). (6) Some of the membrane proteins can no longer be recognized on freeze-fracture replicas because, as a consequence of chemical crosslinking, they are split up into fragments during freeze-cleaving; this fact was established by gel electrophoresis of freeze-fractured membranes (Fisher, 1978; Edwards et al., 1979). (7) Some regular particle arrays, such as the cell membrane of the alga Chlamydoborrys, are considerably distorted after fixation and freeze-fracturing (Lefort-Tran et al., 1978). (8) As discussed by Plattner (1978, 1981), aldehyde prefixation leads to a supravital triggering of exocytotic events, provided secretory granules are positioned close to the plasmalemma. Results obtained with the application of standard techniques to fibroblasts (Bretscher and Whytock, 1977) and capillary endothelial cells (Casley-Smith, 1980) also illustrate this artifact hazard. Beyond this, prefixation by aldehydes cannot guarantee the abolition of all ultrastructural changes caused by the subsequent antifreeze impregnation. ( 1 ) Particle-free blebs were reported to occur despite prefixation in early embryonic cells (Stolinski et al., 19789 and in fibroblasts (Hasty and Hay, 1978; Hay and Hasty, 1979). Only in lymphocytes is this effect stopped by prefixation (McIntyre et al., 1973, 1974). (2) Quite differently, preexisting particle aggregates are dispersed by prefixation in some bacterial membranes (Arancia and Trovalusci , 1978; Arancia et a f . , 1980). Another notorious example of a very labile structure is the “mesosome” of bacteria. Its structural appearance depends largely on the pretreatments used. Whereas mesosomes are only scarcely found in cells frozen directly in the native state, they are abundantly present in cells after chemical fixation (Fooke-Achterrath et a l . , 1974; Ghosh and Nanninga, 1976; Higgins et al., 1976; Lickfeld and Achterrath, 1972; Nanninga, 197I ) , glycerol impregnation (Higgins and DaneoMoore, 1974), or other sudden changes of the physicochemical environment (Lickfeld and Achterrath, 1972). The mesosome was therefore sarcastically designated a “technicosome” (Fooke-Achterrath et a / ., 1974). In fact, it represents a highly reproducible structural equivalent of a specialized cell membrane region which becomes visibly segregated in response to a variety of pretreatments. For a discussion on the present uncertainty on possible side-effects from treatment with polymeric antifreeze agents see Franks (1977) and Section IX,D. In conclusion, one cannot foresee the kind of change which might occur in

278

HELMUT PLATTNER A N D LUIS BACHMANN

connection with prefixation and/or impregnation with the usual cryoprotective agents. This alone would justify all the different attempts which were undertaken up to now in order to avoid any chemical pretreatments by the development of advanced cryofixation techniques. However, we know from our personal experience that such techniques were, and still are, not always fully appreciated. We repeat that our minimum option would be to use standard freezing without chemical pretreatments as a most simple control experiment.

VIII. Tests for the Quality of Cryofixation A. PHYSICAL TESTS In practice there are different ways to check the quality of cryofixation achieved. The simplest way is by freeze-fracturing, which avoids the hazard of secondary ice crystal formation (recrystallization). A dilute solution (5%) of glycerol has widely been accepted as an extremely simple, sensitive, and reproducible specimen to test the quality of cryofixation. It appears that freezedrying, freeze-substitution (chemical fixation at low temperature), and particularly freeze-thawing (melting of frozen tissue in a chilled fixative) are less appropriate as test methods for cryofixation, because some rearrangements of materials must occur with all these techniques. This holds true especially for freeze-thawing which allows for considerable recovery from ice crystal darnage (Baker, 1962; Cummins and Loper, 1965). In contrast freeze-drying and -.substitution are liable to recrystallization. ESR spectroscopy (Briiggeller and Mayer, 1980) has already been mentioned as a sensitive test (Section I). Other test methods used are X-ray (Gulik-Krzywicki, 1975) or electron diffraction (Taylor and Glaser, 1976) and differential thermal analysis ( Rasmussen and Luyet, 1969). TESTS B. VIABILITY At the beginning of freeze-etching, viability tests were made in order to demonstrate that freeze-fracturing represents a true-to-life copy of potentially viable cells (Moor, 1964). However, Mazur (1960, 1977; Mazur et al. 1977; Bank and Mazur 1973), Farrant (1977; Farrant et al., 1977a,b), and others (Luyet, 1970; Morris, 1976a; Sakai et al., 1968; Withers, 1978) have denionstrated that viability after thawing is an interdependent result of freezing and thawing conditions. In many cases slowly frozen cells, which are dehydrated, shrunken, and distorted by extracellular ice crystal growth, yield higher survival rates than quickly frozen aliquots (Bank and Mazur, 1973; Farrant et al., 1977a; Withers and Davey , 1978). With the latter, intracellular recrystallization upon rewarning is a more serious limiting factor for survival. (The postthaw survival rate also depends on the growing conditions: Cottrell, 1981; Morris, 1976b.)

CRYOFIXATION

279

Consequently, there is not necessarily a direct correlation between the intravital structure visualized by freeze-fracturing after rapid cryofixation and the vitality after freezing-thawing (Bank and Mazur, 1972; Plattner et al., 1973; Bank, 1974). Therefore, in the context of cryofixation, according to the criteria relevant for ultrastructure research (Section I), viability tests are feasible only to ascertain whether or not the structural integrity of cells was maintained throughout all the manipulations which might be necessary until the moment of freezing (Plattner et al., 1973). The following viability tests can be applied. Most simply, motile cells (e.g., protozoa, spermatozoa) indicate their vitality directly in the light microscope. Some isolated plant cells (e.g., algae, yeasts) can be plated onto Agar for the purpose of counting the percentage of dividing cells. This allows one to quantify the effects of antifreeze treatment (Moor, 1964; Plattner et al., 1972, 1973: Withers, I978), of different spraying conditions for spray-freezing (Plattner et al. , 1972, 1973), or of the high pressure required for high-pressure freezing (Moor and Hochli, 1970; Moor and Riehle, 1968; Riehle and Hochli, 1973). Similarly, vitality tests are now required to analyze the suitability of polymeric antifreeze agents. Allen and Weatherbee (1979), Echlin et a / . (1977), and Schiller et al. (1978b) have made the first attempts along these lines. The permeability of cells is another criterion for their structural integrity. Tests for permeability can be used at the light microscopic (trypan blue: Phillips, 1973) or electron microscopic (microperoxidases: see Plattner et a/., 1977) level for membrane leaks in the range of 2 1 nm. These compounds, which should not have any toxic side-effects on their own, proved useful to optimize a cryofixation technique. For example, this was done with the combined thin-sandwich and cryogen jet method, applied to cell monolayer cultures (Pscheid et a / . , 1981). When all manipulations (except for freezing) were performed in the presence of these tracers, all cells (except for the most marginal ones along the cutting circle) perfectly excluded the tracers from their cytoplasm. As a positive indication of vitality, microperoxidase was even seen to be actively ingested via coated pits of monolayer cells processed for sandwich-freezing (Pscheid et af.,198I ) . Another critical test would be the determination of the metabolism of small substrate molecules, which are excluded only by an intact cell membrane (succinate: Burr et a l . , 1975). A less rigorous test would be the assay of released soluble enzymes, e.g., lactate dehydrogenase or a change and reattachment to a substratum of the ATP-ADP ratio (Baur et al., 1975). When cell division was compared with a nigrosin dye exclusion test for viability, the division test was superior to the permeability test (Taylor et a l . , 1974). C. GENERAL REMARKS

In analogy, the functional integrity of isolated cell organelle fractions should also be ascertained before freezing. For instance Lang and Bronk (1978) found

280

HELMUT PLATTNER AND LUIS BACHMANN

that isolated mitochondria perform their orthodox-to-condensed state transition within about 20 seconds. Rapid cryofixation facilitated catching the various stages. In practice, however, these demands are only rarely taken into consideration as far as isolated subcellular fractions are concerned. In conclusion, it is important to control the hazard of all the chemical and/or physical manipulations one has to make before cells or organelles can be “optimally” frozen. Such hazards are physical forces, lack of oxygen in thin layers, or contact with toxic fluids (fixatives, cryoprotectants) or metals etc. Viability after freezing and thawing is not a relevant criterion for the preservation of ultrastructure in a close-to-life situation.

IX. Combination of Cryofixation with Other Techniques Table IV as well as Fig. 21 list a variety of secondary techniques which are available for further ultrastructural analysis of materials prepared by cryofixation. Three categories are made: routine (solid frames); nonroutine but applicable (dashed); more “futuristic” but possible techniques (dotted frames), some of which are in the developmental stage. Only a few representative examples can be indicated in the footnote to Table IV and Fig. 21 and the literature compiled there is thought to be a help for further reading. All these methods depend in one way or another on successful cryofixation as the primary step. Special subsections will be devoted to the most important problems relevant to ultrastructure research. Important achievements were made in the past few years on the instrumental sector. Among them are low-temperature transfer and cold stage units for bulk specimens to be analyzed in scanning electron microscopes (Echlin, 1978b; Fuchs, 1979; Fuchs et al.. 1978; Lechene et a l . , 1979; Nei et al., 1974; Piiwley and Norton, 1978; Schafer and Zierold, 1978; Zierold and Schafer, 1978). Nei, Pawley, as well as Schafer and Zierold can directly transfer freeze-fractured and coated material into the microscope; units of this kind are also commercially available. Even more important is the considerable progress which has been recently achieved with the production of thin sections and with the analysis of thin specimens (mostly after freeze-drying) in conventional and scanning transmission electron microscopes. The low-temperature analysis of molecules and molecular assemblies (see point 19 of Fig. 21) and the high-resolution X-ray microanalysis of soluble constituents (see also Section VI) are also possible in conjunction with fast freezing. When cutting is performed at a sufficiently low temperature, no melting occurs (Karp et a l . , 1982). Ultramicrotomes with attachments for cryosectioning are now available from a variety of firms and pertinent literature is indicated in point 13 of Fig. 21. Lowtemperature transfer and cooling stages for (scanning or conventional) transmission electron microscopy were designed by Freeman et al. (1980), Hayward and

28 I

CRYOFIXATION

Glaeser (1980), Hax and Lichtenegger (1981), Heide (1981), Heide and Grund (1974), Hutchinson and Borek (1979), Knapek and Dubochet (1980), Lichtenegger and Hax (1980), Talmon e t a / . (1979, 1980), Taylor and Glaeser (1976), and Zierold et al. (1981). Cooling stages allow one to keep the visible beam damage (monitored by the electron diffraction pattern of hydrated frozen crystals) low during a long time of observation (Freeman et al., 1980; Hayward and Glaeser, 1980; Kellenberger, 1978; Knapek and Dubochet, 1980; Taylor and Glaeser, 1976). Some efforts have also been made to combine such cooling stages with cryolenses in a thermally equilibrated system (Dietrich et al., 1977). A. CRYOFIXATION AS A TOOLIN CYTOCHEMISTRY When combined with different pretreatments, cryofixation can be used to obtain information about local chemical properties. Some of these experiments do not require cryofixation procedures of the advanced type and should, therefore, be discussed here only very briefly. B . COMBINATION WITH FREEZE-FRACTURING When erythrocytes are adsorbed as a monolayer onto a substrate and then frozen and freeze-cleaved, one can analyze selectively outer split membrane halves (Fisher, 1975; Fisher and Branton, 1976). Their biochemical composition such as the cholesterol content, can be correlated with that of the total membrane (Fisher 1976, 1978), so that information about the “sidedness” of membranes can be obtained. Some time ago, Branton (1969, 1971; Branton and Deamer, 1972) started to identify membrane-intercalated particles (MIP) as membrane-integrated proteins which become exposed after splitting up the bilayer by freeze-fracturing. A sensitivity to proteolytic enzymes was one argument (Branton, 197 1). Since the first attempts by Segrest et a / . (1974) and by others, the identity of MIP was established mainly by reconstitution with liposomes. It became clear that MIP represent in most cases oligomers of identical or different polypeptides, which are embedded by hydrophobic bonding within the lipid bilayer. The essential conclusions from this type of analyses have already been discussed in detail by Gulik-Krzywicki (1975), Zingsheim and Plattner (1976), as well as by Verkleij and Ververgaert ( 1978). Recently membrane-integrated Na+ ,K -ATPase has been analyzed very thoroughly (Deguchi e t a l . , 1977; Haase and Koepsell, 1979; Skriver et al., 1978). The oligomeric structure of various types of MIP was analyzed by rotary shadowing at a shallow angle (Margaritis et a / ., 1977); this method, in conjunction with cryofixation and freeze-etching or freeze-fracturing (see Section I), can therefore also reveal some details on the molecular ultrastructure of membrane components. Further identification of intramembranous structures in siru could probably be +

TABLE IV APPLICATION OFADVANCED CRYOFIXATION TECHNIQUES (WITHOUT CHEMICAL PRETREATMENTS) TO ULTRASTRUCTURAL ANALYSES~.~ N N 00

Solutions, suspensions. and emulsions

2 Organelle and membrane fractions

FE

FE

1

Cryofixation method

I. Spray and specimen jet

3 Cell suspension cultures

4

5

Monolayer cell

cultures

Tissues

FE FD

11. Emulsification 111. Sandwich

IV. Sandwichxryogen jet

V. Cryogen jet VI. Metal surface

FE FE XRD FE

-

-

-

C

C FE FE FS

FE FS

FE FE FS NS

FE FS

-

-

FS

FE

FE

FD

FS

FE FS CUM+XMA

VII. High pressure VIII. Cryoballistic method

N

2

FE

Not done

FE

-

-

-

FE FS

" C. Cryomicroscopy; CUM, cryoultramicrotomy: FD, freeze-drying: FE, freeze-etching; FS. freeze-substitution: NS. negative staining; XMA, X-ray microanalysis; XRD, X-ray diffraction; -, not applicable. We indicate here only examples from the literature: other applications would occasionally also appear possible. Examples: (1-1) Bachmann and Schmitt (1971). Bachmann and Schmitt-Fumian (1973a). Bachmann et a / . (1972). Junger and Bachmann (1977). (1-2) Bachmann and Schmitt-Fumian (1973a). Bachmann e t a l . (1972). Gebhardt er a/. (1977). Lang and Bronk (1978). Lefort-Tran et a/.(1978), Schuler et a/. (1978), VanVenetie et 01. (1980). Ververgaert et a / . (1973). (1-3) Bachmann er a/. (1972). Lefort-Tran et a/. (1978). Lickfeld et a/. (1976), Pfaller (1978). ffaller and Rovan (1978). Plattner (1971). Plattner et a/. (1972, 1973). (11-1) Buchheim (1972a.b). (111-1) Costello and Gulik-Krzywicki (1976). Gulik-Krzywicki and Costello (1977, 1978). Talmon et a/.(1979, 1980). (111-2) Costello and Gulik-Krzywicki (1976),Gulik-Krzywicki and Costello (1977. 1978). Lefort-Tran et a/. (1978). Morel et 01. (1980). Talmon et al. (1979, 1980). (111-3) Heide and Grund (1974), Lefort-Tran et ol. (1978). Rebhun and Sander (1971). (111-4) Pscheid and Plattner (1980), Pscheid et a / . (1981). (IV-1) Knoll ef a / . (1982a.b). Moor ef a/. (1976). Miiller er a/. (1980). (IV-2) Knoll et a / . (1982a.b). Miiller et a/. (1980). Meister and Miiller (1980). (1V-3) Knoll etal. (1982a.b). Giddings and Staehelin (1980). Giddings e t a / . (1980). Miiller er a/. (1980). (IV-4) Pscheid and Plattner (1980). Pscheid et o/. (1981). (V-5) Burstein and Maurice (1978). (VI-3) Boyne (1979). Ornberg and Reese (1979, 1980). (VI-4) Artificially attached cells, Chandler and Heuser (1979, 1980);grown monolayers, Heuser and Kirschner (1980). (VI-5) Christensen (1971), Coulter and Terracio (1977). Dempsey and Bullivant (1976a.b). Eranko (1954). Franzini-Armstrong e t a l . (1978). Heuser (1977. 1978, 1980), Heuser and Salpeter (1979), Heuser et a/. (1974, 1975. 1479). Ichikawa et a/. (1980). Raviola et al. (1980). Spriggs and Wynne-Evans (1976), Terracio et a/. (1981). VanHarreveld and Crowell (1964). VanHarreveld et a/. (1965, 1974). (VII-I) Moor and Riehle (1968), Riehle (1968a,b), Riehle and Hochli (1973). (VII-3) Moor and Hiichli (1970). (VII-5) Moor et a/. (1980). (VIII-5) Chang et al. (1980). Monroe era/. (1968).

2

n

freeze-fracture, f reeze-etching 3

9 6

drying

ition ............................

4

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

11 -------

................. :ion pre

~

lcipi tat ...........

--------

5 1 - 1 I staining1 zroup-specific I L -I 'decoration i n I b l---------,,I t r a h i g h vacuum'

- - __

18

............ i............ autoradiogra@iy :

. on replica- ultra-: i thin section sand-: : wi& ............................

285

CRYOFIXATION

Fic. 21, Solid frames, routine methods. Dashed frames, methods with restricted use or experience. Dotted frames, not or only rarely done. EM, Electron microscope. ( I )This review. (2) Only a small selection of reviews can be mentioned, which are relevant either for the method or for the interpretation of results: Benedetti and Favard (1973). Branton and Deamer (1972). Branton and Kirchanski (1977). Bullivant (1973). Moor (1966. 1969). Orci and Perrelet (1975). Rash and Hudson (1979). Sleytr and Robards (1977). Steere (1969), Stolinski and Breathnach (1975). Umrath (1977). (2 + 6) Nei et a / . (1974). Pawley and Norton (1978). Schafer and Zierold (1978). Zierold and Schafer (1978). (2 + 6) Including chemical processing: Haggis (1978). Haggis and Phipps-Todd (1977). (2 -+ 9) Heuser and Kirschner (1980). (2 + 9 + 6) Brooks and Haggis (1973). (2 + 21 + 23) Hereward and Northcote (1972, 1973). Heuser et a / . (1975). Pscheid and Plattner (1980). Pscheid et a / . (198 I ) ; appropriate to control fracture level during freeze-fracturing. (2 + 2 I -+ 23 + 14) (or+ 17, or+ 18) Rash (1979). Rash e t a / . (1980). (3) Fisherand Branton (1973, 1976). Rix et a / . (1977), Schiller et a / . (1978). (4)For proteins see Bachmann and Schmitt-Fumian (1973a). Junger and Bachmann (1977). Bachmann et a/. (1974d,b, 1975); for nucleoproteins see Finch and Klug (1976), Lerman e r a / . (1976), Sperling and Klug (1977). (5)Grosser a / . (1978a,b). Walzthony et a/. (1981). (6) Summarized by Echlin (1978b). (7) Not yet done. (8) The only examples after freeze-fracturing are given by Nermut and Williams (1977) and Nermut et a / . (1978); negative staining after freezing was recently used for liposomes by Meister and Miiller (1980). (9) Historical papers are from Fernandez-Morin (1960), Miiller (1957). Sjostrand and Baker (1958). and Williams (1954); for later work see Nermut (1973, 1977) and Nermut and Frank (1971). for viruses; Rebhun (1972). Rowe (1960). and Sjostrand and Kretzer (1975). for other objects. For combinations with other techniques see below. (9 6) Boyde and Wood (1969) and many other papers. (9 + 23) This combination is frequently used and the contrast is often enhanced by exposing samples to Os04 vapors; Chang era/. (1980). Coulter and Terracio (1977). Edelmann (1978). Pfaller (1978). Pfaller and Rovan (1978). Terracio et al. (1981) are examples. (10) In principle this would be a possible and useful method. (Surface labeling-but without cryofixation-has already been used). ( I 1) The material used was partly hydrated and partly freeze-dried. See Chandler (1977). Echlin (1975, 1978a). Echlin and Galle (1975). Forrest and Marshall (1976). Fuchs (1979), Fuchs e t a / . (1978). Hall (1979). Hall er a/. (1974). Hutchinson (1979). Lechene and Warner (1977, 1979). Lechene et a / . (1979). Schafer and Zierold (1978). Steinbrecht and Zierold (1982). Zierold and Schafer (1978). 2s.-Nagy et a / . (1977). The problems with the use of fully hydrated specimens were analyzed especially by Fuchs (1979) and Fuchs et a/. (1978). (12) More general papers are by Chandler (1977), Echlin and Galle (1975), Forrest and Marshall (1976). Gupta and Hall (1981). Hall (1979). Hall er a / . (1974), Hutchinson (1979). and Lechene and Warner (1977) and the books edited by Hayat (1980) or Lechene and Warner (1979). Specifically for freeze-dried sections see below (13 .+ 9 + 12). (13) Pioneer work was done by Appleton (1973). Bernhard and Viron (1971). Christensen (1971), Tokuyasu (1973, 1976). See also Franzini-Armstrong et a/. (1978). Sirnard (1976). Spriggs 12 and I3 + 14. ( I 3 + 9 and Winne-Evans (l976), Werner ef a/. (1973). etc. as well as I3 9 + 12) Successful examples are Appleton and Newell (1977), Baker and Appleton (1977), Dorge et a / . (1978). Rick et a / . (1979). Roomans and SevCus (1976). Russ (1974). Somlyo and Silcox (1979), Somlyo e t a / . (1977, 1981), Wendt-Gallitelli et al. (1979). (13 12) Ross er a/. (1981). Saubermann e t a / . (1981a.b). Zierold et a/. (1981). (13 + 14) Geuze era/. (1979) and review by Tokuyasu (1980). (14) Geuze et a / . (1979), Painter er a / . (1973), Rash et al. (1980; method used involves 2*+ 23 + 14); Tokuyasu ( 1 980) is working with sucrose infused material and exposing cryosections to antibodies. (15) See critical comments by Rogers (1979). (16) Introduced by Bauer and Sigarlakie (1973. 1975) for delicate enzyme-cytochemical problems with poorly penetrable yeast cells. (17) Rash (1979). Rash era/. (1980); work on motor endplates. (18)Rash (1979), Rash e t a / . (1980); labeled a-bungarotoxin binding on motor endplates after freeze-fracturing and plastic embedding (ultrathin section-replica sandwich). (19) Dubochet et a/. (1981), Freeman et a/. (1980). Hayward and Glaeser (1980). Hayward and Stroud (1981). Heide (1981), Kellenberger (1978). Knapek and .--)

-+

-+

286

HELMUT PLA'ITNER AND LUIS BACHMANN

facilitated by reversible aggregation of MIP, e.g., by changes of pH (PintoDaSilva, 1972), by bivalent cations (Schober et al., 1977), or by forced physical contact (Schuler et al., 1978; Tanaka er al., 1980). Another approach along these lines is the following. Fibroblasts, incubated with glycerol, form cell membrane blisters which are devoid of MIP in freeze-fracture pictures (Hasty and Hay, 1978; Hay and Hasty, 1979). In a study with myoblasts it was found that such blisters lack enzyme activities characteristic of membrane-integrated proteins. (This again documents the need of advanced cryofixation techniques.) Recent reconstitution studies with lipid extracts demonstrated the occurrence of membrane-intercalated particles formed by lipids (DeKruijff et al., 1979; Verkleij and Ververgaert, 1978; Verkleij et al., 1979, 1980); these MIP were interpreted as inverted micelles (although an alternative explanation was presented by Miller, 1980). This led Robertson and Vergara (1980) to warn from a generalized interpretation of MIP as membrane-integrated proteins. In this respect it would appear desirable now to supplement fracturing and replication techniques with lipid extraction (for the feasibility of such attempts see Richter and Sleytr, 1971) or with digestion with various enzymes, including proteases, in order to establish the chemical nature of individual MIP (see Branton, 1971; Vilmart and Plattner, 1982). Another type of cytochemical analysis, which is also executed in conjunction with freeze-fracturing, is the selective complexation of certain membrane components. Some of these complexes can be recognized on replicas as distinct structures (bumps, holes, ripples). In order to reduce redistribution of the complexed membrane constituents, the probes are often added to an aldehyde fixative. For 3-OH-sterols one can use filipin (DeKruijff and Demel, 1974; Elias er al., 1979; Montesano el al., 1979, 1980; Robinson and Karnovski, 1980; Tillack and Kinsky, 1973; Verkleij et al., 1973), digitonin (Elias et al., 1978, 1979), and eventually saponin (see Zingsheim and Plattner, 1976). Polymyxin B reacts visibly with anionic phospholipids (Bearer and Friend, 1980). C. COMBINATIONS WTIH SURFACE ANALYSES Special surface particles can be recognized by deep etching in dilute buffers; the visualization of F,-particles on chromaffin granules is a very recent example of this possibility (Schmidt et al., 1982), which we did in parallel with negative Dubochet (1980), Taylor and Glaeser (1976). (20) Not yet done. (21-23) Bullivant (1965). Dempsey and Bullivant (1976), Feder and Sidman (1958). Franzini-Armstrong et a/. (1978). Handley e t a / . (1981). Heuser et a/. (1975, 1979), Howard and Aist (1979, 1980), Ichikawa et a/. (1980), Mehard (1976), Monroe er al. (1960, Miiller et a / . (1980), Pease (1966, 1967, 1973), Rebhun (1972). VanHarreveld and Crowell (1964). VanHarreveld e t a / . (1965, 1974). Zalokar (1966); Wolosewick and Porter (1979: in combination with high voltage electron microscopy). Woolley (1974; to fix stages of sperm motility). See also combination with (2)! (24) Ornberg and Reese (1980).

CRYOFIXATION

287

staining. It should also be appreciated that glycocalyx and bacterial capsule structures (Fig. 22) are well recognizable after rapid freezing. There is a wide range of marker molecules available which are large enough to be detected on surface replicas or on the surface of bulk specimens, even at the still limited resolution of the secondary electron picture obtained by scanning electron microscopy. The markers available were recently reviewed by Koehler (1978) and Molday and Maher (1980). Large markers are, for instance, virus particles (Hammerling et al., 1969) and hemocyanin (Brown and Revel, 1978; Revel, 1974; Weller, 1974). Among the smaller ones, ferritin is available in a native, polyanionic, or polycationized form and can, therefore, be directly used to probe for surface charge distributions (see reviews by Koehler, 1978; Molday and Maher, 1980). Most markers used in cryobiological analyses are first coupled to specific

FIG.22. Freeze-etched bacteria (Enrerohucrer uerogenes). Structural details of the capsule polysaccharides are most clearly visible after spray-freezing (a), but considerably obscured after standard freezing (b). From Plattner er a / . (1973). Bar = I pm.

288

HELMUT PLATTNER AND LUIS BACHMANN

probes, such as antibody molecules or Fab fragments thereof (Abbas et al., 976; Berzborn et al., 1974; Miller and Staehelin, 1976; Howe and Bachi, 973; Painter e f al., 1973; Perkins and Koehler, 1978; PintoDaSilva et al., 971; Shotton et al., 1978; VanEwijk and DeVries, 1977), lectins (Bachi and Sch ebli, 1975; PintoDaSilva and Nicolson, 1974; Roth et al., 1975; Triche et al., 1973, or specific inhibitors (Hourani et af., 1974). See also the work of Rash (1979; Rash er al., 1980) and Fig. 21 as well as Table IV for further methodical combinations. There is a report that enzyme-labeled ligands, as generally used for ultrathin sectioning, might be used for surface labeling, when their size is increased by the deposition of a reaction product before freezing and replication (Carter and Staehelin, 1979). Although even unlabeled antibodies could be observed on smooth surfaces after fast freezing and deep-etching, as demonstrated with bacterial flagella by Munn et al. (1980), most people work with additional label molecules. When labeling experiments of this type are to be analyzed by freeze-etching, no antifreeze can be used because it would largely impede the exposure of membrane surfaces by deep etching. It is, therefore, common practice now to subject such samples to aldehyde fixation in order to make them resistent enough for a brief wash with distilled water. As this does normally impede a decent cryofixation, advanced cryofixation techniques could represent a valuable alternative also in this case. D. COMBINATIONS WITH OTHERCYTOCHEMICAL PROCEDURES Such procedures are included in Table IV and Fig. 2 1. Examples are autoradiography, element analysis, and cytochemistry (including immunocytochemistry and decoration of functional groups in ultrahigh vacuum). For autoradiography and element analysis rapid cryofixation without chemical pretreatment is a prerequisite in order to avoid removal or dislocation of low molecular components or ions. Although autoradiography was used successfully for this purpose for a long time at the light microscopic level (Kinter and Wilson, 1965; Stirling and Kinter, 1967; Stumpf and Roth, 1966), similar attempts at the electron microscopic level still deserve some scepticism (see also critical evaluation by Rogers, 1979); only very recently has some progress been achieved with a dry emulsion mounting technique (Harris and Salpeter, 198 1). Autoradiography has already been applied to frozen bulk specimens (Schiller et al., 1978a,b; Rix et al., 1977) but it definitely needs more methodical developments in order to reach an acceptable standard. The situation is much more favorable (see Fisher and Branton, 1976) when the chemical composition of membrane remnants from cell monolayers is analyzed, e.g., for selective enrichment of [3H]cholesterol (Fisher, 1976, 1978) or of iodinated membrane proteins (Nermut and Williams, 1980) within one-half of a membrane.

CRYOFIXATION

289

The quality of cryofixation is at present even more important for elemental analysis (X-ray microanalysis) in electron microscopes of different types, with the use of frozen-hydrated or freeze-dried specimens. Efforts along these lines proved very useful in the past few years (see Table IV and Fig. 21). As pointed out in Section V1 cryofixation still can be a resolution-limiting parameter because of the size of the segregation zones. It is obvious that even the slightest chemical prefixation disturbs the intravital ionic distribution (see Holbrook et al., 1976). With this aim, considerable efforts have been made in the last few years to overcome the notorious problem of cryofixation of bulk specimens. The possible potential of high-pressure freezing is discussed in Section 1V,B,7. Another attempt focused on the infiltration of intercellular spaces by nonpenetrating, highpolymeric cryoprotectants (Barnard, 1980; Echlin er al., 1977; Schiller er a / ., 1978b,c; Skaer er al., 1977, 1978, 1979; Wilson and Robards, 1980). However, the occurrence of some side-effects with several cell types (Barnard, 1980; Echlin et al., 1977; Skaer et al., 1978) and the possible redistribution of water across the cell membrane (Allen and Weatherbee, 1979; McGann, 1978; Meryman, 1971; Wilson and Robards, 1982) caution one to expect a general solution from polymeric antifreezes. They might be useful for plant tissues (for which they were indeed primarily used), as one could expect from the presence of a cell wall as a natural, efficient diffusion barrier. Very recently, elements of low atomic weight have been analyzed by electron energy loss spectroscopy (Hutchinson, 1979) but only one application has been published, to our knowledge, with frozen sections (Somlyo and Shuman, 1982). In cytochemistry, including immunocytochemistry, frozen sections are frequently used in order to increase the access of substrates or of relatively large labeled ligands to intracellular structures. Figure 2 1 includes work on ultracryotomy and immunocytochemistry. The pioneer work of Tokuyasu (1973, 1976, 1980) should be especially underscored; he uses sucrose perfusion for cryoprotection and improvement of the cutting. This method is now most appreciated for the localization of intracellular antigens. An earlier alternative was the impregnation with serum albumin followed by aldehyde crosslinking and subsequent freezing (Kuhlmann and Viron, 1972). When Griffiths and Jockusch (1 980) compared these alternative methods, they found that either one has specific disadvantages. In the case of immunocytochemistry freezing serves only to stiffen the material and the application of ‘‘advanced” cryofixation techniques is not necessary. WITH ULTRATHIN SECTIONING E. COMBINATIONS

Table IV and Fig. 21 indicate several possibilities to transfer frozen materials to the ultrathin sectioning techniques. Freeze-substitution is the most popular

290

HELMUT PLATTNER AND LUIS BACHMANN

Fic. 23. Rat liver cells in monolayer culture on Thermanox. sandwich frozen by the one-sided propane-jet method of Fig. 13. That portion of the 15-pm-thick sample which was attached to the Thermanox was subjected to freeze-fracturing (without metal and carbon coating) and subsequently to freeze-substitution fixation (acetone + 1% 0 ~ 0 ~ starting ) . at -85°C. temperature rise 2"Clhour. embedding in Spurr's epoxide resin at room temperature and ultrathin sectioning plus section staini n g (Fig. 17 shows an aliquot after freeze-fracture replication.) C , and C2 are two adjacent cells. GJ, Gap junction; S, substratum (Thermanox). From Pscheid et al. (1981). Bar = I pm.

method (see Fig. 23 for an example). It helps, for instance, to catch rapid events and simultaneously to analyze details which would not be so well recognizable on replicas (see e.g., Heuser er al., 1975, 1979). The water is replaced at low temperature (around -80°C) by a dehydrating agent (acetone, ethanol, methanol), to which various fixatives (osmium tetroxide, glutaraldehyde, acrolein, uranyl acetate) can be added (see mainly Miiller er al., 1980a; VanHarreveld and Crowell, 1964; VanHarreveld et al., 1974; Zalokar, 1966). If properly done, recrystallization seems to be small during freeze-substitution. With some plastic media embedding can be well performed at subzero temperatures (Carlemalm et al., 1982; Kellenberger et al.. 1980). See also the introduction to Section 1X.

X. Suspensions, Emulsions, and Solutions In recent years fast freezing followed by freeze-etching and high-resolution shadowing has been applied to emulsions, suspensions, and solutions of proteins and to micelle-forming surfactants. In these cases, the concentration of the solute is usually much less than one per mille. Cryofixation, therefore, implies the freezing of almost pure water in a way that the suspended or dissolved substances should remain in their statistical distribution. They should not be concentrated in

CRYOFIXATION

29 I

the intercrystalline regions or the size of the ice crystals should not exceed the 10 nm range. Spray-freezing and specimen jet freezing have been used. For solutions, gels etc., which are sensitive to shearing stress, the sandwich-cryogen jet technique is the method of choice (see Section IV,B,5). A. SIZEAND SHAPEOF MACROMOLECULES

Provided that the resolution of the replicating method is high enough, size and shape of hydrated macromolecules can be studied and compared with hydrodynamic data. Tantaludtungsten shadowing and/or rotary shadowing after deep etching have been applied (Fig. 24). It was found that insufficient cryofixation first manifests itself in a nonstatistical distribution of the particles while the dehydration of a protein molecule due to freezing is a relatively slow process. This is indicated by the fact that the size of the highly hydrated fibrinogen molecule is identical in areas of perfect cryofixation and in regions showing segregation (Bachmann et a / . , 1974b. 1975). With the use of his sandwich-freezing technique (see Section IV) Gulik has subsequently also analyzed pure protein (LeMaire et a / ., 198 1) or lipoprotein fractions (Gulik-Krzywicki et a / ., 1979) for their size and shape. Frozen hydrated molecules are now increasingly analyzed by low-temperature, low-dose electron microscopic imaging procedures (for review see Dubochet et al., 198 I ) , including electron diffraction and image reconstruction (see e.g., Hayward and Stroud, 1981). This appears interesting in view of possible narrow-range rearrangements during drying (Lepault and Dubochet, 1980).

FIG.24. Spray-frozen and freeze-etched proteins in solution, shadowed with tantaludtungsten. (a-c) Pyruvate-dehydrogenase complex from pig heart (Junger and Bachmann, 1977); (a, b) natively frozen molecules exposed by deep etching; (c) negatively stained control (after glutaraldehyde stabilization). (d) Fibrinogen molecules after deep etching. From Bachmann ef a / . (1975) and Junger and Bachmann (1977). Bar = 0.1 pm.

292

HELMUT PLATTNER AND LUIS BACHMANN

B. CRYOFIXATION OF HEATEDSOLUTIONS

For the study of some micelle-forming surfactants and of fibrinogen it was necessary to keep the solutions at an elevated constant temperature (up to +80 or 40°C, respectively) prior to freezing (Kutter er al., 1976, Bachmann el al., 1981). Spraying of a warm solution would lead to the evaporation of water and the cooling of the droplets below room temperature before hitting the cryogen. Particle weight determination would, therefore, be impossible and artifacts during cooling could occur. Specimen-jetting (Fig. 10) proved to be suitable: The concentration of the solution is not changed and the jet stream keeps its temperature until it hits the cryogen. VanVenetie et al. (198 1) have developed a thermostat unit for cryogen jetfreezing which allows catching of thermosensitive states (“phases”) of some lipids, which cannot be observed after conventional freezing. c .

DETERMINATION OF PARTICLE

WEIGHT BY COUNTING

A statistical distribution of rapidly frozen particles opens the way to determine their weight by the determination of the number of particles per unit volume of the sublimated solvent (number average). This technique and its underlying rationale have been described in detail by Bachmann and Schmitt-Fumian (1973a), Bachmann et al. (1974a, 1975), and Junger and Bachmann (1977). Identical etching conditions for the sample and a standard are required for this method. This is best obtained by spray-freezing or specimen jet-freezing, where droplets of sample (see Fig. 9) and droplets of standard are processed side by side in the same freeze-etch specimen.

XI. Nonaqueous Systems If solutions of organic substances (e.g., nonane, benzoic acid, polystryrene) in benzene are frozen by the standard technique (Section IV ,A), freeze-etching reveals that phase-segregation effects occur which are similar to those in aqueous systems: Crystals of the solvent, several micrometers in size, are formed, while the solute is concentrated mainly in the intercrystalline space (Kolbusch el al., 1976). Standard-frozen solutions of macromolecules show areas with no particles and others, where the molecules form clusters. Modified spray and specimen jet-freezing improves the quality of cryofixation as in aqueous solutions. Since difficulties occurred in finding a cryomedium and a glue suitable for the organic solvent used, sandwich or sandwich-cryogen jet-freezing is recommended (Section IV). To our knowledge only the polystyrene in benzene system has been investigat-

CRYOFIXATION

293

ed on a quantitative scale so far. We believe, however, that this method would lend itself equally to the investigation of hydrophobic biomolecules.

XII. Concluding Remarks In the past decade some success has been achieved in the cryofixation of biological materials for use in ultrastructure research. The general change in attitude toward the need of advanced cryofixation methods, which allow the processing of biomaterials in the native state (i.e., without any chemical pretreatments), must also be viewed as a success. Previously, attempts along these lines were not always adequately appreciated. Four factors have mainly caused some rethinking: First, evidence has accumulated that any chemical pretreatment introduces the hazard of ultrastructural artifacts. Second, the new techniques for element analysis (from advanced cryosectioning to X-ray microprobe analysis) cannot be fully exploited with chemically altered materials. Third, rapid freezing offers a new possibility to catch fast dynamic processes. Fourth, it allows the investigation of dissolved hydrated proteins and molecular assemblies by electron microscopy. With this in mind, we believe that cryofixation will be of increasing importance for cellular and molecular biology.

ACKNOWLEDGMENTS We thank Mrs. M. L. Hrstka for correcting the manuscript, Dr. P. Pscheid for making some of the drawings, and Mrs. U . Remensperger for her secretarial help. We also gratefully acknowledge the financial support of the “Deutsche Forschungsgemeinschaft” and of the “Sonderforschungsbereich 138” for grants which allowed us to obtain some of the data presented here. All figures were reproduced with the kind permission of authors and publishing companies (Fig. 9, National Academy of Science, Washington; Fig. 13, Blackwell Scientific Publ. Ltd., Oxford; Fig. 14, Springer-Verlag, Wien; Fig. 18, Alan R. Liss Inc., New York; Fig. 22, SOC.F r a y . Microscopie Electronique, Paris; Fig. 24, Huethig & Wapf, Basel).

REFERENCES Abbas, A. K., Dorf, M. E., Kamovsky, M. J . , and Unanue, E. R. (1976). J. Imrnunol. 116, 371-378. Allen, E. D., and Weatherbee, L. (1979). J. Microsc. 117, 381-394. Allen, E. D., and Weatherbee, L. (1980). Cvobiology 17, 448-457. Angell. C. A,, and Sare. E. J. (1970). Science 168, 280-281. Appleton. T. C. (1974). In “Electron Microscopy and Cytochemistry” (E. Wisse, W. T. Daems, I. Molenaar, and P. Van Duijn. eds.), pp. 229-241. North-Holland Publ., Amsterdam. Appleton, T. C., and Newell, P. F. (1977). Nature (London)266, 854-855. and Trovalusci, P. (1978). J. Submicrosc. Cvrol. 10, 106-107. Arancia, 0..

294

HELMUT PLA’ITNER AND LUIS BACHMANN

Arancia, G., Rosati-Valente, F., and Trovalusci-Crateri, P. (1980). J . Microsc. 118, 161-176. Bachi, T., and Schnebli, H. P. (1975). Exp. Cell Res. 91, 285-295. Bachmann. L., and Schmitt. W. W. (1971). Proc. Natl. Acad. Sci. U.S.A. 68, 2149-2152. Bachmann, L., and Schmitt-Fumian, W. W. (1973a). In “Freeze-Etching. Techniques and Applications” (E. L. Benedetti and P. Favard, eds.), pp. 63-72. Socittt Francaise de Microscopie Electronique, Paris. Bachmann, L., and Schmitt-Fumian, W. W. (1973b). In “Freeze-Etching. Techniques and Applications” (E. L. Benedetti and P. Favard, eds.), pp. 73-79. Socittt Francaise de Micaxcopie Electronique, Paris. Bachmann, L., Schmitt, W. W., and Plattner. H.(1972). Proc. Eur. Reg. Congr. Electron Microsc. 5th, pp. 244-245. Bachmann, L., Fritzmann, H.,and Schmitt-Fumian, W. W. (1974a). Proc. Int. Congr. Electron Microsc.. 8th 11, 18-19. Bachmann, L., Junger, E., Lederer, K., and Schmitt-Fumian. W. W. (1974b). Proc. Int. Congr. Electron Microsc., 8th 11, 40-41. Bachmann, L., Schmitt-Fumian, W. W., Hammel, R., and Lederer, K. (1975). Makromol. Chem. 176, 2603-2618. Bachmann, L., Dasch, W., and Kutter, P. (1981). Ber. Bunsenges. Phys. Chem. 85, 883-887. Baker, J . R . J., and Appleton, T. C. (1977). J . Microsc. 108, 307-315. Baker, R. F. (1962). J . Ultrastruct. Res. 7, 173-184. Bank, H. (1974). Exp. Cell Res. 85, 367-376. Bank, H., and Mazur, P. (1972). Exp. Cell Res. 71, 441-454. Bank, H . , and Mazur, P. (1973). J. Cell Biol. 57, 729-742. Barlow, D. I., and Sleigh, M. A. (1979). J . Microsc. 115, 81-95. Barnakov, A. N. (1982). Tsitofogiya 24, 622-624. Barnard, T. (1980). J . Microsc. 120, 93-104. Bauer, H., and Sigarlakie, E. (1973). J . Microsc. 99, 205-218. Bauer, H., and Sigarlakie, E. (1975). J . Ultrastrucr. Res. 50, 208-215. Baur, H . , Kasperek, S., and Pfaff, E. (1975). Hoppe-Seyler’s Z. Physiol. Chem. 356, 827-838. Bearer, E. L., and Friend, D. S. (1980). Proc. Natl. Acad. Sci. U.S.A. 77, 6601-6605. Benedetti, E. L., and Favard, P.. eds. (1973). “Freeze-Etching. Techniques and Applications.” Socitte Francaise de Microscopie Electronique, Paris. Bernhard, W., and Viron, A. (1971). J . CellBiol. 49, 731-746. Berzborn, R. J., Kopp, F., and Muhlethaler, K. (1974). Z. Naturforsch. 29C, 694-699. Boyde, A,, and Wood, C. (1969). J . Microsc. 90, 221-249. Boyne, A. F. (1979). J . Neurosci. Methods 1, 353-364. Branton, D. (1969). Annu. Rev. Plant Physiol. 20, 209-238. Branton, D. (1971). Philos. Trans. R . SOC. Ser. B 261, 133-138. Branton, D., and Deamer, D. W. (1972). “Membrane Structure.” Springer Verlag, Berlin and New York. Branton, D., and Kirchanski, S. (1977). J. Microsc. 111, 117-124. Bretscher, M. S . , and Whytock, S . (1977). J . Ultrastruct. Res. 61, 215-217. Brooks, S. E. H.,and Haggis, G. H. (1973). Lab. Invest. 29, 60-64. Brown, S. S . , and Revel, J. P. (1978). In “Advanced Techniques in Biological Electron Microscopy. 11. Specific Ultrastructural Probes” (J. K. Koehler, ed.), pp. 65-88. Springer-Verlag, Berlin and New York. Briiggeller, P., and Mayer, E. (1980). Nature (London) 288, 569-571. Buchheim, W. (1972a). Proc. Eur. Reg. Congr. Electron Microsc. 5th. pp. 246-247. Buchheim, W. (1972b). Natunvissenschaften 59, 121. Buchheim, W., and Welsch, U. (1977). J . Microsc. 111, 339-349. Bullivant, S. (1965). Lab. Invest. 14, 1178-1 195.

CRYOFIXATION

295

Bullivant, S. (1970). I n “Some Biological Techniques in Electron Microscopy” (D. F. Parsons, ed.). pp. 101-146. Academic Press, New York. Bullivant. S. (1973). I n “Advanced Techniques in Biological Electron Microscopy” (J. K . Koehler. ed.), pp, 67-1 12. Springer-Verlag. Berlin and New York. Burstein, N . L., and Maurice, D. M. (1978). Micron 9, 191-198. Carlemalm, E.. Gravito, R. M., and Villiger, W. (1982). J. Microsc. 126, 123-143. Carter. D. P., and Staehelin, L. A . (1979). J. Microsc. 117, 363-373. Casley-Smith. J. R. (1980). Micron I I , 461-462. Chandler. D. E. (1979). I n “Freeze Fracture: Methods, Artifacts and Interpretations” (J. E. Rash and C. S. Hudson. eds.). pp. 81-88. Raven, New York. Chandler, D. E., and Heuser. J . (1979). J. Cell Biol. 83, 91-108. Chandler. D. E., and Heuser, J. E. (1980). J . Cell Biol. 86, 666-674. Chandler, J . A. (1977). “X-ray Microanalysis in the Electron Microscope.” North-Holland Publ., Amsterdam. Chang, S. H.. Mergner. W. J., Pendergrass, R. E.. Bulger, R. E.. Berezesky, I. K., and Trump, B. F. ( 1980). J. Hisfoehem. Cytochem. 28, 47-5 I . Christensen. A. K. (1971). J. Cell Biol. 51, 772-804. Costello. M. J., and Corless. J. M. (1978). J. Microsc. 112, 17-37. Costello. M. J.. and Gulik-Krzywicki, T. (1976). Biochim. Biophys. Acta 455, 412-432. Cottrell, S. F. (1981). Cpobiology 18, 506-510. Coulter, H. D.. and Terracio, L. (1977). Anar. Rec. 187, 477-494. Crowe, J. H.. and Cooper, A. F. (1971). Sci. Am. 225, 30-36. Cummins. J . M., and Loper, M. N . (1975). J. Micrusc. 104, 121-125. DeGroot. C., and Leene, W. (1979). Eur. J. Cell B i d . 19, 19-25. Deguchi. N.. Jorgensen. P. L., and Maunsbach, A. B. (1977). J. Cell Biol. 75, 619-634. DeKruijff, B.. and Demel. R. A. (1974). Biochim. Biophvs. Acta 339, 57-70. DeKruijff. B.. Verkleij, A . J., VanEchteld, C. J. A., Gerritsen, W. J., Mombers, C., Noordam, P. C.. and DeGier. J. (1979). Biochim. Biophvs. Acra 555, 200-209. Dempsey, G. P., and Bullivant, S . (1976a). J. Microsc. 106, 251-260. Dempsey. G. P., and Bullivant. S. (1976b). J. Microsc. 106, 261-271. Dietrich, I.. Fox. F., Knapek, E., Lefranc, G., Nachtrieb, K.. Weyl, R.. and Zerbst, H. (1977). ultra micros cop^ 2, 24 1-249. Dorge, A , , Rick. R.. Gehring, K.. and Thurau, K. (1978). P’iiger’s Arch. Phvsiol. 373, 85-97. Dubochet, J., and McDowall, A. W. (1981). J. Microsc. 124, RP3-RP4. Dubochet. J., Booy, F. P.. Freeman. R., Jones, A. V.. and Walter. C. A. (1981). Annu. Rev. Biuphy. Bioeng. 10, 133- 149. Echlin. P. (1975). J. Microsc. B i d . Cell. 22, 215-226. Echlin. P. (1977). Nature (London) 267, 312-313. Echlin, P. ( 1978a). “Low Temperature Biological Microscopy and Microanalysis.” Royal Microscopical Society, Oxford. Echlin, P. ( 1978b). In “Advanced Techniques in Biological Electron Microscopy. 11. Specific Ultrastructural Probes“ (1. K. Koehler, ed.), pp. 89- 122. Springer-Verlag. Berlin and New York. Echlin, P., and Galle. P. ( 1975). “Biological Microanalysis.” SociCtC Francaise de Microscopie Electronique, Paris. Echlin, P., Skaer, H. L. B.. Gardiner, B. Q. C.. Franks, F.. and Asquith, M. H. ( 1977).J . Microsc. 110, 239-255. Edelmann. L. (1978). J. Microsc. 112, 243-248. Edwards, H. H., Mueller. T. J., and Morrison, M. (1979). Science 203, 1343-1346. Eger, R., and Rifkin, D. (1980). Biochim Biophvs. Acta 600, 313-319. Elias, P. M.. Goerke, J.. and Friend, D. S. (1978). J. Cell Bid. 78, 577-596. Elias, P. M., Friend, D. S . . and Goerke. J. (1979). J. Histoehem. Cvtochem. 27, 1247-1260.

296

HELMUT PLATTNER AND LUIS BACHMANN

Eranko, 0. (1954). Acru Anar. 22, 331-336. Escaig, J. (1982). J . Microsc. 126, 221-229. Escaig, J., Gtraud, G., and Nicolas, G. ( 1977).C. R. Acud. Sci. Ser. D . , Sci. Nur. 284,2289--2292. Farrant, J. (1977). Philos. Trans. R. SOC. London, Ser. B. 278, 191-205. Farrant, J., and Woolgar, A. E. (1970). In “The Frozen Cell” (G. E. W. Wolstenholme 2nd M. O’Connor, eds.), pp. 97-1 14. Churchill, London. Farrant, J., Walter, C. A., Lee, H., and McGann, L. E. (1977a). Cryobiology 14, 273-286. Farrant, J., Walter, C. A., Lee, H., Moms, G. J., and Clarke, K. J. (1977b). J. Microsc. 111, 17-34. Feder, N., and Sidman, R. L. (1958). J . Biophys. Biochem. Cyrol. 4, 593-602. Fernihdez-Morin, H. (1960). Ann. N.Y. Acud. Sci. 85, 689-713. Finch, J. T., and Klug, A. (1976). Proc. Nurl. Acud. Sci. U.S.A. 73, 1897-1901. Fineran, B. A. (1970). J . Microsc. 92, 85-97. Fisher, K. A. (1975). Science 190, 983-985. Fisher, K. A. (1976). Proc. Nutl. Acud. Sci. U.S.A. 73, 173-177. Fisher, K. A. (1978). Proc. Inr. Congr. Electron Microsc. 9th Ill, 521-532. Fisher, K. A., and Branton, D. (1973).J. Cell Biol. 59, 99a. Fisher, K. A., and Branton, D. (1976). J. Cell Biol. 70, 453-458. Fooke-Achterrath. M., Lickfeld, K. G., Reusch, V. M., Aebi. U.,Tschope, U., and Menge, B. (1974).J. Ulrrustrucr. Res. 49, 270-285. Forrest, Q. G., and Marshall, A. T. (1976).Proc. Evr. Congr. Electron Microsc. 6th 11, 218-220. Franks, F. (1977).J. Microsc. 111, 3-16. Franks, F., Asquith, M. H., Hammond, C. C., Skaer, H. L. B., and Echlin, P. (1977).J . Microsc. 110, 223-238. Franzini-Armstrong, C., Heuser, J. E., Reese, T. S., Somlyo, A. P., and Somlyo, A. V. (1978).J. Physiol. (London) 283, 133- 140. Freeman, R.. Leonard, K. R., and Dubochet, J. (1980).Proc. Eur. Congr. ElectronMicrosc. 7th 11, 640-641.

Fuchs, W. (1979). In “Microbeam Analysis in Biology” (C. P. Lechene and R. R. Warner, eds.), pp. 361-373. Academic Press, New York. Fuchs, W., Lindemann, B., Brombach, J. D., and Trosch, W. (1978). J . Microsc. 112, 3-87. Furcht, L. T., and Scott, R. E. (1974). Exp. Cell Res. 88, 31 1-318. Furcht, L. T., and Scott, R. E. (1975). Biochim. Biophys. Actu 401, 213-220. Gebhardt, C., Gruler. H., and Sackmann, E. (1977). 2. Nuturforsch. 32C, 581-596. Geuze, I . I . , Slot, 1. W., and Tokuyasu, K. T. (1979). J . Cell Biol. 82, 697-707. Ghosh, B. K., and Nanninga, N. (1976). J . Ultrusrrucr. Res. 56, 107-120. Giddings, T. H., and Staehelin, L. A. (1980). J . CeNBiol. 85, 147-152. Giddings, T.H., Brower, D. L., and Staehelin, L. A. (1980). J. Cell Biol. 84, 327-339. Glover, A. J., and Garvitch, Z. S. (1974). Cryobiology 11, 248-254. Grant,C. W.,Hong-Wei, S.,andMcConnell,H. M. (1974).Biochim.Biophys.Acru363, 151-158. Griffiths, G. W., and Jockusch, B. M. (1980). J . Hisrochem. Cytochem. 28, 969-978. Gross, H., Bas, E., and Moor, H. (1978a). J. Cell Biol. 76, 712-728. Gross, H., Kuebler, 0.. Bas, E., and Moor, H. (1978b).J. Cell Biol. 79, 646-656. Gulik-Krzywicki, T. (1975). Biochim. Eiophys. Acra 415, 1-28. Gulik-Krzywicki, T., and Costello, M.J. (1977). Proc. Annu. Meer.. 35th. Electron Microsc. SOC. Am. pp. 330-333. Gulik-Krzywicki, T., and Costello, M. J. (1978).J. Microsc. 112, 103-1 14. Gulik-Krzywicki, T., Yates, M., and Aggerbeck, L. P. (1979). J. Mol. Biol. 131, 475-484. Gupta, B. L., and Hall, T. A. (1981). Tissue Cell 13, 623-643. Haase, W., and Koepsell, H. (1979). Pfluger’s Arch. Physiol. 381, 127-135.

CRYOFIXATION

297

Haggis. G. H. (1978). Proc. Itit. Congr. Electron Microsc. 9th 11, 50-51. Haggis. G. H., and Phipps-Todd. B. (1977). J. Microsc. 111, 193-201. Hall. C. E. (1950). J. Appl. Phps. 21, 61-62. Hall, T. A. (1979). J. Microsc. 117, 145-163. Hall, T. A., Echlin, P., and Kaufmann, R.. eds. (1974). “Microprobe Analysis as Applied to Cells and Tissues.” Academic Press, New York. Hammerling. U.. Aoki. T.. Wood, H. A.. Old, L. J.. Boyse, E. A., and DeHarven. E. (1969). Nature (Londotr)223, 1158- 1159. Handley. D. A.. Alexander, J. T.. and Chien. S. (1981). J. Micruse. 121, 273-282. Harris. W. V.. and Salpeter. M. M. (1981). J. Cell B i d . 91, 416a. Hasty. D. L., and Hay. E. D. (1978). J. Cell B i d . 78, 756-768. Hax. W . M. A., and Lichtenegger, S. (1982). J. Microsc. 126, 275-284. Hay, E. D., and Hasty. D. L. (1979). In “Freeze Fracture: Methods, Artifacts and Interpretations” (J. E. Rash and C. S. Hudson, eds.), pp. 59-66. Raven, New York. Hayat. M. A. (1980). ”X-Ray Microanalysis in Biology.” Univ. Park Press. Baltimore. Maryland. Hayward, S. B.. and Glaeser. R. M. (1980). Ultramicroscupv 5, 3-8. Hayward, S. B., and Stroud. R . M. (1981). J. Mol. B i ~ l 151, . 491-517. Heide, H. G . (1981). Ultrumicroscopp 6, 115-124. Heide. H. G.. and Grund. S. (1974). J. U/trastri~ct.Res. 48, 259-268. Hereward. F. V., and Northcote. D. H. (1972). J. Cell Sci. 10, 555-561. Hereward. F. V . . and Northcote, D. H. (1973). J. Cell Sci. 13, 621-635. Heuser, J . (1980). J. Cell Biol. 84, 560-583. Heuser, J. E. (1977). Proc. Annu. Meet.. 35th. Electron Microsc. Sue. Am. pp. 676-679. Heuser, J. E., and Kirschner, M. W. (1980). J. Cell B i d . 86, 212-234. Heuser, J. E., and Salpeter. S . R. (1979). J. Cell Biol. 82, 150-173. Heuser, J. E.. Reese, T. S., and Landis, D. M. (1974). J. Neurocvtol. 3, 109-131. Heuser, J. E.. Reese. T. S.. and Landis, D. M. D. (1975). Cold Spring Harbor Svmp. Quunt. Biol. 40, 17-24. Heuser. J. E.. Reese. T. S.. Dennis, M. J., Jan, Y.. Jan, L.. and Evans, L. (1979). J. CellBiol. 81, 275-300. Higgins, M. L.. and Daneo-Moore, L. (1974). J. Cell Biol. 61, 288-300. Higgins. M. L., Tsien. H. C., and Daneo-Moore, L. (1976). J. Bucieriol. 127, 1519-1523. Holbrook, K. A.. Holbrook. J. R.. and Odland, G. F. (1976). Proc. Annu. 17T Res. Inst. SEM Symp.. 9th pp. 266-274. Hourani, B. T., Torain, B. F., Henkart, M. P., Carter, R . L.. Marchesi, V. T.. and Fischbach. G. D. (1974). J. Cell Sci. 16, 473-479. Howard, R. J., and Aist, J. R. (1979). J. Ultrastruct. Res. 66, 224-234. Howard, R. J., and Aist. J. R. (1980). J. Cell B i d . 87, 55-64. Howe, C.. and Bachi. T. (1973). Exp. Cell Res. 76, 321-332. Hudson, C. S . , Rash. J. E., and Graham, W. F. (1979).In “Freeze Fracture: Methods, Artifacts and Interpretations” (J. E. Rash and C. S. Hudson. eds.), pp. 1-10, Raven. New York. Humbert, F., Abramow, M., Perrelet. A , . and Orci, L. (1977). Kidnev Int. 12, 66-71. Hutchinson, T. E. (1979). Int. Rev. Cvtol. 58, 115-158. Hutchinson, T. E.. and Borek, J. R. (1979). Ultrumicroscopv 4, 233-239. Ichikawa, A., Ichikawa, M., and Hirokawa, N. (1980). Am. J . Anat. 157, 107-110. Jehl. B.. Bauer. R., Diirge, A,. and Rick, R. (1981). J. Microsc. 123, 307-309. Junger, E., and Bachmann, L. (1977). Biochim. Biophvs. Aria 481, 364-376. Kachar, B., Serrano. J . A.. and PintoDaSilva. P. (1980). Cell Biol. In!. Rep. 4, 347-356. Kanno, H., Speedy, R. J., and Angell, C. A. (1975). Science 189, 880-881. Karp, R. D.. Silcox. J. C., and Somlyo, A. V. (1982). J. Microsc. 125, 157-165.

298

HELMUT PLATTNER AND LUIS BACHMANN

Kellenberger, E. (1978). Trends Biochem. Sci. 3, 135-137. Kellenberger, E., Carlemalm, E., Villiger, W., Roth, J., and Garavito, R. M. (1980). “Low Denaturation Embedding for Electron Microscopy of Thin Sections.” Chem. Werke Lowi, Waldkraitburg, F.R.G. Kinter, W. B., and Wilson, T. H. (1965). J. Cell Biol. 25, 19-39. Kleemann, W., and McConnell, H. M. (1974). Biochim. Biophys. Actu 345, 220-230. Knapek, E., and Dubochet, J. (1980). J. Mol. Biol. 141, 147-161. Knoll, G., Oebel, G., and Plattner, H. (1982a).Proc. Int. Congr. Electron Microsc. loth (in press). Knoll, G., Oebel, G.,and Plattner, H. (1982b). Protoplusma 111, 161-176. Koehler, J . K . (1978). “Advanced Techniques in Biological Electron Microscopy. 11. Specific Ultrastructural Probes.” Springer-Verlag, Berlin and New York. Iohlbusch, P., Schmitt-Fumian, W. W., and Buchmann, L. (1976). Proc. Eur. Congr. Electron Microsc. 6th 11, 116-1 18. Kuhlmann. W. D., and Viron, A. (1972). J. Ultrasrruct. Res. 41, 385-394. Kutter, P., Schmitt-Fumian, W. W., and Bachmann, L. (1976). Proc. Eur. Congr. Electron Mirrosc., 6th 11, 119-121. Lane, L. B. (1925). Ind. Eng. Chem. 17, 924. Lang, R. D. A,, and Bronk, J. R. (1978). J. Cell B i d . 77, 134-147. Lechene, C. P., and Warner, R. R. (1977). Annu. Rev. Biophys. Bioeng. 6 , 57-85. Lechene, C. P., and Warner, R. R., eds. (1979). ”Microbeam Analysis in Biology.” Academic Press, New York. Lechene, C. P., Bonventre, J. V., and Warner, R. R. (1979).In “Microbeam Analysis in Biology” (C. P. Lechene and R. R. Warner, eds.), pp. 409-426. Academic Press, New York. Lefort-Tran, M., Gulik, T., Plattner, H., Beisson, J.. and Wiessner, W. (1978). Proc. Int. Congr. Electron Microsc. 9th 11, 146-147. LeMaire, M., Mdler, J. V., and Gulik-Krzywicki, T. (1981). Biochim. Biophvs. Acta 643, 115-125. Lenard, J., and Singer, S. J. (1968). J . Cell Biol. 37, 117-121. Lepault, J., and Dubochet, J. (1980). J. Ultrasrruct. Res. 72, 223-233. Lennan, L. S., Wilkerson, L. S., Venable, J. H., and Robinson, B. H. (1976).J. Mol. Biol. 108, 27 1-293. Lichtenegger, S., and Hax, W. M. A. (1980). Proc. Eur. Congr. Electron Microsc.. 7th 11, 652-653. Lickfeld, K. G.,and Achterrath, M. (1972). Cvtobiologie 6, 74-85. Lickfeld, K. G.,Almert, U., and Menge, B. (1976). Microsc. Actu 77, 441-444. Luft, J. H. (1976). In:. Rev. Cytol. 45, 291-382. Lusena, C. V. (1960). Ann. N.Y. Acud. Sci. 85, 541-548. Luyet, B. J . (1970). In “The Frozen Cell” ( G . E. W. Wolstenholme and M. O’Connor. eds.), pp. 27-43. Churchill, London. Luyet. B., Tanner, I., and Rapatz, G. ( 1962). Biodynumica 9, 2 1. MacKenzie, A. P. (1969). Biodynumica 10, 341-351. MacKenzie, A. P. (1977). Phil. Trans. R. Soc. London Ser. B 278, 167-189. McGann, L. E. (1978). Cryobiology 15, 382-390. Mclntyre, I . A,, Karnovsky, M. J., and Gilula, N. B. (1973). Nature (London) New Biol. 245, 147- 148. Mclntyre, J. A,, Gilula, N. B., and Karnovsky, M. J. (1974). J. Cell Biol. 60, 192-203. McMillan. J . A,, and Los, S. C. (1965). Nature (London) 206, 806-807. Margaritis, L. H., Elgsaeter, A., and Branton, D. (1977). J. Cell Biol. 72, 47-56. Mayer. E., and Briiggeller, P. (1982). Nurure (London) (in press). Mazur, P. (1960). Ann. N.Y. Acud. Sci. 85, 610-629.

CRYOFIXATION

299

Mazur, P. (1977). Crybiology 14, 251-272. Mazur, P.. Leibo, S. P., Farrant, J., Chu, E. H. Y., Hanna, M. G., and Smith, L. H. (1970). In “The Frozen Cell” ( G . E. W. Wolstenholme and M. O’Connor, eds.), pp. 69-85. Churchill, London. Mazur, P., Leibo, S. P., and Chu, E. H. Y. (1972). Exp. Cell Res. 71, 345-355. Meister, N., and Miiller, M. (1980). Proc. Eur. Congr. Electron Microsc.. 7th 11, 742-743. Mehard, C. W. (1976). J . Cell Biol. 70, 149a. Mendelson, K. (1956). I n “Handbuch der Physik. Encyclopedia of Physics” (S. Fliigge, ed.), XV Low Temperature Physics 11, pp. 370-461. Springer-Verlag. Berlin and New York. Meryman. H. T. (1950). J. Appl. Phys. 21, 68. Meryman, H. T. (1957). Pror. R . Soc. London. Ser. B 147, 452-459. Meryman. H. T. (1971). Cryobiologv 8, 173-183. Meryman. H. T. (1974). Annu. Rev. Biophys. Bioeng. 3, 341-363. Miller, K . R., and Staehelin, L. A. (1976). J. Cell Biol. 68, 30-47. Miller, R. G. (1980). Nature (London) 287, 166-167. Molday. R. S., and Maher. P. (1980). Histochem. J. 12, 273-315. Monroe, R. G.. Gamble. W. I.,LaFarge, C. G., Gamboa. R.. Morgan. C. L., Rosenthal, A., and Bullivant, S. (1968). J. Ultrastruct. Res. 22, 22-36. Montesano, R., Perrelet, A.. Vassalli, P., and Orci, L. (1979). Proc. Nut/. Arud. Sci. U.S.A. 76, 6391-6395. Montesano, R., Vassalli, P., Perrelet. A., and Orci, L. (1980). Cell Biol. Int. Rep. 4, 975-984. Moor, H. (1964). 2. Zellforsch. Mikrosk. Anat. 62, 546-580. Moor, H. (1966). Int. Rev. Exp. Pathol. 5, 179-216. Moor, H. (1969). Inr. Rev. Cvtol. 25, 391-412. Moor, H. ( 197 I ) . Philos. Trans. R . Soc. London Ser. B 261, 12 I - I3 I , Moor, H., and Hoechli, M. (1970). Proc. In!. Congr. Electron Microsr.. 8th I, 445-446. Moor, H., and Riehle, U. (1968). Proc. Eur. Conf. Electron Microsc.. 4th 11, 33-34. Moor, H., Miihlethaler, K., Waldner, H., and Frey-Wyssling. A. (1961). J . Biophvs. Biochem. CVtol. 10, 1-13. Moor, H.. Kistler, J.. and Miiller, M. (1976). Experientiu 32, 805. Moor, H., Bellin, G., Sandri, C., and Akert, K. (1980). Cell Tissue Res. 209, 201-216. Morris, G. J. (1976a). Arch. Microbiol. 107, 57-62. Morris, G. J. (1976b). Arrh. Microbiol. 107, 309-312. Miiller, H. R. (1957). J. Ultrastruct. Res. 1 , 109-137. Miiller, M., Marti, T., and Kriz, S. (1980a). Pror. Eur. Congr. Electron Microsc., 7th 11,720-72 1. Miiller. M., Meister, N., and Moor, H. (1980b). Mikroskopie 36, 129-140. Miiller, W., and Pscheid, P. (1978). J. Microsc. 115, 113-116. Munn, E. A . , Bachmann, L., and Feinstein, A . (1980). Biochim. Biophys. Acta 625, 1-9. Nanninga. N. (1971). J. CellBiol. 48, 219-224. Nei, T. (1973). J . Microsc. 99, 227-233. Nei. T., Yotsumoto, H., Hasegawa. Y., and Hasegawa. M. (1974). J. Electron Microsc. 23, 137- 138. Nermut. M. V. (1973). I n ”Freeze-Etching: Techniques and Applications” (E. L. Benedetti and P. Favard. eds.), pp. 135- 150. Socikte FranCaise de Microscopie Electronique. Paris. Nermut, M. V. (1977). I n “Principles and Techniques for Electron Microscopy” (M. A. Hayat, ed.), Vol. VII, pp. 79-1 17. Van Nostrand-Reinhold, Princeton. New Jersey. Nermut, M. V., and Frank, H. (1971). J. Gen. Virol. 10, 37-51. Nermut. M. V., and Ward. B. J. (1974). 1.Microsc. 102, 29-39. Nermut, M. V., and Williams, L. D. (1977). 1. Microsc. 110, 121-132. Nermut, M. V., and Williams, L. D. (1980). J . Microsc. 118, 453-461.

300

HELMUT PLATTNER AND LUIS BACHMANN

Nermut, M. V., Burdett, 1. D. J., and Williams, L. D. (1978). J. Microsc. 114, 229-239. Niedermeyer, W., and Moor, H. (1976). Proc. Eur. Congr. Electron Microsc., 6rh 11, 108-1 10. Niedermeyer, W., Parish, C. R., and Moor, H. (1977). Proroplusma 92, 177-193. Orci, L., and Perrelet, A. (1975). “Freeze-Etch Histology. A Comparison between Thin Sections and Freeze-Etch Replicas.” Springer-Verlag, Berlin and New York. Ornberg, R. L.,and Reese, T. S. (1979). In “Freeze Fracture: Methods, Artifacts and Interpretations” (J. E. Rash and C. S. Hudson, eds.), pp. 89-97. Raven, New York. Ornberg. R. L., and Reese, T. S. (1980). Fed. Proc. Fed. Am. SOC. Exp. Biol. 39, 2802-2808. O’Toole, T. (1980). Washington Post. 6 May, p. A16. Painter, R. G., Tokuyasu, K. T., and Singer, S. J. (1973). Proc. Natl. Acud. Sci. U.S.A. 70, 1649- 1653. Parish, G. R. (1975). J. Microsc. 104, 245-256. Parsons, D. F., and Subjeck, J.R. (1972). Biochim. Biophvs. Acta 265, 85-113. Pawley, J. B.. and Norton, J. T. (1978). J. Microsc. 112, 169-182. Pease, D. C. (1966). J . Ultrastruct. Res. 14, 356-378. Pease, D. C. (1967). J. Ultrastruct. Res. 21, 98-124. Pease, D. C. (1973). In “Advanced Techniques in Biological Electron Microscopy” (J. K. Koehler, ed.), pp. 35-66. Springer-Verlag, Berlin and New York. Perkins, W. D., and Koehler, J. K. (1978). In “Advanced Techniques in Biological Electron Microscopy. 11. Specific Ultrastructural Probes” (J. K. Koehler, ed.), pp. 39-63. SpringerVerlag, Berlin and New York. Pfaller, W. (1978). Mikroskopie 35, 37-44. Pfaller, W., and Rovan, E. (1978). J . Microsc. 114, 339-351. Pfenninger, K. H. (1976). Balzers High Vac. Rep. DN 6923, 1-8. Phillips, H. J. (1973). In “Tissue Culture, Methods and Applications” (P. F. Kruse and M. K. Patterson, eds.), pp. 406-408. Academic Press, New York. Pinto da Silva, P. (1972). J. Cell Biol. 53, 777-787. Pinto da Silva, P., and Kachar, B. (1980). Cell Biol. In!. Rep. 4, 625-640. Pinto da Silva, P., and Nicolson, G. L. (1974). Biochim. Biophys. Acta 363, 311-319. Pinto da Silva, P., Douglas, S. D., and Branton, D. (1971). Nature (London) 232, 194-196. Plattner, H. (1970). Mikroskopie 26, 233-250. Plattner, H. (1971). J. Su6microsc.-Cytol. 3, 19-32. Plattner, H. (1978). Life Sci. Res. Rep. 11, 465-488. Plattner, H. (1979). Drug Res. 29, 1809-1810. Plattner, H. (1981). Cell Biol. Int. Rep. 5, 435-459. Plattner, H., Fischer, W. M., Schmitt, W. W., and Bachmann, L. (1972). J. Cell Biol. 53, 116-126. Plattner, H., Schmitt-Furnian, W. W., and Bachmann, L. (1973). In “Freeze-Etching. Techniques and Applications” (E. L. Benedetti and P. Favard, eds.), pp. 81-100. SociCtC FranCaise de Microscopie Electronique, Paris. Plattner, H., Wachter, E., and Grobner, P. (1977). Histochemistry 53, 223-242. Prager, E. M., Wilson, A. C., Lowenstein, J. M.. and Sarich, V. M. (1980). Science 209,287--289. Pricam, C., Fisher, K. A., and Friend, D. S. (1977). Anut. Rec. 189, 595-608. Pryde, J. A , , and Jones, G. 0. (1952). Nature (London) 170, 685-688. Pscheid, P., and Plattner, H. (1980). Proc. Eur. Congr. Electron Microsc.. 7th 11, 716-717. Pscheid, P., Schudt, C., and Plattner, H. (1981). J . Microsc. 121, 149-167. Rash, J. E. (1979). In “Freeze Fracture: Methods, Artifacts and Interpretations” (J. E. Rash and C. S. Hudson, eds.), pp. 153-160. Raven, New York.

CRYOFIXATION

30 1

Rash, J. E., and Hudson, C. S., eds. (1979). “Freeze Fracture: Methods, Artifacts and Interpretations.” Raven, New York. Rash, J. E.. Johnson, T. 1. A.. Hudson, C. S., Copio, D. S . , Graham, W. F., Eldefrawi, M. E., and Giddings. F. D. (1980). Proc. Annu. Meet., 38th, Electron Microsc. Soc. Am. pp. 692-695. Rasmussen, D. H.. and Luyet, B. (1969). Biodvnumicu 10, 319-331. Rasmussen, D. H., MacAulay, M. N., and MacKenzie, A. P. (1975). Cpobiology 12, 328-339. Raviola, E.,Goodenough, D. A,. and Raviola, G . (1980). J. Cell Biol. 87, 273-279. Rebhun, L. I. (1972). In “Principles and Techniques of Electron Microscopy” (M. A. Hayat, ed.), Vol. 2, pp. 1-49. Van Hostrand-Reinhold, Princeton, New Jersey. Rebhun. L. I., and GagnC, H. T. (1962). Proc. Int. Congr. Electron Microsc., 5th 11, 2. Rebhun, L. I., and Sander, G. (1971). Am. J. Anut. 130, 1-16. Revel. J. P. (1974). Symp. Soc. Exp. B i d . 28, 447-461. Rey. L. R . (1960).Ann. N.Y. Acud. Sci. 85,510-534. Richter, H. (1968). Protoplusmu 65, 155-166. Richter. H.. and Sleytr. U . (1971). Z. Nuturforsch. 26b, 470-473. Rick, R., Dorge. A.. Gehring. K.. Bauer, R., and Thurau, K.(1979). In “Microbeam Analysis in Biology” (C. P. Lechene and R. R. Warner, eds.), pp. 517-534. Academic Press, New York. Riehle. U . (1968a). Chem. Ing. Tech. 40, 213-218. Riehle, U. (1968b). “Uber die Vitrifizierunp verdunnter waessriger Losungen.” Diss. Nr. 4271, ETH Zurich. Riehle, U . , and Hochli. M. (1973). In “Freeze-Etching. Techniques and Applications” (E. L. Benedetti and P. Favard, eds.), pp. 31-61. SociCtC Franqaise de Microscopie Electronique, Paris. Rix, E., Schiller, A., and Taugner, R. (1977). Hisrochemistry 50, 91-101. Robards, A. W., and Severs, N. J. (1981). Crvo-Lett. 2, 135-144. Robertson, I . D., and Vergara. J. (1980). J . Cell B i d . 86, 514-528. Robinson, J. M., and Kamovsky. M. J. (1980). J. Histochem. Cvtochem. 28, 161-168. Robinson. J. M., Roos, D. S., Davidson. R. L., and Kamovsky. M. J. (1979). J. Cell Sci. 40, 63-75. Rogers, A. W. (1979). “Techniques of Autoradiography, 3rd ed. Elsevier, Amsterdam. Roomans. G . M..and Seveus. L. A. (1976). J. Cell Sci. 21, 119-127. Ross, A , . Sumner. A. T., and Ross, A. R. (1981). J . Microsc. 121, 216-272. Roth, J., Thoss, K.. Wagner, M., and Meyer. H. W. (1975). Histochemistp 43, 275-282. Rowe, T. W. G. (1960). Ann. N.Y. Acud. Sci. 85, 641-679. Russ, J. C. (1974). J. Submicrosc. Cyrol. 6, 55-79. Sakai. A,, Otsuka, K., and Yoshida, S. (1968). Crvobiologv 4, 165-173. Satir. P.. and Satir. B. (1974). Exp. Cell Res. 89, 404-407. Saubermann. A. J., Beeuwkes. R., and Peters, P. D. (1981a). J . CellBiol. 88, 268-273. Saubermann. A. J . , Echlin. P., Peters, P. D., and Beeuwkes, R. (1981b). J. CellBiol. 88. 257267. Schafer. D., and Zierold. K. (1978). Proc. Int. Congr. Electron Microsc.. 9th 11, 112-1 13. Schiller, A,. Rix, E.. and Taugner, R. (1978a). Histochemistrv 59, 9-16. Schiller, A , , Sonnhof. U . , and Taugner, R. (1978b). Mikroskopie 35, 23-30. Schiller, A,. Taugner, R., and Rix, E. (1978~).Mikroskopie 34, 19-23. Schmidt, W., Winkler, H., and Plattner, H. (1982). Eur. J. Cell B i d . 27, 96-104. Schober, R.. Nitsch, C., Rinne, U.. and Morris, S . J. (1977). Science 195, 495-497. Schuler. G . . Plattner, H., Aberer. W., and Winkler. H. (1978). Biochim. Biophvs. Actu 513, 244-254. Schwabe, K . G.,and Terracio, L. (1980). Crvobiologv 17, 571-584.

302

HELMUT PLATTNER AND LUIS BACHMANN

Scott, R. E., Maercklein. P. B., and Furcht, L. T. (1977). J. Cell Sci. 23, 173-192. Segrest. J. P., Gulik-Krzywicki, T.. and Sardet. C. (1974). Proc. Nu//. Acad. Sci. U.S.A. 71, 3294-3298. Severs, N. J. (1980). Trends Neurosci. 1, 4-7. Sevius. L. (1977). J. Microsc. 112, 269-279. Shimshick, E. J.. Kleemann. W.. Hubbel. W. L., and McConnell. H. M. (1973). J. Supramol. Sfruct. 1, 285-294. Shotton. D. (1980). Nature (London) 283, 12-14. Shotton. D.. Thompson, K., Wolfsy. L., and Branton. D. (1978). J. CeIl Biol. 76, 512-53 I . Simard, R. (1976). I n “Principles and Techniques of Electron Microscopy” (M. A. Hayat. ed.). Vol. 6. pp. 290-31 I . Van Nostrand-Reinhold, Princeton, New Jersey. Sitte, H., Fell. H.. Holbl, W.. Kleber. H., and Neumann. K. (1977). J . Microsc. 111, 35-38, Sjostrand, F. S.. and Baker, R. F. (1958). J. Ulfrasrruct. Res. 1, 239-246. Sjostrand, F. S.. and Kretzer, F. (1975). J. Ultrastruct. Res. 53, 1-28. Skaer, H. L. B., Franks, F., Asquith, M. H., and Echlin, P. (1977). J . Microsc. 110, 257--270. Skaer, H. L. B., Franks, F., and Echlin, P. (1978). Crvobiology 15, 589-602. Skaer, H. L. B., Franks, F.. and Echlin. P. (1979). Cryo-Lett. 1, 61-70. Skriver, E., Maunsbach, A. B., and Jorgensen, P. L. (1978). J. Cell Biol. 86, 746-754. Sleytr. U. B.. and Robards, A. W. (1977). J. Microsc. 111, 77-100. Somlyo, A. V., and Shuman, H. (1982). Ultramicroscopy 8, 219-234. Somlyo, A. V.. and Silcox. J. (1979). In “Microbeam Analysis in Biology” (C. P. Lechene and R. R. Warner, eds.), pp. 535-555. Academic Press, New York. Somlyo, A. V.. Shuman. H.. and Somlyo. A. P. (1977). J. Cell Biol. 74, 828-857. Somlyo. A. V., Gonzalez-Serratos. H., Shuman. H.. McClellan. G., and Somlyo. A. P. (1981). J. Cell Biol. 90, 577-594. Sperling, L.. and Klug, A. (1977). J . Mol. Biol. 112, 253-263. Speth. V., and Wunderlich, F. (1973). Biochim. Biophvs. Acta 291, 621-628. Spriggs. T. L. B., and Wynne-Evans. D. (1976). J. Microsc. 107, 35-46. Staehelin, L. A. (1974). Int. Rev. Cytol. 39, 191-283. Steere. R. L. (1957). J. Biophvs. Biochem. CytoI. 3, 45-60. Steere, R. L. (1969). Ctyobiolog?, 5, 306-323. Steinbrecht, R. A. (1980). Tissue Cell 12, 73-100. Steinbrecht. R. A,. and Zierold, K. (1982). Proc. Int. Congr. Electron Microsc.. 10th. (in press). Stephenson, J. L. (1954). Nature (London) 174, 235. Stillinger, F. H. (1980). Science 209, 451-457. Stirling. C. E., and Kinter. W. B. (1967). J. Cell B i d . 35, 585-604. Stolinski. C., and Breathnach, A. S. (1975). In “Freeze-Fracture Replication of Biological Tisues. Techniques, Interpretation and Applications.” Academic Press, New York. Stolinski. C., and Breathnach, A. S. (1976). J . Anaf. 122, 271-281. Stolinski, C.. Breathnach, A. S.. and Bellairs, R. (1978). J . Microsc. 112, 293-299. Stumpf, W. E. and Roth, L. J. (1966). J . Hisrochem. Cvtochem. 15, 274-287. Talmon, Y., Davis, H. T.. Scriven, L. E., and Thomas, E. L. (1979). Rev. Sci. Instrum. 50, 698-704. Talmon. Y.. Davis, H. T., Scnven, L. E., and Thomas, E. L. (1980). Proc. Int. Congr. Electron Microsc.. 7th 2, 718-719. Tanaka, Y., DeCamilli, P.. and Meldolesi. J. (1980). J. Cell Biol. 84, 438-453. Taylor, A. C. (1960). Ann. N . Y . Acad. Sci. 85, 595-609. Taylor. K. A., and Glaeser, R. M. (1976). J. Ultrastruct. Res. 55, 448-456. Taylor, R., Adamo, C. D. J.. Boardman, C. F. B., and Wallis, R. G. (1974). Cryobiologv 11, 430-438.

CRYOFIXATION

303

Terracio. L.. Bankston. P. W., and McAteer, J. A. (1981). Cvobiologv 18, 55-71. Tillack, T.W., and Kinsky, S . C. (1973). Eiochim. Eiophy.s. Acru 323, 43-54. Tokuyasu, K . T. (1973). J. Cell B i d . 57, 551-565. Tokuyasu. K . T. (1976). J. Ulrrustrrtct. Res. 55, 281-287. Tokuyasu. K. T. (1980). Hisrochem. J. 12, 381-403. Triche. T. J.. Tillack. T. W.. and Kornfeld, S . (1975). Biorhim. Eiophvs. Actu 394, 540-549. Umrath. W. (1977). Mikroskopie 33, 11-29, VanDcnBurgh. H. H. (1977). Eiochim. Biophys. Acru 466, 302-314. VanDeurs. 8.. and Luft, J . H. (1979). J. Ultrustrucr. Res. 68, 160-172. VanEwijk. W.,and DeVries, E. (1977). Hisrochem. J. 9, 329-340. VanHarreveld. A.. and Crowell, J. (1964). Anur. Rec. 149, 381-385. VanHarreveld. A.. Crowell, J.. and Malhotra. S. K. (1965). J. Cell B i d . 25, 117-137. VanHarreveld. A.. Trubatch. J., and Steiner. J. (1974). J. Microsc. 100, 189-198. VanVenetie. R.. Leunissen-Bijvelt. J., Verkleij. A. J.. and Ververgaert, P. H. J . (1980).J. Microsc. 118, 401-408. VanVenetie. R.. Hage, W. J., Bluemink. J. G . , and Verkleij, A. J. (1981). J. Microsc. 123, 281-292. VanVenrooij. G . E. P., Aertsen, A. M. H . , Hax, W. M. A., Ververgaert. P. H. J., and Verhoeven, J. J . (1975). Crvobiolop 12, 46-61. Venkatesh. C. G.. Rice. S. A,. and Narten, A. H. (1974). Science 186, 927-928. Verkleij, A. J., and Ververgaert, P. H. J. (1978). Eiochim. Eiophvs. Acru 515, 303-327. Verkleij. A . J.. Ververgaert. P. H. J., VanDeenen. L. L. M., and Elbers, P. F. (1972). Eiochim. Eiophvs. Acra 288, 326-332. Verkleij, A. J., DeKruijff, B., Gerritsen, W. F., Demel. R. A.. VanDeenen, L. L. M., and Ververgaert. P. H. J . (1973). Eiochim. Eiophys. Acru 291, 577-581. Verkleij, A. J.. Mombers, C., Leunissen-Bijvelt. J.. and Ververgaert, P. H. J. (1979). Nature (London) 279, 162- 163. Verkleij, A. J . , VanEchteld, C. J. A , , Gerritsen. W. J.. Cullis, P. R.. and DeKruijff, B. (1980). Eiochirn. Eiophys. Acru 600,620-624. Ververgaert, P. H. J., Verkleij, A. J., Elbers, P. F.. and VanDeenen, L. L. M. (1973a). Biochim. Eiophvs. Aeru 311, 320-329. Ververgaert, P. H. J . , Verkleij, A. J.. Verhoeven. J. J., and Elbers. P. F. (1973b). Eiochim. Eioph>l.y. ACIU 3l,,651-654. Vilmart, J., and Plattner, H. (1982). Proc. hi.Congr. Electron Microsc., IOrh. (in press). Wade, J. B., and Kamovsky, M. J. (1974). J. Cell B i d . 60, 168-180. Walzthony, D., Moor, H., and Gros, H. (1981). Ultrumicroscopv 6, 259-266. Weast. R. C. (1976). “Handbook of Chemistry and Physics,” 57th ed. CRC, Cleveland, Ohio. Weller, N. K. (1974). J. Cell Biol. 63, 699-707. Wendelaar Bonga, S. E., and Veenhuis, M. (1974). J. Cell Sci. 14, 587-609. Wendt-Gallitelli. M. F.. Wolburg, H.. Schwegler. M.. and Schlote. W. (1979). Experienria 35, 1591- 1593. Werner, G., Neumann, K., and Morgenstern, E. (1973). J. Microsc. 99, 219-225. Williams, R. C. (1954). In “Biological Applications of Freezing and Drying to Electron Microscopy” (R.J. C. Harris. ed.), p. 303. Academic Press, New York. Williams, R. J.. and Harris, D. (1977). Cpobiologv 14, 670-680. Wilson, A. J., and Robards, A. W. (1980). Cvo-Letr. 1, 416-425. Wilson, A. J., and Robards, A. W. (1982). J. Microrc. 125, 287-298. Withers, L. A. (1978). Proloplasma 94, 235-247. Withers, L. A,, and Davey. M. R. (1978). frotoplusma 94, 207-219.

304

HELMUT PLATTNER AND LUIS BACHMANN

Wolf, K. V., Stockern, W., and Wohlfdrth-Bottermann, K. E. (1981). Cell Tissue Res. 217, 479-495. Wollenberger. A,, Ristau, 0.. and Schoffa, G . (1960). Pfliiger’s Arch. Phvsiol. 270, 399--412. Wolosewick, J . 1.. and Porter, K. R. (1979). J. CellBiol. 82, 114-139. Woolley. D. M. (1974). J . Microsc. 101, 245-260. Wunderlich. F., Speth. V.. Batz, W.. and Kleinig, H.(1973). Biochim. Biophvs. Actu 298, 39-49. Wunderlich, F., Wdllach, D. F. H., Speth, V., and Fischer, H. (1974). Biochim. Biophvs. Acta 373, 34-43. Yannas. 1. (1968). Science 160, 298-299. Zachariassen, K. E. (1979). J. lnsecr Phvsiol. 25, 29-32. Zalokar, M. (1966). J. Ultrastruct. Res. 15, 469-479. Zierold. K.. and Schafer, D. (1978). J. Microsc. 112, 89-93. Zierold, K., Konig. R., Olech, K. H., Schafer, D., Lubbers, D. W.. Muller, K. H., and Winter, H. ( I98 I). Ultramicrosropy 6 , I8 1- 186. Zingsheim. H. P., and Plattner, H. (1976). I n “Methods in Membrane Biology” (E. D. Korn, ed.), Vol. 7, pp. 1-146. Plenum, New York. ZS-Nagy, I., Pieri, C., Giuli, C.. Bertoni-Freddari, C., and Sz-Nagy, V. (1977). J. Ultrastruct. Res. 58, 22-33.

INTERNATIONAL REVIEW OF CYTOLOGY. VOL. 79

Stress Protein Formation: Gene Expression and Environmental Interaction with Evolutionary Significance C. ADAMSAND R. W. RINNE U.S. Department of Agriculture, Agricultural Research Service, and Department of Agronomy, Universir?, of Illinois. Urbana, Illinois I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 11. Response to Environmental Stresses

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

A. Heat . . . . . . . . . . . B. Chemical Reagents .......................... C. Starvation.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Anoxia . . . . . . . . . . . . . . . .......... E. Wounding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111. Molecular Events in Stress Protein Formation.. . . . . . IV. Function of Stress Proteins. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Evolutionary Significance of Stress Proteins . . . . . . . . . . . . . . . . . . . References . . . . . . . . ....................

305 306 306 308 308 309 309 3 10 31 I 312 3 14

I. Introduction Cells of a metazoan organism have many different phenotypic forms yet usually contain an identical component of information encoded in DNA. The particular characteristics displayed by a given cell are the result of the controlled expression of a portion of this information throughout the course of development. The utilization of this information is controlled through regulating the spectrum of active genes. In eukaryotes there seems to be a diversity of ways by which gene expression is controlled (Brown, 1981). Gene expression can be influenced by a variety of mechanisms such as the loss, amplification, and rearrangement of genes. Differential transcription of genes may occur and there may be variations in the utilization of the RNA transcripts. In addition to the normal changes in gene activity throughout development, changes also occur as a direct consequence of the environment. Both the external environment in which organisms are grown and the internal or cellular environment in which the gene expression occurs can drastically affect phenotypic expression of various genotypes. A dramatic example of environmental influence on gene expression is provided by sun red Zea mays (Emerson, 1921). In sun red genotypes anthocyanin is 305 Copyright t3 I Y X ? by Academic Press. Inc. All righa uf reproduction in any I o n reserved. ISBN 0-12-364479.8

306

C. ADAMS AND R . W. RINNE

formed only in those parts of the plant exposed to sunlight. Temperature can also readily influence phenotype. Abnormal phenotypes from crosses of Phaseolus vulgaris were accentuated by growth temperatures of 30°C day/25"C night. At a 25"C/2OoCgrowth regime the growth rates of the hybrids were similar to those of the parents (Shii et a l . , 1980). The pervasiveness of genotype X environment interaction was demonstrated by Taylor and Condra (1978). They grew 12 strains of Drosophila pseudoobscura on 10 different foods at 2 temperatures and measured 5 components of fitness. Significant genotype X environment interactions were found in all the traits observed. Considerations of gene expression and environment revolve around how the cell notifies the gene which lies within the chromosome, inside the nucleus, when the necessity arises for the production of a particular protein or sequence of proteins. Many phenotypic traits such as size, anatomical shape, color differences, or growth rates are too complex to offer insight into genotype X environment interaction. However deliberate perturbation of the environment by subjecting organisms to environmental stresses produces changes in gene expression of a less complex nature.

11. Response to Environmental Stresses

Specific changes in gene expression in response to environmental stresses are brought about by heat, chemical reagents, starvation, anoxia, and wounding. These alterations of gene activity have been observed in a wide range of eukaryote species. A. HEAT Possibly the most widely studied stress-related example of gene expression is the heat shock syndrome in Drosophila. If Drosophila larvae, excised tissue, or tissue culture cell lines, are subjected to a brief heat shock by increasing the growth temperature from 25 to 37°C there is a dramatic alteration in the pattern of gene transcription (Ashburner and Bonner, 1979). Most of the normal mRNA synthesis is suppressed and specific newly synthesized mRNA appears (Spradling er al., 1975). Specific polypeptides, heat shock proteins (hsp) are produced which may account for 50% of the total protein synthesis at 37°C (Koninkx, 1976). There is a very rapid breakdown of almost the entire population of polysomes and a cessation of normal protein synthesis. About 10 minutes after the start of heat shock new polysomes appear which contain newly synthesized RNA transcribed from induced heat shock genes. These messengers in turn produce the hsp which comprise about eight different polypeptides in D .

STRESS PROTEIN FORMATION

307

mefanogasrer with molecular weights of 22,000, 23,000, 26,000, 27,000, 36,000,68,000,70,000, and 82,000 (Ashburner and Bonner, 1979). There seem to be only six hsp produced by D . hydei (Koninkx, 1976). The basic mechanism of the stress response seems to be that heat shock results in the production of a specific set of mRNAs. Some of these are preferentially translated into hsp. In Drosophila this environmentally induced response is very rapid and hsp can be detected some 10 minutes after heat treatment. The heat shock treatment response seems universal in occurrence and widely disparate species have been shown to produce hsp. Cells of Tetrahymenapyriformis rapidly synthesize a set of specific polypeptides in response to heat shock (Guttman et al., 1980). At least 16 polypeptide species were induced and many accumulated in substantial amounts within 50 minutes of the stress. The molecular weight distributions of these stress proteins were very similar to the hsp of Drosophila. Heat shock in Terrahymena does affect RNA production with rRNA being reduced more severely than mRNA (Hermolin and Zimmerman, 1976). At least two polypeptides were synthesized in response to heat shock in cells of Chironomus tentans (Vincent and Tanguay, 1979). Chick embryo fibroblasts produced three hsp of molecular weights 22,000, 76,000, and 95,000 immediately after heat shock (Kelley et al., 1980). These three polypeptides accounted for about 50% of all proteins synthesized following heat stress. The increased protein synthesis did not occur if actinomycin D was present during heat shock indicating the formation of mRNA was required to code for the hsp. In both avian and mammalian cells the hsp were synthesized in relatively low levels in uninduced cells and the increase in synthesis after induction is directly attributable to an increase in mRNA specific for these proteins (Wang et a f . , 1981). Heat shock response has been observed in Saccharomyces cerevisiae after a shift from 23 to 36°C (Miller er al., 1979; McAlister and Finkelstein, 1980). A coordinate change in the synthetic rates of a large number of cellular proteins occurs which reflects changes in the populations of translatable mRNAs. The preferential synthesis of a group of polypeptides at a particular stage in the response is due to the relative abundance of translatable mRNA for those polypeptides. Higher plants also exhibited heat shock response (Key et al., 198I). When the growth temperature of soybean seedlings was elevated from 30 to 40°C there was a rapid conversion of polyribosomes into monoribosomes, a decreased translation of the 30°C mRNA, initiation and translation of a set of poly(A+) mRNAs which gave rise to a set of hsp. On gel electrophoresis soybean hsp appeared very similar to those of Drosophila. The in vitro translation of poly(A+) mRNA from normal and heat-shocked soybean tissue coupled with the in vivo patterns of protein synthesis at two temperatures indicate that new mRNAs are synthesized in response to heat

308

C. ADAMS AND R . W . RINNE

shock. Also some translational event renders the 30°C mRNA poorly translatable at 40-42.5"C even though many of the abundant poly(A + ) mRNAs persist at the higher temperature for some time. This suggests that heat stress causes a rapid induction of synthesis of a new set of mRNAs.

B. CHEMICAL REAGENTS Kethoxal, a metal-binding ligand, with antitumor activity, induced the synthesis of four proteins in normal chick embryo cells (Levinson et al., 1978a). Induction of these proteins was closely coordinated, and inhibited by actinomycin D and cycloheximide. This again suggests that both de novo RNA and protein synthesis was required for the stress proteins to appear. Disulfiram, a copper-chelating drug, exerted a similar response in chick and human cells (Levinson et al., 1978b). The transition metals, copper, cadmium, zinc, and mercury, can also induce specific protein synthesis in avian and human cells (Levinson et al., 1980). Stress proteins appeared 1 to 3 hours after exposure and were electrophoretically similar to those induced by copper chelating drugs or by heat shock. Chick embryo cells treated with sodium arsenite produced stress proteins which again were quite similar to the hsp of Drosophila (Johnston et al., 1980; Wang et al., 1981). Exposure of the cells to actinomycin D or cycloheximide before addition of arsenite prevented induction of the protein. Translation in vitro of poly(A ) mRNA from arsenite-treated cells demonstrated the presence of increased amounts of mRNA specific for those proteins. Amino acid analogs, canavanine, hydroxynorvaline, and 0-methyl-threonine, caused chick embryo fibroblasts to rapidly increase the synthesis of three proteins of molecular weights 22,000, 76,000, and 95,000. These proteins accumulated to levels which dominated the cell's biosynthetic capacity and exceeded the level of synthesis of the major structural proteins (Kelley and Schlesinger, 1978). Induction of a similar set of proteins was also observed in mouse and hamster cells after similar treatment. Growing cells at elevated temperatures of 41 -45°C could also induce the formation of these same proteins (Kelley and Schlesinger, 1978). +

C. STARVATION The concentration of glucose in the culture medium exerted an effect on protein synthesis in chick embryo cells (Shiu et al., 1977). Glucose-starved cells had an increased capacity for synthesis of two proteins of 78,000 and 95.000 MW. Protein increase was secondary to rapid glucose depletion. The stress response was also connected to glycosylation reactions since feeding glucose derivatives, glucosamine and 2-deoxyglucose, which interfere with glycosylation could also induce stress protein synthesis (Pouyssegur et al., 1977).

STRESS PROTEIN FORMATION

309

Excised pea root tips grown in a carbohydrate-deficient medium showed a prominent increase in the relative amounts of at least one polypeptide species (Webster, 1980). There was a highly preferential synthesis of this polypeptide in cells when cell proliferation was arrested and overall protein synthesis was declining . D. ANOXIA

In roots of rice seedlings major differences were observed between polypeptides synthesized in air and under nitrogen (Bertani et al., 1981). Anaerobic conditions can induce alcohol dehydrogenase proteins in maize seedlings (Sachs and Freeling, 1978). Sachs et al. (1980) showed that anaerobic treatment drastically altered the pattern of protein synthesized by maize primary roots. During the first hour of anaerobiosis, aerobic protein synthesis was halted and there was an increase in polypeptides of molecular weight about 33,000. During the second hour of anaerobic treatment synthesis of another group of polypeptides was initiated. This group accounted for over 70% of total protein synthesis after 5 hours of anaerobiosis. Alcohol dehydrogenase polypeptides were the major proteins produced. The RNAs isolated from roots after 24 hours of anaerobiosis directed the translation only of anaerobic specific proteins. However RNAs from roots treated anaerobically for 5 hours directed translation of both aerobic and anaerobic polypeptides. An early response to anaerobic treatment is the suppression of aerobic message translation. Alcohol dehydrogenase production can be induced in tomato plants by anaerobiosis similar to that observed in maize (Tanksley and Jones, 1981). A single gene accounts for the majority of the alcohol dehydrogenase present in uninduced tissue but expression of this gene and another independent gene is stimulated by anaerobic stress. In tomato seedlings maximum activity was found in roots. Excised tomato leaves though also responded to the stress as well as other tissues which suggests that the response occurs at the cellular level and is not mediated by some hormonal control. E. WOUNDING Deciliation of Tetrahymena pyriformis results in an alteration of patterns of protein synthesis similar to that caused by heat shock (Guttman et al., 1980). Activation of RNA and protein synthesis is initiated in many higher plant tissues in response to mechanical injury (Uritani, 1976). Freshly cut slices of artichoke tuber contained no detectable invertase activity but when they were suspended in aerated distilled water activity developed and reached a maximum after about 3 days (Edelman and Hall, 1965). Ascorbate oxidase activity also increased after slicing. Incubation of slices of pumpkin mesocarp caused a marked increase in

3 10

C. ADAMS AND R. W. RINNE

protein synthesis which began within 3 hours of slicing and was largely complete within 24 hours (Ap Rees, 1969). Slicing of a plant storage tissue into thin discs seems to transmit a signal to the genome triggering increased synthetic activities leading to a coordinated genetic response. Incubation of carrot discs induced the production of a substantial amount of mRNA which coded for cell wall protein, although this mRNA was apparently lacking in fresh carrot discs (Smith, 1981). The slicing and incubation procedures clearly induced or altered gene expression in favor of the production of large amounts of hydroxyproline-rich cell wall protein. Wounding of tomato or potato leaves by attacking pests or other mechanical damage releases a putative wound hormone that is transported throughout the plant where it induces leaves to initiate synthesis and accumulation of two proteinaceous inhibitors of serine endopeptidases (Bishop et al., 1981 ). Wounding seems to release a low-molecular-weight oligosaccharide from plant cell walls that induces massive protein synthesis. Within 48 hours of detachment, young potato leaflets, initially devoid of inhibitors, contained over 3% of the total soluble protein as inhibitor. This represented 12% of all new protein synthesized (Ryan and Huisman, 1970).

111. Molecular Events in Stress Protein Formation

At the molecular level the general stress response seems to be a stimulation of transcription of specific mRNAs and repression of translation of normal mRNAs. This stimulation of transcription however may require cytoplasmic factors as sources of information. The cytoplasm in several tissues can independently mediate and regulate synthesis of some proteins. Both nucleate and enucleate cells of Acetabularia crenuluru can regenerate a cap. During cap regeneration synthesis of a specific alkaline phosphatase occurred in both nucleate and enucleate cells (Spencer and Harris, 1964). Clearly the cytoplasm of Acetabularia cells contained an independent mechanism not only for synthesis of a specific enzyme but also for continuous regulation of this synthesis over a period of several weeks in response to a wounding stress, that is, cap removal by scissors. Activity of the enzyme tryptophan pyrrolase in rat liver increased significantly after administration of tryptophan (Greengard et al., 1963). The enzyme increase was insensitive to both actinomycin and adrenalectomy , which indicated no requirement for RNA synthesis nor hormonal effects. Possibly much of the KNA for the synthesis of the tryptophan pyrrolase was present in liver cells before the accumulation of the enzyme was detectable. Cytoplasmic protein synthesis may exist to a considerable extent without requiring continuous inputs of mRNA from the nuclear genome. A spectrutn of

STRESS PROTEIN FORMATION

31 1

stable mRNAs may exist in the cytoplasm. Perturbation of the environment could result in rapid translation of these messages or lead to translation of modified proteins. Such environmentally stimulated cytoplasmic proteins could then serve as a signal to the nuclear genome for activation of specific genes. Such a concept is supported by the ability of amino acid analogs to induce stress proteins (Kelley and Schlesinger, 1978). These workers suggest that the analog is incorporated into polypeptide chains to give an aberrant protein. This protein acts as a signal to alter gene expression affecting transcriptional and translational events. The wounding response of potato and tomato leaves is brought about by release of pectic polysaccharides from cell walls (Bishop et al., I98 1). These materials then stimulate induction of proteinase inhibitor proteins. In Drosophila cytoplasmic extracts from heat-shocked cells were able to confer characteristics of heat shock on nuclei isolated from non-heat-shocked cells (Craine and Kornberg, 1981). The cytoplasmic factor was found to be a protein which was probably a usual component of cells during normal growth but after the heat shock was modified somehow to become a source of information which altered genome expression. Stress proteins are one of the few systems where a hostile environment mediates the synthesis of a select few polypeptides. Yet the universality of the stress protein responses argues for its importance in some sort of protective measure.

IV. Function of Stress Proteins The unifying theme in the environmental stress responses described above is that a perturbation of the environment seems to influence, directly and immediately, gene expression. This is mediated through the suppression of some genes and activation of others to produce specific mRNAs. These mRNAs are produced in quantity only after the organism or cell has perceived an environmental change, which may be elevation of temperature, exposure to a chemical, deficiency of a nutrient, anoxia, or wounding. The universality of this response which occurs in so many divergent eukaryotic species argues for its importance although precise functions for the stress proteins are not well established. Stress agents may damage DNA and induced proteins may be a part of the DNA repair process (Levinson et al., 1978a). In Drosophila. Velazquez et al. (1980) showed that hsp were rapidly transported to the nucleus where they accumulated. They were clearly not associated with the mitochondria although it has been suggested that mitochondria were the primary target of the induced stress proteins (Ashburner and Bonner, 1979). The hsp may be involved in preservation of the spatial organization of transcriptionally active chromatin (Levinger and Varshavsky, 1981). Some 30-35% of hsp in purified nuclei were associated with nuclease-resistant and high salt-resistant structures. Since tran-

312

C. ADAMS AND R. W. RINNE

scription is one of the processes affected by heat shock response it seems likely that nuclear hsp play a role in protecting the template active portion of the genome from adverse effects of heat shock. If the stress protein response is designed to correct or protect cellular machinery then it is likely that such a system would be widely adopted and highly conserved. A more obvious physiological function of stress protein production can be deduced from root anaerobiosis studies (Sachs and Freeling, 1978; Sachs et al., 1980; Tanksley and Jones, 1981). Here the stress proteins possess alcohol dehydrogenase activity which could be useful in maintaining respiration under conditions where oxygen was limiting. In this case the stress proteins are not protecting the structural integrity of the genome but executing a sparing action on intermediary metabolism. Indeed maize roots can survive up to 72 hours of anaerobiosis before cell death sets in so in this regard induction of stress protein may be a helpful survival mechanism (Sachs et al., 1980).

V. Evolutionary Significance of Stress Proteins Patterns of gene regulation are possibly central to evolutionary change since there is a remarkable similarity of structural gene function between simple and complex organisms (Britten and Davidson, 1971). About 93% of the enzyme activities listed in the International Enzyme Commission tables are known to be present in both prokaryotes and mammals. Therefore evolutionary events leacling to the development of different organisms are unlikely to be the product of numerous mutations in the spectrum of enzymes produced by different organisms. Furthermore it has been suggested that most genetic differences have no physiological significance at all (Kimura and Ohta, 1973; King and Jukes, 1969). This theory explains genetic diversity as due in large part to neutral allzles in the population which are on the way to fixation or elimination by random genetic drift. Support for this theory can be obtained from the widespread occurrence of dispensable proteins in plants (Adams and Rinne, 1981). Several plant proteins such as urease, P-amylase, lectins, proteinase inhibitors, and some proteinases have a variable distribution with no obvious physiological function. These proteins are dispensable in that some plants manage without them or with only extremely low levels whereas other species accumulate significant quantities of the protein. Dispensable proteins seem to have no readily discernible phenotypic effects but are most decidedly genetic traits. Major phenotypic differences must appear from patterns of gene regulation which result in variations in cellular differentiation. Trewavas (1981) has proposed the concept of environmental signals as functional development agents in plants. This ascribes a greater controlling influence of environment than is generally appreciated. To influence development, howev-

STRESS PROTEIN FORMATION

313

er, environmental signals must be able to affect the genomic activity. Stress protein induction illustrates that in certain documented instances this does occur. Nag1 (1979) has taken response to environment even further and postulates that speciation is one adaptive device through which living organisms can master a progressively greater range of environments. The stress protein response does illustrate that environmental parameters can exert a specific influence on gene expression through transcriptional and translational events. Stress protein formation may have some relevance in evolutionary theory since environmental variation can exert an influence on the genome in terms of which genes are switched on or off. Populations exposed to heterogeneous environments maintain more genetic variation than populations living in uniform environments (McDonald and Ayala, 1974). Ho and Saunders (1979) point out that the commonly observed genotype X environmental interaction is a manifestation of the epigenetic systems which actually interact with the environment. This epigenetic or cytoplasmic system ultimately generates phenotypic variations in response to environmental inputs. A possible scenario for environmentally induced evolution is that environmental parameters cause cytoplasmic changes at the cellular level which in turn influence gene expression. An environmentally induced cytoplasmic change is the production of alkaline phosphatase in enucleated cells of Acetabularia (Spencer and Harris, 1964). A cytoplasmic protein is produced by Drosophifa heat-shocked cells which was able to confer characteristics of heat shock on nuclei from non-heat-shocked cells (Crane and Kornberg, 1981). Induction of stress proteins is the result of gene regulation using information from outside the cell to produce specific gene products. Therefore the so-called “normal” gene products may be expressed because of “normal” environmental inputs to the cytoplasm. If the environment changes then an altered genomic program might be followed which would be hereditary. Therefore, conceivably inheritance of acquired characteristics would be observed. In order for the characteristic to be acquired, however, the relevant genes must exist in the first place, but they might only be expressed in response to environmental changes. Eukaryotic genomes in general have much more DNA than is needed to code for the number of proteins actually produced. In higher plants only 5- 10%of the DNA is considered to be necessary to specify the known sequence specific chromosome functions (Flavell. 1980). In view of the large excess of DNA in eukaryotic cells it is difficult to specify the total information available or its possible program of expression. Much of the DNA in higher organisms may have a regulatory function (Britten and Davidson, 197 1) and be capable of responding to diverse environmental inputs. It is conceptually easier to envisage environmental parameters controlling patterns of gene regulation rather than to involve environmentally induced muta-

314

C. ADAMS AND R. W. RlNNE

tion for a specific character. In vitw of the great qualitative similarity of organisms in terms of enzyme complement (Britten and Davidson, I97 1 ), an important aspect of evolutionary progress may be in patterns of gene action. The possibility of environmental effects generating a quasi-Lamarckian inheritance has been discussed in theoretical terms by Ho and Saunders (1979). However. the universal occurrence of stress proteins indicates such a system has some basis in experimental observation.

REtt.KENCtS Adanis. C. A , , and Rinne. R. W. ( I98 I ) . NCW Phyrol. 89, I - 14. Ap Rees, T. (1969). Phytochemistry 8, 1866- 1879. Ashburner. M.. and Bonner. J. J . (1979). Cc4l 17, 241-254. Bcrtani, A,. Menegus. F.. and Bollini. R. (1981 1. Z. l'f7rnienphy.siol. 103, 37-43. Bishop. P. D., Makus. D. J., Pearce, G.. and Ryan, C. A. (1981).Proc. Nut/. Acad. Sci. U.S.A.78, 3536-3540. Britten. R. J . , and Davidson, E. H. (1971). Quurt. R w . Biol. 46, lIl-13X. Brown. D. D. (1981). Science 211, 667-674. Craine. B. L.. and Kornberg, T. (1981). Cell 25, 671-681. Edelman. J.. and Hall. M. A . (1965). Biochem. J. 95, 403-410. Emerson, R. A. (1921). Cornell .!/nil$. Agr. E.rp. Stu. Memoir 39, 156 pp. Flavell. R . (1980). Annu. Rc.17. Plant Physiol. 31, 569-596. Greengard. 0.. Smith, M. A,. and Acs. G. (1963). J. Biol. Chem. 238, 1548-1551. Guttman, S . D., Glover, C. V. C.. Allis, C. D.. and Gorovsky. M. A. (1980). Cell 22, 299--307. Hermolin. J . . and Zimmerman. A. H . (1976). J . Proto:ool. 23, 594-600. Ho, M. W.. and Saunders, P. T. (1979). J . Tlieorer. B i d . 78, 573-591. Johnston. D., Oppermann. H.. Jackson, J., and Levinson, W . ( 1980). J . B i d . Chem. 255,697--698. Kelley. C. M.. and Schlesinger, M. R. (1978). Cell 15, 1277-1286. Kelley, P. M.. Aliperti. G.. and Schlesinger. M . J. (1980). J. Biol. Chem. 255, 3230-3233. Key. J. L., Lin, C. Y., and Chen. Y. M. (1981). Proc. Nurl. Acad. Sci. U.S.A. 78, 3526-,3530. Kimura, M., and Ohta. T. (1971). Nature (London) 229, 467-469. King. I . L.. and Jukes. T. H. (1969). Science 164, 788-798. Koninkx, J. F. J . G. (1976). Biochem. J . 158, 623-628. Levinger. L., and Varshavsky, A. (1981). J. Cell Biol. 90, 793-796. Levinson. W.. Oppermann, H., and Jackson, J. (1978a). Biochim. Biuphys. Acta 518, 401--412. Levinson. W.. Mikelens, P.. Oppermann. H.. and Jackson. J . (1978b). BicJchim.Biophys. Acrc 519, 65-75. Levinson. W.. Oppermann. H.. and Jackson. J. (19x0). Biochim. Biophys. Acra 606, 170--180. McAlister. L.. and Finkelstein, D. B. (1980). J. Bacreriol. 143, 603-612. McDonald, J. F.. and Ayah, F. J. (1974). Nurirrc (London) 250, 572-573. Miller. M. J . . Xuong. N . H.. and Geiduschek. E. P. (1979). Proc. Natl. Acad. Sci. U.S.A. 76, 5222-5225. Nagl, W . (1979). In "Plant Systematics and Evolution. Supplement 2" (W. Nagl. V. Hemleken, and F. Ehrendorf, eds.), pp. 3-25. Springer-Verlag, Berlin and New York. Pouyssegur. J., Shiu, R. P. C., and Pastan. I . (1977). Cell 11, 941-947. Ryan. C. A.. and Huisman. W. (1970). Plant Phvsiol. 45, 484-489.

STRESS PROTEIN FORMATION

315

Sachs. M. M.. and Freeling. M. (1978). Molcc. Ceti. Gemr. 161, 111-115. Sachs. M . M . . Freeling. M.. and Okinioto, R . (1980). Cell 20, 761-767. Shii. C. T . . Mok. M . C.. Temple. S. R . . and Mok. D. W. S. (1980). J. Hered. 71. 218-222. Shiu, R. P. C.. Pouysscgur. J.. and Pastan, I. (1977). Proc. Not/. Acud. Sci. U . S . A . 74.3840-3844. . 9.56-963. Smith. M . A . (19811. Plunt P h ~ s i o l 68, Spencer, T . , and Harris. H . (1964). E i o c ~ h c w i J. . 91. 282-286. Spradling. A , . Penman. S.. and Pardue. M . L. (1975). Cell 4, 395-404. Tanksley. S. D.. and Jones. R . A. (1981 1. Eiochetn. Genc,~.I Y , 397-409. Taylor. C. E . . and Condra. C . (1978).J. Hered. 69, 63-64. Trewavas. A. (1981). Plmi/ Cell Eiiiiron. 4, 203-228. Uritani. I. ( 1976). / t i “Physiological Plant Pathology. Encyclopedia of Plant Physiology New Serieh” ( R . Heitcfuh and P. H. Williams. eds.). Vol. 4. pp. 509-525. Springer-Verlag. Berlin and New York. Velazquci. J . M.. DiDonienico, B. J.. and Lindquist. S. (1980). Cell 20, 679-689. Vincent. M . . and Tanguay. R . M . (1979).Ntr/rrru (London) 281, 501-SO3. Wanp, C.. Gomer. R . H.. and Lazarides, E. (1981 ) . Proc.. Ntrtl. A i d . Sci. U.S.A. 78,3531-3535. Wcbster, P. L. (19x0). Plerrit Sc.i. L e / / . 20, 141-145.

This Page Intentionally Left Blank

Index

A

Anaphase. early. kinetochore fiber and. 30-3 I Anoxia. responses to. 309

C Calniodulin. fertilization and.. I34 Ccll(s). degcncrate. means of elimination.

223-225 cxtrusion into extracellular milieu. 225 phagocytosis. 226- 227 Cell cycle. changes in nuclear organization during G I and S phases: DNA synthesis and chromatin xssenihly. 198-203 (3. pha5c and mitosis. 203-205 Cell death dctcrniination of clegcncl-ation hy cellular interiction control iiicchanisnis. 229-23 I sign;ils derived hy inlcraction. 227-229 Illodes of ;iutoph;igic dcgcncration. 2 19-222 dcgcncration by primary or precocious pycnosis. 223 dilt'u\c dcgcncr;ition. 216-219 Centriole\. fertilization and. 135- I37 Ccnlronicrc. vcrsu\ hinctochorc. 2 - 6 Chemic;il reagents. rcrponses 10, 308 Chroiiiatin intranucleiir distribution of. I80 - I X3 nucIeci\onial fihcr. 17X higher order folding of. I7X- I XO 317

nucleosome and, 175- 177 structural elements in, 183-188 transcription of. 190- I9 I Cryofixation hasic methodological aspects, 240-243 conihinalion with other techniques. 280-28 I free7e-fracturing. 28 1-286 other cytochcmical procedures. 2x8-289 surface analyses. 286-288 as tool in cytocheniistry. 281 ultrathin xctioning!. 289-290 cooling rates. 262-268, 271-273 emulsification technique. 253-254 experimental consequences. 243-247 nonaqueous systems and, 292-293 quality of general remarks. 279-280 physical tests. 27X viability tests. 278-279 resolution of. 273-274 scope and goal of. 237-240 suspensions, ciiiulsions and solutions. 290-29 I dctcrniination of particle weight hy counting, 292 heated solutions. 292 size and shape of niacroniolccules. 291 techniques-advantages and restrictions iidvanced techniques. 2.50-27 I conventional. 248-250 ultrastructural side-effects of chemical pretreatment. 274-278 Cryogcn @-freezing. 255 sandwich. 255-259 Cyclic nucleotidcs. fsnilization and. 130- 134

318

INDEX

Cytochemistry, cryofixation and, 28 I . 288-289

D Deoxyribonucleic acid hynthesis and chromatin assembly. 198-203 topological distribution of. 188

E Egg detection of cytoskeletal elements during fertilization. 86-87 biochemical isolations, 90-94 changes during. 94-95 immunofluorescence microscopy. 89-90 transmission electron microscopy. 88-89 fertilization and. cortical restructuring. 77-80 movements during. 84-86 movements leading to cell division, 83-84 pronuclear migrations, 80-83 sperm incorporation, 64-77

motility. 140-146 regulation of motility. 146-148 overview. 60 perspectives and conclusions. 152- 155 regulation of motility and calmodulin. 134 cyclic nucleotides. 130- I34 ionic controls, 120- I30 onset of fertilization, 119- 120 overview, 134- I35 requirement for movement in. 61-62 Frecze-fracturing. cryofixation and. 28 I -:!86 Freezing o n cold metal surfaces, 259-261. 271 high-pressure. 269-270 sandwich. 254-255

H Heat. responses to, 306-308

I

Interphase, kinetochore and. 8- I2

K F Fertilization consequences for later embryonic development centrioles. 135- I37 establishment of first embryonic axis. 137-138 fate of sperm tail. 138-139 detection of cytoskeletal elements during. 86-87 egg. 88-95 sperm, 87-88 dynamics of egg. 64-86 spermatozoon, 62-64 effects of motility inhibitors microfilament inhibitors. 95- 101 microtubule inhibitors, 101- I17 summary. I 17- I I9 model of regulation of motility during mechanisms for movement. 148- 152

Kinetochore mammalian. structure and chemistry interphase, 8- I2 metaphase, 22-27 prometaphase. 15-22 prophase, 12- 15 organizer, 27-29 types of. 6-8 versus centromere, 2-6 Kinetochore fiber, mammalian. 29-30 general properties during metaphase and early anaphase, 30-31 microtubule formation. 39-53 structure and composition. 31-39

M Metaphabe kinetochore and. 22-27 kinetochore fiber and, 30-31

319

INDEX Microfilaments, inhibitors of. assembly. 95- I00 disassembly. 100- I01 Microtuhule fomiation. kinetochore fiber and, 39-53 inhibitors of. assembly. 101- 10Y dis;i.~cmhly, 109- I17 Mitosis. Gz phase and. 203-205

N Nucleolus. cxtranuclcolar transcription and. 194-196

Nuclcus changes in organization during cell cycle GI and S phases: DNA synthesis and chromatin assenihly. IYX-203 G? phasc and mitosis. 203-105 chromatin. 174- 175 higher order folding of nuclcosomal fiher. 17% I80 nucleosornal fiber. 17X nucleosome. 175- 177 envelope. 166- I67 functional activities ot. 172- 174 lamina, 169- 172 membrane. 167- 169 pore element\. 16Y topological organi7ation of interphasc DNA and chromatin intranuclear distribution of chromatin. I xo- I x3 structural elements in chromatin. 183- I X X topological organi7ation of DNA. I88 transcription. processing and piickaging of RNA. 1x8-I90 cxtrenucleolar transcription. 19 I - 1Y4 nucleolus and nucleolw transcription,

P Prometaphase, kinetochore and. 15-22 Prophase. kinetochore and. 12- I5

R Ribonucleic acid. transport out of nucleus, 197-198

S Sperm. fertilization and. 62-64. X7-XX Spray-freezing and specimen &freezing. 25 1 - 3 3 Starvation. responses to. 308-309 Stress(es). environmental. responhcs to anoxia, 309 chemical reagents. 308 heat. 306-308 starvation. 308-309 wounding. 309-3 10 Stress protein(s) evolutionary significance of. 312-314 formation. molecular events in. 310-31 I function of. 3 I 1-3 I 2 Surface analyses. cryofixation and. 286-288

T Transcription. extranucleolar. 191- 194

U Ultrathin sectioning, cryofixation and. 289-290

IY4-196

transcribing chromatin. 190- I91 franspor! of RNA out of nucleuh. 197-198

W Wounding. responses to. 309-3 10

This Page Intentionally Left Blank

Contents of Recent Volumes and Supplements Volume 60

Biochemistry and Metabolism of Basement Membranes-NIcHoLAs A. KEFALIDES, Transfer RNA-like Structure in Viral GenoROBERTALPER,A N D CHARLES C. CLARK meS-TIMOTHY c. H A L L The Effects of Chemicals and Radiations Cytoplasmic and Cell Surface Deoxyribonuwithin the Cell: An Ultrastructural and cleic Acids with Consideraton of Their Micrurgical Study Using Amoeba profeus Origin-BEVAN L. R E I DA N D ALEXAN- as a Single-Cell Model-M. J. ORD DER J. CHARLSON Growth, Reproduction, and Differentiation Biochemistry of t h e Mitotic Spindlein A c a n t h u m o e b a - T ~ o ~ ~J.s BYERS CHRISTIAN Ps-rzmr SUBJECT I N D E X Alternatives to Classical Mitosis in Hemopoietic Tissues of Vertebrates-VieEKE E. ENGELEBERT Fluidity of Cell Membranes-Current Con- Volume 62 cepts and Trends-M. S H I N I T Z K Y A N D Calcification in Plants- ALLAN PENTECOST P. HENKART Cellular Microinjection by Cell Fusion: Macrophage-Lymphocyte Interactions in FELDMANN, Technique and Applications in Biology Immune Induction-MARC and Medicine-MMrTsuRu FURUSAWA ALANROSENTHAL, A N D PETERERB Immunohistochemistry of Luteinizing Hor- Cytology, Physiology, and Biochemistry of G e r m i n a t i o n of F e r n S p o r e s - V . mone-Releasing Hormone-Producing RAGHAVAN Neurons of the Vertebrates-JurrEN lmmunocytochemical Localization of the BARRY Vertebrate Cyclic Nonapeptide NeurohyCell Reparation of Non-DNA Injury-V. pophyseal H o r m o n e s and N e u r o p h y Y A . ALEXANDROV sins-K. DlERlCKX Ultrastructure of the Carotid Body in the Recent Progress in the Morphology, HistoMammals- ALAINVERNA chemistry, Biochemistry, and Physiology The Cytology and Cytochemistry of the of Developing and Maturing Mammalian WOO1 Fokle-DONALD F. G. O R W I N Testis-SARDUL s. GURAYA SUBJECT INDEX Transitional Cells of Hemopoietic Tissues: Origin, Structure, and Development Potential-JOSEPH M. YOFFEY Volume 61 Human Chromosomal Heteromorphisms: Nature and Clinical SignifiCanCe-RAM The Association of DNA and RNA with s. VERMA A N D HARVEYDOSlK Membranes-MARY PAT MOVER SUBJECT I N D E X Electron Cytochemical Stains Based on Metal Chelation-DAVID E. ALLENA N D DOUGLAS D. P E R R l N Cell Electrophoresis-THOMAS G . PRET- Volume 63 LOW,I1 A N D THERESA P. PRETLOW Physarum polycephalum: A Review of a The Wall of the Growing Plant Cell: Its Model System Using a Structure-FuncThree-Dimensional Organization-JEANtion Approach-EUGENE M. GOODMAN CLAUDEROLANDA N D BRIGITTEVIAN 321

322

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS

Microtubules in Cultured Cells: Indirect Im- Structural Aspects of Brain Barriers, with munofluorescent Staining with Tubulin Special Reference to the Permeability of S. FISTEL,J. Antibody-B. BRINKLEY, the Cerebral Endothelium and Choroidal M. MARCUM, A N D R. L. PARDUE Epithelium-B. V A N DEURS Septate and Scalariform Junctions in Ar- Immunochemistry of Cytoplasmic Contractthropods-CECILE NOIROT-TIMOTHEEile Proteins-UTE GROSCHEL-STEWART A N D CHARLES NOlROT The Ultrastructural Visualization of NucleoThe Cytology of Salivary Glands-CARLIN lar and Extranucleolar RNA Synthesis A. PINKSTAFF A N D E . Puand Distribution-S. FAKAN Development of the Vertebrate CorneaVION ELIZABETH D. HAY Cytological Mechanisms of Calcium CarScanning Electron Microscopy of the Pribonate Excavation by Boring Sponges SHIRLEY A. POMPONI mate Sperm-KENNETH G . GOULD Neuromuscular Disorders with Abnormal Cortical Granules of Mammalian EggsBELAJ. GULYAS Muscle Mitochondria-Z. KAMIENIECKA SUBJECT INDEX A N D H. SCHMALBRUCH SUBJECT INDEX

Volume 64

Volume 66

Variant Mitoses in Lower Eukaryotes: Indicators of the Evolution of Mitosis-I. BRENTHEATH The Centriolar Complex-Scorr P. PETERSON A N D MICHAELW. BERNS The Structural Organization of Mammalian Retinal Disc Membrane-J. OLIVE The Roles of Transport and Phosphorylation in Nutrient Uptake in Cultural Animal Cek-ROBERT M. WOHLHUETER A N D PETERG. W. PLAGEMANN The Contractile Apparatus of Smooth Muscle-J. VICTOR SMALL A N D APOLINARY

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

SOBIESZEK

Cytophysiology of the Adrenal Zona GlomerulOSa-GASTONE G . NUSSDORFER

Y ASOV

CELLSIM: Cell Cycle Simulation Made Easy-CHARLES E. DONAGHEY The Formation of Axonal Sprouts in Organ Culture and Their Relationship to Sprouting in Vivo-I. R. DUCEA N D P. KEEN Volume 65 When Sperm Meets Egg: Biochemical Mechanisms of Gamete lnteractionCell Surface Glycosyltransferase ActiviM. SHAPIRO A N D E. M. EDDY tkS-MICHAEL PIERCE, EVAA. TURLEY, BENNETT A N D STEPHEN ROTH Perisinusoided Stellate Cells (Fat-Sloring The Transport of Steroid Hormones into Cells, Interstitial Cells, Lipocytes), Their Animal Celk-ELEONORA P. GIoRGI Related Structure in and around the Liver

SUBJECT INDEX

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS Sinusoids, and Vitamin A-Storing Cells in Extrahepatic Organs- KENJIRO WAKE SUBJECT I N D E X

323

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

Volume 67

Membrane Circulation in Neurons and Photoreceptors: Some Unresolved IssuesA N D ARTHURM. MERERICHOLTZMAN

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

CURIO

Ultrastructure of Invertebrate Chemo-, Thermo-, and Hygroreceptors and Its Volume 69 Functional SignifiCanCe-HELMUT AL rThe Structures and Functions of the MycoN E R A N D LINDE PRlLLlNGER plasma Membrane--. B. ARCHER Calcium Transport System: A Comparative Metabolic Cooperation between Cells-M. Study in Different CellS-ANNE GODL. HOOPERA N D J. H. SUBAK-SHARPE FRAIND-DE B E C K E RA N D THEOPHILE The Kinetoplast as a Cell Organelle-V. D. GODFRAIND KALLI N I KOVA The Ultrastructure of Skeletogenesis in HerChloroplast DNA Replication in Chlumymatypic Corals-IAN s. JOHNSTON domonus reinhardtii-STEPHEN JAY Protein Turnover in Muscle Cells as VisualKELLERA N D CHlNG HO ized by Autoradiography-J. P. DANucleus-Associated Organelles in FungiDOUNE Identified Serotonin N e u r o n s - N ~ v r ~ ~ ~ I . BRENTHEATH Regulation of the Cell Cycle in Eukaryotic N. OSBORNE A N D VOLKERNEUHOFF CellS-ROSALIND M. YANISHEVSKY A N D Nuclear Proteins in Programming Cell CyGRETCHEN H. STEIN cles-M. v . N A R A S I M H A RAO The Relationship of in Virro Studies to in SUBJECT INDEX vivo H u m a n Aging-EDWARD L. SCHNIEDER A N D JAMESR. SMITH Cell Replacement in Epidermis (KeratoVolume 68 poiesis) via Discrete Units of Proliferation-C. S . POTTEN Moisture Content as a Controlling Factor in Seed Development and Germination-C. A. ADAMSA N D R. w. R l N N E Applications of Protoplasts to the Study of Plant CellS-LARRY c . F O W K EA N D O L U FL. GAMBORC Control of Membrane Morphogenesis in Bacteriophage-GREGORY J. BREWER Scanning Electron Microscopy of IntracelMar StruCtUreS-KEIICHI TANAKA The Relevance of the State of Growth and Transformation of Cells to Their Patterns of Metabolite Uptake-RUTH KOREN lntracellular Source of BioluminescenceBEATRICE M. SWEENEY

INDEX

Volume 70

Cycling 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 in CUkUre-MINA J. BISSELL On the Nature of Oncogenic Transformation Of CCIIS-GERALD L. CHAN

324

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS

Morphological and Biochemical Aspects of Adhesiveness and Dissociation of Cancer Ceh-HIDEO HAYASHI A N D Y A S U J I ISH I M ARU The Cells of the Gastric Mucosa-HERBERT F. HELANDER Ultrastructure and Biology of Female Gametophyte in Flowering Plants-R. N. KA PlL A N D A. K. BHATNAGAR INDEX

DNA Repair-A. R. L E H M A NANN D P. KARRAN Insulin Binding and Glucose TransportRUSSELL HILF, LAURIE K. SORGE, A N D ROGERJ. GAY Cell Interactions and the Control of Development in Myxobacteria PopulationsDAVID WHITE Ultrastructure, Chemistry, and Function of the Bacterial Wall-T. J. BEVERIDOE INDEX

Volume 71

Volume 73

Integration of Oncogenic Viruses in Mammalian Cells-CARL0 M. CRWE Mitochondria1Genetics of Paramecium aurelia-c. H. BEALEA N D A. TAlT Histone Gene Expression: Hybrid Cells and Organisms Establish Complex ControlsPHILIP HOHMANN Gene Expression and Cell Cycle RegulatiOn-sTEVEN J. HOCHHAUSER, JANET L. STEIN, A N D GARY s. STEIN The Diptera as a Model System in Cell and Molecular BiOlOgy-ELENA C. ZEGARELLI-SCHMIDT AND REBAGOODMAN Comments on the Use of Laser Doppler Techniques in Cell Electrophoresis: Reply to Pretlow and Pretlow's ReviewJOELH. KAPLAN A N D E. E. UZGIRIS Comments on the Use of Laser Doppler Techniques as Represented by Kaplan and Uzgiris: Reply t o Kaplan and UZginS-THOMAS G . PRETLOW 11 A N D THERESA P. PRETLOW

Protoplasts of Eukaryotic Algae-MP,RrHA D. BERLINER Polytene Chromosomes of Plants-WALTER NAGL Endosperm -I ts Morphology, Ultrastructure, and Histochemistry-S. P. I~H A TNAGAR A N D VEENA SAWHNEY The Role of Phosphorylated Dolichols in Membrane Glycoprotein Biosynthesis: Relation to Cholesterol BiosynthesisJ O A NTV G EN D H A FT M i u s A N D ANTHONY M. ADAMANY Mechanisms of lntralysosomal Degradation with Special Reference to Autophagocytosis and Heterophagocytosis of Cell OrgandeS-HANS GLAUMANN, JANL. E. ERICSSON, AND LOUIS MARZELLA Membrane Ultrastructure in Urinary Tubules-LELio ORCI, FABIENNE HUMBERT, DENNIS BROWN,A N D A L A I h ' P E R -

INDEX

RELET

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

Volume 72 Volume 74

Microtubule-Membrane Interactions in Cilia and F~age~~a-wILLIAML. DENTLER The Chloroplast Endoplasmic Reticulum: Structure, Function, and Evolutionary Significance-SARAH P. Grms

The Plasma Membrane as a Regulatory Site in Growth and Differentiation of Neuroblastoma Cells-SIEGFRIED w. DE LAAT A N D PAULT. V A N DER SAAG

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS

325

Mechanisms That Regulate the Structural Organization and Expression of Viral Genes in Adenovirus-Transformed Cells-S. J. and Functional Architecture of Cell SurfaCeS-JANET M. OLIVER A N D RICHARD FLINT Highly Repeated Sequences in Mammalian D. BERLIN Genomes-MAXIME F. SINGER Genome Activity and Gene Expression in Avian Erythroid Cek-KARLEN G . GA- Moderately Repetitive DNA in EvolutionSARYAN ROBERTA. BOUCHARD Morphological and Cytological Aspects of Structural Attributes of Membranous OrAlgal CakifiCatiOn-MICHAEL A. BOROganelles in BaCteria-cHARLES C. REMWITZKA

SEN

Naturally Occurring Neuron Death and Its Separated Anterior Pituitary Cells and Their Response to Hypophysiotropic HorRegulation by Developing Neural Pathways-TIMOTHY J. CUNNINGHAM mones-CARL DENEF,Luc SWENNEN, The Brown Fat Cell-JAN NEDERGAARD A N D MARIAANDRIES What Is the Role of Naturally Produced A N D OLOVLINDBERG Electric Current in Vertebrate RegeneraINDEX t i o n and Healing?-RlCHARD B. BORGENS Metabolism of Ethylene by Plants-JOHN Volume 75 DODDSA N D MICHAELA. HALL Mitochondria1 Nuclei-TsuNEYosHI K u - I N D E X ROlWA

Slime Mold LeCtinS-JAMES R. BARTLES, WILLIAMA. FRAZIER, A N D STEVEND. ROSEN Lectin-Resistant Cell Surface Variants of Eukaryotic Cells-EvE BARAK BRILES Cell Division: Key to Cellular Morphogenesis in the Fission Yeast, Schizosaccharomyces-BYRON F. JOHNSON, GODE B. CALLEIA, 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- A N N s. HENDERSON

Volume 77

Calcium-BindingProteins and the Molecular Basis of Calcium Action-LINDA J . VANELDIK, JOSEPHG. ZENDEGUI, DANI E L R. MARSHAK, A N D D. MARTIN WATTERSON

Genetic Predisposition to Cancer in Man: In Vitro Studies-LEVY KOPELOVICH Membrane Flow via the Golgi Apparatus of Higher Plant Celk-DAVID G . ROBINSON A N D UDO KRISTEN Cell Membranes in Sponges-WERNER E. G. MULLER Plant Movements in the Space Environment-DAVID G . HEATHCOTE Chloroplasts and Chloroplast DNA of Acetabularia mediterranea: Facts and Hypotheses-ANGELA L U T T K EA N D SILVANO BONOTTO Structure and Cytochemistry of the ChemiCal SynapSeS-STEPHEN MANALOVA N D WLADIMIR OVTSCHAROFF 1N DEX

326 Volume 78

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS

C h o n d r o c y t e s i n Aging Re se a rc hEDWARD J. MILLERA N D STEFFEN GAY Bioenergetics and Kinetics of Microtubule Growth and Differentiation of Isolated C al v arium C e l l s in a S e r u m - F r e e and Actin Filament Assembly- DisassemblY-TERRELL L. HILLA N D MARC w. Medium-JAMES K. BURKSA N D WILKIRSCHN E R L I A M A. PECK Regulation of the Cell Cycle by Somatome- Studies of Aging in Cultured Nervous Sysdins-HOWARD ROTHSTEIN tem Tissue-DONALD H. SILBERBERG A N D SEUNG U. KIM Epidermal Growth Factor: Mechanisms of Aging of Adrenocortical Cells in CultlureAction-MANJUSRI DAS MICHAELH. SIMONIPETERJ. HORNSBY, Recent Progress in the Structure, Origin, N. GILL AN,A N D GORDON Composition, and Function of Cortical Granules in Animal Egg-SARDUL S. Thyroid Cells in CUkUre-FRANCESCO S. GURAYA AMBESI-IMPIOMBATO A N D HAYDEN G. INDEX COON Permanent Teratocarcinoma-Derived Cell Lines Stabilized by Transformation with Supplement 10: Differentiated Cells in SV40 and SV40tsA Mutant VirusesAging Research W A R R EMNA L T Z M A ND,A N I E LI . H . BROWN,ANGELIKA LINZER,FLORENCE Do Diploid Fibroblasts in Culture Age?K. TERESKY, MAURICEROSENSTRAUS, EUGENE BELL,LOUISMAREK,STEPHA- A N D ARNOLDJ. LEVINE N I E S H E R ,C H A R L O T T M E E R R I L L , Nonreplicating Cultures of Frog Gastric TuAND IAN YOUNG DONALD LEVINSTONE, bular Ce1IS-GERTRUDE H. BLUMENUrinary Tract Epithelial Cells Cultured from THAL AND DINKAR K. KASBEKAR Human Urine-J. S. FELIXA N D J. W. SLIBJECT I N D E X LITTLEFIELD The Role of Terminal Differentiation in the Finite Culture Lifetime of the Human EpiG. dermal Keratinocyte-JAMES Supplement 11A: Perspectives in Plant Cell RHEINWALD and Tissue Culture Long-Term Lymphoid Cell CulturesGEORGEF. S M I T H P , ARVIN JUSTICE, HENRIFRISCHER, LEE KIN C HU,A N D Cell Proliferation and Growth in Callus Cultures-M. M. YEOMAN A N D E. FORCHE JAMES KROC Type I1 Alveolar Pneumonocytes in Virro- Cell Proliferation and Growth in Suspension Cultures-P. J. KING W I L L I A MH. J. DOUGLAS, JAMESA . MCATEER,JAMESR. SMITH,A N D WAL- Cytodifferentiation-RICHARD PHILLIPS Organogenesis in V i m : Structural, PhysioTER R. BRAUNSCHWEIGER logical, and Biochemical AspectsCultured Vascular Endothelial Cells as a TREVOR A. THORPE Model System for the Study of Cellular Senescence-ELLIOT M. LEVINEA N D Chromosomal Variation in Plant Tissues in M. MUELLER Culture-M. W. BAYLISS STEPHEN Vascular Smooth Muscle Cells for Studies Clonal Propagation-INDu K. VASILA N D VIMLA VASIL of Cellular Aging in Vitro; an Examination of Changes in Structural Cell Lip- Control of Morphogenesis by Inherent and ELAINE Exogenously Applied Factors in Thin Cell ids-oLGA 0 . BLUMENFELD, Layers-K. TRANTHANH VAN SCHWARTZ, VERONICA M. HEARN,A N D Androgenetic Haploids-INDRA K. VASIL MARIEJ. KRANEPOOL

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS Isolation, Characterization, and Utilization of Mutant Cell Lines in Higher PlantsPALMALIGA SUBJECT INDEX

327

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. FREED MAANN D Supplement 1lB: Perspectives in Plant Cell PHILIPC. LARIS and Tissue Culture Synthesis and Assembly of Membrane and Organelle PrOteinS-HARVEY F. LODISH, Isolation and Culture of PrOtOphStS-INW I L L I A MA . B R A E L L , A L A N L . DRA K. VASILA N D VIMLA VASIL SCHWARTZ, GER J. A. M. STROUS,A N D Protoplast Fusion and Somatic HybridizaASHER ZILBERSTEIN tiOn-oTTO SCHIEDER A N D I N D R A K. The Importance of Adequate Fixation in VASlI Preservation of Membrane UltrastrucGenetic Modification of Plant Cells Through tUR-RONALD B. LUFTIGA N D PAUL N. Uptake of Foreign DNA-C. 1. KADO MCMILLAN A N D A. KLEINHOFS Liposomes- As Artificial Organelles, ToNitrogen Fixation and Plant Tissue Culpochemical Matrices, and Therapeutic ture-KENNETH L. GILESAND INDRA K. Carrier SyStemS-hTER NICHOLLS VASIL Drug and Chemical Effects on Membrane Preservation of Germplasm-LYNDSEY A. Transport-WILLIAM 0. BERNDT WITHERS INDEX Intraovarian and in Vitro Pollination-M. ZENKTELER Endosperm Culture-B. M. JOHRI,P. S. Supplement 13: Biology of the Rhizobiaceae SRIVASTAVA, A N D A. P. RASTE The Formation of Secondary Metabolites in The Taxonomy of the RhizobiaceaePlant Tissue and Cell Cultures-H. GERALD H. ELKAN BOHM Biology of Agrobacterium tumefaciens: Embryo Culture-V. RACHAVAN Plant Interactions--. W. MOORE A N D The FUtUre-GEORC MELCHERS D. A. COOKSEY SUBJECT INDEX Agrobacterium tumefaciens in Agriculture and Research-FAwzI EL-FIKI A N D KENNETH L. GILES Suppression of, and Recovery from, the NeSupplement 12: Membrane Research: oplastic State-ROBERT TURGEON Classic Origins and Current Concepts Plasmid Studies in Crown Gall TumorigeneAND SiS-STEPHEN L. DELLAPORTA Membrane Events Associated with the GenRICKL. PESANO eration of a Blastocyst-MARTIN H. The Position of Agrobacterium rhizoJOHNSON genes-JESSE M. JAYNESA N D GARYA. Structural and Functional Evidence of STROBEL Cooperativity between Membranes and Cell Wall in Bacteria-MANFRED E. Recognition in Rhizobium -Legume Symbioses-TERRENCE L. GRAHAM BAYER Plant Cell Surface Structure and Recogni- The Rhizobium Bacteroid State-W. D. SUTTON,C. E. PANKHURST, AND A. S. tion Phenomena with Reference to SymbiCRAIG OSeS-PATRICIA s. REISERT

328

CONTENTS OF RECENT VOLUMES AND SUPPLEMENTS

Exchange of Metabolites and Energy be- Nodules Morphogenesis and DifferentiatiOIl-wILLIAM NEWCOMB tween Legume and Rhizobium-JOHN Mutants of Rhizobium That Are Altered in IMSANDE Legume Interaction and Nitrogen FixaThe Genetics of Rhizobium-ADAM tion-L. D. KUYKENDALL A N D ANDREWW. B. JOHNKONDOROSI The Significance and Application of IlhizoSTON bium in AgnCUltUre-HAROLD L. PETERIndigenous Plasmids of Rhizobium- J. DEE. LOYNACHAN E N A R I E EP., B O I S T AR D,F R A N C I N E SON AND THOMAS A. G. ATHERLY, J. 0. INDEX CASSE-DELBART, BERRY,A N D P. RUSSELL