chapter 16: cytoplasmic matrix and cytoskeleton

W. B. Saunders Company: West Washington Square Philadelphia, PA 19 105

Second Edition

THE CELL

DON W . FAWCETT. M.D. Hersey Professor of Anatomy

Harvard Medical School

W. B. SAUNDERS COMPANY Philadelphia London Toronto Mexico City Rio de Janeiro Sydney Tokyo

1 St. Anne's Road Eastbourne, East Sussex BN21 3UN, England

1 Goldthorne Avenue Toronto, Ontario M8Z 5T9, Canada

Apartado 26370 - Cedro 5 12 Mexico 4. D.F.. Mexico

Rua Coronel Cabrita, 8 Sao Cristovao Caixa Postal 21 176 Rio de Janeiro, Brazil

9 Waltham Street Artarmon, N. S. W. 2064, Australia

Ichibancho, Central Bldg., 22-1 Ichibancho Chiyoda-Ku, Tokyo 102, Japan

Library of Congress Cataloging in Publication Data

Fawcett, Don Wayne, 1917-

The cell.

Edition of 1966 published under title: An atlas of fine structure.

Includes bibliographical references. 1. Cytology -Atlases. 2. Ultrastructure (Biology)- Atlases. I. Title. [DNLM: 1. Cells- Ultrastructure- Atlases. 2. Cells- Physiology - Atlases. QH582 F278c]

QH582.F38 1981 591.8'7 80-50297

ISBN 0-7216-3584-9

Listed here is the latest translated edition of this book together with the language of the translation and the publisher.

German (1st Edition)-Urban and Schwarzenberg, Munich, Germany

The Cell ISBN 0-7216-3584-9

© 1981 by W. B. Saunders Company. Copyright 1966 by W. B. Saunders Company. Copyright under the Uniform Copyright Convention. Simultaneously published in Canada. All rights reserved. This book is protected by copyright. No part of it may be reproduced, stored in a retrieval system, or trans- mitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without written permission from the publisher. Made in the United States of America. Press of W. B. Saunders Company. Library of Congress catalog card number 80-50297.

Last digit is the print number: 9 8 7 6 5 4 3 2

cboeke

Line

CONTRIBUTORS OF iv CONTRIBUTORS OF PHOTOMICROGRAPHS

ELECTRON MICROGRAPHS

Dr. John Albright Dr. David Albertini Dr. Nancy Alexander Dr. Winston Anderson Dr. Jacques Auber Dr. Baccio Baccetti Dr. Michael Barrett Dr. Dorothy Bainton Dr. David Begg Dr. Olaf Behnke Dr. Michael Berns Dr. Lester Binder Dr. K. Blinzinger Dr. Gunter Blobel Dr. Robert Bolender Dr. Aiden Breathnach Dr. Susan Brown Dr. Ruth Bulger Dr. Breck Byers Dr. Hektor Chemes Dr. Kent Christensen Dr. Eugene Copeland Dr. Romano Dallai Dr. Jacob Davidowitz Dr. Walter Davis Dr. Igor Dawid Dr. Martin Dym Dr. Edward Eddy Dr. Peter Elias Dr. A. C. Faberge Dr. Dariush Fahimi Dr. Wolf Fahrenbach

Dr. Marilyn Farquhar Dr. Don Fawcett Dr. Richard Folliot Dr. Michael Forbes Dr. Werner Franke Dr. Daniel Friend Dr. Keigi Fujiwara Dr. Penelope Gaddum-Rosse Dr. Joseph Gall Dr. Lawrence Gerace Dr. Ian Gibbon Dr. Norton Gilula Dr. Jean Gouranton Dr. Kiyoshi Hama Dr. Joseph Harb Dr. Etienne de Harven Dr. Elizabeth Hay Dr. Paul Heidger Dr. Arthur Hertig Dr. Marian Hicks Dr. Dixon Hingson Dr. Anita Hoffer Dr. Bessie Huang Dr. Barbara Hull Dr. Richard Hynes Dr. Atsuchi Ichikawa Dr. Susumu It0 Dr. Roy Jones Dr. Arvi Kahri Dr. Vitauts Kalnins Dr. Marvin Kalt Dr. Taku Kanaseki

Dr. Shuichi Karasaki Dr. Morris Karnovsky Dr. Richard Kessel Dr. Toichiro Kuwabara Dr. Ulrich Laemmli Dr. Nancy Lane Dr. Elias Lazarides Dr. Gordon Leedale Dr. Arthur Like Dr. Richard Linck Dr. John Long Dr. Linda Malick Dr. William Massover Dr. A. Gideon Matoltsy Dr. Scott McNutt Dr. Oscar Miller Dr. Mark Mooseker Dr. Enrico Mugnaini Dr. Toichiro Nagano Dr. Marian Neutra Dr. Eldon Newcomb Dr. Ada Olins Dr. Gary Olson Dr. Jan Orenstein Dr. George Palade Dr. Sanford Palay Dr. James Paulson Dr. Lee Peachey Dr. David Phillips Dr. Dorothy Pitelka Dr. Thomas Pollard Dr. Keith Porter

. . . 111

Dr. Jeffrey Pudney Dr. Eli0 Raviola Dr. Giuseppina Raviola Dr. Janardan Reddy Dr. Thomas Reese Dr. Jean Revel Dr. Hans Ris Dr. Joel Rosenbaum Dr. Evans Roth Dr. Thomas Roth Dr. Kogaku Saito Dr. Peter Satir

Dr. Manfred Schliwa Dr. Nicholas Severs Dr. Emma Shelton Dr. Nicholai Simionescu Dr. David Smith Dr. Andrew Somlyo Dr. Sergei Sorokin Dr. Robert Specian Dr. Andrew Staehelin Dr. Fumi Suzuki Dr. Hewson Swift Dr. George Szabo

Dr. John Tersakis Dr. Guy de Th6 Dr. Lewis Tilney Dr. Greta Tyson Dr. Wayne Vogl Dr. Fred Warner Dr. Melvyn Weinstock Dr. Richard Wood Dr. Raymond Wuerker Dr. Eichi Yamada

PREFACE

PREFACE

The history of morphological science is in large measure a chronicle of the discovery of new preparative techniques and the development of more powerful optical instruments. In the middle of the 19th century, improvements in the correction of lenses for the light microscope and the introduction of aniline dyes for selective staining of tissue components ushered in a period of rapid discovery that laid the founda- tions of modern histology and histopathology. The decade around the turn of this century was a golden period in the history of microscopic anatomy, with the leading laboratories using a great variety of fixatives and combinations of dyes to produce histological preparations of exceptional quality. The literature of that period abounds in classical descriptions of tissue structure illustrated by exquisite lithographs. In the decades that followed, the tempo of discovery with the light microscope slackened; interest in innovation in microtechnique declined, and specimen preparation narrowed to a monotonous routine of paraffin sections stained with hematoxylin and eosin.

In the middle of the 20th century, the introduction of the electron microscope suddenly provided access to a vast area of biological structure that had previously been beyond the reach of the compound microscope. Entirely new methods of specimen preparation were required to exploit the resolving power of this new instrument. Once again improvement of fixation, staining, and microtomy commanded the attention of the leading laboratories. Study of the substructure of cells was eagerly pursued with the same excitement and anticipation that attend the geographical exploration of a new continent. Every organ examined yielded a rich reward of new structural information. Unfamiliar cell organelles and inclusions and new macromolecular components of protoplasm were rapidly described and their function almost as quickly established. This bountiful harvest of new structural information brought about an unprecedented convergence of the interests of morphologists, physiologists, and biochemists; this convergence has culminated in the unified new field of science called cell biology.

The first edition of this book (1966) appeared in a period of generous support of science, when scores of laboratories were acquiring electron microscopes and hundreds of investigators were eagerly turning to this instrument to extend their research to the subcellular level. At that time, an extensive text in this rapidly advancing field would have been premature, but there did seem to be a need for an atlas of the ultrastructure of cells to establish acceptable technical standards of electron microscopy and to define and illustrate the cell organelles in a manner that would help novices in the field to interpret their own micrographs. There is reason to believe that the first edition of The Cell: An Atlas of Fine Structure fulfilled this limited objective.

In the 14 years since its publication, dramatic progress has been made in both the morphological and functional aspects of cell biology. The scanning electron microscope and the freeze-fracturing technique have been added to the armamentarium of the miscroscopist, and it seems timely to update the book to incorporate examples of the application of these newer methods, and to correct earlier interpretations that have not withstood the test of time. The text has been completely rewritten and considerably expanded. Drawings and diagrams have been added as text figures. A few of the original transmission electron micrographs to which I have a sentimental attachment have been retained, but the great majority of the micrographs in this edition are new. These changes have inevitably added considerably to the length of the book and therefore to its price, but I hope these will be offset to some extent by its greater informa- tional content.

Twenty years ago, the electron microscope was a solo instrument played by a few virtuosos. Now it is but one among many valuable research tools, and it is most profit-

v

ably used in combination with biochemical, biophysical, and immunocytochemical techniques. Its use has become routine and one begins to detect a decline in the number and quality of published micrographs as other analytical methods increasingly capture the interest of investigators. Although purely descriptive electron microscopic studies now yield diminishing returns, a detailed knowledge of the structural organization of cells continues to be an indispensable foundation for research on cell biology. In under- taking this second edition I have been motivated by a desire to assemble and make easily accessible to students and teachers some of the best of the many informative and aesthetically pleasing transmission and scanning electron micrographs that form the basis of our present understanding of cell structure.

The historical approach employed in the text may not be welcomed by all. In the competitive arena of biological research today investigators tend to be interested only in the current state of knowledge and care little about the steps by which we have arrived at our present position. But to those of us who for the past 25 years have been privileged to participate in one of the most exciting and fruitful periods in the long history of morphology, the young seem to be entering the theater in the middle of an absorbing motion picture without knowing what has gone before. Therefore, in the introduction to each organelle, I have tried to identify, in temporal sequence, a few of the major contributors to our present understanding of its structure and function. In venturing to do this I am cognizant of the hazards inherent in making judgments of priority and significance while many of the dramatis personae are still living. My apologies to any who may feel that their work has not received appropriate recognition.

It is my hope that for students and young investigators entering the field, this book will provide a useful introduction to the architecture of cells and for teachers of cell biology a guide to the literature and a convenient source of illustrative material. The sectional bibliographies include references to many reviews and research papers that are not cited in the text. It is believed that these will prove useful to those readers who wish to go into the subject more deeply.

The omission of magnifications for each of the micrographs will no doubt draw some criticism. Their inclusion was impractical since the original negatives often remained in the hands of the contributing microscopists and micrographs submitted were cropped or copies enlarged to achieve pleasing composition and to focus the reader's attention upon the particular organelle under discussion. Absence was considered preferable to inaccuracy in stated magnification. The majority of readers, I believe, will be interested in form rather than measurement and will not miss this datum.

Assembling these micrographs illustrating the remarkable order and functional design in the structure of cells has been a satisfying experience. I am indebted to more than a hundred cell biologists in this country and abroad who have generously re- sponded to my requests for exceptional micrographs. It is a source of pride that nearly half of the contributors were students, fellows or colleagues in the Department of Anatomy at Harvard Medical School at some time in the past 20 years. I am grateful for their stimulation and for their generosity in sharing prints and negatives. It is a pleasure to express my appreciation for the forbearance of my wife who has had to communicate with me through the door of the darkroom for much of the year while I printed the several hundred micrographs; and for the patience of Helen Deacon who has typed and retyped the manuscript; for the skill of Peter Ley, who has made many copy negatives to gain contrast with minimal loss of detail; and for the artistry of Sylvia Collard Keene whose drawings embellish the text. Special thanks go to Elio and Giuseppina Raviola who read the manuscript and offered many constructive suggestions; and to Albert Meier and the editorial and production staff of the W. B. Saunders Company, the publishers.

And finally I express my gratitude to the Simon Guggenheim Foundation whose commendable policy of encouraging the creativity of the young was relaxed to support my efforts during the later stages of preparation of this work.

DON W. FAWCETT Boston, Massachusetts

CONTENTS CONTENTS

................................................................................... CELL SURFACE 1

........................................................................................ Cell Membrane 1 ....................................................................... Glycocalyx or Surface Coat 35

Basal Lamina .......................................................................................... 45

SPECIALIZATIONS O F THE FREE SURFACE .................................... 65

...................................................... Specializations for Surface Amplification 68

...................................................... Relatively Stable Surface Specializations 80 ....................................................... Specializations Involved in Endocytosis 92

JUNCTIONAL SPECIALIZATIONS ...................................................... 124

.............................................................. Tight Junction (Zonula Occludens) 128 .......................................................... Adhering Junction (Zonula Adherens) 129

................................................................................ Sertoli Cell Junctions 136 Zonula Continua and Septate Junctions of Invertebrates ................................. 148

........................................................................................... Desmosomes 156 ........................................................................... Gap Junctions (Nexuses) 169

................................ Intercalated Discs and Gap Junctions of Cardiac Muscle 187

NUCLEUS ............................................................................................ 195

............................................................................ Nuclear Size and Shape 197 ............................................................................................... Chromatin 204

............................................................................... Mitotic Chromosomes 226 Nucleolus ............................................................................................... 243

.................................................................................. Nucleolar Envelope 266 ................................................................................... Annulate Lamellae 292

............................................................. ENDOPLASMIC RETICULUM 303

................................................................... Rough Endoplasmic Reticulum 303 ................................................................. Smooth Endoplasmic Reticulum 330

Sarcoplasmic Reticulum ............................................................................ 353

GOLGI APPARATUS ............................................................................ 369

..................................................................................... Role in Secretion 372 ......................................... Role in Carbohydrate and Glycoprotein Synthesis 376

............................................................ Contributions to the Cell Membrane 406 vii

................................................................................. MITOCHONDRIA 410

.......................................................................... Structure of Mitochondria 414 ...................................................................................... Matrix Granules 420

................................................................... Mitochondria1 DNA and RNA 424 ........................................................................... Division of Mitochondria 430

............................................................................. Fusion of Mitochondria 438 .................................................................. Variations in Internal Structure 442

........................................................................... Mitochondria1 Inclusions 464 ......................................................................... Numbers and Distribution 468

LYSOSOMES ......................................................................................... 487

............................................................................... Multivesicular Bodies 510

..................................................................................... PEROXISOMES 515

.................................................................... LIPOCHROME PIGMENT 529

MELANIN PIGMENT ........................................................................... 537

....................................................................................... CENTRIOLES 551

Centriolar Adjunct ................................................................................... 568

CILIA AND FLAGELLA ...................................................................... 575

Matrix Components of Cilia ....................................................................... 588 .............................................................................. Aberrant Solitary Cilia 594

.......................................................................................... Modified Cilia 596 ............................................................................................... Stereocilia 598

.......................................................................... SPERM FLAGELLUM 604

..................................................................... Mammalian Sperm Flagellum 604 .......................................................................... Urodele Sperm Flagellum 619

............................................................................. Insect Sperm Flagellum 624

CYTOPLASMIC INCLUSIONS ............................................................. 641

................................................................................................ Glycogen 641 Lipid ...................................................................................................... 655

............................................................................... Crystalline Inclusions 668 ................................................................................... Secretory Products 691

................................................................................................ Synapses 722

CYTOPLASMIC MATRIX AND CYTOSKELETON .............................. 743

Microtubules ........................................................................................... 743 Cytoplasmic Filaments .............................................................................. 784

cindyboeke

Highlight

CYTOPLASMICMATRIX AND

CYTOSKELETON

MICROTUBULES

Microtubules , as now defined , have a relatively short history in cytology, bu t, asmember s of a bro ader category of filamentous components of the cytopl asm , the y werecerta inly observed by man y ofthe early cytologi st s. The spindle fibers and astral rays ofdividing cells were extensivel y studied ; longitudinal filamentous struc tures wereob served in nerve ax ons (F reud , 1882) and in th e ax onemes of spe rm flagell a(Ballowitz, 1890); circumferential band s of filaments were de scribed in amphibianerythrocytes (Meves, 1910). Although the components of the se struc tures are nowknown to be microtubules, the y could be identified onl y as " filaments" with the lightmicroscope . E ven after the introduction of the electron micro scope, the tubular natureof the core structures of cilia and sperm flagella escaped detection in the dissoci ated ,air-dr ied preparations first examined (Harvey and Anderson , 1948; Manton , 1952). Itwas not until it bec ame possible to cut ultrathin transverse sections that the hollo wnatu re of the axonemal fibers was suggested (Fawcett and Porter , 1954). Their crosssectional image s showed a den se rim or wall and a light center, but there was a curiousreluctance to describe them as tubules. It was suspected that incomplete penetrat ion ofa fibril by osmium might res ult in a dense per iphery and a lighter cen ter. Moreover, theaxonemal " fibrils" were con sidered to be the ac tive agents of flagellar motion and itwas difficult to conce ive of a contractile tubule . For several years the y continued to bede scribed as fibrils.

With the dev elopment of satisfactory procedures for embedding tissue and cuttingthin sections thi s became the principal method of biological electron microscopy . Anancillary technique ca lled negative staining was soon introduced (Huxley, 1956; Horneand Brenner, 1958), which invol ved dr ying down of disrupted cells on the specimen gridtogether with a solution of sodium pho sphotungstate . The electron den sity of the metalthen served as a co ntras t med ium for the less de nse cell components and revea led, innegati ve image , details that were otherwise invisible. This stimulated a reviva l ofintere st in the examination of fragmented and dispersed cell components . Observationof the fray ed-out tips of the axonemes of sperm flagella by thi s method showed that thephosphotungstate penetrated into the interior of the axonema l " fibrils" and thi seliminated all lingering doubt about the ir hollo w natu re (Pease, 1963; Andre and Thiery,1963). They appea red to be cylinders whose walls con sisted of about 11 longitudinallyoriented beaded filaments 3.5 nm in diam eter with a repeating ax ial periodicity of about8 nm .

Microtubu les in the cytopl asm of cells were onl y rarely preserved by theosmium-co ntaining fixat ives then in gene ral use . Among the first to be described were

743

744 CYTOPLASMIC MATRIX AND CYTOSKELETO N CYTO PLASMIC MATRIX AND CYTOSKELETON 745

Tubulin dimers of protofilaments

Sch emati c depi cti on of the linear polym er s of tu bulin that make up the pro to filame nts in the wa llof a micr otubul e. Show n be low are three configura tions in which the prot ofilaments may be assembled:indi vidu al cytoplas mic microtu bule s, doubl et s in c iliary and flagellar axone mes, and triplets in thewa ll of ce nt rioles.

work , on ass embly of microtubules from the supernata nt of porcine brain homogenates,it was reported that small disc-like structures of the same diameter as microtubuleswere need ed as nucleation sites . When these were removed by high-speed centrifugation , little polymerization took place (Bori sy and Olmsted , 1972). The que stion remai~s

unsettled , but it is likel y that a nucle ation site or microtubule-organizing center IS

Centriolar tripletCiliarydoublet

Cytoplasmicmicrotubule

~...,v..

>, ,~ ,l>

. . '

: ' ~ " "; ' ~~ ... ,. ....

"x

necessary in vivo for form ation of organized arrays of microtubules such as the mitoticspindle, but it is apparent that the organizing center need not be a centriole.

When tubulin is purified from brain , the preparation contains small amounts (15 percent) of at least two other proteins of high molecular weight (300,000 to 350,000daltons). These microtubule-associated proteins (MAPs) pe rsist through repeatedcycle s of in vitro microtubule disassembly and reassembly (Murphy and Bori sy, 1975;Vallee and Borisy , 1978). Addition of the se proteins to solutions of tubulin devoid ofMAPs promotes microtubule assembly (Sloboda et al., 1976). Fine filamentousprojections from microtubules that were observed earlier in electron micrographs ofperipheral nerve axons and other cell typ es are now believed to correspond to theMAPs that co purify with tubulin. The presence of the se proteins uniformly distributedalong the length of microtubules in situ can be demonstrated immunocytochemically(Connolly et al., 1978). Microtubules ass embled in vitro in the absence of MAPs appea rsmooth surfaced, but when the se proteins are present, the microtubules are deco rat edwith very slender late ral projections (Binder and Rosenbaum , 1978). A 300,000 daltonmicrotubule-associ ated protein (MAP2) isolated from neurotubules promotes in vitroass embly, binding in a constant ratio of I mole of MAP to 9 moles of tubulin dimer, andforms on the re sulting microtubules very slender lateral projections exhibiting a 32 nmaxial periodicity (Kim , Bind er and Rosenbaum, 1979).

The functional significance of the se filamentou s protein s is not well understood ,but it is speculated that where microtubules are laterally ass ociated in organized arrays ,as in the spindle, they may form stabilizing bridges between them (Hepler, McIntoshand Cleland, 1970). They may also be involved in attachment of organelles or othercytoplasmic particles that are translocated along bundles of microtubules (Smith et al. ,1975 ; Heggeness et aI., 1978).

those comprising the manchette of mammalian spermatids (Burgos and Fawcett, 1955),the marginal band of teleost and amphibian erythrocytes (Fawcett and Witebsky , 1959),those forming the centrioles and mitotic spindles (de Harven and Bernhard , 1956; Rothand Daniels, 1962; Harris , 1962), and those in the axons of neurons (Palay , 1960). Theterm microtubule wa s proposed in a paper de scribing those associated with thedeveloping nematocysts of hydra (Slautterback , 1963). It was not until the introductionof glutaraldehyde as a fixative (Sabatini et al., 1963) that microtubules were routinelypreserved and came to be recognized as a nearl y universal constitutent of the cellcytoplasm . They were then observed in the cortex of plant cells , and subunits wereseen in the ir walls in thin transverse sections (Ledbetter and Porter, 1963) at about thesame time that their protofilaments were ob served by negative staining.

It soon became apparent that microtubules are highly versatile organelles involvedin maintenance and change of cell shape , chromosomal movements in mitosis, and indirectional guidance or ac tive translocation of particulate components of the cytoplasm(Porter, 1966). They rapidly bec ame a subject of intense investigative interest.

Selective solubilization of ciliary and flagellar components provided the first methodological approach to isolation of microtubules for investigation of their protein(Gibbons , 1963, 1965). A subunit with a molecular weight of 60,000 daltons wasdescribed which appeared to correspond to the 4 to 5 nm particles responsible for thebeaded appearance of protofilaments in negat ively stained preparations (Grim stone andKlug , 1966). Characterization of the more labile cytoplasmic microtubules was madepossible by taking advantage of the discovery that the antimitotic drug colchicine bindsto the subunits and prevents their polymerization to form spindle microtubules (Ta ylor,1965). A procedure was then worked out for purific ation of microtubule subunits frombrain on the basis of their colchicine-binding activity (Weisenberg, Borisy and Taylor,1968). This became the ba sis for man y of the subsequent chemical studies . Themicrotubule protein wa s named tubulin (Mohri, 1968), and whether it was extractedfrom flagella, brain , or spindles, the isolated native protein had a molecular weight ofabout 110,000 daltons . It proved to be a dimer, and the 55,000 dalton molecular weightmonomer wa s found to occur in two form s, IX and f3 tubulin. The dimers probablycon sist of one molecule each of the IX and f3 form (Bryan and Wilson, 1971) and the seare arranged in tandem along the length of the protofil aments. The possible occurrenceof homodimers and of other arrangements within the wall of the tubule has not beenruled out.

Low temperature , increased pressure , and various antimitotic alkaloids were foundto res ult in disappearance of microtubules or to prevent their form ation . From theseobservations has eme rged the concept that polymerized microtubules in the cytoplasmare in a state of equilibrium with their monomeric subunits (Inoue, 1952; Inoue andSato, 1967). The factors that regul ate the equilibrium to cause microtubule polymerization or depolymerization in different phy siological sta tes of the cell remain largel yunknown . The polymerization of the microtubules of cilia and flagell a can be prevented ,but unlike microtubules of the general cytoplasm , once formed , they are con sidered tobe stabilized by some mechanism yet to be elucidated .

Morphological studies have sugge sted that ordered arrays of microtubules oftenarise in close relationship to centrioles or associ ated small den se bodies,'and it has beenspeculated that these serve as sites of nucleation or initiation of microtubule assembly .However, it remains unclear whether such initiat ion sites are necessary for formation ofthe microtubules that are randomly oriented in the ground cytoplasm . To shed furtherlight on thi s question numerous studies on in vitro ass embly of microtubules have beenundertaken. Cytoplasmic microtubules from brain have been solubilized and there sulting preparations of tubulin have been extensively used for study of in vitroassembly of microtubules (Bori sy and Olmsted , 1972; Weisenberg, 1972).

Partially purified subunits derived from rat brain polymerize in vitro into microtubules nearly indi st ingui shable from those seen in cell s in vivo . Polymerizationproceeded optimally at 35 to 37° C, required the pre sence of guanosine triphosphate andmagne sium ions, and seemingly did not require nucleation (Wei senberg, 1972). In other

746 CYTOPLASMIC MATRIX AND CYTOSKELETON

In the upper figure on the opposite page is a cytoplasmic microtubule in crosssection showing the usual 13 protofilaments. The lower figure is a preparation ofmicrotubules assembled in vitro from flagellar tubulin. The inset presents one of theseat high magnification show ing 15 protofilaments. No ex plana tion is available for theincrease of one or two protofilaments upon ass embly in vitro. A similar phenomenonhas been reported for in vitro assembly of brain tubulin (P ierson et aI., 1978 ). Fl agellarmicrotubules differ from cy toplas mic microtubules in their grea ter resistan ce to solubilization by high pre ssure , low temperatu re , or drugs such as colchicine, which bind totubulin. It is possible , however, to obtain soluble tubulin subunits from flagellar doublet s by extensive sonication (Kuriya ma, 1976). Under suitable conditions , thi s flagellartubulin will assemble in vitro into single microtubules but not into doublets. The ir wallscontain 14 or 15 protofilaments instead of the 13 characteristic of microtubules in vivo .

Figure 399 . Cytoplasmic micro tubul e with the usu al 13 prot ofilamenrs . (Micrograph courte sy ofVitaurs Kalnins.)

Figure 400. Microtubules assembled in vitro from flagellar ru bulin, (From Binder and Rosen baum,].Ce ll BioI. 79: 500- 515, 1978.)

Figure 399, upper Figure 400, lower

747


The accompanying micrograph s permit a compari son of the smooth-walled appea rance of microtubules assembled in the absence of MAPs (top figure) with the fuzzyappea rance of microtubules assembled in their pre sence (middle figure) . In the lowerfigures, the radiating filaments visible in the figure on the right are evidently res ponsiblefor a more uniform spac ing of the microtubules than that seen on the left with microtubul es lack ing MAPs.

Figure 40 1. Microtubules assembled from flagellar tubulin in the abse nce of MAPs.

Figure 402. Microtubul es assembled in the prese nce of MAPs.

Figure 403. A . Cross sec tions of MAP-free microrubules. B. Cross sect ions of MAP- decorated micro tubules. (Figures from Binder and Rosenbaum, J. Cell BioI. 79:500- 515, 1978.)

Figu re 40 1, upper Figure 402, center Figu re 403, lower

749

Microtubules in Particle Movement

Cytoplas mic st rea ming in animal cells is less appare nt than in plant s in which it hasbeen ex tens ive ly studied, particul arl y in the green algae . Nevertheless time-l apsecinematograph y of mammalian cell s in culture has clearly establi shed that a variety ofparticul ate structures in the cytoplasm exhibit line ar di spl acement s along pathsarranged radially with respect to the cell cente r. Examination of the same cells inelectron micrograph s reveals microtubules deplo yed in a radiating pattern centered onthe centrosome. Directed salta tory movements of particles cease afte r treatment ofce lls with drugs that depol ymeri ze microtubules (F reed and Lebo witz, 1970). Perhapsthe most dramatic of the numerous examples of the ass ociation of particle migrationwith microtubules is seen in the melanophores and erythrophores of fish. The se cellshave an extraord inary array of microtubules radiating from the ce ll center. Whenst imulated to pale , the pigment granules rapid ly move to the cell center and ondisaggregation move slow ly outward in a series of salta tory movements. Microtubulespersist du ring aggregation but in reduced number and are re stored during disper sion(Bikle, Tilney and Porter, 1966; Junquiera , Raker and Porter, 1974; Schli wa, Osbornand Weber, 1978). In the se and other examples, there is no doubt that microtubulesdefine path ways along which particles move , but whether the y pro vide a moti ve forceremain s uncertain. There is a possibility that actin filaments associated with themicrotubule are invo lved , but a force-generating interaction between the two has yet tobe demonstrated .

The form of transport that has been the subject of most investigation is that whichoccurs in the pro cesses of nerve ce lls. Two classes of transport are generally accepted,a slow axoplas mic flow at a rate of 1 to 2 mm/day and faste r saltatory movement ofparticles at up to 400 mm/day. Of these only the faster component is affected bymicrotubule disassembly. Close ass ociation between syna ptic vesicle s and microtubules has been reported in axons of some vertebrate species (Smith , Jarlfors andBeranek , 1970; Gray, 1975) and short cro ss-bridges linking the two structures havebeen described (Smith, 1971). It has been establi shed that the close topographicalrelationship between mitochondria and microtubules in cross sections of insect axons issta tistically significant, and linkages were also found between the two co mparable totho se pre viou sly described for the synaptic ves icles (Smith, Jarlfors and Cayer, 1977).Very close associ ation of mitochondria with microtubules has also been demonstratedin several unrelated type s of cultured vertebrate cell s by indirect immunofluorescencetec hniques, which permit their simultaneous visualization . No such ass ociation withactin microfilaments was found . It was sugge sted therefore that microtubules may beinvolved in movement and intracellular distribution of mitochondria (Heggeness ,Simon and Singer, 1978). While there is general agreeme nt that microtubules det erminethe direction along which cytoplasmic strea ming preferentially occurs, no satisfac torymechan ism has been put forward to explain how the y might contribute to the generationof force for translocation of membrane-limited particle s.

751


The interference contrast images at the top of the facing page illustrate the dramatic transformation in a living fish chromatophore from the dispersed (A) to theaggregated state (B) of its pigment granules. This occurs with no significant change inthe outline of the cell (indicated by the arrows). Immunofluorescence micro scopy withantitubulin antibody, in the middle figure , shows the striking array of radially disposedmicro tubules in a chromatophore with dispersed pigment. In the cell with aggregatedpigment in the lower figure , it is evident that the radial dispositon of microtubulesremains unchanged .

The rapid movement of pigment granules along oriented tracts of microtubules hasmade chromatophore s a favo red system for studying the role of microtubules in thetranslocation of particulate components of the cytoplasm. There is no doubt that theyare involved in pigment migration, because their depolymerization with colchicine orlow temperature reduces or abolishes pigment movement, but the exact role microtubules play rem ains uncertain. Although radially oriented microtubules persist in bothstates , pigment aggreg ation is associated with a reduction of about 50 per cent in theirnumbers, suggesting that microtubule assembly and disassembly may be invol ved.

Figure 404 . A . Chro matophore wit h dispersed pigme nt. B . Th e same cell with pigme nt aggregated.

Figure 405. Indirect immunofluor escence of a me lanopho re with dispe rsed pigment.

Fi gure 406. Immunofluorescence of chrom atophore with aggr egat ed pigment. (Figures fro m Schliwa,Osborn and We be r,]. Cell Bio!. 76:229- 23 6, 1978.)

Figure 404, upper Figure 405, center Figure 406, lower

753


Neurons must maintain synaptic terminals which are often loc ated at greatdistan ces from the cell body. Streaming of the cytoplasm , which is particularlycon spicuous both in the perikaryon and along the leng th of the axon, is probablyessential to solve the logistic problems inherent in an extensive sys tem of axonalbranches . A slow bu lk flow of ax oplas m and a mo re ra pid transport of specificsubsta nces are well documented .

In the large dendri te in the acc ompanying micrograph , numerous cro ss-sectionalprofiles of longi tudinally oriented microtubules are uniformly distributed throughout.These are believed to play an essential role in tran sport both in dendrite s andaxons - a conclusion supporte d by the observation that colchicine and other agentsthat dep olymerize microtubules also inhibit axoplas mic translocation. As in other form sof cy toplas mic movem ent, it is not known whether the micro tubules merel y serve astrack s determin ing the direction of movement or whet her they inte ract with some otherfilamentou s component to supply moti ve force.

Also pre sent in the cy toplas m are loose aggregations of 10 nm neurofil aments.These are ass igned no role in transport but are thought to have a supportivecyto skeletal function.

Figure 407 . Proximal de ndrite of an anterior ho rn ce ll. (Micro graph co urt esy of Raymo nd Wu erker. )

•

Figure 407

~ .....~..•.. .....

". 1, ~

." tit;;-~'<I ... ,

"'""

.' ,

.~

; , .. ~".~ ~'O . ~

~

0

""'. ~

~- "'I

(l;

&.

O.

0.

Q

c

o..... .. "

ro.

~ ~ •• ,! " 0" .." .

.....w

o

o

755


It is generally assumed that synaptic vesicles are restricted to nerve terminals andthat synthesis and sequestration of transmitter in their interior occur locally . However, there are observations sugges ting that vesicles and their contents may ari se centrallyand be transported along microtubules to the nerve terminals. Isolated or sparselydistributed vesicle s are also observed at non synapti c sites, and the striking alignment ofves icles along microtubules illustrated in the upper figure on the facing page stronglysuggests that the y are in tr ansit. Schmitt (1968) has speculated that a motive force maybe generated at the site of contact bet ween a micro tubule and a particle which might beresponsible for the observed movement of particles along the axon. No experimentalconfirmation of thi s has been forth coming , but in the materi al illustrated here , smallbridges , or link s , have been demonstrated between the ve sic les and the microtubul es.

Microtubules are not especially abundant at synapses, but the lower figure show svesicles closel y clu stered around a few microtubuJes at the periphery of the synapticarea .

Figures 408 and 409. G iant axons in the ce nt ral ner vous system of the lamp rey tPet romyzon marinus ).(From Smit h, J iirlfors and Beranek. ). Cell Bioi. 46: 199-2 19, 1970. )


757

758

Microtubules in Mitosis

As traditionally de fined by light micro scopi sts, the mitotic apparatus included aspindle composed of fiber s con verging upon a pair of centrioles at both pole s, andas ters made up of fibers radiating from the centrioles. The spindle included chromosomal fib ers from the chro mosomes to the pole s and continuous fib ers, or interzonalfi bers, extending from pole to pole.

Electron micrographs reveal that the spindle " fi bers" and astral " rays" ofclassical cy tology are bundles of microtubules. The microtubules are fixed to localdifferentiation s of the chromosomes called kinetochores . The polar end s of the chromosomal and continuous microtubules terminate in a specialized zone of cytoplasmicmatrix around the ce ntrioles . At anaphase of mito sis, the motion of the chromosomeshas two components: a chromosome-to-pole movement accompanied by a shorteningof the chromosomal microtubules and a movement of the pole s away from one another.The generation of the se movements remains a subject of controversy , with someinvestigators favoring a sliding microtubule mechani sm (McI ntos h et aI. , 1969) andothers favoring a polarized polymerization and depol ymerization res ulting in shorteningof the chromosomal microtubules and lengthening of the pole-to-pole microtubules(Inoue, 1976).

The accompanying micrograph shows groups of microtubules extending from thepole (out of the field at the upper left) to chromosomes (at the lowe r right). Thekine toc hore s are not visible.

Figure 4 10. Chromosomal micro rubules in midanaphase of mitotic divi sion in Pelomyxa. (Micrographcourtesy of Evans Roth .) Figure 4 10

759


When the chromosomes have moved tow ard the poles in late anaphase, thecleavage furrow con stricts around the midpoint of a large bundle of interzon al spindlemicrotubules that are sur rounded by a dense amorphous material. Thi s shaft ofmicrotubules, enclosed at its midregion by the plasma membrane , persists for sometime and connects the daughter cells well into interphase . The den se zone in theintercellular connection corresponds to the midbody (Fleming's body, Z wischenkorper)de scribed by classical cytologist s.

Figure 4 11. Mi crotubules assoc iated wit h the mid bo d y in te lophase of mammalian ce ll div ision.(Micrograph courtesy of D avid Phillips .)

Figure 411

761


The accompan ying micrographs pre sent additional examples of persisting interzonal spindle microtubules in telophase of dividing cell s. When the connection betweendaughter cell s is finally severed , the midbody is incorporated into one daughter cell.The dense matrix material disperses and the microtubules depolymerize.

Figure 4 12. Segment of rhe bridge co nnecri ng daughte r cells in H eLa cell mitosis. (Pho romicrographcourre sy of Br eck Byers.)

Figure 4 13. Per sistin g int er zon al microrubules in a d ividing guinea pig erythro blast. Figure 4 12, upper Figure 4 13, lower

763

764

Microtubules in Cell Motility

Shortening of skeletal muscle depends upon sliding of interdigitating sets of actinand myosin filament s. Their displacement relative to one another is caused bychemomechanical changes in orientation of minute cross bridges between the filaments(Huxley, 1969). The generation of bending movements in cilia and flagella is attributedto a sliding microtubule mechanism comparable to the sliding filament mechanism ofstri ated muscle (Satir, 1968). Sliding results from interaction of arms projecting fromeach doublet microtubule of the axoneme with the adjacent doublet (p. 577).

A comparable interaction of microtubules is in part responsible for large-scalechanges of shape in certain ciliate protozoa, the one most thoroughly studied beingSt ent or coerulus (Huang and Pitelka, 1973). From its trumpet-shaped extended state,this organism can contract to a quarter of its length in a few milliseconds .

Parallel rows of cilia extend the full length of the organisms. Immediately beneaththe cell surface between the rows of cilia, light microscopists described linearstructures which they designated Km fib ers. When examined with the electronmicroscope, these proved to be highly ordered arrays of microtubules. The accompanying micrograph shows a cilium flanked by two such microtubule arrays seen in crosssection. The microtubules are interconnected by short bridges to form discrete layers ,or ribbons, which in turn are attached to basal bodies of neighboring cilia. Adjacentribbons are connected by longer bridges between a few of the microtubules at the endsof the ribbons nearest the cell surface. During extension the overlapping ribbons of stiffmicrotubules slide with respect to one another, decreasing their degree of overlap andcausing the organism to elongate.

More deeply situated in the cell cortex are longitudinal bundles of filamentscomprising the myonem es, one of which is shown here in cross section. Shortening ofthe organism results from contraction of the myonemes.

Figure 4 14. Transvers e section of a portion of the cortex of a partially ext end ed Stentor coerulus. (FromHuang and Pitelka,J. Cell BioI. 57 :704-728, 1973.) Figure 414

765

766

, ,

jj

1-lIl1 i11l 1Uilt &IU:

Figure 4 15

CYTOPLASMIC MATRIX AND CYTOSKELETON

Th e micrograph on the facing page shows a longitudinal sec tion of a Km fiber in aContrac ted S ten tor. T he ribb ons diverge laterally to insert into the basal bodie s ofclosely space d cilia. Forty or more oveelapping miccotubule ribbons ace seen here, butin compamble views of a fully exte nded ce ll, the basal bodies would be spaced at longerinter vaj, and 'he numbe; of oveelapp;ng eibbons would be eeduced to about six. Theseunusually stra ight and rigid microtubules, like tho se of cilia and flagella , are notdepolym erized by colchicine.

Th e microtubule and filament sys tems in this organi sm function as antagonisticelements in con'mct;on and extension. The motive force foe extension depends uponmiccotubule-to-miccotubule slid;ng geneeated by specific ;ntenubu!e be;dges, whe;easthe active shonening appea" to resuh from a confoemational change in the filament scomprising the myonemes. This is of interest because it is possible that the movementsinvolved in shape change in Jess specialized cells of highee oegan;sms may PCOve '0 bedue to s;m;)ae antagonistic effects of micmtubules and contmctile filamentous componen ts of the cytoplasm.

=

Figure ' 15. 1.o"' i'"""'1""i"" of Km fibe, in the mn" of a m"'~'&J s,,"'0' . (From H " ,"g andPirelka,]. Ce l1 BioI. 57:704 -728, 1973.)

767

Cytoskeletal Function of Microtubules

One of the first functions ascribed to microtubules was that of providing an internalframework or cytoskeleton serving to determine and maintain cell form . Later, as cellelongation and the extension of cytoplasmic processes were found to be associated withthe appearance of axial microtubules, greater emphasi s was placed upon their role inshape change. Nevertheless , there remain a number of examples wherein microtubulesare responsible for maintaining a distinctive and stable cell shape.

A dark line called the marginal band (Randreifen ; striae bordante) was observed byRanvier (1875), Dethier (1875), and others around the periphery of nucleated erythrocytes of birds and amphibians. Meves (1904) attributed elastic properties to this structureand sugge sted that it served to maintain the flattened elliptical form of these cells. Heconcluded from studies on supravitally stained preparations that the marginal band wasmade up of fibrillar subunits . These remarkable observations were discounted byinfluential contemporary cytologists who considered the marginal band to be a stainingartifact (Weidenreich, 1905). Interest in this structure therefore rapidly declined.

Fifty years later when fish erythrocytes were examined with the electron microscope, Meves ' findings were substantiated by the finding of a circumferential bundle ofmicrotubules where he had described a marginal band (Fawcett and Witebsky, 1964). Insections perpendicular to the plane of cell flattening , a cluster of cro ss-sectional profilesof microtubules was found at both ends of the cell.

In the anucleate biconcave discoid erythrocytes of mammals, a marginal band isordinarily not present. An interesting exception is found in Camelidae , which haveanucleate erythrocytes that are elliptical rather than circular. These are said to possessa marginal bundle of microtubules (Barclay, 1966).

Figure 4 16. Eryth rocyte from the road fish Opsanus T au. (From Fawcett and Wit eb sky, Z . Zellfor sch.62 :785- 806 , 1964.)

768Figure 416


The assumption that a marginal bundle of microtubules in nucleated erythrocytes isrelated to maintenance of their discoid shape derives additional support from theobservation that the flat blood platelets have a similar disposition of microtubules(Behnke , 1965). In the equatorial section on the facing page , the bundle forms a ring justinside of the membrane but not in contact with it. In sections perpendicular to the disc ,the marginal bundle is seen as a cluster of microtubules at either end of the cross-sectional profile of the platelet (at the arrows and in the inset). Their number varies from 6 to 20or more. It is not known whether the bundle consists of independent closed circles ,open circles overlapping to varying extent, or a single long microtubule in a coil. Themicrotubules form a coherent structural unit which remains intact when the plateletmembrane is disrupted , but if the bundle itself is broken , the microtubules straightenout, indic ating that they have ela stic properties and are under some ten sion when in themarginal bundle.

Cooling of platelets to 0 to 4° C results in disappearance of microtubules , but theyreappear spontaneously upon rewarming and ultimately reconstitute a margin al bundle.Their reappearance is often associated with a resumption of the normal lenticular formof the platelet (Behnke , 1967).

Figure 41 7. Part o f an eq uato rial section of a hu man blood platelet. (M icrograph co urtesy of Ol afBeh nke.)

Figure 418. Sever al platelet s in a glomeru lar capillary of rat. (Microg raph courtesy of Greta Ty sonand Ruth Bulger. Am.]. Anat. 135: 319-343,19 72.)


771


In the organ of Corti in the inner ear, the hair cell s responsible for transduction ofmechanical vibration to a ner ve impul se are suppor ted by rows of cells called pillarcells . The micrographs show n here illustrate the thick walled microtubules in a pillarcell. These highly ordered stable arrays of microtubules interrupted by occ asionalmitochondria and tubular elements of the endoplasmic reticulum are believed toexemplify a purely mech anical cyto skeletal function of microtubules.

Fig ures 4 19 and 420. Electron micrographs of pillar cells o f the mamm alian orga n of Co n i. (Microg raphscourtesy of H ans Engsrro rn.)

Figure 4 19, upper Figure 420, lower

77 3


The columnar Sertoli cells of the seminiferous epithelium function as nur se cell s inmetabolic and mechanical support of the developing male germ cell. Late in spermiogenesis, the heads of maturing spermatozoa occupy deep recesses in the apicalsurface of their supporting cell s . The accompanying micrograph presents a sectiontransve rse to the axis of the Sertoli cells near the. lumen. Several sperm heads aresectioned at different level s, and the surro unding Sertoli cell cytoplasm contains largenumbers of cro ss-sectional profile s of microtubules that are oriented parallel to thevertical axis of the cell. Microtubules vary in abundance in different stages of thespermatogenic cycle , but when the Sertoli cell s are host s to advanced spermatids,cytoplasmic microtubules are always abundant and'consistent in their orientation. It isassumed that the microtubules in such supporting ce lls have an important cytoskeletalfunction.

Figure 42 1. H orizontal sec tio n th ro ugh the apical cyto p lasm of a Serroli ce ll in the testi s of the toad , Bufomartnus,

Figure 421

775

776

Microtubules in Morphogenesis

Studies of mammalian spermatogenes is provided one of the first observationsimplicating microtubules in a change of cell shape. The germ cell s are approximatelyspherical during meiosi s and the early postmeiotic stages of development. Thespermatids then undergo a rapid elongati on. This ch ange is associated with theappearance of a new organelle which classical cytologists called the caudal sheath , ormanchett e. It wa s de scribed as a sleeve-like cylindrical structure arising from anannular thickening of the cell membrane around the equ ator of the nucleus andextending back into the cytoplasm of the spermatid around the base of the flagellum.

In early electron micrographs of thi s region , the manchette wa s found to be acylindrical arra y of parallel microtubules (Burgos and Fawcett, 1954). It is a transientstructure ari sing at the beginning of spermatid elongation and disappearing completelywhen this stage of morphogenesis is completed. The elongation of the microtubulesfrom a fixed origin proximally evidently contributes to lengthening of the cell. Theclose ass ociat ion of vesicular elements with the outer and inner aspect of the manchettesugge st s the pos sibility that it may also participate in transport of particulates or at leastse rve as a framework guiding a caudal flow of cytoplasm .

The microtubules of the manchette are more stable than those of most other celltypes. This may account for the fact that they were observed before the introduction ofaldehyde fixatives.

Figure 422. Transverse section thro ugh th e caudal cytoplasm of a Chinese hamster spermatid. (Micro grap h courte sy of D avid Phillip s.) Figure 422

777


It has frequently been sugge sted that the mobility of certain classes of transmembrane proteins is limited by their attachment to microtubules and other cytoskeletalelements (Nicolson, 1976). Microtubules have also been ass igned a role in guidingsecretory granules to the surface for rele ase (Lacy and Malaisse , 1973). However,convincing micro scopic evidence for termination of microtubules at the inner as pect ofthe membrane has seldom been pre sented . The manchette is one of the few unambiguous examples of a microtubule organelle ending at the cell membrane. Asillustrated in the accompanying micro graph , its con stituent microtubules end in a densesubplasmalemmal matrix seen at the top of the figure . Thi s material occupies theconcav ity in a prominent annular ridge on the cell surface around the ca udal portion ofthe spermatid nucleus . In thi s case , attachment of microtubules doe s not seem to be anexample of cy tos keleta l control of transmembrane prot eins but merely pro vide s a fixedorigin from which pol ymerizing microtubules lengthen to achieve the internal deformation nece ssar y for spermatid elongation .

Figure 423. Longitud inal sec tio n of the rnancherre of a rodent spe rma tid .Figure 423

779


The microtubules composing the manchette are joined together in an irregularpattern by slender cro ss bridges as seen at the arrows in the accompanying micrographs. Similar linkages are seen in various other organelles composed of microtubules,such as the axonemes of Heliozoa , the Km fibers of St entor . There is evidence that thebridges connect to certain specifi ed protofilaments and project at a fixed angle to thetubule. The limited variety of geometric patterns displayed by microtubule arrays isevidently an expression of the se re straints (Fuj iwara and Tilney , 1975). Bridgesprojecting from microtubules at the outer and inner aspects of the manchette frequentlyappear to attach to the membrane of the ass ociated ve sicle s or tubular elements of theendoplasmic reticulum (at asterisks). It is not known whether the se are enduringattachments or transient linkages involved in a transport mech ani sm (Fawcett , Anderson and Phillips , 1971).

Figures 424 and 425. Nucleus and ad jace nt manc het te of rod ent spermatids .Figure 424 , upp er Figure 425, lower

781

782 CYTOPLASMIC MATRIX AND CYTOSKELETO N

In spermiogenes is of various spec ies , there are numerous exampl es of the transientappeara nce of arrays of microtubules during the shaping of the nucleus and ce rta incomponents of the sperm tail. The dispo sition of the microtubules suggests that if the ywere capable of generating movement, they might co ntribute to the ob served shapechange s. The accompanying micrograph s from avian spermiogenes is illustrate thi spoint.

The spe rmatozoa of finches have a corksc rew-shaped head and a very long-pit chedmitochondrial helix extending throughout the greater part of the length of the tail. Earlyin spermiogenesis when the spe rm head is beginning to take on its unu sual form , a largebundle of microtubules pur sues a helical course around the co nde nsing nucleus asshown in A. Th e microtubules later disappear when the nucle ar shaping is co mplete ,but in the surface of the mature sperm head , show n in B, shallow depression s (atarrows) mark the former site of clo se apposition of the helical micro tubule bundl e.Although intrinsic morphogenetic fac tors ass oc iated with chromatin co ndensa tion maybe largely responsible for the shape ass umed by the nucleus , the relationship of thehelical bundle of microtubules sugges ts that the latter may play an ancillary morphogenetic role .

The microtubule bundle continues posteriorly and wind s around the midpiece ,formin g a double helix with the continuous strand of fused mitochondria (Panel C). Themitochondrial helix of the mature sperm (Panel D) ret ain s the same pitch as themicrotubul e bundle after the latter has disapp eared . It is difficul t to escape theco nclusion that thi s transient microtubule organelle is involved in determining the formof the mitochondrial sheath.

Figure 426. Stages in the differe ntiation of the finch spe rmatozoon. (From Fawcett , Anderson andPhi llips, Dev. BioI., 26:22 0-2 51, 1971.)

Mitochondria

Figure 426 A - D

783

784

CYTOPLASMIC FILAMENTS

When the microscop ic study of ce lls shifte d in the later part of the 19th century fromobse rvation of living material to examin ation of chemically pre served and sta ined cells ,man y of the leading cytologi st s of that period held the view that the fundamentalstructure of protoplasm con sisted of delicate fibrils tr aversing a homogeneous groundsubstance . Fleming, a leading proponent of the fibrillar theory of protoplasm , ascribedto these unbranched and discontinuous filar elements a major role in the activities ofcell s. Others, such as Carnoy and Van Beneden , argued just as strongly in favor of thereticular theory , which assumed that the fibrillae formed a continuous branching andanas tomos ing network. Fibrils of muscle were thought to be local modifications of ageneral fibrillar network , and spindles and asters were believed to ar ise around thecentrioles by rearrangement of preexisting filaments. Con sidering the methods used andthe limited re solution of their microscopes, it is likely that their inte rpretations werebased in part upon a relatively coarse pattern of precipitation of proteins by thefixative s. But Rem ak had seen neurofibrils in nerve cell s as ea rly as 1843, and Fromandemonstrated them by use of silver salts in 1864. There is no doubt that some of thecoar ser meshed networks observed in other cell type s represented true filamentou scomponents of the cytoplasm. Heidenhain (1899) concluded that the se were supportingor cytoskeletal structures and therefore called them tonofibril s.

In the next half century , the list of sepa rately named filamentous componentslengthened to include neurofil aments, glial filaments , tonofilaments, myofilaments, andspindle fiber s, but there was no evidence to indicate whether or not the se we re differentfibrou s proteins . Electron micro scopy set apa rt microtubules and revealed two filamenttype s in myofibrils. At least two categories of cytoplasmic filament s of generaloccurrence were distinguishable on the ba sis of size - 6 nm filament s and 10 nmfilaments . It now appears from electrophoretic and immunocytochemical studies thatthere are multiple categorie s of 10 nm filament s. Re search on the filamentouscomponents of cytoplasm and their respective functions is now one of the most activeareas in cell biology.

Filaments 0/Striated Muscle

The cro ss str iations of muscle fiber s were first ob served by Leeuwenhoek (1677),who believed that the striae prob ably represented a spiral elastic coil capabl e ofretraction to produce movement. Myofibril s teased from insect muscle were observedto contract under the micro scope (Rollet , 1885; McDougall, 1897). The fact that the ybulged between successive Z-line s und er these conditions led to the speculation thatcontraction re sulted from swelling (McDougall , 1898). What was known of thebiochemistr y of contraction seemed consistent with the view that the chemical change snecessary for contraction and relaxation were produced by accumulation of lactic acidwithin the muscle and its neutralization by buffer sys tems in the surrounding extracellular space (Meigs, 1912).

When examined at high magnification in stained preparations, the cro ss striationsappeared as alternating dark and light bands. Under the polarizing micro scope , bandsth at were dark in st ained prepar ations appeared light, or anisotropic, and weretherefore calledA-bands , whereas the light bands of sta ined specimens appea red dark , orisotropic , and were designated l-bands (Brucke , 1858). The A-band in stained sectionswas frequentl y bisected by a lighter central region called the H-b and and in favorablepreparations the middle of the H-b and was traversed by a slender M-line. A darktran sverse line in the middle ofthe I-band was called the Z-line . A segment of myofibrilbetween two success ive Z-line s was termed a sarcomere.

Polarization optical studies of muscle were taken up again in the early 1900s, whenits theory was better understood . The positive birefringence of the A-band in polarizedlight was interpreted to mean that it was composed of long molecules or micelle sarranged parallel to the long axis of the myofibril. The I-band wa s also ob served to beweakly birefringent after extraction of pho spholipids and therefore probably alsoconta ined oriented asymmetrical molecules.

The first significant biochemical studies of muscle were undertaken in the 1940swhen Szent-Gyorgyi extracted a protein , actomyosin, which displayed physical properties characteristic of long asy mmetrical molecules, and was shown to be capable ofcontracting in vitro upon addition of ATP . By differential extraction , actomyosin couldbe separated into an asymmetrical protein, myosin, which pos sessed ATPase activity ,and a globular protein, actin . If the ionic strength of the medium was adj usted to that ofthe sarcoplasm of muscle , actin polymerized to a higher molecular weight fibrousprotein (F-actin).

The histochemical approach to the study of tissue s was popular at that time amongmorphologist s. It was shown that extraction of fre sh muscle in solutions known toextract myo sin resulted in loss of the birefringence of the A-band s. These ob servationslocalized myo sin to thi s segment of the sarcomere and provided evid ence that theanisotropism of the myofibril s wa s due in large part to orientation of long molecules ofmyo sin. But the mechanism of contraction was still elu sive . Some postul ated a changein configuration of the myo sin mole cule s involving a folding and cro ss-linking comparable to that of keratin molecules in the shrinkage of wool (Astbury, 1944). Othersenvi sioned a process comparable to blood coagulation and clot retraction , but none ofthe se theorie s could adequate ly expl ain relax ation .

The earliest studies of muscle by electron microscopy confirmed the pre sence offilamentous subunits in myofibrils (Hall et al. , 1946; Draper and Hodge , 1949), but little

785

786 CYTOPLASMIC MATRIX AND CYTOSKELETO N CYTOPLASMIC MATRIX AND CYTOSKELETON 787

more could be learned from dissoci ated , dried , and metal-shadowed preparations . Thedevelopment of methods for cutt ing ultr athin sections made it possible to examine thest ruc tura l co mponents of muscle in their normal spat ial re lat io ns and with highresolution. The myofi bril, the smallest unit of the contractile material visible with thelight microscope , was found by Hu xley and Hanson (1954) to be composed of smallerunit s, the myofil am ents. Th ese were of two kind s, thick 11 nm myosin filam ent s andthin 5 nm actin fi laments . The cross-banded pattern of striated muscle was found to be areflection of the arrangement of these two ca tegories of fil ament s. Arrays of parallelmyosin fil amen ts 1.5 fLm long and 45 nm apart composed the A-band . They alterna tedalong the length of the myofibril with arrays of actin filament s that exte nded about 1 fLmin e ither direction from each Z-line , where they appea red to be att ached laterally. Actinfilaments not only made up the l-band , but the y ex tended some distanc e into theadjacent A-bands , occupying the interstices bet ween the hexagonally packed myosinfilaments. The depth to which the actin fi laments pen etrated into the A-band was foundto vary with the degree of contraction . The central region of the A-band not penetratedby the actin filaments corresponded to the H-band of light microscopy. In the region oftheir interdi gitation at either end of the A-band , the thick and thin filaments were onl y10 to 20 nm apart , and thi s narrow interval wa s spanned by evenly spaced cro ss bridgesthat radiated from each myosin fi lament tow ard the neighboring actin filaments.

When the ultrastructure of muscle was examined in different states of contractionand extension , it was found that the length of the A-band remain s con stant but thelength of the l-b and varies with the degree of contraction . In stretched myofibril s, theactin filaments are drawn out at eith er end of the array of myo sin filaments and theH-b and becomes ver y bro ad , whereas in the contracted state, the actin filamentspenetrate more deepl y into the A-band and the H-b and is shortened. These observation s led Hu xley and Han son to conclude that muscle shortening result s from sliding ofthe ac tin fi laments along the myosin filaments tow ard the center of the sa rcomere. Thecross bridges on the myosin filaments sugges ted further that these produced the slidingby a repetiti ve cycle of attachment and detachment to sites on the actin filament. Witheac h cycle of operation of the cross-linkages , the ac tin fi laments are forced to move ashort distance farther into the A-band.

The sliding-fi lament model of muscle contraction provided , for the first time , alogical and satisfying explanation of many aspects of the general phy siology of muscle .Its evolution is an excellent example of the fundamental importance of morphologicalob serv ations . In several dec ade s of inten sive physiological study and speculation onpossible mech ani sms of muscle contraction , a theory invol ving sliding of filament s ofconstant length had never been advanced , but when the topographical relations of theult rastructural components of myofibril s were revealed by the electron micro scope ,such a mech ani sm at once became obvious and onl y needed experimental confirmation.

Dissoci ation of actin and myosin filaments into their molecul ar subunits and theirexamin ation with the electron microscope soon followed . Isolated myosin filaments are1.5 fLm long with a 0.2 fLm smooth centra l segment , but tow ard each end there arenumerous short lateral projections that corres pond to the cro ss bridges between thickand thin filaments seen in thin sections of muscle. The myosin filaments can be furtherdissoci ated into myosin molecules . These are approximately 160 nm long and con sist ofslender rod s 2 to 4 nm thick and a wider terminal segment.

Brief exposure to protease cle aves the molecule into two fragments calledmeromyosins . The rod -like tail, or backbone, of the molecule is light meromyosin(LMM) and a short segment including the laterally projecting head is hea vy meromyosin(HMM) . The heavy meromyosin fragment retains the adenosine triphosphatase activityof the myo sin molecule and the capacity to bind to actin .

Isolated actin filaments are about 1 fLm long and at high magnification appear toconsist of globular subunits 5.5 nm in diameter arranged in two helically entwinedstrands . Reduction of the ionic strength of the medium result s in further dissoc iation offilamentou s actin (F-actin) into its globular subunits (G-actin). The thin filaments of thel-band of muscle are therefore polymers of G-actin molecules. Myo sin and ac tinfilaments similar to tho se of intact muscle can be reassembled in vitro from theirmolecular subunits .

Two physiologically important regulatory proteins are assoc iated with the actinfilaments, but the se are not resolved in electron micrograph s . Each of the helicallyentwined strands of actin compri sing the thin filaments of muscle is believed to have anassoc iated submicroscopic stra nd of trop omyosin molecules arranged end to end(Ebas hi, E ndo and Ohtsuki , 1969). Each molecule of tropomyosin is 40 nm long ,spanning nine actin molecules. The second regulatory protein trop onin is a globul armolecule attached to each tropomyosin molecule near its end . Troponin bind s calciumthat is released from the sa rcoplas mic reticulum in respon se to nerve stimulation.Calcium binding is belie ved to result in a change in configuration of tropomyosinmolecules, expos ing binding sites for the projecting heads of the myosin molecules ofthe adjacent thick filaments.

z z

~~

CO N TRACT EDI

Ir-;.-AI

---;-11

RESTING LENGTH

HI I

( (

(

< (

STRETCHED

D iagrammatic re prese ntation of the va ry ing de gree of int e rd igitatio n of ac tin a nd myo sin filam en ts(right) that acc ounts fo r the c ha nge s in the patte rn of c ross-ba nd ing o bse rve d in differ en t sta te s ofmyofib ril co nt rac tion (left ). T he A band is of co ns ta nt le ngth bu t the le ngth of the H a nd I ban ds isdep endent upon the dep th of pen et rati o n of the thi n filam ent s int o the A band. (F rom Bloom a ndFa wcett , T extb ook of H istology. 10th ed.. w. B. Sa unde rs Co. , 1975. )


The cross-banded appearance of striated muscle illustrated in the accompanyingmicrograph of cardiac muscle results from the ordered arrangement of actin and myo sinfilaments . The dark A-band s are of con stant length corresponding to the length of themyo sin filaments. The lighter I-band con sisting of actin filaments is bisected by theden se Z-line , where the actin filaments are joined laterally . During contraction, theI-b and becomes narrower as the actin filaments penetrate more deeply into theinte rstice s between the myo sin filaments of the A-band . Higher magnific ation isrequired to re solve the relations of the two sets of filaments.

Figure 427 . Papillary muscle of car hearr.Figure 42 7

789


Here o ne sarc ome re length of two adjacent myofibril s is sho wn at highermagnification . The A-band is seen to be composed of a par allel array of myosinfilaments 10 to 11 nm thick and about 1.5 JLm long . The light l -band consists of actinfi laments about 6 nm thick extend ing in either direction fro m the centra l Z-line . Actinfilaments are not confined to the I-band but penetrate some distance into the A-band ,occ upying the interspaces bet ween the myosin filamen ts. Th e brack et marked Xindicate s the region of overlap of the two sets of inte rdigitat ing filaments. The bracketmark ed Y indicate s the ce ntral region of the A-band, where there are onl y myosinfilaments. This region co rresponds to the H-band of light microscopi sts .

Fig ure 428. Papillar y muscl e of car heart .Figure 428

79 1


The accompanying micro graph shows at high magnification the region of interdigitat ion of actin and myosin filaments at the end of the A-band. The unfixed musclewas quickly frozen using liquid hel ium, then fractured and deep-etched . Cross bridgesbet ween the myosin and acti n filaments are clearly visible . A fine axia l periodicity ca nbe seen in the actin filaments of the I-band at the left of the figure.

Figure 429. Skeletal mu scle prepared by qu ick freezing and deep e tching. (Micrograph co urtesy ofJ oh nH euse r.)

Figu re 429

793

794 CYTOPLASMIC MATRIX AN D CYTOSKELETON

The highly ordered arrangeme nt of thick and thi n filame nts is seen with unu sua lc larit y in insect musc le. Indirect flight musc les shorten very little when they co ntract .The l-bands are shorter at resting length and the A-bands longer than in the striatedmuscle of vertebrates. The thin fi laments penetrate far into the A-ba nd , resu lting in anunusually narrow H-band where only the th ick fi laments are present. In the accompanying micrograph , the cross bridges are barely discernible.

Fi gure 4 30. Flight muscle of Drosophila melanogaster. Section 0.25 p.. m thick, pho tograp hed at amagnification of 83, 000 with a 1 million vo lt electron microscope. (Courtesy of H ans Ris, HVEM Facility,

Un ive rsity of Wisconsin.)

Figure 430

795

796 CYTO PLASMIC MATRIX AND CYTOSKELETON

Tran sver se sec tions through the A-band of insect flight muscle display a remarkably regular latti ce of filaments. The thick filaments are hexagonally arranged and eachis surrounded by six thin filaments. In insect muscle , the actin filaments are in line withthe rows of myosin filaments in all axes. In mammalian muscle , they are situated in thetrigonal positions of the lattice so that each is shared by three equidistant thick fila-

ments.

Fig ure 43 1. Asynchron ou s flight muscle of a fly, Bombylius major. (Micrograph courtesy ofJ. Auber , Am .

Zoo!' 7:451- 456, 1967.)

Figure 4 31

797

798

Filaments 0/Smooth Muscle

With routine methods of specimen preparation , the cytoplasm of smooth mu sclecells often appears surprisingly homogeneous . There are clu sters of ribo somes, a fewmitochondria , and numerous foc al den sities scatt ered throughout the cytoplasm, butthe 6 nm filaments normally pre sent in great abundance are only re solved with somedifficulty. Since it has been shown that osmium tetroxide destroys actin filaments invitro (Maupin-Szamier and Pollard , 1978), it is likely that the full complement offilaments is seldom pre served in tis sue processed for thin sectioning. However , infortuitous preparations viewed at high magnification, the actin filaments can sometimesbe seen . The area enclosed in the rectangle is shown at higher magnification in the nextmicrograph.

Figure 432. Smo o th muscle of mou se epididyma l duct. Figure 432

799


The cut ends of actin filaments appear as dark dots of uniform size . They are notdistributed in a regular pattern as are those of str iate d muscle but occur in irregulargro upings that represent bundles or tracts of parallel filaments. Their small size ca n beappreciated by compari son with the ribo somes (16 nm) at the upper right.

Thick filame nts of myosin are known to be pre sent from observat ions on smoothmuscle glycerinated at pH 6.0 (Kelly and Rice , 1968) and from x-ray diffraction patterns(Small and Squire , 1972), but under conventional conditions of chemical fixation themyosin of smooth muscle is quite labi le. The micrograp h on the facing page was amongthe first (1966) purporting to show thick filaments (at arrows). It has since been learnedthat very carefully balanced conditions of divalent cation and ATP concentration , ionicstrength, and pH must be maintained to preserve smooth muscle myosin in fi lamentousform (Small, 1977). Although the filaments shown here were assumed to be myo sin,their size and sharp cross-sectional outline now make it likely that these are 10 nmfilaments. The observation that they ten d to be clustered aro und the dense nodes in thecytoplasm has sugges ted that the se components may form a complex cyto ske letondistinct from the contractile apparatus (Small and Sobieszek , 1977).

Figure 433. Smooth muscl e of mo use epid idymis.

Figure 433

801

802 CYTO PLASMIC MATRIX AND CYTOSKELETO N

With appropriate preparati ve techniques, it is occasionally possible to preservemyosin filaments in smooth mu scle. In the accompanying micrograph of vascularsmooth muscle, numerous thick filaments are uniformly distributed among the actinfilaments. As can be seen at higher magnification in the inset , the actin filamentsou tnumber the myosin filaments about 15:1 and the latter are rather irregular in outline .

Figure 434. Electro n micrograph of a vascu lar smooth mu scle ce ll in cross section. (Micrograph court esyof An drew Soml yo from Vascular Smooth MttScle, Springer-Verlag, H eidelbe rg, 1972 .)

Figure 434

803

804

Actin and Myosin o/Nonmuscle Cells

Contractility has long been recognized as one of the fund amental properti es ofprotoplasm , ex pressed in cytokines is, protoplasmic strea ming, amoe boid locomotion ,and morphogenetic movement s of eukaryotic cell s. But until recentl y noth ing waskno wn about the molecular mech ani sm involved .

In one of the earl y approaches to thi s problem , Loewy (1952) prepared an extractof the slime mold Physarum , which had some properti es in common with actomyos infrom muscle . The extract was able to hydrolyze ATP , and ATP had a demonstrableeffect on its viscosity. These finding s sugges ted a similarity in the basic mech ani smsunderlying muscle contraction and protoplasmic st rea ming of slime mold s. Crudeactomyos in preparations were soon extracted from human platelets (Bettex-Ga llandand Lu scher, 1959).

The demonstration that nonli ving glycerina ted muscle and isolated myofibril scontrac ted when tre ated with ATP in the presence of MgH ions stimulated HoffmanBerling to undertake similar experiments on cell s othe r than muscle . In a va riety ofglycerin-ex trac ted cell types, he observed feeble contractions upon addition of MgHand ATP. These studies att racted little attention because the highly hetero geneousextracts see med to offe r little po ssibility of isol ation and ana lys is of the ac tivecomponents. But with improved biochemical methods for dealing with such complexmixtures, it was soon possible to purify actin (Hata no and Oo sawa , 1966) and myosin(Hatano and Taz awa , 1966; Adelman and Taylor , 1969) from Physarum and todemonstrate their close relationship to the actin and myosin of muscle. The addition ofgel electrophoresis and column chromato graphy to the armamenta rium of the biochemist has enabled thi s field to move forward rapidl y in the last decade .

Morphological identification of actin in cells was made possible by developm ent ofan ingen iou s technique which involves the use of the heavy merom yosin fragment ofthe myosin molecule as a histochemical reagent. He avy merom yosin will enter ce llsafte r glycerin extraction and will bind to the cytopl asmic actin filaments as it does tomuscl e actin. It binds in a di stinctive configurat ion sugges tive of "arrow hea ds "repeating at regul ar inte rvals along the filament (Ishikawa, Bischoff and Holtzer, 1969).The " decoration" of cytoplasmic filaments with heav y meromyosin is con sidered ahighly specific mean s of identification of filamentous actin.

Cytoplasmic actins from a number of cell types have been purified, and in all oftheir physicochemical parameters they are remarkabl y similar to muscle actin. Inmorphological studies by negati ve sta ining, actin occurs in filament s about 6 nm inthickness . These cons ist of 5 nm globul ar monomers arranged in a doubl e helix. Whendecorated with heavy meromyosin , the arrow heads have a periodicit y of about 38 nmcorres ponding to the half-pitch length of the double helix.

Neither ac tin nor myosin alon e can generate force for movement. The presence ofactin in cell s therefore implie s the coexistence of myosin . As expected , myosin hasbeen found in severa l ac tin-conta ining cell types, including platelet s, granulocytes,fibrobl ast s, and neurons , and it is able to bind to act in and possesses actin-ac tivatedATPase activity . Unlike cytoplasmic actin, the myo sins that have been isolated exhibitcon siderable variation in their phy sical , chemical , and enzy matic properties. Filamentsof the size and shape of myosin filaments have been observed in the cytoplasm of anumber of cell type s, but in the abse nce of a spec ific method comparable to decoration


of actin filaments, their identifi cation is less certain . Antibodies have been prepared tomyosin from str iated muscle , platelets, and granulocytes , and to myosin from smoothmuscle, and these have been used in the immunohisto chemical localization of myosinsin oth er ce ll types. Although useful , these lack the spec ificity of antibody prep aredaga inst pure myosin of the type characte ristic of the cell type being studied .

In most ce lls, actin filaments are present in the cy toplasmic matri x , es pecially in aperipheral ectoplasmic zone immediatel y beneath the ce ll membrane. In some epithelialcells posse ssing a bru sh border, a bundle of actin filament s is found in the core of eachmicrovillus . In the se and other cells, the actin filaments appea r to be att ached to the cellmembrane. Thi s anchoring of the contractile app ar atu s to the cell surface may beessential for motilit y or for sha pe change. In cell s in tissue culture , bundles of actinfilaments are ass oc iate d with the surface that is in contact with the substrate and ex tendout into the ce ll processes. These were designated " stress fi bers" before their chemi calnature was established , and the term continues to be used .

Until more inform ation is available on the form and distribution of myosin in ce lls,little ca n be said about the mechani sm by which actin and myosin interact to producemovement in ce lls other than muscle . Relati vely little progress has been made to date inlocalizing myosin in electron micrograph s of nonmuscle cells . Myosin filaments cansometimes be identified in the cytoplasm of amoebae, but routine methods of fixationapparently do not pre serve myo sin filaments in the cells of vertebrate s and as a ruleonly ac tin filament s are see n. Myo sin purified from such ce lls can be induced to formfi laments in vitro , but it seems probable that in the cy toplas m it occurs inste ad in theform of very small aggrega tes . However, it is conceivable that ifit is pre sent in the formof rand omly ori ent ed filaments in low con centration , these would appea r in sec tions asdots or short linear profiles indistingui shable from oth er cyto plasmic proteins precipitate d by the fixat ive (Pollard and Weihin g, 1974). The straight 6 nm ac tin fi laments thatare see n in elect ron micrograph s probably represent only a fraction of the ac tin pre sentin the living ce ll, fo r it is kno wn that osmium-conta ining fi xati ves extract actin ortransform it into a microfil ament net work that is difficult to identify in thin sections .Tho se straight 6 nm filaments that do resist extraction may have been stabilized byass ociation with other proteins such as o-actinin or tropomyosin . Much of the currentresearch on localization of actin and myosin therefore relie s upon the use of fluore scentantihodies at the light microscope level of reso lution.

805

806 CYTOPLASMIC MATRIX AND CYTOSKE LETO N

When antibody again st puri fied actin is reacted with fl attened cells spread out uponthe substrate in tissue cultures , the fluorescence is assoc iate d with long straight bundlesof filaments travers ing the perinuclear cytoplas m and extending out into cell pro cesses.They co rrespond to the so-called " stres s fibers" see n in living ce lls by phase co ntrastmicro scopy and to bundles of 6 nm filaments seen in electron micrograp hs (Goldman etaI., 1975). Although the rigid appearance of these microfilament bundles might suggestthat they are static cytoskeletal elements, their rapid reorgan ization du ring locomotorchanges in ce ll-to-s ubstrate co ntac t indicates that they are part of a dynamic systeminvolved in ce ll motility .

Figure 4 35. T hinly spread hum an skin fibroblast stained by ind irect immunofluor escen ce with actinspec ific ant ibod y. (From Elias Lazarides ,). Cell Bio!. 65:549- 56 1, 1975.)

Fig ure 435

807

808 CYTOPLASMIC MATRI X AND CYTOSKELETON

In ce lls in their norm al env iro nme nt in vivo 6 to 7 nm actin filam ents occurringind ividu ally or in small groups inte rlace to form a co mplex mesh work in the cy toplasm . Wh en th e same ce ll types a re ex planted an d ma intain ed in culture , th e filame ntstend to aggregate into co a rse bu ndles. In the acco mpa nying mic rograph of a Se rto licell in vitro there a re many ind ivid ua l filaments and se vera l large aggrega tions ofparallel filaments. T he lat te r co rres pond to the "stress fiber s" shown in the previou sfigure of a tis sue cu lture ce ll sta ined with fluore scein conj uga ted antibody agains tac tin.

f igure 436 . Cyroplasm of a Serroli ce ll fro m tes tis of C irell«s lateralis in cultu re. (Micrograph co urtesyof Wayne Vogl.)

Figure 43 6

809

810

Filamentous Cytoskeleton

Now nearly a century after the controversy among light microscopists as towhether the fundamental organization of protoplasm was fibrillar, reticular, or alveolar, investigation of the organization of the cytoplasmic matrix has once againbecome one of the most active frontiers of cell biology. High-voltage electronmicroscopy for three-dimensional imaging of the cytoskeleton and immunocytochemical procedures are rapidly contributing to a better understanding of the structural basisfor cell movement and for the translocation of organelles and other particulatecomponents within the cytoplasm.

A three-dimensional network of fine filaments was described in transmissionelectron micrographs of the axoplasm of giant axons (Metuzals, 1969). From the studyof stereo pairs taken with the high voltage electron microscope, Porter and hiscolleagues (1976) demonstrated in thinly spread tissue-culture cells slender, branchingstrands, or microtrabeculae, 4 to 10 nm in diameter forming a three-dimensional latticethroughout the cytoplasmic matrix. This micro trabecular lattice is attached to theplasmalemma and to membranous cytoplasmic organelles and its meshes are traversedby 10 nm filaments and microtubules.

Schematic presentation of the microtrabecular lattice in the cytoplasm of a thinly spread cell inculture as seen with the high voltage electron microscope. The trabeculae are joined at thicker nodalpoints. The lattice is traversed by microtubules and the trabeculae appear to att ach to these, to mitochondria, and to the cell membrane. (Drawing courtesy of Keith Porter.)


Concurrently with these electron microscopic observations, other cell biologistshave exploited the potential of fluorescent antibodies for lc:caliza~ion of speci~c

structural proteins in tissue-culture cells at the level of the light microscope. Withantiactin, long, straight, fluorescent strands were shown (Lazarides and Weber, 1974),corresponding in distribution to the " stress fibers" originally described in cu~tured cellsby Warren Lewis and to the bundles of 6 nm microfilamen~s encountered m el~ctron

micrographs of such cells (Buckley and Porter, 1973). This pattern of conspicuousbundles of actin filaments was found most commonly in sessile cells firmly attached tothe substrate, whereas in more motile cells or in regions with active undulating ruffles,few if any discrete microfilament bundles were visible and the actin fluorescence wasmore diffuse. The organization of the actin therefore appeared to be clearly related tothe state of activity of the cell. The ultrastructural configuration of ~he ac~in corresponding to diffuse fluorescence at the light microscope level was not Immediately ap-~re~. . .

To observe the cytoskeleton with the resolution of the electron microscope andwith the possibility of parallel immunocytochemical identification of its constituents,Weber and associates (1978) employed an ingenious technique involving removal of themembranes of cultured cells with detergent (Triton) and extraction of soluble proteincomponents, followed by dehydration and critical-point drying. The resulting preparations showed a three-dimensional lattice of microtrabeculae comparable to thatdescribed by Porter. Ferritin-labeled antibody against actin not only decora~ed microfilament bundles but the finer lattice of microtrabeculae as well, suggestmg thepossibility that both consist of actin.

The application of the new technique of rapid freezing of unfixed cells followed bydeep etching and rotary shadowing (Heuser, 1979) has provided further insight into thenature of the cytoskeleton. Such preparations reveal that both the "stress fibers" andthe microtrabeculae are composed of filamentous subunits with an axial periodicitycomparable to that of actin. Thus evidence is accumulating for a major cytoskeletal rolefor actin.

811

812 CYTOPLASMIC MATRIX AND CYTOSKE LETON

Ev ide nce that the micr otrabecul ar meshwork of cy to plasmic matrix con sist smainl y of ac tin has come from a novel approach, in which dernembranat ed , ex trac tedtiss ue-c ulture cell s are qui ck frozen with liquid hel ium , deep etc hed , and rotaryshadowed. The res ulting micrographs viewed as negati ves in stereo pai rs have theth ree -dimen sional qualit y of scanning electro n micrograp hs and the high resolu tion oftran smi ssion electron micro sco py (Heuser, 1979).

The accompany ing low-power micrograph of the " cytos keleton" of a portion of ace ll in tissue culture does not do j ustice to the poten tial of the method for threedimen sion al ana lysis , bu t it clearly shows a co mplex meshwork of microtrabecul aesimilar to that seen in more limited areas of fixed ce lls viewed by high-voltage electronmicroscop y. At higher magn ificati on in succeeding figure s, the strands represe ntingmicrot rabecul ae are seen to have a fine pe riod ic ity correspo nd ing to that of actin.

Figure 437 . Microgr ap h of the microtrabecu lar meshwork of a 3T 3 cell, treated with Trito n, washed ,qui ck fro ze n with liqu id helium, ro ta ry shadowed . (Micrograph courtesy o f J o hn H e user. )

Figure 437

813


At higher magnification of the cyto skeleton in the accompanying micrograph , two" stress fibers" are seen to be composed of bundles of filaments that are continuous attheir margin s with elements of the general microtrabecular meshwork , thus stronglysugges ting that the two are alternative configurations of the same structural protein .

The term cytoskele ton has been used to refer to passive stiffening, or stabilizing,filamentous components of the cytoplasm , involved in the maintenance of cell shapeand providing a framework of fixed origins and insertions for the motor elements of thecytoplasm. If both the stress fiber s and microtrabecular meshwork consist of actin,which under appropriate conditions can also serve as a component of the motorapparatus of the cell , then the appropriateness of referring to the se elements as part ofthe " cytoskeleton" may have to be reconsidered .

Figure 4 38. Port ion of the cytoskeleton of a 3T3 fibro blast, dernembranared, de ple ted of solub leprote in, qui ck fro zen , deep e tched, and rotary shadowed . (Microg raph co urtesy of J ohn H euser.)

Figure 438

815


Shown here at high magnification are some large stress fiber s. The filaments ofwhich they are composed clearly show a 5 to 6 nm periodicity apparently correspondingto the globu lar subunits of fibrou s acti n.

Figure 439. Portion of the fibro us fram ework of the cytoplasm of a 3T3 fibroblast, dememb ranaced ,deple ted of solub le pro te in, qui ck fro zen, dee p e tched, and ro tary shado wed. (Microg raph courte sy of J ohnH euser.)

Figure 439

817


Presented here are two additional examples of the den se cytoskeleton in theperipheral cytoplasm of tissue cul ture cells . In the upper figure , two microtubules ca nbe seen traver sing the field and passing through the meshwork of filaments.

Figures 440 and 44 1. Cyto ske le ton of tissue cultu re ce lls pr ep ared by qu ick freezing, deep e tching, androtary shadowing. (Micro graphs courtesy of Jo hn H eu ser. )

Fig u re 440. upper Figure 44 1. lower

819


Purified actin con sist s of a feltwork of filaments , which shows a periodicity similarto that exhibited by the trabecular meshwork of the cytoplasmic matrix.

In the lower figure , a preparation of ac tin has been decorated with heavymeromyosin, resulting in a rope-like helic al disposition of the attached myosin fragments.

Figures 442 and 443. Purifi ed pr eparation of actin undecorated and decorated with heavy meromyosin.(Microg rap hs cou rtesy of J ohn H eu ser. )


821


It now seems likely that tropomyosin is also ass ociated with actin filaments innon muscle cell s. Immunocytochemical sta ining of vari ou s cell line s in cu lture usingantitropomyos in antibody reveals a linear pattern of fluore scing fiber bundles indi stinguishable from that seen with fluore scent antibody aga ins t actin .

The accompanying photomicrograph of a fiWroblast exposed to antit ropomyosinant ibod y clo se ly resembles the comparable pho~micrograph (Fig. 435) showing thedistribution of actin in the same cell type . Therefore both antibodies are apparentl y

;.'t

binding to the same filamentou s component of th~, cytoplasm.

Figure 444 . Human skin fibroblasr in culrure srained by indirecr imm unofluor escence using tropornyosin anribody . (From Elias Lazarides, ]. Cell BioI. 65:549-561, 1975.)

Figure 444

823

824

Association 0/ Myosinwith Actin Filaments

In electron micrographs of thin sections, myo sin filament s usually cannot beidentified in nonmu scle cells. However, their presence can be demonstrated withfluore scent anti myosin antibodies.

In the two cells illustrated here, antibody against platelet myosin delineates fibers0.5 to 1 fLm in width and up to several hundred micrometers in length . Thesecorrespond in distribution to straight fiber bundles visible in the same cells with thephase contrast microscope and interpreted as stress fibers. The staining is discontinuous , appearing as a series of bright spots at more or less regular inte rvals along thelength of the fibers.

Stress fibers staining in this manner with antimyosin are most prominent in cellsthat are widely spread on the sub strate and judged not to be actively motile. In cells thatare actively moving at the time of fixation , myosin staining is diffuse throughout thecytoplasm. In dividing cells, it is highly concentrated in the mitotic spindle atmetaphase and in the vicinity of the contractile ring during cytokinesis (Fujiwara andPollard, 1976). Thus, the distribution of myosin varies with the activity of the cell,sometimes being clearly associated with the actin in stress fibers, and at other timesdiffuse or associated with configurations of actin that are not visible with the light microscope.

Figure 44 5. H eLa ce l1 s in culture stained with rh odamine-conjugated immune globulin. (Photo micrograph co urt esy of Keigi Fuji wara .) Figure 44 5

825

CYTOPLASMIC MATRIX AND CYTOSKELETO N 827

MICROVILLI MUSCLE

Relationship 0/Actin Filamentsto the Cell Membrane

z-Line

Actinfilaments

Actinfilaments

A mod el prop osed for the func tiona l organiza tion of ac tin and myosin in the intesti nal brus hbo rde r (left ) co mpa red with that of striated mus cle (right). Sho rte ning is att ributed to slid ing of ac tinfilame nts with respect to myosin in the te rmin al web . T he lateral projection s from filam en ts show n her erepresent cross bridge s projec ting from intact myosin molec ules and not the arrowhea ds formed onac tin filaments ex posed in vitro to the S·I fragment of myosin . (Modified afte r Mooseker and Tilney,J. Ce ll BioI. 67 :72 5-743, 1975.)

When decorated with the S-I fragment of myo sin (he av y me romyosin), thearrowhead pattern is in the same direction on all actin filaments and indicate s a polar ityaway from the membrane at the villus tip . These finding s are co nsiste nt with thehypothesis that mov ement of micro villi result s from a sliding mechani sm in which actinfilaments anchored to the villus tip and to the lateral membrane interact with myosinand accessory prot ein s in the terminal web to produce shortening in a mannercompar able to that res ponsible for contraction of ske leta l muscle (Moose ke r andTilney, 1975).

It wa s originally spec ulated that the den se matrix at the villus tip and the crossbridges attaching the filament core to the membrane might consi st of o -actinin, aprotein found at the Z-line of skeleta l muscle where it is bel ieved to cro sslink the actinfilaments (Mooseker and Tilney, 1975). However, immunofluorescence micro scopywith antibody agains t o-actinin has failed to loca lize thi s protein in the microvilli .Instead , another polypeptide of 95,000 dalton molecular weight, ca lled villin, has beenpurified from isolated microvillu s cores (Brets chner and Weber , 1979). Indirectevidence sugges ts that it is thi s protein , peculiar to the micro villi of the bru sh bord er ,which links the actin cores to the inner surface of the membrane .

The highl y ordered actin filament cores of the microvilli appear to lack ass ociatedproteins oth er than villin, but myosin, tropom yosin , filamin , and o-act inin ca n all belocali zed in the terminal web region . Thu s the terminal web not only provides struc turalsupport for the microvillu s cores but its filamentous component s are also believed to becapable of interacting with them to produce shorte ning of the microvilli .

....': ~.

• __ . .. Myosinfilament s

1.11

( ,, ( »»-----

~:~~: ')', 'I\.») i: II;terminal web

« ( <: ,

For the interaction of ac tin and myosin to bring about motility or change in cellshape, one or the oth er of these components must be attached either directl y orindirectly to the surface membrane. There are several examples in electron micrographs where microfil aments appear to end at the plasmalemma. In cardiac muscle , theactin filaments of the myofibril s terminate at the intercalated discs in a den se layerben eath the sarco lemma. In smooth muscle , bundles of act in filaments end in similarden sities in the peripheral cy to plas m. The microfilaments of the terminal web in variousepithelia end in the den se component of the zonula adherens . In all of the se examples,the exact relationship of the filaments to the membrane is obscured by a layer of den seamor pho us mat eri al in th e subs urface cytoplas m. However, there is sugges t iveevidence from fre eze-fracture studies that the filaments reach the membrane and mayattach to some of its integral protein parti cle s (McNutt, 1978). Even in cell type s wherefilaments are not seen ending at the surface in electron micrograph s , actin is often foundto be ass oc iate d with the centrifugally isolated cell membrane fraction (Pollard andKorn , 1973).

Actin filaments ass oc iated with the plasma membrane were first clearly identifi edin the brush border of intestinal epithelial cells (Ishikawa et aI., 1969) and isolated bru shborder s have since been studied in considerable detail. A bundle of longitudinalfilaments that ca n be decorat ed with heavy meromyosin is found in the core of eachmicrovillus . At the tip , the filaments end in a layer of den se mat erial ass ociated with theinner surface of the membrane. At high magnifications , slender cross brid ges can besee n at 33 nm intervals extending laterally from the filament core to the membrane. Atthe base , the actin filaments extend downward into the apical cytoplasm, where somefan out and mingle with the transver sely ori ented filaments of the terminal web .

826


The accompanying micrograph s of longitudinal and transverse sections of inte stinal bru sh border show the actin filaments that form the core of each microvillus. At thevillus tip , the relation of the filaments to the membrane is obscured by a layer of den seamorphous material associated with the inner aspect of the membrane .

Fi gures 44 6 and 447. Microvilli of the bru sh bo rde r of car inrestina l epithelium. (Micrographs courtesyof Susumu Ito. )


829

Figure 44 8 A an d B . Elect ron micrograph of isolated bru sh bo rde r fro m chicke n int estin e. (FromMoosek er and Tiln ey,]. Cell BioI. 67:725, 1975.)

Figure 449A and B. Isolated bru sh border from chicke n, decorated with heavy meromyosin and fixed inthe presen ce of tanni c acid . (From Begg, Rodewald and Rehbun,]. Ce ll Bio I. 79:846, 1978 .)

The termination of the filaments in a density at the villus tip is well shown in aheavily sta ined thin section (upper left). At higher magnification (upper right) slendercro ss bridge s can be seen at 33 nm intervals along the length of the filamentbundle , att aching it to the limiting membrane of the villus. The actin filaments extendinto the apica l cytopl asm , where a few of the transverse microfil aments of the terminalweb are visible .

The polarit y of the actin fil aments is clearly demonstrated in the lower figuresshowing bru sh borders exposed to heav y meromyosin and fixed in the pre sence oftanni c acid. The arrowhead pattern of filament decoration is con stant in direction on allfilaments and indicates a polarit y away from the membrane.


Figure 44 8, upp er Figure 449 , lower

831

Ten NanometerCytoplasmic Filaments

In addition to the 6 nm actin microfilaments and the 25 nm microtubules, most celltypes also contain long unbranched cytoplasmic filaments about 10 nm in diameter. Thedistribution and composition of these intermediate-sized filaments have received muchinve stigative attention. Unfortunately they have been given different names in severalof the tissues in which they have been studied . Elucidation of their chemical nature hasprogressed slowly and major discrepancies persist concerning their constituent polypeptides.

In neurones, where they are called neurojilaments, they appear tubular in crosssection with a 2.5 to 3.5 nm light core (Wuerker, 1970). This ultrastructural feature isless apparent in the so-called glial filam ents of astrocytes. A subunit of glial filamentscomigrates on gels with one of the major neurofilament proteins , has a similar aminoacid composition, and exhibits some immunological cross reactivity (Dahl and Bignami ,1976; Yen et a!., 1976; Schlaepfer, 1977). Nevertheless significant differences arebelieved to exist between the two classes of intermediate filaments in the centralnervous system.

The 10 nm filaments of smooth muscle form an integrated cytoskeletal networkheld together at nodal points by the amorphous material of the dense bodies that arecharacteristic of these cells. The major component of the filaments is a 50,000 daltonprotein named skeletin (Small and Sobierzek, 1977). A subunit of similar molecularweight from intermediate filaments of avian skeletal and cardiac muscle has been calleddesmin (Lazarides and Balzer , 1978). Filaments that comprise the tonofibrils ofstratified squamous epithelia and many other epithelial cell types consist of kerat in, aclass of a-helical polypeptides of molecular weights ranging from 46,000 to 68,000daltons (Culbertson and Freedberg, 1977; Sun , Shik and Green , 1979). The majorprotein of the 10 nm filaments in several fibroblast cell lines has a molecular weight ofabout 58,000 daltons and has been named vimentin from the Latin for wicker or lattice(Franke et al . , 1978; Hynes and Destree, 1978).

There are many inconsistencies and unsolved problems in thi s rapidly advancingarea of cell biology and the terminology is confusing. The number of chemically distincttypes of intermediate-sized filaments remains unclear. The majority of those studiedhave subunits with molecular weights in the range 50,000 to 58,000 daltons. It appearslikely that there are at least three and possibly four immunologically distinct types: onecomprising the tonofibrils of epithelia (keratin), one typical of mesenchymal cells(vimentin), and another found in muscle (desmin, skeletin) . Those in nervous tissuemay constitute a fourth category. Although some cross reactivity has been reported,the bulk of the evidence indicates that neurofilament proteins are not identical withthose of other intermediate-sized filaments. It seems reasonable to assume that allcytoplasmic filaments in this size range have some degree of similarity in their primarystructure and subserve a common cytoskeletal function.

833


The 10 nm filaments follow a straight or gently curving course through thecytoplasm . They may occur individually or associated in loosely organized bundles.Shown here is their typical appea rance in a cultured 3T3 cell . It is not possible to gainfrom electron micrographs an impression of the overall distribution of the se filamentswithin the cell. For thi s purpose use of fluore scent antibodies and light microscopy ismore useful. Several examples of thi s approach are given in the figure s that follow .

Figure 4 50. Cytopl asm of 3T 3 ce ll showing typical 10 nm filaments coursing th rou gh [he meshes of [heendo plasmic reti cul um. (Micrograph courtesy of Scorr McNun , ]. Cell BioI. 50:69 1-708, 1971.)

Figure 450

835


Antibodies raised in guinea pigs against a 57,000 dalton protein extracted from thecytoskeleton of 3T3 fibroblasts selectively bind to the 10 nm filaments of severalmesenchymal cell lines. As previously mentioned , this cytoskeletal protein has beennamed vimentin (Franke et aI. , 1978). The upper micrograph of the facing page illustrate sindirect immunofluorescence with antivimentin in cultured cell s of mesenchymal originderived from the wall of a vein . These filament bundles are not stained by antibodyraised agains t keratin.

The lower figure is a cell from kangaroo rat ren al epithelium (Strain Pt 1\ 2) exposedto antibody against keratin . The con spicuous cytoskeleton in the se epithelial cell s is notsta ined by antibody again st vimentin . The cell s of muscle and brain show no significantfluore scence with either of the se antisera . It is concluded that although the functionallyrel ated 10 nm filaments in va rious cell type s have much the same ultrastructuralappearance and similar solubility properties, the y are immunologically distinct andtherefore probably contain different polypeptides.

Figure 4 51. Immunofluorescence ph oromi crograph of cultured rat rnyofibrobl asts fro m vein wall reactedwith ant ibody against virne ntin.

Fi gure 4 52 . Immunofluorescen ce ph oromi cro graph of a Pr K2 ce ll rea cted with ant ibody againstprek e rat in. (Micrographs from Franke et al., Proc. N at. Acad . Sci. 75:5034-5038, 1978.) Figure 45 1, upper Figure 452, lower

837


Electron micro scopy of thin sections provides little information about the threedimensional arrangement of cytoskeletal elements. By immunofluorescence , it ispossible to compare the distribution of microtubules and 10 nm filaments in cell s spreadupon a tissue culture substrate . In the accompanying photomicrographs , culture s of thesame stra in of ham ster cells were stained with antitubulin (upper figure) and withantibody agains t a 58,000 dalton protein extracted from 10 nm filaments (lower figure).

The microtubules radiate outward from the cell center, extending to the very edgeof the flattened cell periphery . The 10 nm filaments exhibit a rathe r similar fibrillarpattern , but the y are mainly in the thicker central portion of the cell and seldom ex tendto its thin periphery. The patterns may vary somewhat in different cell strains .

More distinctive than the pattern of distribution of these two cytoskeletal elementsis the ir different response to tre atment with colchicine. Dissolution of microtubulesafter expo sure to the drug elimin ate s most of the fibrillar fluore scence with antitubulin.The 10 nm filaments persist and aggregate into conspicuous thick coils around oradjacent to the nucleus.

Figure 4 53. Fluorescen ce photomicrograp h of hamster N lL 8 ce lls exposed to anr itub ulin .

Figure 454. Photomicrograp h of hamster N IL8 cells stained with antiser um to 58 K protein of 10 nmfilaments. (Figures from H ynes and Destree , Ce ll 13:l 5l-1 63, 1978. )

Figure 453, npper Figure 454, lower

839

840

Keratin Filaments

Cells of the epidermal epithelium in all species are rich in cytoplas mic filaments.The accompanying micrograph illust rates their extraordinar y abunda nce in a basal cellof lamp rey ski n. Keratin filaments are about 9 nm in diameter and appear to cons ist ofseveral differ ent polypeptide subunits with molecul ar weights in the ra nge of 40,000 to63,000 daltons (Matoltsy , 1976; Brysk , Gray and Bern stein , 1977). Keratin filamentscan be assembled in vitro from thi s group of polypeptides (Steinert and Guillino , 1976;Ste inert et aI., 1977). Fluorescent antibodies against a pur ified keratin fraction fromhuman skin sta in inten se ly the epidermal ce lls of seve ral ot her spec ies , indicati ng thatthey all possess homologous amino acid sequences. The ce lls contain a fluore scentnetwork of fibers corresponding to aggregations of tonofi laments seen in electronmicrograph s of the sa me cells. Application of antikeratin to a va riety of different cellstra ins of mesench ymal origin yields negati ve re sults , indi cating that the IO nmfilaments of fibroblast s and other mesenchymal deri vat ives are not kerat in (Sun andGreen , 1978).

Figure 4 55. Basal ce ll in the epide rmis of a larval lamprey, Petromyzon fl uviat ilis.Figure 455

841


At higher magnification, the keratin filaments appear as punctate or elongateprofiles depending upon their orientation in the section. They are very uniform in size ,approximately 9 nm thick, and may have a beaded appearance along their length asthough composed of repeating subunits. Among the filaments microtubules (at arrows)are occasionally seen.

Figures 456 an d 457. Basal cell from the epidermis of the amrnocoere of Petromyzon flu viatilis.Figure 456, upper Figure 457 , lower

843


Under certain circumstances, 10 nrn filaments may accumulate in high concentration. The supporting cells of the seminiferous epithelium have a normal complement offilaments and microtubules throughout the spermatogenically active portion of theseminiferous tubules , but near their junction with the tubuli recti and rete testis, wherethere are very few germ' cells, the Sertoli cytoplasm in some species is crowded withfilaments. Numerous microtubules are also found among the 10 nm filaments (atarrows). The physiological significance of thi s striking local increase in the concentration of cytoskeletal elements is not understood.

Figure 4 58 . A portion of the nucleus and adjace nt cyto plasm in a monkey Sertoli cell fro m the terminalsegme nt of the seminifero us tub ule. (Microg raph co urt esy of Martin D yrn.)

Figure 458

845


Neurofilaments are about 10 nm in diameter and of indefinite length. Unlike othercytoplasmic filaments of comparable dimension s, the y appear in transverse section tohave a clear ce nter sur rounded by a 3 nm thick den se wall (Peters and Vaughn , 1967;Weurker and Palay, 1969). They have therefore been interpreted as tubular. A singleprotein subunit called fil arin has been reported by various authors to have a molecularweight bet ween 50,000 and 60,000 daltons (Shelanski, 1973; Davison and Winslow,1974). Neurotubules of invertebrate s are reported to differ from those of vertebrate s inappearance and in the molecular weight of their subunits .

The accompanying micrograph shows axoplas m in a cell from the ciliary ganglioncontaining abundant neurofilaments oriented, for the most part , par allel to the long axisof the axon.

Figure 4 59. Axon fro m the ciliary ganglion of a chicke n. (Micrograph courtes y of Enrico Mugnaini.)Figure 459

847


The processes of neuroglial cell s in the nervous sys tem of both invertebrate s andvertebrates are crowded with filaments that closel y resemble the tonofilaments ofepithelial cell s. They are generally oriented parallel to the long axis of the cell process ,and some of them, like tonofil aments of epithelia, terminate in de smosomes.

Figure 460. Axial nerve cord of [he annel id Aphrodite acaleata ,Figure 460

849


A prominent feature of one class of as trocytes is the presence of large numbers of 9to 10 nm cytopl asmic filaments occurring throughout the perikaryon and extending intothe cell processe s. They are said to be slightly smaller than neurofilaments and insteadof being distributed individually , they tend to be aggregated in co nspicuous bundles . Inthe classical neurocytological methods for demonstration of astroglial filaments , goldwa s deposited on filament bundles and not upon the individual filament s.

A large bundle of clo sely packed filaments coursing along a mitochondrion in a glialcell process is shown in the accompanying micrograph.

Figure 46 1. Filamenrs in an asrrocy re from rhe do rsal coc hlea r nucleus of rat brain . (Micrograph courresyof Enri co Mugnaini .)

Figure 461

851


REFERENCES

Microtubules

Barclay , N . E . Marginal band s in duck and camel erythro cy tes. Anat. Rec . 154:313, 1966.Behnke , O . Incomplete microtu bule s observed in mamm alian blood platelets durin g polymerization . J .

Ce ll BioI. 34:697-7 01, 1967.Behnke , O. Microtubules in disc-shaped blood ce lls . Int. Rev. Ex p. Path ol. 9: 1-92 , 1970. (Rev iew)Bikle , D. , L. G. Tilney and K. R. Por ter. Microtubules and pigment migration in the melanoph ores of

Fundu lus heteroclitus, Protoplasma 61:322-345, 1966.Bind er , L. I. and J . L. Rosenbaum . The in vit ro assembly of flagellar out er doubl et tubulin . J . Ce ll BioI.

79:500-5 15, 1978.Borisy, G. an d J. Olmsted. Nuc lea ted assembly of micr otu bul e s in porcine brain extracts. Sc ience

177: 11 96-1 197, 1972.Bor isy , G. G., J . M. Marcus , J. B. Olmsted , D. b . Mur ph y and K. A. Johnson . Purificat ion of tubu lin

and associated high molecul ar we ight protein s fro m porci ne brain and ch aracterization of microtubuleassembly in vitro . Ann . N .Y. Acad. Sc i. 253: 107-11 9, 1975.

Burnside , B. The form and arra ngemen t of microtubules: An historic al , pr imarily morphological re view.In Ann. N.Y. Acad . Sci. 253:14-26,1 975. (Review)

Byers , B. and K. R. Porter. Orient ed microtubules in elongating cells of the developing lens rud imentafter induction . Proc. Na t. Acad . Sc i. 52: 1091-1099, 1964.

Ca nde, W. Z. , E. Lazarid es and M. J. McIntosh. A co mpa rison of the distribution of ac tin and tubu lin inthe mamm al ian mitotic spindle as seen by indirect immunofluorescence. J . Ce ll BioI. 72:552- 567,1977.

Co nnolly, J. A., V. I. Kalnin s , D. W. Cleve land and M. W. Kir schner. Intracellular localizat ion of highmolecular weight microtu bul e accessory protei n by indirect immu nofluorescen ce . 1. Cell Bi oI.76:R781- R786, 1978.

Dehler, A. Beitrage zur Kenn tnis des feinerem Bau es del' rote n Blutkorperch en beim Huhnerernbryo.Arch Mikr. Anat. 46:414-430, 1895.

Fawcett, D. W. and F. Witeb sk y . Ob servati on s on the ultr a st ructure of nucl eated ery thro cytes andth romb ocyte s, with particular refere nce to the structura l basis of their di scoidal sha pe. Z. Zellforsc h.62:785- 806, 1964.

Freed , 1. J . and M. M. Leb owitz. The asso ciation of a class of sa lta to ry movem en ts with micr otubules incultur ed cell s. J . Ce ll BioI. 45:334-354, 1970.

Harri s, P . Som e structura l and functiona l as pec ts of the mitoti c apparat us in sea urchin embryos. J. CellBioI. 14:475, 1962.

Heggene ss , M. H., M. Simo n and S. J. Singer. Associati on of mitochondria with micro tubules in cul tu redce lls . Proc. Na t. Acad . Sci . 75:3863-3866, 1978.

Inou e , S. The effect of co lchicine on the microscopi c and submicroscopic struct ure of the mitoti c spindle.Ex p. Ce ll Res. Suppl. 2:305- 318, 1952.

Inou e , S. and H . Sato. Ce ll mot ility by labil e associat ion of molecul es. The nature of mitoti c spindlefiber s and thei r role in chromoso me movem ents. J. Gen . Physiol. Suppl. 50:259-288, 1967.

Junq uier a , L. C.; E . Raker and K. R. Port er. Studie s on pigment migration in the melan oph ores of theteleost Fundulus heteroclitus. Arch. Histol. Jpn. 36:339-366, 1974.

Kim , H., L. I. Bind er and J . L. Rosenbaum . The periodic associ ation of MAP 2 with brain microtubule s invitro. J . Ce ll BioI. 80:266-276, 1979.

Ledbette r , M. and K . R. Porter. A microtubule in plant ce ll fine st ruct ure. J. Cell BioI. 19:239-250, 1963.Lockwoo d , A. H. Tubulin assembly pro tein : Irnrnunochernical and immunofluor e scen t st udies on its

funct ion and distribution in microtubules and cultured cell s. Ce ll 13:613- 627, 1978.McI ntos h, J . R . et al. Mo de l for mito sis. Nat ure 224:659-663, 1969.Me ve s, F r. Die Hlinefeld-Hen senchen Bilder del' rot en Blu tk orperch en de l' Amphibien. Ana t. Anz.

24:465-476, 1904.Meves, Fr. Gesammelte Studien an den roten Blutkorperchen del' Amphibien. Arch . Mikr. Anat. 77:465

540, 1911.Murph y, D . B. and G . G. Bori sy . Association of high molecular weight protein s with microtubules and

their role in microtubule asse mbly in vitro. Proc. Na t. Acad . Sci . 72:2696-2700, 1975.Murph y, D . B. and L. G. Tilney . The ro le of microtu bules in the movem ent of pigment gra nules in te leost

melanoph or es. J . Ce ll BioI. 6 1:757-779 , 1974 .Mu rph y, D. B. , R. B. Vallee and G. G. Borisy. Ident ity and polymeri zation-stimulatory ac tivity of the

non tubulin protein associated with microtubul es. Bioc hemist ry. 161:2598-2605, 1977.Olmsted , J . B. and G. G. Boris y. Micro tubules. Ann. Rev. Biochem . 42:507-540, 1973. (Re view)Pease , D . C. Th e ultrastruct ure of flagellar fibrils . J . Ce ll BioI. 18:313,1 963.Porte r, K . R. Motility in ce lls. In Cell Motility, Book A (R. Goldma n, T. Pollard and J . Rosenb aum ,

eds .), pp. 1-28, Cold Spring Harb or Confere nces on Cell Prolifer at ion , Vol. 3, 1976.Ran vier , L. Recherches sur les elemen ts du sa ng. Arch. Phys . D.S. 2:1-15, 1876.Roth , L. E. and E. W. Dani els . Elect ron microsc opic studies of mito sis in Ameba e . II. The giant ameba

Pelomyxa carolinensis . J. Ce ll BioI. 12:57, 1962.

CYTO PLASMIC MATRIX AND CYTOSKELETO N

Schliwa, M. and J . Bereiter-Hahn. Pigment movem ent s in fish melanopho res: Morphological and physiologica l st udies . II . Ce ll shape and microtubule s. Z. Zellfor sch. 147: 107-1 25, 1973.

Schliw a , M., M. Osborn and K. Weber. Microtubule system of isolated fish melanophore s as revealed byimmunofluore scence microscopy. J . Ce ll. BioI. 76:229-236, 1978.

She lans ki, M. L. and E . W. Taylor. Isol ati on of a protein sub unit from micr otub ules. J. Ce ll BioI.34:549-554, 1967.

Slau tterb ack , D . L. Cyto plas mic microtubules. I. Hydr a. J. Ce ll. BioI. 18:367-388, 1963.Sloboda , R. D. , W. C. Dentler and J . L. Rosenbaum. Micro tu bule-associated protein s and the stimulation

of tubulin assembly in vitro . Bioch emi stry 15:4497-4505, 1976.Smith, D. S. , U . Jarlfor s and R. Beranek. The orga nizat ion of sy naptic axoplasm in the lamprey cen tral

nervous sys te m. 1. Ce ll BioI. 46: 199-219, 1970.Smith , D. S. On the significa nce of cross -bridges between microt ubul es and synaptic ves icles . Phil. Tr ans.

Roy. Soc . B. 261:395-405,1 971.Smith, D. S. , U. Jarlfor s and M. L. Cayer. Structu ral cross -br idges between microtubules and mitochon

dria in ce ntra l axo ns of an insect (Periplaneta ame ricana). J. Ce ll Sci. 27:255-272 , 1977.Stephe ns , R. E . Micro tubules. In Biological Macromolecule s (S.N. Timas heff a nd G. D. Fasman , eds .),

pp . 355-391, vol. 4, Dekk er , New York , 1971. (Rev iew)Steph en s , R. E . and K . T . Edds . Microtubules: Structure, chemistry a nd funct ion. Physiol. Rev. 56:709

777, 1976.Taylor, E . The mech ani sm of colchicine inhibiti on of mitosis. I. Kineti cs of inhibiti on and the bindin g of

H3-colchi cin e. 1. Ce ll BioI. 25: 145-160, 1965.Tilney, L. G. Origin and co ntinuity of microtubules . I n Origin and Continuity of Cell O rganelles (J.

Reinert and H. Ursprung, ed s.), pp. 222-259, Springer Verlag, New Yor k, 197 1. (Re view)T ilney , L. G . and K. R. Port er. Studies on microtub ules in Heliozoa. I. Protoplasma 60:3 17, 1965.Tiln ey, L. G. and K. R. Porter. Studies on the microtubules in Heliozoa. II. T he effec t of low temp erature

on the form ation a nd maint en ance ofaxopodia. J. Cell. BioI. 34:327-3 43, 1967.Weidenreich, Fr. Studien libel' das Blut und die blutbildend en und - zerstor ende n Organe . III Uber den

Bau del' Amphibienerythrocy ten. Arch. Mikr. Anat. 66:270-288, 1905.Wein garten , M. D. and A. H . Lockwood , S-Y. Hw o and M. W. Kir schner. A pro tein factor essenti al for

microtubule asse mbly . Proc. Na t. Acad. Sci. 72: 1858-1 863, 1975.Wei senbe rg, R . Mic rot ubule for mat ion in vit ro in so lutions co nta ining low ca lcium ion s . Science

177: 1104-1105, 1972.Weisenb erg, R. C., G. G. Bori sy and E. W. Taylor. The colchi cine binding protein of mamm alian brain

and its relat ion to microtubules. Bioc he mistry. 7:4466-4 479, 1969.Wilson , L. and J. Bryan. Biochemical and pharm acolo gical proper tie s of microtu bules. Adv. Cell Molec .

BioI. 3:22-7 1, 1973. (Rev iew)

Cytoplasmic Filaments

Filaments of Muscle Cells

Ast bury, w. T. and S. Dickinson . X-ray studies of the molecul ar st ructure of myos in. Proc. Ro y. Soc.Lo nd . sB 129:307, 1940.

Bruche, E. Un ter suchungen libel' den Bau del' Muskelfasern mit Hiilfe des polari sitren Lichtes. Den kschr.Kaiserl. Aca d . Weis. Math. - Natur ev. 15:69, 1858.

Cohe n, C. The protei n switc h of muscle contracti on . Sc i. Am. Dec . , pp. 36-45 , 1975.Hall , C. E., 1. A. Jaku s, and F. O. Sc hmitt. An investigation of cross str iations and myosin filament s in

muscle. BioI. Bull. 90:32, 1946.Hu xley, H . E . Mu scle ce lls. In The Cell: Biochemistr y, Physiology, Morphology (J. Brachet and A. E.

Mirsky , eds.), p. 365, vol. 4, Academic Press, New York , 1960. (Rev iew)Hu xley, H . and J . Han son. Changes in the cross -striat ions of muscle duri ng co ntrac tion and stretch and

thei r structura l interpret ation s . Na ture 173:973-976, 1954.Hu xley, H . E. and J. Han son . The molecul ar basis of co ntraction in cross -striated muscles. In Stru cture

and Function of Musc le (G. H. Bou rne , ed .) p. 183, vol. I, Aca de mic Press, New York , 1960.Kelly, R. E . and R. V. Rice. Loc alizat ion of myosin filam ent s in smooth muscle . J. Cell BioI. 37:105-11 6,

1968.Low y, 1. , F. R. Poul sen and P. J. Vibert. Myo sin filaments in ve rteb ra te smoo th muscle . Nature

225: 1053-1 054, 1970.McD ou gall , W. On th e st ructure of cross -st riated mu scle and a suggestion as to th e nature of its

contrac tion. 1. Anat. Ph ysiol. 31:410, 539, 1897.McDougall , W. A theo ry of muscle co ntraction. J . Anat. Physiol. 32: 187, 1898.Meigs , E. B. Contributions to the genera l ph ysiology of smooth and striated muscle . J. Ex p. Zool. 13:497,

1912.Roll et , A. Un te rs uchunge n libel' den Bau del' quergestreiften Muskelfasern . Wien . Math .-Nat. Classe

49:81,1 885.

853


Small, J . Y. Studies on isolated smoo th muscle cells: The co ntrac tile apparatus. J . Ce ll Sci . 24:327-349,1977.

Small, J . Y. and J . M. Squire. Structura l basis of contrac tion in vertebrate smoo th muscle . J. Molec .BioI. 67: 117- 149, 1972.

Small, 1. Y. a nd A. Sobi e szek . Studies on the fun ction a nd co mpositio n of the 10 nm filam en ts ofvertebrate smoo th muscle . J . Ce ll Sci. 23:243-268, 1977.

Somlyo , A. P. St ruct ura l cha racte ristics and mechani sm of co ntr act ion, inner vation and proliferation ofsmooth muscle ce lls . Ultras tructure and function of vasc ular smooth muscle. Adv. Ex p. Med . BIOI.57: 1-80, 1975.

Sornlyo , A. P., C. F . De vine , A. Y. So mlyo and R. Y. Rice . Filame nt organiza tion in ve rte brate smoot hmuscle. Phil. Trans . Roy. Soc. B 265:223-229 , 1973.

Squire , 1. M. Ge neral model of myosin filame nt structure . III. Molecular packing arra ngements in myosinfilame nts . J . Mol. BioI. 77:29 1-323 , 1973.

Weber, A. and J . M. Murray . Molecul ar co ntrol mech ani sms in muscle co ntract ion. Ph ys . Rev. 53:6 12,1973. (Rev iew)

Filaments of Cells Other than Muscle

Begg, D. A., R. Rodewald and L. I. Rehbun . The visualization of ac tin filame nt polarity in thin sec tions .J . Ce ll BioI. 79:846-852 ,1978.

Bignami , A. and D. Dahl. Astrocyte-specific protein and neuroglial differenti ation. An immunofluorescence study with antibodies to the glial fibrillar y acidic prote in . J . Compo Neuro l. 153:27- 38, 1974.

Bret schner , A. and K. Web er. Yillin: T he majo r microfil ament ass oc iated prote in of the inte sti nalmicrovi llus . Proc. Nat. Acad . Sci . 76:2321-2325, 1979.

Cooke, P. H. A filame ntous cytoskeleton in verteb rate smooth muscle fibers. J . Cell BioI. 68:539-556 ,1 ~6 . '

Dah l, D. and A. Bignam i. Immunogenic prop erti es of the glial fibrillary ac id protein . Brain Re s. 116: 150157, 1976.

Davison , P. F . and B. Winslo w. The protein subunit of calf brain neurofilam ent. J . Neurobiol. 5: I19- 137,1974. , •

F ranke , W. W., E. Schmidt, M. Osborn a nd K. Weber. Differ ent intermedi ate sized filament s distinguished by immunofluorescence microscop y. Proc . Nat. Acad . Sci . 75:5034-5 038, 1978.

Franke , W. W., B. Appelha us , E. Schmid a nd C. Freud en steiij. The organization of cy tokeratin filam entsin the inte stinal epithelium. Euro pea n J . Ce ll BioI. 19:2 55~2?8 , 1979.

Franke , W. W., C. Grund , E. Sc hmid and E. Mande lkow. Paracr ystalline arrays of membrane-tomembrane cross brid ge s associated with the inner surface of plasma mem brane . J . Ce ll BioI. 77:323,1978.

Fuj iwara , K. and T. Pollard . Fluorescent a ntibody localizati on of myosin in the cy to plas m, cleavagefur row, and mitot ic spindle. J . Ce ll BioI. 71:848-875 , 1976.

Hynes, R. O. and A. T. Destree . 10 nm filame nts in normal and transformed ce lls. Ce ll 13:151- 163, 1978.Ish ikawa , H., R. Bischo ff and H . Holt zer. Fo rmation of arrowhead complexes with heavy meromyosin in

a variety of ce ll type s. 1. Ce ll BioI. 43:312- 328, 1969.Lazarid es, E . Two general classe s of cy to plasmic actin filam en ts in tissu e cultu re ce lls: The role of

tropom yosin . J. Supramo l. Stru cture 5:531-563, 1976.Lazarid es , E . The distribution of desmin (100 A) filaments in prim ar y cultures of embryonic chick ca rdiac

ce lls. Ex p. Ce ll Res. 112:265- 273, 1978.Laza rides , E. and D . R. Balzer. Specificity of de smi n to av ia n a nd mammalian mu scle cell s. Ce ll

14:429-438 , 1978.Lazarides, F., and K. Weber. Actin antibody . T he specific visualization of actin filame nts in no n-mu scle

cells . Proc. Na t. Aca d. Sci. 7 1:2268, 1974.Matoltsy , A. G . De smosomes , filaments, a nd keratohyaline gra nules: The ir role in the stabilization and

kerat inization of the epidermis. J. Invest. Dermatol. 65:127- 142, 1975.Maupin-Szamier , P. and T . D. Pollard . Actin filament destruc tion by os mium tetroxide . J . Ce ll BioI.

77:837- 852, 1978.McNutt , N . S. A thin sec tion a nd freeze-fracture study of microfil am en t-rnernbrane atta chme nts in

choro id plexus and inte stin al micro villi . J . Cell BioI. 79:774-787 , 1978.Mooseker , M. S. Brush border mot ility: Microvillar contraction in isolated bru sh-border mode ls . J. Ce ll

BioI. 63:23I a , 1974.Moosek er , M. S. Bru sh border motilit y. Microvillar contrac tion in Trito n-t rea ted bru sh border s isolated

fro m inte stinal epithelium. J. Ce ll BioI. 71:417-433 , 1976.Moose ker , M. S. a nd L. G. Tiln ey . Organi zati on of an ac tin-filame nt mem bran e co mplex . Filame nt

polar ity and membrane attachment in the micro villi of intestin al epithelial ce lls. J . Ce ll BioI. 67:725-743,1975.

Mukh erjee , T . M. and L. A. Stae helin. The fine struc tural orga nizat ion of the brus h border of inte stinalepithelia l ce lls. 1. Cell Sci . 8:573, 1971.

Pollard , T . D. F unctional implicat ions of the biochem ical and str uct ural proper tie s of cy toplasmic co ntractile protein s. In Molecule s a nd Ce ll Movem ent (S. Inoue and R. E. Steph en s, eds .), pp . 259-286,Raven Press, New Yo rk , 1975.

CYTOPLASMIC MATRIX AND CYT OSKELETON

Poll ard , T . D. and E . D . Korn . Electro n microscopic identificati on of ac tin associated with isolatedamoe ba pla sma membranes. 1. BioI. Chern. 248:448-4 50, 1973.

Poll ard , T. D . and R. R. Weihing. Actin a nd myosin in cell movem ent s. Cr it. Rev. Biochem. 2:1, 1974.Port er, K. R. and J . J . Wolosewick . Ster eo high voltage elec tro n microscopy of whole ce lls of thehuman

diploid line WI- 38. Am . J . Anat. 147:303- 323, 1976.Schl aepfer , W. W. Immunological and ultrastru ctu ral studies of neurofilam ent s isolated from rat peripher

al ner ve . J. Cell BioI. 74:226-240, 1977.Schl aepfer , W. W. and R. G. Lynch . Immunofluorescence studies of neurofilam ents in the rat and hum an

periphera l and central nervou s sys te m. J . Ce ll BioI. 74:241-250, 1977.Schollmeyer , 1. Y., D. E . GoIl , L. G. Tilney , M. Moo seker, R. Rob son and M. Stromer. Localizat ion of

o-act inin in nonmuscle mate rial. J . Cell BioI. 63:304A, 1974.Shelanski , M. L. Chemi str y of the filame nts and tubules of brain . J . Histoch em . Cy toc hem. 21:529-539,

1973.Tiln ey , L. G . and M. S. Moo sek er . Act in filame nt membrane at tachme nt: Are memb rane particles

involved ? J . Ce ll BioI. 71:402-4 16, 1976.Weber , K . and U . Groe sc hel-Stewart . Ant ibo dy to myo sin : Th e specific visua liza tion of myosin

conta ining filame nts in non-musc le ce lls. Proc . Nat. Acad . Sci . 71:4561-4564 , 1974.Yen , S.-H ., D . Dahl , M. Schachtner and M. L. Shel anski. Biochemi str y of the filaments of the brain .

Proc . Nat. Aca d . Sci. 73:529 . 1976.

Kerat i n Filaments

Brysk , M. M. , R. H. Gray an d I. A. Bernstei n. Tonofi lame nt protein from newborn rat e pidermi s:Iso lation , loc alization and bio synthe sis of marker of epidermal differenti ation. 1. BioI. Chern .252:2127-2133, 1977.

Mato ltsy, A. G. Kerati nization . J . Invest. Derm atol. 67:20-25 , 1976.Steinert, P. M. and M. Guillino . Bovine epidermal ker atin filament ass embly in vitro. Bioc hem. Bioph ys.

Res. Commun . 70:221- 227, 1976.Steinert, P . M., W. W. Idler and S. B. Zimm erm an . Self-assembl y of bovine epid erm al keratin filament s

in vitro . J . Molec . BioI. 108:547- 567, 1977.Sun , T . and H . Green . Immunofluorescent sta ining of keratin fibe rs in cultured cells . Cell 14:469-476,

1978.

855

chapter 16: cytoplasmic matrix and cytoskeleton

Documents