AN ANALYSES OF THE CHEMISTRY AND FUNCTION OF MICROTUBULE PROTEIN AS RELATED TO. CELL DIYTSTON TN ACANTHAMOEBA RHYSODES W Thesis for the Degree of Ph. D ‘ MICHIGAN STATE UNIVERSITY ROBERT WARD mam ‘ 1971 M. LIBRARY Michigan Stat! Unchflity This is to certify that the thesis entitled AN ANALYSIS OF THE CHEMISTRY AND FUNCTION OF MICROTUBULE PROTEIN AS RELATED TO CELL DIVISION IN ACANTHAMOEBA RHYSODES presented by ROBERT WARD RUBIN has been accepted towards fulfillment of the requirements for .__Eh.._D..__ degree in £921.22;— ;?% @146 Major professor Date—Jil—lzJLlQZL 0-7639 ABSTRACT AN ANALYSIS OF THE CHEMISTRY AND FUNCTION OF MICROTUBULE PROTEIN AS RELATED TO CELL DIVISION IN ACANTHAMOEBA RHYSODES BY Robert Ward Rubin Techniques previously devised to induce synchronous mitosis and amitosis in the soil amoeba Acanthamoeba rhysodes were utilized to study the ultrastructure and chemistry of these events. It was found that colchicine would inhibit amitosis when used at the minimal concen- tration effective in inhibiting normal mitotic cell division. This suggested a functional role for micro- tubule protein in both processes. An extensive ultra- structural study revealed that the amitotic cell did not produce a mitotic apparatus although limited microtubule polymerization did occur. Seventy nm diameter actin-like fibers were also seen localized within the amitotic cleavage plane. The mitotic cell possessed typical ultrastructural characteristics with the exception of the presence of electron dense amorphous organizing Robert Ward Rubin centers acting as centrioles and the occasional locali- zation of microtubules parallel to and between sister chromatids at metaphase. An extensive biochemical study of a colchicine binding protein in these cells was then initiated. It was found that a pure protein could be isolated using the techniques previously devised for the isolation of brain microtubule protein. This protein possessed an identical amino acid composition to brain microtubule protein as well as an identical elution profile on Sephadex G-200. However, the amoeba protein lost its colchicine binding characteristics much more rapidly than the brain protein and gave a consistently lower Rf value on acrylamide gel electrophoresis. The amoeba protein also showed GTP binding but SDS gel electro- phoresis demonstrated consistently lower molecular weight values relative to the standard brain protein. The amoeba protein showed a turnover rate in synchronous and non-synchronous cultures equal to that of the total soluble protein pool. No regulation in synthesis of the protein relative to cell division could be detected. It is concluded that amoeba microtubule protein only functions in normal mitosis, possesses a number of characteristics which set it apart from previously de- scribed colchicine binding proteins and is synthesized at a continuous unregulated rate throughout the cell cycle. AN ANALYSIS OF THE CHEMISTRY AND FUNCTION OF MICROTUBULE PROTEIN AS RELATED TO CELL DIVISION IN ACANTHAMOEBA RHYSODES BY Robert Ward Rubin A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Zoology 1971 ACKNOWLEDGMENTS I would like to extend my appreciation to Dr. Neal Band for his consistent encouragement and academic help as well as Dr. Hironobu Ozaki for his assistance and advice in preparing this thesis. I would also like to thank Stew Pankratz for his invaluable aid during my endeavors in electron microscopy. Thanks should also go to Dr. Charles Thornton for his kind help throughout my stay at Michigan State. Lastly, I must acknowledge the late W. C. Fields whose attitudes on life and the nature of man have helped to provide me with the moral stamina to complete my formal academic training. This work conducted under the tenure of a National Institute of Health Grant to R. Neal Band (ROlAlOGll?) and a National Institute of Health Grant to the Depart- ment of Zoology, Michigan State University (T01HD00135). ii TABLE OF CONTENTS Page INTRODUCTION AND LITERATURE REVIEW . . . . . . 1 Composition and Function of the Mitotic Appara- tus. . . . . . . . . . . . . . . . 1 Composition of the Spindle Fiber Microtubules. . 5 Electron Microscopy. . . . . . . . . . 5 Isolation of Microtubule Protein . . . . . 6 Functions of Microtubules and Microfilaments in the Mitotic Process . . . . . . . . . . ll Metabolism and Chemistry of Microtubule Protein . 14 Research Goals . . . . . . . . . . . . 16 MATERIALS AND METHODS . . . . . . . . . . 20 Culture Techniques. . . . . . . . . . . 20 Culture Media. . . . . . . . . . . . 20 Induction of Amitosis . . . . . . . . . 20 Induction of Synchronous Mitosis . . . . . . 21 Filter Paper Assay for Colchicine Binding . . . 21 Flow Sheets for Assays of Major Cellular Fractions . . . . . . . . . . . . . 22 Electron Microscopy . . . . . . . . . . 22 Actin Assay . . . . . . . . . . . . . 29 Gel Filtration . . . . . . . . . . . . 30 iii Page Isolation of Colchicine Binding Protein . . . . 30 Electrophoresis . . . . . . . . . . . . 3O Amino Acid Composition Determination . . . . . 34 Turnover Experiments . . . . . . . . . . 34 RESULTS . . . . . . . . . . . . . . . 36 Effects of Colchicine Application . . . . . . 36 Inhibition of Amitosis . . . . . . . . . 36 Morphology of Treated Cells . . . . . . . 36 Ultrastructure of Mitotic and Amitotic Cells . . 41 Ultrastructure of Mitosis . . . . . . . . 41 Ultrastructure of Amitosis. . . . . . . . 52 Interphase Cells. . . . . . . . . . . . 74 Structure of the Dense Body in Interphase Cells. 74 Comparison of the Nuclear Membrane in Amitotic and Interphase Cells. . . . . . . . . . 74 Localization of the 7 nm Diameter Fibers . . . 79 Amoeba Actin Assay . . . . . . . . . . . 80 Demonstration of a Colchicine Binding Protein in Acanthamoeba Rhysodes . . . . . . . . . . 87 Assays of Major Cellular Fractions . . . . . 87 (a) Soluble and Insoluble Components. . . . 87 (b) Vinblastine Precipipate. . . . . . . 91 (c) Ammonium Sulfate Precipipate . . . . . 91 (d) Stability Over Time . . . . . . . . 92 Purity of Amoeba and Brain Tubulin. . . . . . 92 Comparison of Amoeba and Brain Tubulins . . . . 92 iv Electrophoresis . . . . . . . . . (a) 12 cm Gels. . . . . . . . . (b) 20 cm Gels. . . . . . . . . (c) SDS Electrophoretic Molecular Weight Determinations. . . . . . . . . Amino Acid Analysis. . . . . . . . Comparison of Specific Activities . . . Gel Filtration . . . . . . . . . (a) Brain Tubulin. . . . . . . . (b) Amoeba Tubulin . . . . . . . Turnover Experiments . . . . . . . . Non-Synchronized System . . . . . . Synchronized System. . . . . . . . Vinblastine Precipitation. . . . . . DISCUSSION . . . . . . . . . . . . Comparison of Mitosis and Amitosis . . . Role of Microtubules in Amitosis . . . Function of Microfibers . . . . . . Function of the Dense Bodies. . . . . Microtubule Polymerization and Nuclear Elon- gation . . . . . . . . . . . . Amoeba Actin. . . . . . . . . . . Chochicine Binding to Amoeba Protein. . . Comparison Between Amoeba and Brain Tubulin Electrophoretic Mobility . . . . . . Page 95 95 96 96 99 99 105 105 110 115 115 120 120 123 123 123 124 125 127 129 129 131 131 Page Molecular Weight Comparisons . . . . . . . 132 Amino Acid Analysis . . . . . . . . . . 132 Gel Filtration Column Chromatography . . . . 135 Turnover of Amoeba Tubulin . . . . . . . . 136 LITERATURE CITED . . . . . . . . . . . . 140 vi LIST OF TABLES Table Page 1. Analysis of variance of nuclear membrane dian‘eters O O I C O O O O O I O O 80 2. DEAE filter paper assays for colchicine bind- ing activities . . . . . . . . . . 88 3. Colchicine binding specific activity experi- ments 0 O O C I O O O O I O O O 104 4. Amoeba colchicine binding activity from Sephadex fractions . . . . . . . . . 115 vii Figure 1. 12. 13. 14. 15. 16. LIST OF FIGURES Flow sheet for colchicine binding experiments 1-3 0 o o o o o o o o o o o o o Colchicine binding protein isolation procedure for Acanthamoeba rhysodes. . . . . . . Inhibition of amitosis by .OOSM colchicine. . Random section through interphase cell. X 60,000 0 o o o o o o o o o o o Metaphase cell. x 11,000 . . . . . . . Cortex of metaphase cell. x 50,000 . . . . Periphery of metaphase cell. x 88,000 . . . Composite photograph of the mitotic apparatus of a metaphase cell. x 35,000 . . . . . Metaphase chromosomal alignment. x 84,000. . Dense body from metaphase cell. x 150,000. . Early amitotic cell (phase contrast). x 2,500. O I O O O O O O O O O O Mid-amitotic cell (phase contrast). x 2,500 . Early amitotic cell. x 11,500. . . . . . Early amitotic cleavage plane. x 40,000 . . Early amitotic cleavage plane in longitudinal section. x 59,000 . . . . . . . . . Longitudinal section through early amitotic thread. X 40,000 0 o o o o o o o 0 viii Page 24 32 38 40 43 45 47 49 51 54 56 58 60 62 64 67 Figure 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. Longitudinal section through early amitotic thread. x 35,000 I O O O O O O O 0 Transverse section through late amitotic thread. X 175,000. 0 o o o o o o 0 Longitudinal section through mid-amitotic thread. x 32,000 . . . . . . . . . Dense body from interphase cell in longitudi- nal section. x 164,000 . . . . . . . Interphase cell. x 20,000. . . . . . . Glycerine extracted cell. x 14,000. . . . 10,000 G pellet. x 87,000. . . . . . . Binding of rabbit myosin to a component in an amoeba 10,000 G supernatant. . . . . . Acrylamide gel electrophoresis of brain and amoeba tubulin preparations in BM urea . . Densitometer tracing of reduced and alkylated brain tubulin run on foot-long 8M urea polyacrylamide gels. Actual photograph of section of this gel is also shown. . . . SDS polyacrylemide gel electrophoretic molec- ular weight determination of brain and amoeba tubulins. . . . . . . . . . Amino acid analysis of brain and amoeba tubulin compared to those from neuroblas- toma cells published by Olmstead and Rosenbaum (1970) . . . . . . . . . Sephadex G-200 column chromatography of a brain ammonium sulfate precipitate that had been preincubated in H3-colchicine . . Molecular weight v.s. Kav curve for Sephadex G-200 taken from Andrews (1964) . . . . Sephadex G-200 column chromatography of an amoeba ammonium sulfate precipitate . . . ix Page 69 71 73 76 78 82 84 86 94 98 101 103 107 109 112 Figure Page 32. Sephadex G-200 column chromatography of an amoeba 200,000 G supernatant preincubated with H3-colchicine . . . . . . . . . 114 33. Turnover of amoeba tubulin and total soluble proteins in a non-synchronized culture . . 117 34. Turnover of amoeba tubulin and total soluble proteins in a non-synchronized culture . . 119 35. Turnover of amoeba tubulin and total soluble proteins in a synchronized culture. . . . 122 INTRODUCTION AND LITERATURE REVIEW Few biological processes are as common or as little understood as the phenomenon of mitosis. This eukaryotic process also may be one of the most highly conserved bio- logical events, from an evolutionary standpoint. This is suggested by the remarkable structural similarity of the mitotic machinery, such as the animal centrioles and the microtubules of the spindle fibers, as well as the similar stepwise sequence of events observed for prac- tically all cells. An understanding of the nature of the chemical and structural events which constitute cell division is also essential to the elucidation of such diverse phenomena as carcinogenesis, cell motility, growth and development. Composition and Function of the Mitotic Apparatus The study of the mode of action and molecular con- stituents of the mitotic apparatus has, over the last ten years in particular, absorbed the interest of numerous researchers (Mazia, 1961; Zimmerman, 1963; Sakai and Dan, 1959; Went, 1966; Miki-Noumura, 1968; Bibring and Baxandall, 1969; and many others). Many workers have concentrated on biochemical changes associated with the formation, contraction and dissolution of the visible mitotic apparatus. Mazia (1952, 1961); Kawamura and Dan, (1958) and Stern, (1958), using either 30% ethanol or the less harsh dithiodiglycol technique to stabilize the mitotic apparatus, found this structure to be remarkably homogeneous. They concluded that it was composed of a single protein of high sulfur content plus smaller amounts of RNA, polysaccharides and lipids. This protein was thought by these workers to undergo reversible sulfhydryl redox reactions accompanying chromosome movement. Went (1966) reviews further studies indicating the possible role of alternate oxi- dation and reduction of the mitotic apparatus proteins as the causative factor in chromosomal movement. Recent studies (Sakai, 1966; Kiefer, Sakai, Solari and Mazia, 1966) discussed in succeeding pages, indicate that this protein is probably the microtubule protein of the spindle fibers and it is suggested that electron transfer reactions within the monomers of this system may account for contractions. Taking a different line of investigation concerning the role of the mitotic apparatus, Kawamura (1960), Davis and Smith (1957), Zirkle (1957) and Hiramoto (1956, 1965), using UV microbeam or microsurgical techniques, have demonstrated for various cell types that the mitotic apparatus under normal circumstances plays a role in determining the cleavage plane up to a 'point of no return' (usually mid- to late-anaphase). After this point, cleavage is independent of the presence or location of the mitotic apparatus. Hence, the mitotic apparatus seems to play two interrelated roles: the first involves chromosomal movement; the second, determi- nation of the cleavage plane. One of the major stumbling blocks to an understand- ing of these mechanisms has been the inability to isolate in purified form the various constituents of the mitotic apparatus. In contrast to the supposed homogeneity of the mitotic complex implied by Mazia's earlier studies, electron microscopy has shown this structure actually to be composed of numerous, sizable components. These include chromatin, vesicles, ribosomes, endoplasmic reticulum, an electron-diffuse amorphous material, and microtubules (Bibring and Baxandall, 1969). Detailed research into the structure and mitotic function of only one of these has been investigated to date: this is the microtubule component which is thought to be the major protein constituent of the spindle fibers. The spindle fibers themselves can be classified into four major groupings, (Inoué, 1953): the kinetochore fibers, the interzonal fibers running from chromatid to chromatid at anaphase, the continuous fibers which run directly from pole to pole, and lastly, astral fibers which radiate from the area around each centriole into the cortical regions of the cell. The work of Nicklas (1967), and Takeda and Izutsu (1960), who used microsurgical, UV and pressure techniques to disrupt or selectively break the chromosome-to-pole fibers indicates: (1) that the chromosomes are pulled at metaphase with equal force by both fibers; (2) that this is a pulling, not a pushing force by these fibers; (3) that the fibers themselves have very labile qualities as evidenced by their ability to rejoin to the kinetochore after detachment; and (4) that a UV—induced spot of reduced birefringence on a spindle fiber moves toward the nearest pole at the same rate as the chromosomes (Forer, 1965), indicating that actual fiber movement occurs during mitosis. It is also a generally accepted fact that no change in spindle or microtuble diameter accompanies chromosomal movement; instead, a 'melting' into the centriole region is seen (Dupraw, 1968). This implies that the Spindle fibers or their constituent microtubules are in continuous motion. This could perhaps be caused by polymerization or depolymerization of microtubular subunits at one or both of the two points of connection (Inoué, 1959; Mazia, 1961; Went, 1966) or by the sliding of micro- tubules over one another during anaphase. This latter type of conjecture is, of course, extremely speculative and further points out the need to study directly and individually the components of the mitotic apparatus during the cell cycle. Composition of the Spindle Fiber Microtubules Electron Microscopy As early as 1957, microtubules (200 nm diameter) were observed by Porter in the spindle of various rat tissue cells. These were suggestive of the eleven micro- fibrils, now known to be microtubules, seen by Manton (1952) in flagella. Later, well-known studies by workers such as Fawcett and Porter (1954), Afzelius (1963) and Gibbons (1961, 1963, 1967) on the ultra- structure and function of the now familiar 9-2 arrange- ment of cilia and flagella further established the con- sistency of 200-250 nm microtubules in biological sys- tems. Calcium stabilization techniques developed by researchers such as Roth and Jenkins (1962) demonstrated that spindle microtubules (groups of which comprise the spindle fibers) seen under the light microscope, also show a consistent diameter of about 20.0 +/- 3.0 nm. In 1963, André and Thiéry, using ultrathin sectioning, described 13 protofilaments making up the hollow micro- tubule of sperm flagella. Pease (1963) and Grimstone and Klug (1966) further resolved flagellar microtubules into approximately 45 nm globular subunits which, in a linear array, form the 13 fibers of the 9-2 flagellar (or ciliar) microtubules. Ledbetter and Porter (1964) and Gall (1966) demonstrated that cytoplasmic microtubules also show this globular subunit ultrastructure. More recently, Moor (1967), Barnicot (1966) and others have verified that the Spindle microtubules resolve into similar (3.0-5.0 nm) globular subunits arranged in a helical manner. Hence at the level of the electron microscope the most conspicuous structural component of the mitotic apparatus (the spindle fibers) resolve into bundles of hollow (approximately 20.0-25.0 nm) micro- tubules composed of 4.5 nm globular subunits perhaps helically arranged. Their morphology and localization in motile structures imply the similarity and possible contractile-related role of flagellar, ciliary, cyto- plasmic and spindle microtubules. Isolation of Microtubule Protein A second major line of research into the nature of the mitotic spindle has involved the attempt to isolate and purify the microtubule protein (now known as tubulin). Ever since Mazia (1955) found that there is one homo- geneous major protein constituent of the mitotic appara- tus, much work has been done in an attempt to isolate, characterize and quantify this protein. At this writing numerous workers have successfully isolated a 65 colchicine and GTP binding protein of molecular weight 60,000. For the reasons given below, most workers now believe this to be the major protein constituent of mitotic and cyto- plasmic microtubules. However, reference should also be made to studies (Kane, 1967; Sakai, 1968) in which a different mitotic apparatus protein has been isolated or which support the concept that there is some heterogeneity of protein constituents among different classes of microtubules. In 1962, Kane, using ultracentrifugation, isolated a single protein from the mitotic apparatus of sea urchins with a sedimentation value of 223. Since then, other workers have also isolated a 223 homogeneous protein from the mitotic apparatus (Stephans, 1967). This 223 protein is KCl soluble with an estimated molecular weight of 880,000. However, this 22s protein has not been com- pared in detail to the flagellar microtubule protein which can be readily extracted from a flagellar sus- pension, nor has specificity of colchicine binding been demonstrated. Bibring and Baxandall (1969) prepared an anti-serum against this 223 protein. This anti-serum reacted with all sea urchin egg fractions free of the mitotic appara- tus. Labeled antibody studies utilizing electron micro- scopy demonstrated that the 223 protein was localized within the amorphous material of the mitotic apparatus and not in the microtubules. A different protein can be isolated from sea urchin mitotic apparatus by the technique of Sakai (1966). This protein and a possible polymer of it (3.Ss--Sakai, 1968) can be obtained through ultracen- trifugation of the hexylene glycol stabilized mitotic apparatus. This protein has a molecular weight of 34,700 and a dimer molecular weight of 68,000. Taika Miki- Noumura (1968) has obtained a similar (2.3s) protein from the sea urchin mitotic apparatus and presents evidence based on electrophoretic patterns, centrifu- gation and precipitation purification that this, and not the 223 protein, is the microtubule protein. Kiefer and Sakai (1966) give suggestive evidence that this protein is in fact the 3.3 nm subunit forming the spindle microtubules. These conflicting studies point out one of the major difficulties in evaluating supposed microtubule protein. Until recently, the only known way to demon- strate that the protein isolated from the mitotic appara- tus or cytoplasm was indeed that of microtubules was to compare its properties to the known characteristics of the more easily verifiable ciliary or flagellar micro- tubules. This assumed an identity between these proteins based on similar ultrastructure, colchicine binding activity and common antibody cross reaction (Ruby, 1961). This has been questioned by the recent work by Tilney and Gibbons (1968) and Behnke and Forer (1967), who showed consistent differences in temperature stability, effects of trypsinization and response to hydrostatic pressure and osmium fixation between ciliary, spindle and cyto- plasmic microtubules. Whether this reflects any real differences in protein content between these microtubules, or simply indicates that cytoplasmic position affects the susceptibility of microtubules to these agents is not known. More recently, Adelman, Borisy, Shelanski, Weisen- berg and Taylor (1968), and Weisenberg, Borisy, and Taylor (1968) have carried out a series of detailed studies on the presumed microtubule subunit protein. They isolated by ultracentrifugation, thin layer chromatography and disc electrophoresis a 65 homogeneous protein from a wide variety of cell types. This protein, after careful analysis, showed identical or nearly identical properties, whether isolated from the inner microtubules of sea urchin flagella (electron microsc0pic verification) or from mammalian brain. Its characteristics of a Gs sedi- mentation value; a M.W. of 60,000 or 120,000 depending on degree of polymerization; and GTP binding were also nearly identical to the protein isolated by Renoud, §E_§1. (1968) from Tetrahymena cilia outer doublets. Another 10 interesting feature, common to all of the purified pro- teins, was an amino acid composition similar to that of actin. One further piece of evidence which suggested that these authors had indeed isolated the microtubule protein was the fact that, under proper conditions, as much as one-third of the protein fraction was shown to exist as a 33,000 M.W. subunit which was demonstrated to be derived from the 120,000 M.W. protein (Borisy, Taylor, and Weisenberg, 1968). This puts their protein in the same molecular weight category as Sakai's 2.53 spindle protein. Lastly, Borisy and Taylor (1967) were also successful (using the same techniques) in isolating a 68 protein from the sea urchin mitotic apparatus with nearly identical properties to those listed above. They were further able to show that removal of this colchicine binding protein paralleled the loss of spindle micro- tubules from the hexylene glycol extracted mitotic apparatis. Lastly this protein could also be selectively extracted by precipitation with the Vinca alkaloid vin- blastine which induces crystal formation in cells known to possess high colchicine binding activity (Bensch, Marantz, Wisniewski and Shelanski, 1969). Hence, a homogeneous actin-like colchicine and GTP— binding protein, capable of reversible polymerization, isolated under conditions which selectively remove micro- tubules, has been purified from the mitotic apparatus, ll cytoplasm, flagella and cilia or protozoa, echinoderms and mammalian tissues. The works cited above suggest that the Borisy-Taylor-Weisenberg and Renoud-Gibbons groups (and perhaps Sakai's laboratory) are all dealing with a common subunit protein of microtubules which may be phylogenetically very widespread. The lack of either independent verification from other laboratories or electron microscopic examinations, added to the study by Bibring and Baxandall (1969), cited above, casts serious doubt on the proposed microtubular origin of Kanes 223 protein. It is, therefore, suggested here that the techniques of the Taylor group provide a con- sistent, reliable method of extracting what at this time appears to be the subunit protein of most microtubules. Functions of Microtubules and Microfilaments in the Mitotic Process With all of the voluminous literature available concerning the composition of microtubules little is known concerning their function. Some of the literature cited above indicates that spindle microtubules play two roles in cell division. One concerns their possible interaction with the cell cortex as an initiator of cleavage, or as a factor in the original localization of the future cleavage plane (Hiromato, 1956, 1965). The second possible functional role may involve the various movements associated with cell division. Little is known 12 of the former but a great deal of descriptive work exists concerning the latter. The work of Peters and Vaughn (1967) suggests the inclusion of filaments under the heading microtubules. They present electron microscopic evidence of developing optic nerves and insect flight muscles indicating that micro-neurofilaments and myo- filaments are derived from microtubules and perhaps vice versa. Both of these structures are common integral constituents of numerous motile cell components. This close structural relationship between microfilaments (35-50 g), microtubules and cellular machinery involved in movement is often taken as suggestive evidence that the former are indeed responsible for motility. For example, the association of microtubules or filaments with motile activities has been reported for cell shape differentiation by Gibbons, Tilney and Porter (1969), Tilney and Gibbons (1968); for cytoplasmic streaming by Wblforth-Bottermann (1964); for filopodes by Bowers and Korn (1968); and in the cleavage furrow of numerous cell types (Szollosi, 1968; Schroeder, 1968, and Arnold, 1969). Hence, these structures are commonly associated with motile structures of cell movement in general. This, of course, is only suggestive evidence and it is just as likely that these filaments and microtubules play a solely structural role, providing a supporting framework along which other contractile elements might move. With l3 respect to this problem, however, Sakai (1966, 1968) has demonstrated that a calcium-insoluble 2.53 mitotic appara- tus protein will undergo an electron exchange reaction with a cortical contractile protein identical to that occurring during contraction of a two factor contractile complex. Thus, there is still little understanding of the role of microtubules or filaments in mitosis. Recently a great deal of evidence has been reported in the literature which conclusively demonstrates that the 50-70 A filaments are actually F—actin and have no relationship to microtubules. Ishikawa, Bischoff and Holtzer (1969) demonstrated that cortically localized fibers would bind rabbit heavy meromyocin £2.§£EE in a wide variety of vertebrate cell types, Weihing and Korn (1969) have repeated this using the amoeba Acanthamoeba. Korn and his co-workers have also been successful in isolating amoeba actin which was shown to possess identi- cal properties to rabbit actin (Weihing and Korn, 1970). wessells, Spooner, Ash, Bradley, Luduona, Taylor, Wrenn and Yamada (1971) in their review of the role of micro- filaments further describe numerous studies which con- clusively demonstrate that the drug cytochalasin-B binds to cellular F-actin and thereby inhibits contractile processes (gross cell movement, filopod or cortical undulations, cytokinesis, etc.). Colchicine has no effect on these events and, unlike cytochalasin, does 14 not depolymerize the F-actin fibers. All that will be assumed here is that microtubules probably are essential to the normal chromosomal movement occurring during mitosis and make up the main protein structural com- ponent of the mitotic apparatus. Metabolism and Chemistronf MiCrotubuIér Protein Almost nothing is known about the time of synthesis or general turnover of microtubules during the cell cycle. Went (1960) and others have demonstrated that protein antigens of the mitotic apparatus are present in appreciable quantities prior to division. Recently, Sisken and Wilkens (1967), studying the effects of amino acid analogues, demonstrated that a protein(s) essential to mitosis (its alteration by abnormal amino acid analogues producing prolonged metaphase and death), was synthesized in G2. Hodge, Robbins and Scharff (1969), studying the effects of actinomycin D during the cell cycle, concluded that at least 40% of the mRNA present in late 62 persists through mitosis and functions at 61‘ Gross and Couseneau (1963) found that leucine-H3 is specifically incorporated into the sea urchin mitotic apparatus after fertilization. Inoué and Sato (1967) showed that the increase in amount of spindle micro- tubules (as evidenced by increased birefringence upon addition of D20 was not accompanied by a change in total 15 cellular mitotic apparatus protein. After treatment of cells with colchicine and subsequent washing, they found that protein synthesis inhibitors such as puromycin or chloramphenicol did not in any way affect the speed or amount of mitotic apparatus reorganization. Robbins and Shelanski (1969) have shown that the colchicine binding protein can be synthesized at 61' S, and G2 in rat liver cells and leucocytes. However, their techniques did not permit quantification relative to the total amount of microtubule protein. Within the past 5 years numerous workers have studied the LE.X£XQ chemistry of colchicine binding protein in numerous syStems. The results of these studies have revealed a wide spectrum of interest- ing physiological phenomena associated with this protein. In nervous tissue, for example, colchicine binding protein is associated with or identical to the slow component of axon transport (James and Austin, 1970). Gross (1970) has provided evidence from his laboratory that the syn- thesis of sea urchin microtubule protein is actinomycin insensitive and therefore the protein is read from stable messengers. Coyne and Rosenbaum (1970) have shown that Chlamydomonas flagellar regeneration can occur in the absence of protein synthesis further demonstrating the presence of a relatively large protein pool. Wilt, Sakai and Mazia (1967) have further demonstrated that inhibition of more than 90% of all protein synthesis 16 does not inhibit the first division in sea urchin embryos. Auclair and Siegel (1966), using sea urchin blastulae, were also able to prove that as in the Chlamydomonas flagellar regeneration systems of Coyne and Rosenbaum (1970) ciliar regeneration can occur in the absence of protein synthesis. These and similar studies imply that the bulk of the proteins of the mitotic apparatus may persist through the cell cycle to be used again and again even though new protein synthesis must obviously occur. However, in only one of these studies has the nature of the conserved protein(s) been examined. In no case has microtubule protein synthesis been studied quantitatively. It will be one of the major goals of the present research described here to investigate this problem with spe- cific regard to the turnover of microtubular protein during the cell cycle. Research Goals On the basis of the preceding literature review it can be concluded that the mitotic process is a very complex and poorly understood cellular phenomenon. The main structural component of this process is the mitotic apparatus. It functions in the determination of the cleavage plane and in chromosome movement. An eluci- dation of the mechanisms for the formation and breakdown of this transient organelle may well lead to an 17 understanding of the regulation of cell division in general. Due to the complexity of the division process a study of these regulatory mechanisms must focus on one component at a time. It is very fortunate that a single isolatable component is present in dividing cells. This component is the spindle fiber complex. The constituent microtubules make up the major structural feature of the mitotic apparatus. These tubules are made up of one or perhaps two peptides which can be easily isolated in pure form. The specific function of this protein during mitosis is incompletely understood. However, we do know that the polymerization of this protein into mitotic microtubules is a prerequisite for successful cell division. Hence, an ideal opportunity exists to study a single well-defined molecular aspect of cell division. The major questions that need to be studied include the following: (1) What is the function of microtubules during mitosis? (2) How does the cell regulate the polymerization of microtubule protein (tubulin) into spindle fiber microtubules? (3) Is the regulation of the formation of spindle microtubules also connected with the regulation of cell division in general? (4) Is spindle fiber tubulin the same protein as that of cyto- plasmic microtubules? And lastly, (5) How do the known chemical characteristics of tubulin relate to its function in the cell? 18 The research program described in this thesis was designed specifically to study the second question listed above. Specifically, the question of whether or not tubulin polymerization during mitotic apparatus formation is induced by the rapid synthesis of tubulin at early prophase was investigated. However, in the course of this study, information concerning questions 1-5 was also obtained. The experimental system utilized in this study was the small soil amoeba Acanthamoeba rhysodes. This organism possessed some unique characteristics which were particularly useful for the study of microtubules and cell division. These included the ability to experimentally induce synchronous mitosis and amitotis. The latter phenomenon involves an uncoupling of cyto- kinesis from karyokinesis and as such provided an oppor- tunity to study the role of microtubules in cytokinesis. Furthermore, microtubules serve only a mitotic function in these cells, and therefore any results could be interpreted specifically with respect to mitosis. Lastly, the cells can be grown in axenic cultures in which high yields of cellular material could be easily collected. The research program was carried out in three phases. In the first, experimental manipulations were combined with electron microscopy to determine the 19 localization of microtubules during mitosis and amitosis and to discover whether they play an essential role in these two processes. The second phase involved the isolation and chemical characterization of amoeba tubu- lin. These characteristics were compared to the well- defined mammalian brain tubulin. This aspect of the research program made it possible to state that the protein being studied was indeed microtubule protein and also provided interesting data on the question of the heterogeneity of tubulins which serve different cellular functions. The last phase for which the first two were prerequisites involved a study of the synthesis of the protein in synchronous and asynchronous cell popu- lations. It was hoped that this latter study would help to answer the question concerning the relationship between the synthesis of tubulin and the polymerization of tubu- lin into spindle microtubules. MATERIALS AND METHODS Culture Techniques Culture Media Normal Acanthamoeba rhysodes cultures were grown on the nutrient medium (PPG) or defined medium of Band (1962) employing a static bubbler flask for 5 days fol- lowed by a transfer to a rotary shaker containing normal culture media, starvation media or normal culture media plus an agar surface, depending on the treatment desired. Induction of Amitosis For the most recently devised procedure, cells were cultured as described by Band (1959) but with glucose starvation (Band, Mohrlock and Rubin, 1970). After 48-72 hours of this treatment, cells were placed back in normal culture media in shaker flasks containing 0.1% agar and within 3-4 hours more than 90% of the cells had undergone amitosis. This involved cytokinesis, cutting through the intact nucleus and nucleolus. All cells undergoing amitosis die within 24 hours after division. If the starved cells were placed on shaker flasks in normal 20 21 media lacking agar, a high degree of mitotic synchrony resulted for the following two cell cycles. Induction of Synchronous Mitosis Synchronous nuclear mitotic division lacking cyto- kinesis was accomplished by the starvation precedure described above for which defined medium (Band, 1959) lacking thiamine could be substituted for PPGF or by adding 0.005M colchicine to the normal medium for 24 to 48 hours followed by repeated washing in normal culture medium. They were finally placed in culture media (PPG) in a shaker flask after which most cells underwent apparently normal nuclear division within 1 to 3 hours, producing a binucleate culture. These binucleate cells subsequently all underwent cytokinesis as described above within the following 16 hours if provided with a surface. Filter Paper Assay for Colchicine Binding For the colchicine binding experiments H3 colchicine obtained from New England Nuclear with an activity of Sc/mM in 0.01 M sodium phosphate buffer containing 0.01 M magnesium chloride plus 0.001 M GTP (pH 6.5) was added to an aqueous sample in 0.01 M Na phosphate buffer (pH 6.5) containing .01 M MgCl plus .001 M GTP to make a final concentration of l uc/ml. These samples were incubated for one hour at 37°C after which cold colchicine 22 was added to give a final concentration of 10'4M. A given H3 colchicine labeled sample was then placed in a small gooch crucible in which a 2 cm diameter Whatman DEAE chromedia paper disc was used as a filter. The sample (1-2 ml) was then allowed to slowly drip through the filter paper after which the paper was washed 8 times in 0.01 M sodium phosphate buffer containing 0.01 M mag- nesium chloride and 0.001 M GTP. The filter paper was then dried and placed directly into scintillation vials. For this filter paper assay, modified from that of Wei- senberg, §£_al, (1968), it was often necessary to use two filter paper discs to increase the efficiency of protein binding. Flow Sheets for_Assays of Major Cellular FractIEns The first three experiments on the assays of major cellular fractions (experiments 1-3, Figure 1) were rather complex and flow sheets for these 3 experiments are pro- vided below. Electron Microscopy Cells were prepared for electronmicroscopy by fix- ation in 2 1/2% glutaraldehyde in PPG at room temperature for 1 hour. This was followed, after four washes in .01 M sodium phosphate buffer (PH 6.5), by post fixation in osmium tetroxide in Zetterquvist buffer (PH 7.0). The cells were then washed three times in sodium phosphate 23 Figure 1 Flow Sheet for colchicine binding experiments l-3. 24 Experiment No. l 10 ml packed cells Homogenize in 1 volume Na-phos-Mg-GTP-sucrose buffer in cold 15 min wait Centrifuge 13,000 g x 20 min low speed * ---supernatant 4—% l ‘v pellet resuspend in Na-phos-Mg—GTPT buffer * ---1/2 of pellet sample——————92 1/2 of pellet sample I l w 1 ml sample 1 ml sample Treatme nt 1 TreatmeLnt 2 Add 0.01 M cold A15 min wait vcolchicine llS min wait add 0.01M H3-colchicine add 0.01 M H3-colchicine Jyl hr at 37°C l1 hr at 37°C filter and wash filter and wash through #1 through #1 Whatman filter Whatman filter paper paper Ur labeled protein on filter labeled protein on filter paper to be assayed in paper to be assayed in scintillation counter. scintillation counter. Fig l (Cont.) 25 low speed supernatant (continued from above) centrifuge 150,000 g x 1 hr ---supernatant High speed v pellet resuspend in Nalphos-Mg-GTP buffer i 1 m1 sample Treatment 5 1 .L15 min wait (Ladd H3-colchicine ‘LDEAE filter paper add 0.01M cold colchicine labeled protein on filter paper to be assayed on scin- tillation counter. .1; 1 ml sample Treatment 6 L15 min wait Aadd H3-colchicine iDEAE filter paper labeled protein to be assayed in scintillation counter. 1/2 of resuspended pellet * (continued from above) 2 1 R3 sample I Treatment 3 l ~L15 min wait filter through regular Whatman filtrate #1 filter paper J,add H3-colchicine Jrl hr, 37°C filter through DEAE filter paper labeled protein on filter paper to be assayed in scintillation counter. Add cold colchi- cine Fig 1 1 m1 sample Treatment 4 115 min wait filter through Jregular Whatman #1 filter paper filtrate ladd H3-colchi- cine 11 hr, 370C filter through lDEAE filter paper labeled protein on filter paper to be assayed in scintillation counter. (Cont.) 26 High speed supernatant (continued from above) labeled protein on filter paper to be assayed in V 4’ 1 ml sample 1 ml sample Treatment 7 Treakment 8 \V 3" add cold 0.01 M 115 min wait colchicine ‘ladd H3-colchi- 115 min wait cine ladd H3-colchicine ll hr, 37°C ll hr, 37°C imam filter lDEAE filter paper labeled protein on filter paper to be assayed in scintillation counter. scintillation counter. Fig l (Cont.) 27 Experiment No. 2 10 m1 packed cells Homogenize in Na phos-Mg-GTP-sucrose buffer 14,000 g x 20 min supernatant 1 200,000 g x 20 min pellet resuspend in 5 ml Na phos-Mg- GTP buffer add H3-colchicine J ' ‘ A 2.5 ml 2.5 m1 Treatment 1 Treatment 2 l1 hr wait l200,000 g x 1 hr DEAE filter supernatant paper protein on filter paper to be assayed in scintillation counter. protein on filter paper to be assayed in scintillation counter. DEAE filter paper Fig l (Cont.) 28 Experiment No. 3 10 m1 packed cells Homogenize in Na phos-Mg-GTP-sucrose buffer 14,000 g x 20 min supernatant 1 200,000 g x 1 hr pellet resuspend in Na phos-Mg-GTP buffer to 6 ml 5" A 2 m1 2 m1 TrFatment l Trethent 2 Iadd cold colchicine Ladd H3—colchicine I15 min, 37°C [1 1/4 hr, 37°C ladd H3-colchicine lDEAE filter paper 0 ‘1 hr, 37 C protein on filter paper to be DEAE filter assayed in scintillation paper counter. protein on filter paper to be assayed in scintillation counter. v 2 m add H3-colchicine 1 hr, 370C l 200,000 g x 1 hr A T pellet supernatant resuspend Trektment 3 Treatment 4 \LDEAE filter paper DEAE filter W paper rotein on filter a er to . , ge assayed in scinEiTlation protein on flLteF paper to be counter assayed in sc1ntillation counter. Fig l (Cont.) 29 buffer and rapidly dehydrated with ethanol followed by propylene oxide. The propylene oxide infiltrated cells were then placed in 50% Epon 812 (A & B mixture of Luft, 1961) (Pease, 1965) and 50% propylene oxide overnight. After 24 hours in 100% Epon the cells were centrifuged in conical embedding blocks and polymerized for 24 hours at 60°C. Thin sections were made using glass knives and an MT-2-u1tramicrotome. The sections were stained with uranylacetate (2% aqueous plus 1 drop triton x-lOO) and lead citrate (.25 gm/SOml plus 1 pellet of NaOH and 1 drop of Triton X-lOO). The stained sections were photo- graphed using a Philips 300 electron microscope at an accelerating voltage of 80KV. The statistical analysis of the nuclear envelope size utilized a standard analysis of variance procedure, testing the significance of differences between the intermembranous distance of the amitotic thread verses interphase nuclear envelope. Measurements were made between the outer edge of the outer nuclear membrane and the outer edge of the inner membrane. These measurements were made on non-glossy prints which represented a total magnification of 400,000 x. Actin Assay The presence of F-actin in a 10,000 g supernatent was tested by using the viscosity technique of Weihing and Korn (1969). In this technique the ATP-reversible 30 increase in viscosity upon the addition of rabbit myosin was taken as an indication of the presence of F-actin. Gel Filtration In all cases a 2.5 x 100 cm or a 1.5 x 50 cm column was used with sephadex G-200. Chromatography was carried out at 4°C after equilibrating the column with 3 void volumes of sodium phosphate-Mg-GTP buffer.’ Fractions were measured by absorbance at 280 mu. For the colchi- cine binding experiments a 2 to 4 ml aliquot from each tube was placed directly into scintillation vials con- taining one disc of DEAE chromedia paper. These vials were then dried prior to the addition of cocktail. Isolation of Colchicine Binding Protein Tubulin was isolated from bovine brain using the batch technique of Weisenberg, §E_El° (1968). This pro- tein was stored as a lyophilized powder. For the iso- lation of a colchicine binding protein from amoeba, a modification of this procedure was found to be neces- sary (Figure 2). Electrophoresis Tests of protein purity were done using polyacryl- amide gels. These were made up after the technique of Clark (1964) and include 8M urea in a gel system which is made up to a final gel concentration of 4%. Protein samples to be tested on this gel system were reduced in 31 Figure 2 Colchicine binding protein isolation procedure for Acanthamoeba rhysodes. 32 Wash cells in Na phos-Mg-GTP-sucrose buffer (pH 6.5) Homogenized in 1 vol Na phos-Mg- GTP-sucrose buffer with 10 strokes of Teflon pestle at O0 C 115 min wait 115,000 g x 20 min supernatant slowly add 0.177 gm ammonium sulfate per m1 supernatant (10,000 g x 10 min supernatant add 0.071 gm ammonium sulfate per m1 supernatant £10,000 g x 10 min pellet , Wash pellet surface Na phos-Mg- I WGTP buffer Alternate Rapid solubilize pellet in 10 vol Na technique phos-Mg-GTP buffer solubilize pellet add equal volumes solubilized in 5 vol Na phos-Mg- protein to DEAE Sephadex A-SO GTP buffer previously swollen in Na phos- Mg-GTP buffer and adjusted to ladd to Sephadex G-200 pH 6.5 column Check pH to 6.5 and stir occas- ionally 30 min collect second peak 3,000 g x l min purified protein discard supernatant pellet add equal volume 0.5 M KCl in Na phos-Mg-GTP buffer (pH 6.5) check pH, stir occasionally, 15 min 3,000 g x 1 min Wdiscard supernatant pellet Fig 2 (Cont.) V 33 pellet (cont' from above) lrepeat 0.5 M wash twice more pellet add 10 ml 0.8 M KCl (pH 6.5) in Na phos-Mg-GTP buffer per 30 ml Sephadex, stir 15 min 3,000 g x l min p lletes repeat 0.8 M KCl wash V/ supernatant €> supernatant pool pour through parachute silk dialyze 24 hrs against Na phos-Mg-GTP buffer 15,000 g x 40 min V, supernatant = purified protein Fig 2 (Cont.) 34 0.12% mercaptoethanol for 30 minutes at 37°C followed by the addition of enough urea to make the solution 8 M. This mercaptoethanol-urea reducing preparation was then kept at room temperature for 24 hours. For cases in which complete and permanent reduction was required the protein solutions were reduced and aklylated by the method of Renaud, gp_§l. (1968). SDS-polyacrylamide gel molecular weight determinations were carried out using the method of Weber and Osborn (1969). It was found that the commercially supplied SDS required recrystali- zation to remove impurities prior to initiating a molec- ular weight experiment. RNAase, Lysozyme, Trypsin, insulin, BSA and catalase were used as reference standards. Amino Acid Composition Determination Protein samples at a concentration of 1 mg/ml were hydrolyzed in 6 N HCl at 105°C in a nitrogen environment for 18 hrs. The amino acid composition of these hydro- lyzed samples was determined on a Technicon auto analyzer using C-2 chromobead resin and a sodium citrate buffer varying from pH 2.75-6.1 with the help of the laboratory and staff of Dr. D. Lamport in the AEC Research Center at Michigan State University. Turnover Experiments For experiments on the turnover of amoeba tubulin PPG grown cells were labeled in defined medium (Band, 1959) 35 with H3—1eucine for 24 hours. After one wash in PPG the cells were placed in cold PPG (2L flasks) from which aliquots were removed periodically. Each sample was then washed twice in PPG and frozen. Later all of the frozen samples were thawed and amoeba tubulin isolated by the modified Weisenberg technique. The remaining supernatant and first ammonium sulfate precipitate were collected for each sample. These comprised the fractions from which the total soluble proteins were extracted. These fractions were then precipitated in cold 5% TCA for 1/2 hour. The TCA precipitates were then washed in ethanol, ethanol-ether and hot TCA according to the pro- cedure of Roberts, ep_31. (1951). The resulting precipi- tate was solubilized in 10% NaOH and two 1/2 ml aliquots were removed. One of these was used to quantitate the protein content using the Lowery method (Lowery, Rose- brough, Farr and Randal, 1951). The other aliquot was placed into 10 ml of cocktail in scintillation vials and the cpm counted on a Packard tri-carb scintillation counter No. 3320. Similar aliquots were removed from the purified amoeba tubulin preparations for specific activity determinations. RESULTS Effects of Colchicine Application Inhibition of Amitosis Statically grown cells were placed in starvation media (PPGF) for 24 hours at which time 5 x 10-3 M colchi- cine was added. The cells were kept in the colchicine- PPGF media for another 24 hours after which they were transferred to normal media containing agar (PPGA) plus colchicine. A control culture was taken from the same static culture and treated in an identical manner with the exception that no colchicine was added. As seen in Figure 3 colchicine at the concentration used greatly inhibited amitosis relative to the control culture. This experiment was repeated twice with identical results. Morphology of Treated Cells No microtubules were seen in cells fixed in glutaral- dehydeosmium after varying times of exposure to colchicine. This was true for PPG, PPGF and PPGA grown cells treated with colchicine for 24, 38 or 60 hours. Normal interphase cells possessed a few very short (30-40 nm) microtubule segments (Figure 4). Except for an increase in vacuoles 36 37 mHHmo mo wmmucmoumm mnu mm woudmmme mwmouflfim .Aaooaosc on mcwmmmmmom .Alllv ousuHso omumouuc: Honucoo man on coummeoo mm A¢¢dv manuaao woodman ocwownoaoo on» CA aoflufinflncfi mumamsoo umoaam muoz .mcaoagoaoo Emoo. an mflmoufiem Co aofluflnHEcH m musaflm 38 7.301.035... N 907,, O p ”00an cu qggM SHED O N O n O C On 39 Figure 4 Random section through interphase cell. Note short microtubules (MT). x 6,000 40 41 and loss of microtubules, all colchicine treated cells appeared normal and could be maintained for 85 hours in colchicine, placed back in PPG, and completely recover normal growth within 12-24 hours. Ultrastructure of Mitotic and Amitotic Cells Ultrastructure of Mitosis A typical metaphase cell is shown in Figure 5 at low magnification. Extensive blebbing is evident. These blebs all contain 7 nm diameter fibers oriented parallel to the long axis of the blebs (Figure 6). Sub- cortical bundles of these fibers are seen scattered around the periphery of the metaphase cell (Figure 7). Figure 8 is a composite photograph of the mitotic appara- tus of a metaphase cell, demonstrating 22 nm diameter spindle fiber microtubules (singly and in bundles) and an electron dense structure into which the microtubules insert. Also of interest is the single microtubule positioned in the center of and parallel to the metaphase alignment of chromosomes (Figure 9). The mitotic appara- tus as a whole is typical in ultrastructure of that of most eukaryote cells. No lysosomes, 70 nm fibers or endoplasmic reticulum are present. However, occasional spherical or tubular vesicles, mitochondria as well as ribosome-like particles are observed within the metaphase 42 Figure 5 Metaphase cell. Note cortical blebs (B). x 11,000 43 44 Figure 6 Cortex of metaphase cell. Note microfilaments in cortical blebs. 45 46 Figure 7 Periphery of metaphase cell. Note cortical microfilament bundles (Mb). x 88,000 47 48 ooo.mm x .azv mwuoconooufie can “may moon omawo .Ahmv moadnsuouowfi Hogan wavcwmm .AOV mmEomoEouno ouoz .Haoo ommnmmuwa o no usumnmmmm oaaoufle may mo nmmumouonm ouwmomsoo m ousmwm 49 50 Figure 9 Metaphase chromosomal alignment. Note single microtubule (Mt) in center of chromo- somal diad and vesicles (V). x 84,000 51 52 mitotic apparatus. Figure 10 is an enlargement of the electron dense body showing its bipartite structure. Ultrastructure of Amitosis Cells synchronized by PPGF starvation followed by incubation in PPGA were sectioned in both the longitudinal and transverse planes. At the light microscopic level mitosis is seen as an active cortical constriction of the cytoplasm which ultimately begins to constrict the intact nucleus and nucleolous (Figure 11). This is followed by an opposite migration of both presumptive daughter cells (Figure 12). The migration greatly elongates the cleavage plane produced during early cytokinesis until a very long thin thread 1-0.S u in diameter connects the presumptive daughter cells (Figure 12). This thread ultimately breaks and both daughters die within 48 hours. At the level of electron microscopy, the details of this process are revealed. Figure 13 shows a typical amitotic cell in an early stage of amitotic cytokinesis. The absence of cortical blebbing and presence of the intact nucleus and nucleolus is apparent. Higher magnification of the cleavage plane (Figures 14 and 15) reveals the presence of microtubules just outside of and parallel to the intact nuclear membrane. These microtubules range from 220 to 800 nm in length. With the exception of the microtubules associated with the dense body, no microtubule lengths 53 Figure 10 Dense body (DB) from metaphase cell. Note bipartite structure and attached microtubules (MT). x 150,000 54 55 oom.~ x ocmam mmm>moao vapouweo muoz .A3ouumv msoaose uomucw oucfi mcfluuso .Aumouucoo ommnmv Haoo vapouflfim mason Ha ousowm 56 57 oom.m x .A3ouumv Hounmsmo moo cw msaomaosc manmowuoc mom maflxoma mamaozc emsoumuum 0:» mo Dunn can Lav enema» espoussm 00oz .Haoo owuouHEMToflz NH ousmwm 58 59 uomucfi smoounp mcfluuso woman omm>moao owuouflem muoz oom.aa x .sz msmaosc .Hamo vapoufiEm Manna ma ousmfim 60 61 Figure 14 Early amitotic cleavage plane. Note elongate morphology of stretching nucleus (N) and the presence of short microtubule segments (MT). X 40,000 62 63 Figure 15 Early amitotic cleavage plane in longitudinal section. Note single microtubule (MT) along outside of nuclear membrane (NM). x 59,000 RATE...” {1" 65 in this order of magnitude are observed in interphase. Microtubules are seldom seen in parallel alignment to the interphase nucleus. As amitotic cytokinesis proceeds, the daughter cells pull apart and the cleavage plane becomes stretched out into a thin thread. In longitudinal sections this thread can be seen to contain part of the nucleus and intact nuclear membrane and nucleolus (Figure 16). Another consistent feature of the early amitotic cleavage plane is the presence of a cylindrical sheet of 7 nm fibers lying just under the plasma membrane (Figure 17). These fibers are always found in parallel alignment to the axis of the elongating amitotic thread. As amitosis proceeds, the thread connecting the two daughter cells continues to elongate until it has stretched into a 0.5 to l um diameter thread. By this stage microtubules are no longer present and the 7 nm fibers appear sparse and randomly aligned (Figure 18). Also, the intact nuclear membrane containing nucleolar material is still visible. Occasional mitochondria and randomly spaced vesicles are also observed within the amitotic thread at this time (Figure 18). In sections taken from the termination point of the elongated nucleus, the nuclear membrane is seen to be highly convoluted (Figure 19). When the amitotic thread is reduced in diameter to approximately 0.5 um the force of the separating daughter cells 66 Figure 16 Longitudinal section through early amitotic thread. Note intact nuclear membrane (NM) and nuCleolar material (N). x 40,000 67 68 Figure 17 Longitudinal section through early amitotic thread. Note intact stretched nucleus (N) and sheet of microfilaments (MF). x 35,000 Figure 18 70 Transverse section through late amitotic thread. Note intact nuclear membrane (NM), nucleolar material (N), ribosomes (R) on outer nuclear membrane and single mitochondrion (M). x 175,000 71 72 Figure 19 Longitudinal section through mid-amitotic thread. Note stress lines in the nuclear mem- brane (NM) and the convoluted termination point (TP). x 32,000 74 usually breaks the thread. When thread elongation reaches this final stage occasional longitudinal sections reveal a pronounced involution of the ami- totic nuclear membrane. This is seen as a series of lines within and parallel to the greatly stretched nucleus (Figure 19). Interphase Cells Stgpcture of theDense Body in Interphase Cells Longitudinal and transverse sections (Figures 20 and 21) demonstrate the interphase morphology of the dense bodies. The apparent polarity of microtubule insertion is evident. The unipartite structure of these bodies during interphase is also apparent. A consistent feature of these organelles in interphase as well as mitotic cells is their close association with the Golgi (Figure 21). Comparison of the Nuclear Membrane in Amitotic and Interphase Cells The distance between the proximal edge of the inner nuclear membrane to the distal edge of the outer nuclear membrane was measured from random sections through the normal interphase and amitotic thread portion of the nuclear envelope. There were no sta- tistically significant differences at the 5% level in the intramembranous spaces between amitotic and Figure 20 75 Dense body from interphase cell in longitudinal section. Note microtubules (MT) inserting on one side, and closely associated golgi (G). x 164,000 Figure 21 77 Interphase cell. Note nuclear membrane (NM) with prominent nuclear pores, dense body (DB) with attached microtubules and closely associated golgi (G). x 20,000 ‘ 78 79 interphase cells (Table 1). Figure 20 further shows that a regular feature of the interphase nuclear envelope was the presence of numerous apparently regularly arranged nuclear pores. In an examination of over 50 individual photographs of amitotic threads no nucleous pores were ever observed. However, in the non-elongated region of the amitotic nucleus which is located in the central portion of each daughter cell, numerous nuclear pores were observed. Localization of the 7 nm Diameter Fibers When mitotic or interphase cells were glycerin extracted by the technique of a Ishikawa, gE_al. (1969), it was found that all or most of the cytoplasmic and nuclear structures were broken down. Only the plasma membrane and cortical area comprising a region approxi- mately 100 nm beneath the cell membrane remained intact (Figure 22). Imbedded within this electron diffuse region numerous 7 nm fibers are seen aligned parallel to the surface. If these glycerin extracted cells are homogenized rigorously in a teflon homogenizer in the cold and then spun down at 10,000 g the resulting pellet appears as shown in Figure 23. The presence of sheets of 7 nm fibers is apparent. These sheets are observed within the pellet even if the centrifugation is delayed for 3 hours. 80 Table 1 Analysis of variance of nuclear membrane diameters. Source of Variation Mean Squares Sum of Squares Sampling error .0005 .0501 Experimental error .0081 .1139 Exp. Error MS F 05 = = 16.20 n = 15 ° Sampling Error MS Amoeba Actin Assay Using the technique of Weihing and Korn (1969) an attempt was made to assay for the presence of actin in the soluble fraction of a whole cell homogenate. The binding of amoeba actin to rabbit myosin was demonstrated in the 10,000 g supernatant. This binding, demonstrated by an increase in visosity, was reversible by the addition of 0.00 1M ATP. Figure 24 gives the typical results of one of 4 separate experiments. The rabbit myosin was tested for actin contamination by measuring the change in viscosity on the addition of ATP. No change was detectable. The addition of 0.001 M ATP to whole living cells or to glycerine extracted cells produced no visable effects at the level of the phase contrast microsc0pe. Attempts to isolate amoeba actin using the methods of Miki-Noumura (1969) were not successful and resulted in a heterogeneous protein mixture (as measured by disc electrophoresis) which did not bind to rabbit myosin. 81 Figure 22 Glycerine extracted cell. Note cortical localization of microfilaments. x 14,000 Figure 23 10,0006 pellet. x 87,000 83 Note microfilaments (MF). 1" at. —- a 85 .cflmoafi uHonmu mo GOADAUUM ou Howum mad 0» omcommmu mo Roma muoc omam .maa zaoo. mo coflufieem may no mascawn mag» mo sufiaflawmum>mu muoz .ucmumcnomsm wooo.oa mnmoEm cm Ca unocomaoo C CD camome pagan“ mo mofiocwm em oudmwm 86 00¢ 13.35275: con 00" OO— — A q .3115 nth—oxi... . 0.. wave mh< L C C o... .L _ . C . l C \‘ C O \ 162:. s‘ovtu mp< .0 :3032 O 5 fl sesnugw u! AusoasgA 87 Demonstration of a Colchicine Bindipg Protein in Acanthamoeba Rhysodes Using the DEAE filter paper assay an attempt was made to demonstrate the presence of a protein component which would bind H3-colchicine. The basic assay procedure included comparisons of H3-colchicine binding between identical samples one of which was pretreated with excess cold colchicine. The difference between a given pair or series of these samples gave a roughly quantitative estimate of the presence of colchicine binding com- ponents. ASsays of Major Cellular FractiOns (a) Soluble and insoluble'components.--Experiments l, 2 and 3 (Table 2) were designed to determine whether any cellular components would bind colchicine and, if so, whether the binding site(s) was particulate or part of the soluble fraction. Due to the great variability in colchicine binding activity as well as the unstable nature of the colchicine binding component(s), no attempt to make quantitative estimations was done. Only the question of whether or not a soluble colchicine bind- ing component(s) was present was investigated. Therefore, the experiments listed in Table 2 represent a series of unreplicated single experiments which were varied slightly and designed in such a manner that taken as a whole the 88 Hmm cho pcmumc Tumm9m woman smHn omNHHHnmo>H mo Em no.0 CCNHHHnsHommm m ocHoH£OHoo oHoo mon unnumc m MHm uuwmsm comma anz cmuHHHsmosH no em no.0 emNHHHnsHommm H omOH cho ucmumcuomsm comma anm m omm mcHoHnoHoo oHoo msHm ucmumcuomam women anm H m 5mm ch0 ucmumcnomsm comma anm N HmH mCHoHnoHoo oHoo msHm ucmumcuomom women anm H v mHN umHHmm cmmmm smHm a ham DCMHMCHCQSm woman anm m owe cho uoHHmm ommmm anm m m How ocHownoHoo oHoo maHm uoHHom woman smHm H mHv ucmumcuomsm women anm m mom umHHmm ommmm smHm H m vmv cho ucmumcuomnm women anm m 05H ocHoHnoHoo oHoo msHm ucmumcuomsm women anm n mmw cho uoHHmm unannouomom woman nmwm m Hmm oCHoHCOHoo UHoo msHm umHHmm ucmumcuomsm Comma anm m NmSH cho mumuuHHm umHHmm wmmmm 3oq a H man oconnoHoo oHoo mon mumuuHHw uoHHom omomm 30H m omom cho umHHom voomm 30H m mhmm ocHoHnoHoo oHoo msHm uoHHom momma 30H H Honadz umnEnz Emu usofiumoua ucoaumoua ucoEHuomxm .moHuH>Huom mcHocHn ocHOHSOHoo How phonon momma HouHHm mama N oHnma M 89 .ooumum omquosuo mmoHcs us H x m OOO.OO~ u unnumcuomom voomm anm .mmmmm momma umuHHm on HoHnm Uonm an MC H How ocHoHSOHoonm HE\01H CH owumnaocH mm3 mHmEMm HE NIH m .mmmmo HHm CHM mun em you eHoo :H ummx can “mumsn mHonmznmonmnmz OhH CH coocommsmmu oumuHmHooum CDMMHom ESHCOEEM HE m m man wHoHMHooEEH ooummu oumuHoHoon MHMMHom EsHCOEEm HE m N NH mm cho gunman cH ocHoHnoHooumm H2 N H H: H x m ooo.oo~ ewOsMHnuamo can an: em OOH mHthmHo an wouHMmoolloumuHmHooum mumsdem EdHcoEE< N CCHoHnoHoo pHoo msHm HH Tums H x m ooo.oom OmmsmHuucmo can an: em mHmsHMHe an ONH wouHmmoouloumuHmHomHm DHMMHsm ESHCOE84 mo acmumcuomsm H mHCOIImua OH mnHH mHthoHo an oouHMmoonloumuHmHoon mumsmHsm EdHcoEE¢ m oaHoHnoHoo UHoo msHmTTmun OH OH own mHmmHme an oouHmmmouuopmuHmHooum muMMHSm ESHCOEE< H mHGOTIUouHmmoo OmHH uocunoumuHmHoonm uHMm CDMMHSm EDHCOEEm poocommomom m mcHUHnoHoo oHoo mnHmTToouHmmmp O «NO uocnnwumuHmHooum mDMMHDm EUHCOEEM concommsmom H OMO mcHummHQCH> z HOO.O msHm unnumcumm5m comma anm m mmv tho ucmumcummsm women anm m m «mm cho “mums H5 H H mmm tho ucmumcuomam women anm m omm Humz wow CH nmmz ummHITucmumcnmmsm Comma smHm H h Hoofidz Honfisz Emu ucmEumoHB ucofiumwua ucoEHHomxm AwmscHuaooO m mHnme 90 results would provide a qualitative answer to this question. The flow sheet for experiment 1 (Figure 1) describes how 20 mls of packed cells were fractionated to yield four separate subfractions. From Table 1 it is apparent that there is little, if any, bound colchicine present in either the low speed pellet or the filtrate from this pellet. Some bound colchicine appears to be present in high speed pellet and a marked amount is definitely bound to a component(s) in the high speed supernatant. In experiment 2 (Figure l) a slight increase in cpm is seen in the high speed supernatant relative to the high speed pellet. In experiment 3 (see Figure l) a very slight binding of colchicine in the high speed pellet is observed. It is also particularly interesting to note that the high speed pellet appears upon resuspension to liberate slightly more than 50% of its H3-colchicine (treatments 3 and 4). In experiments 4 and 5, a definite binding of colchi- cine to the soluble fraction (high speed supernatant) of the whole cell homogenate is again demonstrated. Experi- ment 6 shows that the colchicine binding characteristic of the high speed supernatant is maintained after storage as a freeze dried powder. Experiment 7 conclusively demonstrates the loss of the colchicine binding component from the DEAE filter paper when the filter paper is washed in 40% sodium chloride. 91 (b) Vinblastine precipitate.--Experiment 8 indicates that the addition of 0.001M Vinblastine to the high speed supernatant enhances the binding of the colchicine bind- ing component(s) to the filter paper. A visible precipi- tate was observed in the high speed supernatant when this fraction was incubated at 37 C for 1 hour in the presence of 10’3M Vinblastine. However, the precipitate later proved to be highly heterogeneous (as measured by disc electrOphoresis); and, in fact, a small precipitate could often be observed in high speed supernatants incubated at 37 for 1 hour without Vinblastine. (c) Ammonium sulfate precipitate.--In experiments 9 and 10 the colchicine binding activity present in the second ammonium sulfate precipitate (Figure 2) was assayed. It is apparent that the final ammonium sulfate precipitate does contain colchicine binding activity. It is also interesting to note that, although the conditions vary from experiment to experiment, the quantitative amount of bound labeled colchicine is greater than or well within the same order of magnitude as that present in the high speed supernatant (compare experiments 4, 5, 6, 7 and 8). In each of these cases each assay represents 5 ml of packed cells. It is also evident that the presence of residual amounts of ammonium sulfate does not greatly affect colchicine binding. 92 (d) Stability over time.--Experiments 11 and 12 were designed to determine whether the colchicine binding activity was stable over time. In experiment 11 a 24-hour dialysis was used to desalt the ammonium sulfate precipi- tate. It was found that the increased length of dialysis, followed by subsequent centrifugation, appeared to greatly reduce colchicine binding activity. The 24-hour dialysis against sodium phosphate-magnesium-GTP buffer induced the formation of a precipitate. This precipitate was removed by centrifugation at 15,000 g for 20 minutes. The removal of this precipitate was also necessary for the final purification of the protein (Figure 2). Experiment 12 indicates that the loss of colchicine binding activity observed for experiment 11 was probably due to denatur- ation over time and not to the loss of the precipitate. Puripy of Amoeba and Brain Tubulin The protein isolated by the modified Weisenberg technique was tested for purity using reduced fractions on 8M urea polyacrylamide gels. In all cases the protein was 90 to 95% pure as measured by densitometer quanti- tation of fast green stained gels. In Figure 25 a typical gel is shown. Preparations of comparable purity could also be obtained using the sephadex G-200 isolation pro- cedure as indicated in Figure 25. Ten electrophoresis experiments were carried out and, in each case where one of these two procedures was used, the purity was Figure 25 93 Acrylamide gel electrophoresis of brain and amoeba tubulin preparations in BM urea. All samples were first reduced in .12% mercap- toethanol followed by the addition of 8M urea. Gel G61 1 2 DEAE Purified Amoeba Tubulin--slightly overloaded on the gel to show contami- nents. DEAE Amoeba Tubulin. Gels 3, 4 and 5 = DEAE purified amoeba and Gel 6 Gels 7 Gel Gel Gel Gel Gel 10 11 12 13 brain tubulin--overloaded to show contaminents. Brain DEAE purified tubulin. and 8 = DEAE purified amoeba tubulin highly overloaded to show contaminents. Amoeba pooled sephadex G-200 second peak. Brain pooled sephadex G-200 second peak. Brain DEAE purified tubulin. Amoeba ammonium sulfate precipitate-- dialized and centrifuged at 200,000G. Amoeba DEAE purified tubulin. 94 9—-—>l3 8 l 95 comparable to that shown in Figure 25. To guarantee that all of the protein material entered the gels a 3% sample gel was occasionally used. No staining was ever observed at the sample-running gel interface when the protein con- centration was less than 25 ugm/gel. As the con- centration was increased both the brain and amoeba pro- teins formed aggregates which did not enter the gel. If these proteins were first reduced and alkylated no material remained at the top of the gel. Comparison of Amoeba and Brain Tubulins Electrophoresis (a) 12 cm gels.--Bovine brain tubulin isolated by the Weisenberg technique was run against amoeba prepar- ation on normal 8M urea gels as well as SDS molecular weight gels. In all of the 8M urea gel experiments amoeba tubulin did not co-electrophorese with brain tubulin. The brain protein always showed an Rf of .23 to .25 while that from the amoeba preps possessed an Rf of .175 to .195. Figure 25 shows a typical experiment in which the two proteins were run on separate gels of equal length in which the tracking dye (bromophenyl blue) was allowed to reach the end of the gels. When the pro- teins were run on the same gel they could always be sep- arated into two separate but closely running bands with 96 Rf values of .23 to .25 and .175 to .195. These values were obtained from 15 separate electrophoretic runs. (b) 20 cm_gels.--To increase the sensitivity of the electrophoresis experiments 14 inch gels with 8M urea were used to test the purity of amoeba and brain tubulin preparations. It was found that brain tubulin always separated into two closely running bands (Figure 26) amoeba preparations always ran as one single component. These results were always obtained when the proteins were reduced in urea-mercaptoethanol, or reduced and alkylated, with the latter treatment giving the most distinct sep- aration of brain tubulin. When measured quantitatively using fast green staining and densitometer tracings the two brain components were found to be present in equal amounts. PAS staining did not reveal any carbohydrate components of either brain or amoeba tubulin, although a very small contaminant carbohydrate was observed running just behind the tracking dye in the brain tubulin experi- ments. (c) SDS electrophoretic molecular weight determi- nations.--Molecular weight comparisons between the amoeba protein and brain tubulin were carried out on SDS poly- acrylamide gels. This experiment was repeated three times using preparations reduced in mercaptoethanol and urea or reduced and alkylated by the procedure of 97 .a3onm cmHm mH Hom mHnu mo coHuoom mo somumouoam Hmsuod .mHom COHEMHauommHom nous 2m OCOH Tuoom no can CHHsnau CHMHQ ooumHame can oucnomu mo OGHuooHu HmuoEOUHmcoa ON oHoOHm 98 uDOIhoum afioummHnousmc Eoum mmonu ou omummeoo GHHnnsu AI.TV MACOEM pom HTTTO GHMHQ mo mHthmcm oHom OCHEm mm ouanm 103 of. .3 =3 3: 22 :5 .26 o_< :o 2.. 3o :5 :: 3< 9< 1: 1.. _ , o I —- _. —. —- —- —. —- —- —- —- -. —- —- —I —- —- W 2 __ __ __ __ __ ._ __ __ __ __ __ __ __ __ __ __ a __ __ __ __ __ u n. __ __ __ __ __ =1: 1 I . _. __ _ _ _ . . _ _. _ _. _. .1 _ n. __ _. __ __ J __ _ __ __ __ __ __, __ __ . __ __ __ . __ __ __ __ __ __ __ __ __ __ "_ H u. __ "_ __ __ H _ r 2 __ __ __ __ __ __ __ __ __ __ .. __ __ n. __ __ __ __ __ __ "_ u. "_ __ __ _. n 2 _ __ : __ n. ._ . __ _ : __ __ __ __ _ _. __ . _. _ __ _ __ __ __ ._ __ . __ __ 1 _ _. _. __ __ __ __ u. T __ _m Thu. _ _ _ H - 104 colchicine binding activity due primarily to the length of time required to complete the technique. Hence, to make a valid comparison the brain protein which had been stored as a lyophilized powder was solubilized and stored for 28 hours in the cold. Twenty-eight hours was also the time required to obtain the purified amoeba prepar- ation. As can be seen in Table 3, when the proteins were quantitated using the Lowery method the brain protein showed a very high specific activity as compared to the amoeba protein which apparently denatures very rapidly. It should also be mentioned that the values for the amoeba preparation do represent a substantial increase over the controls since the total cpm for each amoeba filter paper disc was greater than 2,000 cpm. Table 3 Colchicine binding Specific activity experiments. Exp # Amoeba CPM/mg Brain H 0 850 1 145 (total) 52,761 311 cpm 930 2 223 (total) 40,359 400 cpm * Control 105 Gel Filtration (a) Brain tubulin.-—A1though a high Specific activity of colchicine binding was demonstrated for brain tubulin, further experiments were carried out to more precisely define the nature of this binding activity. When H3-colchicine (l uc/ml) was bound to the gross ammonium sulfate precipitate from a brain low-speed supernatant and aliquots run over Sephadex G-200, patterns such as that Shown in Figure 29 were consistently observed. This experiment which was repeated 4 times with identical results, compares optical density readings to cpm values from the same tubes. A great deal of important information can be obtained from these experiments. It is evident that the brain preparations separate into two major fractions with the majority of material eluting with the void volume. The second small peak represents an elution volume which indicates a gross molecular weight of 100,000 to 130,000 as measured using the formula, elutant vol - void vol = Kav/bed vol - void vol (Whitaker, 1953) (see Figure 30). This peak, then, probably repre- sents the tubulin dimer, the small shoulder following the second peak may indicate the monomer. Lastly, it is very clear that practically all of the bound counts correspond to this second peak. This experiment was repeated 5 times with identical results. Identical results were also obtained using the purified protein. 106 . “iv xmom .O.O vacuum 0:» nuH3 ucooHoHoo mH Hooov CSHoHnoHoo ocson on» «o HHm mHHMHusmmmo “map ouoz .ocHOHsoHOOT m CH oouonsocHoum soon on: pony mum» m THmHomHm ouMMHdm EDHSOEEM GHMHQ 6 mo mnmmumoumfiouno GESHoo OONIU xoomnmom mm ousmHm 0n 0N OH— H H 3MB; H\414I4u4l4l4|4.4\ \4 . ’4 ‘.|. o:_u_;u_09 1:303:04 [4! 4 \.|. 4 4\ / \M. 0.0\ 0. 4\ 4 43 lo 4\ / m \IO‘OIOIO/ 4 a 4 0/ 4:u. l \ . E 4 U a 00— 0.... o a OS X 'W'd'D pun '00 °‘. 000 l 00¢ 108 .mcHHsndu mnooem pom CHMHQ How cm>Hm omcmu >MM ouoz .AOOOHV madness Eoum coxmu OONTO xoomnmwm mom w>uso >mx .m.> DQOHmz HMHSooHoz om CHSOHE 109 oo— $1.23»; .0 :uo_o .0—2 0— u:__an:. oaooEo ace :_o.a .o 00:03 _X «.00 A 0... 110 (b) Amoeba tubulin.--When similar experiments were run using the gross ammonium sulfate precipitate from an amoeba 15,000 g supernatant, a pattern closely resembling that for the brain experiments was obtained (Figure 31). ” Here the second peak also indicates a molecular weight of 100,000 to 130,000. The amount of colchicine bound in each tube was too low to measure colchicine binding directly from column fractions as was done'for the brain protein. However, if a high Speed supernatant (200,000 g/ 1 hr) was used, the pattern seen in Figure 32 was obtained. Here the bound colchicine came off at the same elution volume as that of the second peak (Figure 30). To determine whether the second peak from the amoeba salt precipitate was pure and contained colchicine bind- ing activity the fractions from several runs were pooled and tested for purity and colchicine binding activity. The seCOnd peak proved to contain a highly purified pro- tein which co-electrophoresed with the protein isolated by the modified Weisenberg technique (Figure 25). The remaining portions of the second peak of two separate experiments were tested for colchicine binding activity (Figure 31 and Table 4). A comparison of the cpm values with the fractions from which they were taken makes it apparent that the highest binding activity is present in the second peak. 111 .14 mHnmav SHH>Huom maHeaHn mcHonOHoo mo mammmm now owns mTH mCOHuome mo mcoHuHmom wuoz .oumuHmHoon ouCMHSm ESHCOEEM mooo€m so no anonymoumfiouno CEdHoo OONIU xoomsmom Hm musmHm u asap 0 Q 0 n 0 N o / o HIAJ _ u _ co_.uo.m :o_.uo.m 010: o]. / o r /o I 0-0 o x) x a“. 0_. 2.3.; ES H _ H :o_.uo.m 113 .oconsoHoou m auH3 onMQSUCHoHQ m ucmumcumm5m 0 OOO.OON 030060 so No mammumoumsonno CSSHoo OONTU xoomnmmm mm musmHm On 114 .____...—-.__— 0? 2.32:3 332...: « on a. an: on 0— 3.... 0N 0? 00 00 XWdD OOZ 115 Turnover Experiments Non-Synchronized System To determine whether the synthesis and turnover of amoeba tubulin was regulated relative to the cell cycle or differed from that of the total soluble protein pool the reduction in specific activity of H3-1eucine pre- labeled protein was studied. The results of these experi- ments are graphed in Figures 33 and 34. In the first two experiments the cells were not synchronized. The values are plotted as a percentage of the initial specific activity, and the magnitude of each point has been increased in proportion to the percentage increase in total cell number. In these first two experiments it is evident that the turnover of the tubulin pool is very close or identical to that of the total soluble pool. Table 4 Amoeba colchicine binding activity from Sephadex Fractions. Fraction Number Exp # (see Fig ) CPM 1 2411 1 2 3023 3 1526 l 910 2 2 2523 3 850 116 .mun om umuHm 0:» Mom .mun ON mo CMHH MHmn muoz .oEHu O um mE\Emo H0000 gnu mo ommucmouom H mm nouuon one nuzoum How omuoonnoo mmsHm> mE\Emo .mHSpHSo omuHcousoahmlsoc m CH Ottdv mchuonm oHQSHom Hmuou one 64440 SHHSQSH mnooem mo Ho>ognsa mm ousmHm 117 «h em 7.301.223 on mu 0 In Bw/w a 3 as 00... 118 .oEHu O um mE\Emo Hou0p on» no omoucoouom m on oouuon poo £u3oum How oouoouuoo mosHo> mE\Emo .zusoum omosm moH CH poo noNHcouaocwmlcoc mos ousuHso one .muc ON mo osHo> oMHH MHmn m o>Hm ou oQOHm one no “TITO SOHuoHomouuxo ouoz A4440 mcHououm oHQSHOm Houou onu mo umau ou oouomeoo ¢¢¢¢V :HHonsu onooso mo Ho>osusa Om ousmHm 119 00 04 on 7.30:.oE: '6 «— mu 0 In 5W/Wd3 as . 00.. 120 Synchronized System In the third experiment (Figure 35) the cells were synchronized by thiamine starvation and again the Slope of the line for tubulin decay is equal to that of the total soluble pool. No change in slope is observed rela- tive to the two mitotic bursts. In all three experiments tubulin possessed a half life of 19-23 hours. Vinblastine Precipitation When Vinblastine at a concentration (10-3M) known to precipitate brain tubulin (Bensch, gt_al,, 1969) was added to the purified amoeba protein and incubated at 37° C for one hour a precipitate formed. This precipi- tate was centrifuged, reduced and run on 8M urea gels. A single band was observed which possessed an Rf of .190. 121 .oEHu O um mE\Emo Hmuou on“ no omoucoouom o no oouuon Ugo nu3oum How oouoouuoo mosHo> mE\Emo .Aooov HE\HoHoss mo gonads mm co>Hm open mHmoqu .ousuHso UoNHcouzocmm o SH “lily ocHououm oHnSHom Hmuou poo edddv :HHSQSD enoOEo mo Ho>ocuse mm musmHm 122 gm )2 Iw/WIMN # o 0 Q n g; 100, N p: O 4 50 37.5 25 Time(HourS) 12.5 DISCUSSION Comparison of Mitosis and Amitosis Role of Microtubules in Amitosis Colchicine, at the minimal concentration that is effective in inhibiting microtubule polymerization and mitosis, also greatly inhibits amitosis (Figure 3). This suggests a relationship at the level of microtubule for- mation between mitotic and amitotic cell division. How- ever, as is evident from the ultrastructure of the amitotic cell, neither mitotic apparatus formation, nuclear membrane and nucleolus breakdown, nor chromo- some condensation occurs during amitosis. Microtubules are present in low numbers within the region of the early amitotic cleavage thread (Figures 14, 15). These micro- tubules are an order of magnitude greater in length than any observed in the interphase cytoplasm (except those associated with the dense body). This suggests that the factors controlling amitosis may induce abnormal micro- tubule polymerization. These microtubules in the amitotic cell are not associated with any organizing center and are very few in number. They are never observed 123 124 terminating in the nuclear envelope. Therefore, the inhibitory effects of this drug on amitosis might be due to toxic side effects. This is in line with the results of workers such as Mueller, Gaulden and Drane (1971), Chakraborty and Biswas (1965) and others who have found that high colchicine concentrations reduce total protein and nucleic acid synthesis as well as inducing structural alterations in interphase and metaphase chromatin structure. For these reasons it is concluded that microtubules probably play little or no role in amitosis. Their only possible function could be their association with the dense bodies. Serial sectioning of amitotic cells will be needed to localize the dense bodies relative to the polarized nuclear elongation which occurs during early amitotic cytokinesis. Studies of this type might provide further evidence on the role of these unusual structures and their attached microtubules in amitosis. Function of Microfibers One general similarity between mitosis and amitosis appears to be the presence of sheets of 7 nm fibers associated with cytokinesis. Arnold (1969), Schroeder (1968) and others have reported similar fibers localized in the cortex of the advancing cleavage furrow of various cell types. Although no mitotic cells showing a telophase 125 configuration have as yet been studied, the presence of localized bundles of 7 nm filaments in the cortex of metaphase cells is suggestive of their possible role in normal cytokinesis. Since considerable motility both in nuclear movement and active cleavage is associated with the early and mid-phases of cytokinesis, some type of contractile system could be expected to be present. Weihing, gp_al. (1970) has demonstrated heavy meromyosin binding to cortical fibers of the same general appearance and diameter as those associated with the amitotic thread. Furthermore, the fact that these fibers are present in parallel alignment in a cylindrical sheet around the amitotic nucleus (Figure 17) also implicates them in the motile events occurring during amitosis. As reported earlier (Band, g£_gl., 1970) normal binucleate cells undergo a type of cytokinesis somewhat similar to amitotic cytokinesis in which initial active furrowing is accompanied by a pronounced migration of both daughter cells away from one another. It is possible that this latter aspect of binucleate cytokinesis is utilized during amitosis. The actin-like fibers might then play a role in daughter cell migration and the narrowing and constriction of the amitotic thread. Function of the Dense Bodies The dense bodies are clearly a consistent feature of both interphase and mitotic cells. They have not been 126 observed in cells actively undergoing amitosis. However, they have been seen in interphase cells from cell popu- lations in which more than 70% of the cells are under- going amitosis. These structures have previously been reported as dense material in interphase cells by Korn, gE_gl. (1969). From the results of this study it is apparent that these structures serve as centriole-like bodies for the insertion of Spindle microtubules (Figure 7). They also possess a bipartite structure as do most typical eukaryote centrioles (Figure 10). However, the fact that only dense bodies from mitotic cells have been observed to possess this bipartite structure suggests that replication of this structure occurs relatively late in interphase or perhaps in early prophase. Lastly, the close association in interphase cells of the nucleus and the dense bodies plus the fact that in almost every case the great majority of all of the microtubules insert on only one Side of this organelle, suggests that these structures may play a role in the determination of premitotic and amitotic nuclear polarity. An unusual characteristic of amitosis is the great amount of nuclear envelope elongation which accompanies amitosis. The amitotic thread often exceeds twenty cell diameters in length yet often the nucleus within the thread represents less than one tenth of the total 127 nuclear volume. The surprising result that there is no statistically significant decrease in the intranuclear envelope space during this extreme elongation implies that the membranes themselves are capable of extreme stretching or that new membrane formation is occurring. When the thread finally breaks, the two strands remain in the elongated state for a number of minutes. There- fore, nuclear and plasms membrane stretching, if they occur, must be of a inelastic nature. Since both outer and inner nuclear membranes were included in the nuclear space measurements it can be concluded that no drastic alterations in the diameters of these membranes occurs. Structural alterations of an order of magnitude of less than 2-3 nm in one of the membranes would not have been resolved by the techniques used. Furthermore, the absence of nuclear pores within the thread region is also difficult to explain. It is possible that membrane stretching induces structural changes which eliminate the molecular organization of the pores. Microtubule Polymerization and Nuclear Elongation There is one structural feature which is shared by both mitotic and amitotic cells. This is the presence of microtubules. In the metaphase cell the microtubules Show the typical eukaryote localizations, the one exception being the occasional unmistakable presence of 128 a single microtubule between sister chromatids. This may simply be an artifact of fixation or could represent a chromatin influence on microtubule polymerization. In amitotic cells microtubules lying outside of and parallel to the early amitotic nuclear thread are a consistent feature (Figures 1, 4, 2 and 15). These microtubules are usually an order of magnitude larger than any (with the exception of those associated with the dense bodies) seen in interphase cells (Figure 4). This implies that they are polymerized dg_poyo or polymerized upon pre-existing short interphase microtubule segments. The work of Stephans (1968) on the requirement of seeder segments for microtubule polymerization implies the latter explanation as being more likely. However, the role of seeder micro- tubules in vivo remains obscure. The colchicine inhibition of amitosis could be due to nonspecific toxicity. The low numbers of microtubules associated with amitosis and the fact that none insert into the nucleus speaks in favor of a non-functional role for these microtubule segments in amitosis. The problem of what provides the force to stretch and maintain the nuclear elongation probably can- not be answered by involving microtubules. The presence of terminal convolutions and longitudinal infolding of the amitotic nuclear membrane does suggest that a positive pulling force is involved. This force does not appear to be transmitted by microtubules. It is apparent that many 129 questions remain to be answered. This system does, how- ever, provide an interesting opportunity for the study of the cellular polarity, physical characteristics of living membranes and the factors associated with microtubule polymerization. The use of different fixation methods which might reveal some of the structural components not visible under the conditions used here might give insight into some of the questions posed by this study. Amoeba Actin It is evident from the work of Weihing, et a1. (1966, 1970) and the experiments reported here (Figure 24) that an actin-like protein is present in Acanthamoeba. The failure to isolate this protein using the methods of Miki-Novmura (1969) could have been due to the fact that much of this protein remained within the cortical matrix after homogenization (Figure 23). This fact might be of more general importance. It implies that the gel-like cortical region is quite stable and may control the localization of F-actin, thereby regulating some of the actin-dependent phenomena such as cytokinesis and pseudo- pod formation. Colchicine Binding to Amoeba Protein The experiments described in Table 1 clearly indi- cate that colchicine does bind to a component(s) largely present in the soluble fraction of the cell. The fact 130 that in all cases the amount of bound colchicine appears to be very low is probably due to a combination of factors. The colchicine binding factor could be present in very low concentrations, or this factor could simply possess a low affinity for colchicine relative to brain tubulin. Lastly, the protein could be very unstable and denature prior to the time of assay. Experiments 10, 11 and 12 (Table 1) indicate that this latter possibility is at least partially responsible for the low binding activity. Experiments 1-6 definitely show that the most of the colchicine binding activity is present in the soluble fraction (200,000 g supernatant) although some binding might be present in the particulate fractions. Shelanski (1971) has found slight colchicine binding activity in membrane fractions from mammalian brain. In experiment 7 the fact that the colchicine binding factor is ionically bound to DEAE paper is demonstrated. The 40% NaCl wash is of the same concentration that Weisenberg, gE_31. (1968) found to be minimal for removal of brain tubulin from DEAE filter paper. Experiment 8 is particularly interesting. Here colchicine binding is enhanced by the addition of vin- blastine. In light of the fact that a small precipitate could be induced in purified amoeba protein preparations by Vinblastine, it would appear that the Vinblastine induced aggregation of the colchicine binding protein 131 and thereby enhanced its retention on the filter paper by simple mechanical filtration. Experiments 9-12 Show that much of the colchicine binding activity is retained in the weisenberg ammonium sulfate precipitate (Figure 2) which further demonstrates the similarity between the colchicine binding activity in amoeba and mammalian brain. Comparison Bepgeen Amoeba and Brain Tubulin Electroph0§ytic Mobility The results of electrophoresis, amino acid analysis and column chromatography definitely demonstrate the similarity between amoeba and brain tubulin. Important differences were also observed. Gel electrophoresis on 8M urea gels (Figure 25) revealed that the proteins obtained from the standard and modified Weisenberg techniques were more than 95% pure and the brain protein possessed an Rf value range identical to that described by Weisenberg, et_gl. (1968). The fact that the amoeba protein showed a consistently lower Rf value indicates that this protein possesses a more positive net charge at pH 8.2 than brain tubulin Since as little as one amino acid substitution on a given protein can be translated into a measurable Rf change (Panyim and Chalkley, 1969) the number of amino acid differences between these two proteins need not be very great. That these differences are not great is 132 further indicated by the similarity of the means of the two Rf values (.195 and .235) and the fact that both proteins are purified using the same Kcl concentrations over DEAE Sephadex. Lastly, by doubling the length of the 8M urea gels the apparently Single brain tubulin band separates into two separate components (Figure 26). The presence of two tubulin components differing by a molecular weight of 3,000 and possessing Slightly different amino acid compositions has been recently investigated by Shelanski (1971), Rosenbaum, gE_gl. (1971) and Bryon (1971). Shelanski (1971) has found that these two components are synthesized at equal rates in mouse brain and he feels that they represent the two components of the 6s dimer. Since the amoeba protein did not separate into two components on long gels, the possibility exists that amoeba tubulin represents the smaller of the two brain components. Molecular Weight Comparisons The molecular weight differences described in the results section are more puzzling. The Weber SDS elec- trophoretic technique did not give a consistent value for brain tubulin (Figure 27). On two occasions the value was 5,000 to 10,000 below the values previously published (Olmstead, e£_21., 1969). On another occasion the brain tubulin split into two components, one at 60,000 and the 133 other at 35,000. In light of the fact that Weisenberg, gE_§l. (1968) also often found a 33,000 M.W. component in their native preparations it is interesting to speculate that the protein might spontaneously break down during storage. The amoeba protein did Show a molecular weight 5,000 to 10,000 below that of brain tubulin except in the experiment in which it was coincident with the 35,000 mw component. The inconsistency of these results casts serious doubt on their validity. The presence of the carbohydrate contaminant mentioned in the results, or differences in a polysaccaride component reported to be covalently bound to brain tubulin (Falxa and Gill, 1970) could interfere with the migration in SDS gels. Shelanski (1971) does obtain consistent values on SDS gels but he finds that when the protein is placed in an SDS spacer gel for 1 week prior to electrophesis the protein splits into two separate bands. The same result was obtained for one of the above described SDS runs. Here the lower brain components is very close to that of the amoeba pro- tein. In light of the fact that the amoeba protein does not separate into two components on extra long urea gels, it may be that in Acanthamoeba, tubulin is present only as the smaller of the two brain components. As mentioned above, Shelanski (1971) has obtained some evidence to indicate that the two brain tubulin components represent 134 the two subunits of the stable dimer. If this is correct then one would expect that there would be distinct molecu- lar differences between the protein which forms a stable dimer with itself and one that forms a dimer with a second slightly different protein. On this basis it is not very surprising that the amoeba protein shows slight charge and molecular weight differences compared to brain tubulin. However, more work needs to be done before any definite conclusions can be drawn from molecular weight determinations of this type. Amino Acid Analysis The close Similarity in amino acid molar ratios of amoeba, brain and previously published brain tubulins (Figure 28) was somewhat unexpected considering the possible molecular weight differences described above. From a careful comparison of the individual amino acid ratios it is apparent that, with the exception of lysine and glutamic acid, the differences between the brain preparations and published tubulin values (e.g., Olmstead and Rosenbaum, 1969) is equal to or greater than the dif- ferences between the amoeba and brain preparations for those amino acids tested. The low proline values are probably due to the amino acid analyzer system which did not completely separate proline from glutamic acid. If the difference in molecular weight indicated by the SDS gel experiments should prove correct then the only way 135 to explain an almost identical amino acid composition would be to postulate that the protein is made up of a number of identical segments with the total amino acid composition uneffected by the loss of one segment. Gel Filtration Column Chroma- tography The results provide a consistent pattern of molecular separations on Sephadex G-200 for both brain and amoeba tubulins. By comparing Figures 29 and 31 it can defi- nitely be seen that the ammonium sulfate precipitates from both brain and amoeba resolve into two components, a large probably heterogeneous void volume and a much smaller (100,000 - 130,000 MW) homogeneous component. That this smaller peak from the brain preparations, which Weisenberg, gE_§l, (1968) demonstrated to be the dimer, actually binds almost all of the bound colchicine is also evident from Figure 29. Since the Specific activity of the purified protein or the gross ammonium sulfate precipitate from amoeba was extremely low (Table 4), the total high speed supernatant had to be used to demonstrate the shape of the bound colchicine peak (Figure 32). It is obvious that colchicine is bound although the actual amount of bound colchicine relative to unbound colchicine is very low. Again the bound colchicine peak is within the 100,000 - 130,000 molecular weight range. Hence, it would appear that both brain and amoeba tubulin bind 136 colchicine to a similar 100,000 to 130,000 MW component. This similarity in colchicine binding is further verified by the experiment Shown in Figure 25 in which the homo- geneity of the amoeba colchicine binding component is clearly demonstrated. It can therefore be concluded that the electrophoretically pure amoeba protein isolated by the modified Weisenberg technique is present (at least in part) in the native form as a 100,000 - 130,000 aggregate which specifically binds colchicine. Turnover of Amoeba Tubulin The results discussed above demonstrate that the protein isolated from Acanthamoeba rhysodes by the modi- fied Weisenberg technique does bind colchicine and that it possesses sufficient characteristics in common with the standardized brain tubulin to justify the assumption that this protein is indeed the structural protein of amoeba microtubules. The last phase of the research pro- gram was then initiated to determine whether the synthesis of amoeba microtubule protein was regulated relative to that of the total soluble pool and to the morphological events associated with mitosis. Studies of the turnover of tubulin in synchronized and non-synchronized cultures provided a method by which both the synthesis and degre- dation of this protein could be monitored. The results plotted in Figures 33, 34 and 35 demonstrate little or no regulation in the synthesis of 137 amoeba tubulin. In the non-synchronized cultures (Figures 33 and 34) the decay in specific activity of tubulin is very close to that of the total soluble pro- tein pool. The fact that the growth-corrected plot is linear for the first 27 hours implies that degradation and reincorporation of labeled proteins is not great enough to affect these results. The tapering off observed after 30 hours is probably due to reduced cellular metab- olism caused by the cultures becoming overly concentrated. This was reflected in the reduction and, finally, cessation of cellular growth after 30 hours. The deviation from linearity for the third point of the second half life experiment (Figure 33) is probably due to an unexplained 10% reduction in cell number which occurred between 9 and 19 hours. A half life of approximately 19 hours for both total soluble proteins and tubulin was recorded during the linear phase. Since the slope of these curves is a function of both the synthesis and breakdown of the cellular proteins, it would appear that within experi- mental error tubulin is synthesized and at the same rate as the rest of the soluble pool. This agrees very well with the recent report of Feit, Dutton, Borondes and Shelanski (1970) which found that brain tubulin possesses a half life equal to that of the total cellular proteins. These are not the results one would expect if the pro- duction of microtubules was regulated by the synthesis 138 of tubulin. In Acanthamoeba the only known function of microtubules is in mitosis. Therefore, the microtubule protein from mitosis would have to be kept well below threshold levels prior to prOphase for regulation at the level of synthesis to occur. This would require that either the protein would break down just after telophase followed by normal synthesis until threshold levels are reached, or that the synthesis of the protein would be reduced until just before prophase. A combi- nation of these events could obviously also be involved. The results described here indicate that rapid turnover greater than that for all soluble proteins is not occur- ring. This is not the expected result if the levels of the protein were controlled by a sudden synthesis at prophase or rapid breakdown after telophase. The experiment plotted in Figure 35 from a syn- chronous culture helps to further define these questions. If tubulin were synthesized at a greatly increased rate just prior to mitosis a dip in the Slope of the curve prior to mitosis would be recorded. From Figure 35 it is apparent that this does not occur. No fluctuation in the turnover of labeled tubulin occurred during the two mitotic bursts. From these results and those of Robbins and Shel- anski, gp_al. (1969), who found that tubulin was synthe- sized throughout the cell cycle, the following two 139 possibilities exist. First, perhaps tubulin synthesis is regulated relative to mitosis but the magnitude of this regulation is too small to be detected given the level of experimental error present in both this and Shelanski's studies. Secondly, tubulin synthesis Simply might not be regulated at all relative to the cell cycle and the production of microtubules might depend upon other factors. These factors could include tubulin locali- zation in specific cellular regions or the localization of catalytic factors. Nevertheless, the studies of Inoué, Gross, Shelanski and Mazia (described and in the literature review) and the research reported here do point to the second explanation as being the most reasonable given the data available at present. LITERATURE CITED Adelman, M., Borisy, G., Shelanski, M., Weisenberg, R. and Taylor, E. (1968). Federation Proceed. 21, 1186-. Afzelius, B. (1963). J. Ultrastruct. Res. 9, 381-. André, J. and Thiéry, J. (1963). J. Microscopie 2, 71-. Andrews, P. (1964). Biochem. J. 21, 222-. Arnold, J. (1969). J. Cell Biol. 41, 894-. Auber, S. (1962). Compt. Rend. 254, 4024-. Auclair, W. and Siegel, B. (1966). Science 154, 913-. Band, R. N. (1959). J. Gen. Microbiol. 31, 80-. Band, R. N. (1962). J. Protozool. g, 377-. Band, R. N., Mohrlok, S. (1969). J. Protozool. 16, 35-. Band, R. N., Mohrlok, S. and Rubin, R. (1970). Nature 227, 379-. Barnicat, N. (1966). J. Cell Sci. 1, 217-. Behnke, O. and Forer, A. (1967). J. Cell Sci. 2, 169-. Bensch, K., Marantz, R., Wisniewski, H. and Shelanski, M. (1969). Science 165, 495-. Bibreing, T. and Baxondall, J. (1969). J. Cell Biol. 41, 577-0 Bowers, B. and Korn, E. (1969). J. Cell Biol. 32, 95-. Chakraborty, A. and Biswas, B. (1965). Exptl. Cell Res. 38, 57-. 140 141 Clark, J. (1964). Annals N.Y. Acad. Sci. 121, 428-. Coyne, B. and Rosenbaum, J. (1970). J. Cell Biol. 41, Davis, M. and Smith, C. (1967). Exptl. Cell Res. 12, 15-. Duprow, E. (1968). In: Cell and Molecular Biology. Academic Press, New York, pp. 591-. Dutton, G. and Barondes, S. (1969). Science.l66, 1637-. Falxa, M. and Gill, T. III (1969). Arch. Biochem. Biophys. 135, 194-. Fawcet, D. and Porter, K. (1954). J. Morphol. 24, 221-. Feit, H., Dutton, G., Barondes, S. and Shelanski, M. (1970). J. Cell Biol. Forer, A. (1965). J. Cell Biol. 2§_(Suppl.), 95-. Gall, T. (1966). Natl. Cancer Inst. Monogr. 23, 475-. Gibbons, I. (1961). Nature. 190, 428-. Gibbons, I. (1963). Proc. Natl. Acad. Sci. U. s. g9, 1002-. Gibbons, I. (1967). In: Molecular Organization and Biological Function. J. M. Ailen, ed., Harper, New York, pp. 211-. Gibbons, J., Tilney, L. and Porter, K. (1969). J. Cell 3101. 5);, 201-0 Grafstein, B., McEwen, B. and Shelanski, M. (1970). Nature. 227, 289-. Grinstone, A. and Klug, A. (1966). J. Cell Sci. 1, 351-. Gross, P. (1970). Presented at the "First International Biological Congress," Detroit. Hatono, O. and Cosaura, S. (1966). Biochem. Biophyp. ACtao 127’ 488-0 Hiramoto, Y. (1956). Exptl. Cell Res. 11, 630-. Hiramoto, Y. (1965). J. Cell Biol. 25, (suppl.), 161-. 142 Hodge, L., Robbins, E. and Scharff, M. (1969). J. Cell Biol. 49, 497-. Inoué, S. (1959). In: Biophysical Sciencg::A Study Pro- gram. J. L. Oncley, ed., Wiley and Sons, New YOEE, pp. 402-. Inoué, S. (1953). Chromosoma. 5, 487-. Inoué, S. and Sato, H. (1967). Symp. N.Y.S. Heart Assn. Ishikawa, H., Bischoff, R. and Holtzen, H. (1969). g. James, K. and Austin, L. (1970). Biochem. J. 117, 773-. Kane, R. (1967). J. Cell Biol. 32, 143-. Kawamura, K. (1960). Exptl. Cell Res. 21, 9-. Kawamura, K. and Dan, K. (1958). J. Biophys. Biochem. Cytol. 4, 615-. Keefer, B., Sakai, H., Salari, A. and Mazia, D. (1966). J. Mol. Biol. 22, 75-. Ledbetter, M. and Porter, K. (1964). Science. 144, 872-. Lowery, 0., Rosebrough, N., Farr, A. and Randall, R. (1951). J. Biol. Chem. 193, 265-. Luft, H. (1961). J. Biophys. Biochem. Cytol. g, 409-. Manton, I. (1952). Symp. Soc. Exptl. Biol. _6_, 306-. Mazia, D. (1952). In: Modern Trends in Physiology and Biochemistry., E. Barron, ed., Academic Press, New York, pp. 77-. Mazia, D. (1961). In: The Cell Vol. III. Bracheford and Mirsky, eds., Academic Press, New York, pp. 80-. Miki-Noumura, T. (1968). Exptl. Cell Res. 52, 54-. Moor, H. (1967). Protoplasma. 64, 89-. 143 Mueller, G., Esther, M. and Drane, W. (1971). J. Cell BiOlo 4_8, 253-0 Nicklas, R. (1967). Chromosoma. 33, 17-. Olmsted, J., Carlson, K., Klebe, R., Ruddle, F. and Rosenbaum, J. (1970). Proc. Natl. Acad. Sci. 33, 129-. Panyim, S. and Chalkley, R. (1969). Biochem. 3, 3972-. Pease, D. (1963). J. Cell Biol. 33, 313-. Peters, A. and Vaughn, J. (1964). J. Cell Biol. 2, 113-. Pollard, T., Shelton, E., Weihing, R. and Korn, E. (1970). J. Mol. Biol. 39, 91-. Renaud, F., Rowe, A. and Gibbons, I. (1968). J. Cell Biol. 33, 79-. Robbins, E. and Shelanski, M. (1969). J. Cell Biol. 33, 371-. Roberts, R., Cowie, D., Abelson, P., Bolton, E. and Britten, R. (1955). In: Studies of Biosynthesis in Escherichia coli. Carnegie Institute, Washing- ton, pp. 13-. Roth, L. and Jenkins, R. (1962). Proc. 5th Intern. Conf. Electron Microscopy. 3, NN-3. Ruby, A. (1961). Doctoral Thesis, University of Cali- fornia, Berkely, California. Sakai, H. (1966). Biochem. Biophys. Acta. 112, 132-. Sakai, H. (1968). Intern. Rev. Cytol. 33, 89-. Sakai, H. and Dan, K. (1959). Exptl. Cell Res. 33, 24-. Schroeder, T. (1968). Exptl. Cell Res. 53, 272-. Shelanski, M. (1971). Personal Communication. Sisken, J. and Wilkes, E. (1967). J. Cell Biol. 33, 97-. Stephans, R. (1967). J. Cell Biol. 33, 255-. Stephans, R. (1968). J. Mol. Biol. 33, 517-. Stephans, R. (1970). Science. 168, 845-. 144 Stephans, R. and Linck, R. (1969). J. Mol. Biol. 33, 497-0 Stern, H. (1958). J. BiOphys. Biochem. Cytol. 3, 157-. Szollosi, D. (1968). J. Cell Biol. 3, 133-. (abstract) Takeda, S. and Izutoso, K. (1960). Symp, Cell Chem. (Japan) 33, 245-. Tilney, L. and Gibbons, J. (1968). ProtOplasma. 33, 167-. Weber, K. and Osborn, M. (1969). J. Biol. Chem. 244, 4406-0 Weihing, R. and Korn, E. (1969). Biochem. Biophys. Res. Com. 33, 906-. Weisenberg, R., Borisy, G. and Taylor, E. (1968). Biochem. 7, 446-. Weisenberg, R. and Timasheff, S. (1970). Biochem. 3, 4110-. Went, H. (1960). Ann. N.Y. Acad. Sci. 32, 422-. went, H. (1966). Protoplasmatologia. 3, 24-. Wessels, N., Spooner, B., Ash, J., Bradley, M., Luduena, M., Taylor, E., Wrenn, J. and Yamada, K. (1971). Science. 171, 135-. Whitaker, J. (1963). Anal. Chem. 33, 1950-. Wilt, F., Sakai, H. and Mazia, D. (1967). J. Mol. Biol. 27' 1-. WOhlfarth-Bottermann, (1964). In: Primitive Motile Systems ipCell Biology. Allen and Kamiya, eds., Academic Press, N.Y. Zimmerman, A. (1963). In: The Cell in Migosis. Livine, ed., Academic Press, New York, pp. 159-. Zirkle, R. (1957). Advances in Biol. Med. Phyg. 3, 104-. AA“ TIT ”N S" R" E THREAT: AT": 4 6112 0317