Citation
Normal and abnormal mitosis in a mammalian cell in vitro : a light and electron microscopic study

Material Information

Title:
Normal and abnormal mitosis in a mammalian cell in vitro : a light and electron microscopic study
Creator:
Roos, Urs-Peter, 1938-
Publication Date:
Language:
English
Physical Description:
xvii, 276 leaves : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Anaphase ( jstor )
Cells ( jstor )
Centrioles ( jstor )
Chromosomes ( jstor )
Kinetochores ( jstor )
Metaphase ( jstor )
Mitosis ( jstor )
Mitotic spindle apparatus ( jstor )
Nuclear membrane ( jstor )
Prophase ( jstor )
Cells ( lcsh )
Dissertations, Academic -- Zoology -- UF
Mitosis ( lcsh )
Zoology thesis Ph. D
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis - University of Florida.
Bibliography:
Bibliography: leaves 260-275.
General Note:
Manuscript copy.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
This item is presumed in the public domain according to the terms of the Retrospective Dissertation Scanning (RDS) policy, which may be viewed at http://ufdc.ufl.edu/AA00007596/00001. The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator(ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
022285031 ( ALEPH )
13639421 ( OCLC )

Downloads

This item has the following downloads:


Full Text














NORMAL AND ABNORMAL MITOSIS IN A MAMMALIAN CELL IN VITRO.
A LIGHT AND ELECTRON MICROSCOPIC STUDY.














By

Urs-Peter Roos


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY





UNIVERSITY OF FLORIDA 1971





























Thus, the task is, not so much to see what nobody has seen yet; but to think what nobody has thought yet, about what everybody sees.

Schopenhauer














ACKNOWLEDGMENTS

I am grateful to Dr. Fred C. Johnson, Chairman of my Supervisory Committee, for his help in procedural matters.

Words cannot express my respect and gratitude for the liberal and stimulating guidance provided by Dr. Henry C. Aldrich, Co-chairman, throughout this study. His generosity regarding time and use of facilities was an important factor for the progress of my work.

I am much indebted to Dr. John W. Cramer and his staff, Department of Pharmacology, for instruction and use of facilities for tissue culture work.

Dr. Lynn H. Larkin and Dr. James L. Nation, members of my

Supervisory Committee, deserve my thanks for their moral support and helpful advice. I appreciate the encouragement provided by Dr. John W. Brookbank, Dr. James H. Gregg, and Dr. Philip B. Morgan.

I am much obliged to Dr. Robert I. De Witt, Chairman of the Department of Zoology, for arranging financial assistance.

I also extend my thanks to Mrs. Rosemary Rumbaugh, and numerous

unnamed persons with whom I was associated in the course of this study, for assistance and forbearance.

Mfy wife, Loan, played a major part in this accomplishment. I am deeply grateful for her never faltering patience, her understanding, encouragement, and help.


iii















PREFACE


Certainly, one cannot argue about the importance and significance of attempts to increase our knowledge of chromosomes and mitosis, both so basic to life in higher organisms. I was fortunate to have the liberty to choose a research project myself, and the study presented evolved mainly out of my curiosity to learn more about these two fundamental aspects of life.

At the time this project began, ultrastructural investigations of eukaryotic chromosomes had essentially reached a standstill. More new questions had been posed than old ones answered. On the other hand, electron microscopic techniques had not yet been applied to the study of chromosomal aberrations. Mitosis in animal cells was fairly well documented at the ultrastructural level, but many of its aspects were the subject of controversy. Like many other students at the beginning of a new road, I had only a limited knowledge of these problems; hence my idealistic assumption that an ultrastructural investigation of normal and abnormal mitosis could answer many of the questions remaining, or at least lead the way to significant experiments. It will become clear from the following presentation how far these hopes were fulfilled and justified.
















TABLE OF CONTENTS


Page

Acknowledgments . . . . . . . . .. i

Preface . * * * *. * * iv List of Tables ......... .. ... .. vii

List of Figures . . . . . ...... viii

Key to Abbreviations and Symbols . * . . . xiv Abstract . . * * * * * * * * * xv Review of Literature ................. 1

General Ultrastructural Features of Vertebrate 1
Cells in Mitosis . . ... . . . ..

Centrioles . . . . . * 5

Spindle Fibers ........... . 11

Chromosomes 9 * * * * * * 20 Kinetochore Structure and Function ........ 24 Chromosomal and Mitotic Aberrations . . .. 33 Statement of Purpose . . . . . . . 44

Materials and Methods .........* *....... 45

Cell Culture . . . * * 45 Chemical and Physical Treatments ... . . 46 Fixation and Embedding . . . . . . . 48

Preparation of Cells for Light and Electron
Microscopy . . . . . . . . 49

Results and Observations . . . . . . 52

Normal Mitosis . . . . . . . . . 52









Page

C-Mitosis . . . ............. 67

Mlitosis in Cold-Treated Cells . . . . . 70

Kinetochore Fine Structure . . . . . . 71

Chromosomal and Mitotic Aberrations . . . 77 Discussion . . . . . . . * *... 225

Centrioles . . . . . . . . . *. 225

Ificrotubules .. . . . .. . . .. .. 230

Chromosomes . . . . . . . . .. 235

Kinetochores .. ...... ........... 240

Nuclear Envelope . . .. . . . . .. 257

Conclusions . . . . . . . . . . .. 259

References . . . ..... ............ 260

Biographical Sketch . . .. . . . . . 276














LIST OF TABLES


Table Page

1 Frequency of chromosomal and mitotic aberra- 79
tions in untreated meta- and anaphase
cells. . . . . .. . . . . .

2 Hitotic indices for streptonigrin-treated and 83
untreated control cells . . . . ...

3 Frequency of chromosomal and mitotic aberra- 84 tions in streptonigrin-treated cells . .















LIST OF FIGURES


Figure Page

1 Karyotype of the male rat kangaroo (Potorous 45
tridactylis) . . ....****

2 The three planes in which blocks were 51
sectioned . . . * * * * #

3 Grazing section of an interphase nucleus . 87 4a-d Centriole of interphase cell . . . . 89 5a-c Transverse sections of interphase nuclei . 91 6 Chromatin fibers of interphase nucleus . . 93 7 Nucleolus of interphase cell . . . . 93 8a Very early prophase cell . . . . .. 94

8b-e Centriole duplication in very early 96
prophase . . . . . . . . .

9a,b Migration of centrioles in very early 97
prophase . . . . . . ....

9c-e Fine structure of centriole in very early 97
prophase . . . . . . .

10 Mid-prophase nucleus ............. 98

lla Mid-prophase nucleus . . . . . .. 99

llb Mid-prophase nucleus, peripheral section . 100 12 Early prophase. Grazing section of the 102
nucleus . . . . . . . . .

13 Early prophase. Transverse section of the 102
nucleus . . . . . . . . .

14 Late prophase chromosome near nuclear 104
envelope . . . . . . . . *


viii










Late prophase chromosomes with
kinetochores . . . . . .

Mid-prophase chromosome with kinetochore


Figure

15 16 17


18a 18b

19 20

21a-c 22a,b 22c,d

23a-c 24a-c

25

26 27 28

29 30 31 32

33a 33b

34a-c 35a-c

36


Mlid-prometaphase chromosome


Mid-prophase chromosome with sister
kinetochores . . .. .

Very early prometaphase cell .... Very early prometaphase cell ... Early prometaphase cell ..... Early prometaphase cell * *.... Early prometaphase. Nuclear envelope Early prometaphase kinetochores . Early prometaphase kinetochores . Early prometaphase kinetochores . Early prometaphase kinetochores . Early prometaphase kinetochores . Mid-prometaphase cell . .. ..... Mid- to late prometaphase cell . . Late prometaphase chromosomes .... Late prometaphase cell . .... Late prometaphase chromosomes .*.. Late prometaphase chromosomes .*. Hid-prometaphase microtubules .... Late prometaphase chromosomes .... Late prometap-ase chromosomes .. Late prometaphase chromosome .. Late prometaphase chromosome .*...


S . . . .


Page


104

106 106


* S *













* . S



* S S

* S *

* S S S





* . .

* S S


107

108 109 110

112 114 116 118

120 121 122

123

124 125 127

127

128

129 130 132

134 136








Figure Page 37a,b Late prometaphase chromosome . . . . 136 38ab Late prometaphase chromosome . . * 138 39a-e Late prometaphase chromosome . . . . 140 40a-c Late prometaphase chromosomes . . . . 141 41 Late prometaphase kinetochore . . . . 142 42 letaphase plate ............... 143

43 Metaphase chromosomes (para-sagittal
section) . . .. . . . . . 144

44 Metaphase plate (para-equatorial section) . 145 45 Very early anaphase cell . . . . . 146

46 Very early anaphase chromosomes . . . 147 47 Early anaphase cell . . . . . .. 148

48 Very early anaphase chromosomes . . . 150 49 Early anaphase kinetochores . . . . 150

50 Mid-anaphase cell (para-equatorial
section) .. . . . . . . . 151

51a,b Late anaphase spindle and chromosomes . . 152 52 Very late anaphase, daughter nucleus . . 153 53 Very late anaphase, kinetochore . . . 155 54 Very late anaphase, stem bodies . . . 155 55a Very late anaphase, stem body . . . . 155 55b Very late anaphase, equatorial region . . 157 56 Early anaphase centrioles . . . . . 157 57 Early telophase nucleus . . . . . 158

58 Early telophase nucleus . . . . . 160

59 Mid-telophase kinetochores . . . . 160 60 Hid-telophase nucleus . . . . . .. 161








Figure

61 62

63 64 65 66 67 68 69 70

71 72 73 74 75 76

77


78a-c


78d, e


Cytokinesis . . * * * Cytokinesis, midbody * * *. C-metaphase * * * *. * C-metaphase . . . . . .

C-mitosis . ... . . . .

C-mitosis . ..........

Colcemid-treated interphase nucleus Colcemnid-treated interphase cell . C-mitosis . .. .... . .

Cold-treated cell in mid-prophase Cold-treated cell in mid-prometaphase Prophase kinetochores of cold-treated Microtubules of cold-treated cell in
metaphase . . . . .....

Cold-fixed control cell in metaphase Cold-treated cell in prometaphase Cold-treated cell in cytokinesis . Centrioles of cold-treated cell in
metaphase . . . . . .

Serial sections of early anaphase
kinetochore . . . . .. Serial sections of early anaphase
kinetochore (cont'd.) . . ..


. . .
. .


. .


Very early anaphase kinetochore .. .


Kinetochore in late prometa- to metaphase . 1etaphase kinetochore . .. . .. .. Kinetochore in very early anaphase ...... Sister kinetochores in very early anaphase .


Page 162 163

164 165 166 167 169 169 169 171 171 173


ce* ll
* *9

cell


80 81

82 83a, b


. *a


. .








Figure

84

85a-c 86a-h 87a,b


88 89 90


91a,b


92 93 94

95a-c

96 97 98 99 100 101


102a,b 103a, b


sections) .a . .


. . . . a .


Kinetochores in c-mitosis . . Kinetochores in c-mitosis .. Kinetochores in c-mnitosis . . Kinetochores in c-mitosis .. Kinetochores in c-mitosis . Kinetochore in c-mitosis ... Kinetochores in c-mitosis .. Untreated anaphase cell, dicentric
bridge . . . . . .. .

Untreated anaphase cell, dicentric
bridge . . . . ...

Untreated anaphase cell, dicentric


bridge .


. . a a a a a a a a a a


kinetochores of
. . . . .n

kinetochore of non. . a a . . .


Untreated anaphase cell,
lagging chromosomes .

Untreated anaphase cell,
lagging chromosome .


xii


Early anaphase kinetochore . . ... Hetaphase kinetochore (para-sagittal
sections) . . . . . . . .

Hetaphase kinetochore (para-equatorial
sections) . . . . . .....

Two metaphase kinetochores (para-equatorial
sections) . . . . . .....

Metaphase kinetochore (para-equatorial
sections) . . . . . . . .

Hletaphase chromosome (para-equatorial
section) . . . . . . .

Mletaphase cell, para-equatorial section of
kinetochore MT . . . . ... ..

Mid-anaphase kinetochore (para-equatorial


Page

184


186 188


190 190 192 192 192


194 194 194 196 198

198

198 199


201 201








Figure

104

105a-c 106a,b 107a-c 108a-f 109


110a,b illa-c 112a,b 113


114


115


116


117a,b 118


119a,b


trast micrographs . . .. .


Streptonigrin-treated cell in early
anaphase . . . . . .

Streptonigrin-treated cell in late
anaphase . . . . . . .

Streptonigrin-treated cell in late
anaphase . . . . . . .

Streptonigrin-treated cells in late
anaphase . . . . . . .

Strcptonigrin-treated cell in late
anaphase . . . . . .

Streptonigrin-treated cell in late
telophase . . . . .

Streptonigrin-treated cell in late
telophase . . . . .

Streptonigrin-treated cell in late
telophase . . . ..... .

Streptonigrin-treated cell in late
telophase . . . . ...

Streptonigrin-treated cell in late
cytokinesis . . . .

Streptonigrin-treated cell in late cytokinesis . . . . .


* * .


* * .








* 0 0 *





* 0 * 0


Diagrammatic representation of kinetochore maturation during prometaphase . .. .. *

Diagrammatic representation of the threedimensional structure of kinetochores .
xiii


Untreated interphase cell, four centrioles Untreated mid-prophase cell, abnormal
centrioles . . . . . . .

Untreated cell in late prometaphase, abnormal
centrioles . . . . . . ... .

Untreated anaphase cell, abnormal
centrioles . . . . . . . .

Streptonigrin-induced aberrations. Phase con-


Page

207 207


209 210 212 214 214 216 217 219 219 219 220 222 222 224 243 248


120 121


. . .














KEY TO ABBREVIATIONS AND SYMBOLS


C centriole(s) CG centromeric granules Ch chromosome(s) Chd chromatid(s) Chr chromatin Ci cisterna(e) Co corona of kinetochore CV intracentriolar vesicle D contamination of thin sections (dirt) DC daughter centriole(s) EOC 8-ethoxycaffeine F fibrillar component of the nucleolus G Golgi complex Gr granular component of the nucleolus H achromatic hole(s) in chromatin or chromosomes HChr heterochromatin IV intranuclear vesicle(s) K kinetochore K1, K2 sister kinetochores KG kinetochore granule KI inner layer of the kinetochore KM middle layer of the kinetochore KO outer layer of the kinetochore MA mitotic apparatus 1MB midbody
-F microfibril(s) Mi mitochondrion III mitotic index .11 micronucleus MT microtubule(s) N nucleus NE nuclear envelope NL nuclear lobe NO nucleolus organizer NP pore-annulus complexes) of the nuclear envelope Nu nucleolus NV membrane vesicle formed by fragmenting nuclear envelope P, P1' P2 spindle pole(s) Pa particle(s) associated with centrioles PC parent centriole(s) R ribosomes (poly- or monosomes) RER rough endoplasmic reticulum S satellite(s), pericentriole body (bodies) Sb stem body Sm stem SN streptonigrin V cytoplasmic vesicle(s) X X chromosome; other chromosomes are numbered arbitrarily


xiv








Abstract of Dissertation Presented to the Graduate Council of
the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

NORMAL AND ABNORMAL MITOSIS IN A MAMMALIAN CELL IN VITRO.
A LIGHT AND ELECTRON MICROSCOPIC STUDY.

By

Urs-Peter Roos

December, 1971

Chairman: Dr. Fred C. Johnson
Co-Chairman: Dr. Henry C. Aldrich Major Department: Zoology

Rat kangaroo cells (PtK2 line), grown as monolayers, were fixed

and embedded in situ. Cells in mitosis were examined and photographed under phase contrast. Serial sections were examined with the electron microscope.

Centrioles duplicate at the onset of prophase. Centriole migration, disintegration of the nucleolus, chromosome condensation$ breakdown of the nuclear envelope (NE), and formation of the mitotic spindle are similar to these processes in other mammalian cells. Prophase kinetochores (Ks) appear as globular, fibrillar bodies in the primary constriction of chromosomes. Detailed observations on prometaphase chromosomes support the pulling theory of amphiorientation and metaphase stability. Amphiorientation is established by unipolar followed by bipolar attachment to spindle microtubules (NT), or by simultaneous bipolar attachment, depending on the position of a chromosome relative to the poles at the time of attachment. In the former case, bipolar attachment is presumably followed by congression. Concomitantly, Ks mature to trilaminar, flat, undulated, concave, or convex discs, 4,0006,000 A in diameter. All the Ks reach maturity at metaphase. Inner and outer layers are 400 A, and the middle layer is 300 A thick. On the

xv








average, 26 bundled MT are anchored in the outer layer. Staining characteristics and behavior in late mitosis suggest that the inner layer is chromatin, the outer layer possibly protein.

Daughter chromosomes are connected by strands of chromatin during very early anaphase. The strands rupture during early anaphase. In late anaphase, when chromosome movement has ceased, the K layers are fuzzy, and fewer, less organized MT are attached. In telophase the inner layer is a very opaque patch of dense material on the inner membrane of the NE; the outer layer is lost in the cytoplasm.

Reconstruction of the NE, cleavage, midbody formation, and

reappearance of the nucleolus, are similar to these processes in other mammalian cells.

Mitotic stages in cells exposed to 0-40C for 1 hr are similar to

untreated cells. Prophase Ks are identical and mature metaphase Ks are typically triple-layered, though slightly fuzzy. Few IT, almost exclusively kinetochore lIT, are preserved. They are coated by an amorphous or finely fibrillar substance.

No spindle MT are present in colcemid-arrested cells (0.05 jg/ml for 2 hr, 1 hr recovery; 0.25 pg/ml for 15 min, no recovery). Chromosomes are scattered in the central area of the cells. Centrioles have duplicated, but not migrated. The higher concentration of colcemid induced decondensation of chromosomes and dispersion of chromatin in all interphase cells. Kinetochores resemble immature prometaphase Ks in control cells. The inner layer is lacking. The outer layer is a convex or undulated, bilaminar plate, 400 A thick and 5,000-9,000 A in diameter, and embedded in a finely fibrillar matrix.


xvi








Streptonigrin (SN; 0.01 pg/ml and 0.05 pg/ml for 4 hr, 48 hr recovery) depressed the mitotic index (2.22 and 0.18, respectively, versus 2.59 in controls) and induced high frequencies of chromosomal and mitotic aberrations in meta- and anaphase cells (49.0. and 100%, respectivelyp versus 14% in controls). The most frequent aberrations in control cells were single dicentric anaphase bridges with a few acentric fragments. multiple bridges and fragments were the most common aberrations in SN-treated cells. The higher concentration of SN produced very complex aberrations, including decondensation of chromosomes. Kinetochores of dicentrics are structurally normal. Micronuclei and very fine nuclear bridges connecting daughter cells in cytokinesis in SN-treated cultures are enveloped by normal NE.


xvii














REVIEW OF LITERATURE


Mitosis has been studied for nearly 100 years (Dietz 1969). Its fundamental importance as a mechanism for the orderly distribution of chromosomes has provoked numerous investigations and experiments. Accordingly, there is a vast amount of literature on both mitosis and chromosomes. I did not intend to compete with the much more knowledgeable and experienced authors of the many reviews that have appeared. Rather, I have tried to survey the literature on those aspects of eukaryotic mitosis and chromosomes which are most closely related to my own observations and results.

General Ultrastructural Features of
Vertebrate Cells in Hitosis

Today, improved methods are available that allow correlated light and electron microscopic observation of single animal cells (e.g., Brinkley etal. 1967). It is a little surprising, therefore, that no detailed study combining the two approaches has been done on animal cells in mitosis. (For a good example of mitosis in plant cells see Bajer 1968, Bajer and Molk-Bajer 1969.) Ultrastructural features of mitosis have been described for a number of animal cells (see Luykx 1970 for references), but light micrographs are equally important for comparison with earlier light microscopic studies.

In the following description the emphasis is on the nucleus.

Cytoplasmic organelles and components are considered only as far as they play a role directly related to the formation of the mitotic








apparatus (MA) and chromosome segregation. Mitosis in newt heart cultures (Barnicot and Huxley 1965), rat thymic lymphocytes (Murray et al. 1965), HeLa cells (Robbins and Gonatas 1964), L strain fibroblasts (Krishan and Buck 1965), and rat hepatoma cells (Chang and Gibley 1968) served as the basis for this summary.


Prophase

The chromatin, which is dispersed during interphase, begins to

condense during early prophase. This process continues until, at the end of this stage, individual chromosomes are clearly recognizable and can be seen to consist of two sister chromatids. Kinetochores (centromeres) may become visible during prophase in some cell types. The centrioles may have duplicated during the previous interphase, or they do so during prophase. So-called procentrioles, or daughter centrioles arise at right angles to the separated parent centrioles. At this stage the centrioles often lie within a pocket of the nuclear envelope. A dense, osmiophilic mass can be seen associated with each parent-daughter pair. As prophase progresses, the pairs migrate to opposite poles. The timing of this event relative to the breakdown of the nuclear envelope varies. In any event, microtubules (MT), which radiate from the centrioles during early prophase, begin to invade the nucleus as the envelope breaks down. Fragments of the latter appear either as double stacks of membranes (Chang and Gibley 1968, Murray et al. 1965), or as vesicles (Barnicot and Huxley 1965). The fibrillar component of the nucleolus disappears (see also Hsu et al. 1965), the granular component disperses and breaks up into smaller masses which often remain attached to chromosomes during meta- and anaphase (Chang and Gibley 1968).








Prometaphase

This stage, which is characterized by metakinesis, i.e.,

chromosome movements that ultimately result in the alignment on the metaphase plate (Mazia 1961), has been neglected by investigators of the ultrastructure of mitosis in animal cells. This may be due partly to difficulties in recognizing and selecting cells in this stage.


Metaphase

The chromosomes are condensed and aligned on the metaphase plate. The mitotic spindle, made up of chromosomal and interpolar (continuous) microtubules, is fully formed between the two poles occupied by the two pairs of centrioles. Asters, consisting of IMT radiating from the centrioles, may be more or less distinct. Larger cell organelles, such as mitochondria and endoplasmic reticulum (ER), are generally excluded from the spindle, but exceptions occur. Ribosomes are abundant within the spindle, mostly as monomers. Spindle MT converge at the poles, where they enter the osmiophilic zone around the centrioles. Bundles of MT connect to the chromosomes at the kinetochores. The latter are seen as straight or crescent-shaped, dark bands, separated from the chromosome by a lighter band. Remnants of the nuclear envelope may be present at the periphery of the spindle, often more concentrated near the poles. Nuclear pore complexes are absent from these fragments.


Anaphase

As the two sets of daughter chromosomes separate, the chromosomal (kinetochore) IT shorten, while the pole-to-pole distance increases (spindle elongation). Later, the chromosomes fuse to one large mass of dense chromatin. Concomitantly, pieces of double membrane, possibly








derived in part from the remnants of the nuclear envelope, appear at the periphery of the chromatin mass, first at the polar face. Ribosomes and nuclear pores are found on some of these pieces. It is not unusual for membranes to get caught between the coalescing chromosomes. The presence of restored nuclear envelope seems incompatible with the occurrence of MT within the chromatin mass. In the interzone, mitochondria and, in some cases, ER can be found, and dense material around the MT in the equatorial region indicates early stages of midbody formation.


Telophase

The nuclear envelope is completely reconstructed by fusion of membrane vesicles and cisternae. Nucleoli are reconstituted from material formed at the nucleolar organizer or by coalescence of small nucleolar bodies, but the process is poorly understood (see Busch and Smetana 1970). Concomitantly, the mass of chromatin disperses. The cytoplasm is divided either by a wedge-shaped constriction in the equatorial region (Robbins and Gonatas 1964), or by the formation of a vesiculated equatorial plate (Murray et al. 1965). In the latter case the vesicles fuse to form the cleavage furrow.


Cytokinesis

Cleavage progresses until the daughter cells are connected only by a cytoplasmic stem. Included in the stem are tightly bundled MT and the midbody, which has formed by fusion of the dense material in the equatorial region (see also Byers and Abramson 1968, Paweletz 1967, Schroeder 1970). The midbody is included in one of the daughter cells or lost. Decondensation of the chromatin may not be completed until late in interphase.








Centrioles


Occurrence

Centrioles occur in most or all animal cells (Brinkley and Stubblefield 1970), in some fungi (Aldrich 1967, Renaud and Swift 1964), and some algae (Ringo 1967a). They are absent in cells of higher plants, either before or during division (Pickett-Heaps 1969, Wilson 1970). The possibility that certain non-dividing, fully differentiated animal cells may also lack centrioles has been suggested by Bernhard and de larven (1960; see also de Harven 1968), but the available evidence is inconclusive.

The apparent de novo origin of centrioles during certain stages of the life cycle in many lower plants is an intriguing problem. A detailed discussion, however, is beyond the scope of this survey. The reader is referred to reviews by de Harven (1968), Luykx (1970), Mazia (1961), and Pickett-Heaps (1969).


Ultrastructure

There is good agreement in the literature that centrioles of interphase cells do not differ structurally from centrioles of the mitotic apparatus (Brinkley and Stubblefield 1970, de Harven 1968, Stubblefield and Brinkley 1967). A centriole is a cylindrical body 1,500-2,500 A in diameter and 3,000-7,000 A long. Nine triplets of fused 240 A MT form its wall. Centriolar MT possibly have 13 subunits in cross section, as do flagellar MT (Ringo 1967b). The triplets are not arranged radially, but slanted, giving the image of a pinwheel. Centrioles are polarized bodies. The so-called distal end is capable of generating the shaft of a flagellum or cilium (Gibbons and








Grimstone 1960). The proximal end, where daughter centrioles are formed during duplication, exhibits a cartwheel structure with spokes radiating from a hub. This cartwheel is recognizable in ideal cross sections only, but it can also be reenforced photographically by the Markham technique (Markham et al. 1963; for an example see de Harven 1968, Figure 2). The spokes seem to connect the hub with the innermost MT of each triplet; other connections possibly occur between the outermost and innermost MT of neighboring triplets. Centrioles are often embedded in an amorphous, osmiophilic matrix (Murray et al. 1965, Robbins and Gonatas 1964, Stubblefield and Brinkley 1967), that may undergo periodic changes in preparation for, and during, cell division (Robbins et al. 1968).

In Chinese hamster fibroblasts the occurrence of a small,

membrane-bound vesicle, termed nucleoid, approximately in the center of the centriole seems to be the rule (Brinkley and Stubblefield 1970, Stubblefield 1968, Stubblefield and Brinkley 1967). These investigators also claimed to have detected a helical filament, 60-70 A in diameter, that winds 8-10 turns just inside the triplets.

In many cells so-called satellites (pericentriole bodies) have been observed (Brinkley and Stubblefield 1970, de Harven 1968, Murray et al. 1965, Robbins et al. 1968). These are osmiophilic bodies in the vicinity of interphase centrioles, but in most cell types studied they seem to be absent or inconspicuous during mitosis (de Harven 1968). They possibly contain PA (Brinkley and Stubblefield 1970). The existence of connections between satellites and centrioles, and the possibly regular number and arrangement of satellites around centrioles are controversial (see de iHarven 1968 for a discussion). A few authors








have claimed that nine satellites form a symmetrical crown around the centrioles (Bessis et al. 1958, Gachet and Thidry 1964). Chemistry

It is obviously a difficult task to determine the elemental or molecular composition of centrioles. Cytochemical methods generally lack sufficient resolution and bulk isolation of reasonably purified centrioles seems virtually impossible. The important information concerning the chemical composition of centrioles is therefore derived from studies on basal bodies (kinetosomes) of cilia, and is based on the assumption that the close structural and developmental relationship between kinetosomes and centrioles (see de Harven 1968 for references) justifies extrapolation of chemical analyses. Seaman's early findings (1960) of 2% RNA and 3% DNA in kinetosomes of Tetrahymena did not remain uncontested (see de Harven 1968 for a detailed discussion). At best, it seems, the possibility that kinetosomes contain RNA and/or DNA cannot be ruled out, but the presence of these important macromolecules in centrioles remains hypothetical. Duplication

As early as 1952 Inou6 concluded from experiments with colchicinetreated Chaetopterus eggs that centrioles duplicate and mature in the absence of a mitotic spindle. In a now classical experiment Mazia et al. (1960) determined the time sequence and mode of duplication of mitotic centers in echinoderm embryos. Although the methods used by the investigators did not allow direct visualization of centrioles (hence the term "mitotic centers"), the results predicted what has since been confirmed by electron microscopy. The results can be summarized as follows: (1) At all times the center is at least a








duplex structure. (2) Duplication occurs in very early interphase or in late telophase of the preceding mitosis. (3) The splitting of the centers is a process distinct from duplication, although the two usually occur at about the same time during mitosis. As the two members of a pair of "old" centers split, each one gives rise to a new one with which it remains associated until the next cycle. (4) The primary duplication event involves only a part of the parent structure.

(5) Mercaptoethanol (ME) inhibits duplication, but not splitting and separation of existing centers. If ME is applied prior to duplication, two daughter cells result from the first division after release from inhibition. iultipolar divisions ensue if ME is applied after duplication.

Went (1966) did a similar experiment with sand dollar eggs.

Benzimidazone (BIA) inhibits cell division, but not duplication of mitotic centers. After release from BIA-inhiibition the cells cleave into as many blastomeres as there had been mitotic centers. Comparing these results with those from ME-treatment, Went proposed there should be a pair of potential mitotic centers at each pole of a BIA-induced tetrapolar mitotic spindle, while in the case of ME-induced tetrapolar cells there should be a single center.

The formation of so-called daughter or procentrioles during

duplication has been confirmed by electron microscopy (Bernhard and de Harven 1960, Brinkley and Stubblefield 1970, Erlandson and de Harven

1971, Murray et al. 1965, Stubblefield and Brinkley 1967). Procentrioles arise at the proximal end of the parent centriole and approximately at a right angle to the latter. They are structurally similar to the parent centrioles, but much shorter during the early stages of duplication.








Stubblefield (1968), using a technique that renders centrioles in fixed cells visible for light microscopy, studied centriole duplication and behavior in colcemid-treated Chinese hamster cells. His findings concerning duplication, maturation, and separation agree well with the above-mentioned inhibitor studies (Mazia et al. 1960, Went 1966). Further confirmation came from an ultrastructural study on colcemidtreated cells (Brinkley et al. 1967).

From the references and reviews mentioned the following picture of the centriole cycle emerges: (1) At some stage between divisions each of the two centrioles normally present in interphase cells produces a daughter centriole by an as yet unknown mechanism. (2) The daughter centrioles undergo maturation, which involves elongation and, possibly, formation of the intracentriolar structures. If we define "'mature" as being capable of generating a procentriole and to participate in spindle formation (see the following paragraph for reservations about the latter), then the timing relative to the cell cycle under normal conditions is so that procentrioles are mature no sooner than the end of karyokinesis following duplication. (3) If spindle formation is inhibited by colcemid or IE, maturation of the procentrioles continues and they may act as mitotic centers after release from the block. As a consequence, multipolar spindles occur more frequently than under normal conditions, and the proportion of these increases with increasing time of exposure to the inhibitor.


Function

Only the role played by centrioles in animal mitosis will be

discussed here. For the involvement of centrioles in the generation of








flagella and cilia see Dirksen and Crocker (1966), Renaud and Swift (1964).

Two principal functions have been assigned to centrioles in

dividing animal cells: (1) Determination of the poles of the mitotic apparatus, and (2) assembly of spindle MT (for detailed reviews see Luykx 1970, Nicklas 1971, Pickett-Heaps 1969).

The axis of the mitotic spindle is not determined by the

orientation of the centrioles at the poles, except, possibly, in special cases (de Harven 1968, Luykx 1970). A number of observations seem to indicate that centrioles are indispensable as poledeterminants. In normal bipolar mitosis there is one pair of centrioles at each pole (Krishan and Buck 1965, Murray et al. 1965, Robbins and Gonatas 1964, Robbins et al. 1968, Stubblefield 1968, Stubblefield and Brinkley 1967). Furthermore, cytasters produced in sea urchin eggs by artificial activation are centered around poles containing centrioles (Dirksen 1961). In contrast to this, the number of poles in hybrid somatic cells often does not correspond to the number of centriole pairs present (e.g., Yamanaka and Okada 1968), and in cranefly spermatocytes mitosis occurs without centrioles under certain conditions (Dietz 1959, 1966). Finally, mitosis in higher plants proceeds without centrioles and many lower animals and plants have pole-determinants other than typical centrioles (see Luykx 1970, Pickett-Heaps 1969).

The idea that centrioles might be active in the assembly of MT draws support from claims that spindle MT are directly connected to centriole walls (Brinkley and Stubblefield 1970, Gall 1961), and from the proposition that the poles may be an area where IT are assembled








and disassembled (Inout 1964, Inout and Sato 1967). It should be noted, however, that the direct connection of MT to centrioles has not been proven unequivocally (see de Harven 1968, Luykx 1970, PickettHeaps 1969). More often, MT seem to connect to pericentriolar bodies or the amorphous, osmiophilic mass surrounding the centrioles. Similar dense masses can be found at the poles of higher plant cells (PickettHeaps 1969, Wilson 1970). Most recent reviews, therefore, consider this material, or part of it, a more likely candidate for a poledeterminant and MT organizer, not only because it occurs almost universally, but it would also bridge the gap between centriolar and acentriolar mitosis (see Luykx 1970, Nicklas 1971, Pickett-Heaps 1969). In this scheme, centrioles are thought to play a much more passive role, being carried along and distributed mainly for use as basal bodies of cilia and flagella (see also FriedlAnder and Wahrman 1970),

Brinkley and Stubblefield (1970; see also Stubblefield and

Brinkley 1967) have presented a different hypothesis. They maintain that centrioles do play a role as pole-determinants and in the assembly of MT. However, in the absence of solid evidence, their detailed mechanistic and molecular model of centriole-MT interaction in mitosis remains highly speculative.


Spindle Fibers

Under favorable conditions fibrillar elements are visible in the mitotic spindle of living cells with the phase contrast microscope (Ris 1955). In fixed and stained cells spindle fibers can be seen without difficulty in the light microscope (e.g., Heneen 1970). The Nomarski interference-phase contrast system also allows visualization of the








fibrous organization in some cells (Bajer and Allen 1966). The most convincing evidence concerning the reality of spindle fibers has come from polarization microscope studies (Inoug 1964, Inoug and Sato 1967).

Early electron microscopic observations by Harris (1962) and Roth and Daniels (1962) revealed numerous MT in the mitotic spindle of sea urchin eggs and amebae, respectively. That the mitotic spindle is a collection of prominent MT has since been substantiated by numerous other workers (e.g., Aldrich 1969, Krishan and Buck 1965, Robbins and Gonatas 1964, Roth et al. 1966). The distribution of spindle MT agrees well with that of spindle fibers as seen in the light microscope (see Luykx 1970, Table III). Further support for the contention that the spindle fibers of the light microscopists consist of bundles of MT came from birefringence and electron microscopic studies on the effect of colchicine, colcemid, vinblastines cold, and ultraviolet light (UV) on spindle 1T. All these agents cause loss of spindle birefringence and a reduction of the number of IT as seen in the electron microscope (Bajer 1969, Inoug 1952, Inou6 and Sato 1967, Malawista et al. 1968, Roth 1967). On the other hand, treatment of mitotic cells with heavy water increases birefringence and the number of MT in the spindle (Inou4 and Sato 1967). The observation by Rebhun and Sander (1967), that MT are not the only birefringent component of the mitotic spindle, imposes limits on the interpretation of such results, but the fact remains that

-T are birefringent elements. The correlation between birefringence and MT is generally regarded as sufficient evidence for the occurrence of MT in living cells, particularly since improved fixatives for electron microscopy have made the demonstration of MT easy (but see Nicklas 1971 for a discussion of this point).








Following common usage in the literature, the term spindle fiber shall hereafter designate the structure seen in the light microscope. The term microtubule (MT) shall be reserved for the fibrils seen in the electron microscope. By this definition a kinetochore fiber

consists of one or several MT.


Fine Structure

Spindle MT are slender cylinders 150-250 A in diameter and several microns long (de Harven 1968, DuPraw 1968, Luykx 1970, Nicklas 1971, Roth 1964). The tubules are hollow and have a wall 40-60 A thick. The substructure of the wall is best resolved by negative staining of isolated MT (Barnicot 1966, Kiefer et al. 1966), or by freeze-etching (Moor 1967). Globular subunits, 30-40 A in diameter, form linear filaments. Ten to thirteen such filaments form the tubule wall. The filaments are arranged in such a way that the subunits form a helix
0
with a pitch.of 10-20. Spindle MT are thought to be as rigid as cytoplasmic MT (Luykx 1970), and therefore should run more or less straight over a certain distance, or only be slightly curved. Wavy MT can be a shrinkage artifact produced during dehydration (Jensen and Bajer 1969).

Kiefer et al. (1966), citing a number of studies on the extraction of proteins from mitotic apparatus, all of which seem to have yielded a protein particle sedimenting at approximately 2.5S (MW approximately 34,000), proposed that this particle corresponds to the 35 A globular subunit of the MT wall. Results from the various laboratories have, however, been interpreted in different ways and the correlation between visible subunits and isolated particles is not firmly established








(see Luykx 1970, Nicklas 1971, for detailed discussions). Inoug and his collaborators (Inoue 1964, Inou6 and Sato 1967) have proposed that MT polymerize from a pool of subunits in a dissociation-association

equilibrium, probably under control of "organizing centers" such as kinetochores and centrioles. Certainly, the rapid loss and reappearance of birefringence in mitotic cells subjected to rapid temperature shifts (Inoue et al. 1970) suggests self assembly of subunits to form MT (see Nicklas 1971).

Fine interconnections or "cross-bridges" between spindle MT, and "arms" on single MT have been reported for a number of plant and animal cells (Hepler and Jackson 1968, Krishan and Buck 1965, Wilson 1969). Hepler et al. (1970) devoted a more detailed and systematic study to these structures in cultured human cells and Haemanthus endosperm, but although they play. a major role in the model of mitosis proposed by Mclntosh et al. (1969), their reality is conjectural at best (see also Nicklas 1971).


Distribution and Classification

Spindle MT can be divided into two main categories: chromosomal and continuous (interpolar) MT (de Harven 1968, Luykx 1970, Nicklas 1971, Roth 1964). In the case of chromosomes with a well-defined kinetochore the chromosomal MT run between the latter and the corresponding spindle pole (Brinkley and Stubblefield 1966, 1970; Harris and Mazia 1962, Jokelainen 1967, Krishan and Buck 1965, Murray et al. 1965, Robbins and Conatas 1964). Chromosomes with so-called diffuse kinetochores are connected with MT at various points along their length (Buck 1967). Luykx (1970) distinguished several








subcategories of spindle MT. Notable among them are the penetrating or transchromosomal MT (Nicklas 1971) observed in many animal cells (e.g., Behnke and Forer 1966, Buck 1967, Jokelainen 1967, Robbins and Gonatas 1964). They may actually be part of the population of

continuous MT.

It is obviously difficult to demonstrate that continuous MT do run from one pole to the other, because they often pass into and out of sections. However, this has been demonstrated in two cases involving relatively short spindles (Aikawa and Beaudoin 1968, Manton et al. 1969b). On the other hand, studies on late stages of mitosis indicate most "continuous" MT overlap in, and terminate beyond, the midbody (Byers and Abramson 1968, Paweletz 1967). Understandably, this point, which is important for an explanation of the function and mechanics of the MA, is still controversial. More detailed studies, involving counts of HT in serial sections cut at a right angle to the spindle axis, have only recently been published (Brinkley and Cartwright 1970, Manton et al. 1969b, McIntosh and Landis 1971). It appears that relatively few MT [107. in the diatom Lithodesmium (Manton et al. 1969b, 1970); 30-40% of the interpolar MT which make up 40-50/. of all the spindle MT in Chinese hamster and rat kangaroo cells (Brinkley and Cartwright 1970)] are truely interpolar. Most "continuous" NIT project across and terminate beyond the equator.

It is generally agreed (see Luykx 1970) that there is no

difference regarding structure and dimensions between the different classes of spindle MT. However, kinetochore MT differ from continuous and astral IlIT in their sensitivity to spindle poisons. Upon exposure of Chaetopterus eggs to colchicine, the continuous fibers lose their








birefringence more rapidly than the chromosomal fibers (Inou4 1952). Sauaia and Mazia (1961) found that in sea urchin eggs the asters, but not the kinetochore fibers, are disorganized by brief exposure to colcemid. Likewise, a low concentration of colcemid disorganizes continuous MT in Chinese hamster cells, but some kinetochore MT remain (Brinkley et al. 1967). Similar differences apply for cold treatment. Kinetochore fibers in lily pollen mother cells in anaphase regain birefringence first when the cells are brought to ambient temperature after chilling (Inoug 1964). In contrast to this, continuous fibers in Chaetopterus eggs are the first to regain birefringence lost during chilling (Inoui 1964). In mammalian cells in vitro Brinkley and Cartwright (1970) found cold shock completely disorganized interpolar MT, while the number of chromosomal M14T was reduced by 30-40%..


Function

It has been proposed that continuous fibers, which form the socalled "central spindle" (Mazia 1961), function in the separation of the pole-determinants (centrioles where applicable. Brinkley et al. 1967, Brinkley and Stubblefield 1970, FriedlaInder and Wahrman 1970, Mazia 1961). In Chinese hamster cells exposed to colcemid, chromosomes arrange more or less radially around the two unseparated pairs of centrioles (Brinkley et al. 1967). The reformation of pole-to-pole MT after release from the inhibitor seems necessary for the separation and migration of the centrioles.

In cells with diffuse kinetochores, the penetrating or

transchromosomal MT may play an important, if not exclusive, role in chromosome movement (Luykx 1970, Nicklas 1971). Whether they also








participate in the movement of chromosomes with a well-defined kinetochore is uncertain. The often cited observations by Carlson (1938) on apparent poleward movement at anaphase of X-ray-induced acentric fragments in grasshopper neuroblasts are not convincing. Bajer (1958) and Bajer and M-ol&-Bajer (1963) presented better evidence based on cinemicrography of /3-irradiated Haemanthus endosperm cells. In the majority of cases, however, acentric fragments do not behave as normal chromosomes (Kihlman 1966, Lea 1962).

Finally, continuous MT are supposed to produce spindle elongation at anaphase in cells where this occurs (Mazia 1961, Roth et al. 1966). Spindle elongation is a process different from poleward movement of chromosomes, as demonstrated by Ris' (1949) observation that chloral hydrate prevents the former, but does not inhibit the latter.

The chromosomal fibers form the "chromosomal spindle" (IMazia

1961). Chromosomes with a distinct kinetochore are firmly attached to the kinetochore fibers. This can be inferred from the above-mentioned study by Ris (1949), and from observations by Inou (1952) that in Chaetopterus eggs treated with colchicine the chromosomes disperse from the metaphase plate only after the birefringence of the chromosomal fibers has completely disappeared. Shimamura (1940) reported that chromosomes at the centripetal pole of centrifuged lily pollen mother cells are firmly anchored by their kinetochore fibers, although the centrifugal force is strong enough to cause uncoiling of the chromosomes themselves. Finally, the most direct evidence came from elegant micromanipulation experiments by Nicklas and Staehly (1967) on grasshopper spermatocytes. Chromosomes can be stretched with a microneedle without changing the kinetochore-to-pole distance








significantly from the beginning of prometaphase to the end of anaphase. Lateral displacement within the spindle is possible, however. The authors found no evidence for interzonal connections between separating chromosomes at anaphase.

There is a large body of evidence that kinetochore fibers play a major, if not exclusive, role during metakinesis, i.e., those movements of the chromosomes that begin with prometaphase and terminate with the alignment of the chromosomes on the equator of the spindle (congression), as well as in the poleward movement of chromosomes during anaphase. Bivalents of spermatocytes in prometaphase often have the shape of an arrowhead with the kinetochore at the tip (Dietz 1969). If one bivalent is stretched to a greater extent than its partner, the movement is always in the direction of the pole towards which the former is oriented. More direct evidence comes from UV microbeam irradiation experiments. In newt fibroblasts irradiation of the kinetochore, but not of other parts of a chromosome, stops prometaphase movement and the irradiated chromosome never reaches the metaphase plate (Bloom et al. 1955, Uretz et al. 1954). In Haemanthus endosperm cells, on the other hand, similarly treated chromosomes do reach the metaphase plate, although their paths of movement during prometaphase may be altered and become very complex (Bajer and 1olt-Bajer 1961).

Chromosome motion at metaphase is very slow and minimal (Dietz

1969, Hazia 1961). The equatorial position of the chromosomes is very likely maintained by a balance of forces on opposed sister kinetochores. If one kinetochore pair of a bivalent in a grasshopper spermatocyte in metaphase is irradiated with a UV microbeam, the bivalent shifts towards the pole closest to the unirradiated








kinetochore (Izutsu 1961). An exception to this rule was reported by Forer (1966) for crane fly spermatocytes. Here, UV microbeam irradiation of the chromosomal fibers on the poleward side of the metaphase plate does not induce movement. A possible explanation for this behavior is the apparent absence of any kind of kinetochore on these chromosomes, as reported by Behnke and Forer (1966). Finally, the necessity of opposing poleward forces for stable metaphase alignment was clearly demonstrated in Nicklas' laboratory (Henderson and Koch 1970, Henderson et al. 1970, Nicklas and Koch 1969).

It is a well-supported conclusion that chromosomes move as

individuals, although they may move synchronously (Luykx 1970, Mazia 1961, Nicklas 1971). Since Ris' study (1949) two types of anaphase movement are distinguished: (1) spindle elongation, and (2) shortening of the chromosome-to-pole distance. The two processes are based on different mechanisms, because the former, but not the latter, is inhibited by chloral hydrate (Ris 1949). In grasshopper spermatocytes the two processes act together (Ris 1949), but in other cells each of the two possible extremes can occur (see Mazia 1961 for references). 1azia (1961), who also discussed the various early hypotheses concerning anaphase movement, summarized the events as follows: "The central spindle is more or less rigid; it moves the poles apart and provides an anchor for the poles which must bear the load of the chromosomes."

The question whether chromosomes are pulled or pushed has engaged the mind of many a biologist. Most reviewers (Dietz 1969, Luykx 1970, lMazia 1961, Nicklas 1971) arrived at the conclusion that a pulling force must be involved in the movement of chromosomes, although a








pushing force (similar to the "Stemmkirper" proposed by B6l1r 1929) may contribute to anaphase separation. Clearly, prometaphase movement

cannot be explained on the basis of pushing forces only. The behavior of chromosomes that are pushed away from the spindle into the cytoplasm by micromanipulation also suggests a pulling force (Nicklas 1967). The results of Forer's (1966) UV microbeam irradiation experiments on crane fly spermatocytes seem to refute the idea that chromosomes are simply pulled by their kinetochore fibers. Nevertheless, the results can be interpreted as supporting a pulling hypothesis (Nicklas 1971).

Even if we accept the pulling hypothesis there remains the problem of how the MT accomplish this, a problem around which center all of the more recent models of mitosis (Dietz 1969, Luykx 1970, McIntosh et al. 1969). There is no change in diameter as MT shorten or lengthen (see Luykx 1970, Nicklas 1971, for references), and a simple contraction mechanism is not compatible with structural observations. Inoue and his collaborators (Inoue 1960, 1964, Inoue and Sato 1967) have consistently explained this phenomenon with their "dynamic equilibrium model," which proposes a pool of MT subunits in the spindle region. Free subunits are in a dynamic equilibrium with subunits bound in MT. Shifts in the equilibrium induce further polymerization or depolimerization. The former would produce lengthening, the latter shortening of MT. Orienting centers are thought to determine the direction of the "growing" MT during polymerization.


Chromosomes

Three basic questions must be considered here: (1) What is the

unit fiber of the mitotic (and interphase) chromosome? (2) What is the








relationship between the DNA double helix and the unit fiber? (3) How is the unit fiber arranged in the highly condensed metaphase chromosome? Data and observations bearing on these questions come from a number of fields of study and so far it has not been possible to reconcile them in one comprehensive model of the eukaryotic chromosome.

In thin sections of interphase nuclei and mitotic chromosomes fibrils approximately 100 A in diameter are visible (Wolfe 1969). Because the fibers are cut at various angles, little can be concluded regarding their three-dimensional organization. Whole-mount preparations of isolated chromosomes seemed initially much more promising (for references see Wolfe 1969). In isolated chromosomes the diameter of the fibers varies from approximately 20 A to 500 A or more, depending on the quality of the preparation, but most investigators agree that the mean diameter is approximately 250 A (DuPraw 1968, Wolfe 1969). However, Ris has demonstrated (1961, 1967; Ris and Kubai 1970) that 100 A fibers can consistently be obtained. Understandably, a lively controversy revolves around these differing results and their interpretation. DuPraw (1968) considers the 250 A fiber as the unit fiber (type B fibril). On the other hand, Wolfe (1969; see also Zirkin and Wolfe 1970) considers this to be an artifact produced by the deposition on a 100 A fiber of contaminating material during preparation. A third view is held by Ris (Ris 1967; Ris and Kubal 1970): after certain chemical treatments the 250 A fibers can be shown to consist of two 100 A fibrils, more or less twisted around each other. Both Ris (Ris and Kubai 1970) and Wolfe (1969) present arguments and evidence supporting their respective hypothesis. At the center of the dispute are the many possible artifacts produced by








differences in the method of preparation. A discussion of these is beyond the scope of this summary. Perhaps Ris and Kubai (1970) are closer to the truth in their conclusion that the 250 A fiber is the structure present in the intact nucleus, but that its relationship to the 100 A fiber is not yet clear.

The second question can be divided into two parts: (a) Does the unit fiber contain a single DNA double helix or several in lateral association? The DNA-histone complex is expected to have a diameter of approximately 30 A (Zubay and Doty 1959). Many investigators have claimed to have observed a fibrillar substructure within the 100 A fiber (see Wolfe 1969, for references). However, enzyme digestion and other treatments that are assumed to remove all or most of the histones from the chromosomal fiber leave a strand 20 A thick, which most probably represents a single DNA double helix (DuPraw 1968, Ris and Kubai 1970, Wolfe 1969). It is thus generally accepted that the unit fiber contains a single DNA double helix. However, there is ample evidence that this helix is highly compacted in the unit fiber (DuPraw 1968, Ris and Kubai 1970). The major role in folding of the DNA has been assigned to the histones, but the molecular basis is not yet clear (see Ris and Kubai 1970, for discussion).

(b) Is the DNA double helix in the unit fiber continuous or does it consist of subunits of variable length, perhaps connected by socalled linkers? Sedimentation coefficients and direct electron microscopic measurements of purified DNA from isolated chromosomes or chromatin yield fragments of variable length (see Ris and Kubai 1970). Similar results are obtained with autoradiography of DNA labeled during replication. Both methods, however, have their limitations and do not








unequivocally support the idea that the DNA is discontinuous. Similarly, the observation of multiple replication points in eukaryotic nuclei is no definite evidence (see DuPraw 1968), and the accepted conclusion is that the DNA molecule in the unit fiber is continuous (DuPraw 1968, Ris and Kubai 1970).

The third question, concerning the organization of the unit fiber in metaphase chromosomes, is at the center of the uninemy-polynemy controversy. The answer is very simple: the problem is not resolved (DuPraw 1968, Ris and Kubai 1970, Wolfe 1969). Bajer (1965) observed the half-chromatid structure of chromosomes in living Haemanthus endosperm cells. When fixed metaphase chromosomes from Vicia cells pretreated with 5-amino uracil are digested with trypsin, each chromatid appears to consist of two subunits (Wolfe and Martin 1968). In whole-mounted, isolated chromosomes, each chromatid contains many fibers (e.g., Abuelo and Hoore 1969, DuPraw 1968, Lampert 1969, Stubblefield and Wray 1971, Wolfe 1965), which can be interpreted to form longitudinal subunits of chromatids, although it is virtually impossible to follow individual strands over a greater distance. Cells exposed to X-ray or certain chromosome-breaking chemicals during G2 exhibit, at the following metaphase, bridges between sister chromatids that involve apparent subunits of chromatids (subchromatid aberrations or "side-arm bridges"; Brinkley and Humphrey 1969, Heddle 1969, Kihlman 1966). Kihlman, who formerly supported polynemy based on his experiments with chromosome-breaking chemicals (e.g., Kihlman 1966), has recently questioned the occurrence of true subchromatid aberrations (Kihlman 1970), and favors the single-stranded "folded fiber" model of DuPraw (1965, 1968).







Isolabeling of chromatids (e.g., Peacock 1963) has always been a strong argument in favor of polynemy,-but it can be accommodated in a single-stranded chromosome model (Comings 1971, Ris and Kubal 1970). On the other hand, the observation that chromosomes replicate semiconservatively (Taylor et al. 1957) is in favor of uninemy. As Ris and Kubai (1970) pointed out, the most compelling evidence against multistrandedness comes from studies on the uniqueness or redundancy of DNA sequences (e.g., Britten and Kohne 1968). In Drosophila and the mouse the majority of DNA sequences are unique. This would exclude the presence of two or more identical DNA strands per chromatid, which is, of course, a prerequisite for multi-strandedness dictated by the orderly segregation of genes at mitosis.

In view of these contradictory observations and results it seems premature to propose a detailed model of the architecture of metaphase chromosomes. For examples of such models the reader is referred to the review by DuPraw (1968) and the paper by Stubblefield and Wray (1970).


Kinetochore Structure and Function Light Microscopic Observations

Chromosomal granules, presumably corresponding to the centromere studied by many later cytologists, were first described by Metzner (1894), who called them "Leitkirperchen." By 1930 a number of different terms were in use (see Schrader 1936). Sharp (1934) introduced the term "kinetochore," which has been used as a synonym of "centromere" by many authors. For clarity I prefer to apply the term centromere to the structure seen with the light microscope in the primary constriction of chromosomes, and to reserve the term








kinetochore for the structure as it appears in the electron microscope. For this review, however, I use the terms in accordance with the authors cited.

For many years most cytologists described the centromere as a

"gap" or non-staining constriction (for references see Schrader 1953). Schrader (1936, 1939) presented a detailed description of the kinetochores of meiotic chromosomes in two species of amphibia and in Tradescantia. He interpreted each tetrad to have two kinetochores, each of which consisted of two spherules lying in a commissural cup. Using various stains to enhance visibility of the spherules, he was able to follow the changes in appearance of the kinetochores during the different stages of meiosis.

Iuch of the early discussion about centromeres concerned the

problem of the presence or absence of DNA in this chromosome region. Applying the Feulgen test to pachytene plant chromosomes, Lima-de-Faria (1950) demonstrated the presence of DNA in the centromere, which he described as consisting of fibrils and chromomeres. Gall (1954) documented that kinetochores of newt lampbrush chromosomes are also Feulgen-positive. He noticed that the kinetochores resemble chromomeres in general appearance, except for the lack of lateral loops. Lima-de-Faria (1956, 1958) furnished additional evidence for his view of kinetochores of plant chromosomes as a specialized region showing Feulgen-positive granules connected by fibrils. Electron Microscopic Observations

The first good electron micrographs of animal kinetochores were

published by Harris (1962; see also Harris 1965, Harris and Mazia 1962)








and by Nebel and Coulon (1962). Satisfactory fixation was still a problem in those days, but the kinetochores in dividing sea urchin eggs could be clearly recognized as irregularly shaped, electron-dense granules at the surface of the chromosomes where they are apparently attached to NT (Harris 1962). lMost remarkable was a difference in staining intensity between the kinetochores and the remainder of the chromosomes in preparations where the fixative had dispersed or partially extracted the chromosomal fibers.

Nebel and Coulon (1962) interpreted the kinetochores of metaphase I pigeon spermatocytes as having the shape of an acorn with MT attached to the convex side of the cup. The details presented in their model (their Figure 12), however, are not all discernible in the only lowpower electron micrograph included (their Figure 11). At best one can distinguish a dense band, 2,000-4,000 A long, following the outline of the chromosome and separated from the latter by a clear zone of approximately the same width. The chromosome proper appears denser at this site; poleward of the dark band a less dense matrix can be seen. The authors interpreted the MT to penetrate the kinetochore and terminate in the chromosome.

Since these early studies a number of papers have appeared

describing the fine structure of kinetochores. The most detailed study of the kinetochores of invertebrate cells was that by Luykx (1965as b) on Urechis eggs. Despite apparent fixation problems the dense-lightdense banding was evident, particularly on melotic chromosomes. The author noted that the density of the kinetochore appeared to increase from prometaphase to anaphase. The dense material was seen to closely follow the curvature of the chromosome surface. The deep layer was








often less dense than the superficial layer and more variable in appearance. Approximately 10-25 MT, some of which seemed to end in the deep layer, could be counted per kinetochore. On mitotic chromosomes the triple-layered structure was less frequent. DNAse treatment indicated little or no DNA in the kinetochore region, or increased resistance of kinetochore DNA to the enzyme (Luykx 1965a).

Wettstein and Sotelo (1965) found that the kinetochores in Gryllus spermatocytes consist of essentially the same 100 A fibrils as the remainder of the chromosome, but the fibrils seemed more densely packed in the kinetochore. The shape of these kinetochores resembles that of a thick nail or screw deeply anchored in the body of the chromosome.

Grasshopper spermatocytes have ovoid kinetochores embedded in a cup (Brinkley and Nicklas 1968, Nicklas 1971, and personal communication; see also Brinkley and Stubblefield 1970, Figures 26 and 27). An electron-dense "axial core" is embedded in a mass of less dense 50-80 A fibrils. In general appearance these kinetochores seem to resemble plant kinetochores (Bajer and Mola-Bajer 1969, Wilson 1968) more than animal kinetochores.

A different kind of kinetochore in invertebrates is the diffuse kinetochore described by Buck (1967) in the bug Rhodnius. He interpreted the finely granular material on the surface of metaphase chromosomes as representing the kinetochore, but noticed the relative paucity of spindle MT attached to this structure. Ris (Ris and Kubal 1970) reported on a preliminary basis that no such material could be found on chromosomes in spermatocytes of the homopteran insect Philaenus. Rather, the MT seemed to penetrate deeply into the chromosome where they end blindly.








Mammalian kinetochores are by far the most intensely studied.

Besides papers devoted specifically to mitosis or kinetochores, there are numerous isolated electron micrographs in the literature showing various profiles of mammalian kinetochores (e.g., Flaks 1971, Hu 1971). Unfortunately, some of the most frequently cited papers on the ultrastructure of mammalian cells in mitosis (e.g., Krishan and Buck 1965, Murray et al. 1965) do not show kinetochores at high magnifications. At best, one can make out that they consist of three layers or bands, like those described by Nebel and Coulon (1962). Robbins and Gonatas (1964, Figure 19) presented a detailed picture of an early anaphase kinetochore in a HeLa cell. The three layers are very clear; MT seem to insert into the outer layer. Barnicot and Huxley (1965) also published electron micrographs showing the kinetochores of cultured newt heart cells as three-layered structures. They interpreted the kinetochores to consist of material different from the chromosomes, based on different stainability. In this context it is interesting to note that the kinetochores of the alga Oedogonium (Pickett-Heaps and Fowke 1969) and the moss hnium (Lambert 1970) are also triple-layered, as are those of Barbulanympha (Hollande and Valentin 1968), which has an intranuclear spindle.

Jokelainen (1965a) described kinetochores in fetal rat kidney as short bands separated from the chromosomes by a clear area, or as two parallel bands, in which case the second band was in direct contact with the chromosomal material. In subsequent papers (Jokelainen 1965b, 1967, 1968) he developed his concept of kinetochore maturation and his kinetochore model. Maturation occurs during prometaphase and is asynchronous for sister kinetochores. It involves the appearance of








the outer layer and a reduction in size of the kinetochoral patch as it becomes anchored to the spindle MT (Jokelainen 1965b). The outer layer is embedded in a moderately dense substance, part of which persists at metaphase as the middle layer of the kinetochore and the so-called corona (Jokelainen 1967). Jokelainen's model (1967) depicts the kinetochore as a trilaminar disk, 2,000-2,500 A in diameter, at the surface of the chromosome, sometimes slightly recessed, sometimes projecting. The outer layer is 300-450 A thick, the middle layer 150-300 A, and the inner layer 150-250 A. The corona over the outer layer measures approximately 300 A in thickness. Evidence for the disk-like appearance came from para-equatorial sections showing kinetochores in face view. The outer disk, which is finely granular or fibrillar, stains consistently, while the density of the chromosomes varies with the staining method employed. The inner layer is highly electron-dense, and roughly granular or fibrillar. Four to seven MT are attached to each kinetochore, apparently penetrating the outer and middle layers and sometimes ending in the chromosome.

Brinkley and Stubblefield (1966, 1970) presented a different model of mammalian kinetochores. In colcemid-arrested Chinese hamster cells the kinetochore appears as a 200-300 A wide band, made up of two 50-80 A threads. This band is embedded in a less dense matrix and follows the curvature of the chromosome surface at a distance of approximately 100 A. In the less dense matrix the authors detected 50-80 A fibrils apparently looping out from the main band. This description was based on kinetochores without attached HT, but according to the authors the kinetochores of untreated cells are very similar. The model of Brinkley and Stubblefield (1966, 1970) describes








the kinetochore as consisting of two lampbrush-like structures, each made up of two closely associated 50-80 A axial filaments from which numerous 50-80 A fibrils loop out laterally. The axial filaments extend along the surface of the chromosome, their ends being inserted into the latter. Hicrotubules attach to the axial filaments in sheets or bundles. According to their most recent report (Brinkley and Stubblefield 1970) there is little change in the structure of the kinetochores in Chinese hamster and rat kangaroo cells from prophase to metaphase. The kinetochores are mature in late prophase or in prometaphase, regardless of whether 1T are attached or not.

An electron micrograph recently published by McIntosh and Landis (1971, Figure 4) supports Jokelainen's model (1967). In this paraequatorial section of the metaphase plate three kinetochores are shown as circular patches of less dense material. On the other hand, the kinetochores of colcemid-treated cells, on which Brinkley and Stubblefield (1966, 1970) mainly based their model, may be atypical. For example, very clear images of a double-banded outer layer embedded in a less dense matrix are produced by the alkaloids vinblastine and vincristine, both of which disorganize MT (George et al. 1965, Journey et al. 1968, Journey and Whaley 1970, Krishan 1968).

Microtubules and kinetochores as seen in thin sections are not preserved in whole-mount preparations of metaphase chromosomes (e.g., DuPraw 1968). Instead, chromosomal fibers can be seen to cross between sister chromatids in the centromere region (see also Abuelo and Moore 1969). Interesting, but unexplained, is the presence of four dense granules at the centromere of Chinese hamster cells subjected to certain treatments during isolation (Stubblefield and Wray 1971).








Function

In addition to the role of the kinetochore in chromosome movement as discussed in a previous section, Luykx (1970) reviewed five other possible functions: (1) Initiation of synapsis and localization of chiasmata; (2) terminalization of chiasmata; (3) chromosome condensation or coiling; (4) association of sister chromatids; and (5) formation or assembly of chromosomal spindle fibers. The latter is the most interesting in the context of this review.

When chromosomal fibers are irradiated between the centromere and the pole in anaphase cells, birefringence disappears from the irradiated region and distal to it. Restoration of birefringence takes place within a few minutes. When, however, the centromere itself is irradiated, birefringence disappears from the whole length of the chromosomal fibers, including the distal non-irradiated portion, and restoration does not occur for a long time (Inou6 1964). Electron micrographs have shown that a great proportion of spindle IT is arranged in bundles associated with the kinetochores (Brinkley and Landis 1970, Brinkley and Stubblefield 1970, Jokelainen 1967). These observations have led to the idea that the kinetochore may play an active role in the assembly and/or orientation of MT (e.g., the "organizing center" of Inoue 1964). Further support for this idea has been drawn from observations of apparent microtubular connections between meiotic sister kinetochores (Luykx 1965b) and mitotic nonsister kinetochores (Bajer 1970). As long as clear evidence from serial sections is lacking, however, such configurations have to be regarded with skepticism. On the other hand, the occurrence of individual chromosomal spindles, either as an anomaly (e.g., Dietz








1969, Koopmans 1958) or as part of normal spindle development (HughesSchrader 1948), seems to indicate that chromosomes alone are capable of organizing spindle NT. Similarly, Roth (1967) suggested that in spindle reformation after cold shock the spindle MT gradually extend from the chromosomes towards the poles. These observations have been interpreted to mean that spindle fiber material is being continuously assembled and oriented at the kinetochore throughout prometaphase and metaphase, but the exact role of the kinetochore is not as yet clear (see Luykx 1970).

Equally interesting in this context are ideas of the kinetochore as a specialized "gene" (Brinkley and Stubblefield 1970) or a gene product (Luykx 1970). Even if we accept the universal occurrence of DFA in the centromere region as a fact, we must bear in mind that the relationship between DNA and the kinetochore at the fine structural level is not at all resolved. Brinkley and Stubblefield (1970) have proposed that the lateral loops of their lampbrush-like kinetochores consist of DNA, which codes for long RNA molecules that bind with protein subunits to form MT. This elaborate hypothesis stands on rather shaky ground. Luykx (1970),on the other hand, proposed the DNA in the kinetochore region "may therefore be viewed as a 'kinetochore organizer', similar to the nucleolar organizing region of the chromosome in a number of ways. It is responsible for the synthesis or assembly of an essential organelle that remains associated with a specific chromosomal site, is often associated with blocks of heterochromatin, remains relatively uncoiled during mitosis, and probably contains a large number of identical genes" (Luykx 1970).








Chromosomal and Mitotic Aberrations

Chromosomal aberrations occur spontaneously in certain organisms (e.g., Brandham 1970, Vig 1970). Far more frequent, however, are induced aberrations. Shaw (1970) presents a long list of agents (clastogens) that can cause chromosome damage. The agents range from nucleic acid analogs, antibiotics, drugs, and pesticides, to ionizing and UV radiation, temperature shock, and weightlessness. Certain viruses are well-known biological clastogens (e.g., Nichols 1970, Nichols et al. 1964).

Only some of the elementary aspects of the induction of chromosomal and mitotic aberrations can be reviewed here. The emphasis will be on chemical clastogens. Radiation-induced aberrations will be referred to only to the extent that they elucidated basic facts that also apply to chemicals (for detailed reviews on radiation-induced aberrations see Bacq and Alexander 1955, Evans 1962, Hollaender 1954, Lea 1962, Wolff 1963).


Cellular Events

Kihlman (1966) distinguished three types of effect of chemicals on dividing cells and chromosomes: (1) Prevention of cells from entering mitosis [e.g., 5-fluorodeoxyuridine, FUdR (Taylor 1963)]; (2) interference with active stages of division [e.g., spindle poisons, such as colchicine (Eigsti and Dustin 1955)]; (3) production of chromosomal aberrations [e.g., streptonigrin (Cohen et al. 1963)]. It is characteristic for many chemicals that they affect mitosis and chromosomes (see also reviews by Biesele 1962, Deysson 1968, Gelfant 1963). For example, FUdR inhibits mitosis and also fragments








chromosomes (Hsu et al. 1964). Other chemicals are primarily chromosome-breaking agents, but at the same time they affect the mitotic rate. To discriminate between chromosomal and mitotic effects, Deysson (1968) classified the cytological effects of antimitotic substances as follows: mitodepressive (lowering the mitotic rate), mitostatic (no proliferation), mitoclasic (disturbances of the mitotic apparatus), and chromatoclasic (induction of aberrations). Generally, the lowest effective concentration of a given chemical is mitodepressive. The same, or a slightly higher concentration produces chromosomal aberrations. Higher concentrations, besides inducing chromosomal anomalies, cause preprophase inhibition (mitostatic effect), and still higher concentrations destroy cells in mitosis.


Chemicals Versus Ionizing Radiation

Aberrations produced by ionizing radiation applied to cells in G are of the chromosome type (e.g., Heddle 1969). The transition from chromosome to chromatid aberrations occurs at the end of GI (Evans and Savage 1963), but some authors maintain that both chromosome and chromatid aberrations can result from irradiation during S, depending on whether unreplicated or replicated parts of the chromosomes are hit (e.g., Casarett 1968). Chromatid aberrations only result from irradiation of cells in G2, i.e., after completion of EVA synthesis. Subchromatid aberrations can be induced during prophase and, possibly, at the end of G2 (see Heddle 1969). The term non-delayed effect used by Kihlman (1966) with reference to chemical clastogens and radiation implies aberrations produced after completion of DNA synthesis (chromatid and subchromatid aberrations), while delayed effects include chromatid and chromosome aberrations (induced during G1 and S).








The often used term "radiomimetic" chemicals for substances that induce chromosome damage is rather misleading. As already mentioned, ionizing radiation produces both non-delayed and delayed effects, while most chemicals (e.g., alkylating agents) produce delayed effects only (Kihlman 1966). Exceptions are streptonigrin (SN) and 8-ethoxycaffeine (EOC), which produce effects very similar to X-rays. Increased oxygen tension, which drastically increases the frequency of aberrations induced by X-rays (Bacq and Alexander 1955, Evans 1962, Lea 1962) has little or no effect with many chemical clastogens (Kihlman 1966). Perhaps the most significant difference between X-rays and chemicals is that the former produce aberrations more or less randomly in a particular chromosome, while aberrations induced by the latter tend to be localized in the heterochromatin (Nichols et al. 1964, Revell 1963). Hypotheses on the Formation of Chromosomal Aberrations

The general or breakage-first hypothesis as described by Kihlman (1966) proposes that the primary event produced by a clastogen is a chromatid or chromosome break in a continuous interphase chromosome. The ends at the point of breakage may rejoin to restore the original configuration (restitution), they may remain open, or they may rejoin with other open ends. Illegitimate fusion of ends from different breaks results in sister-union or various types of exchanges.

According to the exchange hypothesis proposed by Revell (1955; see also Revell 1963) the primary event is not a break, but some other kind of lesion. The lesion may revert to normal or to another state incapable of forming an exchange. If two primary events occur close enough in space and time, an exchange initiation stage may follow.









During subsequent stages of chromosome development, the aberration is transformed into a real chromatid exchange. Revell (1955) presented data that are in good agreement with his hypothesis and Kihlman (1966) came to the same conclusion based on more recent work. However, in experiments especially designed to test the two hypotheses, Heddle and Bodycote (1970) found that neither, as usually interpreted, is entirely correct. Rather, they concluded that deletions are of two types, according to mode of origin, but they were unable to identify the two types morphologically.


Morphology and Hitotic Behavior of Aberrant Chromosomes

The following discussion is restricted to chromosomes with localized kinetochores.


'"Gaps"

"Gaps," or "achromatic lesions" are Feulgen-negative regions of variable size in chromatids (Evans 1963, Scheid and Traut 1970). In Vicia they often resemble the normal nucleolar constriction. Gaps are aberrations of the non-delayed type, since they can be induced by X-rays in prophase nuclei (Evans 1963). The question of chromatid continuity across the gap is important here, because gaps would have to be scored as true breaks if the chromatid were really interrupted. Many investigators (see Evans 1963 for references) have observed some sort of material crossing gaps. Furthermore, the chromosome segment distal to the gap moves normally and seems attached to the main body of the chromosome at anaphase. If gaps were true breaks, they should give rise to chromatid or chromosome aberrations at the second division after exposure, but this is not the case (Evans 1963). Another








interesting point is that in Vicia and Trillium Feulgen-negative regions similar in appearance to gaps can be induced by exposure to low temperature. Electron microscopic examination of whole-mounted and thin-sectioned chromosomes indicates that some gaps are true breaks, while others are traversed by chromosomal fibrils (Brinkley and Shaw 1970). Similar results were obtained by Scheid and Traut (1971) with the scanning electron microscope. They found that gaps represent distinct "notches" which in some cases are traversed by two parallel strands.


Acentric fragments

A break without reunion produces an acentric fragment (Evans 1962, Kaufmann 1954, Kihlman 1966). If the break involves only one chromatid

the fragment is single; it consists of two "sister chromatids" if the entire chromosome is broken. Such fragments are usually lost during division, or they are included at random in daughter cells where they form micronuclei (Humphrey and Brinkley 1969, Kihlman 1966, La Cour 1953). Carlson (1938) attributed apparent migration of acentric fragments towards the poles at anaphase to more than chance movement, but he made his observations on smear preparations. Bajer (1958) and Bajer and Mole-Bajer (1963) recorded the behavior of fragments in irradiated Haemanthus endosperm by time-lapse cinemicrography. The majority of fragments were eliminated from the spindle, either at prometaphase or during ana- and telophase. A few fragments, however, moved in the spindle region in a more than random fashion, sometimes from the equator to one pole and back. The authors attributed these movements to the activity of neocentric fibers. In rat kangaroo cells








exposed to X-rays, Humphrey and Brinkley (1969) confirmed by electron microscopic analysis of thin sections that fragments lack kinetochores. Bridges

An exchange between centric portions of two broken chromatids or chromosomes results in the formation of an anaphase bridge if the two centromeres move to opposite poles (Evans 1962, Hair 1953, Kaufmann 1954, Kihlman 1966, Koller 1953). Subchromatid bridges (side-arm bridges) arise from intrachromosomal exchanges. The unit of breakage and exchange in these aberrations was generally assumed to be a halfchromatid (Heddle 1969, Kihlman 1966). Brinkley and Humphrey (1969) examined X-ray-induced side-arm bridges in rat kangaroo cells with the electron microscope and found that the diameter of these chromatid connections was considerably less than that of a half-chromatid. The authors conceded, however, that chromosome movement during anaphase might have stretched the connections, which appeared to consist of chromosomal fibers of the usual dimensions (see also Brinkley and Shaw 1970).

Anaphase bridges exist as direct connections between two

centromeres, or as interlocked dicentrics (Koller 1953). The thickness and length of these bridges vary; they usually break during ana- or telophase, but may persist into interphase (Hair 1953, Koller 1953). Cinemicrographic studies by Bajer (1963, 1964) on the behavior of dicentrics in Haemanthus endosperm revealed interesting facts: normally, dicentrics break abruptly, but in slightly unhealthy cells they do not break, but form long, thin, sticky bridges. Interlocked dicentrics show two kinds of behavior: they cut one through the other,








and the broken ends rejoin; or they uncoil and do not break at all. Whether or not a bridge breaks or is stretched depends on its length and the position of the kinetochores relative to the equator at metaphase. Breakage at anaphase is due to the pulling force of the chromosomal fibers; at telophase it is caused by the phragmoplast. Sister chromatid bridges tend to break at apparently weak points which are sometimes seen as constrictions. Movement after breakage is scarcely faster than the initial anaphase speed, indicating there is no accumulated tension in the pulling mechanism. Humphrey and Brinkley (1969; see also Brinkley and Shaw 1970) observed apparent gaps within anaphase bridges in rat kangaroo cells. These gaps were constrictions, presumably caused by anaphase tension, within which three classes of chromosomal fibrils could be seen.


Chromosome "stickiness"

When cells are irradiated in late prophase, "sticky" bridges can be observed at the following anaphase (Carlson 1954). Chemical clastogens, e.g., nitrogen mustard, also produce this effect (Koller 1953). Stickiness has been interpreted to be the consequence of surface changes on the chromosomes, changes that make chromosomes adhere to each other if they happen to come in contact (Carlson 1954, Casarett 1968). Apparently, this aberration is reversible: if mitosis is delayed after treatment, stickiness does not occur. The nature of the sticky material remains obscure. Hsu et al. (1965) noted that in Chinese hamster cells nucleolar material sometimes remains attached to the ends of sister chromatids, forming apparent chromatin bridges at anaphase.








Streptonigrin: A Chemical Clastogen

Streptonigrin (SN) is a metabolite of Streptomyces flocculus. It has the formula C24H220 N 4, and the following structure (Rao et al. 1963):

CH3 0

N
COOH
H2N N

H 2 CH


HO





OCH 3

The drug behaves as a weak acid with quinoid properties. It

induces phage release in lysogenic bacteria (Levine and Borthwick 1963) and initiates rapid breakdown of E. coli DNA in vivo (Radding 1963). Inhibition by SN of DNA synthesis and of DNA-dependent RNA synthesis was reported by Koschel et al. (1966) for a cell-free systemp and by Young and Hodas (1965) for tissue culture cells. Streptonigrin caused single strand breaks in calf thymus DNA (Mizuno and Gilboe 1970). The latter authors also found that SN preferentially binds to DNA during the S phase.

Cohen et al. (1963) investigated the effect of SN on cultured

human leukocytes. The mitodepressive effect of SN appeared related to concentration and length of exposure. The mitotic rate was significantly depressed in cells exposed for 36 hr to 0.01 and 0.1 pg/ml SN, but not in cultures treated with 0.001 ,ug/ml SN. The chromosome-








breaking effects of SN were of the delayed and non-delayed type. Aberrations occurred in cells treated as late as 2 hr before fixation. Among the aberrations observed were chromatid and isochromatid breaks, acentric fragments, dicentric chromosomes, cleavage or severe attenuation of the centromere region, telomeric fusion of sister chromatids, "stickiness," uncoiling of chromosomes, and severe fragmentation or degeneration of the entire chromatin material. Chromosomal damage appeared to be non-randomly distributed, chromosomes 19, 20, 21, 22, and the Y being relatively stable. Cohen (1963) further studied the non-randomness of aberrations produced by SN in chromosomes 1, 2, and

3 of cultured human leukocytes. While X-rays induced random breaks, SN preferentially affected the pericentric regions of chromosomes 1 and 2, as well as the area of the secondary constriction of chromosome 1. Breaks in chromosome 3, although fewer in number, were distributed at random. The telomere regions of all three chromosomes, and the short arm of chromosome 2 appeared relatively resistant to SN.

Kihlman (1966) concluded from the work by Cohen et al. (1963) that SN is able to break chromosomes during G 2. A similar finding had been reported for root tip cells of Vicia faba (Kihlman 1964). Exposure of cells to 2-5 pg/ml SN for 1 hr produced subchromatid (cells in early prophase), chromatid (cells in G2 and S), and, possibly, chromosome exchanges (cells in G ). However, Puck (1964) reported that SN does not affect mammalian cells after completion of DNA synthesis. In this respect SN may be similar to 8-ethoxycaffeine (EOC) and other drugs which have non-delayed effects in plant cells, and delayed effects, or hardly any effects at all, in mammalian tissue culture cells (see Kihlman 1966).








Jagiello (1967) described chromosomal aberrations induced by SN in mouse eggs. In vitro, metaphase I chromosomes of ova treated with

1.0 pg/ml SN were agglutinated beyond recognition; 0.1 pg/ml SN induced achromatic gaps and breaks in approximately 40% of the ova. Severe chromosome damage also occurred in ova of mice, to which SN had been administered subcutaneously.

Streptonigrin is still used in in vitro and clinical studies on

tumor chemotherapy (Carter et al. 1968, Oleson et al. 1961). Like many other antitumor drugs it is more active against lymphoma than against solid tumors.


Colcemid: A Spindle Poison

Colchicine is an alkaloid isolated from Colchicum autumnale, whose anti-mitotic effect has been under investigation for over 30 years (see reviews by Biesele 1958, Deysson 1968, Dustin 1963, Eigsti and Dustin 1955, Gelfant 1963, Kihlman 1966). The structural formulas of colchicine (from Kihlman 1966) and its synthetic analog colcemid (demecolcine, N-deacetyl-N-methylcolchicine; from Schar et al. 1954) are given below.


CH 30 NHCOCH CH 30 NHCH



0CH300

3H30 CH30
0
OCH3 C3


Colchicine


Colcemid








The effect of colcemid on plant and animal cells is essentially the same as that of colchicine (see Gelfant 1963), but the former is less toxic and more efficient in animal tissue (Sch~r et al. 1954).

Levan (1938) coined the term "Ic-mitosis" for the peculiar

morphological changes in dividing Allium cells under the influence of colchicine. Chromosomes in c-mitosis are scattered in the cytoplasm, sister chromatids being held together in the centromere region. At "c-anaphase," sister chromatids fall apart and at "c-telophase," because of the absence of a mitotic spindle, all the chromosomes are included in one polyploid restitution nucleus (see also the cinemicrographic analysis of c-mitosis in Haemanthus endosperm by Mola-Bajer 1958). In animal cells the division of the centromere region is delayed until "Ic-telophase," i.e., "c-anaphase" is omitted (Levan 1954).

Besides Levan's (1938) classical scattered metaphase, other chromosomal arrangements, such as "star-mitosis" and "clumped" or "ball" metaphase, can be observed in both animal and plant cells (e.g., Gaulden and Carlson 1951, Deysson 1968). The effect depends on the concentration and the time of exposure.

For many years it was commonly accepted that colchicine and its

analogs arrest dividing cells at metaphase. Sentein (1961) objected to calling colchicine purely a metaphase poison, because he also found arrested prophases, anaphases, and telophases in his material. Brinkley et al. (1967) demonstrated that in Chinese hamster cells arrested in mitosis by treatment with 0.06 pg/ml colcemid the two pairs of centrioles are surrounded by the chromosomes in a configuration different from a typical metaphase. After reversal of the inhibition,








centriole pairs move to opposite poles and a normal metaphase plate is established prior to anaphase segregation. Furthermore, at the concentration used some MT were still present, notably on the kinetochores facing the centrioles (see also Brinkley and Stubblefield 1966). Jokelainen (1968) furnished more evidence that the precise mitotic stage at which cells are arrested may depend on the concentration applied. He found no typical metaphases in fetal rat kidney tissue after treatment of pregnant rats with 0.12 mg/kg colchicine, but metaphases were present in fetuses of rats treated with 0.08 mg/kg.

The conclusion, drawn from birefringence studies (Inoue 1952) and the behavior of chromosomes in colchicine-treated cells (Levan 1938, Mold-Bajer 1958), that this chemical disorganizes spindle MT, has received support from biochemical studies by Borisy and Taylor (1967a, b). They found that colchicine binds to a 6S protein from isolated mitotic apparatus. This protein showed good correlation with the occurrence of MT and was considered to be a subunit of MT.


Statement of Purpose

The objective of this investigation was threefold: (1) To fill

gaps in our knowledge of the structural changes accompanying the buildup of the chromatic and achromatic apparatus during normal mitosis in animal cells. Particular attention was devoted to controversial or less well-studied aspects of mitosis, such as kinetochores and the nuclear envelope. (2) To study possible structural alterations in the mitotic apparatus produced by the agents applied or by the presence of aberrant chromosomes (e.g., dicentrics). (3) To elucidate the fine structure of aberrant chromosomes, with particular reference to kinetochores and chromosome architecture.














MATERIALS AND METHODS


Cell Culture

The PtK2 cell line was initiated in 1961 by Kirsten H. Walen (Walen and Brown 1962) from a kidney of an adult male rat kangaroo, Potorous tridactylis (Marsupialia). The male karyotype consists of ten autosomes, one X chromosome bearing the nucleolus organizer, and two Y chromosomes (Figure 1. Shaw and Krooth 1964s Walen and Brown 1962).









1 3 x




4 5 Y1 Y2



Fig. 1. Karyotype of the male rat kangaroo (Potorous tridactylis). Note the heterochromatic region near the centromere on the X chromosome (nucleolus organizer). Drawn after a micrograph by Shaw and Krooth (1964).





The PtK2 line, however, is aneuploid, the number of chromosomes in the reference stock varying from 11 to 14 (Anonymous 1967). An extrat long subtelocentric chromosome is present in the majority of cells.








According to Walen (1965) the generation time in Eagle's medium with 47. fetal calf serum (FCS) is 28-32 hr. The low number and the individuality of chromosomes make this line particularly suitable for cytogenetic and labeling studies.

I obtained an ampule of frozen cells from the American Type Culture Collection and started a culture in April, 1970. After preliminary experiments this culture was given up in September, 1970, because of a suspected contamination by a microorganism (mycoplasma or fungus). A new culture with fresh cells was initiated in October, 1970. Only cells from this culture were used in the experiments described.

Cells of the stock culture were grown in square glass bottles in Eagle's minimum essential medium (MEM; GIBCO), supplemented with 10% FCS (GIBCO). Cells were harvested with trypsin-versene at weekly intervals and subcultured at appropriate dilutions. Fresh complete medium (HElI with 10% FCS) was supplied once between transfers. Cells to be used for experiments and controls were seeded in Falcon 30-ml flasks and left to attach overnight. Cultures were maintained at 370C, but handling was carried out at room temperature.


Chemical and Physical Treatments


Streptonigrin

Streptonigrin (Pfizer and Co., Inc.) was obtained from Cancer Chemotherapy, NCI, NIH, Bethesda, Md. A stock solution of 250 pg/ml was prepared shortly before use by dissolving the powder in the diluent supplied. The stock solution was diluted volumetrically in distilled water, pH 8.0, to a concentration of 0.25 pg/ml. After sterilization with a millipore filter, the final concentrations of 0.05 pg/ml and








0.01 pg/ml were prepared under sterile conditions by dilution with NEM. No serum was added to avoid possible binding of SN by serum proteins.

Cell monolayers in Falcon 30-ml flasks were rinsed once with MEM before the drug was added. Control cells were treated in the same manner, except that they were exposed to IEII alone instead of the drug. The cells were incubated for 4 hr at 370C. The flasks were then rinsed twice with complete medium and the cells were left to recover in the incubator in a third change of fresh complete medium. Approximately 24 hr later the medium was changed once more. After a total recovery period of 48 hr the cells were fixed in situ according to the standard schedule (see below).


Colcemid

A stock solution of 0.5 pg/ml was prepared by dissolving crystalline colcemid (CIBA) in distilled water, pH 6.4. This solution was sterilized with a millipore filter and further dilutions were prepared with sterile complete medium. Treatment A was 0.05 pg/ml colcemid for 2 hr at 37oC, followed by two rinses with complete medium and recovery for 1 hr at 37 C in a third change of medium. Treatment B was 0.25 pg/ml colcemid for 15 min at 37 C without a recovery period. Control cells were left undisturbed and were fixed at the same time as treated cells.


Cold

Flasks were seeded as usual and the cells left to attach overnight at 37 C. The medium in the experimental flask was then poured off and replaced by complete medium of 0-40 C. The flask was immediately placed in a pan with ice water so that its lower portion was immersed. After








1 hr exposure the medium was replaced by cold glutaraldehyde of the usual strength and buffered as usual (see fixation schedule). Initial fixation was 15 min in the cold, after which the flask was brought to room temperature and the standard fixation schedule was followed from here. One control consisted of cells kept at 37oC and fixed cold as above. The second control consisted of cells not exposed to cold and fixed at room temperature.


Fixation and Embedding

The standard fixation schedule, modified from Brinkley et al. (1967), was as follows:

(I) Decant the culture medium and add 3.1% glutaraldehyde (GA)

buffered with Millonig's phosphate buffer without sucrose (Millonig 1961), pH 7.3. After a few minutes add a fresh change of GA. The cells are fixed for a total of 1 hr at

room temperature.

(2) Rinse with two changes of buffer, 10 min each, at room

temperature.

(3) Postfix for 1 hr at room temperature in similarly buffered 27.

osmium tetroxide (OsO 4).

(4) Dehydrate in 25% and 507. ethanol, 10 min each. Prestain in

cold 2%. uranyl acetate in 70%. ethanol for 2 hr to overnight.

Rinse with two changes of cold 75% ethanol, 10 min each. Ten

min 90%. ethanol; the cold solution is added and the flasks then brought to room temperature. Subsequent steps are all carried out at room temperature; three changes, 10 min each, of 907. hydroxypropyl methacrylate (HPMA); 15 min each of 95%

and 97% HPMA.








Embed in Luft's Epon (Luft 1961) as follows. A ratio of 1 part of mixture A to 2 parts of mixture B gave the best results.

2 parts HPMA : 1 part Epon for 15 min

Equal parts of HPMA and Epon for 15 min

1 part P1~A : 2 parts Epon for 30 min

2 changes of pure Epon, 30 min each

Add fresh pure Epon and burn holes in the top of the culture flasks with a hot glass rod. Drain off excess Epon until a layer about the thickness of a glass slide is left. Leave overnight at room temperature then transfer to 600C for 24 hr or longer.


Preparation of Cells for Light and Electron Microscopy

The basic procedure was adopted from Brinkley et al. (1967).

After curing, the bottom of the flask with the adhering Epon wafer is cut out. 1.afer and flask bottom are separated by alternately cooling the sandwich in liquid N2 and thawing in tap water. Once separation has started at the edge, the Epon wafer can be snapped loose.

For light microscopy (scoring of aberrations, determination of mitotic indices) wafers are placed cell layer up on the stage of a phase contrast microscope. Detailed examination of cells is possible by using oil immersion objectives.

The area around cells selected for thin sectioning is marked with a sharp needle under low power. Several high magnification pictures at different levels of focus are then taken. A disk with the marked area is cut out with a cork borer of suitable diameter. The piece of Epon is roughly trimmed and glued to the tapered end of a plastic peg, cell layer up. Fine trimming to the boundaries of the trapezoidal or








rectangular area containing the cell or cells is done after curing of the resin glue.

As a rule, blocks were sectioned in a plane parallel to the cell monolayer (horizontal plane; see Figure 2). For special sections (para-sagittal = vertical; para-equatorial = orthogonal to the spindle axis; see Figure 2) blocks were mounted differently. A scratch was made on the surface of the Epon wafer, under low power, either parallel to the spindle axis or at a right angle to it and as close to the selected cell as possible. A disk with the marked cell was cut out, roughly trimmed, and mounted in the clamp chuck of an ultramicrotome. By cutting thick sections with a glass knife the block could be faced to the level of the scratch or even closer to the cell, when periodical checking under a microscope was possible. If necessary, the angle of the face relative to the spindle axis could still be corrected. Finally, the block was mounted on a peg with the face up.

Serial sections were cut with a diamond knife on a Porter-Blum MT-2 microtome and picked up on formvar-coated, carbon-stabilized rectangular hole grids (LKB Instruments, Inc.). Sections were poststained with uranyl acetate (Watson 1958) and lead citrate (Reynolds 1963) and examined in a Hitachi HU ll-E electron microscope with a 50-nm objective aperture and operated at 75 kV.
































Fig. 2. The three planes in which blocks were sectioned. C: Cell in mitosis; the spindle axis is assumed to be along a line between the two pointed ends. HO: Horizontal plane. PE: Para-equatorial plane. PS: Para-sagittal plane.















RESULTS ANID OBSERVATIONS


Normal Nitosis

This section is based on light and electron microscopic observations on untreated cells, either used as controls for the various treatments, or fixed directly from the stock culture. Interphase

Although interphase is not a mitotic stage, it makes sense to

describe the structure and spatial orientation in the interphase cell of those components and organelles that take part in the formation and function of the mitotic apparatus.

Figure 3 shows a grazing section of an interphase nucleus. Poreannulus complexes of the nuclear envelope are cut at different levels. They appear as circles at the level of the envelope, and as circles surrounded by a "halo" at the level of the underlying chromatin. Central dense granules can be seen in some of the profiles. Where the section passes below the complexes, their position is indicated by achromatic holes in the chromatin. Whorls of polyribosomes are visible on oblique sections of the nuclear envelope. In the cytoplasm, two centrioles are present, approximately at a right angle to each other. Obliquely sectioned membrane elements, probably of a Golgi complex, can be seen in their vicinity. Microtubules, though more numerous around the centrioles, pass in various directions, without apparent specific orientation to the latter. Details of centriolar structure at different levels are shown in the four serial sections in Figure 4.

52








Osmiophilic, granular or fibrillar material surrounds the centriole at one end (Figure 4a). The short rays radiating from the triplets (Figure 4b) give the whole structure the appearance of a pinwheel. Such images are rare. Dense, spherical particles with a lighter core and aura (Figure 4c), often quite numerous, are exclusively associated with centrioles and are present during all the stages of the cell cycle. An intracentriolar vesicle, shown in an approximately median section in Figure 4d, is often found in the lumen of centrioles. Satellites and microtubules can be seen in all four sections (Figures 4a-d).

The appearance of the interphase nucleus in transverse sections varies. Three representative nuclei are shown in Figure 5. The chromatin is either dispersed (Figure 5a), or it includes heterochromatic chromocenters to a variable extent (Figures Sb and 5c). Nuclei like the one in Figure 5c are very likely in late cytokinesis or very early prophase. Of 34 interphase nuclei examined, 12 had dispersed chromatin, the remainder contained chromocenters of variable size and extent. At higher magnification, chromatin fibers, approximately 130 A and 250 A in diameter, can be seen cut at various angles (Figure 6). A very dense core, 70-80 A in diameter, appears in a few cross sections of 250 A fibers. Nucleoli are large and prominent in interphase cells (Figure 5). Granular and fibrillar components are mixed (Figure 7). Nucleolus-associated chromatin is found within the nucleolus and at its periphery (Figures 5 and 7). Intranuclear vesicles occur in some nuclei (Figure 5a). The wall of these vesicles is finely fibrillar and they often, perhaps always, contain granules or particles 250-400 A in diameter. However, densely packed granules of similar size and appearance also occur free in the karyoplasm.









Prophase

The centrioles duplicate at the onset of prophase. Therefore, two parent-daughter centriole pairs are found in early prophase cells (Figure 8a). Four serial sections of the same centrioles (Figures 8b-e) reveal that each daughter centriole is closely associated with its parent at approximately a right angle. Daughter centrioles appear shorter than mature centrioles (compare Figure 8b with Figure 3), but have the same diameter (approximately 2,200 A). At this stage relatively few MT converge on, or radiate from, each parent centriole. An early stage of centriole migration is illustrated in Figures 9a and 9b. One of the two centrioles shown (C1) is possibly a daughter centriole. It is cut almost perfectly at a right angle to its axis. The structure and arrangement of the tubular triplets is particularly clear (Figures 9c-e). The cartwheel at the proximal end appears in three adjacent sections (Figures 9b-d), but hub and spokes are most distinct in Figure 9c. The triplets seem to be embedded in amorphous osmiophilic material. Bars, approximately 80 by 480 A, possibly cross sections of plates, appear between the triplets (Figure 9e). Numerous satellites are present in the general area of the two centriole pairs (Figures 9a and 9b). In many, perhaps alls mid-prophase cells the centrioles lie in an invagination or pocket of the nuclear envelope (e.g., Figure lla). At this stage numerous MT are present, forming an "aster" around the centrioles. The radial arrangement of these MT also imposes radial orientation on other organelles, notably mitochordria (Figure llb). Migration of centrioles relative to nuclear changes (chromosome condensation, fragmentation of the NE) varies considerably, so that by








the end of prophase the two pairs may have moved a short distance only, or they may lie at opposite poles (see Figures 18a and 19).

Progressive condensation of chromatin is indicated by the

appearance, during very early prophase, of large heterochromatic patches (compare Figure 8a with Figure 5). Discrete chromosomes are present in mid-prophase cells (Figures 10 and lla). Their "mottled" appearance (see also Figure 13) is suggestive of incomplete condensation. During condensation many, possibly all the chromosomes are attached to the nuclear envelope along their entire length or with their telomeres (Figures lla and 13). In transverse sections of nuclei in early prophase the chromosomes appear to be attached by "stalks" (Figure 13), but grazing sections reveal that this appearance is due to

achromatic "holes," enlarged compared to interphase (compare Figure.12 with Figure 3). In late prophase, however, strand-like connections between chromosomes and the nuclear envelope are real (Figure 14).

No kinetochores are discernible in early prophase (Figure 8a).

In sections of nuclei in mid- and late prophase kinetochores appear as roughly circular patches of finely fibrillar material in slight constrictions of chromosomes (Figures lla, 15, and 17). Serial sections revealed that these kinetochores are globular and 5,000-8,000 A in diameter. I cannot state with certainty that the diameter decreases with advancing prophase. The doubleness of the chromosomes is evident in sections showing both sister kinetochores (Figures 15 and 17). The lesser electron density of the kinetochores compared to the chromosomes is very distinct (Figures 10, lla, 15-17). Slightly more opaque kinetochore granules at the surface of the sister chromatids (Figure 17) are very rarely seen.








The nucleolus, still intact in very early prophase (Figure 8a), fragments into several masses of granular material (Figures 10 and 1la), some of which are apparently associated with chromosomes (Figure lla).

The nuclear envelope remains intact until the end of prophase, but the pore-annulus complexes become more fuzzy (Figure 12). Polyribosomes are found on the nuclear envelope throughout prophase (e.g., Figure lla).


Prometaphase

The breakdown of the nuclear envelope, indicating the transition from prophase to prometaphase, is gradual, but structural changes involving MT and kinetochores are more striking. Fragmentation of the nuclear envelope always begins nearest the centrioles (Figures 18a and 19). Fragments are undulated or form vesicles. The telomeres of chromosomes at the periphery of the "nucleus" are often trapped in compartments formed by undulated, still intact portions of the nuclear envelope (Figure 18b). With progressive development of the spindle apparatus, envelope fragments become smaller and scarcer, but some persist at the periphery of the spindle until late prometaphase (Figures 20, 26, and 29). Disappearing central granules indicate early stages of the breakdown of pore-annulus complexes (Figures 21a and 21b). The entire complexes on fragments of the nuclear envelope become fuzzy and disappear by mid-prometaphase (Figure 21). Polyribosomes can be found on clearly identifiable fragments until late prometaphase (Figures 21, 26, and 29).

Chromosome condensation continues throughout prometaphase (Figures 18-20, 26-30). The chromosomes detach from the nuclear envelope as the






57

latter fragments, and they gradually lose their '"mottled" appearance. Most kinetochores in very early prometaphase still resemble prophase kinetochores (Figure 22a), but some are differentiating into more complex structures. For example, the sister kinetochores of the chromosome in Figure 22b, although apparently not attached to MrT, exhibit a slightly denser band within the fibrillar matrix. Such internal structures may be more common than it appears, but due to their relative indistinctness they would not appear in ever so slightly oblique sections. A rather unusual case is illustrated in Figures 22c and 22d. This chromosome of the cell shoin in Figure 18 was situated in the area distant from the centrioles, i.e., near the intact portion of the nuclear envelope. Several obliquely sectioned MT lie outside the envelope opposite the "outer" kinetochore, which also contains a short band. Serial sections revealed no 1T inside the nuclear envelope. The other sister kinetochore, facing in the direction of the distant centrioles, resembles a typical prophase kinetochore.

As prometaphase progresses, kinetochores become more variable in

appearance. Structural differentiation depends in each individual case upon the position and orientation of the chromosome relative to the spindle poles. Sister kinetochores of chromosomes near a pole are dissimilar (Figures 23 and 24). As a rule, the kinetochore facing the near pole is attached to MT and consists of moderately opaque, finely fibrillar material (iK, K2, in Figure 20; K2 in Figures 23 and 24). A banding pattern is discernible in some of these kinetochores (Figures 23a and 23b), which are quite often also stretched (K, K2, in Figure 23a; K2 in Figures 24b and 24c). This makes it impossible to determine where the MIT end. In contrast, the kinetochore facing the distant pole








resembles typical prophase kinetochores (e.g., K in Figure 24), or it consists of a dense band embedded in the fibrillar matrix (KI in Figures 23a and 23c). The band can be seen in several adjacent sections, which suggests it represents a transverse section of a flat or convex plate approximately 300 A thick and 3,500 A in diameter. In each case few or no MT are associated with such a kinetochore.

Sister kinetochores of chromosomes lying near the future equator of the spindle are similar if both are unobstructed, i.e., if there is no nearby obstacle (such as another chromosome) between them and the respective pole (Figure 25). Usually, these kinetochores are more or less stretched, with or without bands, and attached to MT. If one of the sister kinetochores is obstructed by a neighbor chromosome lying very close, it resembles prophase kinetochores (not shown, but similar to K in Figure 24a), while the other kinetochore resembles the unobstructed kinetochores described above.

Hicrotubules are found in the "nucleus" as soon as the nuclear

envelope breaks down (Figures 18, 22, and 23). At first they are more numerous in the vicinity of the centrioles (Figure 18), but as the latter take up their position at opposite poles, MT are abundant in the center of the spindle (Figure 20). Surprisingly, most of the MT are associated with kinetochores and few, if any, continuous tubules seem to be present. However, a more careful, quantitative analysis would be necessary to establish this with certainty.

Remnants of the nucleolus disappear completely and the nucleolar organizer appears on the X chromosome (Figures 19 and 25).

Figures 26-30 illustrate further progression of prometaphase.

Fragments of the nuclear envelope, still found between and around some









chromosomes during mid-prometaphase (Figure 26), become smaller and move to the periphery of the spindle in late prometaphase (Figures 29 and 30a). Nuclear pore complexes have disappeared approximately at the stage shown in Figure 26, but some polyribosomes are still present on the fragments of the nuclear envelope. Peripheral segments of double membrane in late prometaphase cells have ribosomes on both faces, thus resembling rough ER (Figure 29).

As the spindle develops, mitochondria, ER, and other large cellular organelles and components are excluded from its area (compare Figure 20 with Figure 29), but ribosomes, mostly monosomes, are abundant. Bundles of MT in the center of the spindle run more or less straight along the pole-to-pole axis, while peripheral bundles are slightly arched (Figures 26 and 29). An exceptional case is shown in Figure 32, where rather disoriented NT curve sharply towards the centriole. Angular configurations of kinetochore IT, as shown in Figure 20, no longer occur. Still, most of the bundles of MT are associated with kinetochores (Figures 26-29, and 30a). It is very difficult to determine whether the MT of other bundles (e.g., Figure 28) are interpolar, or whether they are associated with kinetochores not visible in the same section. Quite frequently, a bundle of IMT passes near the kinetochores of a maloriented chromosome (e.g., Figure 39). Transchromosomal MT are outstanding in rather thick sections (Figure 30b). Skew MT, as well as "wavy" NT, occur quite frequently, but they always make up a small proportion of the total number of spindle IT (Figures 27, 28, 30a, and 33).

Orientation of the centrioles relative to the axis of the fully formed spindle is variable (Figure 29). This has also been confirmed in cells in meta- and anaphase.









The chromosomes condense further until they appear as solid, very electron dense rods or blocks (compare Figure 26 with Figure 29). Sister chromatids are still tightly joined, but their individuality is apparent in transverse or peripheral longitudinal sections of chromosomes (Figures 27, 29, 30a, 31, and 33), or in situations where the sister kinetochores are stretched (Figure 34). In the latter case, one or several achromatic holes appear in the kinetochore region.

The diversity of kinetochore structure is still remarkable during

mid- and late prometaphase (Figures 26-29, 30a, 31, 33-41). At a glance there seems to be no general pattern of differentiation, but a careful comparison of the structure of sister kinetochores with the position of the chromosomes in the spindle gives a better insight. First, we can distinguish between chromosomes lying on or near the equator of the spindle (Ch1 and Ch2 in Figure 27; Ch3 in Figure 28; the chromosomes in Figures 30a and 33-36), and chromosomes lying closer to one pole (malpositioned chromosomes; Figure 26; Ch3 in Figure 27; Ch1 in Figure 28; Figures 29 and 37-40). Generally speaking, more chromosomes occupy an equatorial position in late than in early prometaphase (Figures 20, 26, and 29). The sister kinetochores of equatorial chromosomes are similar if both are unobstructed, i.e., if no major obstacle lies along a line from the kinetochore to the pole (Figures 27, 30a, 33, 34, and 36); they are dissimilar if one is unobstructed, the other obstructed by a neighbor chromosome (Figures 33 and 35). Unobstructed kinetochores are irregularly shaped and fuzzy (Figures 28 and 36), fuzzy cones with a trace of bands (similar to K in Figure 38), or distinctly triplebanded (Figures 27 and 33-35). It appears that the clarity of the three bands increases with advancing prometaphase. Unobstructed









kinetochores are quite often stretched (Figures 27, 28, 31, 33, and 36), and they are always attached to bundles of MT.

Obstructed kinetochores of equatorial chromosomes appear as patches of finely fibrillar matrix within which a denser band or patch can be seen (Figures 33, 35, and 41). Serial sections make it clear that the bands do not represent fibers, but sections of oddly shaped threedimensional structures (e.g., Figures 35 and 41). The width of the bands varies from 250-400 A. In appearance and dimensions they are comparable to the outer layer of triple-banded kinetochores, except that they often exhibit a partly double-banded, partly beaded substructure (Figure 41). Furthermore, obstructed kinetochores are always attached to very few, if any, MT.

Sister kinetochores of maloriented chromosomes most commonly are also dissimilar. As a rule, the kinetochore facing the near pole resembles the unobstructed kinetochores of equatorial chromosomes (Figures 27, 37, and 38), while the kinetochore facing the far pole resembles obstructed kinetochores (Figures 27, 37, and 38). Figure 40 illustrates a rare exception to this rule. On both these maloriented chromosomes the kinetochore oriented towards the near pole is compact, and attached to liT, while the other kinetochore is fuzzy and stretched, but also attached to LIT.

Figure 39 illustrates another rare case. The sister kinetochores of this maloriented chromosome are very similar in every respect; both are relatively undifferentiated masses of matrix with only a trace of a band. Most remarkable is that the matrix seems to cover almost the entire small side of the chromosome (Figure 39c).









The difference in electron density between the kinetochore matrix and its band on one hand, and the chromosome proper on the other hand, is obvious in many of the figures cited, but particularly in grazing sections of kinetochores (e.g., Figures 33b, 35c) and 36). Interesting, but very rarely seen, are chromatin "strands" between stretched sister kinetochores (Figures 31 and 36). It is difficult to determine, even at high magnification, if the fibers in these strands are finer than normal chromosomal fibers, or if they are more densely packed, and arranged more or less in parallel. Metaphase

Cells in full metaphase are quite rare, probably because this

stage is of very short duration. Most cells judged to be in metaphase, based on light microscopy, turn out, upon examination of thin sections, to be either in very late prometaphase or very early anaphase.

Chromosomes at metaphase are aligned on the equator of the spindle, at least with their kinetochore region (Figure 42). Typically, the non-aligned telomeric portions of the long chromosomes extend beyond the periphery of the spindle into the cytoplasm (insets of Figures 42-44). All the kinetochores in normal metaphase cells are triple-banded (Figures 42 and 43). Bundles of kinetochore IMT converge towards the poles (Figure 42). Within a bundle the MT are more or less parallel, but a few non-kinetochore MT slant across. Wavy IT occur mainly between chromosomes. Overall, the paucity of non-kinetochore MT is remarkable.

Figure 43 represents a para-sagittal section. The plane of sectioning was not precisely at a right angle to the chromosomes,









therefore only one of the two sister kinetochores can be seen. Kinetochore profiles are virtually identical with those seen in horizontal sections (compare Figure 43 with Figure 42). Sister chromatids are separated by grooves relatively devoid of ribosomes and ground substance. However, the separation, which actually indicates late metaphase to very early anaphase (see also inset of Figure 43), is incomplete, for large "bridges" still connect the sister chromatids.

In para-equatorial sections the metaphase plate appears as shown in Figure 44. Three kinetochores with associated MT can be seen. Other MT occur singly or in small clusters, or they are arranged in bundles which may belong to kinetochores not included in this section. Microtubules penetrating chromosomes are surrounded by a clear halo.


Anaphase

Sister chromatids, now more properly called daughter chromosomes, separate from each other at the onset of anaphase. In the light microscope they still appear as parallel rods (Figure 45, inset). Because the chromosomes are somewhat frayed, their separation at very early anaphase (Figure 45) is less distinct in thin sections, except for the kinetochore region. The latter is spaced farther apart than the telomere region (Figures 45 and 46), and this trailing of the telomeres is more pronounced during later anaphase (Figure 47; inset of Figure 50; Figure 51a). The daughter chromosomes lose their individuality during late anaphase and each set forms a large mass of densely packed chromatin near the respective pole (Figure 52). The nucleolus organizer (O10) is enclosed in this mass.









A striking phenomenon is illustrated in Figures 46-48. Daughter chromosomes in very early and early anaphase are connected by electron dense strands. First believed to be an aberration, it was found in six of seven early anaphase cells. The chromosomes of the seventh cell had short strands of similar appearance extending from the rather widely separated kinetochore regions into the interzone. The same was observed in one mid-anaphase cell. In the latter two cases the strands were extremely tapered towards the interzone, giving the impression of connections gradually drawn out and finally ruptured. In the majority of early anaphase cells the strands connect the kinetochore regions of daughter chromosomes. I have counted as many as six strands in serial sections. However, in the cell shown in Figure 47, and also another very similar anaphase, a few strands connected the mid-region of daughter chromosomes.

The diameter of these strands varies from approximately 400 A to 900 A. Short strands are usually thicker, and long strands very frequently are thinnest approximately midway between daughter chromosomes, while they are thicker at their base (i.e., point of attachment; Figures 46 and 47). The staining properties of the strands are identical with those of chromosomes, and the structure is that of fine and/or densely packed fibers.

Kinetochores vary in appearance from more or less straight to

convex, "stalked," or angular (Figures 46-49). The three layers are very distinct except in sections that cut a kinetochore obliquely or peripherally (Figures 46 and 47). In para-equatorial sections the kinetochores appear as circles of moderate electron density, within which cross sections of MT can be seen (Figure 50). In late anaphase









the triple-layered character of the kinetochores is less clear, but still recognizable (Figures 51b and 52). Most kinetochores in very late anaphase cells appear in depressions on the poleward face of the chromatin mass (Figure 53). The less dense band in Figure 53 is approximately 500 A wide and set off from the chromatin by a 250 A wide clear band. The chromatin immediately underlying the kinetochore is denser and less obviously fibrillar than the remainder of the chromosomal mass. Remarkable is the decreasing number and degree of organization of kinetochore MIT in late anaphase (compare Figures 46-49 with Figures 51-53).

Spindle elongation is the rule in PtK2 cells (Figure 51a). The interzone is still free of cytoplasmic organelles even after the chromosomes have almost reached the poles. Pieces of double membrane,

some with ribosomes, are found at the periphery of the spindle area, particularly around the two sets of chromosomes (Figure 51a). Interzonal HT are scarce. Mitochondria, vesicles, and ER invade the interzone in very late anaphase (Figures 52 and 55b). The cytoplasm begins to constrict in the equatorial region (Figure 52, inset), where stem bodies appear. The latter consist of amorphous, osmiophilic material, within which MT are closely packed (Figures 54 and 55a). It is difficult to determine from these sections whether the IT terminate in the stem bodies or beyond. Both possibilities are likely. One MT in Figure 55a clearly passes through the stem body. Its total length visible in this section was 3 p.

It is possible that pieces of double membrane as seen in Figure 51a become involved in the reconstitution of the nuclear envelope. More typically, however, small cisternae and vesicles appear on the









surface of the chromatin masses in late anaphase (Figure 52). Oblique sections of such membranes reveal the presence of nuclear pore complexes. Some of the cisternae obviously are, or have originated from, RER as indicated by the presence of ribosomes (Figure 52). The location of the membrane pieces varies, depending on the level of the section. Usually, the majority of the membrane cisternae and vesicles is apposed to the lateral faces of the chrcatin masses.

Figure 56 shows a pair of centrioles in early anaphase. As

during prometa- and metaphase, the moderately osmiophilic, amorphous material surrounds that centriole on which the IT more or less converge. Direct connections between MiT and any part of the centrioles are never observed.


Telophase and Cytokinesis

Reconstruction of the nuclear envelope continues during telophase, apparently by coalescence of the small vesicles and cisternae seen in anaphase cells. Extensively reformed NE first appears on the lateral faces of the chromatin mass, as well as on the polar face, except directly opposite the centrioles (Figures 57 and 58). Nuclear pore complexes, complete with central granules, are present on chromatinassociated membranes irrespective of size and location (e.g., Figure 58). The spacing between the two membranes of the NE remains irregular until late telophase (Figures 57-60). Quite frequently, pieces of NE are trapped within the reforming nucleus, probably giving rise to the deeply invaginated pockets seen in many interphase cells.

Figures 60 and 61 illustrate progressive decondensation of the chromatin. This process begins even before the NE is completely










restored, but large heterochromatic patches persist until interphase (late stages of cytokinesis).

The less dense component of kinetochores, still present during early telophase (Figure 57), has disappeared by the time the NE is completely reconstructed (Figures 59 and 60). The inner, dense layer becomes an extremely osmiophilic patch on the inner membrane of the NE (Figures 59 and 60). Most commonly, the entire NE is indented at these sites, and quite frequently the inner membrane with the dense patch is deeply invaginated (Figure .60). Microtubules can be seen extending poleward from these kinetochores, but the nature of this association cannot be determined precisely.

During cytokinesis a cleavage furrow in the former equatorial region separates the two daughter cells, except for a stem, whose length and diameter vary depending on the distance between daughter cells (Figures 61 and 62). The stem bodies seen in late anaphase and early telophase have fused to form the midbody, an irregular band of osmiophilic material in the stem. Numerous MT from both daughter cells converge in the midbody, but the dense material obscures details in these paraxial sections. Normally, the midbody remains between the daughter cells even after these have completely separated and moved far apart from each other. Eventually, it seems to be lost. On rare occasions the midbody is included in the peripheral cytoplasm of one of the daughter cells, but this is most likely an abnormal condition.


C-Mitosis

Treatment A (0.05 jg/ml colcemid for 2 hr 1 hr recovery)

produces an accumulation of cells with "scattered" metaphases (Figures









63 and 64). C-mitotic cells are much less compressed along the axis vertical to the growth surface than normal mitotic cells. The extreme is illustrated by the near-spherical cell in Figure 63. Differences among c-mitotic cells are evident in the variable shape at the level of the light microscope (compare the insets of Figures 63 and 64). Ultrastructural variations relate to the distribution of membrane elements and mitochondria. For example, in the cell shown in Figure 63 membrane vesicles and cisternae, as well as mitochondria, are distributed throughout the area occupied by the chromosomes, although the former are more numerous at the periphery. In contrast, the area of the chromosomes in Figure 64 contains very few small vesicles. Numerous vesicles and larger cisternae, some of them rough ER, are arranged almost concentrically at the periphery of this area. Mitochondria are also excluded. In this respect the cell in Figure 64 resembles a prometaphase more than a metaphase, but the degree of condensation of the chromosomes does not bear this out.

The centrioles, embedded in amorphous or finely fibrillar material of moderate electron density, were always found at the periphery of the area occupied by the chromosomes in the few cells examined (Figures 63 and 64). Four centrioles are present in each cell. They do not differ structurally from centrioles in untreated cells.

Kinetochores appear as dense bands embedded in a less dense fibrillar matrix and following the curvature of the chromosomal surface, or as patches of less dense material (Figures 63, 64, 92-98). This depends on the angle and the level of the section. No MT at all were found in such cells. Treatment B (0.25 pg/ml colcemid for 15 min, no recovery) produces strikingly different'effects. As with treatment









A, the usual criteria for determining mitotic stages do not apply. Nevertheless, light microscopic examination of cells subjected to treatment B revealed stages resembling pro- and prometaphase rather than metaphase. One cell was in late telophase or cytokinesis. The chromosomes, fairly distinct in the light micrographs (insets of Figures 65 and 66), were extremely difficult to see by direct observation.

Electron microscopy confirmed the above observations and revealed interesting details. The central area of the cells in Figures 65 and 66 consists of a coarsely granular or fibrillar ground substance in which the chromosomes are embedded. The separation of this area from the cytoplasm is almost perfect; with the exception of small membrane vesicles all the larger organelles are excluded. In the cell in Figure 65 large pieces of double membrane, probably fragments of the NE, are present at the border between central area and cytoplasm. Grazing sections revealed no pore-annulus complexes on these fragments, although colcemid does not destroy their integrity on the NE of interphase nuclei (Figure 68). Chromosomes are highly dispersed, making the recognition of sister chromatids difficult (Figure 65). The dark centromeric granules of chromosomes in the light micrograph are patches or balls of more tightly packed, perhaps also finer, chromosomal fibers. There seems to be one patch per sister chromatid, and, as revealed by serial sections, connections between chromatids exist in this region. Kinetochores in the usual sense are lacking, but a vesicular space relatively poor in fibers and filled with a less dense substance can be recognized adjacent to the dense patches.








The cell in Figure 66 differs from the above description by the absence of large membrane cisternae and the higher degree of chromosome condensation. It is remarkable, in this context, that the chromatin of all the interphase cells subjected to treatment B is completely dispersed (compare Figure 67 with Figure 5). The fibers in the primary constriction of the chromosomes in Figure 66 are also finer, or more tightly packed, or both. In this case, however, a distinct less dense kinetochore band is present in the vesicular space.

Despite the relatively high concentration of colcemid used in

treatment B, a few MT were found in interphase cells (Figure 68), but not in c-mitotic cells. In some of the latter, numerous bundles of microfibrils, approximately 70 A in diameter, were present in the central area (Figure 69).


Mitosis in Cold-Treated Cells

Exposure to 0-4 C does not accumulate metaphase cells. Cells at all stages of mitosis are present in the unsynchronized cultures used. The general ultrastructural features of prophase correspond to control cells, except for the presence of intranuclear clusters of granules approximately the size of ribosomes (Figure 70). Kinetochores of cells in mid-prophase consist of less dense, fibrous material in constrictions of chromosomes (Figure 72). A mitotic spindle in the usual sense does not exist in prometa-, meta-, and anaphase cells, although centrioles are situated at opposite poles (Figures 71 and 75). Most of the few MT present are associated with kinetochores. An amorphous moderately osmiophilic substance coats the microtubules, whose walls appear as stark lines (Figures 73 and









75). Fully differentiated kinetochores exhibit the familiar triplebanded profiles (Figure 75), though less distinctly than in normally fixed control cells. The fuzziness of kinetochores and MT is not caused by the initial cold fixation, because these structures are well preserved in similarly fixed cells not previously exposed to cold (Figure 74). The paucity of MT is also obvious in the midbody region of cells in cytokinesis (Figure 76). Centrioles are not structurally altered by exposure to cold (Figure 77).

Chromosomes are normally condensed, except for achromatic holes, which are prominent at all stages of mitosis (e.g., Figure 71). Remarkable is the great number and increased clarity of chromosomal granules in cold-treated compared to control cells (Figures 70-72, and 75).


Kinetochore Fine Structure

I have already described in detail the fine structure of kinetochores from prophase to late prometaphase. To establish a basis for the discussion of kinetochore models, this section is devoted to an in-depth study of kinetochores in normal meta- and early anaphase, colcemid- and cold-treated cells.

Figure 78 shows five of seven serial sections in the horizontal plane of a kinetochore in a cell in very early anaphase. Most of the sections were too thick (approximately 800 A) to reveal details concerning the attachment of MT. However, two things are very clear: the triple-banding, and the greater length of the kinetochore in the median sections (Figures 78b and 78c) compared to the peripheral sections (Figures 78a, 78d, and 78e). Thinner sections (500-600 A),









also approximately median, of other kinetochores are presented in Figures 81-84. These kinetochores were sectioned transversely, except those shown in Figure 83, where obliquely cut kinetochore MT indicate the sections were tangential. Kinetochore profiles very similar to those in horizontal sections can be seen in para-sagittal sections (Figure 85).

Kinetochores in metaphase cells are rarely flat. Most commonly, they are undulated or S-shaped (Figure 42), more seldom concave (Figure 85a). In very early anaphase the kinetochores of small chromosomes in the center of the spindle are more or less flat (Figures 46, 48, and 82), and at a slightly later stage the kinetochores of such chromosomes are convex (Figure 84). The kinetochores of the long chromosomes at the periphery of the spindle are convex or more irregular in early anaphase (Figures 47 and 49). In mid-anaphase, S-shaped and more exotic profiles of kinetochores are prevalent.

Triple-banded kinetochores in paraxial median sections are 4,0006,700 A long. A faintly staining corona, approximately 400 A wide and consisting of fine fibrils embedded in an amorphous matrix, covers the kinetochores on the poleward side (Figures 78 and 84). The width of the three bands varies, both within and between kinetochores. Average values were 390 A for the outer, 270 A for the middle, and 400 A for the inner band. The outer band consistently stains less intensely than either the inner band or the chromosome proper (Figures 84 and 85a). This is very clear in the image seen on the screen of the electron microscope, but in micrographs printed on contrasty paper the difference is obscured (e.g., Figures 78, 82, and 83). Figure 80 shows a peripheral section of a stretched kinetochore in a late prometa- or









metaphase cell. This section was picked up on an uncoated 200-mesh grid. The difference in contrast between the chromosome and the kinetochore is undeniable.

The basic structure of the outer band is finely granular in highly condensed kinetochores (Figures 789 81-84), but in grazing sections of less condensed kinetochores, 30-50 A fibrils are visible (Figure 80). Superimposed on the fine granularity of condensed kinetochores is a structure of coarse granules or fibers, giving the band a knotted appearance (Figures 78, 81-85). The structure and electron density of the middle band are very similar to the corona (Figures 78b and 84). The inner band is continuous with the chromosome (Figures 78b and 84), but in very early anaphase it may be connected to the main body of the chromosome by a "stalk" of chromatin, giving it the appearance of a mushroom (compare Figure 84 with Figure 46). The fibers of the inner band seem to be identical with the fibers of the chromosome. The greater opacity may be due to denser packing, but an interesting alternative is the presence of a very fine amorphous substance, which is lacking in the remainder of the chromosome.

Microtubules attach to the kinetochore at a variable angle.

Three conditions must be fulfilled in paraxial sections to determine how far the 2T penetrate into the kinetochore: (1) The section must be thin (500 A); (2) the UT in question must not be cut obliquely near the kinetochore; and (3) the section must be approximately median. These conditions are fulfilled in Figure 82. The MT marked with an arrow penetrates the outer layer and ends at the interface with the middle layer. The different impression created by an obliquely sectioned MT is demonstrated in the somewhat thicker section of Figure









81. The straight MT marked with an arrow terminates in the outer layer, while the obliquely sectioned MT marked with an arrowhead seems to penetrate into the inner layer. Skew MT passing in front of a kinetochore occur occasionally (Figure 79).

One metaphase cell was sectioned in a para-equatorial plane from one pole across the metaphase plate into the opposite half-spindle. Of the 24 kinetochores examined, all were apparently cut head-on (e.g., Figures 86 and 87), except one, which was sectioned at a slightly oblique angle. This kinetochore belonged to a chromosome at the periphery of the metaphase plate. To determine which of the T of a chromosomal bundle were kinetochore MT, serial micrographs of eight different chromosomes and associated MT, at final magnifications between 40,000 and 62,500, were analyzed. In the last section of the series, HT in the vicinity of the chromosome were marked with one color. Proceeding poleward, newly emerging MT at the kinetochore were marked with a different color. The average number of kinetochore MT was 26, the range 16-40. This agrees well with estimates from paraxial serial sections. I was unable to ascertain if the number of kinetochore MT is correlated with chromosome size. The average number of MT bypassing the kinetochore was 5 (range 0-9). It is rather arbitrary to choose these bypassing MlT from the population of non-kinetochore tubules, the only criterion being their proximity to the chromosome in question.

The kinetochores in para-equatorial sections are roughly circular patches of variable electron density, depending on the level of the section (Figures 86-88, and 91). The diameter varies from approximately 3,400 A for obviously convex kinetochores to 6,000 A for flatter kinetochores.









The kinetochore of Figure 86 was most likely similar to those in Figures 78, 81, and 82. The first section of the kinetochore itself shows a moderately opaque patch of finely fibrillar material (Figure 86d). I interpret this as the outer layer. Fewer MT are visible than in the preceding sections, indicating they terminate at this level. Remarkable are less opaque circles whose diameter is similar to the inner diameter of MT. They mark the terminals of kinetochore MT, the wall of which cannot be seen because its opacity is the same as that of the outer layer. A few HT can be followed one section farther (Figure 86e), because the kinetochore is not perfectly flat, but as the sections pass through the inner layer (Figure 86f) and through the chromosome (Figures 86g and 86h) all the kinetochore NT have disappeared. The middle layer is always obscured by the more opaque outer or inner layers, because the sections are thicker than the middle layer. Even a very thin section would have to pass just between inner and outer layers of a perfectly flat kinetochore in order to show middle layer only, a very unlikely event.

Figure 87 shows two adjacent sections of two more convex kinetochores. The moderately opaque patches in Figure 87b again represent part of the outer layer in which terminals of IMT are visible. The kinetochore shown in Figure 88 was most likely similar to those in Figure 83. The section grazed the apex of the inner layer (large arrow) which seems embedded in the less opaque outer layer. The latter is clearly set off from the chromosome.

Occasionally, a single MT is found to penetrate the kinetochore

and to extend deeply into the chromosome (Figure 89). Whether such MT pass completely through the chromosome is not clear.









I found no intertubular connections and arms on MT in sections

passing close to the kinetochores. Figure 90 shows what presumably is a bundle of kinetochore MT at a greater distance from the chromosome. Two apparent cross-bridges and two arms can be seen.

Anaphase kinetochores in para-equatorial sections are very

similar to metaphase kinetochores, except that most of them are more convex, i.e., MT at the periphery of the kinetochores terminate one or two sections after the apical MT have disappeared (Figure 91).

Kinetochore profiles in colcemid-treated cells are quite different from the triple-banded structures in untreated cells (Figures 9298). Figure 92 represents an approximately median longitudinal section of two chromosomes. A longitudinal section along line C-D and perpendicular to that of Figure 92 would produce a face-on view of the kinetochore as in Figure 93. A transverse section along line A-B, also perpendicular to that of Figure 92, would produce an image as in Figure 94. The kinetochore bands in sections such as in Figures 92 and 94 are approximately 400 A wide, are embedded in a less opaque, fibrillar matrix, and closely follow the surface of the chromosome. The bands are slightly less electron dense than the chromosomes (Figure 92), but again this characteristic is obscured in contrasty prints (Figures 94, 97, and 98). I consider these bands equivalent to the outer layer of kinetochores in untreated cells. There is no inner layer on any of the chromosomes in colcemid-treated cells, and the kinetochores are never attached to NT. The outer layer is apparently made of two 150 A sheets held together at various points. This accounts for the transverse and longitudinal sections showing two









bands knotted together (Figures 92, 94, 97, and 98), as well as for the fibrous structure in grazing sections (Figures 93, 95, and 96).

Further evidence for the contention that the bands are actually transverse sections of irregularly undulated (K1 in Figure 92; Figures 94 and 97), convex (K in Figure 92), or almost flat sheets (K2 in Figure 98), came from careful analysis of serial sections of chromosomes whose orientation relative to the plane of sections was known from phase contrast micrographs. The diameter of these sheet-like kinetochores is 5,000-9,500 A. Values of 7,000-8,000 A are most

oimon.

Figure 95 illustrates rather unusual, exotic kinetochore profiles. Kinetochore no. 1 seems to consist of two bands converging at their ends. Additional bands were visible at the same locus in three adjacent serial sections. Similar observations were made on the undulated kinetochore no. 2.

In contrast to the single-banded kinetochores of colcemid-treated tells, the metaphase kinetochores of cold-treated cells are triplebanded as in Figure 75. The bands are fuzzier, but their relative opacity is very similar to that of kinetochores in untreated cells. Kinetochores of cold-treated cells are also attached to MT, but these are few in number.


Chromosomal and Mitotic Aberrations Untreated Cells


Frequency of aberrations

To score the frequency of chromosomal and mitotic aberrations, Epon wafers with the embedded cell monolayers were scanned with the









phase contrast microscope at a magnification of 300. Obviously aberrant, as well as doubtful cells were examined at magnifications of 625 and 1,560. Acentric fragments, maloriented chromosomes, and other aberrations (e.g., multipolar spindles) were scored in metaphase cells; fragments, lagging chromosomes, and dicentric bridges were scored in anaphase cells. The results are presented in Table 1.

The cumulative frequency of normal and abnormal cells is less than 100., because cells that could not be clearly identified as normal or abnormal were included in the sample, but not assigned to either category. The difference between the total frequency and 100% is the frequency of these "doubtful" cases. Gross and fine structure of aberrations


Acentric fragments and dicentric bridges.--Figures 99-101

illustrate one of the most common types of aberration in anaphase cells, viz., a single dicentric bridge and fragments. These bridges, formed by a subterminal exchange between sister chromatids, are more or less attenuated, probably depending on the degree of spindle elongation. In both cells shown the kinetochores of the dicentric chromosome were perfectly normal (e.g., Figure 100). There was nothing unusual about the fibrous structure of the bridged chromosomes. The fragments in these cells were located in the cytoplasm at the periphery of the spindle (insets in Figures 99 and 101). Analysis of serial sections of several such cells revealed the truely acentric nature of the fragments.

Dicentric bridges, single or multiple, also occur in telophase cells. In this case the intact bridges cross the 'region of the'

























Table l.--Frequency of chromosomal and mitotic aberrations
in untreated meta- and anaphase cells
(streptonigrin control).


Miti No. of No. of Total No. Percent
Mitotic Normal Abnormal of Cells Normal Abnomal Stage Cells Cells Examined Cells Cells

Metaphase 90 23 135 66.7 17.0 Anaphase 56 5 66 84.9 7.6


Total 146 28 201 72.7 13.9









equatorial constriction. Two cells in cytokinesis were found with "nuclear" bridges connecting the daughter nuclei through the compact midbody. One of these cells was examined in the electron microscope. Nuclear pore complexes were present on the NE wrapping the bridges.


Lagging chromosomes.--The second most common type of aberration in anaphase cells is laggards. Only a few such chromosomes could be examined in serial sections. Representative profiles of kinetochores of two laggards are shown in Figure 102. These chromosomes were lying near or at the periphery of the spindle (inset in Figure 102a). Characteristically, a bundle of arched M14T was present at the kinetochores, which were oriented towards the center of the spindle. Some of these MT apparently bypassed the chromosome, while others seemed to terminate in the fuzzy kinetochores (Figure 102). The bundle of lIT associated with laggard no. 1 (Figure 102a and inset) also seemed to be associated with laggard no. 2 (inset in Figure 102a) in the opposite half-spindle. Numerous non-kinetochore MT near the laggards were oriented haphazardly. Kinetochores and MT of most of the chromosomes near the poles were normal. The chromosome in Figure 103 was an exception. The strange profile of its kinetochore in some of the serial sections may be due to a peculiar angle of sectioning. Some of the kinetochore MTT, particularly in peripheral sections of the kinetochore (Figure 103b), were definitely odd.

In a similar cell one kinetochore of a laggard was very stretched and similar in structure to the one in Figure 102a.









Centriole aberrations.--The three major possible aberrations involving centrioles abnormal number, abnormal structure, and abnormal position are illustrated in Figures 104-107. The presence of four centrioles in interphase (Figure 104) presumably leads to multipolar mitosis. The prophase cell in Figure 105a was very similar to the one in Figure lla, as far as chromosome condensation and kinetochore differentiation are concerned. However, the two pairs of centrioles were positioned on opposite sides of the nucleus (Figure 105a). Despite this obviously axial arrangement, very few MT were associated with each centriole pair (Figures 105b and 105c).

Figure 106 shows two serial sections of the pole no. 2 centrioles of the cell in Figure 29. Portions of three centrioles are visible. The one in the center was deformed to a cup-like structure. Because of missing sections, the precise architecture of, and relationship between, these centrioles could not be reconstructed. Microtubules converged on the osmiophilic masses left and right of center in Figure 106a.

Centriole aberrations also occurred in the anaphase cell of Figure 102. One of a pair of centrioles near pole no. 2 is shown in Figure 107a. The MT, however, converged on a different center (P2), where amorphous, osmiophilic material was present. I am convinced that there was at least a third centriole at this spot, but I could not verify it, because some of the serial sections were missing. On the other hand, one serial section of the centriole shown in Figure 107a revealed a very osmiophilic particle in the lumen of the centriole (Figure 107b). In size and shape this particle was very similar to those found in the vicinity of centrioles at all stages of the cell cycle (e.g., Figure 4). At pole no. 1 of this anaphase cell there were probably also more than two centrioles (Figure 107c).









Streptonigrin-Treated Cells


Mitotic index and frequency of aberrations

The mitotic index was determined from cell counts made with the

phase contrast microscope at a magnification of 200 or higher. Mitotic cells comprised all the stages from pro- to telophase. The control cells were the same used to compute the frequency of aberrations (see Table 1). The results are presented in Table 2.

The procedure for scoring aberrations was as described for untreated cells. The results are presented in Table 3.

The figures for meta- and anaphase cells are of relative value only, because the distinction between these stages is not sharp in highly aberrant cells. As for untreated cells, the cumulative frequency for 0.01 pg/ml SN in less than 1007., due to "doubtful" cases.


Gross and fine structure of aberrations

The cytological effect produced by the two concentrations of SN differs not only quantitatively, but also qualitatively. Generally speakingp the higher concentration induces more complex aberrations and bizarre mitotic figures are common (Figure 108). Ana- and telophase cells usually contain several dicentric bridges of variable diameter, probably depending on the degree of attenuation (Figures 108 c-f). Numerous acentric fragments are located at the periphery of these cells.

Chromatin strands connecting the kinetochore regions of daughter chromosomes occurred in cells in very early anaphase (Figure 109). The kinetochores of these chromosomes appeared normal. An apparent exception is illustrated in Figure 109. This kinetochore resembled certain prometaphase kinetochores (see Figures 35 and 41). Its associated MT passed over the adjacent chromosome.




























Table 2.--Mitotic indices (MI) for streptonigrintreated and untreated control cells.



Treatment Total No. of Cells in Mitosis
Cells Counted no. 7 .MI Control 541 14 2.59 0.01 ug/ml SN 540 12 2.22 0.05 ug/ml SN 541 1 0.18




Full Text
217
Fig. U2a, Id. Streptonigrin-treated cells in late anaphase. Re
construction of the HE. (a) Portion of a polar group of chromosomes.
Note close relationship of membrane cisternae (large arrows) and RER
(small arrows) with chromatin. From the cell in Fig. 108d. x 30,000.
(*) Portions of two chromosomes near a pole. Note membrane cisternae
apposed to chromosome (large arrows), RER in proximity to chromosomes
(small arrows). From the cell in Fig. 108f. x 30,000.


Fig. 96. Kinetochores in c-mitosis. Kq_ obliquely, Kp transversely
sectioned, K grazed. From the same cell as Figs. 92 and 93* x 50,000.
Treatment A.
Fig. 9T Kinetochore in c-mitosis. Approximately median section.
Note doubleness (white arrows), "knots"(black arrows). From the cell
shown in Fig. 66. x 100,000.
Fig. 98. Kinetochores in c-mitosis. Sister kinetochores (Kq_,
Kg) in primary constriction. Note doubleness of Kg. From the
cell shown in Fig. 66. x 67,500.


LIST OF FIGURES
Figure Page
1 Karyotype of the male rat kangaroo (Potorous 45
tridactylis)
2 The three planes in which blocks were 51
sectioned
3 Grazing section of an interphase nucleus ... 87
4a-d Centriole of interphase cell ......... 89
5a-c Transverse sections of interphase nuclei ... 91
6 Chromatin fibers of interphase nucleus .... 93
7 Nucleolus of interphase cell 93
8a Very early prophase cell 94
Sb-e Centriole duplication in very early 96
prophase .
9a,b Migration of centrioles in very early 97
prophase ... ......
9c-e Fine structure of centriole in very early 97
prophase ..
10 Mid-prophase nucleus 98
11a Mid-prophase nucleus 99
lib Mid-prophase nucleus, peripheral section 100
12 Early prophase. Grazing section of the 102
nucleus ........
13 Early prophase. Transverse section of the 102
nucleus
14 Late prophase chromosome near nuclear 104
envelope ........
viii


12
fibrous organization in some cells (Bajer and Allen 1966). The most
convincing evidence concerning the reality of spindle fibers has come
from polarization microscope studies (Inou 1964, Inou and Sato 1967).
Early electron microscopic observations by Harris (1962) and Roth
and Daniels (1962) revealed numerous ITT in the mitotic spindle of sea
urchin eggs and amebae, respectively. That the mitotic spindle is a
collection of prominent MT has since been substantiated by numerous
other workers (e.g., Aldrich 1969, Krishan and Buck 1965, Robbins and
Gonatas 1964, Roth et al. 1966). The distribution of spindle MT agrees
well with that of spindle fibers as seen in the light microscope (see
Luykx 1970, Table III). Further support for the contention that the
spindle fibers of the light microscopists consist of bundles of MT came
from birefringence and electron microscopic studies on the effect of
colchicine, colcemid, vinblastine, cold, and ultraviolet light (UV) on
spindle MT. All these agents cause loss of spindle birefringence and a
reduction of the number of MT as seen in the electron microscope (Bajer
1969, Inou 1952, Inou and Sato 1967, Malawista et al. 1968, Roth
1967). On the other hand, treatment of mitotic cells with heavy water
increases birefringence and the number of MT in the spindle (Inou and
Sato 1967). The observation by Rebhun and Sander (1967), that MT are
not the only birefringent component of the mitotic spindle, imposes
limits on the interpretation of such results, but the fact remains that
MT are birefringent elements. The correlation between birefringence
and MT is generally regarded as sufficient evidence for the occurrence
of MT in living cells, particularly since improved fixatives for
electron microscopy have made the demonstration of MT easy (but see
Nicklas 1971 for a discussion of this point).


Fig. 12. Early prophase. Grazing section of the nucleus. Note
intact nuclear envelope (NE) with pores (NP), chronosorr.es (Ch) with
large achromatic holes (h). x 30,000.
Fig. 13* Early prophase. Transverse (serial) section of the
nucleus shown in Fig. 12. Note chromosomes (Ch) apparently attached
to nuclear envelope (NE) hy "stalks" (arrows), x 22,500
i


178
*
T*5
%s'Tt


157


Fig. 89. Para-equatorial section of metaphase cell. Note
intrachromosoraal MT within clear circle (arrow). From the sane
cell as Fig. 87. x 62,500
Fig. 90. Para-equatorial section of metaphase cell. Bundle
of Icinetochore MT. Note "cross-bridges" (large arrows) and "arms"
(small arrows). From the same cell as Fig. 87. x 75,000.
Fig. 91b., b. Tiro serial sections of mid-anaphase kinetochore
(para-equatorial). From the cell in Fig. 50. x 75,000.


134


Fig. 80. Kinetochore in late prometa- to metaphase. Note
30-50 A fibrils (arrows) in outer layer (KO) of kinetochore, its
lesser electron density compared to the chromosome (Ch). x 100,000.
Fig. 8l. Metaphase kinetochore. A straight MT (arrow) ends
in the outer layer (KO); an obliquely sectioned MT (arrowhead)
seems to penetrate outer and middle layers, x 62,500.
Fig. 82. Kinetochore in very early anaphase. MT marked by
arrow ends in the outer layer (KO). From the cell in Fig. J45.
x 50,000.


167
Fig. 66. C-mitosis. Several chromosomes with kinetochores
(arrowheads) in the central area. A centriole (c) near the cytoplasmic
area, x 11,500 Inset: Phase contrast micrograph of the cell in
plastic, (x l,28o). Treatment B.


Figure Page
104 Untreated interphase cell, four centrioles 207
105a-c Untreated mid-prophase cell, abnormal
centrioles 207
106a,b Untreated cell in late prometaphase, abnormal
centrioles 209
107a-c Untreated anaphase cell, abnormal
centrioles 210
108a-f Streptonigrin-induced aberrations. Phase con
trast micrographs 212
109 Streptonigrin-treated cell in early
anaphase 214
110a,b Streptonigrin-treated cell in late
anaphase 214
llla-c Streptonigrin-treated cell in late
anaphase 216
112a,b Streptonigrin-treated cells in late
anaphase 217
113 Streptonigrin-treated cell in late
anaphase 219
114 Streptonigrin-treated cell in late
telophase 219
115 Streptonigrin-treated cell in late
telophase ............ 219
116 Streptonigrin-treated cell in late
telophase ............ 220
117a,b Streptonigrin-treated cell in late
telophase 222
118 Streptonigrin-treated cell in late
cytokinesis 222
119a,b Streptonigrin-treated cell in late
cytokinesis 224
120 Diagrammatic representation of kinetochore
maturation during prometaphase 243
121 Diagrammatic representation of the three-
dimensional structure of kinetochores ... 248
xiii


164
Fig. 63. C-metaphase. The chromosomes are scattered. Three of
the four centrioles present in this cell can be seen (C). A kinetochore
is opposite the centrioles (arrowhead). Mitochondria (Mi), vesicles (v),
and cisternae (Ci) occur in the central area and at the periphery of
the cell. Black marks are staining artifacts, x 15,750 Inset:
Phase contrast micrograph of the cell in plastic (x l,28o). Treat
ment A.


241
therefore finds immature prophase kinetochores on the chromosomes in
the center of the "nucleus," and more mature kinetochores near the two
poles. 1 could not establish a clear-cut correlation between attach
ment of kinetochores to MT and the appearance of the three bands
typical for mature kinetochores. On the contrary, kinetochores not
attached to MT often exhibit more distinct bands than attached
kinetochores (e.g., Figure 23). However, it is possible that attached
kinetochores are triple-banded, but since they are drawn out into a
cone or more odd structure (Figures 23, 24, and 36) they would
necessarily be cut obliquely by any but ideal median sections and con
sequently the bands would be less distinct. In fact, Figure 23
indicates that these kinetochores are banded. Much of the foregoing
also applies to structural differences between sister kinetochores of
maloriented chromosomes in late prometaphase (Figures 28, 37, and 38).
Figure 120, illustrating asynchronous maturation during prometa
phase of sister kinetochores of a maloriented chromosome, is based on
the above mentioned assumption and observations. Kinetochore no. 1
possibly had a structure similar to that of kinetochore no. 2 in Figure
120b at a stage preceding that shoi/n in Figure 120a. The transition of
kinetochore no. 2 from Figure 120a to Figure 120b is hypothetical.
This kinetochore matured later than its sister oriented towards the
near pole and attached to MT very early. A similar sequence can be
postulated for equatorial chromosomes with one unobstructed and one
obstructed kinetochore. If both kinetochores of an equatorial chromo
some are unobstructed they mature synchronously (e.g., Figure 25).
It appears that the outer layer is formed within the finely
fibrillar material of prophase and immature prometaphase kinetochores


61
kinetochores are quite often stretched (Figures 27, 28, 31, 33, and 36),
and they are always attached to bundles of 11T.
Obstructed kinetochores of equatorial chromosomes appear as patches
of finely fibrillar matrix within which a denser band or patch can be
seen (Figures 33, 35, and 41). Serial sections make it clear that the
bands do not represent fibers, but sections of oddly shaped three-
dimensional structures (e.g., Figures 35 and 41). The width of the
bands varies from 250-400 A. In appearance and dimensions they are
comparable to the outer layer of triple-banded kinetochores, except
that they often exhibit a partly double-banded, partly beaded sub
structure (Figure 41). Furthermore, obstructed kinetochores are always
attached to very few, if any, MT.
Sister kinetochores of maloriented chromosomes most commonly are
also dissimilar. As a rule, the kinetochore facing the near pole
resembles the unobstructed kinetochores of equatorial chromosomes
(Figures 27, 37, and 38), while the kinetochore facing the far pole
resembles obstructed kinetochores (Figures 27, 37, and 38). Figure 40
illustrates a rare exception to this rule. On both these maloriented
chromosomes the kinetochore oriented towards the near pole is compact,
and attached to 1IT, while the other kinetochore is fuzzy and stretched,
but also attached to IIT.
Figure 39 illustrates another rare case. The sister kinetochores
of this maloriented chromosome are very similar in every respect; both
are relatively undifferentiated masses of matrix with only a trace of a
band, 'lost remarkable is that the matrix seems to cover almost the
entire small side of the chromosome (Figure 39c).


42
Jagiello (1967) described chromosomal aberrations induced by SN
O'. 3 o' a o .o
in mouse eggs. _In vitro, metaphase I chromosomes of ova treated with
1.0 ig/ml SN were agglutinated beyond recognition; 0.1 ;jg/ral SN induced
achromatic gaps and breaks in approximately 407. of the ova. Severe
chromosome damage also occurred in ova of mice, to which SN had been
administered subcutaneously.
Streptonigrin is still used in _in vitro and clinical studies on
tumor chemotherapy (Carter et al. 1968, Oleson et al. 1961). Like many
other antitumor drugs it is more active against lymphoma than against
solid tumors.
Colcemid; A Spindle Poison
Colchicine is an alkaloid isolated from Colchicum autumnale, whose
anti-mitotic effect has been under investigation for over 30 years (see
reviews by Biesele 1958, Deysson 1968, Dustin 1963, Eigsti and Dustin
1955, Gelfant 1963, Kihlman 1966). The structural formulas of
colchicine (from Kihlman 1966) and its synthetic analog colcemid
(demecolcine, N-deacetyl-N-methylcolchicine; from Schar et al. 1954)
are given below.
Colchicine
Colcemid


232
granules of approximately the same size in isolated MA of sea urchins
stored for longer periods of time.
Considering these observations one can hypothesize that the
substance coating MT in cold-treated Ptl^ cells is microtubular material
in a dispersed state, but still retaining affinity for persisting MT.
The latter could serve as nucleating centers for rapid polymerization
during recovery from cold. The necessity of nucleating centers for
polymerization of free subunits has been demonstrated by Stephens (1969)
with flagellar outer fibers, but a similar experiment with spindle MT
has not as yet been reported. But it is clear that repolymerization of
MT, as judged by reappearance of birefringence, occurs within seconds
or minutes after exposure to cold, depending on the duration of the
treatment (Inou 1964, Inou et al. 1970).
Brinkley et al. (1967) found persisting MT in Chinese hamster cells
arrested by exposure to 0.06 pg/ml colcemid. These MT were of the
kinetochore type; continuous MT were absent. A normal metaphase
configuration was formed after 15-20 min recovery. In contrast to
this, I found no MT in PtK^ cells after 1 hr recovery following a 2 hr
exposure to 0.05 pg/nl colcemid (e.g., Figures 63 and 64). This
indicates a distinct difference in sensitivity.
The microfibrils observed in PtK^ cells exposed to 0.25 pg/ml
colcemid may be significant in the context of MT depolymerization. In
my opinion these fibrils are not identical with microfibrils that are
numerous mainly near the growth surface in untreated cells. Similar
microfibrils have been reported in cells treated with vincristine by
Journey et al. (1968). Nathaniel et al. (1968) observed 35-50 A fibrils
in great quantity in melanocytes of Harding-Passey tumor treated with


Fig. 92. Kinetochores in c-mitosis. Two chromosomes (Ch-j_,
Chg) in longitudinal section, K2, the sister kinetochores
of Chi* x 50,000. Treatment A.
Fig. 93* Kinetochore in c-mitosis. Grazing section of
kinetochore (k), which is seen in face-view in the depression
of the primary constriction. Portions of the two arms of the
chromatid (Chd). x 50,000. Treatment A.
Fig. 9b. Kinetochores in c-mitosis. Sister kinetochores
(K¡_, K2) of transversely sectioned chromosome (Ch). Note two
centrioles (C). x 50,000. Treatment A.


264
De Harven, E. 1968. The centriole and the mitotic spindle, p. 197-22?.
In A. J. Dalton and F. Haguenau (eds.) The Nucleus. Academic Press,
New York.
De Harven, E., and W. Bernhard. 1956. Etude au microscope electronique
de 1'ultrastructure du centriole chez les vrtebras. Z. Zellforsch.
Mikroskop. Anat. 45: 378-393.
Deysson, G. 1968. Antimitotic substances. Int. Rev. Cytol. 24: 991^8.
Dietz, R. 1959. Centrosomenfreie Spindelpole in Tipuliden-Spermatocyten.
Z. Naturforsch. lib: 749-752.
Dietz, R. 1966. The dispensability of the centrioles in the spermatocyte
divisions of Pales ferruginea (Nematocera), p. l6l-l66. In C. D.
Darlington and K. R. Lewis (eds.) Chromosomes Today, vol. 1.
Plenum Press, New York.
Dietz, R. 1989. Bau und Funktion des Spindelapparats. Naturwissenschaften
56: 237-248.
Dirksen, E. R. 1961. The presence of centrioles in artificially
activated sea urchin eggs. J. Biophys. Biochem. Cytol. 11: 244-247.
Dirksen, E. R., and T. T. Crocker. 1966. Centriole replication in
differentiating ciliated cells of mammalian respiratory epithelium.
An electron microscopic study. J. Microscopie 5- 629-644.
DuPraw, E. J. 1965. Macromole cular organization of nuclei and
chromosomes: A folded-fibre model based on whole-mount electron
microscopy. Nature 206: 338-3^3
DuPraw, E. J. 1968. Cell and Molecular Biology. Academic Press,
New York. 739 P
Dustin, P., Jr. 1963. New aspects of the pharmacology of antimitotic
agents. Pharmacol. Rev. 15: 449-480.
Eigsti, 0. J., and P. Dustin, Jr. 1955* Colchicine in Agriculture
Medicine, Biology, and Chemistry. Iowa State College Press,
Ames. 47 p.
Erlandson, R. A., and E. de Harven. 1971. The ultrastructure of
synchronized HeLa cells. J. Cell Sci. 8: 353397.
Evans, H. J. 1962. Chromosome aberrations induced by ionizing
radiations. Int. Rev. Cytol. 13: 221-321.
Evans, H. J. 1963. Chromosome aberrations and target theory, p. 8-40.
In S. Wolff (ed.) Radiation-Induced Chromosome Aberrations. Columbia
Univ. Press, New York.
Evans, H. J., and J. R. K. Savage. 1963* The relation between DNA
synthesis and chromosome structure as resolved by X-ray damage.
J. Cell Biol. 18: 525-540.


76
I found no intertubular connections and arms on MT in sections
passing close to the kinetochores. Figure 90 shows what presumably is
a bundle of kinetochore MT at a greater distance from the chromosome.
Two apparent cross-bridges and two arms can be seen.
Anaphase kinetochores in para-equatorial sections are very
similar to metaphase kinetochores, except that most of them are more
convex, i.e., MT at the periphery of the kinetochores terminate one or
two sections after the apical MT have disappeared (Figure 91).
Kinetochore profiles in colcemid-treated cells are quite differ
ent from the triple-banded structures in untreated cells (Figures 92-
98). Figure 92 represents an approximately median longitudinal
section of two chromosomes. A longitudinal section along line C-D and
perpendicular to that of Figure 92 would produce a face-on view of the
kinetochore as in Figure 93. A transverse section along line A-B,
also perpendicular to that of Figure 92, would produce an image as in
Figure 94. The kinetochore bands in sections such as in Figures 92
and 94 are approximately 400 A wide, are embedded in a less opaque,
fibrillar matrix, and closely follow the surface of the chromosome.
The bands are slightly less electron dense than the chromosomes
(Figure 92), but again this characteristic is obscured in contrasty
prints (Figures 94, 97, and 98). I consider these bands equivalent to
the outer layer of kinetochores in untreated cells. There is no inner
layer on any of the chromosomes in colcemid-treated cells, and the
kinetochores are never attached to MT. The outer layer is apparently
made of two 150 A sheets held together at various points. This
accounts for the transverse and longitudinal sections showing two


28
Mammalian kinetochores are by far the most intensely studied.
Besides papers devoted specifically to mitosis or kinetochores, there
are numerous isolated electron micrographs in the literature showing
various profiles of mammalian kinetochores (e.g., Flaks 1971, Hu 1971).
Unfortunately, some of the most frequently cited papers on the ultra
structure of mammalian cells in mitosis (e.g., Krishan and Buck 1965,
Murray et al. 1965) do not show kinetochores at high magnifications.
At best, one can make out that they consist of three layers or bands,
like those described by Nebel and Coulon (1962). Robbins and Gonatas
(1964, Figure 19) presented a detailed picture of an early anaphase
kinetochore in a HeLa cell. The three layers are very clear; MT seem
to insert into the outer layer. Bamicot and Huxley (1965) also
published electron micrographs showing the kinetochores of cultured
newt heart cells as three-layered structures. They interpreted the
kinetochores to consist of material different from the chromosomes,
based on different stainability. In this context it is interesting to
note that the kinetochores of the alga Oedogonium (Pickett-Heaps and
Fowke 1969) and the moss Mnium (Lambert 1970) are also triple-layered,
as are those of Barbulanympha (Hollande and Valentin 1968), which has
an intranuclear spindle.
Jokelainen (1965a) described kinetochores in fetal rat kidney as
short bands separated from the chromosomes by a clear area, or as two
parallel bands, in which case the second band was in direct contact
with the chromosomal material. In subsequent papers (Jokelainen 1965b,
1967, 1968) he developed his concept of kinetochore maturation and his
kinetochore model. Maturation occurs during prometaphase and is
asynchronous for sister kinetochores. It involves the appearance of


98
Fig. 10. Mid-prophase. Note dispersed nucleolus (Nu),
chromosomes (Ch), kinetochore (k), intact nuclear envelope (NE).
x 11,500 Inset: Phase contrast micrograph of the cell in plastic,
(x l,28o).


Fig. 104. Four centrioles in an untreated interphase cell,
x
Fig. 105a-c. Abnormal centrioles in an untreated cell in
mid-prophase, (a) Centriole pairs (C-^, C2) on opposite sides
of the nucleus, x 7,750. (b) Higher magnification of the
centriole seen at C]_ in (a). Note paucity of MT. x 30,000.
(c) Serial section of centrioles at C2 in (a). Note paucity
of MT. x 30,000.


32
1969, Koopraans 1958) or as part of normal spindle development (Hughes-
Schrader 1948), seems to indicate that chromosomes alone are capable of
organizing spindle MT. Similarly, Roth (1967) suggested that in
spindle reformation after cold shock the spindle MT gradually extend
from the chromosomes towards the poles. These observations have been
interpreted to mean that spindle fiber material is being continuously
assembled and oriented at the kinetochore throughout prometaphase and
metaphase, but the exact role of the kinetochore is not as yet clear
(see Luykx 1970).
Equally interesting in this context are ideas of the kinetochore
as a specialized "gene" (Brinkley and Stubblefield 1970) or a gene
product (Luykx 1970). Even if we accept the universal occurrence of
DMA in the centromere region as a fact, we must bear in mind that the
relationship between DMA and the kinetochore at the fine structural
level is not at all resolved. Brinkley and Stubblefield (1970) have
proposed that the lateral loops of their lampbrush-like kinetochores
consist of DMA, which codes for long ENA molecules that bind with
protein subunits to form MT. This elaborate hypothesis stands on
rather shaky ground. Luykx (1970),on the other hand, proposed the DMA
in the kinetochore region "may therefore be viewed as a 'kinetochore
organizer', similar to the nucleolar organizing region of the
chromosome in a number of ways. It is responsible for the synthesis or
assembly of an essential organelle that remains associated with a
specific chromosomal site, is often associated with blocks of
heterochromatin, remains relatively uncoiled during mitosis, and
probably contains a large number of identical genes" (Luykx 1970).


127


average, 26 bundled MT are anchored in the outer layer. Staining
3 O i. *3 .9
characteristics and behavior in late mitosis suggest that the inner
layer is chromatin, the outer layer possibly protein.
Daughter chromosomes are connected by strands of chromatin during
very early anaphase. The strands rupture during early anaphase. In
late anaphase, when chromosome movement has ceased, the K layers are
fuzzy, and fewer, less organized MT are attached. In telophase the
inner layer is a very opaque patch of dense material on the inner
membrane of the NE; the outer layer is lost in the cytoplasm.
Reconstruction of the NE, cleavage, midbody formation, and
reappearance of the nucleolus, are similar to these processes in other
mammalian cells.
Mitotic stages in cells exposed to 0-4C for 1 hr are similar to
untreated cells. Prophase Ks are identical and mature metaphase Ks are
typically triple-layered, though slightly fuzzy. Few MT, almost
exclusively kinetochore MT, are preserved. They are coated by an
amorphous or finely fibrillar substance.
No spindle MT are present in colcemid-arrested cells (0.05 pg/ml
for 2 hr, 1 hr recovery; 0.25 pg/ml for 15 min, no recovery). Chromo
somes are scattered in the central area of the cells. Centrioles have
duplicated, but not migrated. The higher concentration of colcemid
induced decondensation of chromosomes and dispersion of chromatin in all
interphase cells. Kinetochores resemble immature prometaphase Ks in
control cells. The inner layer is lacking. The outer layer is a convex
or undulated, bilaminar plate, 400 A thick and 5,000-9,000 A in diameter,
and embedded in a finely fibrillar matrix.


37
interesting point is that in Vicia and Trillium Feulgen-negative
regions similar in appearance to gaps can be induced by exposure to low
temperature. Electron microscopic examination of whole-mounted and
thin-sectioned chromosomes indicates that some gaps are true breaks,
while others are traversed by chromosomal fibrils (Brinkley and Shaw
1970). Similar results were obtained by Scheid and Traut (1971) with
the scanning electron microscope. They found that gaps represent
distinct "notches" which in some cases are traversed by two parallel
strands.
Acentric fragments
A break without reunion produces an acentric fragment (Evans 1962,
Kaufraann 1954, Kihlman 1966). If the break involves only one chromatid
the fragment is single; it consists of two "sister chromatids" if the
entire chromosome is broken. Such fragments are usually lost during
division, or they are included at random in daughter cells where they
form micronuclei (Humphrey and Brinkley 1969, Kihlman 1966, La Cour
1953). Carlson (1938) attributed apparent migration of acentric
fragments towards the poles at anaphase to more than chance movement,
but he made his observations on smear preparations. Bajer (1958) and
Bajer and Mole-Bajer (1963) recorded the behavior of fragments in
irradiated Haemanthus endosperm by time-lapse cinemicrography. The
majority of fragments were eliminated from the spindle, either at
prometaphase or during ana- and telophase. A few fragments, however,
moved in the spindle region in a more than random fashion, sometimes
from the equator to one pole and back. The authors attributed these
movements to the activity of neocentric fibers. In rat kangaroo cells


120


192


82
Streptonigrin-Treated Cells
Mitotic index and frequency of aberrations
The mitotic index was determined from cell counts made with the
phase contrast microscope at a magnification of 200 or higher. Mitotic
cells comprised all the stages from pro- to telophase. The control
cells were the same used to compute the frequency of aberrations (see
Table 1). The results are presented in Table 2.
The procedure for scoring aberrations was as described for un
treated cells. The results are presented in Table 3.
The figures for meta- and anaphase cells are of relative value
only, because the distinction between these stages is not sharp in
highly aberrant cells. As for untreated cells, the cumulative frequency
for 0.01 pg/nl SN in less than 1007., due to "doubtful" cases.
Gross and fine structure of aberrations
The cytological effect produced by the two concentrations of SN
differs not only quantitatively, but also qualitatively. Generally
speaking, the higher concentration induces more complex aberrations and
bizarre mitotic figures are common (Figure 108). Ana- and telophase
cells usually contain several dicentric bridges of variable diameter,
probably depending on the degree of attenuation (Figures 108 c-f).
Numerous acentric fragments are located at the periphery of these cells.
Chromatin strands connecting the kinetochore regions of daughter
chromosomes occurred in cells in very early anaphase (Figure 109). The
kinetochores of these chromosomes appeared normal. An apparent
exception is illustrated in Figure 109. This kinetochore resembled
certain prometaphase kinetochores (see Figures 35 and 41). Its
associated MT passed over the adjacent chromosome.


272
Ris, H. 1955. Cell division, p. 91-125* In B. H. Willier, P. A.
Weiss, and V. Hamburger (eds.) Analysis of Development.
Saunders, Philadelphia.
Ris, H. 1961. The annual invitation lecture. Ultrastructure and
molecular organization of genetic systems. Can. J. Genet. Cytol.
3: 95-120.
Ris, H. 19^7* Ultrastructure of the animal chromosome, p. 11-21.
In V. V. Koningsberger and L. Bosch (eds.) Regulation of Nucleic
Acid and Protein Biosynthesis. Elsevier, New York.
Ris, H., and D. F. Kubai. 1970* Chromosome structure. Annu. Rev.
Genet, 4: 263-294.
Robbins, E., and N. K. Gonatas. 1964. The ultrastructure of a mammalian
cell during the mitotic cycle. J. Cell Biol. 21: 429-463.
Robbins, E., G. Jentzsch, and A. Micali. 1968. The centriole cycle
in synchronized HeLa cells. J. Cell Biol. 36: 329339*
Roth, L. E. 1964. Motile systems with continuous filaments, p. 527-
548. In R. D. Allen and N. Kamiya (eds.) Primitive Motile Systems
in Cell Biology. Academic Press, New York.
Roth, L. E. 1967* Electron microscopy of mitosis in amebae. III.
Cold and urea treatments: A basis for tests of direct effects
of mitotic inhibitors on microtubule formation. J. Cell Biol.
34: 47-60.
Roth, L. E., and E. W. Daniels. 1962. Electron microscopic studies of
mitosis in amebae. II. The giant ameba Pelomyxa carolinensis. J.
Cell Biol. 12: 57-78.
Roth, L. E., H. J. Wilson, and J. Chakraborty. 1966. Anaphase structure
in mitotic cells typified by spindle elongation. J. Ultrastruct.
Res. l4: 460-483.
Sauaia, H., and D. Mazia. 1961. As cited by Luykx, 1970.
Schr, B., P. Loustalot, and F. Gross. 1954. Demecolcin (Substanz F),
ein neues, aus Colchicum autumnale isoliertes Alkaloid mit starker
antimitotischer Wirkung. Klin. Wochenschr. 32: 49-57.
Scheid, W., and H. Traut. 1970. Ultraviolet-microscopical studies on
achromatic lesions ("gaps") induced by X-rays in the chromosomes
of Vicia faba. Mutation Res. 10: 159-161.
Scheid, W., and H. Traut. 1971* Visualization by scanning electron
microscopy of achromatic lesions ("gaps") induced by X-rays in
chromosomes of Vicia faba. Mutation Res. 11: 253-255.
Schrader, F. 1936. The kinetochore or spindle fibre locus in Amphiuma
tridactylum. Biol. Bull. JO: 484-498.


151
Fig. 50. Mid-anaphase. Para-equatorial section of the cell shown
in the inset. Note two circular Itinetochores (k), penetrating MT
(circles), chromosomal granules (arrows). 15,750* Inset: Phase
contrast micrograph of the cell in plastic (x 1,28o).


198


130
Fig. 33b. Late prometaphase. Serial section of the three
chromosomes shown in Fig. 33a x k0,000.


Fig. 58. Early telophase. Otliquely sectioned nuclear lohe
(NL) with pore-annulus complexes (IIP). Transversely sectioned
pieces of KE indicated by large arrows. Serial section of the
cell shown in Fig. 57. x57,500.
Fig. 59- Mid-telophase. Two kinetochores (k) in pockets on
the polar face of the nucleus. Note MT (arrows) associated with
the kinetochores. Serial section of the cell shown in Fig. 60.
x 50>000*


38
exposed to X-rays, Humphrey and Brinkley (1969) confirmed by electron
microscopic analysis of thin sections that fragments lack kinetochores.
Bridges
An exchange between centric portions of two broken chromatids or
chromosomes results in the formation of an anaphase bridge if the two
centromeres move to opposite poles (Evans 1962, Hair 1953, Kaufmann
1954, Kihlman 1966, Koller 1953). Subchromatid bridges (side-arm
bridges) arise from intrachromosomal exchanges. The unit of breakage
and exchange in these aberrations was generally assumed to be a half
chromatid (Heddle 1969, Kihlman 1966). Brinkley and Humphrey (1969)
examined X-ray-induced side-arm bridges in rat kangaroo cells with the
electron microscope and found that the diameter of these chromatid
connections was considerably less than that of a half-chromatid. The
authors conceded, however, that chromosome movement during anaphase
might have stretched the connections, which appeared to consist of
chromosomal fibers of the usual dimensions (see also Brinkley and Shaw
1970).
Anaphase bridges exist as direct connections between two
centromeres, or as interlocked dicentrics (Koller 1953). The thickness
and length of these bridges vary; they usually break during ana- or
telophase, but may persist into interphase (Hair 1953, Koller 1953).
Cinemicrographic studies by Bajer (1963, 1964) on the behavior of
dicentrics in Haeraanthus endosperm revealed interesting facts:
normally, dicentrics break abruptly, but in slightly unhealthy cells
they do not break, but form long, thin, sticky bridges. Interlocked
dicentrics show two kinds of behavior: they cut one through the other,


Fig. 3l¡a-c. Late prometaphase. Three serial sections of
an equatorial chromosome. From the same cell as Fig. 39. Note
triple-handed kinetochores (Kj_, IGj), transchromosomal MT (arrows
x 40,000.


86
layer of kinetochores (Figures 114-116). Numerous MT were observed
penetrating from the pockets into the nuclei, indicating the envelope
was not completely reconstructed. Pieces of double membrane, some
bearing ribosomes (Figure 117a), were also found in the nuclei.
Occasionally, ribosome-like particles appeared on both membranes of the
nuclear envelope (117b).
Cells in late cytokinesis, connected by nuclear bridges across the
midbody, can be identified with the light microscope. Surprisingly,
however, a greater number of daughter cells connected by extremely thin
bridges are found in thin sections. Many of these aberrations are of
truely ultrastructural dimensions. The greater portion of the bridge
shown in Figure 118 was only approximately 500 A in diameter. In this
case no midbody was found between the two daughter cell. The bridge
simply passed across the cleavage furrow into the other cell and
expanded to a large nuclear bleb, which was in turn connected to the
main nucleus by a similarly thin stalk. Both daughter cells also had
micronuclei. A similar case is illustrated in Figure 119. The thin
bridge passed through the compact midbody (Figure 119a). The drop
shaped nuclear bleb was connected to the main nucleus by a very thin
stalk. Micronuclei were also present in both daughter cells (Figure
119b).
Centriole aberrations also occur in SN-treated cells, but
apparently not more frequently than in untreated cells. In one of two
daughter cells in cytokinesis the two centrioles were found unusually
far from the nucleus and also remote from each other.


RESULTS AND OBSERVATIONS
Normal Mitosis
This section is based on light and electron microscopic observa
tions on untreated cells, either used as controls for the various
treatments, or fixed directly from the stock culture.
Interphase
Although interphase is not a mitotic stage, it makes sense to
describe the structure and spatial orientation in the interphase cell
of those components and organelles that take part in the formation and
function of the mitotic apparatus.
Figure 3 shows a grazing section of an interphase nucleus. Pore-
annulus complexes of the nuclear envelope are cut at different levels.
They appear as circles at the level of the envelope, and as circles
surrounded by a "halo" at the level of the underlying chromatin.
Central dense granules can be seen in some of the profiles. Where the
section passes below the complexes, their position is indicated by
achromatic holes in the chromatin. Whorls of polyribosomes are visible
on oblique sections of the nuclear envelope. In the cytoplasm, two
centrioles are present, approximately at a right angle to each other.
Obliquely sectioned membrane elements, probably of a Golgi complex, can
be seen in their vicinity. Microtubules, though more numerous around
the centrioles, pass in various directions, without apparent specific
orientation to the latter. Details of centriolar structure at
different levels are shown in the four serial sections in Figure 4.
52


175


67
restored, but large heterochromatic patches persist until interphase
(late stages of cytokinesis).
The less dense component of kinetochores, still present during
early telophase (Figure 57), has disappeared by the time the NE is
completely reconstructed (Figures 59 and 60). The inner, dense layer
becomes an extremely osmiophilic patch on the inner membrane of the NE
(Figures 59 and 60). Most commonly, the entire NE is indented at
these sites, and quite frequently the inner membrane with the dense
patch is deeply invaginated (Figure 60). Microtubules can be seen
extending poleward from these kinetochores, but the nature of this
association cannot be determined precisely.
During cytokinesis a cleavage furrow in the former equatorial
region separates the two daughter cells, except for a stem, whose
length and diameter vary depending on the distance between daughter
cells (Figures 61 and 62). The stem bodies seen in late anaphase and
early telophase have fused to form the midbody, an irregular band of
osmiophilic material in the stem. Numerous MT from both daughter cells
converge in the midbody, but the dense material obscures details in
these paraxial sections. Normally, the midbody remains between the
daughter cells even after these have completely separated and moved far
apart from each other. Eventually, it seems to be lost. On rare
occasions the midbody is included in the peripheral cytoplasm of one of
the daughter cells, but this is most likely an abnormal condition.
C-Mitosis
Treatment A (0.05 jig/ral colcemid for 2 hr 1 hr recovery)
produces an accumulation of cells with "scattered1' metaphases (Figures


210
Fig. 107a-c. Abnormal centrioles in an untreated anaphase cell,
(a) Pole no. 2 of the cell in Fig. 102. The centriole (c) shown is
one of a pair. Note MT converging near Po, not C. x 30,000. (b)
Serial section of the centriole shown in (a). Note particle in
centriole lumen, x 137,500- (c) Pole no. 1 of the same cell. Note
strange appearance of C2, possibly a third centriole (arrow),
x 22,500.


209


249
the inner layer or, possibly, even deeper in the chromosome. But his
sections were more than 750 A thick and the apparent deep penetration
of MT could well be ascribed to a superposition effect. In the
micrograph of an early anaphase kinetochore published by Robbins and
Gonatas (1964, Figure 19) the MT seem to terminate in the outer layer.
Intuitively one would expect a certain uniformity regarding MT
attachment in cells of different organisms, especially among mammals.
I believe that a careful and detailed reexamination of the cases
apparently contradicting my conclusions would reveal this uniformity.
The purpose of the colcemid treatments was to check the possi
bility that kinetochores of PtK^ cells under these conditions are
different from kinetochores in similarly treated Chinese hamster cells.
This was not the case (Figures 92, 94, and 96-98). The inner layer is
lacking and the outer layer structurally resembles that in c-mitotic
Chinese hamster cells, as well as cells treated with vincristine and
vinblastine (George et al. 1965, Journey et al. 1968, Journey and
Whaley 1970, Krishan 1968). Grazing sections of chromosomes yielded
images compatible with the idea that the outer layer is a circular or
oval plate in a depression of the chromatid (Figures 93, 96, and 121e).
Similar images were obtained by Journey (personal communication) from
vincristine-treated Chinese hamster cells.
Kinetochore profiles in transverse sections of chromosomes in
colcemid-treated PtK2 cells (Figures 92, 94, 97, and 98) suggest the
plate-like outer layer is bilaminar. This substructure is not
discernible in normal, mature kinetochores (e.g., Figures 82 and 84),
but it can be seen in immature prometaphase kinetochores (e.g., Figures
23c and 41). To me these observations indicate that immature
M


LIST OF TABLES
Table Page
1 Frequency of chromosomal and mitotic aberra- 79
tions in untreated meta- and anaphase
cells
2 Mitotic indices for streptonigrin-treated and 83
untreated control cells
3 Frequency of chromosomal and mitotic aberra- 84
tions in streptonigrin-treated cells ....


163
Fig. 62. Cytokinesis. Higher magnification of the stem and
midbody (MB) shown in Fig. 6l. x 30,000.


79
Table 1.Frequency of chromosomal and mitotic aberrations
in untreated meta- and anaphase cells
(streptonigrin control).
Mitotic
Stage
No. of
No. of
Total No.
Percent
Normal
Abnormal
of Cells
Normal
Abnormal
Cells
Cells
Examined
Cells
Cells
Metaphase
90
23
135
66.7
17.0
Anaphase
56
5
66
84.9
7.6
Total
146
28
201
72.7
13.9


180


155


BIOGRAPHICAL SKETCH
Urs-Peter Roos was born August 14, 1938, at Murten, Switzerland.
After completing secondary school in Dietikon, Switzerland, in 1953, he
began apprenticeship as an engraver and successfully passed the Swiss
Federal Examination in 1958. In 1960 he enrolled in the school of the
Akademikergemeinschaft in Zurich. He graduated in March, 1963, and
obtained the "Schweizerisches Maturitatszeugnis Typus C." He enrolled
in the Swiss Federal Institute of Technology (ETH) in Zurich in October,
1963, and graduated from this institution in August, 1967, with the
degree of Ingenieur Agronom. In September, 1967, he enrolled in the
Graduate School of the University of Florida. He worked as a graduate
assistant in the Department of Entomology and Hematology until December,
1968, when he received the degree of Master of Science with a major in
Entomology. In September, 1969, he transferred to the Department of
Zoology. He worked as a teaching and research assistant and was
recipient of a Graduate Fellowship of the College of Arts and Sciences
during the time he pursued his work toward the degree of Doctor of
Philosophy.
Urs-Peter Roos is married to the former Loan Hiang Ghan and is the
father of a son. He is a member of Phi Kappa Phi, Gamma Sigma Delta,
Phi Sigma, an associate member of Sigma Xi, and a student member of the
American Association for the Advancement of Science and the American
Society for Cell Biology.
276


Fig. 85a-c. Three serial sections of a metaphase kinetochore
(parasagittal). Note triple-layered profile, lesser density of
outer layer (arrows) compared to inner layer. From the cell in
Fig. 43. x 62,500.


102


186
4V
8!


273
Schrader, F. 1939* The structure of the kinetochore at meiosis.
Chromosoma 1: 230-237*
Schrader, F. 1953* Mitosis, 2nd ed. Columbia Univ. Press, Hew York.
170 p.
Schroeder, T. E. 1970. The contractile ring. I. Fine structure of
dividing mammalian (HeLa) cells and the effects of cytochalasin
B. Z. Zellforsch. Mikroskop. Anat. 109: 431-449.
Seaman, G. R. i960. Large-scale isolation of kinetosomes from the
ciliated protozoan Tetrahymena pyriformis. Exp. Cell Res. 21:
292-302.
Sentein, P. 1961. As cited by Biesele, 1962.
Sharp, L. W. 1934. As cited by Schrader, 1936.
Shaw, M. W. 1970. Human chromosome damage by chemical agents. Annu.
Rev. Med. 21: 1*09-432.
Shaw, M. W., and R. S. Krooth. 1964. The chromosomes of the Tasmanian
rat-kangaroo (Potorous tridactylis apicalis). Cytogenetics 3 19"33
Shimamura, T. 1940. Studies on the effect of the centrifugal force
upon nuclear division. Cytologia 11: 186-216.
Sisken, J. E., and E. Wilkes. 1967. The time of synthesis and the
conservation of mitosis-related proteins in cultured human amnion
cells. J. Cell Biol. 34: 97-110.
Stephens, R. E. 1969. As cited by Hicklas, 1971.
Stubblefield, E. 1968. Centriole replication in a mammalian cell,
p. 175-93. In The Proliferation and Spread of Neoplastic
Cells. Williams and Wilkins, Baltimore.
Stubblefield, E., and B. R. Brinkley. 1967. Architecture and function
of the mammalian centriole, p. 175-218. In K. B. Warren (ed.)
Formation and Fate of Cell Organelles. Academic Press, Hew York.
Stubblefield, E., and W. Wray. 1971 Architecture of the Chinese hamster
metaphase chromosome. Chromosoma 32: 262-294.
Taylor, J. H. 1963. EHA synthesis in relation to chromosome re
production and the reunion of breaks. J. Cell Comp. Physiol.
62 (suppl): 73-86.
Taylor, J, H., P. S. Woods, and W. L. Hughes. 1957. The organization
and duplication of chromosomes as revealed by autoradiographic
studies using tritium-labeled thymidine. Proc. Hat. Acad. Sci.
43: 122-128.


201
mm
* ^ -n t£
Mv>&
.. m .* _i \ ^


227
Whether centrioles also duplicate during cold treatment could only be
ascertained by direct observation or cineraicrography of living cells.
I have observed parent-daughter pairs in a cold-treated cell in early
prophase. In this case duplication could have occurred prior to treat
ment and the cell may have been completely arrested by cold.
No other organelle in cells of higher animals can match the
esthetic appeal of centrioles. Their orderly, symmetrical structure is
certainly in part the cause for much of the attention they have re
ceived since the early electron microscopic studies of Burgos and
Fawcett (1956) and de Harven and Bernhard (1956). In recent years,
however, the interest provoked by these intriguing organelles has led
to unwarranted speculation about their function (Brinkley and
Stubblefield 1970, Stubblefield and Brinkley 1967; see also discussions
by Pickett-Keaps 1969, 1971).
One such hypothetical function is the generation of continuous
spindle MT (Brinkley and Stubblefield 1970). It is clear that ultra-
structural studies can contribute only circumstantial evidence for this
hypothesis. In prophase cells, the centrioles appear to be the focus
of the increasing number of MT (Figure 11). Some of these will
undoubtedly become astral MT, others probably continuous and, possibly,
kinetochore MT. Except for the relatively rare skew MT, all the MT of
the fully formed spindle converge towards the centrioles (e.g., Figures
29 and 45). Several investigators have claimed to have observed MT
directly connected to the tubules of the centrioles (Krishan and Buck
1965), or inserted into the centriole wall between the triplets
(Brinkley and Stubblefield 1970, Gall 1961). I have examined these
published electron micrographs and found them totallyunconvincing.


176
Fig 77 Centrioles of cold-treated cell in metaphase. Note
clarity of triplets in cross-section, amorphous or fibrillar material
surrounding the same centriole. x 62,500.


25
kinetochore for the structure as it appears in the electron microscope.
o-. 3 o w o o
For this review, however, I use the terms in accordance with the
authors cited.
For many years most cytologists described the centromere as a
"gap" or non-staining constriction (for references see Schrader 1953).
Schrader (1936, 1939) presented a detailed description of the kineto-
chores of meiotic chromosomes in two species of amphibia and in
Tradescantia. He interpreted each tetrad to have two kinetochores,
each of which consisted of two spherules lying in a commissural cup.
Using various stains to enhance visibility of the spherules, he was
able to follow the changes in appearance of the kinetochores during the
different stages of meiosis.
Much of the early discussion about centromeres concerned the
problem of the presence or absence of DUA in this chromosome region.
Applying the Feulgen test to pachytene plant chromosomes, Lima-de-Faria
(1950) demonstrated the presence of DNA in the centromere, which he
described as consisting of fibrils and chromomeres. Gall (1954)
documented that kinetochores of newt lampbrush chromosomes are also
Feulgen-positive. He noticed that the kinetochores resemble
chromomeres in general appearance, except for the lack of lateral
loops. Lima-de-Faria (1956, 1958) furnished additional evidence for
his view of kinetochores of plant chromosomes as a specialized region
showing Feulgen-positive granules connected by fibrils.
Electron Microscopic Observations
The first good electron micrographs of animal kinetochores were
published by Harris (1962; see also Harris 1965, Harris and Mazia 1962)


26
and by Nebel and Coulon (1962). Satisfactory fixation was still a
problem in those days, but the kinetochores in dividing sea urchin eggs
could be clearly recognized as irregularly shaped, electron-dense gran
ules at the surface of the chromosomes where they are apparently
attached to MT (Harris 1962). Host remarkable was a difference in
staining intensity between the kinetochores and the remainder of the
chromosomes in preparations where the fixative had dispersed or
partially extracted the chromosomal fibers.
Nebel and Coulon (1962) interpreted the kinetochores of metaphase
I pigeon spermatocytes as having the shape of an acorn with MT attached
to the convex side of the cup. The details presented in their model
(their Figure 12), however, are not all discernible in the only low-
power electron micrograph included (their Figure 11). At best one can
distinguish a dense band, 2,000-4,000 A long, following the outline of
the chromosome and separated from the latter by a clear zone of
approximately the same width. The chromosome proper appears denser at
this site; poleward of the dark band a less dense matrix can be seen.
The authors interpreted the MT to penetrate the kinetochore and
terminate in the chromosome. .
Since these early studies a number of papers have appeared
describing the fine structure of kinetochores. The most detailed study
of the kinetochores of invertebrate cells was that by Luykx (1965a, b)
on Urechis eggs. Despite apparent fixation problems the dense-light-
dense banding was evident, particularly on meiotic chromosomes. The
author noted that the density of the kinetochore appeared to increase
from prometaphase to anaphase. The dense material was seen to closely
follow the curvature of the chromosome surface. The deep layer was


DISCUSSION
The phase contrast micrographs taken prior to sectioning of
selected cells were extremely helpful in determining mitotic stages.
Equally important, they facilitated the task of locating individual
chromosomes on thin sections, thus safeguarding against utterly false
conclusions. Consider, for example, Figure 27. From the electron
micrograph it would appear there is one long chromosome with four
kinetochores at the equator of the spindle. Phase contrast micrographs
at different levels of focus clearly revealed two different chromosomes
Not only the combination of light and electron microscopy, but
also the rat kangaroo cells themselves were an asset in this study.
The low chromosome number, the individuality of the chromosomes, and
the fact that the axis of the mitotic spindle in these cells is always
more or less parallel to the growth surface, all allowed a more
detailed study of mitosis than has ever been published for animal cells
Centrioles
Structurally, the centrioles of PtK^ cells, whether untreated,
colcemid-, cold-, or SN-treated (Figures 4, 9, 56, 77, and 94), are
similar to those in other cells and organisms (e.g., Brinkley and
Stubblefield 1970, de Harven 1968, Erlandson and de Harven 1971,
Robbins et al. 1968). The nine tubular triplets are embedded in an
osmiophilic matrix, which may be differentiated into ill-defined
structures (Figure 9). This matrix appears to be continuous with the
e o
extensive, slightly less osmiophilic material at the distal end
225


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EYQLTO04P_PV4S42 INGEST_TIME 2015-04-14T18:49:55Z PACKAGE AA00029836_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


166
mam
&&3&1
**&Vv
$ffiS3ga
f 3
Fig. 65. C-mitosis. Chromosome at the periphery of the central
area; Chdi, Chd2 its two sister chromatids. Kinetochores (K-j_, Kg),
centromeric granules (CG) in the primary constriction. The double
membranes (NE) are probably fragments of the nuclear envelope,
x 22,500* Inset: Phase contrast micrograph of the cell in plastic.
Chj_ the chromosome shown in the electron micrograph. Arrows indicate
centromeric granules (x 1,280). Treatment B.


78
phase contrast microscope at a magnification of 300. Obviously
aberrant, as well as doubtful cells were examined at magnifications of
625 and 1,560. Acentric fragments, maloriented chromosomes, and other
aberrations (e.g., multipolar spindles) were scored in metaphase cells
fragments, lagging chromosomes, and dicentric bridges were scored in
anaphase cells. The results are presented in Table 1.
The cumulative frequency of normal and abnormal cells is less
than 100%, because cells that could not be clearly identified as
normal or abnormal were included in the sample, but not assigned to
either category. The difference between the total frequency and 100%
is the frequency of these "doubtful" cases.
Gross and fine structure of aberrations
Acentric fragments and dicentric bridges.Figures 99-101
illustrate one of the most common types of aberration in anaphase
cells, viz., a single dicentric bridge and fragments. These bridges,
formed by a subterminal exchange between sister chromatids, are more
or less attenuated, probably depending on the degree of spindle
elongation. In both cells shown the kinetochores of the dicentric
chromosome were perfectly normal (e.g., Figure 100). There was
nothing unusual about the fibrous structure of the bridged chromosomes
The fragments in these cells were located in the cytoplasm at the
periphery of the spindle (insets in Figures 99 and 101). Analysis of
serial sections of several such cells revealed the truely acentric
nature of the fragments.
Dicentric bridges, single or multiple, also occur in telophase
cells. In this case the intact bridges cross the region of the


91
K& i &C*


NORMAL AND ABNORMAL MITOSIS IN A MAMMALIAN CELL IN VITRO.
A LIGHT AND ELECTRON MICROSCOPIC STUDY.
By
Urs-Peter Roos
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1971

Thus, the task is, not so much to see what nobody
has seen yet; but to think what nobody has thought
yet, about what everybody sees.
Schopenhauer

ACKNOWLEDGMENTS
I am grateful to Dr. Fred C. Johnson, Chairman of my Supervisory
Committee, for his help in procedural matters.
Words cannot express my respect and gratitude for the liberal and
stimulating guidance provided by Dr. Ienry C. Aldrich, Co-chairman,
throughout this study. His generosity regarding time and use of
facilities was an important factor for the progress of my work.
I am much indebted to Dr. John U. Cramer and his staff, Department
of Pharmacology, for instruction and use of facilities for tissue
culture work.
Dr. Lynn H. Larkin and Dr. James L. Nation, members of my
Supervisory Committee, deserve my thanks for their moral support and
helpful advice. I appreciate the encouragement provided by Dr. John W.
Brookbank, Dr. James H. Gregg, and Dr. Philip B. Morgan.
I am much obliged to Dr. Robert M. De Witt, Chairman of the
Department of Zoology, for arranging financial assistance.
I also extend my thanks to Mrs. Rosemary Rumbaugh, and numerous
unnamed persons with whom I was associated in the course of this study,
for assistance and forbearance.
My wife, Loan, played a major part in this accomplishment. I am
deeply grateful for her never faltering patience, her understanding,
encouragement, and help.
iii

PREFACE
Certainly, one cannot argue about the importance and significance
of attempts to increase our knowledge of chromosomes and mitosis, both
so basic to life in higher organisms. I was fortunate to have the
liberty to choose a research project myself, and the study presented
evolved mainly out of my curiosity to learn more about these two
fundamental aspects of life.
At the time this project began, ultrastructural investigations of
eukaryotic chromosomes had essentially reached a standstill. More new
questions had been posed than old ones answered. On the other hand,
electron microscopic techniques had not yet been applied to the study
of chromosomal aberrations. Mitosis in animal cells was fairly well
documented at the ultrastructural level, but many of its aspects were
the subject of controversy. Like many other students at the beginning
of a new road, 1 had only a limited knowledge of these problems; hence
my idealistic assumption that an ultrastructural investigation of
normal and abnormal mitosis could answer many of the questions
remaining, or at least lead the way to significant experiments. It
will become clear from the following presentation how far these hopes
were fulfilled and justified.
iv

TABLE OF CONTENTS
Page
Acknowledgments iii
Preface iv
List of Tables vii
List of Figures viii
Key to Abbreviations and Symbols xiv
Abstract xv
Review of Literature 1
General Ultrastructural Features of Vertebrate l
Cells in Mitosis
Centrioles 5
Spindle Fibers 11
Chromosomes 20
Kinetochore Structure and Function 24
Chromosomal and Mitotic Aberrations 33
Statement of Purpose 44
Materials and Methods 45
Cell Culture 45
Chemical and Physical Treatments 46
Fixation and Embedding 48
Preparation of Cells for Light and Electron
Microscopy 49
Results and Observations 52
Normal Mitosis 52
v

Page
0-. a o W O
C-Mitosis 67
Mitosis in Cold-Treated Cells 70
Kinetochore Fine Structure 71
Chromosomal and Mitotic Aberrations 77
Discussion 225
Centrioles 225
Microtubules 230
Chromosomes 235
Kinetochores 240
Nuclear Envelope 257
Conclusions 259
References 260
Biographical Sketch 276
vi

LIST OF TABLES
Table Page
1 Frequency of chromosomal and mitotic aberra- 79
tions in untreated meta- and anaphase
cells
2 Mitotic indices for streptonigrin-treated and 83
untreated control cells
3 Frequency of chromosomal and mitotic aberra- 84
tions in streptonigrin-treated cells ....

LIST OF FIGURES
Figure Page
1 Karyotype of the male rat kangaroo (Potorous 45
tridactylis)
2 The three planes in which blocks were 51
sectioned
3 Grazing section of an interphase nucleus ... 87
4a-d Centriole of interphase cell ......... 89
5a-c Transverse sections of interphase nuclei ... 91
6 Chromatin fibers of interphase nucleus .... 93
7 Nucleolus of interphase cell 93
8a Very early prophase cell 94
Sb-e Centriole duplication in very early 96
prophase .
9a,b Migration of centrioles in very early 97
prophase ... ......
9c-e Fine structure of centriole in very early 97
prophase ..
10 Mid-prophase nucleus 98
11a Mid-prophase nucleus 99
lib Mid-prophase nucleus, peripheral section 100
12 Early prophase. Grazing section of the 102
nucleus ........
13 Early prophase. Transverse section of the 102
nucleus
14 Late prophase chromosome near nuclear 104
envelope ........
viii

Figure Page
15 Late prophase chromosomes with
kinetochores 104
16 Mid-prophase chromosome with kinetochore ... 106
17 Mid-prophase chromosome with sister 106
kinetochores
18a Very early prometaphase cell ......... 107
18b Very early prometaphase cell 108
19 Early prometaphase cell 109
20 Early prometaphase cell 110
21a-c Early prometaphase. Nuclear envelope .... 112
22a,b Early prometaphase kinetochores 114
22c,d Early prometaphase kinetochores 116
23a-c Early prometaphase kinetochores 118
24a-c Early prometaphase kinetochores 120
25 Early proraetaphase kinetochores 121
26 Mid-proraetaphase cell 122
27 Mid- to late prometaphase cell 123
28 Late prometaphase chromosomes 124
29 Late prometaphase cell 125
30 Late prometaphase chromosomes 127
31 Late prometaphase chromosomes 127
32 Mid-prometaphase microtubules ........ 128
33a Late prometaphase chromosomes 129
33b Late prometaphase chromosomes 130
34a-c Late prometaphase chromosome 132
35a-c Late prometaphase chromosome ......... 134
36 Mid-prometaphase chromosome ......... 136
ix

Figure Page
o'. 3 o c. o
37a,b Late prometaphase chromosome 136
36a,b Late prometaphase chromosome 138
39a-e Late prometaphase chromosome 140
40a-c Late prometaphase chromosomes 141
41 Late prometaphase kinetochore 142
42 lletaphase plate 143
43 lletaphase chromosomes (para-sagittal
section ) 144
44 lletaphase plate (para-equatorial section) 145
45 Very early anaphase cell 146
46 Very early anaphase chromosomes 147
47 Early anaphase cell 148
48 Very early anaphase chromosomes ....... 150
49 Early anaphase kinetochores 150
50 Mid-anaphase cell (para-equatorial
section) 151
51a,b Late anaphase spindle and chromosomes .... 152
52 Very late anaphase, daughter nucleus 153
53 Very late anaphase, kinetochore ....... 155
54 Very late anaphase, stem bodies 155
55a Very late anaphase, stem body 155
55b Very late anaphase, equatorial region .... 157
56 Early anaphase centrioles ..... 157
57 Early telophase nucleus 158
58 Early telophase nucleus 160
59 liid-telophase kinetochores 160
60 liid-telophase nucleus 161
x

Page
Figure
61 Cytokinesis 162
62 Cytokinesis, midbody 163
63 C-metaphase 164
64 C-metaphase 165
65 C-mitosis 166
66 C-mitosis 167
67 Colcemid-treated interphase nucleus 169
68 Colcemid-treated interphase cell 169
69 C-mitosis 169
70 Cold-treated cell in mid-prophase 171
71 Cold-treated cell in mid-prometaphase .... 171
72 Prophase kinetochores of cold-treated cell 173
73 Microtubules of cold-treated cell in
raetaphase 173
74 Cold-fixed control cell in metaphase 173
75 Cold-treated cell in prometaphase 175
76 Cold-treated cell in cytokinesis 175
77 Ccntrioles of cold-treated cell in
metaphase ..... ...... 176
78a-c Serial sections of early anaphase
kinetochore 178
78d,e Serial sections of early anaphase
kinetochore (cont'd.) 180
79 Very early anaphase kinetochore 180
80 Kinetochore in late prometa- to metaphase 182
81 Metaphase kinetochore 182
82 Kinetochore in very early anaphase 182
83a,b Sister kinetochores in very early anaphase 184
xi

Figure Page
84 Early anaphase kinetochore 184
85a-c Metaphase kinetochore (para-sagittal
sections) 186
86a-h Metaphase kinetochore (para-equatorial
sections) ... ..... 188
87a,b TVo metaphase kinetochores (para-equatorial
sections) 190
88 Metaphase kinetochore (para-equatorial
sections) 190
89 Metaphase chromosome (para-equatorial
section) 192
90 Metaphase cell, para-equatorial section of
kinetochore MT 192
91a,b Mid-anaphase kinetochore (para-equatorial
sections) 192
92 Kinetochores in c-mitosis 194
93 Kinetochores in c-mitosis 194
94 Kinetochores in c-mitosis 194
95a-c Kinetochores in c-mitosis 196
96 Kinetochores in c-mitosis 198
97 Kinetochore in c-mitosis 198
98 Kinetochores in c-mitosis 198
99 Untreated anaphase cell, dicentric
bridge 199
100 Untreated anaphase cell, dicentric
bridge 201
101 Untreated anaphase cell, dicentric
bridge 201
102a,b Untreated anaphase cell, kinetochores of
lagging chromosomes 203
103a,b Untreated anaphase cell, kinetochore of non
lagging chromosome 205
xii

Figure Page
104 Untreated interphase cell, four centrioles 207
105a-c Untreated mid-prophase cell, abnormal
centrioles 207
106a,b Untreated cell in late prometaphase, abnormal
centrioles 209
107a-c Untreated anaphase cell, abnormal
centrioles 210
108a-f Streptonigrin-induced aberrations. Phase con
trast micrographs 212
109 Streptonigrin-treated cell in early
anaphase 214
110a,b Streptonigrin-treated cell in late
anaphase 214
llla-c Streptonigrin-treated cell in late
anaphase 216
112a,b Streptonigrin-treated cells in late
anaphase 217
113 Streptonigrin-treated cell in late
anaphase 219
114 Streptonigrin-treated cell in late
telophase 219
115 Streptonigrin-treated cell in late
telophase ............ 219
116 Streptonigrin-treated cell in late
telophase ............ 220
117a,b Streptonigrin-treated cell in late
telophase 222
118 Streptonigrin-treated cell in late
cytokinesis 222
119a,b Streptonigrin-treated cell in late
cytokinesis 224
120 Diagrammatic representation of kinetochore
maturation during prometaphase 243
121 Diagrammatic representation of the three-
dimensional structure of kinetochores ... 248
xiii

KEY TO ABBREVIATIONS AND SYMBOLS
C
CG
Ch
Chd
Chr
Ci
Co
CV
D
DC
EOC
F
G
Gr
H
HChr
IV
K
KI
KM
KO
MA
KB
MF
Mi
MI
MU
MT
N
NE
NL
NO
NP
Nu
NV
PC
R
RER
S
Sb
Sm
SN
V
X
centriole(s)
centromeric granules
chromosome(s)
chromatid(s)
chromatin
cistema(e)
corona of kinetochore
intracentriolar vesicle
contamination of thin sections (dirt)
daughter centriole(s)
8-ethoxycaffeine
fibrillar component of the nucleolus
Golgi complex
granular component of the nucleolus
achromatic hole(s) in chromatin or chromosomes
heterochromatin
intranuclear vesicle(s)
kinetochore
sister kinetochores
kinetochore granule
inner layer of the kinetochore
middle layer of the kinetochore
outer layer of the kinetochore
mitotic apparatus
midbody
microfibril(s)
mitochondrion
mitotic index
micronucleus
microtubule(s)
nucleus
nuclear envelope
nuclear lobe
nucleolus organizer
pore-annulus complex(es) of the nuclear envelope
nucleolus
membrane vesicle formed by fragmenting nuclear envelope
spindle pole(s)
particle(s) associated with centrioles
parent centriole(s)
ribosomes (poly- or monosomes)
rough endoplasmic reticulum
satellite(s), pericentriole body (bodies)
stem body
stem
streptonigrin
cytoplasmic vesicle(s)
X chromosome; other chromosomes are numbered arbitrarily
xiv

Abstract of Dissertation Presented to the Graduate Council of
the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
NORMAL AND ABNORMAL MITOSIS IN A MAMMALIAN CELL IN VITRO.
A LIGHT AND ELECTRON MICROSCOPIC STUDY.
By
Urs-Peter Roos
December, 1971
Chairman: Dr. Fred C. Johnson
Co-Chairman: Dr. Henry C. Aldrich
Major Department: Zoology
Rat kangaroo cells (PtK^ line), grown as monolayers, were fixed
and embedded in situ. Cells in mitosis were examined and photographed
under phase contrast. Serial sections were examined with the electron
microscope.
Centrioles duplicate at the onset of prophase. Centriole migra
tion, disintegration of the nucleolus, chromosome condensation, break
down of the nuclear envelope (NE), and formation of the mitotic spindle
are similar to these processes in other mammalian cells. Prophase
kinetochores (Ks) appear as globular, fibrillar bodies in the primary
constriction of chromosomes. Detailed observations on prometaphase
chromosomes support the pulling theory of amphiorientation and metaphase
stability. Amphiorientation is established by unipolar followed by
bipolar attachment to spindle microtubules (MT), or by simultaneous
bipolar attachment, depending on the position of a chromosome relative
to the poles at the time of attachment. In the former case, bipolar
attachment is presumably followed by congression. Concomitantly, Ks
mature to trilaminar, flat, undulated, concave, or convex discs, 4,000-
6,000 A in diameter. All the Ks reach maturity at metaphase. Inner and
outer layers are 400 A, and the middle layer is 300 A thick. On the
xv

average, 26 bundled MT are anchored in the outer layer. Staining
3 O i. *3 .9
characteristics and behavior in late mitosis suggest that the inner
layer is chromatin, the outer layer possibly protein.
Daughter chromosomes are connected by strands of chromatin during
very early anaphase. The strands rupture during early anaphase. In
late anaphase, when chromosome movement has ceased, the K layers are
fuzzy, and fewer, less organized MT are attached. In telophase the
inner layer is a very opaque patch of dense material on the inner
membrane of the NE; the outer layer is lost in the cytoplasm.
Reconstruction of the NE, cleavage, midbody formation, and
reappearance of the nucleolus, are similar to these processes in other
mammalian cells.
Mitotic stages in cells exposed to 0-4C for 1 hr are similar to
untreated cells. Prophase Ks are identical and mature metaphase Ks are
typically triple-layered, though slightly fuzzy. Few MT, almost
exclusively kinetochore MT, are preserved. They are coated by an
amorphous or finely fibrillar substance.
No spindle MT are present in colcemid-arrested cells (0.05 pg/ml
for 2 hr, 1 hr recovery; 0.25 pg/ml for 15 min, no recovery). Chromo
somes are scattered in the central area of the cells. Centrioles have
duplicated, but not migrated. The higher concentration of colcemid
induced decondensation of chromosomes and dispersion of chromatin in all
interphase cells. Kinetochores resemble immature prometaphase Ks in
control cells. The inner layer is lacking. The outer layer is a convex
or undulated, bilaminar plate, 400 A thick and 5,000-9,000 A in diameter,
and embedded in a finely fibrillar matrix.

Streptonigrin (SN; 0.01 pg/ml and 0.05 >ig/ml for 4 hr, 48 hr
recovery) depressed the mitotic index (2.22 and 0.18, respectively,
versus 2.59 in controls) and induced high frequencies of chromosomal
and mitotic aberrations in meta- and anaphase cells (49.07. and 1007.,
respectively, versus 147. in controls). The most frequent aberrations
in control cells were single dicentric anaphase bridges with a few
acentric fragments. Multiple bridges and fragments were the most
common aberrations in SN-treated cells. The higher concentration of SN
produced very complex aberrations, including decondensation of chromo
somes. Kinetochores of dicentrics are structurally normal. Micro
nuclei and very fine nuclear bridges connecting daughter cells in
cytokinesis in SN-treated cultures are enveloped by normal NE.
xvii

REVIEW OF LITERATURE
Mitosis has been studied for nearly 100 years (Dietz 1969). Its
fundamental importance as a mechanism for the orderly distribution of
chromosomes has provoked numerous investigations and experiments.
Accordingly, there is a vast amount of literature on both mitosis and
chromosomes. I did not intend to compete with the much more knowledge
able and experienced authors of the many reviews that have appeared.
Rather, I have tried to survey the literature on those aspects of
eukaryotic mitosis and chromosomes which are most closely related to
my own observations and results.
General Ultrastructural Features of
Vertebrate Cells in Mitosis
Today, improved methods are available that allow correlated light
and electron microscopic observation of single animal cells (e.g.,
Brinkley et al. 1967). It is a little surprising, therefore, that no
detailed study combining the two approaches has been done on animal
cells in mitosis. (For a good example of mitosis in plant cells see
Bajer 1968, Bajer and Mold-Bajer 1969.) Ultrastructural features of
mitosis have been described for a number of animal cells (see Luykx
1970 for references), but light micrographs are equally important for
comparison with earlier light microscopic studies.
In the following description the emphasis is on the nucleus.
Cytoplasmic organelles and components are considered only as far as
they play a role directly related to the formation of the mitotic
1

2
apparatus (MA) and chromosome segregation. Mitosis in newt heart
cultures (Bamicot and Huxley 1965), rat thymic lymphocytes (Murray
et al. 1965), HeLa cells (Robbins and Gonatas 1964), L strain
fibroblasts (Krishan and Buck 1965), and rat hepatoma cells (Chang and
Gibley 1968) served as the basis for this summary.
Prophase
The chromatin, which is dispersed during interphase, begins to
condense during early prophase. This process continues until, at the
end of this stage, individual chromosomes are clearly recognizable and
can be seen to consist of two sister chromatids. Kinetochores
(centromeres) may become visible during prophase in some cell types.
The centrioles may have duplicated during the previous interphase, or
they do so during prophase. So-called procentrioles, or daughter
centrioles arise at right angles to the separated parent centrioles.
At this stage the centrioles often lie within a pocket of the nuclear
envelope. A dense, osmiophilic mass can be seen associated with each
parent-daughter pair. As prophase progresses, the pairs migrate to
opposite poles. The timing of this event relative to the breakdown of
the nuclear envelope varies. In any event, microtubules (MT), which
radiate from the centrioles during early prophase, begin to invade the
nucleus as the envelope breaks down. Fragments of the latter appear
either as double stacks of membranes (Chang and Gibley 1968, Murray
et al. 1965), or as vesicles (Bamicot and Huxley 1965). The fibrillar
component of the nucleolus disappears (see also Hsu et al. 1965), the
granular component disperses and breaks up into smaller masses which
often remain attached to chromosomes during meta- and anaphase (Chang
and Gibley 1968).

3
Prometaphase
This stage, which is characterized by raetakinesis, i.e.,
chromosome movements that ultimately result in the alignment on the
metaphase plate (Mazia 1961), has been neglected by investigators of
the ultrastructure of mitosis in animal cells. This may be due partly
to difficulties in recognizing and selecting cells in this stage.
Metaphase
The chromosomes are condensed and aligned on the metaphase plate.
The mitotic spindle, made up of chromosomal and interpolar (continuous)
microtubules, is fully formed between the two poles occupied by the two
pairs of centrioles. Asters, consisting of MT radiating from the
centrioles, may be more or less distinct. Larger cell organelles, such
as mitochondria and endoplasmic reticulum (ER), are generally excluded
from the spindle, but exceptions occur. Ribosomes are abundant within
the spindle, mostly as monomers. Spindle MT converge at the poles,
where they enter the osmiophiiic zone around the centrioles. Bundles
of MT connect to the chromosomes at the kinetochores. The latter are
seen as straight or crescent-shaped, dark bands, separated from the
chromosome by a lighter band. Remnants of the nuclear envelope may be
present at the periphery of the spindle, often more concentrated near
the poles. Nuclear pore complexes are absent from these fragments.
Anaphase
As the two sets of daughter chromosomes separate, the chromosomal
(kinetochore) MT shorten, while the pole-to-pole distance increases
(spindle elongation). Later, the chromosomes fuse to one large mass of
dense chromatin. Concomitantly, pieces of double membrane, possibly

4
derived in part from the remnants of the nuclear envelope, appear at
O', J C < O/O
the periphery of the chromatin mass, first at the polar face.
Ribosomes and nuclear pores are found on some of these pieces. It is
not unusual for membranes to get caught between the coalescing
chromosomes. The presence of restored nuclear envelope seems
incompatible with the occurrence of ITT within the chromatin mass. In
the interzone, mitochondria and, in some cases, ER can be found, and
dense material around the MT in the equatorial region indicates early
stages of midbody formation.
Telophase
The nuclear envelope is completely reconstructed by fusion of
membrane vesicles and cistemae. Nucleoli are reconstituted from
material formed at the nucleolar organizer or by coalescence of small
nucleolar bodies, but the process is poorly understood (see Busch and
Smetana 1970). Concomitantly, the mass of chromatin disperses. The
cytoplasm is divided either by a wedge-shaped constriction in the
equatorial region (Robbins and Gonatas 1964), or by the formation of a
vesiculated equatorial plate (Murray et al. 1965). In the latter case
the vesicles fuse to form the cleavage furrow.
Cytokinesis
Cleavage progresses until the daughter cells are connected only by
a cytoplasmic stem. Included in the stem are tightly bundled MT and
the midbody, which has formed by fusion of the dense material in the
equatorial region (see also Byers and Abramson 1968, Paweletz 1967,
Schroeder 1970). The midbody is included in one of the daughter cells
or lost. Decondensation of the chromatin may not be completed until
late in interphase.

5
Centrioles
0 r J 'w' -
Occurrence
Centrioles occur in most or all animal cells (Brinkley and
Stubblefield 1970), in some fungi (Aldrich 1967, Renaud and Swift
1964), and some algae (Ringo 1967a). They are absent in cells of
higher plants, either before or during division (Pickett-Heaps 1969,
Wilson 1970). The possibility that certain non-dividing, fully
differentiated animal cells may also lack centrioles has been suggested
by Bernhard and de Ilarven (1960; see also de Harven 1968), but the
available evidence is inconclusive.
The apparent de novo origin of centrioles during certain stages of
the life cycle in many lower plants is an intriguing problem. A
detailed discussion, however, is beyond the scope of this survey. The
reader is referred to reviews by de Harven (1968), Luykx (1970), Mazia
(1961), and Pickett-Heaps (1969).
Ultrastructure
There is good agreement in the literature that centrioles of
interphase cells do not differ structurally from centrioles of the
mitotic apparatus (Brinkley and Stubblefield 1970, de Harven 1968,
Stubblefield and Brinkley 1967). A centriole is a cylindrical body
1,500-2,500 A in diameter and 3,000-7,000 A long. Nine triplets of
fused 240 A MT form its wall. Centriolar MT possibly have 13 subunits
in cross section, as do flagellar MT (Ringo 1967b). The triplets are
not arranged radially, but slanted, giving the image of a pinwheel.
Centrioles are polarized bodies. The so-called distal end is capable
of generating the shaft of a flagellum or ciliura (Gibbons and

6
Grimstone 1960). The proximal end, where daughter centrioles are
formed during duplication, exhibits a cartwheel structure with spokes
radiating from a hub. This cartwheel is recognizable in ideal cross
sections only, but it can also be reenforced photographically by the
Markham technique (Markham et al. 1963; for an example see de Harven
1968, Figure 2). The spokes seem to connect the hub with the innermost
MT of each triplet; other connections possibly occur between the
outermost and innermost MT of neighboring triplets. Centrioles are
often embedded in an amorphous, osmiophilic matrix (Murray et al. 1965,
Robbins and Gonatas 1964, Stubblefield and Brinkley 1967), that may
undergo periodic changes in preparation for, and during, cell division
(Robbins et al. 1968).
In Chinese hamster fibroblasts the occurrence of a small,
membrane-bound vesicle, termed nucleoid, approximately in the center of
the centriole seems to be the rule (Brinkley and Stubblefield 1970,
Stubblefield 1968, Stubblefield and Brinkley 1967). These
investigators also claimed to have detected a helical filament, 60-70 A
in diameter, that winds 8-10 turns just inside the triplets.
In many cells so-called satellites (pericentriole bodies) have
been observed (Brinkley and Stubblefield 1970, de Harven 1968, Murray
et al. 1965, Robbins et al. 1968). These are osmiophilic bodies in the
vicinity of interphase centrioles, but in most cell types studied they
seem to be absent or inconspicuous during mitosis (de Harven 1968).
They possibly contain RNA (Brinkley and Stubblefield 1970). The
existence of connections between satellites and centrioles, and the
possibly regular number and arrangement of satellites around centrioles
are controversial (see de Harven 1968 for a discussion). A few authors

7
have claimed that nine satellites form a symmetrical crown around the
Centrioles (Bessis et al. 1958, Gachet and Thiry 1964).
Chemistry
It is obviously a difficult task to determine the elemental or
molecular composition of centrioles. Cytochemical methods generally
lack sufficient resolution and bulk isolation of reasonably purified
centrioles seems virtually impossible. The important information
concerning the chemical composition of centrioles is therefore derived
from studies on basal bodies (kinetosomes) of cilia, and is based on
the assumption that the close structural and developmental relationship
between kinetosomes and centrioles (see de Harven 1968 for references)
justifies extrapolation of chemical analyses. Seaman's early findings
(1960) of 27. RNA and 37. DNA in kinetosomes of Tetrahymena did not
remain uncontested (see de Harven 1968 for a detailed discussion). At
best, it seems, the possibility that kinetosomes contain ENA and/or DNA
cannot be ruled out, but the presence of these important macromolecules
in centrioles remains hypothetical.
Duplication
As early as 1952 Inou£ concluded from experiments with colchicine-
treated Chactopterus eggs that centrioles duplicate and mature in the
absence of a mitotic spindle. In a now classical experiment Mazia
et al. (1960) determined the time sequence and mode of duplication of
mitotic centers in echinoderm embryos. Although the methods used by
the investigators did not allow direct visualization of centrioles
(hence the term "mitotic centers"), the results predicted what has
since been confirmed by electron microscopy. The results can be sum
marized as follows: (1) At all times the center is at least a

8
duplex structure. (2) Duplication occurs in very early interphase or
in late telophase of the preceding mitosis. (3) The splitting of the
centers is a process distinct from duplication, although the two
usually occur at about the same time during mitosis. As the two
members of a pair of "old" centers split, each one gives rise to a new
one with which it remains associated until the next cycle. (A) The
primary duplication event involves only a part of the parent structure.
(5) Mercaptoethanol (ME) inhibits duplication, but not splitting and
separation of existing centers. If ME is applied prior to duplication,
two daughter cells result from the first division after release from
inhibition. Multipolar divisions ensue if ME is applied after
duplication.
Went (1966) did a similar experiment with sand dollar eggs.
Benzimidazone (BIA) inhibits cell division, but not duplication of
mitotic centers. After release from BIA-inhibition the cells cleave
into as many blastomeres as there had been mitotic centers. Comparing
these results with those from ME-treatment, Went proposed there should
be a pair of potential mitotic centers at each pole of a BIA-induced
tetrapolar mitotic spindle, while in the case of ME-induced tetrapolar
cells there should be a single center.
The formation of so-called daughter or procentrioles during
duplication has been confirmed by electron microscopy (Bernhard and de
Harven 1960, Brinkley and Stubblefield 1970, Erlandson and de Harven
1971, Murray et al. 1965, Stubblefield and Brinkley 1967). Pro
centrioles arise at the proximal end of the parent centriole and
approximately at a right angle to the latter. They are structurally
similar to the parent centrioles, but much shorter during the early
stages of duplication.

9
Stubblefield (1968), using a technique that renders centrioles in
fixed cells visible for light microscopy, studied centriole duplication
and behavior in colceraid-treated Chinese hamster cells. His findings
concerning duplication, maturation, and separation agree well with the
above-mentioned inhibitor studies (Mazia et al. 1960, Went 1966).
Further confirmation came from an ultrastructural study on colcemid-
treated cells (Brinkley et al. 1967).
From the references and reviews mentioned the following picture of
the centriole cycle emerges: (1) At some stage between divisions each
of the two centrioles normally present in interphase cells produces a
daughter centriole by an as yet unknown mechanism. (2) The daughter
centrioles undergo maturation, which involves elongation and, possibly,
formation of the intracentriolar structures. If wTe define "mature" as
being capable of generating a procentriole and to participate in
spindle formation (see the following paragraph for reservations about
the latter), then the timing relative to the cell cycle under normal
conditions is so that procentrioles are mature no sooner than the end
of karyokinesis following duplication. (3) If spindle formation is
inhibited by colcemid or ME, maturation of the procentrioles continues
and they may act as mitotic centers after release from the block. As a
consequence, multipolar spindles occur more frequently than under
normal conditions, and the proportion of these increases with increasing
time of exposure to the inhibitor.
Function
Only the role played by centrioles in animal mitosis will be
discussed here. For the involvement of centrioles in the generation of

10
flagella and cilia see Dirksen and Crocker (1966), Renaud and Swift
(1964).
Two principal functions have been assigned to centrioles in
dividing animal cells: (1) Determination of the poles of the mitotic
apparatus, and (2) assembly of spindle MT (for detailed reviews see
Luykx 1970, Nicklas 1971, Pickett-Heaps 1969).
The axis of the mitotic spindle is not determined by the
orientation of the centrioles at the poles, except, possibly, in
special cases (de Harven 1968, Luykx 1970). A number of observations
seem to indicate that centrioles are indispensable as pole-
determinants. In normal bipolar mitosis there is one pair of
centrioles at each pole (Krishan and Buck 1965, Hurray et al. 1965,
Robbins and Gonatas 1964, Robbins et al. 1968, Stubblefield 1963,
Stubblefield and Brinkley 1967). Furthermore, cytasters produced in
sea urchin eggs by artificial activation are centered around poles
containing centrioles (Dirksen 1961). In contrast to this, the number
of poles in hybrid somatic cells often does not correspond to the
number of centriole pairs present (e.g., Yamanaka and Okada 1968), and
in cranefly spermatocytes mitosis occurs without centrioles under
certain conditions (Dietz 1959, 1966). Finally, mitosis in higher
plants proceeds without centrioles and many lower animals and plants
have pole-determinants other than typical centrioles (see Luykx 1970,
Pickett-Heaps 1969).
The idea that centrioles might be active in the assembly of MT
draws support from claims that spindle MT are directly connected to
centriole walls (Brinkley and Stubblefield 1970, Gall 1961), and from
the proposition that the poles may be an area where MT are assembled

11
and disassembled (Inou 1964, Inou£ and Sato 1967). It should be
noted, however, that the direct connection of MT to centrioles has not
been proven unequivocally (see de Ilarven 1968, Luykx 1970, Pickett-
Heaps 1969). More often, MT seem to connect to pericentriolar bodies
or the amorphous, osmiophilic mass surrounding the centrioles. Similar
dense masses can be found at the poles of higher plant cells (Pickett-
Heaps 1969, Wilson 1970). Most recent reviews, therefore, consider
this material, or part of it, a more likely candidate for a pole-
determinant and MT organizer, not only because it occurs almost
universally, but it would also bridge the gap between centriolar and
acentriolar mitosis (see Luykx 1970, Nicklas 1971, Pickett-Heaps 1969).
In this scheme, centrioles are thought to play a much more passive
role, being carried along and distributed mainly for use as basal
bodies of cilia and flagella (see also FriedlSnder and Wahrman 1970),
Brinkley and Stubblefield (1970; see also Stubblefield and
Brinkley 1967) have presented a different hypothesis. They maintain
that centrioles do play a role as pole-determinants and in the assembly
of MT. However, in the absence of solid evidence, their detailed
mechanistic and molecular model of centriole-MT interaction in mitosis
remains highly speculative.
Spindle Fibers
Under favorable conditions fibrillar elements are visible in the
mitotic spindle of living cells with the phase contrast microscope (Ris
1955). In fixed and stained cells spindle fibers can be seen without
difficulty in the light microscope (e.g., Heneen 1970). The Nomarski
interference-phase contrast system also allows visualization of the

12
fibrous organization in some cells (Bajer and Allen 1966). The most
convincing evidence concerning the reality of spindle fibers has come
from polarization microscope studies (Inou 1964, Inou and Sato 1967).
Early electron microscopic observations by Harris (1962) and Roth
and Daniels (1962) revealed numerous ITT in the mitotic spindle of sea
urchin eggs and amebae, respectively. That the mitotic spindle is a
collection of prominent MT has since been substantiated by numerous
other workers (e.g., Aldrich 1969, Krishan and Buck 1965, Robbins and
Gonatas 1964, Roth et al. 1966). The distribution of spindle MT agrees
well with that of spindle fibers as seen in the light microscope (see
Luykx 1970, Table III). Further support for the contention that the
spindle fibers of the light microscopists consist of bundles of MT came
from birefringence and electron microscopic studies on the effect of
colchicine, colcemid, vinblastine, cold, and ultraviolet light (UV) on
spindle MT. All these agents cause loss of spindle birefringence and a
reduction of the number of MT as seen in the electron microscope (Bajer
1969, Inou 1952, Inou and Sato 1967, Malawista et al. 1968, Roth
1967). On the other hand, treatment of mitotic cells with heavy water
increases birefringence and the number of MT in the spindle (Inou and
Sato 1967). The observation by Rebhun and Sander (1967), that MT are
not the only birefringent component of the mitotic spindle, imposes
limits on the interpretation of such results, but the fact remains that
MT are birefringent elements. The correlation between birefringence
and MT is generally regarded as sufficient evidence for the occurrence
of MT in living cells, particularly since improved fixatives for
electron microscopy have made the demonstration of MT easy (but see
Nicklas 1971 for a discussion of this point).

13
Following common usage in the literature, the terra spindle fiber
o-, 3 O'. r o O
shall hereafter designate the structure seen in the light microscope.
The term microtubule (MT) shall be reserved for the fibrils seen in
the electron microscope. By this definition a kinetochore fiber
consists of one or several MT.
Fine Structure
Spindle MT are slender cylinders 150-250 A in diameter and several
microns long (de Harven 1968, DuPraw 1968, Luykx 1970, Nicklas 1971,
Roth 1964). The tubules are hollow and have a wall 40-60 A thick. The
substructure of the wall is best resolved by negative staining of
isolated MT (Bamicot 1966, Kiefer et al. 1966), or by freeze-etching
(Moor 1967). Globular subunits, 30-40 A in diameter, form linear
filaments. Ten to thirteen such filaments form the tubule wall. The
filaments are arranged in such a way that the subunits form a helix
with a pitch of 10-20. Spindle MT are thought to be as rigid as
cytoplasmic MT (Luykx 1970), and therefore should run more or less
straight over a certain distance, or only be slightly curved. Wavy MT
can be a shrinkage artifact produced during dehydration (Jensen and
Bajer 1969).
Kiefer et al. (1966), citing a number of studies on the extraction
of proteins from mitotic apparatus, all of which seem to have yielded a
protein particle sedimenting at approximately 2.5S (MW approximately
34,000), proposed that this particle corresponds to the 35 A globular
subunit of the MT wall. Results from the various laboratories have,
however, been interpreted in different ways and the correlation between
visible subunits and isolated particles is not firmly established

14
(see Luykx 1970, Nicklas 1971, for detailed discussions). Inou and
his collaborators (Inou 1964, Inou and Sato 1967) have proposed that
MT polymerize from a pool of subunits in a dissociation-association
equilibrium, probably under control of "organizing centers" such as
kinetochores and centrioles. Certainly, the rapid loss and
reappearance of birefringence in mitotic cells subjected to rapid
temperature shifts (Inou et al. 1970) suggests self assembly of
subunits to form MT (see Nicklas 1971).
Fine interconnections or "cross-bridges" between spindle MT, and
"arms" on single MT have been reported for a number of plant and
animal cells (Hepler and Jackson 1968, Krishan and Buck 1965, Wilson
1969). Hepler et al. (1970) devoted a more detailed and systematic
study to these structures in cultured human cells and Haemanthus
endosperm, but although they play a major role in the model of mitosis
proposed by McIntosh et al. (1969), their reality is conjectural at
best (see also Nicklas 1971).
Distribution and Classification
Spindle MT can be divided into two main categories: chromosomal
and continuous (interpolar) MT (de Harven 1968, Luykx 1970, Nicklas
1971, Roth 1964). In the case of chromosomes with a well-defined
kinetochore the chromosomal MT run between the latter and the
corresponding spindle pole (Brinkley and Stubblefield 1966, 1970;
Harris and Mazia 1962, Jokelainen 1967, Krishan and Buck 1965, Murray
et al 1965, Robbins and Gonatas 1964). Chromosomes with so-called
diffuse kinetochores are connected with MT at various points along
their length (Buck 1967). Luykx (1970) distinguished several

15
subcategories of spindle MT. Notable among them are the penetrating
or transchromosomal MT (Nicklas 1971) observed in many animal cells
(e.g., Behnke and Forer 1966, Buck 1967, Jokelainen 1967, Robbins and
Gonatas 1964). They may actually be part of the population of
continuous MT.
It is obviously difficult to demonstrate that continuous MT do
run from one pole to the other, because they often pass into and out
of sections. However, this has been demonstrated in two cases
involving relatively short spindles (Aikawa and Beaudoin 1968, Mantn
et al. 1969b). On the other hand, studies on late stages of mitosis
indicate most "continuous" MT overlap in, and terminate beyond, the
midbody (Byers and Abramson 1968, Paweletz 1967). Understandably, this
point, which is important for an explanation of the function and
mechanics of the MA, is still controversial. More detailed studies,
involving counts of MT in serial sections cut at a right angle to the
spindle axis, have only recently been published (Brinkley and
Cartwright 1970, Mantn et al. 1969b, McIntosh and Landis 1971). It
appears that relatively few MT [l07. in the diatom Lithodesmium
(Mantn et al. 1969b, 1970); 30-407. of the interpolar MT which make up
40-507. of all the spindle MT in Chinese hamster and rat kangaroo cells
(Brinkley and Cartwright 1970)] are truely interpolar. Most
"continuous" MT project across and terminate beyond the equator.
It is generally agreed (see Luylcx 1970) that there is no
difference regarding structure and dimensions between the different
classes of spindle MT. However, kinetochore MT differ from continuous
and astral MT in their sensitivity to spindle poisons. Upon exposure
f Chaetopterus eggs to colchicine, the continuous fibers lose their

16
birefringence more rapidly than the chromosomal fibers (Inou 1952).
Sauaia and Mazia (1961) found that in sea urchin eggs the asters, but
not the kinetochore fibers, are disorganized by brief exposure to
colcemid. Likewise, a low concentration of colcemid disorganizes
continuous MT in Chinese hamster cells, but some kinetochore MT remain
(Brinkley et al. 1967). Similar differences apply for cold treatment.
Kinetochore fibers in lily pollen mother cells in anaphase regain
birefringence first when the cells are brought to ambient temperature
after chilling (Inou 1964). In contrast to this, continuous fibers
in Chaetopterus eggs are the first to regain birefringence lost during
chilling (Inoul 1964). In mammalian cells in vitro Brinkley and
Cartwright (1970) found cold shock completely disorganized interpolar
MT, while the number of chromosomal MT was reduced by 30-40%.
Function
It has been proposed that continuous fibers, which form the so-
called "central spindle" (Mazia 1961), function in the separation of
the pole-determinants (centrioles where applicable. Brinkley et al.
1967, Brinkley and Stubblefield 1970, Friedlander and Wahrman 1970,
Mazia 1961). In Chinese hamster cells exposed to colcemid, chromosomes
arrange more or less radially around the two unseparated pairs of
centrioles (Brinkley et al. 1967). The reformation of pole-to-pole MT
after release from the inhibitor seems necessary for the separation and
migration of the centrioles.
In cells with diffuse kinetochores, the penetrating or
transchromosomal MT may play an important, if not exclusive, role in
chromosome movement (Luykx 1970, Nicklas 1971). IThether they also

17
participate in the movement of chromosomes with a well-defined
kinetochore is uncertain. The often cited observations by Carlson
(1938) on apparent poleward movement at anaphase of X-ray-induced
acentric fragments in grasshopper neuroblasts are not convincing.
Bajer (1958) and Bajer and Mol-Bajer (1963) presented better evidence
based on cinemicrography of /3-irradiated Haemanthus endosperm cells.
In the majority of cases, however, acentric fragments do not behave as
normal chromosomes (Kihlman 1966, Lea 1962).
Finally, continuous MT are supposed to produce spindle elongation
at anaphase in cells where this occurs (Mazia 1961, Roth et al. 1966).
Spindle elongation is a process different from poleward movement of
chromosomes, as demonstrated by Ris' (1949) observation that chloral
hydrate prevents the former, but does not inhibit the latter.
The chromosomal fibers form the "chromosomal spindle" (Mazia
1961). Chromosomes with a distinct kinetochore are firmly attached to
the kinetochore fibers. This can be inferred from the above-mentioned
study by Ris (1949), and from observations by Inoue (1952) that in
Chaetopterus eggs treated with colchicine the chromosomes disperse from
the metaphase plate only after the birefringence of the chromosomal
fibers has completely disappeared. Shimamura (1940) reported that
chromosomes at the centripetal pole of centrifuged lily pollen mother
cells are firmly anchored by their kinetochore fibers, although the
centrifugal force is strong enough to cause uncoiling of the
chromosomes themselves. Finally, the most direct evidence came from
elegant micromanipulation experiments by Nicklas and Staehly (1967) on
grasshopper spermatocytes. Chromosomes can be stretched with a
microneedle without changing the kinetochore-to-pole distance

18
significantly from the beginning of prometaphase to the end of
anaphase. Lateral displacement within the spindle is possible,
however. The authors found no evidence for interzonal connections
between separating chromosomes at anaphase.
There is a large body of evidence that kinetochore fibers play a
major, if not exclusive, role during metakinesis, i.e., those movements
of the chromosomes that begin with prometaphase and terminate with the
alignment of the chromosomes on the equator of the spindle (congres-
sion), as well as in the poleward movement of chromosomes during
anaphase. Bivalents of spermatocytes in prometaphase often have the
shape of an arrowhead with the kinetochore at the tip (Dietz 1969).
If one bivalent is stretched to a greater extent than its partner, the
movement is always in the direction of the pole towards which the
former is oriented. More direct evidence comes from UV microbeam
irradiation experiments. In newt fibroblasts irradiation of the
kinetochore, but not of other parts of a chromosome, stops prometaphase
movement and the irradiated chromosome never reaches the metaphase
plate (Bloom et al. 1955, Uretz et al. 1954). In Ilaemanthus endosperm
cells, on the other hand, similarly treated chromosomes do reach the
metaphase plate, although their paths of movement during prometaphase
may be altered and become very complex (Bajer and Mol-Bajer 1961).
Chromosome motion at metaphase is very slow and minimal (Dietz
1969, Hazia 1961). The equatorial position of the chromosomes is very
likely maintained by a balance of forces on opposed sister kineto-
chores. If one kinetochore pair of a bivalent in a grasshopper
spermatocyte in metaphase is irradiated with a UV microbeara, the
bivalent shifts towards the pole closest to the unirradiated

19
kinetochore (Izutsu 1961). An exception to this rule was reported by
Forer (1966) for crane fly spermatocytes. Here, UV microbeam
irradiation of the chromosomal fibers on the poleward side of the
metaphase plate does not induce movement. A possible explanation for
this behavior is the apparent absence of any kind of kinetochore on
these chromosomes, as reported by Behnlce and Forer (1966). Finally,
the necessity of opposing poleward forces for stable metaphase
alignment was clearly demonstrated in Nicklas' laboratory (Henderson
and Koch 1970, Henderson et al. 1970, Nicklas and Koch 1969).
It is a well-supported conclusion that chromosomes move as
individuals, although they may move synchronously (Luykx 1970, Mazia
1961, Nicklas 1971). Since Fas' study (1949) two types of anaphase
movement are distinguished: (1) spindle elongation, and (2) shortening
of the chromosome-to-pole distance. The two processes are based on
different mechanisms, because the former, but not the latter, is
inhibited by chloral hydrate (Ris 1949). In grasshopper spermatocytes
the two processes act together (Ris 1949), but in other cells each of
the two possible extremes can occur (see Mazia 1961 for references).
Mazia (1961), who also discussed the various early hypotheses
concerning anaphase movement, summarized the events as follows: "The
central spindle is more or less rigid; it moves the poles apart and
provides an anchor for the poles which must bear the load of the
chromosomes."
The question whether chromosomes are pulled or pushed has engaged
the mind of many a biologist. Most reviewers (Dietz 1969, Luykx 1970,
Mazia 1961, Nicklas 1971) arrived at the conclusion that a pulling
force must be involved in the movement of chromosomes, although a

20
pushing force (similar to the "Stemmkorper" proposed by Blar 1929) may
contribute to anaphase separation. Clearly, prometaphase movement
cannot be explained on the basis of pushing forces only. The behavior
of chromosomes that are pushed away from the spindle into the cytoplasm
by micromanipulation also suggests a pulling force (Nicklas 1967). The
results of Forer's (1966) UV microbeam irradiation experiments on crane
fly spermatocytes seem to refute the idea that chromosomes are simply
pulled by their kinetochore fibers. Nevertheless, the results can be
interpreted as supporting a pulling hypothesis (Nicklas 1971).
Even if we accept the pulling hypothesis there remains the problem
of how the MT accomplish this, a problem around which center all of the
more recent models of mitosis (Dietz 1969, Luykx 1970, McIntosh et al.
1969). There is no change in diameter as MT shorten or lengthen (see
Luykx 1970, Nicklas 1971, for references), and a simple contraction
mechanism is not compatible with structural observations. Inou and
his collaborators (Inou 1960, 1964, Inou and Sato 1967) have
consistently explained this phenomenon with their "dynamic equilibrium
model," which proposes a pool of MT subunits in the spindle region.
Free subunits are in a dynamic equilibrium with subunits bound in MT.
Shifts in the equilibrium induce further polymerization or
depolimerization. The former would produce lengthening, the latter
shortening of MT. Orienting centers are thought to determine the
direction of the "growing" MT during polymerization.
Chromosomes
Three basic questions must be considered here: (1) What is the
unit fiber of the mitotic (and interphase) chromosome? (2) What is the

21
relationship between the DMA double helix and the unit fiber? (3) How
is the unit fiber arranged in the highly condensed metaphase chromo
some? Data and observations bearing on these questions come from a
number of fields of study and so far it has not been possible to
reconcile them in one comprehensive model of the eukaryotic chromosome.
In thin sections of interphase nuclei and mitotic chromosomes
fibrils approximately 100 A in diameter are visible (Wolfe 1969).
Because the fibers are cut at various angles, little can be concluded
regarding their three-dimensional organization. Whole-mount
preparations of isolated chromosomes seemed initially much more
promising (for references see Wolfe 1969). In isolated chromosomes the
diameter of the fibers varies from approximately 20 A to 500 A or more,
depending on the quality of the preparation, but most investigators
agree that the mean diameter is approximately 250 A (DuPraw 1968, Wolfe
1969). However, Ris has demonstrated (1961, 1967; Ris and Kubai 1970)
that 100 A fibers can consistently be obtained. Understandably, a
lively controversy revolves around these differing results and their
interpretation. DuPraw (1968) considers the 250 A fiber as the unit
fiber (type B fibril). On the other hand, Wolfe (1969; see also Zirkin
and Wolfe 1970) considers this to be an artifact produced by the
deposition on a 100 A fiber of contaminating material during
preparation. A third view is held by Ris (Ris 1967; Ris and Kubai
1970): after certain chemical treatments the 250 A fibers can be
shown to consist of two 100 A fibrils, more or less twisted around each
other. Both Ris (Ris and Kubai 1970) and Wolfe (1969) present
arguments and evidence supporting their respective hypothesis. At the
center of the dispute are the many possible artifacts produced by

22
differences in the method of preparation. A discussion of these is
beyond the scope of this summary. Perhaps Ris and Kubai (1970) are
closer to the truth in their conclusion that the 250 A fiber is the
structure present in the intact nucleus, but that its relationship to
the 100 A fiber is not yet clear.
The second question can be divided into two parts: (a) Does the
unit fiber contain a single DNA double helix or several in lateral
association? The DNA-histone complex is expected to have a diameter of
approximately 30 A (Zubay and Doty 1959). Many investigators have
claimed to have observed a fibrillar substructure within the 100 A
fiber (see Wolfe 1969, for references). However, enzyme digestion and
other treatments that are assumed to remove all or most of the histones
from the chromosomal fiber leave a strand 20 A thick, which most
probably represents a single ENA double helix (DuPraw 1968, Ris and
Kubai 1970, Wolfe 1969). It is thus generally accepted that the unit
fiber contains a single TOA double helix. However, there is ample
evidence that this helix is highly compacted in the unit fiber (DuPraw
1968, Ris and Kubai 1970). The major role in folding of the DNA has
been assigned to the histones, but the molecular basis is not yet clear
(see Ris and Kubai 1970, for discussion).
(b) Is the DNA double helix in the unit fiber continuous or does
it consist of subunits of variable length, perhaps connected by so-
called linkers? Sedimentation coefficients and direct electron
microscopic measurements of purified DNA from isolated chromosomes or
chromatin yield fragments of variable length (see Ris and Kubai 1970).
Similar results are obtained with autoradiography of DNA labeled during
replication. Both methods, however, have their limitations and do not

23
unequivocally support the idea that the DNA is discontinuous.
O'. j O '
Similarly, the observation of multiple replication points in eukaryotic
nuclei is no definite evidence (see DuPraw 1968), and the accepted
conclusion is that the DNA molecule in the unit fiber is continuous
(DuPraw 1968, Ris and Kubai 1970).
The third question, concerning the organization of the unit fiber
in metaphase chromosomes, is at the center of the uninemy-polynemy
controversy. The answer is very simple: the problem is not resolved
(DuPraw 1968, Ris and Kubai 1970, Wolfe 1969). Bajer (1965) observed
the half-chromatid structure of chromosomes in living Haemanthus
endosperm cells. When fixed metaphase chromosomes from Vicia cells
pretreated with 5-amino uracil are digested with trypsin, each
chromatid appears to consist of two subunits (Wolfe and Martin 1968).
In whole-mounted, isolated chromosomes, each chromatid contains many
fibers (e.g., Abuelo and Moore 1969, DuPraw 1968, Lampert 1969,
Stubblefield and Wray 1971, Wolfe 1965), which can be interpreted to
form longitudinal subunits of chromatids, although it is virtually
impossible to follow individual strands over a greater distance. Cells
exposed to X-ray or certain chromosome-breaking chemicals during
exhibit, at the following metaphase, bridges between sister chromatids
that involve apparent subunits of chromatids (subchromatid aberrations
or "side-arm bridges"; Brinkley and Humphrey 1969, Heddle 1969, Kihlman
1966). Kihlman, who formerly supported polynemy based on his
experiments with chromosome-breaking chemicals (e.g., Kihlman 1966),
has recently questioned the occurrence of true subchromatid aberrations
(Kihlman 1970), and favors the single-stranded "folded fiber" model of
DuPraw (1965, 1968).

24
Isolabeling of chromatids (e.g., Peacock 1963) has always been a
strong argument in favor of polynemy, but it can be accommodated in a
single-stranded chromosome model (Comings 1971, Ris and Kubai 1970).
On the other hand, the observation that chromosomes replicate semi-
conservatively (Taylor et al. 1957) is in favor of uninemy. As Ris and
Kubai (1970) pointed out, the most compelling evidence against multi
strandedness comes from studies on the uniqueness or redundancy of DNA
sequences (e.g., Britten and Kohne 1968). In Drosophila and the mouse
the majority of DNA sequences are unique. This would exclude the
presence of two or more identical ENA strands per chromatid, which is,
of course, a prerequisite for multi-strandedness dictated by the
orderly segregation of genes at mitosis.
In view of these contradictory observations and results it seems
premature to propose a detailed model of the architecture of metaphase
chromosomes. For examples of such models the reader is referred to the
review by DuPraw (1968) and the paper by Stubblefield and Wray (1970).
Kinetochore Structure and Function
Light Microscopic Observations
Chromosomal granules, presumably corresponding to the centromere
studied by many later cytologists, were first described by Metzner
(1894), who called them "Leitkorperchen." By 1930 a number of
different terms were in use (see Schrader 1936). Sharp (1934)
introduced the term kinetochore," which has been used as a synonym of
"centromere" by many authors. For clarity I prefer to apply the term
centromere to the structure seen with the light microscope in the
primary constriction of chromosomes, and to reserve the term

25
kinetochore for the structure as it appears in the electron microscope.
o-. 3 o w o o
For this review, however, I use the terms in accordance with the
authors cited.
For many years most cytologists described the centromere as a
"gap" or non-staining constriction (for references see Schrader 1953).
Schrader (1936, 1939) presented a detailed description of the kineto-
chores of meiotic chromosomes in two species of amphibia and in
Tradescantia. He interpreted each tetrad to have two kinetochores,
each of which consisted of two spherules lying in a commissural cup.
Using various stains to enhance visibility of the spherules, he was
able to follow the changes in appearance of the kinetochores during the
different stages of meiosis.
Much of the early discussion about centromeres concerned the
problem of the presence or absence of DUA in this chromosome region.
Applying the Feulgen test to pachytene plant chromosomes, Lima-de-Faria
(1950) demonstrated the presence of DNA in the centromere, which he
described as consisting of fibrils and chromomeres. Gall (1954)
documented that kinetochores of newt lampbrush chromosomes are also
Feulgen-positive. He noticed that the kinetochores resemble
chromomeres in general appearance, except for the lack of lateral
loops. Lima-de-Faria (1956, 1958) furnished additional evidence for
his view of kinetochores of plant chromosomes as a specialized region
showing Feulgen-positive granules connected by fibrils.
Electron Microscopic Observations
The first good electron micrographs of animal kinetochores were
published by Harris (1962; see also Harris 1965, Harris and Mazia 1962)

26
and by Nebel and Coulon (1962). Satisfactory fixation was still a
problem in those days, but the kinetochores in dividing sea urchin eggs
could be clearly recognized as irregularly shaped, electron-dense gran
ules at the surface of the chromosomes where they are apparently
attached to MT (Harris 1962). Host remarkable was a difference in
staining intensity between the kinetochores and the remainder of the
chromosomes in preparations where the fixative had dispersed or
partially extracted the chromosomal fibers.
Nebel and Coulon (1962) interpreted the kinetochores of metaphase
I pigeon spermatocytes as having the shape of an acorn with MT attached
to the convex side of the cup. The details presented in their model
(their Figure 12), however, are not all discernible in the only low-
power electron micrograph included (their Figure 11). At best one can
distinguish a dense band, 2,000-4,000 A long, following the outline of
the chromosome and separated from the latter by a clear zone of
approximately the same width. The chromosome proper appears denser at
this site; poleward of the dark band a less dense matrix can be seen.
The authors interpreted the MT to penetrate the kinetochore and
terminate in the chromosome. .
Since these early studies a number of papers have appeared
describing the fine structure of kinetochores. The most detailed study
of the kinetochores of invertebrate cells was that by Luykx (1965a, b)
on Urechis eggs. Despite apparent fixation problems the dense-light-
dense banding was evident, particularly on meiotic chromosomes. The
author noted that the density of the kinetochore appeared to increase
from prometaphase to anaphase. The dense material was seen to closely
follow the curvature of the chromosome surface. The deep layer was

27
often less dense than the superficial layer and more variable in
o 1 m *
appearance. Approximately 10-25 MT, some of which seemed to end in the
deep layer, could be counted per kinetochore. On mitotic chromosomes
the triple-layered structure was less frequent. DNAse treatment in
dicated little or no DNA in the kinetochore region, or increased
resistance of kinetochore DNA to the enzyme (Luykx 1965a).
Wettstein and Sotelo (1965) found that the kinetochores in Gryllus
spermatocytes consist of essentially the same 100 A fibrils as the
remainder of the chromosome, but the fibrils seemed more densely packed
in the kinetochore. The shape of these kinetochores resembles that of
a thick nail or screw deeply anchored in the body of the chromosome.
Grasshopper spermatocytes have ovoid kinetochores embedded in a
cup (Brinkley and Nicklas 1968, Nicklas 1971, and personal communica
tion; see also Brinkley and Stubblefield 1970, Figures 26 and 27). An
electron-dense "axial core" is embedded in a mass of less dense 50-80 A
fibrils. In general appearance these kinetochores seem to resemble
plant kinetochores (Bajer and Mole-Bajer 1969, Wilson 1968) more than
animal kinetochores.
A different kind of kinetochore in invertebrates is the diffuse
kinetochore described by Buck (1967) in the bug Rhodnius. He
interpreted the finely granular material on the surface of metaphase
chromosomes as representing the kinetochore, but noticed the relative
paucity of spindle MT attached to this structure. Ris (Ris and Kubai
1970) reported on a preliminary basis that no such material could be
found on chromosomes in spermatocytes of the homopteran insect
Philaenus. Rather, the MT seemed to penetrate deeply into the
chromosome where they end blindly.

28
Mammalian kinetochores are by far the most intensely studied.
Besides papers devoted specifically to mitosis or kinetochores, there
are numerous isolated electron micrographs in the literature showing
various profiles of mammalian kinetochores (e.g., Flaks 1971, Hu 1971).
Unfortunately, some of the most frequently cited papers on the ultra
structure of mammalian cells in mitosis (e.g., Krishan and Buck 1965,
Murray et al. 1965) do not show kinetochores at high magnifications.
At best, one can make out that they consist of three layers or bands,
like those described by Nebel and Coulon (1962). Robbins and Gonatas
(1964, Figure 19) presented a detailed picture of an early anaphase
kinetochore in a HeLa cell. The three layers are very clear; MT seem
to insert into the outer layer. Bamicot and Huxley (1965) also
published electron micrographs showing the kinetochores of cultured
newt heart cells as three-layered structures. They interpreted the
kinetochores to consist of material different from the chromosomes,
based on different stainability. In this context it is interesting to
note that the kinetochores of the alga Oedogonium (Pickett-Heaps and
Fowke 1969) and the moss Mnium (Lambert 1970) are also triple-layered,
as are those of Barbulanympha (Hollande and Valentin 1968), which has
an intranuclear spindle.
Jokelainen (1965a) described kinetochores in fetal rat kidney as
short bands separated from the chromosomes by a clear area, or as two
parallel bands, in which case the second band was in direct contact
with the chromosomal material. In subsequent papers (Jokelainen 1965b,
1967, 1968) he developed his concept of kinetochore maturation and his
kinetochore model. Maturation occurs during prometaphase and is
asynchronous for sister kinetochores. It involves the appearance of

29
c
the outer layer and a reduction in size of the kinetochoral patch as it
becomes anchored to the spindle MT (Jokelainen 1965b). The outer layer
is embedded in a moderately dense substance, part of which persists at
metaphase as the middle layer of the kinetochore and the so-called
corona (Jokelainen 1967). Jokelainen's model (1967) depicts the
kinetochore as a trilaminar disk, 2,000-2,500 A in diameter, at the
surface of the chromosome, sometimes slightly recessed, sometimes
projecting. The outer layer is 300-450 A thick, the middle layer
150-300 A, and the inner layer 150-250 A. The corona over the outer
layer measures approximately 300 A in thickness. Evidence for the
disk-like appearance came from para-equatorial sections showing
kinetochores in face view. The outer disk, which is finely granular or
fibrillar, stains consistently, while the density of the chromosomes
varies with the staining method employed. The inner layer is highly
electron-dense, and roughly granular or fibrillar. Four to seven MT
are attached to each kinetochore, apparently penetrating the outer and
middle layers and sometimes ending in the chromosome.
Brinkley and Stubblefield (1966, 1970) presented a different model
of mammalian kinetochores. In colcemid-arrested Chinese hamster cells
the kinetochore appears as a 200-300 A wide band, made up of two
50-80 A threads. This band is embedded in a less dense matrix and
follows the curvature of the chromosome surface at a distance of
approximately 100 A. In the less dense matrix the authors detected
50-80 A fibrils apparently looping out from the main band. This
description was based on kinetochores without attached MT, but
according to the authors the kinetochores of untreated cells are very
similar. The model of Brinkley and Stubblefield (1966, 1970) describes

30
the kinetochore as consisting of two lampbrush-like structures, each
O', 3 o' o o .o
made up of two closely associated 50-80 A axial filaments from which
numerous 50-80 A fibrils loop out laterally. The axial filaments
extend along the surface of the chromosome, their ends being inserted
into the latter. Microtubules attach to the axial filaments in sheets
or bundles. According to their most recent report (Brinkley and
Stubblefield 1970) there is little change in the structure of the
kinetochores in Chinese hamster and rat kangaroo cells from prophase to
metaphase. The kinetochores are mature in late prophase or in pro
metaphase, regardless of whether MT are attached or not.
An electron micrograph recently published by McIntosh and Landis
(1971, Figure 4) supports Jokelainen's model (1967). In this para-
equatorial section of the metaphase plate three kinetochores are shown
as circular patches of less dense material. On the other hand, the
kinetochores of colcemid-treated cells, on which Brinkley and
Stubblefield (1966, 1970) mainly based their model, may be atypical.
For example, very clear images of a double-banded outer layer embedded
in a less dense matrix are produced by the alkaloids vinblastine and
vincristine, both of which disorganize MT (George et al. 1965, Journey
et al. 1968, Journey and Whaley 1970, Krishan 1968).
Microtubules and kinetochores as seen in thin sections are not
preserved in whole-mount preparations of metaphase chromosomes (e.g.,
DuPraw 1968). Instead, chromosomal fibers can be seen to cross between
sister chromatids in the centromere region (see also Abuelo and Moore
1969). Interesting, but unexplained, is the presence of four dense
granules at the centromere of Chinese hamster cells subjected to
certain treatments during isolation (Stubblefield and Wray 1971).

31
Function
In addition to the role of the kinetochore in chromosome movement
as discussed in a previous section, Luykx (1970) reviewed five other
possible functions: (1) Initiation of synapsis and localization of
chiasmata; (2) terminalization of chiasmata; (3) chromosome condensa
tion or coiling; (4) association of sister chromatids; and (5)
formation or assembly of chromosomal spindle fibers. The latter is the
most interesting in the context of this review.
When chromosomal fibers are irradiated between the centromere and
the pole in anaphase cells, birefringence disappears from the
irradiated region and distal to it. Restoration of birefringence takes
place within a few minutes. When, however, the centromere itself is
irradiated, birefringence disappears from the whole length of the
chromosomal fibers, including the distal non-irradiated portion, and
restoration does not occur for a long time (Inou 1964). Electron
micrographs have shown that a great proportion of spindle LIT is
arranged in bundles associated with the kinetochores (Brinkley and
Landis 1970, Brinkley and Stubblefield 1970, Jokelainen 1967). These
observations have led to the idea that the kinetochore may play an
active role in the assembly and/or orientation of MT (e.g., the
"organizing center" of Inou 1964). Further support for this idea has
been drawn from observations of apparent microtubular connections
between meiotic sister kinetochores (Luykx 1965b) and mitotic non
sister kinetochores (Bajer 1970). As long as clear evidence from
serial sections is lacking, however, such configurations have to be
regarded with skepticism. On the other hand, the occurrence of
individual chromosomal spindles, either as an anomaly (e.g., Dietz

32
1969, Koopraans 1958) or as part of normal spindle development (Hughes-
Schrader 1948), seems to indicate that chromosomes alone are capable of
organizing spindle MT. Similarly, Roth (1967) suggested that in
spindle reformation after cold shock the spindle MT gradually extend
from the chromosomes towards the poles. These observations have been
interpreted to mean that spindle fiber material is being continuously
assembled and oriented at the kinetochore throughout prometaphase and
metaphase, but the exact role of the kinetochore is not as yet clear
(see Luykx 1970).
Equally interesting in this context are ideas of the kinetochore
as a specialized "gene" (Brinkley and Stubblefield 1970) or a gene
product (Luykx 1970). Even if we accept the universal occurrence of
DMA in the centromere region as a fact, we must bear in mind that the
relationship between DMA and the kinetochore at the fine structural
level is not at all resolved. Brinkley and Stubblefield (1970) have
proposed that the lateral loops of their lampbrush-like kinetochores
consist of DMA, which codes for long ENA molecules that bind with
protein subunits to form MT. This elaborate hypothesis stands on
rather shaky ground. Luykx (1970),on the other hand, proposed the DMA
in the kinetochore region "may therefore be viewed as a 'kinetochore
organizer', similar to the nucleolar organizing region of the
chromosome in a number of ways. It is responsible for the synthesis or
assembly of an essential organelle that remains associated with a
specific chromosomal site, is often associated with blocks of
heterochromatin, remains relatively uncoiled during mitosis, and
probably contains a large number of identical genes" (Luykx 1970).

33
Chromosomal and Mitotic Aberrations
Chromosomal aberrations occur spontaneously in certain organisms
(e.g., Brandham 1970, Vig 1970). Far more frequent, however, are
induced aberrations. Shaw (1970) presents a long list of agents
(clastogens) that can cause chromosome damage. The agents range from
nucleic acid analogs, antibiotics, drugs, and pesticides, to ionizing
and UV radiation, temperature shock, and weightlessness. Certain
viruses are well-known biological clastogens (e.g., Nichols 1970,
Nichols et al. 1964).
Only some of the elementary aspects of the induction of chromo
somal and mitotic aberrations can be reviewed here. The emphasis will
be on chemical clastogens. Radiation-induced aberrations will be
referred to only to the extent that they elucidated basic facts that
also apply to chemicals (for detailed reviews on radiation-induced
aberrations see Bacq and Alexander 1955, Evans 1962, Hollaender 1954,
Lea 1962, Wolff 1963).
Cellular Events
Kihlman (1966) distinguished three types of effect of chemicals on
dividing cells and chromosomes: (1) Prevention of cells from entering
mitosis [e.g., 5-fluorodeoxyuridine, FUdR (Taylor 1963)]; (2) inter
ference with active stages of division [e.g., spindle poisons, such as
colchicine (Eigsti and Dustin 1955)]; (3) production of chromosomal
aberrations [e.g., streptonigrin (Cohen et al. 1963)]. It is
characteristic for many chemicals that they affect mitosis and
chromosomes (see also reviews by Biesele 1962, Deysson 1968, Gelfant
1963). For example, FUdR inhibits mitosis and also fragments

34
chromosomes (Hsu et al. 1964). Other chemicals are primarily
O' 3 9 3 O
chromosome-breaking agents, but at the same time they affect the
mitotic rate. To discriminate between chromosomal and mitotic effects,
Deysson (1968) classified the cytological effects of antimitotic
substances as follows: mitodepressive (lowering the mitotic rate),
mitostatic (no proliferation), mitoclasic (disturbances of the mitotic
apparatus), and chromatoclasic (induction of aberrations). Generally,
the lowest effective concentration of a given chemical is mitodepres
sive. The same, or a slightly higher concentration produces
chromosomal aberrations. Higher concentrations, besides inducing
chromosomal anomalies, cause preprophase inhibition (mitostatic
effect), and still higher concentrations destroy cells in mitosis.
Chemicals Versus Ionizing Radiation
Aberrations produced by ionizing radiation applied to cells in
are of the chromosome type (e.g., Heddle 1969). The transition from
chromosome to chromatid aberrations occurs at the end of G^ (Evans and
Savage 1963), but some authors maintain that both chromosome and
chromatid aberrations can result from irradiation during S, depending
on whether unreplicated or replicated parts of the chromosomes are hit
(e.g., Casarett 1968). Chromatid aberrations only result from
irradiation of cells in G^, i.e., after completion of ENA synthesis.
Subchromatid aberrations can be induced during prophase and, possibly,
at the end of G^ (see Heddle 1969). The term non-delayed effect used
by Kihlman (1966) with reference to chemical clastogens and radiation
implies aberrations produced after completion of ENA synthesis
(chromatid and subchromatid aberrations), while delayed effects include
chromatid and chromosome aberrations (induced during G^ and S).

35
The often used term "radiomimetic" chemicals for substances that
induce chromosome damage is rather misleading. As already mentioned,
ionizing radiation produces both non-delayed and delayed effects, while
most chemicals (e.g., alkylating agents) produce delayed effects only
(Kihlman 1966). Exceptions are streptonigrin (SH) and S-ethoxycaffeine
(EOC), which produce effects very similar to X-rays. Increased oxygen
tension, which drastically increases the frequency of aberrations
induced by X-rays (Bacq and Alexander 1955, Evans 1962, Lea 1962) has
little or no effect with many chemical clastogens (Kihlman 1966).
Perhaps the most significant difference between X-rays and chemicals is
that the former produce aberrations more or less randomly in a
particular chromosome, while aberrations induced by the latter tend to
be localized in the heterochromatin (Nichols et al. 1964, Revell 1963).
Hypotheses on the Formation of Chromosomal Aberrations
The general or breakage-first hypothesis as described by Kihlman
(1966) proposes that the primary event produced by a clastogen is a
chromatid or chromosome break in a continuous interphase chromosome.
The ends at the point of breakage may rejoin to restore the original
configuration (restitution), they may remain open, or they may rejoin
with other open ends. Illegitimate fusion of ends from different
breaks results in sister-union or various types of exchanges.
According to the exchange hypothesis proposed by Revell (1955; see
also Revell 1963) the primary event is not a break, but some other kind
of lesion. The lesion may revert to normal or to another state
incapable of forming an exchange. If two primary events occur close
enough in space and time, an exchange initiation stage may follow.

36
During subsequent stages of chromosome development, the aberration is
transformed into a real chromatid exchange. Revell (1955) presented
data that are in good agreement with his hypothesis and Kihlman (1966)
came to the same conclusion based on more recent work. However, in
experiments especially designed to test the two hypotheses, Heddle and
Bodycote (1970) found that neither, as usually interpreted, is entirely
correct. Rather, they concluded that deletions are of two types,
according to mode of origin, but they were unable to identify the two
types morphologically.
Morphology and Mitotic Behavior of Aberrant Chromosomes
The following discussion is restricted to chromosomes with local
ized kinetochores.
"Gaps"
"Gaps," or "achromatic lesions" are Feulgen-negative regions of
variable size in chromatids (Evans 1963, Scheid and Traut 1970). In
Vicia they often resemble the normal nucleolar constriction. Gaps are
aberrations of the non-delayed type, since they can be induced by X-rays
in prophase nuclei (Evans 1963). The question of chromatid continuity
across the gap is important here, because gaps would have to be scored
as true breaks if the chromatid were really interrupted. Many
investigators (see Evans 1963 for references) have observed some sort of
material crossing gaps. Furthermore, the chromosome segment distal to
the gap moves normally and seems attached to the main body of the
chromosome at anaphase. If gaps were true breaks, they should give
rise to chromatid or chromosome aberrations at the second division
after exposure, but this is not the case (Evans 1963). Another

37
interesting point is that in Vicia and Trillium Feulgen-negative
regions similar in appearance to gaps can be induced by exposure to low
temperature. Electron microscopic examination of whole-mounted and
thin-sectioned chromosomes indicates that some gaps are true breaks,
while others are traversed by chromosomal fibrils (Brinkley and Shaw
1970). Similar results were obtained by Scheid and Traut (1971) with
the scanning electron microscope. They found that gaps represent
distinct "notches" which in some cases are traversed by two parallel
strands.
Acentric fragments
A break without reunion produces an acentric fragment (Evans 1962,
Kaufraann 1954, Kihlman 1966). If the break involves only one chromatid
the fragment is single; it consists of two "sister chromatids" if the
entire chromosome is broken. Such fragments are usually lost during
division, or they are included at random in daughter cells where they
form micronuclei (Humphrey and Brinkley 1969, Kihlman 1966, La Cour
1953). Carlson (1938) attributed apparent migration of acentric
fragments towards the poles at anaphase to more than chance movement,
but he made his observations on smear preparations. Bajer (1958) and
Bajer and Mole-Bajer (1963) recorded the behavior of fragments in
irradiated Haemanthus endosperm by time-lapse cinemicrography. The
majority of fragments were eliminated from the spindle, either at
prometaphase or during ana- and telophase. A few fragments, however,
moved in the spindle region in a more than random fashion, sometimes
from the equator to one pole and back. The authors attributed these
movements to the activity of neocentric fibers. In rat kangaroo cells

38
exposed to X-rays, Humphrey and Brinkley (1969) confirmed by electron
microscopic analysis of thin sections that fragments lack kinetochores.
Bridges
An exchange between centric portions of two broken chromatids or
chromosomes results in the formation of an anaphase bridge if the two
centromeres move to opposite poles (Evans 1962, Hair 1953, Kaufmann
1954, Kihlman 1966, Koller 1953). Subchromatid bridges (side-arm
bridges) arise from intrachromosomal exchanges. The unit of breakage
and exchange in these aberrations was generally assumed to be a half
chromatid (Heddle 1969, Kihlman 1966). Brinkley and Humphrey (1969)
examined X-ray-induced side-arm bridges in rat kangaroo cells with the
electron microscope and found that the diameter of these chromatid
connections was considerably less than that of a half-chromatid. The
authors conceded, however, that chromosome movement during anaphase
might have stretched the connections, which appeared to consist of
chromosomal fibers of the usual dimensions (see also Brinkley and Shaw
1970).
Anaphase bridges exist as direct connections between two
centromeres, or as interlocked dicentrics (Koller 1953). The thickness
and length of these bridges vary; they usually break during ana- or
telophase, but may persist into interphase (Hair 1953, Koller 1953).
Cinemicrographic studies by Bajer (1963, 1964) on the behavior of
dicentrics in Haeraanthus endosperm revealed interesting facts:
normally, dicentrics break abruptly, but in slightly unhealthy cells
they do not break, but form long, thin, sticky bridges. Interlocked
dicentrics show two kinds of behavior: they cut one through the other,

39
and the broken ends rejoin; or they uncoil and do not break at all.
Whether or not a bridge breaks or is stretched depends on its length
and the position of the kinetochores relative to the equator at meta
phase. Breakage at anaphase is due to the pulling force of the
chromosomal fibers; at telophase it is caused by the phragmoplast.
Sister chromatid bridges tend to break at apparently weak points which
are sometimes seen as constrictions. Movement after breakage is
scarcely faster than the initial anaphase speed, indicating there is no
accumulated tension in the pulling mechanism. Humphrey and Brinkley
(1969; see also Brinkley and Shaw 1970) observed apparent gaps within
anaphase bridges in rat kangaroo cells. These gaps were constrictions,
presumably caused by anaphase tension, within which three classes of
chromosomal fibrils could be seen.
Chromosome "stickiness"
When cells are irradiated in late prophase, "sticky" bridges can
be observed at the following anaphase (Carlson 1954). Chemical
clastogens, e.g., nitrogen mustard, also produce this effect (Koller
1953). Stickiness has been interpreted to be the consequence of
surface changes on the chromosomes, changes that make chromosomes
adhere to each other if they happen to come in contact (Carlson 1954,
Casarett 1968). Apparently, this aberration is reversible: if mitosis
is delayed after treatment, stickiness does not occur. The nature of
the sticky material remains obscure. Hsu et al. (1965) noted that in
Chinese hamster cells nucleolar material sometimes remains attached to
the ends of sister chromatids, forming apparent chromatin bridges at
anaphase.

40
Streptonigrin: A Chemical Clastogen
Streptonigrin (SN) is a metabolite of Streptomyces flocculus. It
has the formula C24^22S^4 anc* t*ie fHw:*-nS structure (Rao et al.
The drug behaves as a weak acid with quinoid properties. It
induces phage release in lysogenic bacteria (Levine and Borthwick 1963)
and initiates rapid breakdown of JS. coli DMA _in vivo (Radding 1963).
Inhibition by SN of DNA synthesis and of DNA-dependent KNA synthesis
was reported by Koschei et al. (1966) for a cell-free system, and by
Young and Hodas (1965) for tissue culture cells. Streptonigrin caused
single strand breaks in calf thymus DNA (Mizuno and Gilboe 1970). The
latter authors also found that SN preferentially binds to DNA during
the S phase.
Cohen et al. (1963) investigated the effect of SN on cultured
human leukocytes. The mitodepressive effect of SN appeared related to
concentration and length of exposure. The mitotic rate was signifi
cantly depressed in cells exposed for 36 hr to 0.01 and 0.1 /jg/ml SN,
but not in cultures treated with 0.001 pg/ml SN. The chromosome-

41
breaking effects of SN were of the delayed and non-delayed type.
Aberrations occurred in cells treated as late as 2 hr before fixation.
Among the aberrations observed were chromatid and isochromatid breaks,
acentric fragments, dicentric chromosomes, cleavage or severe attenua
tion of the centromere region, telomeric fusion of sister chromatids,
"stickiness," uncoiling of chromosomes, and severe fragmentation or
degeneration of the entire chromatin material. Chromosomal damage
appeared to be non-randomly distributed, chromosomes 19, 20, 21, 22,
and the Y being relatively stable. Cohen (1963) further studied the
non-randomness of aberrations produced by SN in chromosomes 1, 2, and
3 of cultured human leukocytes. While X-rays induced random breaks, SN
preferentially affected the pericentric regions of chromosomes 1 and 2,
as well as the area of the secondary constriction of chromosome 1.
Breaks in chromosome 3, although fewer in number, were distributed at
random. The telomere regions of all three chromosomes, and the short
arm of chromosome 2 appeared relatively resistant to SN.
Kihlman (1966) concluded from the work by Cohen et al. (1963) that
SN is able to break chromosomes during G^. A similar finding had been
reported for root tip cells of Vicia faba (Kihlman 1964). Exposure of
cells to 2-5 ;jg/ml SN for 1 hr produced subchromatid (cells in early
prophase), chromatid (cells in and S), and, possibly, chromosome
exchanges (cells in G^). However, Puck (1964) reported that SN does
not affect mammalian cells after completion of DNA synthesis. In this
respect SN may be similar to S-ethoxycaffeine (EOC) and other drugs
which have non-delayed effects in plant cells, and delayed effects, or
hardly any effects at all, in mammalian tissue culture cells (see
Kihlman 1966).

42
Jagiello (1967) described chromosomal aberrations induced by SN
O'. 3 o' a o .o
in mouse eggs. _In vitro, metaphase I chromosomes of ova treated with
1.0 ig/ml SN were agglutinated beyond recognition; 0.1 ;jg/ral SN induced
achromatic gaps and breaks in approximately 407. of the ova. Severe
chromosome damage also occurred in ova of mice, to which SN had been
administered subcutaneously.
Streptonigrin is still used in _in vitro and clinical studies on
tumor chemotherapy (Carter et al. 1968, Oleson et al. 1961). Like many
other antitumor drugs it is more active against lymphoma than against
solid tumors.
Colcemid; A Spindle Poison
Colchicine is an alkaloid isolated from Colchicum autumnale, whose
anti-mitotic effect has been under investigation for over 30 years (see
reviews by Biesele 1958, Deysson 1968, Dustin 1963, Eigsti and Dustin
1955, Gelfant 1963, Kihlman 1966). The structural formulas of
colchicine (from Kihlman 1966) and its synthetic analog colcemid
(demecolcine, N-deacetyl-N-methylcolchicine; from Schar et al. 1954)
are given below.
Colchicine
Colcemid

43
The effect of colcemid on plant and animal cells is essentially
the same as that of colchicine (see Gelfant 1963), but the former is
less toxic and more efficient in animal tissue (Schar et al. 1954).
Levan (193£) coined the term "c-mitosis" for the peculiar
morphological changes in dividing Allium cells under the influence of
colchicine. Chromosomes in c-mitosis are scattered in the cytoplasm,
sister chromatids being held together in the centromere region. At
"c-anaphase," sister chromatids fall apart and at "c-telophase,"
because of the absence of a mitotic spindle, all the chromosomes are
included in one polyploid restitution nucleus (see also the cinemicro-
graphic analysis of c-mitosis in Haenanthus endosperm by Mol-Bajer
1958). In animal cells the division of the centromere region is
delayed until "c-telophase," i.e., "c-anaphase" is omitted (Levan
1954).
Besides Levan's (1938) classical scattered metaphase, other
chromosomal arrangements, such as "star-mitosis" and "clumped" or
"ball" metaphase, can be observed in both animal and plant cells (e.g.,
Gaulden and Carlson 1951, Deysson 1968). The effect depends on the
concentration and the time of exposure.
For many years it was commonly accepted that colchicine and its
analogs arrest dividing cells at metaphase. Sentein (1961) objected to
calling colchicine purely a metaphase poison, because he also found
arrested prophases, anaphases, and telophases in his material.
Brinkley et al. (1967) demonstrated that in Chinese hamster cells
arrested in mitosis by treatment xiith 0.06 ;ig/ml colcemid the two pairs
of centrioles are surrounded by the chromosomes in a configuration
different from a typical metaphase. After reversal of the inhibition,

44
centriole pairs move to opposite poles and a normal metaphase plate is
established prior to anaphase segregation. Furthermore, at the
concentration used some MT were still present, notably on the
kinetochores facing the centrioles (see also Brinkley and Stubblefield
1966). Jokelainen (1968) furnished more evidence that the precise
mitotic stage at which cells are arrested may depend on the concentra
tion applied. He found no typical metaphases in fetal rat kidney tis
sue after treatment of pregnant rats with 0.12 mg/kg colchicine, but
metaphases were present in fetuses of rats treated with 0.08 mg/kg.
The conclusion, drawn from birefringence studies (Inou 1952) and
the behavior of chromosomes in colchicine-treated cells (Levan 1938,
Mol-Bajer 1958), that this chemical disorganizes spindle MT, has
received support from biochemical studies by Borisy and Taylor (1967a,
b). They found that colchicine binds to a 6S protein from isolated
mitotic apparatus. This protein showed good correlation with the
occurrence of MT and was considered to be a subunit of MT.
Statement of Purpose
The objective of this investigation was threefold: (1) To fill
gaps in our knowledge of the structural changes accompanying the build
up of the chromatic and achromatic apparatus during normal mitosis in
animal cells. Particular attention was devoted to controversial or
less well-studied aspects of mitosis, such as kinetochores and the
nuclear envelope. (2) To study possible structural alterations in the
mitotic apparatus produced by the agents applied or by the presence of
aberrant chromosomes (e.g., dicentrics). (3) To elucidate the fine
structure of aberrant chromosomes, with particular reference to
kinetochores and chromosome architecture.

MATERIALS AND METHODS
Cell Culture
The PtK^ cell line was initiated in 1961 by Kirsten H. Walen
(Walen and Brown 1962) from a kidney of an adult male rat kangaroo,
Potorous tridactylis (Harsupialia). The male karyotype consists of ten
autosomes, one X chromosome bearing the nucleolus organizer, and two Y
chromosomes (Figure 1. Shaw and Krooth 1964, Walen and Brown 1962).
Fig. 1. Karyotype of the male rat kangaroo (Potorous tridactylis).
Note the heterochromatic region near the centromere on the X chromo
some (nucleolus organizer). Drawn after a micrograph by Shaw and
Krooth (1964).
The Pt^ line, however, is aneuploid, the number of chromosomes in the
reference stock varying from 11 to 14 (Anonymous 1967). An extra, long
subtelocentric chromosome is present in the majority of cells.
45

46
According to Walen (1965) the generation time in Eagle's medium with 4%
fetal calf serum (FCS) is 28-32 hr. The low number and the individual
ity of chromosomes make this line particularly suitable for cytogenetic
and labeling studies.
I obtained an ampule of frozen cells from the American Type
Culture Collection and started a culture in April, 1970. After
preliminary experiments this culture was given up in September, 1970,
because of a suspected contamination by a microorganism (mycoplasma or
fungus). A new culture with fresh cells was initiated in October, 1970.
Only cells from this culture were used in the experiments described.
Cells of the stock culture were grown in square glass bottles in
Eagle's minimum essential medium (MEM; GIBCO), supplemented with 10%
FCS (GIBCO). Cells were harvested with trypsin-versene at weekly
intervals and subcultured at appropriate dilutions. Fresh complete
medium (MEM with 107. FCS) was supplied once between transfers. Cells
to be used for experiments and controls were seeded in Falcon 30-ml
flasks and left to attach overnight. Cultures were maintained at 37C,
but handling was carried out at room temperature.
Chemical and Physical Treatments
Streptonigrin
Streptonigrin (Pfizer and Co., Inc.) was obtained from Cancer
Chemotherapy, NCI, NIH, Bethesda, Md. A stock solution of 250 pg/ml
was prepared shortly before use by dissolving the powder in the diluent
supplied. The stock solution was diluted volumetrically in distilled
water, pH 8.0, to a concentration of 0.25 jjg/ml. After sterilization
with a millipore filter, the final concentrations of 0.05 pg/ml and

0.01 pg/ral were prepared under sterile conditions by dilution with MEM.
No serum was added to avoid possible binding of SN by serum proteins.
Cell monolayers in Falcon 30-ml flasks were rinsed once with MEM
before the drug was added. Control cells were treated in the same
manner, except that they were exposed to MEM alone instead of the drug.
The cells were incubated for 4 hr at 37C. The flasks were then rinsed
twice with complete medium and the cells were left to recover in the
incubator in a third change of fresh complete medium. Approximately 24
hr later the medium was changed once more. After a total recovery
period of 48 hr the cells were fixed _in situ according to the standard
schedule (see below).
Colcemid
A stock solution of 0.5 pg/ml was prepared by dissolving crystal
line colcemid (CIBA) in distilled water, pH 6.4. This solution was
sterilized with a millipore filter and further dilutions were prepared
with sterile complete medium. Treatment A was 0.05 ;ig/ml colcemid for
2 hr at 37C, followed by two rinses with complete medium and recovery
for 1 hr at 37C in a third change of medium. Treatment B was 0.25
^ig/ml colcemid for 15 min at 37C without a recovery period. Control
cells were left undisturbed and were fixed at the same time as treated
cells.
Cold
Flasks were seeded as usual and the cells left to attach overnight
at 37C. The medium in the experimental flask was then poured off and
replaced by complete medium of 0-4 C. The flask was immediately placed
in a pan with ice water so that its lower portion was immersed. After

48
1 hr exposure the medium was replaced by cold glutaraldehyde of the
usual strength and buffered as usual (see fixation schedule). Initial
fixation was 15 min in the cold, after which the flask was brought to
room temperature and the standard fixation schedule was followed from
here. One control consisted of cells kept at 37C and fixed cold as
above. The second control consisted of cells not exposed to cold and
fixed at room temperature.
Fixation and Embedding
The standard fixation schedule, modified from Brinkley et al.
(1967), was as follows:
(1) Decant the culture medium and add 3.17. glutaraldehyde (GA)
buffered with Millonig's phosphate buffer without sucrose
(Millonig 1961), pH 7.3. After a few minutes add a fresh
change of GA. The cells are fixed for a total of 1 hr at
room temperature.
(2) Rinse with two changes of buffer, 10 min each, at room
temperature.
(3) Postfix for 1 hr at room temperature in similarly buffered 27.
osmium tetroxide (OsO^).
(4) Dehydrate in 257. and 507. ethanol, 10 min each. Prestain in
cold 27. uranyl acetate in 707. ethanol for 2 hr to overnight.
Rinse with two changes of cold 757. ethanol, 10 min each. Ten
min 90% ethanol; the cold solution is added and the flasks
then brought to room temperature. Subsequent steps are all
carried out at room temperature; three changes, 10 min each,
of 907. hydroxypropyl methacrylate (HPMA); 15 min each of 957.
and 977. HPMA

49
Enbed in Luft's Epon (Luft 1961) as follows. A ratio of 1 part of
mixture A to 2 parts of mixture B gave the best results.
2 parts HPA : 1 part Epon for 15 min
Equal parts of HPMA and Epon for 15 min
1 part IIPIA : 2 parts Epon for 30 min
2 changes of pure Epon, 30 min each
Add fresh pure Epon and bum holes in the top of the culture flasks
with a hot glass rod. Drain off excess Epon until a layer about the
thickness of a glass slide is left. Leave overnight at room temperature
then transfer to 60C for 24 hr or longer.
Preparation of Cells for Light and Electron Microscopy
The basic procedure was adopted from Brinkley et al. (1967).
After curing, the bottom of the flask with the adhering Epon wafer is
cut out. Wafer and flask bottom are separated by alternately cooling
the sandwich in liquid N_ and thawing in tap water. Once separation
has started at the edge, the Epon wafer can be snapped loose.
For light microscopy (scoring of aberrations, determination of
mitotic indices) wafers are placed cell layer up on the stage of a
phase contrast microscope. Detailed examination of cells is possible
by using oil immersion objectives.
The area around cells selected for thin sectioning is marked with
a sharp needle under low power. Several high magnification pictures at
different levels of focus are then taken. A disk with the marked area
is cut out with a cork borer of suitable diameter. The piece of Epon
is roughly trimmed and glued to the tapered end of a plastic peg, cell
layer up. Fine trimming to the boundaries of the trapezoidal or

50
rectangular area containing the cell or cells is done after curing of
the resin glue.
As a rule, blocks were sectioned in a plane parallel to the cell
monolayer (horizontal plane; see Figure 2). For special sections
(para-sagittal = vertical; para-equatorial = orthogonal to the spindle
axis; see Figure 2) blocks were mounted differently. A scratch was
made on the surface of the Epon wafer, under low power, either parallel
to the spindle axis or at a right angle to it and as close to the
selected cell as possible. A disk with the marked cell was cut out,
roughly trimmed, and mounted in the clamp chuck of an ultramicrotome.
By cutting thick sections with a glass knife the block could be faced
to the level of the scratch or even closer to the cell, when periodical
checking under a microscope was possible. If necessary, the angle of
the face relative to the spindle axis could still be corrected.
Finally, the block was mounted on a peg with the face up.
Serial sections were cut with a diamond knife on a Porter-Blum
MT-2 microtome and picked up on formvar-coated, carbon-stabilized
rectangular hole grids (LKB Instruments, Inc.). Sections were
poststained with uranyl acetate (Watson 1958) and lead citrate
(Reynolds 1963) and examined in a Hitachi HU 11-E electron microscope
with a 50-nm objective aperture and operated at 75 kV.

51
Fig. 2. The three planes in which blocks were sectioned. C: Cell
in mitosis; the spindle axis is assumed to be along a line between the
two pointed ends. HO: Horizontal plane. PE: Para-equatorial plane.
PS: Para-sagittal plane.

RESULTS AND OBSERVATIONS
Normal Mitosis
This section is based on light and electron microscopic observa
tions on untreated cells, either used as controls for the various
treatments, or fixed directly from the stock culture.
Interphase
Although interphase is not a mitotic stage, it makes sense to
describe the structure and spatial orientation in the interphase cell
of those components and organelles that take part in the formation and
function of the mitotic apparatus.
Figure 3 shows a grazing section of an interphase nucleus. Pore-
annulus complexes of the nuclear envelope are cut at different levels.
They appear as circles at the level of the envelope, and as circles
surrounded by a "halo" at the level of the underlying chromatin.
Central dense granules can be seen in some of the profiles. Where the
section passes below the complexes, their position is indicated by
achromatic holes in the chromatin. Whorls of polyribosomes are visible
on oblique sections of the nuclear envelope. In the cytoplasm, two
centrioles are present, approximately at a right angle to each other.
Obliquely sectioned membrane elements, probably of a Golgi complex, can
be seen in their vicinity. Microtubules, though more numerous around
the centrioles, pass in various directions, without apparent specific
orientation to the latter. Details of centriolar structure at
different levels are shown in the four serial sections in Figure 4.
52

53
Osmiophilic, granular or fibrillar material surrounds the centriole at
one end (Figure 4a). The short rays radiating from the triplets
(Figure 4b) give the whole structure the appearance of a pinwheel.
Such images are rare. Dense, spherical particles with a lighter core
and aura (Figure 4c), often quite numerous, are exclusively associated
with centrioles and are present during all the stages of the cell cycle.
An intracentriolar vesicle, shown in an approximately median section in
Figure 4d, is often found in the lumen of centrioles. Satellites and
microtubules can be seen in all four sections (Figures 4a-d).
The appearance of the interphase nucleus in transverse sections
varies. Three representative nuclei are shown in Figure 5. The
chromatin is either dispersed (Figure 5a), or it includes heterochro-
matic chromocenters to a variable extent (Figures 5b and 5c). Nuclei
like the one in Figure 5c are very likely in late cytokinesis or very
early prophase. Of 34 interphase nuclei examined, 12 had dispersed
chromatin, the remainder contained chromocenters of variable size and
extent. At higher magnification, chromatin fibers, approximately 130 A
and 250 A in diameter, can be seen cut at various angles (Figure 6). A
very dense core, 70-80 A in diameter, appears in a few cross sections
of 250 A fibers. Nucleoli are large and prominent in interphase cells
(Figure 5). Granular and fibrillar components are mixed (Figure 7).
Nucleolus-associated chromatin is found within the nucleolus and at its
periphery (Figures 5 and 7). Intranuclear vesicles occur in some
nuclei (Figure 5a). The wall of these vesicles is finely fibrillar and
they often, perhaps always, contain granules or particles 250-400 A in
diameter. However, densely packed granules of similar size and
appearance also occur free in the karyoplasm.

54
Prophase
The centrioles duplicate at the onset of prophase. Therefore, two
parent-daughter centriole pairs are found in early prophase cells
(Figure 8a). Four serial sections of the sane centrioles (Figures
8b-e) reveal that each daughter centriole is closely associated with
its parent at approximately a right angle. Daughter centrioles appear
shorter than nature centrioles (compare Figure 8b with Figure 3), but
have the same diameter (approximately 2,200 A). At this stage rela
tively few MT converge on, or radiate from, each parent centriole. An
early stage of centriole migration is illustrated in Figures 9a and 9b.
One of the two centrioles shown (C^) is possibly a daughter centriole.
It is cut almost perfectly at a right angle to its axis. The structure
and arrangement of the tubular triplets is particularly clear (Figures
9c-e). The cartwheel at the proximal end appears in three adjacent
sections (Figures 9b-d), but hub and spokes are most distinct in
Figure 9c. The triplets seem to be embedded in amorphous osmiophilic
material. Bars, approximately 80 by 480 A, possibly cross sections of
plates, appear between the triplets (Figure 9e). Numerous satellites
are present in the general area of the two centriole pairs (Figures 9a
and 9b). In many, perhaps all, mid-prophase cells the centrioles lie
in an invagination or pocket of the nuclear envelope (e.g., Figure
11a). At this stage numerous ITT are present, forming an "aster"
around the centrioles. The radial arrangement of these MT also imposes
radial orientation on other organelles, notably mitochordria (Figure
lib). Migration of centrioles relative to nuclear changes (chromosome
condensation, fragmentation of the NE) varies considerably, so that by

55
the end of prophase the two pairs may have moved a short distance only,
or they may lie at opposite poles (see Figures 18a and 19).
Progressive condensation of chromatin is indicated by the
appearance, during very early prophase, of large heterochromatic
patches (compare Figure 8a with Figure 5). Discrete chromosomes are
present in mid-prophase cells (Figures 10 and 11a). Their "mottled"
appearance (see also Figure 13) is suggestive of incomplete condensa
tion. During condensation many, possibly all the chromosomes are
attached to the nuclear envelope along their entire length or with
their telomeres (Figures 11a and 13). In transverse sections of nuclei
in early prophase the chromosomes appear to be attached by "stalks"
(Figure 13), but grazing sections reveal that this appearance is due to
achromatic "holes," enlarged compared to interphase (compare Figure 12
with Figure 3). In late prophase, however, strand-like connections
between chromosomes and the nuclear envelope are real (Figure 14).
No kinetochores are discernible in early prophase (Figure 8a).
In sections of nuclei in mid- and late prophase kinetochores appear as
roughly circular patches of finely fibrillar material in slight
constrictions of chromosomes (Figures 11a, 15, and 17). Serial
sections revealed that these kinetochores are globular and 5,000-8,000
A in diameter. I cannot state with certainty that the diameter
decreases with advancing prophase. The doubleness of the chromosomes
is evident in sections showing both sister kinetochores (Figures 15 and
17). The lesser electron density of the kinetochores compared to the
chromosomes is very distinct (Figures 10, 11a, 15-17). Slightly more
opaque kinetochore granules at the surface of the sister chromatids
(Figure 17) are very rarely seen.

56
The nucleolus, still intact in very early prophase (Figure 8a),
fragments into several masses of granular material (Figures 10 and
11a), some of which are apparently associated with chromosomes (Figure
11a).
The nuclear envelope remains intact until the end of prophase, but
the pore-annulus complexes become more fuzzy (Figure 12). Polyribo
somes are found on the nuclear envelope throughout prophase (e.g.,
Figure 11a).
Prometaphase
The breakdown of the nuclear envelope, indicating the transition
from prophase to prometaphase, is gradual, but structural changes
involving MT and lcinetochores are more striking. Fragmentation of the
nuclear envelope always begins nearest the centrioles (Figures 18a and
19). Fragments are undulated or form vesicles. The telomeres of
chromosomes at the periphery of the "nucleus" are often trapped in
compartments formed by undulated, still intact portions of the nuclear
envelope (Figure 18b). With progressive development of the spindle
apparatus, envelope fragments become smaller and scarcer, but some
persist at the periphery of the spindle until late prometaphase
(Figures 20, 26, and 29). Disappearing central granules indicate early
stages of the breakdown of pore-annulus complexes (Figures 21a and
21b). The entire complexes on fragments of the nuclear envelope
become fuzzy and disappear by mid-prometaphase (Figure 21). Polyribo
somes can be found on clearly identifiable fragments until late pro
metaphase (Figures 21, 26, and 29).
Chromosome condensation continues throughout prometaphase (Figures
18-20, 26-30). The chromosomes detach from the nuclear envelope as the

57
latter fragments, and they gradually lose their "mottled" appearance.
Most kinetochores in very early prometaphase still resemble prophase
kinetochores (Figure 22a), but some are differentiating into more com
plex structures. For example, the sister kinetochores of the chromo
some in Figure 22b, although apparently not attached to MT, exhibit a
slightly denser band within the fibrillar matrix. Such internal struc
tures may be more common than it appears, but due to their relative
indistinctness they would not appear in ever so slightly oblique sec
tions. A rather unusual case is illustrated in Figures 22c and 22d.
This chromosome of the cell shown in Figure 18 was situated in the area
distant from the centrioles, i.e., near the intact portion of the
nuclear envelope. Several obliquely sectioned MT lie outside the
envelope opposite the "outer" kinetochore, which also contains a short
band. Serial sections revealed no MT inside the nuclear envelope. The
other sister kinetochore, facing in the direction of the distant
centrioles, resembles a typical prophase kinetochore.
As prometaphase progresses, kinetochores become more variable in
appearance. Structural differentiation depends in each individual case
upon the position and orientation of the chromosome relative to the
spindle poles. Sister kinetochores of chromosomes near a pole are
dissimilar (Figures 23 and 24). As a rule, the kinetochore facing the
near pole is attached to MT and consists of moderately opaque, finely
fibrillar material (K, K^, in Figure 20; K in Figures 23 and 24). A
banding pattern is discernible in some of these kinetochores (Figures
23a and 23b), which are quite often also stretched (K, K in Figure
23a; in Figures 24b and 24c). This makes it impossible to determine
where the MT end. In contrast, the kinetochore facing the distant pole

58
resembles typical prophase kinetochores (e.g., in Figure 24), or it
consists of a dense band embedded in the fibrillar matrix (K^ in
Figures 23a and 23c). The band can be seen in several adjacent sec
tions, which suggests it represents a transverse section of a flat or
convex plate approximately 300 A thick and 3,500 A in diameter. In
each case few or no MT are associated with such a kinetochore.
Sister kinetochores of chromosomes lying near the future equator
of the spindle are similar if both are unobstructed, i.e., if there is
no nearby obstacle (such as another chromosome) between them and the
respective pole (Figure 25). Usually, these kinetochores are more or
less stretched, with or without bands, and attached to MT. If one of
the sister kinetochores is obstructed by a neighbor chromosome lying
very close, it resembles prophase kinetochores (not shorn, but similar
to in Figure 24a), while the other kinetochore resembles the
unobstructed kinetochores described above.
Microtubules are found in the "nucleus" as soon as the nuclear
envelope breaks down (Figures 18, 22, and 23). At first they are more
numerous in the vicinity of the centrioles (Figure 18), but as the lat
ter take up their position at opposite poles, MT are abundant in the
center of the spindle (Figure 20). Surprisingly, most of the MT are
associated with kinetochores and few, if any, continuous tubules seem
to be present. However, a more careful, quantitative analysis would be
necessary to establish this with certainty.
Remnants of the nucleolus disappear completely and the nucleolar
organizer appears on the X chromosome (Figures 19 and 25).
Figures 26-30 illustrate further progression of prometaphase.
Fragments of the nuclear envelope, still found between and around some

59
chromosomes during mid-prometaphase (Figure 26), become smaller and
move to the periphery of the spindle in late prometaphase (Figures 29
and 30a). Nuclear pore complexes have disappeared approximately at the
stage shown in Figure 26, but some polyribosomes are still present on
the fragments of the nuclear envelope. Peripheral segments of double
membrane in late prometaphasc cells have ribosomes on both faces, thus
resembling rough EE (Figure 29).
As the spindle develops, mitochondria, EE, and other large cellu
lar organelles and components are excluded from its area (compare Fig
ure 20 with Figure 29), but ribosomes, mostly monosomes, are abundant.
Bundles of ITT in the center of the spindle run more or less straight
along the pole-to-pole axis, while peripheral bundles are slightly
arched (Figures 26 and 29). An exceptional case is shown in Figure 32,
where rather disoriented MT curve sharply towards the centriole.
Angular configurations of kinetochore Iff, as shown in Figure 20, no
longer occur. Still, most of the bundles of MT are associated with
kinetochores (Figures 26-29, and 30a). It is very difficult to deter
mine whether the MT of other bundles (e.g., Figure 28) are interpolar,
or whether they are associated with kinetochores not visible in the
same section. Quite frequently, a bundle of MT passes near the kineto
chores of a maloriented chromosome (e.g., Figure 39). Transchromosomal
MT are outstanding in rather thick sections (Figure 30b). Skew MT, as
well as "wavy" MT, occur quite frequently, but they always make up a
small proportion of the total number of spindle MT (Figures 27, 28, 30a,
and 33).
Orientation of the centrioles relative to the axis of the fully
formed spindle is variable (figure 29). This has also been confirmed
in cells in meta- and anaphase.

60
The chromosomes condense further until they appear as solid, very
electron dense rods or blocks (compare Figure 26 with Figure 29).
Sister chromatids are still tightly joined, but their individuality is
apparent in transverse or peripheral longitudinal sections of
chromosomes (Figures 27, 29, 30a, 31, and 33), or in situations where
the sister kinetochores are stretched (Figure 34). In the latter case,
one or several achromatic holes appear in the kinetochore region.
The diversity of kinetochore structure is still remarkable during
mid- and late prometaphase (Figures 26-29, 30a, 31, 33-41). At a glance
there seems to be no general pattern of differentiation, but a careful
comparison of the structure of sister kinetochores with the position of
the chromosomes in the spindle gives a better insight. First, we can
distinguish between chromosomes lying on or near the equator of the
spindle (Ch^ and Ch^ in Figure 27; Ch^ in Figure 28; the chromosomes in
Figures 30a and 33-36), and chromosomes lying closer to one pole
(malpositioned chromosomes; Figure 26; Ch^ in Figure 27; Ch^ in Figure
28; Figures 29 and 37-40). Generally speaking, more chromosomes occupy
an equatorial position in late than in early prometaphase (Figures 20,
26, and 29). The sister kinetochores of equatorial chromosomes are
similar if both are unobstructed, i.e., if no major obstacle lies along
a line from the kinetochore to the pole (Figures 27, 30a, 33, 34, and
36); they are dissimilar if one is unobstructed, the other obstructed
by a neighbor chromosome (Figures 33 and 35). Unobstructed kinetochores
are irregularly shaped and fuzzy (Figures 28 and 36), fuzzy cones with
a trace of bands (similar to in Figure 38), or distinctly triple-
banded (Figures 27 and 33-35). It appears that the clarity of the
three bands increases with advancing prometaphase. Unobstructed

61
kinetochores are quite often stretched (Figures 27, 28, 31, 33, and 36),
and they are always attached to bundles of 11T.
Obstructed kinetochores of equatorial chromosomes appear as patches
of finely fibrillar matrix within which a denser band or patch can be
seen (Figures 33, 35, and 41). Serial sections make it clear that the
bands do not represent fibers, but sections of oddly shaped three-
dimensional structures (e.g., Figures 35 and 41). The width of the
bands varies from 250-400 A. In appearance and dimensions they are
comparable to the outer layer of triple-banded kinetochores, except
that they often exhibit a partly double-banded, partly beaded sub
structure (Figure 41). Furthermore, obstructed kinetochores are always
attached to very few, if any, MT.
Sister kinetochores of maloriented chromosomes most commonly are
also dissimilar. As a rule, the kinetochore facing the near pole
resembles the unobstructed kinetochores of equatorial chromosomes
(Figures 27, 37, and 38), while the kinetochore facing the far pole
resembles obstructed kinetochores (Figures 27, 37, and 38). Figure 40
illustrates a rare exception to this rule. On both these maloriented
chromosomes the kinetochore oriented towards the near pole is compact,
and attached to 1IT, while the other kinetochore is fuzzy and stretched,
but also attached to IIT.
Figure 39 illustrates another rare case. The sister kinetochores
of this maloriented chromosome are very similar in every respect; both
are relatively undifferentiated masses of matrix with only a trace of a
band, 'lost remarkable is that the matrix seems to cover almost the
entire small side of the chromosome (Figure 39c).

62
The difference in electron density between the kinetochore matrix
and its band on one hand, and the chromosome proper on the other hand,
is obvious in many of the figures cited, but particularly in grazing
sections of kinetochores (e.g., Figures 33b, 35c, and 36). Interest
ing, but very rarely seen, are chromatin "strands" between stretched
sister kinetochores (Figures 31 and 36). It is difficult to determine,
even at high magnification, if the fibers in these strands are finer
than normal chromosomal fibers, or if they are more densely packed,
and arranged more or less in parallel.
Metaphase
Cells in full metaphase are quite rare, probably because this
stage is of very short duration. Most cells judged to be in metaphase,
based on light microscopy, turn out, upon examination of thin sections,
to be either in very late prometaphase or very early anaphase.
Chromosomes at metaphase are aligned on the equator of the
spindle, at least with their kinetochore region (Figure 42).
Typically, the non-aligned telomeric portions of the long chromosomes
extend beyond the periphery of the spindle into the cytoplasm (insets
of Figures 42-44). All the kinetochores in normal metaphase cells are
triple-banded (Figures 42 and 43). Bundles of kinetochore MT converge
towards the poles (Figure 42). Within a bundle the MT are more or less
parallel, but a few non-kinetochore MT slant across. Wavy MT occur
mainly between chromosomes. Overall, the paucity of non-kinetochore MT
is remarkable.
Figure 43 represents a para-sagittal section. The plane of
sectioning was not precisely at a right angle to the chromosomes,

63
therefore only one of the two sister kinetochores can be seen. Kineto-
chore profiles are virtually identical with those seen in horizontal
sections (compare Figure 43 with Figure 42). Sister chromatids are
separated by grooves relatively devoid of ribosomes and ground sub
stance. However, the separation, which actually indicates late meta
phase to very early anaphase (see also inset of Figure 43), is in
complete, for large "bridges" still connect the sister chromatids.
In para-equatorial sections the metaphase plate appears as shown
in Figure 44. Three kinetochores with associated MT can be seen.
Other MT occur singly or in small clusters, or they are arranged in
bundles which may belong to kinetochores not included in this section.
Microtubules penetrating chromosomes are surrounded by a clear halo.
Anaphase
Sister chromatids, now more properly called daughter chromosomes,
separate from each other at the onset of anaphase. In the light micro
scope they still appear as parallel rods (Figure 45, inset). Because
the chromosomes are somewhat frayed, their separation at very early
anaphase (Figure 45) is less distinct in thin sections, except for the
kinetochore region. The latter is spaced farther apart than the
telomere region (Figures 45 and 46), and this trailing of the telomeres
is more pronounced during later anaphase (Figure 47; inset of Figure
50; Figure 51a). The daughter chromosomes lose their individuality
during late anaphase and each set forms a large mass of densely packed
chromatin near the respective pole (Figure 52). The nucleolus
organizer (NO) is enclosed in this mass.

64
A striking phenomenon is illustrated in Figures 46-48. Daughter
chromosomes in very early and early anaphase are connected by electron
dense strands. First believed to be an aberration, it was found in six
of seven early anaphase cells. The chromosomes of the seventh cell had
short strands of similar appearance extending from the rather widely
separated kinetochore regions into the interzone. The same was
observed in one mid-anaphase cell. In the latter two cases the strands
were extremely tapered towards the interzone, giving the impression of
connections gradually drawn out and finally ruptured. In the majority
of early anaphase cells the strands connect the kinetochore regions of
daughter chromosomes. I have counted as many as six strands in serial
sections. However, in the cell shown in Figure 47, and also another
very similar anaphase, a few strands connected the mid-region of
daughter chromosomes.
The diameter of these strands varies from approximately 400 A to
900 A. Short strands are usually thicker, and long strands very
frequently are thinnest approximately midway between daughter chromo
somes, while they are thicker at their base (i.e., point of attachment;
Figures 46 and 47). The staining properties of the strands are
identical with those of chromosomes, and the structure is that of fine
and/or densely packed fibers.
Kinetochores vary in appearance from more or less straight to
convex, "stalked," or angular (Figures 46-49). The three layers are
very distinct except in sections that cut a kinetochore obliquely or
peripherally (Figures 46 and 47). In para-equatorial sections the
kinetochores appear as circles of moderate electron density, within
which cross sections of MT-can be seen (Figure 50). In late anaphase

65
the triple-layered character of the kinetochores is less clear, but
still recognizable (Figures 51b and 52). Most kinetochores in very
late anaphase cells appear in depressions on the poleward face of the
chromatin mass (Figure 53). The less dense band in Figure 53 is
approximately 500 A wide and set off from the chromatin by a 250 A wide
clear band. The chromatin immediately underlying the kinetochore is
denser and less obviously fibrillar than the remainder of the
chromosomal mass. Remarkable is the decreasing number and degree of
organization of kinetochore MT in late anaphase (compare Figures 46-49
with Figures 51-53).
Spindle elongation is the rule in PtK^ cells (Figure 51a). The
interzone is still free of cytoplasmic organelles even after the
chromosomes have almost reached the poles. Pieces of double membrane,
some with ribosomes, are found at the periphery of the spindle area,
particularly around the two sets of chromosomes (Figure 51a).
Interzonal MT are scarce. Mitochondria, vesicles, and ER invade the
interzone in very late anaphase (Figures 52 and 55b). The cytoplasm
begins to constrict in the equatorial region (Figure 52, inset), where
stem bodies appear. The latter consist of amorphous, osmiophilic
material, within which MT are closely packed (Figures 54 and 55a). It
is difficult to determine from these sections whether the MT terminate
in the stem bodies or beyond. Both possibilities are likely. One MT
in Figure 55a clearly passes through the stem body. Its total length
visible in this section was 3 i.
It is possible that pieces of double membrane as seen in Figure
51a become involved in the reconstitution of the nuclear envelope.
More typically, however, small cistemae and vesicles appear on the

66
surface of the chromatin masses in late anaphase (Figure 52). Oblique
sections of such membranes reveal the presence of nuclear pore
complexes. Some of the cistemae obviously are, or have originated
from, RER as indicated by the presence of ribosomes (Figure 52). The
location of the membrane pieces varies, depending on the level of the
section. Usually, the majority of the membrane cistemae and vesicles
is apposed to the lateral faces of the chrriatin masses.
Figure 56 shows a pair of centrioles in early anaphase. As
during prometa- and metaphase, the moderately osmiophilic, amorphous
material surrounds that centriole on which the MT more or less
converge. Direct connections between MT and any part of the centrioles
are never observed.
Telophase and Cytokinesis
Reconstruction of the nuclear envelope continues during telophase,
apparently by coalescence of the small vesicles and cistemae seen in
anaphase cells. Extensively reformed NE first appears on the lateral
faces of the chromatin mass, as well as on the polar face, except
directly opposite the centrioles (Figures 57 and 58). Nuclear pore
complexes, complete with central granules, are present on chromatin-
associated membranes irrespective of size and location (e.g., Figure
58). The spacing between the two membranes of the NE remains irregular
until late telophase (Figures 57-60). Quite frequently, pieces of NE
are trapped within the reforming nucleus, probably giving rise to the
deeply invaginated pockets seen in many interphase cells.
Figures 60 and 61 illustrate progressive decondensation of the
chromatin. This process begins even before the NE is completely

67
restored, but large heterochromatic patches persist until interphase
(late stages of cytokinesis).
The less dense component of kinetochores, still present during
early telophase (Figure 57), has disappeared by the time the NE is
completely reconstructed (Figures 59 and 60). The inner, dense layer
becomes an extremely osmiophilic patch on the inner membrane of the NE
(Figures 59 and 60). Most commonly, the entire NE is indented at
these sites, and quite frequently the inner membrane with the dense
patch is deeply invaginated (Figure 60). Microtubules can be seen
extending poleward from these kinetochores, but the nature of this
association cannot be determined precisely.
During cytokinesis a cleavage furrow in the former equatorial
region separates the two daughter cells, except for a stem, whose
length and diameter vary depending on the distance between daughter
cells (Figures 61 and 62). The stem bodies seen in late anaphase and
early telophase have fused to form the midbody, an irregular band of
osmiophilic material in the stem. Numerous MT from both daughter cells
converge in the midbody, but the dense material obscures details in
these paraxial sections. Normally, the midbody remains between the
daughter cells even after these have completely separated and moved far
apart from each other. Eventually, it seems to be lost. On rare
occasions the midbody is included in the peripheral cytoplasm of one of
the daughter cells, but this is most likely an abnormal condition.
C-Mitosis
Treatment A (0.05 jig/ral colcemid for 2 hr 1 hr recovery)
produces an accumulation of cells with "scattered1' metaphases (Figures

68
63 and 64). C-mitotic cells are much less compressed along the axis
vertical to the growth surface than normal mitotic cells. The extreme
is illustrated by the near-spherical cell in Figure 63. Differences
among c-mitotic cells are evident in the variable shape at the level
of the light microscope (compare the insets of Figures 63 and 64).
Ultrastructural variations relate to the distribution of membrane
elements and mitochondria. For example, in the cell shown in Figure
63 membrane vesicles and cistemae, as well as mitochondria, are
distributed throughout the area occupied by the chromosomes, although
the former are more numerous at the periphery. In contrast, the area
of the chromosomes in Figure 64 contains very few small vesicles.
Numerous vesicles and larger cistemae, some of them rough ER, are
arranged almost concentrically at the periphery of this area.
Mitochondria are also excluded. In this respect the cell in Figure 64
resembles a prometaphase more than a metaphase, but the degree of
condensation of the chromosomes does not bear this out.
The centrioles, embedded in amorphous or finely fibrillar material
of moderate electron density, were always found at the periphery of
the area occupied by the chromosomes in the few cells examined (Figures
63 and 64). Four centrioles are present in each cell. They do not
differ structurally from centrioles in untreated cells.
Kinetochores appear as dense bands embedded in a less dense
fibrillar matrix and following the curvature of the chromosomal
surface, or as patches of less dense material (Figures 63, 64, 92-98).
This depends on the angle and the level of the section. No MT at all
were found in such cells. Treatment B (0.25 pg/ml colcemid for 15 rain,
no recovery) produces strikingly different effects. As with treatment

69
A, the usual criteria for determining mitotic stages do not apply.
Nevertheless, light microscopic examination of cells subjected to
treatment B revealed stages resembling pro- and prometaphase rather
than metaphase. One cell was in late telophase or cytokinesis. The
chromosomes, fairly distinct in the light micrographs (insets of
Figures 65 and 66), were extremely difficult to see by direct
observation.
Electron microscopy confirmed the above observations and revealed
interesting details. The central area of the cells in Figures 65 and
66 consists of a coarsely granular or fibrillar ground substance in
which the chromosomes are embedded. The separation of this area from
the cytoplasm is almost perfect; with the exception of small membrane
vesicles all the larger organelles are excluded. In the cell in
Figure 65 large pieces of double membrane, probably fragments of the
NE, are present at the border between central area and cytoplasm.
Grazing sections revealed no pore-annulus complexes on these fragments,
although colcemid does not destroy their integrity on the NE of
interphase nuclei (Figure 68). Chromosomes are highly dispersed,
making the recognition of sister chromatids difficult (Figure 65).
The dark centromeric granules of chromosomes in the light micrograph
are patches or balls of more tightly packed, perhaps also finer,
chromosomal fibers. There seems to be one patch per sister chromatid,
and, as revealed by serial sections, connections between chromatids
exist in this region. Kinetochores in the usual sense are lacking,
but a vesicular space relatively poor in fibers and filled with a less
dense substance can be recognized adjacent to the dense patches.

70
The cell in Figure 66 differs from the above description by the
absence of large membrane cistemae and the higher degree of chromo
some condensation. It is remarkable, in this context, that the
chromatin of all the interphase cells subjected to treatment B is
completely dispersed (compare Figure 67 with Figure 5). The fibers in
the primary constriction of the chromosomes in Figure 66 are also
finer, or more tightly packed, or both. In this case, however, a
distinct less dense kinetochore band is present in the vesicular space.
Despite the relatively high concentration of colcemid used in
treatment B, a few 1-T were found in interphase cells (Figure 68), but
not in c-mitotic cells. In some of the latter, numerous bundles of
microfibrils, approximately 70 A in diameter, were present in the
central area (Figure 69).
Mitosis in Cold-Treated Cells
Exposure to 0-4C does not accumulate metaphase cells. Cells at
all stages of mitosis are present in the unsynchronized cultures used.
The general ultrastructural features of prophase correspond to control
cells, except for the presence of intranuclear clusters of granules
approximately the size of ribosomes (Figure 70). Kinetochores of
cells in mid-prophase consist of less dense, fibrous material in
constrictions of chromosomes (Figure 72). Amitotic spindle in the
usual sense does not exist in prometa-, meta-, and anaphase cells,
although centrioles are situated at opposite poles (Figures 71 and
75). Most of the few MT present are associated with kinetochores.
An amorphous moderately osmiophilic substance coats the micro
tubules, whose walls appear as stark lines (Figures 73 and

71
75). Fully differentiated kinetochores exhibit the familiar triple-
banded profiles (Figure 75), though less distinctly than in normally
fixed control cells. The fuzziness of kinetochores and MT is not
caused by the initial cold fixation, because these structures are well
preserved in similarly fixed cells not previously exposed to cold
(Figure 74). The paucity of MT is also obvious in the midbody region
of cells in cytokinesis (Figure 76). Centrioles are not structurally
altered by exposure to cold (Figure 77).
Chromosomes are normally condensed, except for achromatic holes,
which are prominent at all stages of mitosis (e.g., Figure 71).
Remarkable is the great number and increased clarity of chromosomal
granules in cold-treated compared to control cells (Figures 70-72, and
75).
Kinetochore Fine Structure
I have already described in detail the fine structure of kineto
chores from prophase to late prometaphase. To establish a basis for
the discussion of kinetochore models, this section is devoted to an
in-depth study of kinetochores in normal meta- and early anaphase,
colcemid- and cold-treated cells.
Figure 78 shows five of seven serial sections in the horizontal
plane of a kinetochore in a cell in very early anaphase. Most of the
sections were too thick (approximately 800 A) to reveal details
concerning the attachment of MT. However, two things are very clear:
the triple-banding, and the greater length of the kinetochore in the
median sections (Figures 78b and 78c) compared to the peripheral
sections (Figures 78a, 78d, and 78e). Thinner sections (500-600 A),

72
also approximately median, of other kinetochores are presented in
Figures 81-84. These kinetochores were sectioned transversely, except
those shown in Figure 83, where obliquely cut kinetochore MT indicate
the sections were tangential. Kinetochore profiles very similar to
those in horizontal sections can be seen in para-sagittal sections
(Figure 85).
Kinetochores in metaphase cells are rarely flat. Most commonly,
they are undulated or S-shaped (Figure 42), more seldom concave
(Figure 85a). In very early anaphase the kinetochores of small
chromosomes in the center of the spindle are more or less flat (Figures
46, 48, and 82), and at a slightly later stage the kinetochores of such
chromosomes are convex (Figure 84). The kinetochores of the long
chromosomes at the periphery of the spindle are convex or more
irregular in early anaphase (Figures 47 and 49). In mid-anaphase,
S-shaped and more exotic profiles of kinetochores are prevalent.
Triple-banded kinetochores in paraxial median sections are 4,000-
6,700 A long. A faintly staining corona, approximately 400 A wide and
consisting of fine fibrils embedded in an amorphous matrix, covers the
kinetochores on the poleward side (Figures 78 and 84). The width of
the three bands varies, both within and between kinetochores. Average
values were 390 A for the outer, 270 A for the middle, and 400 A for
the inner band. The outer band consistently stains less intensely than
either the inner band or the chromosome proper (Figures 84 and 85a).
This is very clear in the image seen on the screen of the electron
microscope, but in micrographs printed on contrasty paper the differ
ence is obscured (e.g., Figures 78, 82, and 83). Figure 80 shows a
peripheral section of a stretched kinetochore in a late prometa- or

73
metaphase cell. This section was picked up on an uncoated 200-mesh
grid. The difference in contrast between the chromosome and the
kinetochore is undeniable.
The basic structure of the outer band is finely granular in
highly condensed kinetochores (Figures 78, 81-84), but in grazing
sections of less condensed kinetochores, 30-50 A fibrils are visible
(Figure 80). Superimposed on the fine granularity of condensed ki
netochores is a structure of coarse granules or fibers, giving the
band a knotted appearance (Figures 78, 81-85). The structure and
electron density of the middle band are very similar to the corona
(Figures 78b and 84). The inner band is continuous with the chromosome
(Figures 78b and 84), but in very early anaphase it may be connected
to the main body of the chromosome by a "stalk" of chromatin, giving it
the appearance of a mushroom (compare Figure 84 with Figure 46). The
fibers of the inner band seem to be identical with the fibers of the
chromosome. The greater opacity may be due to denser packing, but an
interesting alternative is the presence of a very fine amorphous
substance, which is lacking in the remainder of the chromosome.
Microtubules attach to the kinetochore at a variable angle.
Three conditions must be fulfilled in paraxial sections to determine
how far the MT penetrate into the kinetochore: (1) The section must
be thin (500 A); (2) the ITT in question must not be cut obliquely near
the kinetochore; and (3) the section must be approximately median.
These conditions are fulfilled in Figure 82. The MT marked with an
arrow penetrates the outer layer and ends at the interface with the
middle layer. The different impression created by an obliquely
sectioned MT is demonstrated in the somewhat thicker section of Figure

74
81. The straight MT marked with an arrow terminates in the outer
layer, while the obliquely sectioned MT marked with an arrowhead seems
to penetrate into the inner layer. Skew MT passing in front of a
kinetochore occur occasionally (Figure 79).
One metaphase cell was sectioned in a para-equatorial plane from
one pole across the metaphase plate into the opposite half-spindle.
Of the 24 kinetochores examined, all were apparently cut head-on
(e.g., Figures 86 and 87), except one, which was sectioned at a
slightly oblique angle. This kinetochore belonged to a chromosome at
the periphery of the metaphase plate. To determine which of the MT of
a chromosomal bundle were kinetochore MT, serial micrographs of eight
different chromosomes and associated MT, at final magnifications
between 40,000 and 62,500, were analyzed. In the last section of the
series, MT in the vicinity of the chromosome were marked with one
color. Proceeding poleward, newly emerging MT at the kinetochore were
marked with a different color. The average number of kinetochore MT
was 26, the range 16-40. This agrees well with estimates from paraxial
serial sections. I was unable to ascertain if the number of kineto
chore MT is correlated with chromosome size. The average number of MT
bypassing the kinetochore was 5 (range 0-9). It is rather arbitrary
to choose these bypassing MT from the population of non-kinetochore
tubules, the only criterion being their proximity to the chromosome in
question.
The kinetochores in para-equatorial sections are roughly circular
patches of variable electron density, depending on the level of the
section (Figures 86-88, and 91). The diameter varies from approxi
mately 3,400 A for obviously convex kinetochores to 6,000 A for
flatter kinetochores.

75
The kinetochore of Figure 86 was most likely similar to those in
Figures 78, 81, and 82. The first section of the kinetochore itself
shows a moderately opaque patch of finely fibrillar material (Figure
86d). I interpret this as the outer layer. Fewer ITT are visible than
in the preceding sections, indicating they terminate at this level.
Remarkable are less opaque circles whose diameter is similar to the
inner diameter of MT. They mark the terminals of kinetochore MT, the
wall of which cannot be seen because its opacity is the same as that
of the outer layer. A few MT can be followed one section farther
(Figure 86e), because the kinetochore is not perfectly flat, but as
the sections pass through the inner layer (Figure 86f) and through the
chromosome (Figures 86g and 86h) all the kinetochore MT have dis
appeared. The middle layer is always obscured by the more opaque
outer or inner layers, because the sections are thicker than the
middle layer. Even a very thin section would have to pass just
between inner and outer layers of a perfectly flat kinetochore in
order to show middle layer only, a very unlikely event.
Figure 87 shows two adjacent sections of two more convex kineto-
chores. The moderately opaque patches in Figure 87b again represent
part of the outer layer in which terminals of MT are visible. The
kinetochore shown in Figure 88 was most likely similar to those in
Figure 83. The section grazed the apex of the inner layer (large
arrow) which seems embedded in the less opaque outer layer. The
latter is clearly set off from the chromosome.
Occasionally, a single MT is found to penetrate the kinetochore
and to extend deeply into the chromosome (Figure 89). Whether such MT
pass completely through the chromosome is not clear.

76
I found no intertubular connections and arms on MT in sections
passing close to the kinetochores. Figure 90 shows what presumably is
a bundle of kinetochore MT at a greater distance from the chromosome.
Two apparent cross-bridges and two arms can be seen.
Anaphase kinetochores in para-equatorial sections are very
similar to metaphase kinetochores, except that most of them are more
convex, i.e., MT at the periphery of the kinetochores terminate one or
two sections after the apical MT have disappeared (Figure 91).
Kinetochore profiles in colcemid-treated cells are quite differ
ent from the triple-banded structures in untreated cells (Figures 92-
98). Figure 92 represents an approximately median longitudinal
section of two chromosomes. A longitudinal section along line C-D and
perpendicular to that of Figure 92 would produce a face-on view of the
kinetochore as in Figure 93. A transverse section along line A-B,
also perpendicular to that of Figure 92, would produce an image as in
Figure 94. The kinetochore bands in sections such as in Figures 92
and 94 are approximately 400 A wide, are embedded in a less opaque,
fibrillar matrix, and closely follow the surface of the chromosome.
The bands are slightly less electron dense than the chromosomes
(Figure 92), but again this characteristic is obscured in contrasty
prints (Figures 94, 97, and 98). I consider these bands equivalent to
the outer layer of kinetochores in untreated cells. There is no inner
layer on any of the chromosomes in colcemid-treated cells, and the
kinetochores are never attached to MT. The outer layer is apparently
made of two 150 A sheets held together at various points. This
accounts for the transverse and longitudinal sections showing two

77
bands knotted together (Figures 92, 94, 97, and 98), as well as for
the fibrous structure in grazing sections (Figures 93, 95, and 96).
Further evidence for the contention that the bands are actually
transverse sections of irregularly undulated (K^ in Figure 92; Figures
94 and 97), convex (K in Figure 92), or almost flat sheets (K^ in
Figure 98), came from careful analysis of serial sections of chromo
somes whose orientation relative to the plane of sections was known
from phase contrast micrographs. The diameter of these sheet-like
kinetochores is 5,000-9,500 A. Values of 7,000-8,000 A are most
Common.
Figure 95 illustrates rather unusual, exotic kinetochore profiles.
Kinetochore no. 1 seems to consist of two bands converging at their
ends. Additional bands were visible at the same locus in three
adjacent serial sections. Similar observations were made on the
undulated kinetochore no. 2.
In contrast to the single-handed kinetochores of colcemid-treated
cells, the metaphase kinetochores of cold-treated cells are triple-
banded as in Figure 75. The bands are fuzzier, but their relative
opacity is very similar to that of kinetochores in untreated cells.
Kinetochores of cold-treated cells are also attached to MT, but these
re few in number.
Chromosomal and Mitotic Aberrations
Untreated Cells
Frequency of aberrations
To score the frequency of chromosomal and mitotic aberrations,
Epon wafers with the embedded cell monolayers were scanned with the

78
phase contrast microscope at a magnification of 300. Obviously
aberrant, as well as doubtful cells were examined at magnifications of
625 and 1,560. Acentric fragments, maloriented chromosomes, and other
aberrations (e.g., multipolar spindles) were scored in metaphase cells
fragments, lagging chromosomes, and dicentric bridges were scored in
anaphase cells. The results are presented in Table 1.
The cumulative frequency of normal and abnormal cells is less
than 100%, because cells that could not be clearly identified as
normal or abnormal were included in the sample, but not assigned to
either category. The difference between the total frequency and 100%
is the frequency of these "doubtful" cases.
Gross and fine structure of aberrations
Acentric fragments and dicentric bridges.Figures 99-101
illustrate one of the most common types of aberration in anaphase
cells, viz., a single dicentric bridge and fragments. These bridges,
formed by a subterminal exchange between sister chromatids, are more
or less attenuated, probably depending on the degree of spindle
elongation. In both cells shown the kinetochores of the dicentric
chromosome were perfectly normal (e.g., Figure 100). There was
nothing unusual about the fibrous structure of the bridged chromosomes
The fragments in these cells were located in the cytoplasm at the
periphery of the spindle (insets in Figures 99 and 101). Analysis of
serial sections of several such cells revealed the truely acentric
nature of the fragments.
Dicentric bridges, single or multiple, also occur in telophase
cells. In this case the intact bridges cross the region of the

79
Table 1.Frequency of chromosomal and mitotic aberrations
in untreated meta- and anaphase cells
(streptonigrin control).
Mitotic
Stage
No. of
No. of
Total No.
Percent
Normal
Abnormal
of Cells
Normal
Abnormal
Cells
Cells
Examined
Cells
Cells
Metaphase
90
23
135
66.7
17.0
Anaphase
56
5
66
84.9
7.6
Total
146
28
201
72.7
13.9

80
equatorial constriction. Two cells in cytokinesis were found with
"nuclear" bridges connecting the daughter nuclei through the compact
midbody. One of these cells was examined in the electron microscope.
Nuclear pore complexes were present on the NE wrapping the bridges.
Lagging chromosomes.The second most common type of aberration
in anaphase cells is laggards. Only a few such chromosomes could be
examined in serial sections. Representative profiles of kinetochores
of two laggards are shown in Figure 102. These chromosomes were lying
near or at the periphery of the spindle (inset in Figure 102a).
Characteristically, a bundle of arched MT was present at the kineto
chores, which were oriented towards the center of the spindle. Some
of these MT apparently bypassed the chromosome, while others seemed to
terminate in the fuzzy kinetochores (Figure 102). The bundle of MT
associated with laggard no. 1 (Figure 102a and inset) also seemed to
be associated with laggard no. 2 (inset in Figure 102a) in the
opposite half-spindle. Numerous non-kinetochore MT near the laggards
were oriented haphazardly. Kinetochores and MT of most of the chromo
somes near the poles were normal. The chromosome in Figure 103 was an
exception. The strange profile of its lcinetochore in some of the
serial sections may be due to a peculiar angle of sectioning. Some of
the kinetochore MT, particularly in peripheral sections of the kineto-
chore (Figure 103b), were definitely odd.
In a similar cell one kinetochore of a laggard was very stretched
and similar in structure to the one in Figure 102a.

81
Centriole aberrations.--The three major possible aberrations
involving centrioles abnormal number, abnormal structure, and
abnormal position are illustrated in Figures 104-107. The presence
of four centrioles in interphase (Figure 104) presumably leads to
multipolar mitosis. The prophase cell in Figure 105a was very similar
to the one in Figure 11a, as far as chromosome condensation and
kinetochore differentiation are concerned. However, the two pairs of
centrioles were positioned on opposite sides of the nucleus (Figure
105a). Despite this obviously axial arrangement, very few MT were
associated with each centriole pair (Figures 105b and 105c).
Figure 106 shows two serial sections of the pole no. 2 centrioles
of the cell in Figure 29. Portions of three centrioles are visible.
The one in the center was deformed to a cup-like structure. Because of
missing sections, the precise architecture of, and relationship between,
these centrioles could not be reconstructed. Microtubules converged on
the osmiophilic masses left and right of center in Figure 106a.
Centriole aberrations also occurred in the anaphase cell of Figure
102. One of a pair of centrioles near pole no. 2 is shown in Figure
107a. The MT, however, converged on a different center (P^), where
amorphous, osmiophilic material was present. I am convinced that there
was at least a third centriole at this spot, but I could not verify it,
because some of the serial sections were missing. On the other hand,
one serial section of the centriole shown in Figure 107a revealed a
very osmiophilic particle in the lumen of the centriole (Figure 107b).
In size and shape this particle was very similar to those found in the
vicinity of centrioles at all stages of the cell cycle (e.g., Figure 4).
At pole no. 1 of this anaphase cell there were probably also more than
two centrioles (Figure 107c).

82
Streptonigrin-Treated Cells
Mitotic index and frequency of aberrations
The mitotic index was determined from cell counts made with the
phase contrast microscope at a magnification of 200 or higher. Mitotic
cells comprised all the stages from pro- to telophase. The control
cells were the same used to compute the frequency of aberrations (see
Table 1). The results are presented in Table 2.
The procedure for scoring aberrations was as described for un
treated cells. The results are presented in Table 3.
The figures for meta- and anaphase cells are of relative value
only, because the distinction between these stages is not sharp in
highly aberrant cells. As for untreated cells, the cumulative frequency
for 0.01 pg/nl SN in less than 1007., due to "doubtful" cases.
Gross and fine structure of aberrations
The cytological effect produced by the two concentrations of SN
differs not only quantitatively, but also qualitatively. Generally
speaking, the higher concentration induces more complex aberrations and
bizarre mitotic figures are common (Figure 108). Ana- and telophase
cells usually contain several dicentric bridges of variable diameter,
probably depending on the degree of attenuation (Figures 108 c-f).
Numerous acentric fragments are located at the periphery of these cells.
Chromatin strands connecting the kinetochore regions of daughter
chromosomes occurred in cells in very early anaphase (Figure 109). The
kinetochores of these chromosomes appeared normal. An apparent
exception is illustrated in Figure 109. This kinetochore resembled
certain prometaphase kinetochores (see Figures 35 and 41). Its
associated MT passed over the adjacent chromosome.

83
Table 2.Mitotic
treated
indices (Ml) for streptonigrin-
and untreated control cells.
Treatment
Total No. of
Cells Counted
Cells in Mitosis
no. % Iffi
Control
541
14
2.59
0.01 ug/ml SN
540
12
2.22
0.05 ug/ml SN
541
1
0.18

Table 3.Frequency of chromosomal and mitotic aberrations In streptonlgrin-treated
cells.
No. of Metaphase
No. of Anaphase
Total No.
Percent
Treatment
Cells
Cells
of Cells
Normal
Abnormal
Normal Abnormal
Normal Abnormal
Counted
Cells
Cells
0.01 ug/ml SN
63 65
16 33
200
39.5
49.0
0.05 ug/ml SN
*1
CN
O
0 14
35
0
100.0

85
Many cells treated with 0.05 ig/ml SN contained chromosomes less
condensed than normal (Figure 111). Chromosomal fibrils possibly
cemented together by an amorphous substance formed more electron dense
patches either at the surface of, or within, these chromosomes (Figures
111b and 111c).
Dicentric bridges very often contained strands of densely packed
or finer-than-norraal fibers that extended, at least in a few carefully
studied cases, from one kinetochore to the other (Figures 110 and 111a).
The kinetochores in cells with dicentric chromosomes were structurally
normal, but their distance from the respective pole was more variable
than in normal ana- and telophase cells (Figures 110a and 113). The
kinetochores of dicentrics, in particular, seemed to lag.
Serial sections revealed the truely acentric nature of fragments
irrespective of their position within a cell.
Nuclear envelope reconstruction begins near the poles, along the
sides of the chromosomes. Small cistemae, already with nuclear pore
complexes, are closely apposed to the chromosomes (Figure 112). It
appears there is a gradient for this process from the polar regions to
the interzone. For example, membrane cistemae may be found along the
poleward portions of dicentrics and laggards, but not along the inter
zonal portions.
Late telophase cells with nuclear bridges through the midbody are
more frequent in SN-treated than in control cells. An example is shown
in Figure 116. The nuclei of the daughter cells were polymorphic.
Pockets extended deep into the nuclei on the poleward side. Very
electron dense patches within the nuclei, or on the inner membrane of
the nuclear envelope lining the pockets, are remnants of the inner

86
layer of kinetochores (Figures 114-116). Numerous MT were observed
penetrating from the pockets into the nuclei, indicating the envelope
was not completely reconstructed. Pieces of double membrane, some
bearing ribosomes (Figure 117a), were also found in the nuclei.
Occasionally, ribosome-like particles appeared on both membranes of the
nuclear envelope (117b).
Cells in late cytokinesis, connected by nuclear bridges across the
midbody, can be identified with the light microscope. Surprisingly,
however, a greater number of daughter cells connected by extremely thin
bridges are found in thin sections. Many of these aberrations are of
truely ultrastructural dimensions. The greater portion of the bridge
shown in Figure 118 was only approximately 500 A in diameter. In this
case no midbody was found between the two daughter cell. The bridge
simply passed across the cleavage furrow into the other cell and
expanded to a large nuclear bleb, which was in turn connected to the
main nucleus by a similarly thin stalk. Both daughter cells also had
micronuclei. A similar case is illustrated in Figure 119. The thin
bridge passed through the compact midbody (Figure 119a). The drop
shaped nuclear bleb was connected to the main nucleus by a very thin
stalk. Micronuclei were also present in both daughter cells (Figure
119b).
Centriole aberrations also occur in SN-treated cells, but
apparently not more frequently than in untreated cells. In one of two
daughter cells in cytokinesis the two centrioles were found unusually
far from the nucleus and also remote from each other.

87
Fig. 3. Interphase. Grazing section of the nucleus. Note
centrioles (c), Golgi complex (g), polyrihosor.es (r) on nuclear
envelope, pore-annulus complexes at the level of the envelope (small
arrows) and the level of the nucleus (large arrows), achromatic holes
(H) in the peripheral chromatin (Chr) underlying pores, x 15,750.

Fig. 4a-d. Interphase. Four serial sections of a centriole.
(a) Osmiophilic mass near distal end. (b) Probably the distal
end. Note short rays radiating from triplets (arrows), (c)
Probably near-median section. Note osmiophilic mass forming
cylinder around the centriole, spherical particle (Pa), (d)
Median section revealing intracentriolar vesicle (CV). Note
satellites (s) in all four sections, x 100,000.


Fig. 5a-c. Interphase. Transverse sections of representative
nuclei, (a) Dispersed chromatin (Chr). Note granular (Gr), and
fibrillar (F) components of nucleolus (llu); intranuclear vesicle
containing particles (IV), nuclear envelope (HE). xl5,750. (b)
Moderate proportion of heterochromatin (HChr). x 15,750. (c)
Relatively great proportion of heterochromatin. x 11,500.

91
K& i &C*

Fig. 6. Interphase. Chromatin fibers in the nucleus of
Fig. 5a. Note 250 A fibers (large arrows), 125 A fibers (arrow
heads), 70-80 A core within some 250 A fibers (small arrows),
x 75,000.
Fig. 7 Interphase. Portion of the nucleolus shown in
Fig. 5*>. Note granular (Gr), fibrillar (f) components; nucleolus-
associated chromatin (Chr). x 30,000.

93

94
WtM:
i* sK'JL <>
Fig. 8a. Very early prophase. Note two pairs of centrioles (Cp,
Co), condensing chromosomes (Ch), nucleolus (Nu). Nuclear envelope
(RE) with pores (NP); Golgi complexes (g), rough ER (NCR), and mito
chondria (Mi) in the cytoplasm, x 11,500.

Fig. 8b-e. Very early prophase. Four serial sections of
the two pairs of centrioles shown in Fig. 8a. Daughter centrioles
(DCj_, DC2) closely associated with parent centrioles (PC-^, PCg).
x 40,000.

96

97
Fig. 9a-e. Very early prophase. Migration and structure of
centrioles. (a), (b) Two serial sections shoving one of each pair
of centrioles (C]_, C2). Note satellites (s). Black spots near C]_
in (b) are staining marks. x 1(0,000. (c-e) Three serial sections
of centriole no. 1 (C^). Note cartwheel structure vith hub and
spokes in (c) and (d), skev arrangement of tubular triplets; short,
osmiophilic bars between triplets in (e). x 122,500.

98
Fig. 10. Mid-prophase. Note dispersed nucleolus (Nu),
chromosomes (Ch), kinetochore (k), intact nuclear envelope (NE).
x 11,500 Inset: Phase contrast micrograph of the cell in plastic,
(x l,28o).

99
Fig. 11a. Mid-prophase. Centriole (C) in pocket of the
nuclear envelope (HE). Note chromosomes (Ch) with sister kineto-
chores (Ki, K), remnant of the nucleolus (Nu). x 15,750. Inset:
Phase contrast micrograph of the cell in plastic (x 1,280).
0 O c o

100
Fig. 11b. Mid-prophase. Serial section of the cell in Fig. 11a,
at the periphery of the nucleus. Note microtubules converging at the
location of the centrioles (c); the centrioles themselves not shown.
Mitochondria (Mi) also radially oriented. Grazing sections of the
nuclear envelope (NE). x 11,500. ... ....

Fig. 12. Early prophase. Grazing section of the nucleus. Note
intact nuclear envelope (NE) with pores (NP), chronosorr.es (Ch) with
large achromatic holes (h). x 30,000.
Fig. 13* Early prophase. Transverse (serial) section of the
nucleus shown in Fig. 12. Note chromosomes (Ch) apparently attached
to nuclear envelope (NE) hy "stalks" (arrows), x 22,500
i

102

Fig. 14. Late prophase. Chromosome (Ch) attached to the
intact nuclear envelope (HE) by a "stalk" (arrow), x 57,500.
Fig. 15. Late prophase. Sister kinetochores (Kq_, K2) of
two chromosomes (Ch^, Cl^). From the same cell as Fig. l5.
x 30,000.

104

Fig. 16. Mid-prophase. Oblique section of a chromosome (Ch)
and one of its kinetochores (k). From the cell shovm in Fig. 10.
Note difference in electron density between chromosome and kinetochore.
x 62,500.
Fig. IT* Mid-prophase. Sister kinetochores (KKg) of a
chromosome in the nucleus shown in Fig. 11a. Note doubleness of
chromosome (Chd the two sister chromatids), kinetochore granules
(KG), x 67,500.

106

107
Fig. l8a. Very early prometaphase. Note fragmenting nuclear
envelope (NE) near centrioles (c), intact NE on opposite side of
"nucleus". One fragment formed a vesicle (NV). x 11,500. Inset:
Phase contrast micrograph of the cell in plastic (x l,28o).

108
Fig. 18b. Very early prometaphase. Serial section of the cell
in Fig. l8a. Nuclear envelope (NS) undulated around chromosomes (Ch),
fragments formed vesicles (NV). Note nuclear pores (NP). x 15,750.

109
Fig. 19. Early prometaphase. Breakdown of the nuclear envelope
(KE) near poles (Pj_, Pp). Note nucleolus organizer (NO) on X
chromosome. The white circle near the left margin is due to a hole
in the supporting film; "black spots are dirt and stain marks,
x 1,150. Inset: Phase contrast micrograph of the cell in plastic
(x 1,280).

110
Fig. 20. Early prometaphase. Formation of the spindle. Most
MT connect kinetochores (k) with the poles (Pp, P2). Note vesicle (v)
and elongated cisternae (Ci) at the proximal end of one of the centrioles
(C) at pole no. 1 (P-^). x 11,500. Inset: Phase contrast micrograph
of the cell in plastic; chromosome no. 1 is shown in Fig. 2b (x l,28o).

Fig. 21a-c. Early prometaphase. Grazing sections of the
nuclear envelope at progressively later stages, (a) From the
cell in Fig. 18; note eccentric pore granules (arrows), x 57>500.
(b) x *10,000. (c) x 50,000. Compare pore-annulus complexes (NP),
polyribosomes (R) in the three micrographs.

112
*' K.v.yf
'41 i *- jf^i^s & ** A t i ^ 4
^R 28£f?#P* / -.. /'. ^1 i
#> 21a
^If^AaBSBW *. VKirkfl.*
^ **,
NP
T
n

Fig. 22a-d (continued on the foilwing page). Early pro
metaphase. Kinetochores of the cell in Fig. 18. (a) Note
appearance of kinetochores (large arrows) compared to prophase
(e.g., Fig. 15)* Oblique and cross sections of MT (small arrows),
intrachromosomal MT (circles), x 30,000. (b) Sister kinetochores
(K]_, Kg) with faint band (arrows), x 57*500.


Fig. 22 (contd). (c) and (d) two serial sections of a
chromosome near the intact nuclear envelope (hS). Iiote Ml
outside the HE, opposite one kinetochore (&>); difference be
tween the two kinetochores. x 57,500.


Fig. 23a-c. Early prometaphase. Kinetochores of chromosomes
near pole no. 2 of the cell in Fig. 19. (a) Note difference be
tween kinetochores facing the near pole (K, K2), and kinetochore
facing the far pole (Kj_). Arrows indicate bands, x 30,000. (b)
and (c) two serial sections of another chromosome (Ch). Note
differences, similarities between the kinetochore facing the near
pole (K2) and the kinetochore facing the far pole (Kq_). Arrows
indicate bands, x 30*000

118

Fig. 2ia-c. Early pronetaphase. Three serial sections of
chromosome no. 1 shown in the inset of Fig. 20. Note stretching,
attached MT, of the kinetochore facing the near pole (K2); globular
shape, lack of MT, of the kinetochore facing the far pole (K^).
x 30,000.

120

121
Fig. 25. Early prometaphase. Kinetochores in the equatorial
region. From the cell in Fig. 20. Note unobstructed sister kineto
chores (Kq_, K2) of the X chromosome, its nucleolus organizer (NO),
other kinetochores (k). The direction of the spindle axis is in
dicated by MT marked with large arrows. Other MT are skew (small
arrows), x 30,000.

!*:7. /* ^
122
Fig. 2: Mid-proraetaphase. P^, P^, the two poles of the mitotic
spindle. A centriole is visible at pole no. 1. Several kinetochores
are indicated by arrowheads, x 7,750* Inset: Phase contrast 'micro
graph of the cell in plastic (x l,28o).

123
Fig. 27. Mid- to late prometaphase. Ch^, CI12, two equatorial
chromosomes. Cho a chromosome displaced towards pole no. 1 (see arrow
in inset). Kq., Kp, kinetochores oriented towards pole no 1 and pole
no. 2, respectively, x 7.750* Inset: Phase contrast micrograph'of the
cell in plastic (x l,28o).

124
mBw
Fig. 28. Late prometaphase. Ch2, Ch^, two equatorial chromosomes
Chj_ the chromosome displaced towards pole no. 1 (arrcw in inset). K-i,
Kp, the pole no. 1 and pole no. 2 kinetochores, respectively, x 11,500
Inset: Phase contrast micrograph of the cell in plastic (x l,28o).

125
Fig. 29. Late prometaphase. Note the small chromosome near
pole no. 1 (P-^). Kinetochores are marked by arrowheads. Fragments
of the nuclear envelope (NE) at the periphery of the spindle. Pieces
of double membrane with ribosomes (arrows), x 7,750* Inset: Phase
contrast micrograph of the cell in plastic (x 1,280).

Pig. 30a-, b. Late prometaphase, (a) Equatorial chromosomes
with kinetochores (arrowheads), x 11,500. (b) Relatively thick
section with trar.s chromosomal MT (arrows), x 30,000.
Fig. 31. Late prometaphase. Equatorial chromosomes with
kinetochores (arrowheads). Note chromatin strand connecting
stretched kinetochore regions (arrow), x 15,750.

127

128
Fig. 32. Mid-proraetaphase. Unusual arrangement of MT.
x 22,500.

129
Fig* 33a* Late prometaphase. Three of the chromosomes shown in
Fig. 30a at higher magnification. Unobstructed kinetochores (Kq_ and
Kp of CI13, K£ of Ch2) and obstructed kinetochores (K-j_ of Chp, Kq of
Cn-j_). Note wavy and skew MT. x L0,000.

130
Fig. 33b. Late prometaphase. Serial section of the three
chromosomes shown in Fig. 33a x k0,000.

Fig. 3l¡a-c. Late prometaphase. Three serial sections of
an equatorial chromosome. From the same cell as Fig. 39. Note
triple-handed kinetochores (Kj_, IGj), transchromosomal MT (arrows
x 40,000.

132

Fig. 35&-C. Late prometaphase. Unobstructed (K^) and obstructed
(K2) kinetochore of an equatorial chromosome. From the same cell as
Fig. 4l. Note difference in banding, shape, of the two kinetochores.
x 50,000.

134

Fig. 36. Mid-prometaphase. An equatorial chromosome with
stretched, fuzzy kinetochores (Ki, K2). Note strand of chromosomal
fibrils connecting the two kinetochore regions (arrows). x 30,000.
Fig. 37a, b. Late prometaphase. Two serial sections of the
small chromosome displaced towards pole no. 1 in Fig. 29. Note
fuzzy appearance of Kj_, compact Ko with single band (arrow in
Fig. 3Tb). x 50,000.

136

Fig. 38a, b. Late prometaphase. Higher magnification of
chromosome no. 1 shown in Fig. 23. Note single band of Kg
(arrowhead in Fig. 38a), faint triple-banding of Kp (arrowhead
in Fig. 38b). Arrows point to dark chromosomal granules,
x 30,000.

138

Fig. 39a-e. Late prometaphase. Klnetochores of a chromosome
displaced towards pole no. 2. Kq_, K2, the pole no. 1 and pole no. 2
kinetochores, respectively. Arrowheads indicate bands. Note bundle
of straight, some skew and wavy MT; passing the kinetochores.
x it0,000.

140

141
Fig. hOa-c. Late prometaphase. Two chromosomes (Ch[,, Chi-) of the
cell shown in Fig. 28 displaced towards pole no. 1. K1? the pole
no. 1 and pole no. 2 kinetochores, respectively, x 30,000.

142
Fig. 4l. Late prometaphase. Obstructed kinetochore (k) of an
equatorial chromosome (Ch). Note double (large arrows) and beaded
(small arrows) substructure of the moderately dense band, x 100,000.

143
W
Fig. 42. Metaphase. Chronosorr.es aligned on the metaphase plate
(horizontal section). Sister kinetochores (arrowheads) oriented
twards opposite poles. Bundles of kinetochore MT converge towards
the poles. Note dense chromosomal granules (arrows), x 15,750 Inset:
Phase contrast micrograph of the cell in plastic (x l,28o).

144
s^-vy-;^,
M^mw
*.' <-.wr Yte&'x *V
Fig. 43. Metaphase. Para-sagittal section showing four pairs of
sister chromatids. Three kinetochores (arrowheads), one of them distinctly
triple-handed and concave (upper left). Note dense chromosomal granules
(arrows), x 22,500. Inset: Phase contrast micrograph of the cell in
plastic (x l,28o).

145
ai^? HIWmwL
>V. w* v?-.f .78& .
s*ss§
Fig. 44. Metaphase. Para-equatorial section showing three
circular kinetochores (arrowheads). Note clusters of MT between
chromosomes (*), intrachromosomal MT (circles), dense granules
(arrows). x 22,500* Inset: Phase contrast micrograph of the cell
in plastic (x l,28o).

146
Fig. Very early anaphase. Daughter chromosomes are "pulled"
apart in the kinetochore region, lite "frayed" appearance of chromo
somes; kinetochores (arrowheads), MT, and a centriole at pole no. 1
(P-j_). x 11,500* Inset: Phase contrast micrograph of the cell in
plastic (x 1,28o).

147
Fig. 46. Very early anaphase. Several sister kinetochores (K^,
K) shearing variation of profiles. Kj_ in the upper left comer is
sectioned ohliquely or peripherally. Note the electron dense strand
connecting kinetochores of the chromosomes left of center (arrow).
x 22,500.

148
wmm
'T'-i** .' 4
wm
fcSfcfr&&fcS
sMU
- *
..vv >^1" jai
Fig. 4j. Early anaphase. Telomeric regions of the separating
daughter chromosomes trail the kinetoehore regions. Several kineto-
chores (arrowheads) and connecting strands (arrows) are visible.
Black spots are dirt, x 11,500 Inset: Phase contrast micrograph
of the cell in plastic (x 1,2S0).

Fig. 48. Very early anaphase. Two pairs of daughter
chromosomes (Ch^, Civ,) with their kinetochores (K^, K) and
associated MT. Note electron dense strands connecting kineto-
chore regions of daughter chromosomes (arrows). Serial section
of the cell shown in Fig. 45. x 22,500*
Fig. 49. Early anaphase. Three representative kineto
chores (k). The angular kinetochore near the right margin is
cut obliquely or peripherally. Serial section of the cell shown
in Fig. 4y. x 30,000.

150
:i( -ti cttfiri
\*v-- v* *Jr4% TfJrm
1 Stlijw /Ct3Cf ** ijr^
yV*, *.:
VSkT *m&> v itU^
^iSSfcSfegia:^^ 3sT ^
fe *,ytjp
jf iJ#*.
T 'Til
I *v
&>;*
r- ic-'t?-

151
Fig. 50. Mid-anaphase. Para-equatorial section of the cell shown
in the inset. Note two circular Itinetochores (k), penetrating MT
(circles), chromosomal granules (arrows). 15,750* Inset: Phase
contrast micrograph of the cell in plastic (x 1,28o).

152
Fig. 51. Late anaphase, (a) Survey of elongated spindle showing
two sets of chromosomes near opposite poles (p^, Pp) Note pieces of
double membrane at periphery of spindle area (large arrows), few inter
zonal MT (small arrows), x 7,750 (b) Detail of chromosome group at
pole no. 1. Note rather indistinct kinetochores (arrowheads), MT
crossing in front of chromosomes, x 30,000.

153
Fig. 52. Very late anaphase. Chromosomal mass near pole no. 1
(P-^). Note fuzzy kinetochore (arrowhead), membrane cisternae (large
arrows), chromosomal granules (small arrows), nucleolus organizer (NO),
and MT traversing the chromosomal mass, x 22,500. Inset: Phase
contrast micrograph of the cell in plastic (x 1,000).

Fig. 53* Very late anaphase. Kinetochore (K) in depression
on the poleward face of the chromosomal mass. Note less dense hand
(arrowhead), density of underlying chromatin, kinetochore MT.
Arrows indicate chromosomal granules. Serial section of the cell
shovn in Fig. 52. x 75,000.
Fig. 54. Very late anaphase. Stem bodies (Sb) in the
equatorial region. Serial section of the cell shovn in Fig. 52.
x 30,000.
Fig. 55a. Very late anaphase. Detail of stem body. Most
MT seem to terminate in the stem body or to project a very short
distance beyond it, but one MT (arrows) clearly passes through,
x 57,500.

155

Fig. 55h. Very late anaphase. Mitochondria (Mi), RER, vesicles
(V), and MT in the constricted equatorial region. Serial section of
the same cell as Fig. 55&. x 22,500.
Fig. 56. Centrioles in early anaphase. An unidentified particle
(Pa) and, possibly, a satellite (s) are present. Note MT embedded in
amorphous material surrounding the centriole on the right, x 75000.

157

158
Fig. 57. Early telophase. Reconstitution of the nucleus. Three
obliquely sectioned kinetochores within the chromatin mass (arrowheads).
Note large pieces of chromatin-associated membranes (large arrows),
chromosomal granules (small arrows), and spherical particle (Pa),
x 30*000 Inset: Phase contrast micrograph of the cell in olastic
(x 1,000).

Fig. 58. Early telophase. Otliquely sectioned nuclear lohe
(NL) with pore-annulus complexes (IIP). Transversely sectioned
pieces of KE indicated by large arrows. Serial section of the
cell shown in Fig. 57. x57,500.
Fig. 59- Mid-telophase. Two kinetochores (k) in pockets on
the polar face of the nucleus. Note MT (arrows) associated with
the kinetochores. Serial section of the cell shown in Fig. 60.
x 50>000*

160

161
Fig. 60. Mid-telophase. Nuclear envelope (NE) completely surrounds
the nucleus. A few ribosomes (R) present on the outer membrane of the
NE. Two kinetochores (k) present on the polar face of the nucleus.
Note MT (arrow) extending into the kinetochore pocket, x 30,000.
Inset: Phase contrast micrograph of the cell in plastic, (x 1,280).

162
Fig. 6l. Cytokinesis, Nucleus (ll) of one daughter cell with
large heterochromatic patches. Note cleavage furrow (arrows), stem
(Sm), midbody (MB). Black spot (d) and small dots are stain marks,
x 7,750.

163
Fig. 62. Cytokinesis. Higher magnification of the stem and
midbody (MB) shown in Fig. 6l. x 30,000.

164
Fig. 63. C-metaphase. The chromosomes are scattered. Three of
the four centrioles present in this cell can be seen (C). A kinetochore
is opposite the centrioles (arrowhead). Mitochondria (Mi), vesicles (v),
and cisternae (Ci) occur in the central area and at the periphery of
the cell. Black marks are staining artifacts, x 15,750 Inset:
Phase contrast micrograph of the cell in plastic (x l,28o). Treat
ment A.

165
Fig. 6k, C-metaphase. Two centrioles axe visible (c). A kineto-
chore is indicated by the arrowhead. Vesicles (V) occur in the central
and peripheral areas. Mitochondria (Mi), RER, and cisternae (Ci) axe
restricted to the peripheral area, x 22,500. Inset: Phase contrast
micrograph of the cell in plastic (x 1,280). Treatisent A. ...

166
mam
&&3&1
**&Vv
$ffiS3ga
f 3
Fig. 65. C-mitosis. Chromosome at the periphery of the central
area; Chdi, Chd2 its two sister chromatids. Kinetochores (K-j_, Kg),
centromeric granules (CG) in the primary constriction. The double
membranes (NE) are probably fragments of the nuclear envelope,
x 22,500* Inset: Phase contrast micrograph of the cell in plastic.
Chj_ the chromosome shown in the electron micrograph. Arrows indicate
centromeric granules (x 1,280). Treatment B.

167
Fig. 66. C-mitosis. Several chromosomes with kinetochores
(arrowheads) in the central area. A centriole (c) near the cytoplasmic
area, x 11,500 Inset: Phase contrast micrograph of the cell in
plastic, (x l,28o). Treatment B.

Fig. 67. Colcemid-treated interphase cell. Nucleus with
dispersed chromatin (Chr), nucleolus (Nu), nuclear envelope (NE).
x 15,750* Treatment B.
Fig. 68. Colcemid-treated interphase cell. Grazing section
of the nuclear envelope (HE) with pore-annulus complexes (NP),
helical polyribosomes (R). Note microtubules (MT), and micro
fibrils (MF). x +0,000. Treatment B.
Fig. 69. C-mitosis. Bundle of microfibrils (MF) in the
central area. Same section as in Fig. 66. x 75,000. Treatment B.

169
mm

Fig. TO. Cold-treated cell in mid-prophase. Note granular
component of the dissolving nucleolus (Nu) and its associated
chromosome (Ch), a kinetochore (arrowhead), clusters of intranuclear
granules (arrows), and chromosomal granules (circles), x 11,500.
Inset: Phase contrast micrograph of the cell in plastic. Note
the intact NE (x 1,280).
Fig. Jim Cold-treated cell in mid-prometaphase. Three
kinetochores (arrowheads) and a bundle of MT are visible. Note
achromatic holes in chromosomes, numerous chromosomal granules,
x 11,500. Inset: Phase contract micrograph of the cell in
plastic (x l,28o).

171
mmm$0
mm
. <- it'
;.
-

Fig. 72. Prophase kinetochores of cold-treated cell. Sister
kinetochores (Kq., K) lie in cup-shaped depressions. Fine fibrils
seen to radiate from the less dense mass of Kp. Note chromosomal
granules (circles). Serial section of the cell in Fig. 71. x 140,000.
Fig. 73* Cold-treated cell in netaphase. MT at higher magnifi
cation. Note amorphous coating, stark lines of MT vails, x 100,000.
Fig. 74. Cold-fixed control cell in netaphase. Three kinetochores
(K) and numerous MT are visible. Compare with Fig. 76. x 30,000.

173

Fig. 75. Cold-treated cell in prometaphase. Note fuzzy, "but
triple-banded profiles of kinetochores (K), associated MT, centriole
(C). x 30,000.
Fig. 76. Cold-treated cell in cytokinesis. The midbody (MB)
connected to the daughter cell on the right by the stem (Sm). Serial
sections at a different level revealed connection to the other
daughter cell. Note MT, clumped ribosomes, x 22,500.

175

176
Fig 77 Centrioles of cold-treated cell in metaphase. Note
clarity of triplets in cross-section, amorphous or fibrillar material
surrounding the same centriole. x 62,500.

Fig. 78a-e (continued on the following page). Five of seven
serial sections of a kinetochore in very early anaphase, (a) Section
no. 2. (b) Section no. 4; note corona (Co), outer band (KO), middle
band (KM), and inner band (ICE). (c) Section no. 5.

178
*
T*5
%s'Tt

Fig. 78 (contd.). (d) Section no. 6. (e) Section no. 7.
From the cell in Fig. 45. x 62,500*
Fig. 79* Peripheral section of a kinetochore in very early
anaphase. Note the MT traversing in front of the kinetochore
(arrow). From the same cell as Fig. Jd. x 50,000.

180

Fig. 80. Kinetochore in late prometa- to metaphase. Note
30-50 A fibrils (arrows) in outer layer (KO) of kinetochore, its
lesser electron density compared to the chromosome (Ch). x 100,000.
Fig. 8l. Metaphase kinetochore. A straight MT (arrow) ends
in the outer layer (KO); an obliquely sectioned MT (arrowhead)
seems to penetrate outer and middle layers, x 62,500.
Fig. 82. Kinetochore in very early anaphase. MT marked by
arrow ends in the outer layer (KO). From the cell in Fig. J45.
x 50,000.

t
H
AS.
1 Z'-

r*
-H *TI
391

Fig. 83a, b. Sister IdLnetochores in very early anaphase.
Note obliquely cut kinetochore MT (some marked by arrows), shape
of the kinetochores. From the cell in Fig. 45. x 50,000.
Fig. 84. Early anaphase kinetochore. Note corona (Co),
outer (K0), middle (KM), and inner (KE) layers, lesser density of
KD compared to ICE. From the cell in Fig. 47. x 122,500.

184

Fig. 85a-c. Three serial sections of a metaphase kinetochore
(parasagittal). Note triple-layered profile, lesser density of
outer layer (arrows) compared to inner layer. From the cell in
Fig. 43. x 62,500.

186
4V
8!

Fig. 86a-h. Eight serial sections of a metaphase kinetochore
(para-equatorial). Filled circles mark bypassing ME. (a) Open
circles mark kinetochore MT. (c) MT marked by arrows correspond
to those marked in (d). (d) Less opaque circles (arrows) are
terminals of ME; outer kinetochore layer (KO). (e) Inner layer
(Kl). (f-h) chromosome (Ch). From the cell in Fig. 41. x 62,500.

188
JE*-
A"* ni *
r* > f
9 &
vC> **

Fig. 87a, b. Two serial sections of two metaphase kinetochores
(para-equatorial). Filled circles nark bypassing MT. (a) Open
circles nark kinetochore MT. MT marked by arrows are the sane
marked in (b). (b) Arrows nark terminals of MT in outer layer.
From the sane cell as Fig. 87. x 50,000.
Fig. 88. Metaphase kinetochore. Note electron dense inner
layer (large arrow) within less dense outer layper; intrachronosomal
MT and granule within clear space (circle); other chronosomal
granules (small arrows). From the sane cell as Fig. 87.
x 50,000.

190
re -SnSp'SLi&ss
j. tmS$. E2*r*K. H ...-_i£^_.| __ ||M
3 4\ .,
. ... *'v*_JJ,'' sf*.'
^UKS^agzafcA*($£ WWvV 4* l4.
. *.** ,* -
nsfi- ;~V-
L X mpv
rsr
y.1
Vv£
0
% "*
-
^V
* ^>* I

Fig. 89. Para-equatorial section of metaphase cell. Note
intrachromosoraal MT within clear circle (arrow). From the sane
cell as Fig. 87. x 62,500
Fig. 90. Para-equatorial section of metaphase cell. Bundle
of Icinetochore MT. Note "cross-bridges" (large arrows) and "arms"
(small arrows). From the same cell as Fig. 87. x 75,000.
Fig. 91b., b. Tiro serial sections of mid-anaphase kinetochore
(para-equatorial). From the cell in Fig. 50. x 75,000.

192

Fig. 92. Kinetochores in c-mitosis. Two chromosomes (Ch-j_,
Chg) in longitudinal section, K2, the sister kinetochores
of Chi* x 50,000. Treatment A.
Fig. 93* Kinetochore in c-mitosis. Grazing section of
kinetochore (k), which is seen in face-view in the depression
of the primary constriction. Portions of the two arms of the
chromatid (Chd). x 50,000. Treatment A.
Fig. 9b. Kinetochores in c-mitosis. Sister kinetochores
(K¡_, K2) of transversely sectioned chromosome (Ch). Note two
centrioles (C). x 50,000. Treatment A.

194

Fig. 95a-c. Kinetochores in c-mitosis.
(a, b, adjacent) of sister kinetochores (K^,
profiles. From the same cell as Fig. 9b. x
Three serial sections
K2) with strange
T5,000. Treatment A.

196

Fig. 96. Kinetochores in c-mitosis. Kq_ obliquely, Kp transversely
sectioned, K grazed. From the same cell as Figs. 92 and 93* x 50,000.
Treatment A.
Fig. 9T Kinetochore in c-mitosis. Approximately median section.
Note doubleness (white arrows), "knots"(black arrows). From the cell
shown in Fig. 66. x 100,000.
Fig. 98. Kinetochores in c-mitosis. Sister kinetochores (Kq_,
Kg) in primary constriction. Note doubleness of Kg. From the
cell shown in Fig. 66. x 67,500.

198

199
Fig. 99. Dicentric bridge in an untreated anaphase cell. The
bridge does not appear continuous in this section. Note slight lagging
of one of the daughter X chromosomes (X2), its normal kinetochore (f^),
and the nucleolus organizer (NO). x 11,500* Inset: Phase contrast
micrograph of the cell in plastic. Note attenuation of bridge; frag
ments (arrow), and the X chromosome (X2); (x l,28o).

Fig. 100. Dicentric "bridge in an anaphase cell (serial
section of the cell in Fig. 99)* Note normal kinetochore (K2)
of "bridged chromosome, x 15,750.
Fig. 101. Dicentric "bridge in an untreated anaphase cell.
Note normal kinetochore (K^) of bridged chromosome, x 11,500
Inset: Phase contrast micrograph of the cell in plastic. Arrow
marks location of fragments (x l,28o).

201
mm
* ^ -n t£
Mv>&
.. m .* _i \ ^

Fig. 102a, t>. Kinetochores of lagging chromosomes in un
treated anaphase cell, (a) Chromosome no. l(see inset), (h)
Chromosome X2. Note bypassing MT (large arrows), kinetochore MT
(small arrows). Fuzzy kinetochore material is indicated by
arrowheads. Dark spot left of center in (b) is staining artifact,
x 50,000. Inset: Phase contrast micrograph of the cell in
plastic. Note position of three laggards (no. 1, no. 2, Xp);
(x 1,280).

203

Fig. 103a, b. Kinetochore and MT of a non-lagging chroinosone
in the sane cell as Fig. 102. Note curved and disoriented MT
(arrows), kinetochore (k). x 50,000*

205

Fig. 104. Four centrioles in an untreated interphase cell,
x
Fig. 105a-c. Abnormal centrioles in an untreated cell in
mid-prophase, (a) Centriole pairs (C-^, C2) on opposite sides
of the nucleus, x 7,750. (b) Higher magnification of the
centriole seen at C]_ in (a). Note paucity of MT. x 30,000.
(c) Serial section of centrioles at C2 in (a). Note paucity
of MT. x 30,000.

207

Fig. 106a, b. Abnormal centrioles in an untreated cell in
late prometaphase. Two serial sections. Three centrioles (C-i,
C2, C3) are visible. Note cup-shaped C-j_. MT converge on osmiophilic
masses left and right of Cp. From the cell in Fig. 29, P2*
x 57,500.

209

210
Fig. 107a-c. Abnormal centrioles in an untreated anaphase cell,
(a) Pole no. 2 of the cell in Fig. 102. The centriole (c) shown is
one of a pair. Note MT converging near Po, not C. x 30,000. (b)
Serial section of the centriole shown in (a). Note particle in
centriole lumen, x 137,500- (c) Pole no. 1 of the same cell. Note
strange appearance of C2, possibly a third centriole (arrow),
x 22,500.

Fig. 108a-f. Streptonigrin-induced aberrations. Phase
contrast micrographs, (a) Aberrant "metaphase". Chromosomes
scattered in the spindle, sister chromatids twisted, (b) Very-
early anaphase. Chromosomes not aligned on the equator. Sister
chromatids bifurcated in the stretched centromere regions, (c)
Late anaphase. Dicentric bridges (center), acentric fragments
(arrow), (d) Late anaphase. Two dicentric bridges (center),
acentric fragments (arrow). (e) Late anaphase. Two prominent
bridges. Lagging daughter chromosomes connected by very thin
bridge (center of spindle). Acentric fragments (arrows), (f)
Late anaphase telophase. Three thin bridges (center), two
groups of acentric fragments (arrows). (a), (b), (c), (e), and
(f): 0.05 ^ig/ml SIT; (d): 0.01 ig/ml SN; (all x 1,280).

212

Fig. 109. Streptonigrin-treated cell in early anaphase. Chromatin
strand connecting the kinetochore regions of daughter chromosomes
(arrows). Note profile of kinetochore (k) in a pocket of the adjacent
chromosome. From the cell in Fig. 108b. x 22,500.
Fig. 110a, b. Streptonigrin-treated cell in late anaphase, (a)
Portion of a dicentric chromosome with kinetochore (Kg), dense chromatin
strand (arrows). One other kinetochore is visible (k). x 22,500* (b)
Chromatin strand of a similar bridge at higher magnification. Note
difference in coiling between strand (marked by the two lines) and
peripheral fibers, x 157,500. Both from the cell in Fig. 108f.

214
&¡stk%
^ *T ^ -*J
v3*iKSB
t* >v ^
^
*/#

Fig. llla-c. Streptonigrin-treated cell in late anaphase.
(a) Kinetochore (k) and strand (arrows) of a dicentric chromosome,
x 22,500. (h) Telomeres of two long daughter chromosomes (Chp,
CI12). Note patches of more densely packed fibers (arrows),
x 22,500* (c) Loosely coiled chromosome with patches of more
densely packed fibers (arrows), x 40,000. All from the cell
in Fig. 108c.

216

217
Fig. U2a, Id. Streptonigrin-treated cells in late anaphase. Re
construction of the HE. (a) Portion of a polar group of chromosomes.
Note close relationship of membrane cisternae (large arrows) and RER
(small arrows) with chromatin. From the cell in Fig. 108d. x 30,000.
(*) Portions of two chromosomes near a pole. Note membrane cisternae
apposed to chromosome (large arrows), RER in proximity to chromosomes
(small arrows). From the cell in Fig. 108f. x 30,000.

Fig. 113. Streptonigrin-treated cell in late anaphase. Note
position of the kinetochores (k) relative to the pole (which was
just beyond the left margin). From the cell in Fig. 108d. x 30*000.
Fig. Il4. Streptonigrin-treated cell in late telophase.
Portion of the nucleus shown in Fig. 116. Note nuclear pocket nea,r
left margin. Osmiophilic patches (arrowheads) are presumably
remnants of kinetochores. Another kinetochore (K) within the
nucleus; note its associated MT (arrows), x 50,000.
Fig. 115. Streptonigrin-treated cell in late telophase.
Nuclear pocket with remnant of kinetochore (black arrow). A MT
penetrating into the nucleus (white arrow), x 75*000.

219

220
Fig. 116. Streptonigrin-treated cell in late telophase. Nucleus
of the upper daughter cell in the inset. Note nuclear pore complexes
(NP), kinetochores (arrowheads), nuclear pocket (also at the large
arrow), intranuclear MT (small arrows), x 15,750* Inset: Phase
contrast micrograph of the cell in plastic (x l,28o).

Fig. Ills,, b. Streptonigrin-treated cell in late telophase,
(a) Nuclear pockets; piece of RER trapped in the nucleus (arrear),
x 62,500 (h) Ribosome-like particles on outer and inner membrane
of NE (arrears). x 75,000. Both from the cell in Fig. Il6.
Fig. 118. Streptonigrin-treated cell in late cytokinesis.
Portion of the nuclear bridge (arrows) near the nucleus (N) of
one daughter cell, x 15,750.

222

Fig. 119a, b. Streptonigrin-treated cell in late cytokinesis,
(a) Drop-shaped nuclear "bleb (near left margin) and fine nuclear
bridge (arrows). Note midbody (MB), MT. x 7,750 (b) Micronuclei
(MN) with intact NE (arrows). Note main nucleus (N). x *10,000.
Both from the same cell.

224
'im

DISCUSSION
The phase contrast micrographs taken prior to sectioning of
selected cells were extremely helpful in determining mitotic stages.
Equally important, they facilitated the task of locating individual
chromosomes on thin sections, thus safeguarding against utterly false
conclusions. Consider, for example, Figure 27. From the electron
micrograph it would appear there is one long chromosome with four
kinetochores at the equator of the spindle. Phase contrast micrographs
at different levels of focus clearly revealed two different chromosomes
Not only the combination of light and electron microscopy, but
also the rat kangaroo cells themselves were an asset in this study.
The low chromosome number, the individuality of the chromosomes, and
the fact that the axis of the mitotic spindle in these cells is always
more or less parallel to the growth surface, all allowed a more
detailed study of mitosis than has ever been published for animal cells
Centrioles
Structurally, the centrioles of PtK^ cells, whether untreated,
colcemid-, cold-, or SN-treated (Figures 4, 9, 56, 77, and 94), are
similar to those in other cells and organisms (e.g., Brinkley and
Stubblefield 1970, de Harven 1968, Erlandson and de Harven 1971,
Robbins et al. 1968). The nine tubular triplets are embedded in an
osmiophilic matrix, which may be differentiated into ill-defined
structures (Figure 9). This matrix appears to be continuous with the
e o
extensive, slightly less osmiophilic material at the distal end
225

226
(Figures 56 and 77). At the proximal end the centriole is differen
tiated into the typical cartwheel (Figure 9). Satellites and an intra-
centriolar vesicle occur (Figures 4, 9, and 56), but I have been unable
to follow possible changes during the cell cycle, or to assess the
effect of cold, colcemid, or SN on these structures. The osmiophilic,
globular particles found near the centrioles at all stages of mitosis
are a mystery. They have also been observed by Brinkley (personal
communication). The possibility that they are virus particles lies at
hand, but is not confirmed. One is tempted to speculate that their
consistent association with the centrioles assures distribution to
daughter cells.
Centriole duplication follows the well-known pattern (Brinkley and
Stubblefield 1970, Erlandson and de Harven 1971, Murray et al. 1965).
Daughter centrioles arise at approximately a right angle to parent
centrioles (Figure 8). I have not obtained a sufficient number of
favorable sections of this stage to be able to confirm that daughter
centrioles are formed at the proximal end (see Brinkley and Stubblefield
1970). However, centrioles at the poles of the mitotic spindle are
oriented so that the proximal ends with the cartwheel point away from
each other (Figures 20, 56, and 77). Unless rotation and dislocation
occur following duplication, this configuration must reflect the
original parent-daughter association. It is interesting to note here
that the orthogonal arrangement is not the case in interphase cells
with supernumerary centrioles (Figure 104). Sections of colcemid-
treated cells (Figure 63) confirmed the observation by Brinkley et al.
(1967) that this drug does not inhibit duplication of centrioles. The
same'is true for the spindle poison vincristine (Journey et al.' 1968).

227
Whether centrioles also duplicate during cold treatment could only be
ascertained by direct observation or cineraicrography of living cells.
I have observed parent-daughter pairs in a cold-treated cell in early
prophase. In this case duplication could have occurred prior to treat
ment and the cell may have been completely arrested by cold.
No other organelle in cells of higher animals can match the
esthetic appeal of centrioles. Their orderly, symmetrical structure is
certainly in part the cause for much of the attention they have re
ceived since the early electron microscopic studies of Burgos and
Fawcett (1956) and de Harven and Bernhard (1956). In recent years,
however, the interest provoked by these intriguing organelles has led
to unwarranted speculation about their function (Brinkley and
Stubblefield 1970, Stubblefield and Brinkley 1967; see also discussions
by Pickett-Keaps 1969, 1971).
One such hypothetical function is the generation of continuous
spindle MT (Brinkley and Stubblefield 1970). It is clear that ultra-
structural studies can contribute only circumstantial evidence for this
hypothesis. In prophase cells, the centrioles appear to be the focus
of the increasing number of MT (Figure 11). Some of these will
undoubtedly become astral MT, others probably continuous and, possibly,
kinetochore MT. Except for the relatively rare skew MT, all the MT of
the fully formed spindle converge towards the centrioles (e.g., Figures
29 and 45). Several investigators have claimed to have observed MT
directly connected to the tubules of the centrioles (Krishan and Buck
1965), or inserted into the centriole wall between the triplets
(Brinkley and Stubblefield 1970, Gall 1961). I have examined these
published electron micrographs and found them totallyunconvincing.

228
Furthermore, I have examined numerous of my own micrographs of
centrioles at all stages of mitosis. In reasonably thin sections there
is no evidence whatsoever for direct connections between MT and
centrioles. Rare images where this seems to be the case can be ex
plained by lack of resolution or superposition in rather thick sections.
If the hypothesis were correct, one could reasonably expect such MT-
centriole associations to be frequent and unambiguous. The fact that
even the most ardent proponents of the hypothesis cannot claim this
speaks for itself.
Having dispensed with a popular myth we now can direct our atten
tion to a probably more important polar constituent, the osmiophilic
material surrounding and partly enveloping the distal ends of the
centrioles (e.g., Figures 45 and 56). The possibility that this
material is absent or reduced in amount at least during part of the
interphase (e.g., Figure 3) merits further investigation. It is
certainly present during all stages of mitosis, and it seems to be
arranged mainly around one of the centrioles of a pair, probably the
parent (Figures 56 and 77). I have been unable to observe changes
related to the mitotic cycle, as reported for HeLa cells by Robbins
et al. (1968). The material also occurs in colcemid-, cold-, and SN-
treated cells (Figures 63 and 77). Spindle MT penetrate into this
material in a rather chaotic fashion and apparently terminate there
(Figure 56).
Pickett-Heaps (1969) suggested that the microtubule-organizing-
center (MTOC) of animal cells could be in the osmiophilic material,
rather than the centrioles themselves. He did not explicitly discuss
the possibility that the MTOC, or some other component of the

229
osmiophilic material, could also be the pole-determinant, but the two
functions are obviously related. A rather diffuse, morphologically
less well-defined polar determinant and ItTOC could apply to acentriolar
mitosis in higher plants and other organisms as well.
The concept of MTOC as stated above is amenable to scrutiny in
animal cells. Ultrastructural investigations of the following
possibilities would yield significant information: (1) The number of
mitotic poles is correlated with the number of single centrioles or
pairs of centrioles regardless of the presence or absence of osmiophilic
material. (2) The number of poles is correlated with the number of
masses of osmiophilic material, regardless of the number of centrioles
present. Light microscopic observations have produced some evidence
against point (1). Normal bipolar divisions are more frequent than
multipolar divisions in somatic hybrid cells (e.g., Yamanaka and Okada
1968), and in spermatocytes of certain Diptera mitosis can take place
in cells lacking centrioles (Dietz 1959, 1966). Some of the centriole
aberrations I observed provide additional information. For instance,
the centriole shown in Figure 107a is not at the spindle pole. In a
preliminary study on the effect of the alkylating agent TEPA
(triethylenephosphoramide) I have observed a solitary centriole in the
interzone of a tripolar anaphase. Osmiophilic material was associated
with this centriole and some MT converged at this site, but it was not
a functional mitotic pole. On the other hand, both osmiophilic masses
in Figure 106 are foci of tIT. It appears that the presence of a
centriole in a mitotic cell is not identical with the formation of a
pole. Whether the same applies to osmiophilic masses remains to be
seen. Centriole-free divisions as in dipteran spermatocytes would be
the system of choice to investigate this.

230
In the view of Brinkley and Stubblefield (1970) centriole migra
tion is a consequence of the generation of MT, i.e., the centrioles
"propel themselves to the opposite mitotic poles by pushing against
each other through the spindle which they generate." The same concept
has been adopted by Friedlander and Wahrman (1970) and McIntosh et al.
(1969). This idea is plausible, for no matter what kind of pole-
determinants we propose, we are faced with the problem how they reach
opposite poles. Centriole migration does not occur in colcemid-treated
cells (Figure 63; see also Brinkley et al. 1967). I have seen a few liT
around centrioles in interphase cells treated with 0.25 pg/ml colcemid,
but not in mitotic cells. It would be very interesting to know if
centriole migration occurs in cold-treated cells where almost all of
the MT present are of the kinetochore type (Figure 75; see also
Brinkley and Cartwright 1970). This could only be investigated by
observing living cells, for although all stages of mitosis are found in
unsynchronized populations of cold-treated cells, this may just reflect
the state prior to, not the processes occurring during treatment. What
casts some doubt on this concept of pushing is the paucity and dis
orientation of MT in very early prophase (Figures 8 and 9). This is
even more striking in Figure 105, where the centrioles have apparently
migrated to opposite poles prematurely, before a true spindle was
formed. Alternative explanations of centriole migration are difficult
to conceive. Additional information, preferably from simpler systems,
is certainly necessary.
Microtubules
Microtubules in Ptl<2 cells have the usual dimension, i.e., a
diameter of 200-250 A. Preparations with long, straight MT can be

231
obtained (e.g., Figures lib, 28, and 42). Wavy MT occur most often
between, or near, chromosomes (Figures 20, 25, 27, and 33). This might
well be an artifact, for as Jensen and Bajer (1969) demonstrated in
Haemanthus endosperm, shrinkage of chromosomes during dehydration can
cause adjacent ITT to become wavy.
I do not think that skew MT are an artifact (e.g., Figures 25, 27,
28, 42, and 47). One is tempted to speculate that the number of skew
and otherwise "maloriented" MT decreases from early prometaphase to
metaphase, thereby reflecting the organization of the mitotic spindle.
To substantiate such speculation would require a thorough analysis of a
great number of sections of several cells. For lack of numerical data
we must content ourselves with the explanation that skew MT are part of
the normal spindle and that they do not interfere with orderly chromo
some movement.
Microtubules appear as hollow cylinders whose lumen has the same
electron density as the cytoplasmic matrix (Figure 90). The walls
appear as dark, more or less sharp lines or circles in untreated cells
(Figures 84 and 90). In cold-treated cells the MT are coated by an
amorphous or finely granular, moderately osmiophilic substance (Figures
73 and 75). In dividing amebae (Chaos) exposed to cold the MA is
completely disorganized, but MT reappear after only a few minutes
recovery at normal temperature (Roth 1967). These MT are not coated by
any material. On the other hand, fixatives containing divalent cations
preserve a fine material on the surface of MT in the MA of the same
organism (Roth and Daniels 1962). Amebae fixed in OsO. alone contain
4
no MT, but finely fibrillar, linearly oriented elements in the spindle
region. Kane and Forer (1965) observed no MT, but large numbers of

232
granules of approximately the same size in isolated MA of sea urchins
stored for longer periods of time.
Considering these observations one can hypothesize that the
substance coating MT in cold-treated Ptl^ cells is microtubular material
in a dispersed state, but still retaining affinity for persisting MT.
The latter could serve as nucleating centers for rapid polymerization
during recovery from cold. The necessity of nucleating centers for
polymerization of free subunits has been demonstrated by Stephens (1969)
with flagellar outer fibers, but a similar experiment with spindle MT
has not as yet been reported. But it is clear that repolymerization of
MT, as judged by reappearance of birefringence, occurs within seconds
or minutes after exposure to cold, depending on the duration of the
treatment (Inou 1964, Inou et al. 1970).
Brinkley et al. (1967) found persisting MT in Chinese hamster cells
arrested by exposure to 0.06 pg/ml colcemid. These MT were of the
kinetochore type; continuous MT were absent. A normal metaphase
configuration was formed after 15-20 min recovery. In contrast to
this, I found no MT in PtK^ cells after 1 hr recovery following a 2 hr
exposure to 0.05 pg/nl colcemid (e.g., Figures 63 and 64). This
indicates a distinct difference in sensitivity.
The microfibrils observed in PtK^ cells exposed to 0.25 pg/ml
colcemid may be significant in the context of MT depolymerization. In
my opinion these fibrils are not identical with microfibrils that are
numerous mainly near the growth surface in untreated cells. Similar
microfibrils have been reported in cells treated with vincristine by
Journey et al. (1968). Nathaniel et al. (1968) observed 35-50 A fibrils
in great quantity in melanocytes of Harding-Passey tumor treated with

233
colchicine. Here, then, are three instances where great numbers of
microfibrils were observed in cells exposed to spindle poisons.
Perhaps these fibrils represent a different aggregation state of ITT
subunits induced by the experimental conditions. Naturally, the corre
lation observed may be fortuitous. Further investigations, undoubtedly
necessary, will have to aim at demonstrating a negative correlation
between appearance and disappearance of microfibrils and MT.
There is evidence from immunological (Went 1960) and biochemical
studies (e.g., Sisken and Uilkes 1967, Wilt et al. 1967) that the major
proteins of the MA, among them MT proteins, are synthesized before the
onset of division. The formation of MT in the cytoplasm during
prophase can be explained as an assembly of preexisting subunits (see
also Roth 1964, Hicklas 1971), possibly directed by organizing centers
(Inou 1964, Inou and Sato 1967; the MTOC proposed by Pickett-Heaps
1969, 1971). Purely ultrastructural studies contribute little to the
solution of this problem.
If there is indeed a pool of MT subunits during prophase, the
absence of MT from the nucleus strongly suggests the NE is a real
barrier to these subunits. In all the mammalian cells studied so far
no MT are present within the nucleus during prophase (e.g., Krishan and
Buck 1965, Murray et al. 1965, Robbins and Gonatas 1964). At the time
the NE begins to break down near the centrioles, MT are numerous on the
outside, but very sparse on the inside of the envelope in PtK^ cells
(e.g., Figure 13b). This observation leaves no doubt that the NE is a
barrier to both subunits and assembled MT. In other organisms, how
ever, the spindle is intranuclear, and the ME apparently does not act
as a barrier (e.g.,
Aldrich 1969).

234
Electron micrographs such as Figures 18 and 19 demonstrate yet
another aspect of spindle formation: MT within the "nucleus" are far
more numerous near centrioles than on the opposite side. This could be
interpreted as a "growing process" away from the centrioles, possibly
concomitant with the diffusion of MT subunits into the "nucleus." As
the spindle develops further and the centrioles arrive at opposite
poles, the distribution of MT becomes more uniform (Figures 20, 26, and
29). The apparent paucity of continuous MT compared to kinetochore MT
in fully developed spindles (Figures 29, 42, and 47) is probably
imaginary. Kinetochore MT, arranged in bundles, are more conspicuous
than the dispersed continuous MT.
The absence of MT from the nucleus during prophase has a parallel
in telophase. Microtubules do not occur within the nucleus after the
NE is completely reconstructed (Figure 60). The presence of MT in
abnormal daughter nuclei of SN-treated cells (Figures 114-116) does
not contradict this observation, because in these cases the NE is
incomplete. It is not clear, however, in what way elimination of MT
enclosed in the chromosomal mass in late anaphase (Figure 52) is
accomplished. Robbins and Gonatas (1964) observed incompatibility of
MT and complete NE in HeLa cells, but showed possible exceptions, i.e.,
MT apparently penetrating the NE. The authors conceded, however, that
this could have been a false impression created by superposition.
I have not studied MT in the midbody in detail. It is possible
that most of the "continuous" MT terminate in this body, or near it,
but Figure 55a strongly suggests that at least some MT pass from one
half-spindle into the other. It is interesting to note the abundance
of MT converging in the midbody of cells in late cytokinesis, after the

235
spindle proper has been disorganized (Figures 61 and 62). Apparently
these MT are nore stable than others.
Chromosomes
Aberrations
The frequency of abnormal metaphase cells in the controls is
obviously exaggerated (Table 1). Many late prometaphase cells with
maloriented chromosomes (e.g., Figure 29) were scored as aberrant
metaphases, but most of these cells would probably have divided
normally, otherwise the frequency of aberrations in anaphase cells
would have been similar to that in metaphase cells. Nevertheless, the
frequency of aberrations may reflect instabilities of the cell line,
or it may be due to the presence in the culture of a biological
clastogen (virus or other microorganisms). Ualen (1965) reported
sister chromatid exchanges in the original culture of rat kangaroo
cells. Similarly, Levan (1970) described changes of chromosome number
and structure in the PtK^ line derived from a female rat kangaroo. He
interpreted these changes to reflect adaptation of the cells to the in
vitro environment. Heneen (1970) studied frequency and types of
aberrations in untreated PtK^ cells. Many of these were similar to
aberrations I observed in the controls.
The lower of the two concentrations of SN used (0.01 pg/ml) had
little mitodepressive effect (Table 2), but induced almost 507. aberra
tions (Table 3). In contrast, 0.05 ig/ml SN had a great mitodepressive
effect (Table 2), which accounts for the small sample of cells examined
for aberrations (Table 3). These results confirm the great potency of
SN qs a clastogen.

236
The question whether the SN-treated abnormal cells represented
the first or second generation after treatment cannot be answered.
The 48 hr recovery period was long enough for one round of divisions,
but the nitodepressive and general cytotoxic effects most likely
altered the division cycle. However, it is very unlikely that any of
the aberrations were induced in GIndeed, no subchromatid aberra
tions were observed.
Fine Structure
The analysis of thin sections of dicentric bridges revealed little
about the composition of the chromosomes involved. A comparison of the
number of fibers in dicentrics attenuated to a variable degree might
have provided some insight into the coiling or lateral association of
the fibers, but this was not feasible.
Electron micrographs such as Figures 27 and 30b demonstrate that
in thin sections individual chromosomes do not always appear physically
separated. This posed a certain problem for the analysis of dicentric
bridges such as shown in Figures 99-101. In some of the serial
sections (e.g., Figure 100) the impression was not one of two truely
linked chromosomes, but of long chromosomes whose telomeric ends were
twisted around each other. The attenuation of the bridge (Figure 99)
and the presence of acentric fragments (Figures 99 and 101), however,
were evidence for a real exchange.
In the phase contrast microscope the chromosomes in many of the
cells treated with 0.05 pg/ml SN appear fainter and more slender than
normal chromosomes. Examination of thin sections of these chromosomes
revealed generally looser, but also more irregular packing of fibers
(Figure 111). Most likely, these chromosomes condensed during the

237
recovery period, not during exposure to SN. It therefore appears that
high concentrations of SN alter the bonding properties of the molecules
involved in coiling of chromatin.
Interesting is that high concentrations of colcemid produce a
similar effect (Figures 65 and 66). Considering the short duration of
the treatment (15 min) it is more likely that colcemid dispersed
already condensed chromosomes rather than inhibiting the process of
condensation. Morphological alterations of chromatin in mitotic grass
hopper neuroblasts by colchicine was reported by Gaulden et al. (1970).
Colcemid and colchicine are widely used, often at concentrations
similar to those altering chromosome structure, for the accumulation
of metaphase cells destined to yield a great number of chromosomes for
ultrastructural studies (e.g., Abuelo and Moore 1969). The implica
tions are clear: not only do we have to consider artifacts produced
in whole-mounted chromosomes by the preparation techniques, but the
preceding colcemid or colchicine treatment may have completely altered
the ultrastructure of the chromosomes.
Remarkable is the appearance of centromere granules in chromo
somes of cells that were probably in late prophase or in prometaphase
at the time of treatment (Figure 65). I could not definitely determine
whether there are one or two granules per chromatid. The structure
and position of these granules is very similar to those observed by
Stubblefield and Uray (1971) in whole-mounted Chinese hamster chromo
somes. In the latter case, however, the possibility of an artifact
produced by colcemid is slight, because at the concentration used
(0.06 jug/ml for 4 hr) the drug does not even disperse all the MT
(Brinkley et al. 1967).

238
To my knowledge chromatin strands connecting daughter chromosomes
in early anaphase (Figures 46-48, and 109) have not been previously
reported. In random sections of mitotic cells these strands, as well
as the cells themselves at this stage, can easily escape detection.
Nevertheless, it remains to be proven that this is a universal
phenomenon or that it is restricted to rat kangaroo cells. The densely
fibrillar structure of the strands is most easily explained as a
consequence of stretching. Attenuation could produce parallel align
ment of the fibers involved, which would allow tighter packing than
is possible for coiled fibers. This explanation is supported by
similarly packed fibers in dicentric bridges, where stretching cer
tainly occurs (Figures 110 and 111a).
DuPraw (1965, 1968) proposed with his folded-fiber model of
chromosomes that sister chromatids at metaphase are held together by
short segments of unreplicated ESTA in the centromere region.
Immediately preceding anaphase separation these segments would be
replicated. Numerous autoradiographic studies have demonstrated late-
labeling centromere regions in chromosomes of various cell lines (see
DuPraw 1968 for references), but the short burst of DMA synthesis prior
to anaphase, expected according to DuPraw's model, has not been'proven.
Be this as it may, the fact is that sister chromatids in whole-
mount preparations are connected in the centromere region by chromo
somal fibers (e.g., Abuelo and Moore 1969, DuPraw 1968, Stubblefield
and Uray 1971). It is possible that the chromatin strands I observed
in early anaphase cells are a consequence of the physical separation of
these connections. Logically, the strands can be detected in thin
sections only when the centromere regions are stretched, because in the

239
relaxed state they blend in with the other fibers. This could explain
the occurrence of similar strands of more or less parallel fibers in
stretched prometaphase chromosomes (Figures 31 and 36). The signi
ficance of chromatid connections for stable metaphase orientation of
chromosomes is clear, assuming that orientation stability depends on a
balance of pulling forces acting on sister kinetochores oriented
11
towards opposite poles. This theory was proposed by Ostergren (1951)
for meiosis, and experiments by Henderson and Koch (1970), Nicklas
(1967), and Nicklas and Koch (1969) with grasshopper spermatocytes
strongly support it. Connections between sister chromatids prevent
premature separation, which would lead to unipolar attachment of
chromatids and, consequently, unstable orientation followed by
haphazard chromosome distribution.
The surprisingly frequent occurrence of very fine nuclear bridges
in SN-treated cells in cytokinesis indicates that not all the dicentric
bridges rupture during anaphase (Figures 118 and 119a). Acentric frag
ments form micronuclei if included in daughter cells (Figure 119b).
In one case examined in the electron microscope, however, I found an
"elimination body," consisting of a polymorphic micronucleus and very
little cytoplasm, in the cleavage furrow between daughter cells. This
body possibly contained acentric fragments that were lying near the
equator at the periphery of the cell in metaphase, as shown in Figures
108c and lOSe.
Osmiophilic granules in chromosomes of PtK^ cells have been re
ported by Brinkley and Shaw (1970). I have found these granules in
interphase and all stages of mitosis. They are more clearly visible,
possibly more numerous, in cold-treated than in control cells (Figure 71).

240
Their significance is not at all clear. In size they are distinctly
different (approximately 300-500 A in diameter) from ribosomes and the
granular component of the nucleolus.
Kinctochores
Structural Changes During liitosis
The terra "maturation" suggested by Jokelainen (1968) best de
scribes the structural differentiation of kinetochores from prometa-
to metaphase. 3y definition a mature kinetochore appears in thin
sections as the typical triple-banded structure at the surface of
metaphase chromosomes.
Kinetochores appear "out of nowhere" at the primary constriction
of chromosomes (Figures 10, 11a, 15-17). I could not detect any
relationship between primary constrictions and HE, which nevertheless
does not rule out the possibility that other points of attachment
determine the coiling pattern of chromosomes (see Comings 1968, Comings
and Okada 1970a, b). In stages of prophase earlier than illustrated
in Figure 10 kinetochores are not visible as such, but their future
position is indicated by a slight constriction on the loosely coiled
chromosomes. From mid- to late prophase the electron density of
kinetochores increases slightly, making their detection easier. In any
event, however, the chromosomes always stain much more intensely than
the kinetochores (e.g., Figure 16).
Radical changes in kinetochore structure occur during prometa
phase. These changes first involve the kinetochores of chromosomes
near the centrioles, i.e., in the region where the NE breaks down
first (Figure 23). In cells' such as the one shown in Figure 19 one

241
therefore finds immature prophase kinetochores on the chromosomes in
the center of the "nucleus," and more mature kinetochores near the two
poles. 1 could not establish a clear-cut correlation between attach
ment of kinetochores to MT and the appearance of the three bands
typical for mature kinetochores. On the contrary, kinetochores not
attached to MT often exhibit more distinct bands than attached
kinetochores (e.g., Figure 23). However, it is possible that attached
kinetochores are triple-banded, but since they are drawn out into a
cone or more odd structure (Figures 23, 24, and 36) they would
necessarily be cut obliquely by any but ideal median sections and con
sequently the bands would be less distinct. In fact, Figure 23
indicates that these kinetochores are banded. Much of the foregoing
also applies to structural differences between sister kinetochores of
maloriented chromosomes in late prometaphase (Figures 28, 37, and 38).
Figure 120, illustrating asynchronous maturation during prometa
phase of sister kinetochores of a maloriented chromosome, is based on
the above mentioned assumption and observations. Kinetochore no. 1
possibly had a structure similar to that of kinetochore no. 2 in Figure
120b at a stage preceding that shoi/n in Figure 120a. The transition of
kinetochore no. 2 from Figure 120a to Figure 120b is hypothetical.
This kinetochore matured later than its sister oriented towards the
near pole and attached to MT very early. A similar sequence can be
postulated for equatorial chromosomes with one unobstructed and one
obstructed kinetochore. If both kinetochores of an equatorial chromo
some are unobstructed they mature synchronously (e.g., Figure 25).
It appears that the outer layer is formed within the finely
fibrillar material of prophase and immature prometaphase kinetochores

Fig. 120. Diagrammatic representation of kinetochore maturation
during prometaphase, (a) The chromosome lies near one pole. The
kinetochore oriented towards this pole (Kq_) is drawn out to a cone
and attached to MT. The outer layer has possibly formed. The kineto
chore oriented towards the far pole (Kg) is less mature. As yet
none of the layers has formed and the kinetochore is not attached
to MT. Compare with Fig. 2k. (b) The outer layer of Kg is forming,
but the kinetochore is still not attached to MT. Kj_ is unchanged.
The chromosome is still near the same pole. Compare with Figs. 23
and 37* (c) Kg has reached the stage of maturity corresponding to
in (a); it is attached to MT. The inner layer of is forming.
Both kinetochores are stretched and the chromosome is moving towards
the equator (to the left). Compare with Fig. 40. (d) Both kinetochores
are nearly mature, showing three layers. Both are also stretched. The
chromosome has attained stable bipolar orientation at the equator.
Compare with Fig. 3^

2 A3

244
(Figures 23 and 33). This material is gradually reduced in extent, but
some of it remains present in mature kinetochores as the corona and the
middle layer (Figures 42, 46, 78, and 84). The inner layer, at the
surface of the chromosome, appears last (Figures 33-35).
Kinetochore maturation as described herein essentially agrees
with the observations of Jokelainen (1965). However, his conclusions
that sister kinetochores always mature asynchronously does not apply to
PtK^ cells. A better knowledge of the position of chromosomes in the
spindle and the orientation of sister kinetochores allowed detection of
subtle differences between kinetochores of maloriented chromosomes, as
well as obstructed and unobstructed kinetochores of equatorial chromo
somes .
There is virtually no change in kinetochore structure from meta
phase to mid-anaphase (Figures 42, 43, 45-49). "Dedifferentiation" of
kinetochores begins in late anaphase (Figures 52 and 53). The triple
layered profiles appear less distinct. Concomitantly the number of
kinetochore MT is reduced. During telophase, the reforming NE
separates the remnants of inner and outer layers (Figures 59 and 60).
The former is lost in the cytoplasm, while the latter comes to lie
closely apposed to the inner membrane of the NE (see also Figures 115
and 116). The most convincing evidence for this conclusion came from
an aberrant cell similar to the one in Figure 116, but from a TEPA-
treated culture. The inner layers of several kinetochores, also
situated in deep invaginations of the NE on the polar side, appeared as
very electron dense bumps on the inner membrane of the NE. On the out
side of the outer membrane, opposite these bumps, remnants of the less
opaque outer layers with MT still attached were visible.

245
The Fine Structure of Mature Kinetochores
Several arguments based on published observations and electron
micrographs speak against the filament model of mammalian kinetochores
proposed by Brinkley and Stubblefield (1966, 1970). First, the kireto-
chores in colcemid-treated Chinese hamster cells, on which the original
paper was almost exclusively based (Brinkley and Stubblefield 1966),
are structurally different from kinetochores of normal metaphase
chromosomes. The inner kinetochore layer is absent in c-mitotic cells,
but it is present on metaphase chromosomes in cells recovering from
colccmid (Brinkley et al. 1967). Secondly, random sections of mitotic
cells, particularly if embedded as a pellet (as were the cells used by
Brinkley and Stubblefield 1966), should yield at least some cross
sections of kinetochore filaments. Such profiles have never been
presented. Thirdly, the difference in electron density between outer
and inner bands of normal metaphase kinetochores, observed by a number
of investigators (e.g., Bamicot and Huxley 1965, Luykx 1965a),
suggests the two bands do not represent identical filaments. Fourthly,
Jokelainen (1967) and McIntosh and Landis (1971) have published electron
micrographs of para-equatorial sections of metaphase cells showing
circular kinetochores. Krishan (1968) also presented a micrograph
showing a circular kinetochore in a telophase cell after recovery from
vinblastine. Finally, kinetochores in a variety of cells treated with
the spindle poisons vincristine and vinblastine are virtually identical
with kinetochores in colcemid-treated Chinese hamster cells, and
equally different from normal metaphase kinetochores (George et al.
1965, Journey et al. 1965, Journey and Uhaley 1970, Krishan 1968).

246
Extending their model to PtK^ rat kangaroo cells, Brinkley and
Stubblefield (1970) interpreted the outer and inner bands of metaphase
kinetochores as representing the two kinetochore filaments. My
observations on paraxial (e.g., Figures 78-85) and para-equatorial
(Figures 86, 87, and 91) serial sections leave no doubt that mature
kinetochores in PtK^ cells are trilaminar, roughly circular plates at
the surface of chromosomes. I have not seen any filaments in para-
equatorial sections. Furthermore, kinetochore MT are arranged in
bundles, not in "sheets," as would be expected were they connected to
filaments.
My view of mature kinetochores essentially agrees with the model
of Jokelainen (1967), except for one important aspect: kinetochore
plates are very seldom, if ever, flat. Rather, they are undulated
(Figures 42 and 82), concave (Figure 85), or convex (Figures 46, 47,
and 84) discs. Figure 121a-c diagrammatically represents three-
dimensional reconstructions of kinetochores of various shapes. The
shape of a kinetochore is certainly determined by the curvature of the
chromosomal surface at this locus. It is also possible that both the
shape of the kinetochore and the curvature of the chromosome are
influenced by the differential pulling action of the attached MT.
Ultrathin paraxial sections (Figures 82 and 84) strongly suggest
that kinetochore MT terminate in the outer layer. Even more convincing
are para-equatorial sections (Figures 86 and 87). Nebel and Coulon
(1962) and Luykx (1965a) concluded from their micrographs that kineto
chore MT terminate in the inner layer. These observations, however,
were severely hampered by inadequate preservation of fine structure.
Jokelainen (1967) also concluded that kinetochore MT are anchored in

Fig. 121. Diagrammatic representation of the three-dimensional
structure of kinetochores. (a) Mature metaphase kinetochore
(compare with Figs. 78, 82, and 86). (b) and (c) anaphase
kinetochores (compare with Figs. k6-k$¡). (d) Sister kinetochores
of a telocentric chromosome in a colcemid-treated cell (compare
with Fig. 95)* (e) Sister kinetochores of a metacentric or sub-
metacentric chromosome in a colcemid-treated cell (compare with
Figs. 92-9, 96-98).

248

249
the inner layer or, possibly, even deeper in the chromosome. But his
sections were more than 750 A thick and the apparent deep penetration
of MT could well be ascribed to a superposition effect. In the
micrograph of an early anaphase kinetochore published by Robbins and
Gonatas (1964, Figure 19) the MT seem to terminate in the outer layer.
Intuitively one would expect a certain uniformity regarding MT
attachment in cells of different organisms, especially among mammals.
I believe that a careful and detailed reexamination of the cases
apparently contradicting my conclusions would reveal this uniformity.
The purpose of the colcemid treatments was to check the possi
bility that kinetochores of PtK^ cells under these conditions are
different from kinetochores in similarly treated Chinese hamster cells.
This was not the case (Figures 92, 94, and 96-98). The inner layer is
lacking and the outer layer structurally resembles that in c-mitotic
Chinese hamster cells, as well as cells treated with vincristine and
vinblastine (George et al. 1965, Journey et al. 1968, Journey and
Whaley 1970, Krishan 1968). Grazing sections of chromosomes yielded
images compatible with the idea that the outer layer is a circular or
oval plate in a depression of the chromatid (Figures 93, 96, and 121e).
Similar images were obtained by Journey (personal communication) from
vincristine-treated Chinese hamster cells.
Kinetochore profiles in transverse sections of chromosomes in
colcemid-treated PtK2 cells (Figures 92, 94, 97, and 98) suggest the
plate-like outer layer is bilaminar. This substructure is not
discernible in normal, mature kinetochores (e.g., Figures 82 and 84),
but it can be seen in immature prometaphase kinetochores (e.g., Figures
23c and 41). To me these observations indicate that immature
M

250
proraetaphase kinetochores and kinetochores in colcenid-treated cells
are less condensed than mature kinetochores. The generally greater
diameter of kinetochores and the more extensive matrix in c-mitotic
compared to untreated metaphase cells further supports this interpre
tation.
Intuitively one expects a larger plate to be more easily bent than
a smaller one, particularly if it is also less rigid, as might be the
case with the less condensed outer kinetochore layer in c-mitotic
cells. With this in mind we can explain the strange kinetochore
profiles in Figure 95. They represent oblique sections of the highly
undulated outer layer (Figure 121d). Undulations are more likely to
occur on telocentric chromosomes, or on submetacentric chromosomes bent
at a sharp angle in the kinetochore region. Either of these
possibilities could apply to Figure 95. The near-circular kinetochore
profiles observed by Journey and Uhaley (1970) in vincristine-treated
HeLa and Chinese hamster cells could be the result of superficial
sections through cap-shaped kinetochores such as K in Figure 92, or
they represent near-terminal sections of subtelocentric or telocentric
chromosomes (as in Figure 12Id).
Brinkley and Stubblefield (1966) observed 50-80 A fibrils
apparently looping out from the main band of kinetochores in Chinese
hamster cells recovering from colcemid. I have not observed such loops
in colcemid-treated PtIC2 cells, but the matrix in which the outer
kinetochore layer is embedded does contain fibrils (e.g., Figure 94).
The fibrils observed by Brinkley and Stubblefield (1966) are possibly
also part of the matrix, or they could be characteristic for cells in
which the kinetochore recondenses and the number of kinetochore UT
increases during recovery from colcemid.

251
I more or less expected the kinetochores in cold-treated cells to
resemble those in colcemid-treated cells. This was not the case
(compare Figure 75 with Figure 92). One significant difference was the
presence of kinetochore MT in cold-treated and their complete absence
in colcemid-treated cells. As already explained in the s'ection on
kinetochore maturation, it is not clear if a relationship exists
between microtubular attachment and the appearance of triple-layered
kinetochores. Observations on cold-treated cells might suggest such a
relationship, but in colcemid-treated Chinese hamster cells the inner
layer of kinetochores attached to persisting MT is apparently also
lacking (Brinkley and Stubblefield 1966).
Definite conclusions regarding the molecular composition of
kinetochores are not possible without extensive cytochemical studies.
The fact that the outer layer stains differently than the inner layer
suggests a different molecular composition. I tentatively conclude
that the outer layer is not made of chromatin, and that the inner layer
consists of tightly packed chromatin fibers. This rules out covalent
bonds between the two layers. The outer layer possibly consists of
protein. Since kinetochore MT are attached to the outer layer, the
implications are clear: the bonds between MT and the outer layer on
one hand, and outer and inner layers on the other hand must be very
strong, in fact stronger than the bonds maintaining chromosomal fibers
in a coiled state. This is borne out by the observations that kineto
chores of dicentric chromosomes subjected to considerable pull (Figures
101, 110a, Ilia, and 113) are structurally normal. Also, centrifuga
tion forces sufficient to uncoil chromosomes do not rupture spindle
attachments (Shimamura 1940). How this strong bonding is achieved

252
remains unexplained, but the matrix persisting as the corona and
middle layer of mature kinetochores may play an important role.
Kinetochore-Hicrotubule Interactions
Kinetochores are involved in two important processes during
prometaphase: Chromosome orientation and congression. A third
possible function, assembly and orientation of IfT, is central to most
models of mitosis (Dietz 1969, Inou 1964, Luylcx 1970, McIntosh et al.
1969), but the available evidence is not entirely conclusive (for
detailed discussions see Luykx 1970, Nicklas 1971). Neither light
microscopic nor ultrastructural studies alone can answer the questions
remaining. A cinemicrographic and electron microscopic analysis of the
effect of UV microbeam irradiation and spindle poisons is more likely
to yield significant data. The electron micrographs of PtK^ cells in
very early prometaphase (Figures 18a and 19) can be interpreted as
supporting either the idea that kinetochores organize 1T or that they
attach to existing ITT. Counts of the number of MT before and after
kinetochore attachment, as presented by Mantn et al. (1969a, b) for
the diatom Lithodesmium, might provide clues, but spindle formation in
PtK^ cells is more complex.
Chromosome orientation has been extensively investigated in the
first meiotic division in spermatocytes (e.g., Nicklas 1967). There,
spindle attachments of chromosomes are often broken naturally, or can
be broken experimentally, and subsequently reorientation of chromosomes
and reattachnent occur (see also Nicklas and Koch 1969). This is
necessary for stable bipolar orientation of bivalents initially
attached to one pole only (maloriented bivalents). Mo similar data are
available for mitosis. Host of the current models of mitosis, which,

253
of course, are also designed to explain raeiosis, picture the chromo
somes as orienting within the more or less completely formed central
spindle (Dietz 1969, Luykx 1970, McIntosh et al. 1969). The MT
growing out from the kinetochores are supposed to interact with the
already present "continuous" MT of the central spindle. Whether this
is indeed so in the cells considered by the authors remains to be
proven, because no ultrastructural studies of prometaphase have been
presented. In PtK^ cells the above is certainly not the case, as
demonstrated by Figure 23, and similar micrographs not included.
These pictures clearly show that MT do not grow out from, or attach to,
both kinetochores simultaneously, and that chromosomes may connect to
a pole as soon as the nuclear envelope breaks down, regardless of
whether the central spindle is fully formed or not.
stergren (1951) postulated a pulling theory of chromosome
orientation and congression for meiosis. Briefly, this theory
proposes that bipolar orientation of bivalents is the consequence of
kinetochore polarity. This polarity allox^s each kinetochore to connect
only to that pole towards which it is oriented. Prometaphase movements
are caused by random activity of kinetochores. Ultimately, the
kinetochores exert a pull and the equatorial position of chromosomes at
late prometaphase and metaphase is the result of balanced tension on
kinetochores oriented towards opposite poles.
The following reconstruction of events leading to chromosome
orientation and congression in PtK^ cells fits this theory well, if it
is modified to account for differences between meiotic chromosomes
(bivalents) and mitotic chromosomes (pairs of sister chromatids). As
soon as the nuclear envelope breaks down near the centrioles, the

254
chromosomes near this site become attached to MT by one kinetochore.
Kinetochores are polarized in the sense that the outer, but not the
inner layer, can connect with MT. Therefore, attachment is possible
only to MT approaching from the side the kinetochore faces. Even if a
kinetochore could theoretically attach to MT approaching from the
opposite side ("from the back"), these MT would have to be bent at an
impossible angle. Consequently, for the meta- and submetacentric
chromosomes the initial unipolar attachment immediately establishes
bipolar orientation, especially if combined with tension. Therefore,
when one kinetochore is attached, the other has no choice but to attach
to the opposite pole. Telocentric chromosomes are more likely to
attach with both kinetochores to the same pole. In fact, the only
chromosome I observed displaying this configuration was a telocentric
chromosome in a TEPA-treated cell in mid-prometaphase. Bajer and Mol-
Bajer (1969), however, published a micrograph showing a metacentric
chromosome apparently attached to the same pole with both kinetochores.
Such observations are rare and, though almost certainly due in part to
neglect by many investigators, may reflect the rarity of the event.
For comparison, in crane fly meiosis ten per cent of the bivalents show
unipolar orientation in early prometaphase (Bauer et al. 1961).
If the two poles are still close together at the time the first
attachments are established, one might expect some kinetochores to
connect to both poles. I have never observed this, but Bajer and Mol-
Bajer (1969) reported such a case for Haemanthus endosperm.
It is possible that early unipolar attachments involving only one
kinetochore are maintained during migration of the centrioles and
formation of the definitive spindle axis, with the result that some
r

255
chromosomes lie close to one pole in mid- and late prometaphase (e.g.,
Figures 26-29). Also, these unipolar attachments may be followed by
movement of the chromosome involved towards the near pole. At some
point, however, the sister kinetochore oriented towards the far pole
also becomes attached, and this is followed by movement towards the far
pole (Figure 120c). Figure 40 possibly illustrates the "return point,"
i.e., two chromosomes just beginning to move towards the equator. Once
the tension exerted on sister kinetochores is balanced, the chromosomes
remain stably in an equatorial position.
Chromosomes lying farther from the poles at the time the nuclear
envelope breaks down become attached later than the chromosomes lying
close. Whether this is so because of the greater distance from the
probable pool of MT subunits around the centrioles, or because they are
shielded from the poles by other chromosomes and therefore cannot form
connections until after reshuffling due to movements of already
attached chromosomes has occurred, remains an open question. Neverthe
less, the chance that the kinetochores of these chromosomes attach to
opposite poles simultaneously is greater, because the spindle axis is
established. For example, the chromosome in Figure 24 could have
attached earlier than the chromosome in Figure 25. Consequently the
latter shows bipolar orientation and attachment, whereas the former
shows bipolar orientation, but unipolar attachment. Equatorial
chromosomes immediately attached to both poles can be expected to be
stabilized by tension and to move very little during prometaphase.
However, if one sister kinetochore of an equatorial chromosome is
obstructed, unipolar attachment would precede bipolar attachment.
Again, the latter'could occur only after reshufflinghas cleared the

256
field between the previously unattached kinetochore and its pole. As a
consequence of unipolar attachment the chromosome should travel towards
the pole to which it is attached. This is probably not so, otherwise
many more chromosomes could be found near the poles. Possible ex
planations are connections between non-sister kinetochores (perhaps
between the kinetochores of chromosomes no. 1 and 2 in Figure 33; see
also Bajer 1970, Luykx 1965b), or simply physical hindrance. In fact,
neighbor chromosomes can lie so close to each other that a separation
is not detectable in thin sections (Figure 27).
Chromosomes without definite attachment to either pole (Figure 39)
may simply drift in the spindle or remain stationary until one or both
of its kinetochores become attached. Their subsequent behavior would
depend on the nature of the attachments.
If my interpretation is correct, detachment, reorientation, and
reattachment should occur less frequently for mitotic than for
meiotic chromosomes. This awaits experimental verification. Not only
does the above account agree with Ostergren's theory (1951), but
Nicklas (1967) clearly demonstrated that in grasshopper spermatocytes
kinetochore position after detachment by micromanipulation determines
to which pole the kinetochore will orient and attach. Finally,
stabilization of chromosomes by tension was beautifully demonstrated
by Henderson and Koch (1970), Henderson et al. (1970), and Nicklas and
Koch (1969).
The hypothetical points of my interpretation can be tested
experimentally. Time-lapse cinemicrographic records of the prometa
phase behavior of selected chromosomes followed by an ultrastructural
analysis of kinetochore structure and attachments to MT could easily
reveal which of my assumptions are true or false.

257
Nuclear Envelope
Fragmentation and reconstruction of the nuclear envelope (ME) in
PtK^ cells are similar in many ways to the processes in other mammalian
cells. However, 1 have neither observed stacking of ME fragments as
reported for rat lymphocytes (Murray et al. 1965) and rat hepatoma
cells (Chang and Gibley 1968), nor "polar caps" consisting of numerous
cistemae and vesicles as in HeLa cells in metaphase (Robbins and
Gonatas 1964). There are no membrane elements at metaphase and early
anaphase that could be identified as fragments of the NE (Figures 42,
43 45, and 47). Cistemae possibly representing elements of the
reforming NE occur only in late anaphase at the periphery of the
spindle (Figure 51a). In very late anaphase and early telophase such
cistemae can be clearly identified as portions of the NE mainly
because of their association with the chromosomal mass (Figures 52 and
57).
Comings and Okada (1970a) have described associations of NE
fragments with meta- and anaphase chromosomes as seen in whole mounts.
Their observations can be criticized for two reasons: (1) Most of the
preparations were made from colccmid-arrested cells, and (2) anaphase
was simply judged according to time after colcemid reversal. According
to the authors, however, the results agreed with Comings' hypothesis
(1968; also Comings and Kakefuda 1968, Comings and Okada 1970b, c) that
persisting chromosomal attachments to the NE account for reproducible
patterns of DMA replication and chromosome folding. This is a
tempting hypothesis, but evidence is rather scarce and unconvincing.
In PtK^ cells the reconstruction of the NE seems to be related
spatially to the chromosomal mass as a whole, rather than to specific
sites of individual chromosomes (Figures 52 and 57).

258
The possibility that endoplasmic reticulum (ER) is involved in the
reconstruction of the NE was raised by Porter and Hachado (1960), based
on studies of mitosis in onion root tip cells. Bajer and Mole-Bajer
(1969; Haemanthus endosperm cells) and Robbins and Gonatas (1964;
HeLa cells) agreed that this is a likely possibility. My own observa
tions support this concept (Figures 51a, 52, 112, and.117). Quite
frequently, cistemae xjith ribosomes can be seen apposed to the
chromosomal mass (Figure 52). Other cistemae, in a similar associa
tion with chromatin, appear continuous with rough ER.
The serial sections yielded numerous oblique views of NE fragments
in prometaphase, thus providing a better insight into the fate of
nuclear pore complexes. There is no doubt that the pores disappear
(Figure 21). Remarkable is their reappearance on very small cistemae
associated with chromatin in late anaphase (Figure 58). One gets the
impression that pores are reformed almost immediately upon contact of
cistemae with chromatin, but the process is shrouded in mystery.

CONCLUSIONS
Knowing the question is knowing half the answer. Applied to
science this means if we are able to state a problem and to formulate a
hypothesis, we are able to design experiments with the aim to gather
evidence for or against the hypothesis, or to reach new ideas. Instead
of collecting data more or less at random we could then proceed in an
organized fashion to solve a particular problem.
From the preceding chapters it is clear that not all my expecta
tions stated in the preface were fulfilled. The study of chromosomal
ultrastructure took second place to the investigation of mitosis.
Biophysical and biochemical studies are more likely to provide a better
understanding of such interesting and important problems as polynemy
or uninemy and the orderly coiling of chromosomes, than ultrastructural
studies alone.
My observations on prometaphase and early anaphase of mitosis,
previously neglected by most investigators, have revealed interesting
new data. But again, electron microscopy alone, although a powerful
tool, has its limits. I am firmly convinced, however, that a concerted
and organized effort combining _in vivo observation, particularly
polarization microscopy and cinenicrography, with electron microscopy
will clarify many of the still mysterious aspects of mitosis.
259

REFERENCES
Abuelo, J. G., and D. E. Moore. 1969. The human chromosome. Electron
microscopic observations on chromatin fiber organization. J. Cell
Biol. 1*1: 73-90.
Aikawa, M., and R. L. Beaudoin. 1968. Studies on nuclear division of
a malarial parasite under pyrimethamine treatment. J. Cell Biol.
39: 7^9-75^.
Aldrich, H. C. 1967. The ultrastructure of meiosis in three species
of Physarum. Mycologia 59: 127-148.
Aldrich, H. C. 1969. The ultrastructure of mitosis in myxamoebae
and plasmodia of Physarum flavicomum. Amer. J. Bot. 56: 290-299
Anonymous. 1967. American Type Culture Collection Cell Repository.
Registry of animal cell lines certified by the cell culture
collection committee, suppl. 2.
Bacq, Z. M., and P. Alexander (eds.). 1955 Radiobiology Symposium
1954. Academic Press, New York. 3^2 p.
Bajer, A. 1958 Cine-micrographic studies on chromosome movements
in ^-irradiated cells. Chromosoma 9: 319331
Bajer, A. 1963 Observations on dicentrics in living cells.
Chromosoma 14: 18-30.
Bajer, A. 1964. Cine-micrographic studies on dicentric chromosomes.
Chromosoma 15: 630-651.
Bajer, A. 1965 Subchromatid structure of chromosomes in the living
state. Chromosoma 17: 291-302.
Bajer, A. 1968. Behavior and fine structure of spindle fibers dur
ing mitosis in endosperm. Chromosoma 25: 249-281.
Bajer, A. 1969 Effects of UV microbeam irradiation on chromosome
movements and spindle fine structure (with film demonstration).
Biophys. J. 9: A-151.
Bajer, A. 1970. See Luykx, 1970, p. 54.
Bajer, A. and R. D. Allen. 1966. Structure and organization of the
living mitotic spindle of Haemanthus endosperm. Science 151:
572-574.
260

261
Bajer, A., and J. Mol-Bajer. l$6l. UV microbeam irradiation of
chromosomes during mitosis in endosperm. Exp. Cell Res. 25:
251-267.
Bajer, A., and J. Mol-Bajer. 1963. Cine analysis of some aspects of
mitosis in endosperm, p. 357-409* In G. G. Rose (ed.) Cinemicro-
graphy in Cell Biology. Academic Press, New York.
Bajer, A., and J. Mol-Bajer. 1969 Formation of spindle fibers,
kinetochore orientation, and behavior of the nuclear envelope
during mitosis in endosperm. Fine structural and in vitro
studies. Chromosoma 27: 448-484.
Barnicot, N. A. 1966. A note on the structure of spindle fibers.
J. Cell Sci. 1: 217-222.
Barnicot, N. A., and H. E. Huxley. 1965. Electron microscope
observations on mitotic chromosomes. Quart. J. Microscop.
Sci. 106: 197-214.
Bauer, H., R. Dietz, and C. RSbbelen. 1961. Die Spermatocytenteilungen
der Tipuliden. III. Mitteilung. Das Bev/egungsverhalten der
Chromosomen in Translokationsheterozygoten von Tipula olercea.
Chromosoma 12: U6-I89.
Behnke, 0., and A. Forer. 1966. Some aspects of microtubules in
spermatocyte meiosis in a crane fly (llephrotoma sutural is Loev):
intranuclear and intrachromosomal microtubules. C. R. Trav.
Lab. Carlsberg 35: 437-455-
Belr, K. 1929* Beitrage zur Kausalanalyse der Mitose. II.
Untersuchungen an den Spermatocyten von Chorthippus (Stenobothrus)
lineatus Panz. Wilhelm Roux Arch. Entwicklungsmech. Organ.
118: 359-484.
Bernhard, W., and E. de Harven. i960. L*ultrastructure du centriole
et d'autres elements de lappareil achromatique. Proc. 4th
Int. Conf. Electr. Microscop. 2: 217-227.
Bessis, M., J. Breton-C-orius, and J. P. Thiry. 1958. Centriole,
corps de Golgi et aster des leucocytes. Etude au microscope
lectonique. Rev. Hematol. 13: 363-386.
Biesele, J. J. 1962. Experimental and therapeutic modification of
mitosis. Cancer Res. 22: 779737.
Bloom, W., R. E. Zirkle, and R. B. Uretz. 1955 Irradiation of
parts of individual cells. III. Effects of chromosomal and
extrachromosomal irradiation on chromosome movements. Ann.
N. Y. Acad. Sci. 59: 503-513.
Borisy, G. G., and E. V/. Taylor. 1967a. The mechanism of action of
colchicine. Colchicine binding to sea urchin eggs and the mitotic
apparatus. J. Cell Biol. 34: 535-548.

262
Borisy, G. G., and E. W. Taylor. 196Tb. The mechanism of action of
colchicine. Binding of colchicine-3H to cellular protein. J.
Cell Biol. 34: 525-533-
Brandham, P. E. 1970- Chromosome behaviour in the Aloineae. III.
Correlations between spontaneous chromatid and sub-chromatid
aberrations. Chromosoma 31: 1-17-
Brinhley, B. R., and J. Cartinright, Jr. 1970. Organization of micro
tubules in the mitotic spindle: differential effects of cold
shock on microtubule stability. J. Cell Biol. 47: 25a.
Brinkley, B. R., and R. M. Humphrey. 1969* Evidence for subchromatid
organization in marsupial chromosomes. I. Light and electron
microscopy of X-ray-induced side-arm bridges. J. Cell Biol.
42: 827-836.
Brinkley, R. R., P. Murphy, and L. C. Richardson. 1967. Procedure
for embedding in situ selected cells cultured in vitro. J.
Cell Biol. 35: 279-283.
Brinkley, B. R., and R. B. Uicklas. 1968. Ultrastructure of the
meiotic spindle of grasshopper spermatocytes after chromosome
micromanipulation. J. Cell Biol. 39: l6a-17a.
Brinkley, B. R., and M. W. Shaw. 1970. Ultrastructural aspects of
chromosome damage, p. 313-3I5. In Genetic Concepts and Neoplasia.
Williams and Wilkins, Baltimore.
Brinkley, B. R., and E. Stubblefield. 1966. The fine structure of
the kinetochore of a mammalian cell in vitro. Chromosoma 19:
28-43.
Brinkley, B. R., and E. Stubblefield. 1970. Ultrastructure and inter
action of the kinetochore and centriole in mitosis and meiosis,
P- 119-185- In D. M. Prescott, L. Goldstein, and E. McConkey
(eds.) Advances in Cell Biology, vol. 1. Appleton-Century-
Crofts, New York.
Brinkley, B. R., E. Stubblefield, and T. C. Hsu. 1967. The effects
of colcemid inhibition and reversal on the fine structure of the
mitotic apparatus of Chinese hamster cells in vitro. J. Ultrstruct.
Res. 19: 1-18.
Britten, R. J., and D. E. Kohne. 1968. Repeated sequences in DHA.
Science 161: 529-540.
Buck, R. C. 196( Mitosis and meiosis in Rhodnius prolixus: The
fine structure of the spindle and diffuse kinetochore. J.
Ultrastruct. Res. 18: 189-501.
Burgos, M. H., and D. W. Fawcett. 1956. An electron microscope study
of spermatid differentiation in the toad, Bufo arenarum Hensel.
J. Biophys. Biochem. Cytl. 2: 223-240.

263
Busch, H., and K. Smetana. 1970. The Nucleolus. Academic Press,
New York. 626 p.
Byers, B., and D. H. Abramson. 1968. Cytokinesis in HeLa: post
telophase delay and microtubule-associated motility. Protoplasma
66: 413-435
Carlson, J. G. 1938. Mitotic behavior of induced chromosomal fragments
lacking spindle attachments in the neuroblasts of the grasshopper.
Proc. Nat. Acad. Sci. 2t: 5OO-507.
Carlson, J. G. 1954. Immediate effects on division, morphology, and
viability of the cell, p. 763-824. In A. Hollaender fed.) Radiation
Biology, vol. 1 (pt. 2). McGraw-Hill, New York.
Carter, S. K., D. Rail, P. Schein, R. D. Davis, H. B. Wood, Jr., R.
Engle, J. P. Davignon, J. M. Venditti, S. A. Schepartz, B. R.
Murray, and C. G. Zubrod. 1968. Streptonigrin. NSC 45383
Clinical Brochure. Nat. Cancer Inst. Chemother. 20 p.
Casarett, A. P. 1968. Radiation Biology. Prentice-Hall, Englewood
Cliffs. 368 p.
Chang, J. P., and C. W. Gibley, Jr. 1968. Ultrastructure of tumor
cells during mitosis. Cancer Res. 28: 521-534.
Cohen, M. M. 1983. The specific effects of streptonigrin activity
on human chromosomes in culture. Cytogenetics 2: 271-279
Cohen, M. M., M. W. Shaw, and A. P. Craig. 1983 The effects of
streptonigrin on cultured human leukocytes. Proc. Nat. Acad.
Sci. 50: 16-24.
Comings, D. E. 1968. The rationale for an ordered arrangement of
chromatin in the interphase nucleus. Amer. J. Human Genet.
20: 440-460.
Comings, D. E. 1971 Isolabelling and chromosome strandedness.
Nature New Biol. 229: 24-25.
Comings, D. E., and T. Kakefuda. 1988. Initiation of deoxyribonucleic
acid replication at the nuclear membrane in human cells. J. Mol.
Biol. 33: 225-229.
Comings, D. E., and T. A. Okada. 1970a. Association of nuclear
membrane fragments with metaphase and anaphase chromosomes as
observed by whole mount electron microscopy. Exp. Cell Res.
63: 62-68.
Comings, D. E., and T. A. Okada. 1970b. Association of chromatin
fibers with the annuli of the nuclear membrane. Exp. Cell Res.
62: 293-302.
Comings, D. H., and T. A. Okada. 1970c. Condensation of chromosomes onto
the nuclear membrane during prophase. Exp. Cell Res. 63: 471-473.

264
De Harven, E. 1968. The centriole and the mitotic spindle, p. 197-22?.
In A. J. Dalton and F. Haguenau (eds.) The Nucleus. Academic Press,
New York.
De Harven, E., and W. Bernhard. 1956. Etude au microscope electronique
de 1'ultrastructure du centriole chez les vrtebras. Z. Zellforsch.
Mikroskop. Anat. 45: 378-393.
Deysson, G. 1968. Antimitotic substances. Int. Rev. Cytol. 24: 991^8.
Dietz, R. 1959. Centrosomenfreie Spindelpole in Tipuliden-Spermatocyten.
Z. Naturforsch. lib: 749-752.
Dietz, R. 1966. The dispensability of the centrioles in the spermatocyte
divisions of Pales ferruginea (Nematocera), p. l6l-l66. In C. D.
Darlington and K. R. Lewis (eds.) Chromosomes Today, vol. 1.
Plenum Press, New York.
Dietz, R. 1989. Bau und Funktion des Spindelapparats. Naturwissenschaften
56: 237-248.
Dirksen, E. R. 1961. The presence of centrioles in artificially
activated sea urchin eggs. J. Biophys. Biochem. Cytol. 11: 244-247.
Dirksen, E. R., and T. T. Crocker. 1966. Centriole replication in
differentiating ciliated cells of mammalian respiratory epithelium.
An electron microscopic study. J. Microscopie 5- 629-644.
DuPraw, E. J. 1965. Macromole cular organization of nuclei and
chromosomes: A folded-fibre model based on whole-mount electron
microscopy. Nature 206: 338-3^3
DuPraw, E. J. 1968. Cell and Molecular Biology. Academic Press,
New York. 739 P
Dustin, P., Jr. 1963. New aspects of the pharmacology of antimitotic
agents. Pharmacol. Rev. 15: 449-480.
Eigsti, 0. J., and P. Dustin, Jr. 1955* Colchicine in Agriculture
Medicine, Biology, and Chemistry. Iowa State College Press,
Ames. 47 p.
Erlandson, R. A., and E. de Harven. 1971. The ultrastructure of
synchronized HeLa cells. J. Cell Sci. 8: 353397.
Evans, H. J. 1962. Chromosome aberrations induced by ionizing
radiations. Int. Rev. Cytol. 13: 221-321.
Evans, H. J. 1963. Chromosome aberrations and target theory, p. 8-40.
In S. Wolff (ed.) Radiation-Induced Chromosome Aberrations. Columbia
Univ. Press, New York.
Evans, H. J., and J. R. K. Savage. 1963* The relation between DNA
synthesis and chromosome structure as resolved by X-ray damage.
J. Cell Biol. 18: 525-540.

265
Flaks, B. 1971- Observations on the fine structure of the normal
porcine liver. J. Anat. 108: 563-577*
Forer, A. 1966. Characterization of the mitotic traction system,
and evidence that birefringent spindle fibers neither produce
nor transmit force for chromosome movement. Chromosoma 19: M-98.
Friedlander, M., and J. Wahrman. 1970. The spindle as a basal body
distributor. A study in the meiosis of the male silkworm moth,
Bombyx mori. J. Cell Sci. J: 65-89.
Gachet, J., and J. P. Thiry. I96I*. Application de la mthode de
tirage photographique avec rotations ou translations 1etude de
macromolcules (hmocyanine, hmoglobine, ferritine) et de structures
biologiques (centrioles, fibres de flagelle, nuclocapsides virales).
J. Microscopie 3: 253-268.
Gall, J. G. 195^ Lampbrush chromosomes from oocyte nuclei of the
newt. J. Morphol. 9^: 283-351.
Gall, J. G* 1961. Centriole replication. A study of spermatogenesis
in the snail Viviparus. J. Biophys. Biochem. Cytol. 10: 163-193.
Gaulden, M. E., and J. G. Carlson. 1951 Cytological effects of colchicine
on the grasshopper neuroblast in vitro with special reference to
the origin of the spindle. Exp. Cell Res. 2: I6-U33.
Gaulden, M. E., G. A. Mueller, and W. Drane. 1970. The effects of
varying concentrations of colchicine on the progression of
grasshopper neuroblasts into metaphase. J. Cell. Biol. 1*7: 69a.
Gelfant, S. 1983* Inhibition of cell division: a critical and experi
mental analysis. Int. Rev. Cytol. 14: 1-39*
George, P., L. J. Journey, and M. N. Goldstein. 1965. Effect of
vincristine on the fine structure of HeLa cells during mitosis.
J. Hat. Cancer Inst. 35: 355"375*
Gibbons, I. R., and A. V. Grimstone. i960. On flagellar structure
in certain flagellates. J. Biophys. Biochem. Cytol. J: 697-716.
Hair, J. G. 1953 The origin of new chromosomes in Agronyron.
Heredity 6 (suppl.): 215-233*
Harris, P. 1962. Some structural and functional aspects of the
mitotic apparatus in sea urchin embryos. J. Cell Biol. l4: 1*75-1*87.
Harris, P. 1965* Some observations concerning metakinesis in sea
urchin eggs. J. Cell Biol. 25 (l, pt. 2): 73-77.
Harris, P., and D. Mazia. 1962. The finer structure of the mitotic
apparatus, p. 2793C>5* In R. J. C. Harris (ed.) The Interpretation
of Ultrastructure. Academic Press, Hew York.

266
Heddle, J. A. 1969. The strandedness of chromosomes: Evidence from
chromosomal aberrations. Can. J. Genet. Cytol. 11: 783-793-
Heddle, J. A., and D. J. Bodycote. 1970. On the formation of chromosomal
aberrations. Mutation Res. 9' 117-126.
Henderson, S. A., and C. A. Koch. 1970. Co-orientation stability by
physical tension: a demonstration with experimentally interlocked
bivalents. Chromosoma 29: 207-216.
Henderson, S. A., R. B. Hicklas, and C. A. Koch. 1970. Temperature-
induced orientation instability during meiosis: an experimental
analysis. J. Cell Sci. 6: 323350.
Heneen, W. K. 1970. In situ analysis of normal and abnormal patterns
of the mitotic apparatus in cultured rat-kangaroo cells. Chromosoma
29: 88-117.
Hepler, P. K., and W. T. Jackson. 1968. Microtubules and early stages
of cell-plate formation in the endosperm of Haenanthus katherinae
Baker. J. Cell. Biol. 38: 437t446.
Hepler, P. K., J. R. McIntosh, and S. Cleland. 1970. Intermicrotubule
bridges in mitotic spindle apparatus. J. Cell Biol. 45: 438-1*44.
Hollaender, A. 1954. Radiation Biology, vol. 1 (pt. 2). McGraw-Hill,
Hew York.
Hollande, A. and J. Valentin. 1968. Infrastructure des centromeres et
droulement de la pleuromitose chez les Hypermastigines. C. R.
Acad. Sci. D, 266: 367-370.
Hsu, T. C., F. E. Arrighi, R. R. KLevecz, and B. R. Brinkley. 1965.
The nucleoli in mitotic divisions of mammalian cells in vitro.
J. Cell Biol. 26: 539-553.
Hsu, T. C., R. M. Humphrey, and C. E. Somers. 1964. Responses of
Chinese hamster and L cells to 2'-deoxy-5-fluoro-uridine and
thymidine. J. Nat. Cancer Inst. 32: 839-855.
Hu, F. 1971. Ultrastructural changes in the cell cycle of cultured
melanoma cells. Anat. Rec. 170: 41-56.
Hughes-Schrader, S. 1948. Cytology of coccids (Coccoldea Homoptera)
Adv. Genet. 2: 127-203.
Humphrey, R. M., and B. R. Brinkley, i960. Ultrastructural studies
of radiation-induced chromosome damage. J. Cell Biol. 42: 745-753.
Inoue, S. 1952. The effect of colchicine on the microscopic and
subnicroscopic structure of the mitotic spindle. Exp". Cell Res.
(suppl. 2): 305-314.
Inoue. S. i960. On the physical properties of the mitotic spindle.
Ann. N. Y. Acad. Sci. 90: 529-530.

267
Inou, S. 1964. Organization and function of the mitotic spindle,
p. 949-5914. In R. D. Allen and R. Kamiya (eds.) Primitive
Motile Systems in Cell Biology. Academic Press, New York.
Inou, S., G. W. Ellis, E. D. Salmon, and J. W. Fuseler. 1970.
Rapid measurement of spindle birefringence during controlled
temperature shifts. J. Cell Biol. 47: x95a-96a
Inou, S., and H. Sato. 1967. Cell motility by labile association of
molecules. The nature of mitotic spindle fibers and their role in
chromosome movement. J. Gen. Physiol. 50 (6, pt. 2): 259-288.
Izutsu, K. I96I. As cited by Luykx, 1970.
Jagiello, G. 1987. Streptonigrin: effect on the first meiotic
metaphase of the mouse egg. Science 157: 453-454.
Jensen, C., and A. Bajer. 1969. Effects of dehydration on the
microtubules of the mitotic spindle. Studies in vitro and
with the electron microscope. J. Ultrastruct. Res.26):
367-386.
Jokelainen, P. T. 1965a. Fine structure aspects of kinetochores in
dividing cells from fetal rat kidney. Anat. Rec. 151: 367.
Jokelainen, P. T. 1965b. The differentiation of sister kinetochores
during metakinesis. J. Cell Biol. 27: 48a.
Jokelainen, P. T. 1967* The ultrastructure and spatial organization
of the metaphase kinetochores in mitotic rat cells. J. Ultra
struct. Res. 19: 19-44.
Jokelainen, P. T. 1968. The effect of colchicine on the kinetochores
and mitotic apparatus in the rat. J. Cell Biol. 39: 68a-69a.
Journey, L. J., J. Burdman, and P. George. 1968. Ultrastructural
studies on tissue culture cells treated vith vincristine (RSC-
67574). Cancer Chemother. Rep. 52: 509-517.
Journey, L. J., and A. Whaley. 1970. Kinetochore ultrastructure in
vincristine-treated mammalian cells. J. Cell Sci. J: 49-54.
Kane, R. E., and A. Forer. 1965. The mitotic apparatus. Structural
changes after isolation. J. Cell Biol. 25 (3, pt. 2): 31-39.
Kaufmann, B. P. 1954. Chromosome aberrations induced in animal
cells by ionizing radiations, p. 627-711. In A. Hollaender (ed.)
Radiation Biology, vol. 1 (pt. 2). McGraw-Hill, New York.
Kiefer, B., H. Sakai, A. J. Solari, and D. Mazia. 1966. The molecular
unit of the microtubules of the mitotic apparatus. J. Mol. Biol.
20: 75-79.

268
Kihlman, B. A. 1964. The production of chromosomal aberrations by
streptonigrin in Vicia faba. Mutation Res. 1: ^h-62.
Kihlman, B. A. 1966. Actions of Chemicals on Dividing Cells. Prentice-
Hall, New Jersey. 260 p.
Kihlman, B. A. 1970. Sub-chromatid exchanges and the strandedness of
chromosomes. Hereditas 65: 171-186.
Roller, P. C. 1953* Dicentric chromosomes in a rat tumour induced by
an aromatic nitrogen mustard. Heredity 6 (suppl.): 181-196.
Koopmans, A. 1958. Die Persistenz der Chromosomenspindelfaser.
Naturvissenschaften 45: 66-67.
Koschei, K., G. Hartmann, W. Kersten, and H. Kersten. 1966. Die
Wirkung des Chromomycins und einiger Anthracyclinantibiotica auf
die DHA.- abhang i ge Huele ins Sure-Synthes e. Biochem. Z. 3^' 76-86.
Krishan, A. 1968. Fine structure of the kinetochores in vinblastine
sulfate-treated cells. J. Ultrastruct. Res. 23: 134-143.
Krishan, A., and R. C. Buck. 1965. Structure of the mitotic spindle
in L strain fibroblasts. J. Cell Biol. 2b: *133-1+44.
La Cour, L. F. 1953 The physiology of chromosome breakage and
reunion in Hyacinthus. Heredity 6 (suppl.): 163-179*
Lambert, A, M. 1970. Etude de structures cintiques en rapport avec
la rupture de la membrane nuclaire, en debut de miose chez
Mnium hornum L. Organisation des centromeres. C. R. Acad. Sci. D,
270: 481-484.
Lampert, F. 1969* Feinstruktur und Trockengevicht menschlicher
Chromosomen. Quantitative Elektronenmikroskopie. Naturvissenschaften
56: 629-633.
Lea, D. E. 1962. Actions of Radiations on Living Cells, 2nd ed.
Cambridge Univ. Press, New York. 4l6 p.
Levan, A. 1938. The effect of colchicine on root mitoses in Allium.
Hereditas 24: 471-486.
Levan, A. 1954. Colchicine-induced c-mitosis in tvo mouse ascites
tumours. Hereditas 40: 1-64.
Levan, A. 1970. Contributions to the chromosomal characterization
of the PTK 1 rat-kangaroo cell line. Hereditas 64: 85-96.
Levine, M., and M. Borthvick. 1963. The action of streptonigrin on
bacterial DHA metabolism and on induction of phage production in
lysogenic bacteria. Virology 21: 568-574.
Lima-de-Faria, A. 1950. The Feulgen test applied to centromeric
chromomeres. Hereditas 36: 60-74.

269
Lima-de-Faria, A. 1956. The role of the kinetochore in chromosome
organization. Hereditas 42: 85-160.
Lima-de-Faria, A. 1958. Recent advances in the study of the kinetochore.
Int. Rev. Cytol. 7: 123-157*
Luykx, P. 1965a. The structure of the kinetochore in meiosis and
mitosis in Urechis eggs. Exp. Cell Res. 39: 6^3~65T
Luykx, P. 1965b. Kinetochore-to-pole connections during prometaphase
of the meiotic divisions in Urechis eggs. Exp. Cell Res. 39:
658-668.
Luykx, P. 1970 Cellular mechanisms of chromosome distribution.
Int. Rev. Cytol., suppl. 2. Academic Press, New York. 173 P
Malavista, S. E., H. Sato, and K. G. Bensch. 1968. Vinblastine and
griseofulvine reversibly disrupt the living mitotic spindle.
Science l60: 770-772.
Mantn, I., K. Kowallik, and H. A. van Stosch. 1969a. Observations on
the fine structure and development of the spindle at mitosis and
meiosis in a marine centric diatom (Lithodesmium undulatum) I.
Preliminary survey of mitosis in spermatogonia. J. Microscopy
89: 295-320.
Mantn, I., K. Kowallik, and H. A. von Stosch. 1969b. Observations
on the fine structure and development of the spindle at mitosis
and meiosis in a marine centric diatom (Lithodesmium undulatum).
II. The early meiotic stages in male gametogenesis. J. Cell
Sci. 5: 271-298.
Mantn, I., K. Kowallik, and H. A. van Stosch. 1970. Observations on
the fine structure and development of the spindle at mitosis and
meiosis in a marine centric diatom (Lithodesmium undulatum). III.
The later stages of meiosis I in male gametogenesis. J. Cell Sci.
6: 131-157.
Markham, R., S. Frey, and G. J. Hills. 1963. Methods for the enhance
ment of image detail and accentuation of structure in electron
microscopy. Virology 20: 88-102.
Mazia, D. 1961. Mitosis and the physiology of cell division, p. 77
412. In J. Brachet and A. E. Mirsky (eds.) The Cell, vol. 3.
Academic Press, New York.
Mazia, D., P. J. Harris, and T. Bibring. i960. The multiplicity of
the mitotic centers and the time-course of their duplication and
separation. J. Biophys. Biochem. Cytol. 7: 1-20.
McIntosh, J. R., P. K. Hepler, and D. G. Van Wie. 1969. Model for
mitosis. Nature 224: 659-663.
c .. e *
McIntosh, J. R., S. C. Landis. 1971* The distribution of spindle
microtubules during mitosis in cultured human cells. J. Cell
Biol. 49: 468-497.

270
Metzner, R. 1894. As cited by Schrader, 1936.
Mizuno, N. S., and D. P. Gilboe. 1970. Binding of streptonigrin to
DNA. Biochim. Biophys. Acta 224: 319327
Mole-Bajer, J. 1958. Cine-micrographic analysis of c-mitosis in
endosperm. Chromosoma S>: 332-358.
Moor, H. 1967. Der Feinbau der Mikrotubuli in Hefe nach Gefrier&tzung.
Protoplasma 64: 89-IO3.
Murray, R. G., A. S. Murray, and A. Pizzo. 1965. The fine structure
of mitosis in rat thymic lymphocytes. J. Cell Biol. 26: 6OI-619.
Nathaniel, E. J. H., N. B. Friedman, and H. Rychuk. 1968. Electron
microscopic observations on cells of Harding-Passey melanoma
following colchicine administration. Cancer Res. 28: 1031-1040.
Nebel, B. R., and E. M. Coulon. 1962. The fine structure of chromosomes
in pigeon spermatocytes. Chromosoma 13: 272-291.
Nichols, ¥. W. 1970. Viruses and chromosomal abnormalities. Ann.
New York Acad. Sci. 171: 478-485.
Nichols, ¥. ¥., A. Levan, and B. A. Kihlman. 1964. Chromosome breakage
associated with viruses and DM inhibitors, p. 255-271* In
R. J. C. Harris (ed.) Cytogenetics of Cells in Culture.
Academic Press, New York.
Nicklas, R. B. 1967. Chromosome micromanipulation. II. Induced
reorientation and the experimental control of segregation in
me iosis. Chromosoma 21: 1750.
Nicklas, R. B. 1971. Mitosis. Adv. Cell Biol. vol. 2. In press.
Nicklas, R. B., and C. A. Koch. 1969. Chromosome micromanipulation.
III. Spindle fiber tension and the reorientation of mal-oriented
chromosomes. J. Cell Biol. 43: 40-50.
Nicklas, R. B., and C. A. Staehly. 1967. Chromosome micromanipulation.
I. The mechanics of chromosome attachment to the spindle.
Chromosoma 21: l-l6.
Oleson, J. J., L. A. Calderella, K. J. Mjos, A. R. Reith, R. S. Thie,
and I. Toplin. 1961. The effects of streptonigrin on experimental
tumors. Antibiot. Chemother. 11: 158-164.
Ostergren, G. 1951* The mechanism of co-orientation in bivalents
and multivalents. Tne theory of orientation by pulling.
Hereditas 37: 85-156.
Paweletz, N. 1967. Zur Funktion des "Flemming-Korpers" bei der
Teilung tierischer Zellen. Naturwissenschaften 54: 533-535

271
Peacock, V/. J. 1963. Chromosome duplication and structure as de
termined by autoradiography. Proc. Hat. Acad. Sci 49: 793"801.
Pickett-Heaps, J. D. 1969. The evolution of the mitotic apparatus:
an attempt at comparative ultrastruetural cytology in dividing
plant cells. Cytobios 1: 257-280.
Pickett-Heaps, J. D. 1971* The autonomy of the centriole: fact or
fallacy? Cytobios 3: 205-21It.
Pickett-Heaps, J. D., and L. C. Fowke. 1969 Cell division in
Oedogonium. I. Mitosis, cytokinesis, and cell elongation.
Aust. J. Biol. Sci. 22: 857-894.
Porter, K. R., and R. D. Machado. i960. Studies on the endoplasmic
reticulum. IV. Its form and distribution during mitosis in
cells of onion toot tip. J. Biophys. Biochem. Cytol. 7: 167-180.
Puck, T. T. 1964. Phasing, mitotic delay and chromosomal aberrations
in mammalian cells. Science ihlt: 565-566.
Radding, C. M. 1963. Incorporation of H^ thymidine by K 12 (A )
induced by streptonigrin, p. 22. In S. J. Geerts (ed.)
Genetics Today (Proc. XI Int. Congr. Genet.).
Rao, K. V., K. Biemann, and R. B. Woodward. 1963 The structure of
streptonigrin. J. Arner. Chem. Soc. 85: 2532-2533
Rebhun, L. I., and G. Sander. 1967 Ultrastructure and birefringence
of the isolated mitotic apparatus of marine eggs. J. Cell Biol.
34: 859-883.
Renaud, F. L., and H. Swift. 1964. The development of basal bodies and
flagella in Allomyces arbuscuius. J. Cell Biol. 23: 339354.
Revell, S. H. 1955 A new hypothesis for "chromatid" changes, p. 243-253
In Z. M. Bacq and P. Alexander (eds.) Radiobiology Symposium
1954. Academic Press, Hew York.
Revell, S. H. 1963 Chromatid aberrations the generalized theory,
p. 41-72. In S. Wolff (ed.) Radiation-Induced Chromosome Aberrations.
Columbia Univ. Press, Hew York.
Reynolds, E. S. 1963 The use of lead citrate at high pH as an
electron-opaque stain in electron microscopy. J. Cell Biol.
17: 208-212.
Ringo, D. L. 1967a. Flagellar motion and fine structure of the
flagellar apparatus in Chlanydomonas. J. Cell Biol. 33: 543-571.
Ringo, D. L. 1967b. The arrangement of subunits in flagellar fibers.
J. Ultrastruct. Res. 17: 266-277.
Ris, H. 1949 The anaphase movement of chromosomes in the spermatocytes
of the grasshopper. Biol. Bull. 96: 90-106.

272
Ris, H. 1955. Cell division, p. 91-125* In B. H. Willier, P. A.
Weiss, and V. Hamburger (eds.) Analysis of Development.
Saunders, Philadelphia.
Ris, H. 1961. The annual invitation lecture. Ultrastructure and
molecular organization of genetic systems. Can. J. Genet. Cytol.
3: 95-120.
Ris, H. 19^7* Ultrastructure of the animal chromosome, p. 11-21.
In V. V. Koningsberger and L. Bosch (eds.) Regulation of Nucleic
Acid and Protein Biosynthesis. Elsevier, New York.
Ris, H., and D. F. Kubai. 1970* Chromosome structure. Annu. Rev.
Genet, 4: 263-294.
Robbins, E., and N. K. Gonatas. 1964. The ultrastructure of a mammalian
cell during the mitotic cycle. J. Cell Biol. 21: 429-463.
Robbins, E., G. Jentzsch, and A. Micali. 1968. The centriole cycle
in synchronized HeLa cells. J. Cell Biol. 36: 329339*
Roth, L. E. 1964. Motile systems with continuous filaments, p. 527-
548. In R. D. Allen and N. Kamiya (eds.) Primitive Motile Systems
in Cell Biology. Academic Press, New York.
Roth, L. E. 1967* Electron microscopy of mitosis in amebae. III.
Cold and urea treatments: A basis for tests of direct effects
of mitotic inhibitors on microtubule formation. J. Cell Biol.
34: 47-60.
Roth, L. E., and E. W. Daniels. 1962. Electron microscopic studies of
mitosis in amebae. II. The giant ameba Pelomyxa carolinensis. J.
Cell Biol. 12: 57-78.
Roth, L. E., H. J. Wilson, and J. Chakraborty. 1966. Anaphase structure
in mitotic cells typified by spindle elongation. J. Ultrastruct.
Res. l4: 460-483.
Sauaia, H., and D. Mazia. 1961. As cited by Luykx, 1970.
Schr, B., P. Loustalot, and F. Gross. 1954. Demecolcin (Substanz F),
ein neues, aus Colchicum autumnale isoliertes Alkaloid mit starker
antimitotischer Wirkung. Klin. Wochenschr. 32: 49-57.
Scheid, W., and H. Traut. 1970. Ultraviolet-microscopical studies on
achromatic lesions ("gaps") induced by X-rays in the chromosomes
of Vicia faba. Mutation Res. 10: 159-161.
Scheid, W., and H. Traut. 1971* Visualization by scanning electron
microscopy of achromatic lesions ("gaps") induced by X-rays in
chromosomes of Vicia faba. Mutation Res. 11: 253-255.
Schrader, F. 1936. The kinetochore or spindle fibre locus in Amphiuma
tridactylum. Biol. Bull. JO: 484-498.

273
Schrader, F. 1939* The structure of the kinetochore at meiosis.
Chromosoma 1: 230-237*
Schrader, F. 1953* Mitosis, 2nd ed. Columbia Univ. Press, Hew York.
170 p.
Schroeder, T. E. 1970. The contractile ring. I. Fine structure of
dividing mammalian (HeLa) cells and the effects of cytochalasin
B. Z. Zellforsch. Mikroskop. Anat. 109: 431-449.
Seaman, G. R. i960. Large-scale isolation of kinetosomes from the
ciliated protozoan Tetrahymena pyriformis. Exp. Cell Res. 21:
292-302.
Sentein, P. 1961. As cited by Biesele, 1962.
Sharp, L. W. 1934. As cited by Schrader, 1936.
Shaw, M. W. 1970. Human chromosome damage by chemical agents. Annu.
Rev. Med. 21: 1*09-432.
Shaw, M. W., and R. S. Krooth. 1964. The chromosomes of the Tasmanian
rat-kangaroo (Potorous tridactylis apicalis). Cytogenetics 3 19"33
Shimamura, T. 1940. Studies on the effect of the centrifugal force
upon nuclear division. Cytologia 11: 186-216.
Sisken, J. E., and E. Wilkes. 1967. The time of synthesis and the
conservation of mitosis-related proteins in cultured human amnion
cells. J. Cell Biol. 34: 97-110.
Stephens, R. E. 1969. As cited by Hicklas, 1971.
Stubblefield, E. 1968. Centriole replication in a mammalian cell,
p. 175-93. In The Proliferation and Spread of Neoplastic
Cells. Williams and Wilkins, Baltimore.
Stubblefield, E., and B. R. Brinkley. 1967. Architecture and function
of the mammalian centriole, p. 175-218. In K. B. Warren (ed.)
Formation and Fate of Cell Organelles. Academic Press, Hew York.
Stubblefield, E., and W. Wray. 1971 Architecture of the Chinese hamster
metaphase chromosome. Chromosoma 32: 262-294.
Taylor, J. H. 1963. EHA synthesis in relation to chromosome re
production and the reunion of breaks. J. Cell Comp. Physiol.
62 (suppl): 73-86.
Taylor, J, H., P. S. Woods, and W. L. Hughes. 1957. The organization
and duplication of chromosomes as revealed by autoradiographic
studies using tritium-labeled thymidine. Proc. Hat. Acad. Sci.
43: 122-128.

274
Uretz, R. B., W. Bloom, and R. E. Zirkle. 1954. Irradiation of parts
of individual cells. II. Effects of an ultraviolet microbeam
focused on parts of chromosomes. Science 120: 197-199*
Vig, B. K. 1970. Sub-chromatid aberrations in Haworthla attenuate.
Can. J. Genet. Cytol. 12: l8l-l86.
Walen, K. H. 1965. Spatial relationships in the replication of
chromosomal DM. Genetics 51s 9Y5~929*
Walen, K. H., and S. V/. Brown. 1962. Chromosomes in a marsupial
(Potorous tridactylis) tissue culture. Nature 194: 4o6.
Watson, M. L. 1958. Staining of tissue sections for electron
microscopy with heavy metals. J. Biophys. Biochem. Cytol.
4: 475-478.
Went, H. A. i960. Dynamic aspects of mitotic apparatus protein.
Ann. N. Y. Acad. Sci. 90: 422-429.
Went, H. A. 1966. An indirect method to assay for mitotic centers
in sand dollar (Dendraster excentricus) eggs. J. Cell Biol.
30: 555-562.
Wettstein, R., and J. R. Sotelo. 1965 Fine structure of meiotic
chromosomes. The elementary components of metaphase chromosomes
of Gryllus argentinus. J. Ultrastr. Res. 13: 367-381.
Wilson, H. J. 1968. The fine structure of the kinetochore in meiotic
cells of Tradescantia. Planta 78: 379-385*
Wilson, H. J. 1969* Arms and bridges on microtubules in the mitotic
apparatus. J. Cell Biol. 40: 854-859*
Wilson, H. J. 1970. Endoplasmic reticulum and microtubule formation
in dividing cells of higher plants a postulate. Planta 94:
184-190.
Wilt, F. H., H. Sakai, and D. Mazia. 1967* Old and new protein in the
formation of the mitotic apparatus in cleaving sea urchin eggs.
J. Mol. Biol. 27: 1-7*
Wolfe, S. L. 1965* The fine structure of isolated metaphase chromosomes.
Exp. Cell Res. 37: 45-53*
Wolfe, S. L. 1969* Molecular organization of chromosomes, p. 3_42.
In E. E. Bittar (ed.) The Biological Basis of Medicine, vol. 4.
Academic Press, New York.
Wolfe, S. L., and P. G. Martin. 1968. The ultrastructure and
strandedness of chromosomes from two species of Vicia. Exp.
Cell Res. 50: 140-150.
Wolff, S. (ed.). 1963. Radiation-Induced Chromosome Aberrations.
Columbia Univ. Press, New York. 304 p.

275
Yamanaka, T., and Y. Okada. 1968. Cultivation of fused cells re
sulting from treatment of cells with HVJ. II. Division of
binucleated cells resulting from fusion of two KB cells by
HVJ. Exp. Cell Res. k9: kGl-k6$.
Young, C. W., and S. Hodas. 1965. Acute effects of cytotoxic com
pounds on incorporation of precursors into DM, RIIA. and protein
of HeLa monolayers. Biochem. Pharmacol. l4: 205-21^.
Zirkin, B. R., and S. L. Wolfe. 1970. The chemical composition of
nuclei and chromosomes isolated by the Langmuir trough technique.
Chromosoma 32:162-170.
Zubay, G., and P. Doty. 1959* The isolation and properties of
deoxyribonucleoprotein particles containing single nucleic acid
molecules. J. Mol. Biol. 1: 1-20.

BIOGRAPHICAL SKETCH
Urs-Peter Roos was born August 14, 1938, at Murten, Switzerland.
After completing secondary school in Dietikon, Switzerland, in 1953, he
began apprenticeship as an engraver and successfully passed the Swiss
Federal Examination in 1958. In 1960 he enrolled in the school of the
Akademikergemeinschaft in Zurich. He graduated in March, 1963, and
obtained the "Schweizerisches Maturitatszeugnis Typus C." He enrolled
in the Swiss Federal Institute of Technology (ETH) in Zurich in October,
1963, and graduated from this institution in August, 1967, with the
degree of Ingenieur Agronom. In September, 1967, he enrolled in the
Graduate School of the University of Florida. He worked as a graduate
assistant in the Department of Entomology and Hematology until December,
1968, when he received the degree of Master of Science with a major in
Entomology. In September, 1969, he transferred to the Department of
Zoology. He worked as a teaching and research assistant and was
recipient of a Graduate Fellowship of the College of Arts and Sciences
during the time he pursued his work toward the degree of Doctor of
Philosophy.
Urs-Peter Roos is married to the former Loan Hiang Ghan and is the
father of a son. He is a member of Phi Kappa Phi, Gamma Sigma Delta,
Phi Sigma, an associate member of Sigma Xi, and a student member of the
American Association for the Advancement of Science and the American
Society for Cell Biology.
276

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
F. C. Jo:
Associa
Professor of Zoology
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
JU. C. dtAusL.
H. C. Aldrich, Co-Chairman
Assistant Professor of Botany
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
'yj
fL-i-
L. H. Larkin
Assistant Professor of Anatomical
Sciences
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

This dissertation vas submitted to the Department of Zoology in
the College of Arts and Sciences and to the Graduate Council, and was
accepted as partial fulfillment of the requirements for the degree of
Doctor of Philosophy.
Dean, Graduate School



158
Fig. 57. Early telophase. Reconstitution of the nucleus. Three
obliquely sectioned kinetochores within the chromatin mass (arrowheads).
Note large pieces of chromatin-associated membranes (large arrows),
chromosomal granules (small arrows), and spherical particle (Pa),
x 30*000 Inset: Phase contrast micrograph of the cell in olastic
(x 1,000).


59
chromosomes during mid-prometaphase (Figure 26), become smaller and
move to the periphery of the spindle in late prometaphase (Figures 29
and 30a). Nuclear pore complexes have disappeared approximately at the
stage shown in Figure 26, but some polyribosomes are still present on
the fragments of the nuclear envelope. Peripheral segments of double
membrane in late prometaphasc cells have ribosomes on both faces, thus
resembling rough EE (Figure 29).
As the spindle develops, mitochondria, EE, and other large cellu
lar organelles and components are excluded from its area (compare Fig
ure 20 with Figure 29), but ribosomes, mostly monosomes, are abundant.
Bundles of ITT in the center of the spindle run more or less straight
along the pole-to-pole axis, while peripheral bundles are slightly
arched (Figures 26 and 29). An exceptional case is shown in Figure 32,
where rather disoriented MT curve sharply towards the centriole.
Angular configurations of kinetochore Iff, as shown in Figure 20, no
longer occur. Still, most of the bundles of MT are associated with
kinetochores (Figures 26-29, and 30a). It is very difficult to deter
mine whether the MT of other bundles (e.g., Figure 28) are interpolar,
or whether they are associated with kinetochores not visible in the
same section. Quite frequently, a bundle of MT passes near the kineto
chores of a maloriented chromosome (e.g., Figure 39). Transchromosomal
MT are outstanding in rather thick sections (Figure 30b). Skew MT, as
well as "wavy" MT, occur quite frequently, but they always make up a
small proportion of the total number of spindle MT (Figures 27, 28, 30a,
and 33).
Orientation of the centrioles relative to the axis of the fully
formed spindle is variable (figure 29). This has also been confirmed
in cells in meta- and anaphase.


Fig. 109. Streptonigrin-treated cell in early anaphase. Chromatin
strand connecting the kinetochore regions of daughter chromosomes
(arrows). Note profile of kinetochore (k) in a pocket of the adjacent
chromosome. From the cell in Fig. 108b. x 22,500.
Fig. 110a, b. Streptonigrin-treated cell in late anaphase, (a)
Portion of a dicentric chromosome with kinetochore (Kg), dense chromatin
strand (arrows). One other kinetochore is visible (k). x 22,500* (b)
Chromatin strand of a similar bridge at higher magnification. Note
difference in coiling between strand (marked by the two lines) and
peripheral fibers, x 157,500. Both from the cell in Fig. 108f.


Thus, the task is, not so much to see what nobody
has seen yet; but to think what nobody has thought
yet, about what everybody sees.
Schopenhauer


171
mmm$0
mm
. <- it'
;.
-


244
(Figures 23 and 33). This material is gradually reduced in extent, but
some of it remains present in mature kinetochores as the corona and the
middle layer (Figures 42, 46, 78, and 84). The inner layer, at the
surface of the chromosome, appears last (Figures 33-35).
Kinetochore maturation as described herein essentially agrees
with the observations of Jokelainen (1965). However, his conclusions
that sister kinetochores always mature asynchronously does not apply to
PtK^ cells. A better knowledge of the position of chromosomes in the
spindle and the orientation of sister kinetochores allowed detection of
subtle differences between kinetochores of maloriented chromosomes, as
well as obstructed and unobstructed kinetochores of equatorial chromo
somes .
There is virtually no change in kinetochore structure from meta
phase to mid-anaphase (Figures 42, 43, 45-49). "Dedifferentiation" of
kinetochores begins in late anaphase (Figures 52 and 53). The triple
layered profiles appear less distinct. Concomitantly the number of
kinetochore MT is reduced. During telophase, the reforming NE
separates the remnants of inner and outer layers (Figures 59 and 60).
The former is lost in the cytoplasm, while the latter comes to lie
closely apposed to the inner membrane of the NE (see also Figures 115
and 116). The most convincing evidence for this conclusion came from
an aberrant cell similar to the one in Figure 116, but from a TEPA-
treated culture. The inner layers of several kinetochores, also
situated in deep invaginations of the NE on the polar side, appeared as
very electron dense bumps on the inner membrane of the NE. On the out
side of the outer membrane, opposite these bumps, remnants of the less
opaque outer layers with MT still attached were visible.




93


51
Fig. 2. The three planes in which blocks were sectioned. C: Cell
in mitosis; the spindle axis is assumed to be along a line between the
two pointed ends. HO: Horizontal plane. PE: Para-equatorial plane.
PS: Para-sagittal plane.


Fig. 35&-C. Late prometaphase. Unobstructed (K^) and obstructed
(K2) kinetochore of an equatorial chromosome. From the same cell as
Fig. 4l. Note difference in banding, shape, of the two kinetochores.
x 50,000.


194


81
Centriole aberrations.--The three major possible aberrations
involving centrioles abnormal number, abnormal structure, and
abnormal position are illustrated in Figures 104-107. The presence
of four centrioles in interphase (Figure 104) presumably leads to
multipolar mitosis. The prophase cell in Figure 105a was very similar
to the one in Figure 11a, as far as chromosome condensation and
kinetochore differentiation are concerned. However, the two pairs of
centrioles were positioned on opposite sides of the nucleus (Figure
105a). Despite this obviously axial arrangement, very few MT were
associated with each centriole pair (Figures 105b and 105c).
Figure 106 shows two serial sections of the pole no. 2 centrioles
of the cell in Figure 29. Portions of three centrioles are visible.
The one in the center was deformed to a cup-like structure. Because of
missing sections, the precise architecture of, and relationship between,
these centrioles could not be reconstructed. Microtubules converged on
the osmiophilic masses left and right of center in Figure 106a.
Centriole aberrations also occurred in the anaphase cell of Figure
102. One of a pair of centrioles near pole no. 2 is shown in Figure
107a. The MT, however, converged on a different center (P^), where
amorphous, osmiophilic material was present. I am convinced that there
was at least a third centriole at this spot, but I could not verify it,
because some of the serial sections were missing. On the other hand,
one serial section of the centriole shown in Figure 107a revealed a
very osmiophilic particle in the lumen of the centriole (Figure 107b).
In size and shape this particle was very similar to those found in the
vicinity of centrioles at all stages of the cell cycle (e.g., Figure 4).
At pole no. 1 of this anaphase cell there were probably also more than
two centrioles (Figure 107c).


Fig. 8b-e. Very early prophase. Four serial sections of
the two pairs of centrioles shown in Fig. 8a. Daughter centrioles
(DCj_, DC2) closely associated with parent centrioles (PC-^, PCg).
x 40,000.


256
field between the previously unattached kinetochore and its pole. As a
consequence of unipolar attachment the chromosome should travel towards
the pole to which it is attached. This is probably not so, otherwise
many more chromosomes could be found near the poles. Possible ex
planations are connections between non-sister kinetochores (perhaps
between the kinetochores of chromosomes no. 1 and 2 in Figure 33; see
also Bajer 1970, Luykx 1965b), or simply physical hindrance. In fact,
neighbor chromosomes can lie so close to each other that a separation
is not detectable in thin sections (Figure 27).
Chromosomes without definite attachment to either pole (Figure 39)
may simply drift in the spindle or remain stationary until one or both
of its kinetochores become attached. Their subsequent behavior would
depend on the nature of the attachments.
If my interpretation is correct, detachment, reorientation, and
reattachment should occur less frequently for mitotic than for
meiotic chromosomes. This awaits experimental verification. Not only
does the above account agree with Ostergren's theory (1951), but
Nicklas (1967) clearly demonstrated that in grasshopper spermatocytes
kinetochore position after detachment by micromanipulation determines
to which pole the kinetochore will orient and attach. Finally,
stabilization of chromosomes by tension was beautifully demonstrated
by Henderson and Koch (1970), Henderson et al. (1970), and Nicklas and
Koch (1969).
The hypothetical points of my interpretation can be tested
experimentally. Time-lapse cinemicrographic records of the prometa
phase behavior of selected chromosomes followed by an ultrastructural
analysis of kinetochore structure and attachments to MT could easily
reveal which of my assumptions are true or false.


152
Fig. 51. Late anaphase, (a) Survey of elongated spindle showing
two sets of chromosomes near opposite poles (p^, Pp) Note pieces of
double membrane at periphery of spindle area (large arrows), few inter
zonal MT (small arrows), x 7,750 (b) Detail of chromosome group at
pole no. 1. Note rather indistinct kinetochores (arrowheads), MT
crossing in front of chromosomes, x 30,000.


Fig. 78a-e (continued on the following page). Five of seven
serial sections of a kinetochore in very early anaphase, (a) Section
no. 2. (b) Section no. 4; note corona (Co), outer band (KO), middle
band (KM), and inner band (ICE). (c) Section no. 5.


234
Electron micrographs such as Figures 18 and 19 demonstrate yet
another aspect of spindle formation: MT within the "nucleus" are far
more numerous near centrioles than on the opposite side. This could be
interpreted as a "growing process" away from the centrioles, possibly
concomitant with the diffusion of MT subunits into the "nucleus." As
the spindle develops further and the centrioles arrive at opposite
poles, the distribution of MT becomes more uniform (Figures 20, 26, and
29). The apparent paucity of continuous MT compared to kinetochore MT
in fully developed spindles (Figures 29, 42, and 47) is probably
imaginary. Kinetochore MT, arranged in bundles, are more conspicuous
than the dispersed continuous MT.
The absence of MT from the nucleus during prophase has a parallel
in telophase. Microtubules do not occur within the nucleus after the
NE is completely reconstructed (Figure 60). The presence of MT in
abnormal daughter nuclei of SN-treated cells (Figures 114-116) does
not contradict this observation, because in these cases the NE is
incomplete. It is not clear, however, in what way elimination of MT
enclosed in the chromosomal mass in late anaphase (Figure 52) is
accomplished. Robbins and Gonatas (1964) observed incompatibility of
MT and complete NE in HeLa cells, but showed possible exceptions, i.e.,
MT apparently penetrating the NE. The authors conceded, however, that
this could have been a false impression created by superposition.
I have not studied MT in the midbody in detail. It is possible
that most of the "continuous" MT terminate in this body, or near it,
but Figure 55a strongly suggests that at least some MT pass from one
half-spindle into the other. It is interesting to note the abundance
of MT converging in the midbody of cells in late cytokinesis, after the


This dissertation vas submitted to the Department of Zoology in
the College of Arts and Sciences and to the Graduate Council, and was
accepted as partial fulfillment of the requirements for the degree of
Doctor of Philosophy.
Dean, Graduate School


Fig. 75. Cold-treated cell in prometaphase. Note fuzzy, "but
triple-banded profiles of kinetochores (K), associated MT, centriole
(C). x 30,000.
Fig. 76. Cold-treated cell in cytokinesis. The midbody (MB)
connected to the daughter cell on the right by the stem (Sm). Serial
sections at a different level revealed connection to the other
daughter cell. Note MT, clumped ribosomes, x 22,500.


Fig. 55h. Very late anaphase. Mitochondria (Mi), RER, vesicles
(V), and MT in the constricted equatorial region. Serial section of
the same cell as Fig. 55&. x 22,500.
Fig. 56. Centrioles in early anaphase. An unidentified particle
(Pa) and, possibly, a satellite (s) are present. Note MT embedded in
amorphous material surrounding the centriole on the right, x 75000.


265
Flaks, B. 1971- Observations on the fine structure of the normal
porcine liver. J. Anat. 108: 563-577*
Forer, A. 1966. Characterization of the mitotic traction system,
and evidence that birefringent spindle fibers neither produce
nor transmit force for chromosome movement. Chromosoma 19: M-98.
Friedlander, M., and J. Wahrman. 1970. The spindle as a basal body
distributor. A study in the meiosis of the male silkworm moth,
Bombyx mori. J. Cell Sci. J: 65-89.
Gachet, J., and J. P. Thiry. I96I*. Application de la mthode de
tirage photographique avec rotations ou translations 1etude de
macromolcules (hmocyanine, hmoglobine, ferritine) et de structures
biologiques (centrioles, fibres de flagelle, nuclocapsides virales).
J. Microscopie 3: 253-268.
Gall, J. G. 195^ Lampbrush chromosomes from oocyte nuclei of the
newt. J. Morphol. 9^: 283-351.
Gall, J. G* 1961. Centriole replication. A study of spermatogenesis
in the snail Viviparus. J. Biophys. Biochem. Cytol. 10: 163-193.
Gaulden, M. E., and J. G. Carlson. 1951 Cytological effects of colchicine
on the grasshopper neuroblast in vitro with special reference to
the origin of the spindle. Exp. Cell Res. 2: I6-U33.
Gaulden, M. E., G. A. Mueller, and W. Drane. 1970. The effects of
varying concentrations of colchicine on the progression of
grasshopper neuroblasts into metaphase. J. Cell. Biol. 1*7: 69a.
Gelfant, S. 1983* Inhibition of cell division: a critical and experi
mental analysis. Int. Rev. Cytol. 14: 1-39*
George, P., L. J. Journey, and M. N. Goldstein. 1965. Effect of
vincristine on the fine structure of HeLa cells during mitosis.
J. Hat. Cancer Inst. 35: 355"375*
Gibbons, I. R., and A. V. Grimstone. i960. On flagellar structure
in certain flagellates. J. Biophys. Biochem. Cytol. J: 697-716.
Hair, J. G. 1953 The origin of new chromosomes in Agronyron.
Heredity 6 (suppl.): 215-233*
Harris, P. 1962. Some structural and functional aspects of the
mitotic apparatus in sea urchin embryos. J. Cell Biol. l4: 1*75-1*87.
Harris, P. 1965* Some observations concerning metakinesis in sea
urchin eggs. J. Cell Biol. 25 (l, pt. 2): 73-77.
Harris, P., and D. Mazia. 1962. The finer structure of the mitotic
apparatus, p. 2793C>5* In R. J. C. Harris (ed.) The Interpretation
of Ultrastructure. Academic Press, Hew York.


64
A striking phenomenon is illustrated in Figures 46-48. Daughter
chromosomes in very early and early anaphase are connected by electron
dense strands. First believed to be an aberration, it was found in six
of seven early anaphase cells. The chromosomes of the seventh cell had
short strands of similar appearance extending from the rather widely
separated kinetochore regions into the interzone. The same was
observed in one mid-anaphase cell. In the latter two cases the strands
were extremely tapered towards the interzone, giving the impression of
connections gradually drawn out and finally ruptured. In the majority
of early anaphase cells the strands connect the kinetochore regions of
daughter chromosomes. I have counted as many as six strands in serial
sections. However, in the cell shown in Figure 47, and also another
very similar anaphase, a few strands connected the mid-region of
daughter chromosomes.
The diameter of these strands varies from approximately 400 A to
900 A. Short strands are usually thicker, and long strands very
frequently are thinnest approximately midway between daughter chromo
somes, while they are thicker at their base (i.e., point of attachment;
Figures 46 and 47). The staining properties of the strands are
identical with those of chromosomes, and the structure is that of fine
and/or densely packed fibers.
Kinetochores vary in appearance from more or less straight to
convex, "stalked," or angular (Figures 46-49). The three layers are
very distinct except in sections that cut a kinetochore obliquely or
peripherally (Figures 46 and 47). In para-equatorial sections the
kinetochores appear as circles of moderate electron density, within
which cross sections of MT-can be seen (Figure 50). In late anaphase


229
osmiophilic material, could also be the pole-determinant, but the two
functions are obviously related. A rather diffuse, morphologically
less well-defined polar determinant and ItTOC could apply to acentriolar
mitosis in higher plants and other organisms as well.
The concept of MTOC as stated above is amenable to scrutiny in
animal cells. Ultrastructural investigations of the following
possibilities would yield significant information: (1) The number of
mitotic poles is correlated with the number of single centrioles or
pairs of centrioles regardless of the presence or absence of osmiophilic
material. (2) The number of poles is correlated with the number of
masses of osmiophilic material, regardless of the number of centrioles
present. Light microscopic observations have produced some evidence
against point (1). Normal bipolar divisions are more frequent than
multipolar divisions in somatic hybrid cells (e.g., Yamanaka and Okada
1968), and in spermatocytes of certain Diptera mitosis can take place
in cells lacking centrioles (Dietz 1959, 1966). Some of the centriole
aberrations I observed provide additional information. For instance,
the centriole shown in Figure 107a is not at the spindle pole. In a
preliminary study on the effect of the alkylating agent TEPA
(triethylenephosphoramide) I have observed a solitary centriole in the
interzone of a tripolar anaphase. Osmiophilic material was associated
with this centriole and some MT converged at this site, but it was not
a functional mitotic pole. On the other hand, both osmiophilic masses
in Figure 106 are foci of tIT. It appears that the presence of a
centriole in a mitotic cell is not identical with the formation of a
pole. Whether the same applies to osmiophilic masses remains to be
seen. Centriole-free divisions as in dipteran spermatocytes would be
the system of choice to investigate this.


Fig. 120. Diagrammatic representation of kinetochore maturation
during prometaphase, (a) The chromosome lies near one pole. The
kinetochore oriented towards this pole (Kq_) is drawn out to a cone
and attached to MT. The outer layer has possibly formed. The kineto
chore oriented towards the far pole (Kg) is less mature. As yet
none of the layers has formed and the kinetochore is not attached
to MT. Compare with Fig. 2k. (b) The outer layer of Kg is forming,
but the kinetochore is still not attached to MT. Kj_ is unchanged.
The chromosome is still near the same pole. Compare with Figs. 23
and 37* (c) Kg has reached the stage of maturity corresponding to
in (a); it is attached to MT. The inner layer of is forming.
Both kinetochores are stretched and the chromosome is moving towards
the equator (to the left). Compare with Fig. 40. (d) Both kinetochores
are nearly mature, showing three layers. Both are also stretched. The
chromosome has attained stable bipolar orientation at the equator.
Compare with Fig. 3^


3
Prometaphase
This stage, which is characterized by raetakinesis, i.e.,
chromosome movements that ultimately result in the alignment on the
metaphase plate (Mazia 1961), has been neglected by investigators of
the ultrastructure of mitosis in animal cells. This may be due partly
to difficulties in recognizing and selecting cells in this stage.
Metaphase
The chromosomes are condensed and aligned on the metaphase plate.
The mitotic spindle, made up of chromosomal and interpolar (continuous)
microtubules, is fully formed between the two poles occupied by the two
pairs of centrioles. Asters, consisting of MT radiating from the
centrioles, may be more or less distinct. Larger cell organelles, such
as mitochondria and endoplasmic reticulum (ER), are generally excluded
from the spindle, but exceptions occur. Ribosomes are abundant within
the spindle, mostly as monomers. Spindle MT converge at the poles,
where they enter the osmiophiiic zone around the centrioles. Bundles
of MT connect to the chromosomes at the kinetochores. The latter are
seen as straight or crescent-shaped, dark bands, separated from the
chromosome by a lighter band. Remnants of the nuclear envelope may be
present at the periphery of the spindle, often more concentrated near
the poles. Nuclear pore complexes are absent from these fragments.
Anaphase
As the two sets of daughter chromosomes separate, the chromosomal
(kinetochore) MT shorten, while the pole-to-pole distance increases
(spindle elongation). Later, the chromosomes fuse to one large mass of
dense chromatin. Concomitantly, pieces of double membrane, possibly


NORMAL AND ABNORMAL MITOSIS IN A MAMMALIAN CELL IN VITRO.
A LIGHT AND ELECTRON MICROSCOPIC STUDY.
By
Urs-Peter Roos
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1971


226
(Figures 56 and 77). At the proximal end the centriole is differen
tiated into the typical cartwheel (Figure 9). Satellites and an intra-
centriolar vesicle occur (Figures 4, 9, and 56), but I have been unable
to follow possible changes during the cell cycle, or to assess the
effect of cold, colcemid, or SN on these structures. The osmiophilic,
globular particles found near the centrioles at all stages of mitosis
are a mystery. They have also been observed by Brinkley (personal
communication). The possibility that they are virus particles lies at
hand, but is not confirmed. One is tempted to speculate that their
consistent association with the centrioles assures distribution to
daughter cells.
Centriole duplication follows the well-known pattern (Brinkley and
Stubblefield 1970, Erlandson and de Harven 1971, Murray et al. 1965).
Daughter centrioles arise at approximately a right angle to parent
centrioles (Figure 8). I have not obtained a sufficient number of
favorable sections of this stage to be able to confirm that daughter
centrioles are formed at the proximal end (see Brinkley and Stubblefield
1970). However, centrioles at the poles of the mitotic spindle are
oriented so that the proximal ends with the cartwheel point away from
each other (Figures 20, 56, and 77). Unless rotation and dislocation
occur following duplication, this configuration must reflect the
original parent-daughter association. It is interesting to note here
that the orthogonal arrangement is not the case in interphase cells
with supernumerary centrioles (Figure 104). Sections of colcemid-
treated cells (Figure 63) confirmed the observation by Brinkley et al.
(1967) that this drug does not inhibit duplication of centrioles. The
same'is true for the spindle poison vincristine (Journey et al.' 1968).


230
In the view of Brinkley and Stubblefield (1970) centriole migra
tion is a consequence of the generation of MT, i.e., the centrioles
"propel themselves to the opposite mitotic poles by pushing against
each other through the spindle which they generate." The same concept
has been adopted by Friedlander and Wahrman (1970) and McIntosh et al.
(1969). This idea is plausible, for no matter what kind of pole-
determinants we propose, we are faced with the problem how they reach
opposite poles. Centriole migration does not occur in colcemid-treated
cells (Figure 63; see also Brinkley et al. 1967). I have seen a few liT
around centrioles in interphase cells treated with 0.25 pg/ml colcemid,
but not in mitotic cells. It would be very interesting to know if
centriole migration occurs in cold-treated cells where almost all of
the MT present are of the kinetochore type (Figure 75; see also
Brinkley and Cartwright 1970). This could only be investigated by
observing living cells, for although all stages of mitosis are found in
unsynchronized populations of cold-treated cells, this may just reflect
the state prior to, not the processes occurring during treatment. What
casts some doubt on this concept of pushing is the paucity and dis
orientation of MT in very early prophase (Figures 8 and 9). This is
even more striking in Figure 105, where the centrioles have apparently
migrated to opposite poles prematurely, before a true spindle was
formed. Alternative explanations of centriole migration are difficult
to conceive. Additional information, preferably from simpler systems,
is certainly necessary.
Microtubules
Microtubules in Ptl<2 cells have the usual dimension, i.e., a
diameter of 200-250 A. Preparations with long, straight MT can be


Streptonigrin (SN; 0.01 pg/ml and 0.05 >ig/ml for 4 hr, 48 hr
recovery) depressed the mitotic index (2.22 and 0.18, respectively,
versus 2.59 in controls) and induced high frequencies of chromosomal
and mitotic aberrations in meta- and anaphase cells (49.07. and 1007.,
respectively, versus 147. in controls). The most frequent aberrations
in control cells were single dicentric anaphase bridges with a few
acentric fragments. Multiple bridges and fragments were the most
common aberrations in SN-treated cells. The higher concentration of SN
produced very complex aberrations, including decondensation of chromo
somes. Kinetochores of dicentrics are structurally normal. Micro
nuclei and very fine nuclear bridges connecting daughter cells in
cytokinesis in SN-treated cultures are enveloped by normal NE.
xvii


62
The difference in electron density between the kinetochore matrix
and its band on one hand, and the chromosome proper on the other hand,
is obvious in many of the figures cited, but particularly in grazing
sections of kinetochores (e.g., Figures 33b, 35c, and 36). Interest
ing, but very rarely seen, are chromatin "strands" between stretched
sister kinetochores (Figures 31 and 36). It is difficult to determine,
even at high magnification, if the fibers in these strands are finer
than normal chromosomal fibers, or if they are more densely packed,
and arranged more or less in parallel.
Metaphase
Cells in full metaphase are quite rare, probably because this
stage is of very short duration. Most cells judged to be in metaphase,
based on light microscopy, turn out, upon examination of thin sections,
to be either in very late prometaphase or very early anaphase.
Chromosomes at metaphase are aligned on the equator of the
spindle, at least with their kinetochore region (Figure 42).
Typically, the non-aligned telomeric portions of the long chromosomes
extend beyond the periphery of the spindle into the cytoplasm (insets
of Figures 42-44). All the kinetochores in normal metaphase cells are
triple-banded (Figures 42 and 43). Bundles of kinetochore MT converge
towards the poles (Figure 42). Within a bundle the MT are more or less
parallel, but a few non-kinetochore MT slant across. Wavy MT occur
mainly between chromosomes. Overall, the paucity of non-kinetochore MT
is remarkable.
Figure 43 represents a para-sagittal section. The plane of
sectioning was not precisely at a right angle to the chromosomes,


129
Fig* 33a* Late prometaphase. Three of the chromosomes shown in
Fig. 30a at higher magnification. Unobstructed kinetochores (Kq_ and
Kp of CI13, K£ of Ch2) and obstructed kinetochores (K-j_ of Chp, Kq of
Cn-j_). Note wavy and skew MT. x L0,000.


Fig. Ills,, b. Streptonigrin-treated cell in late telophase,
(a) Nuclear pockets; piece of RER trapped in the nucleus (arrear),
x 62,500 (h) Ribosome-like particles on outer and inner membrane
of NE (arrears). x 75,000. Both from the cell in Fig. Il6.
Fig. 118. Streptonigrin-treated cell in late cytokinesis.
Portion of the nuclear bridge (arrows) near the nucleus (N) of
one daughter cell, x 15,750.


68
63 and 64). C-mitotic cells are much less compressed along the axis
vertical to the growth surface than normal mitotic cells. The extreme
is illustrated by the near-spherical cell in Figure 63. Differences
among c-mitotic cells are evident in the variable shape at the level
of the light microscope (compare the insets of Figures 63 and 64).
Ultrastructural variations relate to the distribution of membrane
elements and mitochondria. For example, in the cell shown in Figure
63 membrane vesicles and cistemae, as well as mitochondria, are
distributed throughout the area occupied by the chromosomes, although
the former are more numerous at the periphery. In contrast, the area
of the chromosomes in Figure 64 contains very few small vesicles.
Numerous vesicles and larger cistemae, some of them rough ER, are
arranged almost concentrically at the periphery of this area.
Mitochondria are also excluded. In this respect the cell in Figure 64
resembles a prometaphase more than a metaphase, but the degree of
condensation of the chromosomes does not bear this out.
The centrioles, embedded in amorphous or finely fibrillar material
of moderate electron density, were always found at the periphery of
the area occupied by the chromosomes in the few cells examined (Figures
63 and 64). Four centrioles are present in each cell. They do not
differ structurally from centrioles in untreated cells.
Kinetochores appear as dense bands embedded in a less dense
fibrillar matrix and following the curvature of the chromosomal
surface, or as patches of less dense material (Figures 63, 64, 92-98).
This depends on the angle and the level of the section. No MT at all
were found in such cells. Treatment B (0.25 pg/ml colcemid for 15 rain,
no recovery) produces strikingly different effects. As with treatment


Fig. 5a-c. Interphase. Transverse sections of representative
nuclei, (a) Dispersed chromatin (Chr). Note granular (Gr), and
fibrillar (F) components of nucleolus (llu); intranuclear vesicle
containing particles (IV), nuclear envelope (HE). xl5,750. (b)
Moderate proportion of heterochromatin (HChr). x 15,750. (c)
Relatively great proportion of heterochromatin. x 11,500.


123
Fig. 27. Mid- to late prometaphase. Ch^, CI12, two equatorial
chromosomes. Cho a chromosome displaced towards pole no. 1 (see arrow
in inset). Kq., Kp, kinetochores oriented towards pole no 1 and pole
no. 2, respectively, x 7.750* Inset: Phase contrast micrograph'of the
cell in plastic (x l,28o).


231
obtained (e.g., Figures lib, 28, and 42). Wavy MT occur most often
between, or near, chromosomes (Figures 20, 25, 27, and 33). This might
well be an artifact, for as Jensen and Bajer (1969) demonstrated in
Haemanthus endosperm, shrinkage of chromosomes during dehydration can
cause adjacent ITT to become wavy.
I do not think that skew MT are an artifact (e.g., Figures 25, 27,
28, 42, and 47). One is tempted to speculate that the number of skew
and otherwise "maloriented" MT decreases from early prometaphase to
metaphase, thereby reflecting the organization of the mitotic spindle.
To substantiate such speculation would require a thorough analysis of a
great number of sections of several cells. For lack of numerical data
we must content ourselves with the explanation that skew MT are part of
the normal spindle and that they do not interfere with orderly chromo
some movement.
Microtubules appear as hollow cylinders whose lumen has the same
electron density as the cytoplasmic matrix (Figure 90). The walls
appear as dark, more or less sharp lines or circles in untreated cells
(Figures 84 and 90). In cold-treated cells the MT are coated by an
amorphous or finely granular, moderately osmiophilic substance (Figures
73 and 75). In dividing amebae (Chaos) exposed to cold the MA is
completely disorganized, but MT reappear after only a few minutes
recovery at normal temperature (Roth 1967). These MT are not coated by
any material. On the other hand, fixatives containing divalent cations
preserve a fine material on the surface of MT in the MA of the same
organism (Roth and Daniels 1962). Amebae fixed in OsO. alone contain
4
no MT, but finely fibrillar, linearly oriented elements in the spindle
region. Kane and Forer (1965) observed no MT, but large numbers of


205


112
*' K.v.yf
'41 i *- jf^i^s & ** A t i ^ 4
^R 28£f?#P* / -.. /'. ^1 i
#> 21a
^If^AaBSBW *. VKirkfl.*
^ **,
NP
T
n


96


136


203


22
differences in the method of preparation. A discussion of these is
beyond the scope of this summary. Perhaps Ris and Kubai (1970) are
closer to the truth in their conclusion that the 250 A fiber is the
structure present in the intact nucleus, but that its relationship to
the 100 A fiber is not yet clear.
The second question can be divided into two parts: (a) Does the
unit fiber contain a single DNA double helix or several in lateral
association? The DNA-histone complex is expected to have a diameter of
approximately 30 A (Zubay and Doty 1959). Many investigators have
claimed to have observed a fibrillar substructure within the 100 A
fiber (see Wolfe 1969, for references). However, enzyme digestion and
other treatments that are assumed to remove all or most of the histones
from the chromosomal fiber leave a strand 20 A thick, which most
probably represents a single ENA double helix (DuPraw 1968, Ris and
Kubai 1970, Wolfe 1969). It is thus generally accepted that the unit
fiber contains a single TOA double helix. However, there is ample
evidence that this helix is highly compacted in the unit fiber (DuPraw
1968, Ris and Kubai 1970). The major role in folding of the DNA has
been assigned to the histones, but the molecular basis is not yet clear
(see Ris and Kubai 1970, for discussion).
(b) Is the DNA double helix in the unit fiber continuous or does
it consist of subunits of variable length, perhaps connected by so-
called linkers? Sedimentation coefficients and direct electron
microscopic measurements of purified DNA from isolated chromosomes or
chromatin yield fragments of variable length (see Ris and Kubai 1970).
Similar results are obtained with autoradiography of DNA labeled during
replication. Both methods, however, have their limitations and do not


145
ai^? HIWmwL
>V. w* v?-.f .78& .
s*ss§
Fig. 44. Metaphase. Para-equatorial section showing three
circular kinetochores (arrowheads). Note clusters of MT between
chromosomes (*), intrachromosomal MT (circles), dense granules
(arrows). x 22,500* Inset: Phase contrast micrograph of the cell
in plastic (x l,28o).


Fig. 14. Late prophase. Chromosome (Ch) attached to the
intact nuclear envelope (HE) by a "stalk" (arrow), x 57,500.
Fig. 15. Late prophase. Sister kinetochores (Kq_, K2) of
two chromosomes (Ch^, Cl^). From the same cell as Fig. l5.
x 30,000.


71
75). Fully differentiated kinetochores exhibit the familiar triple-
banded profiles (Figure 75), though less distinctly than in normally
fixed control cells. The fuzziness of kinetochores and MT is not
caused by the initial cold fixation, because these structures are well
preserved in similarly fixed cells not previously exposed to cold
(Figure 74). The paucity of MT is also obvious in the midbody region
of cells in cytokinesis (Figure 76). Centrioles are not structurally
altered by exposure to cold (Figure 77).
Chromosomes are normally condensed, except for achromatic holes,
which are prominent at all stages of mitosis (e.g., Figure 71).
Remarkable is the great number and increased clarity of chromosomal
granules in cold-treated compared to control cells (Figures 70-72, and
75).
Kinetochore Fine Structure
I have already described in detail the fine structure of kineto
chores from prophase to late prometaphase. To establish a basis for
the discussion of kinetochore models, this section is devoted to an
in-depth study of kinetochores in normal meta- and early anaphase,
colcemid- and cold-treated cells.
Figure 78 shows five of seven serial sections in the horizontal
plane of a kinetochore in a cell in very early anaphase. Most of the
sections were too thick (approximately 800 A) to reveal details
concerning the attachment of MT. However, two things are very clear:
the triple-banding, and the greater length of the kinetochore in the
median sections (Figures 78b and 78c) compared to the peripheral
sections (Figures 78a, 78d, and 78e). Thinner sections (500-600 A),


MATERIALS AND METHODS
Cell Culture
The PtK^ cell line was initiated in 1961 by Kirsten H. Walen
(Walen and Brown 1962) from a kidney of an adult male rat kangaroo,
Potorous tridactylis (Harsupialia). The male karyotype consists of ten
autosomes, one X chromosome bearing the nucleolus organizer, and two Y
chromosomes (Figure 1. Shaw and Krooth 1964, Walen and Brown 1962).
Fig. 1. Karyotype of the male rat kangaroo (Potorous tridactylis).
Note the heterochromatic region near the centromere on the X chromo
some (nucleolus organizer). Drawn after a micrograph by Shaw and
Krooth (1964).
The Pt^ line, however, is aneuploid, the number of chromosomes in the
reference stock varying from 11 to 14 (Anonymous 1967). An extra, long
subtelocentric chromosome is present in the majority of cells.
45


214
&¡stk%
^ *T ^ -*J
v3*iKSB
t* >v ^
^
*/#


53
Osmiophilic, granular or fibrillar material surrounds the centriole at
one end (Figure 4a). The short rays radiating from the triplets
(Figure 4b) give the whole structure the appearance of a pinwheel.
Such images are rare. Dense, spherical particles with a lighter core
and aura (Figure 4c), often quite numerous, are exclusively associated
with centrioles and are present during all the stages of the cell cycle.
An intracentriolar vesicle, shown in an approximately median section in
Figure 4d, is often found in the lumen of centrioles. Satellites and
microtubules can be seen in all four sections (Figures 4a-d).
The appearance of the interphase nucleus in transverse sections
varies. Three representative nuclei are shown in Figure 5. The
chromatin is either dispersed (Figure 5a), or it includes heterochro-
matic chromocenters to a variable extent (Figures 5b and 5c). Nuclei
like the one in Figure 5c are very likely in late cytokinesis or very
early prophase. Of 34 interphase nuclei examined, 12 had dispersed
chromatin, the remainder contained chromocenters of variable size and
extent. At higher magnification, chromatin fibers, approximately 130 A
and 250 A in diameter, can be seen cut at various angles (Figure 6). A
very dense core, 70-80 A in diameter, appears in a few cross sections
of 250 A fibers. Nucleoli are large and prominent in interphase cells
(Figure 5). Granular and fibrillar components are mixed (Figure 7).
Nucleolus-associated chromatin is found within the nucleolus and at its
periphery (Figures 5 and 7). Intranuclear vesicles occur in some
nuclei (Figure 5a). The wall of these vesicles is finely fibrillar and
they often, perhaps always, contain granules or particles 250-400 A in
diameter. However, densely packed granules of similar size and
appearance also occur free in the karyoplasm.


107
Fig. l8a. Very early prometaphase. Note fragmenting nuclear
envelope (NE) near centrioles (c), intact NE on opposite side of
"nucleus". One fragment formed a vesicle (NV). x 11,500. Inset:
Phase contrast micrograph of the cell in plastic (x l,28o).


Fig. 48. Very early anaphase. Two pairs of daughter
chromosomes (Ch^, Civ,) with their kinetochores (K^, K) and
associated MT. Note electron dense strands connecting kineto-
chore regions of daughter chromosomes (arrows). Serial section
of the cell shown in Fig. 45. x 22,500*
Fig. 49. Early anaphase. Three representative kineto
chores (k). The angular kinetochore near the right margin is
cut obliquely or peripherally. Serial section of the cell shown
in Fig. 4y. x 30,000.


40
Streptonigrin: A Chemical Clastogen
Streptonigrin (SN) is a metabolite of Streptomyces flocculus. It
has the formula C24^22S^4 anc* t*ie fHw:*-nS structure (Rao et al.
The drug behaves as a weak acid with quinoid properties. It
induces phage release in lysogenic bacteria (Levine and Borthwick 1963)
and initiates rapid breakdown of JS. coli DMA _in vivo (Radding 1963).
Inhibition by SN of DNA synthesis and of DNA-dependent KNA synthesis
was reported by Koschei et al. (1966) for a cell-free system, and by
Young and Hodas (1965) for tissue culture cells. Streptonigrin caused
single strand breaks in calf thymus DNA (Mizuno and Gilboe 1970). The
latter authors also found that SN preferentially binds to DNA during
the S phase.
Cohen et al. (1963) investigated the effect of SN on cultured
human leukocytes. The mitodepressive effect of SN appeared related to
concentration and length of exposure. The mitotic rate was signifi
cantly depressed in cells exposed for 36 hr to 0.01 and 0.1 /jg/ml SN,
but not in cultures treated with 0.001 pg/ml SN. The chromosome-


138


50
rectangular area containing the cell or cells is done after curing of
the resin glue.
As a rule, blocks were sectioned in a plane parallel to the cell
monolayer (horizontal plane; see Figure 2). For special sections
(para-sagittal = vertical; para-equatorial = orthogonal to the spindle
axis; see Figure 2) blocks were mounted differently. A scratch was
made on the surface of the Epon wafer, under low power, either parallel
to the spindle axis or at a right angle to it and as close to the
selected cell as possible. A disk with the marked cell was cut out,
roughly trimmed, and mounted in the clamp chuck of an ultramicrotome.
By cutting thick sections with a glass knife the block could be faced
to the level of the scratch or even closer to the cell, when periodical
checking under a microscope was possible. If necessary, the angle of
the face relative to the spindle axis could still be corrected.
Finally, the block was mounted on a peg with the face up.
Serial sections were cut with a diamond knife on a Porter-Blum
MT-2 microtome and picked up on formvar-coated, carbon-stabilized
rectangular hole grids (LKB Instruments, Inc.). Sections were
poststained with uranyl acetate (Watson 1958) and lead citrate
(Reynolds 1963) and examined in a Hitachi HU 11-E electron microscope
with a 50-nm objective aperture and operated at 75 kV.




t
H
AS.
1 Z'-

r*
-H *TI
391


13
Following common usage in the literature, the terra spindle fiber
o-, 3 O'. r o O
shall hereafter designate the structure seen in the light microscope.
The term microtubule (MT) shall be reserved for the fibrils seen in
the electron microscope. By this definition a kinetochore fiber
consists of one or several MT.
Fine Structure
Spindle MT are slender cylinders 150-250 A in diameter and several
microns long (de Harven 1968, DuPraw 1968, Luykx 1970, Nicklas 1971,
Roth 1964). The tubules are hollow and have a wall 40-60 A thick. The
substructure of the wall is best resolved by negative staining of
isolated MT (Bamicot 1966, Kiefer et al. 1966), or by freeze-etching
(Moor 1967). Globular subunits, 30-40 A in diameter, form linear
filaments. Ten to thirteen such filaments form the tubule wall. The
filaments are arranged in such a way that the subunits form a helix
with a pitch of 10-20. Spindle MT are thought to be as rigid as
cytoplasmic MT (Luykx 1970), and therefore should run more or less
straight over a certain distance, or only be slightly curved. Wavy MT
can be a shrinkage artifact produced during dehydration (Jensen and
Bajer 1969).
Kiefer et al. (1966), citing a number of studies on the extraction
of proteins from mitotic apparatus, all of which seem to have yielded a
protein particle sedimenting at approximately 2.5S (MW approximately
34,000), proposed that this particle corresponds to the 35 A globular
subunit of the MT wall. Results from the various laboratories have,
however, been interpreted in different ways and the correlation between
visible subunits and isolated particles is not firmly established


106


251
I more or less expected the kinetochores in cold-treated cells to
resemble those in colcemid-treated cells. This was not the case
(compare Figure 75 with Figure 92). One significant difference was the
presence of kinetochore MT in cold-treated and their complete absence
in colcemid-treated cells. As already explained in the s'ection on
kinetochore maturation, it is not clear if a relationship exists
between microtubular attachment and the appearance of triple-layered
kinetochores. Observations on cold-treated cells might suggest such a
relationship, but in colcemid-treated Chinese hamster cells the inner
layer of kinetochores attached to persisting MT is apparently also
lacking (Brinkley and Stubblefield 1966).
Definite conclusions regarding the molecular composition of
kinetochores are not possible without extensive cytochemical studies.
The fact that the outer layer stains differently than the inner layer
suggests a different molecular composition. I tentatively conclude
that the outer layer is not made of chromatin, and that the inner layer
consists of tightly packed chromatin fibers. This rules out covalent
bonds between the two layers. The outer layer possibly consists of
protein. Since kinetochore MT are attached to the outer layer, the
implications are clear: the bonds between MT and the outer layer on
one hand, and outer and inner layers on the other hand must be very
strong, in fact stronger than the bonds maintaining chromosomal fibers
in a coiled state. This is borne out by the observations that kineto
chores of dicentric chromosomes subjected to considerable pull (Figures
101, 110a, Ilia, and 113) are structurally normal. Also, centrifuga
tion forces sufficient to uncoil chromosomes do not rupture spindle
attachments (Shimamura 1940). How this strong bonding is achieved


66
surface of the chromatin masses in late anaphase (Figure 52). Oblique
sections of such membranes reveal the presence of nuclear pore
complexes. Some of the cistemae obviously are, or have originated
from, RER as indicated by the presence of ribosomes (Figure 52). The
location of the membrane pieces varies, depending on the level of the
section. Usually, the majority of the membrane cistemae and vesicles
is apposed to the lateral faces of the chrriatin masses.
Figure 56 shows a pair of centrioles in early anaphase. As
during prometa- and metaphase, the moderately osmiophilic, amorphous
material surrounds that centriole on which the MT more or less
converge. Direct connections between MT and any part of the centrioles
are never observed.
Telophase and Cytokinesis
Reconstruction of the nuclear envelope continues during telophase,
apparently by coalescence of the small vesicles and cistemae seen in
anaphase cells. Extensively reformed NE first appears on the lateral
faces of the chromatin mass, as well as on the polar face, except
directly opposite the centrioles (Figures 57 and 58). Nuclear pore
complexes, complete with central granules, are present on chromatin-
associated membranes irrespective of size and location (e.g., Figure
58). The spacing between the two membranes of the NE remains irregular
until late telophase (Figures 57-60). Quite frequently, pieces of NE
are trapped within the reforming nucleus, probably giving rise to the
deeply invaginated pockets seen in many interphase cells.
Figures 60 and 61 illustrate progressive decondensation of the
chromatin. This process begins even before the NE is completely


8
duplex structure. (2) Duplication occurs in very early interphase or
in late telophase of the preceding mitosis. (3) The splitting of the
centers is a process distinct from duplication, although the two
usually occur at about the same time during mitosis. As the two
members of a pair of "old" centers split, each one gives rise to a new
one with which it remains associated until the next cycle. (A) The
primary duplication event involves only a part of the parent structure.
(5) Mercaptoethanol (ME) inhibits duplication, but not splitting and
separation of existing centers. If ME is applied prior to duplication,
two daughter cells result from the first division after release from
inhibition. Multipolar divisions ensue if ME is applied after
duplication.
Went (1966) did a similar experiment with sand dollar eggs.
Benzimidazone (BIA) inhibits cell division, but not duplication of
mitotic centers. After release from BIA-inhibition the cells cleave
into as many blastomeres as there had been mitotic centers. Comparing
these results with those from ME-treatment, Went proposed there should
be a pair of potential mitotic centers at each pole of a BIA-induced
tetrapolar mitotic spindle, while in the case of ME-induced tetrapolar
cells there should be a single center.
The formation of so-called daughter or procentrioles during
duplication has been confirmed by electron microscopy (Bernhard and de
Harven 1960, Brinkley and Stubblefield 1970, Erlandson and de Harven
1971, Murray et al. 1965, Stubblefield and Brinkley 1967). Pro
centrioles arise at the proximal end of the parent centriole and
approximately at a right angle to the latter. They are structurally
similar to the parent centrioles, but much shorter during the early
stages of duplication.


216


77
bands knotted together (Figures 92, 94, 97, and 98), as well as for
the fibrous structure in grazing sections (Figures 93, 95, and 96).
Further evidence for the contention that the bands are actually
transverse sections of irregularly undulated (K^ in Figure 92; Figures
94 and 97), convex (K in Figure 92), or almost flat sheets (K^ in
Figure 98), came from careful analysis of serial sections of chromo
somes whose orientation relative to the plane of sections was known
from phase contrast micrographs. The diameter of these sheet-like
kinetochores is 5,000-9,500 A. Values of 7,000-8,000 A are most
Common.
Figure 95 illustrates rather unusual, exotic kinetochore profiles.
Kinetochore no. 1 seems to consist of two bands converging at their
ends. Additional bands were visible at the same locus in three
adjacent serial sections. Similar observations were made on the
undulated kinetochore no. 2.
In contrast to the single-handed kinetochores of colcemid-treated
cells, the metaphase kinetochores of cold-treated cells are triple-
banded as in Figure 75. The bands are fuzzier, but their relative
opacity is very similar to that of kinetochores in untreated cells.
Kinetochores of cold-treated cells are also attached to MT, but these
re few in number.
Chromosomal and Mitotic Aberrations
Untreated Cells
Frequency of aberrations
To score the frequency of chromosomal and mitotic aberrations,
Epon wafers with the embedded cell monolayers were scanned with the


Fig. 86a-h. Eight serial sections of a metaphase kinetochore
(para-equatorial). Filled circles mark bypassing ME. (a) Open
circles mark kinetochore MT. (c) MT marked by arrows correspond
to those marked in (d). (d) Less opaque circles (arrows) are
terminals of ME; outer kinetochore layer (KO). (e) Inner layer
(Kl). (f-h) chromosome (Ch). From the cell in Fig. 41. x 62,500.


63
therefore only one of the two sister kinetochores can be seen. Kineto-
chore profiles are virtually identical with those seen in horizontal
sections (compare Figure 43 with Figure 42). Sister chromatids are
separated by grooves relatively devoid of ribosomes and ground sub
stance. However, the separation, which actually indicates late meta
phase to very early anaphase (see also inset of Figure 43), is in
complete, for large "bridges" still connect the sister chromatids.
In para-equatorial sections the metaphase plate appears as shown
in Figure 44. Three kinetochores with associated MT can be seen.
Other MT occur singly or in small clusters, or they are arranged in
bundles which may belong to kinetochores not included in this section.
Microtubules penetrating chromosomes are surrounded by a clear halo.
Anaphase
Sister chromatids, now more properly called daughter chromosomes,
separate from each other at the onset of anaphase. In the light micro
scope they still appear as parallel rods (Figure 45, inset). Because
the chromosomes are somewhat frayed, their separation at very early
anaphase (Figure 45) is less distinct in thin sections, except for the
kinetochore region. The latter is spaced farther apart than the
telomere region (Figures 45 and 46), and this trailing of the telomeres
is more pronounced during later anaphase (Figure 47; inset of Figure
50; Figure 51a). The daughter chromosomes lose their individuality
during late anaphase and each set forms a large mass of densely packed
chromatin near the respective pole (Figure 52). The nucleolus
organizer (NO) is enclosed in this mass.


Fig. 4a-d. Interphase. Four serial sections of a centriole.
(a) Osmiophilic mass near distal end. (b) Probably the distal
end. Note short rays radiating from triplets (arrows), (c)
Probably near-median section. Note osmiophilic mass forming
cylinder around the centriole, spherical particle (Pa), (d)
Median section revealing intracentriolar vesicle (CV). Note
satellites (s) in all four sections, x 100,000.


254
chromosomes near this site become attached to MT by one kinetochore.
Kinetochores are polarized in the sense that the outer, but not the
inner layer, can connect with MT. Therefore, attachment is possible
only to MT approaching from the side the kinetochore faces. Even if a
kinetochore could theoretically attach to MT approaching from the
opposite side ("from the back"), these MT would have to be bent at an
impossible angle. Consequently, for the meta- and submetacentric
chromosomes the initial unipolar attachment immediately establishes
bipolar orientation, especially if combined with tension. Therefore,
when one kinetochore is attached, the other has no choice but to attach
to the opposite pole. Telocentric chromosomes are more likely to
attach with both kinetochores to the same pole. In fact, the only
chromosome I observed displaying this configuration was a telocentric
chromosome in a TEPA-treated cell in mid-prometaphase. Bajer and Mol-
Bajer (1969), however, published a micrograph showing a metacentric
chromosome apparently attached to the same pole with both kinetochores.
Such observations are rare and, though almost certainly due in part to
neglect by many investigators, may reflect the rarity of the event.
For comparison, in crane fly meiosis ten per cent of the bivalents show
unipolar orientation in early prometaphase (Bauer et al. 1961).
If the two poles are still close together at the time the first
attachments are established, one might expect some kinetochores to
connect to both poles. I have never observed this, but Bajer and Mol-
Bajer (1969) reported such a case for Haemanthus endosperm.
It is possible that early unipolar attachments involving only one
kinetochore are maintained during migration of the centrioles and
formation of the definitive spindle axis, with the result that some
r


23
unequivocally support the idea that the DNA is discontinuous.
O'. j O '
Similarly, the observation of multiple replication points in eukaryotic
nuclei is no definite evidence (see DuPraw 1968), and the accepted
conclusion is that the DNA molecule in the unit fiber is continuous
(DuPraw 1968, Ris and Kubai 1970).
The third question, concerning the organization of the unit fiber
in metaphase chromosomes, is at the center of the uninemy-polynemy
controversy. The answer is very simple: the problem is not resolved
(DuPraw 1968, Ris and Kubai 1970, Wolfe 1969). Bajer (1965) observed
the half-chromatid structure of chromosomes in living Haemanthus
endosperm cells. When fixed metaphase chromosomes from Vicia cells
pretreated with 5-amino uracil are digested with trypsin, each
chromatid appears to consist of two subunits (Wolfe and Martin 1968).
In whole-mounted, isolated chromosomes, each chromatid contains many
fibers (e.g., Abuelo and Moore 1969, DuPraw 1968, Lampert 1969,
Stubblefield and Wray 1971, Wolfe 1965), which can be interpreted to
form longitudinal subunits of chromatids, although it is virtually
impossible to follow individual strands over a greater distance. Cells
exposed to X-ray or certain chromosome-breaking chemicals during
exhibit, at the following metaphase, bridges between sister chromatids
that involve apparent subunits of chromatids (subchromatid aberrations
or "side-arm bridges"; Brinkley and Humphrey 1969, Heddle 1969, Kihlman
1966). Kihlman, who formerly supported polynemy based on his
experiments with chromosome-breaking chemicals (e.g., Kihlman 1966),
has recently questioned the occurrence of true subchromatid aberrations
(Kihlman 1970), and favors the single-stranded "folded fiber" model of
DuPraw (1965, 1968).


72
also approximately median, of other kinetochores are presented in
Figures 81-84. These kinetochores were sectioned transversely, except
those shown in Figure 83, where obliquely cut kinetochore MT indicate
the sections were tangential. Kinetochore profiles very similar to
those in horizontal sections can be seen in para-sagittal sections
(Figure 85).
Kinetochores in metaphase cells are rarely flat. Most commonly,
they are undulated or S-shaped (Figure 42), more seldom concave
(Figure 85a). In very early anaphase the kinetochores of small
chromosomes in the center of the spindle are more or less flat (Figures
46, 48, and 82), and at a slightly later stage the kinetochores of such
chromosomes are convex (Figure 84). The kinetochores of the long
chromosomes at the periphery of the spindle are convex or more
irregular in early anaphase (Figures 47 and 49). In mid-anaphase,
S-shaped and more exotic profiles of kinetochores are prevalent.
Triple-banded kinetochores in paraxial median sections are 4,000-
6,700 A long. A faintly staining corona, approximately 400 A wide and
consisting of fine fibrils embedded in an amorphous matrix, covers the
kinetochores on the poleward side (Figures 78 and 84). The width of
the three bands varies, both within and between kinetochores. Average
values were 390 A for the outer, 270 A for the middle, and 400 A for
the inner band. The outer band consistently stains less intensely than
either the inner band or the chromosome proper (Figures 84 and 85a).
This is very clear in the image seen on the screen of the electron
microscope, but in micrographs printed on contrasty paper the differ
ence is obscured (e.g., Figures 78, 82, and 83). Figure 80 shows a
peripheral section of a stretched kinetochore in a late prometa- or


97
Fig. 9a-e. Very early prophase. Migration and structure of
centrioles. (a), (b) Two serial sections shoving one of each pair
of centrioles (C]_, C2). Note satellites (s). Black spots near C]_
in (b) are staining marks. x 1(0,000. (c-e) Three serial sections
of centriole no. 1 (C^). Note cartwheel structure vith hub and
spokes in (c) and (d), skev arrangement of tubular triplets; short,
osmiophilic bars between triplets in (e). x 122,500.




2
apparatus (MA) and chromosome segregation. Mitosis in newt heart
cultures (Bamicot and Huxley 1965), rat thymic lymphocytes (Murray
et al. 1965), HeLa cells (Robbins and Gonatas 1964), L strain
fibroblasts (Krishan and Buck 1965), and rat hepatoma cells (Chang and
Gibley 1968) served as the basis for this summary.
Prophase
The chromatin, which is dispersed during interphase, begins to
condense during early prophase. This process continues until, at the
end of this stage, individual chromosomes are clearly recognizable and
can be seen to consist of two sister chromatids. Kinetochores
(centromeres) may become visible during prophase in some cell types.
The centrioles may have duplicated during the previous interphase, or
they do so during prophase. So-called procentrioles, or daughter
centrioles arise at right angles to the separated parent centrioles.
At this stage the centrioles often lie within a pocket of the nuclear
envelope. A dense, osmiophilic mass can be seen associated with each
parent-daughter pair. As prophase progresses, the pairs migrate to
opposite poles. The timing of this event relative to the breakdown of
the nuclear envelope varies. In any event, microtubules (MT), which
radiate from the centrioles during early prophase, begin to invade the
nucleus as the envelope breaks down. Fragments of the latter appear
either as double stacks of membranes (Chang and Gibley 1968, Murray
et al. 1965), or as vesicles (Bamicot and Huxley 1965). The fibrillar
component of the nucleolus disappears (see also Hsu et al. 1965), the
granular component disperses and breaks up into smaller masses which
often remain attached to chromosomes during meta- and anaphase (Chang
and Gibley 1968).


250
proraetaphase kinetochores and kinetochores in colcenid-treated cells
are less condensed than mature kinetochores. The generally greater
diameter of kinetochores and the more extensive matrix in c-mitotic
compared to untreated metaphase cells further supports this interpre
tation.
Intuitively one expects a larger plate to be more easily bent than
a smaller one, particularly if it is also less rigid, as might be the
case with the less condensed outer kinetochore layer in c-mitotic
cells. With this in mind we can explain the strange kinetochore
profiles in Figure 95. They represent oblique sections of the highly
undulated outer layer (Figure 121d). Undulations are more likely to
occur on telocentric chromosomes, or on submetacentric chromosomes bent
at a sharp angle in the kinetochore region. Either of these
possibilities could apply to Figure 95. The near-circular kinetochore
profiles observed by Journey and Uhaley (1970) in vincristine-treated
HeLa and Chinese hamster cells could be the result of superficial
sections through cap-shaped kinetochores such as K in Figure 92, or
they represent near-terminal sections of subtelocentric or telocentric
chromosomes (as in Figure 12Id).
Brinkley and Stubblefield (1966) observed 50-80 A fibrils
apparently looping out from the main band of kinetochores in Chinese
hamster cells recovering from colcemid. I have not observed such loops
in colcemid-treated PtIC2 cells, but the matrix in which the outer
kinetochore layer is embedded does contain fibrils (e.g., Figure 94).
The fibrils observed by Brinkley and Stubblefield (1966) are possibly
also part of the matrix, or they could be characteristic for cells in
which the kinetochore recondenses and the number of kinetochore UT
increases during recovery from colcemid.


Fig. 78 (contd.). (d) Section no. 6. (e) Section no. 7.
From the cell in Fig. 45. x 62,500*
Fig. 79* Peripheral section of a kinetochore in very early
anaphase. Note the MT traversing in front of the kinetochore
(arrow). From the same cell as Fig. Jd. x 50,000.


261
Bajer, A., and J. Mol-Bajer. l$6l. UV microbeam irradiation of
chromosomes during mitosis in endosperm. Exp. Cell Res. 25:
251-267.
Bajer, A., and J. Mol-Bajer. 1963. Cine analysis of some aspects of
mitosis in endosperm, p. 357-409* In G. G. Rose (ed.) Cinemicro-
graphy in Cell Biology. Academic Press, New York.
Bajer, A., and J. Mol-Bajer. 1969 Formation of spindle fibers,
kinetochore orientation, and behavior of the nuclear envelope
during mitosis in endosperm. Fine structural and in vitro
studies. Chromosoma 27: 448-484.
Barnicot, N. A. 1966. A note on the structure of spindle fibers.
J. Cell Sci. 1: 217-222.
Barnicot, N. A., and H. E. Huxley. 1965. Electron microscope
observations on mitotic chromosomes. Quart. J. Microscop.
Sci. 106: 197-214.
Bauer, H., R. Dietz, and C. RSbbelen. 1961. Die Spermatocytenteilungen
der Tipuliden. III. Mitteilung. Das Bev/egungsverhalten der
Chromosomen in Translokationsheterozygoten von Tipula olercea.
Chromosoma 12: U6-I89.
Behnke, 0., and A. Forer. 1966. Some aspects of microtubules in
spermatocyte meiosis in a crane fly (llephrotoma sutural is Loev):
intranuclear and intrachromosomal microtubules. C. R. Trav.
Lab. Carlsberg 35: 437-455-
Belr, K. 1929* Beitrage zur Kausalanalyse der Mitose. II.
Untersuchungen an den Spermatocyten von Chorthippus (Stenobothrus)
lineatus Panz. Wilhelm Roux Arch. Entwicklungsmech. Organ.
118: 359-484.
Bernhard, W., and E. de Harven. i960. L*ultrastructure du centriole
et d'autres elements de lappareil achromatique. Proc. 4th
Int. Conf. Electr. Microscop. 2: 217-227.
Bessis, M., J. Breton-C-orius, and J. P. Thiry. 1958. Centriole,
corps de Golgi et aster des leucocytes. Etude au microscope
lectonique. Rev. Hematol. 13: 363-386.
Biesele, J. J. 1962. Experimental and therapeutic modification of
mitosis. Cancer Res. 22: 779737.
Bloom, W., R. E. Zirkle, and R. B. Uretz. 1955 Irradiation of
parts of individual cells. III. Effects of chromosomal and
extrachromosomal irradiation on chromosome movements. Ann.
N. Y. Acad. Sci. 59: 503-513.
Borisy, G. G., and E. V/. Taylor. 1967a. The mechanism of action of
colchicine. Colchicine binding to sea urchin eggs and the mitotic
apparatus. J. Cell Biol. 34: 535-548.


70
The cell in Figure 66 differs from the above description by the
absence of large membrane cistemae and the higher degree of chromo
some condensation. It is remarkable, in this context, that the
chromatin of all the interphase cells subjected to treatment B is
completely dispersed (compare Figure 67 with Figure 5). The fibers in
the primary constriction of the chromosomes in Figure 66 are also
finer, or more tightly packed, or both. In this case, however, a
distinct less dense kinetochore band is present in the vesicular space.
Despite the relatively high concentration of colcemid used in
treatment B, a few 1-T were found in interphase cells (Figure 68), but
not in c-mitotic cells. In some of the latter, numerous bundles of
microfibrils, approximately 70 A in diameter, were present in the
central area (Figure 69).
Mitosis in Cold-Treated Cells
Exposure to 0-4C does not accumulate metaphase cells. Cells at
all stages of mitosis are present in the unsynchronized cultures used.
The general ultrastructural features of prophase correspond to control
cells, except for the presence of intranuclear clusters of granules
approximately the size of ribosomes (Figure 70). Kinetochores of
cells in mid-prophase consist of less dense, fibrous material in
constrictions of chromosomes (Figure 72). Amitotic spindle in the
usual sense does not exist in prometa-, meta-, and anaphase cells,
although centrioles are situated at opposite poles (Figures 71 and
75). Most of the few MT present are associated with kinetochores.
An amorphous moderately osmiophilic substance coats the micro
tubules, whose walls appear as stark lines (Figures 73 and


235
spindle proper has been disorganized (Figures 61 and 62). Apparently
these MT are nore stable than others.
Chromosomes
Aberrations
The frequency of abnormal metaphase cells in the controls is
obviously exaggerated (Table 1). Many late prometaphase cells with
maloriented chromosomes (e.g., Figure 29) were scored as aberrant
metaphases, but most of these cells would probably have divided
normally, otherwise the frequency of aberrations in anaphase cells
would have been similar to that in metaphase cells. Nevertheless, the
frequency of aberrations may reflect instabilities of the cell line,
or it may be due to the presence in the culture of a biological
clastogen (virus or other microorganisms). Ualen (1965) reported
sister chromatid exchanges in the original culture of rat kangaroo
cells. Similarly, Levan (1970) described changes of chromosome number
and structure in the PtK^ line derived from a female rat kangaroo. He
interpreted these changes to reflect adaptation of the cells to the in
vitro environment. Heneen (1970) studied frequency and types of
aberrations in untreated PtK^ cells. Many of these were similar to
aberrations I observed in the controls.
The lower of the two concentrations of SN used (0.01 pg/ml) had
little mitodepressive effect (Table 2), but induced almost 507. aberra
tions (Table 3). In contrast, 0.05 ig/ml SN had a great mitodepressive
effect (Table 2), which accounts for the small sample of cells examined
for aberrations (Table 3). These results confirm the great potency of
SN qs a clastogen.


44
centriole pairs move to opposite poles and a normal metaphase plate is
established prior to anaphase segregation. Furthermore, at the
concentration used some MT were still present, notably on the
kinetochores facing the centrioles (see also Brinkley and Stubblefield
1966). Jokelainen (1968) furnished more evidence that the precise
mitotic stage at which cells are arrested may depend on the concentra
tion applied. He found no typical metaphases in fetal rat kidney tis
sue after treatment of pregnant rats with 0.12 mg/kg colchicine, but
metaphases were present in fetuses of rats treated with 0.08 mg/kg.
The conclusion, drawn from birefringence studies (Inou 1952) and
the behavior of chromosomes in colchicine-treated cells (Levan 1938,
Mol-Bajer 1958), that this chemical disorganizes spindle MT, has
received support from biochemical studies by Borisy and Taylor (1967a,
b). They found that colchicine binds to a 6S protein from isolated
mitotic apparatus. This protein showed good correlation with the
occurrence of MT and was considered to be a subunit of MT.
Statement of Purpose
The objective of this investigation was threefold: (1) To fill
gaps in our knowledge of the structural changes accompanying the build
up of the chromatic and achromatic apparatus during normal mitosis in
animal cells. Particular attention was devoted to controversial or
less well-studied aspects of mitosis, such as kinetochores and the
nuclear envelope. (2) To study possible structural alterations in the
mitotic apparatus produced by the agents applied or by the presence of
aberrant chromosomes (e.g., dicentrics). (3) To elucidate the fine
structure of aberrant chromosomes, with particular reference to
kinetochores and chromosome architecture.


238
To my knowledge chromatin strands connecting daughter chromosomes
in early anaphase (Figures 46-48, and 109) have not been previously
reported. In random sections of mitotic cells these strands, as well
as the cells themselves at this stage, can easily escape detection.
Nevertheless, it remains to be proven that this is a universal
phenomenon or that it is restricted to rat kangaroo cells. The densely
fibrillar structure of the strands is most easily explained as a
consequence of stretching. Attenuation could produce parallel align
ment of the fibers involved, which would allow tighter packing than
is possible for coiled fibers. This explanation is supported by
similarly packed fibers in dicentric bridges, where stretching cer
tainly occurs (Figures 110 and 111a).
DuPraw (1965, 1968) proposed with his folded-fiber model of
chromosomes that sister chromatids at metaphase are held together by
short segments of unreplicated ESTA in the centromere region.
Immediately preceding anaphase separation these segments would be
replicated. Numerous autoradiographic studies have demonstrated late-
labeling centromere regions in chromosomes of various cell lines (see
DuPraw 1968 for references), but the short burst of DMA synthesis prior
to anaphase, expected according to DuPraw's model, has not been'proven.
Be this as it may, the fact is that sister chromatids in whole-
mount preparations are connected in the centromere region by chromo
somal fibers (e.g., Abuelo and Moore 1969, DuPraw 1968, Stubblefield
and Uray 1971). It is possible that the chromatin strands I observed
in early anaphase cells are a consequence of the physical separation of
these connections. Logically, the strands can be detected in thin
sections only when the centromere regions are stretched, because in the


Fig. 16. Mid-prophase. Oblique section of a chromosome (Ch)
and one of its kinetochores (k). From the cell shovm in Fig. 10.
Note difference in electron density between chromosome and kinetochore.
x 62,500.
Fig. IT* Mid-prophase. Sister kinetochores (KKg) of a
chromosome in the nucleus shown in Fig. 11a. Note doubleness of
chromosome (Chd the two sister chromatids), kinetochore granules
(KG), x 67,500.


2 A3


233
colchicine. Here, then, are three instances where great numbers of
microfibrils were observed in cells exposed to spindle poisons.
Perhaps these fibrils represent a different aggregation state of ITT
subunits induced by the experimental conditions. Naturally, the corre
lation observed may be fortuitous. Further investigations, undoubtedly
necessary, will have to aim at demonstrating a negative correlation
between appearance and disappearance of microfibrils and MT.
There is evidence from immunological (Went 1960) and biochemical
studies (e.g., Sisken and Uilkes 1967, Wilt et al. 1967) that the major
proteins of the MA, among them MT proteins, are synthesized before the
onset of division. The formation of MT in the cytoplasm during
prophase can be explained as an assembly of preexisting subunits (see
also Roth 1964, Hicklas 1971), possibly directed by organizing centers
(Inou 1964, Inou and Sato 1967; the MTOC proposed by Pickett-Heaps
1969, 1971). Purely ultrastructural studies contribute little to the
solution of this problem.
If there is indeed a pool of MT subunits during prophase, the
absence of MT from the nucleus strongly suggests the NE is a real
barrier to these subunits. In all the mammalian cells studied so far
no MT are present within the nucleus during prophase (e.g., Krishan and
Buck 1965, Murray et al. 1965, Robbins and Gonatas 1964). At the time
the NE begins to break down near the centrioles, MT are numerous on the
outside, but very sparse on the inside of the envelope in PtK^ cells
(e.g., Figure 13b). This observation leaves no doubt that the NE is a
barrier to both subunits and assembled MT. In other organisms, how
ever, the spindle is intranuclear, and the ME apparently does not act
as a barrier (e.g.,
Aldrich 1969).


24
Isolabeling of chromatids (e.g., Peacock 1963) has always been a
strong argument in favor of polynemy, but it can be accommodated in a
single-stranded chromosome model (Comings 1971, Ris and Kubai 1970).
On the other hand, the observation that chromosomes replicate semi-
conservatively (Taylor et al. 1957) is in favor of uninemy. As Ris and
Kubai (1970) pointed out, the most compelling evidence against multi
strandedness comes from studies on the uniqueness or redundancy of DNA
sequences (e.g., Britten and Kohne 1968). In Drosophila and the mouse
the majority of DNA sequences are unique. This would exclude the
presence of two or more identical ENA strands per chromatid, which is,
of course, a prerequisite for multi-strandedness dictated by the
orderly segregation of genes at mitosis.
In view of these contradictory observations and results it seems
premature to propose a detailed model of the architecture of metaphase
chromosomes. For examples of such models the reader is referred to the
review by DuPraw (1968) and the paper by Stubblefield and Wray (1970).
Kinetochore Structure and Function
Light Microscopic Observations
Chromosomal granules, presumably corresponding to the centromere
studied by many later cytologists, were first described by Metzner
(1894), who called them "Leitkorperchen." By 1930 a number of
different terms were in use (see Schrader 1936). Sharp (1934)
introduced the term kinetochore," which has been used as a synonym of
"centromere" by many authors. For clarity I prefer to apply the term
centromere to the structure seen with the light microscope in the
primary constriction of chromosomes, and to reserve the term


18
significantly from the beginning of prometaphase to the end of
anaphase. Lateral displacement within the spindle is possible,
however. The authors found no evidence for interzonal connections
between separating chromosomes at anaphase.
There is a large body of evidence that kinetochore fibers play a
major, if not exclusive, role during metakinesis, i.e., those movements
of the chromosomes that begin with prometaphase and terminate with the
alignment of the chromosomes on the equator of the spindle (congres-
sion), as well as in the poleward movement of chromosomes during
anaphase. Bivalents of spermatocytes in prometaphase often have the
shape of an arrowhead with the kinetochore at the tip (Dietz 1969).
If one bivalent is stretched to a greater extent than its partner, the
movement is always in the direction of the pole towards which the
former is oriented. More direct evidence comes from UV microbeam
irradiation experiments. In newt fibroblasts irradiation of the
kinetochore, but not of other parts of a chromosome, stops prometaphase
movement and the irradiated chromosome never reaches the metaphase
plate (Bloom et al. 1955, Uretz et al. 1954). In Ilaemanthus endosperm
cells, on the other hand, similarly treated chromosomes do reach the
metaphase plate, although their paths of movement during prometaphase
may be altered and become very complex (Bajer and Mol-Bajer 1961).
Chromosome motion at metaphase is very slow and minimal (Dietz
1969, Hazia 1961). The equatorial position of the chromosomes is very
likely maintained by a balance of forces on opposed sister kineto-
chores. If one kinetochore pair of a bivalent in a grasshopper
spermatocyte in metaphase is irradiated with a UV microbeara, the
bivalent shifts towards the pole closest to the unirradiated


Fig. 121. Diagrammatic representation of the three-dimensional
structure of kinetochores. (a) Mature metaphase kinetochore
(compare with Figs. 78, 82, and 86). (b) and (c) anaphase
kinetochores (compare with Figs. k6-k$¡). (d) Sister kinetochores
of a telocentric chromosome in a colcemid-treated cell (compare
with Fig. 95)* (e) Sister kinetochores of a metacentric or sub-
metacentric chromosome in a colcemid-treated cell (compare with
Figs. 92-9, 96-98).


10
flagella and cilia see Dirksen and Crocker (1966), Renaud and Swift
(1964).
Two principal functions have been assigned to centrioles in
dividing animal cells: (1) Determination of the poles of the mitotic
apparatus, and (2) assembly of spindle MT (for detailed reviews see
Luykx 1970, Nicklas 1971, Pickett-Heaps 1969).
The axis of the mitotic spindle is not determined by the
orientation of the centrioles at the poles, except, possibly, in
special cases (de Harven 1968, Luykx 1970). A number of observations
seem to indicate that centrioles are indispensable as pole-
determinants. In normal bipolar mitosis there is one pair of
centrioles at each pole (Krishan and Buck 1965, Hurray et al. 1965,
Robbins and Gonatas 1964, Robbins et al. 1968, Stubblefield 1963,
Stubblefield and Brinkley 1967). Furthermore, cytasters produced in
sea urchin eggs by artificial activation are centered around poles
containing centrioles (Dirksen 1961). In contrast to this, the number
of poles in hybrid somatic cells often does not correspond to the
number of centriole pairs present (e.g., Yamanaka and Okada 1968), and
in cranefly spermatocytes mitosis occurs without centrioles under
certain conditions (Dietz 1959, 1966). Finally, mitosis in higher
plants proceeds without centrioles and many lower animals and plants
have pole-determinants other than typical centrioles (see Luykx 1970,
Pickett-Heaps 1969).
The idea that centrioles might be active in the assembly of MT
draws support from claims that spindle MT are directly connected to
centriole walls (Brinkley and Stubblefield 1970, Gall 1961), and from
the proposition that the poles may be an area where MT are assembled


Fig. 72. Prophase kinetochores of cold-treated cell. Sister
kinetochores (Kq., K) lie in cup-shaped depressions. Fine fibrils
seen to radiate from the less dense mass of Kp. Note chromosomal
granules (circles). Serial section of the cell in Fig. 71. x 140,000.
Fig. 73* Cold-treated cell in netaphase. MT at higher magnifi
cation. Note amorphous coating, stark lines of MT vails, x 100,000.
Fig. 74. Cold-fixed control cell in netaphase. Three kinetochores
(K) and numerous MT are visible. Compare with Fig. 76. x 30,000.


268
Kihlman, B. A. 1964. The production of chromosomal aberrations by
streptonigrin in Vicia faba. Mutation Res. 1: ^h-62.
Kihlman, B. A. 1966. Actions of Chemicals on Dividing Cells. Prentice-
Hall, New Jersey. 260 p.
Kihlman, B. A. 1970. Sub-chromatid exchanges and the strandedness of
chromosomes. Hereditas 65: 171-186.
Roller, P. C. 1953* Dicentric chromosomes in a rat tumour induced by
an aromatic nitrogen mustard. Heredity 6 (suppl.): 181-196.
Koopmans, A. 1958. Die Persistenz der Chromosomenspindelfaser.
Naturvissenschaften 45: 66-67.
Koschei, K., G. Hartmann, W. Kersten, and H. Kersten. 1966. Die
Wirkung des Chromomycins und einiger Anthracyclinantibiotica auf
die DHA.- abhang i ge Huele ins Sure-Synthes e. Biochem. Z. 3^' 76-86.
Krishan, A. 1968. Fine structure of the kinetochores in vinblastine
sulfate-treated cells. J. Ultrastruct. Res. 23: 134-143.
Krishan, A., and R. C. Buck. 1965. Structure of the mitotic spindle
in L strain fibroblasts. J. Cell Biol. 2b: *133-1+44.
La Cour, L. F. 1953 The physiology of chromosome breakage and
reunion in Hyacinthus. Heredity 6 (suppl.): 163-179*
Lambert, A, M. 1970. Etude de structures cintiques en rapport avec
la rupture de la membrane nuclaire, en debut de miose chez
Mnium hornum L. Organisation des centromeres. C. R. Acad. Sci. D,
270: 481-484.
Lampert, F. 1969* Feinstruktur und Trockengevicht menschlicher
Chromosomen. Quantitative Elektronenmikroskopie. Naturvissenschaften
56: 629-633.
Lea, D. E. 1962. Actions of Radiations on Living Cells, 2nd ed.
Cambridge Univ. Press, New York. 4l6 p.
Levan, A. 1938. The effect of colchicine on root mitoses in Allium.
Hereditas 24: 471-486.
Levan, A. 1954. Colchicine-induced c-mitosis in tvo mouse ascites
tumours. Hereditas 40: 1-64.
Levan, A. 1970. Contributions to the chromosomal characterization
of the PTK 1 rat-kangaroo cell line. Hereditas 64: 85-96.
Levine, M., and M. Borthvick. 1963. The action of streptonigrin on
bacterial DHA metabolism and on induction of phage production in
lysogenic bacteria. Virology 21: 568-574.
Lima-de-Faria, A. 1950. The Feulgen test applied to centromeric
chromomeres. Hereditas 36: 60-74.


69
A, the usual criteria for determining mitotic stages do not apply.
Nevertheless, light microscopic examination of cells subjected to
treatment B revealed stages resembling pro- and prometaphase rather
than metaphase. One cell was in late telophase or cytokinesis. The
chromosomes, fairly distinct in the light micrographs (insets of
Figures 65 and 66), were extremely difficult to see by direct
observation.
Electron microscopy confirmed the above observations and revealed
interesting details. The central area of the cells in Figures 65 and
66 consists of a coarsely granular or fibrillar ground substance in
which the chromosomes are embedded. The separation of this area from
the cytoplasm is almost perfect; with the exception of small membrane
vesicles all the larger organelles are excluded. In the cell in
Figure 65 large pieces of double membrane, probably fragments of the
NE, are present at the border between central area and cytoplasm.
Grazing sections revealed no pore-annulus complexes on these fragments,
although colcemid does not destroy their integrity on the NE of
interphase nuclei (Figure 68). Chromosomes are highly dispersed,
making the recognition of sister chromatids difficult (Figure 65).
The dark centromeric granules of chromosomes in the light micrograph
are patches or balls of more tightly packed, perhaps also finer,
chromosomal fibers. There seems to be one patch per sister chromatid,
and, as revealed by serial sections, connections between chromatids
exist in this region. Kinetochores in the usual sense are lacking,
but a vesicular space relatively poor in fibers and filled with a less
dense substance can be recognized adjacent to the dense patches.


43
The effect of colcemid on plant and animal cells is essentially
the same as that of colchicine (see Gelfant 1963), but the former is
less toxic and more efficient in animal tissue (Schar et al. 1954).
Levan (193£) coined the term "c-mitosis" for the peculiar
morphological changes in dividing Allium cells under the influence of
colchicine. Chromosomes in c-mitosis are scattered in the cytoplasm,
sister chromatids being held together in the centromere region. At
"c-anaphase," sister chromatids fall apart and at "c-telophase,"
because of the absence of a mitotic spindle, all the chromosomes are
included in one polyploid restitution nucleus (see also the cinemicro-
graphic analysis of c-mitosis in Haenanthus endosperm by Mol-Bajer
1958). In animal cells the division of the centromere region is
delayed until "c-telophase," i.e., "c-anaphase" is omitted (Levan
1954).
Besides Levan's (1938) classical scattered metaphase, other
chromosomal arrangements, such as "star-mitosis" and "clumped" or
"ball" metaphase, can be observed in both animal and plant cells (e.g.,
Gaulden and Carlson 1951, Deysson 1968). The effect depends on the
concentration and the time of exposure.
For many years it was commonly accepted that colchicine and its
analogs arrest dividing cells at metaphase. Sentein (1961) objected to
calling colchicine purely a metaphase poison, because he also found
arrested prophases, anaphases, and telophases in his material.
Brinkley et al. (1967) demonstrated that in Chinese hamster cells
arrested in mitosis by treatment xiith 0.06 ;ig/ml colcemid the two pairs
of centrioles are surrounded by the chromosomes in a configuration
different from a typical metaphase. After reversal of the inhibition,


199
Fig. 99. Dicentric bridge in an untreated anaphase cell. The
bridge does not appear continuous in this section. Note slight lagging
of one of the daughter X chromosomes (X2), its normal kinetochore (f^),
and the nucleolus organizer (NO). x 11,500* Inset: Phase contrast
micrograph of the cell in plastic. Note attenuation of bridge; frag
ments (arrow), and the X chromosome (X2); (x l,28o).


4
derived in part from the remnants of the nuclear envelope, appear at
O', J C < O/O
the periphery of the chromatin mass, first at the polar face.
Ribosomes and nuclear pores are found on some of these pieces. It is
not unusual for membranes to get caught between the coalescing
chromosomes. The presence of restored nuclear envelope seems
incompatible with the occurrence of ITT within the chromatin mass. In
the interzone, mitochondria and, in some cases, ER can be found, and
dense material around the MT in the equatorial region indicates early
stages of midbody formation.
Telophase
The nuclear envelope is completely reconstructed by fusion of
membrane vesicles and cistemae. Nucleoli are reconstituted from
material formed at the nucleolar organizer or by coalescence of small
nucleolar bodies, but the process is poorly understood (see Busch and
Smetana 1970). Concomitantly, the mass of chromatin disperses. The
cytoplasm is divided either by a wedge-shaped constriction in the
equatorial region (Robbins and Gonatas 1964), or by the formation of a
vesiculated equatorial plate (Murray et al. 1965). In the latter case
the vesicles fuse to form the cleavage furrow.
Cytokinesis
Cleavage progresses until the daughter cells are connected only by
a cytoplasmic stem. Included in the stem are tightly bundled MT and
the midbody, which has formed by fusion of the dense material in the
equatorial region (see also Byers and Abramson 1968, Paweletz 1967,
Schroeder 1970). The midbody is included in one of the daughter cells
or lost. Decondensation of the chromatin may not be completed until
late in interphase.


266
Heddle, J. A. 1969. The strandedness of chromosomes: Evidence from
chromosomal aberrations. Can. J. Genet. Cytol. 11: 783-793-
Heddle, J. A., and D. J. Bodycote. 1970. On the formation of chromosomal
aberrations. Mutation Res. 9' 117-126.
Henderson, S. A., and C. A. Koch. 1970. Co-orientation stability by
physical tension: a demonstration with experimentally interlocked
bivalents. Chromosoma 29: 207-216.
Henderson, S. A., R. B. Hicklas, and C. A. Koch. 1970. Temperature-
induced orientation instability during meiosis: an experimental
analysis. J. Cell Sci. 6: 323350.
Heneen, W. K. 1970. In situ analysis of normal and abnormal patterns
of the mitotic apparatus in cultured rat-kangaroo cells. Chromosoma
29: 88-117.
Hepler, P. K., and W. T. Jackson. 1968. Microtubules and early stages
of cell-plate formation in the endosperm of Haenanthus katherinae
Baker. J. Cell. Biol. 38: 437t446.
Hepler, P. K., J. R. McIntosh, and S. Cleland. 1970. Intermicrotubule
bridges in mitotic spindle apparatus. J. Cell Biol. 45: 438-1*44.
Hollaender, A. 1954. Radiation Biology, vol. 1 (pt. 2). McGraw-Hill,
Hew York.
Hollande, A. and J. Valentin. 1968. Infrastructure des centromeres et
droulement de la pleuromitose chez les Hypermastigines. C. R.
Acad. Sci. D, 266: 367-370.
Hsu, T. C., F. E. Arrighi, R. R. KLevecz, and B. R. Brinkley. 1965.
The nucleoli in mitotic divisions of mammalian cells in vitro.
J. Cell Biol. 26: 539-553.
Hsu, T. C., R. M. Humphrey, and C. E. Somers. 1964. Responses of
Chinese hamster and L cells to 2'-deoxy-5-fluoro-uridine and
thymidine. J. Nat. Cancer Inst. 32: 839-855.
Hu, F. 1971. Ultrastructural changes in the cell cycle of cultured
melanoma cells. Anat. Rec. 170: 41-56.
Hughes-Schrader, S. 1948. Cytology of coccids (Coccoldea Homoptera)
Adv. Genet. 2: 127-203.
Humphrey, R. M., and B. R. Brinkley, i960. Ultrastructural studies
of radiation-induced chromosome damage. J. Cell Biol. 42: 745-753.
Inoue, S. 1952. The effect of colchicine on the microscopic and
subnicroscopic structure of the mitotic spindle. Exp". Cell Res.
(suppl. 2): 305-314.
Inoue. S. i960. On the physical properties of the mitotic spindle.
Ann. N. Y. Acad. Sci. 90: 529-530.


173


19
kinetochore (Izutsu 1961). An exception to this rule was reported by
Forer (1966) for crane fly spermatocytes. Here, UV microbeam
irradiation of the chromosomal fibers on the poleward side of the
metaphase plate does not induce movement. A possible explanation for
this behavior is the apparent absence of any kind of kinetochore on
these chromosomes, as reported by Behnlce and Forer (1966). Finally,
the necessity of opposing poleward forces for stable metaphase
alignment was clearly demonstrated in Nicklas' laboratory (Henderson
and Koch 1970, Henderson et al. 1970, Nicklas and Koch 1969).
It is a well-supported conclusion that chromosomes move as
individuals, although they may move synchronously (Luykx 1970, Mazia
1961, Nicklas 1971). Since Fas' study (1949) two types of anaphase
movement are distinguished: (1) spindle elongation, and (2) shortening
of the chromosome-to-pole distance. The two processes are based on
different mechanisms, because the former, but not the latter, is
inhibited by chloral hydrate (Ris 1949). In grasshopper spermatocytes
the two processes act together (Ris 1949), but in other cells each of
the two possible extremes can occur (see Mazia 1961 for references).
Mazia (1961), who also discussed the various early hypotheses
concerning anaphase movement, summarized the events as follows: "The
central spindle is more or less rigid; it moves the poles apart and
provides an anchor for the poles which must bear the load of the
chromosomes."
The question whether chromosomes are pulled or pushed has engaged
the mind of many a biologist. Most reviewers (Dietz 1969, Luykx 1970,
Mazia 1961, Nicklas 1971) arrived at the conclusion that a pulling
force must be involved in the movement of chromosomes, although a


Fig. 102a, t>. Kinetochores of lagging chromosomes in un
treated anaphase cell, (a) Chromosome no. l(see inset), (h)
Chromosome X2. Note bypassing MT (large arrows), kinetochore MT
(small arrows). Fuzzy kinetochore material is indicated by
arrowheads. Dark spot left of center in (b) is staining artifact,
x 50,000. Inset: Phase contrast micrograph of the cell in
plastic. Note position of three laggards (no. 1, no. 2, Xp);
(x 1,280).


125
Fig. 29. Late prometaphase. Note the small chromosome near
pole no. 1 (P-^). Kinetochores are marked by arrowheads. Fragments
of the nuclear envelope (NE) at the periphery of the spindle. Pieces
of double membrane with ribosomes (arrows), x 7,750* Inset: Phase
contrast micrograph of the cell in plastic (x 1,280).


!*:7. /* ^
122
Fig. 2: Mid-proraetaphase. P^, P^, the two poles of the mitotic
spindle. A centriole is visible at pole no. 1. Several kinetochores
are indicated by arrowheads, x 7,750* Inset: Phase contrast 'micro
graph of the cell in plastic (x l,28o).


REFERENCES
Abuelo, J. G., and D. E. Moore. 1969. The human chromosome. Electron
microscopic observations on chromatin fiber organization. J. Cell
Biol. 1*1: 73-90.
Aikawa, M., and R. L. Beaudoin. 1968. Studies on nuclear division of
a malarial parasite under pyrimethamine treatment. J. Cell Biol.
39: 7^9-75^.
Aldrich, H. C. 1967. The ultrastructure of meiosis in three species
of Physarum. Mycologia 59: 127-148.
Aldrich, H. C. 1969. The ultrastructure of mitosis in myxamoebae
and plasmodia of Physarum flavicomum. Amer. J. Bot. 56: 290-299
Anonymous. 1967. American Type Culture Collection Cell Repository.
Registry of animal cell lines certified by the cell culture
collection committee, suppl. 2.
Bacq, Z. M., and P. Alexander (eds.). 1955 Radiobiology Symposium
1954. Academic Press, New York. 3^2 p.
Bajer, A. 1958 Cine-micrographic studies on chromosome movements
in ^-irradiated cells. Chromosoma 9: 319331
Bajer, A. 1963 Observations on dicentrics in living cells.
Chromosoma 14: 18-30.
Bajer, A. 1964. Cine-micrographic studies on dicentric chromosomes.
Chromosoma 15: 630-651.
Bajer, A. 1965 Subchromatid structure of chromosomes in the living
state. Chromosoma 17: 291-302.
Bajer, A. 1968. Behavior and fine structure of spindle fibers dur
ing mitosis in endosperm. Chromosoma 25: 249-281.
Bajer, A. 1969 Effects of UV microbeam irradiation on chromosome
movements and spindle fine structure (with film demonstration).
Biophys. J. 9: A-151.
Bajer, A. 1970. See Luykx, 1970, p. 54.
Bajer, A. and R. D. Allen. 1966. Structure and organization of the
living mitotic spindle of Haemanthus endosperm. Science 151:
572-574.
260


17
participate in the movement of chromosomes with a well-defined
kinetochore is uncertain. The often cited observations by Carlson
(1938) on apparent poleward movement at anaphase of X-ray-induced
acentric fragments in grasshopper neuroblasts are not convincing.
Bajer (1958) and Bajer and Mol-Bajer (1963) presented better evidence
based on cinemicrography of /3-irradiated Haemanthus endosperm cells.
In the majority of cases, however, acentric fragments do not behave as
normal chromosomes (Kihlman 1966, Lea 1962).
Finally, continuous MT are supposed to produce spindle elongation
at anaphase in cells where this occurs (Mazia 1961, Roth et al. 1966).
Spindle elongation is a process different from poleward movement of
chromosomes, as demonstrated by Ris' (1949) observation that chloral
hydrate prevents the former, but does not inhibit the latter.
The chromosomal fibers form the "chromosomal spindle" (Mazia
1961). Chromosomes with a distinct kinetochore are firmly attached to
the kinetochore fibers. This can be inferred from the above-mentioned
study by Ris (1949), and from observations by Inoue (1952) that in
Chaetopterus eggs treated with colchicine the chromosomes disperse from
the metaphase plate only after the birefringence of the chromosomal
fibers has completely disappeared. Shimamura (1940) reported that
chromosomes at the centripetal pole of centrifuged lily pollen mother
cells are firmly anchored by their kinetochore fibers, although the
centrifugal force is strong enough to cause uncoiling of the
chromosomes themselves. Finally, the most direct evidence came from
elegant micromanipulation experiments by Nicklas and Staehly (1967) on
grasshopper spermatocytes. Chromosomes can be stretched with a
microneedle without changing the kinetochore-to-pole distance


190
re -SnSp'SLi&ss
j. tmS$. E2*r*K. H ...-_i£^_.| __ ||M
3 4\ .,
. ... *'v*_JJ,'' sf*.'
^UKS^agzafcA*($£ WWvV 4* l4.
. *.** ,* -
nsfi- ;~V-
L X mpv
rsr
y.1
Vv£
0
% "*
-
^V
* ^>* I


252
remains unexplained, but the matrix persisting as the corona and
middle layer of mature kinetochores may play an important role.
Kinetochore-Hicrotubule Interactions
Kinetochores are involved in two important processes during
prometaphase: Chromosome orientation and congression. A third
possible function, assembly and orientation of IfT, is central to most
models of mitosis (Dietz 1969, Inou 1964, Luylcx 1970, McIntosh et al.
1969), but the available evidence is not entirely conclusive (for
detailed discussions see Luykx 1970, Nicklas 1971). Neither light
microscopic nor ultrastructural studies alone can answer the questions
remaining. A cinemicrographic and electron microscopic analysis of the
effect of UV microbeam irradiation and spindle poisons is more likely
to yield significant data. The electron micrographs of PtK^ cells in
very early prometaphase (Figures 18a and 19) can be interpreted as
supporting either the idea that kinetochores organize 1T or that they
attach to existing ITT. Counts of the number of MT before and after
kinetochore attachment, as presented by Mantn et al. (1969a, b) for
the diatom Lithodesmium, might provide clues, but spindle formation in
PtK^ cells is more complex.
Chromosome orientation has been extensively investigated in the
first meiotic division in spermatocytes (e.g., Nicklas 1967). There,
spindle attachments of chromosomes are often broken naturally, or can
be broken experimentally, and subsequently reorientation of chromosomes
and reattachnent occur (see also Nicklas and Koch 1969). This is
necessary for stable bipolar orientation of bivalents initially
attached to one pole only (maloriented bivalents). Mo similar data are
available for mitosis. Host of the current models of mitosis, which,


140


Fig. 106a, b. Abnormal centrioles in an untreated cell in
late prometaphase. Two serial sections. Three centrioles (C-i,
C2, C3) are visible. Note cup-shaped C-j_. MT converge on osmiophilic
masses left and right of Cp. From the cell in Fig. 29, P2*
x 57,500.


16
birefringence more rapidly than the chromosomal fibers (Inou 1952).
Sauaia and Mazia (1961) found that in sea urchin eggs the asters, but
not the kinetochore fibers, are disorganized by brief exposure to
colcemid. Likewise, a low concentration of colcemid disorganizes
continuous MT in Chinese hamster cells, but some kinetochore MT remain
(Brinkley et al. 1967). Similar differences apply for cold treatment.
Kinetochore fibers in lily pollen mother cells in anaphase regain
birefringence first when the cells are brought to ambient temperature
after chilling (Inou 1964). In contrast to this, continuous fibers
in Chaetopterus eggs are the first to regain birefringence lost during
chilling (Inoul 1964). In mammalian cells in vitro Brinkley and
Cartwright (1970) found cold shock completely disorganized interpolar
MT, while the number of chromosomal MT was reduced by 30-40%.
Function
It has been proposed that continuous fibers, which form the so-
called "central spindle" (Mazia 1961), function in the separation of
the pole-determinants (centrioles where applicable. Brinkley et al.
1967, Brinkley and Stubblefield 1970, Friedlander and Wahrman 1970,
Mazia 1961). In Chinese hamster cells exposed to colcemid, chromosomes
arrange more or less radially around the two unseparated pairs of
centrioles (Brinkley et al. 1967). The reformation of pole-to-pole MT
after release from the inhibitor seems necessary for the separation and
migration of the centrioles.
In cells with diffuse kinetochores, the penetrating or
transchromosomal MT may play an important, if not exclusive, role in
chromosome movement (Luykx 1970, Nicklas 1971). IThether they also


150
:i( -ti cttfiri
\*v-- v* *Jr4% TfJrm
1 Stlijw /Ct3Cf ** ijr^
yV*, *.:
VSkT *m&> v itU^
^iSSfcSfegia:^^ 3sT ^
fe *,ytjp
jf iJ#*.
T 'Til
I *v
&>;*
r- ic-'t?-


108
Fig. 18b. Very early prometaphase. Serial section of the cell
in Fig. l8a. Nuclear envelope (NS) undulated around chromosomes (Ch),
fragments formed vesicles (NV). Note nuclear pores (NP). x 15,750.


15
subcategories of spindle MT. Notable among them are the penetrating
or transchromosomal MT (Nicklas 1971) observed in many animal cells
(e.g., Behnke and Forer 1966, Buck 1967, Jokelainen 1967, Robbins and
Gonatas 1964). They may actually be part of the population of
continuous MT.
It is obviously difficult to demonstrate that continuous MT do
run from one pole to the other, because they often pass into and out
of sections. However, this has been demonstrated in two cases
involving relatively short spindles (Aikawa and Beaudoin 1968, Mantn
et al. 1969b). On the other hand, studies on late stages of mitosis
indicate most "continuous" MT overlap in, and terminate beyond, the
midbody (Byers and Abramson 1968, Paweletz 1967). Understandably, this
point, which is important for an explanation of the function and
mechanics of the MA, is still controversial. More detailed studies,
involving counts of MT in serial sections cut at a right angle to the
spindle axis, have only recently been published (Brinkley and
Cartwright 1970, Mantn et al. 1969b, McIntosh and Landis 1971). It
appears that relatively few MT [l07. in the diatom Lithodesmium
(Mantn et al. 1969b, 1970); 30-407. of the interpolar MT which make up
40-507. of all the spindle MT in Chinese hamster and rat kangaroo cells
(Brinkley and Cartwright 1970)] are truely interpolar. Most
"continuous" MT project across and terminate beyond the equator.
It is generally agreed (see Luylcx 1970) that there is no
difference regarding structure and dimensions between the different
classes of spindle MT. However, kinetochore MT differ from continuous
and astral MT in their sensitivity to spindle poisons. Upon exposure
f Chaetopterus eggs to colchicine, the continuous fibers lose their


110
Fig. 20. Early prometaphase. Formation of the spindle. Most
MT connect kinetochores (k) with the poles (Pp, P2). Note vesicle (v)
and elongated cisternae (Ci) at the proximal end of one of the centrioles
(C) at pole no. 1 (P-^). x 11,500. Inset: Phase contrast micrograph
of the cell in plastic; chromosome no. 1 is shown in Fig. 2b (x l,28o).


275
Yamanaka, T., and Y. Okada. 1968. Cultivation of fused cells re
sulting from treatment of cells with HVJ. II. Division of
binucleated cells resulting from fusion of two KB cells by
HVJ. Exp. Cell Res. k9: kGl-k6$.
Young, C. W., and S. Hodas. 1965. Acute effects of cytotoxic com
pounds on incorporation of precursors into DM, RIIA. and protein
of HeLa monolayers. Biochem. Pharmacol. l4: 205-21^.
Zirkin, B. R., and S. L. Wolfe. 1970. The chemical composition of
nuclei and chromosomes isolated by the Langmuir trough technique.
Chromosoma 32:162-170.
Zubay, G., and P. Doty. 1959* The isolation and properties of
deoxyribonucleoprotein particles containing single nucleic acid
molecules. J. Mol. Biol. 1: 1-20.


169
mm


57
latter fragments, and they gradually lose their "mottled" appearance.
Most kinetochores in very early prometaphase still resemble prophase
kinetochores (Figure 22a), but some are differentiating into more com
plex structures. For example, the sister kinetochores of the chromo
some in Figure 22b, although apparently not attached to MT, exhibit a
slightly denser band within the fibrillar matrix. Such internal struc
tures may be more common than it appears, but due to their relative
indistinctness they would not appear in ever so slightly oblique sec
tions. A rather unusual case is illustrated in Figures 22c and 22d.
This chromosome of the cell shown in Figure 18 was situated in the area
distant from the centrioles, i.e., near the intact portion of the
nuclear envelope. Several obliquely sectioned MT lie outside the
envelope opposite the "outer" kinetochore, which also contains a short
band. Serial sections revealed no MT inside the nuclear envelope. The
other sister kinetochore, facing in the direction of the distant
centrioles, resembles a typical prophase kinetochore.
As prometaphase progresses, kinetochores become more variable in
appearance. Structural differentiation depends in each individual case
upon the position and orientation of the chromosome relative to the
spindle poles. Sister kinetochores of chromosomes near a pole are
dissimilar (Figures 23 and 24). As a rule, the kinetochore facing the
near pole is attached to MT and consists of moderately opaque, finely
fibrillar material (K, K^, in Figure 20; K in Figures 23 and 24). A
banding pattern is discernible in some of these kinetochores (Figures
23a and 23b), which are quite often also stretched (K, K in Figure
23a; in Figures 24b and 24c). This makes it impossible to determine
where the MT end. In contrast, the kinetochore facing the distant pole


267
Inou, S. 1964. Organization and function of the mitotic spindle,
p. 949-5914. In R. D. Allen and R. Kamiya (eds.) Primitive
Motile Systems in Cell Biology. Academic Press, New York.
Inou, S., G. W. Ellis, E. D. Salmon, and J. W. Fuseler. 1970.
Rapid measurement of spindle birefringence during controlled
temperature shifts. J. Cell Biol. 47: x95a-96a
Inou, S., and H. Sato. 1967. Cell motility by labile association of
molecules. The nature of mitotic spindle fibers and their role in
chromosome movement. J. Gen. Physiol. 50 (6, pt. 2): 259-288.
Izutsu, K. I96I. As cited by Luykx, 1970.
Jagiello, G. 1987. Streptonigrin: effect on the first meiotic
metaphase of the mouse egg. Science 157: 453-454.
Jensen, C., and A. Bajer. 1969. Effects of dehydration on the
microtubules of the mitotic spindle. Studies in vitro and
with the electron microscope. J. Ultrastruct. Res.26):
367-386.
Jokelainen, P. T. 1965a. Fine structure aspects of kinetochores in
dividing cells from fetal rat kidney. Anat. Rec. 151: 367.
Jokelainen, P. T. 1965b. The differentiation of sister kinetochores
during metakinesis. J. Cell Biol. 27: 48a.
Jokelainen, P. T. 1967* The ultrastructure and spatial organization
of the metaphase kinetochores in mitotic rat cells. J. Ultra
struct. Res. 19: 19-44.
Jokelainen, P. T. 1968. The effect of colchicine on the kinetochores
and mitotic apparatus in the rat. J. Cell Biol. 39: 68a-69a.
Journey, L. J., J. Burdman, and P. George. 1968. Ultrastructural
studies on tissue culture cells treated vith vincristine (RSC-
67574). Cancer Chemother. Rep. 52: 509-517.
Journey, L. J., and A. Whaley. 1970. Kinetochore ultrastructure in
vincristine-treated mammalian cells. J. Cell Sci. J: 49-54.
Kane, R. E., and A. Forer. 1965. The mitotic apparatus. Structural
changes after isolation. J. Cell Biol. 25 (3, pt. 2): 31-39.
Kaufmann, B. P. 1954. Chromosome aberrations induced in animal
cells by ionizing radiations, p. 627-711. In A. Hollaender (ed.)
Radiation Biology, vol. 1 (pt. 2). McGraw-Hill, New York.
Kiefer, B., H. Sakai, A. J. Solari, and D. Mazia. 1966. The molecular
unit of the microtubules of the mitotic apparatus. J. Mol. Biol.
20: 75-79.


236
The question whether the SN-treated abnormal cells represented
the first or second generation after treatment cannot be answered.
The 48 hr recovery period was long enough for one round of divisions,
but the nitodepressive and general cytotoxic effects most likely
altered the division cycle. However, it is very unlikely that any of
the aberrations were induced in GIndeed, no subchromatid aberra
tions were observed.
Fine Structure
The analysis of thin sections of dicentric bridges revealed little
about the composition of the chromosomes involved. A comparison of the
number of fibers in dicentrics attenuated to a variable degree might
have provided some insight into the coiling or lateral association of
the fibers, but this was not feasible.
Electron micrographs such as Figures 27 and 30b demonstrate that
in thin sections individual chromosomes do not always appear physically
separated. This posed a certain problem for the analysis of dicentric
bridges such as shown in Figures 99-101. In some of the serial
sections (e.g., Figure 100) the impression was not one of two truely
linked chromosomes, but of long chromosomes whose telomeric ends were
twisted around each other. The attenuation of the bridge (Figure 99)
and the presence of acentric fragments (Figures 99 and 101), however,
were evidence for a real exchange.
In the phase contrast microscope the chromosomes in many of the
cells treated with 0.05 pg/ml SN appear fainter and more slender than
normal chromosomes. Examination of thin sections of these chromosomes
revealed generally looser, but also more irregular packing of fibers
(Figure 111). Most likely, these chromosomes condensed during the


Fig. 113. Streptonigrin-treated cell in late anaphase. Note
position of the kinetochores (k) relative to the pole (which was
just beyond the left margin). From the cell in Fig. 108d. x 30*000.
Fig. Il4. Streptonigrin-treated cell in late telophase.
Portion of the nucleus shown in Fig. 116. Note nuclear pocket nea,r
left margin. Osmiophilic patches (arrowheads) are presumably
remnants of kinetochores. Another kinetochore (K) within the
nucleus; note its associated MT (arrows), x 50,000.
Fig. 115. Streptonigrin-treated cell in late telophase.
Nuclear pocket with remnant of kinetochore (black arrow). A MT
penetrating into the nucleus (white arrow), x 75*000.


128
Fig. 32. Mid-proraetaphase. Unusual arrangement of MT.
x 22,500.


21
relationship between the DMA double helix and the unit fiber? (3) How
is the unit fiber arranged in the highly condensed metaphase chromo
some? Data and observations bearing on these questions come from a
number of fields of study and so far it has not been possible to
reconcile them in one comprehensive model of the eukaryotic chromosome.
In thin sections of interphase nuclei and mitotic chromosomes
fibrils approximately 100 A in diameter are visible (Wolfe 1969).
Because the fibers are cut at various angles, little can be concluded
regarding their three-dimensional organization. Whole-mount
preparations of isolated chromosomes seemed initially much more
promising (for references see Wolfe 1969). In isolated chromosomes the
diameter of the fibers varies from approximately 20 A to 500 A or more,
depending on the quality of the preparation, but most investigators
agree that the mean diameter is approximately 250 A (DuPraw 1968, Wolfe
1969). However, Ris has demonstrated (1961, 1967; Ris and Kubai 1970)
that 100 A fibers can consistently be obtained. Understandably, a
lively controversy revolves around these differing results and their
interpretation. DuPraw (1968) considers the 250 A fiber as the unit
fiber (type B fibril). On the other hand, Wolfe (1969; see also Zirkin
and Wolfe 1970) considers this to be an artifact produced by the
deposition on a 100 A fiber of contaminating material during
preparation. A third view is held by Ris (Ris 1967; Ris and Kubai
1970): after certain chemical treatments the 250 A fibers can be
shown to consist of two 100 A fibrils, more or less twisted around each
other. Both Ris (Ris and Kubai 1970) and Wolfe (1969) present
arguments and evidence supporting their respective hypothesis. At the
center of the dispute are the many possible artifacts produced by


Fig. 100. Dicentric "bridge in an anaphase cell (serial
section of the cell in Fig. 99)* Note normal kinetochore (K2)
of "bridged chromosome, x 15,750.
Fig. 101. Dicentric "bridge in an untreated anaphase cell.
Note normal kinetochore (K^) of bridged chromosome, x 11,500
Inset: Phase contrast micrograph of the cell in plastic. Arrow
marks location of fragments (x l,28o).


35
The often used term "radiomimetic" chemicals for substances that
induce chromosome damage is rather misleading. As already mentioned,
ionizing radiation produces both non-delayed and delayed effects, while
most chemicals (e.g., alkylating agents) produce delayed effects only
(Kihlman 1966). Exceptions are streptonigrin (SH) and S-ethoxycaffeine
(EOC), which produce effects very similar to X-rays. Increased oxygen
tension, which drastically increases the frequency of aberrations
induced by X-rays (Bacq and Alexander 1955, Evans 1962, Lea 1962) has
little or no effect with many chemical clastogens (Kihlman 1966).
Perhaps the most significant difference between X-rays and chemicals is
that the former produce aberrations more or less randomly in a
particular chromosome, while aberrations induced by the latter tend to
be localized in the heterochromatin (Nichols et al. 1964, Revell 1963).
Hypotheses on the Formation of Chromosomal Aberrations
The general or breakage-first hypothesis as described by Kihlman
(1966) proposes that the primary event produced by a clastogen is a
chromatid or chromosome break in a continuous interphase chromosome.
The ends at the point of breakage may rejoin to restore the original
configuration (restitution), they may remain open, or they may rejoin
with other open ends. Illegitimate fusion of ends from different
breaks results in sister-union or various types of exchanges.
According to the exchange hypothesis proposed by Revell (1955; see
also Revell 1963) the primary event is not a break, but some other kind
of lesion. The lesion may revert to normal or to another state
incapable of forming an exchange. If two primary events occur close
enough in space and time, an exchange initiation stage may follow.


143
W
Fig. 42. Metaphase. Chronosorr.es aligned on the metaphase plate
(horizontal section). Sister kinetochores (arrowheads) oriented
twards opposite poles. Bundles of kinetochore MT converge towards
the poles. Note dense chromosomal granules (arrows), x 15,750 Inset:
Phase contrast micrograph of the cell in plastic (x l,28o).


160


TABLE OF CONTENTS
Page
Acknowledgments iii
Preface iv
List of Tables vii
List of Figures viii
Key to Abbreviations and Symbols xiv
Abstract xv
Review of Literature 1
General Ultrastructural Features of Vertebrate l
Cells in Mitosis
Centrioles 5
Spindle Fibers 11
Chromosomes 20
Kinetochore Structure and Function 24
Chromosomal and Mitotic Aberrations 33
Statement of Purpose 44
Materials and Methods 45
Cell Culture 45
Chemical and Physical Treatments 46
Fixation and Embedding 48
Preparation of Cells for Light and Electron
Microscopy 49
Results and Observations 52
Normal Mitosis 52
v


75
The kinetochore of Figure 86 was most likely similar to those in
Figures 78, 81, and 82. The first section of the kinetochore itself
shows a moderately opaque patch of finely fibrillar material (Figure
86d). I interpret this as the outer layer. Fewer ITT are visible than
in the preceding sections, indicating they terminate at this level.
Remarkable are less opaque circles whose diameter is similar to the
inner diameter of MT. They mark the terminals of kinetochore MT, the
wall of which cannot be seen because its opacity is the same as that
of the outer layer. A few MT can be followed one section farther
(Figure 86e), because the kinetochore is not perfectly flat, but as
the sections pass through the inner layer (Figure 86f) and through the
chromosome (Figures 86g and 86h) all the kinetochore MT have dis
appeared. The middle layer is always obscured by the more opaque
outer or inner layers, because the sections are thicker than the
middle layer. Even a very thin section would have to pass just
between inner and outer layers of a perfectly flat kinetochore in
order to show middle layer only, a very unlikely event.
Figure 87 shows two adjacent sections of two more convex kineto-
chores. The moderately opaque patches in Figure 87b again represent
part of the outer layer in which terminals of MT are visible. The
kinetochore shown in Figure 88 was most likely similar to those in
Figure 83. The section grazed the apex of the inner layer (large
arrow) which seems embedded in the less opaque outer layer. The
latter is clearly set off from the chromosome.
Occasionally, a single MT is found to penetrate the kinetochore
and to extend deeply into the chromosome (Figure 89). Whether such MT
pass completely through the chromosome is not clear.


262
Borisy, G. G., and E. W. Taylor. 196Tb. The mechanism of action of
colchicine. Binding of colchicine-3H to cellular protein. J.
Cell Biol. 34: 525-533-
Brandham, P. E. 1970- Chromosome behaviour in the Aloineae. III.
Correlations between spontaneous chromatid and sub-chromatid
aberrations. Chromosoma 31: 1-17-
Brinhley, B. R., and J. Cartinright, Jr. 1970. Organization of micro
tubules in the mitotic spindle: differential effects of cold
shock on microtubule stability. J. Cell Biol. 47: 25a.
Brinkley, B. R., and R. M. Humphrey. 1969* Evidence for subchromatid
organization in marsupial chromosomes. I. Light and electron
microscopy of X-ray-induced side-arm bridges. J. Cell Biol.
42: 827-836.
Brinkley, R. R., P. Murphy, and L. C. Richardson. 1967. Procedure
for embedding in situ selected cells cultured in vitro. J.
Cell Biol. 35: 279-283.
Brinkley, B. R., and R. B. Uicklas. 1968. Ultrastructure of the
meiotic spindle of grasshopper spermatocytes after chromosome
micromanipulation. J. Cell Biol. 39: l6a-17a.
Brinkley, B. R., and M. W. Shaw. 1970. Ultrastructural aspects of
chromosome damage, p. 313-3I5. In Genetic Concepts and Neoplasia.
Williams and Wilkins, Baltimore.
Brinkley, B. R., and E. Stubblefield. 1966. The fine structure of
the kinetochore of a mammalian cell in vitro. Chromosoma 19:
28-43.
Brinkley, B. R., and E. Stubblefield. 1970. Ultrastructure and inter
action of the kinetochore and centriole in mitosis and meiosis,
P- 119-185- In D. M. Prescott, L. Goldstein, and E. McConkey
(eds.) Advances in Cell Biology, vol. 1. Appleton-Century-
Crofts, New York.
Brinkley, B. R., E. Stubblefield, and T. C. Hsu. 1967. The effects
of colcemid inhibition and reversal on the fine structure of the
mitotic apparatus of Chinese hamster cells in vitro. J. Ultrstruct.
Res. 19: 1-18.
Britten, R. J., and D. E. Kohne. 1968. Repeated sequences in DHA.
Science 161: 529-540.
Buck, R. C. 196( Mitosis and meiosis in Rhodnius prolixus: The
fine structure of the spindle and diffuse kinetochore. J.
Ultrastruct. Res. 18: 189-501.
Burgos, M. H., and D. W. Fawcett. 1956. An electron microscope study
of spermatid differentiation in the toad, Bufo arenarum Hensel.
J. Biophys. Biochem. Cytl. 2: 223-240.


85
Many cells treated with 0.05 ig/ml SN contained chromosomes less
condensed than normal (Figure 111). Chromosomal fibrils possibly
cemented together by an amorphous substance formed more electron dense
patches either at the surface of, or within, these chromosomes (Figures
111b and 111c).
Dicentric bridges very often contained strands of densely packed
or finer-than-norraal fibers that extended, at least in a few carefully
studied cases, from one kinetochore to the other (Figures 110 and 111a).
The kinetochores in cells with dicentric chromosomes were structurally
normal, but their distance from the respective pole was more variable
than in normal ana- and telophase cells (Figures 110a and 113). The
kinetochores of dicentrics, in particular, seemed to lag.
Serial sections revealed the truely acentric nature of fragments
irrespective of their position within a cell.
Nuclear envelope reconstruction begins near the poles, along the
sides of the chromosomes. Small cistemae, already with nuclear pore
complexes, are closely apposed to the chromosomes (Figure 112). It
appears there is a gradient for this process from the polar regions to
the interzone. For example, membrane cistemae may be found along the
poleward portions of dicentrics and laggards, but not along the inter
zonal portions.
Late telophase cells with nuclear bridges through the midbody are
more frequent in SN-treated than in control cells. An example is shown
in Figure 116. The nuclei of the daughter cells were polymorphic.
Pockets extended deep into the nuclei on the poleward side. Very
electron dense patches within the nuclei, or on the inner membrane of
the nuclear envelope lining the pockets, are remnants of the inner


Pig. 30a-, b. Late prometaphase, (a) Equatorial chromosomes
with kinetochores (arrowheads), x 11,500. (b) Relatively thick
section with trar.s chromosomal MT (arrows), x 30,000.
Fig. 31. Late prometaphase. Equatorial chromosomes with
kinetochores (arrowheads). Note chromatin strand connecting
stretched kinetochore regions (arrow), x 15,750.


Fig. 39a-e. Late prometaphase. Klnetochores of a chromosome
displaced towards pole no. 2. Kq_, K2, the pole no. 1 and pole no. 2
kinetochores, respectively. Arrowheads indicate bands. Note bundle
of straight, some skew and wavy MT; passing the kinetochores.
x it0,000.


141
Fig. hOa-c. Late prometaphase. Two chromosomes (Ch[,, Chi-) of the
cell shown in Fig. 28 displaced towards pole no. 1. K1? the pole
no. 1 and pole no. 2 kinetochores, respectively, x 30,000.


Fig. 119a, b. Streptonigrin-treated cell in late cytokinesis,
(a) Drop-shaped nuclear "bleb (near left margin) and fine nuclear
bridge (arrows). Note midbody (MB), MT. x 7,750 (b) Micronuclei
(MN) with intact NE (arrows). Note main nucleus (N). x *10,000.
Both from the same cell.


7
have claimed that nine satellites form a symmetrical crown around the
Centrioles (Bessis et al. 1958, Gachet and Thiry 1964).
Chemistry
It is obviously a difficult task to determine the elemental or
molecular composition of centrioles. Cytochemical methods generally
lack sufficient resolution and bulk isolation of reasonably purified
centrioles seems virtually impossible. The important information
concerning the chemical composition of centrioles is therefore derived
from studies on basal bodies (kinetosomes) of cilia, and is based on
the assumption that the close structural and developmental relationship
between kinetosomes and centrioles (see de Harven 1968 for references)
justifies extrapolation of chemical analyses. Seaman's early findings
(1960) of 27. RNA and 37. DNA in kinetosomes of Tetrahymena did not
remain uncontested (see de Harven 1968 for a detailed discussion). At
best, it seems, the possibility that kinetosomes contain ENA and/or DNA
cannot be ruled out, but the presence of these important macromolecules
in centrioles remains hypothetical.
Duplication
As early as 1952 Inou£ concluded from experiments with colchicine-
treated Chactopterus eggs that centrioles duplicate and mature in the
absence of a mitotic spindle. In a now classical experiment Mazia
et al. (1960) determined the time sequence and mode of duplication of
mitotic centers in echinoderm embryos. Although the methods used by
the investigators did not allow direct visualization of centrioles
(hence the term "mitotic centers"), the results predicted what has
since been confirmed by electron microscopy. The results can be sum
marized as follows: (1) At all times the center is at least a


30
the kinetochore as consisting of two lampbrush-like structures, each
O', 3 o' o o .o
made up of two closely associated 50-80 A axial filaments from which
numerous 50-80 A fibrils loop out laterally. The axial filaments
extend along the surface of the chromosome, their ends being inserted
into the latter. Microtubules attach to the axial filaments in sheets
or bundles. According to their most recent report (Brinkley and
Stubblefield 1970) there is little change in the structure of the
kinetochores in Chinese hamster and rat kangaroo cells from prophase to
metaphase. The kinetochores are mature in late prophase or in pro
metaphase, regardless of whether MT are attached or not.
An electron micrograph recently published by McIntosh and Landis
(1971, Figure 4) supports Jokelainen's model (1967). In this para-
equatorial section of the metaphase plate three kinetochores are shown
as circular patches of less dense material. On the other hand, the
kinetochores of colcemid-treated cells, on which Brinkley and
Stubblefield (1966, 1970) mainly based their model, may be atypical.
For example, very clear images of a double-banded outer layer embedded
in a less dense matrix are produced by the alkaloids vinblastine and
vincristine, both of which disorganize MT (George et al. 1965, Journey
et al. 1968, Journey and Whaley 1970, Krishan 1968).
Microtubules and kinetochores as seen in thin sections are not
preserved in whole-mount preparations of metaphase chromosomes (e.g.,
DuPraw 1968). Instead, chromosomal fibers can be seen to cross between
sister chromatids in the centromere region (see also Abuelo and Moore
1969). Interesting, but unexplained, is the presence of four dense
granules at the centromere of Chinese hamster cells subjected to
certain treatments during isolation (Stubblefield and Wray 1971).


Fig. llla-c. Streptonigrin-treated cell in late anaphase.
(a) Kinetochore (k) and strand (arrows) of a dicentric chromosome,
x 22,500. (h) Telomeres of two long daughter chromosomes (Chp,
CI12). Note patches of more densely packed fibers (arrows),
x 22,500* (c) Loosely coiled chromosome with patches of more
densely packed fibers (arrows), x 40,000. All from the cell
in Fig. 108c.


Abstract of Dissertation Presented to the Graduate Council of
the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
NORMAL AND ABNORMAL MITOSIS IN A MAMMALIAN CELL IN VITRO.
A LIGHT AND ELECTRON MICROSCOPIC STUDY.
By
Urs-Peter Roos
December, 1971
Chairman: Dr. Fred C. Johnson
Co-Chairman: Dr. Henry C. Aldrich
Major Department: Zoology
Rat kangaroo cells (PtK^ line), grown as monolayers, were fixed
and embedded in situ. Cells in mitosis were examined and photographed
under phase contrast. Serial sections were examined with the electron
microscope.
Centrioles duplicate at the onset of prophase. Centriole migra
tion, disintegration of the nucleolus, chromosome condensation, break
down of the nuclear envelope (NE), and formation of the mitotic spindle
are similar to these processes in other mammalian cells. Prophase
kinetochores (Ks) appear as globular, fibrillar bodies in the primary
constriction of chromosomes. Detailed observations on prometaphase
chromosomes support the pulling theory of amphiorientation and metaphase
stability. Amphiorientation is established by unipolar followed by
bipolar attachment to spindle microtubules (MT), or by simultaneous
bipolar attachment, depending on the position of a chromosome relative
to the poles at the time of attachment. In the former case, bipolar
attachment is presumably followed by congression. Concomitantly, Ks
mature to trilaminar, flat, undulated, concave, or convex discs, 4,000-
6,000 A in diameter. All the Ks reach maturity at metaphase. Inner and
outer layers are 400 A, and the middle layer is 300 A thick. On the
xv


Fig. 22 (contd). (c) and (d) two serial sections of a
chromosome near the intact nuclear envelope (hS). Iiote Ml
outside the HE, opposite one kinetochore (&>); difference be
tween the two kinetochores. x 57,500.


PREFACE
Certainly, one cannot argue about the importance and significance
of attempts to increase our knowledge of chromosomes and mitosis, both
so basic to life in higher organisms. I was fortunate to have the
liberty to choose a research project myself, and the study presented
evolved mainly out of my curiosity to learn more about these two
fundamental aspects of life.
At the time this project began, ultrastructural investigations of
eukaryotic chromosomes had essentially reached a standstill. More new
questions had been posed than old ones answered. On the other hand,
electron microscopic techniques had not yet been applied to the study
of chromosomal aberrations. Mitosis in animal cells was fairly well
documented at the ultrastructural level, but many of its aspects were
the subject of controversy. Like many other students at the beginning
of a new road, 1 had only a limited knowledge of these problems; hence
my idealistic assumption that an ultrastructural investigation of
normal and abnormal mitosis could answer many of the questions
remaining, or at least lead the way to significant experiments. It
will become clear from the following presentation how far these hopes
were fulfilled and justified.
iv


118


132


153
Fig. 52. Very late anaphase. Chromosomal mass near pole no. 1
(P-^). Note fuzzy kinetochore (arrowhead), membrane cisternae (large
arrows), chromosomal granules (small arrows), nucleolus organizer (NO),
and MT traversing the chromosomal mass, x 22,500. Inset: Phase
contrast micrograph of the cell in plastic (x 1,000).


142
Fig. 4l. Late prometaphase. Obstructed kinetochore (k) of an
equatorial chromosome (Ch). Note double (large arrows) and beaded
(small arrows) substructure of the moderately dense band, x 100,000.


I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
F. C. Jo:
Associa
Professor of Zoology
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
JU. C. dtAusL.
H. C. Aldrich, Co-Chairman
Assistant Professor of Botany
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
'yj
fL-i-
L. H. Larkin
Assistant Professor of Anatomical
Sciences
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.


263
Busch, H., and K. Smetana. 1970. The Nucleolus. Academic Press,
New York. 626 p.
Byers, B., and D. H. Abramson. 1968. Cytokinesis in HeLa: post
telophase delay and microtubule-associated motility. Protoplasma
66: 413-435
Carlson, J. G. 1938. Mitotic behavior of induced chromosomal fragments
lacking spindle attachments in the neuroblasts of the grasshopper.
Proc. Nat. Acad. Sci. 2t: 5OO-507.
Carlson, J. G. 1954. Immediate effects on division, morphology, and
viability of the cell, p. 763-824. In A. Hollaender fed.) Radiation
Biology, vol. 1 (pt. 2). McGraw-Hill, New York.
Carter, S. K., D. Rail, P. Schein, R. D. Davis, H. B. Wood, Jr., R.
Engle, J. P. Davignon, J. M. Venditti, S. A. Schepartz, B. R.
Murray, and C. G. Zubrod. 1968. Streptonigrin. NSC 45383
Clinical Brochure. Nat. Cancer Inst. Chemother. 20 p.
Casarett, A. P. 1968. Radiation Biology. Prentice-Hall, Englewood
Cliffs. 368 p.
Chang, J. P., and C. W. Gibley, Jr. 1968. Ultrastructure of tumor
cells during mitosis. Cancer Res. 28: 521-534.
Cohen, M. M. 1983. The specific effects of streptonigrin activity
on human chromosomes in culture. Cytogenetics 2: 271-279
Cohen, M. M., M. W. Shaw, and A. P. Craig. 1983 The effects of
streptonigrin on cultured human leukocytes. Proc. Nat. Acad.
Sci. 50: 16-24.
Comings, D. E. 1968. The rationale for an ordered arrangement of
chromatin in the interphase nucleus. Amer. J. Human Genet.
20: 440-460.
Comings, D. E. 1971 Isolabelling and chromosome strandedness.
Nature New Biol. 229: 24-25.
Comings, D. E., and T. Kakefuda. 1988. Initiation of deoxyribonucleic
acid replication at the nuclear membrane in human cells. J. Mol.
Biol. 33: 225-229.
Comings, D. E., and T. A. Okada. 1970a. Association of nuclear
membrane fragments with metaphase and anaphase chromosomes as
observed by whole mount electron microscopy. Exp. Cell Res.
63: 62-68.
Comings, D. E., and T. A. Okada. 1970b. Association of chromatin
fibers with the annuli of the nuclear membrane. Exp. Cell Res.
62: 293-302.
Comings, D. H., and T. A. Okada. 1970c. Condensation of chromosomes onto
the nuclear membrane during prophase. Exp. Cell Res. 63: 471-473.


104


60
The chromosomes condense further until they appear as solid, very
electron dense rods or blocks (compare Figure 26 with Figure 29).
Sister chromatids are still tightly joined, but their individuality is
apparent in transverse or peripheral longitudinal sections of
chromosomes (Figures 27, 29, 30a, 31, and 33), or in situations where
the sister kinetochores are stretched (Figure 34). In the latter case,
one or several achromatic holes appear in the kinetochore region.
The diversity of kinetochore structure is still remarkable during
mid- and late prometaphase (Figures 26-29, 30a, 31, 33-41). At a glance
there seems to be no general pattern of differentiation, but a careful
comparison of the structure of sister kinetochores with the position of
the chromosomes in the spindle gives a better insight. First, we can
distinguish between chromosomes lying on or near the equator of the
spindle (Ch^ and Ch^ in Figure 27; Ch^ in Figure 28; the chromosomes in
Figures 30a and 33-36), and chromosomes lying closer to one pole
(malpositioned chromosomes; Figure 26; Ch^ in Figure 27; Ch^ in Figure
28; Figures 29 and 37-40). Generally speaking, more chromosomes occupy
an equatorial position in late than in early prometaphase (Figures 20,
26, and 29). The sister kinetochores of equatorial chromosomes are
similar if both are unobstructed, i.e., if no major obstacle lies along
a line from the kinetochore to the pole (Figures 27, 30a, 33, 34, and
36); they are dissimilar if one is unobstructed, the other obstructed
by a neighbor chromosome (Figures 33 and 35). Unobstructed kinetochores
are irregularly shaped and fuzzy (Figures 28 and 36), fuzzy cones with
a trace of bands (similar to in Figure 38), or distinctly triple-
banded (Figures 27 and 33-35). It appears that the clarity of the
three bands increases with advancing prometaphase. Unobstructed


48
1 hr exposure the medium was replaced by cold glutaraldehyde of the
usual strength and buffered as usual (see fixation schedule). Initial
fixation was 15 min in the cold, after which the flask was brought to
room temperature and the standard fixation schedule was followed from
here. One control consisted of cells kept at 37C and fixed cold as
above. The second control consisted of cells not exposed to cold and
fixed at room temperature.
Fixation and Embedding
The standard fixation schedule, modified from Brinkley et al.
(1967), was as follows:
(1) Decant the culture medium and add 3.17. glutaraldehyde (GA)
buffered with Millonig's phosphate buffer without sucrose
(Millonig 1961), pH 7.3. After a few minutes add a fresh
change of GA. The cells are fixed for a total of 1 hr at
room temperature.
(2) Rinse with two changes of buffer, 10 min each, at room
temperature.
(3) Postfix for 1 hr at room temperature in similarly buffered 27.
osmium tetroxide (OsO^).
(4) Dehydrate in 257. and 507. ethanol, 10 min each. Prestain in
cold 27. uranyl acetate in 707. ethanol for 2 hr to overnight.
Rinse with two changes of cold 757. ethanol, 10 min each. Ten
min 90% ethanol; the cold solution is added and the flasks
then brought to room temperature. Subsequent steps are all
carried out at room temperature; three changes, 10 min each,
of 907. hydroxypropyl methacrylate (HPMA); 15 min each of 957.
and 977. HPMA


ACKNOWLEDGMENTS
I am grateful to Dr. Fred C. Johnson, Chairman of my Supervisory
Committee, for his help in procedural matters.
Words cannot express my respect and gratitude for the liberal and
stimulating guidance provided by Dr. Ienry C. Aldrich, Co-chairman,
throughout this study. His generosity regarding time and use of
facilities was an important factor for the progress of my work.
I am much indebted to Dr. John U. Cramer and his staff, Department
of Pharmacology, for instruction and use of facilities for tissue
culture work.
Dr. Lynn H. Larkin and Dr. James L. Nation, members of my
Supervisory Committee, deserve my thanks for their moral support and
helpful advice. I appreciate the encouragement provided by Dr. John W.
Brookbank, Dr. James H. Gregg, and Dr. Philip B. Morgan.
I am much obliged to Dr. Robert M. De Witt, Chairman of the
Department of Zoology, for arranging financial assistance.
I also extend my thanks to Mrs. Rosemary Rumbaugh, and numerous
unnamed persons with whom I was associated in the course of this study,
for assistance and forbearance.
My wife, Loan, played a major part in this accomplishment. I am
deeply grateful for her never faltering patience, her understanding,
encouragement, and help.
iii


Fig. 2ia-c. Early pronetaphase. Three serial sections of
chromosome no. 1 shown in the inset of Fig. 20. Note stretching,
attached MT, of the kinetochore facing the near pole (K2); globular
shape, lack of MT, of the kinetochore facing the far pole (K^).
x 30,000.


Fig. 38a, b. Late prometaphase. Higher magnification of
chromosome no. 1 shown in Fig. 23. Note single band of Kg
(arrowhead in Fig. 38a), faint triple-banding of Kp (arrowhead
in Fig. 38b). Arrows point to dark chromosomal granules,
x 30,000.


29
c
the outer layer and a reduction in size of the kinetochoral patch as it
becomes anchored to the spindle MT (Jokelainen 1965b). The outer layer
is embedded in a moderately dense substance, part of which persists at
metaphase as the middle layer of the kinetochore and the so-called
corona (Jokelainen 1967). Jokelainen's model (1967) depicts the
kinetochore as a trilaminar disk, 2,000-2,500 A in diameter, at the
surface of the chromosome, sometimes slightly recessed, sometimes
projecting. The outer layer is 300-450 A thick, the middle layer
150-300 A, and the inner layer 150-250 A. The corona over the outer
layer measures approximately 300 A in thickness. Evidence for the
disk-like appearance came from para-equatorial sections showing
kinetochores in face view. The outer disk, which is finely granular or
fibrillar, stains consistently, while the density of the chromosomes
varies with the staining method employed. The inner layer is highly
electron-dense, and roughly granular or fibrillar. Four to seven MT
are attached to each kinetochore, apparently penetrating the outer and
middle layers and sometimes ending in the chromosome.
Brinkley and Stubblefield (1966, 1970) presented a different model
of mammalian kinetochores. In colcemid-arrested Chinese hamster cells
the kinetochore appears as a 200-300 A wide band, made up of two
50-80 A threads. This band is embedded in a less dense matrix and
follows the curvature of the chromosome surface at a distance of
approximately 100 A. In the less dense matrix the authors detected
50-80 A fibrils apparently looping out from the main band. This
description was based on kinetochores without attached MT, but
according to the authors the kinetochores of untreated cells are very
similar. The model of Brinkley and Stubblefield (1966, 1970) describes


257
Nuclear Envelope
Fragmentation and reconstruction of the nuclear envelope (ME) in
PtK^ cells are similar in many ways to the processes in other mammalian
cells. However, 1 have neither observed stacking of ME fragments as
reported for rat lymphocytes (Murray et al. 1965) and rat hepatoma
cells (Chang and Gibley 1968), nor "polar caps" consisting of numerous
cistemae and vesicles as in HeLa cells in metaphase (Robbins and
Gonatas 1964). There are no membrane elements at metaphase and early
anaphase that could be identified as fragments of the NE (Figures 42,
43 45, and 47). Cistemae possibly representing elements of the
reforming NE occur only in late anaphase at the periphery of the
spindle (Figure 51a). In very late anaphase and early telophase such
cistemae can be clearly identified as portions of the NE mainly
because of their association with the chromosomal mass (Figures 52 and
57).
Comings and Okada (1970a) have described associations of NE
fragments with meta- and anaphase chromosomes as seen in whole mounts.
Their observations can be criticized for two reasons: (1) Most of the
preparations were made from colccmid-arrested cells, and (2) anaphase
was simply judged according to time after colcemid reversal. According
to the authors, however, the results agreed with Comings' hypothesis
(1968; also Comings and Kakefuda 1968, Comings and Okada 1970b, c) that
persisting chromosomal attachments to the NE account for reproducible
patterns of DMA replication and chromosome folding. This is a
tempting hypothesis, but evidence is rather scarce and unconvincing.
In PtK^ cells the reconstruction of the NE seems to be related
spatially to the chromosomal mass as a whole, rather than to specific
sites of individual chromosomes (Figures 52 and 57).


Fig. 87a, b. Two serial sections of two metaphase kinetochores
(para-equatorial). Filled circles nark bypassing MT. (a) Open
circles nark kinetochore MT. MT marked by arrows are the sane
marked in (b). (b) Arrows nark terminals of MT in outer layer.
From the sane cell as Fig. 87. x 50,000.
Fig. 88. Metaphase kinetochore. Note electron dense inner
layer (large arrow) within less dense outer layper; intrachronosomal
MT and granule within clear space (circle); other chronosomal
granules (small arrows). From the sane cell as Fig. 87.
x 50,000.


14
(see Luykx 1970, Nicklas 1971, for detailed discussions). Inou and
his collaborators (Inou 1964, Inou and Sato 1967) have proposed that
MT polymerize from a pool of subunits in a dissociation-association
equilibrium, probably under control of "organizing centers" such as
kinetochores and centrioles. Certainly, the rapid loss and
reappearance of birefringence in mitotic cells subjected to rapid
temperature shifts (Inou et al. 1970) suggests self assembly of
subunits to form MT (see Nicklas 1971).
Fine interconnections or "cross-bridges" between spindle MT, and
"arms" on single MT have been reported for a number of plant and
animal cells (Hepler and Jackson 1968, Krishan and Buck 1965, Wilson
1969). Hepler et al. (1970) devoted a more detailed and systematic
study to these structures in cultured human cells and Haemanthus
endosperm, but although they play a major role in the model of mitosis
proposed by McIntosh et al. (1969), their reality is conjectural at
best (see also Nicklas 1971).
Distribution and Classification
Spindle MT can be divided into two main categories: chromosomal
and continuous (interpolar) MT (de Harven 1968, Luykx 1970, Nicklas
1971, Roth 1964). In the case of chromosomes with a well-defined
kinetochore the chromosomal MT run between the latter and the
corresponding spindle pole (Brinkley and Stubblefield 1966, 1970;
Harris and Mazia 1962, Jokelainen 1967, Krishan and Buck 1965, Murray
et al 1965, Robbins and Gonatas 1964). Chromosomes with so-called
diffuse kinetochores are connected with MT at various points along
their length (Buck 1967). Luykx (1970) distinguished several


Fig. 95a-c. Kinetochores in c-mitosis.
(a, b, adjacent) of sister kinetochores (K^,
profiles. From the same cell as Fig. 9b. x
Three serial sections
K2) with strange
T5,000. Treatment A.


Figure Page
o'. 3 o c. o
37a,b Late prometaphase chromosome 136
36a,b Late prometaphase chromosome 138
39a-e Late prometaphase chromosome 140
40a-c Late prometaphase chromosomes 141
41 Late prometaphase kinetochore 142
42 lletaphase plate 143
43 lletaphase chromosomes (para-sagittal
section ) 144
44 lletaphase plate (para-equatorial section) 145
45 Very early anaphase cell 146
46 Very early anaphase chromosomes 147
47 Early anaphase cell 148
48 Very early anaphase chromosomes ....... 150
49 Early anaphase kinetochores 150
50 Mid-anaphase cell (para-equatorial
section) 151
51a,b Late anaphase spindle and chromosomes .... 152
52 Very late anaphase, daughter nucleus 153
53 Very late anaphase, kinetochore ....... 155
54 Very late anaphase, stem bodies 155
55a Very late anaphase, stem body 155
55b Very late anaphase, equatorial region .... 157
56 Early anaphase centrioles ..... 157
57 Early telophase nucleus 158
58 Early telophase nucleus 160
59 liid-telophase kinetochores 160
60 liid-telophase nucleus 161
x


184


Fig. 23a-c. Early prometaphase. Kinetochores of chromosomes
near pole no. 2 of the cell in Fig. 19. (a) Note difference be
tween kinetochores facing the near pole (K, K2), and kinetochore
facing the far pole (Kj_). Arrows indicate bands, x 30,000. (b)
and (c) two serial sections of another chromosome (Ch). Note
differences, similarities between the kinetochore facing the near
pole (K2) and the kinetochore facing the far pole (Kq_). Arrows
indicate bands, x 30*000


Fig. 36. Mid-prometaphase. An equatorial chromosome with
stretched, fuzzy kinetochores (Ki, K2). Note strand of chromosomal
fibrils connecting the two kinetochore regions (arrows). x 30,000.
Fig. 37a, b. Late prometaphase. Two serial sections of the
small chromosome displaced towards pole no. 1 in Fig. 29. Note
fuzzy appearance of Kj_, compact Ko with single band (arrow in
Fig. 3Tb). x 50,000.


Fig. 83a, b. Sister IdLnetochores in very early anaphase.
Note obliquely cut kinetochore MT (some marked by arrows), shape
of the kinetochores. From the cell in Fig. 45. x 50,000.
Fig. 84. Early anaphase kinetochore. Note corona (Co),
outer (K0), middle (KM), and inner (KE) layers, lesser density of
KD compared to ICE. From the cell in Fig. 47. x 122,500.


Fig. TO. Cold-treated cell in mid-prophase. Note granular
component of the dissolving nucleolus (Nu) and its associated
chromosome (Ch), a kinetochore (arrowhead), clusters of intranuclear
granules (arrows), and chromosomal granules (circles), x 11,500.
Inset: Phase contrast micrograph of the cell in plastic. Note
the intact NE (x 1,280).
Fig. Jim Cold-treated cell in mid-prometaphase. Three
kinetochores (arrowheads) and a bundle of MT are visible. Note
achromatic holes in chromosomes, numerous chromosomal granules,
x 11,500. Inset: Phase contract micrograph of the cell in
plastic (x l,28o).


239
relaxed state they blend in with the other fibers. This could explain
the occurrence of similar strands of more or less parallel fibers in
stretched prometaphase chromosomes (Figures 31 and 36). The signi
ficance of chromatid connections for stable metaphase orientation of
chromosomes is clear, assuming that orientation stability depends on a
balance of pulling forces acting on sister kinetochores oriented
11
towards opposite poles. This theory was proposed by Ostergren (1951)
for meiosis, and experiments by Henderson and Koch (1970), Nicklas
(1967), and Nicklas and Koch (1969) with grasshopper spermatocytes
strongly support it. Connections between sister chromatids prevent
premature separation, which would lead to unipolar attachment of
chromatids and, consequently, unstable orientation followed by
haphazard chromosome distribution.
The surprisingly frequent occurrence of very fine nuclear bridges
in SN-treated cells in cytokinesis indicates that not all the dicentric
bridges rupture during anaphase (Figures 118 and 119a). Acentric frag
ments form micronuclei if included in daughter cells (Figure 119b).
In one case examined in the electron microscope, however, I found an
"elimination body," consisting of a polymorphic micronucleus and very
little cytoplasm, in the cleavage furrow between daughter cells. This
body possibly contained acentric fragments that were lying near the
equator at the periphery of the cell in metaphase, as shown in Figures
108c and lOSe.
Osmiophilic granules in chromosomes of PtK^ cells have been re
ported by Brinkley and Shaw (1970). I have found these granules in
interphase and all stages of mitosis. They are more clearly visible,
possibly more numerous, in cold-treated than in control cells (Figure 71).


121
Fig. 25. Early prometaphase. Kinetochores in the equatorial
region. From the cell in Fig. 20. Note unobstructed sister kineto
chores (Kq_, K2) of the X chromosome, its nucleolus organizer (NO),
other kinetochores (k). The direction of the spindle axis is in
dicated by MT marked with large arrows. Other MT are skew (small
arrows), x 30,000.


246
Extending their model to PtK^ rat kangaroo cells, Brinkley and
Stubblefield (1970) interpreted the outer and inner bands of metaphase
kinetochores as representing the two kinetochore filaments. My
observations on paraxial (e.g., Figures 78-85) and para-equatorial
(Figures 86, 87, and 91) serial sections leave no doubt that mature
kinetochores in PtK^ cells are trilaminar, roughly circular plates at
the surface of chromosomes. I have not seen any filaments in para-
equatorial sections. Furthermore, kinetochore MT are arranged in
bundles, not in "sheets," as would be expected were they connected to
filaments.
My view of mature kinetochores essentially agrees with the model
of Jokelainen (1967), except for one important aspect: kinetochore
plates are very seldom, if ever, flat. Rather, they are undulated
(Figures 42 and 82), concave (Figure 85), or convex (Figures 46, 47,
and 84) discs. Figure 121a-c diagrammatically represents three-
dimensional reconstructions of kinetochores of various shapes. The
shape of a kinetochore is certainly determined by the curvature of the
chromosomal surface at this locus. It is also possible that both the
shape of the kinetochore and the curvature of the chromosome are
influenced by the differential pulling action of the attached MT.
Ultrathin paraxial sections (Figures 82 and 84) strongly suggest
that kinetochore MT terminate in the outer layer. Even more convincing
are para-equatorial sections (Figures 86 and 87). Nebel and Coulon
(1962) and Luykx (1965a) concluded from their micrographs that kineto
chore MT terminate in the inner layer. These observations, however,
were severely hampered by inadequate preservation of fine structure.
Jokelainen (1967) also concluded that kinetochore MT are anchored in


253
of course, are also designed to explain raeiosis, picture the chromo
somes as orienting within the more or less completely formed central
spindle (Dietz 1969, Luykx 1970, McIntosh et al. 1969). The MT
growing out from the kinetochores are supposed to interact with the
already present "continuous" MT of the central spindle. Whether this
is indeed so in the cells considered by the authors remains to be
proven, because no ultrastructural studies of prometaphase have been
presented. In PtK^ cells the above is certainly not the case, as
demonstrated by Figure 23, and similar micrographs not included.
These pictures clearly show that MT do not grow out from, or attach to,
both kinetochores simultaneously, and that chromosomes may connect to
a pole as soon as the nuclear envelope breaks down, regardless of
whether the central spindle is fully formed or not.
stergren (1951) postulated a pulling theory of chromosome
orientation and congression for meiosis. Briefly, this theory
proposes that bipolar orientation of bivalents is the consequence of
kinetochore polarity. This polarity allox^s each kinetochore to connect
only to that pole towards which it is oriented. Prometaphase movements
are caused by random activity of kinetochores. Ultimately, the
kinetochores exert a pull and the equatorial position of chromosomes at
late prometaphase and metaphase is the result of balanced tension on
kinetochores oriented towards opposite poles.
The following reconstruction of events leading to chromosome
orientation and congression in PtK^ cells fits this theory well, if it
is modified to account for differences between meiotic chromosomes
(bivalents) and mitotic chromosomes (pairs of sister chromatids). As
soon as the nuclear envelope breaks down near the centrioles, the


58
resembles typical prophase kinetochores (e.g., in Figure 24), or it
consists of a dense band embedded in the fibrillar matrix (K^ in
Figures 23a and 23c). The band can be seen in several adjacent sec
tions, which suggests it represents a transverse section of a flat or
convex plate approximately 300 A thick and 3,500 A in diameter. In
each case few or no MT are associated with such a kinetochore.
Sister kinetochores of chromosomes lying near the future equator
of the spindle are similar if both are unobstructed, i.e., if there is
no nearby obstacle (such as another chromosome) between them and the
respective pole (Figure 25). Usually, these kinetochores are more or
less stretched, with or without bands, and attached to MT. If one of
the sister kinetochores is obstructed by a neighbor chromosome lying
very close, it resembles prophase kinetochores (not shorn, but similar
to in Figure 24a), while the other kinetochore resembles the
unobstructed kinetochores described above.
Microtubules are found in the "nucleus" as soon as the nuclear
envelope breaks down (Figures 18, 22, and 23). At first they are more
numerous in the vicinity of the centrioles (Figure 18), but as the lat
ter take up their position at opposite poles, MT are abundant in the
center of the spindle (Figure 20). Surprisingly, most of the MT are
associated with kinetochores and few, if any, continuous tubules seem
to be present. However, a more careful, quantitative analysis would be
necessary to establish this with certainty.
Remnants of the nucleolus disappear completely and the nucleolar
organizer appears on the X chromosome (Figures 19 and 25).
Figures 26-30 illustrate further progression of prometaphase.
Fragments of the nuclear envelope, still found between and around some


Figure Page
84 Early anaphase kinetochore 184
85a-c Metaphase kinetochore (para-sagittal
sections) 186
86a-h Metaphase kinetochore (para-equatorial
sections) ... ..... 188
87a,b TVo metaphase kinetochores (para-equatorial
sections) 190
88 Metaphase kinetochore (para-equatorial
sections) 190
89 Metaphase chromosome (para-equatorial
section) 192
90 Metaphase cell, para-equatorial section of
kinetochore MT 192
91a,b Mid-anaphase kinetochore (para-equatorial
sections) 192
92 Kinetochores in c-mitosis 194
93 Kinetochores in c-mitosis 194
94 Kinetochores in c-mitosis 194
95a-c Kinetochores in c-mitosis 196
96 Kinetochores in c-mitosis 198
97 Kinetochore in c-mitosis 198
98 Kinetochores in c-mitosis 198
99 Untreated anaphase cell, dicentric
bridge 199
100 Untreated anaphase cell, dicentric
bridge 201
101 Untreated anaphase cell, dicentric
bridge 201
102a,b Untreated anaphase cell, kinetochores of
lagging chromosomes 203
103a,b Untreated anaphase cell, kinetochore of non
lagging chromosome 205
xii


188
JE*-
A"* ni *
r* > f
9 &
vC> **


224
'im


CONCLUSIONS
Knowing the question is knowing half the answer. Applied to
science this means if we are able to state a problem and to formulate a
hypothesis, we are able to design experiments with the aim to gather
evidence for or against the hypothesis, or to reach new ideas. Instead
of collecting data more or less at random we could then proceed in an
organized fashion to solve a particular problem.
From the preceding chapters it is clear that not all my expecta
tions stated in the preface were fulfilled. The study of chromosomal
ultrastructure took second place to the investigation of mitosis.
Biophysical and biochemical studies are more likely to provide a better
understanding of such interesting and important problems as polynemy
or uninemy and the orderly coiling of chromosomes, than ultrastructural
studies alone.
My observations on prometaphase and early anaphase of mitosis,
previously neglected by most investigators, have revealed interesting
new data. But again, electron microscopy alone, although a powerful
tool, has its limits. I am firmly convinced, however, that a concerted
and organized effort combining _in vivo observation, particularly
polarization microscopy and cinenicrography, with electron microscopy
will clarify many of the still mysterious aspects of mitosis.
259


240
Their significance is not at all clear. In size they are distinctly
different (approximately 300-500 A in diameter) from ribosomes and the
granular component of the nucleolus.
Kinctochores
Structural Changes During liitosis
The terra "maturation" suggested by Jokelainen (1968) best de
scribes the structural differentiation of kinetochores from prometa-
to metaphase. 3y definition a mature kinetochore appears in thin
sections as the typical triple-banded structure at the surface of
metaphase chromosomes.
Kinetochores appear "out of nowhere" at the primary constriction
of chromosomes (Figures 10, 11a, 15-17). I could not detect any
relationship between primary constrictions and HE, which nevertheless
does not rule out the possibility that other points of attachment
determine the coiling pattern of chromosomes (see Comings 1968, Comings
and Okada 1970a, b). In stages of prophase earlier than illustrated
in Figure 10 kinetochores are not visible as such, but their future
position is indicated by a slight constriction on the loosely coiled
chromosomes. From mid- to late prophase the electron density of
kinetochores increases slightly, making their detection easier. In any
event, however, the chromosomes always stain much more intensely than
the kinetochores (e.g., Figure 16).
Radical changes in kinetochore structure occur during prometa
phase. These changes first involve the kinetochores of chromosomes
near the centrioles, i.e., in the region where the NE breaks down
first (Figure 23). In cells' such as the one shown in Figure 19 one


Fig. 67. Colcemid-treated interphase cell. Nucleus with
dispersed chromatin (Chr), nucleolus (Nu), nuclear envelope (NE).
x 15,750* Treatment B.
Fig. 68. Colcemid-treated interphase cell. Grazing section
of the nuclear envelope (HE) with pore-annulus complexes (NP),
helical polyribosomes (R). Note microtubules (MT), and micro
fibrils (MF). x +0,000. Treatment B.
Fig. 69. C-mitosis. Bundle of microfibrils (MF) in the
central area. Same section as in Fig. 66. x 75,000. Treatment B.


5
Centrioles
0 r J 'w' -
Occurrence
Centrioles occur in most or all animal cells (Brinkley and
Stubblefield 1970), in some fungi (Aldrich 1967, Renaud and Swift
1964), and some algae (Ringo 1967a). They are absent in cells of
higher plants, either before or during division (Pickett-Heaps 1969,
Wilson 1970). The possibility that certain non-dividing, fully
differentiated animal cells may also lack centrioles has been suggested
by Bernhard and de Ilarven (1960; see also de Harven 1968), but the
available evidence is inconclusive.
The apparent de novo origin of centrioles during certain stages of
the life cycle in many lower plants is an intriguing problem. A
detailed discussion, however, is beyond the scope of this survey. The
reader is referred to reviews by de Harven (1968), Luykx (1970), Mazia
(1961), and Pickett-Heaps (1969).
Ultrastructure
There is good agreement in the literature that centrioles of
interphase cells do not differ structurally from centrioles of the
mitotic apparatus (Brinkley and Stubblefield 1970, de Harven 1968,
Stubblefield and Brinkley 1967). A centriole is a cylindrical body
1,500-2,500 A in diameter and 3,000-7,000 A long. Nine triplets of
fused 240 A MT form its wall. Centriolar MT possibly have 13 subunits
in cross section, as do flagellar MT (Ringo 1967b). The triplets are
not arranged radially, but slanted, giving the image of a pinwheel.
Centrioles are polarized bodies. The so-called distal end is capable
of generating the shaft of a flagellum or ciliura (Gibbons and


9
Stubblefield (1968), using a technique that renders centrioles in
fixed cells visible for light microscopy, studied centriole duplication
and behavior in colceraid-treated Chinese hamster cells. His findings
concerning duplication, maturation, and separation agree well with the
above-mentioned inhibitor studies (Mazia et al. 1960, Went 1966).
Further confirmation came from an ultrastructural study on colcemid-
treated cells (Brinkley et al. 1967).
From the references and reviews mentioned the following picture of
the centriole cycle emerges: (1) At some stage between divisions each
of the two centrioles normally present in interphase cells produces a
daughter centriole by an as yet unknown mechanism. (2) The daughter
centrioles undergo maturation, which involves elongation and, possibly,
formation of the intracentriolar structures. If wTe define "mature" as
being capable of generating a procentriole and to participate in
spindle formation (see the following paragraph for reservations about
the latter), then the timing relative to the cell cycle under normal
conditions is so that procentrioles are mature no sooner than the end
of karyokinesis following duplication. (3) If spindle formation is
inhibited by colcemid or ME, maturation of the procentrioles continues
and they may act as mitotic centers after release from the block. As a
consequence, multipolar spindles occur more frequently than under
normal conditions, and the proportion of these increases with increasing
time of exposure to the inhibitor.
Function
Only the role played by centrioles in animal mitosis will be
discussed here. For the involvement of centrioles in the generation of


Page
0-. a o W O
C-Mitosis 67
Mitosis in Cold-Treated Cells 70
Kinetochore Fine Structure 71
Chromosomal and Mitotic Aberrations 77
Discussion 225
Centrioles 225
Microtubules 230
Chromosomes 235
Kinetochores 240
Nuclear Envelope 257
Conclusions 259
References 260
Biographical Sketch 276
vi


74
81. The straight MT marked with an arrow terminates in the outer
layer, while the obliquely sectioned MT marked with an arrowhead seems
to penetrate into the inner layer. Skew MT passing in front of a
kinetochore occur occasionally (Figure 79).
One metaphase cell was sectioned in a para-equatorial plane from
one pole across the metaphase plate into the opposite half-spindle.
Of the 24 kinetochores examined, all were apparently cut head-on
(e.g., Figures 86 and 87), except one, which was sectioned at a
slightly oblique angle. This kinetochore belonged to a chromosome at
the periphery of the metaphase plate. To determine which of the MT of
a chromosomal bundle were kinetochore MT, serial micrographs of eight
different chromosomes and associated MT, at final magnifications
between 40,000 and 62,500, were analyzed. In the last section of the
series, MT in the vicinity of the chromosome were marked with one
color. Proceeding poleward, newly emerging MT at the kinetochore were
marked with a different color. The average number of kinetochore MT
was 26, the range 16-40. This agrees well with estimates from paraxial
serial sections. I was unable to ascertain if the number of kineto
chore MT is correlated with chromosome size. The average number of MT
bypassing the kinetochore was 5 (range 0-9). It is rather arbitrary
to choose these bypassing MT from the population of non-kinetochore
tubules, the only criterion being their proximity to the chromosome in
question.
The kinetochores in para-equatorial sections are roughly circular
patches of variable electron density, depending on the level of the
section (Figures 86-88, and 91). The diameter varies from approxi
mately 3,400 A for obviously convex kinetochores to 6,000 A for
flatter kinetochores.


237
recovery period, not during exposure to SN. It therefore appears that
high concentrations of SN alter the bonding properties of the molecules
involved in coiling of chromatin.
Interesting is that high concentrations of colcemid produce a
similar effect (Figures 65 and 66). Considering the short duration of
the treatment (15 min) it is more likely that colcemid dispersed
already condensed chromosomes rather than inhibiting the process of
condensation. Morphological alterations of chromatin in mitotic grass
hopper neuroblasts by colchicine was reported by Gaulden et al. (1970).
Colcemid and colchicine are widely used, often at concentrations
similar to those altering chromosome structure, for the accumulation
of metaphase cells destined to yield a great number of chromosomes for
ultrastructural studies (e.g., Abuelo and Moore 1969). The implica
tions are clear: not only do we have to consider artifacts produced
in whole-mounted chromosomes by the preparation techniques, but the
preceding colcemid or colchicine treatment may have completely altered
the ultrastructure of the chromosomes.
Remarkable is the appearance of centromere granules in chromo
somes of cells that were probably in late prophase or in prometaphase
at the time of treatment (Figure 65). I could not definitely determine
whether there are one or two granules per chromatid. The structure
and position of these granules is very similar to those observed by
Stubblefield and Uray (1971) in whole-mounted Chinese hamster chromo
somes. In the latter case, however, the possibility of an artifact
produced by colcemid is slight, because at the concentration used
(0.06 jug/ml for 4 hr) the drug does not even disperse all the MT
(Brinkley et al. 1967).


Fig. 22a-d (continued on the foilwing page). Early pro
metaphase. Kinetochores of the cell in Fig. 18. (a) Note
appearance of kinetochores (large arrows) compared to prophase
(e.g., Fig. 15)* Oblique and cross sections of MT (small arrows),
intrachromosomal MT (circles), x 30,000. (b) Sister kinetochores
(K]_, Kg) with faint band (arrows), x 57*500.


54
Prophase
The centrioles duplicate at the onset of prophase. Therefore, two
parent-daughter centriole pairs are found in early prophase cells
(Figure 8a). Four serial sections of the sane centrioles (Figures
8b-e) reveal that each daughter centriole is closely associated with
its parent at approximately a right angle. Daughter centrioles appear
shorter than nature centrioles (compare Figure 8b with Figure 3), but
have the same diameter (approximately 2,200 A). At this stage rela
tively few MT converge on, or radiate from, each parent centriole. An
early stage of centriole migration is illustrated in Figures 9a and 9b.
One of the two centrioles shown (C^) is possibly a daughter centriole.
It is cut almost perfectly at a right angle to its axis. The structure
and arrangement of the tubular triplets is particularly clear (Figures
9c-e). The cartwheel at the proximal end appears in three adjacent
sections (Figures 9b-d), but hub and spokes are most distinct in
Figure 9c. The triplets seem to be embedded in amorphous osmiophilic
material. Bars, approximately 80 by 480 A, possibly cross sections of
plates, appear between the triplets (Figure 9e). Numerous satellites
are present in the general area of the two centriole pairs (Figures 9a
and 9b). In many, perhaps all, mid-prophase cells the centrioles lie
in an invagination or pocket of the nuclear envelope (e.g., Figure
11a). At this stage numerous ITT are present, forming an "aster"
around the centrioles. The radial arrangement of these MT also imposes
radial orientation on other organelles, notably mitochordria (Figure
lib). Migration of centrioles relative to nuclear changes (chromosome
condensation, fragmentation of the NE) varies considerably, so that by


83
Table 2.Mitotic
treated
indices (Ml) for streptonigrin-
and untreated control cells.
Treatment
Total No. of
Cells Counted
Cells in Mitosis
no. % Iffi
Control
541
14
2.59
0.01 ug/ml SN
540
12
2.22
0.05 ug/ml SN
541
1
0.18


55
the end of prophase the two pairs may have moved a short distance only,
or they may lie at opposite poles (see Figures 18a and 19).
Progressive condensation of chromatin is indicated by the
appearance, during very early prophase, of large heterochromatic
patches (compare Figure 8a with Figure 5). Discrete chromosomes are
present in mid-prophase cells (Figures 10 and 11a). Their "mottled"
appearance (see also Figure 13) is suggestive of incomplete condensa
tion. During condensation many, possibly all the chromosomes are
attached to the nuclear envelope along their entire length or with
their telomeres (Figures 11a and 13). In transverse sections of nuclei
in early prophase the chromosomes appear to be attached by "stalks"
(Figure 13), but grazing sections reveal that this appearance is due to
achromatic "holes," enlarged compared to interphase (compare Figure 12
with Figure 3). In late prophase, however, strand-like connections
between chromosomes and the nuclear envelope are real (Figure 14).
No kinetochores are discernible in early prophase (Figure 8a).
In sections of nuclei in mid- and late prophase kinetochores appear as
roughly circular patches of finely fibrillar material in slight
constrictions of chromosomes (Figures 11a, 15, and 17). Serial
sections revealed that these kinetochores are globular and 5,000-8,000
A in diameter. I cannot state with certainty that the diameter
decreases with advancing prophase. The doubleness of the chromosomes
is evident in sections showing both sister kinetochores (Figures 15 and
17). The lesser electron density of the kinetochores compared to the
chromosomes is very distinct (Figures 10, 11a, 15-17). Slightly more
opaque kinetochore granules at the surface of the sister chromatids
(Figure 17) are very rarely seen.


146
Fig. Very early anaphase. Daughter chromosomes are "pulled"
apart in the kinetochore region, lite "frayed" appearance of chromo
somes; kinetochores (arrowheads), MT, and a centriole at pole no. 1
(P-j_). x 11,500* Inset: Phase contrast micrograph of the cell in
plastic (x 1,28o).


270
Metzner, R. 1894. As cited by Schrader, 1936.
Mizuno, N. S., and D. P. Gilboe. 1970. Binding of streptonigrin to
DNA. Biochim. Biophys. Acta 224: 319327
Mole-Bajer, J. 1958. Cine-micrographic analysis of c-mitosis in
endosperm. Chromosoma S>: 332-358.
Moor, H. 1967. Der Feinbau der Mikrotubuli in Hefe nach Gefrier&tzung.
Protoplasma 64: 89-IO3.
Murray, R. G., A. S. Murray, and A. Pizzo. 1965. The fine structure
of mitosis in rat thymic lymphocytes. J. Cell Biol. 26: 6OI-619.
Nathaniel, E. J. H., N. B. Friedman, and H. Rychuk. 1968. Electron
microscopic observations on cells of Harding-Passey melanoma
following colchicine administration. Cancer Res. 28: 1031-1040.
Nebel, B. R., and E. M. Coulon. 1962. The fine structure of chromosomes
in pigeon spermatocytes. Chromosoma 13: 272-291.
Nichols, ¥. W. 1970. Viruses and chromosomal abnormalities. Ann.
New York Acad. Sci. 171: 478-485.
Nichols, ¥. ¥., A. Levan, and B. A. Kihlman. 1964. Chromosome breakage
associated with viruses and DM inhibitors, p. 255-271* In
R. J. C. Harris (ed.) Cytogenetics of Cells in Culture.
Academic Press, New York.
Nicklas, R. B. 1967. Chromosome micromanipulation. II. Induced
reorientation and the experimental control of segregation in
me iosis. Chromosoma 21: 1750.
Nicklas, R. B. 1971. Mitosis. Adv. Cell Biol. vol. 2. In press.
Nicklas, R. B., and C. A. Koch. 1969. Chromosome micromanipulation.
III. Spindle fiber tension and the reorientation of mal-oriented
chromosomes. J. Cell Biol. 43: 40-50.
Nicklas, R. B., and C. A. Staehly. 1967. Chromosome micromanipulation.
I. The mechanics of chromosome attachment to the spindle.
Chromosoma 21: l-l6.
Oleson, J. J., L. A. Calderella, K. J. Mjos, A. R. Reith, R. S. Thie,
and I. Toplin. 1961. The effects of streptonigrin on experimental
tumors. Antibiot. Chemother. 11: 158-164.
Ostergren, G. 1951* The mechanism of co-orientation in bivalents
and multivalents. Tne theory of orientation by pulling.
Hereditas 37: 85-156.
Paweletz, N. 1967. Zur Funktion des "Flemming-Korpers" bei der
Teilung tierischer Zellen. Naturwissenschaften 54: 533-535


220
Fig. 116. Streptonigrin-treated cell in late telophase. Nucleus
of the upper daughter cell in the inset. Note nuclear pore complexes
(NP), kinetochores (arrowheads), nuclear pocket (also at the large
arrow), intranuclear MT (small arrows), x 15,750* Inset: Phase
contrast micrograph of the cell in plastic (x l,28o).


REVIEW OF LITERATURE
Mitosis has been studied for nearly 100 years (Dietz 1969). Its
fundamental importance as a mechanism for the orderly distribution of
chromosomes has provoked numerous investigations and experiments.
Accordingly, there is a vast amount of literature on both mitosis and
chromosomes. I did not intend to compete with the much more knowledge
able and experienced authors of the many reviews that have appeared.
Rather, I have tried to survey the literature on those aspects of
eukaryotic mitosis and chromosomes which are most closely related to
my own observations and results.
General Ultrastructural Features of
Vertebrate Cells in Mitosis
Today, improved methods are available that allow correlated light
and electron microscopic observation of single animal cells (e.g.,
Brinkley et al. 1967). It is a little surprising, therefore, that no
detailed study combining the two approaches has been done on animal
cells in mitosis. (For a good example of mitosis in plant cells see
Bajer 1968, Bajer and Mold-Bajer 1969.) Ultrastructural features of
mitosis have been described for a number of animal cells (see Luykx
1970 for references), but light micrographs are equally important for
comparison with earlier light microscopic studies.
In the following description the emphasis is on the nucleus.
Cytoplasmic organelles and components are considered only as far as
they play a role directly related to the formation of the mitotic
1


Figure Page
15 Late prophase chromosomes with
kinetochores 104
16 Mid-prophase chromosome with kinetochore ... 106
17 Mid-prophase chromosome with sister 106
kinetochores
18a Very early prometaphase cell ......... 107
18b Very early prometaphase cell 108
19 Early prometaphase cell 109
20 Early prometaphase cell 110
21a-c Early prometaphase. Nuclear envelope .... 112
22a,b Early prometaphase kinetochores 114
22c,d Early prometaphase kinetochores 116
23a-c Early prometaphase kinetochores 118
24a-c Early prometaphase kinetochores 120
25 Early proraetaphase kinetochores 121
26 Mid-proraetaphase cell 122
27 Mid- to late prometaphase cell 123
28 Late prometaphase chromosomes 124
29 Late prometaphase cell 125
30 Late prometaphase chromosomes 127
31 Late prometaphase chromosomes 127
32 Mid-prometaphase microtubules ........ 128
33a Late prometaphase chromosomes 129
33b Late prometaphase chromosomes 130
34a-c Late prometaphase chromosome 132
35a-c Late prometaphase chromosome ......... 134
36 Mid-prometaphase chromosome ......... 136
ix


99
Fig. 11a. Mid-prophase. Centriole (C) in pocket of the
nuclear envelope (HE). Note chromosomes (Ch) with sister kineto-
chores (Ki, K), remnant of the nucleolus (Nu). x 15,750. Inset:
Phase contrast micrograph of the cell in plastic (x 1,280).
0 O c o


162
Fig. 6l. Cytokinesis, Nucleus (ll) of one daughter cell with
large heterochromatic patches. Note cleavage furrow (arrows), stem
(Sm), midbody (MB). Black spot (d) and small dots are stain marks,
x 7,750.


255
chromosomes lie close to one pole in mid- and late prometaphase (e.g.,
Figures 26-29). Also, these unipolar attachments may be followed by
movement of the chromosome involved towards the near pole. At some
point, however, the sister kinetochore oriented towards the far pole
also becomes attached, and this is followed by movement towards the far
pole (Figure 120c). Figure 40 possibly illustrates the "return point,"
i.e., two chromosomes just beginning to move towards the equator. Once
the tension exerted on sister kinetochores is balanced, the chromosomes
remain stably in an equatorial position.
Chromosomes lying farther from the poles at the time the nuclear
envelope breaks down become attached later than the chromosomes lying
close. Whether this is so because of the greater distance from the
probable pool of MT subunits around the centrioles, or because they are
shielded from the poles by other chromosomes and therefore cannot form
connections until after reshuffling due to movements of already
attached chromosomes has occurred, remains an open question. Neverthe
less, the chance that the kinetochores of these chromosomes attach to
opposite poles simultaneously is greater, because the spindle axis is
established. For example, the chromosome in Figure 24 could have
attached earlier than the chromosome in Figure 25. Consequently the
latter shows bipolar orientation and attachment, whereas the former
shows bipolar orientation, but unipolar attachment. Equatorial
chromosomes immediately attached to both poles can be expected to be
stabilized by tension and to move very little during prometaphase.
However, if one sister kinetochore of an equatorial chromosome is
obstructed, unipolar attachment would precede bipolar attachment.
Again, the latter'could occur only after reshufflinghas cleared the


20
pushing force (similar to the "Stemmkorper" proposed by Blar 1929) may
contribute to anaphase separation. Clearly, prometaphase movement
cannot be explained on the basis of pushing forces only. The behavior
of chromosomes that are pushed away from the spindle into the cytoplasm
by micromanipulation also suggests a pulling force (Nicklas 1967). The
results of Forer's (1966) UV microbeam irradiation experiments on crane
fly spermatocytes seem to refute the idea that chromosomes are simply
pulled by their kinetochore fibers. Nevertheless, the results can be
interpreted as supporting a pulling hypothesis (Nicklas 1971).
Even if we accept the pulling hypothesis there remains the problem
of how the MT accomplish this, a problem around which center all of the
more recent models of mitosis (Dietz 1969, Luykx 1970, McIntosh et al.
1969). There is no change in diameter as MT shorten or lengthen (see
Luykx 1970, Nicklas 1971, for references), and a simple contraction
mechanism is not compatible with structural observations. Inou and
his collaborators (Inou 1960, 1964, Inou and Sato 1967) have
consistently explained this phenomenon with their "dynamic equilibrium
model," which proposes a pool of MT subunits in the spindle region.
Free subunits are in a dynamic equilibrium with subunits bound in MT.
Shifts in the equilibrium induce further polymerization or
depolimerization. The former would produce lengthening, the latter
shortening of MT. Orienting centers are thought to determine the
direction of the "growing" MT during polymerization.
Chromosomes
Three basic questions must be considered here: (1) What is the
unit fiber of the mitotic (and interphase) chromosome? (2) What is the


41
breaking effects of SN were of the delayed and non-delayed type.
Aberrations occurred in cells treated as late as 2 hr before fixation.
Among the aberrations observed were chromatid and isochromatid breaks,
acentric fragments, dicentric chromosomes, cleavage or severe attenua
tion of the centromere region, telomeric fusion of sister chromatids,
"stickiness," uncoiling of chromosomes, and severe fragmentation or
degeneration of the entire chromatin material. Chromosomal damage
appeared to be non-randomly distributed, chromosomes 19, 20, 21, 22,
and the Y being relatively stable. Cohen (1963) further studied the
non-randomness of aberrations produced by SN in chromosomes 1, 2, and
3 of cultured human leukocytes. While X-rays induced random breaks, SN
preferentially affected the pericentric regions of chromosomes 1 and 2,
as well as the area of the secondary constriction of chromosome 1.
Breaks in chromosome 3, although fewer in number, were distributed at
random. The telomere regions of all three chromosomes, and the short
arm of chromosome 2 appeared relatively resistant to SN.
Kihlman (1966) concluded from the work by Cohen et al. (1963) that
SN is able to break chromosomes during G^. A similar finding had been
reported for root tip cells of Vicia faba (Kihlman 1964). Exposure of
cells to 2-5 ;jg/ml SN for 1 hr produced subchromatid (cells in early
prophase), chromatid (cells in and S), and, possibly, chromosome
exchanges (cells in G^). However, Puck (1964) reported that SN does
not affect mammalian cells after completion of DNA synthesis. In this
respect SN may be similar to S-ethoxycaffeine (EOC) and other drugs
which have non-delayed effects in plant cells, and delayed effects, or
hardly any effects at all, in mammalian tissue culture cells (see
Kihlman 1966).


Page
Figure
61 Cytokinesis 162
62 Cytokinesis, midbody 163
63 C-metaphase 164
64 C-metaphase 165
65 C-mitosis 166
66 C-mitosis 167
67 Colcemid-treated interphase nucleus 169
68 Colcemid-treated interphase cell 169
69 C-mitosis 169
70 Cold-treated cell in mid-prophase 171
71 Cold-treated cell in mid-prometaphase .... 171
72 Prophase kinetochores of cold-treated cell 173
73 Microtubules of cold-treated cell in
raetaphase 173
74 Cold-fixed control cell in metaphase 173
75 Cold-treated cell in prometaphase 175
76 Cold-treated cell in cytokinesis 175
77 Ccntrioles of cold-treated cell in
metaphase ..... ...... 176
78a-c Serial sections of early anaphase
kinetochore 178
78d,e Serial sections of early anaphase
kinetochore (cont'd.) 180
79 Very early anaphase kinetochore 180
80 Kinetochore in late prometa- to metaphase 182
81 Metaphase kinetochore 182
82 Kinetochore in very early anaphase 182
83a,b Sister kinetochores in very early anaphase 184
xi


109
Fig. 19. Early prometaphase. Breakdown of the nuclear envelope
(KE) near poles (Pj_, Pp). Note nucleolus organizer (NO) on X
chromosome. The white circle near the left margin is due to a hole
in the supporting film; "black spots are dirt and stain marks,
x 1,150. Inset: Phase contrast micrograph of the cell in plastic
(x 1,280).


49
Enbed in Luft's Epon (Luft 1961) as follows. A ratio of 1 part of
mixture A to 2 parts of mixture B gave the best results.
2 parts HPA : 1 part Epon for 15 min
Equal parts of HPMA and Epon for 15 min
1 part IIPIA : 2 parts Epon for 30 min
2 changes of pure Epon, 30 min each
Add fresh pure Epon and bum holes in the top of the culture flasks
with a hot glass rod. Drain off excess Epon until a layer about the
thickness of a glass slide is left. Leave overnight at room temperature
then transfer to 60C for 24 hr or longer.
Preparation of Cells for Light and Electron Microscopy
The basic procedure was adopted from Brinkley et al. (1967).
After curing, the bottom of the flask with the adhering Epon wafer is
cut out. Wafer and flask bottom are separated by alternately cooling
the sandwich in liquid N_ and thawing in tap water. Once separation
has started at the edge, the Epon wafer can be snapped loose.
For light microscopy (scoring of aberrations, determination of
mitotic indices) wafers are placed cell layer up on the stage of a
phase contrast microscope. Detailed examination of cells is possible
by using oil immersion objectives.
The area around cells selected for thin sectioning is marked with
a sharp needle under low power. Several high magnification pictures at
different levels of focus are then taken. A disk with the marked area
is cut out with a cork borer of suitable diameter. The piece of Epon
is roughly trimmed and glued to the tapered end of a plastic peg, cell
layer up. Fine trimming to the boundaries of the trapezoidal or


219


Fig. 6. Interphase. Chromatin fibers in the nucleus of
Fig. 5a. Note 250 A fibers (large arrows), 125 A fibers (arrow
heads), 70-80 A core within some 250 A fibers (small arrows),
x 75,000.
Fig. 7 Interphase. Portion of the nucleolus shown in
Fig. 5*>. Note granular (Gr), fibrillar (f) components; nucleolus-
associated chromatin (Chr). x 30,000.


Table 3.Frequency of chromosomal and mitotic aberrations In streptonlgrin-treated
cells.
No. of Metaphase
No. of Anaphase
Total No.
Percent
Treatment
Cells
Cells
of Cells
Normal
Abnormal
Normal Abnormal
Normal Abnormal
Counted
Cells
Cells
0.01 ug/ml SN
63 65
16 33
200
39.5
49.0
0.05 ug/ml SN
*1
CN
O
0 14
35
0
100.0


271
Peacock, V/. J. 1963. Chromosome duplication and structure as de
termined by autoradiography. Proc. Hat. Acad. Sci 49: 793"801.
Pickett-Heaps, J. D. 1969. The evolution of the mitotic apparatus:
an attempt at comparative ultrastruetural cytology in dividing
plant cells. Cytobios 1: 257-280.
Pickett-Heaps, J. D. 1971* The autonomy of the centriole: fact or
fallacy? Cytobios 3: 205-21It.
Pickett-Heaps, J. D., and L. C. Fowke. 1969 Cell division in
Oedogonium. I. Mitosis, cytokinesis, and cell elongation.
Aust. J. Biol. Sci. 22: 857-894.
Porter, K. R., and R. D. Machado. i960. Studies on the endoplasmic
reticulum. IV. Its form and distribution during mitosis in
cells of onion toot tip. J. Biophys. Biochem. Cytol. 7: 167-180.
Puck, T. T. 1964. Phasing, mitotic delay and chromosomal aberrations
in mammalian cells. Science ihlt: 565-566.
Radding, C. M. 1963. Incorporation of H^ thymidine by K 12 (A )
induced by streptonigrin, p. 22. In S. J. Geerts (ed.)
Genetics Today (Proc. XI Int. Congr. Genet.).
Rao, K. V., K. Biemann, and R. B. Woodward. 1963 The structure of
streptonigrin. J. Arner. Chem. Soc. 85: 2532-2533
Rebhun, L. I., and G. Sander. 1967 Ultrastructure and birefringence
of the isolated mitotic apparatus of marine eggs. J. Cell Biol.
34: 859-883.
Renaud, F. L., and H. Swift. 1964. The development of basal bodies and
flagella in Allomyces arbuscuius. J. Cell Biol. 23: 339354.
Revell, S. H. 1955 A new hypothesis for "chromatid" changes, p. 243-253
In Z. M. Bacq and P. Alexander (eds.) Radiobiology Symposium
1954. Academic Press, Hew York.
Revell, S. H. 1963 Chromatid aberrations the generalized theory,
p. 41-72. In S. Wolff (ed.) Radiation-Induced Chromosome Aberrations.
Columbia Univ. Press, Hew York.
Reynolds, E. S. 1963 The use of lead citrate at high pH as an
electron-opaque stain in electron microscopy. J. Cell Biol.
17: 208-212.
Ringo, D. L. 1967a. Flagellar motion and fine structure of the
flagellar apparatus in Chlanydomonas. J. Cell Biol. 33: 543-571.
Ringo, D. L. 1967b. The arrangement of subunits in flagellar fibers.
J. Ultrastruct. Res. 17: 266-277.
Ris, H. 1949 The anaphase movement of chromosomes in the spermatocytes
of the grasshopper. Biol. Bull. 96: 90-106.


33
Chromosomal and Mitotic Aberrations
Chromosomal aberrations occur spontaneously in certain organisms
(e.g., Brandham 1970, Vig 1970). Far more frequent, however, are
induced aberrations. Shaw (1970) presents a long list of agents
(clastogens) that can cause chromosome damage. The agents range from
nucleic acid analogs, antibiotics, drugs, and pesticides, to ionizing
and UV radiation, temperature shock, and weightlessness. Certain
viruses are well-known biological clastogens (e.g., Nichols 1970,
Nichols et al. 1964).
Only some of the elementary aspects of the induction of chromo
somal and mitotic aberrations can be reviewed here. The emphasis will
be on chemical clastogens. Radiation-induced aberrations will be
referred to only to the extent that they elucidated basic facts that
also apply to chemicals (for detailed reviews on radiation-induced
aberrations see Bacq and Alexander 1955, Evans 1962, Hollaender 1954,
Lea 1962, Wolff 1963).
Cellular Events
Kihlman (1966) distinguished three types of effect of chemicals on
dividing cells and chromosomes: (1) Prevention of cells from entering
mitosis [e.g., 5-fluorodeoxyuridine, FUdR (Taylor 1963)]; (2) inter
ference with active stages of division [e.g., spindle poisons, such as
colchicine (Eigsti and Dustin 1955)]; (3) production of chromosomal
aberrations [e.g., streptonigrin (Cohen et al. 1963)]. It is
characteristic for many chemicals that they affect mitosis and
chromosomes (see also reviews by Biesele 1962, Deysson 1968, Gelfant
1963). For example, FUdR inhibits mitosis and also fragments


212


80
equatorial constriction. Two cells in cytokinesis were found with
"nuclear" bridges connecting the daughter nuclei through the compact
midbody. One of these cells was examined in the electron microscope.
Nuclear pore complexes were present on the NE wrapping the bridges.
Lagging chromosomes.The second most common type of aberration
in anaphase cells is laggards. Only a few such chromosomes could be
examined in serial sections. Representative profiles of kinetochores
of two laggards are shown in Figure 102. These chromosomes were lying
near or at the periphery of the spindle (inset in Figure 102a).
Characteristically, a bundle of arched MT was present at the kineto
chores, which were oriented towards the center of the spindle. Some
of these MT apparently bypassed the chromosome, while others seemed to
terminate in the fuzzy kinetochores (Figure 102). The bundle of MT
associated with laggard no. 1 (Figure 102a and inset) also seemed to
be associated with laggard no. 2 (inset in Figure 102a) in the
opposite half-spindle. Numerous non-kinetochore MT near the laggards
were oriented haphazardly. Kinetochores and MT of most of the chromo
somes near the poles were normal. The chromosome in Figure 103 was an
exception. The strange profile of its lcinetochore in some of the
serial sections may be due to a peculiar angle of sectioning. Some of
the kinetochore MT, particularly in peripheral sections of the kineto-
chore (Figure 103b), were definitely odd.
In a similar cell one kinetochore of a laggard was very stretched
and similar in structure to the one in Figure 102a.


228
Furthermore, I have examined numerous of my own micrographs of
centrioles at all stages of mitosis. In reasonably thin sections there
is no evidence whatsoever for direct connections between MT and
centrioles. Rare images where this seems to be the case can be ex
plained by lack of resolution or superposition in rather thick sections.
If the hypothesis were correct, one could reasonably expect such MT-
centriole associations to be frequent and unambiguous. The fact that
even the most ardent proponents of the hypothesis cannot claim this
speaks for itself.
Having dispensed with a popular myth we now can direct our atten
tion to a probably more important polar constituent, the osmiophilic
material surrounding and partly enveloping the distal ends of the
centrioles (e.g., Figures 45 and 56). The possibility that this
material is absent or reduced in amount at least during part of the
interphase (e.g., Figure 3) merits further investigation. It is
certainly present during all stages of mitosis, and it seems to be
arranged mainly around one of the centrioles of a pair, probably the
parent (Figures 56 and 77). I have been unable to observe changes
related to the mitotic cycle, as reported for HeLa cells by Robbins
et al. (1968). The material also occurs in colcemid-, cold-, and SN-
treated cells (Figures 63 and 77). Spindle MT penetrate into this
material in a rather chaotic fashion and apparently terminate there
(Figure 56).
Pickett-Heaps (1969) suggested that the microtubule-organizing-
center (MTOC) of animal cells could be in the osmiophilic material,
rather than the centrioles themselves. He did not explicitly discuss
the possibility that the MTOC, or some other component of the


0.01 pg/ral were prepared under sterile conditions by dilution with MEM.
No serum was added to avoid possible binding of SN by serum proteins.
Cell monolayers in Falcon 30-ml flasks were rinsed once with MEM
before the drug was added. Control cells were treated in the same
manner, except that they were exposed to MEM alone instead of the drug.
The cells were incubated for 4 hr at 37C. The flasks were then rinsed
twice with complete medium and the cells were left to recover in the
incubator in a third change of fresh complete medium. Approximately 24
hr later the medium was changed once more. After a total recovery
period of 48 hr the cells were fixed _in situ according to the standard
schedule (see below).
Colcemid
A stock solution of 0.5 pg/ml was prepared by dissolving crystal
line colcemid (CIBA) in distilled water, pH 6.4. This solution was
sterilized with a millipore filter and further dilutions were prepared
with sterile complete medium. Treatment A was 0.05 ;ig/ml colcemid for
2 hr at 37C, followed by two rinses with complete medium and recovery
for 1 hr at 37C in a third change of medium. Treatment B was 0.25
^ig/ml colcemid for 15 min at 37C without a recovery period. Control
cells were left undisturbed and were fixed at the same time as treated
cells.
Cold
Flasks were seeded as usual and the cells left to attach overnight
at 37C. The medium in the experimental flask was then poured off and
replaced by complete medium of 0-4 C. The flask was immediately placed
in a pan with ice water so that its lower portion was immersed. After


56
The nucleolus, still intact in very early prophase (Figure 8a),
fragments into several masses of granular material (Figures 10 and
11a), some of which are apparently associated with chromosomes (Figure
11a).
The nuclear envelope remains intact until the end of prophase, but
the pore-annulus complexes become more fuzzy (Figure 12). Polyribo
somes are found on the nuclear envelope throughout prophase (e.g.,
Figure 11a).
Prometaphase
The breakdown of the nuclear envelope, indicating the transition
from prophase to prometaphase, is gradual, but structural changes
involving MT and lcinetochores are more striking. Fragmentation of the
nuclear envelope always begins nearest the centrioles (Figures 18a and
19). Fragments are undulated or form vesicles. The telomeres of
chromosomes at the periphery of the "nucleus" are often trapped in
compartments formed by undulated, still intact portions of the nuclear
envelope (Figure 18b). With progressive development of the spindle
apparatus, envelope fragments become smaller and scarcer, but some
persist at the periphery of the spindle until late prometaphase
(Figures 20, 26, and 29). Disappearing central granules indicate early
stages of the breakdown of pore-annulus complexes (Figures 21a and
21b). The entire complexes on fragments of the nuclear envelope
become fuzzy and disappear by mid-prometaphase (Figure 21). Polyribo
somes can be found on clearly identifiable fragments until late pro
metaphase (Figures 21, 26, and 29).
Chromosome condensation continues throughout prometaphase (Figures
18-20, 26-30). The chromosomes detach from the nuclear envelope as the


161
Fig. 60. Mid-telophase. Nuclear envelope (NE) completely surrounds
the nucleus. A few ribosomes (R) present on the outer membrane of the
NE. Two kinetochores (k) present on the polar face of the nucleus.
Note MT (arrow) extending into the kinetochore pocket, x 30,000.
Inset: Phase contrast micrograph of the cell in plastic, (x 1,280).


73
metaphase cell. This section was picked up on an uncoated 200-mesh
grid. The difference in contrast between the chromosome and the
kinetochore is undeniable.
The basic structure of the outer band is finely granular in
highly condensed kinetochores (Figures 78, 81-84), but in grazing
sections of less condensed kinetochores, 30-50 A fibrils are visible
(Figure 80). Superimposed on the fine granularity of condensed ki
netochores is a structure of coarse granules or fibers, giving the
band a knotted appearance (Figures 78, 81-85). The structure and
electron density of the middle band are very similar to the corona
(Figures 78b and 84). The inner band is continuous with the chromosome
(Figures 78b and 84), but in very early anaphase it may be connected
to the main body of the chromosome by a "stalk" of chromatin, giving it
the appearance of a mushroom (compare Figure 84 with Figure 46). The
fibers of the inner band seem to be identical with the fibers of the
chromosome. The greater opacity may be due to denser packing, but an
interesting alternative is the presence of a very fine amorphous
substance, which is lacking in the remainder of the chromosome.
Microtubules attach to the kinetochore at a variable angle.
Three conditions must be fulfilled in paraxial sections to determine
how far the MT penetrate into the kinetochore: (1) The section must
be thin (500 A); (2) the ITT in question must not be cut obliquely near
the kinetochore; and (3) the section must be approximately median.
These conditions are fulfilled in Figure 82. The MT marked with an
arrow penetrates the outer layer and ends at the interface with the
middle layer. The different impression created by an obliquely
sectioned MT is demonstrated in the somewhat thicker section of Figure


KEY TO ABBREVIATIONS AND SYMBOLS
C
CG
Ch
Chd
Chr
Ci
Co
CV
D
DC
EOC
F
G
Gr
H
HChr
IV
K
KI
KM
KO
MA
KB
MF
Mi
MI
MU
MT
N
NE
NL
NO
NP
Nu
NV
PC
R
RER
S
Sb
Sm
SN
V
X
centriole(s)
centromeric granules
chromosome(s)
chromatid(s)
chromatin
cistema(e)
corona of kinetochore
intracentriolar vesicle
contamination of thin sections (dirt)
daughter centriole(s)
8-ethoxycaffeine
fibrillar component of the nucleolus
Golgi complex
granular component of the nucleolus
achromatic hole(s) in chromatin or chromosomes
heterochromatin
intranuclear vesicle(s)
kinetochore
sister kinetochores
kinetochore granule
inner layer of the kinetochore
middle layer of the kinetochore
outer layer of the kinetochore
mitotic apparatus
midbody
microfibril(s)
mitochondrion
mitotic index
micronucleus
microtubule(s)
nucleus
nuclear envelope
nuclear lobe
nucleolus organizer
pore-annulus complex(es) of the nuclear envelope
nucleolus
membrane vesicle formed by fragmenting nuclear envelope
spindle pole(s)
particle(s) associated with centrioles
parent centriole(s)
ribosomes (poly- or monosomes)
rough endoplasmic reticulum
satellite(s), pericentriole body (bodies)
stem body
stem
streptonigrin
cytoplasmic vesicle(s)
X chromosome; other chromosomes are numbered arbitrarily
xiv


148
wmm
'T'-i** .' 4
wm
fcSfcfr&&fcS
sMU
- *
..vv >^1" jai
Fig. 4j. Early anaphase. Telomeric regions of the separating
daughter chromosomes trail the kinetoehore regions. Several kineto-
chores (arrowheads) and connecting strands (arrows) are visible.
Black spots are dirt, x 11,500 Inset: Phase contrast micrograph
of the cell in plastic (x 1,2S0).


207


94
WtM:
i* sK'JL <>
Fig. 8a. Very early prophase. Note two pairs of centrioles (Cp,
Co), condensing chromosomes (Ch), nucleolus (Nu). Nuclear envelope
(RE) with pores (NP); Golgi complexes (g), rough ER (NCR), and mito
chondria (Mi) in the cytoplasm, x 11,500.


39
and the broken ends rejoin; or they uncoil and do not break at all.
Whether or not a bridge breaks or is stretched depends on its length
and the position of the kinetochores relative to the equator at meta
phase. Breakage at anaphase is due to the pulling force of the
chromosomal fibers; at telophase it is caused by the phragmoplast.
Sister chromatid bridges tend to break at apparently weak points which
are sometimes seen as constrictions. Movement after breakage is
scarcely faster than the initial anaphase speed, indicating there is no
accumulated tension in the pulling mechanism. Humphrey and Brinkley
(1969; see also Brinkley and Shaw 1970) observed apparent gaps within
anaphase bridges in rat kangaroo cells. These gaps were constrictions,
presumably caused by anaphase tension, within which three classes of
chromosomal fibrils could be seen.
Chromosome "stickiness"
When cells are irradiated in late prophase, "sticky" bridges can
be observed at the following anaphase (Carlson 1954). Chemical
clastogens, e.g., nitrogen mustard, also produce this effect (Koller
1953). Stickiness has been interpreted to be the consequence of
surface changes on the chromosomes, changes that make chromosomes
adhere to each other if they happen to come in contact (Carlson 1954,
Casarett 1968). Apparently, this aberration is reversible: if mitosis
is delayed after treatment, stickiness does not occur. The nature of
the sticky material remains obscure. Hsu et al. (1965) noted that in
Chinese hamster cells nucleolar material sometimes remains attached to
the ends of sister chromatids, forming apparent chromatin bridges at
anaphase.


Fig. 21a-c. Early prometaphase. Grazing sections of the
nuclear envelope at progressively later stages, (a) From the
cell in Fig. 18; note eccentric pore granules (arrows), x 57>500.
(b) x *10,000. (c) x 50,000. Compare pore-annulus complexes (NP),
polyribosomes (R) in the three micrographs.


65
the triple-layered character of the kinetochores is less clear, but
still recognizable (Figures 51b and 52). Most kinetochores in very
late anaphase cells appear in depressions on the poleward face of the
chromatin mass (Figure 53). The less dense band in Figure 53 is
approximately 500 A wide and set off from the chromatin by a 250 A wide
clear band. The chromatin immediately underlying the kinetochore is
denser and less obviously fibrillar than the remainder of the
chromosomal mass. Remarkable is the decreasing number and degree of
organization of kinetochore MT in late anaphase (compare Figures 46-49
with Figures 51-53).
Spindle elongation is the rule in PtK^ cells (Figure 51a). The
interzone is still free of cytoplasmic organelles even after the
chromosomes have almost reached the poles. Pieces of double membrane,
some with ribosomes, are found at the periphery of the spindle area,
particularly around the two sets of chromosomes (Figure 51a).
Interzonal MT are scarce. Mitochondria, vesicles, and ER invade the
interzone in very late anaphase (Figures 52 and 55b). The cytoplasm
begins to constrict in the equatorial region (Figure 52, inset), where
stem bodies appear. The latter consist of amorphous, osmiophilic
material, within which MT are closely packed (Figures 54 and 55a). It
is difficult to determine from these sections whether the MT terminate
in the stem bodies or beyond. Both possibilities are likely. One MT
in Figure 55a clearly passes through the stem body. Its total length
visible in this section was 3 i.
It is possible that pieces of double membrane as seen in Figure
51a become involved in the reconstitution of the nuclear envelope.
More typically, however, small cistemae and vesicles appear on the


144
s^-vy-;^,
M^mw
*.' <-.wr Yte&'x *V
Fig. 43. Metaphase. Para-sagittal section showing four pairs of
sister chromatids. Three kinetochores (arrowheads), one of them distinctly
triple-handed and concave (upper left). Note dense chromosomal granules
(arrows), x 22,500. Inset: Phase contrast micrograph of the cell in
plastic (x l,28o).


11
and disassembled (Inou 1964, Inou£ and Sato 1967). It should be
noted, however, that the direct connection of MT to centrioles has not
been proven unequivocally (see de Ilarven 1968, Luykx 1970, Pickett-
Heaps 1969). More often, MT seem to connect to pericentriolar bodies
or the amorphous, osmiophilic mass surrounding the centrioles. Similar
dense masses can be found at the poles of higher plant cells (Pickett-
Heaps 1969, Wilson 1970). Most recent reviews, therefore, consider
this material, or part of it, a more likely candidate for a pole-
determinant and MT organizer, not only because it occurs almost
universally, but it would also bridge the gap between centriolar and
acentriolar mitosis (see Luykx 1970, Nicklas 1971, Pickett-Heaps 1969).
In this scheme, centrioles are thought to play a much more passive
role, being carried along and distributed mainly for use as basal
bodies of cilia and flagella (see also FriedlSnder and Wahrman 1970),
Brinkley and Stubblefield (1970; see also Stubblefield and
Brinkley 1967) have presented a different hypothesis. They maintain
that centrioles do play a role as pole-determinants and in the assembly
of MT. However, in the absence of solid evidence, their detailed
mechanistic and molecular model of centriole-MT interaction in mitosis
remains highly speculative.
Spindle Fibers
Under favorable conditions fibrillar elements are visible in the
mitotic spindle of living cells with the phase contrast microscope (Ris
1955). In fixed and stained cells spindle fibers can be seen without
difficulty in the light microscope (e.g., Heneen 1970). The Nomarski
interference-phase contrast system also allows visualization of the


248


87
Fig. 3. Interphase. Grazing section of the nucleus. Note
centrioles (c), Golgi complex (g), polyrihosor.es (r) on nuclear
envelope, pore-annulus complexes at the level of the envelope (small
arrows) and the level of the nucleus (large arrows), achromatic holes
(H) in the peripheral chromatin (Chr) underlying pores, x 15,750.


46
According to Walen (1965) the generation time in Eagle's medium with 4%
fetal calf serum (FCS) is 28-32 hr. The low number and the individual
ity of chromosomes make this line particularly suitable for cytogenetic
and labeling studies.
I obtained an ampule of frozen cells from the American Type
Culture Collection and started a culture in April, 1970. After
preliminary experiments this culture was given up in September, 1970,
because of a suspected contamination by a microorganism (mycoplasma or
fungus). A new culture with fresh cells was initiated in October, 1970.
Only cells from this culture were used in the experiments described.
Cells of the stock culture were grown in square glass bottles in
Eagle's minimum essential medium (MEM; GIBCO), supplemented with 10%
FCS (GIBCO). Cells were harvested with trypsin-versene at weekly
intervals and subcultured at appropriate dilutions. Fresh complete
medium (MEM with 107. FCS) was supplied once between transfers. Cells
to be used for experiments and controls were seeded in Falcon 30-ml
flasks and left to attach overnight. Cultures were maintained at 37C,
but handling was carried out at room temperature.
Chemical and Physical Treatments
Streptonigrin
Streptonigrin (Pfizer and Co., Inc.) was obtained from Cancer
Chemotherapy, NCI, NIH, Bethesda, Md. A stock solution of 250 pg/ml
was prepared shortly before use by dissolving the powder in the diluent
supplied. The stock solution was diluted volumetrically in distilled
water, pH 8.0, to a concentration of 0.25 jjg/ml. After sterilization
with a millipore filter, the final concentrations of 0.05 pg/ml and


100
Fig. 11b. Mid-prophase. Serial section of the cell in Fig. 11a,
at the periphery of the nucleus. Note microtubules converging at the
location of the centrioles (c); the centrioles themselves not shown.
Mitochondria (Mi) also radially oriented. Grazing sections of the
nuclear envelope (NE). x 11,500. ... ....


Fig. 103a, b. Kinetochore and MT of a non-lagging chroinosone
in the sane cell as Fig. 102. Note curved and disoriented MT
(arrows), kinetochore (k). x 50,000*


245
The Fine Structure of Mature Kinetochores
Several arguments based on published observations and electron
micrographs speak against the filament model of mammalian kinetochores
proposed by Brinkley and Stubblefield (1966, 1970). First, the kireto-
chores in colcemid-treated Chinese hamster cells, on which the original
paper was almost exclusively based (Brinkley and Stubblefield 1966),
are structurally different from kinetochores of normal metaphase
chromosomes. The inner kinetochore layer is absent in c-mitotic cells,
but it is present on metaphase chromosomes in cells recovering from
colccmid (Brinkley et al. 1967). Secondly, random sections of mitotic
cells, particularly if embedded as a pellet (as were the cells used by
Brinkley and Stubblefield 1966), should yield at least some cross
sections of kinetochore filaments. Such profiles have never been
presented. Thirdly, the difference in electron density between outer
and inner bands of normal metaphase kinetochores, observed by a number
of investigators (e.g., Bamicot and Huxley 1965, Luykx 1965a),
suggests the two bands do not represent identical filaments. Fourthly,
Jokelainen (1967) and McIntosh and Landis (1971) have published electron
micrographs of para-equatorial sections of metaphase cells showing
circular kinetochores. Krishan (1968) also presented a micrograph
showing a circular kinetochore in a telophase cell after recovery from
vinblastine. Finally, kinetochores in a variety of cells treated with
the spindle poisons vincristine and vinblastine are virtually identical
with kinetochores in colcemid-treated Chinese hamster cells, and
equally different from normal metaphase kinetochores (George et al.
1965, Journey et al. 1965, Journey and Uhaley 1970, Krishan 1968).


Fig. 108a-f. Streptonigrin-induced aberrations. Phase
contrast micrographs, (a) Aberrant "metaphase". Chromosomes
scattered in the spindle, sister chromatids twisted, (b) Very-
early anaphase. Chromosomes not aligned on the equator. Sister
chromatids bifurcated in the stretched centromere regions, (c)
Late anaphase. Dicentric bridges (center), acentric fragments
(arrow), (d) Late anaphase. Two dicentric bridges (center),
acentric fragments (arrow). (e) Late anaphase. Two prominent
bridges. Lagging daughter chromosomes connected by very thin
bridge (center of spindle). Acentric fragments (arrows), (f)
Late anaphase telophase. Three thin bridges (center), two
groups of acentric fragments (arrows). (a), (b), (c), (e), and
(f): 0.05 ^ig/ml SIT; (d): 0.01 ig/ml SN; (all x 1,280).


222


6
Grimstone 1960). The proximal end, where daughter centrioles are
formed during duplication, exhibits a cartwheel structure with spokes
radiating from a hub. This cartwheel is recognizable in ideal cross
sections only, but it can also be reenforced photographically by the
Markham technique (Markham et al. 1963; for an example see de Harven
1968, Figure 2). The spokes seem to connect the hub with the innermost
MT of each triplet; other connections possibly occur between the
outermost and innermost MT of neighboring triplets. Centrioles are
often embedded in an amorphous, osmiophilic matrix (Murray et al. 1965,
Robbins and Gonatas 1964, Stubblefield and Brinkley 1967), that may
undergo periodic changes in preparation for, and during, cell division
(Robbins et al. 1968).
In Chinese hamster fibroblasts the occurrence of a small,
membrane-bound vesicle, termed nucleoid, approximately in the center of
the centriole seems to be the rule (Brinkley and Stubblefield 1970,
Stubblefield 1968, Stubblefield and Brinkley 1967). These
investigators also claimed to have detected a helical filament, 60-70 A
in diameter, that winds 8-10 turns just inside the triplets.
In many cells so-called satellites (pericentriole bodies) have
been observed (Brinkley and Stubblefield 1970, de Harven 1968, Murray
et al. 1965, Robbins et al. 1968). These are osmiophilic bodies in the
vicinity of interphase centrioles, but in most cell types studied they
seem to be absent or inconspicuous during mitosis (de Harven 1968).
They possibly contain RNA (Brinkley and Stubblefield 1970). The
existence of connections between satellites and centrioles, and the
possibly regular number and arrangement of satellites around centrioles
are controversial (see de Harven 1968 for a discussion). A few authors


124
mBw
Fig. 28. Late prometaphase. Ch2, Ch^, two equatorial chromosomes
Chj_ the chromosome displaced towards pole no. 1 (arrcw in inset). K-i,
Kp, the pole no. 1 and pole no. 2 kinetochores, respectively, x 11,500
Inset: Phase contrast micrograph of the cell in plastic (x l,28o).


165
Fig. 6k, C-metaphase. Two centrioles axe visible (c). A kineto-
chore is indicated by the arrowhead. Vesicles (V) occur in the central
and peripheral areas. Mitochondria (Mi), RER, and cisternae (Ci) axe
restricted to the peripheral area, x 22,500. Inset: Phase contrast
micrograph of the cell in plastic (x 1,280). Treatisent A. ...


258
The possibility that endoplasmic reticulum (ER) is involved in the
reconstruction of the NE was raised by Porter and Hachado (1960), based
on studies of mitosis in onion root tip cells. Bajer and Mole-Bajer
(1969; Haemanthus endosperm cells) and Robbins and Gonatas (1964;
HeLa cells) agreed that this is a likely possibility. My own observa
tions support this concept (Figures 51a, 52, 112, and.117). Quite
frequently, cistemae xjith ribosomes can be seen apposed to the
chromosomal mass (Figure 52). Other cistemae, in a similar associa
tion with chromatin, appear continuous with rough ER.
The serial sections yielded numerous oblique views of NE fragments
in prometaphase, thus providing a better insight into the fate of
nuclear pore complexes. There is no doubt that the pores disappear
(Figure 21). Remarkable is their reappearance on very small cistemae
associated with chromatin in late anaphase (Figure 58). One gets the
impression that pores are reformed almost immediately upon contact of
cistemae with chromatin, but the process is shrouded in mystery.


Fig. 53* Very late anaphase. Kinetochore (K) in depression
on the poleward face of the chromosomal mass. Note less dense hand
(arrowhead), density of underlying chromatin, kinetochore MT.
Arrows indicate chromosomal granules. Serial section of the cell
shovn in Fig. 52. x 75,000.
Fig. 54. Very late anaphase. Stem bodies (Sb) in the
equatorial region. Serial section of the cell shovn in Fig. 52.
x 30,000.
Fig. 55a. Very late anaphase. Detail of stem body. Most
MT seem to terminate in the stem body or to project a very short
distance beyond it, but one MT (arrows) clearly passes through,
x 57,500.


34
chromosomes (Hsu et al. 1964). Other chemicals are primarily
O' 3 9 3 O
chromosome-breaking agents, but at the same time they affect the
mitotic rate. To discriminate between chromosomal and mitotic effects,
Deysson (1968) classified the cytological effects of antimitotic
substances as follows: mitodepressive (lowering the mitotic rate),
mitostatic (no proliferation), mitoclasic (disturbances of the mitotic
apparatus), and chromatoclasic (induction of aberrations). Generally,
the lowest effective concentration of a given chemical is mitodepres
sive. The same, or a slightly higher concentration produces
chromosomal aberrations. Higher concentrations, besides inducing
chromosomal anomalies, cause preprophase inhibition (mitostatic
effect), and still higher concentrations destroy cells in mitosis.
Chemicals Versus Ionizing Radiation
Aberrations produced by ionizing radiation applied to cells in
are of the chromosome type (e.g., Heddle 1969). The transition from
chromosome to chromatid aberrations occurs at the end of G^ (Evans and
Savage 1963), but some authors maintain that both chromosome and
chromatid aberrations can result from irradiation during S, depending
on whether unreplicated or replicated parts of the chromosomes are hit
(e.g., Casarett 1968). Chromatid aberrations only result from
irradiation of cells in G^, i.e., after completion of ENA synthesis.
Subchromatid aberrations can be induced during prophase and, possibly,
at the end of G^ (see Heddle 1969). The term non-delayed effect used
by Kihlman (1966) with reference to chemical clastogens and radiation
implies aberrations produced after completion of ENA synthesis
(chromatid and subchromatid aberrations), while delayed effects include
chromatid and chromosome aberrations (induced during G^ and S).


274
Uretz, R. B., W. Bloom, and R. E. Zirkle. 1954. Irradiation of parts
of individual cells. II. Effects of an ultraviolet microbeam
focused on parts of chromosomes. Science 120: 197-199*
Vig, B. K. 1970. Sub-chromatid aberrations in Haworthla attenuate.
Can. J. Genet. Cytol. 12: l8l-l86.
Walen, K. H. 1965. Spatial relationships in the replication of
chromosomal DM. Genetics 51s 9Y5~929*
Walen, K. H., and S. V/. Brown. 1962. Chromosomes in a marsupial
(Potorous tridactylis) tissue culture. Nature 194: 4o6.
Watson, M. L. 1958. Staining of tissue sections for electron
microscopy with heavy metals. J. Biophys. Biochem. Cytol.
4: 475-478.
Went, H. A. i960. Dynamic aspects of mitotic apparatus protein.
Ann. N. Y. Acad. Sci. 90: 422-429.
Went, H. A. 1966. An indirect method to assay for mitotic centers
in sand dollar (Dendraster excentricus) eggs. J. Cell Biol.
30: 555-562.
Wettstein, R., and J. R. Sotelo. 1965 Fine structure of meiotic
chromosomes. The elementary components of metaphase chromosomes
of Gryllus argentinus. J. Ultrastr. Res. 13: 367-381.
Wilson, H. J. 1968. The fine structure of the kinetochore in meiotic
cells of Tradescantia. Planta 78: 379-385*
Wilson, H. J. 1969* Arms and bridges on microtubules in the mitotic
apparatus. J. Cell Biol. 40: 854-859*
Wilson, H. J. 1970. Endoplasmic reticulum and microtubule formation
in dividing cells of higher plants a postulate. Planta 94:
184-190.
Wilt, F. H., H. Sakai, and D. Mazia. 1967* Old and new protein in the
formation of the mitotic apparatus in cleaving sea urchin eggs.
J. Mol. Biol. 27: 1-7*
Wolfe, S. L. 1965* The fine structure of isolated metaphase chromosomes.
Exp. Cell Res. 37: 45-53*
Wolfe, S. L. 1969* Molecular organization of chromosomes, p. 3_42.
In E. E. Bittar (ed.) The Biological Basis of Medicine, vol. 4.
Academic Press, New York.
Wolfe, S. L., and P. G. Martin. 1968. The ultrastructure and
strandedness of chromosomes from two species of Vicia. Exp.
Cell Res. 50: 140-150.
Wolff, S. (ed.). 1963. Radiation-Induced Chromosome Aberrations.
Columbia Univ. Press, New York. 304 p.


31
Function
In addition to the role of the kinetochore in chromosome movement
as discussed in a previous section, Luykx (1970) reviewed five other
possible functions: (1) Initiation of synapsis and localization of
chiasmata; (2) terminalization of chiasmata; (3) chromosome condensa
tion or coiling; (4) association of sister chromatids; and (5)
formation or assembly of chromosomal spindle fibers. The latter is the
most interesting in the context of this review.
When chromosomal fibers are irradiated between the centromere and
the pole in anaphase cells, birefringence disappears from the
irradiated region and distal to it. Restoration of birefringence takes
place within a few minutes. When, however, the centromere itself is
irradiated, birefringence disappears from the whole length of the
chromosomal fibers, including the distal non-irradiated portion, and
restoration does not occur for a long time (Inou 1964). Electron
micrographs have shown that a great proportion of spindle LIT is
arranged in bundles associated with the kinetochores (Brinkley and
Landis 1970, Brinkley and Stubblefield 1970, Jokelainen 1967). These
observations have led to the idea that the kinetochore may play an
active role in the assembly and/or orientation of MT (e.g., the
"organizing center" of Inou 1964). Further support for this idea has
been drawn from observations of apparent microtubular connections
between meiotic sister kinetochores (Luykx 1965b) and mitotic non
sister kinetochores (Bajer 1970). As long as clear evidence from
serial sections is lacking, however, such configurations have to be
regarded with skepticism. On the other hand, the occurrence of
individual chromosomal spindles, either as an anomaly (e.g., Dietz


27
often less dense than the superficial layer and more variable in
o 1 m *
appearance. Approximately 10-25 MT, some of which seemed to end in the
deep layer, could be counted per kinetochore. On mitotic chromosomes
the triple-layered structure was less frequent. DNAse treatment in
dicated little or no DNA in the kinetochore region, or increased
resistance of kinetochore DNA to the enzyme (Luykx 1965a).
Wettstein and Sotelo (1965) found that the kinetochores in Gryllus
spermatocytes consist of essentially the same 100 A fibrils as the
remainder of the chromosome, but the fibrils seemed more densely packed
in the kinetochore. The shape of these kinetochores resembles that of
a thick nail or screw deeply anchored in the body of the chromosome.
Grasshopper spermatocytes have ovoid kinetochores embedded in a
cup (Brinkley and Nicklas 1968, Nicklas 1971, and personal communica
tion; see also Brinkley and Stubblefield 1970, Figures 26 and 27). An
electron-dense "axial core" is embedded in a mass of less dense 50-80 A
fibrils. In general appearance these kinetochores seem to resemble
plant kinetochores (Bajer and Mole-Bajer 1969, Wilson 1968) more than
animal kinetochores.
A different kind of kinetochore in invertebrates is the diffuse
kinetochore described by Buck (1967) in the bug Rhodnius. He
interpreted the finely granular material on the surface of metaphase
chromosomes as representing the kinetochore, but noticed the relative
paucity of spindle MT attached to this structure. Ris (Ris and Kubai
1970) reported on a preliminary basis that no such material could be
found on chromosomes in spermatocytes of the homopteran insect
Philaenus. Rather, the MT seemed to penetrate deeply into the
chromosome where they end blindly.


269
Lima-de-Faria, A. 1956. The role of the kinetochore in chromosome
organization. Hereditas 42: 85-160.
Lima-de-Faria, A. 1958. Recent advances in the study of the kinetochore.
Int. Rev. Cytol. 7: 123-157*
Luykx, P. 1965a. The structure of the kinetochore in meiosis and
mitosis in Urechis eggs. Exp. Cell Res. 39: 6^3~65T
Luykx, P. 1965b. Kinetochore-to-pole connections during prometaphase
of the meiotic divisions in Urechis eggs. Exp. Cell Res. 39:
658-668.
Luykx, P. 1970 Cellular mechanisms of chromosome distribution.
Int. Rev. Cytol., suppl. 2. Academic Press, New York. 173 P
Malavista, S. E., H. Sato, and K. G. Bensch. 1968. Vinblastine and
griseofulvine reversibly disrupt the living mitotic spindle.
Science l60: 770-772.
Mantn, I., K. Kowallik, and H. A. van Stosch. 1969a. Observations on
the fine structure and development of the spindle at mitosis and
meiosis in a marine centric diatom (Lithodesmium undulatum) I.
Preliminary survey of mitosis in spermatogonia. J. Microscopy
89: 295-320.
Mantn, I., K. Kowallik, and H. A. von Stosch. 1969b. Observations
on the fine structure and development of the spindle at mitosis
and meiosis in a marine centric diatom (Lithodesmium undulatum).
II. The early meiotic stages in male gametogenesis. J. Cell
Sci. 5: 271-298.
Mantn, I., K. Kowallik, and H. A. van Stosch. 1970. Observations on
the fine structure and development of the spindle at mitosis and
meiosis in a marine centric diatom (Lithodesmium undulatum). III.
The later stages of meiosis I in male gametogenesis. J. Cell Sci.
6: 131-157.
Markham, R., S. Frey, and G. J. Hills. 1963. Methods for the enhance
ment of image detail and accentuation of structure in electron
microscopy. Virology 20: 88-102.
Mazia, D. 1961. Mitosis and the physiology of cell division, p. 77
412. In J. Brachet and A. E. Mirsky (eds.) The Cell, vol. 3.
Academic Press, New York.
Mazia, D., P. J. Harris, and T. Bibring. i960. The multiplicity of
the mitotic centers and the time-course of their duplication and
separation. J. Biophys. Biochem. Cytol. 7: 1-20.
McIntosh, J. R., P. K. Hepler, and D. G. Van Wie. 1969. Model for
mitosis. Nature 224: 659-663.
c .. e *
McIntosh, J. R., S. C. Landis. 1971* The distribution of spindle
microtubules during mitosis in cultured human cells. J. Cell
Biol. 49: 468-497.


147
Fig. 46. Very early anaphase. Several sister kinetochores (K^,
K) shearing variation of profiles. Kj_ in the upper left comer is
sectioned ohliquely or peripherally. Note the electron dense strand
connecting kinetochores of the chromosomes left of center (arrow).
x 22,500.


36
During subsequent stages of chromosome development, the aberration is
transformed into a real chromatid exchange. Revell (1955) presented
data that are in good agreement with his hypothesis and Kihlman (1966)
came to the same conclusion based on more recent work. However, in
experiments especially designed to test the two hypotheses, Heddle and
Bodycote (1970) found that neither, as usually interpreted, is entirely
correct. Rather, they concluded that deletions are of two types,
according to mode of origin, but they were unable to identify the two
types morphologically.
Morphology and Mitotic Behavior of Aberrant Chromosomes
The following discussion is restricted to chromosomes with local
ized kinetochores.
"Gaps"
"Gaps," or "achromatic lesions" are Feulgen-negative regions of
variable size in chromatids (Evans 1963, Scheid and Traut 1970). In
Vicia they often resemble the normal nucleolar constriction. Gaps are
aberrations of the non-delayed type, since they can be induced by X-rays
in prophase nuclei (Evans 1963). The question of chromatid continuity
across the gap is important here, because gaps would have to be scored
as true breaks if the chromatid were really interrupted. Many
investigators (see Evans 1963 for references) have observed some sort of
material crossing gaps. Furthermore, the chromosome segment distal to
the gap moves normally and seems attached to the main body of the
chromosome at anaphase. If gaps were true breaks, they should give
rise to chromatid or chromosome aberrations at the second division
after exposure, but this is not the case (Evans 1963). Another


196