MICROFILAMENTS IN EPIDERMAL CANCER CELLS

HARRY L . MALECH and THOMAS L . LENTZ

INTRODUCTION

Cytoplasmic microfilaments variously reportedas 40-90 A in diameter have been found in awide variety of cell types . In many cases, theirpresence has been found to accompany suchdiverse cellular movements as cytokinesis (1, 2),cytoplasmic streaming (3), phagocytosis (4-6),extension of microvilli (7), axon elongation (8),morphogenetic movements (9, 10), and activemigration of cells (3, 4, 11-18) . Cytochalasin B,a mold metabolite, has been shown to disruptmicrofilaments while at the same time alteringcellular motile activities (2, 4, 5, 8, 10, 17, 19) .Biochemical studies have demonstrated the pres-ence of proteins resembling muscle actin inphysical, chemical, and structural propertieswithin cells showing motility or contractilityand containing microfilaments (3, 14, 20-26) .These studies provide evidence that microfila-ments composed of an actin-like protein comprise

From the Department of Anatomy, Yale University School of Medicine, New Haven,Connecticut 06510

ABSTRACT

The occurrence and structure of microfilaments in epidermal cancers induced in mice bytreatment with 3,4-benzpyrene were investigated with the electron microscope . Withmalignant change, pleomorphic, undifferentiated cells with a cortical zone of microfila-ments became increasingly abundant . The microfilaments were 40 A in diameter and occu-pied the cortex of the cells beneath the plasma membrane, extended into cell processes, andwere situated in the cores of microvilli . At high magnification, the filamentous areas wereformed by an interconnected meshwork of filaments which in favorable planes had apolygonal arrangement . When exposed to high concentrations of cytochalasin B, the micro-filaments became clumped and moderately disrupted . At the same time, the processes andmicrovilli of the cells were blunted . The structure of these filaments and their sensitivity tocytochalasin B place them in a class of microfilaments believed to be related to cell motility .Their presence in malignant cells may be correlated with the motile, invasive properties ofthese cells .

THE JOURNAL OF CELL BIOLOGY • VOLUME 60, 1974 . pages 4 73-4ôû

a contractile system responsible for cellular motil-ity.

Invasion and replacement of surrounding tissueis one of the primary behavioral characteristics ofmalignant neoplasms that distinguishes them frombenign neoplasms and normal adult animal tis-sues . Because of the high degree of correlation be-tween the presence of microfilaments and activecellular surface activity and movement, it isconceivable that the increased motility of cancercells might be accompanied by the appearance ofcytoplasmic microfilaments . To investigate theoccurrence of filamentous structures in invasivecancer cells, skin cancers were induced in miceby 3,4-benzpyrene and observed at the finestructural level. Newly appearing microfilamentswere tested for sensitivity to cytochalasin B . Thisstudy reveals the appearance of cytochalasin B-sensitive microfilaments in cancer cells . A com-

473

plete description of other cytological changesoccurring in the epidermis during carcinogenesisis presented elsewhere (27) .

MATERIALS AND METHODS

6-wk old female Balb/C mice were obtained from aninbred colony maintained at the Yale UniversitySchool of Medicine by Dr . William U. Gardner.The polynuclear aromatic hydrocarbons 3,4-benzpyrene and 1,2 :3,4-dibenzanthracene (SigmaChemical Co ., St. Louis, Mo.) were employed ascarcinogen and noncarcinogen control, respec-tively (28) . Both are soluble in acetone and bothenter epidermal cells of exposed mouse skin as shownby binding to cellular proteins (29, 30) .Acetone solutions of 0 .57 3,4-benzpyrene or

0.5% 1,2 :3,4-dibenzanthracene were prepared andstored in 1-ml samples in sealed glass vials . Thesewere kept at 4 °C in the dark for the duration of theexperiment. Previous dose-response studies haveshown that application of 100 µg of 3,4-benzpyrene(0.5 0]0 in acetone) three times a week to mouse skinproduces tumors within 10-15 wk (31) . The solutionswere applied externally onto the inner side of thepinna of the mouse's ear, dropwise using a tuberculinsyringe with a filed-down 26G needle (3 drops =100 µg) .

The pinna proved to be especially advantageousas a site of application because it is thin and hairless,and tumors are easily observable grossly . Further-more, the two epidermal surfaces lie within 0 .5 mmof each other, allowing comparison of treated anduntreated surfaces in a single section . 3,4-Benz-pyrene was applied to the right ear of 12 mice . Thesame volume of pure acetone was applied to the leftear of the same mice . With six mice, 1 , 2 : 3 , 4-dibenzanthracene was applied to the right ear andacetone to the left ear . Chemicals were applied threetimes a week for up to 24 wk .

Tissue samples were taken from untreated miceand from treated mice at 6, 12, and 24 wk. Biopsieswere taken of the part of the pinna containing thedesired region of epidermis or tumor . Except in thecase of deeply invasive carcinomas, this proceduredid not require sacrifice of the mouse and allowedmost mice to be followed for the full duration of theexperiment. The observations and figures are repre-sentative of a survey of tissues from at least threemice at each stage of malignant transformation .

Cytochalasin B (Imperial Chemicals Ltd ., Mac-clesfield, Cheshire, England) dissolved in dimethylsulfoxide (25 mg/ml) was added to culture medium199 (unmodified, Earle's base, Grand Island Bio-logical Co ., Grand Island, N . Y.) to produce finalconcentrations of 50-100 µg/ml . Solutions con-taining dimethyl sulfoxide (0.4%) but lacking cyto-chalasin were also prepared . Thin tissue slices of

4 74

THE JOURNAL OF CELL BIOLOGY . VOLUME 60, 1974

tumors were placed in the cytochalasin B mediumor control medium for I h at room temperature .

All tissue samples were cut into small rectangularpieces to allow orientation and fixed for electron mi-croscopy. Tissue samples were fixed for 2 h at 4 ° C incold 3% glutaraldehyde in 0 .05 M sodium cacodylatebuffer (pH 7 .2) . They were then rinsed in cold bufferseveral times and fixed for 2 h in 2% osmium tetrox-ide buffered with 0.05 M sodium cacodylate (pH7.2) at 4 °C. After osmium tetroxide fixation, thetissue samples were washed in cacodylate buffer(pH 7 .2) and then several times in Michaelis buffer(pH 5.0) . The tissues were stained en bloc withaqueous uranyl acetate by placing them in 0.5/0uranyl acetate in Michaelis buffer (pH 5 .0) for 2h at room temperature (32) . The tissues were thenwashed in Michaelis buffer (pH 5.0) and dehy-drated in a graded series of ethyl alcohol . After care-fully orienting the rectangular tissue blocks to permitsectioning in a plane perpendicular to the surfaceof the epidermis, they were embedded in Maraglas(The Marblette Co., Div. of Allied Products Corp.,Long Island City, N . Y .) .

Thick sections were cut on a Sorvall MT-2 micro-tome (Ivan Sorvall, Inc ., Newtown, Conn .) withglass knives, stained on glass slides with 1 % toluidineblue, and surveyed for areas to be examined at thefine structural level . Thin sections cut with glass ordiamond knives were stained with lead citrate (33)and examined with an Hitachi 8B electron micro-scope . Measurements were performed with a 7Xocular micrometer on micrographs with a magnifi-cation of 100,000 or above. The magnification ofthe microscope was calibrated with a carbon gratingreplica grid .

RESULTS

Light Microscope Structure of Tumors

Exposure of the pinna to benzpyrene resultsin the appearance of focal areas of hyperplasia at6 wk, papillomas as 12 wk, and invasive carci-nomas by 24 wk. The carcinomas underlie hyper-plastic epidermis and invade deep layers of con-nective tissue down to the central layer of cartilage .The boundary between the tumor and underlyingconnective tissue is not readily evident in manyareas. From the main tumor mass, tongues ofcarcinoma cells project into the subcutaneoustissues and surround connective tissue elements,blood vessels, nerves, and muscle . Cells in contactwith the dermis are often elongated with proc-esses extending into the connective tissue (Fig . 1) .Intercellular spaces are present between indi-vidual cells and in the central regions of the tumor

FIGURE 1 Periphery of an invasive carcinoma after 24 wk of exposure to 3,4-benzpyrene . Note thevariation in size, shape, and staining characteristics of individual cells . Many cells at the edge of thetumor project long cytoplasmic processes (arrows) into the adjacent connective tissue (C), while cellsdeeper within the tumor tend to be round or polygonal in shape . Fibroblasts and macrophages occur inthe connective tissue . A mitotic cell is present in the lower left portion of the micrograph . X 800.

they are bridged by thin cytoplasmic extensions .The cells vary in their staining characteristics,some being intensely stained with toluidine blue .Nuclei are often large with multiple nucleoli, andmitotic figures are common .

Fine Structure of Cancer Cells

There is variation in the cytology of the cellscomprising the tumors, although the majorityof cells fit into one of two categories . Central re-gions of tumors are composed largely of polygonal,closely packed cells which appear to be moredifferentiated, containing short bundles of tono-filaments, moderate amounts of rough-surfacedendoplasmic reticulum, and large numbers ofribosomes organized into polysomal clusters .Many small microvilli and pseudopodial proc-

esses arise from the surface of central cells . Thesecond type of cell is more common at the periph-ery of the tumors, especially in the leading edgesof tumor infiltration . These cells are highly irreg-ular in shape with large processes and occasionalmicrovilli (Figs . 2, 3) . Free cytoplasmic ribosomesare common and often occur singly . There is arelative paucity of rough-surfaced endoplasmicreticulum . Tonofibrils are very scarce, but bundlesof microfilaments are abundant in these cells .The tonofibrils are reduced to a few short electron-dense, spike-like structures which are often theonly means of determining the cell's epithelialorigin (Fig . 2) . The tonofilaments comprisingthe tonofibrils are approximately 100 A in diam-eter (50 filaments, mean diameter 105 .9 t SD31 .6 A) .

Intercellular spaces between the less differen-

H. L. MALECH AND T . L. LENTZ Microfilaments in an Epidermal Cancer

475

FIGURE 2 Part of a carcinoma cell at the edge of an invasive squamous cell carcinoma . 24-wk exposureto 3,4-benzpyrene. Exposed en bloc to 0 .4cja dimethyl sulfoxide in vitro for 1 h . The cell extension issituated in the dermis and is irregular in outline with several long, slender microvilli . There is no base-ment lamina although flocculent, amorphous material adheres to the cell surface . Microfilaments (Mf)occur in the peripheral cytoplasm . Some of the microfilament bundles are oriented in the long axis of theprocess (arrows) . Also within the cell are microtubules (311), free ribosomes, and some cisternae of rough-surfaced endoplasmic reticulum . Tonofibrils (T) are scarce, occurring as small short bundles of tono-filaments. X 37,000 .

tiated, peripheral cells are large and irregular . show a reduced number of desmosomes and otherMasses of moderately dense, amorphous material specialized junctions in comparison to normaloccupy the intercellular spaces (Figs. 2-4) . Occa- epidermal cells . A few small hemidesmosomessional desmosomes connect the processes of ad- may occur along the surface of the cell facing thejacent cells (Fig . 3) but in general these cells connective tissue (Fig . 3) . A distinct basement

476

THE JOURNAL OF CELL BIOLOGY • VOLUME 60, 1974

FIGURE 3 Carcinoma cells at the edge of an invasive carcinoma . The cell surfaces are irregular with vil-lous extensions and processes . Note that microfilaments (31f) are present both adjacent to the plasmamembrane as a long cortical band and within the microvilli . There is no distinct basement lamina at thejunction with the connective tissue, but amorphous material (A) adheres in patches to the membrane . Afew small hemidesinosomes (II) and desmosolnes (D) are present . X 47,000.

lamina does not delineate the tumor from sur-rounding connective tissue . However, fine floccu-lent material may occur in patches along thecell surface or form an irregular thin layer on thesurface (Fig . 3) .

Microfilaments

Microfilamentspleomorphic undifferentiated cells at the periph-ery of the tumor (Figs . 2-4) . A small number of

fill most of the cortex of the

similar cells containing microfilaments werefirst observed in papillomas, but have not beenseen in normal or hyperplastic epidermis . Cellsin the central regions of cancers contain few micro-filaments . Commonly, the filaments run in a bandabout 0.1-0.2 µm thick along the periphery ofthe cells beneath the plasmalemma (Figs . 3, 4) .Filaments extend into the cell processes and occupythe cores of microvilli (Fig . 3) . In some places,the filaments appear to be very closely associatedwith the plasma membrane (Fig . 5) . Regions con-

H. L. MALECH AND T. L . LENTZ Microfilaments in an Epidermal Cancer

477

FIGURE 4 a Cortex and portion of the nucleus (N) of a carcinoma cell . Exposed en bloc to 0 .470 di-methyl sulfoxide in vitro for I h . 40 A rnicrofilaments (MD are oriented parallel to the plasma membranein the cortex of the cell . There is no basement lamina separating this cell from the connective tissue (C) .Flocculent material occurs extracellularly. X 47,000 .

FIGURE 4 b High magnification of the blocked portion of Fig . 4 a . Transverse connections can be seenbetween individual longitudinal filaments . X 133,000 .

478

FIGURE 5 Very high magnification of the edge of acarcinoma cell showing a small number of 40 A micro-filaments (M11f) adjacent to the plasma membrane (PM) .The interconnections between microfilaments form apolygonal network. The centers of the polygons aremarked by stars . The plasma membrane above themicrofilanients is sectioned obliquely, and in this regionthe microfilaments appear closely associated with orattached to the membrane . Just below the microfila-ments is a tonofibril (T) . X 294,000 .

taining microfilaments are generally devoid ofother organelles.

Measurement of 100 distinct filaments yielded amean diameter of 37.2 ± SD 12.7 A . Beadlikedensities along the filaments are of larger diam-eter. Low power micrographs give the impressionthat the bundles are composed of parallel arrays

of straight filaments (Fig . 4 a) . However, higher

magnification of different planes of section showsthat the organization of the filaments is more

complex. In sections parallel to a main bundle,filaments appear to run longitudinally, but are

connected to adjacent filaments by cross bridges(Fig . 4 b) . In sections passing obliquely or trans-

versely through the bundles, the interconnectionsbetween filaments are especially apparent . The

filamentous areas thus appear to be comprised

by a meshwork of interconnected filaments . In

some places, a pattern of uniform and delicatepolygons is formed (Fig . 5) .

Effects of Cytochalasin B

In vitro treatment of thin slices of carcinomatissue with 100 µg/ml of cytochalasin B resultsin morphological alteration of the microfilamentsand some of the cell processes . Cytochalasin Bcauses clumping of the long cortical bundles ofmicrofilaments, thickening of filaments, and theappearance of many small focal, irregularlyshaped densities within the disrupted filamentousmaterial (Fig . 6). In addition, the oriented

appearance and polygonal structure of the micro-filament bundles become less apparent . The cellsurface is less irregular and the processes andmicrovilli seen on the untreated cells are bluntedand reduced in number. No other cellular struc-tures are markedly affected by cytochalasin .The effects of 50 µg/ml of cytochalasin are not soapparent or consistent as the effects of 100 µg/ml .

The in vitro controls exposed to 0.4 0]0 dimethyl

sulfoxide are indistinguishable from the carcinomacells fixed immediately after excision (Figs . 2, 4) .Microfilaments and cell processes are unchanged .

1,2 :3 ,4-Dibenzanthraeene and

Acetone Exposure

At 24 wk of exposure to dibenzanthracene theepidermis is hyperplastic, being two to three timesnormal thickness . However, no papillomas or car-cinomas appear, even after I y . Cell contactrelationships and basement lamina are unaltered .

The cytological structure of the cells is almost in-distinguishable from that of the untreated epider-mis, and no microfilaments are seen . Epidermisexposed to acetone for 24 wk is indistinguishablefrom normal at the light and electron microscopelevels .

DISCUSSION

Various theories have been proposed to explainthe origin of the motive force responsible for thespread of tumor cells into adjacent normal tissue .One hypothesis is that increased pressure withinthe growing tumor provides the force for a passiveintrusion of tumor cells into surrounding tissues(34) . This, along with inflammation and per-haps release of proteolytic enzymes or cytotoxicsubstances by tumor cells (35), could provide

H. L. MALECH AND T. L. LENTZ Microfilasnents in an Epidermal Cancer 479

the necessary conditions for penetration of adja-cent tissues.

On the other hand, the most widely acceptedtheory of local tumor invasion suggests that activemotile activity on the part of the tumor cells isresponsible for tumor invasion (36, 37) . It hasbeen shown that placing a malignant tumor be-side a normal tissue in culture results in invasionof the normal tissue by tumor cells (38, 39) .Furthermore, Wood et al . (40), studying in vivomovement of cells using a rabbit ear chamberand time-lapse cinemicrography, found thatunder the same conditions, carcinoma cells andpolymorphonuclear leukocytes moved at similarrates while normal epidermal cells did not moveat all . In addition, the carcinoma cells continu-ally shifted arrangements with respect to oneanother. These studies indicate that the actualforce for local invasion is generated within theindividual cells themselves .

4 80

FIGURE 6 Portion of a carcinoma cell exposed en bloc to cytochalasin B, 100 µg/ml, in vitro for I h .Compare this cell with that seen in Fig. 2 . Within this cell, the microfilaments (Mf) are clumped and dis-rupted, with the usual oriented appearance only vaguely evident . The microvillous processes usuallyseen are blunted and reduced in number. T, tonofibrils . X 33,000 . Inset . High magnification of the blockedportion . The delicate network of microfilaments is largely destroyed . Small, irregularly shaped, electron-dense particles are scattered within the disrupted bundles of microfilaments . X 110,000.

THE JOURNAL OF CELL BIOLOGY • VOLUME 60, 1974

It was found in the present study that duringmalignant transformation induced by benzpyrene,epidermal cells lose their characteristic 100 Atonofilaments and acquire a new population ofthinner microfilaments . Microfilaments were notseen in normal epidermal cells or hyperplasticcells, but were abundant in the cells at the lead-ing edges of invasive cancers . Because of thehigh degree of correlation between the presenceof microfilaments and cellular motility (see In-troduction), the presence of microfilaments inthe pleomorphic cells of invasive carcinomas mayrepresent the mechanism providing the force forthe motile activities of these cells .The microfilaments within the cancer cells

are similar to those described in a variety of othercells (see Introduction) . Close association with theplasma membrane (6, 8, 11, 17) and polygonalconfigurations (8) have been described . The fila-ments of cancer cells seem slightly smaller than

those of most other cells . Because several distinctclasses of microfilaments occur in cells (17, 18), func-tional comparisons on the basis of size alone should

be viewed with caution until the precise functionsof the different types of filaments in each cell typeare determined .

The microfilaments in the pleomorphic, un-differentiated cells were disrupted by exposure tohigh doses of cytochalasin B . Cytochalasin B hasbeen shown to disrupt cytoplasmic microfilamentswhile at the same time altering cell behavior (see

Introduction) . Cytochalasin B may act directlyon microfilaments, causing their disassembly (10),or it may block cell motility by altering muco-

polysaccharide synthesis (41) . While the mecha-

nism of action of cytochalasin B is unknown, thesensitivity of microfilaments in the cancer cellsto cytochalasin suggests that these filaments aresimilar to those associated with motility in othercells. The effect of cytochalasin B on the motilityof the cancer cells was not studied, although itwas observed that cellular extensions and villiwere reduced in number and blunted or retracted

after treatment. Its effect on the microfilamentsystem provides a basis for investigation of theeffects of cytochalasin B on tumors characterized

by invasion and migration of cells .

This work was supported by grants from the Na-tional Cancer Institute, National Institutes of Health,United States Public Health Service (TICA-5055),and from the National Science Foundation (GB-20902) .

Received for publicationform 25 October 1973.

REFERENCES

1 .

2 .

3 .

4 .

27 August 1973, and in revised

SzoLLOSt, D. 1970 . Cortical cytoplasmic fila-ments of cleaving eggs . A structural elementcorresponding to the contractile ring. J. CellBiol. 44 :192 .

SCHROEDER, T. E. 1972 . The contractile ring .II . Determining its brief existence, volumetricchanges, and vital role in cleaving Arbaciaeggs . J. Cell Biol. 53 :419.

POLLARD, T. D., E. SHELTON, R . R. WE[HING,and E. D. KORN . 1970 . Ultrastructural char-acterization of F-actin isolated from Acantha-moeba castellanii and identification of cyto-plasmic filaments as F-actin by reaction withrabbit heavy meromyosin . J . Afol . Biol . 50 :91 .

ALLISON, A . C., P. DAVIES, and S . DEPETRIS .1971 . Role of contractile microfilaments in

macrophage movement and endocytosis.Nat. New Biol. 232:153 .

5. WAGNER, R ., 111 . ROSENBERG, and R. ESTENSEN .1971 . Endocytosis in Chang liver cells . Quan-titation by sucrose- 3H uptake and inhibitionby cytochalasin B . J. Cell Biol. 50 :804 .

6. REAVEN, E . P., and S. G. AXLINE . 1973 . Sub-plasmalemmal microfilaments and microtu-bules in resting and phagocytizing cultivatedmacrophages . J . Cell Biol . 59 :12 .

7 . TILNEY, L . G., and R . R. CARDELL, JR . 1970.Factors controlling the reassembly of themicrovillous border of the small intestine ofthe salamander . J. Cell Biol. 47 :408 .

8. YAMADA, K. M ., B . S. SPOONER, and N . K. WES-SELLS . 1971 . Ultrastructure and function ofgrowth cones and axons of cultured nervecells . J. Cell Biol . 49 :614 .

9. CLONEY, R. A. 1966 . Cytoplasmic filaments andcell movements : Epidermal cells during ascid-ian metamorphosis . J . Ultrastruct. Res . 14 :300.

10. WESSELLS, N . K., B. S. SPOONER, J . F. ASH,n1. 0. BRADLEY, A1. A. LUDUENA, E. L.TAYLOR, J. T. WRENN, and K. AI. YAMADA .1971 . Microfilaments in cellular and develop-mental processes. Science (Wash. D . C.) . 171 :135 .

11 . BUCKLEY, I . K., and K . R. PORTER . 1967 . Cyto-plasmic fibrils in living cultured cells . Alight and electron microscope study . Proto-plasma . 64 :349 .

12. FRANKS, L. M., P. N. RIDDLE, and P. SEAL.1969. Actin-like filaments and cell movementin human ascites tumor cells. An ultrastruc-tural and cinemicrographic study . Exp . CellRes. 54 :157 .

13. GOLDMAN, R. D., and E . A. C. FOLLETT . 1969.The structure of the major cell processes ofisolated BHK21 fibroblasts . Exp. Cell Res.57 :263 .

14. NACHMIAS, V . T., H. E. HUXLEY, and D . KESs-LER. 1970 . Electron microscope observationson aetomyosin and actin preparations fromPhysarum polycephalum, and on their interac-tion with heavy meromyosin subfragment Ifrom muscle myosin . J. Afol. Biol . 50 :83 .

15. ABERCROMBIE, AI ., J . E. A1. HEAYSMAN, andS. M. PEGRUM. 1971 . The locomotion offibroblasts in culture . IV. Electron microscopyof the leading lamella . Exp . Cell Res. 67 :359.

16. KRAWCZYK, W. S . 1971 . A pattern of epidermalcell migration during wound healing . J. CellBiol . 49:247.

17. SPOONER, B . S., K. M. YAMADA, and N. K .WESSELLS . 1971 . Microf laments and celllocomotion . J. Cell Biol. 49:595 .

18. AIcNUTT, N . S ., L . A . CULP, and P. H. BLACK.1973. Contact-inhibited revertant cell lines

It. L . MALECII AND T . L . LENTZ Jlicrofilaments in an Epidermal Cancer

481

isolated from SV40-transformed cells . IV.Microfilament distribution and cell shape inuntransformed, transformed, and revertantBalb/C 3T3 cells. J. Cell Biol . 56 :412 .

19. MALAWISTA, S . E ., J . B. L. GEE, and K. G .BENSCH . 1971 . Cytochalasin B reversibly in-hibits phagocytosis : Functional, metabolic, andultrastructural effects in human blood leuko-cytes and rabbit alveolar macrophages . YaleJ. Biol . 41ed . 44 :286 .

20. ISHIKAWA, H ., R. BISCHOFF, and H . HOLTZER .1969. Formation of arrowhead complexeswith heavy meromyosin in a variety of celltypes . J . Cell Biol . 43 :312.

21 . POLLARD, T . D., and S . ITO . 1970. Cytoplasmicfilaments of Amoeba proteus. I . The role offilaments in consistency changes and move-ment . J. Cell Biol . 46 :267 .

22. MORGAN, J . 1971 . Microfilaments from amoebaproteins . Exp . Cell Res. 65 :7 .

POLLARD, T. D., and E. D. KORN. 1971 . Fila-ments of Amoeba proteus . II. Binding of heavymeromyosin by thin filaments in motile cyto-plasmie extracts . J. Cell Biol. 48 :216 .

24. TILNEY, L. G., and M . MOOSEKER. 1971 . Actinin the brush-border of epithelial cells of thechicken intestine . Proc . Nall . Acad. Sci . Lt. S. A .68 :2611 .

25. COMLY, L. T. 1973 . Microfilaments in Chaoscarolinensis . Membrane association, distribu-tion, and heavy meromyosin binding in theglycerinated cell . J . Cell Biol . 58 :230 .

PERDUE, J . F. 1973 . The distribution, ultra-structure, and chemistry of microfilaments incultured chick embryo fibroblasts . J . CellBiol . 511 :265 .

MALECH, H. M. 1972. Cytological changesduring chemical carcinogenesis in mouseepithelium. The appearance of cytochalasinB-sensitive microfilaments . MD. Thesis, YaleUniversity School of Medicine, New Haven,Conn .

BARRY, G ., J . W. COOK, G. A. D. HASLEWOOD,C. L. HEWETT, 1 . HIEGER, and E . L. KENNA-WAY . 1935. The production of cancer by purehydrocarbons . Proc . R . Soc. Lond. B Biol. Sci .117 Part III . :318 .

29. MILLER, E. C. 19.51 . Studies on the formation

23 .

26 .

27 .

28 .

4 8 2

ToE .IOI'KNAI, OF CELL BIOLOGY . S'oLruE 60, 1974

of protein-bound derivatives of 3,4-benz-pyrene in epidermal fraction of mouse skin .Cancer Res . 11 :100.

30. HEIDELBERGER, C. 1964 . Studies on the molec-ular mechanism of hydrocarbon carcinogene-sis . J. Cell. Comp. Physiol . 64(Suppl . 1) :129 .

31 . WYNDER, E . I .., J . SPRANGER, and lvl . FARK .1960. Dose-response studies with benzo(a)-pyrene. Cancer . 13 :106 .

32. FARQUHAR, M . G., and G . E. PALADE . 1965 .Cell junctions in amphibian skin . J. Cell Biol.26 :263 .

33. REYNOLDS, E. S. 1963 . The use of lead citrateat high pH as an electron-opaque stain inelectron microscopy. J . Cell Biol . 17 :208.

34. HAMPERL, H . 1967 . Early invasive growth asseen in uterine cancer and the role of thebasal membrane . In Mechanisms of Invasionin Cancer. P . Denoix, editor . Springer-Verlag,Berlin . 17-25.

35. SYLVEN, B. 1967 . Some factors relating to theinvasiveness and destructiveness of solidmalignant tumours . In Mechanisms of In-vasion in Cancer. P . Denoix, editor . Springer-Verlag, Berlin . 47-60 .

36. ABERCROMBIE, A1 ., and E . J . AMBROSE. 1962 .The surface properties of cancer cells : Areview . Cancer Res. 22 :525 .

37. ROSENAU, W. 1969. The nature and mecha-nism of metastasis . Oncology (Basel) . 24 :21 .

38. WOLFF, E ., and N . SCHNEIDER . 1957 . La cultured'un sarcome de souris sur des organes depoulet explantes in vitro . Arch . Anat . Microsc ..%Iorphol. Exp . 46 :173 .

39. WOLFF, E., and E. WOLFF . 1'958 . Les resultatsd'une nouvelle methode de culture de cellulecancereuses in vitro . Rev . Fr. Etud . Clin . Biol .3 :945 .

40. WOOD, S . JR ., R . R. BAKER, and B. MARZOCCHI .1967. Locomotion of cancer cells in vivo com-pared with normal cells . In Mechanisms ofInvasion in Cancer . P . Denoix, editor . Sprin-ger-Verlag, Berlin . 26-30 .

41 . SANGER, J . W., and H . HOLTZER . 1972 . Cyto-chalasin B : Effects on cell morphology, celladhesion and Inucopolysaccharide synthesis .Proc . Nail . Acad . Sci. U. S. A . 69 :253 .

481.pdfpage 1page 2