J. Cell Sci. 31, 341-353 (1978) 341Printed in Great Britain © Company of Biologists Limited igyS

CELLULAR ACTIN AND JUNCTION

FORMATION DURING REAGGREGATION OF

ADULT RAT HEPATOCYTES INTO

EPITHELIAL CELL SHEETS

A. MIETTINEN, I. VIRTA.NEN* AND E. LINDERDepartment of Bacteriology and Immunology,

*Third Department of Pathology and Department of Electron Microscopy,University of Helsinki, Helsinki, Finland

SUMMARY

Aggregation of adult rat hepatocytes, isolated by the collagenase perfusion technique, wasstudied by ultrastructural methods and the indirect immunofluorescence technique usinganti-actin antibodies. In a primary culture the cells rapidly made contact with each other byfilopodia-like structures, as seen by scanning electron microscopy. In a few hours this led tostable adhesion of the cells. No identifiable junction formation occurred during the first17 h in culture. Within 48 h the cells had formed epithelial cell sheets with junctional complexesconsisting of tight junctions, bile canaliculus-like structures and zonula adhaerens-typejunctions.

The distribution of cytoplasmic actin fluorescence remained homogeneous in the contactingcells during the first 24 h in culture, as seen with anti-actin antibodies by the indirect immuno-fluorescence technique. The first short, fluorescent actin filaments appeared in the peripheryof the developing lamellipodia of the spreading cell islands. In organized epithelial cell sheetsthese filaments were seen as long fibres ending at the perinuclear region of the marginalcells. In the submarginal cells fluorescent actin fibres were distinct at the junctional regionsof the cells. This filamentous fluorescence seemed to extend throughout the entire cell sheetin an organized manner and correspond to the apical layer of parallel microfilaments seen intransmission electron microscopy. Our results suggest that filamentous actin plays a role inthe contact-induced spreading of the cells and in the maintenance of the internal organizationof the epithelial cell sheets.

INTRODUCTION

Contacts between epithelial cells in culture induce stable adherence of the cellsto each other, as well as polarization and spreading of the adherent cells (Vaughan& Trinkaus, 1966; Garrod & Steinberg, 1975; Middleton, 1976). This leads to theorganization of epithelial cell sheets in which the mutually strongly cohesive cellsare linked by specialized junctional complexes (Middleton, 1973; Garrod & Stein-berg, 1975; Middleton & Pegrum, 1976). In these cell sheets both the locomotoryactivity and the adhesion of the sheet to the growth substratum are confined to themarginal cells (Vaughan & Trinkaus, 1966; DiPasquale, 1975).

Actin and other contractile proteins are known to participate in surface movements,

Address for correspondence: Aaro Miettinen, M.D., Department of Bacteriology andImmunology, University of Helsinki, Haartmaninkatu 3, SF-oozox> Helsinki 29, Finland.

342 A. Miettinen, I. Virtanen and E. Linder

contraction and locomotion of cultured cells (Goldman, Yerna & Schloss, 1976;Nicolson, 1976; Pollard, Fujiwara, Niederman & Maupin-Szamier, 1976). Theorganization of actin in cells seems to reflect their locomotory activity (Pollard &Fujiwara, 1976) and anchorage to the growth substratum (Pollack & Rifkin, 1975;Willingham et al. 1977). These results have been obtained mainly by studyingnormal and transformed fibroblasts, using specific antibodies against contractileproteins and the indirect immunofluorescence technique, but relatively little isknown of the role of these proteins in epithelial cells.

Epithelial cells in vitro form multicellular aggregates and regain some of theorganization of epithelial organs in vivo (cf. Ponton, 1975; Trinkaus, 1976; Vasiliev& Gelfand, 1976). It is not known how, after contact with like cells, epithelial cellsare able to maintain their adherence to each other in spite of the tension created bythe contact-induced spreading of the cells (Middleton, 1976; Middleton & Pegrum,1976). It is possible that the junctional complexes develop rapidly enough to stabilizethe intercellular adherence (Middleton & Pegrum, 1976). On the other hand, thesurface adherence of the contacting plasma membranes as such, might be firmenough to keep the cells together (cf. Obrink, Kuhlenschmidt & Roseman, 1977).

In the present study we used primary cultures of adult rat hepatocytes to investigateactin organization during the reaggregation of epithelial cells. The findings obtainedwith anti-actin antibodies and indirect immunofluorescence were correlated withchanges in the ultrastructure of the cells, as seen by scanning and transmissionelectron microscopy.

MATERIALS AND METHODS

Isolation of hepatocytes

Hepatocytes were isolated from adult, female Sprague-Dawley rats as described earlier(Virtanen & Wartiovaara, 1976), using the collagenase perfusion technique and isolationmedia of Seglen (1976). Briefly, rats were anaesthetized by an intraperitoneal injection ofpentobarbitone (50 mg/kg) and the livers perfused via vena portae starting with the Ca1+-freeperfusion buffer and subsequently with the collagenase buffer containing Ca1+ and 05 %collagenase (Grade I, Sigma, St Louis, Mo., U.S.A.). The cells were then suspended in thesuspension buffer, filtered through 2SO-/tm and ioo-/Jm nylon mesh (Schweitzerische Seiden-gazefabrik Ag, Zurich, Switzerland), incubated with reciprocating shaking at 37 °C for 30 minand filtered again through a ioo-/tm nylon mesh. The liver parenchymal cells were purifiedby 3 subsequent centrifugations in the washing buffer at 100g for 4 min. The final cell pelletswere gently suspended in the culture medium used, and filtered again through 100- and50-/tm nylon mesh. The viability of the suspended cells was estimated by the dye-exclusiontest, using Trypan blue. The cells were used for culture only if the viability exceeded 95 %.

Cell cultures

The cells were cultured at 37 °C in a humidified atmosphere (95 % air, 5 % COS) on plasticPetri dishes (Falcon plastics, Oxnard, Ca., U.S.A.). The dishes and the glass coverslips werecoated with collagen from rat tail according to the method of Michalopoulos & Pitot (1975).The culture medium was HAM F-12 (Flow Laboratories, Rockville, Md., U.S.A.) supple-mented with 15% inactivated (at 56 °C for 30 min) foetal bovine serum (Flow), io/<g/mlcrystalline pork insulin (25 i.u./mg, Medica Pharmaceutical Company, Ltd., Helsinki, Finland),2 /tg/ml corticosterone (Merck AG, Darmstadt, F.R.G.), 100 i.u./ml penicillin and 5O/:g/ml

Actin and junctions in cultured hepatocytes 343

streptomycin. Usually 1 x 10* cells in 4 ml of culture medium were seeded per dish (diameter60 mm). The medium was changed 4 h after plating and thereafter daily.

Anti-actin antibodies

The serum containing smooth muscle antibodies (SMA) was obtained from a patient(R.G.) with chronic active hepatitis and shown to be specific for actin (Kurki, Linder,Miettinen & Alfthan, 1977a; Kurki, Linder, Virtanen & Stenman, 19776). The SMA titreof the antiserum on cryostat sections of rat kidney and stomach was 1/5000 (Kurki et al.1977a). Experimental anti-actin antibodies (SMA titre 1/100) were made by immunizingrabbits as described by Trenchev & Holborow (1976) with bovine skeletal muscle actin,purified by the method of Spudich & Watt (1971). IgG fractions were made from the antisera,and the anti-actin specificity of these fractions was checked in double immunodiffusionexperiments (Norberg, Lidman & Fagraeus, 1975) and by absorption with purified bovineskeletal muscle actin (Kurki et al. 1977a). The SMA titres of the IgG fractions were estimatedand the fractions were tested in doubling dilutions with cultured human embryonic skinfibroblasts for their ability to give the typical filamentous actin staining in immunofluorescence(cf. Kurki et al. 19776). The end-point titres of the human IgG fraction and the rabbit IgGfraction were on fibroblasts 1/128 and 1/32, respectively. Our results were obtained withhuman IgG anti-actin antibodies (dilution 1/50) and later confirmed with rabbit IgG anti-actin antibodies.

Immunofluorescence microscopy

For immunofluorescence, hepatocytes grown on collagen-coated coverslips were washedwith warm Ham F-12 solution and fixed in dry acetone at —20 °C for 15 min. The coverslipswere then washed in phosphate-buffered saline (PBS), pH 7 2 and reacted with the IgGfractions of the antisera at room temperature for 30 min. After washing in PBS the coverslipswere overlaid with fluorescein isothiocyanate (FITC)-conjugated sheep anti-human IgG(F/P molar ratio 2 ,9 ; National Bacteriological Laboratory, Stockholm, Sweden) or withFITC-conjugated sheep anti-rabbit IgG (F/P molar ratio 3,2; Linder & Miettinen, 1976)for 30 min. The coverslips were washed and mounted in buffered glycerol, pH 8-5. For lightmicroscopy a Zeiss Universal microscope was used. For immunofluorescence studies it wasequipped with a Zeiss III RS epi-illuminator, excitation filters KP 500 and KP 490, a dicroidmirror F T 510, an emission filter LP 520 and a high-pressure mercury lamp.

Electron microscopy

For electron microscopy the cultures were washed with warm HAM F-12 solution andfixed in 25 % glutaraldehyde, buffered with 0 1 M Na-cacodylate (pH 72) at room temperaturefor 60 min. For transmission electron microscopy (TEM), the cultures grown on smallculture dishes were postfixed in 1 % osmium tetroxide, buffered with o-i M phosphate buffer(pH 7'2), for 60 min. In some experiments the specimens were stained en bloc with 1 %uranyl acetate in distilled water for 60 min. Thereafter the specimens were dehydrated inethanol and embedded in Epon 812 omitting the propylene oxide stage. Thin sections werecut either vertically or horizontally from the cultures and re-embedded in flat moulds. Post-staining was done with uranyl acetate and lead citrate and the specimens were examined witha Jeol 100B electron microscope at an accelerating voltage of 80 kV.

For scanning electron microscopy (SEM), fixed cultures grown on small coverslips weredehydrated in ethanol and amyl acetate and dried in an Aminco critical-point drying apparatuswith liquid CO,. The dried specimens were covered with a thin layer of carbon and gold andexamined with a Jeol JSM-U3 scanning electron microscope at an accelerating voltage of20 kV.

344 A. Miettinen, I. Virtanen and E. Linder

Fig. i . Scanning electron micrograph of isolated rat liver cells. The entire cell surfaceis densely studded with microviUi without apparent surface specializations, x 3500.

Fig. 2. In thin sections, isolated rat liver cells show typical features of hepatocytessuch as glycogen (g) and microbodies (ntb). Scattered lamellae of rough endoplasmicreticulum are seen, n, nucleus, x 6000.

Fig. 3. After 4 h of culture the cells are still densely covered with microviUi andhave short lamellar extensions at their base. Straight filopodia-like structures (arrow)are seen between the cells (inset), x 3500; inset x 14000.

RESULTS

Isolated hepatocytes

After isolation, adult rat hepatocytes were rounded, densely studded with micro-viUi and without apparent surface specializations (Fig. 1). Most of the cells weresingle, but had a marked tendency to form aggregates in suspension. In thin sections

Actin and junctions in cultured hepatocytes

Fig. 4. After 10-14 h ' n culture the cells, already often in cords of cells, are morespread out, as seen by scanning electron microscopy, x 1000.

Fig. 5. In indirect immunofiuoresccnce with anti-actin antibodies a diffuse cytoplasmicfluorescence is seen in hepatocytes after 10-14 h ' n culture, x 600.

Fig. 6. Within 24 h in culture the cells have formed epithelial islands where theflattened marginal cells have short lamellipodia-like structures (arrow) at theirfree edges, x 600.

Fig. 7. The first fluorescent actin fibres (arrow) are seen in the periphery of thedeveloping lamellipodia of the marginal cells after 24 h of culture, x 2000.

the cells had typical features of hepatocytes such as glycogen granules and micro-bodies, but a somewhat disorganized cytoplasmic structure (Fig. 2). Only homo-geneous cytoplasmic fluorescence could be seen in these cells in indirect immuno-fluorescence with anti-actin antibodies.

A. Miettinen, I. Virtanen and E. Linder

Fig. 8. After 48 h in culture the hepatocytes have organized into large cell sheets. Inclefts of the cell sheets the marginal cells have long lamellipodia (/). x 700.Fig. 9. The cells of the developed epithelial cell sheets are tightly apposed withdistinct junctional ridges (r) between cells, as seen in scanning electron microscopy,x 2000.

Initial reaggregation

After plating, most of the cells rapidly adhered to the growth substratum andaggregated into clumps of a few cells. After 4 h in culture the cells were still denselycovered with microvilli, but already had thin lamellae at their base, as seen in SEM(Fig. 3). Between the cells straight filopodia-like structures, occasionally several /imin length, were seen. At this stage of the culture the cells had reorganized their cyto-plasmic structure, as indicated by the lamellar organization of the rough endo-

Actin and junctions in cultured hepatocytes 34?

Fig. 10. In the developed lamellipodia of the marginal cells long fluorescent actinfibres are seen, x 11 coo.Fig. i i . The fluorescent actin fibres, seen in lamellipodia of the marginal cells, oftenappeared to terminate at the perinuclear area of the cells (arrows), x 2800.Fig. 12. In the epithelial cell sheets (after 48 h in culture), filamentous actin fluorescenceis especially distinct at the junctional regions (arrows) of the cells and appears toextend throughout the cell sheets in an organized manner, x 1000.

plasmic reticulum. Only homogeneous fluorescence could be seen in cells with anti-actin antibodies in indirect immunofluorescence. This corresponded to the lack ofbundles of microfilaments in thin sections.

348 A. Miettinen, I. Virtanen and E. Linder

Formation of epithelial cell sheets

After 10-14 h m culture the hepatocytes, often in long cords of cells, were morespread out, although their free surface was still covered with microvilli (Fig. 4).Cytoplasmic actin fluorescence was diffuse (Fig. 5). After 24 h in culture most ofthe cells had formed small islands where the flattened marginal cells had shortlamellar extensions at their edges (Fig. 6). The first fluorescent actin fibres weredetected in the periphery of these lamellae (Fig. 7).

After 48 h in culture the cells had organized into large cell sheets, often extendingover the entire growth substratum. In the clefts of the cell sheets some of themarginal cells had long lamellipodia, as seen in Fig. 8. The cells in the sheets weretightly apposed, with distinct junctional ridges between the cells (Fig. 9). The freesurface of most of the cells was smooth, but some cells with microvilli were alsodetected.

In indirect immunofluorescence with anti-actin antibodies the leading lamellaeof the marginal cells had at this stage long, parallel fluorescent fibres (Fig. 10). Theseoften seemed to end at the perinuclear area of the cells (Fig. 11). In the submarginalcells filamentous actin fluorescence was especially bright in the junctional areas ofthe cells. This fluorescence seemed to extend throughout the entire cell sheet in anorganized manner, as seen in Fig. 12.

function formation

During the first 17 h in culture no distinct junctional structures were seen in thinsections of the aggregated cells even after thorough examination. Instead, the cellswere closely apposed over the whole lengths of their lateral surfaces with an inter-membrane distance of ca. 20 nm (Fig. 13). Small membrane vesicles were regularlyseen close to the cytoplasmic side of the apposed membranes (Fig. 13, inset). After24 h in culture, junctional structures of intermediate type were seen apically in thecells (Fig. 14). Short bundles of microfilaments seemed to terminate at these areasof the plasma membranes. In the junctional region, a typical microvillous structure

Fig. 13. During the first 17 h in culture no junctional structures are seen in thinsections. Occasional densities are detected apically at the cytoplasmic surface of theapposed plasma membranes. In the junctional region apical microvillous structures(arrow) are consistently seen to bridge the intercellular space, x 480(50. Inset: at thisstage the cells are closely apposed with an intercellular distance of ca. 20 nm, andsmall vesicles are detected regularly close to the cytoplasmic surface of the apposedcells, x 26000.

Fig. 14. The first junctions of intermediate type (arrow) are seen after 24 h inculture apically between the cells. Bundles of microfilaments seem to attach atthese regions. The typical microvillous structure is seen to close the intercellularspace (double arrow), n, nucleus, x 20000.Fig. 15. In the developing epithelial cell sheets bile canaliculus-like formations (be),surrounded by tight junctions (arrows), are seen, x 36000. Inset, junctions ofzonula adhaerens-type are also detected between the cells. Note the cytoplasmicfilaments (/) at this stage of culture, x 45000.

Actin and junctions in cultured hepatocytes

\

349

14

A. Miettinen, I. Virtanen and E. Linder

c~Fig. 16. A layer of microfilaments arranged in parallel and in bundles (arrows) isalways seen at 48 h of culture apically in the submarginal cells of the cell sheets, n,nucleus, x 45 000.

was always seen in TEM (Figs. 13, 14), apparently corresponding to the junctionalridge seen in SEM (cf. Fig. 9).

After 48 h in culture the cells had formed mature junctional complexes, typicalof epithelial cells (cf. Franks & Wilson, 1977). At the apical region of the apposedcells, tight junctions surrounding structures resembling bile canaliculi were seen(Fig. 15). Also zonula adhaerens-like junctions were detected between the cells(Fig. 15, inset). At this stage a distinct apical layer of microfilaments, running parallelto the cell surface was seen in thin sections (Fig. 16). In this layer, bundles ofmicrofilaments were observed, some of which apparently terminated at the junctionalregions of the cells (Figs. 15, 16).

DISCUSSIONIn the present study the initial aggregation of hepatocytes in culture seemed to be

mediated through filopodia-like structures. These were consistently detected betweenthe aggregating cells, but were only seldom seen attached to the growth substratum.In cultured fibroblasts filopodia have a substrate-exploring function, mediate thecell anchorage to substrata, and guide the direction of cell spreading (Albrecht-Buehler, 1976; Bragina, Vasiliev & Gelfand, 1976). Our results indicate that incultured hepatocytes filopodia may provide the initial points of intercellular contactsand facilitate the reaggregation of the cells.

It has been suggested that the rapid formation of specialized cell junctions mightbe needed for stable aggregation of epithelial cells (cf. Middleton & Pegrum, 1976),

Actin and junctions in cultured hepatocytes 351

However, the initial stages of epithelial cell aggregation have not been studied inthis respect. In our culture conditions adult rat hepatocytes formed multicellularaggregates in a few hours, well before cell junctions were detected in thin sections.Junctional complexes consisting of tight junctions surrounding bile canaliculus-likestructures and of zonula adhaerens-type junctions were detected only in the developedepithelial cell sheets after 48 h in culture.

During the first hours of culture only diffuse cytoplasmic fluorescence was seenin the aggregating hepatocytes with anti-actin antibodies. In accordance with thisobservation no bundles of microfilaments were detected in thin sections at thisstage. The first fluorescent actin fibres appeared in the periphery of the developinglamellipodia, at the free edges of the marginal cells of the cell islands. These fibresseemed to grow centripetally from the cell periphery, later forming long fluorescentfibres that seemed to end at the perinuclear area of the marginal cells. Recently Edds(1977) has shown by ultrastructural studies a similar centripetal formation of filamentbundles in sea-urchin coelomocytes. In the established cell islands, filamentous actinfluorescence was also seen at the junctional areas throughout the cell sheet.

The organization of actin-containing microfilaments into bundles in fibroblastscorresponds to the anchorage of the cells to the growth substratum (Pollack &Rifkin, 1975; Willingham et al. 1977). If this is also the case with epithelial cells,then our observations of the distribution of actin-containing bundles of micro-filaments in epithelial islands suggest that the lamellipodia of the marginal cells andthe apical junctional complexes between the cells are the areas of strong anchorage.This suggestion agrees with the results of others (Vaughan & Trinkaus, 1966;DiPasquale, 1975), that epithelial cell sheets attach to the growth substratum mainlyby their marginal cells, whereas the submarginal cells adhere mainly to each other.

Analogously with the immunofluorescence results, microfilaments were seen byelectron microscopy in the submarginal cells of the epithelial cell sheets as an apicallayer of parallel filaments and bundles of filaments that seemed to terminate at thejunctional areas. This apical organization of microfilaments seems to be typical ofepithelial cells, both in vivo (Phillips et al. 1975; Gipson & Anderson, 1977) andin vitro (Middleton & Pegrum, 1976). The function of microfilament bundles is stillcontroversial: a contractile (Isenberg et al. 1976; Weber, 1976) and a cytoskeletalrole (Pollard & Fujiwara, 1976) has been suggested for them in non-muscle cells.Bundles of microfilaments may also be a sign of isometric contraction in cells (Fleischer& Wohlfarth-Bottermann, 1975; Abercrombie, Dunn & Heath, 1976), resulting intension in the epithelial cell sheets. This tension is demonstrated by the shrinkingof epithelial cell sheets when grown on semisolid substrata (Michalopoulos & Pitot,1975) or when detached partially from a solid substratum (Vaughan & Trinkaus, 1966).In the present study the filamentous actin fluorescence found to extend throughout thewhole hepatocyte cell sheet in an organized manner suggests that actin has a rolein maintenance of the internal organization and tension of epithelial cell sheets invitro, as has been suggested also in vivo (Phillips et al. 1975; Gipson & Anderson,1977; Williams, 1977).

23 C E L 31

352 A. Miettinen, I. Virtanen and E. hinder

We are grateful to Ms Jaana Gluschkoff, Ms Pipsa Kaipainen and Ms Pirkko Leikas-Lazanyi for their skilful technical assistance. This study was supported by grants from theSigrid Juse'lius Foundation, The Finnish Medical Research Council, The Finnish CultureFoundation, and the Finnish Foundation for Cancer Research.

REFERENCES

ABERCROMBIE, M., DUNN, G. A. & HEATH, J. P. (1976). Lcomotion and contraction in non-muscle cells. In Contractile Systems in Non-Muscle Tissues (ed. S. V. Perry, A. Margreth &R. S. Adelstein), pp. 3-11. Amsterdam, Oxford, New York: North-Holland.

ALBRECHT-BUEHLER, G. (1976). Filopodia of spreading 3T3 cells. Do they have a substrate-exploring function? J. Cell Biol. 69, 275-286.

BRAGINA, E. E., VASILIEV, JU. M. & GELFAND, I. M. (1976). Formation of bundles of micro-filaments during spreading of fibroblasts on the substrate. Expl Cell Res. 97, 241-248.

DIPASQUALE, A. (1975). Locomotory activity of epithelial cells in culture. Expl Cell Res. 94,191-215-

EDDS, K. T. (1977). Dynamic aspects offilopodial formation by reorganization of microfilaments.J. Cell Biol. 73, 479-491.

FLEISCHER, M. & WOHLFARTH-BOTTERMANN, K. E. (1975). Correlation between tension forcegeneration, fibrillogenesis and ultrastructure of cytoplasmic actomyosin during isometricand isotonic contractions of protoplasmic strands. Cytobiologie 10, 339—365.

FRANKS, L. M. & WILSON, P. D. (1977). Origin and ultrastructure of cells in vitro. Int.Rev. Cytol. 48, 55-139.

GARROD, D. R. & STEINBERG, M. S. (1975). Cell locomotion within a contact-inhibitedmonolayer of chick embryonic liver parenchyma cells. J. Cell Sci. 18, 405-425.

GIPSON, I. K. & ANDERSON, R. A. (1977). Actin filaments in normal and migrating cornealepithelial cells. Invest. Ophthal. 16, 161-166.

GOLDMAN, R. D., YERNA, M.-J. & SCHLOSS, J. A. (1976). Localization and organization ofmicrofilaments and related proteins in normal and virus-transformed cells. J. supramolec.Struct, 5, 155-183.

ISENBERG, G., RATHKE, P. C , HOLSMANN, N., FRANKE, W. W. & WOHLFARTH-BOTTERMANN,K. E. (1976). Cytoplasmic actomyosin fibrils in tissue culture cells. Direct proof of con-tractility by visualization of ATP-induced contraction in fibrils isolated by laser microbeamdissection. Cell Tiss. Res. 166, 427-443.

KURKI, P., LINDER, E., MIETTINEN, A. & ALFTHAN, O. (1977a). Smooth muscle antibodiesof actin and 'non-actin' specificity. Clin. Immun. Immunopath. (in Press).

KURKI, P., LINDER, E., VIRTANEN, I. & STENMAN, S. (19776). Human smooth muscle auto-antibodies reacting with intermediate (100 A) filaments. Nature, Lond. 268, 240—241.

LINDER, E. & MIETTINEN, A. (1976). Prozone effects in indirect immunofluorescence. Scond.J. Immun. 5, 513—519.

MICHALOPOULOS, G. & PITOT, H. C. (1975). Primary culture of parenchymal liver cells oncollagen membranes. Morphological and biochemical observations. Expl Cell Res. 94,70-78.

MIDDLETON, C. A. (1973). The control of epithelial cell locomotion in tissue culture. InLocomotion of Tissue Cells, Ciba Fdn Symp., N.S., No. 14, pp. 251-262. Amsterdam,London, New York: Associated Scientific Publishers.

MIDDLETON, C. A. (1976). Contact-induced spreading is a new phenomenon depending oncell-cell contact. Nature, Lond. 259, 311-313.

MIDDLETON, C. A. & PEGRUM, S. M. (1976). Contacts between pigmented retina epithelialcells in culture. J. Cell Sci. 22, 371-383.

NiCOLSON, G. L. (1976). Transmembrane control of the receptors on normal and tumorcells. I. Cytoplasmic influence over cell surface components. Biochim. biophys. Acta 457,57-io8.

NORBERG, R., LIDMAN, K. & FAGRAEUS, A. (1975). Effects of Cytochalasin B on fibroblasts,lymphoid cells, and platelets revealed by human anti-actin antibodies. Cell 6, 507-512.

OBRINK, B., KUHLENSCHMIDT, M. S. & ROSEMAN, S. (1977). Adhesive specificity of juvenilerat and chicken liver cells and membranes. Proc. natn. Acad. Sci. U.S.A. 74, 1077-1081.

Actin and junctions in cultured hepatocytes 353

PHILLIPS, M.J . , ODA, M., MAK, E., FISHER, M. M. & JEEJEEBHOY, K. N. (1975). Micro-filament dysfunction as a possible cause of intrahepatic cholestasis. Gastroenterology 69,48-58.

POLLACK, R. & RIFKIN, D. (1975). Actin-containing cables within anchorage-dependent ratembryo cells are dissociated by plasmin and trypsin. Cell 6, 495-506.

POLLARD, T. D. & FUJIWARA, K. (1976). Participation of contractile proteins in cytoplasmicstructure and cell division. In Contractile Systems in Non-Muscle Tissues (ed. S. V. Perry,A. Margreth & R. S. Adelstein), pp. 23-28. Amsterdam, Oxford, New York: North-Holland.

POLLARD, T. D., FUJIWARA, K., NIEDERMAN, R. & MAUPIN-SZAMIER, P. (1976). Evidence for

the role of cytoplasmic actin and myosin in cellular structure and motility. Cold SpringHarb. Confs Cell Prolif. 3, 689^724.

PONTEN, J. (1975). Contact inhibition. In Cancer, a Comprehensive Treatise, vol. 4 (ed.F. F. Becker), pp. 55—100. New York: Plenum Press.

SEGLEN, P. O. (1976). Preparation of isolated rat liver cells. Meth. Cell Biol. 13, 29-83.SPUDICH, J. A. & WATT, S. (1971). The regulation of rabbit skeletal muscle contraction. I.

Biochemical studies of the interaction of the tropomyosin-troponin complex with actin andthe proteolytic fragments of myosin. J. biol. Chem. 246, 4866-4871.

TRENCKEV, P. & HOLBOROW, E. J. (1976). The specificity of anti-actin serum. Immunology3i , 5O9-SI7-

TRINKAUS, J. P. (1976). On the mechanism of metazoan cell movements. In The Cell Surfacein Animal Embryogenesis and Development (ed. G. Poste & G. L. Nicolson), pp. 225-329.Amsterdam, New York, Oxford: North-Holland.

VASILIEV, JU. M. & GELFAND, I. M. (1976). Morphogenetic reactions and locomotory behaviourof transformed cells in culture. In Fundamtntal Aspects of Metastasis (ed. L. Weiss),pp. 71-98. Amsterdam, New York, Oxford: North-Holland.

VAUGHAN, R. B. & TRINKAUS, J. P. (1966). Movements of epithelial cell sheets in vitro. J.Cell Sci. 1, 407-413.

VIRTANEN, I. & WARTIOVAARA, J. (i 976). Lectin receptor sites on rat liver cell nuclear membranes.J. Cell Sci. 22, 335-344-

WEBER, K. (1976). Biochemical anatomy of microfilaments in cells in tissue culture usingimmunofluorescence microscopy. In Contractile Systems in Non-Muscle Tissues (ed. S. V.Perry, A. Margreth & R. S. Adelstein), pp. 51-66. Amsterdam, New York, Oxford: North-Holland.

WILLIAMS, J. A. (1977). Effects of Cytochalasin B on pancreatic acinar structure and secretion.Cell Tiss. Res. 179, 453-466.

WILLINGHAM, M. C , YAMADA, K. M., YAMADA, S. S., POUYSSEGUR, J. & PASTAN, I. (1977).Microfilament bundles and cell shape are related to adhesiveness to substratum and aredissociable from growth control in cultured fibroblasts. Cell 10, 375-380.

(Received 29 September 1977)

23-2