The association of statins and taxanes: an efficient combination trigger of cancer cell apoptosis.

BACKGROUND: Cancer cell killing might be achieved by the combined use of available drugs. Statins are major anti-hypercholesterolemia drugs, which also trigger apoptosis of many cancer cell types, while docetaxel is a potent microtubule-stabilising agent. METHODS: Here, we looked at the combined effects of lovastatin and docetaxel in cancer cells. RESULTS: Whole transcriptome microarrays in HGT-1 gastric cancer cells demonstrated that lovastatin strongly suppressed expression of genes involved in cell division, while docetaxel had very little transcriptional effects. Both drugs triggered apoptosis, and their combination was more than additive. A marked rise in the cell-cycle inhibitor p21, together with reduction of aurora kinases A and B, cyclins B1 and D1 proteins was induced by lovastatin alone or in combination with docetaxel. The drug treatments induced the proteolytic cleavage of procaspase-3, a drop of the anti-apoptotic Mcl-1 protein, Poly-ADP-Ribose Polymerase and Bax. Strikingly, docetaxel-resistant HGT-1 cell derivatives overexpressing the MDR-1 gene were much more sensitive to lovastatin than docetaxel-sensitive cells. CONCLUSION: These results suggest that the association of lovastatin and docetaxel, or lovastatin alone, shows promise as plausible anticancer strategies, either as a direct therapeutic approach or following acquired P-glycoprotein-dependent resistance.

The B[a]P-increased intercellular communication via translocation of connexin-43 into gap junctions reduces apoptosis.

Gap junctions are channels in plasma membrane composed of proteins called connexins. These channels are organized in special domains between cells, and provide for direct gap junctional intercellular communication (GJIC), allowing diffusion of signalling molecules <1 kD. GJIC regulates cell homeostasis and notably the balance between proliferation, cell cycle arrest, cell survival and apoptosis. Here, we have investigated benzo[a]pyrene (B[a]P) effects on GJIC and on the subcellular localization of the major protein of gap junction: connexin-43 (Cx43). Our results showed that B[a]P increased GJIC between mouse hepatoma Hepa1c1c7 cells via translocation of Cx43 from Golgi apparatus and lipid rafts into gap junction plaques. Interestingly, inhibition of GJIC by chlordane or small interference RNA directed against Cx43 enhanced B[a]P-induced apoptosis in Hepa1c1c7 cells. The increased apoptosis caused by inhibition of GJIC appeared to be mediated by ERK/MAPK pathway. It is suggested that B[a]P could induce transfer of cell survival signal or dilute cell death signal via regulation of ERK/MAPK through GJIC.

A plasma membrane protein is involved in cell contact-mediated regulation of tissue-specific genes in adult hepatocytes.

We have identified the liver-regulating protein (LRP), a cell surface protein involved in the maintenance of hepatocyte differentiation when cocultured with rat liver epithelial cells (RLEC). LRP was defined by immunoreactivity to a monoclonal antibody (mAb L8) prepared from RLEC. mAb L8 specifically detected two polypeptides of 85 and 73 kD in immunoprecipitation of both hepatocyte- and RLEC-iodinated plasma membranes. The involvement of these polypeptides, which are integral membrane proteins, in cell interaction-mediated regulation of hepatocytes was assessed by evaluating the perturbing effects of the antibody on cocultures with RLEC. Several parameters characteristic of differentiated hepatocytes were studied, such as liver-specific and house-keeping gene expression, cytoskeletal organization and deposition of extracellular matrix (ECM). An early cytoskeletal disturbance was evidenced and a marked alteration of hepatocyte functional capacity was observed in the presence of the antibody, together with a loss of ECM deposition. By contrast, cell-cell aggregation or cell adhesion to various extracellular matrix components were not affected. These findings suggest that LRP is distinct from an extracellular matrix receptor. The fact that early addition of mAb L8 during cell contact establishment was necessary to be effective may indicate that LRP is a novel plasma membrane protein that plays an early pivotal role in the coordinated metabolic changes which lead to the differentiated phenotype of mature hepatocytes.

The mitogen-activated protein kinase kinase/extracellular signal-regulated kinase cascade activation is a key signalling pathway involved in the regulation of G(1) phase progression in proliferating hepatocytes.

In this study, activation of the mitogen-activated protein kinase kinase (MEK)/extracellular signal-regulated kinase (ERK) signalling pathway was analyzed in proliferating rat hepatocytes both in vivo after partial hepatectomy and in vitro following epidermal growth factor (EGF)-pyruvate stimulation. First, a biphasic MEK/ERK activation was evidenced in G(1) phase of hepatocytes from regenerating liver but not from sham-operated control animals. One occurred in early G(1) (30 min to 4 h), and the other occurred in mid-late G(1), peaking at around 10.5 h. Interestingly, the mid-late G(1) activation peak was located just before cyclin D1 induction in both in vivo and in vitro models. Second, the biological role of the MEK/ERK cascade activation in hepatocyte progression through the G(1)/S transition was assessed by adding a MEK inhibitor (PD 98059) to EGF-pyruvate-stimulated hepatocytes in primary culture. In the presence of MEK inhibitor, cyclin D1 mRNA accumulation was inhibited, DNA replication was totally abolished, and the MEK1 isoform was preferentially targeted by this inhibition. This effect was dose dependent and completely reversed by removing the MEK inhibitor. Furthermore, transient transfection of hepatocytes with activated MEK1 construct resulted in increased cyclin D1 mRNA accumulation. Third, a correlation between the mid-late G(1) MEK/ERK activation in hepatocytes in vivo after partial hepatectomy and the mitogen-independent proliferation capacity of these cells in vitro was established. Among hepatocytes isolated either 5, 7, 9, 12 or 15 h after partial hepatectomy, only those isolated from 12- and 15-h regenerating livers were able to replicate DNA without additional growth stimulation in vitro. In addition, PD 98059 intravenous administration in vivo, before MEK activation, was able to inhibit DNA replication in hepatocytes from regenerating livers. Taken together, these results show that (i) early induction of the MEK/ERK cascade is restricted to hepatocytes from hepatectomized animals, allowing an early distinction of primed hepatocytes from those returning to quiescence, and (ii) mid-late G(1) MEK/ERK activation is mainly associated with cyclin D1 accumulation which leads to mitogen-independent progression of hepatocytes to S phase. These results allow us to point to a growth factor dependency in mid-late G(1) phase of proliferating hepatocytes in vivo as observed in vitro in proliferating hepatocytes and argue for a crucial role of the MEK/ERK cascade signalling pathway.

Mechanism in the sequential control of cell morphology and S phase entry by epidermal growth factor involves distinct MEK/ERK activations.

Cell shape plays a role in cell growth, differentiation, and death. Herein, we used the hepatocyte, a normal, highly differentiated cell characterized by a long G1 phase, to understand the mechanisms that link cell shape to growth. First, evidence was provided that the mitogen-activated protein kinase kinase (MEK)/extracellular signal-regulated kinase (ERK) cascade is a key transduction pathway controlling the hepatocyte morphology. MEK2/ERK2 activation in early G1 phase did not lead to cell proliferation but induced cell shape spreading and demonstration was provided that this MAPK-dependent spreading was required for reaching G1/S transition and DNA replication. Moreover, epidermal growth factor (EGF) was found to control this morphogenic signal in addition to its mitogenic effect. Thus, blockade of cell spreading by cytochalasin D or PD98059 treatment resulted in inhibition of EGF-dependent DNA replication. Our data led us to assess the first third of G1, is exclusively devoted to the growth factor-dependent morphogenic events, whereas the mitogenic signal occurred at only approximately mid-G1 phase. Moreover, these two growth factor-related sequential signaling events involved successively activation of MEK2-ERK2 and then MEK1/2-ERK1/2 isoforms. In addition, we demonstrated that inhibition of extracellular matrix receptor, such as integrin beta1 subunit, leads to cell arrest in G1, whereas EGF was found to up-regulated integrin beta1 and fibronectin in a MEK-ERK-dependent manner. This process in relation to cytoskeletal reorganization could induce hepatocyte spreading, making them permissive for DNA replication. Our results provide new insight into the mechanisms by which a growth factor can temporally control dual morphogenic and mitogenic signals during the G1 phase.

Modulation of alkaline phosphodiesterase I in cultured rat hepatocytes.

Alkaline phosphodiesterase I activity was measured in adult and foetal rat hepatocytes maintained in primary culture under various conditions. This enzyme was found to be expressed in both cell populations and could be resolved into two bands having apparent molecular weights of 130,000 and 250,000, respectively. Alkaline phosphodiesterase I activity was already at high levels in 15 day foetal liver and, as early as the 19th day of gestation, it reached adult levels. Alkaline phosphodiesterase I levels were well maintained during culture. In the absence of serum, its level continued to increase with time in foetal cells. It dramatically increased by days 4 and 5, in adult cells maintained on fibronectin and plastic, respectively. Dexamethasone stimulated alkaline phosphodiesterase I activity after a lag phase of 8 h, with a maximum reached after 40 h. As this induction was prevented by addition of actinomycin D or cycloheximide, it could be concluded that it required RNA and protein synthesis. Only the major Mr 250,000 form responded to dexamethasone and was sensitive to serum.

Chronic iron overload inhibits protein secretion by adult rat hepatocytes maintained in long-term primary culture.

Short-term pure cultures and long-term cocultures of adult rat hepatocytes with rat liver epithelial cells, presumably derived from primitive biliary cells, were used to define in vitro models of iron overloaded hepatocytes in order to understand the molecular mechanism responsible for liver damage occurring in patients with hemochromatosis. In vitro iron overload was obtained by daily addition of ferric nitrilotriacetate to the culture medium. A concentration of 20 microM ferric salt induced hepatocyte iron overload with minimal cytotoxicity as evaluated by cell viability, morphological changes of treated cells and cytosolic enzyme leakage into the culture medium. The effects of iron overload on protein biosynthesis and secretion were studied in both short-term pure cultures and long-term cocultures of hepatocytes. The amounts of intracellular and newly synthesized proteins were never modified by the iron treatment. Furthermore, neither the relative amounts of transferrin and albumin mRNAs nor their translational products were altered by iron overload. Moreover, no change in the transferrin isomeric forms were observed in treated cells. In contrast, a prolonged exposure of cocultured hepatocytes to 20 microM ferric salt led to a significant decrease in the amount of proteins secreted in the medium. This decrease included the two major secreted proteins, namely albumin and transferrin, and probably all other secreted proteins. These results demonstrate that iron loading alters neither the total nor the liver specific protein synthesis activity of cultured hepatocytes. They suggest that chronic overload may impede the protein secretion process.

Skp2 induction and phosphorylation is associated with the late G1 phase of proliferating rat hepatocytes.

The changes in phosphoproteins purified with the affinity peptide p9CKShs1 were analyzed from extracts of regenerating rat livers in order to define some G1 and G1/S regulations characteristic of mature hepatocytes stimulated to proliferate. We observed a 47 kDa phosphoprotein that occurred first at the end of G1 before peaking in the S phase. P47 was also found to be phosphorylated in late G1 in primary hepatocyte cultures stimulated with mitogens. P47 was still phosphorylated in extracts depleted of Cdc2, but to a lesser extent after Cdk2 depletion. This phosphoprotein was identified as Skp2. (i) P47 shared the same electrophoretic mobility than Skp2, a cell cycle protein essential for S phase entry in human fibroblasts; (ii) Skp2, like P47, started to be expressed and was highly phosphorylated during the G1/S transition of hepatocytes stimulated to proliferate in vivo and in vitro; (iii) P47 was specifically immunoprecipitated by an antibody directed against Skp2. In addition, cyclin A/Cdk2 complexes from regenerating liver clearly interacted with Skp2. This is the first demonstration that Skp2 is induced and phosphorylated in the late G1 and S phase of hepatocytes in vivo in regenerating liver as well as in vitro in mitogen-stimulated hepatocytes.

Highly selective extraction of spiralin from the Spiroplasma citri cell membrane with alkyl-N-sulfobetaines.

The extraction of proteins from the membrane of the mollicute (mycoplasma) Spiroplasma citri by sodium N-dodecyl-N,N-dimethyl-3-amino-1-propane sulfonate (SB12) and sodium N-tetradecyl-N,N-dimethyl-3-amino-1-propane sulfonate (SB14) was studied with electrophoretic methods. The membranes were prepared by osmotic lysis of the cells and depleted of the bulk of extrinsic proteins. It was possible to extract up to 35 and 45% of membrane proteins with SB12 and SB14, respectively. Maximal yield was obtained in both cases with detergent concentrations greater than or equal to 5 mumoles/mg of membrane protein. Spiralin, the major protein in the S. citri membrane, was highly selectively solubilized without the loss of antigenicity, with a yield of about 90% with SB12 and close to 100% with SB14, for a detergent concentration greater than or equal to 0.2 M. The degree of selectivity in favour of spiralin was higher with SB12 (purity approximately equal to 70%) than with SB14 (purity approximately equal to 50%). Treatment of the S. citri membrane with high concentrations of SB12 is a simple and fast procedure for partial purification of spiralin. This example shows that, in some cases, it should be possible to modulate the selectivity of the extraction of membrane proteins simply by varying the relative concentration of detergent.

Gene regulation by interleukin 6.

Interleukin 6 (IL-6) is a central alarm hormone of the mammalian body. During acute and chronic inflammations, it induces acute phase plasma protein synthesis by liver hepatocytes, modulates the immune response and participates in the regulation of body temperature (fever). In addition, it is a growth factor for certain tumor cells, such as myeloma cells. The details of the IL-6 signal transduction mechanism are unknown. We have contributed to this problem at 2 levels: (a), we have mapped an IL-6-response element (IL-6-RE) in the 5' flanking region of the alpha 2-macroglobulin gene (alpha 2M), a prototype rat liver acute phase gene. This element, CTGGGA, serves as a binding site for nuclear factors that facilitate hormone induced transcription. We have begun to characterize these factors from hepatic cells and demonstrated that they undergo characteristic IL-6-induced changes. Similar factors were also discovered in human Burkitt tumor derived cell lines (B cells). These bound at the IL-6-RE of the rat alpha 1M gene and formed indistinguishable protein DNA complexes, as the corresponding hepatic factors. Thus, common elements probably operate in the IL-6 signal transduction cascade in liver cells and B cells; (b), we have cloned the rat liver IL-6 receptor (IL-6-R) and derived its amino acid sequence. It was 53% identical to the human leukocyte IL-6-R and all functional domains were highly conserved. Therefore, the cell-type specific responses to IL-6 in liver cells and lymphocytes were probably not due to cell-type specific forms of the receptor, but to other so far unknown elements of the signal transduction cascade.(ABSTRACT TRUNCATED AT 250 WORDS)

Immunocytochemical evidence for the maintenance of cytochrome P-450 isozymes, NADPH cytochrome C reductase, and epoxide hydrolase in pure and mixed primary cultures of adult human hepatocytes.

Specific polyclonal antibodies were used to investigate the distribution of two cytochrome P-450 isozymes (5 and 8), NADPH cytochrome c reductase, and epoxide hydrolase in adult human hepatocytes cultured alone or co-cultured with rat liver epithelial cells. The enzymes were localized by the indirect immunoperoxidase technique following fixation with a paraformaldehyde-glutaraldehyde mixture and membrane permeabilization with saponin. The pattern of distribution of the four enzymes after 24 hr in culture was similar to that found in vivo. Virtually all the hepatocytes exhibited nearly homogeneous positive staining for cytochrome P-450-8, whereas only 60-80% were positive for cytochrome P-450-5. Nearly homogeneous staining was also observed in all hepatocytes for NADPH cytochrome c reductase and epoxide hydrolase. During the first 12 days in pure culture, the intensity of staining, as well as the number of positively stained cells, decreased slightly except for epoxide hydrolase, which did not show any obvious change. In contrast, even after 15 days in co-culture the extent of staining for all the enzymes decreased less than in pure culture. These results indicate that adult human hepatocytes continue to express specific drug-metabolizing enzymes for several days in culture and provide further evidence that those cells are more stable than rodent hepatocytes in primary culture.

Biosynthesis of factor V by normal adult rat hepatocytes.

The synthesis of coagulation factor V was investigated in isolated rat hepatocytes maintained in long-term primary culture. Two culture conditions were compared. A clotting assay and an immunoprecipitation experiment with rabbit anti rat factor V IgG were used to demonstrate not only the presence of factor V in the cells but also active secretion into the culture medium. Both the inhibition of the clotting reaction in presence of the antibody and absence of thrombin in culture media confirmed the specificity of the clotting assays. Electron microscopic examination located factor V in the endoplasmic reticulum and Golgi apparatus of hepatocytes in common with other liver specific plasma proteins. Examination of liver tissue sections confirmed the production of factor V in hepatocytes but not in hepatic endothelial cells although it did not exclude a transit pathway of factor V through these cells. Addition of Russell viper venom factor V activating enzyme to the culture medium had no effect on the factor V activity. In contrast, treatment of cell extracts did increase the coagulant activity. This suggests that hepatocytes contained principally an unactivated form or procofactor, whereas factor V present into the culture medium was mainly in an activated form. These data provide evidence for synthesis and secretion of an hepatocytic factor V.

[Role of intercellular communications in hepatic differentiation and carcinogenesis]. / Les communications intercellulaires dans la différenciation et la carcinogenèse hépatiques.

The liver epoch like other tissue epochs, occurs after that different events have induced heterogeneity in embryonic cells which result in distinct evolutionary processes. These events and those of organogenesis like "induction", are deeply dependent on cell-cell communications. Cell-cell interactions involve either soluble factors (hormones, growth factors), extracellular matrix or plasma membrane proteins responsible for cell-cell recognition and/or adhesion. All these plasma membrane signals are transduced to the nucleus and modulate the expression of groups of genes. To be functionally stable along the adult stage the liver has to maintain an ordered activity of cell renewal. This balance between proliferation and differentiation is at least in part, controlled by cell-cell communications. Therefore, it is not surprising that intercellular communications are altered during hepatocarcinogenesis. They involve changes in the distribution of junctions, in the amounts of extracellular matrix components and/or growth factors which all result in modifying the differentiation/proliferation balance. Cell culture models have been used for these different studies; new in vitro systems should be set up in the near future by taking advantage of the targeted hepatocarcinogenesis in transgenic mouse.

Characterization of goldfish fin cells in culture: some evidence of an epithelial cell profile.

Comprehensive characterization of cultured cells in fish was little explored and cell origin is often deduced from morphological analogies with either epithelial of fibroblastic cells. This study aims to characterize cell origin in goldfish fin culture using morphological, immunochemical, and molecular approaches. Time lapse analysis revealed that cultured cell morphology changed within minutes. Therefore, cell morphology cannot predict whether cells are from fibroblastic or epithelial origin. The labeling pattern of heterologous anti-cytokeratin and anti-vimentin antibodies against goldfish epithelial cells and fibroblasts was first tested on skin sections and the corresponding labeling of the cultured cells was analyzed. No cell origin specificity could be obtained with the chosen antibodies. In the molecular approach, detection levels of three cytokeratin (CauK8-IIS, CauK49-IE and CauK50-Ie) and one vimentin transcripts were assessed on skin and fin samples. Specificity for epithelial cells of the most abundant mRNA, CauK49-Ie, was thereafter validated on skin sections by in situ hybridization. The selected markers were used afterwards to characterize fin cultures. CauK49-IE riboprobe labeled every cell in young cultures whereas no labeling was observed in older cultures. Accordingly, CauK49-IE transcript levels decreased after 15 days culture while CauK8-IIS ones increased. The use of homologous marker gave evidence that young cultured cells from goldfish fin are homogeneously of epithelial type and that cell characteristics may change over culture time.

Orientation fields of nonlinear biological fibrils by second harmonic generation microscopy.

The orientation of fibrils within biological tissues is of primary importance. In this study, we propose a simple method based on second harmonic generation (SHG) microscopy to map, pixel by pixel, the orientation of the symmetry axis of the second-order nonlinear susceptibility tensor of fibrils that produce SHG. The method uses only four images acquired at specific polarizations of the input laser beam, and can be easily and cheaply implemented on a confocal microscope. In addition to orientation informations, the method also provides polarization independent images and estimations of the ratio of the nonlinear susceptibility components. We demonstrate the relevance of our concept by studying the orientation fields of the collagen meshwork in a healthy rat liver that provides well separated fibrils. By correlating the mean orientation of the nonlinear susceptibility to the fibril orientation itself for many fibril segments, and using circular statistics, it is shown that both orientations are truly parallel at the fibril scale. Our polarimetric method allows to map fibril orientation fields, independently of individual fibril contrast in the SHG image.

RNAi-mediated ERK2 knockdown inhibits growth of tumor cells in vitro and in vivo.

The MAPK MEK/ERK pathway is often upregulated in cancer cells and represents an attractive target for development of anticancer drugs. Only few data concerning the specific functions of ERK1 and 2 are reported in the literature. In this report, we investigated the specific role of ERK1 and 2 in liver tumor growth both in vitro and in vivo. DNA synthesis and cells in S phase analysed by flow cytometry, correlated with strong inhibition of Cdk1 and cyclin E levels, are strongly reduced after exposure to the MEK inhibitor, U0126. We obtained a significant reduction of colony formation in soft agar assays and a reduction in the size of tumor xenografts in nude mice treated with U0126. Then, we could specifically abolished ERK1 or 2 expression by small-interfering RNA (siRNA) and demonstrated that ERK2 knockdown but not ERK1 interferes with the process of replication. Moreover, we found that colony formation and tumor growth in vivo were significantly inhibited by targeting ERK2 using stable chemically modified siRNA. Taken together, our results emphasize the importance of the MEK/ERK pathway in liver cancer cell growth in vitro and in vivo and argue for a crucial role of ERK2 in this regulation.

The liver-specific response to phenobarbital involves a transient increase in phosphorylation of a 34-kda nuclear protein in rat liver and in hepatocytes in culture.

We have examined the influence of phenobarbital, a liver detoxication gene inducer and a potent tumor promoter, on the phosphorylation status of nuclear-enriched proteins in primary rat hepatocyte cultures and in whole livers. Freshly isolated cells were plated on plastic dishes in presence of serum for 4 h and 2 mM phenobarbital was added for various times, following serum withdrawal. A transient increase in phosphorylation of a 34-kda nuclear protein was detected at 6 h. In whole livers, but not in kidneys, a nuclear protein with the same electrophoretic mobility was also transiently over-phosphorylated, following injection of 80 mg/kg phenobarbital, although the peak activity was attained after 30 min only. No immunological relatedness between major histones and the 34-kda protein was found. Our results demonstrate a specific, yet undescribed, transient effect of phenobarbital on the phosphorylation status of a 34-kda nuclei-enriched protein in rat hepatocytes and in rat liver.

Synthesis and phosphorylation of cytoskeleton components in foetal, regenerating and adult normal rat hepatocytes during culture.

Detergent insoluble material (DIM) was prepared by gentle treatment with detergent from foetal, regenerating and adult normal rat hepatocytes cultured for various times. It retained to some degree the morphology of the cells. After incubation of intact cells with 35S-methionine, most of the labelled DIM proteins were found to be components of the cytoskeleton. They included several cytokeratins, vimentin and actin. The synthesis rate varied with the age of animals and culture conditions. The high synthetic rate of vimentin in foetal and regenerating hepatocytes could be associated with cell proliferation. No correlation was found between cytokeratin synthesis and hepatocyte growth. Most of the cytoskeleton proteins could be phosphorylated in intact cells and in DIM from cultured hepatocytes. However the degree of phosphorylation of these proteins was not related to their synthetic rate. The decreased phosphorylation level in cultured adult rat hepatocytes could be related to the rapid loss of specific functions.

Distinct effects of cell-cell communication and corticosteroids on the synthesis and distribution of cytokeratins in cultured rat hepatocytes.

Cytokeratins CK 8 and CK 18 are the two keratins expressed in the liver. They are known to undergo extensive changes in expression with alteration of the hepatocyte phenotype in vitro. In this study, we have investigated the variation in levels of these two cytokeratins in hepatocytes selected from different situations in vivo. The amounts of corresponding transcripts were compared; cytokeratin 8 and 18 mRNAs were present at similar levels in hepatocytes freshly isolated from adult liver and, unexpectedly, from 17-day-old foetuses and newborn rats, whereas they were markedly higher in regenerating hepatocytes isolated early after partial hepatectomy. In order to investigate whether the different factors that can promote hepatocyte differentiation also produce a similar set of cytoskeletal changes, we have analysed both the expression and the distribution of cytokeratins in hepatocytes under different culture conditions allowing modulation of differentiation. Establishment of cell-cell contacts and addition of glucocorticoids were used as two modulating factors. Coculturing hepatocytes with rat liver epithelial cells (RLEC), which favours active expression of liver-specific genes, resulted in a gradual decline of cytokeratin mRNAs, whereas pure hepatocyte cultures, which exhibit rapid phenotypic changes, expressed increasing levels of CK 8 and CK 18 transcripts. Furthermore, intracellular CK distribution was dramatically modified in parallel: the CK-positive material formed a fine network of fibrils uniformly distributed in the cytoplasm of hepatocytes in pure culture, whereas in cocultured cells CK immunofluorescence appeared principally located at the cellular periphery and it was regularly arranged in long fibrils just beneath the plasma membrane.(ABSTRACT TRUNCATED AT 250 WORDS)