Histological organization in hepatocyte organoid cultures.

Hepatocytes and other cellular elements isolated by collagenase perfusion of the liver and maintained in defined culture conditions undergo a series of complex changes, including apoptosis and cell proliferation, to reconstruct tissue with specific architecture. Cultures in collagen-coated pleated surface roller bottles, with hepatocyte growth medium medium and in the presence of hepatocyte growth factor (HGF) and epidermal growth factor (EGF), form characteristic and reproducible tissue architecture composed of a superficial layer of biliary epithelial cells, an intermediate layer of connective tissue and hepatocytes, and a basal layer of endothelial cells. Dexamethasone, EGF, and HGF are required for the complete histological organization. Analysis of the structures formed demonstrates that the receptor tyrosine kinase ligands HGF and EGF are required for the presence, growth, and phenotypic maturation of the biliary epithelium on the surface of the cultures and for the formation of connective tissue in the cultures. Dexamethasone, in the presence of HGF and EGF, was required for the phenotypic maturation of hepatocytes. The results demonstrate the role of these molecules for the formation and phenotypic maturation of specific histological elements of the liver and suggest roles for these signaling molecules in the formation and structure of the in vivo hepatic architecture.

The processing and utilization of hepatocyte growth factor/scatter factor following partial hepatectomy in the rat.

Hepatocyte growth factor/scatter factor (HGF/SF) is a pluripotent growth factor capable of acting as a motogen, a morphogen, and a mitogen. Originally, HGF/SF was found as a blood-borne mitogen for hepatocytes and has since been determined to be very important in liver repair. Previous studies have established that HGF/SF must be proteolytically cleaved to elicit its effects. After liver injury by toxins such as carbon tetrachloride or after surgical resection, partial hepatectomy (PHX), HGF/SF concentrations increase in the blood. The aims of this study were to examine (1) which form of HGF/SF is present in the normal liver, (2) which form is present in the regenerating liver after PHX, and (3) if the HGF/SF used after PHX is derived from existing liver reservoirs. Both single-chain HGF/SF and active two-chain HGF/SF are present in normal liver, with the former being the dominant form. After PHX, the liver can be described as having two phases with regard to the use of endogenous HGF/SF. The first phase from 0 to 3 hours is the consumptive phase and is characterized by a decrease in both single-chain HGF/SF and active two-chain HGF/SF. The second phase is the productive phase. It is characterized by a pronounced reappearance of both single-chain HGF/SF as well as two-chain HGF/SF. The activation index shows a 5-fold increase over sham operations during the productive phase. The use of radiolabeled HGF/SF showed that during the first 3 hours, HGF/SF is used in part from hepatic stores. Furthermore, during the first 3 hours after PHX, only active two-chain HGF/SF is seen in the plasma.

Changes in WNT/beta-catenin pathway during regulated growth in rat liver regeneration.

The wnt/beta-catenin pathway is important during embryogenesis and carcinogenesis. beta-Catenin interaction with E-cadherin has been shown to be crucial in cell-cell adhesion. We report novel findings in the wnt pathway during rat liver regeneration after 70% partial hepatectomy using Western blot analyses, immunoprecipitation studies, and immunofluorescence. We found wnt-1 and beta-catenin proteins to be predominantly localized in hepatocytes. Immediately following partial hepatectomy, we observed an initial increase in beta-catenin protein during the first 5 minutes with its translocation to the nucleus. We show this increase to be the result of decreased degradation of beta-catenin (decrease in serine phosphorylated beta-catenin) as seen by immunoprecipitation studies. We observed activation of beta-catenin degradation complex comprising of adenomatous polyposis coli gene product (APC) and serine-phosphorylated axin protein, beginning at 5 minutes after hepatectomy, leading to its decreased levels after this time. Quantitative changes observed in E-cadherin protein during liver regeneration are, in general, reverse to those seen in beta-catenin. In addition, using immunoprecipitation, we observe elevated levels of tyrosine-phosphorylated beta-catenin at 6 hours onward. Thus, changes in the wnt pathway during regulated growth seem to tightly regulate cytosolic beta-catenin levels and may be contributing to induce cell proliferation and target gene expression. Furthermore, these changes might also be intended to negatively regulate cell-cell adhesion for structural reorganization during the process of liver regeneration.

Sustained expression of naked plasmid DNA encoding hepatocyte growth factor in mice promotes liver and overall body growth.

To understand the physiological functions of exogenous hepatocyte growth factor (HGF) on normal adult animals, we delivered human HGF gene into mice by a hydrodynamics-based in vivo gene transfection approach using a naked plasmid vector. Systemic administration of naked plasmid containing HGF cDNA driven under cytomegalovirus promoter (pCMV-HGF) by rapid injection via the tail vein produced a remarkable level of human HGF protein in the circulation, beginning to appear at 4 hours and peaking at 12 hours following injection. Tissue distribution studies identified the liver as the organ with the highest level of transgene expression. Through weekly repeated injections of plasmid vector, we achieved sustained, long-term, high levels of exogenous HGF expression in mice for 8 weeks. Increases of more than 31% and 16% in liver and body weights were found, respectively, in the mice that received pCMV-HGF plasmid compared with that given the control vector for 8 weeks. Expression of exogenous HGF in vivo activated mitogen-activated protein kinases and induced proliferating cell nuclear antigen expression in normal adult liver and kidneys. These data suggest that systemic administration of naked plasmid vector is a convenient, safe, and highly efficient approach to introduce and maintain exogenous HGF gene expression in vivo in a controllable fashion. Our results also indicate that long-term expression of human HGF in mice markedly activates growth-related signal transduction events, promotes cell proliferation, and leads to liver and overall body growth in whole adult animals.

Comparative genomic hybridization of hepatocellular carcinoma: correlation with fluorescence in situ hybridization in paraffin-embedded tissue.

BACKGROUND: Proto-oncogene MYC, mapped to chromosomal band 8q24 and the genes for hepatocyte growth factor (HGF at 7q21) and its receptor, MET, at chromosomal band 7q31, have an important role in the biology and growth of normal and neoplastic liver. Comparative genomic hybridization (CGH) and fluorescence in situ hybridization (FISH) studies have reported frequent abnormalities of chromosomes 1 and 8 in hepatocellular carcinomas (HCCs) of various clinical and pathological stages. Chromosome 7 involvement is reported to be less frequent. MATERIALS AND METHODS: Frozen tissue from 17 HCCs was used for CGH analysis and sections of corresponding formalin-fixed, paraffin-embedded HCC tissue were used for dual-color FISH with locus-specific (LSI-cMYC for chromosome 8q24 and LSI D7S486 for chromosome 7q31) and centromeric probes, CEP8 (8p11.1-q11.2) and CEP7 (7p11.1-q11.2) (Vysis, Inc, Downers Grove, IL). This study intended to determine the pattern of chromosomal aberrations in early-stage (incidental) HCC and large surgically resected HCC, and also compared the efficiency and usefulness of the two cytogenetic methods. RESULTS: CGH showed abnormalities on chromosomes 1q, 5q, 7q, 8q, 9, 10, 13q, 15, 16, 17p, 18q, 19, 20, 21, 22, and X. Gains of 8q were noted in 50% of the HCCs, including five cases of incidental HCCs by CGH. Increase in copy numbers of MYC detected by FISH was noted in 25% of tumors that had shown 8q gains by CGH and in five cases with no chromosome abnormalities noted by CGH. Three cases with 7q31 copy number abnormalities were found by FISH in addition to those detected by CGH. CONCLUSION: Combined use of CGH and FISH may provide important information about early and/or primary genetic changes in the development of HCC.

Induction of cytochrome P450 (CYP)1A1, CYP1A2, and CYP3A4 but not of CYP2C9, CYP2C19, multidrug resistance (MDR-1) and multidrug resistance associated protein (MRP-1) by prototypical inducers in human hepatocytes.

Human hepatocytes cultured serum-free for up to 6 weeks were used to study expression and induction of enzymes and membrane transport proteins involved in drug metabolism. Phase I drug metabolizing enzymes cytochrome P450 (CYP)1A1, CYP1A2, CYP2C9, CYP2C19, CYP2E1, and CYP3A4 were detected by Western blot analyses and, when appropriate, by enzymatic assays for ethoxyresorufin-O-deethylase(EROD)-activity and testosterone-6beta-hydroxylase(T6H)-activity. Expression of the membrane transporter multi-drug resistance protein (P-glycoprotein, MDR-1), multidrug resistance-associated protein (MRP-1), and lung-resistance protein (LRP) was maintained during the culture as detected by RT-PCR and Western blot analyses. Model inducers like rifampicin, phenobarbital, or 3-methylcholanthrene and beta-naphtoflavone were able to induce CYP1A or CYP3A4 as well as EROD or T6H activities for up to 30 days. CYP2C9, CYP2C19 and CYP2E1 expression was maintained but not inducible for 48 days. Also, rifampicin and phenobarbital were unable to increase MDR-1 and MRP-1 protein levels significantly.

Serum-free, long-term cultures of human hepatocytes: maintenance of cell morphology, transcription factors, and liver-specific functions.

Since human hepatocytes are available only in limited number, the development of a serum-free culture system for long-term cultivation of differentiated and functional hepatocytes is of great importance. Here we describe the culture of human hepatocytes in a chemically defined serum-free medium for up to 5 weeks. Cell morphology was assayed by light and electron microscopy and revealed a well-preserved cellular morphology. Marker proteins for epithelial and bile duct cells, cytokeratin (CK) 18 and 19, and liver-specific proteins, like phosphoenolpyruvate carboxykinase-2 (PCK2) and serum proteins, were expressed. Liver-enriched transcription factors CCAAT/enhancer binding protein alpha (C/EBPalpha) and hepatocyte nuclear factor-4 (HNF-4), cytokine and mitogen activated factors (nuclear factor kappa B) NFkappaB, and activator protein-1 (AP-1) were maintained and active for several weeks in our cultures. In summary, our serum-free culture system allows the culture of differentiated human hepatocytes for several weeks. It may serve as a model system for metabolic, pharmacologic-toxicologic studies, and studies on human pathogens under defined chemical conditions.

Cross-talk between epidermal growth factor receptor and c-Met signal pathways in transformed cells.

In rat liver epithelial cells constitutively expressing transforming growth factor alpha (TGFalpha), c-Met is constitutively phosphorylated in the absence of its ligand, hepatocyte growth factor. We proposed that TGFalpha and the autocrine activation of its receptor, epidermal growth factor receptor (EGFR), leads to phosphorylation and activation of c-Met. We found that there is constitutive c-Met phosphorylation in human hepatoma cell lines and the human epidermoid carcinoma cell line, A431 which express TGFalpha, but not in normal human hepatocytes. Constitutive c-Met phosphorylation in A431, HepG2, AKN-1, and HuH6 cells was inhibited by neutralizing antibodies against TGFalpha and/or EGFR. Exposure to exogenous TGFalpha or EGF increased the phosphorylation of c-Met in the human epidermoid carcinoma cell line, A431. The increase of c-Met phosphorylation by TGFalpha in A431 cells was inhibited by neutralizing antibodies against TGFalpha and/or EGFR and by the EGFR-specific inhibitor tyrphostin AG1478. These results indicate that constitutive c-Met phosphorylation, and the increase of c-Met phosphorylation by TGFalpha or EGF, in tumor cell lines is the result of the activation via EGFR. We found that c-Met in tumor cells co-immunoprecipitates with EGFR regardless of the existence of their ligands in tumor cells, but not in normal human hepatocytes. We conclude that c-Met associates with EGFR in tumor cells, and this association facilitates the phosphorylation of c-Met in the absence of hepatocyte growth factor. This cross-talk between c-Met and EGFR may have significant implications for altered growth control in tumorigenesis.

Expression and activation of pro-MMP-2 and pro-MMP-9 during rat liver regeneration.

Partial hepatectomy triggers a variety of biological phenomena, which culminate in regeneration of the liver mass. Hepatocyte proliferation is a major feature of the regenerating liver after partial hepatectomy. Previous studies in our laboratory suggested that hepatic matrix remodeling might be a prerequisite process for hepatocyte proliferation in the regenerating liver. In the present study we use immunohistochemical staining, Western blot analysis, and gelatin zymography to show that the inactive matrix metalloproteinases (MMPs), pro-MMP-2 and pro-MMP-9, are elevated at 30 minutes and activated at 6 to 12 hours and at 3 to 6 hours, respectively, after hepatectomy. Sham-operated livers did not show an increase in inactive pro-MMP-2 and pro-MMP-9 and did not contain active MMP-2 or MMP-9. To examine whether tissue inhibitor of matrix metalloproteinase-1 (TIMP-1) is changed to regulate the activities of MMPs after partial hepatectomy, the level of TIMP-1 protein was analyzed by both immunohistochemistry and Western blot analysis. The level of TIMP-1 protein appears to increase by 6 to 18 hours, implying that an increase in TIMP-1 regulates activities of active MMP-2 and MMP-9. Taken together, these results suggest that hepatic matrix remodeling is mediated by activated MMPs, which contribute to modulation of the environment surrounding hepatocytes during rat liver regeneration.

Cationic colloidal silica membrane perturbation as a means of examining changes at the sinusoidal surface during liver regeneration.

By employing the cationic colloidal silica membrane density perturbation technique, we examined growth factor receptor and extracellular matrix (ECM) changes at the sinusoidal surface during rat liver regeneration 72 hours after 70% partial hepatectomy (PHx). At this time after PHx, hepatocyte division has mostly subsided, while sinusoidal endothelial cell (SEC) proliferation is initiating, resulting in avascular hepatocyte islands. Because of the discontinuous nature of the surface of liver SEC, ECM proteins underlying the SEC, as well as SEC luminal membrane proteins, are available to absorption to the charged silica beads when the liver is perfused with the colloid. Subsequent liver homogenization and density centrifugation yield two separate fractions, enriched in SECs as well as hepatocyte basolateral membrane-specific proteins up to 50-fold over whole liver lysates. This technique facilitates examination of changes in protein composition that influence or occur as a result of SEC mitogenesis and migration during regeneration of the liver. When ECM and receptor proteins from SEC-enriched fractions were examined by Western immunoblotting, urokinase plasminogen activator receptor, fibronectin, and plasmin increased at the SEC surface 72 hours after PHx. Epidermal growth factor receptor, plasminogen, SPARC (secreted protein, acidic and rich in cysteine, also called osteonectin or BM40), and collagen IV decreased, and fibrinogen subunits and c-Met expression remained constant 72 hours after PHx when compared to control liver. These results display the usefulness of the cationic colloidal silica membrane isolation protocol. They also show considerable modulation of surface components that may regulate angiogenic processes at the end stage of liver regeneration during the reformation of sinusoids.

STAT 1alpha/1beta, STAT 3 and STAT 5: expression and association with c-MET and EGF-receptor in long-term cultures of human hepatocytes.

Long-term cultures of human hepatocytes were maintained serum-free in a chemically defined medium in the presence of HGF and EGF for up to 30 days. STAT 1alpha/1beta, STAT 3, and STAT 5 were present and tyrosine phosphorylated throughout the culture period in the cytosol as well as the nucleus. We show by co-immunoprecipitation that a portion of the cellular pools of STAT 1alpha/1beta and STAT 5 is physically associated with c-MET and EGF-receptor. Co-immunoprecipitation of STAT 3 with STAT 5 did occur in the cytosol but not in the nucleus, suggesting dimerization of the two STAT family members. The observed differences of protein amounts and tyrosine phosphorylation between cytosol and nucleus, the association of STAT proteins with EGF-receptor and c-MET and with each other may all be involved in regulating the activity of the STAT transcription factors. It is intriguing to speculate that STAT 5 may have a modulating role in the regulation of STAT 3 activity.

Hepatocyte growth factor (HGF) and receptor (c-met) in normal and malignant astrocytic cells.

BACKGROUND: Hepatocyte growth factor (HGF) is a multifunctional peptide that binds to a specific receptor, c-met. Both HGF and c-met have been identified in normal brain and on glial tumors. The purpose of this study is to further define the biologic importance of HGF and c-met on normal and malignant glial cells grown in vitro. MATERIALS AND METHODS: Nine human malignant glioma-derived tumor cell cultures and cultures of astrocytes derived from normal brain were examined for c-met and HGF transcripts using Northern blot or RT-PCR analysis. Cellular invasiveness was quantitated by mechanical assay and mitogenesis was determined by cell count. RESULTS: C-met was expressed in five of seven malignant glioma-derived tumor cell cultures and in both normal astrocyte cultures. HGF transcript was not detected in any of the cell cultures. HGF supplementation enhanced invasiveness in c-met positive cell lines and did not alter cellular mitogenesis in the assayed cultures. CONCLUSIONS: These findings suggest that HGF is a potent stimulator of invasiveness in c-met positive malignant glioma-derived tumor cells and is not an active cytokine with regards to in vitro glial cell proliferation. HGF may therefore stimulate glioma cellular invasion in vivo through binding to its receptor and by activating tyrosine kinase secondary messengers.

Growth factor signal transduction immediately after two-thirds partial hepatectomy in the rat.

Liver regeneration after partial hepatectomy (PHx) of the liver serves as a model for studying normal growth factor signals that become aberrant in cancer. Growth factor signals that may play a role in initiating the proliferation of hepatocytes after 70% PHx in the rat were investigated immediately after surgical resection of the liver. Presumptive activity was evaluated by determining the tyrosine phosphorylation state of receptors for epidermal growth factor (EGF) and hepatocyte growth factor (HGF) in the liver after PHx and after sham operation as a control. Under these conditions, it was determined that the EGF receptor was constitutively phosphorylated. EGF receptor tyrosine phosphorylation, however, was increased over basal levels by 60 min after resection. The HGF receptor, c-Met, was minimally phosphorylated in control livers, but a biphasic increase in phosphorylation was observed at 1-5 min after PHx and 60 min postsurgery. A slight increase in c-Met phosphorylation was observed in the sham-operated livers, but the signal was significantly less when compared with that in resected livers. Furthermore, 1 min after PHx, but not sham operation, urokinase-type plasminogen activator (u-PA) and u-PA receptor were observed in the immunoprecipitates of c-Met. Signaling downstream of growth factor receptor activation was also examined. There were no discernible phosphorylation changes in focal adhesion kinase during the early events after surgery in PHx; however, a rapid and sustained increase in the tyrosine phosphorylation of paxillin beginning 1 min after PHx, and a gradual increase in the phosphorylation beginning 5 min postsham operation, were observed. Changes in the activated state of the small GTP-binding protein Rho A and its associated proteins were seen but only after 3 h after PHx. The results indicate that HGF-related signal transduction cascades, which contribute to hepatocyte proliferation, are initiated within one min after PHx.

Matrix-mediated changes in the expression of HNF-4alpha isoforms and in DNA-binding activity of ARP-1 in primary cultures of rat hepatocytes.

Recently, we have developed a culture system in which rat hepatocytes dedifferentiate and proliferate and after the addition of EHS-gel redifferentiate. During both developmental stages HNF-4alpha2 mRNA was more abundant than HNF-4alpha1 mRNA. However, Western blot analysis using COS-7 cell-expressed HNF-4alpha1 and HNF-4alpha2 proteins as standards revealed that (i) HNF-4alpha2 protein was not expressed in dedifferentiated hepatocytes and (ii) either HNF-4alpha2 protein or a highly phosphorylated HNF-4alpha1 protein was the dominating isoform in redifferentiated hepatocytes. The changes in HNF4-isoform expression could not be mimicked by DMSO, suggesting them to be matrix specific. Furthermore, DMSO was less efficient than EHS-gel in reinducing liver-specific gene expression. EHS-gel overlay also led to reduction of ARP-1 DNA binding activity, while overall ARP-1 protein levels did not change. These results suggest that EHS-matrix overlay regulates the expression of different HNF-4alpha isoforms on a posttranscriptional level while ARP-1 DNA binding activity is regulated by posttranslational mechanisms.

Epidermal growth factor- and hepatocyte growth factor-receptor activity in serum-free cultures of human hepatocytes.

BACKGROUND/AIMS: Serum-free primary cultures of hepatocytes are a useful tool to study factors triggering hepatocyte proliferation and regeneration. We have developed a chemically defined serum-free system that allows human hepatocyte proliferation in the presence of epidermal growth factor and hepatocyte growth factor. METHODS: DNA synthesis and accumulation were determined by [3H]thymidine incorporation and fluorometry, respectively. Western blot analyses and co-immunoprecipitations were used to investigate the association of proteins involved in epidermal growth factor and hepatocyte growth factor activation and signaling: epidermal growth factor receptor, hepatocyte growth factor receptor (MET), urokinase-type plasminogen activator and its receptor, and a member of the signal transducer and activator of transcription family, STAT-3. RESULTS: Primary human hepatocytes proliferated under serum-free conditions in a chemically defined medium for up to 12 days. Epidermal growth factor-receptor and MET were present and functional, decreasing over time. MET, urokinase-type plasminogen activator and urokinase-type plasminogen activator receptor co-precipitated to varying degrees during the culture period. STAT-3 co-precipitated with epidermal growth factor-receptor and MET to varying degrees. CONCLUSIONS: Proliferation of human hepatocytes can improve by modification of a chemically defined medium originally used for rat hepatocyte cultures. In these long-term cultures of human hepatocytes, hepatocyte growth factor and epidermal growth factor can stimulate growth and differentiation by interacting with their receptors and initiating downstream signaling. This involves complex formation of the receptors with other plasma membrane components for MET (urokinase-type plasminogen activator in context of its receptor) and activation of STAT-3 for both receptors.

Expression of the human hepatocyte growth factor cDNA in primary cultures of rat hepatocytes.

Hepatocyte growth factor (HGF) and epidermal growth factor (EGF) are primary mitogens for hepatocytes in culture. hepatocytes express the HGF-receptor MET but not HGF itself. To investigate the influence of autocrine HGF expression on the proliferative potential of hepatocytes, primary cultures were submitted to retrovirus-mediated transduction of the human hgf (huHGF) cDNA. Expression of the transduced cDNA revealed a minimum 2-fold increase in HGF-mRNA, whereas expression of the Escherichia coli beta-galactosidase gene remained even. Estimation of huHGF copy numbers showed there was a minimum 4-fold increase, suggesting an increase in the population of transduced cells. Immunoprecipitation of excreted huHGF and growth bioassays proofed that HGF was present and functional. HGF is excreted into the medium and therefore, by diffusion, available to transduced and non-transduced cells. The increase in huHGF-transduced cells suggests that the autocrine pathway as opposed to the paracrine pathway, which are both present at the same time, confers a growth advantage to these cells.

The five amino acid-deleted isoform of hepatocyte growth factor promotes carcinogenesis in transgenic mice.

Hepatocyte growth factor (HGF) is a polypeptide with mitogenic, motogenic, and morphogenic effects on different cell types including hepatocytes. HGF is expressed as two biologically active isotypes resulting from alternative RNA splicing. The roles of each HGF isoform in development, liver regeneration and tumorigenesis have not yet been well characterized. We report the generation and analysis of transgenic mice overexpressing the five amino acid-deleted variant of HGF (dHGF) in the liver by virtue of an albumin expression vector. These ALB-dHGF transgenic mice develop normally, have an enhanced rate of liver regeneration after partial hepatectomy, and exhibit a threefold higher incidence of hepatocellular carcinoma (HCC) beyond 17 months of age. Moreover, overexpression of dHGF dramatically accelerates diethyl-nitrosamine induced HCC tumorigenesis. These tumors arise faster, are significantly larger, more numerous and more invasive than those appearing in non-transgenic littermates. Approximately 90% of female dHGF-transgenic mice had multiple macroscopic HCCs 40 weeks after injection of DEN; whereas the non-transgenic counterparts had only microscopic nodules. Liver tumors and cultured tumor cell lines from dHGF transgenics showed high levels of HGF and c-Met mRNA and protein. Together, these results reveal that in vivo dHGF plays an active role in liver regeneration and HCC tumorigenesis.

Morphogenetic events in mixed cultures of rat hepatocytes and nonparenchymal cells maintained in biological matrices in the presence of hepatocyte growth factor and epidermal growth factor.

Hepatocytes were grown in chemically defined hepatocyte growth medium (HGM) containing hepatocyte growth factor (HGF) and epidermal growth factor (EGF) on collagen-coated polystyrene beads in roller bottle cultures, forming clusters of beads, and proliferating hepatocytes and nonparenchymal cells, including fenestrated endothelium-forming vascular structures. Desmin-positive cells surrounded hepatocytes. Collagen types I and III were deposited in a diffuse manner whereas collagen type IV surrounded the clusters of the epithelial cells, forming a basement membrane. When the mixed cell clusters were implanted in Matrigel (Collaborative Research, Bedford, MA), hepatocytes grew in three dimensions, forming plates and ducts. Many single, long plates of hepatocytes were seen, suggesting progressive linear assembly guided by hepatocyte specific structural parameters. HGF, EGF, and transforming growth factor-alpha (TGF-alpha) enhance these phenomena. HGF plus EGF elicited maximal response. TGF-beta1 suppressed formation of the ducts and plates. Within three months in Matrigel, the cultures established monolayers composed of plates, ducts, and a well-delineated canalicular network. The mixed cultures expressed albumin, A1AT, AFP, transferrin, and CYPIIB1. Following implantation of the cell clusters in Matrigel, there was decreased expression of c-met, urokinase, urokinase receptor, and TGF-beta1. Electron microscopy showed differentiated hepatocytes with nearly normal ultrastructure. The proliferating cell nuclear antigen (PCNA) labeling index was high (more than 80%) whereas the Bromo-deoxyaridine labeling index of ongoing DNA synthesis varied from 10% to 15%. These results show that the mixed cultures of proliferating hepatocytes and nonparenchymal cells can reproduce the hallmark structures of hepatic histological architecture while maintaining differentiation and the capacity to proliferate. (HEPATOLOGY 1999;29:90-100.)