Non-transformed Hepatocytes Research Articles

Development of a Non-Transformed Human Liver Cell Line with Differentiated-Hepatocyte and Urea-Synthetic Functions: Applicable for Bioartificial Liver

There is a need to develop human hepatocyte cell lines which retain both replicating capacity and highly differentiated functions to facilitate the development of an efficient bioartificial liver. The present study was undertaken to differentiate, using sodium butyrate, the actively replicating immortalized human liver cell line. The effects of butyrate on cell growth and cell cycle were analyzed, and the albumin synthesis, cytochrome P450 and ammonia-detoxifying activity of the butyrate-treated cells were measured. Butyrate treatment resulted in G2/M arrest of the cell cycle and polygonal changes in the cell morphology. Neither the control nor the butyrate-treated cells showed transformed characteristics. Butyrate treatment increased the amount of albumin secretion, cytochrome P450 activity, and the urea production rate of the cells. The present study provides non-transformed human hepatocytes, which can replicate unlimitedly and then restore differentiated hepatocyte-specific functions by butyrate, and therefore, have applications for the development of an efficient bioartificial liver.

Positive and negative regulation of JNK1 by protein kinase C and p42 MAP kinase in adult rat hepatocytes

The role of protein kinase C (PKC) and p42 MAP kinase signaling in the regulation of proliferation and apoptosis was investigated in freshly isolated and primary cultured rat hepatocytes. Acute treatment of freshly isolated hepatocytes with phenylephrine and EGF caused rapid phasic activations of p42 MAP kinase and JNK1. Acute pre-treatment of hepatocytes with the PKC inhibitors sphingosine, chelerythrine and bis-indolylmaleimide abolished the ability of phenylephrine, but not EGF, to activate p42 MAP kinase and JNK1. Acute pre-treatments with all of the PKC inhibitors alone increased JNK1 basal activity ∼2-fold. Acute treatments of primary cultures of hepatocytes with an inhibitor of MEK1 activation (PD98059) also caused inhibition of p42 MAP kinase and a ∼2-fold activation of JNK1. These data demonstrate that PKC can function as both a proximal activator and a distal inhibitor of signaling through the JNK1/SAP kinase pathway. Treatments (4 h) of primary cultured hepatocytes with sphingosine, chelerythrine, bis-indolylmaleimide and PD98059 did not induce apoptosis as judged by propidium iodide staining. Similar acute treatments of HepG2 cells rapidly induced cell death. These data demonstrate that acute inhibition of either PKC or p42 MAP kinase function is sufficient to rapidly induce apoptosis in transformed, but not in non-transformed hepatocytes.

Analysis of the tumorigenicity of the X gene of hepatitis B virus in a nontransformed hepatocyte cell line and the effects of cotransfection with a murine p53 mutant equivalent to human codon 249.

Chronic infection with hepatitis B virus (HBV) is associated with the development of human hepatocellular carcinoma (HCC). One of the HBV genes, HBx, may have transforming potential, but this issue is still the subject of controversy. One of the major difficulties in addressing this question is the lack of a suitable in vitro model. We used a nontransformed, differentiated murine hepatocyte cell line (AML12) to transfect the HBx gene and examine its transforming capabilities. Because mutations of the p53 gene, in particular at codon 249, have been implicated in HCC development in geographical areas with high incidence of the tumor, we also studied the putative cooperative role in transformation between HBx and mutated p53 by cotransfecting HBx with a murine p53 mutant equivalent to human ser249 (ser246p53). Transfection with HBx plasmids containing the HBx gene under the control of two different promoters resulted in fewer colonies than in control plasmids. The toxic effect of HBx on colony formation was abolished by cotransfection with 246p53, suggesting that the inhibitory effect requires functionally intact p53. Clonal cell lines that stably expressed HBx messenger RNA (mRNA) (HBX lines) were tested for their growth characteristics and their ability to grow in soft agar and form tumors in nude mice. At passages 19-27 after transfection, one of four HBx-expressing lines showed the capacity for anchorage-independent growth in soft agar and produced poorly differentiated hepatocellular carcinomas in 8 of 13 sites of injection in nude mice. HBX lines as well as clonal cell lines of cells transfected with 246p53 (246 cell lines), cotransfected with HBx and 246p53 (246x lines) or transfected with control plasmids, were analyzed by flow cytometry to determine the fraction of cells in S phase (SPF). 246 and 246X lines had similar SPFs that were approximately twofold greater than control or HBX lines. 246x lines showed morphological changes in culture such as marked cellular heterogeneity, cell crowding, and the presence of multinucleated giant cells, but their tumorigenicity was not increased compared with the HBX lines. These data show that HBx has a weak tumorigenicity in murine hepatocytes and that the addition of mutation of p53 at codon 249 to HBx expression does not increase tumorigenicity in AML12 cells.

Analysis of the tumorigenicity of the X gene of hepatitis B virus in a nontransformed hepatocyte cell line and the effects of cotransfection with a murine p53 mutant equivalent to human codon 249

Chronic infection with hepatitis B virus (HBV) is associated with the development of human hepatocellular carcinoma (HCC). One of the HBV genes, HBx, may have transforming potential, but this issue is still the subject of controversy. One of the major difficulties in addressing this question is the lack of a suitable in vitro model. We used a nontransformed, differentiated murine hepatocyte cell line (AML12) to transfect the HBx gene and examine its transforming capabilities. Because mutations of the p53 gene, in particular at codon 249, have been implicated in HCC development in geographical areas with high incidence of the tumor, we also studied the putative cooperative role in transformation between HBx and mutated p53 by cotransfecting HBx with a murine p53 mutant equivalent to human ser249 (ser246p53). Transfection with HBx plasmids containing the HBx gene under the control of two different promoters resulted in fewer colonies than in control plasmids. The toxic effect of HBx on colony formation was abolished by cotransfection with 246p53, suggesting that the inhibitory effect requires functionally intact p53. Clonal cell lines that stably expressed HBx messenger RNA (mRNA) (HBX lines) were tested for their growth characteristics and their ability to grow in soft agar and form tumors in nude mice. At passages 19-27 after transfection, one of four HBx-expressing lines showed the capacity for anchorage-independent growth in soft agar and produced poorly differentiated hepatocellular carcinomas in 8 of 13 sites of injection in nude mice. HBX lines as well as clonal cell lines of cells transfected with 246p53 (246 cell lines), cotransfected with HBx and 246p53 (246x lines) or transfected with control plasmids, were analyzed by flow cytometry to determine the fraction of cells in S phase (SPF). 246 and 246X lines had similar SPFs that were approximately twofold greater than control or HBX lines. 246x lines showed morphological changes in culture such as marked cellular heterogeneity, cell crowding, and the presence of multinucleated giant cells, but their tumorigenicity was not increased compared with the HBX lines. These data show that HBx has a weak tumorigenicity in murine hepatocytes and that the addition of mutation of p53 at codon 249 to HBx expression does not increase tumorigenicity in AML12 cells. (Hepatology 1996 Nov;24(5):1024-33)

Introduction of a murine p53 mutation corresponding to human codon 249 into a murine hepatocyte cell line results in growth advantage, but not in transformation

The p53 gene is frequently mutated in human tumors; in hepatocellular carcinomas, there is a high frequency of a specific mutation at codon 249 in regions with significant aflatoxin exposure. To assess the role of this p53 mutation in the development of hepatocellular carcinoma, a mutant murine p53 gene, p53ser246, which corresponds to human codon 249, was transfected into a differentiated, nontransformed hepatocyte cell line AML12. Expression of p53ser246 in this line resulted in a growth advantage when compared with either a control vector (which contains a large p53 deletion) or with a different p53 mutant, val135, not found in hepatocellular carcinoma. Overall, there was a threefold increase in colony formation after transfection with p53ser246 as compared with the control or p53val135 vectors, and the p53ser246 plates developed consistently larger colonies. Whereas clones expressing the control or p53val135 constructs showed no significant morphological changes, clones expressing p53ser246 showed increased heterogeneity (large multinucleated cells and areas of small crowded cells) without focus formation. In addition, the ser246 mutation imparted a growth advantage in serum-free media, suggesting less dependence on specific factors present in serum. None of the mutant p53 or control lines were capable of growth in soft agar or tumor formation in nude mice. Thus in this model, in which endogenous wild-type p53 expression is retained, a high level of mutant p53 expression is not sufficient to transform hepatocytes. Our findings indicate that p53ser246 has effects on hepatocytes that may result in a clonal growth advantage and suggest that additional factors are required for the development of hepatocellular carcinoma. (Hepatology 1995; 22:1279-1288.).

Introduction of a murine p53 mutation corresponding to human codon 249 into a murine hepatocyte cell line results in growth advantage, but not in transformation

The p53 gene is frequently mutated in human tumors; in hepatocellular carcinomas, there is a high frequency of a specific mutation at codon 249 in regions with significant aflatoxin exposure. To assess the role of this p53 mutation in the development of hepatocellular carcinoma, a mutant murine p53 gene, p53ser246, which corresponds to human codon 249, was transfected into a differentiated, nontransformed hepatocyte cell line AML12. Expression of p53ser246 in this line resulted in a growth advantage when compared with either a control vector (which contains a large p53 deletion) or with a different p53 mutant, val135, not found in hepatocellular carcinoma. Overall, there was a threefold increase in colony formation after transfection with p53ser246 as compared with the control or p53val135 vectors, and the p53ser246 plates developed consistently larger colonies. Whereas clones expressing the control or p53val135 constructs showed no significant morphological changes, clones expressing p53ser246 showed increased heterogeneity (large multinucleated cells and areas of small crowded cells) without focus formation. In addition, the ser246 mutation imparted a growth advantage in serum free media, suggesting less dependence on specific factors present in serum. None of the mutant p53 or control lines were capable of growth in soft agar or tumor formation in nude mice. Thus in this model, in which endogenous wild-type p53 expression is retained, a high level of mutant p53 expression is not sufficient to transform hepatocytes. Our findings indicate that p53ser246 has effects on hepatocytes that may result in a clonal growth advantage and suggest that additional factors are required for the development of hepatocellular carcinoma.

Uptake of chylomicron remnants and hepatic lipase-treated chylomicrons by a non-transformed murine hepatocyte cell line in culture

AML 12 is a recently established differentiated, non-transformed hepatocyte cell line derived from mice transgenic for transforming growth factor α (Wu et al. (1994) Proc. Natl. Acad. Sci. 91, 674–678). The ability of these cells to take up [ 3H]cholesterol-labeled in vivo-generated chylomicron remnants, as well as [ 3H]cholesterol-labeled chylomicrons treated with hepatic lipase in vitro was investigated. Both types of lipoprotein particles were taken up by the AML hepatocytes at a much faster rate than intact chylomicrons, and in a saturable and specific manner. Chylomicrons treated with hepatic lipase in vitro competed with in vivo-generated chylomicron remnants for uptake by the AML hepatocytes, and the uptake of both types of lipoproteins was inhibited by lactoferrin, suggesting that they share the same process of cellular recognition and uptake. It is suggested that hepatic lipase-treated chylomicrons may be valuable in studies aimed at gaining a better understanding of the processes involved in the hepatic recognition and uptake of chylomicron remnants. AML hepatocytes, which can be maintained as replicating, untransformed, and differentiated under standard culture conditions, may be useful and practical for such studies.

Colony isolation and secondary culture of fetal porcine hepatocytes on STO feeder cells.

The secondary culture of non-transformed parenchymal hepatocytes has not been possible. STO feeder cell-dependent secondary cultures of fetal pig hepatocytes were established by colony isolation from primary cultures of 26-d fetal livers. The liver cells had the typical polygonal morphology of parenchymal hepatocytes. They also spontaneously differentiated to form small biliary canaliculi between individual cells or progressed further to large multicellular duct-like structures or cells undergoing gross lipid accumulation and secretion. The secondary hepatocyte cultures expressed alpha-fetoprotein (AFP), albumin, and beta-fibrinogen mRNA, and conditioned medium from the cells contained elevated levels of transferrin and albumin. STO feeder cell co-culture may be useful for the sustainable culture of hepatocytes from other species.

Establishment and characterization of differentiated, nontransformed hepatocyte cell lines derived from mice transgenic for transforming growth factor alpha.

Hepatocytes are extensively used in studies of gene regulation but cannot be maintained in long-term culture as replicating, differentiated cells while remaining nontumorigenic. We have derived two hepatocyte lines from livers of transgenic mice overexpressing transforming growth factor alpha, a potent hepatocyte mitogen, which overcome these limitations. The transgenic hepatocytes were maintained for > or = 2 months in serum-supplemented primary culture and gave rise to cell lines, of which two (AML12 and AML14) have been cultured for > 1.5 years (> 80 passages). Both lines have typical hepatocyte features such as peroxisomes and bile canalicular-like structures, do not grow in soft agar, and are nontumorigenic in nude mice. Like normal hepatocytes, AML cells express high levels of mRNA for serum (albumin, alpha 1-antitrypsin, and transferrin) and gap junction (connexins 26 and 32) proteins, secrete albumin, and contain solely isozyme 5 of lactate dehydrogenase. After extensive passaging, AML12 cells continue to strongly coexpress hepatocyte connexin mRNAs but do not display nonparenchymal cell markers. Although mRNA levels for some serum proteins progressively fall, high expression in late AML12 cultures may be regained by passage in serum-free medium. The AML14 line loses expression of both differentiated markers and transgene mRNA with extended passaging, and hepatocytic traits are only partially restored by passage in serum-free medium. These differentiated, nontumorigenic cell lines should serve as models in which to study hepatocyte growth and differentiation.

Cellular and molecular mechanisms of hepatocarcinogenesis.

For colorectal carcinomas, the rate of tumor development is proportional to the fourth to sixth power of elapsed time, suggesting that four to six independent events are necessary. Although similar calculations have not been made for HBV-associated HCCs, it is likely that this is also the case for HCCs, since individuals with persistent HBV infection do not usually develop HCC until they are 45 or greater years old. As evidence for specific genetic and epigenetic changes in HCCs accumulate, the important players in multistep hepatocarcinogenesis are becoming clearer. However, even though Myc family oncogenes are clearly implicated in woodchuck HCC, similar integrations have not been found in human HCCs. Therefore, although rodent and human systems have many similarities, we must realize that important differences may also exist. Regarding tumor suppressor genes, the evidence for p53 alterations in HCC is strong. A growing body of evidence suggests further that alterations in the retinoblastoma gene and one or more tumor suppressor genes on chromosome 11 are also involved in HCC. HBV integrations may certainly play a role in the generation of chromosome aberrations leading to loss of tumor suppressor alleles, since chromosomes 11 and 17 are the most common integration sites. Finally, the role of X proteins as participants in malignant transformation has been demonstrated for certain immortalized, nontransformed hepatocytes. Altered autocrine mechanisms of cell growth control, possibly involving IGF-II, are clearly implicated in HCC. Paracrine mechanisms for the control of hepatocyte growth and differentiated functions may also be altered as a result of the synthesis and secretion of a complex array of interleukins, HGF, and basic and acidic FGFs by cells in the inflammatory and cirrhotic lesions of precancerous livers. Whether the order of molecular changes in the hepatocyte is important for malignant progression is presently not clear. What is clear, however, is that hepatocarcinogenesis involves alterations in the concerted action of protooncogenes, growth factor, and tumor suppressor genes. How these factors interact will provide a more complete understanding of the mechanism or mechanisms of hepatic oncogenesis.

Homocysteine export from cells cultured in the presence of physiological or superfluous levels of methionine: Methionine loading of non‐transformed, transformed, proliferating, and quiescent cells in culture

Determination of the transient increase in plasma homocysteine following administration of excess methionine is an established procedure for the diagnosis of defects in homocysteine metabolism in patients. This so-called methionine loading test has been used for 25 years, but the knowledge of the response of various cell types to excess methionine is limited. In the present paper we investigated homocysteine export from various cell types cultured in the presence of increasing concentrations (15-1,000 microM) of methionine. For comparison of homocysteine export, the export rates per million cells were plotted versus cell density for proliferating cells, and versus time for quiescent cells. The homocysteine export from growing cells was greatest during early to mid-exponential growth phase, and then decreased as a function of cell density. The export rate was higher from phytohemagglutinin-stimulated than non-stimulated lymphocytes, and higher from proliferating than from quiescent fibroblasts. The hepatocytes showed highest export rate among the cell types investigated. The enhancement of homocysteine export by excess methionine ranged from no stimulation to marked enhancement, depending on cell type investigated, and three different response patterns could be distinguished: 1) quiescent fibroblasts and growing murine lymphoma cell showed no significant increase in homocysteine export following methionine loading; export from human lymphocytes was only slightly enhanced in the presence of excess methionine; 2) the homocysteine export from proliferating hepatoma cells and benign and transformed fibroblasts was stimulated three to eightfold by increasing the methionine concentration in the medium from 15 to 1,000 microM; and 3) the response to methionine loading was particularly increased (about 15-fold) in non-transformed primary hepatocytes in stationary culture. The results outline a potentially useful procedure for the comparison of homocysteine export during cell growth in the presence of various concentrations of methionine. The results are discussed in relation to the special feature of homocysteine metabolism in various cell types and tissues including liver, and to the possible source of plasma homocysteine following methionine loading in vivo.

Target polypeptide of a carcinogen is associated with normal mitosis and carcinogen-induced hyperplasias in adult hepatocytes.

Normal rat liver cytosol was found previously to contain a 14,000-dalton polypeptide that is the principal target of the carcinogen N-2-fluorenylacetamide (2-acetyl-aminofluorene) early during hepatocarcinogenesis. By using antiserum that identifies the 14,000-dalton polypeptide in liver cytosol, an immunologically related 17,500-dalton polypeptide was shown to be present in isolated normal liver nuclei, tightly bound to chromatin. We report here that the 14,000-dalton polypeptide is associated with cell multiplication in normal adult hepatocytes. Of 150 hepatocytes in five stages of mitosis, all invariably displayed a greatly increased concentration of the 14,000-dalton polypeptide in cytoplasm, compared to hepatocytes in interphase, by immunostaining with peroxidase-antiperoxidase complex. In contrast, mitosis did not appear to affect the level of the 17,500-dalton polypeptide in nuclei. A small population of high-polyploid hepatocytes with large nuclei had elevated intensities of cytoplasmic immunostain that approached that in mitotic cells. The increased level of discernible 14,000-dalton polypeptide target of the carcinogen in cytoplasm is a marker of the rare mitotic hepatocytes in normal adult rat liver. Ingestion of the liver carcinogen N-2-fluorenylacetamide for 5 wk or an azocarcinogen, 3'-methyl-4-dimethylaminoazobenzene, for 4 wk brought about early foci of hyperplastic hepatocytes in which there was a great overload of detectable 14,000-dalton target polypeptide in their cytoplasm and a near absence of the 17,500-dalton polypeptide in most nuclei. In contrast, the cytoplasm and nuclei in surrounding morphologically nontransformed hepatocytes of livers of the carcinogen-fed rats immunostained to a much smaller degree than in normal adult hepatocytes. Removal of the azo-dye carcinogen for 4 wk resulted in the disappearance of most hyperplastic foci. A few foci did persist and had similar apparent abnormal levels of the two polypeptides. The high concentration of discernible 14,000-dalton target polypeptide in the cytoplasm of hepatocytes in the hyperplastic foci correlates with their known proliferation, whereas the low cytoplasmic level in the surrounding hepatocytes is consistent with the known inhibition of their mitosis by these carcinogens. The collected evidence appears to joint together a chemical carcinogen, a cytoplasmic principal target polypeptide, a chromatin-bound polypeptide, mitosis in normal adult hepatocytes, early and persistent hyperplastic foci caused by carcinogens, and the inhibition of mitosis by carcinogens in surrounding parenchyma in preneoplastic livers.