Hepatocellular carcinoma (HCC) is a common tumor world-wide with extremely poor prognosis. Recent studies have shown that inositol hexaphosphate (IP6), a naturally occurring carbohydrate, has novel anti-cancer function in various in vitro and in vivo models. The aim of this study was to assess whether IP6 could inhibit the growth of human hepatocellular carcinoma. We treated HepG2, a human liver cancer cell line in vitro with IP6 and evaluated its effect on growth and differentiation. IP6 treatment of HepG2 cells caused a dose-dependent growth inhibition. Compared to other cancer cell lines, HepG2 cells were quite sensitive to IP6, IC50 (50% inhibition of cell growth) of IP6 being < 1.0 mM (0.338 mM). Treatment with IP6 decreased the ability of HepG2 cells to form colonies, as assessed in the plating efficiency assay. Morphological changes induced by IP6 were consistent with differentiation of HepG2 cells. Exposure of HepG2 cells to IP6 drastically decreased the rate of production of alpha-fetoprotein (AFP), a tumor marker of HCC, indicating also that IP6 treatment leads to differentiation of malignant liver cells. Further, IP6 treatment caused a decreased expression of mutant p53 protein in HepG2 cells, with no significant change in the expression of wild-type p53. The expression of p21WAF1 protein was increased by 1.5 fold, as determined by immunocytochemical staining and ELISA assay. These data demonstrate that IP6 inhibits the growth, and induces differentiation, and a less aggressive phenotype of HepG2 cells, suggesting a role of IP6 in the treatment of HCC.

Normal human liver tissue and cultured human hepatocytes are valuable models to study xenobiotic metabolism and toxicity, but they only have a limited in vitro life-span and are not readily available. This report describes the establishment of replicative cultures of human adult liver epithelial cells in serum-free medium. The longevity of three of these cultures, derived from different donors, was extended by introduction of the simian virus 40 large T antigen gene. Two cell lines, THLE-2 and -3, established with a recombinant simian virus 40 large T antigen virus have undergone > 100 population doublings, are nontumorigenic when injected into athymic nude mice, have near-diploid karyotypes, and do not express alpha-fetoprotein. The cells express cytokeratin 18 and albumin in early passage, whereas higher-passage cells in logarithmic-phase growth also express cytokeratin 19. THLE-2 and -3 cells metabolize benzo[a]pyrene, N-nitrosodimethylamine, and aflatoxin B1 to their ultimate carcinogenic metabolites that adduct DNA, which indicates functional cytochrome P450 pathways. Other enzymes involved in metabolism of chemical carcinogens, such as epoxide hydrolase, NADPH cytochrome P450 reductase, superoxide dismutase, catalase, glutathione S-transferases, and glutathione peroxidase are also retained by THLE cells. Thus, these immortalized human liver cells constitute an in vitro model for pharmacotoxicological studies and for the investigation of etiology and pathogenesis of human hepatocellular carcinoma.

The hepatocellular cytoprotective effects of 16,16-dimethyl prostaglandin E2 (dmPGE2), an analogue of PGE2, were investigated using primary cultures of rat hepatocytes and aflatoxin B1 as the hepatotoxin. Lactic dehydrogenase (LDH) release by hepatocytes was used as an index of hepatotoxicity. When aflatoxin-treated hepatocytes were co-cultured with 16,16-dmPGE2 (0.01-0.5 micrograms/mL) LDH release was significantly reduced and ultrastructural changes of hepatocellular injury were markedly diminished. The magnitude of the cytoprotective effect was not dependent on the concentration of the prostaglandin over the range tested. A significant cytoprotective effect was also induced when hepatocellular cyclic AMP (cAMP) levels were increased by the addition of dibutyl-cAMP. In contrast to 16,16-dmPGE2, PGF2 alpha Tromethamine, an analogue of PGF2 alpha, which does not stimulate cAMP, induced insignificant changes in cytoprotection. These findings indicate that only a low concentration of 16,16-dmPGE2 (> or = 0.01 micrograms/mL) is necessary to induce a maximal hepatocellular cytoprotective effect and suggest that this effect may be dependent on activation of cAMP.

Exploratory experiments indicate that media containing lipids, phosphoethanolamine, epidermal growth factor, insulin, cholera toxin, bovine pituitary extract, chemically denatured serum, and triiodothyronine will support replicative cultures of normal liver epithelial cells obtained from adult Rhesus monkey and human donors. In addition, we have extended the culture population doubling potential of the human liver epithelial cells by their transfection with a plasmid containing the SV40 virus T-antigen gene. The T-antigen gene-containing cells continued to express albumin through 40 population doublings. Finally, results of preliminary experiments suggest that it may be possible to induce human liver epithelial cells to undergo differentiation to hepatocyte-like cells either by injecting them into the spleen of an athymic nude mouse or by incorporating them into a collagen "tissue equivalent" matrix.

Ultrastructural changes commonly observed in liver cells of rodents exposed to carcinogens in vivo can be induced in hepatocytes exposed to carcinogens in vitro. Human, rat and mouse hepatocytes in primary culture were treated with actinomycin D, aflatoxin B1 (AFB1) and dimethylnitrosamine (DMN). These cultured hepatocytes were examined for ultrastructural alterations following carcinogen exposure for 24 h. Similar to the effects on liver cells in vivo, the most prominent change was a segregation of the nucleolar components. Human, rat and mouse hepatocytes, dosed with 7.9 X 10(-8) M actinomycin D, developed nucleolar segregation in 86%, 98% and 55% of cells, respectively. When incubated with 3.2 X 10(-6) M AFB1, 60% of human and 84% of rat hepatocytes developed nucleolar segregation. However, exposures of mouse hepatocytes less than or equal to 3.2 X 10(-5) M of AFB1 failed to induce segregation of the nucleolus. DMN administered at a dose of 2.0 X 10(-2) M caused segregation in 11% of the rat hepatocytes and in 60% of the mouse hepatocytes. Distinct nucleolar segregation did not occur in human hepatocytes until they were exposed to a concentration of 5.0 X 10(-2) M DMN (31%). Actinomycin D, AFB1, DMN, as well as other compounds that bind to DNA and interfere with template activity cause nucleolar segregation. Morphologic changes observed in cultured rat and mouse hepatocytes correlate well with in vivo experiments with regard to the relative sensitivity of rats and mice to toxicological effects of these carcinogens. Thus, hepatocyte cultures may provide a realistic system to determine the sensitivity of human liver cells to carcinogens.

DNA binding levels were determined and compared in cultured hepatocytes from male and female rats as well as other animal species following exposure to aflatoxin B1 (AFB1) or 2-acetylaminofluorene (2-AAF). When human, rat (both male and female) and mouse hepatocytes in primary culture were exposed to 2.0 X 10(-7) M [3H]AFB1 (sp. act. 2.63 microCi/nmol) for 24 h, male rat hepatocytes had the highest degree of [3H]AFB1-DNA binding (203 pmol/mg DNA) and human hepatocytes contained the next highest binding level (42 pmol/mg DNA). Hepatocytes from female rats contained 38 pmol/mg DNA while cultured mouse hepatocytes contained only 1.4 pmol/mg DNA. When the same dose of [3H]AFB1 was administered to the cultured male rat hepatocytes at 24 h, 48 h, 72 h and 1 week after seeding, and incubated for 24 h, the DNA binding levels were 189, 175, 76, 75 pmol/mg DNA respectively. In parallel experiments to the cultured male rat hepatocytes above, the AFB1-DNA binding levels in the cultured female hepatocytes were 42, 41, 37 and 34 pmol/mg DNA respectively. Human, male and female rat hepatocytes in primary culture were exposed to 5.2 X 10(-5) M 2-acetylamino[9-14C]fluorene (sp. act. 0.0094 microCi/nmol) for 24 h. It was determined that male rat hepatocytes contained the highest amount of radioactively labeled 2-AAF bound to their DNA (1.57 nmol/mg DNA), female rat hepatocytes contained 0.62 nmol/mg DNA and human hepatocytes contained 0.29 nmol/mg DNA. Results from our in vitro hepatocyte culture system correlate well with in vivo animal studies dealing with species and sex differences in DNA binding and carcinogenic susceptibility. This indicates that hepatocytes in vitro maintain many of the biological properties necessary for carcinogen response similar to liver cells in vivo. In addition, comparison of genotoxic effect in cultured hepatocytes from animals as well as humans may be useful in evaluating carcinogenic potential of xenobiotics in human liver.

Human, rat, and mouse hepatocytes in primary culture were treated with aflatoxin B1 (AFB1) and examined for ultrastructural alterations. As early as 1 h following in vitro exposure to AFB1, there were ultrastructural changes in the nuclei of rat and human hepatocytes. The most prominent change in the nuclei was a segregation of nucleolar components that resembled the segregation in liver cells of rats exposed to AFB1 in vivo. The nucleolar segregations were developed by incubating rat hepatocytes for 24 h in a medium containing as little as 0.01 micrograms of AFB1 per ml. The minimum concentration to induce the same change in human hepatocytes was 0.1 micrograms/ml. No distinct nucleolar alteration was observed in mouse hepatocytes incubated in a medium containing 10 micrograms of AFB1 per ml. Irregular nuclear chromatin condensation also developed in the cells exposed to a higher concentration of AFB1, whereas little damage was observed in mitochondria and lysosomes. The similarity in morphological changes between our in vitro model and in vivo models previously investigated indicates that the hepatocytes in primary culture maintain the biological properties necessary for carcinogen responses similar to liver cells in vivo. In addition, the morphological changes in cultured rat and mouse hepatocytes induced by AFB1 correlate with in vivo experiments insofar as mice are relatively resistant, whereas rats are sensitive to AFB1 carcinogenesis. Thus, cultured hepatocyte systems may be a valuable tool to study genetic damage which may lead to hepatocellular carcinomas in human and animal livers.

Human livers were removed at immediate autopsy (IA) from brain death patients within 1 h after cessation of cardiac function. Viable hepatocytes were isolated successfully from these IA livers by perfusion of an intack lobe with collagenase or by digestion of a small tissue wedge with collagenase-dispase. The yields of hepatocytes ranged from 1 to 3 × 106 cells/g liver in the five cases studied. Approximately 70 to 90% of the cells excluded trypan blue dye. In the isolated hepatocytes, 632 pmol/mg protein of cytochromep 450 and 536. pmol/mg protein cytochromeb 5 were measured. The cells attached to the dishes in 4 h and produced monolayer cultures with a high success rate. The cells maintained in primary cultures for several days and developed ultrastructural features characteristic of human hepatocytes in vivo. The cultured hepatocytes can hydroxylate benzo[a]pyrene, conjugate the metabolites, and have a benzo[a]pyrene hydroxylase activity of 48.7 pmol/mg DNA per h, which is comparable to that of rat hepatocytes. The liver cells repaired DNA damage caused by exposures to aminofluorene and acetylaminofluorene in culture.