A retinoblastoma-binding protein that affects cell-cycle control and confers transforming ability.

The retinoblastoma (RB) gene is one of the most extensively studied tumour-suppressor genes. Deletion or inactivation of both RB alleles is an essential, rate-limiting step in the formation of retinoblastoma and osteosarcoma that arise in families that carry mutant RB (ref. 2). RB inactivation is also found in other human tumours. Whereas loss of RB function is associated with the loss of cellular proliferative control, introduction of a wild-type RB can suppress cell growth and tumorigenicity. Thus, identification of factors that interfere with and/or control the function of the RB protein is critical for understanding both cell-cycle control and oncogenesis. Here we describe a new gene, Bog (for B5T over-expressed gene), which was identified and shown to be overexpressed in several transformed rat liver epithelial (RLE) cell lines resistant to the growth-inhibitory effect of TGF-beta1, as well as in primary human liver tumours. The Bog protein shares homology with other retinoblastoma-binding proteins and contains the Rb-binding motif LXCXE. Using the yeast two-hybrid system and co-immunoprecipitation, we demonstrated that Bog binds to Rb. In vivo, Bog/Rb complexes do not contain E2F-1, and Bog can displace E2F-1 from E2F-1/Rb complexes in vitro. Overexpression of Bog in normal RLE cells conferred resistance to the growth-inhibitory effect of TGF-beta1. Furthermore, normal RLE cells are rapidly transformed when Bog is continuously overexpressed and form hepatoblastoma-like tumours when transplanted into nude mice. These data suggest that Bog may be important in the transformation process, in part due to its capacity to confer resistance to the growth-inhibitory effects of TGF-beta1 through interaction with Rb and the subsequent displacement of E2F-1.

Adenovirus-mediated in vivo gene transfer and expression in normal rat liver.

Replication deficient, recombinant adenovirus (Ad) vectors do not require target cell replication for transfer and expression of exogenous genes and thus may be useful for in vivo gene therapy in hepatocytes. In vitro, primary cultures of rat hepatocytes infected with a recombinant Ad containing a human alpha 1-antitrypsin cDNA (Ad-alpha 1AT) synthesized and secreted human alpha 1AT for 4 weeks. In rats, in vivo intraportal administration of a recombinant Ad containing the E. coli lacZ gene, was followed by expression of beta-galactosidase in hepatocytes 3 days after infection. Intraportal infusion of Ad-alpha 1AT produced detectable serum levels of human alpha 1AT for 4 weeks. Thus, targeted gene expression has been achieved in the liver, albeit at low levels, suggesting that adenovirus vectors may be a useful means for in vivo gene therapy in liver disorders.

Epigenetic regulation of cancer stem cells in liver cancer: current concepts and clinical implications.

The two dominant models of carcinogenesis postulate stochastic (clonal evolution) or hierarchic organization of tumor (cancer stem cell model). According to the latter, at the germinal center of tumor evolution is a cancer stem cell (CSC) which, similar to normal adult stem cells, possesses the capacity of self-renewal and a differentiation potential. Over the past few years, compelling evidence has emerged in support of the hierarchic cancer model for many solid tumors including hepatocellular cancers. The CSCs are posited to be responsible not only for tumor initiation but also for the generation of distant metastasis and relapse after therapy. These characteristics are particularly relevant for a multi-resistant tumor entity like human hepatocellular carcinoma and may herald a paradigm shift in the management of this deadly disease. Identification and detailed characterization of liver CSCs is therefore imperative for improving prevention approaches, enhancing early detection, and extending the limited treatment options. Despite the current progress in understanding the contribution of CSCs to the generation of heterogeneity of tumors, the molecular complexity and exact regulation of CSCs is poorly understood. This review focuses on the genetic and epigenetic mechanisms that regulate and define the unique CSC properties with an emphasis on key regulatory pathways of liver CSCs and their clinical significance.

Expression of hepatic transcription factors during liver development and oval cell differentiation.

The oval cells are thought to be the progeny of a liver stem cell compartment and strong evidence now exists indicating that these cells can participate in liver regeneration by differentiating into different hepatic lineages. To better understand the regulation of this process we have studied the expression of liver-enriched transcriptional factors (HNF1 alpha and HNF1 beta, HNF3 alpha, HNF3 beta, and HNF3 gamma, HNF4, C/EBP, C/EBP beta, and DBP) in an experimental model of oval cell proliferation and differentiation and compared the expression of these factors to that observed during late stages of hepatic ontogenesis. The steady-state mRNA levels of four (HNF1 alpha, HNF3 alpha, HNF4, and C/EBP beta) "liver-enriched" transcriptional factors gradually decrease during the late period of embryonic liver development while three factors (HNF1 beta, HNF3 beta, and DBP) increase. In the normal adult rat liver the expression of all the transcription factors are restricted to the hepatocytes. However, during early stages of oval cell proliferation both small and large bile ducts start to express HNF1 alpha and HNF1 beta, HNF3 gamma, C/EBP, and DBP but not HNF4. At the later stages all of these factors are also highly expressed in the proliferating oval cells. Expression of HNF4 is first observed when the oval cells differentiate morphologically and functionally into hepatocytes and form basophilic foci. At that time the expression of some of the other factors is also further increased. Based on these data we suggest that the upregulation of the "establishment" factors (HNF1 and -3) may be an important step in oval cell activation. The high levels of these factors in the oval cells and embryonic hepatoblasts further substantiates the similarity between the two cell compartments. Furthermore, the data suggest that HNF4 may be responsible for the final commitment of a small portion of the oval cells to differentiate into hepatocytes which form the basophilic foci and eventually regenerate the liver parenchyma.

Expression of the multidrug-resistant gene in hepatocarcinogenesis and regenerating rat liver.

Preneoplastic and neoplastic liver nodules and hepatocytes isolated from regenerating rat liver have been shown to be resistant to a broad range of carcinogenic agents. This phenomenon was studied by measuring the expression of the multidrug-resistant (mdr) gene in normal liver cells and in preneoplastic and neoplastic nodules and regenerating liver. Levels of messenger RNA for the mdr gene, which encodes P-glycoprotein, were elevated in both preneoplastic and neoplastic lesions. Expression of the mdr gene also reached high levels in regenerating rat liver 24 to 72 hours after partial hepatectomy. These results show that the expression of the mdr gene can be regulated in liver and is likely to be responsible for part of the multidrug-resistance phenotype of carcinogen-initiated hepatocytes and regenerating liver cells.

Comparative and integrative functional genomics of HCC.

Global gene expression profiling of hepatocellular carcinoma (HCC) is a promising new technology that has already refined the diagnosis and prognostic predictions of HCC patients. This has been accomplished by identifying genes whose expression pattern is associated with clinicopathological features of HCC tumors. Molecular characterization of HCC from gene expression profiling studies will undoubtedly improve the prediction of treatment responses, selection of treatments for specific molecular subtypes of HCC and ultimately the clinical outcome of HCC patients. The research focus is now shifting toward the identification of genetic determinants that are components of the specific regulatory pathways altered in cancers, and that may constitute novel therapeutic targets. Here we review the recent advances in gene expression profiling of HCC and discuss the future strategies for analysing large and complicated data sets from microarray studies and how to integrate these with diverse genomic data.

Cellular distribution of transforming growth factor-beta 1 and procollagen types I, III, and IV transcripts in carbon tetrachloride-induced rat liver fibrosis.

The cellular distribution and temporal expression of transcripts from transforming growth factor-beta 1 (TGF-beta 1) and procollagen alpha 1(I), alpha 1(III), and alpha 1(IV) genes were studied in carbon tetrachloride (CCl4)-induced rat liver fibrosis by using in situ hybridization technique. During the fibrotic process, TGF-beta 1 and procollagen genes were similarly and predominantly expressed in Desmin-positive perisinusoidal cells (e.g., fat-storing cells and myofibroblasts) and fibroblasts and their expression continued to be higher than those observed in control rats. These transcripts were also observed in inflammatory cells mainly granulocytes and macrophage-like cells at the early stages of liver fibrosis. The production of extracellular matrix along small blood vessels and fibrous septa coincided with the expression of these genes. Expression of TGF-beta 1 and procollagen genes were not detected in hepatocytes throughout the experiment. No significant differences in cellular distribution or time course of gene expression among procollagen alpha 1(I), alpha 1(III), and alpha 1(IV) were observed. Desmin-positive perisinusoidal cells and fibroblasts appeared to play the principal role in synthesis of collagens in CCl4-induced hepatic fibrosis. The simultaneous expression of TGF-beta 1 and procollagen genes in mesenchymal cells, including Desmin-positive perisinusoidal cells, during hepatic fibrosis suggests the possibility that TGF-beta 1 may have an important role in the production of fibrosis.

Effect of ethanol on polyamine synthesis during liver regeneration in rats.

Ethanol consumption retards the hepatic regenerative response to injury. This may contribute to the pathogenesis of liver injury in alcoholic individuals. The mechanisms responsible for ethanol-associated inhibition of liver regeneration are poorly understood. To determine if the antiregenerative effects of ethanol involve modulation of polyamine metabolism, parameters of polyamine synthesis were compared before and during surgically induced liver regeneration in ethanol-fed rats and isocalorically maintained controls. After partial hepatectomy, induction of the activity of ornithine decarboxylase (ODC), the rate limiting enzyme for polyamine synthesis, was delayed in rats that had been fed ethanol. This was correlated with reduced levels of putrescine, ODC's immediate product. Increases in hepatic spermidine and spermine were also inhibited. Differences in ODC activity between ethanol-fed and control rats could not be explained by differences in the expression of ODC mRNA or by differences in ODC apoenzyme concentrations, suggesting that chronic ethanol intake inactivates ODC posttranslationally. Supplemental putrescine, administered at partial hepatectomy and 4 and 8 h thereafter, increased hepatic putrescine concentrations and markedly improved DNA synthesis and liver regeneration in ethanol-fed rats. These data suggest that altered polyamine metabolism may contribute to the inhibition of liver regeneration that occurs after chronic exposure to ethanol.

Role of the IkappaB kinase complex in oncogenic Ras- and Raf-mediated transformation of rat liver epithelial cells.

NF-kappaB/Rel factors have been implicated in the regulation of liver cell death during development, after partial hepatectomy, and in hepatocytes in culture. Rat liver epithelial cells (RLEs) display many biochemical and ultrastructural characteristics of oval cells, which are multipotent cells that can differentiate into mature hepatocytes. While untransformed RLEs undergo growth arrest and apoptosis in response to transforming growth factor beta1 (TGF-beta1) treatment, oncogenic Ras- or Raf-transformed RLEs are insensitive to TGF-beta1-mediated growth arrest. Here we have tested the hypothesis that Ras- or Raf-transformed RLEs have altered NF-kappaB regulation, leading to this resistance to TGF-beta1. We show that classical NF-kappaB is aberrantly activated in Ras- or Raf-transformed RLEs, due to increased phosphorylation and degradation of IkappaB-alpha protein. Inhibition of NF-kappaB activity with a dominant negative form of IkappaB-alpha restored TGF-beta1-mediated cell killing of transformed RLEs. IKK activity mediates this hyperphosphorylation of IkappaB-alpha protein. As judged by kinase assays and transfection of dominant negative IKK-1 and IKK-2 expression vectors, NF-kappaB activation by Ras appeared to be mediated by both IKK-1 and IKK-2, while Raf-induced NF-kappaB activation was mediated by IKK-2. NF-kappaB activation in the Ras-transformed cells was mediated by both the Raf and phosphatidylinositol 3-kinase pathways, while in the Raf-transformed cells, NF-kappaB induction was mediated by the mitogen-activated protein kinase cascade. Last, inhibition of either IKK-1 or IKK-2 reduced focus-forming activity in Ras-transformed RLEs. Overall, these studies elucidate a mechanism that contributes to the process of transformation of liver cells by oncogene Ras and Raf through the IkappaB kinase complex leading to constitutive activation of NF-kappaB.

Coinduction of MDR-1 multidrug-resistance and cytochrome P-450 genes in rat liver by xenobiotics.

The levels of mRNA for multidrug-resistance (MDR-1) and selective cytochrome P-450 genes were determined in adult rat liver following administration of various natural and synthetic xenobiotics. MDR-1 (also known as PGY1) was induced following administration of aflatoxin B1, 2-(acetylamino)fluorene (AAF), N-hydroxy-2-(acetylamino)fluorene, isosafrole, phenothiazine, and 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), but not after phenobarbital or 7-hydroxy-2-(acetylamino)fluorene treatment. Cytochrome P-450 isoform d was induced by TCDD, isosafrole, phenothiazine, and AAF, while cytochrome P-450 isoform b was induced by phenobarbital, and to a lesser extent by isosafrole. These observations suggest that both MDR and cytochrome P-450 gene families are evolutionarily selected by the capacity of various xenobiotics to induce their own detoxification either through metabolism to hydrophilic derivatives by the cytochrome P-450 system or direct excretion from the cell by the MDR gene family. Furthermore, the data indicate that induction of selective members of the MDR and the cytochrome P-450 gene families may depend on overlapping regulatory elements.

Induction of monooxygenases in rhesus monkeys by 3-methylcholanthrene: metabolism and mutagenic activation of N-2-acetylaminofluorene and benzo(a)pyrene.

Induction of content and activity of hepatic and lung cytochrome P450-dependent monooxygenases by 3-methylcholanthrene (MCA) was studied in male rhesus monkeys. Hepatic microsomes were concomitantly studied for mutagenic activation of N-2-acetylaminofluorene (AAF) as well as with the use of sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Hepatic N-hydroxylation of AAF was increased sixfold to sevenfold after MCA treatment; aryl hydrocarbon hydroxylase was similarly increased in both liver and lung. The hepatic cytochrome P450 content was increased by 50%, and the Soret maximum in the CO-hemoprotein complex was shifted to the blue by 2 nm after MCA induction. Electrophoresis of hepatic microsomes from MCA-treated monkeys showed an increase in a polypeptide band of 54,000 molecular weight. Mutagenic activation of AAF by liver microsomes from untreated rhesus monkeys was low but was increased 40-fold after MCA treatment. No effect was observed on the mutagenic activation of N-hydroxy-2-acetylaminofluorene by liver microsomes after MCA treatment. Rhesus monkeys may provide a model for evaluation of the role of polycyclic hydrocarbon induction in chemically caused cancer and toxicity in primates.

Developmental pattern of 3-methylcholanthrene-inducible mutagenic activation of N-2-fluorenylacetamide, 2-fluorenamine, and 2,4-diaminoanisole in the rabbit.

The effects of 3-methylcholanthrene (MCA) on mutagenic activation of the carcinogenic arylamines N-2-fluorenylacetamide (FAA), 2-fluorenamine (FA), and 2,4-diaminoanisole (2,4-DAA) by liver homogenates were studied postnatally in Dutch rabbits. These effects were compared with the developmental profiles of cytochrome P448 and aryl hydrocarbon hydroxylase (AHH) activity. Mutagenic activation of FA and 2,4-DAA was increased by MCA as early as 2 days after birth, whereas induction of FAA mutagenicity appeared 6 days after birth. Thereafter, induction of all three arylamines closely paralleled induction of cytochrome P448, which was maintained into adulthood. In contrast, induction of AHH activity by MCA was highest at 2 days of age and decreased to control levels 20 days after birth.

Mutagenic activation of N-2-fluorenylacetamide and N-hydroxy-N-2-fluorenylacetamide in subcellular fractions from X/Gf mice.

The Salmonella mutagenesis test system was used to evaluate the in vitro mutagenic potency of N-2-fluorenylacetamide (2-FAA) and N-hydroxy-N-2-fluorenylacetamide (N-OH-2-FAA) mediated by liver and kidney subcellular fractions from X/Gf mice, a strain resistant to 2-FAA carcinogenesis. Pretreatment of the mice with the microsomal inducers 3-methylcholanthrene (MCA) and 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) increased the number of revertants from both liver and kidney fractions. Mutagenicity of N-OH-2-FAA mediated by liver or kidney microsomes was partially inhibited at 0.001--0.1 microM Paraxon (diethyl-p-nitrophenyl phosphate), an inhibitor of deacetylase enzyme, and the inhibition was complete (98%) in microsomes from control mice (100 microM Paraoxon). Conversely, the liver and kidney microsomal fractions from MCA- and TCDD-treated X/Gf mice were less sensitive to Paraoxon. The inhibition of kidney or liver cytosol-mediated N-OH-2-FAA mutagenicity by Paraoxon was less than that observed with the microsomal fraction (50% inhibition at 1 x 10(-7) and 1 x 10(-5) M Paraoxon, respectively). The mutagenicity of 2-FAA and N-OH-2-FAA mediated by liver or kidney subcellular fractions from X/Gf mice and its response to inducers and inhibitors of mutagenic activation processes appear similar to those observed in species both resistant (cotton rat) and sensitive (Sprague-Dawley rat, NIH Swiss mice) to 2-FAA carcinogenesis.

Histochemical changes in livers from portacaval-shunted rats.

Portacaval shunt (PCS) operations were performed on male inbred SD rats. The activity of liver gamma-glutamyltranspeptidase (GGT) increased a few days after the operation and remained high after several months. However, the activity was only present in periportal areas of liver lobules and was mainly restricted to the endothelial lining cells of periportal blood vessels. Phenobarbital sodium (CAS: 57-03-7) administration did not change the distribution of GGT. The activity of liver glucose-6-phosphatase disappeared in the centrilobular areas a few days after the operation but was present in the periportal areas. The distribution of this activity returned to normal after several months. The capacity of liver cells to store glycogen was significantly decreased following the PCS operation. Morphologic changes in hepatocytes observed after the PCS operation did not show similarities to those seen in preneoplastic livers. The nuclei of the small compact hepatocytes found in the animals with PCS were irregular in shape and stained intensively with hematoxylin and eosin. The large pale-staining cells in the periphery of liver lobes were GGT negative and their nuclei had a normal morphology. Neither morphologic nor histochemical changes consistent with preneoplastic and/or neoplastic stage were observed after 10 months in the liver when the PCS operation was performed on rats having received an ip dose (10 mg/kg) of diethylnitrosamine (CAS: 55-18-5) 2 weeks prior to the operation.

Analysis of c-myc expression in a human hepatoma cell line.

A tumorigenic human hepatoma cell line, Hep G2, has been shown to have high steady-state levels of c-myc transcripts compared to normal human liver. We have now characterized c-myc expression in Hep G2 cells with regard to message stability, gene rearrangements, gene amplification, chromosomal translocations, promoter utilization, and the effects of protein synthesis inhibitors. We have determined that the half-life of the Hep G2 c-myc transcript is approximately 20 min and conclude that the high steady-state level of c-myc mRNA is not the result of a specific stabilization of the c-myc message but probably results from increased c-myc gene transcription. c-myc expression in Hep G2 cells appears to be constitutive, since it remains constant in different cell growth states (log phase versus nondividing cells). The high constitutive expression of the c-myc gene in Hep G2 cells could not be explained by gene amplification, gene rearrangements, or chromosomal translocations. However, based on an S1 nuclease protection assay, the P1/P2 promoter utilization ratio is approximately 1/1 which differs from the 1/5 P1/P2 ratio observed in normal human liver. Treatment with cycloheximide, a protein synthesis inhibitor, does not superinduce Hep G2 c-myc transcription based on transcription "run on" and RNA slot blot analysis. However, cycloheximide treatment does exert a posttranscriptional effect involving the specific stabilization of the c-myc message.

Effect of inhibitors of ornithine decarboxylase on retrovirus induced transformation of murine erythroid precursors in vitro.

alpha-Difluoromethylornithine (DFMO) and methyl acetylene putrescine (MAP) are inhibitors of the rate limiting enzyme in polyamine synthesis, ornithine decarboxylase. We studied the effects of these compounds on the formation of retrovirus transformed erythroid colonies. DFMO was able to effectively reduce the number of transformed colonies at a concentration of 10(-3) M, whereas MAP achieved total inhibition at 10(-4) M. Putrescine, the product of ornithine decarboxylase, did not alter colony number by itself but it was able to overcome the inhibitory effects of both DFMO and MAP. Addition of DFMO at times after the initiation of culture decreased its effectiveness in reducing transformed colony numbers, while the converse was true for the erythroid stimulant, erythropoietin. We concluded from these data that DFMO and MAP probably diminished colony formation by inhibiting proliferation of the target cells for the retroviruses.

Stable expression of human cytochrome P450IIE1 in mammalian cells: metabolic activation of nitrosodimethylamine and formation of adducts with cellular DNA.

To introduce cytochrome P450IIE1 DNA stably into the chromosomal DNA of mammalian cells, we constructed recombinant retroviruses containing the full-length complementary DNA for human cytochrome P450IIE1 and a selectable neo gene. Rat and mouse cells were infected with these viruses, and clones expressing the neo marker gene product were selected in G418-containing medium. Analysis of the DNA of the clones by Southern blotting showed that the viral DNA was integrated into the cellular DNA. Enzymatic analysis of the clones showed that the transduced DNA directed the expression of enzymatically active cytochrome P450IIE1. Treatment of the cells with the carcinogen [14C]-nitrosodimethylamine and analysis of the cellular DNA by CsCl equilibrium density gradients showed incorporation of the label into DNA, indicating the formation of covalent adducts with the cell DNA. Construction of recombinant cell lines constitutively expressing cytochrome P450IIE1 provides a permanent source for this enzyme and can aid in the analysis of its catalytic properties, such as the metabolic activation of chemical mutagens/carcinogens, and the consequent cytotoxic and genotoxic effects of these compounds.

Disruption of the pRb/E2F pathway and inhibition of apoptosis are major oncogenic events in liver constitutively expressing c-myc and transforming growth factor alpha.

The oncogene c-myc and transforming growth factor (TGF) alpha are frequently coexpressed in human cancers, suggesting that their interaction may be a critical step in malignant growth. Consistent with this idea, we recently demonstrated in a transgenic mouse model that TGF-alpha dramatically enhances c-myc-induced hepatocarcinogenesis. To elucidate this synergistic effect, we have now investigated regulation of cell cycle and apoptosis during neoplastic development in the liver of c-myc and c-myc/TGFalpha transgenic mice. Both lines displayed dramatic increases of mitotic and apoptotic rates before the onset of hepatocellular carcinoma (HCC), but only c-myc/TGF-alpha livers showed significant levels of net proliferation (mitosis minus apoptosis). Subsequently, mitosis declined in peritumorous tissues, concomitant with the previously reported induction of TGF-beta1, whereas c-myc and c-myc/TGFalpha HCCs maintained mitotic hyperactivity. The c-myc/TGF-alpha HCCs were also characterized by a particularly strong expression of TGF-alpha and very low apoptotic index in contrast to high levels of apoptosis in peritumorous tissues and c-myc HCCs. The differential levels of cell proliferation in noncancerous and cancerous tissues correlated with a stronger induction of cyclin D1 mRNA and protein in c-myc/TGF-alpha and c-myc HCCs associated with intense pRb hyperphosphorylation. Severe deregulation of G1-S transition was also indicated by the dramatic up-regulation, particularly in the HCCs, of pRb-free E2F1-DP1 and E2F2-DP1 transcription factor heterodimers, as assessed by immunoprecipitation and immunohistochemistry. The existence of increased E2F activity during hepatocarcinogenesis was further indicated by the transcriptional induction of putative E2F target genes involved in cell cycle progression, such as endogenous c-myc, cyclin A, Cdc2, and E2F itself. Cdc2 overexpression and the elevated mitotic indices in the HCCs correlated also with induction of cyclin B steady-state levels. The data suggest that coexpression of c-myc and TGF-alpha leads to a selective growth advantage for hepatic (pre)neoplastic cells by disrupting the pRb/E2F pathway and by TGF-alpha-mediated reduction of apoptosis.

Ploidy and karyotypic alterations associated with early events in the development of hepatocarcinogenesis in transgenic mice harboring c-myc and transforming growth factor alpha transgenes.

The cooperation of the c-myc oncogene with the growth factor transforming growth factor (TGF)-alpha in development of liver tumors in transgenic mice has been demonstrated previously. In this study, we analyzed the ploidy and karyotype of c-myc, TGF-alpha, parental control, and the double transgenic c-myc/TGF-alpha hepatocytes at 3 weeks of age when the liver is histologically normal and at 10 weeks when the c-myc/TGF-alpha liver is dysplastic and contains basophilic foci. Eighty % of the 10-week hepatocytes were aneuploid, and 32% had chromosomal breakage. Statistically significant breakage was observed in six different chromosomes. Breakage at band A5 and at the border of bands C4/5 of chromosome 1 was observed. Fragile sets on chromosome 4 were most frequent in the middle of the chromosome at bands C2 and C6. Chromosome 6 was fragile at band F2. The region of chromosome 7 at bands B5 and D3 was frequently broken and involved in translocations. Chromosome 12 was broken at bands D1 and D3. The breakage sites on chromosomes 1, 4, 7, and 12 correspond to sites of tumor susceptibility genes in the mouse. Although there was no consistent change in copy number, recurrent translocations between chromosomes 1, 4, 7, 12 and 19 were also observed. These studies demonstrate that the development of dysplasia and basophilic foci in the liver is correlated with aneuploidy and chromosome breakage. The specific fragile sites indicate genetic regions that are altered during early stages of hepatocarcinogenesis. Due to the conservation of genetic linkage groups between mice and humans, the identification of genetic alterations in the mouse during hepatocarcinogenesis may provide critical information about tumor susceptibility genes that are important in the early development of human hepatocellular carcinoma.