Genetics of hepatocellular tumors - Nature

Oncogene (2006) 25, 3778–3786

& 2006 Nature Publishing Group All rights reserved 0950-9232/06 $30.00 www.nature.com/onc

REVIEW

Genetics of hepatocellular tumors P Laurent-Puig1,2 and J Zucman-Rossi3,4 1

Inserm, U775, Bases Molećulaires de la re´ponse aux xeńobiotiques, Paris, France; 2Universite´ Paris 5, Paris, France; 3Inserm, U674, Geńomique fonctionnelle des tumeurs solides, Paris, France and 4Universite´ Paris 7 Denis Diderot, Institut Universitaire d’He´matologie, CEPH, Paris, France

Numerous genetic alterations are accumulated during the process of hepatocarcinogenesis. These genetic alterations can be divided into two groups. The first set of genetic alterations is specific of hepatocellular tumor risk factors. It includes integration of hepatitis B virus (HBV) DNA, R249S TP53 (tumor protein p53) mutation in aflatoxin B1-exposed patients, KRAS mutations related to vinyl chloride exposure, hepatocyte nuclear factor 1a (HNF1a) mutations associated to hepatocellular adenomas and adenomatosis polyposis coli (APC) germline mutations predisposing to hepatoblastomas. The second set of genetic alterations are etiological nonspecific, it includes recurrent gains and losses of chromosomes, alteration of TP53 gene, activation of WNT/b-catenin pathway through CTNNB1/b-catenin and AXIN (axis inhibition protein) mutations, inactivation of retinoblastoma and IGF2R (insulin-like growth factor 2 receptor) pathways through inactivation of RB1 (retinoblastoma 1), P16 and IGF2R. Comprehensive analyses of these genetic alterations have defined two pathways of hepatocarcinogenesis according to the presence or the absence of chromosomal instability. Hepatitis B virus and poorly differentiated tumors are related to chromosome instable tumors associated with frequent TP53 mutations, whereas nonHBV and well-differentiated tumors are related to chromosomal stable samples that are frequently b-catenin activated. These classifications have clinical relevance as genetic alterations may also be related to prognosis. Oncogene (2006) 25, 3778–3786. doi:10.1038/sj.onc.1209547 Keywords: hepatocellular carcinoma; hepatocellular adenoma; chromosome instability; gene mutation; tumor suppressor gene

Introduction Cancer is a DNA disease owing to the accumulation of genetic alterations and more specifically of genes that control cell cycle and cell proliferation. Some of the observed genetic alterations are widely shared among Correspondence: Dr J Zucman-Rossi, Inserm, U674, Geńomique fonctionnelle des tumeurs solides, 27 rue Juliette Dodu, Paris 75010, France. E-mail: [email protected]

the different tumor types; however, the association of the recurrent genetic alterations is highly specific of a tumor type. Therefore, the comprehensive knowledge of a broad field of genetic alterations in a tumor type and the study of the correlation between these alterations and the different clinical and histological parameters allow to refine the tumor classification and the understanding of the multistep carcinogenesis process. Dysplastic cirrhotic nodules are considered to be precursors of hepatocellular carcinoma (HCC) because of their frequent association with the HCC occurrence (Furuya et al., 1988; Terada et al., 1993). Overall, HCC development is closely associated with cirrhosis and more than 90% of the tumors are found in a chronic hepatitis or a cirrhotic background (Edmondson and Peters, 1983). The various risk factors that are associated with the development of these lesions are well known. They mainly include infection with the hepatitis B virus (HBV) or hepatitis C virus (HCV), heavy alcohol intake, prolonged dietary exposure to aflatoxin B1 (AFB1) or vinyl chloride and primary hemochromatosis. In 90% of the HCC cases, at least one of these risk factors can be identified, either alone or in combination (detailed in Donato et al., in this issue) and the presence of each risk factor among patients varies according to the geographical origin of the patients (detailed in Seef et al., in this issue). However, in less than 10% of the cases, HCC are observed in noncirrhotic liver and even without inflammatory lesions. These HCC developed in an otherwise normal liver are usually found in patients without well-established risk factors. Some of these cases may correspond to the malignant transformation of liver adenoma that are rare benign hepatocellular tumors found in young women taking oral contraception for more than 3 years (Edmondson et al., 1976). Hepatoblastoma is another particular case of primary hepatocellular tumor developed in otherwise normal liver parenchyma. Hepatoblastoma is a rare embryonic tumors (1:1 000 000 births), but it represents the most frequent liver tumors in children under the age of 3 years (Weinberg and Finegold, 1983). These tumors histologically differ from HCC as they are composed of immature epithelial cells with sometimes immature mesenchymal or teratomatous elements (Weinberg and Finegold, 1983). The HCC as well as precursor benign lesions have been extensively studied in terms of genetic alteration in the past 10 years and our knowledge has dramatically

Genetics of hepatocellular tumors P Laurent-Puig and J Zucman-Rossi

3779

increased in this field leading to the definition of the different altered pathways in hepatocarcinogenesis. As in other solid tumors, a large number of genetic alterations are accumulated during the carcinogenetic process. Indeed, genetic and epigenetic alterations have been observed in cirrhotic nodules and half of them have been found to have a monoclonal origin by examining the X-chromosome methylation pattern (Piao et al., 1997; Paradis et al., 1998; Yeh et al., 2001). Chromosome aberrations with loss of alleles are found in half of cirrhotic nodules and more frequently in nodules with small cell dysplasia (Yeh et al., 2001). However, structural alterations of specific gene, that is, activating mutations of oncogene and inactivating mutations of tumor suppressor genes have not been described in cirrhosis and were only found in HCC and liver cell adenomas. Overall, a very large number of genetic alterations have been described in primary liver tumors, and in this review, we will describe the different somatic genetic alterations acquired by the hepatocytes to be fully transformed in tumor cells. These genetic alterations can be divided in specific or nonspecific aberrations to the HCC etiological factors. Relationship between gene alterations and clinical features such as prognostic and genetic susceptibility will be analysed.

Genetic alterations specific to etiological factors Genetic alterations and hepatitis B virus infection Hepatitis B virus infection can promote carcinogenesis by at least three different mechanisms. First, integration of the viral DNA in the host genome can induce chromosome instability (Aoki et al., 1996). Second, insertional mutations have been described in which HBV integration at specific sites activates endogenous genes such as retinoic acid b-receptor (Dejean et al., 1986), cyclin A (Wang et al., 1990) and mevalonate kinase (Graef et al., 1994). More recently, 15 new genes were found to be altered by an HBV integration in tumors, suggesting that viral integration in the vicinity of genes controlling cell proliferation, viability and differentiation is a mechanism frequently involved in HBV hepatocarcinogenesis (Ferber et al., 2003; Horikawa and Barrett, 2003; Paterlini-Brechot et al., 2003). All of these integration events are unique, except for the human telomerase reverse transcriptase (hTERT) and the inositol 1,4,5-triphosphate receptor (IPR) genes that have been observed to have site-specific HBV integrations in independent tumors (Ferber et al., 2003; Horikawa and Barrett, 2003; Paterlini-Brechot et al., 2003). Insertional mutagenesis seems to be implicated in 20–40% of the HCC cases related to HBV infection. The third mechanism of carcinogenesis linked to HBV infection is based on the expression of viral protein, in particular HBX, to modulate cell proliferation and viability (Andrisani and Barnabas, 1999; Diao et al., 2001). Moreover, HBX binds to p53 and inactivates p53-dependent activities, including p53-mediated apoptosis (Feitelson et al., 1993).

In endemic HBV area, familial aggregations of HCC were observed. Although this clustering of familial HBV is mainly owing to perinatal transmission of HBV infection, some evidences suggest that genetic susceptibility may also be involved (Yu et al., 2000). Furthermore, segregation analysis suggested that this familial aggregation could be owing to a gene with an autosomal mode of inheritance that may act with HBV to increase the risk of HCC (Shen et al., 1991; Cai et al., 2003). Exposure to aflatoxin B1 leads to specific TP53 mutation in hepatocellular carcinoma Aflatoxin B1 requires metabolic conversion to its exo8,9-epoxide in order to damage DNA and proteins and various adducts are formed in vitro and in vivo (Essigmann et al., 1977; Martin and Garner, 1977; Miller, 1978; Groopman et al., 1981). After an exposure to AFB1, mainly in sub-tropical regions, an accumulation of DNA adducts are found in the liver, as a result of conversion of the AFB1 to its active metabolites. Aflatoxin B1 is a very potent mutagen and the AFB1 epoxide reacts with guanine in DNA, leading to genetic changes. The most frequently induced mutation is the GC-TA transversion. The mutational spectrum of TP53 (tumor protein p53) gene in HCC from Qidong and Mozambique, where AFB1 exposure level is high, revealed G-T transversion at codon 249 in more than 50% of the cases (Bressac et al., 1991). This mutation at codon 249 of TP53, leading to the amino-acid substitution R249S, is exceptionally found in HCC from geographical regions without AFB1 exposure. In general, the frequency of the R249S mutation paralleled the estimated level of AFB1 exposure, supporting the hypothesis that the carcinogen has a causative role in hepatocarcinogenesis. Molecular mechanisms of AFB1–DNA binding and mutagenesis have been elucidated in human tumors, animal models and in vitro (see for a review Smela et al., 2001). There is evidence that the carcinogenic potential of AFB1 may vary between individuals, as for the same exposition not all patients will develop HCC. The variation is likely owing to the individual capacity to detoxify the aflatoxin 8,9 epoxide, the mutagenic metabolite of AFB1. A part of this inter-individual variability is owing to different efficacy of the enzymes to metabolize this carcinogene compound. Among the enzymes implicated in the detoxification process, the glutathion transferase GSTM1 and GSTT1 and the epoxide hydrolase (EPHX) have been shown to play an important role. In the human population, these enzymes are polymorphic and some of these polymorphisms have functional consequences. For GSTM1 and GSTT1, the main polymorphism is owing to a gene deletion, which leads to a null allele. Patients are grouped according to the presence or the absence of non-null allele. For the EPHX, the main functional polymorphism is owing to the substitution of a histidine in position 113 by a tyrosine decreasing the EPHX activity of 40%. For GSTM1 and H113T EPHX polymorphisms, there was a statistically significant relationship with detectable levels Oncogene


3780

of AFB1–albumin adducts in the serum. In a case– control study, it has been shown that patients with one EXPH 113 Tyr allele have an increased risk of developing HCC; furthermore, a synergistic effect exists with HBV infection leading to a risk ratio of 77 of developing HCC when patients bearing the at-risk EPHX allele (i.e. one 113 Tyr allele) are infected by HBV. The risk was 15 for HBV patients with no EPHX risk allele and three for patients with the at-risk allele but not infected by HBV (McGlynn et al., 1995). A more recent publication showed that either GSTM1 and GSTT1 homozygous null allele patients are at risk of developing HCC when exposed to AFB1 (Sun et al., 2001). In addition, the effect of aflatoxin exposure on HCC risk was more pronounced among chronic hepatitis B surface antigen carriers with the GSTT1 null genotypes (odds ratio 3.3, confidence interval 95% 1.5–9.3). These results demonstrate the importance of gene–environment interaction in the development of HCC. KRAS mutations are associated with vinyl chloride exposure in hepatocellular carcinoma Vinyl chloride is a carcinogen associated with the development of liver angiosarcomas and rarely with HCC. Recently, the presence of KRAS2 mutations was observed in 33% of 18 vinyl chloride-associated HCCs and three mutations were found in adjacent nonneoplastic liver tissue (Weihrauch et al., 2001). The KRAS mutations observed were probably caused by chloroethylene oxide, a carcinogenic metabolite of vinyl chloride. Because KRAS mutations are rarely observed in HCCs that are not associated with vinyl chloride exposure, these results strongly suggest that KRAS2 mutations play an important role in the carcinogenetic pathway linked to vinyl chloride exposure. Specific genetic alterations associated with liver cell adenomas Recently, biallelic mutation of the transcription factor 1 (TCF1) gene, coding for hepatocyte nuclear factor 1a (HNF1a), were identified in 60% of a sample of liver cell adenoma cases (Bluteau et al., 2002b) (Figure 1). Hepatocyte nuclear factor 1a is a transcription factor that is implicated in hepatocyte differentiation and is required for the liver-specific expression of several genes, including b-fibrinogen, albumin and a1-antitrypsin (Frain et al., 1989; Baumhueter et al., 1990; Cereghini et al., 1990; Chouard et al., 1990). Heterozygous

germline mutations of TCF1/HNF1a have been linked to the occurrence of MODY3 (maturity onset diabetes of the young, OMIM no. 600496) in humans (Yamagata et al., 1996). MODY3 is a rare, dominantly inherited subtype of non-insulin-dependent diabetes mellitus characterized by early onset, usually before the age of 25, and a primary defect in insulin secretion. In liver cell adenomas, inactivation of both TCF1/HNF1a alleles is normally observed; in 90% of the cases both mutations are somatic. However in the other cases, corresponding to MODY3 patients, one mutation is germ line and the second allele inactivation is a somatic event observed only in the tumor (Bluteau et al., 2002b; Bacq et al., 2003). Furthermore, two well-differentiated HCC occurring in normal liver contained somatic biallelic TCF1/HNF1a mutations in a sample of 30 HCCs screened. These results indicate that inactivation of HNF1a, whether sporadic or associated with MODY3, is an important genetic event in the occurrence of human liver adenoma and may be an early step in the development of some HCCs (Bluteau et al., 2002b). In a recent study including 96 adenomas, we showed that HNF1a-mutated adenoma represented the most common type of adenoma which is phenotypically characterized by marked steatosis (Zucman-Rossi et al., 2006). Genetic predisposition to hepatoblastoma Most of the hepatoblastoma are sporadic, but they can also occur in the Beckwith–Wiedemann syndrome (BWS) or in the familial adenomatous polyposis (FAP), two cancers predispositions. The last predisposition is owing to adenomatosis polyposis coli (APC) germline mutation. Immunohistochemical analysis of b-catenin has demonstrated nuclear/cytoplasmic accumulation of the protein in most hepatoblastomas and sequence analysis of the b-catenin N-terminal domain has revealed interstitial deletions or missense mutations in the GSK3b phosphorylation motif in 48–67% of sporadic hepatoblastoma tumors (Koch et al., 1999; Wei et al., 2000; Buendia, 2002). In some hepatoblastoma cases, activation of the WNT/b-catenin pathway may be related to AXIN2 (axis inhibition protein 2) mutations (Koch et al., 2004). However, in a series of five FAP patients with hepatoblastoma, Cetta et al. (2003) noted that neither b-catenin and allelic losses were found at APC locus, suggesting that in patients with germline APC mutation, activation of the Wnt pathway occurs in the absence of b-catenin mutations in the exon 3.

Genetic alterations not related to etiological factors

Figure 1 Transcription factor 1 (TCF1) mutations in liver cell adenoma tumorigenesis (Bluteau et al., 2002b). Oncogene

Gains and losses of chromosomes Tumor hepatocytes in HCC accumulate a large number of chromosome rearrangements leading to highly complex karyotypes. It supports the fact that multistage hepatic carcinogenesis results from accumulation of multiple genetic alterations, like in other solid tumors. Hyperploid DNA content was found by flow cytometry


3781

in tumor cells suggesting a global gain of genetic material, in half of the analysed HCC cases (Ezaki et al., 1988; Fujimoto et al., 1991; Chiu et al., 1992). Hyperploidy is also found in 43% of dysplastic lesions observed in cirrhotic disease (Thomas et al., 1992). This frequency increases in high-grade dysplastic lesions, suggesting that chromosome losses followed by endomitosis are early steps in hepatocarcinogenesis. The sequencing of the human genome has allowed the development of methods to detect quantitative chromosomal aberrations in tumors. These methods are based on the comparison of tumor and non-tumor DNA using microsatellite allelotypes to detect loss of heterozygosity (LOH), or comparative genomic hybridization (CGH) to detect gains and losses of chromosome regions. These two methods were used to analyse large series of human HCC and to obtain a clear consensus regarding recurrent chromosome aberrations. In summary, the most frequently deleted chromosome arms are 17p, 8p, 16q, 16p, 4q, 9p, 13q, lp and 6q and the most frequent gains concerned chromosomes 1q, 7q, 8q and 17q (Fujimoto et al., 1994; Boige et al., 1997; Marchio et al., 1997; Nagai et al., 1997; Piao et al., 1998; Guan et al., 2000; Wong et al., 2000; Balsara et al., 2001; Laurent-Puig et al., 2001; Nishimura et al., 2002). As chromosome loss is a frequent mechanism of inactivation of one allele of a tumor suppressor gene in solid tumors, recurrent chromosome losses at precise locus may point to the presence of a tumor suppressor gene. In HCC, tumor suppressor genes, targeted by LOH, have been identified on chromosome 17p, 13q, 16p, 9p and 6q corresponding to inactivation of TP53, RB1 (retinoblastoma 1), AXIN1 (axis inhibition protein 1), CDKN2A (cyclin-dependent kinase inhibitor 2A, also named p16) and IGF2R (insulin-like growth factor 2 receptor), respectively. In contrast, on chromosome 1p, 4q, 8p and 16q, no targeted tumor suppressor genes have yet been identified and high-resolution allelotyping have been performed to define boundaries and physical map of the deleted chromosome regions (Koyama et al., 1999; Piao et al., 1999; Pineau et al., 1999; Balsara et al., 2001; Yakicier et al., 2001; Bluteau et al., 2002a). Alterations of the TP53 gene To date, the most frequent gene alteration found in human cancer changes the coding properties of the tumor suppressor gene TP53 (Hollstein et al., 1991), a gene located in band p13.1 of chromosome 17 (Isobe et al., 1986; Miller et al., 1986). Germline mutations of this gene are also known to occur in some individuals with a cancer predisposition known as the Li–Fraumeni syndrome (Malkin et al., 1990). P53 behaves as a multifunctional transcription factor involved in the control of the cell cycle, programmed cell death, senescence, differentiation and development, transcription, DNA replication, DNA repair and maintenance of genomic stability. Moreover, analysis of TP53 mutations is an excellent model to study interactions between etiological factors, DNA damage and carcinogenetic pathways. In HCC, the specific TP53 mutation R249S is

found in about 50% of tumors in populations exposed to AFB1 (Bressac et al., 1991; Hsu et al., 1991). In contrast, patients who have not been exposed to this carcinogen have a lower prevalence of TP53 gene mutations (10–30%) and codon 249 is rarely altered. A second hotspot of mutations has been suggested in HCC occurring in patients with hereditary hemochromatosis, at codon 220 (Vautier et al., 1999), but this result is debated in other series (Laurent-Puig et al., 2001) . Genetic alterations activating the WNT/b-catenin pathway b-Catenin is the vertebrate homolog of armadillo in Drosophila melanogaster. It has a dual function in adhesion and Wnt signaling. Indeed, b-catenin, complexes with E-cadherin and catenins in adherent junctions and participates to cell–cell adhesion. b-catenin can acquire oncogenic activity as the result of dominant mutations within its N-terminus (Morin et al., 1997). Many of these mutations lead to the loss of phosphorylation consensus sites implicated in the negative regulation of b-catenin by the GSK3b/APC/axin complex. As a consequence of these mutations, b-catenin level is high in the cytoplasm and in nuclei activating the transcription of Wnt target genes (Figure 2). Such CTNNB1/b-catenin-activating mutations have been found in HCC, both in human and transgenic mouse models (de La Coste et al., 1998; Miyoshi et al., 1998). In human HCC, b-catenin mutations were found in 13–43% of the cases, depending on the series (de La Coste et al., 1998; Huang et al., 1999; Kondo et al., 1999; Legoix et al., 1999; Nhieu et al., 1999; Terris et al., 1999; Hsu et al., 2000; Laurent-Puig et al., 2001). It is also mutated in the majority of hepatoblastomas (Koch et al., 1999; Wei et al., 2000) and in some hepatocellular adenomas (Chen et al., 2002). We recently showed that 15% of 96 adenomas were mutated for b-catenin and these mutated tumors have a higher risk of malignant transformation in HCC (Zucman-Rossi et al., 2006). The AXIN1 gene is located on the short arm of chromosome 16, in 16p13, a chromosome region deleted in about 30% of HCC cases (Laurent-Puig et al., 2001). It codes for a protein of the cytoplamic GSK3b complex that inhibits the Wnt pathway (Figure 2). In HCC, biallelic inactivations of AXIN1 through mutations associated with LOH or homozygous deletion are observed in 5–8% of the HCCs (Satoh et al., 2000; Laurent-Puig et al., 2001). In these tumors, inactivation of Axin1 prevents phosphorylation of b-catenin by GSK3b leading to the accumulation of b-catenin and activation of Wnt target genes. AXIN2 mutations were also found in HCC; however, in contrast with colon cancer, analysis of the APC gene did not show mutation (Legoix et al., 1999; Taniguchi et al., 2002; Ishizaki et al., 2004). Alteration of the retinoblastoma and p16INK4 genes Loss of chromosome 13 is observed in about 30% of HCC (Boige et al., 1997; Nagai et al., 1997; Oncogene


3782

Figure 2 Wnt signaling pathway. Activation of the Wnt pathway (right part) results in b-catenin accumulation and activation of Wnt target genes. In the absence of Wnt signaling (left part), b-catenin is phosphorylated by GSK3b leading to ubiquitinylation and degradation. In hepatocellular carcinoma (HCC), Wnt signaling pathway can be activated by b-catenin-stabilizing mutations or Axin1-inactivating mutations.

Laurent-Puig et al., 2001). The chromosome region commonly deleted in tumors includes the RB1 locus located at 13q14 encoding for the retinoblastoma tumor suppressor gene (RB1) that functions at the core of developmental decisions in cell division, differentiation and apoptosis in almost all cell types. In a series of HCC, LOH at 13q14 correlates with the absence of RB1 expression suggesting biallelic inactivation of the RB1 gene (Zhang et al., 1994). However, point mutations of the RB1 gene have been found in only a few HCC tumors, suggesting that the second allele could be inactivated by epigenetic mechanisms (Lin et al., 1996). The p16INK4 gene codes two different proteins, cyclin D-dependent kinase inhibitor 2 (CDKN2) and ARF implicated in p53-dependent apoptosis. These two proteins function as tumor suppressors in the retinoblastoma pathway (reviewed by Hickman et al., 2002). Loss of heterozygosity at the short arm of chromosome 9, in the region of p16INK4, was found in about 20% of HCC cases (Boige et al., 1997; Nagai et al., 1997; Laurent-Puig et al., 2001). Methylation of the p16INK4 promoter leading to the absence of protein expression was found in 30–70% of the tumors (Liew et al., 1999; Matsuda et al., 1999; Jin et al., 2000; Weihrauch et al., 2001). Somatic inactivating mutations and homozygous deletions of the gene are rarely observed in HCC (Biden et al., 1997; Jin et al., 2000). Finally, Gankyrin was found to be overexpressed in all cases of 34 analysed HCC (Higashitsuji et al., 2000). Moreover, the authors found that Gankyrin induced anchorage-independent growth and tumorigenicity in NIH/3T3 cells. It binds to RB1, increasing RB1 phosphorylation and releasing the activity of the transcription factor E2F-1. Gankyrin Oncogene

accelerated the degradation of RB1 in vivo and in vitro by the proteasome complex. Altogether, frequent alterations of RB1, P16INK4 and Gankyrin emphasize the important role of alteration of the retinoblastoma pathway in hepatocarcinogenesis. Other genetic alterations in hepatocellular carcinoma Different molecular alterations are less frequently described in HCC than those previously mentioned and alteration of the corresponding signaling pathways remained to be evaluated in large series of tumors. In the transforming growth factor b (TGFb) pathway, inactivations of the mannose 6-phosphate/insulin-like growth factor 2 receptor (M6P/IGF2R) have been found in HCC. This gene maps to chomosome 6q25–27 region. It codes for a multifunctional receptor involved in intracellular lysosmal enzyme trafficking, TGF activation and IGF2 degradation. Granzyme B internalization by M6P/IGF2R is also required for cytotoxic T cells to induce apoptosis in cells targeted for death, leading to be referred as a ‘death receptor’ (Motyka et al., 2000). The M6P/IGF2R also plays a crucial role in regulating mammalian fetal growth and development (Filson et al., 1993). Somatic point mutations leading to amino-acid substitution in IGF2R or protein truncation were identified in 10–20% of the screened HCC in two series (De Souza et al., 1995; Oka et al., 2002). From another study (Yakicier et al., 1999), rare mutations leading to amino-acid substitutions were identified in MADH2/ Smad2 and MADH4/Smad4 genes (also known as DPC4 or deleted in pancreatic carcinoma 4). They mediate signaling by the TGF-b/activin/BMP-2/4


3783

cytokine superfamily from receptor serine/threonine protein kinases at the cell surface to the nucleus (Lagna et al., 1996; Liu et al., 1996). Overall, these results suggest alteration of the TGFb pathway in 10–30% of the HCC cases. Recently, mutations activating phosphatidylinositol 3-kinase (PIK3CA gene) have been found in 36% of HCC in Korea (Lee et al., 2005) and these mutations activating the AKT pathway have been found in less than 5% of HCC in France (unpublished results). Suppressor of cytokine signaling (SOCS-1 also known as JAB and SSI-1) inhibits cytokine signaling by its direct interaction with Janus kinase (JAK). SOCS-1 gene is located at the short arm of chromosome 16. Recently, Yoshikawa et al. (2001) identified aberrant methylation in the SOCS-1 promoter leading to transcription silencing of the gene in 65% of 26 tested HCC. Moreover, the authors found that restoration of SOCS-1 suppressed both growth rate and anchorageindependent growth of cells in which SOCS-1 was methylation-silenced and JAK constitutively activated. These results suggest that activation of the JAK/STAT pathway may frequently participate in hepatocarcinogenesis.

Global study of the molecular alterations in hepatocellular carcinoma Genomewide analysis of genetic alterations in hepatocellular carcinoma Genetic alterations accumulated in HCC are numerous with more than 20 different genes involved altering at least four different signaling pathways. To understand relations existing between these different genetic alterations, we recently analysed a large number of genetic alterations in the same series of tumors (Laurent-Puig et al., 2001) (Figure 3). A series of 137 HCC was collected in France in which the different etiological factors were equally represented with alcohol abuse found in 37% of the cases, HBV and HCV infection each in one-third of the cases. Tumors were analysed

Figure 3 Major hepatocarcinogenesis pathways defined by genetic alterations and their relations with clinical parameters. Lines joining boxes indicate significant correlation (Laurent-Puig et al., 2001).

using high-density allelotyping to search for LOH on the entire genome and gene mutations were sought in TP53, AXIN1 and CTNNB1/b-catenin genes. Loss of heterozygosity was the most frequent genetic alteration observed because only seven tumors showed no loss of any chromosome arms. According to the other studies, most frequently deleted chromosome arms were 8p (48%), 17p (45%), 4q (38%), 1p (33%), 13q (30%), 6q (29%), 16p (24%) and 9p (20%). Activation of the Wnt pathway was found in 28% of the case either by activation CTNNB1/b-catenin mutations (19% of the cases) or by AXIN1 inactivation (9% of the cases). Alteration of the TP53 pathway was found at the same frequency with a TP53 gene mutation identified in 26% of the cases. To test associations between genetic alterations, correlations were sought by analysing TP53, AXIN1 and CTNNB1/b-catenin mutations together with the 10 most frequently deleted chromosomes and the FAL index measuring the number of chromosomes deleted per tumor. Half of the tumors showed LOH on more than five chromosome arms (FAL index>0.12), reflecting a chromosomal instability as described by Cahill et al. (1998) in colon cancer. Genetic alterations associated with chromosome instability were LOH at chromosome 1p, 4q, 13q and 16, and mutation of the axin1 and TP53 genes. In the other half of the HCC, tumors were chromosome stable, LOH at the short arm of chromosome 8 and b-catenin mutations were most frequently observed. Thus, this study enabled the definition of two mechanisms of hepatocarcinogenesis. In the first, genetic alterations are accumulated through a chromosome instability, and in the second, activation of the Wnt pathway by CTNNB1/ b-catenin mutation was predominant (Laurent-Puig et al., 2001). Correlations between etiological factors and molecular alterations Associations between etiological factors of HCC and genetic alterations in tumors have been found in several studies. In our series analysing 137 French HCC, chromosome instability together with TP53 and AXIN1 mutations were closely related to HBV infection (Laurent-Puig et al., 2001). This relationship could explain the variation of FAL index values observed in different published series. In contrast, the second hepatocarcinogenesis pathway defined by CTNNB1/bcatenin mutation associated with chromosome 8p deletion in a context of chromosome stability is significantly associated with the absence of HBV infection. This last group of tumors includes HCV and alcohol-related tumors without genetic alteration specifically associated with the alcohol abuse. Furthermore, in this group of chromosome stable tumors, 35% of HCC cases did not show recurrent gene mutation or chromosome LOH. Similar results were also found in other analyses using allelotyping or CGH to search for chromosome abnormalities (Marchio et al., 2000; Okabe et al., 2000; Wong et al., 2000; Collonge-Rame et al., 2001; Bluteau et al., 2002a). Oncogene


3784

Correlations between other clinical parameters and molecular alterations Globally, the number of genetic alterations in HCC tumors is inversely proportional to its degree of differentiation (Nishida et al., 1992). Positive relations were observed between a poor differentiation of the tumor and chromosome losses at 4q (Okabe et al., 2000), 16q (Tsuda et al., 1990; Nishida et al., 1992; Kuroki et al., 1995; Okabe et al., 2000), 17p (Kuroki et al., 1995; Laurent-Puig et al., 2001), 8p (Kuroki et al., 1995; Okabe et al., 2000) and 13q (Nishida et al., 1992; Kuroki et al., 1995; Okabe et al., 2000; Laurent-Puig et al., 2001). Similarly, TP53 mutations are most frequently observed in poorly differentiated tumors (Tanaka et al., 1993; Teramoto et al., 1994; LaurentPuig et al., 2001). These results suggest that chromosome LOH at 4q, 16q, 17p and 13q, and TP53 mutations may be late genetic alterations accumulated at the end of hepatocarcinogenesis progression. Apart from TCF1/HNF1a mutations found in two cases of HCC developed in otherwise normal liver, no specific molecular alteration have been found in this group of tumors (Bluteau et al., 2002a). In our series analysing 137 French tumors, we found that tumors developed in non-cirrhotic liver had significantly most frequent chromosome 8p and 13q LOH (Laurent-Puig et al., 2001). However, this result has to be confirmed in a second study. Finally, b-catenin-activating mutations were found most frequently in large sized (Laurent-Puig et al., 2001; Wong et al., 2001) and well-differentiated tumors (Mao et al., 2001). Although the pathological tumor-node-metastasis staging has been applied clinically to HCC patients, this system is inadequate for predicting recurrence in individuals who undergo

hepatic resection. In our analysis of 137 resected HCC, we found that chromosome instability and more specifically chromosome 9p and 6q losses were associated with poor prognosis (Laurent-Puig et al., 2001). In other studies, different new variables have been suggested to have a prognostic value (Okabe et al., 2001; Shirota et al., 2001; Xu et al., 2001). In the future, additional studies using larger prospective series of patients should clarify the prognostic value of these new molecular markers.

Conclusion Characterization of the genetic alterations associated with HCC tumors is an essential step to increase our knowledge of hepatocarcinogenesis. Systematic search for these alterations in series of tumors, including all grade, stages, etiologies and also pre-neoplastic lesions, will enable to find a chronological order in the accumulation of the genetic alterations during tumor progression. Systematic transcriptome and proteome analyses may also contribute to identify new carcinogenetic pathways altered in these tumors. Finally, comprehensive studies may find clinical application to identify markers useful to early detection of tumors, to predict prognosis or to find new therapeutic targets. Acknowledgements This work was supported by the Inserm and the Fondation de France.

References Andrisani OM, Barnabas S. (1999). Int J Oncol 15: 373–379. Aoki H, Kajino K, Arakawa Y, Hino O. (1996). Proc Natl Acad Sci USA 93: 7300–7304. Bacq T, Jacquemin E, Balabaud C, Jeannot E, Scotto B, Branchereau S et al. (2003). Gastroenterology 125: 1470–1475. Balsara BR, Pei J, De Rienzo A, Simon D, Tosolini A, Lu YY et al. (2001). Gene Chromosome Cancer 30: 245–253. Baumhueter S, Mendel DB, Conley PB, Kuo CJ, Turk C, Graves MK et al. (1990). Genes Dev 4: 372–379. Biden K, Young J, Buttenshaw R, Searle J, Cooksley G, Xu DB et al. (1997). Hepatology 25: 593–597. Bluteau O, Beaudoin J-C, Pasturaud P, Belghiti J, Franco D, Bioulac-Sage P et al. (2002a). Oncogene 21: 1225–1232. Bluteau O, Jeannot E, Bioulac-Sage P, Marques JM, Blanc JF, Bui H et al. (2002b). Nat Genet 32: 312–315. Boige V, Laurent-Puig P, Fouchet P, Flejou JF, Monges G, Bedossa P et al. (1997). Cancer Res 57: 1986–1990. Bressac B, Kew M, Wands J, Ozturk M. (1991). Nature 350: 429–431. Buendia MA. (2002). Med Pediatr Oncol 39: 530–535. Cahill DP, Lengauer C, Yu J, Riggins GJ, Willson JK, Markowitz SD et al. (1998). Nature 392: 300–303. Cai RL, Meng W, Lu HY, Lin WY, Jiang F, Shen FM. (2003). World J Gastroenterol 9: 2428–2432. Oncogene

Cereghini S, Yaniv M, Cortese R. (1990). EMBO J 9: 2257–2263. Cetta F, Mazzarella L, Bon G, Zuckermann M, Casorelli A, Nounga H. (2003). Med Pediatr Oncol 41: 496–497. Chen YW, Jeng YM, Yeh SH, Chen PJ. (2002). Hepatology 36: 927–935. Chiu JH, Kao HL, Wu LH, Chang HM, Lui WY. (1992). J Clin Invest 89: 539–545. Chouard T, Blumenfeld M, Bach I, Vandekerckhove J, Cereghini S, Yaniv M. (1990). Nucleic Acids Res 18: 5853–5863. Collonge-Rame MA, Bresson-Hadni S, Koch S, Carbillet JP, Blagosklonova O, Mantion G et al. (2001). Cancer Genet Cytogenet 127: 49–52. de La Coste A, Romagnolo B, Billuart P, Renard CA, Buendia MA, Soubrane O et al. (1998). Proc Natl Acad Sci USA 95: 8847–8851. De Souza AT, Hankins GR, Washington MK, Orton TC, Jirtle RL. (1995). Nat Genet 11: 447–449. Dejean A, Bougueleret L, Grzeschik KH, Tiollais P. (1986). Nature 322: 70–72. Diao J, Khine AA, Sarangi F, Hsu E, Iorio C, Tibbles LA et al. (2001). J Biol Chem 276: 8328–8340. Edmondson HA, Henderson B, Benton B. (1976). N Engl J Med 294: 470–472.


3785 Edmondson HA, Peters RL. (1983). Semin Roentgenol 18: 75–83. Essigmann JM, Croy RG, Nadzan AM, Busby Jr WF, Reinhold VN, Buchi G et al. (1977). Proc Natl Acad Sci USA 74: 1870–1874. Ezaki T, Kanematsu T, Okamura T, Sonoda T, Sugimachi K. (1988). Cancer 61: 106–109. Feitelson MA, Zhu M, Duan LX, London WT. (1993). Oncogene 8: 1109–1117. Ferber MJ, Montoya DP, Yu C, Aderca I, McGee A, Thorland EC et al. (2003). Oncogene 22: 3813–3820. Filson AJ, Louvi A, Efstratiadis A, Robertson EJ. (1993). Development 118: 731–736. Frain M, Swart G, Monaci P, Nicosia A, Stampfli S, Frank R et al. (1989). Cell 59: 145–157. Fujimoto J, Okamoto E, Yamanaka N, Toyosaka A, Mitsunobu M. (1991). Cancer 67: 939–944. Fujimoto Y, Hampton LL, Wirth PJ, Wang NJ, Xie JP, Thorgeirsson SS. (1994). Cancer Res 54: 281–285. Furuya K, Nakamura M, Yamamoto Y, Togei K, Otsuka H. (1988). Cancer 61: 99–105. Graef E, Caselmann WH, Wells J, Koshy R. (1994). Oncogene 9: 81–87. Groopman JD, Croy RG, Wogan GN. (1981). Proc Natl Acad Sci USA 78: 5445–5449. Guan XY, Fang Y, Sham JS, Kwong DL, Zhang Y, Liang Q et al. (2000). Gene Chromosome Cancer 29: 110–116. Hickman ES, Moroni MC, Helin K. (2002). Curr Opin Genet Dev 12: 60–66. Higashitsuji H, Itoh K, Nagao T, Dawson S, Nonoguchi K, Kido T et al. (2000). Nat Med 6: 96–99. Hollstein M, Sidransky D, Vogelstein B, Harris CC. (1991). Science 253: 49–53. Horikawa I, Barrett JC. (2003). Carcinogenesis 24: 1167–1176. Hsu HC, Jeng YM, Mao TL, Chu JS, Lai PL, Peng SY. (2000). Am J Pathol 157: 763–770. Hsu IC, Metcalf RA, Sun T, Welsh JA, Wang NJ, Harris CC. (1991). Nature 350: 427–428. Huang H, Fujii H, Sankila A, Mahler-Araujo BM, Matsuda M, Cathomas G et al. (1999). Am J Pathol 155: 1795–1801. Ishizaki Y, Ikeda S, Fujimori M, Shimizu Y, Kurihara T, Itamoto T et al. (2004). Int J Oncol 24: 1077–1083. Isobe M, Emanuel BS, Givol D, Oren M, Croce CM. (1986). Nature 320: 84–85. Jin M, Piao Z, Kim NG, Park C, Shin EC, Park JH et al. (2000). Cancer 89: 60–68. Koch A, Denkhaus D, Albrecht S, Leuschner I, von Schweinitz D, Pietsch T. (1999). Cancer Res 59: 269–273. Koch A, Weber N, Waha A, Hartmann W, Denkhaus D, Behrens J et al. (2004). J Pathol 204: 546–554. Kondo Y, Kanai Y, Sakamoto M, Genda T, Mizokami M, Ueda R et al. (1999). Jpn J Cancer Res 90: 1301–1309. Koyama M, Nagai H, Bando K, Ito M, Moriyama Y, Emi M. (1999). Jpn J Cancer Res 90: 951–956. Kuroki T, Fujiwara Y, Tsuchiya E, Nakamori S, Imaoka S, Kanematsu T et al. (1995). Gene Chromosome Cancer 13: 163–167. Lagna G, Hata A, Hemmati-Brivanlou A, Massague J. (1996). Nature 383: 832–836. Laurent-Puig P, Legoix P, Bluteau O, Belghiti J, Franco D, Binot F et al. (2001). Gastroenterology 120: 1763–1773. Lee JW, Soung YH, Kim SY, Lee HW, Park WS, Nam SW et al. (2005). Oncogene 24: 1477–1480. Legoix P, Bluteau O, Bayer J, Perret C, Balabaud C, Belghiti J et al. (1999). Oncogene 18: 4044–4046.

Liew CT, Li HM, Lo KW, Leow CK, Chan JY, Hin LY et al. (1999). Oncogene 18: 789–795. Lin Y, Shi CY, Li B, Soo BH, Mohammed-Ali S, Wee A et al. (1996). Ann Acad Med Singapore 25: 22–30. Liu F, Hata A, Baker JC, Doody J, Carcamo J, Harland RM et al. (1996). Nature 381: 620–623. Malkin D, Li FP, Strong LC, Fraumeni Jr JF, Nelson CE, Kim DH et al. (1990). Science 250: 1233–1238. Mao TL, Chu JS, Jeng YM, Lai PL, Hsu HC. (2001). J Pathol 193: 95–101. Marchio A, Meddeb M, Pineau P, Danglot G, Tiollais P, Bernheim A et al. (1997). Gene Chromosome Cancer 18: 59–65. Marchio A, Pineau P, Meddeb M, Terris B, Tiollais P, Bernheim A et al. (2000). Oncogene 19: 3733–3738. Martin CN, Garner RC. (1977). Nature 267: 863–865. Matsuda Y, Ichida T, Matsuzawa J, Sugimura K, Asakura H. (1999). Gastroenterology 116: 394–400. McGlynn KA, Rosvold EA, Lustbader ED, Hu Y, Clapper ML, Zhou T et al. (1995). Proc Natl Acad Sci USA 92: 2384–2387. Miller C, Mohandas T, Wolf D, Prokocimer M, Rotter V, Koeffler HP. (1986). Nature 319: 783–784. Miller EC. (1978). Cancer Res 38: 1479–1496. Miyoshi Y, Iwao K, Nagasawa Y, Aihara T, Sasaki Y, Imaoka S et al. (1998). Cancer Res 58: 2524–2527. Morin PJ, Sparks AB, Korinek V, Barker N, Clevers H, Vogelstein B et al. (1997). Science 275: 1787–1790. Motyka B, Korbutt G, Pinkoski MJ, Heibein JA, Caputo A, Hobman M et al. (2000). Cell 103: 491–500. Nagai H, Pineau P, Tiollais P, Buendia MA, Dejean A. (1997). Oncogene 14: 2927–2933. Nhieu JT, Renard CA, Wei Y, Cherqui D, Zafrani ES, Buendia MA. (1999). Am J Pathol 155: 703–710. Nishida N, Fukuda Y, Kokuryu H, Sadamoto T, Isowa G, Honda K et al. (1992). Int J Cancer 51: 862–868. Nishimura T, Nishida N, Itoh T, Kuno M, Minata M, Komeda T et al. (2002). Gene Chromosome Cancer 35: 329–339. Oka Y, Waterland RA, Killian JK, Nolan CM, Jang HS, Tohara K et al. (2002). Hepatology 35: 1153–1163. Okabe H, Ikai I, Matsuo K, Satoh S, Momoi H, Kamikawa T et al. (2000). Hepatology 31: 1073–1079. Okabe H, Satoh S, Kato T, Kitahara O, Yanagawa R, Yamaoka Y et al. (2001). Cancer Res 61: 2129–2137. Paradis V, Laurendeau I, Vidaud M, Bedossa P. (1998). Hepatology 28: 953–958. Paterlini-Brechot P, Saigo K, Murakami Y, Chami M, Gozuacik D, Mugnier C et al. (2003). Oncogene 22: 3911–3916. Piao Z, Kim NG, Kim H, Park C. (1999). Cancer Lett 138: 227–232. Piao Z, Park C, Park JH, Kim H. (1998). Int J Cancer 75: 29–33. Piao Z, Park YN, Kim H, Park C. (1997). Liver 17: 251–256. Pineau P, Nagai H, Prigent S, Wei Y, Gyapay G, Weissenbach J et al. (1999). Oncogene 18: 3127–3134. Satoh S, Daigo Y, Furukawa Y, Kato T, Miwa N, Nishiwaki T et al. (2000). Nat Genet 24: 245–250. Shen FM, Lee MK, Gong HM, Cai XQ, King MC. (1991). Am J Hum Genet 49: 88–93. Shirota Y, Kaneko S, Honda M, Kawai HF, Kobayashi K. (2001). Hepatology 33: 832–840. Smela ME, Currier SS, Bailey EA, Essigmann JM. (2001). Carcinogenesis 22: 535–545. Oncogene


3786 Sun CA, Wang LY, Chen CJ, Lu SN, You SL, Wang LW et al. (2001). Carcinogenesis 22: 1289–1294. Tanaka S, Toh Y, Adachi E, Matsumata T, Mori R, Sugimachi K. (1993). Cancer Res 53: 2884–2887. Taniguchi K, Roberts LR, Aderca IN, Dong X, Qian C, Murphy LM et al. (2002). Oncogene 21: 4863–4871. Terada T, Ueda K, Nakanuma Y. (1993). Virchows Arch A 422: 381–388. Teramoto T, Satonaka K, Kitazawa S, Fujimori T, Hayashi K, Maeda S. (1994). Cancer Res 54: 231–235. Terris B, Pineau P, Bregeaud L, Valla D, Belghiti J, Tiollais P et al. (1999). Oncogene 18: 6583–6588. Thomas RM, Berman JJ, Yetter RA, Moore GW, Hutchins GM. (1992). Hum Pathol 23: 496–503. Tsuda H, Zhang WD, Shimosato Y, Yokota J, Terada M, Sugimura T et al. (1990). Proc Natl Acad Sci USA 87: 6791–6794. Vautier G, Bomford AB, Portmann BC, Metivier E, Williams R, Ryder SD. (1999). Gastroenterology 117: 154–160. Wang J, Chenivesse X, Henglein B, Brechot C. (1990). Nature 343: 555–557. Wei Y, Fabre M, Branchereau S, Gauthier F, Perilongo G, Buendia MA. (2000). Oncogene 19: 498–504. Weihrauch M, Benicke M, Lehnert G, Wittekind C, Wrbitzky R, Tannapfel A. (2001). Br J Cancer 84: 982–989.

Oncogene

Weinberg AG, Finegold MJ. (1983). Hum Pathol 14: 512–537. Wong CM, Fan ST, Ng IO. (2001). Cancer 92: 136–145. Wong N, Lai P, Pang E, Leung TW, Lau JW, Johnson PJ. (2000). Hepatology 32: 1060–1068. Xu XR, Huang J, Xu ZG, Qian BZ, Zhu ZD, Yan Q et al. (2001). Proc Natl Acad Sci USA 98: 15089–15094. Yakicier MC, Irmak MB, Romano A, Kew M, Ozturk M. (1999). Oncogene 18: 4879–4883. Yakicier Y, Legoix P, Vaury C, Gressin L, Tubacher E, Capron F et al. (2001). Oncogene 20: 5232–5238. Yamagata K, Oda N, Kaisaki PJ, Menzel S, Furuta H, Vaxillaire M et al. (1996). Nature 384: 455–458. Yeh SH, Chen PJ, Shau WY, Chen YW, Lee PH, Chen JT et al. (2001). Gastroenterology 121: 699–709. Yoshikawa H, Matsubara K, Qian GS, Jackson P, Groopman JD, Manning JE et al. (2001). Nat Genet 28: 29–35. Yu MW, Chang HC, Liaw YF, Lin SM, Lee SD, Liu CJ et al. (2000). J Natl Cancer Inst 92: 1159–1164. Zhang X, Xu HJ, Murakami Y, Sachse R, Yashima K, Hirohashi S et al. (1994). Cancer Res 54: 4177–4182. Zucman-Rossi J, Jeannot E, Tran Van Nhieu J, Scoazec J-Y, Guettier C, Rebouissou S et al. (2006). Hepatology 43: 515–524.