• Share
  • Email
  • Embed
  • Like
  • Save
  • Private Content







Total Views
Views on SlideShare
Embed Views



0 Embeds 0

No embeds



Upload Details

Uploaded via as Adobe PDF

Usage Rights

© All Rights Reserved

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
Post Comment
Edit your comment

    Nrgastro.2011.196 Nrgastro.2011.196 Document Transcript

    • REVIEWS Hepatocellular carcinoma: insight from animal models Yan Li, Zhao-You Tang and Jin-Xuan Hou Abstract | Hepatocellular carcinoma (HCC) ranks as the third most common cause of death from cancer worldwide. Although major risk factors for the development of HCC have been defined, many aspects of the evolution of hepatocellular carcinogenesis and metastasis are still unknown. Suitable animal models are, therefore, essential to promote our understanding of the molecular, cellular and pathophysiological mechanisms of HCC and for the development of new therapeutic strategies. This Review provides an overview of animal models that are relevant to HCC development, metastasis and treatment. For HCC development, this Review focuses on transgenic mouse models of HBV and HCV infection, which provide experimental evidence that viral genes could initiate or promote liver carcinogenesis. Animal models of HCC metastasis provide platforms to elucidate the mechanisms of HCC metastasis, to study the interaction between the microenvironment and HCC invasion and to conduct intervention studies. In addition, animal models have been developed to investigate the effects of new treatment modalities. The criteria for establishing ideal HCC animal models are also discussed. Li, Y. et al. Nat. Rev. Gastroenterol. Hepatol. 9, 32–43 (2012); published online 25 October 2011; doi:10.1038/nrgastro.2011.196 Introduction Department of Oncology, Zhongnan Hospital of Wuhan University, 169 Donghu Road, Wuhan 430071, China (Y. Li, J.‑X. Hou). Liver Cancer Institute and Zhongshan Hospital, Fudan University, Key Laboratory of Carcinogenesis and Cancer Invasion (Fudan University), Ministry of Education, China, 136 Yi Xue Yuan Road, Shanghai 200032, China (Z.‑Y. Tang). Hepatocellular carcinoma (HCC) is one of the most common types of cancer worldwide. Indeed, of the 748,300 new liver cancer cases and 695,900 liver-cancerrelated deaths estimated to have occurred in 2008, 70–85% are believed to have been caused by HCC.1 In addition, half of the total liver-cancer-related deaths and new cases are thought to have occurred in China, rendering HCC the top priority in China’s anticancer campaign.1 Liver carcinogenesis is a multistep process: the presence of specific risk factors promotes gene damage, which leads to a cascade of molecular and cellular deregulations that ultimately result in transformation of hepatocytes (Figure 1). In China and sub-Saharan Africa, the most important risk factors for HCC are HBV infection and exposure to environmental toxins including aflatoxin B1 and diethylnitrosamine.2,3 By contrast, HCV infection, alcohol-related liver cirrhosis and nonalcoholic fatty liver disease are the most important risk factors for HCC in developed countries and other low-risk areas. 2,4,5 Globally speaking, however, viral hepa­titis is the single most important cause of HCC, particularly the HBV and HCV subtypes, which are jointly responsible for up to 80% of HCC cases worldwide.6,7 Moreover, a population-based, long-term, prospective cohort study has provided convincing evidence that an elevated serum HBV DNA level (≥10,000 copies/ml) is a strong independent risk factor for HCC.8 Similarly, clinical and epidemio­logical data also suggest that HCV is an independent cause of HCC.9 Correspondence to: Z.‑Y. Tang zytang88@163.com Competing interests The authors declare no competing interests. 32  |  JANUARY 2012  |  VOLUME 9  Establishing successful animal models of HCC is, therefore, crucial for both basic and translational studies of HCC. A wide range of HCC animal models are currently available, which have provided researchers with the opportunity to assess tumor–host interactions, perform drug screening, mimic the complex multistep process of liver carcinogenesis, and conduct various therapeutic experiments. No model, however, is ideal for all purposes. Consequently, investigators should make a knowledgeable selection from the currently available models, or construct new models, on the basis of several key criteria (Box 1). This Review focuses on the insights obtained from animal models of liver carcinogenesis associated with viral hepatitis. An exhaustive evaluation of all HCC animal models in current use is beyond the scope of this article, although several interesting reviews on this topic have been published elsewhere.10–13 In this article, we focus on the animal models that are relevant to studies of HCC development, metastasis and treatment. Models of HCC development Transgenic mouse models of HBV or HCV infection have provided reliable experimental proof that viral genes could initiate or promote liver carcinogenesis. HBV transgenic mouse models Multiple transgenic mice models that express specific fragments of the HBV genome have been generated for the study of HBV-induced liver carcinogenesis. These transgenes are usually under the control of either the HBV promoter, or liver-specific host promoters, www.nature.com/nrgastro © 2011 Macmillan Publishers Limited. All rights reserved
    • REVIEWS including those for albumin or metallothionein. Most HBV transgenic mouse models focus on the HBx gene, which encodes HBV X protein (HBx)—a transcriptional transactivator that stimulates expression of a broad range of proto-oncogenes, including c‑fos, c‑myc and c‑jun.14,15 Activation of these proto-oncogenes shifts trans­forming growth factor β signaling from tumor-suppressive to oncogenic pathways 16 that stimulate hepatocyte prolifera­tion,17 inhibit apoptosis,18 upregulate protein degradation,19 and induce genetic instability and DNA repair mechanisms.19 Kim et al. 20 investigated the role of the HBx gene in the development of HCC using a transgenic CD1 mouse model (Table 1). In contrast to wild-type CD1 mice, which do not normally develop spontaneous liver tumors and have a lifespan of approximately 24 months, the majority of HBx transgenic mice died from clear cell HCC at 11–15 months of age. Interestingly, male mice died earlier than female mice owing to a faster rate of HCC progression. Multifocal areas of altered hepatocytes with high levels of HBx protein were found in the transgenic mice 4 months after birth, and by months 8–10 these altered hepatocytes had developed into adenomas that expressed high levels of HBx protein and α‑fetoprotein (AFP). Expression of these proteins remained high until the mice died. Two features differed in the findings from this animal model compared with clinical settings—cirrhosis and inflammation were absent in the preneoplastic stages, which signifies a direct contribution of the HBx protein to carcinogenesis.20 Proteomic analysis of liver tissue in the early stages of HCC (dysplasia and adenoma) from HBx transgenic mice21 (Table 1) identified 22 proteins with altered expression levels, the majority of which were involved in the crucial metabolic processes of glyco­ lysis and lipogenesis.22 These observations indicate that consider­able metabolic changes occur in the early stages of liver carcinogenesis.22 Another proteomics study 23 looked at fully developed HCC tissue from HBx transgenic mice24 (Table 1) and found persistent upregulation of the ubiquitin–proteasome and lysosomal pathways.23 These findings indicate the continual presence of cell injury, leading to protracted production of reactive oxygen species (ROS) and liver regeneration. The two most prominent theories of carcinogenesis are the one-hit theory, in which a carcinogen both initiates and promotes cancer formation, and the two-hit theory, in which the first causative factor initiates cell transformation and a second one promotes transformed cells to develop into cancer.25 The HBx transgenic mouse studies suggest that the expression of the HBx gene itself directly, and independently of other factors, causes HCC; these results, therefore, support the one-hit theory of carcinogenesis.26 Lakhtakia et al.27 constructed an HBx15–c-myc transgene model in C57BL/6xSJL mice to determine whether the presence of HBx plus an oncogene would result in substantially accelerated HCC progression (Table 1). The transgene comprised an HBx gene fragment that led to expression of a truncated HBx protein (X1558–154), which is sufficient to provide the crucial transactivating Key points ■■ Suitable animal models are necessary to provide information on the molecular, cellular and pathophysiological mechanisms of hepatocellular carcinoma (HCC) ■■ Transgenic mouse models have provided reliable experimental evidence suggesting that viral hepatitis genes could have a primary role in initiating or promoting liver carcinogenesis ■■ Nonviral factors, including oncogenes and environmental carcinogens, might only have a secondary role in liver carcinogenesis, but they could considerably accelerate the transformation of hepatocytes ■■ An animal model of metastatic human HCC that incorporates the effects of variation in metastatic potential would provide a unique tool for the study of HCC metastasis ■■ Animal models of HCC could be useful for developing and testing novel therapeutic modalities function, along with the mouse c‑myc oncogene. The transgenic mice exhibited a considerable increase in liver size and weight, which was most noticeable 5 months after birth. As noted in the HBx transgenic mouse study by Kim et al.,20 mice in this model also had no cirrhosis or inflammation in the preneoplastic stages.27 c‑myc expression was predominantly cytoplasmic, and more noticeable in neoplastic nodules than in the surrounding normal liver cells. This model, therefore, suggests that synergism occurs between the HBx gene and c‑myc in HCC development,27 and so supports the two-hit theory: after the first hit by HBx protein, the second hit by c‑myc considerably shortens the liver carcinogenesis process, by as much as 4 months. In contrast to the HBx transgenic mouse models, transgenic mice that express the viral S gene, which encodes HBV surface antigen (HBsAg), do develop distinctive inflammation and HCC, especially male mice.24,28 Chisari et al.28,29 developed an HBV transgenic mouse model, designated Tg (Alb‑1 HBV) Bri44, which undergoes a distinct form of liver carcinogenesis (Table 1). These mice have been engineered to contain HBV genomic sequences that encode HBx, HBsAg, and pre‑S proteins (the viral S gene, which encodes the viral capsid protein HBsAg, contains three in-frame start codons that allow three peptides of different lengths [pre‑S1+S, pre‑S2+S, and S respectively] to be transcribed from the same gene). These transgenic mice do not spontaneously develop tumors. At 1 month after birth, Tg (Alb‑1 HBV) Bri44 mice have normal liver histology. By month 2, however, discrete areas of hepatocellular necrosis and inflamma­ tion occur, and by month 3, progressive hepato­ yte c damage, Kupffer cell hyperplasia and mononuclear cell infiltration can be observed, which mimics the features of chronic hepatitis in humans.29–32 By months 4–6, these degenerative alterations30 are followed by cell­ lar u damage caused by an active inflammatory response and elevated compensatory hepatocyte proliferation,32 which constitutes a precancerous state. By months 7–9, adenomas develop, which in 33% of cases are of the clear cell type.31 By month 12, typical HCC with trabecular histology is present.28 Neoplastic lesions progressively grow to macro­scopic nodules that can be observed in all animals by months 16–18. NATURE REVIEWS | GASTROENTEROLOGY & HEPATOLOGY © 2011 Macmillan Publishers Limited. All rights reserved VOLUME 9  |  JANUARY 2012  |  33
    • REVIEWS HBV Hepatocytes HBV surface antigen filaments HBV protein Peroxisome proliferation Endoplasmic reticulum Transactivation No inflammation Glucose and fatty acid metabolism disturbance Lipid droplets Cell injury and inflammation Diethylnitrosamine aflatoxin Oxidative stress Steatosis Alcohol Reactive oxygen species Gene damage Multiple random mutations in hepatocytes Signaling pathway disturbance Adenoma HCC Steatosis Lipid droplets Electron transfer system disturbance Mitochondria Cytoplasm Nuclei Hepatocytes Core protein HCV Figure 1 | Model of liver carcinogenesis based on evidence obtained from transgenic mice. Viral factors exert a dominant role in HCC development through promotion of the central mechanism of increased oxidative stress and ROS generation. This mechanism leads to a vicious cycle of injury, DNA damage and liver regeneration, which render hepatocytes at an increased risk of transforming mutations. External factors, including exposure to chemicals and alcohol, have only a secondary role in the development of HCC. Red boxes indicate major factors or events. Blue boxes indicate secondary factors. Solid arrows indicate major processes, and dashed arrows indicate minor processes. Abbreviation: HCC, hepatocellular carcinoma. Overexpression of the large HBV envelope poly­ peptide in mice initiates a process of hepatocyte injury, inflamma­tion, and regenerative hyperplasia that renders large numbers of hepatocytes at risk of developing trans­forming mutations.28 Interestingly, HBV envelope protein was found in all hepatocytes from these animals by months 2–7, however, it was not expressed in all adenomas, nor in any HCCs. AFP levels also increased with the develop­ ent of adenomas and HCCs. This model, m therefore, mimics many of the pathological events that occur before the development of HCC in humans with chronic HBV infection, and provides a useful tool to study the processes of carcinogenesis and progression of 34  |  JANUARY 2012  |  VOLUME 9  HCC following HBV infection. Again, this model also supports the one-hit hypothesis, as HBV acts as a complete carcinogen that causes HCC by initiating a complex series of events in response to chronic hepato­cyte injury. To study the dysregulatory events that occur during early liver carcinogenesis, Barone et al.33 compared the gene expression profile of 3‑month-old Tg (Alb‑1 HBV) Bri44 transgenic mice with 3‑month-old wild-type animals. Microarray data on a total of 12,600 genes showed that the expression of 45 genes was significantly different in the transgenic mouse—25 genes were upregulated and 20 genes downregulated. The products of many of the upregulated genes have immunological www.nature.com/nrgastro © 2011 Macmillan Publishers Limited. All rights reserved
    • REVIEWS functions, suggesting that the accumulation of viral proteins results in hepatocyte damage and an immune response. 28,32,34 Moreover, several of the genes with altered expression were associated with apoptosis, such as upregulation of the gene encoding the antiapoptotic protein NuprI35 and downregulation of the gene encoding the proapoptotic protein Bnip3.36 This finding supports the conclusion that dysregulation of apoptosis, which facilitates the escape of ‘abnormal’ cells from death, could be a mechanism through which HBV promotes HCC development.27 In a similar fashion, Sell et al.37 constructed HBsAg transgenic C57BL/6 mice (Table 1). The 50‑4 strain of these mice had a high HBsAg content in hepatocytes, premalignant changes, nodules, adenomas and HCC; exposure to diethylnitrosamine or aflatoxin accelerated the development of HCC, and produced considerably more tumor nodules in the liver. This model demonstrates that the HBsAg protein itself could be both an initiator and a promoter of hepatocyte transformation, and that the addition of environmental carcinogens consider­ bly accelerates this process. a HCV transgenic mouse models Several HCV transgenic mouse model systems have been established that are based on the structure of the HCV genome (Table 2). Currently, transgenic mice expressing HCV structural proteins (core, E1, E2 and p7) or nonstructural proteins (NS2, NS3, NS4A, NS4B, NS5A and NS5B) individually or in various combinations are available.38 These proteins are designed to be constitutively expressed under the control of liver-specific promoters. Koike et al. established HCV envelope gene (E1 and E2)39 and core gene40,41 transgenic models, in which the expression of each gene is under the control of the same enhancer or promoter (Table 2). Only mice engineered to express HCV core gene developed HCC.40 As early as 3 months of age, HCV core gene transgenic mice developed hepatic steatosis—a histological feature of chronic HCV infection. In mice aged up to 12 months, steatosis slowly progressed without neoplastic changes. At month 16, one-quarter of the male transgenic mice of the C21 line had developed HCC, but no female mice of the C21 line had developed tumors. Moreover, the livers of transgenic mice aged over 12 months exhibited morpho­ ogical changes, including hepatic nodules l filled with eosinophilic cells containing fat droplets, suggesting an age-dependent increase in oxidative stress. In this study,40 therefore, transgenic mice at ages 6, 12 and 16 months represent the early, medium, and late stages of liver carcino­genesis, respectively. A proteomics study of HCV core gene transgenic mice42 found that in animals over 6 months of age, the expression of apoptosis-promoting proteins was suppressed. In animals over 12 months of age, expression of proteins related to cellular respiration, the electron-transfer system and antioxidative pathways was significantly upregulated. Finally, in animals aged over 16 months, expression of proteins related to defense, oxidation, and apoptosis was also significantly suppressed. This fluctuating expression of Box 1 | Criteria for the selection and design of animal models of HCC ■■ The model should faithfully and stably reproduce the key biological behaviors of HCC. For example, there should be clear and distinctive stages of hepatocyte degeneration, regeneration, proliferation and transformation in spontaneous models of HCC ■■ The model should help to reliably and reproducibly evaluate key molecular and cellular events in HCC development and progression ■■ The model should adequately reflect the full range of interactions between the tumor and the host, and between the primary tumor and metastases ■■ The model should mimic the human tumor microenvironment ■■ The model should be affordable and easy to manipulate Abbreviation: HCC, hepatocellular carcinoma. proteins could explain the stages of liver carcinogenesis. At the initial lesion stage (6 months), major changes are decreased apoptosis and increased β‑oxidation. At the precancer stage (12 months), the most prominent change is the shift from mitochondrial respiration to aerobic glyco­ ysis associated with malignant transformation. At l the final cancer stage (16 months), the most characteristic changes are markedly decreased biological functions including respiration, protein synthesis, defense and metabolism. Another HCV transgenic mouse model has provided convincing evidence to support the direct carcinogenic role of viral proteins in HCC (Table 2).43 This work established full-length HCV polyprotein (FL‑N) and HCV structural protein (S‑N) transgenic models. Adenomas and HCCs developed in these animals from month 13. Major histopathological features in the liver of both FL‑N and S‑N transgenic mice included the absence of inflammatory cell infiltrate, prominent microvesicular and macrovesicular centrilobular steatosis, adenoma and HCC of mixed histological types accompanied by hepatic fibrosis. Once again, this work confirmed the pre-eminent role of HCV structural proteins in HCC development, although nonstructural proteins could also contribute to liver carcinogenesis.43 Kamegaya et  al. 44 developed HCV transgenic mouse models with a common genetic background (FVB×C57BL/6) that expressed either core E1–E2 genes or the HCV core gene alone (Table 2). To accelerate HCC development in these mice, which have a low susceptibility to tumors, they also received intra­ eritoneal injecp tions of diethylnitrosamine once a week for 3 weeks. Interestingly, although similar numbers of liver tumors developed in all three groups of mice (wild-type diethylnitrosamine-treated controls and both transgenic models), the core E1–E2 transgenic mice developed significantly larger tumors (approximately 4 mm in dia­ meter) than wild-type or transgenic mice expressing the HCV core gene only (approximately 1 mm in diameter). Key histopathological features of livers from core E1–E2 mice included well-­ ifferentiated HCC, mild steatosis, d no inflammation and no fibrosis. Of particular note is the finding that E1 and/or E2 protein might accelerate liver carcinogenesis by suppression of apoptosis rather than by enhanced proliferation. NATURE REVIEWS | GASTROENTEROLOGY & HEPATOLOGY © 2011 Macmillan Publishers Limited. All rights reserved VOLUME 9  |  JANUARY 2012  |  35
    • REVIEWS Table 1 | Mouse models of HCC development resulting from HBV infection Mouse model Transgenes expressed Morphological signs of HCC development Reference CD1 HBx gene plus transcription enhancer Month 4: high expression of HBx protein Months 8–10: adenoma formation and high expression of HBx and AFP Months 11–15: death of male mice from clear cell HCC Kim et al. (1991)20 C57BL/6xDBA HBx gene, the HBV transcriptional enhancer and a portion of the pre C–C sequence Month 4: high expression of HBx protein Month 6: small neoplastic nodules form. High expression of HBx and PCNA Months 11–18: grossly identified HCC formation and high expression of HBx and PCNA Month 17: metastatic mesenteric mass formation. High expression of HBx and PCNA Yu et al. (1999)21 Transgenic C57BL/6 (p21HBsAg or p21-HBx) HBsAg or HBx genes Month 5: significantly elevated levels of serum glutamic-pyruvic transaminase in p21-HBsAg transgenic mice Up to month 12: half of the mice had steatosis without inflammatory or neoplastic changes Months 15–24: most male but no female p21-HBsAg mice developed HCC Months 18–24: both male and female p21-HBx mice developed HCC Wang et al. (2004)24 Transgenic C57BL/6xSJL HBx15–c-myc HBx gene encoding amino acids 58–154 plus c‑myc Week 1: increased mitosis, nuclear pleomorphism and multiple nuclei in hepatocytes Months 2–3: focal necrosis; proliferation of Kupffer cells Months 4–5: distinct adenomas; multifocal, well-differentiated HCC with a trabecular pattern. Significant increase in liver size and weight Weeks 24–32: microscopic HCC developed Weeks 28–32: gross HCC developed Lakhtakia et al. (2003)27 Transgenic (Alb‑1 HBV) Bri44 Pre‑S, HBsAg, HBx genes Months 2–7: high expression of HBV envelope protein Month 4: moderately severe chronic hepatitis occurrence Month 6: regenerative nodules and oval cell hyperplasia development Month 8: liver cell adenomas formation and high expression of AFP Months 12–20: HCC development and high expression of AFP Dunsford et al. (1990);28 Chisari et al. (1985);29 Toshkov et al. (1994);31 Huang et al. (1995)32 Female C57BL/6 linage 50-4 HBsAg gene Months 3–15: elevated serum AFP level Month 15: transgenic mice exposed to aflatoxin and diethylnitrosamine developed adneomas or carcinomas. No adenomas or carcinomas developed in transgenic mice not exposed to a carcinogen Sell et al. (1991)37 Abbreviations: AFP, α-fetoprotein; HBx, HBV X protein; HBsAg, HBV surface antigen; HCC, hepatocellular carcinoma; PCNA, proliferating cell nuclear antigen. Models of HCC metastasis Metastasis is a fundamental biological behavior of HCC and the main cause of treatment failure. HCC is prone to both intrahepatic and extrahepatic metastasis. In a clinical setting, the most common site of distant spread is the lung, owing to dissemination of tumor cells via the bloodstream, hemodynamic features of the liver and the intrinsic biological characteristics of the tumor, such as increased prolifera­ ion, invasion and motility 45 (Figure 2a). t Our research group has developed several mouse models of spontaneous HCC metastasis. Preliminary studies, in which tumor tissue from 30 patients with HCC were orthotopically implanted into athymic BALB/c mice, resulted in identification of a HCC xenograft model with high metastasic potential (LCI‑D20).46 Our subsequent studies using the LCI‑D20–BALB/c model system led to the development of MHCC97 cells, which develop lung metastases when inoculated orthotopically into BALB/c nude mice.47 From this cell line, two subclone variants with high (MHCC97H) and low (MHCC97L) metastatic potential were also established. MHCC97H cell clones were then subjected to three, six and nine rounds of in vivo selection of a high potential to metastasize, to produce three HCC cell lines (HCCLM3, HCCLM6 and HCCLM9) with a greater metastatic potential than MHCC97H (Figure 2b).48,49 As the MHCC97 cell line, MHCC97L cell clone, MHCC97H cell clone, HCCLM3, HCCLM6 and HCCLM9 cell lines are derived from the same original tumor tissue model 36  |  JANUARY 2012  |  VOLUME 9  (LCI‑D20), together they make a ‘spontaneous stepwise metastasis model system’. In addition, as all of the cell lines originated from one genetic background, comparisons of gene or protein expression profiles from tumor cells with different metastatic potentials could help to discover metastasis-related markers. For example, overexpression of cytokeratin 19 (CK19) is associated with high metastatic potential,50 and this protein has also been identified as a progenitor-cell marker in a rat HCC model.51 Of practical importance, the MHCC97–BALB/c model system could help to validate metastasis-related or recurrence-related tumor biomarkers. Several important signaling pathways and candidate markers have been identified at the gene49 and protein level.50 We found that several key molecules, including X‑linked inhibitor of apoptosis protein,52 calpain small subunit 1 (Capn4),53 programmed cell death 1 ligand 1 (PD-L1),54 CD24,55 CD15156 and β‑catenin57 were overexpressed in highly metastatic cell lines including HCCLM3 and HCCLM6, and these molecules could, therefore, be candidate biomarkers for future investigation in diagnostic studies and as targets for therapy. Tumor nodules consisting of CD90+ cells appeared in nude mice 3 months after sub­ cutaneous injection of 500 cells from either MHCC97L or MHCC97H cell lines, suggesting that CD90 is a potential marker of liver cancer stem cells.58 This model system could also provide a platform for drug screening. For example, the antitumor and antimetastatic effects of www.nature.com/nrgastro © 2011 Macmillan Publishers Limited. All rights reserved
    • REVIEWS Table 2 | Mouse models of HCC development resulting from HCV infection Transgenes expressed Morphological signs of HCC development Reference HCV envelope gene (E1–E2) Months 0–24: no adenoma or HCC development and high expression of envelope protein Months 1–18: absence of envelope proteins in the sera of mice Koike et al. (1995)39 HCV core gene Month 3: mice transgenic for HCV core genes developed hepatic steatosis Month 12: steatosis slowly progressed without neoplastic change >12 months: liver morphology indicated an age-dependent increase in oxidative stress Month 16: male mice—25.9% of C21 line and 30.8% of C49 line developed HCC. Female mice—0% of C21 line and 14.3% of C49 line developed HCC Moriya et al. (1998);40 Koike et al. (2002)41 HCV polyprotein (FL‑N) or structural protein (S‑N) >10 months: moderate to severe steatosis developed and increased with age ≥13 months: HCC developed in FL‑N/35 strain without hepatic fibrosis >18 months: HCC developed in S‑N/863 strain without hepatic fibrosis Lerat et al. (2002)43 HCV core gene or core E1–E2 gene Months 0–21: no HCC or adenoma development in transgenic mice with the FVBxC57BL/6 background Week 20: after initial treatment with diethylnitrosamine, both transgenic and nontransgenic mice had growth retardation Week 32: both transgenic and nontransgenic mice developed HCC after initial treatment with diethylnitrosamine Kamegaya et al. (2005)44 Abbreviation: HCC, hepatocellular carcinoma. IFN‑α were found using an animal model 10 years ago59 and verified by a randomized controlled trial, in which IFN‑α treatment prolonged overall survival for patients after curative resection.60 Whether a tumor will undergo local and distant spread is defined by determinants of the tumor cells such as proliferating activities and the ability of the tumor cells to respond to growth and survival signals derived from its microenvironment.61 Our animal model suggests that the lung is the most common organ of distant meta­ stasis from HCC,62 which corresponds with data from the clinic.63 Three key factors account for the success of this model. First, HCC cells with the highest metastatic potential were selected and inoculated into a maximally favorable microenvironment for metastatic behavior. To this end, a histologically intact intrahepatic meta­ tasis s from an HBV-positive Chinese patient with HCC was taken as the donor tumor to construct the LCI‑D20 model. Second, the donor tumor tissues were implanted into the livers of recipient nude mice. This maneuver ensures that the tumor has a similar histological environment to that inside the human body, which facilitates the display of its metastatic behavior. Third, tumor subclones were selected in vivo by several cycles of sampling lung metastases derived from these orthotopic xenografts and inoculating them into the livers of further nude mice. This process favors the enrichment of tumor cells with the greatest potential for spontaneous metastasis. The LCI-D20 model system has been very useful for studying the fundamental properties and practical problems of HCC. With regard to the long-standing debate over ‘nature versus nurture’ as the driver of cancer meta­ stasis, our clinical studies found that genetic changes favoring metastasis (nature) had already occurred in primary HCC tumors, even those at very early stages of development.64 By contrast, studies in our animal models have provided convincing evidence that environmental factors favoring the selection of cell clones with higher metastatic potentials (nurture) have a larger role than early genetic changes in driving HCC cells to become increasingly metastatic. Of note, this animal model differs from the clinical situation, as HCC cells undergo several cycles of optimized in vivo selection, which might account for disparities between the mouse model and the clinical picture in humans. HCC metastasis, therefore, is likely to involve the evolutionary interplay of nature and nurture. Animal models of spontaneous HCC development and metastasis can be used to study the mechanism of HCC progression as well as the best mode of intervention. Futakuchi et al.65 have established a rat model of in vivo HCC lung metastasis based on sequential intraperitoneal injection of diethylnitrosamine and administra­tion of drinking water containing the carcino­ gen N‑nitrosomorpholine for 16 weeks, by which time all animals had developed HCC. By week 23, lung meta­ stasis had occurred in 100% of these animals. A subsequent modification of this protocol produced a less aggressive disease model, in which the rats survived for a longer period and developed moderate lung metastasis, making these animals easier to maintain and facilitating both delivery of treatments66 and study of the multi­ step and multifactorial evolutionary process of HCC lung metastasis.67 Experimental models of HCC metastasis are also very useful for the study of the fundamental mechanisms of cancer spread. In investigations focused on the character­ istics of circulating tumor cells, Scatton et al.68 established an experimental model of HCC metastasis, in which the Mahlavu human HCC cell line was introduced into NOD–SCID mice by subcapsular injection, tail vein injection or direct infection into bone marrow. These models have produced some interesting findings. First, when Mahlavu cells were injected into the liver sub­ apsular c region to construct an orthotopic HCC model, the primary tumor continuously produced large numbers of circulating tumor cells, in a manner dependent on tumor size. Moreover, the circulating tumor cells remained viable in the bloodstream for over 60 days without producing any observable meta­stasis in other organs. Second, NATURE REVIEWS | GASTROENTEROLOGY & HEPATOLOGY © 2011 Macmillan Publishers Limited. All rights reserved VOLUME 9  |  JANUARY 2012  |  37
    • REVIEWS a b Pulmonary alveoli Type I epithelial cell HCC intrahepatic lesion Endothelial cell Nude mice liver LCI-D20: metastasis to liver, lung, lymph Cell culture MHCC97: liver and lung metastasis Type II epithelial cell Basal lamina Cell clone MHCC97L: lung metastasis 40% Hepatic veins Central veins Lung metastasis Vena cava Hepatic sinusoid Hepatic cord In vivo selection Round 3: HCCLM3 In vivo selection Bile canaliculi Kupffer cell MHCC97H: lung metastasis 100% Metastasis Subcutaneous tumor Endothelial cell Round 6: HCCLM6 In vivo selection Primary tumor Liver tumor Round 9: HCCLM9 Figure 2 | Theoretical basis for, and description of, a mouse model of HCC incorporating a variation in the propensity for lung metastasis. a | In humans, a primary HCC tumor metastasizes to the lung via the bloodstream. Experimental models of HCC that employ orthotopic implantation of intact tumor tissue into nude mice optimally mimic human primary HCC, as the tumor microenvironment is also transplanted, and are a good model of metastasis because the liver and lung microenvironments create favorable conditions for the growth of metastatic cells. b | A stepwise model of HCC lung metastasis was established by sequential transplantation and cell clone selection. MHCC97 cells contain heterogeneous subpopulations. By cloning culture technique, two subclones designated as MHCC97L with 40% spontaneous lung metastasis rate, and MHCC97H with 100% spontaneous lung metastasis rate, were derived from MHCC97 cells. MHCC97H cells were subjected to three, six and nine rounds of in vivo pulmonary metastasis selection, to produce cells with increasing numbers of lung metastasis nodules and shorter nude mice survival. This in vivo selection signifies the enrichment of metastasis-prone mechanisms. Abbreviation: HCC, hepatocellular carcinoma. when Mahlavu cells were injected into the tail vein, most tumor cells remained circulating in the bloodstream for up to 120 days, albeit without producing any increase in secondary meta­stases. Third, when directly injected into the bone marrow, the Mahlavu cells stayed at the injection site where they proliferated but showed little tendency to circulate in the blood or metastasize. Subsequent experiments using these cells and similar protocols confirmed that circulating Mahlavu cells mainly originated from the primary liver tumor, and that the number of circulating cells was directly proportional to the primary tumor size.69 From this series of experiments, the key message of academic and clinical importance is that the primary tumor is the major source of circulating tumor cells, which can remain viable in the bloodstream for an unexpectedly long time unless the primary tumor is brought under effective control (for example, by hepatectomy, which considerably reduces the number of circulating tumor cells) or the host defense mechanism is powerful enough to curb the spread of tumor cells. Models of HCC treatment The only potentially curative options for patients with HCC are partial hepatectomy (resection) or liver transplantation. However, most patients with HCC are not eligible for resection or transplantation owing to the presence of advanced disease, and these individuals have a poor prognosis (1-year survival of approximately 17.5% 38  |  JANUARY 2012  |  VOLUME 9  and 2‑year survival of 7.3%).70 In addition, even after resection, the recurrence rate of HCC remains high, which leads to poor cure rates and poor long-term survival. Other treatments used in patients with HCC include chemotherapy, radiotherapy, transarterial chemo­ embolization (TACE), 71,72 radiofrequency ablation (RFA), percutaneous alcohol injection73,74 and adjuvant interferon therapy.75,76 In addition, various emerging options, such as gene-based treatments and targeted therapies, are under investigation as potential treatments for HCC. Various animal models have been developed for investigating the effects of each of these therapies. Treatment modalities in clinical use New chemotherapeutic agents Most experiments to test the effects of potential drug treatments in animal models of HCC involve the subcutaneous implantation of human hepatoma cells. Although researchers have expressed doubts and criticisms about the validity of using effects on tumor xenografts to predict clinical activity,77 such human tumor xenografts have proven useful in the development of cytotoxic agents and mechanism-based cytostatic drugs (drugs directed at specific abnormalities that drive the malignant phenotype).78,79 A standard operating procedure should be followed for all xenograft experiments to minimize the limitations of these models and maximize their efficacy.78 www.nature.com/nrgastro © 2011 Macmillan Publishers Limited. All rights reserved
    • REVIEWS The National Cancer Institute has analyzed the usefulness of xenograft models in the development of cytotoxic drugs. Of the 33 agents that had cytotoxic activity in more than one-third of xenograft models tested, 15 (45%) were later proven to be clinically active, and 10— metho­ rexate, chlorambucil, 5‑fluorouracil, cisplatin, t paclitaxel, vinblastine, irinotecan, docetaxel, doxorubicin and cyclophosphamide—were subsquently developed into the now ‘standard’ anticancer drugs.80 A subcutaneous xenograft HCC model was used for the evaluation of chemotherapeutic agents by Huynh et al.,81 who established seven different primary HCC cell lines and corresponding subcutaneous xenograft models. Several key signaling pathways such as the Raf– mitogen-activated protein kinase–extracellular signal regulated kinase (Raf–MAPK–ERK) signaling pathway involved in HCC development and progression were elucidated in these models, which also provided valuable information for the design of molecular targeted agents. For instance, this model system has been used to assess the ability of small molecules, acting either alone or in combination with chemotherapeutic and biologic agents, to target these key signaling pathways in HCC. Small molecule inhibitors and combinations of agents that act on the Raf–MAPK–ERK pathway that have been investigated using these models include AZD6244 plus doxorubicin,82 bevacizumab plus rapamycin,83 brivanib alaninate, 84 everolimus, 85 sunitinib, 86 sorafenib plus rapamycin,87 AZD6244 plus rapamycin88 and AZD6244 plus sorafenib.89 These studies provide strong evidence to support clinical investigations of these drugs and all are currently part of phase II clinical trails in patients with HCC.79,90 In terms of replicating the tumor microenvironment and organ selectivity of human HCC, an orthopic model is superior to subcutaneous xenograft models (Box 2). Establishing high-quality orthotopic models is, however, technically more challenging than the construction of subcutaneous xenograft models. Conventional techniques of intrahepatic tumor implantation involve direct placement of tumor fragments or injection of free tumor cells. These techniques have the major disadvantage of possible inadvertent tumor seeding along the needle track or into the bloodstream, which could considerably compromise the validity of studies of treatment efficacy. To avoid such problems, Yang et al. 91 developed a modi­ ied surgical technique, in which a piece of f Gelfoam ® (Pharmacia & Upjohn Company, North Peapack, USA) is inserted into the liver incision after delivery of HCC cells. The Gelfoam® both facilitates hemostasis and forms a pocket that secures the injected tumor cells. This modified orthotopic model fully displays the progression of HCC from local tumor growth, through adjacent organ invasion, ascites, and eventually spontaneous pulmonary metastasis. The most prominent feature of this model is the substantially reduced rate of early pulmonary metastasis compared with that in models based on direct intrahepatic implantation of either tumor cells or tumor fragments. This technical modification could, therefore, help to avoid artificial Box 2 | Advantages of orthotopic over subcutaneous models of HCC ■■ Maximum mimicry of clinical settings, including tumor location, liver damage and biochemical changes ■■ Allow expression of specific genes and proteins ■■ Allow the development of advanced and metastatic disease ■■ Support an interaction between the tumor and host, in particular between tumor cells and microenvironment ■■ Generate the possibility for testing liver-directed hepatocellular carcinoma therapy Abbreviation: HCC, hepatocellular carcinoma. metastasis. The rat HCC model of Yang et al.91 has been successfully used to evaluate several molecular targeted drugs, including the mTOR inhibitor, sirolimus, and the vascular epidermal growth factor receptor (VEGFR) inhibitor, gefitinib, both of which can be used either alone or in combination with the traditional cytotoxic agent, doxorubicin.92–94 To reproduce the extensive liver disease that is associated with advanced HCC metastasis our group developed another orthotopic mouse model. 95 HCC cells transfected with vectors carrying the gene for the β‑subunit of human choriogonadotropin (β-hCG) were injected into the left liver lobe of SCID mice. In this model, urine levels of β‑hCG can be used as a surrogate marker of tumor burden. This model supports the efficacy of treatments for advanced HCC based on an anti-angiogenic drug combined with metronomic chemotherapy—a sustained, chronic and low-dose cancer chemotherapy—as a survival benefit for the mice was only achieved when the angiogenic drug and metronomic chemotherapy were combined.95 HCC patients often also have liver cirrhosis, a feature that is absent from most animal models of HCC. Schiffer et al.96 have nonetheless established an experimental rat model of cirrhosis that gives rise to HCC. Rats received weekly intraperitoneal injections of diethylnitrosamine for 16 weeks, followed by a 2‑week wash-out period to allow recovery from acute necrosis. This treatment caused cirrhosis in 14 weeks and led to multifocal HCC in 18 weeks. The importance of this model is that it has been used to demonstrate that gefitinib, which blocks epidermal growth factor receptor (EGFR) activity, is useful as a chemopreventive agent because it blocks the transition from cirrhosis to HCC. Similarly, Huang et al.97 used modified versions of this protocol to accelerate the development of rat HCC (from 16–18 weeks to 9–12 weeks). By adjusting the concentration of diethylnitrosamine solution in proportion to the body weight of the animals, the animals could be induced to develop either HCC alone, or HCC and liver cirrhosis simultaneously. This model could be suitable to evaluate the effects of treatment strategies that target both cirrhosis and HCC. New TACE modalities Although several animal models are available, that described by Yang et al. is perhaps the most suitable for NATURE REVIEWS | GASTROENTEROLOGY & HEPATOLOGY © 2011 Macmillan Publishers Limited. All rights reserved VOLUME 9  |  JANUARY 2012  |  39
    • REVIEWS evaluating the efficacy of TACE for HCC treatment, as these rats develop poorly differentiated tumors with a hypervascular property.91 Qian et al.98 used this model to investigate the therapeutic effect of TACE using polylactide-co-glycolide microspheres (PLCG), which have an improved half-life and tissue compatibility compared with present embolic agents. By combining PLCG TACE with mitomycin, the tumor growth rate decreased more than eightfold. The same model was also used to compare TACE alone with TACE combined with other approaches, such as immunotherapy (OK‑432, an anti­ tumor reagent with direct cytotoxic and cytostatic activity) and antiangiogenic therapy (TNP‑470, which selectively inhibit endothelial proliferation). Both the combination therapies led to significantly retarded tumor growth versus TACE alone.99 Investigational treatment modalities Radiolabeled vesicles In an effort to develop novel radiotherapy techniques for HCC that have enhanced efficacy and reduced radiation-induced liver disease, Vanpouille-Box et al.100 generated biomimetic lipid nanocapsules of 50 nm in diameter loaded with 188Re. These nanocapsules can penetrate deeply into tumor blood vessels because of their nanometric scale. This study used a rat model of diethylnitrosamine-induced HCC, which demonstrated that administration of 188Re nanocapsules containing a radiation dose of 80 mBq could improve the median survival of treated animals by 100% and reduce tumor mass by 50%, while keeping liver enzyme levels within twofold of the upper limit of normal values. This study has provided proof-of-principle evidence that tumorselective internal radiotherapy is a feasible strategy for HCC treatment. Gene therapy Gene therapy might provide an alternative treatment modality for established HCC. Graepler et al.101 transferred the genes for soluble vascular endothelial growth factor (VEGF) receptor 1 (sFlt‑1) and endostatin into rat Morris hepatoma (MH) cells, which secrete high levels of VEGF but do not normally express sFlt‑1. MH cells engineered to express both sFlt‑1 and endostatin were then injected subcutaneously into ACI rats, which blocked the activity of VEGF and retarded tumor growth. These results provide evidence that angiostatic gene therapy might be a feasible strategy for the treatment of established HCC. Thereafter, several other gene therapy strategies were used to treat HCC in subcutaneous models, including combined suicide/cytokine gene therapy (HSV-TK/hIL‑2),102 human plasminogen fragment containing five kringle regions (Plgk1‑5),103 DC40 ligand (CD40L)104 and TNF-related apoptosis-inducing ligand (TRAIL) combined with cisplatin,105 all of which provide potential treatement approaches for human HCC. As the vascularization of rabbit liver tumors is similar to that of human liver tumors, a rabbit VX2 tumor model106 has been used in preclinical studies of liver tumor imaging, chemotherapy and tumor etiology. Gu 40  |  JANUARY 2012  |  VOLUME 9  et al.107 developed a transarterially administered gene therapy consisting of TACE and an adenoviral vector carrying the p53 tumor suppressor gene, which showed significant suppression of tumor growth in the orthotopic VX2 model. To evaluate the antitumor effects of the liver-cancer-specific oncolytic adenovirus Ad5WS2 in vivo, Hsieh et al. 108 established a syngeneic HCC model by orthotopic injection of ML‑1 mouse HCC cells. Ad5WS2 significantly increased mice survival, compared with control and liver-cancer-nonspecific adenovirustreated animals. Moreover, Ad5WS2-treated mice only experienced a transient and slight increase in alanine aminotransferase (ALT) levels, in contrast to a significantly longer and higher increase in ALT levels in livercancer-nonspecific adenovirus treated animals. This study brings hope to the possibility of HCC-targeted oncolytic gene therapy. Irreversible electroporation To investigate the feasibility of using irreversible electro­ poration as a liver-directed ablation technique for the treatment of HCC, Guo et al.109 established an orthotopic rat model of HCC by injection of the N1‑S1 hepatoma cell line into the hepatic capsule of the rat. Using serial MRI scans and follow-up histopathological analyses to study the effect, the authors found that irreversible electro­ oration led to a reduction in tumor size of 32% p in one-dimensional maximum diameter, compared with a 110% increase in untreated animals. No major adverse effects were reported. The results suggest that irreversible electroporation could be a safe and effective targeted ablation of liver tumors. RNA interference Therapies for HCC based on RNA interference (RNAi) have shown promising preclinical results, and might be developed into another potentially curative therapeutic approach.110,111 However, almost all in vivo studies have employed subcutaneous xenograft models to evaluate the efficiency of RNAi-mediated HCC prevention and therapy,112–114 which raises concerns that the findings do not reflect the true clinical situation. This Review, therefore, focuses on findings in an orthotopic animal model of HCC (Box 2). Lin et al.115 used an orthotopic model of HCC to investigate the effect of an RNAi-based therapy utilizing an antisense oligonucleotide (cantide) targeted against human telomerase reverse transcriptase on tumor growth in vivo. In this model, a highly metastatic tumor (HCM‑Y89) derived from a human HCC surgical specimen was xenografted into nude mice. Compared with controls, cantide could substantially reduce tumor weight by up to 69%, reduce liver tumor relapse by up to 37.5%, and reduce lung metastasis by up to 58%. This orthotopic HCC model has also been employed to evaluate the anticancer effects of antisense oligonucleotides against survivin116 producing up to 61% tumor weight reduction, and type I insulin-like growth factor receptor 117 producing up to 72% tumor growth inhibition and up to 50% reduction in lung metastasis. Several unique features of these studies merit further comment. First, the tumor www.nature.com/nrgastro © 2011 Macmillan Publishers Limited. All rights reserved
    • REVIEWS xenografts consisted of pieces of previously frozen tumor tissue, which were implanted into the left lobe of the livers of recipient mice. The preservation of tumor tissue in liquid nitrogen rather than as a cell line retains the tumor microenvironment and, therefore, ensures that optimal conditions for displaying the full range of the tumor’s biological behaviors are maintained.77 Second, using an orthotopic rather than a subcutaneous xenograft model of HCC makes the study design as close to the clinical situation as possible, thereby increasing a study’s credibility. Third, this model enables detailed investigation of major issues relevant to HCC treatment, including reductions in the size of the primary tumor, decreases in postoperative tumor recurrence and pulmonary metastasis, and decreases in serum AFP levels. These studies provide proof-of-principle evidence that RNAi-based therapy is a promising strategy for HCC treatment. Conclusions Animal models of HCC have contributed to our improved understanding of liver carcinogenesis and progression, as well as to the design of comprehensive treatment strategies. From the HBV and HCV transgenic mouse models described in this Review, several features are clear. First, tumor development is slower in models based on HBV or HCV gene transfer alone than in models based on transfer of HCV or HBV genes in combination with other genes or chemical exposure. Second, HBV and HCV transgenic mice do not develop liver cirrhosis, suggesting that this symptom observed in patients with HCC could be attributed to factors other than the virus itself. Third, ROS generation and steatosis have a major role in the development of HCC, but 1. Jemal, A. et al. Global cancer statistics. CA Cancer J. Clin. 61, 69–90 (2011). 2. Parkin, D. M. The global health burden of infection-associated cancers in the year 2002. Int. J. Cancer 118, 3030–3044 (2006). 3. Ming, L. et al. Dominant role of hepatitis B virus and cofactor role of aflatoxin in hepatocarcinogenesis in Qidong, China. Hepatology 36, 1214–1220 (2002). 4. El-Serag, H. B. & Mason, A. C. Rising incidence of hepatocellular carcinoma in the United States. N. Engl. J. Med. 340, 745–750 (1999). 5. El-Serag, H. B. Epidemiology of hepatocellular carcinoma in USA. Hepatol. Res. 37 (Suppl. 2), S88–S94 (2007). 6. Tsai, W. L. & Chung, R. T. Viral hepatocarcinogenesis. Oncogene 29, 2309–2324 (2010). 7. Perz, J. F., Armstrong, G. L., Farrington, L. A., Hutin, Y. J. & Bell, B. P The contributions of . hepatitis B virus and hepatitis C virus infections to cirrhosis and primary liver cancer worldwide. J. Hepatol. 45, 529–538 (2006). 8. Chen, C. J. et al. Risk of hepatocellular carcinoma across a biological gradient of serum hepatitis B virus DNA level. JAMA 295, 65–73 (2006). 9. Liang, T. J. & Heller, T. Pathogenesis of hepatitis C‑associated hepatocellular carcinoma. Gastroenterology 127, S62–S71 (2004). 10. Heindryckx, F., Colle, I. & Van Vlierberghe, H. Experimental mouse models for hepatocellular carcinoma research. Int. J. Exp. Pathol. 90, 367–386 (2009). this might not be due to inflammatory cell infiltration. These features suggest that the virus itself could both initiate and promote carcinogenesis, whereas other factors, including exposure to carcinogens, have a secondary role (Figure 1). Thus, HBV and HCV transgenic mouse models provide plausible evidence that a one-hit mechanism of carcinogenesis could be the predominant pattern of HCC development. In the era of molecular medicine, increasing attention has been focused on the identification of key pathways and mediators to enable the development of targeted and individualized cancer treatments. In the rapidly evolving field of HCC model construction, cutting-edge technologies, such as proteomics (and various other ‘omics’ techniques), RNAi, microRNA and molecular imaging will be used to establish target-specific or signal-pathwayspecific models, for the study of disease mechanisms and to pinpoint treatment in the era of individualized medicine. Such models are expected to be the major future direction of HCC study. Meanwhile, technical specifications and unified procedures are important challenges for the establishment of stable yet dynamic animal models of HCC. Review criteria We searched for articles focusing on original research into animal models of HCC, with a particular emphasis on their potential application in HCC development, metastasis and treatment. A PubMed search was performed using the search terms “animal model” and “hepatocellular carcinoma”. All papers identified were English-language full-text papers. We also searched the reference lists of identified articles for further relevant papers. 11. Wu, L., Tang, Z. Y. & Li, Y. Experimental models of hepatocellular carcinoma: developments and evolution. J. Cancer Res. Clin. Oncol. 135, 969–981 (2009). 12. Fausto, N. & Campbell, J. S. Mouse models of hepatocellular carcinoma. Semin. Liver Dis. 30, 87–98 (2010). 13. Aravalli, R. N., Steer, C. J., Sahin, M. B. & Cressman, E. N. Stem cell origins and animal models of hepatocellular carcinoma. Dig. Dis. Sci. 55, 1241–1250 (2010). 14. Kalra, N. & Kumar, V. c‑Fos is a mediator of the c‑myc‑induced apoptotic signaling in serumdeprived hepatoma cells via the p38 mitogenactivated protein kinase pathway. J. Biol. Chem. 279, 25313–25319 (2004). 15. Singh, M. & Kumar, V. Transgenic mouse models of hepatitis B virus-associated hepatocellular carcinoma. Rev. Med. Virol. 13, 243–253 (2003). 16. Murata, M. et al. Hepatitis B virus X protein shifts human hepatic transforming growth factor (TGF)-beta signaling from tumor suppression to oncogenesis in early chronic hepatitis B. Hepatology 49, 1203–1217 (2009). 17. Gearhart, T. L. & Bouchard, M. J. The hepatitis B virus X protein modulates hepatocyte proliferation pathways to stimulate viral replication. J. Virol. 84, 2675–2686 (2010). 18. Zhao, J. et al. Epigenetic silence of ankyrin‑repeat‑containing, SH3‑domain‑containing, and proline‑rich‑regioncontaining protein 1 (ASPP1) and ASPP2 genes promotes tumor growth in hepatitis B virus- NATURE REVIEWS | GASTROENTEROLOGY & HEPATOLOGY © 2011 Macmillan Publishers Limited. All rights reserved 19. 20. 21. 22. 23. 24. 25. 26. positive hepatocellular carcinoma. Hepatology 51, 142–153 (2010). Cheng, B., Zheng, Y., Guo, X., Wang, Y. & Liu, C. Hepatitis B viral X protein alters the biological features and expressions of DNA repair enzymes in LO2 cells. Liver Int. 30, 319–326 (2010). Kim, C. M., Koike, K., Saito, I., Miyamura, T. & Jay, G. HBx gene of hepatitis B virus induces liver cancer in transgenic mice. Nature 351, 317–320 (1991). Yu, D. Y. et al. Incidence of hepatocellular carcinoma in transgenic mice expressing the hepatitis B virus X‑protein. J. Hepatol. 31, 123–132 (1999). Kim, S. Y. et al. Proteomic analysis of liver tissue from HBx-transgenic mice at early stages of hepatocarcinogenesis. Proteomics 9, 5056–5066 (2009). Cui, F. et al. The up-regulation of proteasome subunits and lysosomal proteases in hepatocellular carcinomas of the HBx gene knockin transgenic mice. Proteomics 6, 498–504 (2006). Wang, Y. et al. HBsAg and HBx knocked into the p21 locus causes hepatocellular carcinoma in mice. Hepatology 39, 318–324 (2004). Knudson, A. G. Jr. Mutation and cancer: statistical study of retinoblastoma. Proc. Natl Acad. Sci. USA 68, 820–823 (1971). Huebner, R. J. & Todaro, G. J. Oncogenes of RNA tumor viruses as determinants of cancer. Proc. Natl Acad. Sci. USA 64, 1087–1094 (1969). VOLUME 9  |  JANUARY 2012  |  41
    • REVIEWS 27. Lakhtakia, R. et al. Hepatocellular carcinoma in a hepatitis B ‘x’ transgenic mouse model: a sequential pathological evaluation. J. Gastroenterol. Hepatol. 18, 80–91 (2003). 28. Dunsford, H. A., Sell, S. & Chisari, F. V. Hepatocarcinogenesis due to chronic liver cell injury in hepatitis B virus transgenic mice. Cancer Res. 50, 3400–3407 (1990). 29. Chisari, F. V. et al. A transgenic mouse model of the chronic hepatitis B surface antigen carrier state. Science 230, 1157–1160 (1985). 30. Chisari, F. V. et al. Structural and pathological effects of synthesis of hepatitis B virus large envelope polypeptide in transgenic mice. Proc. Natl Acad. Sci. USA 84, 6909–6913 (1987). 31. Toshkov, I., Chisari, F. V. & Bannasch, P. Hepatic preneoplasia in hepatitis B virus transgenic mice. Hepatology 20, 1162–1172 (1994). 32. Huang, S. N. & Chisari, F. V. Strong, sustained hepatocellular proliferation precedes hepatocarcinogenesis in hepatitis B surface antigen transgenic mice. Hepatology 21, 620–626 (1995). 33. Barone, M. et al. Gene expression analysis in HBV transgenic mouse liver: a model to study early events related to hepatocarcinogenesis. Mol. Med. 12, 115–123 (2006). 34. Nakamoto, Y., Guidotti, L. G., Kuhlen, C. V., Fowler, P. & Chisari, F. V. Immune pathogenesis of hepatocellular carcinoma. J. Exp. Med. 188, 341–350 (1998). 35. Su, S. B. et al. Overexpression of p8 is inversely correlated with apoptosis in pancreatic cancer. Clin. Cancer Res. 7, 1320–1324 (2001). 36. Ray, R. et al. BNIP3 heterodimerizes with Bcl‑2/ Bcl‑X(L) and induces cell death independent of a Bcl‑2 homology 3 (BH3) domain at both mitochondrial and nonmitochondrial sites. J. Biol. Chem. 275, 1439–1448 (2000). 37. Sell, S., Hunt, J. M., Dunsford, H. A. & Chisari, F. V. Synergy between hepatitis B virus expression and chemical hepatocarcinogens in transgenic mice. Cancer Res. 51, 1278–1285 (1991). 38. Koike, K., Moriya, K. & Matsuura, Y. Animal models for hepatitis C and related liver disease. Hepatol. Res. 40, 69–82 (2010). 39. Koike, K. et al. Expression of hepatitis C virus envelope proteins in transgenic mice. J. Gen. Virol. 76 (Pt 12), 3031–3038 (1995). 40. Moriya, K. et al. The core protein of hepatitis C virus induces hepatocellular carcinoma in transgenic mice. Nat. Med. 4, 1065–1067 (1998). 41. Koike, K., Moriya, K. & Kimura, S. Role of hepatitis C virus in the development of hepatocellular carcinoma: transgenic approach to viral hepatocarcinogenesis. J. Gastroenterol. Hepatol. 17, 394–400 (2002). 42. Ichibangase, T., Moriya, K., Koike, K. & Imai, K. A proteomics method revealing disease-related proteins in livers of hepatitis-infected mouse model. J. Proteome Res. 6, 2841–2849 (2007). 43. Lerat, H. et al. Steatosis and liver cancer in transgenic mice expressing the structural and nonstructural proteins of hepatitis C virus. Gastroenterology 122, 352–365 (2002). 44. Kamegaya, Y. et al. Hepatitis C virus acts as a tumor accelerator by blocking apoptosis in a mouse model of hepatocarcinogenesis. Hepatology 41, 660–667 (2005). 45. Li, Y. et al. Stepwise metastatic human hepatocellular carcinoma cell model system with multiple metastatic potentials established through consecutive in vivo selection and studies on metastatic characteristics. J. Cancer Res. Clin. Oncol. 130, 460–468 (2004). 42  |  JANUARY 2012  |  VOLUME 9  46. Sun, F. X. et al. Metastatic models of human liver cancer in nude mice orthotopically constructed by using histologically intact patient specimens. J. Cancer Res. Clin. Oncol. 122, 397–402 (1996). 47. Tian, J. et al. New human hepatocellular carcinoma (HCC) cell line with highly metastatic potential (MHCC97) and its expressions of the factors associated with metastasis. Br. J. Cancer 81, 814–821 (1999). 48. Li, Y. et al. Establishment of cell clones with different metastatic potential from the metastatic hepatocellular carcinoma cell line MHCC97. World J. Gastroenterol. 7, 630–636 (2001). 49. Li, Y. et al. Establishment of a hepatocellular carcinoma cell line with unique metastatic characteristics through in vivo selection and screening for metastasis-related genes through cDNA microarray. J. Cancer Res. Clin. Oncol. 129, 43–51 (2003). 50. Ding, S. J. et al. From proteomic analysis to clinical significance: overexpression of cytokeratin 19 correlates with hepatocellular carcinoma metastasis. Mol. Cell. Proteomics 3, 73–81 (2004). 51. Andersen, J. B. et al. Progenitor-derived hepatocellular carcinoma model in the rat. Hepatology 51, 1401–1409 (2010). 52. Shi, Y. H. et al. Expression of X‑linked inhibitor‑of‑apoptosis protein in hepatocellular carcinoma promotes metastasis and tumor recurrence. Hepatology 48, 497–507 (2008). 53. Bai, D. S. et al. Capn4 overexpression underlies tumor invasion and metastasis after liver transplantation for hepatocellular carcinoma. Hepatology 49, 460–470 (2009). 54. Gao, Q. et al. Overexpression of PD‑L1 significantly associates with tumor aggressiveness and postoperative recurrence in human hepatocellular carcinoma. Clin. Cancer Res. 15, 971–979 (2009). 55. Yang, X. R. et al. CD24 is a novel predictor for poor prognosis of hepatocellular carcinoma after surgery. Clin. Cancer Res. 15, 5518–5527 (2009). 56. Shi, G. M. et al. CD151 modulates expression of matrix metalloproteinase 9 and promotes neoangiogenesis and progression of hepatocellular carcinoma. Hepatology 52, 183–196 (2010). 57. Liu, L. et al. Activation of β‑catenin by hypoxia in hepatocellular carcinoma contributes to enhanced metastatic potential and poor prognosis. Clin. Cancer Res. 16, 2740–2750 (2010). 58. Yang, Z. F. et al. Significance of CD90+ cancer stem cells in human liver cancer. Cancer Cell 13, 153–166 (2008). 59. Wang, L. et al. High-dose and long-term therapy with interferon-alpha inhibits tumor growth and recurrence in nude mice bearing human hepatocellular carcinoma xenografts with high metastatic potential. Hepatology 32, 43–48 (2000). 60. Sun, H. C. et al. Postoperative interferon alpha treatment postponed recurrence and improved overall survival in patients after curative resection of HBV-related hepatocellular carcinoma: a randomized clinical trial. J. Cancer Res. Clin. Oncol. 132, 458–465 (2006). 61. Paget, S. The distribution of secondary growths in cancer of the breast. 1889. Cancer Metastasis Rev. 8, 98–101 (1989). 62. Tang, Z. Y. et al. A decade’s studies on metastasis of hepatocellular carcinoma. J. Cancer Res. Clin. Oncol. 130, 187–196 (2004). 63. Yuki, K., Hirohashi, S., Sakamoto, M., Kanai, T. & Shimosato, Y. Growth and spread of 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. hepatocellular carcinoma. A review of 240 consecutive autopsy cases. Cancer 66, 2174–2179 (1990). Ye, Q. H. et al. Predicting hepatitis B viruspositive metastatic hepatocellular carcinomas using gene expression profiling and supervised machine learning. Nat. Med. 9, 416–423 (2003). Futakuchi, M. et al. Establishment of an in vivo highly metastatic rat hepatocellular carcinoma model. Jpn J. Cancer Res. 90, 1196–1202 (1999). Futakuchi, M., Ogawa, K., Tamano, S., Takahashi, S. & Shirai, T. Suppression of metastasis by nuclear factor kappaB inhibitors in an in vivo lung metastasis model of chemically induced hepatocellular carcinoma. Cancer Sci. 95, 18–24 (2004). Yoshino, H. et al. Modification of an in vivo lung metastasis model of hepatocellular carcinoma by low dose N.‑nitrosomorpholine and diethylnitrosamine. Clin. Exp. Metastasis 22, 441–447 (2005). Scatton, O. et al. Fate and characterization of circulating tumor cells in a NOD/SCID mouse model of human hepatocellular carcinoma. Oncogene 25, 4067–4075 (2006). Scatton, O. et al. Generation and modulation of hepatocellular carcinoma circulating cells: a new experimental model. J. Surg. Res. 150, 183–189 (2008). Cabibbo, G. et al. A meta-analysis of survival rates of untreated patients in randomized clinical trials of hepatocellular carcinoma. Hepatology 51, 1274–1283 (2010). Zhong, J. H. & Li, L. Q. Postoperative adjuvant transarterial chemoembolization for participants with hepatocellular carcinoma: a meta-analysis. Hepatol. Res. 40, 943–953 (2010). Wang, W., Shi, J. & Xie, W. F. Transarterial chemoembolization in combination with percutaneous ablation therapy in unresectable hepatocellular carcinoma: a meta-analysis. Liver Int. 30, 741–749 (2010). Zhou, Y. et al. Meta-analysis of radiofrequency ablation versus hepatic resection for small hepatocellular carcinoma. BMC Gastroenterol. 10, 78 (2010). Germani, G. et al. Clinical outcomes of radiofrequency ablation, percutaneous alcohol and acetic acid injection for hepatocelullar carcinoma: a meta-analysis. J. Hepatol. 52, 380–388 (2010). Shen, Y. C. et al. Adjuvant interferon therapy after curative therapy for hepatocellular carcinoma (HCC): a meta-regression approach. J. Hepatol. 52, 889–894 (2010). Singal, A. K., Freeman, D. H. Jr & Anand, B. S. Meta-analysis: interferon improves outcomes following ablation or resection of hepatocellular carcinoma. Aliment. Pharmacol. Ther. 32, 851–858 (2010). Newell, P Villanueva, A., Friedman, S. L., ., Koike, K. & Llovet, J. M. Experimental models of hepatocellular carcinoma. J. Hepatol. 48, 858–879 (2008). Kelland, L. R. Of mice and men: values and liabilities of the athymic nude mouse model in anticancer drug development. Eur. J. Cancer 40, 827–836 (2004). Newell, P Villanueva, A. & Llovet, J. M. ., Molecular targeted therapies in hepatocellular carcinoma: from pre-clinical models to clinical trials. J. Hepatol. 49, 1–5 (2008). Johnson, J. I. et al. Relationships between drug activity in NCI preclinical in vitro and in vivo models and early clinical trials. Br. J. Cancer 84, 1424–1431 (2001). www.nature.com/nrgastro © 2011 Macmillan Publishers Limited. All rights reserved
    • REVIEWS 81. Huynh, H., Soo, K. C., Chow, P. K., Panasci, L. & Tran, E. Xenografts of human hepatocellular carcinoma: a useful model for testing drugs. Clin. Cancer Res. 12, 4306–4314 (2006). 82. Huynh, H., Chow, P. K. & Soo, K. C. AZD6244 and doxorubicin induce growth suppression and apoptosis in mouse models of hepatocellular carcinoma. Mol. Cancer Ther. 6, 2468–2476 (2007). 83. Huynh, H. et al. Bevacizumab and rapamycin induce growth suppression in mouse models of hepatocellular carcinoma. J. Hepatol. 49, 52–60 (2008). 84. Huynh, H. et al. Brivanib alaninate, a dual inhibitor of vascular endothelial growth factor receptor and fibroblast growth factor receptor tyrosine kinases, induces growth inhibition in mouse models of human hepatocellular carcinoma. Clin. Cancer Res. 14, 6146–6153 (2008). 85. Huynh, H. et al. RAD001 (everolimus) inhibits tumour growth in xenograft models of human hepatocellular carcinoma. J. Cell. Mol. Med. 13, 1371–1380 (2009). 86. Huynh, H. et al. Sunitinib (SUTENT, SU11248) suppresses tumor growth and induces apoptosis in xenograft models of human hepatocellular carcinoma. Curr. Cancer Drug Targets 9, 738–747 (2009). 87. Huynh, H. et al. Sorafenib and rapamycin induce growth suppression in mouse models of hepatocellular carcinoma. J. Cell. Mol. Med. 13, 2673–2683 (2009). 88. Huynh, H. AZD6244 (ARRY‑142886) enhances the antitumor activity of rapamycin in mouse models of human hepatocellular carcinoma. Cancer 116, 1315–1325 (2010). 89. Huynh, H. et al. AZD6244 enhances the antitumor activity of sorafenib in ectopic and orthotopic models of human hepatocellular carcinoma (HCC). J. Hepatol. 52, 79–87 (2010). 90. Park, J. W. et al. Phase II, open-label study of brivanib as first-line therapy in patients with advanced hepatocellular carcinoma. Clin. Cancer Res. 17, 1973–1983 (2011). 91. Yang, R. et al. A reproducible rat liver cancer model for experimental therapy: introducing a technique of intrahepatic tumor implantation. J. Surg. Res. 52, 193–198 (1992). 92. Semela, D. et al. Vascular remodeling and antitumoral effects of mTOR inhibition in a rat model of hepatocellular carcinoma. J. Hepatol. 46, 840–848 (2007). 93. Piguet, A. C. et al. Inhibition of mTOR in combination with doxorubicin in an experimental model of hepatocellular carcinoma. J. Hepatol. 49, 78–87 (2008). 94. Piguet, A. C. et al. Everolimus augments the effects of sorafenib in a syngeneic orthotopic model of hepatocellular carcinoma. Mol. Cancer Ther. 10, 1007–1017 (2011). 95. Tang, T. C., Man, S., Lee, C. R., Xu, P & . Kerbel, R. S. Impact of metronomic UFT/ cyclophosphamide chemotherapy and antiangiogenic drug assessed in a new preclinical model of locally advanced orthotopic hepatocellular carcinoma. Neoplasia 12, 264–274 (2010). 96. Schiffer, E. et al. Gefitinib, an EGFR inhibitor, prevents hepatocellular carcinoma development in the rat liver with cirrhosis. Hepatology 41, 307–314 (2005). 97. Huang, K. W. et al. Dual therapeutic effects of interferon‑α gene therapy in a rat hepatocellular carcinoma model with liver cirrhosis. Mol. Ther. 16, 1681–1687 (2008). 98. Qian, J. et al. Application of poly‑lactide‑co‑glycolide‑microspheres in the transarterial chemoembolization in an animal model of hepatocellular carcinoma. World J. Gastroenterol. 9, 94–98 (2003). 99. Maataoui, A. et al. Transarterial chemoembolization alone and in combination with other therapies: a comparative study in an animal HCC model. Eur. Radiol. 15, 127–133 (2005). 100. Vanpouille-Box, C. et al. Lipid nanocapsules loaded with rhenium‑188 reduce tumor progression in a rat hepatocellular carcinoma model. PLoS ONE 6, e16926 (2011). 101. Graepler, F. et al. Combined endostatin/sFlt‑1 antiangiogenic gene therapy is highly effective in a rat model of HCC. Hepatology 41, 879–886 (2005). 102. Stefani, A. L. et al. Systemic efficacy of combined suicide/cytokine gene therapy in a murine model of hepatocellular carcinoma. J. Hepatol. 42, 728–735 (2005). 103. Schmitz, V. et al. Plasminogen fragment K1–5 improves survival in a murine hepatocellular carcinoma model. Gut 56, 271–278 (2007). 104. Iida, T. et al. Adenovirus-mediated CD40L gene therapy induced both humoral and cellular immunity against rat model of hepatocellular carcinoma. Cancer Sci. 99, 2097–2103 (2008). 105. Wang, Y. et al. The efficacy of combination therapy using adeno-associated virus-TRAIL targeting to telomerase activity and cisplatin in a mice model of hepatocellular carcinoma. J. Cancer Res. Clin. Oncol. 136, 1827–1837 (2010). 106. Ramirez, L. H. et al. Pharmacokinetics and antitumor effects of mitoxantrone after intratumoral or intraarterial hepatic administration in rabbits. Cancer Chemother. Pharmacol. 37, 371–376 (1996). 107. Gu, T. et al. Trans-arterial gene therapy for hepatocellular carcinoma in a rabbit model. World J. Gastroenterol. 13, 2113–2117 (2007). NATURE REVIEWS | GASTROENTEROLOGY & HEPATOLOGY © 2011 Macmillan Publishers Limited. All rights reserved 108. Hsieh, J. L. et al. Transthyretin-driven oncolytic adenovirus suppresses tumor growth in orthotopic and ascites models of hepatocellular carcinoma. Cancer Sci. 100, 537–545 (2009). 109. Guo, Y. et al. Irreversible electroporation therapy in the liver: longitudinal efficacy studies in a rat model of hepatocellular carcinoma. Cancer Res. 70, 1555–1563 (2010). 110. Romano, P . R., McCallus, D. E. & Pachuk, C. J. RNA interference-mediated prevention and therapy for hepatocellular carcinoma. Oncogene 25, 3857–3865 (2006). 111. Arbuthnot, P & Thompson, L. J. Harnessing the . RNA interference pathway to advance treatment and prevention of hepatocellular carcinoma. World J. Gastroenterol. 14, 1670–1681 (2008). 112. Li, H. et al. Use of adenovirus-delivered siRNA to target oncoprotein p28GANK in hepatocellular carcinoma. Gastroenterology 128, 2029–2041 (2005). 113. Cho-Rok, J. et al. Adenovirus-mediated transfer of siRNA against PTTG1 inhibits liver cancer cell growth in vitro and in vivo. Hepatology 43, 1042–1052 (2006). 114. Salvi, A., Arici, B., Alghisi, A., Barlati, S. & De Petro, G. RNA interference against urokinase in hepatocellular carcinoma xenografts in nude mice. Tumour Biol. 28, 16–26 (2007). 115. Lin, R. X., Tuo, C. W., Lu, Q. J., Zhang, W. & Wang, S. Q. Inhibition of tumor growth and metastasis with antisense oligonucleotides (Cantide) targeting hTERT in an in situ human hepatocellular carcinoma model. Acta Pharmacol. Sin. 26, 762–768 (2005). 116. Sun, Y., Lin, R., Dai, J., Jin, D. & Wang, S. Q. Suppression of tumor growth using antisense oligonucleotide against survivin in an orthotopic transplant model of human hepatocellular carcinoma in nude mice. Oligonucleotides 16, 365–374 (2006). 117. Lin, R. X. et al. Inhibition of hepatocellular carcinoma growth by antisense oligonucleotides to type I insulin-like growth factor receptor in vitro and in an orthotopic model. Hepatol. Res. 37, 366–375 (2007). Acknowledgments The authors receive funding from the Science Fund for Creative Research Groups of the National Natural Science Foundation of China (No. 20921062), and the Fundamental Research Funds for the Central Universities of Ministry of Education of China (No. 4103005; both to Y. Li). Author contributions Y. Li, Z.‑Y. Tang and J.‑X. Hou jointly researched data for the article, wrote the manuscript, and made substantial contributions to discussions of the content. In addition, Z.‑Y. Tang reviewed and edited the manuscript before submission. VOLUME 9  |  JANUARY 2012  |  43