haracterization of tumor suppressor genes hDAB2IP and DLEC1 in hepatocellular carcinoma

CHAPTER THREE Expression and Methylation of hDAB2IP Paper I 82 3.3.2 Down-regulation and Promoter Hypermethylation of hDAB2IPA in Liver 3.3.3 Hypermethylation of hDAB2IPA in Primary

CHARACTERIZATION OF

TUMOR SUPPRESSOR GENES hDAB2IP AND DLEC1

IN HEPATOCELLULAR CARCINOMA

QIU GUO-HUA (M.SC., China)

A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

DEPARTMENT OF PHYSIOLOGY NATIONAL UNIVERSITY OF SINGAPORE

2008

ACKNOWLEDGEMENTS

I would like to express my gratitude to all those who have helped me complete this PhD thesis Above all, I am deeply grateful to my supervisor, Associate Professor Hooi Shing Chuan, Head of Department of Physiology, National University of Singapore This work could not have been possible without his patient guidance, valuable discussion and consistent support I appreciated that Prof Hooi had a separate meeting with me every Saturday morning, where some good ideas were developed from the insightful conversations Those times will remain a nice memory for me I learned a lot from Prof Hooi about not only science, but English and life as well More importantly, I would like

to express my sincere appreciation to him for letting me complete the PhD study in his lab at my most difficult time

I thank to my previous supervisor Associate Professor Tao Qian in The Chinese University of Hong Kong for his guidance and support when I was in Johns Hopkins Singapore, where I started to study DNA methylation with inspiration

In addition, I wish to extend my appreciation to my colleagues, Puei Nam, Jianjun, Guodong, Mirtha, Baohua and Colyn for their assistance, discussion and friendship Special thanks to Huangming, Carol and Yuntong, for their effort in this project I appreciated the administrative assistance and friendship of Ms Asha Das, Vasantha Nathan, Jenny and Eileen I would also like to thank the members in my previous lab in Johns Hopkins Singapore, Dr Wen-Sen Hsieh, Dingxie, Zhaohui, Shiguo, Tzer Jing, Fu Li, John, Vivien, Tan Jing and Cai Yan for their assistance, support and

I am grateful as well to the advisory committee members, Associate Professor Yu Qiang and Dr Linda SH Chuang for beneficial suggestions and comments during my PhD Qualifying Exam

I would like to thank my parents in China for their love and support even though I have been so far away from them since beginning my university studies I regret that I am not able to spend more time with them

Most importantly, I wish to express my greatest appreciation to my wife Xie Xiaojin, for her love, continuous support, discussion, encouragement and tolerance throughout the duration of my part-time PhD study Thanks to my lovely daughters, Bi-Qing and Bi-Xin, whom I am proud of

Qiu Guo-Hua National University of Singapore

February 2008

1.1 Overview of Hepatocellular Carcinoma 3

1.2 Genetics of Hepatocellular Carcinoma 19

1.3 Epigenetics of Hepatocellular Carcinoma 31

1.4 Approaches to Screen Tumor Suppressor Genes 50

1.5 Research Objectives 59

CHAPTER TWO Materials and Methods 62

2.1 Cell Lines and Cell Culture 62

2.8 Methylation-specific PCR and Bisulfite Genomic Sequencing 68

2.9 Cloning of DLEC1 Open Reading Frame 68

2.13 Cell Proliferation Assay 70

2.14 Cell Cycle Analysis 70

2.15 Luciferase Reporter Assay 71

CHAPTER THREE Expression and Methylation of hDAB2IP (Paper I) 82

3.3.2 Down-regulation and Promoter Hypermethylation of hDAB2IPA in Liver

3.3.3 Hypermethylation of hDAB2IPA in Primary HCC and Correlation with

4.3.1 DLEC1 Expression is Down-regulated and Correlated to Promoter

4.3.2 CpG Island Methylation of DLEC1 in HCC Primary Tumors and

4.3.3 DLEC1 Inhibits Cell Proliferation, Induces G1 Cell Cycle Arrest and

CHAPTER FIVE Upregulation of p21 by DLEC1 116

5.2.2 Growth Inhibition by DLEC1 is Independent of p53, p21 and DNMT3B 119

CHAPTER SIX General Discussion and Conclusions 126

The first candidate tumor suppressor gene is hDAB2IP, which was screened by an MSP-based approach in HCC The results showed that of the two isoforms, hDAB2IPA

was the predominant one, being expressed in the majority of human normal tissues

examined The expression of hDAB2IPA was silenced or down-regulated but could be

restored by 5-aza-2’-deoxycytidine treatment in liver cancer cell lines The reactivation of

hDAB2IPA was due to the promoter demethylation These results indicate that DNA

methylation is involved in the downregulation of hDAB2IPA in HCC cell lines The correlation between promoter methylation and hDAB2IPA expression was confirmed in

eight pairs of matched HCC samples Furthermore, more than 80% of HCC samples

showed hDAB2IPA promoter methylation, compared to 11.5% in the corresponding adjacent normal tissue (p<0.0001, χ 2

) in the additional 53 pairs of patient samples

Consistent with its role as a tumor suppressor gene, hDAB2IPA is suppressed in HCC

mainly by DNA methylation at promoter region

The second candidate tumor suppressor gene is DLEC1, which was selected by a

RT-PCR-based approach to screen down-regulated genes at 3p21.3, a hot-spot for

chromosomal aberrations and loss of heterozygosity in HCC It was found that DLEC1

was silenced and hypermethylated in nine of 11 HCC cell lines examined Treatment with

5-aza-2'-deoxycytidine reversed the methylation and restored DLEC1 expression in four

cell lines The correlation between hypermethylation and expression was also

demonstrated in ten pairs of HCC and adjacent normal tissues (t-test, p<0.05) Hypermethylation of DLEC1 was detected in 70.6% of HCC tumors, compared to 10.3%

in normal tissues (n=68, p<0.001, χ2

) Of interest, DLEC1 methylation was associated with the AJCC staging of the tumors (p=0.036, χ2

) DLEC1 over-expression in cell lines

decreased colony formation, cell growth and cell size, and induced a G1 arrest in cell

cycle Our data indicate that DLEC1 is a candidate tumor suppressor gene silenced by

DNA methylation in HCC and plays an important role in the development and progression of HCC

The molecular mechanisms by which DLEC1 induces G1 cell cycle arrest were further investigated Using luciferase assay in GAL4 system, we showed that compared

to vector control, the luciferase activity of DLEC1 was activated > 2.5 fold by DLEC1 This indicates that DLEC1 is a transcriptional activator DLEC1 upregulates the transcription of p21 as determined by conventional and real-time RT-PCR and Western Blot Our results suggest that the G1 cell cycle arrest by DLEC1 is likely mediated by the upregulation of p21

LIST OF TABLES

Table 2-1 Primer sequences used for screening of tumor suppressor genes by MSP 73

Table 2-2 Primer sequences used for screening of tumor suppressor genes in

Table 3-1 Methylation status of candidate tumor suppressor genes in HCC 84

Table 3-3 Clinicopathological features and hDAB2IPA promoter methylation 93

Table 4-2 Clinicopathological features and DLEC1 promoter methylation 108

LIST OF FIGURES

Figure 1-1 Geographic variations of HCC by age-standardized mortality rates 4

Figure 1-2 Tumorigenesis and progressive development of HCC induced by

Figure 1-3 Methylation of cytosine is catalyzed by DNA methyltransferases

Figure 1-4 Structural model of nucleosome and major posttranslational

Figure 1-6 Diagram of methylation-sensitive restriction landmark genome

Figure 3-4 Correlation between downregulation and promoter hypermethylation

Figure 4-1 RT-PCR screening of genes or ESTs in 3p21.3 101

Figure 4-2 Correlation between CpG island methylation and DLEC1 expression 105

Figure 4-3 Methylation analysis of DLEC1 in HCC primary tumors 107

Figure 4-4 Inhibition of cell growth by DLEC1 in vitro 110

Figure 4-5 DLEC1 overexpression inhibits proliferation, reduces cell size and

Figure 5-1 The growth inhibition by DLEC1 is independent of p53, p21 and

LIST OF PUBLICATIONS

This thesis is partially based on the following original publications that are

referred in the text by their roman numerals All published papers were reproduced with permission from the publisher

I Qiu GH, Xie H, Wheelhouse N, Harrison D, Chen GG, Salto-Tellez M, Lai P,

Ross JA and Hooi SC Differential expression of hDAB2IPA and hDAB2IPB

in normal tissues and promoter methylation of hDAB2IPA in hepatocellular

carcinoma J.Hepatol 2007, 46:655-663

II Qiu GH, Salto-Tellez M, Ross JA, Yeo W, CuiY, Wheelhouse N, Chen GG,

Harrison D, Lai P,Tao Q and Hooi SC The tumor suppressor gene DLEC1 is

frequently silenced by DNA methylation in hepatocellular carcinoma and

induces G1 arrest in cell cycle J.Hepatol 2008, 48, 433-441.

III Qiu GH, Yun T, Leung CHW and Hooi SC DLEC1 induces G1 cell cycle

arrest through direct transcriptional upregulation of p21 in HCT116 (in

preparation)

AFP-L3 lens culinaris agglutinin (LCA)-reactive isoform of AFP

AXIN1 axis inhibition protein 1

CCRCC clear cell renal cell carcinoma

CDKN2A cyclin-dependent kinase inhibitor 2A

CKI casein kinase I

CRG-L2 cancer related gene-Liver 2

ELISA enzyme linked immunosorbent assays

HNPCC human hereditary nonpolyposis colorectal cancer

IGF2R insulin-like growth factor 2 receptor

IGF-II Insulin-like growth factor-II

Jak Janus kinase

Jak/Stat Janus kinase (Jak)-signal transducer and activator of transcription

factor (Stat)

KO knockout

MAPK mitogen activated protein kinase

MS-APPCR methylation-sensitive arbitrarily primed-polymerase chain reaction

MS-RDA methylation-sensitive representational difference analysis

MTT 3-(4,5-dimethylthiazolyl-2)-2,5-diphenyltetrazolium bromide

NSCLC non-small cell lung cancer

PBMCs Peripheral blood mononuclear cells

PTEN phosphatase and tensin homolog deleted on chromosome 10

RASSF1A RAS association domain family protein 1 A

RDA representational difference analysis

RFTA radiofrequency thermal ablation

RLGS restriction landmark genome scanning

RT-PCR reverse transcription PCR

SOCS suppressors of cytokine signaling

Stat signal transducer and activator of transcription factor

US ultrasound

CHAPTER ONE

Introduction

Cancer is characterized by the uncontrolled cell growth and spread of abnormal cells There are various types of cancer with varied causes, some of them with geographic predominance Cancer may affect every part of the human body and has become one of the most devastating diseases all over the world World Health Organization (WHO) estimated in World Cancer Report that more than 10 million people are diagnosed with cancer every year (Stewart and Kleihues, 2003) Cancer accounts for nearly 12% of annual deaths worldwide, namely six million people and affects almost every family in most countries

Cancer is generally accepted as a genetic disease It has been proposed that tumor usually arises from cells that have accumulated multiple genetic abnormalities, including the chromosomal instabilities, alterations of tumor suppressor genes and proto-oncogenes (Fearon and Vogelstein, 1990).Tumor suppressors are inactivated by ‘loss-of-function’, thus causing the loss of control over cell growth, while protooncogenes are constitutively activated through ‘gain-of-function’, leading to continuous growth signaling which in turn stimulates cell growth

Cancer is also an epigenetic disease An increasing body of evidence shows that besides DNA deletion and mutations, DNA methylation, a major form of epigenetic modification, is an alternative mechanism to inactivate tumor suppressor genes and related genes in cancer Many tumor cells have aberrant, cell-heritable patterns of DNA

methylation that silences tumor suppressor genes and thus enhances the process of

tumorigenesis Therefore, a greater understanding of the mechanisms underlying DNA methylation of tumor suppressor genes is extremely important for unraveling the possible roles of tumor suppressor genes in the progression from normal to tumor cells This has direct complication for early tumor detection and disease monitoring (Jablonka and Lamb, 2002)

Hepatocellular carcinoma (HCC) is the fifth most common cancer in the world, accounting for 5.6% of all human cancers It is the third most common cause of cancer

mortality with 548, 000 deaths in 2000 alone (Bosch et al, 2005) It is a significant public

health problem but knowledge about mechanisms of HCC and treatment is still

insufficient In Asia, it is one of the most common malignancies with high fatality rate Survival rates of primary HCC are very poor even in developed countries because of the lack of markers for the early detection and diagnosis Tumor suppressor genes are good

as diagnostic markers because epigenetic changes of these genes in tumors can be easily detected However, information about methylation alterations of tumor suppressor genes

in HCC is relatively limited, compared to other cancers

The aim of this study is to identify candidate tumor suppressor genes in HCC and then to investigate the methylation status of these genes The thesis will focus on the

candidate tumor suppressor genes hDAB2IP and DLEC1 in HCC and their molecular

mechanisms in tumorigenesis Prior to presenting my results, I will make an overview of HCC and then review our understanding of HCC at the molecular level, including the alterations of oncogenes, tumor suppressor genes and their related genes during

hepatocarcinogenesis

1.1 Overview of Hepatocellular Carcinoma

Hepatocellular carcinoma (HCC) is one of the primary hepatic neoplasms that arise from hepatocytes, the major cell type of the liver (Farazi and DePinho, 2006) It is the most common type of malignant primary liver cancer, representing 75-90% of all cases in most countries (McGlynn and London, 2005)

1.1.1 Incidence and Mortality

HCC is the fifth most common cancer in the world, accounting for 5.6% of all human cancers (7.5% among men and 3.5% among women) Approximately 564,000 new cases are estimated worldwide, corresponding to 398,364 in men and 165,972 in

women in 2000 (Bosch et al, 2005) In Singapore, it is the fourth commonest cancer in

males (Singapore Cancer Society) There is a wide geographic variability in HCC

incidence The majority of HCC (>80%) occurs in either sub-Saharan Africa or Eastern Asia High-rate (age-standardized incidence rate > 20/100,000) areas include Gambia (male, 39.7), South Korea (48.8), China (35.2), Senegal (28.5) and Singapore (28) Medium-rate (5.0 –20.0) areas are countries in Southern European, such as Spain (7.5), Italy (13.5), and Greece (12.1) Low-rate (<5.0) areas are countries in North and South America, Northern Europe, and Oceania Typical low incidence rates are those of Canada (3.2), Colombia (2.2), United Kingdom (2.2) and Australia (3.6) (El-Serag and Rudolph, 2007; McGlynn and London, 2005; Guan, 1996)

HCC is the third most common cause of cancer mortality and 548, 000 deaths due

to liver cancer, corresponding to 383,593 in men and 164,961 in women in 2000 (Bosch

et al, 2005) Survival rates of primary liver cancer are uniformly poor in both high-rate

and low-rate areas (Figure 1-1) The International Agency for Research on Cancer

(IARC) estimates that the age-standardised worldwide incidence rate of primary liver cancer among males is 17.4/100 000 in underdeveloped countries and 8.7/100 000 in developed countries (McGlynn and London, 2005)

Figure 1-1 Geographicvariations of HCC age-standardized mortality rates The rates are

represented per 100,000 persons per year (Adapted from El-Serag and Rudolph, 2007 and prepared with SmartDraw 2008)

In most areas of the world, the HCC incidence among women is a quarter to half

of that among men The biggest differences between male and female rates no longer occur among populations at high-risk HCC, but among the populations in central and

southern Europe (El-Serag and Rudolph, 2007; Bosch et al, 2005) The reasons are not

well understood It might partially be explained by the sex-specific prevalence of risk factors Males are more likely to be infected with HBV and HCV, consume more alcohol and smoke more cigarettes Androgenic hormones and increased genetic susceptibility may also increase risk among males (McGlynn and London, 2005)

1.1.2.3 Race

HCC incidence rates can vary obviously among population of different ethnicities living in the same region For example, the age-adjusted rate of Chinese versus Indian is

21.21 versus 7.86 per 100,000 persons among males in the ethnic Indian, Chinese, and Malay populations in Singapore between 1993 and 1997 The ethnic differences in rates among women are almost the same as rates among men The variation in incidence rates

is almost consistent with the likelihood of infection with hepatitis B viruses (HBV) and hepatitis C viruses (HCV), despite that genetic susceptibility and exposure patterns to other risk factors may also play a role (El-Serag and Rudolph, 2007; McGlynn and London, 2005)

1.1.3 Etiology of HCC

The leading causes of HCC are chronic infection with Hepatitis B viruses (HBV, 50-55%), Hepatitis C viruses (HCV, 25-30%), consumption of aflatoxin B1 (AFB1) or/and alcohol In high-rate HCC areas, HBV and AFB1 are the major factors, whereas HCV and alcohol are more important factors in low- to medium-rate areas Overall, it is estimated that HBV and HCV infections are causally associated with 75% to 80% of

HCC in the world (Bosch et al, 2005).

1.1.3.1 Hepatitis B Virus

The hepatitis B virus (HBV) has a compact, partially double-stranded DNA genome It can cause an acute and transient or a chronic infection of the liver HBV is the most frequent cause of HCC, with an estimated 300 million persons with chronic

infection worldwide Case-control studies have shown that chronic HBV carriers have a 5- to 15-fold increased risk of HCC compared with the general population The majority

(70% to 90%) of HBV-related HCC patients develop with cirrhosis (El-Serag and

Rudolph, 2007; McGlynn and London, 2005) Multiple processes are involved in the HBV-induced hepatocarcinogenesis The HBV genome encodes several viral proteins which are essential to its life cycle and its direct involvement in the functional processes, including inactivation of p53 by HBx binding, host–viral interactions (induction of

oxidative stress), sustained cycles of necrosis–inflammation–regeneration and targeted activation of oncogenic pathways (Farazi and DePinho, 2006)

1.1.3.2 Hepatitis C Virus

Hepatitis C virus (HCV) has a positively oriented, single-stranded RNA genome, without a DNA stage in its lifecycle Hepatocytes in the liver are the main sites of HCV replication Similar to HBV, HCV may cause a chronic infection of a very long duration, accompanied by a slowly evolving liver disease Chronic HCV infection is a major risk factor for the development of HCC HCC risk is increased 17-fold in HCV-infected patients compared with HCV-negative controls by promoting fibrosis and eventually cirrhosis (McGlynn and London, 2005) HCV-induced hepatocarcinogenesis provokes biological processes similar to those by HBV, but is associated with a propensity of HCV

to evade the host’s immune responses and to promote cirrhosis (Farazi and DePinho, 2006)

1.1.3.3 Aflatoxin

Aflatoxin B1 (AFB1) is a mycotoxin produced by the Aspergillus fungi The fungi

grow easily on such foodstuffs as corn and peanuts stored in warm and damp conditions

Although there are four principal aflatoxins, B1, B2, G1, and G2, animal experiments showed that AFB1 is a powerful hepatocarcinogen, leading IARC to classify it as a human carcinogen in 1987 (El-Serag and Rudolph, 2007; McGlynn and London, 2005) Once ingested, AFB1 is metabolized to an active intermediate, which can bind to DNA

and cause DNA damage, producing a characteristic mutation in the TP53

tumor-suppressor gene (249ser) This mutation has been observed in 30%–60% of HCC tumors

in aflatoxin-endemic areas (Hussain et al, 2007) Further, it was estimated that AFB1

consumption increases HCC risk four-fold, HBV infection increases HCC risk fold, and the combination of AFB1 and HBV increases HCC risk 60-fold (El-Serag and Rudolph, 2007; McGlynn and London, 2005)

chronic HBV infection (El-Serag and Rudolph, 2007; Bosch et al, 2005; McGlynn and

London, 2005) Alcohol-induced hepatocarcinogenesis is associated with the production

of proinflammatory cytokine and, consequently, the stimulation of cycles of hepatocyte

necrosis and regeneration, oxidative stress, fibrosis and cirrhosis (Farazi and DePinho, 2006)

The common pathogenetic pathways and processes of the diverse

hepatocarcinogenesis mechanisms have been summarized in Figure 1-2 At the molecular level, p53 inactivation or mutation seems to be a consistent event in HBV-, HCV- and AFB1-induced HCC On the other hand, inflammation, continuous rounds of necrosis and regeneration, and oxidative stress are characteristic of HBV-, HCV- and alcohol-induced hepatocarcinogenesis, suggesting that these processes contribute in fundamental ways to HCC development

1.1.4 Host Factors

1.1.4.1 Cirrhosis

Although there is a great difference in the incidence rates of HCC according to geographical areas in different populations, cirrhosis is a common stage in the

tumorigenesis of the most HCC cases in all areas (Llovet et al, 2003)

Histopathologically, cirrhosis is abnormal nodule surrounded by dense bands of fibrous tissue produced by alteration of the normal hepatocytes The changes must be diffuse throughout the liver Cirrhosis is an end stage of chronic diffuse liver disease and the most advanced stage of fibrosis (Farazi and DePinho, 2006; Okuda, 2007) HCC develops from cirrhosis at a fairly constant rate of about 3% yearly (de and Dell'Era, 2007)

Cirrhosis is observed in the majority (70-80%) of HCC patients and has been considered

as a pre-neoplastic condition of HCC ( Okuda, 2007; McGlynn and London, 2005) The

risk of HCC development from chronic hepatitis or cirrhosis varies according to the degree of fibrosis The risk of cirrhotic patients (F4) is the highest at 5.8% per patient yearly, which is much higher than those who have less fibrosis (F1-F3, 0.5-2.6%)

(Okuda, 2007)

1.1.4.2 Hemochromatosis

Hereditary hemochromatosis (HH) is an autosomal recessive disorder that is characterised by excessive dietary iron absorption and subsequent deposition in the parenchymal cells of the liver, pancreas, heart, joints and pituitary gland The majority of

HH is associated with two missense mutations, C282Y and H63D, in the HFE gene on chromosome 6. Recent studies have demonstrated that the relative risk for development

of HCC in patients with HH is estimated to be close to 20-fold The risk is increased in the presence of a variety of cofactors including male sex, age older than 50 years,

drinking, smoking, and HBV or HCV infections Several studies from different cohorts over a period of 3 decades have found notably similar rates of HCC development in HH patients, approximately 10% overall (McGlynn and London, 2005; Kowdley, 2004) Iron overload has been shown to cause cellular proliferation, DNA damage, inactivation of

tumor suppressor genes such as TP53, formation of reactive oxygen species within the

liver, induction of lipid peroxidation, acceleration of fibrogenesis and immunologic abnormalities (Kowdley, 2004)

1.1.4.3 Obesity and Diabetes

Obesity is now widely documented as a significant risk for the development of many cancers, including HCC A study in US shows that the relative risk of dying from HCC is 1.68 times higher among women and 4.52 times higher for men with a baseline body mass index (BMI) ≥35 kg/m2

than those with BMIs of 18.5 to 24.9 kg/m2

(McGlynn and London, 2005; Caldwell et al, 2004) Obesity is the most significant risk

factor for diabetes, and the two conditions are highly related events Hepatic

inflammation leads to oxidative stress/lipid peroxidation, which can cause hepatic injury, fibrosis, and eventual cirrhosis Several studies have provided evidence that viral

hepatitis, alcohol, and diabetes interact synergistically in affecting the development of HCC(McGlynn and London, 2005; Caldwell et al, 2004; Yu and Yuan, 2004)

As discussed in epidemology, many risk factors are involved in the slow process

of hepatocarcinogenesis (Figure 1-2) During the long preneoplastic stage, the liver is often the site of chronic infection with hepatitis, cirrhosis, or both, and hepatocyte cycling

is accelerated by upregulation of mitogenic pathways This leads to the production of monoclonal populations of abnormal hepatocytes, the development of dysplastic

hepatocytes in foci and nodules with altered phenotype and eventually the appearance of HCC The molecular pathogenesis of this progressive neoplasm, including the genetic and epigenetic alterations will be discussed in the later parts

Figure 1-2 Tumorigenesis and progressive development of HCC induced by various risk factors

(Adapted from Thorgeirsson and Grisham, 2002 and Farazi and DePinho, 2006)

1.1.5 Diagnosis and Therapeutic Options of HCC

1.1.5.1 Serological Markers

Disease-specific serological markers play important roles in the following aspects

in HCC (1) screening of HCC in high-risk persons to add the chance of receiving curative treatment and to improve survival; (2) staging of HCC to provide prognostic information; (3) monitoring therapeutic effectiveness (Hayashi and Di Bisceglie, 2006) Ideally,

serological markers for HCC should possess the following characteristics (1) high

sensitivity and specificity for the diagnosis of HCC; (2) convenience of the assays and (3) inexpensiveness of assays To date, there are three common serological markers for HCC, namely, total -fetoprotein, lens culinaris agglutinin-reactive AFP and protein induced by vitamin K absence or antagonist-II (PIVKA-II) (Yuen and Lai, 2005)

1.1.5.1.1 -fetoprotein

-fetoprotein (AFP) is a glycoprotein with a molecular weight of around 70 kDa and was first described as a marker for HCC in the 1960s AFP gene is highly expressed

in hepatocytes and endodermal cells of the yolk sac during fetal development Its

expression is repressed after birth Pathological elevation of AFP is observed in

hepatocyte regeneration, hepatocarcinogenesis and embryonic carcinomas The serum AFP levels of healthy subjects should normally be less than 20 ng/ml An AFP greater than 400 ng/ml in a cirrhotic is diagnostic, but the percentage of HCC patients with such high levels is only 4.5–22%, especially for patients with small lesions This represents

one of the most important limits for this marker ( Yuen and Lai, 2005; Song et al, 2002)

AFP-L3 is the lens culinaris agglutinin (LCA)-reactive isoform of AFP and the sensitivity of AFP can be improved by measuring this isoform As a marker for HCC, AFP-L3 is superior to the total AFP not only in the accuracy of diagnosing HCC, but also

in the correlation with the stage and the prognosis of the disease The percentage of L3 over the total AFP levels is used as a specific index for HCC (Yuen and Lai, 2005;

Song et al, 2002) However, the conclusion about L3 in HCC is controversial

AFP-L3 cannot be considered as a more reliable marker than AFP in HCC detection and is not therefore useful in surveillance/diagnostic studies since it is very difficult to standardize the conversion of a qualitative by densitometry (Giannelli and Antonaci, 2006)

Since DCP and total AFP levels are independent of each other in the setting of HCC and neither one is ideal as a marker for HCC, combination of these two markers increases the sensitivity, specificity and diagnostic accuracy It has also been shown that

combination of DCP and AFP-L3 is more effective for the early detection of HCC

(Giannelli and Antonaci, 2006; Yuen and Lai, 2005)

immunosorbent assays (ELISA) is detectable in around 40–53% of HCC patients whereas

it is not detected in the serum of healthy individuals The level of GPC3 is independent of total AFP levels in HCC patients Combination of GPC3 and total AFP also increases the sensitivity without affecting the specificity Thus, GPC3 could be a good supplementary molecular marker to AFP in the detection of HCC (Yuen and Lai, 2005)

1.1.5.1.4 Other Novel Markers

Golgi Protein-73 (GP73) is a resident Golgi type II transmembrane protein

expressed primarily in epithelial cells of many human tissues In normal human liver, its expression is detected in biliary epithelial cells, but barely detectable in hepatocytes However, the expression of GP73 is strongly up-regulated in hepatocytes in patients with liver disease GP73 is also elevated in the serum of HCC patient compared to non-

neoplastic liver Thus, GP73 may be a potential marker for the early detection of HCC

(Marrero et al, 2005) Although this is a promising study, more investigations are

required to confirm these data and clarify the role of this marker in clinical practice for

the early detection of HCC (Giannelli and Antonaci, 2006; Marrero et al, 2005)

CRG-L2 (Cancer related gene-Liver 2) is a novel gene identified by

representational difference analysis to compare normal liver and liver tumors obtained from DEN-treated C3H/HeJ mice It is upregulated in both mouse and human HCC Using this mouse model, CRG-L2 was detected in 69% (311/453) of the 32-week tumors examined using in situ hybridization and in 55% of the preneoplastic foci in 20-week-old DEN-treated mice But AFP was found to be upregulated in 30% of preneoplastic foci In comparison with AFP in these studies, CRG-L2 may be a more sensitive marker for the

early detection of HCC (Graveel et al, 2003)

1.1.5.2 Diagnosis

HCC in cirrhotic patients is diagnosed by (1) AFP level, (2) imaging studies, and (3) histologic diagnosis The higher the AFP level, the more specific it is for HCC An AFP greater than 400 ng/mL in a cirrhotic with a vascular hepatic mass on imaging is diagnostic Unfortunately, many HCC cases have only modestly elevated AFP values (Hayashi and Di Bisceglie, 2006)

Imaging studies, including abdominal ultrasound (US), contrast-enhanced

computed tomography (CT), and magnetic resonance imaging (MRI) are powerful tools

in HCC diagnosis US is able to detect earlier HCC, and CT or MRI can assess accurate

tumor burden (Lencioni et al, 2005) Large HCCs are usually easy to be identified, while

small lesions (<2–3 cm) may have subtle vascular markings Once US detects a hepatic nodule, the next clinical step is to characterize the nodule and establish its diagnosis

(Bruix and Sherman, 2005a) Dynamic imaging of tissues is critical for this purpose because it shows the characteristic vascular profile of HCC According to these data, the recent American Association for the Study of Liver Diseases (AASLD) guidelines

proposed the following diagnostic strategy (Bruix and Sherman, 2005b) In brief, minute nodules <1 cm in diameter are proposed for careful follow-up, as current diagnostic techniques are not accurate enough to confidently establish the diagnosis Nodules

between 1 and 2 cm should be characterized by imaging techniques If two of them show coincidental and specific dynamic pattern, the HCC diagnosis can be established Finally, nodules >2cm in size that exhibit a characteristic dynamic profile can be diagnosed as HCC by using a single imaging technique

The European Association for the Study of Liver Disease published the guidelines

in 2001, which are generally accepted currently and provide the diagnostic criteria of

HCC (Bruix et al, 2001) 1 Cytohistopathologic diagnosis; or 2 > 2-cm arterial

hypervascular lesion detected by two coincident imaging techniques in the setting of cirrhosis; or 3 > 2-cm arterial hypervascular lesion detected by one imaging technique with serum AFP >400 ng/mL in the setting of cirrhosis

1.1.5.3 Treatment

Treatment for HCC depends on the stage of the tumor and of the chronic liver disease or cirrhosis Tumor size, hepatic functional reserve or portal hypertension limits indication of surgical or percutaneous ablation, and the success of treatment is burdened

by the high recurrence rate (Avila et al, 2006) HCC is less tolerable to therapeutical

treatment than any other cancers because HCC typically evolves in the setting of

cirrhosis, which increases both operative and chemotherapeutic risks Furthermore, advanced liver failure produces significant baseline mortality unrelated to HCC (Blum, 2005)

Therapies for HCC can be divided into four categories: surgical interventions, percutaneous interventions, transarterial interventions and drug treatment Potentially curative therapies are tumor resection, liver transplantation, and percutaneous

interventions that can result in complete responses and improved survival in a high

proportion of patients (Hayashi and Di Bisceglie, 2006; Kurtovic et al, 2005) Usually,

for patients with preserved liver function, resection is the preferred treatment If resection

is impossible due to poor liver function, liver transplantation is the treatment of choice for the HCC at early stage Other nonsurgical treatments including percutaneous ethanol injection, radiofrequency ablation, and transarterial chemoembolization are employed for

unresectable patients to prevent tumor progression (Schwartz et al, 2007; Avila et al,

to block their transition into chronic hepatitis that carries the risk for developing liver cirrhosis and HCC Stage 3: Interventions at this step are aimed to prevent the

progression of chronic hepatitis to liver cirrhosis that carries a high risk for HCC

development Stage 4: Interventions at this step are aimed at interfering with the

molecular events leading to HCC development (Blum, 2005)

1.1.5.4.2 Secondary HCC Prevention

The prevention of a local recurrence and/or the development of new HCC lesions

in patients after successful surgical or non-surgical HCC treatment is of paramount

importance and is expected to significantly improve disease-free and overall patient

survival ( Avila et al, 2006; Blum, 2005)

1.2 Genetics of HCC

Like in most other solid tumors, tumorigenesis in HCC is a slow and multistep genetic and epigenetic process Accumulation of genetic and epigenetic abnormalities of chromosomes, oncogenes, tumor suppressor genes, and genes in DNA repair, cell cycle regulation, cell adhesion molecules and growth factor/receptor systems progressively alter the hepatocellular phenotype to produce cellular intermediates that evolve into HCC Current evidences indicate that the long preneoplastic process and the early stages in HCC development are characterized by certain common features resulting from both genetic and epigenetic alterations These common traits include the progressive

hepatocyte dedifferentiation because of impaired liver-specific gene expression, and the alteration of numerous signaling pathways leading to autonomous and dysregulated cell proliferation and resistance to cell death (Cha and Dematteo, 2005; Thorgeirsson and

Grisham, 2002; Feitelson et al, 2002) Some genetic alterations in a few genes and

chromosomal loci develop slowly during the early preneoplastic phase, but increase markedly in dysplastic hepatocytes and HCCs Allelic deletions are detected in 30–50%

of livers with chronic hepatitis or cirrhosis, in 70–80% of dysplastic nodules and in almost all HCCs Thus, the multiplicity of allelic deletions in affected cell populations is low in chronic hepatitis but rises sharply in dysplastic hepatocytes, and is highest in HCCs (Thorgeirsson and Grisham, 2002)

1.2.1 Microsatellite and Chromosomal Instability

A large number of chromosomal rearrangements have been found in tumor hepatocytes in HCC, leading to highly complex karyotypes Hyperploid DNA content was found in 43% of dysplastic lesions in cirrhotic disease and in 50% of tumor cells in the analyzed HCC cases, suggesting a global gain of genetic material This frequency increases in high-grade dysplastic lesions, suggesting that chromosome losses followed

by endometriosis are early steps in hepatocarcinogenesis (Laurent-Puig and Rossi, 2006) Microsatellite instability occurs in hepatocytes in some chronic hepatitis, cirrhosis and HCC Microsatellite instability at both identical loci and identical allelic deletions or gene mutations has been described in cirrhotic and dysplastic nodules and adjacent HCCs, indicating that HCCs often arise as clonal outgrowths of cirrhotic

Zucman-(dysplastic) nodules (Thorgeirsson and Grisham, 2002)

Using microsatellite allelotypes to detect loss of heterozygosity (LOH), or

comparative genomic hybridization (CGH) to detect gains and losses of chromosome

samples based on the chromosomal DNA comparison between tumor and non-tumor In summary, LOH presents in 25–45% of HCC patients in 1p, 1q, 4q, 5q, 6q, 8p, 9p, 13q, 16p, 16q, and 17p while the most frequent gains (30–55%) in chromosomes 1q, 7q, 8q

and 17q (Thorgeirsson and Grisham, 2002; Laurent-Puig et al, 2001) As chromosomal

loss is a frequent mechanism to inactivate one allele of a tumor suppressor gene in solid tumors, LOH in a particular chromosomal region in a tumor may indicate the presence of

a tumor suppressor gene that contributes to the tumor In HCC, tumor suppressor genes, targeted by LOH, have been identified on chromosome 17p, 13q, 16p, 9p, 6q and 5q

corresponding to inactivation of TP53, RB (retinoblastoma 1), AXIN1 (axis inhibition protein 1), CDKN2A (cyclin-dependent kinase inhibitor 2A, also named p16), IGF2R (insulin-like growth factor 2 receptor), and APC respectively (Laurent-Puig and Zucman-

Rossi, 2006) More detailed about these genes will be discussed later

1.2.2 Activation of Oncogenes

Oncogenes refer to those genes whose activation can contribute to the

development of tumors The strict definition would require that activation be proven in human primary tumors and that experimental activation of the gene in cultured cells or

animal model could recapitulate the malignancy (Mascaux et al, 2005) They are derived

from normal genes (the proto-oncogene) coding for proteins, which play key roles in physiological cellular processes such as regulations of gene expression or growth signal transduction Their activation can occur through gene amplification or mutation such that more of the oncoprotein encoded by the gene is present; hence, its function is enhanced (Downward, 2003) Other mechanisms of oncogene activation include chromosomal