RICERCA ITALIANA     

Development and progression of hepatocellular carcinoma: molecular mechanisms and therapeutic implications.

Università degli Studi di Palermo

Abstract

Hepatocellular carcinoma (HCC) has been previously considered a rare tumor, but its incidence has considerably risen during the last years with alarming data for the mediterranean countries, including Italy. The actual understanding of the etiology and epidemiology of HCC let us forecast a further increase of HCC in Europe.
HCC prognosis remains unfavorable for the large portion of patients affected by advanced disease, which is deadly aggressive and does not respond to currently available therapies. This emphasizes the need of developing new strategies for the prevention and treatment of the disease; however, a major obstacle in this sense is represented by the incomplete definition of the molecular characteristics of HCC.
Our aim is to thoroughly examine in clinical HCC and in its laboratory models the expression and regulation of a series of factors and pathways, which emerge for their involvement in tumor cell growth and drug resistance, but whose importance is not well defined for HCC yet. These factors are:
1-The Inhibitory of Apotosis Proteins (IAPs) and their splice variants. The research is aimed to better characterize their expression pattern in HCC and in the predisposing pathologies, with relationship to the clinical, histological and prognostic data and to the etiology. We will examine the importance of different pathways (NF-kB, STATs, AP-1, PI3K/Akt, Cdk and aromatase) in their regulation, by treating with specific inhibitors HCC cells and evaluating their antitumor, apoptotic and molecular (e.g., levels of the IAPs and of other genes implicated in apoptosis and drug resistance) effects.
2-The cyclooxygenases (COX-1 and -2) and their products, the prostaglandins, with the relative receptors Eps. Also in this case, we will examine the expression pattern of these molecules in clinical HCC, with relationship to the status of other factors involved in cell invasiveness (E-cadherin), angiogenesis (VEGF), inflammation (TNF-alpha, IL-6), cell growth and apoptosis (like aromatase, estrogens or the IAPs themselves).
3-The DNp73 proteins, which are dominant negative, antiapoptotic and proproliferative variants, which repress p53-, p63- and p73-dependent transcription. We will investigate at molecular level the role of the DNp73 proteins in the activation of the hepatic progenitor cells and in the development of HCC, as well as the signals (NF-kB and beta-catenin) which drive the DNp73s. This study will be carried out in samples from patients with chronic hepatitis, cirrhosis and HCC, by ex vivo experiments and in cultured HCC cells. We will also examine the contribution of the DNp73s to drug resistance and the pharmacological strategies to overcome it.
Overall, the project has a descriptive aspect (corroborated by extensive cDNA microarray analyses) along the HCC trajectory, with clinical and prognostic implications and trying to achieve a better knowledge of the mechanisms responsible for the insurgence and progression of the disease. An important part of the study will be also that of ascertaining the therapeutic potential, also in terms of chemosensitization, of approaches able to interfere with the identified molecular targets. In particular, we will critically evaluate the existing evidences which support that in HCC are activated different cell proliferation and survival pathways, so that is not sufficient to block only one of them to have significant therapeutic results. The therapeutic results will be validated by experiments in mice transplanted with human HCC. <<<

Principal Investigator

Natale D'ALESSANDRO Università degli Studi di PALERMO

Research Objectives

Hepatocellular carcinoma (HCC) has been previously considered a rare tumor, but its incidence has considerably risen during the last years with alarming data for the mediterranean countries, including Italy. This likely reflects the stong impact that the ambiental factors have on the pathogenesis of the tumor. There is in fact a clear relationship between persistent infections fron hepatitis viruses (HBV and HCV) or, at least in some areas, chemical carcinogens and tumor formation. At present, HCC is one of the ten more frequent malignant disease in the world. The actual understanding of the natural history of HCV chronic infections (Pagliaro L. et al. Natural history of chronic hepatitis C. Ital G Gastroenterol Hepatol 31, 28-44, 1999) and the biological and epidemiological consequences of, either manifest or occult, co-infections from HBV in HCC chronic hepatitis let us forecast a further increase of HCC in Europe.
Nonetheless, in spite of the fact that HCC is a quite frequent disease, only a limited percentage of the patients with chronic viral hepatitis and cirrhosis undergo malignant transformation in a given time lap. This suggest the intervention of-co-oncogenic factors in the multistep process of in vivo transformation: one of the possibilities is that, to manifest the oncogenic potential, the viral proteins may require the activation of pro-proliferative and anti-apoptotic oncogenes or/and the inactivation of genes involved in the in vivo execution of apoptosis (Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 100: 57-70, 2000).
The clinical diagnostics and the treatment of the limited stages of HCC have considerably improved, but its prognosis remains highly unfavorable for the high percentage of patients with advanced disease at the first diagnosis or who relapse after "radical" locoregional interventions or resections. In fact, in these case HCC is deadly aggressivwe and does not respond to chemotherapy or radiotherapy. This emphasizes the need of developing new strategies for the prevention and treatment of the disease; however, a major obstacle in this sense is represented by the incomplete definition of the molecular characteristics of HCC.
The aim of this research project is to thoroughly examine along the trajectory of clinical HCC and in its laboratory models the expression and regulation of a series of factors and pathways (discussed further, in the Section "Scientific background"), which emerge for their involvement in tumor cell growth and drug resistance, but whose importance is not well defined for HCC yet. Thus, the project has a descriptive aspect with clinical and prognostic implications and tries to achieve a better knowledge of the mechanisms responsible for the insurgence and progression of HCC. An important part of the study will be also that of ascertaining the therapeutic potential, also in terms of chemosensitization, of interfering with the identified molecular targets. In particular, it will be considered the large evidence that in HCC are active different pathways of cell proliferation and survival together, so that is not sufficient to block only one of them to have significant therapeutic results. The therapeutic results will be validated by experiments in mice transplanted with human HCC.
The attainment of the proposed objectives will be facilitated by the fact that in this project are present and strongly interacting all the structures and expertises (in clinical medicine and pathology, molecular biology and pharmacology, with a previous experience in the field of HCC) necessary to its realization. <<<

Timescale

24 months

National and international background

IAP (INHIBITORY OF APOPTOSIS PROTEINS)
There is a large evidence that an imbalance between unrestrained cell proliferation and low ability to perform apoptosis, either spontaneous or induced from pharmacologic or immunologic agents, is a critical feature of hepatocellular carcinoma (HCC). A major factor which can determine such behavior is the re- or over-expression in HCC of inhibitors of apoptosis like the proteins of the Bcl-2 family and, as also shown by us, the IAPs (1-5).
Human IAPs include NAIP, XIAP, c-IAP-1, c-IAP-2, survivin, livin-alpha and their alternative splice variants (6, 7). They are endowed with a remarkable ability to block apoptosis induced by a wide spectrum of non-related triggers, both through inhibition of caspases and by other mechanisms (8, 9). IAPs exhibit other functions; for example, survivin is required for the execution of cell division and chromosomal segregation. Like the other IAPs, survivin in much more expressed in fetal and tumor tisues than in normal tissues. In tumors like melanoma, the circulating levels of survivin correlate to parameters such as aggressiveness of the disease and survival (10), but nothing is known about the possible correlation between circulating survivin and 1) the expression levels of survivin and the other IAPs in HCC; 2) the clinical-prognostic behavior of HCC; 3) the various stages preceding HCC development (chronic hepatitis, cirrhosis) and 4) the etiology (HCV, HBV, alcohol, others).
Overall, IAPs represent very attractive targets for new therapies against HCC, considering their marked sovraexpression in these tumors. Current attention focuses mainly on survivin only (2, 3, 11-13); it appears however that HCC may exhibit the abundant presence of different IAPs and of their alternative transcripts together (4, 5). As a consequence, it might be useful to develop approaches able to contrast the complex of these factors rather than only one of them.
The regulatory mechanisms of IAP expression are not completely defined yet. Their expression can be promoted by the transcription factor NF-kB, which, in turn, is frequently constitutively activated in HCC (14, 15). Thus, the inhibition of NF-kB might be of help to antagonize IAPs and other NF-kB target genes (e.g., Bcl-XL, c-myc, P-glycoprotein (P-gp) and COX-2) involved in this cancer. However, we have realized that NF-kB activation is not the sole mechanism which may account for the overexpression of the IAPs in HCC (16). Other transcription factors frequently involved in HCC are STATs (mainly STAT3) and AP-1 (15), but their links with IAPs have been scarcely investigated.
Another factor which is frequently activated in HCC and which promotes the survival and drug resistance of the cells by interacting with multiple pathways is the serine/threonine kinase Akt. Many growth stimuli can induce through PI3K/Akt the expression of survivin in different cells and determine their resistance to anticancer agents. Moreover, XIAP is involved in cell resistance to apoptosis induced from cisplatin and this, at least in part, may be due to its ability to activate Akt through PI3K (17-19).
Further, to exert its functions, survivin needs to be phosphorylated by cyclin-dependent kinases (Cdks). Survivin may support cell proliferation by interacting with Cdk4, in particular by releasing the inhibitor p21 from Cdk4 (33, 34). Inhibitors of the Cdks, such as flavopiridol, can induce apoptosis in different tumor cell lines and down-regulate the IAPs and other anti-apoptotic factors, including Bcl-2 (20-22). In preclinical studies flavopiridol has shown a promising antitumor activity, also in terms of synergy with radio- or chemo-therapy. It is currently investigated in the clinical setting. However, there is a lack of information on its possible therapeutic effects on HCC.

AROMATASE AND ESTROGENS
Studies of our group have shown that liver presents key enzymes in steroid hormone metabolism, including aromatase. Though the relationship between aromatase and different solid tumors, including breast cancer, has been strongly underlined, information on the involvement of this enzyme in HCC is scanty. However, it has been shown that non-tumor liver cells are characterized by an androgenic metabolism, with formation of biologically active androgens. On the other hand, transformed liver cells (HepG2) show a metabolic pattern characterized by high levels of estrogen production. Increased levels of expression and activity of the aromatase enzyme have been shown in human liver tumor cells and tissues, by RT-PC and enzymatic analyses on “intact cells” (23). A novel aspect to study regards the possible relationship between local estrogen production and IAP sovraexpression in HCC, a relationship whose existence has been already shown in breast cancer (24).

THE CYCLOOXYGENASES (COX-1 AND COX-2)
It is now well established that the inducible form COX-2 is involved in the development and progression different tumors, including colon, skin and breast cancers (2-4).
Recent evidence suggests that also COX-2 is involved in some tumor types (25-26), but in general the role of the COXs in hepatic carcinogenesis is not clear yet. Some studies showing an increased expression of the two isoforms in various types of liver disease, have suggested their possible role in HCC. Results obtained in our and other laboratories have indicated that selective COX-2 inhibitors are able to inhibit the growth of HCC cells in vitro, while no experimental data on COX-1 are available to date (27-29).
In tumors, over-expression of COX-2 leads to an increase in prostaglandin levels (PGs); PGE2 is the more abundant PG in tumors and acts by binding to the EP receptors (EP1, EP2, EP3, EP4). Recent studies have attempted to clarify the role of PGE2 and of EP receptors in carcinogenesis (30). In particular, EP2 or EP4 promote cell growth via the beta-catenin system or the transactivation of the EGF receptor (31, 32), while receptor EP3 would inhibit cell growth (33). Little is known about EP receptor expression in liver diseases and during HCC progression, as well as about the possible use of EP receptor antagonists and/or agonists as anti-tumor agents.
PGE2 enhances the expression of the gene codifying for aromatase. The regulation of aromatase activity is complex and depends on a number of factors, including interleukin-6 (IL-6) and its soluble receptor (sIL-6R) and TNF-alpha; however, it seems plausible that the COX/PGE2 system may contribute to HCC development also via the local regulation of estrogen synthesis (34).
COX-2 may prevent apoptosis through PGs-mediated induction of the Bcl-2 gene and by removing arachidonic acid, which is pro-apototic and activates the caspases. COX-2 may control apoptosis also through the over-expression of some IAPs, such as c-IAP2 and survivin (35).
HCC is very resistant to anticancer drugs, also because of the over-expression of P-glycoprotein. Recently, multidrug resistance phenotype and P-gp over-expression have been associated to COX-2 in liver cancer cells (36).
The rationale of combining conventional chemotherapy with selective COX-2 inhibitors is further supported by the fact that chemotherapy itself induces COX-2 expression (37).
The finding that the over-expression of COX-2 is associated to alterations of cell mobility and cellular adhesion suggests that COX-2 could also play a role in the control of cell invasiveness (38). Changes associated to these alterations include an increased expression and activation of membrane metalloproteinases and a reduced expression of E-cadherin. In general, the loss of expression of E-cadherin correlates with an increased invasive capacity of cells. Therefore, E-cadherin expression is considered a suppressor of the invasive potential of cancer cells. E-cadherin is under the control of sex steroids, at least in breast (39) and prostate (40) cancers, but this aspect needs to be evaluated in HCC.
HCC is a highly vascularized tumor and COX-2 has been implicated also in this aspect of carcinogenesis (31). High levels of COX-2 expression in colon carcinoma cells have been reported to increase the production of VEGF, endothelial cell migration and the formation of a capillary network (41). The mechanism by which COX-2 may affect angiogenesis in HCC remain to be clarified.

P53, P63, P73 AND REGULATORY MECHANISMS
The p53 tumor suppressor gene is the most frequent site of genetic alterations found in human cancers. Specific mutations at codon 249 of the p53 tumor suppressor gene are found in up to 50% of the HCCs in populations exposed to aflatoxin B1. Among non-exposed patients, however, the prevalence of the p53 gene mutations is much lower (<20%) and codon 249 is exceptionally altered (42). Recently, two genes, referred to as p63 and p73, have been found to encode a number of proteins that share similar, but not identical, structures and functions with p53 (43, 44). P73 and p63, are expressed as multiple splice variants at their COOH-terminus (45, 46). In addition, dominant negative variants, that lack the amino-terminal transactivation domain and function as trans-repressors of p53 as well as p63- and p73-dependent transcription, are encoded from the classical P1p73 promopter (DTAp73 proteins), as well as from a second intragenic promoter (DNp73 proteins). Both p73 and p63 cause cell cycle arrest through the transactivation of p21, induce apoptosis and play an important role in mediating both p53-dependent and p53-independent apoptosis in response to DNA damage and anti-cancer drugs (47, 49).
Mutations in the p73 gene have not been found in HCCs. However, p73 is expressed, the p73 proteins accumulate and this accumulation correlates with poor prognosis in HCCs (43, 44). By real-time RT-PCR we found that TAp73 levels increase from normal liver to chronic hepatitis, cirrhosis and HCC, being TAp73s always higher in tumor vs non-tumor tissue from the same patient and DTAp73 proteins present only in HCC tissue. DNp73 is largely expressed and DNp73 mRNA levels increase in chronic hepatitis, cirrhosis and HCC and always exceed the amount of TAp73 mRNAs (paper submitted). Thus, increased expression of DNp73 and DTAp73 proteins that trans-represses wild type p53 and Tp73 functions is actively selected during liver tumor development. The signals and the molecular pathways that upregulate DNp73 expression in normal and transformed hepatocytes are far from being elucidated. Using small interfering RNAs specific for DNp73 to selectively abrogate its function, we found that DNp73 is required for TNF-alpha induced proliferation of hepatocytes and that, in this context DNp73 expression is controlled by NF-kB (paper submitted). However, the DNp73 promoter is also regulated in different pathological conditions by p53 (50), c-jun (51) and, as we have very recently found, by the beta-catenin pathway (unpublished observations). One may remember here that wnt/beta-catenin pathway is often altered in human HCC. One beta-catenin allele (3p21) has been found mutated in about 25% of adult hepatocellular carcinomas and 48% of childhood hepatoblastomas. Abnormal nuclear accumulation of beta-catenin protein is even more frequent (over 75 % of the HCCs analyzed in one study), due either to mutations in axin or to methylation of the E-cadherin gene and loss of E-cadherin protein expression. Interestingly, accumulation of p53 (52) and p73 (unpublished observations) correlates with nuclear accumulation of beta-catenin. Irrespectively of the underlying mechanism, unregulated nuclear beta-catenin associated with Tcf/LEF causes transcriptional activation of target genes, including c-myc, cyclin D1, fibronectin, matrix metalloproteinase genes and survivin, keeping the cells proliferating.
DNp73 proteins are also targeted by a number of post-translational modifications that regulate their stability and ultimately cooperate with the above mentioned transcriptional regulatory mechanisms to determine the intracellular levels of DNp73 proteins.
In addition the its relevance in the molecular pathogenesis of HCC, DNp73 accumulation may also have important consequences in the therapeutic management of this tumor, since the ratio between TAp73 and DNp73 protein levels has emerged as an important determinant of cell sensitivity to DNA damaging drugs (44-46, 53, 54). We have recently found that the pharmacological modulation of TAp73 and DNp73 protein levels and activity may have a profound impact on cell survival and chemosensitivity (55). <<<