FIELD OF THE INVENTION The invention is in the field of diagnostics and therapeutics for cancer. More specifically, the invention is in the field of diagnostics and therapeutics for hepatocellular carcinoma.

BACKGROUND OF THE INVENTION Hepatocellular carcinoma (HCC) is the most common primary malignant tumor of the liver that accounts for more than 70% of liver cancers worldwide (Parkin et al. , 1999). Many risk factors have been associated with the development of HCC, including hepatitis B (HBV) and hepatitis C (HCV) viral infection, cirrhosis, male gender, exposure to toxins, etc. Death generally occurs due to liver failure associated with cirrhosis and/or rapid outgrowth of multiple nodules. Approximately 0.25-1 million new cases of HCC are diagnosed each year, and the cancer is especially prevalent in Southeast Asia, China, and sub-Saharan Africa. While surgical resection is considered to be the main curative treatment, only 10-15% of cases are suitable for surgery at the time of presentation. This is because either the disease is detected at an advanced stage at presentation or the underlying poor liver functional reserve precluded surgical intervention.

Diagnosis of HCC has included detection of the presence of a liver mass on radiological investigations and the detection of elevated serum alpha fetoprotein (AFP) levels (Yu and Keeffe, 2003). However, elevation of AFP is not exclusive to HCC and has been observed in benign hepatic disease, such as liver cirrhosis, and other cancers such as germ cell cancer (Bosl and Head, 1994). Treatment of HCC has included interferon therapy and antiviral drugs, but the results have proved

unpredictable and the effectiveness maybe limited (Lee, 1997; Yu and Keeffe, 2003).

Microarrays have been used to address changes in gene expression of HCC (Chen et al, 2002, Okabe et al, 2001; Honda et al, 2001 ; Shirota et al, 2001; Tackels-Horne et al, 2001; Xu et al, 2001a; Xu et al, 2001b). However, these reports were restricted to the tissue samples selected for each study and exhibited wide variation in the results, thus limiting the potential significance and utility of the data.

SUMMARY OF THE INVENTION The invention provides in part molecular markers for hepatocellular carcinoma (HCC) that may be used for HCC diagnosis, to assess HCC progression or regression, or the efficacy and/or toxicity of HCC therapeutics, and/or to identify candidate compounds for HCC therapy, with high predictive accuracy.

In one aspect, the invention provides a composition including an addressable collection of two or more nucleic acid molecules, or polypeptides encoded by these nucleic acid molecules, that are differentially expressed in hepatocellular carcinoma, where the nucleic acid molecules consist essentially of the nucleic acid molecules set forth in any one or more of Tables 1 through 4 or complements, fragments, variants, or analogs thereof. The composition may include all of the nucleic acid molecules, or their encoded polypeptides, set forth in any one or more of Tables 1 through 4 or complements, fragments, variants, or analogs thereof, or any subset of these nucleic acid molecules or polypeptides. The nucleic acid molecules or polypeptides may be differentially expressed between hepatocellular carcinoma tissue and non-tumor tissue. The nucleic acid molecules or the polypeptides may be attached to a solid support. The compositions may be used in the preparation of a medicament for diagnosis or therapy of hepatocellular carcinoma.

In other aspects, the invention provides a method of diagnosing hepatocellular carcinoma in a subject by obtaining a sample from the subject and detecting the level of expression of two or more nucleic acid molecules or expression products thereof in the sample, where the nucleic acid molecules consist essentially of the nucleic acid molecules set forth in any one or more of Tables 1 through 4 or complements, fragments, variants, or analogs thereof.

In other aspects, the invention provides a method of monitoring the progression of hepatocellular carcinoma in a subject by obtaining a sample from the subject and detecting the level of expression of two or more nucleic acid molecules or expression products thereof in the sample, where the nucleic acid molecules consist essentially of the nucleic acid molecules set forth in any one or more of Tables 1 through 4, or complements, fragments, variants, or analogs thereof. The sample may be obtained at two or more time points. The method may further include comparing the level of expression of the nucleic acid molecules or expression products at two or more time points.

In other aspects the invention provides a method of monitoring the efficacy of a hepatocellular carcinoma therapy in a subject by administering the therapy to the subject, obtaining a sample from the subject, and detecting the level of expression of two or more nucleic acid molecules or expression products thereof in the sample, where the nucleic acid molecules consist essentially of the nucleic acid molecules set forth in any one or more of Tables 1 through 4, or complements, fragments, variants, or analogs thereof. The therapy may be administered at two or more administration time points. The sample may be obtained at two or more sampling time points. The method may further include comparing the level of expression of the nucleic acid molecules or expression products at two or more administration time points, and/or at two or more sampling time points.

In other aspects, the invention provides a method of screening a compound for treating hepatocellular carcinoma by contacting a sample with a test compound and detecting the level of expression of two or more nucleic acid molecules or expression products thereof in the sample, where the nucleic acid molecules consist essentially of the nucleic acid molecules set forth in any one or more of Tables 1 through 4, or complements, fragments, variants or analogs thereof.

In alternate embodiments of the various aspects, the sample may be liver or serum, or may be suspected of being cancerous, or may be non-cancerous. The methods may further include comparing the level of expression of the nucleic acid molecules or expression products thereof in a non-cancerous sample and in a sample suspected of being cancerous. Differential expression of the nucleic acid molecules or expression products thereof may be indicative of hepatocellular carcinoma, or of

progression of hepatocellular carcinoma, or of the efficacy of the hepatocellular carcinoma therapy. The subject may be suspected of having hepatocellular carcinoma. The subject may be a human.

In alternate embodiments of the various aspects, the method may further include comparing the level of expression of two or more nucleic acid molecules or expression products thereof with a standard, or further include preparing a gene expression profile. The method may be a high throughput method.

In other aspects, the invention provides a solid support including two or more nucleic acid molecules or polypeptides encoded by these nucleic acid molecules that are differentially expressed in hepatocellular carcinoma, where the nucleic acid molecules consist essentially of the nucleic acid molecules set forth in Tables 1 through 4 or complements, fragments, variants, or analogs thereof. The nucleic acid molecules may consist essentially of all the nucleic acid molecules set forth in any one or more of Tables 1 through 4, and/or be differentially expressed between hepatocellular carcinoma tissue and non-tumor tissue. The polypeptides may consist essentially of the polypeptides encoded by all the nucleic acid molecules set forth in any one or more of Tables 1 through 4, and/or be differentially expressed between hepatocellular carcinoma tissue and non-tumor tissue. The nucleic acid molecules or the polypeptides may be covalently or non-covalently attached to the solid support (e. g. , a microarray).

In other aspects, the invention provides a database including information identifying the expression level in liver tissue (e. g., cancerous or non-cancerous tissue) of two or more nucleic acid molecules or expression products thereof, where the nucleic acid molecules consist essentially of the nucleic acid molecules set forth in any one or more of Tables 1 through 4, or complements, fragments, variants, or analogs thereof.

A"composition"as used herein includes a plurality of the nucleic acid molecules described herein, including complements, analogs, variants, and fragments thereof. A composition as used herein also includes a plurality of polypeptides encoded by the nucleic acid molecules described herein, and complements, analogs, variants, and fragments thereof. A composition as used herein also includes a plurality of polypeptides capable of specifically binding to the polypeptides or nucleic acid

molecules described herein (e. g. , antibodies). The composition may include any combination of the nucleic acid molecules described herein, including complements, analogs, variants, and fragments thereof, or polypeptides encoded by these nucleic acid molecules. Accordingly, the composition may include 2,3, 4,5, 6,7, 8, 9,10, etc. up to all of the nucleic acid molecules or polypeptides described herein, e. g. , in any one or more of the Tables or Figures herein. In some embodiments, the composition may include subsets of the nucleic acid molecules or polypeptides described herein, e. g. , in any one or more of the Tables or Figures herein, for example, subsets groups by protein function or characteristics, e. g. , proteins involved in the ubiquitination pathway, or proteins localized to a particular cellular compartment. These nucleic acid molecules or polypeptides may for example be used with a substrate (e. g, a solid substrate or a liquid substrate) in a variety of applications, including the diagnosis of HCC, or monitoring the progression of HCC.

By"addressable collection"is meant a combination of nucleic acid molecules or polypeptides capable of being detected by, for example, the use of hybridization techniques or antibody binding techniques or by any other means of detection known to those of ordinary skill in the art.

The terms"nucleic acid"or"nucleic acid molecule"encompass both RNA (plus and minus strands) and DNA, including cDNA, genomic DNA, and synthetic (e. g. , chemically synthesized) DNA. The nucleic acid may be double-stranded or single-stranded. Where single-stranded, the nucleic acid may be the sense strand or the antisense strand. A nucleic acid molecule may be any chain of two or more covalently bonded nucleotides, including naturally occurring or non-naturally occurring nucleotides, or nucleotide analogs or derivatives. By"RNA"is meant a sequence of two or more covalently bonded, naturally occurring or modified ribonucleotides. One example of a modified RNA included within this term is phosphorothioate RNA. By"DNA"is meant a sequence of two or more covalently bonded, naturally occurring or modified deoxyribonucleotides. By"cDNA"is meant complementary or copy DNA produced from an RNA template by the action of RNA- dependent DNA polymerase (reverse transcriptase). Thus a"cDNA clone"means a duplex DNA sequence complementary to an RNA molecule of interest, carried in a cloning vector. An"oligonucleotide"as used herein is a single stranded molecule

which may be used in hybridization or amplification technologies. In general, an oligonucleotide may be any integer from about 15 to about 100 nucleotides in length, but may also be of greater length. A"probe"or"primer"is a single-stranded DNA or RNA molecule of defined sequence that can base pair to a second DNA or RNA molecule that contains a complementary sequence (the target). The stability of the resulting hybrid molecule depends upon the extent of the base pairing that occurs, and is affected by parameters such as the degree of complementarity between the probe and target molecule, and the degree of stringency of the hybridization conditions. The degree of hybridization stringency is affected by parameters such as the temperature, salt concentration, and concentration of organic molecules, such as formamide, and is determined by methods that are known to those skilled in the art. Probes or primers specific for the nucleic acid sequences described herein, or portions thereof, may vary in length by any integer from at least 8 nucleotides to over 500 nucleotides, including any value in between, depending on the purpose for which, and conditions under which, the probe or primer is used. For example, a probe or primer may be 8,10, 15, 20, or 25 nucleotides in length, or may be at least 30,40, 50, or 60 nucleotides in length, or may be over 100,200, 500, or 1000 nucleotides in length. Probes or primers specific for the nucleic acid molecules described herein may have greater than any integer between 20-30% sequence identity, or at least any integer between 55- 75% sequence identity, or at least any integer between 75-85% sequence identity, or at least any integer between 85-99% sequence identity, or 100% sequence identity to the nucleic acid sequences described herein. Probes or primers can be detectably- labeled, either radioactively or non-radioactively, by methods that are known to those skilled in the art. Probes or primers can be used for methods involving nucleic acid hybridization, such as nucleic acid sequencing, nucleic acid amplification by the polymerase chain reaction, single stranded conformational polymorphism (SSCP) analysis, restriction fragment polymorphism (RFLP) analysis, Southern hybridization, northern hybridization, in situ hybridization, electrophoretic mobility shift assay (EMSA), microarray, and other methods that are known to those skilled in the art.

Probes or primers may be derived from genomic DNA or cDNA, for example, by amplification, or from cloned DNA segments, or may be chemically synthesized.

The"expression product"of a nucleic acid molecule may be any polypeptide encoded by that nucleic acid molecule. Generally, the polypeptide is capable of being expressed.

A"protein,""peptide"or"polypeptide"is any chain of two or more amino acids, including naturally occurring or non-naturally occurring amino acids or amino acid analogues, regardless of post-translational modification (e. g. , glycosylation or phosphorylation). An"amino acid sequence","polypeptide","peptide"or"protein"of the invention may include peptides or proteins that have abnormal linkages, cross links and end caps, non-peptidyl bonds or alternative modifying groups. Such modified peptides are also within the scope of the invention. The term"modifying group"is intended to include structures that are directly attached to the peptidic structure (e. g. , by covalent coupling), as well as those that are indirectly attached to the peptidic structure (e. g. , by a stable non-covalent association or by covalent coupling to additional amino acid residues, or mimetics, analogues or derivatives thereof, which may flank the core peptidic structure). For example, the modifying group can be coupled to the amino-terminus or carboxy-terminus of a peptidic structure, or to a peptidic or peptidomimetic region flanking the core domain.

Alternatively, the modifying group can be coupled to a side chain of at least one amino acid residue of a peptidic structure, or to a peptidic or peptido-mimetic region flanking the core domain (e. g. , through the epsilon amino group of a lysyl residue (s), through the carboxyl group of an aspartic acid residue (s) or a glutamic acid residue (s), through a hydroxy group of a tyrosyl residue (s), a serine residue (s) or a threonine residue (s) or other suitable reactive group on an amino acid side chain). Modifying groups covalently coupled to the peptidic structure can be attached by means and using methods well known in the art for linking chemical structures, including, for example, amide, alkylamino, carbamate or urea bonds. Peptides according to the invention may include peptides encoded by the nucleic acid molecules of Tables 1 through 4 or complements or analogs thereof.

By"differential expression"or"differentially expressed"is meant increased, upregulated or present, or decreased, downregulated or absent, gene expression as detected by the absence, presence, or change (up or down) in the amount of transcribed messenger RNA or translated protein in a sample. For example, the

change may be detected by comparison of the change in gene expression level between a HCC sample and a non-tumor sample. The absolute amount of change of gene expression is not important, as long as the amount of change is reproducible, and measurable. In some embodiments, the change (up or down) in the amount of transcribed messenger or translated protein may be at least 1-fold or at least 1. 5-fold or may be over 2.0, 2.5., 3.0, 3.5, 4.0, 4.5, or 5. 0-fold. In some embodiments, the change in the amount of transcribed messenger or translated protein may be 40%, 50%, 60%, 70%, 80%, 90%, or 100%.

By"detecting"it is intended to include determining the presence or absence, or quantifying the amount, of a nucleic acid molecule or polypeptide of the invention a substance. The term thus refers to the use of the materials, compositions, and methods of the present invention for qualitative and quantitative determinations. For example, detecting an increase in gene expression levels may include quantifying a change of any value between 10% and 90%, or of any value between 30% and 60%, or over 100%, of any of the nucleic acid molecules or polypeptides of the invention when compared to a control. In other embodiments, detecting an increase in gene expression levels may include quantifying a change of any value between 1 to 5 fold or more of any of the nucleic acid molecules or polypeptides of the invention when compared to a control.

"Hepatocellular carcinoma"is cancer that arises from hepatocytes, the major cell type of the liver. It is a form of adenocarcinoma, and is the most common type of liver tumor."Non-tumor"tissue refers to tissue or cells that are non-cancerous. In some embodiments, non-tumor tissue may include tissue or cells from a subject having a liver disorder, such as HBV or HCV infection, cirrhosis, exposure to aflatoxins, etc. The phrase"suspected of being cancerous"as used herein means a HCC tissue sample believed by one of ordinary skill in the art to contain HCC cells.

By"non-cancerous"or"non-tumor"is meant a tissue sample demonstrated by standard diagnostic or other techniques (e. g. , histologic staining, microscopic analysis, immunoassay, etc. ) to contain no HCC cells or evidence of HCC.

A"sample"can be any organ, tissue, cell, or cell extract isolated from a subject, such as a sample isolated from a mammal having a hepatocellular carcinoma or isolated from a mammal not having a hepatocellular carcinoma or a tumor. For

example, a sample can include, without limitation, tissue such as liver tissue (e. g., from a biopsy or autopsy), cells, peripheral blood, whole blood, red cell concentrates, platelet concentrates, leukocyte concentrates, blood cell proteins, blood plasma, platelet-rich plasma, a plasma concentrate, a precipitate from any fractionation of the plasma, a supernatant from any fractionation of the plasma, blood plasma protein fractions, purified or partially purified blood proteins or other components, serum, semen, mammalian colostrum, milk, urine, stool, saliva, placental extracts, amniotic fluid, a cryoprecipitate, a cryosupernatant, a cell lysate, mammalian cell culture or culture medium, products of fermentation, ascitic fluid, proteins present in blood cells, solid tumours isolated from a mammal with a hepatocellular carcinoma, or any other specimen, or any extract thereof, obtained from a patient (human or animal), test subject, or experimental animal. A sample may also include, without limitation, products produced in cell culture by normal, non-tumor, or transformed cells (e. g. , via recombinant DNA technology). A"sample"may also be a cell or cell line created under experimental conditions, that are not directly isolated from a subject. A sample can also be cell-free, artificially derived or synthesised. In some embodiments, samples refer to liver tissue or cells. In some embodiments, the liver tissue may be from a subject having a hepatocellular carcinoma; a subject infected with a hepatitis virus; a subject having a liver disorder e. g. , cirrhosis, or a subject having a normal liver e. g. , not diagnosed with or suspected of having a liver disorder.

As used herein, a subject may be a human, non-human primate, rat, mouse, cow, horse, pig, sheep, goat, dog, cat, etc. The subject may be a clinical patient, a clinical trial volunteer, an experimental animal, etc. The subject may be suspected of having HCC, be diagnosed with HCC, or be a control subject that is confirmed to not have HCC. Diagnostic methods for HCC and the clinical delineation of HCC diagnoses are known to those of ordinary skill in the art, and include biopsy including radiological biopsy by means of a radiological scan, laproscopy, or open surgical biopsy.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a plot showing natural patterns of gene expression differences between HCC tumor and non-tumor liver tissue specimens based on unsupervised

clustering. The plot shows the variance of expression value for each of the gene features across all the HCC tumor and non-tumor liver tissue specimens. The dotted line indicates the 500 most variable gene features.

Figure 2 is a multidimensional scaling plot showing significant gene differential expression between HCC tumor and non-tumor liver tissues, and comparison with liver cancer cell lines (P<lx10-6, approximately 1. 5-fold change).

The plot illustrates the ability of the 218 outlier genes to separate HCC tumor specimens (black circles) from non-tumor liver tissue specimens (dark gray circles).

The plot also shows how different liver cancer cell lines (light gray circles) are from the clinical tissue samples.

Figures 3A-B characterize differentially expressed genes in HCC tumor specimens (P<lx10-6, approximately 1. 5-fold change). Figure 3A is a bar graph showing the chromosomal distribution of the 218 outlier genes. The dark colored and light shaded bars represent genes that are at least 1. 5-fold up-and downregulated, respectively, in HCC tumors relative to non-tumor livers. Figure 3B is a bar graph showing the functional characterization of the outlier genes based on Gene Ontology and published works.

Figure 4 is a bar graph showing the expression of BMI-1 in HCC tumors as determined by cDNA microarray analysis. The data are presented as the level of expression (log base 2) in each HCC tumor specimen with respect to the corresponding non-tumor liver sample.

Figures 5A-D show real-time RT-PCR analysis of IGFBP3, ERBB3, ERBB2 and EGFR in HCC tumor samples. The gene expression patterns for (A) IGFBP3, (B) ERBB3, (C) ERBB2 and (D) EGFR in all the 37 HCC tumor samples and their corresponding non-tumor liver tissue specimens were examined. All data was normalized to the amount of'housekeeping'gene PBGD and are presented as relative fold expression change (log base 2 ratio) in HCC tumor specimens with respect to their corresponding non-tumor liver counterpart. Positive value depicts higher expression level, while negative value depicts lower expression level in the tumor relative to the non-tumor specimen.

Figure 6 lists a panel of genes analyzed using real-time and semi-quantitative RT-PCR analyses, and indicating whether the analysis was conducted in non-tumor human tissues, and in clinical tissue samples or HCC cell lines or both.

Figures 7A-C show gene expression analysis of ARMET. Semi-quantitative RT-PCR analysis (A) of non-tumor tissues and HCC cell lines and real time RT-PCR analysis of non-tumor tissues (B) and patient samples (C) was performed. GAPDH expression level was used as the control in the analyses of HCC cell lines vs. non- tumor human tissues (A). The dotted line in (B) indicates mean expression value of four non-tumor liver tissues i. e. , Fetal/F, Fetal/M, Adult/F, Adult/M.

Figures 8A-C show gene expression analysis of BMI-1. Semi-quantitative RT-PCR analysis (A) of non-tumor tissues and HCC cell lines and real time RT-PCR analysis of non-tumor tissues (B) and patient samples (C) was performed. GAPDH expression level was used as the control in the analyses of HCC cell lines vs. non- tumor human tissues (A). The dotted line in (B) indicates mean expression value of four non-tumor liver tissues i. e., Fetal/F, Fetal/M, Adult/F, Adult/M.

Figures 9A-C show gene expression analysis of CRHBP. Semi-quantitative RT-PCR analysis (A) of non-tumor tissues and HCC cell lines and real time RT-PCR analysis of non-tumor tissues (B) and patient samples (C) was performed. GAPDH expression level was used as the control in the analyses of HCC cell lines vs. non- tumor human tissues (A). The dotted line in (B) indicates mean expression value of four non-tumor liver tissues i. e. , Fetal/F, Fetal/M, Adult/F, Adult/M.

Figures 10A-C show gene expression analysis of CSTB. Semi-quantitative RT-PCR analysis (A) of non-tumor tissues and HCC cell lines and real time RT-PCR analysis of non-tumor tissues (B) and patient samples (C) was performed. GAPDH expression level was used as the control in the analyses of HCC cell lines vs. non- tumor human tissues (A). The dotted line in (B) indicates mean expression value of four non-tumor liver tissues i. e., Fetal/F, Fetal/M, Adult/F, Adult/M.

Figures 11A-C show gene expression analysis of DPT. Semi-quantitative RT-PCR analysis (A) of non-tumor tissues and HCC cell lines and real time RT-PCR analysis of non-tumor tissues (B) and patient samples (C) was performed. GAPDH expression level was used as the control in the analyses of HCC cell lines vs. non-

tumor human tissues (A). The dotted line in (B) indicates mean expression value of four non-tumor liver tissues i. e., Fetal/F, Fetal/M, Adult/F, Adult/M.

Figures 12A-B show gene expression analysis of ERBB3. Real time RT-PCR analysis of non-tumor tissues (A) and patient samples (B) was performed. The dotted line in (A) indicates mean expression value of four non-tumor liver tissues i. e., Fetal/F, Fetal/M, Adult/F, Adult/M.

Figures 13A-B show gene expression analysis of EZH2. Real time RT-PCR analysis of non-tumor tissues (A) and patient samples (B) was performed. The dotted line in (A) indicates mean expression value of four non-tumor liver tissues i. e., Fetal/F, Fetal/M, Adult/F, Adult/M.

Figures 14A-B show gene expression analysis of GPC3. Real time RT-PCR analysis of non-tumor tissues (A) and patient samples (B) was performed. The dotted line in (A) indicates mean expression value of four non-tumor liver tissues i. e., Fetal/F, Fetal/M, Adult/F, Adult/M.

Figures 15A-B show gene expression analysis of HDGF. Real time RT-PCR analysis of non-tumor tissues (A) and patient samples (B) was performed. The dotted line in (A) indicates mean expression value of four non-tumor liver tissues i. e., Fetal/F, Fetal/M, Adult/F, Adult/M.

Figures 16A-B show gene expression analysis of MDK. Real time RT-PCR analysis of non-tumor tissues (A) and patient samples (B) was performed. The dotted line in (A) indicates mean expression value of four non-tumor liver tissues i. e., Fetal/F, Fetal/M, Adult/F, Adult/M.

Figure 17 shows gene expression analysis of D123. Semi-quantitative RT- PCR analysis of non-tumor tissues and HCC cell lines was performed. GAPDH expression level was used as the control in the analyses of HCC cell lines vs. non- tumor human tissues.

Figure 18 shows gene expression analysis of FLJ10326. Semi-quantitative RT-PCR analysis of non-tumor tissues and HCC cell lines was performed. GAPDH expression level was used as the control in the analyses of HCC cell lines vs. non- tumor human tissues.

Figure 19 shows gene expression analysis of ICA-1A. Semi-quantitative RT- PCR analysis of non-tumor tissues and HCC cell lines was performed. GAPDH

expression level was used as the control in the analyses of HCC cell lines vs. non- tumor human tissues.

Figure 20 shows gene expression analysis of LASP1. Semi-quantitative RT- PCR analysis of non-tumor tissues and HCC cell lines was performed. GAPDH expression level was used as the control in the analyses of HCC cell lines vs. non- tumor human tissues.

Figure 21 shows gene expression analysis of PODXL. Semi-quantitative RT- PCR analysis of non-tumor tissues and HCC cell lines was performed. GAPDH expression level was used as the control in the analyses of HCC cell lines vs. non- tumor human tissues.

DETAILED DESCRIPTION OF THE INVENTION Phenotypic changes in cancer may be due to cellular changes at the nucleotide level. Thus, some genes may be expressed, overexpressed, or under-expressed in tumor cells relative to non-tumor cells. However, a wide variation exists in gene expression patterns among cancer patients, including HCC patients. Therefore, examining the regulation or expression of a single gene or target, or even of multiple genes or targets whose regulation or expression vary across different HCC tumors, may be insufficient for accurate diagnosis or treatment of HCC or for screening of HCC therapeutics. Selecting a set of differentially expressed HCC genes, nucleic acid molecules, and/or polypeptides, assists in predictable and accurate diagnosis and therapy, and design of efficacious therapeutics.

The invention provides, in part, nucleic acid molecules and polypeptides that are differentially expressed in HCC cells, when compared to non-HCC tissue, e. g., liver or serum. Thus, the invention provides, in part, molecular markers for HCC derived from the analysis of global changes in gene expression ("gene expression profiles") between HCC tissue and non-HCC tissue. More specifically, cDNA microarrays were used to examine the global cellular changes in matched pairs of HCC tumor and non-tumor tissues of patients diagnosed with HCC. In addition, gene expression patterns between primary HCC tumors and liver cancer cell lines were examined for possible biological variation.

The nucleic acid molecules or polypeptides provided by the invention, as well as subsets thereof, serve as molecular markers that may be used for example for HCC diagnosis; to assess HCC progression or regression; to access the efficacy and/or toxicity of HCC therapeutics; and/or to identify candidate compounds for HCC therapy, with high predictive accuracy. The genes lists identified permit rapid, simple, and reproducible screening of a variety of HCC samples by, for example, nucleic acid microarray hybridization or protein expression technology to determine the expression of the specific genes, or by other means such as differential display, gel electrophoresis, genome mismatch scanning, representational discriminate analysis, clustering, transcript imaging, etc. used singly or in combination. Thus, the selected nucleic acid molecules or polypeptides of the invention define standard and reproducible differential expression patterns against which to compare the expression pattern in a variety of tissue or cells, e. g. , HCC tissue or cells or serum, obtained by biopsy, autopsy, or from cell lines and/or in vitro treatment or assays. The selected nucleic acid molecules or polypeptides of the invention and subsets thereof provide reliable detection of HCC cells or tissue, with reduction or elimination of false positives or false negatives. In some embodiments, the invention provides composite sets of discriminator genes for use as general or global HCC tumor markers. In some embodiments, the nucleic acid molecules or polypeptides of the invention may be used to assess the suitability of a HCC cell line for use as a model for HCC, as gene expression profiles may vary between primary HCC tumors and HCC cell lines.

Various alternative embodiments and examples of the invention are described herein. These embodiments and examples are illustrative and should not be construed as limiting the scope of the invention.

Nucleic Acid Molecules, Polypeptides, And Test Compounds Compounds according to the invention include, without limitation, molecules substantially identical to the nucleic acid molecules of Tables 1 through 4 (e. g. , BMI- 1, ARMET, CRHBP, CSTB, DPT, ERBB3, EZH2, GPC3, HDGF, MDK, D123, FLJ10326, ICAP-1A, LASP1, PODXL) and complements, analogs, fragments, and variants thereof, including, for example, the polypeptides described herein that are encoded by the nucleic acid molecules of Tables 1 through 4, as well as homologs and

fragments thereof. In some embodiments of the invention, compounds of the invention include antibodies that specifically bind to polypeptides encoded by the nucleic acid molecules of Tables 1 through 4. An antibody"specifically binds"an antigen when it recognises and binds the antigen, for example, a polypeptide encoded by any of the nucleic acid molecules described herein, but does not substantially recognise and bind other reference molecules in a sample, for example, a polypeptide that is encoded by a nucleic acid molecule that is not substantially identical to any of the nucleic acid molecules described herein. Such an antibody has, for example, an affinity for the antigen which is at least 10,100, 1000 or 10000 times greater than the affinity of the antibody for another reference molecule in a sample.

A"substantially identical"sequence is an amino acid or nucleotide sequence that differs from a reference sequence only by one or more conservative substitutions, as discussed herein, or by one or more non-conservative substitutions, deletions, or insertions located at positions of the sequence that do not destroy the biological function of the amino acid or nucleic acid molecule, or that do not destroy the detectability (e. g. , by hybridization or specific binding) of the amino acid or nucleic acid molecule. Such a sequence can be any integer from 10% to 99%, or more generally at least 10%, 20%, 30%, 40%, 50,55% or 60%, or at least 65%, 75%, 80%, 85%, 90%, or 95%, or as much as 96%, 97%, 98%, or 99% identical when optimally aligned at the amino acid or nucleotide level to the sequence used for comparison using, for example, the Align Program (Myers and Miller, CABIOS, 1989,4 : 11-17) or FASTA. For polypeptides, the length of comparison sequences may be at least 2, 5,10, or 15 amino acids, or at least 20,25, or 30 amino acids. In alternate embodiments, the length of comparison sequences may be at least 35,40, or 50 amino acids, or over 60,80, or 100 amino acids. For nucleic acid molecules, the length of comparison sequences maybe at least 5,10, 15,20, or 25 nucleotides, or at least 30, 40, or 50 nucleotides. In alternate embodiments, the length of comparison sequences may be at least 60,70, 80, or 90 nucleotides, or over 100,200, or 500 nucleotides.

Sequence identity can be readily measured using publicly available sequence analysis software (e. g., Sequence Analysis Software Package of the Genetics Computer Group, University of Wisconsin Biotechnology Center, 1710 University Avenue, Madison, Wis. 53705, or BLAST software available from the National Library of Medicine, or

as described herein). Examples of useful software include the programs Pile-up and PrettyBox. Such software matches similar sequences by assigning degrees of homology to various substitutions, deletions, substitutions, and other modifications.

Alternatively, or additionally, two nucleic acid sequences may be "substantially identical"if they hybridize under high stringency conditions. In some embodiments, high stringency conditions are, for example, conditions that allow hybridization comparable with the hybridization that occurs using a DNA probe of at least 500 nucleotides in length, in a buffer containing 0.5 M NaHP04, pH 7.2, 7% SDS, 1 mM EDTA, and 1 % BSA (fraction V), at a temperature of 65°C, or a buffer containing 48% formamide, 4.8x SSC, 0.2 M Tris-Cl, pH 7.6, lx Denhardt's solution, 10% dextran sulfate, and 0. 1 % SDS, at a temperature of 42°C. (These are typical conditions for high stringency northern or Southern hybridizations. ) Hybridizations may be carried out over a period of about 20 to 30 minutes, or about 2 to 6 hours, or about 10 to 15 hours, or over 24 hours or more. High stringency hybridization is also relied upon for the success of numerous techniques routinely performed by molecular biologists, such as high stringency PCR, DNA sequencing, single strand conformational polymorphism analysis, and in situ hybridization. In contrast to northern and Southern hybridizations, these techniques are usually performed with relatively short probes (e. g. , usually about 16 nucleotides or longer for PCR or sequencing and about 40 nucleotides or longer for in situ hybridization). The high stringency conditions used in these techniques are well known to those skilled in the art of molecular biology, and examples of them can be found, for example, in Ausubel et al., Current Protocols in Molecular Biology, John Wiley & Sons, New York, N. Y., 1998, which is hereby incorporated by reference.

A"variant"is a nucleic acid molecule that is a recognized variation of a nucleic acid molecule or expression product thereof. Splice variants may be determined for example by using computer programs, e. g, BLAST. Allelic variants have in general a high percent identity to the nucleic acid molecule of interest."Single nucleotide polymorphism" (SNP) refers to a change in a single base as a result of a substitution, insertion or deletion. The change may be conservative (purine for purine) or non-conservative (purine to pyrimidine) and may or may not result in a change in an encoded amino acid. An"analog"is a nucleic acid molecule or polypeptide that

has been subjected to a chemical modification. Nucleic acid analogs can include substitution of a non-traditional base such as queosine or of an analog such as hypoxanthine, or other substitutions known in the art. Analogs in general retain the biological activities of the naturally occurring molecules but may confer advantages such as longer lifespan or enhanced activity. By"complementary'or"complement"is meant that two nucleic acids, e. g. , DNA or RNA, contain a sufficient number of nucleotides which are capable of forming Watson-Crick base pairs to produce a region of double-strandedness between the two nucleic acids. Thus, adenine in one strand of DNA or RNA pairs with thymine in an opposing complementary DNA strand or with uracil in an opposing complementary RNA strand. It will be understood that each nucleotide in a nucleic acid molecule need not form a matched Watson- Crick base pair with a nucleotide in an opposing complementary strand to form a duplex. A nucleic acid molecule is"complementary"to another nucleic acid molecule, or is a"complement"of that other nucleic acid molecule, if it hybridizes, under conditions of high stringency, with the second nucleic acid molecule. The "complement"of a nucleic acid molecule of Tables 1 through 4 may in some embodiments include a nucleic acid molecule that is complementary over the full length of the sequence of a nucleic acid molecule of Tables 1 through 4. A "fragment"may be any portion of a nucleic acid molecule or polypeptide as described herein that is capable of being differentially expressed or detected in an assay or screening method according to the invention.

Various genes and nucleic acid sequences of the invention may be recombinant sequences. The term"recombinant"means that something has been recombined, so that when made in reference to a nucleic acid construct the term refers to a molecule that is comprised of nucleic acid sequences that are joined together or produced by means of molecular biological techniques. The term"recombinant"when made in reference to a protein or a polypeptide refers to a protein or polypeptide molecule which is expressed using a recombinant nucleic acid construct created by means of molecular biological techniques. The term"recombinant"when made in reference to genetic composition refers to a gamete or progeny with new combinations of alleles that did not occur in the parental genomes. Recombinant nucleic acid constructs may include a nucleotide sequence which is ligated to, or is

manipulated to become ligated to, a nucleic acid sequence to which it is not ligated in nature, or to which it is ligated at a different location in nature. Referring to a nucleic acid construct as'recombinant'therefore indicates that the nucleic acid molecule has been manipulated using genetic engineering, i. e. by human intervention.

Recombinant nucleic acid constructs may for example be introduced into a host cell by transformation. Such recombinant nucleic acid constructs may include sequences derived from the same host cell species or from different host cell species, which have been isolated and reintroduced into cells of the host species. Recombinant nucleic acid construct sequences may become integrated into a host cell genome, either as a result of the original transformation of the host cells, or as the result of subsequent recombination and/or repair events.

As used herein,"heterologous"in reference to a nucleic acid or protein is a molecule that has been manipulated by human intervention so that it is located in a place other than the place in which it is naturally found. For example, a nucleic acid sequence from one species may be introduced into the genome of another species, or a nucleic acid sequence from one genomic locus may be moved to another genomic or extrachromasomal locus in the same species. A heterologous protein includes, for example, a protein expressed from a heterologous coding sequence or a protein expressed from a recombinant gene in a cell that would not naturally express the protein.

A compound is"substantially pure"when it is separated from the components that naturally accompany it. Typically, a compound is substantially pure when it is at least 10%, 20%, 30%, 40%, 50%, or 60%, more generally 70%, 75%, 80%, or 85%, or over 90%, 95%, or 99% by weight, of the total material in a sample. Thus, for example, a polypeptide that is chemically synthesised, produced by recombinant technology, isolated by known purification techniques, will be generally be substantially free from its naturally associated components. A substantially pure compound can be obtained, for example, by extraction from a natural source; by expression of a recombinant nucleic acid molecule encoding a polypeptide compound; or by chemical synthesis. Purity can be measured using any appropriate method such as column chromatography, gel electrophoresis, HPLC, etc. A nucleic acid molecule is substantially pure or"isolated"when it is not immediately contiguous with (i. e.,

covalently linked to) the coding sequences with which it is normally contiguous in the naturally occurring genome of the organism from which the DNA of the invention is derived. Therefore, an"isolated"gene or nucleic acid molecule is intended to mean a gene or nucleic acid molecule which is not flanked by nucleic acid molecules which normally (in nature) flank the gene or nucleic acid molecule (such as in genomic sequences) and/or has been completely or partially purified from other transcribed sequences (as in a cDNA or RNA library). For example, an isolated nucleic acid of the invention may be substantially isolated with respect to the complex cellular milieu in which it naturally occurs. In some instances, the isolated material will form part of a composition (for example, a crude extract containing other substances), buffer system or reagent mix. In other circumstance, the material may be purified to essential homogeneity, for example as determined by PAGE or column chromatography such as HPLC. The term therefore includes, e. g. , a recombinant nucleic acid incorporated into a vector, such as an autonomously replicating plasmid or virus ; or into the genomic DNA of a prokaryote or eukaryote, or which exists as a separate molecule (e. g. , a cDNA or a genomic DNA fragment produced by PCR or restriction endonuclease treatment) independent of other sequences. It also includes a recombinant nucleic acid which is part of a hybrid gene encoding additional polypeptide sequences. Preferably, an isolated nucleic acid comprises at least about 40%, 50%, 60%, 70%, 80%, 90%, 95%, or 99% (on a molar basis) of all macromolecular species present. Thus, an isolated gene or nucleic acid molecule can include a gene or nucleic acid molecule which is synthesized chemically or by recombinant means. Recombinant DNA contained in a vector are included in the definition of"isolated"as used herein. Also, isolated nucleic acid molecules include recombinant DNA molecules in heterologous host cells, as well as partially or substantially purified DNA molecules in solution. In vivo and in vitro RNA transcripts of the DNA molecules of the present invention are also encompassed by "isolated"nucleic acid molecules. Such isolated nucleic acid molecules are useful in the manufacture of the encoded polypeptide, as probes for isolating homologous sequences (e. g. , from other mammalian species), for gene mapping (e. g. , by in situ hybridization with chromosomes), or for detecting expression of the gene in tissue (e. g. , human tissue, such as peripheral blood), such as by Northern blot analysis.

Polypeptide compounds can be prepared by, for example, replacing, deleting, or inserting an amino acid residue at any position of a peptide or a peptide analog, for example, a peptide as described herein, with other conservative amino acid residues, i. e. , residues having similar physical, biological, or chemical properties. It is well known in the art that some modifications and changes can be made in the structure of a polypeptide without substantially altering the biological function of that peptide, to obtain a biologically equivalent polypeptide. In one aspect of the invention, polypeptides of the present invention also extend to biologically equivalent peptides that differ from a portion of the sequence of the polypeptides of the present invention by conservative amino acid substitutions. As used herein, the term"conserved amino acid substitutions"refers to the substitution of one amino acid for another at a given location in the peptide, where the substitution can be made without substantial loss of the relevant function. In making such changes, substitutions of like amino acid residues can be made on the basis of relative similarity of side-chain substituents, for example, their size, charge, hydrophobicity, hydrophilicity, and the like, and such substitutions may be assayed for their effect on the function of the peptide by routine testing. Conservative changes can also include the substitution of a chemically derivatised moiety for a non-derivatised residue, by for example, reaction of a functional side group of an amino acid. Peptides or peptide analogs can be synthesised by standard chemical techniques, for example, by automated synthesis using solution or solid phase synthesis methodology. Automated peptide synthesisers are commercially available and use techniques well known in the art. Peptides and peptide analogs can also be prepared using recombinant DNA technology using standard methods such as those described in, for example, Sambrook, et al.

(Molecular Cloning: A Laboratory Manual. 2nd ed., Cold Spring Harbor Laboratory, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N. Y. , 1989) or Ausubel et al. (Current Protocols in Molecular Biology, John Wiley & Sons, 1994). Computer programs such as LASERGENE software (DNASTAR, Madison Wis.), MACVECTOR software (Genetics Computer Group, Madison Wis. ) and RasMol software (www. umass. edu/microbio/rasmol) may be used to determine which and how many amino acid residues in a particular portion of the protein may be

substituted, inserted, or deleted without abolishing biological or immunological activity.

Monitoring changes in gene expression may also be advantageous when screening candidate HCC therapeutics. Often candidate compounds are screened and prescreened for the ability to interact with a major target without regard to other effects they may have on cells or in the subject to be treated, such as toxicity, which prevent the development and use of the potential compound. Thus, the methods of the invention may be used to identify candidate compounds suitable for HCC therapy.

In general, candidate or test compounds are identified from large libraries of both natural products or synthetic (or semi-synthetic) extracts or chemical libraries according to methods known in the art. Those skilled in the field of drug discovery and development will understand that the precise source of test extracts or compounds is not critical to the methods of the invention. Accordingly, virtually any number of chemical extracts or compounds can be screened using the exemplary methods described herein. Examples of such extracts or compounds include, but are not limited to, plant-, fungal-, prokaryotic-or animal-based extracts, fermentation broths, and synthetic compounds, as well as modification of existing compounds. Numerous methods are also available for generating random or directed synthesis (e. g. , semi- synthesis or total synthesis) of any number of chemical compounds, including, but not limited to, saccharide-, lipid-, peptide-, and nucleic acid-based compounds. Synthetic compound libraries are commercially available. Alternatively, libraries of natural compounds in the form of bacterial, fungal, plant, and animal extracts are commercially available from a number of sources, including Biotics (Sussex, UK), Xenova (Slough, UK), Harbor Branch Oceanographic Institute (Ft. Pierce, FL, USA), and PharmaMar, MA, USA. Furthermore, if desired, any library or compound is readily modified using standard chemical, physical, or biochemical methods.

Candidate compounds useful for treating HCC may also be identified by assessing variations in the expression of one or more HCC markers, from Tables 1 through 4, prior to and after contacting HCC cells or tissues with candidate pharmacological agents for the treatment of HCC. The cells may be grown in culture (e. g. from a HCC cell line), or may be obtained from a subject, (e. g. in a clinical trial of candidate pharmaceutical agents to treat HCC). Alterations in expression of one or more of

HCC nucleic acid markers (drug targets), in HCC cells or tissues tested before and after contact with a candidate pharmacological agent to treat HCC, indicate progression, regression, or stasis of the HCC thereby indicating efficacy of candidate agents and concomitant identification of candidate compounds for therapeutic use in HCC. Candidate compounds may also be screened for toxicity, specificity, etc.

When a crude extract is found to modulate expression levels of any of the nucleic acid molecules or polypeptides of the invention, further fractionation of the positive lead extract is necessary to isolate chemical constituents responsible for the observed effect. Thus, the goal of the extraction, fractionation, and purification process is the careful characterization and identification of a chemical entity within the crude extract having the modulatory activities. The same assays described herein for the detection of activities in mixtures of compounds can be used to purify the active component and to test derivatives thereof. Methods of fractionation and purification of such heterogeneous extracts are known in the art. If desired, compounds shown to be useful agents for treatment are chemically modified according to methods known in the art. Compounds identified as being of therapeutic, prophylactic, diagnostic, or other value may be subsequently analyzed using HCC cell lines or a animal model for HCC.

Arrays. Microarrays. Libraries. Databases. And Kits In one aspect, the invention provides nucleic acid or polypeptide arrays and biological assays thereof. Arrays refer to ordered arrangements of at least two nucleic acid molecules or polypeptides on a substrate, which can be any rigid or semi-rigid support to which two nucleic acid molecules or polypeptides may be attached. In some embodiments, a substrate may be a liquid medium. Substrates include membranes, filters, chips, slides, wafers, fibers, beads, gels, capillaries, plates, polymers, and microparticles etc. Because the nucleic acid molecules or polypeptides are located at specified locations on the substrate, the hybridization or binding patterns and intensities create a unique expression profile, which can be interpreted in terms of expression levels of particular genes and can be correlated with HCC progression, regression, therapy, etc.

High density nucleic acid or polypeptide arrays are also referred to as "microarrays, "and may for example be used to monitor the presence or level of

expression of a large number of genes or polypeptides or for detecting sequence variations, mutations and polymorphisms. Arrays and microarrays generally require a solid support (for example, nylon, glass, ceramic, plastic, silica, aluminosilicates, borosilicates, metal oxides such as alumia and nickel oxide, various clays, nitrocellulose, etc. ) to which the nucleic acid molecules or polypeptides are attached in a specified 2-dimensional arrangement, such that the pattern of hybridization or binding to a probe is easily determinable. In some embodiments, at least one of the nucleic acid molecules or polypeptides is a control, standard, or reference molecule, such as a housekeeping gene or portion thereof (e. g. , PBGD, GAPDH), that may assist in the normalization of expression levels or assist in the determining of nucleic acid quality and binding characteristics; reagent quality and effectiveness; hybridization success; analysis thresholds and success, etc.

Nucleic acid molecules or polypeptide probes may be derived from compounds as described herein for example in Tables 1 through 4, and the compositions of the invention may be used as elements on a microarray to analyze gene expression profiles. For the purpose of such arrays,"nucleic acids"may include any polymer or oligomer of nucleosides or nucleotides (polynucleotides or oligonucleotides), which include pyrimidine and purine bases, preferably cytosine, thymine, and uracil, and adenine and guanine, respectively. A variety of methods are known for making and using microarrays, as for example disclosed in Cheung, V. G., et al. 1999; Lipshutz, R.. J. , et al. 1999; Bowtell, D. D. L. , 1999; and, Schweitzer, B., 2002 ; G. MacBeath and S. L. Schreiber, 2000.; all of which are incorporated herein by reference. In some embodiments, the microarray substrate may be coated with a compound to enhance synthesis of the nucleic acid molecule on the substrate as disclosed in, for example, U. S. Pat. No. 4,458, 066. In some embodiments, probes may be synthesized directly on the substrate in a predetermined ordered arrangement.

Methods for storing, querying and analyzing microarray data have for example been disclosed in, for example, United States Patent No. 6, 484, 183; United States Patent No. 6, 188, 783; and Holloway, A. J. , 2002; each of which is incorporated herein by reference. In an alternative aspect, the invention provides nucleic acid or polypeptide microarrays including a number of distinct and selected nucleic acid or polypeptide array sequences of the invention. The number of distinct sequences may for example

be any integer between 2 and 1 x 105, such as at least 102, 103, 104, or 105. The size of the distinct sequences may vary depending on the intended use, and can be determined by a skilled person. For example, the nucleic acid sequences may range from 15 to 5000 bases or more, or any integer between this range.

Microarrays may also be used to examine the expression of all the genes in a tissue or cell such as a liver cell or a HCC cell. Thus, the nucleic acid molecules of Tables 1 through 4 may be attached to a solid support, hybridized with single stranded detectably-labeled cDNAs (corresponding to a"complementary"orientation), and quantified using an appropriate method such that a signal is detected at each location at which hybridization has taken place. The intensity of the signal would then reflect the amount of gene expression. Similarly, protein microarrays may be used according to methods known in the art. Comparison of results from different cells or tissue, for example, hepatocellular carcinoma cells or tissue, hepatitis virus infected cells or tissue, non-tumor cells or tissue, normal cells or tissue, cirrhotic liver cells or tissue, or any combination thereof would elucidate differing levels of expression of specified genes from the different sources.

In one aspect of the invention, libraries may be constructed of bacterial strains each of which bears a plasmid expressing a different nucleic acid molecule of any one or more of Tables 1 through 4 under control of an inducible promoter. ORFs are amplified using PCR and cloned into a vector that enables their expression as N- terminal his-tagged polypeptides. These amplicons are also used to construct hybridization microarrays and enable targeted gene disruption, reducing expenses. A suitable expression host (e. g. E. coli) is selected, and genes encoding particular biochemical activities are identified by screening arrayed pools of his-tagged proteins as described previously (Martzen, M. R. , et al. , 1999).

The invention also provides databases including the nucleic acid and polypeptide sequences described herein, as well as gene expression information in various cancerous and non-cancerous liver and liver cell line samples. Such databases may be used to access information that may aid in diagnosis, prognosis, or other HCC-related methods of the invention. A database as used herein includes any electronic form of the compounds (e. g. , nucleic acid and polypeptide sequences) of

the invention, and information regarding these compounds, and includes computer readable media and any suitable form for storing the information.

The invention also provides kits including for example one or more of the nucleic acid molecules or polypeptides of the invention (or complements, analogs, variants, or fragments thereof), an appropriate buffer, appropriate reagents for detection, and appropriate controls. For example, a kit may include probes or primers (which may or may not be detectably labeled) suitable for hybridization or amplification, or may include antibodies or ligands suitable for specific binding. A kit may also include written or electronic instructions.

Diagnostic and Other Uses A wide variety of detectable labels and conjugation techniques are known by those skilled in the art and may be used in various nucleic acid molecule and polypeptide assays to diagnose HCC. The nucleic acid molecules, proteins, antibodies and other compounds according to the invention may be labeled for purposes of assay by joining them, either covalently or noncovalently, with a detectable label. By"detectably labeled"is meant any means for marking and identifying the presence of a molecule, e. g., an oligonucleotide probe or primer, a gene or fragment thereof, a cDNA molecule, or a polypeptide. Methods for detectably-labeling a molecule are well known in the art and include, without limitation, radioactive labeling (e. g. , with an isotope such as 32p or 35S) and nonradioactive labelling such as, enzymatic labeling (for example, using horseradish peroxidase or alkaline phosphatase), chemiluminescent labeling, fluorescent labeling (for example, using fluorescein), bioluminescent labeling, or antibody detection of a ligand attached to the probe. Also included in this definition is a molecule that is detectably labeled by an indirect means, for example, a molecule that is bound with a first moiety (such as biotin) that is, in turn, bound to a second moiety that may be observed or assayed (such as fluorescein-labeled streptavidin). Labels also include digoxigenin, luciferases, and aequorin. Synthesis of labeled molecules performed by using labels such as 32P-dCTP, Cy3-dCTP or Cy5-dCTP or 35S-methionine.

Compounds according to the invention may also be directly labeled by chemical

conjugation to amines, thiols and other groups present in the molecules using reagents such as BIODIPY or FITC (Molecular Probes, Eugene, OR, USA).

Compounds, compositions, and reagents according to the invention may be used to detect and quantify differential gene expression; absence, presence, or excess expression of nucleic acid molecules (e. g. , mRNAs) or polypeptides; or to monitor nucleic acid molecule (e. g., mRNA) or polypeptide levels during therapeutic intervention in subjects with HCC. The compounds, compositions, and reagents according to the invention can also be utilized as markers of HCC treatment efficacy over a period ranging from days to months to years. The diagnostic assays may use hybridization, amplification, ligand binding, or antibody technologies to compare gene expression in a biological sample from a subject to reference samples or standards, or to cancerous and non-cancerous samples from the subject, in order to detect altered gene expression. Qualitative or quantitative methods for this comparison are known in the art, and any suitable method may be used.

In order to provide a basis for the diagnosis of HCC, a non-tumor or standard gene expression profile may be established. This may be accomplished by combining a biological sample taken from normal or non-tumor subjects or from non-cancerous tissue from a subject with HCC, with a probe under conditions for hybridization or amplification. Standard hybridization may be quantified by comparing the values obtained using non-tumor subjects or non-cancerous tissue with values from an experiment in which a known amount of a substantially purified target sequence is used. Standard values obtained in this manner may be compared with values obtained from samples from patients who are symptomatic for HCC. Deviation from standard values toward those associated with HCC is used to diagnose HCC. Such assays may also be used to monitor the efficacy of a particular HCC therapy in animal studies, in clinical trials, or to monitor the treatment of an individual patient or groups of patients. Once the presence of HCC is established in a subject and a treatment protocol is initiated, assays according to the invention may be repeated on a regular basis to determine if the level of expression in the subject begins to approximate that which is observed in a non-tumor subject, and to monitor the progression of HCC in the subject. The results obtained from successive assays may be used to show the efficacy of treatment over a period ranging from several days to months.

Compounds, compositions, and reagents (e. g., microarrays) according to the invention may be used to monitor the progression or regression of HCC. The differences in gene expression between healthy and diseased tissues or cells can be assessed and cataloged. By analyzing changes in patterns of gene expression, HCC can be diagnosed at earlier stages before the subject is symptomatic. Similarly, by analyzing gene expression profiles and changes therein, prognoses may be formulation, and therapies may be designed. Progression or regression of HCC may be determined by comparison of two or more different HCC samples taken at multiple different times from a subject (e. g. , at least 2,3, 4, or 5 or more time points) over the course of days to months. For example, progression or regression may be evaluated by assessments of expression of sets of two or more, or as many as all, of the nucleic acid molecules of Tables 1 through 4 in a HCC tissue sample from a subject before, during, and following treatment for HCC.

Compounds, compositions, and reagents (e. g. , microarrays) according to the invention can also be used to monitor the efficacy of a therapy. For therapies with known side effects, compounds, compositions, and reagents (e. g., microarrays) according to the invention may be employed to improve the therapeutic regimen. For example, dosages that causes changes in gene profiles that represent efficacious treatment may be determined, and expression profiles associated with the onset of undesirable side effects may be avoided. This approach may be more sensitive and rapid than waiting for the subject to show inadequate improvement, or to manifest side effects, before altering the course of treatment. In another aspect of the invention, pre-and post-treatment alterations in expression of two or more sets of HCC nucleic acid molecules in HCC cells or tissues may be used to assess treatment parameters including, but not limited to: dosage, method of administration, timing of administration, and combination with other known treatments for HCC.

In some aspects, any one or more of the compounds provided herein may be used in therapeutic applications. For example, selected compounds provided herein may be used as therapeutic targets for the identification of agents, that modulate their expression levels and/or activity, that may be used to treat HCC.

EXAMPLES Experimental Procedures RNA isolation, RNA amplification acd cDNA microarray hybridization Paired samples of tumor and corresponding non-tumor tissues were obtained from resected liver specimens from thirty-seven (37) patients who had been diagnosed with hepatitis B virus (HBV)-associated HCC and had undergone curative liver resection. A validation tissue set composed of 58 liver biopsy samples from an independent cohort of twenty-nine (29) patients, who also had HBV-associated HCC and had undergone liver resection, was used. Informed consent from the patient and institutional research and ethics committee approval were obtained. Tissues were snap frozen in liquid nitrogen and stored at-150°C. A small section of each specimen was sampled and total RNA was isolated from tissues using TRIZOL reagent (Life Technologies, Bethesda, MD, USA) according to the manufacturer's instructions. The integrity of the RNA specimen was verified by gel electrophoresis.

The human liver cancer cell lines used in this study were: PLC/PRF/5, HA22T, Huhl, Huh4, Tong, Hep3B, SNU182, SNU449, SNU475, HepG2, Huh6, Huh7, SKHepl, and Mahlavu. All cell lines were cultured under conditions recommended by the American Type Culture Collection (VA, USA).

Total RNA was extracted using TRIZOL reagent (Life Technologies, Bethesda, MD, USA) according to the manufacturer's instructions.

RNA was linearly amplified using a procedure modified from Eberwine and coworkers (Eberwine et al, 1992). Briefly, total RNA was reverse transcribed using a 63-nucleotide synthetic primer containing the T7 polymerase binding site 5'- GGCCAGTGAATTGTAATACGACTCACTATAGGGAGGCGG (T) 24-3'. Full- length double-stranded cDNA synthesis was accomplished in the presence of E. coli DNA polymerase I, DNA ligase and RNase H. The cDNA was made blunt-ended with T4 DNA polymerase, and purified by extraction in a mixture of phenol, chloroform and isoamyl alcohol, and precipitation in the presence of ammonium acetate and ethanol. Purified double-stranded cDNA was then transcribed with T7 polymerase (T7 Megascript Kit, Ambion) to yield linearly amplified antisense RNA, which was subsequently purified with RNeasy mini-columns (Qiagen). Human universal reference RNA (Stratagene, La Jolla, CA), including total RNA from 10

different human cell lines, was amplified and used as the reference for cDNA microarray analysis.

Approximately 9000 human cDNA features (Incyte Genomics, Palo Alto, CA, USA) were spotted onto poly-L-lysine coated slides using OmniGride arrayer (GeneMachines). Probes were generated from the amplified RNA material and hybridized to the chip as described elsewhere (Sotiriou et al, 2002). Briefly, 4 llg amplified RNA was reverse-transcribed using random hexamers and directly labeled with Cy3-conjugated dUTP (reference RNA) or Cy5-conjugated dUTP (sample RNA). Hybridization was performed in the presence of 25% formamide and 5X SSC for 16h at 42°C. Slides were scanned with an Axon 4000b laser scanner (Axon Instruments) after washing and drying. To minimize the effects of labeling biases, reciprocal dye swap labeling experiments were performed for each sample.

Data analysis The 37 paired HCC tumor and non-tumor liver samples, and liver cancer cell lines were processed on the microarray on two separate prints, and the validation tissue set was processed on a third print. Raw data was analyzed on GenePix analysis software version 3.0 (Axon Instruments, Burlingame, CA, USA) and uploaded to a relational database maintained by the Center for Information Technology at the National Institutes of Health (ie. MADB). The cDNA clones used for the microarray are represented by their UniGene identifiers. For each array, the logarithmic expression ratio for a spot on each array was normalized by subtracting the median logarithmic ratio for the same array. Data was filtered to exclude spots with a size of less than 25 urn and any poor quality or missing spots. Since the correlation of the overall data from reciprocal labeling was good, values obtained from reciprocal labeling experiments were averaged. In addition, any gene features that were found to be absent from the data in more than 50% of patient samples in either set of arrays were excluded, and gene features that were common in data from the array print sets were retained. Application of these filters resulted in the inclusion of 8716 of the total 9127 features in subsequent analysis. Statistical comparison of genes between HCC tumors and non-tumors was performed by the Wilcoxon rank- sum non-parametric test. To evaluate gene expression patterns, hierarchical clustering

using one minus Pearson's correlation metric and average linkage (Eisen et al, 1998) and multidimensional scaling was performed on normalized data (mean equals zero, standard deviation equals one). Functional characterization of genes was based on Gene Ontology (The Gene Ontology Consortium, 2000) and other published works known to those of ordinary skill in the art.

The quality of a set of selected gene features to be used as potential markers was measured by estimating the probability that its observed performance, in terms of number of misclassified tissue samples, could occur by chance alone. This was achieved by performing a series of Monte Carlo simulations (Davison and Hinldey, 1997) upon the expression data of the selected genes. In each simulation, the tissues'labels were randomly permuted and the number of misclassifications was noted. A total of one million runs of Monte Carlo simulations were performed. The reported P-value (denoted as Pa) is the fraction of times the permutations generated as few misclassifications as, or fewer than, the original labeling. To determine whether the set of genes observed to have a good performance as tumor discriminators, could appear merely by chance, different Monte Carlo simulations were carried out. In each simulation, an equivalent number of gene features was randomly picked from a designated large population of features, and the performance of the random gene set was evaluated by the number of tissue samples that were misclassified. A total of 10,000 runs of Monte Carlo simulations were performed for each evaluation. The P- value (Pb) is the fraction of times the random gene set performed as good as, or better than, the performance of the selected gene set.

The significance of the number of observed overlapping genes after intersection of the important gene lists, derived as described herein, with gene lists reported previously was approximated by measuring the probabilities that such overlap could occur by chance alone. A separate series of Monte Carlo simulations were employed to estimate the P-values of the two-group comparisons. In each simulation, two lists of genes corresponding to the two groups were generated. Each list was constructed by randomly selecting genes, as many as the number of genes in its corresponding group, from the entire collection of genes of its respective microarray gene set. The two random gene lists were then intersected.

The P-value (Pc) of the comparison was obtained by generating and intersecting the

two random lists 100 million times, and reported as the fraction of times the random overlap is equal or greater than the observed one. To validate the utility of the various expression cassettes to distinguish HCC tumor from non-tumor liver, the prediction accuracy of each discriminator cassette was assessed on an independent tissue set comprising of 58 liver clinical biopsies from 29 patients using a k-Nearest Neighbor (kNN) classification algorithm (k=3) using Pearson correlation to measure the similarity between expression profiles. The algorithm was trained against the dataset comprising 74 tissue samples from 37 patients before testing against the new tissue set.

Real-time semi-quantitative RT-PCR Total RNA from individual tissue samples were analyzed for the expression levels of selected genes by real-time semi-quantitative RT-PCR using the LightCycler RNA amplification kit SYBR Green I on the LightCycler (Roche, Basel, Switzerland) according to the manufacturer's instructions. Briefly, one-step RT-PCR reactions consisted of an initial incubation at 55°C for 10 min, followed by a denaturation step at 95°C for 30 s, and amplification for 40 cycles of 1 s at 95°C, 10 s at 57°C, and 13 s at 72°C. For each reaction, 10 ng of total RNA was analyzed. The gene specific primers designed were, for example, as follows: IGFBP3 5'- ATAATCATCATCAAGAAAGGGCAT-3'and 5'-GAAGGGCGACACTGCTT-3' ; EGFR 5'-GCGTCTCTTGCCGGAATG-3'and 5'-GGCTCACCCTCCAGAAGCTT- 3' ; ERBB2 5'-GGATGTGCGGCTCGTACAC-3'and 5'- TAATTTTGACATGGTTGGGACTCTT-3' ; ERBB3 5'- CGGTTATGTCATGCCAGATACAC-3'and 5'- ACAGAACTGAGACCCACTGAAGAA-3'; PBGD 5'- GAGTGATTCGCGTGGGTACC-3'and 5'-GGCTCCGATGGTGAAGCC-3'. The relative expression level of each gene of interest in individual tissue sample was normalized against that of the"housekeeping"gene PBGD. Data are presented as the level of gene expression in each HCC tumor relative to its corresponding non- tumor liver specimen.

Assessment Of Global Gene Expression Differences Between HCC Tumors And Non-Tumor Liver Specimens The gene expression patterns of primary HCC tumors and the corresponding non-tumor liver tissues from 37 patients were examined by cDNA microarray. Amplified RNA prepared from each experimental sample was labeled with Cy5 and hybridized on the array with pooled human'common reference'amplified RNA labeled with Cy3. Reciprocal dye swap replicate hybridizations were performed to minimize technical noise. Since the overall correlation of reciprocal labeling was good, values obtained from reciprocal labeling experiments were averaged and used in subsequent analyses. Firstly, the overall natural patterns of gene expression in the HCC tumor and non-tumor liver tissues were assessed based on unsupervised hierarchical clustering. Analysis of variance in expression levels for each gene across all the tissues indicated that 500 gene features (containing 493 unique UniGenes) showed the largest variability across both HCC tumor and non-tumor liver tissues (Figure 1). Included in this list are AFP, an often used prognostic marker for HCC, and other genes associated with HCC such as HGF, MYC, and a ras family member RAN. Hierarchical clustering analysis based on these highly variant genes (derived from the 37 pairs of HCC tumor and non-tumor liver samples and using the 500 most variable gene features) separated the tissues into two main clusters, one representing the HCC tumors and the other, the non-tumor liver tissues with only six of 37 HCC tumors misclassified as non-tumors. Thus, the molecular configuration of HCC can be readily distinguished from that of non-tumor liver with minimal data manipulation.

Next, to investigate differential gene expression patterns between HCC tumors and non-tumor livers, the Wilcoxon rank-sum test was used and the top 2.5% candidate genes which displayed the smallest (best) P-value scores (P<lx10-6) and at least 1. 5-fold change in gene expression were identified, resulting in a list of 218 genes (Table 1). For these 218 genes, false discovery rate analysis indicated a false- positive error of less than 0.4%. Multidimensional scaling analysis based on these outlier genes indicated that the HCC tumors were a more heterogeneous population than the non-tumor liver tissues (Figure 2).

Cancer cell lines derived from the primary tumor have traditionally been used as in vitro model systems for investigating the function of genes in the in vivo tumor environment. Using the 218 differentially expressed outlier genes identified in the clinical samples, the expression pattern of the same genes in 14 established human liver cancer cell lines was analysed. These cell lines exhibited gene expression profiles that were different from the clinical HCC tumor tissues (Figure 2), suggesting that they may have accumulated additional genetic or epigenetic alterations in culture and are not entirely reflective of the primary tumor biology.

Identification Of Gene Clusters Differentially Expressed In HCC Tumors Among the statistically significant 218 genes that distinguished HCC tumors from non-tumor liver tissue specimens, more genes were observed to be overexpressed than under-expressed in the malignant tissue specimens relative to the non-tumor tissue specimens. Mapping of the chromosomal location of these 218 unique outlier genes indicated that a disproportionate number of genes was located on chromosome 1 (Figure 3A), particularly in the I q region, and that majority of these genes were more highly expressed in the tumor tissues. Further characterization of these outlier genes revealed that a substantial proportion of genes was involved in transport (e. g., PEA15), RNA processing (e. g., RDBP), and metabolic processes (e. g., NME1) and showed increased expression in HCC tumor specimens, possibly indicating accelerated rates of metabolism (Figure 3B, Table 1). Several outlier genes (e. g., SMT3H1) are members of the ubiquitin-proteasome pathway, suggesting deregulation of this pathway in HCC. A gene cluster associated with lymphocyte infiltrate that included the expression of genes such as IGKC and IGJ was observed, and transcription factors (e. g., ESR1) and genes involved in controlling growth and differentiation (e. g., GRN), and signal transduction (e. g., CSTB) formed the other dominant gene groups. Notably, the polycomb group protein BMI1 was consistently expressed at much higher levels in HCC tumor specimens (Figure 4). Table 1. Genes significantly differentially expressed between HCC tumor and non-tumor liver tissues. Expres- Gene sion Gene Gene Name UniGene change GenBank Identifier in HCC No. tumor* ? AA307289 transcription ILF2 interleukin enhancer Hs. 75117 ? AA307289 factors binding factor 2,45kD BMI1 murine leukemia viral (bmi-Hs. 431? AA884913 1) oncogene homolog TAF9 RNA polymerase 11, Hs. 60679? U21858 TAF9 TATA box binding protein (TBP)-associated factor, 32 kD regulatory factor X, 5 Hs. 166891 ? AL050135 RFX5 (influences HLA class 11 expression) SSRP1 structure specific recognition Hs. 79162? A1635077 protein 1 ZNF146 zinc finger protein 146 Hs. 301819 ? X70394 SREBF2 sterol regulatory element Hs. 108689 ? AA608556 binding transcription factor 2 v-maf musculoaponeurotic Hs. 252229 ? AF059195 MAFG fibrosarcoma (avian) oncogene family, protein G chromodomain helicase Hs.74441 ? BE408958 CHD4 DNA binding protein 4 NR4A1 nuclear receptor subfamily Hs. 1119 ? NM00213 4, group A, member 1 5 ESR1 estrogen receptor 1 Hs. 1657? AL078582 ? AJ223321 ZNF238 zinc finger protein 238 Hs. 69997 ? AJ223321 FOSB FBJ murine osteosarcoma Hs. 75678? L49169 viral oncogene homolo B inhibitor of DNA binding 1, Hs. 75424 ? S78825 ID1 dominant negative helix- loop-helix protein v-fos FBJ murine Hs.25647 ? V01512 FOS osteosarcoma viral oncogene homolog RNA H2AFY H2A histone family, Hs. 75258 ? AA307460 processing member Y small nuclear Hs. 83753? BE252108 SNRPB ribonucleoprotein polypeptides B and B1 RPS7 ribosomal protein S7 Hs. 301547 ? AA315872 MRPS14 mitochondrial ribosomal Hs. 247324? AW973521 protein S14 heterogeneous nuclear Hs. 103804? X65488 HNRPU ribonucleoprotein U (scaffold attachment factor A) SNRPD2 small nuclear Hs. 53125? AA315774 ribonucleoprotein D2 polypeptide (16. 5kD) NCL nucleolin Hs. 79110 AK000250 RPS10 ribosomal protein S10 Hs. 76230 ? AW245775 RPL6 ribosomal protein L6 Hs. 349961? AW675430 splicing factor Hs. 180610 ? X70944 SFPQ proline/glutamine rich (polypyrimidine tract-binding protein-associated) DIM1 similar to S. ombe dim1+ Hs. 5074 ? Al814618 MARS methionine-tRNA Hs. 279946 ? BE299937 synthetase splicing factor, Hs. 77608 ? AL021546 SFRS9 arginine/serine-rich 9 RBM3 RNA binding motif protein 3 Hs. 301404 ? NM_00674 3 U2 small nuclear Hs. 7655 ? AA936430 U2AF65 ribonucleoprotein auxiliary factor (65kD) splicing factor, Hs. 73737 ? M72709 SFRS1 arginine/serine-rich 1 (splicing factor 2, alternate splicing factor) small nuclear Hs. 334612 ? X12466 SNRPE ribonucleoprotein polypeptide E SF3B4 splicing factor 3b, subunit 4, Hs. 25797 ? NM00585 49kD 0 RDBP RD RNA-binding protein Hs. 106061? X16105 small nuclear Hs. 105465? AA649986 SNRPF ribonucleoprotein polypeptide F RRM1 ribonucleotide reductase M1 Hs. 2934 ? X59543 polypeptide RPL38 ribosomal rotein L38 Hs. 2017 ? A1832988 HNRPH1 heterogeneous nuclear Hs. 245710 ? BE296051 ribonucleoprotein H1 (H) U5-U5 snRNP-specific protein, Hs. 151787 ? D21163 116KD 116 kD RPLP1 ribosomal protein, large, P1 Hs. 177592? AW963733 OXA1L oxidase (cytochrome c) Hs. 151134 ? X80695 assembl 1-like DNA ADP-ribosyltransferase Hs. 177766 M18112 replication/ADPRT (NAD+; poly (ADP-ribose) repair polymerase) protein kinase, DNA-Hs. 155637? U34994 PRKDC activated, catalytic polypeptide SMC4 (structural Hs. 50758 ? AB019987 SMC4L1 maintenance of chromosomes 4, yeast)-like 1 H2AV histone H2A. F/Z variant Hs. 301005 ? BE409809 FEN1 flap structure-specific Hs. 4756 ? BE278623 endonuclease 1 minichromosome Hs. 57101 ? BE250461 MCM2 maintenance deficient (S. cerevisiae) 2 (mitotin) HAT1 histone acetyltransferase 1 Hs. 13340 ? AF030424 RAD50 RAD50 (S. cerevisiae) Hs. 41587 ? Z75311 homolo CBX1 chromobox homolog 1 (HP1 Hs. 77254 ? AL046741 beta homolog Drosophila) chondroitin sulfate Hs. 24485? NM00544 proteoglycan 6 (bamacan) 5 FUS fusion, derived from t (12 ; 16) Hs. 99969? BE396632 malignant liposarcoma UNG2 uracil-DNA glycosylase 2 Hs. 3041? AA291356 cell cycle/Hs. 119651 ? U50410 growth/GPC3 glypican 3 differentia- tion cyclin-dependent kinase Hs. 1174 ? Al859822 CDKN2A inhibitor 2A (melanoma, 16, inhibits CDK4) MDK midkine (neurite growth-Hs. 82045 ? AA427949 promoting factor 2 NTRK1 neurotrophic tyrosine Hs. 85844? AA075110 kinase, receptor, type 1 CCNE2 cyclin E2 Hs. 30464 ? NM 00470 2 hepatoma-derived growth Hs. 89525? BE259164 HDGF factor (high-mobility group protein 1-like) TP53BP2 tumor protein p53-binding Hs. 44585? Api123916 Protein, 2 CDC23 CDC23 (cell division cycle Hs. 153546? AF053977 23, yeast, homolo GRN ranulin I Hs. 180577? Al375908 GHR rowth hormone receptor Hs. 125180? X06562 IGFBP3 insulin-like growth factor Hs. 77326? BE336944 binding protein 3 CYR61 cysteine-rich, angiogenic Hs. 8867? Y12084 inducer, 61 hepatocyte growth factor Hs. 809 ? X16323 HGF (hepapoietin A; scatter factor) apoptosis DAP3 death associated protein 3 Hs. 159627? AA207194 PDCD5 programmed cell death 5 Hs. 166468? AA452724 immune PPIA peptidylprolyl isomerase A Hs. 342389? AW732921 response (cyclophilin A TMPO thymopoietin Hs. 11355 ? U09087 PPIB peptidylprolyl isomerase B Hs. 699? BE386706 (cyclophilin B) CD5 antigen-like (scavenger Hs. 52002? NM00589 CD5L receptor cysteine rich 4 family) SCYA14 small inducible cytokine Hs. 20144? NM 00416 subfamily A (Cys-Cys), 6 member 14 SDF1 member 14 Hs. 237356 ? L36033 immunoglobulin heavy Hs. 300697? D78345 IGHG3 constant gamma 3 (G3m marker) C7 complement component 7 Hs. 78065 ? X86328 immunoglobulin J Hs. 76325 ? AW172754 polypeptide, linker protein for immunoglobulin alpha and mu polypeptides IGKC immunoglobulin kappa Hs. 156110 ? AW404507 constant cell Hs. 152931? L25931 adhesion/LBR lamin B receptor cytoskeletal organization integrin, beta 1 (fibronectin Hs. 287797? W38716 ITGB1 receptor, beta polypeptide, antigen CD29 includes MDF2, MSK12) LAMR1 laminin receptor 1 (67kD, Hs. 181357 ? AW328280 ribosomal protein SA) capping protein (actin Hs. 75546 ? U03851 CAPZA2 filament) muscle Z-line, alpha 2 integrin cytoplasmic domain-Hs. 173274? AF012023 associated protein 1 DNCH1 dynein, cytoplasmic, heavy Hs. 7720? AB002323 polypeptide 1 actin related protein 2/3 Hs. 90370 ? Y08999 complex, subunit 1A (41 kD) DPT dermatopontin Hs. 80552 ? AW016451 MMP15 matrix metalloproteinase 15 Hs. 80343 ? D85510 membrane-inserted ARHE ras homolog gene family, Hs. 6838 ? W03441 member E signal CAP2 adenylyl cyclase-associated Hs. 296341 ? AW779995 transduction protein 2 CSTB cystatin B (stefin B) Hs. 695? Al831499 ARMET arginine-rich, mutated in Hs. 75412? AA582041 early stage tumors EFNA1 ephrin-A1 Hs. 1624? 8NM_00442 8 protein phosphatase 2, Hs. 155079 ? AA234460 PPP2R5A regulatory subunit B (B56), alpha isoform RAN, member RAS Hs. 10842 ? NM00632 oncogene family 5 calmodulin 2 Hs. 182278 ? D45887 CALM2 (phosphorylase kinase, delta) LASP1 LIM and SH3 protein 1 Hs. 334851? Al304506 SHC (Src homology 2 Hs. 81972? X68148 SHC1 domain-containing) transforming protein 1 RGS5 regulator of G-protein Hs. 24950 ? Al674877 signalling 5 HAX1 HS1 binding protein Hs. 15318? BE260953 GABRE gamma-aminobutyric acid Hs. 22785 ? NM00496 GABA A receptor, epsilon 1 ADP-ribosylation factor Hs. 118249 ? AA099582 ARFGEF guanine nucleotide- 2 exchange factor 2 (brefeldin A-inhibited MAPK6 mitogen-activated protein Hs. 271980 ? NM00274 kinase 6 8 guanine nucleotide binding Hs. 5662 ? BE206815 GNB2L1 protein (G protein), beta polypeptide 2-like 1 v-erb-b2 avian erythroblastic Hs. 199067 Al565773 ERBB3 leukemia viral oncogene homolo 3 DSCR1 Down syndrome critical Hs. 184222? U85267 region gene 1 CRHBP corticotropin releasing Hs. 115617 ? NM00188 hormone-binding protein 2 serine threonine kinase 39 Hs. 199263 ? F26137 STK39 (STE20/SPS1 homolog, yeast) ubiquitin-Hs. 44532? NM 00639 proteasome UBD diubiquitin 8 pathway proteasome (prosome, Hs. 89545 ? BE336637 PSMB4 macropain) subunit, beta type, 4 Sjogren syndrome antigen Hs. 554 ? NM00460 SSA2 A2 (60kD, ribonucleoprotein 0 autoantiaen SS-A/Ro) ubiquitin specific protease Hs. 75981 ? NM00515 USP14 14 (tRNA-guanine 1 transglycosylase) proteasome (prosome, Hs. 82159? Api889267 PSMA1 macropain) subunit, alpha type, 1 eukaryotic translation Hs. 57783? U62583 EIF3S9 initiation factor 3, subunit 9 (eta, 116kD) SMT3H1 SMT3 (suppressor of mif Hs. 85119? AA160893 two 3, yeast) homolog 1 ubiquitin-conjugating Hs. 108332 ? NM 00333 UBE2D2 enzyme E2D 2 (homologous 9 to yeast UBC4/5) proteasome (prosome, Hs. 90744? AB003102 PSMD11 macropain) 26S subunit, non-ATPase, 11 proteasome (prosome, Hs. 82793? A1028114 PSMB3 macropain) subunit, beta type, 3 proteasome (prosome, Hs. 148495? AA604027 PSMD4 macropain) 26S subunit, non-ATPase, 4 molecular CCT5 chaperonin containing Hs. 1600? D43950 chaperone TCP1, subunit 5 (epsilon CCT3 chaperonin containing Hs. 1708? BE302501 TCP1, subunit 3 (gamma) heat shock 70kD protein 5 Hs. 75410 ? AL043206 HSPA5 (glucose-regulated protein, 78kD CCT4 chaperonin containing Hs. 79150? U38846 TCP1, subunit 4 (delta) HSPA4 heat shock 70kD protein 4 Hs. 90093? AB023420 CCT7 chaperonin containing Hs. 108809? AA314436 TCP1, subunit 7 (eta) HSPA8 heat shock 70kD protein 8 Hs. 180414? AW249010 CCT6A chaperonin containing Hs. 82916 ? L27706 TCP1, subunit 6A (zeta 1) transport ANXA2 annexin A2 Hs. 217493? BE293414 PDZK1 PDZ domain containing 1 Hs. 15456? AF012281 SYPL synaptophysin-like protein Hs. 80919? S72481 translocase of inner Hs. 20716? AW247564 TIMM17A mitochondrial membrane 17 homolog A (yeast) XPO1 exportin 1 (CRM1, yeast, Hs. 79090? D89729 homolog) high mobility group Hs. 236774 ? U90549 HMGN4 nucleosomal binding domain 4 NUCB2 nucleobindin 2 Hs. 3164? AW951523 UGTREL UDP-galactose transporter Hs. 154073? AW192554 1 related PEA15 phosphoprotein enriched in Hs. 194673 ? Y13736 astrocvtes 15 CLTA clathrin, light polypeptide Hs. 104143? AW974204 (Lca) ATPase, H+ transporting, Hs. 6551 ? NM00118 ATP61P1 lysosomal interacting protein 3 1 signal sequence receptor, Hs. 74564? BE313059 SSR2 beta (translocon-associated protein beta) AP3S1 adaptor-related protein Hs. 80917 ? D63643 complex 3,. sigma 1 subunit VDAC2 voltage-dependent anion Hs. 78902 ? Aid 5604 channel 2 VPS45A vacuolar protein sorting 45A Hs. 6650 ? AA702845 easy VCP valosin-containing protein 6 6 SAC2 (suppressor of actin Hs. 169407? AK001725 SACM2L mutations 2, yeast, homolog)-like KPNB1 karyopherin (importin) beta Hs. 180446 ? L38951 1 solute carrier family 21 Hs. 46440? U21943 SLC21A3 (organic anion transporter), member 3 SLC22A1 solute carrier family 22 Hs. 117367? X98332 (organic cation transporter), member 1 Heat shock 70kD protein 5 Hs. 75410 ? (glucose-regulated protein HSPA5 78kD) metabolism GNPAT glyceronephosphate O- Hs.12482 ? AF043937 acyltransferase non-metastatic cells 2, Hs. 275163 ? L16785 NME2 protein (NM23B) expressed in non-metastatic cells 1, Hs. 118638 ? AA147871 NME1 protein (NM23A) expressed in UQCRH ubiquinol-cytochrome c Hs. 73818? Al093521 reductase hinge protein TALD01 transaldolase 1 Hs. 77290 ? AF010400 P5CR2 pyrroline 5-carboxylate Hs. 274287 ? Al161110 reductase isoform GFPT1 glutamine-fructose-6-Hs. 1674? NM00205 phosphate transaminase 1 6 dolichyl-phosphate Hs. 5085 ? AW173486 DPM1 mannosyltransferase polypeptide 1, catalytic subunit ACLY ATP citrate tyase Hs. 174140? AW967351 UDP-Gal : betaGlcNAc beta Hs. 321231? Y12509 B4GALT3 1, 4- galactosyltransferase, polypeptide 3 GCN1 (general control of Hs. 75354 ? D86973 GCN1L1 amino-acid synthesis 1, east-like 1 dolichyl-phosphate (UDP-N-Hs. 26433 ? Z82022 acetylglucosamine) N- DPAGT1 acetylglucosaminephosphotr ansferase 1 (GIcNAc-1-P transferase) acetyl-Coenzyme A Hs. 166160? NM00160 ACM1 acyltransferase 1 7 (peroxisomal 3-oxoacyl- Coenzyme A thiolase) ALDH8A1 aldehyde dehydrogenase 8 Hs. 18443 Al051566 family, member A1 SRD5A2 steroid-5-alpha-reductase, Hs. 1989 ? M74047 alpha polypeptide 2 N-acetyltransferase 2 Hs.2 ? D90040 NAT2 (arylamine N- acetyltransferase) glutathione transferase zeta Hs. 26403 ? U86529 GSTZ1 1 (maleylacetoacetate isomerase) ADH1B alcohol dehydrogenase 1B Hs. 4 ? M24317 (class 1), beta polypeptide cytochrome P450, subfamily Hs. 174220 ? M17398 CYP2C8 IIC (mephenytoin 4- hydroxylase), polypeptide 8 CYP2E cytochrome P450, subfamily Hs. 75183 ? J02843 IIE (ethanol-inducible) ESTs, Highly similar to Hs. 349754? AA313375 Unknown H33_HUMAN HISTONE H3.3 [H. sapiens] ECT2 epithelial cell transforming Hs. 122579? AL137710 sequence 2 oncogene DKFZP56 DKFZP564B167 protein Hs. 76285 ? Al032331 4BI67 translocase of outer Hs. 75187? D13641 KIAA0016 mitochondrial membrane 20 east homolo DXS1357 accessory proteins Hs. 291904? Z31696 E BAP31/BAP29 KIAA0475 KIAA0475 gene product Hs. 5737? AA524523 C rf24 chromosome 20 open Hs. 184062 ? Al340141 reading frame 24 FLJ10326 hypothetical protein Hs. 262823 ? AA665998 FLJ10326 KIAA0117 KIAA0117 protein Hs. 322478? AL133010 Unknown ? NM 00221 1 DEK DEK oncogene (DNA Hs. 110713 ? Al888504 binding) PODXL podocalyxin-like Hs. 16426? BE395330 DSS1 Deleted in split-hand/split- Hs. 333495 ? W79057 foot 1 region PRO1855 hypothetical protein Hs. 283558 ? Al379021 PRO1855 Homo sapiens mRNA ; Hs. 378917? AA425759 cDNA DKFZp4341052 (from clone DKFZp4341052) KIAA0470 KIAA0470 gene product Hs.25132 ? NM_01481 2 MYLE MYLE protein Hs. 11902 ? AA628977 Homo sapiens cDNA Hs. 101810? Al675122 FLJ14232 fis, clone NT2RP4000035 MAGED2 melanoma antigen, family D, Hs. 4943? Z98046 2 FLJ12806 hypothetical protein Hs. 107637 ? BE044582 FLJ12806 tyrosine 3-Hs. 279920 ? AL008725 monooxygenase/tryptophan YWHAB 5-monooxygenase activation protein, beta polypeptide Unknown? AL031775 LOC5123 hypothetical protein Hs. 181444? Al190653 5 KIAA0592 KIAA0592 protein Hs.13273 ? AL080183 KIAA0205 KIAA0205 gene product Hs.3610 ? D86960 C1 rf9 chromosome 1 open Hs. 108636? BE466870 reading frame 9 KIAA0788 KIAA0788 protein Hs.246112 ? AB018331 MGC1955 hypothetical protein Hs. 334787? BE379431 6 MGC19556 KIAA0731 K) AA0731 protein Hs. 6214? AB018274 chromosome 7 open Hs. 84790? D86978 reading frame 14 D123 D123 gene product Hs. 82043 ? U27112 --chromosome 5 open Hs. 75864? BE254013 reading frame 8 Homo sapiens cDNA : Hs. 6127 ? AA054768 FLJ23020 fis, clone LNG00943 AD24 AD24 rotein Hs.74899 ? Al017605 WHIP Werner helicase interacting Hs. 236828? AA481600 protein putative breast Hs. 12107 ? AF042384 BC-2 adenocarcinoma marker (32kD) DKFZP54 DKFZP547E1010 protein Hs. 323817 ? NM_01560 7E101 FLJ22251 hypothetical protein Hs. 289064? AA595663 FLJ22251 ESTs Hs. 89267? AA284067 KIAA0187 KIAA0187 gene product Hs. 10848? D80009 MPV17 MpV17 transgene, murine Hs. 75659 ? NM00243 homolog, lomerulosclerosis 7 MAWBP MAWD binding protein Hs. 16341? Al866254 Homo sapiens cDNA Hs. 346947 ? N44535 FLJ37464 fis, clone BRAWH2011795, weakly similar to LIVER CARBOXYLESTERASE PRECURSOR (EC 3.1. 1. 1) Homo sapiens SNC73 Hs. 293441? AA290845 protein (SNC73) mRNA, complete cds ESTs, Highly similar to Hs. 36102? R99207 SMHU1 B metallothionein 1 B H. sa iens Homo sapiens unknown Hs. 367982 ? H72532 mRNA FLJ12666 hypothetical protein Hs. 23767 ? AW952494 FLJ12666 RNAHP RNA helicase-related Hs. 8765 Al814448 protein *gene expression level showing at least 1. 5-fold change in HCC tumors relative to non-tumor liver tissues (P<1x10-6)

Real-time RT-PCR analysis was performed on a panel of genes, including IGFBP3 and ERBB3 in all the 37 matched HCC tumor and non-tumor liver samples.

Expression of a known"housekeeping"gene porphobilinogen deaminase (PBGD) (Fink et al, 1998) was used as normalizing control. The results of real-time RT-PCR analyses of IGFBP3 and ERBB3 indicated that IGFBP3 expression was diminished in

35 of 37 HCC tumors relative to their corresponding non-tumor liver tissues (Figure SA), while ERBB3 expression was elevated in 34 of 37 tumor samples (Figures 5B).

Since ERBB3 is defective in tyrosine kinase activity and requires dimerization with other receptors, possibly another member of the ERBB family (Riese and Stern, 1998), the hypothesis that HCC tumors expressing high levels of ERBB3 were associated with high expression of ERRB2 or EGFR was tested. The expression of ERBB2 was elevated in 12 of 37 tumors (Figure 5C), while high EGFR expression was found in 15 of 37 tumors (Figure 5D). A significant concomitant increase in ERBB2 expression (t-test P-value-0. 0026), but no association with high EGFR expression (t-test P-value-0. 31) was found in the top fifty percentile of high ERBB3- expressing HCC tumors, indicating that the cognate partners of ERRB3 appeared to be present in those tumors expressing high levels of ERBB3. Real-time and semi- quantitative RT-PCR analyses were also conducted on a panel of genes identified as differentially expressed in HCC (Figures 6-21).

Validation Of HCC Tumor Discriminator Expression Cassettes Changes in gene expression of HCC using microarray technology have been reported (Chen et al, 2002, Okabe et al, 2001; Honda et al, 2001; Shirota et al, 2001; Tackels-Horne et al, 2001 ; Xu et al, 2001a ; Xu et al, 2001b). The intersection of the important gene lists, derived as described herein, with gene lists reported previously was explored, and resulted in the identification of additional gene lists or"expression cassettes" (Tables 2-4) that were capable of distinguishing HCC tumor from non- tumor liver tissues.

In the first gene list, a total of 265 features, containing 245 unique UniGenes from the microarray used herein were observed to overlap (Table 2). Hierarchical clustering analyses based on expression levels of these 265'overlap'features separated the tissue set into two distinct groups of tumor and non-tumor, with five tissue samples misclassified. Such clustering was significant (Pa<lx10-6) based on random permutation testing of sample labels. The likelihood of a randomly chosen set of 265 features producing five or fewer samples misclassified was low (Pb=1. 5xlO-3).

Thus, these 265'overlap'features could distinguish HCC tumor from non-tumor liver with reasonable precision, and the features were unlikely to appear by chance. Among

these genes were smaller subgroups characterized by distinct gene expression signatures involving potential different pathways. A cholesterol biosynthetic pathway was characterized by higher expression in HCC tumors for genes. of the enzymes SQLE, ACLY and FDPS. A subgroup involved in growth and differentiation was characterized in the HCC tumor tissues by lower expression of ESRI, IGFBP3 and PDGFRa, and high expression of PPTB1. A subgroup of bZIP transcription factors ATF3, FOS, JUN, and MYBL2 was characterized to be down-regulated in the HCC tumor tissues.

Table 2. Intersection of microarray expression dataset with HCC Genes UniGene Gene Description GenBank No. Identifier branched chain aminotransferase 2, BE264265, Hs. 101408 BCAT2 mitochondrial AA436410 branched chain aminotransferase 2, NM 001190, Hs.101408 BCAT2 mitochondrial AA436410 AL035296, Hs.102664 VAMP4 vesicle-associated membrane protein 4 AA424813 UDP glycosyltransferase 2 family, J05428, s. polypeptide B7 M746229 AW316760 Hs.10359 ESTs AA630881 Hs. 106061 RDBP RD RNA-binding protein X16105, AA056390 hypothetical protein DKFZp761 F241 AW519080 Hs. 107253 DKFZP761 F241 homo sapiens cDNA : FLJ20925 fis, clone R20416 ADSE00963 Hs. 108441 HAAO 3-hydroxyanthranilate 3, 4-dioxygenase NM_012205, T80846 chromosome 1 open reading frame 9 | BE466870, CH1 MEMBRANE PROTEIN CH1 N36176 Hs.110613 SMG1 PI-3-kinase-related kinase SMG-1 AB007881, KIAA0220 KIAA0220 protein R97225 Hs. 11314 DKFZP564N13 DKFZP564N1363 protein AI360105, 6 T87343 Hs. 115617 CRHBP corticotropin releasing hormone-binding NM 001882, Hs.T15617 CRHBP protein AA286752 AB011182, Hs.118087 KIAA0610 KIAA0610 protein N38860 Hs. 118638 NME1non-metastaticcells1, protein (NM23A) AA147871, expressed in AA644092 hypothetical protein PP591 U79241, human clone 23759 mRNA, partial cds AA626336 Hs. 119651 GPC3 glypican 3 U50410, AA775872 Hs. 12451 EMAPL echinoderm microtubule-associated NM_004434, protein-like AA447196 Hs. 12482 GNPAT glyceronephosphate O-acyltransferase AF043937, AA486845 Hs. 125180 GHR growth hormone receptor X06562, N70358 N94350, Hs. 125359 THY1 Thy-1 cell surface antigen Al346653 AA428836 branched chain keto acid dehydrogenase NM_000056 Hs. 1265 BCKDHB E1, beta polypeptide (maple syrup urine AA427739 disease) M14058, Hs. 1279 C1 R complement component 1, r AA041382 subcomponent AF045649, T69603 Hs. 13999 KIAA0700 KIAA0700 protein Al018400, N55167 Hs. 1430 Fil coagulation factor (plasma AF045649, thromboplastin antecedent) R88990 Hs. 14453 ICSBP1 interferon consensus sequence binding AW964220, rotein 1 N62269 Hs. 14453 ICSBP1 interferon consensus sequence binding AA514545, rotein 1 N62269 AB028970, Hs. 144904 NCOR1 nuclear receptor co-repressor 1 AA085748 AA468619, Hs.144904 NCOR1 nuclear receptor co-repressor 1 AI694342, Hs.145567 AF038169 hypothetical protein AA406301 Hs. 145567 AF038169 hypothetical protein AI963556, AA406301 Hs. 146360 IFITM1 interferon induced transmembrane protein AA428847, 1 (9-27 AA419251 AA044181 likely ortholog of mouse NPC derived Hs.14838 FLJ10773 R93542, proline rich protein 1 AA401264 Hs. ri C1orf16 chromosome 1 open reading frame 16 D87437, KIAA0250 KIAA0250 gene product AA431423 Hs. 151518 TARBP1 TAR (HIV) RNA-binding protein 1 NM_005646, N62244 Hs.15154 SRPX sushi-repeat-containing protein, X NM006307, chromosome AA448569 Hs. 1531 EHHADH enoyl-Coenzyme A, hydratase/3-L07077, R02373 hydroxyac I Coenzyme A dehydrogenase Hs. 1531 EHHADH enoyl-Coenzyme A, hydratase/3- Al800553, hydroxyacyl Coenzyme A dehydrogenase R02373 procollagen-lysine, 2-oxoglutarate 5- AF046889, Hs.153357 PLOD3 dioxygenase 3 AA459305 Hs. 154890 FACL2 fattv-acid-Coenzvme A ligase, long-chain 2 D10040, T73556 Hs. 155079 PPP2R5A protein phosphatase 2, regulatory subunit AA234460, B (B56), alpha isoform R59164 AA203197, Hs. 155560 CANX calnexin 7 protein kinase, DNA-activated, catalytic Hs.155637 PRKDC U34994, R27615 polypeptide N-acetyltransferase 1 (arylamine N- Hs.155956 NAT1 R79401, T67128 acetyltransferase) Hs. 157148 MGC1 3204 hypothetical protein MGC13204 BE262748, Homo sapiens cDNA FLJ11883 fis, clone N62451 HEMBA1007178 Hs. 1578 BIRC5 baculoviral IAP repeat-containing 5 AW247335, survivin AA460685 Hs. 159301 IL18R1 interleukin 18 receptor 1 U43672, AA482489 Hs. 160318 FXYD1 FXYD domain-containing ion transport Al125364, regulator 1 (phospholemman H57136 BE393272, Hs. 160786 ASS argininosuccinate synthetase AA676466 MAWD binding protein Al866254, Hs. 16341 MAWBP ESTs, Weakly similar to predicted using R54416 Genefinder [C. elegans Hs. 16426 PODXL podocalyxin-like BE395330, N64508 regulator factor X, 5 (influences HLA class AL050135, Hs. expression) AA418045 natriuretic peptide receptor A/guanylate Hs. 167382 NPR1 cyclase A (atrionatriuretic peptide receptor AA598841 A) cytochrome P450, subfamily IIC Hs. 167529 CYP2C9 (mephenytoin 4-hydroxylase), polypeptide M61857, R89491 9 Hs. 169517 ALDH1B1 aldehyde dehydrogenase 1 family, member M63967, R93550 Bol complement component 1, s Hs.169756 C1S AA055520, subcomponent T62048 AF025887, Hs. 169907 GSTA4 glutathione S-transferase A4 AA152346 Hs. 170001 EIF2B2 eukaryotic translation initiation factor 2B, AA678061, subunit 2 beta, 39kD R86304 Hs. 170001 EIF2B2 eukaryotic translation initiation factor 2B, AF035280, subunit 2 (beta, 39kD) R86304 Hs. 170133 FOX01A forkheadbox01A (rhabdomyosarcoma) AF032885, AA448277 Hs. 171955 TROAP trophinin associated protein (tastin) U04810, H94949 methylenetetrahydrofolate dehydrogenase Hs.172665 MTHFD1 (NADP+ dependent), NM_005956, methenyltetrahydrofolate cyclohydrolase, H10778 formyltetrahydrofolate synthetase AI458142, Hs.173717 PPAP2B phosphatidic acid phosphatase type 2B T71976 AB006537, Hs. 173880 IL1RAP interleukin 1 receptor accessory protein R35902, AA256132 AW967351, Hs. 174140 ACLY ATP citrate lyase Hou47 cytochrome P450, subfamily IIC Hs. 174220 CYP2C8 (mephenytoin 4-hydroxylase), polypeptide M17398, N53136 8 AW963733, Hs. 177592 RPLP1 ribosomal protein, large, P1 Al732304 Al625594, Hs. 17767 KIAA1554 KIAA1554 protein AA857573, H17860 Hs. 179718 MYBL2 v-myb avian myeloblastosis viral oncogene X13293, homolog-like 2 AA456878 AB013382, Hs.180383 DUSP6 dual specificity phosphatase 6 AA630374 Hs.180919 ID2 inhibitor of DNA binding 2, dominant AI950041, negative helix-loop-helix protein H82442 Hs. 181345 SAH SA (rat hypertension-associated) homolog AI632754, N73827 NM001569, Hs. 182018 IRAK1 interleukin-1 receptor-associated kinase 1 AI202323, AA683550 Hs. 18212 DXS9879E DNA segment on chromosome X (unique) W73156, 9879 expressed sequence AA479062 Hs. 182575 SLC15A2 solute carrier family 15 (H+/peptide S78203, transporter), member 2 AA425352 Hs. 1827 NGFR nerve growth factor receptor (TNFR NM002507, superfamily, member 16) R55303 AF119042, Hs. 183858 TIF1 transcriptional intermediary factor 1 R38345, AA016972 aldehyde dehydrogenase 8 family, member eRR Hs. 18443 ALDH8A1 A1 N70701 ESTs Homo sapiens clone 23785 mRNA Hs.184697 AA041362, AA663440 NM_005842, Hs.18676 SPRY2 sprouty (Drosophila) homolog 2 AA453759 Hs. 194660 CLN3 ceroid-lipofuscinosis, neuronal 3, juvenile AW249073, (Batten, Spielmeyer-Vogt disease) W37752 Hs. 194673 PEA15 phosphoprotein enriched in astrocytes 15 Y13736, AA293211 NM_006589, Hs.19554 C1orf2 chromosome 1 open reading frame 2 AB006746, Hs. 198282 PLSCR1 phospholipid scramblase 1 N25945 Hs. 19904 CTH c stathionase c stathionine amma-I ase S52784, R07167 Hs. 20144 SCYA14 small inducible cytokine subfamily A (Cys- NM_004166, Cys), member 14 R96626 NM_000361, Hs.2030 THBD thrombomodulin H59861 interferon-induced protein with NM_001548, Hs.20315 IFIT1 tetratricopeptide repeats 1 AA157787 NM_004419, Hs.2128 DUSP5 dual specificity phosphatase 5 W65460 Hs. 213289 LDLR low density lipoprotein receptor (familial NM_000527, hypercholesterolemia) AA504461 1 solute carrier family 12, U79245, Hs.21413 SLC12A5 (potassium/chloride transporter member 5 AA166885 NM_001070, Hs. 21635 TUBG1 tubulin, gamma 1 T77732 Hs. 2178 H2BFQ H2B histone family, member Q BE245642, AA010223 Hs. 227656 XPR1 xenotropic and polytropic retrovirus AL137583, receptor AA453474 Hs. 23642 HSU79266 protein predicted by clone 23627 U79266, W95346 Hs. 237356 SDF1 stromal cell-derived factor 1AA442810, AA447115 Hs. 237356 SDF1 stromal cell-derived factor 1 L36033, AA447115 AW952494, hypothetical protein FLJ12666 Hs. 23767 FLJ12666 Homo sapiens cDNA FLJ1266 fis, clone AA115300, NT2RM4002256 U83113, Hs.239 FOXM1 forkhead box M1 AA129552 AA725097, Hs. 239069 FHL1 four and a half LIM domains 1 AA455925 hypothetical protein FLJ12389 similar to acetoacetyl-CoA synthetase AI697801, Hs. 239758 FLJ12389 Homo sapiens cDNA FLJ12389 fis, clone R48270 MAMMA1002671, weakly similar to Acetyl- coenzyme A synthase (EC 6.2. 1. 1) Hs. 241561 PRSS2 protease, serine, 2 (trypsin 2) U66061, AA284528 AA705337, Hs.2430 TCFL1 transcription factor-like 1 AI674877, Hs. 24950 RGS5 regulator of G-protein signalling 5 N34362, AA668470 AA203476, Hs. 252587 PTTG1 pituitary tumor-transforming 1 AA430032 AF068007, Hs. 25313 MCRS1 microspherule protein 1 AA488757 AW779701, Hs. 25475 AQP7 aquaporin 7 H27752, AI075055 Hs.256583 ILF3 interleukin enhancer binding factor 3, 90kD AA449048 Hs. 256583 1LF3 interleukin enhancer binding factor 3, 90kD NM 012218, AA449048 Hs. 25797 SF3B4 splicing factor 3b, subunit 4,49kD NM_005850, AA699361 DKFZP434B04 hypothetical protein DKFZp434B044 AA541776, 4 ESTs AA460304 Hs.26403 GSTZ1 glutathione transferase zeta 1 U86529, meleylacetoacetate isomerase) AA428334 Hs. 264330 ASAHL N-acylsphingosine amidohydrolase (acid BE267007, ceramidase)-like W47576 NM_016937, Hs.267289 POLA polymerase (DNA directed), alpha NM_002081, Hs. 2699 GPC1 glypican 1 AA455895 Homo sapiens clone IMAGE : 1963178, A1355014, Hs. 270256 mRNA sequence R10140 ESTs Hs. 270845 KNSL5 kinesin-like 5 (mitotic kinesin-like protein 1) H63163, AA452513 Hs.279607 CAST calpastatin U38525, H78523 BE256559, Hs.284142 C21orf4 chromosome 21 open reading frame 4 W69668 Hs. 284142 C21orf4 chromosome 21 open reading frame 4 BE142872, W69668 Homo sapiens cDNA : FLJ21869 fis, clone AW582012, HEP02442 R63929 NM_001650, Hs.288650 AQP4 aquaporin 4 H09087 Z31696, Hs.291904 DXS1357E accessory proteins BAP31/BAP29 AA625628 Hs. 2934 RRM1 ribonucleotide reductase M1 polypeptide X59543, AA633549 C00821, H63534, methylmalonate-semialdehyde Hs.293970 ALDH6A1 AA196160, dehydrogenase H63534 M93405, N62179, methylmalonate-semialdehyde Hs.293970 ALDH6A1 AA196160, dehydrogenase BE222511, Hs. 294151 KIAA1917 KIAA1917 protein AA452113 Hs. 295923 SIAH1 seven in absentia (Drosophila) homolog 1 AA935716, T71889 Hs. 296049 MFAP4 microfibrillar-associated protein 4 L38486, AA442695 L48516, R95740, Hs.296259 PON3 paraoxonase 3 T57069 AW779995, Hs.296341 CAP2 adenylyl cyclase-associated protei 2 AA040613 Hs. 296341 CAP2 adenylyl cyclase-associated protein 2 U02390, AA040613 ESTs, Weakly similar to JC5238 AA926994, Hs. 30151 galactosylceramide-like protein, GCP N73570 [H.sapiens] AB032991, Hs.30340 KIAA1165 hypotheticla protein KIAA1165 AA449330 AA770150, Hs.30340 KIAA1165 hypotheticla protein KIAA1165 AA449330 NM001122, Hs. 3416 ADFP adipose differentiation-related protein AA700054, AA142916 Hs. 35120 RFC4 replication factor C (activator 1) 4 (37kD) AA600213, N93924 FUS-interacting protein (serine-arginine AK001656 Hs. 3530 FUSIP2 rich) 2 H11042 TLS-associated serine-arginine protein 2 Hs. 36102 ESTs, Highly similar to SMHU1 B R99207, H72722 metallothionein 1B [H.sapiens] NM_001631, Hs.37009 ALPI alkaline phosphatase, intestinal AA190871 Homo sapiens, Similar to hypothetical Hs. 38163 protein, MGC: 7035, clone MGC: 20737 AW074863, IMAGE : 4563636, mRNA, complete cds H63116 ESTs Hs. 3873 PPT1 palmitoyl-protein thioesterase 1 (ceroid-AL037943, lipofuscinosis, neuronal 1, infantile) AA034250 Hs. 388 NUDT1 nudix (nucleoside diphosphate linked Al656937, moietv X)-type motif 1 AA443998 Hs. 4 ADH1 B alcohol dehydrogenase 1 B (class 1), beta M24317, N93428 ou peptide Hs. 41726 SERPINB8 serine (or cysteine) proteinase inhibitor, NM_002640, clade B (ovalbumin), member 8 W60100 Hs. 4187 LOC55977 hypothetical protein 24636 Al066576, N62562 AW409765, Hs.42650 ZWINT ZW10 interactor NM_006398, Hs. 44532 UBD Diubiquitin N33920 Hs. 460 ATF3 activating transcription factor 3 N39944, H21041 Hs. 4742 GPAA1 anchor attachment protein 1 (Gaa1p, NM_003801, east) homolog AA455301 Hs. 4756 FEN1 flap structure-specific endonuclease 1 BE278623, AA620553 Hs. 4788 NCSTN Nicastrin D87442, KIAA0253 nicastrin NCSTN Nicastrin BE179772, Hs.4788 KIAA0253 nicastrin R96527 hypothetical protein HH114 Hs. 48348 HH114 Homo sapiens clone HH114 unknown M130117 mRNA Hs. 4854 CDKN2C cyclin-dependent kinase inhibitor 2C (p18, AF041248, inhibits CDK4) N72115 AI141174, Hs.49265 ESTs AI140241 Hs. 49912 PXMP2 peroxisomal membrane protein 2 (22kD) BE393339, N70714 SMC4 (structural maintenance of chromosomes 4, yeast)-like 1 AB019987, CAP-C chromosome-associated AA452095 polypeptide C Hs.50966 CPS1 carbamoyl-phosphate synthetase 1, Y15793, N68399 mitochondrial carbamoyl-phosphate synthetase 1, AA113231, Hs.50966 CPS1 mitochondrial N68399 NM_014867, Hs.5333 KIAA0711 KIAA0711 gene product AA702544 Hs. 5719 CNAP1 chromosome condensation-related SMC-D63880, associated protein 1 M668256 Hs. 5719 CNAP1 chromosome condensation-related SMC-NM 014865, s. associated protein 1 M668256 Hs. 572 ORM1 orosomucoid 1 X02544, AA700876 Hs. 574. FBP1 fructose-1, 6-bisphosphatase 1 M19922, AA699427 Homo sapiens mRNA ; cDNA AI383214, Hs. 5897 DKFZp586P1622 (from clone T59658 DKFZp586P1622) Hs. 61638 MYO10 myosinX 87977 AA187977 Hs.6551 ATP6S1 ATPase H+ transporting, lysosomal NM_001183, vacuolar proton pump), subunit 1AA487588 Hs. 6566 TRIP13 thyroid hormone receptor interactor 13 BE090548, AA630784 AB012701, Hs.66 IL1RL1 interleukin 1 receptor-like 1 AA125917 Hs. 6838 ARHE homolog gene family, member E W03441, ESTs AA443302 AF098865, Hs.71465 SQLE squalene epoxidase R01118 SWI/SNF related, matrix associated, actin dependent regulator of chromatin, U66619, Hs.71622 SMARCD3 subfamily d, member 3 AA136103 Weakly similar to KIAA0319 [H. sapiens AA194084, Hs. 737 ETR101 immediate early protein AA496359 Hs. 738 RPL14 ribosomal protein L14 BE410686, early growth response 1 AA486533 NM_003993, Hs.73986 CLK2 CDC-like kinase 2 NM_005607, Hs. 740 PTK2 PTK2 protein tyrosine kinase 2 M291486 Hs. 74120 APM2 adipose specific 2 AI093004, W94684 Hs. 74170 MT1 E metallothionein 1 E (functional) H72532, AA872383 NM_000014, Hs.74561 A2M alpha-2-macroglobulin AA775447 Hs. 74566 DPYSL3 dihydropyrimidinase-like 3 D78014 Ai831083 Hs. 74579 KIAA0263 KIAA0263 gene product D87452, AA634464 Hs. 74615 PDGFRA piatelet-derived growth factor receptor, M21574, H23235 alpha polypeptide Hs. 74615 PDGFRA platelet-derived growth factor receptor, AW887370, alpha polypeptide H23235 DnaJ (Hsp40) homolog, subfamily C, Q Hs. 74711 DNAJC8 member 8 T60163 Splicing factor similar to dnaJ fibroblast growth factor receptor 1 (fms- Hs. 748 FGFR1 related tyrosine kinase 2, Pfeiffer X66945, R54610 syndrome) AA911031, tyrosine 3-monooxygenase/tryptophan 5- AA609598, Hs.75103 YWHAZ monooxygenase activation protein, zeta H94670, polypeptide AA485749 tyrosine 3-monooxygenase/tryptophan 5- BE315169 Hs. 75103 YWHAZ monooxygenase activation protein, zeta AA609598, polypeptide H94670, AA485749 clusterin (complement lysis inhibitor, SP-M25915, Hs. 75106 CLU 40, 40, sulfated glycoprotein 2, AA292226, testosterone-repressed prostate message AA464163 2, apolipoprotein J) AA307289 Hs. 75117 ILF2 interleukin enhancer binding factor 2,45kD AA894687, H95638 AA601029, Hs. 75117 ILF2 interleukin enhancer binding factor 2, 45kD AA894687, H95638 Hs. 75196 BAT8 HLA-B associated transcript 8 NM006709, G9A ankyrin repeat-containing protein AA434117 Hs. 75216 PTPRF protein tyrosine phosphatase, receptor F08552, type, F AA598513 X06956, R36063, Hs.75318 TUBA1 tubulin, alpha 1 (testis specific) AA180742 gene from NF2/meningioma region of AB023200, Hs.75361 PK1.3 22q12 AA700048 AA159812, Hs.75438 QDPR quinoid dihydropteridine reductase R38198 Hs. 75545 IL4R interleukin 4 receptor X52425, AA292025 carboxypeptidase B2 (plasma, NM_001872, Hs.75572 CPB2 carboxypeptidase U) H47837 BE122870, Hs. 75618 RAB11A RAB11A, member RAS oncogene family AA025058 Hs. 75658 PYGB phosphorylase, glycogen ; brain U47025, AA922705 Hs. 75678 FOSB FBJ murine osteosarcoma viral oncogene L49169, T61948 homolo B Hs. 75812 PCK2 phosphoenolpyruvate carboxykinase 2 X92720, mitochondrial AA186901 Hs. 76252 EDNRA endothelin receptor type A D90348, AA450009 ESTs, Highly similar to IGJ_HUMAN AW172754, Hs. 76325 IMMUNOGLOBULIN J CHAIN [H. sapiens] T70057 SLU step II splicing factor SLU7 AW589878, Hs.7645 FGB fibrinogen, B beta polypeptide H91121, T73858 AF074657, Hs. 76461 RBP4 retinol-binding protein 4, plasma T72076 Hs. 7647 MAZ MYC-associated zinc finger protein BE264373, (purine-binding transcription factor) M704613 Hs. 77256 EZH2 enhancer of zeste (Drosophila) homolog 2 U52965, Axa428252 Hs. 77326 IGFBP3 insulin-like growth factor binding protein 3 BE336944, AA598601 farnesyl diphosphate synthase (farnesyl Hs. 77393 FDPS pyrophosphate synthetase, D14697, T65790 dimethylallyltranstransferase, geranyltranstransferase) X75932, Hs.77597 PLK polo (Drosophia)-like kinase AA629262 NM_002346, Hs.77667 LY6E lymphocyte antigen 6 complex, locus E AA865464 AB032064, Hs.77854 RGN regucalcin (senescence marker protein-30) H05140 Hs. 78045 ACTG2 actin, gamma 2, smooth muscle, enteric NM 001615, TFP12 I tissue pathwa inhibitor 2AA293402 Hs. 78465 JUN v-jun avian sarcoma virus 17 oncogene AI885769, homolo W96134 AI263464, Hs. 78524 HTCD37 TcD37 homolog AA022472, AA456635 TAF6 RNA polymerase II, TATA box NM_005641, Hs. 78865 TAF6 binding protein (TBP)-associated factor, 80 R19071 kD Hs.789 GRO1 GRO1 oncogene (melanoma growth NM_001511, stimulating activity, alpha) W46900 Hs. 78996 PCNA proliferating cell nuclear antigen AI624204, AA450264 MAD2 (mitotic arrest deficient, yeast, NM002358, Hs.79078 MAD2L1 homolog)-like 1 AA481076 Hs. 79081 PPP1CC protein phosphatase 1, catalytic subunit, NM_002710, amma isoform AA129930 reticulocalbin 2, EF-hand calcium binding AL120373, Hs.79088 RCN2 domain AA598676 X64318, Hs.79334 NFIL3 nuclear factor, interleukin 3 regulated AA633811 AA975473, Hs.79404 D4S234E neuron-specific protein AA875888 Hs. 80248 RBPMSRNA-bindingprotein gene with multiple D84107, T98807 splicing uncoupling protein 2 (mitochondrial, proton AW192446, carrier H61242 Hs. 81170 PIM1 pim-1 oncogene M54915, AA447730 endothelial PAS domain protein 1 Hs. 8136 EPAS1 Homo sapiens clone 23698mRNA U51626, R24882 sequence Hs. 81687 NME3 non-metastatic cells 3, protein expressed U29656, in AA398218 NM_006265, Hs.81848 RAD21 RAD21 (S. pombe) homolog AA683102 Hs. 81892 KIAA0101 KIAA0101 gene product D14657, W68219 Hs. 82042 SLC23A1 solute carrier family 23 (nucleobase transporters), member 1D87075, N23756 Biglycan NM_001711, Hs. 821 BGN Zinc finger protein homologous to Zfp92 in R77226, N51018 mouse M27492, R56687 Hs. 82112 IL1R1 I interleukin 1 receptor, type I M464525 Hs. 82273 FLJ20152 hypothetical protein AI536745, AA446864 Homo sapiens cDNA FLJ30550 fis, clone Hs.82503 BRAWH2001502 Y09836, Homo sapiens mRNA for 3'UTR of AA670382 unknown protein Hs. 8265 TGM2 transglutaminase 2 (C polypeptide, protein- glutamine-gamma-glutamyltransferase) AA156324 NM_004344, Hs.82794 CETN2 centrin, EF-hand protein, 2 N72193 CDC20 (cell division cycle 20, S. Hs.82906 BE293657 cerevisiae, homolog) Hs. 8294 KIAA0196 KIAA0196 ene roduct NM 014846 Hs. 82962 TYMS thymidylate synthetase NM_001071 Hs. 83164 COL15A1 collagen, type XV, alpha 1 L01697 small nuclear ribonucleoprotein 2108 __polypeptides B and B1 Hs. 86368 CLGN calmegin NM 004362 Hs. 86724 GCH1 GTP cyclohydrolase 1 (dopa-responsive Z29433 dystonia Hs. 87409 THBS1 thrombospondin 1 NM 003246 Hs. 8765 RNAHP RNA helicase-related protein Al127821 Hs. 8867 CYR61 cysteine-rich, angiogenic inducer, 61 Y12084 Hs. 8889 SHMT1 serine hydroxymethyltransferase 1 AI761724 (soluble) Hs.89538 CETP cholesteryl ester transfer protein, plasma M30185 Hs. 89691 UGT2B4 UDP glycosyltransferase 2 family, AF064200 polypeptide B4 Hs. 89771 GCKR glucokinase (hexokinase 4) regulatory NM_001486 protein- Hs. 91813 BTN2A2 butyrophilin, subfamily 2, member A2 AI636514 Hs. 93002 UBE2C ubiquitin-conjugating enzyme E2C AI637467 Hs. 93194 APOA1 apolipoprotein A-1 X00566 Hs. 93210 C8A complement component 8, alpha M16974 olypeptide Hs. 93597 CDK5R1 cyclin-dependent kinase 5, regulatory T04872 subunit 1 35 Hs. 93597 CDK5R1 cyclin-dependent kinase 5, regulatory AW088206 subunit 1 35 Hs. 93832 LOC54499 putative membrane protein AW081809 Hs. 94360 MT1 L metallothionein 1 L F26137 Hs. 94382 ADK adenosine kinase NM 001123 Hs. 9568 ZNF261 zinc finger protein 261 X95808 Hs. 9568 ZNF261 zinc finger protein 261 NM 005096 Hs. 95998 FRDA Friedreich ataxia AW409831 Hs. 9629 PRCC papillary renal cell carcinoma BE258195 translocation-associated Hs. 9670 FLJ10948 hypothetical protein FLJ10948 AA805411

In the second gene list, a total of 230 features, containing 166 unique UniGenes from the 218 significant gene list (containing 213 unique UniGenes) identified herein (Table 1) were observed to overlap (Table 3). Hierarchical clustering analysis based on the expression levels of these 230'overlap'features separated the tissue set into distinct tumor and non-tumor groups, with four tissue samples misclassified. Random permutation of sample labels indicated that the clustering was significant (Pa<lx10-8) and it was unlikely that a randomly chosen set of 230 features could produce four or fewer samples misclassified (Pb<lx104). These 230'overlap'features are therefore able to discern HCC tumor from non-tumor liver.

Table 3. Intersection of Significant Genes Identified Herein with HCC Genes UniGene Gene Description GenBank No. Identifier heterogeneous nuclear ribonucleoprotein U X65488, T97547, Hs.103804 HNRPU (scaffold attachment factor A) AA496741 Hs. 104143 CLTA clathrin, light polypeptide (Lca) AW974204, AA113872 Hs. 105465 SNRPF small nuclear ribonucleoprotein polypeptide AA649986, F AA668189 NM_003339, ubiquitin-conjugating enzyme E2D 2 Hs.108332 UBE2D2 AA159600, (homologous to yeast UBC4/5) AA431868 NM_006325, Hs. 10842 RAN RAN, member RAS oncogene family AA456636 D80009, AA121504, Hs. 10848 KIAA0187 KIAA0187 gene product AA129555, AA402812 Hs.108636 C1orf9 chromosome 1 open reading frame 9 BE466870, N36176 AA608556, sterol regulatory element binding Hs.108689 SREBF2 AA701914, transcription factor 2 AA608556 Hs. 108809 CCT7 chaperonin containing TCP1, subunit 7 (eta) AA314436, AA676588 Hs. 110713 DEK DEK oncogene (DNA binding) Al888504, R25377 nuclear receptor subfamily 4, group A, NM_002135, Hs.1119 NR4A1 member 1 N94487 U09087, H21746, Hs.11355 TMPO thymopoietin AA676998, T63980 Hs.115617 CRHBP corticotropin releasing hormone-binding NM_001882, protein N26546, AA286752 Hs. 117367 SLC22A1 solute carrier family 22 (organic cation X98332, AA702013 transporter), member 1 Hs.1174 CDKN2A cyclin-dependent kinase inhibitor 2A AI859822, (melanoma, p16, inhibits CDK4) AA877595 Hs.118249 ARFGEF2 ADP-ribosylation factor guanine nucleotide- AA099582, N34053 exchange factor 2 (brefeldin A-inhibited) NME1 non-metastatic cells 1, protein (NM23A) AA147871, expressed in AA644092 Hs. 11902 MYLE MYLE protein AA628977, T68845 Hs. 119651 GPC3glypican 3 U50410, AA775872 Hs. 12107 BC-2 putative breast adenocarcinoma marker AF042384, N25578 32kid Hs. 12482 GNPAT glyceronephosphate O-acyltransferase AF043937, W72079 X06562, N70358, Hs.125180 GHR growth hormone receptor AA775738 AF030424 Hs. 13340 HAT1 histone acetyltransferase 1 AA625662 Hs. 148495 PSMD4 proteasome (prosome, macropain) 26S AA604027, subunit, non-ATPase, 4 AA450227 Hs. 151787 U5-116KD U5 snRNP-specific protein, 116 kD D21163, AA779221 Hs. 152931 LBR lamin B receptor L25931, AA099136 Hs.15318 HAX1 HS1 binding protein BE260953, R76263 AW192554, Hs.154073 UGTREL1 UDP-galactose transporter related R41839 Hs. 155079 PPP2R5A protein phosphatase 2, regulatory subunit B AA234460, R59164 (B56), alpha isoform Hs. 155637 PRKDC kinase, DNA-activated, U34994, R27615 polypeptide AW404507, AI732289, Hs.156110 IGKC immunoglobulin kappa constant AA476918, AA486362 Hs. 1600 CCT5 chaperonin containing TC1, subunit 5 D43950, (epsilon) AA126599, AA629692 NM_004428, Hs.1624 EFNA1 ephrin-A1 AA857015 Hs. 16341 MAWBP MAWD binding protein AI866254, R54416 Hs. 16426 PODXL podocalyxin-like BE395330, N64508 AL078582, Hs. 1657 ESR1 estrogen receptor 1 AA164585, AA291702 AA452724, Hs.166468 PDCD5 programmed cell death 5 AA156940 regulatory factor X, 5 (influences HLA class II AL050135, Hs.166891 RFX5 expression AA418045 glutamine-fructose-6-phosphate NM_002056, Hs.1674 GFPT1 transaminase 1 AA478571 SAC2 (suppressor of actin mutations 2, AK001725, Hs.169407 SACM2L yeast, homolog)-like AA454836 Hs.173274 ICAP-1A integrin cytoplasmic domain-associated AF012023, protein 1 AA456882 AW967351, Hs.174140 ACLY ATP citrate lyase H08547, AA126708 cytochrome P450, subfamily IIC Hs.174220 CYP2C8 M17398, N53136 (mephenytoin 4-hydroxylase), polypeptide 8 AW963733, Hs.177592 RPLP1 ribosomal protein, large, P1 AW249010, Hs. 180414 HSPA8 heat shock 70kD protein 8 H64096, AA629567 L38951, AA121732, Hs. 180446 KPNB1 karyopherin (importin) beta 1 AA251527, AA425006 AI375908, Hs. 180577 GRN granulin A1054019, AA496452 splicing factor proline/glutamine rich X70944 R96240 Hs. 180610 SFPQ (polypyrimidine tract-binding protein- N24024, AA425258 associated) Hs. 181357 LAMR1 laminin receptor 1 (67kD, ribosomal protein AW328280, SA) AA629897 Hs. 181444 LOC51235 hypothetical protein AI190653, AA455565 Hs. 184222 DSCR1 Down syndrome critical region gene 1 U85267, AA629707 Hs. 18443 ALDH8A1 aldehyde dehydrogenase 8 family, member AI051566, N70701 Hs. 194673 PEA15 phosphoprotein enriched in astrocytes 15 Y13736, AA293211 Hs. 1989 SRD5A2 steroid-5-alpha-reductase, alpha polypeptide M74074, AI420552 Hs. 199067 ERBB3 v-erb-b2 avian erythroblastic leukemia viral AI565773, N24966, oncogene homolog 3 AA042878 MT1L metallothionein 1 L Hs. 199263 STK39 serinethreoninekinase 39 (STE20/SPS1 F26137, H84871 homolog, yeast) Hs. 2 NAT2 N-acetyltransferase 2 (arylamine N-D90040, Al262683 acetyItransferase Hs. 20144 SCYA14 small inducible cytokine subfamily A (Cys-NM-004166, Ces), member 14 R96626 H 20 1 translocaseofinnermitochondrialmembrane AW247564, 17 homolo A east AA708446 Hs. 22785 GABRE gamma-aminobutyric acid (GABA) A NM004961, receptor, epsilon H63532 Hs.236774 HMG17L3 high-mobility group (nonhistone U90549, R17124 chromosomal) protein 17-like 3 Hs. 236828 WHIP Werner helicase interacting protein AA481600, AA18868 Hs.237356 SDF1 stromal cell-derived factor 1 L36033, AA447115 AW952494, Hs.23767 FLJ12666 hypothetical protein FLJ12666 NM005445, Hs. 24485 CSPG6 chondroitin sulfate proteoglycan 6 (bamacan) W40150, AA463410 Hs.245710 HNRPH1 heterogeneous nuclear ribonucleoprotein H1 BE296051, (H) R11018, W96058 AW973521, Hs.247324 MRPS14 mitochondrial ribosomal protein S14 T51290, AA460831 AI674877, N34362, Hs.24950 RGS5 regulator of G-protein signalling 5 NM014812, Hs. 25132 KIAA0470 KIAA0470 gene product A1049669, AA167129, AA187982 v-maf musculoaponeurotic fibrosarcoma AF059195, avian oncogene family, protein G N21609, Hs. 25647 FOS v-fos FBJ murine osteosarcoma viral V01512, R12840, oncogene homolog N36944, AA485377 Hs. 25797 SF3B4 splicing factor 3b, subunit 4, 49kD NM_005850, AA699361 glutathione transferase zeta 1 Hs. 26403 GSTZ1 male lacetoacetate isomerase U86529, M428334 dolichyl-phosphate (UDP-N- Hs. 26433 DPAGT1 acetylglucosamine) N-Z82022, R55619, acetylglucosaminephosphotransferase 1 AA452517 (GicNAc-1-P transferase) Hs. 271980 MAPK6 mitogen-activated protein kinase 6 NM 002748, AA603152, Ho7504 non-metastatic cells 2, protein (NM23B) L16785, Hs.275163 NME2 expressed in AA422058, AA496512 integrin, beta 1 (fibronectin receptor, beta polypeptide, antigen CD29 includes MDF2, W38716, MSK12) AA037283, Homo sapiens, clone MGC: 17220 W67173 Hs. 291904 DXS1357E accessory proteins BAP31/BAP29 Z31696, AA625628 Hs. 2934 RRM1 ribonucleotide reductase M1 polypeptide X59543, AA633549 Hs.293441 Homo sapiens SNC73 protein (SNC73) AA290845, mRNA, complete cds H28469, H73590 AW779995, Hs.296341 CAP2 adenylyl cyclase-associated protein 2 AA040613 immunoglobulin heavy constant gamma 3 D78345, Hs.300697 IGHG3 (G3m marker) AA740786, N92646, AA465378 Hs. 301005 H2AV histone H2A.F/Z variant BE409809, H97000 Hs. 301404 RBM3 RNA bindin motif protein 3 NM 006743, AA054287 Hs. 301819 ZNF146 zinc finger protein 146 X70394, AA504351 Hs. 3041 UNG2 uracil-DNA glycosylase 2 AA291356, AA425900 AW951523, Hs. 3164 NUCB2 nucleobindin 2 W93954, AA484939 Hs. 321231 B4GALT3 UDP-Gal : betaGIcNAc beta 1, 4- Y12509, AA424578 alactosyItransferase, pol eptide 3 NM_015607, Hs. 323817 DKFZP547E DKFZP547E1010 protein AA406292, 101 AA418004 AA425759, Hs. 332633 BBS2 Bardet-Biedi syndrome 2 AA486738 Hs. 333495 DSS1 Deleted in split-hand/split-foot 1 region W79057, H85464 Hs. 334612 SNRPE small nuclear ribonucleoprotein polypeptide X12466, AA678021 Hs. 334787 MGC19556 hypothetical protein MGC19556 BE379431, AA609463 Hs. 342389 PPIA peptidylprolyl isomerase A (cyclophilin A) AW732921, H72674 AW675430, Hs.349961 RPL6 ribosomal protein L6 AA629808 ESTs, Weakly similar to CNG1_HUMAN Hs. 356525 FLJ12806 cGMP-gated cation channel alpha 1 (CNG BE044582, T73794 channel alpha 1) Hs. 3610 KIAA0205 KIAA0205 gene product D86960, R91263 ESTs, Highly similar to SMHU1B Hs.36102 R99207, H72722 metallothionein 1B [H.sapiens] Hs. 4 ADH1B alcohol dehydrogenase 1B (class I), beta M24317, N93428 polypeptide Z75311, H99196, Hs.41587 RAD50 RAD50 (S. cerevisiae) homolog AA884913, Hs. 431 BMI1 murine leukemia viral (bmi-1) oncogene AA608856, homolog T87514, W90704, AA478036 NM_006398, Hs.44532 UBD diubiquitin N33920 Hs. 44585 TP53BP2 tumor protein p53-binding protein, 2 AI123916, H69077, N34418 Hs. 46440 SLC21A3 solute carrier family 21 (organic anion U21943, N62948 transporter), member 3 BE278623, Hs.4756 FEN1 flap structure-specific endonuclease 1 AA620553 SMC4 (structural maintenance of AB019987, chromosomes 4, yeast-like 1 axa452095 dolichyl-phosphate mannosyltransferase AW173486, olypeptide 1, catalytic subunit AA004759 Hs. 52002 CD5L CD5 antigen-like (scavenger receptor NM_005894, cysteine rich family AA677254 Sjogren syndrome antigen A2 (60kD, NM_004600, Hs.554 SSA2 ribonucleoprotein autoantigen SS-A/Ro) AA010351 Hs.5662 GNB2L1 guanine nucleotide binding protein (G BE206815, protein), beta polypeptide 2-like 1 AA640657, R96220 Hs. 57101 MCM2 minichromosome maintenance deficient (S. BE250461, cerevisiae 2 (mitotin) AA454572 Hs. 5737 KIAA0475 KIAA0475 gene product AA524523, N73927 Hs. 57783 EIF3S9 eukaryotic translation initiation factor 3, U62583, AA676471 subunit 9 (eta, 116kD Hs. 6127 Homo sapiens cDNA: FLJ23020 fis, clone AA054768, T67278 LNG00943 Hs.6551 ATP6S1 ATPase, H+ transporting, lysosomal NM_001183, Hs.6551 ATP6S1 interacting protein 1 AA487588 AA702845, Hs.6650 VPS45B vacuolar protein sorting 45B (yeast homolog) AA885433 Hs. 6838 ARHE ras homolog gene family, member E W03441, W86282, AA443302 Hs. 695 CSTB cystatin B (stefin B) AI831499, 110374 BE386706, Hs. 699 PPIB peptidylprolyl isomerase B (cyclophilin B) N45313, AA481464 Hs. 69997 ZNF238 zinc finger protein 238 AJ223321, R79722 Hs. 74441 CHD4 chromodomain helicase DNA binding protein BE408958, N34372 Hs. 75117 ILF2 interleukin enhancer binding factor 2,45kD AA307289, AA894687, H95638 Hs. 75183 CYP2E cytochrome P450, subfamily IIE (ethanol- J02843, H50500 inducible Hs. 75187 KIAA0016 translocase of outer mitochondrial D13641, AA644550 membrane 20 east) homolo AA307460, Hs. 75258 H2AFY H2A histone family, member Y AA486003 Hs.75354 GCN1L1 GCN1 (general control of amino-acid D86973, R55250 synthesis 1, yeast)-like 1 Hs. 75412 ARMET arginine-rich, mutated in early stage tumors AA582041, R91550 Hs. 75424 ID1 inhibitor of DNA binding 1, dominant S78825, AA457158 negative helix-loop-helix protein Hs. 75546 CAPZA2 capping protein (actin filament) muscle Z- U03851, AA083228 line, alpha 2 Hs. 75659 MPV17 MpV17 transgene, murine homolog, NM 002437, lomerulosclerosis R55046 Hs. 75678 FOSB FBJ murine osteosarcoma viral oncogene L49169, T61948 homolo B Hs. 75981 USP14 ubiquitin specific protease 14 (tRNA-guanine NM005151, transglycosylase) AA039511, T65861 AW245775, AA828564, Hs. 76230 RPS10 ribosomal protein S10 AA828819, A1054003 Hs. 76285 DKFZP564B DKFZP564B167 protein A1032331, 167 AA621342 immunoglobulin J polypeptide, linker protein Hs. 76325 IGJ for immunoglobulin alpha and mu AW172754, polypeptides T90492, T70057 Homo sapiens, clone MGC: 24130 Hs. 7655 U2AF65 U2 small nuclear ribonucleoprotein auxiliary AA936430, factor 65kD AA405748 AB002323, Hs. 7720 DNCH1 dynein, cytoplasmic, heavy polypeptide 1 AA010589, W78967 Hs. 77254 CBX1 chromobox homolog 1 (HP1 beta homolog AL046741, Drosophila) AA448667 Hs. 77326 IGFBP3 insulin-like growth factor binding protein 3 BE336944, AA598601 Hs. 77608 SFRS9 splicing factor, arginine/serine-rich 9 AI021546, N47892, AA490721 Hs. 78065 C7 complement component 7 X86328, AA598478 Hs. 78902 VDAC2 voltage-dependent anion channel 2 A1015604, AA857093, T66813 Hs.79090 XPO1 exportin 1 (CRM1, yeast, homolog) D89729, T59055 AK000250, Hs.79110 NCL nucleolin AA433818 U38846, T98634, Hs.79150 CCT4 chaperonin containing TCP1, subunit 4 AA088226, (delta) AA598637 Hs. 79162 SSRP1 structure specific recognition protein 1 AI635077, R11356 Hs. 80343 MMP15 matrix metalloproteinase 15 (membrane- inserted) AW016451, Hs.80552 DPT dermatopontin R48303 hepatocyte growth factor (hepapoietin A; Hs.809 HGF X16323, R52797 scatter factor) Hs. 80917 AP3S1 adaptor-related protein complex 3, sigma 1 D63643, AA996044 subunit Hs. 80919 SYPL synaptophysin-like protein S72481, AA427447 Hs. 81972 SHC1 SHC (Src homology 2 domain-containing) X68148, R52960, transforming protein 1 T50498 Hs. 82043 D123 D123 gene product U27112, AA448289 Hs. 82159 PSMA1 proteasome (prosome, macropain) subunit, AI889267, R27585 alpha type, 1 Hs. 82793 PSMB3 proteasome (prosome, macropain) subunit, A1028114, beta type, 3 AA620580 Hs. 82916 CCT6A chaperonin containing TCP1, subunit 6A L27706, zeta 1 AA872690, H84286 small nuclear ribonucleoprotein polypeptides BE252108, Hs.83753 SNRPB B and B1 AA599116 Hs. 84790 C7orf14 chromosome 7 open reading frame 14 D86978, AA600190 AA160893 Hs.85119 SMT3H1 SMT3 (suppressor of mif two 3, yeast) AA862529, homolog 1 AA872379 Hs. 8765 RNAHP RNA helicase-related Protein AI814448, T56221 Hs. 8867 CYR61 cysteine-rich, angiogenic inducer, 61 Y12084, AA777187 Hs. 89525 HDGF hepatoma-derived growth factor (high-BE259164, s. mobility group protein 1-like) M453749 AB023420, Hs. 90093 HSPA4 heat shock 70kD protein 4 AA131267, AA433916 Y08999, Hs. 90370 ARPC1A actin related protein 2/3 complex, subunit 1A AA490209, (41 kD) AA016251, AA151930 Hs. 90744 PSMD11 proteasome (prosome, macropain) 26S AB003102 subunit, non-ATPase, 11 Hs. 99969 FUS fusion, derived from t (12 ; 16) malignant BE396632,101207 li osarcoma

In the third gene list, a total of 68 unique UniGenes from the 218 significant gene list (containing 213 unique UniGenes) identified herein (Table 1) were observed to overlap (Table 4), and the likelihood that the overlap would arise by chance if the two gene lists were totally independent was minuscule (Pc<1x10-8).

Table 4. Intersection of Significant Genes Identified Herein with HCC Genes UniGene Gene Description GenBank No. Identifier Hs.119651 GPC3 glypican 3 U50410, AA775872 Hs. 125180 Gar growth hormone receptor X06562, N70358 Hs. 180577 GRN granulin AI375908, AA496452 Hs.44585 TP53BP2 tumor protein p53-binding protein, 2 AI123916, H69077 BE336944, Hs.77326 IGFBP3 insulin-like growth factor binding protein 3 AA598601 Hs. 8867 CYR61 cysteine-rich, angiogenic inducer, 61 Y12084, M777187 CCT5 chaperonin containing TCP1, subunit 5 Hs. 1600 HSEC61 (epsilon) D43950, AA629692 sec 61 homolog heat shock 70kD protein 5 (glucose-AL043206, regulated protein, 78kD) AA962446 Hs.152931 LBR lamin B receptor L25931, AA099136 AB002323, Hs.7720 DNCH1 dynein, cytoplasmic, heavy polypeptide 1 W78967 BE278623, Hs.4756 FEN1 flap structure-specific endonuciease 1 AA620553 SMC4 (structural maintenance of chromosomes 4, yeast)-like 1 AB019987, CAP-C chromosome associated AA452095 polypeptide C Hs. 77254 CBX1 chromobox homolog 1 (HP1 beta homolog AL046741, s. Drosophila) M448667 Hs. 2934 RRM1 ribonucleotide reductase M1 polypeptide X59543, AA633549 AW404507, Hs. 156110 IGKC immunoglobulin kappa constant AA402920, AA486362 Hs. 20144 SCYA14 small inducible cytokine subfamily A (Cys- NM_004166, Ces), member 14 R96626 Hs. 237356 SDF1 stromal cell-derived factor 1 L36033, AA447115 Hs. 78065 C7 complement component X86328, AA598478Hs. 118638 NME1 non-metastatic cells 1, protein (NM23A) AA147871, expressed in AA644092 AF043937, Hs.12482 GNPAT glyceronephosphate O-acyltransferase AA486845 AW967351, Hs.174140 ACLY ATP citrate lyase H08547 cytochrome P450, subfamily IIC Hs. 174220 CYP2C8 (mephenytoin 4-hydroxylase), polypeptide M17398, N53136 8 glutathione transferase zeta 1 Hs. 26403 GSTZ1 maleylacetoacetate isomerase U86529, AA428334 Hs. 4 ADH1B alcohol dehydrogenase 1B (class I), beta M24317, N93428 polypeptide AW963733, Hs.177592 RPLP1 ribosomal protein, large, P1 AI732304 Hs. 25797 SF3B4 splicing factor 3b, subunit 4,49kD NM_005850, AA699361 Hs. 76230 RPS10 ribosomal protein S10 AW245775, Ait54003 IGJ immunoglobulin J polypeptide, linker Hs.76325 protein for immunoglobulin alpha and mu AW172754, SLU7 polypeptides T70057 step 11 splicing factor SLU7 Hs.83753 SNRPB small nuclear ribonucleoprotein BE252108, polypeptides B and B1 AA599116 Hs. 115617 CRHBP corticotropin releasing hormone-binding NM_001882, Protein AA286752 ADP-ribosylation factor guanine ARFGEF2 nucleotide-exchange factor 2 (brefeldin A- Hs. 118249 inhibited) AA099582, N34053 BIG2 Brefeldin A-inhibited guanine nucleotide- exchange protein Hs. 155079 PPP2R5A protein phosphatase 2, regulatory subunit AA234460, R59164 B(B56), alpha isoform protein kinase, DNA-activated, catalystic Hs.155637 PRKDC U34994, R27615 polypeptide NM_004428, Hs.1624 EFNA1 ephrin-A1 AA857015 Hs. 182278 CALM2 calmodulin 2 (phosphorylase kinase, D45887, AA043551 delta STK39 serine threonine kinase 39 (STE20/SPS1 Hs. 199263 SPAK homolog, yeast) F26137, H84871 ste-20 related kinase gamma-aminobutyric acid (GABA) A NM_004961, Hs.22785 GABRE receptor, epsilon H63532 AI674877, N34362, Hs.24950 RGS5 regulator of G-protein signalling 5 AA668470 AW779995, Hs.296341 CAP2 adenylyl cyclase-associated protein 2 AA040613 Hs. 81972 SHC1 SHC (Src homology 2 domain-containing) X68148, R52960, transforming protein 1 Hs. 1119 NR4A1 nuclear receptor subfamily 4, group A, NM002135, member 1 N94487 Hs. 1657 ESR1 estrogen receptor 1 AL078582, AA291702 Hs. 166891 RFX5 regulatory factor X, 5 (influences HLA AL050135, class II expression) AA418045 v-maf musculoaponeurotic fibrosarcoma Hs.252229 MAFG (avian) oncogene family, protein G AF059195, N21609 Hs.25647 FOS v-fos FBJ murine osteosarcoma viral V01512, N36944, oncogene homolog AA485377 AA884913, Hs 431 BMI1 murine leukemia viral (bmi-1) oncogene W90704, homolog AA478036 interleukin enhancer binding factor 2, AA307289, Hs.75117 ILF2 45kD H95638, AA894687 Hs. 75678 FOSBFBJ murine osteosarcoma viral oncogene L49169, T61948 homolo B Hs. 6551 ATP61P1 ATPase, H+ transporting, lysosomai NM 001183, interacting protein 1 AA487588 Hs. 79090 XP01 exportin 1 (CRM1, yeast, homolog) D89729, T59055 Hs. 80917 AP3S1 adaptor-related protein complex 3, sigma D63643, AA996044 1 subunit Hs.80919 SYPL synaptophysin-like protein S72481, AA427447 NM_006398N3392 Hs.44532 UBD diubiquitin 0 Hs. 106061 RDBP RD RNA-bindin rotein X16105, AA056390 C1orf9 chromosome 1 open reading frame 9 Hs.108636 CH1 membrane protein CH1 BE466870, N36176 Hs. 110713 DEK DEK oncogene (DNA binding) AI888504, R25377 Hs. 11355 TMPO Thymopoietin U09087, T63980 ESTs MAWD binding protein Hs. 16341 MAWBP ESTs weakly similar to predicted using Al866254, R54416 enefinder [C. elegans Hs. 16426 PODXL podocalxin-like BE395330, N64508 aldehyde dehydrogenase 8 family, Hs. 18443 ALDH8A1 member A1 A1051566, N70701 ESTs Hs.194673 PEA15 phosphoprotein enriched in astrocytes 15 Y13736, AA293211 AW952494, hypothetical protein FLJ12666 Hs. 23767 FLJ12666 Homo sapiens cDNA FLJ12666 fis, clone AA115300, NT2RM4002256 M131466 Hs. 291904 DXS1357E accessory proteins BAP31/BAP29 Z31696, AA625628 Hs. 3610 KIAA0205 KIAA0205 gene product D86960, R91263 ESTs, Highly similar to SMHU1 B Hs. 36102 metallothionein 1 B [H. sapiens] R99207, H72722 ESTs highly similar to MT1 B Human Metallothionein-IB [H.sapiens] W03441, Hs.6838 ARHE ras homolog gene family, member E AA443302 translocase of outer mitochondrial TOMM20-membrane 20 (yeast) homolog Hs. 75187 PENDI KIAA0016 translocase of outer D13641, AA644550 mitochondrial membrane 20 (yeast) homolog AI814448, T56221, Hs. 8765 RNAHP RNA helicase-related protein N55459

The discriminator cassettes were assessed on an independent tissue set of 58 liver clinical biopsies from 29 patients. Using a kNN prediction algorithm, it was found that all classifier probe cassettes could readily distinguish HCC tumor from non-tumor liver (Table 5), and that the gene discriminators of tumor vs. non-tumor in HCC derived by the intersect analysis of limited tissue sets can be validated in an independent manner. Table 5. Prediction accuracy of gene classifiers using k NN algorithm on 58 liver biopsies from 29 patients. Gene No. of MiscIassiScation No. of No. of false Predictive classifiers gene classifiers rate false negative positive cases* accuracy cases** Table 1 218 4 of 58 4--93% Table 2 265 3 of 58 3 - 95% Table 3 166 3 of 58-3--95% Table 4 68 2 of 58 1 1 96% * False negative cases refer to HCC tumors which were misclassified as non-tumor livers.

**False positive cases refer to non-tumor livers which were misclassified as HCC tumors.

REFERENCES The following publications are incorporated by reference herein.

Balsara, B. R. , Pei, J., de Rienzo, A. , Simon, D. , Tosolini, A. , Lu, Y. Y. , Shen, F-M. , Fan, X., Lin, W-Y., Buetow, K. H. , London, W. T. , and Testa, J. R.

(2001). Human hepatocellular carcinoma is characterized by a highly consistent pattern of genomic imbalances, including frequent loss of 16q23. 1- 24.1. Gene Chrom. Cancer 30,245-253.

Bea, S., Tort, F., Pinyol, M. , Puig, X., Hernandez, L. , Hernandez, S. , Fernandez, P. L. , van Lohuizen, M. , Colomer, D. , and Campo, E. (2001).

BMI-1 gene amplification and overexpression in hematological malignancies occur mainly in mantle cell lymphomas. Cancer Res. 61,2409-2412.

Bosl, G. J. , and Head, M. D. (1994). Serum tumor marker half-life during chemotherapy in patients with germ cell tumors. Int. J. Biol. Markers 9,25-28.

Brown, P. O., and Botstein, D. (1999). Exploring the new world of the genome with DNA microarrays. Nat. Genet. 21 (suppl. ), 33-37.

Carver, R. S., Sliwkowski, M. X., Sitaric, C. , and Russell, W. E. (1996). Insulin regulates heregulin binding and ErbB3 expression in rat hepatocytes. J. Biol.

Chem. 271,13491-13496.

Chen, X. , Cheung, S. T. , So, S. , Fan, S. T. , Barry, C. , Higgins, J. , Lai, K-M.,

Ji, J. , Dudoit, S. , Ng, 1. O. L. , van de Rijn, M. , Botstein, D. , and Brown, P. O.

(2002). Gene expression patterns in human liver cancers. Mol. Biol. Cell 13, 1929-1939.

Davison, A. C. , and Hinkley, D. V. (1997). Bootstrap methods and their application (Cambridge: Cambridge University Press).

Eberwine, J. , Yeh, H., Miyashiro K. , Cao, Y. , Nair, S. , Finnell, R., Zettel, M. , and Coleman, P. (1992). Analysis of gene expression in single live neurons.

Proc. Natl. Acad. Sci. USA 89,3010-3014.

Eisen, M. B. , Spellman, P. T. , Brown, P. O., and Botstein, D. (1998). Cluster analysis and display of genome-wide expression patterns. Proc. Natl. Acad.

Sci. USA 95, 14863-14688.

Filmus, J. , and Selleck, S. B. (2001). Glypicans: proteoglycans with a surprise. J. Clin. Invest. 108,497-501.

Fink, L. , Seeger, W., Ermert, L. , Hanze, J. , Stahl, U. , Grimminger, F., Kummer, W. , and Bohle, R. M. (1998). Real-time quantitative RT-PCR after laser-assisted cell picking. Nat. Med. 4, 1329-1333.

Gonzalez, F. , Delahodde, A. , Kodadek, T. , and Johnston, S. A. (2002).

Recruitment of a 19S proteasome subcomplex to an activated promoter.

Science 296,548-550.

Grimberg, A. , and Cohen, P. (2000). Role of insulin-like growth factors and their binding proteins in growth control and carcinogenesis. J. Cell. Physiol.

183,1-9.

Gucev, Z. S. , Oh, Y., KelleyK. M. , and Rosenfeld, R. G. (1996) Insulin-like growth factor binding protein 3 mediates retinoic acid-and transforming growth factor (32-induced growth inhibition in human breast cancer cells.

Cancer Res. 56,1545-1550.

Honda, M. , Kaneko, S. , Kawa, H., Shirota, Y. , and Kobayashi, K. (2001).

Differential gene expression between chronic hepatitis B and C hepatic lesion.

Gastroenterology 120,955-966.

Huynh, H. , Chow, P. K H. , Ooi, L. L. P. , and Soo K-C. (2002). A possible role for insulin-like growth factor-binding protein-3 autocrine/paracrine loops in controlling hepatocellular carcinoma cell proliferation. Cell Growth Diff. 13, 115-122.

Irniger, S. , and Nasmyth, K. (1997). The anaphase-promoting complex is required in Gl arrested yeast cells to inhibit B-tyoe cyclin accumulation and to prevent uncontrolled entry into S-phase. J. Cell Sci. 110, 1523-1531.

Ito, Y. , Takeda, T. , Sakon, M. , Tsujimoto, M. , Higashiyama, S. , Noda, K., Miyoshi, E. , Monden, M. , and Matsuura, N. (2001). Expression and clinical significance of erb-B receptor family in hepatocellular carcinoma. Br. J.

Cancer 84, 1377-1383.

Johnson, P. J. (2001). The role of serum alpha-fetoprotein estimation in the diagnosis and management of hepatocellular carcinoma. Clin. Liver Dis. 5, 145-159.

Lee, W. M. (1997). Hepatitis B virus infection. New Engl. J. Med. 337,1733- 1745.

Lee, J. -S., and Thorgeirsson, S. S. (2002). Functional and genomic implications of global gene expression profiles in cell lines from human hepatocellular cancer. Hepatology 35, 1134-1143.

Lessard, J. , and Sauvageau, G. (2003). Bmi-1 determines the proliferative capacity of normal and leukaemic stem cells. Nature 423,255-260.

Leung, T. H. Y. , Wong, N. , Lai, P. B. S. , Chan, A. , To, K-F., Liew, C. T. , Lau, W-Y., and Johnson, P. J. (2002). Identification of four distinct regions of allelic imbalances on chromosome 1 by the combined comparative genomic hybridization and microsatellite analysis on hepatocellular carcinoma. Mod.

Pathol. 15,1213-1220.

Mallery, D. L., Vandenberg, C. J. , and Hiom, K. (2002). Activation of the E3 ligase function of the BRCAlBARDl complex by polyubiquitin chains. EMBO J. 21,6755-6762.

Marchio, A. , Meddeb, M. , Pineau, P. , Danglot, G. , Tiollais, P., Bernheim, A. , and Dejean, A. (1997). Recurrent chromosomal abnormalities in hepatocellular carcinoma detected by comparative genomic hybridization.

Genes Chrom. Cancer 18,59-65.

Midorikawa, Y. , Ishikawa, S., Iwanari, H. , Imamura, T. , Sakamoto, H. , Miyazono, K. , Kodama, T., Makuuchi, M. , and Aburatani, H. (2003). Glypican- 3, overexpressed in hepatocellular carcinoma, modulates FGF2 and BMP-7 signaling. Int. J. Cancer 103,455-465.

Murthy, S. S. , Shen, T. , de Rienzo, A. , Lee, W. C., Ferriola, P. C. , Jhanwar, S.

C. , Mossman, B. T. , Filmus, J. , and Testa, J. R. (2000). Expression of GPC3, an X-linked recessive overgrowth gene, is silenced in malignant mesothelioma. Oncogene 19, 410-416.

Nawrocki, S. T., Bruns, C. J. , Harbison, M. T. , Bold, R. J., Gotsch, B. S., Abbruzzese, J. L. , Elliott, P., Adams, J. , and McConkey, D. J. (2002). Effects of the proteasome inhibitor PS-341 on apoptosis and angiogenesis in orthotopic human pancreatic tumor xenografts. Mol. Cancer Therapeutics 1, 1243-1253.

Okabe, H. , Satoh, S. , Kato, T. , Kitahara, O., Yanagawa, R., Yamaoka, Y. ,

Tsunoda, T. , Furukawa, Y. , and Nakamura, Y. (2001). Genome-wide analysis of gene expression in human hepatocellular carcinomas using cDNA microarray : identification of genes involved in viral carcinogenesis and tumor progression. Cancer Res. 61,2129-2137.

Orlando, V. (2003). Polycomb, epigenomes and control of cell identity. Cell, 112,599-606.

Park, ML, Qian, D. , Kiel, M. , Becker, M. W. , Pihalja, M. , Weissman, 1. L. , Morrison, J. , and Clarke, M. F. (2003). Bmi-1 is required for maintenance of adult self-renewing haematopoietic stem cells. Nature 423,302-305.

Parkin, D. M. , Pisani, P. , and Ferlay, J. (1999). Global cancer statistics. CA Cancer J. Clin. 49,33-64.

Pickart, C. M. (2001). Ubiquitin enters the new millennium. Mol. Cell. 8, 499- 504. <BR> <BR> <BR> <BR> <BR> <BR> <BR> <P>Raaphorst, F. M. , van Kemenade, F. J., Blokzijl, T. , Fieret, E. , Hamer, K. M., Satijn, D. P. E., Otte, A. P. , and Meer, C. J. L. M. (2000). Coexpression of BMI-1 and EZH2 polycomb group genes in Reed-Sternberg cells of Hodgkin's disease. Am. J. Pathol. 157,709-715.

Riese, D. J. II, and Stern, D. F. (1998). Specificity within the EGF family/ErbB receptor family signaling network. BioEssays 20,41-48.

Saikali, Z. , and Sinnett, D. (2000). Expression of glypican 3 (GPC3) in embryonal tumors. Int. J. Cancer. (Pred Oncol) 89,418-422.

Sakamoto, K. M. (2002). Ubiquitin-dependent proteolysis: its role in human diseases and the design of therapeutic strategies. Mol. Genet. Metab. 77,44- 56.

Salomon, P. , and Pnadolfi, P. P. (2002). p53 de-ubiquitination: at the edge

between life and death. Nat. Cell Biol. 4, E152-E153.

Salghetti, S. , Caudy, A. A. , Chenoweth, J. G. , and Tansey, W. P. (2001).

Regulation of transcriptional activation domain function by ubiquitin. Science 293,1651-1653.

Shang, Y., Baumrucker, C. R. , and Green, M. H. (1999). Signal relay by retinoic acid receptors a and a in the retinoic acid-induced expression of insulin-like growth factor-binding protein-3 in breast cancer cells. J. Biol.

Chem. 274, 18005-18010.

Shirota, Y. , Kaneko, S. , Honda, M. , Kawai, H. F. , and Kobayashi, K. (2001).

Identification of differentially expressed genes in hepatocellular carcinoma with cDNA microarrays. Hepatology 33, 832-840.

Sotiriou, C. , Khanna, C. , Jazaeri, A. A., Petersen, D. , and Liu, E. T. (2002).

Core biopsies can be used to distinguish differences in expression profiling by cDNA microarrays. J. Mol. Diag. 4,30-36.

Tackels-Horne, D. , Goodman, M. D. , Williams, A. J. , Wilson, D. J. , Eskandari, T. , Vogt, L. M. , Boland, J. F. , Scherf, U. , and Vockley, J. G. (2001).

Identification of differentially expressed genes in hepatocellular carcinoma and metastatic liver tumors by oligonucleotide expression profiling. Cancer 92,395-405.

The Gene Ontology Consortium (2000). Gene Ontology: tool for the unification of biology. Nature Genet. 25,25-29. (World Wide Web URL : http://www. geneontology. org) Toretsky, J. A. , Zitomersky, N. L. , Eskenazi, A. E. , Voigt, R. W. , Strauch, E. D.

Sun, C. C. , Huber, R. , Meltzer, S. J. , and Schlessinger, D. (2001). Glypican-3 expression in Wilms tumor and hepatoblastoma. J. Pediatr. Hematol. Oncol.

23,496-499.

Varambally, S. , Dhanasekaran, S. M. , Zhou, M. , Barrette, T. R. , Kuma-Sinha, C. , Sanda, M. G. , Ghosh, D. , Pienta, K. J., Sewalt, R. G. A. B., Otte, A. P., Rubin, M. A. , and Chinnaiyan, A. M. (2002). The polycomb group protein EZH2 is involved in progression of prostate cancer. Nature 419,624-629.

Vonlanthen, S. , Heighway, J., Altermatt, H. J., Gugger, M. , Kappeler, A., Borner, M. M. , van Lohuizen, M. , and Betticher, D. C. (2001). The bmi-1 oncoprotein is differentially expressed in non-small cell lung cancer and correlates with INK4A-ARF locus expression. Br. J. Cancer 84, 1372-1376.

Wong, N. , Lai, P. Lee, S. W. , Fan, S. , Pang, E. , Liew, C. T., Sheng, Z. , Lau, J.

W. , and Johnson, P. J. (1999). Assessment of genetic changes in hepatocellular carcinoma by comparative genomic hybridization analysis: relationship to disease stage, tumor size, and cirrhosis. Am. J. Pathol. 154, 37-43.

Xiang, Y. Y. , Ladeda, V. , and Filmus, J. (2001). Glypican-3 expression is silenced in human breast cancer. Oncogene 20, 7408-7412.

Xu, L. , Hui, L. , Wang, S. , Gong, J. , Jin, Y. , Wang, Y. , Ji, Y. , Wu, X. , Han, Z., and Hu, G. (2001a). Expression profiling suggested a regulatory role of liverenriched transcription factors in human hepatocellular carcinoma. Cancer Res. 61,3176-3781.

Xu, X-R. , Huang, J. , Xu, Z-G., Qian, B-Z. , Zhu, Z-D. , Yan, Q. , Cai, T., Zhang, X. , Xiao, H-S. , Qu, J. , Liu, F. , Huang, Q-H. , Cheng, Z-H. , Li, N-G. , Du. J-J., Hu, W., Shen, K-T. , Lu, G. , Fu, G. , Zhong, M. , Xu, S-H. , Gu, W-Y. , Huang, W. , Zhao, X-T. , Hu, G-X. , Gu, J-R. , Chen, Z. , and Han, Z-G. (2001b). Insight into hepatocellular carcinogenesis at transcriptome level by comparing gene expression profiles of hepatocellular carcinoma with those of corresponding noncancerous liver. Proc. Natl. Acad. Sci. USA 98,15089-15094.

Yu, A. S. , and Keeffe, E. B. (2003). Management of hepatocellular carcinoma. Rev. Gastroenterol. Disord. 3, 8-24.

Zhang, N. , and Deuel, T. F. (1999). Pleitrophin and midkine, a family of mitogenic and angiogenic heparin-binding growth and differentiation factors.

Curr. Opin. Hemato. 6,44-50.

Zimonjic, D. B., Keck, C. L. , Thorgeirsson, S. S. , and Popescu, N. C. (1999).

Novel recurrent genetic imbalances in human hepatocellular carcinoma cell lines identified by comparative genomic hybridization. Hepatology 29,1208- 1214.

Cheung V. G. , etal., 1999. Nature Genetics Supplement, 21: 15-19.

Lipshutz R. J., et al., 1999. Nature Genetics Supplement, 21: 20-24.

Bowtell D. D. L. , 1999. Nature Genetics Supplement, 21: 25-32.

Schweitzer B., et al., 2002. Nature Biotechnol., 20: 359-365.

HallowayA. J. , et al., 2002. Nature Genetics Supplement, 32: 481-489.

Martzen M. R., McCraith S. M. , Spinelli S. L. , Torres F. M. , Fields S., Grayhack E. J., and Phizicky E. M. , 1999. Science, 286: 1153.

G. MacBeath and S. L. Schreiber,"Printing Proteins as Microarrays for High- Throughput Function Determination,"Science 289 (5485): 1760-1763,2000.

ACKNOWLEDGEMENT Chon Kar Leow contributed to the studies by providing HCC liver tissue samples to the inventors.

OTHER EMBODIMENTS Although various embodiments of the invention are disclosed herein, many adaptations and modifications may be made within the scope of the invention in accordance with the common general knowledge of those skilled in this art. Such modifications include the substitution of known equivalents for any aspect of the invention in order to achieve the same result in substantially the same way. Accession numbers, as used herein, may refer to Accession numbers from multiple databases, including GenBank, the European Molecular Biology Laboratory (EMBL), the DNA Database of Japan (DDBJ), or the Genome Sequence Data Base (GSDB), for nucleotide sequences, and including the Protein Information Resource (PIR), SWISSPROT, Protein Research Foundation (PRF), and Protein Data Bank (PDB) (sequences from solved structures), as well as from translations from annotated coding regions from nucleotide sequences in GenBank, EMBL, DDBJ, or RefSeq, for polypeptide sequences. Accession numbers, as used herein, may also refer to Accession numbers from databases such as UniGene, OMIM, LocusLink, or HomoloGene. Numeric ranges are inclusive of the numbers defining the range. In the specification, the word"comprising"is used as an open-ended term, substantially equivalent to the phrase"including, but not limited to", and the word"comprises"has a corresponding meaning. Citation of references herein shall not be construed as an admission that such references are prior art to the present invention. All publications are incorporated herein by reference as if each individual publication were specifically and individually indicated to be incorporated by reference herein and as though fully set forth herein. The invention includes all embodiments and variations substantially as hereinbefore described and with reference to the examples and drawings.