Skip Navigation


Carcinogenesis Advance Access originally published online on April 2, 2007
Carcinogenesis 2007 28(7):1552-1560; doi:10.1093/carcin/bgm075
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
28/7/1552    most recent
bgm075v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (1)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Budhu, A
Right arrow Articles by Wang, X.W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Budhu, A
Right arrow Articles by Wang, X.W.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Published by Oxford University Press 2007.

Induction of a unique gene expression profile in primary human hepatocytes by hepatitis C virus core, NS3 and NS5A proteins

A Budhu1, Y Chen2, J.W. Kim1,*, M Forgues1, K Valerie3, C.C. Harris1 and X.W. Wang1,4

1 Laboratory of Human Carcinogenesis
2 Genetics Branch, Center for Cancer Research, National Cancer Institute, NIH, Bethesda, MD 20892, USA
3 Department of Radiation Oncology, Medical College of Virginia, Virginia Commonwealth University, Richmond, VA 23298-0058, USA
* Present address: Center for Human Genomics, Wake Forest University School of Medicine, Medical Center Blvd. Winston-Salem, NC 27157, USA

4 To whom correspondence should be addressed. Tel: +301 496 2099; Email: xw3u{at}nih.gov; Fax: 301 496 0497


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 Supplementary material
 References
 
Hepatocellular carcinoma (HCC) is a fatal disease and hepatitis B and C viruses (HBV and HCV) are considered as major causative factors for the development of HCC. We have conducted gene expression profiling studies to search for potential target genes responsible for HCV-mediated HCC. Adenoviruses encoding core (HCV structural protein), NS3 and NS5A [HCV non-structural (NS) proteins] were generated and infected individually or together in freshly isolated primary human hepatocytes. An adenovirus harboring the oncogenic HBV protein, HBx, was included for comparison. A microarray platform of over 22 000 human oligos was analyzed to seek out significant differentially expressed genes among these viral proteins. We also compared these gene expression profiles with those obtained from HCV-infected liver samples from chronic liver disease (CLD) patients and HCV-related HCC. We found that HCV-related proteins largely induce unique genes when compared with HBx. In particular, interferon-inducible gene 27 (IFI27) was highly expressed in HCV or core-infected hepatocytes and HCV-related CLD or HCC, but was not significantly expressed in HBx-infected hepatocytes or HBV-related CLD or HCC, indicating that IFI27 may play a role in HCV-mediated HCC. In conclusion, our results suggest that HBV and HCV promote HCC development mainly through different mechanisms.

Abbreviations: ß-gal, ß-galactosidase; CLD, chronic liver disease; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; IFI27, interferon-inducible gene 27; MOI, multiplicity of infection; NS, non-structural; qRT–PCR, quantitative reverse transcription–polymerase chain reaction


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 Supplementary material
 References
 
Liver cancer ranks as the third most deadly and fifth most common cancer worldwide, with a high prevalence in Asia and sub-Saharan Africa (1). Although the prevalence of liver cancer is relatively low in Europe and North America, its incidence has substantially increased in recent years. Hepatocellular carcinoma (HCC) is the most frequent primary cancer of the liver and is a long-term process that is considered a fatal disease because of its poor prognosis (2). Despite many studies of HCC, information regarding phenotypic and molecular changes associated with the development of this disease is still limited. HCC, however, is one of the human cancers where the underlying etiology can often be identified. Some factors associated with HCC include smoking, alcohol consumption and aflatoxin B1 exposure, which induces G:C to T:A transversions at the third base in codon 249 of p53 (3). Several lines of evidence indicate that chronic infection with hepatitis viruses, hepatitis B virus (HBV) and hepatitis B virus (HCV) are the major risk factors for HCC development and viral co-infection and synergistic effects have been suggested (4). Vaccination is available for HBV, but not for HCV, although studies involving antibody-mediated viral neutralization using HCV-like particles shows promise in characterizing anti-HCV antibodies (5). Despite a strong association between hepatitis infection and HCC, the specific roles of these viruses in the pathogenesis of HCC, whether direct or indirect, remain controversial (6). Of the seven peptides encoded by the DNA virus HBV, the viral transactivator HBx seems to play the most significant role in HCC development (7). In contrast, the contribution of the 10 structural and non-structural (NS) proteins encoded by the RNA virus HCV to HCC are not clearly understood, although the core protein appears to have oncogenic properties (8,9). To understand the molecular mechanisms underlying the development of HCC and to delineate the role of HCV- and HBV-encoded genes in this process, we analyzed gene expression profiles in freshly isolated primary human hepatocytes infected with adenoviruses harboring HCV-encoded genes or HBx. These studies have shown that HCV and HBV affect mainly distinct cellular pathways and have allowed us to identify potential genes that may play a specific role in HCV-related HCC.


   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 Supplementary material
 References
 
Generation of FLAG-HCV proteins and adenovirus
Since the availability of antibodies are limited, flag-tagged HCV proteins were generated from polymerase chain reaction (PCR) amplification of the appropriate regions of pCV-J4l6S, an infectious cDNA clone of HCV genotype 1b, followed by restriction digest and subcloning to pCMV-TAG. Flag-tagged constructs were inserted into the adenoviral expression vector pZERO-TG. Adenovirus generation and purification were conducted by the Massey Cancer Center Virus Vector Shared Resource, Virginia Commonwealth University. Viral amplification was conducted by the Gene Therapy Center Virus Vector Core Facility (The University of North Carolina, Chapel Hill). Adenovirus expressing HBx was described previously (10,11). Detailed subcloning is presented in Supplemental Methods (available at Carcinogenesis Online).

Cell lines and infection
HHT-4 cells were originally derived from a liver transplant donor. Freshly isolated primary hepatocytes from this patient were infected with a replication-defective retrovirus encoding the human telomerase, hTERT. An immortalized clone (HHT-4) was obtained and propagated for >100 population doublings. These cells were not tumorigenic in nude mice and had a near-diploid karyotype (data not shown). HHT-4 cells were grown in (HBM) media supplemented with SingleQuots (Cambrex, Walkersville, MD) and 10% chemically denatured serum (Biofluids, Rockville, MD) and seeded on fibronectin-coated plates. Primary human hepatocytes were freshly isolated from six independent patient donors, HH1019, HH1024, HH1033, HH1046, HH1072 or HH1332, and were managed through the Liver Tissue Procurement and Distribution System at The University of Pittsburgh. Cells were washed twice in 1x PBS and re-suspended in HBM media, followed by trypan blue dye exclusion viability assessment. Cells (7.8 x 106) were seeded on 100 cm2 fibronectin-coated plates and infected with viral stocks 24 h later. In microarray studies, five biological replicates were performed for each viral infection in donor HH1072. After 24 h infection, RNA was extracted by Trizol (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. The concentration of RNA was assayed using a NanoDrop (version 3.1.0; NanoDrop Technologies, Wilmington, DE) and quality was assessed by a 2100 Bioanalyzer (Agilent Technologies, Foster City, CA) according to the manufacturer's instructions. Only RNA with a 260:280 ratio >1.8 was used.

Western blot
Following RNA isolation, protein was isolated according to the manufacturer's protocol. Protein concentration was determined by the Bradford assay according to the manufacturer's protocol (Bio-Rad, Hercules, CA). Equivalent total protein amounts were loaded in each lane of a 16% PAGE gel, followed by transferring and blocking in 5% non-fat dry milk. Blots were probed with anti-M2 antibody (Sigma, St Louis, MO), followed by incubation with horseradish peroxidase-conjugated secondary antibody. Membranes were stripped and reprobed with anti-HA antibody (Roche, Indianapolis, IN) to detect HBx expression. Antibody–antigen complexes were detected by enhanced chemiluminescence according to the manufacturer's protocol (Amersham, Piscataway, NJ).

cDNA and oligo array
Human cDNA and oligo microarrays (version 2.0, Qiagen, Valencia, CA) were generated by the National Cancer Institute (Bethesda, MD) microarray facility at the Advanced Technology Center (ATC). The oligo array platform contains 22 149 70-mer probes that map to 9 799 unique UniGene clusters, whereas the cDNA platform consists of 9180 cDNA clones that map to 8281 unique Unigene clusters. Detailed hybridization, quality control, data acquisition and filtering were performed as described previously (12). For each experiment, fluorescent probes were prepared by an indirect labeling approach of a common reference RNA pool from HHT-4 cells (Cy5) and from total RNA isolated from each infected primary hepatocyte sample (Cy3).

Analysis and statistics
After data extraction, microarray data were normalized via non-linear regression by in-house software (MATLAB, Natick, MA). Statistical characteristics, such as average quality assessment and confidence interval, were also evaluated with this program. Processed data were archived in a FileMaker Pro database (Santa Clara, CA) for significant gene query tasks. Further analysis was performed using the BRB ArrayTools (version 3.1) as described previously (13). This Excel-based platform contains several statistical tools including hierarchical clustering, class comparison and class prediction. Pathway Assist software (version 2.53) was utilized as a demo version (Stratagene, La Jolla, CA). It utilized a database describing >100 000 events of regulation, interaction and modification among 15 000 proteins, cell processes and small molecules based on Pubmed references. Imported unique gene lists for core, NS3, NS5A or HBx were analyzed in Pathway Assist using a ResNet2.5 Mammalian Database to search for molecular interaction networks. Parameters were restricted by searching for only common targets or regulators and filtering for only protein functional classes. Imported up-regulated gene lists for core, NS3, NS5A and down-regulated gene lists for HBx were also analyzed for significant functional pathways using Ingenuity Pathway Analysis (version 3.1, http://www.ingenuity.com) (Redwood City, CA). The analyses were based on the most up-to-date Ingenuity Pathways Knowledge Base of over 20 000 mammalian genes and proteins, 1.4 million biological interactions and 100 canonical pathways incorporating over 6000 discreet gene concepts. Significance was assessed based on the number of gene hits in a pathway compared with the total number of genes assigned to a specific pathway with P < 0.05.

Reverse transcription–PCR
PCR primers and probes (Assays on Demand) were purchased (Applied Biosystems, Foster City, CA) and quantitative reverse transcription (qRT)–PCRs were performed according to the manufacturer's instructions. Human 18S RNA labeled with VICTM reporter dye was used as an endogenous control. Reactions were performed with the ABI PRISM 7700 Sequence Detector System (Applied Biosystems).


   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 Supplementary material
 References
 
Generation and expression of adenovirus harboring HCV proteins
Although HCV is a known risk factor for HCC, the mechanism by which it causes this disease is unknown. We utilized freshly isolated adult normal primary human hepatocytes to study the gene expression changes associated with HCV infection. Since human hepatocytes have very low proliferative capacity, adenoviral vectors encoding N-terminal Flag-tagged-core, NS3 or NS5A were constructed to ensure high efficiency of HCV protein expression. To determine optimal infection conditions, primary human hepatocytes from several independent donors (HH1019, HH1033, HH1024, HH1046, HH1072) were infected with adenovirus expressing core, NS3 or NS5A at multiplicity of infection (MOI) ranging from 5 to 100 for 24 h (Figure 1a). An MOI of 10 resulted in efficient expression of HCV-encoded genes as determined by western blot and was used in further experiments. No visible cytotoxic effects or morphological changes of primary hepatocytes were observed at this MOI (data not shown). To analyze the gene expression profiles associated with these three HCV proteins, freshly isolated human hepatocytes from donor HH1072 were infected, either individually (Figure 1b, lanes 2–4) or in combination (HCV) (Figure 1b, lane 6), with adenovirus harboring core, NS3 or NS5A. An adenovirus encoding HBx was included for comparison, either as an individual infection or in combination with the HCV proteins described above (denoted HBx or HCV + HBx, respectively) (Figure 1b, lanes 5 and 7). Hepatocytes infected with an adenovirus harboring ß-galactosidase (ß-gal) were used as a control (Figure 1b, lane 1). Five individual infections for each viral gene were performed separately as biological replicates.


Figure 1
View larger version (20K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1. Adenovirus-mediated expression of HCV- and HBx-encoded genes in freshly isolated primary human hepatocytes. (a) Primary human hepatocytes (HH1072) were infected with adenovirus harboring HCV core, NS3 or NS5A at 5–100 MOI for 24 h. (b) Primary human hepatocytes were infected with adenovirus harboring HCV core, NS3 or NS5A, a combination of these three proteins (HCV), HBx or HCV + HBx at 10 MOI for 24 h. Asterisks represent non-specific bands. (c) Schematic of gene search strategies for significant and differentially expressed genes in each group. The statistical significance of differentially expressed genes in each group was assessed by three thresholds. Invariant genes were filtered by applying 99% confidence intervals based on either self/self-hybridizations, replicated spots on the array, or performance of biologically replicated experiments. Significant genes were further filtered based on the difference between each sample treatment and ß-gal control. Genes passing the criteria for each method were combined and used for further analysis. (d). Characterization of unique genes in core, NS3, NS5A or HBx. The total number of unique genes (total bar: orange + blue) in each infection group was stratified by expression ratio differences as described in the text to determine the number of up- or down-regulated genes (orange portions or blue portions of total bar, respectively) (e) Multidimensional scaling of core, NS3 NS5A and HBx.

 
A search strategy for statistically significant and differentially expressed genes
Isolated transcripts from each HCV- or HBV-related viral infection were analyzed using a human oligo array that represented 22 149 genes. TERT-immortalized normal human hepatocytes (HHT-4) were used as reference RNA for all array hybridizations. Five independent infections were performed as biological replicates for each virus (core, NS3, NS5A, HBx, HCV + HBx) and ß-gal. The objective of the search criteria was to find differentially expressed genes that were due to the expression of viral hepatitis proteins, while excluding variable genes due to expression of ß-gal. Since each hybridization was performed between infected cells and uninfected cells (HHT-4), differentially expressed genes could easily be determined by comparing a threshold of expression ratios. We chose the thresholding method over the commonly employed t-test technique because of the instability of t-statistics when only a few biological replicates are available (five in this case). To demonstrate the power of the significance assessment by the thresholding method, we determined 99% confidence intervals (expression ratio thresholds) from three independent self-hybridizations of HHT-4 (upper-limit 1.30, 1.58 and 1.42). A query of genes with expression ratio fold changes >1.42 (average upper-limit) in all three self-hybridization yielded no significant genes.

Three 99% confidence intervals for identifying differentially expressed genes were established based on three different invariance conditions (Figure 1c). The first condition was based on expression ratios from HHT-4 self-hybridizations, which provides the lower bond estimation of confidence intervals without the presence of spot-to-spot and array-to-array variations. The second condition utilized expression ratios from replicated spots (spots with the same oligo sequence) from each array that are assumed to produce the same expression ratio, and therefore, account for spot-to-spot variation. The third condition used expression ratios to exclude outlier genes across arrays within the same infection group (biological replicates), which accounted for array-to-array variation. To simplify the search, we took the average threshold within these three estimation methods, 0.522, 0.639 and 0.865 (least to highest stringency, all in log2 scale), respectively.

As illustrated in Figure 1c, three different gene lists were generated based on two criteria for each threshold level, one based on the three estimation methods described above and the other based on expression relative to ß-gal. First, a gene was selected if its expression ratio was beyond the 99% confidence interval (outlier) in at least four out of five samples within the same infection group, but no more than two out of five samples from the ß-gal infection group. Second, the average expression ratio for that gene in the infected group had to be at least 1.5-fold different from that of ß-gal group, whose average gene expression change was 1.3. We then combined the significant and differentially expressed genes from all six infection groups. The resulting number of genes was 365, 304 and 207, from least to most stringent conditions, respectively. These genes were then stratified by generating a high-confidence list (genes satisfying at least three of three threshold criteria, with 146 genes), mid-confidence list (genes satisfying at least two of three, with 275 genes) and low-confidence list (genes satisfying at least one of three, with 421 genes).

Analysis of the gene lists demonstrated that all the cellular genes affected by individual expression of core, NS3 or NS5A were also affected when these proteins were combined (HCV) (data not shown). However, these genes only accounted for ~30% of the significant genes in the HCV group. Since we only included three HCV-encoded genes in our study, the cellular genes affected by this combination was not representative of those affected by the entire HCV genome, but rather distinguishes pathways affected by each of these proteins. Consequently, we also generated three gene lists from four infection groups (core, NS3, NS5A and HBx) that excluded combinatorial viral expression and yielded 204 genes (Figure 1c).

Characterization of the relationship between HCV-encoded genes and HBx
To understand the relationship between HCV core, NS3, NS5A and HBx, we analyzed the number of unique or overlapping genes among these four groups. This analysis revealed that core, NS3, NS5A and HBx form mainly distinctive groups, with minimal overlapping genes (Table I). Only one gene, pregnancy-associated interferon (HTIFN), which shares high sequence identity with IFN-{alpha}, was affected by each HCV-encoded gene and HBx. Core and HBx genes were quite distinct from one another, sharing only seven genes. HBx infection had the greatest number of affected cellular genes followed by core, NS5A and NS3 (Figure 1d). Among the unique genes in each of the four groups, HBx mainly resulted in the down-regulation of cellular genes, whereas HCV-encoded genes up-regulated cellular genes. Multidimensional scaling analysis showed that core, NS3, NS5A and HBx arrays clustered separately (Figure 1e), revealing distinctions among these four groups. Although, these groups have a minimal number of overlapping genes, ~20% of the significant genes found in core, NS3 or NS5A overlap with HBx, indicating that some similarities exist among these viral proteins (Table I). The data suggest that core and NS5A have more overlapping genes than core and NS3, illustrating similarities and differences among the HCV-encoded genes (Table I).


View this table:
[in this window]
[in a new window]

  		Table I. Characterization of the relationship among core, NS3, NS5A and HBx

 
Categorization of the significant and differentially expressed genes affected by HCV or HBx
To further explore the gene sets from this study, the unique genes that were significant and differentially expressed by core, NS3, NS5A, HCV and HBx were categorized according to function (Figure 2 and Supplemental Table 1, available at Carcinogenesis Online). This analysis revealed that HBx and HCV structural or NS proteins affected cell regulation, which includes processes such as proliferation and apoptosis. As expected, given the viral context of these proteins, immune response and host defense genes were also affected by both HCV and HBx. The data also define a distinction between the structural and NS HCV proteins since differentially expressed genes in certain functional groups (e.g. metabolism) demonstrate a relationship among processes solely affected by NS HCV proteins and HBx, but not by core. Transcription-related genes were highly affected by HCV infection. HBx infection affected both GTP-related genes and transport-related genes that were minimally affected by HCV. These results suggest that HCV and HBV affect some genes with similar function that are likely to be genes associated with response to viral infection. However, several functional groups of genes are transcriptionally altered by only particular HCV- or HBV-encoded genes, which serves to highlight significant differences among the cellular pathways that are affected by these viruses. We also used pathway analysis software to assess the functional pathways that were significantly affected by core, NS3, NS5A or HBx. Since only a few genes were significantly altered by NS3 and NS5A in our microarray analysis, these gene sets were combined for pathway analysis. The significantly affected pathways by core, NS3 + NS5A and HBx were differentially affected (Table II). As a control, only four functional pathways, related to adhesion, were significantly altered when all the genes encompassing the entire oligo array were inputted. Only one of these overlapped with significant pathways identified for HBx; however, the specific genes involved in the pathway differed among HBx and the entire oligo list (Supplemental Table 2 is available at Carcinogenesis Online). The analyses above suggested that the affected pathways were not selected by random chance, but rather were unique for specific viral genes. In addition, a comparison among pathways affected by core and NS3 + NS5A or core and HBx revealed a distinction in the pathways affected by structural and NS HCV proteins or the oncogenic proteins of HCV or HBV (Table III).


Figure 2
View larger version (27K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2. Summary of significant differentially expressed unique genes in core, NS3 and NS5A, HCV or HBx. Each pie chart describes the number of genes that reside in each functional category and the percentage attributed by each category.

 


View this table:
[in this window]
[in a new window]

  		Table II. Top 10 significant functions and diseases

 


View this table:
[in this window]
[in a new window]

  		Table III. Unique significant functions and diseases

 
HCV core and HBx of HBV have both been implicated as multifunctional proteins with oncogenic capacity. To further investigate whether core and HBx affect similar or disparate pathways, we analyzed regulator (upstream effector) and target (downstream effector) molecular interaction pathways for these proteins using Pathway Assist (Stratagene) (Figure 3). As a control, five separate randomized gene lists, each composed of 34 genes, were analyzed using this software and resulted in no specific pathway identification (data not shown). The biological area networks generated by Pathway Assist were based on the unique core and HBx gene sets and reveal that functional upstream or downstream effectors related to infection with these viral proteins are mainly distinct. In particular, RAS small monomeric GTPase, affecting the (Raf–MEK–ERK) cascade, plays a role as both a regulator and a target of MYC, a unique gene affected by core infection and a common denominator in many cancers, including HCC. Rho, on the other hand, a member of the GTPase family affecting the (SAPK–JNK) cascade, functions downstream of PKC, a unique gene affected by HBx expression. This analysis demonstrates that the effects on intracellular pathways, including MAPK, can be distinct between core and HBx (14). Only one gene, desmoplakin, a key component of cellular adhesion junctions, serves as a regulator for both core and HBx. However, the specific genes affected by desmoplakin differ when cells are infected with core or HBx, again demonstrating specific differences in the biological networks affected by these two viral genes. In addition, we analyzed the biological area networks affected by the NS HCV proteins, NS3 and NS5A, to assess similarities or differences with HCV core (Figure 3). We found that the up- and downstream functional classes associated with NS3 and NS5A differ from each other and with HCV core. Interestingly, cytokine activity is a common functional class in both NS3 and NS5A interaction networks, although it serves as a regulator for NS3 and a target for NS5A.


Figure 3
View larger version (22K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 3. Biological area network of common regulators and targets of core and HBx. Core unique genes and HBx unique genes were analyzed by Pathway Assist (Stratagene). Red circles represent proteins and yellow hexagons represent protein functional classes. Lines represent biological connections between nodes. Circles or squares along lines represent the biological mode of connection between nodes (gray line with green box = molecular transport, blue line with blue filled box = expression, dashed line with gray filled box = regulation, purple line with purple circle = binding and red line with red circle = protein modification, + or – within boxes along lines represent a positive effect or negative effect, respectively. Imported gene lists for core, NS3, NS5A or HBx were analyzed in Pathway Assist using a ResNet2.5 Mammalian Database. Search parameters for common targets or regulators were restricted by filtering based on functional classes.

 
Comparison among HCV or HBx differentially expressed genes and liver specimens from HCV- or HBV-positive chronic liver disease or HCC patients
The generation of HCC generally takes several decades; however, our analysis was performed to analyze the immediate effects of viral gene expression in primary hepatocytes. To examine the physiological relevance of genes that were differentially regulated by HCV- or HBV-encoded genes, we compared the expression profiles of HCV- or HBx-infected primary human hepatocytes with HCV- or HBV-infected liver samples from patients with chronic liver disease (CLD) or HCC. cDNA array profiles of liver tissue samples from CLD or HCC patients associated with HBV or HCV infection were previously performed (13). Comparison between these liver sample profiles and those found in HCV- or HBx-infected hepatocytes revealed that a subset of genes shared similar expression patterns (Figure 4a and b). Of these genes, those related to transcription events and cell regulation were affected in both human hepatocytes and CLD or HCC patients. Genes related to transport, GTP and metabolism functional groups were mostly affected by HBx.


Figure 4
View larger version (23K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 4. Comparison between HCV or HBx differentially expressed genes and liver specimens from HCV- or HBV-positive chronic liver disease or HCC patients. Gene expression of signature genes is represented as the natural log of the signal intensity ratio of either hepatocytes expressing HCV or HBx (panel a or panel b, respectively, n = 5, yellow bars) compared with HHT-4 cells or HCV- or HBV-infected chronic liver disease (CLD; HCV: n = 11; HBV: n = 7; blue bars) or HCC (HCV: n = 7; HBV: n = 2; red bars) patient samples compared with normal tissue. Asterisks represent samples that were below the detection limit. Data are presented with standard error of mean. (c) RT validation of significant differentially expressed HCV-encoded genes or HBx. Relative expression fold of each gene (n = 4) is shown for HHT4, primary hepatocytes infected with ß-gal, core, HCV or HBx and HCV- or HBV-related CLD (n = 8 or n = 2, respectively) or HCC (n = 7, n = 2, respectively). The numbers 1 and 2 on the x-axis represent two biological replicates. Data are presented with standard error of mean. (d) A linear regression and correlation among expression fold data from qRT–PCR versus either oligo or cDNA microarray is shown with r2 and P-value indicated.

 
To determine the accuracy of the identified signatures, several differentially expressed genes (n = 6) in CLD or HCC liver samples and HCV- or HBx-infected hepatocytes were randomly chosen for validation by qRT–PCR (Figure 4c). These results show that transcript levels measured by qRT–PCR and gene expression oligo array expression from infected primary hepatocytes correlated significantly (P < 0.0001; r2 = 0.7096) (Figure 4d). In addition, qRT–PCR analysis was conducted on tissue samples from patients with HCV- or HBV-related CLD or HCC and was shown to significantly correlate with cDNA expression levels (P < 0.0001; r2 = 0.8128) (Figure 4d). In particular, interferon-inducible gene 27 (IFI27) shows significant up-regulation after infection with core or HCV, but was not significantly differentially expressed by HBx. Since the host transcriptional response to HCV infection can be considerably variable among different donors (15,16), we also analyzed a separate hepatocyte donor (HH1332) and found similar results in the IFI27 gene (Supplemental Figure 1 is available at Carcinogenesis Online).


   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
Supplementary material
 References
 
HCV is a known etiological agent for HCC; however, the mechanisms by which this virus leads to HCC are unknown. To examine the molecular effects of this virus, 3 of the 10 HCV viral proteins, core, NS3 and NS5A, representing either structural or NS groups, were chosen for this study. Core, a structural component of HCV is a multifunctional protein that encodes the viral capsid and has tumorigenic activity (9). NS3 and NS5A are two of the six HCV proteins (i.e. NS2, NS3, NS4A, NS4B, NS5A and NS5B) that together form the replication complex or ‘replicase’ of HCV (17). NS3 is a serine protease that catalyzes a portion of HCV polyprotein cleavage and contains both nucleoside triphosphatase (NTPase) and RNA helicase activity. NS3 activity is essential for HCV replication and is consequently a major target for anti-HCV drug design, which has recently advanced due to the definition of three-dimensional structures (18,19). The NS5A phosphoprotein is implicated in mediating HCV resistance to the cytokine interferon-{gamma} through its ability to inhibit the IFN-induced protein kinase (PKR) (20). NS5A has also been implicated in the activation of nuclear factor-kappa B during oxidative stress (21). These represent direct mechanisms by which HCV can alter host cell responses to persist and replicate after infection.

Several groups have used expression profiling techniques to study the effects of HCV infection in tumor cell lines and patient samples (15,22–25). In our study, freshly isolated adult normal primary human hepatocytes were chosen because they are natural host cells for hepatitis viral infection and their response to infection would most closely resemble human infection. A comparison of the unique or overlapping genes among hepatocytes infected by core, NS3, NS5A and HBx demonstrated that these viral genes mainly affect distinct genes. Only one gene, pregnancy-associated interferon (HTIFN), which shares high sequence identity with IFN-{alpha}, was affected by each HCV-encoded gene and HBx. This finding is in line with our current understanding of viral hepatitis-related mechanisms showing widely documented associations among hepatitis viruses and interferon pathways (26,27). Functional categorization of the gene lists revealed that HBx and HCV structural or NS proteins affected cell regulation, which includes processes such as proliferation and apoptosis. These cellular functions have been documented to be affected by hepatitis viruses (28,29). In addition, transcription-related genes were highly affected by HCV infection. This process appears to originate mainly from core, which has been documented to affect transcription through multiple mechanisms including enhancement of DNA-binding affinity and regulation of p300/CBP (30,31). Pathway analysis further demonstrated that although certain functional pathways are affected commonly by these hepatitis viral-encoded genes, they mainly affect distinct functional pathways. We also showed that specific gene expression changes related to HCV or HBV hepatocyte infection agreed with the expression profiles of liver disease patients, enhancing the significance of the genes found in this study. However, prudence should be taken when interpreting the direct involvement of these significantly affected genes in hepatocarcinogenesis since they were found through altered in vitro-based expression of viral genes. Taken together, these results indicate that despite their similar association as etiological risk factors for HCC, HCV and HBV may cause liver cancer through different mechanisms and while certain similarities exist among the cellular pathways affected by HCV and HBV, these viruses principally affect discrete cellular genes. In this vein, it will be of interest in the future to also compare the gene lists identified by individual viral gene expressions with those following the introduction of HBV or HCV using viral replicon systems.

IFI27, a significantly up-regulated gene following core or HCV infection was among several genes validated by qRT–PCR to provide a measure of confidence for the gene expression profiles identified in this study. Although the up-regulation of this gene was observed in two independent hepatocyte donors, we do recognize that all other genes found to be significantly altered in the original hepatocyte donor will need to be tested in an independent donor, which will be the subject of future study. The expression of IFI27 was shown previously to be specific for HCV-related HCC tumor tissue and not for HBV-related HCC and non-B, non-C HCCs by oligonucleotide microarray (32). In addition, RT–PCR-based studies have shown that IFI27 is up-regulated in early-stage liver fibrosis specifically related to HCV (33). These studies corroborate our findings and strengthen the importance of IFI27 expression in HCV-related liver disease. In particular, our study shows that IFI27 expression is altered in primary hepatocytes and is affected by the expression of HCV core protein. IFI27 is a small hydrophobic protein that has four human isoforms and is significantly conserved in eukaryotes (34). IFI27 is IFN inducible, but some family members do not contain the IFN system. In addition, some members that contain the IFI27 gene are not IFN inducible. This suggests that there are alternative functions for this gene and it has been suggested that IFI27 may be involved in stress responses related to the cellular environment. IFN has been routinely used in HCV treatment and our understanding of an IFN-related gene could improve HCV treatment and provide a potential target for HCV-related therapy. Future exploration of this gene will provide greater understanding of its role in HCV-related liver disease.

In summary, this work has demonstrated that individual HCV proteins, when expressed alone or in combination, affect cellular genes in a primarily distinct manner than HBV-encoded HBx. Analyses of gene expression assays have resulted in the identification of particular genes of interest, which can now be studied to delineate the specific mechanisms involved in HCV- or HBV-related hepatocarcinogenesis. In particular, some of these genes show similar expression in infected hepatocytes and CLD or HCC liver samples, and thus, may be useful as potential clinical markers for early-stage liver disease.


   Supplementary material
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Supplementary material
References
 
Supplementary data can be found at http://carcin.oxfordjournals.org/.


   Acknowledgments
 
We thank Jens Bukh (NIDDK, NIH) for the infectious clone used to generate HCV constructs, Richard Simon for input on experimental design and helpful discussion and Advanced Technology Center for production of the microarrays used in this study. We are grateful to Mike Lipsky (University of Maryland) for the liver donor used to generate HHT-4 cells and Steven Strom (Liver Tissue Procurement and Distribution System, University of Pittsburgh) for primary hepatocytes. Biological area networks were constructed using a demo version of Pathway Assist V2.53 (Stratagene). We also thank Karen MacPherson for bibliographic assistance and Dorothea Dudek-Creaven for editorial assistance.

Conflict of Interest Statement: None declared.


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Supplementary material
 References
 

  1. Parkin DM, et al. Global cancer statistics. CA Cancer J. Clin. (1999) 49:33–64.[Abstract/Free Full Text]
  2. Carr BI, et al. Hepatobiliary cancers: cancer of the liver. In: Cancer Principles & Practice of Oncology—DeVita VT Jr, Hellman S, Rosenberg SA, eds. (1997) Philadelphia: Lippincott-Raven. 1087–1114.
  3. Bressac B, et al. Selective G to T mutations of p53 gene in hepatocellular carcinoma from southern Africa. Nature (1991) 350:429–431.[CrossRef][Medline]
  4. Chen CJ, et al. Epidemiological characteristics and risk factors of hepatocellular carcinoma. J. Gastroenterol. Hepatol. (1997) 12:S294–S308.[Web of Science][Medline]
  5. Steinmann D, et al. Inhibition of hepatitis C virus-like particle binding to target cells by antiviral antibodies in acute and chronic hepatitis C. J. Virol. (2004) 78:9030–9040.[Abstract/Free Full Text]
  6. Budhu A, et al. The role of hepatitis viruses in liver oncogenesis. In: Recent Research Developments in Molecular Biology—Pandalai SG, Fagan J, Davidson JN, Shimizu N, eds. (2003) Trivandrum, Kerala: Research Signpost. 97–118.
  7. Brechot C, et al. Molecular bases for the development of hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC). Semin. Cancer Biol. (2000) 10:211–231.[CrossRef][Web of Science][Medline]
  8. Giannini C, et al. Hepatitis C virus biology. Cell Death. Differ. (2003) 10(suppl. 1):S27–S38.[CrossRef][Web of Science][Medline]
  9. Moriya K, et al. The core protein of hepatitis C virus induces hepatocellular carcinoma in transgenic mice. Nat. Med. (1998) 4:1065–1067.[CrossRef][Web of Science][Medline]
  10. Wu CG, et al. SAGE transcript profiles of normal primary human hepatocytes expressing oncogenic hepatitis B virus X protein. FASEB J. (2002) 16:1665–1667.[Abstract/Free Full Text]
  11. Wu CG, et al. Distinctive gene expression profiles associated with hepatitis B virus x protein. Oncogene (2001) 20:3674–3682.[CrossRef][Web of Science][Medline]
  12. Ye QH, et al. Predicting hepatitis B virus-positive metastatic hepatocellular carcinomas using gene expression profiling and supervised machine learning. Nat. Med. (2003) 9:416–423.[CrossRef][Web of Science][Medline]
  13. Kim JW, et al. Cancer-associated molecular signature in the tissue samples of patients with cirrhosis. Hepatology (2004) 39:518–527.[CrossRef][Web of Science][Medline]
  14. Panteva M, et al. Hepatitis viruses and the MAPK pathway: is this a survival strategy? Virus Res. (2003) 92:131–140.[CrossRef][Web of Science][Medline]
  15. Smith MW, et al. Identification of novel tumor markers in hepatitis C virus-associated hepatocellular carcinoma. Cancer Res. (2003) 63:859–864.[Abstract/Free Full Text]
  16. Smith MW, et al. Gene expression patterns that correlate with hepatitis C and early progression to fibrosis in liver transplant recipients. Gastroenterology (2006) 130:179–187.[CrossRef][Web of Science][Medline]
  17. Neddermann P, et al. The nonstructural proteins of the hepatitis C virus: structure and functions. Biol. Chem. (1997) 378:469–476.[Web of Science][Medline]
  18. Kwong AD, et al. Structure and function of hepatitis C virus NS3 helicase. Curr. Top. Microbiol. Immunol. (2000) 242:171–196.[Web of Science][Medline]
  19. Yan Y, et al. Complex of NS3 protease and NS4A peptide of BK strain hepatitis C virus: a 2.2 A resolution structure in a hexagonal crystal form. Protein Sci. (1998) 7:837–847.[Web of Science][Medline]
  20. Reyes GR. The nonstructural NS5A protein of hepatitis C virus: an expanding, multifunctional role in enhancing hepatitis C virus pathogenesis. J. Biomed. Sci. (2002) 9:187–197.[CrossRef][Web of Science][Medline]
  21. Waris G, et al. Hepatitis C virus NS5A and subgenomic replicon activate NF-kappaB via tyrosine phosphorylation of IkappaBalpha and its degradation by calpain protease. J. Biol. Chem. (2003) 278:40778–40787.[Abstract/Free Full Text]
  22. Shirota Y, et al. Identification of differentially expressed genes in hepatocellular carcinoma with cDNA microarrays. Hepatology (2001) 33:832–840.[CrossRef][Web of Science][Medline]
  23. Smith MW, et al. Hepatitis C virus and liver disease: global transcriptional profiling and identification of potential markers. Hepatology (2003) 38:1458–1467.[Web of Science][Medline]
  24. Aizaki H, et al. Expression profiling of liver cell lines expressing entire or parts of hepatitis C virus open reading frame. Hepatology (2002) 36:1431–1438.[Web of Science][Medline]
  25. Ohkawa K, et al. Changes in gene expression profile by HCV core protein in cultured liver cells: analysis by DNA array assay. Hepatol. Res. (2003) 25:396–408.[CrossRef][Web of Science][Medline]
  26. Guidotti LG. The role of cytotoxic T cells and cytokines in the control of hepatitis B virus infection. Vaccine (2002) 20(suppl. 4):A80–A82.[CrossRef][Web of Science][Medline]
  27. Rosen HR. Hepatitis C pathogenesis: mechanisms of viral clearance and liver injury. Liver Transpl. (2003) 9:S35–S43.[Web of Science][Medline]
  28. Shirakata Y, et al. Hepatitis B virus X protein induces cell death by causing loss of mitochondrial membrane potential. J. Biol. Chem. (2003) 278:22071–22078.[Abstract/Free Full Text]
  29. Jung EY, et al. The repressive activity of hepatitis C virus core protein on the transcription of p21(waf1) is regulated by protein kinase A-mediated phosphorylation. Virus Res. (2001) 79:109–115.[CrossRef][Web of Science][Medline]
  30. Kao CF, et al. Modulation of p53 transcription regulatory activity and post-translational modification by hepatitis C virus core protein. Oncogene (2004) 23:2472–2483.[CrossRef][Web of Science][Medline]
  31. Gomez-Gonzalo M, et al. Hepatitis C virus core protein regulates p300/CBP co-activation function. Possible role in the regulation of NF-AT1 transcriptional activity. Virology (2004) 328:120–130.[CrossRef][Web of Science][Medline]
  32. Iizuka N, et al. Molecular signature in three types of hepatocellular carcinoma with different viral origin by oligonucleotide microarray. Int. J. Oncol. (2004) 24:565–574.[Web of Science][Medline]
  33. Bieche I, et al. Molecular profiling of early stage liver fibrosis in patients with chronic hepatitis C virus infection. Virology (2005) 332:130–144.[CrossRef][Web of Science][Medline]
  34. Parker N, et al. Identification of a novel gene family that includes the interferon-inducible human genes 6-16 and ISG12. BMC Genomics (2004) 5:8.[CrossRef][Medline]
Received February 1, 2007; revised March 26, 2007; accepted March 26, 2007.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
28/7/1552    most recent
bgm075v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (1)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Budhu, A
Right arrow Articles by Wang, X.W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Budhu, A
Right arrow Articles by Wang, X.W.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?