This article is a US Government work and, as such, is in the public domain in the United States of America.
Hepatitis B virus X protein inhibits nucleotide excision repair †
Article first published online: 8 NOV 1999
Copyright © 1999 Wiley-Liss, Inc.
International Journal of Cancer
Volume 80, Issue 6, pages 875–879, 15 March 1999
How to Cite
Jia, L., Wei Wang, X. and Harris, C. C. (1999), Hepatitis B virus X protein inhibits nucleotide excision repair . Int. J. Cancer, 80: 875–879. doi: 10.1002/(SICI)1097-0215(19990315)80:6<875::AID-IJC13>3.0.CO;2-Z
- Issue published online: 8 NOV 1999
- Article first published online: 8 NOV 1999
- Manuscript Revised: 8 OCT 1998
- Manuscript Received: 15 JUN 1998
Hepatocellular carcinoma (HCC) is a common tumor worldwide and ranks as one of the 4 most prevalent malignant diseases of adults in Asia and sub-Saharan Africa. More than 250,000 cases of HCC are diagnosed each year, and less than 3% of these patients will survive 5 or more years (Feitelson and Duan, 1997). Chronic infection with hepatitis B virus (HBV), ingestion of aflatoxin B1 (AFB1)-contaminated foods, alcohol consumption–related cirrhosis and other factors associated with chronic inflammatory and hepatic regenerative changes are important risk factors for hepatocarcinogenesis (Elmore and Harris, 1998; Song et al.,1985; Harris et al.,1987; Beasley et al.,1981).
The X protein encoded by HBV (HBx) is one of the major risk factors associated with primary HCC; it is a protein of 154 amino acids that is highly conserved among the different subtypes of HBV. HBx is multifunctional and is a known pleiotropic transactivator with no DNA-binding activity (Feitelson and Duan, 1997). It can stimulate the cis elements of the promoter of the HBV genome itself, as well as a wide range of other viral promoters such as the herpes simplex virus tk, Simian virus 40, long terminal repeats (LTRs) of human immunodeficiency virus types I and II, LTRs of human T lymphotropic virus type I, LTR of mouse mammary tumor virus and LTRs of Rous sarcoma virus (RSV) (for review: Jia et al.,1997; Feitelson and Duan, 1997). The cellular genes stimulated by HBx include the proto-oncogenes c-myc and c-fos/c-jun, α-anti-trypsin, α-fetoprotein, β-interferon, epidermal growth factor (EGF) receptor, metallothionein, RNA polymerase II and III, genes related to MHC I and II, interleukin 8 and intracellular adhesion molecules (for review: Jia et al.,1997; Feitelson and Duan 1997). HBx is also known to interact with several cellular and viral proteins, including ATF-2, CREB, Oct-1, subunits of RNA polymerase II, TFIIH, TATA-binding protein, the serine protease TL2, a cellular UV-damaged DNA-binding protein and the simian virus large tumor antigen (Jia et al.,1997; Feitelson and Duan, 1997). We and others have shown that HBx binds to p53 both in vivo and in vitro (Wang et al.,1994; Feitelson et al.,1993; Ueda et al.,1995), inhibits p53-mediated sequence-specific transcriptional activity, prevents p53 binding to the transcription-repair factors XPB and XPD (Wang et al.,1995a), reduces p53 entry into the nucleus (Ueda et al.,1995; Elmore et al.,1997; Takada et al.,1997) or abrogates p53-induced apoptosis (Wang et al.,1995a). HBx can stimulate cellular gene expression through activation of the protein kinase C pathway, as well as the ras, raf and MAP kinase signaling cascades (Benn and Schneider, 1994). HBx is oncogenic because it can neoplastically transform rodent cells in vitro and, as a transgene, it induces hepatocellular carcinoma in mice (Feitelson and Duan, 1997). The HBx gene is reported to be frequently integrated into the cellular genome and expressed in HCC, accounting for over 90% of the primary liver cancers (Jia et al.,1997; Feitelson and Duan, 1997). HCC can progress through inactivation of the p53 gene via mutation or interaction with HBx (Feitelson and Duan, 1997).
To better understand and explore the pleiotropic oncogenic mechanism of HBx, we report here that HBx can interfere with the NER pathway, as observed using UVC light-sensitive cell-proliferation and colony-formation assays in addition to a plasmid host cell re-activation assay. Through an in vitro protein–protein binding assay, we also found that HBx can directly bind to either XPB or XPD, which are essential DNA helicases in the TFIIH nucleotide excision–basal transcription complex.
MATERIAL AND METHODS
The pCMV-X1 and GST-X plasmids were described previously (Wang et al.,1994). The plasmid pGL-2 containing the luciferase gene was obtained from Promega (Madison, WI). The pZAP10 plasmid was used for in vitro synthesis of the XPB protein. The pGEM3zf(+)T7XPD plasmid was used for in vitro synthesis of the XPD protein. The pcBLsSE6 plasmid was used for in vitro synthesis of CSB. The pcDNA3-HHR54 plasmid was used for in vitro synthesis of the HHR54 protein. All 4 of these plasmids were kindly provided by Drs. J. Hoeijimakers and G. Weeda (Erasmus University, Rotterdam, The Netherlands). The pSPX46 plasmid was used for in vitro synthesis of HBx. All plasmids were purified using CsCl density gradient centrifugation.
Cell strains and cultures
HepG2 (ATCC, Rockville, MD), RKO, RKO-143ala (p53–143ala) and RKO-E6 cell strains (gifts of Dr. M. Kastan) and culture conditions were essentially as described (Smith et al.,1995). RKO cells were grown in A50 medium supplemented with 10% FBS, 2 mM glutamine, 100 mg/l ampicillin and 100 mg/l streptomycin. HepG2 cells were grown in EMEM supplemented with 10% FBS, non-essential amino acids, 2 mM glutamine, 100 mg/l ampicillin and 100 mg/l streptomycin.
UVC light-sensitive cell-proliferation assay
Twenty-eight hours after HepG2 cells were transfected with different DNA (salmon sperm DNA, pCMV-Neo or pCMV-X1), each using a total of 2 μg DNA by lipofectin (Life Technology, Gaithersburg, MD) in 60-mm dishes, cells were exposed to UVC light (254 nm) at either 7.5 or 15 J/m2. Cells were then cultured in a 96-well plate (5,000 cells/well) and incubated for 72 hr in a humidified 5% CO2 atmosphere. The Cell Titer 96 Non-Radioactive Cell Proliferation Assay kit (Promega) was used to measure cell proliferation.
UVC light-sensitive colony-formation-efficiency assay
Cells were seeded in 60-mm dishes and grown to 85% confluence, then transfected with either control plasmid pCMV-Neo or pCMV-X1 by lipofectin. Twenty-eight hours post-transfection, cells were exposed to UVC light (254 nm) at 0, 10 or 20 J/m2. After culture for 14 days in G418 (600 μg/ml) selective growth medium, cells were fixed and stained with methylene blue and cell colonies were counted.
Plasmid host cell re-activation assay
UVC irradiation of plasmid DNA.
pGL-2 plasmid DNA (0.1 mg/ml in double distilled H2O) was exposed to UVC at 15 KJ/m2.
Human HepG2 cells or RKO cells (2 × 105) were plated in triplicate in 60-mm dishes and transfected the next day using 0.5 μg of either normal or UVC light-damaged pGL-2 plasmid DNA with or without co-transfection of 1 μg pCMV-X1 (total DNA amounts were equalized with salmon sperm DNA to 5 μg for all transfections).
Forty-eight hours post-transfection, cells were rinsed twice in PBS, and then 250 μl of Cell Culture Lysis 1 × Reagent (Promega) were added to each 60-mm culture dish and incubated at room temperature for 15 min. The attached cellular material was scraped from the culture dish and spun briefly (5 sec at 4,800 g) in an Eppendorf microcentrifuge to pellet the large cellular debris. The cell extract (20 μl) was mixed with 100 μl of Luciferase Assay Reagent at room temperature. The reaction was placed in a luminometer (model 2010; Analytical Luminescence Laboratory, San Diego, CA) for luciferase activity assay. The final luciferase activity was determined by using equal amounts of protein.
In vitro translation and protein–protein binding assay
The expression and purification of the recombinant proteins as well as the in vitro protein-binding assays were described previously (Wang et al.,1994). The GST-fusion proteins were produced in Escherichia coli and purified on glutathione-Sepharose 4 B beads according to the manufacturer's instructions (Pharmacia LKB, Piscataway, NJ). Protein concentrations were estimated by Coomassie blue staining of SDS-PAGE and by comparison to m.w. standards (Bio-Rad, Hercules, CA) run on the same gel. To label the in vitro synthesized proteins, the corresponding plasmids were incubated at room temperature for 90 min with [35S]-cysteine (DuPont, Boston, MA) in a 1-step in vitro transcription and translation system (Promega). Binding assays were done in 500 μl IP buffer [50 mM Tris-HCI (pH 8.0), 120 mM NaCl, 0.5% Nonidet P40] by incubating the GST-fusion proteins with in vitro translated proteins for 60 min at room temperature. After washing with IP buffer, bound proteins were released by boiling in Laemmli buffer for 5 min, separated by SDS-PAGE and visualized by autoradiography.
HepG2 or RKO cells expressing HBx are hypersensitive to UVC radiation-induced cell killing
Repair-deficient cells often display hypersensitivity to DNA damage-induced cell killing. Therefore, we determined the sensitivity of various cells expressing HBx to UVC radiation by a cell-proliferation assay or a colony-formation assay. The sensitivity of HepG2 cells to UVC radiation, measured by the cell-proliferation assay, is shown in Figure 1. Without radiation, expression of HBx did not alter the cell-proliferation efficiency of HepG2 cells. Cell-proliferation efficiency was reduced with increasing doses of UVC light and further reduced in cells transfected with the HBx expression vector pCMV-X1 when compared with cells transfected with either salmon sperm DNA or the control vector CMV-Neo. At 7.5 J/m2 of UVC, the proliferation efficiency of HepG2 cells was 76% when transfected with control DNA, 77% with the pCMV-Neo and 56% with pCMV-X1. At the 15 J/m2 UVC dose, the proliferation efficiency of HepG2 cells was 60% when transfected with control DNA, 58% with pCMV-Neo and 30% with pCMV-X1.
The data from the colony-formation efficiency (CFE) assay of the RKO cells are shown in Figure 2. Without radiation, the colony-formation units (CFUs) were similar with or without expression of HBx (mean ± SD: 226 ± 9 CFUs for RKO cells transfected with pCMV-Neo and 216 ± 10 CFUs for RKO cells transfected with pCMV-X1). At 10 J/m2 of UVC, the CFU count in RKO cells transfected with the control vector pCMV-Neo was 170 ± 4 (75% of the non-irradiated control) and that of cells transfected with pCMV-X1 reduced to 105 ± 3 (49%). At 20 J/m2 UVC dose, cells transfected with pCMV-Neo had 47% CFE (107 ± 3 CFUs), whereas cells transfected with pCMV-X1 had 18% CFE (39 ± 3 CFUs). Thus, RKO cells transfected with pCMV-X1 were more sensitive to UVC radiation when compared with the control pCMV-Neo vector.
Cells expressing HBx are impaired in the UV-induced DNA-repair pathways
To determine if the hypersensitivity of HBx-expressing RKO cells is due to a reduction of UV-mediated DNA-repair activity, we used the UVC-irradiated luciferase reporter construct (pGL-2) to measure the DNA-repair ability in different RKO cells that carry wild-type p53, p53 mutant (RKO-143ala) or p53 null (RKO-E6) (Table I). This assay reflects indirectly the nucleotide excision repair (NER) of damaged DNA (for TT or TC dimers) by using transcriptional activity measured as luciferase enzyme activity of the transfected reporter gene. The plasmid pGL-2 was exposed to UVC light (254 nm) at 15 kJ/m2. The parental RKO cells (wild-type p53) rapidly repaired the UVC light-damaged reporter and showed a relatively high luciferase activity. In RKO cells co-transfected with pCMV-X1, luciferase activity was decreased to 52 ± 2%. HBx has a co-transactivation function; i.e., it transactivates the luciferase reporter gene because pGL-2 contains an SV40 promoter. However, when UVC-irradiated pGL-2 was co-transfected with pCMV-X1, the expression level of luciferase activity was only 52 ± 2% (Table I). When UVC-irradiated pGL-2 co-transfected with pCMV-X1 in RKO-143ala cells, the expression level of luciferase activity was further decreased to 46 ± 3% when compared with the control. The status of the p53 gene in RKO-E6 cells is essentially null. When UVC-irradiated pGL-2 co-transfected with pCMV-X1 into the RKO-E6 cells, the expression level of luciferase activity was decreased to 60 ± 3% of the control level. Therefore, HBx may interfere with both p53-dependent and -independent NER pathways.
|Cell type (p53)||Treatment||Relative||Repair|
|RKO (wt/wt)||Control||1 ± 1|
|pCMV-X1||2 ± 0.4|
|pGL-2||19 ± 1|
|pGL-2 and pCMV-X1||51 ± 1|
|UVC-exposed pGL-2||13 ± 0.4|
|UVC-exposed pGL-2||18 ± 1||52 ± 21|
|RKO (mutant||Control||1 ± 0.3|
|pCMV-X1||1 ± 0.2|
|pGL-2||18 ± 1|
|pGL-2 and pCMV-X1||43 ± 3|
|UVC-exposed pGL-2||9 ± 0.2|
|UVC-exposed pGL-2||10 ± 0.2||46 ± 31|
|RKO (E6,||Control||1 ± 1|
|p53-deficient)||pCMV-X1||1 ± 0.1|
|pGL-2||25 ± 2|
|pGL-2 and pCMV-X1||27 ± 1|
|UVC-exposed pGL-2||17 ± 1|
|UVC-exposed pGL-2||11 ± 1||60 ± 31|
|HepG2 (wt/wt)||Control||1 ± 1|
|pCMV-X1||1 ± 0.4|
|pGL-2||163 ± 1|
|pGL-2 and pCMV-X1||279 ± 3|
|UVC-exposed pGL-2||60 ± 1|
|UVC-exposed pGL-2||77 ± 1||75 ± 41|
We also tested a hepatoblastoma cell line, HepG2 (wild-type p53), using the plasmid host cell re-activation assay. HepG2 cells expressing HBx also showed reduced luciferase activity (75 ± 4%) when compared with the control (Table I).
HBx binds to XPB and XPD in vitro
The HBx inhibition of NER in p53-deficient RKO-E6 cells indicates that a protein(s) other than p53 in the NER pathway also may be targeted by HBx. To test this hypothesis, the binding of HBx directly to XPB, XPD, CSB or the human homologue of Rad54 (HHR54) was examined.
The in vitro protein interaction was studied, using either the in vitro translated HBx protein or the HBx fusion protein to glutathione S-transferase (GST). In vitro translated HBx protein (17 kDa) directly binds to GST-XPB (Fig. 3a, lane 3) but not to GST (Fig. 3 a, lane 2). In vitro translated XPD protein (80 kDa) binds to GST-HBx (Fig. 3b, lane 3) but not to GST (Fig. 3b, lane 2). No binding was observed between GST-HBx and in vitro translated XPB (89 kDa), CSB (168 kDa) or HHR54 (78 kDa) (Fig. 3c).
Because the HBx oncoprotein from HBV binds to several cellular proteins that are known to be involved in the DNA-repair pathways (Wang et al.,1994; Lee et al.,1995; Qadri et al.,1996), we have investigated whether HBx may directly interfere with NER using the prototypic DNA-damaging agent UV. By using a cell-proliferation assay and a clonogenic assay, we demonstrate here that both the hepatoblastoma HepG2 cells and the colon carcinoma cells RKO transfected with the HBx expression vector are more sensitive to UVC irradiation than cells transfected with the control DNA. Using a host cell re-activation assay, we also observed that cells expressing HBx showed decreased repair capacity for UV-induced lesions. In addition, HBx binds to damaged DNA, inhibits NER and sensitizes liver cells to UV radiation (Capovilla et al.,1997; Becker et al.,1998). These data indicate that HBx may interfere with the mammalian cell NER pathway. Our results are consistent with findings that HBx may contribute to liver carcinogenesis by disregulation of host cell defense mechanisms, such as cell-cycle transition, DNA repair and apoptosis (Benn and Schneider 1995; Lee et al.,1995; Wang et al.,1995a; Elmore et al.,1997).
Both in vitro and in vivo studies have shown that HBx can bind and disrupt the p53 binding to TFIIH, which is necessary for NER (Feitelson et al.,1993; Wang et al.,1994; 1995a; Ueda et al.,1995). These data suggest that HBx may interfere with NER activity through a p53-dependent mechanism. Therefore, we compared the HBx-expressing cells from a colon carcinoma RKO cell line and its derivative expressing a mutant p53-143ala in response to UVC radiation. The p53 mutant-expressing cells were previously shown to have reduced NER activity when compared with their parent RKO cells, indicating a dominant-negative activity of this mutant to block wild-type p53-mediated NER (Smith et al.,1995). Consistent with this model is our finding that RKO cells expressing mutant p53 displayed greater inhibition of NER activity by HBx than parental cells. The binding domain of p53 required for interaction with HBx has been mapped between residues 293 and 393 (Wang et al.,1995a). This region is also known to bind to XPB and XPD (Wang et al.,1995b). Therefore, it is possible that HBx may interfere in the NER pathway by masking the p53 C-terminal domain and by blocking p53 from binding to XPB and XPD.
A comparison of RKO cells to RKO-E6 cells by response to UVC radiation reveals the complexity of the mechanisms of NER inhibition by HBx. The inhibition of NER by HBx also can be p53-independent because a modest inhibitory effect was observed in the p53-null cells. These results indicate that HBx may inhibit NER through binding to and disrupting other DNA-repair factors. One of the candidates is the human homologue of the UV-DNA-damaged binding protein (UV-DDB) (Lee et al.,1995). In addition, HBx binds to several TFIIH components, including XPB and XPD, which are essential for NER activity, and stimulates TFIIH-mediated DNA helicase activity (Qadri et al.,1996). This is the opposite of p53 because p53 may involve the NER pathway by inhibiting TFIIH-mediated helicase activity during assembly of the repair complex (Wang et al.,1995b; Leveillard et al.,1996). Consistent with these results, we also showed that HBx binds to XPB or XPD in vitro. Interestingly, we found that a GST-tagged HBx recombinant protein bound only to in vitro translated XPD, but not XPB. However, HBx could bind weakly to a GST-tagged XPB recombinant protein. Because there is a discrepancy between our data and those published (Qadri et al.,1996), it is possible that HBx may bind much more strongly to XPD than XPB. Although it is unclear precisely how HBx alters TFIIH-mediated helicase activity, these data imply that HBx could certainly modulate NER activity by targeting TFIIH via either a p53-dependent or -independent mechanism.
Because AFB1-DNA adducts are repaired by NER, its inhibition by HBx could increase the frequency of AFB1-induced mutations, including in the p53 tumor-suppressor gene. The HBx gene is frequently integrated into the host cell genome and expressed in geographic areas where AFB1 also is an important risk factor of HCC (Jia et al.,1997; Feitelson and Duan, 1997). The HBx inhibition of NER could contribute to the high frequency of the AFB1-associated codon 249serp53 mutations in HCC from China and Africa. Thus, inhibition of p53-mediated NER by HBx may contribute to the development of human HCC.
We thank Drs. K. Kraemer and M. Nagashima for valuable support and Ms. D. Dudek for editorial assistance.
- Hepatocellular carcinoma and hepatitis B virus. A prospective study of 22,707 men in Taiwan. Lancet, 2, 1129–1133 (1981). Medline , , , and ,
- Hepatitis B virus X protein interferes with cellular DNA repair. J Virol, 72, 266–272 (1998). Medline , , , and ,
- Hepatitis B virus HBx protein activates Ras-GTP complex formation and establishes a Ras, Raf, MAP kinase signaling cascade. Proc. nat. Acad. Sci. (Wash.), 91, 10350–10354 (1994). , and ,
- Hepatitis B virus HBx protein deregulates cell cycle checkpoint controls. Proc. nat. Acad. Sci (Wash.), 92, 11215–11219 (1995). , and ,
- Hepatitis B virus X-protein binds damaged DNA and sensitizes liver cells to ultraviolet irradiation. Biochem. biophys. Res Comm, 232, 255–260 (1997). Medline , , and ,
- Hepatitis B virus X protein and p53 tumor suppressor interactions in the modulation of apoptosis. Proc. nat. Acad. Sci. (Wash.), 94, 14707–14712 (1997). , , , , , , , , and ,
- Hepatocellular carcinoma. In: B., Vogelstein, and K.WKinzler, (eds.). The genetic basis of human cancer, pp. 682–689, McGraw-Hill, New York (1998). , ,
- Hepatitis B virus X antigen in the pathogenesis of chronic infections and the development of hepatocellular carcinoma. Amer. J Pathol., 150, 1141–1157. (1997). Medline , and ,
- Hepatitis B x antigen and p53 are associated in vitro and in liver tissues from patients with primary hepatocellular carcinoma. Oncogene, 8, 1109–1117 (1993). Medline , , , and ,
- Biochemical and molecular epidemiology of human cancer: indicators of carcinogen exposure, DNA damage, and genetic predisposition. Environ. Hlth. Perspect, 75, 109–119 (1987). , , , , and ,
- Interactive effects of p53 tumor suppressor gene and hepatitis B virus in hepatocellular carcinogenesis. In: E., Tahara, (ed.). Molecular pathology of gastroenterological cancer: application to clinical practice, pp. 209–218, Springer-Verlag, Tokyo (1997). , , , ,
- Hepatitis B virus X protein interacts with a probable cellular DNA repair protein. J. Virol., 69, 1107–1114 (1995). Medline , , and ,
- Functional interactions between p53 and the TFIIH complex are affected by tumour-associated mutations. EMBO J, 15, 1615–1624 (1996). Medline , , , , , , and ,
- Hepatitis B virus transactivator protein, HBx, associates with the components of TFIIH and stimulates the DNA helicase activity of TFIIH. Proc. nat. Acad. Sci (Wash.), 93, 10578–10583 (1996). , , , , and ,
- Involvement of the p53 tumor suppressor in repair of UV-type DNA damage. Oncogene, 10, 1053–1059 (1995). Medline , , , , and ,
- Monoclonal antibody-directed radioimmunoassay detects cytochrome P-450 in human placenta and lymphocytes. Science, 228, 490–492 (1985). Medline , , , , and ,
- Cytoplasmic retention of the p53 tumor suppressor gene product is observed in the hepatitis B virus X gene-transfected cells. Oncogene, 15, 1895–1901 (1997). Medline , , , and ,
- Functional inactivation but not structural mutation of p53 causes liver cancer. Nature (Genet.), 9, 41–47 (1995). Medline , , , , , , and ,
- Hepatitis B virus X protein inhibits p53 sequence-specific DNA binding, transcriptional activity, and association with transcription factor ERCC3. Proc. nat. Acad. Sci (Wash.), 91, 2230–2234 (1994). , , , , , and ,
- Abrogation of p53-induced apoptosis by the hepatitis B virus X gene. Cancer Res, 55, 6012–6016 (1995a). Medline , , , , , , , and ,
- p53 modulation of TFIIH-associated nucleotide excision repair activity. Nature (Genet.), 10, 188–195 (1995b). Medline , and 14 others,