Transcriptional Repression of the Human P53 Gene by Hepatitis B Viral X Protein

SHORT REPORT Transcriptional repression of the human p53 gene by hepatitis B viral X protein

Seok Geun Lee1 and Hyune Mo Rho*,1

1Department of Molecular Biology and Research Center for Cell Dierentiation, Seoul National University, Seoul 151-742, Korea

Hepatitis B viral X protein (HBx) and the human p53 (Wang et al., 1997). Recently, it was reported that protein (p53) have been known as a transactivator and as activation of Src family kinases by HBx is coupled with a tumor suppressor, respectively. These two proteins have the activation of Ras (Klein and Schneider, 1997). also been known to interact with each other to neutralize Wild type p53 is a tumor suppressor protein that their authentic functions and the p53 represses the HBV induces either apoptosis or cell cycle arrest at the G1-S enhancer/X promoter activity. Here we report that the checkpoint in response to DNA damage, thus promoter activity of the human p53 gene was strongly preventing the transmission of genetic mutations to repressed by the HBx using the chloramphenicol acetyl the daughter cells (Lane, 1992). The overall function of transferase (CAT) assay. Analyses of serial deletion, wild-type p53 therefore appears to be maintenance of site-directed mutagenesis and the heterologous promoter the genetic integrity of the cell. Loss of wild-type p53 system showed that the site responsible for the repression activity has been identi®ed in over half of the human was the E-box element in the promoter of the p53 gene. malignancies examined (Levine, 1993), consistent with In addition, HBx as expected also repressed the a role for p53 in genome surveillance and the activation of the p53 promoter by c-Myc through the suppression of oncogenic transformation. Hepatocel- E-box element. Northern blot analyses also showed that lular carcinoma accounts for over 90% of primary liver the expression of the p53 gene in the HepG2-K8 cell line, cancers and may progress through the inactivation of which expresses HBV genes including HBx, was much the p53 gene (Hsu et al., 1991; Aguilar et al., 1994). more repressed than that of the control cell HepG2. HBx and p53 have been known to directly interact These results with previous data suggest that the shift of and inhibit their functions (Wang et al., 1994). In the reciprocal inhibitory activities at the levels of protein- addition, p53 was also known to inhibit the activity of protein interaction and transcription between HBx and the HBV enhancer/X promoter (Ori et al., 1998). Here, p53 may play a decisive role in the HBV-related we investigated if HBx would reciprocally regulate the hepatocarcinogenesis. Oncogene (2000) 19, 468 ± 471. activity of the p53 promoter and the expression of the p53 in HBV-infected cells. Keywords: HBx; repression; human p53 promoter; E- To study the eect of HBx on the promoter activity box element; c-Myc of the human p53 gene, a p53 promoter-CAT construct (p53p1CAT) was transfected into HeLa cells with either the HBx expression vector or adenovirus E1A HBV is a partially double stranded DNA virus that expression vector as a positive control. As shown in replicates through the reverse transcription of prege- Figure 1, the CAT activity of the p53 promoter was nomic RNA (Summers and Mason, 1982). HBV infects gradually repressed up to fourfold by increasing the humans and causes a wide range of clinical manifesta- HBx expression vector, whereas the activity was tions, from acute hepatitis to hepatocellular carcinoma increased by adenovirus E1A as previously reported (HCC) (Ganem and Varmus, 1987). HBx has been (Furlong et al., 1996). These results imply that the p53 implicated in HBV-mediated HCC through its ability gene is a target for HBx-mediated repression. to induce liver cancer in some transgenic mice (Kim et To identify the target sequence of HBx in the al., 1991) and to transactivate a variety of viral and upstream region of the p53 gene, deletion mutants of cellular promoters (Yen, 1996). The activity of HBx is the p53 promoter were prepared and transfected into also known to be mainly mediated through the binding HeLa cells with the HBx expression vector. HBx- sites for other transcription factors such as AP-1, NF- speci®c repression was observed in each construct kB, ATF/CREB and C/EBPa. HBx has been shown to about 4 ± 6-fold regardless of the promoter strength activate AP-1 and NF-kB by means of Ras-mediated (Figure 2a). These results suggested that the target signaling pathways (Yen, 1996; Caselmann, 1996; Choi sequence of HBx-speci®c repression is located in the et al., 1997, 1999). Activation of Ras by HBx was also region of nt 767 to +18, in which there are three shown to increase the cellular level of TBP, which transcriptional factor binding sites (CPEp53, NF-kB induces RNA polymerase III-dependent transcription and E-box element) (Figure 2b). As shown in Figure 2a (lanes 1, 3, 5, 7, and 9), deletion of the DNA sequences to 7128 resulted in an increase of CAT activity, which con®rmed a previous report of a negative transcrip- *Correspondence: HM Rho Received 15 February 1999; revised 11 October 1999; accepted tional control element upstream of the p53 gene (Bienz- 18 October 1999 Tadmor et al., 1985). To determine which site is HBx represses the transcription of p53 SG Lee and HM Rho 469 speci®cally involved in HBx-mediated repression of the dierentiated cell lineages (Weintraub, 1993). BETA2/ p53 promoter activity, we constructed three plasmids NeuroD, BETA2/E47 and USF bind the E-box bearing mutations on each p53 core promoter element sequence of the secretin, insulin, and murine p53 (CPEp53), NF-kB binding site and E-box element (Figure 2c). When transfected into the HeLa cells with the HBx expression vector, HBx-speci®c repression was maintained in the p53mCPE and p53mkB mutants, but not in the plasmid p53mEbox (Figure 2c). These results indicate that the transcriptional repression of the p53 gene by HBx is mediated through the E-box element. The basal promoter activity of p53mCPE was lower than that of the other plasmids (Figure 2c, lane 3) as previously reported (Sun et al., 1995). The eect of HBx was tested on the thymidine kinase (tk) promoter of herpes simplex virus (HSV) with or without the E-box elements. HBx activated the tk promoter about tenfold (Figure 3a, lanes 1 and 2) as previously reported (Caselmann, 1996), but the activation by HBx was not shown with pEboxtk, which suggests the repression of HBx through the E- box element (Figure 3a, lanes 3 and 4). In contrast, HBx activated the pmEboxtk at a level comparable to the pBLCAT2 plasmid (Figure 3a, lanes 5 and 6). These results showed that the transcriptional repression of the p53 promoter by HBx was mediated through the E-box element. The bHLH proteins mediate their eects on cellular gene expression through binding to a consensus DNA recognition sequence (5'-CANNTG-3'), called an E-box element, located in the promoter of target genes (Murre et al., 1989). These proteins have been grouped into three classes based upon the structural characteristic and the pattern of expression, and play an important role in the cell-speci®c expression of many dierent genes and, in some cases, the establishment of

Figure 2 Identi®cation of the HBx-repressive element in the p53 promoter. (a) Serial deletion mutant reporter plasmids of the p53 promoter were transfected with the HBx expression vector as indicated. CAT activities without (7) or with (+) HBx are indicated by the blank bars and solid bars, respectively. The basal CAT activity of p53-416 is designated 1.0 and each CAT activity is shown as a value relative to this basal activity. Error bars indicate standard errors. (b) The schematic representation of the transcription factor binding sites and approximate location in the p53 promoter. EBS, ETS-binding site; C/EBP, CCAAT/enhancer binding protein site; CPEp53, p53 core promoter element; arrow, major transcription start site. (c) Transient cotransfection assays were performed using each of the mutant p53 promoter-CAT Figure 1 The eect of HBx on the promoter of the human p53 reporter plasmids and the HBx expression vector as indicated. gene. Transcriptional activity of the p53-CAT reporter plasmid Methods: Deletion mutants (p53-416, p53-220, p53-128 and p53- (p53p1CAT) was measured in HeLa cells by cotransfection with 67) were prepared by PCR ampli®cation with p53p1CAT as a the HBx expression vector or adenovirus E1A as indicated. template and cloned into the pBLCAT2D plasmid. The p53-320 Methods: The eukaryotic expression plasmid of HBx was plasmid was constructed by insertion of the HindIII-XbaI constructed as described previously (Choi et al., 1999), and the fragment of the p53 promoter into the pBLCAT2D plasmid. p53p1CAT reporter plasmid was kindly provided by David The numbers in the names of the deletion mutants represent the Reisman (Roy et al., 1994). All transient transfections and CAT deletion end points of the 5' region of p53 promoter. The assays were performed as described previously (Choi et al., 1997). pBLCAT2D plasmid was constructed by deletion of the tk Typically, 2.5 mg of each reporter and 1 mg of each eector promoter in pBLCAT2. With p53-67 as a template, the p53 plasmid were used. The conversion of chloramphenicol to its promoter fragments for p53mCPE, p53mkB and p53mEbox were acetylated form was quanti®ed by using the BAS radio-analytic prepared by PCR and cloned into the pBLCAT2D plasmid. The imaging system according to the manufacturer's instructions. All p53mCPE, p53mkB and p53mEbox plasmids carry the 85 bp experiments were repeated at least three times. A representative promoter fragment from the p53 gene, with each mutation: the autoradiograph is shown, and the relative CAT activities CPEp53 (767 -GCAGGTATTGATGCGCTTAGGGTTTT- compared with the basal CAT activity in the absence of HBx 736), NF-kB binding site (749 -ATCGTTTTAGG-739) and are the means of three independent results. Standard deviations E-box element (734 -TCTGTG-729). The mutated sequences were 512% in all cases are underlined

Oncogene HBx represses the transcription of p53 SG Lee and HM Rho 470

Figure 4 The endogenous p53 gene is repressed in the HBV- infected cell line. Total RNA was prepared from exponentially growing control cells HepG2 and HepG2-K8 cells infected with HBV. The endogenous p53 mRNA level was analyzed by northern blot assay. The characteristic bands of 18S and 28S ribosomal RNA were observed. Lane 1, HepG2 and lane 2, HepG2-K8. Methods: Total RNA was extracted by the acid guanidinium thiocyanate-phenol-chloroform method and 20 mgof each sample were electrophoresed and stained with ethidium bromide to assess the quality and quantity of the total RNA in the samples (Chomczynski and Sacchi, 1987). Northern blot assay was performed as described previously (Kim et al., 1992). The a-32P-dCTP-labeled human p53 cDNA probe was used

absence (about ®vefold). The activity of the p53mEbox was neither activated by c-Myc, nor repressed by HBx in the presence of c-Myc (Figure 3b, lanes 4 ± 6). These results suggest that HBx also represses the transcriptional activation of the p53 promoter by the c-Myc, which binds to the E-box element. To determine whether the endogenous p53 gene is repressed in the HBV-infected cell, the p53 mRNA level was measured by northern blot assay in the human hepatoblastoma cell line HepG2 as a control and in the HepG2-K8 cell which was transfected with HBV and expressing HBx (Kim et al., 1992). As shown in Figure 4, the p53 gene was expressed in HepG2, but p53 mRNA was not detectable in HepG2-K8 under the experimental conditions used. In addition, the expres- Figure 3 (a) The plasmid pBLCAT2 has a tk promoter (tkp). sion of p53 gene also repressed in a HBx-transfected The plasmid pEboxtk and pmEboxtk have three copies of the E- cell line (data not shown). These results con®rmed that box and mutated E-box element, respectively, in front of tkp in the endogenous p53 gene was also repressed by HBx. the pBLCAT2. The HBx expression vector was transfected with Inactivation of the p53 gene, through mutations, pBLCAT2, pEboxtk or pmEboxtk. (b) The inhibitory eect of HBx on the transcriptional activation of the p53 promoter by c- direct interaction and transcriptional repression, is an Myc. Each plasmid of p53-67 and p53mEbox was transfected into important event in human cancer, including liver HeLa cells with c-Myc expression vector in the presence or cancer (Levine, 1993). Several DNA tumor viruses absence of HBx. Methods: The pEboxtk and pmEboxtk were have been shown to interact with the p53 protein, constructed by inserting trimerized double stranded E-box (5'- GATCCTCCCATGTGCTCAA-3') and mutated E-box (5'- functionally disabling p53 in the infected cells and GATCCTCCTCTGTGCTCAA-3') oligonucleotides, respectively, possibly contributing to virus-dependent transforma- into the BamHI site of pBLCAT2 tion. These include SV40 large T Ag, HPV E6, CMV IE2 and adenovirus E1B and E4orf6 (Teodoro and Branton, 1997). Recent studies also suggest that genes (Reisman and Rotter, 1993; Naya et al., 1995; retroviruses may dier from DNA tumor viruses by Mutoh et al., 1997). The promoter of the human p53 in¯uencing the transcription of the p53 gene rather tumor suppressor gene is also transactivated by the c- than by directly interacting with p53 protein. Myc/Max heterodimer, which binds to the E-box Transactivator protein Tat of the Human immunode- element (Roy et al., 1994). We therefore investigated ®ciency virus, Tax of HTLV type 1 and the core whether HBx aects the transactivation of the human protein of the hepatitis C virus are known to repress p53 gene by c-Myc. The c-Myc expression vector was the transcription of the p53 gene, suggesting a role in cotransfected with the plasmid containing the E-box the virus-associated pathology and malignant transfor- element in the presence or absence of the HBx mation (Li et al., 1995; Uittenbogaard et al., 1995; Ray expression vector. The c-Myc activated the p53 et al., 1997). promoter about threefold, presumably through the Here we demonstrated that HBx repressed the heterodimerization with endogenous Max protein transcription of the human p53 gene through the E- (Figure 3b, lane 2). However, in the presence of box element. HBx was also known to interact with the HBx, the c-Myc did not activate p53 promoter activity p53 protein, which mutually inhibits their functions (Figure 3b, lane 3). Rather it was more repressed in (Feitelson et al., 1993; Wang et al., 1994). In addition, the presence of c-Myc (about 7.5-fold) than in its the p53 protein was known to bind and repress the

Oncogene HBx represses the transcription of p53 SG Lee and HM Rho 471 HBV enhancer/X promoter (Ori et al., 1998). There- as XAP-1 (Yen, 1996; Caselmann, 1996; Becker et al., fore, HBx and p53 can inhibit each other by the 1998). protein-protein interaction and the transcriptional repression. The balance of the reciprocal inhibitions between these two proteins may play a decisive role in Acknowledgements the development of HBV-related malignancies. We thank Professor David Reisman of University of South Reduced levels of p53 would lead to suppression of Carolina for the kind gift of p53plCAT, p53 promoter- cell cycle arrest, leading to rapid progression through CAT reporter plasmid. This work was supported in part by research grants from the Korean Ministry of Education, the cell cycle without the requisite DNA damage repair from Molecular Medicine Research Program (98-J03-01- prior to genomic replication. This condition may be 01-A-04) of the Ministry of Science and Technology particularly detrimental to HBV-infected cells, as HBx (MOST) and from Korea Science and Engineering may disrupt the maintenance of genomic stability by Foundation (KOSEF) through the Research Center for binding damaged DNA and DNA repair proteins such Cell Dierentiation.

References

Aguilar F, Harris CC, Sun T, Hollstein M and Cerutti P. Murre C, McCaw PS, Vaessin H, Caudy M, Jan LY, Jan YN, (1994). Science, 264, 1317 ± 1319. Cabrera CV, Buskin JN, Hauschka SD and Lassar AB. Becker SA, Lee TH, Butel JS and Slagle BL. (1998). J. Virol., (1989). Cell, 58, 537 ± 544. 72, 266 ± 272. MutohH,FungBP,NayaFJ,TsaiMJ,NishitaniJand Bienz-Tadmor B, Zakut-Houri R, Libresco S, Givol D and Leiter AB. (1997). Proc.Natl.Acad.Sci.USA,94, 3560 ± Oren M. (1985). EMBO J., 4, 3209 ± 3213. 3564. Caselmann WH. (1996). Adv. Virus Res., 47, 303 ± 351. Naya FJ, Stellrecht CM and Tsai MJ. (1995). Genes Dev., 9, Choi BH, Park GT and Rho HM. (1999). J. Biol. Chem., 274, 1009 ± 1019. 2858 ± 2865. Ori A, Zauberman A, Doitsh G, Paran N, Oren M and Shaul Choi CY, Choi BH, Park GT and Rho HM. (1997). J. Biol. Y. (1998). EMBO J., 17, 544 ± 553. Chem., 272, 16934 ± 16939. Ray RB, Steele R, Meyer K and Ray R. (1997). J. Biol. Chomczynski P and Sacchi N. (1987). Anal. Biochem., 162, Chem., 272, 10983 ± 10986. 156 ± 159. Reisman D and Rotter V. (1993). Nucleic Acids Res., 21, Feitelson MA, Zhu M, Duan LX and London WT. (1993). 345 ± 350. Oncogene, 8, 1109 ± 1117. Roy B, Beamon J, Balint E and Reisman D. (1994). Mol. Furlong EE, Rein T and Martin F. (1996). Mol. Cell. Biol., Cell. Biol., 14, 7805 ± 7815. 16, 5933 ± 5945. Summers J and Mason WS. (1982). Cell, 29, 403 ± 415. Ganem D and Varmus HE. (1987). Ann. Rev. Biochem., 56, Sun X, Shimizu H and Yamamoto K. (1995). Mol. Cell. 651 ± 695. Biol., 15, 4489 ± 4496. Hsu IC, Metcalf RA, Sun T, Welsh JA, Wang NJ and Harris Teodoro JG and Branton PE. (1997). J. Virol., 71, 1739 ± CC. (1991). Nature, 350, 427 ± 428. 1746. Kim CM, Koike K, Saito I, Miyamura T and Gilbert J. Uittenbogaard MN, Giebler HA, Reisman D and Nyborg (1991). Nature, 351, 317 ± 320. JK. (1995). J. Biol. Chem., 270, 28503 ± 28506. Kim SH, Hong SP, Kim SK, Lee WS and Rho HM. (1992). J. Wang HD, Trivedi A and Johnson DL. (1997). Mol. Cell. Gen. Virol., 73, 2421 ± 2424. Biol., 17, 6838 ± 6846. Klein NP and Schneider RJ. (1997). Mol. Cell. Biol., 17, WangXW,ForresterK,YehH,FeitelsonMA,GuJRand 6427 ± 6436. Harris CC. (1994). Proc. Natl. Acad. Sci. USA, 91, 2230 ± Lane D. (1992). Nature, 358, 15 ± 16. 2234. Levine AJ. (1993). Ann. Rev. Biochem., 62, 623 ± 651. Weintraub H. (1993). Cell, 75, 1241 ± 1244. Li CJ, Wang C, Friedman DJ and Pardee AB. (1995). Proc. Yen TSB. (1996). J. Biomed. Sci., 3, 20 ± 30. Natl. Acad. Sci. USA, 92, 5461 ± 5464.

Oncogene