Akt augments the oncogenic potential of the HBx protein of hepatitis B virus by phosphorylation

Authors


V. Kumar, Virology Group, International Centre for Genetic Engineering & Biotechnology, Aruna Asaf Ali Marg, New Delhi 110 067, India
Fax: +91 11 26742316
Tel: +91 11 26741680 or +91 11 26742827
E-mail: vijay@icgeb.res.in
Website: http://www.icgeb.org/virology-vijay-kumar.html or http://www.icgeb.org/pi-vkumar.html

Abstract

Hepatitis B virus X protein (HBx) is a putative viral oncoprotein that plays an important role in various cellular processes, including modulation of the phosphatidylinositol 3-kinase/Akt signalling pathway. However, the molecular mechanism of Akt activation remains elusive. Here we show that HBx interacts with Akt1 kinase and is phosphorylated at serine 31 as indicated by mutational analysis of the Akt recognition motif (creating the HBxS31A mutant) or immunoblotting of HBx immunoprecipitates using Akt motif-specific antibody. The Akt-dependent phosphorylation of HBx was abrogated in the presence of the phosphatidylinositol 3-kinase inhibitor LY294002 or Akt1 gene silencing by specific siRNA. Co-immunoprecipitation studies provided evidence for HBx–Akt interaction in a cellular environment. This interaction was also confirmed in hepatoma HepG2.2.15 cells in which HBx was expressed at physiological levels from the integrated hepatitis B viral genome. The HBx–Akt interaction was essential for Akt signalling, and involved displacement of the Akt-bound negative regulator ‘C-terminal modulator protein’ by HBx. HBx-activated Akt phosphorylated its downstream target glycogen synthase kinase 3β, leading to stabilization of β-catenin, while p65 phosphorylation resulted in enhanced promoter recruitment and expression of target genes encoding cyclin D1 and Bcl-XL. Further, the oncogenic potential of HBx was significantly augmented in the presence of Akt in a soft agar colony formation assay. Together, these results suggest that oncogenic co-operation between HBx and Akt may be important for cell proliferation, abrogation of apoptosis and tumorigenic transformation of cells.

Structured digital abstract

Abbreviations
CAT

chloramphenicol acetyl transferase

CREB

cAMP response element-binding

CTMP

C-terminal modulator protein

ERK

extracellular signal regulated kinase

GAPDH

glyceraldehyde 3-phosphate dehydrogenase

GSK-3β

glycogen synthase kinase 3β

HBV

hepatitis B virus

HBx

HBV X protein/gene

NF-κB

nuclear factor κB

PI3K

phosphatidylinositol 3-kinase

RSV-LTR

Rous sarcoma virus-long terminal repeat

Introduction

Chronic infection with hepatitis B virus (HBV) is considered to be a major risk factor for development of hepatocellular carcinoma in humans [1]. One of the viral genes, known as HBx (encoding HBV X protein), is frequently integrated into the host genome and is expressed at low levels in hepatocellular carcinoma [2]. It is also reported to play an important role in HBV infection [3].

HBx is a multifunctional protein that exhibits diverse functions in cells, including activation of transcription, cellular responses to genotoxic stress, cell signalling pathways, cell proliferation or even regulation of proteasomal functions [2]. It is known to stimulate genes encoding proteins that are involved in cell proliferation, such as c-Jun, c-Fos, cyclin D1 and nuclear factor κB (NF-κB), or those associated with angiogenesis, such as vascular endothelial growth factor and interleukin 8 [4–9]. HBx interferes with the DNA repair process of cells and sensitizes cells to UV-induced DNA damage by up-regulating p53 levels [10,11]. HBx stimulates a number of signal transduction pathways, including mitogen-activated protein kinases, p38 kinases, stress-activated protein kinases, Jun kinases and Janus kinase/signal transducers and activators of transcription (JAK/STAT) pathways [3]. In addition, it can activate non-receptor tyrosine kinases of the Src family [12]. The invasive potential of cancerous cells, which is thought to be mediated by the phosphatidylinositol 3-kinase (PI3K)/Akt pathway, is reported to increase in the presence of HBx [13]. In addition, HBx augments the survival signal in synergy with Src kinase and PI3K activities [14]. Although the molecular mechanism of PI3K/Akt activation by HBx is not fully understood, it is suggested that the viral protein may involve cytokines such as interleukin-6, insulin-like growth factor II or epidermal growth factor in the process [15–17].

Activation of Akt (protein kinase B) sends a survival signal that abrogates apoptosis induced by withdrawal of growth factors [18]. Akt is known to phosphorylate pro-apoptotic factor Bad, which prevents the latter from binding and inactivating well-characterized apoptosis guards Bcl-2 and Bcl-XL, and eventually stops apoptotic cell death [19]. The anti-apoptotic action of HBx is mediated by the HBx–PI3K–Akt–Bad axis, which involves phosphorylation of the p85 regulatory subunit of the kinase [18,20]. HBx can also utilize other signalling pathways, such as the Src kinase–Ras–GTP complex and JAK/STAT, to indirectly activate the PI3K/Akt pathway [14,16]. The pleiotropic nature of PI3K/Akt signalling makes it favorable for viral replication, host cell survival and disease pathogenesis [21]. Further, the biological consequences of abnormal PI3K/Akt activation make it an important mediator of cancer progression [22]. The present study focuses on the functional interaction between HBx and Akt, and firmly establishes HBx as a substrate of Akt kinase. The molecular cooperation between HBx and Akt appears to play a key role in survival and oncogenic transformation of cells.

Results

HBx is a substrate of Akt kinase

Bio-informatic analysis of the HBx primary sequence using netphos-k software (http://www.cbs.dtu.dk/cgi-bin/nph-sw_request?netphos) predicted a putative Akt/PKB phosphorylation motif RGRPLSG(26–32) that conforms to the optimal Akt phosphorylation motif (RXRXXS/T) [23,24] (Fig. 1A). To confirm whether HBx is a substrate of Akt kinase, we used bacterially expressed recombinant wild-type HBx protein and the single point mutant HBxS31A (in which serine 31 is substituted by alanine) for in vitro phosphorylation in the presence of recombinant active Akt1. As shown in Fig. 1B, only the wild-type but not the mutant HBx was phosphorylated in the presence of Akt1 (compare lanes 2 and 3), suggesting that serine 31 is the site of phosphorylation. The specificity of the Akt phosphorylation motif on HBx was further confirmed by using recombinant HBx–GST protein (XG), in which 11 amino acids (amino acids 25–35) of HBx are fused to the C-terminus of glutathione S-transferase protein. Fig. 1C shows that XG was specifically phosphorylated by Akt1. As expected, the GST fusion protein with the mutant peptide (S31A) sequence was not phosphorylated (data not shown).

Figure 1.

In vitro phosphorylation of HBx at serine 31 by Akt kinase. (A) Alignment of the consensus Akt phosphorylation motif with the putative Akt motif in HBx. The serine to alanine mutation (HBxS31A) is also indicated. (B) In vitro phosphorylation of recombinant wild-type HBx and mutant HBxS31A proteins in the presence of activated rAkt1 kinase. The levels of phosphorylation of HBx (pHBx) are shown in the top panel (autoradiography) and relative density values are shown below the autoradiograph. The loading controls for rAkt1 and recombinant HBx and HBxS31A (Coomassie blue staining) are shown below. (C) In vitro kinase assay for GST (G) and HBx–GST (XG) (which includes 11 amino acids of HBx) was performed as in panel B. V, vector control.

Akt phosphorylates HBx in a cellular environment and in the HBV genome context

The specificity of HBx phosphorylation by Akt was further validated in a cellular environment using Akt motif-specific antibody [25]. Huh7 cells were transfected with the HBx or HBxS31A expression vectors, and immunoprecipitated using HBx specific IgG, followed by immunoblotting with Akt phosphorylation motif-specific antibody. As shown in Fig. 2A, there was a specific increase in phosphorylation of the Akt motif in HBx compared to the mutant HBxS31A. Pre-treatment of the cells with the PI3K/Akt inhibitor LY294002 abrogated HBx phosphorylation (Fig. 2B). As expected, the phosphorylation of extracellular signal regulated kinase (ERK) was unaffected under these conditions, as an almost threefold increase in its phosphorylation was observed upon serum stimulation. Further, Akt silencing by siRNA specifically attenuated HBx phosphorylation without affecting HBx levels (Fig. 2C).

Figure 2.

 Intracellular phosphorylation of HBx at serine 31 by Akt kinase. (A) Huh7 cells were transfected with control (V), HBx or HBxS31A expression vectors, and the immunoprecipitates (IP) and total lysates were subjected to western blotting (WB) using the indicated antibodies. (B) Cells were transfected with HBx, serum-starved for 48 h (Stv) or serum released in the presence of dimethylsulfoxide (DMSO), LY294002 or U0126, and analysed as in (A). (C) Cells were transfected with control or Akt1 siRNA, followed by HBx transfection, and processed as described in (A). (D) HepG2.2.15 cells were metabolically labelled with orthophosphate and immunoprecipitated using anti-HBx antibody. One group of cells were pretreated with LY294002 and one set of immunoprecipitates were treated with λ-phosphatase. Gels were dried and viewed using PhosphorImager. The relative density of phosphorylated bands is shown below the top panel. Control western blot panels are shown below. PI, pre-immune serum.

To show whether HBx expressed from the HBV genome is phosphorylated by Akt kinase, we used a HepG2.2.15 cell line that has integrated HBV genome and expresses the HBx gene under the control of a natural viral promoter [26]. Cells were metabolically labelled with [32P]-orthophosphate and immunoprecipitated using HBx antibody. As shown in Fig. 2D, the immunoprecipitated samples showed an intense 17 kDa band corresponding to HBx (lane 2) compared to the pre-immune control (lane 1). Treatment of cells with LY294002 drastically reduced the phosphorylation level (lane 3) and the reduction in the phosphorylation level of the immunoprecipitated sample by λ-phosphatase treatment suggests that it is a phosphoprotein (lane 4). Thus, these results indicate that HBx expressed under hepatocytic environment could be a phosphoprotein.

HBx interaction with Akt involves displacement of C-terminal modulator protein

After establishing that HBx is a substrate of Akt, we determined whether the two proteins interact in a cellular environment. The HBx–Akt interaction was investigated in HepG2.2.15 cells in which HBx is expressed at a physiological level. Cells were treated with the PI3K inhibitor LY294002, immunoprecipitated with either anti-HBx or anti-Akt antibodies, and immunoblotted for Akt or HBx binding. The results in Fig. 3A,B clearly indicate that HBx interacts with Akt1 kinase in a cellular environment. As HBx expression is associated with sustained activation of Akt kinase, we speculated whether such activation occurs by disruption of the Akt interaction with its negative regulator (known as the C-terminal modulator protein, CTMP) [27]. To assess this, we immunoprecipitated Akt from HBx- and HBxS31A-transfected cells and performed western blotting for HBx and CTMP. We observed that the HBx–Akt association precluded CTMP interaction with Akt1 (Fig. 3C), which may explain the sustained Akt1 activity in the presence of HBx [13]. Further, the HBxS31A mutant, which does not interact with Akt1, did not affect the Akt–CTMP interaction (Fig. 3C). Interestingly, the interaction of cAMP response element binding (CREB) protein with HBx and HBxS31A was unaffected under these conditions (Fig. 3D). Thus, HBx associates with Akt by displacing its negative regulator CTMP.

Figure 3.

 Association of HBx with Akt. (A) HepG2.2.15 cells were treated with dimethylsulfoxide (DMSO) or LY294002 and immunoprecipitated using anti-HBx antibody followed by western blotting for total and phosphorylated Akt. (B) HepG2.2.15 cells were immunoprecipitated with anti-Akt antibody followed by western blotting for HBx and Akt. (C) Huh7 cells were transiently transfected with vector control, HBx or HBxS31A expression plasmids, immunoprecipitated using anti-Akt1 antibody and western blotted for HBx, Akt1 and CTMP. (D) Huh7 cells were transiently transfected as described in (C), immunoprecipitated with anti-HBx antibody and western blotted for cAMP response element binding (CREB) protein and HBx. PI, pre-immune serum.

HBx and HBxS31A differentially modulate mitogenic pathways

As phosphorylation may have a profound effect on the structure and functions of proteins, we next determined whether HBx phosphorylation by Akt has a modulatory effect on various mitogenic pathways. In accordance with previous reports, we found that wild-type HBx activated the ERK, Akt and Src kinase pathways (Fig. 4A). Interestingly, the mutant HBxS31A did not activate the PI3K/Akt pathway but ERK and Src signalling were similar to that of native HBx (Fig. 4A). This differential effect of HBxS31A on Akt activation was further investigated with respect to downstream targets such as glycogen synthase kinase 3β (GSK-3β) and transcription factor p65 [25,28,29]. Analysis of the phosphorylation status of these proteins suggested that the levels of GSK-3β phosphorylated at serine 9 and of p65 phosphorylated at serine 276 were elevated in HBx-expressing cells but not in HBxS31A-transfected cells (Fig. 4B). These phosphorylation signatures facilitate transcriptional initiation by p65 [28,29]. Further, the level of β-catenin, which is a GSK-3β target, was also stabilized in the presence of native HBx (Fig. 4B).

Figure 4.

 Differential activation of mitogenic pathways by HBx and HBxS31A. (A,B) Huh7 cells were transiently transfected with vector (V), HBx or HBxS31A plasmids, and western blotted for phosphorylated forms of Akt, Erk and Src (A) or GSK--3β, β-catenin and cyclin E (B). (C) Cells were transfected as described in (A) together with RSV–CAT or NF-κB–CAT reporter constructs, and reporter gene activity was measured after 48 h. (D,E) Cells were transfected as described in (A), and the expression of cyclin D1 and Bcl-XL genes was determined by RT-PCR (D) or monitored for the recruitment of p65 to cyclin D1 and Bcl-XL promoters by ChIP/real-time quantitative PCR (E). Values are means ± SD of triplicate determinations. *Statistically significant at < 0.05 with respect to vector control; #not statistically significant (> 0.1).

The differential effect of HBx and HBxS31A on cell signalling was also evident from their ability to transactivate Rous sarcoma virus-long terminal repeat (RSV-LTR) and the NF-kB promoter element. Although RSV-LTR can be driven by multiple cell signalling pathways, the NF-kB element is specifically stimulated by the Akt pathway [29,30]. Accordingly, we observed that, although cells expressing HBx and HBxS31A showed activation of the RSV–CAT (chloramphenicol acetyl transferase) reporter gene to similar levels, HBxS31A did not activate the NF-κB CAT reporter but HBx did (Fig. 4C). Thus, these results suggest that HBxS31A mutation has no effect on the transactivation potential of HBx, but its Akt activation potential is abrogated.

As cyclin D1 and Bcl-XL are not just direct transcriptional targets of p65 but also play a crucial role in various aspects of carcinogenesis [4,31], we next determined the real-time expression of these genes in HBx- and HBxS31A-transfected cells. As shown in Fig. 4D, expression of cyclin D1 and Bcl-XL was significantly increased in the presence of HBx compared to HBxS31A. Further, the chromatin immunoprecipitation/real-time quantitative PCR results confirmed enhanced recruitment of p65 to cyclin D1 and Bcl-XL promoters only in the presence of HBx (Fig. 4E). Thus, the wild-type HBx and mutant HBxS31A differ significantly with regard to Akt activation and its downstream events, without altering the Src and MAPK pathways.

Physiological relevance of Akt activation by HBx

As Akt signalling is critical for cell growth and survival, and the present results suggest differential modulation of this pathway by HBx and HBxS31A, we used a cell transformation assay in soft agar to study the growth-supporting role of Akt in the HBx micro-environment. Figure 5A shows the colonies of HBx and Akt transfectants in the soft agar assay. HBx and Akt appeared to cooperate, resulting in more frequent and larger colonies, perhaps through potentiation of the oncogenic activity of HBx. As shown in Fig. 5B,C, co-transfection of HBx and Akt resulted in a significant increase (< 0.05) in the transformation of NIH-3T3 and AML-12 cells compared to HBxS31A (> 0.1). Thus, HBx and Akt appear to cooperate in directing the cellular machinery towards uncontrolled growth and proliferation.

Figure 5.

 Oncogenic co-operation between HBx and Akt kinase in cell transformation. NIH-3T3 cells were transiently transfected with expression vectors for HBx, HBx with Akt, HBxS31A or HBxS31A with Akt, and used for the soft agar assay after 48 h. (A) Microscopic analysis of colony formation. (B,C) Mean number of foci in 20 fields on each plate (B) and the relative size of the colonies obtained under each condition (C). Values are means ± SD of triplicate determinations. *Statistically significant at < 0.05 in comparison to HBx alone; #not statistically significant (> 0.1).

Discussion

Akt is a serine/threonine kinase that acts downstream of PI3K. Initially, Akt was described as an oncogene that was activated by serum and a variety of growth factors, such as platelet-derived growth factor, epidermal growth factor, bovine fibroblast growth factor, insulin and insulin-like growth factor [25]. The cellular targets of Akt carry a consensus recognition motif (RXRXXS/TB), where X is any amino acid and B is a bulky hydrophobic residue [24,25]. Akt is known inactivate its target, mainly pro-apoptotic molecules, via phosphorylation, resulting in its pro-survival effect. The major targets of Akt include pro-apoptotic factors such as apoptosis signal regulating kinase 1, Bad, Bax and FoxO transcription factors, and human caspase 9 [22,25]. Akt also phosphorylates a variety of other substrates, such as GSK-3β, insulin receptor substrate 1, NF-κB transcription factors, Raf protein kinase, BRCA1 and p21Cip1, and thus is involved in the regulation of cell growth, glucose metabolism, the cell cycle and protein translation [22,25]. HBx is also reported to elicit its anti-apoptotic effect via the PI3K/Akt signalling cascade [13]. However, the molecular mechanisms of cooperation between HBx and Akt are not known. The present study uncovers a new mechanism whereby Akt1 phosphorylates HBx and regulates its diverse functions associated with cell proliferation. Akt naturally exists in three isoforms: Akt 1, 2 and 3. Generally all isoforms have common substrates characterized by their Akt phosphorylation motifs, but they vary in function due to differences in their tissue-specific expression and intracellular localization. Akt1 is the major isoform found in the cytoplasm, while Akt2 is associated with the mitochondria and Akt3 is mostly found in the nuclei of brain tissues. Given the profound influence of HBx on mitochondrial functions, it will be interesting to investigate the role of other Akt isoforms in the HBx micro-environment [25].

The present study was initiated following identification of the putative Akt phosphorylation motif RGRPLSG(26–32) in HBx using netphos-k software (http://www.cbs.dtu.dk/cgi-bin/nph-sw_request?netphos). Our immunoprecipitation studies of metabolically labelled HBx with motif-specific antibody suggested that HBx is a target of Akt phosphorylation in the cellular environment (Fig. 2A). The involvement of Akt in this process was further evident from the inhibition of HBx phosphorylation in the presence of LY294002 (a PI3K inhibitor acting upstream of the Akt pathway) or silencing of Akt1 expression (Fig. 2B,C). Further, our in vitro phosphorylation studies using recombinant Akt1 and HBxS31A or XG mutants confirmed that serine 31 is the Akt-specific phosphorylation site in HBx (Fig. 1). Careful analysis of the HBx sequences in the Genbank database showed that serine 31 of HBx is substituted by proline in some viral isolates. Thus, it will be interesting to study the activation kinetics of Akt on proline-substituted HBx in the context of the hepatitis viral genome.

Akt-dependent phosphorylation of HBx led to an association between HBx and Akt. The HBx–Akt interaction was dependent on the activation status of Akt, and involved competitive displacement of its inhibitory partner CTMP by HBx (Fig. 3). In turn, the activation of Akt kinase created a positive feedback circuit between HBx and Akt signalling. Previous studies have shown that HBx can stimulate many other kinases, including Src and ERK/MAPK [7,14]. However, the present study appears to suggest that phosphorylation of HBx at serine 31 may be a driving mechanism for Akt activation without affecting other signalling pathways [3]. It will be important to determine the identity of the effector kinase triggered by HBx that phosphorylates Akt, and whether the activity of that kinase could be modulated by drug therapies. Thus, use of HBxS31A served to elucidate the contribution of Akt with regard to other signalling pathways in HBx-dependent cellular transformation.

It has been reported previously that HBx forms a complex with 14-3-3 scaffolding proteins through the RXRXXpS (26–31) motif present in HBx, leading to their co-localization in the cytoplasm [32]. The authors discussed the role of an unknown kinase in HBx phosphorylation and its cytoplasmic retention. Further, mutation of the above motif abrogated the interaction between HBx and 14-3-3 proteins. Our observation that HBx interaction with 14-3-3 proteins is lost when cells were pre-treated with PI3K inhibitor suggested the class of kinase involved (Fig. S1). Thus, we identified the unknown kinase associated with HBx phosphorylation as Akt kinase.

A number of viruses, such as HBV, hepatitis C virus, Epstein–Barr virus and human papilloma virus, are known to produce chronic or latent infections, and result in cellular transformation during the course of infection. Viral-encoded proteins, such as HBx from HBV, NS5A from hepatitis C virus, LMP1 and LMP2A from Epstein–Barr virus and E5/E7 from human papilloma virus, have been reported to up-regulate the PI3K/Akt signalling pathway in order to prevent apoptosis of the cells they infect [20,33–36], which may contribute to uncontrolled proliferation of host cells. The PI3K/Akt signalling pathway is frequently up-regulated in many human cancers (including gastric, renal, breast, colo-rectal, tongue and glioma), resulting in poor prognosis and resistance to chemotherapy, radiotherapy and/or target-based therapies [22]. Thus, the PI3K/Akt pathway may be a promising target for developing anti-cancer therapies.

The transcriptional activation of cyclin D1 and Bcl-XL genes in the HBx micro-environment mediated by Akt signalling strongly suggested that both proliferative and anti-apoptotic functions were involved during cell transformation. The oncogenic cooperation between HBx and Akt kinase was further evident from the high frequency of transformation of NIH-3T3 cells and colony formation in a soft agar assay. The oncogenic cooperation between HBx and cellular oncogenes such as c-myc is well established. HBx alone was non-tumorigenic, but hepatic tumour growth was accelerated in the presence of c-myc [37]. Further, HBx expression is associated with anchorage-independent growth of AML-12 murine hepatocytes [38], and, together with H-ras, HBx can induce neoplastic transformation of cells that can overcome transforming growth factor β-induced apoptosis [39]. According to a previous report, the N-terminal region (1–50) of HBx is important for cellular transformation [40]. The present study thus seems to bridge the existing gap in understanding of the oncogenic potential of HBx by identification of an active Akt phosphorylation site in this region. Further, the measurable support for colony formation by mutant HBxS31A suggested the non-exclusive nature of Akt in cell transformation and the involvement of other signalling pathways in this process. Nevertheless, our study provides an important link between the Akt-dependent survival pathway and HBx, which may be crucial for hepatocarcinogenesis.

Experimental procedures

Expression vectors

The expression vector for wild-type HBx has been described previously [30]. The HBxS31A mutant was generated using a QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA, USA) and primers 5′-GACGACCCCTCGCCGGCCCGCTTGG-3′ (forward) and 5′-CCAAGCGGGCCGGCGAGGGGTCGTC-3′ (reverse). The HBx–IRES-EGFP and HBxS31A–IRES-EGFP recombinants were generated by cloning the EcoRI fragments of HBx and HBxS31A into the eukaryotic expression vector pIRES-EGFP (Clontech, Mountain View, CA, USA). These fragments were also cloned into the pET-14b vector (Novagen, EMD4Biosciences, Darmstadt, Germany) to create prokaryotic expression vectors for HBx and HBxS31A. The HBx–GST fusion recombinant XG was produced by in-frame cloning of oligonucleotides corresponding to residues 25–35 of HBx in the pGEX-2T prokaryotic expression vector (Amersham Biosciences, Amersham, UK). The oligonucleotide sequences were 5′-GGATCCGGACGACCCCTCGCCGGCCCGCTTGGGCCCGGG-3′ (forward) and 5′-CCCGGGCCCAAGCGGGCCGGCGAGGGGTCG TCCGGATCC-3′ (reverse). The expression vector for constitutively active Akt (Akt-CA) [41] was kindly provided by Alexandra D. Newton (Department of Pharmacology, University of California, San Diego, CA, USA). The reporter gene RSV–CAT was a kind gift from I. Pastan (Molecular Biology Section, National Institutes of Health, Bethesda, MD, USA) [42],and the interleukin 2 receptor-derived reporter NF-κB–CAT was obtained from Ranjan Sen (National Institutes of Ageing, Baltimore, MD, USA).

Chemicals and other reagents

The chemicals used and their working concentrations are listed below. The MAP kinase inhibitor U0126 (10 μm) and the PI3K inhibitor LY294002 (10 μm) were from Calbiochem (San Diego, CA, USA). [γ-32P]-ATP and [32P]-orthophosphate were supplied by Perkin Elmer Life Sciences (Cambridge, Massachusetts, USA). For affinity chromatography, Ni-NTA, glutathione, protein A and protein A agarose beads were obtained from Amersham Biosciences. Antibodies for Akt, ERK, β-catenin, p65, glyceraldehyde 3-phosphate dehydrogenase, GSK-3β, pGSK-3β (S9), CTMP and proliferating cell nuclear antigen (PC10) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Phosphorylated pAkt (S473 residue), phosphorylated ERK1/2, phosphorylated p65 (S276 residue) and the activated recombinant Akt1 kinase were purchased from Merck (Darmstadt, Germany). The control and Akt1 siRNA, anti-pSrc and anti-pAkt substrate motif antibodies were obtained from Cell Signaling Technology Inc. (Danvers, MA, USA). Monoclonal antibody against HBx has been described previously [40].

Cell culture and transfection

Human hepatoma Huh7 cells, AML-12 immortalized mouse hepatocytes and NIH-3T3 cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) with antibiotics and 10% fetal bovine serum (Hyclone, Barrington, IL, USA), and seeded at a density of 0.7 million per 60 mm dish. Transfections of DNA and siRNA were performed using Lipofectamine (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s protocol. HepG2.2.15 cells were grown in DMEM supplemented with 5 μg·mL−1 bovine pancreatic insulin (Sigma, St Louis, MO, USA).

Metabolic labelling of cells, immunoprecipitation and western blotting

Cells were grown in phosphate-free medium (Dulbecco’s modified Eagle’s medium without sodium phosphate supplemented with 10 mm HEPES and antibiotics) for 2 h, followed by labelling with [32P]-orthophosphate (1 mCi·mL−1, #PB-13; GE Healthcare, Munich, Germany) for a further 2 h. For one set of cells, LY294002 (10 μm) was added during the last 30 min of the labelling period. The cells lysates were immunoprecipitated using the indicated antibodies and resolved by 15% SDS/PAGE followed by western blotting [43]. The protein bands were visualized using PhosphorImager (Amersham Biosciences) and quantified by densitometry.

Purification of HBx, HBxS31A and XG, and in vitro kinase assay

The HBx, HBxS31A and XG recombinants were expressed in Escherichia coli DH5α cells by isopropyl thio-β-d-galactoside induction. Hexahistidine-tagged HBx and HBxS31A recombinant proteins were purified by Ni-NTA affinity chromatography according to the manufacturer’s instructions (Amersham Biosciences). The recombinant XG and GST proteins were purified using glutathione agarose beads according to the manufacturer’s protocol. The in vitro kinase assay was performed using 500 ng recombinant Akt1 kinase (rAkt1) and 5 μg of indicated substrate (recombinant HBx, HBxS31A, recombinant GST or XG proteins) as described previously [43]. Samples were resolved by 15% SDS/PAGE, dried and viewed using PhosphorImager as described above.

Colony formation assay

The colony formation assay was performed as described by Kim et al. [38]. The plates were incubated at 37 °C for 14 days and observed regularly for colony formation under the microscope. Images were taken using a Nikon TE2000S phase-contrast microscope (Tokyo, Japan) at 200× magnification.

Chloramphenicol acetyltransferase assay (CAT assay)

The CAT assay was performed as described previously [40]. CAT activity was normalized against total protein. The relative CAT activity was expressed as means ± SD for three independent observations.

Chromatin immunoprecipitation (ChIP) and real-time quantitative PCR

The methods for chromatin immunoprecipitation (ChIP) and real-time quantitative PCR have been described previously [44]. The levels of cyclin D1 and Bcl-XL transcripts were measured by RT-PCR. The RT-PCR primers were obtained from Qiagen (Hilden, Germany). The ChIP/real-time quantitative PCR primers spanning the NF-κB elements in the cyclin D1 and Bcl-XL promoters were 5′-CCGGCTTTGATCTCTGCTTA-3′ (forward) and 5′-GCTGTACTGCCGGTCTCC-3′ (reverse) for cyclin D1, and 5′-ACAGATCCGAGGCTGTCTTC-3′ (forward) and 5′-CCCGGAGGTATGGGTTTTAGT-3′ (reverse) for Bcl-XL. Real-time quantitative PCR was performed in the presence of SYBR Green (Roche, Indianapolis, IN, USA). Values are means ± SD of three independent experiments.

Statistical analysis

Statistical significance was calculated using Student’s t test. P values < 0.05 were considered significant.

Acknowledgements

This work was supported by a core grant from the International Centre for Genetic Engineering and Biotechnology, New Delhi, India. We are grateful to the following scientists for the generous gift of various recombinants: A.C. Newton (Department of Pharmacology, University of California, San Diego, CA) for Akt-CA, I. Pastan (Molecular Biology Section, National Institutes of Health, Bethesda, MD) for RSV–CAT, and R. Sen (Laboratory of Molecular Biology and Immunology, National Institute of Ageing, Baltimore, MD) for the NF-κB–CAT reporter. E. Khattar is a senior research fellow of the Council of Scientific and Industrial Research, New Delhi, India.

Ancillary