E3 ubiquitin ligase Siah-1 facilitates poly-ubiquitylation and proteasomal degradation of the hepatitis B viral X protein

Abstract

SIAH1 binds to HBx by pull down (View interaction) Hepatocellular carcinoma (HCC) is a common malignancy and chronic hepatitis B virus (HBV) infection is one of the major risks of developing HCC, which accounts for more than 80% of HCC cases in high-incidence region [1]. The pathogenesis and carcinogenesis of HBV-associated HCC mainly involve continuous liver inflammation response to virus infection and the integration of HBV DNA into the host genome to induce direct mutagenesis of diverse cancer-related genes [2]. In many cases of HCC, the hepatitis B × gene is commonly integrated among the HBV genes. Hepatitis B viral X protein (HBx) can associate with many proteins to transactivate a variety of viral and cellular genes, and affect gene transcription, intracellular signal transduction, genotoxic stress response, protein degradation, cell cycle control and apoptosis [3]. Comprehensive studies have suggested that HBx is a promiscuous transactivator and is involved in HBV replication and hepatocarcinogenesis [4, 5]. A high incidence of HCC was observed in transgenic mice expressing HBx protein [6, 7] and knock-down of HBx expression in Hep3B cells markedly reduced xenograft tumor growth [8]. Therefore, HBx expression level is crucial in the pathogenesis of HCC. The stability of HBx protein is regulated in vivo. HBx can be actively ubiquitinated and undergo proteolysis through the ubiquitin-proteasome pathway [9]. It has been discovered that 26S proteasome complex [9], heat shock protein 40 (Hsp40) [10], HBV core proteins [11], tumor suppressor p53 [12], transcriptional factor Id-1 [13] and MDM2 [14] contribute to destabilizing HBx protein in a proteasome-dependent manner. However, no specific E3 ubiquitin ligase(s) accounted for ubiquitylation-dependent proteasomal degradation of HBx protein has been identified until now. Id-1 has been demonstrated to destabilize HBx protein by facilitating the interaction between ubiquitinated HBx and proteasome subunit C8, while Id-1 itself has no influence on the ubiquitylation modification of HBx [13]. Tumor suppressor p53 is capable of increasing HBx ubiquitylation with an unknown mechanism [14]. Our present study attempts to search for the specific E3 ubiquitin ligase that mediates HBx ubiquitylation and degradation. The Siah-1/2 are the mammalian homologues of Drosophila Seven in absentia (Sina) protein [15]. The N-terminus of Siah members encoding Ring-finger domain can associate with ubiquitin conjugating enzymes (E2s), conferring Siah proteins the E3 ubiquitin ligase activity; and the C-terminal regions are the substrate-binding domain (SBD) that can interact with various proteins, some of which can be degraded [16]. Siah-1 and Siah-2 are two human Siah members, which are highly conserved and have some overlapping functions. Through mediating the cellular proteins for ubiquitin-proteasome degradation, Siah-1/2 are involved in diverse biological processes, including apoptosis [17, 18], mitosis [19], estrogen signaling [20], Ras pathway [21, 22], hypoxia [23], DNA damage [24], etc. Siah-1/2 can form dimer with itself or their homologues and regulate their own availability via self ubiquitylation and degradation [16]. Additionally, Siah-1 may act in a complex with Skp1, Eb1, SIP and pAPC to promote β-catenin for ubiquitin-mediated degradation, which may reduce malignant cell proliferation and oncogenesis [25, 26]. Recent study found that ORF45 protein encoded by Kaposi's sarcoma-associated herpesvirus (KSHV) was also the substrate of Siah-1, implicating Siah-1 as a modulator of virus infection [27]. In this study, we identified Siah-1 as a novel E3 ligase to promote HBx poly-ubiquitylation and proteasomal degradation. Moreover, p53-mediated HBx reduction was also dependent on Siah-1 expression. Therefore, our findings suggest that Siah-1 may play an important role in HBx degradation via ubiquitin-dependent pathway and may be involved in suppressing the progression of HCC. Hep3B cells were cultured in RPMI 1640, 293T and HepG2 cells were grown in Dulbecco's modified Eagle's medium (DMEM) at 37 °C in 5% CO2-humidified atmosphere. Both mediums were supplemented with 10% fetal bovine serum (FBS). Hep3B and HepG2 cells are two human HCC cell lines, and Hep3B is deficient in p53 expression whereas HepG2 cell has wild type p53 expression. Cells at 80% confluency were transfected with plasmids using Lipofectamine2000™ reagent (Invitrogen) according to manufacturer's instructions. pCMV-Myc-HBx was kindly provided by Dr. Minghua Zhu (Changhai Hospital, the Second Military Medical University of China). HBx cDNA was subcloned into pcDNA3.1b(−) (Invitrogen) and pET-32a-His vectors (Novagen). The cDNA encoding human Siah-1 and Siah-2 were PCR amplified from human liver marathon cDNA library, and subcloned into PCMV-Myc vector (Clontech). Catalytically inactive mutant Siah-1-C44S were generated using TaKaRa Mutagenesis Kit. The wild type and mutant were also introduced into pCMV-HA (Clontech) and pGEX-4T-2 vectors (Amersham Biosciences) for expressing HA and GST fusion protein, respectively. The glucocorticoid response element (GRE), cAMP response element (CRE) and heat shock response element (HSE)-luc constructs were purchased from Clontech. Specific antibodies used in this paper were as follows: anti-HBx, anti-GFP and DO-1 for anti-p53 (Santa Cruz, USA), anti-Siah-1 (Abcam, USA), anti-HA (Convance, USA), anti-Myc and anti-β-Actin (Sigma, USA), anti-GST and anti-His (Novagen, Germany). The RNAi oligos were purchased from Genepharma (Shanghai, China). The two RNAi oligos sequences for Siah-1 are: 1#-siRNA Siah-1: 5′-ATGAGCCGUCAGACUGCUACAdTdT-3′; 2#-siRNA Siah-1: 5′-GAUAGGAACACGCAAGCAAdTdT-3′; The RNAi oligos sequences for p53 are: 5′-GACUCCAGUGGUAAUCUACdTdT-3′; The RNAi oligos for negative control are: 5′-AACUCCUGCCUCCUUAUGUAUUUdTdT-3′; The siRNA oligos were transfected using the Lipofectamine™ 2000 reagent (Invitrogen, USA) as described below. Diluted 1 μl Lipofectamine™ 2000 and 50 μl Opti-MEM were mixed gently and incubated for 5 min at room temperature. Then, 20 pmol siRNA oligos and the corresponding co-transfected plasmids were diluted in 50 μl Opti-MEM. After 5-min incubation, the diluted oligos and plasmids with the diluted Lipofectamine™ 2000 were mixed gently and incubated for 20 min before adding the complexes to each well. The effect of gene knockdown was detected by Western blot. These experiments were performed as described previously [14]. GST-Siah-1, His-HBx and four His-HBx constructs were expressed in E. coli, and purified according to the manufacture's protocol (Amersham Pharmacia Biotech). Hep3B cells cultured on coverslips were fixed in PBS with 4% paraformaldehyde and permeabilized in 0.2% Triton X-100. The coverslips were blocked with 5% normal goat serum plus 2% BSA, and then incubated with mouse anti-Myc monoclonal antibody or rabbit anti-HA polyclonal antibody, respectively. We then sequentially incubated the coverslips with goat anti-mouse-Alexa 555 secondary antibody or the Alexa Fluor® 488 goat anti-rabbit IgG secondary antibody, and with DAPI. All images were acquired using a LEICA DC 500 camera microscope equipped with LEICA DMRA2 fluorescent optics. The ubiquitylation assay in vivo was performed by transfecting 293T cells in 6-well with 300 ng pGFP–N3, 400 ng HA-ubiquitin, 4 μg pcDNA3.1b(−)His-HBx and together with 6 μg Myc-Siah-1 or Myc-Siah-1-C44S plasmids as indicated. The pcDNA3.1b(−)vector was used to keep equal amounts of plasmid DNA in each transfection. Twenty hours later the cells were treated with proteasome inhibitor MG132 (30 μM) for 6 h and then lysed in buffer A (6 M guanidinium-HCl, 0.1 M Na2HPO4/NaH2PO4, 0.01 M Tris–HCl pH 8.0, 5 mM imidazole, 10 mM β-mercaptoethanol) and incubated with Ni2+-NTA beads for 4 h at room temperature. Beads were washed with buffer A, B (8 M urea, 0.1 M Na2HPO4/NaH2PO4, 0.01 M Tris–HCl pH 8.0, 10 mM β-mercaptoethanol) and C (8 M urea, 0.1 M Na2HPO4/NaH2PO4, 0.01 M Tris–HCl pH 6.3, 10 mM β-mercaptoethanol), and bound proteins were eluted with buffer D (200 mM imidazole, 0.15 M Tris–HCl pH 6.7, 30% glycerol, 0.72 M β-mercaptoethanol, 5% SDS) and analyzed by Western blot. HBx protein can be modified by poly-ubiquitylation and degraded rapidly through ubiquitin-proteasome pathway in vivo [9-13]. However, E3 ubiquitin ligase responsible for HBx ubiquitin-dependent degradation is unknown. Tumor suppressor p53 can increase HBx ubiquitylation and promote HBx degradation with unknown mechanism [12]. Considering that p53 is a transcriptional activator but not E3 ubiquitin ligase itself, p53 can degrade proteins by inducing E3 ubiquitin ligases expression. For example, p53 destabilizes the p21WAF1 by inducing the expression of Ring-finger E3 ubiquitin ligase p53RFP [28]. Siah-1 has been reported to be a p53-inducible E3 ubiquitin ligase [17, 25, 26], and the expression of Siah-1 was downregulated in advanced HCC [29]. Moreover, Siah-1 is capable of mediating herpesvirus-encoded ORF45 protein for ubiquitin degradation [27]. Therefore, we investigated whether Siah-1 was one of the candidate E3 ubiquitin ligases for the ubiquitin-dependent degradation of viral protein HBx. To verify the role of Siah-1 in HBx degradation, we also constructed a Siah-1 mutant with cysteine-to-serine substitution at codon 44 in its N-terminal Ring domain (Siah-1-C44S). Previous studies have reported that the Siah-1-C44S mutant abrogated the E3 ubiquitin ligase activity of Siah-1 to promote substrate DCC degradation and rapid self-degradation, and may have dominant negative effect on endogenous Siah-1 [16, 22]. In Hep3B HCC cells (p53 defective), we found that HBx protein level was significantly reduced in a dose-dependent manner when coexpressed with Siah-1 but not with the catalytically inactive mutant Siah-1-C44S (Fig. 1 A). Additionally, the other homologue Siah-2 also had similar impact on the reduction of HBx protein level (Fig. 1B), while another p53-inducible Ring-domain E3 ubiquitin ligase Pirh2 had no effect on HBx stability (Fig. 1C). These results indicated that Siah-1/2 has potential regulatory effect on HBx protein level and the ubiquitin ligase activity of Siah-1 was required for this process. Because of the high similarity in structure and function between Siah-1 and Siah-2 [15], our present study only focused on Siah-1. Moreover, treatment with proteasome inhibitor MG132 could rescue HBx reduction mediated by Siah-1 (Fig. 1D), indicating that Siah-1 decreased HBx level in a proteasome-dependent manner. Meanwhile, knockdown of endogenous Siah-1 by two specific siRNA resulted in higher HBx protein level compared with the control (Fig. 1E), further supporting that HBx protein is down-regulated by Siah-1. To further demonstrate that Siah-1 regulated HBx abundance at post-translational level, we also determined the effects of Siah-1 and the Siah-1-C44S mutant on HBx protein turnover. The relative stability of the HBx protein was examined at the indicated interval after treatment with cycloheximide (CHX) to block protein synthesis. As shown in Fig. 1F, coexpression of Siah-1 accelerated HBx protein degradation. In contrast, HBx was stabilized by the Siah-1-C44S mutant, perhaps because the Siah-1-C44S mutant may act as a dominant negative mutant abrogating the activity of endogenous Siah-1. Moreover, co-expressed Siah-1 attenuated the glucocorticoid response element (GRE), heat shock response element (HSE), cAMP response element (CRE) signal pathways transactivated by ectopically expressed HBx, while these signal pathways were not affected by Siah-1 over-expression alone (Fig. 1G). This implied that HBx-mediated oncogenic signals could be inhibited by Siah-1. Thus, the above results suggest that E3 ubiquitin ligase Siah-1 promotes the proteasomal degradation of HBx protein in vivo and counteracts HBx transactivation activity. Siah-1 carries out the E3 ubiquitin ligase function by directly or indirectly interacting with substrates [16, 25]. Accordingly we employed a co-immunoprecipitation assay to investigate whether Siah-1 interacted with HBx in vivo. HA-Siah-1 and Myc-HBx were transfected into 293T cells alone or in combination. Cells were treated with the proteasome inhibitor MG132 for 6 h to maintain the proteins levels before harvest. As shown in Fig. 2 A, when immunoprecipitated with anti-HA antibody, Myc-HBx was detected in the co-immunoprecipitated complex from cells co-transfected with HA-Siah-1 (Fig. 2A, middle panel), and consistent result was observed in a reciprocal experiment by using anti-Myc antibody (Fig. 2A, right panel). Meanwhile, we found the Siah-1-C44S mutant also retained the ability to interact with HBx (Fig. 2B). Further, GST pull-down result confirmed that His-HBx directly interacted with GST-Siah-1 but not GST tag in vitro (Fig. 2C). To verify the interaction between Siah-1 and HBx in vivo, we further investigated whether Siah-1/Siah-1-C44S mutant and HBx proteins colocalized in the same subcelluar compartment via immunofluorescence analyses. HA-Siah-1/HA-Siah-1-C44S mutant and Myc-HBx were transfected alone or in combination, into Hep3B cells. Immunofluorescence analysis (Fig. 2D) showed that Myc-HBx alone was mainly detected in cytoplasm at the periphery of the nuclear envelope, and HA-Siah-1 and HA-Siah-1-C44S mutant alone predominantly exhibited the punctuated cytoplasmic staining. The co-expressed HBx and wild type Siah-1 colocalized in cytoplasm with a perinuclear distribution at the small speckle patterns; the HBx and Siah-1-C44S mutant also appeared to have the speckle colocalization distribution in the cytoplasm. These findings provided further evidence to support the interaction between HBx and Siah-1 in vivo. To delineate the Siah-1-interacting domain within HBx protein, four His-HBx truncation mutants were constructed (Fig. 2E, left panel). GST pull-down assay indicated that His-HBxΔ2,Δ3,Δ4 mutant maintained similar capability of binding to GST-Siah-1 as full length one, but Δ1 mutant failed to do so (Fig. 2E, right panel). This result suggested the central transactivation domain of HBx [51–136 amino acid (aa)] is indispensable for the interaction with Siah-1. Siah-1 has been confirmed to catalyze the poly-ubiquitylation of most substrates. HBx protein could be actively ubiquitinated and both ubiquitin-dependent and -independent proteasomal degradation were operative in its turnover [30]. To determine whether Siah-1 affected HBx poly-ubiquitylation modification, we performed an in vivo ubiquitylation assay. As shown in Fig. 3 , a strong smear of HBx poly-ubiquitylation bands was detected when co-expressed with Myc-Siah-1, but a much weaker smear band of HBx was observed in the presence of Myc-Siah-1-C44S, suggesting ubiquitin ligase activity is essential for HBx poly-ubiquitylation. Tumor suppressor p53 was reported to promote HBx ubiquitylation and destabilized HBx protein [12]. The p53-inducible E3 ubiquitin ligase MDM2 has been found to promote HBx proteasomal degradation, but it is unable to induce HBx ubiquitylation [14]. Many studies suggested that Siah-1 is also a p53 transcriptional target. Therefore, we speculated that p53-mediated HBx protein ubiquitylation and degradation might depend on Siah-1 induction. Therefore, we confirmed that wild type p53 induced endogenous Siah-1 expression, whereas the transcriptionally inactive mutant p53 R249S had impaired inducing effect on Siah-1 expression (Fig. 4 A). Moreover, depletion of endogenous p53 protein level by siRNA in HepG2 (p53 wild type) cells increased the HBx level as previously reported [12], and simultaneously reduced Siah-1 expression (Fig. 4B). In Hep3B cells (p53 defective), HBx decreased significantly when transfected with p53 (Fig. 4C, lane 2). However, HBx protein level was increased in response to Siah-1 knockdown even when p53 was exogenously expressed (Fig. 4C, lane 3). Taken together, these results demonstrated that the p53-induced Siah-1 expression participates in p53-mediated HBx degradation. Genomic integration of the HBx gene is highly frequent in HBV-infected HCC cases, and overexpression of HBx protein closely correlates with the progression of HCC [3]. HBx protein can be actively ubiquitinated and subjected to degradation through ubiquitin-proteasome pathway. In this study, we first identified Siah-1 as a novel E3 ubiquitin ligase for HBx. Unlike those proteins which only induced proteasome-dependent degradation of HBx protein in an indirect manner [9-13], Siah-1 could directly facilitate the poly-ubiquitylation of HBx. Additionally, Siah-1 could stimulate the degradation of KSHV-encoded ORF45 protein which is implicated in virus infection and replication [27, 31]. Similar to ORF45, HBx was involved in establishing HBV replication and hepatocarcinogenesis [4, 5]. Therefore, we inferred that Siah-1 may play important roles in fighting virus infection and may be involved in the HCC pathogenesis. Meanwhile, we found that HBx could transactivate the heat shock transcription factors (HSFs) mediated signaling pathway, which is crucial for enhancing the expression of heat shock proteins. The elevated expressions of Hsp90, Hsp70 and Hsp27 can promote cell growth, proliferation, anti-apoptosis, and are associated with high-grade malignant tumors [32]. It is noteworthy that Siah-1 could effectively abrogate HBx-induced elevation of HSE and other signaling pathway. Additionally, Siah-1 expression has been found to be down-regulated in HCC samples [29]. Hence, we hypothesize that the reduction of Siah-1 may facilitate HBx protein to exert its pleiotropic effects. This also provides new evidence that down-regulation of Siah-1 correlates with larger, poorly differentiated and/or advanced staged HCC [29]. The inhibition of tumor suppressor p53 by HBx has been extensively studied. HBx can bind to p53 directly, sequestering p53 protein in the cytoplasm and preventing its entry into nucleus. Thereby, HBx can suppress p53 sequence-specific DNA binding, transcriptional activity, p53-mediated apoptosis and disrupt the association between p53 and other proteins, such as the nucleotide excision repair factors XPB/ERCC3 and XPD [33]. We hypothesized that Siah-1 may be important in fighting with HBV infection by degrading HBx at the early stage of virus infection. Recent studies reported that p53 could increase HBx ubiquitylation and stimulate HBx protein degradation [12]. Whereas HBV in host would induce dysfunctional p53 mutation, and the interaction between p53 and HBx would abrogate p53 function. The inhibition of p53 function may result in the low expression of Siah-1. Siah-1 downregulation may facilitate HBV replication at the latter stage of virus infection. Our previous study also found that HBx protein was rarely detectable in HCC samples with wild type p53 status and p53-induced E3 ubiquitin ligase MDM2 could decrease HBx protein level [14]. Nevertheless, we inferred that MDM2 may not be the only effector responsible for HBx stability. One reason is that MDM2 has no influence on HBx ubiquitylation [14]; more importantly, MDM2 protein level is up-regulated in HCC and over-expression MDM2 correlates with poor prognosis [34]. Our present study demonstrated that wild type p53 but not mutant p53 could induce Siah-1 expression and that E3 ubiquitin ligase Siah-1 participated in p53-mediated HBx ubiquitylation and degradation. Thus, our findings suggest that highly frequent dysfunctional p53 mutation in HCC may be one reason for Siah-1 down-regulation, which may be one possible mechanism for p53 mutation causing up-regulation of HBx protein level in HCC. p53-mediated HBx degradation through the Ub-dependent pathway by Siah-1 or Ub-independent pathway by MDM2 may be linked to different physiological environments and different subcellular localization of HBx. Here, we found Siah-1 co-localized with HBx in the cytoplasm. HBx was also reported to localize in mitochondria [35] and nucleus [36]. Siah proteins contain an important binding groove that recognizes a peptide motif PXAXVXP, within many substrates and adaptors [37]. Although the full-length HBx lacks similar motif, our findings demonstrated that Siah-1 and HBx interacted both in vitro and in vivo. Moreover, truncated HBx containing the only middle transactivating domain (51–136 aa) still maintained the binding capacity to Siah-1. It is also known that damaged-DNA binding complex (DDB1) could bind to HBx within the middle domain (60–101 aa), and protect HBx from proteasome-dependent degradation [38]. Thus, the possible competitive association with HBx in the similar region between Siah-1 and DDB1 may be another regulatory manner for HBx stability. In conclusion, our results demonstrate that Siah-1 is a novel negative regulator for HBx stability and facilitates the ubiquitylation of HBx. Furthermore, Siah-1 can inhibit the transcriptional ability of HBx and participates in p53-mediated HBx degradation. Therefore, our findings shed new light on why down-regulation of Siah-1 and p53 mutation may lead to development and progression of HCC. This work was supported by the National Key Sci-Tech Special Project of and the