|Home | About | Journals | Submit | Contact Us | Français|
HBx, a small regulatory protein of hepatitis B virus (HBV), augments viral DNA replication by stimulating viral transcription. Among numerous reported HBx-binding proteins, DDB1 has drawn attention, because DDB1 acts as a substrate receptor of the Cul4-DDB1 ubiquitin E3 ligase. Previous work reported that the DDB1-HBx interaction is indispensable for HBx-stimulated viral DNA replication, suggesting that the Cul4-DDB1 ubiquitin E3 ligase might target cellular restriction factors for ubiquitination and proteasomal degradation. To gain further insight into the DDB1-HBx interaction, we generated HBx mutants deficient for DDB1 binding (i.e., R96A, L98A, and G99A) and examined whether they support HBx-stimulated viral DNA replication. In contrast to data from previous reports, our results showed that the HBx mutants deficient for DDB1 binding supported viral DNA replication to nearly wild-type levels, revealing that the DDB1-HBx interaction is largely dispensable for HBx-stimulated viral DNA replication. Instead, we found that DDB1 directly stimulates viral transcription regardless of HBx expression. Through an HBV infection study, importantly, we demonstrated that DDB1 stimulates viral transcription from covalently closed circular DNA, a physiological template for viral transcription. Overall, we concluded that DDB1 stimulates viral transcription via a mechanism that does not involve an interaction with HBx.
IMPORTANCE DDB1 constitutes a cullin-based ubiquitin E3 ligase, where DDB1 serves as an adaptor linking the cullin scaffold to the substrate receptor. Previous findings that the DDB1-binding ability of HBx is essential for HBx-stimulated viral DNA replication led to the hypothesis that HBx could downregulate host restriction factors that limit HBV replication through the cullin ubiquitin E3 ligase that requires the DDB1-HBx interaction. Consistent with this hypothesis, recent work identified Smc5/6 as a host restriction factor that is regulated by the viral cullin ubiquitin E3 ligase. In contrast, here we found that the DDB1-HBx interaction is largely dispensable for HBx-stimulated viral DNA replication. Instead, our results clearly showed that DDB1, regardless of HBx expression, enhances viral transcription. Overall, besides its role in the viral cullin ubiquitin E3 ligase, DDB1 itself stimulates viral transcription via HBx-independent mechanisms.
Hepatitis B virus (HBV) infection represents a major global public health concern, with over 300 million chronically infected patients worldwide. Chronic HBV infection carries a great risk of developing into severe liver diseases, including cirrhosis and liver cancer, which result in a million deaths annually (1). HBV carries a small (ca. 3.2-kb), relaxed circular (RC), partially double-stranded DNA. Following entry into the host cell, the viral RC DNA traffics to the nucleus, where it is converted into a covalently closed circular DNA (cccDNA), the template for viral transcription (2). Four viral RNAs are transcribed from their respective promoters and serve as mRNAs for the synthesis of viral proteins: the 3.5-kb RNA encodes the core (C) and P proteins, the 2.4-kb and 2.1-kb RNAs encode HBsAg (S), and the 0.7-kb RNA encodes the X protein (HBx) (2).
HBx, a small regulatory protein with a molecular mass of 17 kDa, is best characterized as a promiscuous transcriptional transactivator (3). HBx stimulates viral genome replication by ~3- to 5-fold, largely at the level of viral transcription. In addition, HBx induces multiple cytoplasmic signaling pathways, such as the Ras–Raf–mitogen-activated protein (MAP) kinase, Src kinase, NF-κB, and Wnt/β-catenin pathways (4). Moreover, HBx reportedly binds to >20 cellular proteins, including damaged DNA binding protein 1 (DDB1), HBXIP, VDAC3, and FLIP (5). Nonetheless, only a few of the many HBx-binding proteins have been independently confirmed. More importantly, the functional relevance of the HBx-interacting proteins for viral DNA replication and the viral life cycle remains largely unresolved (4).
Among the HBx-interacting cellular proteins, evidence of an interaction between HBx and DDB1 is compelling. First, the DDB1-HBx interaction has been observed by four independent laboratories (6,–9). The biological importance of the DDB1-HBx interaction was strengthened by observations in woodchuck animal models, where DDB1-binding-deficient X gene mutants failed to establish productive infection (10). Moreover, the DDB1-HBx interaction has been validated by a structural study. The α-helical motif, referred to as the H box (i.e., residues 88 to 100), of the HBx polypeptide was revealed to be an interacting motif that binds directly to one of three β-propeller domains of DDB1 (i.e., the BPC domain) (11). Importantly, via its interaction with DDB1, HBx was shown to constitute the Cul4A-DDB1 ubiquitin E3 ligase, implicating a role of HBx in ubiquitin-mediated protein degradation (11, 12). An intriguing possibility is that many, if not all, of these attributes of HBx could potentially be related to the DDB1-HBx interaction. In line with this notion, it was reported previously that the DDB1-HBx interaction imparts HBx-induced genetic instability and HBx-induced S-phase progression, potentially contributing to hepatocellular carcinoma development (8, 13, 14). Moreover, the DDB1-HBx interaction was shown to be essential for the ability of HBx to stimulate viral genome replication in an experiment carried out in HepG2 cells, a hepatoma cell line (15, 16).
To gain further insights into the DDB1-HBx interaction in the context of the HBV life cycle, we revisited the question as to the extent to which the DDB1-HBx interaction is critical for HBx-stimulated viral genome replication. In contrast to data from previous reports, the results of our transient-transfection studies performed in HepG2 cells clearly indicate that the DDB1-HBx interaction per se is largely dispensable for HBx-stimulated viral genome replication. Via an experiment carried out in HBV-infected HepG2-NTCP cells, we demonstrate that DDB1 itself augments viral transcription from the cccDNA template, independent of HBx expression.
HBV 1.3-mer replicon constructs (i.e., the 1.3-mer wild type [WT] and its X-null version), Flag-HBx, and hemagglutinin (HA)-luciferase (Luc) were previously described (17, 18). All HBx mutants were generated by using two-step overlapping PCR (19). HA-tagged DDB1 was subcloned into the pcDNA3 vector. The lentivirus plasmid vector pLKO.1-puro containing short hairpin RNA (shRNA) targeting DDB1 was purchased from Sigma-Aldrich. The Luc reporter constructs, including pHBV-Luc, pHBV-S1-Luc, and pHBV-S2-Luc, were previously described (20, 21). pHBV-X/EnhI-Luc was constructed by the insertion of fragments containing the X promoter/enhancer I region.
Lentiviruses expressing green fluorescent protein (GFP) or DDB1-targeting shRNA were obtained by cotransfection of HEK293T cells with the psPAX2, pMD2.G, and pLKO.1-shGFP or pLKO.1-shDDB1 constructs, as previously described (21). Briefly, aliquots of virus supplemented with 4 μg/ml Polybrene (Sigma) were used to transduce HepG2 and HepG2-NTCP cells. One day after transduction, cells were selected with fresh medium containing 3 μg/ml puromycin for 3 days.
HepG2 and HEK293 cells were maintained in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum (FBS). HepG2-NTCP cells were maintained in DMEM containing 10% FBS and 2 mM l-glutamine (21). All cells were maintained in a humidified incubator at 37°C with 5% CO2. Transient transfection was performed by using polyethylenimine (PEI) (linear, 25 kDa) or Lipofectamine 2000 (Invitrogen), as previously described (21).
Intracellular viral core DNA was extracted as described previously (22). For isolation of cccDNAs, protein-free viral DNAs were extracted from HBV-infected cells with the Hirt method, with minor modifications (21). Briefly, cells from one 60-mm dish were lysed in 1 ml of a solution containing 50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 10 mM EDTA, and 1% SDS. After 1 h of incubation at room temperature, the lysate was transferred into a 2-ml tube, and this step was followed by the addition of 0.25 ml of 2.5 M KCl and incubation at 4°C overnight. The lysate was then clarified by centrifugation and extracted with phenol-chloroform. DNA was precipitated with isopropanol overnight and dissolved in Tris-EDTA (TE) buffer. The extracted DNA was treated with Plasmid-Safe ATP-dependent DNase (Epicentre).
Southern blot analysis of HBV DNA replication intermediates was performed as previously described (23). Briefly, 4 days after transfection, cells were harvested, and viral DNA was isolated from cytoplasmic capsids. Viral DNAs were resolved in a 1.3% agarose gel, transferred onto a nylon membrane, and hybridized with a [32P]dCTP-labeled HBV-specific probe.
Northern blot analysis was performed as previously described (21), with modifications. Total RNA was separated on a 1.5% formaldehyde agarose gel and transferred onto a nylon membrane. After UV cross-linking and prehybridization, the membrane was hybridized with a [32P]dCTP-labeled HBV epsilon-specific probe. The viral RNA signals were normalized to values for rRNA.
Western blot analysis was performed as previously described (24). Endogenous DDB1 was revealed with a polyclonal anti-DDB1 antibody (Santa Cruz), and HBx was revealed with a polyclonal anti-HBx antibody (BioVendor).
Immunoprecipitation analysis was performed according to previously established methods (24). Briefly, cells were lysed with lysis buffer (50 mM Tris-Cl [pH 7.5], 50 mM NaCl, 5 mM EDTA, 1% NP-40). The lysate was then incubated with anti-Flag antibody (1:1,000; Sigma) overnight at 4°C. Protein G-Sepharose beads (GE Healthcare) were then added, and the mixture was incubated for 6 h at 4°C. The immune complex was released from the beads by boiling in 1.5× SDS sample buffer.
Briefly, cells were grown in DMEM–F-12 medium supplemented with 5% fetal calf serum (FCS), 500 μg/ml G418, 50 μM hydrocortisone-hemisuccinate, 5 μg/ml insulin, 50 U/ml penicillin, and 50 μg/ml streptomycin (21). For HBV infection, HepG2-NTCP cells were seeded into collagen-coated 12-well plates and infected with WT or X-null virus at 103 GEq (genome equivalents) per cell in medium containing 4% polyethylene glycol 8000 (PEG 8000), as described previously (21).
For RNA analysis, total RNAs were extracted from HepG2-NTCP cells at 7 dpi (days postinfection), and 1 μg of RNA was reverse transcribed and amplified by using reverse transcriptase (New England BioLabs). cDNA was quantified by real-time quantitative reverse transcription-PCR (qRT-PCR) using X-region primers (27). For DNA analysis, levels of viral DNAs in cytoplasmic capsids from HepG2-NTCP cells at 9 dpi were determined by real-time quantitative PCR (qPCR) using core region primers (27).
Chromatin immunoprecipitation (ChIP) experiments were performed on infected HepG2-NTCP cells at 3 dpi. Cells fixed with 1% formaldehyde for 10 min were harvested and incubated in ChIP lysis buffer (10 mM Tris-Cl [pH 7.6], 150 mM NaCl, 1 mM EDTA [pH 8.0], 0.25% Triton X-100, and an EDTA-free protease inhibitor cocktail [Roche]) for 5 min at 4°C. After centrifugation, pellets were resuspended in 1× TE buffer containing an EDTA-free protease inhibitor cocktail (Roche) and sonicated in a UP200S sonication device with 5 cycles at an amplitude of 50%, a 0.5-s pulse during 30 s, and 30 s on ice. Lysates were centrifuged and diluted with 2× radioimmunoprecipitation assay (RIPA) buffer (10 mM Tris-Cl [pH 7.6], 1 mM EDTA [pH 8.0], 0.1% Na-deoxycholate, 1% Triton X-100, and an EDTA-free protease inhibitor cocktail [Roche]) (28). Chromatin was subjected to immunoprecipitation overnight at 4°C with 3 μg of nonspecific immunoglobulins (Sigma) as a negative control, an anti-H4 antibody (Active Motif) as a positive control, or an anti-DDB1 antibody (Santa Cruz). Immunoprecipitated chromatins were processed and analyzed by real-time qPCR (29).
All statistical analyses were performed by using GraphPad Prism version 6.01 (GraphPad Software). A P value of <0.05 was considered statistically significant.
Previous studies reported that the DDB1-HBx interaction is indispensable for HBx-stimulated viral DNA replication (15, 16). To corroborate that finding, we sought to examine whether the ability of HBx to bind DDB1 is essential for HBx-stimulated viral genome replication. Structural analysis revealed that three residues (i.e., the R96, L98, and G99 residues) of the HBx polypeptide are involved in making contact with DDB1 (11). Therefore, we created alanine substitution mutants of these three residues. First, we examined whether the mutants failed to bind DDB1. As anticipated, coimmunoprecipitation with anti-HA antibody showed that each of the substitution mutants failed to bind DDB1, whereas WT HBx could bind DDB1 (Fig. 1A). Reciprocal coimmunoprecipitation (co-IP) with an anti-Flag antibody yielded identical results (data not shown). Thus, as predicted by structural analysis (11), the DDB1-HBx interaction occurring via the H-box region of the HBx polypeptide was experimentally demonstrated.
Next, we examined whether DDB1-binding-defective HBx mutants could support viral DNA replication. HepG2 cells that were transfected with the 1.3-mer WT replicon or a 1.3-mer X-null replicon were complemented by either WT or mutant HBx, as described previously (20). Southern blot analysis of viral DNA replication intermediates isolated from cytoplasmic nucleocapsids was carried out to measure viral DNA replication. First, as expected, we confirmed that HBx stimulated viral DNA replication by ~2- to 3-fold (Fig. 1B, lane 1 versus lane 2). Unexpectedly, the data revealed that viral DNA replication occurred in mutant HBx-transfected cells at a level that was only slightly lower than that in WT HBx-transfected cells (Fig. 1B, lane 3 versus lanes 4 to 6) but significantly higher than that in 1.3-mer X-null replicon-transfected cells (Fig. 1B, lane 2 versus lanes 4 to 6). Western blot analysis performed in parallel showed that the HBx level in the mutants was slightly reduced (Fig. 1B, lane 3 versus lanes 4 to 6), revealing that the slight reduction in the level of viral DNA synthesis was an underestimation. Taken together, the results showed that HBx lacking DDB1-binding ability nonetheless stimulates viral DNA replication albeit to a lesser extent than WT DDB1, a finding that revealed that the DDB1-HBx interaction is largely dispensable for HBx-stimulated HBV DNA replication.
The dispensability of the DDB1-binding ability of HBx for the stimulation of viral genome replication led us to examine the extent to which HBx could stimulate viral DNA replication even in the absence of DDB1. To this end, we sought to deplete DDB1 via shRNA treatment and then examine the effect of HBx on viral DNA replication. As described above, we complemented cells transfected with the 1.3-mer X-null replicon with an increasing amount of the HBx expression plasmid following treatment with an shRNA targeting DDB1 (shDDB1). In parallel, cells were also treated with an shRNA targeting GFP (shGFP) as a control. Southern blot analysis showed that viral DNA replication was increased by HBx in a dose-related manner in shGFP-treated cells (Fig. 2A, lanes 1 to 4), as anticipated. However, viral DNA replication was strikingly reduced by shDDB1 treatment, revealing that DDB1 is critically important for HBx-stimulated viral DNA replication (Fig. 2A, lanes 2 to 4 versus lanes 6 to 8). In addition, the HBx level was considerably reduced by shDDB1 treatment, which is apparently consistent with results of a previous report on DDB1-binding-mediated HBx stabilization (30). However, the observation that the viral DNA level in shDDB1-treated cells was much lower than that in shGFP-treated cells, while the HBx levels were comparable (Fig. 2A, lane 3 versus lane 8), suggested that DDB1 contributes to viral DNA replication via a mechanism that is distinct from HBx stabilization. Moreover, the finding that, even in the absence of HBx, DDB1 supported viral DNA replication led us to conclude that DDB1 stimulates viral DNA replication independently of HBx expression (Fig. 2A, lane 1 versus lane 5). Furthermore, this result strengthened our conclusion about the dispensability of the DDB1-HBx interaction for the stimulation of viral DNA replication. The alternative possibility that the reduction of viral replication (Fig. 2A, lanes 5 to 8) is attributable to cell death induced by DDB1 depletion, as reported previously (31), was excluded by the observation of no significant cell death under our experimental conditions (data not shown).
Although the above-described experiment clearly demonstrated that endogenous DDB1 is critical for viral DNA replication, the extent to which the DDB1-HBx interaction contributes to HBx-stimulated viral DNA replication remained uncertain. To address this question, we employed a DDB1-binding-defective HBx mutant and examined whether the DDB1-binding-defective HBx mutant can fully support viral DNA replication. Cells were transfected as described above but complemented with a DDB1-binding-defective HBx mutant (i.e., R96A). Southern blot analysis showed that the DDB1-binding-defective HBx mutant nonetheless supported HBx-stimulated viral DNA replication to near-WT levels (Fig. 2B, lanes 1 to 4), which is consistent with the data shown in Fig. 1B. In DDB1-depleted cells, however, viral DNA replication was greatly reduced (Fig. 2B, lanes 5 to 8). Notably, in addition to the viral DNA level, the HBx protein level was considerably reduced in the absence of DDB1 (Fig. 2B, lanes 5 to 8). Importantly, the fact that the reduced HBx level was not ascribable to the lack of DDB1 binding in this experiment suggested that the mechanism underlying the reduced HBx level is distinct from DDB1-mediated HBx stabilization, as previously reported (13, 32). Furthermore, HBx stability measured following cycloheximide treatment, as previously described (26), revealed that HBx stability remained unaltered by the depletion of DDB1, corroborating the above-mentioned interpretation (Fig. 2C). Overall, the critical importance of DDB1 for viral DNA synthesis led us to investigate the mechanism underlying the impact of DDB1 on viral genome replication.
The results described above clearly demonstrated that both HBx and DDB1 are needed to gain maximal viral DNA replication (Fig. 2). To measure the impacts of DDB1 and HBx separately, we examined the impact of DDB1 on viral DNA replication with respect to HBx expression. In other words, we examined the impact of ectopic DDB1 expression on cells transfected with the WT or X-null replicon (Fig. 3A). Cells were transfected as indicated, and the data showed that ectopically expressed DDB1 augmented HBV DNA replication in WT replicon-transfected cells (Fig. 3A, lane 1 versus lane 2). Likewise, DDB1 augmented HBV DNA replication in X-null replicon-transfected cells to a similar extent (Fig. 3A, lane 3 versus lane 4), pointing out clearly that DDB1 augments viral DNA replication regardless of HBx expression. The observation that ectopic DDB1 transfection augments viral DNA replication in not only WT replicon-transfected cells but also X-null replicon-transfected cells strengthens our conclusion that the DDB1-HBx interaction is largely dispensable for HBx-stimulated viral genome replication (Fig. 1B).
A question, then, is whether DDB1 augments viral DNA replication at the transcriptional level or at the posttranscriptional level. To address this issue, we examined the impact of DDB1 on viral RNAs in WT replicon-transfected cells and in X-null replicon-transfected cells (Fig. 3B). Northern blot analysis showed that DDB1 elevated the levels of viral transcripts in both WT replicon-transfected cells and X-null replicon-transfected cells, indicating a direct effect of DDB1 on viral transcription, regardless of HBx expression (Fig. 3B, lane 1 versus lane 2 and lane 3 versus lane 4). Conversely, the depletion of endogenous DDB1 via shRNA treatment resulted in a concomitant reduction of the levels of viral transcripts not only in WT replicon-transfected cells but also in X-null replicon-transfected cells (Fig. 3C, lane 1 versus lane 2 and lane 3 versus lane 4). Thus, both the ectopic expression and the depletion of DDB1 consistently demonstrated a stimulatory effect of DDB1 on the level of viral RNAs, regardless of HBx expression. Importantly, the lack of either DDB1 or HBx led to a decrease in the level of viral RNAs in parallel (Fig. 3C, lanes 2 and 3), while the lack of both DDB1 and HBx led to a considerable decrease in the level of viral RNAs (Fig. 3C, lane 4). The additive effect of DDB1 depletion in the absence of HBx suggests that DDB1 acts to elevate the level of viral RNAs, regardless of HBx expression.
To assess the possibility that DDB1 augments the level of viral RNAs at the posttranscriptional level by enhancing the stability of viral RNAs, we measured viral RNA stability. Viral RNAs were examined following actinomycin D treatment in either shGFP-treated cells or shDDB1-treated cells (Fig. 4A). Quantitation revealed that the stability of the viral RNAs remained unaltered in shDDB1-treated cells (Fig. 4B). Thus, the data pointed out that DDB1 augments viral RNAs most likely at the transcriptional level.
The results described above suggest that DDB1 augments viral RNAs at the transcriptional level. HBV possesses four viral transcripts, of 3.5 kb, 2.4 kb, 2.1 kb, and 0.7 kb, each of which is transcribed from its own promoter (2). The Northern blot analysis shown in Fig. 3B clearly showed an elevation of the levels of three transcripts by DDB1: the 3.5-kb RNA band and the S transcripts (i.e., 2.4-kb and 2.1-kb RNAs). However, due to the limit of the resolution of agarose gel electrophoresis, whether the level of the X transcript (i.e., the 0.7-kb RNA) was elevated remained uncertain. Therefore, we carried out a luciferase reporter assay that measures viral promoter activity, as described previously (21). Four luciferase reporter constructs were employed (Fig. 5A). The luciferase reporter assay showed that transcription from all four viral promoters was concomitantly reduced by ~3- to 4-fold following shDDB1 treatment (Fig. 5B), but transcription from the herpes simplex virus thymidine kinase promoter remained unaltered by shDDB1 treatment (data not shown), confirming the specific effect of DDB1 on HBV promoters. Intriguingly, the magnitude of the reduction shown by the reporter assay depicted in Fig. 5B parallels that shown by the Northern blot analysis depicted in Fig. 3C. Therefore, we concluded that DDB1 augments viral transcription from all four promoters.
The data described above consistently indicated that DDB1 augments viral RNAs at the transcriptional level. One caveat is that all of the above-described experiments were carried out by transfection of replicon plasmid DNA, which then served as a template for viral transcription in transfected cells. As a matter of fact, in HBV-infected cells, cccDNA serves as the template for viral transcription (33). Hence, the experimental findings from the transfection studies needed to be validated by an HBV infection study (34). To substantiate the above-described findings in the context of HBV infection, we employed the HepG2-NTCP cell line, which is susceptible to HBV infection (27, 34). Following infection of HepG2-NTCP cells that had been transduced by a lentivirus encoding shGFP or shDDB1, we detected HBV cccDNA by Southern blotting of viral DNA extracted according to the Hirt extraction method. The results showed that cccDNA could be measurably detected in WT HBV-infected cells, regardless of DDB1 depletion (Fig. 6A, lanes 1 and 2). Likewise, a comparable level of cccDNA was detected in X-null HBV-infected cells, regardless of DDB1 depletion (Fig. 6A, lanes 3 and 4), indicating that both HBx and DDB1 are dispensable for the establishment of HBV infection. The observed dispensability of HBx for the establishment of HBV infection is consistent with results from a previous report (35, 36). In contrast, the amount of viral RNAs measured by real-time qRT-PCR was substantially decreased by the depletion of DDB1 in WT HBV-infected cells (Fig. 6B); an essentially similar result was obtained by Northern blotting (data not shown). Consistently, a parallel reduction of viral RNAs was observed following the depletion of DDB1, even in X-null HBV-infected cells, pointing out that DDB1 stimulates viral transcription independently of HBx expression (Fig. 6B). Importantly, the additive effects of DDB1 and HBx on viral transcription shown by the transfection study (Fig. 3C) were recapitulated in HBV-infected cells (Fig. 6B). Therefore, we concluded that DDB1 augments viral transcription from cccDNA, a genuine template for viral transcription. Consistently, the amount of nucleocapsid-associated viral DNA was accordingly decreased following shDDB1 treatment, as measured by real-time qRT-PCR (Fig. 6C) and by Southern blotting (data not shown).
A question remained as to how DDB1 augments viral transcription. One possibility is that DDB1 augments viral transcription via its association with the chromatin structure of cccDNA. To examine that possibility, we carried out a ChIP analysis of the cccDNA isolated from HBV-infected cells, as detailed in Materials and Methods. The data showed that DDB1 is recruited to the cccDNA to a level that is substantially higher than that of IgG (negative control) but lower than that of anti-histone H4 (positive control) (Fig. 6D, left). Not surprisingly, the recruitment of DDB1 to cccDNA was dramatically reduced in shDDB1-treated cells (Fig. 6D, right). Moreover, the finding that there was no significant recruitment of DDB1 to the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene revealed the specific association of DDB1 with cccDNA chromatin (Fig. 6D, right). Overall, we demonstrated that DDB1 stimulates viral DNA replication at the level of viral transcription via its association with cccDNA chromatin in HBV-infected cells (Fig. 7).
Among the numerous HBx-binding cellular proteins, DDB1 has drawn significant attention, perhaps because HBx was hypothesized to serve as a DDB1-Cul4-associated factor (DCAF) in HBV-infected cells, potentially regulating numerous cellular proteins (11). Previous studies reported that the DDB1-HBx interaction is indispensable for HBx-stimulated viral DNA replication (15, 16), suggesting that a host restriction factor that limits HBV replication is regulated by the Cul4-DDB1 ubiquitin E3 ligase (Fig. 7). In contrast to that hypothesis, we found that DDB1-binding-defective HBx mutants nonetheless significantly stimulate viral genome replication albeit to a lesser extent than WT HBx (Fig. 1), suggesting that the DDB1-HBx interaction is largely dispensable for HBx-stimulated viral DNA replication. A subsequent DDB1 depletion study revealed the critical importance of DDB1 to viral genome replication in the absence of an interaction with HBx (Fig. 2B). Mechanistically, we found that DDB1 itself augments viral transcription, regardless of HBx expression (Fig. 3). The experimental observations from transfection studies were further validated by the demonstration of DDB1-augmented viral transcription from cccDNA in HBV-infected cells (Fig. 6).
An intriguing speculation coming from previous reports that the HBx-DDB1 interaction is critical for viral genome replication (15, 16) was that HBx-targeted host factors that restrict viral DNA replication are regulated by the Cul4-DDB1 ubiquitin E3 ligase (12). Contrary to the expectation of that hypothesis, our findings consistently indicated that the DDB1-HBx interaction is largely dispensable for HBx-stimulated viral genome replication. First, our data showed that three DDB1-binding-defective HBx mutants support HBV DNA replication (Fig. 1) and viral transcription (data not shown). Furthermore, an HBx-binding-defective DDB1 mutant (i.e., the A381E F382D double-substitution mutant ) supports HBV replication, further supporting the dispensability of the DDB1-HBx interaction (data not shown). Second, even in the absence of HBx expression, DDB1 stimulated viral DNA replication by augmenting viral transcription (Fig. 3A and andB).B). Conversely, the depletion of DDB1 led to a decrease in viral transcription (Fig. 3C).
Previous studies reported that HBx protein stability is greatly enhanced by the ectopic expression of DDB1 (30, 32), suggesting that HBx stability is significantly enhanced via interactions with DDB1. Similarly, we observed an elevated HBx level in DDB1-transfected cells (Fig. 3A and andB)B) and a parallel reduction of HBx in DDB1-depleted cells (Fig. 2A and and3C).3C). However, our findings suggested that the elevation of the HBx level by DDB1 is not regulated at the level of protein stability.
First, the observation that HBx stability measured after cycloheximide treatment remained unaltered in DDB1-depleted cells disputes the hypothesis of DDB1-mediated HBx stabilization (Fig. 2C). Second, the observation that the HBx level in the DDB1-binding-defective HBx mutant (i.e., R96A) was greatly reduced by DDB1 depletion excludes the possibility that DDB1 enhances HBx stability via DDB1-HBx interactions (Fig. 2B). Rather, our findings led us to conclude that the HBx level is augmented by DDB1 at the level of transcription (Fig. 3, ,5,5, and and66).
DDB1 is a multifunctional protein that was first isolated as a subunit of a heterodimeric complex (i.e., the DDB1/DDB2 complex) that recognizes UV-induced DNA lesions in the nucleotide excision-repair pathway (37). What is the mechanism underlying transcriptional stimulation on the viral cccDNA by DDB1? In fact, cccDNA is assembled with histones into a chromatin structure, and HBx was shown to bind cccDNA chromatin and induce epigenetic modifications (38). In addition to HBx, our data revealed that DDB1 is recruited to cccDNA in HBV-infected cells (Fig. 6D). The data suggest that DDB1 itself, regardless of its interaction with HBx, is recruited to the cccDNA, thereby stimulating viral transcription (Fig. 7). At the moment, it remains unclear how DDB1 carries out transcriptional transactivation in the absence of HBx. For instance, it is possible that DDB1, in association with DDB2, could induce transcriptional activation on cccDNA via the recruitment of the p300/CBP histone acetyltransferase, as described previously (39). Alternatively, DDB1, in association with DDB2, could function as a transcription factor. DDB1 could function as a binding partner of E2F1, a member of the E2F transcription factors which plays a crucial role in cell cycle control (40). In this scenario, DDB1 could facilitate transcription from the viral cccDNA through an interaction with one of the numerous cellular transcription factors (e.g., Sp1) bound to the viral promoters (3), perhaps via the formation of an Sp1-E2F1-DDB1 interaction (41).
Recently, the “structural maintenance of chromosomes” (Smc) complex Smc5/6 was identified as a host restriction factor that is targeted by the HBx-associated cullin RING ubiquitin E3 ligase (42). An interpretation of this finding is that the impact of HBx on viral transcription is attributable mainly to the degradation of Smc5/6 by the HBx-associated cullin RING ubiquitin E3 ligase (Fig. 7). According to this scenario, the DDB1-HBx interaction is critical for HBx-mediated stimulation of viral transcription, at least in the primary human hepatocytes (PHHs) and HepaRG cells that were used in a previous HBV infection study (42). In contrast to results of that report, through experiments carried out with HepG2 cells, we showed here that the DDB1-HBx interaction is largely dispensable for the HBx-mediated stimulation of viral transcription. This discrepancy may be accounted for by differences in the extent of host restriction between cells employed for experiments. For instance, the restriction imposed by Smc5/6 could be minimal in HepG2 cells so that the DDB1-HBx interaction necessary for the degradation of Smc5/6 is largely dispensable. In an agreement with above-mentioned notion, our preliminary results showed that the Smc5/6 level remained unaltered in both HBV replicon-transfected HepG2 cells and HBV-infected HepG2 cells (data not shown). The reason why Smc5/6 is not targeted by the HBx-associated cullin RING ubiquitin E3 ligase in HBV-infected HepG2 cells merits further investigation.
Regarding the role of DDB1 in viral transcription, we speculated that DDB1 contributes to viral transcription via two distinct mechanisms with respect to the DDB1-HBx interaction: one that does not require the DDB1-HBx interaction and another that requires the DDB1-HBx interaction. The extent to which the two distinct mechanisms contribute to viral transcription depends on the strength of host restriction imposed in a given cell line. We speculate that the former mechanism manifests mainly in HepG2 cells, as we observed here, while the latter mechanism overwhelmingly manifests in PHHs and HepaRG cells (42). Overall, given the importance of DDB1 for viral transcription in HBV-infected cells (Fig. 6), the mechanisms by which DDB1 enhances transcription from a cccDNA template merit further investigation.
This work was supported by National Research Foundation of Korea (NRF) grant NRF-2015R1A2A2A01005938 funded by the South Korean government.
W.K., S.L., Y.S., and C.K. have been awarded fellowships from the BK PLUS program.