Deubiquitylation of hepatitis B virus X protein (HBx) by ubiquitin-specific peptidase 15 (USP15) increases HBx stability and its transactivation activity

Hepatitis B virus X protein (HBx) plays important roles in viral replication and the development of hepatocellular carcinoma. HBx is a rapid turnover protein and ubiquitin-proteasome pathway has been suggested to influence HBx stability as treatment with proteasome inhibitors increases the levels of HBx protein and causes accumulation of the polyubiquitinated forms of HBx. Deubiquitinases (DUBs) are known to act by removing ubiquitin moieties from proteins and thereby reverse their stability and/or activity. However, no information is available regarding the involvement of DUBs in regulation of ubiquitylation-dependent proteasomal degradation of HBx protein. This study identified the deubiquitylating enzyme USP15 as a critical regulator of HBx protein level. USP15 was found to directly interact with HBx via binding to the HBx region between amino acid residues 51 and 80. USP15 increased HBx protein levels in a dose-dependent manner and siRNA-mediated knockdown of endogenous USP15 reduced HBx protein levels. Increased HBx stability and steady-state level by USP15 were attributable to reduced HBx ubiquitination and proteasomal degradation. Importantly, the transcriptional transactivation function of HBx is enhanced by overexpression of USP15. These results suggest that USP15 plays an essential role in stabilizing HBx and subsequently affects the biological function of HBx.

have been shown to contribute to destabilizing HBx protein in a proteasome-dependent manner. For instances, Id-1 destabilized HBx protein by facilitating the interaction between ubiquitinated HBx and proteasome subunit C8 while Id-1 itself had no influence on the ubiquitylation modification of HBx 13 . Tumor suppressor p53 is capable of increasing HBx ubiquitylation with an unknown mechanism 14 . A previous study demonstrated that HBx can be actively ubiquitinated with the host cell and undergo proteolysis, suggesting that HBx degradation is ubiquitin-dependent 9 . Removal of ubiquitin from mono-and poly-ubiquitylated proteins is also critical for rescuing them from degradative pathways or leading to the reversion of ubiquitin signaling. However, no deubiquitylating enzymes or deubiquitinases (DUBs) have been identified so far that might account for regulation of ubiquitylation-dependent proteasomal degradation of HBx protein. One DUB that has recently received a lot of attention, primarily through its well-documented association with various cancer-signaling pathways, is the ubiquitin-specific peptidase 15 (USP15). For example, USP15 was one of twelve DUBs identified from an siRNA screen that regulates the hepatocyte growth factor (HGF)-dependent cell scattering response in non-small cell lung cancer and pancreatic cancer cells 16 . A number of specific USP15 substrates have also been described, including the human papilloma virus (HPV) E6 oncoprotein 17 , the RING-box protein Rbx1 18 , the adenomatous polyposis coli (APC) tumour suppressor 19 , and the NF-κ B inhibitor Iκ Bα 20 .
Many of the HBx binding partners have been identified using immunoprecipitation or using classical yeast two-hybrid (Y2H) screenings. While some of these interactions with HBx have been validated, the physiologic relevance of such interactions remains largely uncertain. In an effort to identify HBx interacting proteins, we previously employed the CytoTrap yeast two hybrid (Y2H) system to screen for cellular proteins that may interact with HBx and have identified several new candidate proteins including USP15 (unpublished data). This study further explored and confirmed HBx as a novel USP15-binding partner and substrate. We found that USP15-mediated deubiquitylation protects HBx from proteasomal degradation thus increasing the stability and level of HBx protein as well as its transactivation activity. These results suggest that interventions directed at suppressing the level or functional activity of USP15 may be of therapeutic value in HBx-related hepatocarcinogenesis.

Results
HBx associates with USP15 in vitro and in vivo. To confirm the interaction of HBx with USP15 and map the domains of HBx involved in this interaction, HBx and a series of deletion mutants (Fig. 1A) were examined by CytoTrap two-hybrid assay for their ability to interact with USP15. Yeast transformant colonies harboring pMyr-USP15 with deletion mutant of HBxΔ 51-80 or HBxΔ 81-120 could not grow in galactose at 37 °C (Fig. 1B, rows 8 and 9), which indicates that the region between HBx amino acid residues 51 and 120 is required for the interaction with USP15. To further confirm the interaction between HBx and USP15 in vitro, the GST pull-down assay was performed. As shown in Fig. 1C and D, GST and GST-HBx protein were well expressed, and 35 S-labeled USP15 was retained on the GST-HBx-conjugated sepharose whereas GST alone was not, indicating that USP15 could interact with HBx directly in vitro. To address a potential interaction between the two proteins in hepatocytes, we conducted an in vivo co-immunoprecipitation (Co-IP) study with the Huh7 cells transiently transfected with HBx expression vector pHBx-flag or empty vector. Figures 1E and F show that HBx was able to efficiently co-precipitate with endogenous USP15, and vice versa. Taken together, these results strongly suggested that HBx and USP15 interact specifically both in vitro and in vivo.
HBx does not affect USP15 degradation or peptidase activities. To assess the functional consequence of the interaction between USP15 and HBx, we first examined the effects of HBx on USP15 protein level and peptidase activities. As shown in Fig. 2A-C, the steady-state level of USP15 was not affected by HBx expression in Huh7, HepG2 and Hep3B cells. We also performed an in vitro USP15 peptidase activities by DUB-Glo Protease assay using a human recombinant USP15 with the addition of increasing amount of purified HBx. The results demonstrated that peptidase activities of USP15 was not affected by HBx (Fig. 2D). We therefore conclude that HBx does not impact USP15 protein level and its peptidase activity. USP15 affects HBx protein levels. We then tested whether USP15 could affect the HBx protein level.
Huh7 cells were transfected with HBx alone or together with USP15. As shown in Fig. 3A, USP15 overexpression significantly increased HBx protein levels in a dose-dependent manner. A similar result was obtained in HepG2 and Hep3B cells ( Fig. 3B and C). In contrast, knocking down endogenous USP15 by siRNA resulted in a significant reduction of HBx protein levels in Huh7, HepG2 and Hep3B cells (Fig. 3D). To examine whether over-expression or knockdown of USP15 affects the HBx protein levels in a HCC cell line with "endogenous" HBx we ectopically expressed a carboxyl-terminal truncated form of HBx (Ct-HBx, depletion of amino acids 130-154) in Huh7 cells (the stable cell line is designated here as Huh7/Ct-HBx). As shown in Fig. 3E, overexpression or knockdown of USP15 in the Huh7/Ct-HBx cells correspondingly increased or reduced Ct-HBx protein expression. To assess the specificity of USP15 in increasing HBx levels, we performed similar cotransfection HBx had no effect on USP15 degradation. Endogenous USP15 protein level was analyzed in Huh7(A), HepG2 (B) and Hep3B (C) cell lines transfected with empty vector pcDNA3.1/myc-His(− ) A or increasing amount of pHBx-FLAG, β -actin is included as a loading control. (D) HBx had no effect on USP15 peptidase activities. In vitro US15 peptidase activities was assessed by DUB-Glo Protease assay using a human recombinant USP15 with the addition of increasing amount of purified HBx. 1 nM of DUB protease UCH-L3 was used as a positive control. The luminescence was measured and was recorded as relative light units (RLU). Values were mean ± S.D of three separate experiments. *P < 0.05 vs blank control. experiments with another DUB, USP5 (also known as isopeptidase) 21 . We observed only a moderate increase of 1.8-fold in HBx protein levels in Huh7 cells expressing USP5 (Fig. 3F), as compared with 8.5-fold increase in USP15-expressing Huh7 cells. However, HBx and USP5 did not co-immunoprecipitate with each other in the cell extracts prepared from the co-transfected Huh7 cells (Fig. 3G,H), suggesting that the moderate increase in HBx levels by USP5 might be owing to the general ability of USP5 to disassemble polyubiquitin chains 22 rather than the binding of USP5 to HBx.

USP15 stabilizes HBx.
To determine whether the increased levels of HBx by co-expression of USP15 resulted from an extended half-life, we performed a cycloheximide chase experiment. Huh7 cells were transfected with HBx alone or in combination with USP15, then continually exposed to cycloheximide for different time periods up to 120 mins. Cell extracts were analyzed by Western blot with the specific antibodies. Figure 4A and B shows that USP15 expression led to a longer degradation time for the HBx-transfected cells, increasing HBx half-life from approximate 45 mins to 113 mins. To examine the effect of endogenous USP15 on the exogenously expressed HBx protein, Huh7 cells were co-transfected with HBx and USP15-targeting siRNA. As shown in Fig. 4C and D, knockdown of endogenous USP15 resulted in a significant decrease in the half-life of HBx from around 45 mins to 18 mins. There results clearly indicate that USP15 attenuates the degradation of HBx.
USP15 attenuates HBx degradation through deubiquitination of HBx. USP15 has been well documented to catalyze the deubiquitylation of most substrates 17,20 . Ubiquitination of HBx has been previously reported and both ubiquitin-dependent and -independent proteasomal degradation processes appears to be operative in its turnover 23 . To test if the effect of USP15 on HBx expression is proteasome-dependent, Huh7 cells were USP15 and HBx were detected by anti-myc or anti-FLAG antibody, respectively. β -actin is included as a loading control. Densitometry analysis of band signals was carried out using Image J software and the levels of HBx protein in cells transfected with pHBx-FLAG + pUSP15-myc were expressed as the relative intensity (RI) to that in the empty vector control pUSP5-myc or pHBx-FLAG after normalization to β -actin. A: Huh7cells cotransfected with 0.75 μ g pHBx-FLAG and 2.25 μ g pUSP15-myc; B: HepG2 cell lines co-transfected with 0.75 μ g pHBx-FLAG and 2.25 μ g pUSP15-myc; C: Hep3B cell lines co-transfected with 0.75 μ g pHBx-FLAG and 2.25 μg pUSP15-myc. (D) Knocking down of endogenous USP15 decreased HBx protein level. Huh7, HepG2 and Hep3B cells were separately transfected with siRNA targeting USP15. HBx and USP15 were detected by anti-FLAG and anti-USP15 antibody. β -actin was included as a loading control. (E) Overexpression or knockdown of USP15 correspondingly increased or reduced HBx expression in Huh7/Ct-HBx cells that had been stably transfected with a carboxyl-terminal truncated form of HBx (Ct-HBx, depletion of amino acids 130-154). HBx and USP15 were detected by anti-FLAG and anti-USP15 antibody. β -actin was included as a loading control. (F) USP5 overexpression increased HBx protein level. Huh7 cells were co-transfected with 0.75 μ g pHBx-FLAG and 2.15 μ g pUSP5-myc (equal molar amount of 2.25 μ g pUSP15-myc). HBx was analyzed by western blot, and relative intensity was calculated as mentioned above. (G,H) Co-IP assay of USP5 and HBx. Huh7 cells were transfected with empty vector pcDNA3.1/myc-His(− ) A or pHBx -FLAG. The immunoprecipitation of HBx was detected for USP5 (F) and vice versa (G) by western blot analysis. The input served as expression and specificity control for the individual proteins.
Scientific RepoRts | 7:40246 | DOI: 10.1038/srep40246 co-transfected with pHBx-FLAG and pUSP15-myc or USP15 targeting siRNA then treated with the proteasome inhibitor MG132. As shown in Fig. 5A, increasing HBx expression by ectopic expression of USP15 was further enhanced by treatment with the proteasome inhibitor MG132 whereas the reduced HBx protein levels resulting from knockdown of USP15 was rescued by treatment with the MG132 (Fig. 5B). Since a proteasome inhibitor can block the rapid breakdown of proteins by the ubiquitin pathway without affecting protein synthesis, it should cause a short-lived protein to accumulate in the cell if the protein is degraded through the ubiquitin-proteasome pathway. The observation that addition of MG132 led to higher HBx level at USP15 overexpression or rescued the reduction of HBx from USP15 knockdown indicates that the proteasome-dependent function of USP15 plays a primary role in regulation of HBx turnover. To confirm the deubiquitination of HBx by USP15, we performed an in vivo ubiquitination assay in which HBx was immunoprecipitated and detected by antibody against ubiquitin. Indeed, overexpression of USP15 diminished K48-linked polyubiquitination of HBx whereas inhibition of endogenous USP15 expression by siRNA-mediated depletion markedly increased K48-linked HBx polyubiquitination even though the protein level of HBx itself had been substantially reduced (Fig. 5C). In addition, for the reasons that E3 ubiquitin ligase has been shown to facilitate polyubiquitylation and proteasomal degradation of HBx 15 , we considered the possibility that USP15 may interfere with E3 ubiquitin ligase in its ability to target HBx for proteasomal degradation. As shown in Fig. 5D, despite the fact that USP15 dose-dependently increases HBx levels, it did not affect the expression levels of DDB1, a core component of E3 ubiquitin ligase complexes. Instead, with the increase of HBx levels, an association of USP15 with DDB1 was also increased. Therefore, these results indicate that USP15 protects HBx from proteasome-mediated degradation through reducing HBx ubiquitination rather than competing with HBx to bind to E3 ubiquitin ligase complexes.
USP15 enhances the transactivation activity of HBx. One of fundamental functions of DUBs is the specific deconjugation of ubiquitin from targeted proteins, which may rescue them from proteasome-dependent destruction and thus resume their original designated functions. Given the observations that USP15 binds to HBx, trims ubiquitin from HBx, and consequently increases HBx stability and protein level, we went on to explore whether the transcriptional transactivation function of HBx could be increased as a result of elevated HBx protein level due to the interaction between USP15 and HBx. As shown in Fig. 6, co-expression of USP15 enhanced the AP-1, AP-2, AP-3, SP-1 and NF-κ B signal pathways transactivated by ectopically expressed HBx although these signal pathways were also affected to a lesser extent by USP15 overexpression alone (Fig. 6). This implied that the association of USP15 with HBx not only increases the stability of HBx but also augments HBx-mediated oncogenic signals.

Discussion
HBx is an unstable protein having a short half-life, whose instability is considered to be attributed to rapid degradation through the ubiquitin-proteasome pathway 9,22 . Since ubiquitination is a dynamic and reversible process,   Huh7 cells were co-transfected with pHBx-FLAG, and 0.4 μ g each of cis-element luciferase reporter plasmid including pAP-1-luc, pAP-2-luc, pAP-3-luc, pSP-1-luc, and pNF-κ B-luc. 48 h after transfection, cells were lysed and 30 μ g protein were used for the detection of intracellular luciferase activity. The light intensity was measured and the relative luciferase unit (RLU) were obtained by comparison to that from the empty vector control pCDNA3.1/myc-His(− )A. Each transfection was performed in duplicate and repeated three times. *P < 0.05 vs empty vector, **P < 0.05 vs HBx or USP15. and prolonged expression of the viral regulatory protein HBx is required for dysregulation of cell transcription and proliferation control, cells must possess a specific and also rapid deubiquitination machinery capable of such a tight control of HBx expression. However, mechanisms underlying HBx deubiquitination remains unknown. In our present study, we have identified the deubiquitylating enzyme USP15 as a key regulator of HBx protein levels. We demonstrate for the first time that USP15 specifically interacts with HBx and consequently increases HBx stability and the steady-state level by inhibition of HBx ubiquitination and degradation.
The physical and specific interaction between HBx and USP15 was confirmed from a series of comprehensive binding studies. To determine whether there is a possible association between HBx and USP15, perhaps under more physiologic condition, the CytoTrap yeast two-hybrid system was employed, and the results demonstrated that HBx could interact with USP15, and the region between HBx amino acid residues 51 and 80 is required for the interaction with USP15. Furthermore, by using the GST pull-down assay, we demonstrated that HBx was able to bind to USP15 in the absence of a cellular context. To assess the biological relevance of this interaction, HBx was transfected in Huh7 cells, and coimmunoprecipitation analyses showed that HBx interacted with endogenous USP15 and vice versa. These results clearly indicate that HBx and USP15 interact specifically both in vitro and in vivo.
USP15 was first cloned and characterized in 1999 and belongs to the largest ubiquitin specific protease (USP) group of deubiquitinating enzymes (DUBs) 23 . However, only recently some of its function and targets are being elucidated. USP15 has been reported to associate with COP9 signalosome (CSN), a multiprotein complex that regulates the ubiquitin-proteasome pathway predominantly through interaction with cullin-based E3 ligases 18 . In this scenario, USP was shown to protect Rbx1 from autoubiquitylation. CSN-associated USP15 can deubiquitinate Iκ Bα after TNFα -mediated stimulation of the NF-κ B pathway 24 . In contrary, another study found no interaction between USP15 and Iκ Bα , and it was proposed that USP11 inhibits the ubiquitination and degradation of Iκ Bα in the early stage while USP15 functions at a later time point in the TNFα -induced NF-κ B activation 25 . USP15 also acts as a key component of the transforming growth factor β (TGF-β ) signaling pathway 26 , and is a DUB for R-SMADs 27 . The DUB activity of USP15 has also been reported to be involved in the regulation of Tip110 protein degradation 28 , parkin-mediated mitochondrial ubiquitination and mitophagy 29 , and ALK3/BMPR1A in bone morphogenetic protein signaling 30 as well as the retinoic acid-inducible gene I (RIG-I)-dependent type I IFN induction pathway 31 . Based on the insights into the characteristics of USP15 gained so far, it is conceivable that USP15 is a multifunctional protein by acting as a DUB for a wide variety of molecules in different signaling pathway. In contrast to our emerging knowledge about the ubiquitylation of HBx, little or none is known about the role of deubiquitylation for regulating HBx stability and expression levels. Here, we have identified USP15 as a critical regulator of HBx. USP15 binds to HBx and stabilizes HBx through removal of Lys 48 -linked polyubiquitin moieties from HBx, thus extending its half-life and increasing its steady-state level. The observation that treatment with the proteasome inhibitor MG132 enhanced USP15-induced HBx levels or rescued the reduction of HBx due to USP15 knockdown further supports the concept that USP15 is important for maintaining cellular pools of HBx and that upon loss of this protection HBx is targeted for proteasomal degradation. However, it should be recognized that while we observed that ectopic expression of USP15 induced a significantly higher levels of HBx than USP5, we cannot rule out the involvement of other USPs such as USP4 and USP11 which share 71 and 60% similarity at the amino acid level, respectively 32 , contributing to HBx deubiquitylation and subsequent increase of HBx levels. It is also noteworthy that in this study, HBx itself does not have any impact on USP15 protein level and its peptidase activity although the HBx protein is known capable of targeting several components of the UPS, including DDB1, the CSN, and distinct subunits of the 26 S 33 .
DUBs have been classified as oncogenes or tumor suppressors dependent on their regulatory functions on the activity of targeted proteins involved in tumor development. Since we have shown that USP15 is essential for maintaining HBx stability and that USP15 augments HBx-mediated oncogenic signals, one inference from our work is that compromising USP15 might be a novel approach to abrogate cellular transformation and serve as a target for anti-cancer therapy. Interestingly, USP15 was found to be active in various human tumor cell lines including cervical, colon, lung, brain and kidney cancers as well as lymphomas 34 . It has been recently demonstrated that inhibition of USP15 both induced tumor cell apoptosis and boosted antitumor T cell responses, and thus have important clinical applications 35 . HBx is a promiscuous transactivator that functions to regulate HBV replication, disrupt host gene expression, affect intracellular signal transduction, accelerate cell proliferation, inhibit apoptosis, and drive HCC cell migration and invasion 36 . Full length HBx is a short-lived protein in vivo, and many factors such as 26S proteasome complex 9 , Hsp40 10 , HBV core proteins 11 , tumor suppressor p53 12 , and transcriptional factor Id-1 13 can decrease its protein level. Two E3 ligases, MDM2 and Siah-1, have been reported to be able to destabilize HBx 14,15 . Despite the ability of MDM2 to induce HBx degradation through the proteasome-dependent mechanism, MDM2 had no influence on HBx ubiquitination 14 . Conversely, Siah-1 was found to facilitate the poly-ubiquitylation modification of HBx thus predisposing it towards proteasomal degradation, and therefore attenuate its transcriptional activity 15 . Irrespective of various factors that can downregulate HBx in cells, high-level HBx is frequently observed in HCC patients and is associated with HCC progression 37 . The mechanism that sustains HBx expression in HCC at a high level is largely unknown. Our findings suggest that USP15 could protect HBx from ubiquitin-dependent proteasomal degradation and may provide a novel mechanism for the elevation of HBx that is important in the pathogenesis of HBV-related hepatocarcinoma. USP15 is a global ubiquitylation suppressor that protects many other regulatory proteins such as caspase 3, R-SMAD, TRIM25 and Nrf1 from proteasomal degradation thus affecting a wide range of signaling pathways those protein involve 27,32,38,39 . Moreover, while USP15 is not a transactivator protein accumulating evidence has shown that it can engage in promoter occupancy and stimulation on its own 27,30 . Our results obtained from the promoter reporter assays (Fig. 6) suggest that USP15 alone can stimulate various reporter genes equally well as or better than HBx. This implicates that USP15 can influence the activity of other regulatory molecules independent of HBx for transactivation function, which is further evident from the combined effect of HBx and USP15 on the Scientific RepoRts | 7:40246 | DOI: 10.1038/srep40246 promoter activation that was not even additive. Regardless, considering the ability of the viral oncoprotein HBx to affect cell functions, activate oncogenic pathways and sensitize liver cells to mutagens, it is tempting to speculate that inhibition of USP15 would interrupt chronic HBV infection, prevent the development and progression of HCC as well as foster the development of USP15-targeting strategies to expand the repertoire of molecular therapies against HCC.
CytoTrap yeast two-hybrid assay. Yeast two-hybrid verification was performed according to the manufacturer's instructions. Briefly, pMyr-USP15 was co-transformed with each of the HBx constructs into temperature-sensitive mutant yeast strain cdc25Hα , which grows normally at permissive temperature (24 °C) but needs complementation by Sos protein for survival at restrictive temperature (37 °C). After replica plating, clones that grew on SD/galactose (-UL) but not on SD/glucose (-UL) plates at 37 °C were defined as "positive" which indicates the interaction of USP15 and HBx or HBx deletion mutants. Positive controls (pSosMAFB + pMyrSB and pSosMAFB + pMyrMAFB) and negative controls (pSosMAFB + pMyrLaminC and pSosColI + pMyrMAFB) were as described previously 43 .
GST pull-down assay. E. coli Rosetta (DE3)( Novagen) transformed with pGEX-HBx or empty vector pGEX-4T-1 was grown and induced with 0.5 mM isopropyl-β -D-thiogalactopyranoside (IPTG). The cells were harvested and disrupted by sonication in interaction buffer (phosphate buffer saline (PBS) containing 5 mM EDTA, 1 mM DTT, 1 mM PMSF and protease inhibitor cocktail (Roche Diagnostics). After centrifuging, the supernatant was incubated with glutathione-sepharose 4B beads (GE Healthcare) and the GST immobilized beads were washed with interaction buffer. The purity and quantity of the bound GST and GST-HBx proteins were determined by examining SDS-PAGE gels stained with coomassie blue. TNT T7 Quick Coupled Transcription/ Translation System (Promega) was employed to express 35 S-labeled USP15 protein according to the manufacturer's instructions. In brief, 2 μ g of pCMVTNT-USP15 and 50 μ Ci of 35 S-methionine (Amersham Biosciences) were incubated with 40 μ l rabbit reticulocyte lysate for 90 min at 30 °C. For the GST-pull down assay, translated 35 S-labeled USP15 was incubated with immobilized GST-HBx or GST beads overnight at 4 °C, and then the beads were washed five times with interaction buffer. The bound proteins were subjected to 12% SDS-PAGE. After drying the gel for 15 min, the presence of 35 S-USP15 was detected by autoradiography.
Western blot analysis. Cell lysates were prepared using RIPA lysis buffer (Pierce) containing a proteinase inhibitor cocktail (Roche Diagnostics). A total of 30 μ g protein extracts were quantified and then subjected to electrophoresis on a 12% SDS-PAGE gel. The proteins were transferred to polyvinylidene difluoride (PVDF) membranes (Amersham Biosciences) and blocked in Tris-buffered saline (TBS) containing 2% bovine serum albumin (BSA). The specific antibodies used including anti-FLAG (Cell Signaling Technology), anti-USP15 (Santa Cruz), anti-K48-linkage Specific Polyubiquitin (Cell Signaling Technology), anti-ubiquitin (Cell Signaling Technology), anti-DDB1(Cell Signaling Technology), anti-USP5 (Santa Cruz), and anti-β -actin(Sigma Aldrich). Proteins were detected by addition of alkaline phosphatase (AP)-conjugated secondary antibody. Visualization of the immunoreactive proteins was performed by addition of CDP STAR reagents (Roche). Densitometry analysis of band signals was carried out using Image J software and the levels of HBx protein in cells transfected with pHBx-FLAG + pUSP15-myc were expressed as the relative intensity (RI) to that in the empty vector control pUSP5-myc or pHBx-FLAG after normalization to β -actin. inhibitor cocktail (Roche Diagnostics) and 1 mM PMSF. After centrifuging at 12,000 g for 10 min at 4 °C, 40 μ l of EZview Red ANTI-FLAG M2 Affinity Gel (Sigma Aldrich) were incubated with the supernatant overnight. For the reverse immunoprecipitation, the supernatant was pre-cleaned with 80 μ l of Protein A&G Agarose (Santa Cruz) and 0.8 μ g normal mouse IgG (Santa Cruz) for 2 hours at 4 °C with gentle rotation, and then incubated with another 80 μ l of Protein A&G Agarose and 1 μ g USP15 antibody (Santa Cruz). After washing with cell lysis buffer for three times, the immunoprecipitated complexes were separated by 12% SDS-PAGE and analyzed by western blotting using specific antibodies as described above.

Co-immunoprecipitation (Co
RNA interference and siRNA transfection. 40 pmol of USP15 siRNA mix (Santa Cruz) or the control siRNAs, together with the co-transfected plasmids were introduced into cells using lipofectamine 3000 Reagent (Invitrogen) following the manufacturer's protocol. The cells were lysed with RIPA lysis buffer (Pierce) containing a proteinase inhibitor cocktail (Roche Diagnostics) and 1 mM PMSF, and analyzed by Western blot.
Measurements of USP15 peptidase activity in vitro. The DUB-Glo Protease Assay (Promega) was used to measure the peptidase activity of USP15 in vitro following the manufacturer's protocol. Briefly, the Z-RLRGG-Glo substrate was mixed with DUB-Glo buffer and luciferin detection reagent to form DUB-Glo reagent. The DUB-Glo TM reagent was then incubated with 80 nM USP15 (Enzo Life Sciences) and increasing amount of HBx (1-80 nM) (Abcam) for 30 minutes at 22 °C. A DUB protease UCH-L3 (Boston Biochem) was used as positive control. At the end of the incubation the luminescence was measured with an Orion Microplate Luminometer (Berthold Technologies), and was recorded as relative light units (RLU). Each reaction was performed in duplicate and repeated three times.
Determination of HBx half-life. 0.75 μ g of pHBx-FLAG were co-transfected into Huh7 cells with 2.25 μg of pUSP15-myc or 40 pmol of USP15 siRNA. The equal mole amounts of pcDNA3.1/myc-His(− )A or control siRNA were used as a negative control. 48 hours posttransfection, cells were treated with 100 μ g/ml of cycloheximide (Cell Signaling Technology) at indicated times. Cells were washed three times with prechilled PBS and The cells were lysed with RIPA lysis buffer (Pierce) containing a proteinase inhibitor cocktail (Roche Diagnostics) and 1 mM PMSF, and followed by SDS-PAGE and Western blot analysis.
In vivo ubiquitylation assay. Huh7 cells in 6-well plates were transfected with 0.4 μ g pUb-HA, 0.75 μg pHBx-FLAG together with 2.14 μ g pUSP15-myc or 40 pmol of USP15 siRNA. The equal mole amounts of pcDNA3.1/myc-His(− )A or control siRNA were used as a negative control. 20 h after transfection, the cells were treated with 20 μ M proteasome inhibitor MG132 (Sigma Aldrich) for 6 h and then lysed with RIPA lysis buffer (Pierce) containing a proteinase inhibitor cocktail (Roche Diagnostics) and 1 mM PMSF. After incubation of the lysates with EZview Red ANTI-FLAG M2 Affinity Gel (Sigma Aldrich) for overnight at 4 °C, beads were washed with lysis buffer and the proteins were separated by 12% SDS-PAGE and analyzed by western blotting using specific antibodies including anti-K48 (1:2 000 dilution), anti-Flag (1:1 000 dilution) and anti-USP15 (1:500 dilution).
Cis-element luciferase reporter assay. 5 × 10 5 Huh7 cells were co-transfected with 0.75 μ g of pHBx-FLAG, 2.25 μ g of pUSP15-myc and 0.4 μ g each of pAP-1-luc, pAP-2-luc, pAP-3-luc, pSP-1-luc, pNF-κ B-luc. Equal molar amounts of pcDNA3.1/myc-His(− )A were used as a negative control. 48 h after transfection, cells were lysed and 30 μ g protein were used for the detection of intracellular luciferase activity (Bright-Glo Luciferase Assay System; Promega) following the manufacturer's protocol. The light intensity was measured by a luminometer (Berthold Technologies). The relative luciferase unit (RLU) was obtained by comparison to the empty vector pCDNA3.1/myc-His(− )A and was set to '1' in each experiment. Each transfection was performed in duplicate and repeated three times. Statistical analysis. Statistical analyses were performed with a two-tailed unpaired t test. p < 0.05 was considered statistically significant. Experiments were performed at least three times, and representative results were shown.