Screening of the HBx transactivation domain interacting proteins and the function of interactor Pin1 in HBV replication

Hepatitis B virus (HBV) X protein (HBx) has been determined to play a crucial role in the replication and transcription of HBV, and its biological functions mainly depend on the interaction with other host proteins. This study aims at screening the proteins that bind to the key functional domain of HBx by integrated proteomics. Proteins that specifically bind to the transactivation domain of HBx were selected by comparing interactors of full-length HBx and HBx-D5 truncation determined by glutathione-S-transferase (GST) pull-down assay combined with mass spectrometry (MS). The function of HBx interactor Pin1 in HBV replication was further investigated by in vitro experiments. In this study, a total of 189 proteins were identified from HepG2 cells that specifically bind to the transactivation domain of HBx by GST pull-down and subsequent MS. After gene ontology (GO) analysis, Pin1 was selected as the protein with the highest score in the largest cluster functioning in protein binding, and also classified into the cluster of proteins with the function of structural molecule activity, which is of great potential to be involved in HBV life cycle. The interaction between Pin1 and HBx has been further confirmed by Ni2+-NTA pulldown assay, co-immunoprecipitation, and immunofluorescence microscopy. HBsAg and HBeAg levels significantly decreased in Pin1 expression inhibited HepG2.2.15 cells. Besides, the inhibition of Pin1 expression in HepG2 cells impeded the restored replication of HBx-deficient HBV repaired by ectopic HBx expression. In conclusion, our study identified Pin1 as an interactor binds to the transactivation domain of HBx, and suggested the potential association between Pin1 and the function of HBx in HBV replication.


Results
Profiling of proteins interacting with HBx. GST pull-down screening was performed to analyze the potential proteins interacts with full-length HBx or HBx-D5 truncated HBx. The GST protein, GST-HBx, and GST-HBx-D5 expressed in E. coli were purified by GST agarose affinity chromatography ( Supplementary Figure 1). Purified proteins were incubated with HepG2 cell lysate and glutathione-agarose beads. Samples of elution from three parallel experiments for each group were combined together and separated with SDS-PAGE gel following silver staining. Visible differences were detected between the GST lane, GST-HBx lane, and GST-HBx-D5 lane (Supplementary Figure 2), which were further confirmed by mass spectrometry analysis. Finally, a total of 402 proteins were identified by full-length HBx (Fig. 1A) and 351 proteins were detected to potentially interact with HBx-D5 truncation (Fig. 1B) after compared with GST protein to exclude non-specific bindings. www.nature.com/scientificreports/ Comparing the two groups of proteins, we filtered out 189 proteins that interact with full-length HBx but not belong to the HBx-D5 identified group (Fig. 1C,D), which refer to those bind to the transactivation domain of HBx and have great potential to affect the transactivation and replication of HBV. To further investigate the functions of the HBx interacted proteins, an enrichment analysis was carried out using DAVID online software (https:// david. ncifc rf. gov/ home. jsp). A total of 75 biological process terms were identified from Gene Otology (GO) analysis (Supplementary Table 1), such as regulation of mRNA stability, regulation of cellular amino acid metabolic process, Wnt signaling, NIK/NF-kappaB signaling, and so on. 28 significantly enriched molecular function terms were analyzed (Supplementary Table 2), such as protein binding, RNA binding, poly(A) RNA binding, structural molecule activity, and so on (Fig. 1E). Thus, except participating in the cell signaling pathways, this group of proteins specifically identified by HBx transactivation domain have high potential to participate in the life cycle of HBV, including transcription of virus, protein translation, assembly of the virus capsule, and so on (Fig. 2). Among these interactors, Pin1 attracted our attention with the highest score in the largest cluster functioning in protein binding. As a member of peptidyl-prolyl cis/trans isomerase family, Pin1 was also reported to participate in various cell signal pathways and transcriptional regulation. Besides, Pin1 was classified into the cluster of proteins with the function of structural molecule activity as well. Considering that HBx is easily to degrade with a short half-life, Pin1 protein, which could contribute to the structural integrity of proteins, is of great potential to affect the life cycle of HBV virus.
Verification of the interaction between HBx and Pin1. The interaction between Pin1 and HBx was first confirmed using Ni 2+ -NTA pull-down assay. His 6 -tagged Pin1 and Flag-tagged HBx expression plasmids were transfected into BL21 cells respectively and the cell lysates were incubated with Ni 2+ -NTA to pull-down the His 6 -tagged Pin1 protein. HBx was detected in whole cell extract as well as the eluents of the bead. As shown in Fig. 3A, HBx expression was detected in both the whole cell lysate isolated from Ni 2+ -NTA beads and high concentration of imidazole eluent, but not detected in low concentration of imidazole fluent which was for elution of unbound proteins and non-specific binding. No HBx was detected in eluent without Pin1 incubation, suggesting that HBx could interact directly with purified Pin1 in vitro. Using co-immunoprecipitation assay, the interaction between HBx and Pin1 was further confirmed in a cellular context of HepG2 cells. With the transfection of Flag-tagged HBx, endogenous Pin1 protein pulled down by HBx was detected in eluent after incubated with anti-Flag beads. In contrast, HBx was detected in eluent after incubated with anti-Pin1 beads (Fig. 3B). These results indicated the direction interaction between HBx and Pin1 in HepG2 cells. To further address the physiologically relevant interaction, immunofluorescence microscopy was conducted. The results showed the common position of Pin1 and HBx in both cytoplasm and nucleus, which further confirmed the possibility of the interaction of Pin1 and HBx in HepG2 cells (Fig. 3C).   (Fig. 4A). Considering that Pin1 was reported to enhance cell proliferation 22 , which could affect HBV propagation as HBV replication is dependent on cell cycle 23   Co-immunoprecipitation assay showed that with the transfection of Flag-tagged HBx, endogenous Pin1 protein pulled down by HBx was detected in eluent after incubated with anti-Flag beads, while HBx was detected in eluent after incubated with anti-Pin1 beads (B). Immunofluorescence microscopy shows the common position of Pin1 and HBx in both cytoplasm and nucleus (C).

Role of Pin1 in HBx augmenting HBV replication.
In addition, the potential role of Pin1 expression in process of HBx augmenting HBV replication was further investigated in HepG2 cells. As shown by southern blotting in Fig. 5A, compared with wild-type HBV, the HBx-deficient HBV genome exhibited lower levels of HBV DNA replication intermediates which could be restored with the transfection of exogenous HBx expression plasmids, indicating that HBx participates in the transcription and replication of HBV with a positive regulative function. However, with the transfection of shRNA-Pin1, HBx-deficient virus exhibited lower levels of replication intermediates comparing with in wild-type HBV, which could not be restored even with the supplement of

Role of Pin1 in Wnt signaling involved in HBV replication. Considering that the HBx interactors
identified in this study were significantly enriched in several biological process terms which have great correlation with Wnt signaling, combined with the important role that Pin1 participates in, the expression of several proteins that involved in Wnt signaling were further investigated in Pin1 shRNA transfected cells. Western blot results showed that the expression of b-catenin was induced in Pin1 shRNA transfected cells while the expression of cyclin D1 was reduced (Fig. 6A,C). However, HBx was not observed to participate in regulating the expression of b-catenin and cyclin D1 (Fig. 6B,D). Coincide with previously published studies which showed that Pin1 promotes cyclin D1 overexpression directly or through intranuclear accumulation of beta-catenin in cancer cells, our results indicated that Pin1 was involved in regulating cyclin D1 and beta-catenin expres-  www.nature.com/scientificreports/ sion in HBV replication models. However, HBx was not observed to participate in regulating the expression of b-catenin and cyclin D1. Besides, expression of correlated proteins including c-Myc and c-Jun were also investigated in Pin1 shRNA transfected cells. Expression of both c-Myc and c-Jun were observed to reduce in HepG2.2.15 cells (Fig. 6E,G). However, no significant difference was observed HepG2 cells (Fig. 6F,H), indicating that the function of HBx was not dependent on these proteins. Further investigations into the detailed manner of HBx enhancing HBV replication are still needed in the future.

Discussion
As an important determinant during the life cycle of HBV, HBx has been characterized to mediate the pathological effects of HBV according to interaction with various proteins in host cells 17 . Published studies have identified abundant proteins interact with HBx, such as jumonji C-domain-containing 5 (JMJD5) 24 , focal adhesion protein 25 , and Hsp40 26 , which participated in the HBV replication and development of HBV related hepatocellular carcinoma. However, this study is to our knowledge the first study to select the proteins that specifically interact with the transactivation domain of HBx, which can be regarded as the "key" host proteins that involved in the functions of HBx. According to GST pull-down and following MS analysis, our study identified 402 proteins interact with full-length HBx and 351 proteins interact with the N-terminal truncation of HBx which was not required in augmenting HBV replication. As a result, 189 proteins interact with the functional domain of HBx were further selected according to comparison, which are of great potential to affect the transcription and replication of HBx.
Previously published studies regarding the identification of HBx interacting proteins were mainly focused on yeast two-hybrid screen, which was criticized to have high possibility of both false positive and false negative identifications 27 . With the reduced false identification and improved sensitivity, tag-based pull-down assays conducted in vitro exhibited more advantages in researches of protein-protein interactions 28,29 . Even though expression microarray profiling and chromatin immunoprecipitation have also been applied to identify HBx target proteins in nuclear, which were mainly transcriptional regulators 19 , other proteins located in cytoplasm were likely missed. Consistent with previously published literatures that HBx located predominantly in cytoplasm 30 , our study also identified that more than half of the HBx interacted proteins located in cytoplasm (Supplementary Table 3). Considering that the interaction between proteins identified by modified target protein may introduce artificial deviations, the interactions need to be further validated. Among the 189 proteins identified in this study, there were 14 proteins reported to bind to HBx in other studies, such as DNA damage-binding protein 1 (DDB1) 31 , mitochondrial 60 kDa heat shock protein (HSPD1) 32 , peroxiredoxin-4 (PRDX4) 17 , and peptidyl-prolyl cis-trans isomerase (Pin1) 33 , which contributed to the reliability of the GST pull-down assay results.
HBx is easily to degrade with a short half-life 34,35 , which is also regarded as the barrier in the research of crystal structure of HBx and complex of its interacting proteins. The stability of the structure has been proved to associated with its transactivation ability 35,36 . Considering that Pin1 has been shown to bind to specific proteins and regulate its function according to stabilization of the targets, such as p53 and ß-catenin 37,38 , Pin1 may interact with HBx in a similar manner. Pang et al. has also identified the interaction between Pin1 and HBx by GST pull-down and Co-IP analysis, and provided evidence to the role of Pin1-HBx interaction in increasing the steady-state level of HBx and luciferase activity of HBx, as well as enhancing hepatocarcinogenesis in HBV-infected hepatocytes, which was relied on the interaction between the two proteins 22 . However, the effect of Pin1 in HBx augmenting HBV replication has not been defined yet. Based on their study, our study further confirmed the interaction between HBx and Pin1 in HepG2 cells, showed a significantly decreased HBsAg and HBeAg levels in Pin1 expression inhibited HepG2.2.15 cells, indicating that Pin1 has a potential function of regulating HBV replication. Besides, the HBV replication ability in HepG2 cells transfected with HBx-deficient HBV could be restored by ectopic HBx expression, but the inhibition of Pin1 expression could not be restored to wild-type of HBV anymore.
As a regulator participates in ß-catenin signaling and cyclin D1, the interaction between Pin1 and ß-catenin inhibits the degradation and transportation resulting the stabilization and accumulation of ß-catenin in nuclear and subsequent activation of transcriptional factors 37 . Besides, Pin1 can enhance the transcriptional activity of cyclin D1 by binding directly to cyclin D1 or activating the protein 1 (AP-1) site in promoter, to participate in the subcellular localization and stabilization of cyclin D1 39 . As described in previous studies, HBx participates in the transactivation of transcriptional factors or target genes such as c-myc, AP-1, and cyclin D1 16,40-42 , suggesting the similar pathway and common targets of Pin1 and HBx in molecular biological process, which is also of great potential to be further investigated in pathogenic of HBV and HCC. However, HBx was not observed to regulate the expression of cyclin D1 and ß-catenin in either control group or Pin1 inhibited cells in our study, as well as two other correlated proteins including c-Myc and c-Jun. Further studies investigating the mechanism of HBx regulating HBV replication are highly recommended in the future.
Recent study has also investigated the function of Pin1 on regulating HBV replication. Nishi et al. 43 showed that shPin1 was associated with reduced expression of HBcAg and inhibited HBV replication in HepG2.2.15 cells, which is coincide with our results. They also demonstrated that Pin1 binds and stabilizes hepatitis B virus core protein (HBc) in a phosphorylation-dependent manner, and promotes the efficient viral propagation. However, according to their findings, Pin1 is very likely to participated in HBV replication through different manners. Other study proved that pin-1 enhances cell proliferation and also induces several proteins in the host cell, which could also be a possible pattern to regulating HBV propagation as it is critical that HBV replication is dependent on cell cycle. Although no influence of Pin1 on cell proliferation was observed in the current study, we showed that Pin1 might be associated with the function of HBx in HBV replication, which also provided new insights into further studies of HBV replication. www.nature.com/scientificreports/ In addition to HBV, Pin1 has been proved to participate in the replication of other virus. Lim et al. showed that Pin1 interacts directly with NS5A and NS5B proteins of HCV and plays unique roles in HCV replication 44 . Shogo et al. has demonstrated that in the life cycle of HIV-1, the uncoating event of capsid core was promoted by virion incorporated extracellular signal-regulated kinase 2 (ERK2) 45 and requires an interaction of the capsid protein with Pin1 at a specific phosphorylated Ser16-Pro17 motif, which also affect the replication efficiency of the virus 46 . Besides, Pin1 regulates Epstein-Barr virus (EBV) DNA replication through interacting with EBV DNA polymerase catalytic subunit 47 and participates in epstein barr virus (EBV)-related nasopharyngeal carcinoma 48 . In addition, Pin1 has also been proved to modulate the replication and propagation of feline coronavirus (FCoV) 49 , and cyprinid herpesvirus 2 (CyHV-2) 50 . Given that most of the interactions between Pin1 and other proteins rely on the Thr/Ser-Pro motif, the binding site of Pin1 and HBx are of highly potential to be the Ser39-Pro40 or Ser144-Pro145 region. Even though Pang et al. demonstrated the binding at Ser39-Pro40 of HBx 22 , our study suggesting the interacting region was more likely to locate on the 51-154 amino acid of HBx. Thus, further researches regarding the binding sites of Pin1 and HBx are highly recommended in the future.
In conclusion, we identified 401 proteins bind to full-length HBx by GST pull-down assay combined with MS analysis, including 189 proteins interact specifically with the transactivation domain of HBx. The interaction between Pin1 and HBx was further confirmed, and Pin1 was suggested to be associated with the function of HBx in HBV replication.
The plasmid payw1.2 (1.2 wt, subtype ayw) was used to investigate HBV replication, which has been described in previous study 12 and contains 1.2 copies of the wild-type HBV genome. The HBx-minusmutant vector payw*7(1.2x( −)) was used to investigate the function of HBx in HBV replication, which was also described previously and contains 1.2 copies of HBx-minus HBV genome. Preparation of GST fusion proteins. pGEX-6p-1-HBx and pGEX-6p-1-HBx-D5 plasmids containing coding sequences for full-length HBx or HBx N-terminal truncation HBx-D5 were transformed into E. coli. strain BL21. Single colonies were grown on Luria-Bertani (LB) plate overnight at 37 °C and selected for enlarge cultivation in LB medium at 37 °C until the optical density (600 nM) reached 0.6 following induction with 0.1 mM isopropyl-β-D-thiogalactopyranoside (IPTG) overnight at 20 °C. Proteins were obtained by sonication of the cells and purified by agarose affinity chromatography, and then solubilized to glutathione-Sepharose beads (GE Healthcare, USA) according to the instructions provided by manufacturer.
GST pull down assay. HBx and HBx-D5 proteins solubilized to glutathione-Sepharose beads were incubated with HepG2 cell lysate overnight at 4 °C with the GST protein as the negative control. 3 sets of parallel experiment were carried out at the same time. Beads were washed with respective incubation buffer to remove the unbound proteins. Proteins bound to the beads were collected in elution buffer by centrifugation. Protein samples obtained from the parallel experiments were mixed together and concentrated for SDS-PAGE separation followed by silver staining.
In-gel digestion and MS/MS identification. Bands of each samples were cut out of the gel into 6 parts and decolored with 50 mM NaS 2 O 3 and 15 mM K 3 Fe(CN) 6 following dehydration with 100% acetonitrile (ACN). Samples were firstly incubated with 10 mM DTT in 25 mM ammonium bicarbonate (ABC) at 60 °C for 1 h and then with 55 mM indole-3-acetic acid (IAA) at room temperature for 45 min, following washing step with 50% ACN and 25 mM ABC. Proteins were digested by sequencing grade modified trypsin (Promega, USA) in 50 mM ABC at 37 °C for 12-16 h and extracted by double extraction with 75% ACN, 0.1% TFA. Samples were dried and redissolved in 0.1% fomic acid for mass spectrometry analysis.
Ni2 + -NTA pulldown assay. E Detection of HBV cccDNA and pgRNA. pgRNA in cell-culture supernatant was detected by RT-PCR.
Total RNA from the supernatants was extracted by using RNA extraction kit (Takara, China) and reverse transcribed using a PrimeScript™ RT reagent Kit to produce cDNA; RT-PCR was performed to detect pgRNA using Premix Ex Taq™ (Takara, China) using the primers 5'-CAC CTC TGC CTA ATC ATC -3' and 5'-GGA AAG AAG TCA GAA GGC AA-3' . The cccDNA levels in cells were quantified by quantitative polymerase chain reaction (qPCR). Total DNA was extracted from the cells using a TIANamp Genomic DNA Kit (Tiangen, China) following DNase reagent (Takara, China) to enhance the efficiency of the specific extraction of cccDNA. qPCR was performed using primers 5'-GAC TCC CCG TCT GTG CCT TCT CAT C' , and 5'-AGA CCA ATT TAT GCC TAC AGC CTC C-3' for cccDNA amplification using SYBR Premix Ex Taq (Takara, China).
Amplification was performed as follows: 95 °C for 30 s, 40 cycles of 95 °C for 5 s and 60 °C for 34 s, 95 °C for 15 s and 60 °C for 1 min. Southern blotting. HBV DNA replication intermediates were extracted and determined according to the protocol described in previous literatures 13 . Samples were resuspended in 30 μL of tris-ethylene diamine tetraacetic acid (TDTA) buffer and separated with 1% agarose gel. HBV replication intermediates samples were transferred to Hybond-N + membrane (Amersham Biosciences, UK) and detected using digoxigenin-labeled full-length HBV DNA sequence as probe following further detection with the DIG Luminescent Detection Kit for Nucleic Acids (Roche, Germany).
Bioinformatics and statistical analysis. The HBx PPI network was analyzed and built using the Cytoscape software 51 (v3.8.2, Washington, USA, https:// cytos cape. org/). Gene ontology analysis was conducted using the Database for Annotation, Visualization and Integrated Discovery (DAVID) v6. 8