In this study, we enrolled 10 healthy individuals and 20 HBV-infected patients (Supplementary Table 1), and their serum and peripheral blood mononuclear cells (PBMCs) were collected from the venous blood samples. The protocol of this prospective clinical sampling study was approved by the institutional review board of the First Hospital, Jilin University.

HepG2 and HEK293T cells were maintained in Dulbecco’s Modified Eagle’s Medium supplemented with 10% inactivated fetal bovine serum, including penicillin (100 IU/mL) and streptomycin (100 mg/mL) under a 5% CO₂ atmosphere at 37°C. PHH cells were pruchased from RILD (Research Institute for liver Diseases, Shanghai, China), HBx antibody (GS968942, Gilead Sciences) was kindly provided as a gift from Dr. Simon Fletcher. The expression constructs of TRIM5γ were generated by cloning the sequence of the coding region into a VR1012 expression vector. Anti-GAPDH antibody, anti-HA-tag, and GST-tag antibody were obtained from Proteintech. Human IFN-α was purchased from Peprotech (Jiangsu, China). DDB1, STAT1, and STAT3 antibodies were procured from Cell Signaling Technology (Danvers, MA, United States). Anti-tubulin antibody was obtained from Santa Cruz Biotechnology (Santa Cruz, CA, United States), Antibodies against TRIM5γ and Smc6 were purchased from Abcam (Shanghai, China), Stattic was obtained from Sigma (Danvers, MA, United States).

The supernatants of HepG2.2.15 cells, which were derived from the human hepatoma cell line HepG2 transfected with the full genome of HBV, were collected and concentrated. PHH cells were seeded into collagen-coated 60-mm plates and cultured overnight. Cells were inoculated for 24 h with a multiplicity of infection (MOI) of 500 genome equivalents (Geq) per cell; after infection, cells were washed three times with PBS and were maintained in maintenance medium containing 2% DMSO and 2% FBS, and the cells were then incubated at the incubator. 4 days later, cells were collected and whole cell lysates were made and subjected to pull down assay, briefly, for immunoprecipitation, dynabeads protein G (Selleck) were incubated with 10 μg of monoclonal rabbit anti-HBx antibody (GS968942, Gilead Sciences), DDB1 (#5428, CST) or IgG for 10 min at room temperature. 2 mg of protein lysate was added to the dynabeads antibody complex and rotated overnight at 4°C. The beads were washed for 5 times, and boiled with 30 ul protein loading buffer and subjected to immunoblotting on the next day.

Total RNA was extracted from the cells by using the EasyPure RNA Kit (Transgen, China) according to the manufacturer’s instructions and then converted to first-strand cDNA using the TransScript First-Strand cDNA Synthesis SuperMix (Transgen). HBV DNA was isolated from supernatants as per the manufacturer’s instructions (Transgen). A housekeeping gene, GAPDH, was used as an internal control for quantitation, and the gene expression was quantified as previously described. The gene-specific primer sequences used for qRT-PCR are shown in Supplementary Table 2.

Between 24 and 48 h after the transfection of the expression plasmids, the cells were lysed with 50 mM Tris–HCl (pH 8.0), 150 mM NaCl, and 1% NP-40 containing cocktail inhibitors (Millipore). The cell lysates were immunoprecipitated and then incubated with the ANTI-FLAG^® M2 Affinity Gel (Sigma) for overnight. Immunoblotting was performed as previously described (Tan et al., 2018c). Briefly, the cells were collected and lysed in ice-cold cell lysis buffer for 30 min, with tapping of the tubes performed every 10 min. The protein concentration was quantified by the Coomassie Plus^TM Protein Assay Reagent (Thermo Scientific). The band intensities were quantified with the ChemiDoc^TM XRS + Molecular Imager software (Bio-Rad). The samples were separated by SDS-PAGE and transferred onto polyvinylidene difluoride membranes. The blots were blocked in Tris-buffered saline containing 0.1% Tween-20 and 5% skimmed milk and then probed with the relevant antibodies.

The HepG2 cells were mock-transfected or transfected with TRIM5γ-expression plasmids together with the pHBV1.3-HBV expression plasmids. The supernatant was collected after 72 h and then subjected to ELISA to detect the levels of HBV e-antigen (HBeAg) and HBV surface antigen (HBsAg) (Kehua Biotech, China).

Plasmids expressing Cas9 and sgRNA were co-transfected into HepG2 cells using lip2000 transfection reagent (Invitrogen). At 36-h post-transfection, the cells were selected with puromycin at 2 μg/mL. After 2 days, the living cells were added to a 96-well plate at a cell density of 1 cell/well. Immunoblotting was performed with TRIM5γ-specific antibodies to ensure the gene knockout (KO) results and DNA sequencing was performed to further confirm the results of gene knockout. The sequences of the sgRNAs are listed in Supplementary Table 2.

The TRIM5γ promoter reporters were generated by cloning the promoter region sequence into a pGL4.1 expression vector. The HepG2 cells were transfected with the respective TRIM5γ promoter reporters together with a pGL4.74 Tk-Rluc reporter. After 16 h, the cells were treated with IFN-α. After treatment, the cells were lysed with a passive lysis buffer and the activities of firefly luciferase and Renilla luciferase in the lysates were measured with the Dual-Luciferase Reporter Assay System (Promega).

The results were presented as mean ± SD and analyzed using the Student’s t-test. P < 0.05 was considered indicative of statistically significant differences.

In our recent study, we reported that TRIM5γ inhibited HBV replication (Tan et al., 2019). To further investigate the mechanism of TRIM5γ induction by IFN in the present study, we first stimulated PHH or HepG2 cells with IFN-α in a dose- and time-dependent manner. As expected, the TRIM5γ induction as confirmed by both q-PCR and western blotting, IFIT2, a classical Interferon Simulated Gene, was used as a positive control (Figures 1A–D). STAT1 was the major response transcriptional factor in the type-I IFN signaling pathway. Thus, we questioned whether STAT1 is critical in the induction of TRIM5γ; however, TRIM5γ remained upregulated in the STAT1 knockout cells, which was the same cell line we used previously (Tan et al., 2018b). Interestingly, the induction was blocked in the STAT3 knockout cells (Figures 1E,F), which we used as the control cells. To confirm this finding, PolydAdT, which was used as a DNA virus stimulation and induced huge amounts of interferon, was transfected into the WT or STAT1 or 3 KO cells, and as expected, TRIM5γ induction was noted in STAT1 but not in STAT3 KO cells (Figure 1G). In addition, the STAT3 inhibitor Stattic was used to treat the HepG2 cells together with IFN and the induction was inhibited (Figures 1H,I). Collectively, all these data suggested that STAT3, and not STAT1, was the major factor in the IFN-induced-TRIM5γ upregulation.

To further determine the mechanism of TRIM5γ induction in response to IFN-α stimulation, we predicted the potential STAT3-binding sites in the human TRIM5γ promoter region (5000 bps) by JASPAR (Mathelier et al., 2016). Three regions of the promoter (P1, P2, and P3), which included 1 or 2 predicted STAT3-binding sites, were inserted into the pGL4.1 luciferase reporter construct (Figure 2A). We found that IFN-α stimulation significantly induced the P3-luc activity (Figure 2B); however, this induction was blocked in the STAT3 Knockout cells (Figure 2C). This event indicated that STAT3 is the activator of TRIM5γ promoter luciferase activity. In order to further confirm the STAT3-binding site, we mutated the potential sites individually in the P3 region (Figure 2D). Interestingly, only Site #2 mutation blocked the luciferase activity induced by IFN treatment. Collectively, all these data indicated that STAT3 may directly bind to the promoter of TRIM5γ and induced its transcription.

In this study, we confirmed the TRIM5γ-HBx-DDB1 interactions in the HBV-infected PHH cells by HBx or DDB1 pulldown assay (Figures 3A,B and Supplementary Figure 1), and in addition to our research on TRIM5γ BBox domain promoting HBx degradation, we wanted to determine the binding sites of TRIM5γ by HBx. IFN promoted HBx ubiquitination and degradation by inducing TRIM5γ expression (Figures 3C,D). The BBox domain of TRIM5γ was found to be sufficient to trigger HBx (WT or R96E, an HBx mutation blocks the interaction of HBx-DDB1-Smc complexes) degradation (Figure 3E). In order to further investigate how this small peptide interacted with HBx, truncated BBox constructs were generated and CoIP was performed by co-transfection with an HBx construct. Our results indicated that HBx interacted with B2 truncation. However, both B1 and B3 protein was not stable in the cells and degraded into small fragments, therefore the interaction was not detected (Figure 3F and Supplementary Figure 2). In addition, after we truncated BBox into 4 peptides, we found that the interaction region was at the end of the BBox domain (shown as B7 in Figure 3G). Thus, these data cumulatively indicated that HBx bind to the C-terminal region of BBox.

To identify the critical site on the BBox domain that interacts with HBx, individual plasmids encoding mutated zinc binding sites of the BBox domain was generated. Interestingly, the BBox-induced degradation of HBx was rescued after H123 site mutation, but not the others in HepG2 cells. Moreover, we further confirmed that the interaction between HBx and BBox was almost blocked after H123 mutation (Figures 4A–C). Small peptides are believed to act as good candidates for drug development (Crook et al., 2020). We wondered whether a much smaller peptide can promote HBx degradation and accordingly generated constructs encoding two peptides (as indicated in Figure 4D). Surprisingly, HBx was degraded by BZ2, but not BZ1 (Figure 4D), which was consistent with our results that H123 is critical in HBx degradation. We further investigated the function of H123 in the regulation of the HBV replication. HepG2 TRIM5 KO cells were generated by the CRISPR/Cas9 technology, and WT or mutants of BBox was co-transfected with pHBV1.3 plasmids into the HepG2 WT and KO cells. As expected, HBV replication was significantly inhibited in WT BBox and mutants, including C95A, H98A, C117A plasmids transfected HepG2 cells, but not that significant in H123A mutant-transfected cells (Figures 4E,F), suggesting that H123 in the BBox domain was indispensable in the interaction and degradation of HBx as well as in the inhibition of HBV replication. In addition, on investigating the role of BZ1 and BZ2 in restricting HBV replication, we found that BZ2 was consistent with our expectation (Figures 4G,H).

In our previous study, high TRIM5γ induction was noted, which indicated a better therapeutic effect in IFN-a treatment of HBV patients. Here, we investigated the expression of TRIM5γ in HBV patients. Interestingly, TRIM5γ expression demonstrated a much different pattern in the ALT low and high groups. It was highly expressed in the ALT low group, but less expressed in the ALT high group (Figure 5A), which indicated the different immune responses during different periods of hepatitis B virus infection. In addition, the plasma of HBV patients significantly inhibited the TRIM5γ expression in HepG2 cells (Figure 5B), which was consistent with our results. In fact, TRIM5γ expression was much lower in HepG2.2.15 cells (Figure 5C), which is a cell line with the HBV genome inserted. Finally, IFN-induced TRIM5γ expression was significantly inhibited when cells (Monocytes or HepG2) were cultured with the supernatant from HepG2.2.15 cells (Figures 5D,E). Cumulatively, these data suggest that the TRIM5γ expression was inhibited by HBV.