To generate an HBV replication competent plasmid, a 1.3-fold over length fragment of HBV wild-type (WT) subtype adw R9 genome was cloned into the pGEM4z (Promega) vector and then subcloned to pcDNA3; the resulting construct was designated HBV WT 1.3mer (adw). HBV WT, the priming reaction-deficient TP-Y65F mutant, the RT reaction-deficient RT-YMHA mutant, and the P-deficient mutant, in which transcription of pgRNA is controlled by the cytomegalovirus immediate early (CMV IE) promoter (Kim et al., 2004), were used for the study. Also, an HBx-deficient mutant, in which HBx expression was abolished completely by mutation of three ATGs in the HBx open reading frame, was placed under the control of the CMV IE promoter (Yoon et al., 2011). A subtype ayw 1.3mer HBV WT plasmid was kindly provided by Dr. Ryu WS (Yonsei Uni, South Korea). With the exception of the ayw 1.3mer HBV WT plasmid, all of the HBV constructs used in this study had the adw HBV background. Myc- or HA-tagged HBc WT was constructed by PCR-amplification of an HBc sequence on a P-deficient mutant background (Kim et al., 2004; Jung et al., 2014); the primers used are listed in the Table. The amplified PCR product was digested with Msc I and Kpn I, gel-purified, and inserted into a pCMV10-Myc or pCMV-HA plasmid to generate pCMV10-Myc-HBc WT (SST-STSSSS) or pCMV10-HA-HBc WT, respectively. Myc-tagged HBc mutants were constructed by PCR-generated site directed mutagenesis of pCMV10-Myc-HBc WT, thereby generating pCMV10-Myc-HBc-Y132A, pCMV10-Myc-HBc-SST-SAAAAA, pCMV10-Myc-HBc-SST-AASAAA, pCMV10-Myc-HBc-SST-AAAASA, pCMV10-Myc-HBc-S157A, pCMV10-Myc-HBc-T162A, pCMV10-Myc-HBc-S164A, pCMV10-Myc-HBc-S172A, pCMV10-Myc-HBc-S44A, pCMV10-Myc-HBc-S49A, pCMV10-Myc-HBc-T128A, and pCMV10-Myc-HBc-AAA-STSSSS. The sequences of the primers used to construct the HBc WT and HBc mutants are listed in the Table. To generate the Myc-tagged HBc-SST-ATAAAA and HBc-SST-AAAAAA mutants, mutant HBc DNA fragments from previously published ATAAAA and AAAAAA mutants of HBc on a P-deficient mutant background (Jung et al., 2014) were inserted into a Msc I/Kpn I-linearized pCMV10-Myc plasmid to yield pCMV10-Myc-HBc-SST-ATAAAA and pCMV10-Myc-HBc-SST-AAAAAA, respectively. To generate the Myc-tagged HBc-AAA-AAAAA and -AAA-AAASAS mutants, site directed mutagenesis was conducted using the primers listed in the Table, and the mutant HBc DNA PCR fragment was inserted into a Msc I/Kpn I-linearized pCMV10-Myc plasmid to yield pCMV10-Myc-HBc-AAA-AAAAA and pCMV10-Myc-HBc-AAA-AAASAS, respectively. HBc-ΔCTD comprising the NTD 145 amino acids of HBc has been described (Jung et al., 2012). To generate a WT Pin1 expression construct, the PIN1 cDNA was synthesized from total RNA of HepG2 cells using superscript III reverse transcriptase (18080-044, Invitrogen™) and a reverse primer specific for PIN1 WT. Next, the cDNA was amplified by PCR using the forward and reverse primers for PIN1 WT (Table). The amplified DNA fragment was inserted into Hind III/BamH I linearized pcDNA3.1 (Invitrogen™). PCR-generated site directed mutagenesis was used to construct dephosphorylation mimetic substrate binding-positive S16A (AGC → GCC), phosphorylation mimetic substrate binding-deficient S16E (AGC → GAA), substrate binding-deficient W34A (TGG → GCT), PPIase-active S71A (TCG → GCG), and PPIase-inactive C113A (TGC → GCC) mutants (Table). The resulting Pin1 mutant plasmids were designated as pcDNA3-Pin1-S16A, pcDNA3-Pin1-S16E, pcDNA3-Pin1-W34A, pcDNA3-Pin1-S71A, and pcDNA3-Pin1-C113A. The pCDH-CMV-PIN1-EF1-Puro plasmid encoding PIN1 in a lentiviral vector was generated by inserting the amplified PIN1 WT DNA fragment into the EcoR I and BamH I sites of pCDH-CMV-MCS-EF1-Puro (System Biosciences). Four short hairpin (sh)RNA (shPIN1 #1, 2, 3, and 4) constructs in the pLK0.1 vector, which target multiple regions of the PIN1 gene, were purchased from Sigma-Aldrich (SHCLNG-NM_006221). Control shRNA harboring random sequences (5’-ACG TGA CAC GTT CGG AGA AC) was inserted into pLK0.1 to generate a scrambled shRNA control. The primer sequences used to construct PIN1 WT and its mutants, and the shRNAs targeting PIN1 are listed in the Table. All constructs were sequenced to confirm the presence of the introduced mutations and the absence of extraneous mutations.

HepG2, Huh7, or HEK293T cells were grown in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS; Gibco-BRL) and 1% penicillin/streptomycin under a humidified atmosphere at 37°C/5% CO₂. HepG2 and Huh7 cells and HEK293T cells were passaged every third and second day, respectively. HepG2.2.15 cells were grown in the same culture medium as HepG2 cells, except for addition of 0.2 mg/ml G418 (G0175.0005, Duchefa Biochemie). HepAD38 cells (kindly provided by Christopher Seeger, Fox Chase Cancer Center) were also grown in the same culture medium as HepG2.2.15 cells, except that tetracycline (1 µg/ml, T7660-5G, Sigma-Aldrich) was added. The tetracycline was removed to induce HBV transcription and production of HBV virions. For transfection into 1×10⁶ HEK293T cells in 6 cm culture plates, 2 µg of each construct was mixed with 4 µg/µl polyethylenimine (PEI; Polysciences) in 200 µl of OPTI-MEM (Gibco). For transfection into 2×10⁶ HepG2 or HepG2.2.15 cells in 6 cm culture plates, 3 µg of each construct was mixed with 9 µg/µl PEI in 200 µl of OPTI-MEM and poured onto the cells. For transfection into 1×10⁶ Huh7-Pin1 cells in 6 cm culture plates, 3 µg of each construct was mixed with 9 µg/µl PEI in 200 µl of OPTI-MEM and poured onto the cells. For cotransfection into HEK293T cells in 6 cm culture plates, 2 µg of Pin1-expressing plasmid plus 2 µg of HBV WT, HBc WT, HBc-ΔCTD, Myc-tagged HBc WT, Myc-tagged HBc S/TP site mutants, Myc-tagged-HBc-R133D mutant, Myc-tagged-HBc-R133E mutant, or Myc-tagged HBc-Y132A mutant were mixed with 8 µg/µl PEI in 200 µl of OPTI-MEM and poured onto the cells. For cotransfection into HepG2 cells in 6 cm culture plates, 3 µg of 1.3mer HBV WT plus 3 µg of Pin1 WT or empty pcDNA3.1 vector was mixed with 18 µg/µl PEI in 200 µl of OPTI-MEM and poured onto the cells. Also, HepG2 cells were cotransfected with a mixture of 3 µg of Pin1 WT plus 3 µg of HBV WT (adw) or HBx-deficient HBV (adw) mutant, plus 18 µg/µl PEI in 200 µl of OPTI-MEM. The amount of pcDNA3 was adjusted to the amount of transfected DNA. At 24 h post-transfection, the culture medium containing transfected DNA was exchanged for fresh complete medium.

Huh7 and HepAD38 cells stably expressing human Pin1 were generated using a lentiviral expression system. Briefly, 2×10⁶ HEK293T cells were seeded on 10 cm plates and transfected with 1 µg of pVSV-G, 3 µg of pGag-Pol, and 4 µg of pCDH-CMV-Pin1-EF1-Puro in 16 µg of PEI/500 µl of OPTI-MEM. One day after transfection, the medium was replaced, and 2 days after transfection, the supernatant containing the pseudoviral particles containing Pin1-transcripts was harvested. Next, 1 ml of supernatant containing pseudoviral particles was mixed with 3 ml of fresh complete medium containing 10 µg/ml polybrene (Hexadimethrine bromide, Sigma-Aldrich) and inoculated onto Huh7 or HepAD38 cells in 6 cm plates. After transduction with lentiviral pseudoviral particles, human Pin1 WT-expressing Huh7 or HepAD38 cells were established by selection with 10 µg/ml puromycin (Sigma-Aldrich) for 72 h. To generate PIN1 KD HepG2 and PIN4 KD HEK293T cells, HEK293T cells were transfected with 1 µg of pVSV-G, 3 µg of pGAG-Pol, and 4 µg of pLK0.1-shPIN1 (#1, 2, 3, and 4); pLK0.1-shPIN4 #1; or pLK0.1-shControl in 16 µg of PEI/500 µl of OPTI-MEM. Then, HepG2 or HEK293T cells were transduced with supernatant containing pseudoviral particles and shRNAs (shPIN1 or shPIN4) or control shRNA, as mentioned above, and then selected with puromycin. Stable HepG2-hNTCP-C9 cells were generated as described previously using lentiviral expression vector pCDH-CMV-hNTCP-C9-EF1-Puro (Kim et al., 2022). HepG2-hNTCP-C9-derived PIN1 knockdown cells were generated as described above.

For immunoprecipitation, HEK293T cells were (co)transfected with Myc-HBc WT, Pin1 WT, or Myc-HBc WT plus Pin1 WT. Also, HEK293T cells were cotransfected with Pin1 WT plus the Myc-HBc WT (SST-STSSSS) or Myc-HBc mutants (Y132A, AAA-STSSSS, SST-AAAAAA). At 2 days post-transfection, cells were harvested, lysed in 0.2% NP-40 (IGEPAL, Sigma-Aldrich)-TNE buffer (10 mM Tris-HCl [pH 8.0], 50 mM NaCl, and 1 mM EDTA) containing 1 mM PMSF (P7626, Sigma-Aldrich), as previously described (Kim et al., 2004); immunoprecipitated with a rabbit polyclonal anti-HBc (produced in-house; 1:1,000) (Jung et al., 2012); and then immunoblotted with mouse monoclonal anti-Pin1 (1:1,000; sc-46660; Santa Cruz) antibodies, or vice versa. Rabbit or mouse normal IgG was used as a negative control for immunoprecipitation (12-370, 12-371, Merck-Millipore). To analyze the structural stability of the HBc NTD S/TP motif mutants, the cell lysates were exposed to freezing temperatures (approximately -20°C) for 16 h, followed by thawing at room temperature for 1 h. This freeze-thaw cycle was repeated two or four times. For reducing SDS-PAGE, sample buffer containing 125 mM Tris-HCl [pH 6.8], 20% glycerol, 4% SDS, 0.1% bromophenol blue, and 5% β-mercaptoethanol was used. For nonreducing PAGE, the sample buffer was as described above, except that the 5% β-mercaptoethanol was omitted (Saeed et al., 2021). For SDS-PAGE and immunoblotting, equal quantities of protein from cell lysates were subjected to SDS-PAGE on 13.5% gels, transferred to polyvinylidene fluoride (PVDF) membranes (Millipore), and then incubated with the appropriate primary antibodies: rabbit polyclonal anti-HBc (produced in-house; 1:1,000) (Jung et al., 2012), mouse monoclonal anti-Myc (1:1,000; sc-40; Santa Cruz), mouse monoclonal anti-Pin1 (1:1,000; Santa Cruz), rabbit monoclonal anti-CypA (1:1,000; ab126738; Abcam), mouse monoclonal anti-FKBP12 (1:1,000; sc-133067; Santa Cruz), mouse monoclonal anti-GAPDH (1:1,000; sc-32233; Santa Cruz), rabbit monoclonal anti-PIN4 (1:1,000; ab155283; Abcam), rabbit polyclonal anti-HBs (1:1,000; 1811; Virostat) for NAGE, mouse monoclonal anti-HBs (1:1,000; sc-53299; Santa Cruz) for SDS-PAGE, mouse monoclonal anti-PreS1 (1:1,000; sc-57761; Santa Cruz), or mouse monoclonal anti-rhodopsin C9 (1:1,000; MAB5356; Millipore). Primary antibodies were followed by anti-rabbit (HRP; 1:5,000 dilution; #31460; Thermo Fisher Scientific) or anti-mouse (HRP; 1:5,000 dilution; 5220-0460; Seracare) secondary antibodies coupled to horseradish peroxidase. The blots were then visualized by enhanced chemiluminescence (ECL™ western blotting detection reagent, Amersham™). Relative band intensities were measured using ImageJ. 1.50b.

To analyze HBV core particles by immunoblotting, cell lysates in 0.2% NP-40-TNE buffer were electrophoresed on 1% native agarose gels, and the resolved core particles were transferred to PVDF membrane for immunoblotting with a rabbit polyclonal anti-HBc primary antibody (produced in-house; 1:1,000) (Jung et al., 2012), followed by an anti-rabbit secondary antibody conjugated to HRP (1:5,000; Thermo Fisher Scientific). The bound secondary antibodies were visualized by ECL. Relative band intensities were measured using ImageJ. 1.50b.

HBV particles were prepared from culture supernatants of HepAD38-pCDH, HepAD38-Pin1, and HBV-infected HepG2-hNTCP-C9 cells as described previously (Watashi et al., 2013; Ko et al., 2014; Kim et al., 2022), with minor modifications. Culture supernatants comprising F-12 DMEM (11320-033, Gibco) supplemented with 10% FBS, 1% penicillin/streptomycin, 5 μg/ml insulin (#I9278, Sigma-Aldrich), and 50 μM hydrocortisone hemisuccinate (#1319002, Sigma-Aldrich) were collected 4 days after tetracycline removal, centrifuged briefly, and filtered through a 0.45 μm bottle top filter to remove cell debris. The filtered supernatants were layered onto a 20% sucrose cushion (w/w in TNE buffer) and then ultracentrifuged (Optima L-90K, Beckman Coulter) at 26,000 rpm (115,633 g) for 2 h at 4°C. Pellets containing virions, subviral particles, and naked core particles were resuspended in TNE buffer, electrophoresed on 1% native agarose gels, transferred to a PVDF membrane, and immunoblotted. Core particles, virions, Pin1-bound-core particles, and/or subviral particles were detected using rabbit polyclonal anti-HBc (produced in-house; 1:1,000) (Jung et al., 2012), mouse monoclonal anti-PreS1 (1:1,000; Santa Cruz), rabbit polyclonal anti-HBs (for NAGE; 1:1,000; Virostat), mouse monoclonal anti-HBs (for SDS-PAGE; 1:1,000; Santa Cruz), or mouse monoclonal anti-Pin1 (1:1,000; Santa Cruz) primary antibodies, followed by HRP–conjugated anti-rabbit (1:5,000; Thermo Fisher Scientific) or anti-mouse (1:5,000; Seracare) secondary antibodies. The bound secondary antibodies were visualized by ECL. The relative band intensities were measured using ImageJ. 1.50b.

The HBV inoculum used for infection was harvested from the culture supernatant of HepAD38 cells as described previously, with minor modifications (Watashi et al., 2013; Ko et al., 2014; Kim et al., 2022). Briefly, the culture supernatants were collected every 2-3 days, centrifuged, and filtered as described above. For partial purification of HBV particles, the filtered supernatants were ultracentrifuged through a 20–60% discontinuous sucrose gradient at 26,000 rpm for 3 h at 4°C. Next, the HBV fraction at the interface (between 30–50%) was collected and then pelleted through a 20% sucrose cushion by ultracentrifugation at 26,000 rpm for 1.5 h at 4°C. The amount of HBV DNA in the precipitate after ultracentrifugation was quantitated by southern blotting. For HBV infection, 2×10⁵ HepG2-hNTCP-C9 cells were seeded onto collagen-coated (#354249, Corning) 6-well plates and infected with 2×10³ genome equivalents (GEq) of HBV per cell in medium containing 4% polyethylene glycol (#25322-68-3, Affymetrix) and 2.5% dimethyl sulfoxide (DMSO), as described previously (Ni et al., 2014). One day after infection, the cells were washed thoroughly with 1× PBS (at least three times) to remove residual HBV and then maintained in the same fresh medium containing 2.5% DMSO, as described previously (Ni et al., 2014). The medium was changed every second day. Lysates were prepared at 9 days postinfection and subjected to SDS-PAGE and immunoblotting, native agarose gel electrophoresis (NAGE) and immunoblotting, and southern blot analyses. For northern blot analysis, total cell lysates were prepared at 5 days postinfection.

To study the interaction between Pin1 and the core particle, HEK293T cells were cotransfected with Myc-HBc plus empty vector or with Pin1 WT. At 48 h post-transfection, the lysates were either left untreated (control) or heated at 65°C for 2 h to dissociate bound Pin1 from the core particle, followed by incubation for 5 min on ice. Then, lysates were subjected to NAGE plus immunoblotting, and SDS-PAGE plus immunoblotting, as described above. To detect Pin1 inside the core particle, the core particle on the PVDF membrane was treated with 0.2 N NaOH for 1 min to open it up, followed by UV-crosslinking (XL-1500 UV CROSSLINKER, Spectrolinker™) and immunoblotting with anti-Pin1, anti-HBc, and anti-Myc antibodies.

HBV WT-transfected HEK293T cell lysates containing core particles were prepared as described above and treated for 1 h at 37°C with 1 unit of calf intestinal alkaline phosphatase (CIAP) (#290S, NEB) to remove phosphates from the core particles. Mock-treated HBV WT-transfected HEK293T cell lysates incubated for 1 h at 37°C in the absence of CIAP were used as a positive control. The reactions were stopped by heating to 75°C for 10 min. Pin1-overexpressing lysates were mixed with either mock-treated lysates or CIAP-treated lysates and then further incubated for 0, 12, 24, and 48 h at 37°C, followed by NAGE plus immunoblotting and SDS-PAGE plus immunoblotting, as described above.

PIN4-KD HEK293T cells (5×10⁵) were plated onto 6-well plates and transfected with HBV WT (Kim et al., 2004) plus the mock, Pin1 WT, or Pin1 W34A mutant. At 12 h post-transfection, 50 µg/ml cycloheximide (C1988-1G, Sigma-Aldrich) was added to inhibit protein synthesis, and cell lysates were prepared at 0, 12, 24, and 48 h post-treatment for NAGE plus immunoblotting (for core particle analysis) and SDS-PAGE plus immunoblotting (for analysis of HBc and Pin1 proteins).

To analyze HBV DNA synthesis by southern blotting, HBV DNA extracted from isolated core particles was separated by agarose gel electrophoresis, transferred to a nylon membrane (11417240001, Sigma-Aldrich), and hybridized to a ³²P-labeled random-primed probe specific for the full-length HBV sequence, as described previously (Jung et al., 2014). To analyze HBV RNA expression by northern blotting, total RNA was extracted using TRIzol reagent (15596026, Thermo Fisher Scientific). In brief, 20 μg of total RNA was denatured at 65°C for 10 min and then electrophoresed on 1% agarose gels (16500500, Thermo Fisher Scientific) containing formaldehyde (F8775, Sigma-Aldrich) and 1× morpholinepropanesulfonic acid (MOPS) buffer [200 mM MOPS, 10 mM EDTA, 50 mM sodium acetate (pH 7.0)]. RNA was then transferred to a nylon membrane and hybridized at 68°C for 4 h to a ³²P-labeled random-primed probe specific for the full-length HBV sequence. In situ nucleic acid blotting was performed to analyze HBV DNAs inside virions and/or naked core particles. Briefly, pellets containing virions and naked core particles were resuspended in TNE buffer, and cell lysates containing cytoplasmic core particles in 0.2% NP-40-TNE buffer were electrophoresed on 1% native agarose gels prior to transfer to a PVDF membrane. The PVDF membrane was treated with 1% SDS for 1 h, followed by 0.2 N NaOH for 30 s. It was then washed quickly with distilled water, UV-crosslinked (XL-1500 UV CROSSLINKER, Spectrolinker™), and hybridized with a ³²P-labeled random-primed probe specific for the full-length HBV sequence. In situ nucleic acid blotting was also performed to detect minus-strand DNA in cytoplasmic core particles using a synthesized Dig-labeled plus-strand RNA probe specific for the full-length HBV sequence. Relative band intensities were measured using ImageJ. 1.50b.

To examine the effects of Pin1 on formation of HBV cccDNA, the latter was extracted using a Hirt protein-free DNA extraction procedure, as described previously (Cai et al., 2013) but with minor modifications. In brief, 3×10⁶ HepG2-hNTCP-C9 cells were seeded into collagen-coated 10 cm dishes and infected with 2×10³ GEq HBV, as described above. At 7 days postinfection, the cells were lysed for 30 min at room temperature (on a shaker) with 0.6% SDS-TE buffer (10 mM Tris-HCl [pH 7.5], 10 mM EDTA). The lysates were treated with 5 M NaCl to adjust the final NaCl concentration to 1 M, incubated for 16 h at 4°C, and centrifuged at 14,500 g for 30 min. The supernatants were extracted twice with phenol, followed by once with phenol-chloroform, and then precipitated with ethanol for 16 h at room temperature. Finally, the precipitates were analyzed by southern blotting. To further verify the authenticity of HBV cccDNA, the Hirt DNA sample was heated to 85°C for 5 min. Among the three types of HBV DNAs [cccDNA (2.1 Kbp), DL DNA (3.2 Kbp), and protein-free RC DNA (more than 3.2 Kbp)] in the Hirt protein-free DNA extracted sample, RC and DL DNA, but not cccDNA, were denatured. The electrophoretic mobility of cccDNA is not altered by heat treatment. The heat-treated DNA samples were then digested with EcoR I to linearize the cccDNA and yield genome-length double-stranded DNA (3.2 Kbp).

The four HBc CTD S/TP motifs (¹⁵⁷SP, ¹⁶²TP, ¹⁶⁴SP, and ¹⁷²SP) are phosphoacceptor sites ( Figure 1A ) (Kann and Gerlich, 1994; Lan et al., 1999; Gazina et al., 2000; Melegari et al., 2005; Lewellyn and Loeb, 2011b; Ludgate et al., 2012; Jung et al., 2014) that can serve as Pin1-binding sites (Nishi et al., 2020). To verify the physical interaction between Pin and HBc proteins (Nishi et al., 2020), we performed coimmunoprecipitation experiments ( Figure 1B ) using (co)transfected cell lysates. Lysates were immunoprecipitated with an anti-HBc antibody and immunoblotted with an anti-Pin1 antibody, and vice versa ( Figure 1B ). The anti-HBc and anti-Pin1 antibodies coimmunoprecipitated Pin1 and HBc, respectively ( Figure 1B , lane 3), confirming the interaction between Pin1 and HBV HBc protein (Nishi et al., 2020). However, coimmunoprecipitation, SDS-PAGE, and immunoblotting cannot discriminate Pin1/HBc interactions from Pin1/core particle interactions.

Therefore, to examine whether the HBV core particle is a binding substrate of Pin1, we performed NAGE of cotransfected cell lysates, followed by immunoblotting ( Figure 1C ). We found that Pin1 bound to the core particle ( Figure 1C , top panel; lane 3). Thus, Pin1 can interact physically with HBV core particles, a finding not reported before. However, we showed previously that parvulin proteins Par14 and Par17 interact with both HBc and the core particle (Saeed et al., 2021). It should be noted that when we overexpressed Pin1, the levels of both HBc and the core particle increased ( Figure 1C , second, third, fifth, and sixth panels; lane 2 vs. 3).

Since PPIases bind to Xaa-Pro motifs (Hanes, 2015), and HBc contains 16 Xaa-Pro motifs, we performed NAGE and immunoblotting to examine whether the core particle binds to cyclophilins and/or FKBPs. The Pin1/core particle interaction was detected after an exposure time of only 5 sec ( Figure 1D , top panel); however, the cyclophilin A/core particle interaction could not be detected even after a long exposure (3 h) ( Figure 1D , second panel), indicating that cyclophilin A cannot bind to the core particle. By contrast, FKBP12/core particle binding was weakly detectable after a long exposure (3 h) ( Figure 1D , third panel). These results indicate that the interaction between Pin1 and the core particle is more specific than that between the core particle and other PPIases, suggesting that Pin1 may contribute to HBV proliferation via its interaction with the core particle.

Although we demonstrated the physical interaction between Pin1 and the core particle for the first time, we did not differentiate the Pin1/HBc interaction from the Pin1/core particle interaction ( Figures 1B−D ). Therefore, to check whether Pin1 interacts with HBc, we used a core particle assembly-defective, dimer-positive ¹³²Tyr to Ala (Y132A) mutant of HBc (Bourne et al., 2009) in a coimmunoprecipitation experiment with an anti-HBc antibody, followed by immunoblotting with anti-Pin1 antibody. Although expression of the Myc-HBc-Y132A mutant protein was comparable with that of Myc-HBc WT ( Figure 1E , third panel; lane 1 vs. 2), core particle assembly by Myc-HBc-Y132A was defective, whereas that by Myc-HBc WT was intact ( Figure 1E , bottom panel; lane 1 vs. 2) (Bourne et al., 2009). As shown in Figure 1B , Pin1 coimmunoprecipitated with Myc-HBc WT but not with the core particle-defective Myc-HBc-Y132A mutant ( Figure 1E , top panel; lane 1 vs. 2), demonstrating that neither dimers nor monomers of HBc are binding partners of Pin1. This clearly demonstrates that the HBV core particle is a bona fide binding substrate of Pin1.

To further identify the importance of the S/TP motifs in the NTD and CTD of HBc WT (⁴⁴S⁴⁹S¹²⁸T-¹⁵⁷S¹⁶²T¹⁶⁴S¹⁷⁰S¹⁷²S¹⁷⁸S, shortened to SST-STSSSS), Myc-tagged HBc-AAA-STSSSS- or Myc-tagged HBc-SST-AAAAAA-transfected lysates were immunoprecipitated with the anti-HBc antibody and then immunoblotted with the anti-Pin1 antibody ( Figure 1E , top panel; lanes 1 vs. 3 and 4). Unlike Myc-tagged HBc WT and Myc-tagged HBc-AAA-STSSSS, Myc-tagged HBc-SST-AAAAAA was not immunoprecipitated, demonstrating that CTD phosphoacceptor S/TP motifs are important for Pin1-binding ( Figure 1E , top panel; lanes 1 and 3 vs. 4). It should be noted here that the intensity of the Myc-HBc-Y132A and Myc-HBc-SST-AAAAAA bands detected by the anti-HBc and anti-Myc antibodies was weaker than that of the Myc-HBc WT and Myc-HBc-AAA-STSSSS bands ( Figure 1E , second panel; lanes 1, 3 vs. 2, 4).

Previously, we showed that the HBc ¹³³Arg to Asp (HBc-R133D) and HBc ¹³³Arg to Glu (HBc-R133E) mutants were particle assembly-defective and dimer-positive (Saeed et al., 2021), as is HBc-Y132A. Consistent with this, NAGE and immunoblotting revealed that HBc-R133D and -R133E are particle-defective mutants (Saeed et al., 2021) ( Figure 1F , top panel; lane(s) 2 and 4 vs. 5, 6, and 9). Additionally, when we analyzed the HBc-AAA-AAASAS and HBc-AAA-AAAAAA mutants in which seven S/TP motifs were substituted with AP, we found that they were particle-defective ( Figure 1F , top panel; lane(s) 2 and 4 vs. 5-9), indicating that HBc S/TP motifs may play a role in core particle assembly. Coimmunoprecipitation with an anti-HBc antibody followed by immunoblotting with an anti-Pin1 antibody, and vice versa, did not detect particle-defective HBc proteins ( Figure 1F , sixth and ninth panels; lanes 4 vs. 5-9), further demonstrating that neither monomers nor dimers of HBc are binding partners of Pin1.

To characterize Pin1/core particle binding, we treated transfected cell lysates with 0.5 or 1 M sodium chloride and immediately subjected them to NAGE, followed by immunoblotting with anti-Pin1, anti-HBc, and anti-Myc antibodies ( Figure 2A ). As mentioned above, HBc levels in Pin1 WT and Myc-HBc WT cotransfected cells ( Figure 2A , fourth and fifth panels; lane 2 vs. 3) were slightly higher than Myc-HBc WT transfected cells, and core particle levels increased accordingly ( Figure 2A , second panel; lane 2 vs. 3). In the presence of high concentrations of sodium chloride, Pin1/core particle binding was significantly weakened ( Figure 2A , top panel; lanes 3 vs. 7 and 11), indicating that Pin1/core particle binding is a dynamic on-off interaction. Also, core particle levels increased, and the core particle in the presence of high concentrations of sodium chloride migrated more slowly on NAGE than the core particle in untreated lysates ( Figure 2A , second panel; lanes 2 and 3 vs. 6 and 7 vs. 10 and 11). We think that this is due to increased levels of unstable core particles in the presence of high concentrations of sodium chloride (Jung et al., 2012).

Furthermore, when we incubated transfected lysates at 65°C for 5, 15, and 30 min, the Pin/core particle interaction began to break down at 5 min, and the complex dissociated almost completely after 30 min, further demonstrating that the Pin1/core particle interaction is dynamic ( Figure 2B , top panel).

Since we used cotransfected lysates to detect the physical Pin1/core particle interaction, we next tried to examine the interaction in vitro. To do this, we separately transfected HBV WT and Pin1, and the respective lysates were mixed and incubated for 0, 12, 24, and 48 h at 37°C. Although HBV WT-transfected lysates may contain endogenous Pin1, which might interfere with exogenously added Pin1 in Pin1-transfected lysates, we detected slowly migrating Pin1 by NAGE conducted after 24 and 48 h of incubation; Pin1 detection was independent of transfection, indicating that Pin1 binds to core particles in vitro ( Figure 2C , top panel; lanes 7 and 8). To further show that this in vitro interaction depends on the phosphorylation status of the core particle, we treated HBV WT-transfected lysates for 1 h with CIAP. We found that we could hardly detect slowly migrating Pin1 ( Figure 2C , top panel; lanes 11 and 12), indicating that the physical Pin1/core particle interaction is phosphorylation-dependent. This dynamic on-off and in vitro interaction between the core particle and Pin1 ( Figures 2A-C ) indicates that Pin1 binds to the outside of the core particle. This result suggests that some CTDs within HBc, if not all, may face outward from the center of the core particle and serve as binding sites for Pin1.

Although Pin1 can bind to the outside of the core particle through a dynamic on-off interaction ( Figures 2A-C ), we could not exclude the possibility that Pin1 may also be incorporated into the core particle, like Par14 and Par17 (Saeed et al., 2021). To test this, lysates were mock-treated or heated at 65°C for 2 h. Consistent with the above results ( Figure 2B ), Pin1 dissociated almost completely from the core particle after heat treatment ( Figure 2D , top panel; lane 3 vs. 6). To allow the specific antibody to bind to the inside of the core particle on immunoblots, the PVDF membrane was treated with 0.2 N NaOH for 1 min to open up the core particle. It was then UV-crosslinked prior to immunoblotting (Jung et al., 2012; Jung et al., 2014; Saeed et al., 2021). Although heat treatment dissociated Pin1 from the outside of the core particle, it was also detected on the opened-up core particle ( Figure 2D , second and third panels; lanes 2 and 3 vs. 5 and 6), with no accompanying changes in core particle levels ( Figure 2D , fourth panel; lanes 2, 3, 5, and 6). This suggests that Pin1 is incorporated into the core particle. Thus, like Par14 and Par17 (Saeed et al., 2021), Pin1 can bind to the outside and the inside of the core particle.

Figure 1 shows the importance of the S/TP motifs within the HBc CTD for the Pin1 interaction ( Figures 1A, E ). Amino acid sequence alignment that HBc ¹²⁸TP within the NTD, and ¹⁶²TP, ¹⁶⁴SP, and ¹⁷²SP within the CTD, are conserved completely, and that HBc ⁴⁴SP and ⁴⁹SP within the NTD and ¹⁵⁷SP within the CTD are highly conserved, among 20 isolates of 10 genotypes of human HBc protein (including human HBV [adw subtype (HBc WT)]) in the National Center for Biotechnology Information (NCBI) database ( Figure 3A ). Consistent with the HBc CTD alignment results ( Figure 3A ), previous reports show that ¹⁵⁷S is less essential for HBV replication (¹⁵⁵S of ayw subtype) (Lan et al., 1999; Melegari et al., 2005; Lewellyn and Loeb, 2011b; Jung et al., 2014; Nishi et al., 2020), suggesting that the ¹⁶²TP, ¹⁶⁴SP, and ¹⁷²SP motifs may be more important for the Pin1/core particle interaction. Although HBc NTD seemed not to be important for Pin1/core particle interaction ( Figure 1E , top panel, lane 3), the completely conserved ¹²⁸TP motif resides within the irregular proline-rich loop 6 (¹²⁸ TP PAYRPPN¹³⁶) of HBc ( Figure 3A ) (Wynne et al., 1999), suggesting that this motif might play a role in core particle assembly.

HBc amino acid sequence alignment of a further seven isolates of mammalian hepadnaviruses revealed that HBc ⁴⁹SP and ¹²⁸TP within the NTD, and ¹⁶²TP, ¹⁶⁴SP, ¹⁷²SP, and ¹⁸⁰SP within the CTD (human HBc WT adw subtype numbering is used for convenience), are completely conserved, and that HBc ¹⁵⁷SP within the CTD is highly conserved ( Figure 3B ). The HBc of mammalian hepadnaviruses shows a similar S/TP motif pattern to that of 20 isolates of human HBc, with the exception of the first ⁴⁴SP motif in the NTD and the last ¹⁸⁰SP motif in the CTD. Unlike human HBc, ⁴⁴SP in the NTD is conserved within the HBc amino acid sequences of gorilla HBV (GoHBV), orangutan HBV (OrHBV), and chimpanzee HBV (ChHBV) but not in those of woolly monkey HBV (WMHBV), ground squirrel hepatitis virus (GSHV), woodchuck hepatitis virus (WHV), and bat HBV (batHBV) ( Figure 3B ). The mammalian hepadnavirus HBc ¹⁸⁰SP motif in the CTD is completely conserved, but human HBc harbors either 11 SR motifs or 9 SP motifs ( Figures 3A, B ). Avian hepadnaviruses HBc ¹⁶⁴TP in the NTD and ²⁴⁵SP and ²⁵⁹SP in the CTD are completely conserved, and HBc ²⁵⁷SP in the CTD is highly conserved, among five isolates of avian hepadnaviruses ( Figure 3C ). Taken together, we can speculate that the HBc S/TP motifs of orthohepadnaviruses (including human and mammalian hepadnaviruses) and avian hepadnaviruses may play a positive role in replication ( Figure 3 ; Supplementary Figure S1 ).

Figures 1 , 2 show that Pin1 interacts with the core particle through pS/TP motifs in the HBc CTD (Kann and Gerlich, 1994; Kau and Ting, 1998; Daub et al., 2002; Chen et al., 2011; Ludgate et al., 2012; Jung et al., 2014; Nishi et al., 2020). We also examined the CTD-deleted HBc (HBc-ΔCTD) (Jung et al., 2012) by NAGE and immunoblotting to confirm the importance of the HBc CTD for Pin1/core particle binding. As expected, Pin1 did not bind to the core particle of HBc-ΔCTD, although the levels of core particles and HBc were comparable with those in HBc WT ( Figure 4A , second and fourth panels; lane 5 vs. 6), confirming that HBc CTD is important for Pin1 binding.

To identify the S/TP motif(s) important for the Pin1/core particle interaction, we constructed HBc mutants that retain only a single phosphoacceptor S/TP site, and in which the corresponding S/TP motifs were substituted with AP motifs by PCR-generated site directed mutagenesis. To discriminate the respective mutants, we denote HBc-WT as bracketed with SST-STSSSS; for example, HBc-SST-AAAAAA (= ⁴⁴S⁴⁹S¹²⁸T-¹⁵⁷A¹⁶²A¹⁶⁴A¹⁷⁰A¹⁷²A¹⁷⁸A) has sextuple serine and threonine (S/T)-to-alanine (A) mutations at the CTD phosphoacceptor sites and as such is Pin1-binding defective ( Figure 1E , lane 4; Figures 4B, C , lane 7). Even though the ¹⁷⁰S and ¹⁷⁸S residues do not belong to an S/TP motif ( Figure 1A ), we included these two sites in the HBc WT and HBc S/T-to-A substituted mutants because both ¹⁷⁰S and ¹⁷⁸S are phosphorylated and take part in HBV replication (Jung et al., 2014); thus, these constructs are designated as HBc-WT (SST-STSSSS) and HBc-SST-AAAAAA, instead of HBc-WT (SST-STSS) and HBc-SST-AAAA. Pin1/core particle binding by HBc mutants harboring only a single phosphoacceptor S/TP site was examined by NAGE and immunoblotting ( Figure 4B ). The results showed that the Pin1/core particle interaction was severely impaired by the core particle with only a single S/TP motif on HBc, similar to the Pin binding-defective HBc-SST-AAAAAA mutant ( Figure 4B ; lanes 3−6 vs. lane 7). This suggests that a single S/TP motif on HBc is not sufficient for the Pin1/core particle interaction. In addition, the levels of the mutant HBc proteins and the core particles were less than those of the HBc WT and the core particle; also, the level of HBc sextuple mutant and the core particle was the least ( Figure 4B , second and third panels; lane(s) 2 vs. 3−6 vs. 7). This result was similar to that reported previously for a T160A and S162A double mutant (Nishi et al., 2020).

Next, we constructed single S/T-to-A mutants by PCR-generated site directed mutagenesis (HBc-S157A, HBc-T162A, HBc-S164A, and HBc-S172A); these mutants retain three S/TP motifs in the CTD. Surprisingly, NAGE and immunoblotting showed that the Pin1/core particle interaction was also severely impaired in the HBc-T162A, HBc-S164A, and HBc-S172A mutants but not the HBc-S157A mutant, demonstrating that the ¹⁶²TP, ¹⁶⁴SP, and ¹⁷²SP motifs are important ( Figure 4C , top panel; lane 2 vs. lanes 4−6), whereas the ¹⁵⁷SP motif is less important ( Figure 4C , top panel; lane 2 vs. 3), for the Pin1/core particle interaction. These data are in accordance with amino acid sequence alignment results ( Figure 3 ). Unlike a previous report that emphasized the importance of ¹⁶²TP and ¹⁶⁴SP (¹⁶⁰TP and ¹⁶²SP) for Pin1/HBc binding (Nishi et al., 2020), we found that the ¹⁶²TP, ¹⁶⁴SP, and ¹⁷²SP motifs are important ( Figure 4C ). Although the amount of HBc protein by these single S/TP motif mutants was comparable ( Figure 4C , third panel; lane 2 vs. 3−6), the amount of core particle by HBc-T162A, HBc-S164A, and HBc-S172A mutants was less, although not as significantly less as in the case of the HBc sextuple mutant ( Figure 4C , second panel; lane 2 vs. 4−6 vs. 7).

Since we constructed HBc mutants harboring S/T-to-A substitutions at the S/TP motifs, the newly generated AP motifs may serve as binding sites for other proline-directed PPIases such as cyclophilins or FKBPs. To examine whether the AP motifs serve as binding sites for cyclophilins or FKBPs, we performed NAGE followed by immunoblotting with anti-cyclophilin A or anti-FKBP12 antibodies ( Supplementary Figure S2 ). The result showed that the HBc-SST-AAAAAA mutant does not bind to cyclophilin A or FKBP12.

Although we already demonstrated that HBc NTD S/TP motifs are not important for the Pin1/core particle interaction ( Figure 1E , lane 3 vs. 4), we further investigated the HBc NTD single and triple S/T-to-A substitution mutants by NAGE and immunoblotting ( Figure 5A ). We included a particle-defective HBc-Y132A mutant as a negative control. With the exception of the core particles from the Myc-HBc-T128A mutant, the amount of core particles from HBc NTD mutants was comparable with that from HBc-WT ( Figure 5A , second panel; lane(s) 2−4, 6 vs. 5). Although the core particle level of Myc-HBc-T128A was significantly lower than that of Myc-HBc WT and other HBc NTD mutants, binding of Pin1 binding to the core particles was unaffected ( Figure 5A , bottom graph; lanes 2 vs. 3−6). Consistent with this, a previous study shows that the completely conserved ¹²⁸TP motif ( Figure 3A ) in an irregular proline-rich loop (Wynne et al., 1999) plays a role in core particle assembly.

To scrutinize the roles of the HBc NTD S/TP motifs in core particle assembly and/or structural stability, we subjected Myc-HBc-WT- and Myc-HBc-NTD S/TP motif mutant-transfected HEK293T cell lysates to repeated freeze/thaw cycles ( Figure 5B ). Expression of HBc monomers was comparable under reducing conditions ( Figure 5B , right top, right third, right fifth panels; lane(s) 8 vs. 9−12). As shown above ( Figure 5A , second panel), core particle levels from fresh lysates were comparable (the exception was the core particle of Myc-HBc-T128A) on NAGE and immunoblot analysis ( Figure 5B , left third panel). Repeated freezing and thawing led to almost complete disassociation of Myc-HBc-S49A and Myc-HBc-AAA-STSSSS core particles ( Figure 5B , left sixth and ninth panels; lanes 2, 3 vs. 4, 6). Similarly, dimers produced by the Myc-HBc-S49A and Myc-HBc-AAA-STSSSS mutants disappeared after repeated freezing and thawing under nonreducing conditions ( Figure 5B , seventh panel; lanes 4 and 6). As shown for the Myc-HBc-S49A and Myc-HBc-AAA-STSSSS mutants, the HBc NTD S/TP motifs may play a role in HBV core particle stability in the absence of Pin1-binding activity. In the case of Myc-HBc-T128A ( Figure 5B , left third panel; lane(s) 5 vs. 3, 4 and 6), the levels of core particle and dimer were not affected by repeated freezing and thawing ( Figure 5B , lane 5), further supporting the notion that the ¹²⁸TP motif plays a role in core particle assembly. When we cotransfected with HBc-deficient mutant HBV plus HBc WT or HBc NTD S/TP motif mutants, HBc-deficient mutant HBV replication was not rescued well when trans-complemented with HBc NTD S/TP motif mutants such as Myc-HBc-S49A, Myc-HBc-T128A, and Myc-HBc-AAA-STSSSS, especially by Myc-HBc-T128A ( Supplementary Figure S1 ). In accordance with above notions, HBc NTD S/TP motifs are important for HBV replication.

The Pin1 residues that are important for substrate binding and PPIase activity have been identified (Lu and Hunter, 2014; Zhou and Lu, 2016). Indeed, ¹⁶S and ⁷¹S act as phosphoacceptors that affect substrate binding and PPIase activity, respectively; tryptophan 34 (³⁴Trp: ³⁴W) required for substrate binding and cysteine 113 (¹¹³Cys: ¹¹³C) required for PPIase activity are indicated in Figure 6A (Rippmann et al., 2000; Watashi et al., 2008; Lu and Hunter, 2014). To identify the Pin1 residues involved in HBV core particle binding, we constructed a substrate binding-positive dephosphorylated mimetic S16A mutant, a substrate binding-negative phosphorylated mimetic ¹⁶S-to-Glu (S16E) mutant, a substrate binding-deficient W34A mutant, a PPIase-active S71A mutant, and a PPIase-inactive C113A mutant ( Figure 6A ). We then examined their ability to bind core particles by conducting NAGE and immunoblotting. As expected, the substrate binding-negative S16E and W34A mutants showed impaired core particle binding, whereas the S16A, S71A, and C113A mutants bound to core particles as efficiently as or more efficiently than Pin1 WT ( Figure 6B ; lane(s) 4 vs. 6, 7 vs. 5, 8, 9). This demonstrates that the substrate binding ¹⁶S and ³⁴W residues of Pin1 are important for the Pin1/core particle interaction. The results for the PPIase-inactive C113A mutant show that PPIase activity is not required for the Pin1/core particle interaction.

HCC and other cancer cells express high levels of Par14/Par17 protein (Saeed et al., 2019) ( Figure 6C , fourth panel; lane 1) and Par14/Par17 bind to both HBc and the core particle to increase their stability (Saeed et al., 2021); in addition, Par14 might compensate for a lack of Pin1 (Uchida et al., 2003). Therefore, we hypothesized that the effect of Pin1 might be less than optimal in the presence of Par14/Par17. To test this, we transduced HEK293 cells with lentiviral PIN4 shRNA to generate PIN4 KD HEK293T cells (pLK0.1-shPIN4#1) (Table) (Saeed et al., 2019). Since Pin1 proteins are involved in the stability of many substrate proteins (Liou et al., 2011; Lu and Hunter, 2014; Zhou and Lu, 2016), we investigated the levels of HBc and core particles in cycloheximide-treated PIN4 KD HEK293T cells transfected with HBV WT plus mock-, Pin1 WT-, or Pin1 W34A. At 12 h post-transfection, cells were treated with 50 µg/ml of cycloheximide to inhibit protein synthesis ( Figure 6C ; lanes 8−19). In PIN4 KD cells ( Figure 6C , fourth panel, lanes 2−19), we detected endogenous Pin1 (third panel; lanes 1−3, 8−10), overexpression of Pin1 WT (third panel; lanes 4, 5, 12−15), and overexpression of the Pin1 W34A mutant (third panel; lanes 6, 7, 16−19). At the same time, we compared HBc protein levels ( Figure 6C , second panel). At 12 h after cycloheximide treatment (i.e., 24 h post-transfection), there was no significant change in HBc protein levels ( Figure 6C , second panel; lanes 9, 13, 17; Figure 6D , lower panel), confirming that HBc protein is stable. At 24 and 48 h postcycloheximide treatment (i.e., at 36 and 60 h post-transfection), HBc protein levels fell significantly; however, HBc levels in the Pin1 WT-cotransfected cells decreased to a lesser extent than those in mock- or Pin1 W34A mutant-cotransfected cells ( Figure 6C , second panel; lanes 10, 11 vs. 14, 15 vs. 18, 19; Figure 6D , lower panel). Accordingly, in the absence of a new supply of HBc protein, core particle levels in Pin1 WT-overexpressing cells decreased to a lesser extent than those in mock- or Pin1 W34A mutant-cotransfected cells ( Figure 6C , top panel; lanes 12−15 vs. 8−11 vs. 16−19; Figure 6D , upper panel), demonstrating that Pin1 WT, but not the substrate binding negative Pin1 W34A mutant, increases core particle stability, possibly through the Pin1/core particle interaction. Since Pin1 WT alone increased the stability of the core particle, supposedly through the Pin1/core particle interaction, in the absence of PIN4 (i.e., in PIN4 KD cells). Therefore, increasing the stability of the core particles increases the level of HBc proteins ( Figures 6C, D ).

Alternatively, we can hypothesize that Pin1 WT might facilitate core particle assembly. Since Pin1 does not bind to monomeric or dimeric HBc proteins ( Figure 1E ), Pin1 WT may facilitate core particle assembly via a partial intermediate form. This hypothesis is supported by the data in Figure 2D .

To further examine the function of Pin1 during HBV replication, we used parvulin inhibitors PiB (Diethyl-1,3,6,8-tetrahydro-1,3,6,8-tetraoxobenzo[lmn][3,8]phenanthroline-2,7-diacetic acid; 20 µM in DMSO; CAS 64005-90-9, Calbiochem) ( Figure 7A ) and Juglone (5-hydroxy-1,4-naphthoquinone, 20µM in ethanol, AG17724, Sigma-Aldrich) ( Figure 7B ). At 48 h post-transfection, 1.3mer HBV WT (adw) transfected HepG2 cells were treated with these inhibitors for 24 h. PiB is a competitive reversible inhibitor of parvulin PPIases such as Pin1 and Par14/Par17 (Uchida et al., 2003). Juglone, a natural compound derived from walnut trees, is a competitive irreversible inhibitor that inhibits and/or inactivates members of the parvulin PPIase family; it does not affect other PPIase families (Hennig et al., 1998). Consistent with our previous report (Saeed et al., 2019), we found that HBc protein expression, core particle assembly, and HBV replicative intermediate (RI) DNA synthesis fell significantly after 24 h treatment with PiB ( Figure 7A ; lane 2 vs. 3) and Juglone ( Figure 7B ; lane 2 vs. 3). Unlike our previous study, in which we treated cells with PiB and Juglone for 72 h (Saeed et al., 2019), we exposed cells to PiB and Juglone for 24 h at 48 h post-transfection. The results of these experiments indicate that parvulin proteins affect HBV replication in multiple ways (Saeed et al., 2019; Saeed et al., 2021; Kim, 2023).

To demonstrate that Pin1 affects HBV replication, we generated PIN1 KD HepG2 cells using a lentiviral shRNA system (Table). Briefly, HepG2 cells were transduced with lentiviral control shRNA (pLK0.1-shControl) or PIN1 shRNAs (pLK0.1-shPIN1#1−4). Lentivirus-transduced, puromycin-selected PIN1 KD HepG2 cells were then transfected with 1.3mer ayw HBV WT ( Figure 7C ). HBc protein expression, core particle assembly, and HBV RI DNA synthesis in PIN1 KD HepG2 cells ( Figure 7C , top panel; lane 3 vs. 4−7), but not in shPIN1#3-transduced cells, decreased significantly ( Figure 7C ; lanes 4−7), suggesting that Pin1 may support efficient HBV replication.

To examine the effects of Pin1 on HBV infection, we transduced HepG2-hNTCP-C9 cells with lentiviral control shRNA or PIN1 shRNAs (shPIN1-#1 or #4). Following HBV infection, HBc protein expression, core particle assembly, and HBV RI DNA synthesis decreased significantly upon PIN1 KD ( Figure 7D , third, eighth, and bottom panels; lane 3 vs. 4 and 5); however, the levels of HBV pgRNA, subgenomic RNAs, and cccDNA were unaffected ( Figure 7D , fifth and sixth panels; lane 3 vs. 4 and 5). This result clearly demonstrates that Pin1 affects HBV replication after HBV RNA synthesis.

Since HBV replication was downregulated by parvulin inhibitors and PIN1 KD ( Figure 7 ), we further investigated the effects of Pin1 overexpression on the HBV life cycle. When HepG2 cells were cotransfected with Pin1 and 1.3mer HBV WT (adw), HBc protein expression, core particle assembly, and HBV RI DNA synthesis increased significantly ( Figure 8A , top, second, sixth, bottom panels; lane 2 vs. 3). When Pin1 was transiently transfected into HBV-replicating stable HepG2.2.15 cells, HBV RI DNA synthesis was upregulated ( Figure 8B ; lane 4). Even though pcDNA3-transfected HepG2.2.15 cells (usual mock to adjust transfected DNA amount) showed higher expression of HBV RI DNA than no DNA transfected cells ( Figure 8B , bottom panel, lane 2 vs. 3), core particle assembly and HBV RI DNA synthesis were upregulated to a greater extent in Pin1-overexpressing cells ( Figure 8B , second, sixth, bottom panels; lane 2 vs. 3 vs. 4). Unlike the above results ( Figures 1C , 2A, B , 8A ), HBc protein expression in HepG2.2.15 cells was not significantly increased by Pin1 overexpression. Consistent with the PIN1 KD experiment ( Figure 7D ), Pin1 had no significant effect on HBV transcription ( Figures 8A, B , fourth panel).

Since HBx interacts with Pin1 (Pang et al., 2007) and Par14/Par17 (Saeed et al., 2019), and Par14/Par17 upregulate HBV replication in an HBx-dependent manner (Saeed et al., 2019), it may be that Pin1 overexpression also upregulates HBV replication in an HBx-dependent manner. To examine this, HepG2 cells were cotransfected with Pin1 plus CMV-HBV WT (adw) or CMV-HBx-deficient HBV (adw) (Yoon et al., 2011). Consistent with previous reports (Yoon et al., 2011; Piracha et al., 2018; Kim et al., 2022), we found that replication of HBx-deficient HBV was lower than that of HBV WT ( Figure 8C ; lane 2 vs. 4). Unlike Par14/Par17 (Saeed et al., 2019), overexpression of Pin1 upregulated HBV replication in the absence of HBx ( Figure 8C ; lane 4 vs. 5), demonstrating that Pin1-mediated effects are independent of HBx.

To elucidate the effects of Pin1 overexpression in an HBV infection system, we transduced HepG2-hNTCP-C9 cells with lentiviral vectors pCDH or pCDH-PIN1, as described previously (Kim et al., 2022). Following HBV infection of HepG2-hNTCP-C9 cells that had been transduced with the vectors or PIN1, we found that Pin1 overexpression increased core particle and HBV RI DNA synthesis significantly ( Figure 8D , eighth and bottom panels; lane 3 vs. 4), whereas HBc protein levels, cccDNA formation, and HBV pgRNA and subgenomic RNA levels did not change significantly ( Figure 8D , third, fifth, and sixth panels; lane 3 vs. 4). Consistent with the data for Pin1-overexpressing HepG2.2.15 cells ( Figure 8B , second panel), HBc levels in HBV-infected HepG2-hNTCP-C9 cells did not increase ( Figure 8D , third panel). It is extraordinary that core particle levels increased without concomitant increases in pgRNA and HBc protein levels. Usually, protein levels increase along with RNA levels. Since Pin1 cannot interact with HBc, we speculate that Pin1 stabilizes core particles, thereby increasing core particle levels. Consequently, the increase in HBc levels observed on SDS-PAGE may be due to increased stability of the core particle.

Immature core particles are heavily phosphorylated by series of kinases, including proline-directed serine/threonine kinases, CDK2, and/or a CDK2-like kinase, which phosphorylate S/TP motifs (Ludgate et al., 2012). Also, mature core particles are dephosphorylated by as-yet-unidentified phosphatases, although protein phosphatase 2A (PP2A) (Ning et al., 2017) or PDP2 (Nishi et al., 2020) have been suggested. Since we demonstrated that Pin1 interacts physically with core particles but not with HBc monomers and dimers ( Figures 1E, F ), we hypothesized that Pin1 may function in this dephosphorylation process by binding to pS/TP on core particles, thereby catalyzing isomerization and conformational changes that facilitate dephosphorylation of these pS/TP motifs; it may then detach from the mature core particles prior to budding or recycling ( Figure 9 ).

To test this hypothesis, we examined binding of Pin1 to the core particle at different stages of the HBV life cycle. We examined this in stable Pin1-overexpressing Huh7 cells transiently transfected with a P-deficient mutant harboring empty core particles, a reverse transcription-defective RT-YMHA mutant, the oligomeric minus-strand translocation-defective priming-deficient TP-Y65F mutant (Kim et al., 2004), and HBV WT ( Figure 10A ). When we compared Pin1 binding to immature core particles from the RT-YMHA mutant- and TP-Y65F mutant-transfected cells (Kim et al., 2004) with that to core particles from HBV WT-transfected cells, we found that Pin1 bound more efficiently to immature core particles from RT-YMHA and TP-Y65F mutants than to those from HBV WT ( Figure 10A , top and bottom panels; lane 5 vs. 3 and 4). Core particles from the P-deficient mutant, supposedly harboring empty core particles, bound the most Pin1 ( Figure 10A , top and bottom panels; lane 2 vs. 3−5), and core particles from HBV WT bound the least Pin1 ( Figure 10A , top and bottom panels; lane 5 vs. 2−4). More Pin1 proteins seem to bind to core particles from the RT-YMHA mutant than from the TP-Y65F mutant; however, the difference was insignificant ( Figure 10A , top and bottom panels; lane 3 vs. 4). Of note, in HBV WT-transfected cells, core particles at multiple replicative stages (immature through the mature) were present. As expected, in situ DNA hybridization to detect minus-strand DNA in core particles showed that minus-strand DNA was present only in HBV WT core particles ( Figure 10A , third panel; lane 5). As expected, in situ nucleic acid hybridization detected encapsidated pgRNAs from RT-YMHA and TP-Y65F mutants weakly but detected encapsidated pgRNAs and RI DNAs from HBV WT strongly ( Figure 10A , fourth panel; lane(s) 3, 4 vs. 5). This indicates that our Pin1/core particle interaction model based on HBV replication stage is plausible ( Figure 9 ).

To further extend this model, we examined culture supernatants from Pin1 WT- and control empty vector-transduced HepAD38 cells by NAGE, followed by immunoblotting with anti-PreS1, anti-Pin1, and anti-HBc antibodies and in situ DNA hybridization ( Figure 10B ). More virions and subviral particles containing HBs (LHBs, MHBs, and SHBs) were released, thus the more virion DNA and PreS1 were detected, in Pin1 WT stable cells than control cells ( Figure 10B , top and bottom panels; lane 2 vs. 3). The anti-Pin1 antibody reacted strongly with virions and subviral particles from Pin1 stable cells ( Figure 10B , second panel; lane 3), indicating that HBs in virions and subviral particles may also be binding substrates of Pin1. Consistent with this, amino acid sequence analysis of HBs from ten genotypes of human HBV revealed that HBs proteins (including LHBs, MHBs, and SHBs) harbor seven conserved S/TP motifs (according to the amino acid sequence of LHBs in the NCBI database) ( Supplementary Figure S3 ). Consistent with the data in Figure 10B , Pin1-overexpressing, HBV-infected HepG2-hNTCP-C9 cells showed increased secretion of virions and subviral particles, as demonstrated by increased levels of extracellular HBs and virion DNA ( Figure 10C ; lane 3 vs. 4). By contrast, fewer extracellular HBs and virion DNA were detected in HBV-infected PIN1 KD cells ( Figure 10D ; lane 3 vs. 4 and 5). Pin1 upregulates HBV RI DNA synthesis as well as secretion of virions and subviral particles. Taken together, we propose that the Pin1/core particle interaction facilitates assembly of HBV core particles, stabilizes them, helps them to mature during the intracellular HBV replicative stage, and facilitates virion secretion ( Figure 9 ).