Porcine Circovirus Type 2 Hijacks Host IPO5 to Sustain the Intracytoplasmic Stability of Its Capsid Protein

ABSTRACT

Nuclear entrance and stability of porcine circovirus type 2 (PCV2), the smallest virus in mammals, are crucial for its infection and replication. However, the mechanisms are not fully understood. Here, we found that the PCV2 virion maintains self-stability via the host importin 5 (IPO5) during infection. Coimmunoprecipitation combined with mass spectrometry and glutathione S-transferase pulldown assays showed that the capsid protein (Cap) of PCV2 binds directly to IPO5. Fine identification demonstrated that the N-terminal residue arginine²⁴ of Cap is the most critical to efficient binding to the proline⁷⁰⁹ residue of IPO5. Detection of replication ability further showed that IPO5 supports PCV2 replication by promoting the nuclear import of incoming PCV2 virions. Knockdown of IPO5 delayed the nuclear transport of incoming PCV2 virions and significantly decreased the intracellular levels of overexpressed PCV2 Cap, which was reversed by treatment with a proteasome inhibitor or by rescuing IPO5 expression. Cycloheximide treatment showed that IPO5 increases the stability of the PCV2 Cap protein. Taken together, our findings demonstrated that during infection, IPO5 facilitates PCV2 replication by directly binding to the nuclear localization signal of Cap to block proteasome degradation.

IMPORTANCE Circovirus is the smallest virus to cause immune suppression in pigs. The capsid protein (Cap) is the only viral structural protein that is closely related to viral infection. The nuclear entry and stability of Cap are necessary for PCV2 replication. However, the molecular mechanism maintaining the stability of Cap during nuclear trafficking of PCV2 is unknown. Here, we report that IPO5 aggregates within the nuclear periphery and combines with incoming PCV2 capsids to promote their nuclear entry. Concurrently, IPO5 inhibits the degradation of newly synthesized Cap protein, which facilitates the synthesis of virus proteins and virus replication. These findings highlight a mechanism whereby IPO5 plays a dual role in PCV2 infection, which not only enriches our understanding of the virus replication cycle but also lays the foundation for the subsequent development of antiviral drugs.

INTRODUCTION

Porcine circovirus (PCV), the smallest known virus that can replicate autonomously in mammalian cells, can be subdivided into the following four genotypes: porcine circovirus type 1 (PCV1), PCV2, PCV3, and PCV4 (1–4). Over the past 30 years, five viral proteins of PCV2 have been identified (5–10). The capsid (Cap) protein, as the unique viral structural protein, is not only the main viral immunogen but also the main packaging protein of the PCV genome (11, 12). Therefore, an increased understanding of the functions and mechanisms of Cap could enrich our understanding of the virus replication cycle and provide new insights into therapeutic and prophylactic interventions for PCV2 infection.

The capsid of PCV2 comprises an icosahedral T = 1 structure containing 60 Cap protein molecules arranged in 12 pentamer cluster units (12). The Cap protein, as the only component of the virus shell, is potentially involved in the PCV2 replication cycle via interactions with host proteins (13). Viral capsids bind to cell surface receptors heparan sulfate and chondroitin sulfate B and are internalized either through clathrin-mediated endocytosis or through an actin/small-GTPase-dependent pathway (14–17). Subsequently, PCV2 virions escape from endosomes into the cytosol (18). Afterward, PCV2 particles recruit microtubule-associated molecular motors to affect their nuclear trafficking (19). After entering the nucleus, PCV2 genomic DNA undergoes ring-rolling replication, transcription, and translation of viral proteins with the assistance of host cell enzymes (20). Unfortunately, although the distribution of the Cap protein is involved in the dynamic changes during PCV2 infection (21), the mechanism underlying PCV2 capsid and Cap protein nuclear translocation remains poorly defined.

To gain access to the nucleus, viral proteins have evolved to attach to appropriate components of the nuclear transport machinery to overcome the barrier of the nuclear membrane. To date, classical and nonclassical nuclear import pathways have been discovered to assist the nuclear transport of proteins. The classical nuclear pathway comprises karyopherin β1 (importin β1) interacting either directly with the cargo or via an adaptor protein, known as karyopherin α (importin α), to facilitate nuclear import of proteins containing a nuclear localization signal (NLS). The nonclassical nuclear import pathway includes diverse patterns of nuclear transportation (22–24). Importin β homologs, like importin β2 and importin β3 (importin 5 [IPO5]), make contact with the cargo protein to form cargo/importin β complexes that execute nuclear import (25–29). In recent years, researchers have discovered that nonclassical pathways assist the nuclear import of viral proteins. Previous studies have shown that the NLS, comprising 41 amino acid residues at the N terminus of the PCV2 Cap protein, is important for its nuclear import (30). However, which pathways are exploited by PCV2 capsids and Cap and which nuclear transport receptors are utilized by PCV2 remain unknown. Understanding the mechanism of the nucleocytoplasmic transport system during viral infection is especially important for the PCV2 life cycle and will advance the development of antiviral therapies to control viral diseases.

In the present study, we found that both PCV2 capsids and Cap could bind to IPO5 during PCV2 infection and that the nuclear entry of viral particles was mediated by IPO5. Although an IPO5 deficiency in cells had no influence on the nuclear transportation of newly synthesized viral Cap protein, IPO5 could inhibit Cap protein degradation via the proteasome pathway to accelerate the replication and proliferation of the virus at the later stage of infection. In summary, IPO5 plays an important role in the process of PCV2 replication.

RESULTS

IPO5 is necessary for virus progeny production during PCV2 replication.

Identifying the interaction between viral proteins and potential host proteins will provide useful information to accelerate the investigation of the underlying mechanism of PCV2 infection and pathogenesis. Therefore, we performed immunoprecipitation with anti-Cap antibodies in PK15 cells infected with PCV2 for 48 h, followed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and silver staining (Fig. 1A). Subsequently, peptide sequencing and liquid chromatography-mass spectrometry (LC-MS) identified IPO5 as a protein that physically associated with PCV2 Cap (Fig. 1B). To assess the function of IPO5 during PCV2 infection, a short hairpin RNA (shRNA) specific for IPO5 (shIPO5) and a nontargeting control shRNA (shCON) were transferred via lentivirus-mediated shRNA to produce cells stably expressing green fluorescent protein (GFP)-shRNAs. Western blot assays showed that endogenous IPO5 expression was significantly decreased in the shIPO5-transfected PK15 cells compared with that in the shCON-transfected cells (P < 0.01) (Fig. 1C). In addition, cell counting kit-8 (CCK-8) assays showed that the viability and proliferation of the shIPO5-transfected cells were not affected significantly (P > 0.05) (Fig. 1D). To define the physiological role of IPO5 in PCV2 infection, IPO5-silenced cells were inoculated with PCV2 at a multiplicity of infection (MOI) of 1, and the RNA and protein levels of the PCV2 Cap gene were measured. The RNA and protein levels of Cap decreased significantly in IPO5-silenced cells compared with the levels in PCV2-infected shCON cells (P < 0.01) (Fig. 1E and F), and the replicative ability of PCV2 was almost 10-fold lower than that in the control cells over the course of the infection (Fig. 1G). Moreover, the rescue of IPO5 in IPO5-deficient PK15 cells and IPO5-overexpressed PK15 cells reversed the PCV2 Cap protein levels (Fig. 1H and I), and the replication ability of PCV2 in IPO5-overexpressed PK15 cells was significantly increased compared with that in control cells (Fig. 1J). Collectively, these results demonstrated that IPO5 promoted the production of PCV2 progeny virions.

FIG 1.

IPO5 expression increases PCV2 Cap expression and facilitates progeny virus replication. (A and B) PK15 cells were infected with PCV2 at an MOI of 1. After 48 h, cell lysates were immunoprecipitated (Ip) with anti-Cap antibody or IgG, and then immune complexes were analyzed by silver staining (A) and the MS/MS spectrum of IPO5 in the Cap immune complex was detected (B). PEP, posterior error probability. (C) IPO5 expression in shIPO5-transfected PK15 cells was detected using Western blotting. (D) The viability of PK15 cells stably expressing shIPO5 was analyzed using a CCK-8 assay. (E and F) IPO5-silenced cells were infected with PCV2 at an MOI of 1 for the indicated times. Cap, β-actin, and IPO5 levels were analyzed using Western blotting with their corresponding antibodies (E), and Cap mRNA transcripts were determined using real-time qPCR and analyzed using the threshold cycle relative-quantification method (2^−ΔΔCT) (F). GAPDH, glyceraldehyde-3-phosphate dehydrogenase. (G) PCV2 titers in IPO5-silenced PK15 cells. IPO5-silenced PK15 cells were infected with PCV2 at an MOI of 1, 10, or 25. Virus stocks were harvested from 24 to 72 hpi at 12-h intervals, and titers were determined. The results are presented as the mean TCID₅₀ ± SD (n = 3). Three independent biological experiments were performed. (H and I) IPO5-silenced PK15 cells (H) or PK15 cells (I) were transfected with empty vector or Flag-mIPO5 (a novel IPO5 non-sense mutant) or Flag-IPO5 for 18 h and then infected with PCV2 at an MOI of 1 for the indicated times. The cells were then collected and subjected to immunoblotting (IB) with anti-Cap MAb, anti-β-actin MAb, and anti-Flag MAb. (J) The samples from the experiment whose results are shown in panel I were used to measure PCV2 replication by determining TCID₅₀ values. Three independent biological experiments were performed. Data are presented as the mean values ± standard deviations (SD) from three independent biological experiments.

IPO5 facilitates the nuclear import of incoming PCV2 virions.

To explore the critical role of IPO5 in the PCV2 life cycle, whether IPO5 was required for the attachment, internalization, nuclear targeting transportation, or nuclear import of viral capsids was investigated. IPO5-silenced cells were incubated with PCV2 at a high MOI at 4°C or 37°C for 2 h. Western blotting showed that the level of Cap in lysates of IPO5 knockdown cells had no significant difference from that of the shCON cells (Fig. 2A), indicating that IPO5 does not affect the attachment and internalization of PCV2 virions. Confocal microscopy analysis also revealed that PCV2 virions moved to the perinuclear region in both IPO5-silenced and shCON-transfected cells at 12 h postinfection (Fig. 2B), indicating that IPO5 did not affect nuclear-targeting transportation of PCV2 virions. To further explore whether IPO5 is involved in the nuclear import of PCV2 virions, a cell fractionation experiment was conducted. The IPO5-silenced cells and empty-retrovirus-transduced control cells were inoculated with PCV2 at an MOI of 50 for 12, 14, and 16 h. The Cap level in the nucleus of IPO5-deficient cells was always less than that in the control cells at each time point of PCV2 infection (P < 0.05) (Fig. 2C). Correspondingly, the Cap levels in the cytoplasm of IPO5-silenced cells were always higher than those in the control cells. Similarly, confocal microscopy analysis revealed that most of the incoming PCV2 virions had delayed migration into the nucleus in IPO5-silenced PK15 cells compared with that in the control cells (Fig. 2D). Six levels (level 0 to level 5) were defined to evaluate the corresponding degrees of virion entry into the nucleus (Fig. 2E). Based on these criteria, 109 IPO5-silenced cells and 81 control cells were counted at 14 h after PCV2 infection. Positive cells with PCV2 in the nucleus (above level 3) represented 54.33% (44/81) of cells in the control group and 31.19% (34/119) of cells in the IPO5-deficient group. In contrast, positive cells with PCV2 outside the nucleus (below level 3) were 45.67% (34/81) of cells in the control group and 68.81% (75/109) of cells in the IPO5-silenced group. These data demonstrated that IPO5 facilitated the nuclear import of newly invading PCV2 capsids.

FIG 2.

IPO5 is required for nuclear import of incoming viral capsids, but not for cell binding and cytoplasmic trafficking to the nuclear pore. (A) PCV2 virions attached to IPO5-silenced PK15 cell membranes. IPO5-silenced PK15 cells and control cells were exposed to PCV2 at an MOI of 25 or 50 at 4°C or 37°C for 2 h. The cells were then collected and subjected to immunoblotting with anti-Cap antibodies, anti-β-actin antibodies, and anti-IPO5 antibodies. (B) Intracellular migration of PCV2 virions in IPO5-silenced PK15 cells. IPO5-silenced PK15 cells and control cells were infected with PCV2 at an MOI of 50 for 12 h, incubated with mouse anti-Cap IgG overnight at 4°C, and then stained with Alexa Fluor 546 anti-mouse antibody. Green fluorescence represents cells expressing GFP-shRNA. The subcellular localization of viral particles was examined using confocal microscopy. Viral particles were inverted into gray-scale images to make them easier to view. The white arrows show the positions of PCV2 virus capsids. (C) Cap protein in the cytoplasm and nucleus of IPO5-silenced PK15 cells. IPO5-silenced PK15 cells and control PK15 cells were infected with PCV2 at an MOI of 50 for 12, 14, and 16 h. Cells were then collected and subjected to separation of subcellular components and detection of Cap. Data are presented as mean values ± SD from three independent biological experiments. (D) Intracellular localization of PCV2 virions. IPO5-silenced PK15 cells and control cells were infected with PCV2 at an MOI of 50 for 14 h, and then the cells were fixed and stained for capsids (inverted into gray scale from red) and nuclei (blue). *, signal observed inside the nucleus; #, signal observed outside the nucleus. (E) Five levels of nuclear entrance of PCV2 capsids were defined. IPO5-silenced cells and control cells infected with PCV2 were counted according to this rule.

PCV2 Cap binds directly to IPO5.

To further detect the relationship between Cap and IPO5 in the process of Cap nuclear shuttling, Flag-IPO5 and Myc-Cap expression plasmids were cotransfected into 293T cells, and immunoprecipitation was conducted using FLAG beads or purified anti-Myc monoclonal antibody (MAb). As shown by the results in Fig. 3A and B, IPO5 interacted with the Cap protein. Furthermore, glutathione S-transferase (GST) affinity isolation revealed a direct interaction between IPO5 and Cap (Fig. 3C). Moreover, to evaluate whether endogenous IPO5 interacted with PCV2 Cap protein, PK15 cells were infected with PCV2 at an MOI of 10 for 24 h, and the cell lysates were immunoprecipitated with purified anti-Cap IgG. The data in Fig. 3D showed that IPO5 was only detected in the precipitates of PCV2-infected cells, and not in those of mock-infected cells. Furthermore, the subcellular localization of IPO5 with PCV2 was observed in infected cells. Confocal microscopy revealed that the endogenous IPO5 protein was colocalized at the nuclear periphery with the PCV2 capsids (Fig. 3E). To identify the critical amino acid residues responsible for the binding of IPO5 to Cap protein, the HDOCK server (http://hdock.phys.hust.edu.cn/) was used to simulate the Cap and IPO5 binding interfaces. The prediction results displayed 20 amino acid residues in IPO5 that might participate in the interaction (data not shown). Furthermore, coimmunoprecipitation (co-IP) assays showed that the IPO5 P709A mutant (bearing a change of P to A at position 709) completely lost the ability to bind to Cap and that the IPO5 F715A and M745A mutants had weakened binding ability to Cap, indicating that the residue proline⁷⁰⁹ of IPO5 is the most critical to the interaction with Cap (Fig. 3F). Taken together, these results clearly demonstrated that IPO5 could bind directly to the Cap protein of PCV2.

FIG 3.

IPO5 binds directly to the PCV2 Cap protein. (A) Plasmids expressing Flag-tagged IPO5 were cotransfected with those expressing Myc-tagged Cap protein into 293T cells for 36 h. Cell lysates were immunoprecipitated (IP) using anti-Flag MAb, and the Cap protein was detected using anti-Myc PAb. (B) In reciprocal immunoprecipitation assays, cell lysates were immunoprecipitated with anti-Myc MAb and Flag-IPO5 was detected using anti-Flag MAb. (C) GST-importin β3 protein was immobilized on glutathione-Sepharose beads and incubated with His-Sumo-Cap. Proteins in the GST pulldown assays were examined using immunoblotting with anti-His antibodies and anti-GST antibodies. The input proteins were determined using Coomassie blue staining. (D) PK15 cells were infected with PCV2 at an MOI of 10 for 24 h. The cell lysates were immunoprecipitated using mouse anti-Cap IgG or anti-IgG antibody. Immunoblotting was then performed to determine the presence of IPO5 in the Cap immunoprecipitates. (E) PK15 cells were infected with PCV2 at an MOI of 50 for 14 h, and the cells were immunostained using mouse anti-Cap IgG and rabbit anti-IPO5 antibody and then stained with FITC anti-rabbit antibody and Alexa Fluor 546 anti-mouse antibody, respectively. The colocalization of capsids with IPO5 was visualized using a superresolution confocal microscope. The intensities of the three channels along the white lines are shown in graphs a and b to the left. (F) Various mutants of IPO5 were cotransfected with Myc-Cap expression plasmid into 293T cells for 36 h, and cell lysates were collected, immunoprecipitated with an anti-Flag MAb, and immunoblotted (IB) with an anti-Flag or anti-Myc MAb.

IPO5 binding to Cap maintains PCV2 replication.

To identify which domain of Cap was involved in its interaction, plasmids encoding the NLS and a Cap mutant with the NLS deleted (dCap) were constructed (Fig. 4A). We measured the interactions in 293T cells transfected with IPO5 and one of the two domain deletion mutants of Cap or full-length Cap. We observed that the NLS deletion mutant could not be coimmunoprecipitated with IPO5 (Fig. 4B), confirming that amino acids 1 to 41 of Cap might be sufficient for IPO5 binding. To further identify the critical amino acid residues in the NLS that are responsible for binding to IPO5, we constructed 41 Cap mutants in which residues 1 to 41 within the NLS were individually mutated to alanine (A). All forty-one mutants, along with wild-type Cap, were subjected to co-IP assay. To our surprise, no specific amino acid in the NLS of Cap was identified as being required for their interaction (data not shown), but the interaction between Flag-IPO5 and Myc-CapR24A was weakest (Fig. 4C). These data indicated that the N-terminal R²⁴ residue was involved in the binding of Cap to IPO5. To explore the influence of the N-terminal R²⁴ residue of Cap on viral replication, the wild-type PCV2 and R²⁴A mutant PCV2 (24m-PCV2) infectious DNA clones were constructed and used to infect PK15 cells. As shown by the results in Fig. 4D, Cap signals could be detected both in wild-type (wt)-PCV2-infected and in 24m-PCV2-infected PK15 cells when immunostained with the anti-Cap MAb. Subsequently, the replicative abilities of wt PCV2 and 24m-PCV2 were determined. As shown by the results in Fig. 4E and F, the expression level of the Cap viral protein was significantly decreased in 24m-PCV2-infected cells compared with that in wt-PCV2-infected cells (P < 0.05) (Fig. 4E). Consistently, the replicative ability of 24m-PCV2 was markedly decreased in comparison with that of wt PCV2 (P < 0.05) (Fig. 4F). These results indicated that the R²⁴ residue within the NLS was significantly involved in PCV2 replication.

FIG 4.

The 24 N-terminal amino acid residues of Cap protein are important for PCV2 replication. (A) Schematic representation of truncated forms of PCV2 Cap. (B) 293T cells were cotransfected with plasmids expressing Cap, dCap, or NLS together with the Flag-IPO5 expression plasmid. Immunoprecipitation and immunoblotting were then performed to examine the interactions between various forms of Cap and Flag-IPO5. (C) Residues R²⁴ and L¹⁹ of Cap are critical for the interaction. 293T cells were cotransfected with Flag-IPO5 and Myc-Cap or its mutants for 36 h. Cellular lysates were subjected to co-IP assays. (D) Wild-type PCV2 and 24m-PCV2, in which Arg²⁴ is replaced by alanine, were rescued from PK15 cells, and the recovered wt PCV2 and 24m-PCV2 were validated by using an anti-Cap MAb and sequencing. Images were captured with a ×10 eyepiece and 20× lens objective. (E) PK15 cells were infected with wt PCV2 or 24m-PCV2 at an MOI of 1 and cultured for the indicated times. Cap and β-actin levels were analyzed with their corresponding antibodies in Western blotting. (F) One-step growth curves of wt PCV2 and 24m-PCV2. TCID₅₀ values were detected in wt-PCV2- or 24m-PCV2-infected samples. Data are presented as mean values ± SD from three independent biological experiments.

IPO5 has no influence on nuclear import of the newly synthesized Cap protein.

To investigate the role of IPO5’s binding to PCV2 in the nuclear import of PCV2 virions, it was determined whether IPO5 affected the nuclear location of Cap protein. Cap in the nuclear fraction was detected in Cap-transfected and PCV2-infected IPO5-silenced cells. No significant differences in the nuclear accumulation of Cap were observed between the IPO5-depleted group and the control group (Fig. 5A and B), indicating that substantial attenuation of IPO5 levels did not affect the nuclear localization of Cap. Furthermore, viral RNA transcription and viral protein levels were detected at 21 h postinfection (hpi) and 24 hpi. The results showed equal levels of Cap mRNA transcription in wild-type and IPO5-silenced cells at 21 hpi and 24 hpi (Fig. 5C); however, the Cap protein level was decreased in IPO5-deficient cells at 24 hpi compared with that in the wild-type cells (Fig. 5D). To further confirm these results, IPO5-silenced cells and control cells were transfected with the Cap-encoding plasmid, and Rep-encoding plasmids were used as a negative control. As shown by Western blotting, IPO5 attenuation resulted in the decrease of Cap but not Rep (Fig. 5E). Additionally, transcriptional analysis demonstrated that Cap mRNA was not affected in IPO5-silenced cells transfected with Cap (Fig. 5F). These observations suggested that the decrease of Cap protein in IPO5-deficient cells was not related to Cap gene transcription but might be due to the intracellular degradation of Cap.

FIG 5.

IPO5 is not involved in PCV2 Cap nuclear import and Cap gene expression. (A) IPO5-silenced and control cells were transfected with pcDNA3.0-Cap and pmCherry-Cap for 36 h. Then, the resultant cells were fixed, and cells transfected with pmCherry-Cap were stained with DAPI, while cells transfected with pcDNA3.0-Cap were stained for Cap (red) and nuclei (blue). Stained cells were subjected to confocal microscopy observation. (B) IPO5-deficient cells were infected with PCV2 at an MOI of 25 for 21 h and 24 h. Cells were then collected and subjected to separation of subcellular components and detection of Cap as in the experiment whose results are shown in Fig. 2C. (C and D) IPO5-silenced PK15 cells and control cells were infected with PCV2 at MOIs of 25 and 50 for 21 h and 24 h. (C) Cell lysates were probed with anti-Cap, anti-β-actin, and anti-IPO5 antibodies in immunoblotting experiments. (D) Cap mRNA transcripts were determined using real-time qPCR and analyzed using the 2^−ΔΔCT method. (E and F) IPO5-silenced cells were transfected with pcDNA3.0-Cap or pcDNA3.0-Rep for 36 h. (E) Cell lysates were probed with anti-Cap, anti-Rep, anti-IPO5, and anti-β-actin antibodies in immunoblotting experiments. (F) Cap and Rep mRNA transcripts were analyzed by comparative real-time qPCR. Data are presented as mean values ± SD from three independent biological experiments.

IPO5 plays a critical role in the stability of Cap protein through binding to the NLS.

To assess whether IPO5 was involved in the stability of the Cap protein, we monitored the stability of Cap protein in cells treated with cycloheximide (CHX). The results showed that the Cap protein level decreased gradually in all CHX-treated cells but decreased more rapidly in IPO5-silenced cells than in control cells (Fig. 6A). To assess whether proteasome-dependent degradation was responsible for the decreased Cap levels in the setting of IPO5 depletion, cells were treated with MG132 to inhibit the proteasome pathway. Western blot analyses of cell lysates showed that the reduction in the Cap protein level was prevented by MG132 in IPO5-silenced cells (P < 0.05) (Fig. 6B). Moreover, the rescue of IPO5 in IPO5-deficient cells reversed the reduction in Cap protein levels (P < 0.05) (Fig. 6C). Taken together, these results showed that IPO5 protects the Cap protein from proteasomal degradation. To identify the binding domain of Cap that is responsible for its stability, dCap and the NLS were detected by Western blotting in both the IPO5-deficient cell line and the control cell line following transfection. IPO5 silencing led to decreased levels of the NLS, similar to the results for Cap (P < 0.05) (Fig. 6D). The level of the NLS in the IPO5-silenced cells was rescued by proteasome inhibition using MG132 (P < 0.05) (Fig. 6E). Conversely, dCap was present at the same low levels with or without IPO5 attenuation (Fig. 6D). These results indicated that IPO5 binds to Cap to protect Cap from degradation and that the NLS domain of Cap is necessary for IPO5-mediated protection of Cap from proteasomal degradation.

FIG 6.

IPO5 interacts with the NLS domain of Cap to inhibit Cap degradation via the proteasome pathway. (A) IPO5-silenced cells and control cells were transfected with pcDNA3.0-cap for 36 h. Cells were then treated with CHX (100 μg/mL) for the indicated times. (B) Cap was overexpressed in shIPO5- or shCON-transduced PK15 cells treated with DMSO or MG132 (20 μM) for 8 h at 36 h posttransfection. (C) Different doses of Flag-IPO5 plasmids (0, 0.5, and 1 μg) were cotransfected with a fixed amount of Cap expression vector into the IPO5-silenced cells. (D) IPO5-silenced cells were transfected with Myc-tagged Cap, dCap, and NLS for 36 h. (E) NLS was overexpressed in shIPO5- or shCON-transduced PK15 cells treated with DMSO or MG132 (20 μM) for 8 h at 36 posttransfection. (A to E) The cell lysates were probed with anti-Myc, anti-Cap, anti-β-actin, and anti-IPO5 antibodies in the immunoblotting experiments. Data are presented as mean values ± SD from three independent biological experiments.

Both PCV2 incoming capsid and newly synthesized Cap protein enter the nucleus.

Circovirus particles are targeted to the nucleus by traveling along microtubules and might enter the nucleus without disassembly (31). As the smallest animal virus (32), previous studies have observed the nuclear location of PCV2 Cap protein in the late stage of infection (21). When and how newly synthesized Cap protein migrates into the nucleus at the early stage of infection is unknown. To determine the duration for which Cap exists as a virus particle in the cells for the first round of infection and the definite time point of Cap protein synthesis, PK15 cells were infected with PCV2 at MOIs of 1, 10, 25, and 50, and cells were collected at 12, 14, 16, 18, 21, and 24 hpi. As shown by the results in Fig. 7A, regardless of the initial dose of PCV2, Cap expression was significantly increased at 21 hpi (P < 0.05) (Fig. 7A). Interestingly, the Cap mRNA transcripts increased markedly at 18 hpi (P < 0.05) (Fig. 7B). The transcription of Cap increased slightly at 16 hpi in PK15 cells infected with PCV2 at a high MOI of 25 or 50 (Fig. 7B), indicating that Cap detected before 18 hpi might exist as PCV2 capsids. Subsequently, new Cap proteins were synthesized. PK15 cells infected with PCV2 at an MOI of 50 for 12 h, 14 h, and 16 h were collected, while cytoplasmic and nuclear fraction extractions were performed to measure the nuclear import of PCV2 capsids. As shown by the results in Fig. 7D, a small amount of Cap protein was detected in the nucleus at 12 hpi, and subsequently, nuclear entry of the Cap protein gradually increased at 14 hpi and 16 hpi (P < 0.05). Consistent with the Western blotting results, visible scattered dots, which indicated the viral particles, were observed in the nucleus using confocal microscopy at 14 hpi (Fig. 7C-b), indicating that the incoming viral capsids executed nuclear migration in the first round of PCV2 infection from 12 to 16 hpi. Newly synthesized Cap protein was diffusely distributed in the nucleus, the cytoplasm, or both the cytoplasm and nucleus at the late stage of PCV2 infection (Fig. 7C-c). Up to 70% of Cap protein was transported into the nucleus once it had been synthesized, with the level of nuclear-localized Cap protein gradually increasing with prolonged infection time (Fig. 7E). These results showed that both incoming viral capsids and newly synthesized Cap protein underwent nuclear transportation, with viral capsids undergoing nuclear migration at 12 to 16 hpi, while new Cap protein first appeared after 18 hpi and was transported into the nucleus instantaneously.

FIG 7.

Nuclear entrance of the PCV2 capsid and newly synthesized Cap protein. (A and B) PK15 cells were infected with PCV2 at MOIs of 1, 10, 25, and 50 for 12, 14, 16, 18, 21, and 24 h and then treated with CHX (100 μg/mL) (or not) at 12 h postinfection. (A) Cell lysates were probed using anti-Cap and anti-β-actin antibodies in immunoblotting experiments. (B) Cap mRNA transcript levels were determined using real-time qPCR and analyzed using the 2^−ΔΔCT method. (C) PK15 cells were infected with PCV2 at an MOI of 50 and cultured for 0 h (a), 14 h (b), and 21 h (c). The subcellular localizations of viral particles and Cap protein were examined using mouse anti-Cap IgG by confocal microscopy. The white outlines show the cell contour, the white arrows show the positions of PCV2 capsids. (D and E) PK15 cells infected with PCV2 at an MOI of 50 for 12, 14, and 16 h (D) or at an MOI of 10 for 21, 24, and 36 h (E) were collected and subjected to separation of subcellular components. Cap was detected in the cytoplasm and nucleus to validate the intranuclear colocalization of Cap. β-Tubulin served as a cytoplasmic marker, and histone H3 was used as a nuclear marker. Data are presented as mean values ± SD from three independent biological experiments.

DISCUSSION

The viral life cycle comprises cell entry, replication, assembly, and egress. Viruses usurp cellular factors and machinery to complete the process. For many viruses that replicate in the cell nucleus, the cellular and nuclear membranes are two major barriers that need to be overcome. Increasing evidence confirms that the components of the nuclear import machinery play a critical role in virus-host adaptation (33), host antiviral response regulation (34), and efficient virus replication and spread. Nuclear import receptors (NIRs), including 7 importin α genes and 20 importin β genes in humans (35), are essential components that are required for effective nuclear transport (26, 36). The use of NIRs in the host cell by viral proteins can improve virus infection. PCV2 is a nuclear replication-dependent virus. How it makes use of NIRs during infection is unclear. In this study, initial observation by liquid chromatography-tandem mass spectrometry (LC-MS/MS), confocal laser microscopy, and Western blotting confirmed that PCV2 Cap bound to IPO5 (Fig. 1A and 3). IPO5 attenuation decreased Cap expression and inhibited progeny virus production (Fig. 1), indicating that IPO5 facilitates viral replication during PCV2 infection.

Studies on the nucleocytoplasmic trafficking system and viruses perfectly complement each other. Viruses are used as invaluable tools for the discovery of novel nuclear transport pathways. Conversely, disruption or usage of specific nuclear transport components has been shown to be crucial for the successful life cycle of viruses and the discovery of novel antiviral defense mechanisms (37). NIRs are required to recognize and facilitate cargos to shuttle into the nucleus through nuclear pores; therefore, viruses generally exploit their nuclear import function to achieve effective infection (38). Viral proteins containing an NLS interact directly with NIRs to move into the nucleus to facilitate viral replication. For example, the nonclassical NLS of herpes simplex virus 1 (HSV-1) VP19C binds with importin β1 for nuclear import, and an NLS-mutated recombinant virus replicates less efficiently than wild-type HSV-1 (39). Damage to the nuclear import of influenza A virus nucleoprotein suppresses virus replication (40). Meanwhile, several viruses are known to escape the immune response by inhibiting nuclear import of transcription factors through competitive combination with NIRs. Ebola virus VP24 protein binds importin α1 to inhibit PY-STAT1 nuclear import to counteract the antiviral effects of interferon (IFN) (41). Chronic hepatitis B virus polymerase binding to importin α5 interferes with the nuclear transportation of STAT1/2 (42). In this study, IPO5 had no effect on Cap protein nuclear import, but it maintained Cap stability (Fig. 5 and 6) and accelerated the transfer of virions into the nucleus (Fig. 2), indicating a novel mechanism in which a viral protein usurps a host protein to favor viral replication. This phenomenon was first discovered in viral infection and is likely to provide new insights into key regulatory steps of NIRs in viral infection, which might serve as potential targets for viral therapy.

The Cap protein is the main structural protein of PCV. It is not only related to host immunity and virus virulence but is also indispensable in virus replication and shows the nuclear shuttle phenomenon of “cytoplasm→nucleoplasm→nucleolus→cytosol” (11, 21, 43–45). Our knowledge concerning the replication and maturation of PCV is rather limited; therefore, investigation of the subnuclear localization of Cap will provide additional information to understand these processes. NPM1 bound with Cap protein to promote its nucleolar location (46). Based on the nuclear localization of Cap during early infection, it was speculated that regulation of the cell cycle, induction of mitosis stagnation in the S phase, and modification of host protein expression might be executed when Cap is nuclearly located (21). Classical and nonclassical nuclear pathways have been clarified based on the types of cargos and NIRs (22–25). IPO5, a member of the importin β superfamily, was first described as a part of the trimeric RanBP1-Ran-RanBP5 complex, which is predominantly distributed in the cytoplasm, can bind to nuclear pore complexes, and is reported to mediate in influenza A virus PB1-PA and human papillomavirus 11 (HPV11) L2 protein import as the mediator of a nucleocytoplasmic nonclassical transport pathway (47–50). In the present study, IPO5 had no effect on Cap protein nuclear import (Fig. 5A and B). We attempted to search for other nuclear transport receptors (NTRs) that might assist Cap to target the nucleus. However, Cap did not interact with importin β1 or with importin β2 (unpublished data). Thus, other transport factors involved in Cap migration need to be further investigated.

IPO5 has been reported to possess multiple functions and is expected to have a role in nucleoplasmic transport (51–53). Recently, IPO5 was discovered to bind to the intracellular protein FLIL33 to inhibit its proteasome-dependent degradation, but it was not required for the nuclear localization of FLIL33 (51). No other research on viral proteins or other exogenous proteins reported similar influences on IPO5 expression. In this study, we found that IPO5 bound to the NLS region of Cap but did not mediate the nuclear import of Cap; instead, it increased the stability of the Cap protein by inhibiting its proteasome-dependent degradation (Fig. 4, 5, and 6). A recent study showed that an E3 ubiquitin ligase pMKRN1 promotes the degradation of Cap (54). Our research might extend our knowledge about how PCV2 Cap proteins escape the cellular degradation machinery. Previous reports proposed that PCV2-driven endoplasmic reticulum (ER) stress is Cap dependent (55, 56). Exogenous pathogens, such as viruses and toxins, always hijack protein degradation machinery in endoplasmic reticulum-associated degradation (ERAD) for their invasion and evasion of immunological surveillance (57). Besides, importin family members have been implicated in the regulation of the degradation of targets through ERAD (58). Thus, it is possible that IPO5 might act as a novel component of ERAD complexes, similar to previously reported importin β, which stabilizes ERAD substrates. Another possibility is that the binding of IPO5 to Cap might promote Cap ubiquitination or inhibit Cap deubiquitination. A third possibility is that IPO5 blocks the delivery of Cap to the proteasome. However, these speculations require further experimental verification.

In summary, we discovered that hijacking of IPO5 by PCV2 promotes the nuclear transport of the incoming capsids at an early stage of infection, inhibits the proteasome-dependent degradation of Cap protein, and facilitates Cap expression and viral propagation. Our data highlight a mechanism whereby IPO5 promotes nuclear transport of the incoming PCV2 capsids and inhibits proteasome-dependent degradation of newly synthesized Cap protein in PCV2 infection, which not only enriches our understanding of the virus replication cycle but also lays the foundation for the subsequent development of antiviral drugs.

MATERIALS AND METHODS

Cells and virus.

Porcine kidney epithelial cell line PK15 (CCL-33; ATCC, Manassas, VA, USA) and the human embryonic kidney epithelial cell line HEK 293T (CRL-3216; ATCC) were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Carlsbad, CA, USA) supplemented with 5% fetal bovine serum. PCV2 strain HZ0201 (accession no. AY188355) (10^8.25 50% tissue culture infective doses [TCID₅₀]/mL; stored in our laboratory) was originally isolated from a pig with naturally occurring postweaning multisystemic wasting syndrome (59) and was propagated in PK15 cells. To investigate the nuclear import of the virus, cells were infected with PCV2 at an MOI of 50 for the indicated times. To detect virus replication, cells were infected with PCV2 at MOIs of 1, 10, and 25 and cultured for the indicated times.

Plasmid construction and cell transfection.

The Cap gene, cloned from the PCV2 HZ0201 genome, was inserted into the multiple cloning site of pCMV-N-myc (Clontech, Palo Alto, CA, USA), pmCherry-C1 (Clontech), and pET-sumo/CAT (Invitrogen, Carlsbad, CA, USA) to construct recombinant expression vectors. The full-length IPO5 cDNA was amplified by PCR from the cDNA of PK15 cells and inserted into plasmids pCMV-N-flag and pGEX4T-1 (GE Healthcare Biosciences, Piscataway, NJ, USA), resulting in pCMV-N-flag-IPO5 and pGEX4T-1-IPO5. IPO5 mutant variants were constructed using inverse PCR with pCMV-N-flag-IPO5 as the template. Briefly, the DNA fragment of full-length IPO5 was amplified from pCMV-N-flag-IPO5, purified, digested with DpnI, and transformed into Escherichia coli strain DH5a. The truncated and site-directed-mutation mutants of Cap were constructed using the same methods, with pCMV-N-myc-Cap as the template. The resultant plasmids were termed Myc-dCap, Myc-NLS, Myc-L19A-Cap, and Myc-R24A-Cap. All constructs were confirmed by sequencing. The primer sequences used for plasmid constructions are available on request.

For cell transfection, PK15 cells were plated into 35-mm-diameter glass bottom dishes at a suitable density according to the experimental schemes and grown to 40% to 50% confluence. JetPrime transfection reagent (Polyplus Transfection, New York, NY, USA) was used for PK15 cell transfection. 293T cells were seeded into culture dishes and grown to 80% to 90% confluence. ExFect transfection reagent (catalog number T101-01/02; Vazyme Biotechnology, Nanjing, China) was used for 293T cell transfection.

SDS-PAGE, immunoblotting, and silver staining.

Cells were lysed using lysis buffer (2% SDS, 1% Triton X-100, 50 mM Tris-HCl, pH 7.4, 8 M urea, 5% mercaptoethanol, and a protease inhibitor mixture) as previously described (9), and proteins were separated using SDS-PAGE gels (10% for IPO5, 12% for Cap, β-actin, β-tubulin, and histone H3, and 15% for truncated Cap fragments). The separated proteins were transferred to nitrocellulose membranes (Amersham Biosciences, Piscataway, NJ, USA). After blocking with 5% skim milk in phosphate-buffered saline–Tween 20 (PBST) at room temperature for 1 h, the membranes were incubated with primary antibodies overnight at 4°C. Finally, an appropriate secondary antibody, horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG or goat anti-rabbit IgG (Kirkegaard & Perry Laboratories, Inc., Gaithersburg, MD, USA), was reacted with the immunoreactive proteins on the membrane at 37°C for 1 h. The immunoreactive protein bands were visualized using a SuperSignal West Femto maximum-sensitivity substrate (Thermo Fisher Scientific, Rockford, IL, USA) and captured using an AI680 chemiluminescent imaging system (GE Healthcare).

Silver staining was performed according to a previously described protocol (60). Briefly, the polyacrylamide gel was fixed using a 4:1:5 ratio of ethanol, glacial acetic acid, and double-distilled water at room temperature for 30 min and then washed with double-distilled water three times. After incubation with sensitizing solution containing sodium acetate, sodium thiosulfate, and absolute ethanol, the gel was transferred to the silver stain solution and incubated in the dark at room temperature for 30 min. Finally, a color developer containing anhydrous sodium carbonate, formaldehyde, and sodium thiosulfate was added and the mixture incubated in the dark with gentle agitation. The reaction was terminated by adding 5% glacial acetic acid. The stained bands were scanned using an ImageScanner III instrument (GE Healthcare, Chicago, IL, USA), and the bands of interest were excised from the gel for liquid chromatography-mass spectrometry (LC-MS) analysis.

IFA and confocal microscopy.

To detect the viral titer, an indirect immunofluorescence assay (IFA) was performed according to a previously described protocol (61). Mouse MAb against Cap was used as the primary antibody to indicate the virus.

Confocal microscopy was used to observe the protein localization in cases of plasmid transfection or PCV2 infection. To detect the distribution of the Cap protein under the influence of IPO5, Cap expression plasmids were transfected into IPO5-silenced cells or control cells. At 36 h after transfection, cells were fixed using 4% polyformaldehyde for 20 min at room temperature, permeabilized with 0.2% Triton X-100 for 2 min, and blocked with 5% skim milk for 30 min at 37°C. After washing thrice with PBS, the membranes were incubated with anti-Cap MAb as the primary antibody and Alexa Fluor 546-conjugated donkey anti-mouse antibody as the secondary antibody. Finally, the cell nuclei were stained using DAPI (4′,6-diamidino-2-phenylindole; Sigma-Aldrich, St. Louis, MO, USA) for 10 min and observed under a Zeiss LSM 780 confocal laser scanning microscope (Carl Zeiss, Jena, Germany). To detect the localization of capsids during the early stage of PCV2 infection and the colocalization between capsids and IPO5, PK15 cells grown on glass coverslips (In Vitro Scientific, Sunnyvale, CA, USA) were subsequently infected with PCV2 at an MOI of 50 for the indicated time. Anti-Cap MAb (61) alone or rabbit anti-IPO5 antibody (ab187175; Abcam, Cambridge, MA, USA) was used as the primary antibody, and Alexa Fluor 546-conjugated donkey anti-mouse IgG (Life Technologies, Carlsbad, CA, USA) and fluorescein isothiocyanate (FITC)-conjugated goat anti-rabbit IgG (KPL, Milford, MA, USA) were used as secondary antibodies. The nuclei were stained using DAPI at room temperature for 10 min. The stained cells were observed using a Zeiss LSM 980 Airyscan 2 confocal microscope.

Coimmunoprecipitation and glutathione S-transferase (GST) pulldown assays.

To confirm the interaction between Cap and IPO5 in vivo and in vitro, the vector expressing Myc-tagged Cap was cotransfected into 293T cells with the vector encoding Flag-tagged IPO5. Vectors expressing truncated Cap were cotransfected with those expressing IPO5 into 293T cells to screen the binding domain of Cap. At 36 h after transfection, cells were lysed using NP-40 lysis buffer (P0013F; Beyotime, Jiangsu, China) supplemented with 1 mM phenylmethylsulfonyl fluoride (PMSF) (ST506; Beyotime) at 4°C for 2 h. The supernatant was first pretreated with protein A/G Plus-agarose (catalog number sc-2002; Santa Cruz Biotechnology, Santa Cruz, CA, USA) for 1 h at 4°C and then incubated with Flag-agarose beads (Sigma-Aldrich) or anti-Myc MAb (2712325; Millipore, Billerica, MA, USA) preincubated with protein A/G Plus-agarose for 4 h at 4°C. After three stringent washes in NP-40 lysis buffer, the immunoprecipitated proteins were analyzed using Western blotting with mouse anti-Flag MAb (F1804; Sigma) and rabbit anti-Myc polyclonal antibody (PAb) (R1208-1; Huaan Biological Technology, Hangzhou, China). To confirm the interaction between Cap and endogenous IPO5 during viral infection, cells infected with PCV2 at the indicated times were collected. After eliminating nonspecific binding using protein A/G Plus-agarose, the supernatant was incubated with mouse anti-Cap MAb at 4°C for 4 h, and immune complexes were then precipitated by incubation with protein A/G Plus-agarose for another 4 h at 4°C. The samples were subjected to SDS-PAGE, followed by immunoblotting using mouse anti-Cap MAb and rabbit anti-IPO5 PAb.

For the GST pulldown assay, the expression of His-Sumo-Cap, GST, and GST-IPO5 in Escherichia coli strain BL21 was induced using 0.2 mM IPTG (isopropyl-β-d-thiogalactopyranoside; Amersham) for 16 h at 16°C. GST and GST-IPO5 were obtained through incubation with Pierce glutathione agarose (Thermo, Rockford, IL, USA) and were eluted with 1 mL of cold 1× PBS per 10 mg of reduced glutathione. His-Sumo-Cap was purified using nitrilotriacetic acid (NTA)-agarose affinity resin (catalog number 30210; Qiagen, Hilden, Germany) and eluted using an imidazole concentration gradient. Equal amounts of purified GST or GST fusion proteins, which were immobilized on glutathione agarose beads, were incubated with the corresponding prey proteins at 4°C for 6 h. The precipitated proteins were washed with PBS, subjected to SDS-PAGE, and detected using immunoblotting with mouse anti-GST MAb (M0807-1; Huaan Biological Technology) and mouse anti-His MAb (produced and stored in our laboratory).

Nuclear and cytoplasmic extraction.

Isolation of nuclear and cytoplasmic components was performed as stated in a previous study (9). According to the protocol, PCV2-infected cells were lysed using 0.1% NP-40 lysis buffer containing 1 mM PMSF at 4°C for 5 min. After centrifugation at 1,000 × g for 5 min, the supernatants of the samples were collected as the cytoplasmic fractions, while the precipitates were lysed with strong lysis buffer and used as the nuclear fractions. Western blotting was performed using mouse MAb to Cap, mouse MAb against histone H3 (R1105-1; Huaan Biological Technology), and rabbit MAb against beta-tubulin (ET1602-4; Huaan Biological Technology).

qPCR.

According to the manufacturer’s instructions, total cellular RNA was extracted by adding TRIzol reagent (Invitrogen). Based on the manufacturer’s protocol, reverse transcription of total RNA was carried out using a RevertAid RT reverse transcription kit (catalog number K1691; Thermo Fisher Scientific). Quantitative real-time PCR (qPCR) was performed to detect the relative abundances of Cap transcripts using ChamQ universal SYBR qPCR master mix (catalog number Q711-02/03; Vazyme Biotechnology) in a LightCycler 96 sequence detector system (Roche, Basel, Switzerland). The primer sequences for the real-time qPCR step were the same as those used previously (19).

Construction of a mutant PCV2 clone and virus rescue.

The whole genome of PCV2 strain HZ0201 cloned into vector pMD-18T was stored in our laboratory. Mutant PCV2 with residue Arg²⁴ replaced by alanine (24m-PCV2) was constructed by a site-specific mutation experiment using the pMD-18T-HZ0201 plasmid as the template. The linear genomes of wt and mutant PCV2 were extracted by digesting with EcoRI, cyclized using T4 DNA ligase, and transfected into PK15 cells for virus rescue with JetPrime transfection reagent. The transfected cells were cultured for 48 h, and then serial passage of PK15 cells was performed eight times to rescue viruses. The recovered viruses were analyzed by IFA with an anti-Cap MAb, and the genome sequences were confirmed by sequencing.

One-step growth curve.

To analyze the replicative ability of the rescued virus, PK15 cells were individually infected with the wt and 24m-PCV2 at an MOI of 1. The cells were harvested at the indicated times and freeze-thawed three times. Then, the supernatant fractions were collected to perform the titration of the TCID₅₀ values.

IPO5 knockdown by lentivirus-mediated RNA interference.

Three pairs of shRNAs targeting porcine IPO5 (shIPO5-1, -2, and -3) were designed using small interfering RNA (siRNA) design software and cloned into the pGreenPuro shRNA lentivector (catalog number SI505A-1; System Biosciences, Palo Alto, CA, USA). After transfection and selection, an effective shIPO5 (targeting sequence GCGTCCTCATTTGGAAGCTACTTTA) and the control shRNA shCON, which was developed previously (9), were cotransfected using a ViraPower lentiviral packaging mix (Thermo Fisher) into 293T cells for 48 h. PK15 cells were then infected with the resultant lentiviruses, cultured for 24 h, and selected using puromycin (5 μg/mL) (catalog number A1113803; Invitrogen) for a week to obtain IPO5 knockdown cells. Finally, Western blotting analysis was performed using rabbit PAb against IPO5 and mouse MAb against β-actin. Cell viability was determined using a cell counting kit-8 (CCK-8) assay (Beyotime) according to the manufacturer’s protocol.

Cell treatment.

To inhibit the newly synthesized Cap protein during PCV2 infection, PK15 cells infected with PCV2 at the indicated MOI were treated with the protein synthesis inhibitor cycloheximide (CHX) (100 μg/mL; MedChemExpress, Monmouth Junction, NJ, USA) at 12 hpi for the indicated times. To monitor the stability of Cap in IPO5-deficient cells, the vector expressing Cap was transfected into IPO5-silenced cells or control cells for 36 h, and then CHX (100 μg/mL) was added for 3, 6, 9, and 12 h. To analyze IPO5-mediated degradation of Cap and Cap truncation mutants via the proteasome pathway, transfected cells were treated with MG132 (20 μM) and dimethyl sulfoxide (DMSO) for 8 h. Cell lysates were collected and subjected to immunoblotting using an anti-Cap MAb, an anti-IPO5 PAb, and an anti-β-actin MAb.

Cell viability.

Cell viability was measured using a cell counting kit-8 (CCK-8) assay (Beyotime) according to the manufacturer’s protocol. To exclude the possibility that drug pretreatment affected cell viability, cells were plated in 96-well plates at a concentration of 1 × 10³ cells/mL per well and divided into a control (Con) group, a CHX group, an MG132 group, and a CHX-plus-MG132 group. After treatment for 3, 6, 9 and 12 h, 10 μL CCK-8 solution was added to each well for 2 h. The absorbance of these wells was detected using a microplate reader at 450 nm.

Statistical analysis.

All results are presented as mean values ± standard deviations. Statistical analysis was performed using the two-tailed Student’s t test. P values of ≤0.05 were considered significant.