PRMT5 Facilitates Infectious Bursal Disease Virus Replication through Arginine Methylation of VP1

ABSTRACT

The infectious bursal diseases virus (IBDV) polymerase, VP1 protein, is responsible for transcription, initial translation and viral genomic replication. Knowledge about the new kind of post-translational modification of VP1 supports identification of novel drugs against the virus. Because the arginine residue is known to be methylated by protein arginine methyltransferase (PRMT) enzyme, we investigated whether IBDV VP1 is a substrate for known PRMTs. In this study, we show that VP1 is specifically associated with and methylated by PRMT5 at the arginine 426 (R426) residue. IBDV infection causes the accumulation of PRMT5 in the cytoplasm, which colocalizes with VP1 as a punctate structure. In addition, ectopic expression of PRMT5 significantly enhances the viral replication. In the presence of PMRT5, enzyme inhibitor and knockout of PRMT5 remarkably decreased viral replication. The polymerase activity of VP1 was severely damaged when R426 mutated to alanine, resulting in impaired viral replication. Our study reports a novel form of post-translational modification of VP1, which supports its polymerase function to facilitate the viral replication.

IMPORTANCE Post-translational modification of infectious bursal disease virus (IBDV) VP1 is important for the regulation of its polymerase activity. Investigation of the significance of specific modification of VP1 can lead to better understanding of viral replication and can probably also help in identifying novel targets for antiviral compounds. Our work demonstrates the molecular mechanism of VP1 methylation mediated by PRMT5, which is critical for viral polymerase activity, as well as viral replication. Our study expands a novel insight into the function of arginine methylation of VP1, which might be useful for limiting the replication of IBDV.

INTRODUCTION

Infectious bursal disease virus (IBDV) is a representative member of the Birnaviridae family (1). IBDV is the causative agent of the highly contagious disease disrupting the maturation of antibody-producing B-lymphocytes in the bursal of Fabricius, which results in serious immunosuppression and high mortality in young chickens (2). The genome of IBDV contains two segments of double-strand RNA (dsRNA). Segment A consists of two overlapping open reading frames (ORFs) that encode a nonstructural protein VP5 and a polyprotein VP2-4-3, which is autocatalytically processed into the viral capsid protein VP2, multifunctional protein VP3, and protease protein VP4 (3–5). Segment B encodes protein VP1 with a molecular weight of approximate 90 kDa includes six classical motifs (designated G, F, C, A, B, and E) that is highly similar to RNA-dependent RNA polymerase (RdRp) of other RNA viruses (6–9). VP1 protein was found to exist as two forms: free VP1 (also named “VP1 alone”) and genomic conjugated VP1 (also named “VPg”), both of which are encapsulated into viral particles (10, 11). During the process of viral replication, VPg along with VP3 is responsible for the viral genome replication, while the free VP1 is involved in the virion packaging (12, 13). VP1, VP3, and dsRNA together form the viral replication complex of IBDV, which is responsible for viral replication (13, 14).

The polymerase activity of VP1 is regulated by multiple factors, including VP3 protein, several host proteins, and especially the post-translational modifications, such as SUMOylation and ubiquitination (15–18). VP3, which is also considered polymerase helper protein, has been reported to facilitate the polymerase activity of VP1 in vitro (13). Initiation of viral replication by eIF4AII depends on the interaction between eIF4AII and VP1, which supports the polymerase activity of VP1 (19). SUMOylation of VP1 inhibits the K48-linked ubiquitination and degradation of VP1 via the proteasome pathway (17). As an exception of SUMOylation, the K63-linked ubiquitination of VP1 plays the essential role of promoting the polymerase activity of VP1 (18). In addition to SUMOylation and ubiquitination, arginine methylation of multiple viral proteins has been reported in previous literature (20, 21). Hence, it may be possible that IBDV VP1 is an arginine methylation-modified protein that is mediated by the cellular protein arginine methyltransferases (PRMTs).

Arginine methylation is a common post-translational modification controlled by the enzymes of the PRMT family (22). Arginine methylation is categorized into asymmetric dimethylarginines (aDMAs), in which two methyl molecules are attached to the same nitrogen atom, and symmetric dimethylarginines (sDMA), in which two methyl molecules are added to the different nitrogen atoms. Type I PRMTs, including PRMT1, PRMT3, CARM1, and PRMT6, are responsible for the formation of aDMA. The PRMT5 is the only known type II enzyme that is responsible for catalyzing the formation of sDMA. Another class of PRMTs, including PRMT7, PRMT8, and PRMT9, catalyze the formation of monomermethylarginines (23). The arginine methylation is a dynamic modification. It has been shown that certain lysine demethylases, such as KDM3A and KDM5C, also play the role of arginine methylation activity (24, 25). Based on previous reports, arginine methylation is known to affect several host signaling pathways. For example, arginine methylation of MAVS mediated by PRMT7 and PRM9 negatively regulates IFN-β production (26, 27). PRMT1 inhibits TBK1 activation by arginine methylation of TBK1 at multiple arginine residues (28). In addition, arginine methylation of several viral proteins has been reported to affect viral replication. PRMT6 targets the Tat protein of HIV for arginine methylation, which supports viral gene expression (21, 29, 30). Arginine methylation of the hepatitis C virus (HCV) NS3 protein on the R1493 residue mediated by PRMT1 reduces its RNA-binding capacity (31). Arginine methylation of ICP27 RGG box is necessary for HSV-1 replication (32, 33). Whether IBDV VP1 could be modified by arginine methylation is still unclear.

In this report, we investigate the role of PRMT5 on arginine methylation of VP1. IBDV infection causes the accumulation of PRMT5 in the cytoplasm, which colocalized with VP1 as a punctate structure. Overexpression of PRMT5 facilitates viral replication. PRMT5 inhibitor treatment, as well as PRMT5 knockout, significantly blocks IBDV replication. PRMT5 specifically interacts with VP1 and supports symmetric dimethylation of VP1 at arginine 426 (R426), which then promoted polymerase activity. Further, we rescued R426A mutant in recombinant IBDV and examined its effects on viral growth kinetics. IBDV VP1 is the first viral polymerase identified as the substrate for PRMT5. Our data indicate that arginine methylation is a novel post-translational modification of the polymerase protein of IBDV.

RESULTS

IBDV VP1 is a symmetrical dimethylation (sDMA)-modified protein.

The role of viral protein arginine methylation has been reported previously in several viruses (34). Arginine methylation is classified into monomer methylation (MMA), symmetrical dimethylation (sDMA), and asymmetric dimethylation (aDMA) (35). In order to find out whether VP1 undergoes sDMA during transfection and IBDV infection, immunoprecipitated (IP) complexes of 293T cells transfected with Flag-VP1 plasmids were probed by sDMA antibody. We found that sDMA antibody could detect sDMA moiety of VP1, suggesting that VP1 could be modified by sDMA (Fig. 1A, lane 2). To further confirm this modification of VP1 during IBDV infection, immunoblotting of purified VP1 from infected DF-1 cells shows that VP1 could be modified by sDMA during viral replication (Fig. 1B, lane 2). These data demonstrate that IBDV VP1 could be modified by sDMA.

FIG 1.

Infectious bursal diseases virus (IBDV) VP1 is a symmetric dimethylation arginine (sDMA) modified protein. (A) 293T cells were transiently transfected with Flag-VP1 (empty vectors as a negative control) for 36 h. Flag-VP1 was immunoprecipitated (IP) with anti-Flag agarose, which was subsequently analyzed by Western blotting with indicated antibodies. (B) DF-1 cells infected with 1 multiplicity of infection (MOI) IBDV for 12 h were subjected to IP with VP1 antibodies. Immunoblotting (IB) of IP complex and cell lysates by Western blotting was performed using the indicated antibodies. Met, methylated.

PRMT5 interacts with VP1.

Protein arginine methylation is carried out by a group of protein arginine methyltransferase enzymes, also called the PRMT family (36). To identify which PRMT causes the sDMA of VP1, we attempted to identify the potential interacting partners of VP1 using liquid chromatography-mass spectrometry (LC-MS). Purified Flag-VP1 from transfected 293T cells was separated by SDS-PAGE and visualized by CBB staining; then, several individual bands were subjected to LC-MS analysis. The results showed that PRMT5 was one potential interacting partner of VP1 (Fig. 2A). There were a total 10 peptides of PRMT5 with approximately 19% sequence coverage (Fig. 2B and C). The function of others partner of VP1, such as CDK1 and hnRNPU, have been confirmed by our published work (37, 38). In this study, we focused on the role of PRMT5. Co-IP analysis further confirmed the interaction of VP1 and PRMT5, which is independent of the enzyme activity of PRMT5 (Fig. 2D, lane 3, and E, lanes 3 and 4). Additionally, the association between VP1 and PRMT5 was also detected under the condition of IBDV infection via IP assay (Fig. 2E, lane 2). To exclude the possibility of indirect association between VP1 and PRMT5, glutathione S-transferase (GST) pulldown assay with purified GST-VP1 and His-PRMT5 (and its mutant) showed that VP1 interacted with PRMT5, as well as its R375A mutant directly (Fig. 2F, lanes 2 and 3). In humans, the family of PRMT contains PRMT1 to PRMT9, but in chickens, the PRMT2, PRMT6, PRMT8, and PRMT9 enzymes are absent (39). To validate the LC-MS data of PRMT5 interacting with VP1, all the chicken PRMTs were synthesized for co-IP analysis. We found that only PRMT5 could efficiently interact with VP1, while other PRMTs failed to do so (Fig. 2G, lane 6). Altogether, these data suggest that PRMT5 might be the protein arginine methyltransferase supporting sDMA of VP1.

FIG 2.

PRMT5 interacts with VP1. (A) Flag-VP1 was immunoprecipitated (IP) with anti-Flag agarose from 293T cells transfected with Flag-VP1. The samples were separated by SDS-PAGE and stained with Coommassie Brilliant Blue (CBB). Separated specific bands were excised and analyzed by mass spectrometry. Several host proteins were associated with VP1, including PRMT5. (B) The coverage of identified peptides in the complete length of PRMT5. (C) The identified peptides located in the sequence of PRMT5. (D, E) Immunoblotting (IB) of whole-cell lysates (bottom panel) and IP complex (top panel) with from 293T cells transfected with indicated plasmids for 36 h. (F) Immunoblotting (IB) of whole-cell lysates (bottom panel) and IP complex (top panel) with from DF-1 cells infected with 1 MOI IBDV for 12 h. (G) Purified glutathione S-transferase (GST)-VP1 proteins were individually incubated with purified His-PRMT5 (or R375A mutant PRMT5 [RA]), followed by IP with GST resin. IP complex was analyzed by Western blotting with indicated antibodies. The whole protein was examined by SDS-PAGE and visualized by Coomassie Brilliant Blue (CBB) staining. (H) Flag-VP1 was individually cotransfected with different Myc-PRMTs plasmids into 293T cells for 36 h. Immunoblotting of IP complex and whole-cell lysates was performed by Western blotting with indicated antibodies. β-Actin protein was used as the loading control. WT, wild type.

PRMT5 contributes viral replication.

To elucidate the role of PRMT5 during the IBDV life cycle, we detected the protein level and mRNA transcription level of PRMT5 upon IBDV infection. The results showed that both the protein and mRNA levels were not changed significantly during the IBDV infection (Fig. 3A and B). Our previous work showed that multiple host proteins, such as CDK1, cyclin B1, and hnRNPU, were all upregulated after IBDV infection (37, 38). Prior work suggests that PRMT5 is diffusely distributed in both the nucleus and the cytoplasm (40). However, protein extraction analysis demonstrated that IBDV infection causes increased accumulation of PRMT5 into the cytoplasm and reduced levels in the nucleus (Fig. 3C and D). Since IBDV replication primarily takes place into the cytoplasm without any nuclear stage (41), we hypothesize that IBDV infection recruits PRMT5 to VP1 into the cytoplasm to support the viral replication. To confirm this, confocal microscopy analysis showed that PRMT5 diffusely distributed in cytoplasm and nucleus in resting state, whereas during the IBDV infection, PRMT5 completely accumulated in the cytoplasm and colocalized with VP1 as a punctate structure (Fig. 3E and F). These data prompted us to investigate the role of PRMT5 in viral replication. Quantitative reverse transcription-PCR (RT-qPCR) results indicate that the copy number of viral RNA increased nearly 5-fold in 6 hours postinfection (hpi) and 2-fold in 12 hpi in DF-1 cells transfected with PRMT5 compared to the control cells (Fig. 3G). Consistent with the result of RT-qPCR, Western blot analysis also demonstrated that ectopic expression of PRMT5 significantly increased the expression of major viral proteins, such as VP1, VP2, and VP3, compared to the control cells (Fig. 3H and I). In addition, the viral titers in DF-1 cells transfected with PRMT5 were remarkably higher than that in control cells (Fig. 3J). In summary, IBDV infection causes the accumulation of PRMT5 into the cytoplasm that colocalizes with VP1, and overexpression of PRMT5 facilitates IBDV replication.

FIG 3.

PRMT5 contributes IBDV replication. (A, B) DF-1 cells were infected with 1 MOI IBDV for different time. The cells were subjected to Western blot with corresponding antibodies (A) and quantitative reverse transcription-PCR (RT-qPCR) analysis (B). (C) DF-1 cells infected with 1 MOI IBDV for different time were subjected to subcellular proteins extraction, which were followed by Western blotting with indicated antibodies. (D) Comparison of the expression of PRMT5 in different lanes in panel C. The density of PRMT5, β-actin, and H3 was quantified by ImageJ software. (E) Myc-PRMT5 was transfected into DF-1 cells for 12 h, which were further infected with 1 MOI IBDV for another 12 h. The cells were subjected to immunofluorescence assay (IFA) and microscopy scanning. PRMT5 (green), VP1 (red), and DAPI (blue) showed in the presented images. Bars, 5 μm. (F) Quantification of PRMT5 positive cells showed in panel E. (G to I) DF-1 cells were transfected with Flag-PRMT5 plasmids (empty vector for control) for 12 h, which were subsequently infected with I MOI IBDV for different time. Total viral RNA expression (G) and major viral proteins (VP1/2/3) expression (H) were analyzed by RT-qPCR and Western blotting, respectively. (I) Comparison of the expression of VP1/2/3 in different lanes in panel H. The density of VP1/2/3 and β-actin was quantified by ImageJ software. (J) The DF-1 cells were transfected with Myc-PRMT5 for 12 h and then infected with 0.1 MOI IBDV for additional different time. Viral titers of whole cells were examined by 50% tissue culture infective dose (TCID₅₀) assay. The Western blot data were semiquantified and normalized against β-actin protein loading control. The data were obtained from three independent experiments and presented as means ± standard deviation (SD). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant.

Inhibitor of PRMT5 suppresses IBDV replication.

Since PRMT5 is a crucial enzyme responsible for arginine methylation of various substrates, its role in IBDV replication is still unclear. HLCL-61, a well characterized inhibitor of PRMT5, was chosen to study the role of PRMT5 activity on viral replication (42). According to enzyme assays in vitro, HLCL-61 was determined as one of the most potent of the selected inhibitors. HLCL-61 possessed no inhibitory activity against the PRMT1/4/7 proteins but had specificity for PRMT5. This inhibitory function of HLCL-61 significantly suppressed the symmetric arginine demethylation (sDMA) of histones H3 and H4 in mammalian cells (43). Subsequent experiments showed that HLCL-61 can inhibit PRMT5 activity by overlapping the enzymatic active site of the double-E loop or pyridine ring in the S-adenosylmethionine (SAM) domain that is highly conserved in different species’ PRMT5 (44). Thus, we think that HLCL-61 can also inhibit the activity of chicken PRMT5. First, we assessed the cell viability in the presence of different concentrations of HLCL-61. CCK8 analysis showed that there was no severe toxicity at the 4 μg/mL of HLCL-61 (Fig. 4A), which also did not affect the PRMT5 expression level at different times (Fig. 4B), but HLCL-61 treatment noticeably inhibited arginine methylation of VP1 (Fig. 4C, line 3). We found that HLCL-61 treatment significantly decreases the viral genome copies with 10-fold reduction of at 6 hpi and 30-fold reduction at 12 hpi compared with the cells treated with dimethyl sulfoxide (DMSO) (Fig. 4D). In line with the RT-qPCR analysis, Western blot analysis also demonstrated that the expression of major viral proteins (VP1, VP2, and VP3) was reduced to 50% in HLCL-61-treated cells compared with the DMSO-treated cells (Fig. 4E and F). In addition, the viral titer in HLCL-61-treated cells was also significantly lower than that in DMSO-treated cells (Fig. 4G). Taken together, our data showed that inhibition of the activity of PRMT5 significantly disrupts IBDV replication.

FIG 4.

Inhibitor of PRMT5 suppresses viral replication. (A) DF-1 cells were treated with different concentrations of HLCL-61, which were followed with cell viability analysis. (B) DF-1 cells were treated with 4 μg/mL HLCL-61 with different time. The treated cells were subjected to Western blot with indicated antibodies. (C) Flag-VP1 was transfected into DF-1 cells for 36 h, and the transfected cells were treated with 4 μg/mL HLCL-61 for addition 6 h before harvesting cells. The cells were subjected to IP and Western blot with indicated antibodies. (D to F) The DF-1 cells were treated with 4 μg/mL HLCL-61 for 6 h and then infected with 1 MOI IBDV for additional different time. Total viral RNA expression (D) and major viral proteins (VP1/2/3) expression (E) were analyzed by RT-qPCR and Western blotting, respectively. (F) Comparison of the expression of VP1/2/3 in different lanes in panel E. The density of VP1/2/3 and β-actin was quantified by ImageJ software. (G) The DF-1 cells were treated with 4 μg/mL HLCL-61 for 6 h and then infected with 0.1 MOI IBDV for additional different time. Viral titers of whole cells were examined by TCID₅₀ assay. Western blot data were semiquantified and normalized against β-actin protein loading control. The data were obtained from three independent experiments and presented as means ± SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001. DMSO, dimethyl sulfoxide; OD, optical density.

PRMT5 knockout inhibits viral replication.

To better understand the role of PRMT5 in IBDV replication, PRMT5 deletion DF-1 cells were prepared. Two single guide RNAs (sgRNAs) targeting exon3 of chicken PRMT5 were designed to generate the PRMT5 knockout cells. T7E1 digestion showed that sgRNA1, rather than sgRNA2, exhibited the high editing efficiency (Fig. 5A, lane 2). Four lines of monoclonal cells were subjected to Western blot, and we observed that PRMT5 expression in #1 and #2 monoclonal cells was completely abolished (Fig. 5B, lanes 2 and 3). Sequencing of PRMT5 gene showed that there was a 14-bp deletion in the genome of PRMT5 knockout cells compared with the genome of the WT cells (Fig. 5C). The deletion of PRMT5 in #1 cells (named “prmt5−/−”) was stable even after 10 passages (Fig. 5D, lane 2). Next, we used the prmt5−/− cells to examine the role of PRMT5 in viral replication. The viral genome copy in prmt5−/− cells was reduced nearly 6-fold at both 6 and 12 hpi in comparison with that of WT cells (Fig. 5E). In addition, the Western blot analysis demonstrated that the expression of major viral protein (VP1, VP2, and VP3) in prmt5−/− cells was also significantly lower than in WT cells (Fig. 5F and G). A 50% tissue culture infective dose (TCID₅₀) assay also showed that the viral titers in prmt5−/− cells were decreased at least 10- to 100-fold in comparison with that in WT cells (Fig. 5H). Altogether, our data suggest that PRMT5 deletion significantly diminishes the IBDV replication, indicating that PRMT5 is essential for viral replication.

FIG 5.

Deletion of PRMT5 inhibits viral proliferation. (A) Digestion analysis of the PCR products of the target for single guide RNA 1 (sgRNA1) and sgRNA2 using T7E1 assay. Following electrophoresis on 1.5% agarose gel, the digestions were imaged using a Bio-Rad XR+ Imagelab (Bio-Rad, USA). (B) Lysates of four monoclonal chPRMT5 knockout DF-1 cells and WT cells were analyzed by Western blotting using PRMT5 antibodies. (C) PCR products of PRMT5 knockout cells were inserted into pMD18T. Sequence alignment of PCR products of WT PRMT5 genome and PRMT5 knockout genome. (D) PRMT5 deletion cells (clone 1) were passaged 10 times, and then, the cell lysates were subjected to Western blot using PRMT5 antibodies. (E to G) WT and prmt5−/− DF-1 cells were infected with 1 MOI IBDV for different time. Viral genome (E) and major viral proteins VP1/2/3 level (F) were analyzed by RT-qPCR and Western blotting, respectively. (G) Comparison of the expression of VP1/2/3 in different lanes in panel F. The density of VP1/2/3 and β-actin was quantified by ImageJ software. (H) WT and prmt5−/− DF-1 cells were infected with 0.1 MOI IBDV for different time. Viral titers of whole cells were determined by TCID₅₀ assay. Western blot data were semiquantified and normalized against β-actin protein loading control. The data were obtained from three independent experiments and presented as means ± SD. ns, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001.

R426 residue of VP1 is the sDMA site.

PRMT5 is the class II protein arginine methyltransferase that is responsible for sDMA (40). In order to confirm that the sDMA of VP1 is mediated by PRMT5, Myc-PRMT1/3/4/5/7 derived from chicken were cotransfected with Flag-VP1 into 293T cells. IP and immunoblot analysis showed that the sDMA of VP1 was carried out only by PRMT5, and other PRMTs could not modify the VP1 protein (Fig. 6A, lane 5). In order to identify the exact arginine residue in VP1 protein that is methylated by the PRMT5, we cotransfected Flag-VP1 and Myc-PRMT5 into 293T cells and then purified the Flag-VP1 by immunoprecipitation and subjected it to LC-MS analysis (Fig. 6B). Arginine 7 (R7) (R426 residue of VP1) of peptide GEANCTRQHMQAAMYYILTR was identified with dimethylation with high confidence (Fig. 6C and D). Next, we examined whether the sDMA of VP1 on R426 is indeed mediated by PRMT5. Indicated plasmids were transfected into 293T cells. IP and Western blot analysis showed that WT PRMT5, not R375A mutant PRMT5, promoted the sDMA of VP1 (Fig. 6E, lanes 2 and 3), R426A mutant VP1 did not show the sDMA signal, even in the presence of WT PRMT5 (Fig. 6E, lanes 4 and 5). Furthermore, sDMA modification of VP1 was completely absent in prmt5−/− cells in the presence of R375A mutant PRMT5; however, sDMA modification was restored when wild-type (WT) PRMT5 was expressed (Fig. 6F, lanes 3 to 5). In vitro methylation analysis showed that WT PRMT5, rather than the R375A mutant PRMT5, causes the sDMA of VP1 at R426 residue (Fig. 6G, lane 4). RNA-dependent RNA polymerase of IBDV contains seven classical motifs, and each of them plays a critical role in the polymerase activity (45). R426 is located in motif A and is responsible for viral RNA entering during RNA replication (46). Additionally, the R426 residue is highly conserved among different IBDV strains, including the virulent, attenuated, and vaccine strains, of which the sequences are available in GenBank (Fig. 6H). Taken together, our results suggest that sDMA modification of VP1 at R426 is mediated by PRMT5, which might be crucial for polymerase activity and viral replication.

FIG 6.

Arginine 426 (R426) of VP1 is the symmetric dimethylation (sDMA) site mediated by PRMT5. (A) Indicated plasmids were cotransfected into 293T cells for 36 h. The transfected cells were subjected to IP. Immunoblotting of IP complex and cell lysates of the transfected cells using the corresponding antibodies. (B) DF-1 cells were cotransfected with Flag-VP1 and Myc-PRMT5 for 36 h. Purification of Flag-VP1 from the transfected cells. Purified Flag-VP1 were separated by SDS-PAGE and visualized by CBB staining. The specific Flag-VP1 band was excised, digested, and analyzed by liquid chromatography-mass spectrometry (LC-MS). (C) The information of identified peptide by LC-MS described in panel B. (D) The MS spectra of identified VP1 peptide, GEANCTRQHMQAAMYYILTR. (E) 293T cells were transfected with indicated plasmids for 36 h. Then, the cells were subjected to IP by anti-Flag agarose. Immunoblotting of IP complex and cell lysates was performed using corresponding antibodies. (F) WT and prmt5−/− DF-1 cells were transfected with plasmids for 36 h. Then, the cells were subjected to IP by anti-Flag agarose. Immunoblotting of IP complex and cell lysates was performed using corresponding antibodies. (G) The indicated proteins were mixed for arginine methylation reaction in vitro, which were subsequently to analyze by Western blotting with the indicated antibodies. (H) The diagram of architectures of VP1 polymerase. R426 (Blue), located in motif A, is highly conserved among vv IBDV, attenuated IBDV and vaccine IBDV. The black arrows indicate two Mg²⁺-binding sites that are essential for polymerase activity.

Symmetric dimethylation of VP1 contributes polymerase activity.

The location of R426 in motif A prompted us to investigate the effect of arginine methylation of VP1 on its polymerase activity. Here, transfection of different combination of VP1, VP3, and pUC-mA plasmids was used to investigate whether R426 arginine methylation affects the VP1 polymerase activity that is required for viral genomic RNA translation initiation. The pUC-mA plasmid, producing the minus-sense segment A transcript after transfection, along with VP1 and VP3 plasmids were applied to analyze the polymerase activity via detecting level of VP2 expression (Fig. 7A and B) (47). Only the WT VP1, not the D402 and D416 mutant VP1, also called the “polymerase dead mutants,” could efficiently induce the expression of VP2 encoded by segment A (Fig. 7B, lane 2) (9). The VP2 expression was significantly reduced with ~60% in prmt5−/− cells compared with that in WT cells (Fig. 7C, lane 3 versus lane 2). This reduction was restored by the ectopic expression of WT PRMT5, not R375A mutant PRMT5, in prmt5−/− cells (Fig. 7C, lanes 4 and 5 versus lane 2). In comparison to the WT VP1, the R426A mutant VP1 causes ~70% reduced expression of VP2, which was not restored by the overexpression of PRMT5 (Fig. 7D, lane 5 versus lanes 2 and 6). Additionally, the PRMT5 inhibitor HLCL-61 also decreased the expression of VP1 in a dose-dependent manner (Fig. 7E), suggesting that HLCL-61 significantly inhibited VP1 polymerase activity. Next, we expressed the WT, as well as the R375A mutant PRMT5, in prmt5−/− cells to determine whether the supporting of viral IBDV genome replication depends on the methyltransferase activity of PRMT5. Consistent with the result of Fig. 3 (G to J), Western blot analysis demonstrated that overexpression of WT PRMT5, not R375A mutant PRMT5, promotes the expression of major viral proteins (VP1/VP2/VP3) in WT cells (Fig. 7F, lanes 3 and 4 versus lane 2). Furthermore, ectopic expression of WT PRMT5, not the R375A mutant PRMT5, restored the reduced expression of major viral proteins in prmt5−/− cells (Fig. 7F, lanes 6 and 7 versus lane 2). These results were in agreement with the RT-qPCR data showed in Fig. 7G, suggesting that the positive effect of PRMT5 on viral replication depends on its methyltransferase activity. Taken together, our data demonstrate that PRMT5-mediated methylation of R416 residue of VP1 supports its polymerase activity.

FIG 7.

Arginine methylation of VP1 contributes its polymerase activity. (A) The evaluation system of polymerase activity. (B) VP1 (or D402 and D416 mutant VP1), VP3, and pUC-mA-expressing plasmids were transfected into 293T cells for 72 h. The cell lysates were subjected to analyze VP2, VP1 and VP3 by Western blotting. (C) WT and prmt5−/− DF1-1 cells were transfected with indicated expressing plasmids for 72 h. (Top panel) Immunoblotting of cell lysates was performed by Western blotting with indicated antibodies. (Bottom panel) Comparison of the expression of VP2 in different lanes in the top panel. The density of VP2 and β-actin was quantified by ImageJ software. (D) 293T cells were transfected with indicated expressing plasmids for 72 h. (Top panel) Immunoblotting of cell lysates was performed by Western blotting with the indicated antibodies. (Bottom panel) Comparison of the expression of VP2 in different lanes in the top panel. The density of VP2 and β-actin was quantified by ImageJ software. (E) 293T cells were transfected with indicated expressing plasmids for 72 h. Different concentrations of and 4 μg/mL HLCL-61-treated the transfected cells for 6 h before collecting cells. (Top panel) Immunoblotting of cell lysates was performed by Western blotting with indicated antibodies. (Bottom panel) Comparison of the expression of VP2 in different lanes in the top panel. (F, G) WT DF-1 and prmt5−/− DF-1 cells were transfected with indicated plasmids for 12 h. Then, the transfected cells were infected with 1 MOI IBDV for another 12 h. Major viral proteins, including VP1, VP2, and VP3, were analyzed by Western blotting (F), and the viral genome was analyzed by RT-qPCR (G). Western blot data were semiquantified and normalized against β-actin protein loading control. The data were obtained from three independent experiments and presented as means ± SD. **, P < 0.01; ***, P < 0.001. CMV, cytomegalovirus.

R426A mutant IBDV disrupts virus multiplication.

Mutant virus rescue is a useful technique to study the role of important residue in virus multiplication (48). For this, a high efficiency IBDV rescue system with dual promoters was established to generate R426A mutant IBDV (Fig. 8A) (47). A clear cytopathic effect (CPE) was observed in the DF-1 cells infected with supernatants of 293T cells transfected with pCMV-mA and pCMV-mB (Fig. 8B, panel b) or pCMV-mB R426A (Fig. 8B, panel c). The rescued IBDVs were further confirmed by the immunofluorescence assay (IFA) with VP3 antibodies (Fig. 8B, panels e and f), indicating that R426A mutant IBDV was rescued successfully. Since IBDV VP1 is the polymerase protein that affects the efficiency of viral rescue, we found that R426A mutant VP1 severely disrupted the viral rescue (Fig. 8C). Next, the R426A mutant IBDV was also used to assess the viral replication. RT-qPCR analysis showed that in comparison with the WT IBDV, the viral RNA copy number of R426A mutant IBDV was reduced (Fig. 8D). Consistent with the RT-qPCR results, Western blot detection also demonstrated that the expression of major viral proteins were decreased at 6, 12, and even 18 hpi in R426A mutant IBDV compared with the WT IBDV (Fig. 8E, lanes 5 to 7 versus lanes 2 to 4, and Fig. 8F). Consistent with the RT-qPCR and Western blot results, there was a significant reduction of viral titer of R426A mutant IBDV compared to the WT IBDV (Fig. 8G). Altogether, these data suggest that the R426 residue of VP1 is essential for viral replication.

FIG 8.

R426 mutant reduces viral replication. (A) Schematic model of rescuing IBDV. The system of viral rescuing with dual promoters was established in previous work. (B) Clear cytopathic effect (CPE) images of DF-1 cells infected (or mock infected) with 1 MOI WT IBDV or R426A IBDV for 36 h. Bars, 100 μm (panels a to c). DF-1 cells infected (or mock infected) with 1 MOI WT IBDV or R426A IBDV for 12 h were detected with anti-VP3 (green) antibodies. The nucleus were stained by DAPI. The images were scanned by confocal microscopy with 20-μm bars (panels d to f). (C) The IBDV titers in the supernatant of 293T cells transfected with indicated plasmids were examined by TCID₅₀ assay at 72 and 96 h of post-transfection. (D to E) DF-1 cells were infected with WT IBDV and R426A mutant IBDV (1 MOI each) for different time. Viral genome level was determined by RT-qPCR. The results were normalized to GAPDH mRNA in the same samples. (E) The cell lysates were subjected to Western blot with anti-VP1, anti-VP2, and anti-VP3 antibodies. The β-actin was used as the loading control. (F) Comparison of the expression of VP1/2/3 in different lanes in panel E. The density of VP1/2/3 and β-actin was quantified by ImageJ software. (G) DF-1 cells were infected with 0.1 MOI WT IBDV or R426A IBDV for different times. Viral titers of supernatants of DF-1 cells were determined by TCID₅₀ assay. Western blot data were semiquantified and normalized against β-actin protein loading control. The data were obtained from three independent experiments and presented as means ± SD. **, P < 0.01; ***, P < 0.001.

DISCUSSION

VP1 is the polymerase protein of IBDV that is responsible for viral genome replication (49). Post-translational modification of VP1 plays a key role in regulating its polymerase activity (17). In this report, we have provided multiple lines of evidence such as LC-MS analysis, ectopic expression, panel of inhibitor, gene knockout, and specific arginine methylation site mutant and showed that cellular PRMT5 can carry out the symmetric dimethylation of VP1 at R426 both in vivo and in vitro. First, IP coupled with LC-MS identified that PRMT5 interacts with VP1, which was further validated by co-IP, GST pulldown, and confocal microscopy analysis. Second, we showed the critical role of PRMT5 during viral replication by PRMT5 overexpression and knockout, as well as the PRMT5 specific inhibitor treatment. Third, LC-MS analysis of purified VP1 combined with methylation reaction in vitro showed that the arginine 426 (R426) residue of VP1 was the substrate of PRMT5. However, the percentage of R426 methylation that occurred is unclear. Generation of sDMA-R426 antibody could resolve this important problem. Fourth, polymerase reaction measurement analysis showed that PRMT5-mediated R426 methylation was crucial for the polymerase activity of VP1. Finally, we also showed that R426A mutant recombinant IBDV had significantly decreased viral replication compared with the WT IBDV. This is the first report that IBDV VP1 protein is the substrate for arginine methylation of host PRMT5.

Proteins may undergo multiple types of post-translational modification, such as phosphorylation, acetylation, ubiquitination, and SUMOylation (50). Arginine methylation involves the formation of covalent bond between methyl groups to the side chain of arginine residues in the substrate protein. Arginine methylation majorly includes three types: mono, symmetric, and asymmetric (51). This type of modification often takes place on histone proteins and nonhistone proteins, especially on RNA-binding proteins, which contain canonical GAR, RGG, RG, and RXR motifs (52, 53). Our data showed that the R426 of VP1 did not conform these consensus motifs. PRMT5 is a type II methyltransferase that catalyzes the formation of symmetric dimethylarginine in a variety of substrates at multiple sites, which are not the constituents of the above-mentioned consensus sequences (54). Additionally, PRMT5 is a critical cellular protein that affect multiple functions through arginine methylation (55). Loss of PRMT5 might affect cellular signaling transduction, which then further regulates viral replication. However, IBDV VP1 is the target of PRMT5. Deletion of PRMT5 significantly disrupted virus replication. This inhibitory role of PRMT5 in virus replication was also verified in other virus, such as human immunodeficiency virus type 1 (HIV-1), Epstein Barr virus (EBV), and hepatitis B virus (HBV) (56–58). We could not determine whether PRMT5 affects virus replication via regulating arginine methylation of host proteins or only via affecting VP1 arginine methylation. The effecting cellular proteins on virus replication in PRMT5-deficient cells need to be further explored. In addition, during the process of IBDV replication, there are two forms of VP1, free VP1 and dsRNA-conjugated VP1 (VPg), in the host cells. Purified Flag-VP1 could be detected with sDMA signal, suggesting that at least free VP1 could be modified by arginine methylation. However, whether VPg contained this methylation is unclear. We need sDMA-R426 antibody to resolve this issue. In addition, both free VP1 and VPg could be enclosed into matured virion. Whether PRMT5 and methylated VP1 could be also enclosed into matured virion must be further investigated.

Virus-encoded RNA-dependent RNA polymerases (RdRps) interact with other auxiliary viral and host proteins to form virus replication complexes (59). Most RdRp sequence motifs include G-F-C-A-B-D-E, in which A, B, and C are the most conserved (Fig. 6H) (46). Typically, these motifs are in A-B-C order, which is characteristic of template-dependent palm-based polymerase. Nevertheless, 3D structure analysis showed that in RdRp of IBDV, as well as in other members of the Birnaviridae family, this motif was arranged in C-A-B order. Previous work showed that the predominant functional role of three highly conserved residues (D402 and N403 in motif C and D416 in motif A) are pairing of metal ions, which is absolutely essential for polymerase function (46). R426, also located in the motif A, is close to D416, which might be critical for polymerase activity. Our data showed that R426 mutation severely disrupted the VP1-mediated polymerase activity, as well the viral genome replication, but this significant reduction of polymerase activity in R426A mutant VP1 might not be solely due to the lack of arginine methylation. R426 and the entire motif A are not located in the surface of polymerase, but R426 is highly closed with the Mg²⁺ binding sites (D416 and E421) (Fig. 6H) (46). R426A might not only lack of methylation but also affect the Mg²⁺ binding ability of VP1, thereby regulating polymerase activity. Thus, we speculate that the mutation of this residue may affect the function of motif A directly, which then decreases the function of polymerase activity. This assumption requires to be further exploration. At least, the essential role of R426 needs its arginine methylation. Although the region of VP1-associated PRMT5 was not determined in this study, methylated site R426 might not be essential for interaction between VP1 and PRMT5 because R375A mutant PRMT5 did not affect their interaction (Fig. 1E). Thus, their interaction might not affect the polymerase activity, but R426 methylation is critical for polymerase activity. We postulate that IBDV might hijack host methyltransferase for methylation of VP1 at its R426 residue, which is necessary for the IBDV replication. Therefore, this study is the first report showing the crucial role of R426 methylation of VP1 in motif A for its polymerase activity.

In summary, we have identified the first viral polymerase as the substrate of host PRMT5 and have shown that the IBDV VP1 protein activity is regulated by its methylation. This is also the first post-translational modification of IBDV protein of its kind. Further research on the PRMT5-mediated VP1 methylation may lead to new insights in the regulation of IBDV replication and new potential targets for IBDV therapeutics. Indeed, small-molecule regulators of PRMT activity have recently been identified (60). Thus, compounds that activate PRMT5 or inhibit PRMT5 could function as a supporter or blocker of IBDV replication. These findings not only increase our understanding of the role of VP1 arginine methylation in IBDV replication but also help identify novel antiviral targets to suppress the virus replication.

MATERIALS AND METHODS

Cell culture and virus.

Chicken fibroblast line DF-1 cells (IM-C027, IMMOCELL) were maintained in minimal essential medium (MEM) containing 10% fetal bovine serum (FBS) (16000-044, Gibco, Carlsbad, CA). 293T cells (ATCC, CRL-11268) were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco) supplemented with 10% FBS (FSP500; ExCell Bio, Uruguay). All cell lines were cultured at 37°C supplemented with 5% CO₂. IBDV strain JX was isolated from chicken bursal infected with IBDV and stored in our laboratory. DF-1 cells were used to amplify IBDV, and we evaluated the viral titers by TCID₅₀ assay.

Antibodies, reagents, and plasmids construction.

Mouse Anti-Flag monoclonal antibody (MAb) (H3663) was purchased from Sigma-Aldrich (St. Louis, MO). Mouse anti-Myc (M200212F) MAb, anti-β-actin (M20009F) MAb, anti-His (M3011) MAb, anti-GST (M20007) MAb, rabbit anti-PRMT5 polyclonal antibody (pAb) (T55454), and anti-Flag agarose (M20012F) were purchased from Abmart (Shanghai, China). Rabbit or mouse anti-VP1, VP2, and VP3 pAb were all generated from the serum of rabbit or mice by immunization with purified protein, respectively. Symmetric di-methyl arginine motif rabbit MAb (catalog No. 13222) was purchased from Cell Signaling Technology. Subcellular extraction kits (P0027) were purchased from Beyotime (Shanghai, China). PRMT5 inhibitor HLCL-61 (HY-100025A) was purchased from MedchemExpress. Horseradish peroxidase (HRP)-labeled anti-mouse and anti-rabbit (IgG) antibodies were purchased from KPL (Millford, MA, USA). Cell lysis buffer NP-40 (50 mM Tris [pH 7.4], 150 mM NaCl, 1% NP-40) was purchased from Beyotime (P0013F). Immunofluorescence secondary antibodies, including fluorescein isothiocyanate (FITC)-labeled goat anti-rabbit antibody (A0562) and Alexa Fluor 555-conjugated anti-mouse (A0460) antibody, were purchased from Beyotime. Exfect transfection reagent (T101-01/02) were purchased from Vazyme Biotechnology (Nanjing, China). pUC-mA, pCMV-mA, pCMV-mB, Flag-VP1, pCDNA3.0-VP1, and pCDNA3.0-VP3 were stored in our laboratory. Chicken PRMT1-7-expressing plasmids were constructed by Transship Biotechnology (Shanghai, China). The kinase-dead mutant PRMT5R375A and VP1 mutant plasmids were generated by site-directed mutagenesis assay. All recombinant plasmids were confirmed via Sanger sequencing. All plasmids DNA were transfected into cells using Exfect transfection reagent based on the protocols.

Reverse transcription and qRT-PCR.

The total RNA was extracted from the cells using the TRIzol reagent (Beyotime) based on the manufacturer’s instructions. 1 μg RNA was performed to reverse transcription by using SynScript III cDNA synthesis mix (TSK322S, Tsingke Biotech, Beijing, China). The amplification of the target genes (prmt5 and viral genome) (Table 1) was used to analyze the transcript abundance. The qRT-PCR was performed using 2×TSINGKE Master qPCR mix (SYBR green I) (TSE20, Tsingke Biotech) in QuantStudio 7 flex real-time PCR Detection System (ABI7900, Applied Biosystems, USA). The primer sequences for qRT-PCR are listed in Table 1. Gene fold change was determined by the comparative cycle threshold (CT) (2^−ΔΔCT) analysis.

TABLE 1.

Primers used in this study

Primers	Sequences
chPRMT5F	GATTGCTCAAGCTGGAGGTG
chPRMT5R	CTCCAGGTACTGCAGGTAGG
SgRNA1	GGGAATCGGGTCGGATCCAT
SgRNA2	CGCTTCCGAGTTGCGCCTCA
IBDVF	CCTCTGGGAGTCACGAATTAAC
IBDVR	ACTCATGGTGGCAGAATCATC
chGAPDHF	CCCAGCAACATGAAATGGGCAGAT
chGAPDHR	TGATAACACGCTTAGCACC

Immunofluorescence confocal laser scanning microscopy.

Immunofluorescence was performed as described in prior work (17). Briefly, the DF-1 cells were fixed with 4% paraformaldehyde and penetrated with 0.2% Triton X-100. The fixed cells were subsequently incubated with primary antibodies for 2 h at room temperature. The cells were subjected to stain with secondary antibodies for 1 h at 37°C. Finally, fluorescence signals were scanned using Olympus laser scanning confocal microscope (Olympus Corporation, Tokyo, Japan).

Western blot, Co-IP analysis, and LC-MS.

Whole-cell lysate was extracted by NP-40 buffer on ice and subsequently centrifuged at 12,000 rpm/min for 10 min. The protein samples were loaded into SDS-PAGE gels and transferred onto nitrocellulose membranes (GE Healthcare, USA). The transferred membranes were blocked by 5% skim milk for 30 min at room temperature and then incubated with the indicated primary antibodies at 4°C for 12 h, followed by incubation with the corresponding secondary antibodies at room temperature for 1 h. After incubation with enhanced chemiluminescence (ECL) reagent, the signals of the membranes were scanned by AMERSH Amersham ImageQuant 800 (AI800) (GE Healthcare, USA).

For the coimmunoprecipitation (co-IP) analysis, cell lysates supernatant was incubated with the indicated antibodies at 4°C for 4 h. The protein-antibody complex was captured with protein A/G-agarose (P2028, Beyotime) at 4°C for 2 h. The IP complex was obtained by centrifugation at 1,500 × g for 5 min and washed five times with cold NP-40 buffer, followed by SDS-PAGE and Western blotting.

For LC-MS analysis, NP-40 buffer was used to extract the cells transfected with Flag-VP1. The whole lysates were incubated with anti-Flag agarose at 4°C for 4 h. Then, the IP complexes were obtained by centrifugation at 1,500 × g for 5 min and washed five times. After that, the protein samples were subjected to SDS-PAGE, and Coommassie Brilliant Blue staining (CBB) was performed to visualize differential proteins that were delivered to liquid chromatography-tandem mass spectrometry (LC-MS) analysis (Applied Protein Technology, Shanghai, China) to identify the interacting proteins and arginine modification site.

GST pulldown.

100 ng GST-VP1 proteins were individually added to incubate with the purified His-PRMT5 (or R375A) proteins for 4 h at 4°C. After that, 100 μL of glutathione-agarose beads was added to capture the bait proteins. After washing with cold phosphate-buffered saline (PBS) three times (5 min each time), the beads were boiled with 4× SDS loading buffer. The protein samples were subjected to SDS-PAGE and immunoblotted with anti-GST and anti-His antibodies.

In vitro methylation analysis.

Methylation assay reactions in vitro were performed in 20 μL of methylation reaction buffer (20 mM Tris-HCl, pH 8.0, 200 mM NaCl, 0.4 mM EDTA). This reaction mixture contained 200 ng VP1 or R426A mutant VP1, 100 ng His-PRMT5, and 1 μg SAM, which were incubated at 30°C for 30 min and then analyzed by SDS-PAGE and Western blotting.

Generation of PRMT5 DF-1 cells.

CRISPR-Cas9 gene editing system was used to generate prmt5 deletion DF-1 cell lines. A specific single guide RNA (sgRNA) was designed according to the prediction program (crispr.tefor.net) (Table 1). Then, the sgRNA was inserted into lentiCRISPRv2 vector (Addgene catalog No. 52961). The recombinant plasmids were transiently transfected into DF-1 cells for 48 h. The DF-1 cells were selected with 2 μg/mL of puromycin for 7 days. Selected cells were subjected to continued culture, and genomic DNA was extracted to determine the effectiveness of knockout. To create a single cell for stable knockout cell lines, selected DF-1 cells were serially diluted in 96-well plates. Individual cells were subjected to PCR after reaching the size of a well, and the production was inserted into a pMD18T for sequencing. The T7EI digestion assay was performed in accordance with the established methods. T7E1 mixed buffer was briefly annealed at 95°C for 5 min and then touched down at a temperature reduction of 0.1°C/s. After that, T7E1 endonuclease was used to digest the PCR products for 30 min at 37°C. Further analysis of the digested fragments was analyzed using 1.5% agarose gel electrophoresis.

Cellular fractionation.

Mock- or IBDV-infected DF-1 cells were treated with 200 μL buffer A containing protease inhibitor on ice for 10 min. Then, buffer B was added to the vigorous vortex mix with buffer A, which was subjected to centrifugation at 12,000 × g for another 10 min. The nuclear-debris pellet was resuspended in 100 μL nuclear protein extraction buffer for 20 min. Western blot analysis were conducted on the supernatant with anti-PRMT5 and anti-VP1 antibodies. Anti-H3 and anti-β-actin band was detected as nuclear marker and cytoplasmic marker, respectively.

Rescue of mutant IBDV.

The dual-promoter rescue system of IBDV established in prior work (Fig. 8A) (38). The viral rescue plasmids pCMV-mA and pCMV-mB were constructed and maintained in our laboratory. The mutant plasmid pCMV-mBR426A was constructed using s direct mutation assay. The pCMV-mB or pCMV-mBR426A plasmids were individually cotransfected with pCMV-mA into 293T cells for 72 h. The resultant cells were freeze-thawed three times, followed by centrifuging with 12,000 × g for 10 min. The supernatants were transferred into DF-1 cells for culturing addition 48 h, until the cytopathic effect (CPE) was observed. IFA with VP3 antibody was used to confirm whether IBDV rescued successfully.

TCID₅₀ assay.

The 50% tissue culture infective dose (TCID₅₀) assay was used to determine the viral titers. The viral solution was serially diluted 10-fold in MEM containing 2% FBS. Each diluted sample of 100 μL was added to infect DF-1 cells in 96-well plates. Tissue culture wells with cytopathic effects were considered positive.

Statistical analysis.

All data are presented as means ± standard deviation (SD) for each group and analyzed by using GraphPad Prism 5.0. The statistical significance of differences between groups was determined using Student’s t test. P values less than 0.05 are considered statistically significant.