E3 ligase RNF2 inhibits porcine circovirus type 3 replication by targeting its capsid protein for ubiquitination-dependent degradation

ABSTRACT

Porcine circovirus type 3 (PCV3) is closely associated with various diseases, such as the porcine dermatitis, nephropathy syndrome, and multisystemic clinicopathological diseases. PCV3-associated diseases are increasingly recognized as severe diseases in the global swine industry. Ring finger protein 2 (RNF2), an E3 ubiquitin ligase exclusively located in the nucleus, contributes to various biological processes. This ligase interacts with the PCV3 Cap. However, its role in PCV3 replication remains unclear. This study confirmed that the nuclear localization signal domain of the Cap and the RNF2 N-terminal RING domain facilitate the interaction between the Cap and RNF2. Furthermore, RNF2 promoted the binding of K48-linked polyubiquitination chains to lysine at positions 139 and 140 (K139 and K140) of the PCV3 Cap, thereby degrading the Cap. RNF2 knockdown and overexpression increased or decreased PCV3 replication, respectively. Moreover, the RING domain-deleted RNF2 mutant eliminated the RNF2-induced degradation of the PCV3 Cap and RNF2-mediated inhibition of viral replication. This indicates that both processes were associated with its E3 ligase activity. Our findings demonstrate that RNF2 can interact with and degrade the PCV3 Cap via its N-terminal RING domain in a ubiquitination-dependent manner, thereby inhibiting PCV3 replication.

IMPORTANCE

Porcine circovirus type 3 is a recently described pathogen that is prevalent worldwide, causing substantial economic losses to the swine industry. However, the mechanisms through which host proteins regulate its replication remain unclear. Here, we demonstrate that ring finger protein 2 inhibits porcine circovirus type 3 replication by interacting with and degrading the Cap of this pathogen in a ubiquitination-dependent manner, requiring its N-terminal RING domain. Ring finger protein 2-mediated degradation of the Cap relies on its E3 ligase activity and the simultaneous existence of K139 and K140 within the Cap. These findings reveal the mechanism by which this protein interacts with and degrades the Cap to inhibit porcine circovirus type 3 replication. This consequently provides novel insights into porcine circovirus type 3 pathogenesis and facilitates the development of preventative measures against this pathogen.

INTRODUCTION

Porcine circovirus (PCV) is a circular, single-stranded DNA virus belonging to the genus Circovirus within the family Circoviridae. Four PCV genotypes (PCV1, PCV2, PCV3, and PCV4) have been recognized (1–4). PCV3 was initially detected in a case of porcine dermatitis and nephropathy syndrome (PDNS) in the United States and was subsequently reported in multiple countries (3–10). PCV3 has been associated with multiple other clinical conditions, such as reproductive failure, respiratory disease, and multi-organ inflammation (7, 11, 12). The PCV3 genome is 2,000 bp in length and contains three main open-reading frames (ORF): ORF1, ORF2, and ORF3. ORF1 encodes a viral replicase associated with viral replication (Rep). ORF2 encodes a major host-protective structural Cap, whereas ORF3 encodes a protein of unidentified function (13). Inoculation of piglets with PCV3 stock obtained from DNA infectious clones induced clinicopathological signs of PDNS-like disease (14). Obvious lesions in the myocardium and lungs have also been observed in PCV3-infected Kunming mice (11, 15). Isobaric tags for relative and absolute quantification labeling combined with liquid chromatography-tandem mass spectrometric analysis showed that PCV3 infection is primarily associated with metabolic processes, immune responses, MHC-I and MHC-II components, and phagosome pathways (16). Thus, PCV3 has been accepted as an emerging and important disease in major pig-rearing countries and regions worldwide since its first description in the USA in 2017 (3).

Ring finger protein 2 (RNF2), a key component of polycomb repressive complex 1, mediates histone H2A mono-ubiquitination on lysine 119 (K119; H2K119ub), resulting in the repression of target gene transcription (17). H2K119ub is associated with stem cell identity, silencing of several developmental genes, and genomic imprinting (17, 18). In addition, it is closely associated with tumorigenesis and tumor growth. For example, RNF2 can increase H2K119ub levels at the E-cadherin promoter region, which promotes the metastasis of hepatocellular carcinoma via epithelial-mesenchymal transition (19). RNF2 induces the ubiquitination and degradation of IRF4, thereby promoting cell proliferation, migration, and invasion in colon cancer (20). Homeobox (HOX) genes are classical targets of polycomb repressive complex 1, and their regulation is usually related to embryonic development and tumorigenesis (17, 18). The induction of HOX genes may be associated with hepatitis C virus-related diseases (21, 22). Moreover, the hepatitis C virus core protein induces RNF2 degradation, leading to the induction of HOX genes (23). This indicates that RNF2 is involved in viral infection. As a member of the RING-domain E3 ligase family, RNF2 regulates the polyubiquitination of several target proteins such as p53 and AMBRA1. Thus, it participates in tumor development and autophagy (24, 25). In addition, this protein promotes K33-linked polyubiquitination of STAT1 and subsequently induces the disassociation of STAT1/STAT2 from DNA, which inhibits the interferon-dependent antiviral response (26). RNF2 can inhibit the K48-linked polyubiquitination of estrogen receptor α. This may increase estrogen receptor α stability and facilitate the development of breast cancer (27). Our previous studies revealed that RNF2 interacts with the PCV3 Cap (28, 29). However, the relationship between RNF2 and this Cap and whether RNF2 regulates PCV3 replication in cultured cells remain unclear.

This study confirms that RNF2 interacts with the PCV3 Cap and promotes its degradation in a ubiquitin-dependent manner. This study further indicated that lysine residues 139 and 140 of the Cap are the corresponding ubiquitination sites. Moreover, the E3 ligase activity of RNF2 was identified as a crucial component of the restriction of RNF2-induced PCV3 Cap expression. These findings indicate that RNF2 can effectively inhibit PCV3 replication, which relies on E3 ligase activity. This study reveals the mechanism through which RNF2 degrades the Cap to inhibit PCV3 replication, thereby providing novel insights into PCV3 pathogenesis.

RESULTS

RNF2 inhibits PCV3 replication

Our recent study showed that the PCV3 Cap can recruit E3 ligases to degrade target proteins, which facilitates viral replication (28). To verify whether E3 ligases have a direct effect on Cap expression, we first selected three different E3 ligases from the top-ranking protein-protein interaction network map. We then used western blotting to determine PCV3 Cap expression in HEK-293T cells co-transfected with green fluorescent protein (GFP)-tagged RNF2, RNF34, TRIML2 (GFP-RNF2, GFP-RNF34, and GFP-TRIML2), and FLAG-tagged PCV3 Cap (FLAG-Cap). As shown in Fig. 1A, Cap expression was considerably reduced in the presence of RNF2 but not in that of RNF34 and TRIML2. Furthermore, RNF2 reduced the expression of the PCV3 Cap in a dose-dependent manner following co-transfection with the GFP-RNF2 and FLAG-Cap plasmids into HEK-293T cells (Fig. 1B). Additionally, quantitative real time PCR (qRT-PCR) and western blot analyses were used to detect the levels of RNF2 mRNA and protein during PCV3 replication in PK-15 cells, respectively. RNF2 protein levels steadily decreased during PCV3 replication, whereas the mRNA levels did not change (Fig. 1C and D). To further evaluate the effect of PCV3 infection on the RNF2 expression, we detected the levels of RNF2 protein in PK-15 cells infected with UV-inactivated PCV3 which has no infectivity. As presented in Fig. 1E, UV-inactivated PCV3 had no effect on RNF2 expression compared with PCV3, indicating that the inhibition role of PCV3 in RNF2 expression depends on its infectivity. As RNF2 expression was suppressed by PCV3 and RNF2 can degrade the PCV3 Cap, we explored the effect of RNF2 expression on PCV3 replication. Transfection of small interfering RNA (siRNA)-targeting (siRNA-RNF2) or FLAG-tagged (FLAG-RNF2) RNF2 plasmids into PK-15 cells was followed by infection with PCV3. Expression levels of the PCV3 Rep protein were enhanced in siRNA-RNF2 and PCV3-infected cells, whereas western blotting revealed decreased levels of Rep protein expression in FLAG-RNF2 and PCV3-infected cells (Fig. 1F through H). This was further verified using a viral titer assay, which showed that RNF2 negatively regulates PCV3 replication (Fig. 1I). These results above suggested that RNF2 inhibits PCV3 replication.

Fig 1.

RNF2 negatively regulated PCV3 replication. (A) HEK-293T cells expressing GFP-RNF2, GFP-RNF34, GFP-TRIML2, or FLAG-Cap were detected using anti-FLAG, GFP, or β-actin antibodies. (B) HEK-293T cells were co-transfected with FLAG-Cap and increasing amounts of GFP-RNF2 plasmids and then detected with anti-GFP, anti-FLAG, or anti-β-actin antibodies. The ratios of FLAG-Cap to β-actin were analyzed using Image J and GraphPad software in panels A and B. (C) PK-15 cells infected with PCV3 for 0, 12, 24, 36, and 48 h, followed by processed and detection with anti-RNF2, anti-Rep, and anti-β-actin antibodies. (D) Relative levels of RNF2 mRNA were measured by qRT-PCR in PCV3-infected cells at 12, 24, 36, and 48 h. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as internal references. (E) The levels of RNF2 and Rep were detected in PK-15 cells infected with PCV3 or UV-inactivated PCV3 for 24 h using anti-RNF2 and anti-Rep antibodies, and β-actin was measured as internal control. (F) PK-15 cells transfected various siRNF2 (20 µM) or control siRNA (NC) for 24 h were detected with anti-RNF2 and anti-β-actin antibodies. (G and H) PK-15 cells transfected with siRNF2 (G) or FLAG-RFN2 plasmids (H) were infected with PCV3 for 36 h and then analyzed RNF2 expression and PCV3 replication level through detecting endogenous, exogenous RNF2, and Rep expression levels. RNF2 and Rep levels were quantified using image J and GraphPad software. (I) The viral titer in PK-15 cells under RNF2 overexpression or knockdown was measured using the 50% tissue culture infective dose (TCID₅₀) assay. Data are expressed as the means ± SD of three independent experiments. *P < 0.05; **P < 0.01.

The nuclear location signal domain of the PCV3 Cap interacts with RNF2

As demonstrated by the protein-protein interaction network analysis, RNF2 interacts with the Cap (29). To further verify whether the PCV3 Cap and RNF2 interact, a co-immunoprecipitation (co-IP) assay was performed on HEK-293T cells transfected with GFP-RNF2 and FLAG-Cap plasmids for 36 h using GFP-Nanoab-Agarose beads or an anti-FLAG affinity resin. As shown in Fig. 2A and B, the PCV3 Cap interacted with RNF2. Confocal microscopy revealed co-localization of RNF2 and the PCV3 Cap in the nucleolus (Fig. 2C). The Cap has a nuclear location signal (NLS) domain at its N-terminal and was co-localized with RNF2 in the nucleolus. To further identify the precise PCV3 Cap region that interacts with RNF2, two mutant plasmids of the GFP-tagged PCV3 Cap (GFP-Cap) that encode the NLS and dCap regions of the Cap were generated (Fig. 2D). Co-IP analysis was conducted on HEK-293T cells containing the GFP-Cap, GFP-Cap-NLS, GFP-Cap-dCap, and FLAG-RNF2 plasmids. The precise region of the PCV3 Cap that interacted with RNF2 was the NLS domain (Fig. 2E).

Fig 2.

NLS domain of PCV3 Cap is responsible for interaction with RNF2. (A and B) HEK-293T cells co-expressed GFP-RNF2 and FLAG-Cap for 36 h were immunoprecipitated with GFP-Nanoab-Agarose beads (A) or anti-FLAG affinity resin (B) and then analyzed with anti-GFP, anti-FLAG, or anti-β-actin antibodies. (C) PK-15 cells co-expressed with Cherry-RNF2 (red signals) and GFP-Cap (green signals) or Cherry-RNF2 and GFP were fixed and stained with 2,4-diamidino-2-phenylindole (DAPI; blue signals), and then observed. Scale bar, 10 µm. (D) Schematic representation in lengths of various truncated PCV3 Cap. (E) HEK-293T cells co-expressed with various truncated PCV3 Cap and Flag-RNF2 were lysed and immunoprecipitated with anti-GFP antibody, followed by analysis by co-IP.

RNF2 facilitates the degradation of the PCV3 Cap

RNF2 inhibited the expression of the PCV3 Cap (Fig. 1A and B). Therefore, we hypothesized that it may mediate the instability of the protein Cap. Cycloheximide (CHX), an inhibitor of protein synthesis, was used to confirm this hypothesis. As shown in Fig. 3A and B, RNF2 accelerated the CHX-mediated degradation of the PCV3 Cap in PK-15 cells. Contrastingly, the CHX-induced degradation of this Cap was inhibited in PK-15 cells when RNF2 expression was suppressed by siRNA (Fig. 3C and D). RNF2 is a RING-domain E3 ligase, and the ubiquitin-proteasome pathway mediates the degradation of substrate proteins. Therefore, we next investigated whether RNF2 can degrade the PCV3 Cap in a ubiquitin-dependent manner. To ascertain whether RNF2 was involved in PCV3 Cap ubiquitination, an immunoprecipitation (IP) assay was performed using an anti-FLAG affinity resin in HEK-293T cells co-transfected with HA-tagged ubiquitin (HA-Ub), GFP-Cap, and FLAG-RNF2 plasmids. Figure 3E shows that RNF2 increases PCV3 Cap ubiquitination. Subsequently, an IP assay was performed on whole HEK-293 cell lysates containing HA-Ub-K48 (with K48 as the only Lys residue), HA-Ub-K63 (with K63 as the only Lys residue), GFP-RNF2, and FLAG-Cap to further verify RNF2-mediated PCV3 Cap ubiquitination. A lower ubiquitination level was observed in Lane 4 of Fig. 3F compared with Lanes 3 and 2, indicating that RNF2 promotes K48-linked ubiquitination of the PCV3 Cap. Finally, the proteasome inhibitor MG132 and the autophagy inhibitor chloroquine were used in PK-15 cells to further verify whether RNF2 degrades the PCV3 Cap through ubiquitination. As shown in Fig. 3G, MG132 treatment, but not chloroquine treatment, restored the inhibition of RNF2-induced PCV3 Cap expression. Moreover, the CHX-mediated degradation of the PCV3 Cap was accelerated in the presence of RNF2 (Fig. 3A), whereas MG132 suppressed the RNF2-accelerated degradation of this Cap (Fig. 3H and I). These results indicated that RNF2 promotes the degradation of the PCV3 Cap via ubiquitination.

Fig 3.

RNF2 degrades PCV3 cap through ubiquitination. (A) PK-15 cells co-expressed with FLAG-Cap and GFP-C1 or GFP-RNF2 for 24 h were treated with CHX for 0, 4, or 8 h and then detected by western blotting. (B) The ratios of FLAG-Cap to β-actin were quantified by Image J and GraphPad software. (C) PK-15 cells treated with 20 µM of NC or siRNA-RNF2 were transfected with FLAG-Cap and treated with CHX, and then detected by western blotting. (D) The ratios of FLAG-Cap to β-actin were quantified by Image J and GraphPad software. (E) HEK-293T cells transfected with HA-Ub, GFP-RNF2, and/or FLAG -Cap were immunoprecipitated with an anti-FLAG affinity resin and then analyzed with various antibodies. (F) HEK-293T cells transfected with HA-Ub and its mutants (HA-Ub-K48 or HA-Ub-K63), GFP-RNF2, and/or, FLAG-Cap were immunoprecipitated with an anti-FLAG affinity resin and then analyzed with various antibodies. (G) HEK-293T cells co-expressed with GFP-C1 or GFP-RNF2 and FLAG-Cap for 24 h were treated with chloroquine (CQ) or MG132 for 6 h and then analyzed by western blotting. (H) PK-15 cells co-expressed with FLAG-Cap and GFP-C1 or GFP-RNF2 for 24 h were treated with CHX for 0, 4, or 8 h in the presence of CQ or MG132 and then detected by western blot. (I) The ratios of FLAG-Cap to β-actin were quantified by Image J and GraphPad software. Data are expressed as the means ± SD of three independent experiments. *P < 0.05; **P < 0.01.

Lysine positions 139 and 140 are important for the RNF2-induced ubiquitination and degradation of the PCV3 Cap

E3 ligases recognize lysine (Lys) residues of the target protein and mediate the binding of ubiquitin to its corresponding Lys position, which is an indispensable step in ubiquitination (30). The Bayesian discriminant method (BDM-PUB) online tool was used to predict the potential ubiquitination sites of the PCV3 Cap. BDM-PUB identified the Lys residues at positions 42, 133, 138, 139, 140, and 211 of the Cap as possible ubiquitination sites (Fig. 4A). To determine the precise ubiquitination site (s), several FLAG-Cap mutant plasmids containing Lys at positions 42, 133, 138, 139, 140, and 211 were replaced by Arg (FLAG-Cap-K42R, FLAG-Cap-K133R, FLAG-Cap-K138R, FLAG-Cap-K139R, FLAG-Cap-K140R, and FLAG-Cap-K211R). Subsequently, western blotting was performed on HEK-293T cells co-transfected with GFP-RNF2 and FLAG-Cap mutant plasmids. As shown in Fig. 4B, the RNF2-decreased expression of the PCV3 Cap was restored when Lys at positions 139 and 140 of the Cap was replaced with Arg. To verify whether the mutant PCV3 Cap (K139R and K140R) interacted with RNF2, a co-IP assay was performed on HEK-293T cells that co-expressed FLAG-RNF2, GFP-Cap, GFP-Cap-K139R, and GFP-Cap-K140R. Specific bands were observed in the presence of GFP-Cap, GFP-Cap-K139R, or GFP-Cap-K140R, indicating that the PCV3 Cap could still interact with RNF2 when Lys was replaced with Arg at positions 139 and 140 (Fig. 4C). We confirmed that RNF2 could facilitate PCV3 Cap degradation (Fig. 3A), and the RNF2-accelerated degradation of the PCV3 Cap was suppressed when Lys at positions 139 and 140 of this Cap was replaced by Arg (Fig. 4D and E). Therefore, we predicted that positions 139 and 140 of the Cap may be ubiquitination sites. To verify this hypothesis, an IP assay was conducted using an anti-FLAG affinity resin in HEK-293T cell lysates containing GFP-RNF2, HA-Ub, and FLAG-Cap mutant plasmids. As shown in Fig. 4F, the RNF2-increased ubiquitination of the PCV3 Cap was inhibited when Lys at position 139 or 140 of the PCV3 Cap was replaced with Arg. These results confirmed that Lys positions 139 and 140 of the PCV3 Cap are important for the RNF2-induced ubiquitination and degradation of this Cap.

Fig 4.

The 139th and 140th lysine residues of PCV3 Cap are responsible for RNF2-mediated PCV3 Cap degradation. (A) Schematic representation of the predicted lysine sites of PCV3 Cap. (B) HEK-293T cells co-expressed with GFP-C1 or GFP-RNF2 and various FLAG-Cap mutants (FLAG-Cap-K42R, FLAG-Cap-K133R, FLAG-Cap-K138R, FLAG-Cap-K139R, FLAG-Cap-K140R, and FLAG-Cap-K211R) for 24 h were analyzed by western blotting. (C) HEK-293T cells co-expressed with GFP-RNF2 and the indicated FLAG-Cap mutant were immunoprecipitated with GFP-Nanoab-Agarose beads and then analyzed by western blotting. (D) PK-15 cells co-expressed with GFP-C1 or GFP-RNF2 and the indicated FLAG-Cap mutants for 24 h were treated with CHX for 0, 4, or 8 h and then analyzed by western blotting. (E) The ratios of FLAG-Cap or its mutants to β-actin were quantified by Image J and GraphPad software. (F) HEK-293T cells were co-expressed GFP-RNF2, HA-Ub, and FLAG-Cap or its mutants, followed by IP with an anti-FLAG affinity resin, and then immunoblotted with anti-HA, anti-GFP, anti-FLAG, and anti-β-actin antibodies. Data are expressed as the means ± SD of three independent experiments. *P < 0.05; **P < 0.01.

The simultaneous existence of K139 and K140 is essential for the RNF2-mediated ubiquitination degradation of the PCV3 Cap

To further determine the role of Lys at positions 139 and 140 of the PCV3 Cap in its ubiquitination and degradation, PCV3 Cap mutant plasmids with Lys at position 139 as the only Lys residue (K139), Lys at position 140 as the only Lys residue (K140), and Lys at positions 139 and 140 were constructed (K139, K140, and K139K140). To analyze the role of K139 and K140 in RNF2-induced Cap degradation, western blotting was used to detect the expression of the PCV3 Cap in HEK-293T cells co-transfected with FLAG-Cap-K139, FLAG-Cap-K140, FLAG-Cap-K139K140, and GFP-RNF2. Decreased Cap expression was observed in Lanes 2 and 8 of Fig. 5A than in Lanes 1 and 7, indicating that K139 and K140 are important for the RNF2-mediated degradation of the PCV3 Cap. To verify whether K139, K140, or K139K140 affected the interaction between RNF2 and the Cap, HEK-293T cells co-transfected with FLAG-Cap-K139, FLAG-Cap-K140, FLAG-Cap-K139K140, or GFP-RNF2 were analyzed using a co-IP assay. As shown in Fig. 5B, wild-type PCV3 Cap, K139, K140, and K130K140 Caps interacted with RNF2. This indicates that K139, K140, and K139K140 did not affect the interaction between the PCV3 Cap and RNF2. Finally, HEK-293T cells were co-transfected with GFP-RNF2, HA-Ub-K48, FLAG-Cap-K139, FLAG-Cap-K140, and FLAG-Cap-K139K140 plasmids. Then, an IP assay was performed on HEK-293T cells using an anti-FLAG affinity resin to verify the effect of K139, K140, and K139K140 on RNF2-induced PCV3 Cap K48-linked ubiquitination. Figure 5C shows the enhanced K48-linked ubiquitination of FLAG-Cap and FLAG-Cap-K139K140 by RNF2. These results showed that the RNF2-induced ubiquitination degradation of the PCV3 Cap requires the simultaneous presence of K139 and K140.

Fig 5.

The simultaneous existence of K139 and K140 in PCV3 Cap is essential for the RNF2-mediated ubiquitination degradation of PCV3 Cap. (A) HEK-293T cells were co-expressed with GFP-C1 or GFP-RNF2 and FLAG-Cap or its mutants (FLAG-Cap-K139, FLAG-Cap-K140, and FLAG-Cap-K139K140) for 24 h, followed by detection with various antibodies. (B) HEK-293T cells co-expressed with GFP-RNF2 and FLAG-Cap or its mutants were immunoprecipitated with an anti-FLAG affinity resin and detected by western blotting. (C) HEK-293T cells were co-expressed HA-Ub-K48, FLAG-Cap and its mutant plasmids, or GFP-RNF2, followed by IP with an anti-FLAG affinity resin, and then detected with various antibodies.

RNF2 E3 ligase activity is necessary for the RNF2-mediated ubiquitination and degradation of the PCV3 Cap and the inhibition of PCV3 replication

We generated different RNF2 mutations containing a deficiency in the C-terminal, N-terminal, or RING domain to identify the region involved in the ubiquitination and degradation of the PCV3 Cap (Fig. 6A). The corresponding RNF2 mutant and FLAG-Cap plasmids were co-transfected into HEK-293T cells. Next, western blotting was performed to determine the effect of different RNF2 mutations on PCV3 Cap expression. The absence of the RING domain (GFP-RNF2-ΔRING) and deficiency of the N-Terminal of RNF2 (GFP-RNF2-C) had no effect on PCV3 Cap expression (Fig. 6B). Furthermore, we constructed a C51W/C54S double amino acid point mutation (GFP-RNF2-C51W/C54S) in the RING domain of RNF2, which eliminates E3 ligase activity. The Cap expression levels in Lane 6 of Fig. 6B were higher than those in Lane 2, indicating that E3 ligase activity is necessary for RNF2-mediated Cap degradation. Next, we co-transfected HEK-293T cells with the corresponding RNF2 mutant and FLAG-Cap plasmids, and a co-IP assay was used to analyze the role of the RING domain in the interaction between the PCV3 Cap and RNF2. Specific bands were observed in the presence of GFP-RNF2 and GFP-RNF2-N, indicating that the RNF2 RING domain is necessary for the interaction between the PCV3 Cap and RNF2 (Fig. 6C). To verify the role of the RNF2 RING domain in RNF-induced PCV3 Cap ubiquitination, HEK-293T cells were co-transfected with FLAG-Cap, HA-Ub, and the corresponding RNF2 mutant plasmids. IP analysis was subsequently performed using an anti-FLAG affinity resin. As shown in Fig. 6D, RNF2 could not induce ubiquitination of the PCV3 Cap when the RING domain was deleted, or E3 ligase activity was inhibited. This indicates that the E3 ligase activity of RNF2 is necessary for RNF2-induced ubiquitination of the PCV3 Cap. Moreover, the RNF2-accelerated degradation of this Cap in PK-15 cells was inhibited when the RNF2 RING domain was deleted or when E3 ligase activity was abolished. This indicates that the E3 ligase activity of RNF2 is necessary for RNF2-induced degradation of the PCV3 Cap (Fig. 6E and F). These results confirm that RNF2 cannot degrade the PCV3 Cap when its E3 ligase activity is inhibited (Fig. 6B). Additionally, PK-15 cells transfected with RNF2 mutant plasmids were infected with PCV3 to evaluate viral replication. The viral replication was lower in cells expressing GFP-RNF2 and GFP-RNF2-N plasmids compared to that in cells expressing GFP-C1, GFP-RNF2-C, GFP-RNF2-ΔRING plasmids, and GFP-RNF2-C51W/C54S (Fig. 6G). Subsequently, a viral titer assay was used to further verify the effect of the indicated RNF2 mutants on the PCV3 replication (Fig. 6H). These results showed that E3 ligase activity is necessary for the RNF2-mediated ubiquitination and degradation of the PCV3 Cap and inhibition of PCV3 replication.

Fig 6.

The E3 ligase activity of RNF2 plays an important role in RNF2-mediated ubiquitination degradation of PCV3 Cap and inhibition of PCV3 replication. (A) Schematic representation of various RNF2 mutants. (B) HEK-293T cells were co-expressed with FLAG-Cap and various RNF2 mutants and then detected with anti-FLAG, anti-GFP, and anti-β-actin antibodies. (C) HEK-293T cells co-expressed with various RNF2 mutants and FLAG-Cap were immunoprecipitated with GFP-Nanoab-Agarose beads and then analyzed by western blotting. (D) HEK-293T cells co-expressed HA-Ub, GFP-RNF2, and various RNF2 mutants were immunoprecipitated with an anti-FLAG affinity resin and then measured for FLAG, GFP, and β-actin expression. (E) PK-15 cells were co-transfected with FLAG-Cap and the GFP-RNF2 and its mutant plasmids for 24 h, followed by treatment with CHX for 0, 4, or 8 h, and then detected by western blotting. (F) The ratios of FLAG-Cap to β-actin were quantified by Image J and GraphPad software. (G) PK-15 cells transfected with RNF2 and its mutant plasmids for 24 h, followed by infection with PCV3, and then detected with anti-GFP, anti-Rep, and anti-β-actin antibodies. (H) The viral titers in PK-15 cells under overexpression of the indicated RNF2 mutant were measured using the 50% tissue culture infective dose (TCID₅₀) assay. Data are expressed as the means ± SD of three independent experiments. *P < 0.05; **P < 0.01.

DISCUSSION

As an important post-translational modification system, the ubiquitin-proteasome system (UPS) modifies substrate proteins. It mediates various biological processes, such as apoptosis, immune regulation, tumorigenesis, inflammation, and autophagy (20, 26, 31–35). Our previous studies demonstrated that the PCV3 Cap can utilize host proteins, including E3 ligase, to facilitate self-replication (28, 29), indicating that the UPS may be involved in PCV3 replication. The present study confirmed that the nuclear E3 ligase RNF2 inhibits PCV3 replication by degrading the PCV3 Cap via ubiquitination. RNF2-mediated degradation and ubiquitination of the PCV3 Cap rely on the simultaneous presence of Lys positions 139 and 140 in this Cap and on the E3 ligase activity of RNF2.

The UPS is involved in various viral replication processes. For example, the E2 ubiquitin-conjugating enzyme UBE2L6 facilitates Senecavirus A replication by stabilizing the RNA-dependent RNA polymerase of this virus (the 3D protein of Senecavirus A) (36). This is achieved by mediating the ubiquitination of the 3D protein. The influenza A virus PB1 protein can self-replicate by evading innate immunity, which occurs through the recruitment of the E3 ligase RNF5 to promote mitochondrial antiviral signaling protein (MAVS) ubiquitination and degradation (37). The PCV3 Cap self-replicates by promoting the ubiquitinated degradation of nucleolin, suggesting that the UPS may regulate PCV3 replication (28). In the present study, three different E3 ligases in the protein-protein interaction network mapping were selected to explore the relationship between PCV3 replication and the UPS. Here, RNF2 decreased Cap expression (Fig. 1A and B). Furthermore, PCV3 infection affected RNF2 expression, and RNF2 inhibited PCV3 replication (Fig. 1C, F and G). These results suggest that RNF2 inhibits PCV3 replication by affecting PCV3 Cap expression. The recognition of target proteins by E3 ligase is necessary for the ubiquitination of substrates. For example, RNF5 interacts with the severe acute respiratory syndrome coronavirus 2 envelope protein (E) (38). This interaction catalyzes the ubiquitination of E, leading to its degradation. The E3 ligase pRNF114 interacts with and catalyzes the K27-linked polyubiquitination of classical swine fever virus NS4B, leading to its degradation (39). Our results showed that the PCV3 Cap interacted and was colocalized with RNF2 in the nucleolus (Fig. 2A through C). Furthermore, the primary region of the PCV3 Cap that interacted with RNF2 was the NLS region (Fig. 2D and E). In our previous studies, this Cap was colocalized with NPM1, nucleolin, and fibrillarin in the nucleolus, and its NLS was the primary region that interacted with NPM1, the nucleolin, and fibrillarin (28, 29). The present study further emphasizes the importance of the NLS region of the PCV3 Cap in the interaction between this Cap and host proteins. The UPS involves three steps: ubiquitin activation, ubiquitin conjugation, and ubiquitin-protein ligation. Here, the ubiquitinated levels of the PCV3 Cap increased in the presence of RNF2. In addition, the proteasome inhibitor MG132 prevented RNF2-mediated degradation of the Cap (Fig. 3G through I). These results further illustrate that RNF2 induces PCV3 Cap degradation through ubiquitination.

Seven Lys residues (K6, K11, K27, K29, K33, K48, and K63) are found in ubiquitin molecules. These residues combine to form ubiquitin chains on substrate proteins (40–42). K48-linked ubiquitination of target proteins usually degrades them, whereas K63-linked ubiquitination of target proteins stabilizes or activates them (43–46). Our findings demonstrate that RNF2 induces the K48-linked polyubiquitination and consequent degradation of the PCV3 Cap (Fig. 3F). The Lys residue of substrate proteins is important for ubiquitin-protein ligation because ubiquitin is always recruited to the Lys residue of the substrate protein. For example, the avian influenza virus (AIV) PA protein promotes the ubiquitination and subsequent degradation of Janus kinase 1 at Lys 249 (47). Similarly, the PB1 protein of the AIV H7N9 virus catalyzes the ubiquitination and subsequent degradation of MAVS at Lys residues 362 and 461 (37). BDM-PUB identified Lys residues at positions 42, 133, 138, 139, 140, and 211 of the PCV3 Cap (K42, K133, K138, K139, K140, and K211, respectively) as potential ubiquitination sites (Fig. 5A). Additionally, mutations in K139 and K140 decreased the RNF2-induced ubiquitination and degradation of the PCV3 Cap (Fig. 4B, D, E and F). In contrast, this process was not affected in the presence of both K139 and K140 in the PCV3 Cap (Fig. 5). Thus, the functions of K139 and K140 in this Cap warrant further elucidation and may provide new targets for PCV3 prevention and treatment.

The RING domain of RNF2 is essential for E3 ligase activity. RNF2 interacts with and enhances the activity of a subunit of the proteasomal 19 S regulatory complex (S6’) through the RING domain (48). Furthermore, this protein inhibits the antiviral response mediated by STAT1 via ubiquitination; E3 ligase activity is necessary for this inhibition (26). Here, we confirmed that the RNF2 RING domain is necessary for the Cap-RNF2 interaction (Fig. 6C). Furthermore, the E3 ligase activity of RNF2 regulates PCV3 Cap degradation through ubiquitination, consequently inhibiting PCV3 replication (Fig. 6B, D and G). Our results are consistent with earlier findings that the RNF2 RING domain is necessary for its function and provide evidence that the E3 ligase activity of this protein regulates viral replication.

This study shows that the nuclear E3 ligase RNF2 can promote K48-linked ubiquitination of the PCV3 Cap at K139 and K140. This degrades the Cap and subsequently inhibits PCV3 replication. In addition, the RING domain of RNF2 is essential for the RNF2-Cap interaction and RNF-mediated ubiquitinated degradation of the PCV3 Cap. These results reveal the precise mechanism through which RNF2 degrades this Cap and provide new insights into the relationship between ubiquitination and viral replication.

MATERIALS AND METHODS

Cells and virus determination

PK-15 and HEK-293T were originally purchased from the American Type Culture Collection and grown in Dulbecco’s modified Eagle’s medium (DMEM; Gibco; 12100061) containing 5% or 10% fetal bovine serum (LONSERA; S711-001s), penicillin, and streptomycin (Solarbio; P1400). PCV3 strain was stored and used in our laboratory (14).

The procedure of the 50% tissue culture infective dose (TCID₅₀) assay was performed as follows: PK-15 cells transfected with plasmids or siRNA (FLAG-Vector, FLAG-RNF2, NC, or siRNA-#C) were infected with PCV3 at a multiplicity of infection of 0.5 for 1 h at 37°C, and the cell supernatants were collected at the indicated times. The harvested cell supernatants were subjected to 10-fold serial and then cultured in PK-15 monolayers in 96-well cell culture plates. Immunofluorescence assay was utilized for the detection of viral titers, which were expressed as TCID₅₀ per 0.1 mL as previously described (28).

Antibodies

All antibodies used are listed below: rabbit anti-RNF2 antibody (ABclonal Technology; A5563), mouse anti-FLAG antibody (Sigma; F1804), mouse anti-actin antibody (BBI; D191048), rabbit anti-GFP antibody (BBI; D110008), mouse anti-HA antibody (Sigma; H3663), pig anti-PCV3 Rep antibody was obtained from our laboratory, horseradish peroxidase-conjugated goat anti-rabbit or anti-mouse immunoglobulin G (IgG; BBI; D110058 or D110087); and rabbit anti-pig IgG (Bioss; bs-0309Rs).

Plasmids and siRNA transfection

Porcine RNF2 gene (Gene ID: 100517846) was cloned into plasmids to generate FLAG-RNF2 plasmids, and all other plasmids (HA-Ub, HA-Ub-K48R, HA-Ub-K63R, GFP-PCV3-Cap, and FLAG-PCV3-Cap) were stored in our laboratory. Cells grown to 70%–80% confluence were transfected with various plasmids using the Lipofectamine 2000 (Invitrogen Corp; 11668027) according to the manufacturer's protocol for 24 h, followed by other process and analysis. The primers used are listed in Table 1.

TABLE 1.

Primer sequences (5′−3′) used for plasmid construction and quantitative real-time PCR

Name	Forward	Reverse
GFP-RNF2	tcgaattctatgtctcaggctgtgc	gtggatcctcatttgtgctcctttg
GFP-RNF34	tcgaattctatgaaggcgggtgccac	gtggatcctcaggacttgaacacgtg
GFP-TRIML2	tcgaattctatgtccaaaaggctcc	ccgtcgactcacaccaaagaagac
FLAG-Cap-K42R	acatactacacaaagaggtactccaccatgaac	gttcatggtggagtacctctttgtgtagtatgt
FLAG-Cap-K133R	gaaagttccactcgtagggttatgacttctaaa	tttagaagtcataaccctacgagtggaactttc
FLAG-Cap-K138R	aaagttatgacttctaggaaaaaacacagccgt	acggctgtgttttttcctagaagtcataacttt
FLAG-Cap-K139R	gttatgacttctaaaaggaaacacagccgttac	gtaacggctgtgtttccttttagaagtcataac
FLAG-Cap-K140R	atgacttctaaaaaaaggcacagccgttacttc	gaagtaacggctgtgcctttttttagaagtcat
FLAG-Cap-211R	gtttggattcgttacaggtccgttctctaa	ttagagaacggacctgtaacgaatccaaac
qPCR-GAPDH	tcggagtgaacggatttggc	tgacaagcttcccgttctcc
qPCR-RNF2	agcgaggatcaacaagcaca	aggtcctgcttcctgattgc

The siRNAs targeting RNF2 were designed by GengPharma and then transfected into PK-15 cells using the Lipofectamine RNAiMAX Transfection Reagent (Invitrogen; 13778150) for subsequent experiments. siRNAs are as shown below: siRNF2-A (sense, 5′-GCCCUCAGAAGUGGCAAUATT-3′ and antisense, 5′-UAUUGCCACUUCUGAGGGCTT-3′); siRNF2-B (sense, 5′-CUGGGCUUGAGCUUGAUAATT-3′ and antisense, 5′-UUAUCAAGCUCAAGCCCAGTT-3′); siRNF2-C (sense, 5′-GCACAGACAAGAUACAUAATT-3′ and antisense, 5′-UUAUGUAUCUUGUCUGUGCTT-3′).

Co-immunoprecipitation and western blotting

HEK-293 cells co-expressed PCV3-Cap and GFP-RNF2 were lysed by NP-40 lysis solution containing 1 mM protease inhibitor cocktail (Beyotime Biotechnology; P1005) and subsequently treated with the GFP-Nanoab-Agarose beads (LABLEDA; GNA-250–5K) or anti-FLAG affinity resin (Sigma; A2220) for 12 h at 4°C. The protein samples were separated by SDS-PAGE and then transferred onto a nitrocellulose membrane. The membrane was blocked in tris buffered saline with Tween-20 (TBST) buffer with 5% (wt/vol) nonfat milk for 2 h at 37°C, followed by incubation with various primary antibodies and secondary antibodies. After washing, the membranes were exposed using a SuperSignal West Femto Substrate Trial Kit (Thermo Fisher Scientific; 34096) in a chemiluminescence device.

CHX chase assay

PK-15 cells co-expressed with GFP-RNF2 and FLAG-Cap were treated with 100 µg/mL of CHX (MedChemExpress; HY-12320) for 0, 4, and 8 h and then processed and analyzed using western blotting.

Quantitative real-time reverse transcription PCR

Total RNA obtained from PK-15 cells was reversely transcribed into cDNA using a Vazyme cDNA Synthesis Kit (Vazyme; R323-01), followed by a real-time RT-PCR using AceQTM Universal SYBR qPCR Master Mix Kit (Vazyme Biotech, Q712-02). Glyceraldehyde 3-phosphate dehydrogenase was detected as an internal reference. The primers are listed in Table 1.

Confocal microscopy

PK-15 cells transfected with GFP-PCV3-Cap and pmCherry-RNF2 plasmids for 36 h, followed by fix with 4% paraformaldehyde (Solarbio; P1110) for 30 min at room temperature, and then stained with 2,4-diamidino-2-phenylindole (Sigma; D9542) and visualized using a confocal immunofluorescence microscope (Leica TCS SP8 STED, Germany).

Statistical analysis

The statistical analysis was conducted using GraphPad Prism Version 7.0 (GraphPad Software). Three separate studies were conducted to obtained the data. The differences in the results between the two groups were tested for statistical significance using independent samples t tests, and a P value < 0.05 was considered statistically significant.

DATA AVAILABILITY

All data sets generated for this study are within the paper.