Porcine RING Finger Protein 114 Inhibits Classical Swine Fever Virus Replication via K27-Linked Polyubiquitination of Viral NS4B

Porcine RING finger protein 114 (pRNF114) is a member of the RING domain E3 ligases. In this study, it was shown that pRNF114 is a potential anti-CSFV factor and the anti-CSFV effect of pRNF114 depends on its E3 ligase activity. Notably, pRNF114 targets and catalyzes the K27-linked polyubiquitination of the NS4B protein and then promotes proteasome-dependent degradation of NS4B, inhibiting the replication of CSFV. To our knowledge, pRNF114 is the first E3 ligase to be identified as being involved in anti-CSFV activity, and targeting NS4B could be a crucial route for antiviral development.


RESULTS
The mRNA transcription level of pRNF114 is upregulated upon CSFV infection. To evaluate the effects of pRNF114 during CSFV infection in vivo and in vitro, we detected the mRNA transcription level of pRNF114 upon CSFV (strain Shimen) infection by a real-time reverse transcription-quantitative PCR (RT-qPCR) assay. The results showed that the pRNF114 mRNA transcription level was increased significantly in peripheral blood mononuclear cells (PBMCs) of CSFV-infected pigs at 1, 3, or 5 days postinfection (Fig. 1A). Consistently, the mRNA transcription level of pRNF114 was increased in the isolated PBMCs (Fig. 1B), PK-15 cells (Fig. 1C), and porcine alveolar macrophages (PAMs) (Fig. 1D) infected with Shimen (multiplicity of infection [MOI] ϭ 0.1) at 6 or 12 h postinfection (hpi), as well as in PAMs at an MOI of 0.01, 0.1, or 1 at 6 hpi (Fig. 1E). These data suggest that CSFV infection can induce the expression of pRNF114, indicating a potential role for pRNF114 in CSFV infection.
pRNF114 inhibits CSFV replication. Considering that pRNF114 potentially inhibits CSFV replication (31), we established PK-pRNF114 or PK-EGFP cells stably expressing pRNF114 or enhanced green fluorescent protein (EGFP) to confirm the antiviral activity of pRNF114 against CSFV. The cell growth and viability of PK-pRNF114 were similar to those of PK-EGFP cells ( Fig. 2A). We demonstrated that overexpression of pRNF114 inhibits rCSFV-Fluc replication at an MOI of 0.01 or 0.1 at 24 or 48 hpi ( Fig. 2B and C). The anti-CSFV activity of pRNF114 was examined in PK-pRNF114 and PK-EGFP cells upon infection with Shimen (MOI ϭ 0.1 or 0.01). The CSFV RNA ( Fig. 2D and E) and the extracellular titers of the progeny virus ( Fig. 2F and G) were significantly reduced in PK-pRNF114 cells compared with those in PK-EGFP cells at 24 or 48 hpi.
To further examine the anti-CSFV effects of pRNF114, we established a PK-RNF114-KO cell line by CRISPR/Cas9 system and used PK-Cas9 as a control. We examined the protein expression level of RNF114 in PK-RNF114-KO and PK-Cas9 cells. The results indicated that RNF114 protein expression was completely abrogated in PK-RNF114-KO cells (Fig. 3A). Subsequently, the replication efficiency of rCSFV-Fluc or Shimen was assessed in PK-RNF114-KO and PK-Cas9 cells. The viability of PK-RNF114-KO cells was indistinguishable from that of PK-Cas9 cells (Fig. 3B). Compared with those in PK-Cas9 cells, the intracellular Fluc activities (Fig. 3C), viral titers (Fig. 3D), and viral RNA copies (Fig. 3E) were increased in PK-RNF114-KO cells. To further demonstrate that the observed phenomenon was not due to off-target effects, Lenti-pRNF114(NGG mut), expressing pRNF114 with the NGG consensus mutation, was transduced into PK-RNF114-KO cells. We found that pRNF114 restoration inhibited CSFV infection compared with that in PK-RNF114-KO cells transduced with Lenti-EGFP (Fig. 3F). Together, these results demonstrate that knockout of RNF114 promotes CSFV replication, suggesting that pRNF114 inhibits CSFV replication.
E3 ligase activity of pRNF114 is essential for its inhibition of CSFV replication. The pRNF114 protein belongs to the RING-UIM family, containing a RING domain (19). To investigate whether the E3 ligase activity of pRNF114 is involved in the inhibition of CSFV, we generated a pRNF114 RING domain mutant plasmid, pGST-pRNF114(C64/ 67A), in which the cysteines (C) at positions 64 and 67 on the pRNF114 RING domain were mutated to alanine (A) (Fig. 4A), and the lack of ubiquitin ligase activity for pRNF114(C64/67A) was verified by an in vitro ubiquitination experiment (Fig. 4B). We established PK-pRNF114(C64/67A) cells stably expressing pRNF114(C64/67A), with no difference in cell growth and viability from PK-pRNF114 and PK-EGFP cells (Fig. 4C). The anti-CSFV effects of pRNF114 were detected in those cells. Compared with those in PK-EGFP cells, the intracellular Fluc activities (Fig. 4D), CSFV RNA (Fig. 4E), and viral titers (Fig. 4F) were reduced in PK-pRNF114 cells. In PK-pRNF114(C64/67A) cells, however, Influence of overexpression of pRNF114 on CSFV growth. PK-pRNF114 and PK-EGFP cells were infected with rCSFV-Fluc or Shimen (MOI ϭ 0.01 or 0.1) for 24 or 48 h. The luciferase activities were assayed by a luciferase reporter assay system (Promega) and expressed as relative light units (RLU) (B and C). The CSFV genomic RNA copy numbers were quantified by real-time reverse transcription-quantitative PCR (D and E). Viral titers were determined as median tissue culture infective doses (TCID 50 ) per milliliter (F and G). Bars represent SDs. *, P Ͻ 0.05; **, P Ͻ 0.01. pRNF114(C64/67A) failed to display an anti-CSFV action, as demonstrated by higher Fluc activities, viral genome RNA, and virus yields. These data suggest that the anti-CSFV activity of pRNF114 is dependent on the E3 ligase activity.
pRNF114 interacts with the CSFV NS4B protein. Many RING E3s have been reported to restrict the replication of various viruses through interaction with viral nonstructural proteins (21,23,24). To clarify the specific targets of pRNF114, we examined the association between pRNF114 and CSFV nonstructural proteins by coimmunoprecipitation (co-IP) and identified NS4B as a pRNF114-interacting partner (Fig. 5A). Furthermore, the interaction between pRNF114 and NS4B was demonstrated by co-IP (Fig. 5B), reciprocal co-IP (Fig. 5C), and glutathione S-transferase (GST) pulldown assays (Fig. 5D). Next, we verified the interaction of pRNF114 with NS4B in the context of CSFV infection (Fig. 5E).
To further prove the association pRNF114 and NS4B, we also evaluated whether the pRNF114 protein colocalizes with NS4B in HEK293T cells cotransfected with pFlag-pRNF114 and pMyc-NS4B. The results indicated the colocalization of pRNF114 and NS4B in the cytoplasm (Fig. 5F). We further examined the colocalization of the pRNF114 protein and NS4B in CSFV-infected PK-pRNF114 cells. Confocal images revealed that pRNF114 colocalizes with NS4B in the cytoplasm of the CSFV-infected cells (Fig. 5G). These results together suggest that pRNF114 interacts with the CSFV NS4B protein.

The pRNF114 C-terminal region interacts with the NS4B C-terminal region.
To further investigate the key regions responsible for the NS4B-pRNF114 interaction, we constructed truncated mutants of pRNF114 and NS4B ( Fig. 6A and B). HEK293T cells were transfected with the desired plasmids, and the domain responsible for interaction of pRNF114 with NS4B was determined using co-IP and GST pulldown assays. The results showed that pRNF114(69-228) but not pRNF114(1-68) was involved in the interaction (Fig. 6C) and NS4B(180-347) exhibited a strong interaction with pRNF114 ( Fig. 6D). Further results confirmed that NS4B(180-347) interacted with pRNF114(69-228) (Fig. 6E). These findings suggest that the C-terminal region of pRNF114 is necessary for interaction with the C-terminal region of NS4B.
pRNF114 promotes proteasomal degradation of NS4B. Considering that pRNF114 is an E3 ubiquitin ligase and its enzymatic activity is critical for inhibition of the CSFV replication, we hypothesized that pRNF114 could promote NS4B degradation. To this end, pMyc-NS4B and increasing amounts of pFlag-pRNF114 were cotransfected into HEK293T cells. The results showed that pRNF114 did not influence mRNA transcription ( Fig. 7A) but reduced protein expression of NS4B in a dose-dependent manner (Fig. 7B). Upon CSFV infection, the protein expression level of NS4B was decreased in PK-15 cells transfected with pFlag-pRNF114 (Fig. 7C). There are two major pathways responsible for protein degradation, proteasome pathway and lysosomal proteolysis. To test whether NS4B degradation results from these two pathways, we treated cells with the proteasome inhibitor MG132 or lysosome inhibitor NH 4 Cl. The results showed that pRNF114-induced NS4B degradation was rescued by MG132 but not by NH 4 Cl (Fig. 7D). Next, we determined whether NS4B degradation is dependent on the E3 ubiquitin ligase activity of pRNF114. The results showed that pRNF114(C64/67A) attenuated NS4B degradation when expressed in HEK293T cells compared with the wild-type pRNF114 (Fig. 7E). Collectively, these results indicate that the interaction of pRNF114 with NS4B enhances NS4B degradation in a proteasome pathway, which is dependent on its E3 ubiquitin ligase activity.
pRNF114 mediates ubiquitination of NS4B. Protein ubiquitination is a critical step in the proteasome degradation pathway (33). Therefore, we hypothesized that the antiviral potential of pRNF114 could be due to ubiquitination of NS4B. For this purpose, pMyc-NS4B, pHA-Ub, and pFlag-pRNF114 were cotransfected into HEK293T cells. The results showed that NS4B ubiquitination was markedly increased in the presence of pRNF114 (Fig. 8A). We then validated the role of pRNF114 in NS4B ubiquitination in the context of CSFV infection. The results showed that the NS4B ubiquitination was increased in PK-15 cells transfected with pFlag-pRNF114 (Fig. 8B). Next, pMyc-NS4B and Myc-NS4B protein. The proteins bound to beads were analyzed by immunoblotting using the indicated antibodies. (E) Endogenous co-IP assay. PK-15 cells transfected with pFlag-pRNF114 were infected with Shimen (MOI ϭ 1) for 48 h and subjected to a co-IP assay using anti-Flag MAb. (F and G) Confocal assay. HEK293T cells were cotransfected with pFlag-pRNF114 and pMyc-NS4B (F), and PK-pRNF114 cells were infected with Shimen (MOI ϭ 1) (G). The cells were subjected to a confocal assay. DAPI, 4=,6-diamidino-2-phenylindole. increasing doses of pRNF114 (0, 1, or 2 g) were cotransfected in HEK293T cells, and immunoprecipitation and immunoblotting analyses showed that pRNF114 ubiquitinates NS4B in a dose-dependent manner (Fig. 8C). To determine the role of the E3 ligase activity in the ubiquitination of NS4B, HEK293T cells were cotransfected with pMyc-NS4B and pFlag-pRNF114, pRNF114(C64/67A), or pRNF114(69-228). The ubiquitination analyses revealed that the mutation of the RING domain on pRNF114 reduced NS4B ubiquitination (Fig. 8D). We also detected more endogenous ubiquitin conjugated to NS4B in the presence of pRNF114 but not mutated pRNF114(C64/67A) and pRNF114(69-228) (Fig. 8E). These data suggest that the ubiquitination of NS4B induced by pRNF114 depends on its E3 ligase activity.
pRNF114 catalyzes the K27-linked polyubiquitination of NS4B. To verify the effects of pRNF114 on the types of polyubiquitination on NS4B, a panel of Ub mutants was included. We first generated an HA-Ub mutant (K0), in which all lysine (K) residues were replaced with arginine (R), and then individual lysine residues (K6, K11, K27, K29, K33, K48, or K63) were reintroduced into the K0 mutant to generate single-lysine mutants (Fig. 9A). pMyc-NS4B and pFlag-pRNF114 together with wild-type HA-Ub or its mutants were transfected into HEK293T cells. The results demonstrated that pRNF114mediated NS4B polyubiquitination could be detected in the presence of HA-K27-Ub but not of other HA-Ub mutants (Fig. 9B). In contrast, the ubiquitination of NS4B was impaired when using HA-K27R-Ub (HA-Ub with only K27 mutated to R), indicating that pRNF114 mediates the K27-linked polyubiquitination of NS4B (Fig. 9C). Finally, in vitro ubiquitination assay further confirmed that pRNF114 mediated K27-linked polyubiquitination of NS4B in the cell-free system (Fig. 9D). These results collectively show that pRNF114 catalyzes K27-linked polyubiquitination of NS4B.

DISCUSSION
In this study, we clarified the antiviral mechanism of pRNF114, which is involved in interaction with CSFV NS4B and the degradation of NS4B via a proteasome-dependent pathway. We found that pRNF114 efficiently suppressed CSFV growth in PK-15 cells (Fig. 2 and 3). Interestingly, we found that the anti-CSFV effects of pRNF114 depended on its E3 ligase activity ( Fig. 4 and 8). Notably, we demonstrated that pRNF114 binds to NS4B  A and B) The ubiquitination level of NS4B is potentiated by expression of pRNF114. HEK293T cells were cotransfected with the indicated plasmids for 24 h and treated with 10 M MG132 (Sigma-Aldrich) for an additional 24 h. Cell extracts were prepared and analyzed by co-IP assay using an anti-Myc monoclonal antibody (MAb) followed by Western blotting (WB) with the indicated antibodies (A). PK-15 cells transfected with pFlag-pRNF114 were infected with Shimen (MOI ϭ 1) for 48 h. Lysates derived from these cells were subjected to IP using anti-NS4B polyclonal antibody and WB with the indicated antibodies (B). (C) pRNF114 ubiquitinates NS4B in a dose-dependent manner. HEK293T cells were cotransfected with pMyc-NS4B and increasing concentrations of pFlag-pRNF114 (0, 1, or 2 g), followed by immunoprecipitation with anti-Myc MAb, and analyzed via immunoblotting with an anti-Ub MAb. (D and E) pRNF114 ubiquitinates NS4B depending on its E3 ubiquitin ligase. Lysates from HEK293T cells which were cotransfected with pMyc-NS4B and pFlag-pRNF114, pRNF114(C64/67A), and pRNF114(69-228), together with pHA-Ub (D) or without pHA-Ub (E), were subjected to immunoblotting followed by co-IP. Ubiquitination of NS4B was detected with anti-HA (D) or anti-Ub (E) MAb. and promotes NS4B K27-linked polyubiquitination, leading to the proteasomal degradation of NS4B (Fig. 5 to 8). Taken together, our findings demonstrate that pRNF114 is a novel anti-CSFV protein that acts through targeting NS4B for its degradation.
Our results indicate that pRNF114 inhibits CSFV replication through interacting with NS4B. Since the interaction mediates NS4B degradation, we suppose the disruption of pRNF114 and NS4B would inhibit CSFV replication. It has previously been shown that pRNF114 is involved in T-cell activation (30) and that overexpression of pRNF114 enhances NF-B and IRF3 activities and increases the expression of type I and III interferons (IFNs). Our findings that pRNF114 directly inhibits CSFV replication have not been previously reported and may contribute to a better understanding of the roles of pRNF114 in antiviral response beyond immunology regulation.  Protein modifications are intimately linked to the protein's function. It has been demonstrated that ubiquitin-modification regulates the replication of flaviviruses. The ubiquitin ligase CBLL1 and the proteasome-ubiquitin system are required for the cellular internalization of West Nile virus (WNV) and the entry and replication of other flaviviruses (34,35). The E3 ubiquitin ligase MKRN1 induces degradation of WNV capsid protein in a proteasome-dependent manner (36). It has been reported that the K48linked polyubiquitination of the DENV NS1 protein inhibits its interaction with the viral partner NS4B, suggesting that NS1 ubiquitination may affect virus replication (37). In addition, it has been shown that the CSFV N pro induces degradation of IRF3 via the proteasome pathway; possibly, N pro recruits E3 ubiquitin ligase complex to degrade IRF3 (38). Since proteasome inhibitor MG132 increases ubiquitination of CSFV NS4B, our study further supports a role for the ubiquitin-proteasome system in flaviviral replication.
The well-understood functions of polyubiquitination are mediated by chains composed of K48 and K63 linkages; however, little is known about the functions of the other noncanonical ubiquitin chains (linked via K6, K11, K27, K29, or K33) (39). Increasing studies suggest that K27-linked polyubiquitination contributes the regulation of proteasomal degradation (40). TRIM40 interacts with RIG-I and MDA5 and promotes their K27/K48-linked polyubiquitination, leading to their proteasomal degradation (41). IpaH9.8, a Shigella bacterial ubiquitin E3 ligase, catalyzes K27-linked polyubiquitination of NEMO and undergoes proteasomal degradation (42). Here we demonstrate that the K27-linked polyubiquitination of NS4B is also associated with the ubiquitin-proteasome pathway, providing an additional role for polyubiquitination with K27 linkage. In addition, CSFV NS4B has 23 lysine residues. Six (K48, K63, K178, K206, K258, and K262) lysine residues were predicted to be potential ubiquitination sites by the UbPred program (http://www .ubpred.org/). The specific lysine residues deserve future investigations, which would be extremely valuable in constructing NS4B mutants to support CSFV replication.
Currently, few studies have reported on the interactions between flaviviral NS4B and cellular proteins, possibly due to the lower significance of NS4B in the flavivirus life cycle than of other nonstructural proteins, such as NS3 and NS5. Previous studies have provided clues that the formation of the K27-linked polyubiquitination on target proteins serves as an anchoring platform for its interaction with other functional proteins. Upon microbial DNA challenge, E3 ubiquitin ligase complex INSIG1 and AMFR catalyze the K27-linked polyubiquitination of STING, creating an anchoring platform for recruiting and activating TBK1 (43). TRIM21 positively regulates innate immunity by catalyzing K27-linked polyubiquitination of MAVS and promoting the recruitment of TBK1 to MAVS (44). During DENV replication, the K27-linked polyubiquitination of NS3 is necessary to interact with NS2B to form the NS2B3 protease complex (45). pRNF114 is the first E3 ligase to be identified as being involved in the K27-linked polyubiquitination of the CSFV NS4B, which is a component of the replication complex, together with NS4A, NS3, NS4B, NS5A, and NS5B (7). Therefore, it is possible that this modification will influence the formation of replication complex and, in turn, regulate the replication of CSFV. Our study provides new insights on the role of the host ubiquitin system during CSFV replication. A more comprehensive understanding of this virushost interplay will be crucial for our understanding of flavivirus pathogenesis and the development of antiviral strategies and treatments.
In summary, here we have reported a novel function of the host E3-ubqiuitin ligase pRNF114 in inhibiting CSFV replication. Because the nonstructural protein NS4B is indispensable for processing the CSFV polyprotein, targeting NS4B could be a crucial route for antiviral development.

Cells and viruses.
HEK293T and PK-15 cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (FBS). PBMCs and PAMs were isolated and maintained in RPMI 1640 medium supplemented with 10% FBS, and all these cells were cultured in a 5% CO 2 humidified atmosphere. The rCSFV-Fluc strain and the CSFV Shimen strain were propagated in PK-15 cells.
Generation of PK-RNF114-KO cells. To generate PK-RNF114-KO cells, the single guide RNA (sgRNA) was designed using an online CRISPR design tool (http://crispr.mit.edu/) to target exons 1 and 2 of pRNF114 ( Table 2). The lentiCRISPRv2 plasmid was digested using BsmBI (New England BioLabs [NEB]), and pCRISPRv2-sgRNA-RNF114-1 and pCRISPRv2-sgRNA-RNF114-2 were constructed. All of the constructs in this study were verified by sequencing. PK-15 cells were transduced and selected in the presence of 2.5 g/ml of puromycin (Solarbio) and seeded on 96-well plates with serial dilutions to obtain a single-cell-derived colony. Two weeks later, the cells were collected and genomic DNA was extracted; the pRNF114 gene sequence was detected using PCR and sequencing analysis.
Cell viability assay. Cell viability was determined by using Cell Counting Kit-8 (CCK-8; Dojindo) according to the manufacturer's instructions.
IFA. Culture medium was collected and measured for titers of CSFV by indirect immunofluorescence assay (IFA) as reported previously (46).
Real-time RT-qPCR. Total RNA extracted from CSFV-infected PK-15 cells or other cells was treated with TRIzol reagent (Invitrogen) and was reverse transcribed into the cDNA using avian myeloblastosis virus (AMV) reverse transcriptase XL (TaKaRa). Viral RNA copies were quantified using a previously described RT-qPCR assay (47). The transcription levels of RNF114 and NS4B in PK-15 cells with or without Shimen infection were quantified by the threshold cycle (2 ϪΔΔCT ) method (48). The mRNA level of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was set as an internal loading control. The primers used for real-time RT-PCR are listed in Table 1.
Western blotting and co-IP assay. Samples were lysed with NP-40 lysis buffer containing a protease inhibitor (1 mM phenylmethylsulfonyl fluoride [PMSF]; Beyotime) for 30 min and then centrifuged for 20 min at 4°C; the supernatants were denatured in 5ϫ sample loading buffer for 7 min at 100°C. For immunoprecipitation assays, the supernatants were mixed with the respective primary antibodies in the presence of protein G-agarose (Roche) at 4°C overnight, followed by Western blotting.
GST pulldown assay. GST pulldown was performed as described previously (46). Briefly, purified GST or GST-pRNF114 fusion protein was mixed with lysates from cells transfected with pMyc-NS4B at 4°C overnight. The proteins pulled down by GST-agarose were detected by Western blotting using a rabbit anti-Myc monoclonal antibody (MAb) and a mouse anti-GST polyclonal antibody (PAb) (Tiangen).
Confocal imaging. The cells were seeded in glass-bottom dishes. HEK293T cells were cotransfected with the pFlag-pRNF114 and pMyc-NS4B (1 g each) for 36 h. PK-pRNF114 cells were infected with Shimen at an MOI of 1 for 48 h. Then the cells were stained with the relevant antibodies and images were acquired by using a Leica SP2 confocal system (Leica Microsystems) (31).
Ubiquitination assay. pMyc-NS4B and pHA-Ub were cotransfected into HEK293T cells with or without pFlag-pRNF114. The cells were then treated with 10 M MG132 (Sigma-Aldrich) for 24 h and were lysed by NP-40 lysis buffer containing a deubiquitinase inhibitor (10 M N-ethylmaleimide [NEM]; Sigma-Aldrich). Then the anti-Myc MAb-conjugated agarose beads were added into the samples separately. Following incubation overnight at 4°C, the samples were examined by Western blotting.
For the in vitro ubiquitination assay, bacterially expressed GST fusion proteins of pRNF114 and pRNF114(C64/67A) were captured on glutathione-Sepharose beads and incubated for 90 min at 30°C with purified Ub-activating enzyme UBE1, His-tagged Ub-conjugating enzyme UbcH5A, Flag-Ub, and Ub conjugation reaction buffer (R&D Systems). The ubiquitin ligase activity of pRNF114 was analyzed by Western blotting with an anti-Flag MAb. For in vitro ubiquitination of NS4B by pRNF114, purified NS4B and pRNF114 were incubated with UBE1, UbcH5A, His-Ub, His-K27, or His-K27R in the presence of ATP. The in vitro ubiquitination of NS4B was analyzed by Western blotting.
Statistical analysis. All experiments were performed at least three times. Data were statistically analyzed by Student's t test using SPSS 22.0 software. Differences were considered significant if the unadjusted P value was less than 0.05.

ACKNOWLEDGMENTS
This work was supported by the National Natural Science Foundation of China (grants 31702220, 31630080, and 31672537), the Natural Science Foundation of Heilongjiang Province of China (grant QC2017027), and the China Postdoctoral Science Foundation (grant 2017M620979).