Rhesus Macaque Rhadinovirus Encodes a Viral Interferon Regulatory Factor To Disrupt Promyelocytic Leukemia Nuclear Bodies and Antagonize Type I Interferon Signaling

KSHV and RRV encode a unique set of homologs of cellular IFN regulatory factors, termed vIRFs, which are hypothesized to help these viruses evade the innate immune response and establish infections in their respective hosts. Our work elucidates the role of one RRV vIRF, R12, and demonstrates that RRV can dampen the type I IFN response downstream of IFN signaling, which would be important for establishing a successful infection in vivo.

ABSTRACT

Interferon (IFN) production and the subsequent induction of IFN-stimulated genes (ISGs) are highly effective innate strategies utilized by cells to protect against invading pathogens, including viruses. Critical components involved in this innate process are promyelocytic leukemia nuclear bodies (PML-NBs), which are subnuclear structures required for the development of a robust IFN response. As such, PML-NBs serve as an important hurdle for viruses to overcome to successfully establish an infection. Both Kaposi’s sarcoma-associated herpesvirus (KSHV) and the closely related rhesus macaque rhadinovirus (RRV) are unique for encoding viral homologs of IFN regulatory factors (termed vIRFs) that can manipulate the host immune response by multiple mechanisms. All four KSHV vIRFs inhibit the induction of IFN, while vIRF1 and vIRF2 can inhibit ISG induction downstream of the IFN receptor. Less is known about the RRV vIRFs. RRV vIRF R6 can inhibit the induction of IFN by IRF3; however, it is not known whether any RRV vIRFs inhibit ISG induction following IFN receptor signaling. In our present study, we demonstrate that the RRV vIRF R12 aids viral replication in the presence of the type I IFN response. This is achieved in part through the disruption of PML-NBs and the inhibition of robust ISG transcription.

IMPORTANCE KSHV and RRV encode a unique set of homologs of cellular IFN regulatory factors, termed vIRFs, which are hypothesized to help these viruses evade the innate immune response and establish infections in their respective hosts. Our work elucidates the role of one RRV vIRF, R12, and demonstrates that RRV can dampen the type I IFN response downstream of IFN signaling, which would be important for establishing a successful infection in vivo.

INTRODUCTION

All viruses must employ tactics to evade or counteract the host immune response, if they are to successfully establish an infection. The cell’s first line of defense against infection is the innate immune system. Recent studies have found that viruses can antagonize the innate immune system by disrupting promyelocytic leukemia nuclear bodies (PML-NBs) (1–3). PML-NBs are constitutively expressed, multiprotein punctate structures located within the nucleus of most cell types (4). Proteins found in PML-NBs are either permanent residents (PML, SP100, and Daxx) or transiently present (CBP, P300, and p53) (2–8). The PML proteins are absolutely required to form PML-NBs, and evidence suggests that SUMOylation of PML and some of the other PML-resident proteins is important to form these structures (9, 10). Multiple functions have been attributed to PML-NBs, including cell cycle regulation, antiviral defense, as sites where proteins are SUMO modified, and transcriptional regulation (1–3). The antiviral functions of PML-NBs have been shown to restrict herpesvirus infections. PML-NBs can epigenetically silence herpes simplex virus 1 (HSV-1) and human cytomegalovirus (HCMV) genomes once they enter the nucleus, and PML-NBs composed of PML isoform IV can trap varicella zoster virus (VZV) nucleocapsids to prevent nuclear egress (11–14). More generally, PML-NBs enhance the induction of interferon (IFN)-stimulated genes (ISGs), aiding the establishment of an antiviral state within the cell (15). While PML and SP100 are constitutively expressed, type I IFN upregulates their protein expression (16, 17). Several viruses, including human herpesviruses, have been shown to modulate or disperse PML-NBs in order to evade antiviral defenses, regulate viral gene transcription, and replicate (3, 18–24). For example, HSV-1 encodes ICP0, which disrupts PML-NBs by degrading SUMO-1-modified forms of PML and SP100 proteins (25). HCMV encodes pp71 and IE1, both of which aid in the dispersal of PML-NBs (14, 21). Gammaherpesviruses have also been reported to antagonize PML-NBs. Kaposi’s sarcoma (KS)-associated herpesvirus (KSHV), a human gamma-2 herpesvirus (γ2HV), encodes open reading frame 75 (ORF75) and viral IFN regulatory factor 3 (vIRF3) to mediate displacement of human Daxx (hDaxx) from PML-NBs and induce the degradation of PML protein, respectively (18, 23). ORF75 homologs of other gammaherpesviruses (murine gammaherpesvirus 68 [MHV-68], Epstein-Barr virus [EBV], herpesvirus saimiri [HVS], and rhesus macaque [RM] rhadinovirus [RRV]) have also been implicated in the disruption of PML-NBs (24, 26–29).

KSHV infection is generally asymptomatic in healthy individuals; however, the virus can promote the development of KS, primary effusion lymphoma (PEL), multicentric Castleman’s disease (MCD), and some non-Hodgkin’s lymphomas (NHLs) in immunocompromised individuals, including AIDS patients (30–33). Unfortunately, establishing animal models for KSHV have proven difficult. Infection of RMs with KSHV proved ineffective, with no KSHV transcripts or pathologies being detected in infected animals (34). Several mouse models have also been explored, with limited success (35–37). In vitro, KSHV produces a predominantly latent infection in cell culture, making the study of virus replication and expression of viral genes difficult (38, 39). A viable alternative is to study an animal virus that can induce similar disease manifestations in its natural host, such as the γ2HV RRV, which naturally infects RMs (40, 41). RRV provides a powerful animal model with which to study infection and KSHV-like disease development, as experimental RRV infection of simian immunodeficiency virus (SIV)-infected RMs can lead to the development of MCD, NHL, and retroperitoneal fibromatosis (a mesenchymal proliferative lesion that possesses cellular features that resemble those of KS) (42, 43). In vitro, RRV establishes a robust lytic infection in primary RM fibroblast cells, facilitating the growth and propagation of the virus (41, 44–46). Importantly, the genomes of KSHV and RRV are essentially colinear, with both RRV and KSHV encoding vIRFs, with KSHV encoding four vIRF ORFs (vIRF-1 through vIRF-4) and RRV encoding eight (R6 through R13). The vIRFs share some homology to cellular IRFs, which are transcription factors involved in regulating the IFN response (40, 47–52). Research on the KSHV vIRFs, which has mostly involved overexpression systems or chemically induced reactivation of KSHV from latently infected cells in vitro, has revealed multiple functions of these molecules, including an ability to antagonize both the innate and adaptive immune responses and the inhibition of apoptosis (49, 53). RRV provides the opportunity to investigate the function of vIRFs not only during in vitro infection but also during infection in vivo in an established primate model system. Previously, an RRV bacterial artificial chromosome (BAC) system was used to investigate the vIRF ORFs in the context of de novo lytic infection both in vitro and in vivo (51, 52). It was shown that while wild-type (WT) RRV and an RRV_BAC-derived mutant lacking all 8 vIRF ORFs (RRV_vIRF-KO) grew similarly in vitro, there was a significant growth defect observed in vivo for RRV_vIRF-KO early during infection, indicating that the type I IFN response is overcome more efficiently when the vIRF ORFs are expressed during viral infection in vivo (51, 52). In this study, we show that the RRV vIRFs help establish a more efficient infection in the presence of type I IFN through the disruption of PML-NBs. Furthermore, we show that the vIRF R12 is necessary but not sufficient for the RRV disruption of PML-NBs.

RESULTS

RRV vIRFs enhance infection in the presence of IFN.

To determine whether the difference in viral growth between WT RRV and RRV_vIRF-KO observed in vivo is due to the type I IFN response, a single-step growth curve was performed using primary rhesus fibroblast (RhF) cells that were treated with rhesus IFN alpha 2 (RhIFN-α2) for 18 h prior to infection (Fig. 1A). Overall, growths of WT RRV and RRV_vIRF-KO in the absence of IFN were not significantly different, while the difference between either virus grown in the absence of IFN compared to either virus grown in the presence of IFN was significant (adjusted P value of <0.001) (Fig. 1A). However, we were interested in comparing WT RRV to RRV_vIRF-KO in the presence of RhIFN-α2. In the presence of RhIFN-α2, an almost 1-log growth reduction was measured during RRV_vIRF-KO infection compared to WT RRV infections at 12, 24, 48, and 72 h postinfection (hpi) (Fig. 1A). Furthermore, the growth kinetics of WT RRV was significantly different than that of RRV_vIRF-KO when grown in the presence of IFN (adjusted P value of 0.0201). Residual RhIFN-α2 from the growth curves did not negatively affect the plaque assays, as this was verified by performing plaque assays in the presence of the highest concentration of residual IFN. This indicated that vIRFs are capable of lessening the negative effects of IFN on viral lytic replication in vitro but do not restore viral growth to levels attained in the absence of IFN.

FIG 1.

RRV is sensitive to IFN, and the vIRFs are necessary for PML-NB disruption. (A) Primary RhFs were infected with WT RRV or RRV_vIRF-KO at an MOI of 2 in the presence or absence of 100 U/ml RhIFN-α2. Viral titers were measured at the indicated times postinfection by a plaque assay and are presented as PFU per milliliter. Assays were performed in duplicate, replicates were averaged, and data were analyzed by repeated-measures ANOVA with a post hoc Tukey-Kramer test. Adjusted P values of less than 0.05 were considered significant, and asterisks denote significant P values (*, P ≤ 0.05; **, P ≤ 0.01). (B) tRhFs were infected with WT RRV or RRV_vIRF-KO at an MOI of 2 in the presence or absence of 100 U/ml RhIFN-α2. At the indicated times postinfection, cells were fixed with methanol and stained with PML (red)- and RRV-gB (green)-specific antibodies, while nuclei were stained with DAPI (blue). Arrows indicate RRV-gB-positive (gB⁺) cells that lack PML-NB staining within nuclei. The experiment was performed twice in duplicate, and representative images at a ×40 magnification are shown. (C) PML/RRV-gB-double-positive cells as well as PML-negative/RRV-gB-positive cells were counted at 36 hpi from cells treated as described above for panel B. Data were analyzed by two-tailed Fisher’s exact test, and P values of less than 0.05 were considered significant. (D) For the RRV-gB-positive cells that were positive for PML-NBs in panel C, the number of PML-NBs was counted within the nuclei of each cell. Data were analyzed by unpaired Student’s t test, and P values of less than 0.05 were considered significant. (C and D) Experiments were performed twice, and data from one representative experiment are shown. (E) tRhFs were infected with WT RRV-GFP or RRV_vIRF-KO-GFP at an MOI of 5 for 24 h, before sorting the GFP-positive cells by flow cytometry for fluorescein isothiocyanate (FITC). Cytoplasmic and nuclear fractions were prepared from sorted cells, untreated cells, and cells treated with 100 U/ml RhIFN-α2 for 18 h. Cytoplasmic and nuclear proteins were separated before analysis by SDS-PAGE and probed with PML-, lamin A/C-, RRV-ORF52-, and GAPDH-specific antibodies. Densitometry on the PML isoform I/II band was performed and normalized to the lamin A/C loading control. The experiment was performed three times, and data from one representative experiment are shown. IB, immunoblotting.

RRV vIRFs are required for the dispersal of PML-NBs.

As PML-NBs enhance the type I IFN response, we next investigated whether RRV infection affected PML-NBs to evade the type I IFN response. Telomerized RhFs (tRhFs) were infected at a multiplicity of infection (MOI) of 2 in the presence or absence of exogenous IFN, and cultures were analyzed at 18, 24, and 36 hpi by immunofluorescence analysis (IFA) for PML and RRV glycoprotein B (RRV-gB) to detect virus infection and production (Fig. 1B). We observed that by 18 hpi, the punctate PML-NB staining pattern present in uninfected cells was no longer detectable in WT RRV-infected cells, while cells infected with RRV_vIRF-KO retained PML-NB staining even out to 36 hpi. These results demonstrate that WT RRV infection can disrupt PML-NBs and suggest that one or more of the RRV vIRFs are necessary for this disruption. Compared to the untreated and WT RRV-infected cultures, we found that RhIFN-α2 treatment delayed the loss of PML-NBs during WT RRV infection by 6 h (PML-NB loss observed at 24 hpi), while PML-NBs remained intact throughout the RRV_vIRF-KO infection (Fig. 1B). To quantify the loss of PML-NBs observed by IFA, the numbers of gB-positive (gB⁺) cells with and without PML-NBs were counted 36 h after infection (several hours after PML-NB disruption by WT RRV had begun but before cytopathic effects of viral infection could obscure results) with either WT RRV or RRV_vIRF-KO (Fig. 1C). Following WT RRV infection, 11% of the gB⁺ cells still contained PML-NBs, whereas 90% of gB⁺ RRV_vIRF-KO-infected cells retained PML-NBs. The difference between the two virus infections was statistically significant (P < 0.0001) (Fig. 1C). RhIFN-α2 treatment did not alter this phenotype, as 21% of WT RRV-infected cells had PML-NBs, while 96% of RRV_vIRF-KO-infected cells contained PML-NBs (Fig. 1C). In addition, the number of PML-NBs within nuclei of gB⁺ cells in the same samples at 36 hpi was also counted, and a statistically significant (P < 0.01) difference was found between the two types of virus-infected cells in the absence and presence of RhIFN-α2 (Fig. 1D). Specifically, there were fewer PML-NB punctate structures within the nuclei of WT RRV-infected cells (2 or 7, with or without RhIFN-α2, respectively) than in cells infected with RRV_vIRF-KO (16 or 21, with or without RhIFN-α2, respectively) (Fig. 1D).

To determine whether the loss of PML-NBs observed by IFA was due to degradation of PML proteins or to dispersal of PML-NBs, tRhFs were infected with green fluorescent protein (GFP)-expressing WT RRV (WT RRV-GFP) or RRV_vIRF-KO (RRV_vIRF-KO-GFP) at an MOI of 5 for 24 h, followed by sorting of the infected cells based on GFP expression. Nuclear and cytoplasmic protein extracts were prepared from the GFP⁺ cells and resolved by SDS-PAGE for Western blot analysis (Fig. 1E). This revealed that 24 h after infection with WT RRV-GFP, there was a reduction in PML isoform I and II protein levels by approximately 75% compared to those in mock-infected cells. This differed from RRV_vIRF-KO-GFP infection, where increases in PML isoform I and II protein levels were detected compared to those in mock-infected cells, suggesting that the expression of the vIRFs induces the degradation of PML isoform I and II proteins.

RRV vIRFs are not necessary for the loss of SP100 or Daxx punctate structures in RRV-infected nuclei.

We subsequently investigated whether SP100 or Daxx, two PML-NB-resident proteins, remain associated with PML-NBs after RRV infection. At 24 hpi, cells infected with either WT RRV or RRV_vIRF-KO were analyzed for SP100 and Daxx by IFA (Fig. 2A). As expected, PML protein no longer stained with the characteristic punctate structures in WT RRV-infected cells but was still present in RRV_vIRF-KO-infected cells (Fig. 2A, left panels). However, both SP100 and Daxx were absent in WT RRV- and RRV_vIRF-KO-infected cells, indicating that an RRV protein other than the vIRFs is responsible for the loss of SP100 and Daxx (Fig. 2A, middle and right panels). Quantification of the SP100 and Daxx punctate structures from the IFA images revealed no statistical difference between WT RRV and RRV_vIRF-KO infection for either SP100 or Daxx (Fig. 2B and C), in contrast to uninfected cells, which averaged 13.78 SP100 dots per nucleus and 10.72 Daxx dots per nucleus (Fig. 2D).

FIG 2.

RRV vIRFs do not affect localization of SP100 or Daxx proteins. (A) tRhFs were mock infected or infected with WT RRV_BAC or RRV_vIRF-KO at an MOI of 2 for 24 h. Cells were fixed with methanol and stained with PML (red)-, SP100 (red)-, Daxx (red)-, and RRV-gB (green)-specific antibodies. Nuclei were stained with DAPI (blue). Arrows indicate gB-positive cells that lack nuclear staining for PML, SP100, or Daxx. Images at a ×40 magnification are presented. (B and C) IFAs were performed as described in the legends of Fig. 1C and D for RRV-gB-positive cells with or without SP100 (B) or Daxx (C). Data were analyzed by two-tailed Fisher’s exact test, and P values of less than 0.05 were considered significant. NS, not significant. (D) SP100 and Daxx dots within uninfected cells were quantified from IFA images and are displayed as the number of punctate dots per nucleus.

RRV vIRF R12 colocalizes with PML-NBs.

Previous work from our laboratory showed that during transient-transfection assays, R12 localized to the nucleus in punctate structures (data not shown), while the other seven vIRFs were localized diffusely throughout the nucleus and/or cytoplasm. Thus, we first investigated R12 as the possible vIRF involved in the disruption of PML-NBs. A doxycycline (Dox)-inducible tRhF cell line that expressed R12 tagged with a C-terminal FLAG epitope (tRhF-R12) was constructed to assess R12 involvement in PML-NB disruption. While R12 localizes to PML-NBs, the expression of R12 alone (in the presence or absence of RhIFN-α2) does not lead to the complete disappearance of PML-NBs by IFA, even up to 72 h after Dox treatment (Fig. 3A and data not shown). However, there was a statistically significant change in the number of PML-NBs, as untreated cells had an average of 16 PML-NBs per nucleus, while R12-FLAG-expressing cells averaged 7.5 PML-NBs (Fig. 3B). A statistically significant reduction in the number of PML-NBs following R12 expression was also observed in the presence of RhIFN-α2 treatment, with averages of 19.5 PML-NBs in RhIFN-α2-treated cells and 13 PML-NBs in Dox- and RhIFN-α2-treated cells (Fig. 3B). The reduction in PML-NBs following Dox-induced R12 expression was not due to Dox treatment alone, as an empty vector tRhF cell line actually resulted in a statistically significant increase in the number of PML-NBs following Dox treatment, compared to mock-treated cells (Fig. 3B).

FIG 3.

RRV vIRF R12 colocalizes with PML-NBs. (A) Doxycycline (Dox)-inducible tRhFs expressing R12-FLAG (tRhF-R12) were grown in the presence or absence of 100 U/ml RhIFN-α2 with or without 2 μg/ml Dox for the indicated hours before fixation with methanol. Fixed cells were stained with FLAG (green)- and PML (red)-specific antibodies, while nuclei were stained with DAPI (blue). White arrows indicate cells with enlarged PML-NBs. Images at a ×40 magnification are presented. The experiment was performed three times, and representative images are shown. (B) Numbers of PML-NBs within nuclei of cells under each culture condition were counted and are presented in a graph. Data were analyzed by unpaired Student’s t test, and P values of less than 0.05 were considered significant. Experiments were performed twice, and data from a representative experiment are shown. (C) tRhF-R12 cells were grown as described above for panel A for the indicated times. Nuclear fractions were isolated, and protein lysates were analyzed by SDS-PAGE and probed with PML-, FLAG-, lamin A/C-, and GAPDH-specific antibodies. Experiments were performed three times, and data from representative experiments are presented. (D) tRhF-R12 cells were treated with Dox for 18 h and then fixed with methanol and stained with FLAG (green)- and PML (red)-specific antibodies. Images at a ×40 magnification are presented, along with a ×63 magnification and z-axis analysis of the cell, highlighted by the white box.

While the number of PML-NBs per nucleus declined in the presence of R12 protein expression, the size of the PML-NBs appeared to increase, suggesting that R12 may promote the reorganization of these subnuclear structures (Fig. 3A, white arrows). Western blot analysis of tRhF-R12 cells showed that levels of un-SUMOylated PML isoforms I and II decreased when R12 was expressed (Fig. 3C, first 3 lanes from the left). The apparent decrease in PML isoform I/II in Fig. 3C was not always reproducible and cannot be inferred. In addition to the presence of a protein of the predicted size of R12 (37 kDa), an additional larger protein species was also observed at approximately 75 kDa (Fig. 3C). While this may be a nonspecific cross-reaction with the FLAG antibody, it is also conceivable that R12 protein is posttranslationally modified.

To further investigate the interaction between R12-FLAG protein and PML-NBs, we analyzed the IFA results by z-stack analysis and confocal microscopy to better visualize the potential nuclear localization and interaction. Eighteen hours after Dox treatment, R12-FLAG protein is detectable, and z-stack analysis reveals that R12-FLAG colocalizes with PML-NBs and appears to be surrounded by and encased in PML protein (Fig. 3D).

To determine whether the localization of R12 with PML-NBs was due to a direct interaction between PML protein and R12 protein, we performed coimmunoprecipitation (co-IP) assays using the tRhF-R12 cell line treated with or without Dox to induce R12-FLAG expression and in the presence or absence of RhIFN-α2 to upregulate PML protein expression. Following separation of cytoplasmic and nuclear lysates, PML protein was immunoprecipitated from the nuclear lysates, resolved by SDS-PAGE, and then analyzed by Western blot analysis for R12-FLAG protein (Fig. 4A). By this analysis, we observed that FLAG-tagged R12 protein coimmunoprecipitated with PML protein only after Dox induction, regardless of RhIFN-α2 treatment (Fig. 4A). The ∼37-kDa R12-FLAG protein was the only form of R12-FLAG that coimmunoprecipitated with PML protein. Of note, the presence of R12-FLAG did not inhibit PML protein from coimmunoprecipitating with SP100 or Daxx, as both were detected (data not shown).

FIG 4.

R12 protein coimmunoprecipitates with PML protein. (A) tRhF-R12 cells were grown and nuclear fractions were isolated as described in the legend of Fig. 3C. Nuclear protein lysates were immunoprecipitated (IP) with a PML-specific antibody before resolving proteins on SDS-PAGE gels. Western blots were probed with PML- and FLAG-specific antibodies. (B) Nuclear protein lysates from tRhF-R12 cells cultured as described above for panel A were immunoprecipitated with a FLAG-specific antibody before resolving the proteins on SDS-PAGE gels. Western blots were probed with PML- and FLAG-specific antibodies. < denotes high-molecular-weight R12-FLAG protein, and * denotes the predicted size of the R12-FLAG protein. (C) Input control for panels A and B. Total nuclear lysates were subjected to SDS-PAGE, and Western blots were probed with PML-, FLAG-, lamin A/C-, and GAPDH-specific antibodies. Experiments were performed at least twice, and representative Western blots are shown.

A reciprocal co-IP assay was also performed to substantiate these results. Immunoprecipitation of FLAG-tagged R12 resulted in the detection of high-molecular-weight PML protein by Western blot analysis (Fig. 4B). Importantly, the interaction of R12 with PML protein appears to be specific, as R12-FLAG does not copurify with SP100 or Daxx (data not shown). Input control Western blotting of total nuclear lysates from this experiment confirmed that PML proteins were expressed under every condition tested and that both the 37-kDa and the 75- to 80-kDa R12 proteins were expressed upon Dox treatment (Fig. 4C).

R12 protein is SUMO-1 modified.

To further investigate the interaction between PML protein and R12-FLAG protein, tRhF-R12 cells were transiently transfected with a His₆–SUMO-1 expression plasmid for 2 days before cells were treated with or without Dox. Total cell lysates were then applied to nickel columns under denaturing conditions in order to isolate any cellular protein modified by His₆–SUMO-1. Isolated proteins were resolved on SDS-PAGE gels, and Western blot analysis with anti-FLAG antibody revealed that R12-FLAG protein is SUMO-1 modified. Under these conditions, only the ∼80-kDa R12-FLAG protein was SUMO-1 modified (Fig. 5A). Similarly, we confirmed that PML proteins are SUMO-1 modified, including the high-molecular-weight bands that appear to coimmunoprecipitate with R12-FLAG protein (Fig. 5A). Input protein levels are displayed in Fig. 5B. This is consistent with previous publications that have shown that SUMOylation of PML is necessary for PML to interact with other proteins and to form PML-NBs (54). Additional studies have also found that a SUMO-interacting motif (SIM) contained within the PML protein may be necessary for the interaction of PML with other SUMOylated proteins (55). While prediction software identified four possible SIMs in the R12 protein sequence, our analysis of immunoprecipitated R12-FLAG protein found that R12 protein was SUMO-1 modified (Fig. 5A and data not shown).

FIG 5.

R12 protein is SUMO-1 modified. (A) tRhF-R12 cells were either mock transfected or transfected with a His₆–SUMO-1 plasmid and grown with or without 2 μg/ml Dox. His₆–SUMO-1 proteins were purified from nuclear lysates using nickel columns and resolved on SDS-PAGE gels, followed by Western blot analysis with FLAG-, PML-, and SUMO-1-specific antibodies. (B) Input controls for panel A. Total nuclear lysates treated as described above for panel A were resolved on SDS-PAGE gels, and Western blots were probed with PML-, SUMO-1-, FLAG-, lamin A/C-, and GAPDH-specific antibodies. < denotes high-molecular-weight R12-FLAG protein, and * denotes the predicted size of the R12-FLAG protein.

Exogenous R12 protein expression during RRV_vIRF-KO infection results in the loss of PML-NBs.

Because RRV_vIRF-KO infection did not result in the loss of PML-NBs and R12 appears to interact with PML-NBs, we next asked whether exogenous expression of R12 could rescue the WT RRV phenotype (loss of PML-NBs) during infection with RRV_vIRF-KO. To test this, the tRhF-R12 cell line was infected with RRV_vIRF-KO at an MOI of 2 for 24 h in the presence or absence of Dox (Fig. 6). IFA revealed that only when R12-FLAG expression was induced by Dox treatment did RRV_vIRF-KO infection result in a loss of PML-NBs, similar to what is observed during WT RRV infection (Fig. 6A). Quantification of IFA images confirmed that in the absence of Dox, RRV_vIRF-KO infection did not result in any gB⁺/PML-NB-negative cells (Fig. 6B). However, Dox treatment and R12-FLAG expression during RRV_vIRF-KO infection resulted in 91% of gB⁺ cells lacking PML-NBs (Fig. 6B). Similar to what is observed during WT RRV infection, the number of PML-NBs within nuclei of RRV_vIRF-KO gB⁺ cells with R12 expression was significantly lower (average of 2.2) than during RRV_vIRF-KO infection without R12 expression (average of 18.64) (Fig. 6C).

FIG 6.

Stable expression of R12 protein during RRV_vIRF-KO infection results in a loss of PML-NB structures. (A) tRhF-R12 cells were grown in the presence or absence of 2 μg/ml Dox for 18 h, before mock infection or infection with RRV_vIRF-KO at an MOI of 2 for 24 h. Cells were fixed with methanol and stained with antibodies specific for PML (red), FLAG (green), and RRV-gB (purple), and nuclei were stained with DAPI (blue). Images at a ×63 magnification were obtained by confocal microscopy. White boxes indicate RRV-gB-positive cells, and yellow boxes indicate R12-expressing cells within uninfected cultures. The experiment was performed four times, and representative images are shown. (B) IFAs were performed as described in the legend of Fig. 1C. Data were analyzed by two-tailed Fisher’s exact test, and P values of less than 0.05 were considered significant. (C) IFAs were performed as described in the legend of Fig. 1D. Data were analyzed by unpaired Student’s t test, and P values of less than 0.05 were considered significant. (D) R12-FLAG-inducible cells were grown and treated as described above for panel A and infected with RRV_vIRF-KO at an MOI of 2 for 24 h. Nuclear and cytoplasmic fractions were isolated, and proteins were separated and resolved on SDS-PAGE gels, followed by Western blot analysis. Western blots were probed with PML-, FLAG-, lamin A/C-, RRV-ORF52-, and GAPDH-specific antibodies.

Finally, the tRhF-R12 cell line was mock treated or treated with RhIFN-α2, in the presence or absence of Dox, and infected with RRV_vIRF-KO at an MOI of 2 for 24 h. Nuclear and cytoplasmic protein lysates were separated and resolved on SDS-PAGE gels. Western blot analysis revealed that PML protein levels during RRV_vIRF-KO infection were much lower when R12-FLAG was expressed than when R12-FLAG was not expressed during infection (Fig. 6D). This confirmed that R12 is involved in the loss of PML-NBs and PML protein levels during RRV infection.

R12 protein expression inhibits ISG transcription.

Our data imply that levels of PML isoforms I and/or II are decreased following WT RRV infection, and this is dependent on R12 protein expression. We have also shown that exogenous R12 expression, alone, can reduce the number of PML-NBs. Thus, we next wanted to define the downstream effects of R12 protein expression on the cell. Previous reports state that PML isoform II is necessary for efficient ISG transcription induction, in particular targets of the ISGF3 transcription complex (15). As such, we analyzed ISG induction, downstream of IFN receptor signaling, during exogenous R12-FLAG protein expression. To investigate this, the tRhF-R12 cell line was treated with or without Dox for 6 h and then further cultured in the presence or absence of 100 U/ml of RhIFN-α2. Empty vector cells were treated similarly as a control for off-target effects of Dox and utilized as a comparison for the analysis (Fig. 7). RNA was purified from the cells and analyzed by quantitative reverse transcriptase (qRT)-PCR. We found that expression of R12 protein during RhIFN-α2 treatment led to statistically significant reductions in IP-10, IRF7, and IFN-β transcripts compared to those during RhIFN-α2 treatment alone, while the empty vector cell line treated identically did not display similar decreases in the presence of Dox (Fig. 7). Surprisingly, Dox treatment of the empty vector cells in the presence of RhIFN-α2 led to statistically significant increases in IRF7 and IFN-β levels, compared to RhIFN-α2 alone (Fig. 7).

FIG 7.

R12 protein expression reduces ISG transcription in the presence of type I IFN. (A) tRF-R12 cells were mock treated or treated with 100 U/ml RhIFN-α2 for 6 h, 2 μg/ml Dox for 12 h, or Dox for 6 h before RhIFN-α2 was added for an additional 6 h. The same experiment was repeated in a tRhF empty vector cell line. Transcript levels were normalized to GAPDH transcript levels and are presented as fold changes over values with RhIFN-α2 treatment. Data were analyzed by unpaired Student’s t test. P values of less than 0.05 were considered significant, and asterisks denote significant P values (*, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001). Experiments were performed in duplicate, and the averaged results are displayed.

Construction of recombinant R12 mutant RRV.

To examine the contributions of R12 in the context of viral infection, we constructed two recombinant viruses using the RRV₁₇₅₇₇ BAC (RRV_BAC) system. An R12 nonsense mutant RRV (RRV_R12ns) converting the start codon into a stop codon in the R12 ORF and encoding a FLAG epitope tag at the C terminus of the R12 ORF was constructed. Additionally, an R12 rescue mutant RRV (RRV_R12FLAG) encoding FLAG-tagged R12 but lacking any other mutations was constructed so that endogenous R12 could be detected following infection (Fig. 8A). Following the construction of the mutant RRV BAC DNA clones using recombination to insert the mutated R12 sequences, the clones were screened by restriction endonuclease digestion and Southern blot analysis (Fig. 8B and C). Bacmid clones were also screened by PCR and DNA sequence analysis to confirm the correct insertion into the RRV_BAC, and the resulting bacmid clones were used to generate infectious recombinant virus by transfection into RhF cells. The BAC cassette was removed from the recombinant viruses using CRE recombinase and the loxP sites flanking the BAC cassette. After growth and purification of recombinant viruses, Western blot analysis was performed on nuclear lysates from tRhFs infected with both recombinant viruses and indicated that RRV_R12FLAG expresses R12 FLAG-tagged protein, while RRV_R12ns lacks R12 protein expression (Fig. 8D). Furthermore, RT-PCR analysis of RNA from cells infected with RRV_R12FLAG or RRV_R12ns indicates that the ORFs immediately upstream and downstream of R12 (R11 and R13) are expressed and that the altered R12 sequence in both viruses does not affect the transcription of neighboring genes (Fig. 8E). Finally, one-step and multistep growth curve analyses of WT RRV, RRV_vIRF-KO, RRV_R12ns, and RRV_R12FLAG revealed that all four viruses display the same growth kinetics in vitro, indicating the mutations do not affect the growth of virus in vitro (Fig. 8F).

FIG 8.

RRV_R12FLAG and RRV_R12ns construction and characterization. (A) Schematic of the mutations introduced into the RRV genome to generate RRV_R12ns and RRV_R12FLAG. (B) BamHI restriction digests of WT RRV_BAC, RRV-R12-GalK-KO BAC, RRV_R12ns BAC, and RRV_R12FLAG BAC DNA clones. * denotes the location of the digest band containing the R12 sequence, and < denotes the increased size of the digest band that contains the GalK cassette. (C) Southern blot analysis of BamHI-digested BAC clones probed with GalK- and R12-specific probes. (D) Nuclear and cytoplasmic protein lysates from mock-infected and WT RRV_BAC-, RRV_vIRF-KO-, RRV_R12ns-, and RRV_R12FLAG-infected cells (MOI of 2) were collected at 24 h; resolved on SDS-PAGE gels; and probed with FLAG-, lamin A/C-, RRV-ORF52-, and GAPDH-specific antibodies. (E) RT-PCR analysis of total cellular RNA isolated from WT RRV_BAC-, RRV_R12ns-, or RRV_R12FLAG-infected cells with or without reverse transcriptase enzyme using R11-, R13-, and GAPDH-specific primers. PCR products were subjected to agarose gel electrophoresis. (F) One-step (MOI = 2) or multistep (MOI = 0.1) growth analysis of WT RRV_BAC, RRV_vIRF-KO, RRV_R12ns, or RRV_R12FLAG. Viral titers at each time point were determined by a plaque assay and are presented as PFU per milliliter. Experiments were performed twice, and data from representative experiments are shown, except for panel F, where the average titer from both experiments is graphed.

Endogenous R12 complexes with PML protein and is necessary for PML-NB disruption during viral infection.

To characterize the kinetics of endogenous R12 expression during RRV infection, we infected tRhFs with RRV_R12FLAG at an MOI of 2 and harvested RNA from infected cells every 2 h during the first 24 h of infection. RT-PCR revealed R12 transcripts at every time point tested, even as early as 2 hpi (Fig. 9A). Analysis of R12 protein expression by Western blotting, during a time course of infection with RRV_R12FLAG, revealed that endogenous R12 protein can be detected by 6 hpi, with the 80-kDa R12-FLAG protein being the most abundant form. As expected, infection with RRV_R12ns did not result in R12-FLAG protein expression (Fig. 9B).

FIG 9.

Virus-expressed R12 coimmunoprecipitates with PML protein and is required for RRV disruption of PML-NBs. (A) RNA was purified from mock-infected, WT RRV_BAC-infected (24 hpi), or RRV_R12FLAG-infected cells at the indicated hours postinfection. RT-PCR was performed with or without reverse transcriptase enzyme using R12- and GAPDH-specific primers, and PCR products were subjected to agarose gel electrophoresis. (B) Nuclear lysates from mock-infected, RRV_R12FLAG-infected, and RRV_R12ns-infected cells (MOI of 2) were collected at the indicated times postinfection; resolved on SDS-PAGE gels; and probed with FLAG-, lamin A/C-, and GAPDH-specific antibodies. (C) WT RRV_BAC-, RRV_R12FLAG-, RRV_R12ns-, or RRV_vIRF-KO-infected cells were fixed after 24 hpi and stained with RRV-gB- and PML-specific antibodies, while nuclei were stained with DAPI. IFAs were performed as described in the legend of Fig. 1C, and data were analyzed by two-tailed Fisher’s exact test, with P values of less than 0.05 being considered significant. (D) IFAs were further performed as described in the legend of Fig. 1D, and data were analyzed by unpaired Student’s t test, with P values of less than 0.05 being considered significant. (E) Nuclear and cytoplasmic fractions were isolated from mock-infected or RRV_R12FLAG-infected cells at 6, 10, 14, 16, 18, and 24 hpi. Nuclear lysates were immunoprecipitated with a FLAG-specific antibody, resolved by SDS-PAGE, and analyzed by Western blotting using a PML-specific antibody. Nuclear and cytoplasmic proteins were analyzed by SDS-PAGE and Western blotting using FLAG-, PML-, lamin A/C-, RRV-ORF52-, and GAPDH-specific antibodies as controls for protein analysis. The experiment in panel A was performed twice, and experiments in panels B to E were performed at least three times. Data from representative experiments and representative images are shown.

Next, tRhF cells were infected with RRV_R12ns, RRV_vIRF-KO, RRV_R12FLAG, or WT RRV at an MOI of 2 for 24 h before fixation and IFA to detect PML-NBs (data not shown). Quantification of the IFA images revealed that RRV_R12ns infection was similar to RRV_vIRF-KO infection, with 96.4% and 97.3% of gB⁺ cells containing PML-NBs, respectively, whereas RRV_R12FLAG infection and WT RRV infection were similar, with 11.5% and 19.5% of gB⁺ cells containing PML-NBs, respectively (Fig. 9C). The difference in gB⁺/PML-NB⁺ cells between RRV_R12ns and RRV_R12FLAG infections was statistically significant, as was that for RRV_vIRF-KO compared to WT RRV (Fig. 9C). Additionally, the numbers of PML-NBs left in the gB⁺ cells were significantly higher in the RRV_R12ns and RRV_vIRF-KO infections than in the RRV_R12FLAG and WT RRV infections (Fig. 9D).

To determine whether endogenous R12 protein produced during RRV infection interacts with PML protein, coimmunoprecipitation assays were performed in RRV_R12FLAG-infected cells. We found that when R12-FLAG was immunoprecipitated using a FLAG-specific antibody, PML protein could be detected in the immunoprecipitation lysates (Fig. 9E). The PML protein that copurifies with the R12-FLAG protein was of a high molecular weight, indicative of SUMOylation.

Disruption of PML-NBs during RRV infection inhibits ISG induction and aids RRV replication in the presence of type I IFN.

We have shown that exogenous R12-FLAG can inhibit ISG transcription following RhIFN-α2 treatment. To investigate whether endogenous R12 protein could serve a similar role during RRV infection, we infected tRhFs with WT RRV, RRV_R12FLAG, RRV_vIRF-KO, or RRV_R12ns for 18 h at an MOI of 2, followed by the addition of 100 U/ml of RhIFN-α2 to the culture medium. The infections continued for an additional 6 h before harvesting of total RNA to determine the induction of IP-10, IRF7, IFN-β, and a non-IFN-regulated gene (RPL32) (Fig. 10A). The qRT-PCR analysis revealed that WT RRV and RRV_R12FLAG were able to suppress the induction of IP-10, IRF7, and IFN-β following the addition of type I IFN to the infected cells (Fig. 10A). However, both RRV_vIRF-KO and RRV_R12ns infections induced significantly higher levels of IP-10, IRF7, and IFN-β transcripts than did WT RRV and RRV_R12FLAG infections (Fig. 10A). This revealed that WT RRV blocks ISG induction after IFN-α signals through the IFN-α/β receptor and is dependent on R12 expression.

FIG 10.

RRV R12 inhibits ISG induction downstream of IFN signaling and aids RRV replication in the presence of IFN. (A) tRhFs were infected with WT RRV_BAC, RRV_vIRF-KO, RRV_R12FLAG, or RRV_R12ns at an MOI of 2 for 18 h. Afterwards, 100 U/ml RhIFN-α2 was added to the infected cell culture media for an additional 6 h. RNA was purified, and cDNA was synthesized before transcript levels were measured by quantitative PCR. Transcript levels were normalized to GAPDH transcript levels and are presented as fold changes relative to the values for the RhIFN-α2 sample. Data were analyzed by unpaired Student’s t test. P values of less than 0.05 were considered significant, and asterisks denote significant P values (*, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001). (B) PML protein levels were knocked down using shRNA specific for PML. Nuclear lysates from tRhF cells, shPML-tRhF cells, and shControl-tRhF cells treated with 100 U/ml RhIFN-α2 for 18 h were resolved on SDS-PAGE gels, and Western blots were probed with PML- and lamin A/C-specific antibodies. PML protein bands were quantified by densitometry. (C) The shPML cell line was infected with WT RRV_BAC, RRV_vIRF-KO, RRV_R12FLAG, or RRV_R12ns identically as described above for panel A and treated with or without 100 U/ml RhIFN-α2 for an additional 6 h. As controls, tRhF or shControl tRhF cells were treated with 100 U/ml RhIFN-α2 for 6 h. RNA was purified, and cDNA was synthesized before transcript levels were measured by quantitative PCR. Transcript levels were normalized to GAPDH transcript levels and are presented as fold changes over values for shControl cells treated with RhIFN-α2 for 6 h. Data were analyzed by unpaired Student’s t test. P values of less than 0.05 were considered significant, and asterisks denote significant P values. Experiments were performed in duplicate, and averaged results are shown. (D) Primary RhFs were infected with WT RRV, RRV_vIRF-KO, RRV_R12FLAG, or RRV_R12ns at an MOI of 2.5 in the presence of 100 U/ml RhIFN-α2. Viral titers were measured at the indicated times postinfection by a plaque assay and are presented as PFU per milliliter. Data were analyzed by repeated-measures ANOVA with a post hoc Tukey-Kramer test. Adjusted P values of less than 0.05 were considered significant, and asterisks denote significant P values. 1, WT RRV versus RRV_vIRF-KO and RRV_R12FLAG versus RRV_vIRF-KO; 2, WT RRV versus RRV_R12ns and RRV_R12ns versus RRV_R12FLAG; 3, RRV_R12ns versus WT RRV; 4, WT RRV versus RRV_vIRF-KO and RRV_R12ns versus RRV_R12FLAG; 5, RRV_R12FLAG versus RRV_vIRF-KO. The experiment was performed twice, and the averages from both experiments are graphed.

We next wanted to determine whether the RRV R12 block in ISG transcription was due to the loss of PML observed during WT RRV and RRV_R12FLAG infections. To accomplish this, we utilized short hairpin RNA (shRNA) against PML (shPML) to knock down PML protein in tRhF cells (Fig. 10B). Using this shPML cell line, we were able to determine if the inability of RRV_vIRF-KO and RRV_R12ns infections to inhibit ISG induction was connected to the lack of PML-NB disruption by these two viruses. The shPML cell line was subsequently infected with WT RRV, RRV_R12FLAG, RRV_vIRF-KO, or RRV_R12ns for 24 h at an MOI of 2 with or without treatment with RhIFN-α2 for the last 6 h (Fig. 10C). qRT-PCR analysis showed that when PML protein was knocked down, every virus failed to robustly induce IP-10, IRF7, and IFN-β transcripts. Thus, the difference in transcript induction observed in the tRhF (PML-intact) cells was due to the differential abilities of the mutant RRVs to disrupt PML-NBs.

Finally, we wanted to determine whether the expression of R12 would aid viral replication in the presence of IFN. Thus, we performed a one-step growth curve analysis in the presence RhIFN-α2 with WT RRV, RRV_vIRF-KO, RRV_R12ns, and RRV_R12FLAG (Fig. 10D). Both the RRV_vIRF-KO and RRV_R12ns growth curves had statistically significantly lower viral titers (1/2 to 1 log lower) at 24 and 48 hpi than the WT RRV and RRV_R12FLAG growth curves.

DISCUSSION

Accumulating evidence indicates that PML-NBs are important and essential for the efficient and robust induction of the innate immune response upon pathogen infection. For viruses to establish a successful acute or chronic infection, they must employ mechanisms to circumvent the intrinsic cellular host innate response. We previously reported that animals infected with recombinant RRV_vIRF-KO displayed reduced viral loads in the periphery and an increased type I IFN response in serum compared to WT RRV_BAC-infected RMs, even though WT RRV_BAC and RRV_vIRF-KO showed similar growth kinetics in vitro. Because of this, we decided to investigate the role that type I IFN and PML-NBs played in restricting RRV growth and how the vIRFs, specifically the vIRF R12, counteracted this host defense. Here, we provide evidence that RRV encodes a protein that interacts with PML-NBs to facilitate disruption of PML-NBs during RRV infection.

Utilizing inducible viral gene expression and molecular bacterial artificial chromosome (bacmid) virology, we found that vIRF R12 is necessary for the RRV-induced disruption of PML-NBs during de novo lytic infection. This is important because in the natural host infection setting, RRV will encounter not only the viral restriction capacity of PML-NBs but also type I IFN signaling from the type I IFN produced from neighboring uninfected cells. The capacity of RRV to disrupt PML-NBs has a 2-fold effect: (i) it overcomes the viral genetic silencing exerted by PML-NB-resident proteins, and (ii) it inhibits the type I IFN signaling cascade to prevent ISGs (with their own antiviral effects) from being expressed. The second effect, inhibiting type I IFN signaling, would explain why WT RRV could grow to significantly higher titers in the presence of RhIFN-α2 than RRV_vIRF-KO and RRV_R12ns. While we found that R12 is necessary for RRV to disrupt PML-NBs, R12 expression alone did not result in the complete loss of PML-NBs. R12 protein expression outside the context of infection appeared to affect PML-NB organization, as there were fewer but larger PML-NBs, and R12 was sufficient for the correlated decrease in ISG expression.

HSV-1 ICP0 and HCMV IE1 have been shown to modulate PML proteins and inhibit PML-NB formation. In our inducible stable expression culture, R12 appeared to modify PML-NB formation or stability, as there were fewer PML-NBs when R12 protein was expressed. Additionally, R12 protein expression during RRV_vIRF-KO infection could restore the full disruption and loss of PML-NBs, suggesting a critical role for R12 in this process. A previous publication on PML-NB disruption by RRV implicated the RRV tegument protein encoded by ORF75 and found that while RRV infection of RhF cells resulted in a loss of SP100 protein by 8 hpi, PML protein was not lost until 24 hpi, and this could be rescued with cycloheximide treatment (24). Thus, the inhibition of R12 expression by cycloheximide treatment could explain why PML protein levels were rescued. The involvement of both ORF75 and a vIRF in the disruption of PML-NBs by RRV is similar to what is observed for KSHV disruption of PML-NBs. KSHV ORF75 induces the loss of ATRX protein and causes the dispersal of Daxx protein from PML-NBs (23). KSHV vIRF3 increases SUMO-modified PML protein levels, leading to SUMO-dependent ubiquitination and, eventually, degradation of PML protein (18). Therefore, while RRV may be similar to KSHV in utilizing at least two viral proteins to disrupt PML-NBs, the mechanisms may have diverged. Additional hypotheses to explain the R12 requirement for PML-NB disruption by RRV include the following: R12 expression may be necessary for ORF75 transcription, R12 may be required for ORF75 localization to PML-NBs by complexing with both ORF75 and PML proteins, or R12 may be required for the expression of another unknown protein that works with ORF75 to disrupt PML-NBs. While further investigation is needed to decipher the mechanism of PML-NB disruption by R12, we have analyzed infected tRhFs for ORF75 transcripts. Using RT-PCR, we found that WT RRV, RRV_vIRF-KO, RRV_R12FLAG, and RRV_R12ns infections produced ORF75 transcripts (data not shown). Therefore, R12 (or any other vIRF) is not necessary for the transcription of ORF75 to occur.

Our data provide evidence that RRV disrupts PML-NBs by reducing protein levels of PML isoforms I and II in an R12-dependent manner. PML isoform II has been implicated in the efficient induction of ISGs that are transcriptionally regulated by the ISGF3 transcription complex. When PML isoform II is absent from cells, ISGF3 targets are not as strongly induced following stimulation with IFN-α or poly(I·C) (15). In line with this role of PML isoform II, we found that only WT RRV and RRV_R12FLAG, and not RRV_vIRF-KO or RRV_R12ns, were able to inhibit IP-10, IRF7, and IFN-β induction following the addition of RhIFN-α2 to the infected cell culture medium (Fig. 10A). The same experiment performed on cells with PML knocked down by shRNA resulted in a similar reduction of ISG transcription for all viruses tested. This supports the theory that the inhibition of ISG transcription following RhIFN-α2 treatment during WT RRV infection is a result of the loss of PML-NBs. Additionally, R12 expression outside the context of viral infection could reduce ISG transcription following treatment with RhIFN-α2, providing further evidence that R12 is involved in this phenotype during RRV infection.

Inhibition of ISG induction when faced with type I IFN would maintain a cellular environment more conducive to viral infection and replication. We were able to demonstrate this using viral growth curve analysis in the presence of type I IFN. The viruses that were able to disrupt PML-NBs and inhibit ISG transcription (WT RRV and RRV_R12FLAG) displayed significantly increased viral titers at 24 and 48 hpi compared to those of viruses that were unable to disrupt PML-NBs (RRV_vIRF-KO and RRV_R12ns) (Fig. 10D). Taken together, we conclude that R12 expression during de novo lytic infection leads to a reduction in protein levels of PML isoform I/II, aiding the disruption of PML-NBs, which results in the inhibition of transcription of ISGs regulated by the ISGF3 transcription complex, even in the presence of type I IFN signaling. This inhibition of ISGs allows RRV to effectively establish infection when the type I IFN response is activated, as we would expect during in vivo infection. The data that we have presented in this study provide evidence that the RRV vIRF R12 plays a role in the disruption of PML-NBs during infection with RRV, and while R12 is necessary, it is not sufficient for PML-NB loss.

MATERIALS AND METHODS

Cells, virus, drugs, and cytokines.

Primary rhesus fibroblasts (RhFs) and telomerized RhFs (tRhFs) (32) were grown in Dulbecco’s modified Eagle’s medium (DMEM) (Mediatech, Manassas, VA) supplemented with 10% fetal bovine serum (FBS) (HyClone, Ogden, UT). Inducible R12-FLAG cells (tRhF-R12) were grown in DMEM supplemented with 10% tetracycline-free FBS, 3 μg/ml puromycin (Sigma-Aldrich, St. Louis, MO), and 300 μg/ml hygromycin B (Invitrogen, Carlsbad, CA). Human BJAB cells were grown in RPMI 1640 (Mediatech) supplemented with 10% FBS. RRV infections were performed in complete DMEM or RPMI medium with 5 μg/ml Polybrene; following a 2-h adsorption period, cells were washed twice with Dulbecco’s phosphate-buffered saline (DPBS) (Mediatech) to remove unbound virus, and fresh medium was added. These studies utilized plaque-purified isolates of bacterial artificial chromosome (BAC)-derived RRV₁₇₅₇₇ (WT RRV_BAC) (16), WT RRV-GFP, RRV_vIRF-KO, and RRV_vIRF-KO-GFP. RRV_vIRF-KO, WT RRV-GFP, and RRV_vIRF-KO-GFP were previously reported (51). All RRV stocks were purified through a 30% sorbitol cushion and resuspended in PBS, and viral titers were determined by using a serial dilution plaque assay with RhFs.

Doxycycline hydrochloride (ThermoFisher Scientific, Waltham, MA) was resuspended in dimethyl sulfoxide (DMSO) (ThermoFisher Scientific) and added to cell culture media at 2 μg/ml every 24 h. Rhesus interferon alpha 2 (RhIFN-α2) (PBL Assay Science, Piscataway, NJ) was used at a final concentration of 100 U/ml.

Construction of RRV_R12ns and RRV_R12FLAG.

The RRV₁₇₅₇₇ BAC galK positive/negative-selection system was utilized to create two mutant RRVs; the first was engineered to replace the start codon of R12 with a stop codon to create a nonsense mutation to prevent the expression of R12 (RRV_R12ns). A C-terminal FLAG epitope tag was also introduced just before the native termination codon of R12 to ensure no readthrough expression. The second recombinant replaced the nonsense mutation with a start codon, to essentially create a revertant of RRV_R12ns. The revertant harbors the C-terminal FLAG epitope tag to follow the R12 protein, and we termed this virus RRV_R12FLAG. This system of generating mutant RRVs was previously described (56). Briefly, R12 and R13 ORFs (nucleotides 87625 to 90478) in the RRV₁₇₅₇₇ BAC were replaced with galK, as described above (R12 5′-flanking primer sequence 5′-TTTATTGCAGGGACAGGGCAAAAGCAAGCTGTGCACGGTAACAGTATGTGTCAGCACTGTCCTGCTCCT-3′ and R13 3′-flanking sequence 5′-TAGGGGAGTGGTGAGGGCTTTTGAGTTAGTTTTCGTGGACCAAGTTCACACCTGTTGACAATTAATCATCGGCA-3′ [sequences homologous to the galK cassette are underlined]). Next, we cloned the R12 and R13 ORFs with 250 bp of flanking regions into the psP73 vector using EcoRI and HindIII restriction sites engineered into the primers (250 bp upstream of R12 primer 5′-CGGAATTCGCCTAACTATATACGCCCACGGG-3′ and 250 bp downstream of R13 primer 5′-CGAAGCTTGCTTGGTGCCCTTTAAATTGAACG-3′ [restriction sites are underlined]). Using QuikChange II XL site-directed mutagenesis (Agilent Technologies, La Jolla, CA) according to the manufacturer’s specifications, we inserted a FLAG epitope just before the stop codon of R12 using the following primers: forward primer 5′-GTATGTGTCACTTGTCATCGTCATCCTTGTAGTCCTGGGCCGCATCC-3′ and reverse primer 5′-GGATGCGGCCCAGGACTACAAGGATGACGATGACAAGTGACACATAC-3′ (the FLAG epitope sequence is underlined). The nonsense mutation to the R12 ORF was accomplished using the following primers once the R12-FLAG-tagged plasmid was obtained: forward primer 5′-GCCCGTCCTTCCGCTCACTCTGAGGGTCCGCTCGC-3′ and reverse primer 5′-GCGAGCGGACCCTCAGAGTGAGCGGAAGGACGGGC-3′ (the mutated start codon is underlined). Finally, the R12-mutated R12-R13 repair cassette with flanking regions was digested out of the psP73 plasmid and used to repair the galK R12-R13 knockout RRV₁₇₅₇₇ BAC as described previously (57). After identification of repaired BAC clones by Southern blotting, PCR, and sequencing, a clone was used to transfect RhFs to make virus, and the BAC cassette was removed by CRE recombination as described previously (57). Each virus was plaque purified twice before viral stocks were grown and purified over a 30% sorbitol cushion and resuspended in PBS. Insertion junctions and the R12 ORF were PCR amplified and sequenced from each virus to confirm that the correct mutations were present and that there were no other alterations to the viral genome in these locations.

In vitro growth curves.

One-step (MOI = 2.5) and multistep (MOI = 0.1) growth curve analyses were carried out with RhFs, essentially as described previously (51). For growth curves in the presence of RhIFN-α2, cells were seeded in culture tubes in the presence of 100 U/ml of RhIFN-α2, which was kept on the cells throughout the infection time course. Every 24 h, an additional 50 U/ml RhIFN-α2 was added to the culture tubes to ensure active IFN signaling throughout. Residual IFN carried over from the culture tubes onto the titer plates was measured by an IFN bioassay and IFN enzyme-linked immunosorbent assays (ELISAs). Briefly, the UV-inactivated supernatant from the viral growth curve tubes was added to telomerized rhesus fibroblasts expressing luciferase under the control of the interferon-stimulated response element (ISRE) promoter. Measurements of luciferase expression were converted to units per milliliter with the use of a standard curve on the same 96-well plate as for the tested samples. Two ELISAs were performed on culture media from the viral growth curve tubes, one for IFN beta and one for pan-IFN alpha, according to the manufacturers’ protocols (IFN beta, R&D Systems Inc., Minneapolis, MN; IFN alpha, Mabtech AB, Sweden). The highest concentrations of IFN were measured with the ELISAs, and this concentration was tested in a plaque assay to determine effects on virus growth.

RNA isolation, RT-PCR, and real-time RT-PCR.

RNA was isolated from uninfected or infected tRhF cells using the Quick-RNA miniprep kit and the RNA Clean & Concentrator-5 kit (Zymo Research, Irvine, CA). DNA was removed following in-column DNase I enzyme treatment followed by a second out-of-column DNase I treatment according to the protocols included in the kits. RT-PCR was performed by using Superscript III one-step RT-PCR with Platinum Taq (Invitrogen). Transcripts were detected with the following specific oligonucleotide pairs: ORF R13 forward primer 5′-GGCGGCCCTGGCATATACGG-3′ and ORF R13 reverse primer 5′-CCGAGGTATGAGTGGCATGCAACC-3′, ORF R11 forward primer 5′-AACCGGTGCACCGACAGTCGC-3′ and ORF R11 reverse primer 5′-CCGTGTCCTCTCGAAAACATC-3′, ORF R12 forward primer 5′-ATTGTTGCGATAATGATAAGC-3′ and ORF R12 reverse primer 5′-CCGGTGGCATCCGCTTCGTTA-3′, and ORF75 forward primer 5′-GCGGACATGACAGTTTCCCCGTGGG-3′ and ORF75 reverse primer 5′-TTACTGTCTGTTTCTTATGC-3′.

First-strand cDNA synthesis was carried out using Superscript III reverse transcriptase for qRT-PCR (Invitrogen), and cDNA was subsequently amplified using Power SYBR green master mix (Applied Biosystems, Waltham, MA). Concentrations of target transcripts were determined using a standard curve included on each plate, consisting of serial dilutions of cDNA obtained from RhIFN-α2-stimulated tRhFs. All data were normalized to the levels of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in each sample, and normalized levels of target transcripts are presented as fold changes over values for mock-treated cells.

Immunoprecipitation, PAGE analysis, and immunoblotting.

Nuclear and cytoplasmic cell lysates were separated according to kit protocols (NE-PER; ThermoFisher Scientific). Nuclear lysates were immunoprecipitated with an anti-FLAG M2 monoclonal antibody (mAb) (catalog number F3165; Sigma-Aldrich) or an anti-PML (H238) polyclonal antibody (pAb) (Santa Cruz Biotechnology, Dallas, TX) in native lysis buffer (50 mM Tris-Cl [pH 8.0], 1% NP-40, and 150 mM NaCl supplemented with protease inhibitors [100× cocktail; Sigma-Aldrich]), followed by incubation with protein A/G Plus-agarose (Santa Cruz Biotechnology), and lysates were finally collected in radioimmunoprecipitation assay (RIPA) buffer (native lysis buffer with 0.1% SDS and 0.5% sodium deoxycholate). Immunoprecipitation (IP) assays to analyze SUMO modifications on PML and R12 proteins were performed by harvesting whole-cell lysates in RIPA buffer with 1% SDS and immediately boiling the samples for 5 min. RIPA buffer with no SDS was added to the protein lysates to bring the final SDS concentration to 0.1% before PML and R12-FLAG proteins were immunoprecipitated with protein A/G Plus-agarose. Whole-cell extracts were also collected in RIPA buffer, nuclear and cytoplasmic lysates were collected according to kit protocols (NE-PER; ThermoFisher Scientific), and all samples were analyzed on Bolt 4-to-12% gradient Bis-Tris Plus protein gels (Invitrogen). Proteins were then transferred onto a polyvinylidene difluoride (PVDF) membrane (Bio-Rad Laboratories, Hercules, CA) via semidry transfer (60 min at 15 V at room temperature).

Membranes were probed with anti-human PML mAb (PG-M3; Santa Cruz Biotechnology), anti-FLAG M2 mAb (catalog number F3165; Sigma-Aldrich), anti-FLAG M2 (horseradish peroxidase [HRP]) mAb (catalog number A8592; Sigma-Aldrich), anti-human GADPH mAb (catalog number 51906; Santa Cruz Biotechnology), anti-RRV glycoprotein B (gB) mAb (clone 10B5.2; VGTI Monoclonal Antibody Core), anti-RRV-ORF52 mAb (clone 3G9.2; VGTI Monoclonal Antibody Core), anti-human SP100 pAb (catalog number 43151; Abcam, Cambridge, UK), anti-human Daxx pAb (catalog number 105173; Abcam), anti-lamin A/C mAb (E-1) (catalog number 376248; Santa Cruz Biotechnology), anti-SUMO-1 (Y299) mAb (catalog number 32058; Abcam), and anti-SUMO-2/3 (18H8) (Cell Signaling Technology, Danvers, MA). Densitometry was performed using ImageJ software.

SUMO-1 protein conjugate purification.

The pSG5-his-SUMO1 plasmid (plasmid 17271; Addgene, Watertown, MA) was transiently transfected into tRhF-R12 cells for 48 h before doxycycline was added to the cells for an additional 18 h. Proteins that were posttranslationally modified by the transfected His–SUMO-1 were isolated using a His SpinTrap column according to the manufacturer’s protocol (GE Healthcare, Chicago, IL). Briefly, cells were mechanically lysed by repeated freeze-thawing and sonication in binding buffer (20 mM Tris-HCl, 8 M urea, 500 mM NaCl, 5 mM imidazole [pH 8.0], 1 mM β-mercaptoethanol) and cleared of insoluble material by centrifugation. Supernatants were added to the His SpinTrap column and centrifuged at low speed, and the columns were then washed 3 times with binding buffer. SUMO-1-modified proteins that were retained within the nickel column were eluted with elution buffer (20 mM Tris-HCl, 8 M urea, 500 mM NaCl, 500 mM imidazole [pH 8.0], 1 mM β-mercaptoethanol).

Immunofluorescence analysis.

Cells were grown on glass coverslips in 12-well plates and fixed with methanol (20 min at −20°C). Cells were then blocked in 1% bovine serum albumin (BSA) in Tris-buffered saline (TBS) (1 h at room temperature) prior to staining, and all subsequent steps were performed with 1% BSA–TBS. Cells on coverslips were stained with rabbit anti-PML (H238) (Santa Cruz Biotechnology), goat anti-FLAG pAb (catalog number 1257; Abcam), mouse anti-RRV gB mAb (clone 10B5.2; VGTI Monoclonal Antibody Core), rabbit anti-SP100 (Abcam), or rabbit anti-Daxx (catalog number 07-471; MilliporeSigma, Burlington, MA) overnight at 4°C and subsequently stained with Alexa Fluor 594 anti-mouse IgG (catalog number A11020; Invitrogen), Alexa Fluor 633 anti-rabbit IgG (catalog number A21071; Invitrogen), and Alexa Fluor 488 anti-goat IgG (catalog number A11055; Invitrogen) (1 h at room temperature), and nuclei were stained with DAPI (4′,6-diamidino-2-phenylindole). Cells on coverslips were mounted onto slides by using Vectashield (Vector Labs) and examined on a Zeiss Axio Imager.M1 microscope (Zeiss Imaging Solutions, Thornwood, NY) or a Leica SP5 Acousto-Optical Beam Splitter spectral confocal microscope (Leica Microsystems, Buffalo Grove, IL). Images were acquired by using a Zeiss AxioCam camera (MRm) with AxioVision software (version 4.6) and subsequently processed by using Adobe Photoshop (Adobe Systems, San Jose, CA).

Generation of a doxycycline-inducible stable cell line.

The pLVX lentiviral vector system (Clontech, Mountain View, CA) was utilized for constructing a stable doxycycline (Dox)-inducible cell line as described previously (50). Briefly, the pLVX-R12FLAG plasmid was constructed by subcloning full-length FLAG-tagged R12 from WT RRV_BAC DNA into the pLVX-Tight-Puro retroviral vector. Replication-defective recombinant retrovirus was produced in HEK 293T/17 cells and used to transduce the target cells (tRhFs containing a Dox-responsive transactivator [tRF-rtTAs]). Dox-inducible R12-FLAG cells (tRhF-R12) were maintained in DMEM plus 10% Tet-free FBS containing 3 μg/ml puromycin and 300 μg/ml hygromycin B. In order to determine the optimal concentration of Dox and duration of Dox treatment, cells were experimentally treated with Dox at various concentrations and for various lengths of time. We found that 2 μg/ml of Dox yielded half-maximal R12-FLAG expression at 18 h posttreatment.

Generation of short-hairpin RNA stable cell lines.

The PML knockdown stable cell line was constructed using Mission shRNA plasmid (pLKO.1) DNA purchased from Sigma-Aldrich. The shPML sequence was CCGGGTGTACCGGCAGATTGTGGATCTCGAGATCCACAATCTGCCGGTACACTTTTT. Nontarget control cells were constructed using the Mission pLKO.1-puro nontarget shRNA control plasmid from Sigma-Aldrich. The nontarget sequence was CCGGGCGCGATAGCGCTAATAATTTCTCGAGAAATTATTAGCGCTATCGCGCTTTTT.

Replication-defective recombinant retrovirus was produced in HEK 293T/17 cells and used to transduce the target cells (tRhFs). shPML and shControl cells were maintained in DMEM plus 10% FBS containing 3 μg/ml puromycin.

Statistical analysis.

Data from viral growth curve assays were analyzed using mixed-model, repeated-measures analysis of variance (ANOVA), with genotype (mutant/WT) and the presence of IFN as between-group factors and time (hours) as a within-group factor. Prior to applying repeated-measures ANOVA, titer values were transformed with a logarithmic function with base 10, due to the skewed distribution. The Bayesian information criteria (BIC) were used to assess the optimal covariance structure to account for within-subject correlation. Autoregressive order AR(1) was chosen to be the covariance structure. Tukey-Kramer multiple-comparison correction was used for controlling the type I error rate. Other data were analyzed by using GraphPad Instat (GraphPad Software, La Jolla, CA), and significant differences were determined by unpaired Student’s t test or two-tailed Fisher’s exact test. P values of <0.05 were considered significant.