Discovery of a Coregulatory Interaction between Kaposi's Sarcoma-Associated Herpesvirus ORF45 and the Viral Protein Kinase ORF36

ABSTRACT

Kaposi's sarcoma-associated herpesvirus (KSHV) is the causative agent of three human malignancies. KSHV ORF36 encodes a serine/threonine viral protein kinase, which is conserved throughout all herpesviruses. Although several studies have identified the viral and cellular substrates of conserved herpesvirus protein kinases (CHPKs), the precise functions of KSHV ORF36 during lytic replication remain elusive. Here, we report that ORF36 interacts with another lytic protein, ORF45, in a manner dependent on ORF36 kinase activity. We mapped the regions of ORF36 and ORF45 involved in the binding. Their association appears to be mediated by electrostatic interactions, since deletion of either the highly basic N terminus of ORF36 or an acidic patch of ORF45 abolished the binding. In addition, the dephosphorylation of ORF45 protein dramatically reduced its association with ORF36. Importantly, ORF45 enhances both the stability and kinase activity of ORF36. Consistent with previous studies of CHPK homologs, we detected ORF36 protein in extracellular virions. To investigate the roles of ORF36 in the context of KSHV lytic replication, we used bacterial artificial chromosome mutagenesis to engineer both ORF36-null and kinase-dead mutants. We found that ORF36-null/mutant virions are moderately defective in viral particle production and are further deficient in primary infection. In summary, our results uncover a functionally important interaction between ORF36 and ORF45 and indicate a significant role of ORF36 in the production of infectious progeny virions.

IMPORTANCE Kaposi's sarcoma-associated herpesvirus (KSHV) is a human tumor virus with a significant public health burden. KSHV ORF36 encodes a serine/threonine viral protein kinase, whose functions throughout the viral life cycle have not been elucidated. Here, we report that ORF36 interacts with another KSHV protein, ORF45. We mapped the regions of ORF36 and ORF45 involved in their association and further characterized the consequences of this interaction. We engineered ORF36 mutant viruses in order to investigate the functional roles of ORF36 in the context of KSHV lytic replication, and we confirmed that ORF36 is a component of KSHV virions. Moreover, we found that ORF36 mutants are defective in virion production and primary infection. In summary, we discovered and characterized a functionally important interaction between KSHV ORF36 and ORF45, and our results suggest a significant role of ORF36 in the production of infectious progeny virions, a process critical for KSHV pathogenesis.

INTRODUCTION

Kaposi's sarcoma-associated herpesvirus (KSHV) is a human tumor virus and the causative agent of Kaposi's sarcoma (KS), as well as two lymphoproliferative disorders (1–3). All herpesviruses encode at least one serine/threonine protein kinase that is conserved throughout the three subfamilies (alpha-, beta-, and gammaherpesviruses), collectively referred to as conserved herpesvirus protein kinases (CHPKs) (reviewed in references 4 and 5). Orthologs of CHPKs include UL13 of herpes simplex virus 1 (HSV-1), UL97 of human cytomegalovirus (HCMV), U69 of HHV-6, ORF47 of varicella-zoster virus, BGLF4 of Epstein-Barr virus (EBV), and ORF36 of KSHV, and murine herpesvirus 68 (MHV-68). Although there is considerable sequence divergence between CHPKs, certain features and functions, including autophosphorylation activity, tegument incorporation, nuclear localization, phosphorylation of cellular elongation factor 1δ (EF-1δ) (6–9), subversion of the interferon response (10, 11), and phosphorylation of ganciclovir, are conserved to various extents (reviewed in references 4 and 5). In addition, phosphorylation/disruption of the nuclear lamina and cyclin-dependent kinase activity has been detected for members of the beta- and gammaherpesvirus subfamilies (12, 13). KSHV ORF36 was originally identified as a serine protein kinase based on its sequence homology to known viral/cellular kinases (14). It was later found to activate the c-Jun N-terminal kinase (JNK) pathway (15). Since then, several viral and cellular proteins have been reported to be phosphorylated by ORF36. These include MKK4/7 (15), KSHV K8/K-bZIP (16), Kruppel-associated box domain-associated protein-1 (KAP-1/TRIM28) (17), retinoblastoma (Rb) (12), lamin A/C (12), histone H3 (18), and KSHV ORF59/PF-8 (19). However, compared to its homologs in HSV-1 (UL13), HCMV (UL97), and EBV (BGLF4), relatively little is known regarding the functional roles of KSHV ORF36 during viral lytic replication.

We have previously described the mechanism of sustained activation of the cellular p90 ribosomal S6 kinase (RSK) by the KSHV lytic protein ORF45 and revealed the importance of this activation during the lytic cycle (20–23). In a recent phosphoproteomic screen, we identified KSHV ORF36 as a potential substrate of KSHV ORF45-activated RSK (24). Here, we sought to confirm this finding and, in doing so, discovered the formation of a complex between ORF36, ORF45, and RSK that is dependent upon ORF36 kinase activity. We mapped the regions of both ORF36 and ORF45 that are critical for their association. We also found that ORF45 stabilizes ORF36 posttranslationally by protecting it from proteasome-dependent degradation. Importantly, coexpression of ORF45 and ORF36 in cells enhances the in vitro kinase activity of ORF36. To investigate the functional significance of KSHV ORF36 during lytic replication, we used bacterial artificial chromosome (BAC) mutagenesis to engineer ORF36-null or kinase-dead (KD) mutations in KSHV BAC16. Upon KSHV lytic reactivation, these mutant viruses are moderately defective in progeny virion production, suggesting that ORF36 plays an important role during the late lytic cycle. Consistent with studies of the ORF36 homologs HSV UL13, HCMV UL97, and EBV BGLF4, we detected KSHV ORF36 in extracellular viral particles. Furthermore, we found that virion-contained ORF36 is enzymatically active and is capable of phosphorylating KSHV virion proteins. Finally, we observed that ORF36 mutant viruses exhibit reduced infectivity, indicating that virion-contained ORF36 contributes to the optimal efficiency of primary infection. Taken together, these findings shed light on the functional significance of KSHV ORF36 during lytic replication.

MATERIALS AND METHODS

Antibodies, chemicals, and plasmids.

Anti-RxRxxS*/T* and anti-pJNK antibodies were ordered from Cell Signaling Technology. Anti-vPK (ORF36) rabbit polyclonal antibody, anti-RTA (ORF50) monoclonal mouse antibody, anti-ORF75 rabbit polyclonal antibody, and anti-PF8 (ORF59) rabbit polyclonal antibody were provided by Yoshi Izumiya, Ke Lan, Pinghui Feng, and Robert Ricciardi, respectively. Anti-ORF26, anti-ORF33, anti-ORF52, anti-ORF65, anti-ORF38, and 8B8 anti-ORF45 monoclonal antibodies were generated by the Florida State University hybridoma facility. All other antibodies and chemicals used in this study have been described previously (20, 21, 22, 25). Plasmids pCR3.1-ORF45 and derivatives have been described previously (21, 22, 26). Deletion mutants spanning the length of pCR3.1-ORF45 were also described previously (27). Other mutants of ORF45 or ORF36 were generated using the QuikChange mutagenesis kit (Stratagene, La Jolla, CA). Primer sequences used for cloning and mutagenesis are available upon request. All clones were verified by DNA sequencing.

Expression/preparation of glutathione S-transferase (GST) fusion proteins and GST pulldown assays.

Escherichia coli BL21 cultures transformed with plasmids encoding GST-ORF36, GST-ORF45 1-115, GST-K8, or GST-S6 fusion proteins were induced with 200 μM IPTG (isopropyl-β-d-thiogalactopyranoside) overnight at 18°C. Then, 500-ml cultures were pelleted by centrifugation at 6,500 × g for 10 min and resuspended in 50 ml of E. coli resuspension buffer (20 mM sodium phosphate buffer [pH 7.0], 500 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, 10% glycerol). Triton X-100 was added to a final concentration of 1%, and the cell suspension was sonicated. After 30 min of incubation at 4°C with gentle agitation, the cell debris was removed by centrifugation at 10,000 × g for 10 min. The supernatant was incubated with 1 ml of a 50% slurry of glutathione agarose beads for 3 h at 4°C. The beads were washed ten times with 20 ml of E. coli resuspension buffer. For in vitro kinase assays, proteins were eluted with 10 mM reduced glutathione, dialyzed overnight in phosphate-buffered saline (PBS), and concentrated using Amicon Ultra centrifugal filters. For GST pulldown assays, washed beads were incubated with precleared lysates of HEK293T cells for 3 h or overnight at 4°C with gentle agitation. The beads were washed five times with whole-cell lysis buffer, and bound proteins were eluted by addition of 2× loading dye and boiling for 5 min. Eluted proteins were separated by SDS-PAGE and analyzed by Western blotting with the indicated antibodies.

Cell culture and transfection.

HEK293 and HEK293T cells were cultured under 5% CO₂ at 37°C in Dulbecco modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and antibiotics. iSLK.BAC16 cells were cultured in DMEM containing 10% FBS, 250 μg/ml G418 sulfate, 400 μg/ml hygromycin B, and 1 μg/ml puromycin, as previously described (26, 28). These cells were induced by the addition of 2 μg/ml doxycycline and 1 mM sodium butyrate. Transient transfections were performed in six-well plates with Lipofectamine 2000 (Invitrogen, Carlsbad, CA) or 100-mm dishes by calcium phosphate methods.

Immunoprecipitation and Western blot analysis.

Immunoprecipitation and Western blot analysis were performed as previously described (21, 22, 26). For immunoprecipitation with anti-FLAG or anti-hemagglutinin (anti-HA) antibodies, the cell lysates were incubated with EZview red anti-Flag M2 or anti-HA affinity resin for 3 h at 4°C. After two washes with lysis buffer and three washes with Tris-buffered saline (TBS), proteins were eluted by incubation with 150 μg of 3×Flag peptide/ml for 1 h at 4°C, respectively. For immunoprecipitation of ORF45 from iSLK.BAC16 cells, we used a monoclonal anti-ORF45 antibody (8B8) conjugated to CNBr-activated Sepharose 4B (GE Life Sciences). Clarified lysates were bound to the beads for 2 h at 4°C and washed three times each with lysis buffer and TBS, and the bound complexes were eluted by boiling. For Western blotting, ∼20-μg portions of the proteins were resolved by SDS-PAGE and transferred to nitrocellulose membranes. The membranes were blocked in 5% dried milk in 1× PBS plus 0.2% Tween 20, followed by incubation with diluted primary antibodies for 2 h at room temperature or overnight at 4°C. Anti-rabbit, anti-rat, or anti-mouse IgG antibodies conjugated to horseradish peroxidase (Pierce) were used as the secondary antibodies. SuperSignal chemiluminescence reagents (Pierce) were used for detection.

In vitro kinase assays.

293T cells seeded into 100-mm dishes were cotransfected with 8 μg each of pCMV-ORF36 (wild type [WT] or KD mutant) and pCR3.1 (empty vector or ORF45 WT). At 24 h posttransfection (hpt), whole-cell lysates were made, and FLAG-ORF36 was immunoprecipitated with 50 μl of EZview red anti-FLAG affinity beads as described above. The kinase reaction was performed by incubation of 5 μl of immunoprecipitation (IP) complexes with 2 μg substrate in 25 μl of 1× kinase assay buffer (25 mM HEPES [pH 7.5], 50 mM NaCl, 20 mM β-glycerophosphate, 10 mM MnCl₂, 1 mM dithiothreitol, 1 mM Na₃VO₄, 100 μg of bovine serum albumin [BSA]/ml, 2 μM ATP, and 5 to 10 μCi of [γ-³²P]ATP). The reaction mixtures were kept at 30°C for 30 min and stopped by the addition of PAGE loading buffer. FLAG-Rb protein was purified from baculovirus-infected insect cells and was a gift from Yoshi Izumiya. For kinase assays of purified virions, ∼10⁸ extracellular virions were used per reaction. Virions were incubated at 37°C for 45 min in the presence of 1% Triton X-100 in 25 μl of 1× kinase assay buffer (without BSA), and reactions were stopped by the addition of PAGE loading buffer. After fractionation of samples by SDS-PAGE, the gels were dried and analyzed with a PhosphorImager.

Genome editing.

Potential guide RNAs (gRNAs) targeting the first exon of RSK1 (RPS6KA1; NM_002953; chr1: 26856249-26901520) or RSK2 (RPS6KA3; NM_004586; chrX: 20168029-20284750) were analyzed using the CRISPR Design tool (crispr.mit.edu) (29). Primers used for cloning of the target sequences have been previously described (24). Double-stranded oligonucleotides were generated and cloned into the lentiCRISPRv1 vector, which was then transfected into HEK293T cells (30, 31). At 2, 3, and 4 days after transfection, medium samples from the cells were collected, clarified by centrifugation, and filtered through a 0.45-μm-pore size filter to collect lentiviral particles. These were used to transduce HEK293T in a 24-well plate at a multiplicity of infection (MOI) of 0.5 (lentiCRISPRs carrying gRNAs targeting RSK1 or RSK2 were used as a 50:50 mix). The cells were supplemented with media containing this lentiviral mix and 4 μg of Polybrene/ml, immediately centrifuged at 800 × g for 1 h at 37°C, and cultured under normal conditions for 24 h. The day after transduction, the cells were trypsinized and grown in the presence of 1 μg of puromycin/ml. Genomic DNA surrounding the Cas9 cleavage sites were extracted using QuickExtract solution and PCR amplified using the Herculase II fusion DNA polymerase (Agilent). Cleavage was confirmed using a Surveyor mutation detection kit (Transgenomic), as well as Sanger sequencing. For the isolation of HEK293T ΔRSK1/2 clones, cells were diluted to a concentration of ∼0.5 cells/100 μl, seeded to 96-well plates, and grown for 2 to 3 weeks in the presence of 1 μg of puromycin/ml. Thirty single clones were isolated and assayed by Western blot analysis for RSK1/RSK2 protein expression. Of these, two that showed a complete loss of signal (i.e., clones 22 and 27) were chosen for further analyses.

Genetic manipulation of KSHV BAC16 genome.

Mutagenesis of BAC16 (32) was performed using a recombineering system as Tischer et al. described (33, 34). In brief, the Kan/I-SceI cassettes were amplified from plasmid pEPKan-S by PCR with various primer pairs: KS36-K108Q-5′ (5′-CTTTGGCATAATCGTCCCTATCTCCGAGGATCTGTGTGTGCAGCAGTTTGATAGCCGCCGAGGATGACGACGATAAGTAGG-3′) and KS36-K108Q-3′ (5′-ATTGCCTCGTAGAAAAACTCCCGGCGGCTATCAAACTGCTGCACACACAGATCCTCGGAGAGCCAGTGTTACAACCAATTAACC-3′) were used for the K108Q (kinase-dead; KD) mutant; KS36-2xStop-5′ (5′-AATGGAGAGGAGACCCCCACTCACTCCTCTTCGGAGATAAAGGACACAATCGTGAGGTGGAGGCTTGAAGGATGACGACGATAAGTAGGG-3′) and KS36-2xStop-3′ (5′-CAGGGCACACCGGGGGCAAATCGTCAAGCCTCCACCTCACGATTGTGTCCTTTATCTCCGAAGAGGACCAGTGTTACAACCAATTAACC-3′) were used for the 2xStop36 mutant; and KS36-dStart-5′ (5′-CCGTCGCGGACCTCAAAGAAGAGGTGGCCGTGCGCCTAGACGCGCTGGAAGAGAACGGAGAGGAGACCAGGATGACGACGATAAGTAGGG-3′) and KS36-dStart-3′ (5′-TGATCTCCGAAGAGGAGTGAGTGGGGGTCTCCTCTCCGTTCTCTTCCAGCGCGTCTAGGCGCACGGCGCCAGTGTTACAACCAATTAACC-3′) were used for the ΔStart36 mutant. The purified PCR fragments were electroporated into BAC16-containing GS1783 cells (34) that had been induced at 42°C for 15 min. The recombinant clones were selected at 32°C on Luria-Bertani (LB) plates containing 34 μg of chloramphenicol/ml and 50 μg of kanamycin/ml and then analyzed by restriction enzyme digestion. Positive clones were cultured with 1% l-arabinose, induced at 42°C again, and plated on LB plates containing 1% l-arabinose for secondary recombination. Colonies which survived on the l-arabinose plates were replicated on plates with 34 μg/ml chloramphenicol alone and on plates with both 34 μg/ml chloramphenicol and 50 μg/ml kanamycin. The kanamycin-sensitive clones were analyzed by restriction enzyme digestion, and proper mutations were further confirmed by DNA sequencing.

Reconstitution of recombinant KSHVs.

Briefly, iSLK cells seeded in a 24-well plate were transfected with 1 μg of BAC DNAs by Effectene (Qiagen). One day after transfection, cells were subcultured into a T150 flask with fresh medium containing 450 μg/ml G418 and 1 μg/ml puromycin. The next day, hygromycin was added to a final concentration of 500 μg/ml for selection. After 12 days of selection, hygromycin-resistant colonies were trypsinized and subcultured with 1:9 dilution every 3 days. Two clones, designated KD1 and KD2, were chosen for the K108Q mutant. To induce viral lytic replication, BAC-containing iSLK cells were seeded into six-well plates or a T150 flask and, 1 day later (when the cells reached ∼90% confluence), the medium was replaced with fresh medium containing 2 μg/ml doxycycline and 1 mM sodium butyrate.

Real-time quantitative PCR analysis of virion DNA.

The virion DNA was prepared as previously described (20, 25). The medium from induced BAC-iSLK cells was collected, centrifuged, and passed through a 0.45-μm-pore size filter to clear cell debris. Treatment of 100 μl of the cleared supernatant with 10 U of Turbo DNase (Ambion, Austin, TX) at 37°C for 1 h degraded the extravirion DNA. The reaction was stopped by the addition of EDTA to a final concentration of 5 mM, followed by heat inactivation at 70°C. Next, 20 μl of proteinase K solution and 200 μl of buffer AL from a DNeasy kit (Qiagen, Valencia, CA) were added. The mixture was kept at 70°C for 15 min and then extracted with phenol-chloroform. The DNA was precipitated by the addition of 2 volumes of ethanol with glycogen as a carrier, and the DNA pellet was dissolved in 40 μl of Tris-EDTA buffer. Next, 2 μl of DNA was used in SYBR green real-time quantitative PCR analyses using a Bio-Rad CFX96 real-time detection system and a C1000 thermal cycler. Thermal amplification was performed according to the following parameters: 10 min at 96°C, followed by incubation 45 amplification cycles, each with denaturation (95°C for 10 s), annealing (60°C for 20 s), and extension (72°C for 15 s) periods. The cycle threshold (C_T) value was determined as the point (cycle) at which the amplification plot crossed the threshold line. The threshold line was automatically set at ten times the standard deviation of the baseline by the program. Viral DNA copy numbers were calculated with external standards of known concentrations of serially diluted BAC16 DNA.

Virus stock preparation and infection.

Six T150 flasks of cells were induced for 5 days, and then the medium was collected and centrifuged to remove cell debris. Virions were pelleted at 100,000 × g for 1 h on a 25% sucrose cushion with a Beckman SW28 rotor. The virus pellets were dissolved in 1/100 of the original volume of TNE buffer (10 mM Tris-HCl [pH 8.0], 150 mM NaCl, 1 mM EDTA) and stored at −80°C. The viral genome copy number was quantified by quantitative reverse transcription-PCR. Infection was carried out as previously described (20, 25). Briefly, HEK293 cells plated in 24-well plates were incubated with 2-fold serial dilutions of concentrated virus plus Polybrene (4 μg/ml) and spun at 800 × g for 1 h at room temperature. The plates were incubated at 37°C for 2 h, and the inocula were then removed and replaced with fresh medium with 5% FBS. At 24 h postinduction, the cells were washed twice and resuspended in PBS, and then the green fluorescent protein (GFP) expression was measured by using a BD FACSCanto analyzer.

RESULTS

KSHV ORF36 interacts with ORF45 and RSK in cells.

We recently identified ORF36 as one of the few viral substrates of ORF45-activated RSK using a mass spectrometry-based phosphoproteomic approach (24). In order to confirm ORF45/RSK-mediated phosphorylation of ORF36, we coexpressed ORF45 and FLAG-ORF36 in cells and immunoprecipitated FLAG-ORF36. As expected, we observed that overexpression of ORF45 increased the phosphorylation of ORF36 at the consensus RSK phosphorylation motif: RxRxxS*/T* (Fig. 1; right panel, compare lanes 1 and 2, FLAG-ORF36 migrates at ∼60 kDa). We also detected a diffuse signal at ∼75 to 90 kDa upon coexpression of ORF36 WT and ORF45 (lane 2). The size range of this signal matches perfectly to the expected mobility of ORF45, whose phosphorylation at one of two putative RSK phosphorylation sites is detectable by the RxRxxS*/T* motif antibody. Interestingly, ORF45 coprecipitated with ORF36, suggesting that they form a stable complex in cells (Fig. 1, right panel). RSK also coprecipitated with ORF36, but only in the presence of ORF45 (Fig. 1, right panel, lane 2). In addition, we observed that ORF36 can induce phosphorylation of RSK at a site indicative of its activation (Ser380) independently of ORF45 (Fig. 1, left panel, compare lanes 1 and 3). RSK phosphorylation is not induced by a KD mutant of ORF36 in which the catalytic lysine has been mutated to a glutamine (K108Q) (15). KD ORF36 also did not stably interact with ORF45 and was not phosphorylated at its RxRxxS*/T* motif, indicating that ORF36 kinase activity is required for these phenotypes (Fig. 1). Overexpression of ORF36 WT but not the KD mutant induced a clear mobility shift in ORF45 signal indicative of phosphorylation (Fig. 1, left panel, compare lanes 2 and 4). This could represent direct phosphorylation of ORF45 by ORF36 and/or increased phosphorylation by RSK or other cellular kinases.

FIG 1.

KSHV ORF36 and ORF45 stably interact in cells. 293T cells were transfected with the indicated plasmids. At 24 hpt, cell lysates were collected, and FLAG-ORF36 was immunoprecipitated with anti-FLAG resin. Input whole-cell lysates (WCL) and eluates (IP:FLAG) were subjected to SDS-PAGE and then analyzed by Western blotting with the indicated antibodies. The data are representative of three biological replicates.

Mapping the regions of ORF45 and ORF36 required for their interaction.

To further characterize ORF36-ORF45 interaction, we next sought to identify the region(s) of ORF36 involved in the binding. To do so, we expressed and purified GST-ORF36 truncation mutants from E. coli (Fig. 2A). We then performed GST pulldown assays after incubation of these mutants with the lysates of ORF45-expressing cells. We found that the N-terminal 83 amino acids of ORF36 were critical for binding to ORF45 (Fig. 2B), whereas the kinase domain (positions 83 to 244) and the C-terminal domain (positions 244 to 444) appeared to be dispensable for the interaction (Fig. 2B, lanes 4 and 5). We next mapped the region(s) of ORF45 required for the interaction. Using a set of internal deletion mutants spanning the full length of ORF45, we found that deletion of only one region, amino acids (aa) 90 to 115, dramatically reduced the binding of ORF45 to the ORF36 N-terminal fragment (Fig. 2C). The absence of signal in lanes 9 and 11 is due to deletion of the epitope for anti-ORF45 antibodies 2D4 and 4C3, respectively. We confirmed our GST pulldown results by performing a FLAG-IP following coexpression of ORF36 mutants with WT ORF45 (Fig. 3A). ORF36 mutants lacking the N-terminal 83 aa exhibited decreased binding to ORF45, whereas ORF36 1-244 (aa 1 to 244) bound almost as well as the WT, further suggesting that the C-terminal domain of ORF36 (positions 244 to 444) is not critical for this interaction. Interestingly, only the full-length ORF36 induced a mobility shift in ORF45 signal (Fig. 3A, compare lanes 5 and 8). We next sought to validate the importance of ORF45 90-115 (aa 90 to 115), as well as address the contribution(s) of ERK/RSK activation for ORF45-ORF36 interaction. To do so, we performed a FLAG-IP after coexpression of FLAG-ORF36 WT with one of four ORF45 mutants: (i) Δ19-77, in which the binding to and activation of both ERK and RSK is abrogated (22); (ii) F66A, a point mutant deficient in RSK activation (23); (iii) F32/34A, a double point mutant deficient in ERK activation (unpublished data); and (iv) Δ90-115. The pattern of ERK/RSK phosphorylation upon overexpression of these various ORF45 mutants is as would be expected (Fig. 3B). Strikingly, the data indicate that deletion of the region from aa 90 to 115 or ablation of ORF45-mediated RSK activation abolishes the interaction between ORF45 and ORF36 in cells. To further validate the importance of RSK, we generated RSK1/2-knockout cells using the CRISPR/Cas9 system. Consistent with our previous findings, knocking out RSK significantly decreased the association of ORF45 with FLAG-ORF36 (Fig. 3C). Phosphorylation of ORF36 at its RxRxxS*/T* motif is also decreased in the RSK knockout cells (Fig. 3C, right panel, ORF36 signal denoted by arrow).

FIG 2.

Mapping the critical regions for ORF36-ORF45 interaction. (A) The N-terminal 83 amino acids of ORF36 are necessary and sufficient for binding to ORF45 in vitro. GST and the indicated GST-ORF36 truncation mutants were expressed, purified, and then incubated with the lysates of 293T cells expressing WT ORF45. After GST pulldown assays, the bound proteins were analyzed by Western blotting with the indicated antibodies. (B) The deletion of aa 90 to 115 of ORF45 abolishes its binding to GST-ORF36 1-83. ORF45 internal deletion mutants were expressed in 293T cells, the lysates of which were incubated with GST-ORF36 1-83. GST pulldown assays, and Western blot analyses were performed as described in panel A. The data are representative of two biological replicates.

FIG 3.

Characterization of the requirements for interaction between ORF36 and ORF45. (A) 293T cells were transfected with plasmids encoding FLAG-ORF36 WT or one of the indicated truncation mutants in the presence or absence of ORF45. FLAG-IP and Western blot analyses were performed as in Fig. 1. (B) 293T cells were transfected with FLAG-ORF36 in the presence or absence of ORF45 WT or the indicated deletion/point mutants. FLAG-IP was performed as described previously. (C) FLAG-ORF36 and ORF45 were coexpressed in 293T WT or RSK1/2 knockout cells (clones 22 and 27 represent single cell clones). FLAG-IP was performed as described previously. (D) Purified ORF45 protein was mock treated or treated with lambda phosphatase (λ_pp) in the presence or absence of phosphatase inhibitors (inh.) and then subjected to GST pulldown assays as in Fig. 2B. The data are representative of two biological replicates.

Phosphorylation of ORF45 is critical for optimal binding to ORF36.

The region from aa 90 to 115 of ORF45 is highly acidic; 19 of its 26 aa are either aspartate or glutamate. This acidic patch contributes to the low predicted pI of ORF45: ∼4.3. On the other hand, ORF36 is basic, with a predicted pI of ∼9.5, and the pI of the region from aa 1 to 83 alone is ∼11.6. Furthermore, ORF45 has a remarkably high proportion of serines and threonines (∼24% of its amino acids), many of which are phosphorylated in cells. Thus, we hypothesized that ORF36-ORF45 interaction may be enhanced by ionic bond formation between the negatively charged residues of ORF45 and the positively charged residues of ORF36. To test this, we first expressed and purified ORF45 from insect cells. We treated the protein with λ phosphatase (λ_pp) in the presence and absence of phosphatase inhibitors, then assessed its binding to the GST-ORF36 1-83 fragment. As expected, we observed a clear reduction in the binding of phosphatase-treated ORF45 to ORF36, which could be almost completely recovered by including phosphatase inhibitors in the reaction (Fig. 3D). This finding strongly supports the notion that the phosphorylation status of ORF45 is of critical importance for its interaction with ORF36. This is consistent with the previously observed decreased association between ORF45 and ORF36 upon (i) KD mutation of ORF36 (Fig. 1), (ii) deletion of the acidic region of ORF45 90-115 (Fig. 2B and 3B), (iii) deletion/mutation of residues involved in ORF45-mediated RSK activation (Fig. 3B), and (iv) RSK1/2 knockout (Fig. 3C).

ORF45 protects ORF36 from proteasome-dependent degradation.

We previously noticed that ORF36 protein level was noticeably reduced in iSLK.BAC16 Stop45 (ORF45-null) cells (24). We have also consistently observed that coexpression with ORF45 induces a slight increase in ORF36 protein level (i.e., Fig. 3B). While this phenomenon could be explained by ORF45's roles in transcriptional and translational regulation (23, 24, 26), the effect was reminiscent of the ORF45-dependent stabilization of another KSHV tegument protein, ORF33 (27). To assess whether ORF45 could enhance the stability of ORF36 protein posttranslationally, we cotransfected cells with ORF36 and ORF45, treated them with cycloheximide, and measured the levels of both proteins over a 12-hour time course. Indeed, we found that ORF45 increased the half-life of ORF36 from ∼9 to >24 h (Fig. 4, left panel). Interestingly, it had no apparent effect on KD ORF36, which also is inherently less stable than the wild type. These results were suggestive of an interaction-dependent stabilization of ORF36 by ORF45. In further support of this notion, ORF36 truncation mutants containing the N-terminal 83 amino acids that are critical for ORF36-ORF45 interaction (Fig. 3A) were also stabilized by ORF45 (Fig. 4, middle panel). Conversely, ORF36 83-244 and 83-444, which do not appreciably bind to ORF45 (Fig. 3A), are not stabilized by ORF45 coexpression under these conditions (Fig. 4, right panel). The decrease in ORF36 protein level can be attributed to proteasome-dependent degradation, since the level remains unaffected when cells are treated with cycloheximide in the presence of the proteasome inhibitor MG132 (Fig. 4, lower left panel).

FIG 4.

ORF45 enhances the stability of ORF36. 293T cells were transfected with FLAG-ORF36 WT or the indicated mutant in the presence and absence of ORF45. At 24 hpt, cells were harvested or treated with cycloheximide (CHX) for the indicated times in hours (h). Lysates were collected, subjected to SDS-PAGE, and analyzed by Western blotting with the indicated antibodies. The relative intensities for FLAG signal (normalized to actin) were quantified using ImageJ and are shown below α-FLAG immunoblots. The data are representative of at least two biological replicates.

ORF45 enhances ORF36 kinase activity in a RSK-independent manner.

After our characterization of ORF45-ORF36 interaction, the next logical question was “what are the functional consequences?” We have previously shown that ORF45 enhances RSK kinase activity, as evidenced by increased in vitro phosphorylation of the RSK substrate S6 (21). We thus performed in vitro kinase assays to examine the potential effect(s) of ORF45 on ORF36 kinase activity. Importantly, phosphorylation of FLAG-ORF36 itself was detected in lane 1 (ORF36 WT) and increased in lane 2 (ORF36 WT + ORF45) but was absent in lanes 3 (ORF36 KD) and 4 (ORF36 KD + ORF45) (Fig. 5, all panels, ∼60 kDa). In addition, we found that N-terminal fragments of both ORF36 and ORF45 are phosphorylated in a manner dependent on ORF36 kinase activity (Fig. 5A, bottom two panels). Moreover, coexpression with ORF45 slightly enhanced their ORF36-dependent phosphorylation (Fig. 5A, compare lanes 1 and 2). Strikingly, ORF36-mediated phosphorylation of its previously identified substrates, cellular Rb (12) and viral K8/K-bzip (16), is dramatically increased by ORF45 coexpression (Fig. 5B, compare lanes 1 and 3). Because of the apparent importance of RSK for ORF45-ORF36 interaction (Fig. 3C), we also assessed whether RSK activation affected the kinase activity of ORF36. Surprisingly, ORF45 F66A mutation had no effect on the ORF36-dependent phosphorylation of its substrates (Fig. 5B, compare lanes 3 and 5), indicating that the binding to and activation of RSK is dispensable for the ORF45-dependent enhancement of ORF36 kinase activity.

FIG 5.

ORF45 enhances ORF36 kinase activity in a RSK-independent manner. (A and B) 293T cells were transfected with the indicated plasmids, and FLAG-ORF36 was immunoprecipitated and then used for in vitro kinase assays with the indicated substrate as described in Materials and Methods. Red asterisks denote the expected size of the indicated substrate. The bands at ∼72 kDa are BSA. The data are representative of three biological replicates.

ORF36-null/KD mutants were generated in KSHV BAC16.

Previous studies have identified individual substrates of KSHV ORF36 and its homologs, EBV BGLF4 and HCMV UL97. However, the roles of KSHV ORF36 in lytic reactivation have not been comprehensively investigated. We made use of BAC mutagenesis (33, 34) to engineer the kinase-inactivating point mutation (K108Q or KD), or one of two ORF36-null mutations, in the KSHV genome. ORF36 deletion mutants were made by either introducing two stop codons just downstream of the start codon (2xStop36) or mutating the start codon (ΔStart36). Neither of these mutations affected the amino acid sequence of the overlapping ORF35. We then reconstituted these BACs in iSLK cells as previously described (23, 32) to produce the corresponding iSLK.BAC16 mutant cell lines.

Characterization of BAC16 ORF36 mutants.

We confirmed the loss of ORF36 expression in the ORF36-null mutants by Western blot analysis (Fig. 6A). The level of ORF36 protein detected after lytic reactivation of the KD mutants (KD1 and KD2 are two single clones) was lower than that of the WT, presumably due to lower stability of KD ORF36 (Fig. 4). Although the expression of immediate-early (RTA, ORF45, and PF-8) and early (K8) lytic genes was not significantly affected by ORF36 mutation or deletion, the levels of several late genes were noticeably decreased (Fig. 6A). Among those affected are several tegument proteins, such as ORF33, ORF38, and ORF75, as well as the small capsid protein, ORF65. Because we found that ORF36 was capable of inducing phosphorylation of RSK under cotransfection conditions (Fig. 1), we wondered whether this effect was relevant during lytic replication. We observed a mild decrease in RSK phosphorylation at Ser380 in the KD mutants, although there was no apparent decrease in the ORF36-null mutants (Fig. 6B). This is indicative of a dominant negative effect of the KD mutation. We also detected a substantial decrease in pJNK upon ORF36 deletion or KD mutation, a finding consistent with a previous study showing that JNK is a substrate of ORF36 (15). Because of the apparent downregulation of RSK phosphorylation by ORF36 mutation, we assessed the phosphorylation status of two RSK substrates, eukaryotic translation initiation factor 4B (eIF4B) and S6 ribosomal protein. We observed a reduction of eIF4B phosphorylation that correlated well with decreased pRSK (Fig. 6B). On the other hand, we saw no reduction in signal of pS6, indicating that the activation of other AGC kinases (including S6K and/or Akt, which can both phosphorylate S6) is likely sufficient to compensate for the lower RSK activity. These data are in line with our previous results suggesting that, in induced iSLK.BAC16 cells, RSK is the predominant kinase of eIF4B, while S6K is the predominant kinase of S6 (24).

FIG 6.

Profiling of ORF36 mutant virus-infected cells during a time course of lytic reactivation. iSLK.BAC16 cells stably expressing BAC16 WT or the indicated ORF36 mutants were uninduced (U) or induced to undergo lytic reactivation. Cell lysates were collected at the indicated day postinduction, subjected to SDS-PAGE, and analyzed by Western blotting with the indicated antibodies. (A) KSHV genes; (B) cellular genes. The data are representative of at least two biological replicates.

Importantly, we confirmed the interaction between ORF45 and ORF36 in the context of lytic reactivation and, in concordance with our previous results, observed decreased interaction upon ORF45 F66A mutation or ORF36 KD mutation (Fig. 7A). We also performed stability assays of KSHV-infected cells and found that endogenous ORF36 is less stable in the absence of ORF45 (Stop45) or upon KD mutation (Fig. 7B). Western blot analysis of extracellular virions confirmed that ORF36 is a virion protein, and also indicated that deletion/mutation of ORF36 has no apparent effect on the packaging of other virion components (Fig. 7C). We next examined the potential importance of ORF36 for the production of progeny viral particles. As shown in Fig. 7D, ORF36 deletion resulted in an ∼3-fold reduction in virion production, whereas KD mutation was accompanied by an approximately 5- to 10-fold decrease. To identify the putative virion protein substrates of ORF36, we performed in vitro kinase assays of purified extracellular viral particles. Strikingly, a strong signal could be observed at ca. 70 to 85 kDa in a manner dependent on Triton X-100-induced lysis of virions and the addition of MnCl₂ (Fig. 7E, compare lanes 1 to 4). Furthermore, the intensity of phosphorylation was dramatically reduced upon ORF36 deletion or KD mutation (Fig. 7E). Based on the molecular weight and pattern of phosphorylation, we can predict that the identity of this major band is most likely ORF45. We can also detect signal at ∼55 kDa (only for WT virions), which presumably represents ORF36 autophosphorylation activity (Fig. 7E). The finding that virion-contained ORF36 is enzymatically active led us to speculate that ORF36 may play an important role during primary infection. To investigate this, we infected 293T cells with 2-fold dilutions of KSHV WT or ORF36-null/KD mutants (normalized by viral genome copy number) and then measured the percentage of GFP-positive cells by fluorescence-activated cell sorting (FACS). Interestingly, all mutants were defective in infectivity, with the KD mutants displaying an ∼5-fold reduction in GFP-positive cells compared to the WT (Fig. 8). Altogether, these results suggest that the kinase activity of KSHV ORF36 is important for optimal production of infectious progeny viruses.

FIG 7.

Analysis of the functional consequences of ORF36 deletion or KD mutation. (A to E) iSLK.BAC16 cells were induced as in Fig. 6. (A) At 48 hpi, cell lysates were collected and immunoprecipitated with anti-ORF45 antibody-conjugated beads. The input and eluates were subjected to SDS-PAGE and analyzed by Western blotting with the indicated antibodies. (B) At 48 hpi, cells were untreated (0 h) or treated with cycloheximide (CHX). At the indicated times after treatment, lysates were collected and analyzed as in Fig. 4. (C) Virions collected from culture media at 5 days postinfection were concentrated by ultracentrifugation. After normalization by viral genome copy number, the samples were subjected to SDS-PAGE and analyzed by Western blotting with the indicated antibodies. (D) At the indicated hour postinduction, the culture medium was collected. Virion DNA was extracted and quantified by qPCR. The results shown are the averages of three biological replicates. (E) The same virions analyzed in panel C were used for in vitro kinase assays, as described in Materials and Methods. The data are representative of two biological replicates.

FIG 8.

ORF36 is critical for efficient KSHV primary infection. 293T cells were infected with 2-fold serial dilutions of the indicated virus (WT or one of four ORF36 mutants; normalized by viral genome copy number) and then assessed at 24 hpi by FACS for GFP-positive percentage. The chart illustrates the percentage value for each dilution tested (the average of two biological replicates). Representative histograms and fluorescence micrographs are shown for dilutions at an MOI of 25.

DISCUSSION

All herpesviruses encode at least one serine/threonine protein kinase, collectively referred to as CHPKs. The CHPK ortholog of the human tumor virus KSHV is encoded by ORF36. Here, we shed light on the important roles of this viral kinase during lytic replication and primary infection. We first became interested in ORF36 because it is one of the few viral substrates of ORF45-activated RSK (24), and subsequently made the serendipitous finding that it interacts with the multifunctional tegument protein, ORF45 (Fig. 1). In an effort to better understand this viral-viral protein interaction, we mapped the critical regions for their binding (Fig. 2 and 3). Interestingly, ORF45 posttranslationally stabilizes ORF36 in a manner reminiscent of its stabilization of another KSHV tegument protein, ORF33 (27) (Fig. 4). Importantly, ORF45 enhances ORF36 kinase activity in vitro (Fig. 5). We developed novel KSHV ORF36-null/KD mutants using BAC mutagenesis as a means to further investigate the functional roles of ORF36 during the viral life cycle. We confirmed that ORF36 and ORF45 interact in KSHV-infected cells (Fig. 7A). We also found that ORF36 is present in extracellular virions (Fig. 7C) and detected a significant reduction in viral particle production upon ORF36 mutation (Fig. 7D). By performing in vitro kinase assays of purified virions, we show evidence which suggests that ORF45 is phosphorylated in a manner dependent on ORF36 kinase activity (Fig. 7E). Finally, we found that ORF36 mutant virions exhibit reduced infection rate, indicating a requirement of ORF36 kinase activity for optimal primary infection (Fig. 8).

ORF36 has been shown to localize to the nucleus (35), while ORF45 exhibits both nuclear and cytoplasmic localization (25, 35). We recently found that ORF45 induces an interaction-dependent relocalization of ORF33 (which is normally exclusively nuclear), to the cytoplasm of cotransfected cells (27). Thus, it would be interesting to investigate the extent to which ORF45-ORF36 interaction may alter their subcellular localizations both in transfected and virus-infected cells. We would expect that at least a portion of ORF36 in cells is associated with ORF45 in the cytoplasm, thereby effectively broadening the range of putative ORF36 substrates. Conversely, through phosphorylating ORF45, ORF36 may affect ORF45 localization. Indeed, a phosphorylation-dependent regulation of ORF45 is probable, since cytoplasmic and nuclear fractions of ORF45 in virus-infected cells are distinguishable by a distinct mobility shift indicative of phosphorylation (25). If there are in fact coregulatory effects on ORF45/ORF36 localization that are mediated by their interaction, these would invariably have important functional consequences during KSHV lytic replication and primary infection.

Perhaps the most physiologically relevant findings of this study are the apparent effects of ORF36 deletion or KD mutation on (i) production of progeny virions and (ii) primary infection efficiency. Below, we propose several non-mutually exclusive models to explain these phenomena. (i) Regarding the defect in virion production, it is possible that ORF36 deletion/mutation reduces the efficiency of KSHV late gene expression. Although the level of capsid protein ORF26 was not affected by ORF36 mutation, the levels of the tegument proteins ORF33, ORF38, ORF52, and ORF75, as well as the small capsid protein ORF65, were noticeably reduced (Fig. 6A). It is plausible that the optimal expression of ORF33, as well as other KSHV genes with roles in viral DNA replication or virion assembly, may depend on ORF36 kinase activity. Indeed, ORF36-dependent phosphorylation of viral K-bzip has been shown to modulate viral transcription (16). Furthermore, EBV BGLF4 has been shown to play a critical role in the efficient expression of several EBV late lytic genes (36). (ii) ORF36 is critical for efficient viral DNA synthesis. This notion is consistent with recent work showing that ORF36 can phosphorylate the KSHV processivity factor (PF-8, encoded by ORF59) and that ablation of these phosphorylation sites on PF-8 impairs viral particle production (19). (iii) ORF36 is required for packaging of newly assembled viral particles. However, this is unlikely, since virion protein composition was apparently unaffected by ORF36 mutation (Fig. 7C). (iv) ORF36 is involved in the nuclear egress of nucleocapsids. This function has been suggested for other CHPKs, including HSV-2 UL13 (37), HCMV UL97 (12, 38–40), and EBV BGLF4 (41), due to their conserved ability to phosphorylate and consequently induce disassembly of the nuclear lamina. Although KSHV ORF36 has been shown to induce only mild disassembly of lamins after transient transfection (41), we suspect that its role during lytic replication is more significant. Furthermore, a related yet distinct mechanism by which ORF36 may play a role in nuclear egress is by phosphorylating components of the viral nuclear egress complex (NEC). This has recently been suggested as an additional mechanism by which HCMV UL97 modulates nuclear egress (38, 42). The KSHV NEC, comprised of ORF67 and ORF69, has only recently begun to be studied (43, 44). ORF36 may interact with and/or phosphorylate these proteins to facilitate efficient egress of new viral particles. (v) Finally, it is conceivable that ORF36 plays a role in the cytoplasmic secondary envelopment of virions, as has been suggested for HCMV UL97 (45, 46).

As for the defect of ORF36 mutant viruses in primary infection rate, (i) ORF36 may be required for the efficient disassembly of incoming virions immediately after entry. The dissociation of tegument proteins from capsids following infection has important consequences on the roles of these proteins, as well as the intracellular trafficking of incoming nucleocapsids to the nucleus. Indeed, a phosphorylation-dependent dissociation of certain virion components has been reported to be mediated by HSV-1 UL13 (47) and EBV BGLF4 (48). Because virion-contained ORF36 possesses enzymatic activity (Fig. 7E), it stands to reason that certain tegument proteins (including ORF45) are phosphorylated in an ORF36-dependent manner during primary infection. This could affect ORF45's functions in modulation of cellular kinase signaling (21–24, 26) and/or evasion of the innate immune response (49–52). (ii) ORF36 itself may play a role in the subversion of the immune response to KSHV infection. MHV-68 ORF36 and its homologs in HSV-1, HCMV, EBV, and KSHV have been shown to suppress the innate immune response by inhibiting the production of type I interferon (10). Whether this conserved inhibition occurs during primary infection remains to be determined. (iii) ORF36 might regulate the nuclear translocation of virions. Recently, it was reported that BGLF4 induces the nuclear redistribution of several EBV lytic proteins, including the major capsid protein, VCA (ORF25) (53). Interestingly, the same study found that KSHV ORF36 was also capable of inducing the nuclear translocation of VCA. Future studies are required to understand the mechanisms by which conserved herpesvirus protein kinases regulate virion disassembly, innate immunity, and nuclear import during primary infection.

There are several additional questions to which we still do not have clear answers, such as what are viral and cellular substrates of ORF36? How is ORF36 substrate specificity affected by ORF45? What is or are the phosphorylation motifs of ORF36 substrates? Is the ORF36-ORF45 interaction conserved throughout gammaherpesviruses or unique to KSHV? Many of these questions and others could be addressed by performing an unbiased MS-based phosphoproteomic analysis of ORF36-expressing and/or KSHV-infected cells. Previously, researchers from the Hayward lab identified the viral (13) and cellular (54) substrates of EBV BGLF4. In the latter of these studies, cellular substrates of HSV-1 UL13, HCMV UL97, and KSHV ORF36 were also identified. However, these studies relied on in vitro assays, and consequently, the substrates identified might not be representative of the true substrates during viral replication. Recently, the same group published data obtained from phosphoproteomic profiling of BGLF4-expressing cells (55). Such experiments would be invaluable in delineating the substrates of KSHV ORF36, as well as other herpesviral kinases. Despite the questions that remain, the data presented herein mark a significant advancement of our understanding of the viral protein kinase, ORF36. Future studies aimed at further investigating ORF36-ORF45 interaction and localization, characterizing the roles of ORF36 during lytic replication and primary infection, and identifying its functionally relevant substrates will yield new insights into the important functions of ORF36 throughout the life cycle of KSHV.