ORF45 of Kaposi's sarcoma–associated herpesvirus (KSHV) is a gamma herpesvirus-specific, immediate-early, and tegument protein. Our previous studies have revealed its crucial roles in both early and late stages of KSHV infection. In this study, we surveyed the interactome of ORF45 using a panel of monoclonal antibodies. In addition to the previously identified extracellular regulated kinase (ERK) and p90 ribosomal S6 kinase (RSK) proteins, we found several other co-purified proteins, including prominent ones of ~38 kDa and ~130 kDa. Mass spectrometry revealed that the 38 kDa protein is viral ORF33 and the 130 kDa protein is cellular USP7 (ubiquitin-specific protease 7). We mapped the ORF33-binding domain to the highly conserved carboxyl terminal 19-aa of ORF45, and the USP7-binding domain to the reported consensus motif in the central region of ORF45. Using immunofluorescence staining, we observed colocalization of ORF45 with ORF33 or USP7, in both transfected conditions and KSHV-infected cells. Moreover, we noticed an ORF45-dependent relocalization of a portion of ORF33/USP7 from the nucleus to the cytoplasm. We found that ORF45 caused an increase in ORF33 protein accumulation, which was abolished if either the ORF33- or USP7-binding domain in ORF45 was deleted. Furthermore, deletion of the conserved carboxyl terminus of ORF45 in the KSHV genome drastically reduced the level of ORF33 protein in KSHV-infected cells and abolished production of progeny virions. To determine if the binding of ORF33 is a critical function of C19, we used co-precipitation with point mutants of the C19 region and identified two required residues: tryptophan 403 and tryptophan 405. We then engineered KSHV genomes containing these mutants and transfected them into iSLK cells. Similar to the C19 deletion mutant, we found that both binding-deficient mutants exhibited decreased ORF33 accumulation and viral particle production. Since C19 is sufficient for binding ORF33, we hypothesized that introduction of a C19 analogue could inhibit binding and may lead to a similar decrease in viral particle production. We used ELISA to measure the binding of ORF33 to ORF45 in the presence of TAT-C19 and found that TAT-C19 inhibited binding in a dose-dependent manner. To measure the analogue's effect on viral particle production, we treated KSHV-infected cells with TAT-C19 and found that TAT-C19 inhibited viral particle production in a dose-dependent manner. In conclusion, binding of ORF33 and ORF45 during the lytic cycle is required for accumulation of the ORF33 and production of viral particles. In addition to forming a complex with ORF33 and USP7, we also found a strong association of ORF45 to RSK and ERK during the lytic cycle, matching previous reports that ORF45 bound ERK and RSK in a single complex. During that study, ORF45 was found to form a complex with ERK and RSK and the formation of this complex lead to accumulation of active ERK and RSK. While the binding site of RSK was mapped to aa 55-70, it was unclear if ERK bound to ORF45 directly or through RSK. Using in vitro binding analysis, we identified an ERK binding site in aa 16-35. Using T-Coffee analysis, we compared the sequences of gamma-2 herpesvirus homologues of ORF45 and found two highly conserved phenylalanine residues at aa 32 and 34. After generating point mutants of each residue to alanine, we measured their effect on the activation of ERK and RSK induced by ORF45 and found that either mutation lead to decreased activation of ERK and RSK. We are currently evaluating their effect upon the activation of ERK and RSK during the lytic cycle and the production of progeny virions.
