PRC1-independent binding and activity of RYBP on the KSHV genome during de novo infection

Abstract

Kaposi's sarcoma-associated herpesvirus (KSHV) is an oncogenic virus that causes lifelong infection in humans by establishing latency after primary infection. Latent infection is a prerequisite for both persistent infection and the development of KSHV-associated cancers. While viral lytic genes are transiently expressed after primary infection, their expression is significantly restricted and concomitant with the binding of host epigenetic repressors Polycomb Repressive Complex 1 and 2 (PRC1 and PRC2) to lytic genes. PRC1 and PRC2 mediate the repressive histone marks H2AK119ub and H3K27me3, respectively, and maintain heterochromatin structure on KSHV lytic genes to inhibit their expression. In contrast to PRC2, little is known about the recruitment and role of PRC1 factors on the KSHV genome following de novo infection. Thus, the goal of this study was to examine the function of PRC1 factors in the establishment of KSHV latency. To address this question, we performed an shRNA screen targeting 7 different components of the canonical and non-canonical PRC1 complexes during primary KSHV infection. We found that RYBP, a main subunit of the non-canonical PRC1 complexes, is a potent repressor of KSHV lytic genes that can bind to the viral genome and inhibit lytic genes as early as 4 hours post infection. Surprisingly, our ChIP analyses showed that RYBP binds to lytic viral gene promoters in a PRC1-independent manner, does not affect PRC1 activity on the KSHV genome, and can reduce the level of histone marks associated with transcription elongation. Our data also suggest that RYBP can repress the viral lytic cycle after primary infection by inhibiting the transcription elongation of the lytic cycle inducer KSHV gene RTA. Based on our results we propose that RYBP uses a PRC1-independent mechanism to block KSHV RTA expression thereby promoting the establishment of KSHV latency following de novo infection.

Fig 1. Analysis of viral gene expression changes upon shRNA knockdown of PRC1 factors during primary KSHV infection.

(A) Components of the canonical and non-canonical Polycomb Repressive Complex 1. (B-D) SLK cells were transduced with lentiviruses expressing shRNAs that were specific for different subunits of PRC1. Three days post-transduction, the cells were infected with KSHV for 72 hours. (B) RT-qPCR and immunoblot analyses of the expression of PRC1 factors after shRNA knockdown. The percentage of remaining transcript expression is shown relative to the shControl. (C) RT-qPCR analysis of the expression of viral genes including immediate early (IE), early (E), and late (L) genes. (D) Immunoblot analysis of viral protein expression. Tubulin was used as a loading control.

Fig 2. RYBP inhibits KSHV lytic genes during de novo infection, which are dependent on RTA expression.

(A-B) shControl- and shRYBP-treated SLK cells were infected with BAC16-3xFLAG-RTA KSHV for 4, 8, 16, 24, 48, and 72 hours. (A) Immunoblot analysis of RYBP and viral protein expression. (B) RT-qPCR analysis of viral gene expression. The relative fold change represents the induction of viral gene expression in shRYBP-treated cells relative to shControl-treated cells at different time points of KSHV infection. (C) Viral DNA level at 72 hpi was measured by qPCR, normalized to the host DNA level, and then calculated relative to 2 hpi in shControl- and shRYBP-treated SLK cells. (D-E) shControl- and shRYBP-treated SLK cells were infected with wild-type (WT) and RTA knockout (RTA-KO) KSHV for 72 hours. (D) Immunoblots analysis of RYBP and viral protein expression. (E) Measuring viral gene expression by RT-qPCR. The relative fold change represents the induction of viral gene expression in shRYBP-treated cells relative to shControl-treated cells.

Fig 3. Analysis of the enrichment of histone modifications and the binding of PRC1 factors on KSHV promoters during de novo KSHV infection.

SLK cells were infected with KSHV, and ChIP-qPCR analysis was performed on LANA, RTA, K2, ORF25 promoters at the indicated time points for (A) H3K4me3, (B) H2AK119ub, (C) H3K27me3, (D) RYBP, and (E) RING1B.

Fig 4. Testing the effect of PRC1 and PRC1-recruiting transcription factors on RYBP binding to KSHV promoters.

(A-D) SLK cells were transduced with shControl, shRING1A, shRING1B, shKDM2B, and shYY1 lentiviruses for 72 hours followed by KSHV infection for 24 hours. (A) Immunoblot analysis to confirm the shRNA knockdown of RING1A, RING1B, and YY1 at the protein level. (B) Confirming the shRNA knockdown of KDM2B by RT-qPCR. The percentage of remaining RNA transcript level is shown relative to the shControl. (C) ChIP for H2AK119ub. (D) ChIP for RYBP. (E-F) SLK cells were co-transduced with shRING1A and shRING1B lentiviruses for 72 hours, followed by KSHV infection for 24 hours. (E) The shRNA knockdown of RING1A/B was confirmed by immunoblot. (F) ChIP for the indicated factors on viral promoters and RTA gene body. The t-test was performed between shControl and shRING1A, shRING1B, shKDM2B, shYY1 and between shControl and shRING1A/B where p<0.05 (*) was considered statistically significant.

Fig 5. Identification of the functional domain of RYBP that is required for its binding to KSHV promoters.

(A) Schematic representation of the domain structure of RYBP. NZF, Npl4 Zinc finger motif possessing ubiquitin binding activity; RBD, RING1-binding domain; NLS1-3, nuclear localization signals. Point mutations in NZF mutant (NZFm) are indicated. (B-C) SLK cells were transduced with lentiviruses to express GFP (negative control), 3xFLAG-RYBP WT, N158, N79, or aa80-158 RYBP truncation mutants for 3 days, followed by KSHV infection for 24 hours. (B) FLAG immunoblot analysis of the expression of WT and truncation mutants of 3xFLAG-RYBP in SLK cells. (C) FLAG-ChIP analysis for testing the binding of 3xFLAG-tagged RYBP proteins on viral lytic promoters and RTA gene body. The t-test was performed between 3xFLAG-RYBP and 3xFLAG-RYBP N79 samples, and p<0.05 (*) was considered statistically significant. (D) FLAG IP using 293T cells expressing 3xFLAG-RYBP WT or truncation mutants. (E-F) SLK cells were transduced with lentiviruses expressing GFP, WT, NZFm, or dRBD 3xFLAG-RYBP for 3 days, followed by KSHV infection for 24 hours. (E) Immunoblot analysis of the expression of WT and mutants of 3xFLAG-RYBP in SLK cells. (F) FLAG ChIP analysis to test the binding of 3xFLAG-tagged RYBP proteins on viral lytic promoters and RTA gene body. Lenti-GFP was used as a negative control in the experiments. (G) FLAG IP using 293T cells expressing 3xFLAG-RYBP WT or mutants. EV, empty vector was transfected.

Fig 6. Testing the effect of RYBP depletion on the KSHV epigenome at 24 hpi.

shControl- and shRYBP-treated SLK cells were infected with RTA-KO KSHV for 24 hours. (A) Immunoblot analysis of host epigenetic factors. (B) FAIRE assay to analyze the chromatinization of KSHV promoters during de novo KSHV infection. (C) ChIP analysis of the enrichment of H3K4me3, H2AK119ub, and RING1B on viral promoters. (D) RING1B ChIP on host promoters. (E) ChIP analysis of H3K36me3, H3K79me2, and H2BK120ub at viral promoters and on RTA gene body. The t-tests were performed between shControl and shRYBP samples. p<0.05 (*) was considered statistically significant.

Fig 7. The effect of RYBP on transcription initiation and elongation at viral genes.

(A) Schematic representation of the structure of RTA and ORF57 genes. The location of RT-qPCR primers targeting the proximal (prox) and distal (dist) regions are indicated. (B) SLK cells were transduced with shControl- and shRYBP-expressing lentiviruses for 72 hours, followed by KSHV infection for 8 hours. RT-qPCR analysis of viral mRNA levels at the proximal and distal gene regions of RTA and ORF57 upon RYBP depletion relative to shControl. (C) shControl- and shRYBP-treated SLK cells were treated with 30 nM AZD4573 and infected with KSHV for 8 hours. RT-qPCR analysis of viral mRNA levels at the proximal and distal gene regions of RTA and ORF57 upon RYBP depletion relative to shControl samples. The t-test was performed between DMSO-treated and AZD4573-treated shRYBP samples and p<0.05 (*) was considered statistically significant. (D) Immunoblots analysis of RYBP and Pol II expression as well as Ser2 phosphorylation levels of Pol II.