Fig 1.
Diagram of the upstream signaling pathways that converge on activation of AGC kinases.
In dark blue are the KSHV proteins that have been shown to affect various nodes of the PI3K/Akt/mTOR and/or Raf/MEK/ERK signaling pathways. In red are the kinase inhibitors used in this study. AGC kinases, including Akt, p70S6K (S6K), and p90RSK (RSK), are shown in light blue, and phosphorylate their substrates at the RxRxxS*/T* motif, leading to regulation of key biological processes.
Fig 2.
The increased phosphorylation of putative RSK substrates induced by KSHV lytic reactivation is dramatically reduced by ORF45 mutation or deletion.
Stable iSLK.219 or iSLK.BAC16 cells carrying the indicated BAC (F66A, Stop45, or the F66A revertant (A66F)), were uninduced (U) or induced (I) with doxycycline and butyrate. Cell lysates were harvested at 0, 6, 12, 24, 48, and 72 hours post-induction (hpi) and analyzed by western blot using the indicated antibodies.
Fig 3.
Phosphoproteomic analysis allows for the identification of putative substrates of ORF45-activated RSK during KSHV lytic replication.
(A) Flowchart of the experimental design. (B) Phosphorylation motif for RSK, the 110B7E antibody used for the PhosphoScan, and the experimental data. (C) Scatterplot representation of the unique phosphopeptides identified by the screening that met the cutoff for deltaCN, intensity, and spectral counts. (D) Pie chart of the protein type/function distribution among the unique phosphopeptides with at least a 2.5-fold change (A66F:F66A).
Table 1.
List of genes representing the phosphopeptides identified by the phosphoproteomic screening with the highest fold-change in phosphorylation between A66F and F66A.
Fig 4.
DAVID bioinformatic analysis of the phosphoproteomic screening results.
The enrichment of Gene Ontology (GO) terms associated with the putative substrates of ORF45-activated RSK (121 unique phosphopeptides with ≥2.5-fold change in phosphorylation between A66F and F66A) was assessed in the background of the human proteome. The x-axes values denote the percentage of the 98 unique proteins that were enriched (black bars) or the fold enrichment (gray bars) for the indicated GO term, functional category, or protein domain. P-values were all <0.05. The exact p-values and additional information can be found in S2 File.
Fig 5.
The phosphorylation of several substrates with roles in translational regulation is mediated by ORF45-activated RSK.
(A) The indicated cell lines were uninduced (U) or induced (I) as described previously. Two days post-induction (dpi), cells lysates were collected and subjected to western blot analysis with the indicated antibodies. (B) Pharmacological inhibition confirms that KSHV-induced phosphorylation of eIF4B is dependent on RSK activity. iSLK.BAC16 A66F cells were uninduced (U) or induced (I) as described previously. At 46 hpi, cells were treated for 2 h with the indicated kinase inhibitor (the target of which is listed above in parentheses), then cells were harvested and lysates were analyzed by western blot with the indicated antibodies.
Fig 6.
Nuclear IP further validated ORF45/RSK-dependent phosphorylation of a subset of substrates.
Nuclei were isolated from the indicated cell lines at 48 hpi, then subjected to IP using anti-RxRxxS*/T* magnetic beads. The input, eluate (IP), and flow-through (FT) were analyzed by western blot with the indicated antibodies.
Fig 7.
CRISPR/Cas9-mediated knockout of RSK1/2 results in the reduction of RSK substrate phosphorylation, viral late lytic gene expression, and virion production.
(A and B) SLK-iBAC cells transduced with empty lentiCRISPR or single cell clones with RSK1/2 knockout (#1 and #15) were uninduced (U) or induced (I) with dox/butyrate, and cells were harvested at the indicated dpi. Lysates were analyzed by western blot with the indicated antibodies. (C) The indicated cell lines were induced as in (A and B). Twenty-four hpi, cells were harvested and RNAs were extracted, reverse transcribed, and subjected to qRT-PCR analysis using the indicated primers. Relative mRNA level is shown normalized to β-Actin. (D) KSHV progeny virion production is compromised by RSK knockout. SLK-iBAC cells were induced as described in (A and B). At the indicated hpi, media were collected from cells, extracellular virion DNA was extracted, and viral genome copy number was measured by qPCR.
Fig 8.
CRISPR screen characterizes the contributions of putative RSK substrates to ORF45/RSK-mediated transactivation and KSHV progeny virion production.
(A) NH1 (top) and NH2 (top) reporter cells were stably transduced with the indicated lentiCRISPR, transfected with pCR3.1 or pCR3.1-ORF45, and subjected to luciferase assays. Shown are the relative light units (RLU) for firefly luciferase (FL) normalized to the NH1 FL WT sample. Values shown are the average of three biological replicates. (B) SLK-iBAC cells were stably transduced with the indicated lentiCRISPR, then induced as described previously. Extracellular virions were collected from the media at 5 dpi, and virion DNA was extracted and quantified by qPCR. Values shown are the average of three technical replicates from each of two biological replicates. Knockout of TAF3 in SLK-iBAC cells was not viable. (A and B) For RSK substrate knockout, values shown are the average of knockout cells generated from two independent gRNAs. Significance compared to empty lentiCRISPR sample: *p < 0.05; **p < 0.01; ***p <0.001; N.S.–not significant.
Fig 9.
ORF45/eIF4B-dependent translational control of mRNAs with complex 5’ UTR structure.
(A and B) 293T WT cells or cells stably transduced with the indicated lentiCRISPR were transfected with the indicated plasmids, as well as HCV IRES-FL (firefly luciferase), and subjected to luciferase assays (A) or western blot analysis (B) at 24 hpt. Values for (A) are represented as the renilla/firefly luciferase ratio normalized to the WT sample. (C) iSLK.BAC16 A66F, F66A, or Stop45 cells were uninduced or induced with dox/butyrate, and cells were harvested at 48 hpi. Lysates were fractionated by sucrose gradient centrifugation and RNAs were extracted. RNAs associated with either monosome or polysome fractions were pooled, reverse transcribed, and subjected to qRT-PCR using the indicated primers. Values shown are for mRNA level relative to the uninduced control, as a ratio of polysome:monosome (normalized to β-Actin). Genes are arranged in order of decreasing minimum free energy (left to right), as predicted by RNAFold software [109].
Fig 10.
Summary of the mechanisms by which KSHV ORF45 manipulates ERK/RSK signaling to regulate translation.
ORF45 regulates translation at multiple levels through direct and indirect mechanisms, including: (1) RSK-dependent phosphorylation of eIF4B and S6, (2) ERK/RSK-mediated inhibition of TSC2, relieving its inhibition of mTORC1, and (3) crosstalk with Akt/mTOR, thereby enhancing S6K-dependent phosphorylation of S6. Several other factors, such as GSK-3β, PRAS40, and PDCD4 (all of which were identified by our analyses to have decreased phosphorylation at key regulatory residues upon ORF45 F66A mutation), also contribute to this phenomenon.