The viral packaging motor potentiates Kaposi's sarcoma-associated herpesvirus gene expression late in infection

Abstract

β- and γ-herpesviruses transcribe their late genes in a manner distinct from host transcription. This process is directed by a complex of viral transcriptional activator proteins that hijack cellular RNA polymerase II and an unknown set of additional factors. We employed proximity labeling coupled with mass spectrometry, followed by CRISPR and siRNA screening to identify proteins functionally associated with the Kaposi's sarcoma-associated herpesvirus (KSHV) late gene transcriptional complex. These data revealed that the catalytic subunit of the viral DNA packaging motor, ORF29, is both dynamically associated with the viral transcriptional activator complex and potentiates gene expression late in infection. Through genetic mutation and deletion of ORF29, we establish that its catalytic activity potentiates viral transcription and is required for robust accumulation of essential late proteins during infection. Thus, we propose an expanded role for ORF29 that encompasses its established function in viral packaging and its newly discovered contributions to viral transcription and late gene expression in KSHV.

Fig 1. The vTA complex is associated with numerous cellular and viral factors during infection.

(A) TurboID was tethered to the C-terminus of ORF18. ORF18 directly contacts ORF30, ORF31, ORF66, and ORF34 and is proximal to ORF24 and RNA polymerase II. (B) Infectious virion production from reactivated iSLKs was measured by viral supernatant transfer onto HEK293T cells using flow cytometry for the GFP-expressing virus. Data are from three independent biological replicates and statistics are calculated using an unpaired t test. * = P<0.05. (C) Overlap between proteins identified by mass spectrometry with ≥2 unique peptides that demonstrated >2-fold increase in the ORF18-TurboID + biotin test condition compared to the three negative controls, across 2 biological replicates. Forty-five high confidence hits were specifically enriched in ORF18-TurboID + biotin sample using these filtering criteria. (D) STRING protein-protein interaction network of high-confidence cellular proteins (above) and KSHV proteins (below). ORF18 is shown in dark blue. Known interactors of ORF18 are in light blue, and all other proteins are shown in grey. Proteins with nuclear localization annotated in UniProt are outlined in gold. (E) Gene ontology enrichment analysis of cellular ORF18-TurboID hits. All major enriched functional classes are shown.

Fig 2. Functional analysis of ORF18-interacting proteins.

(A) Schematic of the late gene reporter virus used to measure late gene expression by flow cytometry. The reporter construct was cloned into a region of the vector backbone of the KSHV BAC16. The K8.1 late gene promoter drives expression of eGFP, and mIFP is constitutively expressed from the EF-1a promoter. (B) Late gene reporter signal for CRISPR knockouts of viral genes at 72 h post reactivation. Knockouts were normalized against the non-targeting control (NT) within each replicate. Data are from three independent biological replicates, **** = P<0.0001. P values calculated from ordinary one-way ANOVA test. (C) Late gene reporter signal for CRISPR knockouts of cellular genes at 72 h post reactivation. Knockouts were normalized against the non-targeting control (NT) within each replicate. Data are from three independent biological replicates, **** = P<0.0001, * = P<0.05. P values calculated from ordinary one-way ANOVA test. (D) Early gene reporter signal for CRISPR knockouts of viral and cellular genes at 48 h post reactivation. Knockouts were normalized against the non-targeting control (NT) within each replicate. Data are from three independent biological replicates, **** = P<0.0001. P values calculated from ordinary one-way ANOVA test. (E) Viral DNA replication was measured using qPCR to compare reactivated to non-reactivated cells for each CRISPR knockout line. Data are from three independent biological replicates, * = P<0.05. P values calculated from ordinary one-way ANOVA test.

Fig 3. siRNA-based depletion of select TurboID hits suggests a generalized role for PABC proteins in viral gene expression.

(A) Representative western blots of siRNA-treated WT iSLK cell lysates at 48 h post reactivation showing the efficiency of XRN1, CASK, and PABPC1 depletion as well as levels of the representative late protein, K8.1, and early protein, ORF6. Vinculin serves as a loading control. (B) Representative western blots of WT iSLK cells at 48 h post reactivation that were depleted of PABPC1 and PABPC4 individually or in tandem by siRNA treatment. Levels of late proteins K8.1 and ORF26 are shown, as are levels of early proteins ORF6, ORF68, and ORF59. Vinculin serves as a loading control.

Fig 4. Loss of ORF29 impairs late gene transcription.

(A) Western blot showing the expression kinetics of ORF29 relative to the early proteins ORF59, ORF6, and ORF68, and the late protein K8.1 in reactivated WT iSLK cells. (B) Infectious virion production from reactivated iSLK cells was measured by viral supernatant transfer using flow cytometry. Data are from three independent biological replicates, **** = P<0.0001. P values calculated from ordinary one-way ANOVA test. (C) Representative western blots of lysates harvested from iSLK cells at 72 h post reactivation. Levels of the late proteins, K8.1 and ORF26, are shown, as are levels of the early proteins, ORF6, ORF59, and ORF68. GAPDH serves as a loading control. (D) RNA levels of the indicated genes were measured by RTqPCR. Reactivated cells were harvested at 48 h post reactivation. Data are from three independent biological replicates, normalized to WT for each transcript, **** = P<0.0001, ** = P<0.005, * = P<0.05. P values calculated from ordinary one-way ANOVA test. (E) RNA levels of the indicated genes were measured by RTqPCR. Reactivated cells were harvested at 72 h post reactivation. Data are from three independent biological replicates, normalized to WT for each transcript, **** = P<0.0001, ** = P<0.005, * = P<0.05. P values calculated from ordinary one-way ANOVA test. (F) RNA levels of the indicated genes were measured by RTqPCR. Reactivated cells were harvested at 24 h post reactivation. Data are from three independent biological replicates, normalized to WT for each transcript, **** = P<0.0001, ** = P<0.005, * = P<0.05. P values calculated from ordinary one-way ANOVA test. (G) Viral DNA replication was measured using qPCR in cells harvested 24, 48, and 72 h post reactivation. Fold DNA replication values were calculated by comparing reactivated to non-reactivated cells. Data are from three independent biological replicates, * = P<0.05. P values calculated from ordinary one-way ANOVA test.

Fig 5. Complementation with ORF29 functional mutants fails to rescue late gene transcription in the background of an ORF29.stop virus.

(A) Schematic of the ORF29 coding sequence. Positions of the Walker A and Walker B sequences are indicated, as are the residues that have previously been implicated in magnesium coordination and nuclease activity. The point mutants used in subsequent experiments are indicated in red. (B) Infectious virion production from reactivated iSLK cells was measured by viral supernatant transfer using flow cytometry. Data are from three independent biological replicates, **** = P<0.0001. P values calculated from ordinary one-way ANOVA test. (C) Representative western blots of lysates harvested from iSLK cells harboring the ORF29.stop virus complemented with ORF29 variants at 72 h post reactivation. Levels of the late protein K8.1 and early proteins ORF59 and ORF6 are shown. GAPDH serves as a loading control. (D) RNA levels of the indicated genes were measured by RTqPCR. Reactivated cells were harvested at 72 h post reactivation. Data are from three independent biological replicates, normalized to WT for each transcript, ** = P<0.005. P values calculated from ordinary one-way ANOVA test. (E) Viral DNA replication was measured using qPCR in cells harvested 72 h post reactivation. Fold DNA replication values were calculated by comparing reactivated to non-reactivated cells. Data are from three independent biological replicates. P values calculated from ordinary one-way ANOVA test.

Fig 6. Models for the contribution of ORF29 to viral transcription.

(A) Newly replicated concatemeric herpesvirus genomes are shown with aberrant DNA structures that are thought to arise during replication. As a factor that can both move along and cleave DNA, ORF29 may function to resolve DNA structures that could stall RNA polymerase or are otherwise detrimental to transcription. Light purple regions on the DNA denote terminal repeats at the end of each unit length genome. (B) DNA replication and transcription can generate torsional strain shown as supercoiled DNA. If left unresolved, torsional strain can impair transcription. ORF29 activity may relieve accumulated torsional strain to allow for full viral transcription.