Essential function of the integrator complex in Kaposi's sarcoma-associated herpesvirus lytic replication

Abstract

The integrator complex (INT) is an essential regulator of RNA biogenesis across evolution. Most current findings describe INT's function in states of equilibrium, presenting a research gap in INT's role in dynamic states, such as in infections and cancers. Viruses hijack cellular RNA machinery to transcribe their genes and produce viral progeny, presenting a unique condition to investigate INT-dependent RNA regulation under perturbation. Kaposi's sarcoma-associated herpesvirus (KSHV) is an oncogenic DNA virus that causes two deadly cancers, Kaposi's sarcoma and primary effusion lymphoma. KSHV undergoes a highly regulated and robust transcription of viral genes upon lytic reactivation, providing a complex and dynamic system to investigate integrator-mediated viral/host RNA regulation. We find that integrator subunit 11 (INTS11), the enzymatic core of INT, is essential for KSHV lytic replication triggered by reactivation or primary infection. Further RNA-seq analyses revealed a dynamic and unique signature of human transcriptomes during each lytic stage, respectively. Although the knockdown of INTS11 resulted in selective upregulation and downregulation of certain human gene transcription, INTS11's loss globally repressed the KSHV transcriptome throughout KSHV lytic replication. This inhibited viral lytic gene expression, viral genome replication, and virion production. Integrator subunits 9 and 6 are also important for KSHV lytic replication. Mechanistically, ChIP-seq analysis showed that INTS11 is increasingly recruited to the KSHV genome with some unique binding patterns as the lytic cycle progresses, suggesting that KSHV hijacks INTS11 during lytic gene transcriptions. In all, our findings reveal the essential roles of the Integrator complex in KSHV lytic replication.IMPORTANCEThe integrator complex (INT) is essential for RNA metabolism and is fundamental to all organisms, but its function and regulation during viral infection are not well described. Kaposi's sarcoma-associated herpesvirus (KSHV) infection establishes lifelong infection and causes two deadly cancers; however, no vaccine is available. Using KSHV as a model, we found that integrator subunit 11 (INTS11), the enzymatic core of INT, is recruited to the KSHV genome under lytic phases and plays an essential role in facilitating global KSHV lytic mRNA transcription and viral production. This reveals the critical role of INT in viral infection, a common and inevitable event in human life.

Keywords: INTS11; Kaposi's sarcoma-associated herpesvirus; herpesviruses; integrator; lytic; reactivation.

Fig 1

INTS11 deficiency causes massive and global repression of KSHV lytic gene transcription. iSLK.219 cells were transfected with control non-specific (NS) siRNA or INTS11 siRNA for 48 h and then treated with doxycycline (Dox, 0.2 µg/mL). (A) iSLK.219 cell construct contains a reporter element constitutively expressing GFP and RFP when treated with Dox. (B) Experiment schematic. (C and D) RNA was extracted, cDNA was synthesized, and the knockdown efficiency of (C) INTS11 and mRNA expression levels of (D) Pre-U1 were measured by RT-PCR. (E) Treated iSLK.219 cells were imaged at 0, 24, 48, and 72 h post reactivation. (F) RFP and GFP fluorescence values were measured using a microplate reader and graphed as a GFP/RFP ratio. mRNA expression levels of KSHV (G) ORF57, (H) ORF39, and (I) K8.1 were measured by RT-PCR. Target mRNA expression was normalized to GAPDH mRNA and presented as fold induction. (J) iSLK.219 cells were transfected with siNS or siINTS11 as described in the text and for panel B. A real-time qPCR-based KSHV transcriptome array was performed. Higher transcript expression levels are indicated by red and lower expression levels by blue as shown in the key. Details about this assay are described in Materials and Methods. hpr = hours post-reactivation. Error bars are representative of the standard deviation from three experiments analyzed with unpaired t-test statistics. (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

Fig 2

INTS11 is required for optimal KSHV genome replication, protein production, and virion production. Experiments were performed as described in Fig. 1B, (A) Genomic DNA was extracted and measured for the number of KSHV ORF39 copies/mL, normalized to genome β-Actin copies with real-time PCR. (B) Cell lysates were collected at the indicated time points, and western blots were performed with the indicated antibodies. (C) The supernatants were collected from treated iSLK.219 cells, and extracellular virion production was measured using the KSHV ORF39 gene. (D–G) Equal volumes of supernatant from 72 h reactivated iSLK.219 cells transfected with siNS or siINTS11 were added to naïve HEK293 cells to assess de novo KSHV infection. (D) Infection assay experiment schematic. (E) KSHV infection was visualized for GFP-positive cells at 24 hpi. (F) Ct (cycle threshold) of Actin qPCR in the infected HEK293 cells. (G) KSHV genome copy number was measured using the ORF39 genome primers and normalized to genome β-Actin copies with qPCR. hpr = hours post-reactivation. Error bars are representative of the standard deviation from three experiments analyzed with unpaired t-test statistics. (*P < 0.05, **P < 0.01, ***P < 0.001).

Fig 3

INTS11 is dispensable for the default KSHV primary infection to the latency route. SLK cells were treated with siNS or siINTS11 siRNA for 24 h. Equal volumes of supernatant containing KSHV virions from Dox-treated iSLK.219 cells were harvested and added to siINTS11-treated SLK cells. (A) Experiment schematic. (B–F) RNA was extracted, and cDNA was synthesized at the indicated time points post-infection. The knockdown efficiency of (B) INTS11 and the mRNA expression levels of (C) Pre-U1 were measured by real-time PCR. The mRNA expression levels of KSHV genes (D) ORF57, (E) ORF39, and (F) K8.1 were measured using real-time PCR. Target mRNA expression was normalized to GAPDH mRNA and presented as fold induction. (G) Genomic DNA was extracted and measured for the number of KSHV ORF39 copies/mL normalized to β-Actin with real-time PCR. (H) Cell lysates were collected at the indicated time points, and western blots were performed with the indicated antibodies hpi = hours post-infection. Error bars are representative of the standard deviation from three experiments analyzed with unpaired t-test statistics. (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

Fig 4

INTS11 is necessary for non-canonical KSHV primary infection to the lytic route. iSLK.RTA cells were treated with siNS or siINTS11 siRNA for 24 h. Equal volumes of supernatant containing KSHV virions from Dox-treated iSLK.219 cells were harvested and added to the treated SLK cells. (A) Experiment schematic. (B–F) RNA was extracted, and cDNA was synthesized at the indicated time points post-infection. The knockdown efficiency of (B) INTS11 and the mRNA expression levels of (C) Pre-U1 transcripts were measured by real-time PCR. The mRNA expression levels of KSHV genes (D) ORF57, (E) ORF39, and (F) K8.1 were measured using real-time PCR. Target mRNA expression was normalized to GAPDH mRNA and presented as fold induction. (G) Genomic DNA was extracted and measured for the number of KSHV ORF39 copies/mL normalized to β-Actin, with real-time PCR. (H) Cell lysates were collected at the indicated time points, and western blots were performed with the indicated antibodies hpi = hours post-infection. Error bars are representative of the standard deviation from three experiments analyzed with unpaired t-test statistics. (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

Fig 5

INTS9 and INTS6 facilitate KSHV lytic replication. (A) iSLK.219 cells were transfected with a control non-specific (NS) siRNA, siINTS9, or siINTS6 siRNA for 48 h and then treated with doxycycline (Dox, 0.2 µg/mL). (B–G) RNA was extracted, cDNA was synthesized, and the knockdown efficiency of (A) INTS9, (B) INTS6, and the mRNA expression levels of (C) Pre-U1 and the KSHV (E) ORF57, (F) ORF39, and (G) K8.1 were measured by real-time PCR. Target mRNA expression was normalized to GAPDH mRNA and presented as fold induction. (H) Genomic DNA was extracted and measured for the number of KSHV ORF39 copies/mL with real-time PCR. (I) The supernatants were harvested and measured for the number of extracellular KSHV ORF39 copies/mL with real-time PCR. (J) Cell lysates were collected at the indicated time points, and western blots were performed with the indicated antibodies. hpr = hours post-reactivation. Error bars are representative of the standard deviation from three experiments analyzed with unpaired t-test statistics. (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

Fig 6

INTS11 distinctly regulates KSHV and human gene transcription. iSLK.219 cells were transfected with control non-specific (NS) siRNA or INTS11 siRNA for 48 h and then treated with doxycycline (Dox, 0.2 µg/mL). RNAs were collected from 0, 24, 48, and 72 h after Dox treatment and subjected to Ribo-depletion RNA-seq. (A–H) Volcano plots demonstrate the change in gene expression for INTS11 knockdown cells compared with control cells at latency and 24, 48, and 72 hpr for (A–D) KSHV genes and (E–H) human genes. Genes exhibiting significantly different gene expression are marked in red. (I–L) Venn diagrams comparing the number of differentially expressed human genes at latency and immediate early (IE, 24 h), early (E, 48 h), and late (L, 72 h) lytic time points for (I and J) upregulated genes and (K and L) downregulated genes. (M–P) GO analysis demonstrates the biological processes most enriched in differentially expressed genes upon INTS11 knockdown. hpr = hours post-reactivation.

Fig 7

Dynamics of INTS11-dependent human gene expression in lytic stages upon reactivation from latency. iSLK.219 cells were transfected with control non-specific (NS) siRNA or INTS11 siRNA for 48 h and then treated with doxycycline (Dox, 0.2 µg/mL). RNAs were collected from 0 h, 24 h, 48 h, and 72 h after Dox treatment and subjected to Ribo-depletion RNA-seq. (A–C) Volcano plots demonstrating human genes that responded differently to INTS11 knockdown at (A) 24 h, (B) 48 h, and (C) 72 h post-reactivation compared with INTS11 knockdown during latency. Genes exhibiting significantly different gene expression are marked in red. (D and E) Venn diagrams demonstrate the overlap of genes that are (D) upregulated and (E) downregulated at immediate early (IE, 24 h), early (E, 48 h), and late (L, 72 h) lytic time points relative to the latent expression. (F–H) GO analysis demonstrates the biological processes most enriched in differentially expressed genes upon INTS11 knockdown.

Fig 8

ChIP-seq data reveal the preferential binding sites of INTS11 on the KSHV genome. iSLK.219 cells were transfected with control non-specific (NS) siRNA or INTS11 siRNA for 48 h and then treated with doxycycline (Dox, 0.2 µg/mL). IgG served as a control for normalization. (A) Coverage of INTS11 ChIP-seq reads across the KSHV genome. (B) Coverage of INTS11 ChIP-seq reads across the KSHV ORFs. (C and E) Metagene plots show the average binding position of INTS11 across all genes in the (C) KSHV and (E) human genomes. (D and F) Heatmaps showing the abundance of INTS11 reads across (D) KSHV and (F) human genes. All panels (A–F) use INTS11 read abundance normalized to the abundance of KSHV viral genomes in the ChIP input samples.