Transcriptional Elongation of HSV Immediate Early Genes by the Super Elongation Complex Drives Lytic Infection and Reactivation from Latency

Abstract

The cellular transcriptional coactivator HCF-1 is required for initiation of herpes simplex virus (HSV) lytic infection and for reactivation from latency in sensory neurons. HCF-1 stabilizes the viral Immediate Early (IE) gene enhancer complex and mediates chromatin transitions to promote IE transcription initiation. In infected cells, HCF-1 was also found to be associated with a network of transcription elongation components including the super elongation complex (SEC). IE genes exhibit characteristics of genes controlled by transcriptional elongation, and the SEC-P-TEFb complex is specifically required to drive the levels of productive IE mRNAs. Significantly, compounds that enhance the levels of SEC-P-TEFb also potently stimulated HSV reactivation from latency both in a sensory ganglia model system and in vivo. Thus, transcriptional elongation of HSV IE genes is a key limiting parameter governing both the initiation of HSV infection and reactivation of latent genomes.

Keywords: P-TEFb; herpes simplex virus; host cell factor-1; latency; super elongation complex; transcriptional elongation.

Figure 1. Transcription related protein complexes associated with HCF-1

(A) Subunits of complexes identified in MS analyses of HCF-1 immunoisolates from Mock infected and HSV-1 infected (2 hpi) are indicated with black dots. Targets and functions relative to transcription initiation and elongation are depicted. SET complex, KMT2A, KMT2C, KMT2E (histone H3K4 methyltransferases); DSIF and NELF (RNAPII pausing factors); SEC, P-TEFb (phosphorylates RNAPII CTD, DSIF, and NELF), SETD2 (histone H3K36 methyltransferase); FACT (histone chaperone/remodeling complex); TSS (transcription start site). (B) Functional interaction network of HCF-1-associated complexes involved in transcription initiation and elongation. The nodes represent baits identified during HCF-1 immunoaffinity purification and the edges indicate interactions from the Reactome database. See also Table S1.

Figure 2. Coimmunoprecipitation confirmations of HCF-1 MS partner identifications

(A) HCF-1 and control IgG precipitates from uninfected and HSV infected HEK293 cells were probed for protein partners identified by MS. (B) Schematic of RNAPII preinitiation complex, initiating RNAPII, and elongating RNAPII. CTD S5 phosphorylation by TFIIH promotes promoter clearance and transcription initiation while S2 phosphorylation by CDK9 promotes elongation. (C) HCF-1 and control IgG immunoprecipitates from Mock infected and HSV infected HEK293 cells were probed for forms of RNAPII.

Figure 3. AFF4 localizes to early HSV transcriptional foci and is required for efficient IE expression

(A) MRC5 cells were mock infected or infected with HSV-1 for 2 h. Cells were co-stained with antibodies to the SEC scaffold component AFF4 and the HSV-1 lytic activator ICP4. (B–C) MRC5 cells were transfected with siControl or siRNAs to AFF4 or BRD4. Cells were infected with HSV-1 for 2 h and mRNA levels of the siRNA target (AFF4, BRD4), control cellular gene (GAPDH), and viral IE genes (ICP4, ICP27, ICP22) were determined. Data are levels in siAFF4 or siBRD4 cells relative to control siRNA cells. (Means +/− s.e.m., n > = 6). (D) JQ1+, IBET-762, and HMBA bind to the bromodomains of BRD4 and inhibit its binding to chromatin. These compounds also induce release of 7SK snRNP sequestered P-TEFb. (E–F) MRC5 cells treated with JQ1+ 1 uM, IBET-762 1 uM, or HMBA 5 mM and infected with HSV-1 for 2 h. The mRNA levels of control cellular gene GAPDH and viral IE genes (ICP4, ICP27, ICP22) are relative to those in vehicle treated cells. (Means +/− s.e.m., n > = 6). (G–H) mRNA levels of HSV IE genes and controls (GAPDH, HPRT) in cells treated with the indicated concentrations of CDK9 inhibitors. (Means +/− s.e.m., n = 3). (I–J) MRC5 cells treated with vehicle, JQ1+, or dBET1 and infected with HSV-1 for 2 h. (I) Western blot of ICP4, BRD4, and control GAPDH protein levels. (J) mRNA levels of viral IE and GAPDH are relative to those in vehicle treated cells. (Means +/− s.e.m., n = 3). See also Figures S1–S6.

Figure 4. Depletion of AFF4 or BRD4 differentially impacts elongation of HSV IE gene transcription

(A) Schematic of small and large IE mRNAs RT-qPCR primer sets (product center). (B–D) The levels of large and small IE RNAs in AFF4 (B) or BRD4 (C) depleted MRC5 cells relative to siControl transfected cells. Cells were infected with HSV-1 for 2 h. BTUB, cellular control gene. The ratios of large to small RNA levels are shown for AFF4 and BRD4 depleted cells (D). (E) The levels of ICP4 large and small RNAs are shown in cells treated with vehicle, JQ1+, or the CDK9 inhibitor flavopiridol (FV) and infected with HSV-1 for 2 h. (B–E) Data are means +/− s.e.m., n = 4.

Figure 5. ChIP assays show pausing and elongation factors at viral IE genes

(A–B) ChIP assays illustrating the occupancy of RNAPII and NELF at viral IE genes (ICP4, ICP0) and the cellular MYC gene. Data are means +/− s.e.m., n = 6. (C) HCF-1, RNAPII-S2P, AFF4, and CDK9 occupancy levels of the ICP4 promoter proximal/TSS region in vehicle and JQ1+ treated cells at 2 hpi. (D) Fold occupancy in JQ1+ treated relative to vehicle treated cells. Data are means +/− s.e.m., n > = 3.

Figure 6. BET inhibitors drive viral reactivation in the mouse ganglia explant model system

(A) Viral yields from latently infected trigeminal ganglia explanted in the presence of the BRD4-BET inhibitor JQ1+ or control compounds (Vehicle, JQ1−) for 24 or 48 h. Data are yields from individual ganglia, n > = 15. (B) Trigeminal ganglia from latently infected mice were explanted in the presence of JQ1+ or controls (DMSO, JQ1−) for 48 h. Sections were stained for UL29 (HSV lytic ss-DNA binding replication protein, ICP8). (C) The numbers of UL29+ cells per ganglia, n = 12. (D–E) Trigeminal ganglia from latently infected mice were explanted in the presence of acyclovir (ACV) alone or in combination with JQ1+ or JQ1− for 48 h. Ganglia sections were stained for UL29 and the numbers of individual UL29+ cells per ganglia were quantitated, n = 12. (F) Trigeminal ganglia from latently infected mice were explanted in the presence of Vehicle, JQ1+, or JQ1+ and ACV. Viral (ICP27, gC) mRNA levels were determined at 4, 8, and 12 h post explant. (G) Trigeminal ganglia from latently infected mice were explanted in the presence of JQ1+ (1 uM), iCDK9 (0.1 uM), or JQ1+ and iCDK9. Viral mRNA levels were determined at 12 h post explant. (F–G) Samples were normalized based on the levels of cellular GAPDH mRNA. Data are means +/− s.e.m., n = 3 pools of 5 ganglia per group. See also Figure S7.

Figure 7. BET inhibitors drive HSV-1 viral reactivation from latency in vivo

(A) Latently infected mice were injected twice daily with JQ1+ or control vehicle and mRNA levels of viral lytic genes (ICP27, gC) and control GAPDH were determined at 24 and 48 h post injection. (B) Latently infected mice were injected with JQ1+ or control JQ1− and mRNA levels of viral lytic genes (ICP27, gC) are shown at the indicated times post injection. (A–B) Data are means +/− s.e.m., n = 3 pools of 5 ganglia per group. (C) Viral DNA at 48 h post injection in eyes of latently infected mice injected with vehicle or JQ1+. Data are means +/− s.e.m., n = 8. (A–C) Viral mRNA or DNA levels in samples were normalized based on the levels of cellular GAPDH mRNA or DNA respectively. *p < 0.05; **p < 0.01.