Brd4 Activates Early Viral Transcription upon Human Papillomavirus 18 Infection of Primary Keratinocytes

Abstract

Human papillomaviruses (HPVs) replicate in the cutaneous and mucosal epithelia, and the infectious cycle is synchronous with the differentiation program of the host keratinocytes. The virus initially infects dividing cells in the lower layers of the epithelium, where it establishes a persistent infection. The viral genome is maintained as a low-copy-number, extrachromosomal element in these proliferating cells but switches to the late stage of the life cycle in differentiated cells. The cellular chromatin adaptor protein Brd4 is involved in several stages and processes of the viral life cycle. In concert with the viral transcriptional regulator E2, Brd4 can repress transcription from the early viral promoter. Brd4 and E2 form a complex with the viral genome that associates with host chromosomes to partition the viral genome in dividing cells; Brd4 also localizes to active sites of productive HPV DNA replication. However, because of the difficulties in producing HPV viral particles, the role of Brd4 in modulating viral transcription and replication at the initial stage of infection is unclear. In this study, we have used an HPV18 quasivirus-based genome delivery system to assess the role of Brd4 in the initial infectivity of primary human keratinocytes. We show that, upon infection of primary human keratinocytes with HPV18 quasivirus, Brd4 activates viral transcription and replication. Furthermore, this activation is independent of the functional interaction between Brd4 and the HPV18 E2 protein.

Importance: HPVs lack encapsidated proteins and so rely exquisitely on host cellular factors to initiate their gene expression programs in newly infected cells. Brd4 is an important cellular chromatin adaptor molecule that normally activates host transcription initiation and elongation. In this study, we further optimize and utilize a quasivirus infection system to show that Brd4 activates HPV18 transcription at early stages of infection. HPVs are important human pathogens causing a wide range of cutaneous and tumorigenic morbidities. Therefore, specifically targeting this protein could provide a new target of therapeutic prevention of establishment of HPV infections.

FIG 1

Overview of quasivirus stock production and analysis. (A) 293TT cells were cotransfected with recircularized HPV18 genome and an HPV16 L1/L2 capsid protein expression plasmid. Forty-eight hours posttransfection, cells were lysed and virions were extracted in high-salt buffer. Clarified lysates were separated through a 27 to 39% OptiPrep density gradient by high-speed centrifugation. Fractions were collected, and total protein in each was resolved by SDS-PAGE and visualized by SYPRO Ruby staining. Fractions that contained predominantly L1 and L2 and cellular histones were pooled as a virus stock. (B) Pooled virus stocks were diluted and resolved by SDS-PAGE to qualitatively and quantitatively determine capsid amount and L1/L2 ratio. (C) Electron micrograph of 50-nm quasivirus particles.

FIG 2

Infection of primary keratinocytes with HPV18 quasivirus and neutralization of infection with Cervarix antiserum. (A) Primary HFKs were infected with HPV18 quasivirus at 100 VGE/cell and incubated for the indicated times. E1^E4 and E6*I spliced early transcripts were detected by qRT-PCR and normalized to human TATA binding protein (TBP) transcripts. (B) Quasivirus was preincubated with the indicated dilutions of rabbit serum obtained pre- and post-vaccination with Cervarix. Primary keratinocytes were infected at 100 VGE/cell, and at 72 h postinfection, viral E1^E4 transcript abundance was determined with qRT-PCR and normalized to human TATA binding protein. Percent neutralization compared to preimmune serum is shown on the right. n = 3; error bars show standard errors of the means.

FIG 3

HPV18 replicates in 293TT cells only when E1 and E2 are coexpressed, enabling a DpnI resistance replication assay in quasivirus-infected keratinocytes. (A and B) Recircularized HPV18 genomes were cotransfected into 293TT cells along with pMEP E1 and E2 expression plasmids. Replicated DNA is resistant to DpnI and sensitive to MboI cleavage. 293TT cells replicate the HPV18 genome only in the presence of exogenously expressed E1 and E2 proteins, as shown by Southern blot analysis (n = 2) (A) and qPCR (n = 3) (B). (C) E1 and E2 expression enhances HPV18 replication in cells cotransfected with the pSHELL plasmid and actively packaging viral genome (n = 3). (D) HPV18 genome-containing capsids are enriched in replicated viral DNA when produced from E1/E2-expressing cells (n = 3). (E) The packaging efficiency of HPV18 genomes is greatly increased when produced from E1/E2-expressing cells (n = 3). (F) Quasiviruses containing replicated and unreplicated genomes have similar transcriptional activities upon HFK infection (n = 3). (G) Quasiviruses containing unreplicated viral genomes allow analysis of nascently produced viral DNA after infection of primary HFKs. Total and replicated (DpnI-resistant) DNA was detected by qPCR and normalized to β-actin (n = 3). (A to E) Forty-eight hours posttransfection. (F and G) Seventy-two hours postinfection. Error bars show standard errors of the means. A paired t test was used for statistical analysis.

FIG 4

Brd4 depletion reduces HPV18 transcript abundance and viral DNA replication in infected HFKs. (A and B) HFKs were transfected with 20 nM Brd4-targeting or All* negative-control siRNA. Twenty-four hours posttransfection, cells were infected with 100 VGE/cell of HPV18 quasivirus. Seventy-two hours postinfection, cells were harvested for RNA, DNA, and protein. Knockdown efficiency was determined by Brd4 RNA (A) and Brd4 protein abundance (B). Brd4 protein levels were determined by immunoblotting, and α-tubulin is provided as a control for equal loading. Cell growth was monitored in an IncuCyte microscope to assess toxicity of the Brd4 downregulation (see Fig. S1 and Movie S1 in the supplemental material). (C and D) HPV18 E1^E4 and E6*1 transcripts were detected by qRT-PCR, corrected to TATA binding protein (TBP) transcripts (C), and normalized to siCtrl (D). n = 5. (E and F) Total DNA was digested with DpnI, and the abundance of nascently replicated HPV18 genomes was detected by qPCR, corrected to β-actin (E), and normalized to siCtrl (F). n = 3. Error bars show standard errors of the means. A paired t test was used for statistical analysis (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001).

FIG 5

Expression and Brd4 binding of E2 mutants. (A) Western blot analysis of protein extracts from C33A cell lines stably transfected with pMEP4 (no E2), wild-type pMEP-E2 (WT), or mutated pMEP-E2 (R41A, I77A, and R41A/I77A). E2 proteins were tagged with a FLAG epitope and detected with FLAG-M2 monoclonal antisera. α-Tubulin is provided as a control for equal cell number. (B) E2 expression was corrected to tubulin expression and normalized to WT. Raw band volume was quantified by Gene Tools software (Syngene). n = 3. Error bars represent standard errors of the means. (C) ³⁵S-labeled in vitro-translated E2 proteins were mixed with in vitro-translated Brd4 and immunoprecipitated with the Brd4-specific antibody 2290. Immune complexes were eluted and resolved by SDS-PAGE. Shown is a representative audioradiograph. (D) E2 bands were detected by using a Typhoon phosphorimager and quantified with Gene Tools software (Syngene). Brd4-bound E2 was corrected for background binding to rabbit IgG and normalized to WT. n = 3. A paired t test was used for statistical analysis (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001).

FIG 6

Efficient gene expression is not repressed by the Brd4 binding function of HPV18 E2 upon de novo infection. Quasivirus inocula containing WT HPV18 genomes or genomes mutated in one or both of the key Brd4 binding E2 transactivation domain residues were used to infect HFKs at 100 VGE/cell. (A and B) E1^E4 and E6*1 early spliced transcripts were measured by qRT-PCR 72 h postinfection, corrected to TATA binding protein (TBP) transcripts (A), and normalized to WT (B). n = 3. Error bars show standard errors of the means. (C and D) Total DNA was digested with DpnI, and the abundance of newly replicated HPV18 genomes was detected by qPCR and normalized to β-actin. n = 3. Error bars show standard errors of the means. A paired t test was used for statistical analysis; ns, not significant.