Deregulation of Host Gene Expression and Control of Wart Formation by Papillomavirus E8^E2

Abstract

Persistent infections with high-risk human papillomaviruses such as HPV16 can cause anogenital and oropharyngeal cancers. HPV infect keratinocytes in the basal layer of cutaneous or mucosal epithelia. The viral E8^E2 protein is a highly conserved repressor of PV replication and gene expression, but the advantage of the expression of E8^E2 for PV remains unclear. HPV16 genomes lacking E8^E2 expression (E8-) show also enhanced genome replication and viral late protein expression in differentiated cells. A small number of differentially expressed genes were identified by global transcriptome analysis of differentiated HPV16 wild-type (wt) and E8- cell lines. The analysis of selected genes suggested that their deregulation requires keratinocyte differentiation and correlated with viral late transcription. Consistent with this, the additional knock-out of the viral E4 and E5 genes, which are known to enhance productive replication, alleviated the deregulation of these host cell genes. These data reveal for the first time that productive HPV16 replication modulates host cell transcription. Through the expression and purification of the HPV16 E2 protein for subsequent polyclonal antibody production in rabbits, I was able to characterize the expression of E2 and E8^E2 proteins in HPV16 positive cell lines. A smaller number of wt compared to E8- cells display nuclear E2 foci and these colocalize with the replication protein RPA32 and thus represent most likely viral replication centers. E2 foci-positive cells also expressed the viral late E4 protein indicating that these cells have entered the productive replication cycle. This confirms that HPV16 E8^E2 regulates productive replication and makes it possible to further explore the roles of E2 and E8^E2 in the life cycle of HPV16. Since PV are highly species-specific, a suitable model to investigate E8^E2 function in vivo is the Mus musculus (Mmu) PV1-mouse model. Recent studies revealed that a MmuPV1 E8^E2 knock-out genome displays increased viral gene expression in cultured keratinocytes, but failed to form warts in T-cell deficient Foxn1nu/nu mice. I was able to show that the MmuPV1 E8^E2 protein interacts with NCoR/SMRT co-repressor complex components in an E8-domain dependent manner in order to repress transcription and replication, indicating that this interaction is conserved between MmuPV1 and HPV. Interestingly, MmuPV1 E8^E2 mt genomes display greatly increased expression of the viral late E4 protein in undifferentiated murine keratinocytes. MmuPV1 E4 caused an arrest in the G2 phase of the cell cycle, similar to its HPV counterparts, and this most likely prevents the expansion of virus-infected keratinocytes in the basal layer of the epithelium which explains the lack of wart formation in vivo. In summary, these data indicate that the main function of E8^E2 is to prevent the unscheduled induction of the productive replication cycle in basal-like keratinocytes.