Heterogeneous Polycomb Silencing of Incoming Herpes Simplex Virus Genomes During Productive Infection

Herpes simplex virus 1 (HSV-1) establishes life-long infection with distinct lytic and latent phases in the epithelial mucosa and peripheral nervous system respectively. Reactivation is triggered by factors including stress, fever, sun exposure and hormonal changes, leading to renewed lytic replication. How lytic and latent infection are shaped in these distinct contexts is unclear, but must be understood to inform the goal of preventing reactivation.

During latency, viral DNA exists in the nucleus as repressive heterochromatin. Latent HSV-1 chromatin features the histone post-translational modification H3K27me3, laid by Polycomb repressor complex 2 (PRC2). In non-neuronal cells, viral chromatin is instead permissive to lytic gene expression. We tested the assumption in the field that H3K27me3 is rapidly deposited and removed from the viral genome in fibroblasts, and determined that H3K27me3 does not associate with lytic HSV-1 genomes. Using a novel assay imaging individual viral genomes, we found that H3K27me2 forms on a subset of genomes and represses lytic gene expression. We determined that H3K27me2 reader PHF20L1 co-localizes with incoming viral DNA, indicating that transcriptional repression by H3K27me2 could be mediated via PHF20L1 activity.

We then explored whether the other branch of Polycomb silencing, Polycomb repressor complex 1 (PRC1)-mediated H2AK119ub, regulates HSV-1 gene expression. H2AK119ub was detected on viral chromatin from latency established in vivo, providing an updated picture of latent chromatin that features both Polycomb-associated modifications. We did not detect H2AK119ub on lytic chromatin, but found that PRC1 activity is pro-transcriptional to lytic gene expression. PRC1 subunits RING1A/B both co-localized with lytic viral genomes, and pro-transcriptional H2AK119ub reader Zuotin- related factor 1 (ZRF1) co-localized with incoming genomes and replication compartments. These findings point to PRC1 promoting lytic gene expression through reader protein ZRF1.

We conclude that both branches of Polycomb silencing contribute to the transcriptional regulation of HSV-1 infection. PRC2 activity represses lytic gene expression, whereas PRC1 activity is pro-transcriptional in lytic infection of fibroblasts. We also identify reader proteins PHF20L1 and ZRF1 that may mediate the transcriptional regulation by PRC2 and PRC1 activity respectively. We propose these two axes of transcriptional regulation by Polycomb complexes, aspects of which may differ between non-neuronal and neuronal infection and shape the cell type-specific outcomes. We also illustrate the value of capturing heterogeneity between viral genomes, an approach that revealed previously unobserved roles for Polycomb regulation of HSV-1 infection.