Human Papillomavirus Infection Inhibits Sting Phosphorylation To Limit cGAS/Sting Activity

Abstract

Human Papillomavirus (HPV) is a small double-strand DNA virus that infects the cutaneous and mucosal epithelium. A wound or micro-abrasion permits access of the virion to the basal stem cell keratinocytes to establish infection and begin the stable expression of a small number of viral genome copies as epichromosomes (episomes). Over time, a reservoir of HPV+ basal stem cells grow, evading the host immune system’s attempts at clearing most of the infection, which is critical for long-term HPV persistence: the main risk factor for high-risk HPV-dependent oncogenesis and the cause of ~5% of cancers worldwide. Therefore, studying the mechanisms of immune evasion, and especially early detection by innate immune sensors, during HPV persistence is important for the discovery of HPV-specific therapeutics that are sorely lacking in the medical field. Here, we present results of our studies on HPV inhibition of the cytosolic DNA-sensing cGAS/STING pathway. Briefly, the enzyme cGAS was initially discovered to be a cytosolic DNA sensor that binds dsDNA and catalyzes the production of cGAMP which binds ER-resident STING. Activated STING is transported to the Golgi, where it recruits TBK1, which phosphorylates STING and IRF3. Phospho-IRF3 translocates into the nucleus to act as a transcription factor of type-I IFN genes, promoting a proinflammatory response. Previous studies by Dr. Brittany Uhlorn showed that cGAS/STING activity is disrupted during mitosis, and that during HPV entry and initial establishment of infection, the virus may have possibly evolved to time genome transport into the nucleus until the onset of mitosis (1,2). Dr. Uhlorn then proceeded to study cGAS/STING immune evasion post-establishment in our novel model of HPV persistent infection that consists of primary human foreskin keratinocyte (HFKs) cell lines co-cultured with irradiated mouse fibroblasts. Multiple studies have been done by other groups observe HPV16/18 inhibition of cGAS/STING and other innate immune pathways that promote an antiviral response, albeit these studies were often done with immortalized cell lines with integrated HPV genomes (e.g., HeLa cells), or in cell lines that overexpress HPV oncoproteins. These models are not as representative of HPV persistence as our primary co-culture model, since they do not lead to the expression of a small genome copy number of episomal HPV DNA in basal stem cell keratinocytes. Our overarching hypothesis is that HPV early proteins (E5, E6, and E7) dampen the cGAS/STING response to promote persistence and maintenance of episomal viral DNA (vDNA). Here, we show initial results by Dr. Uhlorn that suggest, suggest HPV18+ HFKs (compared to the parental HPV- cell line) inhibit STING activation by dampening the production of cGAMP in response to DNA transfection, and that exogenous cGAMP treatment rescues STING/IRF3 activation. This data implied HPV inhibited cGAS, and so we focused our attention on elucidating the mechanism. We first showcase our attempts at studying the localization of cGAS, which a few previous studies indicate is dynamic between nucleus and cytosol, which can regulate its function: many studies suggest that nuclear cGAS is tethered to host chromatin, which restrains cGAMP production ~200 fold and prevents autoimmunity. Therefore, localization is critical to DNA sensing (3–5). We also observed through preliminary immunofluorescence and subcellular fractionation experiments that cGAS localization is highly nuclear (as shown by many studies in various cell lines), but not altered in an HPV-dependent manner. We then present our results from studying the post-translational control of cGAS which many studies have shown is the predominant mechanism of regulating cGAMP production. Every step of cGAS mechanism (binding DNA, oligomerizing via phase separation, and catalyzing cGAMP) has been shown to be regulated by multiple post-translation modifications (PTMs) (6). We show our attempts at studying the DNA binding ability of cGAS via transfection of biotinylated-DNA oligos and streptavidin bead pulldown, observing cGAS to be pulled down equivalently in HPV18-/+ cells, suggesting HPV18 does not inhibit DNA binding ability of cGAS. To directly identify the potential HPV-dependent binding partners/regulators of cGAS, we did immunoprecipitation/mass spectrometry (IP-MS). Our initial attempts at pulling down endogenous cGAS were unsuccessful, however, due to non-specific antibody issues. To overcome that issue, we shifted towards pulling down GFP-cGAS by using a commercial GFP-pull down kit. We first transduced our HPV18-/+ cell lines with lentivirus encoding GFP-cGAS. The IP was successful, and we had adequate peptide coverage (~50% of cGAS) for interactome analysis, identifying a few proteins that were upregulated in our HPV18+ pulldown such as known cGAS inhibitors: DNA-PKcs and CDK1 (kinases that are known to phosphorylate and inhibit cGAS’ DNA binding ability), as well as TRIM26 (part of a family of E3 ubiquitin ligases, some of which are known to promote ubiquitination and degradation of various proteins in the cGAS/STING pathway). Further biological replicates would be necessary, as well as validation of those target proteins’ effect on HPV-dependent inhibition of the pathway. While waiting for MS results, we also attempted to replicate the earlier findings of Dr. Uhlorn. Even though we similarly observed reduced STING phosphorylation, we were surprised to find that cGAMP production measured by ELISA with a different commercial kit was not inhibited by HPV18. The difference in results is likely because the cGAMP ELISA kit Dr. Uhlorn used has since been discontinued and so we had to use a different commercial kit. We also observed that STING levels were reduced in an HPV-dependent manner that occurred at baseline, but more so after DNA transfection, suggesting another possible HPV mechanism of the cGAS/STING pathway. This was also noticed by Isabelle Tobey in her studies of cGAS/STING pathway inhibition in HPV16+ human tonsil keratinocytes (HTKs) that have a mixed pool of episomal and integrated HPV genome copies. She observed that STING levels were reduced in a similar manner, and when water/DNA transfected HTKs were treated with proteasome and lysosomal inhibitors (MG135 and Bafilomycin A respectively), she observed that STING were recovered in HPV16+ HTKs, as well as phospho-STING in DNA transfected cells. Curiously, we observed that the recovery of STING protein levels led to phospho-STING occurring even without DNA transfection solely in HPV16+ HTKs. Similarly, we also noticed in our more recent ELISA repeats that cGAS-overexpressing HPV18+ cells uniquely produced low levels of cGAMP in our negative control water-transfected cells. These more recent results suggest that the presence and expression of the HPV genome alone is enough to produce DNA of unknown origin that leads to a cGAS/STING response when either cGAS or STING levels are increased, but at baseline levels of those proteins, HPV increases STING turnover to dampen phosphorylation of STING and inhibit the immune pathway. Lastly, we discuss our results, hypothesizing the origin of that possible cGAS-detectable DNA and that we must shift our focus to not just studying cGAS, but also STING regulation by HPV to properly elucidate the main HPV mechanisms of inhibiting cGAS/STING.