Oncogenic human papillomaviruses (HPV) are the causative agents of uterine cervical and an increasing portion of head and neck squamous cell carcinomas (HNSCC), but HNSCC is almost exclusively associated with HPV type 16. The oncogenic properties of HPV type 16 are largely attributed to two major HPV oncogenes, E6 and E7, that degrade p53 and retinoblastoma (RB) family members, respectively. Degradation of these tumor suppressors by E6 and E7 results in uncontrolled proliferation, diminished apoptosis and increased genomic instability that predisposes to malignant transformation. The crucial roles of E6/E7 in HPV-related carcinogenesis make them an attractive target for anti-cancer therapy, since methods for decreasing their expression or activities would restore p53 and RB activity in tumors driven by HPV. Our preliminary results indicate that treatment of HPV-positive (HPV+) HNSCC with the demethylating agent, 5-azacytidine (5-aza) at clinically relevant concentrations, resulted in remarkable downregulation of all HPV gene expression, including E6 and E7. 5-aza treatment restored p53 expression and activity in HPV+ head and neck cancer cells, which was partially responsible for the sensitivity of these cells to 5-aza. In addition to restoration of p53, 5-aza also was toxic to HPV+ HNSCC through creation of DNA double strand breaks (DSBs). Mechanistically, 5-aza-induced DNA DSBs in HPV+ HNSCC were dependent on transcription and replication and on overexpression of the cytidine deaminase, APOBEC3B (A3B). Experimental depletion of A3B inhibited 5-aza toxicity and diminished DSB formation, but also indicated that untreated HPV+ HNSCC depend on A3B for clonogenic growth. The observations that untreated HPV+ HNSCC dependent on A3B, but that A3B contributes to 5-aza toxicity and DSBs, suggests an A3B-dependent synthetic lethality upon treatment with 5-aza, and that A3B may serve as a biomarker of response. Treatment of mice bearing HPV+ tumors with 5-aza revealed significant tumor growth inhibition and prevented detection of circulating tumor cells. A window clinical trial in patients with HNSCC using standard dosing for 5 days achieved demethylation (LINE-1) similar to that seen in our in vitro experiments and confirmed that 5-aza treatment: 1) significantly decreased expression of HPV genes; 2) reactivated p53; 3) activated caspases, 4) and inhibited matrix metalloproteinase expression in HPV+ HNSCCs. This proposal is designed to determine molecular mechanisms of demethylation-induced downregulation of HPV oncogenes, elucidate the role of A3B in 5-aza-induced synthetic lethality and DNA DSBs formation, determine effect of demethylation on immune cell infiltration in HPV+ HNSCC, and explore the potential of 5-aza alone or in combination with chemotherapeutic agents to suppress HPV-associated HNSCC metastasis and inhibit growth using patient-derived xenografts. These studies will provide a basis for a new rational targeted therapy for HPV+ HNSCC, which is desperately needed to treat patients with recurrent or metastatic HPV+ HNSCC and to decrease the toxicity associated with current therapy.
. Here, we propose to explore demethylation as a potential new therapy for HPV-associated head and neck cancer by determining mechanisms of its toxicity that are specific to HPV-positive tumors. Demethylation of HPV-associated head and neck cancers reactivates tumor suppressors and causes DNA damage through mechanisms that are not common to currently used therapies. In addition to providing improved understanding of activity, we will determine if FDA approved demethylating drug is effective against HPV-positive patient- derived xenografts as a single agent and in combination with chemotherapeutic drugs, providing a basis for rational targeted therapy as an alternative to standard cytotoxic treatment for primary, metastatic, or recurrent HPV-positive head and neck cancer.