DNA Demethylation Switches Oncogenic ΔNp63 to Tumor Suppressive TAp63 in Squamous Cell Carcinoma

Abstract

The TP63 gene encodes two major protein variants; TAp63 contains a p53-like transcription domain and consequently has tumor suppressor activities whereas ΔNp63 lacks this domain and acts as an oncogene. The two variants show distinct expression patterns in normal tissues and tumors, with lymphocytes and lymphomas/leukemias expressing TAp63, and basal epithelial cells and some carcinomas expressing high levels of ΔNp63, most notably squamous cell carcinomas (SCC). Whilst the transcriptional functions of TAp63 and ΔNp63 isoforms are known, the mechanisms involved in their regulation are poorly understood. Using squamous epithelial cells that contain high levels of ΔNp63 and low/undetectable TAp63, the DNA demethylating agent decitabine (5-aza-2’-deoxycytidine, 5-dAza) caused a dose-dependent increase in TAp63, with a simultaneous reduction in ΔNp63, indicating DNA methylation-dependent regulation at the isoform-specific promoters. The basal cytokeratin KRT5, a direct ΔNp63 transcriptional target, was also reduced, confirming functional alteration of p63 activity after DNA demethylation. We also showed high level methylation of three CpG sites in the TAP63 promoter in these cells, which was reduced by decitabine. DNMT1 depletion using inducible shRNAs partially replicated these effects, including an increase in the ratio of TAP63:ΔNP63 mRNAs, a reduction in ΔNp63 protein and reduced KRT5 mRNA levels. Finally, high DNA methylation levels were found at the TAP63 promoter in clinical SCC samples and matched normal tissues. We conclude that DNA methylation at the TAP63 promoter normally silences transcription in squamous epithelial cells, indicating DNA methylation as a therapeutic approach to induce this tumor suppressor in cancer. That decitabine simultaneously reduced the oncogenic activity of ΔNp63 provides a “double whammy” for SCC and other p63-positive carcinomas. Whilst a variety of mechanisms may be involved in producing the opposite effects of DNA demethylation on TAp63 and ΔNp63, we propose an “either or” mechanism in which TAP63 transcription physically interferes with the ability to initiate transcription from the downstream ΔNP63 promoter on the same DNA strand. This mechanism can explain the observed inverse expression of p63 isoforms in normal cells and cancer.