Epigenetic Regulation of S100A4/Metastasin

Epigenetic modulation of gene expression is an important mediator of cancer progression. It is well understood that tumor suppressors can be inactivated by the methylation of their promoters, which can contribute to tumor progression. However, select promoters of pro-invasive and pro-metastatic genes can be selectively demethylated, and thus activated, during tumor progression. S100A4 is one such gene that contributes to tumor metastasis in several cancer types and its expression is tightly controlled by promoter methylation. We have found that the integrin á6â4 can control S100A4 expression in breast cancer by stimulating the demethylation of the S100A4 promoter. In this proposal, we provide evidence that this demethylation status is dynamic and is controlled by DICER generated small non-coding RNAs (sncRNAs). We provide evidence to support the hypothesis that these sncRNAs are generated by double stranded RNAs mediated by sense and antisense transcripts from a given promoter. In the case of the S100A4 promoter, the region that governs the methylation status of the promoter is located in a non-coding alternatively spliced exon. Importantly, integrin á6â4 signaling can stimulate the expression of this antisense transcript. Accordingly, the central hypothesis of this application is integrin á6â4 stimulates S100A4 promoter demethylation through non-coding RNA-directed modification of methylated cytosines and subsequent DNA mismatch repair. We will test our central hypothesis and achieve our long-term goal through completion of the following two aims: 1) Define how integrin á6â4 controls DNA demethylation through to generation of antisense RNAs and sncRNAs, 2) Determine the involvement of DNA repair in the active DNA demethylation of the S100A4 promoter, and 3) Define changes in DNA methylation cause by integrin á6â4 signaling on a genomic scale. Understanding how epigenetic modulation of an invasive and metastatic phenotype is controlled will give us the power to target it therapeutically. If our model is proven largely correct, it will represent a new paradigm of how integrins, non-coding RNAs and RNA processing transmit oncogenic signals.