We have mined datasets from microarray analyses of gene expression from large numbers of cell lines of particular human cancer types to determine relationships between levels of particular transcription factors and their impact on target gene expression. Using a set of defined conditions and by excluding confounder data points, this enabled us to then clearly observe significant correlations existing between sets of transcription factors and their target genes (1). These results were consistently confirmed and validated using other independent methods for analysis of transcription factor binding, including chromosomal immunoprecipitation, RT-PCR, 5’-RACE, gene reporter, knockdown of gene expression or gel shift assays. We have applied our microarray analysis of several genes to provide novel links not previously recognised including some genes involved in, or regulated by the interferon response. For example, IFNg activation of STAT5 signaling was shown to bind and activate MYC gene expression in human colon cancer. We propose that our approaches may be more widely applicable, as a general method for refining and determining novel relationships existing between different transcription factors and the genes that they regulate.

1) REST negatively and ISGF3 positively regulate the human STAT1 gene in melanoma. Amalraj J, Cutler SJ, Ghazawi I, Boyle GM, Ralph SJ. Mol Cancer Ther. 2013 Jul;12(7):1288-98.