Impact of transcriptional and posttranscriptional regulation of HNF4A and its target genes on diabetes and cancer

HNF4A is a susceptibility gene for diabetes and is considered a tumor suppressor in certain cancers including RCC. Although several different HNF4A mutations have been linked to diabetes, in the majority of cases including RCC the reason for HNF4A dysregulation is unknown. In this study, regulation of proliferation relevant target genes of this transcription factor as well as transcriptional and posttranscriptional mechanisms that regulate the expression of HNF4A itself were investigated with the intention to illuminate how disruption of those processes could impact on diabetes and RCC. An inducible HNF4A8 expression cell line was established, which in contrast to HNF4A2 has no impact on cell proliferation decrease and morphology. To deduce proliferation relevant genes from the 1411 potential HNF4A2 target genes identified previously, an HNF4A8 dependent microarray was performed. 111 from only 191 HNF4A8 target genes deemed to be irrelevant for proliferation control were also controlled by HNF4A2 and excluded, leaving 1300 potential HNF4A2 target genes. qRT-PCR validated that the apoptosis and metabolism gene CIDEB is highly upregulated by HNF4A2 in contrast to HNF4A8 and HNF4A mutants. The impact of CIDEB on proliferation control was reasoned to be dependent on a network triggered by HNF4A2 as determined by RNAi and Cideb inducible cell lines. The two novel mutations -136A>G and -169C>T identified in the P2 promoter of patients with symptoms of HNF4A monogenic β-cell diabetes together with a previously reported -192C>G promoter mutation linked to late-onset diabetes in several families, were shown to impair the function of the HNF4A P2 promoter in vitro using a luciferase reporter assay system. Furthermore, evidence of miRNAs enhancing gene expression by targeting the HNF4A P1 promoter was provided by Dicer dependent luciferase reporter assays. To elucidate the so far unrecognized posttranscriptional regulation of HNF4A, the predicted 1.7 kb HNF4A 3’UTR was validated and an additional 3.2 kb long, predominantly used 3’UTR was identified. Both 3’UTRs conferred a repressive effect in HEK293 and INS-1 cells, which was even more pronounced in two distinct, previously unknown elements of about 400 nt located within the 3’UTR as determined by luciferase assays. These negative elements A and B were counteracted by an element located within the 5’ 630 nt. Dicer knock-down reporter assays inferred negative regulation of the 3’UTRs by miRNAs. More detailed overexpression experiments of selected miRNAs upregulated in RCC revealed a modest effect of miR-21 dependent on several sites within the HNF4A 3’UTR. miR-34a negatively regulated HNF4A by targeting at least two sites located at the 5’ end of both 3’UTRs. In conclusion, dysfunction of transcriptional and posttranscriptional regulation of the two promoters and 3’UTRs of HNF4A, respectively, mediated by proteins and miRNAs, alters the HNF4A dependent network cascade and likely contributes to the development and/or progression of diabetes and RCC