Regulatory DNA Elements Modulating Hypoxia-Inducible Erythropoietin Gene Expression

Abstract

producing Epo are the fetal liver and the adult kidney, but ectopic Epo expression can be found also in other organs, including the brain. In response to anemia or inspiratory hypoxia, Epo synthesis is transcriptionally induced by activation of the hypoxia-inducible factor (HIF) pathway, allowing the binding of HIF-2 to a hypoxia response element (HRE) within the EPO gene. The availability of liver-derived cell models as well as in vivo studies permitted the characterization of the Epo transcriptional regulation in the liver, leading to the finding that the 3' HRE is the regulatory element responsible for the induction of hepatic Epo expression in hypoxia. Reporter transgenic mouse models demonstrated that a region encompassing -14 to -6 kb upstream of the transcription start site (TSS) of EPO gene contains the renal regulatory enhancers involved in the regulation of renal Epo and, it has been termed kidney inducible element (KIE). Within the KIE, we previously identified by in silico analysis a novel 5' HRE located at -9.2 kb from the TSS of EPO. We showed that this 5' HRE is highly conserved among species and that drives exogenous reporter gene expression in hypoxia. Because the liver-specific 3' HRE is dispensable for renal Epo regulation and a renal cell culture model was missing until recently, we investigated the contribution of the novel 5' HRE and the well-established 3' HRE by endogenous gene editing (CRISPR/Cas9) of both HREs in Kelly as well as Hep3B and HepG2 cells, capable of producing Epo in a hypoxia-inducible manner. We showed that while in hepatic cells the 3' but not the 5' HRE is required, in neuronal cells both the 5' and the 3' HREs contribute to regulate hypoxia-inducible endogenous Epo expression. Moreover, we identified two not yet reported HREs within the EPO promoter, termed pHRE1 and pHRE2, in line with a weak but significant promoter-driven reporter gene induction. While in hepatoma cells HIF interacted mainly with the distal 3' HRE, in neuronal cells, HIF bound most strongly the EPO promoter, weaker the 3' HRE and not at all the 5' HRE, suggesting that the 5' HRE acts in a non-canonical way by recruiting other transcription factors rather than HIF. We additionally generated 5'/3' double mutant Kelly cells and could show that both distal 5' and 3' HREs functionally cooperate with HIF binding to promoter HREs. These findings provide new insights into brain Epo regulation and they might be of help to recapitulate renal Epo regulation as renal Epo-producing (REP) cells show neuronal features. Due to the lack of a renal cell culture model able to produce Epo in a hypoxia-inducible manner, the study of renal Epo transcriptional regulation has been constrained. Therefore, we generated a novel transgenic mouse model to isolate fresh primary REP cells and we established a new method to culture both, primary cells and immortalized clonal cell lines, which we named fibroblastoid atypical interstitial kidney (FAIK) cells. By two simultaneous stimuli, hypoxia and tamoxifen, we were able to temporally and conditionally tag the REP cells with red fluorescence due to permanent activation of tdTomato reporter gene and negatively select the CreERT2 expressing cells through inactivation of the diphtheria toxin receptor (DTR) as non-CreERT2 expressing cells died in presence of diphtheria toxin. This mouse model is, thus, useful to generate primary REP culture and clonal REP cell lines that can be adopted as model to study hypoxic renal Epo regulation. We also applied a novel chromogenic in situ hybridization method to visualize on a single cell level Epo mRNA molecules as well as other HIF target genes in neuroblastoma, hepatoma and FAIK cells. We provided initial evidence for an inherent cell-to-cell variability of the Epo hypoxic transcript pattern not observed with constitutively expressed transcripts. This PhD work focussed on the transcriptional regulation of erythropoietin and on the DNA regulatory elements enhancing Epo expression in neuroblastoma and hepatoma cell lines. We could show that, in contrast to the liver, in the brain, Epo regulation occurs through a complex interplay between the distal 5' and 3' HREs and Epo promoter, demonstrating that several HREs can cooperate in oxygen-regulated gene expression in a cell-type specific manner.