Endocrine Abstracts

Glucocorticoids (Gc) act on most cells and tissues, but their effects vary in a cell-type specific manner. A number of genes show divergent responses to glucocorticoid action, dependent on their tissue of origin. One such gene is MIF, a proinflammatory cytokine associated with Gc resistance in inflammation, and linked with susceptibility to chronic, human inflammatory disease. Importantly, MIF gene expression is inhibited by Gc in lymphoblasts, but not in epithelial cells; a finding with implications for understanding the effects of Gc in inflammation. Multiple truncated MIF promoter constructs from −775 to −71 all retained Gc inhibition in lymphoblast cells CEMC7A, and were not affected by Gc exposure in the A549 cell line, despite both cell lines expressing glucocorticoid receptor (GR), and being capable of Gc transactivation of simple reporter genes. Analysis of MIF −71 to +85 revealed no potential GR binding sites; however, three putative AP-1 binding sites were found, one upstream (MIF1) and two downstream (MIF2 and MIF3) of the transcription start site. Since activated GR can suppress gene transcription by interacting with DNA bound AP-1, these sites were deleted from the minimal (−71 to +85) MIF reporter construct, individually, and in combination. Single site deletions did not prevent Gc repression of promoter activity. However, Gc suppression was completely abolished when MIF1 and MIF3 were deleted.

EMSA and antibody supershifts identified a complex composed of ATF-1/CREB binding MIF1, in nuclear extracts from both cell lines. In contrast, MIF3 binds a distinct complex present only in CEMC7A cells, and is not AP-1 or CREB-like. Crucially, we show time dependent recruitment of activated GR to the MIF1/MIF3 composite element in lymphoblasts, suggesting a tethering mechanism for Gc action at the MIF gene. This suggests that the unique, lymphoblast complex binding MIF3, interacting with the ubiquitous MIF1 complex, is essential for cell-type specific modulation of Gc response, either by altering GR binding or the subsequent effects on RNA PolII activation. These findings have further implications for understanding Gc action in inflammation.