Discovered in the 1930s in adrenal gland extracts, glucoof the attention until that time has focused on positively corticoid hormones (GCs) were amongst the first steroid regulated genes. However, in the late 1980s several hormones to be found. The study of GC action provided groups began to analyze genes that are negatively regumany important and critical insights to the mechanisms lated by GC. The first genes subjected to such scrutiny used by nuclear receptors and other sequence-specific included those coding for proopiomelanocortin (POMC), transcription factors to regulate gene transcription. Durthe α subunit of glycoprotein hormones, and collagening the 1970s it was realized that GCs exert their biologi- ase type I. These genes represent important physiological effects through a specific receptor, GR, a DNA-bindcal targets for GC. ing protein whose nuclear translocation is induced upon In the case of theα subunit gene GC-mediated represligand binding (Yamamoto, 1985; Beato et al., 1995). sion is caused by interference with the cAMP-respon-GCs were found to induce transcription of target genes, sive transcription factor CREB (Akerblom et al., 1988), the most immediate being those whose induction is in- while repression of collagenase transcription was attribsensitive to inhibitors of protein synthesis. Cloning and uted to GR-mediated interference with transcription analysis of such genes, the murine mammary tumor factor AP-1 (Jonat et al., 1990; Yang-Yen et al., 1990). virus (MMTV) genome and the human metallothionein AP-1 activity is stimulated by many agonists of which IIA (hMTIIA) gene, led to identification of the first hormone proinflammatory cytokines are most relevant in this conresponse elements (HREs), in this case called GREs, text. Negative regulation of the POMC gene, on the other which serve as inducible enhancer elements and GR- hand, is exerted through direct interaction of GR with binding sites (reviewed by Yamamoto, 1985; Beato et its promoter, which contains a negative GRE (nGRE; al., 1995). These findings obtained in the early 1980s Drouin et al., 1993). Whereas the exact mechanism by forged the notion that GR is a sequence-specific ligand- which GR interferes with CREB or AP-1 activity is yet regulated transcription factor, one of the first eukaryotic to be elucidated, it is almost certain that it does not transcription factors to be identified. This model was require direct binding to DNA. Mutations within the DBD considered proven when the GR was molecularly cloned or substitutions of the GR DBD with DBDs of other tranand the mechanism by which it activates transcription scription factors do not abolish GR’s ability to interfere ofGRE-containinggenes wasstudied indetail (reviewed with AP-1 (Jonat et al., 1990; Yang-Yen et al., 1990). by Beato et al., 1995). It was even considered that all Importantly, a dimerization-defective GR, generated by of the physiological effects of GC are mediated through a mutation within the D loop (A458T), does not bind gene induction (Yamamoto, 1985). Results published in cooperatively to GREs but can still repress AP-1 reguthis issue of Cell (Reichardt et al., 1998) question lated genes (Heck et al., 1994). whether DNA binding is required at all for many of the Another transcription factor whose activity is negaphysiological functions of GR. tively regulated by GC is NF-κB. While transient trans-Distinct Modes of Transcriptional Control fection experiments suggest that this case of negative The GR, a prototypical nuclear receptor, is composed regulation is also based on direct interaction between of three functional domains: a constitutive N-terminal GR and the affected transcription factor …