Glucocorticoid Receptor Dimers are Necessary but not Sufficient for Chromatin Occupancy and Activity

THOMAS A. JOHNSON; VILLE PAAKINAHO; GORDON L. HAGER; DIEGO M. PRESMAN

Bethesda

Simposio; Center for Cancer Research Fellows & Young Investigators Colloquium; 2018

Center for Cancer Research, NIH

Background and HypothesesSynthetic glucocorticoids (GCs) are one of the most prescribed pharmaceuticals world-wide due to their powerful anti-inflammatory and immunosuppressive activities. The action of GCs is mediated by the glucocorticoid receptor (GR), a member of the nuclear receptor superfamily of transcription factors. Historically, GR transcriptional activity and clinical outcomes have been linked to its dimeric/monomeric state. A widely discussed model suggests that dimeric GR regulates unfavorable metabolic pathways, while monomeric GR is responsible for anti-inflammatory activities. Hence, GR ligands that preferentially induce the monomeric rather than the dimeric pathway should retain the desired pharmacological effects, but will lack the undesired adverse reactions. However, the search for improved ligands under this paradigm has produced no significant results for the last 20 years. Consistently, recent studies have indicated that dimeric GR is equally active in anti-inflammatory activities. Furthermore, our recent findings show that GR can form tetramers on a response element during gene activation. These results suggest that higher oligomerization states are important for the receptor?s activity.Study Design and MethodsTo unveil the relationship between GR?s oligomerization state and transcriptional outcome, we removed the endogenous GR from our model cells using CRISPR-Cas9 technology. Subsequently, we reintroduced four different forms of GR into the cells; wild-type (GRwt), DNA-binding domain dimerization defective (GRdim, A465T), fully monomeric (GRmon, A465T/I634A), and tetrameric (GRtetra, P483R) receptor, which mimics the DNA-bound conformation and forms tetramers independent of DNA binding. We performed a genome-wide comparison of binding profiles for GR and the above mutants. We also characterized chromatin accessibility by ATAC-sec and assessed the transcriptional response of each mutant by RNA-seq.Results and ConclusionsWhen compared to GRwt, global gene expression mediated by GRdim is defective both in the number of GC-regulated genes and the relative hormone response. GRmon is essentially a non-functional receptor, as almost no genes respond to hormone. These effects were partially mirrored by the genome-wide binding analysis. While GRmon binds only a few sites, a significant number are occupied by both GRdim and GRwt, suggesting that GRdim activity is affected downstream of chromatin binding. In remarkable contrast, GRtetra regulates more genes upon glucocorticoid exposure than the GRwt, and binds thousands of new sites on chromatin, as measured by ChIP-seq. Interestingly, GRdim and GRmon binding are both significantly dependent on the pre-accessible chromatin status. In comparison, GRtetra binds hundreds of sites in seemingly closed chromatin regions. Our results indicate that GR must minimally form dimers for proper glucocorticoid mediated action, due to the non-functional nature of the monomeric GR. A DNA-independent tetrameric GR acts a super receptor by binding to response elements not accessible to wild-type receptor and induces or represses more genes. Relevance and importanceThese results argue a common mode of GR binding via higher order structures that drives both the activating and repressive actions of glucocorticoids, and suggests that the dimeric/monomeric dichotomy concept is outdated. This has important implications for the therapeutic uses of steroid hormones and the goal of finding selective anti-inflammatory drugs that do not create unwanted side-effects.