Use of gene expression profiling to identify a novel glucocorticoid sensitivity determining gene, BMPRII.

Wide variation in glucocorticoid (Gc) sensitivity exists between individuals which may influence susceptibility to, and treatment response of, inflammatory diseases. To determine a genetic fingerprint of Gc sensitivity 100 healthy human volunteers were polarized into the 10% most Gc-sensitive and 10% most Gc-resistant following a low dose dexamethasone (0.25 mg) suppression test. Gene expression profiling of primary lymphocytes identified the 98 most significantly Gc regulated genes. These genes were used to build a subnetwork of Gc signaling, with 54 genes mapping as nodes, and 6 non-Gc regulated genes inferred as signaling nodes. Twenty four of the 98 genes showed a difference in Gc response in vitro dependent on the Gc sensitivity of their donor individuals in vivo. A predictive model was built using both partial least squares discriminate analysis and support vector machines that predicted donor glucocorticoid sensitivity with 87% accuracy. Discriminating genes included bone morphogenetic protein receptor, type II (BMPRII). Transfection studies showed that BMPRII modulated Gc action. These studies reveal a broad base of gene expression that predicts Gc sensitivity and determine a Gc signaling network in human primary T lymphocytes. Furthermore, this combined gene profiling, and functional analysis approach has identified BMPRII as a modulator of Gc signaling.