Glucocorticoids inhibit osteocalcin transcription in osteoblasts by suppressing Egr2/Krox20-binding enhancer.

OBJECTIVE: Glucocorticoids are widely used for the management of rheumatoid arthritis. Osteoporosis is a major side effect of glucocorticoid therapy and is attributable to inhibition of bone formation. We developed an osteoblast culture system in which glucocorticoids strongly inhibit development of the osteoblast phenotype, including expression of the bone-specific osteocalcin (OC) gene. Using this gene as a model, the goal of this study was to discover glucocorticoid-sensitive transcriptional mechanisms in osteoblasts.
METHODS: Dexamethasone (DEX; 1 microM) was administered to murine MC3T3-E1 osteoblastic cultures under conditions that inhibit mineralized extracellular matrix formation and OC messenger RNA levels by >10-fold. Because standard (short-term) transient transfection assays with OC promoter-reporter constructs did not recapitulate the strong DEX-mediated repression, mapping of OC negative glucocorticoid response elements (GREs) was performed initially by stable transfection and then with long-term transient transfection assays. Transcription factor binding to the OC negative GRE was studied by electrophoretic mobility shift assays.
RESULTS: Several-fold repression of OC-luciferase constructs was recapitulated in stable and long-term transient transfection assays, in which the transfected cells were allowed to progress to a sufficiently advanced developmental stage. Analysis of a 5' promoter deletion series mapped an OC negative GRE to a 15-bp G/C-rich motif (-161/-147) located just upstream of the binding site for the osteoblast master transcription factor Runx2. Oligonucleotides encompassing this element and MC3T3-E1 cell extracts formed a protein-DNA complex that contained an Egr/Krox family member(s). Complex formation was competed by either an oligonucleotide containing 2 consensus Egr motifs or by anti-Egr2/Krox20 antibodies. Three copies of this Krox-binding element conferred 20-fold transcriptional activation on the 147-bp basal OC promoter in osteoblasts, and the enhancer activity was inhibited by DEX. Enhancer activity was not observed in 10T1/2 fibroblasts unless these cells were cotransfected with Runx2.
CONCLUSION: An Egr2/Krox20-binding site located immediately upstream of the Runx2 site of the mouse OC promoter was identified as an enhancer in osteoblasts, whose activity is repressed by glucocorticoids. Sequence similarity suggests that such a mechanism is likely operative in both murine and human cells. Because glucocorticoids inhibit Egr2/Krox20 expression in osteoblasts, and because trabecular bone formation is arrested in Egr2/Krox20-knockout mice, the inhibition of Egr2/Krox20 activity likely contributes to glucocorticoid-induced osteoporosis.