Control by basal phosphorylation of cell cycle-dependent, hormone-induced glucocorticoid receptor hyperphosphorylation.

Mouse glucocorticoid receptors (GRs) are phosphorylated in the N-terminal domain at serine/ threonine residues, most lying in consensus sequences for cell cycle-associated kinases. Glucocorticoid agonists, but not antagonists, induce hyperphosphorylation. Phosphorylation of GRs overexpressed in Chinese hamster ovary (CHO) cells is cell cycle-dependent: basal phosphorylation in S phase is one third that in G2/M; glucocorticoids induce hyperphosphorylation in S but not G2/M, paralleling the reported sensitivity in S and resistance in G2/M of proliferating cells to transcriptional activation by glucocorticoids. This parallel led us to investigate what controls hyperphosphorylation. We tested three hypotheses: hyperphosphorylation is controlled by 1) negative charge due to basal GR phosphorylation, being permitted in S by low charge and blocked in G2/M by high charge; 2) presence in S and absence in G2/M of required kinases; 3) availability in S and lack in G2/M of unoccupied phosphorylatable sites. Our results are inconsistent with 2) and 3), but strongly support 1). GR mutants with alanines (A7GR) or glutamates (E7GR) replacing all but one phosphorylated site were overexpressed in CHO cells. Serine 122 remained intact to report GR phosphorylation. Consistent with hypothesis 1, with A7GRs hormone-induced hyperphosphorylation occurred in both S and G2/M (thus revealing kinase activity for hyperphosphorylation of at least serine 122 in both phases), whereas with E7GRs it occurred in neither phase. We conclude that basal GR phosphorylation controls hormone-induced GR hyperphosphorylation by modulating negative charge in the N-terminal domain and could potentially control other cell cycle-dependent GR properties.