Steroid Receptor Associated Proteins: Heat Shock Protein 90 and P59 Immunophilin

Abstract

Nearly ten years ago, hoping to obtain a monoclonal antibody against “native” 8S chick progesterone receptor prepared in the now classical way in the presence of molybdate, we obtained an antibody which displaced progesterone binding in sucrose gradients (Joab et al, 1984). It also displaced estrogen, androgen and glucocorticosteroid binding, and we now know that it can also displace mineralocorticosteroid receptors (Rafestin-Oblin,1989). In low salt gradients, where the 8S complex was conserved, the peak of bound steroid was shifted by the BF4 antibody, but in high salt, the complex dissociates and the 4S peak was not shifted, indicating that BF4 interacted not with the hormone-binding receptor molecule, but with another protein, which was part of the 8S, but not of the 4S peak. We had discovered a new receptor-associated protein, which turned out to be an abundant, cytosol, non-steroid-binding protein of 90 kDa. We soon found out that it was a well known heat shock protein (hsp 90) (Catelli et al, 1985), which could interact with other intracellular proteins such as oncogenic protein kinases (reviewed in Brugge, 1986, Zimiecki et al, 1986). Chick hsp 90 was cloned using BF4 antibody and then sequenced. Among other features it has a negatively charged region, which was called “zone A” (which is predicted to contain two α helices with a proline containing loop in the middle, and where the negative charges could be aligned by the computer on the phosphate backbone of the B-DNA helix)(Binart et al, 1989). At about the same time, the structure of steroid receptors had been described (Hollenberg et al, 1985, Walter et al, 1985, Gronemeyer et al, 1987), and we hypothesized that zone A could be engaged in an ionic interaction with the DNA binding domain (DBD) of steroid receptors, and that somehow hsp 90 could cap the DBD and maintain the receptor in an inactive conformation (Baulieu, 1987).