Immunologic analysis of human breast cancer progesterone receptors. 2. Structure, phosporylation, and processing

Abstract

We have used a monoclonal antibody (MAb) directed against chick oviduct progesterone receptors (PR), that cross-reacts with human PR, to analyze PR structure and phosphorylation. This MAb, designated PR-6, interacts only with B receptors (Mr 120,000) of T47D human breast cancer cells; it has no affinity for A receptors (Mr 94,000) or for proteolytic fragments from either protein. The antibody immunoprecipitates native B receptors and was used to study the structure of native untransformed 8S and transformed 4S receptors, using sucrose density gradient analysis, photoaffinity labeling, and gel electrophoresis. On molybdate-containing low-salt gradients, PR-6 complexes with 8S B receptors, causing their shift to the bottom of the gradient while A receptors remain at 8 S. Therefore, A and B receptors from separate 8S complexes, and we conclude that A and B do not dimerize in the holoreceptor. Similar gradient studies using salt-containing, molydbate-free buffers show that there are two forms of salt-transformed 4S receptors, comprising either A proteins or B proteins, suggesting that A and B are also not linked to one another in transformed PR. The independence of A- and B-receptor complexes was confirmed by the finding that purified, transformed B receptors bind well to DNA-cellulose. Since PR-6 cross-reacts with nuclear PR, it was used to analyze nuclear PR processing.sbd.a down-regulation step associated with receptor loss as measured by hormone binding. Insoluble nuclear receptors and soluble cytosol receptors were measured by immunoblotting following treatment of T47D cells for 5 min to 48 h with either R5020 or progesterone. From 8 to 48 h after R5020 treatment, immunoassayable receptors decreased in nuclei and were not recovered in cytosols. Nuclear receptors also decreased after progesterone treatment but replenished in cytosols between 8 and 24 h after the start of treatment. Thus, processing involves a true loss of nuclear receptor protein, and not just loss of hormone binding activity, and occurs after progesterone or R5020 treatment. This loss is chronic, however, only in R5020-treated cells. Additional studies focused on the covalent modifications of receptors. We previously described shifts in apparent molecular weight of nuclear PR following R5020 treatment using in situ photoaffinity labeling. To show whether these shifts can be explained by receptor phosphorylation, untreated cells and hormone-treated cells were metabolically labeled with [32P]orthophosphate, and the B receptors were isolated by immunoprecipitation with PR-6 and analyzed by sodium dodecyl sulfate (SDS) gel electrophoresis. In both treatment states, B receptors were labeled in vivo with 32P, thus demonstrating directly that human PR are phosphoproteins. Since B receptors were labeled in the absence of hormone and also after their in vivo transformation by hormone, they appear to be substances for two phosphorylation reactions, one in the untransformed state and another after they are tightly bound to chromatin. The second phosphorylation may account for the mobility shift of the receptors on SDS gels. On the basis of these data and those in the preceding paper [Estes, P.A., Suba, E. J., Lawler-Heavner, J., Elashry-Stowers, D., Wei, L. L., Toft, D. O., Sullivan, W. P., Horwitz, K. B., and Edwards, D. P. (1987) Biochemistry (preceding paper in this issue)], a model of human PR structure and subcellular receptor dynamics is presented.