Search

To elucidate the earliest molecular steps in the activation of transcription by the progesterone receptor (PR), we investigated its activity in a cell-free transcription system utilizing chromatin templates. PR prepared as a ligand-free, recombinant protein failed to induce transcription on chromatin templates. However, transcriptional competence could be restored by coincubation with rabbit reticulocyte lysate (RRL). The interaction of PR with chaperones results in a receptor conformation competent to bind ligand and RRL contains abundant chaperone-mediated protein folding activity. Blocking this activity with the specific inhibitor geldanamycin inhibited receptor-dependent transcriptional activity. However, recombinant chaperones could not replace RRL in the restoration of transcriptional activity on chromatin templates, suggesting the presence of an additional activity in the lysate. Under chromatin assembly conditions, PR could bind naked DNA and RRL did not increase that binding. In contrast, PR bound to a chromatin template only poorly. Interestingly, RRL stimulated sequence-specific binding by PR to target sites in chromatin and the concomitant recruitment of the steroid receptor coactivator 1 to the promoter. Thus, our results indicate that a novel protein-mediated activity in RRL is involved in an additional, heretofore unrecognized, activation step required for PR to become transcriptionally competent on chromatin templates.

T47D human breast carcinoma cells and the chicken oviduct were used to study the structure of the nonactivated progesterone receptor (PR) complex. Immunoprecipitation of PR (B form) from cytosol extracts was performed using monoclonal antibody PR6, a cross-reactive antibody prepared to chicken PR. Analysis of the PR complex by sodium dodecyl sulfate gels and Western immuno-blotting revealed the presence of several specific copurifying proteins. Consistent with previous reports, the two heat shock proteins, hsp90 and hsp70, were shown to be present. A third 59-kilodalton (kDa) protein observed previously was confirmed to be p59 (also called hsp56 or FKBP52), which has been shown to bind the immunosuppressant drug FK506. Two additional PR-associated proteins were observed that had not been previously recognized with human PR. These have molecular masses of 54-kDa and 23-kDa and have been shown by Western blotting to be related to the proteins p54 and p23 that are associated with chicken PR. P23 is a novel protein of unknown function and p54 or FKBP54 has been recently shown to be another FK506-binding protein related to p59. Finally, the cyclosporin A-binding protein, CyP-40, could be detected in isolated chicken PR complexes and in PR complexes that were reconstituted in vitro, but this protein was not detected in human PR complexes, which are less stable than chicken PR complexes in cytosol extracts. The functional significance of FK506 and cyclosporin A-binding proteins to hormone action was tested using a T47D cell line that contained a progestin reporter gene, MMTV-CAT. Treatment with cyclosporin A had no effect on the basal level of CAT expression, but it caused a dramatic increase in the sensitivity and magnitude of the response to the synthetic progestin, R5020. The enhanced response elicited by drug treatment was blocked by the antiprogestin RU486 indicating that this effect was receptor-mediated. While cyclosporin A enhanced progestin action in T47D cells, it inhibited a PR/reporter gene system in L cells. The drugs FK506 and rapamycin had no effect on progestin action in T47D cells, but they stimulated glucocorticoid action in T47D cells. Thus, the effects of these immunosuppressant drugs vary with the cell type and hormonal system that is tested. Whether these drug effects relate directly to the immunophilins bound in receptor complexes remains unknown.

We have cloned and sequenced 4.5 kilobases (Kb) of cDNA encoding the chicken progesterone receptor. The complete cDNA contains an open reading frame of 2361 nucleotides in length and encodes a polypeptide of 787 amino acids with a mol wt of 85.9 K. At least four mRNA species have been detected in chick oviduct cells. Direct sequencing of variant cDNAs has suggested that two of the mRNAs (4.5 Kb and 3.6 Kb) differ only in the length of their 3'-untranslated regions. A third mRNA (1.8 Kb) produces a truncated polypeptide which encodes the immunoreactive NH2 terminal sequence of the receptor but lacks the hormone binding regional and half of the DNA-binding domain. The polypeptide expressed from the receptor cDNA in progesterone receptor negative Cos M-6 cells is indistinguishable from oviduct progesterone receptor in terms of hormone binding and antibody reactivity. Furthermore, the cloned receptor is capable of activating transcription of a target gene. This activation is progesterone dependent (with half-maximal stimulation at approximately 3.3 x 10(-10) M) and specific for the target gene.

The chicken progesterone receptor (PR) cDNA has been cloned and sequenced in our laboratory. Functional receptor A was synthesized from cDNA in two independent systems, by transient transfection of receptor-negative COS M6 cells and by in vitro transcription and translation. These receptors exhibited DNA and hormone binding properties similar to the native PR from oviduct. The ability of receptor to induce target gene transcription was measured by cotransfection of receptor-negative CV-1 cells with expression vectors containing the receptor A cDNA and a progesterone-inducible promotor linked to the chloramphenicol acetyl transferase (CAT) gene. In these assays, receptor A produced hormone-dependent induction of CAT activity. In order to define the functional domains of receptor A, expression constructs coding for C-terminal deletion proteins were prepared. Deletion of the C-terminus resulted in loss of hormone binding activity as well as a loss of CAT induction. However, when 290 amino acids were removed from the C-terminus, this severely truncated receptor protein produced hormone-independent target gene activation. Mutant receptor proteins which retained the highly conserved cysteine-rich (C1) region were able to bind to DNA-cellulose, although removal of 290 amino acids from the C-terminus resulted in reduced affinity for DNA. Deletion of part or all of the C1 region resulted in loss of both DNA-binding and transcriptional activation capacities. These results confirm that C1 functions in DNA binding and transcriptional activation and that hormone binding activity can be localized to the C-terminal half of the protein.