We previously reported that the chromatin high-mobility group protein 1 (HMG-1) enhances the sequencespecific DNA binding activity of progesterone receptor (PR) in vitro, thus providing the first evidence that HMG-1 may have a coregulatory role in steroid receptor-mediated gene transcription. Here we show that HMG-1 and the highly related HMG-2 stimulate DNA binding by other steroid receptors, including estrogen, androgen, and glucocorticoid receptors, but have no effect on DNA binding by several nonsteroid nuclear receptors, including retinoid acid receptor (RAR), retinoic X receptor (RXR), and vitamin D receptor (VDR). As highly purified recombinant full-length proteins, all steroid receptors tested exhibited weak binding affinity for their optimal palindromic hormone response elements (HREs), and the addition of purified HMG-1 or -2 substantially increased their affinity for HREs. Purified RAR, RXR, and VDR also exhibited little to no detectable binding to their cognate direct repeat HREs but, in contrast to results with steroid receptors, the addition of HMG-1 or HMG-2 had no stimulatory effect. Instead, the addition of purified RXR enhanced RAR and VDR DNA binding through a heterodimerization mechanism and HMG-1 or HMG-2 had no further effect on DNA binding by RXR-RAR or RXR-VDR heterodimers. HMG-1 and HMG-2 (HMG-1/-2) themselves do not bind to progesterone response elements, but in the presence of PR they were detected as part of an HMG-PR-DNA ternary complex. HMG-1/-2 can also interact transiently in vitro with PR in the absence of DNA; however, no direct protein interaction was detected with VDR. These results, taken together with the fact that PR can bend its target DNA and that HMG-1/-2 are non-sequence-specific DNA binding proteins that recognize DNA structure, suggest that HMG-1/-2 are recruited to the PR-DNA complex by the combined effect of transient protein interaction and DNA bending. In transient-transfection assays, coexpression of HMG-1 or HMG-2 increased PR-mediated transcription in mammalian cells by as much as 7-to 10-fold without altering the basal promoter activity of target reporter genes. This increase in PR-mediated gene activation by coexpression of HMG-1/-2 was observed in different cell types and with different target promoters, suggesting a generality to the functional interaction between HMG-1/-2 and PR in vivo. Cotransfection of HMG-1 also increased reporter gene activation mediated by other steroid receptors, including glucocorticoid and androgen receptors, but it had a minimal influence on VDR-dependent transcription in vivo. These results support the conclusion that HMG-1/-2 are coregulatory proteins that increase the DNA binding and transcriptional activity of the steroid hormone class of receptors but that do not functionally interact with certain nonsteroid classes of nuclear receptors.Steroid hormone receptors are members of a superfamily of ligand-dependent transcriptional activators which direct the expression of specific gene networks involved in regulating the differen...

Steroid hormone receptors are ligand-dependent transcriptional activators that exert their effects by binding as dimers to cis-acting DNA sequences termed hormone response elements. When human progesterone receptor (PR), expressed as a full-length protein in a baculovirus system, was purified to homogeneity, it retained its ability to bind hormonal ligand and to dimerize but exhibited a dramatic loss in DNA binding activity for specific progesterone response elements (PREs). Addition of nuclear extracts from several cellular sources restored DNA binding activity, suggesting that PR requires a ubiquitous accessory protein for efficient interaction with specific DNA sequences. Here we have demonstrated that the high-mobility-group chromatin protein HMG-1, as a highly purified protein, dramatically enhanced binding of purified PR to PREs in gel mobility shift assays. This effect appeared to be highly selective for HMG-1, since a number of other nonspecific proteins failed to enhance PRE binding. Moreover, HMG-1 was effective when added in stoichiometric amounts with receptor, and it was capable of enhancing the DNA binding of both the A and B amino-terminal variants of PR. The presence of HMG-1 measurably increased the binding affinity of purified PR by 10-fold when a synthetic palindromic PRE was the target DNA. The increase in binding affinity for a partial palindromic PRE present in natural target genes was greater than 10-fold. Coimmunoprecipitation assays using anti-PR or anti-HMG-1 antibodies demonstrated that both PR and HMG-1 are present in the enhanced complex with PRE. HMG-1 protein has two conserved DNA binding domains (A and B), which recognize DNA structure rather than specific sequences. The A-or B-box domain expressed and purified from Escherichia coli independently stimulated the binding of PR to PRE, and the B box was able to functionally substitute for HMG-1 in enhancing PR binding. DNA ligase-mediated ring closure assays demonstrated that both the A and B binding domains mediate DNA flexure. It was also demonstrated in competition binding studies that the intact HMG-1 protein binds to tightly curved covalently closed or relaxed DNA sequences in preference to the same sequence in linear form. The finding that enhanced PRE binding was intrinsic to the HMG-1 box, combined with the demonstration that HMG-1 or its DNA binding boxes can flex DNA, suggests that HMG-1 facilitates the binding of PR by inducing a structural change in the target DNA.

We have prepared a monoclonal antibody, C-262, to a synthetic peptide that contains the carboxy-terminal 14 amino acids from progesterone receptors (PR). This sequence is 100% conserved in all species of PRs that have been cloned to date, suggesting that this antibody will recognize all mammalian and avian PR. The C-262 antibody recognizes both native and denatured forms of the receptor. However, it does not recognize PR when they are bound to the hormone agonists progesterone or R5020. Surprisingly the antibody does recognize PR when they are bound to the steroid antagonist RU486. This suggests that progestin agonists induce a conformational change in the receptor that occludes the C-262 epitope in the carboxyl-terminus, whereas unliganded receptors and receptors bound with RU486 assume distinct conformations that leaves the C-terminal tail accessible to the C-262 antibody.

No abstract

We have used circular permutation and phasing electrophoretic mobility shift assays to determine the ability of the A and B forms of human progesterone receptor (PR) to bend target DNA. Studies were done with baculovirus-expressed full-length receptors purified to apparent homogeneity. By circular permutation analysis, both forms of PR induced substantial distortions in the structure of target DNA with calculated distortion angles (alpha D) of 57 degrees for PR-A and 84 degrees for PR-B. The apparent bend centers for both forms of PR were similarly located a few base pairs (-4 to -2 bp) from the middle of the progesterone response element. No differences were detected in the magnitude of distortion or apparent bend centers when PR was bound to hormone agonist (R5020) or the antagonist RU486. Phasing analysis, which can determine the orientation of a DNA bend, revealed that both forms of PR mediated directional bends toward the major groove of the DNA helix. Calculated directed bend angles (alpha B) were 40 degrees for PR-B and 31 degrees for PR-A. The chromatin high mobility group protein HMG-1, which acts as an accessory factor to enhance the binding affinity of purified PR for progesterone response elements, had minimal influence on PR-mediated DNA bending. This result, taken together with the fact that HMG-1 can form a ternary complex with PR and DNA, is consistent with the conclusion that HMG-1 facilitates PR binding by stabilizing a receptor-induced DNA conformation that is required for assembly of a high affinity PR-DNA complex. The results of this study also suggest that DNA bending may be coupled to transcriptional regulation since PR-B is generally a stronger transcriptional activator than PR-A and also mediates a larger bend in target DNA than PR-A.