ANTAGONIST-OCCUPIED HUMAN PROGESTERONE B-RECEPTORS ACTIVATE TRANSCRIPTION WITHOUT BINDING TO PROGESTERONE RESPONSE ELEMENTS AND ARE DOMINANTLY INHIBITED BY A-RECEPTORS

When antagonist-occupied steroid receptors have agonist-like effects, the clinical consequences are grave. We present evidence that human pr ogesterone B-receptors (hPR(B)) when occupied by progesterone antagoni sts, inappropriately activate transcription by an unusual mechanism th at does not require the canonical progesterone response element (PRE). In HeLa cells cotransfected with a PRE-tk-chloramphenicol acetyltrans ferase reporter and a hPR(B) expression vector, strong transcription i s seen not only when receptors are activated by the agonist R5020, but also in the presence of the three antiprogestins, RU486, ZK112993, an d ZK98299. Human PR(B) occupied by ZK98299 do not bind to a PRE, sugge sting that the transcriptional stimulation is independent of DNA bindi ng. Indeed, a tk-chloramphenicol acetyltransferase promoter-reporter l acking the PRE loses transcriptional activation by the agonist, but re tains transactivation by the three antagonists. The PRE-independent an tagonist-induced transcription requires that hPR(B) have an intact DNA -binding domain, but hPR target gene specificity is not required, beca use a hPR(B) mutant that binds an estrogen response element still acti vates transcription. It appears that antagonist-occupied hPR activate transcription without binding to a PRE, perhaps by interacting with te thering proteins instead. Even a gene that is not a normal progesteron e target could be aberrantly activated. Human cells contain equimolar amounts of hPR(B) and the N-terminally truncated natural isotype, hPR( A). Unlike hPR(B), hPR(A) are not transcriptionally activated by proge sterone antagonists. We, therefore, tested the effects of antagonists when the two receptor isotypes are coexpressed and found that A-recept ors can annul the inappropriate transcription by B-receptors. Thus, wh en both receptor forms are present, the hPR(A) phenotype is dominant. Moreover, pure hPR(B)/hPR(A) heterodimers, produced by fos/jun leucine zipper domain-hPR chimeras, also have the inactive transcriptional ph enotype of hPR(A). Our studies suggest not only that the two hPR isoty pes are functionally quite different, but also that some of the agonis t-like transcriptional effects of antagonist-occupied B-receptors proc eed through novel mechanisms.