This article refers to:
Volume 17, no. 10, p. 6131–6138. In attempts to extend the experiments reported in this paper, we discovered unanticipated additional mutations in a number of the clones analyzed. Most relevant to our conclusions, the actions of the mutations that we initially attributed to a single lesion at proline 453 were the combined effects of two distinct mutations: the P453H lesion and an artifactual L451F substitution. This observation does not alter our overall conclusion that conformational changes to helix 12 appear to play an important role in corepressor release; it does, however, alter our interpretation of the precise locations of the genetic lesions that disrupt this phenomenon. We apologize for any confusion this error may have caused.