Broad binding‐site specificity and affinity properties of octamer 1 and brain octamer‐binding proteins

Abstract

The ubiquitous Pit‐1‐Oct‐1‐Unc‐1 (POU)‐domain protein octamer 1 (Oct‐1) has been observed to bind specifically to a number of degenerate and dissimilar sequences. We have used antibodies directed against a C‐terminal Oct‐1 peptide to immunoselect binding sequences for HeLa cell Oct‐1 from random‐sequence oligonucleotides and we describe the isolation of binding sequences of considerable heterogeneity. Although our consensus alignment indicated a 9‐bp TATGCAAAT motif with AT‐rich flanking sequences, this binding motif is not immediately obvious in the population of sequences and no clone actually contained this sequence. Screening these Oct‐1‐binding sequences with a mouse whole‐brain extract demonstrated that the neuronal octamer‐binding proteins exhibit similar but distinct DNA sequence specificities. Unlike the reported selection of binding sequences for other transcription factors, the dependence of Oct‐1‐binding affinity upon sequence did not correspond tightly to the degree of conservation at particular positions of the consensus sequence. Our results suggest that either base‐specific hydrogen bonding is not the only major determinant of binding affinity and specificity, or that Oct‐1 binding to some sequences is mechanistically different from its binding to an octamer. These results exemplify the potential to overlook binding sites for some factors by searching gene sequences with a consensus nucleotide sequence. Copyright © 1993, Wiley Blackwell. All rights reserved