Abstract
We have used the 100-kD HeLa cell octamer-binding protein OBP100 as a model to study flexible DNA sequence recognition by promoter-binding proteins. OBP100 binds to the conserved octamer motif ATGCAAAT found in numerous promoters and additionally to two degenerate octamer motifs (sites I and II) within the SV40 enhancer region. We show here that OBP100 binds the herpes simplex virus immediate early promoter TAATGARAT (R = purine) motif itself, extending the flexibility of OBP100 sequence recognition to sequences that bear very little resemblance (four matches over a 14-bp region). Nevertheless, a progression of OBP100-binding sites can be established that links the sequences of these two apparently unrelated binding sites by incremental steps. Mutational and chemical modification interference analyses of a degenerate octamer binding site (SV40 site II) show that specific sequences, which are not normally conserved but flank the degenerate octamer motif, can compensate for the degeneracy in the octamer core sequence. Thus, different regions of the binding site sequence (core or flanking) can diverge separately but not independently of one another. These results suggest that flexible DNA sequence recognition arises because there are few obligatory contact sites for OBP100 binding, but, rather, specific binding reflects the sum of many independent interactions.