Interaction of Oct-1 with TFIIB. Implications for a novel response elicited through the proximal octamer site of the lipoprotein lipase promoter.

The ubiquitous human POU domain protein, Oct-1, and the related B-cell protein, Oct-2, regulate transcription from a variety of eukaryotic genes by binding to a common cis-acting octamer element, 5'-ATTTGCAT-3'. The binding of Oct-1 and Oct-2 to the functionally important lipoprotein lipase (LPL) promoter octamer site was stimulated by the general transcription factor, TFIIB. Comparative analysis of the LPL, histone H2B (H2B), and herpes simplex virus ICPO gene promoter octamer sites revealed that nucleotide sequences within and flanking the octamer sequence determined the degree of TFIIB-mediated stimulation of Oct-1 DNA binding. TFIIB was found to decrease the rate of dissociation of Oct-1 from the LPL octamer site, whereas it increased the rate of association, as well as decreased the rate of dissociation, of Oct-1 from the H2B octamer site. A monoclonal antibody against TFIIB immunoprecipitated a ternary complex containing TFIIB, Oct-1, and the LPL and H2B octamer binding sites. TFIIB did not alter the DNase I footprints generated by Oct-1 on the LPL and H2B promoters. However, Oct-1 on the TATA-binding protein and TFIIB from footprinting the perfect TATA box sequence located 5' of the LPL, NF-Y binding site. In transfection experiments, transcription from the reporters containing the LPL octamer, and either the SV40 or the yeast transcription factor GAL4-dependent enhancers, initiated at a precise position within the octamer sequence. Transcription from reporters containing the H2B octamer and the SV40 enhancer initiated at several positions within and flanking the octamer site, whereas transcription initiated at a precise position within the octamer from reporters with both the H2B octamer and the GAL4-dependent enhancer. These results suggest that octamers and their flanking sequences play an important role in positioning the site of transcription initiation, and that this could be a function of the interaction of Oct-1 with TFIIB.