Protein binding to simple repetitive sequences depends on DNA secondary structure(s)

Simple repetitive DNA sequences are ubiquitous constituents of eukaryotic c hromosomes. The properties of simple repeats generate increased interest as expansions of certain trinucleotide blocks cause human diseases. We studie d protein binding and structural features of(gaa.ttc)(n) tracts e.g. in the polymorphic frataxin intron 1 and (gt)(n)(ga)(m) stretches from different HLA-DRB1 alleles in their original genomic environments. Electrophoretic mo bility shift assays revealed that HeLa nuclear proteins bind to DNA fragmen ts containing these simple repeat blocks. The major retarded protein/DNA co mplexes comprise, in both cases, zinc-dependent proteins present in nuclear extracts from different cell types. Competition experiments using various simple repeats differing in length and flanking regions demonstrate specifi c interactions. DNase I footprinting shows protein-binding sites located ei ther within the repeats alone or within the repeats as well as their flanki ng regions, often with preference for one strand. Comparing different (gt)( n)(ga)(m) alleles, a regular pattern of footprints was not detectable in th e (gt)(n) part indicating that the zinc-dependent protein recognizes struct ural rather than sequence-specific features. OsO4 and DEPC modifications fo llowed by electrophoretic and electron microscopical analyses demonstrate t hat the homopurine blocks often form different types of intramolecular trip le helices. A similar situation was evident using (gaa.ttc)(n) blocks of di fferent lengths within frataxin intron 1 as targets. These data have functi onal implications for non-coding (gaa.ttc)(n) and (gt)(n)(ga)(m) tracts wit h regard to gene expression in vivo.