Multiple collagen I gene regulatory elements have sites of stress‐induced DNA duplex destabilization and nuclear scaffold/matrix association potential

Abstract

The availability of the complete nucleotide sequences of numerous prokaryotic and eukaryotic organisms should stimulate the development and application of computer-based approaches for studying genome organization and function. Earlier work has shown that distinct regulatory DNA elements can be identified by computational analysis as sites of stress-induced DNA duplex destabilization (SIDD). Here we report the results of computational and experimental analyses of previously identified regulatory elements in the murine alpha1(I) collagen (Col1a1) gene domain. We found that several distal 5' DNase I-hypersensitive sites (HSs) which function in the chromatin loop organization of the Col1a1 gene are characterized by strongly destabilized SIDD profiles. Elements in the proximal 5' promoter and first intron which differentially regulate Col1a1 promoter activity in different collagen-producing cell types also contain SIDD sites. All 5' elements associated with destabilized sites are shown to have nuclear matrix binding activity in an in vitro binding assay. Other putative regulatory elements in the transcribed and 3'-flanking regions of the Col1a1 gene, including both of its polyadenylation sites, are also associated with SIDD peaks. The human COL1A1 gene has periodic SIDD peaks within the transcribed region, suggesting that abundantly expressed genes may require SIDDs acting as topological sinks during transcription. The 5' ends of the murine Col1a1 and the homologous human gene revealed similar SIDD profiles, but limited DNA sequence similarity, indicating that some DNA functions are evolutionarily conserved by preserving higher order DNA structural properties rather than nucleotide sequence. Our results show that destabilized SIDD profiles are a common feature of eukaryotic regulatory DNA elements with such diverse functions as chromatin organization, cell-specific transcriptional enhancement, and initiation and termination of transcription. They demonstrate the usefulness of computational analyses that predict SIDD properties in reliably identifying DNA elements involved in the structural organization of the eukaryotic genome and the regulation of its expression.