The long-term goal of this application is to understand how chromatin structure influences gene regulation. The simple idea that all genes are packaged, irrespective of their sequence, into uniform solenoid-like chromatin structures that must be disrupted in order for gene activation to occur may not be correct. Instead, DNA may be packaged in a sequence-specific fashion. Different nucleosome arrangements may lead to the formation of different chromatin structures that facilitate gene regulation. DNA sequence variations in certain regions of non-coding DNA among individuals might influence gene expression and disease susceptibility. Preliminary findings suggest that there are signals in vertebrate DNA that influence nucleosome array formation. Specifically, there is evidence that the nucleotide triplet consensus non-T, A/T, G (VWG) existing at ten base pair multiples is a strong nucleosome localization signal. The hypothesis that oscillations specifically in periodic VWGs influences nucleosome alignment into ordered arrays will be tested using an in vitro chromatin assembly system. The VWG signal in regions of genes that strongly align nucleosomes will be disrupted by insertions, deletions, and by site-directed mutagenesis. The hypothesis that DNA sequences that exhibit strong oscillations in the average period- 10 VWG count with a dinucleosome period strongly align nucleosomes into ordered arrays, while sequences that are aperiodic in VWG do not, will be tested in vitro and in nuclei. The hypothesis that VWGs influence nucleosome positioning through anisotropic bendability will be tested. The hypothesis that the degree of regularity of the nucleosome array influences the chromatin higher-order structure will be tested using analytical sedimentation and electron microscopy. It will be determined whether the presence or absence of strong nucleosome alignment signals in transgenes influences their transcription in the mouse. Information obtained here might ultimately enable one to predict functionally important aspects of chromatin structure computationally from DNA sequence data.
