This project uses a variety of approaches to address fundamental questions relating to chromatin structure, transcription and replication. We previously reported that the yeast alpha-2 protein organized chromatin structure; positioned nucleosomes flanked the binding site for this protein in a minichromosome. Using primer extension, we have now mapped these nucleosomes at base level resolution with the finding that the positioning is exquisitely accurate. Identical positioning is observed for the single genomic copy of a gene repressed by alpha-2; since the sequences flanking the operator differ completely for the two situations, this suggests an active organization of chromatin by the repressor. We have addressed the possible mechanism of positioning with the finding that deletions of amino acids 4-19 or 4-23 of histone H4 abolish positioning around the alpha-2 operator, suggesting a direct interaction of the repressor with nucleosomal histones. Alpha-2 also functions to repress the expression of haploid specific genes in yeast diploid cells; another homeobox containing protein, al, is required for this activity. We have conclusively demonstrated that a heterodimer, alpha-2/a1, is the structural entity which binds to haploid specific operators to effect this repression. Mapping of DNA contacts for the heterodimer shows a markedly different pattern from that for the ax homodimer + MCM1 dimer that represses a-specific genes. The data explain the dual repressive activities of the alpha-2 protein. This is the first demonstration of differential function of homo- and heterodimers of homeobox containing proteins and suggests the possible importance of such dimers in combinatorial regulation of genes in larger organisms. Studies of the chromatin structure and transcription of RNA polymerase III genes in yeast are continuing in attempts to correlate nucleosome positioning and interactions of regulatory factors with transcriptional initiation and elongation. A region replete with DNA sequences similar or identical to those bound by known transcriptional regulatory factors, I kb 5' to a heat shock gene of yeast, seems likely to function in regulation of transcription of the gene. Deletion of less than 150 bp in the putative regulatory region leads to a phenotype similar to deletion of the entire structural gene. Collaborative studies of high resolution core particle structure and higher order chromatin structure are continuing.
