Genomewide identification of transcription factor binding sites by DNAseI footprinting
The complement of DNA-binding proteins and their occupancy of sites throughout the genome determine an organism’s programs of gene expression, DNA replication and other chromosome-based processes. A detailed picture of factor binding on a genome-wide basis exists for Saccharomyces cerevisiae, obtained by a combination of transcriptional profiles, chromatin immunoprecipitation of more than 200 transcription factors, computational analyses and other assays. In an alternative approach, we have used digestion of chromatin by DNase I followed by high throughput DNA sequencing to identify sites of increased nuclease accessibility throughout the yeast genome. The resulting set of more than 10 million sequence reads provides both a global view of chromatin architecture as well as a gene-by-gene view of regulatory sequences protected from digestion by the presence of bound proteins. Unlike the case with results from chromatin immunoprecipitation, these gene-by-gene DNase I footprints can be used to directly identify transcription factor binding sites, and thereby infer their motifs. We found previously unknown binding sites in the genome for well-characterized factors, and observed other annotated binding sites that appear not to be protected from nuclease digestion under our conditions. This approach has the potential to characterize the transcriptional regulatory network of a poorly characterized organism given only its genome sequence.
•Jay Hesselberth, Zhihong Zhang
DrnI is a novel debranching enzyme-associated nuclease with a role in intron turnover
The turnover of introns spliced from pre-mRNA occurs first by debranching the lariat intron followed by destruction of the linear intron by other nucleases. We have identified a novel component of intron turnover, DRN1 (YGR093W). Using yeast two-hybrid screens, we found that Drn1 interacts with the debranching enzyme, Dbr1, and another spliceosomal component, Syf1. Sequence alignments revealed that Drn1 is homologous to the metallophosterase domain of Dbr1, and Drn1 has RNA endonuclease activity in vitro. Deletion of DRN1 results in the accumulation of lariat introns spliced from some ribosomal protein genes. We have identified genetic interactions between DRN1 and mutant alleles of PRP43, suggesting that Drn1 plays a role in the turnover of the spliceosomal complexes containing lariat introns. Intriguingly, the subset of Drn1-effected introns use RNA structural elements to stabilize conformations productive for splicing. We propose a model in which the nuclease activity of Drn1 is required for the efficient turnover of these large, structured introns, whose hyperstability may hinder the dissociation of lariat intron complexes by the spliceosomal DEAD/H-box ATPases.
•Jay Hesselberth
Isolating RNAs containing 3' cyclic phosphate moeities
Standard techniques used to isolate and identify RNA from cellular extracts have traditionally relied upon poly-thymidine oligonucleotide hybridization or T4 RNA ligase based methodologies. These methods have been successful in isolating populations of RNAs that are modified with poly-adenosine tracts or have hydroxyl moieties (-OH) at their 3’ terminus. It is possible that these two classes represent the majority of the cellular ‘RNA universe’. However it is clear that the RNA universe is more complex than previously appreciated. Specifically RNAs that are the product of ribonucleolytic cleavage possess cyclic phosphate moieties at their 3’ terminus. These molecules are not substrates for isolation via either of the two previously mentioned methods. As such we are currently working on optimizing a strategy for the isolation and identification of RNAs with 3’ cyclic phosphate moieties.
•Kevin Schutz
