We propose to use yeast as a model system to evaluate methodologies for characterizing quantitative trait genes and to understand the complexity of the genetics that underlies this class of traits. Because quantitative traits are poorly understood, we have developed S. cerevisiae as a model for the study of quantitative traits and have chosen drug resistance because it is easily phenotyped by quantitatively measuring growth rate and our preliminary results suggest that it is genetically complex. Our combination of technologies designed, developed in our laboratory and already applied to dissect two quantitative traits, growth at high-temperature and efficiency of sporulation, will be extended here to a larger scale to identify most of the protein-coding and ncRNAs that condition variation in a complex trait. A new application named reciprocal hemizygosity scanning has been developed in our laboratory to identify alleles that contribute to a trait. A collection of two sets of reciprocal hemizygous yeast deletion mutants has been generated in a diploid SK1/S288c hybrid, representing 76% of all ORFs. Here, quantitative trait loci (QTL) will be identified by combining three innovative strategies: reciprocal hemizygosity scanning (RHS), bulk segregant analysis (BSA) and classical linkage mapping (CLM) using whole genome sequencing. We expect that this combination of approaches will yield the most comprehensive list to date of genetic factors underlying a quantitative trait. In parallel we propose to extend our RHS technology to the entire genome, which will allow analyzing the impact of genes, non-coding RNAs and intergenic regions. Since a major question is how QTLs mediate their effects on phenotype, allele replacement strains will be constructed for all combinations of five major effect QTLs, and the consequences will be monitored at the molecular level by transcriptome profiling using tagseq, a sequencing protocol being developed in our lab, based on the capture of polyadenylated transcripts and sequencing of their 3'ends. The establishment of tools for identifying quantitative trait genes and results on their functional reconstitution and genetic engineering will provide the best data for approaching complex traits for medical studies.
Most phenotypes in natural populations are quantitative, including disease susceptibility and drug response variability in humans. Unfortunately, this class of phenotypes, termed quantitative traits, is genetically complex and difficult to study using traditional techniques. We will use yeast as a model system to discover general principles regarding the genetic architecture of quantitative traits and uncover limitations in current methodologies to dissect quantitative traits, including those used in humans.
Showing the most recent 10 out of 27 publications
