Quantitative genetics of protein traits

28/05/2013 - Room C209 at 14:00 - External Seminar
Leopold Parts
(The Donnelly Centre for Cellular and Biomolecular Research, University of Toronto)
Studies in humans as well as a number of model systems have emphasized the complexity of an individual’s genetic landscape, with many combinations of alleles that underlie phenotypic variability. Linking genotype to phenotype remains a major challenge in large part due to a dearth of systematic data about fundamental mechanisms underlying most complex inherited traits. The heritability of variation in mRNA levels has been thoroughly studied, but how protein localization and abundance varies with genetic variation has been less scrutinized. This largely reflects the lack of suitable technologies for accurate quantification of protein traits, while methods to measure nucleic acids are widely available. Uniquely in yeast, protein level and localization can be readily studied on the genome scale by using the Green Fluorescent Protein (GFP) collection, in which each strain produces a different full-length protein tagged with GFP at its C-terminus under the control of its endogenous promoter. We used the GFP collection to measure the level of nearly 4,000 proteins in hundreds of thousands of individual random progeny from a two-parent yeast cross. We compared the distribution of protein abundance in this segregating population to that of a parental clone to discover the extent of genetic influence on these traits. Only a minority of proteins (372) exhibited change of at least 25% between parents and the progeny, and over a half of them (206) were previously shown to be under genetic control on the transcript level. We then measured population allele frequencies in large sets of segregants sorted for high and low protein level to pinpoint genomic regions responsible for the heritable signal. We mapped 56 loci controlling protein abundance, and validated the causal effect of one of them. Protein traits tend to be complex, with three to fourteen loci contributing, and an average protein influenced by nine regions. We saw that many alleles previously found to influence mRNA levels also affected protein levels; however, the broad sense heritabilities for the mRNA and protein level traits appeared largely uncorrelated, partly due to our estimates being dominated by cell-to-cell variation. Further, we measured the localization pattern changes of yeast proteins by analyzing microscopy images of over a thousand cells, and found few changes due to the genotype of distal regulators. We are currently mapping and validating further genetic determinants of these traits, as well as quantifying the extent of cis versus trans control, and the dominance of the causal alleles.
Hosted by: Oliver Stegle