Examining The Predictability of Genetic Background Effects in The Drosophila Wing

Abstract

Background dependence is a ubiquitous attribute of eukaryotic gene systems that modulates the phenotypic effects of a mutant allele due to segregating genetic variation among different wildtype strains. Despite the wealth of literature demonstrating the presence of genetic background effects, very little is known about how they functionally or mechanistically contribute to the relationship between genetic variation and phenotypic expression. It has been postulated that background dependent effects may be highly specific to the activity of individual alleles or genes. A recent examination of mutant alleles in two interacting genes in the Drosophila wing network demonstrated the magnitude of phenotypic effect of a mutant allele may predict it’s sensitivity to the genetic background. To further understand this, I examined the background dependence of many alleles for genes across the regulatory network of Drosophila wing development in many inbred strains. Our goal was to understand whether effects of the genetic background are an attribute of individual alleles, alleles of the same gene, or genes with similar phenotypes or developmental roles. Our analysis suggests that background dependence is highly positively correlated among alleles of the same gene, especially between alleles with similar magnitudes of phenotypic effect. Similarly, the background dependence of genes within the same regulatory network were also positively correlated. Alleles from different genes, but of the same magnitude of phenotypic effect, generally demonstrated the highest degree of intergenic correlation. However, the background dependence of mutant alleles were generally not well correlated with the wildtype allele. Interestingly, we also found no recovery of any lethal alleles, despite thousands of individuals screened and evident suppression of mutant effects in some strains. We also analyzed the magnitude of intra-line variance in among a subset of our genes. This demonstrated a strong positive relationship between the magnitude of intra-line variation and the severity of phenotypic effects, regardless of the identity of the mutant allele. However, we show no correlation between intra-line variability in the wildtype and the magnitude of perturbation for a given mutant allele. To confirm the quantitative estimates of mean wing size accurately reflected subtle perturbations to wing tissue, we conducted a semi-quantitative analysis and compared it to our quantitative estimates. We demonstrate a high degree of correlation between the quantitative and semi-quantitative approaches, indicating semi-quantitative analysis is a useful way to capture subtle phenotypic effects. In addition, we repeated the quantitative analysis with a subset of the genes and inbred strains from the original data. Importantly, results of the repeated study largely recapitulate our original results.