Combining mapping and arraying: An approach to candidate gene identification

Abstract

A combination of quantitative trait locus (QTL) mapping and microarray analysis was developed and used to identify 34 candidate genes for ovariole number, a quantitative trait, in Drosophila melanogaster. Ovariole number is related to evolutionary fitness, which has been extensively studied, but for which few a priori candidate genes exist. A set of recombinant inbred lines were assayed for ovariole number, and QTL analyses for this trait identified 5,286 positional candidate loci. Forty deletions spanning the QTL were employed to further refine the map position of genes contributing to variation in this trait between parental lines, with six deficiencies showing significant effects and reducing the number of positional candidates to 548. Parental lines were then assayed for expression differences by using Affymetrix microarray technology, and ANOVA was used to identify differentially expressed genes in these deletions. Thirty-four genes were identified that showed evidence for differential expression between the parental lines, one of which was significant even after a conservative Bonferroni correction. The list of potential candidates includes 5 genes for which previous annotations did not exist, and therefore would have been unlikely choices for follow-up from mapping studies alone. The use of microarray technology in this context allows an efficient, objective, quantitative evaluation of genes in the QTL and has the potential to reduce the overall effort needed in identifying genes causally associated with quantitative traits of interest.

Fig 1.

For each of the 294 genes under consideration, the mean expression of the Oregon-R replicates is plotted on the x axis and the mean of the 2b replicates is plotted on the y axis. Means were calculated after normalization.

Fig 2.

(Upper) The black arrow highlights the recombinational map position of the candidate genes CG17327, yellow-f, and Su(fu). Red curves indicate the value of the test statistic for the presence of QTL (7). Blue triangles indicate cytological markers used in the QTL experiment; the purple star represents the centromere. Horizontal bars are the deficiencies that were tested; gold bars showed a significant interaction across parents and genotypes, whereas green bars did not (M.L.W., L. Jacobs, A. Kuntz, and L.-Y. Shen, unpublished work). Deficiencies are staggered in height for presentation purposes. (Lower) Venn diagram showing systematic narrowing of list of candidate genes from all genes in the genome (open circle), to genes mapped in the QTL (red circle), to genes identified by deficiency mapping present on the array (gold circle), to genes differentially expressed between the parental lines by Bonferroni criteria (CG17327, black dot).