Dissecting anxiety-related QTLs in mice by univariate and multivariate mapping

To detect genes underlying anxiety-related traits in mice, we performed univariate and multivariate QTL mapping analyses of phenotypes obtained from 71 mice of the BXD recombinant inbred (RI) strains (n=528 mice) and their parental strains (C57BL/6J and DBA/2J). Separate and joint mapping analyses were carried out using a linkage map composed of 506 simple sequence repeats (SSRs). The main QTL effects, interactions between pairs of QTLs (epistasis), and their environmental interactions were estimated. The results showed that anxiety-related traits were influenced by multiple QTLs (five main effect QTLs and three epistatic QTLs). Ten potential anxiety-related candidate genes within the QTL intervals on chromosomes 5, 13 and 15 were identified. Some of these genes have been reported previously to be associated with the anxiety response. Based on our results, it is suggested that the multivariate QTL mapping approach improves the statistical power for detecting QTL and the precision of parameter estimation. Moreover, multivariate mapping can also detect pleiotropic QTL effects.