Quantitative trait loci and global expression patterns offer insight into phenotypic plasticity

Abstract

Phenotypic plasticity, i.e. changes in phenotype with environment, may allow organisms to produce optimal phenotypes in all environments (adaptive plasticity). There is considerable discussion about the genetic mechanisms for phenotypic plasticity. However, there is some agreement that changes in gene expression must be involved. In order to examine how natural selection has acted on phenotypic plasticity on the trait bolting time trait (transition to reproduction) and gene expression patterns in past populations, a set of Recombinant Inbred Lines (RILs) of Arabidopsis thaliana were used. We tested the effect of cold-treating seeds (stratification) on bolting time in Arabidopsis thaliana. Cold stratification of seeds may be one environmental factor contributing to variation in spring versus fall germination in A. thaliana populations. Variation in both the direction and degree of plasticity was observed in a set of 120 RILs screened; genotypic selection analysis showed that past selection had favored bolting earlier in both environments. Three RILs that displayed extreme plasticities for bolting time in opposing directions across cold treatments were identified and global gene expression patterns were measured in a microarray experiment. A total of 294 genes were identified as being differentially expressed across cold treatments for the three extreme RILs (Fold change of >2; p-value<0.05). In addition, Quantitative Trait Loci (QTL) for bolting time were mapped in the complete set of RILs. Five QTL were mapped in the cold environment and three QTL were mapped in the no-cold environment explaining 40% and 30% of the observed phenotypic variation, respectively. QTL underlying variation in bolting time were screened for differentially expressed genes from the microarray study to identify candidate genes. Ninety-three of the identified genes colocalized to bolting time QTL, with kinesis and transferases being overrepresented in the expressed QTLs. Few studies have combined microarray and QTL data, and this study will offer insight into the genetic mechanism of phenotypic plasticity.