Unifying Genetic Canalization, Genetic Constraint, and Genotype-by-Environment Interaction: QTL by Genomic Background by Environment Interaction of Flowering Time in Boechera stricta

Natural populations exhibit substantial variation in quantitative traits. A quantitative trait is typically defined by its mean and variance, and to date most genetic mapping studies focus on loci altering trait means but not (co)variances. For single traits, the control of trait variance across genetic backgrounds is referred to as genetic canalization. With multiple traits, the genetic covariance among different traits in the same environment indicates the magnitude of potential genetic constraint, while genotype-by-environment interaction (GxE) concerns the same trait across different environments. While some have suggested that these three attributes of quantitative traits are different views of similar concepts, it is not yet clear, however, whether they have the same underlying genetic mechanism. Here, we detect quantitative trait loci (QTL) influencing the (co)variance of phenological traits in six distinct environments in Boechera stricta, a close relative of Arabidopsis. We identified nFT as the QTL altering the magnitude of phenological trait canalization, genetic constraint, and GxE. Both the magnitude and direction of nFT's canalization effects depend on the environment, and to our knowledge, this reversibility of canalization across environments has not been reported previously. nFT's effects on trait covariance structure (genetic constraint and GxE) likely result from the variable and reversible canalization effects across different traits and environments, which can be explained by the interaction among nFT, genomic backgrounds, and environmental stimuli. This view is supported by experiments demonstrating significant nFT by genomic background epistatic interactions affecting phenological traits and expression of the candidate gene, FT. In contrast to the well-known canalization gene Hsp90, the case of nFT may exemplify an alternative mechanism: Our results suggest that (at least in traits with major signal integrators such as flowering time) genetic canalization, genetic constraint, and GxE may have related genetic mechanisms resulting from interactions among major QTL, genomic backgrounds, and environments.Author Summary: Biological traits often display large amounts of genetic variability as well as genetic correlations among traits. This variability provides the raw material for evolutionary change and may alter the direction of trait evolution under selection. Despite this importance, it is unclear whether the genetic controls of variability in single traits and relationships among multiple traits have related mechanisms. Using the flowering time of a plant species as model, here we performed genetic mapping and identified a locus altering single-trait variability and multi-trait relationships. The effect likely results from the distinct thresholds required by its different alleles to trigger flowering, which can be explained by the interaction among this major locus, the variable genomic backgrounds, and the distinct environments. This view is supported by experiments showing epistatic effects of this major locus on flowering time and expression pattern of the candidate gene. Together, our results show that, at least for traits with major signal integrator genes such as flowering time, the genetic control of single-trait variability and multi-trait relationships may have a common underlying mechanism that may be generalizable to other genes or pathways, mediated by interaction among major loci, genomic backgrounds, and surrounding environments.