PANTOTHENATE KINASE4, LOSS OF GDU2, and TRANSPOSON PROTEIN1 affect the canalization of tomato fruit metabolism.

Most studies investigating quantitative traits focus on mean levels per genotype rather than the variation between different individuals of one genotype or the variation elicited by different environments. Consequently, the genes that govern this effect are not well understood. The concept, named canalization, which describes a lack of variation, is well known in the context of developmental processes but is poorly studied for quantitative traits such as metabolism. In this study, we selected 8 putative candidate genes from previously identified canalized metabolic quantitative trait loci and created genome-edited tomato (Solanum lycopersicum) mutants of these genes for experimental validation. Most lines showed wild-type morphology, except for an ARF-like protein mutant showing aberrant phenotypes in the form of scarred fruit cuticles. In greenhouse trials with different irrigation conditions, whole-plant traits showed a general increase of their level toward the more optimal irrigation conditions, whereas most metabolic traits showed an increase toward the other end of the gradient. Mutants of a PANTOTHENATE KINASE 4, the AIRP ubiquitin gene LOSS OF GDU2, and TRANSPOSON PROTEIN 1 grown under these conditions showed an overall improved plant performance. Additional effects, on both target and other metabolites in tomato fruits, regarding the mean level at specific conditions and, ergo, the cross-environment coefficient of variation, were observed. However, variation between individuals remained unaffected. In conclusion, this study supports the idea of distinct sets of genes regulating different types of variation.