Comparative Study Of Morphological And Physiological Evolution In Arabidopsis

One of the fundamental questions in biology is how phenotypic trait diversity evolves. The evolution of form has been proposed to be mechanistically different from the evolution of physiology: morphological evolution occurs primarily through changes in gene expression, while physiological evolution occurs primarily through changes in coding sequences. However, these claims are thought to be unjustified and remain controversial. Here we systemically contrast the evolutionary pattern of the genes controlling morphological traits(morphogenes) with that of the genes controlling physiological traits(physiogenes) in Arabidopsis thaliana. We compared the evolutionary rate of morphological and physiological genes in protein-coding regions, cis-regulatory sequences, gene expression, gene duplications/deletion and alternative splicing. Our findings provide important insights into the genetic mechanisms underlying morphological and physiological trait evolution in Arabidopsis thaliana. The main results are as follows:(1) Morphogenes and physiogenes have different functional properties. Morphogenes and physiogenes involved in different GO Biological Process terms. But no GO Molecular Function term is found to over- or under-represented when comparing morphogenes with physiogenes. It shows that morphogenes and physiogenes have different functional properties.(2) Morphogenes generally evolve more slowly than physiogenes in Arabidopsis thaliana. Our comparative genomic analyses yield a rather consistent picture: morphogenes evolve more slowly than physiogenes in terms of coding and cis-regulatory sequences, gene expression, and gene family size. After we controlled gene essentiality, the pattern that morphogenes evolve more slowly than physiogenes still holds. The d N/d S ratios of morphogenes(mean=0.16) and physiogenes(mean=0.19) are generally low. Only AGG1 gene(d N/d S=1.06) and DEP gene(d N/d S=1.05) exhibit positive selection signal. These results suggest that relaxed functional constraints, rather than positive selection, are the dominant force resulting in the rapid evolution of physiogenes than morphogenes. Consistently, we found that morphogenes are more pleiotropic than physiogenes; the higher pleiotropic level of morphogenes might partially explain their lower rate of evolution. Physiogenes evolve faster in gene expression than morphogenes. Consistently, cis-regulatory sequences of morphogenes evolve slower than these of physiogenes, which might partially explain the slower gene expression evolution in morphogenes. The difference of cis-regulatory sequence divergence between morphogenes and physiogenes might be explained by mutation rates(d S), because the difference in gene expression divergence disappeared after we controlled mutation rates indicated by d S.(3) There is no difference on evolutionary genetic mechanism between morphological and physiological trait of Arabidopsis thaliana. Our analysis robustly shows that both coding sequence and gene expression, physiogenes evolved faster than morphogenes. Though the results are not identical to previous studies, at least we can say that, in Arabidopsis thaliana, the evolution of morphological and physiological traits may not significantly different on the genetic mechanisms. The independent origin of multicellularity which morphological traits are based on in land plants and animals might contribute to the genetic basis difference.(4) Alternative splicing events may play an important role in the evolution of the phenotypes in Arabidopsis thaliana. While physiogenes generally evolve more rapidly than morphogenes in many genetic mechanisms, it is observe that physiogenes and morphogenes have a similar rate of alternative splicing evolution. It is possible that alternative splicing substantially contribute to both morphological and physiological trait evolution. We found that alternative splicing of both physiogenes and morphogenes evolves more rapidly than random genes. These results indicate that alternative splicing might play an important role in phenotypic trait evolution.