DIVERGENCE IN GENE EXPRESSION IS UNCOUPLED FROM DIVERGENCE IN CODING SEQUENCE IN A SECONDARILY WOODY SUNFLOWER

摘要

Premise of research. The transition from herbaceous annual to woody perennial has occurred numerous times during angiosperm evolution, but the suite of genetic changes that accompanies this life history shift remains poorly understood. Here we analyze genetic sequence and gene expression divergence between the annual Helianthus annuus and a recently diverged woody perennial ecotype from California. Methodology. We grew plants from populations of typical H. annuus and the woody perennial ecotype in a common garden and measured days to flower, mature height, and basal stem density. We looked for evidence of genome size evolution in the woody ecotype using flow cytometry and conducted anatomical observations of stem development in seedlings. We sequenced the aboveground seedling transcriptomes of six individuals of each ecotype and identified genes with high fixation index (F_(ST)) and significant divergence in gene expression. Finally, we examined the gene ontology annotation of differentiated genes and assessed the extent of overlap between genes with divergent sequence and those with divergent expression. Pivotal results. Plants from the novel woody ecotype flowered later and were taller at maturity than typical H. annuus plants. Despite differences in the initiation and extent of secondary growth, the two sunflowers were similar in basal stem density at 6 mo. The ecotypes do not differ in genome size. We discovered 575 genes (3.5%) with significantly different expression between the ecotypes and identified seven biological processes associated with divergent gene expression. Surprisingly, only five of the differentially expressed genes were also present among genes with the most coding sequence divergence (top 3.5% of F_(ST) distribution). Conclusions. The recent transition in H. annuus from annual to woody perennial has been accompanied by numerous genomic changes, including divergence in gene expression and in coding sequence. These two mechanisms may represent alternative pathways toward the same adaptive optimum, or genes may be constrained to evolve via one mechanism or the other over short timescales.