This project relies on a selection experiment with Drosophila simulans populations that evolve in one of two different thermal environments (hot, fluctuating between 18 and 28°C and cold, fluctuating between 10 and 20°C). From this work, we derived multiple experiments in order to investigate the genetic and phenotypic changes occurring during evolution. First, we designed a “reverse selection” experiment: flies that initially were selected in our cold treatment were then shifted after four years to the other environment. Then, using common garden experiments we analyzed the gene expression profiles of evolved populations and contrast them with their ancestral value. In parallel, we investigated rapid genomic changes in the main experiment in order to link phenotypic changes to genetic target of selection using NGS technologies.
We showed that phenotypic plasticity of gene expression in a natural population of Drosophila simulans was evolving during adaptation in novel thermal environments and managed to explore some of the molecular basis of its rapid evolution. In particular, we demonstrated that:
- Selection in hot environments led to rapid changes in the metabolism regulation of the fruit flies, leading to profound changes of gene expression associated with the main biological pathways involved in energy production
- Phenotypic evolution relies on the evolution of phenotypic plasticity and not on constitutive changes (independent of the temperature).
- Phenotypic plasticity in the ancestral populations is adaptive: the direction of evolution in experiment is in the same direction than the ancestral plasticity. This suggests that selection for thermal plasticity occurs in natural populations, potentially due to seasonal and local variations
- These complex phenotypic changes rely on a simple genetic basis. We identified two evolving genes involved in the activity of a single enzyme, the AMP-Kinase, a key regulator of cellular homeostasis. Contrary to theoretical expectations, complex phenotypic traits can evolve rapidly due to changes in frequency of a small number of genomic regions, each having an important impact on the phenotype.
- The same genetic regions show pronounced latitudinal variations in North America, demonstrating that replicated laboratory experiments are well-suited to dissect and functionally characterize adaptive processes relevant to natural populations.
These results have been presented at two international conferences during the duration of the project and led to two research manuscripts that will rapidly be published.