Evolutionary theory suggests that fitness-associated alleles should be purged from the gene pool, yet, decades of GWAS and model organism studies have shown that they persist. And, although several evolutionary forces are thought to maintain such genetic variation segregating in wild populations, experimental support is limited. Focusing on lifespan, a trait tightly linked to fitness in fruit flies, we address one of such potential explanations: the idea that the alleles that regulate lifespan do so only in certain conditions. To do this, we exposed outbred Drosophila populations to two dietary conditions: control and high sugar diets. High sugar diets are associated with detrimental effects, including shortened lifespan, in flies, mice, humans, and other organisms. We then tracked genome-wide allele frequency changes over the lifetime of these populations, recording in real time the changes in the genomic composition of each population as flies aged. Using DNA sequences from over 10,000 individual flies, we identified thousands of lifespan-altering alleles, some associated with early vs late life tradeoffs, late-onset effects, and notably, a third of these loci showed genotype-by-environment interactions – all factors that are predicted to maintain genetic variation for lifespan. We find that lifespan-reducing alleles are most likely to be recently derived, have stronger effects on high sugar diets, and have a history of positive selection. These patterns are consistent with the mismatch evolutionary theory that predicts that historically beneficial alleles can become detrimental in novel conditions. Our results provide insight into the polygenic and context-dependent genetic architecture of lifespan, and the evolutionary processes that shape this key trait.
Of thousands of species of mosquitoes, just a handful present the greatest threats to public health: those species that have recently evolved a strong preference for human hosts and habitats. I will discuss the results of a set of continent-scale studies documenting massive diversity and rapid evolution in Aedes aegypti, the globally invasive primary vector of dengue, Zika, chikungunya, and yellow fever, across 27 sites from its native range in sub-Saharan Africa. Behavioral preference for human hosts was largely explained by two main ecological factors: the intensity of dry seasons, and human population density. By combining our ecological and behavioral analyses with analysis of 375 whole genomes spanning sub-Saharan Africa, we can identify three major phases in the history of this vector. Specialization on human hosts likely first evolved several thousand years ago in the highly seasonal habitats of the Sahel, where natural habitat for aquatic mosquito larvae is scarce across the long, hot dry season, but human water storage provides an ideal niche. Next, these human-specialist populations spread to the Americas, where they were responsible for devastating outbreaks of yellow fever and other arboviruses; these invasive populations subsequently spread across the global tropics. Now, in rapidly growing cities across sub-Saharan Africa, urbanization effects are beginning to dominate, and our models suggest that Aedes aegypti populations are becoming more specialized on human hosts. A large number of genes are involved in specialization, but they are concentrated in just a few key genomic loci. One of these loci is also associated with increased Zika competence, potentially helping explain why Zika has explosively emerged in the Americas and Pacific, while remaining less of a threat in sub-Saharan Africa. The expansion of cities into nearby forest areas, combined with changes in underlying vector genetics, present a major and growing risk for the zoonotic emergence of arboviruses.