The Evolutionary Implications of Adaptations to Stressful Environments on Time Scales Spanning Days to Millennia

Life is a perpetual onslaught of less-than-optimal conditions. In order for life to continue despite the ubiquity of otherwise-lethal stressors, organisms develop physiological and behavioral adaptations to escape, tolerate or avoid the stresses they face. Broadly, my dissertation covers the genetic signatures of evolutionary responses to stressful conditions. In the first portion of my dissertation, I investigate the evolutionary consequences of an extreme stress tolerance strategy in Firmicutes bacteria. For the remaining parts of my dissertation, I turn to the humble fruit fly, Drosophila melanogaster. Using this model organism, I investigate genetic mechanisms related to survival in starvation and desiccation conditions, such as those encountered at the onset of winter when food becomes scarce. Using both systems, I address how strategies of adapting to stress influences each organism’s genome. Using various inferential techniques, I make use of present-day observations to recapitulate the evolutionary trajectories of organisms—trajectories that vary in length from thousands of years to a single day.

In chapter one, I revisit the generation-time hypothesis using the Firmicutes bacteria system, increasing bacteria representation from 23 taxa to 197 taxa. With this improved taxonomic sampling, I reevaluated the relationship between generation time and evolutionary rate within the Firmicutes phylum. My comparative phylogenomic analysis of the Firmicutes phylum was capable of estimating putative endospore-forming ability from the history of gene content across time. I show strong evidence that the rate of molecular evolution is negatively correlated with generation time in bacteria, in line with observations throughout the rest of the tree of life. In order to understand the evolutionary history of endospore formation in ancestral nodes of the Firmicutes tree, I conducted ancestral state reconstruction of spore-forming ability. My results indicate that endospore formation arose once near the root of the Firmicutes tree, and the trait was subsequently lost in multiple lineages—never to be regained.

In chapter two, I evaluate methods of reconstructing forward-simulated Hybrid Swarm genomes from ultra-low coverage sequencing data. After optimizing parameters of the reconstruction pipeline, I show that it is feasible to generate high-quality genotype estimates while sampling as few as 1/20 to 1/200 variable sites. However, it is not sufficient to solely show that genome reconstructions are accurate—it is further necessary to show that Hybrid Swarm populations are capable of resolving genotype-phenotype relationships. In order to evaluate the effectiveness of the Hybrid Swarm for genetic association mapping, I developed a high-throughput pipeline for simulating GWAS, and then compared my newly developed mapping population to modern alternatives. I show that although inbred populations exhibit greater intrinsic power in mapping additive traits, a Hybrid Swarm population performs similarly to a highly outbred population, e.g. individuals from wild sampling. My results suggest that outbred mapping populations can be quickly generated using the Hybrid Swarm method, as a simpler and cheaper alternative to wild sampling.

In chapter three, I put my low-coverage genome reconstruction pipeline to use to study the mechanisms of rapid adaptation by D. melanogaster to desiccation and starvation conditions. After reconstructing nearly 700 individual fly genomes, I conduct a GWAS of survival, with significant associations found for multiple mitochondrial proteins—a class that has been previously implicated in lifespan expansion and tolerance to stress. Using expression data, I categorize genes that are differentially expressed and biological processes differentially represented in fed or starvation conditions. As expected, fed flies exhibit increased expression of reproductive genes. In the starvation condition, flies exhibit upregulation of multiple metabolic processes, particularly purine synthesis—which has been previously implicated in the tradeoff between fecundity and lifespan. Interestingly, we see that genes primarily expressed in different tissues exhibit variable rates of ASE, with increased representation of ASE in the carcass, salivary glands, heart, and fat body; the only tissue class to show reduced ASE is the ovary. Previous work has shown reduced levels of genetic variation in D. melanogaster reproductive genes, as a potential signature of purifying selection. My results suggest that genetic variation in and around reproductive genes themselves do not underly the tradeoff between survival and reproduction.