Patterns of adaptive and purifying selection in the genomes of phocid seals

• Main Author: Gaughran, Stephen John
• Format: Text
• Language: unknown
• Published: EliScholar – A Digital Platform for Scholarly Publishing at Yale 2021
• Subjects: Genetics; Weddell; Elephant Seal; Elephant Seals; ringed seal; Southern Elephant Seal; Weddell Seal; Weddell Seals
• Online Access: https://elischolar.library.yale.edu/gsas_dissertations/51; https://elischolar.library.yale.edu/cgi/viewcontent.cgi?article=1050&context=gsas_dissertations

Modern genomic sequencing technologies provide the opportunity to address long-standing questions in molecular evolution with empirical data. In this dissertation, I combine this new technology with advances in statistical population genetics to describe how deleterious mutations and adaptive evolution have shaped the genomic evolution of phocid seals. In Chapter 1, I model historical demographic processes using whole genome sequences of eight seal taxa: the Hawaiian monk seal, the Mediterranean monk seal, the northern elephant seal, the southern elephant seal, the Weddell seal, the grey seal, the Baltic ringed seal, and the Saimaa ringed seal. Through this, I establish that the endangered monk seal species have long-term small population sizes, as do grey seals. On the other hand, the elephant seals, Weddell seal, and ringed seals had much larger populations in the distant past. Notably, the most recent glaciation (c. 12,000-120,000 years ago) appeared to have a dramatic effect on phocid populations throughout the world. With this knowledge of historical population sizes, I test a fundamental premise of molecular evolution: that the rate of mutation accumulation will be higher in smaller populations due to less efficient purifying selection. I show that there is not a higher substitution rate or overall rate of mutation accumulation in the long-term small populations of monk seals compared to other seal species. On the contrary, overall rates of mutation accumulation appear to be lower in monk seals and grey seals, both of which show smaller long-term population sizes compared to the other species. This suggests either that the distribution of fitness effects may differ across seal species in a way that depends on population size and history. In Chapter 2, I use population genomic data and a newly developed statistical model to detect positive selection in the protein coding genes of phocid seals (monk seals, elephant seals, Weddell seals, grey seals, and ringed seals). In addition, I use a phylogenetic framework to detect parallel evolution across multiple lineages of seals, relating to traits such as polar adaptations, hypoxia tolerance during long dives, and mating behavior. I develop a new bioinformatic tool to process raw BAM files and transform them into useable input for MASS-PRF, a tool to detect selection from polymorphism and divergence data. Through these analyses, I identify thousands of genes that show positive selection across multiple seal lineages. Genes associated with immune function, sperm competition, and blubber composition show positive selection in all lineages, highlighting how complex and important these traits are in seals. In the deep-diving elephant seals, the list of positively selected genes was enriched for genes relating to cardiac muscle development and function, providing important insight into how adaptive protein evolution has helped allow these seals to survive sustained bradycardia during dives that last over an hour. Weddell seals, on the other hand, showed enrichment for genes relating to neuronal development, which may relate to molecular adaptations that allow their neurons to survive hypoxic conditions during long dives. Because MASS-PRF allows for site-specific tests of selection, I am able to show how parallel evolution in the same genes across lineages sometimes may or may not involve positive selection at the same genic site. In Chapter 3, I use the population genomic data from Chapter 2 to model the distribution of fitness effects (DFE) of segregating alleles in each population. Due to sample size issues, only parameters for the Hawaiian monk seal were confidently estimated. Using the site frequency spectrum of synonymous sites, I show that the Hawaiian monk seal has had a long-term effective population size below 5000, in agreement with the results from Chapter 1. In addition, I should that after the arrival of humans in Hawaii, the monk seal experienced a 95% decline in effective population size, in line with the current census size of fewer than 1500 individuals. Conditioning the model on the Hawaiian monk seal demographic parameters, I am able to estimate the shape of DFE in Hawaiian monk seals using the site frequency spectrum of nonsynonymous sites. I estimate a DFE for the Hawaiian monk seal that is nearly identical to the one estimated in humans. This DFE, however, is different than the one estimated for mouse, with the seal and human DFEs having a higher proportion of more strongly deleterious alleles. This pattern cannot be explained by phylogenetic relatedness or differences in phenotypic complexity, but instead is likely related to differences in effective population size. I discuss how the geometric model of evolution predicts such a shift in DFE in response to the epistatic effect of fixed deleterious mutations in smaller populations.