Adaptation is one of the central processes in evolution. Not only does adaptation drive phenotypic change in organisms, it also underlies many human diseases, such as cancer and the emergence of drug resistance in their treatment. Yet our current understanding of adaptation suffers from a major limitation.
While many organisms of interest to biology and medicine are diploid, many studies characterize adaptive mutations—the agent of evolutionary change—only in haploids. Mutations driving adaptation likely have very different properties depending on whether they arose in diploids or haploids. For instance, diploid adaptive mutations cannot be fully recessive and may be gain-of-function more often than haploid ones. This can lead to profoundly different adaptation dynamics between diploids and haploids. We previously investigated this difference using a theoretical model [Sellis et al. 2011], though little empirical data on adaptation in diploids exist to date. Thus, our understanding of adaptation in diploid populations often requires extrapolating parameters from haploid experiments. This is particularly troubling in studies of cancer—an adaptive process in diploids.
To overcome this limitation, we are isolating and characterizing a statistically representative number of adaptive mutations in both diploids and haploids. To achieve this, we have used experimental evolution in yeast coupled with a high-resolution molecular barcoding system that enables us to measure and track the fitness of ~5x105 lineages over time [Levy et al. 2015]. We carried out experimental evolution in barcoded S. cerevisiae diploid and haploid populations in multiple evolutionary conditions. Results from haploids provide a basis of comparison, while multiple conditions take us towards generalizability of results.
As a first step towards understanding how adaptation differs between diploids and haploids, we will identify lineages that acquire adaptive mutation by increases in the frequencies of their barcodes. We will then estimate the distribution of fitness among thousands of adaptive clones for each evolution experiment. These adaptive clones will be taken from time point(s) in which most have one causal adaptive mutation. This data will be the largest and most precise to date on the dynamics of adaptation in diploid populations. The results will inform us on how diploids differ from haploids in terms of speed of adaption and effect size of adaptive change. In the future, we will identify the causal adaptive mutations to deeply characterize how diploids differ from haploids in terms of target and pathway of adaptation, dominance, and environmental pleiotropy.