The distribution of fitness effects during adaptive walks using a simple genetic network

Abstract

The tempo and mode of adaptation is dependent on the availability of beneficial alleles. Genetic interactions arising from gene networks can restrict this availability. However, the extent to which networks affect adaptation remains inconclusive. Current models of evolution consider additive genotype-phenotype relationships while often ignoring the contribution to phenotypic variance arising from gene interactions. In this study, we present a novel approach to modeling quantitative traits as products of genetic networks via systems of ordinary differential equations so we can mechanistically explore the effects of network structures on adaptation. We present the first step of this approach by studying a simple gene regulatory network, the negative autoregulation motif (NAR). Using forward-time genetic simulations, we measure adaptive walks towards a phenotypic optimum in both additive and NAR models. A key expectation from adaptive walk theory is that the distribution of fitness effects (DFE) of new beneficial mutations is exponential. We found that although both models strayed from expectation, harboring fewer beneficial mutations than expected, they took a similar number of adaptive steps to reach the optimum. NAR populations were 10% less likely to reach the optimum and took longer to adapt. This was partly due to a complex and largely deleterious DFE from new mutations, and by requiring coordinated changes among its molecular components. This behavior is reminiscent of the cost of complexity, where correlations among traits constrain adaptation. Our results suggest that the interactions emerging from genetic networks can hinder adaptation by generating complex and multi-modal distributions of fitness effects.