Margie Kinnersley1, Katja Schwartz2, Dong-Dong Yang3, Gavin Sherlock4, Frank Rosenzweig5,6. 1. Division of Biological Sciences, The University of Montana, Missoula, MT, 59812, USA. 2. Department of Genetics, Stanford University School of Medicine, Stanford, CA, 94305-5120, USA. 3. School of Biological Sciences, Georgia Institute of Technology, Atlanta, GA, 30332, USA. 4. Department of Genetics, Stanford University School of Medicine, Stanford, CA, 94305-5120, USA. gsherloc@stanford.edu. 5. Division of Biological Sciences, The University of Montana, Missoula, MT, 59812, USA. frank.rosenzweig@biology.gatech.edu. 6. School of Biological Sciences, Georgia Institute of Technology, Atlanta, GA, 30332, USA. frank.rosenzweig@biology.gatech.edu.
Abstract
BACKGROUND: Microbial evolution experiments can be used to study the tempo and dynamics of evolutionary change in asexual populations, founded from single clones and growing into large populations with multiple clonal lineages. High-throughput sequencing can be used to catalog de novo mutations as potential targets of selection, determine in which lineages they arise, and track the fates of those lineages. Here, we describe a long-term experimental evolution study to identify targets of selection and to determine when, where, and how often those targets are hit. RESULTS: We experimentally evolved replicate Escherichia coli populations that originated from a mutator/nonsense suppressor ancestor under glucose limitation for between 300 and 500 generations. Whole-genome, whole-population sequencing enabled us to catalog 3346 de novo mutations that reached > 1% frequency. We sequenced the genomes of 96 clones from each population when allelic diversity was greatest in order to establish whether mutations were in the same or different lineages and to depict lineage dynamics. Operon-specific mutations that enhance glucose uptake were the first to rise to high frequency, followed by global regulatory mutations. Mutations related to energy conservation, membrane biogenesis, and mitigating the impact of nonsense mutations, both ancestral and derived, arose later. New alleles were confined to relatively few loci, with many instances of identical mutations arising independently in multiple lineages, among and within replicate populations. However, most never exceeded 10% in frequency and were at a lower frequency at the end of the experiment than at their maxima, indicating clonal interference. Many alleles mapped to key structures within the proteins that they mutated, providing insight into their functional consequences. CONCLUSIONS: Overall, we find that when mutational input is increased by an ancestral defect in DNA repair, the spectrum of high-frequency beneficial mutations in a simple, constant resource-limited environment is narrow, resulting in extreme parallelism where many adaptive mutations arise but few ever go to fixation.
BACKGROUND: Microbial evolution experiments can be used to study the tempo and dynamics of evolutionary change in asexual populations, founded from single clones and growing into large populations with multiple clonal lineages. High-throughput sequencing can be used to catalog de novo mutations as potential targets of selection, determine in which lineages they arise, and track the fates of those lineages. Here, we describe a long-term experimental evolution study to identify targets of selection and to determine when, where, and how often those targets are hit. RESULTS: We experimentally evolved replicate Escherichia coli populations that originated from a mutator/nonsense suppressor ancestor under glucose limitation for between 300 and 500 generations. Whole-genome, whole-population sequencing enabled us to catalog 3346 de novo mutations that reached > 1% frequency. We sequenced the genomes of 96 clones from each population when allelic diversity was greatest in order to establish whether mutations were in the same or different lineages and to depict lineage dynamics. Operon-specific mutations that enhance glucose uptake were the first to rise to high frequency, followed by global regulatory mutations. Mutations related to energy conservation, membrane biogenesis, and mitigating the impact of nonsense mutations, both ancestral and derived, arose later. New alleles were confined to relatively few loci, with many instances of identical mutations arising independently in multiple lineages, among and within replicate populations. However, most never exceeded 10% in frequency and were at a lower frequency at the end of the experiment than at their maxima, indicating clonal interference. Many alleles mapped to key structures within the proteins that they mutated, providing insight into their functional consequences. CONCLUSIONS: Overall, we find that when mutational input is increased by an ancestral defect in DNA repair, the spectrum of high-frequency beneficial mutations in a simple, constant resource-limited environment is narrow, resulting in extreme parallelism where many adaptive mutations arise but few ever go to fixation.
Authors: Rembrandt J F Haft; David H Keating; Tyler Schwaegler; Michael S Schwalbach; Jeffrey Vinokur; Mary Tremaine; Jason M Peters; Matthew V Kotlajich; Edward L Pohlmann; Irene M Ong; Jeffrey A Grass; Patricia J Kiley; Robert Landick Journal: Proc Natl Acad Sci U S A Date: 2014-06-09 Impact factor: 11.205
Authors: Lei Wang; Beny Spira; Zhemin Zhou; Lu Feng; Ram P Maharjan; Xiaomin Li; Fangfang Li; Christopher McKenzie; Peter R Reeves; Thomas Ferenci Journal: Genome Biol Evol Date: 2010-07-16 Impact factor: 3.416